diff --git "a/checkpoint-18830/trainer_state.json" "b/checkpoint-18830/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-18830/trainer_state.json"
@@ -0,0 +1,131843 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.9988052568697725,
+  "eval_steps": 500,
+  "global_step": 18830,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00026549847338377806,
+      "grad_norm": 1.4133170389245202,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.9082,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005309969467675561,
+      "grad_norm": 1.4044375677052006,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.9046,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007964954201513341,
+      "grad_norm": 1.3102037815788463,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.871,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010619938935351122,
+      "grad_norm": 1.3366471282054486,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.8782,
+      "step": 4
+    },
+    {
+      "epoch": 0.0013274923669188902,
+      "grad_norm": 1.3635185920921395,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.9146,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015929908403026682,
+      "grad_norm": 1.4014578584835566,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.911,
+      "step": 6
+    },
+    {
+      "epoch": 0.0018584893136864462,
+      "grad_norm": 1.3549748295314084,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 0.8819,
+      "step": 7
+    },
+    {
+      "epoch": 0.0021239877870702245,
+      "grad_norm": 1.3135635642099046,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.893,
+      "step": 8
+    },
+    {
+      "epoch": 0.0023894862604540022,
+      "grad_norm": 1.2926227219867545,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 0.8847,
+      "step": 9
+    },
+    {
+      "epoch": 0.0026549847338377805,
+      "grad_norm": 1.251755492618311,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.8509,
+      "step": 10
+    },
+    {
+      "epoch": 0.0029204832072215587,
+      "grad_norm": 1.3284034155858548,
+      "learning_rate": 5.5e-07,
+      "loss": 0.8776,
+      "step": 11
+    },
+    {
+      "epoch": 0.0031859816806053365,
+      "grad_norm": 1.3011186118234537,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.8948,
+      "step": 12
+    },
+    {
+      "epoch": 0.0034514801539891147,
+      "grad_norm": 1.2341704129726287,
+      "learning_rate": 6.5e-07,
+      "loss": 0.8831,
+      "step": 13
+    },
+    {
+      "epoch": 0.0037169786273728925,
+      "grad_norm": 1.2514230159054196,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.8712,
+      "step": 14
+    },
+    {
+      "epoch": 0.00398247710075667,
+      "grad_norm": 1.1987764538870713,
+      "learning_rate": 7.5e-07,
+      "loss": 0.8823,
+      "step": 15
+    },
+    {
+      "epoch": 0.004247975574140449,
+      "grad_norm": 1.2672326543360966,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.8759,
+      "step": 16
+    },
+    {
+      "epoch": 0.004513474047524227,
+      "grad_norm": 1.2320004691691897,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.8802,
+      "step": 17
+    },
+    {
+      "epoch": 0.0047789725209080045,
+      "grad_norm": 1.1676684676908597,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.854,
+      "step": 18
+    },
+    {
+      "epoch": 0.005044470994291783,
+      "grad_norm": 1.1050533839495187,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.8827,
+      "step": 19
+    },
+    {
+      "epoch": 0.005309969467675561,
+      "grad_norm": 1.0752828509499528,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.8508,
+      "step": 20
+    },
+    {
+      "epoch": 0.005575467941059339,
+      "grad_norm": 1.0120183920536465,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.8603,
+      "step": 21
+    },
+    {
+      "epoch": 0.005840966414443117,
+      "grad_norm": 0.994980565363457,
+      "learning_rate": 1.1e-06,
+      "loss": 0.8676,
+      "step": 22
+    },
+    {
+      "epoch": 0.006106464887826895,
+      "grad_norm": 0.8764882048485012,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.8489,
+      "step": 23
+    },
+    {
+      "epoch": 0.006371963361210673,
+      "grad_norm": 0.8550001678984946,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.8422,
+      "step": 24
+    },
+    {
+      "epoch": 0.006637461834594451,
+      "grad_norm": 0.7578437022717379,
+      "learning_rate": 1.25e-06,
+      "loss": 0.8232,
+      "step": 25
+    },
+    {
+      "epoch": 0.006902960307978229,
+      "grad_norm": 0.6992717109150458,
+      "learning_rate": 1.3e-06,
+      "loss": 0.8427,
+      "step": 26
+    },
+    {
+      "epoch": 0.007168458781362007,
+      "grad_norm": 0.6689093642420254,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.7961,
+      "step": 27
+    },
+    {
+      "epoch": 0.007433957254745785,
+      "grad_norm": 0.6399612312421933,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.8,
+      "step": 28
+    },
+    {
+      "epoch": 0.007699455728129564,
+      "grad_norm": 0.572695626409587,
+      "learning_rate": 1.45e-06,
+      "loss": 0.7684,
+      "step": 29
+    },
+    {
+      "epoch": 0.00796495420151334,
+      "grad_norm": 0.6000840388722052,
+      "learning_rate": 1.5e-06,
+      "loss": 0.8152,
+      "step": 30
+    },
+    {
+      "epoch": 0.00823045267489712,
+      "grad_norm": 0.5831757904515951,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.825,
+      "step": 31
+    },
+    {
+      "epoch": 0.008495951148280898,
+      "grad_norm": 0.5376764754780556,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.7904,
+      "step": 32
+    },
+    {
+      "epoch": 0.008761449621664676,
+      "grad_norm": 0.544894260928256,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.7981,
+      "step": 33
+    },
+    {
+      "epoch": 0.009026948095048453,
+      "grad_norm": 0.5044098805336641,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.8049,
+      "step": 34
+    },
+    {
+      "epoch": 0.009292446568432231,
+      "grad_norm": 0.5105087348699734,
+      "learning_rate": 1.75e-06,
+      "loss": 0.8189,
+      "step": 35
+    },
+    {
+      "epoch": 0.009557945041816009,
+      "grad_norm": 0.5211090086716426,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.7974,
+      "step": 36
+    },
+    {
+      "epoch": 0.009823443515199787,
+      "grad_norm": 0.5131137326253041,
+      "learning_rate": 1.85e-06,
+      "loss": 0.7662,
+      "step": 37
+    },
+    {
+      "epoch": 0.010088941988583566,
+      "grad_norm": 0.49907189290405146,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.747,
+      "step": 38
+    },
+    {
+      "epoch": 0.010354440461967344,
+      "grad_norm": 0.47771973021953945,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.7463,
+      "step": 39
+    },
+    {
+      "epoch": 0.010619938935351122,
+      "grad_norm": 0.4785638397033305,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8147,
+      "step": 40
+    },
+    {
+      "epoch": 0.0108854374087349,
+      "grad_norm": 0.4558671534536234,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7741,
+      "step": 41
+    },
+    {
+      "epoch": 0.011150935882118677,
+      "grad_norm": 0.44950473746412367,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.7629,
+      "step": 42
+    },
+    {
+      "epoch": 0.011416434355502455,
+      "grad_norm": 0.4443003313102011,
+      "learning_rate": 2.15e-06,
+      "loss": 0.7771,
+      "step": 43
+    },
+    {
+      "epoch": 0.011681932828886235,
+      "grad_norm": 0.3910766012475798,
+      "learning_rate": 2.2e-06,
+      "loss": 0.7125,
+      "step": 44
+    },
+    {
+      "epoch": 0.011947431302270013,
+      "grad_norm": 0.4008723150350752,
+      "learning_rate": 2.25e-06,
+      "loss": 0.7306,
+      "step": 45
+    },
+    {
+      "epoch": 0.01221292977565379,
+      "grad_norm": 0.3754687761913381,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.7331,
+      "step": 46
+    },
+    {
+      "epoch": 0.012478428249037568,
+      "grad_norm": 0.36958796475999844,
+      "learning_rate": 2.35e-06,
+      "loss": 0.7114,
+      "step": 47
+    },
+    {
+      "epoch": 0.012743926722421346,
+      "grad_norm": 0.36066586543548546,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.7646,
+      "step": 48
+    },
+    {
+      "epoch": 0.013009425195805124,
+      "grad_norm": 0.377952015244848,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.7574,
+      "step": 49
+    },
+    {
+      "epoch": 0.013274923669188901,
+      "grad_norm": 0.35033265192734125,
+      "learning_rate": 2.5e-06,
+      "loss": 0.7175,
+      "step": 50
+    },
+    {
+      "epoch": 0.013540422142572681,
+      "grad_norm": 0.32967169997783835,
+      "learning_rate": 2.55e-06,
+      "loss": 0.7142,
+      "step": 51
+    },
+    {
+      "epoch": 0.013805920615956459,
+      "grad_norm": 0.3128738432355881,
+      "learning_rate": 2.6e-06,
+      "loss": 0.772,
+      "step": 52
+    },
+    {
+      "epoch": 0.014071419089340237,
+      "grad_norm": 0.3121467711239632,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.7435,
+      "step": 53
+    },
+    {
+      "epoch": 0.014336917562724014,
+      "grad_norm": 0.30406168256466826,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.7307,
+      "step": 54
+    },
+    {
+      "epoch": 0.014602416036107792,
+      "grad_norm": 0.30253139985404487,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.764,
+      "step": 55
+    },
+    {
+      "epoch": 0.01486791450949157,
+      "grad_norm": 0.2932645101598996,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.7545,
+      "step": 56
+    },
+    {
+      "epoch": 0.015133412982875348,
+      "grad_norm": 0.28874763194548436,
+      "learning_rate": 2.85e-06,
+      "loss": 0.7204,
+      "step": 57
+    },
+    {
+      "epoch": 0.015398911456259127,
+      "grad_norm": 0.3043644175395429,
+      "learning_rate": 2.9e-06,
+      "loss": 0.7431,
+      "step": 58
+    },
+    {
+      "epoch": 0.015664409929642905,
+      "grad_norm": 0.29277216204800016,
+      "learning_rate": 2.95e-06,
+      "loss": 0.6851,
+      "step": 59
+    },
+    {
+      "epoch": 0.01592990840302668,
+      "grad_norm": 0.2999037795611065,
+      "learning_rate": 3e-06,
+      "loss": 0.7458,
+      "step": 60
+    },
+    {
+      "epoch": 0.01619540687641046,
+      "grad_norm": 0.28603682629407434,
+      "learning_rate": 3.05e-06,
+      "loss": 0.743,
+      "step": 61
+    },
+    {
+      "epoch": 0.01646090534979424,
+      "grad_norm": 0.28812321057239954,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.7232,
+      "step": 62
+    },
+    {
+      "epoch": 0.016726403823178016,
+      "grad_norm": 0.2941773752215629,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.7664,
+      "step": 63
+    },
+    {
+      "epoch": 0.016991902296561796,
+      "grad_norm": 0.28683774669729817,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.7103,
+      "step": 64
+    },
+    {
+      "epoch": 0.01725740076994557,
+      "grad_norm": 0.268635186496577,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.7105,
+      "step": 65
+    },
+    {
+      "epoch": 0.01752289924332935,
+      "grad_norm": 0.27286564198875624,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.7309,
+      "step": 66
+    },
+    {
+      "epoch": 0.017788397716713127,
+      "grad_norm": 0.27685825537587966,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.6991,
+      "step": 67
+    },
+    {
+      "epoch": 0.018053896190096907,
+      "grad_norm": 0.26542577831742614,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.7258,
+      "step": 68
+    },
+    {
+      "epoch": 0.018319394663480686,
+      "grad_norm": 0.25455554558379656,
+      "learning_rate": 3.45e-06,
+      "loss": 0.6762,
+      "step": 69
+    },
+    {
+      "epoch": 0.018584893136864462,
+      "grad_norm": 0.26551700128237027,
+      "learning_rate": 3.5e-06,
+      "loss": 0.7118,
+      "step": 70
+    },
+    {
+      "epoch": 0.018850391610248242,
+      "grad_norm": 0.26405604223669027,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.74,
+      "step": 71
+    },
+    {
+      "epoch": 0.019115890083632018,
+      "grad_norm": 0.2682710044475517,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7176,
+      "step": 72
+    },
+    {
+      "epoch": 0.019381388557015797,
+      "grad_norm": 0.26835845009771225,
+      "learning_rate": 3.65e-06,
+      "loss": 0.7033,
+      "step": 73
+    },
+    {
+      "epoch": 0.019646887030399574,
+      "grad_norm": 0.27186708180710134,
+      "learning_rate": 3.7e-06,
+      "loss": 0.7531,
+      "step": 74
+    },
+    {
+      "epoch": 0.019912385503783353,
+      "grad_norm": 0.26739347296891547,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.7331,
+      "step": 75
+    },
+    {
+      "epoch": 0.020177883977167133,
+      "grad_norm": 0.2620288921624337,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.6777,
+      "step": 76
+    },
+    {
+      "epoch": 0.02044338245055091,
+      "grad_norm": 0.27361490719897,
+      "learning_rate": 3.85e-06,
+      "loss": 0.766,
+      "step": 77
+    },
+    {
+      "epoch": 0.020708880923934688,
+      "grad_norm": 0.2532185746262512,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.7097,
+      "step": 78
+    },
+    {
+      "epoch": 0.020974379397318464,
+      "grad_norm": 0.2567082880948565,
+      "learning_rate": 3.95e-06,
+      "loss": 0.6954,
+      "step": 79
+    },
+    {
+      "epoch": 0.021239877870702244,
+      "grad_norm": 0.2566665605997135,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7383,
+      "step": 80
+    },
+    {
+      "epoch": 0.021505376344086023,
+      "grad_norm": 0.25736368043811986,
+      "learning_rate": 4.05e-06,
+      "loss": 0.7464,
+      "step": 81
+    },
+    {
+      "epoch": 0.0217708748174698,
+      "grad_norm": 0.2524846582503469,
+      "learning_rate": 4.1e-06,
+      "loss": 0.7511,
+      "step": 82
+    },
+    {
+      "epoch": 0.02203637329085358,
+      "grad_norm": 0.25627792119137593,
+      "learning_rate": 4.15e-06,
+      "loss": 0.745,
+      "step": 83
+    },
+    {
+      "epoch": 0.022301871764237355,
+      "grad_norm": 0.2637730516534178,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.75,
+      "step": 84
+    },
+    {
+      "epoch": 0.022567370237621134,
+      "grad_norm": 0.25510395075205117,
+      "learning_rate": 4.25e-06,
+      "loss": 0.6979,
+      "step": 85
+    },
+    {
+      "epoch": 0.02283286871100491,
+      "grad_norm": 0.2890559540691289,
+      "learning_rate": 4.3e-06,
+      "loss": 0.7075,
+      "step": 86
+    },
+    {
+      "epoch": 0.02309836718438869,
+      "grad_norm": 0.2509012922405937,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.7087,
+      "step": 87
+    },
+    {
+      "epoch": 0.02336386565777247,
+      "grad_norm": 0.25550464739881684,
+      "learning_rate": 4.4e-06,
+      "loss": 0.7016,
+      "step": 88
+    },
+    {
+      "epoch": 0.023629364131156246,
+      "grad_norm": 0.2556178702239233,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.6977,
+      "step": 89
+    },
+    {
+      "epoch": 0.023894862604540025,
+      "grad_norm": 0.24845133311886222,
+      "learning_rate": 4.5e-06,
+      "loss": 0.6902,
+      "step": 90
+    },
+    {
+      "epoch": 0.0241603610779238,
+      "grad_norm": 0.2487833268099019,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.7192,
+      "step": 91
+    },
+    {
+      "epoch": 0.02442585955130758,
+      "grad_norm": 0.2523603689649045,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.6753,
+      "step": 92
+    },
+    {
+      "epoch": 0.024691358024691357,
+      "grad_norm": 0.24635668385065182,
+      "learning_rate": 4.65e-06,
+      "loss": 0.6986,
+      "step": 93
+    },
+    {
+      "epoch": 0.024956856498075136,
+      "grad_norm": 0.26107306703307287,
+      "learning_rate": 4.7e-06,
+      "loss": 0.7201,
+      "step": 94
+    },
+    {
+      "epoch": 0.025222354971458916,
+      "grad_norm": 0.24664587374623534,
+      "learning_rate": 4.75e-06,
+      "loss": 0.6522,
+      "step": 95
+    },
+    {
+      "epoch": 0.025487853444842692,
+      "grad_norm": 0.2520881808354904,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.7131,
+      "step": 96
+    },
+    {
+      "epoch": 0.02575335191822647,
+      "grad_norm": 0.25598702780366006,
+      "learning_rate": 4.85e-06,
+      "loss": 0.7467,
+      "step": 97
+    },
+    {
+      "epoch": 0.026018850391610247,
+      "grad_norm": 0.2616550173587426,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.7336,
+      "step": 98
+    },
+    {
+      "epoch": 0.026284348864994027,
+      "grad_norm": 0.25891740874006586,
+      "learning_rate": 4.95e-06,
+      "loss": 0.7221,
+      "step": 99
+    },
+    {
+      "epoch": 0.026549847338377803,
+      "grad_norm": 0.26549726719468714,
+      "learning_rate": 5e-06,
+      "loss": 0.7071,
+      "step": 100
+    },
+    {
+      "epoch": 0.026815345811761582,
+      "grad_norm": 0.25874316429601724,
+      "learning_rate": 4.99999997562194e-06,
+      "loss": 0.755,
+      "step": 101
+    },
+    {
+      "epoch": 0.027080844285145362,
+      "grad_norm": 0.2511929232702433,
+      "learning_rate": 4.999999902487759e-06,
+      "loss": 0.7244,
+      "step": 102
+    },
+    {
+      "epoch": 0.027346342758529138,
+      "grad_norm": 0.2558497677742771,
+      "learning_rate": 4.999999780597459e-06,
+      "loss": 0.7185,
+      "step": 103
+    },
+    {
+      "epoch": 0.027611841231912918,
+      "grad_norm": 0.25423090267132625,
+      "learning_rate": 4.999999609951042e-06,
+      "loss": 0.7142,
+      "step": 104
+    },
+    {
+      "epoch": 0.027877339705296694,
+      "grad_norm": 0.25520571574035367,
+      "learning_rate": 4.9999993905485115e-06,
+      "loss": 0.7254,
+      "step": 105
+    },
+    {
+      "epoch": 0.028142838178680473,
+      "grad_norm": 0.2577386731233315,
+      "learning_rate": 4.999999122389871e-06,
+      "loss": 0.7032,
+      "step": 106
+    },
+    {
+      "epoch": 0.02840833665206425,
+      "grad_norm": 0.24645487363162205,
+      "learning_rate": 4.999998805475128e-06,
+      "loss": 0.677,
+      "step": 107
+    },
+    {
+      "epoch": 0.02867383512544803,
+      "grad_norm": 0.2505991297766978,
+      "learning_rate": 4.999998439804287e-06,
+      "loss": 0.7279,
+      "step": 108
+    },
+    {
+      "epoch": 0.028939333598831808,
+      "grad_norm": 0.24612817265821374,
+      "learning_rate": 4.999998025377354e-06,
+      "loss": 0.6821,
+      "step": 109
+    },
+    {
+      "epoch": 0.029204832072215584,
+      "grad_norm": 0.2559063653494428,
+      "learning_rate": 4.99999756219434e-06,
+      "loss": 0.7362,
+      "step": 110
+    },
+    {
+      "epoch": 0.029470330545599364,
+      "grad_norm": 0.2529304236380953,
+      "learning_rate": 4.999997050255252e-06,
+      "loss": 0.7276,
+      "step": 111
+    },
+    {
+      "epoch": 0.02973582901898314,
+      "grad_norm": 0.25661217078094695,
+      "learning_rate": 4.9999964895601e-06,
+      "loss": 0.7101,
+      "step": 112
+    },
+    {
+      "epoch": 0.03000132749236692,
+      "grad_norm": 0.25593901891463355,
+      "learning_rate": 4.999995880108896e-06,
+      "loss": 0.7209,
+      "step": 113
+    },
+    {
+      "epoch": 0.030266825965750695,
+      "grad_norm": 0.2468092183814833,
+      "learning_rate": 4.999995221901651e-06,
+      "loss": 0.674,
+      "step": 114
+    },
+    {
+      "epoch": 0.030532324439134475,
+      "grad_norm": 0.25989065906974634,
+      "learning_rate": 4.9999945149383785e-06,
+      "loss": 0.6863,
+      "step": 115
+    },
+    {
+      "epoch": 0.030797822912518254,
+      "grad_norm": 0.2501700260472302,
+      "learning_rate": 4.999993759219091e-06,
+      "loss": 0.7129,
+      "step": 116
+    },
+    {
+      "epoch": 0.03106332138590203,
+      "grad_norm": 0.2477100927680114,
+      "learning_rate": 4.999992954743805e-06,
+      "loss": 0.6745,
+      "step": 117
+    },
+    {
+      "epoch": 0.03132881985928581,
+      "grad_norm": 0.25731972201226244,
+      "learning_rate": 4.999992101512536e-06,
+      "loss": 0.7153,
+      "step": 118
+    },
+    {
+      "epoch": 0.03159431833266959,
+      "grad_norm": 0.25648861667989575,
+      "learning_rate": 4.999991199525299e-06,
+      "loss": 0.6942,
+      "step": 119
+    },
+    {
+      "epoch": 0.03185981680605336,
+      "grad_norm": 0.25543098235740036,
+      "learning_rate": 4.999990248782113e-06,
+      "loss": 0.7018,
+      "step": 120
+    },
+    {
+      "epoch": 0.03212531527943714,
+      "grad_norm": 0.2573491732459776,
+      "learning_rate": 4.999989249282995e-06,
+      "loss": 0.721,
+      "step": 121
+    },
+    {
+      "epoch": 0.03239081375282092,
+      "grad_norm": 0.24897639401490088,
+      "learning_rate": 4.999988201027967e-06,
+      "loss": 0.6665,
+      "step": 122
+    },
+    {
+      "epoch": 0.0326563122262047,
+      "grad_norm": 0.25353835471321695,
+      "learning_rate": 4.999987104017048e-06,
+      "loss": 0.6535,
+      "step": 123
+    },
+    {
+      "epoch": 0.03292181069958848,
+      "grad_norm": 0.25324921439833153,
+      "learning_rate": 4.999985958250259e-06,
+      "loss": 0.6652,
+      "step": 124
+    },
+    {
+      "epoch": 0.03318730917297225,
+      "grad_norm": 0.2531800167423538,
+      "learning_rate": 4.999984763727622e-06,
+      "loss": 0.6899,
+      "step": 125
+    },
+    {
+      "epoch": 0.03345280764635603,
+      "grad_norm": 0.25245484453871886,
+      "learning_rate": 4.999983520449162e-06,
+      "loss": 0.7255,
+      "step": 126
+    },
+    {
+      "epoch": 0.03371830611973981,
+      "grad_norm": 0.24636543509983638,
+      "learning_rate": 4.999982228414902e-06,
+      "loss": 0.7039,
+      "step": 127
+    },
+    {
+      "epoch": 0.03398380459312359,
+      "grad_norm": 0.2596679023066606,
+      "learning_rate": 4.999980887624868e-06,
+      "loss": 0.7585,
+      "step": 128
+    },
+    {
+      "epoch": 0.03424930306650737,
+      "grad_norm": 0.2544471820006133,
+      "learning_rate": 4.999979498079085e-06,
+      "loss": 0.7226,
+      "step": 129
+    },
+    {
+      "epoch": 0.03451480153989114,
+      "grad_norm": 0.25541840400438764,
+      "learning_rate": 4.999978059777582e-06,
+      "loss": 0.7151,
+      "step": 130
+    },
+    {
+      "epoch": 0.03478030001327492,
+      "grad_norm": 0.25798389533674054,
+      "learning_rate": 4.999976572720384e-06,
+      "loss": 0.7209,
+      "step": 131
+    },
+    {
+      "epoch": 0.0350457984866587,
+      "grad_norm": 0.25504242933602034,
+      "learning_rate": 4.999975036907523e-06,
+      "loss": 0.7069,
+      "step": 132
+    },
+    {
+      "epoch": 0.03531129696004248,
+      "grad_norm": 0.2509343509012928,
+      "learning_rate": 4.999973452339028e-06,
+      "loss": 0.7057,
+      "step": 133
+    },
+    {
+      "epoch": 0.035576795433426255,
+      "grad_norm": 0.2513894881269105,
+      "learning_rate": 4.99997181901493e-06,
+      "loss": 0.7033,
+      "step": 134
+    },
+    {
+      "epoch": 0.035842293906810034,
+      "grad_norm": 0.2584358731096311,
+      "learning_rate": 4.9999701369352595e-06,
+      "loss": 0.6961,
+      "step": 135
+    },
+    {
+      "epoch": 0.036107792380193814,
+      "grad_norm": 0.25256173256109765,
+      "learning_rate": 4.99996840610005e-06,
+      "loss": 0.7059,
+      "step": 136
+    },
+    {
+      "epoch": 0.03637329085357759,
+      "grad_norm": 0.27139559456131557,
+      "learning_rate": 4.999966626509336e-06,
+      "loss": 0.7153,
+      "step": 137
+    },
+    {
+      "epoch": 0.03663878932696137,
+      "grad_norm": 0.26317533069134746,
+      "learning_rate": 4.999964798163153e-06,
+      "loss": 0.728,
+      "step": 138
+    },
+    {
+      "epoch": 0.036904287800345145,
+      "grad_norm": 0.2523361779869315,
+      "learning_rate": 4.9999629210615345e-06,
+      "loss": 0.6856,
+      "step": 139
+    },
+    {
+      "epoch": 0.037169786273728925,
+      "grad_norm": 0.25623775469647725,
+      "learning_rate": 4.999960995204518e-06,
+      "loss": 0.6928,
+      "step": 140
+    },
+    {
+      "epoch": 0.037435284747112704,
+      "grad_norm": 0.2606127602795901,
+      "learning_rate": 4.999959020592141e-06,
+      "loss": 0.7049,
+      "step": 141
+    },
+    {
+      "epoch": 0.037700783220496484,
+      "grad_norm": 0.25596727411162024,
+      "learning_rate": 4.999956997224443e-06,
+      "loss": 0.7175,
+      "step": 142
+    },
+    {
+      "epoch": 0.03796628169388026,
+      "grad_norm": 0.26383740632476566,
+      "learning_rate": 4.999954925101462e-06,
+      "loss": 0.6857,
+      "step": 143
+    },
+    {
+      "epoch": 0.038231780167264036,
+      "grad_norm": 0.26891725396566324,
+      "learning_rate": 4.99995280422324e-06,
+      "loss": 0.6928,
+      "step": 144
+    },
+    {
+      "epoch": 0.038497278640647815,
+      "grad_norm": 0.24704940977635542,
+      "learning_rate": 4.999950634589815e-06,
+      "loss": 0.6439,
+      "step": 145
+    },
+    {
+      "epoch": 0.038762777114031595,
+      "grad_norm": 0.25891770187188756,
+      "learning_rate": 4.9999484162012335e-06,
+      "loss": 0.7084,
+      "step": 146
+    },
+    {
+      "epoch": 0.039028275587415374,
+      "grad_norm": 0.2747765622828544,
+      "learning_rate": 4.999946149057536e-06,
+      "loss": 0.71,
+      "step": 147
+    },
+    {
+      "epoch": 0.03929377406079915,
+      "grad_norm": 0.2629632566296953,
+      "learning_rate": 4.999943833158769e-06,
+      "loss": 0.7176,
+      "step": 148
+    },
+    {
+      "epoch": 0.03955927253418293,
+      "grad_norm": 0.26013386317185416,
+      "learning_rate": 4.999941468504975e-06,
+      "loss": 0.7096,
+      "step": 149
+    },
+    {
+      "epoch": 0.039824771007566706,
+      "grad_norm": 0.2602432005734629,
+      "learning_rate": 4.999939055096203e-06,
+      "loss": 0.7376,
+      "step": 150
+    },
+    {
+      "epoch": 0.040090269480950486,
+      "grad_norm": 0.255755143030168,
+      "learning_rate": 4.999936592932497e-06,
+      "loss": 0.7107,
+      "step": 151
+    },
+    {
+      "epoch": 0.040355767954334265,
+      "grad_norm": 0.26459441298069336,
+      "learning_rate": 4.9999340820139065e-06,
+      "loss": 0.6986,
+      "step": 152
+    },
+    {
+      "epoch": 0.04062126642771804,
+      "grad_norm": 0.2580175240666269,
+      "learning_rate": 4.999931522340482e-06,
+      "loss": 0.6941,
+      "step": 153
+    },
+    {
+      "epoch": 0.04088676490110182,
+      "grad_norm": 0.2548066104486532,
+      "learning_rate": 4.999928913912271e-06,
+      "loss": 0.7073,
+      "step": 154
+    },
+    {
+      "epoch": 0.0411522633744856,
+      "grad_norm": 0.2566359870885854,
+      "learning_rate": 4.999926256729325e-06,
+      "loss": 0.695,
+      "step": 155
+    },
+    {
+      "epoch": 0.041417761847869376,
+      "grad_norm": 0.2549048866184483,
+      "learning_rate": 4.9999235507916966e-06,
+      "loss": 0.6859,
+      "step": 156
+    },
+    {
+      "epoch": 0.041683260321253156,
+      "grad_norm": 0.2632131747724388,
+      "learning_rate": 4.999920796099438e-06,
+      "loss": 0.7316,
+      "step": 157
+    },
+    {
+      "epoch": 0.04194875879463693,
+      "grad_norm": 0.2545162111395541,
+      "learning_rate": 4.9999179926526035e-06,
+      "loss": 0.7016,
+      "step": 158
+    },
+    {
+      "epoch": 0.04221425726802071,
+      "grad_norm": 0.2705893544135468,
+      "learning_rate": 4.999915140451245e-06,
+      "loss": 0.7202,
+      "step": 159
+    },
+    {
+      "epoch": 0.04247975574140449,
+      "grad_norm": 0.26185974573938753,
+      "learning_rate": 4.999912239495424e-06,
+      "loss": 0.6786,
+      "step": 160
+    },
+    {
+      "epoch": 0.04274525421478827,
+      "grad_norm": 0.26128486781124644,
+      "learning_rate": 4.9999092897851905e-06,
+      "loss": 0.69,
+      "step": 161
+    },
+    {
+      "epoch": 0.043010752688172046,
+      "grad_norm": 0.27173877465739715,
+      "learning_rate": 4.9999062913206065e-06,
+      "loss": 0.7127,
+      "step": 162
+    },
+    {
+      "epoch": 0.04327625116155582,
+      "grad_norm": 0.2623584024641717,
+      "learning_rate": 4.999903244101728e-06,
+      "loss": 0.679,
+      "step": 163
+    },
+    {
+      "epoch": 0.0435417496349396,
+      "grad_norm": 0.26064668777175376,
+      "learning_rate": 4.999900148128617e-06,
+      "loss": 0.7104,
+      "step": 164
+    },
+    {
+      "epoch": 0.04380724810832338,
+      "grad_norm": 0.2620858617788341,
+      "learning_rate": 4.999897003401331e-06,
+      "loss": 0.7191,
+      "step": 165
+    },
+    {
+      "epoch": 0.04407274658170716,
+      "grad_norm": 0.2592932893814249,
+      "learning_rate": 4.999893809919934e-06,
+      "loss": 0.6995,
+      "step": 166
+    },
+    {
+      "epoch": 0.04433824505509093,
+      "grad_norm": 0.26211056697252927,
+      "learning_rate": 4.999890567684485e-06,
+      "loss": 0.6786,
+      "step": 167
+    },
+    {
+      "epoch": 0.04460374352847471,
+      "grad_norm": 0.26822117591960615,
+      "learning_rate": 4.99988727669505e-06,
+      "loss": 0.7037,
+      "step": 168
+    },
+    {
+      "epoch": 0.04486924200185849,
+      "grad_norm": 0.2717335949331894,
+      "learning_rate": 4.999883936951693e-06,
+      "loss": 0.6964,
+      "step": 169
+    },
+    {
+      "epoch": 0.04513474047524227,
+      "grad_norm": 0.26646268613952157,
+      "learning_rate": 4.999880548454478e-06,
+      "loss": 0.6981,
+      "step": 170
+    },
+    {
+      "epoch": 0.04540023894862605,
+      "grad_norm": 0.26895211181167006,
+      "learning_rate": 4.999877111203471e-06,
+      "loss": 0.7185,
+      "step": 171
+    },
+    {
+      "epoch": 0.04566573742200982,
+      "grad_norm": 0.26596194964860426,
+      "learning_rate": 4.99987362519874e-06,
+      "loss": 0.6934,
+      "step": 172
+    },
+    {
+      "epoch": 0.0459312358953936,
+      "grad_norm": 0.2638074081092255,
+      "learning_rate": 4.999870090440352e-06,
+      "loss": 0.6726,
+      "step": 173
+    },
+    {
+      "epoch": 0.04619673436877738,
+      "grad_norm": 0.2685766772130088,
+      "learning_rate": 4.999866506928377e-06,
+      "loss": 0.7,
+      "step": 174
+    },
+    {
+      "epoch": 0.04646223284216116,
+      "grad_norm": 0.26125994565790756,
+      "learning_rate": 4.999862874662885e-06,
+      "loss": 0.6729,
+      "step": 175
+    },
+    {
+      "epoch": 0.04672773131554494,
+      "grad_norm": 0.2585324922913285,
+      "learning_rate": 4.999859193643946e-06,
+      "loss": 0.6717,
+      "step": 176
+    },
+    {
+      "epoch": 0.04699322978892871,
+      "grad_norm": 0.2719323796706383,
+      "learning_rate": 4.999855463871631e-06,
+      "loss": 0.7109,
+      "step": 177
+    },
+    {
+      "epoch": 0.04725872826231249,
+      "grad_norm": 0.28112974850028344,
+      "learning_rate": 4.999851685346015e-06,
+      "loss": 0.7035,
+      "step": 178
+    },
+    {
+      "epoch": 0.04752422673569627,
+      "grad_norm": 0.263044613805147,
+      "learning_rate": 4.9998478580671694e-06,
+      "loss": 0.6709,
+      "step": 179
+    },
+    {
+      "epoch": 0.04778972520908005,
+      "grad_norm": 0.26340405459611804,
+      "learning_rate": 4.99984398203517e-06,
+      "loss": 0.6843,
+      "step": 180
+    },
+    {
+      "epoch": 0.04805522368246382,
+      "grad_norm": 0.2717026054338667,
+      "learning_rate": 4.999840057250092e-06,
+      "loss": 0.7074,
+      "step": 181
+    },
+    {
+      "epoch": 0.0483207221558476,
+      "grad_norm": 0.26018460501123286,
+      "learning_rate": 4.999836083712014e-06,
+      "loss": 0.6874,
+      "step": 182
+    },
+    {
+      "epoch": 0.04858622062923138,
+      "grad_norm": 0.2962560946041906,
+      "learning_rate": 4.99983206142101e-06,
+      "loss": 0.6984,
+      "step": 183
+    },
+    {
+      "epoch": 0.04885171910261516,
+      "grad_norm": 0.26542573047174617,
+      "learning_rate": 4.999827990377161e-06,
+      "loss": 0.6746,
+      "step": 184
+    },
+    {
+      "epoch": 0.04911721757599894,
+      "grad_norm": 0.2627478069344486,
+      "learning_rate": 4.999823870580545e-06,
+      "loss": 0.6983,
+      "step": 185
+    },
+    {
+      "epoch": 0.04938271604938271,
+      "grad_norm": 0.2695317200701367,
+      "learning_rate": 4.999819702031243e-06,
+      "loss": 0.6904,
+      "step": 186
+    },
+    {
+      "epoch": 0.04964821452276649,
+      "grad_norm": 0.2692151455364245,
+      "learning_rate": 4.999815484729337e-06,
+      "loss": 0.7121,
+      "step": 187
+    },
+    {
+      "epoch": 0.04991371299615027,
+      "grad_norm": 0.2670591361605896,
+      "learning_rate": 4.999811218674907e-06,
+      "loss": 0.6779,
+      "step": 188
+    },
+    {
+      "epoch": 0.05017921146953405,
+      "grad_norm": 0.2665842036825122,
+      "learning_rate": 4.999806903868039e-06,
+      "loss": 0.6778,
+      "step": 189
+    },
+    {
+      "epoch": 0.05044470994291783,
+      "grad_norm": 0.26415858006089876,
+      "learning_rate": 4.999802540308815e-06,
+      "loss": 0.6705,
+      "step": 190
+    },
+    {
+      "epoch": 0.050710208416301604,
+      "grad_norm": 0.2699353100740373,
+      "learning_rate": 4.999798127997322e-06,
+      "loss": 0.6625,
+      "step": 191
+    },
+    {
+      "epoch": 0.050975706889685383,
+      "grad_norm": 0.276360043882742,
+      "learning_rate": 4.999793666933643e-06,
+      "loss": 0.7146,
+      "step": 192
+    },
+    {
+      "epoch": 0.05124120536306916,
+      "grad_norm": 0.27731778712619215,
+      "learning_rate": 4.999789157117868e-06,
+      "loss": 0.7244,
+      "step": 193
+    },
+    {
+      "epoch": 0.05150670383645294,
+      "grad_norm": 0.28021136346995323,
+      "learning_rate": 4.999784598550085e-06,
+      "loss": 0.7223,
+      "step": 194
+    },
+    {
+      "epoch": 0.051772202309836715,
+      "grad_norm": 0.27363447721763196,
+      "learning_rate": 4.99977999123038e-06,
+      "loss": 0.6751,
+      "step": 195
+    },
+    {
+      "epoch": 0.052037700783220495,
+      "grad_norm": 0.27293263360243547,
+      "learning_rate": 4.999775335158845e-06,
+      "loss": 0.666,
+      "step": 196
+    },
+    {
+      "epoch": 0.052303199256604274,
+      "grad_norm": 0.26892809104063314,
+      "learning_rate": 4.999770630335571e-06,
+      "loss": 0.6627,
+      "step": 197
+    },
+    {
+      "epoch": 0.052568697729988054,
+      "grad_norm": 0.2655011939742062,
+      "learning_rate": 4.999765876760649e-06,
+      "loss": 0.6376,
+      "step": 198
+    },
+    {
+      "epoch": 0.05283419620337183,
+      "grad_norm": 0.27378030490933,
+      "learning_rate": 4.999761074434172e-06,
+      "loss": 0.699,
+      "step": 199
+    },
+    {
+      "epoch": 0.053099694676755606,
+      "grad_norm": 0.2797580892313466,
+      "learning_rate": 4.999756223356233e-06,
+      "loss": 0.6817,
+      "step": 200
+    },
+    {
+      "epoch": 0.053365193150139385,
+      "grad_norm": 0.27676547571532695,
+      "learning_rate": 4.999751323526928e-06,
+      "loss": 0.7059,
+      "step": 201
+    },
+    {
+      "epoch": 0.053630691623523165,
+      "grad_norm": 0.2800957514600047,
+      "learning_rate": 4.999746374946351e-06,
+      "loss": 0.7007,
+      "step": 202
+    },
+    {
+      "epoch": 0.053896190096906944,
+      "grad_norm": 0.2658748217580449,
+      "learning_rate": 4.9997413776146006e-06,
+      "loss": 0.7133,
+      "step": 203
+    },
+    {
+      "epoch": 0.054161688570290724,
+      "grad_norm": 0.27165720985917746,
+      "learning_rate": 4.999736331531772e-06,
+      "loss": 0.7057,
+      "step": 204
+    },
+    {
+      "epoch": 0.054427187043674496,
+      "grad_norm": 0.27110457882180444,
+      "learning_rate": 4.9997312366979644e-06,
+      "loss": 0.6923,
+      "step": 205
+    },
+    {
+      "epoch": 0.054692685517058276,
+      "grad_norm": 0.26608928525249387,
+      "learning_rate": 4.999726093113277e-06,
+      "loss": 0.6808,
+      "step": 206
+    },
+    {
+      "epoch": 0.054958183990442055,
+      "grad_norm": 0.27284820287434014,
+      "learning_rate": 4.9997209007778115e-06,
+      "loss": 0.6769,
+      "step": 207
+    },
+    {
+      "epoch": 0.055223682463825835,
+      "grad_norm": 0.2754244175594125,
+      "learning_rate": 4.9997156596916675e-06,
+      "loss": 0.6949,
+      "step": 208
+    },
+    {
+      "epoch": 0.05548918093720961,
+      "grad_norm": 0.280054498798005,
+      "learning_rate": 4.999710369854948e-06,
+      "loss": 0.7043,
+      "step": 209
+    },
+    {
+      "epoch": 0.05575467941059339,
+      "grad_norm": 0.27321589443694444,
+      "learning_rate": 4.9997050312677555e-06,
+      "loss": 0.6894,
+      "step": 210
+    },
+    {
+      "epoch": 0.05602017788397717,
+      "grad_norm": 0.2783392162014966,
+      "learning_rate": 4.999699643930194e-06,
+      "loss": 0.6887,
+      "step": 211
+    },
+    {
+      "epoch": 0.056285676357360946,
+      "grad_norm": 0.27440601906566764,
+      "learning_rate": 4.99969420784237e-06,
+      "loss": 0.6768,
+      "step": 212
+    },
+    {
+      "epoch": 0.056551174830744726,
+      "grad_norm": 0.2734307791646763,
+      "learning_rate": 4.999688723004388e-06,
+      "loss": 0.6838,
+      "step": 213
+    },
+    {
+      "epoch": 0.0568166733041285,
+      "grad_norm": 0.27394668345734086,
+      "learning_rate": 4.999683189416357e-06,
+      "loss": 0.6711,
+      "step": 214
+    },
+    {
+      "epoch": 0.05708217177751228,
+      "grad_norm": 0.28115025983760983,
+      "learning_rate": 4.999677607078382e-06,
+      "loss": 0.7068,
+      "step": 215
+    },
+    {
+      "epoch": 0.05734767025089606,
+      "grad_norm": 0.26899994312737147,
+      "learning_rate": 4.999671975990574e-06,
+      "loss": 0.6776,
+      "step": 216
+    },
+    {
+      "epoch": 0.05761316872427984,
+      "grad_norm": 0.2817314356169039,
+      "learning_rate": 4.999666296153042e-06,
+      "loss": 0.7044,
+      "step": 217
+    },
+    {
+      "epoch": 0.057878667197663616,
+      "grad_norm": 0.27902869905165584,
+      "learning_rate": 4.999660567565897e-06,
+      "loss": 0.6744,
+      "step": 218
+    },
+    {
+      "epoch": 0.05814416567104739,
+      "grad_norm": 0.27022592081224195,
+      "learning_rate": 4.9996547902292514e-06,
+      "loss": 0.7199,
+      "step": 219
+    },
+    {
+      "epoch": 0.05840966414443117,
+      "grad_norm": 0.27208455882650584,
+      "learning_rate": 4.999648964143217e-06,
+      "loss": 0.6908,
+      "step": 220
+    },
+    {
+      "epoch": 0.05867516261781495,
+      "grad_norm": 0.2737163124143047,
+      "learning_rate": 4.999643089307907e-06,
+      "loss": 0.6652,
+      "step": 221
+    },
+    {
+      "epoch": 0.05894066109119873,
+      "grad_norm": 0.2688280646468157,
+      "learning_rate": 4.999637165723437e-06,
+      "loss": 0.705,
+      "step": 222
+    },
+    {
+      "epoch": 0.05920615956458251,
+      "grad_norm": 0.26927072313433303,
+      "learning_rate": 4.999631193389922e-06,
+      "loss": 0.6798,
+      "step": 223
+    },
+    {
+      "epoch": 0.05947165803796628,
+      "grad_norm": 0.26469848070271984,
+      "learning_rate": 4.999625172307479e-06,
+      "loss": 0.6719,
+      "step": 224
+    },
+    {
+      "epoch": 0.05973715651135006,
+      "grad_norm": 0.27456219654052877,
+      "learning_rate": 4.9996191024762235e-06,
+      "loss": 0.6746,
+      "step": 225
+    },
+    {
+      "epoch": 0.06000265498473384,
+      "grad_norm": 0.27871327711391836,
+      "learning_rate": 4.999612983896277e-06,
+      "loss": 0.7,
+      "step": 226
+    },
+    {
+      "epoch": 0.06026815345811762,
+      "grad_norm": 0.27110710717724956,
+      "learning_rate": 4.999606816567756e-06,
+      "loss": 0.6953,
+      "step": 227
+    },
+    {
+      "epoch": 0.06053365193150139,
+      "grad_norm": 0.2755922210275337,
+      "learning_rate": 4.999600600490783e-06,
+      "loss": 0.6938,
+      "step": 228
+    },
+    {
+      "epoch": 0.06079915040488517,
+      "grad_norm": 0.28005633334714464,
+      "learning_rate": 4.999594335665477e-06,
+      "loss": 0.7015,
+      "step": 229
+    },
+    {
+      "epoch": 0.06106464887826895,
+      "grad_norm": 0.2825554641299085,
+      "learning_rate": 4.999588022091963e-06,
+      "loss": 0.6875,
+      "step": 230
+    },
+    {
+      "epoch": 0.06133014735165273,
+      "grad_norm": 0.28230308518275543,
+      "learning_rate": 4.999581659770362e-06,
+      "loss": 0.674,
+      "step": 231
+    },
+    {
+      "epoch": 0.06159564582503651,
+      "grad_norm": 0.28030786780792255,
+      "learning_rate": 4.999575248700797e-06,
+      "loss": 0.6727,
+      "step": 232
+    },
+    {
+      "epoch": 0.06186114429842028,
+      "grad_norm": 0.28322905286732064,
+      "learning_rate": 4.999568788883396e-06,
+      "loss": 0.6759,
+      "step": 233
+    },
+    {
+      "epoch": 0.06212664277180406,
+      "grad_norm": 0.28063162337237035,
+      "learning_rate": 4.999562280318284e-06,
+      "loss": 0.6507,
+      "step": 234
+    },
+    {
+      "epoch": 0.06239214124518784,
+      "grad_norm": 0.2803099234303004,
+      "learning_rate": 4.999555723005588e-06,
+      "loss": 0.6762,
+      "step": 235
+    },
+    {
+      "epoch": 0.06265763971857162,
+      "grad_norm": 0.29590886163225577,
+      "learning_rate": 4.9995491169454345e-06,
+      "loss": 0.7273,
+      "step": 236
+    },
+    {
+      "epoch": 0.0629231381919554,
+      "grad_norm": 0.2824885396276655,
+      "learning_rate": 4.999542462137953e-06,
+      "loss": 0.6792,
+      "step": 237
+    },
+    {
+      "epoch": 0.06318863666533918,
+      "grad_norm": 0.2816431483123683,
+      "learning_rate": 4.999535758583274e-06,
+      "loss": 0.69,
+      "step": 238
+    },
+    {
+      "epoch": 0.06345413513872296,
+      "grad_norm": 0.2884562386922022,
+      "learning_rate": 4.999529006281527e-06,
+      "loss": 0.7017,
+      "step": 239
+    },
+    {
+      "epoch": 0.06371963361210672,
+      "grad_norm": 0.28273311581933935,
+      "learning_rate": 4.999522205232846e-06,
+      "loss": 0.7186,
+      "step": 240
+    },
+    {
+      "epoch": 0.0639851320854905,
+      "grad_norm": 0.2783167115742715,
+      "learning_rate": 4.999515355437361e-06,
+      "loss": 0.6668,
+      "step": 241
+    },
+    {
+      "epoch": 0.06425063055887428,
+      "grad_norm": 0.2816825131413215,
+      "learning_rate": 4.999508456895207e-06,
+      "loss": 0.6843,
+      "step": 242
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 0.2801353687084795,
+      "learning_rate": 4.999501509606518e-06,
+      "loss": 0.6968,
+      "step": 243
+    },
+    {
+      "epoch": 0.06478162750564184,
+      "grad_norm": 0.2835734150455514,
+      "learning_rate": 4.9994945135714305e-06,
+      "loss": 0.6862,
+      "step": 244
+    },
+    {
+      "epoch": 0.06504712597902562,
+      "grad_norm": 0.27741725505020737,
+      "learning_rate": 4.999487468790079e-06,
+      "loss": 0.6647,
+      "step": 245
+    },
+    {
+      "epoch": 0.0653126244524094,
+      "grad_norm": 0.26681654360115015,
+      "learning_rate": 4.999480375262603e-06,
+      "loss": 0.643,
+      "step": 246
+    },
+    {
+      "epoch": 0.06557812292579318,
+      "grad_norm": 0.2789170195456692,
+      "learning_rate": 4.999473232989139e-06,
+      "loss": 0.7272,
+      "step": 247
+    },
+    {
+      "epoch": 0.06584362139917696,
+      "grad_norm": 0.27742709910456664,
+      "learning_rate": 4.999466041969828e-06,
+      "loss": 0.7021,
+      "step": 248
+    },
+    {
+      "epoch": 0.06610911987256073,
+      "grad_norm": 0.2911592528989723,
+      "learning_rate": 4.999458802204809e-06,
+      "loss": 0.7011,
+      "step": 249
+    },
+    {
+      "epoch": 0.0663746183459445,
+      "grad_norm": 0.2704968024069899,
+      "learning_rate": 4.999451513694224e-06,
+      "loss": 0.6421,
+      "step": 250
+    },
+    {
+      "epoch": 0.06664011681932829,
+      "grad_norm": 0.28089468200732376,
+      "learning_rate": 4.999444176438214e-06,
+      "loss": 0.7055,
+      "step": 251
+    },
+    {
+      "epoch": 0.06690561529271206,
+      "grad_norm": 0.2767204642624905,
+      "learning_rate": 4.999436790436924e-06,
+      "loss": 0.6503,
+      "step": 252
+    },
+    {
+      "epoch": 0.06717111376609584,
+      "grad_norm": 0.2922773807732511,
+      "learning_rate": 4.999429355690496e-06,
+      "loss": 0.7075,
+      "step": 253
+    },
+    {
+      "epoch": 0.06743661223947962,
+      "grad_norm": 0.2825676930712853,
+      "learning_rate": 4.999421872199076e-06,
+      "loss": 0.6934,
+      "step": 254
+    },
+    {
+      "epoch": 0.0677021107128634,
+      "grad_norm": 0.2875336732999721,
+      "learning_rate": 4.99941433996281e-06,
+      "loss": 0.6761,
+      "step": 255
+    },
+    {
+      "epoch": 0.06796760918624718,
+      "grad_norm": 0.29005402811931774,
+      "learning_rate": 4.999406758981845e-06,
+      "loss": 0.6926,
+      "step": 256
+    },
+    {
+      "epoch": 0.06823310765963096,
+      "grad_norm": 0.28761923552882396,
+      "learning_rate": 4.99939912925633e-06,
+      "loss": 0.7085,
+      "step": 257
+    },
+    {
+      "epoch": 0.06849860613301474,
+      "grad_norm": 0.28727780363272737,
+      "learning_rate": 4.999391450786409e-06,
+      "loss": 0.6802,
+      "step": 258
+    },
+    {
+      "epoch": 0.06876410460639851,
+      "grad_norm": 0.28205517649545725,
+      "learning_rate": 4.999383723572238e-06,
+      "loss": 0.6708,
+      "step": 259
+    },
+    {
+      "epoch": 0.06902960307978229,
+      "grad_norm": 0.28491849512760403,
+      "learning_rate": 4.999375947613963e-06,
+      "loss": 0.6706,
+      "step": 260
+    },
+    {
+      "epoch": 0.06929510155316607,
+      "grad_norm": 0.2829201389826739,
+      "learning_rate": 4.999368122911739e-06,
+      "loss": 0.6926,
+      "step": 261
+    },
+    {
+      "epoch": 0.06956060002654985,
+      "grad_norm": 0.2878773147526592,
+      "learning_rate": 4.9993602494657166e-06,
+      "loss": 0.6837,
+      "step": 262
+    },
+    {
+      "epoch": 0.06982609849993363,
+      "grad_norm": 0.28952392627424006,
+      "learning_rate": 4.999352327276049e-06,
+      "loss": 0.6871,
+      "step": 263
+    },
+    {
+      "epoch": 0.0700915969733174,
+      "grad_norm": 0.2780781868581628,
+      "learning_rate": 4.999344356342893e-06,
+      "loss": 0.6665,
+      "step": 264
+    },
+    {
+      "epoch": 0.07035709544670118,
+      "grad_norm": 0.2837049791196424,
+      "learning_rate": 4.999336336666402e-06,
+      "loss": 0.6594,
+      "step": 265
+    },
+    {
+      "epoch": 0.07062259392008496,
+      "grad_norm": 0.2821916104605583,
+      "learning_rate": 4.999328268246732e-06,
+      "loss": 0.7093,
+      "step": 266
+    },
+    {
+      "epoch": 0.07088809239346874,
+      "grad_norm": 0.2837950250863991,
+      "learning_rate": 4.999320151084043e-06,
+      "loss": 0.6961,
+      "step": 267
+    },
+    {
+      "epoch": 0.07115359086685251,
+      "grad_norm": 0.2832447193649773,
+      "learning_rate": 4.99931198517849e-06,
+      "loss": 0.703,
+      "step": 268
+    },
+    {
+      "epoch": 0.07141908934023629,
+      "grad_norm": 0.2812061141776504,
+      "learning_rate": 4.999303770530235e-06,
+      "loss": 0.6981,
+      "step": 269
+    },
+    {
+      "epoch": 0.07168458781362007,
+      "grad_norm": 0.26950781781942745,
+      "learning_rate": 4.9992955071394365e-06,
+      "loss": 0.6412,
+      "step": 270
+    },
+    {
+      "epoch": 0.07195008628700385,
+      "grad_norm": 0.28606325968475227,
+      "learning_rate": 4.999287195006257e-06,
+      "loss": 0.6802,
+      "step": 271
+    },
+    {
+      "epoch": 0.07221558476038763,
+      "grad_norm": 0.28217180124962293,
+      "learning_rate": 4.999278834130857e-06,
+      "loss": 0.6704,
+      "step": 272
+    },
+    {
+      "epoch": 0.0724810832337714,
+      "grad_norm": 0.2799545605253567,
+      "learning_rate": 4.999270424513402e-06,
+      "loss": 0.6739,
+      "step": 273
+    },
+    {
+      "epoch": 0.07274658170715519,
+      "grad_norm": 0.2948366125568834,
+      "learning_rate": 4.999261966154053e-06,
+      "loss": 0.6996,
+      "step": 274
+    },
+    {
+      "epoch": 0.07301208018053897,
+      "grad_norm": 0.28526141854662124,
+      "learning_rate": 4.999253459052977e-06,
+      "loss": 0.6786,
+      "step": 275
+    },
+    {
+      "epoch": 0.07327757865392275,
+      "grad_norm": 0.2851150240842429,
+      "learning_rate": 4.99924490321034e-06,
+      "loss": 0.6933,
+      "step": 276
+    },
+    {
+      "epoch": 0.07354307712730652,
+      "grad_norm": 0.28820089750565403,
+      "learning_rate": 4.999236298626307e-06,
+      "loss": 0.681,
+      "step": 277
+    },
+    {
+      "epoch": 0.07380857560069029,
+      "grad_norm": 0.2846131009168272,
+      "learning_rate": 4.999227645301047e-06,
+      "loss": 0.6672,
+      "step": 278
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 0.2928403434083448,
+      "learning_rate": 4.999218943234729e-06,
+      "loss": 0.6856,
+      "step": 279
+    },
+    {
+      "epoch": 0.07433957254745785,
+      "grad_norm": 0.29824152148219807,
+      "learning_rate": 4.999210192427523e-06,
+      "loss": 0.6768,
+      "step": 280
+    },
+    {
+      "epoch": 0.07460507102084163,
+      "grad_norm": 0.2840864286946336,
+      "learning_rate": 4.999201392879598e-06,
+      "loss": 0.6833,
+      "step": 281
+    },
+    {
+      "epoch": 0.07487056949422541,
+      "grad_norm": 0.28658971140766376,
+      "learning_rate": 4.999192544591128e-06,
+      "loss": 0.6767,
+      "step": 282
+    },
+    {
+      "epoch": 0.07513606796760919,
+      "grad_norm": 0.28735745839725857,
+      "learning_rate": 4.999183647562283e-06,
+      "loss": 0.6775,
+      "step": 283
+    },
+    {
+      "epoch": 0.07540156644099297,
+      "grad_norm": 0.28250300851086513,
+      "learning_rate": 4.999174701793239e-06,
+      "loss": 0.6529,
+      "step": 284
+    },
+    {
+      "epoch": 0.07566706491437675,
+      "grad_norm": 0.2872554066283219,
+      "learning_rate": 4.999165707284169e-06,
+      "loss": 0.6548,
+      "step": 285
+    },
+    {
+      "epoch": 0.07593256338776053,
+      "grad_norm": 0.29246682414100555,
+      "learning_rate": 4.999156664035248e-06,
+      "loss": 0.6973,
+      "step": 286
+    },
+    {
+      "epoch": 0.07619806186114429,
+      "grad_norm": 0.2900221113157214,
+      "learning_rate": 4.999147572046654e-06,
+      "loss": 0.6885,
+      "step": 287
+    },
+    {
+      "epoch": 0.07646356033452807,
+      "grad_norm": 0.2944016883960358,
+      "learning_rate": 4.9991384313185635e-06,
+      "loss": 0.6914,
+      "step": 288
+    },
+    {
+      "epoch": 0.07672905880791185,
+      "grad_norm": 0.2878252740936143,
+      "learning_rate": 4.9991292418511536e-06,
+      "loss": 0.6938,
+      "step": 289
+    },
+    {
+      "epoch": 0.07699455728129563,
+      "grad_norm": 0.2929662685483102,
+      "learning_rate": 4.999120003644606e-06,
+      "loss": 0.6996,
+      "step": 290
+    },
+    {
+      "epoch": 0.07726005575467941,
+      "grad_norm": 0.3000498782620618,
+      "learning_rate": 4.999110716699097e-06,
+      "loss": 0.6668,
+      "step": 291
+    },
+    {
+      "epoch": 0.07752555422806319,
+      "grad_norm": 0.2935909465153773,
+      "learning_rate": 4.999101381014813e-06,
+      "loss": 0.6538,
+      "step": 292
+    },
+    {
+      "epoch": 0.07779105270144697,
+      "grad_norm": 0.29460974819047786,
+      "learning_rate": 4.999091996591931e-06,
+      "loss": 0.6922,
+      "step": 293
+    },
+    {
+      "epoch": 0.07805655117483075,
+      "grad_norm": 0.2888230041157458,
+      "learning_rate": 4.999082563430637e-06,
+      "loss": 0.6873,
+      "step": 294
+    },
+    {
+      "epoch": 0.07832204964821453,
+      "grad_norm": 0.2946906069535359,
+      "learning_rate": 4.9990730815311144e-06,
+      "loss": 0.6367,
+      "step": 295
+    },
+    {
+      "epoch": 0.0785875481215983,
+      "grad_norm": 0.29255274301221523,
+      "learning_rate": 4.999063550893548e-06,
+      "loss": 0.6142,
+      "step": 296
+    },
+    {
+      "epoch": 0.07885304659498207,
+      "grad_norm": 0.2932435372434208,
+      "learning_rate": 4.999053971518123e-06,
+      "loss": 0.6708,
+      "step": 297
+    },
+    {
+      "epoch": 0.07911854506836585,
+      "grad_norm": 0.29885523057090346,
+      "learning_rate": 4.999044343405027e-06,
+      "loss": 0.6917,
+      "step": 298
+    },
+    {
+      "epoch": 0.07938404354174963,
+      "grad_norm": 0.295942948386189,
+      "learning_rate": 4.999034666554448e-06,
+      "loss": 0.6771,
+      "step": 299
+    },
+    {
+      "epoch": 0.07964954201513341,
+      "grad_norm": 0.3141460477901498,
+      "learning_rate": 4.999024940966575e-06,
+      "loss": 0.6724,
+      "step": 300
+    },
+    {
+      "epoch": 0.07991504048851719,
+      "grad_norm": 0.29660845458593904,
+      "learning_rate": 4.999015166641595e-06,
+      "loss": 0.6973,
+      "step": 301
+    },
+    {
+      "epoch": 0.08018053896190097,
+      "grad_norm": 0.2970126436601212,
+      "learning_rate": 4.999005343579703e-06,
+      "loss": 0.6648,
+      "step": 302
+    },
+    {
+      "epoch": 0.08044603743528475,
+      "grad_norm": 0.29744317293632544,
+      "learning_rate": 4.9989954717810865e-06,
+      "loss": 0.6891,
+      "step": 303
+    },
+    {
+      "epoch": 0.08071153590866853,
+      "grad_norm": 0.2901410497684151,
+      "learning_rate": 4.99898555124594e-06,
+      "loss": 0.6735,
+      "step": 304
+    },
+    {
+      "epoch": 0.08097703438205231,
+      "grad_norm": 0.29798459544629646,
+      "learning_rate": 4.9989755819744565e-06,
+      "loss": 0.6539,
+      "step": 305
+    },
+    {
+      "epoch": 0.08124253285543608,
+      "grad_norm": 0.2929904256856655,
+      "learning_rate": 4.998965563966831e-06,
+      "loss": 0.6562,
+      "step": 306
+    },
+    {
+      "epoch": 0.08150803132881985,
+      "grad_norm": 0.2937270688048953,
+      "learning_rate": 4.998955497223258e-06,
+      "loss": 0.663,
+      "step": 307
+    },
+    {
+      "epoch": 0.08177352980220363,
+      "grad_norm": 0.2931852456506224,
+      "learning_rate": 4.9989453817439345e-06,
+      "loss": 0.6823,
+      "step": 308
+    },
+    {
+      "epoch": 0.08203902827558741,
+      "grad_norm": 0.2912478762560809,
+      "learning_rate": 4.998935217529058e-06,
+      "loss": 0.6713,
+      "step": 309
+    },
+    {
+      "epoch": 0.0823045267489712,
+      "grad_norm": 0.30838628503323395,
+      "learning_rate": 4.998925004578826e-06,
+      "loss": 0.6922,
+      "step": 310
+    },
+    {
+      "epoch": 0.08257002522235497,
+      "grad_norm": 0.2967990541380092,
+      "learning_rate": 4.9989147428934385e-06,
+      "loss": 0.6448,
+      "step": 311
+    },
+    {
+      "epoch": 0.08283552369573875,
+      "grad_norm": 0.2953837409606851,
+      "learning_rate": 4.998904432473095e-06,
+      "loss": 0.6708,
+      "step": 312
+    },
+    {
+      "epoch": 0.08310102216912253,
+      "grad_norm": 0.29801587193664514,
+      "learning_rate": 4.998894073317996e-06,
+      "loss": 0.6945,
+      "step": 313
+    },
+    {
+      "epoch": 0.08336652064250631,
+      "grad_norm": 0.29225932313117575,
+      "learning_rate": 4.998883665428345e-06,
+      "loss": 0.6492,
+      "step": 314
+    },
+    {
+      "epoch": 0.08363201911589008,
+      "grad_norm": 0.29959881567048147,
+      "learning_rate": 4.998873208804343e-06,
+      "loss": 0.6806,
+      "step": 315
+    },
+    {
+      "epoch": 0.08389751758927386,
+      "grad_norm": 0.29960446019233394,
+      "learning_rate": 4.998862703446196e-06,
+      "loss": 0.6573,
+      "step": 316
+    },
+    {
+      "epoch": 0.08416301606265764,
+      "grad_norm": 0.2946962090420051,
+      "learning_rate": 4.998852149354108e-06,
+      "loss": 0.656,
+      "step": 317
+    },
+    {
+      "epoch": 0.08442851453604142,
+      "grad_norm": 0.307819842583063,
+      "learning_rate": 4.998841546528284e-06,
+      "loss": 0.664,
+      "step": 318
+    },
+    {
+      "epoch": 0.0846940130094252,
+      "grad_norm": 0.2940226762339422,
+      "learning_rate": 4.998830894968934e-06,
+      "loss": 0.6504,
+      "step": 319
+    },
+    {
+      "epoch": 0.08495951148280897,
+      "grad_norm": 0.3010880006100906,
+      "learning_rate": 4.9988201946762615e-06,
+      "loss": 0.6499,
+      "step": 320
+    },
+    {
+      "epoch": 0.08522500995619275,
+      "grad_norm": 0.29970401482842796,
+      "learning_rate": 4.998809445650476e-06,
+      "loss": 0.678,
+      "step": 321
+    },
+    {
+      "epoch": 0.08549050842957653,
+      "grad_norm": 0.3121225152546364,
+      "learning_rate": 4.99879864789179e-06,
+      "loss": 0.6821,
+      "step": 322
+    },
+    {
+      "epoch": 0.08575600690296031,
+      "grad_norm": 0.3040775724714788,
+      "learning_rate": 4.998787801400412e-06,
+      "loss": 0.6874,
+      "step": 323
+    },
+    {
+      "epoch": 0.08602150537634409,
+      "grad_norm": 0.2974808855776655,
+      "learning_rate": 4.998776906176553e-06,
+      "loss": 0.7015,
+      "step": 324
+    },
+    {
+      "epoch": 0.08628700384972786,
+      "grad_norm": 0.29641748546634916,
+      "learning_rate": 4.998765962220428e-06,
+      "loss": 0.6909,
+      "step": 325
+    },
+    {
+      "epoch": 0.08655250232311164,
+      "grad_norm": 0.28834263017403083,
+      "learning_rate": 4.998754969532247e-06,
+      "loss": 0.6707,
+      "step": 326
+    },
+    {
+      "epoch": 0.08681800079649542,
+      "grad_norm": 0.30874417666950094,
+      "learning_rate": 4.998743928112226e-06,
+      "loss": 0.6893,
+      "step": 327
+    },
+    {
+      "epoch": 0.0870834992698792,
+      "grad_norm": 0.3023373305283153,
+      "learning_rate": 4.998732837960582e-06,
+      "loss": 0.7246,
+      "step": 328
+    },
+    {
+      "epoch": 0.08734899774326298,
+      "grad_norm": 0.2986985311476253,
+      "learning_rate": 4.998721699077528e-06,
+      "loss": 0.6878,
+      "step": 329
+    },
+    {
+      "epoch": 0.08761449621664676,
+      "grad_norm": 0.30676169244288776,
+      "learning_rate": 4.998710511463284e-06,
+      "loss": 0.6941,
+      "step": 330
+    },
+    {
+      "epoch": 0.08787999469003054,
+      "grad_norm": 0.30787895260330744,
+      "learning_rate": 4.998699275118066e-06,
+      "loss": 0.6797,
+      "step": 331
+    },
+    {
+      "epoch": 0.08814549316341432,
+      "grad_norm": 0.2923955549513281,
+      "learning_rate": 4.998687990042096e-06,
+      "loss": 0.6725,
+      "step": 332
+    },
+    {
+      "epoch": 0.0884109916367981,
+      "grad_norm": 0.29892236532536787,
+      "learning_rate": 4.998676656235591e-06,
+      "loss": 0.66,
+      "step": 333
+    },
+    {
+      "epoch": 0.08867649011018186,
+      "grad_norm": 0.29981490267370287,
+      "learning_rate": 4.998665273698774e-06,
+      "loss": 0.6555,
+      "step": 334
+    },
+    {
+      "epoch": 0.08894198858356564,
+      "grad_norm": 0.2937419301980639,
+      "learning_rate": 4.998653842431865e-06,
+      "loss": 0.6445,
+      "step": 335
+    },
+    {
+      "epoch": 0.08920748705694942,
+      "grad_norm": 0.2900337493308749,
+      "learning_rate": 4.99864236243509e-06,
+      "loss": 0.6574,
+      "step": 336
+    },
+    {
+      "epoch": 0.0894729855303332,
+      "grad_norm": 0.29057820345685953,
+      "learning_rate": 4.998630833708671e-06,
+      "loss": 0.6315,
+      "step": 337
+    },
+    {
+      "epoch": 0.08973848400371698,
+      "grad_norm": 0.29714366006288034,
+      "learning_rate": 4.998619256252832e-06,
+      "loss": 0.6929,
+      "step": 338
+    },
+    {
+      "epoch": 0.09000398247710076,
+      "grad_norm": 0.2991233791943958,
+      "learning_rate": 4.998607630067801e-06,
+      "loss": 0.6658,
+      "step": 339
+    },
+    {
+      "epoch": 0.09026948095048454,
+      "grad_norm": 0.3024435971160214,
+      "learning_rate": 4.998595955153803e-06,
+      "loss": 0.674,
+      "step": 340
+    },
+    {
+      "epoch": 0.09053497942386832,
+      "grad_norm": 0.28226917416513464,
+      "learning_rate": 4.998584231511067e-06,
+      "loss": 0.6463,
+      "step": 341
+    },
+    {
+      "epoch": 0.0908004778972521,
+      "grad_norm": 0.3001637870421723,
+      "learning_rate": 4.99857245913982e-06,
+      "loss": 0.692,
+      "step": 342
+    },
+    {
+      "epoch": 0.09106597637063586,
+      "grad_norm": 0.29754947156555495,
+      "learning_rate": 4.998560638040292e-06,
+      "loss": 0.6416,
+      "step": 343
+    },
+    {
+      "epoch": 0.09133147484401964,
+      "grad_norm": 0.30565826213510777,
+      "learning_rate": 4.998548768212716e-06,
+      "loss": 0.6686,
+      "step": 344
+    },
+    {
+      "epoch": 0.09159697331740342,
+      "grad_norm": 0.2998450329350609,
+      "learning_rate": 4.998536849657319e-06,
+      "loss": 0.6331,
+      "step": 345
+    },
+    {
+      "epoch": 0.0918624717907872,
+      "grad_norm": 0.3048929107122252,
+      "learning_rate": 4.998524882374338e-06,
+      "loss": 0.6475,
+      "step": 346
+    },
+    {
+      "epoch": 0.09212797026417098,
+      "grad_norm": 0.30975926255815844,
+      "learning_rate": 4.9985128663640035e-06,
+      "loss": 0.6633,
+      "step": 347
+    },
+    {
+      "epoch": 0.09239346873755476,
+      "grad_norm": 0.3168439032314714,
+      "learning_rate": 4.99850080162655e-06,
+      "loss": 0.6846,
+      "step": 348
+    },
+    {
+      "epoch": 0.09265896721093854,
+      "grad_norm": 0.29738742881475877,
+      "learning_rate": 4.9984886881622146e-06,
+      "loss": 0.6838,
+      "step": 349
+    },
+    {
+      "epoch": 0.09292446568432232,
+      "grad_norm": 0.2963089371897067,
+      "learning_rate": 4.998476525971232e-06,
+      "loss": 0.6365,
+      "step": 350
+    },
+    {
+      "epoch": 0.0931899641577061,
+      "grad_norm": 0.3120127203908184,
+      "learning_rate": 4.998464315053841e-06,
+      "loss": 0.6847,
+      "step": 351
+    },
+    {
+      "epoch": 0.09345546263108988,
+      "grad_norm": 0.30644369963379364,
+      "learning_rate": 4.998452055410276e-06,
+      "loss": 0.6804,
+      "step": 352
+    },
+    {
+      "epoch": 0.09372096110447364,
+      "grad_norm": 0.3166463471381308,
+      "learning_rate": 4.998439747040781e-06,
+      "loss": 0.6245,
+      "step": 353
+    },
+    {
+      "epoch": 0.09398645957785742,
+      "grad_norm": 0.30056253488107737,
+      "learning_rate": 4.9984273899455924e-06,
+      "loss": 0.6619,
+      "step": 354
+    },
+    {
+      "epoch": 0.0942519580512412,
+      "grad_norm": 0.29551938579257553,
+      "learning_rate": 4.998414984124953e-06,
+      "loss": 0.6793,
+      "step": 355
+    },
+    {
+      "epoch": 0.09451745652462498,
+      "grad_norm": 0.30577293764977503,
+      "learning_rate": 4.998402529579104e-06,
+      "loss": 0.685,
+      "step": 356
+    },
+    {
+      "epoch": 0.09478295499800876,
+      "grad_norm": 0.3136027154810347,
+      "learning_rate": 4.998390026308288e-06,
+      "loss": 0.6522,
+      "step": 357
+    },
+    {
+      "epoch": 0.09504845347139254,
+      "grad_norm": 0.3097738609974774,
+      "learning_rate": 4.998377474312751e-06,
+      "loss": 0.6835,
+      "step": 358
+    },
+    {
+      "epoch": 0.09531395194477632,
+      "grad_norm": 0.29881152239543424,
+      "learning_rate": 4.998364873592734e-06,
+      "loss": 0.6532,
+      "step": 359
+    },
+    {
+      "epoch": 0.0955794504181601,
+      "grad_norm": 0.30803760769027594,
+      "learning_rate": 4.998352224148487e-06,
+      "loss": 0.7106,
+      "step": 360
+    },
+    {
+      "epoch": 0.09584494889154388,
+      "grad_norm": 0.3146232150958707,
+      "learning_rate": 4.998339525980254e-06,
+      "loss": 0.7225,
+      "step": 361
+    },
+    {
+      "epoch": 0.09611044736492765,
+      "grad_norm": 0.30707379260872547,
+      "learning_rate": 4.998326779088283e-06,
+      "loss": 0.6821,
+      "step": 362
+    },
+    {
+      "epoch": 0.09637594583831142,
+      "grad_norm": 0.308760345968221,
+      "learning_rate": 4.998313983472823e-06,
+      "loss": 0.6844,
+      "step": 363
+    },
+    {
+      "epoch": 0.0966414443116952,
+      "grad_norm": 0.3042905996368259,
+      "learning_rate": 4.998301139134123e-06,
+      "loss": 0.6572,
+      "step": 364
+    },
+    {
+      "epoch": 0.09690694278507898,
+      "grad_norm": 0.30086485484817665,
+      "learning_rate": 4.998288246072434e-06,
+      "loss": 0.6521,
+      "step": 365
+    },
+    {
+      "epoch": 0.09717244125846276,
+      "grad_norm": 0.30234981928885984,
+      "learning_rate": 4.9982753042880076e-06,
+      "loss": 0.684,
+      "step": 366
+    },
+    {
+      "epoch": 0.09743793973184654,
+      "grad_norm": 0.30072324839807785,
+      "learning_rate": 4.998262313781096e-06,
+      "loss": 0.6916,
+      "step": 367
+    },
+    {
+      "epoch": 0.09770343820523032,
+      "grad_norm": 0.3078805070343228,
+      "learning_rate": 4.998249274551952e-06,
+      "loss": 0.6932,
+      "step": 368
+    },
+    {
+      "epoch": 0.0979689366786141,
+      "grad_norm": 0.3106113547728943,
+      "learning_rate": 4.9982361866008306e-06,
+      "loss": 0.6991,
+      "step": 369
+    },
+    {
+      "epoch": 0.09823443515199788,
+      "grad_norm": 0.3069038054416314,
+      "learning_rate": 4.998223049927987e-06,
+      "loss": 0.7023,
+      "step": 370
+    },
+    {
+      "epoch": 0.09849993362538166,
+      "grad_norm": 0.30210262430895374,
+      "learning_rate": 4.998209864533677e-06,
+      "loss": 0.6651,
+      "step": 371
+    },
+    {
+      "epoch": 0.09876543209876543,
+      "grad_norm": 0.31127245322981645,
+      "learning_rate": 4.998196630418159e-06,
+      "loss": 0.6575,
+      "step": 372
+    },
+    {
+      "epoch": 0.0990309305721492,
+      "grad_norm": 0.30865037988370164,
+      "learning_rate": 4.9981833475816886e-06,
+      "loss": 0.6797,
+      "step": 373
+    },
+    {
+      "epoch": 0.09929642904553299,
+      "grad_norm": 0.3056326792569424,
+      "learning_rate": 4.998170016024526e-06,
+      "loss": 0.6576,
+      "step": 374
+    },
+    {
+      "epoch": 0.09956192751891677,
+      "grad_norm": 0.30830297591571204,
+      "learning_rate": 4.9981566357469325e-06,
+      "loss": 0.6684,
+      "step": 375
+    },
+    {
+      "epoch": 0.09982742599230054,
+      "grad_norm": 0.31003718799398505,
+      "learning_rate": 4.998143206749168e-06,
+      "loss": 0.6819,
+      "step": 376
+    },
+    {
+      "epoch": 0.10009292446568432,
+      "grad_norm": 0.3104855221548017,
+      "learning_rate": 4.998129729031495e-06,
+      "loss": 0.6804,
+      "step": 377
+    },
+    {
+      "epoch": 0.1003584229390681,
+      "grad_norm": 0.3079010557016698,
+      "learning_rate": 4.998116202594175e-06,
+      "loss": 0.6551,
+      "step": 378
+    },
+    {
+      "epoch": 0.10062392141245188,
+      "grad_norm": 0.2990541945875933,
+      "learning_rate": 4.998102627437472e-06,
+      "loss": 0.6565,
+      "step": 379
+    },
+    {
+      "epoch": 0.10088941988583566,
+      "grad_norm": 0.31555616099171835,
+      "learning_rate": 4.998089003561653e-06,
+      "loss": 0.6869,
+      "step": 380
+    },
+    {
+      "epoch": 0.10115491835921943,
+      "grad_norm": 0.2949840548334904,
+      "learning_rate": 4.99807533096698e-06,
+      "loss": 0.6668,
+      "step": 381
+    },
+    {
+      "epoch": 0.10142041683260321,
+      "grad_norm": 0.3158090992458187,
+      "learning_rate": 4.9980616096537235e-06,
+      "loss": 0.6974,
+      "step": 382
+    },
+    {
+      "epoch": 0.10168591530598699,
+      "grad_norm": 0.30802755793824466,
+      "learning_rate": 4.998047839622149e-06,
+      "loss": 0.6868,
+      "step": 383
+    },
+    {
+      "epoch": 0.10195141377937077,
+      "grad_norm": 0.30873942546894295,
+      "learning_rate": 4.998034020872526e-06,
+      "loss": 0.6974,
+      "step": 384
+    },
+    {
+      "epoch": 0.10221691225275455,
+      "grad_norm": 0.31268243841582133,
+      "learning_rate": 4.9980201534051214e-06,
+      "loss": 0.7004,
+      "step": 385
+    },
+    {
+      "epoch": 0.10248241072613833,
+      "grad_norm": 0.30737025093264747,
+      "learning_rate": 4.998006237220209e-06,
+      "loss": 0.6492,
+      "step": 386
+    },
+    {
+      "epoch": 0.1027479091995221,
+      "grad_norm": 0.3067303152805426,
+      "learning_rate": 4.997992272318058e-06,
+      "loss": 0.6869,
+      "step": 387
+    },
+    {
+      "epoch": 0.10301340767290589,
+      "grad_norm": 0.3125290276549197,
+      "learning_rate": 4.997978258698942e-06,
+      "loss": 0.6474,
+      "step": 388
+    },
+    {
+      "epoch": 0.10327890614628966,
+      "grad_norm": 0.30298746561845363,
+      "learning_rate": 4.9979641963631344e-06,
+      "loss": 0.6521,
+      "step": 389
+    },
+    {
+      "epoch": 0.10354440461967343,
+      "grad_norm": 0.3185722182882203,
+      "learning_rate": 4.997950085310907e-06,
+      "loss": 0.7171,
+      "step": 390
+    },
+    {
+      "epoch": 0.10380990309305721,
+      "grad_norm": 0.3056792173121192,
+      "learning_rate": 4.997935925542539e-06,
+      "loss": 0.663,
+      "step": 391
+    },
+    {
+      "epoch": 0.10407540156644099,
+      "grad_norm": 0.3162449557105779,
+      "learning_rate": 4.997921717058304e-06,
+      "loss": 0.684,
+      "step": 392
+    },
+    {
+      "epoch": 0.10434090003982477,
+      "grad_norm": 0.3114840922452039,
+      "learning_rate": 4.997907459858479e-06,
+      "loss": 0.6651,
+      "step": 393
+    },
+    {
+      "epoch": 0.10460639851320855,
+      "grad_norm": 0.3160506470893945,
+      "learning_rate": 4.997893153943343e-06,
+      "loss": 0.7194,
+      "step": 394
+    },
+    {
+      "epoch": 0.10487189698659233,
+      "grad_norm": 0.3106758543108377,
+      "learning_rate": 4.997878799313174e-06,
+      "loss": 0.6599,
+      "step": 395
+    },
+    {
+      "epoch": 0.10513739545997611,
+      "grad_norm": 0.3086954997584318,
+      "learning_rate": 4.997864395968252e-06,
+      "loss": 0.6911,
+      "step": 396
+    },
+    {
+      "epoch": 0.10540289393335989,
+      "grad_norm": 0.3114318196879341,
+      "learning_rate": 4.99784994390886e-06,
+      "loss": 0.6593,
+      "step": 397
+    },
+    {
+      "epoch": 0.10566839240674367,
+      "grad_norm": 0.31406721651161573,
+      "learning_rate": 4.9978354431352785e-06,
+      "loss": 0.6338,
+      "step": 398
+    },
+    {
+      "epoch": 0.10593389088012745,
+      "grad_norm": 0.3177758176653506,
+      "learning_rate": 4.997820893647789e-06,
+      "loss": 0.7053,
+      "step": 399
+    },
+    {
+      "epoch": 0.10619938935351121,
+      "grad_norm": 0.3027956837248792,
+      "learning_rate": 4.997806295446676e-06,
+      "loss": 0.6804,
+      "step": 400
+    },
+    {
+      "epoch": 0.10646488782689499,
+      "grad_norm": 0.3020081713951576,
+      "learning_rate": 4.9977916485322255e-06,
+      "loss": 0.6412,
+      "step": 401
+    },
+    {
+      "epoch": 0.10673038630027877,
+      "grad_norm": 0.32617568438218675,
+      "learning_rate": 4.997776952904723e-06,
+      "loss": 0.6756,
+      "step": 402
+    },
+    {
+      "epoch": 0.10699588477366255,
+      "grad_norm": 0.30597090699286184,
+      "learning_rate": 4.9977622085644525e-06,
+      "loss": 0.6723,
+      "step": 403
+    },
+    {
+      "epoch": 0.10726138324704633,
+      "grad_norm": 0.3144766515591238,
+      "learning_rate": 4.997747415511705e-06,
+      "loss": 0.6653,
+      "step": 404
+    },
+    {
+      "epoch": 0.10752688172043011,
+      "grad_norm": 0.31378962465347404,
+      "learning_rate": 4.997732573746765e-06,
+      "loss": 0.6941,
+      "step": 405
+    },
+    {
+      "epoch": 0.10779238019381389,
+      "grad_norm": 0.31097367436263995,
+      "learning_rate": 4.997717683269926e-06,
+      "loss": 0.624,
+      "step": 406
+    },
+    {
+      "epoch": 0.10805787866719767,
+      "grad_norm": 0.31114714417062406,
+      "learning_rate": 4.997702744081477e-06,
+      "loss": 0.6889,
+      "step": 407
+    },
+    {
+      "epoch": 0.10832337714058145,
+      "grad_norm": 0.3278466064162668,
+      "learning_rate": 4.9976877561817086e-06,
+      "loss": 0.6977,
+      "step": 408
+    },
+    {
+      "epoch": 0.10858887561396521,
+      "grad_norm": 0.31784295440346955,
+      "learning_rate": 4.997672719570913e-06,
+      "loss": 0.6588,
+      "step": 409
+    },
+    {
+      "epoch": 0.10885437408734899,
+      "grad_norm": 0.31856573937167837,
+      "learning_rate": 4.997657634249384e-06,
+      "loss": 0.712,
+      "step": 410
+    },
+    {
+      "epoch": 0.10911987256073277,
+      "grad_norm": 0.3086321126823493,
+      "learning_rate": 4.997642500217417e-06,
+      "loss": 0.694,
+      "step": 411
+    },
+    {
+      "epoch": 0.10938537103411655,
+      "grad_norm": 0.3233180475823364,
+      "learning_rate": 4.997627317475305e-06,
+      "loss": 0.6972,
+      "step": 412
+    },
+    {
+      "epoch": 0.10965086950750033,
+      "grad_norm": 0.31087689547222547,
+      "learning_rate": 4.997612086023345e-06,
+      "loss": 0.653,
+      "step": 413
+    },
+    {
+      "epoch": 0.10991636798088411,
+      "grad_norm": 0.3176995483680813,
+      "learning_rate": 4.997596805861835e-06,
+      "loss": 0.6807,
+      "step": 414
+    },
+    {
+      "epoch": 0.11018186645426789,
+      "grad_norm": 0.31386266013051656,
+      "learning_rate": 4.997581476991071e-06,
+      "loss": 0.6618,
+      "step": 415
+    },
+    {
+      "epoch": 0.11044736492765167,
+      "grad_norm": 0.3152387993743806,
+      "learning_rate": 4.997566099411354e-06,
+      "loss": 0.686,
+      "step": 416
+    },
+    {
+      "epoch": 0.11071286340103545,
+      "grad_norm": 0.33351140546506874,
+      "learning_rate": 4.997550673122983e-06,
+      "loss": 0.6772,
+      "step": 417
+    },
+    {
+      "epoch": 0.11097836187441922,
+      "grad_norm": 0.3186827810679576,
+      "learning_rate": 4.997535198126258e-06,
+      "loss": 0.662,
+      "step": 418
+    },
+    {
+      "epoch": 0.111243860347803,
+      "grad_norm": 0.31009634384997614,
+      "learning_rate": 4.997519674421483e-06,
+      "loss": 0.6655,
+      "step": 419
+    },
+    {
+      "epoch": 0.11150935882118677,
+      "grad_norm": 0.3268830658834569,
+      "learning_rate": 4.997504102008957e-06,
+      "loss": 0.693,
+      "step": 420
+    },
+    {
+      "epoch": 0.11177485729457055,
+      "grad_norm": 0.30954548187647873,
+      "learning_rate": 4.997488480888988e-06,
+      "loss": 0.671,
+      "step": 421
+    },
+    {
+      "epoch": 0.11204035576795433,
+      "grad_norm": 0.30894394142603065,
+      "learning_rate": 4.997472811061879e-06,
+      "loss": 0.6645,
+      "step": 422
+    },
+    {
+      "epoch": 0.11230585424133811,
+      "grad_norm": 0.31069961094923226,
+      "learning_rate": 4.997457092527934e-06,
+      "loss": 0.6619,
+      "step": 423
+    },
+    {
+      "epoch": 0.11257135271472189,
+      "grad_norm": 0.31267417835999584,
+      "learning_rate": 4.997441325287461e-06,
+      "loss": 0.6625,
+      "step": 424
+    },
+    {
+      "epoch": 0.11283685118810567,
+      "grad_norm": 0.31286268667247885,
+      "learning_rate": 4.997425509340767e-06,
+      "loss": 0.6898,
+      "step": 425
+    },
+    {
+      "epoch": 0.11310234966148945,
+      "grad_norm": 0.3133878370932067,
+      "learning_rate": 4.997409644688162e-06,
+      "loss": 0.6498,
+      "step": 426
+    },
+    {
+      "epoch": 0.11336784813487323,
+      "grad_norm": 0.30696063627302694,
+      "learning_rate": 4.997393731329954e-06,
+      "loss": 0.6828,
+      "step": 427
+    },
+    {
+      "epoch": 0.113633346608257,
+      "grad_norm": 0.30531189678202303,
+      "learning_rate": 4.997377769266454e-06,
+      "loss": 0.6434,
+      "step": 428
+    },
+    {
+      "epoch": 0.11389884508164078,
+      "grad_norm": 0.30062330676572224,
+      "learning_rate": 4.997361758497972e-06,
+      "loss": 0.6208,
+      "step": 429
+    },
+    {
+      "epoch": 0.11416434355502456,
+      "grad_norm": 0.2946536636599198,
+      "learning_rate": 4.997345699024821e-06,
+      "loss": 0.6227,
+      "step": 430
+    },
+    {
+      "epoch": 0.11442984202840834,
+      "grad_norm": 0.3069850692030948,
+      "learning_rate": 4.9973295908473146e-06,
+      "loss": 0.6576,
+      "step": 431
+    },
+    {
+      "epoch": 0.11469534050179211,
+      "grad_norm": 0.31148890419045955,
+      "learning_rate": 4.997313433965767e-06,
+      "loss": 0.6678,
+      "step": 432
+    },
+    {
+      "epoch": 0.1149608389751759,
+      "grad_norm": 0.324529924104944,
+      "learning_rate": 4.997297228380492e-06,
+      "loss": 0.7094,
+      "step": 433
+    },
+    {
+      "epoch": 0.11522633744855967,
+      "grad_norm": 0.3155063789449674,
+      "learning_rate": 4.997280974091808e-06,
+      "loss": 0.6392,
+      "step": 434
+    },
+    {
+      "epoch": 0.11549183592194345,
+      "grad_norm": 0.3120980596384352,
+      "learning_rate": 4.997264671100029e-06,
+      "loss": 0.6454,
+      "step": 435
+    },
+    {
+      "epoch": 0.11575733439532723,
+      "grad_norm": 0.31139341298472606,
+      "learning_rate": 4.997248319405475e-06,
+      "loss": 0.6834,
+      "step": 436
+    },
+    {
+      "epoch": 0.116022832868711,
+      "grad_norm": 0.3220778362463673,
+      "learning_rate": 4.997231919008466e-06,
+      "loss": 0.6931,
+      "step": 437
+    },
+    {
+      "epoch": 0.11628833134209478,
+      "grad_norm": 0.3140339484239711,
+      "learning_rate": 4.997215469909319e-06,
+      "loss": 0.6612,
+      "step": 438
+    },
+    {
+      "epoch": 0.11655382981547856,
+      "grad_norm": 0.3130511631833597,
+      "learning_rate": 4.997198972108356e-06,
+      "loss": 0.6604,
+      "step": 439
+    },
+    {
+      "epoch": 0.11681932828886234,
+      "grad_norm": 0.31838552167395007,
+      "learning_rate": 4.997182425605899e-06,
+      "loss": 0.6819,
+      "step": 440
+    },
+    {
+      "epoch": 0.11708482676224612,
+      "grad_norm": 0.3231736062338604,
+      "learning_rate": 4.9971658304022705e-06,
+      "loss": 0.7176,
+      "step": 441
+    },
+    {
+      "epoch": 0.1173503252356299,
+      "grad_norm": 0.31796270168634794,
+      "learning_rate": 4.997149186497795e-06,
+      "loss": 0.6435,
+      "step": 442
+    },
+    {
+      "epoch": 0.11761582370901368,
+      "grad_norm": 0.31256477602255317,
+      "learning_rate": 4.997132493892796e-06,
+      "loss": 0.6753,
+      "step": 443
+    },
+    {
+      "epoch": 0.11788132218239745,
+      "grad_norm": 0.3187404644494588,
+      "learning_rate": 4.997115752587598e-06,
+      "loss": 0.6952,
+      "step": 444
+    },
+    {
+      "epoch": 0.11814682065578123,
+      "grad_norm": 0.3297731732543038,
+      "learning_rate": 4.99709896258253e-06,
+      "loss": 0.6911,
+      "step": 445
+    },
+    {
+      "epoch": 0.11841231912916501,
+      "grad_norm": 0.31247350710985655,
+      "learning_rate": 4.997082123877918e-06,
+      "loss": 0.6803,
+      "step": 446
+    },
+    {
+      "epoch": 0.11867781760254878,
+      "grad_norm": 0.3297760667742294,
+      "learning_rate": 4.997065236474092e-06,
+      "loss": 0.6922,
+      "step": 447
+    },
+    {
+      "epoch": 0.11894331607593256,
+      "grad_norm": 0.30422819180050326,
+      "learning_rate": 4.997048300371377e-06,
+      "loss": 0.6466,
+      "step": 448
+    },
+    {
+      "epoch": 0.11920881454931634,
+      "grad_norm": 0.3199451761919143,
+      "learning_rate": 4.997031315570109e-06,
+      "loss": 0.6569,
+      "step": 449
+    },
+    {
+      "epoch": 0.11947431302270012,
+      "grad_norm": 0.32478782472338646,
+      "learning_rate": 4.997014282070615e-06,
+      "loss": 0.6674,
+      "step": 450
+    },
+    {
+      "epoch": 0.1197398114960839,
+      "grad_norm": 0.3316660848846396,
+      "learning_rate": 4.9969971998732294e-06,
+      "loss": 0.7,
+      "step": 451
+    },
+    {
+      "epoch": 0.12000530996946768,
+      "grad_norm": 0.31690024112628196,
+      "learning_rate": 4.996980068978285e-06,
+      "loss": 0.6643,
+      "step": 452
+    },
+    {
+      "epoch": 0.12027080844285146,
+      "grad_norm": 0.3267511446290268,
+      "learning_rate": 4.996962889386116e-06,
+      "loss": 0.6668,
+      "step": 453
+    },
+    {
+      "epoch": 0.12053630691623524,
+      "grad_norm": 0.31948431087726,
+      "learning_rate": 4.9969456610970555e-06,
+      "loss": 0.6709,
+      "step": 454
+    },
+    {
+      "epoch": 0.12080180538961902,
+      "grad_norm": 0.32208238773189896,
+      "learning_rate": 4.996928384111442e-06,
+      "loss": 0.6874,
+      "step": 455
+    },
+    {
+      "epoch": 0.12106730386300278,
+      "grad_norm": 0.3148108500947568,
+      "learning_rate": 4.996911058429611e-06,
+      "loss": 0.6862,
+      "step": 456
+    },
+    {
+      "epoch": 0.12133280233638656,
+      "grad_norm": 0.3119940304171294,
+      "learning_rate": 4.996893684051901e-06,
+      "loss": 0.6406,
+      "step": 457
+    },
+    {
+      "epoch": 0.12159830080977034,
+      "grad_norm": 0.3232865917452015,
+      "learning_rate": 4.996876260978651e-06,
+      "loss": 0.7013,
+      "step": 458
+    },
+    {
+      "epoch": 0.12186379928315412,
+      "grad_norm": 0.3252102414886518,
+      "learning_rate": 4.9968587892102e-06,
+      "loss": 0.6559,
+      "step": 459
+    },
+    {
+      "epoch": 0.1221292977565379,
+      "grad_norm": 0.3269747897933381,
+      "learning_rate": 4.99684126874689e-06,
+      "loss": 0.6155,
+      "step": 460
+    },
+    {
+      "epoch": 0.12239479622992168,
+      "grad_norm": 0.3179398124119427,
+      "learning_rate": 4.996823699589062e-06,
+      "loss": 0.6369,
+      "step": 461
+    },
+    {
+      "epoch": 0.12266029470330546,
+      "grad_norm": 0.3280930124454813,
+      "learning_rate": 4.996806081737058e-06,
+      "loss": 0.7001,
+      "step": 462
+    },
+    {
+      "epoch": 0.12292579317668924,
+      "grad_norm": 0.3370923358901789,
+      "learning_rate": 4.996788415191223e-06,
+      "loss": 0.6858,
+      "step": 463
+    },
+    {
+      "epoch": 0.12319129165007302,
+      "grad_norm": 0.3295220353946801,
+      "learning_rate": 4.9967706999519e-06,
+      "loss": 0.6602,
+      "step": 464
+    },
+    {
+      "epoch": 0.12345679012345678,
+      "grad_norm": 0.3152774696864616,
+      "learning_rate": 4.996752936019435e-06,
+      "loss": 0.6837,
+      "step": 465
+    },
+    {
+      "epoch": 0.12372228859684056,
+      "grad_norm": 0.3239072427931697,
+      "learning_rate": 4.996735123394175e-06,
+      "loss": 0.7113,
+      "step": 466
+    },
+    {
+      "epoch": 0.12398778707022434,
+      "grad_norm": 0.31285507970163806,
+      "learning_rate": 4.996717262076467e-06,
+      "loss": 0.6255,
+      "step": 467
+    },
+    {
+      "epoch": 0.12425328554360812,
+      "grad_norm": 0.3425777000470296,
+      "learning_rate": 4.996699352066659e-06,
+      "loss": 0.6906,
+      "step": 468
+    },
+    {
+      "epoch": 0.1245187840169919,
+      "grad_norm": 0.3226226899231145,
+      "learning_rate": 4.9966813933651005e-06,
+      "loss": 0.6753,
+      "step": 469
+    },
+    {
+      "epoch": 0.12478428249037568,
+      "grad_norm": 0.31435598476016197,
+      "learning_rate": 4.996663385972143e-06,
+      "loss": 0.6144,
+      "step": 470
+    },
+    {
+      "epoch": 0.12504978096375946,
+      "grad_norm": 0.3193846061113134,
+      "learning_rate": 4.996645329888135e-06,
+      "loss": 0.6238,
+      "step": 471
+    },
+    {
+      "epoch": 0.12531527943714324,
+      "grad_norm": 0.31361568411029606,
+      "learning_rate": 4.996627225113432e-06,
+      "loss": 0.6552,
+      "step": 472
+    },
+    {
+      "epoch": 0.12558077791052702,
+      "grad_norm": 0.309880081345077,
+      "learning_rate": 4.996609071648384e-06,
+      "loss": 0.6661,
+      "step": 473
+    },
+    {
+      "epoch": 0.1258462763839108,
+      "grad_norm": 0.32315101335746027,
+      "learning_rate": 4.996590869493347e-06,
+      "loss": 0.7002,
+      "step": 474
+    },
+    {
+      "epoch": 0.12611177485729458,
+      "grad_norm": 0.3149890808320547,
+      "learning_rate": 4.996572618648675e-06,
+      "loss": 0.6614,
+      "step": 475
+    },
+    {
+      "epoch": 0.12637727333067836,
+      "grad_norm": 0.3322823792242008,
+      "learning_rate": 4.996554319114724e-06,
+      "loss": 0.6387,
+      "step": 476
+    },
+    {
+      "epoch": 0.12664277180406214,
+      "grad_norm": 0.3371307654023024,
+      "learning_rate": 4.996535970891852e-06,
+      "loss": 0.6466,
+      "step": 477
+    },
+    {
+      "epoch": 0.12690827027744592,
+      "grad_norm": 0.32510819181993517,
+      "learning_rate": 4.9965175739804165e-06,
+      "loss": 0.7028,
+      "step": 478
+    },
+    {
+      "epoch": 0.1271737687508297,
+      "grad_norm": 0.3288285496065907,
+      "learning_rate": 4.996499128380775e-06,
+      "loss": 0.6783,
+      "step": 479
+    },
+    {
+      "epoch": 0.12743926722421345,
+      "grad_norm": 0.3090508055553086,
+      "learning_rate": 4.996480634093287e-06,
+      "loss": 0.6767,
+      "step": 480
+    },
+    {
+      "epoch": 0.12770476569759723,
+      "grad_norm": 0.32731228442771315,
+      "learning_rate": 4.9964620911183154e-06,
+      "loss": 0.6456,
+      "step": 481
+    },
+    {
+      "epoch": 0.127970264170981,
+      "grad_norm": 0.323293904647429,
+      "learning_rate": 4.996443499456221e-06,
+      "loss": 0.6439,
+      "step": 482
+    },
+    {
+      "epoch": 0.1282357626443648,
+      "grad_norm": 0.3316511232865672,
+      "learning_rate": 4.996424859107365e-06,
+      "loss": 0.7088,
+      "step": 483
+    },
+    {
+      "epoch": 0.12850126111774857,
+      "grad_norm": 0.32521408570883387,
+      "learning_rate": 4.9964061700721124e-06,
+      "loss": 0.6177,
+      "step": 484
+    },
+    {
+      "epoch": 0.12876675959113235,
+      "grad_norm": 0.3316638909021025,
+      "learning_rate": 4.996387432350827e-06,
+      "loss": 0.6778,
+      "step": 485
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 0.3299686641445629,
+      "learning_rate": 4.996368645943874e-06,
+      "loss": 0.6897,
+      "step": 486
+    },
+    {
+      "epoch": 0.1292977565378999,
+      "grad_norm": 0.32230058363824565,
+      "learning_rate": 4.996349810851621e-06,
+      "loss": 0.6731,
+      "step": 487
+    },
+    {
+      "epoch": 0.12956325501128368,
+      "grad_norm": 0.31931572173577116,
+      "learning_rate": 4.996330927074434e-06,
+      "loss": 0.6659,
+      "step": 488
+    },
+    {
+      "epoch": 0.12982875348466746,
+      "grad_norm": 0.3400639593522898,
+      "learning_rate": 4.996311994612683e-06,
+      "loss": 0.658,
+      "step": 489
+    },
+    {
+      "epoch": 0.13009425195805124,
+      "grad_norm": 0.3169094401640569,
+      "learning_rate": 4.996293013466735e-06,
+      "loss": 0.6712,
+      "step": 490
+    },
+    {
+      "epoch": 0.13035975043143502,
+      "grad_norm": 0.32357816964838454,
+      "learning_rate": 4.996273983636961e-06,
+      "loss": 0.6663,
+      "step": 491
+    },
+    {
+      "epoch": 0.1306252489048188,
+      "grad_norm": 0.32940241014043,
+      "learning_rate": 4.996254905123734e-06,
+      "loss": 0.6579,
+      "step": 492
+    },
+    {
+      "epoch": 0.13089074737820258,
+      "grad_norm": 0.3193910004394228,
+      "learning_rate": 4.9962357779274225e-06,
+      "loss": 0.6527,
+      "step": 493
+    },
+    {
+      "epoch": 0.13115624585158636,
+      "grad_norm": 0.3432696084572552,
+      "learning_rate": 4.996216602048402e-06,
+      "loss": 0.6496,
+      "step": 494
+    },
+    {
+      "epoch": 0.13142174432497014,
+      "grad_norm": 0.33351583885342173,
+      "learning_rate": 4.996197377487047e-06,
+      "loss": 0.6813,
+      "step": 495
+    },
+    {
+      "epoch": 0.13168724279835392,
+      "grad_norm": 0.34081496762738256,
+      "learning_rate": 4.996178104243731e-06,
+      "loss": 0.6772,
+      "step": 496
+    },
+    {
+      "epoch": 0.1319527412717377,
+      "grad_norm": 0.33491355000505957,
+      "learning_rate": 4.9961587823188294e-06,
+      "loss": 0.662,
+      "step": 497
+    },
+    {
+      "epoch": 0.13221823974512145,
+      "grad_norm": 0.33301061936692605,
+      "learning_rate": 4.996139411712721e-06,
+      "loss": 0.6918,
+      "step": 498
+    },
+    {
+      "epoch": 0.13248373821850523,
+      "grad_norm": 0.3400842904392137,
+      "learning_rate": 4.996119992425782e-06,
+      "loss": 0.6647,
+      "step": 499
+    },
+    {
+      "epoch": 0.132749236691889,
+      "grad_norm": 0.3102669536299804,
+      "learning_rate": 4.9961005244583915e-06,
+      "loss": 0.6541,
+      "step": 500
+    },
+    {
+      "epoch": 0.1330147351652728,
+      "grad_norm": 0.328665079455763,
+      "learning_rate": 4.99608100781093e-06,
+      "loss": 0.6683,
+      "step": 501
+    },
+    {
+      "epoch": 0.13328023363865657,
+      "grad_norm": 0.33072326613560726,
+      "learning_rate": 4.996061442483776e-06,
+      "loss": 0.6707,
+      "step": 502
+    },
+    {
+      "epoch": 0.13354573211204035,
+      "grad_norm": 0.3359127670041347,
+      "learning_rate": 4.996041828477314e-06,
+      "loss": 0.6756,
+      "step": 503
+    },
+    {
+      "epoch": 0.13381123058542413,
+      "grad_norm": 0.34455880718341786,
+      "learning_rate": 4.996022165791924e-06,
+      "loss": 0.6917,
+      "step": 504
+    },
+    {
+      "epoch": 0.1340767290588079,
+      "grad_norm": 0.32900522206732946,
+      "learning_rate": 4.99600245442799e-06,
+      "loss": 0.6471,
+      "step": 505
+    },
+    {
+      "epoch": 0.1343422275321917,
+      "grad_norm": 0.3104006173223105,
+      "learning_rate": 4.995982694385898e-06,
+      "loss": 0.6592,
+      "step": 506
+    },
+    {
+      "epoch": 0.13460772600557547,
+      "grad_norm": 0.3134989782521258,
+      "learning_rate": 4.995962885666031e-06,
+      "loss": 0.6551,
+      "step": 507
+    },
+    {
+      "epoch": 0.13487322447895925,
+      "grad_norm": 0.33422565458562004,
+      "learning_rate": 4.9959430282687785e-06,
+      "loss": 0.6428,
+      "step": 508
+    },
+    {
+      "epoch": 0.13513872295234303,
+      "grad_norm": 0.32444176124655094,
+      "learning_rate": 4.995923122194523e-06,
+      "loss": 0.6525,
+      "step": 509
+    },
+    {
+      "epoch": 0.1354042214257268,
+      "grad_norm": 0.33384089903040765,
+      "learning_rate": 4.995903167443658e-06,
+      "loss": 0.6828,
+      "step": 510
+    },
+    {
+      "epoch": 0.13566971989911059,
+      "grad_norm": 0.34164211670121414,
+      "learning_rate": 4.99588316401657e-06,
+      "loss": 0.6701,
+      "step": 511
+    },
+    {
+      "epoch": 0.13593521837249437,
+      "grad_norm": 0.3323764019600334,
+      "learning_rate": 4.995863111913648e-06,
+      "loss": 0.6707,
+      "step": 512
+    },
+    {
+      "epoch": 0.13620071684587814,
+      "grad_norm": 0.3228690168718786,
+      "learning_rate": 4.9958430111352855e-06,
+      "loss": 0.6482,
+      "step": 513
+    },
+    {
+      "epoch": 0.13646621531926192,
+      "grad_norm": 0.3210241856808829,
+      "learning_rate": 4.995822861681873e-06,
+      "loss": 0.6383,
+      "step": 514
+    },
+    {
+      "epoch": 0.1367317137926457,
+      "grad_norm": 0.32995295199460256,
+      "learning_rate": 4.995802663553803e-06,
+      "loss": 0.6645,
+      "step": 515
+    },
+    {
+      "epoch": 0.13699721226602948,
+      "grad_norm": 0.32683288941278976,
+      "learning_rate": 4.995782416751472e-06,
+      "loss": 0.6647,
+      "step": 516
+    },
+    {
+      "epoch": 0.13726271073941324,
+      "grad_norm": 0.3196919907438924,
+      "learning_rate": 4.995762121275272e-06,
+      "loss": 0.6881,
+      "step": 517
+    },
+    {
+      "epoch": 0.13752820921279701,
+      "grad_norm": 0.33328796423808194,
+      "learning_rate": 4.9957417771256e-06,
+      "loss": 0.6933,
+      "step": 518
+    },
+    {
+      "epoch": 0.1377937076861808,
+      "grad_norm": 0.3349629222708952,
+      "learning_rate": 4.995721384302853e-06,
+      "loss": 0.6785,
+      "step": 519
+    },
+    {
+      "epoch": 0.13805920615956457,
+      "grad_norm": 0.34465169753892805,
+      "learning_rate": 4.995700942807429e-06,
+      "loss": 0.6657,
+      "step": 520
+    },
+    {
+      "epoch": 0.13832470463294835,
+      "grad_norm": 0.3135898169384583,
+      "learning_rate": 4.995680452639725e-06,
+      "loss": 0.6412,
+      "step": 521
+    },
+    {
+      "epoch": 0.13859020310633213,
+      "grad_norm": 0.35602453516277327,
+      "learning_rate": 4.995659913800143e-06,
+      "loss": 0.6343,
+      "step": 522
+    },
+    {
+      "epoch": 0.1388557015797159,
+      "grad_norm": 0.3204032153635913,
+      "learning_rate": 4.995639326289081e-06,
+      "loss": 0.675,
+      "step": 523
+    },
+    {
+      "epoch": 0.1391212000530997,
+      "grad_norm": 0.3277898414945331,
+      "learning_rate": 4.995618690106942e-06,
+      "loss": 0.6886,
+      "step": 524
+    },
+    {
+      "epoch": 0.13938669852648347,
+      "grad_norm": 0.3393355770985635,
+      "learning_rate": 4.995598005254129e-06,
+      "loss": 0.6717,
+      "step": 525
+    },
+    {
+      "epoch": 0.13965219699986725,
+      "grad_norm": 0.32347596138398255,
+      "learning_rate": 4.995577271731044e-06,
+      "loss": 0.644,
+      "step": 526
+    },
+    {
+      "epoch": 0.13991769547325103,
+      "grad_norm": 0.3266478883557009,
+      "learning_rate": 4.995556489538091e-06,
+      "loss": 0.6767,
+      "step": 527
+    },
+    {
+      "epoch": 0.1401831939466348,
+      "grad_norm": 0.3381961614423876,
+      "learning_rate": 4.995535658675678e-06,
+      "loss": 0.6889,
+      "step": 528
+    },
+    {
+      "epoch": 0.1404486924200186,
+      "grad_norm": 0.3283413859213309,
+      "learning_rate": 4.995514779144208e-06,
+      "loss": 0.6768,
+      "step": 529
+    },
+    {
+      "epoch": 0.14071419089340237,
+      "grad_norm": 0.3281449282605594,
+      "learning_rate": 4.995493850944091e-06,
+      "loss": 0.6835,
+      "step": 530
+    },
+    {
+      "epoch": 0.14097968936678615,
+      "grad_norm": 0.32330376989056236,
+      "learning_rate": 4.9954728740757336e-06,
+      "loss": 0.6586,
+      "step": 531
+    },
+    {
+      "epoch": 0.14124518784016993,
+      "grad_norm": 0.34094390843593775,
+      "learning_rate": 4.995451848539545e-06,
+      "loss": 0.662,
+      "step": 532
+    },
+    {
+      "epoch": 0.1415106863135537,
+      "grad_norm": 0.31373142731202014,
+      "learning_rate": 4.9954307743359355e-06,
+      "loss": 0.6439,
+      "step": 533
+    },
+    {
+      "epoch": 0.1417761847869375,
+      "grad_norm": 0.33183275332426326,
+      "learning_rate": 4.995409651465316e-06,
+      "loss": 0.6567,
+      "step": 534
+    },
+    {
+      "epoch": 0.14204168326032127,
+      "grad_norm": 0.3390225618878591,
+      "learning_rate": 4.995388479928098e-06,
+      "loss": 0.6446,
+      "step": 535
+    },
+    {
+      "epoch": 0.14230718173370502,
+      "grad_norm": 0.32369220295055845,
+      "learning_rate": 4.995367259724696e-06,
+      "loss": 0.6731,
+      "step": 536
+    },
+    {
+      "epoch": 0.1425726802070888,
+      "grad_norm": 0.33924915117232796,
+      "learning_rate": 4.9953459908555215e-06,
+      "loss": 0.6767,
+      "step": 537
+    },
+    {
+      "epoch": 0.14283817868047258,
+      "grad_norm": 0.3521701665111978,
+      "learning_rate": 4.995324673320991e-06,
+      "loss": 0.657,
+      "step": 538
+    },
+    {
+      "epoch": 0.14310367715385636,
+      "grad_norm": 0.3343091569613087,
+      "learning_rate": 4.99530330712152e-06,
+      "loss": 0.6953,
+      "step": 539
+    },
+    {
+      "epoch": 0.14336917562724014,
+      "grad_norm": 0.33082562613574634,
+      "learning_rate": 4.995281892257525e-06,
+      "loss": 0.6555,
+      "step": 540
+    },
+    {
+      "epoch": 0.14363467410062392,
+      "grad_norm": 0.3360100839446358,
+      "learning_rate": 4.9952604287294236e-06,
+      "loss": 0.6534,
+      "step": 541
+    },
+    {
+      "epoch": 0.1439001725740077,
+      "grad_norm": 0.3196775960507627,
+      "learning_rate": 4.995238916537634e-06,
+      "loss": 0.6713,
+      "step": 542
+    },
+    {
+      "epoch": 0.14416567104739147,
+      "grad_norm": 0.33537763471186866,
+      "learning_rate": 4.995217355682576e-06,
+      "loss": 0.676,
+      "step": 543
+    },
+    {
+      "epoch": 0.14443116952077525,
+      "grad_norm": 0.34609653703087356,
+      "learning_rate": 4.9951957461646705e-06,
+      "loss": 0.6683,
+      "step": 544
+    },
+    {
+      "epoch": 0.14469666799415903,
+      "grad_norm": 0.35264005734812376,
+      "learning_rate": 4.995174087984339e-06,
+      "loss": 0.6619,
+      "step": 545
+    },
+    {
+      "epoch": 0.1449621664675428,
+      "grad_norm": 0.3245915297253958,
+      "learning_rate": 4.995152381142003e-06,
+      "loss": 0.66,
+      "step": 546
+    },
+    {
+      "epoch": 0.1452276649409266,
+      "grad_norm": 0.342351782543807,
+      "learning_rate": 4.995130625638087e-06,
+      "loss": 0.6909,
+      "step": 547
+    },
+    {
+      "epoch": 0.14549316341431037,
+      "grad_norm": 0.34685925800788514,
+      "learning_rate": 4.995108821473014e-06,
+      "loss": 0.6886,
+      "step": 548
+    },
+    {
+      "epoch": 0.14575866188769415,
+      "grad_norm": 0.33818507932197633,
+      "learning_rate": 4.99508696864721e-06,
+      "loss": 0.7143,
+      "step": 549
+    },
+    {
+      "epoch": 0.14602416036107793,
+      "grad_norm": 0.32524686758940563,
+      "learning_rate": 4.995065067161101e-06,
+      "loss": 0.6731,
+      "step": 550
+    },
+    {
+      "epoch": 0.1462896588344617,
+      "grad_norm": 0.32694976818783966,
+      "learning_rate": 4.9950431170151145e-06,
+      "loss": 0.6794,
+      "step": 551
+    },
+    {
+      "epoch": 0.1465551573078455,
+      "grad_norm": 0.33462789426067663,
+      "learning_rate": 4.995021118209679e-06,
+      "loss": 0.665,
+      "step": 552
+    },
+    {
+      "epoch": 0.14682065578122927,
+      "grad_norm": 0.34655446891220265,
+      "learning_rate": 4.994999070745222e-06,
+      "loss": 0.6797,
+      "step": 553
+    },
+    {
+      "epoch": 0.14708615425461305,
+      "grad_norm": 0.326030150865353,
+      "learning_rate": 4.994976974622175e-06,
+      "loss": 0.6561,
+      "step": 554
+    },
+    {
+      "epoch": 0.1473516527279968,
+      "grad_norm": 0.3414446163531893,
+      "learning_rate": 4.9949548298409665e-06,
+      "loss": 0.6751,
+      "step": 555
+    },
+    {
+      "epoch": 0.14761715120138058,
+      "grad_norm": 0.3539416165824423,
+      "learning_rate": 4.994932636402032e-06,
+      "loss": 0.6449,
+      "step": 556
+    },
+    {
+      "epoch": 0.14788264967476436,
+      "grad_norm": 0.3316388913017657,
+      "learning_rate": 4.994910394305801e-06,
+      "loss": 0.6387,
+      "step": 557
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 0.32747030798146964,
+      "learning_rate": 4.994888103552709e-06,
+      "loss": 0.6707,
+      "step": 558
+    },
+    {
+      "epoch": 0.14841364662153192,
+      "grad_norm": 0.37322744776731953,
+      "learning_rate": 4.994865764143192e-06,
+      "loss": 0.6819,
+      "step": 559
+    },
+    {
+      "epoch": 0.1486791450949157,
+      "grad_norm": 0.33672340444534893,
+      "learning_rate": 4.994843376077683e-06,
+      "loss": 0.6607,
+      "step": 560
+    },
+    {
+      "epoch": 0.14894464356829948,
+      "grad_norm": 0.3298195074729293,
+      "learning_rate": 4.994820939356619e-06,
+      "loss": 0.6569,
+      "step": 561
+    },
+    {
+      "epoch": 0.14921014204168326,
+      "grad_norm": 0.3498886832736364,
+      "learning_rate": 4.994798453980439e-06,
+      "loss": 0.65,
+      "step": 562
+    },
+    {
+      "epoch": 0.14947564051506704,
+      "grad_norm": 0.3371708895587616,
+      "learning_rate": 4.994775919949582e-06,
+      "loss": 0.6969,
+      "step": 563
+    },
+    {
+      "epoch": 0.14974113898845082,
+      "grad_norm": 0.3342634727667992,
+      "learning_rate": 4.994753337264484e-06,
+      "loss": 0.6767,
+      "step": 564
+    },
+    {
+      "epoch": 0.1500066374618346,
+      "grad_norm": 0.3260166590996332,
+      "learning_rate": 4.994730705925589e-06,
+      "loss": 0.6432,
+      "step": 565
+    },
+    {
+      "epoch": 0.15027213593521838,
+      "grad_norm": 0.3291001951665862,
+      "learning_rate": 4.994708025933337e-06,
+      "loss": 0.6365,
+      "step": 566
+    },
+    {
+      "epoch": 0.15053763440860216,
+      "grad_norm": 0.3310069934220363,
+      "learning_rate": 4.994685297288171e-06,
+      "loss": 0.6412,
+      "step": 567
+    },
+    {
+      "epoch": 0.15080313288198594,
+      "grad_norm": 0.3323823688607964,
+      "learning_rate": 4.994662519990533e-06,
+      "loss": 0.6277,
+      "step": 568
+    },
+    {
+      "epoch": 0.15106863135536971,
+      "grad_norm": 0.33134140774263565,
+      "learning_rate": 4.9946396940408674e-06,
+      "loss": 0.6284,
+      "step": 569
+    },
+    {
+      "epoch": 0.1513341298287535,
+      "grad_norm": 0.3366495795396674,
+      "learning_rate": 4.994616819439621e-06,
+      "loss": 0.6159,
+      "step": 570
+    },
+    {
+      "epoch": 0.15159962830213727,
+      "grad_norm": 0.34612904693930246,
+      "learning_rate": 4.994593896187238e-06,
+      "loss": 0.6755,
+      "step": 571
+    },
+    {
+      "epoch": 0.15186512677552105,
+      "grad_norm": 0.3325808273913825,
+      "learning_rate": 4.9945709242841656e-06,
+      "loss": 0.6744,
+      "step": 572
+    },
+    {
+      "epoch": 0.1521306252489048,
+      "grad_norm": 0.33475212875983545,
+      "learning_rate": 4.994547903730853e-06,
+      "loss": 0.6713,
+      "step": 573
+    },
+    {
+      "epoch": 0.15239612372228858,
+      "grad_norm": 0.3419898738743717,
+      "learning_rate": 4.994524834527748e-06,
+      "loss": 0.6991,
+      "step": 574
+    },
+    {
+      "epoch": 0.15266162219567236,
+      "grad_norm": 0.3359194326215708,
+      "learning_rate": 4.994501716675303e-06,
+      "loss": 0.635,
+      "step": 575
+    },
+    {
+      "epoch": 0.15292712066905614,
+      "grad_norm": 0.3310403527739758,
+      "learning_rate": 4.994478550173965e-06,
+      "loss": 0.6285,
+      "step": 576
+    },
+    {
+      "epoch": 0.15319261914243992,
+      "grad_norm": 0.32681290994687995,
+      "learning_rate": 4.994455335024188e-06,
+      "loss": 0.6506,
+      "step": 577
+    },
+    {
+      "epoch": 0.1534581176158237,
+      "grad_norm": 0.335434627448189,
+      "learning_rate": 4.9944320712264254e-06,
+      "loss": 0.6811,
+      "step": 578
+    },
+    {
+      "epoch": 0.15372361608920748,
+      "grad_norm": 0.3315355703841914,
+      "learning_rate": 4.994408758781129e-06,
+      "loss": 0.6841,
+      "step": 579
+    },
+    {
+      "epoch": 0.15398911456259126,
+      "grad_norm": 0.453893517786475,
+      "learning_rate": 4.994385397688756e-06,
+      "loss": 0.5861,
+      "step": 580
+    },
+    {
+      "epoch": 0.15425461303597504,
+      "grad_norm": 0.35520556295894407,
+      "learning_rate": 4.994361987949759e-06,
+      "loss": 0.6726,
+      "step": 581
+    },
+    {
+      "epoch": 0.15452011150935882,
+      "grad_norm": 0.3597665130331063,
+      "learning_rate": 4.994338529564597e-06,
+      "loss": 0.6865,
+      "step": 582
+    },
+    {
+      "epoch": 0.1547856099827426,
+      "grad_norm": 0.3598421860319184,
+      "learning_rate": 4.994315022533725e-06,
+      "loss": 0.6689,
+      "step": 583
+    },
+    {
+      "epoch": 0.15505110845612638,
+      "grad_norm": 0.3335391109966114,
+      "learning_rate": 4.994291466857604e-06,
+      "loss": 0.6286,
+      "step": 584
+    },
+    {
+      "epoch": 0.15531660692951016,
+      "grad_norm": 0.35510010063761116,
+      "learning_rate": 4.994267862536691e-06,
+      "loss": 0.6747,
+      "step": 585
+    },
+    {
+      "epoch": 0.15558210540289394,
+      "grad_norm": 0.345721931026568,
+      "learning_rate": 4.994244209571449e-06,
+      "loss": 0.6901,
+      "step": 586
+    },
+    {
+      "epoch": 0.15584760387627772,
+      "grad_norm": 0.3329197069443996,
+      "learning_rate": 4.994220507962338e-06,
+      "loss": 0.6862,
+      "step": 587
+    },
+    {
+      "epoch": 0.1561131023496615,
+      "grad_norm": 0.3260484470975118,
+      "learning_rate": 4.99419675770982e-06,
+      "loss": 0.6617,
+      "step": 588
+    },
+    {
+      "epoch": 0.15637860082304528,
+      "grad_norm": 0.33040165821619644,
+      "learning_rate": 4.9941729588143586e-06,
+      "loss": 0.6654,
+      "step": 589
+    },
+    {
+      "epoch": 0.15664409929642906,
+      "grad_norm": 0.3479610458896437,
+      "learning_rate": 4.994149111276417e-06,
+      "loss": 0.6471,
+      "step": 590
+    },
+    {
+      "epoch": 0.15690959776981284,
+      "grad_norm": 0.3371334049835157,
+      "learning_rate": 4.994125215096462e-06,
+      "loss": 0.6469,
+      "step": 591
+    },
+    {
+      "epoch": 0.1571750962431966,
+      "grad_norm": 0.3356169532422175,
+      "learning_rate": 4.994101270274957e-06,
+      "loss": 0.6708,
+      "step": 592
+    },
+    {
+      "epoch": 0.15744059471658037,
+      "grad_norm": 0.33618341704576754,
+      "learning_rate": 4.994077276812372e-06,
+      "loss": 0.6622,
+      "step": 593
+    },
+    {
+      "epoch": 0.15770609318996415,
+      "grad_norm": 0.33281710205959186,
+      "learning_rate": 4.994053234709173e-06,
+      "loss": 0.6735,
+      "step": 594
+    },
+    {
+      "epoch": 0.15797159166334793,
+      "grad_norm": 0.3571218236860454,
+      "learning_rate": 4.99402914396583e-06,
+      "loss": 0.6611,
+      "step": 595
+    },
+    {
+      "epoch": 0.1582370901367317,
+      "grad_norm": 0.3339588523505776,
+      "learning_rate": 4.994005004582811e-06,
+      "loss": 0.6447,
+      "step": 596
+    },
+    {
+      "epoch": 0.15850258861011549,
+      "grad_norm": 0.3199764131162705,
+      "learning_rate": 4.993980816560589e-06,
+      "loss": 0.6297,
+      "step": 597
+    },
+    {
+      "epoch": 0.15876808708349927,
+      "grad_norm": 0.37021465038246504,
+      "learning_rate": 4.993956579899635e-06,
+      "loss": 0.6348,
+      "step": 598
+    },
+    {
+      "epoch": 0.15903358555688304,
+      "grad_norm": 0.3270374220899211,
+      "learning_rate": 4.993932294600421e-06,
+      "loss": 0.6613,
+      "step": 599
+    },
+    {
+      "epoch": 0.15929908403026682,
+      "grad_norm": 0.33111837493896173,
+      "learning_rate": 4.9939079606634214e-06,
+      "loss": 0.6502,
+      "step": 600
+    },
+    {
+      "epoch": 0.1595645825036506,
+      "grad_norm": 0.3311759761687824,
+      "learning_rate": 4.99388357808911e-06,
+      "loss": 0.6713,
+      "step": 601
+    },
+    {
+      "epoch": 0.15983008097703438,
+      "grad_norm": 0.3365585056743753,
+      "learning_rate": 4.993859146877963e-06,
+      "loss": 0.653,
+      "step": 602
+    },
+    {
+      "epoch": 0.16009557945041816,
+      "grad_norm": 0.34668796669168633,
+      "learning_rate": 4.993834667030456e-06,
+      "loss": 0.7061,
+      "step": 603
+    },
+    {
+      "epoch": 0.16036107792380194,
+      "grad_norm": 0.32664222917788455,
+      "learning_rate": 4.993810138547067e-06,
+      "loss": 0.6335,
+      "step": 604
+    },
+    {
+      "epoch": 0.16062657639718572,
+      "grad_norm": 0.33890684011208744,
+      "learning_rate": 4.993785561428275e-06,
+      "loss": 0.6339,
+      "step": 605
+    },
+    {
+      "epoch": 0.1608920748705695,
+      "grad_norm": 0.35136215572926227,
+      "learning_rate": 4.993760935674559e-06,
+      "loss": 0.6745,
+      "step": 606
+    },
+    {
+      "epoch": 0.16115757334395328,
+      "grad_norm": 0.3203309936280434,
+      "learning_rate": 4.993736261286399e-06,
+      "loss": 0.6084,
+      "step": 607
+    },
+    {
+      "epoch": 0.16142307181733706,
+      "grad_norm": 0.3452582678950179,
+      "learning_rate": 4.993711538264275e-06,
+      "loss": 0.6735,
+      "step": 608
+    },
+    {
+      "epoch": 0.16168857029072084,
+      "grad_norm": 0.3354994569214719,
+      "learning_rate": 4.993686766608672e-06,
+      "loss": 0.6628,
+      "step": 609
+    },
+    {
+      "epoch": 0.16195406876410462,
+      "grad_norm": 0.334237430848364,
+      "learning_rate": 4.99366194632007e-06,
+      "loss": 0.6781,
+      "step": 610
+    },
+    {
+      "epoch": 0.16221956723748837,
+      "grad_norm": 0.3313359117252434,
+      "learning_rate": 4.993637077398955e-06,
+      "loss": 0.6456,
+      "step": 611
+    },
+    {
+      "epoch": 0.16248506571087215,
+      "grad_norm": 0.34060902405580806,
+      "learning_rate": 4.993612159845812e-06,
+      "loss": 0.6532,
+      "step": 612
+    },
+    {
+      "epoch": 0.16275056418425593,
+      "grad_norm": 0.3255501853576438,
+      "learning_rate": 4.993587193661126e-06,
+      "loss": 0.606,
+      "step": 613
+    },
+    {
+      "epoch": 0.1630160626576397,
+      "grad_norm": 0.3507908793390629,
+      "learning_rate": 4.993562178845384e-06,
+      "loss": 0.6698,
+      "step": 614
+    },
+    {
+      "epoch": 0.1632815611310235,
+      "grad_norm": 0.34343872187243774,
+      "learning_rate": 4.993537115399075e-06,
+      "loss": 0.6901,
+      "step": 615
+    },
+    {
+      "epoch": 0.16354705960440727,
+      "grad_norm": 0.33413101033583026,
+      "learning_rate": 4.993512003322687e-06,
+      "loss": 0.6741,
+      "step": 616
+    },
+    {
+      "epoch": 0.16381255807779105,
+      "grad_norm": 0.34354965288666184,
+      "learning_rate": 4.9934868426167095e-06,
+      "loss": 0.6784,
+      "step": 617
+    },
+    {
+      "epoch": 0.16407805655117483,
+      "grad_norm": 0.34372422112041834,
+      "learning_rate": 4.993461633281633e-06,
+      "loss": 0.6296,
+      "step": 618
+    },
+    {
+      "epoch": 0.1643435550245586,
+      "grad_norm": 0.3334250118427838,
+      "learning_rate": 4.993436375317951e-06,
+      "loss": 0.5963,
+      "step": 619
+    },
+    {
+      "epoch": 0.1646090534979424,
+      "grad_norm": 0.37412542083164385,
+      "learning_rate": 4.993411068726153e-06,
+      "loss": 0.6439,
+      "step": 620
+    },
+    {
+      "epoch": 0.16487455197132617,
+      "grad_norm": 0.34083625580390686,
+      "learning_rate": 4.993385713506735e-06,
+      "loss": 0.6488,
+      "step": 621
+    },
+    {
+      "epoch": 0.16514005044470995,
+      "grad_norm": 0.32910774835941825,
+      "learning_rate": 4.9933603096601904e-06,
+      "loss": 0.636,
+      "step": 622
+    },
+    {
+      "epoch": 0.16540554891809373,
+      "grad_norm": 0.38172097171318353,
+      "learning_rate": 4.993334857187015e-06,
+      "loss": 0.6628,
+      "step": 623
+    },
+    {
+      "epoch": 0.1656710473914775,
+      "grad_norm": 0.37361327501254793,
+      "learning_rate": 4.993309356087706e-06,
+      "loss": 0.6785,
+      "step": 624
+    },
+    {
+      "epoch": 0.16593654586486128,
+      "grad_norm": 0.3354208369872467,
+      "learning_rate": 4.993283806362759e-06,
+      "loss": 0.688,
+      "step": 625
+    },
+    {
+      "epoch": 0.16620204433824506,
+      "grad_norm": 0.33450601458476165,
+      "learning_rate": 4.9932582080126734e-06,
+      "loss": 0.7019,
+      "step": 626
+    },
+    {
+      "epoch": 0.16646754281162884,
+      "grad_norm": 0.4924071901734736,
+      "learning_rate": 4.9932325610379485e-06,
+      "loss": 0.6463,
+      "step": 627
+    },
+    {
+      "epoch": 0.16673304128501262,
+      "grad_norm": 0.3301536451004917,
+      "learning_rate": 4.993206865439084e-06,
+      "loss": 0.6683,
+      "step": 628
+    },
+    {
+      "epoch": 0.1669985397583964,
+      "grad_norm": 0.35214494911561767,
+      "learning_rate": 4.993181121216582e-06,
+      "loss": 0.6737,
+      "step": 629
+    },
+    {
+      "epoch": 0.16726403823178015,
+      "grad_norm": 0.3996492927577036,
+      "learning_rate": 4.9931553283709425e-06,
+      "loss": 0.6626,
+      "step": 630
+    },
+    {
+      "epoch": 0.16752953670516393,
+      "grad_norm": 0.3472386845192621,
+      "learning_rate": 4.99312948690267e-06,
+      "loss": 0.647,
+      "step": 631
+    },
+    {
+      "epoch": 0.1677950351785477,
+      "grad_norm": 0.3284775082396095,
+      "learning_rate": 4.993103596812269e-06,
+      "loss": 0.6481,
+      "step": 632
+    },
+    {
+      "epoch": 0.1680605336519315,
+      "grad_norm": 0.3420332625692997,
+      "learning_rate": 4.993077658100243e-06,
+      "loss": 0.692,
+      "step": 633
+    },
+    {
+      "epoch": 0.16832603212531527,
+      "grad_norm": 0.3747111543822854,
+      "learning_rate": 4.993051670767098e-06,
+      "loss": 0.6747,
+      "step": 634
+    },
+    {
+      "epoch": 0.16859153059869905,
+      "grad_norm": 0.34727129132010565,
+      "learning_rate": 4.993025634813342e-06,
+      "loss": 0.6625,
+      "step": 635
+    },
+    {
+      "epoch": 0.16885702907208283,
+      "grad_norm": 0.33104329412752553,
+      "learning_rate": 4.992999550239482e-06,
+      "loss": 0.674,
+      "step": 636
+    },
+    {
+      "epoch": 0.1691225275454666,
+      "grad_norm": 0.34417715343099703,
+      "learning_rate": 4.992973417046028e-06,
+      "loss": 0.6273,
+      "step": 637
+    },
+    {
+      "epoch": 0.1693880260188504,
+      "grad_norm": 0.389161476418092,
+      "learning_rate": 4.992947235233488e-06,
+      "loss": 0.6468,
+      "step": 638
+    },
+    {
+      "epoch": 0.16965352449223417,
+      "grad_norm": 0.3382559855533501,
+      "learning_rate": 4.992921004802372e-06,
+      "loss": 0.5964,
+      "step": 639
+    },
+    {
+      "epoch": 0.16991902296561795,
+      "grad_norm": 0.34217674700171413,
+      "learning_rate": 4.992894725753193e-06,
+      "loss": 0.6689,
+      "step": 640
+    },
+    {
+      "epoch": 0.17018452143900173,
+      "grad_norm": 0.39400178355254273,
+      "learning_rate": 4.9928683980864634e-06,
+      "loss": 0.6622,
+      "step": 641
+    },
+    {
+      "epoch": 0.1704500199123855,
+      "grad_norm": 0.42784826089298833,
+      "learning_rate": 4.992842021802697e-06,
+      "loss": 0.6255,
+      "step": 642
+    },
+    {
+      "epoch": 0.1707155183857693,
+      "grad_norm": 0.33397660471822366,
+      "learning_rate": 4.992815596902407e-06,
+      "loss": 0.6612,
+      "step": 643
+    },
+    {
+      "epoch": 0.17098101685915307,
+      "grad_norm": 0.3660443714679061,
+      "learning_rate": 4.99278912338611e-06,
+      "loss": 0.66,
+      "step": 644
+    },
+    {
+      "epoch": 0.17124651533253685,
+      "grad_norm": 0.3506234120068022,
+      "learning_rate": 4.992762601254322e-06,
+      "loss": 0.6927,
+      "step": 645
+    },
+    {
+      "epoch": 0.17151201380592063,
+      "grad_norm": 0.4752779091611214,
+      "learning_rate": 4.992736030507559e-06,
+      "loss": 0.6296,
+      "step": 646
+    },
+    {
+      "epoch": 0.1717775122793044,
+      "grad_norm": 0.34570403940780386,
+      "learning_rate": 4.99270941114634e-06,
+      "loss": 0.6307,
+      "step": 647
+    },
+    {
+      "epoch": 0.17204301075268819,
+      "grad_norm": 0.33898176223347,
+      "learning_rate": 4.992682743171185e-06,
+      "loss": 0.6862,
+      "step": 648
+    },
+    {
+      "epoch": 0.17230850922607194,
+      "grad_norm": 0.3527315562887804,
+      "learning_rate": 4.992656026582613e-06,
+      "loss": 0.6435,
+      "step": 649
+    },
+    {
+      "epoch": 0.17257400769945572,
+      "grad_norm": 0.5591109976004349,
+      "learning_rate": 4.992629261381146e-06,
+      "loss": 0.6455,
+      "step": 650
+    },
+    {
+      "epoch": 0.1728395061728395,
+      "grad_norm": 0.3511782949125343,
+      "learning_rate": 4.992602447567304e-06,
+      "loss": 0.6937,
+      "step": 651
+    },
+    {
+      "epoch": 0.17310500464622328,
+      "grad_norm": 0.3660381170727164,
+      "learning_rate": 4.992575585141612e-06,
+      "loss": 0.6492,
+      "step": 652
+    },
+    {
+      "epoch": 0.17337050311960706,
+      "grad_norm": 0.3770987884631617,
+      "learning_rate": 4.992548674104594e-06,
+      "loss": 0.6823,
+      "step": 653
+    },
+    {
+      "epoch": 0.17363600159299084,
+      "grad_norm": 0.3761289256131638,
+      "learning_rate": 4.992521714456774e-06,
+      "loss": 0.6396,
+      "step": 654
+    },
+    {
+      "epoch": 0.17390150006637461,
+      "grad_norm": 0.3475518003285641,
+      "learning_rate": 4.992494706198678e-06,
+      "loss": 0.6428,
+      "step": 655
+    },
+    {
+      "epoch": 0.1741669985397584,
+      "grad_norm": 0.338760464616784,
+      "learning_rate": 4.992467649330831e-06,
+      "loss": 0.6868,
+      "step": 656
+    },
+    {
+      "epoch": 0.17443249701314217,
+      "grad_norm": 0.329468348433597,
+      "learning_rate": 4.992440543853763e-06,
+      "loss": 0.6339,
+      "step": 657
+    },
+    {
+      "epoch": 0.17469799548652595,
+      "grad_norm": 0.3520208363642314,
+      "learning_rate": 4.992413389768001e-06,
+      "loss": 0.6196,
+      "step": 658
+    },
+    {
+      "epoch": 0.17496349395990973,
+      "grad_norm": 0.3526400854331627,
+      "learning_rate": 4.992386187074078e-06,
+      "loss": 0.6451,
+      "step": 659
+    },
+    {
+      "epoch": 0.1752289924332935,
+      "grad_norm": 0.35797450081758125,
+      "learning_rate": 4.992358935772519e-06,
+      "loss": 0.6789,
+      "step": 660
+    },
+    {
+      "epoch": 0.1754944909066773,
+      "grad_norm": 0.3350384684829453,
+      "learning_rate": 4.99233163586386e-06,
+      "loss": 0.6483,
+      "step": 661
+    },
+    {
+      "epoch": 0.17575998938006107,
+      "grad_norm": 0.33898553335758785,
+      "learning_rate": 4.992304287348632e-06,
+      "loss": 0.6805,
+      "step": 662
+    },
+    {
+      "epoch": 0.17602548785344485,
+      "grad_norm": 0.33291920836012545,
+      "learning_rate": 4.992276890227368e-06,
+      "loss": 0.6401,
+      "step": 663
+    },
+    {
+      "epoch": 0.17629098632682863,
+      "grad_norm": 0.3507502389049936,
+      "learning_rate": 4.992249444500601e-06,
+      "loss": 0.6369,
+      "step": 664
+    },
+    {
+      "epoch": 0.1765564848002124,
+      "grad_norm": 0.3528312083787476,
+      "learning_rate": 4.992221950168869e-06,
+      "loss": 0.6696,
+      "step": 665
+    },
+    {
+      "epoch": 0.1768219832735962,
+      "grad_norm": 0.33847661618514874,
+      "learning_rate": 4.992194407232708e-06,
+      "loss": 0.6583,
+      "step": 666
+    },
+    {
+      "epoch": 0.17708748174697994,
+      "grad_norm": 0.33546638130976375,
+      "learning_rate": 4.9921668156926526e-06,
+      "loss": 0.6779,
+      "step": 667
+    },
+    {
+      "epoch": 0.17735298022036372,
+      "grad_norm": 0.34310485759623466,
+      "learning_rate": 4.992139175549243e-06,
+      "loss": 0.678,
+      "step": 668
+    },
+    {
+      "epoch": 0.1776184786937475,
+      "grad_norm": 0.35026130026948477,
+      "learning_rate": 4.992111486803017e-06,
+      "loss": 0.6064,
+      "step": 669
+    },
+    {
+      "epoch": 0.17788397716713128,
+      "grad_norm": 0.3408774225154651,
+      "learning_rate": 4.992083749454516e-06,
+      "loss": 0.6478,
+      "step": 670
+    },
+    {
+      "epoch": 0.17814947564051506,
+      "grad_norm": 0.34073396016254826,
+      "learning_rate": 4.99205596350428e-06,
+      "loss": 0.6463,
+      "step": 671
+    },
+    {
+      "epoch": 0.17841497411389884,
+      "grad_norm": 0.35756027313020583,
+      "learning_rate": 4.992028128952851e-06,
+      "loss": 0.6413,
+      "step": 672
+    },
+    {
+      "epoch": 0.17868047258728262,
+      "grad_norm": 0.3336875362140239,
+      "learning_rate": 4.992000245800772e-06,
+      "loss": 0.6461,
+      "step": 673
+    },
+    {
+      "epoch": 0.1789459710606664,
+      "grad_norm": 0.3445936259159989,
+      "learning_rate": 4.991972314048587e-06,
+      "loss": 0.6645,
+      "step": 674
+    },
+    {
+      "epoch": 0.17921146953405018,
+      "grad_norm": 0.33654721153260087,
+      "learning_rate": 4.99194433369684e-06,
+      "loss": 0.6289,
+      "step": 675
+    },
+    {
+      "epoch": 0.17947696800743396,
+      "grad_norm": 0.34037008807256686,
+      "learning_rate": 4.991916304746077e-06,
+      "loss": 0.6733,
+      "step": 676
+    },
+    {
+      "epoch": 0.17974246648081774,
+      "grad_norm": 0.34539272142794913,
+      "learning_rate": 4.991888227196846e-06,
+      "loss": 0.6486,
+      "step": 677
+    },
+    {
+      "epoch": 0.18000796495420152,
+      "grad_norm": 0.3306497540608407,
+      "learning_rate": 4.9918601010496925e-06,
+      "loss": 0.6784,
+      "step": 678
+    },
+    {
+      "epoch": 0.1802734634275853,
+      "grad_norm": 0.3453371982415876,
+      "learning_rate": 4.991831926305166e-06,
+      "loss": 0.6757,
+      "step": 679
+    },
+    {
+      "epoch": 0.18053896190096907,
+      "grad_norm": 0.3506783353640043,
+      "learning_rate": 4.991803702963815e-06,
+      "loss": 0.6471,
+      "step": 680
+    },
+    {
+      "epoch": 0.18080446037435285,
+      "grad_norm": 0.3460571746922703,
+      "learning_rate": 4.9917754310261925e-06,
+      "loss": 0.6468,
+      "step": 681
+    },
+    {
+      "epoch": 0.18106995884773663,
+      "grad_norm": 0.3455399902951912,
+      "learning_rate": 4.991747110492847e-06,
+      "loss": 0.667,
+      "step": 682
+    },
+    {
+      "epoch": 0.1813354573211204,
+      "grad_norm": 0.3335499569300848,
+      "learning_rate": 4.991718741364333e-06,
+      "loss": 0.6478,
+      "step": 683
+    },
+    {
+      "epoch": 0.1816009557945042,
+      "grad_norm": 0.3570524927092075,
+      "learning_rate": 4.991690323641202e-06,
+      "loss": 0.6727,
+      "step": 684
+    },
+    {
+      "epoch": 0.18186645426788797,
+      "grad_norm": 0.3401655496067992,
+      "learning_rate": 4.991661857324009e-06,
+      "loss": 0.6437,
+      "step": 685
+    },
+    {
+      "epoch": 0.18213195274127172,
+      "grad_norm": 0.34351058102025905,
+      "learning_rate": 4.991633342413309e-06,
+      "loss": 0.6719,
+      "step": 686
+    },
+    {
+      "epoch": 0.1823974512146555,
+      "grad_norm": 0.33841071957533675,
+      "learning_rate": 4.991604778909659e-06,
+      "loss": 0.6371,
+      "step": 687
+    },
+    {
+      "epoch": 0.18266294968803928,
+      "grad_norm": 0.3420778911786694,
+      "learning_rate": 4.991576166813614e-06,
+      "loss": 0.6825,
+      "step": 688
+    },
+    {
+      "epoch": 0.18292844816142306,
+      "grad_norm": 0.3654088014459768,
+      "learning_rate": 4.991547506125735e-06,
+      "loss": 0.705,
+      "step": 689
+    },
+    {
+      "epoch": 0.18319394663480684,
+      "grad_norm": 0.34959108681412265,
+      "learning_rate": 4.991518796846578e-06,
+      "loss": 0.6296,
+      "step": 690
+    },
+    {
+      "epoch": 0.18345944510819062,
+      "grad_norm": 0.33902705185959686,
+      "learning_rate": 4.9914900389767044e-06,
+      "loss": 0.6555,
+      "step": 691
+    },
+    {
+      "epoch": 0.1837249435815744,
+      "grad_norm": 0.35297183061504295,
+      "learning_rate": 4.991461232516675e-06,
+      "loss": 0.6221,
+      "step": 692
+    },
+    {
+      "epoch": 0.18399044205495818,
+      "grad_norm": 0.3469568147576771,
+      "learning_rate": 4.991432377467052e-06,
+      "loss": 0.6346,
+      "step": 693
+    },
+    {
+      "epoch": 0.18425594052834196,
+      "grad_norm": 0.3386616592406038,
+      "learning_rate": 4.991403473828397e-06,
+      "loss": 0.6138,
+      "step": 694
+    },
+    {
+      "epoch": 0.18452143900172574,
+      "grad_norm": 0.3407264654457628,
+      "learning_rate": 4.9913745216012735e-06,
+      "loss": 0.6777,
+      "step": 695
+    },
+    {
+      "epoch": 0.18478693747510952,
+      "grad_norm": 0.345545417846684,
+      "learning_rate": 4.991345520786248e-06,
+      "loss": 0.6874,
+      "step": 696
+    },
+    {
+      "epoch": 0.1850524359484933,
+      "grad_norm": 0.34674174814963427,
+      "learning_rate": 4.991316471383885e-06,
+      "loss": 0.6467,
+      "step": 697
+    },
+    {
+      "epoch": 0.18531793442187708,
+      "grad_norm": 0.3454970518855923,
+      "learning_rate": 4.99128737339475e-06,
+      "loss": 0.6797,
+      "step": 698
+    },
+    {
+      "epoch": 0.18558343289526086,
+      "grad_norm": 0.33639062026906313,
+      "learning_rate": 4.991258226819412e-06,
+      "loss": 0.6562,
+      "step": 699
+    },
+    {
+      "epoch": 0.18584893136864464,
+      "grad_norm": 0.35528926257682564,
+      "learning_rate": 4.991229031658439e-06,
+      "loss": 0.5918,
+      "step": 700
+    },
+    {
+      "epoch": 0.18611442984202842,
+      "grad_norm": 0.361808065424919,
+      "learning_rate": 4.9911997879124e-06,
+      "loss": 0.675,
+      "step": 701
+    },
+    {
+      "epoch": 0.1863799283154122,
+      "grad_norm": 0.3459245861361517,
+      "learning_rate": 4.991170495581866e-06,
+      "loss": 0.682,
+      "step": 702
+    },
+    {
+      "epoch": 0.18664542678879598,
+      "grad_norm": 0.3462120423392343,
+      "learning_rate": 4.991141154667408e-06,
+      "loss": 0.6423,
+      "step": 703
+    },
+    {
+      "epoch": 0.18691092526217976,
+      "grad_norm": 0.3436493823715349,
+      "learning_rate": 4.991111765169599e-06,
+      "loss": 0.6518,
+      "step": 704
+    },
+    {
+      "epoch": 0.1871764237355635,
+      "grad_norm": 0.35167741744699027,
+      "learning_rate": 4.99108232708901e-06,
+      "loss": 0.6606,
+      "step": 705
+    },
+    {
+      "epoch": 0.1874419222089473,
+      "grad_norm": 0.34327635583975075,
+      "learning_rate": 4.991052840426216e-06,
+      "loss": 0.639,
+      "step": 706
+    },
+    {
+      "epoch": 0.18770742068233107,
+      "grad_norm": 0.3453640625141677,
+      "learning_rate": 4.991023305181793e-06,
+      "loss": 0.6354,
+      "step": 707
+    },
+    {
+      "epoch": 0.18797291915571485,
+      "grad_norm": 0.3567185085773874,
+      "learning_rate": 4.990993721356317e-06,
+      "loss": 0.6372,
+      "step": 708
+    },
+    {
+      "epoch": 0.18823841762909863,
+      "grad_norm": 0.364478705086117,
+      "learning_rate": 4.990964088950364e-06,
+      "loss": 0.6461,
+      "step": 709
+    },
+    {
+      "epoch": 0.1885039161024824,
+      "grad_norm": 0.3431707251325505,
+      "learning_rate": 4.990934407964511e-06,
+      "loss": 0.6308,
+      "step": 710
+    },
+    {
+      "epoch": 0.18876941457586618,
+      "grad_norm": 0.3661209655624326,
+      "learning_rate": 4.990904678399341e-06,
+      "loss": 0.6622,
+      "step": 711
+    },
+    {
+      "epoch": 0.18903491304924996,
+      "grad_norm": 0.3431422937762494,
+      "learning_rate": 4.9908749002554295e-06,
+      "loss": 0.6791,
+      "step": 712
+    },
+    {
+      "epoch": 0.18930041152263374,
+      "grad_norm": 0.35905228630684316,
+      "learning_rate": 4.990845073533358e-06,
+      "loss": 0.68,
+      "step": 713
+    },
+    {
+      "epoch": 0.18956590999601752,
+      "grad_norm": 0.3461910167092669,
+      "learning_rate": 4.99081519823371e-06,
+      "loss": 0.6636,
+      "step": 714
+    },
+    {
+      "epoch": 0.1898314084694013,
+      "grad_norm": 0.3386006082998523,
+      "learning_rate": 4.990785274357067e-06,
+      "loss": 0.6434,
+      "step": 715
+    },
+    {
+      "epoch": 0.19009690694278508,
+      "grad_norm": 0.3453548364134816,
+      "learning_rate": 4.990755301904012e-06,
+      "loss": 0.6684,
+      "step": 716
+    },
+    {
+      "epoch": 0.19036240541616886,
+      "grad_norm": 0.35014637101935864,
+      "learning_rate": 4.9907252808751304e-06,
+      "loss": 0.6771,
+      "step": 717
+    },
+    {
+      "epoch": 0.19062790388955264,
+      "grad_norm": 0.35004893115165964,
+      "learning_rate": 4.990695211271008e-06,
+      "loss": 0.6489,
+      "step": 718
+    },
+    {
+      "epoch": 0.19089340236293642,
+      "grad_norm": 0.342933936524258,
+      "learning_rate": 4.990665093092231e-06,
+      "loss": 0.6549,
+      "step": 719
+    },
+    {
+      "epoch": 0.1911589008363202,
+      "grad_norm": 0.3637709812659784,
+      "learning_rate": 4.990634926339385e-06,
+      "loss": 0.6753,
+      "step": 720
+    },
+    {
+      "epoch": 0.19142439930970398,
+      "grad_norm": 0.34613321478661163,
+      "learning_rate": 4.99060471101306e-06,
+      "loss": 0.644,
+      "step": 721
+    },
+    {
+      "epoch": 0.19168989778308776,
+      "grad_norm": 0.36246048423430655,
+      "learning_rate": 4.990574447113847e-06,
+      "loss": 0.6373,
+      "step": 722
+    },
+    {
+      "epoch": 0.19195539625647154,
+      "grad_norm": 0.3430804737712154,
+      "learning_rate": 4.990544134642333e-06,
+      "loss": 0.6776,
+      "step": 723
+    },
+    {
+      "epoch": 0.1922208947298553,
+      "grad_norm": 0.3704339499448267,
+      "learning_rate": 4.9905137735991104e-06,
+      "loss": 0.6225,
+      "step": 724
+    },
+    {
+      "epoch": 0.19248639320323907,
+      "grad_norm": 0.3585600683814105,
+      "learning_rate": 4.990483363984771e-06,
+      "loss": 0.6703,
+      "step": 725
+    },
+    {
+      "epoch": 0.19275189167662285,
+      "grad_norm": 0.35002134355805975,
+      "learning_rate": 4.990452905799909e-06,
+      "loss": 0.6273,
+      "step": 726
+    },
+    {
+      "epoch": 0.19301739015000663,
+      "grad_norm": 0.358388032460886,
+      "learning_rate": 4.990422399045117e-06,
+      "loss": 0.6375,
+      "step": 727
+    },
+    {
+      "epoch": 0.1932828886233904,
+      "grad_norm": 0.34611033141067893,
+      "learning_rate": 4.990391843720992e-06,
+      "loss": 0.6412,
+      "step": 728
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 0.3553189733335606,
+      "learning_rate": 4.990361239828127e-06,
+      "loss": 0.6648,
+      "step": 729
+    },
+    {
+      "epoch": 0.19381388557015797,
+      "grad_norm": 0.340952722084266,
+      "learning_rate": 4.990330587367121e-06,
+      "loss": 0.6608,
+      "step": 730
+    },
+    {
+      "epoch": 0.19407938404354175,
+      "grad_norm": 0.43374978272153053,
+      "learning_rate": 4.990299886338571e-06,
+      "loss": 0.6073,
+      "step": 731
+    },
+    {
+      "epoch": 0.19434488251692553,
+      "grad_norm": 0.33795144738861543,
+      "learning_rate": 4.990269136743076e-06,
+      "loss": 0.6459,
+      "step": 732
+    },
+    {
+      "epoch": 0.1946103809903093,
+      "grad_norm": 0.37392102483203854,
+      "learning_rate": 4.990238338581236e-06,
+      "loss": 0.699,
+      "step": 733
+    },
+    {
+      "epoch": 0.19487587946369309,
+      "grad_norm": 0.3760566762842307,
+      "learning_rate": 4.99020749185365e-06,
+      "loss": 0.662,
+      "step": 734
+    },
+    {
+      "epoch": 0.19514137793707687,
+      "grad_norm": 0.34528152011058866,
+      "learning_rate": 4.9901765965609225e-06,
+      "loss": 0.6681,
+      "step": 735
+    },
+    {
+      "epoch": 0.19540687641046064,
+      "grad_norm": 0.344179669061022,
+      "learning_rate": 4.990145652703653e-06,
+      "loss": 0.6497,
+      "step": 736
+    },
+    {
+      "epoch": 0.19567237488384442,
+      "grad_norm": 0.34965044920948035,
+      "learning_rate": 4.990114660282447e-06,
+      "loss": 0.6755,
+      "step": 737
+    },
+    {
+      "epoch": 0.1959378733572282,
+      "grad_norm": 0.3535454904841323,
+      "learning_rate": 4.990083619297908e-06,
+      "loss": 0.6609,
+      "step": 738
+    },
+    {
+      "epoch": 0.19620337183061198,
+      "grad_norm": 0.34515881193454195,
+      "learning_rate": 4.990052529750642e-06,
+      "loss": 0.6535,
+      "step": 739
+    },
+    {
+      "epoch": 0.19646887030399576,
+      "grad_norm": 0.35290041799419597,
+      "learning_rate": 4.990021391641255e-06,
+      "loss": 0.6688,
+      "step": 740
+    },
+    {
+      "epoch": 0.19673436877737954,
+      "grad_norm": 0.35948043134213553,
+      "learning_rate": 4.989990204970354e-06,
+      "loss": 0.662,
+      "step": 741
+    },
+    {
+      "epoch": 0.19699986725076332,
+      "grad_norm": 0.34688437244572223,
+      "learning_rate": 4.989958969738548e-06,
+      "loss": 0.6561,
+      "step": 742
+    },
+    {
+      "epoch": 0.19726536572414707,
+      "grad_norm": 0.3423806942407409,
+      "learning_rate": 4.989927685946444e-06,
+      "loss": 0.6564,
+      "step": 743
+    },
+    {
+      "epoch": 0.19753086419753085,
+      "grad_norm": 0.3671254216720733,
+      "learning_rate": 4.989896353594655e-06,
+      "loss": 0.6922,
+      "step": 744
+    },
+    {
+      "epoch": 0.19779636267091463,
+      "grad_norm": 0.3426208067836131,
+      "learning_rate": 4.989864972683791e-06,
+      "loss": 0.6478,
+      "step": 745
+    },
+    {
+      "epoch": 0.1980618611442984,
+      "grad_norm": 0.37198162066372836,
+      "learning_rate": 4.989833543214463e-06,
+      "loss": 0.6948,
+      "step": 746
+    },
+    {
+      "epoch": 0.1983273596176822,
+      "grad_norm": 0.34879592548399513,
+      "learning_rate": 4.989802065187285e-06,
+      "loss": 0.664,
+      "step": 747
+    },
+    {
+      "epoch": 0.19859285809106597,
+      "grad_norm": 0.34837159432268605,
+      "learning_rate": 4.989770538602871e-06,
+      "loss": 0.6731,
+      "step": 748
+    },
+    {
+      "epoch": 0.19885835656444975,
+      "grad_norm": 0.365626296978299,
+      "learning_rate": 4.989738963461835e-06,
+      "loss": 0.66,
+      "step": 749
+    },
+    {
+      "epoch": 0.19912385503783353,
+      "grad_norm": 0.3601140620745634,
+      "learning_rate": 4.989707339764794e-06,
+      "loss": 0.6238,
+      "step": 750
+    },
+    {
+      "epoch": 0.1993893535112173,
+      "grad_norm": 0.34474706919274484,
+      "learning_rate": 4.989675667512363e-06,
+      "loss": 0.6743,
+      "step": 751
+    },
+    {
+      "epoch": 0.1996548519846011,
+      "grad_norm": 0.3540948265947992,
+      "learning_rate": 4.989643946705161e-06,
+      "loss": 0.6437,
+      "step": 752
+    },
+    {
+      "epoch": 0.19992035045798487,
+      "grad_norm": 0.3569867242462055,
+      "learning_rate": 4.989612177343807e-06,
+      "loss": 0.6532,
+      "step": 753
+    },
+    {
+      "epoch": 0.20018584893136865,
+      "grad_norm": 0.3641200112332836,
+      "learning_rate": 4.98958035942892e-06,
+      "loss": 0.66,
+      "step": 754
+    },
+    {
+      "epoch": 0.20045134740475243,
+      "grad_norm": 0.3441067788826454,
+      "learning_rate": 4.98954849296112e-06,
+      "loss": 0.6535,
+      "step": 755
+    },
+    {
+      "epoch": 0.2007168458781362,
+      "grad_norm": 0.3451350430287015,
+      "learning_rate": 4.989516577941029e-06,
+      "loss": 0.6891,
+      "step": 756
+    },
+    {
+      "epoch": 0.20098234435152,
+      "grad_norm": 0.35665774669038647,
+      "learning_rate": 4.989484614369269e-06,
+      "loss": 0.6535,
+      "step": 757
+    },
+    {
+      "epoch": 0.20124784282490377,
+      "grad_norm": 0.3735976922013897,
+      "learning_rate": 4.989452602246464e-06,
+      "loss": 0.6074,
+      "step": 758
+    },
+    {
+      "epoch": 0.20151334129828755,
+      "grad_norm": 0.35283173086183617,
+      "learning_rate": 4.989420541573238e-06,
+      "loss": 0.6821,
+      "step": 759
+    },
+    {
+      "epoch": 0.20177883977167133,
+      "grad_norm": 0.34824192261422704,
+      "learning_rate": 4.989388432350216e-06,
+      "loss": 0.6458,
+      "step": 760
+    },
+    {
+      "epoch": 0.20204433824505508,
+      "grad_norm": 0.355889090178465,
+      "learning_rate": 4.989356274578025e-06,
+      "loss": 0.6517,
+      "step": 761
+    },
+    {
+      "epoch": 0.20230983671843886,
+      "grad_norm": 0.3392175251337885,
+      "learning_rate": 4.989324068257292e-06,
+      "loss": 0.6519,
+      "step": 762
+    },
+    {
+      "epoch": 0.20257533519182264,
+      "grad_norm": 0.3407751555251443,
+      "learning_rate": 4.989291813388644e-06,
+      "loss": 0.6643,
+      "step": 763
+    },
+    {
+      "epoch": 0.20284083366520642,
+      "grad_norm": 0.3504841174357293,
+      "learning_rate": 4.98925950997271e-06,
+      "loss": 0.6825,
+      "step": 764
+    },
+    {
+      "epoch": 0.2031063321385902,
+      "grad_norm": 0.35525779440856015,
+      "learning_rate": 4.989227158010123e-06,
+      "loss": 0.6477,
+      "step": 765
+    },
+    {
+      "epoch": 0.20337183061197397,
+      "grad_norm": 0.34858526803908774,
+      "learning_rate": 4.9891947575015095e-06,
+      "loss": 0.6434,
+      "step": 766
+    },
+    {
+      "epoch": 0.20363732908535775,
+      "grad_norm": 0.33648804934110765,
+      "learning_rate": 4.989162308447505e-06,
+      "loss": 0.6159,
+      "step": 767
+    },
+    {
+      "epoch": 0.20390282755874153,
+      "grad_norm": 0.3568903495343007,
+      "learning_rate": 4.98912981084874e-06,
+      "loss": 0.6558,
+      "step": 768
+    },
+    {
+      "epoch": 0.2041683260321253,
+      "grad_norm": 0.36416058886348546,
+      "learning_rate": 4.98909726470585e-06,
+      "loss": 0.6623,
+      "step": 769
+    },
+    {
+      "epoch": 0.2044338245055091,
+      "grad_norm": 0.35405775315654897,
+      "learning_rate": 4.989064670019469e-06,
+      "loss": 0.6569,
+      "step": 770
+    },
+    {
+      "epoch": 0.20469932297889287,
+      "grad_norm": 0.35255045795860585,
+      "learning_rate": 4.9890320267902325e-06,
+      "loss": 0.6332,
+      "step": 771
+    },
+    {
+      "epoch": 0.20496482145227665,
+      "grad_norm": 0.3606279396471135,
+      "learning_rate": 4.988999335018777e-06,
+      "loss": 0.6518,
+      "step": 772
+    },
+    {
+      "epoch": 0.20523031992566043,
+      "grad_norm": 0.3476308603367875,
+      "learning_rate": 4.988966594705741e-06,
+      "loss": 0.6659,
+      "step": 773
+    },
+    {
+      "epoch": 0.2054958183990442,
+      "grad_norm": 0.3425688513013799,
+      "learning_rate": 4.988933805851762e-06,
+      "loss": 0.6932,
+      "step": 774
+    },
+    {
+      "epoch": 0.205761316872428,
+      "grad_norm": 0.3768492089354647,
+      "learning_rate": 4.98890096845748e-06,
+      "loss": 0.6548,
+      "step": 775
+    },
+    {
+      "epoch": 0.20602681534581177,
+      "grad_norm": 0.34867139824243465,
+      "learning_rate": 4.988868082523535e-06,
+      "loss": 0.6545,
+      "step": 776
+    },
+    {
+      "epoch": 0.20629231381919555,
+      "grad_norm": 0.3618989818520878,
+      "learning_rate": 4.988835148050568e-06,
+      "loss": 0.6257,
+      "step": 777
+    },
+    {
+      "epoch": 0.20655781229257933,
+      "grad_norm": 0.35329490061661384,
+      "learning_rate": 4.988802165039222e-06,
+      "loss": 0.6504,
+      "step": 778
+    },
+    {
+      "epoch": 0.2068233107659631,
+      "grad_norm": 0.34555989750148103,
+      "learning_rate": 4.988769133490141e-06,
+      "loss": 0.6103,
+      "step": 779
+    },
+    {
+      "epoch": 0.20708880923934686,
+      "grad_norm": 0.3431675375187069,
+      "learning_rate": 4.988736053403968e-06,
+      "loss": 0.6251,
+      "step": 780
+    },
+    {
+      "epoch": 0.20735430771273064,
+      "grad_norm": 0.3555657917148854,
+      "learning_rate": 4.988702924781349e-06,
+      "loss": 0.6636,
+      "step": 781
+    },
+    {
+      "epoch": 0.20761980618611442,
+      "grad_norm": 0.36124070993240076,
+      "learning_rate": 4.988669747622929e-06,
+      "loss": 0.6601,
+      "step": 782
+    },
+    {
+      "epoch": 0.2078853046594982,
+      "grad_norm": 0.3477640473044727,
+      "learning_rate": 4.988636521929355e-06,
+      "loss": 0.6304,
+      "step": 783
+    },
+    {
+      "epoch": 0.20815080313288198,
+      "grad_norm": 0.35789720292635246,
+      "learning_rate": 4.988603247701277e-06,
+      "loss": 0.6639,
+      "step": 784
+    },
+    {
+      "epoch": 0.20841630160626576,
+      "grad_norm": 0.35820990851418305,
+      "learning_rate": 4.9885699249393415e-06,
+      "loss": 0.631,
+      "step": 785
+    },
+    {
+      "epoch": 0.20868180007964954,
+      "grad_norm": 0.3823713400187527,
+      "learning_rate": 4.9885365536442e-06,
+      "loss": 0.6304,
+      "step": 786
+    },
+    {
+      "epoch": 0.20894729855303332,
+      "grad_norm": 0.3506296950107694,
+      "learning_rate": 4.9885031338165025e-06,
+      "loss": 0.6921,
+      "step": 787
+    },
+    {
+      "epoch": 0.2092127970264171,
+      "grad_norm": 0.3870833129338183,
+      "learning_rate": 4.988469665456901e-06,
+      "loss": 0.6486,
+      "step": 788
+    },
+    {
+      "epoch": 0.20947829549980088,
+      "grad_norm": 0.347488569930697,
+      "learning_rate": 4.988436148566047e-06,
+      "loss": 0.691,
+      "step": 789
+    },
+    {
+      "epoch": 0.20974379397318466,
+      "grad_norm": 0.358713841867121,
+      "learning_rate": 4.988402583144597e-06,
+      "loss": 0.668,
+      "step": 790
+    },
+    {
+      "epoch": 0.21000929244656844,
+      "grad_norm": 0.35299913733330485,
+      "learning_rate": 4.988368969193204e-06,
+      "loss": 0.64,
+      "step": 791
+    },
+    {
+      "epoch": 0.21027479091995221,
+      "grad_norm": 0.34859776950733684,
+      "learning_rate": 4.9883353067125225e-06,
+      "loss": 0.6553,
+      "step": 792
+    },
+    {
+      "epoch": 0.210540289393336,
+      "grad_norm": 0.34686055149021544,
+      "learning_rate": 4.98830159570321e-06,
+      "loss": 0.6258,
+      "step": 793
+    },
+    {
+      "epoch": 0.21080578786671977,
+      "grad_norm": 0.35949850530490796,
+      "learning_rate": 4.988267836165924e-06,
+      "loss": 0.6669,
+      "step": 794
+    },
+    {
+      "epoch": 0.21107128634010355,
+      "grad_norm": 0.35366808769806707,
+      "learning_rate": 4.988234028101324e-06,
+      "loss": 0.6593,
+      "step": 795
+    },
+    {
+      "epoch": 0.21133678481348733,
+      "grad_norm": 0.3547574223059748,
+      "learning_rate": 4.988200171510067e-06,
+      "loss": 0.6633,
+      "step": 796
+    },
+    {
+      "epoch": 0.2116022832868711,
+      "grad_norm": 0.34981892699610107,
+      "learning_rate": 4.988166266392815e-06,
+      "loss": 0.6665,
+      "step": 797
+    },
+    {
+      "epoch": 0.2118677817602549,
+      "grad_norm": 0.34759057078754807,
+      "learning_rate": 4.988132312750229e-06,
+      "loss": 0.6492,
+      "step": 798
+    },
+    {
+      "epoch": 0.21213328023363864,
+      "grad_norm": 0.3526274083157686,
+      "learning_rate": 4.98809831058297e-06,
+      "loss": 0.6691,
+      "step": 799
+    },
+    {
+      "epoch": 0.21239877870702242,
+      "grad_norm": 0.36962844118627614,
+      "learning_rate": 4.988064259891703e-06,
+      "loss": 0.6573,
+      "step": 800
+    },
+    {
+      "epoch": 0.2126642771804062,
+      "grad_norm": 0.3495562841744903,
+      "learning_rate": 4.988030160677091e-06,
+      "loss": 0.6276,
+      "step": 801
+    },
+    {
+      "epoch": 0.21292977565378998,
+      "grad_norm": 0.35906869205672887,
+      "learning_rate": 4.987996012939799e-06,
+      "loss": 0.6419,
+      "step": 802
+    },
+    {
+      "epoch": 0.21319527412717376,
+      "grad_norm": 0.3545749367955905,
+      "learning_rate": 4.987961816680493e-06,
+      "loss": 0.6626,
+      "step": 803
+    },
+    {
+      "epoch": 0.21346077260055754,
+      "grad_norm": 0.35503241613586267,
+      "learning_rate": 4.98792757189984e-06,
+      "loss": 0.6307,
+      "step": 804
+    },
+    {
+      "epoch": 0.21372627107394132,
+      "grad_norm": 0.3352386894697989,
+      "learning_rate": 4.987893278598507e-06,
+      "loss": 0.6316,
+      "step": 805
+    },
+    {
+      "epoch": 0.2139917695473251,
+      "grad_norm": 0.35190578272262724,
+      "learning_rate": 4.987858936777165e-06,
+      "loss": 0.6156,
+      "step": 806
+    },
+    {
+      "epoch": 0.21425726802070888,
+      "grad_norm": 0.3513033240075226,
+      "learning_rate": 4.987824546436482e-06,
+      "loss": 0.6549,
+      "step": 807
+    },
+    {
+      "epoch": 0.21452276649409266,
+      "grad_norm": 0.34204140216555323,
+      "learning_rate": 4.987790107577128e-06,
+      "loss": 0.641,
+      "step": 808
+    },
+    {
+      "epoch": 0.21478826496747644,
+      "grad_norm": 0.3490871460559712,
+      "learning_rate": 4.987755620199776e-06,
+      "loss": 0.6796,
+      "step": 809
+    },
+    {
+      "epoch": 0.21505376344086022,
+      "grad_norm": 0.35347875353423974,
+      "learning_rate": 4.987721084305101e-06,
+      "loss": 0.5716,
+      "step": 810
+    },
+    {
+      "epoch": 0.215319261914244,
+      "grad_norm": 0.3483940521143135,
+      "learning_rate": 4.987686499893771e-06,
+      "loss": 0.6487,
+      "step": 811
+    },
+    {
+      "epoch": 0.21558476038762778,
+      "grad_norm": 0.37216731718694335,
+      "learning_rate": 4.987651866966464e-06,
+      "loss": 0.649,
+      "step": 812
+    },
+    {
+      "epoch": 0.21585025886101156,
+      "grad_norm": 0.3581584548987224,
+      "learning_rate": 4.9876171855238554e-06,
+      "loss": 0.6365,
+      "step": 813
+    },
+    {
+      "epoch": 0.21611575733439534,
+      "grad_norm": 0.3596825816868976,
+      "learning_rate": 4.98758245556662e-06,
+      "loss": 0.648,
+      "step": 814
+    },
+    {
+      "epoch": 0.21638125580777912,
+      "grad_norm": 0.3480720817879569,
+      "learning_rate": 4.987547677095437e-06,
+      "loss": 0.6254,
+      "step": 815
+    },
+    {
+      "epoch": 0.2166467542811629,
+      "grad_norm": 0.3693355122012089,
+      "learning_rate": 4.987512850110983e-06,
+      "loss": 0.6331,
+      "step": 816
+    },
+    {
+      "epoch": 0.21691225275454667,
+      "grad_norm": 0.3603395444162195,
+      "learning_rate": 4.987477974613939e-06,
+      "loss": 0.6283,
+      "step": 817
+    },
+    {
+      "epoch": 0.21717775122793043,
+      "grad_norm": 0.3737800526817902,
+      "learning_rate": 4.987443050604983e-06,
+      "loss": 0.6747,
+      "step": 818
+    },
+    {
+      "epoch": 0.2174432497013142,
+      "grad_norm": 0.3440344614094484,
+      "learning_rate": 4.987408078084797e-06,
+      "loss": 0.6304,
+      "step": 819
+    },
+    {
+      "epoch": 0.21770874817469799,
+      "grad_norm": 0.3690454107199134,
+      "learning_rate": 4.987373057054064e-06,
+      "loss": 0.6687,
+      "step": 820
+    },
+    {
+      "epoch": 0.21797424664808177,
+      "grad_norm": 0.3678878059485673,
+      "learning_rate": 4.987337987513467e-06,
+      "loss": 0.605,
+      "step": 821
+    },
+    {
+      "epoch": 0.21823974512146554,
+      "grad_norm": 0.34335471426275843,
+      "learning_rate": 4.987302869463687e-06,
+      "loss": 0.6352,
+      "step": 822
+    },
+    {
+      "epoch": 0.21850524359484932,
+      "grad_norm": 0.37952789084664285,
+      "learning_rate": 4.987267702905413e-06,
+      "loss": 0.6591,
+      "step": 823
+    },
+    {
+      "epoch": 0.2187707420682331,
+      "grad_norm": 0.340934874626806,
+      "learning_rate": 4.987232487839328e-06,
+      "loss": 0.6322,
+      "step": 824
+    },
+    {
+      "epoch": 0.21903624054161688,
+      "grad_norm": 0.3592082283497444,
+      "learning_rate": 4.9871972242661195e-06,
+      "loss": 0.6554,
+      "step": 825
+    },
+    {
+      "epoch": 0.21930173901500066,
+      "grad_norm": 0.35312258794824203,
+      "learning_rate": 4.987161912186476e-06,
+      "loss": 0.6844,
+      "step": 826
+    },
+    {
+      "epoch": 0.21956723748838444,
+      "grad_norm": 0.39253561740637266,
+      "learning_rate": 4.9871265516010854e-06,
+      "loss": 0.6342,
+      "step": 827
+    },
+    {
+      "epoch": 0.21983273596176822,
+      "grad_norm": 0.37013801640971983,
+      "learning_rate": 4.9870911425106375e-06,
+      "loss": 0.6448,
+      "step": 828
+    },
+    {
+      "epoch": 0.220098234435152,
+      "grad_norm": 0.37332415971943467,
+      "learning_rate": 4.987055684915824e-06,
+      "loss": 0.6408,
+      "step": 829
+    },
+    {
+      "epoch": 0.22036373290853578,
+      "grad_norm": 0.3391307278832495,
+      "learning_rate": 4.9870201788173334e-06,
+      "loss": 0.6455,
+      "step": 830
+    },
+    {
+      "epoch": 0.22062923138191956,
+      "grad_norm": 0.3462273975496499,
+      "learning_rate": 4.986984624215861e-06,
+      "loss": 0.6425,
+      "step": 831
+    },
+    {
+      "epoch": 0.22089472985530334,
+      "grad_norm": 0.3546835803282198,
+      "learning_rate": 4.986949021112098e-06,
+      "loss": 0.6414,
+      "step": 832
+    },
+    {
+      "epoch": 0.22116022832868712,
+      "grad_norm": 0.35926255487376796,
+      "learning_rate": 4.986913369506742e-06,
+      "loss": 0.6453,
+      "step": 833
+    },
+    {
+      "epoch": 0.2214257268020709,
+      "grad_norm": 0.346809525634212,
+      "learning_rate": 4.986877669400485e-06,
+      "loss": 0.6378,
+      "step": 834
+    },
+    {
+      "epoch": 0.22169122527545468,
+      "grad_norm": 0.3623921896573435,
+      "learning_rate": 4.9868419207940255e-06,
+      "loss": 0.7046,
+      "step": 835
+    },
+    {
+      "epoch": 0.22195672374883843,
+      "grad_norm": 0.3633208444851114,
+      "learning_rate": 4.986806123688059e-06,
+      "loss": 0.6769,
+      "step": 836
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.35061516901527795,
+      "learning_rate": 4.986770278083284e-06,
+      "loss": 0.6449,
+      "step": 837
+    },
+    {
+      "epoch": 0.222487720695606,
+      "grad_norm": 0.3571100916897149,
+      "learning_rate": 4.986734383980401e-06,
+      "loss": 0.6548,
+      "step": 838
+    },
+    {
+      "epoch": 0.22275321916898977,
+      "grad_norm": 0.3632731411894179,
+      "learning_rate": 4.9866984413801075e-06,
+      "loss": 0.6793,
+      "step": 839
+    },
+    {
+      "epoch": 0.22301871764237355,
+      "grad_norm": 0.3657427335552278,
+      "learning_rate": 4.986662450283107e-06,
+      "loss": 0.6692,
+      "step": 840
+    },
+    {
+      "epoch": 0.22328421611575733,
+      "grad_norm": 0.34989572876436803,
+      "learning_rate": 4.9866264106900995e-06,
+      "loss": 0.6341,
+      "step": 841
+    },
+    {
+      "epoch": 0.2235497145891411,
+      "grad_norm": 0.35678405687180387,
+      "learning_rate": 4.98659032260179e-06,
+      "loss": 0.6096,
+      "step": 842
+    },
+    {
+      "epoch": 0.2238152130625249,
+      "grad_norm": 0.34616644977846645,
+      "learning_rate": 4.986554186018879e-06,
+      "loss": 0.6152,
+      "step": 843
+    },
+    {
+      "epoch": 0.22408071153590867,
+      "grad_norm": 0.3472065104355087,
+      "learning_rate": 4.986518000942075e-06,
+      "loss": 0.6326,
+      "step": 844
+    },
+    {
+      "epoch": 0.22434621000929245,
+      "grad_norm": 0.3429874846006786,
+      "learning_rate": 4.986481767372081e-06,
+      "loss": 0.6527,
+      "step": 845
+    },
+    {
+      "epoch": 0.22461170848267623,
+      "grad_norm": 0.3397592737642725,
+      "learning_rate": 4.986445485309606e-06,
+      "loss": 0.6397,
+      "step": 846
+    },
+    {
+      "epoch": 0.22487720695606,
+      "grad_norm": 0.35275167602090624,
+      "learning_rate": 4.986409154755354e-06,
+      "loss": 0.6615,
+      "step": 847
+    },
+    {
+      "epoch": 0.22514270542944378,
+      "grad_norm": 0.3661359137649159,
+      "learning_rate": 4.9863727757100375e-06,
+      "loss": 0.6733,
+      "step": 848
+    },
+    {
+      "epoch": 0.22540820390282756,
+      "grad_norm": 0.3590313219117273,
+      "learning_rate": 4.986336348174364e-06,
+      "loss": 0.6185,
+      "step": 849
+    },
+    {
+      "epoch": 0.22567370237621134,
+      "grad_norm": 0.37040577861478885,
+      "learning_rate": 4.986299872149043e-06,
+      "loss": 0.6885,
+      "step": 850
+    },
+    {
+      "epoch": 0.22593920084959512,
+      "grad_norm": 0.3522247074086537,
+      "learning_rate": 4.986263347634787e-06,
+      "loss": 0.6409,
+      "step": 851
+    },
+    {
+      "epoch": 0.2262046993229789,
+      "grad_norm": 0.3626158025539011,
+      "learning_rate": 4.986226774632309e-06,
+      "loss": 0.6617,
+      "step": 852
+    },
+    {
+      "epoch": 0.22647019779636268,
+      "grad_norm": 0.3619975206918967,
+      "learning_rate": 4.986190153142321e-06,
+      "loss": 0.6815,
+      "step": 853
+    },
+    {
+      "epoch": 0.22673569626974646,
+      "grad_norm": 0.36106204110325646,
+      "learning_rate": 4.986153483165539e-06,
+      "loss": 0.7225,
+      "step": 854
+    },
+    {
+      "epoch": 0.2270011947431302,
+      "grad_norm": 0.35954909376182703,
+      "learning_rate": 4.9861167647026755e-06,
+      "loss": 0.6373,
+      "step": 855
+    },
+    {
+      "epoch": 0.227266693216514,
+      "grad_norm": 0.3654694196099264,
+      "learning_rate": 4.986079997754449e-06,
+      "loss": 0.6383,
+      "step": 856
+    },
+    {
+      "epoch": 0.22753219168989777,
+      "grad_norm": 0.36021768240263363,
+      "learning_rate": 4.986043182321575e-06,
+      "loss": 0.6642,
+      "step": 857
+    },
+    {
+      "epoch": 0.22779769016328155,
+      "grad_norm": 0.36383655510780455,
+      "learning_rate": 4.986006318404772e-06,
+      "loss": 0.6248,
+      "step": 858
+    },
+    {
+      "epoch": 0.22806318863666533,
+      "grad_norm": 0.36525947002795617,
+      "learning_rate": 4.985969406004759e-06,
+      "loss": 0.6456,
+      "step": 859
+    },
+    {
+      "epoch": 0.2283286871100491,
+      "grad_norm": 0.3847121537010074,
+      "learning_rate": 4.985932445122257e-06,
+      "loss": 0.6528,
+      "step": 860
+    },
+    {
+      "epoch": 0.2285941855834329,
+      "grad_norm": 0.37182492589245225,
+      "learning_rate": 4.985895435757985e-06,
+      "loss": 0.6892,
+      "step": 861
+    },
+    {
+      "epoch": 0.22885968405681667,
+      "grad_norm": 0.3712810265763539,
+      "learning_rate": 4.9858583779126645e-06,
+      "loss": 0.6591,
+      "step": 862
+    },
+    {
+      "epoch": 0.22912518253020045,
+      "grad_norm": 0.36232619618674444,
+      "learning_rate": 4.9858212715870205e-06,
+      "loss": 0.6492,
+      "step": 863
+    },
+    {
+      "epoch": 0.22939068100358423,
+      "grad_norm": 0.3725415819789804,
+      "learning_rate": 4.985784116781774e-06,
+      "loss": 0.6659,
+      "step": 864
+    },
+    {
+      "epoch": 0.229656179476968,
+      "grad_norm": 0.3413921760507849,
+      "learning_rate": 4.985746913497652e-06,
+      "loss": 0.6554,
+      "step": 865
+    },
+    {
+      "epoch": 0.2299216779503518,
+      "grad_norm": 0.3836353768850874,
+      "learning_rate": 4.985709661735379e-06,
+      "loss": 0.6287,
+      "step": 866
+    },
+    {
+      "epoch": 0.23018717642373557,
+      "grad_norm": 0.36842816359691,
+      "learning_rate": 4.9856723614956815e-06,
+      "loss": 0.6662,
+      "step": 867
+    },
+    {
+      "epoch": 0.23045267489711935,
+      "grad_norm": 0.4069655588649439,
+      "learning_rate": 4.985635012779288e-06,
+      "loss": 0.6567,
+      "step": 868
+    },
+    {
+      "epoch": 0.23071817337050313,
+      "grad_norm": 0.3586906177489775,
+      "learning_rate": 4.985597615586924e-06,
+      "loss": 0.6457,
+      "step": 869
+    },
+    {
+      "epoch": 0.2309836718438869,
+      "grad_norm": 0.3690248569199568,
+      "learning_rate": 4.985560169919321e-06,
+      "loss": 0.6836,
+      "step": 870
+    },
+    {
+      "epoch": 0.23124917031727069,
+      "grad_norm": 0.3618169009342208,
+      "learning_rate": 4.9855226757772105e-06,
+      "loss": 0.7102,
+      "step": 871
+    },
+    {
+      "epoch": 0.23151466879065447,
+      "grad_norm": 0.45226740809634364,
+      "learning_rate": 4.985485133161321e-06,
+      "loss": 0.6265,
+      "step": 872
+    },
+    {
+      "epoch": 0.23178016726403824,
+      "grad_norm": 0.38149227214737613,
+      "learning_rate": 4.985447542072386e-06,
+      "loss": 0.6584,
+      "step": 873
+    },
+    {
+      "epoch": 0.232045665737422,
+      "grad_norm": 0.42137441504054346,
+      "learning_rate": 4.985409902511138e-06,
+      "loss": 0.6359,
+      "step": 874
+    },
+    {
+      "epoch": 0.23231116421080578,
+      "grad_norm": 0.3629050435358434,
+      "learning_rate": 4.985372214478313e-06,
+      "loss": 0.6712,
+      "step": 875
+    },
+    {
+      "epoch": 0.23257666268418956,
+      "grad_norm": 0.3797761774400396,
+      "learning_rate": 4.985334477974643e-06,
+      "loss": 0.6328,
+      "step": 876
+    },
+    {
+      "epoch": 0.23284216115757334,
+      "grad_norm": 0.3683308015560075,
+      "learning_rate": 4.985296693000866e-06,
+      "loss": 0.6919,
+      "step": 877
+    },
+    {
+      "epoch": 0.23310765963095711,
+      "grad_norm": 0.37672600814985063,
+      "learning_rate": 4.985258859557719e-06,
+      "loss": 0.6541,
+      "step": 878
+    },
+    {
+      "epoch": 0.2333731581043409,
+      "grad_norm": 0.3635422343734699,
+      "learning_rate": 4.985220977645939e-06,
+      "loss": 0.6449,
+      "step": 879
+    },
+    {
+      "epoch": 0.23363865657772467,
+      "grad_norm": 0.340922674674232,
+      "learning_rate": 4.985183047266265e-06,
+      "loss": 0.6319,
+      "step": 880
+    },
+    {
+      "epoch": 0.23390415505110845,
+      "grad_norm": 0.3490076534717787,
+      "learning_rate": 4.985145068419437e-06,
+      "loss": 0.6418,
+      "step": 881
+    },
+    {
+      "epoch": 0.23416965352449223,
+      "grad_norm": 0.3774554872420336,
+      "learning_rate": 4.9851070411061954e-06,
+      "loss": 0.6166,
+      "step": 882
+    },
+    {
+      "epoch": 0.234435151997876,
+      "grad_norm": 0.3570349216253811,
+      "learning_rate": 4.985068965327281e-06,
+      "loss": 0.6441,
+      "step": 883
+    },
+    {
+      "epoch": 0.2347006504712598,
+      "grad_norm": 0.36709361727630363,
+      "learning_rate": 4.985030841083439e-06,
+      "loss": 0.6676,
+      "step": 884
+    },
+    {
+      "epoch": 0.23496614894464357,
+      "grad_norm": 0.36607196029775896,
+      "learning_rate": 4.98499266837541e-06,
+      "loss": 0.6355,
+      "step": 885
+    },
+    {
+      "epoch": 0.23523164741802735,
+      "grad_norm": 0.373504831992073,
+      "learning_rate": 4.9849544472039394e-06,
+      "loss": 0.6617,
+      "step": 886
+    },
+    {
+      "epoch": 0.23549714589141113,
+      "grad_norm": 0.3573539490018993,
+      "learning_rate": 4.984916177569774e-06,
+      "loss": 0.6118,
+      "step": 887
+    },
+    {
+      "epoch": 0.2357626443647949,
+      "grad_norm": 0.35686848155853435,
+      "learning_rate": 4.984877859473658e-06,
+      "loss": 0.6522,
+      "step": 888
+    },
+    {
+      "epoch": 0.2360281428381787,
+      "grad_norm": 0.35608679872998955,
+      "learning_rate": 4.98483949291634e-06,
+      "loss": 0.6628,
+      "step": 889
+    },
+    {
+      "epoch": 0.23629364131156247,
+      "grad_norm": 0.3637007806136096,
+      "learning_rate": 4.984801077898568e-06,
+      "loss": 0.6506,
+      "step": 890
+    },
+    {
+      "epoch": 0.23655913978494625,
+      "grad_norm": 0.3545252456352333,
+      "learning_rate": 4.984762614421092e-06,
+      "loss": 0.6983,
+      "step": 891
+    },
+    {
+      "epoch": 0.23682463825833003,
+      "grad_norm": 0.36288115661550346,
+      "learning_rate": 4.98472410248466e-06,
+      "loss": 0.6717,
+      "step": 892
+    },
+    {
+      "epoch": 0.23709013673171378,
+      "grad_norm": 0.3626678375509858,
+      "learning_rate": 4.9846855420900245e-06,
+      "loss": 0.6132,
+      "step": 893
+    },
+    {
+      "epoch": 0.23735563520509756,
+      "grad_norm": 0.3484005456555979,
+      "learning_rate": 4.984646933237938e-06,
+      "loss": 0.6565,
+      "step": 894
+    },
+    {
+      "epoch": 0.23762113367848134,
+      "grad_norm": 0.34605002196912016,
+      "learning_rate": 4.984608275929153e-06,
+      "loss": 0.6166,
+      "step": 895
+    },
+    {
+      "epoch": 0.23788663215186512,
+      "grad_norm": 0.34730696193486404,
+      "learning_rate": 4.984569570164423e-06,
+      "loss": 0.6599,
+      "step": 896
+    },
+    {
+      "epoch": 0.2381521306252489,
+      "grad_norm": 0.353993737368934,
+      "learning_rate": 4.984530815944502e-06,
+      "loss": 0.6451,
+      "step": 897
+    },
+    {
+      "epoch": 0.23841762909863268,
+      "grad_norm": 0.3515130354430692,
+      "learning_rate": 4.984492013270148e-06,
+      "loss": 0.6616,
+      "step": 898
+    },
+    {
+      "epoch": 0.23868312757201646,
+      "grad_norm": 0.36635953389940396,
+      "learning_rate": 4.9844531621421165e-06,
+      "loss": 0.6187,
+      "step": 899
+    },
+    {
+      "epoch": 0.23894862604540024,
+      "grad_norm": 0.34987838664873055,
+      "learning_rate": 4.984414262561166e-06,
+      "loss": 0.6582,
+      "step": 900
+    },
+    {
+      "epoch": 0.23921412451878402,
+      "grad_norm": 0.36813857307007014,
+      "learning_rate": 4.984375314528052e-06,
+      "loss": 0.6537,
+      "step": 901
+    },
+    {
+      "epoch": 0.2394796229921678,
+      "grad_norm": 0.36019520717900866,
+      "learning_rate": 4.984336318043539e-06,
+      "loss": 0.6505,
+      "step": 902
+    },
+    {
+      "epoch": 0.23974512146555157,
+      "grad_norm": 0.35914257245516107,
+      "learning_rate": 4.984297273108385e-06,
+      "loss": 0.6441,
+      "step": 903
+    },
+    {
+      "epoch": 0.24001061993893535,
+      "grad_norm": 0.36124637472865423,
+      "learning_rate": 4.984258179723351e-06,
+      "loss": 0.6503,
+      "step": 904
+    },
+    {
+      "epoch": 0.24027611841231913,
+      "grad_norm": 0.37851229317535584,
+      "learning_rate": 4.9842190378892e-06,
+      "loss": 0.6443,
+      "step": 905
+    },
+    {
+      "epoch": 0.2405416168857029,
+      "grad_norm": 0.3473299876317927,
+      "learning_rate": 4.984179847606695e-06,
+      "loss": 0.6199,
+      "step": 906
+    },
+    {
+      "epoch": 0.2408071153590867,
+      "grad_norm": 0.3730397722899552,
+      "learning_rate": 4.9841406088766005e-06,
+      "loss": 0.6886,
+      "step": 907
+    },
+    {
+      "epoch": 0.24107261383247047,
+      "grad_norm": 0.3672165214242251,
+      "learning_rate": 4.9841013216996825e-06,
+      "loss": 0.6966,
+      "step": 908
+    },
+    {
+      "epoch": 0.24133811230585425,
+      "grad_norm": 0.3816250965958184,
+      "learning_rate": 4.984061986076707e-06,
+      "loss": 0.6443,
+      "step": 909
+    },
+    {
+      "epoch": 0.24160361077923803,
+      "grad_norm": 0.3638965681825285,
+      "learning_rate": 4.98402260200844e-06,
+      "loss": 0.6707,
+      "step": 910
+    },
+    {
+      "epoch": 0.2418691092526218,
+      "grad_norm": 0.3865422142358914,
+      "learning_rate": 4.983983169495651e-06,
+      "loss": 0.6047,
+      "step": 911
+    },
+    {
+      "epoch": 0.24213460772600556,
+      "grad_norm": 0.348430739502864,
+      "learning_rate": 4.983943688539108e-06,
+      "loss": 0.6376,
+      "step": 912
+    },
+    {
+      "epoch": 0.24240010619938934,
+      "grad_norm": 0.3440800363082338,
+      "learning_rate": 4.983904159139581e-06,
+      "loss": 0.6056,
+      "step": 913
+    },
+    {
+      "epoch": 0.24266560467277312,
+      "grad_norm": 0.3602285592115701,
+      "learning_rate": 4.983864581297841e-06,
+      "loss": 0.6562,
+      "step": 914
+    },
+    {
+      "epoch": 0.2429311031461569,
+      "grad_norm": 0.36508533233465384,
+      "learning_rate": 4.983824955014661e-06,
+      "loss": 0.7068,
+      "step": 915
+    },
+    {
+      "epoch": 0.24319660161954068,
+      "grad_norm": 0.37488541240290535,
+      "learning_rate": 4.983785280290812e-06,
+      "loss": 0.6669,
+      "step": 916
+    },
+    {
+      "epoch": 0.24346210009292446,
+      "grad_norm": 0.36745750796129384,
+      "learning_rate": 4.9837455571270696e-06,
+      "loss": 0.6554,
+      "step": 917
+    },
+    {
+      "epoch": 0.24372759856630824,
+      "grad_norm": 0.347919831926423,
+      "learning_rate": 4.983705785524207e-06,
+      "loss": 0.6433,
+      "step": 918
+    },
+    {
+      "epoch": 0.24399309703969202,
+      "grad_norm": 0.3645222567372667,
+      "learning_rate": 4.983665965483001e-06,
+      "loss": 0.623,
+      "step": 919
+    },
+    {
+      "epoch": 0.2442585955130758,
+      "grad_norm": 0.36734959172373594,
+      "learning_rate": 4.983626097004227e-06,
+      "loss": 0.677,
+      "step": 920
+    },
+    {
+      "epoch": 0.24452409398645958,
+      "grad_norm": 0.3597583035018811,
+      "learning_rate": 4.9835861800886635e-06,
+      "loss": 0.6362,
+      "step": 921
+    },
+    {
+      "epoch": 0.24478959245984336,
+      "grad_norm": 0.36723122963508614,
+      "learning_rate": 4.983546214737088e-06,
+      "loss": 0.6605,
+      "step": 922
+    },
+    {
+      "epoch": 0.24505509093322714,
+      "grad_norm": 0.3575193735159897,
+      "learning_rate": 4.9835062009502815e-06,
+      "loss": 0.6969,
+      "step": 923
+    },
+    {
+      "epoch": 0.24532058940661092,
+      "grad_norm": 0.3639065644092349,
+      "learning_rate": 4.983466138729023e-06,
+      "loss": 0.655,
+      "step": 924
+    },
+    {
+      "epoch": 0.2455860878799947,
+      "grad_norm": 0.3523416143981236,
+      "learning_rate": 4.983426028074095e-06,
+      "loss": 0.643,
+      "step": 925
+    },
+    {
+      "epoch": 0.24585158635337848,
+      "grad_norm": 0.36972737936054123,
+      "learning_rate": 4.983385868986277e-06,
+      "loss": 0.6543,
+      "step": 926
+    },
+    {
+      "epoch": 0.24611708482676226,
+      "grad_norm": 0.3628422006374738,
+      "learning_rate": 4.983345661466356e-06,
+      "loss": 0.6751,
+      "step": 927
+    },
+    {
+      "epoch": 0.24638258330014604,
+      "grad_norm": 0.378487498998904,
+      "learning_rate": 4.983305405515113e-06,
+      "loss": 0.6578,
+      "step": 928
+    },
+    {
+      "epoch": 0.24664808177352981,
+      "grad_norm": 0.35773346585319693,
+      "learning_rate": 4.983265101133335e-06,
+      "loss": 0.6425,
+      "step": 929
+    },
+    {
+      "epoch": 0.24691358024691357,
+      "grad_norm": 0.3567053830313769,
+      "learning_rate": 4.983224748321808e-06,
+      "loss": 0.6457,
+      "step": 930
+    },
+    {
+      "epoch": 0.24717907872029735,
+      "grad_norm": 0.35787311418723283,
+      "learning_rate": 4.983184347081318e-06,
+      "loss": 0.5949,
+      "step": 931
+    },
+    {
+      "epoch": 0.24744457719368113,
+      "grad_norm": 0.34942920908399994,
+      "learning_rate": 4.9831438974126525e-06,
+      "loss": 0.6474,
+      "step": 932
+    },
+    {
+      "epoch": 0.2477100756670649,
+      "grad_norm": 0.3726541359601728,
+      "learning_rate": 4.983103399316602e-06,
+      "loss": 0.6706,
+      "step": 933
+    },
+    {
+      "epoch": 0.24797557414044868,
+      "grad_norm": 0.4419698443655577,
+      "learning_rate": 4.983062852793956e-06,
+      "loss": 0.6362,
+      "step": 934
+    },
+    {
+      "epoch": 0.24824107261383246,
+      "grad_norm": 0.3730662683997837,
+      "learning_rate": 4.983022257845504e-06,
+      "loss": 0.6489,
+      "step": 935
+    },
+    {
+      "epoch": 0.24850657108721624,
+      "grad_norm": 0.39203459093004006,
+      "learning_rate": 4.982981614472039e-06,
+      "loss": 0.6303,
+      "step": 936
+    },
+    {
+      "epoch": 0.24877206956060002,
+      "grad_norm": 0.3746997194288708,
+      "learning_rate": 4.982940922674353e-06,
+      "loss": 0.6475,
+      "step": 937
+    },
+    {
+      "epoch": 0.2490375680339838,
+      "grad_norm": 0.3774628011618854,
+      "learning_rate": 4.98290018245324e-06,
+      "loss": 0.6794,
+      "step": 938
+    },
+    {
+      "epoch": 0.24930306650736758,
+      "grad_norm": 0.3851350044540439,
+      "learning_rate": 4.982859393809493e-06,
+      "loss": 0.6605,
+      "step": 939
+    },
+    {
+      "epoch": 0.24956856498075136,
+      "grad_norm": 0.36840943795916076,
+      "learning_rate": 4.98281855674391e-06,
+      "loss": 0.6229,
+      "step": 940
+    },
+    {
+      "epoch": 0.24983406345413514,
+      "grad_norm": 0.3552009871115495,
+      "learning_rate": 4.982777671257285e-06,
+      "loss": 0.5982,
+      "step": 941
+    },
+    {
+      "epoch": 0.2500995619275189,
+      "grad_norm": 0.3707382212266125,
+      "learning_rate": 4.982736737350418e-06,
+      "loss": 0.6646,
+      "step": 942
+    },
+    {
+      "epoch": 0.2503650604009027,
+      "grad_norm": 0.3543486718509332,
+      "learning_rate": 4.982695755024105e-06,
+      "loss": 0.6414,
+      "step": 943
+    },
+    {
+      "epoch": 0.2506305588742865,
+      "grad_norm": 0.38513378319201824,
+      "learning_rate": 4.982654724279146e-06,
+      "loss": 0.6485,
+      "step": 944
+    },
+    {
+      "epoch": 0.25089605734767023,
+      "grad_norm": 0.3636189188787609,
+      "learning_rate": 4.982613645116342e-06,
+      "loss": 0.6496,
+      "step": 945
+    },
+    {
+      "epoch": 0.25116155582105404,
+      "grad_norm": 0.36571575378837284,
+      "learning_rate": 4.9825725175364926e-06,
+      "loss": 0.6856,
+      "step": 946
+    },
+    {
+      "epoch": 0.2514270542944378,
+      "grad_norm": 0.3807450704812266,
+      "learning_rate": 4.9825313415404005e-06,
+      "loss": 0.6756,
+      "step": 947
+    },
+    {
+      "epoch": 0.2516925527678216,
+      "grad_norm": 0.36818288389234677,
+      "learning_rate": 4.98249011712887e-06,
+      "loss": 0.6881,
+      "step": 948
+    },
+    {
+      "epoch": 0.25195805124120535,
+      "grad_norm": 0.3549349496432932,
+      "learning_rate": 4.982448844302704e-06,
+      "loss": 0.6183,
+      "step": 949
+    },
+    {
+      "epoch": 0.25222354971458916,
+      "grad_norm": 0.36306646432366935,
+      "learning_rate": 4.982407523062707e-06,
+      "loss": 0.6808,
+      "step": 950
+    },
+    {
+      "epoch": 0.2524890481879729,
+      "grad_norm": 0.3394703612759071,
+      "learning_rate": 4.982366153409685e-06,
+      "loss": 0.6189,
+      "step": 951
+    },
+    {
+      "epoch": 0.2527545466613567,
+      "grad_norm": 0.3465033805123829,
+      "learning_rate": 4.982324735344446e-06,
+      "loss": 0.6086,
+      "step": 952
+    },
+    {
+      "epoch": 0.25302004513474047,
+      "grad_norm": 0.3614620012877914,
+      "learning_rate": 4.982283268867796e-06,
+      "loss": 0.6588,
+      "step": 953
+    },
+    {
+      "epoch": 0.2532855436081243,
+      "grad_norm": 0.3609814306439208,
+      "learning_rate": 4.9822417539805445e-06,
+      "loss": 0.6627,
+      "step": 954
+    },
+    {
+      "epoch": 0.253551042081508,
+      "grad_norm": 0.3664856599554348,
+      "learning_rate": 4.982200190683502e-06,
+      "loss": 0.6118,
+      "step": 955
+    },
+    {
+      "epoch": 0.25381654055489183,
+      "grad_norm": 0.3551510979475306,
+      "learning_rate": 4.9821585789774775e-06,
+      "loss": 0.6709,
+      "step": 956
+    },
+    {
+      "epoch": 0.2540820390282756,
+      "grad_norm": 0.3559223188494908,
+      "learning_rate": 4.982116918863285e-06,
+      "loss": 0.6481,
+      "step": 957
+    },
+    {
+      "epoch": 0.2543475375016594,
+      "grad_norm": 0.37500353329046965,
+      "learning_rate": 4.9820752103417335e-06,
+      "loss": 0.6362,
+      "step": 958
+    },
+    {
+      "epoch": 0.25461303597504314,
+      "grad_norm": 0.3468948169563887,
+      "learning_rate": 4.982033453413638e-06,
+      "loss": 0.6088,
+      "step": 959
+    },
+    {
+      "epoch": 0.2548785344484269,
+      "grad_norm": 0.3539323668269743,
+      "learning_rate": 4.981991648079814e-06,
+      "loss": 0.6213,
+      "step": 960
+    },
+    {
+      "epoch": 0.2551440329218107,
+      "grad_norm": 0.391904199579984,
+      "learning_rate": 4.981949794341077e-06,
+      "loss": 0.6606,
+      "step": 961
+    },
+    {
+      "epoch": 0.25540953139519446,
+      "grad_norm": 0.37079725181686624,
+      "learning_rate": 4.98190789219824e-06,
+      "loss": 0.6235,
+      "step": 962
+    },
+    {
+      "epoch": 0.25567502986857826,
+      "grad_norm": 0.3498087643698131,
+      "learning_rate": 4.981865941652124e-06,
+      "loss": 0.6164,
+      "step": 963
+    },
+    {
+      "epoch": 0.255940528341962,
+      "grad_norm": 0.37674344852061253,
+      "learning_rate": 4.981823942703544e-06,
+      "loss": 0.6377,
+      "step": 964
+    },
+    {
+      "epoch": 0.2562060268153458,
+      "grad_norm": 0.3659909389340991,
+      "learning_rate": 4.9817818953533215e-06,
+      "loss": 0.6428,
+      "step": 965
+    },
+    {
+      "epoch": 0.2564715252887296,
+      "grad_norm": 0.3686889799009494,
+      "learning_rate": 4.981739799602275e-06,
+      "loss": 0.6735,
+      "step": 966
+    },
+    {
+      "epoch": 0.2567370237621134,
+      "grad_norm": 0.3754154691933635,
+      "learning_rate": 4.981697655451228e-06,
+      "loss": 0.658,
+      "step": 967
+    },
+    {
+      "epoch": 0.25700252223549713,
+      "grad_norm": 0.3494085335346896,
+      "learning_rate": 4.981655462900998e-06,
+      "loss": 0.6506,
+      "step": 968
+    },
+    {
+      "epoch": 0.25726802070888094,
+      "grad_norm": 0.3617112473409996,
+      "learning_rate": 4.981613221952412e-06,
+      "loss": 0.6292,
+      "step": 969
+    },
+    {
+      "epoch": 0.2575335191822647,
+      "grad_norm": 0.35222431431476064,
+      "learning_rate": 4.981570932606291e-06,
+      "loss": 0.6287,
+      "step": 970
+    },
+    {
+      "epoch": 0.2577990176556485,
+      "grad_norm": 0.358091030551002,
+      "learning_rate": 4.981528594863461e-06,
+      "loss": 0.6463,
+      "step": 971
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.35783571511857615,
+      "learning_rate": 4.981486208724748e-06,
+      "loss": 0.6451,
+      "step": 972
+    },
+    {
+      "epoch": 0.25833001460241606,
+      "grad_norm": 0.3376080650157559,
+      "learning_rate": 4.981443774190978e-06,
+      "loss": 0.6119,
+      "step": 973
+    },
+    {
+      "epoch": 0.2585955130757998,
+      "grad_norm": 0.37354453677979915,
+      "learning_rate": 4.981401291262979e-06,
+      "loss": 0.6466,
+      "step": 974
+    },
+    {
+      "epoch": 0.2588610115491836,
+      "grad_norm": 0.3585901221168928,
+      "learning_rate": 4.981358759941579e-06,
+      "loss": 0.625,
+      "step": 975
+    },
+    {
+      "epoch": 0.25912651002256737,
+      "grad_norm": 0.3606103589636798,
+      "learning_rate": 4.981316180227609e-06,
+      "loss": 0.6434,
+      "step": 976
+    },
+    {
+      "epoch": 0.2593920084959511,
+      "grad_norm": 0.3568169059932706,
+      "learning_rate": 4.981273552121896e-06,
+      "loss": 0.6422,
+      "step": 977
+    },
+    {
+      "epoch": 0.25965750696933493,
+      "grad_norm": 0.3612424596771187,
+      "learning_rate": 4.981230875625275e-06,
+      "loss": 0.6385,
+      "step": 978
+    },
+    {
+      "epoch": 0.2599230054427187,
+      "grad_norm": 0.3654440661113346,
+      "learning_rate": 4.9811881507385766e-06,
+      "loss": 0.6624,
+      "step": 979
+    },
+    {
+      "epoch": 0.2601885039161025,
+      "grad_norm": 0.3744325849965256,
+      "learning_rate": 4.9811453774626335e-06,
+      "loss": 0.6301,
+      "step": 980
+    },
+    {
+      "epoch": 0.26045400238948624,
+      "grad_norm": 0.36838728225673273,
+      "learning_rate": 4.981102555798282e-06,
+      "loss": 0.6492,
+      "step": 981
+    },
+    {
+      "epoch": 0.26071950086287005,
+      "grad_norm": 0.4095904456767099,
+      "learning_rate": 4.981059685746355e-06,
+      "loss": 0.6332,
+      "step": 982
+    },
+    {
+      "epoch": 0.2609849993362538,
+      "grad_norm": 0.3657123204519917,
+      "learning_rate": 4.9810167673076896e-06,
+      "loss": 0.6447,
+      "step": 983
+    },
+    {
+      "epoch": 0.2612504978096376,
+      "grad_norm": 0.37768044353818037,
+      "learning_rate": 4.980973800483123e-06,
+      "loss": 0.649,
+      "step": 984
+    },
+    {
+      "epoch": 0.26151599628302136,
+      "grad_norm": 0.4065192180685813,
+      "learning_rate": 4.980930785273492e-06,
+      "loss": 0.6574,
+      "step": 985
+    },
+    {
+      "epoch": 0.26178149475640516,
+      "grad_norm": 0.3681415226963581,
+      "learning_rate": 4.980887721679637e-06,
+      "loss": 0.6555,
+      "step": 986
+    },
+    {
+      "epoch": 0.2620469932297889,
+      "grad_norm": 0.36216962817587833,
+      "learning_rate": 4.980844609702397e-06,
+      "loss": 0.6256,
+      "step": 987
+    },
+    {
+      "epoch": 0.2623124917031727,
+      "grad_norm": 0.3752251781540997,
+      "learning_rate": 4.9808014493426124e-06,
+      "loss": 0.6248,
+      "step": 988
+    },
+    {
+      "epoch": 0.2625779901765565,
+      "grad_norm": 0.36627485508812824,
+      "learning_rate": 4.9807582406011265e-06,
+      "loss": 0.6692,
+      "step": 989
+    },
+    {
+      "epoch": 0.2628434886499403,
+      "grad_norm": 0.3638792629169123,
+      "learning_rate": 4.9807149834787806e-06,
+      "loss": 0.6736,
+      "step": 990
+    },
+    {
+      "epoch": 0.26310898712332403,
+      "grad_norm": 0.37848625346615206,
+      "learning_rate": 4.980671677976419e-06,
+      "loss": 0.6286,
+      "step": 991
+    },
+    {
+      "epoch": 0.26337448559670784,
+      "grad_norm": 0.3748855594569195,
+      "learning_rate": 4.980628324094885e-06,
+      "loss": 0.6538,
+      "step": 992
+    },
+    {
+      "epoch": 0.2636399840700916,
+      "grad_norm": 0.35586075229113756,
+      "learning_rate": 4.9805849218350255e-06,
+      "loss": 0.6697,
+      "step": 993
+    },
+    {
+      "epoch": 0.2639054825434754,
+      "grad_norm": 0.39699081477204445,
+      "learning_rate": 4.980541471197687e-06,
+      "loss": 0.6472,
+      "step": 994
+    },
+    {
+      "epoch": 0.26417098101685915,
+      "grad_norm": 0.366446635023391,
+      "learning_rate": 4.980497972183716e-06,
+      "loss": 0.6594,
+      "step": 995
+    },
+    {
+      "epoch": 0.2644364794902429,
+      "grad_norm": 0.3796150359783703,
+      "learning_rate": 4.980454424793961e-06,
+      "loss": 0.6424,
+      "step": 996
+    },
+    {
+      "epoch": 0.2647019779636267,
+      "grad_norm": 0.3723896820854948,
+      "learning_rate": 4.980410829029272e-06,
+      "loss": 0.628,
+      "step": 997
+    },
+    {
+      "epoch": 0.26496747643701046,
+      "grad_norm": 0.35886436852751274,
+      "learning_rate": 4.980367184890499e-06,
+      "loss": 0.6445,
+      "step": 998
+    },
+    {
+      "epoch": 0.26523297491039427,
+      "grad_norm": 0.35835065043728165,
+      "learning_rate": 4.980323492378491e-06,
+      "loss": 0.6441,
+      "step": 999
+    },
+    {
+      "epoch": 0.265498473383778,
+      "grad_norm": 0.3591313042147608,
+      "learning_rate": 4.980279751494104e-06,
+      "loss": 0.6866,
+      "step": 1000
+    },
+    {
+      "epoch": 0.26576397185716183,
+      "grad_norm": 0.3656836899949334,
+      "learning_rate": 4.980235962238188e-06,
+      "loss": 0.6926,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2660294703305456,
+      "grad_norm": 0.3607855738162133,
+      "learning_rate": 4.980192124611599e-06,
+      "loss": 0.6272,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2662949688039294,
+      "grad_norm": 0.36855282842246273,
+      "learning_rate": 4.9801482386151905e-06,
+      "loss": 0.6606,
+      "step": 1003
+    },
+    {
+      "epoch": 0.26656046727731314,
+      "grad_norm": 0.3699974943566117,
+      "learning_rate": 4.980104304249819e-06,
+      "loss": 0.6967,
+      "step": 1004
+    },
+    {
+      "epoch": 0.26682596575069695,
+      "grad_norm": 0.35427505302748685,
+      "learning_rate": 4.9800603215163415e-06,
+      "loss": 0.613,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2670914642240807,
+      "grad_norm": 0.3620278170370366,
+      "learning_rate": 4.9800162904156155e-06,
+      "loss": 0.6374,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2673569626974645,
+      "grad_norm": 0.3569586726428625,
+      "learning_rate": 4.979972210948499e-06,
+      "loss": 0.6423,
+      "step": 1007
+    },
+    {
+      "epoch": 0.26762246117084826,
+      "grad_norm": 0.35097849073088916,
+      "learning_rate": 4.979928083115853e-06,
+      "loss": 0.6223,
+      "step": 1008
+    },
+    {
+      "epoch": 0.26788795964423207,
+      "grad_norm": 0.3558841188227152,
+      "learning_rate": 4.979883906918536e-06,
+      "loss": 0.6142,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2681534581176158,
+      "grad_norm": 0.36359631099266737,
+      "learning_rate": 4.979839682357414e-06,
+      "loss": 0.6596,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2684189565909996,
+      "grad_norm": 0.3636149284226135,
+      "learning_rate": 4.979795409433344e-06,
+      "loss": 0.6702,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2686844550643834,
+      "grad_norm": 0.3722364173082872,
+      "learning_rate": 4.979751088147192e-06,
+      "loss": 0.6375,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2689499535377672,
+      "grad_norm": 0.38001659256946785,
+      "learning_rate": 4.979706718499824e-06,
+      "loss": 0.6302,
+      "step": 1013
+    },
+    {
+      "epoch": 0.26921545201115094,
+      "grad_norm": 0.37155343314022443,
+      "learning_rate": 4.9796623004921016e-06,
+      "loss": 0.6795,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2694809504845347,
+      "grad_norm": 0.36957715696091586,
+      "learning_rate": 4.979617834124895e-06,
+      "loss": 0.6315,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2697464489579185,
+      "grad_norm": 0.36786882803706994,
+      "learning_rate": 4.979573319399068e-06,
+      "loss": 0.6483,
+      "step": 1016
+    },
+    {
+      "epoch": 0.27001194743130225,
+      "grad_norm": 0.36288275679902826,
+      "learning_rate": 4.979528756315491e-06,
+      "loss": 0.6397,
+      "step": 1017
+    },
+    {
+      "epoch": 0.27027744590468605,
+      "grad_norm": 0.3549232309789082,
+      "learning_rate": 4.979484144875032e-06,
+      "loss": 0.6611,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2705429443780698,
+      "grad_norm": 0.3576446391815027,
+      "learning_rate": 4.9794394850785616e-06,
+      "loss": 0.6114,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2708084428514536,
+      "grad_norm": 0.36090274675872946,
+      "learning_rate": 4.97939477692695e-06,
+      "loss": 0.6024,
+      "step": 1020
+    },
+    {
+      "epoch": 0.27107394132483736,
+      "grad_norm": 0.3572624736441068,
+      "learning_rate": 4.979350020421071e-06,
+      "loss": 0.6561,
+      "step": 1021
+    },
+    {
+      "epoch": 0.27133943979822117,
+      "grad_norm": 0.3715741116524624,
+      "learning_rate": 4.979305215561795e-06,
+      "loss": 0.661,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2716049382716049,
+      "grad_norm": 0.36784880588918306,
+      "learning_rate": 4.9792603623499965e-06,
+      "loss": 0.6703,
+      "step": 1023
+    },
+    {
+      "epoch": 0.27187043674498873,
+      "grad_norm": 0.40290401087707023,
+      "learning_rate": 4.979215460786552e-06,
+      "loss": 0.6533,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2721359352183725,
+      "grad_norm": 0.35855405719324956,
+      "learning_rate": 4.979170510872334e-06,
+      "loss": 0.6329,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2724014336917563,
+      "grad_norm": 0.3647908558652933,
+      "learning_rate": 4.979125512608223e-06,
+      "loss": 0.6547,
+      "step": 1026
+    },
+    {
+      "epoch": 0.27266693216514004,
+      "grad_norm": 0.36947883787403,
+      "learning_rate": 4.979080465995093e-06,
+      "loss": 0.6467,
+      "step": 1027
+    },
+    {
+      "epoch": 0.27293243063852385,
+      "grad_norm": 0.3846719010338243,
+      "learning_rate": 4.979035371033824e-06,
+      "loss": 0.6712,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2731979291119076,
+      "grad_norm": 0.37234047371660695,
+      "learning_rate": 4.978990227725297e-06,
+      "loss": 0.6811,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2734634275852914,
+      "grad_norm": 0.3682858925006891,
+      "learning_rate": 4.97894503607039e-06,
+      "loss": 0.6696,
+      "step": 1030
+    },
+    {
+      "epoch": 0.27372892605867516,
+      "grad_norm": 0.3553102914397915,
+      "learning_rate": 4.978899796069986e-06,
+      "loss": 0.6348,
+      "step": 1031
+    },
+    {
+      "epoch": 0.27399442453205897,
+      "grad_norm": 0.38445335797009916,
+      "learning_rate": 4.9788545077249655e-06,
+      "loss": 0.6416,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2742599230054427,
+      "grad_norm": 0.3540641688007243,
+      "learning_rate": 4.978809171036213e-06,
+      "loss": 0.6436,
+      "step": 1033
+    },
+    {
+      "epoch": 0.27452542147882647,
+      "grad_norm": 0.35701320811905085,
+      "learning_rate": 4.978763786004613e-06,
+      "loss": 0.6522,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2747909199522103,
+      "grad_norm": 0.3816791247486726,
+      "learning_rate": 4.97871835263105e-06,
+      "loss": 0.6384,
+      "step": 1035
+    },
+    {
+      "epoch": 0.27505641842559403,
+      "grad_norm": 0.3729395285778984,
+      "learning_rate": 4.97867287091641e-06,
+      "loss": 0.6667,
+      "step": 1036
+    },
+    {
+      "epoch": 0.27532191689897784,
+      "grad_norm": 0.3533855518978844,
+      "learning_rate": 4.9786273408615805e-06,
+      "loss": 0.6406,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2755874153723616,
+      "grad_norm": 0.3622612832585471,
+      "learning_rate": 4.978581762467449e-06,
+      "loss": 0.6295,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2758529138457454,
+      "grad_norm": 0.37672182638995366,
+      "learning_rate": 4.978536135734905e-06,
+      "loss": 0.6302,
+      "step": 1039
+    },
+    {
+      "epoch": 0.27611841231912915,
+      "grad_norm": 0.3657574771076464,
+      "learning_rate": 4.9784904606648366e-06,
+      "loss": 0.6105,
+      "step": 1040
+    },
+    {
+      "epoch": 0.27638391079251295,
+      "grad_norm": 0.42636879444707215,
+      "learning_rate": 4.978444737258137e-06,
+      "loss": 0.6249,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2766494092658967,
+      "grad_norm": 0.3674160896616571,
+      "learning_rate": 4.9783989655156965e-06,
+      "loss": 0.6643,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2769149077392805,
+      "grad_norm": 0.3682225329604301,
+      "learning_rate": 4.978353145438407e-06,
+      "loss": 0.6556,
+      "step": 1043
+    },
+    {
+      "epoch": 0.27718040621266427,
+      "grad_norm": 0.39590198616690214,
+      "learning_rate": 4.978307277027165e-06,
+      "loss": 0.6404,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2774459046860481,
+      "grad_norm": 0.4081112800566932,
+      "learning_rate": 4.9782613602828614e-06,
+      "loss": 0.6114,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2777114031594318,
+      "grad_norm": 0.3838676655816886,
+      "learning_rate": 4.978215395206394e-06,
+      "loss": 0.6332,
+      "step": 1046
+    },
+    {
+      "epoch": 0.27797690163281563,
+      "grad_norm": 0.3815210844192773,
+      "learning_rate": 4.978169381798659e-06,
+      "loss": 0.6487,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2782424001061994,
+      "grad_norm": 0.3605995293936653,
+      "learning_rate": 4.978123320060552e-06,
+      "loss": 0.6422,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2785078985795832,
+      "grad_norm": 0.35616745318355486,
+      "learning_rate": 4.9780772099929755e-06,
+      "loss": 0.5991,
+      "step": 1049
+    },
+    {
+      "epoch": 0.27877339705296694,
+      "grad_norm": 0.3736046260647422,
+      "learning_rate": 4.978031051596824e-06,
+      "loss": 0.6171,
+      "step": 1050
+    },
+    {
+      "epoch": 0.27903889552635075,
+      "grad_norm": 0.3576515022274845,
+      "learning_rate": 4.977984844873e-06,
+      "loss": 0.6189,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2793043939997345,
+      "grad_norm": 0.37415264077762256,
+      "learning_rate": 4.977938589822405e-06,
+      "loss": 0.6481,
+      "step": 1052
+    },
+    {
+      "epoch": 0.27956989247311825,
+      "grad_norm": 0.38740232408265945,
+      "learning_rate": 4.97789228644594e-06,
+      "loss": 0.6465,
+      "step": 1053
+    },
+    {
+      "epoch": 0.27983539094650206,
+      "grad_norm": 0.36488961523079405,
+      "learning_rate": 4.977845934744509e-06,
+      "loss": 0.6551,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2801008894198858,
+      "grad_norm": 0.36671510752598824,
+      "learning_rate": 4.977799534719014e-06,
+      "loss": 0.6578,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2803663878932696,
+      "grad_norm": 0.36863121575333246,
+      "learning_rate": 4.977753086370363e-06,
+      "loss": 0.6543,
+      "step": 1056
+    },
+    {
+      "epoch": 0.28063188636665337,
+      "grad_norm": 0.3673835879709623,
+      "learning_rate": 4.977706589699459e-06,
+      "loss": 0.647,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2808973848400372,
+      "grad_norm": 0.3805719174915159,
+      "learning_rate": 4.977660044707211e-06,
+      "loss": 0.6461,
+      "step": 1058
+    },
+    {
+      "epoch": 0.28116288331342093,
+      "grad_norm": 0.3738617837889614,
+      "learning_rate": 4.977613451394525e-06,
+      "loss": 0.6605,
+      "step": 1059
+    },
+    {
+      "epoch": 0.28142838178680474,
+      "grad_norm": 0.36895272556611924,
+      "learning_rate": 4.977566809762312e-06,
+      "loss": 0.6855,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2816938802601885,
+      "grad_norm": 0.3663668110974325,
+      "learning_rate": 4.977520119811479e-06,
+      "loss": 0.6584,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2819593787335723,
+      "grad_norm": 0.37707507476680496,
+      "learning_rate": 4.977473381542936e-06,
+      "loss": 0.6175,
+      "step": 1062
+    },
+    {
+      "epoch": 0.28222487720695605,
+      "grad_norm": 0.3734079207074507,
+      "learning_rate": 4.977426594957598e-06,
+      "loss": 0.6599,
+      "step": 1063
+    },
+    {
+      "epoch": 0.28249037568033986,
+      "grad_norm": 0.3861721946751422,
+      "learning_rate": 4.977379760056376e-06,
+      "loss": 0.6004,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2827558741537236,
+      "grad_norm": 0.3547710138333417,
+      "learning_rate": 4.977332876840182e-06,
+      "loss": 0.6164,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2830213726271074,
+      "grad_norm": 0.3849148616938619,
+      "learning_rate": 4.977285945309932e-06,
+      "loss": 0.6784,
+      "step": 1066
+    },
+    {
+      "epoch": 0.28328687110049117,
+      "grad_norm": 0.3799132968612521,
+      "learning_rate": 4.97723896546654e-06,
+      "loss": 0.5987,
+      "step": 1067
+    },
+    {
+      "epoch": 0.283552369573875,
+      "grad_norm": 0.36704655517993257,
+      "learning_rate": 4.977191937310923e-06,
+      "loss": 0.6373,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2838178680472587,
+      "grad_norm": 0.36616281253825506,
+      "learning_rate": 4.977144860843999e-06,
+      "loss": 0.6402,
+      "step": 1069
+    },
+    {
+      "epoch": 0.28408336652064253,
+      "grad_norm": 0.43383765197262214,
+      "learning_rate": 4.977097736066684e-06,
+      "loss": 0.6709,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2843488649940263,
+      "grad_norm": 0.3747998695242347,
+      "learning_rate": 4.977050562979899e-06,
+      "loss": 0.6496,
+      "step": 1071
+    },
+    {
+      "epoch": 0.28461436346741004,
+      "grad_norm": 0.43774417732764065,
+      "learning_rate": 4.977003341584562e-06,
+      "loss": 0.5989,
+      "step": 1072
+    },
+    {
+      "epoch": 0.28487986194079384,
+      "grad_norm": 0.3694306445203735,
+      "learning_rate": 4.976956071881596e-06,
+      "loss": 0.6353,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2851453604141776,
+      "grad_norm": 0.3748486580016225,
+      "learning_rate": 4.976908753871922e-06,
+      "loss": 0.6374,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2854108588875614,
+      "grad_norm": 0.40236012775431534,
+      "learning_rate": 4.976861387556462e-06,
+      "loss": 0.6497,
+      "step": 1075
+    },
+    {
+      "epoch": 0.28567635736094515,
+      "grad_norm": 0.3619853947522983,
+      "learning_rate": 4.976813972936141e-06,
+      "loss": 0.669,
+      "step": 1076
+    },
+    {
+      "epoch": 0.28594185583432896,
+      "grad_norm": 0.3849930432274775,
+      "learning_rate": 4.976766510011883e-06,
+      "loss": 0.5857,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2862073543077127,
+      "grad_norm": 0.40420789974038696,
+      "learning_rate": 4.976718998784614e-06,
+      "loss": 0.6152,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2864728527810965,
+      "grad_norm": 0.4047890114990248,
+      "learning_rate": 4.976671439255261e-06,
+      "loss": 0.6075,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2867383512544803,
+      "grad_norm": 0.36417790362112606,
+      "learning_rate": 4.97662383142475e-06,
+      "loss": 0.6725,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2870038497278641,
+      "grad_norm": 0.37279288917560943,
+      "learning_rate": 4.976576175294012e-06,
+      "loss": 0.6688,
+      "step": 1081
+    },
+    {
+      "epoch": 0.28726934820124783,
+      "grad_norm": 0.3769711148414339,
+      "learning_rate": 4.9765284708639735e-06,
+      "loss": 0.651,
+      "step": 1082
+    },
+    {
+      "epoch": 0.28753484667463164,
+      "grad_norm": 0.36201838478819254,
+      "learning_rate": 4.976480718135567e-06,
+      "loss": 0.6295,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2878003451480154,
+      "grad_norm": 0.36618405010112043,
+      "learning_rate": 4.976432917109722e-06,
+      "loss": 0.6507,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2880658436213992,
+      "grad_norm": 0.36136083902163096,
+      "learning_rate": 4.9763850677873735e-06,
+      "loss": 0.6319,
+      "step": 1085
+    },
+    {
+      "epoch": 0.28833134209478295,
+      "grad_norm": 0.38406693990346413,
+      "learning_rate": 4.9763371701694515e-06,
+      "loss": 0.6262,
+      "step": 1086
+    },
+    {
+      "epoch": 0.28859684056816676,
+      "grad_norm": 0.3888302360216497,
+      "learning_rate": 4.976289224256892e-06,
+      "loss": 0.6442,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2888623390415505,
+      "grad_norm": 0.3611850453000929,
+      "learning_rate": 4.97624123005063e-06,
+      "loss": 0.6431,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2891278375149343,
+      "grad_norm": 0.3589076257694757,
+      "learning_rate": 4.9761931875516e-06,
+      "loss": 0.6237,
+      "step": 1089
+    },
+    {
+      "epoch": 0.28939333598831807,
+      "grad_norm": 0.3581585044999295,
+      "learning_rate": 4.976145096760741e-06,
+      "loss": 0.626,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2896588344617018,
+      "grad_norm": 0.3666506277965257,
+      "learning_rate": 4.9760969576789905e-06,
+      "loss": 0.6325,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2899243329350856,
+      "grad_norm": 0.3646360957648641,
+      "learning_rate": 4.976048770307286e-06,
+      "loss": 0.6253,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2901898314084694,
+      "grad_norm": 0.3557008343438929,
+      "learning_rate": 4.976000534646568e-06,
+      "loss": 0.6068,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2904553298818532,
+      "grad_norm": 0.38629588484800864,
+      "learning_rate": 4.975952250697778e-06,
+      "loss": 0.6755,
+      "step": 1094
+    },
+    {
+      "epoch": 0.29072082835523694,
+      "grad_norm": 0.36941376555216937,
+      "learning_rate": 4.975903918461856e-06,
+      "loss": 0.6559,
+      "step": 1095
+    },
+    {
+      "epoch": 0.29098632682862074,
+      "grad_norm": 0.36868040224675036,
+      "learning_rate": 4.975855537939746e-06,
+      "loss": 0.6524,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2912518253020045,
+      "grad_norm": 0.3779000553077218,
+      "learning_rate": 4.9758071091323905e-06,
+      "loss": 0.607,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2915173237753883,
+      "grad_norm": 0.41871886278418036,
+      "learning_rate": 4.975758632040734e-06,
+      "loss": 0.6527,
+      "step": 1098
+    },
+    {
+      "epoch": 0.29178282224877206,
+      "grad_norm": 0.37608653378257667,
+      "learning_rate": 4.975710106665724e-06,
+      "loss": 0.6361,
+      "step": 1099
+    },
+    {
+      "epoch": 0.29204832072215586,
+      "grad_norm": 0.39800491607222016,
+      "learning_rate": 4.975661533008305e-06,
+      "loss": 0.6666,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2923138191955396,
+      "grad_norm": 0.3747362995726176,
+      "learning_rate": 4.975612911069425e-06,
+      "loss": 0.6554,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2925793176689234,
+      "grad_norm": 0.3628819752775224,
+      "learning_rate": 4.975564240850032e-06,
+      "loss": 0.6571,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2928448161423072,
+      "grad_norm": 0.3633263237254101,
+      "learning_rate": 4.975515522351073e-06,
+      "loss": 0.6166,
+      "step": 1103
+    },
+    {
+      "epoch": 0.293110314615691,
+      "grad_norm": 0.3714301141730258,
+      "learning_rate": 4.975466755573503e-06,
+      "loss": 0.6155,
+      "step": 1104
+    },
+    {
+      "epoch": 0.29337581308907473,
+      "grad_norm": 0.37697531633451764,
+      "learning_rate": 4.975417940518268e-06,
+      "loss": 0.6652,
+      "step": 1105
+    },
+    {
+      "epoch": 0.29364131156245854,
+      "grad_norm": 0.3665382088093708,
+      "learning_rate": 4.975369077186324e-06,
+      "loss": 0.6735,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2939068100358423,
+      "grad_norm": 0.3881447317983678,
+      "learning_rate": 4.975320165578621e-06,
+      "loss": 0.6678,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2941723085092261,
+      "grad_norm": 0.3681517437141483,
+      "learning_rate": 4.975271205696115e-06,
+      "loss": 0.6268,
+      "step": 1108
+    },
+    {
+      "epoch": 0.29443780698260985,
+      "grad_norm": 0.3717262403608333,
+      "learning_rate": 4.975222197539758e-06,
+      "loss": 0.6017,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2947033054559936,
+      "grad_norm": 0.36193710046561917,
+      "learning_rate": 4.97517314111051e-06,
+      "loss": 0.6389,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2949688039293774,
+      "grad_norm": 0.3711770934901428,
+      "learning_rate": 4.975124036409325e-06,
+      "loss": 0.6873,
+      "step": 1111
+    },
+    {
+      "epoch": 0.29523430240276116,
+      "grad_norm": 0.38119641481426275,
+      "learning_rate": 4.975074883437161e-06,
+      "loss": 0.6423,
+      "step": 1112
+    },
+    {
+      "epoch": 0.29549980087614497,
+      "grad_norm": 0.3710720262270734,
+      "learning_rate": 4.975025682194976e-06,
+      "loss": 0.6482,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2957652993495287,
+      "grad_norm": 0.37143480055752387,
+      "learning_rate": 4.974976432683732e-06,
+      "loss": 0.667,
+      "step": 1114
+    },
+    {
+      "epoch": 0.29603079782291253,
+      "grad_norm": 0.3795667856621345,
+      "learning_rate": 4.974927134904386e-06,
+      "loss": 0.6367,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 0.36287774793276917,
+      "learning_rate": 4.974877788857903e-06,
+      "loss": 0.6384,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2965617947696801,
+      "grad_norm": 0.35103145336870095,
+      "learning_rate": 4.9748283945452415e-06,
+      "loss": 0.5954,
+      "step": 1117
+    },
+    {
+      "epoch": 0.29682729324306384,
+      "grad_norm": 0.36495536794758493,
+      "learning_rate": 4.974778951967368e-06,
+      "loss": 0.6254,
+      "step": 1118
+    },
+    {
+      "epoch": 0.29709279171644765,
+      "grad_norm": 0.4147833945623723,
+      "learning_rate": 4.974729461125246e-06,
+      "loss": 0.6221,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2973582901898314,
+      "grad_norm": 0.3687917710658798,
+      "learning_rate": 4.974679922019839e-06,
+      "loss": 0.667,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2976237886632152,
+      "grad_norm": 0.368437127448733,
+      "learning_rate": 4.974630334652116e-06,
+      "loss": 0.6405,
+      "step": 1121
+    },
+    {
+      "epoch": 0.29788928713659896,
+      "grad_norm": 0.3765791219071444,
+      "learning_rate": 4.974580699023042e-06,
+      "loss": 0.6424,
+      "step": 1122
+    },
+    {
+      "epoch": 0.29815478560998276,
+      "grad_norm": 0.39218884689709,
+      "learning_rate": 4.9745310151335855e-06,
+      "loss": 0.6235,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2984202840833665,
+      "grad_norm": 0.38378639476074716,
+      "learning_rate": 4.974481282984716e-06,
+      "loss": 0.6474,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2986857825567503,
+      "grad_norm": 0.42629351755264794,
+      "learning_rate": 4.9744315025774015e-06,
+      "loss": 0.6358,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2989512810301341,
+      "grad_norm": 0.3764995616646047,
+      "learning_rate": 4.9743816739126146e-06,
+      "loss": 0.6435,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2992167795035179,
+      "grad_norm": 0.3699827961388043,
+      "learning_rate": 4.974331796991328e-06,
+      "loss": 0.6279,
+      "step": 1127
+    },
+    {
+      "epoch": 0.29948227797690163,
+      "grad_norm": 0.3829003820104978,
+      "learning_rate": 4.974281871814513e-06,
+      "loss": 0.6758,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2997477764502854,
+      "grad_norm": 0.37089225016504546,
+      "learning_rate": 4.974231898383142e-06,
+      "loss": 0.6474,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3000132749236692,
+      "grad_norm": 0.3521625524575173,
+      "learning_rate": 4.974181876698192e-06,
+      "loss": 0.592,
+      "step": 1130
+    },
+    {
+      "epoch": 0.30027877339705294,
+      "grad_norm": 0.4093245197064176,
+      "learning_rate": 4.974131806760637e-06,
+      "loss": 0.6153,
+      "step": 1131
+    },
+    {
+      "epoch": 0.30054427187043675,
+      "grad_norm": 0.3747038691130968,
+      "learning_rate": 4.974081688571455e-06,
+      "loss": 0.6516,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3008097703438205,
+      "grad_norm": 0.3728344093015197,
+      "learning_rate": 4.974031522131621e-06,
+      "loss": 0.637,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3010752688172043,
+      "grad_norm": 0.36801062969007814,
+      "learning_rate": 4.973981307442116e-06,
+      "loss": 0.6617,
+      "step": 1134
+    },
+    {
+      "epoch": 0.30134076729058806,
+      "grad_norm": 0.37359690290288783,
+      "learning_rate": 4.973931044503918e-06,
+      "loss": 0.6944,
+      "step": 1135
+    },
+    {
+      "epoch": 0.30160626576397187,
+      "grad_norm": 0.3616515553973182,
+      "learning_rate": 4.973880733318007e-06,
+      "loss": 0.6712,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3018717642373556,
+      "grad_norm": 0.35759397663643866,
+      "learning_rate": 4.973830373885365e-06,
+      "loss": 0.5899,
+      "step": 1137
+    },
+    {
+      "epoch": 0.30213726271073943,
+      "grad_norm": 0.3649548838358172,
+      "learning_rate": 4.973779966206974e-06,
+      "loss": 0.6367,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3024027611841232,
+      "grad_norm": 0.35978835326881975,
+      "learning_rate": 4.973729510283816e-06,
+      "loss": 0.6402,
+      "step": 1139
+    },
+    {
+      "epoch": 0.302668259657507,
+      "grad_norm": 0.3574868347892176,
+      "learning_rate": 4.973679006116876e-06,
+      "loss": 0.6389,
+      "step": 1140
+    },
+    {
+      "epoch": 0.30293375813089074,
+      "grad_norm": 0.3592349428526638,
+      "learning_rate": 4.973628453707139e-06,
+      "loss": 0.6339,
+      "step": 1141
+    },
+    {
+      "epoch": 0.30319925660427455,
+      "grad_norm": 0.36513790765039866,
+      "learning_rate": 4.973577853055591e-06,
+      "loss": 0.6356,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3034647550776583,
+      "grad_norm": 0.35776671262425036,
+      "learning_rate": 4.973527204163218e-06,
+      "loss": 0.6463,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3037302535510421,
+      "grad_norm": 0.39145078042075204,
+      "learning_rate": 4.9734765070310085e-06,
+      "loss": 0.6121,
+      "step": 1144
+    },
+    {
+      "epoch": 0.30399575202442586,
+      "grad_norm": 0.38454986651587336,
+      "learning_rate": 4.973425761659952e-06,
+      "loss": 0.6419,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3042612504978096,
+      "grad_norm": 0.3651720822252698,
+      "learning_rate": 4.973374968051035e-06,
+      "loss": 0.6789,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3045267489711934,
+      "grad_norm": 0.37266762022796446,
+      "learning_rate": 4.973324126205251e-06,
+      "loss": 0.6573,
+      "step": 1147
+    },
+    {
+      "epoch": 0.30479224744457717,
+      "grad_norm": 0.36583017222514574,
+      "learning_rate": 4.973273236123592e-06,
+      "loss": 0.6436,
+      "step": 1148
+    },
+    {
+      "epoch": 0.305057745917961,
+      "grad_norm": 0.36984308792531545,
+      "learning_rate": 4.973222297807048e-06,
+      "loss": 0.654,
+      "step": 1149
+    },
+    {
+      "epoch": 0.30532324439134473,
+      "grad_norm": 0.36296298997708737,
+      "learning_rate": 4.9731713112566144e-06,
+      "loss": 0.6213,
+      "step": 1150
+    },
+    {
+      "epoch": 0.30558874286472854,
+      "grad_norm": 0.3969882923530544,
+      "learning_rate": 4.973120276473284e-06,
+      "loss": 0.6303,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3058542413381123,
+      "grad_norm": 0.3709599941947518,
+      "learning_rate": 4.9730691934580545e-06,
+      "loss": 0.6151,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3061197398114961,
+      "grad_norm": 0.3712133967684035,
+      "learning_rate": 4.973018062211919e-06,
+      "loss": 0.6542,
+      "step": 1153
+    },
+    {
+      "epoch": 0.30638523828487985,
+      "grad_norm": 0.3686728439426148,
+      "learning_rate": 4.972966882735876e-06,
+      "loss": 0.6572,
+      "step": 1154
+    },
+    {
+      "epoch": 0.30665073675826365,
+      "grad_norm": 0.36155246211334735,
+      "learning_rate": 4.972915655030926e-06,
+      "loss": 0.633,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3069162352316474,
+      "grad_norm": 0.3647944810341329,
+      "learning_rate": 4.972864379098064e-06,
+      "loss": 0.6331,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3071817337050312,
+      "grad_norm": 0.3969342707637075,
+      "learning_rate": 4.972813054938293e-06,
+      "loss": 0.6688,
+      "step": 1157
+    },
+    {
+      "epoch": 0.30744723217841496,
+      "grad_norm": 0.37933894707453275,
+      "learning_rate": 4.972761682552613e-06,
+      "loss": 0.6509,
+      "step": 1158
+    },
+    {
+      "epoch": 0.30771273065179877,
+      "grad_norm": 0.3952402592545023,
+      "learning_rate": 4.9727102619420265e-06,
+      "loss": 0.6603,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3079782291251825,
+      "grad_norm": 0.43888487689848305,
+      "learning_rate": 4.972658793107534e-06,
+      "loss": 0.6456,
+      "step": 1160
+    },
+    {
+      "epoch": 0.30824372759856633,
+      "grad_norm": 0.38915812338524325,
+      "learning_rate": 4.972607276050142e-06,
+      "loss": 0.6755,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3085092260719501,
+      "grad_norm": 0.4565404388718119,
+      "learning_rate": 4.972555710770854e-06,
+      "loss": 0.5976,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3087747245453339,
+      "grad_norm": 0.38584294050783047,
+      "learning_rate": 4.9725040972706755e-06,
+      "loss": 0.6396,
+      "step": 1163
+    },
+    {
+      "epoch": 0.30904022301871764,
+      "grad_norm": 0.3904123684938697,
+      "learning_rate": 4.9724524355506135e-06,
+      "loss": 0.6219,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3093057214921014,
+      "grad_norm": 0.39446221328058956,
+      "learning_rate": 4.9724007256116755e-06,
+      "loss": 0.6141,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3095712199654852,
+      "grad_norm": 0.34872890334357637,
+      "learning_rate": 4.9723489674548695e-06,
+      "loss": 0.6082,
+      "step": 1166
+    },
+    {
+      "epoch": 0.30983671843886895,
+      "grad_norm": 0.3779781026484708,
+      "learning_rate": 4.972297161081205e-06,
+      "loss": 0.6445,
+      "step": 1167
+    },
+    {
+      "epoch": 0.31010221691225276,
+      "grad_norm": 0.4612015309967254,
+      "learning_rate": 4.972245306491694e-06,
+      "loss": 0.6357,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3103677153856365,
+      "grad_norm": 0.376288731974902,
+      "learning_rate": 4.9721934036873455e-06,
+      "loss": 0.6445,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3106332138590203,
+      "grad_norm": 0.39029455733963414,
+      "learning_rate": 4.972141452669172e-06,
+      "loss": 0.6836,
+      "step": 1170
+    },
+    {
+      "epoch": 0.31089871233240407,
+      "grad_norm": 0.41205089113636817,
+      "learning_rate": 4.972089453438189e-06,
+      "loss": 0.6622,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3111642108057879,
+      "grad_norm": 0.3806204577440277,
+      "learning_rate": 4.972037405995408e-06,
+      "loss": 0.6608,
+      "step": 1172
+    },
+    {
+      "epoch": 0.31142970927917163,
+      "grad_norm": 0.37423120088296086,
+      "learning_rate": 4.971985310341845e-06,
+      "loss": 0.6389,
+      "step": 1173
+    },
+    {
+      "epoch": 0.31169520775255544,
+      "grad_norm": 0.38961077240672787,
+      "learning_rate": 4.971933166478516e-06,
+      "loss": 0.6384,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3119607062259392,
+      "grad_norm": 0.39326611404176093,
+      "learning_rate": 4.971880974406438e-06,
+      "loss": 0.6493,
+      "step": 1175
+    },
+    {
+      "epoch": 0.312226204699323,
+      "grad_norm": 0.3773472668067406,
+      "learning_rate": 4.971828734126628e-06,
+      "loss": 0.6402,
+      "step": 1176
+    },
+    {
+      "epoch": 0.31249170317270675,
+      "grad_norm": 0.37634222257849803,
+      "learning_rate": 4.971776445640107e-06,
+      "loss": 0.6436,
+      "step": 1177
+    },
+    {
+      "epoch": 0.31275720164609055,
+      "grad_norm": 0.3971968746908418,
+      "learning_rate": 4.9717241089478925e-06,
+      "loss": 0.6312,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3130227001194743,
+      "grad_norm": 0.36898188875095717,
+      "learning_rate": 4.971671724051006e-06,
+      "loss": 0.6319,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3132881985928581,
+      "grad_norm": 0.3735814886195571,
+      "learning_rate": 4.9716192909504695e-06,
+      "loss": 0.6995,
+      "step": 1180
+    },
+    {
+      "epoch": 0.31355369706624187,
+      "grad_norm": 0.43807928720477224,
+      "learning_rate": 4.971566809647305e-06,
+      "loss": 0.64,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3138191955396257,
+      "grad_norm": 0.3554185719143483,
+      "learning_rate": 4.971514280142536e-06,
+      "loss": 0.653,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3140846940130094,
+      "grad_norm": 0.3600671399514037,
+      "learning_rate": 4.971461702437188e-06,
+      "loss": 0.6499,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3143501924863932,
+      "grad_norm": 0.42884570095607055,
+      "learning_rate": 4.9714090765322855e-06,
+      "loss": 0.5977,
+      "step": 1184
+    },
+    {
+      "epoch": 0.314615690959777,
+      "grad_norm": 0.3771822640580587,
+      "learning_rate": 4.971356402428855e-06,
+      "loss": 0.6146,
+      "step": 1185
+    },
+    {
+      "epoch": 0.31488118943316074,
+      "grad_norm": 0.36555143380876054,
+      "learning_rate": 4.971303680127923e-06,
+      "loss": 0.6494,
+      "step": 1186
+    },
+    {
+      "epoch": 0.31514668790654454,
+      "grad_norm": 0.35469507178908677,
+      "learning_rate": 4.971250909630519e-06,
+      "loss": 0.6228,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3154121863799283,
+      "grad_norm": 0.37383246878433135,
+      "learning_rate": 4.971198090937671e-06,
+      "loss": 0.6685,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3156776848533121,
+      "grad_norm": 0.38633289261138815,
+      "learning_rate": 4.971145224050411e-06,
+      "loss": 0.6502,
+      "step": 1189
+    },
+    {
+      "epoch": 0.31594318332669585,
+      "grad_norm": 0.37760264930069487,
+      "learning_rate": 4.971092308969767e-06,
+      "loss": 0.6211,
+      "step": 1190
+    },
+    {
+      "epoch": 0.31620868180007966,
+      "grad_norm": 0.3923110144964888,
+      "learning_rate": 4.971039345696775e-06,
+      "loss": 0.6457,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3164741802734634,
+      "grad_norm": 0.41798910756637,
+      "learning_rate": 4.970986334232464e-06,
+      "loss": 0.6165,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3167396787468472,
+      "grad_norm": 0.37896907857521517,
+      "learning_rate": 4.970933274577869e-06,
+      "loss": 0.6131,
+      "step": 1193
+    },
+    {
+      "epoch": 0.31700517722023097,
+      "grad_norm": 0.3929800177031569,
+      "learning_rate": 4.970880166734026e-06,
+      "loss": 0.6109,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3172706756936148,
+      "grad_norm": 0.38625252598325116,
+      "learning_rate": 4.9708270107019695e-06,
+      "loss": 0.6512,
+      "step": 1195
+    },
+    {
+      "epoch": 0.31753617416699853,
+      "grad_norm": 0.396773031617555,
+      "learning_rate": 4.970773806482738e-06,
+      "loss": 0.6369,
+      "step": 1196
+    },
+    {
+      "epoch": 0.31780167264038234,
+      "grad_norm": 0.3928328348883509,
+      "learning_rate": 4.970720554077366e-06,
+      "loss": 0.658,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3180671711137661,
+      "grad_norm": 0.36965240472489613,
+      "learning_rate": 4.970667253486895e-06,
+      "loss": 0.6504,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3183326695871499,
+      "grad_norm": 0.3882045898977093,
+      "learning_rate": 4.970613904712363e-06,
+      "loss": 0.6731,
+      "step": 1199
+    },
+    {
+      "epoch": 0.31859816806053365,
+      "grad_norm": 0.37814287715143236,
+      "learning_rate": 4.970560507754811e-06,
+      "loss": 0.6302,
+      "step": 1200
+    },
+    {
+      "epoch": 0.31886366653391746,
+      "grad_norm": 0.40819492057896173,
+      "learning_rate": 4.970507062615279e-06,
+      "loss": 0.626,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3191291650073012,
+      "grad_norm": 0.37598809384418796,
+      "learning_rate": 4.970453569294812e-06,
+      "loss": 0.67,
+      "step": 1202
+    },
+    {
+      "epoch": 0.31939466348068496,
+      "grad_norm": 0.3820873853598865,
+      "learning_rate": 4.970400027794451e-06,
+      "loss": 0.6703,
+      "step": 1203
+    },
+    {
+      "epoch": 0.31966016195406877,
+      "grad_norm": 0.3840650953228521,
+      "learning_rate": 4.970346438115241e-06,
+      "loss": 0.7007,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3199256604274525,
+      "grad_norm": 0.3833331826383565,
+      "learning_rate": 4.970292800258227e-06,
+      "loss": 0.6881,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3201911589008363,
+      "grad_norm": 0.3883360600368711,
+      "learning_rate": 4.970239114224455e-06,
+      "loss": 0.6386,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3204566573742201,
+      "grad_norm": 0.38546498480970454,
+      "learning_rate": 4.9701853800149715e-06,
+      "loss": 0.6408,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3207221558476039,
+      "grad_norm": 0.39736338249726827,
+      "learning_rate": 4.970131597630826e-06,
+      "loss": 0.6383,
+      "step": 1208
+    },
+    {
+      "epoch": 0.32098765432098764,
+      "grad_norm": 0.3750662089816997,
+      "learning_rate": 4.970077767073066e-06,
+      "loss": 0.6497,
+      "step": 1209
+    },
+    {
+      "epoch": 0.32125315279437144,
+      "grad_norm": 0.37614692491962076,
+      "learning_rate": 4.970023888342742e-06,
+      "loss": 0.6481,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3215186512677552,
+      "grad_norm": 0.39398809434318954,
+      "learning_rate": 4.969969961440904e-06,
+      "loss": 0.6603,
+      "step": 1211
+    },
+    {
+      "epoch": 0.321784149741139,
+      "grad_norm": 0.37903198746645655,
+      "learning_rate": 4.969915986368604e-06,
+      "loss": 0.6394,
+      "step": 1212
+    },
+    {
+      "epoch": 0.32204964821452275,
+      "grad_norm": 0.3806943211152248,
+      "learning_rate": 4.969861963126896e-06,
+      "loss": 0.6852,
+      "step": 1213
+    },
+    {
+      "epoch": 0.32231514668790656,
+      "grad_norm": 0.378411613698888,
+      "learning_rate": 4.9698078917168315e-06,
+      "loss": 0.6596,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 0.37293848988846345,
+      "learning_rate": 4.969753772139466e-06,
+      "loss": 0.6292,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3228461436346741,
+      "grad_norm": 0.3871294240192617,
+      "learning_rate": 4.969699604395856e-06,
+      "loss": 0.6813,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3231116421080579,
+      "grad_norm": 0.38562512359951995,
+      "learning_rate": 4.969645388487056e-06,
+      "loss": 0.6322,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3233771405814417,
+      "grad_norm": 0.37858135306273044,
+      "learning_rate": 4.969591124414124e-06,
+      "loss": 0.6487,
+      "step": 1218
+    },
+    {
+      "epoch": 0.32364263905482543,
+      "grad_norm": 0.38742082721166776,
+      "learning_rate": 4.969536812178119e-06,
+      "loss": 0.6275,
+      "step": 1219
+    },
+    {
+      "epoch": 0.32390813752820924,
+      "grad_norm": 0.3949888770742023,
+      "learning_rate": 4.969482451780099e-06,
+      "loss": 0.6512,
+      "step": 1220
+    },
+    {
+      "epoch": 0.324173636001593,
+      "grad_norm": 0.37362786748858295,
+      "learning_rate": 4.969428043221126e-06,
+      "loss": 0.6462,
+      "step": 1221
+    },
+    {
+      "epoch": 0.32443913447497674,
+      "grad_norm": 0.3854571710583223,
+      "learning_rate": 4.969373586502258e-06,
+      "loss": 0.6899,
+      "step": 1222
+    },
+    {
+      "epoch": 0.32470463294836055,
+      "grad_norm": 0.37609556159438146,
+      "learning_rate": 4.9693190816245615e-06,
+      "loss": 0.6941,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3249701314217443,
+      "grad_norm": 0.376576559340508,
+      "learning_rate": 4.969264528589095e-06,
+      "loss": 0.6507,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3252356298951281,
+      "grad_norm": 0.3759310587412502,
+      "learning_rate": 4.9692099273969255e-06,
+      "loss": 0.6347,
+      "step": 1225
+    },
+    {
+      "epoch": 0.32550112836851186,
+      "grad_norm": 0.3677685461171089,
+      "learning_rate": 4.9691552780491165e-06,
+      "loss": 0.642,
+      "step": 1226
+    },
+    {
+      "epoch": 0.32576662684189567,
+      "grad_norm": 0.43429630087021026,
+      "learning_rate": 4.969100580546734e-06,
+      "loss": 0.655,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3260321253152794,
+      "grad_norm": 0.3972913905485784,
+      "learning_rate": 4.969045834890844e-06,
+      "loss": 0.6222,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3262976237886632,
+      "grad_norm": 0.39411628890650474,
+      "learning_rate": 4.968991041082516e-06,
+      "loss": 0.6367,
+      "step": 1229
+    },
+    {
+      "epoch": 0.326563122262047,
+      "grad_norm": 0.4121478949385552,
+      "learning_rate": 4.968936199122818e-06,
+      "loss": 0.6576,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3268286207354308,
+      "grad_norm": 0.39351689900550935,
+      "learning_rate": 4.968881309012817e-06,
+      "loss": 0.671,
+      "step": 1231
+    },
+    {
+      "epoch": 0.32709411920881454,
+      "grad_norm": 0.40431089792915353,
+      "learning_rate": 4.968826370753588e-06,
+      "loss": 0.6297,
+      "step": 1232
+    },
+    {
+      "epoch": 0.32735961768219835,
+      "grad_norm": 0.4155017267907915,
+      "learning_rate": 4.968771384346199e-06,
+      "loss": 0.6343,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3276251161555821,
+      "grad_norm": 0.39505094820926934,
+      "learning_rate": 4.968716349791723e-06,
+      "loss": 0.6508,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3278906146289659,
+      "grad_norm": 0.3900439205266144,
+      "learning_rate": 4.968661267091234e-06,
+      "loss": 0.6651,
+      "step": 1235
+    },
+    {
+      "epoch": 0.32815611310234966,
+      "grad_norm": 0.3838567287234332,
+      "learning_rate": 4.968606136245806e-06,
+      "loss": 0.6784,
+      "step": 1236
+    },
+    {
+      "epoch": 0.32842161157573346,
+      "grad_norm": 0.3692972630077756,
+      "learning_rate": 4.968550957256515e-06,
+      "loss": 0.6597,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3286871100491172,
+      "grad_norm": 0.3787752976099921,
+      "learning_rate": 4.968495730124435e-06,
+      "loss": 0.5861,
+      "step": 1238
+    },
+    {
+      "epoch": 0.328952608522501,
+      "grad_norm": 0.385960339456465,
+      "learning_rate": 4.968440454850645e-06,
+      "loss": 0.6413,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3292181069958848,
+      "grad_norm": 0.38108364058089694,
+      "learning_rate": 4.968385131436222e-06,
+      "loss": 0.6373,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3294836054692685,
+      "grad_norm": 0.37361168543067974,
+      "learning_rate": 4.968329759882245e-06,
+      "loss": 0.63,
+      "step": 1241
+    },
+    {
+      "epoch": 0.32974910394265233,
+      "grad_norm": 0.38063184677826034,
+      "learning_rate": 4.968274340189795e-06,
+      "loss": 0.6595,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3300146024160361,
+      "grad_norm": 0.38266301382281803,
+      "learning_rate": 4.968218872359952e-06,
+      "loss": 0.6829,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3302801008894199,
+      "grad_norm": 0.3743457762433329,
+      "learning_rate": 4.968163356393797e-06,
+      "loss": 0.642,
+      "step": 1244
+    },
+    {
+      "epoch": 0.33054559936280364,
+      "grad_norm": 0.37673658524483494,
+      "learning_rate": 4.968107792292414e-06,
+      "loss": 0.6515,
+      "step": 1245
+    },
+    {
+      "epoch": 0.33081109783618745,
+      "grad_norm": 0.399175358291409,
+      "learning_rate": 4.9680521800568856e-06,
+      "loss": 0.6282,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3310765963095712,
+      "grad_norm": 0.3575619463721458,
+      "learning_rate": 4.967996519688298e-06,
+      "loss": 0.6574,
+      "step": 1247
+    },
+    {
+      "epoch": 0.331342094782955,
+      "grad_norm": 0.374835351902063,
+      "learning_rate": 4.967940811187733e-06,
+      "loss": 0.6934,
+      "step": 1248
+    },
+    {
+      "epoch": 0.33160759325633876,
+      "grad_norm": 0.3718438462752692,
+      "learning_rate": 4.967885054556283e-06,
+      "loss": 0.6369,
+      "step": 1249
+    },
+    {
+      "epoch": 0.33187309172972257,
+      "grad_norm": 0.3915516002224567,
+      "learning_rate": 4.96782924979503e-06,
+      "loss": 0.6553,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3321385902031063,
+      "grad_norm": 0.3931536697264864,
+      "learning_rate": 4.967773396905065e-06,
+      "loss": 0.6474,
+      "step": 1251
+    },
+    {
+      "epoch": 0.33240408867649013,
+      "grad_norm": 0.35748610097975514,
+      "learning_rate": 4.967717495887476e-06,
+      "loss": 0.6158,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3326695871498739,
+      "grad_norm": 0.37916638694158805,
+      "learning_rate": 4.967661546743354e-06,
+      "loss": 0.6333,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3329350856232577,
+      "grad_norm": 0.3634201334581195,
+      "learning_rate": 4.96760554947379e-06,
+      "loss": 0.6352,
+      "step": 1254
+    },
+    {
+      "epoch": 0.33320058409664144,
+      "grad_norm": 0.3775606423853063,
+      "learning_rate": 4.967549504079876e-06,
+      "loss": 0.6559,
+      "step": 1255
+    },
+    {
+      "epoch": 0.33346608257002525,
+      "grad_norm": 0.3823992935295063,
+      "learning_rate": 4.967493410562705e-06,
+      "loss": 0.638,
+      "step": 1256
+    },
+    {
+      "epoch": 0.333731581043409,
+      "grad_norm": 0.3752163493940841,
+      "learning_rate": 4.967437268923372e-06,
+      "loss": 0.6747,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3339970795167928,
+      "grad_norm": 0.3737054649051383,
+      "learning_rate": 4.96738107916297e-06,
+      "loss": 0.5964,
+      "step": 1258
+    },
+    {
+      "epoch": 0.33426257799017656,
+      "grad_norm": 0.38471152194156755,
+      "learning_rate": 4.967324841282596e-06,
+      "loss": 0.613,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3345280764635603,
+      "grad_norm": 0.37944769648714166,
+      "learning_rate": 4.967268555283346e-06,
+      "loss": 0.6418,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3347935749369441,
+      "grad_norm": 0.3651988063579846,
+      "learning_rate": 4.967212221166319e-06,
+      "loss": 0.6454,
+      "step": 1261
+    },
+    {
+      "epoch": 0.33505907341032787,
+      "grad_norm": 0.36961299995464575,
+      "learning_rate": 4.967155838932613e-06,
+      "loss": 0.6338,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3353245718837117,
+      "grad_norm": 0.377312244983752,
+      "learning_rate": 4.967099408583327e-06,
+      "loss": 0.6377,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3355900703570954,
+      "grad_norm": 0.3877479803620617,
+      "learning_rate": 4.967042930119562e-06,
+      "loss": 0.6924,
+      "step": 1264
+    },
+    {
+      "epoch": 0.33585556883047923,
+      "grad_norm": 0.3722360404758459,
+      "learning_rate": 4.9669864035424195e-06,
+      "loss": 0.6402,
+      "step": 1265
+    },
+    {
+      "epoch": 0.336121067303863,
+      "grad_norm": 0.3861384724525087,
+      "learning_rate": 4.966929828853002e-06,
+      "loss": 0.6144,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3363865657772468,
+      "grad_norm": 0.38293062568547903,
+      "learning_rate": 4.966873206052413e-06,
+      "loss": 0.6273,
+      "step": 1267
+    },
+    {
+      "epoch": 0.33665206425063054,
+      "grad_norm": 0.37043997405393947,
+      "learning_rate": 4.966816535141756e-06,
+      "loss": 0.6153,
+      "step": 1268
+    },
+    {
+      "epoch": 0.33691756272401435,
+      "grad_norm": 0.3865129056209508,
+      "learning_rate": 4.9667598161221374e-06,
+      "loss": 0.6197,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3371830611973981,
+      "grad_norm": 0.3913776831297648,
+      "learning_rate": 4.966703048994662e-06,
+      "loss": 0.628,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3374485596707819,
+      "grad_norm": 0.3926460415222075,
+      "learning_rate": 4.9666462337604386e-06,
+      "loss": 0.6403,
+      "step": 1271
+    },
+    {
+      "epoch": 0.33771405814416566,
+      "grad_norm": 0.3821552611169409,
+      "learning_rate": 4.966589370420573e-06,
+      "loss": 0.6668,
+      "step": 1272
+    },
+    {
+      "epoch": 0.33797955661754947,
+      "grad_norm": 0.3695640890804788,
+      "learning_rate": 4.966532458976176e-06,
+      "loss": 0.6226,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3382450550909332,
+      "grad_norm": 0.34457327673863775,
+      "learning_rate": 4.966475499428357e-06,
+      "loss": 0.6027,
+      "step": 1274
+    },
+    {
+      "epoch": 0.33851055356431703,
+      "grad_norm": 0.37633149095083684,
+      "learning_rate": 4.9664184917782275e-06,
+      "loss": 0.6922,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3387760520377008,
+      "grad_norm": 0.35822937825563533,
+      "learning_rate": 4.966361436026898e-06,
+      "loss": 0.6536,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3390415505110846,
+      "grad_norm": 0.36824771473329787,
+      "learning_rate": 4.966304332175482e-06,
+      "loss": 0.6236,
+      "step": 1277
+    },
+    {
+      "epoch": 0.33930704898446834,
+      "grad_norm": 0.4191537794706296,
+      "learning_rate": 4.966247180225092e-06,
+      "loss": 0.6761,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3395725474578521,
+      "grad_norm": 0.36977002305745066,
+      "learning_rate": 4.966189980176844e-06,
+      "loss": 0.6244,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3398380459312359,
+      "grad_norm": 0.3697939815838604,
+      "learning_rate": 4.966132732031854e-06,
+      "loss": 0.6382,
+      "step": 1280
+    },
+    {
+      "epoch": 0.34010354440461965,
+      "grad_norm": 0.3647216931300155,
+      "learning_rate": 4.966075435791237e-06,
+      "loss": 0.6665,
+      "step": 1281
+    },
+    {
+      "epoch": 0.34036904287800346,
+      "grad_norm": 0.3877494990176763,
+      "learning_rate": 4.96601809145611e-06,
+      "loss": 0.6025,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3406345413513872,
+      "grad_norm": 0.37051861749202386,
+      "learning_rate": 4.965960699027594e-06,
+      "loss": 0.65,
+      "step": 1283
+    },
+    {
+      "epoch": 0.340900039824771,
+      "grad_norm": 0.40310083557088616,
+      "learning_rate": 4.965903258506806e-06,
+      "loss": 0.611,
+      "step": 1284
+    },
+    {
+      "epoch": 0.34116553829815477,
+      "grad_norm": 0.3805140330371282,
+      "learning_rate": 4.965845769894867e-06,
+      "loss": 0.6752,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3414310367715386,
+      "grad_norm": 0.3686700457380233,
+      "learning_rate": 4.965788233192898e-06,
+      "loss": 0.6036,
+      "step": 1286
+    },
+    {
+      "epoch": 0.34169653524492233,
+      "grad_norm": 0.38530850242444353,
+      "learning_rate": 4.965730648402021e-06,
+      "loss": 0.6322,
+      "step": 1287
+    },
+    {
+      "epoch": 0.34196203371830614,
+      "grad_norm": 0.39369770405184756,
+      "learning_rate": 4.965673015523359e-06,
+      "loss": 0.6518,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3422275321916899,
+      "grad_norm": 0.37569563153017255,
+      "learning_rate": 4.9656153345580375e-06,
+      "loss": 0.6552,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3424930306650737,
+      "grad_norm": 0.3853621467619779,
+      "learning_rate": 4.9655576055071786e-06,
+      "loss": 0.6294,
+      "step": 1290
+    },
+    {
+      "epoch": 0.34275852913845745,
+      "grad_norm": 0.38634565588761144,
+      "learning_rate": 4.96549982837191e-06,
+      "loss": 0.626,
+      "step": 1291
+    },
+    {
+      "epoch": 0.34302402761184125,
+      "grad_norm": 0.36609973965410436,
+      "learning_rate": 4.965442003153359e-06,
+      "loss": 0.6097,
+      "step": 1292
+    },
+    {
+      "epoch": 0.343289526085225,
+      "grad_norm": 0.38447990687036715,
+      "learning_rate": 4.965384129852652e-06,
+      "loss": 0.6676,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3435550245586088,
+      "grad_norm": 0.3835162570357552,
+      "learning_rate": 4.965326208470917e-06,
+      "loss": 0.5892,
+      "step": 1294
+    },
+    {
+      "epoch": 0.34382052303199256,
+      "grad_norm": 0.3680716196904,
+      "learning_rate": 4.965268239009286e-06,
+      "loss": 0.6001,
+      "step": 1295
+    },
+    {
+      "epoch": 0.34408602150537637,
+      "grad_norm": 0.38077278109458834,
+      "learning_rate": 4.965210221468888e-06,
+      "loss": 0.694,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3443515199787601,
+      "grad_norm": 0.3874965431285721,
+      "learning_rate": 4.965152155850855e-06,
+      "loss": 0.6675,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3446170184521439,
+      "grad_norm": 0.3739594745659148,
+      "learning_rate": 4.965094042156319e-06,
+      "loss": 0.6409,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3448825169255277,
+      "grad_norm": 0.3797731927777758,
+      "learning_rate": 4.965035880386414e-06,
+      "loss": 0.6611,
+      "step": 1299
+    },
+    {
+      "epoch": 0.34514801539891143,
+      "grad_norm": 0.37648059503370057,
+      "learning_rate": 4.9649776705422735e-06,
+      "loss": 0.6424,
+      "step": 1300
+    },
+    {
+      "epoch": 0.34541351387229524,
+      "grad_norm": 0.3926410453862713,
+      "learning_rate": 4.9649194126250334e-06,
+      "loss": 0.6594,
+      "step": 1301
+    },
+    {
+      "epoch": 0.345679012345679,
+      "grad_norm": 0.36104628330663807,
+      "learning_rate": 4.964861106635829e-06,
+      "loss": 0.6184,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3459445108190628,
+      "grad_norm": 0.3823712939991956,
+      "learning_rate": 4.964802752575799e-06,
+      "loss": 0.6558,
+      "step": 1303
+    },
+    {
+      "epoch": 0.34621000929244655,
+      "grad_norm": 0.3854073436797071,
+      "learning_rate": 4.96474435044608e-06,
+      "loss": 0.6405,
+      "step": 1304
+    },
+    {
+      "epoch": 0.34647550776583036,
+      "grad_norm": 0.36896677281817614,
+      "learning_rate": 4.964685900247811e-06,
+      "loss": 0.6548,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3467410062392141,
+      "grad_norm": 0.37470596844469567,
+      "learning_rate": 4.964627401982133e-06,
+      "loss": 0.6731,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3470065047125979,
+      "grad_norm": 0.3701356096650525,
+      "learning_rate": 4.964568855650186e-06,
+      "loss": 0.6402,
+      "step": 1307
+    },
+    {
+      "epoch": 0.34727200318598167,
+      "grad_norm": 0.376241418411062,
+      "learning_rate": 4.9645102612531116e-06,
+      "loss": 0.6328,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3475375016593655,
+      "grad_norm": 0.3661483028009873,
+      "learning_rate": 4.964451618792053e-06,
+      "loss": 0.62,
+      "step": 1309
+    },
+    {
+      "epoch": 0.34780300013274923,
+      "grad_norm": 0.3624725041571129,
+      "learning_rate": 4.964392928268154e-06,
+      "loss": 0.625,
+      "step": 1310
+    },
+    {
+      "epoch": 0.34806849860613304,
+      "grad_norm": 0.3870982359232708,
+      "learning_rate": 4.964334189682559e-06,
+      "loss": 0.6771,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3483339970795168,
+      "grad_norm": 0.37174773107257353,
+      "learning_rate": 4.964275403036414e-06,
+      "loss": 0.6319,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3485994955529006,
+      "grad_norm": 0.36899297737022696,
+      "learning_rate": 4.964216568330864e-06,
+      "loss": 0.6098,
+      "step": 1313
+    },
+    {
+      "epoch": 0.34886499402628435,
+      "grad_norm": 0.3804981128632557,
+      "learning_rate": 4.9641576855670585e-06,
+      "loss": 0.6761,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3491304924996681,
+      "grad_norm": 0.3804053549888035,
+      "learning_rate": 4.9640987547461445e-06,
+      "loss": 0.6596,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3493959909730519,
+      "grad_norm": 0.39034962730016304,
+      "learning_rate": 4.964039775869271e-06,
+      "loss": 0.6487,
+      "step": 1316
+    },
+    {
+      "epoch": 0.34966148944643566,
+      "grad_norm": 0.3741099395770308,
+      "learning_rate": 4.96398074893759e-06,
+      "loss": 0.6605,
+      "step": 1317
+    },
+    {
+      "epoch": 0.34992698791981947,
+      "grad_norm": 0.36271656310405986,
+      "learning_rate": 4.963921673952251e-06,
+      "loss": 0.5872,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3501924863932032,
+      "grad_norm": 0.3995062957591318,
+      "learning_rate": 4.963862550914407e-06,
+      "loss": 0.661,
+      "step": 1319
+    },
+    {
+      "epoch": 0.350457984866587,
+      "grad_norm": 0.37305610009600587,
+      "learning_rate": 4.96380337982521e-06,
+      "loss": 0.6496,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3507234833399708,
+      "grad_norm": 0.3801367848683364,
+      "learning_rate": 4.963744160685815e-06,
+      "loss": 0.6137,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3509889818133546,
+      "grad_norm": 0.3955362303938955,
+      "learning_rate": 4.963684893497377e-06,
+      "loss": 0.6405,
+      "step": 1322
+    },
+    {
+      "epoch": 0.35125448028673834,
+      "grad_norm": 0.37893790310795683,
+      "learning_rate": 4.963625578261051e-06,
+      "loss": 0.6393,
+      "step": 1323
+    },
+    {
+      "epoch": 0.35151997876012214,
+      "grad_norm": 0.3788019859897635,
+      "learning_rate": 4.963566214977994e-06,
+      "loss": 0.6382,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3517854772335059,
+      "grad_norm": 0.38654949916511866,
+      "learning_rate": 4.963506803649364e-06,
+      "loss": 0.5996,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3520509757068897,
+      "grad_norm": 0.40307788144519574,
+      "learning_rate": 4.9634473442763195e-06,
+      "loss": 0.6813,
+      "step": 1326
+    },
+    {
+      "epoch": 0.35231647418027345,
+      "grad_norm": 0.3704607316775162,
+      "learning_rate": 4.963387836860021e-06,
+      "loss": 0.6259,
+      "step": 1327
+    },
+    {
+      "epoch": 0.35258197265365726,
+      "grad_norm": 0.3735058113907317,
+      "learning_rate": 4.963328281401627e-06,
+      "loss": 0.6488,
+      "step": 1328
+    },
+    {
+      "epoch": 0.352847471127041,
+      "grad_norm": 0.3811243356459918,
+      "learning_rate": 4.963268677902302e-06,
+      "loss": 0.6577,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3531129696004248,
+      "grad_norm": 0.38674222498284955,
+      "learning_rate": 4.963209026363206e-06,
+      "loss": 0.6521,
+      "step": 1330
+    },
+    {
+      "epoch": 0.35337846807380857,
+      "grad_norm": 0.361601234266494,
+      "learning_rate": 4.963149326785502e-06,
+      "loss": 0.6282,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3536439665471924,
+      "grad_norm": 0.3979335219653062,
+      "learning_rate": 4.963089579170356e-06,
+      "loss": 0.6428,
+      "step": 1332
+    },
+    {
+      "epoch": 0.35390946502057613,
+      "grad_norm": 0.3678971039569373,
+      "learning_rate": 4.963029783518933e-06,
+      "loss": 0.6606,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3541749634939599,
+      "grad_norm": 0.36543972095556243,
+      "learning_rate": 4.962969939832398e-06,
+      "loss": 0.6243,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3544404619673437,
+      "grad_norm": 0.3927679539886193,
+      "learning_rate": 4.962910048111919e-06,
+      "loss": 0.6293,
+      "step": 1335
+    },
+    {
+      "epoch": 0.35470596044072744,
+      "grad_norm": 0.4277322929158258,
+      "learning_rate": 4.962850108358664e-06,
+      "loss": 0.6122,
+      "step": 1336
+    },
+    {
+      "epoch": 0.35497145891411125,
+      "grad_norm": 0.3697604390986625,
+      "learning_rate": 4.962790120573801e-06,
+      "loss": 0.6617,
+      "step": 1337
+    },
+    {
+      "epoch": 0.355236957387495,
+      "grad_norm": 0.4369151743487042,
+      "learning_rate": 4.962730084758501e-06,
+      "loss": 0.6256,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3555024558608788,
+      "grad_norm": 0.39076793071676813,
+      "learning_rate": 4.962670000913935e-06,
+      "loss": 0.6485,
+      "step": 1339
+    },
+    {
+      "epoch": 0.35576795433426256,
+      "grad_norm": 0.37779970531006934,
+      "learning_rate": 4.962609869041275e-06,
+      "loss": 0.6773,
+      "step": 1340
+    },
+    {
+      "epoch": 0.35603345280764637,
+      "grad_norm": 0.3876902671823763,
+      "learning_rate": 4.962549689141692e-06,
+      "loss": 0.6102,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3562989512810301,
+      "grad_norm": 0.3764840908133353,
+      "learning_rate": 4.962489461216361e-06,
+      "loss": 0.6278,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3565644497544139,
+      "grad_norm": 0.38494621232872744,
+      "learning_rate": 4.962429185266457e-06,
+      "loss": 0.6462,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3568299482277977,
+      "grad_norm": 0.37072142522297324,
+      "learning_rate": 4.962368861293153e-06,
+      "loss": 0.6499,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3570954467011815,
+      "grad_norm": 0.3717158724599488,
+      "learning_rate": 4.962308489297628e-06,
+      "loss": 0.6571,
+      "step": 1345
+    },
+    {
+      "epoch": 0.35736094517456524,
+      "grad_norm": 0.40404331801254534,
+      "learning_rate": 4.962248069281059e-06,
+      "loss": 0.6072,
+      "step": 1346
+    },
+    {
+      "epoch": 0.35762644364794904,
+      "grad_norm": 0.36336814573957393,
+      "learning_rate": 4.962187601244625e-06,
+      "loss": 0.631,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3578919421213328,
+      "grad_norm": 0.3749020997535401,
+      "learning_rate": 4.9621270851895035e-06,
+      "loss": 0.6523,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3581574405947166,
+      "grad_norm": 0.380765478062254,
+      "learning_rate": 4.962066521116875e-06,
+      "loss": 0.6101,
+      "step": 1349
+    },
+    {
+      "epoch": 0.35842293906810035,
+      "grad_norm": 0.3858701462383024,
+      "learning_rate": 4.962005909027922e-06,
+      "loss": 0.6673,
+      "step": 1350
+    },
+    {
+      "epoch": 0.35868843754148416,
+      "grad_norm": 0.3842253412518075,
+      "learning_rate": 4.961945248923825e-06,
+      "loss": 0.6508,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3589539360148679,
+      "grad_norm": 0.3682174188889406,
+      "learning_rate": 4.961884540805767e-06,
+      "loss": 0.5865,
+      "step": 1352
+    },
+    {
+      "epoch": 0.35921943448825167,
+      "grad_norm": 0.38771166101276655,
+      "learning_rate": 4.961823784674934e-06,
+      "loss": 0.6445,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3594849329616355,
+      "grad_norm": 0.3707244021777986,
+      "learning_rate": 4.96176298053251e-06,
+      "loss": 0.587,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3597504314350192,
+      "grad_norm": 0.3725300854127343,
+      "learning_rate": 4.9617021283796795e-06,
+      "loss": 0.6345,
+      "step": 1355
+    },
+    {
+      "epoch": 0.36001592990840303,
+      "grad_norm": 0.37645402418603974,
+      "learning_rate": 4.961641228217631e-06,
+      "loss": 0.6342,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3602814283817868,
+      "grad_norm": 0.3734347605020842,
+      "learning_rate": 4.96158028004755e-06,
+      "loss": 0.6655,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3605469268551706,
+      "grad_norm": 0.37275528379918504,
+      "learning_rate": 4.961519283870628e-06,
+      "loss": 0.6428,
+      "step": 1358
+    },
+    {
+      "epoch": 0.36081242532855434,
+      "grad_norm": 0.3695152405972768,
+      "learning_rate": 4.961458239688053e-06,
+      "loss": 0.6119,
+      "step": 1359
+    },
+    {
+      "epoch": 0.36107792380193815,
+      "grad_norm": 0.36621105763711564,
+      "learning_rate": 4.9613971475010145e-06,
+      "loss": 0.6152,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3613434222753219,
+      "grad_norm": 0.3719351382979484,
+      "learning_rate": 4.961336007310706e-06,
+      "loss": 0.6226,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3616089207487057,
+      "grad_norm": 0.3894427564540057,
+      "learning_rate": 4.961274819118319e-06,
+      "loss": 0.6001,
+      "step": 1362
+    },
+    {
+      "epoch": 0.36187441922208946,
+      "grad_norm": 0.37337631169965396,
+      "learning_rate": 4.961213582925046e-06,
+      "loss": 0.6212,
+      "step": 1363
+    },
+    {
+      "epoch": 0.36213991769547327,
+      "grad_norm": 0.3753722374202282,
+      "learning_rate": 4.961152298732083e-06,
+      "loss": 0.6508,
+      "step": 1364
+    },
+    {
+      "epoch": 0.362405416168857,
+      "grad_norm": 0.37578329555809076,
+      "learning_rate": 4.961090966540623e-06,
+      "loss": 0.6528,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3626709146422408,
+      "grad_norm": 0.37205930681028027,
+      "learning_rate": 4.961029586351865e-06,
+      "loss": 0.6723,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3629364131156246,
+      "grad_norm": 0.37545613857899063,
+      "learning_rate": 4.960968158167004e-06,
+      "loss": 0.6283,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3632019115890084,
+      "grad_norm": 0.36725230843341256,
+      "learning_rate": 4.960906681987238e-06,
+      "loss": 0.6508,
+      "step": 1368
+    },
+    {
+      "epoch": 0.36346741006239214,
+      "grad_norm": 0.3638765124628343,
+      "learning_rate": 4.960845157813766e-06,
+      "loss": 0.6502,
+      "step": 1369
+    },
+    {
+      "epoch": 0.36373290853577595,
+      "grad_norm": 0.382484017705505,
+      "learning_rate": 4.96078358564779e-06,
+      "loss": 0.6503,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3639984070091597,
+      "grad_norm": 0.38719996839925136,
+      "learning_rate": 4.960721965490507e-06,
+      "loss": 0.6506,
+      "step": 1371
+    },
+    {
+      "epoch": 0.36426390548254345,
+      "grad_norm": 0.36338076391017987,
+      "learning_rate": 4.960660297343122e-06,
+      "loss": 0.6311,
+      "step": 1372
+    },
+    {
+      "epoch": 0.36452940395592726,
+      "grad_norm": 0.35058788659765905,
+      "learning_rate": 4.9605985812068355e-06,
+      "loss": 0.6115,
+      "step": 1373
+    },
+    {
+      "epoch": 0.364794902429311,
+      "grad_norm": 0.39396344960920776,
+      "learning_rate": 4.960536817082853e-06,
+      "loss": 0.6346,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3650604009026948,
+      "grad_norm": 0.37178981434123,
+      "learning_rate": 4.9604750049723775e-06,
+      "loss": 0.6965,
+      "step": 1375
+    },
+    {
+      "epoch": 0.36532589937607857,
+      "grad_norm": 0.3707160699657333,
+      "learning_rate": 4.960413144876615e-06,
+      "loss": 0.6204,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3655913978494624,
+      "grad_norm": 0.36689060022358805,
+      "learning_rate": 4.960351236796772e-06,
+      "loss": 0.6772,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3658568963228461,
+      "grad_norm": 0.3891078559027146,
+      "learning_rate": 4.960289280734056e-06,
+      "loss": 0.5861,
+      "step": 1378
+    },
+    {
+      "epoch": 0.36612239479622993,
+      "grad_norm": 0.36991824090341346,
+      "learning_rate": 4.960227276689676e-06,
+      "loss": 0.651,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3663878932696137,
+      "grad_norm": 0.37552427922271137,
+      "learning_rate": 4.96016522466484e-06,
+      "loss": 0.661,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3666533917429975,
+      "grad_norm": 0.37116303316060584,
+      "learning_rate": 4.960103124660759e-06,
+      "loss": 0.6068,
+      "step": 1381
+    },
+    {
+      "epoch": 0.36691889021638124,
+      "grad_norm": 0.3817038291614676,
+      "learning_rate": 4.960040976678643e-06,
+      "loss": 0.6144,
+      "step": 1382
+    },
+    {
+      "epoch": 0.36718438868976505,
+      "grad_norm": 0.3805664191708131,
+      "learning_rate": 4.959978780719704e-06,
+      "loss": 0.6423,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3674498871631488,
+      "grad_norm": 0.37140788716933093,
+      "learning_rate": 4.959916536785157e-06,
+      "loss": 0.6314,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3677153856365326,
+      "grad_norm": 0.4224130419842736,
+      "learning_rate": 4.959854244876214e-06,
+      "loss": 0.6218,
+      "step": 1385
+    },
+    {
+      "epoch": 0.36798088410991636,
+      "grad_norm": 0.37485670103515806,
+      "learning_rate": 4.959791904994091e-06,
+      "loss": 0.6236,
+      "step": 1386
+    },
+    {
+      "epoch": 0.36824638258330017,
+      "grad_norm": 0.4620162539370904,
+      "learning_rate": 4.959729517140003e-06,
+      "loss": 0.6082,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3685118810566839,
+      "grad_norm": 0.3955744303491755,
+      "learning_rate": 4.959667081315167e-06,
+      "loss": 0.6317,
+      "step": 1388
+    },
+    {
+      "epoch": 0.36877737953006773,
+      "grad_norm": 0.3812914244875187,
+      "learning_rate": 4.9596045975208e-06,
+      "loss": 0.6167,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3690428780034515,
+      "grad_norm": 0.4354104908718216,
+      "learning_rate": 4.959542065758123e-06,
+      "loss": 0.6583,
+      "step": 1390
+    },
+    {
+      "epoch": 0.36930837647683523,
+      "grad_norm": 0.39167267251467797,
+      "learning_rate": 4.959479486028351e-06,
+      "loss": 0.6386,
+      "step": 1391
+    },
+    {
+      "epoch": 0.36957387495021904,
+      "grad_norm": 0.3743109799078965,
+      "learning_rate": 4.95941685833271e-06,
+      "loss": 0.5948,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3698393734236028,
+      "grad_norm": 0.40758134817294345,
+      "learning_rate": 4.959354182672417e-06,
+      "loss": 0.661,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3701048718969866,
+      "grad_norm": 0.421962544737718,
+      "learning_rate": 4.959291459048696e-06,
+      "loss": 0.6468,
+      "step": 1394
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 0.37499074470520627,
+      "learning_rate": 4.95922868746277e-06,
+      "loss": 0.6458,
+      "step": 1395
+    },
+    {
+      "epoch": 0.37063586884375416,
+      "grad_norm": 0.382070034558638,
+      "learning_rate": 4.9591658679158635e-06,
+      "loss": 0.6779,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3709013673171379,
+      "grad_norm": 0.4111615078041875,
+      "learning_rate": 4.959103000409202e-06,
+      "loss": 0.6291,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3711668657905217,
+      "grad_norm": 0.38482984054466113,
+      "learning_rate": 4.95904008494401e-06,
+      "loss": 0.6357,
+      "step": 1398
+    },
+    {
+      "epoch": 0.37143236426390547,
+      "grad_norm": 0.3935396036411042,
+      "learning_rate": 4.958977121521516e-06,
+      "loss": 0.6506,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3716978627372893,
+      "grad_norm": 0.4085701958297275,
+      "learning_rate": 4.9589141101429485e-06,
+      "loss": 0.6383,
+      "step": 1400
+    },
+    {
+      "epoch": 0.371963361210673,
+      "grad_norm": 0.3767915921767564,
+      "learning_rate": 4.958851050809534e-06,
+      "loss": 0.6143,
+      "step": 1401
+    },
+    {
+      "epoch": 0.37222885968405683,
+      "grad_norm": 0.3687676905981951,
+      "learning_rate": 4.958787943522505e-06,
+      "loss": 0.6469,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3724943581574406,
+      "grad_norm": 0.4003614931457942,
+      "learning_rate": 4.9587247882830905e-06,
+      "loss": 0.6387,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3727598566308244,
+      "grad_norm": 0.4035927947510916,
+      "learning_rate": 4.958661585092523e-06,
+      "loss": 0.6268,
+      "step": 1404
+    },
+    {
+      "epoch": 0.37302535510420815,
+      "grad_norm": 0.38131953814588676,
+      "learning_rate": 4.958598333952034e-06,
+      "loss": 0.6516,
+      "step": 1405
+    },
+    {
+      "epoch": 0.37329085357759195,
+      "grad_norm": 0.3764012535717099,
+      "learning_rate": 4.958535034862857e-06,
+      "loss": 0.6515,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3735563520509757,
+      "grad_norm": 0.4153391344046764,
+      "learning_rate": 4.9584716878262285e-06,
+      "loss": 0.6385,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3738218505243595,
+      "grad_norm": 0.3697704117982249,
+      "learning_rate": 4.958408292843383e-06,
+      "loss": 0.6251,
+      "step": 1408
+    },
+    {
+      "epoch": 0.37408734899774326,
+      "grad_norm": 0.38153125902827845,
+      "learning_rate": 4.958344849915556e-06,
+      "loss": 0.5958,
+      "step": 1409
+    },
+    {
+      "epoch": 0.374352847471127,
+      "grad_norm": 0.3764632124972295,
+      "learning_rate": 4.958281359043985e-06,
+      "loss": 0.6227,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3746183459445108,
+      "grad_norm": 0.3699439558545008,
+      "learning_rate": 4.958217820229909e-06,
+      "loss": 0.6523,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3748838444178946,
+      "grad_norm": 0.3787900759276024,
+      "learning_rate": 4.958154233474567e-06,
+      "loss": 0.6762,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3751493428912784,
+      "grad_norm": 0.3850429812791596,
+      "learning_rate": 4.958090598779199e-06,
+      "loss": 0.6577,
+      "step": 1413
+    },
+    {
+      "epoch": 0.37541484136466213,
+      "grad_norm": 0.37732513930131806,
+      "learning_rate": 4.9580269161450455e-06,
+      "loss": 0.5914,
+      "step": 1414
+    },
+    {
+      "epoch": 0.37568033983804594,
+      "grad_norm": 0.3908931497346101,
+      "learning_rate": 4.957963185573349e-06,
+      "loss": 0.649,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3759458383114297,
+      "grad_norm": 0.369642313827512,
+      "learning_rate": 4.957899407065352e-06,
+      "loss": 0.6427,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3762113367848135,
+      "grad_norm": 0.3696632006440851,
+      "learning_rate": 4.957835580622298e-06,
+      "loss": 0.6002,
+      "step": 1417
+    },
+    {
+      "epoch": 0.37647683525819725,
+      "grad_norm": 0.3866127348878271,
+      "learning_rate": 4.9577717062454336e-06,
+      "loss": 0.6607,
+      "step": 1418
+    },
+    {
+      "epoch": 0.37674233373158106,
+      "grad_norm": 0.3780464404386548,
+      "learning_rate": 4.957707783936002e-06,
+      "loss": 0.6527,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3770078322049648,
+      "grad_norm": 0.3912925028662332,
+      "learning_rate": 4.9576438136952525e-06,
+      "loss": 0.636,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3772733306783486,
+      "grad_norm": 0.38704762134675663,
+      "learning_rate": 4.9575797955244306e-06,
+      "loss": 0.6182,
+      "step": 1421
+    },
+    {
+      "epoch": 0.37753882915173237,
+      "grad_norm": 0.38299667748331434,
+      "learning_rate": 4.9575157294247855e-06,
+      "loss": 0.6587,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3778043276251162,
+      "grad_norm": 0.39125377495109814,
+      "learning_rate": 4.957451615397567e-06,
+      "loss": 0.6757,
+      "step": 1423
+    },
+    {
+      "epoch": 0.37806982609849993,
+      "grad_norm": 0.37718180543445884,
+      "learning_rate": 4.957387453444024e-06,
+      "loss": 0.6643,
+      "step": 1424
+    },
+    {
+      "epoch": 0.37833532457188374,
+      "grad_norm": 0.3735075398585431,
+      "learning_rate": 4.95732324356541e-06,
+      "loss": 0.6618,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3786008230452675,
+      "grad_norm": 0.3714641315208511,
+      "learning_rate": 4.957258985762976e-06,
+      "loss": 0.6656,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3788663215186513,
+      "grad_norm": 0.37303200700408046,
+      "learning_rate": 4.957194680037976e-06,
+      "loss": 0.6345,
+      "step": 1427
+    },
+    {
+      "epoch": 0.37913181999203505,
+      "grad_norm": 0.3769124858903932,
+      "learning_rate": 4.957130326391662e-06,
+      "loss": 0.6449,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3793973184654188,
+      "grad_norm": 0.4037711755596157,
+      "learning_rate": 4.957065924825291e-06,
+      "loss": 0.6188,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3796628169388026,
+      "grad_norm": 0.38385396358345264,
+      "learning_rate": 4.957001475340119e-06,
+      "loss": 0.6648,
+      "step": 1430
+    },
+    {
+      "epoch": 0.37992831541218636,
+      "grad_norm": 0.3855857786816034,
+      "learning_rate": 4.956936977937402e-06,
+      "loss": 0.632,
+      "step": 1431
+    },
+    {
+      "epoch": 0.38019381388557016,
+      "grad_norm": 0.37470720168500926,
+      "learning_rate": 4.956872432618399e-06,
+      "loss": 0.6476,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3804593123589539,
+      "grad_norm": 0.3691855564432916,
+      "learning_rate": 4.956807839384367e-06,
+      "loss": 0.6117,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3807248108323377,
+      "grad_norm": 0.3742803590304269,
+      "learning_rate": 4.9567431982365675e-06,
+      "loss": 0.6191,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3809903093057215,
+      "grad_norm": 0.38440285333850405,
+      "learning_rate": 4.956678509176262e-06,
+      "loss": 0.6245,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3812558077791053,
+      "grad_norm": 0.3766892786388694,
+      "learning_rate": 4.956613772204708e-06,
+      "loss": 0.6795,
+      "step": 1436
+    },
+    {
+      "epoch": 0.38152130625248903,
+      "grad_norm": 0.3912542692411445,
+      "learning_rate": 4.956548987323172e-06,
+      "loss": 0.6114,
+      "step": 1437
+    },
+    {
+      "epoch": 0.38178680472587284,
+      "grad_norm": 0.42336261428352623,
+      "learning_rate": 4.956484154532916e-06,
+      "loss": 0.6225,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3820523031992566,
+      "grad_norm": 0.3935056723083644,
+      "learning_rate": 4.956419273835204e-06,
+      "loss": 0.6151,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3823178016726404,
+      "grad_norm": 0.3802837261012137,
+      "learning_rate": 4.956354345231302e-06,
+      "loss": 0.6485,
+      "step": 1440
+    },
+    {
+      "epoch": 0.38258330014602415,
+      "grad_norm": 0.3835400112370889,
+      "learning_rate": 4.956289368722476e-06,
+      "loss": 0.6717,
+      "step": 1441
+    },
+    {
+      "epoch": 0.38284879861940796,
+      "grad_norm": 0.404962853402544,
+      "learning_rate": 4.956224344309993e-06,
+      "loss": 0.6466,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3831142970927917,
+      "grad_norm": 0.38818383690924424,
+      "learning_rate": 4.9561592719951225e-06,
+      "loss": 0.6598,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3833797955661755,
+      "grad_norm": 0.3898081719218938,
+      "learning_rate": 4.956094151779132e-06,
+      "loss": 0.6235,
+      "step": 1444
+    },
+    {
+      "epoch": 0.38364529403955927,
+      "grad_norm": 0.38436450299378305,
+      "learning_rate": 4.9560289836632914e-06,
+      "loss": 0.692,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3839107925129431,
+      "grad_norm": 0.48203113579878776,
+      "learning_rate": 4.955963767648872e-06,
+      "loss": 0.6295,
+      "step": 1446
+    },
+    {
+      "epoch": 0.38417629098632683,
+      "grad_norm": 0.3761191864930641,
+      "learning_rate": 4.955898503737147e-06,
+      "loss": 0.6532,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3844417894597106,
+      "grad_norm": 0.39325727838410857,
+      "learning_rate": 4.955833191929388e-06,
+      "loss": 0.6678,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3847072879330944,
+      "grad_norm": 0.5168610566283248,
+      "learning_rate": 4.955767832226868e-06,
+      "loss": 0.6481,
+      "step": 1449
+    },
+    {
+      "epoch": 0.38497278640647814,
+      "grad_norm": 0.38494809132804086,
+      "learning_rate": 4.9557024246308635e-06,
+      "loss": 0.6514,
+      "step": 1450
+    },
+    {
+      "epoch": 0.38523828487986195,
+      "grad_norm": 0.3860619969425166,
+      "learning_rate": 4.955636969142649e-06,
+      "loss": 0.6581,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3855037833532457,
+      "grad_norm": 0.4665927604403339,
+      "learning_rate": 4.9555714657635e-06,
+      "loss": 0.6122,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3857692818266295,
+      "grad_norm": 0.523630248204926,
+      "learning_rate": 4.955505914494696e-06,
+      "loss": 0.6307,
+      "step": 1453
+    },
+    {
+      "epoch": 0.38603478030001326,
+      "grad_norm": 0.38482052933862343,
+      "learning_rate": 4.9554403153375145e-06,
+      "loss": 0.6662,
+      "step": 1454
+    },
+    {
+      "epoch": 0.38630027877339707,
+      "grad_norm": 0.4378683106307078,
+      "learning_rate": 4.955374668293234e-06,
+      "loss": 0.6424,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3865657772467808,
+      "grad_norm": 0.3910977708505808,
+      "learning_rate": 4.9553089733631375e-06,
+      "loss": 0.65,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3868312757201646,
+      "grad_norm": 0.43031474737653463,
+      "learning_rate": 4.955243230548503e-06,
+      "loss": 0.6245,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.3614439227720914,
+      "learning_rate": 4.955177439850615e-06,
+      "loss": 0.6025,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3873622726669322,
+      "grad_norm": 0.39897427541295477,
+      "learning_rate": 4.955111601270754e-06,
+      "loss": 0.6153,
+      "step": 1459
+    },
+    {
+      "epoch": 0.38762777114031594,
+      "grad_norm": 0.41460765512030234,
+      "learning_rate": 4.955045714810207e-06,
+      "loss": 0.6834,
+      "step": 1460
+    },
+    {
+      "epoch": 0.38789326961369974,
+      "grad_norm": 0.387853719192773,
+      "learning_rate": 4.954979780470257e-06,
+      "loss": 0.6246,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3881587680870835,
+      "grad_norm": 0.37773562675084454,
+      "learning_rate": 4.954913798252191e-06,
+      "loss": 0.6633,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3884242665604673,
+      "grad_norm": 0.3997877485273931,
+      "learning_rate": 4.954847768157296e-06,
+      "loss": 0.5861,
+      "step": 1463
+    },
+    {
+      "epoch": 0.38868976503385105,
+      "grad_norm": 0.39331225458899316,
+      "learning_rate": 4.954781690186857e-06,
+      "loss": 0.6347,
+      "step": 1464
+    },
+    {
+      "epoch": 0.38895526350723486,
+      "grad_norm": 0.37548258213028635,
+      "learning_rate": 4.954715564342166e-06,
+      "loss": 0.67,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3892207619806186,
+      "grad_norm": 0.3730080748435241,
+      "learning_rate": 4.95464939062451e-06,
+      "loss": 0.632,
+      "step": 1466
+    },
+    {
+      "epoch": 0.38948626045400236,
+      "grad_norm": 0.3979567934107656,
+      "learning_rate": 4.954583169035183e-06,
+      "loss": 0.5809,
+      "step": 1467
+    },
+    {
+      "epoch": 0.38975175892738617,
+      "grad_norm": 0.3845961397805626,
+      "learning_rate": 4.9545168995754735e-06,
+      "loss": 0.6291,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3900172574007699,
+      "grad_norm": 0.36817228101345933,
+      "learning_rate": 4.954450582246674e-06,
+      "loss": 0.5853,
+      "step": 1469
+    },
+    {
+      "epoch": 0.39028275587415373,
+      "grad_norm": 0.38502100501420783,
+      "learning_rate": 4.954384217050079e-06,
+      "loss": 0.6273,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3905482543475375,
+      "grad_norm": 0.38305800890396074,
+      "learning_rate": 4.954317803986983e-06,
+      "loss": 0.6529,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3908137528209213,
+      "grad_norm": 0.3872047844134107,
+      "learning_rate": 4.95425134305868e-06,
+      "loss": 0.6275,
+      "step": 1472
+    },
+    {
+      "epoch": 0.39107925129430504,
+      "grad_norm": 0.39045498310880516,
+      "learning_rate": 4.954184834266467e-06,
+      "loss": 0.6491,
+      "step": 1473
+    },
+    {
+      "epoch": 0.39134474976768885,
+      "grad_norm": 0.38814459311763694,
+      "learning_rate": 4.954118277611641e-06,
+      "loss": 0.6307,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3916102482410726,
+      "grad_norm": 0.3856182405209275,
+      "learning_rate": 4.954051673095499e-06,
+      "loss": 0.6636,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3918757467144564,
+      "grad_norm": 0.36713649195996256,
+      "learning_rate": 4.953985020719342e-06,
+      "loss": 0.66,
+      "step": 1476
+    },
+    {
+      "epoch": 0.39214124518784016,
+      "grad_norm": 0.36360960011877747,
+      "learning_rate": 4.953918320484468e-06,
+      "loss": 0.6551,
+      "step": 1477
+    },
+    {
+      "epoch": 0.39240674366122397,
+      "grad_norm": 0.3693469414807645,
+      "learning_rate": 4.953851572392179e-06,
+      "loss": 0.6334,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3926722421346077,
+      "grad_norm": 0.3750301406185971,
+      "learning_rate": 4.9537847764437755e-06,
+      "loss": 0.6606,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3929377406079915,
+      "grad_norm": 0.379412127001905,
+      "learning_rate": 4.953717932640562e-06,
+      "loss": 0.6449,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3932032390813753,
+      "grad_norm": 0.3709039163120122,
+      "learning_rate": 4.95365104098384e-06,
+      "loss": 0.6671,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3934687375547591,
+      "grad_norm": 0.3712835013747682,
+      "learning_rate": 4.953584101474915e-06,
+      "loss": 0.6322,
+      "step": 1482
+    },
+    {
+      "epoch": 0.39373423602814284,
+      "grad_norm": 0.37516913432924626,
+      "learning_rate": 4.953517114115093e-06,
+      "loss": 0.6342,
+      "step": 1483
+    },
+    {
+      "epoch": 0.39399973450152664,
+      "grad_norm": 0.3832057761510323,
+      "learning_rate": 4.95345007890568e-06,
+      "loss": 0.6203,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3942652329749104,
+      "grad_norm": 0.38388009158934117,
+      "learning_rate": 4.953382995847984e-06,
+      "loss": 0.6668,
+      "step": 1485
+    },
+    {
+      "epoch": 0.39453073144829415,
+      "grad_norm": 0.40003373474073817,
+      "learning_rate": 4.953315864943312e-06,
+      "loss": 0.6346,
+      "step": 1486
+    },
+    {
+      "epoch": 0.39479622992167795,
+      "grad_norm": 0.37682658361850974,
+      "learning_rate": 4.9532486861929745e-06,
+      "loss": 0.6524,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3950617283950617,
+      "grad_norm": 0.3767846630718036,
+      "learning_rate": 4.953181459598281e-06,
+      "loss": 0.6459,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3953272268684455,
+      "grad_norm": 0.3812495439451251,
+      "learning_rate": 4.9531141851605426e-06,
+      "loss": 0.6193,
+      "step": 1489
+    },
+    {
+      "epoch": 0.39559272534182927,
+      "grad_norm": 0.3846612978316415,
+      "learning_rate": 4.953046862881071e-06,
+      "loss": 0.6557,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3958582238152131,
+      "grad_norm": 0.37315842183341713,
+      "learning_rate": 4.952979492761179e-06,
+      "loss": 0.6171,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3961237222885968,
+      "grad_norm": 0.3956807307227741,
+      "learning_rate": 4.952912074802182e-06,
+      "loss": 0.6503,
+      "step": 1492
+    },
+    {
+      "epoch": 0.39638922076198063,
+      "grad_norm": 0.3757489994884287,
+      "learning_rate": 4.952844609005393e-06,
+      "loss": 0.643,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3966547192353644,
+      "grad_norm": 0.3888396186642242,
+      "learning_rate": 4.9527770953721285e-06,
+      "loss": 0.657,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3969202177087482,
+      "grad_norm": 0.3714727391950251,
+      "learning_rate": 4.952709533903706e-06,
+      "loss": 0.6379,
+      "step": 1495
+    },
+    {
+      "epoch": 0.39718571618213194,
+      "grad_norm": 0.3741917030539026,
+      "learning_rate": 4.9526419246014415e-06,
+      "loss": 0.6292,
+      "step": 1496
+    },
+    {
+      "epoch": 0.39745121465551575,
+      "grad_norm": 0.3694705283973955,
+      "learning_rate": 4.952574267466655e-06,
+      "loss": 0.5826,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3977167131288995,
+      "grad_norm": 0.3706702336109769,
+      "learning_rate": 4.9525065625006645e-06,
+      "loss": 0.6458,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3979822116022833,
+      "grad_norm": 0.3569730233874185,
+      "learning_rate": 4.952438809704791e-06,
+      "loss": 0.6174,
+      "step": 1499
+    },
+    {
+      "epoch": 0.39824771007566706,
+      "grad_norm": 0.37982738830536933,
+      "learning_rate": 4.9523710090803576e-06,
+      "loss": 0.6178,
+      "step": 1500
+    },
+    {
+      "epoch": 0.39851320854905087,
+      "grad_norm": 0.3743163991683717,
+      "learning_rate": 4.952303160628684e-06,
+      "loss": 0.6524,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3987787070224346,
+      "grad_norm": 0.37351588481447573,
+      "learning_rate": 4.952235264351095e-06,
+      "loss": 0.6393,
+      "step": 1502
+    },
+    {
+      "epoch": 0.39904420549581837,
+      "grad_norm": 0.3965425276450105,
+      "learning_rate": 4.952167320248914e-06,
+      "loss": 0.6372,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3993097039692022,
+      "grad_norm": 0.37300389314962357,
+      "learning_rate": 4.952099328323466e-06,
+      "loss": 0.6502,
+      "step": 1504
+    },
+    {
+      "epoch": 0.39957520244258593,
+      "grad_norm": 0.3730643040164551,
+      "learning_rate": 4.952031288576078e-06,
+      "loss": 0.619,
+      "step": 1505
+    },
+    {
+      "epoch": 0.39984070091596974,
+      "grad_norm": 0.37593313637519654,
+      "learning_rate": 4.9519632010080765e-06,
+      "loss": 0.558,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4001061993893535,
+      "grad_norm": 0.40376294370648996,
+      "learning_rate": 4.951895065620789e-06,
+      "loss": 0.6297,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4003716978627373,
+      "grad_norm": 0.3827676631681475,
+      "learning_rate": 4.951826882415544e-06,
+      "loss": 0.6064,
+      "step": 1508
+    },
+    {
+      "epoch": 0.40063719633612105,
+      "grad_norm": 0.3902324075672343,
+      "learning_rate": 4.951758651393672e-06,
+      "loss": 0.64,
+      "step": 1509
+    },
+    {
+      "epoch": 0.40090269480950486,
+      "grad_norm": 0.37491893902961715,
+      "learning_rate": 4.951690372556502e-06,
+      "loss": 0.633,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4011681932828886,
+      "grad_norm": 0.37754628982436195,
+      "learning_rate": 4.9516220459053685e-06,
+      "loss": 0.6528,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4014336917562724,
+      "grad_norm": 0.3725784806878786,
+      "learning_rate": 4.951553671441603e-06,
+      "loss": 0.6338,
+      "step": 1512
+    },
+    {
+      "epoch": 0.40169919022965617,
+      "grad_norm": 0.3866060610989434,
+      "learning_rate": 4.951485249166537e-06,
+      "loss": 0.6666,
+      "step": 1513
+    },
+    {
+      "epoch": 0.40196468870304,
+      "grad_norm": 0.37658539033153926,
+      "learning_rate": 4.951416779081507e-06,
+      "loss": 0.6554,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4022301871764237,
+      "grad_norm": 0.38017348322180383,
+      "learning_rate": 4.951348261187847e-06,
+      "loss": 0.6725,
+      "step": 1515
+    },
+    {
+      "epoch": 0.40249568564980753,
+      "grad_norm": 0.3592714783102323,
+      "learning_rate": 4.951279695486895e-06,
+      "loss": 0.6006,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4027611841231913,
+      "grad_norm": 0.38051905290563476,
+      "learning_rate": 4.951211081979987e-06,
+      "loss": 0.6384,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4030266825965751,
+      "grad_norm": 0.363536852764823,
+      "learning_rate": 4.951142420668461e-06,
+      "loss": 0.6224,
+      "step": 1518
+    },
+    {
+      "epoch": 0.40329218106995884,
+      "grad_norm": 0.3890642636995573,
+      "learning_rate": 4.951073711553657e-06,
+      "loss": 0.6358,
+      "step": 1519
+    },
+    {
+      "epoch": 0.40355767954334265,
+      "grad_norm": 0.39445351972956,
+      "learning_rate": 4.951004954636913e-06,
+      "loss": 0.6595,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4038231780167264,
+      "grad_norm": 0.3982213687881686,
+      "learning_rate": 4.950936149919572e-06,
+      "loss": 0.6027,
+      "step": 1521
+    },
+    {
+      "epoch": 0.40408867649011015,
+      "grad_norm": 0.3854249110118144,
+      "learning_rate": 4.950867297402976e-06,
+      "loss": 0.643,
+      "step": 1522
+    },
+    {
+      "epoch": 0.40435417496349396,
+      "grad_norm": 0.3581356017525054,
+      "learning_rate": 4.950798397088465e-06,
+      "loss": 0.6192,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4046196734368777,
+      "grad_norm": 0.3773246357129126,
+      "learning_rate": 4.950729448977386e-06,
+      "loss": 0.6732,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4048851719102615,
+      "grad_norm": 0.3815722803658837,
+      "learning_rate": 4.9506604530710825e-06,
+      "loss": 0.6361,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4051506703836453,
+      "grad_norm": 0.3664922401679366,
+      "learning_rate": 4.9505914093709e-06,
+      "loss": 0.5954,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4054161688570291,
+      "grad_norm": 0.37563904913174356,
+      "learning_rate": 4.950522317878184e-06,
+      "loss": 0.5986,
+      "step": 1527
+    },
+    {
+      "epoch": 0.40568166733041283,
+      "grad_norm": 0.3892051459536696,
+      "learning_rate": 4.950453178594283e-06,
+      "loss": 0.6038,
+      "step": 1528
+    },
+    {
+      "epoch": 0.40594716580379664,
+      "grad_norm": 0.39694643329554125,
+      "learning_rate": 4.950383991520547e-06,
+      "loss": 0.6585,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4062126642771804,
+      "grad_norm": 0.4083925886587904,
+      "learning_rate": 4.950314756658321e-06,
+      "loss": 0.6524,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4064781627505642,
+      "grad_norm": 0.3759360707159121,
+      "learning_rate": 4.95024547400896e-06,
+      "loss": 0.65,
+      "step": 1531
+    },
+    {
+      "epoch": 0.40674366122394795,
+      "grad_norm": 0.387354723117181,
+      "learning_rate": 4.9501761435738115e-06,
+      "loss": 0.6549,
+      "step": 1532
+    },
+    {
+      "epoch": 0.40700915969733176,
+      "grad_norm": 0.3726027294176786,
+      "learning_rate": 4.95010676535423e-06,
+      "loss": 0.6367,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4072746581707155,
+      "grad_norm": 0.3741693104713276,
+      "learning_rate": 4.950037339351568e-06,
+      "loss": 0.6649,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4075401566440993,
+      "grad_norm": 0.3760665944438138,
+      "learning_rate": 4.949967865567178e-06,
+      "loss": 0.6768,
+      "step": 1535
+    },
+    {
+      "epoch": 0.40780565511748307,
+      "grad_norm": 0.37782839437812066,
+      "learning_rate": 4.949898344002417e-06,
+      "loss": 0.6394,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4080711535908669,
+      "grad_norm": 0.37257726173594624,
+      "learning_rate": 4.949828774658639e-06,
+      "loss": 0.6229,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4083366520642506,
+      "grad_norm": 0.3770754472393843,
+      "learning_rate": 4.949759157537203e-06,
+      "loss": 0.6557,
+      "step": 1538
+    },
+    {
+      "epoch": 0.40860215053763443,
+      "grad_norm": 0.37313936514262486,
+      "learning_rate": 4.949689492639465e-06,
+      "loss": 0.6122,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4088676490110182,
+      "grad_norm": 0.39078374874637534,
+      "learning_rate": 4.949619779966783e-06,
+      "loss": 0.6443,
+      "step": 1540
+    },
+    {
+      "epoch": 0.40913314748440194,
+      "grad_norm": 0.3888752356200754,
+      "learning_rate": 4.949550019520518e-06,
+      "loss": 0.6644,
+      "step": 1541
+    },
+    {
+      "epoch": 0.40939864595778575,
+      "grad_norm": 0.3973044570124634,
+      "learning_rate": 4.949480211302031e-06,
+      "loss": 0.6582,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4096641444311695,
+      "grad_norm": 0.38694257415978767,
+      "learning_rate": 4.949410355312681e-06,
+      "loss": 0.6428,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4099296429045533,
+      "grad_norm": 0.3749325764698057,
+      "learning_rate": 4.949340451553834e-06,
+      "loss": 0.6349,
+      "step": 1544
+    },
+    {
+      "epoch": 0.41019514137793706,
+      "grad_norm": 0.3744298373161061,
+      "learning_rate": 4.94927050002685e-06,
+      "loss": 0.6093,
+      "step": 1545
+    },
+    {
+      "epoch": 0.41046063985132086,
+      "grad_norm": 0.3936594556322945,
+      "learning_rate": 4.949200500733093e-06,
+      "loss": 0.6514,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4107261383247046,
+      "grad_norm": 0.3760700440461478,
+      "learning_rate": 4.949130453673932e-06,
+      "loss": 0.6306,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4109916367980884,
+      "grad_norm": 0.37535800885008996,
+      "learning_rate": 4.949060358850729e-06,
+      "loss": 0.6239,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4112571352714722,
+      "grad_norm": 0.37688638991759393,
+      "learning_rate": 4.948990216264853e-06,
+      "loss": 0.6442,
+      "step": 1549
+    },
+    {
+      "epoch": 0.411522633744856,
+      "grad_norm": 0.369621358837807,
+      "learning_rate": 4.9489200259176715e-06,
+      "loss": 0.6354,
+      "step": 1550
+    },
+    {
+      "epoch": 0.41178813221823973,
+      "grad_norm": 0.4001845431668011,
+      "learning_rate": 4.948849787810554e-06,
+      "loss": 0.606,
+      "step": 1551
+    },
+    {
+      "epoch": 0.41205363069162354,
+      "grad_norm": 0.3801373776920019,
+      "learning_rate": 4.9487795019448695e-06,
+      "loss": 0.6328,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4123191291650073,
+      "grad_norm": 0.38123105172776506,
+      "learning_rate": 4.948709168321989e-06,
+      "loss": 0.5982,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4125846276383911,
+      "grad_norm": 0.37581282507433045,
+      "learning_rate": 4.948638786943285e-06,
+      "loss": 0.6146,
+      "step": 1554
+    },
+    {
+      "epoch": 0.41285012611177485,
+      "grad_norm": 0.38060785345567816,
+      "learning_rate": 4.948568357810128e-06,
+      "loss": 0.6446,
+      "step": 1555
+    },
+    {
+      "epoch": 0.41311562458515866,
+      "grad_norm": 0.37141398664019587,
+      "learning_rate": 4.9484978809238945e-06,
+      "loss": 0.6678,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4133811230585424,
+      "grad_norm": 0.37935883805892023,
+      "learning_rate": 4.948427356285956e-06,
+      "loss": 0.6352,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4136466215319262,
+      "grad_norm": 0.3816316230215305,
+      "learning_rate": 4.94835678389769e-06,
+      "loss": 0.6057,
+      "step": 1558
+    },
+    {
+      "epoch": 0.41391212000530997,
+      "grad_norm": 0.39538492076129317,
+      "learning_rate": 4.9482861637604725e-06,
+      "loss": 0.6067,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4141776184786937,
+      "grad_norm": 0.36542160119434036,
+      "learning_rate": 4.94821549587568e-06,
+      "loss": 0.6211,
+      "step": 1560
+    },
+    {
+      "epoch": 0.41444311695207753,
+      "grad_norm": 0.39548725338996055,
+      "learning_rate": 4.94814478024469e-06,
+      "loss": 0.5854,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4147086154254613,
+      "grad_norm": 0.3836340813617557,
+      "learning_rate": 4.948074016868884e-06,
+      "loss": 0.6707,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4149741138988451,
+      "grad_norm": 0.40700202796436696,
+      "learning_rate": 4.94800320574964e-06,
+      "loss": 0.6146,
+      "step": 1563
+    },
+    {
+      "epoch": 0.41523961237222884,
+      "grad_norm": 0.3900921802915412,
+      "learning_rate": 4.94793234688834e-06,
+      "loss": 0.6699,
+      "step": 1564
+    },
+    {
+      "epoch": 0.41550511084561265,
+      "grad_norm": 0.38383660877028974,
+      "learning_rate": 4.947861440286365e-06,
+      "loss": 0.6066,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4157706093189964,
+      "grad_norm": 0.3811570065982231,
+      "learning_rate": 4.9477904859450994e-06,
+      "loss": 0.6778,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4160361077923802,
+      "grad_norm": 0.39264849903297744,
+      "learning_rate": 4.947719483865927e-06,
+      "loss": 0.6468,
+      "step": 1567
+    },
+    {
+      "epoch": 0.41630160626576396,
+      "grad_norm": 0.3753239643561131,
+      "learning_rate": 4.947648434050231e-06,
+      "loss": 0.626,
+      "step": 1568
+    },
+    {
+      "epoch": 0.41656710473914776,
+      "grad_norm": 0.4050929356879002,
+      "learning_rate": 4.947577336499396e-06,
+      "loss": 0.6342,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4168326032125315,
+      "grad_norm": 0.39462571695589943,
+      "learning_rate": 4.947506191214812e-06,
+      "loss": 0.6337,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4170981016859153,
+      "grad_norm": 0.38905594357205586,
+      "learning_rate": 4.947434998197864e-06,
+      "loss": 0.6384,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4173636001592991,
+      "grad_norm": 0.3804850730806265,
+      "learning_rate": 4.947363757449941e-06,
+      "loss": 0.6482,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4176290986326829,
+      "grad_norm": 0.3818377216918663,
+      "learning_rate": 4.947292468972433e-06,
+      "loss": 0.6422,
+      "step": 1573
+    },
+    {
+      "epoch": 0.41789459710606663,
+      "grad_norm": 0.39699057776833674,
+      "learning_rate": 4.94722113276673e-06,
+      "loss": 0.6565,
+      "step": 1574
+    },
+    {
+      "epoch": 0.41816009557945044,
+      "grad_norm": 0.42960285906165224,
+      "learning_rate": 4.947149748834223e-06,
+      "loss": 0.6426,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4184255940528342,
+      "grad_norm": 0.37287793567883765,
+      "learning_rate": 4.947078317176304e-06,
+      "loss": 0.6169,
+      "step": 1576
+    },
+    {
+      "epoch": 0.418691092526218,
+      "grad_norm": 0.3733878120090327,
+      "learning_rate": 4.9470068377943665e-06,
+      "loss": 0.6249,
+      "step": 1577
+    },
+    {
+      "epoch": 0.41895659099960175,
+      "grad_norm": 0.40996754506660293,
+      "learning_rate": 4.9469353106898046e-06,
+      "loss": 0.6497,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4192220894729855,
+      "grad_norm": 0.3808394070889554,
+      "learning_rate": 4.946863735864013e-06,
+      "loss": 0.6373,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4194875879463693,
+      "grad_norm": 0.4072587890017732,
+      "learning_rate": 4.9467921133183864e-06,
+      "loss": 0.605,
+      "step": 1580
+    },
+    {
+      "epoch": 0.41975308641975306,
+      "grad_norm": 0.3963611100653441,
+      "learning_rate": 4.946720443054324e-06,
+      "loss": 0.6004,
+      "step": 1581
+    },
+    {
+      "epoch": 0.42001858489313687,
+      "grad_norm": 0.37052249405576043,
+      "learning_rate": 4.9466487250732224e-06,
+      "loss": 0.6336,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4202840833665206,
+      "grad_norm": 0.38606938014253195,
+      "learning_rate": 4.946576959376479e-06,
+      "loss": 0.6435,
+      "step": 1583
+    },
+    {
+      "epoch": 0.42054958183990443,
+      "grad_norm": 0.4154998029038964,
+      "learning_rate": 4.946505145965496e-06,
+      "loss": 0.6658,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4208150803132882,
+      "grad_norm": 0.38187431149946094,
+      "learning_rate": 4.946433284841672e-06,
+      "loss": 0.6683,
+      "step": 1585
+    },
+    {
+      "epoch": 0.421080578786672,
+      "grad_norm": 0.38575420325722015,
+      "learning_rate": 4.946361376006409e-06,
+      "loss": 0.6506,
+      "step": 1586
+    },
+    {
+      "epoch": 0.42134607726005574,
+      "grad_norm": 0.3689311821011402,
+      "learning_rate": 4.94628941946111e-06,
+      "loss": 0.6713,
+      "step": 1587
+    },
+    {
+      "epoch": 0.42161157573343955,
+      "grad_norm": 0.3651451466414451,
+      "learning_rate": 4.946217415207177e-06,
+      "loss": 0.6379,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4218770742068233,
+      "grad_norm": 0.39532985642846613,
+      "learning_rate": 4.946145363246015e-06,
+      "loss": 0.6169,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4221425726802071,
+      "grad_norm": 0.36428085953941064,
+      "learning_rate": 4.94607326357903e-06,
+      "loss": 0.5861,
+      "step": 1590
+    },
+    {
+      "epoch": 0.42240807115359086,
+      "grad_norm": 0.3899354145866775,
+      "learning_rate": 4.946001116207627e-06,
+      "loss": 0.5937,
+      "step": 1591
+    },
+    {
+      "epoch": 0.42267356962697467,
+      "grad_norm": 0.38357067102826686,
+      "learning_rate": 4.9459289211332134e-06,
+      "loss": 0.6555,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4229390681003584,
+      "grad_norm": 0.3804977641869738,
+      "learning_rate": 4.945856678357197e-06,
+      "loss": 0.5971,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4232045665737422,
+      "grad_norm": 0.38452688135530994,
+      "learning_rate": 4.9457843878809875e-06,
+      "loss": 0.6161,
+      "step": 1594
+    },
+    {
+      "epoch": 0.423470065047126,
+      "grad_norm": 0.37575089193213984,
+      "learning_rate": 4.9457120497059926e-06,
+      "loss": 0.6672,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4237355635205098,
+      "grad_norm": 0.39086886708302315,
+      "learning_rate": 4.945639663833626e-06,
+      "loss": 0.6511,
+      "step": 1596
+    },
+    {
+      "epoch": 0.42400106199389354,
+      "grad_norm": 0.3942861296902854,
+      "learning_rate": 4.945567230265298e-06,
+      "loss": 0.7053,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4242665604672773,
+      "grad_norm": 0.3867998815391152,
+      "learning_rate": 4.945494749002421e-06,
+      "loss": 0.6077,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4245320589406611,
+      "grad_norm": 0.3766080561335471,
+      "learning_rate": 4.945422220046409e-06,
+      "loss": 0.5895,
+      "step": 1599
+    },
+    {
+      "epoch": 0.42479755741404485,
+      "grad_norm": 0.3835422444876443,
+      "learning_rate": 4.945349643398675e-06,
+      "loss": 0.6336,
+      "step": 1600
+    },
+    {
+      "epoch": 0.42506305588742865,
+      "grad_norm": 0.4052010590526513,
+      "learning_rate": 4.945277019060637e-06,
+      "loss": 0.6608,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4253285543608124,
+      "grad_norm": 0.4151652082319791,
+      "learning_rate": 4.94520434703371e-06,
+      "loss": 0.6407,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4255940528341962,
+      "grad_norm": 0.3783049756753611,
+      "learning_rate": 4.945131627319311e-06,
+      "loss": 0.6689,
+      "step": 1603
+    },
+    {
+      "epoch": 0.42585955130757996,
+      "grad_norm": 0.3964485947247106,
+      "learning_rate": 4.945058859918858e-06,
+      "loss": 0.6417,
+      "step": 1604
+    },
+    {
+      "epoch": 0.42612504978096377,
+      "grad_norm": 0.3893080871437303,
+      "learning_rate": 4.944986044833772e-06,
+      "loss": 0.654,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4263905482543475,
+      "grad_norm": 0.3930411477685514,
+      "learning_rate": 4.94491318206547e-06,
+      "loss": 0.6768,
+      "step": 1606
+    },
+    {
+      "epoch": 0.42665604672773133,
+      "grad_norm": 0.38547954004412693,
+      "learning_rate": 4.944840271615376e-06,
+      "loss": 0.6231,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4269215452011151,
+      "grad_norm": 0.3743913916645727,
+      "learning_rate": 4.94476731348491e-06,
+      "loss": 0.6513,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4271870436744989,
+      "grad_norm": 0.3869653809582928,
+      "learning_rate": 4.944694307675497e-06,
+      "loss": 0.6269,
+      "step": 1609
+    },
+    {
+      "epoch": 0.42745254214788264,
+      "grad_norm": 0.37519303574952695,
+      "learning_rate": 4.944621254188558e-06,
+      "loss": 0.6372,
+      "step": 1610
+    },
+    {
+      "epoch": 0.42771804062126645,
+      "grad_norm": 0.3921145753836486,
+      "learning_rate": 4.94454815302552e-06,
+      "loss": 0.594,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4279835390946502,
+      "grad_norm": 0.39545167671228976,
+      "learning_rate": 4.944475004187806e-06,
+      "loss": 0.6411,
+      "step": 1612
+    },
+    {
+      "epoch": 0.428249037568034,
+      "grad_norm": 0.37658542065485245,
+      "learning_rate": 4.944401807676846e-06,
+      "loss": 0.6153,
+      "step": 1613
+    },
+    {
+      "epoch": 0.42851453604141776,
+      "grad_norm": 0.3690547653104833,
+      "learning_rate": 4.944328563494066e-06,
+      "loss": 0.641,
+      "step": 1614
+    },
+    {
+      "epoch": 0.42878003451480157,
+      "grad_norm": 0.3816290597461041,
+      "learning_rate": 4.944255271640893e-06,
+      "loss": 0.6428,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4290455329881853,
+      "grad_norm": 0.3846315553712388,
+      "learning_rate": 4.944181932118759e-06,
+      "loss": 0.6326,
+      "step": 1616
+    },
+    {
+      "epoch": 0.42931103146156907,
+      "grad_norm": 0.3822047578943089,
+      "learning_rate": 4.944108544929091e-06,
+      "loss": 0.6101,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4295765299349529,
+      "grad_norm": 0.3772806855454925,
+      "learning_rate": 4.944035110073324e-06,
+      "loss": 0.5933,
+      "step": 1618
+    },
+    {
+      "epoch": 0.42984202840833663,
+      "grad_norm": 0.3813315082184482,
+      "learning_rate": 4.943961627552888e-06,
+      "loss": 0.6484,
+      "step": 1619
+    },
+    {
+      "epoch": 0.43010752688172044,
+      "grad_norm": 0.39679355084632945,
+      "learning_rate": 4.943888097369216e-06,
+      "loss": 0.6393,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4303730253551042,
+      "grad_norm": 0.38731039460318645,
+      "learning_rate": 4.943814519523742e-06,
+      "loss": 0.6418,
+      "step": 1621
+    },
+    {
+      "epoch": 0.430638523828488,
+      "grad_norm": 0.38140765610596006,
+      "learning_rate": 4.943740894017902e-06,
+      "loss": 0.6612,
+      "step": 1622
+    },
+    {
+      "epoch": 0.43090402230187175,
+      "grad_norm": 0.37436354173900305,
+      "learning_rate": 4.9436672208531315e-06,
+      "loss": 0.6436,
+      "step": 1623
+    },
+    {
+      "epoch": 0.43116952077525555,
+      "grad_norm": 0.4107710848666306,
+      "learning_rate": 4.943593500030867e-06,
+      "loss": 0.6232,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4314350192486393,
+      "grad_norm": 0.3930193351242004,
+      "learning_rate": 4.943519731552546e-06,
+      "loss": 0.6473,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4317005177220231,
+      "grad_norm": 0.3925443383035921,
+      "learning_rate": 4.943445915419607e-06,
+      "loss": 0.6219,
+      "step": 1626
+    },
+    {
+      "epoch": 0.43196601619540687,
+      "grad_norm": 0.4028359292563561,
+      "learning_rate": 4.94337205163349e-06,
+      "loss": 0.5811,
+      "step": 1627
+    },
+    {
+      "epoch": 0.4322315146687907,
+      "grad_norm": 0.3753775744625516,
+      "learning_rate": 4.943298140195637e-06,
+      "loss": 0.6316,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4324970131421744,
+      "grad_norm": 0.3714009589504834,
+      "learning_rate": 4.9432241811074875e-06,
+      "loss": 0.6593,
+      "step": 1629
+    },
+    {
+      "epoch": 0.43276251161555823,
+      "grad_norm": 0.39135741376656336,
+      "learning_rate": 4.943150174370483e-06,
+      "loss": 0.6096,
+      "step": 1630
+    },
+    {
+      "epoch": 0.433028010088942,
+      "grad_norm": 0.3656532922497972,
+      "learning_rate": 4.94307611998607e-06,
+      "loss": 0.659,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4332935085623258,
+      "grad_norm": 0.38751350689028435,
+      "learning_rate": 4.94300201795569e-06,
+      "loss": 0.6614,
+      "step": 1632
+    },
+    {
+      "epoch": 0.43355900703570954,
+      "grad_norm": 0.44445393479364587,
+      "learning_rate": 4.94292786828079e-06,
+      "loss": 0.6134,
+      "step": 1633
+    },
+    {
+      "epoch": 0.43382450550909335,
+      "grad_norm": 0.3709095673236794,
+      "learning_rate": 4.942853670962815e-06,
+      "loss": 0.6643,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4340900039824771,
+      "grad_norm": 0.38598086806624593,
+      "learning_rate": 4.942779426003211e-06,
+      "loss": 0.6083,
+      "step": 1635
+    },
+    {
+      "epoch": 0.43435550245586085,
+      "grad_norm": 0.47269179740172845,
+      "learning_rate": 4.942705133403429e-06,
+      "loss": 0.5798,
+      "step": 1636
+    },
+    {
+      "epoch": 0.43462100092924466,
+      "grad_norm": 0.39146512560079444,
+      "learning_rate": 4.942630793164916e-06,
+      "loss": 0.632,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4348864994026284,
+      "grad_norm": 0.3929720894985973,
+      "learning_rate": 4.94255640528912e-06,
+      "loss": 0.6353,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4351519978760122,
+      "grad_norm": 0.3906718711554545,
+      "learning_rate": 4.942481969777496e-06,
+      "loss": 0.6829,
+      "step": 1639
+    },
+    {
+      "epoch": 0.43541749634939597,
+      "grad_norm": 0.39182488585389164,
+      "learning_rate": 4.942407486631492e-06,
+      "loss": 0.6656,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4356829948227798,
+      "grad_norm": 0.39462909347772407,
+      "learning_rate": 4.942332955852563e-06,
+      "loss": 0.6509,
+      "step": 1641
+    },
+    {
+      "epoch": 0.43594849329616353,
+      "grad_norm": 0.3940803765074498,
+      "learning_rate": 4.94225837744216e-06,
+      "loss": 0.6401,
+      "step": 1642
+    },
+    {
+      "epoch": 0.43621399176954734,
+      "grad_norm": 0.3941480372471297,
+      "learning_rate": 4.94218375140174e-06,
+      "loss": 0.6237,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4364794902429311,
+      "grad_norm": 0.38749929224259866,
+      "learning_rate": 4.942109077732757e-06,
+      "loss": 0.6449,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4367449887163149,
+      "grad_norm": 0.38312830400750875,
+      "learning_rate": 4.942034356436668e-06,
+      "loss": 0.6073,
+      "step": 1645
+    },
+    {
+      "epoch": 0.43701048718969865,
+      "grad_norm": 0.3862796603844449,
+      "learning_rate": 4.9419595875149294e-06,
+      "loss": 0.6843,
+      "step": 1646
+    },
+    {
+      "epoch": 0.43727598566308246,
+      "grad_norm": 0.3873473046706782,
+      "learning_rate": 4.941884770969e-06,
+      "loss": 0.6698,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4375414841364662,
+      "grad_norm": 0.38685093293955375,
+      "learning_rate": 4.9418099068003386e-06,
+      "loss": 0.6379,
+      "step": 1648
+    },
+    {
+      "epoch": 0.43780698260985,
+      "grad_norm": 0.38332926684140883,
+      "learning_rate": 4.941734995010406e-06,
+      "loss": 0.6275,
+      "step": 1649
+    },
+    {
+      "epoch": 0.43807248108323377,
+      "grad_norm": 0.3724629035094914,
+      "learning_rate": 4.941660035600663e-06,
+      "loss": 0.6679,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4383379795566176,
+      "grad_norm": 0.43837860308373383,
+      "learning_rate": 4.941585028572571e-06,
+      "loss": 0.5832,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4386034780300013,
+      "grad_norm": 0.38785275403811387,
+      "learning_rate": 4.941509973927591e-06,
+      "loss": 0.6672,
+      "step": 1652
+    },
+    {
+      "epoch": 0.43886897650338513,
+      "grad_norm": 0.37736372400950186,
+      "learning_rate": 4.9414348716671915e-06,
+      "loss": 0.6422,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4391344749767689,
+      "grad_norm": 0.40750027812390716,
+      "learning_rate": 4.941359721792832e-06,
+      "loss": 0.657,
+      "step": 1654
+    },
+    {
+      "epoch": 0.43939997345015264,
+      "grad_norm": 0.4073221870315453,
+      "learning_rate": 4.941284524305982e-06,
+      "loss": 0.6192,
+      "step": 1655
+    },
+    {
+      "epoch": 0.43966547192353644,
+      "grad_norm": 0.38725402474887244,
+      "learning_rate": 4.9412092792081055e-06,
+      "loss": 0.6892,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4399309703969202,
+      "grad_norm": 0.3950293089845777,
+      "learning_rate": 4.941133986500671e-06,
+      "loss": 0.6196,
+      "step": 1657
+    },
+    {
+      "epoch": 0.440196468870304,
+      "grad_norm": 0.41446212489566236,
+      "learning_rate": 4.941058646185148e-06,
+      "loss": 0.6061,
+      "step": 1658
+    },
+    {
+      "epoch": 0.44046196734368775,
+      "grad_norm": 0.3828708722445977,
+      "learning_rate": 4.9409832582630036e-06,
+      "loss": 0.6443,
+      "step": 1659
+    },
+    {
+      "epoch": 0.44072746581707156,
+      "grad_norm": 0.40221057778277863,
+      "learning_rate": 4.940907822735709e-06,
+      "loss": 0.6342,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4409929642904553,
+      "grad_norm": 0.40136523177880523,
+      "learning_rate": 4.9408323396047365e-06,
+      "loss": 0.596,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4412584627638391,
+      "grad_norm": 0.38910095784161536,
+      "learning_rate": 4.9407568088715565e-06,
+      "loss": 0.6162,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4415239612372229,
+      "grad_norm": 0.4184623613488529,
+      "learning_rate": 4.940681230537643e-06,
+      "loss": 0.6483,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4417894597106067,
+      "grad_norm": 0.3815204813473002,
+      "learning_rate": 4.940605604604469e-06,
+      "loss": 0.631,
+      "step": 1664
+    },
+    {
+      "epoch": 0.44205495818399043,
+      "grad_norm": 0.3791833852379832,
+      "learning_rate": 4.94052993107351e-06,
+      "loss": 0.6341,
+      "step": 1665
+    },
+    {
+      "epoch": 0.44232045665737424,
+      "grad_norm": 0.3956966500277488,
+      "learning_rate": 4.940454209946242e-06,
+      "loss": 0.6043,
+      "step": 1666
+    },
+    {
+      "epoch": 0.442585955130758,
+      "grad_norm": 0.38571208556788744,
+      "learning_rate": 4.940378441224143e-06,
+      "loss": 0.6706,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4428514536041418,
+      "grad_norm": 0.3749164199841067,
+      "learning_rate": 4.940302624908689e-06,
+      "loss": 0.6124,
+      "step": 1668
+    },
+    {
+      "epoch": 0.44311695207752555,
+      "grad_norm": 0.3915145469847177,
+      "learning_rate": 4.9402267610013575e-06,
+      "loss": 0.6295,
+      "step": 1669
+    },
+    {
+      "epoch": 0.44338245055090936,
+      "grad_norm": 0.3759360062483427,
+      "learning_rate": 4.94015084950363e-06,
+      "loss": 0.6737,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4436479490242931,
+      "grad_norm": 0.37937027922396005,
+      "learning_rate": 4.940074890416987e-06,
+      "loss": 0.6301,
+      "step": 1671
+    },
+    {
+      "epoch": 0.44391344749767686,
+      "grad_norm": 0.374164587013809,
+      "learning_rate": 4.9399988837429085e-06,
+      "loss": 0.6339,
+      "step": 1672
+    },
+    {
+      "epoch": 0.44417894597106067,
+      "grad_norm": 0.3801139949754767,
+      "learning_rate": 4.939922829482878e-06,
+      "loss": 0.6178,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.38613321793831296,
+      "learning_rate": 4.9398467276383786e-06,
+      "loss": 0.6465,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4447099429178282,
+      "grad_norm": 0.3864302965127435,
+      "learning_rate": 4.939770578210894e-06,
+      "loss": 0.6505,
+      "step": 1675
+    },
+    {
+      "epoch": 0.444975441391212,
+      "grad_norm": 0.3671960895719421,
+      "learning_rate": 4.939694381201909e-06,
+      "loss": 0.5898,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4452409398645958,
+      "grad_norm": 0.3785366184114263,
+      "learning_rate": 4.939618136612911e-06,
+      "loss": 0.6634,
+      "step": 1677
+    },
+    {
+      "epoch": 0.44550643833797954,
+      "grad_norm": 0.37229760246352267,
+      "learning_rate": 4.9395418444453865e-06,
+      "loss": 0.6529,
+      "step": 1678
+    },
+    {
+      "epoch": 0.44577193681136335,
+      "grad_norm": 0.3808553843853277,
+      "learning_rate": 4.9394655047008225e-06,
+      "loss": 0.617,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4460374352847471,
+      "grad_norm": 0.36503412428572335,
+      "learning_rate": 4.939389117380708e-06,
+      "loss": 0.5868,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4463029337581309,
+      "grad_norm": 0.3797407090372478,
+      "learning_rate": 4.9393126824865335e-06,
+      "loss": 0.6363,
+      "step": 1681
+    },
+    {
+      "epoch": 0.44656843223151466,
+      "grad_norm": 0.38620167463900623,
+      "learning_rate": 4.939236200019789e-06,
+      "loss": 0.638,
+      "step": 1682
+    },
+    {
+      "epoch": 0.44683393070489846,
+      "grad_norm": 0.3765492387858599,
+      "learning_rate": 4.939159669981966e-06,
+      "loss": 0.6152,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4470994291782822,
+      "grad_norm": 0.37395862179561223,
+      "learning_rate": 4.939083092374558e-06,
+      "loss": 0.6482,
+      "step": 1684
+    },
+    {
+      "epoch": 0.447364927651666,
+      "grad_norm": 0.38391228111425557,
+      "learning_rate": 4.939006467199057e-06,
+      "loss": 0.6012,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4476304261250498,
+      "grad_norm": 0.3897952675436452,
+      "learning_rate": 4.93892979445696e-06,
+      "loss": 0.6611,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4478959245984336,
+      "grad_norm": 0.36778256560587486,
+      "learning_rate": 4.938853074149757e-06,
+      "loss": 0.6073,
+      "step": 1687
+    },
+    {
+      "epoch": 0.44816142307181733,
+      "grad_norm": 0.38277630659821105,
+      "learning_rate": 4.93877630627895e-06,
+      "loss": 0.6338,
+      "step": 1688
+    },
+    {
+      "epoch": 0.44842692154520114,
+      "grad_norm": 0.38469641726248494,
+      "learning_rate": 4.938699490846035e-06,
+      "loss": 0.6429,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4486924200185849,
+      "grad_norm": 0.3700129046701821,
+      "learning_rate": 4.938622627852507e-06,
+      "loss": 0.6135,
+      "step": 1690
+    },
+    {
+      "epoch": 0.44895791849196864,
+      "grad_norm": 0.383356261722202,
+      "learning_rate": 4.938545717299867e-06,
+      "loss": 0.6788,
+      "step": 1691
+    },
+    {
+      "epoch": 0.44922341696535245,
+      "grad_norm": 0.3749427085145032,
+      "learning_rate": 4.938468759189616e-06,
+      "loss": 0.6629,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4494889154387362,
+      "grad_norm": 0.3726686152622098,
+      "learning_rate": 4.938391753523254e-06,
+      "loss": 0.6326,
+      "step": 1693
+    },
+    {
+      "epoch": 0.44975441391212,
+      "grad_norm": 0.3948298784210491,
+      "learning_rate": 4.938314700302282e-06,
+      "loss": 0.6273,
+      "step": 1694
+    },
+    {
+      "epoch": 0.45001991238550376,
+      "grad_norm": 0.3863984004680951,
+      "learning_rate": 4.938237599528203e-06,
+      "loss": 0.6069,
+      "step": 1695
+    },
+    {
+      "epoch": 0.45028541085888757,
+      "grad_norm": 0.3866378844432379,
+      "learning_rate": 4.938160451202521e-06,
+      "loss": 0.6497,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4505509093322713,
+      "grad_norm": 0.41075334584857653,
+      "learning_rate": 4.938083255326741e-06,
+      "loss": 0.5576,
+      "step": 1697
+    },
+    {
+      "epoch": 0.45081640780565513,
+      "grad_norm": 0.3920227584745048,
+      "learning_rate": 4.938006011902368e-06,
+      "loss": 0.6544,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4510819062790389,
+      "grad_norm": 0.3595136813858587,
+      "learning_rate": 4.937928720930908e-06,
+      "loss": 0.6244,
+      "step": 1699
+    },
+    {
+      "epoch": 0.4513474047524227,
+      "grad_norm": 0.3950605631766553,
+      "learning_rate": 4.937851382413869e-06,
+      "loss": 0.6171,
+      "step": 1700
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 0.40046136331571747,
+      "learning_rate": 4.937773996352759e-06,
+      "loss": 0.6234,
+      "step": 1701
+    },
+    {
+      "epoch": 0.45187840169919025,
+      "grad_norm": 0.3803613330253262,
+      "learning_rate": 4.937696562749088e-06,
+      "loss": 0.6403,
+      "step": 1702
+    },
+    {
+      "epoch": 0.452143900172574,
+      "grad_norm": 0.37269232995298995,
+      "learning_rate": 4.937619081604366e-06,
+      "loss": 0.6076,
+      "step": 1703
+    },
+    {
+      "epoch": 0.4524093986459578,
+      "grad_norm": 0.37804299250572265,
+      "learning_rate": 4.937541552920102e-06,
+      "loss": 0.63,
+      "step": 1704
+    },
+    {
+      "epoch": 0.45267489711934156,
+      "grad_norm": 0.4115677556718682,
+      "learning_rate": 4.937463976697811e-06,
+      "loss": 0.5989,
+      "step": 1705
+    },
+    {
+      "epoch": 0.45294039559272536,
+      "grad_norm": 0.3771962252099528,
+      "learning_rate": 4.937386352939004e-06,
+      "loss": 0.6346,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4532058940661091,
+      "grad_norm": 0.433898767443415,
+      "learning_rate": 4.937308681645196e-06,
+      "loss": 0.5951,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4534713925394929,
+      "grad_norm": 0.378806177547712,
+      "learning_rate": 4.9372309628178995e-06,
+      "loss": 0.6457,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4537368910128767,
+      "grad_norm": 0.38432939642619474,
+      "learning_rate": 4.937153196458633e-06,
+      "loss": 0.6478,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4540023894862604,
+      "grad_norm": 0.3952040727996697,
+      "learning_rate": 4.937075382568912e-06,
+      "loss": 0.5839,
+      "step": 1710
+    },
+    {
+      "epoch": 0.45426788795964423,
+      "grad_norm": 0.3757951589192852,
+      "learning_rate": 4.936997521150254e-06,
+      "loss": 0.6377,
+      "step": 1711
+    },
+    {
+      "epoch": 0.454533386433028,
+      "grad_norm": 0.3735648170534193,
+      "learning_rate": 4.936919612204177e-06,
+      "loss": 0.5973,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4547988849064118,
+      "grad_norm": 0.3884318126431452,
+      "learning_rate": 4.9368416557322014e-06,
+      "loss": 0.6507,
+      "step": 1713
+    },
+    {
+      "epoch": 0.45506438337979555,
+      "grad_norm": 0.3940904861093302,
+      "learning_rate": 4.936763651735847e-06,
+      "loss": 0.6573,
+      "step": 1714
+    },
+    {
+      "epoch": 0.45532988185317935,
+      "grad_norm": 0.3795500691170738,
+      "learning_rate": 4.936685600216635e-06,
+      "loss": 0.6689,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4555953803265631,
+      "grad_norm": 0.38041657880412216,
+      "learning_rate": 4.936607501176088e-06,
+      "loss": 0.6452,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4558608787999469,
+      "grad_norm": 0.379827072104167,
+      "learning_rate": 4.936529354615729e-06,
+      "loss": 0.6325,
+      "step": 1717
+    },
+    {
+      "epoch": 0.45612637727333066,
+      "grad_norm": 0.37834522292738343,
+      "learning_rate": 4.9364511605370815e-06,
+      "loss": 0.5989,
+      "step": 1718
+    },
+    {
+      "epoch": 0.45639187574671447,
+      "grad_norm": 0.3896691716148016,
+      "learning_rate": 4.936372918941671e-06,
+      "loss": 0.622,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4566573742200982,
+      "grad_norm": 0.3750019835504782,
+      "learning_rate": 4.9362946298310236e-06,
+      "loss": 0.6508,
+      "step": 1720
+    },
+    {
+      "epoch": 0.45692287269348203,
+      "grad_norm": 0.378092227330637,
+      "learning_rate": 4.936216293206666e-06,
+      "loss": 0.6175,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4571883711668658,
+      "grad_norm": 0.372360338643961,
+      "learning_rate": 4.936137909070125e-06,
+      "loss": 0.6479,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4574538696402496,
+      "grad_norm": 0.38238014554140415,
+      "learning_rate": 4.936059477422931e-06,
+      "loss": 0.6047,
+      "step": 1723
+    },
+    {
+      "epoch": 0.45771936811363334,
+      "grad_norm": 0.3777027803451668,
+      "learning_rate": 4.935980998266613e-06,
+      "loss": 0.6004,
+      "step": 1724
+    },
+    {
+      "epoch": 0.45798486658701715,
+      "grad_norm": 0.3951776977852749,
+      "learning_rate": 4.9359024716027e-06,
+      "loss": 0.6384,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4582503650604009,
+      "grad_norm": 0.3845661945307065,
+      "learning_rate": 4.935823897432725e-06,
+      "loss": 0.636,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4585158635337847,
+      "grad_norm": 0.40118180045020735,
+      "learning_rate": 4.93574527575822e-06,
+      "loss": 0.6176,
+      "step": 1727
+    },
+    {
+      "epoch": 0.45878136200716846,
+      "grad_norm": 0.391039266777863,
+      "learning_rate": 4.935666606580719e-06,
+      "loss": 0.6185,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4590468604805522,
+      "grad_norm": 0.37819717057976776,
+      "learning_rate": 4.935587889901756e-06,
+      "loss": 0.6573,
+      "step": 1729
+    },
+    {
+      "epoch": 0.459312358953936,
+      "grad_norm": 0.41544670685909424,
+      "learning_rate": 4.935509125722865e-06,
+      "loss": 0.5862,
+      "step": 1730
+    },
+    {
+      "epoch": 0.45957785742731977,
+      "grad_norm": 0.4080586799532462,
+      "learning_rate": 4.935430314045583e-06,
+      "loss": 0.6383,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4598433559007036,
+      "grad_norm": 0.40190361658617757,
+      "learning_rate": 4.935351454871447e-06,
+      "loss": 0.6706,
+      "step": 1732
+    },
+    {
+      "epoch": 0.46010885437408733,
+      "grad_norm": 0.41063212068772775,
+      "learning_rate": 4.9352725482019944e-06,
+      "loss": 0.6384,
+      "step": 1733
+    },
+    {
+      "epoch": 0.46037435284747114,
+      "grad_norm": 0.3950823442438883,
+      "learning_rate": 4.935193594038765e-06,
+      "loss": 0.6355,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4606398513208549,
+      "grad_norm": 0.38032576246327926,
+      "learning_rate": 4.935114592383298e-06,
+      "loss": 0.6205,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4609053497942387,
+      "grad_norm": 0.3862929803763625,
+      "learning_rate": 4.9350355432371345e-06,
+      "loss": 0.6521,
+      "step": 1736
+    },
+    {
+      "epoch": 0.46117084826762245,
+      "grad_norm": 0.3897244594003261,
+      "learning_rate": 4.934956446601815e-06,
+      "loss": 0.5909,
+      "step": 1737
+    },
+    {
+      "epoch": 0.46143634674100625,
+      "grad_norm": 0.37387682523316074,
+      "learning_rate": 4.9348773024788845e-06,
+      "loss": 0.576,
+      "step": 1738
+    },
+    {
+      "epoch": 0.46170184521439,
+      "grad_norm": 0.3837862860957116,
+      "learning_rate": 4.934798110869884e-06,
+      "loss": 0.643,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4619673436877738,
+      "grad_norm": 0.3998133847095514,
+      "learning_rate": 4.934718871776359e-06,
+      "loss": 0.6454,
+      "step": 1740
+    },
+    {
+      "epoch": 0.46223284216115756,
+      "grad_norm": 0.37805680537467495,
+      "learning_rate": 4.934639585199855e-06,
+      "loss": 0.6441,
+      "step": 1741
+    },
+    {
+      "epoch": 0.46249834063454137,
+      "grad_norm": 0.38788744008372245,
+      "learning_rate": 4.934560251141918e-06,
+      "loss": 0.6234,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4627638391079251,
+      "grad_norm": 0.3654820128679984,
+      "learning_rate": 4.934480869604095e-06,
+      "loss": 0.5979,
+      "step": 1743
+    },
+    {
+      "epoch": 0.46302933758130893,
+      "grad_norm": 0.39218850405121114,
+      "learning_rate": 4.934401440587935e-06,
+      "loss": 0.6566,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4632948360546927,
+      "grad_norm": 0.37947823451585755,
+      "learning_rate": 4.934321964094985e-06,
+      "loss": 0.6392,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4635603345280765,
+      "grad_norm": 0.38657237651505744,
+      "learning_rate": 4.934242440126798e-06,
+      "loss": 0.6319,
+      "step": 1746
+    },
+    {
+      "epoch": 0.46382583300146024,
+      "grad_norm": 0.37588575359767307,
+      "learning_rate": 4.934162868684923e-06,
+      "loss": 0.6073,
+      "step": 1747
+    },
+    {
+      "epoch": 0.464091331474844,
+      "grad_norm": 0.3982070933536065,
+      "learning_rate": 4.934083249770912e-06,
+      "loss": 0.6504,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4643568299482278,
+      "grad_norm": 0.40661072542218,
+      "learning_rate": 4.934003583386319e-06,
+      "loss": 0.6043,
+      "step": 1749
+    },
+    {
+      "epoch": 0.46462232842161155,
+      "grad_norm": 0.3819824929010193,
+      "learning_rate": 4.933923869532695e-06,
+      "loss": 0.6274,
+      "step": 1750
+    },
+    {
+      "epoch": 0.46488782689499536,
+      "grad_norm": 0.42403335588596847,
+      "learning_rate": 4.933844108211598e-06,
+      "loss": 0.6432,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4651533253683791,
+      "grad_norm": 0.3818754144442884,
+      "learning_rate": 4.933764299424581e-06,
+      "loss": 0.6221,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4654188238417629,
+      "grad_norm": 0.3766156996849648,
+      "learning_rate": 4.9336844431732e-06,
+      "loss": 0.6292,
+      "step": 1753
+    },
+    {
+      "epoch": 0.46568432231514667,
+      "grad_norm": 0.4119487407929856,
+      "learning_rate": 4.933604539459015e-06,
+      "loss": 0.6177,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4659498207885305,
+      "grad_norm": 0.3793213603442324,
+      "learning_rate": 4.933524588283583e-06,
+      "loss": 0.6511,
+      "step": 1755
+    },
+    {
+      "epoch": 0.46621531926191423,
+      "grad_norm": 0.37576310847605426,
+      "learning_rate": 4.933444589648463e-06,
+      "loss": 0.6446,
+      "step": 1756
+    },
+    {
+      "epoch": 0.46648081773529804,
+      "grad_norm": 0.38357642146509346,
+      "learning_rate": 4.9333645435552156e-06,
+      "loss": 0.6271,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4667463162086818,
+      "grad_norm": 0.43516532010809733,
+      "learning_rate": 4.9332844500054015e-06,
+      "loss": 0.6514,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4670118146820656,
+      "grad_norm": 0.37990546112442525,
+      "learning_rate": 4.933204309000583e-06,
+      "loss": 0.6489,
+      "step": 1759
+    },
+    {
+      "epoch": 0.46727731315544935,
+      "grad_norm": 0.38457986689359225,
+      "learning_rate": 4.933124120542323e-06,
+      "loss": 0.6121,
+      "step": 1760
+    },
+    {
+      "epoch": 0.46754281162883315,
+      "grad_norm": 0.3704122932587119,
+      "learning_rate": 4.933043884632185e-06,
+      "loss": 0.5994,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4678083101022169,
+      "grad_norm": 0.3899208147014975,
+      "learning_rate": 4.932963601271734e-06,
+      "loss": 0.6903,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4680738085756007,
+      "grad_norm": 0.40387845271885775,
+      "learning_rate": 4.932883270462536e-06,
+      "loss": 0.6355,
+      "step": 1763
+    },
+    {
+      "epoch": 0.46833930704898447,
+      "grad_norm": 0.3899688593253954,
+      "learning_rate": 4.932802892206158e-06,
+      "loss": 0.6193,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4686048055223683,
+      "grad_norm": 0.4211207461379197,
+      "learning_rate": 4.932722466504167e-06,
+      "loss": 0.5933,
+      "step": 1765
+    },
+    {
+      "epoch": 0.468870303995752,
+      "grad_norm": 0.4004879803929037,
+      "learning_rate": 4.932641993358132e-06,
+      "loss": 0.6347,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4691358024691358,
+      "grad_norm": 0.39776362486682454,
+      "learning_rate": 4.93256147276962e-06,
+      "loss": 0.6449,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4694013009425196,
+      "grad_norm": 0.38379232001516417,
+      "learning_rate": 4.932480904740206e-06,
+      "loss": 0.6231,
+      "step": 1768
+    },
+    {
+      "epoch": 0.46966679941590334,
+      "grad_norm": 0.41869550785919246,
+      "learning_rate": 4.932400289271456e-06,
+      "loss": 0.6213,
+      "step": 1769
+    },
+    {
+      "epoch": 0.46993229788928714,
+      "grad_norm": 0.3780403678137773,
+      "learning_rate": 4.932319626364947e-06,
+      "loss": 0.6081,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4701977963626709,
+      "grad_norm": 0.3849828428519195,
+      "learning_rate": 4.932238916022249e-06,
+      "loss": 0.5926,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4704632948360547,
+      "grad_norm": 0.3954503159594688,
+      "learning_rate": 4.9321581582449365e-06,
+      "loss": 0.6444,
+      "step": 1772
+    },
+    {
+      "epoch": 0.47072879330943845,
+      "grad_norm": 0.38946945581332887,
+      "learning_rate": 4.932077353034585e-06,
+      "loss": 0.6281,
+      "step": 1773
+    },
+    {
+      "epoch": 0.47099429178282226,
+      "grad_norm": 0.38524665261405866,
+      "learning_rate": 4.931996500392772e-06,
+      "loss": 0.6241,
+      "step": 1774
+    },
+    {
+      "epoch": 0.471259790256206,
+      "grad_norm": 0.3856937094015711,
+      "learning_rate": 4.931915600321071e-06,
+      "loss": 0.5861,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4715252887295898,
+      "grad_norm": 0.382896892369575,
+      "learning_rate": 4.931834652821063e-06,
+      "loss": 0.6539,
+      "step": 1776
+    },
+    {
+      "epoch": 0.47179078720297357,
+      "grad_norm": 0.42227343563372355,
+      "learning_rate": 4.9317536578943235e-06,
+      "loss": 0.6164,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4720562856763574,
+      "grad_norm": 0.3792211153051177,
+      "learning_rate": 4.931672615542434e-06,
+      "loss": 0.6097,
+      "step": 1778
+    },
+    {
+      "epoch": 0.47232178414974113,
+      "grad_norm": 0.40435135045283704,
+      "learning_rate": 4.931591525766976e-06,
+      "loss": 0.6529,
+      "step": 1779
+    },
+    {
+      "epoch": 0.47258728262312494,
+      "grad_norm": 0.3840802542840774,
+      "learning_rate": 4.931510388569528e-06,
+      "loss": 0.6399,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4728527810965087,
+      "grad_norm": 0.3986134311166046,
+      "learning_rate": 4.9314292039516755e-06,
+      "loss": 0.6221,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4731182795698925,
+      "grad_norm": 0.3897418395704273,
+      "learning_rate": 4.931347971914999e-06,
+      "loss": 0.6734,
+      "step": 1782
+    },
+    {
+      "epoch": 0.47338377804327625,
+      "grad_norm": 0.40374280754299857,
+      "learning_rate": 4.931266692461085e-06,
+      "loss": 0.649,
+      "step": 1783
+    },
+    {
+      "epoch": 0.47364927651666006,
+      "grad_norm": 0.4150821891451634,
+      "learning_rate": 4.9311853655915175e-06,
+      "loss": 0.6364,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4739147749900438,
+      "grad_norm": 0.3730492094868012,
+      "learning_rate": 4.931103991307883e-06,
+      "loss": 0.64,
+      "step": 1785
+    },
+    {
+      "epoch": 0.47418027346342756,
+      "grad_norm": 0.3793974882897054,
+      "learning_rate": 4.931022569611768e-06,
+      "loss": 0.6702,
+      "step": 1786
+    },
+    {
+      "epoch": 0.47444577193681137,
+      "grad_norm": 0.3992342594127169,
+      "learning_rate": 4.930941100504761e-06,
+      "loss": 0.6377,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4747112704101951,
+      "grad_norm": 0.4613974070846031,
+      "learning_rate": 4.93085958398845e-06,
+      "loss": 0.6258,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4749767688835789,
+      "grad_norm": 0.3869638506793003,
+      "learning_rate": 4.930778020064426e-06,
+      "loss": 0.6334,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4752422673569627,
+      "grad_norm": 0.43960079157031906,
+      "learning_rate": 4.930696408734278e-06,
+      "loss": 0.6033,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4755077658303465,
+      "grad_norm": 0.39071061415164876,
+      "learning_rate": 4.9306147499995996e-06,
+      "loss": 0.6632,
+      "step": 1791
+    },
+    {
+      "epoch": 0.47577326430373024,
+      "grad_norm": 0.39240737775505524,
+      "learning_rate": 4.930533043861982e-06,
+      "loss": 0.6597,
+      "step": 1792
+    },
+    {
+      "epoch": 0.47603876277711404,
+      "grad_norm": 0.3850044943362015,
+      "learning_rate": 4.930451290323019e-06,
+      "loss": 0.6189,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4763042612504978,
+      "grad_norm": 0.4371740593999468,
+      "learning_rate": 4.9303694893843065e-06,
+      "loss": 0.639,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4765697597238816,
+      "grad_norm": 0.40516604641583126,
+      "learning_rate": 4.930287641047437e-06,
+      "loss": 0.6383,
+      "step": 1795
+    },
+    {
+      "epoch": 0.47683525819726535,
+      "grad_norm": 0.38546254096400157,
+      "learning_rate": 4.9302057453140085e-06,
+      "loss": 0.6293,
+      "step": 1796
+    },
+    {
+      "epoch": 0.47710075667064916,
+      "grad_norm": 0.4034326852028295,
+      "learning_rate": 4.930123802185618e-06,
+      "loss": 0.6505,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4773662551440329,
+      "grad_norm": 0.40494528447050565,
+      "learning_rate": 4.9300418116638625e-06,
+      "loss": 0.6287,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4776317536174167,
+      "grad_norm": 0.3821707943680172,
+      "learning_rate": 4.929959773750342e-06,
+      "loss": 0.6446,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4778972520908005,
+      "grad_norm": 0.39275736714185777,
+      "learning_rate": 4.929877688446657e-06,
+      "loss": 0.6369,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4781627505641843,
+      "grad_norm": 0.42733990423350493,
+      "learning_rate": 4.929795555754407e-06,
+      "loss": 0.6399,
+      "step": 1801
+    },
+    {
+      "epoch": 0.47842824903756803,
+      "grad_norm": 0.3981819973597323,
+      "learning_rate": 4.929713375675196e-06,
+      "loss": 0.6239,
+      "step": 1802
+    },
+    {
+      "epoch": 0.47869374751095184,
+      "grad_norm": 0.36912415854551983,
+      "learning_rate": 4.929631148210624e-06,
+      "loss": 0.6336,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4789592459843356,
+      "grad_norm": 0.42297729082895735,
+      "learning_rate": 4.929548873362297e-06,
+      "loss": 0.6176,
+      "step": 1804
+    },
+    {
+      "epoch": 0.47922474445771934,
+      "grad_norm": 0.4065210066685864,
+      "learning_rate": 4.929466551131816e-06,
+      "loss": 0.6289,
+      "step": 1805
+    },
+    {
+      "epoch": 0.47949024293110315,
+      "grad_norm": 0.38213165971721497,
+      "learning_rate": 4.929384181520791e-06,
+      "loss": 0.6381,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4797557414044869,
+      "grad_norm": 0.3827911908875227,
+      "learning_rate": 4.929301764530826e-06,
+      "loss": 0.6069,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4800212398778707,
+      "grad_norm": 0.4111573647662111,
+      "learning_rate": 4.929219300163528e-06,
+      "loss": 0.621,
+      "step": 1808
+    },
+    {
+      "epoch": 0.48028673835125446,
+      "grad_norm": 0.37363367624712823,
+      "learning_rate": 4.929136788420507e-06,
+      "loss": 0.6097,
+      "step": 1809
+    },
+    {
+      "epoch": 0.48055223682463827,
+      "grad_norm": 0.38187753033938593,
+      "learning_rate": 4.92905422930337e-06,
+      "loss": 0.6182,
+      "step": 1810
+    },
+    {
+      "epoch": 0.480817735298022,
+      "grad_norm": 0.3928287873363872,
+      "learning_rate": 4.928971622813728e-06,
+      "loss": 0.6448,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4810832337714058,
+      "grad_norm": 0.4317048263471141,
+      "learning_rate": 4.928888968953193e-06,
+      "loss": 0.623,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4813487322447896,
+      "grad_norm": 0.3839940387112986,
+      "learning_rate": 4.9288062677233756e-06,
+      "loss": 0.6157,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4816142307181734,
+      "grad_norm": 0.38336127182567864,
+      "learning_rate": 4.92872351912589e-06,
+      "loss": 0.6468,
+      "step": 1814
+    },
+    {
+      "epoch": 0.48187972919155714,
+      "grad_norm": 0.43882607404615626,
+      "learning_rate": 4.9286407231623476e-06,
+      "loss": 0.6676,
+      "step": 1815
+    },
+    {
+      "epoch": 0.48214522766494095,
+      "grad_norm": 0.43762663715889866,
+      "learning_rate": 4.928557879834366e-06,
+      "loss": 0.6156,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4824107261383247,
+      "grad_norm": 0.3991874039101594,
+      "learning_rate": 4.92847498914356e-06,
+      "loss": 0.6445,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4826762246117085,
+      "grad_norm": 0.43988804884879057,
+      "learning_rate": 4.928392051091545e-06,
+      "loss": 0.5833,
+      "step": 1818
+    },
+    {
+      "epoch": 0.48294172308509226,
+      "grad_norm": 0.40037173837794393,
+      "learning_rate": 4.92830906567994e-06,
+      "loss": 0.6617,
+      "step": 1819
+    },
+    {
+      "epoch": 0.48320722155847606,
+      "grad_norm": 0.5218895118820784,
+      "learning_rate": 4.9282260329103614e-06,
+      "loss": 0.6237,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4834727200318598,
+      "grad_norm": 0.39367543447004505,
+      "learning_rate": 4.928142952784431e-06,
+      "loss": 0.6131,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4837382185052436,
+      "grad_norm": 0.4803276478832675,
+      "learning_rate": 4.928059825303768e-06,
+      "loss": 0.5896,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4840037169786274,
+      "grad_norm": 0.40419417536708496,
+      "learning_rate": 4.927976650469993e-06,
+      "loss": 0.6523,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4842692154520111,
+      "grad_norm": 0.3851814826644902,
+      "learning_rate": 4.927893428284728e-06,
+      "loss": 0.6691,
+      "step": 1824
+    },
+    {
+      "epoch": 0.48453471392539493,
+      "grad_norm": 0.4364195716483101,
+      "learning_rate": 4.927810158749597e-06,
+      "loss": 0.5915,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4848002123987787,
+      "grad_norm": 0.4587611935637998,
+      "learning_rate": 4.927726841866225e-06,
+      "loss": 0.6119,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4850657108721625,
+      "grad_norm": 0.38441463311340074,
+      "learning_rate": 4.927643477636234e-06,
+      "loss": 0.6193,
+      "step": 1827
+    },
+    {
+      "epoch": 0.48533120934554624,
+      "grad_norm": 0.3699694153904709,
+      "learning_rate": 4.927560066061251e-06,
+      "loss": 0.6252,
+      "step": 1828
+    },
+    {
+      "epoch": 0.48559670781893005,
+      "grad_norm": 0.446749391341119,
+      "learning_rate": 4.927476607142904e-06,
+      "loss": 0.6393,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4858622062923138,
+      "grad_norm": 0.4180574613486697,
+      "learning_rate": 4.92739310088282e-06,
+      "loss": 0.599,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4861277047656976,
+      "grad_norm": 0.375267555779444,
+      "learning_rate": 4.927309547282626e-06,
+      "loss": 0.6389,
+      "step": 1831
+    },
+    {
+      "epoch": 0.48639320323908136,
+      "grad_norm": 0.400155794848709,
+      "learning_rate": 4.927225946343954e-06,
+      "loss": 0.6436,
+      "step": 1832
+    },
+    {
+      "epoch": 0.48665870171246517,
+      "grad_norm": 0.4662098221918315,
+      "learning_rate": 4.927142298068432e-06,
+      "loss": 0.607,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4869242001858489,
+      "grad_norm": 0.3877638533614041,
+      "learning_rate": 4.927058602457694e-06,
+      "loss": 0.5983,
+      "step": 1834
+    },
+    {
+      "epoch": 0.48718969865923273,
+      "grad_norm": 0.39973654597842606,
+      "learning_rate": 4.92697485951337e-06,
+      "loss": 0.6361,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4874551971326165,
+      "grad_norm": 0.4067946433566237,
+      "learning_rate": 4.926891069237093e-06,
+      "loss": 0.6393,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4877206956060003,
+      "grad_norm": 0.42188160943189607,
+      "learning_rate": 4.926807231630499e-06,
+      "loss": 0.612,
+      "step": 1837
+    },
+    {
+      "epoch": 0.48798619407938404,
+      "grad_norm": 0.4354009373050952,
+      "learning_rate": 4.926723346695222e-06,
+      "loss": 0.6072,
+      "step": 1838
+    },
+    {
+      "epoch": 0.48825169255276785,
+      "grad_norm": 0.4223277098623755,
+      "learning_rate": 4.926639414432898e-06,
+      "loss": 0.6073,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4885171910261516,
+      "grad_norm": 0.4497404463001629,
+      "learning_rate": 4.926555434845164e-06,
+      "loss": 0.6079,
+      "step": 1840
+    },
+    {
+      "epoch": 0.48878268949953535,
+      "grad_norm": 0.4038820148101401,
+      "learning_rate": 4.926471407933658e-06,
+      "loss": 0.6057,
+      "step": 1841
+    },
+    {
+      "epoch": 0.48904818797291916,
+      "grad_norm": 0.38044712431509375,
+      "learning_rate": 4.926387333700018e-06,
+      "loss": 0.6292,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4893136864463029,
+      "grad_norm": 0.4162725292710651,
+      "learning_rate": 4.926303212145885e-06,
+      "loss": 0.6521,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4895791849196867,
+      "grad_norm": 0.38821500859141916,
+      "learning_rate": 4.926219043272899e-06,
+      "loss": 0.605,
+      "step": 1844
+    },
+    {
+      "epoch": 0.48984468339307047,
+      "grad_norm": 0.3952703625765734,
+      "learning_rate": 4.9261348270827004e-06,
+      "loss": 0.6602,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4901101818664543,
+      "grad_norm": 0.37552225562544594,
+      "learning_rate": 4.926050563576932e-06,
+      "loss": 0.6489,
+      "step": 1846
+    },
+    {
+      "epoch": 0.490375680339838,
+      "grad_norm": 0.388530643313539,
+      "learning_rate": 4.925966252757238e-06,
+      "loss": 0.6347,
+      "step": 1847
+    },
+    {
+      "epoch": 0.49064117881322183,
+      "grad_norm": 0.4114483732608687,
+      "learning_rate": 4.925881894625263e-06,
+      "loss": 0.6333,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4909066772866056,
+      "grad_norm": 0.3793648338354744,
+      "learning_rate": 4.925797489182651e-06,
+      "loss": 0.6559,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4911721757599894,
+      "grad_norm": 0.37267369016981056,
+      "learning_rate": 4.925713036431049e-06,
+      "loss": 0.6239,
+      "step": 1850
+    },
+    {
+      "epoch": 0.49143767423337315,
+      "grad_norm": 0.3938862689139422,
+      "learning_rate": 4.9256285363721025e-06,
+      "loss": 0.6173,
+      "step": 1851
+    },
+    {
+      "epoch": 0.49170317270675695,
+      "grad_norm": 0.3859814816879519,
+      "learning_rate": 4.925543989007462e-06,
+      "loss": 0.6213,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4919686711801407,
+      "grad_norm": 0.39389672074854576,
+      "learning_rate": 4.925459394338774e-06,
+      "loss": 0.6036,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4922341696535245,
+      "grad_norm": 0.37010200453351183,
+      "learning_rate": 4.92537475236769e-06,
+      "loss": 0.5774,
+      "step": 1854
+    },
+    {
+      "epoch": 0.49249966812690826,
+      "grad_norm": 0.38274026855741344,
+      "learning_rate": 4.925290063095859e-06,
+      "loss": 0.592,
+      "step": 1855
+    },
+    {
+      "epoch": 0.49276516660029207,
+      "grad_norm": 0.3965939670089687,
+      "learning_rate": 4.925205326524934e-06,
+      "loss": 0.5765,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4930306650736758,
+      "grad_norm": 0.3815564350233024,
+      "learning_rate": 4.925120542656567e-06,
+      "loss": 0.6389,
+      "step": 1857
+    },
+    {
+      "epoch": 0.49329616354705963,
+      "grad_norm": 0.3938485307172248,
+      "learning_rate": 4.925035711492413e-06,
+      "loss": 0.6484,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4935616620204434,
+      "grad_norm": 0.3957464540329845,
+      "learning_rate": 4.924950833034124e-06,
+      "loss": 0.6379,
+      "step": 1859
+    },
+    {
+      "epoch": 0.49382716049382713,
+      "grad_norm": 0.41618181645568386,
+      "learning_rate": 4.924865907283356e-06,
+      "loss": 0.5952,
+      "step": 1860
+    },
+    {
+      "epoch": 0.49409265896721094,
+      "grad_norm": 0.38897976703843057,
+      "learning_rate": 4.924780934241766e-06,
+      "loss": 0.6365,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4943581574405947,
+      "grad_norm": 0.3909324650568712,
+      "learning_rate": 4.924695913911011e-06,
+      "loss": 0.6289,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4946236559139785,
+      "grad_norm": 0.394899926105948,
+      "learning_rate": 4.924610846292751e-06,
+      "loss": 0.6087,
+      "step": 1863
+    },
+    {
+      "epoch": 0.49488915438736225,
+      "grad_norm": 0.41135468729534785,
+      "learning_rate": 4.924525731388641e-06,
+      "loss": 0.5911,
+      "step": 1864
+    },
+    {
+      "epoch": 0.49515465286074606,
+      "grad_norm": 0.39475059271360924,
+      "learning_rate": 4.924440569200343e-06,
+      "loss": 0.6319,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4954201513341298,
+      "grad_norm": 0.44822777827399035,
+      "learning_rate": 4.924355359729518e-06,
+      "loss": 0.5802,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4956856498075136,
+      "grad_norm": 0.37686364393220034,
+      "learning_rate": 4.924270102977828e-06,
+      "loss": 0.6389,
+      "step": 1867
+    },
+    {
+      "epoch": 0.49595114828089737,
+      "grad_norm": 0.37756133814720716,
+      "learning_rate": 4.924184798946935e-06,
+      "loss": 0.6604,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4962166467542812,
+      "grad_norm": 0.3806651847473416,
+      "learning_rate": 4.924099447638503e-06,
+      "loss": 0.6396,
+      "step": 1869
+    },
+    {
+      "epoch": 0.49648214522766493,
+      "grad_norm": 0.40260430574696265,
+      "learning_rate": 4.924014049054197e-06,
+      "loss": 0.6566,
+      "step": 1870
+    },
+    {
+      "epoch": 0.49674764370104874,
+      "grad_norm": 0.3993604109813291,
+      "learning_rate": 4.923928603195682e-06,
+      "loss": 0.5899,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4970131421744325,
+      "grad_norm": 0.3932748029941721,
+      "learning_rate": 4.923843110064624e-06,
+      "loss": 0.6216,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4972786406478163,
+      "grad_norm": 0.3870168233713288,
+      "learning_rate": 4.9237575696626905e-06,
+      "loss": 0.631,
+      "step": 1873
+    },
+    {
+      "epoch": 0.49754413912120005,
+      "grad_norm": 0.3930017732892036,
+      "learning_rate": 4.92367198199155e-06,
+      "loss": 0.6544,
+      "step": 1874
+    },
+    {
+      "epoch": 0.49780963759458385,
+      "grad_norm": 0.37701820531803265,
+      "learning_rate": 4.923586347052872e-06,
+      "loss": 0.6274,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4980751360679676,
+      "grad_norm": 0.3823095560780843,
+      "learning_rate": 4.923500664848327e-06,
+      "loss": 0.6287,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4983406345413514,
+      "grad_norm": 0.3864494476918735,
+      "learning_rate": 4.923414935379584e-06,
+      "loss": 0.6253,
+      "step": 1877
+    },
+    {
+      "epoch": 0.49860613301473516,
+      "grad_norm": 0.3948185098943959,
+      "learning_rate": 4.9233291586483165e-06,
+      "loss": 0.6448,
+      "step": 1878
+    },
+    {
+      "epoch": 0.4988716314881189,
+      "grad_norm": 0.3773610927653805,
+      "learning_rate": 4.923243334656198e-06,
+      "loss": 0.5942,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4991371299615027,
+      "grad_norm": 0.3764972842969289,
+      "learning_rate": 4.9231574634049005e-06,
+      "loss": 0.6336,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4994026284348865,
+      "grad_norm": 0.39257912049596666,
+      "learning_rate": 4.9230715448960995e-06,
+      "loss": 0.6022,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4996681269082703,
+      "grad_norm": 0.37839307617243384,
+      "learning_rate": 4.922985579131471e-06,
+      "loss": 0.6447,
+      "step": 1882
+    },
+    {
+      "epoch": 0.49993362538165403,
+      "grad_norm": 0.3746841193048756,
+      "learning_rate": 4.922899566112691e-06,
+      "loss": 0.5758,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5001991238550378,
+      "grad_norm": 0.39589685535459257,
+      "learning_rate": 4.922813505841438e-06,
+      "loss": 0.6123,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5004646223284216,
+      "grad_norm": 0.3998132706267363,
+      "learning_rate": 4.922727398319389e-06,
+      "loss": 0.606,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5007301208018053,
+      "grad_norm": 0.37631711209643176,
+      "learning_rate": 4.922641243548223e-06,
+      "loss": 0.6253,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5009956192751892,
+      "grad_norm": 0.372388653748478,
+      "learning_rate": 4.9225550415296226e-06,
+      "loss": 0.6033,
+      "step": 1887
+    },
+    {
+      "epoch": 0.501261117748573,
+      "grad_norm": 0.3939052723338677,
+      "learning_rate": 4.922468792265267e-06,
+      "loss": 0.6563,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5015266162219567,
+      "grad_norm": 0.3772137272142795,
+      "learning_rate": 4.922382495756838e-06,
+      "loss": 0.635,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5017921146953405,
+      "grad_norm": 0.38930701121697076,
+      "learning_rate": 4.92229615200602e-06,
+      "loss": 0.6286,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5020576131687243,
+      "grad_norm": 0.3956069772498084,
+      "learning_rate": 4.9222097610144955e-06,
+      "loss": 0.6391,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5023231116421081,
+      "grad_norm": 0.3923718045233908,
+      "learning_rate": 4.922123322783951e-06,
+      "loss": 0.6124,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5025886101154918,
+      "grad_norm": 0.37514500669708245,
+      "learning_rate": 4.922036837316071e-06,
+      "loss": 0.619,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5028541085888756,
+      "grad_norm": 0.40192421736446754,
+      "learning_rate": 4.921950304612543e-06,
+      "loss": 0.6053,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5031196070622594,
+      "grad_norm": 0.379298999507021,
+      "learning_rate": 4.9218637246750535e-06,
+      "loss": 0.6295,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5033851055356432,
+      "grad_norm": 0.37048689247873723,
+      "learning_rate": 4.921777097505291e-06,
+      "loss": 0.6595,
+      "step": 1896
+    },
+    {
+      "epoch": 0.503650604009027,
+      "grad_norm": 0.3694561367715235,
+      "learning_rate": 4.921690423104947e-06,
+      "loss": 0.6218,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5039161024824107,
+      "grad_norm": 0.3730893334920156,
+      "learning_rate": 4.921603701475709e-06,
+      "loss": 0.6303,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5041816009557945,
+      "grad_norm": 0.3766107856638912,
+      "learning_rate": 4.92151693261927e-06,
+      "loss": 0.6239,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5044470994291783,
+      "grad_norm": 0.3908428042761929,
+      "learning_rate": 4.9214301165373225e-06,
+      "loss": 0.6026,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5047125979025621,
+      "grad_norm": 0.3899714867806938,
+      "learning_rate": 4.921343253231558e-06,
+      "loss": 0.6405,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5049780963759458,
+      "grad_norm": 0.3958765178921134,
+      "learning_rate": 4.9212563427036735e-06,
+      "loss": 0.6587,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5052435948493296,
+      "grad_norm": 0.3939974386685129,
+      "learning_rate": 4.921169384955361e-06,
+      "loss": 0.6699,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5055090933227134,
+      "grad_norm": 0.38011296113512727,
+      "learning_rate": 4.921082379988317e-06,
+      "loss": 0.65,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5057745917960972,
+      "grad_norm": 0.3797311677169585,
+      "learning_rate": 4.92099532780424e-06,
+      "loss": 0.6685,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5060400902694809,
+      "grad_norm": 0.37817001612046675,
+      "learning_rate": 4.9209082284048245e-06,
+      "loss": 0.6321,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5063055887428647,
+      "grad_norm": 0.3856365392957159,
+      "learning_rate": 4.9208210817917735e-06,
+      "loss": 0.6152,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5065710872162486,
+      "grad_norm": 0.3867586074534019,
+      "learning_rate": 4.920733887966783e-06,
+      "loss": 0.6274,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5068365856896323,
+      "grad_norm": 0.3766034581587804,
+      "learning_rate": 4.920646646931554e-06,
+      "loss": 0.6002,
+      "step": 1909
+    },
+    {
+      "epoch": 0.507102084163016,
+      "grad_norm": 0.3811016647787012,
+      "learning_rate": 4.92055935868779e-06,
+      "loss": 0.6141,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5073675826363998,
+      "grad_norm": 0.386717603058035,
+      "learning_rate": 4.920472023237191e-06,
+      "loss": 0.6163,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5076330811097837,
+      "grad_norm": 0.3927583755703415,
+      "learning_rate": 4.9203846405814616e-06,
+      "loss": 0.6331,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5078985795831674,
+      "grad_norm": 0.3815325638823032,
+      "learning_rate": 4.920297210722306e-06,
+      "loss": 0.6331,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5081640780565512,
+      "grad_norm": 0.38050871211737247,
+      "learning_rate": 4.920209733661427e-06,
+      "loss": 0.651,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5084295765299349,
+      "grad_norm": 0.38264302361312386,
+      "learning_rate": 4.920122209400535e-06,
+      "loss": 0.6577,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5086950750033188,
+      "grad_norm": 0.3858956475533456,
+      "learning_rate": 4.920034637941334e-06,
+      "loss": 0.6397,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5089605734767025,
+      "grad_norm": 0.39106068747768324,
+      "learning_rate": 4.91994701928553e-06,
+      "loss": 0.6389,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5092260719500863,
+      "grad_norm": 0.3962571386974248,
+      "learning_rate": 4.919859353434836e-06,
+      "loss": 0.6127,
+      "step": 1918
+    },
+    {
+      "epoch": 0.50949157042347,
+      "grad_norm": 0.3919897578088484,
+      "learning_rate": 4.91977164039096e-06,
+      "loss": 0.6568,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5097570688968538,
+      "grad_norm": 0.3885273390315128,
+      "learning_rate": 4.9196838801556115e-06,
+      "loss": 0.617,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5100225673702377,
+      "grad_norm": 0.3644433260512407,
+      "learning_rate": 4.919596072730504e-06,
+      "loss": 0.5982,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5102880658436214,
+      "grad_norm": 0.37282171537928127,
+      "learning_rate": 4.919508218117348e-06,
+      "loss": 0.6375,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5105535643170052,
+      "grad_norm": 0.3799837899155145,
+      "learning_rate": 4.919420316317858e-06,
+      "loss": 0.6373,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5108190627903889,
+      "grad_norm": 0.37946793023993824,
+      "learning_rate": 4.919332367333748e-06,
+      "loss": 0.6057,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5110845612637728,
+      "grad_norm": 0.40069051623420443,
+      "learning_rate": 4.919244371166733e-06,
+      "loss": 0.608,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5113500597371565,
+      "grad_norm": 0.3802355423096846,
+      "learning_rate": 4.919156327818529e-06,
+      "loss": 0.643,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5116155582105403,
+      "grad_norm": 0.39186629265062317,
+      "learning_rate": 4.919068237290855e-06,
+      "loss": 0.6291,
+      "step": 1927
+    },
+    {
+      "epoch": 0.511881056683924,
+      "grad_norm": 0.4238564062674733,
+      "learning_rate": 4.9189800995854266e-06,
+      "loss": 0.6116,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5121465551573079,
+      "grad_norm": 0.3834861960942196,
+      "learning_rate": 4.918891914703964e-06,
+      "loss": 0.6637,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5124120536306916,
+      "grad_norm": 0.39934349711203193,
+      "learning_rate": 4.918803682648186e-06,
+      "loss": 0.6033,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5126775521040754,
+      "grad_norm": 0.4200366020980555,
+      "learning_rate": 4.918715403419813e-06,
+      "loss": 0.5966,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5129430505774591,
+      "grad_norm": 0.376407814148563,
+      "learning_rate": 4.91862707702057e-06,
+      "loss": 0.6287,
+      "step": 1932
+    },
+    {
+      "epoch": 0.513208549050843,
+      "grad_norm": 0.40309078261311976,
+      "learning_rate": 4.918538703452176e-06,
+      "loss": 0.6408,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5134740475242268,
+      "grad_norm": 0.3819125400165194,
+      "learning_rate": 4.9184502827163555e-06,
+      "loss": 0.6342,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5137395459976105,
+      "grad_norm": 0.40529513749499674,
+      "learning_rate": 4.918361814814833e-06,
+      "loss": 0.6229,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5140050444709943,
+      "grad_norm": 0.38053152431488924,
+      "learning_rate": 4.918273299749334e-06,
+      "loss": 0.6413,
+      "step": 1936
+    },
+    {
+      "epoch": 0.514270542944378,
+      "grad_norm": 0.38757573142446594,
+      "learning_rate": 4.918184737521585e-06,
+      "loss": 0.6348,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5145360414177619,
+      "grad_norm": 0.3818057490279428,
+      "learning_rate": 4.918096128133313e-06,
+      "loss": 0.5663,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5148015398911456,
+      "grad_norm": 0.4005751737622086,
+      "learning_rate": 4.918007471586246e-06,
+      "loss": 0.6036,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5150670383645294,
+      "grad_norm": 0.39270978192568645,
+      "learning_rate": 4.917918767882113e-06,
+      "loss": 0.6767,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5153325368379131,
+      "grad_norm": 0.3763477916252793,
+      "learning_rate": 4.917830017022643e-06,
+      "loss": 0.6385,
+      "step": 1941
+    },
+    {
+      "epoch": 0.515598035311297,
+      "grad_norm": 0.39086733670678614,
+      "learning_rate": 4.917741219009569e-06,
+      "loss": 0.632,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5158635337846808,
+      "grad_norm": 0.40244996628122104,
+      "learning_rate": 4.917652373844621e-06,
+      "loss": 0.6464,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.39273291272186367,
+      "learning_rate": 4.917563481529532e-06,
+      "loss": 0.6733,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5163945307314483,
+      "grad_norm": 0.39209331331117075,
+      "learning_rate": 4.9174745420660365e-06,
+      "loss": 0.6217,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5166600292048321,
+      "grad_norm": 0.3839363807793837,
+      "learning_rate": 4.917385555455868e-06,
+      "loss": 0.6422,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5169255276782159,
+      "grad_norm": 0.43240063995311945,
+      "learning_rate": 4.917296521700763e-06,
+      "loss": 0.6074,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5171910261515996,
+      "grad_norm": 0.41508983897325225,
+      "learning_rate": 4.917207440802456e-06,
+      "loss": 0.6111,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5174565246249834,
+      "grad_norm": 0.41016295935876546,
+      "learning_rate": 4.9171183127626865e-06,
+      "loss": 0.6234,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5177220230983672,
+      "grad_norm": 0.3954904487286936,
+      "learning_rate": 4.917029137583192e-06,
+      "loss": 0.5951,
+      "step": 1950
+    },
+    {
+      "epoch": 0.517987521571751,
+      "grad_norm": 0.394923398565474,
+      "learning_rate": 4.916939915265711e-06,
+      "loss": 0.6198,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5182530200451347,
+      "grad_norm": 0.48502633921023514,
+      "learning_rate": 4.916850645811984e-06,
+      "loss": 0.6094,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 0.3786627461567855,
+      "learning_rate": 4.916761329223752e-06,
+      "loss": 0.6312,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5187840169919022,
+      "grad_norm": 0.40261587125930187,
+      "learning_rate": 4.916671965502757e-06,
+      "loss": 0.6369,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5190495154652861,
+      "grad_norm": 0.41534193470608577,
+      "learning_rate": 4.916582554650741e-06,
+      "loss": 0.6197,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5193150139386699,
+      "grad_norm": 0.46011395269336497,
+      "learning_rate": 4.916493096669449e-06,
+      "loss": 0.6118,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5195805124120536,
+      "grad_norm": 0.40225877249385233,
+      "learning_rate": 4.916403591560625e-06,
+      "loss": 0.6054,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5198460108854374,
+      "grad_norm": 0.415958633591973,
+      "learning_rate": 4.916314039326014e-06,
+      "loss": 0.6552,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5201115093588212,
+      "grad_norm": 0.46886199065225426,
+      "learning_rate": 4.916224439967364e-06,
+      "loss": 0.6294,
+      "step": 1959
+    },
+    {
+      "epoch": 0.520377007832205,
+      "grad_norm": 0.40599814251803173,
+      "learning_rate": 4.91613479348642e-06,
+      "loss": 0.6553,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5206425063055887,
+      "grad_norm": 0.3836047734668553,
+      "learning_rate": 4.916045099884934e-06,
+      "loss": 0.6135,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5209080047789725,
+      "grad_norm": 0.3856827309329937,
+      "learning_rate": 4.915955359164652e-06,
+      "loss": 0.6171,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5211735032523563,
+      "grad_norm": 0.40526618476236925,
+      "learning_rate": 4.915865571327324e-06,
+      "loss": 0.6311,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5214390017257401,
+      "grad_norm": 0.38913229193634735,
+      "learning_rate": 4.915775736374704e-06,
+      "loss": 0.659,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5217045001991238,
+      "grad_norm": 0.38549491006600756,
+      "learning_rate": 4.915685854308542e-06,
+      "loss": 0.6519,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5219699986725076,
+      "grad_norm": 0.3960830305499255,
+      "learning_rate": 4.915595925130591e-06,
+      "loss": 0.6061,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5222354971458915,
+      "grad_norm": 0.3766447961195249,
+      "learning_rate": 4.915505948842604e-06,
+      "loss": 0.5812,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5225009956192752,
+      "grad_norm": 0.5873945390769875,
+      "learning_rate": 4.915415925446338e-06,
+      "loss": 0.586,
+      "step": 1968
+    },
+    {
+      "epoch": 0.522766494092659,
+      "grad_norm": 0.3947862157472956,
+      "learning_rate": 4.9153258549435474e-06,
+      "loss": 0.6375,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5230319925660427,
+      "grad_norm": 0.4009150621680005,
+      "learning_rate": 4.915235737335989e-06,
+      "loss": 0.6542,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5232974910394266,
+      "grad_norm": 0.39546563680275615,
+      "learning_rate": 4.915145572625419e-06,
+      "loss": 0.6377,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5235629895128103,
+      "grad_norm": 0.39279509113501543,
+      "learning_rate": 4.9150553608135985e-06,
+      "loss": 0.6694,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5238284879861941,
+      "grad_norm": 0.3912170569135838,
+      "learning_rate": 4.914965101902285e-06,
+      "loss": 0.6344,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5240939864595778,
+      "grad_norm": 0.39161873637847233,
+      "learning_rate": 4.914874795893239e-06,
+      "loss": 0.6263,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5243594849329616,
+      "grad_norm": 0.3850736598857089,
+      "learning_rate": 4.914784442788222e-06,
+      "loss": 0.6469,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5246249834063454,
+      "grad_norm": 0.3826200214861477,
+      "learning_rate": 4.914694042588995e-06,
+      "loss": 0.6107,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5248904818797292,
+      "grad_norm": 0.41095595612741176,
+      "learning_rate": 4.914603595297324e-06,
+      "loss": 0.5918,
+      "step": 1977
+    },
+    {
+      "epoch": 0.525155980353113,
+      "grad_norm": 0.37871380973928637,
+      "learning_rate": 4.914513100914969e-06,
+      "loss": 0.6429,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5254214788264967,
+      "grad_norm": 0.38368522679277045,
+      "learning_rate": 4.914422559443698e-06,
+      "loss": 0.5959,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5256869772998806,
+      "grad_norm": 0.3828986814526875,
+      "learning_rate": 4.914331970885275e-06,
+      "loss": 0.6302,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5259524757732643,
+      "grad_norm": 0.3981471908610477,
+      "learning_rate": 4.914241335241468e-06,
+      "loss": 0.6474,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5262179742466481,
+      "grad_norm": 0.41024766901173804,
+      "learning_rate": 4.914150652514043e-06,
+      "loss": 0.6081,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5264834727200318,
+      "grad_norm": 0.4124643430761794,
+      "learning_rate": 4.914059922704769e-06,
+      "loss": 0.6388,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5267489711934157,
+      "grad_norm": 0.40345225455287465,
+      "learning_rate": 4.913969145815417e-06,
+      "loss": 0.6567,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5270144696667994,
+      "grad_norm": 0.38991503410506784,
+      "learning_rate": 4.9138783218477556e-06,
+      "loss": 0.6512,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5272799681401832,
+      "grad_norm": 0.39642400756437307,
+      "learning_rate": 4.913787450803557e-06,
+      "loss": 0.6559,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5275454666135669,
+      "grad_norm": 0.40064833637557923,
+      "learning_rate": 4.913696532684593e-06,
+      "loss": 0.6459,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5278109650869508,
+      "grad_norm": 0.38159091614178725,
+      "learning_rate": 4.913605567492636e-06,
+      "loss": 0.6303,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5280764635603346,
+      "grad_norm": 0.37690842226897187,
+      "learning_rate": 4.913514555229463e-06,
+      "loss": 0.6044,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5283419620337183,
+      "grad_norm": 0.40763192542970106,
+      "learning_rate": 4.913423495896845e-06,
+      "loss": 0.652,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5286074605071021,
+      "grad_norm": 0.37128921995742165,
+      "learning_rate": 4.91333238949656e-06,
+      "loss": 0.6263,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5288729589804858,
+      "grad_norm": 0.38138438301490124,
+      "learning_rate": 4.9132412360303855e-06,
+      "loss": 0.6325,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5291384574538697,
+      "grad_norm": 0.389781491818443,
+      "learning_rate": 4.9131500355000975e-06,
+      "loss": 0.6023,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5294039559272534,
+      "grad_norm": 0.3809007368154659,
+      "learning_rate": 4.913058787907476e-06,
+      "loss": 0.6644,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5296694544006372,
+      "grad_norm": 0.37908606140321605,
+      "learning_rate": 4.9129674932542996e-06,
+      "loss": 0.6367,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5299349528740209,
+      "grad_norm": 0.38096417942253313,
+      "learning_rate": 4.912876151542349e-06,
+      "loss": 0.6415,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5302004513474048,
+      "grad_norm": 0.38040443064803364,
+      "learning_rate": 4.912784762773406e-06,
+      "loss": 0.6529,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5304659498207885,
+      "grad_norm": 0.389670144775674,
+      "learning_rate": 4.912693326949252e-06,
+      "loss": 0.597,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5307314482941723,
+      "grad_norm": 0.36797849256191056,
+      "learning_rate": 4.912601844071671e-06,
+      "loss": 0.6218,
+      "step": 1999
+    },
+    {
+      "epoch": 0.530996946767556,
+      "grad_norm": 0.3887401473220355,
+      "learning_rate": 4.912510314142448e-06,
+      "loss": 0.6188,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5312624452409399,
+      "grad_norm": 0.394670833903728,
+      "learning_rate": 4.912418737163366e-06,
+      "loss": 0.6261,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5315279437143237,
+      "grad_norm": 0.3991550639507201,
+      "learning_rate": 4.912327113136212e-06,
+      "loss": 0.6313,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5317934421877074,
+      "grad_norm": 0.41440807456952067,
+      "learning_rate": 4.912235442062773e-06,
+      "loss": 0.6428,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5320589406610912,
+      "grad_norm": 0.39260958449403605,
+      "learning_rate": 4.9121437239448376e-06,
+      "loss": 0.6329,
+      "step": 2004
+    },
+    {
+      "epoch": 0.532324439134475,
+      "grad_norm": 0.4215609059821753,
+      "learning_rate": 4.912051958784193e-06,
+      "loss": 0.6451,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5325899376078588,
+      "grad_norm": 0.44110540921989233,
+      "learning_rate": 4.911960146582629e-06,
+      "loss": 0.6123,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5328554360812425,
+      "grad_norm": 0.3919226211228559,
+      "learning_rate": 4.911868287341936e-06,
+      "loss": 0.6093,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5331209345546263,
+      "grad_norm": 0.39614996621878684,
+      "learning_rate": 4.911776381063908e-06,
+      "loss": 0.5863,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5333864330280101,
+      "grad_norm": 0.3842995675771415,
+      "learning_rate": 4.911684427750335e-06,
+      "loss": 0.6377,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5336519315013939,
+      "grad_norm": 0.38487675830179735,
+      "learning_rate": 4.91159242740301e-06,
+      "loss": 0.6142,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5339174299747776,
+      "grad_norm": 0.3886019426889275,
+      "learning_rate": 4.911500380023728e-06,
+      "loss": 0.632,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5341829284481614,
+      "grad_norm": 0.39223741962400865,
+      "learning_rate": 4.911408285614286e-06,
+      "loss": 0.6526,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5344484269215451,
+      "grad_norm": 0.41792845694869524,
+      "learning_rate": 4.911316144176477e-06,
+      "loss": 0.6127,
+      "step": 2013
+    },
+    {
+      "epoch": 0.534713925394929,
+      "grad_norm": 0.37886528393274843,
+      "learning_rate": 4.911223955712099e-06,
+      "loss": 0.5881,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5349794238683128,
+      "grad_norm": 0.3922125535634023,
+      "learning_rate": 4.91113172022295e-06,
+      "loss": 0.6373,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5352449223416965,
+      "grad_norm": 0.4099652113635852,
+      "learning_rate": 4.91103943771083e-06,
+      "loss": 0.6486,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5355104208150803,
+      "grad_norm": 0.3925792103145467,
+      "learning_rate": 4.910947108177537e-06,
+      "loss": 0.6365,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5357759192884641,
+      "grad_norm": 0.3914181019984941,
+      "learning_rate": 4.910854731624873e-06,
+      "loss": 0.6489,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5360414177618479,
+      "grad_norm": 0.4027266335702039,
+      "learning_rate": 4.910762308054638e-06,
+      "loss": 0.5776,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5363069162352316,
+      "grad_norm": 0.3698376211538782,
+      "learning_rate": 4.910669837468637e-06,
+      "loss": 0.6194,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5365724147086154,
+      "grad_norm": 0.3894126970083076,
+      "learning_rate": 4.91057731986867e-06,
+      "loss": 0.6479,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5368379131819992,
+      "grad_norm": 0.3885370586836856,
+      "learning_rate": 4.910484755256544e-06,
+      "loss": 0.6217,
+      "step": 2022
+    },
+    {
+      "epoch": 0.537103411655383,
+      "grad_norm": 0.38024985391777083,
+      "learning_rate": 4.910392143634064e-06,
+      "loss": 0.6163,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5373689101287668,
+      "grad_norm": 0.3978649928960978,
+      "learning_rate": 4.910299485003034e-06,
+      "loss": 0.5969,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5376344086021505,
+      "grad_norm": 0.38999713662535207,
+      "learning_rate": 4.9102067793652645e-06,
+      "loss": 0.6124,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5378999070755344,
+      "grad_norm": 0.3771815764071492,
+      "learning_rate": 4.91011402672256e-06,
+      "loss": 0.6173,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5381654055489181,
+      "grad_norm": 0.4011553085079414,
+      "learning_rate": 4.9100212270767315e-06,
+      "loss": 0.6015,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5384309040223019,
+      "grad_norm": 0.3810920171325633,
+      "learning_rate": 4.909928380429588e-06,
+      "loss": 0.6359,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5386964024956856,
+      "grad_norm": 0.3767688501902981,
+      "learning_rate": 4.9098354867829415e-06,
+      "loss": 0.6153,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5389619009690694,
+      "grad_norm": 0.38881017925996,
+      "learning_rate": 4.909742546138602e-06,
+      "loss": 0.6388,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5392273994424532,
+      "grad_norm": 0.38149688393532777,
+      "learning_rate": 4.909649558498383e-06,
+      "loss": 0.5973,
+      "step": 2031
+    },
+    {
+      "epoch": 0.539492897915837,
+      "grad_norm": 0.37580474143004144,
+      "learning_rate": 4.909556523864098e-06,
+      "loss": 0.5982,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5397583963892207,
+      "grad_norm": 0.3882417055617691,
+      "learning_rate": 4.909463442237561e-06,
+      "loss": 0.6263,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5400238948626045,
+      "grad_norm": 0.38715921572958895,
+      "learning_rate": 4.909370313620587e-06,
+      "loss": 0.6442,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5402893933359884,
+      "grad_norm": 0.3839340050720667,
+      "learning_rate": 4.909277138014994e-06,
+      "loss": 0.6274,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5405548918093721,
+      "grad_norm": 0.3865694160339814,
+      "learning_rate": 4.909183915422596e-06,
+      "loss": 0.576,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5408203902827559,
+      "grad_norm": 0.3855764536735899,
+      "learning_rate": 4.909090645845214e-06,
+      "loss": 0.6458,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5410858887561396,
+      "grad_norm": 0.381334965046924,
+      "learning_rate": 4.9089973292846665e-06,
+      "loss": 0.5939,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5413513872295235,
+      "grad_norm": 0.3890453294774196,
+      "learning_rate": 4.908903965742772e-06,
+      "loss": 0.6255,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5416168857029072,
+      "grad_norm": 0.41455624731884266,
+      "learning_rate": 4.908810555221352e-06,
+      "loss": 0.6487,
+      "step": 2040
+    },
+    {
+      "epoch": 0.541882384176291,
+      "grad_norm": 0.3987133890260368,
+      "learning_rate": 4.90871709772223e-06,
+      "loss": 0.6385,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5421478826496747,
+      "grad_norm": 0.3966554965078316,
+      "learning_rate": 4.9086235932472254e-06,
+      "loss": 0.5804,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5424133811230586,
+      "grad_norm": 0.3891784830525446,
+      "learning_rate": 4.908530041798164e-06,
+      "loss": 0.6051,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5426788795964423,
+      "grad_norm": 0.37050107770847845,
+      "learning_rate": 4.908436443376869e-06,
+      "loss": 0.6068,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5429443780698261,
+      "grad_norm": 0.39616810997144736,
+      "learning_rate": 4.908342797985167e-06,
+      "loss": 0.6365,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5432098765432098,
+      "grad_norm": 0.3992223883897869,
+      "learning_rate": 4.908249105624884e-06,
+      "loss": 0.6466,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5434753750165937,
+      "grad_norm": 0.40806638014668023,
+      "learning_rate": 4.908155366297846e-06,
+      "loss": 0.6417,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5437408734899775,
+      "grad_norm": 0.41048772467987144,
+      "learning_rate": 4.908061580005884e-06,
+      "loss": 0.6069,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5440063719633612,
+      "grad_norm": 0.3807003999786117,
+      "learning_rate": 4.907967746750824e-06,
+      "loss": 0.6143,
+      "step": 2049
+    },
+    {
+      "epoch": 0.544271870436745,
+      "grad_norm": 0.3906568534344132,
+      "learning_rate": 4.907873866534498e-06,
+      "loss": 0.6267,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5445373689101287,
+      "grad_norm": 0.38566281253909396,
+      "learning_rate": 4.907779939358735e-06,
+      "loss": 0.6033,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5448028673835126,
+      "grad_norm": 0.3758823380254977,
+      "learning_rate": 4.907685965225369e-06,
+      "loss": 0.6175,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5450683658568963,
+      "grad_norm": 0.4113044262717281,
+      "learning_rate": 4.907591944136231e-06,
+      "loss": 0.6416,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5453338643302801,
+      "grad_norm": 0.3938740571600203,
+      "learning_rate": 4.907497876093155e-06,
+      "loss": 0.5973,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5455993628036638,
+      "grad_norm": 0.38334057621592116,
+      "learning_rate": 4.9074037610979765e-06,
+      "loss": 0.6369,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5458648612770477,
+      "grad_norm": 0.37107462150204695,
+      "learning_rate": 4.90730959915253e-06,
+      "loss": 0.6142,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5461303597504314,
+      "grad_norm": 0.38365278644512046,
+      "learning_rate": 4.907215390258653e-06,
+      "loss": 0.6447,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5463958582238152,
+      "grad_norm": 0.3943919668941426,
+      "learning_rate": 4.907121134418181e-06,
+      "loss": 0.5708,
+      "step": 2058
+    },
+    {
+      "epoch": 0.546661356697199,
+      "grad_norm": 0.4163368103225318,
+      "learning_rate": 4.907026831632953e-06,
+      "loss": 0.6066,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5469268551705828,
+      "grad_norm": 0.38527796730582103,
+      "learning_rate": 4.906932481904809e-06,
+      "loss": 0.6412,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5471923536439666,
+      "grad_norm": 0.37312812386696503,
+      "learning_rate": 4.906838085235589e-06,
+      "loss": 0.6116,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5474578521173503,
+      "grad_norm": 0.3874072289231269,
+      "learning_rate": 4.9067436416271315e-06,
+      "loss": 0.6487,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5477233505907341,
+      "grad_norm": 0.4022103791020934,
+      "learning_rate": 4.906649151081282e-06,
+      "loss": 0.6378,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5479888490641179,
+      "grad_norm": 0.38061999121785595,
+      "learning_rate": 4.906554613599881e-06,
+      "loss": 0.6464,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5482543475375017,
+      "grad_norm": 0.3859123159817981,
+      "learning_rate": 4.906460029184773e-06,
+      "loss": 0.6227,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5485198460108854,
+      "grad_norm": 0.3961735599825678,
+      "learning_rate": 4.906365397837803e-06,
+      "loss": 0.6385,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5487853444842692,
+      "grad_norm": 0.394029028948993,
+      "learning_rate": 4.906270719560815e-06,
+      "loss": 0.6126,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5490508429576529,
+      "grad_norm": 0.3693685697958605,
+      "learning_rate": 4.906175994355656e-06,
+      "loss": 0.6442,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5493163414310368,
+      "grad_norm": 0.39211302516438096,
+      "learning_rate": 4.906081222224174e-06,
+      "loss": 0.5986,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5495818399044206,
+      "grad_norm": 0.3830786351532926,
+      "learning_rate": 4.905986403168218e-06,
+      "loss": 0.5719,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5498473383778043,
+      "grad_norm": 0.3782755714287307,
+      "learning_rate": 4.9058915371896354e-06,
+      "loss": 0.623,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5501128368511881,
+      "grad_norm": 0.3990298814224943,
+      "learning_rate": 4.9057966242902775e-06,
+      "loss": 0.6315,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5503783353245719,
+      "grad_norm": 0.3939806321713868,
+      "learning_rate": 4.905701664471994e-06,
+      "loss": 0.6092,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5506438337979557,
+      "grad_norm": 0.37623921441529895,
+      "learning_rate": 4.90560665773664e-06,
+      "loss": 0.6366,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5509093322713394,
+      "grad_norm": 0.3757539305206352,
+      "learning_rate": 4.905511604086065e-06,
+      "loss": 0.6525,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5511748307447232,
+      "grad_norm": 0.39633723753870215,
+      "learning_rate": 4.905416503522124e-06,
+      "loss": 0.6233,
+      "step": 2076
+    },
+    {
+      "epoch": 0.551440329218107,
+      "grad_norm": 0.388027801508874,
+      "learning_rate": 4.905321356046672e-06,
+      "loss": 0.6111,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5517058276914908,
+      "grad_norm": 0.3899109684092773,
+      "learning_rate": 4.905226161661564e-06,
+      "loss": 0.6576,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5519713261648745,
+      "grad_norm": 0.40443956482179805,
+      "learning_rate": 4.905130920368657e-06,
+      "loss": 0.6243,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5522368246382583,
+      "grad_norm": 0.38594194174105484,
+      "learning_rate": 4.905035632169808e-06,
+      "loss": 0.6381,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5525023231116422,
+      "grad_norm": 0.3889138022398403,
+      "learning_rate": 4.904940297066876e-06,
+      "loss": 0.62,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5527678215850259,
+      "grad_norm": 0.3705549552623768,
+      "learning_rate": 4.9048449150617195e-06,
+      "loss": 0.631,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5530333200584097,
+      "grad_norm": 0.36827424623469257,
+      "learning_rate": 4.9047494861562e-06,
+      "loss": 0.5987,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5532988185317934,
+      "grad_norm": 0.3741083791691221,
+      "learning_rate": 4.9046540103521765e-06,
+      "loss": 0.6235,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5535643170051773,
+      "grad_norm": 0.386302888384375,
+      "learning_rate": 4.904558487651513e-06,
+      "loss": 0.5965,
+      "step": 2085
+    },
+    {
+      "epoch": 0.553829815478561,
+      "grad_norm": 0.3910143789167576,
+      "learning_rate": 4.904462918056071e-06,
+      "loss": 0.6102,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5540953139519448,
+      "grad_norm": 0.3806114225783564,
+      "learning_rate": 4.904367301567715e-06,
+      "loss": 0.6444,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5543608124253285,
+      "grad_norm": 0.3758951461070809,
+      "learning_rate": 4.90427163818831e-06,
+      "loss": 0.6482,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5546263108987123,
+      "grad_norm": 0.38715614700989387,
+      "learning_rate": 4.904175927919722e-06,
+      "loss": 0.6361,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5548918093720961,
+      "grad_norm": 0.38367039724574203,
+      "learning_rate": 4.904080170763816e-06,
+      "loss": 0.634,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5551573078454799,
+      "grad_norm": 0.4168224051384316,
+      "learning_rate": 4.903984366722461e-06,
+      "loss": 0.6188,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5554228063188636,
+      "grad_norm": 0.3817421936850519,
+      "learning_rate": 4.903888515797524e-06,
+      "loss": 0.6048,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5556883047922474,
+      "grad_norm": 0.3670280478052143,
+      "learning_rate": 4.903792617990876e-06,
+      "loss": 0.6051,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5559538032656313,
+      "grad_norm": 0.3938453401105219,
+      "learning_rate": 4.903696673304387e-06,
+      "loss": 0.6186,
+      "step": 2094
+    },
+    {
+      "epoch": 0.556219301739015,
+      "grad_norm": 0.39360141206843813,
+      "learning_rate": 4.903600681739926e-06,
+      "loss": 0.6329,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5564848002123988,
+      "grad_norm": 0.3954338510193279,
+      "learning_rate": 4.903504643299368e-06,
+      "loss": 0.6368,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5567502986857825,
+      "grad_norm": 0.3799079005061654,
+      "learning_rate": 4.903408557984585e-06,
+      "loss": 0.6386,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5570157971591664,
+      "grad_norm": 0.37192170883761927,
+      "learning_rate": 4.90331242579745e-06,
+      "loss": 0.6183,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5572812956325501,
+      "grad_norm": 0.38815571622690004,
+      "learning_rate": 4.903216246739837e-06,
+      "loss": 0.6415,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5575467941059339,
+      "grad_norm": 0.38535358872207337,
+      "learning_rate": 4.903120020813625e-06,
+      "loss": 0.6452,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5578122925793176,
+      "grad_norm": 0.39592321936836095,
+      "learning_rate": 4.903023748020688e-06,
+      "loss": 0.6255,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5580777910527015,
+      "grad_norm": 0.3863115053098059,
+      "learning_rate": 4.9029274283629046e-06,
+      "loss": 0.6427,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5583432895260853,
+      "grad_norm": 0.38221772361703116,
+      "learning_rate": 4.902831061842153e-06,
+      "loss": 0.6506,
+      "step": 2103
+    },
+    {
+      "epoch": 0.558608787999469,
+      "grad_norm": 0.3973981667806005,
+      "learning_rate": 4.9027346484603116e-06,
+      "loss": 0.6004,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5588742864728528,
+      "grad_norm": 0.3907938965791774,
+      "learning_rate": 4.902638188219262e-06,
+      "loss": 0.6335,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5591397849462365,
+      "grad_norm": 0.3860874537813031,
+      "learning_rate": 4.902541681120886e-06,
+      "loss": 0.6203,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5594052834196204,
+      "grad_norm": 0.39445636651628896,
+      "learning_rate": 4.9024451271670635e-06,
+      "loss": 0.5878,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5596707818930041,
+      "grad_norm": 0.3991038587018792,
+      "learning_rate": 4.9023485263596795e-06,
+      "loss": 0.6103,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5599362803663879,
+      "grad_norm": 0.38404543983183625,
+      "learning_rate": 4.902251878700618e-06,
+      "loss": 0.6428,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5602017788397716,
+      "grad_norm": 0.3758713989457986,
+      "learning_rate": 4.902155184191762e-06,
+      "loss": 0.6637,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5604672773131555,
+      "grad_norm": 0.39684714170555097,
+      "learning_rate": 4.902058442834999e-06,
+      "loss": 0.6469,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5607327757865392,
+      "grad_norm": 0.3955355601685318,
+      "learning_rate": 4.901961654632214e-06,
+      "loss": 0.6513,
+      "step": 2112
+    },
+    {
+      "epoch": 0.560998274259923,
+      "grad_norm": 0.38149509209917476,
+      "learning_rate": 4.901864819585297e-06,
+      "loss": 0.6326,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5612637727333067,
+      "grad_norm": 0.38726352688714627,
+      "learning_rate": 4.901767937696135e-06,
+      "loss": 0.6567,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5615292712066906,
+      "grad_norm": 0.3800742017223432,
+      "learning_rate": 4.901671008966618e-06,
+      "loss": 0.6582,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5617947696800744,
+      "grad_norm": 0.3746328300006796,
+      "learning_rate": 4.901574033398635e-06,
+      "loss": 0.6445,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5620602681534581,
+      "grad_norm": 0.38260282871000356,
+      "learning_rate": 4.901477010994079e-06,
+      "loss": 0.5919,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5623257666268419,
+      "grad_norm": 0.3740965623998434,
+      "learning_rate": 4.901379941754841e-06,
+      "loss": 0.6254,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5625912651002257,
+      "grad_norm": 0.3761963427338876,
+      "learning_rate": 4.901282825682815e-06,
+      "loss": 0.6224,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5628567635736095,
+      "grad_norm": 0.38083963318217334,
+      "learning_rate": 4.901185662779895e-06,
+      "loss": 0.612,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5631222620469932,
+      "grad_norm": 0.392352340225909,
+      "learning_rate": 4.901088453047975e-06,
+      "loss": 0.6406,
+      "step": 2121
+    },
+    {
+      "epoch": 0.563387760520377,
+      "grad_norm": 0.37933525053159506,
+      "learning_rate": 4.900991196488951e-06,
+      "loss": 0.5823,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5636532589937607,
+      "grad_norm": 0.38008886707506945,
+      "learning_rate": 4.90089389310472e-06,
+      "loss": 0.6119,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5639187574671446,
+      "grad_norm": 0.38651929782239175,
+      "learning_rate": 4.90079654289718e-06,
+      "loss": 0.644,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5641842559405283,
+      "grad_norm": 0.39176422515446785,
+      "learning_rate": 4.900699145868228e-06,
+      "loss": 0.6371,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5644497544139121,
+      "grad_norm": 0.393499491297263,
+      "learning_rate": 4.900601702019767e-06,
+      "loss": 0.587,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5647152528872958,
+      "grad_norm": 0.3877468380146798,
+      "learning_rate": 4.900504211353694e-06,
+      "loss": 0.5753,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5649807513606797,
+      "grad_norm": 0.39194630719423856,
+      "learning_rate": 4.900406673871912e-06,
+      "loss": 0.6076,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5652462498340635,
+      "grad_norm": 0.3865805070743415,
+      "learning_rate": 4.900309089576321e-06,
+      "loss": 0.6458,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5655117483074472,
+      "grad_norm": 0.413154720498655,
+      "learning_rate": 4.900211458468827e-06,
+      "loss": 0.6264,
+      "step": 2130
+    },
+    {
+      "epoch": 0.565777246780831,
+      "grad_norm": 0.37762177320125556,
+      "learning_rate": 4.900113780551332e-06,
+      "loss": 0.6636,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5660427452542148,
+      "grad_norm": 0.39938122548014515,
+      "learning_rate": 4.900016055825743e-06,
+      "loss": 0.5983,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5663082437275986,
+      "grad_norm": 0.3825194381566569,
+      "learning_rate": 4.899918284293964e-06,
+      "loss": 0.6577,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5665737422009823,
+      "grad_norm": 0.41054618584625147,
+      "learning_rate": 4.899820465957903e-06,
+      "loss": 0.6314,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5668392406743661,
+      "grad_norm": 0.37313881058907006,
+      "learning_rate": 4.899722600819467e-06,
+      "loss": 0.6299,
+      "step": 2135
+    },
+    {
+      "epoch": 0.56710473914775,
+      "grad_norm": 0.40422916207703785,
+      "learning_rate": 4.899624688880564e-06,
+      "loss": 0.6348,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5673702376211337,
+      "grad_norm": 0.374248197860821,
+      "learning_rate": 4.899526730143104e-06,
+      "loss": 0.6403,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5676357360945175,
+      "grad_norm": 0.37025737340891063,
+      "learning_rate": 4.8994287246089985e-06,
+      "loss": 0.6463,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5679012345679012,
+      "grad_norm": 0.3769084840508759,
+      "learning_rate": 4.899330672280158e-06,
+      "loss": 0.6328,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5681667330412851,
+      "grad_norm": 0.3858461436635138,
+      "learning_rate": 4.899232573158495e-06,
+      "loss": 0.6214,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5684322315146688,
+      "grad_norm": 0.3824468158039468,
+      "learning_rate": 4.899134427245922e-06,
+      "loss": 0.5963,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5686977299880526,
+      "grad_norm": 0.3880901580951896,
+      "learning_rate": 4.899036234544354e-06,
+      "loss": 0.6617,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5689632284614363,
+      "grad_norm": 0.4064318711168065,
+      "learning_rate": 4.898937995055706e-06,
+      "loss": 0.6264,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5692287269348201,
+      "grad_norm": 0.3923192035373891,
+      "learning_rate": 4.898839708781893e-06,
+      "loss": 0.6457,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5694942254082039,
+      "grad_norm": 0.3888934408786419,
+      "learning_rate": 4.8987413757248316e-06,
+      "loss": 0.609,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5697597238815877,
+      "grad_norm": 0.3908300467173408,
+      "learning_rate": 4.898642995886441e-06,
+      "loss": 0.6109,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5700252223549714,
+      "grad_norm": 0.3863165067932937,
+      "learning_rate": 4.898544569268639e-06,
+      "loss": 0.637,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5702907208283552,
+      "grad_norm": 0.3935153573179178,
+      "learning_rate": 4.898446095873345e-06,
+      "loss": 0.6266,
+      "step": 2148
+    },
+    {
+      "epoch": 0.570556219301739,
+      "grad_norm": 0.37662569809214685,
+      "learning_rate": 4.8983475757024805e-06,
+      "loss": 0.6219,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5708217177751228,
+      "grad_norm": 0.38826483250184635,
+      "learning_rate": 4.898249008757965e-06,
+      "loss": 0.6317,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5710872162485066,
+      "grad_norm": 0.4313419777251204,
+      "learning_rate": 4.898150395041723e-06,
+      "loss": 0.6306,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5713527147218903,
+      "grad_norm": 0.3847667107954513,
+      "learning_rate": 4.898051734555676e-06,
+      "loss": 0.6373,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5716182131952742,
+      "grad_norm": 0.4021383500058913,
+      "learning_rate": 4.897953027301748e-06,
+      "loss": 0.644,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5718837116686579,
+      "grad_norm": 0.4219288427264436,
+      "learning_rate": 4.897854273281866e-06,
+      "loss": 0.6115,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5721492101420417,
+      "grad_norm": 0.38071541982270457,
+      "learning_rate": 4.897755472497954e-06,
+      "loss": 0.5962,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5724147086154254,
+      "grad_norm": 0.4540849896264911,
+      "learning_rate": 4.89765662495194e-06,
+      "loss": 0.6148,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5726802070888093,
+      "grad_norm": 0.3842505208160628,
+      "learning_rate": 4.897557730645751e-06,
+      "loss": 0.6364,
+      "step": 2157
+    },
+    {
+      "epoch": 0.572945705562193,
+      "grad_norm": 0.3844551975074067,
+      "learning_rate": 4.897458789581317e-06,
+      "loss": 0.6173,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5732112040355768,
+      "grad_norm": 0.38431361074353193,
+      "learning_rate": 4.897359801760565e-06,
+      "loss": 0.6352,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5734767025089605,
+      "grad_norm": 0.43270885048710706,
+      "learning_rate": 4.8972607671854276e-06,
+      "loss": 0.6047,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5737422009823443,
+      "grad_norm": 0.3846691149620548,
+      "learning_rate": 4.897161685857836e-06,
+      "loss": 0.6176,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5740076994557282,
+      "grad_norm": 0.4007762887021661,
+      "learning_rate": 4.897062557779722e-06,
+      "loss": 0.6244,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5742731979291119,
+      "grad_norm": 0.3961438131633055,
+      "learning_rate": 4.89696338295302e-06,
+      "loss": 0.6412,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5745386964024957,
+      "grad_norm": 0.390097268979398,
+      "learning_rate": 4.8968641613796634e-06,
+      "loss": 0.6479,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5748041948758794,
+      "grad_norm": 0.38810299744104254,
+      "learning_rate": 4.896764893061586e-06,
+      "loss": 0.6425,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5750696933492633,
+      "grad_norm": 0.3959714972015471,
+      "learning_rate": 4.8966655780007246e-06,
+      "loss": 0.641,
+      "step": 2166
+    },
+    {
+      "epoch": 0.575335191822647,
+      "grad_norm": 0.39502538544557075,
+      "learning_rate": 4.896566216199017e-06,
+      "loss": 0.6516,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5756006902960308,
+      "grad_norm": 0.3893632869954923,
+      "learning_rate": 4.896466807658401e-06,
+      "loss": 0.6515,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5758661887694145,
+      "grad_norm": 0.3749742772574467,
+      "learning_rate": 4.8963673523808146e-06,
+      "loss": 0.5914,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5761316872427984,
+      "grad_norm": 0.3693946699712746,
+      "learning_rate": 4.896267850368196e-06,
+      "loss": 0.6232,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5763971857161821,
+      "grad_norm": 0.405383443756792,
+      "learning_rate": 4.896168301622489e-06,
+      "loss": 0.6182,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5766626841895659,
+      "grad_norm": 0.3732202033074253,
+      "learning_rate": 4.896068706145632e-06,
+      "loss": 0.5644,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5769281826629497,
+      "grad_norm": 0.3793821430485284,
+      "learning_rate": 4.89596906393957e-06,
+      "loss": 0.6167,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5771936811363335,
+      "grad_norm": 0.4111046814110193,
+      "learning_rate": 4.895869375006244e-06,
+      "loss": 0.6211,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5774591796097173,
+      "grad_norm": 0.4054227692967559,
+      "learning_rate": 4.895769639347598e-06,
+      "loss": 0.6486,
+      "step": 2175
+    },
+    {
+      "epoch": 0.577724678083101,
+      "grad_norm": 0.37646422034201293,
+      "learning_rate": 4.895669856965581e-06,
+      "loss": 0.6548,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5779901765564848,
+      "grad_norm": 0.41723005095630505,
+      "learning_rate": 4.895570027862133e-06,
+      "loss": 0.6085,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5782556750298686,
+      "grad_norm": 0.40438159811930047,
+      "learning_rate": 4.895470152039206e-06,
+      "loss": 0.6463,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5785211735032524,
+      "grad_norm": 0.3897077639387259,
+      "learning_rate": 4.895370229498746e-06,
+      "loss": 0.6097,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5787866719766361,
+      "grad_norm": 0.3800236733526297,
+      "learning_rate": 4.895270260242701e-06,
+      "loss": 0.6326,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5790521704500199,
+      "grad_norm": 0.3874467382028264,
+      "learning_rate": 4.895170244273022e-06,
+      "loss": 0.5995,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5793176689234036,
+      "grad_norm": 0.38017708406839484,
+      "learning_rate": 4.895070181591658e-06,
+      "loss": 0.6276,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5795831673967875,
+      "grad_norm": 0.38278814329780914,
+      "learning_rate": 4.894970072200561e-06,
+      "loss": 0.6501,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5798486658701713,
+      "grad_norm": 0.3958442998239954,
+      "learning_rate": 4.894869916101685e-06,
+      "loss": 0.6083,
+      "step": 2184
+    },
+    {
+      "epoch": 0.580114164343555,
+      "grad_norm": 0.38684358144848,
+      "learning_rate": 4.894769713296981e-06,
+      "loss": 0.6409,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5803796628169388,
+      "grad_norm": 0.388722478341223,
+      "learning_rate": 4.8946694637884045e-06,
+      "loss": 0.6105,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 0.3754123202087565,
+      "learning_rate": 4.8945691675779104e-06,
+      "loss": 0.6206,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5809106597637064,
+      "grad_norm": 0.3926535967431366,
+      "learning_rate": 4.894468824667454e-06,
+      "loss": 0.6424,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5811761582370901,
+      "grad_norm": 0.37486026021945246,
+      "learning_rate": 4.894368435058993e-06,
+      "loss": 0.6491,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5814416567104739,
+      "grad_norm": 0.3863608593671605,
+      "learning_rate": 4.894267998754486e-06,
+      "loss": 0.633,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5817071551838577,
+      "grad_norm": 0.3884998634533141,
+      "learning_rate": 4.894167515755891e-06,
+      "loss": 0.6061,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5819726536572415,
+      "grad_norm": 0.38409327701132057,
+      "learning_rate": 4.894066986065167e-06,
+      "loss": 0.6324,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5822381521306252,
+      "grad_norm": 0.3714906413935847,
+      "learning_rate": 4.893966409684274e-06,
+      "loss": 0.6231,
+      "step": 2193
+    },
+    {
+      "epoch": 0.582503650604009,
+      "grad_norm": 0.39370885532732425,
+      "learning_rate": 4.893865786615176e-06,
+      "loss": 0.6145,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5827691490773929,
+      "grad_norm": 0.38889061772731237,
+      "learning_rate": 4.893765116859833e-06,
+      "loss": 0.5917,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5830346475507766,
+      "grad_norm": 0.37937530144908327,
+      "learning_rate": 4.89366440042021e-06,
+      "loss": 0.6295,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5833001460241604,
+      "grad_norm": 0.37975776473212,
+      "learning_rate": 4.89356363729827e-06,
+      "loss": 0.6051,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5835656444975441,
+      "grad_norm": 0.39773401945617287,
+      "learning_rate": 4.893462827495979e-06,
+      "loss": 0.6453,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5838311429709279,
+      "grad_norm": 0.3764356020152879,
+      "learning_rate": 4.893361971015302e-06,
+      "loss": 0.6339,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5840966414443117,
+      "grad_norm": 0.3968692273333052,
+      "learning_rate": 4.8932610678582076e-06,
+      "loss": 0.6714,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5843621399176955,
+      "grad_norm": 0.3703178505898537,
+      "learning_rate": 4.893160118026662e-06,
+      "loss": 0.6337,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5846276383910792,
+      "grad_norm": 0.3930140117563288,
+      "learning_rate": 4.8930591215226355e-06,
+      "loss": 0.622,
+      "step": 2202
+    },
+    {
+      "epoch": 0.584893136864463,
+      "grad_norm": 0.3816700564419683,
+      "learning_rate": 4.892958078348096e-06,
+      "loss": 0.6051,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5851586353378468,
+      "grad_norm": 0.3870884558521837,
+      "learning_rate": 4.892856988505015e-06,
+      "loss": 0.6211,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5854241338112306,
+      "grad_norm": 0.41377118989595385,
+      "learning_rate": 4.892755851995364e-06,
+      "loss": 0.6265,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5856896322846143,
+      "grad_norm": 0.39182707672478273,
+      "learning_rate": 4.8926546688211155e-06,
+      "loss": 0.6161,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5859551307579981,
+      "grad_norm": 0.38699158023072017,
+      "learning_rate": 4.892553438984243e-06,
+      "loss": 0.6379,
+      "step": 2207
+    },
+    {
+      "epoch": 0.586220629231382,
+      "grad_norm": 0.38068111580536557,
+      "learning_rate": 4.8924521624867204e-06,
+      "loss": 0.6017,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5864861277047657,
+      "grad_norm": 0.3752111789320971,
+      "learning_rate": 4.8923508393305224e-06,
+      "loss": 0.6447,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5867516261781495,
+      "grad_norm": 0.3851509876776998,
+      "learning_rate": 4.892249469517626e-06,
+      "loss": 0.6231,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5870171246515332,
+      "grad_norm": 0.3948533170299444,
+      "learning_rate": 4.892148053050008e-06,
+      "loss": 0.6378,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5872826231249171,
+      "grad_norm": 0.398749040086564,
+      "learning_rate": 4.892046589929645e-06,
+      "loss": 0.599,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5875481215983008,
+      "grad_norm": 0.38272011041291865,
+      "learning_rate": 4.891945080158518e-06,
+      "loss": 0.6075,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5878136200716846,
+      "grad_norm": 0.3949612619072957,
+      "learning_rate": 4.891843523738605e-06,
+      "loss": 0.6475,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5880791185450683,
+      "grad_norm": 0.38584000416633896,
+      "learning_rate": 4.8917419206718866e-06,
+      "loss": 0.595,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5883446170184522,
+      "grad_norm": 0.3859712641239847,
+      "learning_rate": 4.891640270960345e-06,
+      "loss": 0.6273,
+      "step": 2216
+    },
+    {
+      "epoch": 0.588610115491836,
+      "grad_norm": 0.39724845837803974,
+      "learning_rate": 4.891538574605962e-06,
+      "loss": 0.6298,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5888756139652197,
+      "grad_norm": 0.3978337776338302,
+      "learning_rate": 4.8914368316107215e-06,
+      "loss": 0.6375,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5891411124386035,
+      "grad_norm": 0.37927679100241796,
+      "learning_rate": 4.891335041976608e-06,
+      "loss": 0.6444,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5894066109119872,
+      "grad_norm": 0.3850657177831566,
+      "learning_rate": 4.891233205705606e-06,
+      "loss": 0.6335,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5896721093853711,
+      "grad_norm": 0.39558329617006666,
+      "learning_rate": 4.891131322799701e-06,
+      "loss": 0.6047,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5899376078587548,
+      "grad_norm": 0.4014748284874707,
+      "learning_rate": 4.891029393260882e-06,
+      "loss": 0.6587,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5902031063321386,
+      "grad_norm": 0.38448678396113367,
+      "learning_rate": 4.890927417091135e-06,
+      "loss": 0.6375,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5904686048055223,
+      "grad_norm": 0.3918819994005177,
+      "learning_rate": 4.890825394292449e-06,
+      "loss": 0.6437,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5907341032789062,
+      "grad_norm": 0.379902795113804,
+      "learning_rate": 4.8907233248668145e-06,
+      "loss": 0.5903,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5909996017522899,
+      "grad_norm": 0.3878970032464588,
+      "learning_rate": 4.890621208816222e-06,
+      "loss": 0.6575,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5912651002256737,
+      "grad_norm": 0.39848748754990604,
+      "learning_rate": 4.890519046142662e-06,
+      "loss": 0.5976,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5915305986990574,
+      "grad_norm": 0.3774780289070293,
+      "learning_rate": 4.890416836848128e-06,
+      "loss": 0.5948,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5917960971724413,
+      "grad_norm": 0.3853094810340306,
+      "learning_rate": 4.890314580934612e-06,
+      "loss": 0.6249,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5920615956458251,
+      "grad_norm": 0.3764111814497864,
+      "learning_rate": 4.890212278404111e-06,
+      "loss": 0.6345,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5923270941192088,
+      "grad_norm": 0.41719540411039174,
+      "learning_rate": 4.890109929258616e-06,
+      "loss": 0.6243,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 0.38991606571766907,
+      "learning_rate": 4.890007533500127e-06,
+      "loss": 0.6322,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5928580910659764,
+      "grad_norm": 0.386810740284029,
+      "learning_rate": 4.889905091130639e-06,
+      "loss": 0.6245,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5931235895393602,
+      "grad_norm": 0.3957528074562135,
+      "learning_rate": 4.889802602152151e-06,
+      "loss": 0.5907,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5933890880127439,
+      "grad_norm": 0.4026387439350841,
+      "learning_rate": 4.889700066566659e-06,
+      "loss": 0.6793,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5936545864861277,
+      "grad_norm": 0.38501982758376346,
+      "learning_rate": 4.889597484376166e-06,
+      "loss": 0.6314,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5939200849595114,
+      "grad_norm": 0.4095472745716219,
+      "learning_rate": 4.8894948555826714e-06,
+      "loss": 0.6493,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5941855834328953,
+      "grad_norm": 0.37240788797093644,
+      "learning_rate": 4.889392180188177e-06,
+      "loss": 0.6136,
+      "step": 2238
+    },
+    {
+      "epoch": 0.594451081906279,
+      "grad_norm": 0.3729662364141783,
+      "learning_rate": 4.889289458194684e-06,
+      "loss": 0.6159,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5947165803796628,
+      "grad_norm": 0.36713331053545234,
+      "learning_rate": 4.889186689604197e-06,
+      "loss": 0.591,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5949820788530465,
+      "grad_norm": 0.3893915458181373,
+      "learning_rate": 4.88908387441872e-06,
+      "loss": 0.6591,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5952475773264304,
+      "grad_norm": 0.3718382314454185,
+      "learning_rate": 4.888981012640257e-06,
+      "loss": 0.6121,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5955130757998142,
+      "grad_norm": 0.3867013733578012,
+      "learning_rate": 4.888878104270816e-06,
+      "loss": 0.6369,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5957785742731979,
+      "grad_norm": 0.3803969294872666,
+      "learning_rate": 4.888775149312403e-06,
+      "loss": 0.6076,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5960440727465817,
+      "grad_norm": 0.394707896190095,
+      "learning_rate": 4.888672147767025e-06,
+      "loss": 0.6186,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5963095712199655,
+      "grad_norm": 0.3851378577706324,
+      "learning_rate": 4.888569099636692e-06,
+      "loss": 0.6166,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5965750696933493,
+      "grad_norm": 0.3969387739544168,
+      "learning_rate": 4.888466004923413e-06,
+      "loss": 0.6316,
+      "step": 2247
+    },
+    {
+      "epoch": 0.596840568166733,
+      "grad_norm": 0.3847333963298914,
+      "learning_rate": 4.8883628636292e-06,
+      "loss": 0.6065,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5971060666401168,
+      "grad_norm": 0.41493574810457373,
+      "learning_rate": 4.888259675756063e-06,
+      "loss": 0.6137,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5973715651135006,
+      "grad_norm": 0.3994941140972394,
+      "learning_rate": 4.888156441306014e-06,
+      "loss": 0.6411,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5976370635868844,
+      "grad_norm": 0.40092443510912823,
+      "learning_rate": 4.888053160281069e-06,
+      "loss": 0.6552,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5979025620602682,
+      "grad_norm": 0.3882861571211597,
+      "learning_rate": 4.8879498326832385e-06,
+      "loss": 0.6154,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5981680605336519,
+      "grad_norm": 0.3713948020326159,
+      "learning_rate": 4.887846458514541e-06,
+      "loss": 0.5616,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5984335590070358,
+      "grad_norm": 0.38384861987675445,
+      "learning_rate": 4.88774303777699e-06,
+      "loss": 0.633,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5986990574804195,
+      "grad_norm": 0.3839817470691487,
+      "learning_rate": 4.887639570472604e-06,
+      "loss": 0.5817,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5989645559538033,
+      "grad_norm": 0.39361941153049085,
+      "learning_rate": 4.887536056603401e-06,
+      "loss": 0.5861,
+      "step": 2256
+    },
+    {
+      "epoch": 0.599230054427187,
+      "grad_norm": 0.371897654657084,
+      "learning_rate": 4.887432496171398e-06,
+      "loss": 0.5875,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5994955529005708,
+      "grad_norm": 0.3888488630457417,
+      "learning_rate": 4.887328889178617e-06,
+      "loss": 0.6195,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5997610513739546,
+      "grad_norm": 0.3828262064183697,
+      "learning_rate": 4.887225235627077e-06,
+      "loss": 0.6017,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6000265498473384,
+      "grad_norm": 0.38484680256981363,
+      "learning_rate": 4.887121535518799e-06,
+      "loss": 0.6761,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6002920483207221,
+      "grad_norm": 0.3840856986296762,
+      "learning_rate": 4.887017788855808e-06,
+      "loss": 0.6035,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6005575467941059,
+      "grad_norm": 0.3849416262305795,
+      "learning_rate": 4.8869139956401255e-06,
+      "loss": 0.632,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6008230452674898,
+      "grad_norm": 0.385096199829097,
+      "learning_rate": 4.8868101558737755e-06,
+      "loss": 0.6467,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6010885437408735,
+      "grad_norm": 0.38411824941438083,
+      "learning_rate": 4.886706269558783e-06,
+      "loss": 0.5553,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6013540422142573,
+      "grad_norm": 0.3801440224246112,
+      "learning_rate": 4.886602336697175e-06,
+      "loss": 0.63,
+      "step": 2265
+    },
+    {
+      "epoch": 0.601619540687641,
+      "grad_norm": 0.38733630432591415,
+      "learning_rate": 4.886498357290978e-06,
+      "loss": 0.6047,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6018850391610249,
+      "grad_norm": 0.41059978073098796,
+      "learning_rate": 4.88639433134222e-06,
+      "loss": 0.6292,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6021505376344086,
+      "grad_norm": 0.3855283486466279,
+      "learning_rate": 4.88629025885293e-06,
+      "loss": 0.6129,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6024160361077924,
+      "grad_norm": 0.41265246113613696,
+      "learning_rate": 4.886186139825137e-06,
+      "loss": 0.6255,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6026815345811761,
+      "grad_norm": 0.40886195111049,
+      "learning_rate": 4.8860819742608715e-06,
+      "loss": 0.6148,
+      "step": 2270
+    },
+    {
+      "epoch": 0.60294703305456,
+      "grad_norm": 0.3764460429812742,
+      "learning_rate": 4.885977762162166e-06,
+      "loss": 0.6421,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6032125315279437,
+      "grad_norm": 0.38148557590078197,
+      "learning_rate": 4.885873503531052e-06,
+      "loss": 0.6015,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6034780300013275,
+      "grad_norm": 0.37988106953348527,
+      "learning_rate": 4.885769198369562e-06,
+      "loss": 0.6073,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6037435284747112,
+      "grad_norm": 0.38697233761955335,
+      "learning_rate": 4.885664846679733e-06,
+      "loss": 0.6403,
+      "step": 2274
+    },
+    {
+      "epoch": 0.604009026948095,
+      "grad_norm": 0.3883099861193893,
+      "learning_rate": 4.885560448463598e-06,
+      "loss": 0.5837,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6042745254214789,
+      "grad_norm": 0.39095583833319403,
+      "learning_rate": 4.885456003723192e-06,
+      "loss": 0.6,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6045400238948626,
+      "grad_norm": 0.38923836797913375,
+      "learning_rate": 4.885351512460554e-06,
+      "loss": 0.6394,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6048055223682464,
+      "grad_norm": 0.392686111275325,
+      "learning_rate": 4.885246974677722e-06,
+      "loss": 0.5906,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6050710208416301,
+      "grad_norm": 0.39871754092907447,
+      "learning_rate": 4.885142390376734e-06,
+      "loss": 0.6354,
+      "step": 2279
+    },
+    {
+      "epoch": 0.605336519315014,
+      "grad_norm": 0.39714503012304886,
+      "learning_rate": 4.885037759559629e-06,
+      "loss": 0.6732,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6056020177883977,
+      "grad_norm": 0.38248809310209814,
+      "learning_rate": 4.884933082228448e-06,
+      "loss": 0.6383,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6058675162617815,
+      "grad_norm": 0.382362656022168,
+      "learning_rate": 4.884828358385232e-06,
+      "loss": 0.6416,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6061330147351652,
+      "grad_norm": 0.40072667467094,
+      "learning_rate": 4.884723588032025e-06,
+      "loss": 0.5973,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6063985132085491,
+      "grad_norm": 0.3930848269379804,
+      "learning_rate": 4.884618771170869e-06,
+      "loss": 0.6469,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6066640116819328,
+      "grad_norm": 0.3866725265930605,
+      "learning_rate": 4.884513907803808e-06,
+      "loss": 0.6217,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6069295101553166,
+      "grad_norm": 0.3952857127342665,
+      "learning_rate": 4.884408997932888e-06,
+      "loss": 0.6326,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6071950086287004,
+      "grad_norm": 0.39260635351597767,
+      "learning_rate": 4.884304041560154e-06,
+      "loss": 0.6259,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6074605071020842,
+      "grad_norm": 0.4073982009483649,
+      "learning_rate": 4.884199038687653e-06,
+      "loss": 0.6402,
+      "step": 2288
+    },
+    {
+      "epoch": 0.607726005575468,
+      "grad_norm": 0.37151038893093485,
+      "learning_rate": 4.8840939893174346e-06,
+      "loss": 0.6137,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6079915040488517,
+      "grad_norm": 0.3764410770156596,
+      "learning_rate": 4.883988893451545e-06,
+      "loss": 0.5987,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6082570025222355,
+      "grad_norm": 0.3988277215546898,
+      "learning_rate": 4.883883751092035e-06,
+      "loss": 0.6097,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6085225009956192,
+      "grad_norm": 0.38839529823626273,
+      "learning_rate": 4.883778562240956e-06,
+      "loss": 0.6456,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6087879994690031,
+      "grad_norm": 0.3885724256697228,
+      "learning_rate": 4.8836733269003574e-06,
+      "loss": 0.6078,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6090534979423868,
+      "grad_norm": 0.4111759276077352,
+      "learning_rate": 4.8835680450722936e-06,
+      "loss": 0.6468,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6093189964157706,
+      "grad_norm": 0.3807975493978897,
+      "learning_rate": 4.883462716758817e-06,
+      "loss": 0.656,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6095844948891543,
+      "grad_norm": 0.3923694304730504,
+      "learning_rate": 4.88335734196198e-06,
+      "loss": 0.6475,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6098499933625382,
+      "grad_norm": 0.37477259929688456,
+      "learning_rate": 4.8832519206838414e-06,
+      "loss": 0.6318,
+      "step": 2297
+    },
+    {
+      "epoch": 0.610115491835922,
+      "grad_norm": 0.3838689728476556,
+      "learning_rate": 4.883146452926455e-06,
+      "loss": 0.6224,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6103809903093057,
+      "grad_norm": 0.3657888431749063,
+      "learning_rate": 4.8830409386918766e-06,
+      "loss": 0.6015,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6106464887826895,
+      "grad_norm": 0.39157567946203004,
+      "learning_rate": 4.882935377982166e-06,
+      "loss": 0.6444,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6109119872560733,
+      "grad_norm": 0.3791873823184957,
+      "learning_rate": 4.882829770799381e-06,
+      "loss": 0.6111,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6111774857294571,
+      "grad_norm": 0.3758602781250536,
+      "learning_rate": 4.88272411714558e-06,
+      "loss": 0.6124,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6114429842028408,
+      "grad_norm": 0.3925975205540793,
+      "learning_rate": 4.882618417022827e-06,
+      "loss": 0.622,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6117084826762246,
+      "grad_norm": 0.3899158034417881,
+      "learning_rate": 4.88251267043318e-06,
+      "loss": 0.6238,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6119739811496084,
+      "grad_norm": 0.38634268746897243,
+      "learning_rate": 4.882406877378703e-06,
+      "loss": 0.6152,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6122394796229922,
+      "grad_norm": 0.3809589119286443,
+      "learning_rate": 4.882301037861459e-06,
+      "loss": 0.5864,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6125049780963759,
+      "grad_norm": 0.3857995001561622,
+      "learning_rate": 4.882195151883511e-06,
+      "loss": 0.5764,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6127704765697597,
+      "grad_norm": 0.3913976214194172,
+      "learning_rate": 4.882089219446925e-06,
+      "loss": 0.6648,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6130359750431436,
+      "grad_norm": 0.38831120435461774,
+      "learning_rate": 4.8819832405537675e-06,
+      "loss": 0.6469,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6133014735165273,
+      "grad_norm": 0.49249891976627497,
+      "learning_rate": 4.881877215206104e-06,
+      "loss": 0.6067,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6135669719899111,
+      "grad_norm": 0.39022262201084135,
+      "learning_rate": 4.881771143406004e-06,
+      "loss": 0.6406,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6138324704632948,
+      "grad_norm": 0.41752508991924614,
+      "learning_rate": 4.881665025155535e-06,
+      "loss": 0.6215,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6140979689366786,
+      "grad_norm": 0.3862454665198474,
+      "learning_rate": 4.881558860456765e-06,
+      "loss": 0.5983,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6143634674100624,
+      "grad_norm": 0.38444100504837136,
+      "learning_rate": 4.881452649311768e-06,
+      "loss": 0.6294,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6146289658834462,
+      "grad_norm": 0.3857585870410081,
+      "learning_rate": 4.881346391722613e-06,
+      "loss": 0.6513,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6148944643568299,
+      "grad_norm": 0.41067777735844124,
+      "learning_rate": 4.881240087691372e-06,
+      "loss": 0.6406,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6151599628302137,
+      "grad_norm": 0.3887479327214694,
+      "learning_rate": 4.88113373722012e-06,
+      "loss": 0.5949,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6154254613035975,
+      "grad_norm": 0.40226059323754443,
+      "learning_rate": 4.88102734031093e-06,
+      "loss": 0.6234,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6156909597769813,
+      "grad_norm": 0.39721682665934654,
+      "learning_rate": 4.880920896965877e-06,
+      "loss": 0.6123,
+      "step": 2319
+    },
+    {
+      "epoch": 0.615956458250365,
+      "grad_norm": 0.39369038462415645,
+      "learning_rate": 4.880814407187037e-06,
+      "loss": 0.6271,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6162219567237488,
+      "grad_norm": 0.3897859329053599,
+      "learning_rate": 4.880707870976486e-06,
+      "loss": 0.6068,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6164874551971327,
+      "grad_norm": 0.42543408190189935,
+      "learning_rate": 4.880601288336304e-06,
+      "loss": 0.6471,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6167529536705164,
+      "grad_norm": 0.39625575285272974,
+      "learning_rate": 4.880494659268567e-06,
+      "loss": 0.6388,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6170184521439002,
+      "grad_norm": 0.371069958328103,
+      "learning_rate": 4.880387983775356e-06,
+      "loss": 0.6013,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6172839506172839,
+      "grad_norm": 0.39029632802901953,
+      "learning_rate": 4.880281261858751e-06,
+      "loss": 0.6347,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6175494490906678,
+      "grad_norm": 0.41170146729381074,
+      "learning_rate": 4.880174493520834e-06,
+      "loss": 0.6275,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6178149475640515,
+      "grad_norm": 0.3864660763060855,
+      "learning_rate": 4.8800676787636865e-06,
+      "loss": 0.658,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6180804460374353,
+      "grad_norm": 0.41771132209064465,
+      "learning_rate": 4.879960817589392e-06,
+      "loss": 0.6564,
+      "step": 2328
+    },
+    {
+      "epoch": 0.618345944510819,
+      "grad_norm": 0.4069991024607502,
+      "learning_rate": 4.879853910000034e-06,
+      "loss": 0.6413,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6186114429842028,
+      "grad_norm": 0.3868430910284126,
+      "learning_rate": 4.879746955997698e-06,
+      "loss": 0.6354,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6188769414575866,
+      "grad_norm": 0.39251733973127556,
+      "learning_rate": 4.87963995558447e-06,
+      "loss": 0.5813,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6191424399309704,
+      "grad_norm": 0.3891066975473416,
+      "learning_rate": 4.879532908762436e-06,
+      "loss": 0.6261,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6194079384043542,
+      "grad_norm": 0.3965593417905505,
+      "learning_rate": 4.8794258155336845e-06,
+      "loss": 0.6196,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6196734368777379,
+      "grad_norm": 0.38956380180103933,
+      "learning_rate": 4.879318675900304e-06,
+      "loss": 0.6374,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6199389353511218,
+      "grad_norm": 0.3928432233142457,
+      "learning_rate": 4.879211489864383e-06,
+      "loss": 0.6188,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6202044338245055,
+      "grad_norm": 0.4422949602136624,
+      "learning_rate": 4.8791042574280135e-06,
+      "loss": 0.588,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6204699322978893,
+      "grad_norm": 0.39925491957786974,
+      "learning_rate": 4.878996978593286e-06,
+      "loss": 0.6214,
+      "step": 2337
+    },
+    {
+      "epoch": 0.620735430771273,
+      "grad_norm": 0.38079109358165275,
+      "learning_rate": 4.878889653362292e-06,
+      "loss": 0.611,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6210009292446569,
+      "grad_norm": 0.46654279735645504,
+      "learning_rate": 4.878782281737125e-06,
+      "loss": 0.621,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6212664277180406,
+      "grad_norm": 0.40210668345158673,
+      "learning_rate": 4.878674863719879e-06,
+      "loss": 0.5821,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6215319261914244,
+      "grad_norm": 0.3770331725948,
+      "learning_rate": 4.87856739931265e-06,
+      "loss": 0.6089,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6217974246648081,
+      "grad_norm": 0.38523267183865273,
+      "learning_rate": 4.878459888517533e-06,
+      "loss": 0.6494,
+      "step": 2342
+    },
+    {
+      "epoch": 0.622062923138192,
+      "grad_norm": 0.391862533871917,
+      "learning_rate": 4.878352331336624e-06,
+      "loss": 0.5861,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6223284216115758,
+      "grad_norm": 0.39001758829827077,
+      "learning_rate": 4.8782447277720225e-06,
+      "loss": 0.6192,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6225939200849595,
+      "grad_norm": 0.39191717983325164,
+      "learning_rate": 4.8781370778258254e-06,
+      "loss": 0.6415,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6228594185583433,
+      "grad_norm": 0.3792087930924814,
+      "learning_rate": 4.878029381500132e-06,
+      "loss": 0.6257,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6231249170317271,
+      "grad_norm": 0.38707098623135416,
+      "learning_rate": 4.877921638797044e-06,
+      "loss": 0.6411,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6233904155051109,
+      "grad_norm": 0.3853638096747392,
+      "learning_rate": 4.877813849718661e-06,
+      "loss": 0.6032,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6236559139784946,
+      "grad_norm": 0.3898622760666724,
+      "learning_rate": 4.877706014267086e-06,
+      "loss": 0.6281,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6239214124518784,
+      "grad_norm": 0.3822552211577024,
+      "learning_rate": 4.877598132444423e-06,
+      "loss": 0.5934,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6241869109252621,
+      "grad_norm": 0.40283323302067503,
+      "learning_rate": 4.877490204252775e-06,
+      "loss": 0.5896,
+      "step": 2351
+    },
+    {
+      "epoch": 0.624452409398646,
+      "grad_norm": 0.3879238698124411,
+      "learning_rate": 4.877382229694247e-06,
+      "loss": 0.6444,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6247179078720297,
+      "grad_norm": 0.39407605100023513,
+      "learning_rate": 4.8772742087709435e-06,
+      "loss": 0.623,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6249834063454135,
+      "grad_norm": 0.41825717516977834,
+      "learning_rate": 4.877166141484973e-06,
+      "loss": 0.6239,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6252489048187972,
+      "grad_norm": 0.39924390668571913,
+      "learning_rate": 4.877058027838443e-06,
+      "loss": 0.5883,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6255144032921811,
+      "grad_norm": 0.38975037913708366,
+      "learning_rate": 4.8769498678334605e-06,
+      "loss": 0.6402,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6257799017655649,
+      "grad_norm": 0.3961743296798766,
+      "learning_rate": 4.876841661472136e-06,
+      "loss": 0.5923,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6260454002389486,
+      "grad_norm": 0.3749225849145829,
+      "learning_rate": 4.8767334087565795e-06,
+      "loss": 0.6059,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6263108987123324,
+      "grad_norm": 0.39068089806455725,
+      "learning_rate": 4.876625109688903e-06,
+      "loss": 0.653,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6265763971857162,
+      "grad_norm": 0.4281047437820375,
+      "learning_rate": 4.8765167642712175e-06,
+      "loss": 0.6203,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6268418956591,
+      "grad_norm": 0.3866583387083701,
+      "learning_rate": 4.876408372505636e-06,
+      "loss": 0.6348,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6271073941324837,
+      "grad_norm": 0.39484408716558617,
+      "learning_rate": 4.876299934394274e-06,
+      "loss": 0.6721,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6273728926058675,
+      "grad_norm": 0.39982151261862975,
+      "learning_rate": 4.876191449939244e-06,
+      "loss": 0.6193,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6276383910792513,
+      "grad_norm": 0.410190825250382,
+      "learning_rate": 4.8760829191426626e-06,
+      "loss": 0.6488,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6279038895526351,
+      "grad_norm": 0.3840322930565453,
+      "learning_rate": 4.875974342006647e-06,
+      "loss": 0.6174,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6281693880260188,
+      "grad_norm": 0.37532539966358164,
+      "learning_rate": 4.875865718533315e-06,
+      "loss": 0.6687,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6284348864994026,
+      "grad_norm": 0.43432916636481106,
+      "learning_rate": 4.875757048724784e-06,
+      "loss": 0.6141,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6287003849727864,
+      "grad_norm": 0.3839813673259421,
+      "learning_rate": 4.875648332583174e-06,
+      "loss": 0.6498,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6289658834461702,
+      "grad_norm": 0.37749868778090734,
+      "learning_rate": 4.875539570110604e-06,
+      "loss": 0.5948,
+      "step": 2369
+    },
+    {
+      "epoch": 0.629231381919554,
+      "grad_norm": 0.3790653920541351,
+      "learning_rate": 4.875430761309197e-06,
+      "loss": 0.6334,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6294968803929377,
+      "grad_norm": 0.3914305185793328,
+      "learning_rate": 4.875321906181073e-06,
+      "loss": 0.6317,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6297623788663215,
+      "grad_norm": 0.38487283781364384,
+      "learning_rate": 4.8752130047283565e-06,
+      "loss": 0.6152,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6300278773397053,
+      "grad_norm": 0.3890328379433162,
+      "learning_rate": 4.875104056953172e-06,
+      "loss": 0.6278,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6302933758130891,
+      "grad_norm": 0.3780682485002683,
+      "learning_rate": 4.874995062857641e-06,
+      "loss": 0.6052,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6305588742864728,
+      "grad_norm": 0.3964397805161175,
+      "learning_rate": 4.874886022443893e-06,
+      "loss": 0.6295,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6308243727598566,
+      "grad_norm": 0.400707792006883,
+      "learning_rate": 4.8747769357140515e-06,
+      "loss": 0.6554,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6310898712332405,
+      "grad_norm": 0.3745922323646813,
+      "learning_rate": 4.874667802670246e-06,
+      "loss": 0.5956,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6313553697066242,
+      "grad_norm": 0.3802428317650885,
+      "learning_rate": 4.874558623314604e-06,
+      "loss": 0.6002,
+      "step": 2378
+    },
+    {
+      "epoch": 0.631620868180008,
+      "grad_norm": 0.4018762375198942,
+      "learning_rate": 4.874449397649256e-06,
+      "loss": 0.6311,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6318863666533917,
+      "grad_norm": 0.38163789447817625,
+      "learning_rate": 4.87434012567633e-06,
+      "loss": 0.6713,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6321518651267756,
+      "grad_norm": 0.37485428652108665,
+      "learning_rate": 4.8742308073979584e-06,
+      "loss": 0.5963,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6324173636001593,
+      "grad_norm": 0.4083955630109161,
+      "learning_rate": 4.874121442816273e-06,
+      "loss": 0.6268,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6326828620735431,
+      "grad_norm": 0.4070481869629904,
+      "learning_rate": 4.874012031933407e-06,
+      "loss": 0.6142,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6329483605469268,
+      "grad_norm": 0.3918489236527451,
+      "learning_rate": 4.873902574751493e-06,
+      "loss": 0.645,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6332138590203107,
+      "grad_norm": 0.3857532718122673,
+      "learning_rate": 4.8737930712726665e-06,
+      "loss": 0.6117,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6334793574936944,
+      "grad_norm": 0.400455466140325,
+      "learning_rate": 4.873683521499064e-06,
+      "loss": 0.6635,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6337448559670782,
+      "grad_norm": 0.39315283241470816,
+      "learning_rate": 4.87357392543282e-06,
+      "loss": 0.6267,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6340103544404619,
+      "grad_norm": 0.375461837976201,
+      "learning_rate": 4.873464283076074e-06,
+      "loss": 0.5915,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6342758529138457,
+      "grad_norm": 0.382130824836503,
+      "learning_rate": 4.873354594430963e-06,
+      "loss": 0.6489,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6345413513872296,
+      "grad_norm": 0.38855594608669636,
+      "learning_rate": 4.873244859499625e-06,
+      "loss": 0.6349,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6348068498606133,
+      "grad_norm": 0.37666399182589605,
+      "learning_rate": 4.873135078284203e-06,
+      "loss": 0.6137,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6350723483339971,
+      "grad_norm": 0.38143451438472825,
+      "learning_rate": 4.873025250786836e-06,
+      "loss": 0.6406,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6353378468073808,
+      "grad_norm": 0.37615911290761034,
+      "learning_rate": 4.872915377009666e-06,
+      "loss": 0.6218,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6356033452807647,
+      "grad_norm": 0.3849287713753922,
+      "learning_rate": 4.872805456954837e-06,
+      "loss": 0.619,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6358688437541484,
+      "grad_norm": 0.3945313230420764,
+      "learning_rate": 4.8726954906244915e-06,
+      "loss": 0.6451,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6361343422275322,
+      "grad_norm": 0.4192730294881194,
+      "learning_rate": 4.872585478020776e-06,
+      "loss": 0.6402,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6363998407009159,
+      "grad_norm": 0.3879844510767756,
+      "learning_rate": 4.872475419145832e-06,
+      "loss": 0.6363,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6366653391742998,
+      "grad_norm": 0.3771814158584119,
+      "learning_rate": 4.87236531400181e-06,
+      "loss": 0.6697,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6369308376476835,
+      "grad_norm": 0.3794289928880538,
+      "learning_rate": 4.872255162590857e-06,
+      "loss": 0.6053,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6371963361210673,
+      "grad_norm": 0.3914351725253561,
+      "learning_rate": 4.872144964915118e-06,
+      "loss": 0.6366,
+      "step": 2400
+    },
+    {
+      "epoch": 0.637461834594451,
+      "grad_norm": 0.3903557114148793,
+      "learning_rate": 4.872034720976745e-06,
+      "loss": 0.62,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6377273330678349,
+      "grad_norm": 0.3947328295929092,
+      "learning_rate": 4.871924430777888e-06,
+      "loss": 0.6098,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6379928315412187,
+      "grad_norm": 0.3850221782244195,
+      "learning_rate": 4.871814094320696e-06,
+      "loss": 0.6343,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6382583300146024,
+      "grad_norm": 0.4028496713885383,
+      "learning_rate": 4.871703711607322e-06,
+      "loss": 0.664,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6385238284879862,
+      "grad_norm": 0.36735043987302124,
+      "learning_rate": 4.871593282639919e-06,
+      "loss": 0.6447,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6387893269613699,
+      "grad_norm": 0.3851391222934083,
+      "learning_rate": 4.87148280742064e-06,
+      "loss": 0.5908,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6390548254347538,
+      "grad_norm": 0.3893816450710175,
+      "learning_rate": 4.87137228595164e-06,
+      "loss": 0.6276,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6393203239081375,
+      "grad_norm": 0.39413184248496463,
+      "learning_rate": 4.871261718235075e-06,
+      "loss": 0.6278,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6395858223815213,
+      "grad_norm": 0.4365868367662229,
+      "learning_rate": 4.8711511042731e-06,
+      "loss": 0.6267,
+      "step": 2409
+    },
+    {
+      "epoch": 0.639851320854905,
+      "grad_norm": 0.3810323792503826,
+      "learning_rate": 4.871040444067873e-06,
+      "loss": 0.6209,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6401168193282889,
+      "grad_norm": 0.39784333252611237,
+      "learning_rate": 4.870929737621551e-06,
+      "loss": 0.6409,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6403823178016727,
+      "grad_norm": 0.38942566771250076,
+      "learning_rate": 4.870818984936296e-06,
+      "loss": 0.6378,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6406478162750564,
+      "grad_norm": 0.4087388819997767,
+      "learning_rate": 4.8707081860142645e-06,
+      "loss": 0.6171,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6409133147484402,
+      "grad_norm": 0.36240731231871115,
+      "learning_rate": 4.8705973408576205e-06,
+      "loss": 0.6004,
+      "step": 2414
+    },
+    {
+      "epoch": 0.641178813221824,
+      "grad_norm": 0.39722425919833493,
+      "learning_rate": 4.870486449468522e-06,
+      "loss": 0.6066,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6414443116952078,
+      "grad_norm": 0.4163633823795304,
+      "learning_rate": 4.870375511849136e-06,
+      "loss": 0.6402,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6417098101685915,
+      "grad_norm": 0.36712661710311617,
+      "learning_rate": 4.870264528001623e-06,
+      "loss": 0.6108,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6419753086419753,
+      "grad_norm": 0.397016529454373,
+      "learning_rate": 4.870153497928147e-06,
+      "loss": 0.6484,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6422408071153591,
+      "grad_norm": 0.38553671497935776,
+      "learning_rate": 4.8700424216308765e-06,
+      "loss": 0.5517,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6425063055887429,
+      "grad_norm": 0.40996167485293356,
+      "learning_rate": 4.869931299111974e-06,
+      "loss": 0.6491,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6427718040621266,
+      "grad_norm": 0.38722113113089673,
+      "learning_rate": 4.8698201303736105e-06,
+      "loss": 0.6395,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6430373025355104,
+      "grad_norm": 0.4000129523578866,
+      "learning_rate": 4.869708915417951e-06,
+      "loss": 0.6179,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6433028010088943,
+      "grad_norm": 0.40020045379966307,
+      "learning_rate": 4.869597654247166e-06,
+      "loss": 0.6324,
+      "step": 2423
+    },
+    {
+      "epoch": 0.643568299482278,
+      "grad_norm": 0.3850289895582323,
+      "learning_rate": 4.8694863468634255e-06,
+      "loss": 0.6296,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6438337979556618,
+      "grad_norm": 0.3797961163216567,
+      "learning_rate": 4.869374993268898e-06,
+      "loss": 0.6131,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6440992964290455,
+      "grad_norm": 0.3965706741086633,
+      "learning_rate": 4.8692635934657585e-06,
+      "loss": 0.6176,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6443647949024293,
+      "grad_norm": 0.4042120244882442,
+      "learning_rate": 4.869152147456177e-06,
+      "loss": 0.6599,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6446302933758131,
+      "grad_norm": 0.381712068532955,
+      "learning_rate": 4.8690406552423294e-06,
+      "loss": 0.6246,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6448957918491969,
+      "grad_norm": 0.3758611799165009,
+      "learning_rate": 4.868929116826388e-06,
+      "loss": 0.5742,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.39409171170112794,
+      "learning_rate": 4.868817532210528e-06,
+      "loss": 0.631,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6454267887959644,
+      "grad_norm": 0.37446255195683514,
+      "learning_rate": 4.868705901396926e-06,
+      "loss": 0.5964,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6456922872693482,
+      "grad_norm": 0.38023788410347037,
+      "learning_rate": 4.86859422438776e-06,
+      "loss": 0.6272,
+      "step": 2432
+    },
+    {
+      "epoch": 0.645957785742732,
+      "grad_norm": 0.3706862445659322,
+      "learning_rate": 4.868482501185208e-06,
+      "loss": 0.622,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6462232842161157,
+      "grad_norm": 0.39954781948419105,
+      "learning_rate": 4.868370731791447e-06,
+      "loss": 0.6237,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6464887826894995,
+      "grad_norm": 0.3917522697853562,
+      "learning_rate": 4.868258916208658e-06,
+      "loss": 0.6468,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6467542811628834,
+      "grad_norm": 0.3816753365170755,
+      "learning_rate": 4.868147054439022e-06,
+      "loss": 0.633,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6470197796362671,
+      "grad_norm": 0.3877572153449095,
+      "learning_rate": 4.8680351464847206e-06,
+      "loss": 0.6265,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6472852781096509,
+      "grad_norm": 0.3753767609207334,
+      "learning_rate": 4.867923192347936e-06,
+      "loss": 0.5868,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6475507765830346,
+      "grad_norm": 0.40940122388600325,
+      "learning_rate": 4.86781119203085e-06,
+      "loss": 0.6212,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6478162750564185,
+      "grad_norm": 0.38740799178512,
+      "learning_rate": 4.86769914553565e-06,
+      "loss": 0.6205,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6480817735298022,
+      "grad_norm": 0.4109288749924628,
+      "learning_rate": 4.867587052864519e-06,
+      "loss": 0.6439,
+      "step": 2441
+    },
+    {
+      "epoch": 0.648347272003186,
+      "grad_norm": 0.3935390604153536,
+      "learning_rate": 4.867474914019643e-06,
+      "loss": 0.6322,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6486127704765697,
+      "grad_norm": 0.3936567995133444,
+      "learning_rate": 4.86736272900321e-06,
+      "loss": 0.65,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6488782689499535,
+      "grad_norm": 0.38473693777235074,
+      "learning_rate": 4.867250497817407e-06,
+      "loss": 0.6052,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6491437674233373,
+      "grad_norm": 0.40568914204433476,
+      "learning_rate": 4.8671382204644245e-06,
+      "loss": 0.6185,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6494092658967211,
+      "grad_norm": 0.42111090788991906,
+      "learning_rate": 4.86702589694645e-06,
+      "loss": 0.5368,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6496747643701049,
+      "grad_norm": 0.4185580111244889,
+      "learning_rate": 4.866913527265676e-06,
+      "loss": 0.6461,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6499402628434886,
+      "grad_norm": 0.42930425429914,
+      "learning_rate": 4.8668011114242916e-06,
+      "loss": 0.621,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6502057613168725,
+      "grad_norm": 0.4021288392418631,
+      "learning_rate": 4.8666886494244916e-06,
+      "loss": 0.6738,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6504712597902562,
+      "grad_norm": 0.39124077599819274,
+      "learning_rate": 4.866576141268468e-06,
+      "loss": 0.6422,
+      "step": 2450
+    },
+    {
+      "epoch": 0.65073675826364,
+      "grad_norm": 0.40667606076307977,
+      "learning_rate": 4.866463586958415e-06,
+      "loss": 0.6491,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6510022567370237,
+      "grad_norm": 0.38649628067677233,
+      "learning_rate": 4.8663509864965286e-06,
+      "loss": 0.6452,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6512677552104076,
+      "grad_norm": 0.4401364031459709,
+      "learning_rate": 4.866238339885004e-06,
+      "loss": 0.6321,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6515332536837913,
+      "grad_norm": 0.38044106896229735,
+      "learning_rate": 4.866125647126038e-06,
+      "loss": 0.6282,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6517987521571751,
+      "grad_norm": 0.48513941071869,
+      "learning_rate": 4.866012908221829e-06,
+      "loss": 0.5852,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6520642506305588,
+      "grad_norm": 0.4096964037537831,
+      "learning_rate": 4.865900123174574e-06,
+      "loss": 0.6311,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6523297491039427,
+      "grad_norm": 0.4013675193995615,
+      "learning_rate": 4.8657872919864756e-06,
+      "loss": 0.6262,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6525952475773265,
+      "grad_norm": 0.40322131157403657,
+      "learning_rate": 4.8656744146597315e-06,
+      "loss": 0.617,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6528607460507102,
+      "grad_norm": 0.4021442945770837,
+      "learning_rate": 4.8655614911965445e-06,
+      "loss": 0.5883,
+      "step": 2459
+    },
+    {
+      "epoch": 0.653126244524094,
+      "grad_norm": 0.4001518030032774,
+      "learning_rate": 4.865448521599116e-06,
+      "loss": 0.64,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6533917429974777,
+      "grad_norm": 0.392434863142415,
+      "learning_rate": 4.86533550586965e-06,
+      "loss": 0.6122,
+      "step": 2461
+    },
+    {
+      "epoch": 0.6536572414708616,
+      "grad_norm": 0.3944349165766381,
+      "learning_rate": 4.86522244401035e-06,
+      "loss": 0.629,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6539227399442453,
+      "grad_norm": 0.3884606087919453,
+      "learning_rate": 4.865109336023422e-06,
+      "loss": 0.538,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6541882384176291,
+      "grad_norm": 0.402793762278969,
+      "learning_rate": 4.864996181911072e-06,
+      "loss": 0.6368,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6544537368910128,
+      "grad_norm": 0.3866759782295091,
+      "learning_rate": 4.8648829816755035e-06,
+      "loss": 0.6094,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6547192353643967,
+      "grad_norm": 0.38169281248894205,
+      "learning_rate": 4.864769735318929e-06,
+      "loss": 0.6391,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6549847338377804,
+      "grad_norm": 0.3824012639897559,
+      "learning_rate": 4.864656442843553e-06,
+      "loss": 0.5972,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6552502323111642,
+      "grad_norm": 0.3939800833629133,
+      "learning_rate": 4.864543104251587e-06,
+      "loss": 0.6094,
+      "step": 2468
+    },
+    {
+      "epoch": 0.655515730784548,
+      "grad_norm": 0.40144669559925655,
+      "learning_rate": 4.864429719545241e-06,
+      "loss": 0.6595,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6557812292579318,
+      "grad_norm": 0.3853803866206681,
+      "learning_rate": 4.864316288726727e-06,
+      "loss": 0.6412,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6560467277313156,
+      "grad_norm": 0.3883359918727682,
+      "learning_rate": 4.864202811798257e-06,
+      "loss": 0.639,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6563122262046993,
+      "grad_norm": 0.3761311457924535,
+      "learning_rate": 4.864089288762042e-06,
+      "loss": 0.5927,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6565777246780831,
+      "grad_norm": 0.3781772054186187,
+      "learning_rate": 4.863975719620298e-06,
+      "loss": 0.6301,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6568432231514669,
+      "grad_norm": 0.39547772531661374,
+      "learning_rate": 4.8638621043752406e-06,
+      "loss": 0.6462,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6571087216248507,
+      "grad_norm": 0.3772884194350595,
+      "learning_rate": 4.863748443029084e-06,
+      "loss": 0.6164,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6573742200982344,
+      "grad_norm": 0.3939065243459478,
+      "learning_rate": 4.863634735584045e-06,
+      "loss": 0.6083,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6576397185716182,
+      "grad_norm": 0.3892537674893566,
+      "learning_rate": 4.863520982042341e-06,
+      "loss": 0.6283,
+      "step": 2477
+    },
+    {
+      "epoch": 0.657905217045002,
+      "grad_norm": 0.37374868197468986,
+      "learning_rate": 4.863407182406192e-06,
+      "loss": 0.6056,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6581707155183858,
+      "grad_norm": 0.39387086314838793,
+      "learning_rate": 4.863293336677815e-06,
+      "loss": 0.6126,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6584362139917695,
+      "grad_norm": 0.4092623613598788,
+      "learning_rate": 4.8631794448594325e-06,
+      "loss": 0.6825,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6587017124651533,
+      "grad_norm": 0.378173667013933,
+      "learning_rate": 4.863065506953265e-06,
+      "loss": 0.6548,
+      "step": 2481
+    },
+    {
+      "epoch": 0.658967210938537,
+      "grad_norm": 0.37910481845293986,
+      "learning_rate": 4.862951522961534e-06,
+      "loss": 0.6229,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6592327094119209,
+      "grad_norm": 0.3849573192672843,
+      "learning_rate": 4.862837492886463e-06,
+      "loss": 0.6437,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6594982078853047,
+      "grad_norm": 0.38643860083625237,
+      "learning_rate": 4.862723416730275e-06,
+      "loss": 0.6016,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6597637063586884,
+      "grad_norm": 0.38429315220821314,
+      "learning_rate": 4.8626092944951965e-06,
+      "loss": 0.6487,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6600292048320722,
+      "grad_norm": 0.39503680112549455,
+      "learning_rate": 4.862495126183451e-06,
+      "loss": 0.6416,
+      "step": 2486
+    },
+    {
+      "epoch": 0.660294703305456,
+      "grad_norm": 0.3689026852007654,
+      "learning_rate": 4.8623809117972665e-06,
+      "loss": 0.619,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6605602017788398,
+      "grad_norm": 0.3866011676925736,
+      "learning_rate": 4.86226665133887e-06,
+      "loss": 0.6535,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6608257002522235,
+      "grad_norm": 0.38050250486608783,
+      "learning_rate": 4.862152344810491e-06,
+      "loss": 0.6343,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6610911987256073,
+      "grad_norm": 0.38632021616597995,
+      "learning_rate": 4.862037992214357e-06,
+      "loss": 0.6215,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6613566971989912,
+      "grad_norm": 0.38711932869121424,
+      "learning_rate": 4.861923593552699e-06,
+      "loss": 0.6334,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6616221956723749,
+      "grad_norm": 0.37789754896138056,
+      "learning_rate": 4.861809148827747e-06,
+      "loss": 0.6601,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6618876941457587,
+      "grad_norm": 0.38116298789751557,
+      "learning_rate": 4.8616946580417344e-06,
+      "loss": 0.6237,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6621531926191424,
+      "grad_norm": 0.38789035611558487,
+      "learning_rate": 4.861580121196894e-06,
+      "loss": 0.6037,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6624186910925263,
+      "grad_norm": 0.38684696372789545,
+      "learning_rate": 4.861465538295459e-06,
+      "loss": 0.6267,
+      "step": 2495
+    },
+    {
+      "epoch": 0.66268418956591,
+      "grad_norm": 0.38752623864801644,
+      "learning_rate": 4.861350909339664e-06,
+      "loss": 0.6366,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6629496880392938,
+      "grad_norm": 0.4059626560122765,
+      "learning_rate": 4.861236234331745e-06,
+      "loss": 0.6642,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6632151865126775,
+      "grad_norm": 0.3871700790127218,
+      "learning_rate": 4.861121513273938e-06,
+      "loss": 0.6233,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6634806849860613,
+      "grad_norm": 0.39138099142193133,
+      "learning_rate": 4.861006746168479e-06,
+      "loss": 0.5794,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6637461834594451,
+      "grad_norm": 0.38879068707973874,
+      "learning_rate": 4.86089193301761e-06,
+      "loss": 0.6356,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6640116819328289,
+      "grad_norm": 0.37596873234278433,
+      "learning_rate": 4.860777073823567e-06,
+      "loss": 0.6215,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6642771804062126,
+      "grad_norm": 0.383533070368256,
+      "learning_rate": 4.86066216858859e-06,
+      "loss": 0.6416,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6645426788795964,
+      "grad_norm": 0.3982564508673109,
+      "learning_rate": 4.860547217314922e-06,
+      "loss": 0.6029,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6648081773529803,
+      "grad_norm": 0.378744352411347,
+      "learning_rate": 4.860432220004803e-06,
+      "loss": 0.606,
+      "step": 2504
+    },
+    {
+      "epoch": 0.665073675826364,
+      "grad_norm": 0.3732731941170594,
+      "learning_rate": 4.860317176660477e-06,
+      "loss": 0.6209,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6653391742997478,
+      "grad_norm": 0.38207672202431,
+      "learning_rate": 4.8602020872841865e-06,
+      "loss": 0.6204,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6656046727731315,
+      "grad_norm": 0.3994957365386139,
+      "learning_rate": 4.860086951878177e-06,
+      "loss": 0.6453,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6658701712465154,
+      "grad_norm": 0.3856215522084647,
+      "learning_rate": 4.859971770444692e-06,
+      "loss": 0.6513,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6661356697198991,
+      "grad_norm": 0.41097373652956726,
+      "learning_rate": 4.859856542985981e-06,
+      "loss": 0.5832,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6664011681932829,
+      "grad_norm": 0.38595628458347075,
+      "learning_rate": 4.859741269504289e-06,
+      "loss": 0.5778,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.3992401585977327,
+      "learning_rate": 4.859625950001864e-06,
+      "loss": 0.619,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6669321651400505,
+      "grad_norm": 0.39287580036457476,
+      "learning_rate": 4.859510584480956e-06,
+      "loss": 0.6259,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6671976636134342,
+      "grad_norm": 0.38021736579129695,
+      "learning_rate": 4.859395172943815e-06,
+      "loss": 0.6263,
+      "step": 2513
+    },
+    {
+      "epoch": 0.667463162086818,
+      "grad_norm": 0.3838993766133399,
+      "learning_rate": 4.859279715392691e-06,
+      "loss": 0.6107,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6677286605602017,
+      "grad_norm": 0.38181431090714285,
+      "learning_rate": 4.8591642118298355e-06,
+      "loss": 0.6186,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6679941590335856,
+      "grad_norm": 0.37836799081787303,
+      "learning_rate": 4.8590486622575015e-06,
+      "loss": 0.6305,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6682596575069694,
+      "grad_norm": 0.38835025871021733,
+      "learning_rate": 4.858933066677944e-06,
+      "loss": 0.6106,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6685251559803531,
+      "grad_norm": 0.3775874377775798,
+      "learning_rate": 4.858817425093415e-06,
+      "loss": 0.6459,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6687906544537369,
+      "grad_norm": 0.39040841830586814,
+      "learning_rate": 4.85870173750617e-06,
+      "loss": 0.6496,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6690561529271206,
+      "grad_norm": 0.38147813795752034,
+      "learning_rate": 4.858586003918469e-06,
+      "loss": 0.6072,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6693216514005045,
+      "grad_norm": 0.3865106030914664,
+      "learning_rate": 4.8584702243325635e-06,
+      "loss": 0.6387,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6695871498738882,
+      "grad_norm": 0.36842893630068424,
+      "learning_rate": 4.858354398750715e-06,
+      "loss": 0.6074,
+      "step": 2522
+    },
+    {
+      "epoch": 0.669852648347272,
+      "grad_norm": 0.3822321973649649,
+      "learning_rate": 4.858238527175181e-06,
+      "loss": 0.6423,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6701181468206557,
+      "grad_norm": 0.3735881108709939,
+      "learning_rate": 4.858122609608222e-06,
+      "loss": 0.6357,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6703836452940396,
+      "grad_norm": 0.3908279670644259,
+      "learning_rate": 4.858006646052098e-06,
+      "loss": 0.6357,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6706491437674234,
+      "grad_norm": 0.38800731462938065,
+      "learning_rate": 4.857890636509072e-06,
+      "loss": 0.6179,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6709146422408071,
+      "grad_norm": 0.38994344794696584,
+      "learning_rate": 4.8577745809814046e-06,
+      "loss": 0.6182,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6711801407141909,
+      "grad_norm": 0.374108861793892,
+      "learning_rate": 4.85765847947136e-06,
+      "loss": 0.5971,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6714456391875747,
+      "grad_norm": 0.38951660976775165,
+      "learning_rate": 4.857542331981203e-06,
+      "loss": 0.6474,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6717111376609585,
+      "grad_norm": 0.3873093174653202,
+      "learning_rate": 4.857426138513198e-06,
+      "loss": 0.5944,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6719766361343422,
+      "grad_norm": 0.47293909425983943,
+      "learning_rate": 4.857309899069612e-06,
+      "loss": 0.5405,
+      "step": 2531
+    },
+    {
+      "epoch": 0.672242134607726,
+      "grad_norm": 0.416257384381464,
+      "learning_rate": 4.857193613652711e-06,
+      "loss": 0.6393,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6725076330811098,
+      "grad_norm": 0.40259090507413703,
+      "learning_rate": 4.857077282264763e-06,
+      "loss": 0.628,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6727731315544936,
+      "grad_norm": 0.39750122403523114,
+      "learning_rate": 4.856960904908037e-06,
+      "loss": 0.6022,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6730386300278773,
+      "grad_norm": 0.38010415065804315,
+      "learning_rate": 4.856844481584801e-06,
+      "loss": 0.6291,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6733041285012611,
+      "grad_norm": 0.38542806440509086,
+      "learning_rate": 4.856728012297329e-06,
+      "loss": 0.6237,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6735696269746448,
+      "grad_norm": 0.3873558433018967,
+      "learning_rate": 4.85661149704789e-06,
+      "loss": 0.6305,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6738351254480287,
+      "grad_norm": 0.47909573788052434,
+      "learning_rate": 4.856494935838757e-06,
+      "loss": 0.6323,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6741006239214125,
+      "grad_norm": 0.39494063248857897,
+      "learning_rate": 4.856378328672202e-06,
+      "loss": 0.579,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6743661223947962,
+      "grad_norm": 0.38522195779640217,
+      "learning_rate": 4.856261675550501e-06,
+      "loss": 0.6539,
+      "step": 2540
+    },
+    {
+      "epoch": 0.67463162086818,
+      "grad_norm": 0.398179815084743,
+      "learning_rate": 4.856144976475928e-06,
+      "loss": 0.6416,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6748971193415638,
+      "grad_norm": 0.3836072796857305,
+      "learning_rate": 4.85602823145076e-06,
+      "loss": 0.6114,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6751626178149476,
+      "grad_norm": 0.385671427520204,
+      "learning_rate": 4.855911440477272e-06,
+      "loss": 0.6208,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6754281162883313,
+      "grad_norm": 0.3879075507199383,
+      "learning_rate": 4.8557946035577426e-06,
+      "loss": 0.6268,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6756936147617151,
+      "grad_norm": 0.3852121785704072,
+      "learning_rate": 4.8556777206944506e-06,
+      "loss": 0.624,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6759591132350989,
+      "grad_norm": 0.3740846958512857,
+      "learning_rate": 4.8555607918896755e-06,
+      "loss": 0.6048,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6762246117084827,
+      "grad_norm": 0.37518597880135274,
+      "learning_rate": 4.8554438171456974e-06,
+      "loss": 0.5879,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6764901101818664,
+      "grad_norm": 0.38795221060220025,
+      "learning_rate": 4.855326796464798e-06,
+      "loss": 0.6211,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6767556086552502,
+      "grad_norm": 0.39484116607055514,
+      "learning_rate": 4.855209729849259e-06,
+      "loss": 0.6239,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6770211071286341,
+      "grad_norm": 0.3926309328666225,
+      "learning_rate": 4.8550926173013634e-06,
+      "loss": 0.6356,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6772866056020178,
+      "grad_norm": 0.3880984084489336,
+      "learning_rate": 4.854975458823396e-06,
+      "loss": 0.6169,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6775521040754016,
+      "grad_norm": 0.3833201663137394,
+      "learning_rate": 4.854858254417642e-06,
+      "loss": 0.6241,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6778176025487853,
+      "grad_norm": 0.38820548935411137,
+      "learning_rate": 4.854741004086384e-06,
+      "loss": 0.62,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6780831010221692,
+      "grad_norm": 0.38958891269785556,
+      "learning_rate": 4.854623707831912e-06,
+      "loss": 0.6407,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6783485994955529,
+      "grad_norm": 0.3872380183920155,
+      "learning_rate": 4.854506365656514e-06,
+      "loss": 0.6505,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6786140979689367,
+      "grad_norm": 0.3976817989003109,
+      "learning_rate": 4.854388977562476e-06,
+      "loss": 0.5915,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6788795964423204,
+      "grad_norm": 0.38741463647113006,
+      "learning_rate": 4.8542715435520885e-06,
+      "loss": 0.6612,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6791450949157042,
+      "grad_norm": 0.38786176003940787,
+      "learning_rate": 4.854154063627642e-06,
+      "loss": 0.6679,
+      "step": 2558
+    },
+    {
+      "epoch": 0.679410593389088,
+      "grad_norm": 0.3831634590879999,
+      "learning_rate": 4.854036537791426e-06,
+      "loss": 0.5804,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6796760918624718,
+      "grad_norm": 0.4029791035561383,
+      "learning_rate": 4.853918966045735e-06,
+      "loss": 0.6045,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6799415903358556,
+      "grad_norm": 0.3939396834785739,
+      "learning_rate": 4.85380134839286e-06,
+      "loss": 0.6094,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6802070888092393,
+      "grad_norm": 0.3883102941178923,
+      "learning_rate": 4.853683684835097e-06,
+      "loss": 0.5842,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6804725872826232,
+      "grad_norm": 0.39129053138260844,
+      "learning_rate": 4.853565975374738e-06,
+      "loss": 0.6411,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6807380857560069,
+      "grad_norm": 0.374001069584491,
+      "learning_rate": 4.85344822001408e-06,
+      "loss": 0.6024,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6810035842293907,
+      "grad_norm": 0.37645585029846396,
+      "learning_rate": 4.85333041875542e-06,
+      "loss": 0.619,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6812690827027744,
+      "grad_norm": 0.3778577425661308,
+      "learning_rate": 4.853212571601055e-06,
+      "loss": 0.6425,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6815345811761583,
+      "grad_norm": 0.391038252382859,
+      "learning_rate": 4.8530946785532825e-06,
+      "loss": 0.6227,
+      "step": 2567
+    },
+    {
+      "epoch": 0.681800079649542,
+      "grad_norm": 0.3763444120502296,
+      "learning_rate": 4.852976739614403e-06,
+      "loss": 0.6154,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6820655781229258,
+      "grad_norm": 0.3894312190850422,
+      "learning_rate": 4.852858754786716e-06,
+      "loss": 0.616,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6823310765963095,
+      "grad_norm": 0.37619742731414907,
+      "learning_rate": 4.852740724072523e-06,
+      "loss": 0.6362,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6825965750696934,
+      "grad_norm": 0.3811860358247419,
+      "learning_rate": 4.852622647474125e-06,
+      "loss": 0.6251,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6828620735430772,
+      "grad_norm": 0.3827484116118257,
+      "learning_rate": 4.852504524993824e-06,
+      "loss": 0.6624,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6831275720164609,
+      "grad_norm": 0.3909718121010186,
+      "learning_rate": 4.852386356633926e-06,
+      "loss": 0.5929,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6833930704898447,
+      "grad_norm": 0.3914324040933539,
+      "learning_rate": 4.852268142396734e-06,
+      "loss": 0.6272,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6836585689632284,
+      "grad_norm": 0.3852289015281856,
+      "learning_rate": 4.852149882284554e-06,
+      "loss": 0.6328,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6839240674366123,
+      "grad_norm": 0.38668677917707833,
+      "learning_rate": 4.852031576299693e-06,
+      "loss": 0.6178,
+      "step": 2576
+    },
+    {
+      "epoch": 0.684189565909996,
+      "grad_norm": 0.3954760639322168,
+      "learning_rate": 4.851913224444456e-06,
+      "loss": 0.658,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6844550643833798,
+      "grad_norm": 0.4294103355890287,
+      "learning_rate": 4.851794826721153e-06,
+      "loss": 0.6365,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6847205628567635,
+      "grad_norm": 0.3963387839243065,
+      "learning_rate": 4.851676383132093e-06,
+      "loss": 0.6214,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6849860613301474,
+      "grad_norm": 0.4015125487870929,
+      "learning_rate": 4.851557893679586e-06,
+      "loss": 0.6058,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6852515598035311,
+      "grad_norm": 0.45015400148145274,
+      "learning_rate": 4.8514393583659425e-06,
+      "loss": 0.5626,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6855170582769149,
+      "grad_norm": 0.3991252721447173,
+      "learning_rate": 4.851320777193474e-06,
+      "loss": 0.592,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6857825567502986,
+      "grad_norm": 0.3933441331229182,
+      "learning_rate": 4.851202150164493e-06,
+      "loss": 0.6199,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6860480552236825,
+      "grad_norm": 0.44163082552696814,
+      "learning_rate": 4.851083477281313e-06,
+      "loss": 0.6181,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6863135536970663,
+      "grad_norm": 0.41899120795505423,
+      "learning_rate": 4.85096475854625e-06,
+      "loss": 0.592,
+      "step": 2585
+    },
+    {
+      "epoch": 0.68657905217045,
+      "grad_norm": 0.4186786494049867,
+      "learning_rate": 4.850845993961617e-06,
+      "loss": 0.6148,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6868445506438338,
+      "grad_norm": 0.5655822471740156,
+      "learning_rate": 4.850727183529732e-06,
+      "loss": 0.5874,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6871100491172176,
+      "grad_norm": 0.3981542733599476,
+      "learning_rate": 4.8506083272529105e-06,
+      "loss": 0.6319,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6873755475906014,
+      "grad_norm": 0.4261072246034151,
+      "learning_rate": 4.8504894251334725e-06,
+      "loss": 0.582,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6876410460639851,
+      "grad_norm": 0.4008182960922975,
+      "learning_rate": 4.850370477173735e-06,
+      "loss": 0.6304,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6879065445373689,
+      "grad_norm": 0.3826105897046771,
+      "learning_rate": 4.8502514833760185e-06,
+      "loss": 0.638,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6881720430107527,
+      "grad_norm": 0.3908007000737641,
+      "learning_rate": 4.850132443742643e-06,
+      "loss": 0.6277,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6884375414841365,
+      "grad_norm": 0.38731619213074264,
+      "learning_rate": 4.850013358275931e-06,
+      "loss": 0.5537,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6887030399575202,
+      "grad_norm": 0.41059076033361336,
+      "learning_rate": 4.849894226978205e-06,
+      "loss": 0.6136,
+      "step": 2594
+    },
+    {
+      "epoch": 0.688968538430904,
+      "grad_norm": 0.3923793363302241,
+      "learning_rate": 4.849775049851788e-06,
+      "loss": 0.6258,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6892340369042878,
+      "grad_norm": 0.4143458298244493,
+      "learning_rate": 4.849655826899005e-06,
+      "loss": 0.6242,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6894995353776716,
+      "grad_norm": 0.38346814001939306,
+      "learning_rate": 4.84953655812218e-06,
+      "loss": 0.6289,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6897650338510554,
+      "grad_norm": 0.3977828338130154,
+      "learning_rate": 4.849417243523639e-06,
+      "loss": 0.61,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6900305323244391,
+      "grad_norm": 0.38353593247447937,
+      "learning_rate": 4.84929788310571e-06,
+      "loss": 0.6037,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6902960307978229,
+      "grad_norm": 0.40118915263536675,
+      "learning_rate": 4.84917847687072e-06,
+      "loss": 0.6471,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6905615292712067,
+      "grad_norm": 0.4059740296864397,
+      "learning_rate": 4.849059024820998e-06,
+      "loss": 0.6632,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6908270277445905,
+      "grad_norm": 0.41762988105274723,
+      "learning_rate": 4.8489395269588725e-06,
+      "loss": 0.6171,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6910925262179742,
+      "grad_norm": 0.3899823197542196,
+      "learning_rate": 4.848819983286677e-06,
+      "loss": 0.6133,
+      "step": 2603
+    },
+    {
+      "epoch": 0.691358024691358,
+      "grad_norm": 0.3848599299908165,
+      "learning_rate": 4.848700393806739e-06,
+      "loss": 0.6239,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6916235231647418,
+      "grad_norm": 0.38733771313066306,
+      "learning_rate": 4.848580758521393e-06,
+      "loss": 0.6302,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6918890216381256,
+      "grad_norm": 0.39635704743554306,
+      "learning_rate": 4.848461077432973e-06,
+      "loss": 0.6182,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6921545201115094,
+      "grad_norm": 0.39039677922944666,
+      "learning_rate": 4.848341350543811e-06,
+      "loss": 0.63,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6924200185848931,
+      "grad_norm": 0.3852633028118449,
+      "learning_rate": 4.848221577856244e-06,
+      "loss": 0.6088,
+      "step": 2608
+    },
+    {
+      "epoch": 0.692685517058277,
+      "grad_norm": 0.39322254504668075,
+      "learning_rate": 4.848101759372605e-06,
+      "loss": 0.609,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6929510155316607,
+      "grad_norm": 0.3971663647336428,
+      "learning_rate": 4.847981895095234e-06,
+      "loss": 0.6183,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6932165140050445,
+      "grad_norm": 0.3837183314663157,
+      "learning_rate": 4.8478619850264675e-06,
+      "loss": 0.6198,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6934820124784282,
+      "grad_norm": 0.392346323756682,
+      "learning_rate": 4.847742029168643e-06,
+      "loss": 0.6466,
+      "step": 2612
+    },
+    {
+      "epoch": 0.693747510951812,
+      "grad_norm": 0.3698956181295069,
+      "learning_rate": 4.8476220275241e-06,
+      "loss": 0.6219,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6940130094251958,
+      "grad_norm": 0.3773324830349831,
+      "learning_rate": 4.847501980095181e-06,
+      "loss": 0.5703,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6942785078985796,
+      "grad_norm": 0.38836175044527804,
+      "learning_rate": 4.8473818868842246e-06,
+      "loss": 0.6545,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6945440063719633,
+      "grad_norm": 0.4007302920292224,
+      "learning_rate": 4.8472617478935744e-06,
+      "loss": 0.6249,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6948095048453471,
+      "grad_norm": 0.39216784343425254,
+      "learning_rate": 4.8471415631255725e-06,
+      "loss": 0.619,
+      "step": 2617
+    },
+    {
+      "epoch": 0.695075003318731,
+      "grad_norm": 0.38146862062044684,
+      "learning_rate": 4.847021332582564e-06,
+      "loss": 0.643,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6953405017921147,
+      "grad_norm": 0.3844625924455284,
+      "learning_rate": 4.846901056266893e-06,
+      "loss": 0.6569,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6956060002654985,
+      "grad_norm": 0.3870792179509595,
+      "learning_rate": 4.846780734180905e-06,
+      "loss": 0.6362,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6958714987388822,
+      "grad_norm": 0.3878343142627452,
+      "learning_rate": 4.846660366326947e-06,
+      "loss": 0.6493,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6961369972122661,
+      "grad_norm": 0.38824424518080686,
+      "learning_rate": 4.846539952707366e-06,
+      "loss": 0.6372,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6964024956856498,
+      "grad_norm": 0.381097332062954,
+      "learning_rate": 4.84641949332451e-06,
+      "loss": 0.6238,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6966679941590336,
+      "grad_norm": 0.39249737648919003,
+      "learning_rate": 4.8462989881807295e-06,
+      "loss": 0.6033,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6969334926324173,
+      "grad_norm": 0.38492329963418653,
+      "learning_rate": 4.846178437278375e-06,
+      "loss": 0.6448,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6971989911058012,
+      "grad_norm": 0.3803116438683266,
+      "learning_rate": 4.846057840619795e-06,
+      "loss": 0.6274,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6974644895791849,
+      "grad_norm": 0.3875639142413487,
+      "learning_rate": 4.845937198207344e-06,
+      "loss": 0.6433,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6977299880525687,
+      "grad_norm": 0.39390029475279664,
+      "learning_rate": 4.8458165100433725e-06,
+      "loss": 0.6401,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6979954865259524,
+      "grad_norm": 0.3774646506851741,
+      "learning_rate": 4.845695776130236e-06,
+      "loss": 0.597,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6982609849993362,
+      "grad_norm": 0.39145801555943405,
+      "learning_rate": 4.845574996470289e-06,
+      "loss": 0.65,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6985264834727201,
+      "grad_norm": 0.37923453762103676,
+      "learning_rate": 4.845454171065887e-06,
+      "loss": 0.6119,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6987919819461038,
+      "grad_norm": 0.3912627709400642,
+      "learning_rate": 4.8453332999193854e-06,
+      "loss": 0.6154,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6990574804194876,
+      "grad_norm": 0.39647054375916135,
+      "learning_rate": 4.845212383033142e-06,
+      "loss": 0.5926,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6993229788928713,
+      "grad_norm": 0.38543941146569916,
+      "learning_rate": 4.845091420409515e-06,
+      "loss": 0.5537,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6995884773662552,
+      "grad_norm": 0.37478166965820847,
+      "learning_rate": 4.844970412050864e-06,
+      "loss": 0.608,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6998539758396389,
+      "grad_norm": 0.3936287744363812,
+      "learning_rate": 4.8448493579595485e-06,
+      "loss": 0.616,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7001194743130227,
+      "grad_norm": 0.3915730473874248,
+      "learning_rate": 4.844728258137929e-06,
+      "loss": 0.6132,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7003849727864064,
+      "grad_norm": 0.38685701677178286,
+      "learning_rate": 4.8446071125883674e-06,
+      "loss": 0.6361,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7006504712597903,
+      "grad_norm": 0.3790822534511941,
+      "learning_rate": 4.844485921313227e-06,
+      "loss": 0.6417,
+      "step": 2639
+    },
+    {
+      "epoch": 0.700915969733174,
+      "grad_norm": 0.39605015319172515,
+      "learning_rate": 4.844364684314871e-06,
+      "loss": 0.6304,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7011814682065578,
+      "grad_norm": 0.3925159358194844,
+      "learning_rate": 4.844243401595663e-06,
+      "loss": 0.6234,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7014469666799416,
+      "grad_norm": 0.3914855293098387,
+      "learning_rate": 4.84412207315797e-06,
+      "loss": 0.622,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7017124651533254,
+      "grad_norm": 0.38446146141623927,
+      "learning_rate": 4.844000699004156e-06,
+      "loss": 0.6129,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7019779636267092,
+      "grad_norm": 0.3955420711556819,
+      "learning_rate": 4.843879279136589e-06,
+      "loss": 0.6085,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7022434621000929,
+      "grad_norm": 0.3722152369503793,
+      "learning_rate": 4.843757813557639e-06,
+      "loss": 0.6188,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7025089605734767,
+      "grad_norm": 0.39288394958209916,
+      "learning_rate": 4.843636302269672e-06,
+      "loss": 0.6134,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7027744590468605,
+      "grad_norm": 0.3780279065230019,
+      "learning_rate": 4.843514745275059e-06,
+      "loss": 0.6159,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7030399575202443,
+      "grad_norm": 0.40262741603223956,
+      "learning_rate": 4.84339314257617e-06,
+      "loss": 0.6396,
+      "step": 2648
+    },
+    {
+      "epoch": 0.703305455993628,
+      "grad_norm": 0.3997449926090783,
+      "learning_rate": 4.843271494175378e-06,
+      "loss": 0.6559,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7035709544670118,
+      "grad_norm": 0.4129818372307495,
+      "learning_rate": 4.843149800075054e-06,
+      "loss": 0.6005,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7038364529403955,
+      "grad_norm": 0.3859111695168442,
+      "learning_rate": 4.8430280602775725e-06,
+      "loss": 0.5729,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7041019514137794,
+      "grad_norm": 0.3785485467842736,
+      "learning_rate": 4.842906274785306e-06,
+      "loss": 0.5954,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7043674498871632,
+      "grad_norm": 0.38978516180672673,
+      "learning_rate": 4.842784443600632e-06,
+      "loss": 0.6203,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7046329483605469,
+      "grad_norm": 0.3996638681439292,
+      "learning_rate": 4.842662566725924e-06,
+      "loss": 0.644,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7048984468339307,
+      "grad_norm": 0.3821948974207942,
+      "learning_rate": 4.8425406441635615e-06,
+      "loss": 0.5864,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7051639453073145,
+      "grad_norm": 0.37975802340295783,
+      "learning_rate": 4.84241867591592e-06,
+      "loss": 0.6202,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7054294437806983,
+      "grad_norm": 0.3865295040695601,
+      "learning_rate": 4.8422966619853795e-06,
+      "loss": 0.6403,
+      "step": 2657
+    },
+    {
+      "epoch": 0.705694942254082,
+      "grad_norm": 0.38345848440660857,
+      "learning_rate": 4.842174602374319e-06,
+      "loss": 0.6259,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7059604407274658,
+      "grad_norm": 0.38894517327788974,
+      "learning_rate": 4.842052497085119e-06,
+      "loss": 0.6221,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7062259392008496,
+      "grad_norm": 0.4014850168709462,
+      "learning_rate": 4.841930346120161e-06,
+      "loss": 0.5752,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7064914376742334,
+      "grad_norm": 0.3779393000447171,
+      "learning_rate": 4.841808149481827e-06,
+      "loss": 0.6369,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7067569361476171,
+      "grad_norm": 0.3980667358981174,
+      "learning_rate": 4.841685907172501e-06,
+      "loss": 0.6006,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7070224346210009,
+      "grad_norm": 0.3965787010329716,
+      "learning_rate": 4.841563619194565e-06,
+      "loss": 0.5865,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7072879330943848,
+      "grad_norm": 0.4149449600830029,
+      "learning_rate": 4.841441285550407e-06,
+      "loss": 0.6416,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7075534315677685,
+      "grad_norm": 0.3986300544804875,
+      "learning_rate": 4.84131890624241e-06,
+      "loss": 0.6033,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7078189300411523,
+      "grad_norm": 0.40950327115232465,
+      "learning_rate": 4.841196481272962e-06,
+      "loss": 0.6375,
+      "step": 2666
+    },
+    {
+      "epoch": 0.708084428514536,
+      "grad_norm": 0.39344157737685936,
+      "learning_rate": 4.841074010644451e-06,
+      "loss": 0.6013,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7083499269879198,
+      "grad_norm": 0.3885292628877199,
+      "learning_rate": 4.840951494359264e-06,
+      "loss": 0.6394,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7086154254613036,
+      "grad_norm": 0.39807870604794077,
+      "learning_rate": 4.840828932419792e-06,
+      "loss": 0.6394,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7088809239346874,
+      "grad_norm": 0.3940794565330124,
+      "learning_rate": 4.840706324828423e-06,
+      "loss": 0.6408,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7091464224080711,
+      "grad_norm": 0.389525565444789,
+      "learning_rate": 4.840583671587551e-06,
+      "loss": 0.6206,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7094119208814549,
+      "grad_norm": 0.39086680902995535,
+      "learning_rate": 4.840460972699567e-06,
+      "loss": 0.63,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 0.38162221825724835,
+      "learning_rate": 4.840338228166862e-06,
+      "loss": 0.6207,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7099429178282225,
+      "grad_norm": 0.38631093654671855,
+      "learning_rate": 4.840215437991833e-06,
+      "loss": 0.5902,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7102084163016062,
+      "grad_norm": 0.40197529788488806,
+      "learning_rate": 4.840092602176872e-06,
+      "loss": 0.657,
+      "step": 2675
+    },
+    {
+      "epoch": 0.71047391477499,
+      "grad_norm": 0.3910634987928013,
+      "learning_rate": 4.839969720724377e-06,
+      "loss": 0.6357,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7107394132483739,
+      "grad_norm": 0.3926464229659296,
+      "learning_rate": 4.839846793636742e-06,
+      "loss": 0.6051,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7110049117217576,
+      "grad_norm": 0.3786085783996354,
+      "learning_rate": 4.839723820916366e-06,
+      "loss": 0.609,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7112704101951414,
+      "grad_norm": 0.37515048093952635,
+      "learning_rate": 4.839600802565647e-06,
+      "loss": 0.5779,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7115359086685251,
+      "grad_norm": 0.4016594721468733,
+      "learning_rate": 4.839477738586984e-06,
+      "loss": 0.5829,
+      "step": 2680
+    },
+    {
+      "epoch": 0.711801407141909,
+      "grad_norm": 0.38247376456700155,
+      "learning_rate": 4.839354628982777e-06,
+      "loss": 0.5868,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7120669056152927,
+      "grad_norm": 0.3904012100676597,
+      "learning_rate": 4.839231473755427e-06,
+      "loss": 0.6004,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7123324040886765,
+      "grad_norm": 0.3939250292403363,
+      "learning_rate": 4.839108272907336e-06,
+      "loss": 0.5925,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7125979025620602,
+      "grad_norm": 0.37986215662453254,
+      "learning_rate": 4.8389850264409055e-06,
+      "loss": 0.6062,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7128634010354441,
+      "grad_norm": 0.39323207411072897,
+      "learning_rate": 4.838861734358541e-06,
+      "loss": 0.6057,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7131288995088279,
+      "grad_norm": 0.38716468749046573,
+      "learning_rate": 4.8387383966626465e-06,
+      "loss": 0.5998,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7133943979822116,
+      "grad_norm": 0.3867185961010731,
+      "learning_rate": 4.838615013355627e-06,
+      "loss": 0.64,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7136598964555954,
+      "grad_norm": 0.4484737722836541,
+      "learning_rate": 4.838491584439888e-06,
+      "loss": 0.5948,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7139253949289791,
+      "grad_norm": 0.3810109151145234,
+      "learning_rate": 4.8383681099178385e-06,
+      "loss": 0.6127,
+      "step": 2689
+    },
+    {
+      "epoch": 0.714190893402363,
+      "grad_norm": 0.3888128776298206,
+      "learning_rate": 4.838244589791885e-06,
+      "loss": 0.6102,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7144563918757467,
+      "grad_norm": 0.3909530990167454,
+      "learning_rate": 4.8381210240644375e-06,
+      "loss": 0.63,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7147218903491305,
+      "grad_norm": 0.38630949045402846,
+      "learning_rate": 4.837997412737905e-06,
+      "loss": 0.611,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7149873888225142,
+      "grad_norm": 0.39549871748641846,
+      "learning_rate": 4.837873755814698e-06,
+      "loss": 0.6378,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7152528872958981,
+      "grad_norm": 0.37527015893172583,
+      "learning_rate": 4.837750053297229e-06,
+      "loss": 0.6327,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7155183857692818,
+      "grad_norm": 0.4213494559443399,
+      "learning_rate": 4.83762630518791e-06,
+      "loss": 0.5699,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7157838842426656,
+      "grad_norm": 0.38760136451103044,
+      "learning_rate": 4.837502511489155e-06,
+      "loss": 0.5828,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7160493827160493,
+      "grad_norm": 0.4034611300192151,
+      "learning_rate": 4.837378672203378e-06,
+      "loss": 0.597,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7163148811894332,
+      "grad_norm": 0.4034669076033467,
+      "learning_rate": 4.837254787332993e-06,
+      "loss": 0.6251,
+      "step": 2698
+    },
+    {
+      "epoch": 0.716580379662817,
+      "grad_norm": 0.4115041210496319,
+      "learning_rate": 4.837130856880418e-06,
+      "loss": 0.6053,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7168458781362007,
+      "grad_norm": 0.4061381852681957,
+      "learning_rate": 4.837006880848069e-06,
+      "loss": 0.6161,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7171113766095845,
+      "grad_norm": 0.3922997461013148,
+      "learning_rate": 4.8368828592383635e-06,
+      "loss": 0.6068,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7173768750829683,
+      "grad_norm": 0.3817760610409912,
+      "learning_rate": 4.836758792053721e-06,
+      "loss": 0.596,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7176423735563521,
+      "grad_norm": 0.3945996356900288,
+      "learning_rate": 4.83663467929656e-06,
+      "loss": 0.6483,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7179078720297358,
+      "grad_norm": 0.39435325139445204,
+      "learning_rate": 4.8365105209693015e-06,
+      "loss": 0.6582,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7181733705031196,
+      "grad_norm": 0.3665830778055999,
+      "learning_rate": 4.836386317074368e-06,
+      "loss": 0.6147,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7184388689765033,
+      "grad_norm": 0.39212906815583454,
+      "learning_rate": 4.8362620676141805e-06,
+      "loss": 0.622,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7187043674498872,
+      "grad_norm": 0.38888920059635756,
+      "learning_rate": 4.8361377725911625e-06,
+      "loss": 0.6221,
+      "step": 2707
+    },
+    {
+      "epoch": 0.718969865923271,
+      "grad_norm": 0.3838288585316257,
+      "learning_rate": 4.836013432007738e-06,
+      "loss": 0.609,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7192353643966547,
+      "grad_norm": 0.3766154712694823,
+      "learning_rate": 4.835889045866332e-06,
+      "loss": 0.6175,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7195008628700384,
+      "grad_norm": 0.3935365236909523,
+      "learning_rate": 4.83576461416937e-06,
+      "loss": 0.6386,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7197663613434223,
+      "grad_norm": 0.3746494541735743,
+      "learning_rate": 4.83564013691928e-06,
+      "loss": 0.6039,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7200318598168061,
+      "grad_norm": 0.3958855733104673,
+      "learning_rate": 4.835515614118488e-06,
+      "loss": 0.6206,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7202973582901898,
+      "grad_norm": 0.3990273793543752,
+      "learning_rate": 4.8353910457694235e-06,
+      "loss": 0.5893,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7205628567635736,
+      "grad_norm": 0.3825074603959132,
+      "learning_rate": 4.835266431874516e-06,
+      "loss": 0.632,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7208283552369574,
+      "grad_norm": 0.38485740587732975,
+      "learning_rate": 4.835141772436194e-06,
+      "loss": 0.6317,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7210938537103412,
+      "grad_norm": 0.4067854496688009,
+      "learning_rate": 4.835017067456892e-06,
+      "loss": 0.6057,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7213593521837249,
+      "grad_norm": 0.38906319773523507,
+      "learning_rate": 4.834892316939038e-06,
+      "loss": 0.6125,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7216248506571087,
+      "grad_norm": 0.40826904243452194,
+      "learning_rate": 4.834767520885069e-06,
+      "loss": 0.6612,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7218903491304925,
+      "grad_norm": 0.3927443248306872,
+      "learning_rate": 4.834642679297415e-06,
+      "loss": 0.6076,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7221558476038763,
+      "grad_norm": 0.47996324107007193,
+      "learning_rate": 4.834517792178513e-06,
+      "loss": 0.5771,
+      "step": 2720
+    },
+    {
+      "epoch": 0.72242134607726,
+      "grad_norm": 0.37420339276518444,
+      "learning_rate": 4.8343928595307986e-06,
+      "loss": 0.6255,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7226868445506438,
+      "grad_norm": 0.4346393954653237,
+      "learning_rate": 4.834267881356708e-06,
+      "loss": 0.6164,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7229523430240277,
+      "grad_norm": 0.598646683483842,
+      "learning_rate": 4.834142857658678e-06,
+      "loss": 0.6173,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7232178414974114,
+      "grad_norm": 0.379375202482494,
+      "learning_rate": 4.834017788439148e-06,
+      "loss": 0.6317,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7234833399707952,
+      "grad_norm": 0.4396162719192362,
+      "learning_rate": 4.833892673700556e-06,
+      "loss": 0.6448,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7237488384441789,
+      "grad_norm": 0.46814974281808003,
+      "learning_rate": 4.833767513445342e-06,
+      "loss": 0.5925,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7240143369175627,
+      "grad_norm": 0.6267303996799323,
+      "learning_rate": 4.833642307675948e-06,
+      "loss": 0.598,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7242798353909465,
+      "grad_norm": 0.41578917951553057,
+      "learning_rate": 4.833517056394816e-06,
+      "loss": 0.6344,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7245453338643303,
+      "grad_norm": 0.5404076902512648,
+      "learning_rate": 4.833391759604386e-06,
+      "loss": 0.6511,
+      "step": 2729
+    },
+    {
+      "epoch": 0.724810832337714,
+      "grad_norm": 0.5384994277964352,
+      "learning_rate": 4.833266417307105e-06,
+      "loss": 0.6271,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7250763308110978,
+      "grad_norm": 0.4374745481304502,
+      "learning_rate": 4.833141029505417e-06,
+      "loss": 0.5907,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7253418292844817,
+      "grad_norm": 0.4138675087051396,
+      "learning_rate": 4.8330155962017645e-06,
+      "loss": 0.6143,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7256073277578654,
+      "grad_norm": 0.4640266889654643,
+      "learning_rate": 4.832890117398596e-06,
+      "loss": 0.597,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7258728262312492,
+      "grad_norm": 0.4441762398041202,
+      "learning_rate": 4.8327645930983595e-06,
+      "loss": 0.651,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7261383247046329,
+      "grad_norm": 0.47555521385135624,
+      "learning_rate": 4.832639023303501e-06,
+      "loss": 0.6317,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7264038231780168,
+      "grad_norm": 0.4006585151692151,
+      "learning_rate": 4.832513408016471e-06,
+      "loss": 0.5759,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7266693216514005,
+      "grad_norm": 0.4202307412152153,
+      "learning_rate": 4.832387747239717e-06,
+      "loss": 0.5972,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7269348201247843,
+      "grad_norm": 0.4216843139057117,
+      "learning_rate": 4.832262040975692e-06,
+      "loss": 0.6194,
+      "step": 2738
+    },
+    {
+      "epoch": 0.727200318598168,
+      "grad_norm": 0.5441528042431764,
+      "learning_rate": 4.832136289226847e-06,
+      "loss": 0.595,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7274658170715519,
+      "grad_norm": 0.3886247515099562,
+      "learning_rate": 4.832010491995634e-06,
+      "loss": 0.6416,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7277313155449356,
+      "grad_norm": 0.39390311568534647,
+      "learning_rate": 4.831884649284507e-06,
+      "loss": 0.6182,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7279968140183194,
+      "grad_norm": 0.38143248397905566,
+      "learning_rate": 4.83175876109592e-06,
+      "loss": 0.5992,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7282623124917031,
+      "grad_norm": 0.3990434085906802,
+      "learning_rate": 4.8316328274323275e-06,
+      "loss": 0.6382,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7285278109650869,
+      "grad_norm": 0.4152186638628756,
+      "learning_rate": 4.831506848296187e-06,
+      "loss": 0.625,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7287933094384708,
+      "grad_norm": 0.39717738614322656,
+      "learning_rate": 4.831380823689954e-06,
+      "loss": 0.6854,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7290588079118545,
+      "grad_norm": 0.4064110160412851,
+      "learning_rate": 4.8312547536160855e-06,
+      "loss": 0.6421,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7293243063852383,
+      "grad_norm": 0.394455839798419,
+      "learning_rate": 4.831128638077044e-06,
+      "loss": 0.6162,
+      "step": 2747
+    },
+    {
+      "epoch": 0.729589804858622,
+      "grad_norm": 0.37564715334029924,
+      "learning_rate": 4.831002477075284e-06,
+      "loss": 0.6246,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7298553033320059,
+      "grad_norm": 0.397268055489803,
+      "learning_rate": 4.830876270613269e-06,
+      "loss": 0.6154,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7301208018053896,
+      "grad_norm": 0.39290058760195923,
+      "learning_rate": 4.830750018693461e-06,
+      "loss": 0.635,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7303863002787734,
+      "grad_norm": 0.417533315899868,
+      "learning_rate": 4.830623721318319e-06,
+      "loss": 0.5752,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7306517987521571,
+      "grad_norm": 0.3816486465129925,
+      "learning_rate": 4.8304973784903094e-06,
+      "loss": 0.595,
+      "step": 2752
+    },
+    {
+      "epoch": 0.730917297225541,
+      "grad_norm": 0.3919818387608073,
+      "learning_rate": 4.830370990211896e-06,
+      "loss": 0.6128,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7311827956989247,
+      "grad_norm": 0.39466680240663665,
+      "learning_rate": 4.830244556485541e-06,
+      "loss": 0.641,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7314482941723085,
+      "grad_norm": 0.40178825699033777,
+      "learning_rate": 4.830118077313711e-06,
+      "loss": 0.5925,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7317137926456923,
+      "grad_norm": 0.3917799700294571,
+      "learning_rate": 4.829991552698875e-06,
+      "loss": 0.6573,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7319792911190761,
+      "grad_norm": 0.3963976162008505,
+      "learning_rate": 4.829864982643498e-06,
+      "loss": 0.6097,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7322447895924599,
+      "grad_norm": 0.3834633489838258,
+      "learning_rate": 4.82973836715005e-06,
+      "loss": 0.6348,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7325102880658436,
+      "grad_norm": 0.3858137908361253,
+      "learning_rate": 4.829611706220999e-06,
+      "loss": 0.6223,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7327757865392274,
+      "grad_norm": 0.402427907671124,
+      "learning_rate": 4.829484999858816e-06,
+      "loss": 0.59,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7330412850126112,
+      "grad_norm": 0.39852846856995233,
+      "learning_rate": 4.829358248065972e-06,
+      "loss": 0.6413,
+      "step": 2761
+    },
+    {
+      "epoch": 0.733306783485995,
+      "grad_norm": 0.3968234651476136,
+      "learning_rate": 4.829231450844939e-06,
+      "loss": 0.6253,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7335722819593787,
+      "grad_norm": 0.3878410739598387,
+      "learning_rate": 4.8291046081981895e-06,
+      "loss": 0.6117,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7338377804327625,
+      "grad_norm": 0.3832351555693445,
+      "learning_rate": 4.828977720128198e-06,
+      "loss": 0.6264,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7341032789061462,
+      "grad_norm": 0.3887570327875792,
+      "learning_rate": 4.828850786637438e-06,
+      "loss": 0.602,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7343687773795301,
+      "grad_norm": 0.39188880916219765,
+      "learning_rate": 4.828723807728386e-06,
+      "loss": 0.6316,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7346342758529139,
+      "grad_norm": 0.4009026649474955,
+      "learning_rate": 4.8285967834035186e-06,
+      "loss": 0.5947,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7348997743262976,
+      "grad_norm": 0.394562762404538,
+      "learning_rate": 4.8284697136653115e-06,
+      "loss": 0.6423,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7351652727996814,
+      "grad_norm": 0.3911591541800308,
+      "learning_rate": 4.828342598516244e-06,
+      "loss": 0.6325,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7354307712730652,
+      "grad_norm": 0.3805036859948306,
+      "learning_rate": 4.828215437958796e-06,
+      "loss": 0.5986,
+      "step": 2770
+    },
+    {
+      "epoch": 0.735696269746449,
+      "grad_norm": 0.3946856110402343,
+      "learning_rate": 4.828088231995446e-06,
+      "loss": 0.6102,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7359617682198327,
+      "grad_norm": 0.40555422649461653,
+      "learning_rate": 4.827960980628675e-06,
+      "loss": 0.6308,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7362272666932165,
+      "grad_norm": 0.3987021479530642,
+      "learning_rate": 4.827833683860966e-06,
+      "loss": 0.6459,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7364927651666003,
+      "grad_norm": 0.3902535314183136,
+      "learning_rate": 4.8277063416948e-06,
+      "loss": 0.6567,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7367582636399841,
+      "grad_norm": 0.38156721487418765,
+      "learning_rate": 4.827578954132661e-06,
+      "loss": 0.6234,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7370237621133678,
+      "grad_norm": 0.395287256562999,
+      "learning_rate": 4.827451521177033e-06,
+      "loss": 0.6515,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7372892605867516,
+      "grad_norm": 0.3850843482038379,
+      "learning_rate": 4.827324042830403e-06,
+      "loss": 0.6254,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7375547590601355,
+      "grad_norm": 0.37816520076211557,
+      "learning_rate": 4.8271965190952555e-06,
+      "loss": 0.5893,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7378202575335192,
+      "grad_norm": 0.38186737655543224,
+      "learning_rate": 4.8270689499740776e-06,
+      "loss": 0.6036,
+      "step": 2779
+    },
+    {
+      "epoch": 0.738085756006903,
+      "grad_norm": 0.384296951487842,
+      "learning_rate": 4.826941335469357e-06,
+      "loss": 0.6115,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7383512544802867,
+      "grad_norm": 0.40834758981838704,
+      "learning_rate": 4.8268136755835835e-06,
+      "loss": 0.6286,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7386167529536705,
+      "grad_norm": 0.3877035623227378,
+      "learning_rate": 4.826685970319247e-06,
+      "loss": 0.6293,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7388822514270543,
+      "grad_norm": 0.39568721704849596,
+      "learning_rate": 4.826558219678837e-06,
+      "loss": 0.6583,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7391477499004381,
+      "grad_norm": 0.377479796213409,
+      "learning_rate": 4.8264304236648455e-06,
+      "loss": 0.6137,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7394132483738218,
+      "grad_norm": 0.3965867465751187,
+      "learning_rate": 4.826302582279764e-06,
+      "loss": 0.6053,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7396787468472056,
+      "grad_norm": 0.39723326621145977,
+      "learning_rate": 4.826174695526086e-06,
+      "loss": 0.6367,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7399442453205894,
+      "grad_norm": 0.39248336432761927,
+      "learning_rate": 4.826046763406307e-06,
+      "loss": 0.6014,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7402097437939732,
+      "grad_norm": 0.3804681684215838,
+      "learning_rate": 4.825918785922921e-06,
+      "loss": 0.6133,
+      "step": 2788
+    },
+    {
+      "epoch": 0.740475242267357,
+      "grad_norm": 0.38887933143631737,
+      "learning_rate": 4.825790763078423e-06,
+      "loss": 0.5993,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.38083737374050763,
+      "learning_rate": 4.825662694875312e-06,
+      "loss": 0.5987,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7410062392141246,
+      "grad_norm": 0.3675860024020678,
+      "learning_rate": 4.825534581316083e-06,
+      "loss": 0.6026,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7412717376875083,
+      "grad_norm": 0.37996911363114316,
+      "learning_rate": 4.825406422403235e-06,
+      "loss": 0.6402,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7415372361608921,
+      "grad_norm": 0.3976011733980962,
+      "learning_rate": 4.82527821813927e-06,
+      "loss": 0.651,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7418027346342758,
+      "grad_norm": 0.42514282046473584,
+      "learning_rate": 4.825149968526686e-06,
+      "loss": 0.6119,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7420682331076597,
+      "grad_norm": 0.3837758468126048,
+      "learning_rate": 4.8250216735679844e-06,
+      "loss": 0.6138,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7423337315810434,
+      "grad_norm": 0.3797692586416595,
+      "learning_rate": 4.824893333265667e-06,
+      "loss": 0.6113,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7425992300544272,
+      "grad_norm": 0.38163235535132634,
+      "learning_rate": 4.824764947622238e-06,
+      "loss": 0.6073,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7428647285278109,
+      "grad_norm": 0.37031221252924607,
+      "learning_rate": 4.8246365166402005e-06,
+      "loss": 0.5934,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7431302270011948,
+      "grad_norm": 0.3933555108740605,
+      "learning_rate": 4.824508040322059e-06,
+      "loss": 0.6373,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7433957254745786,
+      "grad_norm": 0.3919329621284459,
+      "learning_rate": 4.824379518670319e-06,
+      "loss": 0.6325,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7436612239479623,
+      "grad_norm": 0.38737802704491076,
+      "learning_rate": 4.824250951687488e-06,
+      "loss": 0.6119,
+      "step": 2801
+    },
+    {
+      "epoch": 0.743926722421346,
+      "grad_norm": 0.3892074664217247,
+      "learning_rate": 4.824122339376072e-06,
+      "loss": 0.6687,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7441922208947298,
+      "grad_norm": 0.3848385033170652,
+      "learning_rate": 4.8239936817385815e-06,
+      "loss": 0.5734,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7444577193681137,
+      "grad_norm": 0.38198065376022733,
+      "learning_rate": 4.823864978777522e-06,
+      "loss": 0.5928,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7447232178414974,
+      "grad_norm": 0.375840257615797,
+      "learning_rate": 4.823736230495407e-06,
+      "loss": 0.6078,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7449887163148812,
+      "grad_norm": 0.38617472866324026,
+      "learning_rate": 4.823607436894746e-06,
+      "loss": 0.6557,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7452542147882649,
+      "grad_norm": 0.3964718064426852,
+      "learning_rate": 4.823478597978049e-06,
+      "loss": 0.6415,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7455197132616488,
+      "grad_norm": 0.38162919943729473,
+      "learning_rate": 4.823349713747832e-06,
+      "loss": 0.6236,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7457852117350325,
+      "grad_norm": 0.39130304025533047,
+      "learning_rate": 4.823220784206608e-06,
+      "loss": 0.6366,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7460507102084163,
+      "grad_norm": 0.37975547292941775,
+      "learning_rate": 4.823091809356889e-06,
+      "loss": 0.5996,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7463162086818,
+      "grad_norm": 0.3899161626263363,
+      "learning_rate": 4.822962789201192e-06,
+      "loss": 0.5977,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7465817071551839,
+      "grad_norm": 0.3813481635025675,
+      "learning_rate": 4.822833723742033e-06,
+      "loss": 0.5883,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7468472056285677,
+      "grad_norm": 0.40208793653077435,
+      "learning_rate": 4.8227046129819295e-06,
+      "loss": 0.6243,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7471127041019514,
+      "grad_norm": 0.40462108997191293,
+      "learning_rate": 4.8225754569234e-06,
+      "loss": 0.6369,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7473782025753352,
+      "grad_norm": 0.39903228538964913,
+      "learning_rate": 4.822446255568961e-06,
+      "loss": 0.6009,
+      "step": 2815
+    },
+    {
+      "epoch": 0.747643701048719,
+      "grad_norm": 0.38714760650706337,
+      "learning_rate": 4.822317008921133e-06,
+      "loss": 0.6141,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7479091995221028,
+      "grad_norm": 0.3896480079381763,
+      "learning_rate": 4.82218771698244e-06,
+      "loss": 0.6239,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7481746979954865,
+      "grad_norm": 0.38876274722907134,
+      "learning_rate": 4.822058379755399e-06,
+      "loss": 0.5795,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7484401964688703,
+      "grad_norm": 0.38595288152480245,
+      "learning_rate": 4.821928997242534e-06,
+      "loss": 0.5969,
+      "step": 2819
+    },
+    {
+      "epoch": 0.748705694942254,
+      "grad_norm": 0.3877007577720642,
+      "learning_rate": 4.821799569446368e-06,
+      "loss": 0.6484,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7489711934156379,
+      "grad_norm": 0.37798400269211985,
+      "learning_rate": 4.821670096369427e-06,
+      "loss": 0.6363,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7492366918890216,
+      "grad_norm": 0.3849291661633291,
+      "learning_rate": 4.821540578014234e-06,
+      "loss": 0.6673,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7495021903624054,
+      "grad_norm": 0.40099800685563236,
+      "learning_rate": 4.821411014383317e-06,
+      "loss": 0.6039,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7497676888357891,
+      "grad_norm": 0.3833907503353607,
+      "learning_rate": 4.8212814054792e-06,
+      "loss": 0.6504,
+      "step": 2824
+    },
+    {
+      "epoch": 0.750033187309173,
+      "grad_norm": 0.38224448187116683,
+      "learning_rate": 4.821151751304412e-06,
+      "loss": 0.6218,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7502986857825568,
+      "grad_norm": 0.38988209626322307,
+      "learning_rate": 4.821022051861482e-06,
+      "loss": 0.5858,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7505641842559405,
+      "grad_norm": 0.3754834798575984,
+      "learning_rate": 4.8208923071529394e-06,
+      "loss": 0.6076,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7508296827293243,
+      "grad_norm": 0.39057354478944306,
+      "learning_rate": 4.820762517181314e-06,
+      "loss": 0.5906,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7510951812027081,
+      "grad_norm": 0.3871768544576605,
+      "learning_rate": 4.820632681949138e-06,
+      "loss": 0.6351,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7513606796760919,
+      "grad_norm": 0.3948262560785759,
+      "learning_rate": 4.820502801458942e-06,
+      "loss": 0.6198,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7516261781494756,
+      "grad_norm": 0.38015628806143265,
+      "learning_rate": 4.82037287571326e-06,
+      "loss": 0.6152,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7518916766228594,
+      "grad_norm": 0.38438109287268973,
+      "learning_rate": 4.820242904714626e-06,
+      "loss": 0.5904,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7521571750962432,
+      "grad_norm": 0.4081007225432161,
+      "learning_rate": 4.820112888465574e-06,
+      "loss": 0.6372,
+      "step": 2833
+    },
+    {
+      "epoch": 0.752422673569627,
+      "grad_norm": 0.3821812290906295,
+      "learning_rate": 4.8199828269686395e-06,
+      "loss": 0.639,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7526881720430108,
+      "grad_norm": 0.381324449389942,
+      "learning_rate": 4.819852720226361e-06,
+      "loss": 0.6761,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7529536705163945,
+      "grad_norm": 0.3730766585898743,
+      "learning_rate": 4.819722568241274e-06,
+      "loss": 0.5566,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7532191689897783,
+      "grad_norm": 0.3901910047325791,
+      "learning_rate": 4.8195923710159165e-06,
+      "loss": 0.5964,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7534846674631621,
+      "grad_norm": 0.3888273246622894,
+      "learning_rate": 4.81946212855283e-06,
+      "loss": 0.6175,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7537501659365459,
+      "grad_norm": 0.39070985990796303,
+      "learning_rate": 4.819331840854551e-06,
+      "loss": 0.6407,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7540156644099296,
+      "grad_norm": 0.37407735286231825,
+      "learning_rate": 4.819201507923623e-06,
+      "loss": 0.6268,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7542811628833134,
+      "grad_norm": 0.39490825929126455,
+      "learning_rate": 4.819071129762588e-06,
+      "loss": 0.6218,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7545466613566972,
+      "grad_norm": 0.3904636117438276,
+      "learning_rate": 4.818940706373988e-06,
+      "loss": 0.6264,
+      "step": 2842
+    },
+    {
+      "epoch": 0.754812159830081,
+      "grad_norm": 0.3758749674362929,
+      "learning_rate": 4.8188102377603655e-06,
+      "loss": 0.664,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7550776583034647,
+      "grad_norm": 0.39353049150138575,
+      "learning_rate": 4.818679723924266e-06,
+      "loss": 0.6318,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7553431567768485,
+      "grad_norm": 0.38782710599549575,
+      "learning_rate": 4.818549164868236e-06,
+      "loss": 0.6206,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7556086552502324,
+      "grad_norm": 0.38531966142503277,
+      "learning_rate": 4.818418560594819e-06,
+      "loss": 0.6318,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7558741537236161,
+      "grad_norm": 0.3796290069349175,
+      "learning_rate": 4.8182879111065635e-06,
+      "loss": 0.6292,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7561396521969999,
+      "grad_norm": 0.39733783086606916,
+      "learning_rate": 4.818157216406019e-06,
+      "loss": 0.6142,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7564051506703836,
+      "grad_norm": 0.41336730869352195,
+      "learning_rate": 4.818026476495732e-06,
+      "loss": 0.6049,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7566706491437675,
+      "grad_norm": 0.39396539375502837,
+      "learning_rate": 4.817895691378253e-06,
+      "loss": 0.6301,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7569361476171512,
+      "grad_norm": 0.402771897323766,
+      "learning_rate": 4.817764861056134e-06,
+      "loss": 0.6421,
+      "step": 2851
+    },
+    {
+      "epoch": 0.757201646090535,
+      "grad_norm": 0.4261722074461022,
+      "learning_rate": 4.817633985531924e-06,
+      "loss": 0.6338,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7574671445639187,
+      "grad_norm": 0.39311857332916716,
+      "learning_rate": 4.817503064808178e-06,
+      "loss": 0.6399,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7577326430373026,
+      "grad_norm": 0.37774626411213613,
+      "learning_rate": 4.8173720988874475e-06,
+      "loss": 0.6141,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7579981415106863,
+      "grad_norm": 0.4015965677893637,
+      "learning_rate": 4.817241087772287e-06,
+      "loss": 0.6217,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7582636399840701,
+      "grad_norm": 0.397607753910022,
+      "learning_rate": 4.817110031465252e-06,
+      "loss": 0.6223,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7585291384574538,
+      "grad_norm": 0.38283352590978686,
+      "learning_rate": 4.816978929968897e-06,
+      "loss": 0.6668,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7587946369308376,
+      "grad_norm": 0.3908908754959361,
+      "learning_rate": 4.816847783285782e-06,
+      "loss": 0.6343,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7590601354042215,
+      "grad_norm": 0.39230612259317565,
+      "learning_rate": 4.816716591418461e-06,
+      "loss": 0.6514,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7593256338776052,
+      "grad_norm": 0.4327028388240448,
+      "learning_rate": 4.816585354369494e-06,
+      "loss": 0.5903,
+      "step": 2860
+    },
+    {
+      "epoch": 0.759591132350989,
+      "grad_norm": 0.37954733886715813,
+      "learning_rate": 4.816454072141442e-06,
+      "loss": 0.6224,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7598566308243727,
+      "grad_norm": 0.40265519749244033,
+      "learning_rate": 4.816322744736863e-06,
+      "loss": 0.6228,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7601221292977566,
+      "grad_norm": 0.4099725995712369,
+      "learning_rate": 4.816191372158318e-06,
+      "loss": 0.6268,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7603876277711403,
+      "grad_norm": 0.4045979862349543,
+      "learning_rate": 4.816059954408372e-06,
+      "loss": 0.6428,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7606531262445241,
+      "grad_norm": 0.41086632410810203,
+      "learning_rate": 4.815928491489587e-06,
+      "loss": 0.6136,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7609186247179078,
+      "grad_norm": 0.3972164821152107,
+      "learning_rate": 4.815796983404524e-06,
+      "loss": 0.6267,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7611841231912917,
+      "grad_norm": 0.3902024330743728,
+      "learning_rate": 4.81566543015575e-06,
+      "loss": 0.5963,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7614496216646754,
+      "grad_norm": 0.40330659855741036,
+      "learning_rate": 4.8155338317458315e-06,
+      "loss": 0.5927,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7617151201380592,
+      "grad_norm": 0.38084131951564604,
+      "learning_rate": 4.815402188177333e-06,
+      "loss": 0.6365,
+      "step": 2869
+    },
+    {
+      "epoch": 0.761980618611443,
+      "grad_norm": 0.4221286842267405,
+      "learning_rate": 4.815270499452823e-06,
+      "loss": 0.6455,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7622461170848268,
+      "grad_norm": 0.4028283526570253,
+      "learning_rate": 4.815138765574869e-06,
+      "loss": 0.6418,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7625116155582106,
+      "grad_norm": 0.4105999221222162,
+      "learning_rate": 4.815006986546042e-06,
+      "loss": 0.6315,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7627771140315943,
+      "grad_norm": 0.3939269638636593,
+      "learning_rate": 4.81487516236891e-06,
+      "loss": 0.5824,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7630426125049781,
+      "grad_norm": 0.38876964883627213,
+      "learning_rate": 4.814743293046044e-06,
+      "loss": 0.6286,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7633081109783618,
+      "grad_norm": 0.3913028918686446,
+      "learning_rate": 4.814611378580016e-06,
+      "loss": 0.6384,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7635736094517457,
+      "grad_norm": 0.3811802259981648,
+      "learning_rate": 4.8144794189734e-06,
+      "loss": 0.6157,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7638391079251294,
+      "grad_norm": 0.39224969117165753,
+      "learning_rate": 4.814347414228768e-06,
+      "loss": 0.6242,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7641046063985132,
+      "grad_norm": 0.38710563455825564,
+      "learning_rate": 4.814215364348695e-06,
+      "loss": 0.6321,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7643701048718969,
+      "grad_norm": 0.40701079071417173,
+      "learning_rate": 4.814083269335757e-06,
+      "loss": 0.6321,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7646356033452808,
+      "grad_norm": 0.4154013311391526,
+      "learning_rate": 4.813951129192528e-06,
+      "loss": 0.6331,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7649011018186646,
+      "grad_norm": 0.3926129204062203,
+      "learning_rate": 4.813818943921587e-06,
+      "loss": 0.5864,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7651666002920483,
+      "grad_norm": 0.378144922737672,
+      "learning_rate": 4.813686713525512e-06,
+      "loss": 0.6198,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7654320987654321,
+      "grad_norm": 0.3874636487519552,
+      "learning_rate": 4.813554438006881e-06,
+      "loss": 0.6232,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7656975972388159,
+      "grad_norm": 0.3997421492654902,
+      "learning_rate": 4.813422117368273e-06,
+      "loss": 0.6425,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7659630957121997,
+      "grad_norm": 0.43148803941850866,
+      "learning_rate": 4.81328975161227e-06,
+      "loss": 0.6392,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7662285941855834,
+      "grad_norm": 0.39429568287485456,
+      "learning_rate": 4.8131573407414534e-06,
+      "loss": 0.6,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7664940926589672,
+      "grad_norm": 0.39363939075189897,
+      "learning_rate": 4.813024884758405e-06,
+      "loss": 0.594,
+      "step": 2887
+    },
+    {
+      "epoch": 0.766759591132351,
+      "grad_norm": 0.40246296057980246,
+      "learning_rate": 4.812892383665708e-06,
+      "loss": 0.6163,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7670250896057348,
+      "grad_norm": 0.41114581068698275,
+      "learning_rate": 4.812759837465946e-06,
+      "loss": 0.6502,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7672905880791185,
+      "grad_norm": 0.4113119839828481,
+      "learning_rate": 4.812627246161705e-06,
+      "loss": 0.6266,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7675560865525023,
+      "grad_norm": 0.3899321213433095,
+      "learning_rate": 4.812494609755571e-06,
+      "loss": 0.6638,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7678215850258862,
+      "grad_norm": 0.38660059296230714,
+      "learning_rate": 4.812361928250129e-06,
+      "loss": 0.5976,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7680870834992699,
+      "grad_norm": 0.38584939706221605,
+      "learning_rate": 4.812229201647967e-06,
+      "loss": 0.6211,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7683525819726537,
+      "grad_norm": 0.42629618970007627,
+      "learning_rate": 4.812096429951677e-06,
+      "loss": 0.5858,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7686180804460374,
+      "grad_norm": 0.3880003573730656,
+      "learning_rate": 4.811963613163844e-06,
+      "loss": 0.6134,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7688835789194212,
+      "grad_norm": 0.37682593883713944,
+      "learning_rate": 4.81183075128706e-06,
+      "loss": 0.6453,
+      "step": 2896
+    },
+    {
+      "epoch": 0.769149077392805,
+      "grad_norm": 0.39006645981172405,
+      "learning_rate": 4.811697844323916e-06,
+      "loss": 0.6055,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7694145758661888,
+      "grad_norm": 0.3858947263046375,
+      "learning_rate": 4.811564892277003e-06,
+      "loss": 0.6142,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7696800743395725,
+      "grad_norm": 0.4054651077962247,
+      "learning_rate": 4.811431895148917e-06,
+      "loss": 0.6412,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7699455728129563,
+      "grad_norm": 0.39216972756435564,
+      "learning_rate": 4.811298852942248e-06,
+      "loss": 0.6252,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7702110712863401,
+      "grad_norm": 0.39892334901550697,
+      "learning_rate": 4.811165765659593e-06,
+      "loss": 0.6281,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7704765697597239,
+      "grad_norm": 0.38647458628122094,
+      "learning_rate": 4.811032633303547e-06,
+      "loss": 0.6486,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7707420682331076,
+      "grad_norm": 0.388502486175942,
+      "learning_rate": 4.810899455876706e-06,
+      "loss": 0.5853,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7710075667064914,
+      "grad_norm": 0.38794512919805835,
+      "learning_rate": 4.8107662333816675e-06,
+      "loss": 0.6283,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7712730651798753,
+      "grad_norm": 0.378637211836451,
+      "learning_rate": 4.81063296582103e-06,
+      "loss": 0.5796,
+      "step": 2905
+    },
+    {
+      "epoch": 0.771538563653259,
+      "grad_norm": 0.3895036340486329,
+      "learning_rate": 4.810499653197393e-06,
+      "loss": 0.5938,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7718040621266428,
+      "grad_norm": 0.39562898790582673,
+      "learning_rate": 4.8103662955133544e-06,
+      "loss": 0.6298,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7720695606000265,
+      "grad_norm": 0.3873624183380144,
+      "learning_rate": 4.810232892771516e-06,
+      "loss": 0.6062,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7723350590734104,
+      "grad_norm": 0.3930627184834584,
+      "learning_rate": 4.810099444974481e-06,
+      "loss": 0.6229,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7726005575467941,
+      "grad_norm": 0.38845285356878767,
+      "learning_rate": 4.80996595212485e-06,
+      "loss": 0.616,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7728660560201779,
+      "grad_norm": 0.4057450413210804,
+      "learning_rate": 4.809832414225227e-06,
+      "loss": 0.6203,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7731315544935616,
+      "grad_norm": 0.39552285158857,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.6509,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7733970529669454,
+      "grad_norm": 0.39503753899891747,
+      "learning_rate": 4.809565203286425e-06,
+      "loss": 0.6496,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7736625514403292,
+      "grad_norm": 0.3881227733150097,
+      "learning_rate": 4.809431530252456e-06,
+      "loss": 0.6337,
+      "step": 2914
+    },
+    {
+      "epoch": 0.773928049913713,
+      "grad_norm": 0.39464067899109084,
+      "learning_rate": 4.809297812178918e-06,
+      "loss": 0.5801,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.4063830918132964,
+      "learning_rate": 4.809164049068419e-06,
+      "loss": 0.5893,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7744590468604805,
+      "grad_norm": 0.4114952777520471,
+      "learning_rate": 4.809030240923566e-06,
+      "loss": 0.6089,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7747245453338644,
+      "grad_norm": 0.385084434659045,
+      "learning_rate": 4.808896387746971e-06,
+      "loss": 0.619,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7749900438072481,
+      "grad_norm": 0.38481500748273706,
+      "learning_rate": 4.808762489541242e-06,
+      "loss": 0.6394,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7752555422806319,
+      "grad_norm": 0.3945167554852316,
+      "learning_rate": 4.808628546308993e-06,
+      "loss": 0.6578,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7755210407540156,
+      "grad_norm": 0.3866927988729723,
+      "learning_rate": 4.808494558052834e-06,
+      "loss": 0.5707,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7757865392273995,
+      "grad_norm": 0.3927322070675447,
+      "learning_rate": 4.80836052477538e-06,
+      "loss": 0.615,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7760520377007832,
+      "grad_norm": 0.3790505710213684,
+      "learning_rate": 4.808226446479242e-06,
+      "loss": 0.573,
+      "step": 2923
+    },
+    {
+      "epoch": 0.776317536174167,
+      "grad_norm": 0.3855940690441728,
+      "learning_rate": 4.808092323167038e-06,
+      "loss": 0.6013,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7765830346475507,
+      "grad_norm": 0.38730582839123984,
+      "learning_rate": 4.807958154841383e-06,
+      "loss": 0.6087,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7768485331209346,
+      "grad_norm": 0.3768157541034743,
+      "learning_rate": 4.807823941504893e-06,
+      "loss": 0.6005,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7771140315943184,
+      "grad_norm": 0.3878411660495298,
+      "learning_rate": 4.807689683160185e-06,
+      "loss": 0.6224,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7773795300677021,
+      "grad_norm": 0.3811076776944274,
+      "learning_rate": 4.807555379809878e-06,
+      "loss": 0.5965,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7776450285410859,
+      "grad_norm": 0.39920255595996806,
+      "learning_rate": 4.807421031456592e-06,
+      "loss": 0.6436,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7779105270144697,
+      "grad_norm": 0.3815713523887147,
+      "learning_rate": 4.807286638102945e-06,
+      "loss": 0.5779,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7781760254878535,
+      "grad_norm": 0.39670412837126307,
+      "learning_rate": 4.80715219975156e-06,
+      "loss": 0.6039,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7784415239612372,
+      "grad_norm": 0.3994763151178164,
+      "learning_rate": 4.807017716405058e-06,
+      "loss": 0.633,
+      "step": 2932
+    },
+    {
+      "epoch": 0.778707022434621,
+      "grad_norm": 0.3816882153271918,
+      "learning_rate": 4.806883188066063e-06,
+      "loss": 0.6364,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7789725209080047,
+      "grad_norm": 0.3994844839005148,
+      "learning_rate": 4.806748614737197e-06,
+      "loss": 0.6462,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7792380193813886,
+      "grad_norm": 0.3899630272401776,
+      "learning_rate": 4.806613996421085e-06,
+      "loss": 0.6274,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7795035178547723,
+      "grad_norm": 0.39350037441691904,
+      "learning_rate": 4.806479333120352e-06,
+      "loss": 0.6488,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7797690163281561,
+      "grad_norm": 0.39549162194149007,
+      "learning_rate": 4.806344624837626e-06,
+      "loss": 0.5834,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7800345148015398,
+      "grad_norm": 0.3902836543896936,
+      "learning_rate": 4.8062098715755315e-06,
+      "loss": 0.6165,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7803000132749237,
+      "grad_norm": 0.37882458615267006,
+      "learning_rate": 4.806075073336699e-06,
+      "loss": 0.6119,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7805655117483075,
+      "grad_norm": 0.3945470338531427,
+      "learning_rate": 4.805940230123755e-06,
+      "loss": 0.5608,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7808310102216912,
+      "grad_norm": 0.38574600185236063,
+      "learning_rate": 4.805805341939333e-06,
+      "loss": 0.6063,
+      "step": 2941
+    },
+    {
+      "epoch": 0.781096508695075,
+      "grad_norm": 0.3932120520679073,
+      "learning_rate": 4.805670408786059e-06,
+      "loss": 0.6494,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7813620071684588,
+      "grad_norm": 0.387579262601144,
+      "learning_rate": 4.805535430666568e-06,
+      "loss": 0.6258,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7816275056418426,
+      "grad_norm": 0.38110120339450104,
+      "learning_rate": 4.805400407583491e-06,
+      "loss": 0.6331,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7818930041152263,
+      "grad_norm": 0.39119423003978865,
+      "learning_rate": 4.805265339539461e-06,
+      "loss": 0.6603,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7821585025886101,
+      "grad_norm": 0.3810595461997376,
+      "learning_rate": 4.805130226537112e-06,
+      "loss": 0.6058,
+      "step": 2946
+    },
+    {
+      "epoch": 0.782424001061994,
+      "grad_norm": 0.37985340674377616,
+      "learning_rate": 4.804995068579082e-06,
+      "loss": 0.6855,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7826894995353777,
+      "grad_norm": 0.3964583029794383,
+      "learning_rate": 4.804859865668002e-06,
+      "loss": 0.6404,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7829549980087614,
+      "grad_norm": 0.3906598558363303,
+      "learning_rate": 4.804724617806512e-06,
+      "loss": 0.6937,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7832204964821452,
+      "grad_norm": 0.3827250015372093,
+      "learning_rate": 4.8045893249972495e-06,
+      "loss": 0.6098,
+      "step": 2950
+    },
+    {
+      "epoch": 0.783485994955529,
+      "grad_norm": 0.3957584088405087,
+      "learning_rate": 4.804453987242853e-06,
+      "loss": 0.6129,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7837514934289128,
+      "grad_norm": 0.37950326853201494,
+      "learning_rate": 4.804318604545961e-06,
+      "loss": 0.641,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7840169919022966,
+      "grad_norm": 0.3822250393835054,
+      "learning_rate": 4.804183176909214e-06,
+      "loss": 0.6068,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7842824903756803,
+      "grad_norm": 0.3848591378719966,
+      "learning_rate": 4.804047704335254e-06,
+      "loss": 0.6412,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7845479888490641,
+      "grad_norm": 0.3881268395726478,
+      "learning_rate": 4.803912186826722e-06,
+      "loss": 0.6312,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7848134873224479,
+      "grad_norm": 0.39554484008479746,
+      "learning_rate": 4.803776624386262e-06,
+      "loss": 0.6383,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7850789857958317,
+      "grad_norm": 0.3763153558720562,
+      "learning_rate": 4.803641017016516e-06,
+      "loss": 0.6031,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7853444842692154,
+      "grad_norm": 0.39052628837475367,
+      "learning_rate": 4.803505364720131e-06,
+      "loss": 0.63,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7856099827425992,
+      "grad_norm": 0.36963789589009133,
+      "learning_rate": 4.803369667499751e-06,
+      "loss": 0.6048,
+      "step": 2959
+    },
+    {
+      "epoch": 0.785875481215983,
+      "grad_norm": 0.3815686570433188,
+      "learning_rate": 4.803233925358023e-06,
+      "loss": 0.6339,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7861409796893668,
+      "grad_norm": 0.46078915740846943,
+      "learning_rate": 4.803098138297594e-06,
+      "loss": 0.576,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7864064781627506,
+      "grad_norm": 0.39342741939552883,
+      "learning_rate": 4.802962306321113e-06,
+      "loss": 0.6011,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7866719766361343,
+      "grad_norm": 0.39106682727992903,
+      "learning_rate": 4.802826429431227e-06,
+      "loss": 0.6083,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7869374751095182,
+      "grad_norm": 0.40584316504966633,
+      "learning_rate": 4.802690507630588e-06,
+      "loss": 0.6072,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7872029735829019,
+      "grad_norm": 0.3804267793705631,
+      "learning_rate": 4.8025545409218465e-06,
+      "loss": 0.6144,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7874684720562857,
+      "grad_norm": 0.4229561074639001,
+      "learning_rate": 4.802418529307654e-06,
+      "loss": 0.6032,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7877339705296694,
+      "grad_norm": 0.41496474949956363,
+      "learning_rate": 4.802282472790663e-06,
+      "loss": 0.6078,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7879994690030533,
+      "grad_norm": 0.4162004586289964,
+      "learning_rate": 4.802146371373525e-06,
+      "loss": 0.6331,
+      "step": 2968
+    },
+    {
+      "epoch": 0.788264967476437,
+      "grad_norm": 0.4515606887735363,
+      "learning_rate": 4.802010225058898e-06,
+      "loss": 0.5855,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7885304659498208,
+      "grad_norm": 0.3960843628246435,
+      "learning_rate": 4.801874033849435e-06,
+      "loss": 0.619,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7887959644232045,
+      "grad_norm": 0.4092054627378958,
+      "learning_rate": 4.801737797747792e-06,
+      "loss": 0.6023,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7890614628965883,
+      "grad_norm": 0.41951841062801887,
+      "learning_rate": 4.8016015167566265e-06,
+      "loss": 0.6289,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7893269613699722,
+      "grad_norm": 0.4559277053067011,
+      "learning_rate": 4.801465190878596e-06,
+      "loss": 0.6386,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7895924598433559,
+      "grad_norm": 0.38011225556715383,
+      "learning_rate": 4.80132882011636e-06,
+      "loss": 0.6047,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7898579583167397,
+      "grad_norm": 0.3932016851471219,
+      "learning_rate": 4.8011924044725765e-06,
+      "loss": 0.6126,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7901234567901234,
+      "grad_norm": 0.38368297969397236,
+      "learning_rate": 4.801055943949907e-06,
+      "loss": 0.624,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7903889552635073,
+      "grad_norm": 0.39175187781741116,
+      "learning_rate": 4.800919438551012e-06,
+      "loss": 0.6221,
+      "step": 2977
+    },
+    {
+      "epoch": 0.790654453736891,
+      "grad_norm": 0.3904633241427886,
+      "learning_rate": 4.800782888278556e-06,
+      "loss": 0.6207,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7909199522102748,
+      "grad_norm": 0.38512573140653733,
+      "learning_rate": 4.8006462931351995e-06,
+      "loss": 0.591,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7911854506836585,
+      "grad_norm": 0.40702726558338137,
+      "learning_rate": 4.800509653123607e-06,
+      "loss": 0.6233,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7914509491570424,
+      "grad_norm": 0.3901216916858801,
+      "learning_rate": 4.800372968246444e-06,
+      "loss": 0.6272,
+      "step": 2981
+    },
+    {
+      "epoch": 0.7917164476304261,
+      "grad_norm": 0.38754953990316465,
+      "learning_rate": 4.800236238506376e-06,
+      "loss": 0.6047,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7919819461038099,
+      "grad_norm": 0.40160248185350533,
+      "learning_rate": 4.80009946390607e-06,
+      "loss": 0.6607,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7922474445771936,
+      "grad_norm": 0.3824913621186493,
+      "learning_rate": 4.799962644448191e-06,
+      "loss": 0.6089,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7925129430505775,
+      "grad_norm": 0.38890634812615466,
+      "learning_rate": 4.799825780135411e-06,
+      "loss": 0.593,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7927784415239613,
+      "grad_norm": 0.38539566727529206,
+      "learning_rate": 4.799688870970396e-06,
+      "loss": 0.6189,
+      "step": 2986
+    },
+    {
+      "epoch": 0.793043939997345,
+      "grad_norm": 0.38677878228576895,
+      "learning_rate": 4.799551916955818e-06,
+      "loss": 0.6331,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7933094384707288,
+      "grad_norm": 0.40154225892810613,
+      "learning_rate": 4.799414918094347e-06,
+      "loss": 0.6414,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7935749369441125,
+      "grad_norm": 0.4186541670024924,
+      "learning_rate": 4.799277874388656e-06,
+      "loss": 0.6111,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7938404354174964,
+      "grad_norm": 0.37130180790833417,
+      "learning_rate": 4.799140785841417e-06,
+      "loss": 0.6013,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7941059338908801,
+      "grad_norm": 0.3745858589150191,
+      "learning_rate": 4.799003652455302e-06,
+      "loss": 0.5941,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7943714323642639,
+      "grad_norm": 0.38701578660458447,
+      "learning_rate": 4.7988664742329875e-06,
+      "loss": 0.6069,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7946369308376476,
+      "grad_norm": 0.3941868397833378,
+      "learning_rate": 4.7987292511771474e-06,
+      "loss": 0.6286,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7949024293110315,
+      "grad_norm": 0.38929898414732367,
+      "learning_rate": 4.798591983290461e-06,
+      "loss": 0.6376,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7951679277844153,
+      "grad_norm": 0.39449662455560097,
+      "learning_rate": 4.7984546705756e-06,
+      "loss": 0.6434,
+      "step": 2995
+    },
+    {
+      "epoch": 0.795433426257799,
+      "grad_norm": 0.3778611954502537,
+      "learning_rate": 4.7983173130352475e-06,
+      "loss": 0.5915,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7956989247311828,
+      "grad_norm": 0.4011326295665181,
+      "learning_rate": 4.798179910672079e-06,
+      "loss": 0.6001,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7959644232045666,
+      "grad_norm": 0.382780846431489,
+      "learning_rate": 4.798042463488775e-06,
+      "loss": 0.6339,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7962299216779504,
+      "grad_norm": 0.38543224192435804,
+      "learning_rate": 4.797904971488019e-06,
+      "loss": 0.6133,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7964954201513341,
+      "grad_norm": 0.37802472578784047,
+      "learning_rate": 4.797767434672488e-06,
+      "loss": 0.6252,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7967609186247179,
+      "grad_norm": 0.38678043336886053,
+      "learning_rate": 4.797629853044865e-06,
+      "loss": 0.6455,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7970264170981017,
+      "grad_norm": 0.40459052522651207,
+      "learning_rate": 4.797492226607836e-06,
+      "loss": 0.6175,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7972919155714855,
+      "grad_norm": 0.37964100233076326,
+      "learning_rate": 4.797354555364082e-06,
+      "loss": 0.6287,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7975574140448692,
+      "grad_norm": 0.38341809554558787,
+      "learning_rate": 4.797216839316291e-06,
+      "loss": 0.6112,
+      "step": 3004
+    },
+    {
+      "epoch": 0.797822912518253,
+      "grad_norm": 0.3923328729463677,
+      "learning_rate": 4.797079078467145e-06,
+      "loss": 0.6249,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7980884109916367,
+      "grad_norm": 0.4243101334360585,
+      "learning_rate": 4.7969412728193345e-06,
+      "loss": 0.6099,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7983539094650206,
+      "grad_norm": 0.39538420941156294,
+      "learning_rate": 4.796803422375544e-06,
+      "loss": 0.5722,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7986194079384044,
+      "grad_norm": 0.3942206599636837,
+      "learning_rate": 4.796665527138465e-06,
+      "loss": 0.6254,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7988849064117881,
+      "grad_norm": 0.38938382369631763,
+      "learning_rate": 4.796527587110783e-06,
+      "loss": 0.6455,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7991504048851719,
+      "grad_norm": 0.41905754132214856,
+      "learning_rate": 4.7963896022951915e-06,
+      "loss": 0.6177,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7994159033585557,
+      "grad_norm": 0.4065932144561058,
+      "learning_rate": 4.796251572694379e-06,
+      "loss": 0.6143,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7996814018319395,
+      "grad_norm": 0.37800178823170866,
+      "learning_rate": 4.79611349831104e-06,
+      "loss": 0.605,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7999469003053232,
+      "grad_norm": 0.40352141037193645,
+      "learning_rate": 4.795975379147865e-06,
+      "loss": 0.6152,
+      "step": 3013
+    },
+    {
+      "epoch": 0.800212398778707,
+      "grad_norm": 0.4275551389495503,
+      "learning_rate": 4.795837215207549e-06,
+      "loss": 0.5919,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8004778972520908,
+      "grad_norm": 0.3963273680155388,
+      "learning_rate": 4.7956990064927865e-06,
+      "loss": 0.6397,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8007433957254746,
+      "grad_norm": 0.37794494686677665,
+      "learning_rate": 4.795560753006272e-06,
+      "loss": 0.5899,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8010088941988583,
+      "grad_norm": 0.3839474054456509,
+      "learning_rate": 4.795422454750703e-06,
+      "loss": 0.6336,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8012743926722421,
+      "grad_norm": 0.4085013943268837,
+      "learning_rate": 4.7952841117287746e-06,
+      "loss": 0.6243,
+      "step": 3018
+    },
+    {
+      "epoch": 0.801539891145626,
+      "grad_norm": 0.4033356716760012,
+      "learning_rate": 4.795145723943187e-06,
+      "loss": 0.601,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8018053896190097,
+      "grad_norm": 0.3801297064742477,
+      "learning_rate": 4.7950072913966386e-06,
+      "loss": 0.6081,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8020708880923935,
+      "grad_norm": 0.40534575296039577,
+      "learning_rate": 4.794868814091828e-06,
+      "loss": 0.627,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8023363865657772,
+      "grad_norm": 0.4170024787657662,
+      "learning_rate": 4.794730292031457e-06,
+      "loss": 0.6199,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8026018850391611,
+      "grad_norm": 0.3877307879508813,
+      "learning_rate": 4.794591725218227e-06,
+      "loss": 0.5959,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8028673835125448,
+      "grad_norm": 0.385260307210276,
+      "learning_rate": 4.794453113654841e-06,
+      "loss": 0.6484,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8031328819859286,
+      "grad_norm": 0.3958538084046821,
+      "learning_rate": 4.794314457344001e-06,
+      "loss": 0.6455,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8033983804593123,
+      "grad_norm": 0.49005574904683974,
+      "learning_rate": 4.794175756288411e-06,
+      "loss": 0.6302,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8036638789326961,
+      "grad_norm": 0.38787925883372026,
+      "learning_rate": 4.794037010490777e-06,
+      "loss": 0.6124,
+      "step": 3027
+    },
+    {
+      "epoch": 0.80392937740608,
+      "grad_norm": 0.41030266965415774,
+      "learning_rate": 4.793898219953804e-06,
+      "loss": 0.6451,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8041948758794637,
+      "grad_norm": 0.4338130007013215,
+      "learning_rate": 4.7937593846802e-06,
+      "loss": 0.6171,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8044603743528475,
+      "grad_norm": 0.4287021615609794,
+      "learning_rate": 4.793620504672673e-06,
+      "loss": 0.5859,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8047258728262312,
+      "grad_norm": 0.3792493028037663,
+      "learning_rate": 4.7934815799339285e-06,
+      "loss": 0.6322,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8049913712996151,
+      "grad_norm": 0.44063062639660205,
+      "learning_rate": 4.793342610466678e-06,
+      "loss": 0.5823,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8052568697729988,
+      "grad_norm": 0.44977973699506585,
+      "learning_rate": 4.793203596273632e-06,
+      "loss": 0.5843,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8055223682463826,
+      "grad_norm": 0.39183705702570304,
+      "learning_rate": 4.793064537357502e-06,
+      "loss": 0.6619,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8057878667197663,
+      "grad_norm": 0.3853020531081753,
+      "learning_rate": 4.792925433720998e-06,
+      "loss": 0.6281,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8060533651931502,
+      "grad_norm": 0.40941377101087645,
+      "learning_rate": 4.792786285366834e-06,
+      "loss": 0.6207,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8063188636665339,
+      "grad_norm": 0.39984320559078634,
+      "learning_rate": 4.792647092297724e-06,
+      "loss": 0.6414,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8065843621399177,
+      "grad_norm": 0.38670405384388634,
+      "learning_rate": 4.792507854516383e-06,
+      "loss": 0.6016,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8068498606133014,
+      "grad_norm": 0.38147800309502916,
+      "learning_rate": 4.792368572025525e-06,
+      "loss": 0.6208,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8071153590866853,
+      "grad_norm": 0.386319652032236,
+      "learning_rate": 4.792229244827867e-06,
+      "loss": 0.5851,
+      "step": 3040
+    },
+    {
+      "epoch": 0.807380857560069,
+      "grad_norm": 0.4053514067391913,
+      "learning_rate": 4.792089872926127e-06,
+      "loss": 0.5931,
+      "step": 3041
+    },
+    {
+      "epoch": 0.8076463560334528,
+      "grad_norm": 0.3863874156057099,
+      "learning_rate": 4.791950456323023e-06,
+      "loss": 0.6266,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8079118545068366,
+      "grad_norm": 0.394406759418258,
+      "learning_rate": 4.791810995021271e-06,
+      "loss": 0.6011,
+      "step": 3043
+    },
+    {
+      "epoch": 0.8081773529802203,
+      "grad_norm": 0.3934387726763986,
+      "learning_rate": 4.791671489023595e-06,
+      "loss": 0.6039,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8084428514536042,
+      "grad_norm": 0.39613581161129774,
+      "learning_rate": 4.791531938332714e-06,
+      "loss": 0.6487,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8087083499269879,
+      "grad_norm": 0.3826439695550616,
+      "learning_rate": 4.79139234295135e-06,
+      "loss": 0.6022,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8089738484003717,
+      "grad_norm": 0.38489715956190756,
+      "learning_rate": 4.7912527028822246e-06,
+      "loss": 0.6346,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8092393468737554,
+      "grad_norm": 0.40076861754465976,
+      "learning_rate": 4.7911130181280605e-06,
+      "loss": 0.6041,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8095048453471393,
+      "grad_norm": 0.38486072268242794,
+      "learning_rate": 4.790973288691585e-06,
+      "loss": 0.6315,
+      "step": 3049
+    },
+    {
+      "epoch": 0.809770343820523,
+      "grad_norm": 0.3890072312613263,
+      "learning_rate": 4.790833514575519e-06,
+      "loss": 0.6338,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8100358422939068,
+      "grad_norm": 0.40063802865243636,
+      "learning_rate": 4.790693695782592e-06,
+      "loss": 0.6212,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8103013407672905,
+      "grad_norm": 0.3805255829024729,
+      "learning_rate": 4.790553832315529e-06,
+      "loss": 0.6088,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8105668392406744,
+      "grad_norm": 0.38052266688049624,
+      "learning_rate": 4.790413924177058e-06,
+      "loss": 0.6271,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8108323377140582,
+      "grad_norm": 0.3891470925112324,
+      "learning_rate": 4.790273971369906e-06,
+      "loss": 0.6033,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8110978361874419,
+      "grad_norm": 0.39702011828000344,
+      "learning_rate": 4.790133973896806e-06,
+      "loss": 0.6173,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8113633346608257,
+      "grad_norm": 0.3871313317783108,
+      "learning_rate": 4.789993931760486e-06,
+      "loss": 0.6416,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8116288331342095,
+      "grad_norm": 0.3881859027432485,
+      "learning_rate": 4.789853844963677e-06,
+      "loss": 0.6218,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8118943316075933,
+      "grad_norm": 0.3868948385668171,
+      "learning_rate": 4.789713713509112e-06,
+      "loss": 0.6059,
+      "step": 3058
+    },
+    {
+      "epoch": 0.812159830080977,
+      "grad_norm": 0.40151321349355845,
+      "learning_rate": 4.7895735373995225e-06,
+      "loss": 0.5989,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8124253285543608,
+      "grad_norm": 0.37704399463899213,
+      "learning_rate": 4.789433316637644e-06,
+      "loss": 0.6125,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8126908270277446,
+      "grad_norm": 0.3837903681264029,
+      "learning_rate": 4.789293051226211e-06,
+      "loss": 0.6498,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8129563255011284,
+      "grad_norm": 0.4075595595495999,
+      "learning_rate": 4.789152741167956e-06,
+      "loss": 0.5957,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8132218239745121,
+      "grad_norm": 0.40736896177681364,
+      "learning_rate": 4.78901238646562e-06,
+      "loss": 0.6468,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8134873224478959,
+      "grad_norm": 0.3881459920702017,
+      "learning_rate": 4.788871987121937e-06,
+      "loss": 0.598,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8137528209212797,
+      "grad_norm": 0.38756965470984517,
+      "learning_rate": 4.788731543139646e-06,
+      "loss": 0.6308,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8140183193946635,
+      "grad_norm": 0.38766933747438825,
+      "learning_rate": 4.788591054521486e-06,
+      "loss": 0.6319,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8142838178680473,
+      "grad_norm": 0.4253101243468969,
+      "learning_rate": 4.788450521270198e-06,
+      "loss": 0.5777,
+      "step": 3067
+    },
+    {
+      "epoch": 0.814549316341431,
+      "grad_norm": 0.3811032855153689,
+      "learning_rate": 4.78830994338852e-06,
+      "loss": 0.6002,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 0.387554693583662,
+      "learning_rate": 4.788169320879197e-06,
+      "loss": 0.6345,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8150803132881986,
+      "grad_norm": 0.40243570860162275,
+      "learning_rate": 4.788028653744969e-06,
+      "loss": 0.613,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8153458117615824,
+      "grad_norm": 0.3985264563294363,
+      "learning_rate": 4.787887941988581e-06,
+      "loss": 0.6387,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8156113102349661,
+      "grad_norm": 0.3927562343922202,
+      "learning_rate": 4.787747185612776e-06,
+      "loss": 0.6154,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8158768087083499,
+      "grad_norm": 0.38124945600392735,
+      "learning_rate": 4.7876063846203e-06,
+      "loss": 0.6446,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8161423071817338,
+      "grad_norm": 0.40006847227750797,
+      "learning_rate": 4.787465539013897e-06,
+      "loss": 0.6038,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8164078056551175,
+      "grad_norm": 0.39533456184373295,
+      "learning_rate": 4.787324648796316e-06,
+      "loss": 0.6187,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8166733041285013,
+      "grad_norm": 0.4061119856435276,
+      "learning_rate": 4.787183713970304e-06,
+      "loss": 0.5932,
+      "step": 3076
+    },
+    {
+      "epoch": 0.816938802601885,
+      "grad_norm": 0.38498866901121387,
+      "learning_rate": 4.787042734538611e-06,
+      "loss": 0.586,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8172043010752689,
+      "grad_norm": 0.38604882667142726,
+      "learning_rate": 4.786901710503984e-06,
+      "loss": 0.623,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8174697995486526,
+      "grad_norm": 0.39481088801397074,
+      "learning_rate": 4.786760641869174e-06,
+      "loss": 0.6211,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8177352980220364,
+      "grad_norm": 0.38895916363318683,
+      "learning_rate": 4.786619528636933e-06,
+      "loss": 0.633,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8180007964954201,
+      "grad_norm": 0.390630967657958,
+      "learning_rate": 4.786478370810012e-06,
+      "loss": 0.6079,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8182662949688039,
+      "grad_norm": 0.3998832046417734,
+      "learning_rate": 4.7863371683911644e-06,
+      "loss": 0.5956,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8185317934421877,
+      "grad_norm": 0.4311476377342335,
+      "learning_rate": 4.786195921383145e-06,
+      "loss": 0.5949,
+      "step": 3083
+    },
+    {
+      "epoch": 0.8187972919155715,
+      "grad_norm": 0.3999998209449262,
+      "learning_rate": 4.7860546297887065e-06,
+      "loss": 0.6349,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8190627903889552,
+      "grad_norm": 0.38991995662173307,
+      "learning_rate": 4.785913293610607e-06,
+      "loss": 0.564,
+      "step": 3085
+    },
+    {
+      "epoch": 0.819328288862339,
+      "grad_norm": 0.4045768827897694,
+      "learning_rate": 4.7857719128516e-06,
+      "loss": 0.6189,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8195937873357229,
+      "grad_norm": 0.39984430576344626,
+      "learning_rate": 4.785630487514444e-06,
+      "loss": 0.5802,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8198592858091066,
+      "grad_norm": 0.3852432445950339,
+      "learning_rate": 4.785489017601899e-06,
+      "loss": 0.6612,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8201247842824904,
+      "grad_norm": 0.3818704358266958,
+      "learning_rate": 4.785347503116721e-06,
+      "loss": 0.5754,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8203902827558741,
+      "grad_norm": 0.3681576688362433,
+      "learning_rate": 4.7852059440616715e-06,
+      "loss": 0.627,
+      "step": 3090
+    },
+    {
+      "epoch": 0.820655781229258,
+      "grad_norm": 0.393464260171437,
+      "learning_rate": 4.78506434043951e-06,
+      "loss": 0.6264,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8209212797026417,
+      "grad_norm": 0.378244620539641,
+      "learning_rate": 4.7849226922530006e-06,
+      "loss": 0.6095,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8211867781760255,
+      "grad_norm": 0.38019313440902974,
+      "learning_rate": 4.784780999504903e-06,
+      "loss": 0.6331,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8214522766494092,
+      "grad_norm": 0.4074252468749414,
+      "learning_rate": 4.784639262197983e-06,
+      "loss": 0.6605,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8217177751227931,
+      "grad_norm": 0.38585300792180804,
+      "learning_rate": 4.784497480335003e-06,
+      "loss": 0.6291,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8219832735961768,
+      "grad_norm": 0.3876931187738847,
+      "learning_rate": 4.784355653918729e-06,
+      "loss": 0.6453,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8222487720695606,
+      "grad_norm": 0.36332652523024367,
+      "learning_rate": 4.784213782951926e-06,
+      "loss": 0.5722,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8225142705429443,
+      "grad_norm": 0.3887959834336381,
+      "learning_rate": 4.784071867437362e-06,
+      "loss": 0.5863,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8227797690163282,
+      "grad_norm": 0.38212226745974903,
+      "learning_rate": 4.783929907377805e-06,
+      "loss": 0.5543,
+      "step": 3099
+    },
+    {
+      "epoch": 0.823045267489712,
+      "grad_norm": 0.38555522311114687,
+      "learning_rate": 4.783787902776021e-06,
+      "loss": 0.6055,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8233107659630957,
+      "grad_norm": 0.38390738971620475,
+      "learning_rate": 4.783645853634782e-06,
+      "loss": 0.6126,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8235762644364795,
+      "grad_norm": 0.390567561772916,
+      "learning_rate": 4.783503759956858e-06,
+      "loss": 0.6273,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8238417629098632,
+      "grad_norm": 0.38487795571260974,
+      "learning_rate": 4.78336162174502e-06,
+      "loss": 0.6588,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8241072613832471,
+      "grad_norm": 0.38965158774468656,
+      "learning_rate": 4.783219439002038e-06,
+      "loss": 0.5916,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8243727598566308,
+      "grad_norm": 0.37582919918188046,
+      "learning_rate": 4.7830772117306875e-06,
+      "loss": 0.5936,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8246382583300146,
+      "grad_norm": 0.38482312988662143,
+      "learning_rate": 4.782934939933742e-06,
+      "loss": 0.6116,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8249037568033983,
+      "grad_norm": 0.4039339620970389,
+      "learning_rate": 4.7827926236139754e-06,
+      "loss": 0.5906,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8251692552767822,
+      "grad_norm": 0.40081933198032615,
+      "learning_rate": 4.782650262774164e-06,
+      "loss": 0.6003,
+      "step": 3108
+    },
+    {
+      "epoch": 0.825434753750166,
+      "grad_norm": 0.3776685090144088,
+      "learning_rate": 4.782507857417083e-06,
+      "loss": 0.5958,
+      "step": 3109
+    },
+    {
+      "epoch": 0.8257002522235497,
+      "grad_norm": 0.38276870586488,
+      "learning_rate": 4.782365407545509e-06,
+      "loss": 0.6117,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8259657506969335,
+      "grad_norm": 0.3884070994966659,
+      "learning_rate": 4.782222913162223e-06,
+      "loss": 0.5878,
+      "step": 3111
+    },
+    {
+      "epoch": 0.8262312491703173,
+      "grad_norm": 0.39344114122789803,
+      "learning_rate": 4.782080374270002e-06,
+      "loss": 0.612,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8264967476437011,
+      "grad_norm": 0.3851587563270667,
+      "learning_rate": 4.781937790871626e-06,
+      "loss": 0.5895,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8267622461170848,
+      "grad_norm": 0.39150494638776406,
+      "learning_rate": 4.781795162969875e-06,
+      "loss": 0.593,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8270277445904686,
+      "grad_norm": 0.3872467177140655,
+      "learning_rate": 4.781652490567533e-06,
+      "loss": 0.6437,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8272932430638524,
+      "grad_norm": 0.3865888914631022,
+      "learning_rate": 4.78150977366738e-06,
+      "loss": 0.6057,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8275587415372362,
+      "grad_norm": 0.3980511947596024,
+      "learning_rate": 4.7813670122722015e-06,
+      "loss": 0.6116,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8278242400106199,
+      "grad_norm": 0.37673976668592946,
+      "learning_rate": 4.781224206384779e-06,
+      "loss": 0.608,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8280897384840037,
+      "grad_norm": 0.38914364170088517,
+      "learning_rate": 4.781081356007901e-06,
+      "loss": 0.6177,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8283552369573874,
+      "grad_norm": 0.3962658556356825,
+      "learning_rate": 4.7809384611443496e-06,
+      "loss": 0.6527,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8286207354307713,
+      "grad_norm": 0.38940717741034075,
+      "learning_rate": 4.780795521796914e-06,
+      "loss": 0.6262,
+      "step": 3121
+    },
+    {
+      "epoch": 0.8288862339041551,
+      "grad_norm": 0.40627445492556524,
+      "learning_rate": 4.780652537968382e-06,
+      "loss": 0.6396,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8291517323775388,
+      "grad_norm": 0.3889625376418175,
+      "learning_rate": 4.780509509661541e-06,
+      "loss": 0.6138,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8294172308509226,
+      "grad_norm": 0.394228670778925,
+      "learning_rate": 4.780366436879181e-06,
+      "loss": 0.6125,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8296827293243064,
+      "grad_norm": 0.38510912767694816,
+      "learning_rate": 4.780223319624092e-06,
+      "loss": 0.6233,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8299482277976902,
+      "grad_norm": 0.3905933874343222,
+      "learning_rate": 4.780080157899066e-06,
+      "loss": 0.629,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8302137262710739,
+      "grad_norm": 0.40519786211821524,
+      "learning_rate": 4.7799369517068935e-06,
+      "loss": 0.6182,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8304792247444577,
+      "grad_norm": 0.4109984779764674,
+      "learning_rate": 4.779793701050369e-06,
+      "loss": 0.6109,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8307447232178415,
+      "grad_norm": 0.3799831099592614,
+      "learning_rate": 4.779650405932285e-06,
+      "loss": 0.6174,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8310102216912253,
+      "grad_norm": 0.3932220390496914,
+      "learning_rate": 4.779507066355437e-06,
+      "loss": 0.6151,
+      "step": 3130
+    },
+    {
+      "epoch": 0.831275720164609,
+      "grad_norm": 0.3866879722158227,
+      "learning_rate": 4.779363682322619e-06,
+      "loss": 0.6373,
+      "step": 3131
+    },
+    {
+      "epoch": 0.8315412186379928,
+      "grad_norm": 0.3809366409326176,
+      "learning_rate": 4.779220253836629e-06,
+      "loss": 0.598,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8318067171113767,
+      "grad_norm": 0.3849597152962915,
+      "learning_rate": 4.779076780900264e-06,
+      "loss": 0.6174,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8320722155847604,
+      "grad_norm": 0.3905174099243532,
+      "learning_rate": 4.77893326351632e-06,
+      "loss": 0.6035,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8323377140581442,
+      "grad_norm": 0.3917365719394999,
+      "learning_rate": 4.7787897016876e-06,
+      "loss": 0.6056,
+      "step": 3135
+    },
+    {
+      "epoch": 0.8326032125315279,
+      "grad_norm": 0.37880885390080776,
+      "learning_rate": 4.778646095416899e-06,
+      "loss": 0.5776,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8328687110049118,
+      "grad_norm": 0.3848780838439989,
+      "learning_rate": 4.778502444707022e-06,
+      "loss": 0.5827,
+      "step": 3137
+    },
+    {
+      "epoch": 0.8331342094782955,
+      "grad_norm": 0.40166021223746573,
+      "learning_rate": 4.778358749560767e-06,
+      "loss": 0.5953,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8333997079516793,
+      "grad_norm": 0.38804772882322447,
+      "learning_rate": 4.778215009980939e-06,
+      "loss": 0.6234,
+      "step": 3139
+    },
+    {
+      "epoch": 0.833665206425063,
+      "grad_norm": 0.39274118005395675,
+      "learning_rate": 4.77807122597034e-06,
+      "loss": 0.6135,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8339307048984468,
+      "grad_norm": 0.3835964630609637,
+      "learning_rate": 4.777927397531774e-06,
+      "loss": 0.6446,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8341962033718306,
+      "grad_norm": 0.39095336947029047,
+      "learning_rate": 4.7777835246680474e-06,
+      "loss": 0.6388,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8344617018452144,
+      "grad_norm": 0.3880369156107484,
+      "learning_rate": 4.7776396073819644e-06,
+      "loss": 0.6282,
+      "step": 3143
+    },
+    {
+      "epoch": 0.8347272003185982,
+      "grad_norm": 0.3938324515828304,
+      "learning_rate": 4.777495645676332e-06,
+      "loss": 0.6383,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8349926987919819,
+      "grad_norm": 0.390770988372887,
+      "learning_rate": 4.7773516395539585e-06,
+      "loss": 0.5779,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8352581972653658,
+      "grad_norm": 0.3868153973519988,
+      "learning_rate": 4.777207589017653e-06,
+      "loss": 0.5991,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8355236957387495,
+      "grad_norm": 0.39428931749911,
+      "learning_rate": 4.777063494070224e-06,
+      "loss": 0.6196,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8357891942121333,
+      "grad_norm": 0.386015801009133,
+      "learning_rate": 4.7769193547144796e-06,
+      "loss": 0.6096,
+      "step": 3148
+    },
+    {
+      "epoch": 0.836054692685517,
+      "grad_norm": 0.38137073746130273,
+      "learning_rate": 4.776775170953235e-06,
+      "loss": 0.5956,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8363201911589009,
+      "grad_norm": 0.37427746563359643,
+      "learning_rate": 4.776630942789299e-06,
+      "loss": 0.5862,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8365856896322846,
+      "grad_norm": 0.3666149995254103,
+      "learning_rate": 4.776486670225486e-06,
+      "loss": 0.6312,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8368511881056684,
+      "grad_norm": 0.40489049933255256,
+      "learning_rate": 4.776342353264609e-06,
+      "loss": 0.6218,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8371166865790521,
+      "grad_norm": 0.40845708285976867,
+      "learning_rate": 4.776197991909482e-06,
+      "loss": 0.6024,
+      "step": 3153
+    },
+    {
+      "epoch": 0.837382185052436,
+      "grad_norm": 0.3818684657208028,
+      "learning_rate": 4.7760535861629224e-06,
+      "loss": 0.6286,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8376476835258198,
+      "grad_norm": 0.395799031268732,
+      "learning_rate": 4.775909136027744e-06,
+      "loss": 0.6188,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8379131819992035,
+      "grad_norm": 0.3887917180500662,
+      "learning_rate": 4.7757646415067655e-06,
+      "loss": 0.605,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8381786804725873,
+      "grad_norm": 0.3691750625253155,
+      "learning_rate": 4.7756201026028045e-06,
+      "loss": 0.5931,
+      "step": 3157
+    },
+    {
+      "epoch": 0.838444178945971,
+      "grad_norm": 0.39371689343704186,
+      "learning_rate": 4.775475519318679e-06,
+      "loss": 0.6165,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 0.39668220311237906,
+      "learning_rate": 4.7753308916572105e-06,
+      "loss": 0.6363,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8389751758927386,
+      "grad_norm": 0.4355400970748323,
+      "learning_rate": 4.775186219621219e-06,
+      "loss": 0.5781,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8392406743661224,
+      "grad_norm": 0.3943198735658743,
+      "learning_rate": 4.775041503213525e-06,
+      "loss": 0.6506,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8395061728395061,
+      "grad_norm": 0.4012652048985837,
+      "learning_rate": 4.774896742436951e-06,
+      "loss": 0.6253,
+      "step": 3162
+    },
+    {
+      "epoch": 0.83977167131289,
+      "grad_norm": 0.4292686082847599,
+      "learning_rate": 4.774751937294321e-06,
+      "loss": 0.6335,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8400371697862737,
+      "grad_norm": 0.4069525415717263,
+      "learning_rate": 4.774607087788459e-06,
+      "loss": 0.6076,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8403026682596575,
+      "grad_norm": 0.38379359031761656,
+      "learning_rate": 4.7744621939221895e-06,
+      "loss": 0.599,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8405681667330412,
+      "grad_norm": 0.398286940665663,
+      "learning_rate": 4.774317255698339e-06,
+      "loss": 0.6181,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8408336652064251,
+      "grad_norm": 0.3917339772562374,
+      "learning_rate": 4.774172273119732e-06,
+      "loss": 0.6525,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8410991636798089,
+      "grad_norm": 0.4021805520194308,
+      "learning_rate": 4.774027246189198e-06,
+      "loss": 0.5924,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8413646621531926,
+      "grad_norm": 0.394687312393228,
+      "learning_rate": 4.773882174909565e-06,
+      "loss": 0.6059,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8416301606265764,
+      "grad_norm": 0.393902015013949,
+      "learning_rate": 4.773737059283663e-06,
+      "loss": 0.6391,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8418956590999602,
+      "grad_norm": 0.3926874567469572,
+      "learning_rate": 4.77359189931432e-06,
+      "loss": 0.6377,
+      "step": 3171
+    },
+    {
+      "epoch": 0.842161157573344,
+      "grad_norm": 0.38647410589710635,
+      "learning_rate": 4.7734466950043686e-06,
+      "loss": 0.6461,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8424266560467277,
+      "grad_norm": 0.39205605490187023,
+      "learning_rate": 4.773301446356642e-06,
+      "loss": 0.5811,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8426921545201115,
+      "grad_norm": 0.39220506787691517,
+      "learning_rate": 4.773156153373969e-06,
+      "loss": 0.6032,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8429576529934952,
+      "grad_norm": 0.3900570347682403,
+      "learning_rate": 4.773010816059186e-06,
+      "loss": 0.6157,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8432231514668791,
+      "grad_norm": 0.38721075626040485,
+      "learning_rate": 4.772865434415127e-06,
+      "loss": 0.6215,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8434886499402628,
+      "grad_norm": 0.37977109955804594,
+      "learning_rate": 4.772720008444627e-06,
+      "loss": 0.6251,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8437541484136466,
+      "grad_norm": 0.37568541974395564,
+      "learning_rate": 4.7725745381505224e-06,
+      "loss": 0.6136,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8440196468870304,
+      "grad_norm": 0.38619740189382623,
+      "learning_rate": 4.772429023535649e-06,
+      "loss": 0.6187,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8442851453604142,
+      "grad_norm": 0.4080744813654574,
+      "learning_rate": 4.772283464602848e-06,
+      "loss": 0.615,
+      "step": 3180
+    },
+    {
+      "epoch": 0.844550643833798,
+      "grad_norm": 0.39936099484356835,
+      "learning_rate": 4.772137861354954e-06,
+      "loss": 0.638,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8448161423071817,
+      "grad_norm": 0.3864016093507137,
+      "learning_rate": 4.771992213794809e-06,
+      "loss": 0.6085,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8450816407805655,
+      "grad_norm": 0.40017897090729204,
+      "learning_rate": 4.771846521925254e-06,
+      "loss": 0.5825,
+      "step": 3183
+    },
+    {
+      "epoch": 0.8453471392539493,
+      "grad_norm": 0.4003068420345821,
+      "learning_rate": 4.771700785749128e-06,
+      "loss": 0.653,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8456126377273331,
+      "grad_norm": 0.3769586394165981,
+      "learning_rate": 4.771555005269276e-06,
+      "loss": 0.6215,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8458781362007168,
+      "grad_norm": 0.38817502344943944,
+      "learning_rate": 4.7714091804885386e-06,
+      "loss": 0.6301,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8461436346741006,
+      "grad_norm": 0.3942822918160568,
+      "learning_rate": 4.7712633114097615e-06,
+      "loss": 0.6495,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8464091331474844,
+      "grad_norm": 0.38853609691553753,
+      "learning_rate": 4.7711173980357886e-06,
+      "loss": 0.65,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8466746316208682,
+      "grad_norm": 0.39466017653366076,
+      "learning_rate": 4.770971440369465e-06,
+      "loss": 0.6282,
+      "step": 3189
+    },
+    {
+      "epoch": 0.846940130094252,
+      "grad_norm": 0.37338410635228586,
+      "learning_rate": 4.770825438413639e-06,
+      "loss": 0.5808,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8472056285676357,
+      "grad_norm": 0.3784067564181116,
+      "learning_rate": 4.770679392171156e-06,
+      "loss": 0.6213,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8474711270410196,
+      "grad_norm": 0.43627897176446817,
+      "learning_rate": 4.770533301644866e-06,
+      "loss": 0.5834,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8477366255144033,
+      "grad_norm": 0.3889817148927191,
+      "learning_rate": 4.770387166837618e-06,
+      "loss": 0.5672,
+      "step": 3193
+    },
+    {
+      "epoch": 0.8480021239877871,
+      "grad_norm": 0.3792133359844391,
+      "learning_rate": 4.77024098775226e-06,
+      "loss": 0.6271,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8482676224611708,
+      "grad_norm": 0.3888465738791731,
+      "learning_rate": 4.770094764391645e-06,
+      "loss": 0.564,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8485331209345546,
+      "grad_norm": 0.3920221713124069,
+      "learning_rate": 4.769948496758624e-06,
+      "loss": 0.6149,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8487986194079384,
+      "grad_norm": 0.42496371198852184,
+      "learning_rate": 4.7698021848560495e-06,
+      "loss": 0.6031,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8490641178813222,
+      "grad_norm": 0.38540140272960993,
+      "learning_rate": 4.769655828686775e-06,
+      "loss": 0.6429,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8493296163547059,
+      "grad_norm": 0.4314829857035156,
+      "learning_rate": 4.769509428253655e-06,
+      "loss": 0.6424,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8495951148280897,
+      "grad_norm": 0.39222652379073375,
+      "learning_rate": 4.769362983559544e-06,
+      "loss": 0.5902,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8498606133014736,
+      "grad_norm": 0.4012979678910393,
+      "learning_rate": 4.769216494607298e-06,
+      "loss": 0.6151,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8501261117748573,
+      "grad_norm": 0.38349179695624713,
+      "learning_rate": 4.7690699613997745e-06,
+      "loss": 0.6028,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8503916102482411,
+      "grad_norm": 0.41836772041991915,
+      "learning_rate": 4.768923383939832e-06,
+      "loss": 0.5919,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8506571087216248,
+      "grad_norm": 0.39372308711913767,
+      "learning_rate": 4.768776762230328e-06,
+      "loss": 0.6112,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8509226071950087,
+      "grad_norm": 0.3951894127325376,
+      "learning_rate": 4.768630096274121e-06,
+      "loss": 0.6077,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8511881056683924,
+      "grad_norm": 0.39088160061797267,
+      "learning_rate": 4.7684833860740746e-06,
+      "loss": 0.6161,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8514536041417762,
+      "grad_norm": 0.38585445036304644,
+      "learning_rate": 4.768336631633046e-06,
+      "loss": 0.6118,
+      "step": 3207
+    },
+    {
+      "epoch": 0.8517191026151599,
+      "grad_norm": 0.38547069026437336,
+      "learning_rate": 4.7681898329539004e-06,
+      "loss": 0.6158,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8519846010885438,
+      "grad_norm": 0.39136239017861296,
+      "learning_rate": 4.768042990039499e-06,
+      "loss": 0.6758,
+      "step": 3209
+    },
+    {
+      "epoch": 0.8522500995619275,
+      "grad_norm": 0.3811011756459646,
+      "learning_rate": 4.767896102892706e-06,
+      "loss": 0.6718,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8525155980353113,
+      "grad_norm": 0.38935438062194233,
+      "learning_rate": 4.767749171516387e-06,
+      "loss": 0.6134,
+      "step": 3211
+    },
+    {
+      "epoch": 0.852781096508695,
+      "grad_norm": 0.3931317395063905,
+      "learning_rate": 4.767602195913407e-06,
+      "loss": 0.6247,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8530465949820788,
+      "grad_norm": 0.40413026725746803,
+      "learning_rate": 4.767455176086631e-06,
+      "loss": 0.617,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8533120934554627,
+      "grad_norm": 0.3768135899049395,
+      "learning_rate": 4.767308112038928e-06,
+      "loss": 0.6218,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8535775919288464,
+      "grad_norm": 0.4069314027079967,
+      "learning_rate": 4.767161003773165e-06,
+      "loss": 0.6224,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8538430904022302,
+      "grad_norm": 0.39850720930153566,
+      "learning_rate": 4.767013851292213e-06,
+      "loss": 0.6111,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8541085888756139,
+      "grad_norm": 0.4005002568480204,
+      "learning_rate": 4.766866654598938e-06,
+      "loss": 0.6544,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8543740873489978,
+      "grad_norm": 0.44980922914111277,
+      "learning_rate": 4.766719413696215e-06,
+      "loss": 0.6142,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8546395858223815,
+      "grad_norm": 0.38353505496742896,
+      "learning_rate": 4.7665721285869124e-06,
+      "loss": 0.5871,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8549050842957653,
+      "grad_norm": 0.3764916703833999,
+      "learning_rate": 4.766424799273905e-06,
+      "loss": 0.5943,
+      "step": 3220
+    },
+    {
+      "epoch": 0.855170582769149,
+      "grad_norm": 0.38866086443464615,
+      "learning_rate": 4.766277425760063e-06,
+      "loss": 0.6284,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8554360812425329,
+      "grad_norm": 0.392596700224273,
+      "learning_rate": 4.7661300080482635e-06,
+      "loss": 0.6329,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8557015797159166,
+      "grad_norm": 0.4082030501237004,
+      "learning_rate": 4.76598254614138e-06,
+      "loss": 0.5855,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8559670781893004,
+      "grad_norm": 0.41373048475368934,
+      "learning_rate": 4.765835040042288e-06,
+      "loss": 0.5874,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8562325766626842,
+      "grad_norm": 0.38405175079068105,
+      "learning_rate": 4.765687489753866e-06,
+      "loss": 0.6002,
+      "step": 3225
+    },
+    {
+      "epoch": 0.856498075136068,
+      "grad_norm": 0.3854283651083716,
+      "learning_rate": 4.765539895278991e-06,
+      "loss": 0.6149,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8567635736094518,
+      "grad_norm": 0.4031568729841069,
+      "learning_rate": 4.7653922566205414e-06,
+      "loss": 0.5757,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8570290720828355,
+      "grad_norm": 0.3836051323887502,
+      "learning_rate": 4.765244573781394e-06,
+      "loss": 0.6389,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8572945705562193,
+      "grad_norm": 0.3867823167628796,
+      "learning_rate": 4.765096846764433e-06,
+      "loss": 0.6013,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8575600690296031,
+      "grad_norm": 0.4094187066410639,
+      "learning_rate": 4.764949075572537e-06,
+      "loss": 0.6035,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8578255675029869,
+      "grad_norm": 0.4031456084406773,
+      "learning_rate": 4.764801260208589e-06,
+      "loss": 0.6202,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8580910659763706,
+      "grad_norm": 0.39175898887486005,
+      "learning_rate": 4.7646534006754705e-06,
+      "loss": 0.6309,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8583565644497544,
+      "grad_norm": 0.3905216451561184,
+      "learning_rate": 4.764505496976066e-06,
+      "loss": 0.6104,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8586220629231381,
+      "grad_norm": 0.4059328610186368,
+      "learning_rate": 4.764357549113261e-06,
+      "loss": 0.621,
+      "step": 3234
+    },
+    {
+      "epoch": 0.858887561396522,
+      "grad_norm": 0.3959113807772086,
+      "learning_rate": 4.764209557089938e-06,
+      "loss": 0.6541,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8591530598699058,
+      "grad_norm": 0.37569948041243706,
+      "learning_rate": 4.764061520908986e-06,
+      "loss": 0.6351,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8594185583432895,
+      "grad_norm": 0.3933463708060713,
+      "learning_rate": 4.76391344057329e-06,
+      "loss": 0.5878,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8596840568166733,
+      "grad_norm": 0.3967413311947283,
+      "learning_rate": 4.763765316085739e-06,
+      "loss": 0.6263,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8599495552900571,
+      "grad_norm": 0.38374454956475107,
+      "learning_rate": 4.763617147449222e-06,
+      "loss": 0.6187,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8602150537634409,
+      "grad_norm": 0.39735130900058785,
+      "learning_rate": 4.763468934666629e-06,
+      "loss": 0.6155,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8604805522368246,
+      "grad_norm": 0.3858536539765493,
+      "learning_rate": 4.763320677740849e-06,
+      "loss": 0.6341,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8607460507102084,
+      "grad_norm": 0.37599079648790223,
+      "learning_rate": 4.763172376674775e-06,
+      "loss": 0.6122,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8610115491835922,
+      "grad_norm": 0.38521902853267986,
+      "learning_rate": 4.763024031471298e-06,
+      "loss": 0.5676,
+      "step": 3243
+    },
+    {
+      "epoch": 0.861277047656976,
+      "grad_norm": 0.39153599125625715,
+      "learning_rate": 4.762875642133311e-06,
+      "loss": 0.5694,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8615425461303597,
+      "grad_norm": 0.3823507852988624,
+      "learning_rate": 4.7627272086637085e-06,
+      "loss": 0.5846,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8618080446037435,
+      "grad_norm": 0.40803509453668707,
+      "learning_rate": 4.762578731065387e-06,
+      "loss": 0.6291,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8620735430771274,
+      "grad_norm": 0.40715214908263025,
+      "learning_rate": 4.762430209341239e-06,
+      "loss": 0.6618,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8623390415505111,
+      "grad_norm": 0.3945799129136733,
+      "learning_rate": 4.762281643494163e-06,
+      "loss": 0.6323,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8626045400238949,
+      "grad_norm": 0.3882279641623884,
+      "learning_rate": 4.762133033527056e-06,
+      "loss": 0.6015,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8628700384972786,
+      "grad_norm": 0.38923969820469684,
+      "learning_rate": 4.761984379442816e-06,
+      "loss": 0.6155,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8631355369706624,
+      "grad_norm": 0.38452506604652875,
+      "learning_rate": 4.761835681244342e-06,
+      "loss": 0.6291,
+      "step": 3251
+    },
+    {
+      "epoch": 0.8634010354440462,
+      "grad_norm": 0.39722003852402576,
+      "learning_rate": 4.761686938934534e-06,
+      "loss": 0.6235,
+      "step": 3252
+    },
+    {
+      "epoch": 0.86366653391743,
+      "grad_norm": 0.39554783917357644,
+      "learning_rate": 4.761538152516294e-06,
+      "loss": 0.6092,
+      "step": 3253
+    },
+    {
+      "epoch": 0.8639320323908137,
+      "grad_norm": 0.4056793370090963,
+      "learning_rate": 4.761389321992522e-06,
+      "loss": 0.6271,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8641975308641975,
+      "grad_norm": 0.39004068585232793,
+      "learning_rate": 4.7612404473661225e-06,
+      "loss": 0.6239,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8644630293375813,
+      "grad_norm": 0.37458859211804146,
+      "learning_rate": 4.7610915286399976e-06,
+      "loss": 0.6052,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8647285278109651,
+      "grad_norm": 0.3854551154278075,
+      "learning_rate": 4.760942565817051e-06,
+      "loss": 0.5855,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8649940262843488,
+      "grad_norm": 0.39172330553477347,
+      "learning_rate": 4.7607935589001884e-06,
+      "loss": 0.5993,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8652595247577326,
+      "grad_norm": 0.38955921923922726,
+      "learning_rate": 4.760644507892317e-06,
+      "loss": 0.6043,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8655250232311165,
+      "grad_norm": 0.38910605908125334,
+      "learning_rate": 4.7604954127963425e-06,
+      "loss": 0.6279,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8657905217045002,
+      "grad_norm": 0.38661001591718847,
+      "learning_rate": 4.7603462736151725e-06,
+      "loss": 0.6092,
+      "step": 3261
+    },
+    {
+      "epoch": 0.866056020177884,
+      "grad_norm": 0.3775927695644392,
+      "learning_rate": 4.760197090351716e-06,
+      "loss": 0.6236,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8663215186512677,
+      "grad_norm": 0.38930640013674983,
+      "learning_rate": 4.760047863008883e-06,
+      "loss": 0.5734,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8665870171246516,
+      "grad_norm": 0.39780941997169983,
+      "learning_rate": 4.759898591589582e-06,
+      "loss": 0.5673,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8668525155980353,
+      "grad_norm": 0.383324042561951,
+      "learning_rate": 4.759749276096726e-06,
+      "loss": 0.6068,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8671180140714191,
+      "grad_norm": 0.3968484934702401,
+      "learning_rate": 4.759599916533226e-06,
+      "loss": 0.6243,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8673835125448028,
+      "grad_norm": 0.37545830597412905,
+      "learning_rate": 4.759450512901995e-06,
+      "loss": 0.5757,
+      "step": 3267
+    },
+    {
+      "epoch": 0.8676490110181867,
+      "grad_norm": 0.4104189431074705,
+      "learning_rate": 4.759301065205947e-06,
+      "loss": 0.6375,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8679145094915705,
+      "grad_norm": 0.3834421916652933,
+      "learning_rate": 4.759151573447996e-06,
+      "loss": 0.6068,
+      "step": 3269
+    },
+    {
+      "epoch": 0.8681800079649542,
+      "grad_norm": 0.393487425142642,
+      "learning_rate": 4.75900203763106e-06,
+      "loss": 0.6218,
+      "step": 3270
+    },
+    {
+      "epoch": 0.868445506438338,
+      "grad_norm": 0.3808668639964955,
+      "learning_rate": 4.7588524577580505e-06,
+      "loss": 0.6066,
+      "step": 3271
+    },
+    {
+      "epoch": 0.8687110049117217,
+      "grad_norm": 0.38748737323108323,
+      "learning_rate": 4.75870283383189e-06,
+      "loss": 0.5993,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8689765033851056,
+      "grad_norm": 0.3936400929341232,
+      "learning_rate": 4.758553165855492e-06,
+      "loss": 0.6576,
+      "step": 3273
+    },
+    {
+      "epoch": 0.8692420018584893,
+      "grad_norm": 0.40152461480424106,
+      "learning_rate": 4.758403453831778e-06,
+      "loss": 0.6477,
+      "step": 3274
+    },
+    {
+      "epoch": 0.8695075003318731,
+      "grad_norm": 0.3761131885945365,
+      "learning_rate": 4.758253697763668e-06,
+      "loss": 0.5842,
+      "step": 3275
+    },
+    {
+      "epoch": 0.8697729988052568,
+      "grad_norm": 0.3909990154254638,
+      "learning_rate": 4.758103897654081e-06,
+      "loss": 0.625,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8700384972786407,
+      "grad_norm": 0.38964355475994894,
+      "learning_rate": 4.757954053505939e-06,
+      "loss": 0.5957,
+      "step": 3277
+    },
+    {
+      "epoch": 0.8703039957520244,
+      "grad_norm": 0.3880481326523315,
+      "learning_rate": 4.757804165322165e-06,
+      "loss": 0.6075,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8705694942254082,
+      "grad_norm": 0.3894186819851033,
+      "learning_rate": 4.7576542331056814e-06,
+      "loss": 0.6033,
+      "step": 3279
+    },
+    {
+      "epoch": 0.8708349926987919,
+      "grad_norm": 0.38956140968114766,
+      "learning_rate": 4.757504256859412e-06,
+      "loss": 0.5867,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8711004911721758,
+      "grad_norm": 0.39246319443338373,
+      "learning_rate": 4.757354236586283e-06,
+      "loss": 0.6222,
+      "step": 3281
+    },
+    {
+      "epoch": 0.8713659896455596,
+      "grad_norm": 0.39588894736088304,
+      "learning_rate": 4.75720417228922e-06,
+      "loss": 0.6361,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8716314881189433,
+      "grad_norm": 0.3897914130351932,
+      "learning_rate": 4.757054063971148e-06,
+      "loss": 0.6113,
+      "step": 3283
+    },
+    {
+      "epoch": 0.8718969865923271,
+      "grad_norm": 0.3939227850561614,
+      "learning_rate": 4.756903911634996e-06,
+      "loss": 0.6071,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8721624850657109,
+      "grad_norm": 0.4076397357244799,
+      "learning_rate": 4.7567537152836915e-06,
+      "loss": 0.6469,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8724279835390947,
+      "grad_norm": 0.4119187367460943,
+      "learning_rate": 4.756603474920165e-06,
+      "loss": 0.5813,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8726934820124784,
+      "grad_norm": 0.39519397283206353,
+      "learning_rate": 4.756453190547344e-06,
+      "loss": 0.6334,
+      "step": 3287
+    },
+    {
+      "epoch": 0.8729589804858622,
+      "grad_norm": 0.3871710551227271,
+      "learning_rate": 4.756302862168163e-06,
+      "loss": 0.6347,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8732244789592459,
+      "grad_norm": 0.38669326400346216,
+      "learning_rate": 4.7561524897855504e-06,
+      "loss": 0.6172,
+      "step": 3289
+    },
+    {
+      "epoch": 0.8734899774326298,
+      "grad_norm": 0.39639256465687545,
+      "learning_rate": 4.756002073402441e-06,
+      "loss": 0.6305,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8737554759060135,
+      "grad_norm": 0.39930090714223104,
+      "learning_rate": 4.755851613021767e-06,
+      "loss": 0.6039,
+      "step": 3291
+    },
+    {
+      "epoch": 0.8740209743793973,
+      "grad_norm": 0.3699246895708019,
+      "learning_rate": 4.7557011086464634e-06,
+      "loss": 0.5902,
+      "step": 3292
+    },
+    {
+      "epoch": 0.874286472852781,
+      "grad_norm": 0.3915798091559982,
+      "learning_rate": 4.755550560279465e-06,
+      "loss": 0.6154,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8745519713261649,
+      "grad_norm": 0.41677536311013336,
+      "learning_rate": 4.755399967923709e-06,
+      "loss": 0.6049,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8748174697995487,
+      "grad_norm": 0.3949964007316087,
+      "learning_rate": 4.755249331582131e-06,
+      "loss": 0.6511,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8750829682729324,
+      "grad_norm": 0.39726565234475836,
+      "learning_rate": 4.755098651257669e-06,
+      "loss": 0.5981,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8753484667463162,
+      "grad_norm": 0.4127253070033059,
+      "learning_rate": 4.7549479269532615e-06,
+      "loss": 0.608,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8756139652197,
+      "grad_norm": 0.38828082155939475,
+      "learning_rate": 4.754797158671849e-06,
+      "loss": 0.6173,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8758794636930838,
+      "grad_norm": 0.38841361287621073,
+      "learning_rate": 4.754646346416371e-06,
+      "loss": 0.5959,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8761449621664675,
+      "grad_norm": 0.38566806534210335,
+      "learning_rate": 4.754495490189769e-06,
+      "loss": 0.6195,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8764104606398513,
+      "grad_norm": 0.3907819137410776,
+      "learning_rate": 4.754344589994985e-06,
+      "loss": 0.6131,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8766759591132351,
+      "grad_norm": 0.38308247558070607,
+      "learning_rate": 4.754193645834963e-06,
+      "loss": 0.598,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8769414575866189,
+      "grad_norm": 0.40440391073340404,
+      "learning_rate": 4.754042657712643e-06,
+      "loss": 0.5883,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8772069560600027,
+      "grad_norm": 0.39233713811219384,
+      "learning_rate": 4.753891625630975e-06,
+      "loss": 0.6387,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8774724545333864,
+      "grad_norm": 0.39130254882310145,
+      "learning_rate": 4.7537405495929e-06,
+      "loss": 0.6092,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8777379530067703,
+      "grad_norm": 0.39412519967687676,
+      "learning_rate": 4.753589429601367e-06,
+      "loss": 0.5972,
+      "step": 3306
+    },
+    {
+      "epoch": 0.878003451480154,
+      "grad_norm": 0.40428041354450067,
+      "learning_rate": 4.753438265659322e-06,
+      "loss": 0.5991,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8782689499535378,
+      "grad_norm": 0.37504294957402573,
+      "learning_rate": 4.753287057769714e-06,
+      "loss": 0.6327,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8785344484269215,
+      "grad_norm": 0.3839356147912363,
+      "learning_rate": 4.753135805935492e-06,
+      "loss": 0.6204,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8787999469003053,
+      "grad_norm": 0.40923237791155004,
+      "learning_rate": 4.752984510159604e-06,
+      "loss": 0.592,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8790654453736891,
+      "grad_norm": 0.39458197109693294,
+      "learning_rate": 4.752833170445002e-06,
+      "loss": 0.6846,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8793309438470729,
+      "grad_norm": 0.3956103479023119,
+      "learning_rate": 4.752681786794637e-06,
+      "loss": 0.6084,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8795964423204566,
+      "grad_norm": 0.3800192075546216,
+      "learning_rate": 4.752530359211462e-06,
+      "loss": 0.601,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8798619407938404,
+      "grad_norm": 0.38879181208594316,
+      "learning_rate": 4.75237888769843e-06,
+      "loss": 0.616,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8801274392672243,
+      "grad_norm": 0.388194199082558,
+      "learning_rate": 4.752227372258495e-06,
+      "loss": 0.6089,
+      "step": 3315
+    },
+    {
+      "epoch": 0.880392937740608,
+      "grad_norm": 0.3725905196810677,
+      "learning_rate": 4.752075812894611e-06,
+      "loss": 0.5965,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8806584362139918,
+      "grad_norm": 0.39300519418278473,
+      "learning_rate": 4.751924209609735e-06,
+      "loss": 0.6254,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8809239346873755,
+      "grad_norm": 0.4064612647247102,
+      "learning_rate": 4.751772562406824e-06,
+      "loss": 0.6028,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8811894331607594,
+      "grad_norm": 0.38658665977171214,
+      "learning_rate": 4.751620871288834e-06,
+      "loss": 0.6104,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8814549316341431,
+      "grad_norm": 0.39343046388128616,
+      "learning_rate": 4.7514691362587245e-06,
+      "loss": 0.6232,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8817204301075269,
+      "grad_norm": 0.4038024334947061,
+      "learning_rate": 4.751317357319454e-06,
+      "loss": 0.6324,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8819859285809106,
+      "grad_norm": 0.4241980139379008,
+      "learning_rate": 4.751165534473983e-06,
+      "loss": 0.6009,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8822514270542945,
+      "grad_norm": 0.3860880812837124,
+      "learning_rate": 4.7510136677252724e-06,
+      "loss": 0.5575,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8825169255276782,
+      "grad_norm": 0.39684502196861987,
+      "learning_rate": 4.7508617570762836e-06,
+      "loss": 0.6574,
+      "step": 3324
+    },
+    {
+      "epoch": 0.882782424001062,
+      "grad_norm": 0.4076841287759156,
+      "learning_rate": 4.750709802529979e-06,
+      "loss": 0.6226,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8830479224744457,
+      "grad_norm": 0.49457788706324485,
+      "learning_rate": 4.7505578040893236e-06,
+      "loss": 0.5857,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8833134209478295,
+      "grad_norm": 0.3664050501234694,
+      "learning_rate": 4.75040576175728e-06,
+      "loss": 0.6097,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8835789194212134,
+      "grad_norm": 0.3796588890867391,
+      "learning_rate": 4.750253675536815e-06,
+      "loss": 0.6101,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8838444178945971,
+      "grad_norm": 0.4578942762009108,
+      "learning_rate": 4.750101545430893e-06,
+      "loss": 0.6186,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8841099163679809,
+      "grad_norm": 0.44615251759776303,
+      "learning_rate": 4.749949371442481e-06,
+      "loss": 0.6068,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8843754148413646,
+      "grad_norm": 0.38671244008324474,
+      "learning_rate": 4.749797153574548e-06,
+      "loss": 0.6161,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8846409133147485,
+      "grad_norm": 0.4137124586297112,
+      "learning_rate": 4.749644891830063e-06,
+      "loss": 0.6128,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8849064117881322,
+      "grad_norm": 0.47396625417379273,
+      "learning_rate": 4.749492586211994e-06,
+      "loss": 0.6194,
+      "step": 3333
+    },
+    {
+      "epoch": 0.885171910261516,
+      "grad_norm": 0.4136404060666191,
+      "learning_rate": 4.749340236723311e-06,
+      "loss": 0.6364,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8854374087348997,
+      "grad_norm": 0.39737095985235527,
+      "learning_rate": 4.749187843366987e-06,
+      "loss": 0.575,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8857029072082836,
+      "grad_norm": 0.4016915774106672,
+      "learning_rate": 4.7490354061459926e-06,
+      "loss": 0.6214,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8859684056816673,
+      "grad_norm": 0.3923140730107079,
+      "learning_rate": 4.7488829250633015e-06,
+      "loss": 0.6213,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8862339041550511,
+      "grad_norm": 0.3950996626156589,
+      "learning_rate": 4.748730400121887e-06,
+      "loss": 0.6503,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8864994026284349,
+      "grad_norm": 0.3865926499544235,
+      "learning_rate": 4.748577831324723e-06,
+      "loss": 0.6139,
+      "step": 3339
+    },
+    {
+      "epoch": 0.8867649011018187,
+      "grad_norm": 0.3876237606170458,
+      "learning_rate": 4.748425218674787e-06,
+      "loss": 0.6486,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8870303995752025,
+      "grad_norm": 0.39219201650548463,
+      "learning_rate": 4.748272562175054e-06,
+      "loss": 0.6845,
+      "step": 3341
+    },
+    {
+      "epoch": 0.8872958980485862,
+      "grad_norm": 0.4229854213010335,
+      "learning_rate": 4.748119861828501e-06,
+      "loss": 0.5981,
+      "step": 3342
+    },
+    {
+      "epoch": 0.88756139652197,
+      "grad_norm": 0.3867845738529824,
+      "learning_rate": 4.747967117638107e-06,
+      "loss": 0.6451,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8878268949953537,
+      "grad_norm": 0.4103182291847476,
+      "learning_rate": 4.7478143296068494e-06,
+      "loss": 0.6363,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8880923934687376,
+      "grad_norm": 0.3923960939700569,
+      "learning_rate": 4.747661497737709e-06,
+      "loss": 0.5832,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8883578919421213,
+      "grad_norm": 0.39655784312828,
+      "learning_rate": 4.747508622033666e-06,
+      "loss": 0.629,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8886233904155051,
+      "grad_norm": 0.39563218181553156,
+      "learning_rate": 4.747355702497702e-06,
+      "loss": 0.6217,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.3894663617075761,
+      "learning_rate": 4.7472027391328e-06,
+      "loss": 0.6236,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8891543873622727,
+      "grad_norm": 0.3794840489106913,
+      "learning_rate": 4.747049731941943e-06,
+      "loss": 0.6108,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8894198858356565,
+      "grad_norm": 0.406072941096683,
+      "learning_rate": 4.746896680928113e-06,
+      "loss": 0.6348,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8896853843090402,
+      "grad_norm": 0.3906669065421325,
+      "learning_rate": 4.746743586094297e-06,
+      "loss": 0.5816,
+      "step": 3351
+    },
+    {
+      "epoch": 0.889950882782424,
+      "grad_norm": 0.4005635751735138,
+      "learning_rate": 4.7465904474434795e-06,
+      "loss": 0.6032,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8902163812558078,
+      "grad_norm": 0.40568533560202646,
+      "learning_rate": 4.746437264978649e-06,
+      "loss": 0.6506,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8904818797291916,
+      "grad_norm": 0.3867746098250967,
+      "learning_rate": 4.7462840387027905e-06,
+      "loss": 0.6336,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8907473782025753,
+      "grad_norm": 0.4134811903339993,
+      "learning_rate": 4.746130768618894e-06,
+      "loss": 0.5994,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8910128766759591,
+      "grad_norm": 0.3959572858923188,
+      "learning_rate": 4.745977454729947e-06,
+      "loss": 0.5902,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8912783751493429,
+      "grad_norm": 0.36989418847389577,
+      "learning_rate": 4.7458240970389415e-06,
+      "loss": 0.5872,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8915438736227267,
+      "grad_norm": 0.39633422708202654,
+      "learning_rate": 4.7456706955488674e-06,
+      "loss": 0.6096,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8918093720961104,
+      "grad_norm": 0.40682819594508984,
+      "learning_rate": 4.745517250262717e-06,
+      "loss": 0.6215,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8920748705694942,
+      "grad_norm": 0.39040005074295636,
+      "learning_rate": 4.7453637611834805e-06,
+      "loss": 0.5893,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8923403690428781,
+      "grad_norm": 0.38470472898196983,
+      "learning_rate": 4.7452102283141545e-06,
+      "loss": 0.5662,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8926058675162618,
+      "grad_norm": 0.38319021046409607,
+      "learning_rate": 4.745056651657731e-06,
+      "loss": 0.584,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8928713659896456,
+      "grad_norm": 0.3924579676587756,
+      "learning_rate": 4.744903031217206e-06,
+      "loss": 0.6517,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8931368644630293,
+      "grad_norm": 0.3940493281798428,
+      "learning_rate": 4.744749366995576e-06,
+      "loss": 0.5955,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8934023629364131,
+      "grad_norm": 0.39069459108657206,
+      "learning_rate": 4.744595658995837e-06,
+      "loss": 0.6208,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8936678614097969,
+      "grad_norm": 0.41369043512590986,
+      "learning_rate": 4.744441907220987e-06,
+      "loss": 0.641,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8939333598831807,
+      "grad_norm": 0.3848688934330599,
+      "learning_rate": 4.744288111674023e-06,
+      "loss": 0.5991,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8941988583565644,
+      "grad_norm": 0.39001922379455345,
+      "learning_rate": 4.744134272357948e-06,
+      "loss": 0.624,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8944643568299482,
+      "grad_norm": 0.3940172810090595,
+      "learning_rate": 4.74398038927576e-06,
+      "loss": 0.6004,
+      "step": 3369
+    },
+    {
+      "epoch": 0.894729855303332,
+      "grad_norm": 0.37253415916519955,
+      "learning_rate": 4.743826462430459e-06,
+      "loss": 0.5825,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8949953537767158,
+      "grad_norm": 0.3808815856012623,
+      "learning_rate": 4.7436724918250485e-06,
+      "loss": 0.6093,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8952608522500995,
+      "grad_norm": 0.39620878451794717,
+      "learning_rate": 4.743518477462531e-06,
+      "loss": 0.6272,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8955263507234833,
+      "grad_norm": 0.39922070142204164,
+      "learning_rate": 4.7433644193459104e-06,
+      "loss": 0.6125,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8957918491968672,
+      "grad_norm": 0.3908614419935643,
+      "learning_rate": 4.7432103174781906e-06,
+      "loss": 0.6639,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8960573476702509,
+      "grad_norm": 0.38921458118132785,
+      "learning_rate": 4.743056171862378e-06,
+      "loss": 0.6039,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8963228461436347,
+      "grad_norm": 0.3945253800621873,
+      "learning_rate": 4.742901982501476e-06,
+      "loss": 0.6185,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8965883446170184,
+      "grad_norm": 0.3828735579396043,
+      "learning_rate": 4.742747749398497e-06,
+      "loss": 0.6274,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8968538430904023,
+      "grad_norm": 0.3895632511687009,
+      "learning_rate": 4.742593472556442e-06,
+      "loss": 0.6353,
+      "step": 3378
+    },
+    {
+      "epoch": 0.897119341563786,
+      "grad_norm": 0.3901858908952081,
+      "learning_rate": 4.742439151978326e-06,
+      "loss": 0.605,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8973848400371698,
+      "grad_norm": 0.40535561330640746,
+      "learning_rate": 4.7422847876671565e-06,
+      "loss": 0.6002,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8976503385105535,
+      "grad_norm": 0.3874398179605058,
+      "learning_rate": 4.742130379625942e-06,
+      "loss": 0.5923,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8979158369839373,
+      "grad_norm": 0.3998501987431258,
+      "learning_rate": 4.741975927857696e-06,
+      "loss": 0.601,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8981813354573212,
+      "grad_norm": 0.37363983255522465,
+      "learning_rate": 4.74182143236543e-06,
+      "loss": 0.6063,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8984468339307049,
+      "grad_norm": 0.3855054525446527,
+      "learning_rate": 4.741666893152157e-06,
+      "loss": 0.6124,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8987123324040887,
+      "grad_norm": 0.3905471841090725,
+      "learning_rate": 4.741512310220891e-06,
+      "loss": 0.6232,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8989778308774724,
+      "grad_norm": 0.3794340488347295,
+      "learning_rate": 4.7413576835746474e-06,
+      "loss": 0.5728,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8992433293508563,
+      "grad_norm": 0.39805627838134156,
+      "learning_rate": 4.741203013216441e-06,
+      "loss": 0.6178,
+      "step": 3387
+    },
+    {
+      "epoch": 0.89950882782424,
+      "grad_norm": 0.38134711964687046,
+      "learning_rate": 4.741048299149289e-06,
+      "loss": 0.5873,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8997743262976238,
+      "grad_norm": 0.3874373189477116,
+      "learning_rate": 4.740893541376207e-06,
+      "loss": 0.5988,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9000398247710075,
+      "grad_norm": 0.39219368851332287,
+      "learning_rate": 4.740738739900215e-06,
+      "loss": 0.6251,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9003053232443914,
+      "grad_norm": 0.37806396193996067,
+      "learning_rate": 4.740583894724331e-06,
+      "loss": 0.5789,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9005708217177751,
+      "grad_norm": 0.38366356775642907,
+      "learning_rate": 4.740429005851574e-06,
+      "loss": 0.6234,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9008363201911589,
+      "grad_norm": 0.3809104078142495,
+      "learning_rate": 4.740274073284968e-06,
+      "loss": 0.5959,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9011018186645426,
+      "grad_norm": 0.37604806239771565,
+      "learning_rate": 4.740119097027531e-06,
+      "loss": 0.5905,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9013673171379265,
+      "grad_norm": 0.38587494226103675,
+      "learning_rate": 4.739964077082287e-06,
+      "loss": 0.6012,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9016328156113103,
+      "grad_norm": 0.3753780403412113,
+      "learning_rate": 4.73980901345226e-06,
+      "loss": 0.6302,
+      "step": 3396
+    },
+    {
+      "epoch": 0.901898314084694,
+      "grad_norm": 0.3802983510698151,
+      "learning_rate": 4.739653906140473e-06,
+      "loss": 0.6305,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9021638125580778,
+      "grad_norm": 0.3881969020196794,
+      "learning_rate": 4.73949875514995e-06,
+      "loss": 0.6142,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9024293110314616,
+      "grad_norm": 0.40957065563154593,
+      "learning_rate": 4.7393435604837195e-06,
+      "loss": 0.6184,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9026948095048454,
+      "grad_norm": 0.3796227686844207,
+      "learning_rate": 4.739188322144806e-06,
+      "loss": 0.5881,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9029603079782291,
+      "grad_norm": 0.3912403746909842,
+      "learning_rate": 4.739033040136238e-06,
+      "loss": 0.6012,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 0.3953662526039944,
+      "learning_rate": 4.738877714461043e-06,
+      "loss": 0.6003,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9034913049249966,
+      "grad_norm": 0.3906456290994063,
+      "learning_rate": 4.7387223451222515e-06,
+      "loss": 0.604,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9037568033983805,
+      "grad_norm": 0.37655622906030306,
+      "learning_rate": 4.738566932122893e-06,
+      "loss": 0.5846,
+      "step": 3404
+    },
+    {
+      "epoch": 0.9040223018717642,
+      "grad_norm": 0.39770680349420695,
+      "learning_rate": 4.738411475465998e-06,
+      "loss": 0.6327,
+      "step": 3405
+    },
+    {
+      "epoch": 0.904287800345148,
+      "grad_norm": 0.3878671438040319,
+      "learning_rate": 4.738255975154599e-06,
+      "loss": 0.6029,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9045532988185317,
+      "grad_norm": 0.4007513607231948,
+      "learning_rate": 4.7381004311917284e-06,
+      "loss": 0.5809,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9048187972919156,
+      "grad_norm": 0.37434710592084747,
+      "learning_rate": 4.7379448435804184e-06,
+      "loss": 0.5807,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9050842957652994,
+      "grad_norm": 0.40287565951271437,
+      "learning_rate": 4.7377892123237055e-06,
+      "loss": 0.6223,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9053497942386831,
+      "grad_norm": 0.4027691649707259,
+      "learning_rate": 4.737633537424624e-06,
+      "loss": 0.6194,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9056152927120669,
+      "grad_norm": 0.383371165090648,
+      "learning_rate": 4.7374778188862105e-06,
+      "loss": 0.6313,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9058807911854507,
+      "grad_norm": 0.414209314183217,
+      "learning_rate": 4.7373220567115e-06,
+      "loss": 0.6125,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9061462896588345,
+      "grad_norm": 0.38664384582759515,
+      "learning_rate": 4.737166250903532e-06,
+      "loss": 0.6299,
+      "step": 3413
+    },
+    {
+      "epoch": 0.9064117881322182,
+      "grad_norm": 0.4126284430607286,
+      "learning_rate": 4.737010401465344e-06,
+      "loss": 0.6024,
+      "step": 3414
+    },
+    {
+      "epoch": 0.906677286605602,
+      "grad_norm": 0.39055349680356477,
+      "learning_rate": 4.736854508399977e-06,
+      "loss": 0.6471,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9069427850789858,
+      "grad_norm": 0.38250749709782844,
+      "learning_rate": 4.73669857171047e-06,
+      "loss": 0.6004,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9072082835523696,
+      "grad_norm": 0.4311919139039721,
+      "learning_rate": 4.736542591399864e-06,
+      "loss": 0.5911,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9074737820257534,
+      "grad_norm": 0.389457151028843,
+      "learning_rate": 4.736386567471202e-06,
+      "loss": 0.6431,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9077392804991371,
+      "grad_norm": 0.37651842436689276,
+      "learning_rate": 4.736230499927525e-06,
+      "loss": 0.618,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9080047789725209,
+      "grad_norm": 0.38980235627361093,
+      "learning_rate": 4.736074388771879e-06,
+      "loss": 0.5787,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9082702774459047,
+      "grad_norm": 0.3935241800958016,
+      "learning_rate": 4.735918234007309e-06,
+      "loss": 0.6011,
+      "step": 3421
+    },
+    {
+      "epoch": 0.9085357759192885,
+      "grad_norm": 0.37928593590440823,
+      "learning_rate": 4.735762035636857e-06,
+      "loss": 0.5975,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9088012743926722,
+      "grad_norm": 0.39108513017901275,
+      "learning_rate": 4.735605793663572e-06,
+      "loss": 0.6064,
+      "step": 3423
+    },
+    {
+      "epoch": 0.909066772866056,
+      "grad_norm": 0.39422797484336725,
+      "learning_rate": 4.7354495080904996e-06,
+      "loss": 0.6124,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9093322713394398,
+      "grad_norm": 0.38992106166335516,
+      "learning_rate": 4.735293178920689e-06,
+      "loss": 0.6122,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9095977698128236,
+      "grad_norm": 0.3811299990487383,
+      "learning_rate": 4.735136806157189e-06,
+      "loss": 0.6011,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9098632682862073,
+      "grad_norm": 0.3985867372643103,
+      "learning_rate": 4.734980389803048e-06,
+      "loss": 0.6066,
+      "step": 3427
+    },
+    {
+      "epoch": 0.9101287667595911,
+      "grad_norm": 0.3906465069650725,
+      "learning_rate": 4.734823929861317e-06,
+      "loss": 0.5955,
+      "step": 3428
+    },
+    {
+      "epoch": 0.910394265232975,
+      "grad_norm": 0.3959664146770712,
+      "learning_rate": 4.734667426335048e-06,
+      "loss": 0.6121,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9106597637063587,
+      "grad_norm": 0.40644770166392485,
+      "learning_rate": 4.734510879227293e-06,
+      "loss": 0.5867,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9109252621797425,
+      "grad_norm": 0.40123775455126887,
+      "learning_rate": 4.734354288541104e-06,
+      "loss": 0.637,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9111907606531262,
+      "grad_norm": 0.3817635893801134,
+      "learning_rate": 4.734197654279537e-06,
+      "loss": 0.6204,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9114562591265101,
+      "grad_norm": 0.38367590525080303,
+      "learning_rate": 4.734040976445645e-06,
+      "loss": 0.5631,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9117217575998938,
+      "grad_norm": 0.3918928428740401,
+      "learning_rate": 4.733884255042484e-06,
+      "loss": 0.5998,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9119872560732776,
+      "grad_norm": 0.3838734025692161,
+      "learning_rate": 4.7337274900731105e-06,
+      "loss": 0.6316,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9122527545466613,
+      "grad_norm": 0.3960968956208628,
+      "learning_rate": 4.733570681540581e-06,
+      "loss": 0.6327,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9125182530200452,
+      "grad_norm": 0.3898099642594238,
+      "learning_rate": 4.7334138294479554e-06,
+      "loss": 0.5759,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9127837514934289,
+      "grad_norm": 0.379897145027227,
+      "learning_rate": 4.733256933798292e-06,
+      "loss": 0.5834,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9130492499668127,
+      "grad_norm": 0.3919447951120241,
+      "learning_rate": 4.7330999945946496e-06,
+      "loss": 0.6302,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9133147484401964,
+      "grad_norm": 0.40362527250079516,
+      "learning_rate": 4.73294301184009e-06,
+      "loss": 0.5834,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9135802469135802,
+      "grad_norm": 0.38717567999686087,
+      "learning_rate": 4.732785985537675e-06,
+      "loss": 0.6519,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9138457453869641,
+      "grad_norm": 0.39825444352731054,
+      "learning_rate": 4.732628915690466e-06,
+      "loss": 0.6372,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9141112438603478,
+      "grad_norm": 0.40355891459094184,
+      "learning_rate": 4.732471802301527e-06,
+      "loss": 0.6332,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9143767423337316,
+      "grad_norm": 0.38584603358776676,
+      "learning_rate": 4.732314645373922e-06,
+      "loss": 0.5981,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9146422408071153,
+      "grad_norm": 0.39800617543877354,
+      "learning_rate": 4.732157444910716e-06,
+      "loss": 0.6018,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9149077392804992,
+      "grad_norm": 0.4027673253626878,
+      "learning_rate": 4.732000200914973e-06,
+      "loss": 0.5593,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9151732377538829,
+      "grad_norm": 0.39155546175687,
+      "learning_rate": 4.7318429133897625e-06,
+      "loss": 0.6538,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9154387362272667,
+      "grad_norm": 0.41234558121547255,
+      "learning_rate": 4.73168558233815e-06,
+      "loss": 0.5976,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9157042347006504,
+      "grad_norm": 0.3828924708686552,
+      "learning_rate": 4.731528207763205e-06,
+      "loss": 0.6136,
+      "step": 3449
+    },
+    {
+      "epoch": 0.9159697331740343,
+      "grad_norm": 0.37303850616839945,
+      "learning_rate": 4.731370789667996e-06,
+      "loss": 0.6211,
+      "step": 3450
+    },
+    {
+      "epoch": 0.916235231647418,
+      "grad_norm": 0.4000994455317232,
+      "learning_rate": 4.731213328055594e-06,
+      "loss": 0.6245,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9165007301208018,
+      "grad_norm": 0.4181567365359792,
+      "learning_rate": 4.7310558229290685e-06,
+      "loss": 0.6184,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9167662285941856,
+      "grad_norm": 0.3854924393243661,
+      "learning_rate": 4.730898274291492e-06,
+      "loss": 0.6234,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9170317270675694,
+      "grad_norm": 0.4100989351001039,
+      "learning_rate": 4.7307406821459364e-06,
+      "loss": 0.607,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9172972255409532,
+      "grad_norm": 0.4339623701841104,
+      "learning_rate": 4.730583046495477e-06,
+      "loss": 0.626,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9175627240143369,
+      "grad_norm": 0.3911102828515388,
+      "learning_rate": 4.730425367343186e-06,
+      "loss": 0.5989,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9178282224877207,
+      "grad_norm": 0.4067230998516061,
+      "learning_rate": 4.730267644692139e-06,
+      "loss": 0.6503,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9180937209611044,
+      "grad_norm": 0.40571298322688043,
+      "learning_rate": 4.7301098785454124e-06,
+      "loss": 0.617,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9183592194344883,
+      "grad_norm": 0.391348128842071,
+      "learning_rate": 4.729952068906084e-06,
+      "loss": 0.6389,
+      "step": 3459
+    },
+    {
+      "epoch": 0.918624717907872,
+      "grad_norm": 0.39287424854153175,
+      "learning_rate": 4.729794215777229e-06,
+      "loss": 0.6264,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9188902163812558,
+      "grad_norm": 0.3866999746446962,
+      "learning_rate": 4.729636319161928e-06,
+      "loss": 0.6513,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9191557148546395,
+      "grad_norm": 0.39238238205886716,
+      "learning_rate": 4.72947837906326e-06,
+      "loss": 0.5746,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9194212133280234,
+      "grad_norm": 0.390307595239034,
+      "learning_rate": 4.729320395484304e-06,
+      "loss": 0.6036,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9196867118014072,
+      "grad_norm": 0.3843938231362498,
+      "learning_rate": 4.729162368428143e-06,
+      "loss": 0.5853,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9199522102747909,
+      "grad_norm": 0.37906986999352327,
+      "learning_rate": 4.729004297897856e-06,
+      "loss": 0.6009,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9202177087481747,
+      "grad_norm": 0.39581950280073036,
+      "learning_rate": 4.72884618389653e-06,
+      "loss": 0.6245,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9204832072215585,
+      "grad_norm": 0.37726386807208134,
+      "learning_rate": 4.728688026427245e-06,
+      "loss": 0.6221,
+      "step": 3467
+    },
+    {
+      "epoch": 0.9207487056949423,
+      "grad_norm": 0.3869628549915921,
+      "learning_rate": 4.7285298254930865e-06,
+      "loss": 0.5813,
+      "step": 3468
+    },
+    {
+      "epoch": 0.921014204168326,
+      "grad_norm": 0.39145864774959555,
+      "learning_rate": 4.728371581097141e-06,
+      "loss": 0.6027,
+      "step": 3469
+    },
+    {
+      "epoch": 0.9212797026417098,
+      "grad_norm": 0.38039074340293794,
+      "learning_rate": 4.7282132932424926e-06,
+      "loss": 0.6149,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9215452011150936,
+      "grad_norm": 0.3999847100913083,
+      "learning_rate": 4.72805496193223e-06,
+      "loss": 0.6304,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9218106995884774,
+      "grad_norm": 0.39635395889112085,
+      "learning_rate": 4.72789658716944e-06,
+      "loss": 0.6612,
+      "step": 3472
+    },
+    {
+      "epoch": 0.9220761980618611,
+      "grad_norm": 0.3991550414705968,
+      "learning_rate": 4.7277381689572125e-06,
+      "loss": 0.6343,
+      "step": 3473
+    },
+    {
+      "epoch": 0.9223416965352449,
+      "grad_norm": 0.3798025780095887,
+      "learning_rate": 4.727579707298636e-06,
+      "loss": 0.623,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9226071950086288,
+      "grad_norm": 0.37644007998739926,
+      "learning_rate": 4.727421202196802e-06,
+      "loss": 0.6183,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9228726934820125,
+      "grad_norm": 0.3939535516255843,
+      "learning_rate": 4.727262653654799e-06,
+      "loss": 0.5954,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9231381919553963,
+      "grad_norm": 0.38868606381712056,
+      "learning_rate": 4.727104061675722e-06,
+      "loss": 0.6121,
+      "step": 3477
+    },
+    {
+      "epoch": 0.92340369042878,
+      "grad_norm": 0.3985793206729489,
+      "learning_rate": 4.726945426262662e-06,
+      "loss": 0.6002,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9236691889021638,
+      "grad_norm": 0.3871942692114719,
+      "learning_rate": 4.726786747418715e-06,
+      "loss": 0.6089,
+      "step": 3479
+    },
+    {
+      "epoch": 0.9239346873755476,
+      "grad_norm": 0.39136421575701336,
+      "learning_rate": 4.726628025146974e-06,
+      "loss": 0.6353,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9242001858489314,
+      "grad_norm": 0.3840020408134472,
+      "learning_rate": 4.726469259450534e-06,
+      "loss": 0.5982,
+      "step": 3481
+    },
+    {
+      "epoch": 0.9244656843223151,
+      "grad_norm": 0.40077545721314484,
+      "learning_rate": 4.726310450332493e-06,
+      "loss": 0.6326,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9247311827956989,
+      "grad_norm": 0.39902835161834627,
+      "learning_rate": 4.726151597795947e-06,
+      "loss": 0.6385,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9249966812690827,
+      "grad_norm": 0.3879133606947771,
+      "learning_rate": 4.725992701843994e-06,
+      "loss": 0.606,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9252621797424665,
+      "grad_norm": 0.41207641225646685,
+      "learning_rate": 4.725833762479733e-06,
+      "loss": 0.6193,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9255276782158502,
+      "grad_norm": 0.3849465427336246,
+      "learning_rate": 4.725674779706264e-06,
+      "loss": 0.6276,
+      "step": 3486
+    },
+    {
+      "epoch": 0.925793176689234,
+      "grad_norm": 0.40921801976671845,
+      "learning_rate": 4.725515753526688e-06,
+      "loss": 0.5777,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9260586751626179,
+      "grad_norm": 0.3824138209234103,
+      "learning_rate": 4.725356683944106e-06,
+      "loss": 0.586,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9263241736360016,
+      "grad_norm": 0.3880008217806217,
+      "learning_rate": 4.725197570961619e-06,
+      "loss": 0.5813,
+      "step": 3489
+    },
+    {
+      "epoch": 0.9265896721093854,
+      "grad_norm": 0.3988359982285484,
+      "learning_rate": 4.725038414582331e-06,
+      "loss": 0.5844,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9268551705827691,
+      "grad_norm": 0.41519921823890016,
+      "learning_rate": 4.724879214809347e-06,
+      "loss": 0.6233,
+      "step": 3491
+    },
+    {
+      "epoch": 0.927120669056153,
+      "grad_norm": 0.3898478397602767,
+      "learning_rate": 4.724719971645771e-06,
+      "loss": 0.6502,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9273861675295367,
+      "grad_norm": 0.40898787006693804,
+      "learning_rate": 4.724560685094708e-06,
+      "loss": 0.6206,
+      "step": 3493
+    },
+    {
+      "epoch": 0.9276516660029205,
+      "grad_norm": 0.38256678885372414,
+      "learning_rate": 4.724401355159265e-06,
+      "loss": 0.6306,
+      "step": 3494
+    },
+    {
+      "epoch": 0.9279171644763042,
+      "grad_norm": 0.39631642627366565,
+      "learning_rate": 4.724241981842549e-06,
+      "loss": 0.6371,
+      "step": 3495
+    },
+    {
+      "epoch": 0.928182662949688,
+      "grad_norm": 0.3744770038402985,
+      "learning_rate": 4.724082565147669e-06,
+      "loss": 0.5989,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9284481614230718,
+      "grad_norm": 0.3959072486033822,
+      "learning_rate": 4.723923105077733e-06,
+      "loss": 0.6623,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9287136598964556,
+      "grad_norm": 0.4235396707397403,
+      "learning_rate": 4.7237636016358515e-06,
+      "loss": 0.6276,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9289791583698394,
+      "grad_norm": 0.40392340729904874,
+      "learning_rate": 4.723604054825135e-06,
+      "loss": 0.6115,
+      "step": 3499
+    },
+    {
+      "epoch": 0.9292446568432231,
+      "grad_norm": 0.38258362505809135,
+      "learning_rate": 4.723444464648696e-06,
+      "loss": 0.6022,
+      "step": 3500
+    },
+    {
+      "epoch": 0.929510155316607,
+      "grad_norm": 0.3790529906787704,
+      "learning_rate": 4.723284831109645e-06,
+      "loss": 0.5948,
+      "step": 3501
+    },
+    {
+      "epoch": 0.9297756537899907,
+      "grad_norm": 0.3840540623424871,
+      "learning_rate": 4.723125154211097e-06,
+      "loss": 0.6179,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9300411522633745,
+      "grad_norm": 0.408530282230814,
+      "learning_rate": 4.7229654339561645e-06,
+      "loss": 0.5846,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9303066507367582,
+      "grad_norm": 0.3960898881412602,
+      "learning_rate": 4.722805670347963e-06,
+      "loss": 0.5946,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9305721492101421,
+      "grad_norm": 0.38204270054388545,
+      "learning_rate": 4.72264586338961e-06,
+      "loss": 0.6027,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9308376476835258,
+      "grad_norm": 0.38131700057876367,
+      "learning_rate": 4.72248601308422e-06,
+      "loss": 0.593,
+      "step": 3506
+    },
+    {
+      "epoch": 0.9311031461569096,
+      "grad_norm": 0.3896561283967658,
+      "learning_rate": 4.722326119434911e-06,
+      "loss": 0.6047,
+      "step": 3507
+    },
+    {
+      "epoch": 0.9313686446302933,
+      "grad_norm": 0.3915952652728005,
+      "learning_rate": 4.722166182444801e-06,
+      "loss": 0.6572,
+      "step": 3508
+    },
+    {
+      "epoch": 0.9316341431036772,
+      "grad_norm": 0.42333469278626573,
+      "learning_rate": 4.72200620211701e-06,
+      "loss": 0.6157,
+      "step": 3509
+    },
+    {
+      "epoch": 0.931899641577061,
+      "grad_norm": 0.39543587323385704,
+      "learning_rate": 4.721846178454658e-06,
+      "loss": 0.5714,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9321651400504447,
+      "grad_norm": 0.3845789903687893,
+      "learning_rate": 4.721686111460866e-06,
+      "loss": 0.6068,
+      "step": 3511
+    },
+    {
+      "epoch": 0.9324306385238285,
+      "grad_norm": 0.41802367665992934,
+      "learning_rate": 4.721526001138754e-06,
+      "loss": 0.6061,
+      "step": 3512
+    },
+    {
+      "epoch": 0.9326961369972122,
+      "grad_norm": 0.40571467836760294,
+      "learning_rate": 4.7213658474914465e-06,
+      "loss": 0.6382,
+      "step": 3513
+    },
+    {
+      "epoch": 0.9329616354705961,
+      "grad_norm": 0.3928869829662483,
+      "learning_rate": 4.721205650522066e-06,
+      "loss": 0.6213,
+      "step": 3514
+    },
+    {
+      "epoch": 0.9332271339439798,
+      "grad_norm": 0.3919667381453046,
+      "learning_rate": 4.721045410233737e-06,
+      "loss": 0.6364,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9334926324173636,
+      "grad_norm": 0.3972783164202163,
+      "learning_rate": 4.720885126629584e-06,
+      "loss": 0.6023,
+      "step": 3516
+    },
+    {
+      "epoch": 0.9337581308907473,
+      "grad_norm": 0.407462513465511,
+      "learning_rate": 4.720724799712734e-06,
+      "loss": 0.621,
+      "step": 3517
+    },
+    {
+      "epoch": 0.9340236293641312,
+      "grad_norm": 0.43232414172633815,
+      "learning_rate": 4.720564429486312e-06,
+      "loss": 0.5754,
+      "step": 3518
+    },
+    {
+      "epoch": 0.9342891278375149,
+      "grad_norm": 0.3905489312484764,
+      "learning_rate": 4.720404015953448e-06,
+      "loss": 0.6405,
+      "step": 3519
+    },
+    {
+      "epoch": 0.9345546263108987,
+      "grad_norm": 0.38075055483398174,
+      "learning_rate": 4.720243559117268e-06,
+      "loss": 0.6125,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9348201247842824,
+      "grad_norm": 0.44666240024584875,
+      "learning_rate": 4.720083058980903e-06,
+      "loss": 0.5968,
+      "step": 3521
+    },
+    {
+      "epoch": 0.9350856232576663,
+      "grad_norm": 0.38000039164090604,
+      "learning_rate": 4.719922515547483e-06,
+      "loss": 0.6356,
+      "step": 3522
+    },
+    {
+      "epoch": 0.9353511217310501,
+      "grad_norm": 0.39400085528512446,
+      "learning_rate": 4.719761928820139e-06,
+      "loss": 0.6301,
+      "step": 3523
+    },
+    {
+      "epoch": 0.9356166202044338,
+      "grad_norm": 0.37905520336141546,
+      "learning_rate": 4.719601298802002e-06,
+      "loss": 0.5994,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9358821186778176,
+      "grad_norm": 0.3853796510866678,
+      "learning_rate": 4.719440625496205e-06,
+      "loss": 0.6522,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9361476171512014,
+      "grad_norm": 0.3854573658366218,
+      "learning_rate": 4.719279908905882e-06,
+      "loss": 0.6196,
+      "step": 3526
+    },
+    {
+      "epoch": 0.9364131156245852,
+      "grad_norm": 0.3783878744935739,
+      "learning_rate": 4.719119149034166e-06,
+      "loss": 0.6185,
+      "step": 3527
+    },
+    {
+      "epoch": 0.9366786140979689,
+      "grad_norm": 0.38443751951282223,
+      "learning_rate": 4.718958345884195e-06,
+      "loss": 0.6151,
+      "step": 3528
+    },
+    {
+      "epoch": 0.9369441125713527,
+      "grad_norm": 0.3899471549058231,
+      "learning_rate": 4.718797499459102e-06,
+      "loss": 0.6013,
+      "step": 3529
+    },
+    {
+      "epoch": 0.9372096110447365,
+      "grad_norm": 0.42267009026821517,
+      "learning_rate": 4.718636609762025e-06,
+      "loss": 0.646,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9374751095181203,
+      "grad_norm": 0.39049097494025076,
+      "learning_rate": 4.718475676796103e-06,
+      "loss": 0.6459,
+      "step": 3531
+    },
+    {
+      "epoch": 0.937740607991504,
+      "grad_norm": 0.39948159005145745,
+      "learning_rate": 4.718314700564473e-06,
+      "loss": 0.626,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9380061064648878,
+      "grad_norm": 0.38862287755708075,
+      "learning_rate": 4.718153681070276e-06,
+      "loss": 0.6236,
+      "step": 3533
+    },
+    {
+      "epoch": 0.9382716049382716,
+      "grad_norm": 0.3844487538483707,
+      "learning_rate": 4.717992618316649e-06,
+      "loss": 0.6045,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9385371034116554,
+      "grad_norm": 0.3840615286530079,
+      "learning_rate": 4.717831512306737e-06,
+      "loss": 0.5714,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9388026018850392,
+      "grad_norm": 0.38626787538083596,
+      "learning_rate": 4.71767036304368e-06,
+      "loss": 0.6267,
+      "step": 3536
+    },
+    {
+      "epoch": 0.9390681003584229,
+      "grad_norm": 0.3831602323571539,
+      "learning_rate": 4.7175091705306206e-06,
+      "loss": 0.6181,
+      "step": 3537
+    },
+    {
+      "epoch": 0.9393335988318067,
+      "grad_norm": 0.381564441987076,
+      "learning_rate": 4.7173479347707044e-06,
+      "loss": 0.6013,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9395990973051905,
+      "grad_norm": 0.39919109369221556,
+      "learning_rate": 4.717186655767073e-06,
+      "loss": 0.6025,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9398645957785743,
+      "grad_norm": 0.3831387321488225,
+      "learning_rate": 4.717025333522874e-06,
+      "loss": 0.6001,
+      "step": 3540
+    },
+    {
+      "epoch": 0.940130094251958,
+      "grad_norm": 0.3785704153450314,
+      "learning_rate": 4.716863968041252e-06,
+      "loss": 0.6078,
+      "step": 3541
+    },
+    {
+      "epoch": 0.9403955927253418,
+      "grad_norm": 0.38582635301175494,
+      "learning_rate": 4.716702559325356e-06,
+      "loss": 0.5849,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9406610911987257,
+      "grad_norm": 0.39162354418307566,
+      "learning_rate": 4.716541107378332e-06,
+      "loss": 0.6242,
+      "step": 3543
+    },
+    {
+      "epoch": 0.9409265896721094,
+      "grad_norm": 0.3846172482049233,
+      "learning_rate": 4.716379612203329e-06,
+      "loss": 0.6014,
+      "step": 3544
+    },
+    {
+      "epoch": 0.9411920881454932,
+      "grad_norm": 0.38280507041972966,
+      "learning_rate": 4.716218073803498e-06,
+      "loss": 0.602,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9414575866188769,
+      "grad_norm": 0.3747513761882635,
+      "learning_rate": 4.716056492181987e-06,
+      "loss": 0.6126,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9417230850922608,
+      "grad_norm": 0.42223261439510645,
+      "learning_rate": 4.715894867341949e-06,
+      "loss": 0.6025,
+      "step": 3547
+    },
+    {
+      "epoch": 0.9419885835656445,
+      "grad_norm": 0.38803614841312795,
+      "learning_rate": 4.715733199286536e-06,
+      "loss": 0.6299,
+      "step": 3548
+    },
+    {
+      "epoch": 0.9422540820390283,
+      "grad_norm": 0.39536417979035315,
+      "learning_rate": 4.715571488018901e-06,
+      "loss": 0.6409,
+      "step": 3549
+    },
+    {
+      "epoch": 0.942519580512412,
+      "grad_norm": 0.38119715322976605,
+      "learning_rate": 4.715409733542196e-06,
+      "loss": 0.6297,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9427850789857958,
+      "grad_norm": 0.3912170819860612,
+      "learning_rate": 4.7152479358595785e-06,
+      "loss": 0.6347,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9430505774591796,
+      "grad_norm": 0.4024684886020273,
+      "learning_rate": 4.7150860949742e-06,
+      "loss": 0.6081,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9433160759325634,
+      "grad_norm": 0.4092764083558984,
+      "learning_rate": 4.714924210889221e-06,
+      "loss": 0.5879,
+      "step": 3553
+    },
+    {
+      "epoch": 0.9435815744059471,
+      "grad_norm": 0.37695650207026915,
+      "learning_rate": 4.714762283607797e-06,
+      "loss": 0.6049,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9438470728793309,
+      "grad_norm": 0.4243002297754266,
+      "learning_rate": 4.714600313133085e-06,
+      "loss": 0.5838,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9441125713527148,
+      "grad_norm": 0.41261227594966987,
+      "learning_rate": 4.714438299468245e-06,
+      "loss": 0.6082,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9443780698260985,
+      "grad_norm": 0.38174420139911713,
+      "learning_rate": 4.714276242616437e-06,
+      "loss": 0.6132,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9446435682994823,
+      "grad_norm": 0.3871002909739619,
+      "learning_rate": 4.71411414258082e-06,
+      "loss": 0.6121,
+      "step": 3558
+    },
+    {
+      "epoch": 0.944909066772866,
+      "grad_norm": 0.3986292354860389,
+      "learning_rate": 4.7139519993645555e-06,
+      "loss": 0.6046,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9451745652462499,
+      "grad_norm": 0.3881081172474212,
+      "learning_rate": 4.713789812970807e-06,
+      "loss": 0.6115,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9454400637196336,
+      "grad_norm": 0.3868471666006034,
+      "learning_rate": 4.713627583402737e-06,
+      "loss": 0.5882,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9457055621930174,
+      "grad_norm": 0.3793587646580147,
+      "learning_rate": 4.713465310663509e-06,
+      "loss": 0.6272,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9459710606664011,
+      "grad_norm": 0.3721766806443814,
+      "learning_rate": 4.713302994756288e-06,
+      "loss": 0.5809,
+      "step": 3563
+    },
+    {
+      "epoch": 0.946236559139785,
+      "grad_norm": 0.37253670227246405,
+      "learning_rate": 4.71314063568424e-06,
+      "loss": 0.6213,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9465020576131687,
+      "grad_norm": 0.3856079119107004,
+      "learning_rate": 4.712978233450531e-06,
+      "loss": 0.6173,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9467675560865525,
+      "grad_norm": 0.37916228674395314,
+      "learning_rate": 4.712815788058327e-06,
+      "loss": 0.5932,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9470330545599362,
+      "grad_norm": 0.3683105933076519,
+      "learning_rate": 4.712653299510798e-06,
+      "loss": 0.5897,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9472985530333201,
+      "grad_norm": 0.3812396209762039,
+      "learning_rate": 4.712490767811111e-06,
+      "loss": 0.6413,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9475640515067039,
+      "grad_norm": 0.388790949165087,
+      "learning_rate": 4.712328192962438e-06,
+      "loss": 0.6533,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9478295499800876,
+      "grad_norm": 0.3969476032613877,
+      "learning_rate": 4.712165574967949e-06,
+      "loss": 0.6078,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9480950484534714,
+      "grad_norm": 0.3753287846734859,
+      "learning_rate": 4.712002913830815e-06,
+      "loss": 0.5838,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9483605469268551,
+      "grad_norm": 0.396556859423283,
+      "learning_rate": 4.711840209554207e-06,
+      "loss": 0.6016,
+      "step": 3572
+    },
+    {
+      "epoch": 0.948626045400239,
+      "grad_norm": 0.4065613362655618,
+      "learning_rate": 4.711677462141299e-06,
+      "loss": 0.5923,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9488915438736227,
+      "grad_norm": 0.3901875041360443,
+      "learning_rate": 4.711514671595267e-06,
+      "loss": 0.647,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9491570423470065,
+      "grad_norm": 0.3806874918242015,
+      "learning_rate": 4.711351837919283e-06,
+      "loss": 0.606,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9494225408203902,
+      "grad_norm": 0.3737996868771382,
+      "learning_rate": 4.7111889611165254e-06,
+      "loss": 0.5689,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9496880392937741,
+      "grad_norm": 0.3845205504765776,
+      "learning_rate": 4.711026041190168e-06,
+      "loss": 0.6045,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9499535377671579,
+      "grad_norm": 0.3820802234953147,
+      "learning_rate": 4.710863078143389e-06,
+      "loss": 0.6104,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9502190362405416,
+      "grad_norm": 0.3823997520337332,
+      "learning_rate": 4.710700071979367e-06,
+      "loss": 0.6073,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9504845347139254,
+      "grad_norm": 0.39558462192664146,
+      "learning_rate": 4.710537022701281e-06,
+      "loss": 0.5872,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9507500331873092,
+      "grad_norm": 0.383874294746993,
+      "learning_rate": 4.710373930312311e-06,
+      "loss": 0.6032,
+      "step": 3581
+    },
+    {
+      "epoch": 0.951015531660693,
+      "grad_norm": 0.39124421725543956,
+      "learning_rate": 4.710210794815637e-06,
+      "loss": 0.6331,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9512810301340767,
+      "grad_norm": 0.39622892804406773,
+      "learning_rate": 4.710047616214441e-06,
+      "loss": 0.6118,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9515465286074605,
+      "grad_norm": 0.37741913373008423,
+      "learning_rate": 4.709884394511906e-06,
+      "loss": 0.622,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9518120270808443,
+      "grad_norm": 0.38102130190077316,
+      "learning_rate": 4.7097211297112145e-06,
+      "loss": 0.6144,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9520775255542281,
+      "grad_norm": 0.38955337923468664,
+      "learning_rate": 4.70955782181555e-06,
+      "loss": 0.6316,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9523430240276118,
+      "grad_norm": 0.396638900754623,
+      "learning_rate": 4.709394470828099e-06,
+      "loss": 0.6534,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9526085225009956,
+      "grad_norm": 0.391803800569157,
+      "learning_rate": 4.709231076752045e-06,
+      "loss": 0.613,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9528740209743793,
+      "grad_norm": 0.3984269122357259,
+      "learning_rate": 4.709067639590576e-06,
+      "loss": 0.6321,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9531395194477632,
+      "grad_norm": 0.395700076282968,
+      "learning_rate": 4.708904159346881e-06,
+      "loss": 0.6013,
+      "step": 3590
+    },
+    {
+      "epoch": 0.953405017921147,
+      "grad_norm": 0.38068629903417356,
+      "learning_rate": 4.708740636024145e-06,
+      "loss": 0.6488,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9536705163945307,
+      "grad_norm": 0.39973489616064506,
+      "learning_rate": 4.708577069625559e-06,
+      "loss": 0.5966,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9539360148679145,
+      "grad_norm": 0.3909092586396629,
+      "learning_rate": 4.708413460154313e-06,
+      "loss": 0.6394,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9542015133412983,
+      "grad_norm": 0.40284589959583805,
+      "learning_rate": 4.708249807613597e-06,
+      "loss": 0.6098,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9544670118146821,
+      "grad_norm": 0.4022975164531979,
+      "learning_rate": 4.708086112006604e-06,
+      "loss": 0.5786,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9547325102880658,
+      "grad_norm": 0.37800906327107797,
+      "learning_rate": 4.707922373336524e-06,
+      "loss": 0.6603,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9549980087614496,
+      "grad_norm": 0.3904826018311421,
+      "learning_rate": 4.707758591606552e-06,
+      "loss": 0.6495,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9552635072348334,
+      "grad_norm": 0.38674877034434985,
+      "learning_rate": 4.7075947668198825e-06,
+      "loss": 0.593,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9555290057082172,
+      "grad_norm": 0.3914397071460796,
+      "learning_rate": 4.70743089897971e-06,
+      "loss": 0.5818,
+      "step": 3599
+    },
+    {
+      "epoch": 0.955794504181601,
+      "grad_norm": 0.3933866827836264,
+      "learning_rate": 4.70726698808923e-06,
+      "loss": 0.5861,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9560600026549847,
+      "grad_norm": 0.382664339751929,
+      "learning_rate": 4.707103034151639e-06,
+      "loss": 0.5691,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9563255011283686,
+      "grad_norm": 0.38044517979027814,
+      "learning_rate": 4.706939037170135e-06,
+      "loss": 0.5911,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9565909996017523,
+      "grad_norm": 0.39137606815892434,
+      "learning_rate": 4.706774997147917e-06,
+      "loss": 0.6073,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9568564980751361,
+      "grad_norm": 0.39381494532615724,
+      "learning_rate": 4.7066109140881824e-06,
+      "loss": 0.5978,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9571219965485198,
+      "grad_norm": 0.3825296235402476,
+      "learning_rate": 4.706446787994134e-06,
+      "loss": 0.6115,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9573874950219037,
+      "grad_norm": 0.38969170898807953,
+      "learning_rate": 4.706282618868969e-06,
+      "loss": 0.6323,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9576529934952874,
+      "grad_norm": 0.40403364898354477,
+      "learning_rate": 4.706118406715892e-06,
+      "loss": 0.6156,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9579184919686712,
+      "grad_norm": 0.38786804309152906,
+      "learning_rate": 4.7059541515381045e-06,
+      "loss": 0.6305,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9581839904420549,
+      "grad_norm": 0.38961955782479457,
+      "learning_rate": 4.70578985333881e-06,
+      "loss": 0.6283,
+      "step": 3609
+    },
+    {
+      "epoch": 0.9584494889154387,
+      "grad_norm": 0.38773616697041347,
+      "learning_rate": 4.705625512121213e-06,
+      "loss": 0.6125,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9587149873888225,
+      "grad_norm": 0.3865579178052979,
+      "learning_rate": 4.705461127888517e-06,
+      "loss": 0.601,
+      "step": 3611
+    },
+    {
+      "epoch": 0.9589804858622063,
+      "grad_norm": 0.39613489912379074,
+      "learning_rate": 4.70529670064393e-06,
+      "loss": 0.6227,
+      "step": 3612
+    },
+    {
+      "epoch": 0.95924598433559,
+      "grad_norm": 0.39270094445322157,
+      "learning_rate": 4.705132230390657e-06,
+      "loss": 0.607,
+      "step": 3613
+    },
+    {
+      "epoch": 0.9595114828089738,
+      "grad_norm": 0.3919473897579516,
+      "learning_rate": 4.704967717131907e-06,
+      "loss": 0.6279,
+      "step": 3614
+    },
+    {
+      "epoch": 0.9597769812823577,
+      "grad_norm": 0.3802434003911743,
+      "learning_rate": 4.704803160870888e-06,
+      "loss": 0.6512,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9600424797557414,
+      "grad_norm": 0.40046601659455683,
+      "learning_rate": 4.704638561610809e-06,
+      "loss": 0.5767,
+      "step": 3616
+    },
+    {
+      "epoch": 0.9603079782291252,
+      "grad_norm": 0.3880968097212626,
+      "learning_rate": 4.70447391935488e-06,
+      "loss": 0.6055,
+      "step": 3617
+    },
+    {
+      "epoch": 0.9605734767025089,
+      "grad_norm": 0.39250409904499617,
+      "learning_rate": 4.704309234106312e-06,
+      "loss": 0.6154,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9608389751758928,
+      "grad_norm": 0.38450001318202004,
+      "learning_rate": 4.704144505868318e-06,
+      "loss": 0.594,
+      "step": 3619
+    },
+    {
+      "epoch": 0.9611044736492765,
+      "grad_norm": 0.40137595561487843,
+      "learning_rate": 4.703979734644107e-06,
+      "loss": 0.5621,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9613699721226603,
+      "grad_norm": 0.38427758191234046,
+      "learning_rate": 4.703814920436897e-06,
+      "loss": 0.5736,
+      "step": 3621
+    },
+    {
+      "epoch": 0.961635470596044,
+      "grad_norm": 0.38947722918312283,
+      "learning_rate": 4.7036500632498995e-06,
+      "loss": 0.6168,
+      "step": 3622
+    },
+    {
+      "epoch": 0.9619009690694279,
+      "grad_norm": 0.39007761189984497,
+      "learning_rate": 4.703485163086331e-06,
+      "loss": 0.5841,
+      "step": 3623
+    },
+    {
+      "epoch": 0.9621664675428117,
+      "grad_norm": 0.3985843518177514,
+      "learning_rate": 4.7033202199494055e-06,
+      "loss": 0.6365,
+      "step": 3624
+    },
+    {
+      "epoch": 0.9624319660161954,
+      "grad_norm": 0.3891574803335972,
+      "learning_rate": 4.703155233842342e-06,
+      "loss": 0.6074,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9626974644895792,
+      "grad_norm": 0.3961272823075828,
+      "learning_rate": 4.7029902047683565e-06,
+      "loss": 0.6323,
+      "step": 3626
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 0.39100088914724596,
+      "learning_rate": 4.702825132730669e-06,
+      "loss": 0.5924,
+      "step": 3627
+    },
+    {
+      "epoch": 0.9632284614363468,
+      "grad_norm": 0.3890924820441184,
+      "learning_rate": 4.702660017732497e-06,
+      "loss": 0.6515,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9634939599097305,
+      "grad_norm": 0.38355615738109367,
+      "learning_rate": 4.7024948597770625e-06,
+      "loss": 0.62,
+      "step": 3629
+    },
+    {
+      "epoch": 0.9637594583831143,
+      "grad_norm": 0.3954073823720711,
+      "learning_rate": 4.702329658867585e-06,
+      "loss": 0.6397,
+      "step": 3630
+    },
+    {
+      "epoch": 0.964024956856498,
+      "grad_norm": 0.37543761854241386,
+      "learning_rate": 4.702164415007288e-06,
+      "loss": 0.5817,
+      "step": 3631
+    },
+    {
+      "epoch": 0.9642904553298819,
+      "grad_norm": 0.3854728665577839,
+      "learning_rate": 4.701999128199392e-06,
+      "loss": 0.6234,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9645559538032656,
+      "grad_norm": 0.40362742567571847,
+      "learning_rate": 4.701833798447121e-06,
+      "loss": 0.6091,
+      "step": 3633
+    },
+    {
+      "epoch": 0.9648214522766494,
+      "grad_norm": 0.3933445105415557,
+      "learning_rate": 4.701668425753701e-06,
+      "loss": 0.6076,
+      "step": 3634
+    },
+    {
+      "epoch": 0.9650869507500331,
+      "grad_norm": 0.380807769650326,
+      "learning_rate": 4.701503010122356e-06,
+      "loss": 0.5941,
+      "step": 3635
+    },
+    {
+      "epoch": 0.965352449223417,
+      "grad_norm": 0.39856075635397675,
+      "learning_rate": 4.701337551556312e-06,
+      "loss": 0.5767,
+      "step": 3636
+    },
+    {
+      "epoch": 0.9656179476968008,
+      "grad_norm": 0.37781771022776084,
+      "learning_rate": 4.701172050058795e-06,
+      "loss": 0.5825,
+      "step": 3637
+    },
+    {
+      "epoch": 0.9658834461701845,
+      "grad_norm": 0.3889588409640584,
+      "learning_rate": 4.701006505633035e-06,
+      "loss": 0.642,
+      "step": 3638
+    },
+    {
+      "epoch": 0.9661489446435683,
+      "grad_norm": 0.3896655514807649,
+      "learning_rate": 4.700840918282258e-06,
+      "loss": 0.5552,
+      "step": 3639
+    },
+    {
+      "epoch": 0.9664144431169521,
+      "grad_norm": 0.391716893041241,
+      "learning_rate": 4.700675288009695e-06,
+      "loss": 0.6201,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9666799415903359,
+      "grad_norm": 0.3936588621740142,
+      "learning_rate": 4.7005096148185765e-06,
+      "loss": 0.6386,
+      "step": 3641
+    },
+    {
+      "epoch": 0.9669454400637196,
+      "grad_norm": 0.37889689828095957,
+      "learning_rate": 4.700343898712132e-06,
+      "loss": 0.6298,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9672109385371034,
+      "grad_norm": 0.4091429640071916,
+      "learning_rate": 4.7001781396935945e-06,
+      "loss": 0.6398,
+      "step": 3643
+    },
+    {
+      "epoch": 0.9674764370104872,
+      "grad_norm": 0.3883545639185824,
+      "learning_rate": 4.700012337766196e-06,
+      "loss": 0.6239,
+      "step": 3644
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.38225271118778803,
+      "learning_rate": 4.69984649293317e-06,
+      "loss": 0.5958,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9680074339572547,
+      "grad_norm": 0.39137614154896344,
+      "learning_rate": 4.699680605197751e-06,
+      "loss": 0.6224,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9682729324306385,
+      "grad_norm": 0.3846780556433604,
+      "learning_rate": 4.6995146745631746e-06,
+      "loss": 0.6126,
+      "step": 3647
+    },
+    {
+      "epoch": 0.9685384309040223,
+      "grad_norm": 0.3912733934965504,
+      "learning_rate": 4.699348701032678e-06,
+      "loss": 0.634,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9688039293774061,
+      "grad_norm": 0.40683308534234996,
+      "learning_rate": 4.699182684609495e-06,
+      "loss": 0.6115,
+      "step": 3649
+    },
+    {
+      "epoch": 0.9690694278507899,
+      "grad_norm": 0.3889780902116469,
+      "learning_rate": 4.699016625296866e-06,
+      "loss": 0.5928,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9693349263241736,
+      "grad_norm": 0.3866846816002472,
+      "learning_rate": 4.69885052309803e-06,
+      "loss": 0.6016,
+      "step": 3651
+    },
+    {
+      "epoch": 0.9696004247975574,
+      "grad_norm": 0.3870270892617259,
+      "learning_rate": 4.698684378016223e-06,
+      "loss": 0.6454,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9698659232709412,
+      "grad_norm": 0.3793337172442884,
+      "learning_rate": 4.698518190054688e-06,
+      "loss": 0.5917,
+      "step": 3653
+    },
+    {
+      "epoch": 0.970131421744325,
+      "grad_norm": 0.3943488726168635,
+      "learning_rate": 4.698351959216666e-06,
+      "loss": 0.6099,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9703969202177087,
+      "grad_norm": 0.3949298885078147,
+      "learning_rate": 4.698185685505397e-06,
+      "loss": 0.6284,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9706624186910925,
+      "grad_norm": 0.40927389400452,
+      "learning_rate": 4.698019368924126e-06,
+      "loss": 0.634,
+      "step": 3656
+    },
+    {
+      "epoch": 0.9709279171644764,
+      "grad_norm": 0.3960272796831328,
+      "learning_rate": 4.697853009476096e-06,
+      "loss": 0.5853,
+      "step": 3657
+    },
+    {
+      "epoch": 0.9711934156378601,
+      "grad_norm": 0.3836826859623769,
+      "learning_rate": 4.69768660716455e-06,
+      "loss": 0.5875,
+      "step": 3658
+    },
+    {
+      "epoch": 0.9714589141112439,
+      "grad_norm": 0.3849274510743423,
+      "learning_rate": 4.697520161992735e-06,
+      "loss": 0.6226,
+      "step": 3659
+    },
+    {
+      "epoch": 0.9717244125846276,
+      "grad_norm": 0.39910614750810963,
+      "learning_rate": 4.6973536739638965e-06,
+      "loss": 0.6174,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9719899110580115,
+      "grad_norm": 0.4185146628531675,
+      "learning_rate": 4.697187143081281e-06,
+      "loss": 0.6227,
+      "step": 3661
+    },
+    {
+      "epoch": 0.9722554095313952,
+      "grad_norm": 0.3910465796237792,
+      "learning_rate": 4.697020569348136e-06,
+      "loss": 0.6226,
+      "step": 3662
+    },
+    {
+      "epoch": 0.972520908004779,
+      "grad_norm": 0.3917665963744819,
+      "learning_rate": 4.696853952767711e-06,
+      "loss": 0.6157,
+      "step": 3663
+    },
+    {
+      "epoch": 0.9727864064781627,
+      "grad_norm": 0.3969503845979608,
+      "learning_rate": 4.6966872933432565e-06,
+      "loss": 0.6252,
+      "step": 3664
+    },
+    {
+      "epoch": 0.9730519049515465,
+      "grad_norm": 0.4105211003899395,
+      "learning_rate": 4.69652059107802e-06,
+      "loss": 0.5931,
+      "step": 3665
+    },
+    {
+      "epoch": 0.9733174034249303,
+      "grad_norm": 0.3944654723818686,
+      "learning_rate": 4.696353845975253e-06,
+      "loss": 0.6015,
+      "step": 3666
+    },
+    {
+      "epoch": 0.9735829018983141,
+      "grad_norm": 0.38634470571483776,
+      "learning_rate": 4.69618705803821e-06,
+      "loss": 0.5941,
+      "step": 3667
+    },
+    {
+      "epoch": 0.9738484003716978,
+      "grad_norm": 0.3918507831216075,
+      "learning_rate": 4.696020227270142e-06,
+      "loss": 0.6074,
+      "step": 3668
+    },
+    {
+      "epoch": 0.9741138988450816,
+      "grad_norm": 0.4524254406535516,
+      "learning_rate": 4.6958533536743025e-06,
+      "loss": 0.5836,
+      "step": 3669
+    },
+    {
+      "epoch": 0.9743793973184655,
+      "grad_norm": 0.39216272913247,
+      "learning_rate": 4.695686437253946e-06,
+      "loss": 0.5988,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9746448957918492,
+      "grad_norm": 0.39774952287863624,
+      "learning_rate": 4.6955194780123285e-06,
+      "loss": 0.6198,
+      "step": 3671
+    },
+    {
+      "epoch": 0.974910394265233,
+      "grad_norm": 0.5575594701577312,
+      "learning_rate": 4.695352475952706e-06,
+      "loss": 0.5818,
+      "step": 3672
+    },
+    {
+      "epoch": 0.9751758927386167,
+      "grad_norm": 0.44469383728107376,
+      "learning_rate": 4.695185431078334e-06,
+      "loss": 0.6164,
+      "step": 3673
+    },
+    {
+      "epoch": 0.9754413912120006,
+      "grad_norm": 0.3867963703664239,
+      "learning_rate": 4.695018343392472e-06,
+      "loss": 0.5924,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9757068896853843,
+      "grad_norm": 0.43739384560926065,
+      "learning_rate": 4.694851212898378e-06,
+      "loss": 0.6295,
+      "step": 3675
+    },
+    {
+      "epoch": 0.9759723881587681,
+      "grad_norm": 0.5094909347727954,
+      "learning_rate": 4.694684039599311e-06,
+      "loss": 0.6198,
+      "step": 3676
+    },
+    {
+      "epoch": 0.9762378866321518,
+      "grad_norm": 0.5472220226418029,
+      "learning_rate": 4.694516823498532e-06,
+      "loss": 0.6027,
+      "step": 3677
+    },
+    {
+      "epoch": 0.9765033851055357,
+      "grad_norm": 0.40021750006934775,
+      "learning_rate": 4.694349564599301e-06,
+      "loss": 0.6225,
+      "step": 3678
+    },
+    {
+      "epoch": 0.9767688835789194,
+      "grad_norm": 0.5479388666932355,
+      "learning_rate": 4.694182262904882e-06,
+      "loss": 0.5762,
+      "step": 3679
+    },
+    {
+      "epoch": 0.9770343820523032,
+      "grad_norm": 0.5719635976414318,
+      "learning_rate": 4.6940149184185355e-06,
+      "loss": 0.5885,
+      "step": 3680
+    },
+    {
+      "epoch": 0.977299880525687,
+      "grad_norm": 0.4073858968239127,
+      "learning_rate": 4.693847531143528e-06,
+      "loss": 0.6071,
+      "step": 3681
+    },
+    {
+      "epoch": 0.9775653789990707,
+      "grad_norm": 0.3951570872686861,
+      "learning_rate": 4.69368010108312e-06,
+      "loss": 0.6493,
+      "step": 3682
+    },
+    {
+      "epoch": 0.9778308774724546,
+      "grad_norm": 0.4835789842734578,
+      "learning_rate": 4.69351262824058e-06,
+      "loss": 0.6511,
+      "step": 3683
+    },
+    {
+      "epoch": 0.9780963759458383,
+      "grad_norm": 0.3988660593653177,
+      "learning_rate": 4.693345112619173e-06,
+      "loss": 0.6318,
+      "step": 3684
+    },
+    {
+      "epoch": 0.9783618744192221,
+      "grad_norm": 0.43644186892211223,
+      "learning_rate": 4.693177554222167e-06,
+      "loss": 0.5959,
+      "step": 3685
+    },
+    {
+      "epoch": 0.9786273728926058,
+      "grad_norm": 0.3849105238605806,
+      "learning_rate": 4.6930099530528275e-06,
+      "loss": 0.5945,
+      "step": 3686
+    },
+    {
+      "epoch": 0.9788928713659897,
+      "grad_norm": 0.39202125675770977,
+      "learning_rate": 4.692842309114425e-06,
+      "loss": 0.6155,
+      "step": 3687
+    },
+    {
+      "epoch": 0.9791583698393734,
+      "grad_norm": 0.3937344635542589,
+      "learning_rate": 4.692674622410229e-06,
+      "loss": 0.629,
+      "step": 3688
+    },
+    {
+      "epoch": 0.9794238683127572,
+      "grad_norm": 0.43389821032805437,
+      "learning_rate": 4.692506892943508e-06,
+      "loss": 0.5864,
+      "step": 3689
+    },
+    {
+      "epoch": 0.9796893667861409,
+      "grad_norm": 0.4155245302247665,
+      "learning_rate": 4.6923391207175356e-06,
+      "loss": 0.629,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9799548652595248,
+      "grad_norm": 0.39114439522055067,
+      "learning_rate": 4.692171305735582e-06,
+      "loss": 0.6099,
+      "step": 3691
+    },
+    {
+      "epoch": 0.9802203637329086,
+      "grad_norm": 0.3828856583766105,
+      "learning_rate": 4.692003448000921e-06,
+      "loss": 0.5884,
+      "step": 3692
+    },
+    {
+      "epoch": 0.9804858622062923,
+      "grad_norm": 0.3993874435271273,
+      "learning_rate": 4.691835547516826e-06,
+      "loss": 0.6002,
+      "step": 3693
+    },
+    {
+      "epoch": 0.980751360679676,
+      "grad_norm": 0.43845103690590476,
+      "learning_rate": 4.691667604286571e-06,
+      "loss": 0.5995,
+      "step": 3694
+    },
+    {
+      "epoch": 0.9810168591530599,
+      "grad_norm": 0.39052978718718556,
+      "learning_rate": 4.691499618313431e-06,
+      "loss": 0.6299,
+      "step": 3695
+    },
+    {
+      "epoch": 0.9812823576264437,
+      "grad_norm": 0.37746046189224663,
+      "learning_rate": 4.691331589600683e-06,
+      "loss": 0.6394,
+      "step": 3696
+    },
+    {
+      "epoch": 0.9815478560998274,
+      "grad_norm": 0.40066545179781904,
+      "learning_rate": 4.691163518151604e-06,
+      "loss": 0.6313,
+      "step": 3697
+    },
+    {
+      "epoch": 0.9818133545732112,
+      "grad_norm": 0.3766838231710626,
+      "learning_rate": 4.690995403969471e-06,
+      "loss": 0.6082,
+      "step": 3698
+    },
+    {
+      "epoch": 0.982078853046595,
+      "grad_norm": 0.39961014282509283,
+      "learning_rate": 4.690827247057563e-06,
+      "loss": 0.5669,
+      "step": 3699
+    },
+    {
+      "epoch": 0.9823443515199788,
+      "grad_norm": 0.40501165983081905,
+      "learning_rate": 4.69065904741916e-06,
+      "loss": 0.6391,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9826098499933625,
+      "grad_norm": 0.3857307020929534,
+      "learning_rate": 4.690490805057543e-06,
+      "loss": 0.6152,
+      "step": 3701
+    },
+    {
+      "epoch": 0.9828753484667463,
+      "grad_norm": 0.4099902283354911,
+      "learning_rate": 4.69032251997599e-06,
+      "loss": 0.6263,
+      "step": 3702
+    },
+    {
+      "epoch": 0.98314084694013,
+      "grad_norm": 0.3856394137656872,
+      "learning_rate": 4.690154192177787e-06,
+      "loss": 0.5846,
+      "step": 3703
+    },
+    {
+      "epoch": 0.9834063454135139,
+      "grad_norm": 0.3989168924172415,
+      "learning_rate": 4.6899858216662136e-06,
+      "loss": 0.6062,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9836718438868977,
+      "grad_norm": 0.39889635541321233,
+      "learning_rate": 4.689817408444555e-06,
+      "loss": 0.6378,
+      "step": 3705
+    },
+    {
+      "epoch": 0.9839373423602814,
+      "grad_norm": 0.39406652466850894,
+      "learning_rate": 4.689648952516095e-06,
+      "loss": 0.6028,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9842028408336652,
+      "grad_norm": 0.4033355488192236,
+      "learning_rate": 4.68948045388412e-06,
+      "loss": 0.6072,
+      "step": 3707
+    },
+    {
+      "epoch": 0.984468339307049,
+      "grad_norm": 0.37862305697802934,
+      "learning_rate": 4.689311912551915e-06,
+      "loss": 0.606,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9847338377804328,
+      "grad_norm": 0.3878418648235052,
+      "learning_rate": 4.689143328522767e-06,
+      "loss": 0.5976,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9849993362538165,
+      "grad_norm": 0.3978761070416158,
+      "learning_rate": 4.688974701799964e-06,
+      "loss": 0.5996,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9852648347272003,
+      "grad_norm": 0.3820619528836613,
+      "learning_rate": 4.688806032386796e-06,
+      "loss": 0.6028,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9855303332005841,
+      "grad_norm": 0.3796453250847225,
+      "learning_rate": 4.6886373202865506e-06,
+      "loss": 0.6338,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9857958316739679,
+      "grad_norm": 0.3968556621124762,
+      "learning_rate": 4.688468565502519e-06,
+      "loss": 0.6341,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9860613301473516,
+      "grad_norm": 0.404800194864264,
+      "learning_rate": 4.688299768037991e-06,
+      "loss": 0.6604,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9863268286207354,
+      "grad_norm": 0.36930134446805696,
+      "learning_rate": 4.688130927896261e-06,
+      "loss": 0.5828,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9865923270941193,
+      "grad_norm": 0.3922966539266864,
+      "learning_rate": 4.687962045080619e-06,
+      "loss": 0.5727,
+      "step": 3716
+    },
+    {
+      "epoch": 0.986857825567503,
+      "grad_norm": 0.3955857287808288,
+      "learning_rate": 4.687793119594361e-06,
+      "loss": 0.6415,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9871233240408868,
+      "grad_norm": 0.3949074082518782,
+      "learning_rate": 4.6876241514407806e-06,
+      "loss": 0.6329,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9873888225142705,
+      "grad_norm": 0.39585244516211565,
+      "learning_rate": 4.687455140623173e-06,
+      "loss": 0.5627,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9876543209876543,
+      "grad_norm": 0.416463933185347,
+      "learning_rate": 4.687286087144833e-06,
+      "loss": 0.6316,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9879198194610381,
+      "grad_norm": 0.3867486743578083,
+      "learning_rate": 4.687116991009061e-06,
+      "loss": 0.6101,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9881853179344219,
+      "grad_norm": 0.4012355026380244,
+      "learning_rate": 4.6869478522191515e-06,
+      "loss": 0.6185,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9884508164078056,
+      "grad_norm": 0.4085551394800617,
+      "learning_rate": 4.686778670778404e-06,
+      "loss": 0.6213,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9887163148811894,
+      "grad_norm": 0.39663432013046435,
+      "learning_rate": 4.686609446690119e-06,
+      "loss": 0.6044,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9889818133545732,
+      "grad_norm": 0.4018089291913098,
+      "learning_rate": 4.686440179957597e-06,
+      "loss": 0.6178,
+      "step": 3725
+    },
+    {
+      "epoch": 0.989247311827957,
+      "grad_norm": 0.39496287783320855,
+      "learning_rate": 4.686270870584137e-06,
+      "loss": 0.607,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9895128103013408,
+      "grad_norm": 0.3938774541356936,
+      "learning_rate": 4.686101518573041e-06,
+      "loss": 0.5869,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9897783087747245,
+      "grad_norm": 0.3918321207536686,
+      "learning_rate": 4.685932123927615e-06,
+      "loss": 0.6078,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9900438072481084,
+      "grad_norm": 0.39201053561114496,
+      "learning_rate": 4.685762686651159e-06,
+      "loss": 0.6552,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9903093057214921,
+      "grad_norm": 0.3766261080464049,
+      "learning_rate": 4.685593206746979e-06,
+      "loss": 0.5849,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9905748041948759,
+      "grad_norm": 0.3826967061947365,
+      "learning_rate": 4.685423684218381e-06,
+      "loss": 0.5673,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9908403026682596,
+      "grad_norm": 0.37498966330526406,
+      "learning_rate": 4.68525411906867e-06,
+      "loss": 0.6033,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9911058011416435,
+      "grad_norm": 0.3780543886684855,
+      "learning_rate": 4.685084511301154e-06,
+      "loss": 0.6057,
+      "step": 3733
+    },
+    {
+      "epoch": 0.9913712996150272,
+      "grad_norm": 0.3892072551607135,
+      "learning_rate": 4.684914860919138e-06,
+      "loss": 0.6136,
+      "step": 3734
+    },
+    {
+      "epoch": 0.991636798088411,
+      "grad_norm": 0.37365922048188355,
+      "learning_rate": 4.684745167925934e-06,
+      "loss": 0.5994,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9919022965617947,
+      "grad_norm": 0.39626044475214023,
+      "learning_rate": 4.6845754323248505e-06,
+      "loss": 0.6305,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9921677950351786,
+      "grad_norm": 0.3937902793503907,
+      "learning_rate": 4.684405654119196e-06,
+      "loss": 0.6043,
+      "step": 3737
+    },
+    {
+      "epoch": 0.9924332935085624,
+      "grad_norm": 0.39432004616059513,
+      "learning_rate": 4.684235833312284e-06,
+      "loss": 0.5921,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9926987919819461,
+      "grad_norm": 0.3819503364307295,
+      "learning_rate": 4.684065969907425e-06,
+      "loss": 0.6173,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9929642904553299,
+      "grad_norm": 0.3754592203801274,
+      "learning_rate": 4.683896063907932e-06,
+      "loss": 0.6069,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9932297889287136,
+      "grad_norm": 0.38545667597499994,
+      "learning_rate": 4.683726115317119e-06,
+      "loss": 0.613,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9934952874020975,
+      "grad_norm": 0.39499805200038546,
+      "learning_rate": 4.6835561241382985e-06,
+      "loss": 0.6031,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9937607858754812,
+      "grad_norm": 0.3794613043776615,
+      "learning_rate": 4.683386090374789e-06,
+      "loss": 0.6436,
+      "step": 3743
+    },
+    {
+      "epoch": 0.994026284348865,
+      "grad_norm": 0.3845896428935738,
+      "learning_rate": 4.683216014029905e-06,
+      "loss": 0.604,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9942917828222487,
+      "grad_norm": 0.3886443680703139,
+      "learning_rate": 4.6830458951069625e-06,
+      "loss": 0.6094,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9945572812956326,
+      "grad_norm": 0.3815840078596797,
+      "learning_rate": 4.68287573360928e-06,
+      "loss": 0.6372,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9948227797690163,
+      "grad_norm": 0.3820926262645276,
+      "learning_rate": 4.682705529540177e-06,
+      "loss": 0.6035,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9950882782424001,
+      "grad_norm": 0.3762459209247919,
+      "learning_rate": 4.6825352829029705e-06,
+      "loss": 0.6073,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9953537767157838,
+      "grad_norm": 0.3843193911083839,
+      "learning_rate": 4.682364993700983e-06,
+      "loss": 0.5902,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9956192751891677,
+      "grad_norm": 0.38812051434919137,
+      "learning_rate": 4.682194661937535e-06,
+      "loss": 0.6504,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9958847736625515,
+      "grad_norm": 0.4053891827606106,
+      "learning_rate": 4.682024287615948e-06,
+      "loss": 0.6063,
+      "step": 3751
+    },
+    {
+      "epoch": 0.9961502721359352,
+      "grad_norm": 0.3757484200398291,
+      "learning_rate": 4.681853870739545e-06,
+      "loss": 0.6037,
+      "step": 3752
+    },
+    {
+      "epoch": 0.996415770609319,
+      "grad_norm": 0.3974556813432687,
+      "learning_rate": 4.681683411311649e-06,
+      "loss": 0.6236,
+      "step": 3753
+    },
+    {
+      "epoch": 0.9966812690827028,
+      "grad_norm": 0.38323425943369327,
+      "learning_rate": 4.681512909335585e-06,
+      "loss": 0.6031,
+      "step": 3754
+    },
+    {
+      "epoch": 0.9969467675560866,
+      "grad_norm": 0.37955441716052646,
+      "learning_rate": 4.681342364814678e-06,
+      "loss": 0.598,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9972122660294703,
+      "grad_norm": 0.38406770095360143,
+      "learning_rate": 4.681171777752255e-06,
+      "loss": 0.5885,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9974777645028541,
+      "grad_norm": 0.38121222912205954,
+      "learning_rate": 4.68100114815164e-06,
+      "loss": 0.6079,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9977432629762378,
+      "grad_norm": 0.3986173792183184,
+      "learning_rate": 4.680830476016164e-06,
+      "loss": 0.6106,
+      "step": 3758
+    },
+    {
+      "epoch": 0.9980087614496217,
+      "grad_norm": 0.4013301795791731,
+      "learning_rate": 4.680659761349154e-06,
+      "loss": 0.5945,
+      "step": 3759
+    },
+    {
+      "epoch": 0.9982742599230054,
+      "grad_norm": 0.3969451492757828,
+      "learning_rate": 4.680489004153939e-06,
+      "loss": 0.6177,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9985397583963892,
+      "grad_norm": 0.37550632993955513,
+      "learning_rate": 4.6803182044338506e-06,
+      "loss": 0.612,
+      "step": 3761
+    },
+    {
+      "epoch": 0.998805256869773,
+      "grad_norm": 0.3903113437702318,
+      "learning_rate": 4.680147362192218e-06,
+      "loss": 0.6479,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9990707553431568,
+      "grad_norm": 0.3737328244921276,
+      "learning_rate": 4.679976477432373e-06,
+      "loss": 0.5976,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9993362538165406,
+      "grad_norm": 0.3707473898268587,
+      "learning_rate": 4.679805550157651e-06,
+      "loss": 0.5925,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9996017522899243,
+      "grad_norm": 0.4049322640665022,
+      "learning_rate": 4.679634580371383e-06,
+      "loss": 0.5632,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9998672507633081,
+      "grad_norm": 0.3858564461362987,
+      "learning_rate": 4.679463568076904e-06,
+      "loss": 0.5944,
+      "step": 3766
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3858564461362987,
+      "learning_rate": 4.6792925132775495e-06,
+      "loss": 0.6066,
+      "step": 3767
+    },
+    {
+      "epoch": 1.0002654984733839,
+      "grad_norm": 0.6259057377833318,
+      "learning_rate": 4.679121415976654e-06,
+      "loss": 0.5663,
+      "step": 3768
+    },
+    {
+      "epoch": 1.0005309969467675,
+      "grad_norm": 0.3963461731153748,
+      "learning_rate": 4.678950276177556e-06,
+      "loss": 0.6094,
+      "step": 3769
+    },
+    {
+      "epoch": 1.0007964954201514,
+      "grad_norm": 0.41829513680315344,
+      "learning_rate": 4.678779093883593e-06,
+      "loss": 0.5822,
+      "step": 3770
+    },
+    {
+      "epoch": 1.001061993893535,
+      "grad_norm": 0.3897118635509286,
+      "learning_rate": 4.6786078690981026e-06,
+      "loss": 0.6003,
+      "step": 3771
+    },
+    {
+      "epoch": 1.0013274923669189,
+      "grad_norm": 0.40909010762029085,
+      "learning_rate": 4.678436601824426e-06,
+      "loss": 0.5824,
+      "step": 3772
+    },
+    {
+      "epoch": 1.0015929908403027,
+      "grad_norm": 0.4079357338221315,
+      "learning_rate": 4.678265292065901e-06,
+      "loss": 0.6177,
+      "step": 3773
+    },
+    {
+      "epoch": 1.0018584893136864,
+      "grad_norm": 0.3975615536380815,
+      "learning_rate": 4.678093939825869e-06,
+      "loss": 0.5731,
+      "step": 3774
+    },
+    {
+      "epoch": 1.0021239877870702,
+      "grad_norm": 0.39046478516719413,
+      "learning_rate": 4.677922545107673e-06,
+      "loss": 0.6115,
+      "step": 3775
+    },
+    {
+      "epoch": 1.002389486260454,
+      "grad_norm": 0.40044517494805526,
+      "learning_rate": 4.677751107914654e-06,
+      "loss": 0.6184,
+      "step": 3776
+    },
+    {
+      "epoch": 1.0026549847338377,
+      "grad_norm": 0.4019656685191064,
+      "learning_rate": 4.677579628250157e-06,
+      "loss": 0.6468,
+      "step": 3777
+    },
+    {
+      "epoch": 1.0029204832072216,
+      "grad_norm": 0.3869316355567388,
+      "learning_rate": 4.677408106117526e-06,
+      "loss": 0.6013,
+      "step": 3778
+    },
+    {
+      "epoch": 1.0031859816806052,
+      "grad_norm": 0.3951489761742382,
+      "learning_rate": 4.677236541520105e-06,
+      "loss": 0.6157,
+      "step": 3779
+    },
+    {
+      "epoch": 1.003451480153989,
+      "grad_norm": 0.39593968440594485,
+      "learning_rate": 4.677064934461241e-06,
+      "loss": 0.5898,
+      "step": 3780
+    },
+    {
+      "epoch": 1.003716978627373,
+      "grad_norm": 0.4085811771712458,
+      "learning_rate": 4.676893284944281e-06,
+      "loss": 0.6061,
+      "step": 3781
+    },
+    {
+      "epoch": 1.0039824771007566,
+      "grad_norm": 0.408787436234934,
+      "learning_rate": 4.676721592972571e-06,
+      "loss": 0.5558,
+      "step": 3782
+    },
+    {
+      "epoch": 1.0042479755741405,
+      "grad_norm": 0.38579083766709615,
+      "learning_rate": 4.67654985854946e-06,
+      "loss": 0.5695,
+      "step": 3783
+    },
+    {
+      "epoch": 1.0045134740475243,
+      "grad_norm": 0.3855617336310054,
+      "learning_rate": 4.676378081678299e-06,
+      "loss": 0.571,
+      "step": 3784
+    },
+    {
+      "epoch": 1.004778972520908,
+      "grad_norm": 0.37781350078508324,
+      "learning_rate": 4.676206262362436e-06,
+      "loss": 0.6223,
+      "step": 3785
+    },
+    {
+      "epoch": 1.0050444709942918,
+      "grad_norm": 0.40473186389736987,
+      "learning_rate": 4.676034400605223e-06,
+      "loss": 0.6438,
+      "step": 3786
+    },
+    {
+      "epoch": 1.0053099694676755,
+      "grad_norm": 0.3964697621523235,
+      "learning_rate": 4.6758624964100104e-06,
+      "loss": 0.5797,
+      "step": 3787
+    },
+    {
+      "epoch": 1.0055754679410593,
+      "grad_norm": 0.40271319237949815,
+      "learning_rate": 4.675690549780152e-06,
+      "loss": 0.5945,
+      "step": 3788
+    },
+    {
+      "epoch": 1.0058409664144432,
+      "grad_norm": 0.3879779861067464,
+      "learning_rate": 4.675518560719002e-06,
+      "loss": 0.604,
+      "step": 3789
+    },
+    {
+      "epoch": 1.0061064648878268,
+      "grad_norm": 0.39868356294679413,
+      "learning_rate": 4.675346529229912e-06,
+      "loss": 0.603,
+      "step": 3790
+    },
+    {
+      "epoch": 1.0063719633612107,
+      "grad_norm": 0.39674484222218653,
+      "learning_rate": 4.67517445531624e-06,
+      "loss": 0.6126,
+      "step": 3791
+    },
+    {
+      "epoch": 1.0066374618345943,
+      "grad_norm": 0.4022661282035222,
+      "learning_rate": 4.675002338981339e-06,
+      "loss": 0.5996,
+      "step": 3792
+    },
+    {
+      "epoch": 1.0069029603079782,
+      "grad_norm": 0.3975232938458579,
+      "learning_rate": 4.674830180228568e-06,
+      "loss": 0.6382,
+      "step": 3793
+    },
+    {
+      "epoch": 1.007168458781362,
+      "grad_norm": 0.3928531816249291,
+      "learning_rate": 4.6746579790612824e-06,
+      "loss": 0.6114,
+      "step": 3794
+    },
+    {
+      "epoch": 1.0074339572547457,
+      "grad_norm": 0.39539341044791426,
+      "learning_rate": 4.674485735482843e-06,
+      "loss": 0.5988,
+      "step": 3795
+    },
+    {
+      "epoch": 1.0076994557281296,
+      "grad_norm": 0.3965256695581769,
+      "learning_rate": 4.674313449496607e-06,
+      "loss": 0.6046,
+      "step": 3796
+    },
+    {
+      "epoch": 1.0079649542015134,
+      "grad_norm": 0.37929220201522545,
+      "learning_rate": 4.674141121105935e-06,
+      "loss": 0.631,
+      "step": 3797
+    },
+    {
+      "epoch": 1.008230452674897,
+      "grad_norm": 0.3862540210375224,
+      "learning_rate": 4.673968750314189e-06,
+      "loss": 0.5596,
+      "step": 3798
+    },
+    {
+      "epoch": 1.008495951148281,
+      "grad_norm": 0.3873002440554336,
+      "learning_rate": 4.673796337124729e-06,
+      "loss": 0.6062,
+      "step": 3799
+    },
+    {
+      "epoch": 1.0087614496216646,
+      "grad_norm": 0.40964134645451533,
+      "learning_rate": 4.673623881540917e-06,
+      "loss": 0.6488,
+      "step": 3800
+    },
+    {
+      "epoch": 1.0090269480950484,
+      "grad_norm": 0.4012541396609028,
+      "learning_rate": 4.6734513835661184e-06,
+      "loss": 0.6168,
+      "step": 3801
+    },
+    {
+      "epoch": 1.0092924465684323,
+      "grad_norm": 0.3885834564617068,
+      "learning_rate": 4.673278843203697e-06,
+      "loss": 0.593,
+      "step": 3802
+    },
+    {
+      "epoch": 1.009557945041816,
+      "grad_norm": 0.38139183602894294,
+      "learning_rate": 4.673106260457015e-06,
+      "loss": 0.604,
+      "step": 3803
+    },
+    {
+      "epoch": 1.0098234435151998,
+      "grad_norm": 0.39949221812252345,
+      "learning_rate": 4.6729336353294415e-06,
+      "loss": 0.5734,
+      "step": 3804
+    },
+    {
+      "epoch": 1.0100889419885837,
+      "grad_norm": 0.40253401981329706,
+      "learning_rate": 4.6727609678243426e-06,
+      "loss": 0.6145,
+      "step": 3805
+    },
+    {
+      "epoch": 1.0103544404619673,
+      "grad_norm": 0.38671036359841193,
+      "learning_rate": 4.672588257945085e-06,
+      "loss": 0.5973,
+      "step": 3806
+    },
+    {
+      "epoch": 1.0106199389353512,
+      "grad_norm": 0.3972410132459281,
+      "learning_rate": 4.672415505695035e-06,
+      "loss": 0.6014,
+      "step": 3807
+    },
+    {
+      "epoch": 1.0108854374087348,
+      "grad_norm": 0.3938474800282929,
+      "learning_rate": 4.672242711077565e-06,
+      "loss": 0.5835,
+      "step": 3808
+    },
+    {
+      "epoch": 1.0111509358821187,
+      "grad_norm": 0.4080790671272506,
+      "learning_rate": 4.672069874096044e-06,
+      "loss": 0.5673,
+      "step": 3809
+    },
+    {
+      "epoch": 1.0114164343555025,
+      "grad_norm": 0.38773639414212874,
+      "learning_rate": 4.671896994753841e-06,
+      "loss": 0.6199,
+      "step": 3810
+    },
+    {
+      "epoch": 1.0116819328288862,
+      "grad_norm": 0.3855485014693545,
+      "learning_rate": 4.67172407305433e-06,
+      "loss": 0.5991,
+      "step": 3811
+    },
+    {
+      "epoch": 1.01194743130227,
+      "grad_norm": 0.38890547216898175,
+      "learning_rate": 4.671551109000882e-06,
+      "loss": 0.6324,
+      "step": 3812
+    },
+    {
+      "epoch": 1.0122129297756537,
+      "grad_norm": 0.4100699456509582,
+      "learning_rate": 4.67137810259687e-06,
+      "loss": 0.6201,
+      "step": 3813
+    },
+    {
+      "epoch": 1.0124784282490376,
+      "grad_norm": 0.3871235114915657,
+      "learning_rate": 4.671205053845669e-06,
+      "loss": 0.6018,
+      "step": 3814
+    },
+    {
+      "epoch": 1.0127439267224214,
+      "grad_norm": 0.3848621701251693,
+      "learning_rate": 4.671031962750653e-06,
+      "loss": 0.577,
+      "step": 3815
+    },
+    {
+      "epoch": 1.013009425195805,
+      "grad_norm": 0.3979228799548935,
+      "learning_rate": 4.670858829315198e-06,
+      "loss": 0.5473,
+      "step": 3816
+    },
+    {
+      "epoch": 1.013274923669189,
+      "grad_norm": 0.38593714215860553,
+      "learning_rate": 4.670685653542682e-06,
+      "loss": 0.6116,
+      "step": 3817
+    },
+    {
+      "epoch": 1.0135404221425728,
+      "grad_norm": 0.40542261226495707,
+      "learning_rate": 4.670512435436479e-06,
+      "loss": 0.5714,
+      "step": 3818
+    },
+    {
+      "epoch": 1.0138059206159564,
+      "grad_norm": 0.4013574318406041,
+      "learning_rate": 4.670339174999971e-06,
+      "loss": 0.6037,
+      "step": 3819
+    },
+    {
+      "epoch": 1.0140714190893403,
+      "grad_norm": 0.3883892623864832,
+      "learning_rate": 4.670165872236534e-06,
+      "loss": 0.5923,
+      "step": 3820
+    },
+    {
+      "epoch": 1.014336917562724,
+      "grad_norm": 0.3886093085221166,
+      "learning_rate": 4.6699925271495495e-06,
+      "loss": 0.5849,
+      "step": 3821
+    },
+    {
+      "epoch": 1.0146024160361078,
+      "grad_norm": 0.4035935907573044,
+      "learning_rate": 4.669819139742398e-06,
+      "loss": 0.6267,
+      "step": 3822
+    },
+    {
+      "epoch": 1.0148679145094917,
+      "grad_norm": 0.3972272760666861,
+      "learning_rate": 4.66964571001846e-06,
+      "loss": 0.5686,
+      "step": 3823
+    },
+    {
+      "epoch": 1.0151334129828753,
+      "grad_norm": 0.38617956951169913,
+      "learning_rate": 4.669472237981118e-06,
+      "loss": 0.6274,
+      "step": 3824
+    },
+    {
+      "epoch": 1.0153989114562592,
+      "grad_norm": 0.3893845128129732,
+      "learning_rate": 4.669298723633757e-06,
+      "loss": 0.594,
+      "step": 3825
+    },
+    {
+      "epoch": 1.0156644099296428,
+      "grad_norm": 0.3768802731056577,
+      "learning_rate": 4.669125166979758e-06,
+      "loss": 0.5987,
+      "step": 3826
+    },
+    {
+      "epoch": 1.0159299084030267,
+      "grad_norm": 0.4173704274274717,
+      "learning_rate": 4.668951568022508e-06,
+      "loss": 0.5663,
+      "step": 3827
+    },
+    {
+      "epoch": 1.0161954068764105,
+      "grad_norm": 0.4189146773867367,
+      "learning_rate": 4.668777926765392e-06,
+      "loss": 0.6065,
+      "step": 3828
+    },
+    {
+      "epoch": 1.0164609053497942,
+      "grad_norm": 0.39468865392240277,
+      "learning_rate": 4.668604243211797e-06,
+      "loss": 0.6022,
+      "step": 3829
+    },
+    {
+      "epoch": 1.016726403823178,
+      "grad_norm": 0.4012418917127769,
+      "learning_rate": 4.668430517365109e-06,
+      "loss": 0.5912,
+      "step": 3830
+    },
+    {
+      "epoch": 1.016991902296562,
+      "grad_norm": 0.3997728918042466,
+      "learning_rate": 4.668256749228717e-06,
+      "loss": 0.6109,
+      "step": 3831
+    },
+    {
+      "epoch": 1.0172574007699455,
+      "grad_norm": 0.3857273297766302,
+      "learning_rate": 4.668082938806009e-06,
+      "loss": 0.5956,
+      "step": 3832
+    },
+    {
+      "epoch": 1.0175228992433294,
+      "grad_norm": 0.40515089971101936,
+      "learning_rate": 4.667909086100375e-06,
+      "loss": 0.6109,
+      "step": 3833
+    },
+    {
+      "epoch": 1.017788397716713,
+      "grad_norm": 0.40120298631566786,
+      "learning_rate": 4.667735191115207e-06,
+      "loss": 0.6109,
+      "step": 3834
+    },
+    {
+      "epoch": 1.018053896190097,
+      "grad_norm": 0.4098672779003959,
+      "learning_rate": 4.667561253853894e-06,
+      "loss": 0.594,
+      "step": 3835
+    },
+    {
+      "epoch": 1.0183193946634808,
+      "grad_norm": 0.39335799588059045,
+      "learning_rate": 4.66738727431983e-06,
+      "loss": 0.5745,
+      "step": 3836
+    },
+    {
+      "epoch": 1.0185848931368644,
+      "grad_norm": 0.40248586436280626,
+      "learning_rate": 4.667213252516408e-06,
+      "loss": 0.6454,
+      "step": 3837
+    },
+    {
+      "epoch": 1.0188503916102483,
+      "grad_norm": 0.4016739251164001,
+      "learning_rate": 4.667039188447021e-06,
+      "loss": 0.5908,
+      "step": 3838
+    },
+    {
+      "epoch": 1.0191158900836321,
+      "grad_norm": 0.3831738281956251,
+      "learning_rate": 4.666865082115064e-06,
+      "loss": 0.6099,
+      "step": 3839
+    },
+    {
+      "epoch": 1.0193813885570158,
+      "grad_norm": 0.40143125602335383,
+      "learning_rate": 4.666690933523932e-06,
+      "loss": 0.556,
+      "step": 3840
+    },
+    {
+      "epoch": 1.0196468870303996,
+      "grad_norm": 0.40232850105449164,
+      "learning_rate": 4.666516742677022e-06,
+      "loss": 0.6159,
+      "step": 3841
+    },
+    {
+      "epoch": 1.0199123855037833,
+      "grad_norm": 0.39724733564801323,
+      "learning_rate": 4.666342509577731e-06,
+      "loss": 0.5946,
+      "step": 3842
+    },
+    {
+      "epoch": 1.0201778839771671,
+      "grad_norm": 0.3942266242092512,
+      "learning_rate": 4.666168234229457e-06,
+      "loss": 0.6218,
+      "step": 3843
+    },
+    {
+      "epoch": 1.020443382450551,
+      "grad_norm": 0.39671314798907953,
+      "learning_rate": 4.665993916635599e-06,
+      "loss": 0.6053,
+      "step": 3844
+    },
+    {
+      "epoch": 1.0207088809239346,
+      "grad_norm": 0.394670254581841,
+      "learning_rate": 4.665819556799556e-06,
+      "loss": 0.6352,
+      "step": 3845
+    },
+    {
+      "epoch": 1.0209743793973185,
+      "grad_norm": 0.400681560274759,
+      "learning_rate": 4.66564515472473e-06,
+      "loss": 0.5699,
+      "step": 3846
+    },
+    {
+      "epoch": 1.0212398778707021,
+      "grad_norm": 0.3883498993718981,
+      "learning_rate": 4.66547071041452e-06,
+      "loss": 0.6036,
+      "step": 3847
+    },
+    {
+      "epoch": 1.021505376344086,
+      "grad_norm": 0.4089043937265339,
+      "learning_rate": 4.665296223872328e-06,
+      "loss": 0.5819,
+      "step": 3848
+    },
+    {
+      "epoch": 1.0217708748174699,
+      "grad_norm": 0.4073348056144677,
+      "learning_rate": 4.66512169510156e-06,
+      "loss": 0.5983,
+      "step": 3849
+    },
+    {
+      "epoch": 1.0220363732908535,
+      "grad_norm": 0.4043560168257078,
+      "learning_rate": 4.664947124105617e-06,
+      "loss": 0.6051,
+      "step": 3850
+    },
+    {
+      "epoch": 1.0223018717642374,
+      "grad_norm": 0.4039619958705381,
+      "learning_rate": 4.664772510887905e-06,
+      "loss": 0.6379,
+      "step": 3851
+    },
+    {
+      "epoch": 1.0225673702376212,
+      "grad_norm": 0.3965065767538858,
+      "learning_rate": 4.664597855451828e-06,
+      "loss": 0.578,
+      "step": 3852
+    },
+    {
+      "epoch": 1.0228328687110049,
+      "grad_norm": 0.38789913747237514,
+      "learning_rate": 4.664423157800793e-06,
+      "loss": 0.6055,
+      "step": 3853
+    },
+    {
+      "epoch": 1.0230983671843887,
+      "grad_norm": 0.41978173442823347,
+      "learning_rate": 4.6642484179382075e-06,
+      "loss": 0.5887,
+      "step": 3854
+    },
+    {
+      "epoch": 1.0233638656577724,
+      "grad_norm": 0.3968057102416469,
+      "learning_rate": 4.664073635867479e-06,
+      "loss": 0.6011,
+      "step": 3855
+    },
+    {
+      "epoch": 1.0236293641311562,
+      "grad_norm": 0.38760434076691996,
+      "learning_rate": 4.663898811592015e-06,
+      "loss": 0.6018,
+      "step": 3856
+    },
+    {
+      "epoch": 1.02389486260454,
+      "grad_norm": 0.40156851207975036,
+      "learning_rate": 4.663723945115227e-06,
+      "loss": 0.5885,
+      "step": 3857
+    },
+    {
+      "epoch": 1.0241603610779237,
+      "grad_norm": 0.39866949319640016,
+      "learning_rate": 4.663549036440523e-06,
+      "loss": 0.6113,
+      "step": 3858
+    },
+    {
+      "epoch": 1.0244258595513076,
+      "grad_norm": 0.38432397411316843,
+      "learning_rate": 4.663374085571317e-06,
+      "loss": 0.5956,
+      "step": 3859
+    },
+    {
+      "epoch": 1.0246913580246915,
+      "grad_norm": 0.39993289466861925,
+      "learning_rate": 4.663199092511019e-06,
+      "loss": 0.5908,
+      "step": 3860
+    },
+    {
+      "epoch": 1.024956856498075,
+      "grad_norm": 0.40924468509791817,
+      "learning_rate": 4.663024057263042e-06,
+      "loss": 0.6431,
+      "step": 3861
+    },
+    {
+      "epoch": 1.025222354971459,
+      "grad_norm": 0.40891860498148186,
+      "learning_rate": 4.662848979830801e-06,
+      "loss": 0.612,
+      "step": 3862
+    },
+    {
+      "epoch": 1.0254878534448426,
+      "grad_norm": 0.3970686004743134,
+      "learning_rate": 4.662673860217708e-06,
+      "loss": 0.5941,
+      "step": 3863
+    },
+    {
+      "epoch": 1.0257533519182265,
+      "grad_norm": 0.4283242202155573,
+      "learning_rate": 4.66249869842718e-06,
+      "loss": 0.6059,
+      "step": 3864
+    },
+    {
+      "epoch": 1.0260188503916103,
+      "grad_norm": 0.39952309425662036,
+      "learning_rate": 4.662323494462633e-06,
+      "loss": 0.5891,
+      "step": 3865
+    },
+    {
+      "epoch": 1.026284348864994,
+      "grad_norm": 0.39890012480983744,
+      "learning_rate": 4.662148248327484e-06,
+      "loss": 0.5883,
+      "step": 3866
+    },
+    {
+      "epoch": 1.0265498473383778,
+      "grad_norm": 0.39388272422894277,
+      "learning_rate": 4.661972960025149e-06,
+      "loss": 0.6069,
+      "step": 3867
+    },
+    {
+      "epoch": 1.0268153458117615,
+      "grad_norm": 0.4108952603539573,
+      "learning_rate": 4.661797629559048e-06,
+      "loss": 0.6498,
+      "step": 3868
+    },
+    {
+      "epoch": 1.0270808442851453,
+      "grad_norm": 0.4183665697277097,
+      "learning_rate": 4.661622256932601e-06,
+      "loss": 0.5946,
+      "step": 3869
+    },
+    {
+      "epoch": 1.0273463427585292,
+      "grad_norm": 0.40502224138892995,
+      "learning_rate": 4.661446842149228e-06,
+      "loss": 0.575,
+      "step": 3870
+    },
+    {
+      "epoch": 1.0276118412319128,
+      "grad_norm": 0.44647338276446036,
+      "learning_rate": 4.661271385212348e-06,
+      "loss": 0.6272,
+      "step": 3871
+    },
+    {
+      "epoch": 1.0278773397052967,
+      "grad_norm": 0.3888651813594647,
+      "learning_rate": 4.6610958861253844e-06,
+      "loss": 0.5596,
+      "step": 3872
+    },
+    {
+      "epoch": 1.0281428381786806,
+      "grad_norm": 0.383801902184117,
+      "learning_rate": 4.66092034489176e-06,
+      "loss": 0.5966,
+      "step": 3873
+    },
+    {
+      "epoch": 1.0284083366520642,
+      "grad_norm": 0.3906959120853525,
+      "learning_rate": 4.660744761514899e-06,
+      "loss": 0.5992,
+      "step": 3874
+    },
+    {
+      "epoch": 1.028673835125448,
+      "grad_norm": 0.40496452264374355,
+      "learning_rate": 4.660569135998224e-06,
+      "loss": 0.6177,
+      "step": 3875
+    },
+    {
+      "epoch": 1.0289393335988317,
+      "grad_norm": 0.3984213404972599,
+      "learning_rate": 4.66039346834516e-06,
+      "loss": 0.59,
+      "step": 3876
+    },
+    {
+      "epoch": 1.0292048320722156,
+      "grad_norm": 0.39626090226395955,
+      "learning_rate": 4.660217758559135e-06,
+      "loss": 0.5982,
+      "step": 3877
+    },
+    {
+      "epoch": 1.0294703305455994,
+      "grad_norm": 0.39995172426394393,
+      "learning_rate": 4.660042006643574e-06,
+      "loss": 0.6101,
+      "step": 3878
+    },
+    {
+      "epoch": 1.029735829018983,
+      "grad_norm": 0.3985513612786605,
+      "learning_rate": 4.659866212601906e-06,
+      "loss": 0.5726,
+      "step": 3879
+    },
+    {
+      "epoch": 1.030001327492367,
+      "grad_norm": 0.40356926884575833,
+      "learning_rate": 4.659690376437558e-06,
+      "loss": 0.5837,
+      "step": 3880
+    },
+    {
+      "epoch": 1.0302668259657506,
+      "grad_norm": 0.40091577644762116,
+      "learning_rate": 4.65951449815396e-06,
+      "loss": 0.6156,
+      "step": 3881
+    },
+    {
+      "epoch": 1.0305323244391345,
+      "grad_norm": 0.4139339054772377,
+      "learning_rate": 4.659338577754541e-06,
+      "loss": 0.587,
+      "step": 3882
+    },
+    {
+      "epoch": 1.0307978229125183,
+      "grad_norm": 0.4034515787037412,
+      "learning_rate": 4.659162615242734e-06,
+      "loss": 0.6472,
+      "step": 3883
+    },
+    {
+      "epoch": 1.031063321385902,
+      "grad_norm": 0.4065605403973272,
+      "learning_rate": 4.658986610621969e-06,
+      "loss": 0.5891,
+      "step": 3884
+    },
+    {
+      "epoch": 1.0313288198592858,
+      "grad_norm": 0.4000505817178667,
+      "learning_rate": 4.6588105638956785e-06,
+      "loss": 0.5814,
+      "step": 3885
+    },
+    {
+      "epoch": 1.0315943183326697,
+      "grad_norm": 0.3951311139365383,
+      "learning_rate": 4.658634475067297e-06,
+      "loss": 0.5969,
+      "step": 3886
+    },
+    {
+      "epoch": 1.0318598168060533,
+      "grad_norm": 0.3804157236603994,
+      "learning_rate": 4.658458344140258e-06,
+      "loss": 0.5734,
+      "step": 3887
+    },
+    {
+      "epoch": 1.0321253152794372,
+      "grad_norm": 0.39420004854683033,
+      "learning_rate": 4.658282171117996e-06,
+      "loss": 0.6018,
+      "step": 3888
+    },
+    {
+      "epoch": 1.0323908137528208,
+      "grad_norm": 0.4162983486621894,
+      "learning_rate": 4.6581059560039475e-06,
+      "loss": 0.5929,
+      "step": 3889
+    },
+    {
+      "epoch": 1.0326563122262047,
+      "grad_norm": 0.3959197888088827,
+      "learning_rate": 4.657929698801549e-06,
+      "loss": 0.6028,
+      "step": 3890
+    },
+    {
+      "epoch": 1.0329218106995885,
+      "grad_norm": 0.39309517282903017,
+      "learning_rate": 4.657753399514238e-06,
+      "loss": 0.6028,
+      "step": 3891
+    },
+    {
+      "epoch": 1.0331873091729722,
+      "grad_norm": 0.3990977660770755,
+      "learning_rate": 4.657577058145451e-06,
+      "loss": 0.6233,
+      "step": 3892
+    },
+    {
+      "epoch": 1.033452807646356,
+      "grad_norm": 0.4091445503169717,
+      "learning_rate": 4.657400674698631e-06,
+      "loss": 0.5942,
+      "step": 3893
+    },
+    {
+      "epoch": 1.03371830611974,
+      "grad_norm": 0.4000936986328691,
+      "learning_rate": 4.657224249177215e-06,
+      "loss": 0.5487,
+      "step": 3894
+    },
+    {
+      "epoch": 1.0339838045931236,
+      "grad_norm": 0.3829671456833252,
+      "learning_rate": 4.657047781584643e-06,
+      "loss": 0.5863,
+      "step": 3895
+    },
+    {
+      "epoch": 1.0342493030665074,
+      "grad_norm": 0.387430340212757,
+      "learning_rate": 4.656871271924359e-06,
+      "loss": 0.5785,
+      "step": 3896
+    },
+    {
+      "epoch": 1.034514801539891,
+      "grad_norm": 0.4017994748236961,
+      "learning_rate": 4.656694720199805e-06,
+      "loss": 0.6127,
+      "step": 3897
+    },
+    {
+      "epoch": 1.034780300013275,
+      "grad_norm": 0.4037994730988607,
+      "learning_rate": 4.656518126414422e-06,
+      "loss": 0.5753,
+      "step": 3898
+    },
+    {
+      "epoch": 1.0350457984866588,
+      "grad_norm": 0.39784196668284155,
+      "learning_rate": 4.656341490571657e-06,
+      "loss": 0.5983,
+      "step": 3899
+    },
+    {
+      "epoch": 1.0353112969600424,
+      "grad_norm": 0.4018079959932046,
+      "learning_rate": 4.656164812674952e-06,
+      "loss": 0.566,
+      "step": 3900
+    },
+    {
+      "epoch": 1.0355767954334263,
+      "grad_norm": 0.40493489957738565,
+      "learning_rate": 4.655988092727754e-06,
+      "loss": 0.594,
+      "step": 3901
+    },
+    {
+      "epoch": 1.03584229390681,
+      "grad_norm": 0.40057723838207704,
+      "learning_rate": 4.65581133073351e-06,
+      "loss": 0.5898,
+      "step": 3902
+    },
+    {
+      "epoch": 1.0361077923801938,
+      "grad_norm": 0.39390566921995784,
+      "learning_rate": 4.655634526695666e-06,
+      "loss": 0.6105,
+      "step": 3903
+    },
+    {
+      "epoch": 1.0363732908535777,
+      "grad_norm": 0.39611816245475645,
+      "learning_rate": 4.655457680617672e-06,
+      "loss": 0.6102,
+      "step": 3904
+    },
+    {
+      "epoch": 1.0366387893269613,
+      "grad_norm": 0.4199728709384419,
+      "learning_rate": 4.655280792502975e-06,
+      "loss": 0.6051,
+      "step": 3905
+    },
+    {
+      "epoch": 1.0369042878003452,
+      "grad_norm": 0.3889982763926899,
+      "learning_rate": 4.655103862355026e-06,
+      "loss": 0.5806,
+      "step": 3906
+    },
+    {
+      "epoch": 1.037169786273729,
+      "grad_norm": 0.373123833365217,
+      "learning_rate": 4.654926890177274e-06,
+      "loss": 0.615,
+      "step": 3907
+    },
+    {
+      "epoch": 1.0374352847471127,
+      "grad_norm": 0.39676172673818993,
+      "learning_rate": 4.6547498759731725e-06,
+      "loss": 0.6125,
+      "step": 3908
+    },
+    {
+      "epoch": 1.0377007832204965,
+      "grad_norm": 0.39299389999840884,
+      "learning_rate": 4.654572819746172e-06,
+      "loss": 0.6378,
+      "step": 3909
+    },
+    {
+      "epoch": 1.0379662816938802,
+      "grad_norm": 0.3967926998517405,
+      "learning_rate": 4.6543957214997265e-06,
+      "loss": 0.578,
+      "step": 3910
+    },
+    {
+      "epoch": 1.038231780167264,
+      "grad_norm": 0.3940682206990731,
+      "learning_rate": 4.65421858123729e-06,
+      "loss": 0.6314,
+      "step": 3911
+    },
+    {
+      "epoch": 1.038497278640648,
+      "grad_norm": 0.3929337954727868,
+      "learning_rate": 4.654041398962315e-06,
+      "loss": 0.5954,
+      "step": 3912
+    },
+    {
+      "epoch": 1.0387627771140315,
+      "grad_norm": 0.3914247113949542,
+      "learning_rate": 4.653864174678261e-06,
+      "loss": 0.6066,
+      "step": 3913
+    },
+    {
+      "epoch": 1.0390282755874154,
+      "grad_norm": 0.39633798836215656,
+      "learning_rate": 4.653686908388581e-06,
+      "loss": 0.5987,
+      "step": 3914
+    },
+    {
+      "epoch": 1.0392937740607993,
+      "grad_norm": 0.3945795448179149,
+      "learning_rate": 4.6535096000967325e-06,
+      "loss": 0.5705,
+      "step": 3915
+    },
+    {
+      "epoch": 1.039559272534183,
+      "grad_norm": 0.39286220128118704,
+      "learning_rate": 4.653332249806175e-06,
+      "loss": 0.6431,
+      "step": 3916
+    },
+    {
+      "epoch": 1.0398247710075668,
+      "grad_norm": 0.3855911053260191,
+      "learning_rate": 4.653154857520367e-06,
+      "loss": 0.5996,
+      "step": 3917
+    },
+    {
+      "epoch": 1.0400902694809504,
+      "grad_norm": 0.40293359210332946,
+      "learning_rate": 4.652977423242766e-06,
+      "loss": 0.6033,
+      "step": 3918
+    },
+    {
+      "epoch": 1.0403557679543343,
+      "grad_norm": 0.39345500365371766,
+      "learning_rate": 4.652799946976835e-06,
+      "loss": 0.5939,
+      "step": 3919
+    },
+    {
+      "epoch": 1.0406212664277181,
+      "grad_norm": 0.40907690282966475,
+      "learning_rate": 4.652622428726033e-06,
+      "loss": 0.6065,
+      "step": 3920
+    },
+    {
+      "epoch": 1.0408867649011018,
+      "grad_norm": 0.39850429172018254,
+      "learning_rate": 4.652444868493824e-06,
+      "loss": 0.6126,
+      "step": 3921
+    },
+    {
+      "epoch": 1.0411522633744856,
+      "grad_norm": 0.39506095836641636,
+      "learning_rate": 4.65226726628367e-06,
+      "loss": 0.5997,
+      "step": 3922
+    },
+    {
+      "epoch": 1.0414177618478693,
+      "grad_norm": 0.3868743181514466,
+      "learning_rate": 4.652089622099034e-06,
+      "loss": 0.5669,
+      "step": 3923
+    },
+    {
+      "epoch": 1.0416832603212531,
+      "grad_norm": 0.3988298060522133,
+      "learning_rate": 4.651911935943381e-06,
+      "loss": 0.6048,
+      "step": 3924
+    },
+    {
+      "epoch": 1.041948758794637,
+      "grad_norm": 0.39294241088626475,
+      "learning_rate": 4.651734207820177e-06,
+      "loss": 0.5982,
+      "step": 3925
+    },
+    {
+      "epoch": 1.0422142572680206,
+      "grad_norm": 0.39674976460430234,
+      "learning_rate": 4.6515564377328875e-06,
+      "loss": 0.5689,
+      "step": 3926
+    },
+    {
+      "epoch": 1.0424797557414045,
+      "grad_norm": 0.39706026828222624,
+      "learning_rate": 4.651378625684979e-06,
+      "loss": 0.6184,
+      "step": 3927
+    },
+    {
+      "epoch": 1.0427452542147884,
+      "grad_norm": 0.3975827023725938,
+      "learning_rate": 4.651200771679921e-06,
+      "loss": 0.5954,
+      "step": 3928
+    },
+    {
+      "epoch": 1.043010752688172,
+      "grad_norm": 0.39035044752191134,
+      "learning_rate": 4.65102287572118e-06,
+      "loss": 0.6116,
+      "step": 3929
+    },
+    {
+      "epoch": 1.0432762511615559,
+      "grad_norm": 0.4154047059386226,
+      "learning_rate": 4.650844937812227e-06,
+      "loss": 0.5963,
+      "step": 3930
+    },
+    {
+      "epoch": 1.0435417496349395,
+      "grad_norm": 0.4149390956710434,
+      "learning_rate": 4.65066695795653e-06,
+      "loss": 0.5663,
+      "step": 3931
+    },
+    {
+      "epoch": 1.0438072481083234,
+      "grad_norm": 0.3862636405477702,
+      "learning_rate": 4.650488936157564e-06,
+      "loss": 0.6071,
+      "step": 3932
+    },
+    {
+      "epoch": 1.0440727465817072,
+      "grad_norm": 0.42258029916521017,
+      "learning_rate": 4.650310872418797e-06,
+      "loss": 0.6158,
+      "step": 3933
+    },
+    {
+      "epoch": 1.0443382450550909,
+      "grad_norm": 0.39155900609428707,
+      "learning_rate": 4.650132766743703e-06,
+      "loss": 0.6031,
+      "step": 3934
+    },
+    {
+      "epoch": 1.0446037435284747,
+      "grad_norm": 0.4000197891652038,
+      "learning_rate": 4.649954619135755e-06,
+      "loss": 0.6021,
+      "step": 3935
+    },
+    {
+      "epoch": 1.0448692420018584,
+      "grad_norm": 0.39435809474804107,
+      "learning_rate": 4.6497764295984295e-06,
+      "loss": 0.5934,
+      "step": 3936
+    },
+    {
+      "epoch": 1.0451347404752422,
+      "grad_norm": 0.3932886718245807,
+      "learning_rate": 4.649598198135199e-06,
+      "loss": 0.5785,
+      "step": 3937
+    },
+    {
+      "epoch": 1.045400238948626,
+      "grad_norm": 0.3856928734277222,
+      "learning_rate": 4.649419924749541e-06,
+      "loss": 0.585,
+      "step": 3938
+    },
+    {
+      "epoch": 1.0456657374220097,
+      "grad_norm": 0.397321231385105,
+      "learning_rate": 4.649241609444931e-06,
+      "loss": 0.5848,
+      "step": 3939
+    },
+    {
+      "epoch": 1.0459312358953936,
+      "grad_norm": 0.3959339300378574,
+      "learning_rate": 4.6490632522248485e-06,
+      "loss": 0.6104,
+      "step": 3940
+    },
+    {
+      "epoch": 1.0461967343687775,
+      "grad_norm": 0.4174649763601148,
+      "learning_rate": 4.64888485309277e-06,
+      "loss": 0.6006,
+      "step": 3941
+    },
+    {
+      "epoch": 1.046462232842161,
+      "grad_norm": 0.39572924729913433,
+      "learning_rate": 4.648706412052176e-06,
+      "loss": 0.6005,
+      "step": 3942
+    },
+    {
+      "epoch": 1.046727731315545,
+      "grad_norm": 0.4050920069964945,
+      "learning_rate": 4.648527929106545e-06,
+      "loss": 0.5916,
+      "step": 3943
+    },
+    {
+      "epoch": 1.0469932297889286,
+      "grad_norm": 0.40300721216178076,
+      "learning_rate": 4.64834940425936e-06,
+      "loss": 0.6268,
+      "step": 3944
+    },
+    {
+      "epoch": 1.0472587282623125,
+      "grad_norm": 0.41351572312069973,
+      "learning_rate": 4.648170837514101e-06,
+      "loss": 0.6354,
+      "step": 3945
+    },
+    {
+      "epoch": 1.0475242267356963,
+      "grad_norm": 0.41676418583413,
+      "learning_rate": 4.647992228874252e-06,
+      "loss": 0.6013,
+      "step": 3946
+    },
+    {
+      "epoch": 1.04778972520908,
+      "grad_norm": 0.42266597449064286,
+      "learning_rate": 4.6478135783432945e-06,
+      "loss": 0.6302,
+      "step": 3947
+    },
+    {
+      "epoch": 1.0480552236824638,
+      "grad_norm": 0.4046531734540872,
+      "learning_rate": 4.647634885924713e-06,
+      "loss": 0.613,
+      "step": 3948
+    },
+    {
+      "epoch": 1.0483207221558477,
+      "grad_norm": 0.40537812088325564,
+      "learning_rate": 4.647456151621994e-06,
+      "loss": 0.6054,
+      "step": 3949
+    },
+    {
+      "epoch": 1.0485862206292313,
+      "grad_norm": 0.41246438042069344,
+      "learning_rate": 4.647277375438621e-06,
+      "loss": 0.6388,
+      "step": 3950
+    },
+    {
+      "epoch": 1.0488517191026152,
+      "grad_norm": 0.395286229298196,
+      "learning_rate": 4.647098557378082e-06,
+      "loss": 0.5996,
+      "step": 3951
+    },
+    {
+      "epoch": 1.0491172175759989,
+      "grad_norm": 0.4242085118227341,
+      "learning_rate": 4.646919697443865e-06,
+      "loss": 0.5566,
+      "step": 3952
+    },
+    {
+      "epoch": 1.0493827160493827,
+      "grad_norm": 0.3948842848693423,
+      "learning_rate": 4.646740795639457e-06,
+      "loss": 0.6018,
+      "step": 3953
+    },
+    {
+      "epoch": 1.0496482145227666,
+      "grad_norm": 0.3971262371796073,
+      "learning_rate": 4.646561851968348e-06,
+      "loss": 0.623,
+      "step": 3954
+    },
+    {
+      "epoch": 1.0499137129961502,
+      "grad_norm": 0.404608872432813,
+      "learning_rate": 4.646382866434025e-06,
+      "loss": 0.6155,
+      "step": 3955
+    },
+    {
+      "epoch": 1.050179211469534,
+      "grad_norm": 0.4391896833401388,
+      "learning_rate": 4.646203839039983e-06,
+      "loss": 0.5745,
+      "step": 3956
+    },
+    {
+      "epoch": 1.050444709942918,
+      "grad_norm": 0.4086218137284621,
+      "learning_rate": 4.6460247697897104e-06,
+      "loss": 0.5631,
+      "step": 3957
+    },
+    {
+      "epoch": 1.0507102084163016,
+      "grad_norm": 0.38294863021613434,
+      "learning_rate": 4.6458456586867005e-06,
+      "loss": 0.5427,
+      "step": 3958
+    },
+    {
+      "epoch": 1.0509757068896854,
+      "grad_norm": 0.3965531203963847,
+      "learning_rate": 4.645666505734446e-06,
+      "loss": 0.6063,
+      "step": 3959
+    },
+    {
+      "epoch": 1.051241205363069,
+      "grad_norm": 0.45196652342293364,
+      "learning_rate": 4.645487310936442e-06,
+      "loss": 0.5812,
+      "step": 3960
+    },
+    {
+      "epoch": 1.051506703836453,
+      "grad_norm": 0.4229436816368648,
+      "learning_rate": 4.645308074296182e-06,
+      "loss": 0.5793,
+      "step": 3961
+    },
+    {
+      "epoch": 1.0517722023098368,
+      "grad_norm": 0.4039056274450764,
+      "learning_rate": 4.645128795817162e-06,
+      "loss": 0.6321,
+      "step": 3962
+    },
+    {
+      "epoch": 1.0520377007832205,
+      "grad_norm": 0.4078630946423279,
+      "learning_rate": 4.644949475502878e-06,
+      "loss": 0.5885,
+      "step": 3963
+    },
+    {
+      "epoch": 1.0523031992566043,
+      "grad_norm": 0.4281255109122617,
+      "learning_rate": 4.644770113356827e-06,
+      "loss": 0.6252,
+      "step": 3964
+    },
+    {
+      "epoch": 1.052568697729988,
+      "grad_norm": 0.42492414306687437,
+      "learning_rate": 4.644590709382508e-06,
+      "loss": 0.6275,
+      "step": 3965
+    },
+    {
+      "epoch": 1.0528341962033718,
+      "grad_norm": 0.3994959922056805,
+      "learning_rate": 4.64441126358342e-06,
+      "loss": 0.6098,
+      "step": 3966
+    },
+    {
+      "epoch": 1.0530996946767557,
+      "grad_norm": 0.430295370667777,
+      "learning_rate": 4.644231775963061e-06,
+      "loss": 0.6003,
+      "step": 3967
+    },
+    {
+      "epoch": 1.0533651931501393,
+      "grad_norm": 0.4025262604522123,
+      "learning_rate": 4.6440522465249325e-06,
+      "loss": 0.5705,
+      "step": 3968
+    },
+    {
+      "epoch": 1.0536306916235232,
+      "grad_norm": 0.40095141611684554,
+      "learning_rate": 4.643872675272536e-06,
+      "loss": 0.5755,
+      "step": 3969
+    },
+    {
+      "epoch": 1.053896190096907,
+      "grad_norm": 0.3855975982673741,
+      "learning_rate": 4.6436930622093736e-06,
+      "loss": 0.5841,
+      "step": 3970
+    },
+    {
+      "epoch": 1.0541616885702907,
+      "grad_norm": 0.3763769382386772,
+      "learning_rate": 4.6435134073389476e-06,
+      "loss": 0.5582,
+      "step": 3971
+    },
+    {
+      "epoch": 1.0544271870436746,
+      "grad_norm": 0.3898341243635561,
+      "learning_rate": 4.643333710664761e-06,
+      "loss": 0.6118,
+      "step": 3972
+    },
+    {
+      "epoch": 1.0546926855170582,
+      "grad_norm": 0.4072152099711562,
+      "learning_rate": 4.64315397219032e-06,
+      "loss": 0.5526,
+      "step": 3973
+    },
+    {
+      "epoch": 1.054958183990442,
+      "grad_norm": 0.39817168003443787,
+      "learning_rate": 4.642974191919129e-06,
+      "loss": 0.6171,
+      "step": 3974
+    },
+    {
+      "epoch": 1.055223682463826,
+      "grad_norm": 0.39582321590602365,
+      "learning_rate": 4.642794369854695e-06,
+      "loss": 0.5945,
+      "step": 3975
+    },
+    {
+      "epoch": 1.0554891809372096,
+      "grad_norm": 0.39713867375513295,
+      "learning_rate": 4.642614506000524e-06,
+      "loss": 0.591,
+      "step": 3976
+    },
+    {
+      "epoch": 1.0557546794105934,
+      "grad_norm": 0.39985740999871866,
+      "learning_rate": 4.642434600360124e-06,
+      "loss": 0.6041,
+      "step": 3977
+    },
+    {
+      "epoch": 1.056020177883977,
+      "grad_norm": 0.40421810047672957,
+      "learning_rate": 4.642254652937003e-06,
+      "loss": 0.5846,
+      "step": 3978
+    },
+    {
+      "epoch": 1.056285676357361,
+      "grad_norm": 0.4135223492425173,
+      "learning_rate": 4.6420746637346716e-06,
+      "loss": 0.5858,
+      "step": 3979
+    },
+    {
+      "epoch": 1.0565511748307448,
+      "grad_norm": 0.39687827660462077,
+      "learning_rate": 4.64189463275664e-06,
+      "loss": 0.569,
+      "step": 3980
+    },
+    {
+      "epoch": 1.0568166733041284,
+      "grad_norm": 0.3977025868764313,
+      "learning_rate": 4.641714560006418e-06,
+      "loss": 0.5852,
+      "step": 3981
+    },
+    {
+      "epoch": 1.0570821717775123,
+      "grad_norm": 0.40460146590018226,
+      "learning_rate": 4.641534445487519e-06,
+      "loss": 0.6124,
+      "step": 3982
+    },
+    {
+      "epoch": 1.0573476702508962,
+      "grad_norm": 0.3801516478573456,
+      "learning_rate": 4.641354289203455e-06,
+      "loss": 0.6004,
+      "step": 3983
+    },
+    {
+      "epoch": 1.0576131687242798,
+      "grad_norm": 0.4046332239665869,
+      "learning_rate": 4.6411740911577386e-06,
+      "loss": 0.6091,
+      "step": 3984
+    },
+    {
+      "epoch": 1.0578786671976637,
+      "grad_norm": 0.4187030458623271,
+      "learning_rate": 4.640993851353885e-06,
+      "loss": 0.5953,
+      "step": 3985
+    },
+    {
+      "epoch": 1.0581441656710473,
+      "grad_norm": 0.38875551995967106,
+      "learning_rate": 4.640813569795411e-06,
+      "loss": 0.5498,
+      "step": 3986
+    },
+    {
+      "epoch": 1.0584096641444312,
+      "grad_norm": 0.3942155624605592,
+      "learning_rate": 4.640633246485828e-06,
+      "loss": 0.5853,
+      "step": 3987
+    },
+    {
+      "epoch": 1.058675162617815,
+      "grad_norm": 0.4051403453068996,
+      "learning_rate": 4.6404528814286575e-06,
+      "loss": 0.6415,
+      "step": 3988
+    },
+    {
+      "epoch": 1.0589406610911987,
+      "grad_norm": 0.4110702761747505,
+      "learning_rate": 4.640272474627414e-06,
+      "loss": 0.5783,
+      "step": 3989
+    },
+    {
+      "epoch": 1.0592061595645825,
+      "grad_norm": 0.38866081164380356,
+      "learning_rate": 4.6400920260856184e-06,
+      "loss": 0.6032,
+      "step": 3990
+    },
+    {
+      "epoch": 1.0594716580379662,
+      "grad_norm": 0.4053554002886409,
+      "learning_rate": 4.639911535806787e-06,
+      "loss": 0.5987,
+      "step": 3991
+    },
+    {
+      "epoch": 1.05973715651135,
+      "grad_norm": 0.39492411529486454,
+      "learning_rate": 4.639731003794442e-06,
+      "loss": 0.5614,
+      "step": 3992
+    },
+    {
+      "epoch": 1.060002654984734,
+      "grad_norm": 0.4114656610011355,
+      "learning_rate": 4.639550430052103e-06,
+      "loss": 0.6325,
+      "step": 3993
+    },
+    {
+      "epoch": 1.0602681534581175,
+      "grad_norm": 0.4020381155232606,
+      "learning_rate": 4.639369814583292e-06,
+      "loss": 0.6457,
+      "step": 3994
+    },
+    {
+      "epoch": 1.0605336519315014,
+      "grad_norm": 0.39583993671126844,
+      "learning_rate": 4.6391891573915324e-06,
+      "loss": 0.6063,
+      "step": 3995
+    },
+    {
+      "epoch": 1.0607991504048853,
+      "grad_norm": 0.39995343768694624,
+      "learning_rate": 4.639008458480346e-06,
+      "loss": 0.5923,
+      "step": 3996
+    },
+    {
+      "epoch": 1.061064648878269,
+      "grad_norm": 0.3962011462952487,
+      "learning_rate": 4.638827717853258e-06,
+      "loss": 0.5937,
+      "step": 3997
+    },
+    {
+      "epoch": 1.0613301473516528,
+      "grad_norm": 0.3997531766832948,
+      "learning_rate": 4.638646935513792e-06,
+      "loss": 0.5774,
+      "step": 3998
+    },
+    {
+      "epoch": 1.0615956458250364,
+      "grad_norm": 0.38683666221284196,
+      "learning_rate": 4.638466111465474e-06,
+      "loss": 0.5843,
+      "step": 3999
+    },
+    {
+      "epoch": 1.0618611442984203,
+      "grad_norm": 0.4108375489718136,
+      "learning_rate": 4.638285245711832e-06,
+      "loss": 0.5789,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0621266427718041,
+      "grad_norm": 0.3936205346293354,
+      "learning_rate": 4.638104338256392e-06,
+      "loss": 0.5664,
+      "step": 4001
+    },
+    {
+      "epoch": 1.0623921412451878,
+      "grad_norm": 0.39522341551196094,
+      "learning_rate": 4.6379233891026825e-06,
+      "loss": 0.6,
+      "step": 4002
+    },
+    {
+      "epoch": 1.0626576397185716,
+      "grad_norm": 0.3915513163077133,
+      "learning_rate": 4.637742398254232e-06,
+      "loss": 0.598,
+      "step": 4003
+    },
+    {
+      "epoch": 1.0629231381919555,
+      "grad_norm": 0.3895599518886543,
+      "learning_rate": 4.637561365714571e-06,
+      "loss": 0.5873,
+      "step": 4004
+    },
+    {
+      "epoch": 1.0631886366653391,
+      "grad_norm": 0.41083908834248734,
+      "learning_rate": 4.63738029148723e-06,
+      "loss": 0.5793,
+      "step": 4005
+    },
+    {
+      "epoch": 1.063454135138723,
+      "grad_norm": 0.3924829090295871,
+      "learning_rate": 4.63719917557574e-06,
+      "loss": 0.6103,
+      "step": 4006
+    },
+    {
+      "epoch": 1.0637196336121066,
+      "grad_norm": 0.4037497914256407,
+      "learning_rate": 4.6370180179836325e-06,
+      "loss": 0.5627,
+      "step": 4007
+    },
+    {
+      "epoch": 1.0639851320854905,
+      "grad_norm": 0.40973722859646755,
+      "learning_rate": 4.6368368187144415e-06,
+      "loss": 0.599,
+      "step": 4008
+    },
+    {
+      "epoch": 1.0642506305588744,
+      "grad_norm": 0.4088499627365484,
+      "learning_rate": 4.636655577771701e-06,
+      "loss": 0.5812,
+      "step": 4009
+    },
+    {
+      "epoch": 1.064516129032258,
+      "grad_norm": 0.3923980546324671,
+      "learning_rate": 4.636474295158946e-06,
+      "loss": 0.5899,
+      "step": 4010
+    },
+    {
+      "epoch": 1.0647816275056419,
+      "grad_norm": 0.40326377613545966,
+      "learning_rate": 4.63629297087971e-06,
+      "loss": 0.5754,
+      "step": 4011
+    },
+    {
+      "epoch": 1.0650471259790257,
+      "grad_norm": 0.40855081579138747,
+      "learning_rate": 4.636111604937531e-06,
+      "loss": 0.623,
+      "step": 4012
+    },
+    {
+      "epoch": 1.0653126244524094,
+      "grad_norm": 0.40059093995638795,
+      "learning_rate": 4.635930197335946e-06,
+      "loss": 0.5479,
+      "step": 4013
+    },
+    {
+      "epoch": 1.0655781229257932,
+      "grad_norm": 0.3994927713733089,
+      "learning_rate": 4.635748748078493e-06,
+      "loss": 0.6155,
+      "step": 4014
+    },
+    {
+      "epoch": 1.0658436213991769,
+      "grad_norm": 0.4137964448651529,
+      "learning_rate": 4.635567257168708e-06,
+      "loss": 0.5958,
+      "step": 4015
+    },
+    {
+      "epoch": 1.0661091198725607,
+      "grad_norm": 0.41072798377557773,
+      "learning_rate": 4.635385724610134e-06,
+      "loss": 0.6297,
+      "step": 4016
+    },
+    {
+      "epoch": 1.0663746183459446,
+      "grad_norm": 0.395282013123413,
+      "learning_rate": 4.635204150406311e-06,
+      "loss": 0.592,
+      "step": 4017
+    },
+    {
+      "epoch": 1.0666401168193282,
+      "grad_norm": 0.4005879453539046,
+      "learning_rate": 4.635022534560778e-06,
+      "loss": 0.6176,
+      "step": 4018
+    },
+    {
+      "epoch": 1.066905615292712,
+      "grad_norm": 0.3991568310917474,
+      "learning_rate": 4.634840877077078e-06,
+      "loss": 0.5751,
+      "step": 4019
+    },
+    {
+      "epoch": 1.0671711137660957,
+      "grad_norm": 0.3945368982901071,
+      "learning_rate": 4.634659177958754e-06,
+      "loss": 0.5831,
+      "step": 4020
+    },
+    {
+      "epoch": 1.0674366122394796,
+      "grad_norm": 0.38894346471271196,
+      "learning_rate": 4.634477437209351e-06,
+      "loss": 0.606,
+      "step": 4021
+    },
+    {
+      "epoch": 1.0677021107128635,
+      "grad_norm": 0.40818729118424146,
+      "learning_rate": 4.63429565483241e-06,
+      "loss": 0.5838,
+      "step": 4022
+    },
+    {
+      "epoch": 1.0679676091862471,
+      "grad_norm": 0.4020142420116359,
+      "learning_rate": 4.6341138308314795e-06,
+      "loss": 0.63,
+      "step": 4023
+    },
+    {
+      "epoch": 1.068233107659631,
+      "grad_norm": 0.40365132656974057,
+      "learning_rate": 4.633931965210103e-06,
+      "loss": 0.5859,
+      "step": 4024
+    },
+    {
+      "epoch": 1.0684986061330148,
+      "grad_norm": 0.4131720649015307,
+      "learning_rate": 4.633750057971829e-06,
+      "loss": 0.5625,
+      "step": 4025
+    },
+    {
+      "epoch": 1.0687641046063985,
+      "grad_norm": 0.41242328153385804,
+      "learning_rate": 4.6335681091202035e-06,
+      "loss": 0.6144,
+      "step": 4026
+    },
+    {
+      "epoch": 1.0690296030797823,
+      "grad_norm": 0.39882614654718024,
+      "learning_rate": 4.633386118658777e-06,
+      "loss": 0.5778,
+      "step": 4027
+    },
+    {
+      "epoch": 1.069295101553166,
+      "grad_norm": 0.40762356706951475,
+      "learning_rate": 4.6332040865910975e-06,
+      "loss": 0.6065,
+      "step": 4028
+    },
+    {
+      "epoch": 1.0695606000265498,
+      "grad_norm": 0.4094125740556995,
+      "learning_rate": 4.633022012920716e-06,
+      "loss": 0.6325,
+      "step": 4029
+    },
+    {
+      "epoch": 1.0698260984999337,
+      "grad_norm": 0.4146056476165413,
+      "learning_rate": 4.632839897651182e-06,
+      "loss": 0.5754,
+      "step": 4030
+    },
+    {
+      "epoch": 1.0700915969733173,
+      "grad_norm": 0.40852967614332164,
+      "learning_rate": 4.6326577407860476e-06,
+      "loss": 0.565,
+      "step": 4031
+    },
+    {
+      "epoch": 1.0703570954467012,
+      "grad_norm": 0.3899578116984801,
+      "learning_rate": 4.632475542328866e-06,
+      "loss": 0.6179,
+      "step": 4032
+    },
+    {
+      "epoch": 1.0706225939200849,
+      "grad_norm": 0.397426110489581,
+      "learning_rate": 4.63229330228319e-06,
+      "loss": 0.5967,
+      "step": 4033
+    },
+    {
+      "epoch": 1.0708880923934687,
+      "grad_norm": 0.42837850860275356,
+      "learning_rate": 4.632111020652574e-06,
+      "loss": 0.5747,
+      "step": 4034
+    },
+    {
+      "epoch": 1.0711535908668526,
+      "grad_norm": 0.41237492707396645,
+      "learning_rate": 4.631928697440573e-06,
+      "loss": 0.6062,
+      "step": 4035
+    },
+    {
+      "epoch": 1.0714190893402362,
+      "grad_norm": 0.3887398665098403,
+      "learning_rate": 4.631746332650743e-06,
+      "loss": 0.5838,
+      "step": 4036
+    },
+    {
+      "epoch": 1.07168458781362,
+      "grad_norm": 0.4034091547286304,
+      "learning_rate": 4.631563926286639e-06,
+      "loss": 0.6039,
+      "step": 4037
+    },
+    {
+      "epoch": 1.071950086287004,
+      "grad_norm": 0.4087668165035704,
+      "learning_rate": 4.63138147835182e-06,
+      "loss": 0.6066,
+      "step": 4038
+    },
+    {
+      "epoch": 1.0722155847603876,
+      "grad_norm": 0.4166615103916408,
+      "learning_rate": 4.631198988849843e-06,
+      "loss": 0.6068,
+      "step": 4039
+    },
+    {
+      "epoch": 1.0724810832337714,
+      "grad_norm": 0.43581208237862434,
+      "learning_rate": 4.631016457784269e-06,
+      "loss": 0.6402,
+      "step": 4040
+    },
+    {
+      "epoch": 1.072746581707155,
+      "grad_norm": 0.4001290711025062,
+      "learning_rate": 4.630833885158656e-06,
+      "loss": 0.6372,
+      "step": 4041
+    },
+    {
+      "epoch": 1.073012080180539,
+      "grad_norm": 0.40624572138555,
+      "learning_rate": 4.630651270976564e-06,
+      "loss": 0.6018,
+      "step": 4042
+    },
+    {
+      "epoch": 1.0732775786539228,
+      "grad_norm": 0.391118087302682,
+      "learning_rate": 4.630468615241556e-06,
+      "loss": 0.5695,
+      "step": 4043
+    },
+    {
+      "epoch": 1.0735430771273065,
+      "grad_norm": 0.4027048041466175,
+      "learning_rate": 4.630285917957195e-06,
+      "loss": 0.5849,
+      "step": 4044
+    },
+    {
+      "epoch": 1.0738085756006903,
+      "grad_norm": 0.4530159977697565,
+      "learning_rate": 4.630103179127041e-06,
+      "loss": 0.5549,
+      "step": 4045
+    },
+    {
+      "epoch": 1.074074074074074,
+      "grad_norm": 0.40028759639303,
+      "learning_rate": 4.629920398754661e-06,
+      "loss": 0.5684,
+      "step": 4046
+    },
+    {
+      "epoch": 1.0743395725474578,
+      "grad_norm": 0.4210963743094575,
+      "learning_rate": 4.629737576843617e-06,
+      "loss": 0.5887,
+      "step": 4047
+    },
+    {
+      "epoch": 1.0746050710208417,
+      "grad_norm": 0.4128290360089716,
+      "learning_rate": 4.629554713397476e-06,
+      "loss": 0.5944,
+      "step": 4048
+    },
+    {
+      "epoch": 1.0748705694942253,
+      "grad_norm": 0.40629500661141016,
+      "learning_rate": 4.629371808419805e-06,
+      "loss": 0.6134,
+      "step": 4049
+    },
+    {
+      "epoch": 1.0751360679676092,
+      "grad_norm": 0.41294341940243207,
+      "learning_rate": 4.629188861914169e-06,
+      "loss": 0.619,
+      "step": 4050
+    },
+    {
+      "epoch": 1.075401566440993,
+      "grad_norm": 0.39827831627336674,
+      "learning_rate": 4.629005873884137e-06,
+      "loss": 0.6145,
+      "step": 4051
+    },
+    {
+      "epoch": 1.0756670649143767,
+      "grad_norm": 0.38773555444590707,
+      "learning_rate": 4.6288228443332786e-06,
+      "loss": 0.5963,
+      "step": 4052
+    },
+    {
+      "epoch": 1.0759325633877606,
+      "grad_norm": 0.40527077192532657,
+      "learning_rate": 4.628639773265162e-06,
+      "loss": 0.6337,
+      "step": 4053
+    },
+    {
+      "epoch": 1.0761980618611442,
+      "grad_norm": 0.3985440810812894,
+      "learning_rate": 4.628456660683358e-06,
+      "loss": 0.5574,
+      "step": 4054
+    },
+    {
+      "epoch": 1.076463560334528,
+      "grad_norm": 0.39393920952517114,
+      "learning_rate": 4.628273506591439e-06,
+      "loss": 0.5839,
+      "step": 4055
+    },
+    {
+      "epoch": 1.076729058807912,
+      "grad_norm": 0.3954995575609508,
+      "learning_rate": 4.628090310992975e-06,
+      "loss": 0.5815,
+      "step": 4056
+    },
+    {
+      "epoch": 1.0769945572812956,
+      "grad_norm": 0.4053163964827301,
+      "learning_rate": 4.627907073891538e-06,
+      "loss": 0.5851,
+      "step": 4057
+    },
+    {
+      "epoch": 1.0772600557546794,
+      "grad_norm": 0.4087608732005017,
+      "learning_rate": 4.627723795290704e-06,
+      "loss": 0.594,
+      "step": 4058
+    },
+    {
+      "epoch": 1.0775255542280633,
+      "grad_norm": 0.41284048213269314,
+      "learning_rate": 4.6275404751940465e-06,
+      "loss": 0.6287,
+      "step": 4059
+    },
+    {
+      "epoch": 1.077791052701447,
+      "grad_norm": 0.4075695186606942,
+      "learning_rate": 4.627357113605141e-06,
+      "loss": 0.5922,
+      "step": 4060
+    },
+    {
+      "epoch": 1.0780565511748308,
+      "grad_norm": 0.4081051314218715,
+      "learning_rate": 4.627173710527563e-06,
+      "loss": 0.5752,
+      "step": 4061
+    },
+    {
+      "epoch": 1.0783220496482144,
+      "grad_norm": 0.4187394365033306,
+      "learning_rate": 4.626990265964889e-06,
+      "loss": 0.5719,
+      "step": 4062
+    },
+    {
+      "epoch": 1.0785875481215983,
+      "grad_norm": 0.4112295225806692,
+      "learning_rate": 4.626806779920697e-06,
+      "loss": 0.6388,
+      "step": 4063
+    },
+    {
+      "epoch": 1.0788530465949822,
+      "grad_norm": 0.4100271543627126,
+      "learning_rate": 4.626623252398566e-06,
+      "loss": 0.6075,
+      "step": 4064
+    },
+    {
+      "epoch": 1.0791185450683658,
+      "grad_norm": 0.41631544639993207,
+      "learning_rate": 4.626439683402074e-06,
+      "loss": 0.5845,
+      "step": 4065
+    },
+    {
+      "epoch": 1.0793840435417497,
+      "grad_norm": 0.40470529300569175,
+      "learning_rate": 4.6262560729348025e-06,
+      "loss": 0.5815,
+      "step": 4066
+    },
+    {
+      "epoch": 1.0796495420151335,
+      "grad_norm": 0.3973788458301625,
+      "learning_rate": 4.6260724210003315e-06,
+      "loss": 0.6086,
+      "step": 4067
+    },
+    {
+      "epoch": 1.0799150404885172,
+      "grad_norm": 0.39142176023047,
+      "learning_rate": 4.6258887276022425e-06,
+      "loss": 0.5778,
+      "step": 4068
+    },
+    {
+      "epoch": 1.080180538961901,
+      "grad_norm": 0.40524163985892986,
+      "learning_rate": 4.625704992744118e-06,
+      "loss": 0.6152,
+      "step": 4069
+    },
+    {
+      "epoch": 1.0804460374352847,
+      "grad_norm": 0.3999464567193446,
+      "learning_rate": 4.625521216429543e-06,
+      "loss": 0.5923,
+      "step": 4070
+    },
+    {
+      "epoch": 1.0807115359086685,
+      "grad_norm": 0.40619490805430397,
+      "learning_rate": 4.6253373986620985e-06,
+      "loss": 0.5934,
+      "step": 4071
+    },
+    {
+      "epoch": 1.0809770343820524,
+      "grad_norm": 0.3931726727012223,
+      "learning_rate": 4.625153539445371e-06,
+      "loss": 0.5917,
+      "step": 4072
+    },
+    {
+      "epoch": 1.081242532855436,
+      "grad_norm": 0.3857731390731141,
+      "learning_rate": 4.624969638782946e-06,
+      "loss": 0.5842,
+      "step": 4073
+    },
+    {
+      "epoch": 1.08150803132882,
+      "grad_norm": 0.39874269463638917,
+      "learning_rate": 4.6247856966784115e-06,
+      "loss": 0.5879,
+      "step": 4074
+    },
+    {
+      "epoch": 1.0817735298022035,
+      "grad_norm": 0.40993896377612815,
+      "learning_rate": 4.624601713135353e-06,
+      "loss": 0.5853,
+      "step": 4075
+    },
+    {
+      "epoch": 1.0820390282755874,
+      "grad_norm": 0.4205507913908924,
+      "learning_rate": 4.624417688157359e-06,
+      "loss": 0.6193,
+      "step": 4076
+    },
+    {
+      "epoch": 1.0823045267489713,
+      "grad_norm": 0.39417773110000665,
+      "learning_rate": 4.624233621748019e-06,
+      "loss": 0.5884,
+      "step": 4077
+    },
+    {
+      "epoch": 1.082570025222355,
+      "grad_norm": 0.39894065437122356,
+      "learning_rate": 4.624049513910922e-06,
+      "loss": 0.6197,
+      "step": 4078
+    },
+    {
+      "epoch": 1.0828355236957388,
+      "grad_norm": 0.4062838970200478,
+      "learning_rate": 4.623865364649659e-06,
+      "loss": 0.6042,
+      "step": 4079
+    },
+    {
+      "epoch": 1.0831010221691226,
+      "grad_norm": 0.40102762554442134,
+      "learning_rate": 4.623681173967821e-06,
+      "loss": 0.5829,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0833665206425063,
+      "grad_norm": 0.38647607303578524,
+      "learning_rate": 4.623496941869e-06,
+      "loss": 0.5842,
+      "step": 4081
+    },
+    {
+      "epoch": 1.0836320191158901,
+      "grad_norm": 0.4074918791365656,
+      "learning_rate": 4.623312668356791e-06,
+      "loss": 0.5584,
+      "step": 4082
+    },
+    {
+      "epoch": 1.0838975175892738,
+      "grad_norm": 0.4098707063384309,
+      "learning_rate": 4.623128353434785e-06,
+      "loss": 0.6217,
+      "step": 4083
+    },
+    {
+      "epoch": 1.0841630160626576,
+      "grad_norm": 0.3963502183367459,
+      "learning_rate": 4.622943997106578e-06,
+      "loss": 0.6107,
+      "step": 4084
+    },
+    {
+      "epoch": 1.0844285145360415,
+      "grad_norm": 0.39437021559905555,
+      "learning_rate": 4.6227595993757655e-06,
+      "loss": 0.5812,
+      "step": 4085
+    },
+    {
+      "epoch": 1.0846940130094251,
+      "grad_norm": 0.3948263729511105,
+      "learning_rate": 4.622575160245944e-06,
+      "loss": 0.5957,
+      "step": 4086
+    },
+    {
+      "epoch": 1.084959511482809,
+      "grad_norm": 0.39799456259942495,
+      "learning_rate": 4.622390679720709e-06,
+      "loss": 0.6196,
+      "step": 4087
+    },
+    {
+      "epoch": 1.0852250099561926,
+      "grad_norm": 0.4061189324122895,
+      "learning_rate": 4.622206157803659e-06,
+      "loss": 0.5821,
+      "step": 4088
+    },
+    {
+      "epoch": 1.0854905084295765,
+      "grad_norm": 0.4047046102093522,
+      "learning_rate": 4.622021594498393e-06,
+      "loss": 0.5982,
+      "step": 4089
+    },
+    {
+      "epoch": 1.0857560069029604,
+      "grad_norm": 0.3858058221957461,
+      "learning_rate": 4.6218369898085105e-06,
+      "loss": 0.6049,
+      "step": 4090
+    },
+    {
+      "epoch": 1.086021505376344,
+      "grad_norm": 0.3951353034048394,
+      "learning_rate": 4.621652343737611e-06,
+      "loss": 0.6026,
+      "step": 4091
+    },
+    {
+      "epoch": 1.0862870038497279,
+      "grad_norm": 0.39666346551425696,
+      "learning_rate": 4.621467656289297e-06,
+      "loss": 0.6027,
+      "step": 4092
+    },
+    {
+      "epoch": 1.0865525023231117,
+      "grad_norm": 0.40686510723042235,
+      "learning_rate": 4.621282927467169e-06,
+      "loss": 0.6458,
+      "step": 4093
+    },
+    {
+      "epoch": 1.0868180007964954,
+      "grad_norm": 0.4139525417350658,
+      "learning_rate": 4.6210981572748305e-06,
+      "loss": 0.6135,
+      "step": 4094
+    },
+    {
+      "epoch": 1.0870834992698792,
+      "grad_norm": 0.39562366154066525,
+      "learning_rate": 4.620913345715884e-06,
+      "loss": 0.5636,
+      "step": 4095
+    },
+    {
+      "epoch": 1.0873489977432629,
+      "grad_norm": 0.39662945248108944,
+      "learning_rate": 4.620728492793934e-06,
+      "loss": 0.5956,
+      "step": 4096
+    },
+    {
+      "epoch": 1.0876144962166467,
+      "grad_norm": 0.40064245378846836,
+      "learning_rate": 4.620543598512587e-06,
+      "loss": 0.5605,
+      "step": 4097
+    },
+    {
+      "epoch": 1.0878799946900306,
+      "grad_norm": 0.40143468481824596,
+      "learning_rate": 4.620358662875448e-06,
+      "loss": 0.6084,
+      "step": 4098
+    },
+    {
+      "epoch": 1.0881454931634142,
+      "grad_norm": 0.40287424263354066,
+      "learning_rate": 4.620173685886123e-06,
+      "loss": 0.6268,
+      "step": 4099
+    },
+    {
+      "epoch": 1.088410991636798,
+      "grad_norm": 0.4098648566606787,
+      "learning_rate": 4.6199886675482195e-06,
+      "loss": 0.5562,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0886764901101817,
+      "grad_norm": 0.3864751162833137,
+      "learning_rate": 4.619803607865348e-06,
+      "loss": 0.5767,
+      "step": 4101
+    },
+    {
+      "epoch": 1.0889419885835656,
+      "grad_norm": 0.39170264086987516,
+      "learning_rate": 4.6196185068411145e-06,
+      "loss": 0.6308,
+      "step": 4102
+    },
+    {
+      "epoch": 1.0892074870569495,
+      "grad_norm": 0.3981892055421346,
+      "learning_rate": 4.619433364479131e-06,
+      "loss": 0.6012,
+      "step": 4103
+    },
+    {
+      "epoch": 1.0894729855303331,
+      "grad_norm": 0.4061562311151115,
+      "learning_rate": 4.619248180783007e-06,
+      "loss": 0.6128,
+      "step": 4104
+    },
+    {
+      "epoch": 1.089738484003717,
+      "grad_norm": 0.4023449299639812,
+      "learning_rate": 4.619062955756355e-06,
+      "loss": 0.5969,
+      "step": 4105
+    },
+    {
+      "epoch": 1.0900039824771008,
+      "grad_norm": 0.3977184108415491,
+      "learning_rate": 4.618877689402787e-06,
+      "loss": 0.5763,
+      "step": 4106
+    },
+    {
+      "epoch": 1.0902694809504845,
+      "grad_norm": 0.3961411915816232,
+      "learning_rate": 4.618692381725916e-06,
+      "loss": 0.6095,
+      "step": 4107
+    },
+    {
+      "epoch": 1.0905349794238683,
+      "grad_norm": 0.3917624753300087,
+      "learning_rate": 4.618507032729357e-06,
+      "loss": 0.6034,
+      "step": 4108
+    },
+    {
+      "epoch": 1.0908004778972522,
+      "grad_norm": 0.39639060661463604,
+      "learning_rate": 4.618321642416722e-06,
+      "loss": 0.5721,
+      "step": 4109
+    },
+    {
+      "epoch": 1.0910659763706358,
+      "grad_norm": 0.40190680241737353,
+      "learning_rate": 4.618136210791631e-06,
+      "loss": 0.5951,
+      "step": 4110
+    },
+    {
+      "epoch": 1.0913314748440197,
+      "grad_norm": 0.39548479513380613,
+      "learning_rate": 4.617950737857696e-06,
+      "loss": 0.6094,
+      "step": 4111
+    },
+    {
+      "epoch": 1.0915969733174034,
+      "grad_norm": 0.401331861874179,
+      "learning_rate": 4.617765223618536e-06,
+      "loss": 0.5679,
+      "step": 4112
+    },
+    {
+      "epoch": 1.0918624717907872,
+      "grad_norm": 0.3916152837209789,
+      "learning_rate": 4.6175796680777695e-06,
+      "loss": 0.5906,
+      "step": 4113
+    },
+    {
+      "epoch": 1.092127970264171,
+      "grad_norm": 0.4029644015801374,
+      "learning_rate": 4.617394071239014e-06,
+      "loss": 0.6126,
+      "step": 4114
+    },
+    {
+      "epoch": 1.0923934687375547,
+      "grad_norm": 0.40721124317873747,
+      "learning_rate": 4.617208433105891e-06,
+      "loss": 0.5757,
+      "step": 4115
+    },
+    {
+      "epoch": 1.0926589672109386,
+      "grad_norm": 0.3918991687289311,
+      "learning_rate": 4.617022753682019e-06,
+      "loss": 0.6086,
+      "step": 4116
+    },
+    {
+      "epoch": 1.0929244656843222,
+      "grad_norm": 0.3966487156633864,
+      "learning_rate": 4.61683703297102e-06,
+      "loss": 0.594,
+      "step": 4117
+    },
+    {
+      "epoch": 1.093189964157706,
+      "grad_norm": 0.4056797273915033,
+      "learning_rate": 4.616651270976516e-06,
+      "loss": 0.5956,
+      "step": 4118
+    },
+    {
+      "epoch": 1.09345546263109,
+      "grad_norm": 0.3897052293265004,
+      "learning_rate": 4.61646546770213e-06,
+      "loss": 0.5925,
+      "step": 4119
+    },
+    {
+      "epoch": 1.0937209611044736,
+      "grad_norm": 0.38918292541649774,
+      "learning_rate": 4.616279623151485e-06,
+      "loss": 0.5926,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0939864595778575,
+      "grad_norm": 0.3915180438517737,
+      "learning_rate": 4.616093737328206e-06,
+      "loss": 0.6066,
+      "step": 4121
+    },
+    {
+      "epoch": 1.0942519580512413,
+      "grad_norm": 0.4032178103254784,
+      "learning_rate": 4.615907810235918e-06,
+      "loss": 0.5918,
+      "step": 4122
+    },
+    {
+      "epoch": 1.094517456524625,
+      "grad_norm": 0.40296944404598123,
+      "learning_rate": 4.615721841878247e-06,
+      "loss": 0.5664,
+      "step": 4123
+    },
+    {
+      "epoch": 1.0947829549980088,
+      "grad_norm": 0.4109243861097142,
+      "learning_rate": 4.61553583225882e-06,
+      "loss": 0.6136,
+      "step": 4124
+    },
+    {
+      "epoch": 1.0950484534713925,
+      "grad_norm": 0.4008234294995445,
+      "learning_rate": 4.6153497813812645e-06,
+      "loss": 0.6106,
+      "step": 4125
+    },
+    {
+      "epoch": 1.0953139519447763,
+      "grad_norm": 0.4075644704495962,
+      "learning_rate": 4.615163689249209e-06,
+      "loss": 0.5858,
+      "step": 4126
+    },
+    {
+      "epoch": 1.0955794504181602,
+      "grad_norm": 0.47906943921243966,
+      "learning_rate": 4.614977555866283e-06,
+      "loss": 0.5796,
+      "step": 4127
+    },
+    {
+      "epoch": 1.0958449488915438,
+      "grad_norm": 0.401882044679437,
+      "learning_rate": 4.614791381236116e-06,
+      "loss": 0.6101,
+      "step": 4128
+    },
+    {
+      "epoch": 1.0961104473649277,
+      "grad_norm": 0.4225134550070061,
+      "learning_rate": 4.614605165362339e-06,
+      "loss": 0.5785,
+      "step": 4129
+    },
+    {
+      "epoch": 1.0963759458383113,
+      "grad_norm": 0.41874741040074764,
+      "learning_rate": 4.6144189082485845e-06,
+      "loss": 0.6114,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0966414443116952,
+      "grad_norm": 0.41127132375103914,
+      "learning_rate": 4.614232609898483e-06,
+      "loss": 0.6138,
+      "step": 4131
+    },
+    {
+      "epoch": 1.096906942785079,
+      "grad_norm": 0.40405057586061777,
+      "learning_rate": 4.614046270315671e-06,
+      "loss": 0.607,
+      "step": 4132
+    },
+    {
+      "epoch": 1.0971724412584627,
+      "grad_norm": 0.3952100579595354,
+      "learning_rate": 4.613859889503779e-06,
+      "loss": 0.5574,
+      "step": 4133
+    },
+    {
+      "epoch": 1.0974379397318466,
+      "grad_norm": 0.4123951325547869,
+      "learning_rate": 4.613673467466444e-06,
+      "loss": 0.6101,
+      "step": 4134
+    },
+    {
+      "epoch": 1.0977034382052304,
+      "grad_norm": 0.42640950665596883,
+      "learning_rate": 4.613487004207302e-06,
+      "loss": 0.6331,
+      "step": 4135
+    },
+    {
+      "epoch": 1.097968936678614,
+      "grad_norm": 0.42120442898790655,
+      "learning_rate": 4.613300499729988e-06,
+      "loss": 0.598,
+      "step": 4136
+    },
+    {
+      "epoch": 1.098234435151998,
+      "grad_norm": 0.3941650373034087,
+      "learning_rate": 4.61311395403814e-06,
+      "loss": 0.5913,
+      "step": 4137
+    },
+    {
+      "epoch": 1.0984999336253816,
+      "grad_norm": 0.4280069565805221,
+      "learning_rate": 4.612927367135396e-06,
+      "loss": 0.6019,
+      "step": 4138
+    },
+    {
+      "epoch": 1.0987654320987654,
+      "grad_norm": 0.3945544039784049,
+      "learning_rate": 4.6127407390253944e-06,
+      "loss": 0.5817,
+      "step": 4139
+    },
+    {
+      "epoch": 1.0990309305721493,
+      "grad_norm": 0.3899870461460316,
+      "learning_rate": 4.612554069711777e-06,
+      "loss": 0.5725,
+      "step": 4140
+    },
+    {
+      "epoch": 1.099296429045533,
+      "grad_norm": 0.40030549711494107,
+      "learning_rate": 4.612367359198182e-06,
+      "loss": 0.6193,
+      "step": 4141
+    },
+    {
+      "epoch": 1.0995619275189168,
+      "grad_norm": 0.4184377233830072,
+      "learning_rate": 4.612180607488251e-06,
+      "loss": 0.6379,
+      "step": 4142
+    },
+    {
+      "epoch": 1.0998274259923004,
+      "grad_norm": 0.4247987828366459,
+      "learning_rate": 4.611993814585628e-06,
+      "loss": 0.5804,
+      "step": 4143
+    },
+    {
+      "epoch": 1.1000929244656843,
+      "grad_norm": 0.3950831529379711,
+      "learning_rate": 4.611806980493954e-06,
+      "loss": 0.6401,
+      "step": 4144
+    },
+    {
+      "epoch": 1.1003584229390682,
+      "grad_norm": 0.4384869221097367,
+      "learning_rate": 4.611620105216872e-06,
+      "loss": 0.5582,
+      "step": 4145
+    },
+    {
+      "epoch": 1.1006239214124518,
+      "grad_norm": 0.40799503340071896,
+      "learning_rate": 4.611433188758029e-06,
+      "loss": 0.579,
+      "step": 4146
+    },
+    {
+      "epoch": 1.1008894198858357,
+      "grad_norm": 0.4006083766317225,
+      "learning_rate": 4.6112462311210695e-06,
+      "loss": 0.6246,
+      "step": 4147
+    },
+    {
+      "epoch": 1.1011549183592195,
+      "grad_norm": 0.40156418204366257,
+      "learning_rate": 4.611059232309639e-06,
+      "loss": 0.6143,
+      "step": 4148
+    },
+    {
+      "epoch": 1.1014204168326032,
+      "grad_norm": 0.39776408041954836,
+      "learning_rate": 4.610872192327385e-06,
+      "loss": 0.5993,
+      "step": 4149
+    },
+    {
+      "epoch": 1.101685915305987,
+      "grad_norm": 0.41781668304688074,
+      "learning_rate": 4.610685111177954e-06,
+      "loss": 0.566,
+      "step": 4150
+    },
+    {
+      "epoch": 1.1019514137793707,
+      "grad_norm": 0.40190497953337084,
+      "learning_rate": 4.610497988864996e-06,
+      "loss": 0.6125,
+      "step": 4151
+    },
+    {
+      "epoch": 1.1022169122527545,
+      "grad_norm": 0.393687751210559,
+      "learning_rate": 4.610310825392159e-06,
+      "loss": 0.5782,
+      "step": 4152
+    },
+    {
+      "epoch": 1.1024824107261384,
+      "grad_norm": 0.39370126201878136,
+      "learning_rate": 4.610123620763095e-06,
+      "loss": 0.5889,
+      "step": 4153
+    },
+    {
+      "epoch": 1.102747909199522,
+      "grad_norm": 0.4067620143883082,
+      "learning_rate": 4.609936374981454e-06,
+      "loss": 0.6114,
+      "step": 4154
+    },
+    {
+      "epoch": 1.103013407672906,
+      "grad_norm": 0.40541603036720636,
+      "learning_rate": 4.6097490880508875e-06,
+      "loss": 0.5635,
+      "step": 4155
+    },
+    {
+      "epoch": 1.1032789061462898,
+      "grad_norm": 0.40218555063238504,
+      "learning_rate": 4.609561759975047e-06,
+      "loss": 0.6061,
+      "step": 4156
+    },
+    {
+      "epoch": 1.1035444046196734,
+      "grad_norm": 0.43697665314011647,
+      "learning_rate": 4.609374390757589e-06,
+      "loss": 0.6246,
+      "step": 4157
+    },
+    {
+      "epoch": 1.1038099030930573,
+      "grad_norm": 0.4184297102929878,
+      "learning_rate": 4.609186980402165e-06,
+      "loss": 0.5659,
+      "step": 4158
+    },
+    {
+      "epoch": 1.104075401566441,
+      "grad_norm": 0.4053245845440966,
+      "learning_rate": 4.60899952891243e-06,
+      "loss": 0.6146,
+      "step": 4159
+    },
+    {
+      "epoch": 1.1043409000398248,
+      "grad_norm": 0.38890252961138627,
+      "learning_rate": 4.6088120362920416e-06,
+      "loss": 0.5911,
+      "step": 4160
+    },
+    {
+      "epoch": 1.1046063985132086,
+      "grad_norm": 0.40299281327320086,
+      "learning_rate": 4.6086245025446544e-06,
+      "loss": 0.5902,
+      "step": 4161
+    },
+    {
+      "epoch": 1.1048718969865923,
+      "grad_norm": 0.38227546403827056,
+      "learning_rate": 4.608436927673927e-06,
+      "loss": 0.6026,
+      "step": 4162
+    },
+    {
+      "epoch": 1.1051373954599761,
+      "grad_norm": 0.40271843035801363,
+      "learning_rate": 4.608249311683517e-06,
+      "loss": 0.58,
+      "step": 4163
+    },
+    {
+      "epoch": 1.10540289393336,
+      "grad_norm": 0.39399921190873083,
+      "learning_rate": 4.608061654577084e-06,
+      "loss": 0.6008,
+      "step": 4164
+    },
+    {
+      "epoch": 1.1056683924067436,
+      "grad_norm": 0.3894925190695123,
+      "learning_rate": 4.607873956358286e-06,
+      "loss": 0.6032,
+      "step": 4165
+    },
+    {
+      "epoch": 1.1059338908801275,
+      "grad_norm": 0.4099472601363349,
+      "learning_rate": 4.6076862170307865e-06,
+      "loss": 0.6161,
+      "step": 4166
+    },
+    {
+      "epoch": 1.1061993893535111,
+      "grad_norm": 0.4024169834283145,
+      "learning_rate": 4.607498436598244e-06,
+      "loss": 0.5597,
+      "step": 4167
+    },
+    {
+      "epoch": 1.106464887826895,
+      "grad_norm": 0.38204717576175373,
+      "learning_rate": 4.607310615064323e-06,
+      "loss": 0.5775,
+      "step": 4168
+    },
+    {
+      "epoch": 1.1067303863002789,
+      "grad_norm": 0.3989533946769941,
+      "learning_rate": 4.607122752432685e-06,
+      "loss": 0.6101,
+      "step": 4169
+    },
+    {
+      "epoch": 1.1069958847736625,
+      "grad_norm": 0.3876225944596314,
+      "learning_rate": 4.606934848706994e-06,
+      "loss": 0.564,
+      "step": 4170
+    },
+    {
+      "epoch": 1.1072613832470464,
+      "grad_norm": 0.4050181428459777,
+      "learning_rate": 4.606746903890914e-06,
+      "loss": 0.5484,
+      "step": 4171
+    },
+    {
+      "epoch": 1.10752688172043,
+      "grad_norm": 0.39445860470503064,
+      "learning_rate": 4.606558917988113e-06,
+      "loss": 0.5662,
+      "step": 4172
+    },
+    {
+      "epoch": 1.1077923801938139,
+      "grad_norm": 0.4192376845576687,
+      "learning_rate": 4.606370891002254e-06,
+      "loss": 0.6145,
+      "step": 4173
+    },
+    {
+      "epoch": 1.1080578786671977,
+      "grad_norm": 0.40777181982949945,
+      "learning_rate": 4.606182822937006e-06,
+      "loss": 0.5729,
+      "step": 4174
+    },
+    {
+      "epoch": 1.1083233771405814,
+      "grad_norm": 0.40447915141031576,
+      "learning_rate": 4.605994713796036e-06,
+      "loss": 0.6128,
+      "step": 4175
+    },
+    {
+      "epoch": 1.1085888756139652,
+      "grad_norm": 0.40851286920143803,
+      "learning_rate": 4.605806563583012e-06,
+      "loss": 0.6629,
+      "step": 4176
+    },
+    {
+      "epoch": 1.108854374087349,
+      "grad_norm": 0.43285146242118144,
+      "learning_rate": 4.605618372301605e-06,
+      "loss": 0.5715,
+      "step": 4177
+    },
+    {
+      "epoch": 1.1091198725607327,
+      "grad_norm": 0.42470963243260174,
+      "learning_rate": 4.605430139955484e-06,
+      "loss": 0.6126,
+      "step": 4178
+    },
+    {
+      "epoch": 1.1093853710341166,
+      "grad_norm": 0.411138195282551,
+      "learning_rate": 4.6052418665483204e-06,
+      "loss": 0.5907,
+      "step": 4179
+    },
+    {
+      "epoch": 1.1096508695075002,
+      "grad_norm": 0.40846999919174365,
+      "learning_rate": 4.605053552083786e-06,
+      "loss": 0.6031,
+      "step": 4180
+    },
+    {
+      "epoch": 1.109916367980884,
+      "grad_norm": 0.4038190181208699,
+      "learning_rate": 4.604865196565553e-06,
+      "loss": 0.5914,
+      "step": 4181
+    },
+    {
+      "epoch": 1.110181866454268,
+      "grad_norm": 0.41855550063869873,
+      "learning_rate": 4.604676799997295e-06,
+      "loss": 0.609,
+      "step": 4182
+    },
+    {
+      "epoch": 1.1104473649276516,
+      "grad_norm": 0.39641413680209786,
+      "learning_rate": 4.6044883623826866e-06,
+      "loss": 0.5857,
+      "step": 4183
+    },
+    {
+      "epoch": 1.1107128634010355,
+      "grad_norm": 0.3887209537307622,
+      "learning_rate": 4.604299883725401e-06,
+      "loss": 0.6112,
+      "step": 4184
+    },
+    {
+      "epoch": 1.1109783618744191,
+      "grad_norm": 0.3997013232689751,
+      "learning_rate": 4.604111364029118e-06,
+      "loss": 0.5903,
+      "step": 4185
+    },
+    {
+      "epoch": 1.111243860347803,
+      "grad_norm": 0.4085127662316873,
+      "learning_rate": 4.60392280329751e-06,
+      "loss": 0.6012,
+      "step": 4186
+    },
+    {
+      "epoch": 1.1115093588211868,
+      "grad_norm": 0.41870204612786327,
+      "learning_rate": 4.603734201534257e-06,
+      "loss": 0.6106,
+      "step": 4187
+    },
+    {
+      "epoch": 1.1117748572945705,
+      "grad_norm": 0.4047874010785604,
+      "learning_rate": 4.603545558743035e-06,
+      "loss": 0.611,
+      "step": 4188
+    },
+    {
+      "epoch": 1.1120403557679543,
+      "grad_norm": 0.4086861990194707,
+      "learning_rate": 4.603356874927525e-06,
+      "loss": 0.578,
+      "step": 4189
+    },
+    {
+      "epoch": 1.1123058542413382,
+      "grad_norm": 0.4185807284404185,
+      "learning_rate": 4.603168150091406e-06,
+      "loss": 0.634,
+      "step": 4190
+    },
+    {
+      "epoch": 1.1125713527147219,
+      "grad_norm": 0.4091818793904963,
+      "learning_rate": 4.602979384238359e-06,
+      "loss": 0.5847,
+      "step": 4191
+    },
+    {
+      "epoch": 1.1128368511881057,
+      "grad_norm": 0.4168123995392763,
+      "learning_rate": 4.602790577372064e-06,
+      "loss": 0.5905,
+      "step": 4192
+    },
+    {
+      "epoch": 1.1131023496614894,
+      "grad_norm": 0.4003934481060536,
+      "learning_rate": 4.602601729496204e-06,
+      "loss": 0.6077,
+      "step": 4193
+    },
+    {
+      "epoch": 1.1133678481348732,
+      "grad_norm": 0.4031037033735679,
+      "learning_rate": 4.602412840614463e-06,
+      "loss": 0.5864,
+      "step": 4194
+    },
+    {
+      "epoch": 1.113633346608257,
+      "grad_norm": 0.40682334686164157,
+      "learning_rate": 4.602223910730525e-06,
+      "loss": 0.5964,
+      "step": 4195
+    },
+    {
+      "epoch": 1.1138988450816407,
+      "grad_norm": 0.4145387287475848,
+      "learning_rate": 4.602034939848072e-06,
+      "loss": 0.6053,
+      "step": 4196
+    },
+    {
+      "epoch": 1.1141643435550246,
+      "grad_norm": 0.40240899343747966,
+      "learning_rate": 4.601845927970792e-06,
+      "loss": 0.587,
+      "step": 4197
+    },
+    {
+      "epoch": 1.1144298420284082,
+      "grad_norm": 0.40101775629577413,
+      "learning_rate": 4.60165687510237e-06,
+      "loss": 0.5692,
+      "step": 4198
+    },
+    {
+      "epoch": 1.114695340501792,
+      "grad_norm": 0.3935512473913428,
+      "learning_rate": 4.601467781246492e-06,
+      "loss": 0.5788,
+      "step": 4199
+    },
+    {
+      "epoch": 1.114960838975176,
+      "grad_norm": 0.4044175962249614,
+      "learning_rate": 4.601278646406849e-06,
+      "loss": 0.5709,
+      "step": 4200
+    },
+    {
+      "epoch": 1.1152263374485596,
+      "grad_norm": 0.42054058538944394,
+      "learning_rate": 4.601089470587126e-06,
+      "loss": 0.5743,
+      "step": 4201
+    },
+    {
+      "epoch": 1.1154918359219435,
+      "grad_norm": 0.41230296755811063,
+      "learning_rate": 4.600900253791015e-06,
+      "loss": 0.631,
+      "step": 4202
+    },
+    {
+      "epoch": 1.1157573343953273,
+      "grad_norm": 0.40862673492388485,
+      "learning_rate": 4.6007109960222045e-06,
+      "loss": 0.5977,
+      "step": 4203
+    },
+    {
+      "epoch": 1.116022832868711,
+      "grad_norm": 0.3962651528454553,
+      "learning_rate": 4.600521697284387e-06,
+      "loss": 0.5681,
+      "step": 4204
+    },
+    {
+      "epoch": 1.1162883313420948,
+      "grad_norm": 0.41969620413148845,
+      "learning_rate": 4.600332357581252e-06,
+      "loss": 0.6299,
+      "step": 4205
+    },
+    {
+      "epoch": 1.1165538298154785,
+      "grad_norm": 0.41104348682358244,
+      "learning_rate": 4.600142976916494e-06,
+      "loss": 0.6239,
+      "step": 4206
+    },
+    {
+      "epoch": 1.1168193282888623,
+      "grad_norm": 0.39792093294818626,
+      "learning_rate": 4.599953555293807e-06,
+      "loss": 0.614,
+      "step": 4207
+    },
+    {
+      "epoch": 1.1170848267622462,
+      "grad_norm": 0.41806139008693155,
+      "learning_rate": 4.599764092716883e-06,
+      "loss": 0.588,
+      "step": 4208
+    },
+    {
+      "epoch": 1.1173503252356298,
+      "grad_norm": 0.4290109737688029,
+      "learning_rate": 4.599574589189419e-06,
+      "loss": 0.5726,
+      "step": 4209
+    },
+    {
+      "epoch": 1.1176158237090137,
+      "grad_norm": 0.40931412337661655,
+      "learning_rate": 4.599385044715109e-06,
+      "loss": 0.5904,
+      "step": 4210
+    },
+    {
+      "epoch": 1.1178813221823976,
+      "grad_norm": 0.41267087407676684,
+      "learning_rate": 4.5991954592976504e-06,
+      "loss": 0.5555,
+      "step": 4211
+    },
+    {
+      "epoch": 1.1181468206557812,
+      "grad_norm": 0.395245460021575,
+      "learning_rate": 4.5990058329407405e-06,
+      "loss": 0.6214,
+      "step": 4212
+    },
+    {
+      "epoch": 1.118412319129165,
+      "grad_norm": 0.40110320470953037,
+      "learning_rate": 4.5988161656480785e-06,
+      "loss": 0.6143,
+      "step": 4213
+    },
+    {
+      "epoch": 1.1186778176025487,
+      "grad_norm": 0.3890728206523804,
+      "learning_rate": 4.5986264574233615e-06,
+      "loss": 0.6062,
+      "step": 4214
+    },
+    {
+      "epoch": 1.1189433160759326,
+      "grad_norm": 0.3981589273212458,
+      "learning_rate": 4.598436708270291e-06,
+      "loss": 0.6165,
+      "step": 4215
+    },
+    {
+      "epoch": 1.1192088145493164,
+      "grad_norm": 0.40279097620937515,
+      "learning_rate": 4.5982469181925675e-06,
+      "loss": 0.6039,
+      "step": 4216
+    },
+    {
+      "epoch": 1.1194743130227,
+      "grad_norm": 0.402564655796646,
+      "learning_rate": 4.59805708719389e-06,
+      "loss": 0.6026,
+      "step": 4217
+    },
+    {
+      "epoch": 1.119739811496084,
+      "grad_norm": 0.3979570934000749,
+      "learning_rate": 4.597867215277964e-06,
+      "loss": 0.6059,
+      "step": 4218
+    },
+    {
+      "epoch": 1.1200053099694678,
+      "grad_norm": 0.4145037547485939,
+      "learning_rate": 4.59767730244849e-06,
+      "loss": 0.614,
+      "step": 4219
+    },
+    {
+      "epoch": 1.1202708084428514,
+      "grad_norm": 0.4095470127521354,
+      "learning_rate": 4.597487348709172e-06,
+      "loss": 0.556,
+      "step": 4220
+    },
+    {
+      "epoch": 1.1205363069162353,
+      "grad_norm": 0.398477200715919,
+      "learning_rate": 4.597297354063716e-06,
+      "loss": 0.5985,
+      "step": 4221
+    },
+    {
+      "epoch": 1.120801805389619,
+      "grad_norm": 0.4088072729297395,
+      "learning_rate": 4.597107318515826e-06,
+      "loss": 0.6221,
+      "step": 4222
+    },
+    {
+      "epoch": 1.1210673038630028,
+      "grad_norm": 0.3893122622904414,
+      "learning_rate": 4.596917242069209e-06,
+      "loss": 0.5848,
+      "step": 4223
+    },
+    {
+      "epoch": 1.1213328023363867,
+      "grad_norm": 0.4255312349984939,
+      "learning_rate": 4.596727124727572e-06,
+      "loss": 0.5795,
+      "step": 4224
+    },
+    {
+      "epoch": 1.1215983008097703,
+      "grad_norm": 0.40142865884755996,
+      "learning_rate": 4.596536966494622e-06,
+      "loss": 0.5786,
+      "step": 4225
+    },
+    {
+      "epoch": 1.1218637992831542,
+      "grad_norm": 0.40047260435012183,
+      "learning_rate": 4.596346767374068e-06,
+      "loss": 0.5806,
+      "step": 4226
+    },
+    {
+      "epoch": 1.1221292977565378,
+      "grad_norm": 0.42233976959533925,
+      "learning_rate": 4.5961565273696194e-06,
+      "loss": 0.6002,
+      "step": 4227
+    },
+    {
+      "epoch": 1.1223947962299217,
+      "grad_norm": 0.40616003543509005,
+      "learning_rate": 4.595966246484986e-06,
+      "loss": 0.5399,
+      "step": 4228
+    },
+    {
+      "epoch": 1.1226602947033055,
+      "grad_norm": 0.4098600200834036,
+      "learning_rate": 4.595775924723879e-06,
+      "loss": 0.6076,
+      "step": 4229
+    },
+    {
+      "epoch": 1.1229257931766892,
+      "grad_norm": 0.4134511063073729,
+      "learning_rate": 4.59558556209001e-06,
+      "loss": 0.5387,
+      "step": 4230
+    },
+    {
+      "epoch": 1.123191291650073,
+      "grad_norm": 0.38366695581028937,
+      "learning_rate": 4.595395158587093e-06,
+      "loss": 0.6061,
+      "step": 4231
+    },
+    {
+      "epoch": 1.123456790123457,
+      "grad_norm": 0.4034998142744694,
+      "learning_rate": 4.595204714218838e-06,
+      "loss": 0.6211,
+      "step": 4232
+    },
+    {
+      "epoch": 1.1237222885968405,
+      "grad_norm": 0.3951363817294016,
+      "learning_rate": 4.595014228988962e-06,
+      "loss": 0.5799,
+      "step": 4233
+    },
+    {
+      "epoch": 1.1239877870702244,
+      "grad_norm": 0.3979057271878824,
+      "learning_rate": 4.594823702901179e-06,
+      "loss": 0.5857,
+      "step": 4234
+    },
+    {
+      "epoch": 1.124253285543608,
+      "grad_norm": 0.38888340991930065,
+      "learning_rate": 4.594633135959205e-06,
+      "loss": 0.6034,
+      "step": 4235
+    },
+    {
+      "epoch": 1.124518784016992,
+      "grad_norm": 0.3955025703873869,
+      "learning_rate": 4.594442528166756e-06,
+      "loss": 0.5922,
+      "step": 4236
+    },
+    {
+      "epoch": 1.1247842824903758,
+      "grad_norm": 0.39095324444771157,
+      "learning_rate": 4.594251879527551e-06,
+      "loss": 0.5933,
+      "step": 4237
+    },
+    {
+      "epoch": 1.1250497809637594,
+      "grad_norm": 0.41100452779958446,
+      "learning_rate": 4.594061190045305e-06,
+      "loss": 0.6418,
+      "step": 4238
+    },
+    {
+      "epoch": 1.1253152794371433,
+      "grad_norm": 0.4002378528278809,
+      "learning_rate": 4.593870459723738e-06,
+      "loss": 0.5866,
+      "step": 4239
+    },
+    {
+      "epoch": 1.125580777910527,
+      "grad_norm": 0.41591673394122,
+      "learning_rate": 4.5936796885665725e-06,
+      "loss": 0.5953,
+      "step": 4240
+    },
+    {
+      "epoch": 1.1258462763839108,
+      "grad_norm": 0.4015448820218692,
+      "learning_rate": 4.593488876577526e-06,
+      "loss": 0.6108,
+      "step": 4241
+    },
+    {
+      "epoch": 1.1261117748572946,
+      "grad_norm": 0.3976611257233658,
+      "learning_rate": 4.5932980237603195e-06,
+      "loss": 0.586,
+      "step": 4242
+    },
+    {
+      "epoch": 1.1263772733306783,
+      "grad_norm": 0.3977989655418618,
+      "learning_rate": 4.593107130118678e-06,
+      "loss": 0.5393,
+      "step": 4243
+    },
+    {
+      "epoch": 1.1266427718040621,
+      "grad_norm": 0.4022883236779997,
+      "learning_rate": 4.592916195656322e-06,
+      "loss": 0.6202,
+      "step": 4244
+    },
+    {
+      "epoch": 1.126908270277446,
+      "grad_norm": 0.3966396273278027,
+      "learning_rate": 4.592725220376976e-06,
+      "loss": 0.614,
+      "step": 4245
+    },
+    {
+      "epoch": 1.1271737687508296,
+      "grad_norm": 0.39230237441538535,
+      "learning_rate": 4.592534204284364e-06,
+      "loss": 0.5658,
+      "step": 4246
+    },
+    {
+      "epoch": 1.1274392672242135,
+      "grad_norm": 0.40309995569915774,
+      "learning_rate": 4.592343147382211e-06,
+      "loss": 0.6273,
+      "step": 4247
+    },
+    {
+      "epoch": 1.1277047656975971,
+      "grad_norm": 0.39282030743083224,
+      "learning_rate": 4.592152049674245e-06,
+      "loss": 0.6398,
+      "step": 4248
+    },
+    {
+      "epoch": 1.127970264170981,
+      "grad_norm": 0.39215854767023556,
+      "learning_rate": 4.59196091116419e-06,
+      "loss": 0.5592,
+      "step": 4249
+    },
+    {
+      "epoch": 1.1282357626443649,
+      "grad_norm": 0.4024622746384169,
+      "learning_rate": 4.591769731855777e-06,
+      "loss": 0.6334,
+      "step": 4250
+    },
+    {
+      "epoch": 1.1285012611177485,
+      "grad_norm": 0.39404603695945745,
+      "learning_rate": 4.591578511752733e-06,
+      "loss": 0.5959,
+      "step": 4251
+    },
+    {
+      "epoch": 1.1287667595911324,
+      "grad_norm": 0.40931903864851577,
+      "learning_rate": 4.591387250858785e-06,
+      "loss": 0.6112,
+      "step": 4252
+    },
+    {
+      "epoch": 1.129032258064516,
+      "grad_norm": 0.4150392582605503,
+      "learning_rate": 4.591195949177666e-06,
+      "loss": 0.6164,
+      "step": 4253
+    },
+    {
+      "epoch": 1.1292977565378999,
+      "grad_norm": 0.4067786036855695,
+      "learning_rate": 4.591004606713105e-06,
+      "loss": 0.6024,
+      "step": 4254
+    },
+    {
+      "epoch": 1.1295632550112837,
+      "grad_norm": 0.40194099726084326,
+      "learning_rate": 4.590813223468836e-06,
+      "loss": 0.6233,
+      "step": 4255
+    },
+    {
+      "epoch": 1.1298287534846674,
+      "grad_norm": 0.3902873137041122,
+      "learning_rate": 4.590621799448589e-06,
+      "loss": 0.5662,
+      "step": 4256
+    },
+    {
+      "epoch": 1.1300942519580512,
+      "grad_norm": 0.429037891563439,
+      "learning_rate": 4.5904303346560984e-06,
+      "loss": 0.6001,
+      "step": 4257
+    },
+    {
+      "epoch": 1.130359750431435,
+      "grad_norm": 0.40232873097465577,
+      "learning_rate": 4.590238829095098e-06,
+      "loss": 0.5579,
+      "step": 4258
+    },
+    {
+      "epoch": 1.1306252489048187,
+      "grad_norm": 0.3838087291615794,
+      "learning_rate": 4.590047282769322e-06,
+      "loss": 0.5901,
+      "step": 4259
+    },
+    {
+      "epoch": 1.1308907473782026,
+      "grad_norm": 0.4049598737557053,
+      "learning_rate": 4.589855695682508e-06,
+      "loss": 0.602,
+      "step": 4260
+    },
+    {
+      "epoch": 1.1311562458515865,
+      "grad_norm": 0.40249650479098137,
+      "learning_rate": 4.5896640678383895e-06,
+      "loss": 0.5591,
+      "step": 4261
+    },
+    {
+      "epoch": 1.1314217443249701,
+      "grad_norm": 0.3848978703388012,
+      "learning_rate": 4.589472399240707e-06,
+      "loss": 0.5842,
+      "step": 4262
+    },
+    {
+      "epoch": 1.131687242798354,
+      "grad_norm": 0.4147613415923532,
+      "learning_rate": 4.589280689893195e-06,
+      "loss": 0.5776,
+      "step": 4263
+    },
+    {
+      "epoch": 1.1319527412717376,
+      "grad_norm": 0.4241405181007222,
+      "learning_rate": 4.589088939799596e-06,
+      "loss": 0.5243,
+      "step": 4264
+    },
+    {
+      "epoch": 1.1322182397451215,
+      "grad_norm": 0.4002949365799512,
+      "learning_rate": 4.588897148963647e-06,
+      "loss": 0.5627,
+      "step": 4265
+    },
+    {
+      "epoch": 1.1324837382185051,
+      "grad_norm": 0.4269931837453073,
+      "learning_rate": 4.588705317389088e-06,
+      "loss": 0.6095,
+      "step": 4266
+    },
+    {
+      "epoch": 1.132749236691889,
+      "grad_norm": 0.41039316719331886,
+      "learning_rate": 4.588513445079663e-06,
+      "loss": 0.6113,
+      "step": 4267
+    },
+    {
+      "epoch": 1.1330147351652728,
+      "grad_norm": 0.403715450210771,
+      "learning_rate": 4.588321532039111e-06,
+      "loss": 0.6006,
+      "step": 4268
+    },
+    {
+      "epoch": 1.1332802336386565,
+      "grad_norm": 0.40248862211140435,
+      "learning_rate": 4.5881295782711764e-06,
+      "loss": 0.593,
+      "step": 4269
+    },
+    {
+      "epoch": 1.1335457321120403,
+      "grad_norm": 0.4179157629800408,
+      "learning_rate": 4.5879375837796026e-06,
+      "loss": 0.6275,
+      "step": 4270
+    },
+    {
+      "epoch": 1.1338112305854242,
+      "grad_norm": 0.3936150303902074,
+      "learning_rate": 4.5877455485681336e-06,
+      "loss": 0.6174,
+      "step": 4271
+    },
+    {
+      "epoch": 1.1340767290588079,
+      "grad_norm": 0.4191352127077562,
+      "learning_rate": 4.587553472640515e-06,
+      "loss": 0.6029,
+      "step": 4272
+    },
+    {
+      "epoch": 1.1343422275321917,
+      "grad_norm": 0.39581654999444527,
+      "learning_rate": 4.5873613560004925e-06,
+      "loss": 0.6259,
+      "step": 4273
+    },
+    {
+      "epoch": 1.1346077260055756,
+      "grad_norm": 0.40456689687387054,
+      "learning_rate": 4.5871691986518115e-06,
+      "loss": 0.5986,
+      "step": 4274
+    },
+    {
+      "epoch": 1.1348732244789592,
+      "grad_norm": 0.395604897033609,
+      "learning_rate": 4.586977000598223e-06,
+      "loss": 0.5928,
+      "step": 4275
+    },
+    {
+      "epoch": 1.135138722952343,
+      "grad_norm": 0.40476824186348964,
+      "learning_rate": 4.586784761843471e-06,
+      "loss": 0.5922,
+      "step": 4276
+    },
+    {
+      "epoch": 1.1354042214257267,
+      "grad_norm": 0.3995343438607479,
+      "learning_rate": 4.586592482391309e-06,
+      "loss": 0.5745,
+      "step": 4277
+    },
+    {
+      "epoch": 1.1356697198991106,
+      "grad_norm": 0.41523553225996046,
+      "learning_rate": 4.586400162245483e-06,
+      "loss": 0.6398,
+      "step": 4278
+    },
+    {
+      "epoch": 1.1359352183724944,
+      "grad_norm": 0.39780621096452673,
+      "learning_rate": 4.586207801409746e-06,
+      "loss": 0.5837,
+      "step": 4279
+    },
+    {
+      "epoch": 1.136200716845878,
+      "grad_norm": 0.40245607014516194,
+      "learning_rate": 4.586015399887849e-06,
+      "loss": 0.5653,
+      "step": 4280
+    },
+    {
+      "epoch": 1.136466215319262,
+      "grad_norm": 0.4074364049507209,
+      "learning_rate": 4.585822957683544e-06,
+      "loss": 0.5964,
+      "step": 4281
+    },
+    {
+      "epoch": 1.1367317137926456,
+      "grad_norm": 0.4288557939725421,
+      "learning_rate": 4.585630474800585e-06,
+      "loss": 0.6124,
+      "step": 4282
+    },
+    {
+      "epoch": 1.1369972122660295,
+      "grad_norm": 0.41661816147268377,
+      "learning_rate": 4.585437951242725e-06,
+      "loss": 0.5748,
+      "step": 4283
+    },
+    {
+      "epoch": 1.1372627107394133,
+      "grad_norm": 0.44452198381336505,
+      "learning_rate": 4.585245387013719e-06,
+      "loss": 0.6009,
+      "step": 4284
+    },
+    {
+      "epoch": 1.137528209212797,
+      "grad_norm": 0.4003785475367137,
+      "learning_rate": 4.585052782117322e-06,
+      "loss": 0.5803,
+      "step": 4285
+    },
+    {
+      "epoch": 1.1377937076861808,
+      "grad_norm": 0.40938385164890345,
+      "learning_rate": 4.584860136557291e-06,
+      "loss": 0.5789,
+      "step": 4286
+    },
+    {
+      "epoch": 1.1380592061595647,
+      "grad_norm": 0.39615381997230703,
+      "learning_rate": 4.584667450337382e-06,
+      "loss": 0.5914,
+      "step": 4287
+    },
+    {
+      "epoch": 1.1383247046329483,
+      "grad_norm": 0.40140944385360927,
+      "learning_rate": 4.584474723461355e-06,
+      "loss": 0.5996,
+      "step": 4288
+    },
+    {
+      "epoch": 1.1385902031063322,
+      "grad_norm": 0.3964996407844935,
+      "learning_rate": 4.5842819559329655e-06,
+      "loss": 0.5643,
+      "step": 4289
+    },
+    {
+      "epoch": 1.1388557015797158,
+      "grad_norm": 0.39738695020501213,
+      "learning_rate": 4.584089147755976e-06,
+      "loss": 0.5917,
+      "step": 4290
+    },
+    {
+      "epoch": 1.1391212000530997,
+      "grad_norm": 0.3954365491929587,
+      "learning_rate": 4.5838962989341445e-06,
+      "loss": 0.5658,
+      "step": 4291
+    },
+    {
+      "epoch": 1.1393866985264836,
+      "grad_norm": 0.3827228472520975,
+      "learning_rate": 4.583703409471234e-06,
+      "loss": 0.5762,
+      "step": 4292
+    },
+    {
+      "epoch": 1.1396521969998672,
+      "grad_norm": 0.3934912099010245,
+      "learning_rate": 4.583510479371004e-06,
+      "loss": 0.5857,
+      "step": 4293
+    },
+    {
+      "epoch": 1.139917695473251,
+      "grad_norm": 0.3983600647238615,
+      "learning_rate": 4.583317508637219e-06,
+      "loss": 0.6211,
+      "step": 4294
+    },
+    {
+      "epoch": 1.1401831939466347,
+      "grad_norm": 0.40078534572319097,
+      "learning_rate": 4.583124497273641e-06,
+      "loss": 0.6402,
+      "step": 4295
+    },
+    {
+      "epoch": 1.1404486924200186,
+      "grad_norm": 0.41170336961598936,
+      "learning_rate": 4.582931445284037e-06,
+      "loss": 0.6171,
+      "step": 4296
+    },
+    {
+      "epoch": 1.1407141908934024,
+      "grad_norm": 0.3994540264969489,
+      "learning_rate": 4.582738352672168e-06,
+      "loss": 0.5929,
+      "step": 4297
+    },
+    {
+      "epoch": 1.140979689366786,
+      "grad_norm": 0.40304779540207886,
+      "learning_rate": 4.582545219441802e-06,
+      "loss": 0.5906,
+      "step": 4298
+    },
+    {
+      "epoch": 1.14124518784017,
+      "grad_norm": 0.4118781115665021,
+      "learning_rate": 4.582352045596705e-06,
+      "loss": 0.5615,
+      "step": 4299
+    },
+    {
+      "epoch": 1.1415106863135538,
+      "grad_norm": 0.4073933136467304,
+      "learning_rate": 4.582158831140645e-06,
+      "loss": 0.5571,
+      "step": 4300
+    },
+    {
+      "epoch": 1.1417761847869374,
+      "grad_norm": 0.4191341167443241,
+      "learning_rate": 4.58196557607739e-06,
+      "loss": 0.5872,
+      "step": 4301
+    },
+    {
+      "epoch": 1.1420416832603213,
+      "grad_norm": 0.40363087802261094,
+      "learning_rate": 4.581772280410709e-06,
+      "loss": 0.5876,
+      "step": 4302
+    },
+    {
+      "epoch": 1.142307181733705,
+      "grad_norm": 0.40088202686285845,
+      "learning_rate": 4.581578944144371e-06,
+      "loss": 0.5992,
+      "step": 4303
+    },
+    {
+      "epoch": 1.1425726802070888,
+      "grad_norm": 0.39802560390802516,
+      "learning_rate": 4.581385567282147e-06,
+      "loss": 0.5569,
+      "step": 4304
+    },
+    {
+      "epoch": 1.1428381786804727,
+      "grad_norm": 0.4385536958866902,
+      "learning_rate": 4.581192149827808e-06,
+      "loss": 0.6067,
+      "step": 4305
+    },
+    {
+      "epoch": 1.1431036771538563,
+      "grad_norm": 0.39831625863198367,
+      "learning_rate": 4.580998691785127e-06,
+      "loss": 0.5754,
+      "step": 4306
+    },
+    {
+      "epoch": 1.1433691756272402,
+      "grad_norm": 0.40644696416310405,
+      "learning_rate": 4.580805193157876e-06,
+      "loss": 0.6019,
+      "step": 4307
+    },
+    {
+      "epoch": 1.1436346741006238,
+      "grad_norm": 0.4108739499394583,
+      "learning_rate": 4.580611653949829e-06,
+      "loss": 0.5787,
+      "step": 4308
+    },
+    {
+      "epoch": 1.1439001725740077,
+      "grad_norm": 0.41938891674170853,
+      "learning_rate": 4.580418074164761e-06,
+      "loss": 0.5959,
+      "step": 4309
+    },
+    {
+      "epoch": 1.1441656710473915,
+      "grad_norm": 0.4124085100535883,
+      "learning_rate": 4.580224453806447e-06,
+      "loss": 0.578,
+      "step": 4310
+    },
+    {
+      "epoch": 1.1444311695207752,
+      "grad_norm": 0.4181290215766447,
+      "learning_rate": 4.580030792878662e-06,
+      "loss": 0.6051,
+      "step": 4311
+    },
+    {
+      "epoch": 1.144696667994159,
+      "grad_norm": 0.39095242675070696,
+      "learning_rate": 4.5798370913851846e-06,
+      "loss": 0.5811,
+      "step": 4312
+    },
+    {
+      "epoch": 1.144962166467543,
+      "grad_norm": 0.4033025462178715,
+      "learning_rate": 4.579643349329791e-06,
+      "loss": 0.5924,
+      "step": 4313
+    },
+    {
+      "epoch": 1.1452276649409265,
+      "grad_norm": 0.39758849391784024,
+      "learning_rate": 4.5794495667162605e-06,
+      "loss": 0.617,
+      "step": 4314
+    },
+    {
+      "epoch": 1.1454931634143104,
+      "grad_norm": 0.40855575100791214,
+      "learning_rate": 4.579255743548372e-06,
+      "loss": 0.5881,
+      "step": 4315
+    },
+    {
+      "epoch": 1.1457586618876943,
+      "grad_norm": 0.4060475760244103,
+      "learning_rate": 4.579061879829906e-06,
+      "loss": 0.5855,
+      "step": 4316
+    },
+    {
+      "epoch": 1.146024160361078,
+      "grad_norm": 0.39626878897437523,
+      "learning_rate": 4.578867975564642e-06,
+      "loss": 0.6316,
+      "step": 4317
+    },
+    {
+      "epoch": 1.1462896588344618,
+      "grad_norm": 0.4009063840990518,
+      "learning_rate": 4.578674030756364e-06,
+      "loss": 0.5865,
+      "step": 4318
+    },
+    {
+      "epoch": 1.1465551573078454,
+      "grad_norm": 0.4184239460125252,
+      "learning_rate": 4.578480045408852e-06,
+      "loss": 0.5692,
+      "step": 4319
+    },
+    {
+      "epoch": 1.1468206557812293,
+      "grad_norm": 0.394082915045355,
+      "learning_rate": 4.578286019525889e-06,
+      "loss": 0.6147,
+      "step": 4320
+    },
+    {
+      "epoch": 1.1470861542546131,
+      "grad_norm": 0.4063715073134393,
+      "learning_rate": 4.578091953111262e-06,
+      "loss": 0.588,
+      "step": 4321
+    },
+    {
+      "epoch": 1.1473516527279968,
+      "grad_norm": 0.4088097849868308,
+      "learning_rate": 4.5778978461687525e-06,
+      "loss": 0.5907,
+      "step": 4322
+    },
+    {
+      "epoch": 1.1476171512013806,
+      "grad_norm": 0.39819479412364894,
+      "learning_rate": 4.577703698702149e-06,
+      "loss": 0.557,
+      "step": 4323
+    },
+    {
+      "epoch": 1.1478826496747643,
+      "grad_norm": 0.4131593160183521,
+      "learning_rate": 4.577509510715234e-06,
+      "loss": 0.6154,
+      "step": 4324
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 0.4197682165234953,
+      "learning_rate": 4.577315282211799e-06,
+      "loss": 0.577,
+      "step": 4325
+    },
+    {
+      "epoch": 1.148413646621532,
+      "grad_norm": 0.4071254271727446,
+      "learning_rate": 4.577121013195629e-06,
+      "loss": 0.6078,
+      "step": 4326
+    },
+    {
+      "epoch": 1.1486791450949156,
+      "grad_norm": 0.4383429135156656,
+      "learning_rate": 4.576926703670513e-06,
+      "loss": 0.596,
+      "step": 4327
+    },
+    {
+      "epoch": 1.1489446435682995,
+      "grad_norm": 0.394587245595195,
+      "learning_rate": 4.576732353640242e-06,
+      "loss": 0.6,
+      "step": 4328
+    },
+    {
+      "epoch": 1.1492101420416834,
+      "grad_norm": 0.4137329699443566,
+      "learning_rate": 4.576537963108606e-06,
+      "loss": 0.6188,
+      "step": 4329
+    },
+    {
+      "epoch": 1.149475640515067,
+      "grad_norm": 0.397942619021064,
+      "learning_rate": 4.576343532079395e-06,
+      "loss": 0.6067,
+      "step": 4330
+    },
+    {
+      "epoch": 1.1497411389884509,
+      "grad_norm": 0.3919396977565792,
+      "learning_rate": 4.5761490605564005e-06,
+      "loss": 0.6117,
+      "step": 4331
+    },
+    {
+      "epoch": 1.1500066374618345,
+      "grad_norm": 0.40139265544110847,
+      "learning_rate": 4.575954548543416e-06,
+      "loss": 0.5983,
+      "step": 4332
+    },
+    {
+      "epoch": 1.1502721359352184,
+      "grad_norm": 0.39496110889879826,
+      "learning_rate": 4.575759996044236e-06,
+      "loss": 0.6232,
+      "step": 4333
+    },
+    {
+      "epoch": 1.1505376344086022,
+      "grad_norm": 0.4107838187004183,
+      "learning_rate": 4.575565403062653e-06,
+      "loss": 0.6256,
+      "step": 4334
+    },
+    {
+      "epoch": 1.1508031328819859,
+      "grad_norm": 0.4092664147597602,
+      "learning_rate": 4.575370769602463e-06,
+      "loss": 0.5984,
+      "step": 4335
+    },
+    {
+      "epoch": 1.1510686313553697,
+      "grad_norm": 0.4101930266650931,
+      "learning_rate": 4.575176095667462e-06,
+      "loss": 0.5701,
+      "step": 4336
+    },
+    {
+      "epoch": 1.1513341298287534,
+      "grad_norm": 0.38430619673382416,
+      "learning_rate": 4.5749813812614455e-06,
+      "loss": 0.6357,
+      "step": 4337
+    },
+    {
+      "epoch": 1.1515996283021372,
+      "grad_norm": 0.40583016322142557,
+      "learning_rate": 4.574786626388212e-06,
+      "loss": 0.5767,
+      "step": 4338
+    },
+    {
+      "epoch": 1.151865126775521,
+      "grad_norm": 0.4114533009190886,
+      "learning_rate": 4.574591831051559e-06,
+      "loss": 0.5689,
+      "step": 4339
+    },
+    {
+      "epoch": 1.1521306252489047,
+      "grad_norm": 0.39756160439719124,
+      "learning_rate": 4.574396995255286e-06,
+      "loss": 0.609,
+      "step": 4340
+    },
+    {
+      "epoch": 1.1523961237222886,
+      "grad_norm": 0.4018294389530276,
+      "learning_rate": 4.574202119003192e-06,
+      "loss": 0.628,
+      "step": 4341
+    },
+    {
+      "epoch": 1.1526616221956725,
+      "grad_norm": 0.4148272578167303,
+      "learning_rate": 4.57400720229908e-06,
+      "loss": 0.6139,
+      "step": 4342
+    },
+    {
+      "epoch": 1.1529271206690561,
+      "grad_norm": 0.3966406167337366,
+      "learning_rate": 4.573812245146748e-06,
+      "loss": 0.582,
+      "step": 4343
+    },
+    {
+      "epoch": 1.15319261914244,
+      "grad_norm": 0.40722377067530713,
+      "learning_rate": 4.573617247550001e-06,
+      "loss": 0.6002,
+      "step": 4344
+    },
+    {
+      "epoch": 1.1534581176158236,
+      "grad_norm": 0.4016143861509407,
+      "learning_rate": 4.573422209512639e-06,
+      "loss": 0.627,
+      "step": 4345
+    },
+    {
+      "epoch": 1.1537236160892075,
+      "grad_norm": 0.3936139569586051,
+      "learning_rate": 4.573227131038468e-06,
+      "loss": 0.5365,
+      "step": 4346
+    },
+    {
+      "epoch": 1.1539891145625913,
+      "grad_norm": 0.38811120907461716,
+      "learning_rate": 4.5730320121312915e-06,
+      "loss": 0.6201,
+      "step": 4347
+    },
+    {
+      "epoch": 1.154254613035975,
+      "grad_norm": 0.42063094122751454,
+      "learning_rate": 4.572836852794915e-06,
+      "loss": 0.5914,
+      "step": 4348
+    },
+    {
+      "epoch": 1.1545201115093588,
+      "grad_norm": 0.41055565884154815,
+      "learning_rate": 4.5726416530331455e-06,
+      "loss": 0.6083,
+      "step": 4349
+    },
+    {
+      "epoch": 1.1547856099827425,
+      "grad_norm": 0.40469347811889916,
+      "learning_rate": 4.572446412849789e-06,
+      "loss": 0.5849,
+      "step": 4350
+    },
+    {
+      "epoch": 1.1550511084561264,
+      "grad_norm": 0.3961204287237384,
+      "learning_rate": 4.572251132248652e-06,
+      "loss": 0.5864,
+      "step": 4351
+    },
+    {
+      "epoch": 1.1553166069295102,
+      "grad_norm": 0.4056444506190797,
+      "learning_rate": 4.572055811233545e-06,
+      "loss": 0.587,
+      "step": 4352
+    },
+    {
+      "epoch": 1.1555821054028939,
+      "grad_norm": 0.40742181248702347,
+      "learning_rate": 4.571860449808276e-06,
+      "loss": 0.5758,
+      "step": 4353
+    },
+    {
+      "epoch": 1.1558476038762777,
+      "grad_norm": 0.4082464463596411,
+      "learning_rate": 4.571665047976657e-06,
+      "loss": 0.6179,
+      "step": 4354
+    },
+    {
+      "epoch": 1.1561131023496616,
+      "grad_norm": 0.42270839953522177,
+      "learning_rate": 4.5714696057424955e-06,
+      "loss": 0.6128,
+      "step": 4355
+    },
+    {
+      "epoch": 1.1563786008230452,
+      "grad_norm": 0.40302720730968244,
+      "learning_rate": 4.571274123109606e-06,
+      "loss": 0.5868,
+      "step": 4356
+    },
+    {
+      "epoch": 1.156644099296429,
+      "grad_norm": 0.39506001108151667,
+      "learning_rate": 4.5710786000817984e-06,
+      "loss": 0.5628,
+      "step": 4357
+    },
+    {
+      "epoch": 1.156909597769813,
+      "grad_norm": 0.41900861998223315,
+      "learning_rate": 4.5708830366628884e-06,
+      "loss": 0.6159,
+      "step": 4358
+    },
+    {
+      "epoch": 1.1571750962431966,
+      "grad_norm": 0.4082948086040363,
+      "learning_rate": 4.5706874328566886e-06,
+      "loss": 0.6009,
+      "step": 4359
+    },
+    {
+      "epoch": 1.1574405947165805,
+      "grad_norm": 0.4256093228088509,
+      "learning_rate": 4.570491788667014e-06,
+      "loss": 0.5685,
+      "step": 4360
+    },
+    {
+      "epoch": 1.157706093189964,
+      "grad_norm": 0.3947565997542352,
+      "learning_rate": 4.570296104097679e-06,
+      "loss": 0.5794,
+      "step": 4361
+    },
+    {
+      "epoch": 1.157971591663348,
+      "grad_norm": 0.4220567093035895,
+      "learning_rate": 4.570100379152503e-06,
+      "loss": 0.5844,
+      "step": 4362
+    },
+    {
+      "epoch": 1.1582370901367316,
+      "grad_norm": 0.3946991063016911,
+      "learning_rate": 4.5699046138353e-06,
+      "loss": 0.581,
+      "step": 4363
+    },
+    {
+      "epoch": 1.1585025886101155,
+      "grad_norm": 0.4013250793148427,
+      "learning_rate": 4.569708808149889e-06,
+      "loss": 0.6133,
+      "step": 4364
+    },
+    {
+      "epoch": 1.1587680870834993,
+      "grad_norm": 0.4111893879446417,
+      "learning_rate": 4.569512962100088e-06,
+      "loss": 0.6288,
+      "step": 4365
+    },
+    {
+      "epoch": 1.159033585556883,
+      "grad_norm": 0.3897881244853749,
+      "learning_rate": 4.569317075689718e-06,
+      "loss": 0.5526,
+      "step": 4366
+    },
+    {
+      "epoch": 1.1592990840302668,
+      "grad_norm": 0.39158440073797823,
+      "learning_rate": 4.5691211489225985e-06,
+      "loss": 0.5549,
+      "step": 4367
+    },
+    {
+      "epoch": 1.1595645825036507,
+      "grad_norm": 0.3878985636000898,
+      "learning_rate": 4.56892518180255e-06,
+      "loss": 0.6051,
+      "step": 4368
+    },
+    {
+      "epoch": 1.1598300809770343,
+      "grad_norm": 0.40838214922032773,
+      "learning_rate": 4.568729174333396e-06,
+      "loss": 0.5849,
+      "step": 4369
+    },
+    {
+      "epoch": 1.1600955794504182,
+      "grad_norm": 0.3877566228637658,
+      "learning_rate": 4.568533126518957e-06,
+      "loss": 0.6097,
+      "step": 4370
+    },
+    {
+      "epoch": 1.160361077923802,
+      "grad_norm": 0.42041684240148997,
+      "learning_rate": 4.568337038363058e-06,
+      "loss": 0.5981,
+      "step": 4371
+    },
+    {
+      "epoch": 1.1606265763971857,
+      "grad_norm": 0.40960859424338986,
+      "learning_rate": 4.568140909869522e-06,
+      "loss": 0.6155,
+      "step": 4372
+    },
+    {
+      "epoch": 1.1608920748705696,
+      "grad_norm": 0.3981388219702541,
+      "learning_rate": 4.567944741042175e-06,
+      "loss": 0.613,
+      "step": 4373
+    },
+    {
+      "epoch": 1.1611575733439532,
+      "grad_norm": 0.4081867519957751,
+      "learning_rate": 4.567748531884841e-06,
+      "loss": 0.5763,
+      "step": 4374
+    },
+    {
+      "epoch": 1.161423071817337,
+      "grad_norm": 0.41679817998867136,
+      "learning_rate": 4.567552282401349e-06,
+      "loss": 0.6126,
+      "step": 4375
+    },
+    {
+      "epoch": 1.161688570290721,
+      "grad_norm": 0.39021181631666174,
+      "learning_rate": 4.567355992595526e-06,
+      "loss": 0.572,
+      "step": 4376
+    },
+    {
+      "epoch": 1.1619540687641046,
+      "grad_norm": 0.4053955432990037,
+      "learning_rate": 4.567159662471199e-06,
+      "loss": 0.5794,
+      "step": 4377
+    },
+    {
+      "epoch": 1.1622195672374884,
+      "grad_norm": 0.4158627036409467,
+      "learning_rate": 4.5669632920321965e-06,
+      "loss": 0.5517,
+      "step": 4378
+    },
+    {
+      "epoch": 1.162485065710872,
+      "grad_norm": 0.4154756497949518,
+      "learning_rate": 4.56676688128235e-06,
+      "loss": 0.5759,
+      "step": 4379
+    },
+    {
+      "epoch": 1.162750564184256,
+      "grad_norm": 0.4153698012457725,
+      "learning_rate": 4.566570430225488e-06,
+      "loss": 0.6073,
+      "step": 4380
+    },
+    {
+      "epoch": 1.1630160626576398,
+      "grad_norm": 0.4118697437215938,
+      "learning_rate": 4.566373938865443e-06,
+      "loss": 0.5752,
+      "step": 4381
+    },
+    {
+      "epoch": 1.1632815611310234,
+      "grad_norm": 0.4039111215126782,
+      "learning_rate": 4.566177407206047e-06,
+      "loss": 0.6119,
+      "step": 4382
+    },
+    {
+      "epoch": 1.1635470596044073,
+      "grad_norm": 0.4248429675101417,
+      "learning_rate": 4.5659808352511334e-06,
+      "loss": 0.5743,
+      "step": 4383
+    },
+    {
+      "epoch": 1.1638125580777912,
+      "grad_norm": 0.39609242733633576,
+      "learning_rate": 4.565784223004535e-06,
+      "loss": 0.612,
+      "step": 4384
+    },
+    {
+      "epoch": 1.1640780565511748,
+      "grad_norm": 0.4099478144199655,
+      "learning_rate": 4.565587570470085e-06,
+      "loss": 0.6174,
+      "step": 4385
+    },
+    {
+      "epoch": 1.1643435550245587,
+      "grad_norm": 0.38693110294079974,
+      "learning_rate": 4.565390877651621e-06,
+      "loss": 0.5715,
+      "step": 4386
+    },
+    {
+      "epoch": 1.1646090534979423,
+      "grad_norm": 0.40842807643642776,
+      "learning_rate": 4.565194144552978e-06,
+      "loss": 0.5989,
+      "step": 4387
+    },
+    {
+      "epoch": 1.1648745519713262,
+      "grad_norm": 0.4048322965176205,
+      "learning_rate": 4.564997371177992e-06,
+      "loss": 0.5884,
+      "step": 4388
+    },
+    {
+      "epoch": 1.16514005044471,
+      "grad_norm": 0.4107886105100827,
+      "learning_rate": 4.564800557530502e-06,
+      "loss": 0.6111,
+      "step": 4389
+    },
+    {
+      "epoch": 1.1654055489180937,
+      "grad_norm": 0.41103771323376853,
+      "learning_rate": 4.564603703614345e-06,
+      "loss": 0.6045,
+      "step": 4390
+    },
+    {
+      "epoch": 1.1656710473914775,
+      "grad_norm": 0.3987996301871185,
+      "learning_rate": 4.564406809433362e-06,
+      "loss": 0.5892,
+      "step": 4391
+    },
+    {
+      "epoch": 1.1659365458648612,
+      "grad_norm": 0.41081796026287204,
+      "learning_rate": 4.56420987499139e-06,
+      "loss": 0.6071,
+      "step": 4392
+    },
+    {
+      "epoch": 1.166202044338245,
+      "grad_norm": 0.3985794969791896,
+      "learning_rate": 4.564012900292273e-06,
+      "loss": 0.5723,
+      "step": 4393
+    },
+    {
+      "epoch": 1.166467542811629,
+      "grad_norm": 0.4030640916780077,
+      "learning_rate": 4.563815885339849e-06,
+      "loss": 0.5664,
+      "step": 4394
+    },
+    {
+      "epoch": 1.1667330412850125,
+      "grad_norm": 0.43623304154587245,
+      "learning_rate": 4.563618830137964e-06,
+      "loss": 0.5597,
+      "step": 4395
+    },
+    {
+      "epoch": 1.1669985397583964,
+      "grad_norm": 0.39797416940200103,
+      "learning_rate": 4.563421734690458e-06,
+      "loss": 0.5577,
+      "step": 4396
+    },
+    {
+      "epoch": 1.1672640382317803,
+      "grad_norm": 0.4093438242962241,
+      "learning_rate": 4.563224599001176e-06,
+      "loss": 0.6026,
+      "step": 4397
+    },
+    {
+      "epoch": 1.167529536705164,
+      "grad_norm": 0.4280372928856646,
+      "learning_rate": 4.563027423073963e-06,
+      "loss": 0.6009,
+      "step": 4398
+    },
+    {
+      "epoch": 1.1677950351785478,
+      "grad_norm": 0.4262371076640645,
+      "learning_rate": 4.562830206912664e-06,
+      "loss": 0.572,
+      "step": 4399
+    },
+    {
+      "epoch": 1.1680605336519314,
+      "grad_norm": 0.41452503322674533,
+      "learning_rate": 4.562632950521126e-06,
+      "loss": 0.5813,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1683260321253153,
+      "grad_norm": 0.40887694032556754,
+      "learning_rate": 4.562435653903195e-06,
+      "loss": 0.5658,
+      "step": 4401
+    },
+    {
+      "epoch": 1.1685915305986991,
+      "grad_norm": 0.43168322887375643,
+      "learning_rate": 4.562238317062719e-06,
+      "loss": 0.5888,
+      "step": 4402
+    },
+    {
+      "epoch": 1.1688570290720828,
+      "grad_norm": 0.41546602270343747,
+      "learning_rate": 4.562040940003546e-06,
+      "loss": 0.5852,
+      "step": 4403
+    },
+    {
+      "epoch": 1.1691225275454666,
+      "grad_norm": 0.4033632677397616,
+      "learning_rate": 4.561843522729526e-06,
+      "loss": 0.5897,
+      "step": 4404
+    },
+    {
+      "epoch": 1.1693880260188503,
+      "grad_norm": 0.42295209094946473,
+      "learning_rate": 4.561646065244509e-06,
+      "loss": 0.6243,
+      "step": 4405
+    },
+    {
+      "epoch": 1.1696535244922341,
+      "grad_norm": 0.3915228058390058,
+      "learning_rate": 4.5614485675523465e-06,
+      "loss": 0.5652,
+      "step": 4406
+    },
+    {
+      "epoch": 1.169919022965618,
+      "grad_norm": 0.39424830298244506,
+      "learning_rate": 4.561251029656889e-06,
+      "loss": 0.6068,
+      "step": 4407
+    },
+    {
+      "epoch": 1.1701845214390016,
+      "grad_norm": 0.4023095949127277,
+      "learning_rate": 4.561053451561991e-06,
+      "loss": 0.6163,
+      "step": 4408
+    },
+    {
+      "epoch": 1.1704500199123855,
+      "grad_norm": 0.4074628522585257,
+      "learning_rate": 4.560855833271503e-06,
+      "loss": 0.551,
+      "step": 4409
+    },
+    {
+      "epoch": 1.1707155183857694,
+      "grad_norm": 0.40475474746346707,
+      "learning_rate": 4.560658174789281e-06,
+      "loss": 0.598,
+      "step": 4410
+    },
+    {
+      "epoch": 1.170981016859153,
+      "grad_norm": 0.4039940095318477,
+      "learning_rate": 4.560460476119178e-06,
+      "loss": 0.5967,
+      "step": 4411
+    },
+    {
+      "epoch": 1.1712465153325369,
+      "grad_norm": 0.432137768569034,
+      "learning_rate": 4.560262737265053e-06,
+      "loss": 0.5884,
+      "step": 4412
+    },
+    {
+      "epoch": 1.1715120138059207,
+      "grad_norm": 0.4048461335656455,
+      "learning_rate": 4.560064958230759e-06,
+      "loss": 0.6042,
+      "step": 4413
+    },
+    {
+      "epoch": 1.1717775122793044,
+      "grad_norm": 0.40193364323770364,
+      "learning_rate": 4.559867139020154e-06,
+      "loss": 0.6044,
+      "step": 4414
+    },
+    {
+      "epoch": 1.1720430107526882,
+      "grad_norm": 0.4177572101740509,
+      "learning_rate": 4.5596692796370965e-06,
+      "loss": 0.5479,
+      "step": 4415
+    },
+    {
+      "epoch": 1.1723085092260719,
+      "grad_norm": 0.4053928878582181,
+      "learning_rate": 4.559471380085446e-06,
+      "loss": 0.5931,
+      "step": 4416
+    },
+    {
+      "epoch": 1.1725740076994557,
+      "grad_norm": 0.39152537400138876,
+      "learning_rate": 4.559273440369061e-06,
+      "loss": 0.57,
+      "step": 4417
+    },
+    {
+      "epoch": 1.1728395061728394,
+      "grad_norm": 0.4060633572060415,
+      "learning_rate": 4.5590754604918005e-06,
+      "loss": 0.6121,
+      "step": 4418
+    },
+    {
+      "epoch": 1.1731050046462232,
+      "grad_norm": 0.40490156316901305,
+      "learning_rate": 4.558877440457529e-06,
+      "loss": 0.63,
+      "step": 4419
+    },
+    {
+      "epoch": 1.1733705031196071,
+      "grad_norm": 0.41274850894603254,
+      "learning_rate": 4.5586793802701055e-06,
+      "loss": 0.5992,
+      "step": 4420
+    },
+    {
+      "epoch": 1.1736360015929908,
+      "grad_norm": 0.3977290201611633,
+      "learning_rate": 4.558481279933393e-06,
+      "loss": 0.6018,
+      "step": 4421
+    },
+    {
+      "epoch": 1.1739015000663746,
+      "grad_norm": 0.40782755391328357,
+      "learning_rate": 4.558283139451256e-06,
+      "loss": 0.6141,
+      "step": 4422
+    },
+    {
+      "epoch": 1.1741669985397585,
+      "grad_norm": 0.4093455930594219,
+      "learning_rate": 4.558084958827559e-06,
+      "loss": 0.6222,
+      "step": 4423
+    },
+    {
+      "epoch": 1.1744324970131421,
+      "grad_norm": 0.4167049326855498,
+      "learning_rate": 4.557886738066165e-06,
+      "loss": 0.5963,
+      "step": 4424
+    },
+    {
+      "epoch": 1.174697995486526,
+      "grad_norm": 0.41377064322024276,
+      "learning_rate": 4.5576884771709414e-06,
+      "loss": 0.5889,
+      "step": 4425
+    },
+    {
+      "epoch": 1.1749634939599098,
+      "grad_norm": 0.40552841499279046,
+      "learning_rate": 4.557490176145755e-06,
+      "loss": 0.6068,
+      "step": 4426
+    },
+    {
+      "epoch": 1.1752289924332935,
+      "grad_norm": 0.405054856497394,
+      "learning_rate": 4.557291834994472e-06,
+      "loss": 0.6395,
+      "step": 4427
+    },
+    {
+      "epoch": 1.1754944909066773,
+      "grad_norm": 0.4122727360553481,
+      "learning_rate": 4.557093453720961e-06,
+      "loss": 0.6244,
+      "step": 4428
+    },
+    {
+      "epoch": 1.175759989380061,
+      "grad_norm": 0.39312919854500455,
+      "learning_rate": 4.556895032329092e-06,
+      "loss": 0.5762,
+      "step": 4429
+    },
+    {
+      "epoch": 1.1760254878534449,
+      "grad_norm": 0.4261987839984862,
+      "learning_rate": 4.556696570822733e-06,
+      "loss": 0.6217,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1762909863268287,
+      "grad_norm": 0.3961696856133243,
+      "learning_rate": 4.5564980692057545e-06,
+      "loss": 0.5837,
+      "step": 4431
+    },
+    {
+      "epoch": 1.1765564848002124,
+      "grad_norm": 0.39374528085339133,
+      "learning_rate": 4.556299527482029e-06,
+      "loss": 0.6204,
+      "step": 4432
+    },
+    {
+      "epoch": 1.1768219832735962,
+      "grad_norm": 0.3954590479198233,
+      "learning_rate": 4.556100945655428e-06,
+      "loss": 0.6104,
+      "step": 4433
+    },
+    {
+      "epoch": 1.1770874817469799,
+      "grad_norm": 0.3735118609040518,
+      "learning_rate": 4.555902323729825e-06,
+      "loss": 0.5994,
+      "step": 4434
+    },
+    {
+      "epoch": 1.1773529802203637,
+      "grad_norm": 0.4164002595732277,
+      "learning_rate": 4.555703661709092e-06,
+      "loss": 0.5678,
+      "step": 4435
+    },
+    {
+      "epoch": 1.1776184786937476,
+      "grad_norm": 0.4114315005319558,
+      "learning_rate": 4.5555049595971054e-06,
+      "loss": 0.5768,
+      "step": 4436
+    },
+    {
+      "epoch": 1.1778839771671312,
+      "grad_norm": 0.4010844841730783,
+      "learning_rate": 4.555306217397739e-06,
+      "loss": 0.6069,
+      "step": 4437
+    },
+    {
+      "epoch": 1.178149475640515,
+      "grad_norm": 0.4179267793227562,
+      "learning_rate": 4.5551074351148685e-06,
+      "loss": 0.6171,
+      "step": 4438
+    },
+    {
+      "epoch": 1.178414974113899,
+      "grad_norm": 0.4053431723984752,
+      "learning_rate": 4.5549086127523725e-06,
+      "loss": 0.6048,
+      "step": 4439
+    },
+    {
+      "epoch": 1.1786804725872826,
+      "grad_norm": 0.41097058314698387,
+      "learning_rate": 4.554709750314126e-06,
+      "loss": 0.5922,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1789459710606665,
+      "grad_norm": 0.4134809534029421,
+      "learning_rate": 4.554510847804009e-06,
+      "loss": 0.5874,
+      "step": 4441
+    },
+    {
+      "epoch": 1.17921146953405,
+      "grad_norm": 0.4474319315573684,
+      "learning_rate": 4.5543119052259e-06,
+      "loss": 0.6015,
+      "step": 4442
+    },
+    {
+      "epoch": 1.179476968007434,
+      "grad_norm": 0.3989674882312488,
+      "learning_rate": 4.554112922583679e-06,
+      "loss": 0.5816,
+      "step": 4443
+    },
+    {
+      "epoch": 1.1797424664808178,
+      "grad_norm": 0.4284139969314759,
+      "learning_rate": 4.553913899881227e-06,
+      "loss": 0.5982,
+      "step": 4444
+    },
+    {
+      "epoch": 1.1800079649542015,
+      "grad_norm": 0.410263729605568,
+      "learning_rate": 4.553714837122425e-06,
+      "loss": 0.6215,
+      "step": 4445
+    },
+    {
+      "epoch": 1.1802734634275853,
+      "grad_norm": 0.43039607111057654,
+      "learning_rate": 4.553515734311155e-06,
+      "loss": 0.5971,
+      "step": 4446
+    },
+    {
+      "epoch": 1.180538961900969,
+      "grad_norm": 0.3976128116246067,
+      "learning_rate": 4.5533165914513015e-06,
+      "loss": 0.6068,
+      "step": 4447
+    },
+    {
+      "epoch": 1.1808044603743528,
+      "grad_norm": 0.42687000644298945,
+      "learning_rate": 4.5531174085467455e-06,
+      "loss": 0.5925,
+      "step": 4448
+    },
+    {
+      "epoch": 1.1810699588477367,
+      "grad_norm": 0.43996845001643287,
+      "learning_rate": 4.552918185601374e-06,
+      "loss": 0.6569,
+      "step": 4449
+    },
+    {
+      "epoch": 1.1813354573211203,
+      "grad_norm": 0.4012139379490353,
+      "learning_rate": 4.552718922619071e-06,
+      "loss": 0.6206,
+      "step": 4450
+    },
+    {
+      "epoch": 1.1816009557945042,
+      "grad_norm": 0.40865989646963835,
+      "learning_rate": 4.552519619603724e-06,
+      "loss": 0.6334,
+      "step": 4451
+    },
+    {
+      "epoch": 1.181866454267888,
+      "grad_norm": 0.3915229794382577,
+      "learning_rate": 4.552320276559218e-06,
+      "loss": 0.583,
+      "step": 4452
+    },
+    {
+      "epoch": 1.1821319527412717,
+      "grad_norm": 0.408867311475484,
+      "learning_rate": 4.552120893489441e-06,
+      "loss": 0.5875,
+      "step": 4453
+    },
+    {
+      "epoch": 1.1823974512146556,
+      "grad_norm": 0.4041688088258742,
+      "learning_rate": 4.551921470398284e-06,
+      "loss": 0.6328,
+      "step": 4454
+    },
+    {
+      "epoch": 1.1826629496880392,
+      "grad_norm": 0.3907784636482234,
+      "learning_rate": 4.5517220072896335e-06,
+      "loss": 0.5628,
+      "step": 4455
+    },
+    {
+      "epoch": 1.182928448161423,
+      "grad_norm": 0.4215109317205618,
+      "learning_rate": 4.55152250416738e-06,
+      "loss": 0.5901,
+      "step": 4456
+    },
+    {
+      "epoch": 1.183193946634807,
+      "grad_norm": 0.44120316888269473,
+      "learning_rate": 4.551322961035415e-06,
+      "loss": 0.5816,
+      "step": 4457
+    },
+    {
+      "epoch": 1.1834594451081906,
+      "grad_norm": 0.40244274378616046,
+      "learning_rate": 4.55112337789763e-06,
+      "loss": 0.5775,
+      "step": 4458
+    },
+    {
+      "epoch": 1.1837249435815744,
+      "grad_norm": 0.41561805639551425,
+      "learning_rate": 4.550923754757917e-06,
+      "loss": 0.6064,
+      "step": 4459
+    },
+    {
+      "epoch": 1.183990442054958,
+      "grad_norm": 0.44607153269247085,
+      "learning_rate": 4.550724091620169e-06,
+      "loss": 0.5926,
+      "step": 4460
+    },
+    {
+      "epoch": 1.184255940528342,
+      "grad_norm": 0.4001639537003755,
+      "learning_rate": 4.5505243884882804e-06,
+      "loss": 0.6111,
+      "step": 4461
+    },
+    {
+      "epoch": 1.1845214390017258,
+      "grad_norm": 0.41067182389608814,
+      "learning_rate": 4.550324645366145e-06,
+      "loss": 0.5833,
+      "step": 4462
+    },
+    {
+      "epoch": 1.1847869374751094,
+      "grad_norm": 0.4054665697570779,
+      "learning_rate": 4.55012486225766e-06,
+      "loss": 0.6399,
+      "step": 4463
+    },
+    {
+      "epoch": 1.1850524359484933,
+      "grad_norm": 0.41083439509682923,
+      "learning_rate": 4.549925039166721e-06,
+      "loss": 0.57,
+      "step": 4464
+    },
+    {
+      "epoch": 1.1853179344218772,
+      "grad_norm": 0.43795726955714775,
+      "learning_rate": 4.549725176097223e-06,
+      "loss": 0.5816,
+      "step": 4465
+    },
+    {
+      "epoch": 1.1855834328952608,
+      "grad_norm": 0.3786445238361183,
+      "learning_rate": 4.549525273053067e-06,
+      "loss": 0.5946,
+      "step": 4466
+    },
+    {
+      "epoch": 1.1858489313686447,
+      "grad_norm": 0.4202558449348351,
+      "learning_rate": 4.549325330038149e-06,
+      "loss": 0.6061,
+      "step": 4467
+    },
+    {
+      "epoch": 1.1861144298420285,
+      "grad_norm": 0.39885877883740245,
+      "learning_rate": 4.54912534705637e-06,
+      "loss": 0.6382,
+      "step": 4468
+    },
+    {
+      "epoch": 1.1863799283154122,
+      "grad_norm": 0.41303146492407405,
+      "learning_rate": 4.548925324111629e-06,
+      "loss": 0.6269,
+      "step": 4469
+    },
+    {
+      "epoch": 1.186645426788796,
+      "grad_norm": 0.4076530912489785,
+      "learning_rate": 4.548725261207828e-06,
+      "loss": 0.6292,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1869109252621797,
+      "grad_norm": 0.4171350147421362,
+      "learning_rate": 4.548525158348869e-06,
+      "loss": 0.6066,
+      "step": 4471
+    },
+    {
+      "epoch": 1.1871764237355635,
+      "grad_norm": 0.40412164642943715,
+      "learning_rate": 4.548325015538653e-06,
+      "loss": 0.6467,
+      "step": 4472
+    },
+    {
+      "epoch": 1.1874419222089472,
+      "grad_norm": 0.3919507147783421,
+      "learning_rate": 4.548124832781084e-06,
+      "loss": 0.6009,
+      "step": 4473
+    },
+    {
+      "epoch": 1.187707420682331,
+      "grad_norm": 0.3997728557657679,
+      "learning_rate": 4.547924610080065e-06,
+      "loss": 0.5912,
+      "step": 4474
+    },
+    {
+      "epoch": 1.187972919155715,
+      "grad_norm": 0.40546725997533944,
+      "learning_rate": 4.547724347439504e-06,
+      "loss": 0.5858,
+      "step": 4475
+    },
+    {
+      "epoch": 1.1882384176290985,
+      "grad_norm": 0.37537981933255915,
+      "learning_rate": 4.547524044863303e-06,
+      "loss": 0.5615,
+      "step": 4476
+    },
+    {
+      "epoch": 1.1885039161024824,
+      "grad_norm": 0.3949913308540101,
+      "learning_rate": 4.54732370235537e-06,
+      "loss": 0.5611,
+      "step": 4477
+    },
+    {
+      "epoch": 1.1887694145758663,
+      "grad_norm": 0.42383223108378926,
+      "learning_rate": 4.547123319919613e-06,
+      "loss": 0.5784,
+      "step": 4478
+    },
+    {
+      "epoch": 1.18903491304925,
+      "grad_norm": 0.40104666067330436,
+      "learning_rate": 4.546922897559939e-06,
+      "loss": 0.6221,
+      "step": 4479
+    },
+    {
+      "epoch": 1.1893004115226338,
+      "grad_norm": 0.41261503353019885,
+      "learning_rate": 4.546722435280256e-06,
+      "loss": 0.604,
+      "step": 4480
+    },
+    {
+      "epoch": 1.1895659099960176,
+      "grad_norm": 0.4081502130888106,
+      "learning_rate": 4.546521933084475e-06,
+      "loss": 0.5861,
+      "step": 4481
+    },
+    {
+      "epoch": 1.1898314084694013,
+      "grad_norm": 0.400822806896701,
+      "learning_rate": 4.546321390976505e-06,
+      "loss": 0.5936,
+      "step": 4482
+    },
+    {
+      "epoch": 1.1900969069427851,
+      "grad_norm": 0.42768810762957216,
+      "learning_rate": 4.546120808960258e-06,
+      "loss": 0.5998,
+      "step": 4483
+    },
+    {
+      "epoch": 1.1903624054161688,
+      "grad_norm": 0.4011064979656986,
+      "learning_rate": 4.545920187039645e-06,
+      "loss": 0.6071,
+      "step": 4484
+    },
+    {
+      "epoch": 1.1906279038895526,
+      "grad_norm": 0.40875027623805554,
+      "learning_rate": 4.54571952521858e-06,
+      "loss": 0.6205,
+      "step": 4485
+    },
+    {
+      "epoch": 1.1908934023629365,
+      "grad_norm": 0.41132805619660584,
+      "learning_rate": 4.545518823500974e-06,
+      "loss": 0.5892,
+      "step": 4486
+    },
+    {
+      "epoch": 1.1911589008363201,
+      "grad_norm": 0.40585142412641545,
+      "learning_rate": 4.545318081890744e-06,
+      "loss": 0.5939,
+      "step": 4487
+    },
+    {
+      "epoch": 1.191424399309704,
+      "grad_norm": 0.407943189320469,
+      "learning_rate": 4.545117300391803e-06,
+      "loss": 0.6116,
+      "step": 4488
+    },
+    {
+      "epoch": 1.1916898977830876,
+      "grad_norm": 0.4318494921232628,
+      "learning_rate": 4.544916479008068e-06,
+      "loss": 0.6027,
+      "step": 4489
+    },
+    {
+      "epoch": 1.1919553962564715,
+      "grad_norm": 0.3828940604674715,
+      "learning_rate": 4.544715617743454e-06,
+      "loss": 0.6242,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1922208947298554,
+      "grad_norm": 0.41038753962187513,
+      "learning_rate": 4.544514716601879e-06,
+      "loss": 0.5905,
+      "step": 4491
+    },
+    {
+      "epoch": 1.192486393203239,
+      "grad_norm": 0.41139164763241187,
+      "learning_rate": 4.544313775587262e-06,
+      "loss": 0.6135,
+      "step": 4492
+    },
+    {
+      "epoch": 1.1927518916766229,
+      "grad_norm": 0.4235348602624653,
+      "learning_rate": 4.54411279470352e-06,
+      "loss": 0.6007,
+      "step": 4493
+    },
+    {
+      "epoch": 1.1930173901500067,
+      "grad_norm": 0.40420641580806965,
+      "learning_rate": 4.543911773954575e-06,
+      "loss": 0.5693,
+      "step": 4494
+    },
+    {
+      "epoch": 1.1932828886233904,
+      "grad_norm": 0.3939052860726097,
+      "learning_rate": 4.543710713344345e-06,
+      "loss": 0.5322,
+      "step": 4495
+    },
+    {
+      "epoch": 1.1935483870967742,
+      "grad_norm": 0.39059419670150747,
+      "learning_rate": 4.543509612876752e-06,
+      "loss": 0.5753,
+      "step": 4496
+    },
+    {
+      "epoch": 1.1938138855701579,
+      "grad_norm": 0.41546161555275146,
+      "learning_rate": 4.543308472555719e-06,
+      "loss": 0.5806,
+      "step": 4497
+    },
+    {
+      "epoch": 1.1940793840435417,
+      "grad_norm": 0.4334850931043781,
+      "learning_rate": 4.543107292385168e-06,
+      "loss": 0.5885,
+      "step": 4498
+    },
+    {
+      "epoch": 1.1943448825169256,
+      "grad_norm": 0.4089692095857566,
+      "learning_rate": 4.542906072369022e-06,
+      "loss": 0.568,
+      "step": 4499
+    },
+    {
+      "epoch": 1.1946103809903093,
+      "grad_norm": 0.4063070216243931,
+      "learning_rate": 4.542704812511205e-06,
+      "loss": 0.6113,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1948758794636931,
+      "grad_norm": 0.4221650231976977,
+      "learning_rate": 4.542503512815644e-06,
+      "loss": 0.6173,
+      "step": 4501
+    },
+    {
+      "epoch": 1.1951413779370768,
+      "grad_norm": 0.40758090839154787,
+      "learning_rate": 4.542302173286263e-06,
+      "loss": 0.576,
+      "step": 4502
+    },
+    {
+      "epoch": 1.1954068764104606,
+      "grad_norm": 0.41290803497968254,
+      "learning_rate": 4.542100793926989e-06,
+      "loss": 0.5814,
+      "step": 4503
+    },
+    {
+      "epoch": 1.1956723748838445,
+      "grad_norm": 0.4208714973677631,
+      "learning_rate": 4.54189937474175e-06,
+      "loss": 0.557,
+      "step": 4504
+    },
+    {
+      "epoch": 1.1959378733572281,
+      "grad_norm": 0.5011782559685238,
+      "learning_rate": 4.541697915734474e-06,
+      "loss": 0.56,
+      "step": 4505
+    },
+    {
+      "epoch": 1.196203371830612,
+      "grad_norm": 0.4085847322666289,
+      "learning_rate": 4.541496416909089e-06,
+      "loss": 0.5869,
+      "step": 4506
+    },
+    {
+      "epoch": 1.1964688703039958,
+      "grad_norm": 0.3958313087239738,
+      "learning_rate": 4.541294878269526e-06,
+      "loss": 0.5786,
+      "step": 4507
+    },
+    {
+      "epoch": 1.1967343687773795,
+      "grad_norm": 0.39454111088478866,
+      "learning_rate": 4.541093299819714e-06,
+      "loss": 0.6109,
+      "step": 4508
+    },
+    {
+      "epoch": 1.1969998672507633,
+      "grad_norm": 0.405234064488423,
+      "learning_rate": 4.540891681563586e-06,
+      "loss": 0.6118,
+      "step": 4509
+    },
+    {
+      "epoch": 1.197265365724147,
+      "grad_norm": 0.41491462236196663,
+      "learning_rate": 4.540690023505073e-06,
+      "loss": 0.5486,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1975308641975309,
+      "grad_norm": 0.3855837997128597,
+      "learning_rate": 4.540488325648109e-06,
+      "loss": 0.5963,
+      "step": 4511
+    },
+    {
+      "epoch": 1.1977963626709147,
+      "grad_norm": 0.41037098836540603,
+      "learning_rate": 4.540286587996626e-06,
+      "loss": 0.5522,
+      "step": 4512
+    },
+    {
+      "epoch": 1.1980618611442984,
+      "grad_norm": 0.42924493232429467,
+      "learning_rate": 4.540084810554559e-06,
+      "loss": 0.6041,
+      "step": 4513
+    },
+    {
+      "epoch": 1.1983273596176822,
+      "grad_norm": 0.40355437810198835,
+      "learning_rate": 4.5398829933258425e-06,
+      "loss": 0.5728,
+      "step": 4514
+    },
+    {
+      "epoch": 1.1985928580910659,
+      "grad_norm": 0.3928276784461158,
+      "learning_rate": 4.539681136314414e-06,
+      "loss": 0.6263,
+      "step": 4515
+    },
+    {
+      "epoch": 1.1988583565644497,
+      "grad_norm": 0.4164995084427747,
+      "learning_rate": 4.539479239524209e-06,
+      "loss": 0.565,
+      "step": 4516
+    },
+    {
+      "epoch": 1.1991238550378336,
+      "grad_norm": 0.4282678988989548,
+      "learning_rate": 4.539277302959166e-06,
+      "loss": 0.5586,
+      "step": 4517
+    },
+    {
+      "epoch": 1.1993893535112172,
+      "grad_norm": 0.4091829418395631,
+      "learning_rate": 4.539075326623221e-06,
+      "loss": 0.5984,
+      "step": 4518
+    },
+    {
+      "epoch": 1.199654851984601,
+      "grad_norm": 0.44798887697912254,
+      "learning_rate": 4.538873310520316e-06,
+      "loss": 0.5976,
+      "step": 4519
+    },
+    {
+      "epoch": 1.199920350457985,
+      "grad_norm": 0.4235865839724445,
+      "learning_rate": 4.538671254654388e-06,
+      "loss": 0.5843,
+      "step": 4520
+    },
+    {
+      "epoch": 1.2001858489313686,
+      "grad_norm": 0.3960958371071703,
+      "learning_rate": 4.538469159029381e-06,
+      "loss": 0.5857,
+      "step": 4521
+    },
+    {
+      "epoch": 1.2004513474047525,
+      "grad_norm": 0.4145149146278742,
+      "learning_rate": 4.5382670236492335e-06,
+      "loss": 0.6093,
+      "step": 4522
+    },
+    {
+      "epoch": 1.2007168458781363,
+      "grad_norm": 0.40765250268742015,
+      "learning_rate": 4.538064848517888e-06,
+      "loss": 0.63,
+      "step": 4523
+    },
+    {
+      "epoch": 1.20098234435152,
+      "grad_norm": 0.4143048055970985,
+      "learning_rate": 4.5378626336392885e-06,
+      "loss": 0.5871,
+      "step": 4524
+    },
+    {
+      "epoch": 1.2012478428249038,
+      "grad_norm": 0.4454744517716582,
+      "learning_rate": 4.537660379017378e-06,
+      "loss": 0.5848,
+      "step": 4525
+    },
+    {
+      "epoch": 1.2015133412982875,
+      "grad_norm": 0.4074511184369045,
+      "learning_rate": 4.537458084656101e-06,
+      "loss": 0.5807,
+      "step": 4526
+    },
+    {
+      "epoch": 1.2017788397716713,
+      "grad_norm": 0.414583462594911,
+      "learning_rate": 4.537255750559403e-06,
+      "loss": 0.6341,
+      "step": 4527
+    },
+    {
+      "epoch": 1.202044338245055,
+      "grad_norm": 0.42959940060825563,
+      "learning_rate": 4.5370533767312296e-06,
+      "loss": 0.5525,
+      "step": 4528
+    },
+    {
+      "epoch": 1.2023098367184388,
+      "grad_norm": 0.45575786058567525,
+      "learning_rate": 4.536850963175528e-06,
+      "loss": 0.6377,
+      "step": 4529
+    },
+    {
+      "epoch": 1.2025753351918227,
+      "grad_norm": 0.4402112900444621,
+      "learning_rate": 4.536648509896245e-06,
+      "loss": 0.5758,
+      "step": 4530
+    },
+    {
+      "epoch": 1.2028408336652063,
+      "grad_norm": 0.38074566253122516,
+      "learning_rate": 4.536446016897329e-06,
+      "loss": 0.5889,
+      "step": 4531
+    },
+    {
+      "epoch": 1.2031063321385902,
+      "grad_norm": 0.4401624468798666,
+      "learning_rate": 4.536243484182731e-06,
+      "loss": 0.6074,
+      "step": 4532
+    },
+    {
+      "epoch": 1.203371830611974,
+      "grad_norm": 0.43027291202695106,
+      "learning_rate": 4.536040911756399e-06,
+      "loss": 0.6053,
+      "step": 4533
+    },
+    {
+      "epoch": 1.2036373290853577,
+      "grad_norm": 0.42253011725451534,
+      "learning_rate": 4.535838299622284e-06,
+      "loss": 0.5569,
+      "step": 4534
+    },
+    {
+      "epoch": 1.2039028275587416,
+      "grad_norm": 0.3907656598436608,
+      "learning_rate": 4.535635647784338e-06,
+      "loss": 0.5928,
+      "step": 4535
+    },
+    {
+      "epoch": 1.2041683260321254,
+      "grad_norm": 0.4112153085011214,
+      "learning_rate": 4.5354329562465115e-06,
+      "loss": 0.5911,
+      "step": 4536
+    },
+    {
+      "epoch": 1.204433824505509,
+      "grad_norm": 0.4040585164261546,
+      "learning_rate": 4.53523022501276e-06,
+      "loss": 0.5748,
+      "step": 4537
+    },
+    {
+      "epoch": 1.204699322978893,
+      "grad_norm": 0.40644159671718444,
+      "learning_rate": 4.535027454087036e-06,
+      "loss": 0.6354,
+      "step": 4538
+    },
+    {
+      "epoch": 1.2049648214522766,
+      "grad_norm": 0.4110424534487759,
+      "learning_rate": 4.534824643473293e-06,
+      "loss": 0.5583,
+      "step": 4539
+    },
+    {
+      "epoch": 1.2052303199256604,
+      "grad_norm": 0.40955423635184196,
+      "learning_rate": 4.534621793175488e-06,
+      "loss": 0.5952,
+      "step": 4540
+    },
+    {
+      "epoch": 1.2054958183990443,
+      "grad_norm": 0.4289755889051613,
+      "learning_rate": 4.534418903197576e-06,
+      "loss": 0.5712,
+      "step": 4541
+    },
+    {
+      "epoch": 1.205761316872428,
+      "grad_norm": 0.4120154199246293,
+      "learning_rate": 4.534215973543515e-06,
+      "loss": 0.6286,
+      "step": 4542
+    },
+    {
+      "epoch": 1.2060268153458118,
+      "grad_norm": 0.41588980666684294,
+      "learning_rate": 4.534013004217262e-06,
+      "loss": 0.6193,
+      "step": 4543
+    },
+    {
+      "epoch": 1.2062923138191954,
+      "grad_norm": 0.39906101138682426,
+      "learning_rate": 4.533809995222774e-06,
+      "loss": 0.6124,
+      "step": 4544
+    },
+    {
+      "epoch": 1.2065578122925793,
+      "grad_norm": 0.4145166725200401,
+      "learning_rate": 4.533606946564013e-06,
+      "loss": 0.5961,
+      "step": 4545
+    },
+    {
+      "epoch": 1.2068233107659632,
+      "grad_norm": 0.42495458727778007,
+      "learning_rate": 4.5334038582449356e-06,
+      "loss": 0.6129,
+      "step": 4546
+    },
+    {
+      "epoch": 1.2070888092393468,
+      "grad_norm": 0.41361553015640895,
+      "learning_rate": 4.5332007302695055e-06,
+      "loss": 0.5928,
+      "step": 4547
+    },
+    {
+      "epoch": 1.2073543077127307,
+      "grad_norm": 0.4029178366872803,
+      "learning_rate": 4.532997562641683e-06,
+      "loss": 0.5361,
+      "step": 4548
+    },
+    {
+      "epoch": 1.2076198061861145,
+      "grad_norm": 0.3996784170604655,
+      "learning_rate": 4.532794355365431e-06,
+      "loss": 0.5916,
+      "step": 4549
+    },
+    {
+      "epoch": 1.2078853046594982,
+      "grad_norm": 0.4011109933262168,
+      "learning_rate": 4.532591108444712e-06,
+      "loss": 0.621,
+      "step": 4550
+    },
+    {
+      "epoch": 1.208150803132882,
+      "grad_norm": 0.3945229566377231,
+      "learning_rate": 4.532387821883488e-06,
+      "loss": 0.5801,
+      "step": 4551
+    },
+    {
+      "epoch": 1.2084163016062657,
+      "grad_norm": 0.398654024028116,
+      "learning_rate": 4.532184495685726e-06,
+      "loss": 0.5832,
+      "step": 4552
+    },
+    {
+      "epoch": 1.2086818000796495,
+      "grad_norm": 0.4157788965184638,
+      "learning_rate": 4.531981129855392e-06,
+      "loss": 0.6173,
+      "step": 4553
+    },
+    {
+      "epoch": 1.2089472985530334,
+      "grad_norm": 0.4051148222248255,
+      "learning_rate": 4.53177772439645e-06,
+      "loss": 0.6115,
+      "step": 4554
+    },
+    {
+      "epoch": 1.209212797026417,
+      "grad_norm": 0.41810120374116055,
+      "learning_rate": 4.531574279312867e-06,
+      "loss": 0.6308,
+      "step": 4555
+    },
+    {
+      "epoch": 1.209478295499801,
+      "grad_norm": 0.40198744311657575,
+      "learning_rate": 4.531370794608612e-06,
+      "loss": 0.586,
+      "step": 4556
+    },
+    {
+      "epoch": 1.2097437939731845,
+      "grad_norm": 0.39062583238353016,
+      "learning_rate": 4.531167270287653e-06,
+      "loss": 0.5855,
+      "step": 4557
+    },
+    {
+      "epoch": 1.2100092924465684,
+      "grad_norm": 0.40915740354957764,
+      "learning_rate": 4.530963706353959e-06,
+      "loss": 0.5712,
+      "step": 4558
+    },
+    {
+      "epoch": 1.2102747909199523,
+      "grad_norm": 0.40037791420706736,
+      "learning_rate": 4.5307601028115e-06,
+      "loss": 0.6025,
+      "step": 4559
+    },
+    {
+      "epoch": 1.210540289393336,
+      "grad_norm": 0.39933653746219744,
+      "learning_rate": 4.5305564596642475e-06,
+      "loss": 0.5988,
+      "step": 4560
+    },
+    {
+      "epoch": 1.2108057878667198,
+      "grad_norm": 0.400864297190555,
+      "learning_rate": 4.530352776916172e-06,
+      "loss": 0.586,
+      "step": 4561
+    },
+    {
+      "epoch": 1.2110712863401036,
+      "grad_norm": 0.41490692534316537,
+      "learning_rate": 4.530149054571245e-06,
+      "loss": 0.6253,
+      "step": 4562
+    },
+    {
+      "epoch": 1.2113367848134873,
+      "grad_norm": 0.40539192908506905,
+      "learning_rate": 4.529945292633442e-06,
+      "loss": 0.5514,
+      "step": 4563
+    },
+    {
+      "epoch": 1.2116022832868711,
+      "grad_norm": 0.389057409179555,
+      "learning_rate": 4.529741491106735e-06,
+      "loss": 0.5641,
+      "step": 4564
+    },
+    {
+      "epoch": 1.211867781760255,
+      "grad_norm": 0.39969140548337406,
+      "learning_rate": 4.5295376499950995e-06,
+      "loss": 0.58,
+      "step": 4565
+    },
+    {
+      "epoch": 1.2121332802336386,
+      "grad_norm": 0.41620549492660186,
+      "learning_rate": 4.529333769302511e-06,
+      "loss": 0.5511,
+      "step": 4566
+    },
+    {
+      "epoch": 1.2123987787070225,
+      "grad_norm": 0.40172141795373945,
+      "learning_rate": 4.529129849032945e-06,
+      "loss": 0.6071,
+      "step": 4567
+    },
+    {
+      "epoch": 1.2126642771804061,
+      "grad_norm": 0.41048727528800466,
+      "learning_rate": 4.5289258891903784e-06,
+      "loss": 0.5761,
+      "step": 4568
+    },
+    {
+      "epoch": 1.21292977565379,
+      "grad_norm": 0.39947229916316845,
+      "learning_rate": 4.52872188977879e-06,
+      "loss": 0.6092,
+      "step": 4569
+    },
+    {
+      "epoch": 1.2131952741271737,
+      "grad_norm": 0.3992775681904612,
+      "learning_rate": 4.528517850802158e-06,
+      "loss": 0.5893,
+      "step": 4570
+    },
+    {
+      "epoch": 1.2134607726005575,
+      "grad_norm": 0.4248472191629426,
+      "learning_rate": 4.52831377226446e-06,
+      "loss": 0.5768,
+      "step": 4571
+    },
+    {
+      "epoch": 1.2137262710739414,
+      "grad_norm": 0.41341739925748106,
+      "learning_rate": 4.528109654169678e-06,
+      "loss": 0.5866,
+      "step": 4572
+    },
+    {
+      "epoch": 1.213991769547325,
+      "grad_norm": 0.40606116812831267,
+      "learning_rate": 4.527905496521792e-06,
+      "loss": 0.5787,
+      "step": 4573
+    },
+    {
+      "epoch": 1.2142572680207089,
+      "grad_norm": 0.39154015159768674,
+      "learning_rate": 4.527701299324783e-06,
+      "loss": 0.6088,
+      "step": 4574
+    },
+    {
+      "epoch": 1.2145227664940927,
+      "grad_norm": 0.3914008201477086,
+      "learning_rate": 4.527497062582635e-06,
+      "loss": 0.5969,
+      "step": 4575
+    },
+    {
+      "epoch": 1.2147882649674764,
+      "grad_norm": 0.40571045446581433,
+      "learning_rate": 4.52729278629933e-06,
+      "loss": 0.5737,
+      "step": 4576
+    },
+    {
+      "epoch": 1.2150537634408602,
+      "grad_norm": 0.4111999284430291,
+      "learning_rate": 4.527088470478851e-06,
+      "loss": 0.6127,
+      "step": 4577
+    },
+    {
+      "epoch": 1.215319261914244,
+      "grad_norm": 0.40429646665507607,
+      "learning_rate": 4.526884115125184e-06,
+      "loss": 0.6201,
+      "step": 4578
+    },
+    {
+      "epoch": 1.2155847603876277,
+      "grad_norm": 0.4006407886039044,
+      "learning_rate": 4.526679720242315e-06,
+      "loss": 0.607,
+      "step": 4579
+    },
+    {
+      "epoch": 1.2158502588610116,
+      "grad_norm": 0.39365647895764416,
+      "learning_rate": 4.526475285834229e-06,
+      "loss": 0.5369,
+      "step": 4580
+    },
+    {
+      "epoch": 1.2161157573343953,
+      "grad_norm": 0.4047812942037202,
+      "learning_rate": 4.526270811904913e-06,
+      "loss": 0.5636,
+      "step": 4581
+    },
+    {
+      "epoch": 1.2163812558077791,
+      "grad_norm": 0.4108136995666403,
+      "learning_rate": 4.526066298458355e-06,
+      "loss": 0.5828,
+      "step": 4582
+    },
+    {
+      "epoch": 1.216646754281163,
+      "grad_norm": 0.4128774609606118,
+      "learning_rate": 4.525861745498543e-06,
+      "loss": 0.5875,
+      "step": 4583
+    },
+    {
+      "epoch": 1.2169122527545466,
+      "grad_norm": 0.4066142453827169,
+      "learning_rate": 4.525657153029467e-06,
+      "loss": 0.608,
+      "step": 4584
+    },
+    {
+      "epoch": 1.2171777512279305,
+      "grad_norm": 0.4198649312306408,
+      "learning_rate": 4.525452521055117e-06,
+      "loss": 0.5892,
+      "step": 4585
+    },
+    {
+      "epoch": 1.2174432497013141,
+      "grad_norm": 0.41533679139931573,
+      "learning_rate": 4.525247849579484e-06,
+      "loss": 0.6364,
+      "step": 4586
+    },
+    {
+      "epoch": 1.217708748174698,
+      "grad_norm": 0.43139295959642077,
+      "learning_rate": 4.525043138606559e-06,
+      "loss": 0.5895,
+      "step": 4587
+    },
+    {
+      "epoch": 1.2179742466480818,
+      "grad_norm": 0.4056367209519541,
+      "learning_rate": 4.524838388140335e-06,
+      "loss": 0.6038,
+      "step": 4588
+    },
+    {
+      "epoch": 1.2182397451214655,
+      "grad_norm": 0.415044173530926,
+      "learning_rate": 4.524633598184805e-06,
+      "loss": 0.5698,
+      "step": 4589
+    },
+    {
+      "epoch": 1.2185052435948494,
+      "grad_norm": 0.41697069069839054,
+      "learning_rate": 4.524428768743961e-06,
+      "loss": 0.614,
+      "step": 4590
+    },
+    {
+      "epoch": 1.2187707420682332,
+      "grad_norm": 0.4115570086337441,
+      "learning_rate": 4.524223899821802e-06,
+      "loss": 0.6154,
+      "step": 4591
+    },
+    {
+      "epoch": 1.2190362405416169,
+      "grad_norm": 0.4166193044486813,
+      "learning_rate": 4.524018991422318e-06,
+      "loss": 0.5928,
+      "step": 4592
+    },
+    {
+      "epoch": 1.2193017390150007,
+      "grad_norm": 0.410511604327231,
+      "learning_rate": 4.523814043549509e-06,
+      "loss": 0.6006,
+      "step": 4593
+    },
+    {
+      "epoch": 1.2195672374883844,
+      "grad_norm": 0.4039501558846791,
+      "learning_rate": 4.523609056207372e-06,
+      "loss": 0.5889,
+      "step": 4594
+    },
+    {
+      "epoch": 1.2198327359617682,
+      "grad_norm": 0.41508578310649125,
+      "learning_rate": 4.523404029399903e-06,
+      "loss": 0.5829,
+      "step": 4595
+    },
+    {
+      "epoch": 1.220098234435152,
+      "grad_norm": 0.3970011813756224,
+      "learning_rate": 4.523198963131101e-06,
+      "loss": 0.6332,
+      "step": 4596
+    },
+    {
+      "epoch": 1.2203637329085357,
+      "grad_norm": 0.41313031555925156,
+      "learning_rate": 4.522993857404966e-06,
+      "loss": 0.6023,
+      "step": 4597
+    },
+    {
+      "epoch": 1.2206292313819196,
+      "grad_norm": 0.39739995783304405,
+      "learning_rate": 4.522788712225498e-06,
+      "loss": 0.5933,
+      "step": 4598
+    },
+    {
+      "epoch": 1.2208947298553032,
+      "grad_norm": 0.4167065395080637,
+      "learning_rate": 4.522583527596696e-06,
+      "loss": 0.6148,
+      "step": 4599
+    },
+    {
+      "epoch": 1.221160228328687,
+      "grad_norm": 0.4144098222409626,
+      "learning_rate": 4.522378303522566e-06,
+      "loss": 0.5566,
+      "step": 4600
+    },
+    {
+      "epoch": 1.221425726802071,
+      "grad_norm": 0.3998625675016481,
+      "learning_rate": 4.522173040007104e-06,
+      "loss": 0.5745,
+      "step": 4601
+    },
+    {
+      "epoch": 1.2216912252754546,
+      "grad_norm": 0.39702683324226407,
+      "learning_rate": 4.521967737054319e-06,
+      "loss": 0.5604,
+      "step": 4602
+    },
+    {
+      "epoch": 1.2219567237488385,
+      "grad_norm": 0.41059343956295374,
+      "learning_rate": 4.521762394668212e-06,
+      "loss": 0.6153,
+      "step": 4603
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.4009489866586967,
+      "learning_rate": 4.521557012852787e-06,
+      "loss": 0.6155,
+      "step": 4604
+    },
+    {
+      "epoch": 1.222487720695606,
+      "grad_norm": 0.41834760473973975,
+      "learning_rate": 4.521351591612052e-06,
+      "loss": 0.6347,
+      "step": 4605
+    },
+    {
+      "epoch": 1.2227532191689898,
+      "grad_norm": 0.4079427954416437,
+      "learning_rate": 4.521146130950011e-06,
+      "loss": 0.589,
+      "step": 4606
+    },
+    {
+      "epoch": 1.2230187176423735,
+      "grad_norm": 0.3954929039564125,
+      "learning_rate": 4.520940630870672e-06,
+      "loss": 0.629,
+      "step": 4607
+    },
+    {
+      "epoch": 1.2232842161157573,
+      "grad_norm": 0.4085706251219328,
+      "learning_rate": 4.520735091378042e-06,
+      "loss": 0.5406,
+      "step": 4608
+    },
+    {
+      "epoch": 1.2235497145891412,
+      "grad_norm": 0.4110493967941192,
+      "learning_rate": 4.520529512476132e-06,
+      "loss": 0.616,
+      "step": 4609
+    },
+    {
+      "epoch": 1.2238152130625248,
+      "grad_norm": 0.3983089461016142,
+      "learning_rate": 4.520323894168948e-06,
+      "loss": 0.5994,
+      "step": 4610
+    },
+    {
+      "epoch": 1.2240807115359087,
+      "grad_norm": 0.40665554619408245,
+      "learning_rate": 4.520118236460501e-06,
+      "loss": 0.6026,
+      "step": 4611
+    },
+    {
+      "epoch": 1.2243462100092923,
+      "grad_norm": 0.4403881455444084,
+      "learning_rate": 4.519912539354803e-06,
+      "loss": 0.558,
+      "step": 4612
+    },
+    {
+      "epoch": 1.2246117084826762,
+      "grad_norm": 0.42054227170701386,
+      "learning_rate": 4.519706802855866e-06,
+      "loss": 0.5818,
+      "step": 4613
+    },
+    {
+      "epoch": 1.22487720695606,
+      "grad_norm": 0.41161084844160967,
+      "learning_rate": 4.519501026967699e-06,
+      "loss": 0.5766,
+      "step": 4614
+    },
+    {
+      "epoch": 1.2251427054294437,
+      "grad_norm": 0.39775434452622377,
+      "learning_rate": 4.519295211694319e-06,
+      "loss": 0.6018,
+      "step": 4615
+    },
+    {
+      "epoch": 1.2254082039028276,
+      "grad_norm": 0.4171902766664939,
+      "learning_rate": 4.519089357039738e-06,
+      "loss": 0.5896,
+      "step": 4616
+    },
+    {
+      "epoch": 1.2256737023762114,
+      "grad_norm": 0.4255779402347447,
+      "learning_rate": 4.5188834630079705e-06,
+      "loss": 0.6361,
+      "step": 4617
+    },
+    {
+      "epoch": 1.225939200849595,
+      "grad_norm": 0.4100721490623379,
+      "learning_rate": 4.518677529603032e-06,
+      "loss": 0.603,
+      "step": 4618
+    },
+    {
+      "epoch": 1.226204699322979,
+      "grad_norm": 0.4062782809134659,
+      "learning_rate": 4.518471556828939e-06,
+      "loss": 0.6254,
+      "step": 4619
+    },
+    {
+      "epoch": 1.2264701977963628,
+      "grad_norm": 0.4091156502775502,
+      "learning_rate": 4.51826554468971e-06,
+      "loss": 0.6022,
+      "step": 4620
+    },
+    {
+      "epoch": 1.2267356962697464,
+      "grad_norm": 0.41707692683216596,
+      "learning_rate": 4.5180594931893605e-06,
+      "loss": 0.6062,
+      "step": 4621
+    },
+    {
+      "epoch": 1.2270011947431303,
+      "grad_norm": 0.4056566322567737,
+      "learning_rate": 4.5178534023319096e-06,
+      "loss": 0.6122,
+      "step": 4622
+    },
+    {
+      "epoch": 1.227266693216514,
+      "grad_norm": 0.41111474368024437,
+      "learning_rate": 4.517647272121377e-06,
+      "loss": 0.6156,
+      "step": 4623
+    },
+    {
+      "epoch": 1.2275321916898978,
+      "grad_norm": 0.41366505204896453,
+      "learning_rate": 4.517441102561783e-06,
+      "loss": 0.5562,
+      "step": 4624
+    },
+    {
+      "epoch": 1.2277976901632814,
+      "grad_norm": 0.405011097796717,
+      "learning_rate": 4.517234893657148e-06,
+      "loss": 0.5697,
+      "step": 4625
+    },
+    {
+      "epoch": 1.2280631886366653,
+      "grad_norm": 0.4013322854132762,
+      "learning_rate": 4.5170286454114934e-06,
+      "loss": 0.5999,
+      "step": 4626
+    },
+    {
+      "epoch": 1.2283286871100492,
+      "grad_norm": 0.4131120973257043,
+      "learning_rate": 4.516822357828841e-06,
+      "loss": 0.5804,
+      "step": 4627
+    },
+    {
+      "epoch": 1.2285941855834328,
+      "grad_norm": 0.4046586142140177,
+      "learning_rate": 4.516616030913214e-06,
+      "loss": 0.6332,
+      "step": 4628
+    },
+    {
+      "epoch": 1.2288596840568167,
+      "grad_norm": 0.4001701879071022,
+      "learning_rate": 4.5164096646686385e-06,
+      "loss": 0.5561,
+      "step": 4629
+    },
+    {
+      "epoch": 1.2291251825302005,
+      "grad_norm": 0.40580430907131293,
+      "learning_rate": 4.516203259099137e-06,
+      "loss": 0.6,
+      "step": 4630
+    },
+    {
+      "epoch": 1.2293906810035842,
+      "grad_norm": 0.40016235897519387,
+      "learning_rate": 4.515996814208735e-06,
+      "loss": 0.5774,
+      "step": 4631
+    },
+    {
+      "epoch": 1.229656179476968,
+      "grad_norm": 0.3934421227732901,
+      "learning_rate": 4.51579033000146e-06,
+      "loss": 0.5922,
+      "step": 4632
+    },
+    {
+      "epoch": 1.229921677950352,
+      "grad_norm": 0.4024494322905055,
+      "learning_rate": 4.515583806481337e-06,
+      "loss": 0.6046,
+      "step": 4633
+    },
+    {
+      "epoch": 1.2301871764237355,
+      "grad_norm": 0.4048726091100463,
+      "learning_rate": 4.5153772436523955e-06,
+      "loss": 0.6374,
+      "step": 4634
+    },
+    {
+      "epoch": 1.2304526748971194,
+      "grad_norm": 0.4167136721780543,
+      "learning_rate": 4.515170641518663e-06,
+      "loss": 0.5964,
+      "step": 4635
+    },
+    {
+      "epoch": 1.230718173370503,
+      "grad_norm": 0.4107196437185372,
+      "learning_rate": 4.514964000084169e-06,
+      "loss": 0.5774,
+      "step": 4636
+    },
+    {
+      "epoch": 1.230983671843887,
+      "grad_norm": 0.3949272655823079,
+      "learning_rate": 4.514757319352944e-06,
+      "loss": 0.612,
+      "step": 4637
+    },
+    {
+      "epoch": 1.2312491703172708,
+      "grad_norm": 0.42061832336763316,
+      "learning_rate": 4.514550599329017e-06,
+      "loss": 0.6127,
+      "step": 4638
+    },
+    {
+      "epoch": 1.2315146687906544,
+      "grad_norm": 0.40004602168281184,
+      "learning_rate": 4.514343840016422e-06,
+      "loss": 0.6193,
+      "step": 4639
+    },
+    {
+      "epoch": 1.2317801672640383,
+      "grad_norm": 0.41111003842226984,
+      "learning_rate": 4.5141370414191895e-06,
+      "loss": 0.6096,
+      "step": 4640
+    },
+    {
+      "epoch": 1.232045665737422,
+      "grad_norm": 0.4069227657412981,
+      "learning_rate": 4.513930203541354e-06,
+      "loss": 0.6078,
+      "step": 4641
+    },
+    {
+      "epoch": 1.2323111642108058,
+      "grad_norm": 0.3902501374962566,
+      "learning_rate": 4.513723326386948e-06,
+      "loss": 0.5724,
+      "step": 4642
+    },
+    {
+      "epoch": 1.2325766626841896,
+      "grad_norm": 0.4082978881839397,
+      "learning_rate": 4.5135164099600074e-06,
+      "loss": 0.5692,
+      "step": 4643
+    },
+    {
+      "epoch": 1.2328421611575733,
+      "grad_norm": 0.41004955212249095,
+      "learning_rate": 4.513309454264566e-06,
+      "loss": 0.5849,
+      "step": 4644
+    },
+    {
+      "epoch": 1.2331076596309571,
+      "grad_norm": 0.40945634756922294,
+      "learning_rate": 4.5131024593046606e-06,
+      "loss": 0.5997,
+      "step": 4645
+    },
+    {
+      "epoch": 1.233373158104341,
+      "grad_norm": 0.4057072883324477,
+      "learning_rate": 4.512895425084329e-06,
+      "loss": 0.6041,
+      "step": 4646
+    },
+    {
+      "epoch": 1.2336386565777246,
+      "grad_norm": 0.40020944826483784,
+      "learning_rate": 4.512688351607608e-06,
+      "loss": 0.5559,
+      "step": 4647
+    },
+    {
+      "epoch": 1.2339041550511085,
+      "grad_norm": 0.3992843360135841,
+      "learning_rate": 4.512481238878536e-06,
+      "loss": 0.5671,
+      "step": 4648
+    },
+    {
+      "epoch": 1.2341696535244921,
+      "grad_norm": 0.413861323136047,
+      "learning_rate": 4.512274086901153e-06,
+      "loss": 0.5934,
+      "step": 4649
+    },
+    {
+      "epoch": 1.234435151997876,
+      "grad_norm": 0.4131061998339838,
+      "learning_rate": 4.512066895679498e-06,
+      "loss": 0.613,
+      "step": 4650
+    },
+    {
+      "epoch": 1.2347006504712599,
+      "grad_norm": 0.411082721705838,
+      "learning_rate": 4.511859665217612e-06,
+      "loss": 0.5991,
+      "step": 4651
+    },
+    {
+      "epoch": 1.2349661489446435,
+      "grad_norm": 0.4129553397487316,
+      "learning_rate": 4.511652395519537e-06,
+      "loss": 0.6307,
+      "step": 4652
+    },
+    {
+      "epoch": 1.2352316474180274,
+      "grad_norm": 0.4005651475073838,
+      "learning_rate": 4.511445086589315e-06,
+      "loss": 0.5705,
+      "step": 4653
+    },
+    {
+      "epoch": 1.235497145891411,
+      "grad_norm": 0.40416961106595234,
+      "learning_rate": 4.511237738430988e-06,
+      "loss": 0.632,
+      "step": 4654
+    },
+    {
+      "epoch": 1.2357626443647949,
+      "grad_norm": 0.4094538232150745,
+      "learning_rate": 4.511030351048602e-06,
+      "loss": 0.621,
+      "step": 4655
+    },
+    {
+      "epoch": 1.2360281428381787,
+      "grad_norm": 0.41203220445226707,
+      "learning_rate": 4.5108229244462e-06,
+      "loss": 0.5656,
+      "step": 4656
+    },
+    {
+      "epoch": 1.2362936413115624,
+      "grad_norm": 0.3974064204139351,
+      "learning_rate": 4.510615458627827e-06,
+      "loss": 0.624,
+      "step": 4657
+    },
+    {
+      "epoch": 1.2365591397849462,
+      "grad_norm": 0.4053512805953914,
+      "learning_rate": 4.51040795359753e-06,
+      "loss": 0.5914,
+      "step": 4658
+    },
+    {
+      "epoch": 1.2368246382583301,
+      "grad_norm": 0.38385587230019147,
+      "learning_rate": 4.510200409359356e-06,
+      "loss": 0.6076,
+      "step": 4659
+    },
+    {
+      "epoch": 1.2370901367317138,
+      "grad_norm": 0.39302575054313876,
+      "learning_rate": 4.509992825917352e-06,
+      "loss": 0.6034,
+      "step": 4660
+    },
+    {
+      "epoch": 1.2373556352050976,
+      "grad_norm": 0.38631989647523873,
+      "learning_rate": 4.509785203275566e-06,
+      "loss": 0.5872,
+      "step": 4661
+    },
+    {
+      "epoch": 1.2376211336784813,
+      "grad_norm": 0.3947814513041379,
+      "learning_rate": 4.509577541438048e-06,
+      "loss": 0.6012,
+      "step": 4662
+    },
+    {
+      "epoch": 1.2378866321518651,
+      "grad_norm": 0.4049232620209587,
+      "learning_rate": 4.509369840408848e-06,
+      "loss": 0.6128,
+      "step": 4663
+    },
+    {
+      "epoch": 1.238152130625249,
+      "grad_norm": 0.3963694798022003,
+      "learning_rate": 4.5091621001920166e-06,
+      "loss": 0.5964,
+      "step": 4664
+    },
+    {
+      "epoch": 1.2384176290986326,
+      "grad_norm": 0.4114874854497047,
+      "learning_rate": 4.508954320791605e-06,
+      "loss": 0.5651,
+      "step": 4665
+    },
+    {
+      "epoch": 1.2386831275720165,
+      "grad_norm": 0.395159427292642,
+      "learning_rate": 4.508746502211664e-06,
+      "loss": 0.5826,
+      "step": 4666
+    },
+    {
+      "epoch": 1.2389486260454001,
+      "grad_norm": 0.4036992794775691,
+      "learning_rate": 4.508538644456249e-06,
+      "loss": 0.6182,
+      "step": 4667
+    },
+    {
+      "epoch": 1.239214124518784,
+      "grad_norm": 0.3969933140920868,
+      "learning_rate": 4.508330747529413e-06,
+      "loss": 0.6273,
+      "step": 4668
+    },
+    {
+      "epoch": 1.2394796229921679,
+      "grad_norm": 0.40312899372428723,
+      "learning_rate": 4.50812281143521e-06,
+      "loss": 0.6178,
+      "step": 4669
+    },
+    {
+      "epoch": 1.2397451214655515,
+      "grad_norm": 0.4043137724542079,
+      "learning_rate": 4.5079148361776945e-06,
+      "loss": 0.585,
+      "step": 4670
+    },
+    {
+      "epoch": 1.2400106199389354,
+      "grad_norm": 0.45024355811210065,
+      "learning_rate": 4.5077068217609246e-06,
+      "loss": 0.6055,
+      "step": 4671
+    },
+    {
+      "epoch": 1.2402761184123192,
+      "grad_norm": 0.41368168926650717,
+      "learning_rate": 4.5074987681889555e-06,
+      "loss": 0.565,
+      "step": 4672
+    },
+    {
+      "epoch": 1.2405416168857029,
+      "grad_norm": 0.42062874365501995,
+      "learning_rate": 4.5072906754658454e-06,
+      "loss": 0.601,
+      "step": 4673
+    },
+    {
+      "epoch": 1.2408071153590867,
+      "grad_norm": 0.5015529124767025,
+      "learning_rate": 4.5070825435956525e-06,
+      "loss": 0.6023,
+      "step": 4674
+    },
+    {
+      "epoch": 1.2410726138324706,
+      "grad_norm": 0.398246879631349,
+      "learning_rate": 4.506874372582435e-06,
+      "loss": 0.6171,
+      "step": 4675
+    },
+    {
+      "epoch": 1.2413381123058542,
+      "grad_norm": 0.41955983213594866,
+      "learning_rate": 4.506666162430254e-06,
+      "loss": 0.5944,
+      "step": 4676
+    },
+    {
+      "epoch": 1.241603610779238,
+      "grad_norm": 0.5200679786373662,
+      "learning_rate": 4.50645791314317e-06,
+      "loss": 0.5587,
+      "step": 4677
+    },
+    {
+      "epoch": 1.2418691092526217,
+      "grad_norm": 0.41630508089961454,
+      "learning_rate": 4.506249624725244e-06,
+      "loss": 0.6151,
+      "step": 4678
+    },
+    {
+      "epoch": 1.2421346077260056,
+      "grad_norm": 0.40955551028286946,
+      "learning_rate": 4.506041297180538e-06,
+      "loss": 0.5864,
+      "step": 4679
+    },
+    {
+      "epoch": 1.2424001061993892,
+      "grad_norm": 0.43170902491721946,
+      "learning_rate": 4.505832930513115e-06,
+      "loss": 0.5926,
+      "step": 4680
+    },
+    {
+      "epoch": 1.242665604672773,
+      "grad_norm": 0.4083531845328568,
+      "learning_rate": 4.505624524727039e-06,
+      "loss": 0.5993,
+      "step": 4681
+    },
+    {
+      "epoch": 1.242931103146157,
+      "grad_norm": 0.4386420105478396,
+      "learning_rate": 4.5054160798263735e-06,
+      "loss": 0.5785,
+      "step": 4682
+    },
+    {
+      "epoch": 1.2431966016195406,
+      "grad_norm": 0.3951430743520727,
+      "learning_rate": 4.5052075958151856e-06,
+      "loss": 0.576,
+      "step": 4683
+    },
+    {
+      "epoch": 1.2434621000929245,
+      "grad_norm": 0.40548669423749983,
+      "learning_rate": 4.504999072697539e-06,
+      "loss": 0.596,
+      "step": 4684
+    },
+    {
+      "epoch": 1.2437275985663083,
+      "grad_norm": 0.4588289431459512,
+      "learning_rate": 4.504790510477501e-06,
+      "loss": 0.6228,
+      "step": 4685
+    },
+    {
+      "epoch": 1.243993097039692,
+      "grad_norm": 0.4090080408112996,
+      "learning_rate": 4.50458190915914e-06,
+      "loss": 0.593,
+      "step": 4686
+    },
+    {
+      "epoch": 1.2442585955130758,
+      "grad_norm": 0.41576918542041946,
+      "learning_rate": 4.5043732687465245e-06,
+      "loss": 0.5503,
+      "step": 4687
+    },
+    {
+      "epoch": 1.2445240939864597,
+      "grad_norm": 0.4008314511939019,
+      "learning_rate": 4.504164589243721e-06,
+      "loss": 0.5777,
+      "step": 4688
+    },
+    {
+      "epoch": 1.2447895924598433,
+      "grad_norm": 0.4073114351363417,
+      "learning_rate": 4.503955870654803e-06,
+      "loss": 0.6074,
+      "step": 4689
+    },
+    {
+      "epoch": 1.2450550909332272,
+      "grad_norm": 0.4140349938233657,
+      "learning_rate": 4.503747112983838e-06,
+      "loss": 0.6218,
+      "step": 4690
+    },
+    {
+      "epoch": 1.2453205894066108,
+      "grad_norm": 0.42281289291644614,
+      "learning_rate": 4.503538316234898e-06,
+      "loss": 0.6192,
+      "step": 4691
+    },
+    {
+      "epoch": 1.2455860878799947,
+      "grad_norm": 0.4067179694581884,
+      "learning_rate": 4.503329480412055e-06,
+      "loss": 0.6112,
+      "step": 4692
+    },
+    {
+      "epoch": 1.2458515863533786,
+      "grad_norm": 0.4128378596988542,
+      "learning_rate": 4.503120605519383e-06,
+      "loss": 0.621,
+      "step": 4693
+    },
+    {
+      "epoch": 1.2461170848267622,
+      "grad_norm": 0.4210900128113515,
+      "learning_rate": 4.502911691560954e-06,
+      "loss": 0.5675,
+      "step": 4694
+    },
+    {
+      "epoch": 1.246382583300146,
+      "grad_norm": 0.40758204383424324,
+      "learning_rate": 4.502702738540844e-06,
+      "loss": 0.594,
+      "step": 4695
+    },
+    {
+      "epoch": 1.2466480817735297,
+      "grad_norm": 0.39190155252852305,
+      "learning_rate": 4.502493746463126e-06,
+      "loss": 0.6229,
+      "step": 4696
+    },
+    {
+      "epoch": 1.2469135802469136,
+      "grad_norm": 0.3956725827925965,
+      "learning_rate": 4.502284715331878e-06,
+      "loss": 0.5718,
+      "step": 4697
+    },
+    {
+      "epoch": 1.2471790787202974,
+      "grad_norm": 0.39501824097640387,
+      "learning_rate": 4.502075645151175e-06,
+      "loss": 0.5612,
+      "step": 4698
+    },
+    {
+      "epoch": 1.247444577193681,
+      "grad_norm": 0.39803292433372905,
+      "learning_rate": 4.501866535925096e-06,
+      "loss": 0.5822,
+      "step": 4699
+    },
+    {
+      "epoch": 1.247710075667065,
+      "grad_norm": 0.3984842903381202,
+      "learning_rate": 4.501657387657716e-06,
+      "loss": 0.6264,
+      "step": 4700
+    },
+    {
+      "epoch": 1.2479755741404488,
+      "grad_norm": 0.3967076500272702,
+      "learning_rate": 4.501448200353118e-06,
+      "loss": 0.5803,
+      "step": 4701
+    },
+    {
+      "epoch": 1.2482410726138324,
+      "grad_norm": 0.3976296019879218,
+      "learning_rate": 4.501238974015379e-06,
+      "loss": 0.6121,
+      "step": 4702
+    },
+    {
+      "epoch": 1.2485065710872163,
+      "grad_norm": 0.4145519121192118,
+      "learning_rate": 4.50102970864858e-06,
+      "loss": 0.6072,
+      "step": 4703
+    },
+    {
+      "epoch": 1.2487720695606,
+      "grad_norm": 0.39764640397743245,
+      "learning_rate": 4.500820404256803e-06,
+      "loss": 0.6249,
+      "step": 4704
+    },
+    {
+      "epoch": 1.2490375680339838,
+      "grad_norm": 0.41122615024325393,
+      "learning_rate": 4.500611060844129e-06,
+      "loss": 0.6099,
+      "step": 4705
+    },
+    {
+      "epoch": 1.2493030665073677,
+      "grad_norm": 0.3890616770628102,
+      "learning_rate": 4.50040167841464e-06,
+      "loss": 0.6027,
+      "step": 4706
+    },
+    {
+      "epoch": 1.2495685649807513,
+      "grad_norm": 0.4058409505074949,
+      "learning_rate": 4.500192256972421e-06,
+      "loss": 0.576,
+      "step": 4707
+    },
+    {
+      "epoch": 1.2498340634541352,
+      "grad_norm": 0.40915680828716045,
+      "learning_rate": 4.499982796521556e-06,
+      "loss": 0.6163,
+      "step": 4708
+    },
+    {
+      "epoch": 1.2500995619275188,
+      "grad_norm": 0.4053049845103103,
+      "learning_rate": 4.49977329706613e-06,
+      "loss": 0.6318,
+      "step": 4709
+    },
+    {
+      "epoch": 1.2503650604009027,
+      "grad_norm": 0.41696705173949994,
+      "learning_rate": 4.499563758610228e-06,
+      "loss": 0.568,
+      "step": 4710
+    },
+    {
+      "epoch": 1.2506305588742865,
+      "grad_norm": 0.415157862556153,
+      "learning_rate": 4.499354181157936e-06,
+      "loss": 0.5802,
+      "step": 4711
+    },
+    {
+      "epoch": 1.2508960573476702,
+      "grad_norm": 0.40249290308362196,
+      "learning_rate": 4.499144564713343e-06,
+      "loss": 0.5692,
+      "step": 4712
+    },
+    {
+      "epoch": 1.251161555821054,
+      "grad_norm": 0.39639656091276615,
+      "learning_rate": 4.498934909280537e-06,
+      "loss": 0.6373,
+      "step": 4713
+    },
+    {
+      "epoch": 1.251427054294438,
+      "grad_norm": 0.4021407436121963,
+      "learning_rate": 4.498725214863605e-06,
+      "loss": 0.6079,
+      "step": 4714
+    },
+    {
+      "epoch": 1.2516925527678215,
+      "grad_norm": 0.4063696027207103,
+      "learning_rate": 4.498515481466639e-06,
+      "loss": 0.5864,
+      "step": 4715
+    },
+    {
+      "epoch": 1.2519580512412054,
+      "grad_norm": 0.4025966927960479,
+      "learning_rate": 4.498305709093726e-06,
+      "loss": 0.5816,
+      "step": 4716
+    },
+    {
+      "epoch": 1.2522235497145893,
+      "grad_norm": 0.41638340641391497,
+      "learning_rate": 4.49809589774896e-06,
+      "loss": 0.6278,
+      "step": 4717
+    },
+    {
+      "epoch": 1.252489048187973,
+      "grad_norm": 0.40860106293670095,
+      "learning_rate": 4.497886047436432e-06,
+      "loss": 0.6149,
+      "step": 4718
+    },
+    {
+      "epoch": 1.2527545466613568,
+      "grad_norm": 0.4062239507832847,
+      "learning_rate": 4.4976761581602335e-06,
+      "loss": 0.5809,
+      "step": 4719
+    },
+    {
+      "epoch": 1.2530200451347404,
+      "grad_norm": 0.40019229522876015,
+      "learning_rate": 4.497466229924459e-06,
+      "loss": 0.608,
+      "step": 4720
+    },
+    {
+      "epoch": 1.2532855436081243,
+      "grad_norm": 0.40499221575238076,
+      "learning_rate": 4.497256262733203e-06,
+      "loss": 0.5626,
+      "step": 4721
+    },
+    {
+      "epoch": 1.253551042081508,
+      "grad_norm": 0.4083037706323773,
+      "learning_rate": 4.497046256590559e-06,
+      "loss": 0.5813,
+      "step": 4722
+    },
+    {
+      "epoch": 1.2538165405548918,
+      "grad_norm": 0.3988592553494657,
+      "learning_rate": 4.4968362115006245e-06,
+      "loss": 0.5634,
+      "step": 4723
+    },
+    {
+      "epoch": 1.2540820390282756,
+      "grad_norm": 0.41194910918504213,
+      "learning_rate": 4.4966261274674934e-06,
+      "loss": 0.5874,
+      "step": 4724
+    },
+    {
+      "epoch": 1.2543475375016593,
+      "grad_norm": 0.4098327613785817,
+      "learning_rate": 4.496416004495265e-06,
+      "loss": 0.5952,
+      "step": 4725
+    },
+    {
+      "epoch": 1.2546130359750431,
+      "grad_norm": 0.40154345997679974,
+      "learning_rate": 4.496205842588036e-06,
+      "loss": 0.5944,
+      "step": 4726
+    },
+    {
+      "epoch": 1.254878534448427,
+      "grad_norm": 0.4047970712517502,
+      "learning_rate": 4.495995641749907e-06,
+      "loss": 0.6039,
+      "step": 4727
+    },
+    {
+      "epoch": 1.2551440329218106,
+      "grad_norm": 0.5090464261553893,
+      "learning_rate": 4.495785401984975e-06,
+      "loss": 0.623,
+      "step": 4728
+    },
+    {
+      "epoch": 1.2554095313951945,
+      "grad_norm": 0.40659137936963125,
+      "learning_rate": 4.49557512329734e-06,
+      "loss": 0.5657,
+      "step": 4729
+    },
+    {
+      "epoch": 1.2556750298685784,
+      "grad_norm": 0.37876860421427405,
+      "learning_rate": 4.495364805691105e-06,
+      "loss": 0.5948,
+      "step": 4730
+    },
+    {
+      "epoch": 1.255940528341962,
+      "grad_norm": 0.4097370145139313,
+      "learning_rate": 4.495154449170371e-06,
+      "loss": 0.5725,
+      "step": 4731
+    },
+    {
+      "epoch": 1.2562060268153459,
+      "grad_norm": 0.4016076593630768,
+      "learning_rate": 4.49494405373924e-06,
+      "loss": 0.5599,
+      "step": 4732
+    },
+    {
+      "epoch": 1.2564715252887295,
+      "grad_norm": 0.41112434906347095,
+      "learning_rate": 4.494733619401815e-06,
+      "loss": 0.5867,
+      "step": 4733
+    },
+    {
+      "epoch": 1.2567370237621134,
+      "grad_norm": 0.41230364602881003,
+      "learning_rate": 4.4945231461622015e-06,
+      "loss": 0.6118,
+      "step": 4734
+    },
+    {
+      "epoch": 1.257002522235497,
+      "grad_norm": 0.3939069021295258,
+      "learning_rate": 4.494312634024502e-06,
+      "loss": 0.5986,
+      "step": 4735
+    },
+    {
+      "epoch": 1.2572680207088809,
+      "grad_norm": 0.3974222344781593,
+      "learning_rate": 4.494102082992825e-06,
+      "loss": 0.5912,
+      "step": 4736
+    },
+    {
+      "epoch": 1.2575335191822647,
+      "grad_norm": 0.4137566663149835,
+      "learning_rate": 4.493891493071274e-06,
+      "loss": 0.6081,
+      "step": 4737
+    },
+    {
+      "epoch": 1.2577990176556484,
+      "grad_norm": 0.39977547756883264,
+      "learning_rate": 4.4936808642639565e-06,
+      "loss": 0.5977,
+      "step": 4738
+    },
+    {
+      "epoch": 1.2580645161290323,
+      "grad_norm": 0.416984841331229,
+      "learning_rate": 4.493470196574981e-06,
+      "loss": 0.5927,
+      "step": 4739
+    },
+    {
+      "epoch": 1.2583300146024161,
+      "grad_norm": 0.3934255723016643,
+      "learning_rate": 4.493259490008457e-06,
+      "loss": 0.5806,
+      "step": 4740
+    },
+    {
+      "epoch": 1.2585955130757998,
+      "grad_norm": 0.4151678926674046,
+      "learning_rate": 4.493048744568491e-06,
+      "loss": 0.5773,
+      "step": 4741
+    },
+    {
+      "epoch": 1.2588610115491836,
+      "grad_norm": 0.4085956237297976,
+      "learning_rate": 4.4928379602591955e-06,
+      "loss": 0.5093,
+      "step": 4742
+    },
+    {
+      "epoch": 1.2591265100225675,
+      "grad_norm": 0.3979179667069224,
+      "learning_rate": 4.49262713708468e-06,
+      "loss": 0.5915,
+      "step": 4743
+    },
+    {
+      "epoch": 1.2593920084959511,
+      "grad_norm": 0.4055209220212614,
+      "learning_rate": 4.492416275049057e-06,
+      "loss": 0.631,
+      "step": 4744
+    },
+    {
+      "epoch": 1.259657506969335,
+      "grad_norm": 0.4099440898415954,
+      "learning_rate": 4.492205374156438e-06,
+      "loss": 0.5806,
+      "step": 4745
+    },
+    {
+      "epoch": 1.2599230054427186,
+      "grad_norm": 0.4244296166019007,
+      "learning_rate": 4.491994434410937e-06,
+      "loss": 0.5447,
+      "step": 4746
+    },
+    {
+      "epoch": 1.2601885039161025,
+      "grad_norm": 0.40926102246167423,
+      "learning_rate": 4.4917834558166665e-06,
+      "loss": 0.6181,
+      "step": 4747
+    },
+    {
+      "epoch": 1.2604540023894861,
+      "grad_norm": 0.41411071143827854,
+      "learning_rate": 4.4915724383777425e-06,
+      "loss": 0.5786,
+      "step": 4748
+    },
+    {
+      "epoch": 1.26071950086287,
+      "grad_norm": 0.431605614908769,
+      "learning_rate": 4.491361382098279e-06,
+      "loss": 0.5808,
+      "step": 4749
+    },
+    {
+      "epoch": 1.2609849993362539,
+      "grad_norm": 0.4126530945020734,
+      "learning_rate": 4.491150286982394e-06,
+      "loss": 0.6124,
+      "step": 4750
+    },
+    {
+      "epoch": 1.2612504978096375,
+      "grad_norm": 0.41152550735466387,
+      "learning_rate": 4.490939153034203e-06,
+      "loss": 0.597,
+      "step": 4751
+    },
+    {
+      "epoch": 1.2615159962830214,
+      "grad_norm": 0.42812263358301217,
+      "learning_rate": 4.490727980257823e-06,
+      "loss": 0.6013,
+      "step": 4752
+    },
+    {
+      "epoch": 1.2617814947564052,
+      "grad_norm": 0.4087351316592823,
+      "learning_rate": 4.490516768657373e-06,
+      "loss": 0.5725,
+      "step": 4753
+    },
+    {
+      "epoch": 1.2620469932297889,
+      "grad_norm": 0.4222095885754993,
+      "learning_rate": 4.4903055182369735e-06,
+      "loss": 0.5928,
+      "step": 4754
+    },
+    {
+      "epoch": 1.2623124917031727,
+      "grad_norm": 0.41816055562302085,
+      "learning_rate": 4.490094229000743e-06,
+      "loss": 0.6029,
+      "step": 4755
+    },
+    {
+      "epoch": 1.2625779901765566,
+      "grad_norm": 0.38859145446784293,
+      "learning_rate": 4.489882900952802e-06,
+      "loss": 0.5834,
+      "step": 4756
+    },
+    {
+      "epoch": 1.2628434886499402,
+      "grad_norm": 0.4920456543950286,
+      "learning_rate": 4.489671534097273e-06,
+      "loss": 0.5923,
+      "step": 4757
+    },
+    {
+      "epoch": 1.263108987123324,
+      "grad_norm": 0.38975299000872626,
+      "learning_rate": 4.489460128438276e-06,
+      "loss": 0.5658,
+      "step": 4758
+    },
+    {
+      "epoch": 1.263374485596708,
+      "grad_norm": 0.4086697348407216,
+      "learning_rate": 4.489248683979937e-06,
+      "loss": 0.5668,
+      "step": 4759
+    },
+    {
+      "epoch": 1.2636399840700916,
+      "grad_norm": 0.4091965681306532,
+      "learning_rate": 4.489037200726378e-06,
+      "loss": 0.6212,
+      "step": 4760
+    },
+    {
+      "epoch": 1.2639054825434755,
+      "grad_norm": 0.44470399037483943,
+      "learning_rate": 4.488825678681724e-06,
+      "loss": 0.5685,
+      "step": 4761
+    },
+    {
+      "epoch": 1.264170981016859,
+      "grad_norm": 0.42906645057804543,
+      "learning_rate": 4.488614117850098e-06,
+      "loss": 0.6041,
+      "step": 4762
+    },
+    {
+      "epoch": 1.264436479490243,
+      "grad_norm": 0.4030515200006183,
+      "learning_rate": 4.48840251823563e-06,
+      "loss": 0.5984,
+      "step": 4763
+    },
+    {
+      "epoch": 1.2647019779636266,
+      "grad_norm": 0.41071995572515413,
+      "learning_rate": 4.488190879842443e-06,
+      "loss": 0.5844,
+      "step": 4764
+    },
+    {
+      "epoch": 1.2649674764370105,
+      "grad_norm": 0.41879161853034597,
+      "learning_rate": 4.487979202674667e-06,
+      "loss": 0.5673,
+      "step": 4765
+    },
+    {
+      "epoch": 1.2652329749103943,
+      "grad_norm": 0.4073524914173789,
+      "learning_rate": 4.487767486736428e-06,
+      "loss": 0.5909,
+      "step": 4766
+    },
+    {
+      "epoch": 1.265498473383778,
+      "grad_norm": 0.4310325096706651,
+      "learning_rate": 4.487555732031856e-06,
+      "loss": 0.5652,
+      "step": 4767
+    },
+    {
+      "epoch": 1.2657639718571618,
+      "grad_norm": 0.4088101291713759,
+      "learning_rate": 4.487343938565082e-06,
+      "loss": 0.5817,
+      "step": 4768
+    },
+    {
+      "epoch": 1.2660294703305457,
+      "grad_norm": 0.40114012494791745,
+      "learning_rate": 4.4871321063402345e-06,
+      "loss": 0.6111,
+      "step": 4769
+    },
+    {
+      "epoch": 1.2662949688039293,
+      "grad_norm": 0.4390390891353396,
+      "learning_rate": 4.486920235361446e-06,
+      "loss": 0.5953,
+      "step": 4770
+    },
+    {
+      "epoch": 1.2665604672773132,
+      "grad_norm": 0.4107415947510306,
+      "learning_rate": 4.486708325632847e-06,
+      "loss": 0.5918,
+      "step": 4771
+    },
+    {
+      "epoch": 1.266825965750697,
+      "grad_norm": 0.4041596575008259,
+      "learning_rate": 4.486496377158572e-06,
+      "loss": 0.5535,
+      "step": 4772
+    },
+    {
+      "epoch": 1.2670914642240807,
+      "grad_norm": 0.3952041202089783,
+      "learning_rate": 4.486284389942755e-06,
+      "loss": 0.6087,
+      "step": 4773
+    },
+    {
+      "epoch": 1.2673569626974646,
+      "grad_norm": 0.4069445701758252,
+      "learning_rate": 4.486072363989528e-06,
+      "loss": 0.5532,
+      "step": 4774
+    },
+    {
+      "epoch": 1.2676224611708482,
+      "grad_norm": 0.46798041103016264,
+      "learning_rate": 4.4858602993030275e-06,
+      "loss": 0.5893,
+      "step": 4775
+    },
+    {
+      "epoch": 1.267887959644232,
+      "grad_norm": 0.39629517895838134,
+      "learning_rate": 4.4856481958873885e-06,
+      "loss": 0.5718,
+      "step": 4776
+    },
+    {
+      "epoch": 1.2681534581176157,
+      "grad_norm": 0.4088267200524566,
+      "learning_rate": 4.485436053746749e-06,
+      "loss": 0.6095,
+      "step": 4777
+    },
+    {
+      "epoch": 1.2684189565909996,
+      "grad_norm": 0.4206337488572424,
+      "learning_rate": 4.4852238728852435e-06,
+      "loss": 0.5816,
+      "step": 4778
+    },
+    {
+      "epoch": 1.2686844550643834,
+      "grad_norm": 0.4248247975976921,
+      "learning_rate": 4.485011653307013e-06,
+      "loss": 0.5917,
+      "step": 4779
+    },
+    {
+      "epoch": 1.268949953537767,
+      "grad_norm": 0.4065412804836255,
+      "learning_rate": 4.4847993950161955e-06,
+      "loss": 0.627,
+      "step": 4780
+    },
+    {
+      "epoch": 1.269215452011151,
+      "grad_norm": 0.411719771782179,
+      "learning_rate": 4.4845870980169294e-06,
+      "loss": 0.546,
+      "step": 4781
+    },
+    {
+      "epoch": 1.2694809504845348,
+      "grad_norm": 0.4191350242941418,
+      "learning_rate": 4.484374762313356e-06,
+      "loss": 0.5656,
+      "step": 4782
+    },
+    {
+      "epoch": 1.2697464489579184,
+      "grad_norm": 0.4050154944718186,
+      "learning_rate": 4.484162387909616e-06,
+      "loss": 0.6091,
+      "step": 4783
+    },
+    {
+      "epoch": 1.2700119474313023,
+      "grad_norm": 0.41182283144403287,
+      "learning_rate": 4.483949974809852e-06,
+      "loss": 0.6256,
+      "step": 4784
+    },
+    {
+      "epoch": 1.2702774459046862,
+      "grad_norm": 0.41344406733221284,
+      "learning_rate": 4.483737523018206e-06,
+      "loss": 0.6095,
+      "step": 4785
+    },
+    {
+      "epoch": 1.2705429443780698,
+      "grad_norm": 0.401132951206163,
+      "learning_rate": 4.48352503253882e-06,
+      "loss": 0.5828,
+      "step": 4786
+    },
+    {
+      "epoch": 1.2708084428514537,
+      "grad_norm": 0.40306996783221555,
+      "learning_rate": 4.4833125033758404e-06,
+      "loss": 0.5794,
+      "step": 4787
+    },
+    {
+      "epoch": 1.2710739413248373,
+      "grad_norm": 0.40600983052603384,
+      "learning_rate": 4.48309993553341e-06,
+      "loss": 0.5955,
+      "step": 4788
+    },
+    {
+      "epoch": 1.2713394397982212,
+      "grad_norm": 0.39566826374726843,
+      "learning_rate": 4.482887329015677e-06,
+      "loss": 0.5818,
+      "step": 4789
+    },
+    {
+      "epoch": 1.2716049382716048,
+      "grad_norm": 0.40271988676629156,
+      "learning_rate": 4.4826746838267855e-06,
+      "loss": 0.5918,
+      "step": 4790
+    },
+    {
+      "epoch": 1.2718704367449887,
+      "grad_norm": 0.3869110575668333,
+      "learning_rate": 4.4824619999708825e-06,
+      "loss": 0.6297,
+      "step": 4791
+    },
+    {
+      "epoch": 1.2721359352183725,
+      "grad_norm": 0.3988670578931698,
+      "learning_rate": 4.482249277452117e-06,
+      "loss": 0.5732,
+      "step": 4792
+    },
+    {
+      "epoch": 1.2724014336917562,
+      "grad_norm": 0.40337604236371577,
+      "learning_rate": 4.482036516274637e-06,
+      "loss": 0.5877,
+      "step": 4793
+    },
+    {
+      "epoch": 1.27266693216514,
+      "grad_norm": 0.3965852107471716,
+      "learning_rate": 4.4818237164425926e-06,
+      "loss": 0.6058,
+      "step": 4794
+    },
+    {
+      "epoch": 1.272932430638524,
+      "grad_norm": 0.39467361845498866,
+      "learning_rate": 4.481610877960133e-06,
+      "loss": 0.5946,
+      "step": 4795
+    },
+    {
+      "epoch": 1.2731979291119075,
+      "grad_norm": 0.41405476701881155,
+      "learning_rate": 4.48139800083141e-06,
+      "loss": 0.5655,
+      "step": 4796
+    },
+    {
+      "epoch": 1.2734634275852914,
+      "grad_norm": 0.39727062147722964,
+      "learning_rate": 4.481185085060574e-06,
+      "loss": 0.5744,
+      "step": 4797
+    },
+    {
+      "epoch": 1.2737289260586753,
+      "grad_norm": 0.4051504859500576,
+      "learning_rate": 4.480972130651779e-06,
+      "loss": 0.5939,
+      "step": 4798
+    },
+    {
+      "epoch": 1.273994424532059,
+      "grad_norm": 0.4110274549838971,
+      "learning_rate": 4.480759137609177e-06,
+      "loss": 0.5798,
+      "step": 4799
+    },
+    {
+      "epoch": 1.2742599230054428,
+      "grad_norm": 0.4041875531959028,
+      "learning_rate": 4.480546105936921e-06,
+      "loss": 0.6093,
+      "step": 4800
+    },
+    {
+      "epoch": 1.2745254214788264,
+      "grad_norm": 0.3970489992123183,
+      "learning_rate": 4.4803330356391675e-06,
+      "loss": 0.599,
+      "step": 4801
+    },
+    {
+      "epoch": 1.2747909199522103,
+      "grad_norm": 0.40995236191142226,
+      "learning_rate": 4.4801199267200705e-06,
+      "loss": 0.6095,
+      "step": 4802
+    },
+    {
+      "epoch": 1.275056418425594,
+      "grad_norm": 0.4015925545741892,
+      "learning_rate": 4.479906779183788e-06,
+      "loss": 0.6153,
+      "step": 4803
+    },
+    {
+      "epoch": 1.2753219168989778,
+      "grad_norm": 0.4093549856330689,
+      "learning_rate": 4.479693593034475e-06,
+      "loss": 0.5716,
+      "step": 4804
+    },
+    {
+      "epoch": 1.2755874153723616,
+      "grad_norm": 0.38151222400251716,
+      "learning_rate": 4.479480368276289e-06,
+      "loss": 0.6065,
+      "step": 4805
+    },
+    {
+      "epoch": 1.2758529138457453,
+      "grad_norm": 0.40350724378341696,
+      "learning_rate": 4.47926710491339e-06,
+      "loss": 0.6121,
+      "step": 4806
+    },
+    {
+      "epoch": 1.2761184123191291,
+      "grad_norm": 0.4075373516368296,
+      "learning_rate": 4.479053802949936e-06,
+      "loss": 0.6211,
+      "step": 4807
+    },
+    {
+      "epoch": 1.276383910792513,
+      "grad_norm": 0.41150806857921224,
+      "learning_rate": 4.478840462390088e-06,
+      "loss": 0.5906,
+      "step": 4808
+    },
+    {
+      "epoch": 1.2766494092658967,
+      "grad_norm": 0.40762919214732546,
+      "learning_rate": 4.478627083238005e-06,
+      "loss": 0.5958,
+      "step": 4809
+    },
+    {
+      "epoch": 1.2769149077392805,
+      "grad_norm": 0.3930084019369108,
+      "learning_rate": 4.4784136654978505e-06,
+      "loss": 0.5839,
+      "step": 4810
+    },
+    {
+      "epoch": 1.2771804062126644,
+      "grad_norm": 0.41015098535345956,
+      "learning_rate": 4.478200209173785e-06,
+      "loss": 0.5903,
+      "step": 4811
+    },
+    {
+      "epoch": 1.277445904686048,
+      "grad_norm": 0.41187644322477424,
+      "learning_rate": 4.4779867142699715e-06,
+      "loss": 0.5507,
+      "step": 4812
+    },
+    {
+      "epoch": 1.2777114031594319,
+      "grad_norm": 0.40852280586238554,
+      "learning_rate": 4.477773180790574e-06,
+      "loss": 0.5943,
+      "step": 4813
+    },
+    {
+      "epoch": 1.2779769016328157,
+      "grad_norm": 0.4030066484287729,
+      "learning_rate": 4.477559608739758e-06,
+      "loss": 0.5891,
+      "step": 4814
+    },
+    {
+      "epoch": 1.2782424001061994,
+      "grad_norm": 0.40570442453337285,
+      "learning_rate": 4.477345998121687e-06,
+      "loss": 0.5892,
+      "step": 4815
+    },
+    {
+      "epoch": 1.2785078985795832,
+      "grad_norm": 0.40639671976606295,
+      "learning_rate": 4.477132348940528e-06,
+      "loss": 0.6073,
+      "step": 4816
+    },
+    {
+      "epoch": 1.2787733970529669,
+      "grad_norm": 0.41492400827927484,
+      "learning_rate": 4.476918661200447e-06,
+      "loss": 0.583,
+      "step": 4817
+    },
+    {
+      "epoch": 1.2790388955263507,
+      "grad_norm": 0.41355270732709376,
+      "learning_rate": 4.476704934905612e-06,
+      "loss": 0.5681,
+      "step": 4818
+    },
+    {
+      "epoch": 1.2793043939997344,
+      "grad_norm": 0.40671603678749557,
+      "learning_rate": 4.476491170060191e-06,
+      "loss": 0.5812,
+      "step": 4819
+    },
+    {
+      "epoch": 1.2795698924731183,
+      "grad_norm": 0.40960390021168025,
+      "learning_rate": 4.476277366668353e-06,
+      "loss": 0.6126,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2798353909465021,
+      "grad_norm": 0.3989253870870453,
+      "learning_rate": 4.476063524734268e-06,
+      "loss": 0.5956,
+      "step": 4821
+    },
+    {
+      "epoch": 1.2801008894198858,
+      "grad_norm": 0.3959694536636935,
+      "learning_rate": 4.4758496442621055e-06,
+      "loss": 0.565,
+      "step": 4822
+    },
+    {
+      "epoch": 1.2803663878932696,
+      "grad_norm": 0.4050948502117333,
+      "learning_rate": 4.475635725256038e-06,
+      "loss": 0.5907,
+      "step": 4823
+    },
+    {
+      "epoch": 1.2806318863666535,
+      "grad_norm": 0.3985265120580581,
+      "learning_rate": 4.475421767720235e-06,
+      "loss": 0.5825,
+      "step": 4824
+    },
+    {
+      "epoch": 1.2808973848400371,
+      "grad_norm": 0.3984796997375991,
+      "learning_rate": 4.4752077716588725e-06,
+      "loss": 0.6085,
+      "step": 4825
+    },
+    {
+      "epoch": 1.281162883313421,
+      "grad_norm": 0.38902982064431796,
+      "learning_rate": 4.474993737076121e-06,
+      "loss": 0.5868,
+      "step": 4826
+    },
+    {
+      "epoch": 1.2814283817868048,
+      "grad_norm": 0.4055795435691279,
+      "learning_rate": 4.474779663976157e-06,
+      "loss": 0.5712,
+      "step": 4827
+    },
+    {
+      "epoch": 1.2816938802601885,
+      "grad_norm": 0.39817241335539527,
+      "learning_rate": 4.474565552363153e-06,
+      "loss": 0.5937,
+      "step": 4828
+    },
+    {
+      "epoch": 1.2819593787335724,
+      "grad_norm": 0.39704626074569316,
+      "learning_rate": 4.474351402241288e-06,
+      "loss": 0.5746,
+      "step": 4829
+    },
+    {
+      "epoch": 1.282224877206956,
+      "grad_norm": 0.42386760846994787,
+      "learning_rate": 4.474137213614735e-06,
+      "loss": 0.5539,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2824903756803399,
+      "grad_norm": 0.41060071853683094,
+      "learning_rate": 4.473922986487674e-06,
+      "loss": 0.594,
+      "step": 4831
+    },
+    {
+      "epoch": 1.2827558741537235,
+      "grad_norm": 0.39749026576463337,
+      "learning_rate": 4.473708720864281e-06,
+      "loss": 0.6217,
+      "step": 4832
+    },
+    {
+      "epoch": 1.2830213726271074,
+      "grad_norm": 0.4205963811592871,
+      "learning_rate": 4.473494416748736e-06,
+      "loss": 0.5945,
+      "step": 4833
+    },
+    {
+      "epoch": 1.2832868711004912,
+      "grad_norm": 0.4213602542027498,
+      "learning_rate": 4.473280074145219e-06,
+      "loss": 0.5607,
+      "step": 4834
+    },
+    {
+      "epoch": 1.2835523695738749,
+      "grad_norm": 0.40042092281192254,
+      "learning_rate": 4.473065693057908e-06,
+      "loss": 0.6233,
+      "step": 4835
+    },
+    {
+      "epoch": 1.2838178680472587,
+      "grad_norm": 0.41131366887398746,
+      "learning_rate": 4.472851273490985e-06,
+      "loss": 0.5799,
+      "step": 4836
+    },
+    {
+      "epoch": 1.2840833665206426,
+      "grad_norm": 0.4023613343319866,
+      "learning_rate": 4.472636815448631e-06,
+      "loss": 0.5956,
+      "step": 4837
+    },
+    {
+      "epoch": 1.2843488649940262,
+      "grad_norm": 0.4149846313082392,
+      "learning_rate": 4.47242231893503e-06,
+      "loss": 0.5907,
+      "step": 4838
+    },
+    {
+      "epoch": 1.28461436346741,
+      "grad_norm": 0.41737786085224055,
+      "learning_rate": 4.472207783954365e-06,
+      "loss": 0.6129,
+      "step": 4839
+    },
+    {
+      "epoch": 1.284879861940794,
+      "grad_norm": 0.4089426490047283,
+      "learning_rate": 4.471993210510819e-06,
+      "loss": 0.5601,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2851453604141776,
+      "grad_norm": 0.40589244994955054,
+      "learning_rate": 4.471778598608577e-06,
+      "loss": 0.6024,
+      "step": 4841
+    },
+    {
+      "epoch": 1.2854108588875615,
+      "grad_norm": 0.4130906605683283,
+      "learning_rate": 4.471563948251824e-06,
+      "loss": 0.5867,
+      "step": 4842
+    },
+    {
+      "epoch": 1.285676357360945,
+      "grad_norm": 0.40454329764588115,
+      "learning_rate": 4.471349259444748e-06,
+      "loss": 0.6083,
+      "step": 4843
+    },
+    {
+      "epoch": 1.285941855834329,
+      "grad_norm": 0.40150588487408206,
+      "learning_rate": 4.471134532191534e-06,
+      "loss": 0.5751,
+      "step": 4844
+    },
+    {
+      "epoch": 1.2862073543077126,
+      "grad_norm": 0.39357217461506994,
+      "learning_rate": 4.470919766496371e-06,
+      "loss": 0.5676,
+      "step": 4845
+    },
+    {
+      "epoch": 1.2864728527810965,
+      "grad_norm": 0.3974261930678769,
+      "learning_rate": 4.470704962363446e-06,
+      "loss": 0.5918,
+      "step": 4846
+    },
+    {
+      "epoch": 1.2867383512544803,
+      "grad_norm": 0.4216397648345008,
+      "learning_rate": 4.47049011979695e-06,
+      "loss": 0.5696,
+      "step": 4847
+    },
+    {
+      "epoch": 1.287003849727864,
+      "grad_norm": 0.3991238213415633,
+      "learning_rate": 4.470275238801072e-06,
+      "loss": 0.5609,
+      "step": 4848
+    },
+    {
+      "epoch": 1.2872693482012478,
+      "grad_norm": 0.39755954531684756,
+      "learning_rate": 4.470060319380002e-06,
+      "loss": 0.5833,
+      "step": 4849
+    },
+    {
+      "epoch": 1.2875348466746317,
+      "grad_norm": 0.410318689773213,
+      "learning_rate": 4.4698453615379325e-06,
+      "loss": 0.6427,
+      "step": 4850
+    },
+    {
+      "epoch": 1.2878003451480153,
+      "grad_norm": 0.4119998414391094,
+      "learning_rate": 4.469630365279056e-06,
+      "loss": 0.6049,
+      "step": 4851
+    },
+    {
+      "epoch": 1.2880658436213992,
+      "grad_norm": 0.415713469107909,
+      "learning_rate": 4.469415330607565e-06,
+      "loss": 0.579,
+      "step": 4852
+    },
+    {
+      "epoch": 1.288331342094783,
+      "grad_norm": 0.4008093507352382,
+      "learning_rate": 4.469200257527652e-06,
+      "loss": 0.5921,
+      "step": 4853
+    },
+    {
+      "epoch": 1.2885968405681667,
+      "grad_norm": 0.4204746050607293,
+      "learning_rate": 4.468985146043514e-06,
+      "loss": 0.6103,
+      "step": 4854
+    },
+    {
+      "epoch": 1.2888623390415506,
+      "grad_norm": 0.4144420026529189,
+      "learning_rate": 4.468769996159343e-06,
+      "loss": 0.6077,
+      "step": 4855
+    },
+    {
+      "epoch": 1.2891278375149344,
+      "grad_norm": 0.41346886902265734,
+      "learning_rate": 4.468554807879339e-06,
+      "loss": 0.5892,
+      "step": 4856
+    },
+    {
+      "epoch": 1.289393335988318,
+      "grad_norm": 0.4130070518978181,
+      "learning_rate": 4.468339581207694e-06,
+      "loss": 0.608,
+      "step": 4857
+    },
+    {
+      "epoch": 1.2896588344617017,
+      "grad_norm": 0.41133939447795226,
+      "learning_rate": 4.468124316148609e-06,
+      "loss": 0.6089,
+      "step": 4858
+    },
+    {
+      "epoch": 1.2899243329350856,
+      "grad_norm": 0.3990111320182523,
+      "learning_rate": 4.467909012706281e-06,
+      "loss": 0.6017,
+      "step": 4859
+    },
+    {
+      "epoch": 1.2901898314084694,
+      "grad_norm": 0.39591399252536497,
+      "learning_rate": 4.467693670884909e-06,
+      "loss": 0.6099,
+      "step": 4860
+    },
+    {
+      "epoch": 1.290455329881853,
+      "grad_norm": 0.4245395063765497,
+      "learning_rate": 4.467478290688693e-06,
+      "loss": 0.6037,
+      "step": 4861
+    },
+    {
+      "epoch": 1.290720828355237,
+      "grad_norm": 0.4131157545958495,
+      "learning_rate": 4.4672628721218325e-06,
+      "loss": 0.5918,
+      "step": 4862
+    },
+    {
+      "epoch": 1.2909863268286208,
+      "grad_norm": 0.4241970294972353,
+      "learning_rate": 4.467047415188529e-06,
+      "loss": 0.6079,
+      "step": 4863
+    },
+    {
+      "epoch": 1.2912518253020044,
+      "grad_norm": 0.40836007528589496,
+      "learning_rate": 4.466831919892986e-06,
+      "loss": 0.6031,
+      "step": 4864
+    },
+    {
+      "epoch": 1.2915173237753883,
+      "grad_norm": 0.40460961855468014,
+      "learning_rate": 4.466616386239403e-06,
+      "loss": 0.6057,
+      "step": 4865
+    },
+    {
+      "epoch": 1.2917828222487722,
+      "grad_norm": 0.3983367622313922,
+      "learning_rate": 4.466400814231987e-06,
+      "loss": 0.6104,
+      "step": 4866
+    },
+    {
+      "epoch": 1.2920483207221558,
+      "grad_norm": 0.44122615558805117,
+      "learning_rate": 4.46618520387494e-06,
+      "loss": 0.5715,
+      "step": 4867
+    },
+    {
+      "epoch": 1.2923138191955397,
+      "grad_norm": 0.39461543050577097,
+      "learning_rate": 4.465969555172468e-06,
+      "loss": 0.6042,
+      "step": 4868
+    },
+    {
+      "epoch": 1.2925793176689235,
+      "grad_norm": 0.4008858405893098,
+      "learning_rate": 4.465753868128775e-06,
+      "loss": 0.5769,
+      "step": 4869
+    },
+    {
+      "epoch": 1.2928448161423072,
+      "grad_norm": 0.41956821905745156,
+      "learning_rate": 4.465538142748069e-06,
+      "loss": 0.5602,
+      "step": 4870
+    },
+    {
+      "epoch": 1.293110314615691,
+      "grad_norm": 0.39954694162905713,
+      "learning_rate": 4.4653223790345575e-06,
+      "loss": 0.6066,
+      "step": 4871
+    },
+    {
+      "epoch": 1.2933758130890747,
+      "grad_norm": 0.392956173024187,
+      "learning_rate": 4.465106576992447e-06,
+      "loss": 0.5755,
+      "step": 4872
+    },
+    {
+      "epoch": 1.2936413115624585,
+      "grad_norm": 0.39229841630418266,
+      "learning_rate": 4.464890736625947e-06,
+      "loss": 0.5887,
+      "step": 4873
+    },
+    {
+      "epoch": 1.2939068100358422,
+      "grad_norm": 0.4020361585980307,
+      "learning_rate": 4.464674857939267e-06,
+      "loss": 0.5925,
+      "step": 4874
+    },
+    {
+      "epoch": 1.294172308509226,
+      "grad_norm": 0.40567876691463367,
+      "learning_rate": 4.464458940936617e-06,
+      "loss": 0.5922,
+      "step": 4875
+    },
+    {
+      "epoch": 1.29443780698261,
+      "grad_norm": 0.4013564549878668,
+      "learning_rate": 4.464242985622207e-06,
+      "loss": 0.6106,
+      "step": 4876
+    },
+    {
+      "epoch": 1.2947033054559935,
+      "grad_norm": 0.4038216727943326,
+      "learning_rate": 4.46402699200025e-06,
+      "loss": 0.5887,
+      "step": 4877
+    },
+    {
+      "epoch": 1.2949688039293774,
+      "grad_norm": 0.3978614122089539,
+      "learning_rate": 4.463810960074958e-06,
+      "loss": 0.6061,
+      "step": 4878
+    },
+    {
+      "epoch": 1.2952343024027613,
+      "grad_norm": 0.3977618233500997,
+      "learning_rate": 4.463594889850543e-06,
+      "loss": 0.6316,
+      "step": 4879
+    },
+    {
+      "epoch": 1.295499800876145,
+      "grad_norm": 0.4218751705702764,
+      "learning_rate": 4.4633787813312216e-06,
+      "loss": 0.6167,
+      "step": 4880
+    },
+    {
+      "epoch": 1.2957652993495288,
+      "grad_norm": 0.4058847872883744,
+      "learning_rate": 4.463162634521206e-06,
+      "loss": 0.6037,
+      "step": 4881
+    },
+    {
+      "epoch": 1.2960307978229126,
+      "grad_norm": 0.42217733066166013,
+      "learning_rate": 4.462946449424713e-06,
+      "loss": 0.5571,
+      "step": 4882
+    },
+    {
+      "epoch": 1.2962962962962963,
+      "grad_norm": 0.4098362634149795,
+      "learning_rate": 4.462730226045957e-06,
+      "loss": 0.5729,
+      "step": 4883
+    },
+    {
+      "epoch": 1.2965617947696801,
+      "grad_norm": 0.4232138768948184,
+      "learning_rate": 4.462513964389157e-06,
+      "loss": 0.5639,
+      "step": 4884
+    },
+    {
+      "epoch": 1.2968272932430638,
+      "grad_norm": 0.39686602524610415,
+      "learning_rate": 4.462297664458529e-06,
+      "loss": 0.6282,
+      "step": 4885
+    },
+    {
+      "epoch": 1.2970927917164476,
+      "grad_norm": 0.4138818690910932,
+      "learning_rate": 4.462081326258292e-06,
+      "loss": 0.5851,
+      "step": 4886
+    },
+    {
+      "epoch": 1.2973582901898313,
+      "grad_norm": 0.4194989149152848,
+      "learning_rate": 4.4618649497926655e-06,
+      "loss": 0.581,
+      "step": 4887
+    },
+    {
+      "epoch": 1.2976237886632151,
+      "grad_norm": 0.40558023529005793,
+      "learning_rate": 4.461648535065869e-06,
+      "loss": 0.5991,
+      "step": 4888
+    },
+    {
+      "epoch": 1.297889287136599,
+      "grad_norm": 0.41331403939075745,
+      "learning_rate": 4.461432082082123e-06,
+      "loss": 0.6123,
+      "step": 4889
+    },
+    {
+      "epoch": 1.2981547856099827,
+      "grad_norm": 0.4174597695987421,
+      "learning_rate": 4.461215590845649e-06,
+      "loss": 0.6309,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2984202840833665,
+      "grad_norm": 0.4111726198608028,
+      "learning_rate": 4.460999061360668e-06,
+      "loss": 0.6279,
+      "step": 4891
+    },
+    {
+      "epoch": 1.2986857825567504,
+      "grad_norm": 0.43028637961062544,
+      "learning_rate": 4.460782493631405e-06,
+      "loss": 0.6217,
+      "step": 4892
+    },
+    {
+      "epoch": 1.298951281030134,
+      "grad_norm": 0.41265769455231094,
+      "learning_rate": 4.460565887662083e-06,
+      "loss": 0.5778,
+      "step": 4893
+    },
+    {
+      "epoch": 1.2992167795035179,
+      "grad_norm": 0.4202122052221382,
+      "learning_rate": 4.4603492434569255e-06,
+      "loss": 0.5451,
+      "step": 4894
+    },
+    {
+      "epoch": 1.2994822779769017,
+      "grad_norm": 0.43001248873158865,
+      "learning_rate": 4.460132561020158e-06,
+      "loss": 0.5641,
+      "step": 4895
+    },
+    {
+      "epoch": 1.2997477764502854,
+      "grad_norm": 0.4372149661039394,
+      "learning_rate": 4.459915840356006e-06,
+      "loss": 0.5949,
+      "step": 4896
+    },
+    {
+      "epoch": 1.3000132749236692,
+      "grad_norm": 0.4021836231967466,
+      "learning_rate": 4.4596990814686966e-06,
+      "loss": 0.5682,
+      "step": 4897
+    },
+    {
+      "epoch": 1.300278773397053,
+      "grad_norm": 0.4057342123872635,
+      "learning_rate": 4.459482284362457e-06,
+      "loss": 0.5711,
+      "step": 4898
+    },
+    {
+      "epoch": 1.3005442718704368,
+      "grad_norm": 0.41845207031620923,
+      "learning_rate": 4.459265449041516e-06,
+      "loss": 0.5919,
+      "step": 4899
+    },
+    {
+      "epoch": 1.3008097703438204,
+      "grad_norm": 0.4048730667788804,
+      "learning_rate": 4.4590485755101e-06,
+      "loss": 0.5988,
+      "step": 4900
+    },
+    {
+      "epoch": 1.3010752688172043,
+      "grad_norm": 0.3970598634688899,
+      "learning_rate": 4.458831663772441e-06,
+      "loss": 0.5992,
+      "step": 4901
+    },
+    {
+      "epoch": 1.3013407672905881,
+      "grad_norm": 0.3907628960114115,
+      "learning_rate": 4.458614713832768e-06,
+      "loss": 0.6216,
+      "step": 4902
+    },
+    {
+      "epoch": 1.3016062657639718,
+      "grad_norm": 0.40237661235405225,
+      "learning_rate": 4.458397725695314e-06,
+      "loss": 0.5517,
+      "step": 4903
+    },
+    {
+      "epoch": 1.3018717642373556,
+      "grad_norm": 0.4042654953387462,
+      "learning_rate": 4.458180699364308e-06,
+      "loss": 0.5938,
+      "step": 4904
+    },
+    {
+      "epoch": 1.3021372627107395,
+      "grad_norm": 0.4148204077744537,
+      "learning_rate": 4.457963634843985e-06,
+      "loss": 0.6185,
+      "step": 4905
+    },
+    {
+      "epoch": 1.3024027611841231,
+      "grad_norm": 0.4157144792489109,
+      "learning_rate": 4.457746532138576e-06,
+      "loss": 0.5866,
+      "step": 4906
+    },
+    {
+      "epoch": 1.302668259657507,
+      "grad_norm": 0.4103066982869338,
+      "learning_rate": 4.457529391252317e-06,
+      "loss": 0.5722,
+      "step": 4907
+    },
+    {
+      "epoch": 1.3029337581308909,
+      "grad_norm": 0.3958075038029031,
+      "learning_rate": 4.457312212189442e-06,
+      "loss": 0.6335,
+      "step": 4908
+    },
+    {
+      "epoch": 1.3031992566042745,
+      "grad_norm": 0.4118756023541191,
+      "learning_rate": 4.457094994954186e-06,
+      "loss": 0.5822,
+      "step": 4909
+    },
+    {
+      "epoch": 1.3034647550776584,
+      "grad_norm": 0.40902517113752285,
+      "learning_rate": 4.456877739550787e-06,
+      "loss": 0.6125,
+      "step": 4910
+    },
+    {
+      "epoch": 1.3037302535510422,
+      "grad_norm": 0.4024089878684346,
+      "learning_rate": 4.456660445983479e-06,
+      "loss": 0.6175,
+      "step": 4911
+    },
+    {
+      "epoch": 1.3039957520244259,
+      "grad_norm": 0.4007104200870501,
+      "learning_rate": 4.456443114256503e-06,
+      "loss": 0.5798,
+      "step": 4912
+    },
+    {
+      "epoch": 1.3042612504978095,
+      "grad_norm": 0.40492257298770873,
+      "learning_rate": 4.4562257443740955e-06,
+      "loss": 0.5647,
+      "step": 4913
+    },
+    {
+      "epoch": 1.3045267489711934,
+      "grad_norm": 0.41318478381146634,
+      "learning_rate": 4.456008336340497e-06,
+      "loss": 0.6189,
+      "step": 4914
+    },
+    {
+      "epoch": 1.3047922474445772,
+      "grad_norm": 0.42449931760978266,
+      "learning_rate": 4.455790890159946e-06,
+      "loss": 0.5838,
+      "step": 4915
+    },
+    {
+      "epoch": 1.3050577459179609,
+      "grad_norm": 0.41283085691953864,
+      "learning_rate": 4.455573405836684e-06,
+      "loss": 0.6002,
+      "step": 4916
+    },
+    {
+      "epoch": 1.3053232443913447,
+      "grad_norm": 0.42067519724069574,
+      "learning_rate": 4.455355883374953e-06,
+      "loss": 0.5906,
+      "step": 4917
+    },
+    {
+      "epoch": 1.3055887428647286,
+      "grad_norm": 0.3966990758298903,
+      "learning_rate": 4.455138322778995e-06,
+      "loss": 0.6081,
+      "step": 4918
+    },
+    {
+      "epoch": 1.3058542413381122,
+      "grad_norm": 0.39692347523808585,
+      "learning_rate": 4.454920724053052e-06,
+      "loss": 0.5817,
+      "step": 4919
+    },
+    {
+      "epoch": 1.306119739811496,
+      "grad_norm": 0.40854683486509485,
+      "learning_rate": 4.454703087201368e-06,
+      "loss": 0.581,
+      "step": 4920
+    },
+    {
+      "epoch": 1.30638523828488,
+      "grad_norm": 0.40358569446086373,
+      "learning_rate": 4.4544854122281885e-06,
+      "loss": 0.5978,
+      "step": 4921
+    },
+    {
+      "epoch": 1.3066507367582636,
+      "grad_norm": 0.39539448833647567,
+      "learning_rate": 4.454267699137759e-06,
+      "loss": 0.5813,
+      "step": 4922
+    },
+    {
+      "epoch": 1.3069162352316475,
+      "grad_norm": 0.4178932170650583,
+      "learning_rate": 4.454049947934323e-06,
+      "loss": 0.602,
+      "step": 4923
+    },
+    {
+      "epoch": 1.3071817337050313,
+      "grad_norm": 0.4087833790590675,
+      "learning_rate": 4.453832158622129e-06,
+      "loss": 0.58,
+      "step": 4924
+    },
+    {
+      "epoch": 1.307447232178415,
+      "grad_norm": 0.4315773571172743,
+      "learning_rate": 4.4536143312054246e-06,
+      "loss": 0.6021,
+      "step": 4925
+    },
+    {
+      "epoch": 1.3077127306517988,
+      "grad_norm": 0.43304482672326006,
+      "learning_rate": 4.453396465688458e-06,
+      "loss": 0.5878,
+      "step": 4926
+    },
+    {
+      "epoch": 1.3079782291251825,
+      "grad_norm": 0.40648341954478884,
+      "learning_rate": 4.4531785620754765e-06,
+      "loss": 0.6036,
+      "step": 4927
+    },
+    {
+      "epoch": 1.3082437275985663,
+      "grad_norm": 0.4104017402563507,
+      "learning_rate": 4.452960620370731e-06,
+      "loss": 0.5634,
+      "step": 4928
+    },
+    {
+      "epoch": 1.30850922607195,
+      "grad_norm": 0.4215156051170924,
+      "learning_rate": 4.452742640578472e-06,
+      "loss": 0.5798,
+      "step": 4929
+    },
+    {
+      "epoch": 1.3087747245453338,
+      "grad_norm": 0.4293369353442428,
+      "learning_rate": 4.45252462270295e-06,
+      "loss": 0.595,
+      "step": 4930
+    },
+    {
+      "epoch": 1.3090402230187177,
+      "grad_norm": 0.40963147419785767,
+      "learning_rate": 4.452306566748418e-06,
+      "loss": 0.5794,
+      "step": 4931
+    },
+    {
+      "epoch": 1.3093057214921013,
+      "grad_norm": 0.40718378492322854,
+      "learning_rate": 4.452088472719127e-06,
+      "loss": 0.5882,
+      "step": 4932
+    },
+    {
+      "epoch": 1.3095712199654852,
+      "grad_norm": 0.41386040653446554,
+      "learning_rate": 4.451870340619331e-06,
+      "loss": 0.5912,
+      "step": 4933
+    },
+    {
+      "epoch": 1.309836718438869,
+      "grad_norm": 0.43873713573961204,
+      "learning_rate": 4.4516521704532855e-06,
+      "loss": 0.5983,
+      "step": 4934
+    },
+    {
+      "epoch": 1.3101022169122527,
+      "grad_norm": 0.41559467475908973,
+      "learning_rate": 4.4514339622252434e-06,
+      "loss": 0.5746,
+      "step": 4935
+    },
+    {
+      "epoch": 1.3103677153856366,
+      "grad_norm": 0.3894541434040422,
+      "learning_rate": 4.451215715939461e-06,
+      "loss": 0.5968,
+      "step": 4936
+    },
+    {
+      "epoch": 1.3106332138590204,
+      "grad_norm": 0.4036207110914407,
+      "learning_rate": 4.450997431600196e-06,
+      "loss": 0.5797,
+      "step": 4937
+    },
+    {
+      "epoch": 1.310898712332404,
+      "grad_norm": 0.41484750551669364,
+      "learning_rate": 4.450779109211703e-06,
+      "loss": 0.5812,
+      "step": 4938
+    },
+    {
+      "epoch": 1.311164210805788,
+      "grad_norm": 0.4103433857138673,
+      "learning_rate": 4.450560748778242e-06,
+      "loss": 0.5606,
+      "step": 4939
+    },
+    {
+      "epoch": 1.3114297092791716,
+      "grad_norm": 0.40268329385275714,
+      "learning_rate": 4.45034235030407e-06,
+      "loss": 0.566,
+      "step": 4940
+    },
+    {
+      "epoch": 1.3116952077525554,
+      "grad_norm": 0.4195615678342612,
+      "learning_rate": 4.450123913793446e-06,
+      "loss": 0.5839,
+      "step": 4941
+    },
+    {
+      "epoch": 1.311960706225939,
+      "grad_norm": 0.4075118580038868,
+      "learning_rate": 4.449905439250632e-06,
+      "loss": 0.5909,
+      "step": 4942
+    },
+    {
+      "epoch": 1.312226204699323,
+      "grad_norm": 0.4077988711345542,
+      "learning_rate": 4.449686926679888e-06,
+      "loss": 0.584,
+      "step": 4943
+    },
+    {
+      "epoch": 1.3124917031727068,
+      "grad_norm": 0.39843862997833607,
+      "learning_rate": 4.449468376085475e-06,
+      "loss": 0.6331,
+      "step": 4944
+    },
+    {
+      "epoch": 1.3127572016460904,
+      "grad_norm": 0.40758777115073547,
+      "learning_rate": 4.449249787471655e-06,
+      "loss": 0.6179,
+      "step": 4945
+    },
+    {
+      "epoch": 1.3130227001194743,
+      "grad_norm": 0.4170488038253358,
+      "learning_rate": 4.4490311608426915e-06,
+      "loss": 0.6002,
+      "step": 4946
+    },
+    {
+      "epoch": 1.3132881985928582,
+      "grad_norm": 0.4137659366462854,
+      "learning_rate": 4.448812496202849e-06,
+      "loss": 0.5762,
+      "step": 4947
+    },
+    {
+      "epoch": 1.3135536970662418,
+      "grad_norm": 0.41173212727099967,
+      "learning_rate": 4.448593793556391e-06,
+      "loss": 0.5936,
+      "step": 4948
+    },
+    {
+      "epoch": 1.3138191955396257,
+      "grad_norm": 0.41220219762569343,
+      "learning_rate": 4.448375052907583e-06,
+      "loss": 0.6073,
+      "step": 4949
+    },
+    {
+      "epoch": 1.3140846940130095,
+      "grad_norm": 0.41164444333141387,
+      "learning_rate": 4.448156274260692e-06,
+      "loss": 0.5952,
+      "step": 4950
+    },
+    {
+      "epoch": 1.3143501924863932,
+      "grad_norm": 0.41336092340573943,
+      "learning_rate": 4.447937457619982e-06,
+      "loss": 0.5786,
+      "step": 4951
+    },
+    {
+      "epoch": 1.314615690959777,
+      "grad_norm": 0.4132578017693204,
+      "learning_rate": 4.447718602989723e-06,
+      "loss": 0.5486,
+      "step": 4952
+    },
+    {
+      "epoch": 1.3148811894331607,
+      "grad_norm": 0.40171552999094035,
+      "learning_rate": 4.447499710374184e-06,
+      "loss": 0.5641,
+      "step": 4953
+    },
+    {
+      "epoch": 1.3151466879065445,
+      "grad_norm": 0.3990453125849152,
+      "learning_rate": 4.447280779777629e-06,
+      "loss": 0.5577,
+      "step": 4954
+    },
+    {
+      "epoch": 1.3154121863799282,
+      "grad_norm": 0.4196073706130943,
+      "learning_rate": 4.447061811204334e-06,
+      "loss": 0.5552,
+      "step": 4955
+    },
+    {
+      "epoch": 1.315677684853312,
+      "grad_norm": 0.40245043178713646,
+      "learning_rate": 4.446842804658566e-06,
+      "loss": 0.5824,
+      "step": 4956
+    },
+    {
+      "epoch": 1.315943183326696,
+      "grad_norm": 0.4074182932135894,
+      "learning_rate": 4.446623760144595e-06,
+      "loss": 0.5883,
+      "step": 4957
+    },
+    {
+      "epoch": 1.3162086818000795,
+      "grad_norm": 0.41514255806711453,
+      "learning_rate": 4.446404677666696e-06,
+      "loss": 0.5774,
+      "step": 4958
+    },
+    {
+      "epoch": 1.3164741802734634,
+      "grad_norm": 0.41226552621481066,
+      "learning_rate": 4.446185557229139e-06,
+      "loss": 0.6007,
+      "step": 4959
+    },
+    {
+      "epoch": 1.3167396787468473,
+      "grad_norm": 0.4174646477327288,
+      "learning_rate": 4.445966398836199e-06,
+      "loss": 0.5885,
+      "step": 4960
+    },
+    {
+      "epoch": 1.317005177220231,
+      "grad_norm": 0.4175165796360001,
+      "learning_rate": 4.44574720249215e-06,
+      "loss": 0.6381,
+      "step": 4961
+    },
+    {
+      "epoch": 1.3172706756936148,
+      "grad_norm": 0.4179511888457544,
+      "learning_rate": 4.445527968201267e-06,
+      "loss": 0.5988,
+      "step": 4962
+    },
+    {
+      "epoch": 1.3175361741669986,
+      "grad_norm": 0.4145385412129239,
+      "learning_rate": 4.445308695967824e-06,
+      "loss": 0.5728,
+      "step": 4963
+    },
+    {
+      "epoch": 1.3178016726403823,
+      "grad_norm": 0.42404094443982626,
+      "learning_rate": 4.445089385796099e-06,
+      "loss": 0.6115,
+      "step": 4964
+    },
+    {
+      "epoch": 1.3180671711137661,
+      "grad_norm": 0.4031924679634836,
+      "learning_rate": 4.444870037690368e-06,
+      "loss": 0.5817,
+      "step": 4965
+    },
+    {
+      "epoch": 1.31833266958715,
+      "grad_norm": 0.40885603633534584,
+      "learning_rate": 4.444650651654911e-06,
+      "loss": 0.6037,
+      "step": 4966
+    },
+    {
+      "epoch": 1.3185981680605336,
+      "grad_norm": 0.45789135321367413,
+      "learning_rate": 4.444431227694004e-06,
+      "loss": 0.6013,
+      "step": 4967
+    },
+    {
+      "epoch": 1.3188636665339175,
+      "grad_norm": 0.409986144554615,
+      "learning_rate": 4.444211765811927e-06,
+      "loss": 0.5808,
+      "step": 4968
+    },
+    {
+      "epoch": 1.3191291650073012,
+      "grad_norm": 0.4080317409770435,
+      "learning_rate": 4.443992266012961e-06,
+      "loss": 0.6323,
+      "step": 4969
+    },
+    {
+      "epoch": 1.319394663480685,
+      "grad_norm": 0.40962624701312145,
+      "learning_rate": 4.443772728301385e-06,
+      "loss": 0.5617,
+      "step": 4970
+    },
+    {
+      "epoch": 1.3196601619540687,
+      "grad_norm": 0.4184075206212948,
+      "learning_rate": 4.4435531526814825e-06,
+      "loss": 0.5714,
+      "step": 4971
+    },
+    {
+      "epoch": 1.3199256604274525,
+      "grad_norm": 0.40541038067775204,
+      "learning_rate": 4.443333539157535e-06,
+      "loss": 0.6065,
+      "step": 4972
+    },
+    {
+      "epoch": 1.3201911589008364,
+      "grad_norm": 0.3979943404378569,
+      "learning_rate": 4.443113887733824e-06,
+      "loss": 0.561,
+      "step": 4973
+    },
+    {
+      "epoch": 1.32045665737422,
+      "grad_norm": 0.40468721169834854,
+      "learning_rate": 4.442894198414636e-06,
+      "loss": 0.5863,
+      "step": 4974
+    },
+    {
+      "epoch": 1.3207221558476039,
+      "grad_norm": 0.41360532501049635,
+      "learning_rate": 4.4426744712042545e-06,
+      "loss": 0.5477,
+      "step": 4975
+    },
+    {
+      "epoch": 1.3209876543209877,
+      "grad_norm": 0.3980882953464089,
+      "learning_rate": 4.442454706106963e-06,
+      "loss": 0.5848,
+      "step": 4976
+    },
+    {
+      "epoch": 1.3212531527943714,
+      "grad_norm": 0.39465377928410006,
+      "learning_rate": 4.4422349031270495e-06,
+      "loss": 0.5909,
+      "step": 4977
+    },
+    {
+      "epoch": 1.3215186512677553,
+      "grad_norm": 0.41398691594786546,
+      "learning_rate": 4.4420150622688e-06,
+      "loss": 0.6067,
+      "step": 4978
+    },
+    {
+      "epoch": 1.3217841497411391,
+      "grad_norm": 0.4169460212147881,
+      "learning_rate": 4.441795183536502e-06,
+      "loss": 0.5964,
+      "step": 4979
+    },
+    {
+      "epoch": 1.3220496482145228,
+      "grad_norm": 0.4032021127670633,
+      "learning_rate": 4.441575266934444e-06,
+      "loss": 0.5971,
+      "step": 4980
+    },
+    {
+      "epoch": 1.3223151466879066,
+      "grad_norm": 0.4114418771584214,
+      "learning_rate": 4.4413553124669126e-06,
+      "loss": 0.5583,
+      "step": 4981
+    },
+    {
+      "epoch": 1.3225806451612903,
+      "grad_norm": 0.39697519598617365,
+      "learning_rate": 4.441135320138201e-06,
+      "loss": 0.59,
+      "step": 4982
+    },
+    {
+      "epoch": 1.3228461436346741,
+      "grad_norm": 0.3990293758275327,
+      "learning_rate": 4.440915289952598e-06,
+      "loss": 0.6205,
+      "step": 4983
+    },
+    {
+      "epoch": 1.3231116421080578,
+      "grad_norm": 0.4020691031854009,
+      "learning_rate": 4.440695221914394e-06,
+      "loss": 0.5803,
+      "step": 4984
+    },
+    {
+      "epoch": 1.3233771405814416,
+      "grad_norm": 0.4102046584337675,
+      "learning_rate": 4.440475116027882e-06,
+      "loss": 0.595,
+      "step": 4985
+    },
+    {
+      "epoch": 1.3236426390548255,
+      "grad_norm": 0.39274017988004467,
+      "learning_rate": 4.440254972297354e-06,
+      "loss": 0.5708,
+      "step": 4986
+    },
+    {
+      "epoch": 1.3239081375282091,
+      "grad_norm": 0.3920600022470943,
+      "learning_rate": 4.4400347907271024e-06,
+      "loss": 0.5414,
+      "step": 4987
+    },
+    {
+      "epoch": 1.324173636001593,
+      "grad_norm": 0.38421972792100684,
+      "learning_rate": 4.4398145713214234e-06,
+      "loss": 0.5774,
+      "step": 4988
+    },
+    {
+      "epoch": 1.3244391344749769,
+      "grad_norm": 0.41529836990010277,
+      "learning_rate": 4.439594314084611e-06,
+      "loss": 0.5571,
+      "step": 4989
+    },
+    {
+      "epoch": 1.3247046329483605,
+      "grad_norm": 0.4060323850989439,
+      "learning_rate": 4.43937401902096e-06,
+      "loss": 0.6157,
+      "step": 4990
+    },
+    {
+      "epoch": 1.3249701314217444,
+      "grad_norm": 0.40949474269586444,
+      "learning_rate": 4.439153686134768e-06,
+      "loss": 0.5963,
+      "step": 4991
+    },
+    {
+      "epoch": 1.3252356298951282,
+      "grad_norm": 0.4029031408444375,
+      "learning_rate": 4.438933315430331e-06,
+      "loss": 0.5712,
+      "step": 4992
+    },
+    {
+      "epoch": 1.3255011283685119,
+      "grad_norm": 0.4043024787791668,
+      "learning_rate": 4.438712906911946e-06,
+      "loss": 0.5846,
+      "step": 4993
+    },
+    {
+      "epoch": 1.3257666268418957,
+      "grad_norm": 0.40919937957356767,
+      "learning_rate": 4.438492460583913e-06,
+      "loss": 0.602,
+      "step": 4994
+    },
+    {
+      "epoch": 1.3260321253152794,
+      "grad_norm": 0.41866175490329155,
+      "learning_rate": 4.438271976450531e-06,
+      "loss": 0.5969,
+      "step": 4995
+    },
+    {
+      "epoch": 1.3262976237886632,
+      "grad_norm": 0.4040783534475649,
+      "learning_rate": 4.4380514545161e-06,
+      "loss": 0.649,
+      "step": 4996
+    },
+    {
+      "epoch": 1.3265631222620469,
+      "grad_norm": 0.4070636098216918,
+      "learning_rate": 4.437830894784921e-06,
+      "loss": 0.5958,
+      "step": 4997
+    },
+    {
+      "epoch": 1.3268286207354307,
+      "grad_norm": 0.405725607606133,
+      "learning_rate": 4.437610297261294e-06,
+      "loss": 0.6041,
+      "step": 4998
+    },
+    {
+      "epoch": 1.3270941192088146,
+      "grad_norm": 0.41216796382968224,
+      "learning_rate": 4.437389661949522e-06,
+      "loss": 0.5959,
+      "step": 4999
+    },
+    {
+      "epoch": 1.3273596176821982,
+      "grad_norm": 0.42580833447783706,
+      "learning_rate": 4.4371689888539085e-06,
+      "loss": 0.5918,
+      "step": 5000
+    },
+    {
+      "epoch": 1.327625116155582,
+      "grad_norm": 0.4022898238282934,
+      "learning_rate": 4.436948277978756e-06,
+      "loss": 0.5387,
+      "step": 5001
+    },
+    {
+      "epoch": 1.327890614628966,
+      "grad_norm": 0.4243664722750062,
+      "learning_rate": 4.4367275293283706e-06,
+      "loss": 0.6421,
+      "step": 5002
+    },
+    {
+      "epoch": 1.3281561131023496,
+      "grad_norm": 0.40452875420533846,
+      "learning_rate": 4.436506742907056e-06,
+      "loss": 0.6072,
+      "step": 5003
+    },
+    {
+      "epoch": 1.3284216115757335,
+      "grad_norm": 0.4006932644712524,
+      "learning_rate": 4.436285918719118e-06,
+      "loss": 0.6004,
+      "step": 5004
+    },
+    {
+      "epoch": 1.3286871100491173,
+      "grad_norm": 0.41270472684735265,
+      "learning_rate": 4.4360650567688646e-06,
+      "loss": 0.6129,
+      "step": 5005
+    },
+    {
+      "epoch": 1.328952608522501,
+      "grad_norm": 0.41761017984070553,
+      "learning_rate": 4.435844157060602e-06,
+      "loss": 0.5781,
+      "step": 5006
+    },
+    {
+      "epoch": 1.3292181069958848,
+      "grad_norm": 0.4014641418120313,
+      "learning_rate": 4.435623219598638e-06,
+      "loss": 0.6019,
+      "step": 5007
+    },
+    {
+      "epoch": 1.3294836054692685,
+      "grad_norm": 0.4158333117016189,
+      "learning_rate": 4.435402244387284e-06,
+      "loss": 0.5686,
+      "step": 5008
+    },
+    {
+      "epoch": 1.3297491039426523,
+      "grad_norm": 0.39589892370960766,
+      "learning_rate": 4.435181231430845e-06,
+      "loss": 0.6102,
+      "step": 5009
+    },
+    {
+      "epoch": 1.330014602416036,
+      "grad_norm": 0.41354345770561685,
+      "learning_rate": 4.4349601807336354e-06,
+      "loss": 0.6024,
+      "step": 5010
+    },
+    {
+      "epoch": 1.3302801008894198,
+      "grad_norm": 0.39990629929221533,
+      "learning_rate": 4.434739092299964e-06,
+      "loss": 0.5865,
+      "step": 5011
+    },
+    {
+      "epoch": 1.3305455993628037,
+      "grad_norm": 0.3960230668392498,
+      "learning_rate": 4.434517966134143e-06,
+      "loss": 0.6226,
+      "step": 5012
+    },
+    {
+      "epoch": 1.3308110978361873,
+      "grad_norm": 0.4163560407375716,
+      "learning_rate": 4.4342968022404855e-06,
+      "loss": 0.599,
+      "step": 5013
+    },
+    {
+      "epoch": 1.3310765963095712,
+      "grad_norm": 0.410894748073801,
+      "learning_rate": 4.4340756006233045e-06,
+      "loss": 0.5679,
+      "step": 5014
+    },
+    {
+      "epoch": 1.331342094782955,
+      "grad_norm": 0.39352752492836857,
+      "learning_rate": 4.4338543612869146e-06,
+      "loss": 0.5641,
+      "step": 5015
+    },
+    {
+      "epoch": 1.3316075932563387,
+      "grad_norm": 0.41061439661062427,
+      "learning_rate": 4.433633084235629e-06,
+      "loss": 0.5435,
+      "step": 5016
+    },
+    {
+      "epoch": 1.3318730917297226,
+      "grad_norm": 0.39333883740131115,
+      "learning_rate": 4.433411769473763e-06,
+      "loss": 0.5366,
+      "step": 5017
+    },
+    {
+      "epoch": 1.3321385902031064,
+      "grad_norm": 0.3977195000961838,
+      "learning_rate": 4.433190417005635e-06,
+      "loss": 0.5777,
+      "step": 5018
+    },
+    {
+      "epoch": 1.33240408867649,
+      "grad_norm": 0.39975439408090524,
+      "learning_rate": 4.4329690268355606e-06,
+      "loss": 0.6084,
+      "step": 5019
+    },
+    {
+      "epoch": 1.332669587149874,
+      "grad_norm": 0.40345153781685766,
+      "learning_rate": 4.432747598967857e-06,
+      "loss": 0.561,
+      "step": 5020
+    },
+    {
+      "epoch": 1.3329350856232578,
+      "grad_norm": 0.4031615410430819,
+      "learning_rate": 4.432526133406843e-06,
+      "loss": 0.5165,
+      "step": 5021
+    },
+    {
+      "epoch": 1.3332005840966414,
+      "grad_norm": 0.4126099355864836,
+      "learning_rate": 4.4323046301568374e-06,
+      "loss": 0.5437,
+      "step": 5022
+    },
+    {
+      "epoch": 1.3334660825700253,
+      "grad_norm": 0.42318155647259525,
+      "learning_rate": 4.432083089222161e-06,
+      "loss": 0.589,
+      "step": 5023
+    },
+    {
+      "epoch": 1.333731581043409,
+      "grad_norm": 0.42853829159879614,
+      "learning_rate": 4.431861510607134e-06,
+      "loss": 0.5966,
+      "step": 5024
+    },
+    {
+      "epoch": 1.3339970795167928,
+      "grad_norm": 0.4108754486994434,
+      "learning_rate": 4.431639894316076e-06,
+      "loss": 0.5928,
+      "step": 5025
+    },
+    {
+      "epoch": 1.3342625779901764,
+      "grad_norm": 0.4028532830464729,
+      "learning_rate": 4.431418240353312e-06,
+      "loss": 0.6179,
+      "step": 5026
+    },
+    {
+      "epoch": 1.3345280764635603,
+      "grad_norm": 0.41811562947826264,
+      "learning_rate": 4.431196548723164e-06,
+      "loss": 0.6042,
+      "step": 5027
+    },
+    {
+      "epoch": 1.3347935749369442,
+      "grad_norm": 0.4118665234306777,
+      "learning_rate": 4.430974819429954e-06,
+      "loss": 0.5901,
+      "step": 5028
+    },
+    {
+      "epoch": 1.3350590734103278,
+      "grad_norm": 0.4049832608414355,
+      "learning_rate": 4.430753052478007e-06,
+      "loss": 0.5831,
+      "step": 5029
+    },
+    {
+      "epoch": 1.3353245718837117,
+      "grad_norm": 0.41570376109355267,
+      "learning_rate": 4.430531247871649e-06,
+      "loss": 0.6395,
+      "step": 5030
+    },
+    {
+      "epoch": 1.3355900703570955,
+      "grad_norm": 0.4134855453681411,
+      "learning_rate": 4.430309405615204e-06,
+      "loss": 0.6081,
+      "step": 5031
+    },
+    {
+      "epoch": 1.3358555688304792,
+      "grad_norm": 0.4195997411148105,
+      "learning_rate": 4.430087525712999e-06,
+      "loss": 0.6097,
+      "step": 5032
+    },
+    {
+      "epoch": 1.336121067303863,
+      "grad_norm": 0.39883310142252193,
+      "learning_rate": 4.429865608169363e-06,
+      "loss": 0.5666,
+      "step": 5033
+    },
+    {
+      "epoch": 1.336386565777247,
+      "grad_norm": 0.3978944358633572,
+      "learning_rate": 4.4296436529886216e-06,
+      "loss": 0.5701,
+      "step": 5034
+    },
+    {
+      "epoch": 1.3366520642506305,
+      "grad_norm": 0.4281636841120658,
+      "learning_rate": 4.429421660175105e-06,
+      "loss": 0.5882,
+      "step": 5035
+    },
+    {
+      "epoch": 1.3369175627240144,
+      "grad_norm": 0.3975590630072414,
+      "learning_rate": 4.429199629733141e-06,
+      "loss": 0.5765,
+      "step": 5036
+    },
+    {
+      "epoch": 1.337183061197398,
+      "grad_norm": 0.4062196905582909,
+      "learning_rate": 4.428977561667062e-06,
+      "loss": 0.5673,
+      "step": 5037
+    },
+    {
+      "epoch": 1.337448559670782,
+      "grad_norm": 0.39482637350780103,
+      "learning_rate": 4.428755455981196e-06,
+      "loss": 0.5857,
+      "step": 5038
+    },
+    {
+      "epoch": 1.3377140581441656,
+      "grad_norm": 0.4134638636681826,
+      "learning_rate": 4.428533312679878e-06,
+      "loss": 0.6154,
+      "step": 5039
+    },
+    {
+      "epoch": 1.3379795566175494,
+      "grad_norm": 0.4026259070920607,
+      "learning_rate": 4.4283111317674375e-06,
+      "loss": 0.6218,
+      "step": 5040
+    },
+    {
+      "epoch": 1.3382450550909333,
+      "grad_norm": 0.40904344092777983,
+      "learning_rate": 4.4280889132482085e-06,
+      "loss": 0.5972,
+      "step": 5041
+    },
+    {
+      "epoch": 1.338510553564317,
+      "grad_norm": 0.42274703242304396,
+      "learning_rate": 4.427866657126525e-06,
+      "loss": 0.6346,
+      "step": 5042
+    },
+    {
+      "epoch": 1.3387760520377008,
+      "grad_norm": 0.39950268110867015,
+      "learning_rate": 4.427644363406721e-06,
+      "loss": 0.5905,
+      "step": 5043
+    },
+    {
+      "epoch": 1.3390415505110846,
+      "grad_norm": 0.40976211167144805,
+      "learning_rate": 4.4274220320931334e-06,
+      "loss": 0.6348,
+      "step": 5044
+    },
+    {
+      "epoch": 1.3393070489844683,
+      "grad_norm": 0.4193727489853488,
+      "learning_rate": 4.427199663190097e-06,
+      "loss": 0.5917,
+      "step": 5045
+    },
+    {
+      "epoch": 1.3395725474578521,
+      "grad_norm": 0.41160967623833183,
+      "learning_rate": 4.426977256701948e-06,
+      "loss": 0.5844,
+      "step": 5046
+    },
+    {
+      "epoch": 1.339838045931236,
+      "grad_norm": 0.4356531804738442,
+      "learning_rate": 4.426754812633024e-06,
+      "loss": 0.597,
+      "step": 5047
+    },
+    {
+      "epoch": 1.3401035444046197,
+      "grad_norm": 0.4324168009310734,
+      "learning_rate": 4.426532330987664e-06,
+      "loss": 0.5648,
+      "step": 5048
+    },
+    {
+      "epoch": 1.3403690428780035,
+      "grad_norm": 0.39502224872010605,
+      "learning_rate": 4.4263098117702075e-06,
+      "loss": 0.5821,
+      "step": 5049
+    },
+    {
+      "epoch": 1.3406345413513872,
+      "grad_norm": 0.4133142833839653,
+      "learning_rate": 4.426087254984992e-06,
+      "loss": 0.6146,
+      "step": 5050
+    },
+    {
+      "epoch": 1.340900039824771,
+      "grad_norm": 0.45347721868925234,
+      "learning_rate": 4.42586466063636e-06,
+      "loss": 0.5783,
+      "step": 5051
+    },
+    {
+      "epoch": 1.3411655382981547,
+      "grad_norm": 0.44019193437927234,
+      "learning_rate": 4.425642028728652e-06,
+      "loss": 0.6004,
+      "step": 5052
+    },
+    {
+      "epoch": 1.3414310367715385,
+      "grad_norm": 0.41238131588091564,
+      "learning_rate": 4.4254193592662085e-06,
+      "loss": 0.6087,
+      "step": 5053
+    },
+    {
+      "epoch": 1.3416965352449224,
+      "grad_norm": 0.4083800972192628,
+      "learning_rate": 4.425196652253374e-06,
+      "loss": 0.5711,
+      "step": 5054
+    },
+    {
+      "epoch": 1.341962033718306,
+      "grad_norm": 0.4251367749526427,
+      "learning_rate": 4.424973907694491e-06,
+      "loss": 0.6202,
+      "step": 5055
+    },
+    {
+      "epoch": 1.3422275321916899,
+      "grad_norm": 0.4134648872232151,
+      "learning_rate": 4.424751125593903e-06,
+      "loss": 0.5515,
+      "step": 5056
+    },
+    {
+      "epoch": 1.3424930306650737,
+      "grad_norm": 0.4391411207469455,
+      "learning_rate": 4.4245283059559565e-06,
+      "loss": 0.6309,
+      "step": 5057
+    },
+    {
+      "epoch": 1.3427585291384574,
+      "grad_norm": 0.4114403926087078,
+      "learning_rate": 4.424305448784995e-06,
+      "loss": 0.5553,
+      "step": 5058
+    },
+    {
+      "epoch": 1.3430240276118413,
+      "grad_norm": 0.48383354083900076,
+      "learning_rate": 4.424082554085366e-06,
+      "loss": 0.5618,
+      "step": 5059
+    },
+    {
+      "epoch": 1.3432895260852251,
+      "grad_norm": 0.4000636094710507,
+      "learning_rate": 4.423859621861417e-06,
+      "loss": 0.5891,
+      "step": 5060
+    },
+    {
+      "epoch": 1.3435550245586088,
+      "grad_norm": 0.40022518286199704,
+      "learning_rate": 4.423636652117494e-06,
+      "loss": 0.5351,
+      "step": 5061
+    },
+    {
+      "epoch": 1.3438205230319926,
+      "grad_norm": 0.4221018952087181,
+      "learning_rate": 4.423413644857947e-06,
+      "loss": 0.5786,
+      "step": 5062
+    },
+    {
+      "epoch": 1.3440860215053765,
+      "grad_norm": 0.4124671380774633,
+      "learning_rate": 4.423190600087124e-06,
+      "loss": 0.6412,
+      "step": 5063
+    },
+    {
+      "epoch": 1.3443515199787601,
+      "grad_norm": 0.4258627339237869,
+      "learning_rate": 4.422967517809377e-06,
+      "loss": 0.6627,
+      "step": 5064
+    },
+    {
+      "epoch": 1.3446170184521438,
+      "grad_norm": 0.4131704741384072,
+      "learning_rate": 4.422744398029054e-06,
+      "loss": 0.6202,
+      "step": 5065
+    },
+    {
+      "epoch": 1.3448825169255276,
+      "grad_norm": 0.39329717059421276,
+      "learning_rate": 4.422521240750507e-06,
+      "loss": 0.584,
+      "step": 5066
+    },
+    {
+      "epoch": 1.3451480153989115,
+      "grad_norm": 0.41269799981370214,
+      "learning_rate": 4.42229804597809e-06,
+      "loss": 0.5749,
+      "step": 5067
+    },
+    {
+      "epoch": 1.3454135138722951,
+      "grad_norm": 0.4212370407606538,
+      "learning_rate": 4.422074813716153e-06,
+      "loss": 0.579,
+      "step": 5068
+    },
+    {
+      "epoch": 1.345679012345679,
+      "grad_norm": 0.4210303616633975,
+      "learning_rate": 4.421851543969052e-06,
+      "loss": 0.5801,
+      "step": 5069
+    },
+    {
+      "epoch": 1.3459445108190629,
+      "grad_norm": 0.38243691527271945,
+      "learning_rate": 4.42162823674114e-06,
+      "loss": 0.576,
+      "step": 5070
+    },
+    {
+      "epoch": 1.3462100092924465,
+      "grad_norm": 0.4131632383889058,
+      "learning_rate": 4.421404892036773e-06,
+      "loss": 0.556,
+      "step": 5071
+    },
+    {
+      "epoch": 1.3464755077658304,
+      "grad_norm": 0.39017119518826265,
+      "learning_rate": 4.421181509860306e-06,
+      "loss": 0.6197,
+      "step": 5072
+    },
+    {
+      "epoch": 1.3467410062392142,
+      "grad_norm": 0.40189399659200686,
+      "learning_rate": 4.420958090216095e-06,
+      "loss": 0.5918,
+      "step": 5073
+    },
+    {
+      "epoch": 1.3470065047125979,
+      "grad_norm": 0.4143724895933705,
+      "learning_rate": 4.420734633108499e-06,
+      "loss": 0.5741,
+      "step": 5074
+    },
+    {
+      "epoch": 1.3472720031859817,
+      "grad_norm": 0.4080873981595074,
+      "learning_rate": 4.420511138541874e-06,
+      "loss": 0.5559,
+      "step": 5075
+    },
+    {
+      "epoch": 1.3475375016593656,
+      "grad_norm": 0.4056024414237227,
+      "learning_rate": 4.4202876065205805e-06,
+      "loss": 0.5988,
+      "step": 5076
+    },
+    {
+      "epoch": 1.3478030001327492,
+      "grad_norm": 0.403401918130962,
+      "learning_rate": 4.4200640370489764e-06,
+      "loss": 0.6079,
+      "step": 5077
+    },
+    {
+      "epoch": 1.348068498606133,
+      "grad_norm": 0.39958445292036104,
+      "learning_rate": 4.419840430131422e-06,
+      "loss": 0.6197,
+      "step": 5078
+    },
+    {
+      "epoch": 1.3483339970795167,
+      "grad_norm": 0.41459340641791764,
+      "learning_rate": 4.419616785772279e-06,
+      "loss": 0.5922,
+      "step": 5079
+    },
+    {
+      "epoch": 1.3485994955529006,
+      "grad_norm": 0.43002389749473247,
+      "learning_rate": 4.4193931039759095e-06,
+      "loss": 0.6362,
+      "step": 5080
+    },
+    {
+      "epoch": 1.3488649940262842,
+      "grad_norm": 0.4050876883830849,
+      "learning_rate": 4.419169384746675e-06,
+      "loss": 0.624,
+      "step": 5081
+    },
+    {
+      "epoch": 1.349130492499668,
+      "grad_norm": 0.4051989824459654,
+      "learning_rate": 4.418945628088937e-06,
+      "loss": 0.5596,
+      "step": 5082
+    },
+    {
+      "epoch": 1.349395990973052,
+      "grad_norm": 0.3948948661453604,
+      "learning_rate": 4.418721834007062e-06,
+      "loss": 0.5623,
+      "step": 5083
+    },
+    {
+      "epoch": 1.3496614894464356,
+      "grad_norm": 0.39304582750769324,
+      "learning_rate": 4.418498002505413e-06,
+      "loss": 0.5945,
+      "step": 5084
+    },
+    {
+      "epoch": 1.3499269879198195,
+      "grad_norm": 0.4044159936110484,
+      "learning_rate": 4.418274133588356e-06,
+      "loss": 0.5872,
+      "step": 5085
+    },
+    {
+      "epoch": 1.3501924863932033,
+      "grad_norm": 0.4075175743729038,
+      "learning_rate": 4.418050227260257e-06,
+      "loss": 0.5534,
+      "step": 5086
+    },
+    {
+      "epoch": 1.350457984866587,
+      "grad_norm": 0.40925555442412337,
+      "learning_rate": 4.417826283525481e-06,
+      "loss": 0.5932,
+      "step": 5087
+    },
+    {
+      "epoch": 1.3507234833399708,
+      "grad_norm": 0.38691779675751636,
+      "learning_rate": 4.417602302388398e-06,
+      "loss": 0.5976,
+      "step": 5088
+    },
+    {
+      "epoch": 1.3509889818133547,
+      "grad_norm": 0.41979258821905574,
+      "learning_rate": 4.417378283853374e-06,
+      "loss": 0.5946,
+      "step": 5089
+    },
+    {
+      "epoch": 1.3512544802867383,
+      "grad_norm": 0.39871698278174694,
+      "learning_rate": 4.4171542279247795e-06,
+      "loss": 0.5715,
+      "step": 5090
+    },
+    {
+      "epoch": 1.3515199787601222,
+      "grad_norm": 0.40728571312735545,
+      "learning_rate": 4.416930134606984e-06,
+      "loss": 0.62,
+      "step": 5091
+    },
+    {
+      "epoch": 1.3517854772335058,
+      "grad_norm": 0.4130698278443358,
+      "learning_rate": 4.416706003904357e-06,
+      "loss": 0.5696,
+      "step": 5092
+    },
+    {
+      "epoch": 1.3520509757068897,
+      "grad_norm": 0.459513594535085,
+      "learning_rate": 4.41648183582127e-06,
+      "loss": 0.5671,
+      "step": 5093
+    },
+    {
+      "epoch": 1.3523164741802733,
+      "grad_norm": 0.39243996224763655,
+      "learning_rate": 4.416257630362095e-06,
+      "loss": 0.588,
+      "step": 5094
+    },
+    {
+      "epoch": 1.3525819726536572,
+      "grad_norm": 0.42102018332415797,
+      "learning_rate": 4.416033387531204e-06,
+      "loss": 0.5552,
+      "step": 5095
+    },
+    {
+      "epoch": 1.352847471127041,
+      "grad_norm": 0.4231024813055664,
+      "learning_rate": 4.4158091073329714e-06,
+      "loss": 0.6053,
+      "step": 5096
+    },
+    {
+      "epoch": 1.3531129696004247,
+      "grad_norm": 0.41444342980929527,
+      "learning_rate": 4.415584789771769e-06,
+      "loss": 0.5758,
+      "step": 5097
+    },
+    {
+      "epoch": 1.3533784680738086,
+      "grad_norm": 0.410161314351242,
+      "learning_rate": 4.415360434851975e-06,
+      "loss": 0.615,
+      "step": 5098
+    },
+    {
+      "epoch": 1.3536439665471924,
+      "grad_norm": 0.4054365677030745,
+      "learning_rate": 4.4151360425779615e-06,
+      "loss": 0.6066,
+      "step": 5099
+    },
+    {
+      "epoch": 1.353909465020576,
+      "grad_norm": 0.43503654516844664,
+      "learning_rate": 4.414911612954107e-06,
+      "loss": 0.5993,
+      "step": 5100
+    },
+    {
+      "epoch": 1.35417496349396,
+      "grad_norm": 0.4332953614553006,
+      "learning_rate": 4.414687145984787e-06,
+      "loss": 0.6001,
+      "step": 5101
+    },
+    {
+      "epoch": 1.3544404619673438,
+      "grad_norm": 0.407582719266609,
+      "learning_rate": 4.41446264167438e-06,
+      "loss": 0.6012,
+      "step": 5102
+    },
+    {
+      "epoch": 1.3547059604407274,
+      "grad_norm": 0.4008306934284124,
+      "learning_rate": 4.414238100027264e-06,
+      "loss": 0.5881,
+      "step": 5103
+    },
+    {
+      "epoch": 1.3549714589141113,
+      "grad_norm": 0.4208095853075956,
+      "learning_rate": 4.414013521047817e-06,
+      "loss": 0.5485,
+      "step": 5104
+    },
+    {
+      "epoch": 1.355236957387495,
+      "grad_norm": 0.43405413338415744,
+      "learning_rate": 4.413788904740422e-06,
+      "loss": 0.6103,
+      "step": 5105
+    },
+    {
+      "epoch": 1.3555024558608788,
+      "grad_norm": 0.4024702046819154,
+      "learning_rate": 4.4135642511094566e-06,
+      "loss": 0.5844,
+      "step": 5106
+    },
+    {
+      "epoch": 1.3557679543342624,
+      "grad_norm": 0.3964452290433775,
+      "learning_rate": 4.413339560159303e-06,
+      "loss": 0.5837,
+      "step": 5107
+    },
+    {
+      "epoch": 1.3560334528076463,
+      "grad_norm": 0.4488612374596151,
+      "learning_rate": 4.413114831894344e-06,
+      "loss": 0.553,
+      "step": 5108
+    },
+    {
+      "epoch": 1.3562989512810302,
+      "grad_norm": 0.42676583857580014,
+      "learning_rate": 4.412890066318961e-06,
+      "loss": 0.5698,
+      "step": 5109
+    },
+    {
+      "epoch": 1.3565644497544138,
+      "grad_norm": 0.39670193834007356,
+      "learning_rate": 4.412665263437538e-06,
+      "loss": 0.5554,
+      "step": 5110
+    },
+    {
+      "epoch": 1.3568299482277977,
+      "grad_norm": 0.38910047758471217,
+      "learning_rate": 4.41244042325446e-06,
+      "loss": 0.6195,
+      "step": 5111
+    },
+    {
+      "epoch": 1.3570954467011815,
+      "grad_norm": 0.4736034088407191,
+      "learning_rate": 4.412215545774111e-06,
+      "loss": 0.5712,
+      "step": 5112
+    },
+    {
+      "epoch": 1.3573609451745652,
+      "grad_norm": 0.43170925827530643,
+      "learning_rate": 4.4119906310008765e-06,
+      "loss": 0.628,
+      "step": 5113
+    },
+    {
+      "epoch": 1.357626443647949,
+      "grad_norm": 0.41175074062244466,
+      "learning_rate": 4.411765678939144e-06,
+      "loss": 0.5905,
+      "step": 5114
+    },
+    {
+      "epoch": 1.357891942121333,
+      "grad_norm": 0.41182863150885163,
+      "learning_rate": 4.4115406895933e-06,
+      "loss": 0.6135,
+      "step": 5115
+    },
+    {
+      "epoch": 1.3581574405947165,
+      "grad_norm": 0.39681398882634666,
+      "learning_rate": 4.411315662967732e-06,
+      "loss": 0.5708,
+      "step": 5116
+    },
+    {
+      "epoch": 1.3584229390681004,
+      "grad_norm": 0.4029371644670817,
+      "learning_rate": 4.411090599066829e-06,
+      "loss": 0.6264,
+      "step": 5117
+    },
+    {
+      "epoch": 1.3586884375414843,
+      "grad_norm": 0.42340763942115595,
+      "learning_rate": 4.410865497894979e-06,
+      "loss": 0.6171,
+      "step": 5118
+    },
+    {
+      "epoch": 1.358953936014868,
+      "grad_norm": 0.41988373333397255,
+      "learning_rate": 4.410640359456575e-06,
+      "loss": 0.573,
+      "step": 5119
+    },
+    {
+      "epoch": 1.3592194344882516,
+      "grad_norm": 0.40488482694374284,
+      "learning_rate": 4.410415183756005e-06,
+      "loss": 0.5962,
+      "step": 5120
+    },
+    {
+      "epoch": 1.3594849329616354,
+      "grad_norm": 0.4153589898806118,
+      "learning_rate": 4.4101899707976615e-06,
+      "loss": 0.5836,
+      "step": 5121
+    },
+    {
+      "epoch": 1.3597504314350193,
+      "grad_norm": 0.4108910256101074,
+      "learning_rate": 4.409964720585937e-06,
+      "loss": 0.578,
+      "step": 5122
+    },
+    {
+      "epoch": 1.360015929908403,
+      "grad_norm": 0.40191295331559923,
+      "learning_rate": 4.409739433125223e-06,
+      "loss": 0.5719,
+      "step": 5123
+    },
+    {
+      "epoch": 1.3602814283817868,
+      "grad_norm": 0.4098072327025475,
+      "learning_rate": 4.4095141084199155e-06,
+      "loss": 0.5847,
+      "step": 5124
+    },
+    {
+      "epoch": 1.3605469268551706,
+      "grad_norm": 0.4142253200508539,
+      "learning_rate": 4.409288746474407e-06,
+      "loss": 0.6066,
+      "step": 5125
+    },
+    {
+      "epoch": 1.3608124253285543,
+      "grad_norm": 0.39511170463658823,
+      "learning_rate": 4.409063347293093e-06,
+      "loss": 0.5791,
+      "step": 5126
+    },
+    {
+      "epoch": 1.3610779238019381,
+      "grad_norm": 0.423496425849535,
+      "learning_rate": 4.40883791088037e-06,
+      "loss": 0.6097,
+      "step": 5127
+    },
+    {
+      "epoch": 1.361343422275322,
+      "grad_norm": 0.40114354325139334,
+      "learning_rate": 4.408612437240633e-06,
+      "loss": 0.5823,
+      "step": 5128
+    },
+    {
+      "epoch": 1.3616089207487057,
+      "grad_norm": 0.40411237735566014,
+      "learning_rate": 4.408386926378282e-06,
+      "loss": 0.6187,
+      "step": 5129
+    },
+    {
+      "epoch": 1.3618744192220895,
+      "grad_norm": 0.4004335352018841,
+      "learning_rate": 4.4081613782977125e-06,
+      "loss": 0.6027,
+      "step": 5130
+    },
+    {
+      "epoch": 1.3621399176954734,
+      "grad_norm": 0.41121731232989656,
+      "learning_rate": 4.407935793003324e-06,
+      "loss": 0.6067,
+      "step": 5131
+    },
+    {
+      "epoch": 1.362405416168857,
+      "grad_norm": 0.40741065321744613,
+      "learning_rate": 4.407710170499517e-06,
+      "loss": 0.5947,
+      "step": 5132
+    },
+    {
+      "epoch": 1.3626709146422409,
+      "grad_norm": 0.40662582732184704,
+      "learning_rate": 4.40748451079069e-06,
+      "loss": 0.5945,
+      "step": 5133
+    },
+    {
+      "epoch": 1.3629364131156245,
+      "grad_norm": 0.40149074464183626,
+      "learning_rate": 4.407258813881244e-06,
+      "loss": 0.5979,
+      "step": 5134
+    },
+    {
+      "epoch": 1.3632019115890084,
+      "grad_norm": 0.39889904614270105,
+      "learning_rate": 4.407033079775583e-06,
+      "loss": 0.6051,
+      "step": 5135
+    },
+    {
+      "epoch": 1.363467410062392,
+      "grad_norm": 0.4001566859196351,
+      "learning_rate": 4.4068073084781075e-06,
+      "loss": 0.5835,
+      "step": 5136
+    },
+    {
+      "epoch": 1.363732908535776,
+      "grad_norm": 0.4043102328551009,
+      "learning_rate": 4.40658149999322e-06,
+      "loss": 0.5957,
+      "step": 5137
+    },
+    {
+      "epoch": 1.3639984070091598,
+      "grad_norm": 0.3980156532579141,
+      "learning_rate": 4.406355654325325e-06,
+      "loss": 0.5925,
+      "step": 5138
+    },
+    {
+      "epoch": 1.3642639054825434,
+      "grad_norm": 0.41100335983671743,
+      "learning_rate": 4.4061297714788275e-06,
+      "loss": 0.6124,
+      "step": 5139
+    },
+    {
+      "epoch": 1.3645294039559273,
+      "grad_norm": 0.40137178260222345,
+      "learning_rate": 4.405903851458133e-06,
+      "loss": 0.6285,
+      "step": 5140
+    },
+    {
+      "epoch": 1.3647949024293111,
+      "grad_norm": 0.41479490289086995,
+      "learning_rate": 4.405677894267646e-06,
+      "loss": 0.5526,
+      "step": 5141
+    },
+    {
+      "epoch": 1.3650604009026948,
+      "grad_norm": 0.40607094222167517,
+      "learning_rate": 4.405451899911775e-06,
+      "loss": 0.5956,
+      "step": 5142
+    },
+    {
+      "epoch": 1.3653258993760786,
+      "grad_norm": 0.4124953942465148,
+      "learning_rate": 4.405225868394926e-06,
+      "loss": 0.6073,
+      "step": 5143
+    },
+    {
+      "epoch": 1.3655913978494625,
+      "grad_norm": 0.4122150155113502,
+      "learning_rate": 4.404999799721508e-06,
+      "loss": 0.5639,
+      "step": 5144
+    },
+    {
+      "epoch": 1.3658568963228461,
+      "grad_norm": 0.4088992678580032,
+      "learning_rate": 4.4047736938959295e-06,
+      "loss": 0.5794,
+      "step": 5145
+    },
+    {
+      "epoch": 1.36612239479623,
+      "grad_norm": 0.39888140573982966,
+      "learning_rate": 4.404547550922601e-06,
+      "loss": 0.6099,
+      "step": 5146
+    },
+    {
+      "epoch": 1.3663878932696136,
+      "grad_norm": 0.4042769177880839,
+      "learning_rate": 4.404321370805931e-06,
+      "loss": 0.5796,
+      "step": 5147
+    },
+    {
+      "epoch": 1.3666533917429975,
+      "grad_norm": 0.3997113478974893,
+      "learning_rate": 4.404095153550332e-06,
+      "loss": 0.5715,
+      "step": 5148
+    },
+    {
+      "epoch": 1.3669188902163811,
+      "grad_norm": 0.40237218998174806,
+      "learning_rate": 4.4038688991602165e-06,
+      "loss": 0.5729,
+      "step": 5149
+    },
+    {
+      "epoch": 1.367184388689765,
+      "grad_norm": 0.3963382812760406,
+      "learning_rate": 4.403642607639995e-06,
+      "loss": 0.5365,
+      "step": 5150
+    },
+    {
+      "epoch": 1.3674498871631489,
+      "grad_norm": 0.4024347914345515,
+      "learning_rate": 4.403416278994081e-06,
+      "loss": 0.5909,
+      "step": 5151
+    },
+    {
+      "epoch": 1.3677153856365325,
+      "grad_norm": 0.41489547589206194,
+      "learning_rate": 4.403189913226891e-06,
+      "loss": 0.6142,
+      "step": 5152
+    },
+    {
+      "epoch": 1.3679808841099164,
+      "grad_norm": 0.4144344201689968,
+      "learning_rate": 4.402963510342837e-06,
+      "loss": 0.562,
+      "step": 5153
+    },
+    {
+      "epoch": 1.3682463825833002,
+      "grad_norm": 0.4035158287419469,
+      "learning_rate": 4.402737070346335e-06,
+      "loss": 0.5604,
+      "step": 5154
+    },
+    {
+      "epoch": 1.3685118810566839,
+      "grad_norm": 0.4018852196902614,
+      "learning_rate": 4.402510593241802e-06,
+      "loss": 0.6218,
+      "step": 5155
+    },
+    {
+      "epoch": 1.3687773795300677,
+      "grad_norm": 0.4359921044630224,
+      "learning_rate": 4.402284079033654e-06,
+      "loss": 0.5748,
+      "step": 5156
+    },
+    {
+      "epoch": 1.3690428780034516,
+      "grad_norm": 0.4237437493228621,
+      "learning_rate": 4.402057527726309e-06,
+      "loss": 0.5933,
+      "step": 5157
+    },
+    {
+      "epoch": 1.3693083764768352,
+      "grad_norm": 0.41906126999846177,
+      "learning_rate": 4.4018309393241845e-06,
+      "loss": 0.594,
+      "step": 5158
+    },
+    {
+      "epoch": 1.369573874950219,
+      "grad_norm": 0.40130761670913456,
+      "learning_rate": 4.401604313831701e-06,
+      "loss": 0.6085,
+      "step": 5159
+    },
+    {
+      "epoch": 1.3698393734236027,
+      "grad_norm": 0.4375434043708791,
+      "learning_rate": 4.4013776512532775e-06,
+      "loss": 0.5894,
+      "step": 5160
+    },
+    {
+      "epoch": 1.3701048718969866,
+      "grad_norm": 0.44560980520280286,
+      "learning_rate": 4.401150951593334e-06,
+      "loss": 0.5901,
+      "step": 5161
+    },
+    {
+      "epoch": 1.3703703703703702,
+      "grad_norm": 0.41522144859204435,
+      "learning_rate": 4.400924214856292e-06,
+      "loss": 0.6097,
+      "step": 5162
+    },
+    {
+      "epoch": 1.370635868843754,
+      "grad_norm": 0.41000841916376307,
+      "learning_rate": 4.4006974410465745e-06,
+      "loss": 0.6107,
+      "step": 5163
+    },
+    {
+      "epoch": 1.370901367317138,
+      "grad_norm": 0.43472481653103207,
+      "learning_rate": 4.400470630168602e-06,
+      "loss": 0.5831,
+      "step": 5164
+    },
+    {
+      "epoch": 1.3711668657905216,
+      "grad_norm": 0.4860283206483714,
+      "learning_rate": 4.4002437822267995e-06,
+      "loss": 0.5841,
+      "step": 5165
+    },
+    {
+      "epoch": 1.3714323642639055,
+      "grad_norm": 0.4065099474621937,
+      "learning_rate": 4.400016897225591e-06,
+      "loss": 0.5832,
+      "step": 5166
+    },
+    {
+      "epoch": 1.3716978627372893,
+      "grad_norm": 0.39128508150702834,
+      "learning_rate": 4.399789975169401e-06,
+      "loss": 0.593,
+      "step": 5167
+    },
+    {
+      "epoch": 1.371963361210673,
+      "grad_norm": 0.4068634251361246,
+      "learning_rate": 4.399563016062654e-06,
+      "loss": 0.6045,
+      "step": 5168
+    },
+    {
+      "epoch": 1.3722288596840568,
+      "grad_norm": 0.45924541259772284,
+      "learning_rate": 4.399336019909778e-06,
+      "loss": 0.5575,
+      "step": 5169
+    },
+    {
+      "epoch": 1.3724943581574407,
+      "grad_norm": 0.41493654768151256,
+      "learning_rate": 4.399108986715199e-06,
+      "loss": 0.6114,
+      "step": 5170
+    },
+    {
+      "epoch": 1.3727598566308243,
+      "grad_norm": 0.4035868879952453,
+      "learning_rate": 4.398881916483345e-06,
+      "loss": 0.5826,
+      "step": 5171
+    },
+    {
+      "epoch": 1.3730253551042082,
+      "grad_norm": 0.3975413196561161,
+      "learning_rate": 4.398654809218644e-06,
+      "loss": 0.606,
+      "step": 5172
+    },
+    {
+      "epoch": 1.373290853577592,
+      "grad_norm": 0.4245881673087071,
+      "learning_rate": 4.398427664925526e-06,
+      "loss": 0.6202,
+      "step": 5173
+    },
+    {
+      "epoch": 1.3735563520509757,
+      "grad_norm": 0.43707577390240415,
+      "learning_rate": 4.398200483608421e-06,
+      "loss": 0.5882,
+      "step": 5174
+    },
+    {
+      "epoch": 1.3738218505243596,
+      "grad_norm": 0.4221363096323231,
+      "learning_rate": 4.397973265271758e-06,
+      "loss": 0.594,
+      "step": 5175
+    },
+    {
+      "epoch": 1.3740873489977432,
+      "grad_norm": 0.3882554117741956,
+      "learning_rate": 4.39774600991997e-06,
+      "loss": 0.5557,
+      "step": 5176
+    },
+    {
+      "epoch": 1.374352847471127,
+      "grad_norm": 0.4028669981011816,
+      "learning_rate": 4.397518717557487e-06,
+      "loss": 0.5828,
+      "step": 5177
+    },
+    {
+      "epoch": 1.3746183459445107,
+      "grad_norm": 0.41465818363684664,
+      "learning_rate": 4.397291388188744e-06,
+      "loss": 0.6095,
+      "step": 5178
+    },
+    {
+      "epoch": 1.3748838444178946,
+      "grad_norm": 0.4068593304795824,
+      "learning_rate": 4.397064021818174e-06,
+      "loss": 0.5993,
+      "step": 5179
+    },
+    {
+      "epoch": 1.3751493428912784,
+      "grad_norm": 0.395238036353193,
+      "learning_rate": 4.39683661845021e-06,
+      "loss": 0.601,
+      "step": 5180
+    },
+    {
+      "epoch": 1.375414841364662,
+      "grad_norm": 0.40823737926322445,
+      "learning_rate": 4.396609178089289e-06,
+      "loss": 0.6143,
+      "step": 5181
+    },
+    {
+      "epoch": 1.375680339838046,
+      "grad_norm": 0.41417551719357265,
+      "learning_rate": 4.3963817007398435e-06,
+      "loss": 0.592,
+      "step": 5182
+    },
+    {
+      "epoch": 1.3759458383114298,
+      "grad_norm": 0.41741590668108636,
+      "learning_rate": 4.396154186406313e-06,
+      "loss": 0.5865,
+      "step": 5183
+    },
+    {
+      "epoch": 1.3762113367848134,
+      "grad_norm": 0.3911179426916859,
+      "learning_rate": 4.395926635093133e-06,
+      "loss": 0.5533,
+      "step": 5184
+    },
+    {
+      "epoch": 1.3764768352581973,
+      "grad_norm": 0.3999646422309801,
+      "learning_rate": 4.395699046804741e-06,
+      "loss": 0.5747,
+      "step": 5185
+    },
+    {
+      "epoch": 1.3767423337315812,
+      "grad_norm": 0.4150982840876382,
+      "learning_rate": 4.3954714215455765e-06,
+      "loss": 0.5868,
+      "step": 5186
+    },
+    {
+      "epoch": 1.3770078322049648,
+      "grad_norm": 0.4121907759175896,
+      "learning_rate": 4.395243759320078e-06,
+      "loss": 0.5915,
+      "step": 5187
+    },
+    {
+      "epoch": 1.3772733306783487,
+      "grad_norm": 0.40075637999370006,
+      "learning_rate": 4.3950160601326865e-06,
+      "loss": 0.5486,
+      "step": 5188
+    },
+    {
+      "epoch": 1.3775388291517323,
+      "grad_norm": 0.3960970869949558,
+      "learning_rate": 4.394788323987842e-06,
+      "loss": 0.5909,
+      "step": 5189
+    },
+    {
+      "epoch": 1.3778043276251162,
+      "grad_norm": 0.39499631804329505,
+      "learning_rate": 4.394560550889986e-06,
+      "loss": 0.6211,
+      "step": 5190
+    },
+    {
+      "epoch": 1.3780698260984998,
+      "grad_norm": 0.42576540200857177,
+      "learning_rate": 4.3943327408435595e-06,
+      "loss": 0.5977,
+      "step": 5191
+    },
+    {
+      "epoch": 1.3783353245718837,
+      "grad_norm": 0.40123813178607043,
+      "learning_rate": 4.394104893853007e-06,
+      "loss": 0.5716,
+      "step": 5192
+    },
+    {
+      "epoch": 1.3786008230452675,
+      "grad_norm": 0.4084780525420066,
+      "learning_rate": 4.393877009922772e-06,
+      "loss": 0.5905,
+      "step": 5193
+    },
+    {
+      "epoch": 1.3788663215186512,
+      "grad_norm": 0.41111851101974783,
+      "learning_rate": 4.393649089057298e-06,
+      "loss": 0.5996,
+      "step": 5194
+    },
+    {
+      "epoch": 1.379131819992035,
+      "grad_norm": 0.40484709527646406,
+      "learning_rate": 4.39342113126103e-06,
+      "loss": 0.5698,
+      "step": 5195
+    },
+    {
+      "epoch": 1.379397318465419,
+      "grad_norm": 0.402256232823181,
+      "learning_rate": 4.393193136538414e-06,
+      "loss": 0.6086,
+      "step": 5196
+    },
+    {
+      "epoch": 1.3796628169388025,
+      "grad_norm": 0.409708999293741,
+      "learning_rate": 4.392965104893897e-06,
+      "loss": 0.561,
+      "step": 5197
+    },
+    {
+      "epoch": 1.3799283154121864,
+      "grad_norm": 0.40512371843060374,
+      "learning_rate": 4.3927370363319255e-06,
+      "loss": 0.5754,
+      "step": 5198
+    },
+    {
+      "epoch": 1.3801938138855703,
+      "grad_norm": 0.4159106102704585,
+      "learning_rate": 4.392508930856948e-06,
+      "loss": 0.5572,
+      "step": 5199
+    },
+    {
+      "epoch": 1.380459312358954,
+      "grad_norm": 0.40693072266653496,
+      "learning_rate": 4.392280788473412e-06,
+      "loss": 0.6428,
+      "step": 5200
+    },
+    {
+      "epoch": 1.3807248108323378,
+      "grad_norm": 0.43367558769797493,
+      "learning_rate": 4.392052609185768e-06,
+      "loss": 0.5888,
+      "step": 5201
+    },
+    {
+      "epoch": 1.3809903093057214,
+      "grad_norm": 0.41551843956692247,
+      "learning_rate": 4.391824392998466e-06,
+      "loss": 0.6278,
+      "step": 5202
+    },
+    {
+      "epoch": 1.3812558077791053,
+      "grad_norm": 0.4195867693713215,
+      "learning_rate": 4.391596139915954e-06,
+      "loss": 0.5902,
+      "step": 5203
+    },
+    {
+      "epoch": 1.381521306252489,
+      "grad_norm": 0.41489556331257904,
+      "learning_rate": 4.391367849942689e-06,
+      "loss": 0.5602,
+      "step": 5204
+    },
+    {
+      "epoch": 1.3817868047258728,
+      "grad_norm": 0.3963284122015101,
+      "learning_rate": 4.391139523083118e-06,
+      "loss": 0.6444,
+      "step": 5205
+    },
+    {
+      "epoch": 1.3820523031992566,
+      "grad_norm": 0.40405803538589546,
+      "learning_rate": 4.390911159341697e-06,
+      "loss": 0.5545,
+      "step": 5206
+    },
+    {
+      "epoch": 1.3823178016726403,
+      "grad_norm": 0.42129643175926523,
+      "learning_rate": 4.390682758722879e-06,
+      "loss": 0.5969,
+      "step": 5207
+    },
+    {
+      "epoch": 1.3825833001460242,
+      "grad_norm": 0.40080826826494015,
+      "learning_rate": 4.390454321231116e-06,
+      "loss": 0.5926,
+      "step": 5208
+    },
+    {
+      "epoch": 1.382848798619408,
+      "grad_norm": 0.40771326198367613,
+      "learning_rate": 4.390225846870867e-06,
+      "loss": 0.5779,
+      "step": 5209
+    },
+    {
+      "epoch": 1.3831142970927917,
+      "grad_norm": 0.4083610487619787,
+      "learning_rate": 4.389997335646586e-06,
+      "loss": 0.6063,
+      "step": 5210
+    },
+    {
+      "epoch": 1.3833797955661755,
+      "grad_norm": 0.40602251783352133,
+      "learning_rate": 4.389768787562728e-06,
+      "loss": 0.6117,
+      "step": 5211
+    },
+    {
+      "epoch": 1.3836452940395594,
+      "grad_norm": 0.397016564639049,
+      "learning_rate": 4.3895402026237525e-06,
+      "loss": 0.6045,
+      "step": 5212
+    },
+    {
+      "epoch": 1.383910792512943,
+      "grad_norm": 0.398031226496274,
+      "learning_rate": 4.389311580834116e-06,
+      "loss": 0.6,
+      "step": 5213
+    },
+    {
+      "epoch": 1.3841762909863269,
+      "grad_norm": 0.41573413135455856,
+      "learning_rate": 4.389082922198279e-06,
+      "loss": 0.5528,
+      "step": 5214
+    },
+    {
+      "epoch": 1.3844417894597105,
+      "grad_norm": 0.40069920513836227,
+      "learning_rate": 4.388854226720699e-06,
+      "loss": 0.6154,
+      "step": 5215
+    },
+    {
+      "epoch": 1.3847072879330944,
+      "grad_norm": 0.411186195683949,
+      "learning_rate": 4.388625494405837e-06,
+      "loss": 0.612,
+      "step": 5216
+    },
+    {
+      "epoch": 1.384972786406478,
+      "grad_norm": 0.42290654487713875,
+      "learning_rate": 4.388396725258154e-06,
+      "loss": 0.6032,
+      "step": 5217
+    },
+    {
+      "epoch": 1.385238284879862,
+      "grad_norm": 0.4324741418431151,
+      "learning_rate": 4.388167919282111e-06,
+      "loss": 0.621,
+      "step": 5218
+    },
+    {
+      "epoch": 1.3855037833532458,
+      "grad_norm": 0.40589506391927377,
+      "learning_rate": 4.387939076482171e-06,
+      "loss": 0.5726,
+      "step": 5219
+    },
+    {
+      "epoch": 1.3857692818266294,
+      "grad_norm": 0.40004561454762105,
+      "learning_rate": 4.387710196862796e-06,
+      "loss": 0.5743,
+      "step": 5220
+    },
+    {
+      "epoch": 1.3860347803000133,
+      "grad_norm": 0.4185177741539191,
+      "learning_rate": 4.387481280428449e-06,
+      "loss": 0.6067,
+      "step": 5221
+    },
+    {
+      "epoch": 1.3863002787733971,
+      "grad_norm": 0.39380742461438356,
+      "learning_rate": 4.387252327183598e-06,
+      "loss": 0.6017,
+      "step": 5222
+    },
+    {
+      "epoch": 1.3865657772467808,
+      "grad_norm": 0.3992993598117144,
+      "learning_rate": 4.387023337132705e-06,
+      "loss": 0.5929,
+      "step": 5223
+    },
+    {
+      "epoch": 1.3868312757201646,
+      "grad_norm": 0.4190835212476741,
+      "learning_rate": 4.386794310280236e-06,
+      "loss": 0.6207,
+      "step": 5224
+    },
+    {
+      "epoch": 1.3870967741935485,
+      "grad_norm": 0.41231246584060016,
+      "learning_rate": 4.386565246630659e-06,
+      "loss": 0.5732,
+      "step": 5225
+    },
+    {
+      "epoch": 1.3873622726669321,
+      "grad_norm": 0.3981583509387078,
+      "learning_rate": 4.3863361461884406e-06,
+      "loss": 0.5723,
+      "step": 5226
+    },
+    {
+      "epoch": 1.387627771140316,
+      "grad_norm": 0.4045155093797978,
+      "learning_rate": 4.386107008958048e-06,
+      "loss": 0.5672,
+      "step": 5227
+    },
+    {
+      "epoch": 1.3878932696136999,
+      "grad_norm": 0.406821326535879,
+      "learning_rate": 4.385877834943952e-06,
+      "loss": 0.5763,
+      "step": 5228
+    },
+    {
+      "epoch": 1.3881587680870835,
+      "grad_norm": 0.40400137811210296,
+      "learning_rate": 4.38564862415062e-06,
+      "loss": 0.5535,
+      "step": 5229
+    },
+    {
+      "epoch": 1.3884242665604674,
+      "grad_norm": 0.4306187064563657,
+      "learning_rate": 4.385419376582523e-06,
+      "loss": 0.5789,
+      "step": 5230
+    },
+    {
+      "epoch": 1.388689765033851,
+      "grad_norm": 0.4099761592932349,
+      "learning_rate": 4.385190092244132e-06,
+      "loss": 0.5859,
+      "step": 5231
+    },
+    {
+      "epoch": 1.3889552635072349,
+      "grad_norm": 0.42042871322118036,
+      "learning_rate": 4.384960771139918e-06,
+      "loss": 0.5796,
+      "step": 5232
+    },
+    {
+      "epoch": 1.3892207619806185,
+      "grad_norm": 0.4276004321752673,
+      "learning_rate": 4.384731413274355e-06,
+      "loss": 0.6205,
+      "step": 5233
+    },
+    {
+      "epoch": 1.3894862604540024,
+      "grad_norm": 0.41051195787550787,
+      "learning_rate": 4.3845020186519134e-06,
+      "loss": 0.5801,
+      "step": 5234
+    },
+    {
+      "epoch": 1.3897517589273862,
+      "grad_norm": 0.41765827688835866,
+      "learning_rate": 4.384272587277069e-06,
+      "loss": 0.6001,
+      "step": 5235
+    },
+    {
+      "epoch": 1.3900172574007699,
+      "grad_norm": 0.41303481892769406,
+      "learning_rate": 4.384043119154296e-06,
+      "loss": 0.5814,
+      "step": 5236
+    },
+    {
+      "epoch": 1.3902827558741537,
+      "grad_norm": 0.4088262874845921,
+      "learning_rate": 4.383813614288068e-06,
+      "loss": 0.5706,
+      "step": 5237
+    },
+    {
+      "epoch": 1.3905482543475376,
+      "grad_norm": 0.41649243462503616,
+      "learning_rate": 4.383584072682864e-06,
+      "loss": 0.6,
+      "step": 5238
+    },
+    {
+      "epoch": 1.3908137528209212,
+      "grad_norm": 0.39287278721369573,
+      "learning_rate": 4.383354494343157e-06,
+      "loss": 0.6184,
+      "step": 5239
+    },
+    {
+      "epoch": 1.391079251294305,
+      "grad_norm": 0.3986868006834712,
+      "learning_rate": 4.383124879273427e-06,
+      "loss": 0.5915,
+      "step": 5240
+    },
+    {
+      "epoch": 1.391344749767689,
+      "grad_norm": 0.4106235709483954,
+      "learning_rate": 4.3828952274781515e-06,
+      "loss": 0.5924,
+      "step": 5241
+    },
+    {
+      "epoch": 1.3916102482410726,
+      "grad_norm": 0.40646694030420844,
+      "learning_rate": 4.382665538961809e-06,
+      "loss": 0.5819,
+      "step": 5242
+    },
+    {
+      "epoch": 1.3918757467144565,
+      "grad_norm": 0.4112161565798037,
+      "learning_rate": 4.382435813728878e-06,
+      "loss": 0.5801,
+      "step": 5243
+    },
+    {
+      "epoch": 1.39214124518784,
+      "grad_norm": 0.4049221955762099,
+      "learning_rate": 4.38220605178384e-06,
+      "loss": 0.5817,
+      "step": 5244
+    },
+    {
+      "epoch": 1.392406743661224,
+      "grad_norm": 0.40905885815627957,
+      "learning_rate": 4.381976253131176e-06,
+      "loss": 0.5804,
+      "step": 5245
+    },
+    {
+      "epoch": 1.3926722421346076,
+      "grad_norm": 0.40874363693064714,
+      "learning_rate": 4.381746417775366e-06,
+      "loss": 0.5631,
+      "step": 5246
+    },
+    {
+      "epoch": 1.3929377406079915,
+      "grad_norm": 0.4089193276785924,
+      "learning_rate": 4.381516545720895e-06,
+      "loss": 0.5817,
+      "step": 5247
+    },
+    {
+      "epoch": 1.3932032390813753,
+      "grad_norm": 0.43167715992544775,
+      "learning_rate": 4.381286636972244e-06,
+      "loss": 0.5838,
+      "step": 5248
+    },
+    {
+      "epoch": 1.393468737554759,
+      "grad_norm": 0.39500581663931006,
+      "learning_rate": 4.381056691533897e-06,
+      "loss": 0.6011,
+      "step": 5249
+    },
+    {
+      "epoch": 1.3937342360281428,
+      "grad_norm": 0.4028414347896397,
+      "learning_rate": 4.38082670941034e-06,
+      "loss": 0.5703,
+      "step": 5250
+    },
+    {
+      "epoch": 1.3939997345015267,
+      "grad_norm": 0.4136080586054172,
+      "learning_rate": 4.380596690606056e-06,
+      "loss": 0.6054,
+      "step": 5251
+    },
+    {
+      "epoch": 1.3942652329749103,
+      "grad_norm": 0.40149146172991584,
+      "learning_rate": 4.380366635125533e-06,
+      "loss": 0.5902,
+      "step": 5252
+    },
+    {
+      "epoch": 1.3945307314482942,
+      "grad_norm": 0.416163080127752,
+      "learning_rate": 4.380136542973256e-06,
+      "loss": 0.5898,
+      "step": 5253
+    },
+    {
+      "epoch": 1.394796229921678,
+      "grad_norm": 0.396558188045198,
+      "learning_rate": 4.379906414153713e-06,
+      "loss": 0.5887,
+      "step": 5254
+    },
+    {
+      "epoch": 1.3950617283950617,
+      "grad_norm": 0.42481901318735066,
+      "learning_rate": 4.379676248671392e-06,
+      "loss": 0.5691,
+      "step": 5255
+    },
+    {
+      "epoch": 1.3953272268684456,
+      "grad_norm": 0.4145932863931775,
+      "learning_rate": 4.3794460465307825e-06,
+      "loss": 0.5956,
+      "step": 5256
+    },
+    {
+      "epoch": 1.3955927253418292,
+      "grad_norm": 0.3977345376481052,
+      "learning_rate": 4.379215807736373e-06,
+      "loss": 0.5933,
+      "step": 5257
+    },
+    {
+      "epoch": 1.395858223815213,
+      "grad_norm": 0.40879445235316325,
+      "learning_rate": 4.3789855322926545e-06,
+      "loss": 0.598,
+      "step": 5258
+    },
+    {
+      "epoch": 1.3961237222885967,
+      "grad_norm": 0.40717392334343067,
+      "learning_rate": 4.378755220204117e-06,
+      "loss": 0.6068,
+      "step": 5259
+    },
+    {
+      "epoch": 1.3963892207619806,
+      "grad_norm": 0.3869425074525864,
+      "learning_rate": 4.378524871475253e-06,
+      "loss": 0.6149,
+      "step": 5260
+    },
+    {
+      "epoch": 1.3966547192353644,
+      "grad_norm": 0.40452703949006136,
+      "learning_rate": 4.3782944861105545e-06,
+      "loss": 0.5633,
+      "step": 5261
+    },
+    {
+      "epoch": 1.396920217708748,
+      "grad_norm": 0.41214806232486756,
+      "learning_rate": 4.378064064114515e-06,
+      "loss": 0.6269,
+      "step": 5262
+    },
+    {
+      "epoch": 1.397185716182132,
+      "grad_norm": 0.4178515447823688,
+      "learning_rate": 4.377833605491629e-06,
+      "loss": 0.6087,
+      "step": 5263
+    },
+    {
+      "epoch": 1.3974512146555158,
+      "grad_norm": 0.41279575162953963,
+      "learning_rate": 4.377603110246388e-06,
+      "loss": 0.5848,
+      "step": 5264
+    },
+    {
+      "epoch": 1.3977167131288994,
+      "grad_norm": 0.4054764645833888,
+      "learning_rate": 4.37737257838329e-06,
+      "loss": 0.6044,
+      "step": 5265
+    },
+    {
+      "epoch": 1.3979822116022833,
+      "grad_norm": 0.40834561977882916,
+      "learning_rate": 4.3771420099068295e-06,
+      "loss": 0.5615,
+      "step": 5266
+    },
+    {
+      "epoch": 1.3982477100756672,
+      "grad_norm": 0.40490491799432443,
+      "learning_rate": 4.376911404821504e-06,
+      "loss": 0.5945,
+      "step": 5267
+    },
+    {
+      "epoch": 1.3985132085490508,
+      "grad_norm": 0.413022600097997,
+      "learning_rate": 4.376680763131811e-06,
+      "loss": 0.6112,
+      "step": 5268
+    },
+    {
+      "epoch": 1.3987787070224347,
+      "grad_norm": 0.4089796095265697,
+      "learning_rate": 4.376450084842249e-06,
+      "loss": 0.6115,
+      "step": 5269
+    },
+    {
+      "epoch": 1.3990442054958183,
+      "grad_norm": 0.42262412206284095,
+      "learning_rate": 4.376219369957315e-06,
+      "loss": 0.5912,
+      "step": 5270
+    },
+    {
+      "epoch": 1.3993097039692022,
+      "grad_norm": 0.4304380950985587,
+      "learning_rate": 4.3759886184815096e-06,
+      "loss": 0.6099,
+      "step": 5271
+    },
+    {
+      "epoch": 1.3995752024425858,
+      "grad_norm": 0.3979478574613287,
+      "learning_rate": 4.375757830419332e-06,
+      "loss": 0.5859,
+      "step": 5272
+    },
+    {
+      "epoch": 1.3998407009159697,
+      "grad_norm": 0.40454692791838376,
+      "learning_rate": 4.375527005775285e-06,
+      "loss": 0.5864,
+      "step": 5273
+    },
+    {
+      "epoch": 1.4001061993893535,
+      "grad_norm": 0.4405928165143641,
+      "learning_rate": 4.37529614455387e-06,
+      "loss": 0.6055,
+      "step": 5274
+    },
+    {
+      "epoch": 1.4003716978627372,
+      "grad_norm": 0.42149400487582067,
+      "learning_rate": 4.375065246759588e-06,
+      "loss": 0.5874,
+      "step": 5275
+    },
+    {
+      "epoch": 1.400637196336121,
+      "grad_norm": 0.39690124135541516,
+      "learning_rate": 4.3748343123969425e-06,
+      "loss": 0.6018,
+      "step": 5276
+    },
+    {
+      "epoch": 1.400902694809505,
+      "grad_norm": 0.4101053211535445,
+      "learning_rate": 4.3746033414704385e-06,
+      "loss": 0.6074,
+      "step": 5277
+    },
+    {
+      "epoch": 1.4011681932828886,
+      "grad_norm": 0.41938765628301966,
+      "learning_rate": 4.374372333984578e-06,
+      "loss": 0.5746,
+      "step": 5278
+    },
+    {
+      "epoch": 1.4014336917562724,
+      "grad_norm": 0.4139352872300997,
+      "learning_rate": 4.374141289943869e-06,
+      "loss": 0.5996,
+      "step": 5279
+    },
+    {
+      "epoch": 1.4016991902296563,
+      "grad_norm": 0.4168765139329988,
+      "learning_rate": 4.373910209352816e-06,
+      "loss": 0.5917,
+      "step": 5280
+    },
+    {
+      "epoch": 1.40196468870304,
+      "grad_norm": 0.40484974493577286,
+      "learning_rate": 4.373679092215925e-06,
+      "loss": 0.5643,
+      "step": 5281
+    },
+    {
+      "epoch": 1.4022301871764238,
+      "grad_norm": 0.4067886246181902,
+      "learning_rate": 4.3734479385377045e-06,
+      "loss": 0.5952,
+      "step": 5282
+    },
+    {
+      "epoch": 1.4024956856498076,
+      "grad_norm": 0.4071437567241423,
+      "learning_rate": 4.373216748322663e-06,
+      "loss": 0.5963,
+      "step": 5283
+    },
+    {
+      "epoch": 1.4027611841231913,
+      "grad_norm": 0.4130238143307736,
+      "learning_rate": 4.372985521575307e-06,
+      "loss": 0.6022,
+      "step": 5284
+    },
+    {
+      "epoch": 1.4030266825965751,
+      "grad_norm": 0.4035257034397141,
+      "learning_rate": 4.372754258300148e-06,
+      "loss": 0.6177,
+      "step": 5285
+    },
+    {
+      "epoch": 1.4032921810699588,
+      "grad_norm": 0.4074638211158301,
+      "learning_rate": 4.3725229585016965e-06,
+      "loss": 0.5848,
+      "step": 5286
+    },
+    {
+      "epoch": 1.4035576795433427,
+      "grad_norm": 0.4158302905645629,
+      "learning_rate": 4.372291622184462e-06,
+      "loss": 0.5869,
+      "step": 5287
+    },
+    {
+      "epoch": 1.4038231780167263,
+      "grad_norm": 0.39694763723980575,
+      "learning_rate": 4.372060249352956e-06,
+      "loss": 0.5652,
+      "step": 5288
+    },
+    {
+      "epoch": 1.4040886764901102,
+      "grad_norm": 0.39655098340209066,
+      "learning_rate": 4.371828840011693e-06,
+      "loss": 0.583,
+      "step": 5289
+    },
+    {
+      "epoch": 1.404354174963494,
+      "grad_norm": 0.4091042832231682,
+      "learning_rate": 4.3715973941651836e-06,
+      "loss": 0.6206,
+      "step": 5290
+    },
+    {
+      "epoch": 1.4046196734368777,
+      "grad_norm": 0.4053651809520837,
+      "learning_rate": 4.371365911817943e-06,
+      "loss": 0.6026,
+      "step": 5291
+    },
+    {
+      "epoch": 1.4048851719102615,
+      "grad_norm": 0.40941064842613534,
+      "learning_rate": 4.371134392974486e-06,
+      "loss": 0.6361,
+      "step": 5292
+    },
+    {
+      "epoch": 1.4051506703836454,
+      "grad_norm": 0.415346579813271,
+      "learning_rate": 4.370902837639325e-06,
+      "loss": 0.6199,
+      "step": 5293
+    },
+    {
+      "epoch": 1.405416168857029,
+      "grad_norm": 0.40667694688698347,
+      "learning_rate": 4.3706712458169795e-06,
+      "loss": 0.591,
+      "step": 5294
+    },
+    {
+      "epoch": 1.4056816673304129,
+      "grad_norm": 0.40934496579058594,
+      "learning_rate": 4.370439617511963e-06,
+      "loss": 0.6115,
+      "step": 5295
+    },
+    {
+      "epoch": 1.4059471658037968,
+      "grad_norm": 0.40727328604552504,
+      "learning_rate": 4.370207952728796e-06,
+      "loss": 0.6051,
+      "step": 5296
+    },
+    {
+      "epoch": 1.4062126642771804,
+      "grad_norm": 0.408943781476209,
+      "learning_rate": 4.369976251471995e-06,
+      "loss": 0.5886,
+      "step": 5297
+    },
+    {
+      "epoch": 1.4064781627505643,
+      "grad_norm": 0.41548593915236476,
+      "learning_rate": 4.369744513746077e-06,
+      "loss": 0.5629,
+      "step": 5298
+    },
+    {
+      "epoch": 1.406743661223948,
+      "grad_norm": 0.39960240060407004,
+      "learning_rate": 4.369512739555564e-06,
+      "loss": 0.5957,
+      "step": 5299
+    },
+    {
+      "epoch": 1.4070091596973318,
+      "grad_norm": 0.41430665662236077,
+      "learning_rate": 4.369280928904975e-06,
+      "loss": 0.5872,
+      "step": 5300
+    },
+    {
+      "epoch": 1.4072746581707154,
+      "grad_norm": 0.40361720832664233,
+      "learning_rate": 4.3690490817988305e-06,
+      "loss": 0.5963,
+      "step": 5301
+    },
+    {
+      "epoch": 1.4075401566440993,
+      "grad_norm": 0.4098336163375993,
+      "learning_rate": 4.368817198241654e-06,
+      "loss": 0.6168,
+      "step": 5302
+    },
+    {
+      "epoch": 1.4078056551174831,
+      "grad_norm": 0.40695955961714264,
+      "learning_rate": 4.368585278237966e-06,
+      "loss": 0.5861,
+      "step": 5303
+    },
+    {
+      "epoch": 1.4080711535908668,
+      "grad_norm": 0.4088482339140311,
+      "learning_rate": 4.368353321792289e-06,
+      "loss": 0.6039,
+      "step": 5304
+    },
+    {
+      "epoch": 1.4083366520642506,
+      "grad_norm": 0.40111906182858825,
+      "learning_rate": 4.368121328909149e-06,
+      "loss": 0.5584,
+      "step": 5305
+    },
+    {
+      "epoch": 1.4086021505376345,
+      "grad_norm": 0.3995935637648144,
+      "learning_rate": 4.367889299593069e-06,
+      "loss": 0.6107,
+      "step": 5306
+    },
+    {
+      "epoch": 1.4088676490110181,
+      "grad_norm": 0.4027004276828151,
+      "learning_rate": 4.367657233848574e-06,
+      "loss": 0.5606,
+      "step": 5307
+    },
+    {
+      "epoch": 1.409133147484402,
+      "grad_norm": 0.4055271806434678,
+      "learning_rate": 4.36742513168019e-06,
+      "loss": 0.5814,
+      "step": 5308
+    },
+    {
+      "epoch": 1.4093986459577859,
+      "grad_norm": 0.3986650374895936,
+      "learning_rate": 4.367192993092443e-06,
+      "loss": 0.608,
+      "step": 5309
+    },
+    {
+      "epoch": 1.4096641444311695,
+      "grad_norm": 0.4055650011720386,
+      "learning_rate": 4.366960818089862e-06,
+      "loss": 0.5736,
+      "step": 5310
+    },
+    {
+      "epoch": 1.4099296429045534,
+      "grad_norm": 0.40806863755128475,
+      "learning_rate": 4.366728606676974e-06,
+      "loss": 0.563,
+      "step": 5311
+    },
+    {
+      "epoch": 1.410195141377937,
+      "grad_norm": 0.39639167088312977,
+      "learning_rate": 4.366496358858307e-06,
+      "loss": 0.5699,
+      "step": 5312
+    },
+    {
+      "epoch": 1.4104606398513209,
+      "grad_norm": 0.41173710187012796,
+      "learning_rate": 4.366264074638391e-06,
+      "loss": 0.5993,
+      "step": 5313
+    },
+    {
+      "epoch": 1.4107261383247045,
+      "grad_norm": 0.40364060277847263,
+      "learning_rate": 4.366031754021757e-06,
+      "loss": 0.5619,
+      "step": 5314
+    },
+    {
+      "epoch": 1.4109916367980884,
+      "grad_norm": 0.3945677982795281,
+      "learning_rate": 4.365799397012934e-06,
+      "loss": 0.584,
+      "step": 5315
+    },
+    {
+      "epoch": 1.4112571352714722,
+      "grad_norm": 0.4053811827837774,
+      "learning_rate": 4.365567003616455e-06,
+      "loss": 0.6125,
+      "step": 5316
+    },
+    {
+      "epoch": 1.4115226337448559,
+      "grad_norm": 0.4127711039071285,
+      "learning_rate": 4.365334573836851e-06,
+      "loss": 0.5948,
+      "step": 5317
+    },
+    {
+      "epoch": 1.4117881322182397,
+      "grad_norm": 0.41381557860950585,
+      "learning_rate": 4.365102107678657e-06,
+      "loss": 0.5808,
+      "step": 5318
+    },
+    {
+      "epoch": 1.4120536306916236,
+      "grad_norm": 0.39775087318588587,
+      "learning_rate": 4.364869605146405e-06,
+      "loss": 0.5871,
+      "step": 5319
+    },
+    {
+      "epoch": 1.4123191291650072,
+      "grad_norm": 0.4120491014172842,
+      "learning_rate": 4.364637066244629e-06,
+      "loss": 0.5882,
+      "step": 5320
+    },
+    {
+      "epoch": 1.412584627638391,
+      "grad_norm": 0.40479997226818665,
+      "learning_rate": 4.364404490977865e-06,
+      "loss": 0.5641,
+      "step": 5321
+    },
+    {
+      "epoch": 1.412850126111775,
+      "grad_norm": 0.3963891092680847,
+      "learning_rate": 4.364171879350649e-06,
+      "loss": 0.5876,
+      "step": 5322
+    },
+    {
+      "epoch": 1.4131156245851586,
+      "grad_norm": 0.40607422557176354,
+      "learning_rate": 4.363939231367517e-06,
+      "loss": 0.611,
+      "step": 5323
+    },
+    {
+      "epoch": 1.4133811230585425,
+      "grad_norm": 0.4152950605828711,
+      "learning_rate": 4.363706547033006e-06,
+      "loss": 0.6077,
+      "step": 5324
+    },
+    {
+      "epoch": 1.4136466215319263,
+      "grad_norm": 0.4061887548601726,
+      "learning_rate": 4.363473826351654e-06,
+      "loss": 0.605,
+      "step": 5325
+    },
+    {
+      "epoch": 1.41391212000531,
+      "grad_norm": 0.40335911343303776,
+      "learning_rate": 4.363241069328001e-06,
+      "loss": 0.6235,
+      "step": 5326
+    },
+    {
+      "epoch": 1.4141776184786936,
+      "grad_norm": 0.4178917424227247,
+      "learning_rate": 4.363008275966584e-06,
+      "loss": 0.5695,
+      "step": 5327
+    },
+    {
+      "epoch": 1.4144431169520775,
+      "grad_norm": 0.40552455661420167,
+      "learning_rate": 4.362775446271944e-06,
+      "loss": 0.6083,
+      "step": 5328
+    },
+    {
+      "epoch": 1.4147086154254613,
+      "grad_norm": 0.41041762619987465,
+      "learning_rate": 4.362542580248622e-06,
+      "loss": 0.5633,
+      "step": 5329
+    },
+    {
+      "epoch": 1.414974113898845,
+      "grad_norm": 0.411152773868579,
+      "learning_rate": 4.36230967790116e-06,
+      "loss": 0.5839,
+      "step": 5330
+    },
+    {
+      "epoch": 1.4152396123722288,
+      "grad_norm": 0.393379430056764,
+      "learning_rate": 4.362076739234099e-06,
+      "loss": 0.5985,
+      "step": 5331
+    },
+    {
+      "epoch": 1.4155051108456127,
+      "grad_norm": 0.41925976968021955,
+      "learning_rate": 4.361843764251983e-06,
+      "loss": 0.6014,
+      "step": 5332
+    },
+    {
+      "epoch": 1.4157706093189963,
+      "grad_norm": 0.4162452389113423,
+      "learning_rate": 4.361610752959355e-06,
+      "loss": 0.5684,
+      "step": 5333
+    },
+    {
+      "epoch": 1.4160361077923802,
+      "grad_norm": 0.41766182715221006,
+      "learning_rate": 4.361377705360758e-06,
+      "loss": 0.5963,
+      "step": 5334
+    },
+    {
+      "epoch": 1.416301606265764,
+      "grad_norm": 0.40308721273131726,
+      "learning_rate": 4.361144621460741e-06,
+      "loss": 0.5735,
+      "step": 5335
+    },
+    {
+      "epoch": 1.4165671047391477,
+      "grad_norm": 0.416531654864316,
+      "learning_rate": 4.360911501263845e-06,
+      "loss": 0.596,
+      "step": 5336
+    },
+    {
+      "epoch": 1.4168326032125316,
+      "grad_norm": 0.402475051645599,
+      "learning_rate": 4.360678344774619e-06,
+      "loss": 0.5922,
+      "step": 5337
+    },
+    {
+      "epoch": 1.4170981016859154,
+      "grad_norm": 0.4103172913816098,
+      "learning_rate": 4.360445151997609e-06,
+      "loss": 0.5866,
+      "step": 5338
+    },
+    {
+      "epoch": 1.417363600159299,
+      "grad_norm": 0.4067747266708178,
+      "learning_rate": 4.360211922937364e-06,
+      "loss": 0.5763,
+      "step": 5339
+    },
+    {
+      "epoch": 1.417629098632683,
+      "grad_norm": 0.4303578690942572,
+      "learning_rate": 4.359978657598432e-06,
+      "loss": 0.5983,
+      "step": 5340
+    },
+    {
+      "epoch": 1.4178945971060666,
+      "grad_norm": 0.41425841064424457,
+      "learning_rate": 4.359745355985361e-06,
+      "loss": 0.5949,
+      "step": 5341
+    },
+    {
+      "epoch": 1.4181600955794504,
+      "grad_norm": 0.4121203876653073,
+      "learning_rate": 4.359512018102703e-06,
+      "loss": 0.574,
+      "step": 5342
+    },
+    {
+      "epoch": 1.418425594052834,
+      "grad_norm": 0.40959548419616476,
+      "learning_rate": 4.359278643955008e-06,
+      "loss": 0.6094,
+      "step": 5343
+    },
+    {
+      "epoch": 1.418691092526218,
+      "grad_norm": 0.423015636493266,
+      "learning_rate": 4.359045233546827e-06,
+      "loss": 0.5818,
+      "step": 5344
+    },
+    {
+      "epoch": 1.4189565909996018,
+      "grad_norm": 0.401945569769227,
+      "learning_rate": 4.3588117868827115e-06,
+      "loss": 0.6068,
+      "step": 5345
+    },
+    {
+      "epoch": 1.4192220894729854,
+      "grad_norm": 0.40915017011413707,
+      "learning_rate": 4.358578303967217e-06,
+      "loss": 0.597,
+      "step": 5346
+    },
+    {
+      "epoch": 1.4194875879463693,
+      "grad_norm": 0.40799379446061695,
+      "learning_rate": 4.358344784804892e-06,
+      "loss": 0.6024,
+      "step": 5347
+    },
+    {
+      "epoch": 1.4197530864197532,
+      "grad_norm": 0.42379208375936706,
+      "learning_rate": 4.358111229400296e-06,
+      "loss": 0.5782,
+      "step": 5348
+    },
+    {
+      "epoch": 1.4200185848931368,
+      "grad_norm": 0.41992365656012554,
+      "learning_rate": 4.357877637757981e-06,
+      "loss": 0.6092,
+      "step": 5349
+    },
+    {
+      "epoch": 1.4202840833665207,
+      "grad_norm": 0.4094398581320978,
+      "learning_rate": 4.357644009882503e-06,
+      "loss": 0.5828,
+      "step": 5350
+    },
+    {
+      "epoch": 1.4205495818399045,
+      "grad_norm": 0.4134424558294693,
+      "learning_rate": 4.357410345778419e-06,
+      "loss": 0.5951,
+      "step": 5351
+    },
+    {
+      "epoch": 1.4208150803132882,
+      "grad_norm": 0.40875161300908747,
+      "learning_rate": 4.357176645450285e-06,
+      "loss": 0.6119,
+      "step": 5352
+    },
+    {
+      "epoch": 1.421080578786672,
+      "grad_norm": 0.4196709278126285,
+      "learning_rate": 4.356942908902659e-06,
+      "loss": 0.5819,
+      "step": 5353
+    },
+    {
+      "epoch": 1.4213460772600557,
+      "grad_norm": 0.4088574346368782,
+      "learning_rate": 4.3567091361401e-06,
+      "loss": 0.5793,
+      "step": 5354
+    },
+    {
+      "epoch": 1.4216115757334395,
+      "grad_norm": 0.41445792450823365,
+      "learning_rate": 4.356475327167167e-06,
+      "loss": 0.5795,
+      "step": 5355
+    },
+    {
+      "epoch": 1.4218770742068232,
+      "grad_norm": 0.41088643333676433,
+      "learning_rate": 4.35624148198842e-06,
+      "loss": 0.561,
+      "step": 5356
+    },
+    {
+      "epoch": 1.422142572680207,
+      "grad_norm": 0.40542690902190504,
+      "learning_rate": 4.356007600608419e-06,
+      "loss": 0.6155,
+      "step": 5357
+    },
+    {
+      "epoch": 1.422408071153591,
+      "grad_norm": 0.4297874724677386,
+      "learning_rate": 4.355773683031725e-06,
+      "loss": 0.6235,
+      "step": 5358
+    },
+    {
+      "epoch": 1.4226735696269746,
+      "grad_norm": 0.40599526949998954,
+      "learning_rate": 4.355539729262901e-06,
+      "loss": 0.563,
+      "step": 5359
+    },
+    {
+      "epoch": 1.4229390681003584,
+      "grad_norm": 0.40884888221244214,
+      "learning_rate": 4.355305739306508e-06,
+      "loss": 0.5671,
+      "step": 5360
+    },
+    {
+      "epoch": 1.4232045665737423,
+      "grad_norm": 0.3997016891896433,
+      "learning_rate": 4.355071713167111e-06,
+      "loss": 0.5475,
+      "step": 5361
+    },
+    {
+      "epoch": 1.423470065047126,
+      "grad_norm": 0.402385255102974,
+      "learning_rate": 4.354837650849275e-06,
+      "loss": 0.5535,
+      "step": 5362
+    },
+    {
+      "epoch": 1.4237355635205098,
+      "grad_norm": 0.42051078246384344,
+      "learning_rate": 4.354603552357562e-06,
+      "loss": 0.6228,
+      "step": 5363
+    },
+    {
+      "epoch": 1.4240010619938936,
+      "grad_norm": 0.41455231571053097,
+      "learning_rate": 4.3543694176965394e-06,
+      "loss": 0.5993,
+      "step": 5364
+    },
+    {
+      "epoch": 1.4242665604672773,
+      "grad_norm": 0.41806032887970623,
+      "learning_rate": 4.354135246870773e-06,
+      "loss": 0.629,
+      "step": 5365
+    },
+    {
+      "epoch": 1.4245320589406612,
+      "grad_norm": 0.42407248250406526,
+      "learning_rate": 4.353901039884829e-06,
+      "loss": 0.6323,
+      "step": 5366
+    },
+    {
+      "epoch": 1.4247975574140448,
+      "grad_norm": 0.4024201674732508,
+      "learning_rate": 4.353666796743276e-06,
+      "loss": 0.5977,
+      "step": 5367
+    },
+    {
+      "epoch": 1.4250630558874287,
+      "grad_norm": 0.4036344051520522,
+      "learning_rate": 4.353432517450682e-06,
+      "loss": 0.5672,
+      "step": 5368
+    },
+    {
+      "epoch": 1.4253285543608123,
+      "grad_norm": 0.41543869187316923,
+      "learning_rate": 4.3531982020116155e-06,
+      "loss": 0.5912,
+      "step": 5369
+    },
+    {
+      "epoch": 1.4255940528341962,
+      "grad_norm": 0.40188075825548764,
+      "learning_rate": 4.352963850430647e-06,
+      "loss": 0.5869,
+      "step": 5370
+    },
+    {
+      "epoch": 1.42585955130758,
+      "grad_norm": 0.403358019921609,
+      "learning_rate": 4.3527294627123464e-06,
+      "loss": 0.5653,
+      "step": 5371
+    },
+    {
+      "epoch": 1.4261250497809637,
+      "grad_norm": 0.3935130719948093,
+      "learning_rate": 4.352495038861285e-06,
+      "loss": 0.6267,
+      "step": 5372
+    },
+    {
+      "epoch": 1.4263905482543475,
+      "grad_norm": 0.41265078077525935,
+      "learning_rate": 4.352260578882035e-06,
+      "loss": 0.5926,
+      "step": 5373
+    },
+    {
+      "epoch": 1.4266560467277314,
+      "grad_norm": 0.39721978921417833,
+      "learning_rate": 4.352026082779168e-06,
+      "loss": 0.5622,
+      "step": 5374
+    },
+    {
+      "epoch": 1.426921545201115,
+      "grad_norm": 0.3919134791898912,
+      "learning_rate": 4.351791550557259e-06,
+      "loss": 0.6206,
+      "step": 5375
+    },
+    {
+      "epoch": 1.427187043674499,
+      "grad_norm": 0.39853708208547006,
+      "learning_rate": 4.35155698222088e-06,
+      "loss": 0.5795,
+      "step": 5376
+    },
+    {
+      "epoch": 1.4274525421478828,
+      "grad_norm": 0.4007002271504504,
+      "learning_rate": 4.351322377774606e-06,
+      "loss": 0.5688,
+      "step": 5377
+    },
+    {
+      "epoch": 1.4277180406212664,
+      "grad_norm": 0.394594869417455,
+      "learning_rate": 4.351087737223013e-06,
+      "loss": 0.6253,
+      "step": 5378
+    },
+    {
+      "epoch": 1.4279835390946503,
+      "grad_norm": 0.38511800117446227,
+      "learning_rate": 4.350853060570678e-06,
+      "loss": 0.5768,
+      "step": 5379
+    },
+    {
+      "epoch": 1.4282490375680341,
+      "grad_norm": 0.4077100419432239,
+      "learning_rate": 4.350618347822175e-06,
+      "loss": 0.6031,
+      "step": 5380
+    },
+    {
+      "epoch": 1.4285145360414178,
+      "grad_norm": 0.40662903595869726,
+      "learning_rate": 4.350383598982084e-06,
+      "loss": 0.6137,
+      "step": 5381
+    },
+    {
+      "epoch": 1.4287800345148016,
+      "grad_norm": 0.4181155945222628,
+      "learning_rate": 4.350148814054982e-06,
+      "loss": 0.6298,
+      "step": 5382
+    },
+    {
+      "epoch": 1.4290455329881853,
+      "grad_norm": 0.4241685652598452,
+      "learning_rate": 4.349913993045448e-06,
+      "loss": 0.617,
+      "step": 5383
+    },
+    {
+      "epoch": 1.4293110314615691,
+      "grad_norm": 0.41344417837681113,
+      "learning_rate": 4.349679135958062e-06,
+      "loss": 0.6182,
+      "step": 5384
+    },
+    {
+      "epoch": 1.4295765299349528,
+      "grad_norm": 0.4186245162057323,
+      "learning_rate": 4.349444242797406e-06,
+      "loss": 0.5878,
+      "step": 5385
+    },
+    {
+      "epoch": 1.4298420284083366,
+      "grad_norm": 0.4203660862473339,
+      "learning_rate": 4.349209313568057e-06,
+      "loss": 0.5846,
+      "step": 5386
+    },
+    {
+      "epoch": 1.4301075268817205,
+      "grad_norm": 0.4118107730038185,
+      "learning_rate": 4.348974348274599e-06,
+      "loss": 0.6418,
+      "step": 5387
+    },
+    {
+      "epoch": 1.4303730253551041,
+      "grad_norm": 0.4235576926585897,
+      "learning_rate": 4.348739346921613e-06,
+      "loss": 0.5852,
+      "step": 5388
+    },
+    {
+      "epoch": 1.430638523828488,
+      "grad_norm": 0.3964713478502504,
+      "learning_rate": 4.348504309513685e-06,
+      "loss": 0.5905,
+      "step": 5389
+    },
+    {
+      "epoch": 1.4309040223018719,
+      "grad_norm": 0.4072504004544051,
+      "learning_rate": 4.348269236055396e-06,
+      "loss": 0.5795,
+      "step": 5390
+    },
+    {
+      "epoch": 1.4311695207752555,
+      "grad_norm": 0.39577504724901286,
+      "learning_rate": 4.348034126551333e-06,
+      "loss": 0.59,
+      "step": 5391
+    },
+    {
+      "epoch": 1.4314350192486394,
+      "grad_norm": 0.4066690657217721,
+      "learning_rate": 4.347798981006078e-06,
+      "loss": 0.6043,
+      "step": 5392
+    },
+    {
+      "epoch": 1.4317005177220232,
+      "grad_norm": 0.39525308153088334,
+      "learning_rate": 4.347563799424219e-06,
+      "loss": 0.563,
+      "step": 5393
+    },
+    {
+      "epoch": 1.4319660161954069,
+      "grad_norm": 0.3941996342086955,
+      "learning_rate": 4.347328581810343e-06,
+      "loss": 0.5973,
+      "step": 5394
+    },
+    {
+      "epoch": 1.4322315146687907,
+      "grad_norm": 0.3954181035257636,
+      "learning_rate": 4.347093328169036e-06,
+      "loss": 0.5797,
+      "step": 5395
+    },
+    {
+      "epoch": 1.4324970131421744,
+      "grad_norm": 0.40958793499614976,
+      "learning_rate": 4.346858038504888e-06,
+      "loss": 0.604,
+      "step": 5396
+    },
+    {
+      "epoch": 1.4327625116155582,
+      "grad_norm": 0.4125830091648663,
+      "learning_rate": 4.3466227128224845e-06,
+      "loss": 0.6023,
+      "step": 5397
+    },
+    {
+      "epoch": 1.4330280100889419,
+      "grad_norm": 0.41549334230852086,
+      "learning_rate": 4.346387351126418e-06,
+      "loss": 0.6011,
+      "step": 5398
+    },
+    {
+      "epoch": 1.4332935085623257,
+      "grad_norm": 0.3952119580353629,
+      "learning_rate": 4.3461519534212774e-06,
+      "loss": 0.5814,
+      "step": 5399
+    },
+    {
+      "epoch": 1.4335590070357096,
+      "grad_norm": 0.40641000231135405,
+      "learning_rate": 4.345916519711653e-06,
+      "loss": 0.5883,
+      "step": 5400
+    },
+    {
+      "epoch": 1.4338245055090932,
+      "grad_norm": 0.3958097134277731,
+      "learning_rate": 4.3456810500021365e-06,
+      "loss": 0.5672,
+      "step": 5401
+    },
+    {
+      "epoch": 1.434090003982477,
+      "grad_norm": 0.41847262900874216,
+      "learning_rate": 4.345445544297321e-06,
+      "loss": 0.5684,
+      "step": 5402
+    },
+    {
+      "epoch": 1.434355502455861,
+      "grad_norm": 0.40777697246168537,
+      "learning_rate": 4.3452100026018e-06,
+      "loss": 0.6054,
+      "step": 5403
+    },
+    {
+      "epoch": 1.4346210009292446,
+      "grad_norm": 0.41588291312127035,
+      "learning_rate": 4.344974424920164e-06,
+      "loss": 0.5764,
+      "step": 5404
+    },
+    {
+      "epoch": 1.4348864994026285,
+      "grad_norm": 0.3930706260191455,
+      "learning_rate": 4.344738811257011e-06,
+      "loss": 0.5686,
+      "step": 5405
+    },
+    {
+      "epoch": 1.4351519978760123,
+      "grad_norm": 0.40908641818940145,
+      "learning_rate": 4.344503161616934e-06,
+      "loss": 0.5711,
+      "step": 5406
+    },
+    {
+      "epoch": 1.435417496349396,
+      "grad_norm": 0.399902104508832,
+      "learning_rate": 4.344267476004529e-06,
+      "loss": 0.5829,
+      "step": 5407
+    },
+    {
+      "epoch": 1.4356829948227798,
+      "grad_norm": 0.40441707037823643,
+      "learning_rate": 4.344031754424394e-06,
+      "loss": 0.5973,
+      "step": 5408
+    },
+    {
+      "epoch": 1.4359484932961635,
+      "grad_norm": 0.3972826785019991,
+      "learning_rate": 4.343795996881123e-06,
+      "loss": 0.5508,
+      "step": 5409
+    },
+    {
+      "epoch": 1.4362139917695473,
+      "grad_norm": 0.41285746563624176,
+      "learning_rate": 4.3435602033793166e-06,
+      "loss": 0.5814,
+      "step": 5410
+    },
+    {
+      "epoch": 1.436479490242931,
+      "grad_norm": 0.4113427960837754,
+      "learning_rate": 4.343324373923572e-06,
+      "loss": 0.5656,
+      "step": 5411
+    },
+    {
+      "epoch": 1.4367449887163148,
+      "grad_norm": 0.3936720183865051,
+      "learning_rate": 4.343088508518489e-06,
+      "loss": 0.5762,
+      "step": 5412
+    },
+    {
+      "epoch": 1.4370104871896987,
+      "grad_norm": 0.40742200774348103,
+      "learning_rate": 4.342852607168668e-06,
+      "loss": 0.5954,
+      "step": 5413
+    },
+    {
+      "epoch": 1.4372759856630823,
+      "grad_norm": 0.4080860404982624,
+      "learning_rate": 4.3426166698787085e-06,
+      "loss": 0.5944,
+      "step": 5414
+    },
+    {
+      "epoch": 1.4375414841364662,
+      "grad_norm": 0.40244449202505966,
+      "learning_rate": 4.342380696653212e-06,
+      "loss": 0.5952,
+      "step": 5415
+    },
+    {
+      "epoch": 1.43780698260985,
+      "grad_norm": 0.4039381803811445,
+      "learning_rate": 4.342144687496782e-06,
+      "loss": 0.5686,
+      "step": 5416
+    },
+    {
+      "epoch": 1.4380724810832337,
+      "grad_norm": 0.41499291111739744,
+      "learning_rate": 4.34190864241402e-06,
+      "loss": 0.5945,
+      "step": 5417
+    },
+    {
+      "epoch": 1.4383379795566176,
+      "grad_norm": 0.4177973178536254,
+      "learning_rate": 4.34167256140953e-06,
+      "loss": 0.5785,
+      "step": 5418
+    },
+    {
+      "epoch": 1.4386034780300014,
+      "grad_norm": 0.4128393275701462,
+      "learning_rate": 4.341436444487915e-06,
+      "loss": 0.6134,
+      "step": 5419
+    },
+    {
+      "epoch": 1.438868976503385,
+      "grad_norm": 0.41300285253860947,
+      "learning_rate": 4.341200291653781e-06,
+      "loss": 0.5908,
+      "step": 5420
+    },
+    {
+      "epoch": 1.439134474976769,
+      "grad_norm": 0.40712441379318276,
+      "learning_rate": 4.340964102911734e-06,
+      "loss": 0.6207,
+      "step": 5421
+    },
+    {
+      "epoch": 1.4393999734501526,
+      "grad_norm": 0.4185747047257086,
+      "learning_rate": 4.340727878266378e-06,
+      "loss": 0.6129,
+      "step": 5422
+    },
+    {
+      "epoch": 1.4396654719235364,
+      "grad_norm": 0.40830765602805963,
+      "learning_rate": 4.340491617722322e-06,
+      "loss": 0.636,
+      "step": 5423
+    },
+    {
+      "epoch": 1.43993097039692,
+      "grad_norm": 0.4026988027973634,
+      "learning_rate": 4.340255321284174e-06,
+      "loss": 0.5891,
+      "step": 5424
+    },
+    {
+      "epoch": 1.440196468870304,
+      "grad_norm": 0.405317914422975,
+      "learning_rate": 4.340018988956541e-06,
+      "loss": 0.5918,
+      "step": 5425
+    },
+    {
+      "epoch": 1.4404619673436878,
+      "grad_norm": 0.41181399368716043,
+      "learning_rate": 4.339782620744033e-06,
+      "loss": 0.5947,
+      "step": 5426
+    },
+    {
+      "epoch": 1.4407274658170715,
+      "grad_norm": 0.4039368767437816,
+      "learning_rate": 4.3395462166512596e-06,
+      "loss": 0.5927,
+      "step": 5427
+    },
+    {
+      "epoch": 1.4409929642904553,
+      "grad_norm": 0.40637221806875323,
+      "learning_rate": 4.33930977668283e-06,
+      "loss": 0.598,
+      "step": 5428
+    },
+    {
+      "epoch": 1.4412584627638392,
+      "grad_norm": 0.4178281724738597,
+      "learning_rate": 4.339073300843356e-06,
+      "loss": 0.6057,
+      "step": 5429
+    },
+    {
+      "epoch": 1.4415239612372228,
+      "grad_norm": 0.39391549625748806,
+      "learning_rate": 4.338836789137451e-06,
+      "loss": 0.5723,
+      "step": 5430
+    },
+    {
+      "epoch": 1.4417894597106067,
+      "grad_norm": 0.40899828343454736,
+      "learning_rate": 4.338600241569726e-06,
+      "loss": 0.608,
+      "step": 5431
+    },
+    {
+      "epoch": 1.4420549581839905,
+      "grad_norm": 0.40565685463235096,
+      "learning_rate": 4.338363658144794e-06,
+      "loss": 0.595,
+      "step": 5432
+    },
+    {
+      "epoch": 1.4423204566573742,
+      "grad_norm": 0.4118582770191195,
+      "learning_rate": 4.33812703886727e-06,
+      "loss": 0.6299,
+      "step": 5433
+    },
+    {
+      "epoch": 1.442585955130758,
+      "grad_norm": 0.42700345287573316,
+      "learning_rate": 4.337890383741769e-06,
+      "loss": 0.6084,
+      "step": 5434
+    },
+    {
+      "epoch": 1.442851453604142,
+      "grad_norm": 0.40277256222908836,
+      "learning_rate": 4.337653692772904e-06,
+      "loss": 0.5807,
+      "step": 5435
+    },
+    {
+      "epoch": 1.4431169520775255,
+      "grad_norm": 0.40443995077301803,
+      "learning_rate": 4.337416965965294e-06,
+      "loss": 0.6309,
+      "step": 5436
+    },
+    {
+      "epoch": 1.4433824505509094,
+      "grad_norm": 0.3951468780229571,
+      "learning_rate": 4.337180203323553e-06,
+      "loss": 0.5886,
+      "step": 5437
+    },
+    {
+      "epoch": 1.443647949024293,
+      "grad_norm": 0.41984358426849466,
+      "learning_rate": 4.336943404852301e-06,
+      "loss": 0.6055,
+      "step": 5438
+    },
+    {
+      "epoch": 1.443913447497677,
+      "grad_norm": 0.40774160937653453,
+      "learning_rate": 4.336706570556155e-06,
+      "loss": 0.5817,
+      "step": 5439
+    },
+    {
+      "epoch": 1.4441789459710606,
+      "grad_norm": 0.407572516765919,
+      "learning_rate": 4.3364697004397334e-06,
+      "loss": 0.5862,
+      "step": 5440
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.43177515256798255,
+      "learning_rate": 4.336232794507657e-06,
+      "loss": 0.6095,
+      "step": 5441
+    },
+    {
+      "epoch": 1.4447099429178283,
+      "grad_norm": 0.4173660347566688,
+      "learning_rate": 4.335995852764544e-06,
+      "loss": 0.6126,
+      "step": 5442
+    },
+    {
+      "epoch": 1.444975441391212,
+      "grad_norm": 0.4181341854904306,
+      "learning_rate": 4.335758875215018e-06,
+      "loss": 0.5978,
+      "step": 5443
+    },
+    {
+      "epoch": 1.4452409398645958,
+      "grad_norm": 0.4133611250793086,
+      "learning_rate": 4.335521861863698e-06,
+      "loss": 0.5723,
+      "step": 5444
+    },
+    {
+      "epoch": 1.4455064383379796,
+      "grad_norm": 0.411445085282278,
+      "learning_rate": 4.335284812715208e-06,
+      "loss": 0.5939,
+      "step": 5445
+    },
+    {
+      "epoch": 1.4457719368113633,
+      "grad_norm": 0.409531687595904,
+      "learning_rate": 4.335047727774171e-06,
+      "loss": 0.6025,
+      "step": 5446
+    },
+    {
+      "epoch": 1.4460374352847472,
+      "grad_norm": 0.4041967861213368,
+      "learning_rate": 4.334810607045209e-06,
+      "loss": 0.5762,
+      "step": 5447
+    },
+    {
+      "epoch": 1.446302933758131,
+      "grad_norm": 0.3916044838431704,
+      "learning_rate": 4.334573450532949e-06,
+      "loss": 0.5749,
+      "step": 5448
+    },
+    {
+      "epoch": 1.4465684322315147,
+      "grad_norm": 0.39981302592637735,
+      "learning_rate": 4.3343362582420144e-06,
+      "loss": 0.5651,
+      "step": 5449
+    },
+    {
+      "epoch": 1.4468339307048985,
+      "grad_norm": 0.4107400754193761,
+      "learning_rate": 4.334099030177032e-06,
+      "loss": 0.6287,
+      "step": 5450
+    },
+    {
+      "epoch": 1.4470994291782822,
+      "grad_norm": 0.4104590770130155,
+      "learning_rate": 4.333861766342627e-06,
+      "loss": 0.6355,
+      "step": 5451
+    },
+    {
+      "epoch": 1.447364927651666,
+      "grad_norm": 0.4166844764771257,
+      "learning_rate": 4.333624466743429e-06,
+      "loss": 0.5715,
+      "step": 5452
+    },
+    {
+      "epoch": 1.4476304261250497,
+      "grad_norm": 0.39199725133534785,
+      "learning_rate": 4.333387131384063e-06,
+      "loss": 0.5926,
+      "step": 5453
+    },
+    {
+      "epoch": 1.4478959245984335,
+      "grad_norm": 0.3911890903739129,
+      "learning_rate": 4.333149760269159e-06,
+      "loss": 0.582,
+      "step": 5454
+    },
+    {
+      "epoch": 1.4481614230718174,
+      "grad_norm": 0.3961386406296229,
+      "learning_rate": 4.332912353403348e-06,
+      "loss": 0.5566,
+      "step": 5455
+    },
+    {
+      "epoch": 1.448426921545201,
+      "grad_norm": 0.4071531075080342,
+      "learning_rate": 4.3326749107912565e-06,
+      "loss": 0.6022,
+      "step": 5456
+    },
+    {
+      "epoch": 1.448692420018585,
+      "grad_norm": 0.416994450285505,
+      "learning_rate": 4.332437432437518e-06,
+      "loss": 0.5732,
+      "step": 5457
+    },
+    {
+      "epoch": 1.4489579184919688,
+      "grad_norm": 0.41562233857098557,
+      "learning_rate": 4.332199918346763e-06,
+      "loss": 0.5871,
+      "step": 5458
+    },
+    {
+      "epoch": 1.4492234169653524,
+      "grad_norm": 0.42038990433640216,
+      "learning_rate": 4.331962368523623e-06,
+      "loss": 0.6306,
+      "step": 5459
+    },
+    {
+      "epoch": 1.4494889154387363,
+      "grad_norm": 0.4024305790045775,
+      "learning_rate": 4.331724782972732e-06,
+      "loss": 0.563,
+      "step": 5460
+    },
+    {
+      "epoch": 1.4497544139121201,
+      "grad_norm": 0.4188294426436788,
+      "learning_rate": 4.331487161698723e-06,
+      "loss": 0.6077,
+      "step": 5461
+    },
+    {
+      "epoch": 1.4500199123855038,
+      "grad_norm": 0.392663338312878,
+      "learning_rate": 4.33124950470623e-06,
+      "loss": 0.5897,
+      "step": 5462
+    },
+    {
+      "epoch": 1.4502854108588876,
+      "grad_norm": 0.4125620926323638,
+      "learning_rate": 4.331011811999887e-06,
+      "loss": 0.5943,
+      "step": 5463
+    },
+    {
+      "epoch": 1.4505509093322713,
+      "grad_norm": 0.4097834564106724,
+      "learning_rate": 4.330774083584332e-06,
+      "loss": 0.5922,
+      "step": 5464
+    },
+    {
+      "epoch": 1.4508164078056551,
+      "grad_norm": 0.42159881845540426,
+      "learning_rate": 4.330536319464198e-06,
+      "loss": 0.5948,
+      "step": 5465
+    },
+    {
+      "epoch": 1.4510819062790388,
+      "grad_norm": 0.4230387523286175,
+      "learning_rate": 4.330298519644126e-06,
+      "loss": 0.5959,
+      "step": 5466
+    },
+    {
+      "epoch": 1.4513474047524226,
+      "grad_norm": 0.4168478177364035,
+      "learning_rate": 4.330060684128751e-06,
+      "loss": 0.5926,
+      "step": 5467
+    },
+    {
+      "epoch": 1.4516129032258065,
+      "grad_norm": 0.41229228265412776,
+      "learning_rate": 4.32982281292271e-06,
+      "loss": 0.6202,
+      "step": 5468
+    },
+    {
+      "epoch": 1.4518784016991901,
+      "grad_norm": 0.40178209280550553,
+      "learning_rate": 4.329584906030646e-06,
+      "loss": 0.5662,
+      "step": 5469
+    },
+    {
+      "epoch": 1.452143900172574,
+      "grad_norm": 0.3980591501056759,
+      "learning_rate": 4.329346963457196e-06,
+      "loss": 0.593,
+      "step": 5470
+    },
+    {
+      "epoch": 1.4524093986459579,
+      "grad_norm": 0.40434182895856163,
+      "learning_rate": 4.329108985207002e-06,
+      "loss": 0.5754,
+      "step": 5471
+    },
+    {
+      "epoch": 1.4526748971193415,
+      "grad_norm": 0.39006608758269024,
+      "learning_rate": 4.328870971284703e-06,
+      "loss": 0.561,
+      "step": 5472
+    },
+    {
+      "epoch": 1.4529403955927254,
+      "grad_norm": 0.39959009725792033,
+      "learning_rate": 4.328632921694944e-06,
+      "loss": 0.5972,
+      "step": 5473
+    },
+    {
+      "epoch": 1.4532058940661092,
+      "grad_norm": 0.4138017002470512,
+      "learning_rate": 4.328394836442364e-06,
+      "loss": 0.5417,
+      "step": 5474
+    },
+    {
+      "epoch": 1.4534713925394929,
+      "grad_norm": 0.40348164169717077,
+      "learning_rate": 4.328156715531608e-06,
+      "loss": 0.5726,
+      "step": 5475
+    },
+    {
+      "epoch": 1.4537368910128767,
+      "grad_norm": 0.39324190531619213,
+      "learning_rate": 4.327918558967321e-06,
+      "loss": 0.5981,
+      "step": 5476
+    },
+    {
+      "epoch": 1.4540023894862604,
+      "grad_norm": 0.4144204931628961,
+      "learning_rate": 4.327680366754147e-06,
+      "loss": 0.5924,
+      "step": 5477
+    },
+    {
+      "epoch": 1.4542678879596442,
+      "grad_norm": 0.41712205163051763,
+      "learning_rate": 4.3274421388967305e-06,
+      "loss": 0.6086,
+      "step": 5478
+    },
+    {
+      "epoch": 1.4545333864330279,
+      "grad_norm": 0.42511867942034154,
+      "learning_rate": 4.327203875399717e-06,
+      "loss": 0.6086,
+      "step": 5479
+    },
+    {
+      "epoch": 1.4547988849064117,
+      "grad_norm": 0.39937049794595736,
+      "learning_rate": 4.326965576267755e-06,
+      "loss": 0.5834,
+      "step": 5480
+    },
+    {
+      "epoch": 1.4550643833797956,
+      "grad_norm": 0.3973577042092701,
+      "learning_rate": 4.32672724150549e-06,
+      "loss": 0.596,
+      "step": 5481
+    },
+    {
+      "epoch": 1.4553298818531792,
+      "grad_norm": 0.4013106328113882,
+      "learning_rate": 4.326488871117572e-06,
+      "loss": 0.6089,
+      "step": 5482
+    },
+    {
+      "epoch": 1.455595380326563,
+      "grad_norm": 0.4039276485466298,
+      "learning_rate": 4.3262504651086495e-06,
+      "loss": 0.5913,
+      "step": 5483
+    },
+    {
+      "epoch": 1.455860878799947,
+      "grad_norm": 0.3995375908847534,
+      "learning_rate": 4.326012023483371e-06,
+      "loss": 0.5787,
+      "step": 5484
+    },
+    {
+      "epoch": 1.4561263772733306,
+      "grad_norm": 0.4263769137357294,
+      "learning_rate": 4.325773546246387e-06,
+      "loss": 0.5842,
+      "step": 5485
+    },
+    {
+      "epoch": 1.4563918757467145,
+      "grad_norm": 0.41494019247186775,
+      "learning_rate": 4.325535033402349e-06,
+      "loss": 0.6301,
+      "step": 5486
+    },
+    {
+      "epoch": 1.4566573742200983,
+      "grad_norm": 0.404144692901755,
+      "learning_rate": 4.325296484955908e-06,
+      "loss": 0.5614,
+      "step": 5487
+    },
+    {
+      "epoch": 1.456922872693482,
+      "grad_norm": 0.41933221855008823,
+      "learning_rate": 4.325057900911716e-06,
+      "loss": 0.5875,
+      "step": 5488
+    },
+    {
+      "epoch": 1.4571883711668658,
+      "grad_norm": 0.4167875042132777,
+      "learning_rate": 4.324819281274427e-06,
+      "loss": 0.5988,
+      "step": 5489
+    },
+    {
+      "epoch": 1.4574538696402497,
+      "grad_norm": 0.39476948305622933,
+      "learning_rate": 4.324580626048694e-06,
+      "loss": 0.6044,
+      "step": 5490
+    },
+    {
+      "epoch": 1.4577193681136333,
+      "grad_norm": 0.40857762026441824,
+      "learning_rate": 4.324341935239171e-06,
+      "loss": 0.6044,
+      "step": 5491
+    },
+    {
+      "epoch": 1.4579848665870172,
+      "grad_norm": 0.4049614863481209,
+      "learning_rate": 4.324103208850514e-06,
+      "loss": 0.5639,
+      "step": 5492
+    },
+    {
+      "epoch": 1.4582503650604008,
+      "grad_norm": 0.41736952126280746,
+      "learning_rate": 4.323864446887378e-06,
+      "loss": 0.5811,
+      "step": 5493
+    },
+    {
+      "epoch": 1.4585158635337847,
+      "grad_norm": 0.4160546325409148,
+      "learning_rate": 4.32362564935442e-06,
+      "loss": 0.5786,
+      "step": 5494
+    },
+    {
+      "epoch": 1.4587813620071683,
+      "grad_norm": 0.4030609696339856,
+      "learning_rate": 4.323386816256297e-06,
+      "loss": 0.6011,
+      "step": 5495
+    },
+    {
+      "epoch": 1.4590468604805522,
+      "grad_norm": 0.4117288046483035,
+      "learning_rate": 4.323147947597667e-06,
+      "loss": 0.5981,
+      "step": 5496
+    },
+    {
+      "epoch": 1.459312358953936,
+      "grad_norm": 0.4067457796775077,
+      "learning_rate": 4.322909043383186e-06,
+      "loss": 0.5866,
+      "step": 5497
+    },
+    {
+      "epoch": 1.4595778574273197,
+      "grad_norm": 0.4077089464034527,
+      "learning_rate": 4.322670103617518e-06,
+      "loss": 0.6138,
+      "step": 5498
+    },
+    {
+      "epoch": 1.4598433559007036,
+      "grad_norm": 0.41988084076100224,
+      "learning_rate": 4.322431128305319e-06,
+      "loss": 0.5707,
+      "step": 5499
+    },
+    {
+      "epoch": 1.4601088543740874,
+      "grad_norm": 0.4036926688403312,
+      "learning_rate": 4.322192117451252e-06,
+      "loss": 0.5925,
+      "step": 5500
+    },
+    {
+      "epoch": 1.460374352847471,
+      "grad_norm": 0.40916019952040755,
+      "learning_rate": 4.321953071059976e-06,
+      "loss": 0.582,
+      "step": 5501
+    },
+    {
+      "epoch": 1.460639851320855,
+      "grad_norm": 0.4110460054913804,
+      "learning_rate": 4.321713989136154e-06,
+      "loss": 0.5838,
+      "step": 5502
+    },
+    {
+      "epoch": 1.4609053497942388,
+      "grad_norm": 0.41757126431902813,
+      "learning_rate": 4.321474871684449e-06,
+      "loss": 0.5882,
+      "step": 5503
+    },
+    {
+      "epoch": 1.4611708482676224,
+      "grad_norm": 0.42378763710163636,
+      "learning_rate": 4.321235718709525e-06,
+      "loss": 0.6033,
+      "step": 5504
+    },
+    {
+      "epoch": 1.4614363467410063,
+      "grad_norm": 0.4086764788524873,
+      "learning_rate": 4.320996530216045e-06,
+      "loss": 0.5869,
+      "step": 5505
+    },
+    {
+      "epoch": 1.46170184521439,
+      "grad_norm": 0.4049745795585868,
+      "learning_rate": 4.320757306208674e-06,
+      "loss": 0.5654,
+      "step": 5506
+    },
+    {
+      "epoch": 1.4619673436877738,
+      "grad_norm": 0.3941631829128186,
+      "learning_rate": 4.3205180466920784e-06,
+      "loss": 0.5845,
+      "step": 5507
+    },
+    {
+      "epoch": 1.4622328421611575,
+      "grad_norm": 0.411320031751828,
+      "learning_rate": 4.320278751670922e-06,
+      "loss": 0.5668,
+      "step": 5508
+    },
+    {
+      "epoch": 1.4624983406345413,
+      "grad_norm": 0.4240298495466126,
+      "learning_rate": 4.320039421149874e-06,
+      "loss": 0.6176,
+      "step": 5509
+    },
+    {
+      "epoch": 1.4627638391079252,
+      "grad_norm": 0.42734390919595755,
+      "learning_rate": 4.319800055133602e-06,
+      "loss": 0.5973,
+      "step": 5510
+    },
+    {
+      "epoch": 1.4630293375813088,
+      "grad_norm": 0.4011125372527518,
+      "learning_rate": 4.319560653626772e-06,
+      "loss": 0.5421,
+      "step": 5511
+    },
+    {
+      "epoch": 1.4632948360546927,
+      "grad_norm": 0.40815814109542436,
+      "learning_rate": 4.3193212166340535e-06,
+      "loss": 0.5808,
+      "step": 5512
+    },
+    {
+      "epoch": 1.4635603345280765,
+      "grad_norm": 0.40871228740343035,
+      "learning_rate": 4.319081744160119e-06,
+      "loss": 0.5859,
+      "step": 5513
+    },
+    {
+      "epoch": 1.4638258330014602,
+      "grad_norm": 0.4074011815085495,
+      "learning_rate": 4.318842236209635e-06,
+      "loss": 0.6056,
+      "step": 5514
+    },
+    {
+      "epoch": 1.464091331474844,
+      "grad_norm": 0.4052406086605638,
+      "learning_rate": 4.3186026927872736e-06,
+      "loss": 0.6182,
+      "step": 5515
+    },
+    {
+      "epoch": 1.464356829948228,
+      "grad_norm": 0.4008152565339012,
+      "learning_rate": 4.318363113897709e-06,
+      "loss": 0.5776,
+      "step": 5516
+    },
+    {
+      "epoch": 1.4646223284216116,
+      "grad_norm": 0.4180290683632047,
+      "learning_rate": 4.31812349954561e-06,
+      "loss": 0.5684,
+      "step": 5517
+    },
+    {
+      "epoch": 1.4648878268949954,
+      "grad_norm": 0.4151488415050085,
+      "learning_rate": 4.3178838497356515e-06,
+      "loss": 0.5506,
+      "step": 5518
+    },
+    {
+      "epoch": 1.465153325368379,
+      "grad_norm": 0.40188978367624584,
+      "learning_rate": 4.317644164472508e-06,
+      "loss": 0.6391,
+      "step": 5519
+    },
+    {
+      "epoch": 1.465418823841763,
+      "grad_norm": 0.40669239146877795,
+      "learning_rate": 4.317404443760851e-06,
+      "loss": 0.6353,
+      "step": 5520
+    },
+    {
+      "epoch": 1.4656843223151466,
+      "grad_norm": 0.42104072195555886,
+      "learning_rate": 4.317164687605359e-06,
+      "loss": 0.6036,
+      "step": 5521
+    },
+    {
+      "epoch": 1.4659498207885304,
+      "grad_norm": 0.399939047938581,
+      "learning_rate": 4.3169248960107054e-06,
+      "loss": 0.5626,
+      "step": 5522
+    },
+    {
+      "epoch": 1.4662153192619143,
+      "grad_norm": 0.4260673896839088,
+      "learning_rate": 4.316685068981569e-06,
+      "loss": 0.5985,
+      "step": 5523
+    },
+    {
+      "epoch": 1.466480817735298,
+      "grad_norm": 0.40033047321187815,
+      "learning_rate": 4.316445206522625e-06,
+      "loss": 0.5901,
+      "step": 5524
+    },
+    {
+      "epoch": 1.4667463162086818,
+      "grad_norm": 0.4163852826881481,
+      "learning_rate": 4.3162053086385515e-06,
+      "loss": 0.6048,
+      "step": 5525
+    },
+    {
+      "epoch": 1.4670118146820657,
+      "grad_norm": 0.41807279093668753,
+      "learning_rate": 4.315965375334029e-06,
+      "loss": 0.6199,
+      "step": 5526
+    },
+    {
+      "epoch": 1.4672773131554493,
+      "grad_norm": 0.39260821331302076,
+      "learning_rate": 4.315725406613735e-06,
+      "loss": 0.5873,
+      "step": 5527
+    },
+    {
+      "epoch": 1.4675428116288332,
+      "grad_norm": 0.4226695052785415,
+      "learning_rate": 4.315485402482349e-06,
+      "loss": 0.5824,
+      "step": 5528
+    },
+    {
+      "epoch": 1.467808310102217,
+      "grad_norm": 0.4210150617194616,
+      "learning_rate": 4.315245362944553e-06,
+      "loss": 0.6075,
+      "step": 5529
+    },
+    {
+      "epoch": 1.4680738085756007,
+      "grad_norm": 0.4140100363396281,
+      "learning_rate": 4.315005288005029e-06,
+      "loss": 0.5696,
+      "step": 5530
+    },
+    {
+      "epoch": 1.4683393070489845,
+      "grad_norm": 0.402729836590436,
+      "learning_rate": 4.314765177668456e-06,
+      "loss": 0.5956,
+      "step": 5531
+    },
+    {
+      "epoch": 1.4686048055223684,
+      "grad_norm": 0.4068745389019842,
+      "learning_rate": 4.31452503193952e-06,
+      "loss": 0.5878,
+      "step": 5532
+    },
+    {
+      "epoch": 1.468870303995752,
+      "grad_norm": 0.41315535902769135,
+      "learning_rate": 4.314284850822904e-06,
+      "loss": 0.5948,
+      "step": 5533
+    },
+    {
+      "epoch": 1.4691358024691357,
+      "grad_norm": 0.3973011188192901,
+      "learning_rate": 4.31404463432329e-06,
+      "loss": 0.5965,
+      "step": 5534
+    },
+    {
+      "epoch": 1.4694013009425195,
+      "grad_norm": 0.4007996448295039,
+      "learning_rate": 4.313804382445364e-06,
+      "loss": 0.601,
+      "step": 5535
+    },
+    {
+      "epoch": 1.4696667994159034,
+      "grad_norm": 0.41312856893792915,
+      "learning_rate": 4.313564095193812e-06,
+      "loss": 0.5947,
+      "step": 5536
+    },
+    {
+      "epoch": 1.469932297889287,
+      "grad_norm": 0.41232622730959095,
+      "learning_rate": 4.31332377257332e-06,
+      "loss": 0.5871,
+      "step": 5537
+    },
+    {
+      "epoch": 1.470197796362671,
+      "grad_norm": 0.41991647379237185,
+      "learning_rate": 4.313083414588574e-06,
+      "loss": 0.5644,
+      "step": 5538
+    },
+    {
+      "epoch": 1.4704632948360548,
+      "grad_norm": 0.3912478548717292,
+      "learning_rate": 4.312843021244263e-06,
+      "loss": 0.6015,
+      "step": 5539
+    },
+    {
+      "epoch": 1.4707287933094384,
+      "grad_norm": 0.40103338117562687,
+      "learning_rate": 4.312602592545073e-06,
+      "loss": 0.5993,
+      "step": 5540
+    },
+    {
+      "epoch": 1.4709942917828223,
+      "grad_norm": 0.41101252231516816,
+      "learning_rate": 4.312362128495697e-06,
+      "loss": 0.5663,
+      "step": 5541
+    },
+    {
+      "epoch": 1.4712597902562061,
+      "grad_norm": 0.40098559661083966,
+      "learning_rate": 4.31212162910082e-06,
+      "loss": 0.5907,
+      "step": 5542
+    },
+    {
+      "epoch": 1.4715252887295898,
+      "grad_norm": 0.4074430144974285,
+      "learning_rate": 4.311881094365134e-06,
+      "loss": 0.5553,
+      "step": 5543
+    },
+    {
+      "epoch": 1.4717907872029736,
+      "grad_norm": 0.39736619126303796,
+      "learning_rate": 4.311640524293332e-06,
+      "loss": 0.5693,
+      "step": 5544
+    },
+    {
+      "epoch": 1.4720562856763575,
+      "grad_norm": 0.41380392173836616,
+      "learning_rate": 4.3113999188901036e-06,
+      "loss": 0.5846,
+      "step": 5545
+    },
+    {
+      "epoch": 1.4723217841497411,
+      "grad_norm": 0.408249144675621,
+      "learning_rate": 4.311159278160142e-06,
+      "loss": 0.541,
+      "step": 5546
+    },
+    {
+      "epoch": 1.472587282623125,
+      "grad_norm": 0.4062292276377927,
+      "learning_rate": 4.31091860210814e-06,
+      "loss": 0.5312,
+      "step": 5547
+    },
+    {
+      "epoch": 1.4728527810965086,
+      "grad_norm": 0.4076363580452722,
+      "learning_rate": 4.31067789073879e-06,
+      "loss": 0.6293,
+      "step": 5548
+    },
+    {
+      "epoch": 1.4731182795698925,
+      "grad_norm": 0.4171891999021032,
+      "learning_rate": 4.31043714405679e-06,
+      "loss": 0.5758,
+      "step": 5549
+    },
+    {
+      "epoch": 1.4733837780432761,
+      "grad_norm": 0.40277192134787976,
+      "learning_rate": 4.310196362066831e-06,
+      "loss": 0.5979,
+      "step": 5550
+    },
+    {
+      "epoch": 1.47364927651666,
+      "grad_norm": 0.4177233212049592,
+      "learning_rate": 4.309955544773612e-06,
+      "loss": 0.5682,
+      "step": 5551
+    },
+    {
+      "epoch": 1.4739147749900439,
+      "grad_norm": 0.4102520481735466,
+      "learning_rate": 4.309714692181827e-06,
+      "loss": 0.623,
+      "step": 5552
+    },
+    {
+      "epoch": 1.4741802734634275,
+      "grad_norm": 0.39713997465951567,
+      "learning_rate": 4.3094738042961755e-06,
+      "loss": 0.5825,
+      "step": 5553
+    },
+    {
+      "epoch": 1.4744457719368114,
+      "grad_norm": 0.40795216533220846,
+      "learning_rate": 4.309232881121355e-06,
+      "loss": 0.5722,
+      "step": 5554
+    },
+    {
+      "epoch": 1.4747112704101952,
+      "grad_norm": 0.41206610467356236,
+      "learning_rate": 4.308991922662063e-06,
+      "loss": 0.5992,
+      "step": 5555
+    },
+    {
+      "epoch": 1.4749767688835789,
+      "grad_norm": 0.40942208276612935,
+      "learning_rate": 4.308750928922999e-06,
+      "loss": 0.5878,
+      "step": 5556
+    },
+    {
+      "epoch": 1.4752422673569627,
+      "grad_norm": 0.3881866358538866,
+      "learning_rate": 4.308509899908864e-06,
+      "loss": 0.6051,
+      "step": 5557
+    },
+    {
+      "epoch": 1.4755077658303466,
+      "grad_norm": 0.4133789310269571,
+      "learning_rate": 4.308268835624358e-06,
+      "loss": 0.6081,
+      "step": 5558
+    },
+    {
+      "epoch": 1.4757732643037302,
+      "grad_norm": 0.3975243911554642,
+      "learning_rate": 4.3080277360741805e-06,
+      "loss": 0.6048,
+      "step": 5559
+    },
+    {
+      "epoch": 1.476038762777114,
+      "grad_norm": 0.41559021125329,
+      "learning_rate": 4.307786601263037e-06,
+      "loss": 0.5606,
+      "step": 5560
+    },
+    {
+      "epoch": 1.4763042612504977,
+      "grad_norm": 0.405777975127508,
+      "learning_rate": 4.307545431195629e-06,
+      "loss": 0.6047,
+      "step": 5561
+    },
+    {
+      "epoch": 1.4765697597238816,
+      "grad_norm": 0.39733511192242815,
+      "learning_rate": 4.3073042258766586e-06,
+      "loss": 0.6069,
+      "step": 5562
+    },
+    {
+      "epoch": 1.4768352581972652,
+      "grad_norm": 0.41274048030382415,
+      "learning_rate": 4.30706298531083e-06,
+      "loss": 0.5628,
+      "step": 5563
+    },
+    {
+      "epoch": 1.477100756670649,
+      "grad_norm": 0.39774215958369075,
+      "learning_rate": 4.30682170950285e-06,
+      "loss": 0.5775,
+      "step": 5564
+    },
+    {
+      "epoch": 1.477366255144033,
+      "grad_norm": 0.4076207794860899,
+      "learning_rate": 4.306580398457423e-06,
+      "loss": 0.5623,
+      "step": 5565
+    },
+    {
+      "epoch": 1.4776317536174166,
+      "grad_norm": 0.39960304642648214,
+      "learning_rate": 4.306339052179254e-06,
+      "loss": 0.5978,
+      "step": 5566
+    },
+    {
+      "epoch": 1.4778972520908005,
+      "grad_norm": 0.41066224982186994,
+      "learning_rate": 4.306097670673051e-06,
+      "loss": 0.5464,
+      "step": 5567
+    },
+    {
+      "epoch": 1.4781627505641843,
+      "grad_norm": 0.41006629351145496,
+      "learning_rate": 4.305856253943522e-06,
+      "loss": 0.5546,
+      "step": 5568
+    },
+    {
+      "epoch": 1.478428249037568,
+      "grad_norm": 0.4015994292117701,
+      "learning_rate": 4.3056148019953735e-06,
+      "loss": 0.608,
+      "step": 5569
+    },
+    {
+      "epoch": 1.4786937475109518,
+      "grad_norm": 0.40961989751249295,
+      "learning_rate": 4.305373314833317e-06,
+      "loss": 0.5865,
+      "step": 5570
+    },
+    {
+      "epoch": 1.4789592459843357,
+      "grad_norm": 0.39969620914788334,
+      "learning_rate": 4.305131792462058e-06,
+      "loss": 0.5941,
+      "step": 5571
+    },
+    {
+      "epoch": 1.4792247444577193,
+      "grad_norm": 0.41959562060935424,
+      "learning_rate": 4.3048902348863116e-06,
+      "loss": 0.592,
+      "step": 5572
+    },
+    {
+      "epoch": 1.4794902429311032,
+      "grad_norm": 0.39824942342312053,
+      "learning_rate": 4.3046486421107854e-06,
+      "loss": 0.5705,
+      "step": 5573
+    },
+    {
+      "epoch": 1.4797557414044868,
+      "grad_norm": 0.41327026109964204,
+      "learning_rate": 4.3044070141401925e-06,
+      "loss": 0.5912,
+      "step": 5574
+    },
+    {
+      "epoch": 1.4800212398778707,
+      "grad_norm": 0.39486956605967033,
+      "learning_rate": 4.304165350979246e-06,
+      "loss": 0.6338,
+      "step": 5575
+    },
+    {
+      "epoch": 1.4802867383512543,
+      "grad_norm": 0.4158632378033139,
+      "learning_rate": 4.303923652632656e-06,
+      "loss": 0.6324,
+      "step": 5576
+    },
+    {
+      "epoch": 1.4805522368246382,
+      "grad_norm": 0.40948179239793003,
+      "learning_rate": 4.303681919105138e-06,
+      "loss": 0.5849,
+      "step": 5577
+    },
+    {
+      "epoch": 1.480817735298022,
+      "grad_norm": 0.41436358661688566,
+      "learning_rate": 4.3034401504014076e-06,
+      "loss": 0.6025,
+      "step": 5578
+    },
+    {
+      "epoch": 1.4810832337714057,
+      "grad_norm": 0.41119008776859284,
+      "learning_rate": 4.303198346526178e-06,
+      "loss": 0.5943,
+      "step": 5579
+    },
+    {
+      "epoch": 1.4813487322447896,
+      "grad_norm": 0.4043209080477481,
+      "learning_rate": 4.302956507484166e-06,
+      "loss": 0.5538,
+      "step": 5580
+    },
+    {
+      "epoch": 1.4816142307181734,
+      "grad_norm": 0.43261837570592837,
+      "learning_rate": 4.302714633280087e-06,
+      "loss": 0.5961,
+      "step": 5581
+    },
+    {
+      "epoch": 1.481879729191557,
+      "grad_norm": 0.40956763973232396,
+      "learning_rate": 4.30247272391866e-06,
+      "loss": 0.5865,
+      "step": 5582
+    },
+    {
+      "epoch": 1.482145227664941,
+      "grad_norm": 0.4113442762061393,
+      "learning_rate": 4.3022307794046006e-06,
+      "loss": 0.6032,
+      "step": 5583
+    },
+    {
+      "epoch": 1.4824107261383248,
+      "grad_norm": 0.4342482374146037,
+      "learning_rate": 4.301988799742629e-06,
+      "loss": 0.5536,
+      "step": 5584
+    },
+    {
+      "epoch": 1.4826762246117084,
+      "grad_norm": 0.41119700625275757,
+      "learning_rate": 4.301746784937464e-06,
+      "loss": 0.5671,
+      "step": 5585
+    },
+    {
+      "epoch": 1.4829417230850923,
+      "grad_norm": 0.4052253487530817,
+      "learning_rate": 4.3015047349938244e-06,
+      "loss": 0.5695,
+      "step": 5586
+    },
+    {
+      "epoch": 1.4832072215584762,
+      "grad_norm": 0.41187273250251494,
+      "learning_rate": 4.301262649916432e-06,
+      "loss": 0.5651,
+      "step": 5587
+    },
+    {
+      "epoch": 1.4834727200318598,
+      "grad_norm": 0.4095504145663514,
+      "learning_rate": 4.301020529710009e-06,
+      "loss": 0.5888,
+      "step": 5588
+    },
+    {
+      "epoch": 1.4837382185052437,
+      "grad_norm": 0.4113066764240779,
+      "learning_rate": 4.3007783743792735e-06,
+      "loss": 0.5786,
+      "step": 5589
+    },
+    {
+      "epoch": 1.4840037169786273,
+      "grad_norm": 0.42666649927641165,
+      "learning_rate": 4.3005361839289514e-06,
+      "loss": 0.5807,
+      "step": 5590
+    },
+    {
+      "epoch": 1.4842692154520112,
+      "grad_norm": 0.4091737005822595,
+      "learning_rate": 4.300293958363766e-06,
+      "loss": 0.588,
+      "step": 5591
+    },
+    {
+      "epoch": 1.4845347139253948,
+      "grad_norm": 0.4026541211913795,
+      "learning_rate": 4.3000516976884406e-06,
+      "loss": 0.5903,
+      "step": 5592
+    },
+    {
+      "epoch": 1.4848002123987787,
+      "grad_norm": 0.4132137671109537,
+      "learning_rate": 4.2998094019076996e-06,
+      "loss": 0.6284,
+      "step": 5593
+    },
+    {
+      "epoch": 1.4850657108721625,
+      "grad_norm": 0.4172679460127461,
+      "learning_rate": 4.299567071026268e-06,
+      "loss": 0.6206,
+      "step": 5594
+    },
+    {
+      "epoch": 1.4853312093455462,
+      "grad_norm": 0.4103474329947075,
+      "learning_rate": 4.299324705048873e-06,
+      "loss": 0.5795,
+      "step": 5595
+    },
+    {
+      "epoch": 1.48559670781893,
+      "grad_norm": 0.4097596880368327,
+      "learning_rate": 4.299082303980241e-06,
+      "loss": 0.601,
+      "step": 5596
+    },
+    {
+      "epoch": 1.485862206292314,
+      "grad_norm": 0.4211418978380807,
+      "learning_rate": 4.298839867825098e-06,
+      "loss": 0.598,
+      "step": 5597
+    },
+    {
+      "epoch": 1.4861277047656976,
+      "grad_norm": 0.39386632711745845,
+      "learning_rate": 4.298597396588175e-06,
+      "loss": 0.5865,
+      "step": 5598
+    },
+    {
+      "epoch": 1.4863932032390814,
+      "grad_norm": 0.43396563815849093,
+      "learning_rate": 4.298354890274198e-06,
+      "loss": 0.6054,
+      "step": 5599
+    },
+    {
+      "epoch": 1.4866587017124653,
+      "grad_norm": 0.41935494872928014,
+      "learning_rate": 4.298112348887897e-06,
+      "loss": 0.5791,
+      "step": 5600
+    },
+    {
+      "epoch": 1.486924200185849,
+      "grad_norm": 0.4075637480750658,
+      "learning_rate": 4.297869772434004e-06,
+      "loss": 0.5909,
+      "step": 5601
+    },
+    {
+      "epoch": 1.4871896986592328,
+      "grad_norm": 0.417287421316621,
+      "learning_rate": 4.297627160917247e-06,
+      "loss": 0.5819,
+      "step": 5602
+    },
+    {
+      "epoch": 1.4874551971326164,
+      "grad_norm": 0.4189200875337072,
+      "learning_rate": 4.29738451434236e-06,
+      "loss": 0.5842,
+      "step": 5603
+    },
+    {
+      "epoch": 1.4877206956060003,
+      "grad_norm": 0.4133173901574044,
+      "learning_rate": 4.297141832714073e-06,
+      "loss": 0.5375,
+      "step": 5604
+    },
+    {
+      "epoch": 1.487986194079384,
+      "grad_norm": 0.4153016247916373,
+      "learning_rate": 4.296899116037122e-06,
+      "loss": 0.5975,
+      "step": 5605
+    },
+    {
+      "epoch": 1.4882516925527678,
+      "grad_norm": 0.3949821939060139,
+      "learning_rate": 4.296656364316238e-06,
+      "loss": 0.5388,
+      "step": 5606
+    },
+    {
+      "epoch": 1.4885171910261517,
+      "grad_norm": 0.40581855738849826,
+      "learning_rate": 4.296413577556155e-06,
+      "loss": 0.5976,
+      "step": 5607
+    },
+    {
+      "epoch": 1.4887826894995353,
+      "grad_norm": 0.4024073387300952,
+      "learning_rate": 4.296170755761609e-06,
+      "loss": 0.5855,
+      "step": 5608
+    },
+    {
+      "epoch": 1.4890481879729192,
+      "grad_norm": 0.41911761598701847,
+      "learning_rate": 4.295927898937337e-06,
+      "loss": 0.5958,
+      "step": 5609
+    },
+    {
+      "epoch": 1.489313686446303,
+      "grad_norm": 0.4240188732259072,
+      "learning_rate": 4.2956850070880725e-06,
+      "loss": 0.602,
+      "step": 5610
+    },
+    {
+      "epoch": 1.4895791849196867,
+      "grad_norm": 0.4084060144334041,
+      "learning_rate": 4.295442080218554e-06,
+      "loss": 0.603,
+      "step": 5611
+    },
+    {
+      "epoch": 1.4898446833930705,
+      "grad_norm": 0.41550603533829994,
+      "learning_rate": 4.295199118333519e-06,
+      "loss": 0.577,
+      "step": 5612
+    },
+    {
+      "epoch": 1.4901101818664544,
+      "grad_norm": 0.4073914364998427,
+      "learning_rate": 4.294956121437706e-06,
+      "loss": 0.5901,
+      "step": 5613
+    },
+    {
+      "epoch": 1.490375680339838,
+      "grad_norm": 0.3979286568584305,
+      "learning_rate": 4.2947130895358535e-06,
+      "loss": 0.6033,
+      "step": 5614
+    },
+    {
+      "epoch": 1.490641178813222,
+      "grad_norm": 0.4065132966537831,
+      "learning_rate": 4.2944700226327014e-06,
+      "loss": 0.5816,
+      "step": 5615
+    },
+    {
+      "epoch": 1.4909066772866055,
+      "grad_norm": 0.41421185621331674,
+      "learning_rate": 4.29422692073299e-06,
+      "loss": 0.6022,
+      "step": 5616
+    },
+    {
+      "epoch": 1.4911721757599894,
+      "grad_norm": 0.41242826486158735,
+      "learning_rate": 4.293983783841461e-06,
+      "loss": 0.5619,
+      "step": 5617
+    },
+    {
+      "epoch": 1.491437674233373,
+      "grad_norm": 0.4019956423160651,
+      "learning_rate": 4.293740611962856e-06,
+      "loss": 0.6135,
+      "step": 5618
+    },
+    {
+      "epoch": 1.491703172706757,
+      "grad_norm": 0.39929011099356776,
+      "learning_rate": 4.293497405101917e-06,
+      "loss": 0.5938,
+      "step": 5619
+    },
+    {
+      "epoch": 1.4919686711801408,
+      "grad_norm": 0.39606208367172996,
+      "learning_rate": 4.293254163263388e-06,
+      "loss": 0.5931,
+      "step": 5620
+    },
+    {
+      "epoch": 1.4922341696535244,
+      "grad_norm": 0.41259615063719823,
+      "learning_rate": 4.293010886452012e-06,
+      "loss": 0.5544,
+      "step": 5621
+    },
+    {
+      "epoch": 1.4924996681269083,
+      "grad_norm": 0.41513473088555203,
+      "learning_rate": 4.292767574672532e-06,
+      "loss": 0.5432,
+      "step": 5622
+    },
+    {
+      "epoch": 1.4927651666002921,
+      "grad_norm": 0.4007579597252526,
+      "learning_rate": 4.2925242279296965e-06,
+      "loss": 0.5993,
+      "step": 5623
+    },
+    {
+      "epoch": 1.4930306650736758,
+      "grad_norm": 0.43662004007940564,
+      "learning_rate": 4.29228084622825e-06,
+      "loss": 0.5723,
+      "step": 5624
+    },
+    {
+      "epoch": 1.4932961635470596,
+      "grad_norm": 0.42985856343764267,
+      "learning_rate": 4.292037429572937e-06,
+      "loss": 0.6088,
+      "step": 5625
+    },
+    {
+      "epoch": 1.4935616620204435,
+      "grad_norm": 0.4215415654202423,
+      "learning_rate": 4.291793977968508e-06,
+      "loss": 0.5864,
+      "step": 5626
+    },
+    {
+      "epoch": 1.4938271604938271,
+      "grad_norm": 0.40658037424266164,
+      "learning_rate": 4.291550491419709e-06,
+      "loss": 0.6058,
+      "step": 5627
+    },
+    {
+      "epoch": 1.494092658967211,
+      "grad_norm": 0.41499367657110897,
+      "learning_rate": 4.291306969931289e-06,
+      "loss": 0.5511,
+      "step": 5628
+    },
+    {
+      "epoch": 1.4943581574405946,
+      "grad_norm": 0.4171361813430919,
+      "learning_rate": 4.291063413507997e-06,
+      "loss": 0.5857,
+      "step": 5629
+    },
+    {
+      "epoch": 1.4946236559139785,
+      "grad_norm": 0.4032400113309654,
+      "learning_rate": 4.2908198221545824e-06,
+      "loss": 0.6097,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4948891543873621,
+      "grad_norm": 0.43116954204269214,
+      "learning_rate": 4.2905761958757975e-06,
+      "loss": 0.5851,
+      "step": 5631
+    },
+    {
+      "epoch": 1.495154652860746,
+      "grad_norm": 0.4099033401462102,
+      "learning_rate": 4.290332534676393e-06,
+      "loss": 0.6219,
+      "step": 5632
+    },
+    {
+      "epoch": 1.4954201513341299,
+      "grad_norm": 0.4225885236505941,
+      "learning_rate": 4.2900888385611205e-06,
+      "loss": 0.5766,
+      "step": 5633
+    },
+    {
+      "epoch": 1.4956856498075135,
+      "grad_norm": 0.4261721246243504,
+      "learning_rate": 4.289845107534732e-06,
+      "loss": 0.5972,
+      "step": 5634
+    },
+    {
+      "epoch": 1.4959511482808974,
+      "grad_norm": 0.43151428229846045,
+      "learning_rate": 4.289601341601982e-06,
+      "loss": 0.5847,
+      "step": 5635
+    },
+    {
+      "epoch": 1.4962166467542812,
+      "grad_norm": 0.397997685381986,
+      "learning_rate": 4.289357540767625e-06,
+      "loss": 0.5915,
+      "step": 5636
+    },
+    {
+      "epoch": 1.4964821452276649,
+      "grad_norm": 0.40833929534906654,
+      "learning_rate": 4.289113705036413e-06,
+      "loss": 0.5814,
+      "step": 5637
+    },
+    {
+      "epoch": 1.4967476437010487,
+      "grad_norm": 0.40920675231347575,
+      "learning_rate": 4.288869834413104e-06,
+      "loss": 0.6153,
+      "step": 5638
+    },
+    {
+      "epoch": 1.4970131421744326,
+      "grad_norm": 0.4001000938239612,
+      "learning_rate": 4.288625928902454e-06,
+      "loss": 0.5514,
+      "step": 5639
+    },
+    {
+      "epoch": 1.4972786406478162,
+      "grad_norm": 0.4003037292230063,
+      "learning_rate": 4.288381988509219e-06,
+      "loss": 0.5851,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4975441391212,
+      "grad_norm": 0.4093169466843017,
+      "learning_rate": 4.288138013238155e-06,
+      "loss": 0.6148,
+      "step": 5641
+    },
+    {
+      "epoch": 1.497809637594584,
+      "grad_norm": 0.41305181203290936,
+      "learning_rate": 4.287894003094024e-06,
+      "loss": 0.5959,
+      "step": 5642
+    },
+    {
+      "epoch": 1.4980751360679676,
+      "grad_norm": 0.40953119504945756,
+      "learning_rate": 4.2876499580815816e-06,
+      "loss": 0.6023,
+      "step": 5643
+    },
+    {
+      "epoch": 1.4983406345413515,
+      "grad_norm": 0.4269553731026164,
+      "learning_rate": 4.287405878205587e-06,
+      "loss": 0.5959,
+      "step": 5644
+    },
+    {
+      "epoch": 1.498606133014735,
+      "grad_norm": 0.4038855634942212,
+      "learning_rate": 4.287161763470802e-06,
+      "loss": 0.5732,
+      "step": 5645
+    },
+    {
+      "epoch": 1.498871631488119,
+      "grad_norm": 0.41406221287188316,
+      "learning_rate": 4.286917613881987e-06,
+      "loss": 0.6278,
+      "step": 5646
+    },
+    {
+      "epoch": 1.4991371299615026,
+      "grad_norm": 0.40379250161406066,
+      "learning_rate": 4.286673429443904e-06,
+      "loss": 0.5852,
+      "step": 5647
+    },
+    {
+      "epoch": 1.4994026284348865,
+      "grad_norm": 0.4069517055843132,
+      "learning_rate": 4.286429210161313e-06,
+      "loss": 0.6056,
+      "step": 5648
+    },
+    {
+      "epoch": 1.4996681269082703,
+      "grad_norm": 0.3965435612616709,
+      "learning_rate": 4.28618495603898e-06,
+      "loss": 0.613,
+      "step": 5649
+    },
+    {
+      "epoch": 1.499933625381654,
+      "grad_norm": 0.40007288627181575,
+      "learning_rate": 4.285940667081667e-06,
+      "loss": 0.5757,
+      "step": 5650
+    },
+    {
+      "epoch": 1.5001991238550378,
+      "grad_norm": 0.39495602156663484,
+      "learning_rate": 4.2856963432941376e-06,
+      "loss": 0.5975,
+      "step": 5651
+    },
+    {
+      "epoch": 1.5004646223284217,
+      "grad_norm": 0.39340035894210856,
+      "learning_rate": 4.285451984681157e-06,
+      "loss": 0.5915,
+      "step": 5652
+    },
+    {
+      "epoch": 1.5007301208018053,
+      "grad_norm": 0.40184427972918263,
+      "learning_rate": 4.285207591247492e-06,
+      "loss": 0.5773,
+      "step": 5653
+    },
+    {
+      "epoch": 1.5009956192751892,
+      "grad_norm": 0.4057439125318073,
+      "learning_rate": 4.284963162997908e-06,
+      "loss": 0.5936,
+      "step": 5654
+    },
+    {
+      "epoch": 1.501261117748573,
+      "grad_norm": 0.3986417073629942,
+      "learning_rate": 4.284718699937172e-06,
+      "loss": 0.5497,
+      "step": 5655
+    },
+    {
+      "epoch": 1.5015266162219567,
+      "grad_norm": 0.415730078997179,
+      "learning_rate": 4.284474202070052e-06,
+      "loss": 0.5706,
+      "step": 5656
+    },
+    {
+      "epoch": 1.5017921146953404,
+      "grad_norm": 0.39952821522491205,
+      "learning_rate": 4.284229669401314e-06,
+      "loss": 0.5805,
+      "step": 5657
+    },
+    {
+      "epoch": 1.5020576131687244,
+      "grad_norm": 0.42519066384807086,
+      "learning_rate": 4.28398510193573e-06,
+      "loss": 0.5801,
+      "step": 5658
+    },
+    {
+      "epoch": 1.502323111642108,
+      "grad_norm": 0.4308340176687171,
+      "learning_rate": 4.283740499678069e-06,
+      "loss": 0.5639,
+      "step": 5659
+    },
+    {
+      "epoch": 1.5025886101154917,
+      "grad_norm": 0.4123144627485876,
+      "learning_rate": 4.2834958626331006e-06,
+      "loss": 0.6236,
+      "step": 5660
+    },
+    {
+      "epoch": 1.5028541085888756,
+      "grad_norm": 0.42368446770308654,
+      "learning_rate": 4.283251190805596e-06,
+      "loss": 0.582,
+      "step": 5661
+    },
+    {
+      "epoch": 1.5031196070622594,
+      "grad_norm": 0.4292458098889344,
+      "learning_rate": 4.283006484200327e-06,
+      "loss": 0.618,
+      "step": 5662
+    },
+    {
+      "epoch": 1.503385105535643,
+      "grad_norm": 0.4420354203264926,
+      "learning_rate": 4.282761742822066e-06,
+      "loss": 0.5935,
+      "step": 5663
+    },
+    {
+      "epoch": 1.503650604009027,
+      "grad_norm": 0.4040574009279775,
+      "learning_rate": 4.282516966675586e-06,
+      "loss": 0.5504,
+      "step": 5664
+    },
+    {
+      "epoch": 1.5039161024824108,
+      "grad_norm": 0.4161243728781349,
+      "learning_rate": 4.282272155765661e-06,
+      "loss": 0.582,
+      "step": 5665
+    },
+    {
+      "epoch": 1.5041816009557945,
+      "grad_norm": 0.42018307719944964,
+      "learning_rate": 4.282027310097066e-06,
+      "loss": 0.5947,
+      "step": 5666
+    },
+    {
+      "epoch": 1.5044470994291783,
+      "grad_norm": 0.41669890656749975,
+      "learning_rate": 4.281782429674574e-06,
+      "loss": 0.5719,
+      "step": 5667
+    },
+    {
+      "epoch": 1.5047125979025622,
+      "grad_norm": 0.43574601468676943,
+      "learning_rate": 4.281537514502962e-06,
+      "loss": 0.5602,
+      "step": 5668
+    },
+    {
+      "epoch": 1.5049780963759458,
+      "grad_norm": 0.4146290532136355,
+      "learning_rate": 4.281292564587009e-06,
+      "loss": 0.5518,
+      "step": 5669
+    },
+    {
+      "epoch": 1.5052435948493295,
+      "grad_norm": 0.43743126605669674,
+      "learning_rate": 4.2810475799314864e-06,
+      "loss": 0.6038,
+      "step": 5670
+    },
+    {
+      "epoch": 1.5055090933227135,
+      "grad_norm": 0.4222172900325046,
+      "learning_rate": 4.2808025605411775e-06,
+      "loss": 0.6315,
+      "step": 5671
+    },
+    {
+      "epoch": 1.5057745917960972,
+      "grad_norm": 0.41348021273746804,
+      "learning_rate": 4.280557506420857e-06,
+      "loss": 0.5809,
+      "step": 5672
+    },
+    {
+      "epoch": 1.5060400902694808,
+      "grad_norm": 0.41289531247665456,
+      "learning_rate": 4.2803124175753064e-06,
+      "loss": 0.602,
+      "step": 5673
+    },
+    {
+      "epoch": 1.5063055887428647,
+      "grad_norm": 0.4180819198976172,
+      "learning_rate": 4.280067294009305e-06,
+      "loss": 0.5733,
+      "step": 5674
+    },
+    {
+      "epoch": 1.5065710872162486,
+      "grad_norm": 0.40901871899706793,
+      "learning_rate": 4.279822135727631e-06,
+      "loss": 0.5808,
+      "step": 5675
+    },
+    {
+      "epoch": 1.5068365856896322,
+      "grad_norm": 0.4074075843669644,
+      "learning_rate": 4.27957694273507e-06,
+      "loss": 0.6026,
+      "step": 5676
+    },
+    {
+      "epoch": 1.507102084163016,
+      "grad_norm": 0.4089481149917887,
+      "learning_rate": 4.2793317150364e-06,
+      "loss": 0.5734,
+      "step": 5677
+    },
+    {
+      "epoch": 1.5073675826364,
+      "grad_norm": 0.4079017601675781,
+      "learning_rate": 4.279086452636406e-06,
+      "loss": 0.629,
+      "step": 5678
+    },
+    {
+      "epoch": 1.5076330811097836,
+      "grad_norm": 0.40210392018110563,
+      "learning_rate": 4.27884115553987e-06,
+      "loss": 0.6176,
+      "step": 5679
+    },
+    {
+      "epoch": 1.5078985795831674,
+      "grad_norm": 0.4169193803859788,
+      "learning_rate": 4.2785958237515755e-06,
+      "loss": 0.5864,
+      "step": 5680
+    },
+    {
+      "epoch": 1.5081640780565513,
+      "grad_norm": 0.414436725305469,
+      "learning_rate": 4.278350457276308e-06,
+      "loss": 0.58,
+      "step": 5681
+    },
+    {
+      "epoch": 1.508429576529935,
+      "grad_norm": 0.4249506065618703,
+      "learning_rate": 4.2781050561188526e-06,
+      "loss": 0.636,
+      "step": 5682
+    },
+    {
+      "epoch": 1.5086950750033188,
+      "grad_norm": 0.4331827324970614,
+      "learning_rate": 4.277859620283995e-06,
+      "loss": 0.5802,
+      "step": 5683
+    },
+    {
+      "epoch": 1.5089605734767026,
+      "grad_norm": 0.40661131321595523,
+      "learning_rate": 4.277614149776522e-06,
+      "loss": 0.5961,
+      "step": 5684
+    },
+    {
+      "epoch": 1.5092260719500863,
+      "grad_norm": 0.42209748960133614,
+      "learning_rate": 4.27736864460122e-06,
+      "loss": 0.5759,
+      "step": 5685
+    },
+    {
+      "epoch": 1.50949157042347,
+      "grad_norm": 0.40884949798421116,
+      "learning_rate": 4.277123104762878e-06,
+      "loss": 0.6121,
+      "step": 5686
+    },
+    {
+      "epoch": 1.5097570688968538,
+      "grad_norm": 0.4037067491065205,
+      "learning_rate": 4.276877530266284e-06,
+      "loss": 0.5943,
+      "step": 5687
+    },
+    {
+      "epoch": 1.5100225673702377,
+      "grad_norm": 0.4074580698543477,
+      "learning_rate": 4.276631921116229e-06,
+      "loss": 0.6042,
+      "step": 5688
+    },
+    {
+      "epoch": 1.5102880658436213,
+      "grad_norm": 0.4111341664884331,
+      "learning_rate": 4.2763862773175005e-06,
+      "loss": 0.5979,
+      "step": 5689
+    },
+    {
+      "epoch": 1.5105535643170052,
+      "grad_norm": 0.4249869181036253,
+      "learning_rate": 4.276140598874891e-06,
+      "loss": 0.5879,
+      "step": 5690
+    },
+    {
+      "epoch": 1.510819062790389,
+      "grad_norm": 0.4052432796809206,
+      "learning_rate": 4.2758948857931904e-06,
+      "loss": 0.5819,
+      "step": 5691
+    },
+    {
+      "epoch": 1.5110845612637727,
+      "grad_norm": 0.416293544723965,
+      "learning_rate": 4.275649138077191e-06,
+      "loss": 0.5805,
+      "step": 5692
+    },
+    {
+      "epoch": 1.5113500597371565,
+      "grad_norm": 0.41801282225434344,
+      "learning_rate": 4.275403355731687e-06,
+      "loss": 0.6356,
+      "step": 5693
+    },
+    {
+      "epoch": 1.5116155582105404,
+      "grad_norm": 0.4329096586032273,
+      "learning_rate": 4.27515753876147e-06,
+      "loss": 0.5745,
+      "step": 5694
+    },
+    {
+      "epoch": 1.511881056683924,
+      "grad_norm": 0.43026700494688946,
+      "learning_rate": 4.2749116871713345e-06,
+      "loss": 0.6215,
+      "step": 5695
+    },
+    {
+      "epoch": 1.512146555157308,
+      "grad_norm": 0.4087202495096165,
+      "learning_rate": 4.274665800966076e-06,
+      "loss": 0.5805,
+      "step": 5696
+    },
+    {
+      "epoch": 1.5124120536306918,
+      "grad_norm": 0.40682328917061755,
+      "learning_rate": 4.2744198801504875e-06,
+      "loss": 0.5683,
+      "step": 5697
+    },
+    {
+      "epoch": 1.5126775521040754,
+      "grad_norm": 0.39125673378603715,
+      "learning_rate": 4.274173924729369e-06,
+      "loss": 0.5659,
+      "step": 5698
+    },
+    {
+      "epoch": 1.512943050577459,
+      "grad_norm": 0.4119223366222529,
+      "learning_rate": 4.273927934707514e-06,
+      "loss": 0.6062,
+      "step": 5699
+    },
+    {
+      "epoch": 1.5132085490508431,
+      "grad_norm": 0.40881559721118427,
+      "learning_rate": 4.273681910089721e-06,
+      "loss": 0.5922,
+      "step": 5700
+    },
+    {
+      "epoch": 1.5134740475242268,
+      "grad_norm": 0.40717292570261626,
+      "learning_rate": 4.273435850880788e-06,
+      "loss": 0.6317,
+      "step": 5701
+    },
+    {
+      "epoch": 1.5137395459976104,
+      "grad_norm": 0.40586660086368254,
+      "learning_rate": 4.273189757085514e-06,
+      "loss": 0.5974,
+      "step": 5702
+    },
+    {
+      "epoch": 1.5140050444709943,
+      "grad_norm": 0.4144374750707454,
+      "learning_rate": 4.272943628708698e-06,
+      "loss": 0.5633,
+      "step": 5703
+    },
+    {
+      "epoch": 1.5142705429443781,
+      "grad_norm": 0.40129072911171604,
+      "learning_rate": 4.27269746575514e-06,
+      "loss": 0.5963,
+      "step": 5704
+    },
+    {
+      "epoch": 1.5145360414177618,
+      "grad_norm": 0.3923367085830904,
+      "learning_rate": 4.272451268229642e-06,
+      "loss": 0.5636,
+      "step": 5705
+    },
+    {
+      "epoch": 1.5148015398911456,
+      "grad_norm": 0.40976214723567633,
+      "learning_rate": 4.272205036137005e-06,
+      "loss": 0.5965,
+      "step": 5706
+    },
+    {
+      "epoch": 1.5150670383645295,
+      "grad_norm": 0.4182397749786371,
+      "learning_rate": 4.271958769482029e-06,
+      "loss": 0.6112,
+      "step": 5707
+    },
+    {
+      "epoch": 1.5153325368379131,
+      "grad_norm": 0.4114383139556441,
+      "learning_rate": 4.271712468269519e-06,
+      "loss": 0.6215,
+      "step": 5708
+    },
+    {
+      "epoch": 1.515598035311297,
+      "grad_norm": 0.41693883299641066,
+      "learning_rate": 4.271466132504279e-06,
+      "loss": 0.5964,
+      "step": 5709
+    },
+    {
+      "epoch": 1.5158635337846809,
+      "grad_norm": 0.40690146982240954,
+      "learning_rate": 4.271219762191111e-06,
+      "loss": 0.6151,
+      "step": 5710
+    },
+    {
+      "epoch": 1.5161290322580645,
+      "grad_norm": 0.42503877756829644,
+      "learning_rate": 4.270973357334821e-06,
+      "loss": 0.5879,
+      "step": 5711
+    },
+    {
+      "epoch": 1.5163945307314481,
+      "grad_norm": 0.40951056623422405,
+      "learning_rate": 4.270726917940216e-06,
+      "loss": 0.6076,
+      "step": 5712
+    },
+    {
+      "epoch": 1.5166600292048322,
+      "grad_norm": 0.42331275595948265,
+      "learning_rate": 4.270480444012098e-06,
+      "loss": 0.5758,
+      "step": 5713
+    },
+    {
+      "epoch": 1.5169255276782159,
+      "grad_norm": 0.40193668472741895,
+      "learning_rate": 4.270233935555278e-06,
+      "loss": 0.5894,
+      "step": 5714
+    },
+    {
+      "epoch": 1.5171910261515995,
+      "grad_norm": 0.39882600344474484,
+      "learning_rate": 4.2699873925745626e-06,
+      "loss": 0.5939,
+      "step": 5715
+    },
+    {
+      "epoch": 1.5174565246249834,
+      "grad_norm": 0.3954701330949738,
+      "learning_rate": 4.269740815074758e-06,
+      "loss": 0.5595,
+      "step": 5716
+    },
+    {
+      "epoch": 1.5177220230983672,
+      "grad_norm": 0.3953214868975649,
+      "learning_rate": 4.269494203060675e-06,
+      "loss": 0.5541,
+      "step": 5717
+    },
+    {
+      "epoch": 1.5179875215717509,
+      "grad_norm": 0.41101915563990105,
+      "learning_rate": 4.269247556537123e-06,
+      "loss": 0.6199,
+      "step": 5718
+    },
+    {
+      "epoch": 1.5182530200451347,
+      "grad_norm": 0.42744969013463696,
+      "learning_rate": 4.269000875508911e-06,
+      "loss": 0.5958,
+      "step": 5719
+    },
+    {
+      "epoch": 1.5185185185185186,
+      "grad_norm": 0.40595400291397254,
+      "learning_rate": 4.268754159980851e-06,
+      "loss": 0.5732,
+      "step": 5720
+    },
+    {
+      "epoch": 1.5187840169919022,
+      "grad_norm": 0.40696951479015514,
+      "learning_rate": 4.2685074099577545e-06,
+      "loss": 0.5584,
+      "step": 5721
+    },
+    {
+      "epoch": 1.519049515465286,
+      "grad_norm": 0.40036171005748317,
+      "learning_rate": 4.268260625444433e-06,
+      "loss": 0.5927,
+      "step": 5722
+    },
+    {
+      "epoch": 1.51931501393867,
+      "grad_norm": 0.4143565478246165,
+      "learning_rate": 4.2680138064457e-06,
+      "loss": 0.5797,
+      "step": 5723
+    },
+    {
+      "epoch": 1.5195805124120536,
+      "grad_norm": 0.404597709231146,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.5663,
+      "step": 5724
+    },
+    {
+      "epoch": 1.5198460108854372,
+      "grad_norm": 0.4157517574357017,
+      "learning_rate": 4.267520065011255e-06,
+      "loss": 0.6102,
+      "step": 5725
+    },
+    {
+      "epoch": 1.5201115093588213,
+      "grad_norm": 0.41114391194162436,
+      "learning_rate": 4.267273142585171e-06,
+      "loss": 0.5724,
+      "step": 5726
+    },
+    {
+      "epoch": 1.520377007832205,
+      "grad_norm": 0.4164424914595022,
+      "learning_rate": 4.2670261856929335e-06,
+      "loss": 0.5973,
+      "step": 5727
+    },
+    {
+      "epoch": 1.5206425063055886,
+      "grad_norm": 0.4249100794165623,
+      "learning_rate": 4.26677919433936e-06,
+      "loss": 0.5673,
+      "step": 5728
+    },
+    {
+      "epoch": 1.5209080047789725,
+      "grad_norm": 0.4101077677346282,
+      "learning_rate": 4.2665321685292655e-06,
+      "loss": 0.5814,
+      "step": 5729
+    },
+    {
+      "epoch": 1.5211735032523563,
+      "grad_norm": 0.4082484490267254,
+      "learning_rate": 4.266285108267468e-06,
+      "loss": 0.5258,
+      "step": 5730
+    },
+    {
+      "epoch": 1.52143900172574,
+      "grad_norm": 0.40682684243895173,
+      "learning_rate": 4.266038013558788e-06,
+      "loss": 0.6033,
+      "step": 5731
+    },
+    {
+      "epoch": 1.5217045001991238,
+      "grad_norm": 0.43040652090542003,
+      "learning_rate": 4.265790884408042e-06,
+      "loss": 0.5905,
+      "step": 5732
+    },
+    {
+      "epoch": 1.5219699986725077,
+      "grad_norm": 0.4087725821179951,
+      "learning_rate": 4.26554372082005e-06,
+      "loss": 0.5415,
+      "step": 5733
+    },
+    {
+      "epoch": 1.5222354971458913,
+      "grad_norm": 0.43616185181429407,
+      "learning_rate": 4.2652965227996325e-06,
+      "loss": 0.5873,
+      "step": 5734
+    },
+    {
+      "epoch": 1.5225009956192752,
+      "grad_norm": 0.40345535276470496,
+      "learning_rate": 4.265049290351611e-06,
+      "loss": 0.5676,
+      "step": 5735
+    },
+    {
+      "epoch": 1.522766494092659,
+      "grad_norm": 0.39985845694822797,
+      "learning_rate": 4.264802023480806e-06,
+      "loss": 0.5905,
+      "step": 5736
+    },
+    {
+      "epoch": 1.5230319925660427,
+      "grad_norm": 0.40959956238084144,
+      "learning_rate": 4.264554722192042e-06,
+      "loss": 0.5812,
+      "step": 5737
+    },
+    {
+      "epoch": 1.5232974910394266,
+      "grad_norm": 0.4003863360931367,
+      "learning_rate": 4.264307386490139e-06,
+      "loss": 0.6032,
+      "step": 5738
+    },
+    {
+      "epoch": 1.5235629895128104,
+      "grad_norm": 0.4246943408432233,
+      "learning_rate": 4.264060016379924e-06,
+      "loss": 0.563,
+      "step": 5739
+    },
+    {
+      "epoch": 1.523828487986194,
+      "grad_norm": 0.40813454209444033,
+      "learning_rate": 4.263812611866218e-06,
+      "loss": 0.5831,
+      "step": 5740
+    },
+    {
+      "epoch": 1.5240939864595777,
+      "grad_norm": 0.4188415604131818,
+      "learning_rate": 4.2635651729538485e-06,
+      "loss": 0.6085,
+      "step": 5741
+    },
+    {
+      "epoch": 1.5243594849329616,
+      "grad_norm": 0.4031768279681957,
+      "learning_rate": 4.263317699647639e-06,
+      "loss": 0.6094,
+      "step": 5742
+    },
+    {
+      "epoch": 1.5246249834063454,
+      "grad_norm": 0.40269882335524293,
+      "learning_rate": 4.263070191952418e-06,
+      "loss": 0.6232,
+      "step": 5743
+    },
+    {
+      "epoch": 1.524890481879729,
+      "grad_norm": 0.41501177730664696,
+      "learning_rate": 4.262822649873011e-06,
+      "loss": 0.6433,
+      "step": 5744
+    },
+    {
+      "epoch": 1.525155980353113,
+      "grad_norm": 0.4033869910141432,
+      "learning_rate": 4.262575073414247e-06,
+      "loss": 0.5687,
+      "step": 5745
+    },
+    {
+      "epoch": 1.5254214788264968,
+      "grad_norm": 0.41613654408175105,
+      "learning_rate": 4.262327462580953e-06,
+      "loss": 0.5986,
+      "step": 5746
+    },
+    {
+      "epoch": 1.5256869772998805,
+      "grad_norm": 0.40565225424344326,
+      "learning_rate": 4.262079817377958e-06,
+      "loss": 0.5849,
+      "step": 5747
+    },
+    {
+      "epoch": 1.5259524757732643,
+      "grad_norm": 0.43132185406208706,
+      "learning_rate": 4.261832137810093e-06,
+      "loss": 0.6044,
+      "step": 5748
+    },
+    {
+      "epoch": 1.5262179742466482,
+      "grad_norm": 0.4225896186641911,
+      "learning_rate": 4.2615844238821864e-06,
+      "loss": 0.603,
+      "step": 5749
+    },
+    {
+      "epoch": 1.5264834727200318,
+      "grad_norm": 0.40717071804314,
+      "learning_rate": 4.261336675599072e-06,
+      "loss": 0.6102,
+      "step": 5750
+    },
+    {
+      "epoch": 1.5267489711934157,
+      "grad_norm": 0.4164106350921646,
+      "learning_rate": 4.261088892965579e-06,
+      "loss": 0.5976,
+      "step": 5751
+    },
+    {
+      "epoch": 1.5270144696667995,
+      "grad_norm": 0.41520198117847645,
+      "learning_rate": 4.260841075986542e-06,
+      "loss": 0.5741,
+      "step": 5752
+    },
+    {
+      "epoch": 1.5272799681401832,
+      "grad_norm": 0.43780824406482194,
+      "learning_rate": 4.2605932246667914e-06,
+      "loss": 0.6134,
+      "step": 5753
+    },
+    {
+      "epoch": 1.5275454666135668,
+      "grad_norm": 0.4050832055131384,
+      "learning_rate": 4.260345339011163e-06,
+      "loss": 0.6281,
+      "step": 5754
+    },
+    {
+      "epoch": 1.527810965086951,
+      "grad_norm": 0.4012649197372533,
+      "learning_rate": 4.2600974190244896e-06,
+      "loss": 0.5465,
+      "step": 5755
+    },
+    {
+      "epoch": 1.5280764635603346,
+      "grad_norm": 0.424276300870772,
+      "learning_rate": 4.259849464711607e-06,
+      "loss": 0.6334,
+      "step": 5756
+    },
+    {
+      "epoch": 1.5283419620337182,
+      "grad_norm": 0.4309303162873798,
+      "learning_rate": 4.259601476077352e-06,
+      "loss": 0.6129,
+      "step": 5757
+    },
+    {
+      "epoch": 1.528607460507102,
+      "grad_norm": 0.40009392847866054,
+      "learning_rate": 4.259353453126559e-06,
+      "loss": 0.5734,
+      "step": 5758
+    },
+    {
+      "epoch": 1.528872958980486,
+      "grad_norm": 0.4005208881435323,
+      "learning_rate": 4.2591053958640666e-06,
+      "loss": 0.5864,
+      "step": 5759
+    },
+    {
+      "epoch": 1.5291384574538696,
+      "grad_norm": 0.4249982754784965,
+      "learning_rate": 4.258857304294712e-06,
+      "loss": 0.602,
+      "step": 5760
+    },
+    {
+      "epoch": 1.5294039559272534,
+      "grad_norm": 0.4141603237565349,
+      "learning_rate": 4.258609178423333e-06,
+      "loss": 0.5363,
+      "step": 5761
+    },
+    {
+      "epoch": 1.5296694544006373,
+      "grad_norm": 0.40413341864793056,
+      "learning_rate": 4.258361018254769e-06,
+      "loss": 0.5438,
+      "step": 5762
+    },
+    {
+      "epoch": 1.529934952874021,
+      "grad_norm": 0.4200763894297239,
+      "learning_rate": 4.258112823793861e-06,
+      "loss": 0.5974,
+      "step": 5763
+    },
+    {
+      "epoch": 1.5302004513474048,
+      "grad_norm": 0.4197995183917972,
+      "learning_rate": 4.2578645950454475e-06,
+      "loss": 0.6084,
+      "step": 5764
+    },
+    {
+      "epoch": 1.5304659498207887,
+      "grad_norm": 0.42058011282831137,
+      "learning_rate": 4.257616332014372e-06,
+      "loss": 0.5796,
+      "step": 5765
+    },
+    {
+      "epoch": 1.5307314482941723,
+      "grad_norm": 0.406127012476845,
+      "learning_rate": 4.257368034705473e-06,
+      "loss": 0.5966,
+      "step": 5766
+    },
+    {
+      "epoch": 1.530996946767556,
+      "grad_norm": 0.39841963386881635,
+      "learning_rate": 4.257119703123594e-06,
+      "loss": 0.5635,
+      "step": 5767
+    },
+    {
+      "epoch": 1.53126244524094,
+      "grad_norm": 0.4175207304720907,
+      "learning_rate": 4.25687133727358e-06,
+      "loss": 0.5921,
+      "step": 5768
+    },
+    {
+      "epoch": 1.5315279437143237,
+      "grad_norm": 0.4074352266903592,
+      "learning_rate": 4.256622937160273e-06,
+      "loss": 0.566,
+      "step": 5769
+    },
+    {
+      "epoch": 1.5317934421877073,
+      "grad_norm": 0.40506171798647617,
+      "learning_rate": 4.256374502788517e-06,
+      "loss": 0.5833,
+      "step": 5770
+    },
+    {
+      "epoch": 1.5320589406610912,
+      "grad_norm": 0.3832598521750574,
+      "learning_rate": 4.25612603416316e-06,
+      "loss": 0.5672,
+      "step": 5771
+    },
+    {
+      "epoch": 1.532324439134475,
+      "grad_norm": 0.4393559869051958,
+      "learning_rate": 4.255877531289043e-06,
+      "loss": 0.5876,
+      "step": 5772
+    },
+    {
+      "epoch": 1.5325899376078587,
+      "grad_norm": 0.4223043351269334,
+      "learning_rate": 4.255628994171016e-06,
+      "loss": 0.5959,
+      "step": 5773
+    },
+    {
+      "epoch": 1.5328554360812425,
+      "grad_norm": 0.42145960012246664,
+      "learning_rate": 4.255380422813925e-06,
+      "loss": 0.598,
+      "step": 5774
+    },
+    {
+      "epoch": 1.5331209345546264,
+      "grad_norm": 0.41480488591362813,
+      "learning_rate": 4.255131817222619e-06,
+      "loss": 0.58,
+      "step": 5775
+    },
+    {
+      "epoch": 1.53338643302801,
+      "grad_norm": 0.40980118405916865,
+      "learning_rate": 4.254883177401944e-06,
+      "loss": 0.6268,
+      "step": 5776
+    },
+    {
+      "epoch": 1.533651931501394,
+      "grad_norm": 0.42656038660278023,
+      "learning_rate": 4.25463450335675e-06,
+      "loss": 0.5875,
+      "step": 5777
+    },
+    {
+      "epoch": 1.5339174299747778,
+      "grad_norm": 0.39327612686813335,
+      "learning_rate": 4.254385795091888e-06,
+      "loss": 0.5823,
+      "step": 5778
+    },
+    {
+      "epoch": 1.5341829284481614,
+      "grad_norm": 0.4194634308998548,
+      "learning_rate": 4.254137052612207e-06,
+      "loss": 0.59,
+      "step": 5779
+    },
+    {
+      "epoch": 1.534448426921545,
+      "grad_norm": 0.4297277855058897,
+      "learning_rate": 4.2538882759225595e-06,
+      "loss": 0.5557,
+      "step": 5780
+    },
+    {
+      "epoch": 1.5347139253949291,
+      "grad_norm": 0.3957925631192835,
+      "learning_rate": 4.253639465027796e-06,
+      "loss": 0.5631,
+      "step": 5781
+    },
+    {
+      "epoch": 1.5349794238683128,
+      "grad_norm": 0.4107791274312506,
+      "learning_rate": 4.2533906199327685e-06,
+      "loss": 0.5838,
+      "step": 5782
+    },
+    {
+      "epoch": 1.5352449223416964,
+      "grad_norm": 0.3995563226634287,
+      "learning_rate": 4.2531417406423315e-06,
+      "loss": 0.5918,
+      "step": 5783
+    },
+    {
+      "epoch": 1.5355104208150803,
+      "grad_norm": 0.41867451064553973,
+      "learning_rate": 4.252892827161338e-06,
+      "loss": 0.569,
+      "step": 5784
+    },
+    {
+      "epoch": 1.5357759192884641,
+      "grad_norm": 0.42245589469063777,
+      "learning_rate": 4.252643879494642e-06,
+      "loss": 0.6175,
+      "step": 5785
+    },
+    {
+      "epoch": 1.5360414177618478,
+      "grad_norm": 0.4218231256229024,
+      "learning_rate": 4.252394897647101e-06,
+      "loss": 0.5801,
+      "step": 5786
+    },
+    {
+      "epoch": 1.5363069162352316,
+      "grad_norm": 0.42133253523658604,
+      "learning_rate": 4.252145881623567e-06,
+      "loss": 0.621,
+      "step": 5787
+    },
+    {
+      "epoch": 1.5365724147086155,
+      "grad_norm": 0.4257051952221697,
+      "learning_rate": 4.2518968314288985e-06,
+      "loss": 0.5368,
+      "step": 5788
+    },
+    {
+      "epoch": 1.5368379131819991,
+      "grad_norm": 0.40312463098060375,
+      "learning_rate": 4.251647747067953e-06,
+      "loss": 0.5941,
+      "step": 5789
+    },
+    {
+      "epoch": 1.537103411655383,
+      "grad_norm": 0.4033068783655608,
+      "learning_rate": 4.251398628545588e-06,
+      "loss": 0.6304,
+      "step": 5790
+    },
+    {
+      "epoch": 1.5373689101287669,
+      "grad_norm": 0.42136270751274707,
+      "learning_rate": 4.25114947586666e-06,
+      "loss": 0.5503,
+      "step": 5791
+    },
+    {
+      "epoch": 1.5376344086021505,
+      "grad_norm": 0.42328521480102993,
+      "learning_rate": 4.250900289036032e-06,
+      "loss": 0.6142,
+      "step": 5792
+    },
+    {
+      "epoch": 1.5378999070755344,
+      "grad_norm": 0.4348095111276332,
+      "learning_rate": 4.250651068058559e-06,
+      "loss": 0.6094,
+      "step": 5793
+    },
+    {
+      "epoch": 1.5381654055489182,
+      "grad_norm": 0.40941184330175634,
+      "learning_rate": 4.250401812939106e-06,
+      "loss": 0.5909,
+      "step": 5794
+    },
+    {
+      "epoch": 1.5384309040223019,
+      "grad_norm": 0.41099698682452285,
+      "learning_rate": 4.25015252368253e-06,
+      "loss": 0.5962,
+      "step": 5795
+    },
+    {
+      "epoch": 1.5386964024956855,
+      "grad_norm": 0.42019408909110584,
+      "learning_rate": 4.2499032002936955e-06,
+      "loss": 0.6372,
+      "step": 5796
+    },
+    {
+      "epoch": 1.5389619009690694,
+      "grad_norm": 0.431586458670598,
+      "learning_rate": 4.249653842777465e-06,
+      "loss": 0.5749,
+      "step": 5797
+    },
+    {
+      "epoch": 1.5392273994424532,
+      "grad_norm": 0.44193927000098066,
+      "learning_rate": 4.2494044511387e-06,
+      "loss": 0.5617,
+      "step": 5798
+    },
+    {
+      "epoch": 1.5394928979158369,
+      "grad_norm": 0.3975668084401015,
+      "learning_rate": 4.249155025382265e-06,
+      "loss": 0.6112,
+      "step": 5799
+    },
+    {
+      "epoch": 1.5397583963892207,
+      "grad_norm": 0.4143076607183988,
+      "learning_rate": 4.248905565513023e-06,
+      "loss": 0.5796,
+      "step": 5800
+    },
+    {
+      "epoch": 1.5400238948626046,
+      "grad_norm": 0.4864694341534911,
+      "learning_rate": 4.248656071535842e-06,
+      "loss": 0.5872,
+      "step": 5801
+    },
+    {
+      "epoch": 1.5402893933359882,
+      "grad_norm": 0.4351270774596731,
+      "learning_rate": 4.2484065434555855e-06,
+      "loss": 0.574,
+      "step": 5802
+    },
+    {
+      "epoch": 1.540554891809372,
+      "grad_norm": 0.4064014103644156,
+      "learning_rate": 4.24815698127712e-06,
+      "loss": 0.5622,
+      "step": 5803
+    },
+    {
+      "epoch": 1.540820390282756,
+      "grad_norm": 0.42124028804181896,
+      "learning_rate": 4.247907385005314e-06,
+      "loss": 0.6258,
+      "step": 5804
+    },
+    {
+      "epoch": 1.5410858887561396,
+      "grad_norm": 0.47537886182196665,
+      "learning_rate": 4.247657754645035e-06,
+      "loss": 0.6112,
+      "step": 5805
+    },
+    {
+      "epoch": 1.5413513872295235,
+      "grad_norm": 0.40091313115791594,
+      "learning_rate": 4.24740809020115e-06,
+      "loss": 0.5904,
+      "step": 5806
+    },
+    {
+      "epoch": 1.5416168857029073,
+      "grad_norm": 0.409271462210794,
+      "learning_rate": 4.2471583916785286e-06,
+      "loss": 0.5797,
+      "step": 5807
+    },
+    {
+      "epoch": 1.541882384176291,
+      "grad_norm": 0.4323446005740633,
+      "learning_rate": 4.246908659082041e-06,
+      "loss": 0.5897,
+      "step": 5808
+    },
+    {
+      "epoch": 1.5421478826496746,
+      "grad_norm": 0.41291909640193863,
+      "learning_rate": 4.246658892416557e-06,
+      "loss": 0.6087,
+      "step": 5809
+    },
+    {
+      "epoch": 1.5424133811230587,
+      "grad_norm": 0.4100678291779503,
+      "learning_rate": 4.246409091686949e-06,
+      "loss": 0.5939,
+      "step": 5810
+    },
+    {
+      "epoch": 1.5426788795964423,
+      "grad_norm": 0.41200681198697153,
+      "learning_rate": 4.246159256898087e-06,
+      "loss": 0.5434,
+      "step": 5811
+    },
+    {
+      "epoch": 1.542944378069826,
+      "grad_norm": 0.40083888715965993,
+      "learning_rate": 4.2459093880548454e-06,
+      "loss": 0.5984,
+      "step": 5812
+    },
+    {
+      "epoch": 1.5432098765432098,
+      "grad_norm": 0.41168998752501446,
+      "learning_rate": 4.245659485162094e-06,
+      "loss": 0.5641,
+      "step": 5813
+    },
+    {
+      "epoch": 1.5434753750165937,
+      "grad_norm": 0.4219655027983877,
+      "learning_rate": 4.24540954822471e-06,
+      "loss": 0.5843,
+      "step": 5814
+    },
+    {
+      "epoch": 1.5437408734899773,
+      "grad_norm": 0.40222172974883863,
+      "learning_rate": 4.245159577247566e-06,
+      "loss": 0.5937,
+      "step": 5815
+    },
+    {
+      "epoch": 1.5440063719633612,
+      "grad_norm": 0.41471412291120374,
+      "learning_rate": 4.244909572235537e-06,
+      "loss": 0.5673,
+      "step": 5816
+    },
+    {
+      "epoch": 1.544271870436745,
+      "grad_norm": 0.41427992842533257,
+      "learning_rate": 4.2446595331934994e-06,
+      "loss": 0.6154,
+      "step": 5817
+    },
+    {
+      "epoch": 1.5445373689101287,
+      "grad_norm": 0.41141856956047185,
+      "learning_rate": 4.2444094601263294e-06,
+      "loss": 0.5963,
+      "step": 5818
+    },
+    {
+      "epoch": 1.5448028673835126,
+      "grad_norm": 0.41672842956731926,
+      "learning_rate": 4.244159353038903e-06,
+      "loss": 0.5678,
+      "step": 5819
+    },
+    {
+      "epoch": 1.5450683658568964,
+      "grad_norm": 0.3978426799042112,
+      "learning_rate": 4.243909211936099e-06,
+      "loss": 0.6011,
+      "step": 5820
+    },
+    {
+      "epoch": 1.54533386433028,
+      "grad_norm": 0.42027792805681297,
+      "learning_rate": 4.243659036822797e-06,
+      "loss": 0.608,
+      "step": 5821
+    },
+    {
+      "epoch": 1.5455993628036637,
+      "grad_norm": 0.4126590542300162,
+      "learning_rate": 4.243408827703873e-06,
+      "loss": 0.5786,
+      "step": 5822
+    },
+    {
+      "epoch": 1.5458648612770478,
+      "grad_norm": 0.41610498361438586,
+      "learning_rate": 4.243158584584208e-06,
+      "loss": 0.6142,
+      "step": 5823
+    },
+    {
+      "epoch": 1.5461303597504314,
+      "grad_norm": 0.4129868549044019,
+      "learning_rate": 4.242908307468684e-06,
+      "loss": 0.6091,
+      "step": 5824
+    },
+    {
+      "epoch": 1.546395858223815,
+      "grad_norm": 0.4043845470339631,
+      "learning_rate": 4.24265799636218e-06,
+      "loss": 0.6179,
+      "step": 5825
+    },
+    {
+      "epoch": 1.546661356697199,
+      "grad_norm": 0.4157360122665706,
+      "learning_rate": 4.242407651269578e-06,
+      "loss": 0.57,
+      "step": 5826
+    },
+    {
+      "epoch": 1.5469268551705828,
+      "grad_norm": 0.4048007360187365,
+      "learning_rate": 4.242157272195762e-06,
+      "loss": 0.5828,
+      "step": 5827
+    },
+    {
+      "epoch": 1.5471923536439665,
+      "grad_norm": 0.40256862972778257,
+      "learning_rate": 4.241906859145611e-06,
+      "loss": 0.5625,
+      "step": 5828
+    },
+    {
+      "epoch": 1.5474578521173503,
+      "grad_norm": 0.4016118657995483,
+      "learning_rate": 4.2416564121240135e-06,
+      "loss": 0.5954,
+      "step": 5829
+    },
+    {
+      "epoch": 1.5477233505907342,
+      "grad_norm": 0.4037782828613498,
+      "learning_rate": 4.2414059311358516e-06,
+      "loss": 0.5909,
+      "step": 5830
+    },
+    {
+      "epoch": 1.5479888490641178,
+      "grad_norm": 0.41781049004740584,
+      "learning_rate": 4.2411554161860095e-06,
+      "loss": 0.5459,
+      "step": 5831
+    },
+    {
+      "epoch": 1.5482543475375017,
+      "grad_norm": 0.4019925451988796,
+      "learning_rate": 4.240904867279374e-06,
+      "loss": 0.6088,
+      "step": 5832
+    },
+    {
+      "epoch": 1.5485198460108855,
+      "grad_norm": 0.4055333673670718,
+      "learning_rate": 4.240654284420831e-06,
+      "loss": 0.6078,
+      "step": 5833
+    },
+    {
+      "epoch": 1.5487853444842692,
+      "grad_norm": 0.4186215945187696,
+      "learning_rate": 4.240403667615268e-06,
+      "loss": 0.6023,
+      "step": 5834
+    },
+    {
+      "epoch": 1.5490508429576528,
+      "grad_norm": 0.405876514870244,
+      "learning_rate": 4.240153016867572e-06,
+      "loss": 0.5933,
+      "step": 5835
+    },
+    {
+      "epoch": 1.549316341431037,
+      "grad_norm": 0.41630704898642606,
+      "learning_rate": 4.2399023321826316e-06,
+      "loss": 0.6102,
+      "step": 5836
+    },
+    {
+      "epoch": 1.5495818399044206,
+      "grad_norm": 0.3978300852038402,
+      "learning_rate": 4.239651613565335e-06,
+      "loss": 0.593,
+      "step": 5837
+    },
+    {
+      "epoch": 1.5498473383778042,
+      "grad_norm": 0.4072855438589772,
+      "learning_rate": 4.239400861020574e-06,
+      "loss": 0.6128,
+      "step": 5838
+    },
+    {
+      "epoch": 1.550112836851188,
+      "grad_norm": 0.4139393078405513,
+      "learning_rate": 4.239150074553236e-06,
+      "loss": 0.6129,
+      "step": 5839
+    },
+    {
+      "epoch": 1.550378335324572,
+      "grad_norm": 0.42332588918592934,
+      "learning_rate": 4.238899254168214e-06,
+      "loss": 0.5804,
+      "step": 5840
+    },
+    {
+      "epoch": 1.5506438337979556,
+      "grad_norm": 0.40530174807263925,
+      "learning_rate": 4.238648399870399e-06,
+      "loss": 0.5945,
+      "step": 5841
+    },
+    {
+      "epoch": 1.5509093322713394,
+      "grad_norm": 0.41599619653155445,
+      "learning_rate": 4.238397511664683e-06,
+      "loss": 0.5572,
+      "step": 5842
+    },
+    {
+      "epoch": 1.5511748307447233,
+      "grad_norm": 0.4035182662794045,
+      "learning_rate": 4.238146589555959e-06,
+      "loss": 0.5615,
+      "step": 5843
+    },
+    {
+      "epoch": 1.551440329218107,
+      "grad_norm": 0.42470619214545546,
+      "learning_rate": 4.2378956335491216e-06,
+      "loss": 0.6011,
+      "step": 5844
+    },
+    {
+      "epoch": 1.5517058276914908,
+      "grad_norm": 0.4190231135994607,
+      "learning_rate": 4.237644643649064e-06,
+      "loss": 0.5819,
+      "step": 5845
+    },
+    {
+      "epoch": 1.5519713261648747,
+      "grad_norm": 0.39722209428262373,
+      "learning_rate": 4.237393619860681e-06,
+      "loss": 0.603,
+      "step": 5846
+    },
+    {
+      "epoch": 1.5522368246382583,
+      "grad_norm": 0.43323014455335324,
+      "learning_rate": 4.237142562188869e-06,
+      "loss": 0.5904,
+      "step": 5847
+    },
+    {
+      "epoch": 1.5525023231116422,
+      "grad_norm": 0.415443619871322,
+      "learning_rate": 4.236891470638524e-06,
+      "loss": 0.6152,
+      "step": 5848
+    },
+    {
+      "epoch": 1.552767821585026,
+      "grad_norm": 0.4098809981824822,
+      "learning_rate": 4.236640345214541e-06,
+      "loss": 0.5871,
+      "step": 5849
+    },
+    {
+      "epoch": 1.5530333200584097,
+      "grad_norm": 0.405467005488627,
+      "learning_rate": 4.236389185921821e-06,
+      "loss": 0.5832,
+      "step": 5850
+    },
+    {
+      "epoch": 1.5532988185317933,
+      "grad_norm": 0.4034172674951067,
+      "learning_rate": 4.236137992765259e-06,
+      "loss": 0.6032,
+      "step": 5851
+    },
+    {
+      "epoch": 1.5535643170051774,
+      "grad_norm": 0.4320247273570949,
+      "learning_rate": 4.235886765749757e-06,
+      "loss": 0.5925,
+      "step": 5852
+    },
+    {
+      "epoch": 1.553829815478561,
+      "grad_norm": 0.41077635104360777,
+      "learning_rate": 4.235635504880212e-06,
+      "loss": 0.5762,
+      "step": 5853
+    },
+    {
+      "epoch": 1.5540953139519447,
+      "grad_norm": 0.3926395662340222,
+      "learning_rate": 4.235384210161525e-06,
+      "loss": 0.5661,
+      "step": 5854
+    },
+    {
+      "epoch": 1.5543608124253285,
+      "grad_norm": 0.4046592172890656,
+      "learning_rate": 4.235132881598597e-06,
+      "loss": 0.5823,
+      "step": 5855
+    },
+    {
+      "epoch": 1.5546263108987124,
+      "grad_norm": 0.39844549991809963,
+      "learning_rate": 4.23488151919633e-06,
+      "loss": 0.566,
+      "step": 5856
+    },
+    {
+      "epoch": 1.554891809372096,
+      "grad_norm": 0.43014457763183267,
+      "learning_rate": 4.2346301229596255e-06,
+      "loss": 0.5721,
+      "step": 5857
+    },
+    {
+      "epoch": 1.55515730784548,
+      "grad_norm": 0.39965980908173704,
+      "learning_rate": 4.234378692893386e-06,
+      "loss": 0.5692,
+      "step": 5858
+    },
+    {
+      "epoch": 1.5554228063188638,
+      "grad_norm": 0.4075590108772619,
+      "learning_rate": 4.234127229002516e-06,
+      "loss": 0.6058,
+      "step": 5859
+    },
+    {
+      "epoch": 1.5556883047922474,
+      "grad_norm": 0.4023674993181511,
+      "learning_rate": 4.2338757312919184e-06,
+      "loss": 0.6049,
+      "step": 5860
+    },
+    {
+      "epoch": 1.5559538032656313,
+      "grad_norm": 0.4014781129877831,
+      "learning_rate": 4.233624199766499e-06,
+      "loss": 0.5843,
+      "step": 5861
+    },
+    {
+      "epoch": 1.5562193017390151,
+      "grad_norm": 0.4235314768565759,
+      "learning_rate": 4.233372634431164e-06,
+      "loss": 0.5637,
+      "step": 5862
+    },
+    {
+      "epoch": 1.5564848002123988,
+      "grad_norm": 0.4087101517562995,
+      "learning_rate": 4.233121035290818e-06,
+      "loss": 0.568,
+      "step": 5863
+    },
+    {
+      "epoch": 1.5567502986857824,
+      "grad_norm": 0.40102443397848125,
+      "learning_rate": 4.232869402350368e-06,
+      "loss": 0.5987,
+      "step": 5864
+    },
+    {
+      "epoch": 1.5570157971591665,
+      "grad_norm": 0.4285702642848026,
+      "learning_rate": 4.232617735614723e-06,
+      "loss": 0.5869,
+      "step": 5865
+    },
+    {
+      "epoch": 1.5572812956325501,
+      "grad_norm": 0.41656109106347255,
+      "learning_rate": 4.232366035088789e-06,
+      "loss": 0.5616,
+      "step": 5866
+    },
+    {
+      "epoch": 1.5575467941059338,
+      "grad_norm": 0.41884577853795607,
+      "learning_rate": 4.232114300777476e-06,
+      "loss": 0.6016,
+      "step": 5867
+    },
+    {
+      "epoch": 1.5578122925793176,
+      "grad_norm": 0.4013414373630112,
+      "learning_rate": 4.231862532685693e-06,
+      "loss": 0.6185,
+      "step": 5868
+    },
+    {
+      "epoch": 1.5580777910527015,
+      "grad_norm": 0.4124732319704868,
+      "learning_rate": 4.23161073081835e-06,
+      "loss": 0.6101,
+      "step": 5869
+    },
+    {
+      "epoch": 1.5583432895260851,
+      "grad_norm": 0.41176002739281914,
+      "learning_rate": 4.231358895180358e-06,
+      "loss": 0.5694,
+      "step": 5870
+    },
+    {
+      "epoch": 1.558608787999469,
+      "grad_norm": 0.41132520728093724,
+      "learning_rate": 4.23110702577663e-06,
+      "loss": 0.5927,
+      "step": 5871
+    },
+    {
+      "epoch": 1.5588742864728529,
+      "grad_norm": 0.41557830903833526,
+      "learning_rate": 4.2308551226120745e-06,
+      "loss": 0.5622,
+      "step": 5872
+    },
+    {
+      "epoch": 1.5591397849462365,
+      "grad_norm": 0.40339539707210914,
+      "learning_rate": 4.230603185691607e-06,
+      "loss": 0.558,
+      "step": 5873
+    },
+    {
+      "epoch": 1.5594052834196204,
+      "grad_norm": 0.40176190724417543,
+      "learning_rate": 4.230351215020141e-06,
+      "loss": 0.6058,
+      "step": 5874
+    },
+    {
+      "epoch": 1.5596707818930042,
+      "grad_norm": 0.41817120041728995,
+      "learning_rate": 4.230099210602588e-06,
+      "loss": 0.6008,
+      "step": 5875
+    },
+    {
+      "epoch": 1.5599362803663879,
+      "grad_norm": 0.4010513109388064,
+      "learning_rate": 4.229847172443866e-06,
+      "loss": 0.6136,
+      "step": 5876
+    },
+    {
+      "epoch": 1.5602017788397715,
+      "grad_norm": 0.39602537916768554,
+      "learning_rate": 4.229595100548888e-06,
+      "loss": 0.5022,
+      "step": 5877
+    },
+    {
+      "epoch": 1.5604672773131556,
+      "grad_norm": 0.4044732716956632,
+      "learning_rate": 4.229342994922571e-06,
+      "loss": 0.5858,
+      "step": 5878
+    },
+    {
+      "epoch": 1.5607327757865392,
+      "grad_norm": 0.4217154112969024,
+      "learning_rate": 4.2290908555698315e-06,
+      "loss": 0.5923,
+      "step": 5879
+    },
+    {
+      "epoch": 1.5609982742599229,
+      "grad_norm": 0.4132991401352576,
+      "learning_rate": 4.228838682495587e-06,
+      "loss": 0.6226,
+      "step": 5880
+    },
+    {
+      "epoch": 1.5612637727333067,
+      "grad_norm": 0.4091460801184228,
+      "learning_rate": 4.228586475704753e-06,
+      "loss": 0.5588,
+      "step": 5881
+    },
+    {
+      "epoch": 1.5615292712066906,
+      "grad_norm": 0.4042625894178759,
+      "learning_rate": 4.228334235202253e-06,
+      "loss": 0.5985,
+      "step": 5882
+    },
+    {
+      "epoch": 1.5617947696800742,
+      "grad_norm": 0.41057629366254517,
+      "learning_rate": 4.228081960993003e-06,
+      "loss": 0.5645,
+      "step": 5883
+    },
+    {
+      "epoch": 1.562060268153458,
+      "grad_norm": 0.4027692636260605,
+      "learning_rate": 4.227829653081924e-06,
+      "loss": 0.604,
+      "step": 5884
+    },
+    {
+      "epoch": 1.562325766626842,
+      "grad_norm": 0.41716933357947744,
+      "learning_rate": 4.227577311473936e-06,
+      "loss": 0.5976,
+      "step": 5885
+    },
+    {
+      "epoch": 1.5625912651002256,
+      "grad_norm": 0.41674950240607905,
+      "learning_rate": 4.22732493617396e-06,
+      "loss": 0.6123,
+      "step": 5886
+    },
+    {
+      "epoch": 1.5628567635736095,
+      "grad_norm": 0.4081189975922936,
+      "learning_rate": 4.2270725271869195e-06,
+      "loss": 0.5792,
+      "step": 5887
+    },
+    {
+      "epoch": 1.5631222620469933,
+      "grad_norm": 0.41343739306195226,
+      "learning_rate": 4.226820084517736e-06,
+      "loss": 0.568,
+      "step": 5888
+    },
+    {
+      "epoch": 1.563387760520377,
+      "grad_norm": 0.4002282207509223,
+      "learning_rate": 4.226567608171332e-06,
+      "loss": 0.5809,
+      "step": 5889
+    },
+    {
+      "epoch": 1.5636532589937606,
+      "grad_norm": 0.41760442619703514,
+      "learning_rate": 4.226315098152633e-06,
+      "loss": 0.5817,
+      "step": 5890
+    },
+    {
+      "epoch": 1.5639187574671447,
+      "grad_norm": 0.4160844121834743,
+      "learning_rate": 4.226062554466562e-06,
+      "loss": 0.6201,
+      "step": 5891
+    },
+    {
+      "epoch": 1.5641842559405283,
+      "grad_norm": 0.398262143753868,
+      "learning_rate": 4.225809977118046e-06,
+      "loss": 0.5656,
+      "step": 5892
+    },
+    {
+      "epoch": 1.564449754413912,
+      "grad_norm": 0.41444698291950066,
+      "learning_rate": 4.2255573661120095e-06,
+      "loss": 0.5772,
+      "step": 5893
+    },
+    {
+      "epoch": 1.5647152528872958,
+      "grad_norm": 0.4063777185920357,
+      "learning_rate": 4.22530472145338e-06,
+      "loss": 0.6136,
+      "step": 5894
+    },
+    {
+      "epoch": 1.5649807513606797,
+      "grad_norm": 0.41199836514667787,
+      "learning_rate": 4.225052043147083e-06,
+      "loss": 0.564,
+      "step": 5895
+    },
+    {
+      "epoch": 1.5652462498340634,
+      "grad_norm": 0.4068644710297008,
+      "learning_rate": 4.224799331198049e-06,
+      "loss": 0.6002,
+      "step": 5896
+    },
+    {
+      "epoch": 1.5655117483074472,
+      "grad_norm": 0.4170324572753706,
+      "learning_rate": 4.2245465856112044e-06,
+      "loss": 0.5929,
+      "step": 5897
+    },
+    {
+      "epoch": 1.565777246780831,
+      "grad_norm": 0.4076004639658558,
+      "learning_rate": 4.2242938063914784e-06,
+      "loss": 0.5926,
+      "step": 5898
+    },
+    {
+      "epoch": 1.5660427452542147,
+      "grad_norm": 0.4136370170116393,
+      "learning_rate": 4.2240409935438024e-06,
+      "loss": 0.5934,
+      "step": 5899
+    },
+    {
+      "epoch": 1.5663082437275986,
+      "grad_norm": 0.4006854615382548,
+      "learning_rate": 4.2237881470731044e-06,
+      "loss": 0.5873,
+      "step": 5900
+    },
+    {
+      "epoch": 1.5665737422009824,
+      "grad_norm": 0.40757664191477816,
+      "learning_rate": 4.2235352669843185e-06,
+      "loss": 0.59,
+      "step": 5901
+    },
+    {
+      "epoch": 1.566839240674366,
+      "grad_norm": 0.4030121772108053,
+      "learning_rate": 4.223282353282374e-06,
+      "loss": 0.6202,
+      "step": 5902
+    },
+    {
+      "epoch": 1.56710473914775,
+      "grad_norm": 0.43258334978943097,
+      "learning_rate": 4.223029405972204e-06,
+      "loss": 0.6048,
+      "step": 5903
+    },
+    {
+      "epoch": 1.5673702376211338,
+      "grad_norm": 0.4218620751956037,
+      "learning_rate": 4.222776425058743e-06,
+      "loss": 0.5808,
+      "step": 5904
+    },
+    {
+      "epoch": 1.5676357360945175,
+      "grad_norm": 0.418534280963697,
+      "learning_rate": 4.222523410546924e-06,
+      "loss": 0.5855,
+      "step": 5905
+    },
+    {
+      "epoch": 1.567901234567901,
+      "grad_norm": 0.43832169659519216,
+      "learning_rate": 4.2222703624416795e-06,
+      "loss": 0.5931,
+      "step": 5906
+    },
+    {
+      "epoch": 1.5681667330412852,
+      "grad_norm": 0.4219943324386712,
+      "learning_rate": 4.222017280747947e-06,
+      "loss": 0.6069,
+      "step": 5907
+    },
+    {
+      "epoch": 1.5684322315146688,
+      "grad_norm": 0.4150603646712336,
+      "learning_rate": 4.221764165470661e-06,
+      "loss": 0.6103,
+      "step": 5908
+    },
+    {
+      "epoch": 1.5686977299880525,
+      "grad_norm": 0.4295903044425061,
+      "learning_rate": 4.221511016614757e-06,
+      "loss": 0.566,
+      "step": 5909
+    },
+    {
+      "epoch": 1.5689632284614363,
+      "grad_norm": 0.4356518803526646,
+      "learning_rate": 4.221257834185175e-06,
+      "loss": 0.5799,
+      "step": 5910
+    },
+    {
+      "epoch": 1.5692287269348202,
+      "grad_norm": 0.4157975325252088,
+      "learning_rate": 4.22100461818685e-06,
+      "loss": 0.6036,
+      "step": 5911
+    },
+    {
+      "epoch": 1.5694942254082038,
+      "grad_norm": 0.4052430668887788,
+      "learning_rate": 4.220751368624722e-06,
+      "loss": 0.5994,
+      "step": 5912
+    },
+    {
+      "epoch": 1.5697597238815877,
+      "grad_norm": 0.428700519188456,
+      "learning_rate": 4.220498085503728e-06,
+      "loss": 0.5947,
+      "step": 5913
+    },
+    {
+      "epoch": 1.5700252223549716,
+      "grad_norm": 0.41918695540082573,
+      "learning_rate": 4.2202447688288095e-06,
+      "loss": 0.5877,
+      "step": 5914
+    },
+    {
+      "epoch": 1.5702907208283552,
+      "grad_norm": 0.4153244902174952,
+      "learning_rate": 4.219991418604906e-06,
+      "loss": 0.6168,
+      "step": 5915
+    },
+    {
+      "epoch": 1.570556219301739,
+      "grad_norm": 0.42847546456605407,
+      "learning_rate": 4.2197380348369585e-06,
+      "loss": 0.5769,
+      "step": 5916
+    },
+    {
+      "epoch": 1.570821717775123,
+      "grad_norm": 0.4196710552880703,
+      "learning_rate": 4.219484617529909e-06,
+      "loss": 0.5765,
+      "step": 5917
+    },
+    {
+      "epoch": 1.5710872162485066,
+      "grad_norm": 0.41646039173416527,
+      "learning_rate": 4.2192311666887e-06,
+      "loss": 0.6288,
+      "step": 5918
+    },
+    {
+      "epoch": 1.5713527147218902,
+      "grad_norm": 0.42930802234671644,
+      "learning_rate": 4.2189776823182725e-06,
+      "loss": 0.5775,
+      "step": 5919
+    },
+    {
+      "epoch": 1.5716182131952743,
+      "grad_norm": 0.45519510858263085,
+      "learning_rate": 4.218724164423572e-06,
+      "loss": 0.5958,
+      "step": 5920
+    },
+    {
+      "epoch": 1.571883711668658,
+      "grad_norm": 0.43972190133326644,
+      "learning_rate": 4.218470613009542e-06,
+      "loss": 0.5628,
+      "step": 5921
+    },
+    {
+      "epoch": 1.5721492101420416,
+      "grad_norm": 0.389497709810723,
+      "learning_rate": 4.218217028081128e-06,
+      "loss": 0.584,
+      "step": 5922
+    },
+    {
+      "epoch": 1.5724147086154254,
+      "grad_norm": 0.452147906159794,
+      "learning_rate": 4.217963409643274e-06,
+      "loss": 0.6105,
+      "step": 5923
+    },
+    {
+      "epoch": 1.5726802070888093,
+      "grad_norm": 0.43340560000034933,
+      "learning_rate": 4.217709757700928e-06,
+      "loss": 0.5949,
+      "step": 5924
+    },
+    {
+      "epoch": 1.572945705562193,
+      "grad_norm": 0.43131757974404655,
+      "learning_rate": 4.217456072259036e-06,
+      "loss": 0.6204,
+      "step": 5925
+    },
+    {
+      "epoch": 1.5732112040355768,
+      "grad_norm": 0.4044872744621555,
+      "learning_rate": 4.2172023533225455e-06,
+      "loss": 0.6001,
+      "step": 5926
+    },
+    {
+      "epoch": 1.5734767025089607,
+      "grad_norm": 0.4109284103509459,
+      "learning_rate": 4.216948600896404e-06,
+      "loss": 0.5759,
+      "step": 5927
+    },
+    {
+      "epoch": 1.5737422009823443,
+      "grad_norm": 0.40287804903654656,
+      "learning_rate": 4.2166948149855625e-06,
+      "loss": 0.5871,
+      "step": 5928
+    },
+    {
+      "epoch": 1.5740076994557282,
+      "grad_norm": 0.4044696521998476,
+      "learning_rate": 4.216440995594968e-06,
+      "loss": 0.5743,
+      "step": 5929
+    },
+    {
+      "epoch": 1.574273197929112,
+      "grad_norm": 0.39533156802861547,
+      "learning_rate": 4.216187142729571e-06,
+      "loss": 0.5603,
+      "step": 5930
+    },
+    {
+      "epoch": 1.5745386964024957,
+      "grad_norm": 0.40899463646742273,
+      "learning_rate": 4.2159332563943236e-06,
+      "loss": 0.5852,
+      "step": 5931
+    },
+    {
+      "epoch": 1.5748041948758793,
+      "grad_norm": 0.41925956785571183,
+      "learning_rate": 4.215679336594176e-06,
+      "loss": 0.6013,
+      "step": 5932
+    },
+    {
+      "epoch": 1.5750696933492634,
+      "grad_norm": 0.4189328959409041,
+      "learning_rate": 4.2154253833340806e-06,
+      "loss": 0.5815,
+      "step": 5933
+    },
+    {
+      "epoch": 1.575335191822647,
+      "grad_norm": 0.40386936387151323,
+      "learning_rate": 4.21517139661899e-06,
+      "loss": 0.5757,
+      "step": 5934
+    },
+    {
+      "epoch": 1.5756006902960307,
+      "grad_norm": 0.4170304415279702,
+      "learning_rate": 4.2149173764538585e-06,
+      "loss": 0.5919,
+      "step": 5935
+    },
+    {
+      "epoch": 1.5758661887694145,
+      "grad_norm": 0.40576731645784875,
+      "learning_rate": 4.214663322843639e-06,
+      "loss": 0.6074,
+      "step": 5936
+    },
+    {
+      "epoch": 1.5761316872427984,
+      "grad_norm": 0.40562659783285193,
+      "learning_rate": 4.2144092357932866e-06,
+      "loss": 0.5477,
+      "step": 5937
+    },
+    {
+      "epoch": 1.576397185716182,
+      "grad_norm": 0.40788260610754606,
+      "learning_rate": 4.214155115307755e-06,
+      "loss": 0.6253,
+      "step": 5938
+    },
+    {
+      "epoch": 1.576662684189566,
+      "grad_norm": 0.40154662351773246,
+      "learning_rate": 4.2139009613920035e-06,
+      "loss": 0.5475,
+      "step": 5939
+    },
+    {
+      "epoch": 1.5769281826629498,
+      "grad_norm": 0.4115552774617222,
+      "learning_rate": 4.213646774050986e-06,
+      "loss": 0.5875,
+      "step": 5940
+    },
+    {
+      "epoch": 1.5771936811363334,
+      "grad_norm": 0.4144542959024575,
+      "learning_rate": 4.2133925532896604e-06,
+      "loss": 0.5713,
+      "step": 5941
+    },
+    {
+      "epoch": 1.5774591796097173,
+      "grad_norm": 0.41059310937891563,
+      "learning_rate": 4.213138299112986e-06,
+      "loss": 0.6193,
+      "step": 5942
+    },
+    {
+      "epoch": 1.5777246780831011,
+      "grad_norm": 0.4147022724289457,
+      "learning_rate": 4.21288401152592e-06,
+      "loss": 0.6137,
+      "step": 5943
+    },
+    {
+      "epoch": 1.5779901765564848,
+      "grad_norm": 0.40987596971745355,
+      "learning_rate": 4.212629690533421e-06,
+      "loss": 0.6239,
+      "step": 5944
+    },
+    {
+      "epoch": 1.5782556750298686,
+      "grad_norm": 0.41194164881596446,
+      "learning_rate": 4.212375336140451e-06,
+      "loss": 0.5555,
+      "step": 5945
+    },
+    {
+      "epoch": 1.5785211735032525,
+      "grad_norm": 0.40441671216400144,
+      "learning_rate": 4.2121209483519685e-06,
+      "loss": 0.5932,
+      "step": 5946
+    },
+    {
+      "epoch": 1.5787866719766361,
+      "grad_norm": 0.4129208472987318,
+      "learning_rate": 4.211866527172935e-06,
+      "loss": 0.591,
+      "step": 5947
+    },
+    {
+      "epoch": 1.5790521704500198,
+      "grad_norm": 0.4188609381846575,
+      "learning_rate": 4.211612072608314e-06,
+      "loss": 0.5388,
+      "step": 5948
+    },
+    {
+      "epoch": 1.5793176689234036,
+      "grad_norm": 0.40912721622715353,
+      "learning_rate": 4.211357584663066e-06,
+      "loss": 0.6259,
+      "step": 5949
+    },
+    {
+      "epoch": 1.5795831673967875,
+      "grad_norm": 0.4060790567069496,
+      "learning_rate": 4.2111030633421556e-06,
+      "loss": 0.5751,
+      "step": 5950
+    },
+    {
+      "epoch": 1.5798486658701711,
+      "grad_norm": 0.41568493885817687,
+      "learning_rate": 4.210848508650546e-06,
+      "loss": 0.6012,
+      "step": 5951
+    },
+    {
+      "epoch": 1.580114164343555,
+      "grad_norm": 0.4237085114846836,
+      "learning_rate": 4.210593920593201e-06,
+      "loss": 0.6048,
+      "step": 5952
+    },
+    {
+      "epoch": 1.5803796628169389,
+      "grad_norm": 0.4087352108341222,
+      "learning_rate": 4.2103392991750865e-06,
+      "loss": 0.5951,
+      "step": 5953
+    },
+    {
+      "epoch": 1.5806451612903225,
+      "grad_norm": 0.44598374346046615,
+      "learning_rate": 4.210084644401169e-06,
+      "loss": 0.5876,
+      "step": 5954
+    },
+    {
+      "epoch": 1.5809106597637064,
+      "grad_norm": 0.4029743827997855,
+      "learning_rate": 4.209829956276413e-06,
+      "loss": 0.5981,
+      "step": 5955
+    },
+    {
+      "epoch": 1.5811761582370902,
+      "grad_norm": 0.4121599418975682,
+      "learning_rate": 4.209575234805786e-06,
+      "loss": 0.5556,
+      "step": 5956
+    },
+    {
+      "epoch": 1.5814416567104739,
+      "grad_norm": 0.425859398239993,
+      "learning_rate": 4.2093204799942565e-06,
+      "loss": 0.5854,
+      "step": 5957
+    },
+    {
+      "epoch": 1.5817071551838577,
+      "grad_norm": 0.43541061108265505,
+      "learning_rate": 4.2090656918467935e-06,
+      "loss": 0.5773,
+      "step": 5958
+    },
+    {
+      "epoch": 1.5819726536572416,
+      "grad_norm": 0.40816877271914176,
+      "learning_rate": 4.208810870368364e-06,
+      "loss": 0.632,
+      "step": 5959
+    },
+    {
+      "epoch": 1.5822381521306252,
+      "grad_norm": 0.4155544602766432,
+      "learning_rate": 4.2085560155639385e-06,
+      "loss": 0.5858,
+      "step": 5960
+    },
+    {
+      "epoch": 1.5825036506040089,
+      "grad_norm": 0.3944216205024293,
+      "learning_rate": 4.208301127438487e-06,
+      "loss": 0.5781,
+      "step": 5961
+    },
+    {
+      "epoch": 1.582769149077393,
+      "grad_norm": 0.40462766703041153,
+      "learning_rate": 4.208046205996983e-06,
+      "loss": 0.5528,
+      "step": 5962
+    },
+    {
+      "epoch": 1.5830346475507766,
+      "grad_norm": 0.4379330632387086,
+      "learning_rate": 4.207791251244394e-06,
+      "loss": 0.5587,
+      "step": 5963
+    },
+    {
+      "epoch": 1.5833001460241602,
+      "grad_norm": 0.4188592762053884,
+      "learning_rate": 4.207536263185695e-06,
+      "loss": 0.5745,
+      "step": 5964
+    },
+    {
+      "epoch": 1.5835656444975441,
+      "grad_norm": 0.3981961743754073,
+      "learning_rate": 4.207281241825858e-06,
+      "loss": 0.5647,
+      "step": 5965
+    },
+    {
+      "epoch": 1.583831142970928,
+      "grad_norm": 0.4189791875112949,
+      "learning_rate": 4.207026187169856e-06,
+      "loss": 0.6078,
+      "step": 5966
+    },
+    {
+      "epoch": 1.5840966414443116,
+      "grad_norm": 0.4063913160380496,
+      "learning_rate": 4.206771099222665e-06,
+      "loss": 0.5822,
+      "step": 5967
+    },
+    {
+      "epoch": 1.5843621399176955,
+      "grad_norm": 0.4149144916651215,
+      "learning_rate": 4.206515977989258e-06,
+      "loss": 0.603,
+      "step": 5968
+    },
+    {
+      "epoch": 1.5846276383910793,
+      "grad_norm": 0.4016819064469472,
+      "learning_rate": 4.206260823474612e-06,
+      "loss": 0.5855,
+      "step": 5969
+    },
+    {
+      "epoch": 1.584893136864463,
+      "grad_norm": 0.4057378993228367,
+      "learning_rate": 4.206005635683701e-06,
+      "loss": 0.5834,
+      "step": 5970
+    },
+    {
+      "epoch": 1.5851586353378468,
+      "grad_norm": 0.38687882269812934,
+      "learning_rate": 4.205750414621503e-06,
+      "loss": 0.5702,
+      "step": 5971
+    },
+    {
+      "epoch": 1.5854241338112307,
+      "grad_norm": 0.4060709000613773,
+      "learning_rate": 4.205495160292996e-06,
+      "loss": 0.5782,
+      "step": 5972
+    },
+    {
+      "epoch": 1.5856896322846143,
+      "grad_norm": 0.4145604152503478,
+      "learning_rate": 4.205239872703158e-06,
+      "loss": 0.5931,
+      "step": 5973
+    },
+    {
+      "epoch": 1.585955130757998,
+      "grad_norm": 0.39135702850820037,
+      "learning_rate": 4.204984551856968e-06,
+      "loss": 0.563,
+      "step": 5974
+    },
+    {
+      "epoch": 1.586220629231382,
+      "grad_norm": 0.4001288730345016,
+      "learning_rate": 4.204729197759403e-06,
+      "loss": 0.5772,
+      "step": 5975
+    },
+    {
+      "epoch": 1.5864861277047657,
+      "grad_norm": 0.41946455918791337,
+      "learning_rate": 4.204473810415446e-06,
+      "loss": 0.6101,
+      "step": 5976
+    },
+    {
+      "epoch": 1.5867516261781494,
+      "grad_norm": 0.3999852785381553,
+      "learning_rate": 4.204218389830076e-06,
+      "loss": 0.5842,
+      "step": 5977
+    },
+    {
+      "epoch": 1.5870171246515332,
+      "grad_norm": 0.4215461550554838,
+      "learning_rate": 4.203962936008275e-06,
+      "loss": 0.6235,
+      "step": 5978
+    },
+    {
+      "epoch": 1.587282623124917,
+      "grad_norm": 0.42373764730836094,
+      "learning_rate": 4.203707448955024e-06,
+      "loss": 0.5966,
+      "step": 5979
+    },
+    {
+      "epoch": 1.5875481215983007,
+      "grad_norm": 0.429760373243519,
+      "learning_rate": 4.203451928675306e-06,
+      "loss": 0.6217,
+      "step": 5980
+    },
+    {
+      "epoch": 1.5878136200716846,
+      "grad_norm": 0.4129260216197234,
+      "learning_rate": 4.203196375174106e-06,
+      "loss": 0.5545,
+      "step": 5981
+    },
+    {
+      "epoch": 1.5880791185450684,
+      "grad_norm": 0.39771812223466274,
+      "learning_rate": 4.202940788456405e-06,
+      "loss": 0.6288,
+      "step": 5982
+    },
+    {
+      "epoch": 1.588344617018452,
+      "grad_norm": 0.3903532495377373,
+      "learning_rate": 4.202685168527191e-06,
+      "loss": 0.5778,
+      "step": 5983
+    },
+    {
+      "epoch": 1.588610115491836,
+      "grad_norm": 0.39689511354125345,
+      "learning_rate": 4.202429515391446e-06,
+      "loss": 0.5216,
+      "step": 5984
+    },
+    {
+      "epoch": 1.5888756139652198,
+      "grad_norm": 0.41033234328965706,
+      "learning_rate": 4.2021738290541585e-06,
+      "loss": 0.5454,
+      "step": 5985
+    },
+    {
+      "epoch": 1.5891411124386035,
+      "grad_norm": 0.4057344824731183,
+      "learning_rate": 4.201918109520312e-06,
+      "loss": 0.6122,
+      "step": 5986
+    },
+    {
+      "epoch": 1.589406610911987,
+      "grad_norm": 0.40231300412898074,
+      "learning_rate": 4.201662356794896e-06,
+      "loss": 0.5632,
+      "step": 5987
+    },
+    {
+      "epoch": 1.5896721093853712,
+      "grad_norm": 0.4265490622011692,
+      "learning_rate": 4.201406570882898e-06,
+      "loss": 0.6224,
+      "step": 5988
+    },
+    {
+      "epoch": 1.5899376078587548,
+      "grad_norm": 0.40712687510895884,
+      "learning_rate": 4.201150751789305e-06,
+      "loss": 0.6053,
+      "step": 5989
+    },
+    {
+      "epoch": 1.5902031063321385,
+      "grad_norm": 0.4091628103122109,
+      "learning_rate": 4.2008948995191085e-06,
+      "loss": 0.571,
+      "step": 5990
+    },
+    {
+      "epoch": 1.5904686048055223,
+      "grad_norm": 0.41336670515306734,
+      "learning_rate": 4.200639014077297e-06,
+      "loss": 0.5782,
+      "step": 5991
+    },
+    {
+      "epoch": 1.5907341032789062,
+      "grad_norm": 0.4046226993967342,
+      "learning_rate": 4.2003830954688605e-06,
+      "loss": 0.5894,
+      "step": 5992
+    },
+    {
+      "epoch": 1.5909996017522898,
+      "grad_norm": 0.4128853195306421,
+      "learning_rate": 4.20012714369879e-06,
+      "loss": 0.6186,
+      "step": 5993
+    },
+    {
+      "epoch": 1.5912651002256737,
+      "grad_norm": 0.4313063298193808,
+      "learning_rate": 4.199871158772077e-06,
+      "loss": 0.5918,
+      "step": 5994
+    },
+    {
+      "epoch": 1.5915305986990576,
+      "grad_norm": 0.3914005958264913,
+      "learning_rate": 4.199615140693716e-06,
+      "loss": 0.546,
+      "step": 5995
+    },
+    {
+      "epoch": 1.5917960971724412,
+      "grad_norm": 0.400408273087365,
+      "learning_rate": 4.199359089468697e-06,
+      "loss": 0.582,
+      "step": 5996
+    },
+    {
+      "epoch": 1.592061595645825,
+      "grad_norm": 0.41062738685374334,
+      "learning_rate": 4.199103005102016e-06,
+      "loss": 0.6125,
+      "step": 5997
+    },
+    {
+      "epoch": 1.592327094119209,
+      "grad_norm": 0.4246184671474157,
+      "learning_rate": 4.198846887598666e-06,
+      "loss": 0.5955,
+      "step": 5998
+    },
+    {
+      "epoch": 1.5925925925925926,
+      "grad_norm": 0.3967529096668717,
+      "learning_rate": 4.198590736963642e-06,
+      "loss": 0.5416,
+      "step": 5999
+    },
+    {
+      "epoch": 1.5928580910659764,
+      "grad_norm": 0.4133297856797672,
+      "learning_rate": 4.19833455320194e-06,
+      "loss": 0.639,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5931235895393603,
+      "grad_norm": 0.4070327314079298,
+      "learning_rate": 4.198078336318555e-06,
+      "loss": 0.5758,
+      "step": 6001
+    },
+    {
+      "epoch": 1.593389088012744,
+      "grad_norm": 0.4202513466676356,
+      "learning_rate": 4.197822086318486e-06,
+      "loss": 0.5555,
+      "step": 6002
+    },
+    {
+      "epoch": 1.5936545864861276,
+      "grad_norm": 0.39350257469331734,
+      "learning_rate": 4.197565803206729e-06,
+      "loss": 0.5931,
+      "step": 6003
+    },
+    {
+      "epoch": 1.5939200849595114,
+      "grad_norm": 0.40212818904911024,
+      "learning_rate": 4.197309486988283e-06,
+      "loss": 0.5873,
+      "step": 6004
+    },
+    {
+      "epoch": 1.5941855834328953,
+      "grad_norm": 0.4026076412576511,
+      "learning_rate": 4.197053137668145e-06,
+      "loss": 0.6058,
+      "step": 6005
+    },
+    {
+      "epoch": 1.594451081906279,
+      "grad_norm": 0.4421551757263256,
+      "learning_rate": 4.196796755251317e-06,
+      "loss": 0.5852,
+      "step": 6006
+    },
+    {
+      "epoch": 1.5947165803796628,
+      "grad_norm": 0.4147659196130696,
+      "learning_rate": 4.1965403397427965e-06,
+      "loss": 0.5753,
+      "step": 6007
+    },
+    {
+      "epoch": 1.5949820788530467,
+      "grad_norm": 0.40457771278237803,
+      "learning_rate": 4.196283891147587e-06,
+      "loss": 0.574,
+      "step": 6008
+    },
+    {
+      "epoch": 1.5952475773264303,
+      "grad_norm": 0.39508156118232685,
+      "learning_rate": 4.196027409470688e-06,
+      "loss": 0.6126,
+      "step": 6009
+    },
+    {
+      "epoch": 1.5955130757998142,
+      "grad_norm": 0.40919573593044006,
+      "learning_rate": 4.195770894717101e-06,
+      "loss": 0.6301,
+      "step": 6010
+    },
+    {
+      "epoch": 1.595778574273198,
+      "grad_norm": 0.4107629359874985,
+      "learning_rate": 4.19551434689183e-06,
+      "loss": 0.5823,
+      "step": 6011
+    },
+    {
+      "epoch": 1.5960440727465817,
+      "grad_norm": 0.41731025839725505,
+      "learning_rate": 4.195257765999879e-06,
+      "loss": 0.6038,
+      "step": 6012
+    },
+    {
+      "epoch": 1.5963095712199655,
+      "grad_norm": 0.38997182093140825,
+      "learning_rate": 4.1950011520462484e-06,
+      "loss": 0.5811,
+      "step": 6013
+    },
+    {
+      "epoch": 1.5965750696933494,
+      "grad_norm": 0.40432310650264053,
+      "learning_rate": 4.194744505035948e-06,
+      "loss": 0.554,
+      "step": 6014
+    },
+    {
+      "epoch": 1.596840568166733,
+      "grad_norm": 0.4201659505765884,
+      "learning_rate": 4.194487824973979e-06,
+      "loss": 0.5991,
+      "step": 6015
+    },
+    {
+      "epoch": 1.5971060666401167,
+      "grad_norm": 0.3993082821757404,
+      "learning_rate": 4.194231111865348e-06,
+      "loss": 0.5555,
+      "step": 6016
+    },
+    {
+      "epoch": 1.5973715651135008,
+      "grad_norm": 0.41904984827321556,
+      "learning_rate": 4.193974365715064e-06,
+      "loss": 0.621,
+      "step": 6017
+    },
+    {
+      "epoch": 1.5976370635868844,
+      "grad_norm": 0.4090905886977012,
+      "learning_rate": 4.1937175865281295e-06,
+      "loss": 0.5936,
+      "step": 6018
+    },
+    {
+      "epoch": 1.597902562060268,
+      "grad_norm": 0.4029763981599839,
+      "learning_rate": 4.193460774309557e-06,
+      "loss": 0.6192,
+      "step": 6019
+    },
+    {
+      "epoch": 1.598168060533652,
+      "grad_norm": 0.3937157037597357,
+      "learning_rate": 4.1932039290643534e-06,
+      "loss": 0.5982,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5984335590070358,
+      "grad_norm": 0.41644591247177665,
+      "learning_rate": 4.192947050797527e-06,
+      "loss": 0.6061,
+      "step": 6021
+    },
+    {
+      "epoch": 1.5986990574804194,
+      "grad_norm": 0.4342225597248977,
+      "learning_rate": 4.192690139514088e-06,
+      "loss": 0.5792,
+      "step": 6022
+    },
+    {
+      "epoch": 1.5989645559538033,
+      "grad_norm": 0.41188019422404626,
+      "learning_rate": 4.192433195219047e-06,
+      "loss": 0.5776,
+      "step": 6023
+    },
+    {
+      "epoch": 1.5992300544271871,
+      "grad_norm": 0.4246544101093638,
+      "learning_rate": 4.1921762179174145e-06,
+      "loss": 0.5981,
+      "step": 6024
+    },
+    {
+      "epoch": 1.5994955529005708,
+      "grad_norm": 0.42876932450653327,
+      "learning_rate": 4.191919207614203e-06,
+      "loss": 0.5863,
+      "step": 6025
+    },
+    {
+      "epoch": 1.5997610513739546,
+      "grad_norm": 0.40472826607595974,
+      "learning_rate": 4.191662164314424e-06,
+      "loss": 0.5915,
+      "step": 6026
+    },
+    {
+      "epoch": 1.6000265498473385,
+      "grad_norm": 0.41168054434141055,
+      "learning_rate": 4.19140508802309e-06,
+      "loss": 0.564,
+      "step": 6027
+    },
+    {
+      "epoch": 1.6002920483207221,
+      "grad_norm": 0.3986275230207836,
+      "learning_rate": 4.191147978745218e-06,
+      "loss": 0.5798,
+      "step": 6028
+    },
+    {
+      "epoch": 1.6005575467941058,
+      "grad_norm": 0.40809307681144,
+      "learning_rate": 4.1908908364858185e-06,
+      "loss": 0.5786,
+      "step": 6029
+    },
+    {
+      "epoch": 1.6008230452674899,
+      "grad_norm": 0.4026854828048606,
+      "learning_rate": 4.190633661249907e-06,
+      "loss": 0.6128,
+      "step": 6030
+    },
+    {
+      "epoch": 1.6010885437408735,
+      "grad_norm": 0.41094788097923096,
+      "learning_rate": 4.1903764530425005e-06,
+      "loss": 0.639,
+      "step": 6031
+    },
+    {
+      "epoch": 1.6013540422142571,
+      "grad_norm": 0.4049171393338981,
+      "learning_rate": 4.190119211868614e-06,
+      "loss": 0.5963,
+      "step": 6032
+    },
+    {
+      "epoch": 1.601619540687641,
+      "grad_norm": 0.41232436452152416,
+      "learning_rate": 4.189861937733265e-06,
+      "loss": 0.6036,
+      "step": 6033
+    },
+    {
+      "epoch": 1.6018850391610249,
+      "grad_norm": 0.42445584696695715,
+      "learning_rate": 4.1896046306414705e-06,
+      "loss": 0.5707,
+      "step": 6034
+    },
+    {
+      "epoch": 1.6021505376344085,
+      "grad_norm": 0.40834564499552445,
+      "learning_rate": 4.189347290598249e-06,
+      "loss": 0.5594,
+      "step": 6035
+    },
+    {
+      "epoch": 1.6024160361077924,
+      "grad_norm": 0.41454216488797413,
+      "learning_rate": 4.1890899176086194e-06,
+      "loss": 0.575,
+      "step": 6036
+    },
+    {
+      "epoch": 1.6026815345811762,
+      "grad_norm": 0.4105445653650947,
+      "learning_rate": 4.1888325116776004e-06,
+      "loss": 0.5912,
+      "step": 6037
+    },
+    {
+      "epoch": 1.6029470330545599,
+      "grad_norm": 0.41284401700146206,
+      "learning_rate": 4.1885750728102135e-06,
+      "loss": 0.5906,
+      "step": 6038
+    },
+    {
+      "epoch": 1.6032125315279437,
+      "grad_norm": 0.42300140448743107,
+      "learning_rate": 4.1883176010114764e-06,
+      "loss": 0.6081,
+      "step": 6039
+    },
+    {
+      "epoch": 1.6034780300013276,
+      "grad_norm": 0.3989886367238546,
+      "learning_rate": 4.188060096286414e-06,
+      "loss": 0.5505,
+      "step": 6040
+    },
+    {
+      "epoch": 1.6037435284747112,
+      "grad_norm": 0.4086036100972658,
+      "learning_rate": 4.187802558640046e-06,
+      "loss": 0.576,
+      "step": 6041
+    },
+    {
+      "epoch": 1.6040090269480949,
+      "grad_norm": 0.4034879193494364,
+      "learning_rate": 4.187544988077395e-06,
+      "loss": 0.5413,
+      "step": 6042
+    },
+    {
+      "epoch": 1.604274525421479,
+      "grad_norm": 0.3935537817437524,
+      "learning_rate": 4.187287384603486e-06,
+      "loss": 0.5744,
+      "step": 6043
+    },
+    {
+      "epoch": 1.6045400238948626,
+      "grad_norm": 0.4040200452183954,
+      "learning_rate": 4.187029748223341e-06,
+      "loss": 0.5938,
+      "step": 6044
+    },
+    {
+      "epoch": 1.6048055223682463,
+      "grad_norm": 0.41149813239860483,
+      "learning_rate": 4.186772078941986e-06,
+      "loss": 0.5564,
+      "step": 6045
+    },
+    {
+      "epoch": 1.6050710208416301,
+      "grad_norm": 0.40439805725675165,
+      "learning_rate": 4.186514376764446e-06,
+      "loss": 0.5449,
+      "step": 6046
+    },
+    {
+      "epoch": 1.605336519315014,
+      "grad_norm": 0.40787848283245315,
+      "learning_rate": 4.186256641695745e-06,
+      "loss": 0.5689,
+      "step": 6047
+    },
+    {
+      "epoch": 1.6056020177883976,
+      "grad_norm": 0.3926588912516516,
+      "learning_rate": 4.1859988737409114e-06,
+      "loss": 0.5778,
+      "step": 6048
+    },
+    {
+      "epoch": 1.6058675162617815,
+      "grad_norm": 0.41539247350104747,
+      "learning_rate": 4.185741072904972e-06,
+      "loss": 0.5916,
+      "step": 6049
+    },
+    {
+      "epoch": 1.6061330147351653,
+      "grad_norm": 0.4298756280854937,
+      "learning_rate": 4.185483239192954e-06,
+      "loss": 0.6093,
+      "step": 6050
+    },
+    {
+      "epoch": 1.606398513208549,
+      "grad_norm": 0.41554725769750595,
+      "learning_rate": 4.185225372609886e-06,
+      "loss": 0.585,
+      "step": 6051
+    },
+    {
+      "epoch": 1.6066640116819328,
+      "grad_norm": 0.4038829145919097,
+      "learning_rate": 4.184967473160797e-06,
+      "loss": 0.5752,
+      "step": 6052
+    },
+    {
+      "epoch": 1.6069295101553167,
+      "grad_norm": 0.4057289541482558,
+      "learning_rate": 4.184709540850717e-06,
+      "loss": 0.5947,
+      "step": 6053
+    },
+    {
+      "epoch": 1.6071950086287004,
+      "grad_norm": 0.4034778598210146,
+      "learning_rate": 4.184451575684677e-06,
+      "loss": 0.5968,
+      "step": 6054
+    },
+    {
+      "epoch": 1.6074605071020842,
+      "grad_norm": 0.4092400161315561,
+      "learning_rate": 4.1841935776677055e-06,
+      "loss": 0.531,
+      "step": 6055
+    },
+    {
+      "epoch": 1.607726005575468,
+      "grad_norm": 0.41883555978447745,
+      "learning_rate": 4.183935546804836e-06,
+      "loss": 0.5784,
+      "step": 6056
+    },
+    {
+      "epoch": 1.6079915040488517,
+      "grad_norm": 0.4083646428369843,
+      "learning_rate": 4.183677483101101e-06,
+      "loss": 0.5878,
+      "step": 6057
+    },
+    {
+      "epoch": 1.6082570025222354,
+      "grad_norm": 0.39761457937439987,
+      "learning_rate": 4.183419386561532e-06,
+      "loss": 0.6052,
+      "step": 6058
+    },
+    {
+      "epoch": 1.6085225009956192,
+      "grad_norm": 0.40667292836753893,
+      "learning_rate": 4.183161257191163e-06,
+      "loss": 0.5998,
+      "step": 6059
+    },
+    {
+      "epoch": 1.608787999469003,
+      "grad_norm": 0.4000448845102318,
+      "learning_rate": 4.18290309499503e-06,
+      "loss": 0.597,
+      "step": 6060
+    },
+    {
+      "epoch": 1.6090534979423867,
+      "grad_norm": 0.4153649457019075,
+      "learning_rate": 4.1826448999781646e-06,
+      "loss": 0.617,
+      "step": 6061
+    },
+    {
+      "epoch": 1.6093189964157706,
+      "grad_norm": 0.4097173663044732,
+      "learning_rate": 4.182386672145604e-06,
+      "loss": 0.5489,
+      "step": 6062
+    },
+    {
+      "epoch": 1.6095844948891544,
+      "grad_norm": 0.40404261362103244,
+      "learning_rate": 4.1821284115023845e-06,
+      "loss": 0.5666,
+      "step": 6063
+    },
+    {
+      "epoch": 1.609849993362538,
+      "grad_norm": 0.4312765154231968,
+      "learning_rate": 4.181870118053543e-06,
+      "loss": 0.5771,
+      "step": 6064
+    },
+    {
+      "epoch": 1.610115491835922,
+      "grad_norm": 0.427891179051653,
+      "learning_rate": 4.181611791804115e-06,
+      "loss": 0.6098,
+      "step": 6065
+    },
+    {
+      "epoch": 1.6103809903093058,
+      "grad_norm": 0.40751219237961833,
+      "learning_rate": 4.1813534327591405e-06,
+      "loss": 0.6018,
+      "step": 6066
+    },
+    {
+      "epoch": 1.6106464887826895,
+      "grad_norm": 0.4150417504875353,
+      "learning_rate": 4.1810950409236575e-06,
+      "loss": 0.5792,
+      "step": 6067
+    },
+    {
+      "epoch": 1.6109119872560733,
+      "grad_norm": 0.40469706168129826,
+      "learning_rate": 4.180836616302704e-06,
+      "loss": 0.5881,
+      "step": 6068
+    },
+    {
+      "epoch": 1.6111774857294572,
+      "grad_norm": 0.39557641769068447,
+      "learning_rate": 4.180578158901323e-06,
+      "loss": 0.5948,
+      "step": 6069
+    },
+    {
+      "epoch": 1.6114429842028408,
+      "grad_norm": 0.40207327005497584,
+      "learning_rate": 4.180319668724551e-06,
+      "loss": 0.5793,
+      "step": 6070
+    },
+    {
+      "epoch": 1.6117084826762245,
+      "grad_norm": 0.3932159742067176,
+      "learning_rate": 4.180061145777433e-06,
+      "loss": 0.5839,
+      "step": 6071
+    },
+    {
+      "epoch": 1.6119739811496085,
+      "grad_norm": 0.41540833788977327,
+      "learning_rate": 4.1798025900650086e-06,
+      "loss": 0.6116,
+      "step": 6072
+    },
+    {
+      "epoch": 1.6122394796229922,
+      "grad_norm": 0.4045738217370343,
+      "learning_rate": 4.179544001592321e-06,
+      "loss": 0.5687,
+      "step": 6073
+    },
+    {
+      "epoch": 1.6125049780963758,
+      "grad_norm": 0.4098193846988828,
+      "learning_rate": 4.179285380364413e-06,
+      "loss": 0.5994,
+      "step": 6074
+    },
+    {
+      "epoch": 1.6127704765697597,
+      "grad_norm": 0.4160047075260533,
+      "learning_rate": 4.179026726386329e-06,
+      "loss": 0.622,
+      "step": 6075
+    },
+    {
+      "epoch": 1.6130359750431436,
+      "grad_norm": 0.4121566456455558,
+      "learning_rate": 4.178768039663111e-06,
+      "loss": 0.6006,
+      "step": 6076
+    },
+    {
+      "epoch": 1.6133014735165272,
+      "grad_norm": 0.4213164058973677,
+      "learning_rate": 4.178509320199808e-06,
+      "loss": 0.5687,
+      "step": 6077
+    },
+    {
+      "epoch": 1.613566971989911,
+      "grad_norm": 0.41204682901178885,
+      "learning_rate": 4.1782505680014625e-06,
+      "loss": 0.5979,
+      "step": 6078
+    },
+    {
+      "epoch": 1.613832470463295,
+      "grad_norm": 0.41544248228873504,
+      "learning_rate": 4.177991783073123e-06,
+      "loss": 0.5945,
+      "step": 6079
+    },
+    {
+      "epoch": 1.6140979689366786,
+      "grad_norm": 0.4021412627030649,
+      "learning_rate": 4.177732965419834e-06,
+      "loss": 0.5315,
+      "step": 6080
+    },
+    {
+      "epoch": 1.6143634674100624,
+      "grad_norm": 0.4047176103043243,
+      "learning_rate": 4.177474115046645e-06,
+      "loss": 0.6031,
+      "step": 6081
+    },
+    {
+      "epoch": 1.6146289658834463,
+      "grad_norm": 0.42015720900913267,
+      "learning_rate": 4.177215231958604e-06,
+      "loss": 0.5623,
+      "step": 6082
+    },
+    {
+      "epoch": 1.61489446435683,
+      "grad_norm": 0.4019744291178643,
+      "learning_rate": 4.176956316160759e-06,
+      "loss": 0.5612,
+      "step": 6083
+    },
+    {
+      "epoch": 1.6151599628302136,
+      "grad_norm": 0.4203494210920179,
+      "learning_rate": 4.176697367658159e-06,
+      "loss": 0.6212,
+      "step": 6084
+    },
+    {
+      "epoch": 1.6154254613035977,
+      "grad_norm": 0.41243660741486005,
+      "learning_rate": 4.176438386455856e-06,
+      "loss": 0.6063,
+      "step": 6085
+    },
+    {
+      "epoch": 1.6156909597769813,
+      "grad_norm": 0.40795909220205034,
+      "learning_rate": 4.1761793725588995e-06,
+      "loss": 0.5536,
+      "step": 6086
+    },
+    {
+      "epoch": 1.615956458250365,
+      "grad_norm": 0.42411925222603436,
+      "learning_rate": 4.175920325972341e-06,
+      "loss": 0.5951,
+      "step": 6087
+    },
+    {
+      "epoch": 1.6162219567237488,
+      "grad_norm": 0.4036072881164683,
+      "learning_rate": 4.175661246701234e-06,
+      "loss": 0.5804,
+      "step": 6088
+    },
+    {
+      "epoch": 1.6164874551971327,
+      "grad_norm": 0.42126595905352393,
+      "learning_rate": 4.175402134750629e-06,
+      "loss": 0.5979,
+      "step": 6089
+    },
+    {
+      "epoch": 1.6167529536705163,
+      "grad_norm": 0.4089282972814364,
+      "learning_rate": 4.175142990125581e-06,
+      "loss": 0.5587,
+      "step": 6090
+    },
+    {
+      "epoch": 1.6170184521439002,
+      "grad_norm": 0.40842373125551007,
+      "learning_rate": 4.174883812831142e-06,
+      "loss": 0.6072,
+      "step": 6091
+    },
+    {
+      "epoch": 1.617283950617284,
+      "grad_norm": 0.41307635953059874,
+      "learning_rate": 4.174624602872369e-06,
+      "loss": 0.6028,
+      "step": 6092
+    },
+    {
+      "epoch": 1.6175494490906677,
+      "grad_norm": 0.4041236531775181,
+      "learning_rate": 4.174365360254316e-06,
+      "loss": 0.5764,
+      "step": 6093
+    },
+    {
+      "epoch": 1.6178149475640515,
+      "grad_norm": 0.4160419026301076,
+      "learning_rate": 4.174106084982038e-06,
+      "loss": 0.6037,
+      "step": 6094
+    },
+    {
+      "epoch": 1.6180804460374354,
+      "grad_norm": 0.4084376416196963,
+      "learning_rate": 4.173846777060593e-06,
+      "loss": 0.5799,
+      "step": 6095
+    },
+    {
+      "epoch": 1.618345944510819,
+      "grad_norm": 0.42601488847702534,
+      "learning_rate": 4.1735874364950364e-06,
+      "loss": 0.6252,
+      "step": 6096
+    },
+    {
+      "epoch": 1.6186114429842027,
+      "grad_norm": 0.40899829897887763,
+      "learning_rate": 4.173328063290429e-06,
+      "loss": 0.5371,
+      "step": 6097
+    },
+    {
+      "epoch": 1.6188769414575868,
+      "grad_norm": 0.41847364874322346,
+      "learning_rate": 4.173068657451827e-06,
+      "loss": 0.5855,
+      "step": 6098
+    },
+    {
+      "epoch": 1.6191424399309704,
+      "grad_norm": 0.43040027552388593,
+      "learning_rate": 4.172809218984289e-06,
+      "loss": 0.6033,
+      "step": 6099
+    },
+    {
+      "epoch": 1.619407938404354,
+      "grad_norm": 0.42051275205866556,
+      "learning_rate": 4.1725497478928765e-06,
+      "loss": 0.5483,
+      "step": 6100
+    },
+    {
+      "epoch": 1.619673436877738,
+      "grad_norm": 0.40658005765987104,
+      "learning_rate": 4.172290244182647e-06,
+      "loss": 0.5613,
+      "step": 6101
+    },
+    {
+      "epoch": 1.6199389353511218,
+      "grad_norm": 0.4144136657336713,
+      "learning_rate": 4.172030707858665e-06,
+      "loss": 0.5799,
+      "step": 6102
+    },
+    {
+      "epoch": 1.6202044338245054,
+      "grad_norm": 0.414170883594481,
+      "learning_rate": 4.1717711389259905e-06,
+      "loss": 0.5949,
+      "step": 6103
+    },
+    {
+      "epoch": 1.6204699322978893,
+      "grad_norm": 0.4137007605600363,
+      "learning_rate": 4.171511537389684e-06,
+      "loss": 0.6115,
+      "step": 6104
+    },
+    {
+      "epoch": 1.6207354307712731,
+      "grad_norm": 0.4074563715173919,
+      "learning_rate": 4.171251903254812e-06,
+      "loss": 0.5686,
+      "step": 6105
+    },
+    {
+      "epoch": 1.6210009292446568,
+      "grad_norm": 0.40335924884148944,
+      "learning_rate": 4.170992236526434e-06,
+      "loss": 0.5591,
+      "step": 6106
+    },
+    {
+      "epoch": 1.6212664277180406,
+      "grad_norm": 0.4070786816692419,
+      "learning_rate": 4.170732537209617e-06,
+      "loss": 0.562,
+      "step": 6107
+    },
+    {
+      "epoch": 1.6215319261914245,
+      "grad_norm": 0.40881955865293984,
+      "learning_rate": 4.170472805309424e-06,
+      "loss": 0.5731,
+      "step": 6108
+    },
+    {
+      "epoch": 1.6217974246648081,
+      "grad_norm": 0.4050546172988652,
+      "learning_rate": 4.1702130408309226e-06,
+      "loss": 0.6221,
+      "step": 6109
+    },
+    {
+      "epoch": 1.622062923138192,
+      "grad_norm": 0.39869275073453253,
+      "learning_rate": 4.169953243779177e-06,
+      "loss": 0.5734,
+      "step": 6110
+    },
+    {
+      "epoch": 1.6223284216115759,
+      "grad_norm": 0.4057329653986426,
+      "learning_rate": 4.169693414159255e-06,
+      "loss": 0.5969,
+      "step": 6111
+    },
+    {
+      "epoch": 1.6225939200849595,
+      "grad_norm": 0.4232718662098156,
+      "learning_rate": 4.169433551976222e-06,
+      "loss": 0.593,
+      "step": 6112
+    },
+    {
+      "epoch": 1.6228594185583431,
+      "grad_norm": 0.41817767050446764,
+      "learning_rate": 4.1691736572351485e-06,
+      "loss": 0.6183,
+      "step": 6113
+    },
+    {
+      "epoch": 1.6231249170317272,
+      "grad_norm": 0.4242438733109001,
+      "learning_rate": 4.168913729941101e-06,
+      "loss": 0.5964,
+      "step": 6114
+    },
+    {
+      "epoch": 1.6233904155051109,
+      "grad_norm": 0.41650494557179485,
+      "learning_rate": 4.1686537700991505e-06,
+      "loss": 0.6018,
+      "step": 6115
+    },
+    {
+      "epoch": 1.6236559139784945,
+      "grad_norm": 0.41738744719303306,
+      "learning_rate": 4.168393777714364e-06,
+      "loss": 0.5742,
+      "step": 6116
+    },
+    {
+      "epoch": 1.6239214124518784,
+      "grad_norm": 0.40680178549473783,
+      "learning_rate": 4.168133752791815e-06,
+      "loss": 0.5983,
+      "step": 6117
+    },
+    {
+      "epoch": 1.6241869109252622,
+      "grad_norm": 0.4031023316747236,
+      "learning_rate": 4.167873695336575e-06,
+      "loss": 0.5914,
+      "step": 6118
+    },
+    {
+      "epoch": 1.6244524093986459,
+      "grad_norm": 0.42387610979480694,
+      "learning_rate": 4.167613605353712e-06,
+      "loss": 0.6312,
+      "step": 6119
+    },
+    {
+      "epoch": 1.6247179078720297,
+      "grad_norm": 0.42011859217972486,
+      "learning_rate": 4.167353482848302e-06,
+      "loss": 0.5983,
+      "step": 6120
+    },
+    {
+      "epoch": 1.6249834063454136,
+      "grad_norm": 0.4129244650761646,
+      "learning_rate": 4.167093327825417e-06,
+      "loss": 0.6035,
+      "step": 6121
+    },
+    {
+      "epoch": 1.6252489048187972,
+      "grad_norm": 0.3974062132472111,
+      "learning_rate": 4.166833140290128e-06,
+      "loss": 0.5511,
+      "step": 6122
+    },
+    {
+      "epoch": 1.625514403292181,
+      "grad_norm": 0.41200410883365246,
+      "learning_rate": 4.166572920247514e-06,
+      "loss": 0.574,
+      "step": 6123
+    },
+    {
+      "epoch": 1.625779901765565,
+      "grad_norm": 0.4030772574830196,
+      "learning_rate": 4.166312667702647e-06,
+      "loss": 0.581,
+      "step": 6124
+    },
+    {
+      "epoch": 1.6260454002389486,
+      "grad_norm": 0.40076109867172627,
+      "learning_rate": 4.166052382660604e-06,
+      "loss": 0.5719,
+      "step": 6125
+    },
+    {
+      "epoch": 1.6263108987123323,
+      "grad_norm": 0.4301905162364889,
+      "learning_rate": 4.165792065126459e-06,
+      "loss": 0.6099,
+      "step": 6126
+    },
+    {
+      "epoch": 1.6265763971857163,
+      "grad_norm": 0.4014776446171743,
+      "learning_rate": 4.1655317151052906e-06,
+      "loss": 0.5874,
+      "step": 6127
+    },
+    {
+      "epoch": 1.6268418956591,
+      "grad_norm": 0.4083216500620904,
+      "learning_rate": 4.1652713326021756e-06,
+      "loss": 0.5822,
+      "step": 6128
+    },
+    {
+      "epoch": 1.6271073941324836,
+      "grad_norm": 0.4242217715503586,
+      "learning_rate": 4.165010917622193e-06,
+      "loss": 0.5885,
+      "step": 6129
+    },
+    {
+      "epoch": 1.6273728926058675,
+      "grad_norm": 0.43271239320364385,
+      "learning_rate": 4.16475047017042e-06,
+      "loss": 0.5656,
+      "step": 6130
+    },
+    {
+      "epoch": 1.6276383910792513,
+      "grad_norm": 0.41445016707339577,
+      "learning_rate": 4.164489990251937e-06,
+      "loss": 0.5795,
+      "step": 6131
+    },
+    {
+      "epoch": 1.627903889552635,
+      "grad_norm": 0.4263278103983279,
+      "learning_rate": 4.164229477871824e-06,
+      "loss": 0.598,
+      "step": 6132
+    },
+    {
+      "epoch": 1.6281693880260188,
+      "grad_norm": 0.4155773499362417,
+      "learning_rate": 4.163968933035162e-06,
+      "loss": 0.5871,
+      "step": 6133
+    },
+    {
+      "epoch": 1.6284348864994027,
+      "grad_norm": 0.4208296658953933,
+      "learning_rate": 4.16370835574703e-06,
+      "loss": 0.6264,
+      "step": 6134
+    },
+    {
+      "epoch": 1.6287003849727864,
+      "grad_norm": 0.41994182283039133,
+      "learning_rate": 4.163447746012513e-06,
+      "loss": 0.5781,
+      "step": 6135
+    },
+    {
+      "epoch": 1.6289658834461702,
+      "grad_norm": 0.4024937743711282,
+      "learning_rate": 4.163187103836692e-06,
+      "loss": 0.5724,
+      "step": 6136
+    },
+    {
+      "epoch": 1.629231381919554,
+      "grad_norm": 0.4092982827039093,
+      "learning_rate": 4.1629264292246505e-06,
+      "loss": 0.5783,
+      "step": 6137
+    },
+    {
+      "epoch": 1.6294968803929377,
+      "grad_norm": 0.4144124204534514,
+      "learning_rate": 4.162665722181472e-06,
+      "loss": 0.6145,
+      "step": 6138
+    },
+    {
+      "epoch": 1.6297623788663214,
+      "grad_norm": 0.4235307821490689,
+      "learning_rate": 4.162404982712241e-06,
+      "loss": 0.5564,
+      "step": 6139
+    },
+    {
+      "epoch": 1.6300278773397054,
+      "grad_norm": 0.421892964490945,
+      "learning_rate": 4.1621442108220425e-06,
+      "loss": 0.5626,
+      "step": 6140
+    },
+    {
+      "epoch": 1.630293375813089,
+      "grad_norm": 0.42406328523486675,
+      "learning_rate": 4.161883406515962e-06,
+      "loss": 0.6085,
+      "step": 6141
+    },
+    {
+      "epoch": 1.6305588742864727,
+      "grad_norm": 0.4105191092035432,
+      "learning_rate": 4.161622569799086e-06,
+      "loss": 0.605,
+      "step": 6142
+    },
+    {
+      "epoch": 1.6308243727598566,
+      "grad_norm": 0.4006804805908787,
+      "learning_rate": 4.1613617006765016e-06,
+      "loss": 0.5936,
+      "step": 6143
+    },
+    {
+      "epoch": 1.6310898712332405,
+      "grad_norm": 0.38996310795407324,
+      "learning_rate": 4.161100799153297e-06,
+      "loss": 0.5865,
+      "step": 6144
+    },
+    {
+      "epoch": 1.631355369706624,
+      "grad_norm": 0.41671128644001837,
+      "learning_rate": 4.160839865234558e-06,
+      "loss": 0.6017,
+      "step": 6145
+    },
+    {
+      "epoch": 1.631620868180008,
+      "grad_norm": 0.4001074347833242,
+      "learning_rate": 4.160578898925377e-06,
+      "loss": 0.5526,
+      "step": 6146
+    },
+    {
+      "epoch": 1.6318863666533918,
+      "grad_norm": 0.3977811634343154,
+      "learning_rate": 4.160317900230841e-06,
+      "loss": 0.5923,
+      "step": 6147
+    },
+    {
+      "epoch": 1.6321518651267755,
+      "grad_norm": 0.4067900207900259,
+      "learning_rate": 4.160056869156041e-06,
+      "loss": 0.5642,
+      "step": 6148
+    },
+    {
+      "epoch": 1.6324173636001593,
+      "grad_norm": 0.4027236458197424,
+      "learning_rate": 4.1597958057060675e-06,
+      "loss": 0.5976,
+      "step": 6149
+    },
+    {
+      "epoch": 1.6326828620735432,
+      "grad_norm": 0.40172676177857874,
+      "learning_rate": 4.159534709886012e-06,
+      "loss": 0.5961,
+      "step": 6150
+    },
+    {
+      "epoch": 1.6329483605469268,
+      "grad_norm": 0.40478471902325897,
+      "learning_rate": 4.159273581700965e-06,
+      "loss": 0.5945,
+      "step": 6151
+    },
+    {
+      "epoch": 1.6332138590203107,
+      "grad_norm": 0.40050833918555934,
+      "learning_rate": 4.159012421156023e-06,
+      "loss": 0.5799,
+      "step": 6152
+    },
+    {
+      "epoch": 1.6334793574936946,
+      "grad_norm": 0.40050030821461574,
+      "learning_rate": 4.158751228256275e-06,
+      "loss": 0.603,
+      "step": 6153
+    },
+    {
+      "epoch": 1.6337448559670782,
+      "grad_norm": 0.3939591489093314,
+      "learning_rate": 4.158490003006817e-06,
+      "loss": 0.5409,
+      "step": 6154
+    },
+    {
+      "epoch": 1.6340103544404618,
+      "grad_norm": 0.4025942853463516,
+      "learning_rate": 4.158228745412745e-06,
+      "loss": 0.5724,
+      "step": 6155
+    },
+    {
+      "epoch": 1.6342758529138457,
+      "grad_norm": 0.4312021363071985,
+      "learning_rate": 4.1579674554791505e-06,
+      "loss": 0.6074,
+      "step": 6156
+    },
+    {
+      "epoch": 1.6345413513872296,
+      "grad_norm": 0.41423615721707946,
+      "learning_rate": 4.1577061332111316e-06,
+      "loss": 0.5903,
+      "step": 6157
+    },
+    {
+      "epoch": 1.6348068498606132,
+      "grad_norm": 0.41035197940726176,
+      "learning_rate": 4.157444778613784e-06,
+      "loss": 0.5845,
+      "step": 6158
+    },
+    {
+      "epoch": 1.635072348333997,
+      "grad_norm": 0.41212990153842805,
+      "learning_rate": 4.157183391692206e-06,
+      "loss": 0.5487,
+      "step": 6159
+    },
+    {
+      "epoch": 1.635337846807381,
+      "grad_norm": 0.41695461209939216,
+      "learning_rate": 4.156921972451494e-06,
+      "loss": 0.5738,
+      "step": 6160
+    },
+    {
+      "epoch": 1.6356033452807646,
+      "grad_norm": 0.41434622391201165,
+      "learning_rate": 4.1566605208967465e-06,
+      "loss": 0.6107,
+      "step": 6161
+    },
+    {
+      "epoch": 1.6358688437541484,
+      "grad_norm": 0.40577748694263566,
+      "learning_rate": 4.156399037033063e-06,
+      "loss": 0.5594,
+      "step": 6162
+    },
+    {
+      "epoch": 1.6361343422275323,
+      "grad_norm": 0.4125891360352178,
+      "learning_rate": 4.156137520865542e-06,
+      "loss": 0.5804,
+      "step": 6163
+    },
+    {
+      "epoch": 1.636399840700916,
+      "grad_norm": 0.42704791950403886,
+      "learning_rate": 4.1558759723992845e-06,
+      "loss": 0.589,
+      "step": 6164
+    },
+    {
+      "epoch": 1.6366653391742998,
+      "grad_norm": 0.4106778160593796,
+      "learning_rate": 4.155614391639392e-06,
+      "loss": 0.5958,
+      "step": 6165
+    },
+    {
+      "epoch": 1.6369308376476837,
+      "grad_norm": 0.4121664073279707,
+      "learning_rate": 4.155352778590964e-06,
+      "loss": 0.5783,
+      "step": 6166
+    },
+    {
+      "epoch": 1.6371963361210673,
+      "grad_norm": 0.41224868821400934,
+      "learning_rate": 4.155091133259105e-06,
+      "loss": 0.6216,
+      "step": 6167
+    },
+    {
+      "epoch": 1.637461834594451,
+      "grad_norm": 0.41527339349457415,
+      "learning_rate": 4.154829455648916e-06,
+      "loss": 0.5909,
+      "step": 6168
+    },
+    {
+      "epoch": 1.637727333067835,
+      "grad_norm": 0.4078061501894263,
+      "learning_rate": 4.154567745765501e-06,
+      "loss": 0.5594,
+      "step": 6169
+    },
+    {
+      "epoch": 1.6379928315412187,
+      "grad_norm": 0.3956864337507178,
+      "learning_rate": 4.154306003613965e-06,
+      "loss": 0.5724,
+      "step": 6170
+    },
+    {
+      "epoch": 1.6382583300146023,
+      "grad_norm": 0.3981378344183963,
+      "learning_rate": 4.15404422919941e-06,
+      "loss": 0.5263,
+      "step": 6171
+    },
+    {
+      "epoch": 1.6385238284879862,
+      "grad_norm": 0.4179824345946913,
+      "learning_rate": 4.153782422526943e-06,
+      "loss": 0.5875,
+      "step": 6172
+    },
+    {
+      "epoch": 1.63878932696137,
+      "grad_norm": 0.39134723518773457,
+      "learning_rate": 4.15352058360167e-06,
+      "loss": 0.5515,
+      "step": 6173
+    },
+    {
+      "epoch": 1.6390548254347537,
+      "grad_norm": 0.39961521046893517,
+      "learning_rate": 4.153258712428697e-06,
+      "loss": 0.6104,
+      "step": 6174
+    },
+    {
+      "epoch": 1.6393203239081375,
+      "grad_norm": 0.41425705749840463,
+      "learning_rate": 4.152996809013131e-06,
+      "loss": 0.5731,
+      "step": 6175
+    },
+    {
+      "epoch": 1.6395858223815214,
+      "grad_norm": 0.45333256468870214,
+      "learning_rate": 4.15273487336008e-06,
+      "loss": 0.584,
+      "step": 6176
+    },
+    {
+      "epoch": 1.639851320854905,
+      "grad_norm": 0.3958538958844591,
+      "learning_rate": 4.152472905474653e-06,
+      "loss": 0.5849,
+      "step": 6177
+    },
+    {
+      "epoch": 1.640116819328289,
+      "grad_norm": 0.4015134633521911,
+      "learning_rate": 4.152210905361958e-06,
+      "loss": 0.5867,
+      "step": 6178
+    },
+    {
+      "epoch": 1.6403823178016728,
+      "grad_norm": 0.40141478694589433,
+      "learning_rate": 4.151948873027105e-06,
+      "loss": 0.6068,
+      "step": 6179
+    },
+    {
+      "epoch": 1.6406478162750564,
+      "grad_norm": 0.42398501794559523,
+      "learning_rate": 4.151686808475204e-06,
+      "loss": 0.5731,
+      "step": 6180
+    },
+    {
+      "epoch": 1.64091331474844,
+      "grad_norm": 0.39558618040866333,
+      "learning_rate": 4.151424711711366e-06,
+      "loss": 0.6024,
+      "step": 6181
+    },
+    {
+      "epoch": 1.6411788132218241,
+      "grad_norm": 0.4220607933628933,
+      "learning_rate": 4.151162582740703e-06,
+      "loss": 0.5831,
+      "step": 6182
+    },
+    {
+      "epoch": 1.6414443116952078,
+      "grad_norm": 0.3968453065135876,
+      "learning_rate": 4.150900421568328e-06,
+      "loss": 0.5617,
+      "step": 6183
+    },
+    {
+      "epoch": 1.6417098101685914,
+      "grad_norm": 0.41703052185711126,
+      "learning_rate": 4.150638228199352e-06,
+      "loss": 0.5792,
+      "step": 6184
+    },
+    {
+      "epoch": 1.6419753086419753,
+      "grad_norm": 0.4094710152008768,
+      "learning_rate": 4.150376002638888e-06,
+      "loss": 0.5735,
+      "step": 6185
+    },
+    {
+      "epoch": 1.6422408071153591,
+      "grad_norm": 0.4041939781155795,
+      "learning_rate": 4.150113744892053e-06,
+      "loss": 0.5687,
+      "step": 6186
+    },
+    {
+      "epoch": 1.6425063055887428,
+      "grad_norm": 0.4049060040768663,
+      "learning_rate": 4.1498514549639575e-06,
+      "loss": 0.5877,
+      "step": 6187
+    },
+    {
+      "epoch": 1.6427718040621266,
+      "grad_norm": 0.4113525493215451,
+      "learning_rate": 4.149589132859721e-06,
+      "loss": 0.6165,
+      "step": 6188
+    },
+    {
+      "epoch": 1.6430373025355105,
+      "grad_norm": 0.40495989894501877,
+      "learning_rate": 4.149326778584457e-06,
+      "loss": 0.6204,
+      "step": 6189
+    },
+    {
+      "epoch": 1.6433028010088941,
+      "grad_norm": 0.41350056409386493,
+      "learning_rate": 4.149064392143282e-06,
+      "loss": 0.5889,
+      "step": 6190
+    },
+    {
+      "epoch": 1.643568299482278,
+      "grad_norm": 0.40710807425726075,
+      "learning_rate": 4.1488019735413145e-06,
+      "loss": 0.5504,
+      "step": 6191
+    },
+    {
+      "epoch": 1.6438337979556619,
+      "grad_norm": 0.41068614620172866,
+      "learning_rate": 4.148539522783671e-06,
+      "loss": 0.5608,
+      "step": 6192
+    },
+    {
+      "epoch": 1.6440992964290455,
+      "grad_norm": 0.40957905468107303,
+      "learning_rate": 4.148277039875471e-06,
+      "loss": 0.5783,
+      "step": 6193
+    },
+    {
+      "epoch": 1.6443647949024291,
+      "grad_norm": 0.40382362750619155,
+      "learning_rate": 4.1480145248218325e-06,
+      "loss": 0.5468,
+      "step": 6194
+    },
+    {
+      "epoch": 1.6446302933758132,
+      "grad_norm": 0.4195215216697413,
+      "learning_rate": 4.147751977627876e-06,
+      "loss": 0.5671,
+      "step": 6195
+    },
+    {
+      "epoch": 1.6448957918491969,
+      "grad_norm": 0.4106618374739023,
+      "learning_rate": 4.147489398298722e-06,
+      "loss": 0.6221,
+      "step": 6196
+    },
+    {
+      "epoch": 1.6451612903225805,
+      "grad_norm": 0.40584045928985996,
+      "learning_rate": 4.14722678683949e-06,
+      "loss": 0.5791,
+      "step": 6197
+    },
+    {
+      "epoch": 1.6454267887959644,
+      "grad_norm": 0.40636200741855466,
+      "learning_rate": 4.146964143255304e-06,
+      "loss": 0.5786,
+      "step": 6198
+    },
+    {
+      "epoch": 1.6456922872693482,
+      "grad_norm": 0.40943823554126807,
+      "learning_rate": 4.146701467551284e-06,
+      "loss": 0.5888,
+      "step": 6199
+    },
+    {
+      "epoch": 1.6459577857427319,
+      "grad_norm": 0.40840696304599533,
+      "learning_rate": 4.1464387597325525e-06,
+      "loss": 0.5617,
+      "step": 6200
+    },
+    {
+      "epoch": 1.6462232842161157,
+      "grad_norm": 0.4158556573842827,
+      "learning_rate": 4.1461760198042355e-06,
+      "loss": 0.6274,
+      "step": 6201
+    },
+    {
+      "epoch": 1.6464887826894996,
+      "grad_norm": 0.414135298042347,
+      "learning_rate": 4.145913247771455e-06,
+      "loss": 0.6094,
+      "step": 6202
+    },
+    {
+      "epoch": 1.6467542811628832,
+      "grad_norm": 0.43872145995368317,
+      "learning_rate": 4.145650443639336e-06,
+      "loss": 0.5528,
+      "step": 6203
+    },
+    {
+      "epoch": 1.6470197796362671,
+      "grad_norm": 0.40543703759873323,
+      "learning_rate": 4.1453876074130046e-06,
+      "loss": 0.5588,
+      "step": 6204
+    },
+    {
+      "epoch": 1.647285278109651,
+      "grad_norm": 0.4043137008848618,
+      "learning_rate": 4.145124739097585e-06,
+      "loss": 0.5521,
+      "step": 6205
+    },
+    {
+      "epoch": 1.6475507765830346,
+      "grad_norm": 0.40171402098633396,
+      "learning_rate": 4.1448618386982065e-06,
+      "loss": 0.5914,
+      "step": 6206
+    },
+    {
+      "epoch": 1.6478162750564185,
+      "grad_norm": 0.42962939831299696,
+      "learning_rate": 4.144598906219995e-06,
+      "loss": 0.5761,
+      "step": 6207
+    },
+    {
+      "epoch": 1.6480817735298023,
+      "grad_norm": 0.4137795924604853,
+      "learning_rate": 4.144335941668077e-06,
+      "loss": 0.5954,
+      "step": 6208
+    },
+    {
+      "epoch": 1.648347272003186,
+      "grad_norm": 0.41575333958277927,
+      "learning_rate": 4.144072945047582e-06,
+      "loss": 0.593,
+      "step": 6209
+    },
+    {
+      "epoch": 1.6486127704765696,
+      "grad_norm": 0.395273843277485,
+      "learning_rate": 4.1438099163636406e-06,
+      "loss": 0.5566,
+      "step": 6210
+    },
+    {
+      "epoch": 1.6488782689499535,
+      "grad_norm": 0.41298271518847307,
+      "learning_rate": 4.1435468556213804e-06,
+      "loss": 0.6236,
+      "step": 6211
+    },
+    {
+      "epoch": 1.6491437674233373,
+      "grad_norm": 0.4125252242818947,
+      "learning_rate": 4.1432837628259325e-06,
+      "loss": 0.5707,
+      "step": 6212
+    },
+    {
+      "epoch": 1.649409265896721,
+      "grad_norm": 0.4253046340483282,
+      "learning_rate": 4.143020637982428e-06,
+      "loss": 0.5664,
+      "step": 6213
+    },
+    {
+      "epoch": 1.6496747643701049,
+      "grad_norm": 0.40589552261925216,
+      "learning_rate": 4.1427574810959966e-06,
+      "loss": 0.58,
+      "step": 6214
+    },
+    {
+      "epoch": 1.6499402628434887,
+      "grad_norm": 0.4038100950343004,
+      "learning_rate": 4.142494292171774e-06,
+      "loss": 0.5882,
+      "step": 6215
+    },
+    {
+      "epoch": 1.6502057613168724,
+      "grad_norm": 0.40530702140073943,
+      "learning_rate": 4.14223107121489e-06,
+      "loss": 0.5724,
+      "step": 6216
+    },
+    {
+      "epoch": 1.6504712597902562,
+      "grad_norm": 0.4203119053990801,
+      "learning_rate": 4.1419678182304805e-06,
+      "loss": 0.5779,
+      "step": 6217
+    },
+    {
+      "epoch": 1.65073675826364,
+      "grad_norm": 0.41514364892484984,
+      "learning_rate": 4.141704533223678e-06,
+      "loss": 0.5872,
+      "step": 6218
+    },
+    {
+      "epoch": 1.6510022567370237,
+      "grad_norm": 0.4045781292121557,
+      "learning_rate": 4.141441216199617e-06,
+      "loss": 0.5948,
+      "step": 6219
+    },
+    {
+      "epoch": 1.6512677552104076,
+      "grad_norm": 0.415084793543445,
+      "learning_rate": 4.141177867163433e-06,
+      "loss": 0.5741,
+      "step": 6220
+    },
+    {
+      "epoch": 1.6515332536837914,
+      "grad_norm": 0.42018925226627823,
+      "learning_rate": 4.140914486120263e-06,
+      "loss": 0.6233,
+      "step": 6221
+    },
+    {
+      "epoch": 1.651798752157175,
+      "grad_norm": 0.40602602960171846,
+      "learning_rate": 4.140651073075243e-06,
+      "loss": 0.5402,
+      "step": 6222
+    },
+    {
+      "epoch": 1.6520642506305587,
+      "grad_norm": 0.41734485862591525,
+      "learning_rate": 4.14038762803351e-06,
+      "loss": 0.559,
+      "step": 6223
+    },
+    {
+      "epoch": 1.6523297491039428,
+      "grad_norm": 0.39546839570875353,
+      "learning_rate": 4.140124151000201e-06,
+      "loss": 0.6167,
+      "step": 6224
+    },
+    {
+      "epoch": 1.6525952475773265,
+      "grad_norm": 0.4119933688878461,
+      "learning_rate": 4.1398606419804565e-06,
+      "loss": 0.6,
+      "step": 6225
+    },
+    {
+      "epoch": 1.65286074605071,
+      "grad_norm": 0.39970356730416085,
+      "learning_rate": 4.139597100979413e-06,
+      "loss": 0.5859,
+      "step": 6226
+    },
+    {
+      "epoch": 1.653126244524094,
+      "grad_norm": 0.40138469202406823,
+      "learning_rate": 4.139333528002213e-06,
+      "loss": 0.5551,
+      "step": 6227
+    },
+    {
+      "epoch": 1.6533917429974778,
+      "grad_norm": 0.4000983143834468,
+      "learning_rate": 4.139069923053995e-06,
+      "loss": 0.5891,
+      "step": 6228
+    },
+    {
+      "epoch": 1.6536572414708615,
+      "grad_norm": 0.4098431275921327,
+      "learning_rate": 4.1388062861399004e-06,
+      "loss": 0.5655,
+      "step": 6229
+    },
+    {
+      "epoch": 1.6539227399442453,
+      "grad_norm": 0.4178427771067677,
+      "learning_rate": 4.1385426172650715e-06,
+      "loss": 0.5642,
+      "step": 6230
+    },
+    {
+      "epoch": 1.6541882384176292,
+      "grad_norm": 0.3981866275904858,
+      "learning_rate": 4.138278916434648e-06,
+      "loss": 0.5802,
+      "step": 6231
+    },
+    {
+      "epoch": 1.6544537368910128,
+      "grad_norm": 0.40325194615294147,
+      "learning_rate": 4.138015183653775e-06,
+      "loss": 0.5712,
+      "step": 6232
+    },
+    {
+      "epoch": 1.6547192353643967,
+      "grad_norm": 0.39890934406055234,
+      "learning_rate": 4.137751418927596e-06,
+      "loss": 0.6222,
+      "step": 6233
+    },
+    {
+      "epoch": 1.6549847338377806,
+      "grad_norm": 0.42068599178016536,
+      "learning_rate": 4.137487622261254e-06,
+      "loss": 0.5974,
+      "step": 6234
+    },
+    {
+      "epoch": 1.6552502323111642,
+      "grad_norm": 0.4098890913665752,
+      "learning_rate": 4.137223793659894e-06,
+      "loss": 0.594,
+      "step": 6235
+    },
+    {
+      "epoch": 1.6555157307845478,
+      "grad_norm": 0.42397914087060007,
+      "learning_rate": 4.1369599331286615e-06,
+      "loss": 0.5374,
+      "step": 6236
+    },
+    {
+      "epoch": 1.655781229257932,
+      "grad_norm": 0.41304192083496677,
+      "learning_rate": 4.136696040672702e-06,
+      "loss": 0.6274,
+      "step": 6237
+    },
+    {
+      "epoch": 1.6560467277313156,
+      "grad_norm": 0.41278506832165685,
+      "learning_rate": 4.136432116297164e-06,
+      "loss": 0.578,
+      "step": 6238
+    },
+    {
+      "epoch": 1.6563122262046992,
+      "grad_norm": 0.4094875937980845,
+      "learning_rate": 4.136168160007191e-06,
+      "loss": 0.6053,
+      "step": 6239
+    },
+    {
+      "epoch": 1.656577724678083,
+      "grad_norm": 0.4035127617560246,
+      "learning_rate": 4.135904171807934e-06,
+      "loss": 0.5596,
+      "step": 6240
+    },
+    {
+      "epoch": 1.656843223151467,
+      "grad_norm": 0.4204748616645181,
+      "learning_rate": 4.1356401517045395e-06,
+      "loss": 0.5335,
+      "step": 6241
+    },
+    {
+      "epoch": 1.6571087216248506,
+      "grad_norm": 0.40997332975393685,
+      "learning_rate": 4.135376099702158e-06,
+      "loss": 0.544,
+      "step": 6242
+    },
+    {
+      "epoch": 1.6573742200982344,
+      "grad_norm": 0.3992096663527178,
+      "learning_rate": 4.135112015805938e-06,
+      "loss": 0.5609,
+      "step": 6243
+    },
+    {
+      "epoch": 1.6576397185716183,
+      "grad_norm": 0.407378546327697,
+      "learning_rate": 4.1348479000210305e-06,
+      "loss": 0.5875,
+      "step": 6244
+    },
+    {
+      "epoch": 1.657905217045002,
+      "grad_norm": 0.4154709615900247,
+      "learning_rate": 4.1345837523525865e-06,
+      "loss": 0.5688,
+      "step": 6245
+    },
+    {
+      "epoch": 1.6581707155183858,
+      "grad_norm": 0.40770356163791593,
+      "learning_rate": 4.134319572805756e-06,
+      "loss": 0.5302,
+      "step": 6246
+    },
+    {
+      "epoch": 1.6584362139917697,
+      "grad_norm": 0.3880617767005316,
+      "learning_rate": 4.134055361385694e-06,
+      "loss": 0.5859,
+      "step": 6247
+    },
+    {
+      "epoch": 1.6587017124651533,
+      "grad_norm": 0.4196745597244992,
+      "learning_rate": 4.1337911180975506e-06,
+      "loss": 0.6098,
+      "step": 6248
+    },
+    {
+      "epoch": 1.658967210938537,
+      "grad_norm": 0.4129199285056074,
+      "learning_rate": 4.133526842946481e-06,
+      "loss": 0.572,
+      "step": 6249
+    },
+    {
+      "epoch": 1.659232709411921,
+      "grad_norm": 0.41145264862768005,
+      "learning_rate": 4.133262535937637e-06,
+      "loss": 0.5706,
+      "step": 6250
+    },
+    {
+      "epoch": 1.6594982078853047,
+      "grad_norm": 0.4148045692625393,
+      "learning_rate": 4.132998197076175e-06,
+      "loss": 0.5707,
+      "step": 6251
+    },
+    {
+      "epoch": 1.6597637063586883,
+      "grad_norm": 0.3957170698282108,
+      "learning_rate": 4.13273382636725e-06,
+      "loss": 0.5821,
+      "step": 6252
+    },
+    {
+      "epoch": 1.6600292048320722,
+      "grad_norm": 0.4118281769956654,
+      "learning_rate": 4.132469423816019e-06,
+      "loss": 0.5864,
+      "step": 6253
+    },
+    {
+      "epoch": 1.660294703305456,
+      "grad_norm": 0.4226261841297954,
+      "learning_rate": 4.132204989427636e-06,
+      "loss": 0.607,
+      "step": 6254
+    },
+    {
+      "epoch": 1.6605602017788397,
+      "grad_norm": 0.425985370340082,
+      "learning_rate": 4.1319405232072596e-06,
+      "loss": 0.5963,
+      "step": 6255
+    },
+    {
+      "epoch": 1.6608257002522235,
+      "grad_norm": 0.3996984870942285,
+      "learning_rate": 4.131676025160047e-06,
+      "loss": 0.5826,
+      "step": 6256
+    },
+    {
+      "epoch": 1.6610911987256074,
+      "grad_norm": 0.4063701415508513,
+      "learning_rate": 4.131411495291158e-06,
+      "loss": 0.6088,
+      "step": 6257
+    },
+    {
+      "epoch": 1.661356697198991,
+      "grad_norm": 0.406081896704152,
+      "learning_rate": 4.131146933605749e-06,
+      "loss": 0.5785,
+      "step": 6258
+    },
+    {
+      "epoch": 1.661622195672375,
+      "grad_norm": 0.4230551885595814,
+      "learning_rate": 4.130882340108981e-06,
+      "loss": 0.556,
+      "step": 6259
+    },
+    {
+      "epoch": 1.6618876941457588,
+      "grad_norm": 0.4149994063430953,
+      "learning_rate": 4.130617714806016e-06,
+      "loss": 0.587,
+      "step": 6260
+    },
+    {
+      "epoch": 1.6621531926191424,
+      "grad_norm": 0.4257823328443519,
+      "learning_rate": 4.130353057702012e-06,
+      "loss": 0.6051,
+      "step": 6261
+    },
+    {
+      "epoch": 1.6624186910925263,
+      "grad_norm": 0.40865431395602214,
+      "learning_rate": 4.130088368802132e-06,
+      "loss": 0.5961,
+      "step": 6262
+    },
+    {
+      "epoch": 1.6626841895659101,
+      "grad_norm": 0.42533649853128425,
+      "learning_rate": 4.129823648111537e-06,
+      "loss": 0.6312,
+      "step": 6263
+    },
+    {
+      "epoch": 1.6629496880392938,
+      "grad_norm": 0.39085554844775955,
+      "learning_rate": 4.129558895635391e-06,
+      "loss": 0.5607,
+      "step": 6264
+    },
+    {
+      "epoch": 1.6632151865126774,
+      "grad_norm": 0.4207350142253288,
+      "learning_rate": 4.129294111378857e-06,
+      "loss": 0.6031,
+      "step": 6265
+    },
+    {
+      "epoch": 1.6634806849860613,
+      "grad_norm": 0.39123248553015144,
+      "learning_rate": 4.129029295347097e-06,
+      "loss": 0.5895,
+      "step": 6266
+    },
+    {
+      "epoch": 1.6637461834594451,
+      "grad_norm": 0.4148025621611677,
+      "learning_rate": 4.128764447545278e-06,
+      "loss": 0.5855,
+      "step": 6267
+    },
+    {
+      "epoch": 1.6640116819328288,
+      "grad_norm": 0.4021487051768757,
+      "learning_rate": 4.1284995679785655e-06,
+      "loss": 0.6107,
+      "step": 6268
+    },
+    {
+      "epoch": 1.6642771804062126,
+      "grad_norm": 0.41992299218051976,
+      "learning_rate": 4.128234656652122e-06,
+      "loss": 0.6095,
+      "step": 6269
+    },
+    {
+      "epoch": 1.6645426788795965,
+      "grad_norm": 0.4116384967035457,
+      "learning_rate": 4.127969713571119e-06,
+      "loss": 0.5836,
+      "step": 6270
+    },
+    {
+      "epoch": 1.6648081773529801,
+      "grad_norm": 0.4119548148414317,
+      "learning_rate": 4.127704738740719e-06,
+      "loss": 0.5876,
+      "step": 6271
+    },
+    {
+      "epoch": 1.665073675826364,
+      "grad_norm": 0.398825976868628,
+      "learning_rate": 4.127439732166091e-06,
+      "loss": 0.5589,
+      "step": 6272
+    },
+    {
+      "epoch": 1.6653391742997479,
+      "grad_norm": 0.4090683604248852,
+      "learning_rate": 4.127174693852404e-06,
+      "loss": 0.6122,
+      "step": 6273
+    },
+    {
+      "epoch": 1.6656046727731315,
+      "grad_norm": 0.405529061429767,
+      "learning_rate": 4.126909623804825e-06,
+      "loss": 0.6155,
+      "step": 6274
+    },
+    {
+      "epoch": 1.6658701712465154,
+      "grad_norm": 0.4126587093699257,
+      "learning_rate": 4.126644522028526e-06,
+      "loss": 0.5725,
+      "step": 6275
+    },
+    {
+      "epoch": 1.6661356697198992,
+      "grad_norm": 0.4124320374860765,
+      "learning_rate": 4.1263793885286775e-06,
+      "loss": 0.5382,
+      "step": 6276
+    },
+    {
+      "epoch": 1.6664011681932829,
+      "grad_norm": 0.3889636497590304,
+      "learning_rate": 4.1261142233104475e-06,
+      "loss": 0.564,
+      "step": 6277
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.42026093390239305,
+      "learning_rate": 4.125849026379009e-06,
+      "loss": 0.6312,
+      "step": 6278
+    },
+    {
+      "epoch": 1.6669321651400506,
+      "grad_norm": 0.4236148123260068,
+      "learning_rate": 4.125583797739534e-06,
+      "loss": 0.5489,
+      "step": 6279
+    },
+    {
+      "epoch": 1.6671976636134342,
+      "grad_norm": 0.4012570714568701,
+      "learning_rate": 4.125318537397195e-06,
+      "loss": 0.5674,
+      "step": 6280
+    },
+    {
+      "epoch": 1.6674631620868179,
+      "grad_norm": 0.4078922674826175,
+      "learning_rate": 4.125053245357164e-06,
+      "loss": 0.5979,
+      "step": 6281
+    },
+    {
+      "epoch": 1.6677286605602017,
+      "grad_norm": 0.41892003489855617,
+      "learning_rate": 4.124787921624617e-06,
+      "loss": 0.5884,
+      "step": 6282
+    },
+    {
+      "epoch": 1.6679941590335856,
+      "grad_norm": 0.40938248580799247,
+      "learning_rate": 4.124522566204727e-06,
+      "loss": 0.5889,
+      "step": 6283
+    },
+    {
+      "epoch": 1.6682596575069693,
+      "grad_norm": 0.42011994660105495,
+      "learning_rate": 4.124257179102668e-06,
+      "loss": 0.6041,
+      "step": 6284
+    },
+    {
+      "epoch": 1.6685251559803531,
+      "grad_norm": 0.4191644433253345,
+      "learning_rate": 4.1239917603236195e-06,
+      "loss": 0.5809,
+      "step": 6285
+    },
+    {
+      "epoch": 1.668790654453737,
+      "grad_norm": 0.40211573031454656,
+      "learning_rate": 4.123726309872754e-06,
+      "loss": 0.6034,
+      "step": 6286
+    },
+    {
+      "epoch": 1.6690561529271206,
+      "grad_norm": 0.41975737406106656,
+      "learning_rate": 4.12346082775525e-06,
+      "loss": 0.5695,
+      "step": 6287
+    },
+    {
+      "epoch": 1.6693216514005045,
+      "grad_norm": 0.39337615849979457,
+      "learning_rate": 4.123195313976286e-06,
+      "loss": 0.5824,
+      "step": 6288
+    },
+    {
+      "epoch": 1.6695871498738883,
+      "grad_norm": 0.4303546203791357,
+      "learning_rate": 4.122929768541038e-06,
+      "loss": 0.5812,
+      "step": 6289
+    },
+    {
+      "epoch": 1.669852648347272,
+      "grad_norm": 0.4164017797520493,
+      "learning_rate": 4.122664191454686e-06,
+      "loss": 0.5779,
+      "step": 6290
+    },
+    {
+      "epoch": 1.6701181468206556,
+      "grad_norm": 0.4097090631863583,
+      "learning_rate": 4.12239858272241e-06,
+      "loss": 0.5981,
+      "step": 6291
+    },
+    {
+      "epoch": 1.6703836452940397,
+      "grad_norm": 0.416787858211982,
+      "learning_rate": 4.1221329423493885e-06,
+      "loss": 0.6036,
+      "step": 6292
+    },
+    {
+      "epoch": 1.6706491437674234,
+      "grad_norm": 0.4292695720492115,
+      "learning_rate": 4.121867270340803e-06,
+      "loss": 0.5547,
+      "step": 6293
+    },
+    {
+      "epoch": 1.670914642240807,
+      "grad_norm": 0.41061784102903787,
+      "learning_rate": 4.121601566701836e-06,
+      "loss": 0.5737,
+      "step": 6294
+    },
+    {
+      "epoch": 1.6711801407141909,
+      "grad_norm": 0.41479910673626835,
+      "learning_rate": 4.121335831437667e-06,
+      "loss": 0.6225,
+      "step": 6295
+    },
+    {
+      "epoch": 1.6714456391875747,
+      "grad_norm": 0.39647962614135507,
+      "learning_rate": 4.121070064553479e-06,
+      "loss": 0.5911,
+      "step": 6296
+    },
+    {
+      "epoch": 1.6717111376609584,
+      "grad_norm": 0.4036427208188757,
+      "learning_rate": 4.120804266054457e-06,
+      "loss": 0.579,
+      "step": 6297
+    },
+    {
+      "epoch": 1.6719766361343422,
+      "grad_norm": 0.4056796986607875,
+      "learning_rate": 4.120538435945782e-06,
+      "loss": 0.5496,
+      "step": 6298
+    },
+    {
+      "epoch": 1.672242134607726,
+      "grad_norm": 0.4080395713944233,
+      "learning_rate": 4.120272574232641e-06,
+      "loss": 0.5717,
+      "step": 6299
+    },
+    {
+      "epoch": 1.6725076330811097,
+      "grad_norm": 0.4015556865683174,
+      "learning_rate": 4.1200066809202175e-06,
+      "loss": 0.5959,
+      "step": 6300
+    },
+    {
+      "epoch": 1.6727731315544936,
+      "grad_norm": 0.41772861638897374,
+      "learning_rate": 4.119740756013697e-06,
+      "loss": 0.6041,
+      "step": 6301
+    },
+    {
+      "epoch": 1.6730386300278774,
+      "grad_norm": 0.4121327627603656,
+      "learning_rate": 4.119474799518266e-06,
+      "loss": 0.5957,
+      "step": 6302
+    },
+    {
+      "epoch": 1.673304128501261,
+      "grad_norm": 0.41607194339937,
+      "learning_rate": 4.119208811439111e-06,
+      "loss": 0.5918,
+      "step": 6303
+    },
+    {
+      "epoch": 1.6735696269746447,
+      "grad_norm": 0.41607659660043184,
+      "learning_rate": 4.11894279178142e-06,
+      "loss": 0.5651,
+      "step": 6304
+    },
+    {
+      "epoch": 1.6738351254480288,
+      "grad_norm": 0.40703456582321434,
+      "learning_rate": 4.11867674055038e-06,
+      "loss": 0.5779,
+      "step": 6305
+    },
+    {
+      "epoch": 1.6741006239214125,
+      "grad_norm": 0.39641516024729906,
+      "learning_rate": 4.11841065775118e-06,
+      "loss": 0.6025,
+      "step": 6306
+    },
+    {
+      "epoch": 1.674366122394796,
+      "grad_norm": 0.41827020132395565,
+      "learning_rate": 4.118144543389011e-06,
+      "loss": 0.585,
+      "step": 6307
+    },
+    {
+      "epoch": 1.67463162086818,
+      "grad_norm": 0.4082423123129381,
+      "learning_rate": 4.117878397469062e-06,
+      "loss": 0.5751,
+      "step": 6308
+    },
+    {
+      "epoch": 1.6748971193415638,
+      "grad_norm": 0.3940808378832142,
+      "learning_rate": 4.117612219996522e-06,
+      "loss": 0.5856,
+      "step": 6309
+    },
+    {
+      "epoch": 1.6751626178149475,
+      "grad_norm": 0.408874741784608,
+      "learning_rate": 4.117346010976583e-06,
+      "loss": 0.6043,
+      "step": 6310
+    },
+    {
+      "epoch": 1.6754281162883313,
+      "grad_norm": 0.40966683365669165,
+      "learning_rate": 4.117079770414436e-06,
+      "loss": 0.6044,
+      "step": 6311
+    },
+    {
+      "epoch": 1.6756936147617152,
+      "grad_norm": 0.4305904864209752,
+      "learning_rate": 4.1168134983152756e-06,
+      "loss": 0.5756,
+      "step": 6312
+    },
+    {
+      "epoch": 1.6759591132350988,
+      "grad_norm": 0.40715643782085553,
+      "learning_rate": 4.116547194684293e-06,
+      "loss": 0.5931,
+      "step": 6313
+    },
+    {
+      "epoch": 1.6762246117084827,
+      "grad_norm": 0.4096397716811137,
+      "learning_rate": 4.1162808595266825e-06,
+      "loss": 0.574,
+      "step": 6314
+    },
+    {
+      "epoch": 1.6764901101818666,
+      "grad_norm": 0.4025989393357566,
+      "learning_rate": 4.1160144928476366e-06,
+      "loss": 0.5715,
+      "step": 6315
+    },
+    {
+      "epoch": 1.6767556086552502,
+      "grad_norm": 0.42442510676230444,
+      "learning_rate": 4.115748094652352e-06,
+      "loss": 0.6085,
+      "step": 6316
+    },
+    {
+      "epoch": 1.677021107128634,
+      "grad_norm": 0.4121507733462747,
+      "learning_rate": 4.115481664946024e-06,
+      "loss": 0.5848,
+      "step": 6317
+    },
+    {
+      "epoch": 1.677286605602018,
+      "grad_norm": 0.3941377941422974,
+      "learning_rate": 4.115215203733848e-06,
+      "loss": 0.5657,
+      "step": 6318
+    },
+    {
+      "epoch": 1.6775521040754016,
+      "grad_norm": 0.41219243346894474,
+      "learning_rate": 4.11494871102102e-06,
+      "loss": 0.5647,
+      "step": 6319
+    },
+    {
+      "epoch": 1.6778176025487852,
+      "grad_norm": 0.43268918353710634,
+      "learning_rate": 4.114682186812739e-06,
+      "loss": 0.5848,
+      "step": 6320
+    },
+    {
+      "epoch": 1.6780831010221693,
+      "grad_norm": 0.40701781828082384,
+      "learning_rate": 4.1144156311142025e-06,
+      "loss": 0.5764,
+      "step": 6321
+    },
+    {
+      "epoch": 1.678348599495553,
+      "grad_norm": 0.40048312256767066,
+      "learning_rate": 4.114149043930607e-06,
+      "loss": 0.6017,
+      "step": 6322
+    },
+    {
+      "epoch": 1.6786140979689366,
+      "grad_norm": 0.4255823562558479,
+      "learning_rate": 4.113882425267154e-06,
+      "loss": 0.5919,
+      "step": 6323
+    },
+    {
+      "epoch": 1.6788795964423204,
+      "grad_norm": 0.41201070477762003,
+      "learning_rate": 4.113615775129042e-06,
+      "loss": 0.5839,
+      "step": 6324
+    },
+    {
+      "epoch": 1.6791450949157043,
+      "grad_norm": 0.4111270706763679,
+      "learning_rate": 4.113349093521472e-06,
+      "loss": 0.5922,
+      "step": 6325
+    },
+    {
+      "epoch": 1.679410593389088,
+      "grad_norm": 0.41365891011181766,
+      "learning_rate": 4.113082380449644e-06,
+      "loss": 0.613,
+      "step": 6326
+    },
+    {
+      "epoch": 1.6796760918624718,
+      "grad_norm": 0.4116880308580051,
+      "learning_rate": 4.11281563591876e-06,
+      "loss": 0.5837,
+      "step": 6327
+    },
+    {
+      "epoch": 1.6799415903358557,
+      "grad_norm": 0.42748097326883955,
+      "learning_rate": 4.112548859934021e-06,
+      "loss": 0.5838,
+      "step": 6328
+    },
+    {
+      "epoch": 1.6802070888092393,
+      "grad_norm": 0.4114554528539254,
+      "learning_rate": 4.112282052500632e-06,
+      "loss": 0.5831,
+      "step": 6329
+    },
+    {
+      "epoch": 1.6804725872826232,
+      "grad_norm": 0.41978008815731305,
+      "learning_rate": 4.112015213623796e-06,
+      "loss": 0.5986,
+      "step": 6330
+    },
+    {
+      "epoch": 1.680738085756007,
+      "grad_norm": 0.4155852396845013,
+      "learning_rate": 4.111748343308717e-06,
+      "loss": 0.6073,
+      "step": 6331
+    },
+    {
+      "epoch": 1.6810035842293907,
+      "grad_norm": 0.40970980697927045,
+      "learning_rate": 4.111481441560598e-06,
+      "loss": 0.5686,
+      "step": 6332
+    },
+    {
+      "epoch": 1.6812690827027743,
+      "grad_norm": 0.41972736975756464,
+      "learning_rate": 4.1112145083846455e-06,
+      "loss": 0.5641,
+      "step": 6333
+    },
+    {
+      "epoch": 1.6815345811761584,
+      "grad_norm": 0.4028261859791443,
+      "learning_rate": 4.110947543786065e-06,
+      "loss": 0.5802,
+      "step": 6334
+    },
+    {
+      "epoch": 1.681800079649542,
+      "grad_norm": 0.41341272091908965,
+      "learning_rate": 4.110680547770063e-06,
+      "loss": 0.5456,
+      "step": 6335
+    },
+    {
+      "epoch": 1.6820655781229257,
+      "grad_norm": 0.3977026297832417,
+      "learning_rate": 4.110413520341847e-06,
+      "loss": 0.5823,
+      "step": 6336
+    },
+    {
+      "epoch": 1.6823310765963095,
+      "grad_norm": 0.39243486546212913,
+      "learning_rate": 4.1101464615066245e-06,
+      "loss": 0.5759,
+      "step": 6337
+    },
+    {
+      "epoch": 1.6825965750696934,
+      "grad_norm": 0.43224512890557176,
+      "learning_rate": 4.109879371269604e-06,
+      "loss": 0.5964,
+      "step": 6338
+    },
+    {
+      "epoch": 1.682862073543077,
+      "grad_norm": 0.41444387962089035,
+      "learning_rate": 4.109612249635993e-06,
+      "loss": 0.5783,
+      "step": 6339
+    },
+    {
+      "epoch": 1.683127572016461,
+      "grad_norm": 0.3935197047036064,
+      "learning_rate": 4.109345096611004e-06,
+      "loss": 0.6114,
+      "step": 6340
+    },
+    {
+      "epoch": 1.6833930704898448,
+      "grad_norm": 0.42624563039776986,
+      "learning_rate": 4.1090779121998435e-06,
+      "loss": 0.5435,
+      "step": 6341
+    },
+    {
+      "epoch": 1.6836585689632284,
+      "grad_norm": 0.4230824834732495,
+      "learning_rate": 4.108810696407725e-06,
+      "loss": 0.5779,
+      "step": 6342
+    },
+    {
+      "epoch": 1.6839240674366123,
+      "grad_norm": 0.4054209585045214,
+      "learning_rate": 4.108543449239858e-06,
+      "loss": 0.5661,
+      "step": 6343
+    },
+    {
+      "epoch": 1.6841895659099961,
+      "grad_norm": 0.44395284151553416,
+      "learning_rate": 4.108276170701456e-06,
+      "loss": 0.5984,
+      "step": 6344
+    },
+    {
+      "epoch": 1.6844550643833798,
+      "grad_norm": 0.41026685187487105,
+      "learning_rate": 4.108008860797731e-06,
+      "loss": 0.5893,
+      "step": 6345
+    },
+    {
+      "epoch": 1.6847205628567634,
+      "grad_norm": 0.43986690557286967,
+      "learning_rate": 4.107741519533895e-06,
+      "loss": 0.5737,
+      "step": 6346
+    },
+    {
+      "epoch": 1.6849860613301475,
+      "grad_norm": 0.3969660905177006,
+      "learning_rate": 4.107474146915164e-06,
+      "loss": 0.5712,
+      "step": 6347
+    },
+    {
+      "epoch": 1.6852515598035311,
+      "grad_norm": 0.41012061916371767,
+      "learning_rate": 4.107206742946751e-06,
+      "loss": 0.5807,
+      "step": 6348
+    },
+    {
+      "epoch": 1.6855170582769148,
+      "grad_norm": 0.41499185474366906,
+      "learning_rate": 4.106939307633871e-06,
+      "loss": 0.5554,
+      "step": 6349
+    },
+    {
+      "epoch": 1.6857825567502986,
+      "grad_norm": 0.4450379986225015,
+      "learning_rate": 4.106671840981741e-06,
+      "loss": 0.5689,
+      "step": 6350
+    },
+    {
+      "epoch": 1.6860480552236825,
+      "grad_norm": 0.426476334776983,
+      "learning_rate": 4.1064043429955755e-06,
+      "loss": 0.5829,
+      "step": 6351
+    },
+    {
+      "epoch": 1.6863135536970661,
+      "grad_norm": 0.42478931827500366,
+      "learning_rate": 4.106136813680593e-06,
+      "loss": 0.5912,
+      "step": 6352
+    },
+    {
+      "epoch": 1.68657905217045,
+      "grad_norm": 0.4096086951383353,
+      "learning_rate": 4.1058692530420096e-06,
+      "loss": 0.598,
+      "step": 6353
+    },
+    {
+      "epoch": 1.6868445506438339,
+      "grad_norm": 0.4373314025858963,
+      "learning_rate": 4.105601661085044e-06,
+      "loss": 0.5864,
+      "step": 6354
+    },
+    {
+      "epoch": 1.6871100491172175,
+      "grad_norm": 0.40282744081911964,
+      "learning_rate": 4.105334037814915e-06,
+      "loss": 0.5894,
+      "step": 6355
+    },
+    {
+      "epoch": 1.6873755475906014,
+      "grad_norm": 0.42586938814405817,
+      "learning_rate": 4.105066383236841e-06,
+      "loss": 0.5836,
+      "step": 6356
+    },
+    {
+      "epoch": 1.6876410460639852,
+      "grad_norm": 0.4302025906478626,
+      "learning_rate": 4.104798697356044e-06,
+      "loss": 0.6025,
+      "step": 6357
+    },
+    {
+      "epoch": 1.6879065445373689,
+      "grad_norm": 0.4074057891769,
+      "learning_rate": 4.104530980177742e-06,
+      "loss": 0.5973,
+      "step": 6358
+    },
+    {
+      "epoch": 1.6881720430107527,
+      "grad_norm": 0.4229883427210265,
+      "learning_rate": 4.104263231707157e-06,
+      "loss": 0.6113,
+      "step": 6359
+    },
+    {
+      "epoch": 1.6884375414841366,
+      "grad_norm": 0.4287571724094451,
+      "learning_rate": 4.103995451949512e-06,
+      "loss": 0.5975,
+      "step": 6360
+    },
+    {
+      "epoch": 1.6887030399575202,
+      "grad_norm": 0.42496527565666037,
+      "learning_rate": 4.103727640910028e-06,
+      "loss": 0.5863,
+      "step": 6361
+    },
+    {
+      "epoch": 1.6889685384309039,
+      "grad_norm": 0.4016826548906657,
+      "learning_rate": 4.103459798593928e-06,
+      "loss": 0.5978,
+      "step": 6362
+    },
+    {
+      "epoch": 1.6892340369042878,
+      "grad_norm": 0.40552571059681175,
+      "learning_rate": 4.103191925006438e-06,
+      "loss": 0.5661,
+      "step": 6363
+    },
+    {
+      "epoch": 1.6894995353776716,
+      "grad_norm": 0.4165922795596941,
+      "learning_rate": 4.102924020152777e-06,
+      "loss": 0.6139,
+      "step": 6364
+    },
+    {
+      "epoch": 1.6897650338510553,
+      "grad_norm": 0.42427768752913114,
+      "learning_rate": 4.102656084038176e-06,
+      "loss": 0.6129,
+      "step": 6365
+    },
+    {
+      "epoch": 1.6900305323244391,
+      "grad_norm": 0.40137273718291483,
+      "learning_rate": 4.102388116667856e-06,
+      "loss": 0.6225,
+      "step": 6366
+    },
+    {
+      "epoch": 1.690296030797823,
+      "grad_norm": 0.42317600116874127,
+      "learning_rate": 4.102120118047044e-06,
+      "loss": 0.5697,
+      "step": 6367
+    },
+    {
+      "epoch": 1.6905615292712066,
+      "grad_norm": 0.40658546027517756,
+      "learning_rate": 4.101852088180967e-06,
+      "loss": 0.6202,
+      "step": 6368
+    },
+    {
+      "epoch": 1.6908270277445905,
+      "grad_norm": 0.42287643576618905,
+      "learning_rate": 4.101584027074854e-06,
+      "loss": 0.6062,
+      "step": 6369
+    },
+    {
+      "epoch": 1.6910925262179743,
+      "grad_norm": 0.410552811964971,
+      "learning_rate": 4.101315934733929e-06,
+      "loss": 0.6238,
+      "step": 6370
+    },
+    {
+      "epoch": 1.691358024691358,
+      "grad_norm": 0.4278473776348954,
+      "learning_rate": 4.101047811163424e-06,
+      "loss": 0.5711,
+      "step": 6371
+    },
+    {
+      "epoch": 1.6916235231647418,
+      "grad_norm": 0.42162881407903335,
+      "learning_rate": 4.100779656368566e-06,
+      "loss": 0.6005,
+      "step": 6372
+    },
+    {
+      "epoch": 1.6918890216381257,
+      "grad_norm": 0.42086775835127227,
+      "learning_rate": 4.100511470354585e-06,
+      "loss": 0.5609,
+      "step": 6373
+    },
+    {
+      "epoch": 1.6921545201115094,
+      "grad_norm": 0.4168206115018423,
+      "learning_rate": 4.100243253126712e-06,
+      "loss": 0.5981,
+      "step": 6374
+    },
+    {
+      "epoch": 1.692420018584893,
+      "grad_norm": 0.43982691121191747,
+      "learning_rate": 4.099975004690178e-06,
+      "loss": 0.5925,
+      "step": 6375
+    },
+    {
+      "epoch": 1.692685517058277,
+      "grad_norm": 0.4070570648371935,
+      "learning_rate": 4.0997067250502134e-06,
+      "loss": 0.5768,
+      "step": 6376
+    },
+    {
+      "epoch": 1.6929510155316607,
+      "grad_norm": 0.41347871930142327,
+      "learning_rate": 4.099438414212051e-06,
+      "loss": 0.5498,
+      "step": 6377
+    },
+    {
+      "epoch": 1.6932165140050444,
+      "grad_norm": 0.41634321760258314,
+      "learning_rate": 4.099170072180923e-06,
+      "loss": 0.57,
+      "step": 6378
+    },
+    {
+      "epoch": 1.6934820124784282,
+      "grad_norm": 0.43542058241940884,
+      "learning_rate": 4.098901698962063e-06,
+      "loss": 0.5903,
+      "step": 6379
+    },
+    {
+      "epoch": 1.693747510951812,
+      "grad_norm": 0.4229693905665481,
+      "learning_rate": 4.098633294560707e-06,
+      "loss": 0.5431,
+      "step": 6380
+    },
+    {
+      "epoch": 1.6940130094251957,
+      "grad_norm": 0.4096668223370534,
+      "learning_rate": 4.098364858982085e-06,
+      "loss": 0.5871,
+      "step": 6381
+    },
+    {
+      "epoch": 1.6942785078985796,
+      "grad_norm": 0.41890765798353896,
+      "learning_rate": 4.098096392231437e-06,
+      "loss": 0.5661,
+      "step": 6382
+    },
+    {
+      "epoch": 1.6945440063719635,
+      "grad_norm": 0.41402569766025216,
+      "learning_rate": 4.097827894313995e-06,
+      "loss": 0.571,
+      "step": 6383
+    },
+    {
+      "epoch": 1.694809504845347,
+      "grad_norm": 0.42179137110450254,
+      "learning_rate": 4.097559365234998e-06,
+      "loss": 0.5534,
+      "step": 6384
+    },
+    {
+      "epoch": 1.695075003318731,
+      "grad_norm": 0.404547654102909,
+      "learning_rate": 4.097290804999681e-06,
+      "loss": 0.5778,
+      "step": 6385
+    },
+    {
+      "epoch": 1.6953405017921148,
+      "grad_norm": 0.4149106010204333,
+      "learning_rate": 4.097022213613283e-06,
+      "loss": 0.5862,
+      "step": 6386
+    },
+    {
+      "epoch": 1.6956060002654985,
+      "grad_norm": 0.415602586451017,
+      "learning_rate": 4.096753591081041e-06,
+      "loss": 0.5767,
+      "step": 6387
+    },
+    {
+      "epoch": 1.695871498738882,
+      "grad_norm": 0.40146208630611885,
+      "learning_rate": 4.096484937408195e-06,
+      "loss": 0.5506,
+      "step": 6388
+    },
+    {
+      "epoch": 1.6961369972122662,
+      "grad_norm": 0.43260958103687774,
+      "learning_rate": 4.096216252599984e-06,
+      "loss": 0.5553,
+      "step": 6389
+    },
+    {
+      "epoch": 1.6964024956856498,
+      "grad_norm": 0.40107115383249686,
+      "learning_rate": 4.095947536661648e-06,
+      "loss": 0.5973,
+      "step": 6390
+    },
+    {
+      "epoch": 1.6966679941590335,
+      "grad_norm": 0.40664575830641714,
+      "learning_rate": 4.0956787895984276e-06,
+      "loss": 0.5655,
+      "step": 6391
+    },
+    {
+      "epoch": 1.6969334926324173,
+      "grad_norm": 0.42616997752416796,
+      "learning_rate": 4.095410011415563e-06,
+      "loss": 0.5847,
+      "step": 6392
+    },
+    {
+      "epoch": 1.6971989911058012,
+      "grad_norm": 0.4325767776143576,
+      "learning_rate": 4.095141202118297e-06,
+      "loss": 0.5978,
+      "step": 6393
+    },
+    {
+      "epoch": 1.6974644895791848,
+      "grad_norm": 0.41061979151503814,
+      "learning_rate": 4.094872361711872e-06,
+      "loss": 0.5908,
+      "step": 6394
+    },
+    {
+      "epoch": 1.6977299880525687,
+      "grad_norm": 0.41320415833809454,
+      "learning_rate": 4.094603490201532e-06,
+      "loss": 0.6026,
+      "step": 6395
+    },
+    {
+      "epoch": 1.6979954865259526,
+      "grad_norm": 0.4252031107905295,
+      "learning_rate": 4.094334587592519e-06,
+      "loss": 0.5992,
+      "step": 6396
+    },
+    {
+      "epoch": 1.6982609849993362,
+      "grad_norm": 0.4171249749714777,
+      "learning_rate": 4.0940656538900785e-06,
+      "loss": 0.5963,
+      "step": 6397
+    },
+    {
+      "epoch": 1.69852648347272,
+      "grad_norm": 0.42979606864536424,
+      "learning_rate": 4.093796689099453e-06,
+      "loss": 0.5785,
+      "step": 6398
+    },
+    {
+      "epoch": 1.698791981946104,
+      "grad_norm": 0.3998948301091688,
+      "learning_rate": 4.093527693225891e-06,
+      "loss": 0.6166,
+      "step": 6399
+    },
+    {
+      "epoch": 1.6990574804194876,
+      "grad_norm": 0.3999341502225336,
+      "learning_rate": 4.093258666274639e-06,
+      "loss": 0.5772,
+      "step": 6400
+    },
+    {
+      "epoch": 1.6993229788928712,
+      "grad_norm": 0.42143661748813177,
+      "learning_rate": 4.0929896082509395e-06,
+      "loss": 0.6068,
+      "step": 6401
+    },
+    {
+      "epoch": 1.6995884773662553,
+      "grad_norm": 0.40074130024828597,
+      "learning_rate": 4.092720519160044e-06,
+      "loss": 0.5657,
+      "step": 6402
+    },
+    {
+      "epoch": 1.699853975839639,
+      "grad_norm": 0.43700781928529603,
+      "learning_rate": 4.092451399007198e-06,
+      "loss": 0.562,
+      "step": 6403
+    },
+    {
+      "epoch": 1.7001194743130226,
+      "grad_norm": 0.4050483358338015,
+      "learning_rate": 4.09218224779765e-06,
+      "loss": 0.5783,
+      "step": 6404
+    },
+    {
+      "epoch": 1.7003849727864064,
+      "grad_norm": 0.4125435779686136,
+      "learning_rate": 4.0919130655366504e-06,
+      "loss": 0.5425,
+      "step": 6405
+    },
+    {
+      "epoch": 1.7006504712597903,
+      "grad_norm": 0.40616842361924405,
+      "learning_rate": 4.091643852229449e-06,
+      "loss": 0.5794,
+      "step": 6406
+    },
+    {
+      "epoch": 1.700915969733174,
+      "grad_norm": 0.4062933433129822,
+      "learning_rate": 4.091374607881294e-06,
+      "loss": 0.5627,
+      "step": 6407
+    },
+    {
+      "epoch": 1.7011814682065578,
+      "grad_norm": 0.4192116361728902,
+      "learning_rate": 4.0911053324974395e-06,
+      "loss": 0.6051,
+      "step": 6408
+    },
+    {
+      "epoch": 1.7014469666799417,
+      "grad_norm": 0.4137945760551017,
+      "learning_rate": 4.0908360260831345e-06,
+      "loss": 0.6003,
+      "step": 6409
+    },
+    {
+      "epoch": 1.7017124651533253,
+      "grad_norm": 0.41050716920187064,
+      "learning_rate": 4.090566688643632e-06,
+      "loss": 0.6169,
+      "step": 6410
+    },
+    {
+      "epoch": 1.7019779636267092,
+      "grad_norm": 0.42012432549972434,
+      "learning_rate": 4.090297320184185e-06,
+      "loss": 0.5621,
+      "step": 6411
+    },
+    {
+      "epoch": 1.702243462100093,
+      "grad_norm": 0.40871221296596166,
+      "learning_rate": 4.090027920710046e-06,
+      "loss": 0.5968,
+      "step": 6412
+    },
+    {
+      "epoch": 1.7025089605734767,
+      "grad_norm": 0.4090334544123141,
+      "learning_rate": 4.08975849022647e-06,
+      "loss": 0.5975,
+      "step": 6413
+    },
+    {
+      "epoch": 1.7027744590468605,
+      "grad_norm": 0.41812633835746,
+      "learning_rate": 4.089489028738711e-06,
+      "loss": 0.5654,
+      "step": 6414
+    },
+    {
+      "epoch": 1.7030399575202444,
+      "grad_norm": 0.4116453932312294,
+      "learning_rate": 4.089219536252024e-06,
+      "loss": 0.5729,
+      "step": 6415
+    },
+    {
+      "epoch": 1.703305455993628,
+      "grad_norm": 0.41001971523236247,
+      "learning_rate": 4.0889500127716645e-06,
+      "loss": 0.5363,
+      "step": 6416
+    },
+    {
+      "epoch": 1.7035709544670117,
+      "grad_norm": 0.411511221648785,
+      "learning_rate": 4.088680458302891e-06,
+      "loss": 0.5983,
+      "step": 6417
+    },
+    {
+      "epoch": 1.7038364529403955,
+      "grad_norm": 0.39979271999330074,
+      "learning_rate": 4.0884108728509566e-06,
+      "loss": 0.5591,
+      "step": 6418
+    },
+    {
+      "epoch": 1.7041019514137794,
+      "grad_norm": 0.4105004758211771,
+      "learning_rate": 4.088141256421122e-06,
+      "loss": 0.5765,
+      "step": 6419
+    },
+    {
+      "epoch": 1.704367449887163,
+      "grad_norm": 0.41020123287357524,
+      "learning_rate": 4.087871609018645e-06,
+      "loss": 0.636,
+      "step": 6420
+    },
+    {
+      "epoch": 1.704632948360547,
+      "grad_norm": 0.40750140931572637,
+      "learning_rate": 4.087601930648783e-06,
+      "loss": 0.569,
+      "step": 6421
+    },
+    {
+      "epoch": 1.7048984468339308,
+      "grad_norm": 0.417438342071532,
+      "learning_rate": 4.087332221316797e-06,
+      "loss": 0.5523,
+      "step": 6422
+    },
+    {
+      "epoch": 1.7051639453073144,
+      "grad_norm": 0.41578059154919833,
+      "learning_rate": 4.087062481027946e-06,
+      "loss": 0.5877,
+      "step": 6423
+    },
+    {
+      "epoch": 1.7054294437806983,
+      "grad_norm": 0.41194798600497096,
+      "learning_rate": 4.08679270978749e-06,
+      "loss": 0.6104,
+      "step": 6424
+    },
+    {
+      "epoch": 1.7056949422540821,
+      "grad_norm": 0.4268203060982029,
+      "learning_rate": 4.086522907600692e-06,
+      "loss": 0.5771,
+      "step": 6425
+    },
+    {
+      "epoch": 1.7059604407274658,
+      "grad_norm": 0.4270837848572207,
+      "learning_rate": 4.086253074472812e-06,
+      "loss": 0.5695,
+      "step": 6426
+    },
+    {
+      "epoch": 1.7062259392008496,
+      "grad_norm": 0.40103513994611456,
+      "learning_rate": 4.085983210409114e-06,
+      "loss": 0.5901,
+      "step": 6427
+    },
+    {
+      "epoch": 1.7064914376742335,
+      "grad_norm": 0.4040823716747356,
+      "learning_rate": 4.08571331541486e-06,
+      "loss": 0.5956,
+      "step": 6428
+    },
+    {
+      "epoch": 1.7067569361476171,
+      "grad_norm": 0.425820221533074,
+      "learning_rate": 4.085443389495314e-06,
+      "loss": 0.5642,
+      "step": 6429
+    },
+    {
+      "epoch": 1.7070224346210008,
+      "grad_norm": 0.4227111587250706,
+      "learning_rate": 4.08517343265574e-06,
+      "loss": 0.59,
+      "step": 6430
+    },
+    {
+      "epoch": 1.7072879330943849,
+      "grad_norm": 0.4077812801578376,
+      "learning_rate": 4.084903444901402e-06,
+      "loss": 0.5754,
+      "step": 6431
+    },
+    {
+      "epoch": 1.7075534315677685,
+      "grad_norm": 0.4231708024750626,
+      "learning_rate": 4.084633426237568e-06,
+      "loss": 0.5746,
+      "step": 6432
+    },
+    {
+      "epoch": 1.7078189300411522,
+      "grad_norm": 0.4181944281392462,
+      "learning_rate": 4.084363376669501e-06,
+      "loss": 0.567,
+      "step": 6433
+    },
+    {
+      "epoch": 1.708084428514536,
+      "grad_norm": 0.4112539176711787,
+      "learning_rate": 4.0840932962024695e-06,
+      "loss": 0.5812,
+      "step": 6434
+    },
+    {
+      "epoch": 1.7083499269879199,
+      "grad_norm": 0.39504209909000465,
+      "learning_rate": 4.083823184841741e-06,
+      "loss": 0.5287,
+      "step": 6435
+    },
+    {
+      "epoch": 1.7086154254613035,
+      "grad_norm": 0.4077716554327273,
+      "learning_rate": 4.083553042592581e-06,
+      "loss": 0.528,
+      "step": 6436
+    },
+    {
+      "epoch": 1.7088809239346874,
+      "grad_norm": 0.41785135524504974,
+      "learning_rate": 4.08328286946026e-06,
+      "loss": 0.5905,
+      "step": 6437
+    },
+    {
+      "epoch": 1.7091464224080712,
+      "grad_norm": 0.393431350335522,
+      "learning_rate": 4.083012665450047e-06,
+      "loss": 0.597,
+      "step": 6438
+    },
+    {
+      "epoch": 1.7094119208814549,
+      "grad_norm": 0.3964504355407762,
+      "learning_rate": 4.08274243056721e-06,
+      "loss": 0.5872,
+      "step": 6439
+    },
+    {
+      "epoch": 1.7096774193548387,
+      "grad_norm": 0.40600021646242535,
+      "learning_rate": 4.082472164817021e-06,
+      "loss": 0.5958,
+      "step": 6440
+    },
+    {
+      "epoch": 1.7099429178282226,
+      "grad_norm": 0.38750031990339173,
+      "learning_rate": 4.082201868204751e-06,
+      "loss": 0.5506,
+      "step": 6441
+    },
+    {
+      "epoch": 1.7102084163016062,
+      "grad_norm": 0.4055901711825753,
+      "learning_rate": 4.08193154073567e-06,
+      "loss": 0.5962,
+      "step": 6442
+    },
+    {
+      "epoch": 1.71047391477499,
+      "grad_norm": 0.4067784155990374,
+      "learning_rate": 4.081661182415051e-06,
+      "loss": 0.576,
+      "step": 6443
+    },
+    {
+      "epoch": 1.710739413248374,
+      "grad_norm": 0.416429523114435,
+      "learning_rate": 4.081390793248166e-06,
+      "loss": 0.5791,
+      "step": 6444
+    },
+    {
+      "epoch": 1.7110049117217576,
+      "grad_norm": 0.4072772005433632,
+      "learning_rate": 4.081120373240288e-06,
+      "loss": 0.5927,
+      "step": 6445
+    },
+    {
+      "epoch": 1.7112704101951413,
+      "grad_norm": 0.39880803456168584,
+      "learning_rate": 4.080849922396692e-06,
+      "loss": 0.5874,
+      "step": 6446
+    },
+    {
+      "epoch": 1.7115359086685251,
+      "grad_norm": 0.4019850215024897,
+      "learning_rate": 4.080579440722653e-06,
+      "loss": 0.5968,
+      "step": 6447
+    },
+    {
+      "epoch": 1.711801407141909,
+      "grad_norm": 0.39670634230166185,
+      "learning_rate": 4.080308928223444e-06,
+      "loss": 0.559,
+      "step": 6448
+    },
+    {
+      "epoch": 1.7120669056152926,
+      "grad_norm": 0.410121864042993,
+      "learning_rate": 4.080038384904341e-06,
+      "loss": 0.6348,
+      "step": 6449
+    },
+    {
+      "epoch": 1.7123324040886765,
+      "grad_norm": 0.4080685061246559,
+      "learning_rate": 4.0797678107706215e-06,
+      "loss": 0.593,
+      "step": 6450
+    },
+    {
+      "epoch": 1.7125979025620603,
+      "grad_norm": 0.406464197577668,
+      "learning_rate": 4.079497205827563e-06,
+      "loss": 0.5889,
+      "step": 6451
+    },
+    {
+      "epoch": 1.712863401035444,
+      "grad_norm": 0.3972498954039883,
+      "learning_rate": 4.0792265700804405e-06,
+      "loss": 0.6046,
+      "step": 6452
+    },
+    {
+      "epoch": 1.7131288995088279,
+      "grad_norm": 0.4041997439336237,
+      "learning_rate": 4.078955903534533e-06,
+      "loss": 0.5807,
+      "step": 6453
+    },
+    {
+      "epoch": 1.7133943979822117,
+      "grad_norm": 0.4094804300767878,
+      "learning_rate": 4.0786852061951195e-06,
+      "loss": 0.5833,
+      "step": 6454
+    },
+    {
+      "epoch": 1.7136598964555954,
+      "grad_norm": 0.3990464385465636,
+      "learning_rate": 4.07841447806748e-06,
+      "loss": 0.6179,
+      "step": 6455
+    },
+    {
+      "epoch": 1.713925394928979,
+      "grad_norm": 0.4211035646587186,
+      "learning_rate": 4.078143719156893e-06,
+      "loss": 0.5843,
+      "step": 6456
+    },
+    {
+      "epoch": 1.714190893402363,
+      "grad_norm": 0.3862936986970333,
+      "learning_rate": 4.0778729294686395e-06,
+      "loss": 0.5577,
+      "step": 6457
+    },
+    {
+      "epoch": 1.7144563918757467,
+      "grad_norm": 0.4065149957277128,
+      "learning_rate": 4.077602109008001e-06,
+      "loss": 0.5903,
+      "step": 6458
+    },
+    {
+      "epoch": 1.7147218903491304,
+      "grad_norm": 0.4114984045862392,
+      "learning_rate": 4.077331257780257e-06,
+      "loss": 0.6215,
+      "step": 6459
+    },
+    {
+      "epoch": 1.7149873888225142,
+      "grad_norm": 0.41412438434444615,
+      "learning_rate": 4.077060375790694e-06,
+      "loss": 0.5886,
+      "step": 6460
+    },
+    {
+      "epoch": 1.715252887295898,
+      "grad_norm": 0.4145478491544096,
+      "learning_rate": 4.076789463044591e-06,
+      "loss": 0.6185,
+      "step": 6461
+    },
+    {
+      "epoch": 1.7155183857692817,
+      "grad_norm": 0.4079812481134643,
+      "learning_rate": 4.076518519547232e-06,
+      "loss": 0.5242,
+      "step": 6462
+    },
+    {
+      "epoch": 1.7157838842426656,
+      "grad_norm": 0.40181893687288417,
+      "learning_rate": 4.076247545303904e-06,
+      "loss": 0.5752,
+      "step": 6463
+    },
+    {
+      "epoch": 1.7160493827160495,
+      "grad_norm": 0.4280001907689151,
+      "learning_rate": 4.075976540319888e-06,
+      "loss": 0.5889,
+      "step": 6464
+    },
+    {
+      "epoch": 1.716314881189433,
+      "grad_norm": 0.40462215549039166,
+      "learning_rate": 4.075705504600471e-06,
+      "loss": 0.6138,
+      "step": 6465
+    },
+    {
+      "epoch": 1.716580379662817,
+      "grad_norm": 0.4202453038672166,
+      "learning_rate": 4.075434438150939e-06,
+      "loss": 0.6127,
+      "step": 6466
+    },
+    {
+      "epoch": 1.7168458781362008,
+      "grad_norm": 0.40610066860651095,
+      "learning_rate": 4.075163340976578e-06,
+      "loss": 0.5867,
+      "step": 6467
+    },
+    {
+      "epoch": 1.7171113766095845,
+      "grad_norm": 0.41606926502164887,
+      "learning_rate": 4.074892213082676e-06,
+      "loss": 0.6096,
+      "step": 6468
+    },
+    {
+      "epoch": 1.7173768750829683,
+      "grad_norm": 0.4070920623401771,
+      "learning_rate": 4.074621054474519e-06,
+      "loss": 0.5792,
+      "step": 6469
+    },
+    {
+      "epoch": 1.7176423735563522,
+      "grad_norm": 0.41991387058373475,
+      "learning_rate": 4.0743498651573966e-06,
+      "loss": 0.5878,
+      "step": 6470
+    },
+    {
+      "epoch": 1.7179078720297358,
+      "grad_norm": 0.4031444925611946,
+      "learning_rate": 4.074078645136597e-06,
+      "loss": 0.5925,
+      "step": 6471
+    },
+    {
+      "epoch": 1.7181733705031195,
+      "grad_norm": 0.40787428006592125,
+      "learning_rate": 4.073807394417409e-06,
+      "loss": 0.5918,
+      "step": 6472
+    },
+    {
+      "epoch": 1.7184388689765033,
+      "grad_norm": 0.4021743957758243,
+      "learning_rate": 4.073536113005125e-06,
+      "loss": 0.5677,
+      "step": 6473
+    },
+    {
+      "epoch": 1.7187043674498872,
+      "grad_norm": 0.40637385902873846,
+      "learning_rate": 4.073264800905032e-06,
+      "loss": 0.5862,
+      "step": 6474
+    },
+    {
+      "epoch": 1.7189698659232708,
+      "grad_norm": 0.4082521927974009,
+      "learning_rate": 4.072993458122425e-06,
+      "loss": 0.5639,
+      "step": 6475
+    },
+    {
+      "epoch": 1.7192353643966547,
+      "grad_norm": 0.4088434797337089,
+      "learning_rate": 4.072722084662593e-06,
+      "loss": 0.5662,
+      "step": 6476
+    },
+    {
+      "epoch": 1.7195008628700386,
+      "grad_norm": 0.4129241905146231,
+      "learning_rate": 4.072450680530831e-06,
+      "loss": 0.5906,
+      "step": 6477
+    },
+    {
+      "epoch": 1.7197663613434222,
+      "grad_norm": 0.4127538434151883,
+      "learning_rate": 4.07217924573243e-06,
+      "loss": 0.5959,
+      "step": 6478
+    },
+    {
+      "epoch": 1.720031859816806,
+      "grad_norm": 0.4299661159149051,
+      "learning_rate": 4.071907780272685e-06,
+      "loss": 0.5646,
+      "step": 6479
+    },
+    {
+      "epoch": 1.72029735829019,
+      "grad_norm": 0.42131397914061647,
+      "learning_rate": 4.071636284156888e-06,
+      "loss": 0.5905,
+      "step": 6480
+    },
+    {
+      "epoch": 1.7205628567635736,
+      "grad_norm": 0.4040503166510062,
+      "learning_rate": 4.071364757390337e-06,
+      "loss": 0.5559,
+      "step": 6481
+    },
+    {
+      "epoch": 1.7208283552369574,
+      "grad_norm": 0.4229742119781497,
+      "learning_rate": 4.071093199978325e-06,
+      "loss": 0.559,
+      "step": 6482
+    },
+    {
+      "epoch": 1.7210938537103413,
+      "grad_norm": 0.4091464314884923,
+      "learning_rate": 4.070821611926149e-06,
+      "loss": 0.6037,
+      "step": 6483
+    },
+    {
+      "epoch": 1.721359352183725,
+      "grad_norm": 0.41133828087437535,
+      "learning_rate": 4.070549993239106e-06,
+      "loss": 0.5789,
+      "step": 6484
+    },
+    {
+      "epoch": 1.7216248506571086,
+      "grad_norm": 0.4103944419430368,
+      "learning_rate": 4.070278343922492e-06,
+      "loss": 0.5738,
+      "step": 6485
+    },
+    {
+      "epoch": 1.7218903491304927,
+      "grad_norm": 0.4053549285718851,
+      "learning_rate": 4.070006663981606e-06,
+      "loss": 0.5963,
+      "step": 6486
+    },
+    {
+      "epoch": 1.7221558476038763,
+      "grad_norm": 0.4095061100300723,
+      "learning_rate": 4.069734953421746e-06,
+      "loss": 0.6,
+      "step": 6487
+    },
+    {
+      "epoch": 1.72242134607726,
+      "grad_norm": 0.41881698139462603,
+      "learning_rate": 4.0694632122482125e-06,
+      "loss": 0.58,
+      "step": 6488
+    },
+    {
+      "epoch": 1.7226868445506438,
+      "grad_norm": 0.40162987009694756,
+      "learning_rate": 4.069191440466302e-06,
+      "loss": 0.5906,
+      "step": 6489
+    },
+    {
+      "epoch": 1.7229523430240277,
+      "grad_norm": 0.40125551385564623,
+      "learning_rate": 4.0689196380813164e-06,
+      "loss": 0.5859,
+      "step": 6490
+    },
+    {
+      "epoch": 1.7232178414974113,
+      "grad_norm": 0.4096344823890862,
+      "learning_rate": 4.0686478050985565e-06,
+      "loss": 0.5923,
+      "step": 6491
+    },
+    {
+      "epoch": 1.7234833399707952,
+      "grad_norm": 0.4172744880573066,
+      "learning_rate": 4.068375941523325e-06,
+      "loss": 0.5975,
+      "step": 6492
+    },
+    {
+      "epoch": 1.723748838444179,
+      "grad_norm": 0.4183005841209401,
+      "learning_rate": 4.068104047360922e-06,
+      "loss": 0.5991,
+      "step": 6493
+    },
+    {
+      "epoch": 1.7240143369175627,
+      "grad_norm": 0.3965036355412523,
+      "learning_rate": 4.06783212261665e-06,
+      "loss": 0.5643,
+      "step": 6494
+    },
+    {
+      "epoch": 1.7242798353909465,
+      "grad_norm": 0.4265467477525225,
+      "learning_rate": 4.0675601672958126e-06,
+      "loss": 0.5802,
+      "step": 6495
+    },
+    {
+      "epoch": 1.7245453338643304,
+      "grad_norm": 0.42550491000281565,
+      "learning_rate": 4.067288181403715e-06,
+      "loss": 0.6092,
+      "step": 6496
+    },
+    {
+      "epoch": 1.724810832337714,
+      "grad_norm": 0.4056888465983238,
+      "learning_rate": 4.067016164945659e-06,
+      "loss": 0.5989,
+      "step": 6497
+    },
+    {
+      "epoch": 1.7250763308110977,
+      "grad_norm": 0.41399737876617976,
+      "learning_rate": 4.066744117926952e-06,
+      "loss": 0.5669,
+      "step": 6498
+    },
+    {
+      "epoch": 1.7253418292844818,
+      "grad_norm": 0.41084712423096337,
+      "learning_rate": 4.066472040352899e-06,
+      "loss": 0.5943,
+      "step": 6499
+    },
+    {
+      "epoch": 1.7256073277578654,
+      "grad_norm": 0.4254045083340015,
+      "learning_rate": 4.066199932228805e-06,
+      "loss": 0.5726,
+      "step": 6500
+    },
+    {
+      "epoch": 1.725872826231249,
+      "grad_norm": 0.41213205181867074,
+      "learning_rate": 4.065927793559978e-06,
+      "loss": 0.551,
+      "step": 6501
+    },
+    {
+      "epoch": 1.726138324704633,
+      "grad_norm": 0.40424425759441435,
+      "learning_rate": 4.065655624351725e-06,
+      "loss": 0.5483,
+      "step": 6502
+    },
+    {
+      "epoch": 1.7264038231780168,
+      "grad_norm": 0.4226921073138124,
+      "learning_rate": 4.065383424609354e-06,
+      "loss": 0.5804,
+      "step": 6503
+    },
+    {
+      "epoch": 1.7266693216514004,
+      "grad_norm": 0.3952102821170808,
+      "learning_rate": 4.065111194338174e-06,
+      "loss": 0.6039,
+      "step": 6504
+    },
+    {
+      "epoch": 1.7269348201247843,
+      "grad_norm": 0.4125175990813606,
+      "learning_rate": 4.064838933543494e-06,
+      "loss": 0.5747,
+      "step": 6505
+    },
+    {
+      "epoch": 1.7272003185981681,
+      "grad_norm": 0.40701259448877025,
+      "learning_rate": 4.064566642230622e-06,
+      "loss": 0.6064,
+      "step": 6506
+    },
+    {
+      "epoch": 1.7274658170715518,
+      "grad_norm": 0.4197157276715124,
+      "learning_rate": 4.06429432040487e-06,
+      "loss": 0.5983,
+      "step": 6507
+    },
+    {
+      "epoch": 1.7277313155449356,
+      "grad_norm": 0.4020441473168711,
+      "learning_rate": 4.0640219680715485e-06,
+      "loss": 0.5662,
+      "step": 6508
+    },
+    {
+      "epoch": 1.7279968140183195,
+      "grad_norm": 0.4079702267667392,
+      "learning_rate": 4.06374958523597e-06,
+      "loss": 0.6116,
+      "step": 6509
+    },
+    {
+      "epoch": 1.7282623124917031,
+      "grad_norm": 0.427835795562143,
+      "learning_rate": 4.063477171903446e-06,
+      "loss": 0.6076,
+      "step": 6510
+    },
+    {
+      "epoch": 1.7285278109650868,
+      "grad_norm": 0.41371501948257755,
+      "learning_rate": 4.063204728079289e-06,
+      "loss": 0.5818,
+      "step": 6511
+    },
+    {
+      "epoch": 1.7287933094384709,
+      "grad_norm": 0.41644671838677544,
+      "learning_rate": 4.062932253768812e-06,
+      "loss": 0.5681,
+      "step": 6512
+    },
+    {
+      "epoch": 1.7290588079118545,
+      "grad_norm": 0.40854132416092004,
+      "learning_rate": 4.062659748977329e-06,
+      "loss": 0.585,
+      "step": 6513
+    },
+    {
+      "epoch": 1.7293243063852382,
+      "grad_norm": 0.4370516538734631,
+      "learning_rate": 4.062387213710155e-06,
+      "loss": 0.5907,
+      "step": 6514
+    },
+    {
+      "epoch": 1.729589804858622,
+      "grad_norm": 0.42561001434745294,
+      "learning_rate": 4.0621146479726055e-06,
+      "loss": 0.5476,
+      "step": 6515
+    },
+    {
+      "epoch": 1.7298553033320059,
+      "grad_norm": 0.4169885273394613,
+      "learning_rate": 4.061842051769995e-06,
+      "loss": 0.5692,
+      "step": 6516
+    },
+    {
+      "epoch": 1.7301208018053895,
+      "grad_norm": 0.42042977366952405,
+      "learning_rate": 4.061569425107641e-06,
+      "loss": 0.5741,
+      "step": 6517
+    },
+    {
+      "epoch": 1.7303863002787734,
+      "grad_norm": 0.4201185458931557,
+      "learning_rate": 4.06129676799086e-06,
+      "loss": 0.5888,
+      "step": 6518
+    },
+    {
+      "epoch": 1.7306517987521572,
+      "grad_norm": 0.442603475118607,
+      "learning_rate": 4.061024080424969e-06,
+      "loss": 0.595,
+      "step": 6519
+    },
+    {
+      "epoch": 1.7309172972255409,
+      "grad_norm": 0.4096136870681202,
+      "learning_rate": 4.060751362415286e-06,
+      "loss": 0.5837,
+      "step": 6520
+    },
+    {
+      "epoch": 1.7311827956989247,
+      "grad_norm": 0.39845176144785643,
+      "learning_rate": 4.060478613967131e-06,
+      "loss": 0.5815,
+      "step": 6521
+    },
+    {
+      "epoch": 1.7314482941723086,
+      "grad_norm": 0.4172787874499624,
+      "learning_rate": 4.060205835085821e-06,
+      "loss": 0.5576,
+      "step": 6522
+    },
+    {
+      "epoch": 1.7317137926456923,
+      "grad_norm": 0.40926243043631466,
+      "learning_rate": 4.059933025776678e-06,
+      "loss": 0.591,
+      "step": 6523
+    },
+    {
+      "epoch": 1.7319792911190761,
+      "grad_norm": 0.40419760902700513,
+      "learning_rate": 4.059660186045023e-06,
+      "loss": 0.6001,
+      "step": 6524
+    },
+    {
+      "epoch": 1.73224478959246,
+      "grad_norm": 0.42930322248224045,
+      "learning_rate": 4.059387315896173e-06,
+      "loss": 0.5926,
+      "step": 6525
+    },
+    {
+      "epoch": 1.7325102880658436,
+      "grad_norm": 0.4186394730990391,
+      "learning_rate": 4.059114415335453e-06,
+      "loss": 0.5695,
+      "step": 6526
+    },
+    {
+      "epoch": 1.7327757865392273,
+      "grad_norm": 0.4286327221476898,
+      "learning_rate": 4.058841484368186e-06,
+      "loss": 0.5832,
+      "step": 6527
+    },
+    {
+      "epoch": 1.7330412850126113,
+      "grad_norm": 0.40727958260971603,
+      "learning_rate": 4.0585685229996915e-06,
+      "loss": 0.5557,
+      "step": 6528
+    },
+    {
+      "epoch": 1.733306783485995,
+      "grad_norm": 0.4232909270065776,
+      "learning_rate": 4.058295531235296e-06,
+      "loss": 0.5453,
+      "step": 6529
+    },
+    {
+      "epoch": 1.7335722819593786,
+      "grad_norm": 0.4088641700793539,
+      "learning_rate": 4.058022509080322e-06,
+      "loss": 0.5624,
+      "step": 6530
+    },
+    {
+      "epoch": 1.7338377804327625,
+      "grad_norm": 0.4241057977537865,
+      "learning_rate": 4.0577494565400944e-06,
+      "loss": 0.5991,
+      "step": 6531
+    },
+    {
+      "epoch": 1.7341032789061464,
+      "grad_norm": 0.4075900696694043,
+      "learning_rate": 4.057476373619938e-06,
+      "loss": 0.5869,
+      "step": 6532
+    },
+    {
+      "epoch": 1.73436877737953,
+      "grad_norm": 0.3968186468150876,
+      "learning_rate": 4.05720326032518e-06,
+      "loss": 0.5819,
+      "step": 6533
+    },
+    {
+      "epoch": 1.7346342758529139,
+      "grad_norm": 0.4289654645204103,
+      "learning_rate": 4.056930116661144e-06,
+      "loss": 0.5416,
+      "step": 6534
+    },
+    {
+      "epoch": 1.7348997743262977,
+      "grad_norm": 0.4277504534371669,
+      "learning_rate": 4.05665694263316e-06,
+      "loss": 0.5946,
+      "step": 6535
+    },
+    {
+      "epoch": 1.7351652727996814,
+      "grad_norm": 0.4229210958706269,
+      "learning_rate": 4.056383738246554e-06,
+      "loss": 0.5883,
+      "step": 6536
+    },
+    {
+      "epoch": 1.7354307712730652,
+      "grad_norm": 0.4053274095420921,
+      "learning_rate": 4.056110503506654e-06,
+      "loss": 0.5768,
+      "step": 6537
+    },
+    {
+      "epoch": 1.735696269746449,
+      "grad_norm": 0.3984035050388826,
+      "learning_rate": 4.055837238418791e-06,
+      "loss": 0.5626,
+      "step": 6538
+    },
+    {
+      "epoch": 1.7359617682198327,
+      "grad_norm": 0.43063569711169525,
+      "learning_rate": 4.05556394298829e-06,
+      "loss": 0.596,
+      "step": 6539
+    },
+    {
+      "epoch": 1.7362272666932164,
+      "grad_norm": 0.4163124255485391,
+      "learning_rate": 4.0552906172204845e-06,
+      "loss": 0.6004,
+      "step": 6540
+    },
+    {
+      "epoch": 1.7364927651666004,
+      "grad_norm": 0.40868346600138583,
+      "learning_rate": 4.055017261120704e-06,
+      "loss": 0.5904,
+      "step": 6541
+    },
+    {
+      "epoch": 1.736758263639984,
+      "grad_norm": 0.4076593578635097,
+      "learning_rate": 4.054743874694279e-06,
+      "loss": 0.5341,
+      "step": 6542
+    },
+    {
+      "epoch": 1.7370237621133677,
+      "grad_norm": 0.412052739537721,
+      "learning_rate": 4.054470457946542e-06,
+      "loss": 0.5846,
+      "step": 6543
+    },
+    {
+      "epoch": 1.7372892605867516,
+      "grad_norm": 0.4125055410366082,
+      "learning_rate": 4.054197010882826e-06,
+      "loss": 0.5922,
+      "step": 6544
+    },
+    {
+      "epoch": 1.7375547590601355,
+      "grad_norm": 0.4133294661631091,
+      "learning_rate": 4.053923533508462e-06,
+      "loss": 0.5897,
+      "step": 6545
+    },
+    {
+      "epoch": 1.737820257533519,
+      "grad_norm": 0.4145985645294253,
+      "learning_rate": 4.053650025828785e-06,
+      "loss": 0.5565,
+      "step": 6546
+    },
+    {
+      "epoch": 1.738085756006903,
+      "grad_norm": 0.4081788673106127,
+      "learning_rate": 4.0533764878491285e-06,
+      "loss": 0.5634,
+      "step": 6547
+    },
+    {
+      "epoch": 1.7383512544802868,
+      "grad_norm": 0.40319963446551593,
+      "learning_rate": 4.0531029195748265e-06,
+      "loss": 0.588,
+      "step": 6548
+    },
+    {
+      "epoch": 1.7386167529536705,
+      "grad_norm": 0.4061629001496505,
+      "learning_rate": 4.052829321011215e-06,
+      "loss": 0.5771,
+      "step": 6549
+    },
+    {
+      "epoch": 1.7388822514270543,
+      "grad_norm": 0.4084395782841003,
+      "learning_rate": 4.0525556921636296e-06,
+      "loss": 0.6114,
+      "step": 6550
+    },
+    {
+      "epoch": 1.7391477499004382,
+      "grad_norm": 0.4218094049712499,
+      "learning_rate": 4.052282033037408e-06,
+      "loss": 0.5761,
+      "step": 6551
+    },
+    {
+      "epoch": 1.7394132483738218,
+      "grad_norm": 0.43864736949046174,
+      "learning_rate": 4.052008343637885e-06,
+      "loss": 0.5658,
+      "step": 6552
+    },
+    {
+      "epoch": 1.7396787468472055,
+      "grad_norm": 0.4092254357176569,
+      "learning_rate": 4.0517346239704e-06,
+      "loss": 0.5668,
+      "step": 6553
+    },
+    {
+      "epoch": 1.7399442453205896,
+      "grad_norm": 0.4181490669998728,
+      "learning_rate": 4.05146087404029e-06,
+      "loss": 0.5832,
+      "step": 6554
+    },
+    {
+      "epoch": 1.7402097437939732,
+      "grad_norm": 0.43196287879216094,
+      "learning_rate": 4.051187093852895e-06,
+      "loss": 0.5782,
+      "step": 6555
+    },
+    {
+      "epoch": 1.7404752422673568,
+      "grad_norm": 0.4770861667131978,
+      "learning_rate": 4.050913283413553e-06,
+      "loss": 0.5812,
+      "step": 6556
+    },
+    {
+      "epoch": 1.7407407407407407,
+      "grad_norm": 0.4199658479211904,
+      "learning_rate": 4.050639442727606e-06,
+      "loss": 0.5751,
+      "step": 6557
+    },
+    {
+      "epoch": 1.7410062392141246,
+      "grad_norm": 0.3996931497967501,
+      "learning_rate": 4.050365571800392e-06,
+      "loss": 0.5668,
+      "step": 6558
+    },
+    {
+      "epoch": 1.7412717376875082,
+      "grad_norm": 0.4678040977474167,
+      "learning_rate": 4.050091670637254e-06,
+      "loss": 0.5583,
+      "step": 6559
+    },
+    {
+      "epoch": 1.741537236160892,
+      "grad_norm": 0.4243455377755792,
+      "learning_rate": 4.049817739243532e-06,
+      "loss": 0.5754,
+      "step": 6560
+    },
+    {
+      "epoch": 1.741802734634276,
+      "grad_norm": 0.43202984752398316,
+      "learning_rate": 4.049543777624571e-06,
+      "loss": 0.6082,
+      "step": 6561
+    },
+    {
+      "epoch": 1.7420682331076596,
+      "grad_norm": 0.40720103442299777,
+      "learning_rate": 4.0492697857857115e-06,
+      "loss": 0.5994,
+      "step": 6562
+    },
+    {
+      "epoch": 1.7423337315810434,
+      "grad_norm": 0.42362962610328386,
+      "learning_rate": 4.048995763732298e-06,
+      "loss": 0.6032,
+      "step": 6563
+    },
+    {
+      "epoch": 1.7425992300544273,
+      "grad_norm": 0.4488404055454519,
+      "learning_rate": 4.048721711469675e-06,
+      "loss": 0.5417,
+      "step": 6564
+    },
+    {
+      "epoch": 1.742864728527811,
+      "grad_norm": 0.41165074162444915,
+      "learning_rate": 4.048447629003186e-06,
+      "loss": 0.599,
+      "step": 6565
+    },
+    {
+      "epoch": 1.7431302270011948,
+      "grad_norm": 0.4077958636446163,
+      "learning_rate": 4.048173516338178e-06,
+      "loss": 0.6019,
+      "step": 6566
+    },
+    {
+      "epoch": 1.7433957254745787,
+      "grad_norm": 0.4092268470827822,
+      "learning_rate": 4.047899373479994e-06,
+      "loss": 0.5874,
+      "step": 6567
+    },
+    {
+      "epoch": 1.7436612239479623,
+      "grad_norm": 0.5179927698481733,
+      "learning_rate": 4.047625200433984e-06,
+      "loss": 0.5448,
+      "step": 6568
+    },
+    {
+      "epoch": 1.743926722421346,
+      "grad_norm": 0.42831109605807205,
+      "learning_rate": 4.047350997205492e-06,
+      "loss": 0.5797,
+      "step": 6569
+    },
+    {
+      "epoch": 1.7441922208947298,
+      "grad_norm": 0.42182548063428943,
+      "learning_rate": 4.047076763799869e-06,
+      "loss": 0.5863,
+      "step": 6570
+    },
+    {
+      "epoch": 1.7444577193681137,
+      "grad_norm": 0.4975277214813528,
+      "learning_rate": 4.04680250022246e-06,
+      "loss": 0.5956,
+      "step": 6571
+    },
+    {
+      "epoch": 1.7447232178414973,
+      "grad_norm": 0.42253041402695946,
+      "learning_rate": 4.046528206478615e-06,
+      "loss": 0.5696,
+      "step": 6572
+    },
+    {
+      "epoch": 1.7449887163148812,
+      "grad_norm": 0.41363595365217243,
+      "learning_rate": 4.046253882573683e-06,
+      "loss": 0.5485,
+      "step": 6573
+    },
+    {
+      "epoch": 1.745254214788265,
+      "grad_norm": 0.40893550641369186,
+      "learning_rate": 4.045979528513015e-06,
+      "loss": 0.5715,
+      "step": 6574
+    },
+    {
+      "epoch": 1.7455197132616487,
+      "grad_norm": 0.4154769475342251,
+      "learning_rate": 4.0457051443019606e-06,
+      "loss": 0.59,
+      "step": 6575
+    },
+    {
+      "epoch": 1.7457852117350325,
+      "grad_norm": 0.40984150144553694,
+      "learning_rate": 4.045430729945872e-06,
+      "loss": 0.6362,
+      "step": 6576
+    },
+    {
+      "epoch": 1.7460507102084164,
+      "grad_norm": 0.41910663541223325,
+      "learning_rate": 4.0451562854501e-06,
+      "loss": 0.5349,
+      "step": 6577
+    },
+    {
+      "epoch": 1.7463162086818,
+      "grad_norm": 0.4032734944892639,
+      "learning_rate": 4.044881810819996e-06,
+      "loss": 0.5586,
+      "step": 6578
+    },
+    {
+      "epoch": 1.746581707155184,
+      "grad_norm": 0.4346664314469867,
+      "learning_rate": 4.044607306060916e-06,
+      "loss": 0.5962,
+      "step": 6579
+    },
+    {
+      "epoch": 1.7468472056285678,
+      "grad_norm": 0.41575913957684874,
+      "learning_rate": 4.044332771178211e-06,
+      "loss": 0.6091,
+      "step": 6580
+    },
+    {
+      "epoch": 1.7471127041019514,
+      "grad_norm": 0.4102087393071203,
+      "learning_rate": 4.044058206177236e-06,
+      "loss": 0.5669,
+      "step": 6581
+    },
+    {
+      "epoch": 1.747378202575335,
+      "grad_norm": 0.42439716602294586,
+      "learning_rate": 4.0437836110633455e-06,
+      "loss": 0.6089,
+      "step": 6582
+    },
+    {
+      "epoch": 1.7476437010487191,
+      "grad_norm": 0.3952857331589499,
+      "learning_rate": 4.043508985841894e-06,
+      "loss": 0.6174,
+      "step": 6583
+    },
+    {
+      "epoch": 1.7479091995221028,
+      "grad_norm": 0.3998263039604155,
+      "learning_rate": 4.043234330518239e-06,
+      "loss": 0.5714,
+      "step": 6584
+    },
+    {
+      "epoch": 1.7481746979954864,
+      "grad_norm": 0.40169484274428324,
+      "learning_rate": 4.042959645097735e-06,
+      "loss": 0.5509,
+      "step": 6585
+    },
+    {
+      "epoch": 1.7484401964688703,
+      "grad_norm": 0.406623158965972,
+      "learning_rate": 4.04268492958574e-06,
+      "loss": 0.5843,
+      "step": 6586
+    },
+    {
+      "epoch": 1.7487056949422541,
+      "grad_norm": 0.4061533543998768,
+      "learning_rate": 4.042410183987614e-06,
+      "loss": 0.5658,
+      "step": 6587
+    },
+    {
+      "epoch": 1.7489711934156378,
+      "grad_norm": 0.4211592982661729,
+      "learning_rate": 4.042135408308711e-06,
+      "loss": 0.5963,
+      "step": 6588
+    },
+    {
+      "epoch": 1.7492366918890216,
+      "grad_norm": 0.40864661096244187,
+      "learning_rate": 4.041860602554392e-06,
+      "loss": 0.5803,
+      "step": 6589
+    },
+    {
+      "epoch": 1.7495021903624055,
+      "grad_norm": 0.4254625447919162,
+      "learning_rate": 4.041585766730016e-06,
+      "loss": 0.5877,
+      "step": 6590
+    },
+    {
+      "epoch": 1.7497676888357891,
+      "grad_norm": 0.41918740410262323,
+      "learning_rate": 4.041310900840944e-06,
+      "loss": 0.5882,
+      "step": 6591
+    },
+    {
+      "epoch": 1.750033187309173,
+      "grad_norm": 0.4184930025211725,
+      "learning_rate": 4.041036004892534e-06,
+      "loss": 0.5499,
+      "step": 6592
+    },
+    {
+      "epoch": 1.7502986857825569,
+      "grad_norm": 0.41521741050048006,
+      "learning_rate": 4.040761078890151e-06,
+      "loss": 0.5744,
+      "step": 6593
+    },
+    {
+      "epoch": 1.7505641842559405,
+      "grad_norm": 0.40365113020255894,
+      "learning_rate": 4.040486122839154e-06,
+      "loss": 0.5667,
+      "step": 6594
+    },
+    {
+      "epoch": 1.7508296827293242,
+      "grad_norm": 0.40818413024588884,
+      "learning_rate": 4.040211136744905e-06,
+      "loss": 0.5677,
+      "step": 6595
+    },
+    {
+      "epoch": 1.7510951812027082,
+      "grad_norm": 0.4076273584168611,
+      "learning_rate": 4.039936120612767e-06,
+      "loss": 0.5729,
+      "step": 6596
+    },
+    {
+      "epoch": 1.7513606796760919,
+      "grad_norm": 0.40528917919580143,
+      "learning_rate": 4.0396610744481055e-06,
+      "loss": 0.5872,
+      "step": 6597
+    },
+    {
+      "epoch": 1.7516261781494755,
+      "grad_norm": 0.3987071554952435,
+      "learning_rate": 4.0393859982562834e-06,
+      "loss": 0.5426,
+      "step": 6598
+    },
+    {
+      "epoch": 1.7518916766228594,
+      "grad_norm": 0.41900645372489553,
+      "learning_rate": 4.039110892042664e-06,
+      "loss": 0.5809,
+      "step": 6599
+    },
+    {
+      "epoch": 1.7521571750962432,
+      "grad_norm": 0.4069892406717045,
+      "learning_rate": 4.038835755812615e-06,
+      "loss": 0.5912,
+      "step": 6600
+    },
+    {
+      "epoch": 1.7524226735696269,
+      "grad_norm": 0.4176666863118596,
+      "learning_rate": 4.038560589571501e-06,
+      "loss": 0.5838,
+      "step": 6601
+    },
+    {
+      "epoch": 1.7526881720430108,
+      "grad_norm": 0.42499834184545204,
+      "learning_rate": 4.038285393324689e-06,
+      "loss": 0.5626,
+      "step": 6602
+    },
+    {
+      "epoch": 1.7529536705163946,
+      "grad_norm": 0.4184014937857046,
+      "learning_rate": 4.038010167077544e-06,
+      "loss": 0.5967,
+      "step": 6603
+    },
+    {
+      "epoch": 1.7532191689897783,
+      "grad_norm": 0.4229606510310555,
+      "learning_rate": 4.0377349108354355e-06,
+      "loss": 0.5593,
+      "step": 6604
+    },
+    {
+      "epoch": 1.7534846674631621,
+      "grad_norm": 0.4190484197388039,
+      "learning_rate": 4.037459624603731e-06,
+      "loss": 0.5682,
+      "step": 6605
+    },
+    {
+      "epoch": 1.753750165936546,
+      "grad_norm": 0.4040382952238776,
+      "learning_rate": 4.037184308387801e-06,
+      "loss": 0.5687,
+      "step": 6606
+    },
+    {
+      "epoch": 1.7540156644099296,
+      "grad_norm": 0.39656698413284036,
+      "learning_rate": 4.036908962193011e-06,
+      "loss": 0.5708,
+      "step": 6607
+    },
+    {
+      "epoch": 1.7542811628833133,
+      "grad_norm": 0.41894603382728196,
+      "learning_rate": 4.036633586024735e-06,
+      "loss": 0.5727,
+      "step": 6608
+    },
+    {
+      "epoch": 1.7545466613566973,
+      "grad_norm": 0.4368774187877,
+      "learning_rate": 4.036358179888341e-06,
+      "loss": 0.5592,
+      "step": 6609
+    },
+    {
+      "epoch": 1.754812159830081,
+      "grad_norm": 0.4053271824935084,
+      "learning_rate": 4.036082743789202e-06,
+      "loss": 0.5742,
+      "step": 6610
+    },
+    {
+      "epoch": 1.7550776583034646,
+      "grad_norm": 0.4025499347636373,
+      "learning_rate": 4.0358072777326865e-06,
+      "loss": 0.5453,
+      "step": 6611
+    },
+    {
+      "epoch": 1.7553431567768485,
+      "grad_norm": 0.41175185466705605,
+      "learning_rate": 4.0355317817241705e-06,
+      "loss": 0.5881,
+      "step": 6612
+    },
+    {
+      "epoch": 1.7556086552502324,
+      "grad_norm": 0.42963526717887235,
+      "learning_rate": 4.035256255769024e-06,
+      "loss": 0.6022,
+      "step": 6613
+    },
+    {
+      "epoch": 1.755874153723616,
+      "grad_norm": 0.4084833324235432,
+      "learning_rate": 4.034980699872622e-06,
+      "loss": 0.5866,
+      "step": 6614
+    },
+    {
+      "epoch": 1.7561396521969999,
+      "grad_norm": 0.40609272449252576,
+      "learning_rate": 4.0347051140403385e-06,
+      "loss": 0.5548,
+      "step": 6615
+    },
+    {
+      "epoch": 1.7564051506703837,
+      "grad_norm": 0.4247669150585388,
+      "learning_rate": 4.034429498277546e-06,
+      "loss": 0.5975,
+      "step": 6616
+    },
+    {
+      "epoch": 1.7566706491437674,
+      "grad_norm": 0.42602218826048216,
+      "learning_rate": 4.034153852589623e-06,
+      "loss": 0.5491,
+      "step": 6617
+    },
+    {
+      "epoch": 1.7569361476171512,
+      "grad_norm": 0.4029172410004458,
+      "learning_rate": 4.033878176981943e-06,
+      "loss": 0.5851,
+      "step": 6618
+    },
+    {
+      "epoch": 1.757201646090535,
+      "grad_norm": 0.41703193602472705,
+      "learning_rate": 4.033602471459884e-06,
+      "loss": 0.5651,
+      "step": 6619
+    },
+    {
+      "epoch": 1.7574671445639187,
+      "grad_norm": 0.42727637831045134,
+      "learning_rate": 4.033326736028821e-06,
+      "loss": 0.5676,
+      "step": 6620
+    },
+    {
+      "epoch": 1.7577326430373026,
+      "grad_norm": 0.41000351509458977,
+      "learning_rate": 4.033050970694133e-06,
+      "loss": 0.6082,
+      "step": 6621
+    },
+    {
+      "epoch": 1.7579981415106865,
+      "grad_norm": 0.39783516034211625,
+      "learning_rate": 4.032775175461196e-06,
+      "loss": 0.5912,
+      "step": 6622
+    },
+    {
+      "epoch": 1.75826363998407,
+      "grad_norm": 0.3967768396833056,
+      "learning_rate": 4.0324993503353915e-06,
+      "loss": 0.5979,
+      "step": 6623
+    },
+    {
+      "epoch": 1.7585291384574537,
+      "grad_norm": 0.41285797653939565,
+      "learning_rate": 4.0322234953220965e-06,
+      "loss": 0.5515,
+      "step": 6624
+    },
+    {
+      "epoch": 1.7587946369308376,
+      "grad_norm": 0.4150974076269747,
+      "learning_rate": 4.031947610426693e-06,
+      "loss": 0.5872,
+      "step": 6625
+    },
+    {
+      "epoch": 1.7590601354042215,
+      "grad_norm": 0.40716120170655684,
+      "learning_rate": 4.03167169565456e-06,
+      "loss": 0.5872,
+      "step": 6626
+    },
+    {
+      "epoch": 1.759325633877605,
+      "grad_norm": 0.407629409510305,
+      "learning_rate": 4.031395751011078e-06,
+      "loss": 0.5818,
+      "step": 6627
+    },
+    {
+      "epoch": 1.759591132350989,
+      "grad_norm": 0.4087098730370454,
+      "learning_rate": 4.03111977650163e-06,
+      "loss": 0.6263,
+      "step": 6628
+    },
+    {
+      "epoch": 1.7598566308243728,
+      "grad_norm": 0.42017328199788445,
+      "learning_rate": 4.030843772131597e-06,
+      "loss": 0.5782,
+      "step": 6629
+    },
+    {
+      "epoch": 1.7601221292977565,
+      "grad_norm": 0.39945350672825264,
+      "learning_rate": 4.030567737906363e-06,
+      "loss": 0.5714,
+      "step": 6630
+    },
+    {
+      "epoch": 1.7603876277711403,
+      "grad_norm": 0.4104803283690374,
+      "learning_rate": 4.03029167383131e-06,
+      "loss": 0.5762,
+      "step": 6631
+    },
+    {
+      "epoch": 1.7606531262445242,
+      "grad_norm": 0.4093183971814103,
+      "learning_rate": 4.030015579911823e-06,
+      "loss": 0.6214,
+      "step": 6632
+    },
+    {
+      "epoch": 1.7609186247179078,
+      "grad_norm": 0.40562923557932873,
+      "learning_rate": 4.0297394561532856e-06,
+      "loss": 0.5867,
+      "step": 6633
+    },
+    {
+      "epoch": 1.7611841231912917,
+      "grad_norm": 0.40816175435277885,
+      "learning_rate": 4.029463302561084e-06,
+      "loss": 0.6062,
+      "step": 6634
+    },
+    {
+      "epoch": 1.7614496216646756,
+      "grad_norm": 0.41015111420931805,
+      "learning_rate": 4.029187119140603e-06,
+      "loss": 0.6167,
+      "step": 6635
+    },
+    {
+      "epoch": 1.7617151201380592,
+      "grad_norm": 0.4235618497935881,
+      "learning_rate": 4.028910905897229e-06,
+      "loss": 0.5712,
+      "step": 6636
+    },
+    {
+      "epoch": 1.7619806186114428,
+      "grad_norm": 0.40568308901574224,
+      "learning_rate": 4.028634662836349e-06,
+      "loss": 0.5911,
+      "step": 6637
+    },
+    {
+      "epoch": 1.762246117084827,
+      "grad_norm": 0.4133862341077795,
+      "learning_rate": 4.02835838996335e-06,
+      "loss": 0.5633,
+      "step": 6638
+    },
+    {
+      "epoch": 1.7625116155582106,
+      "grad_norm": 0.41431337075236846,
+      "learning_rate": 4.028082087283621e-06,
+      "loss": 0.5846,
+      "step": 6639
+    },
+    {
+      "epoch": 1.7627771140315942,
+      "grad_norm": 0.4089988875488357,
+      "learning_rate": 4.027805754802549e-06,
+      "loss": 0.585,
+      "step": 6640
+    },
+    {
+      "epoch": 1.763042612504978,
+      "grad_norm": 0.40003121162604544,
+      "learning_rate": 4.0275293925255245e-06,
+      "loss": 0.6021,
+      "step": 6641
+    },
+    {
+      "epoch": 1.763308110978362,
+      "grad_norm": 0.41519379930250433,
+      "learning_rate": 4.027253000457937e-06,
+      "loss": 0.5901,
+      "step": 6642
+    },
+    {
+      "epoch": 1.7635736094517456,
+      "grad_norm": 0.41866266336762176,
+      "learning_rate": 4.026976578605176e-06,
+      "loss": 0.5641,
+      "step": 6643
+    },
+    {
+      "epoch": 1.7638391079251294,
+      "grad_norm": 0.4062009426567029,
+      "learning_rate": 4.026700126972633e-06,
+      "loss": 0.5777,
+      "step": 6644
+    },
+    {
+      "epoch": 1.7641046063985133,
+      "grad_norm": 0.40409256120154463,
+      "learning_rate": 4.0264236455657005e-06,
+      "loss": 0.6167,
+      "step": 6645
+    },
+    {
+      "epoch": 1.764370104871897,
+      "grad_norm": 0.40694387183136166,
+      "learning_rate": 4.026147134389769e-06,
+      "loss": 0.5743,
+      "step": 6646
+    },
+    {
+      "epoch": 1.7646356033452808,
+      "grad_norm": 0.42303410484491405,
+      "learning_rate": 4.025870593450232e-06,
+      "loss": 0.6023,
+      "step": 6647
+    },
+    {
+      "epoch": 1.7649011018186647,
+      "grad_norm": 0.43716108426476313,
+      "learning_rate": 4.025594022752483e-06,
+      "loss": 0.5843,
+      "step": 6648
+    },
+    {
+      "epoch": 1.7651666002920483,
+      "grad_norm": 0.45202150160351534,
+      "learning_rate": 4.025317422301914e-06,
+      "loss": 0.5743,
+      "step": 6649
+    },
+    {
+      "epoch": 1.765432098765432,
+      "grad_norm": 0.41025441912786664,
+      "learning_rate": 4.025040792103922e-06,
+      "loss": 0.6046,
+      "step": 6650
+    },
+    {
+      "epoch": 1.765697597238816,
+      "grad_norm": 0.4124058776081248,
+      "learning_rate": 4.024764132163899e-06,
+      "loss": 0.5696,
+      "step": 6651
+    },
+    {
+      "epoch": 1.7659630957121997,
+      "grad_norm": 0.4108617134818245,
+      "learning_rate": 4.024487442487242e-06,
+      "loss": 0.6156,
+      "step": 6652
+    },
+    {
+      "epoch": 1.7662285941855833,
+      "grad_norm": 0.4105477491847005,
+      "learning_rate": 4.02421072307935e-06,
+      "loss": 0.5805,
+      "step": 6653
+    },
+    {
+      "epoch": 1.7664940926589672,
+      "grad_norm": 0.42833907831174295,
+      "learning_rate": 4.023933973945615e-06,
+      "loss": 0.579,
+      "step": 6654
+    },
+    {
+      "epoch": 1.766759591132351,
+      "grad_norm": 0.4171271465893289,
+      "learning_rate": 4.023657195091436e-06,
+      "loss": 0.5553,
+      "step": 6655
+    },
+    {
+      "epoch": 1.7670250896057347,
+      "grad_norm": 0.40330478697567573,
+      "learning_rate": 4.023380386522211e-06,
+      "loss": 0.5996,
+      "step": 6656
+    },
+    {
+      "epoch": 1.7672905880791185,
+      "grad_norm": 0.41116862166087287,
+      "learning_rate": 4.02310354824334e-06,
+      "loss": 0.5758,
+      "step": 6657
+    },
+    {
+      "epoch": 1.7675560865525024,
+      "grad_norm": 0.39769075787674824,
+      "learning_rate": 4.02282668026022e-06,
+      "loss": 0.5986,
+      "step": 6658
+    },
+    {
+      "epoch": 1.767821585025886,
+      "grad_norm": 0.39000250716463525,
+      "learning_rate": 4.02254978257825e-06,
+      "loss": 0.5825,
+      "step": 6659
+    },
+    {
+      "epoch": 1.76808708349927,
+      "grad_norm": 0.40864099756177,
+      "learning_rate": 4.022272855202833e-06,
+      "loss": 0.5736,
+      "step": 6660
+    },
+    {
+      "epoch": 1.7683525819726538,
+      "grad_norm": 0.4053326510770391,
+      "learning_rate": 4.021995898139367e-06,
+      "loss": 0.5744,
+      "step": 6661
+    },
+    {
+      "epoch": 1.7686180804460374,
+      "grad_norm": 0.4136214850527255,
+      "learning_rate": 4.021718911393256e-06,
+      "loss": 0.5621,
+      "step": 6662
+    },
+    {
+      "epoch": 1.768883578919421,
+      "grad_norm": 0.4069967911742805,
+      "learning_rate": 4.021441894969899e-06,
+      "loss": 0.5657,
+      "step": 6663
+    },
+    {
+      "epoch": 1.7691490773928051,
+      "grad_norm": 0.40290058691802394,
+      "learning_rate": 4.0211648488747005e-06,
+      "loss": 0.6243,
+      "step": 6664
+    },
+    {
+      "epoch": 1.7694145758661888,
+      "grad_norm": 0.4142998862472786,
+      "learning_rate": 4.020887773113063e-06,
+      "loss": 0.5847,
+      "step": 6665
+    },
+    {
+      "epoch": 1.7696800743395724,
+      "grad_norm": 0.3995043314169401,
+      "learning_rate": 4.02061066769039e-06,
+      "loss": 0.6061,
+      "step": 6666
+    },
+    {
+      "epoch": 1.7699455728129563,
+      "grad_norm": 0.4053016745797228,
+      "learning_rate": 4.020333532612086e-06,
+      "loss": 0.5872,
+      "step": 6667
+    },
+    {
+      "epoch": 1.7702110712863401,
+      "grad_norm": 0.42178003999694946,
+      "learning_rate": 4.020056367883556e-06,
+      "loss": 0.5754,
+      "step": 6668
+    },
+    {
+      "epoch": 1.7704765697597238,
+      "grad_norm": 0.4285221713624082,
+      "learning_rate": 4.019779173510204e-06,
+      "loss": 0.598,
+      "step": 6669
+    },
+    {
+      "epoch": 1.7707420682331076,
+      "grad_norm": 0.42580296352122304,
+      "learning_rate": 4.019501949497438e-06,
+      "loss": 0.5709,
+      "step": 6670
+    },
+    {
+      "epoch": 1.7710075667064915,
+      "grad_norm": 0.404118880417849,
+      "learning_rate": 4.019224695850664e-06,
+      "loss": 0.5765,
+      "step": 6671
+    },
+    {
+      "epoch": 1.7712730651798752,
+      "grad_norm": 0.4124397369750885,
+      "learning_rate": 4.018947412575287e-06,
+      "loss": 0.5776,
+      "step": 6672
+    },
+    {
+      "epoch": 1.771538563653259,
+      "grad_norm": 0.4231230463740868,
+      "learning_rate": 4.018670099676718e-06,
+      "loss": 0.5689,
+      "step": 6673
+    },
+    {
+      "epoch": 1.7718040621266429,
+      "grad_norm": 0.39437917859929883,
+      "learning_rate": 4.018392757160363e-06,
+      "loss": 0.5943,
+      "step": 6674
+    },
+    {
+      "epoch": 1.7720695606000265,
+      "grad_norm": 0.41780077712783203,
+      "learning_rate": 4.018115385031632e-06,
+      "loss": 0.5705,
+      "step": 6675
+    },
+    {
+      "epoch": 1.7723350590734104,
+      "grad_norm": 0.40785215920560286,
+      "learning_rate": 4.017837983295934e-06,
+      "loss": 0.5825,
+      "step": 6676
+    },
+    {
+      "epoch": 1.7726005575467942,
+      "grad_norm": 0.43008972547146007,
+      "learning_rate": 4.017560551958679e-06,
+      "loss": 0.5783,
+      "step": 6677
+    },
+    {
+      "epoch": 1.7728660560201779,
+      "grad_norm": 0.40849145662135616,
+      "learning_rate": 4.0172830910252765e-06,
+      "loss": 0.6143,
+      "step": 6678
+    },
+    {
+      "epoch": 1.7731315544935615,
+      "grad_norm": 0.4278505762267433,
+      "learning_rate": 4.01700560050114e-06,
+      "loss": 0.5647,
+      "step": 6679
+    },
+    {
+      "epoch": 1.7733970529669454,
+      "grad_norm": 0.3951472079430473,
+      "learning_rate": 4.01672808039168e-06,
+      "loss": 0.587,
+      "step": 6680
+    },
+    {
+      "epoch": 1.7736625514403292,
+      "grad_norm": 0.4141906323868477,
+      "learning_rate": 4.016450530702309e-06,
+      "loss": 0.6067,
+      "step": 6681
+    },
+    {
+      "epoch": 1.773928049913713,
+      "grad_norm": 0.4153852526456724,
+      "learning_rate": 4.01617295143844e-06,
+      "loss": 0.5737,
+      "step": 6682
+    },
+    {
+      "epoch": 1.7741935483870968,
+      "grad_norm": 0.41014387610959574,
+      "learning_rate": 4.015895342605485e-06,
+      "loss": 0.5804,
+      "step": 6683
+    },
+    {
+      "epoch": 1.7744590468604806,
+      "grad_norm": 0.4009121828074509,
+      "learning_rate": 4.01561770420886e-06,
+      "loss": 0.5796,
+      "step": 6684
+    },
+    {
+      "epoch": 1.7747245453338643,
+      "grad_norm": 0.4196681349617996,
+      "learning_rate": 4.015340036253979e-06,
+      "loss": 0.5752,
+      "step": 6685
+    },
+    {
+      "epoch": 1.7749900438072481,
+      "grad_norm": 0.4170897476613166,
+      "learning_rate": 4.015062338746256e-06,
+      "loss": 0.5579,
+      "step": 6686
+    },
+    {
+      "epoch": 1.775255542280632,
+      "grad_norm": 0.3999841141295786,
+      "learning_rate": 4.014784611691109e-06,
+      "loss": 0.53,
+      "step": 6687
+    },
+    {
+      "epoch": 1.7755210407540156,
+      "grad_norm": 0.416260651993525,
+      "learning_rate": 4.014506855093952e-06,
+      "loss": 0.6065,
+      "step": 6688
+    },
+    {
+      "epoch": 1.7757865392273995,
+      "grad_norm": 0.4090105779456124,
+      "learning_rate": 4.014229068960205e-06,
+      "loss": 0.5788,
+      "step": 6689
+    },
+    {
+      "epoch": 1.7760520377007833,
+      "grad_norm": 0.40350553304500164,
+      "learning_rate": 4.013951253295283e-06,
+      "loss": 0.6055,
+      "step": 6690
+    },
+    {
+      "epoch": 1.776317536174167,
+      "grad_norm": 0.41113286065162047,
+      "learning_rate": 4.013673408104604e-06,
+      "loss": 0.6118,
+      "step": 6691
+    },
+    {
+      "epoch": 1.7765830346475506,
+      "grad_norm": 0.40203800212551655,
+      "learning_rate": 4.0133955333935884e-06,
+      "loss": 0.5919,
+      "step": 6692
+    },
+    {
+      "epoch": 1.7768485331209347,
+      "grad_norm": 0.40985005754769566,
+      "learning_rate": 4.013117629167653e-06,
+      "loss": 0.5653,
+      "step": 6693
+    },
+    {
+      "epoch": 1.7771140315943184,
+      "grad_norm": 0.3946564108519016,
+      "learning_rate": 4.012839695432221e-06,
+      "loss": 0.5996,
+      "step": 6694
+    },
+    {
+      "epoch": 1.777379530067702,
+      "grad_norm": 0.4115639074175485,
+      "learning_rate": 4.01256173219271e-06,
+      "loss": 0.5994,
+      "step": 6695
+    },
+    {
+      "epoch": 1.7776450285410859,
+      "grad_norm": 0.4175388887118231,
+      "learning_rate": 4.012283739454542e-06,
+      "loss": 0.6014,
+      "step": 6696
+    },
+    {
+      "epoch": 1.7779105270144697,
+      "grad_norm": 0.41057168848001685,
+      "learning_rate": 4.012005717223138e-06,
+      "loss": 0.5651,
+      "step": 6697
+    },
+    {
+      "epoch": 1.7781760254878534,
+      "grad_norm": 0.4199201558044919,
+      "learning_rate": 4.01172766550392e-06,
+      "loss": 0.5969,
+      "step": 6698
+    },
+    {
+      "epoch": 1.7784415239612372,
+      "grad_norm": 0.39833980333063823,
+      "learning_rate": 4.0114495843023125e-06,
+      "loss": 0.5796,
+      "step": 6699
+    },
+    {
+      "epoch": 1.778707022434621,
+      "grad_norm": 0.4185999555153697,
+      "learning_rate": 4.011171473623736e-06,
+      "loss": 0.5964,
+      "step": 6700
+    },
+    {
+      "epoch": 1.7789725209080047,
+      "grad_norm": 0.43427690545700415,
+      "learning_rate": 4.010893333473617e-06,
+      "loss": 0.5547,
+      "step": 6701
+    },
+    {
+      "epoch": 1.7792380193813886,
+      "grad_norm": 0.417941478982828,
+      "learning_rate": 4.010615163857379e-06,
+      "loss": 0.634,
+      "step": 6702
+    },
+    {
+      "epoch": 1.7795035178547725,
+      "grad_norm": 0.4144390171721843,
+      "learning_rate": 4.010336964780445e-06,
+      "loss": 0.5437,
+      "step": 6703
+    },
+    {
+      "epoch": 1.779769016328156,
+      "grad_norm": 0.43920459699861164,
+      "learning_rate": 4.0100587362482435e-06,
+      "loss": 0.5964,
+      "step": 6704
+    },
+    {
+      "epoch": 1.7800345148015397,
+      "grad_norm": 0.41801525737566153,
+      "learning_rate": 4.009780478266199e-06,
+      "loss": 0.5545,
+      "step": 6705
+    },
+    {
+      "epoch": 1.7803000132749238,
+      "grad_norm": 0.4114136717501727,
+      "learning_rate": 4.009502190839739e-06,
+      "loss": 0.5627,
+      "step": 6706
+    },
+    {
+      "epoch": 1.7805655117483075,
+      "grad_norm": 0.4067294335395838,
+      "learning_rate": 4.00922387397429e-06,
+      "loss": 0.5929,
+      "step": 6707
+    },
+    {
+      "epoch": 1.780831010221691,
+      "grad_norm": 0.43645275526314004,
+      "learning_rate": 4.008945527675281e-06,
+      "loss": 0.5867,
+      "step": 6708
+    },
+    {
+      "epoch": 1.781096508695075,
+      "grad_norm": 0.4289717337771052,
+      "learning_rate": 4.008667151948138e-06,
+      "loss": 0.578,
+      "step": 6709
+    },
+    {
+      "epoch": 1.7813620071684588,
+      "grad_norm": 0.41452371991451675,
+      "learning_rate": 4.008388746798293e-06,
+      "loss": 0.5771,
+      "step": 6710
+    },
+    {
+      "epoch": 1.7816275056418425,
+      "grad_norm": 0.4124151606917401,
+      "learning_rate": 4.008110312231175e-06,
+      "loss": 0.5485,
+      "step": 6711
+    },
+    {
+      "epoch": 1.7818930041152263,
+      "grad_norm": 0.43148046562706127,
+      "learning_rate": 4.007831848252212e-06,
+      "loss": 0.5779,
+      "step": 6712
+    },
+    {
+      "epoch": 1.7821585025886102,
+      "grad_norm": 0.41277836263795215,
+      "learning_rate": 4.007553354866835e-06,
+      "loss": 0.5757,
+      "step": 6713
+    },
+    {
+      "epoch": 1.7824240010619938,
+      "grad_norm": 0.39845628701136115,
+      "learning_rate": 4.007274832080479e-06,
+      "loss": 0.5624,
+      "step": 6714
+    },
+    {
+      "epoch": 1.7826894995353777,
+      "grad_norm": 0.4023912322385927,
+      "learning_rate": 4.006996279898572e-06,
+      "loss": 0.6067,
+      "step": 6715
+    },
+    {
+      "epoch": 1.7829549980087616,
+      "grad_norm": 0.4145733079155911,
+      "learning_rate": 4.006717698326548e-06,
+      "loss": 0.6021,
+      "step": 6716
+    },
+    {
+      "epoch": 1.7832204964821452,
+      "grad_norm": 0.4324421611238071,
+      "learning_rate": 4.00643908736984e-06,
+      "loss": 0.5844,
+      "step": 6717
+    },
+    {
+      "epoch": 1.7834859949555288,
+      "grad_norm": 0.4033355141476237,
+      "learning_rate": 4.006160447033881e-06,
+      "loss": 0.5926,
+      "step": 6718
+    },
+    {
+      "epoch": 1.783751493428913,
+      "grad_norm": 0.4090873830746304,
+      "learning_rate": 4.005881777324106e-06,
+      "loss": 0.5773,
+      "step": 6719
+    },
+    {
+      "epoch": 1.7840169919022966,
+      "grad_norm": 0.41451345115315874,
+      "learning_rate": 4.005603078245949e-06,
+      "loss": 0.6093,
+      "step": 6720
+    },
+    {
+      "epoch": 1.7842824903756802,
+      "grad_norm": 0.4176807508533954,
+      "learning_rate": 4.005324349804845e-06,
+      "loss": 0.5491,
+      "step": 6721
+    },
+    {
+      "epoch": 1.784547988849064,
+      "grad_norm": 0.41376046114487586,
+      "learning_rate": 4.005045592006232e-06,
+      "loss": 0.5884,
+      "step": 6722
+    },
+    {
+      "epoch": 1.784813487322448,
+      "grad_norm": 0.4144505435566118,
+      "learning_rate": 4.004766804855544e-06,
+      "loss": 0.5413,
+      "step": 6723
+    },
+    {
+      "epoch": 1.7850789857958316,
+      "grad_norm": 0.4212963988314677,
+      "learning_rate": 4.00448798835822e-06,
+      "loss": 0.5839,
+      "step": 6724
+    },
+    {
+      "epoch": 1.7853444842692154,
+      "grad_norm": 0.4156929550650738,
+      "learning_rate": 4.0042091425196946e-06,
+      "loss": 0.594,
+      "step": 6725
+    },
+    {
+      "epoch": 1.7856099827425993,
+      "grad_norm": 0.4231670247745961,
+      "learning_rate": 4.00393026734541e-06,
+      "loss": 0.5716,
+      "step": 6726
+    },
+    {
+      "epoch": 1.785875481215983,
+      "grad_norm": 0.4032690912634656,
+      "learning_rate": 4.003651362840803e-06,
+      "loss": 0.5574,
+      "step": 6727
+    },
+    {
+      "epoch": 1.7861409796893668,
+      "grad_norm": 0.41497692317316043,
+      "learning_rate": 4.003372429011312e-06,
+      "loss": 0.584,
+      "step": 6728
+    },
+    {
+      "epoch": 1.7864064781627507,
+      "grad_norm": 0.41700836564965743,
+      "learning_rate": 4.003093465862377e-06,
+      "loss": 0.5593,
+      "step": 6729
+    },
+    {
+      "epoch": 1.7866719766361343,
+      "grad_norm": 0.44354906312780357,
+      "learning_rate": 4.00281447339944e-06,
+      "loss": 0.5661,
+      "step": 6730
+    },
+    {
+      "epoch": 1.7869374751095182,
+      "grad_norm": 0.41971133136509464,
+      "learning_rate": 4.002535451627941e-06,
+      "loss": 0.5912,
+      "step": 6731
+    },
+    {
+      "epoch": 1.787202973582902,
+      "grad_norm": 0.4046616678096043,
+      "learning_rate": 4.002256400553323e-06,
+      "loss": 0.5963,
+      "step": 6732
+    },
+    {
+      "epoch": 1.7874684720562857,
+      "grad_norm": 0.4068544354994896,
+      "learning_rate": 4.001977320181025e-06,
+      "loss": 0.5511,
+      "step": 6733
+    },
+    {
+      "epoch": 1.7877339705296693,
+      "grad_norm": 0.4124776198168891,
+      "learning_rate": 4.001698210516494e-06,
+      "loss": 0.5772,
+      "step": 6734
+    },
+    {
+      "epoch": 1.7879994690030534,
+      "grad_norm": 0.4231214718807104,
+      "learning_rate": 4.001419071565169e-06,
+      "loss": 0.598,
+      "step": 6735
+    },
+    {
+      "epoch": 1.788264967476437,
+      "grad_norm": 0.40523243305685236,
+      "learning_rate": 4.001139903332497e-06,
+      "loss": 0.5751,
+      "step": 6736
+    },
+    {
+      "epoch": 1.7885304659498207,
+      "grad_norm": 0.4227323818227764,
+      "learning_rate": 4.000860705823921e-06,
+      "loss": 0.61,
+      "step": 6737
+    },
+    {
+      "epoch": 1.7887959644232045,
+      "grad_norm": 0.4103721905254452,
+      "learning_rate": 4.000581479044887e-06,
+      "loss": 0.5788,
+      "step": 6738
+    },
+    {
+      "epoch": 1.7890614628965884,
+      "grad_norm": 0.407688642094616,
+      "learning_rate": 4.00030222300084e-06,
+      "loss": 0.6018,
+      "step": 6739
+    },
+    {
+      "epoch": 1.789326961369972,
+      "grad_norm": 0.40338591151942643,
+      "learning_rate": 4.000022937697226e-06,
+      "loss": 0.5533,
+      "step": 6740
+    },
+    {
+      "epoch": 1.789592459843356,
+      "grad_norm": 0.4125465844528528,
+      "learning_rate": 3.999743623139493e-06,
+      "loss": 0.5632,
+      "step": 6741
+    },
+    {
+      "epoch": 1.7898579583167398,
+      "grad_norm": 0.4202690967650945,
+      "learning_rate": 3.9994642793330865e-06,
+      "loss": 0.5664,
+      "step": 6742
+    },
+    {
+      "epoch": 1.7901234567901234,
+      "grad_norm": 0.41986284840292776,
+      "learning_rate": 3.999184906283455e-06,
+      "loss": 0.5673,
+      "step": 6743
+    },
+    {
+      "epoch": 1.7903889552635073,
+      "grad_norm": 0.4233748241497013,
+      "learning_rate": 3.998905503996048e-06,
+      "loss": 0.6221,
+      "step": 6744
+    },
+    {
+      "epoch": 1.7906544537368911,
+      "grad_norm": 0.41751933140947617,
+      "learning_rate": 3.998626072476314e-06,
+      "loss": 0.6071,
+      "step": 6745
+    },
+    {
+      "epoch": 1.7909199522102748,
+      "grad_norm": 0.401131030236606,
+      "learning_rate": 3.998346611729701e-06,
+      "loss": 0.5586,
+      "step": 6746
+    },
+    {
+      "epoch": 1.7911854506836584,
+      "grad_norm": 0.408037665025668,
+      "learning_rate": 3.998067121761661e-06,
+      "loss": 0.5597,
+      "step": 6747
+    },
+    {
+      "epoch": 1.7914509491570425,
+      "grad_norm": 0.4109719261712508,
+      "learning_rate": 3.997787602577644e-06,
+      "loss": 0.5689,
+      "step": 6748
+    },
+    {
+      "epoch": 1.7917164476304261,
+      "grad_norm": 0.4191907985438587,
+      "learning_rate": 3.997508054183102e-06,
+      "loss": 0.5607,
+      "step": 6749
+    },
+    {
+      "epoch": 1.7919819461038098,
+      "grad_norm": 0.4025224186409604,
+      "learning_rate": 3.997228476583487e-06,
+      "loss": 0.6024,
+      "step": 6750
+    },
+    {
+      "epoch": 1.7922474445771936,
+      "grad_norm": 0.42400758948127865,
+      "learning_rate": 3.99694886978425e-06,
+      "loss": 0.605,
+      "step": 6751
+    },
+    {
+      "epoch": 1.7925129430505775,
+      "grad_norm": 0.4088776238221803,
+      "learning_rate": 3.996669233790845e-06,
+      "loss": 0.5648,
+      "step": 6752
+    },
+    {
+      "epoch": 1.7927784415239612,
+      "grad_norm": 0.4005017286578978,
+      "learning_rate": 3.996389568608725e-06,
+      "loss": 0.5738,
+      "step": 6753
+    },
+    {
+      "epoch": 1.793043939997345,
+      "grad_norm": 0.409950117296982,
+      "learning_rate": 3.9961098742433456e-06,
+      "loss": 0.5992,
+      "step": 6754
+    },
+    {
+      "epoch": 1.7933094384707289,
+      "grad_norm": 0.4094387153339892,
+      "learning_rate": 3.99583015070016e-06,
+      "loss": 0.5652,
+      "step": 6755
+    },
+    {
+      "epoch": 1.7935749369441125,
+      "grad_norm": 0.4160710863725598,
+      "learning_rate": 3.995550397984624e-06,
+      "loss": 0.5789,
+      "step": 6756
+    },
+    {
+      "epoch": 1.7938404354174964,
+      "grad_norm": 0.41428489369513855,
+      "learning_rate": 3.9952706161021935e-06,
+      "loss": 0.5942,
+      "step": 6757
+    },
+    {
+      "epoch": 1.7941059338908802,
+      "grad_norm": 0.4239792117802188,
+      "learning_rate": 3.994990805058325e-06,
+      "loss": 0.5833,
+      "step": 6758
+    },
+    {
+      "epoch": 1.7943714323642639,
+      "grad_norm": 0.42909582328902346,
+      "learning_rate": 3.994710964858475e-06,
+      "loss": 0.602,
+      "step": 6759
+    },
+    {
+      "epoch": 1.7946369308376475,
+      "grad_norm": 0.40918627492853654,
+      "learning_rate": 3.994431095508102e-06,
+      "loss": 0.5934,
+      "step": 6760
+    },
+    {
+      "epoch": 1.7949024293110316,
+      "grad_norm": 0.408092007214423,
+      "learning_rate": 3.994151197012664e-06,
+      "loss": 0.5941,
+      "step": 6761
+    },
+    {
+      "epoch": 1.7951679277844153,
+      "grad_norm": 0.4061522536338927,
+      "learning_rate": 3.993871269377619e-06,
+      "loss": 0.5774,
+      "step": 6762
+    },
+    {
+      "epoch": 1.795433426257799,
+      "grad_norm": 0.410127219541861,
+      "learning_rate": 3.993591312608427e-06,
+      "loss": 0.5948,
+      "step": 6763
+    },
+    {
+      "epoch": 1.7956989247311828,
+      "grad_norm": 0.39740529421018345,
+      "learning_rate": 3.9933113267105465e-06,
+      "loss": 0.5614,
+      "step": 6764
+    },
+    {
+      "epoch": 1.7959644232045666,
+      "grad_norm": 0.41163835686945005,
+      "learning_rate": 3.993031311689439e-06,
+      "loss": 0.5817,
+      "step": 6765
+    },
+    {
+      "epoch": 1.7962299216779503,
+      "grad_norm": 0.4048328335509438,
+      "learning_rate": 3.992751267550567e-06,
+      "loss": 0.549,
+      "step": 6766
+    },
+    {
+      "epoch": 1.7964954201513341,
+      "grad_norm": 0.4157182148542579,
+      "learning_rate": 3.992471194299388e-06,
+      "loss": 0.5716,
+      "step": 6767
+    },
+    {
+      "epoch": 1.796760918624718,
+      "grad_norm": 0.43080367008851167,
+      "learning_rate": 3.9921910919413675e-06,
+      "loss": 0.5826,
+      "step": 6768
+    },
+    {
+      "epoch": 1.7970264170981016,
+      "grad_norm": 0.40936816547465427,
+      "learning_rate": 3.9919109604819675e-06,
+      "loss": 0.6329,
+      "step": 6769
+    },
+    {
+      "epoch": 1.7972919155714855,
+      "grad_norm": 0.41647967557843796,
+      "learning_rate": 3.991630799926651e-06,
+      "loss": 0.5814,
+      "step": 6770
+    },
+    {
+      "epoch": 1.7975574140448694,
+      "grad_norm": 0.415466295232353,
+      "learning_rate": 3.991350610280881e-06,
+      "loss": 0.5805,
+      "step": 6771
+    },
+    {
+      "epoch": 1.797822912518253,
+      "grad_norm": 0.4173964290795593,
+      "learning_rate": 3.991070391550124e-06,
+      "loss": 0.5932,
+      "step": 6772
+    },
+    {
+      "epoch": 1.7980884109916366,
+      "grad_norm": 0.3989217392871118,
+      "learning_rate": 3.990790143739842e-06,
+      "loss": 0.5892,
+      "step": 6773
+    },
+    {
+      "epoch": 1.7983539094650207,
+      "grad_norm": 0.42301929733052934,
+      "learning_rate": 3.9905098668555034e-06,
+      "loss": 0.5827,
+      "step": 6774
+    },
+    {
+      "epoch": 1.7986194079384044,
+      "grad_norm": 0.4186676923609016,
+      "learning_rate": 3.990229560902573e-06,
+      "loss": 0.5876,
+      "step": 6775
+    },
+    {
+      "epoch": 1.798884906411788,
+      "grad_norm": 0.40775126557563107,
+      "learning_rate": 3.989949225886516e-06,
+      "loss": 0.5896,
+      "step": 6776
+    },
+    {
+      "epoch": 1.7991504048851719,
+      "grad_norm": 0.4082324733188587,
+      "learning_rate": 3.989668861812802e-06,
+      "loss": 0.6246,
+      "step": 6777
+    },
+    {
+      "epoch": 1.7994159033585557,
+      "grad_norm": 0.40622031691397353,
+      "learning_rate": 3.989388468686898e-06,
+      "loss": 0.5561,
+      "step": 6778
+    },
+    {
+      "epoch": 1.7996814018319394,
+      "grad_norm": 0.4134371356650234,
+      "learning_rate": 3.989108046514272e-06,
+      "loss": 0.5888,
+      "step": 6779
+    },
+    {
+      "epoch": 1.7999469003053232,
+      "grad_norm": 0.41897793178465764,
+      "learning_rate": 3.988827595300394e-06,
+      "loss": 0.6024,
+      "step": 6780
+    },
+    {
+      "epoch": 1.800212398778707,
+      "grad_norm": 0.40808028101470173,
+      "learning_rate": 3.988547115050731e-06,
+      "loss": 0.6041,
+      "step": 6781
+    },
+    {
+      "epoch": 1.8004778972520907,
+      "grad_norm": 0.398651550143708,
+      "learning_rate": 3.988266605770756e-06,
+      "loss": 0.6135,
+      "step": 6782
+    },
+    {
+      "epoch": 1.8007433957254746,
+      "grad_norm": 0.4260135063692521,
+      "learning_rate": 3.987986067465938e-06,
+      "loss": 0.6182,
+      "step": 6783
+    },
+    {
+      "epoch": 1.8010088941988585,
+      "grad_norm": 0.4243534815522036,
+      "learning_rate": 3.9877055001417476e-06,
+      "loss": 0.5587,
+      "step": 6784
+    },
+    {
+      "epoch": 1.801274392672242,
+      "grad_norm": 0.4097221488500463,
+      "learning_rate": 3.9874249038036585e-06,
+      "loss": 0.5204,
+      "step": 6785
+    },
+    {
+      "epoch": 1.801539891145626,
+      "grad_norm": 0.39020504548859486,
+      "learning_rate": 3.987144278457141e-06,
+      "loss": 0.581,
+      "step": 6786
+    },
+    {
+      "epoch": 1.8018053896190098,
+      "grad_norm": 0.4252682499755787,
+      "learning_rate": 3.98686362410767e-06,
+      "loss": 0.6026,
+      "step": 6787
+    },
+    {
+      "epoch": 1.8020708880923935,
+      "grad_norm": 0.4149469807642028,
+      "learning_rate": 3.986582940760717e-06,
+      "loss": 0.5668,
+      "step": 6788
+    },
+    {
+      "epoch": 1.802336386565777,
+      "grad_norm": 0.4623670980457604,
+      "learning_rate": 3.9863022284217575e-06,
+      "loss": 0.5974,
+      "step": 6789
+    },
+    {
+      "epoch": 1.8026018850391612,
+      "grad_norm": 0.41451994504363865,
+      "learning_rate": 3.986021487096265e-06,
+      "loss": 0.571,
+      "step": 6790
+    },
+    {
+      "epoch": 1.8028673835125448,
+      "grad_norm": 0.422431533507994,
+      "learning_rate": 3.985740716789715e-06,
+      "loss": 0.5804,
+      "step": 6791
+    },
+    {
+      "epoch": 1.8031328819859285,
+      "grad_norm": 0.4829477816671888,
+      "learning_rate": 3.985459917507584e-06,
+      "loss": 0.571,
+      "step": 6792
+    },
+    {
+      "epoch": 1.8033983804593123,
+      "grad_norm": 0.4825420029996147,
+      "learning_rate": 3.985179089255347e-06,
+      "loss": 0.5519,
+      "step": 6793
+    },
+    {
+      "epoch": 1.8036638789326962,
+      "grad_norm": 0.4076926917949993,
+      "learning_rate": 3.984898232038481e-06,
+      "loss": 0.5867,
+      "step": 6794
+    },
+    {
+      "epoch": 1.8039293774060798,
+      "grad_norm": 0.43630892186683706,
+      "learning_rate": 3.984617345862465e-06,
+      "loss": 0.6022,
+      "step": 6795
+    },
+    {
+      "epoch": 1.8041948758794637,
+      "grad_norm": 0.42421521316490074,
+      "learning_rate": 3.9843364307327745e-06,
+      "loss": 0.567,
+      "step": 6796
+    },
+    {
+      "epoch": 1.8044603743528476,
+      "grad_norm": 0.41776301918404113,
+      "learning_rate": 3.98405548665489e-06,
+      "loss": 0.6336,
+      "step": 6797
+    },
+    {
+      "epoch": 1.8047258728262312,
+      "grad_norm": 0.413140844851692,
+      "learning_rate": 3.983774513634289e-06,
+      "loss": 0.5809,
+      "step": 6798
+    },
+    {
+      "epoch": 1.804991371299615,
+      "grad_norm": 0.41053898584161114,
+      "learning_rate": 3.983493511676453e-06,
+      "loss": 0.596,
+      "step": 6799
+    },
+    {
+      "epoch": 1.805256869772999,
+      "grad_norm": 0.40652542068363795,
+      "learning_rate": 3.983212480786861e-06,
+      "loss": 0.5409,
+      "step": 6800
+    },
+    {
+      "epoch": 1.8055223682463826,
+      "grad_norm": 0.4080604089336657,
+      "learning_rate": 3.982931420970995e-06,
+      "loss": 0.5672,
+      "step": 6801
+    },
+    {
+      "epoch": 1.8057878667197662,
+      "grad_norm": 0.4071026067531037,
+      "learning_rate": 3.982650332234335e-06,
+      "loss": 0.599,
+      "step": 6802
+    },
+    {
+      "epoch": 1.8060533651931503,
+      "grad_norm": 0.41925125987130984,
+      "learning_rate": 3.982369214582362e-06,
+      "loss": 0.5809,
+      "step": 6803
+    },
+    {
+      "epoch": 1.806318863666534,
+      "grad_norm": 0.40424297393236863,
+      "learning_rate": 3.982088068020562e-06,
+      "loss": 0.5716,
+      "step": 6804
+    },
+    {
+      "epoch": 1.8065843621399176,
+      "grad_norm": 0.41457434040207436,
+      "learning_rate": 3.981806892554414e-06,
+      "loss": 0.6056,
+      "step": 6805
+    },
+    {
+      "epoch": 1.8068498606133014,
+      "grad_norm": 0.42759855334938374,
+      "learning_rate": 3.981525688189405e-06,
+      "loss": 0.5938,
+      "step": 6806
+    },
+    {
+      "epoch": 1.8071153590866853,
+      "grad_norm": 0.41430977852499373,
+      "learning_rate": 3.981244454931017e-06,
+      "loss": 0.5638,
+      "step": 6807
+    },
+    {
+      "epoch": 1.807380857560069,
+      "grad_norm": 0.4059039155523686,
+      "learning_rate": 3.9809631927847345e-06,
+      "loss": 0.583,
+      "step": 6808
+    },
+    {
+      "epoch": 1.8076463560334528,
+      "grad_norm": 0.4081727818607107,
+      "learning_rate": 3.980681901756046e-06,
+      "loss": 0.6182,
+      "step": 6809
+    },
+    {
+      "epoch": 1.8079118545068367,
+      "grad_norm": 0.4160574014519053,
+      "learning_rate": 3.980400581850433e-06,
+      "loss": 0.5759,
+      "step": 6810
+    },
+    {
+      "epoch": 1.8081773529802203,
+      "grad_norm": 0.40529282691449536,
+      "learning_rate": 3.980119233073384e-06,
+      "loss": 0.5634,
+      "step": 6811
+    },
+    {
+      "epoch": 1.8084428514536042,
+      "grad_norm": 0.42052593207965144,
+      "learning_rate": 3.979837855430387e-06,
+      "loss": 0.5583,
+      "step": 6812
+    },
+    {
+      "epoch": 1.808708349926988,
+      "grad_norm": 0.40796553657244394,
+      "learning_rate": 3.979556448926929e-06,
+      "loss": 0.5914,
+      "step": 6813
+    },
+    {
+      "epoch": 1.8089738484003717,
+      "grad_norm": 0.4024568507645281,
+      "learning_rate": 3.979275013568496e-06,
+      "loss": 0.5751,
+      "step": 6814
+    },
+    {
+      "epoch": 1.8092393468737553,
+      "grad_norm": 0.41620334266469783,
+      "learning_rate": 3.97899354936058e-06,
+      "loss": 0.5919,
+      "step": 6815
+    },
+    {
+      "epoch": 1.8095048453471394,
+      "grad_norm": 0.41714351909591146,
+      "learning_rate": 3.9787120563086675e-06,
+      "loss": 0.5854,
+      "step": 6816
+    },
+    {
+      "epoch": 1.809770343820523,
+      "grad_norm": 0.4140405639599087,
+      "learning_rate": 3.97843053441825e-06,
+      "loss": 0.5868,
+      "step": 6817
+    },
+    {
+      "epoch": 1.8100358422939067,
+      "grad_norm": 0.4044607528187589,
+      "learning_rate": 3.978148983694817e-06,
+      "loss": 0.5555,
+      "step": 6818
+    },
+    {
+      "epoch": 1.8103013407672905,
+      "grad_norm": 0.4138167916681698,
+      "learning_rate": 3.9778674041438605e-06,
+      "loss": 0.5706,
+      "step": 6819
+    },
+    {
+      "epoch": 1.8105668392406744,
+      "grad_norm": 0.4139509457082379,
+      "learning_rate": 3.9775857957708695e-06,
+      "loss": 0.572,
+      "step": 6820
+    },
+    {
+      "epoch": 1.810832337714058,
+      "grad_norm": 0.3952554492369363,
+      "learning_rate": 3.977304158581339e-06,
+      "loss": 0.5733,
+      "step": 6821
+    },
+    {
+      "epoch": 1.811097836187442,
+      "grad_norm": 0.4216797146380888,
+      "learning_rate": 3.97702249258076e-06,
+      "loss": 0.6087,
+      "step": 6822
+    },
+    {
+      "epoch": 1.8113633346608258,
+      "grad_norm": 0.40768296117858194,
+      "learning_rate": 3.976740797774626e-06,
+      "loss": 0.5728,
+      "step": 6823
+    },
+    {
+      "epoch": 1.8116288331342094,
+      "grad_norm": 0.4063751869081823,
+      "learning_rate": 3.97645907416843e-06,
+      "loss": 0.554,
+      "step": 6824
+    },
+    {
+      "epoch": 1.8118943316075933,
+      "grad_norm": 0.4096184952112304,
+      "learning_rate": 3.976177321767668e-06,
+      "loss": 0.5935,
+      "step": 6825
+    },
+    {
+      "epoch": 1.8121598300809771,
+      "grad_norm": 0.42072672693211105,
+      "learning_rate": 3.975895540577835e-06,
+      "loss": 0.5523,
+      "step": 6826
+    },
+    {
+      "epoch": 1.8124253285543608,
+      "grad_norm": 0.4165263028415445,
+      "learning_rate": 3.975613730604423e-06,
+      "loss": 0.5696,
+      "step": 6827
+    },
+    {
+      "epoch": 1.8126908270277446,
+      "grad_norm": 0.4202931615903253,
+      "learning_rate": 3.975331891852931e-06,
+      "loss": 0.591,
+      "step": 6828
+    },
+    {
+      "epoch": 1.8129563255011285,
+      "grad_norm": 0.406871159146028,
+      "learning_rate": 3.975050024328856e-06,
+      "loss": 0.5602,
+      "step": 6829
+    },
+    {
+      "epoch": 1.8132218239745121,
+      "grad_norm": 0.4085316530523902,
+      "learning_rate": 3.974768128037693e-06,
+      "loss": 0.5709,
+      "step": 6830
+    },
+    {
+      "epoch": 1.8134873224478958,
+      "grad_norm": 0.42312446102361884,
+      "learning_rate": 3.9744862029849404e-06,
+      "loss": 0.5622,
+      "step": 6831
+    },
+    {
+      "epoch": 1.8137528209212797,
+      "grad_norm": 0.42682122154259966,
+      "learning_rate": 3.974204249176098e-06,
+      "loss": 0.6021,
+      "step": 6832
+    },
+    {
+      "epoch": 1.8140183193946635,
+      "grad_norm": 0.42775275627796105,
+      "learning_rate": 3.973922266616662e-06,
+      "loss": 0.6152,
+      "step": 6833
+    },
+    {
+      "epoch": 1.8142838178680472,
+      "grad_norm": 0.4107629138001381,
+      "learning_rate": 3.973640255312134e-06,
+      "loss": 0.5763,
+      "step": 6834
+    },
+    {
+      "epoch": 1.814549316341431,
+      "grad_norm": 0.4122845505727306,
+      "learning_rate": 3.973358215268013e-06,
+      "loss": 0.6115,
+      "step": 6835
+    },
+    {
+      "epoch": 1.8148148148148149,
+      "grad_norm": 0.41054725684345034,
+      "learning_rate": 3.973076146489798e-06,
+      "loss": 0.6355,
+      "step": 6836
+    },
+    {
+      "epoch": 1.8150803132881985,
+      "grad_norm": 0.4168830083094273,
+      "learning_rate": 3.972794048982992e-06,
+      "loss": 0.5602,
+      "step": 6837
+    },
+    {
+      "epoch": 1.8153458117615824,
+      "grad_norm": 0.40859993742626444,
+      "learning_rate": 3.9725119227530965e-06,
+      "loss": 0.5994,
+      "step": 6838
+    },
+    {
+      "epoch": 1.8156113102349662,
+      "grad_norm": 0.40494888425964376,
+      "learning_rate": 3.972229767805613e-06,
+      "loss": 0.5842,
+      "step": 6839
+    },
+    {
+      "epoch": 1.8158768087083499,
+      "grad_norm": 0.4060388916853308,
+      "learning_rate": 3.9719475841460444e-06,
+      "loss": 0.5736,
+      "step": 6840
+    },
+    {
+      "epoch": 1.8161423071817338,
+      "grad_norm": 0.39772669619632506,
+      "learning_rate": 3.971665371779894e-06,
+      "loss": 0.6278,
+      "step": 6841
+    },
+    {
+      "epoch": 1.8164078056551176,
+      "grad_norm": 0.40206305843109685,
+      "learning_rate": 3.971383130712665e-06,
+      "loss": 0.5825,
+      "step": 6842
+    },
+    {
+      "epoch": 1.8166733041285013,
+      "grad_norm": 0.41548644846716803,
+      "learning_rate": 3.971100860949863e-06,
+      "loss": 0.5862,
+      "step": 6843
+    },
+    {
+      "epoch": 1.816938802601885,
+      "grad_norm": 0.42362314828560854,
+      "learning_rate": 3.970818562496993e-06,
+      "loss": 0.5783,
+      "step": 6844
+    },
+    {
+      "epoch": 1.817204301075269,
+      "grad_norm": 0.4076203223985314,
+      "learning_rate": 3.9705362353595585e-06,
+      "loss": 0.5686,
+      "step": 6845
+    },
+    {
+      "epoch": 1.8174697995486526,
+      "grad_norm": 0.42422761470994286,
+      "learning_rate": 3.970253879543067e-06,
+      "loss": 0.59,
+      "step": 6846
+    },
+    {
+      "epoch": 1.8177352980220363,
+      "grad_norm": 0.41515227694722934,
+      "learning_rate": 3.969971495053026e-06,
+      "loss": 0.534,
+      "step": 6847
+    },
+    {
+      "epoch": 1.8180007964954201,
+      "grad_norm": 0.40005750625970493,
+      "learning_rate": 3.969689081894941e-06,
+      "loss": 0.5336,
+      "step": 6848
+    },
+    {
+      "epoch": 1.818266294968804,
+      "grad_norm": 0.42563449381958635,
+      "learning_rate": 3.96940664007432e-06,
+      "loss": 0.5893,
+      "step": 6849
+    },
+    {
+      "epoch": 1.8185317934421876,
+      "grad_norm": 0.40709190852716304,
+      "learning_rate": 3.9691241695966735e-06,
+      "loss": 0.5901,
+      "step": 6850
+    },
+    {
+      "epoch": 1.8187972919155715,
+      "grad_norm": 0.4208219527175216,
+      "learning_rate": 3.968841670467507e-06,
+      "loss": 0.603,
+      "step": 6851
+    },
+    {
+      "epoch": 1.8190627903889554,
+      "grad_norm": 0.42438763403004487,
+      "learning_rate": 3.968559142692333e-06,
+      "loss": 0.6115,
+      "step": 6852
+    },
+    {
+      "epoch": 1.819328288862339,
+      "grad_norm": 0.4444673240053476,
+      "learning_rate": 3.968276586276659e-06,
+      "loss": 0.5403,
+      "step": 6853
+    },
+    {
+      "epoch": 1.8195937873357229,
+      "grad_norm": 0.40843547394315566,
+      "learning_rate": 3.9679940012259965e-06,
+      "loss": 0.5718,
+      "step": 6854
+    },
+    {
+      "epoch": 1.8198592858091067,
+      "grad_norm": 0.40557208046900034,
+      "learning_rate": 3.967711387545858e-06,
+      "loss": 0.5759,
+      "step": 6855
+    },
+    {
+      "epoch": 1.8201247842824904,
+      "grad_norm": 0.43608630065980597,
+      "learning_rate": 3.967428745241752e-06,
+      "loss": 0.545,
+      "step": 6856
+    },
+    {
+      "epoch": 1.820390282755874,
+      "grad_norm": 0.4456599856711164,
+      "learning_rate": 3.967146074319194e-06,
+      "loss": 0.6156,
+      "step": 6857
+    },
+    {
+      "epoch": 1.820655781229258,
+      "grad_norm": 0.41807563761382394,
+      "learning_rate": 3.966863374783695e-06,
+      "loss": 0.5844,
+      "step": 6858
+    },
+    {
+      "epoch": 1.8209212797026417,
+      "grad_norm": 0.42270989477356485,
+      "learning_rate": 3.966580646640768e-06,
+      "loss": 0.5792,
+      "step": 6859
+    },
+    {
+      "epoch": 1.8211867781760254,
+      "grad_norm": 0.4205151943187361,
+      "learning_rate": 3.966297889895929e-06,
+      "loss": 0.5996,
+      "step": 6860
+    },
+    {
+      "epoch": 1.8214522766494092,
+      "grad_norm": 0.4435206697567981,
+      "learning_rate": 3.96601510455469e-06,
+      "loss": 0.6189,
+      "step": 6861
+    },
+    {
+      "epoch": 1.821717775122793,
+      "grad_norm": 0.42922459160521326,
+      "learning_rate": 3.965732290622567e-06,
+      "loss": 0.5643,
+      "step": 6862
+    },
+    {
+      "epoch": 1.8219832735961767,
+      "grad_norm": 0.4166631525914852,
+      "learning_rate": 3.965449448105076e-06,
+      "loss": 0.619,
+      "step": 6863
+    },
+    {
+      "epoch": 1.8222487720695606,
+      "grad_norm": 0.41910531231153036,
+      "learning_rate": 3.965166577007733e-06,
+      "loss": 0.618,
+      "step": 6864
+    },
+    {
+      "epoch": 1.8225142705429445,
+      "grad_norm": 0.416901957180022,
+      "learning_rate": 3.9648836773360535e-06,
+      "loss": 0.6001,
+      "step": 6865
+    },
+    {
+      "epoch": 1.822779769016328,
+      "grad_norm": 0.41027350467885454,
+      "learning_rate": 3.9646007490955565e-06,
+      "loss": 0.5964,
+      "step": 6866
+    },
+    {
+      "epoch": 1.823045267489712,
+      "grad_norm": 0.42278800924118876,
+      "learning_rate": 3.9643177922917584e-06,
+      "loss": 0.5577,
+      "step": 6867
+    },
+    {
+      "epoch": 1.8233107659630958,
+      "grad_norm": 0.4649399626440613,
+      "learning_rate": 3.9640348069301785e-06,
+      "loss": 0.5729,
+      "step": 6868
+    },
+    {
+      "epoch": 1.8235762644364795,
+      "grad_norm": 0.4193467019799003,
+      "learning_rate": 3.963751793016335e-06,
+      "loss": 0.5742,
+      "step": 6869
+    },
+    {
+      "epoch": 1.823841762909863,
+      "grad_norm": 0.4285518145735066,
+      "learning_rate": 3.963468750555748e-06,
+      "loss": 0.5665,
+      "step": 6870
+    },
+    {
+      "epoch": 1.8241072613832472,
+      "grad_norm": 0.4425197104083943,
+      "learning_rate": 3.963185679553936e-06,
+      "loss": 0.6098,
+      "step": 6871
+    },
+    {
+      "epoch": 1.8243727598566308,
+      "grad_norm": 0.44767889063167665,
+      "learning_rate": 3.962902580016422e-06,
+      "loss": 0.5552,
+      "step": 6872
+    },
+    {
+      "epoch": 1.8246382583300145,
+      "grad_norm": 0.4129172959137986,
+      "learning_rate": 3.962619451948726e-06,
+      "loss": 0.5912,
+      "step": 6873
+    },
+    {
+      "epoch": 1.8249037568033983,
+      "grad_norm": 0.4123136336702218,
+      "learning_rate": 3.962336295356368e-06,
+      "loss": 0.618,
+      "step": 6874
+    },
+    {
+      "epoch": 1.8251692552767822,
+      "grad_norm": 0.4021082096493839,
+      "learning_rate": 3.962053110244873e-06,
+      "loss": 0.574,
+      "step": 6875
+    },
+    {
+      "epoch": 1.8254347537501658,
+      "grad_norm": 0.4737960529429846,
+      "learning_rate": 3.961769896619763e-06,
+      "loss": 0.607,
+      "step": 6876
+    },
+    {
+      "epoch": 1.8257002522235497,
+      "grad_norm": 0.41459691231585427,
+      "learning_rate": 3.9614866544865605e-06,
+      "loss": 0.5959,
+      "step": 6877
+    },
+    {
+      "epoch": 1.8259657506969336,
+      "grad_norm": 0.41804812351626613,
+      "learning_rate": 3.96120338385079e-06,
+      "loss": 0.6043,
+      "step": 6878
+    },
+    {
+      "epoch": 1.8262312491703172,
+      "grad_norm": 0.411606232216746,
+      "learning_rate": 3.9609200847179765e-06,
+      "loss": 0.5671,
+      "step": 6879
+    },
+    {
+      "epoch": 1.826496747643701,
+      "grad_norm": 0.4741741936047053,
+      "learning_rate": 3.960636757093643e-06,
+      "loss": 0.6091,
+      "step": 6880
+    },
+    {
+      "epoch": 1.826762246117085,
+      "grad_norm": 0.4204405750069081,
+      "learning_rate": 3.960353400983317e-06,
+      "loss": 0.5885,
+      "step": 6881
+    },
+    {
+      "epoch": 1.8270277445904686,
+      "grad_norm": 0.4248732639820394,
+      "learning_rate": 3.9600700163925245e-06,
+      "loss": 0.5854,
+      "step": 6882
+    },
+    {
+      "epoch": 1.8272932430638524,
+      "grad_norm": 0.39879096356030924,
+      "learning_rate": 3.959786603326792e-06,
+      "loss": 0.583,
+      "step": 6883
+    },
+    {
+      "epoch": 1.8275587415372363,
+      "grad_norm": 0.4032168028135756,
+      "learning_rate": 3.959503161791645e-06,
+      "loss": 0.6018,
+      "step": 6884
+    },
+    {
+      "epoch": 1.82782424001062,
+      "grad_norm": 0.4428247410526083,
+      "learning_rate": 3.959219691792614e-06,
+      "loss": 0.6365,
+      "step": 6885
+    },
+    {
+      "epoch": 1.8280897384840036,
+      "grad_norm": 0.4302132972119312,
+      "learning_rate": 3.958936193335227e-06,
+      "loss": 0.565,
+      "step": 6886
+    },
+    {
+      "epoch": 1.8283552369573874,
+      "grad_norm": 0.42115244794776907,
+      "learning_rate": 3.958652666425011e-06,
+      "loss": 0.5825,
+      "step": 6887
+    },
+    {
+      "epoch": 1.8286207354307713,
+      "grad_norm": 0.3989890541183796,
+      "learning_rate": 3.958369111067497e-06,
+      "loss": 0.5671,
+      "step": 6888
+    },
+    {
+      "epoch": 1.828886233904155,
+      "grad_norm": 0.3989471951017054,
+      "learning_rate": 3.958085527268215e-06,
+      "loss": 0.566,
+      "step": 6889
+    },
+    {
+      "epoch": 1.8291517323775388,
+      "grad_norm": 0.431697655434134,
+      "learning_rate": 3.957801915032695e-06,
+      "loss": 0.5799,
+      "step": 6890
+    },
+    {
+      "epoch": 1.8294172308509227,
+      "grad_norm": 0.42087917328398955,
+      "learning_rate": 3.957518274366468e-06,
+      "loss": 0.5698,
+      "step": 6891
+    },
+    {
+      "epoch": 1.8296827293243063,
+      "grad_norm": 0.4078086518688895,
+      "learning_rate": 3.957234605275066e-06,
+      "loss": 0.5999,
+      "step": 6892
+    },
+    {
+      "epoch": 1.8299482277976902,
+      "grad_norm": 0.4143377134359836,
+      "learning_rate": 3.9569509077640215e-06,
+      "loss": 0.5829,
+      "step": 6893
+    },
+    {
+      "epoch": 1.830213726271074,
+      "grad_norm": 0.40664548226008695,
+      "learning_rate": 3.956667181838867e-06,
+      "loss": 0.602,
+      "step": 6894
+    },
+    {
+      "epoch": 1.8304792247444577,
+      "grad_norm": 0.4072543374160619,
+      "learning_rate": 3.956383427505136e-06,
+      "loss": 0.5962,
+      "step": 6895
+    },
+    {
+      "epoch": 1.8307447232178415,
+      "grad_norm": 0.4132660895429606,
+      "learning_rate": 3.956099644768363e-06,
+      "loss": 0.5885,
+      "step": 6896
+    },
+    {
+      "epoch": 1.8310102216912254,
+      "grad_norm": 0.4123343765840562,
+      "learning_rate": 3.955815833634081e-06,
+      "loss": 0.5859,
+      "step": 6897
+    },
+    {
+      "epoch": 1.831275720164609,
+      "grad_norm": 0.409588982080439,
+      "learning_rate": 3.955531994107825e-06,
+      "loss": 0.548,
+      "step": 6898
+    },
+    {
+      "epoch": 1.8315412186379927,
+      "grad_norm": 0.4056604959574261,
+      "learning_rate": 3.955248126195133e-06,
+      "loss": 0.5578,
+      "step": 6899
+    },
+    {
+      "epoch": 1.8318067171113768,
+      "grad_norm": 0.4120319183354142,
+      "learning_rate": 3.954964229901539e-06,
+      "loss": 0.5953,
+      "step": 6900
+    },
+    {
+      "epoch": 1.8320722155847604,
+      "grad_norm": 0.40359523613022386,
+      "learning_rate": 3.95468030523258e-06,
+      "loss": 0.575,
+      "step": 6901
+    },
+    {
+      "epoch": 1.832337714058144,
+      "grad_norm": 0.4114259200271428,
+      "learning_rate": 3.954396352193792e-06,
+      "loss": 0.6483,
+      "step": 6902
+    },
+    {
+      "epoch": 1.832603212531528,
+      "grad_norm": 0.41812609480393964,
+      "learning_rate": 3.954112370790716e-06,
+      "loss": 0.5516,
+      "step": 6903
+    },
+    {
+      "epoch": 1.8328687110049118,
+      "grad_norm": 0.42278747761979335,
+      "learning_rate": 3.953828361028888e-06,
+      "loss": 0.5683,
+      "step": 6904
+    },
+    {
+      "epoch": 1.8331342094782954,
+      "grad_norm": 0.4145092316195848,
+      "learning_rate": 3.953544322913847e-06,
+      "loss": 0.5647,
+      "step": 6905
+    },
+    {
+      "epoch": 1.8333997079516793,
+      "grad_norm": 0.4269310386880556,
+      "learning_rate": 3.953260256451133e-06,
+      "loss": 0.61,
+      "step": 6906
+    },
+    {
+      "epoch": 1.8336652064250631,
+      "grad_norm": 0.409618894668987,
+      "learning_rate": 3.952976161646286e-06,
+      "loss": 0.5804,
+      "step": 6907
+    },
+    {
+      "epoch": 1.8339307048984468,
+      "grad_norm": 0.41819619456935553,
+      "learning_rate": 3.9526920385048465e-06,
+      "loss": 0.6074,
+      "step": 6908
+    },
+    {
+      "epoch": 1.8341962033718306,
+      "grad_norm": 0.43048840819700895,
+      "learning_rate": 3.952407887032355e-06,
+      "loss": 0.6119,
+      "step": 6909
+    },
+    {
+      "epoch": 1.8344617018452145,
+      "grad_norm": 0.3930225028037828,
+      "learning_rate": 3.952123707234354e-06,
+      "loss": 0.5589,
+      "step": 6910
+    },
+    {
+      "epoch": 1.8347272003185982,
+      "grad_norm": 0.4069389233453311,
+      "learning_rate": 3.951839499116384e-06,
+      "loss": 0.5952,
+      "step": 6911
+    },
+    {
+      "epoch": 1.8349926987919818,
+      "grad_norm": 0.4134311064961768,
+      "learning_rate": 3.95155526268399e-06,
+      "loss": 0.6046,
+      "step": 6912
+    },
+    {
+      "epoch": 1.8352581972653659,
+      "grad_norm": 0.41619515966378,
+      "learning_rate": 3.951270997942716e-06,
+      "loss": 0.6382,
+      "step": 6913
+    },
+    {
+      "epoch": 1.8355236957387495,
+      "grad_norm": 0.40929604319969626,
+      "learning_rate": 3.950986704898102e-06,
+      "loss": 0.5571,
+      "step": 6914
+    },
+    {
+      "epoch": 1.8357891942121332,
+      "grad_norm": 0.3946239114268539,
+      "learning_rate": 3.950702383555694e-06,
+      "loss": 0.5509,
+      "step": 6915
+    },
+    {
+      "epoch": 1.836054692685517,
+      "grad_norm": 0.4005028180358469,
+      "learning_rate": 3.95041803392104e-06,
+      "loss": 0.567,
+      "step": 6916
+    },
+    {
+      "epoch": 1.8363201911589009,
+      "grad_norm": 0.414259523949154,
+      "learning_rate": 3.950133655999681e-06,
+      "loss": 0.6307,
+      "step": 6917
+    },
+    {
+      "epoch": 1.8365856896322845,
+      "grad_norm": 0.4139833930567431,
+      "learning_rate": 3.9498492497971665e-06,
+      "loss": 0.5957,
+      "step": 6918
+    },
+    {
+      "epoch": 1.8368511881056684,
+      "grad_norm": 0.3996265316737662,
+      "learning_rate": 3.949564815319043e-06,
+      "loss": 0.5743,
+      "step": 6919
+    },
+    {
+      "epoch": 1.8371166865790522,
+      "grad_norm": 0.4173874486594135,
+      "learning_rate": 3.949280352570854e-06,
+      "loss": 0.599,
+      "step": 6920
+    },
+    {
+      "epoch": 1.837382185052436,
+      "grad_norm": 0.4189102319089356,
+      "learning_rate": 3.948995861558151e-06,
+      "loss": 0.6028,
+      "step": 6921
+    },
+    {
+      "epoch": 1.8376476835258198,
+      "grad_norm": 0.40938282620613214,
+      "learning_rate": 3.94871134228648e-06,
+      "loss": 0.5929,
+      "step": 6922
+    },
+    {
+      "epoch": 1.8379131819992036,
+      "grad_norm": 0.40507414414913434,
+      "learning_rate": 3.948426794761392e-06,
+      "loss": 0.5866,
+      "step": 6923
+    },
+    {
+      "epoch": 1.8381786804725873,
+      "grad_norm": 0.4117766308302032,
+      "learning_rate": 3.948142218988434e-06,
+      "loss": 0.6118,
+      "step": 6924
+    },
+    {
+      "epoch": 1.838444178945971,
+      "grad_norm": 0.41251748518821607,
+      "learning_rate": 3.9478576149731575e-06,
+      "loss": 0.583,
+      "step": 6925
+    },
+    {
+      "epoch": 1.838709677419355,
+      "grad_norm": 0.4509170725551388,
+      "learning_rate": 3.947572982721112e-06,
+      "loss": 0.5789,
+      "step": 6926
+    },
+    {
+      "epoch": 1.8389751758927386,
+      "grad_norm": 0.4169512986291828,
+      "learning_rate": 3.94728832223785e-06,
+      "loss": 0.5827,
+      "step": 6927
+    },
+    {
+      "epoch": 1.8392406743661223,
+      "grad_norm": 0.41734339719976643,
+      "learning_rate": 3.947003633528922e-06,
+      "loss": 0.5764,
+      "step": 6928
+    },
+    {
+      "epoch": 1.8395061728395061,
+      "grad_norm": 0.42631442640618766,
+      "learning_rate": 3.94671891659988e-06,
+      "loss": 0.6186,
+      "step": 6929
+    },
+    {
+      "epoch": 1.83977167131289,
+      "grad_norm": 0.41159493904399747,
+      "learning_rate": 3.946434171456277e-06,
+      "loss": 0.585,
+      "step": 6930
+    },
+    {
+      "epoch": 1.8400371697862736,
+      "grad_norm": 0.4099996179284077,
+      "learning_rate": 3.946149398103666e-06,
+      "loss": 0.6312,
+      "step": 6931
+    },
+    {
+      "epoch": 1.8403026682596575,
+      "grad_norm": 0.41920342230074054,
+      "learning_rate": 3.9458645965476005e-06,
+      "loss": 0.5982,
+      "step": 6932
+    },
+    {
+      "epoch": 1.8405681667330414,
+      "grad_norm": 0.410247917285299,
+      "learning_rate": 3.9455797667936365e-06,
+      "loss": 0.5704,
+      "step": 6933
+    },
+    {
+      "epoch": 1.840833665206425,
+      "grad_norm": 0.4026126901284573,
+      "learning_rate": 3.945294908847327e-06,
+      "loss": 0.6081,
+      "step": 6934
+    },
+    {
+      "epoch": 1.8410991636798089,
+      "grad_norm": 0.4020790591491528,
+      "learning_rate": 3.945010022714228e-06,
+      "loss": 0.589,
+      "step": 6935
+    },
+    {
+      "epoch": 1.8413646621531927,
+      "grad_norm": 0.4174677658329594,
+      "learning_rate": 3.944725108399896e-06,
+      "loss": 0.5861,
+      "step": 6936
+    },
+    {
+      "epoch": 1.8416301606265764,
+      "grad_norm": 0.41613573831651435,
+      "learning_rate": 3.944440165909886e-06,
+      "loss": 0.6048,
+      "step": 6937
+    },
+    {
+      "epoch": 1.8418956590999602,
+      "grad_norm": 0.405590446337733,
+      "learning_rate": 3.944155195249757e-06,
+      "loss": 0.5784,
+      "step": 6938
+    },
+    {
+      "epoch": 1.842161157573344,
+      "grad_norm": 0.4173842754338806,
+      "learning_rate": 3.943870196425066e-06,
+      "loss": 0.5831,
+      "step": 6939
+    },
+    {
+      "epoch": 1.8424266560467277,
+      "grad_norm": 0.41833973656606505,
+      "learning_rate": 3.943585169441371e-06,
+      "loss": 0.5818,
+      "step": 6940
+    },
+    {
+      "epoch": 1.8426921545201114,
+      "grad_norm": 0.4166588182995895,
+      "learning_rate": 3.94330011430423e-06,
+      "loss": 0.5665,
+      "step": 6941
+    },
+    {
+      "epoch": 1.8429576529934952,
+      "grad_norm": 0.4066253100791406,
+      "learning_rate": 3.943015031019204e-06,
+      "loss": 0.6069,
+      "step": 6942
+    },
+    {
+      "epoch": 1.843223151466879,
+      "grad_norm": 0.4158627441513552,
+      "learning_rate": 3.942729919591851e-06,
+      "loss": 0.5742,
+      "step": 6943
+    },
+    {
+      "epoch": 1.8434886499402627,
+      "grad_norm": 0.41128856398162067,
+      "learning_rate": 3.942444780027732e-06,
+      "loss": 0.5581,
+      "step": 6944
+    },
+    {
+      "epoch": 1.8437541484136466,
+      "grad_norm": 0.4069034956182217,
+      "learning_rate": 3.942159612332408e-06,
+      "loss": 0.5637,
+      "step": 6945
+    },
+    {
+      "epoch": 1.8440196468870305,
+      "grad_norm": 0.415807308141061,
+      "learning_rate": 3.941874416511441e-06,
+      "loss": 0.6002,
+      "step": 6946
+    },
+    {
+      "epoch": 1.844285145360414,
+      "grad_norm": 0.4144794052635019,
+      "learning_rate": 3.941589192570392e-06,
+      "loss": 0.573,
+      "step": 6947
+    },
+    {
+      "epoch": 1.844550643833798,
+      "grad_norm": 0.4149018012203615,
+      "learning_rate": 3.941303940514826e-06,
+      "loss": 0.567,
+      "step": 6948
+    },
+    {
+      "epoch": 1.8448161423071818,
+      "grad_norm": 0.4095783353325972,
+      "learning_rate": 3.941018660350302e-06,
+      "loss": 0.5538,
+      "step": 6949
+    },
+    {
+      "epoch": 1.8450816407805655,
+      "grad_norm": 0.39799016740997856,
+      "learning_rate": 3.940733352082387e-06,
+      "loss": 0.5661,
+      "step": 6950
+    },
+    {
+      "epoch": 1.8453471392539493,
+      "grad_norm": 0.42955056674073594,
+      "learning_rate": 3.940448015716644e-06,
+      "loss": 0.5729,
+      "step": 6951
+    },
+    {
+      "epoch": 1.8456126377273332,
+      "grad_norm": 0.4163976796612784,
+      "learning_rate": 3.940162651258637e-06,
+      "loss": 0.5954,
+      "step": 6952
+    },
+    {
+      "epoch": 1.8458781362007168,
+      "grad_norm": 0.41032931712941006,
+      "learning_rate": 3.939877258713933e-06,
+      "loss": 0.5901,
+      "step": 6953
+    },
+    {
+      "epoch": 1.8461436346741005,
+      "grad_norm": 0.4090675884744846,
+      "learning_rate": 3.939591838088097e-06,
+      "loss": 0.5996,
+      "step": 6954
+    },
+    {
+      "epoch": 1.8464091331474846,
+      "grad_norm": 0.4142928798379635,
+      "learning_rate": 3.939306389386696e-06,
+      "loss": 0.5669,
+      "step": 6955
+    },
+    {
+      "epoch": 1.8466746316208682,
+      "grad_norm": 0.4213378709982677,
+      "learning_rate": 3.939020912615295e-06,
+      "loss": 0.5718,
+      "step": 6956
+    },
+    {
+      "epoch": 1.8469401300942518,
+      "grad_norm": 0.39484293134250065,
+      "learning_rate": 3.938735407779464e-06,
+      "loss": 0.5098,
+      "step": 6957
+    },
+    {
+      "epoch": 1.8472056285676357,
+      "grad_norm": 0.42292218393672387,
+      "learning_rate": 3.93844987488477e-06,
+      "loss": 0.6282,
+      "step": 6958
+    },
+    {
+      "epoch": 1.8474711270410196,
+      "grad_norm": 0.39724106047794144,
+      "learning_rate": 3.938164313936781e-06,
+      "loss": 0.5649,
+      "step": 6959
+    },
+    {
+      "epoch": 1.8477366255144032,
+      "grad_norm": 0.4189608291655136,
+      "learning_rate": 3.937878724941066e-06,
+      "loss": 0.5761,
+      "step": 6960
+    },
+    {
+      "epoch": 1.848002123987787,
+      "grad_norm": 0.41197466655443343,
+      "learning_rate": 3.937593107903196e-06,
+      "loss": 0.59,
+      "step": 6961
+    },
+    {
+      "epoch": 1.848267622461171,
+      "grad_norm": 0.41261599546037114,
+      "learning_rate": 3.93730746282874e-06,
+      "loss": 0.5965,
+      "step": 6962
+    },
+    {
+      "epoch": 1.8485331209345546,
+      "grad_norm": 0.4103523178610754,
+      "learning_rate": 3.93702178972327e-06,
+      "loss": 0.5908,
+      "step": 6963
+    },
+    {
+      "epoch": 1.8487986194079384,
+      "grad_norm": 0.4059384165149419,
+      "learning_rate": 3.936736088592357e-06,
+      "loss": 0.5788,
+      "step": 6964
+    },
+    {
+      "epoch": 1.8490641178813223,
+      "grad_norm": 0.4159782179573111,
+      "learning_rate": 3.936450359441572e-06,
+      "loss": 0.5652,
+      "step": 6965
+    },
+    {
+      "epoch": 1.849329616354706,
+      "grad_norm": 0.4133402380836053,
+      "learning_rate": 3.936164602276488e-06,
+      "loss": 0.5645,
+      "step": 6966
+    },
+    {
+      "epoch": 1.8495951148280896,
+      "grad_norm": 0.3940295484067086,
+      "learning_rate": 3.935878817102679e-06,
+      "loss": 0.5685,
+      "step": 6967
+    },
+    {
+      "epoch": 1.8498606133014737,
+      "grad_norm": 0.4082797448803069,
+      "learning_rate": 3.935593003925716e-06,
+      "loss": 0.5554,
+      "step": 6968
+    },
+    {
+      "epoch": 1.8501261117748573,
+      "grad_norm": 0.4012126929322681,
+      "learning_rate": 3.935307162751174e-06,
+      "loss": 0.5789,
+      "step": 6969
+    },
+    {
+      "epoch": 1.850391610248241,
+      "grad_norm": 0.40915072726009466,
+      "learning_rate": 3.935021293584629e-06,
+      "loss": 0.5589,
+      "step": 6970
+    },
+    {
+      "epoch": 1.8506571087216248,
+      "grad_norm": 0.409323474463736,
+      "learning_rate": 3.934735396431656e-06,
+      "loss": 0.612,
+      "step": 6971
+    },
+    {
+      "epoch": 1.8509226071950087,
+      "grad_norm": 0.40803397077971276,
+      "learning_rate": 3.934449471297829e-06,
+      "loss": 0.6003,
+      "step": 6972
+    },
+    {
+      "epoch": 1.8511881056683923,
+      "grad_norm": 0.4044914220403835,
+      "learning_rate": 3.934163518188725e-06,
+      "loss": 0.5878,
+      "step": 6973
+    },
+    {
+      "epoch": 1.8514536041417762,
+      "grad_norm": 0.41849556901391216,
+      "learning_rate": 3.933877537109921e-06,
+      "loss": 0.5893,
+      "step": 6974
+    },
+    {
+      "epoch": 1.85171910261516,
+      "grad_norm": 0.4121817583274806,
+      "learning_rate": 3.933591528066995e-06,
+      "loss": 0.5838,
+      "step": 6975
+    },
+    {
+      "epoch": 1.8519846010885437,
+      "grad_norm": 0.4245968786430536,
+      "learning_rate": 3.933305491065524e-06,
+      "loss": 0.6225,
+      "step": 6976
+    },
+    {
+      "epoch": 1.8522500995619275,
+      "grad_norm": 0.41767815062103897,
+      "learning_rate": 3.933019426111086e-06,
+      "loss": 0.5724,
+      "step": 6977
+    },
+    {
+      "epoch": 1.8525155980353114,
+      "grad_norm": 0.40692991333474743,
+      "learning_rate": 3.932733333209261e-06,
+      "loss": 0.5894,
+      "step": 6978
+    },
+    {
+      "epoch": 1.852781096508695,
+      "grad_norm": 0.4297095729670086,
+      "learning_rate": 3.932447212365627e-06,
+      "loss": 0.5968,
+      "step": 6979
+    },
+    {
+      "epoch": 1.8530465949820787,
+      "grad_norm": 0.4038910888247878,
+      "learning_rate": 3.9321610635857675e-06,
+      "loss": 0.5839,
+      "step": 6980
+    },
+    {
+      "epoch": 1.8533120934554628,
+      "grad_norm": 0.39807978095272545,
+      "learning_rate": 3.931874886875259e-06,
+      "loss": 0.551,
+      "step": 6981
+    },
+    {
+      "epoch": 1.8535775919288464,
+      "grad_norm": 0.4181460716959522,
+      "learning_rate": 3.931588682239684e-06,
+      "loss": 0.5938,
+      "step": 6982
+    },
+    {
+      "epoch": 1.85384309040223,
+      "grad_norm": 0.4072785093978776,
+      "learning_rate": 3.931302449684625e-06,
+      "loss": 0.5822,
+      "step": 6983
+    },
+    {
+      "epoch": 1.854108588875614,
+      "grad_norm": 0.41679217851508005,
+      "learning_rate": 3.931016189215664e-06,
+      "loss": 0.5978,
+      "step": 6984
+    },
+    {
+      "epoch": 1.8543740873489978,
+      "grad_norm": 0.4198293604984194,
+      "learning_rate": 3.930729900838384e-06,
+      "loss": 0.6019,
+      "step": 6985
+    },
+    {
+      "epoch": 1.8546395858223814,
+      "grad_norm": 0.39854310088267186,
+      "learning_rate": 3.930443584558368e-06,
+      "loss": 0.5692,
+      "step": 6986
+    },
+    {
+      "epoch": 1.8549050842957653,
+      "grad_norm": 0.41363365602146657,
+      "learning_rate": 3.930157240381199e-06,
+      "loss": 0.5915,
+      "step": 6987
+    },
+    {
+      "epoch": 1.8551705827691491,
+      "grad_norm": 0.40609414164861235,
+      "learning_rate": 3.929870868312463e-06,
+      "loss": 0.6193,
+      "step": 6988
+    },
+    {
+      "epoch": 1.8554360812425328,
+      "grad_norm": 0.4102358930830493,
+      "learning_rate": 3.929584468357743e-06,
+      "loss": 0.5886,
+      "step": 6989
+    },
+    {
+      "epoch": 1.8557015797159166,
+      "grad_norm": 0.4120011596357434,
+      "learning_rate": 3.929298040522626e-06,
+      "loss": 0.5691,
+      "step": 6990
+    },
+    {
+      "epoch": 1.8559670781893005,
+      "grad_norm": 0.39642712371795763,
+      "learning_rate": 3.929011584812698e-06,
+      "loss": 0.5878,
+      "step": 6991
+    },
+    {
+      "epoch": 1.8562325766626842,
+      "grad_norm": 0.4229512224512087,
+      "learning_rate": 3.928725101233544e-06,
+      "loss": 0.6005,
+      "step": 6992
+    },
+    {
+      "epoch": 1.856498075136068,
+      "grad_norm": 0.419682157187313,
+      "learning_rate": 3.928438589790755e-06,
+      "loss": 0.6292,
+      "step": 6993
+    },
+    {
+      "epoch": 1.8567635736094519,
+      "grad_norm": 0.4311370340813145,
+      "learning_rate": 3.928152050489915e-06,
+      "loss": 0.5986,
+      "step": 6994
+    },
+    {
+      "epoch": 1.8570290720828355,
+      "grad_norm": 0.4188002309867436,
+      "learning_rate": 3.927865483336613e-06,
+      "loss": 0.5549,
+      "step": 6995
+    },
+    {
+      "epoch": 1.8572945705562192,
+      "grad_norm": 0.40987962762826796,
+      "learning_rate": 3.927578888336437e-06,
+      "loss": 0.5774,
+      "step": 6996
+    },
+    {
+      "epoch": 1.8575600690296032,
+      "grad_norm": 0.4123701598687084,
+      "learning_rate": 3.927292265494979e-06,
+      "loss": 0.5967,
+      "step": 6997
+    },
+    {
+      "epoch": 1.8578255675029869,
+      "grad_norm": 0.43184928741569023,
+      "learning_rate": 3.9270056148178265e-06,
+      "loss": 0.5483,
+      "step": 6998
+    },
+    {
+      "epoch": 1.8580910659763705,
+      "grad_norm": 0.4177628607055168,
+      "learning_rate": 3.926718936310571e-06,
+      "loss": 0.5819,
+      "step": 6999
+    },
+    {
+      "epoch": 1.8583565644497544,
+      "grad_norm": 0.4137772822737616,
+      "learning_rate": 3.926432229978803e-06,
+      "loss": 0.5778,
+      "step": 7000
+    },
+    {
+      "epoch": 1.8586220629231383,
+      "grad_norm": 0.40507119077028475,
+      "learning_rate": 3.926145495828113e-06,
+      "loss": 0.5789,
+      "step": 7001
+    },
+    {
+      "epoch": 1.858887561396522,
+      "grad_norm": 0.4366827296791418,
+      "learning_rate": 3.925858733864095e-06,
+      "loss": 0.5826,
+      "step": 7002
+    },
+    {
+      "epoch": 1.8591530598699058,
+      "grad_norm": 0.4225870782102806,
+      "learning_rate": 3.92557194409234e-06,
+      "loss": 0.5446,
+      "step": 7003
+    },
+    {
+      "epoch": 1.8594185583432896,
+      "grad_norm": 0.409338830054313,
+      "learning_rate": 3.9252851265184426e-06,
+      "loss": 0.5718,
+      "step": 7004
+    },
+    {
+      "epoch": 1.8596840568166733,
+      "grad_norm": 0.4147793387598263,
+      "learning_rate": 3.924998281147995e-06,
+      "loss": 0.6116,
+      "step": 7005
+    },
+    {
+      "epoch": 1.8599495552900571,
+      "grad_norm": 0.4242447271527274,
+      "learning_rate": 3.924711407986591e-06,
+      "loss": 0.6009,
+      "step": 7006
+    },
+    {
+      "epoch": 1.860215053763441,
+      "grad_norm": 0.4243040590806271,
+      "learning_rate": 3.924424507039828e-06,
+      "loss": 0.5839,
+      "step": 7007
+    },
+    {
+      "epoch": 1.8604805522368246,
+      "grad_norm": 0.4083765144251573,
+      "learning_rate": 3.924137578313299e-06,
+      "loss": 0.5757,
+      "step": 7008
+    },
+    {
+      "epoch": 1.8607460507102083,
+      "grad_norm": 0.4344873741278483,
+      "learning_rate": 3.9238506218125995e-06,
+      "loss": 0.6167,
+      "step": 7009
+    },
+    {
+      "epoch": 1.8610115491835924,
+      "grad_norm": 0.42440515609996415,
+      "learning_rate": 3.923563637543327e-06,
+      "loss": 0.5723,
+      "step": 7010
+    },
+    {
+      "epoch": 1.861277047656976,
+      "grad_norm": 0.4117864765771132,
+      "learning_rate": 3.923276625511078e-06,
+      "loss": 0.5164,
+      "step": 7011
+    },
+    {
+      "epoch": 1.8615425461303596,
+      "grad_norm": 0.398080969531183,
+      "learning_rate": 3.92298958572145e-06,
+      "loss": 0.5698,
+      "step": 7012
+    },
+    {
+      "epoch": 1.8618080446037435,
+      "grad_norm": 0.4152261717406292,
+      "learning_rate": 3.922702518180042e-06,
+      "loss": 0.6084,
+      "step": 7013
+    },
+    {
+      "epoch": 1.8620735430771274,
+      "grad_norm": 0.41227070117414727,
+      "learning_rate": 3.92241542289245e-06,
+      "loss": 0.5941,
+      "step": 7014
+    },
+    {
+      "epoch": 1.862339041550511,
+      "grad_norm": 0.399566862205925,
+      "learning_rate": 3.922128299864276e-06,
+      "loss": 0.554,
+      "step": 7015
+    },
+    {
+      "epoch": 1.8626045400238949,
+      "grad_norm": 0.4023805576115783,
+      "learning_rate": 3.9218411491011175e-06,
+      "loss": 0.5715,
+      "step": 7016
+    },
+    {
+      "epoch": 1.8628700384972787,
+      "grad_norm": 0.427954927302662,
+      "learning_rate": 3.921553970608575e-06,
+      "loss": 0.5953,
+      "step": 7017
+    },
+    {
+      "epoch": 1.8631355369706624,
+      "grad_norm": 0.4072849853449334,
+      "learning_rate": 3.92126676439225e-06,
+      "loss": 0.5737,
+      "step": 7018
+    },
+    {
+      "epoch": 1.8634010354440462,
+      "grad_norm": 0.4117371358774136,
+      "learning_rate": 3.920979530457743e-06,
+      "loss": 0.5186,
+      "step": 7019
+    },
+    {
+      "epoch": 1.86366653391743,
+      "grad_norm": 0.41760085787178824,
+      "learning_rate": 3.920692268810656e-06,
+      "loss": 0.5994,
+      "step": 7020
+    },
+    {
+      "epoch": 1.8639320323908137,
+      "grad_norm": 0.43754121540634955,
+      "learning_rate": 3.9204049794565916e-06,
+      "loss": 0.5334,
+      "step": 7021
+    },
+    {
+      "epoch": 1.8641975308641974,
+      "grad_norm": 0.4066923846218359,
+      "learning_rate": 3.920117662401153e-06,
+      "loss": 0.6017,
+      "step": 7022
+    },
+    {
+      "epoch": 1.8644630293375815,
+      "grad_norm": 0.425835622217567,
+      "learning_rate": 3.919830317649942e-06,
+      "loss": 0.603,
+      "step": 7023
+    },
+    {
+      "epoch": 1.864728527810965,
+      "grad_norm": 0.41243683298308464,
+      "learning_rate": 3.919542945208563e-06,
+      "loss": 0.5594,
+      "step": 7024
+    },
+    {
+      "epoch": 1.8649940262843487,
+      "grad_norm": 0.4323968409259387,
+      "learning_rate": 3.919255545082622e-06,
+      "loss": 0.5988,
+      "step": 7025
+    },
+    {
+      "epoch": 1.8652595247577326,
+      "grad_norm": 0.41124068422260446,
+      "learning_rate": 3.918968117277722e-06,
+      "loss": 0.5965,
+      "step": 7026
+    },
+    {
+      "epoch": 1.8655250232311165,
+      "grad_norm": 0.41688159374831313,
+      "learning_rate": 3.91868066179947e-06,
+      "loss": 0.5406,
+      "step": 7027
+    },
+    {
+      "epoch": 1.8657905217045,
+      "grad_norm": 0.4205942850696052,
+      "learning_rate": 3.918393178653472e-06,
+      "loss": 0.6348,
+      "step": 7028
+    },
+    {
+      "epoch": 1.866056020177884,
+      "grad_norm": 0.42770247507352255,
+      "learning_rate": 3.918105667845333e-06,
+      "loss": 0.6007,
+      "step": 7029
+    },
+    {
+      "epoch": 1.8663215186512678,
+      "grad_norm": 0.40042531361537104,
+      "learning_rate": 3.9178181293806625e-06,
+      "loss": 0.566,
+      "step": 7030
+    },
+    {
+      "epoch": 1.8665870171246515,
+      "grad_norm": 0.42314818105008406,
+      "learning_rate": 3.9175305632650665e-06,
+      "loss": 0.5958,
+      "step": 7031
+    },
+    {
+      "epoch": 1.8668525155980353,
+      "grad_norm": 0.4151923153351884,
+      "learning_rate": 3.917242969504154e-06,
+      "loss": 0.5921,
+      "step": 7032
+    },
+    {
+      "epoch": 1.8671180140714192,
+      "grad_norm": 0.40717137058083086,
+      "learning_rate": 3.916955348103533e-06,
+      "loss": 0.5954,
+      "step": 7033
+    },
+    {
+      "epoch": 1.8673835125448028,
+      "grad_norm": 0.4010803302552354,
+      "learning_rate": 3.916667699068815e-06,
+      "loss": 0.6084,
+      "step": 7034
+    },
+    {
+      "epoch": 1.8676490110181867,
+      "grad_norm": 0.4177206524164597,
+      "learning_rate": 3.916380022405606e-06,
+      "loss": 0.6482,
+      "step": 7035
+    },
+    {
+      "epoch": 1.8679145094915706,
+      "grad_norm": 0.4064869796423884,
+      "learning_rate": 3.91609231811952e-06,
+      "loss": 0.5894,
+      "step": 7036
+    },
+    {
+      "epoch": 1.8681800079649542,
+      "grad_norm": 0.4117399834446077,
+      "learning_rate": 3.915804586216166e-06,
+      "loss": 0.5939,
+      "step": 7037
+    },
+    {
+      "epoch": 1.8684455064383378,
+      "grad_norm": 0.4038886545348688,
+      "learning_rate": 3.915516826701157e-06,
+      "loss": 0.5296,
+      "step": 7038
+    },
+    {
+      "epoch": 1.8687110049117217,
+      "grad_norm": 0.4128136495153341,
+      "learning_rate": 3.9152290395801034e-06,
+      "loss": 0.5759,
+      "step": 7039
+    },
+    {
+      "epoch": 1.8689765033851056,
+      "grad_norm": 0.40882796094160617,
+      "learning_rate": 3.914941224858619e-06,
+      "loss": 0.6188,
+      "step": 7040
+    },
+    {
+      "epoch": 1.8692420018584892,
+      "grad_norm": 0.4236903690793271,
+      "learning_rate": 3.914653382542316e-06,
+      "loss": 0.6061,
+      "step": 7041
+    },
+    {
+      "epoch": 1.869507500331873,
+      "grad_norm": 0.45522785024099915,
+      "learning_rate": 3.914365512636808e-06,
+      "loss": 0.5333,
+      "step": 7042
+    },
+    {
+      "epoch": 1.869772998805257,
+      "grad_norm": 0.40912051091098667,
+      "learning_rate": 3.91407761514771e-06,
+      "loss": 0.5738,
+      "step": 7043
+    },
+    {
+      "epoch": 1.8700384972786406,
+      "grad_norm": 0.42416505063028453,
+      "learning_rate": 3.913789690080636e-06,
+      "loss": 0.5897,
+      "step": 7044
+    },
+    {
+      "epoch": 1.8703039957520244,
+      "grad_norm": 0.45966633961184605,
+      "learning_rate": 3.913501737441202e-06,
+      "loss": 0.5864,
+      "step": 7045
+    },
+    {
+      "epoch": 1.8705694942254083,
+      "grad_norm": 0.46243260498498207,
+      "learning_rate": 3.913213757235023e-06,
+      "loss": 0.5317,
+      "step": 7046
+    },
+    {
+      "epoch": 1.870834992698792,
+      "grad_norm": 0.41045049137262035,
+      "learning_rate": 3.912925749467715e-06,
+      "loss": 0.584,
+      "step": 7047
+    },
+    {
+      "epoch": 1.8711004911721758,
+      "grad_norm": 0.4265956650495861,
+      "learning_rate": 3.912637714144897e-06,
+      "loss": 0.5848,
+      "step": 7048
+    },
+    {
+      "epoch": 1.8713659896455597,
+      "grad_norm": 0.42875699279050344,
+      "learning_rate": 3.9123496512721835e-06,
+      "loss": 0.5396,
+      "step": 7049
+    },
+    {
+      "epoch": 1.8716314881189433,
+      "grad_norm": 0.4485515056738511,
+      "learning_rate": 3.912061560855194e-06,
+      "loss": 0.5734,
+      "step": 7050
+    },
+    {
+      "epoch": 1.871896986592327,
+      "grad_norm": 0.3963442275472542,
+      "learning_rate": 3.911773442899547e-06,
+      "loss": 0.5685,
+      "step": 7051
+    },
+    {
+      "epoch": 1.872162485065711,
+      "grad_norm": 0.40815834074072116,
+      "learning_rate": 3.911485297410861e-06,
+      "loss": 0.5848,
+      "step": 7052
+    },
+    {
+      "epoch": 1.8724279835390947,
+      "grad_norm": 0.4177885606628742,
+      "learning_rate": 3.9111971243947555e-06,
+      "loss": 0.6002,
+      "step": 7053
+    },
+    {
+      "epoch": 1.8726934820124783,
+      "grad_norm": 0.3982488207452386,
+      "learning_rate": 3.910908923856851e-06,
+      "loss": 0.5551,
+      "step": 7054
+    },
+    {
+      "epoch": 1.8729589804858622,
+      "grad_norm": 0.4073962251072794,
+      "learning_rate": 3.910620695802768e-06,
+      "loss": 0.6004,
+      "step": 7055
+    },
+    {
+      "epoch": 1.873224478959246,
+      "grad_norm": 0.4427886926565442,
+      "learning_rate": 3.9103324402381285e-06,
+      "loss": 0.5414,
+      "step": 7056
+    },
+    {
+      "epoch": 1.8734899774326297,
+      "grad_norm": 0.41073338129047515,
+      "learning_rate": 3.910044157168552e-06,
+      "loss": 0.5758,
+      "step": 7057
+    },
+    {
+      "epoch": 1.8737554759060135,
+      "grad_norm": 0.41259637289069107,
+      "learning_rate": 3.909755846599663e-06,
+      "loss": 0.573,
+      "step": 7058
+    },
+    {
+      "epoch": 1.8740209743793974,
+      "grad_norm": 0.4290440138050512,
+      "learning_rate": 3.909467508537082e-06,
+      "loss": 0.5761,
+      "step": 7059
+    },
+    {
+      "epoch": 1.874286472852781,
+      "grad_norm": 0.4248638059398351,
+      "learning_rate": 3.909179142986436e-06,
+      "loss": 0.5467,
+      "step": 7060
+    },
+    {
+      "epoch": 1.874551971326165,
+      "grad_norm": 0.4014542453981169,
+      "learning_rate": 3.908890749953344e-06,
+      "loss": 0.5637,
+      "step": 7061
+    },
+    {
+      "epoch": 1.8748174697995488,
+      "grad_norm": 0.41432015219835344,
+      "learning_rate": 3.908602329443434e-06,
+      "loss": 0.5928,
+      "step": 7062
+    },
+    {
+      "epoch": 1.8750829682729324,
+      "grad_norm": 0.42709130461589084,
+      "learning_rate": 3.90831388146233e-06,
+      "loss": 0.6149,
+      "step": 7063
+    },
+    {
+      "epoch": 1.875348466746316,
+      "grad_norm": 0.41988608307090197,
+      "learning_rate": 3.908025406015656e-06,
+      "loss": 0.6249,
+      "step": 7064
+    },
+    {
+      "epoch": 1.8756139652197001,
+      "grad_norm": 0.41175358541659673,
+      "learning_rate": 3.907736903109041e-06,
+      "loss": 0.5573,
+      "step": 7065
+    },
+    {
+      "epoch": 1.8758794636930838,
+      "grad_norm": 0.43369499182599835,
+      "learning_rate": 3.907448372748109e-06,
+      "loss": 0.575,
+      "step": 7066
+    },
+    {
+      "epoch": 1.8761449621664674,
+      "grad_norm": 0.38989814639866516,
+      "learning_rate": 3.907159814938487e-06,
+      "loss": 0.568,
+      "step": 7067
+    },
+    {
+      "epoch": 1.8764104606398513,
+      "grad_norm": 0.42665951948656816,
+      "learning_rate": 3.906871229685803e-06,
+      "loss": 0.5758,
+      "step": 7068
+    },
+    {
+      "epoch": 1.8766759591132351,
+      "grad_norm": 0.3906605363540033,
+      "learning_rate": 3.906582616995686e-06,
+      "loss": 0.6332,
+      "step": 7069
+    },
+    {
+      "epoch": 1.8769414575866188,
+      "grad_norm": 0.4502266554381831,
+      "learning_rate": 3.906293976873764e-06,
+      "loss": 0.5759,
+      "step": 7070
+    },
+    {
+      "epoch": 1.8772069560600027,
+      "grad_norm": 0.4401907751667689,
+      "learning_rate": 3.9060053093256664e-06,
+      "loss": 0.5683,
+      "step": 7071
+    },
+    {
+      "epoch": 1.8774724545333865,
+      "grad_norm": 0.3887070679862068,
+      "learning_rate": 3.905716614357023e-06,
+      "loss": 0.5801,
+      "step": 7072
+    },
+    {
+      "epoch": 1.8777379530067702,
+      "grad_norm": 0.426074423699116,
+      "learning_rate": 3.905427891973463e-06,
+      "loss": 0.582,
+      "step": 7073
+    },
+    {
+      "epoch": 1.878003451480154,
+      "grad_norm": 0.4311525127479367,
+      "learning_rate": 3.905139142180619e-06,
+      "loss": 0.5939,
+      "step": 7074
+    },
+    {
+      "epoch": 1.8782689499535379,
+      "grad_norm": 0.4289219634781661,
+      "learning_rate": 3.904850364984121e-06,
+      "loss": 0.5718,
+      "step": 7075
+    },
+    {
+      "epoch": 1.8785344484269215,
+      "grad_norm": 0.41533920716406125,
+      "learning_rate": 3.904561560389601e-06,
+      "loss": 0.5811,
+      "step": 7076
+    },
+    {
+      "epoch": 1.8787999469003052,
+      "grad_norm": 0.41417032333409726,
+      "learning_rate": 3.9042727284026915e-06,
+      "loss": 0.5811,
+      "step": 7077
+    },
+    {
+      "epoch": 1.8790654453736892,
+      "grad_norm": 0.4119648391998343,
+      "learning_rate": 3.903983869029025e-06,
+      "loss": 0.5629,
+      "step": 7078
+    },
+    {
+      "epoch": 1.8793309438470729,
+      "grad_norm": 0.4338637584952911,
+      "learning_rate": 3.9036949822742364e-06,
+      "loss": 0.5871,
+      "step": 7079
+    },
+    {
+      "epoch": 1.8795964423204565,
+      "grad_norm": 0.4248311757657316,
+      "learning_rate": 3.903406068143959e-06,
+      "loss": 0.5396,
+      "step": 7080
+    },
+    {
+      "epoch": 1.8798619407938404,
+      "grad_norm": 0.4436117104678151,
+      "learning_rate": 3.903117126643828e-06,
+      "loss": 0.5837,
+      "step": 7081
+    },
+    {
+      "epoch": 1.8801274392672243,
+      "grad_norm": 0.40926219767764765,
+      "learning_rate": 3.902828157779475e-06,
+      "loss": 0.5778,
+      "step": 7082
+    },
+    {
+      "epoch": 1.880392937740608,
+      "grad_norm": 0.44262614321451976,
+      "learning_rate": 3.90253916155654e-06,
+      "loss": 0.5845,
+      "step": 7083
+    },
+    {
+      "epoch": 1.8806584362139918,
+      "grad_norm": 0.4208902684691188,
+      "learning_rate": 3.902250137980656e-06,
+      "loss": 0.5783,
+      "step": 7084
+    },
+    {
+      "epoch": 1.8809239346873756,
+      "grad_norm": 0.4314078678057738,
+      "learning_rate": 3.901961087057463e-06,
+      "loss": 0.5801,
+      "step": 7085
+    },
+    {
+      "epoch": 1.8811894331607593,
+      "grad_norm": 0.4070660062455029,
+      "learning_rate": 3.901672008792595e-06,
+      "loss": 0.5939,
+      "step": 7086
+    },
+    {
+      "epoch": 1.8814549316341431,
+      "grad_norm": 0.41904395539529654,
+      "learning_rate": 3.901382903191691e-06,
+      "loss": 0.5527,
+      "step": 7087
+    },
+    {
+      "epoch": 1.881720430107527,
+      "grad_norm": 0.41560623587894185,
+      "learning_rate": 3.901093770260389e-06,
+      "loss": 0.5851,
+      "step": 7088
+    },
+    {
+      "epoch": 1.8819859285809106,
+      "grad_norm": 0.4192905403772401,
+      "learning_rate": 3.900804610004329e-06,
+      "loss": 0.6059,
+      "step": 7089
+    },
+    {
+      "epoch": 1.8822514270542945,
+      "grad_norm": 0.43038459333408474,
+      "learning_rate": 3.900515422429149e-06,
+      "loss": 0.6114,
+      "step": 7090
+    },
+    {
+      "epoch": 1.8825169255276784,
+      "grad_norm": 0.41069500608507187,
+      "learning_rate": 3.900226207540489e-06,
+      "loss": 0.5919,
+      "step": 7091
+    },
+    {
+      "epoch": 1.882782424001062,
+      "grad_norm": 0.41503715019740184,
+      "learning_rate": 3.899936965343989e-06,
+      "loss": 0.6078,
+      "step": 7092
+    },
+    {
+      "epoch": 1.8830479224744456,
+      "grad_norm": 0.42023327605317484,
+      "learning_rate": 3.899647695845291e-06,
+      "loss": 0.5638,
+      "step": 7093
+    },
+    {
+      "epoch": 1.8833134209478295,
+      "grad_norm": 0.39592249521843,
+      "learning_rate": 3.899358399050036e-06,
+      "loss": 0.5636,
+      "step": 7094
+    },
+    {
+      "epoch": 1.8835789194212134,
+      "grad_norm": 0.41660178126431174,
+      "learning_rate": 3.899069074963866e-06,
+      "loss": 0.5839,
+      "step": 7095
+    },
+    {
+      "epoch": 1.883844417894597,
+      "grad_norm": 0.4080299981450624,
+      "learning_rate": 3.8987797235924225e-06,
+      "loss": 0.6028,
+      "step": 7096
+    },
+    {
+      "epoch": 1.8841099163679809,
+      "grad_norm": 0.40829228703046844,
+      "learning_rate": 3.898490344941352e-06,
+      "loss": 0.5658,
+      "step": 7097
+    },
+    {
+      "epoch": 1.8843754148413647,
+      "grad_norm": 0.3937678374315934,
+      "learning_rate": 3.898200939016295e-06,
+      "loss": 0.5685,
+      "step": 7098
+    },
+    {
+      "epoch": 1.8846409133147484,
+      "grad_norm": 0.42225679619736156,
+      "learning_rate": 3.897911505822894e-06,
+      "loss": 0.5641,
+      "step": 7099
+    },
+    {
+      "epoch": 1.8849064117881322,
+      "grad_norm": 0.4124490971575039,
+      "learning_rate": 3.897622045366798e-06,
+      "loss": 0.5664,
+      "step": 7100
+    },
+    {
+      "epoch": 1.885171910261516,
+      "grad_norm": 0.41383272080816186,
+      "learning_rate": 3.8973325576536504e-06,
+      "loss": 0.604,
+      "step": 7101
+    },
+    {
+      "epoch": 1.8854374087348997,
+      "grad_norm": 0.42296562095170526,
+      "learning_rate": 3.897043042689095e-06,
+      "loss": 0.5908,
+      "step": 7102
+    },
+    {
+      "epoch": 1.8857029072082836,
+      "grad_norm": 0.41033353709716786,
+      "learning_rate": 3.896753500478781e-06,
+      "loss": 0.5726,
+      "step": 7103
+    },
+    {
+      "epoch": 1.8859684056816675,
+      "grad_norm": 0.42002942515730557,
+      "learning_rate": 3.896463931028354e-06,
+      "loss": 0.6214,
+      "step": 7104
+    },
+    {
+      "epoch": 1.886233904155051,
+      "grad_norm": 0.40460757940480924,
+      "learning_rate": 3.896174334343461e-06,
+      "loss": 0.5417,
+      "step": 7105
+    },
+    {
+      "epoch": 1.8864994026284347,
+      "grad_norm": 0.4135518491071666,
+      "learning_rate": 3.89588471042975e-06,
+      "loss": 0.6152,
+      "step": 7106
+    },
+    {
+      "epoch": 1.8867649011018188,
+      "grad_norm": 0.4054098715824253,
+      "learning_rate": 3.89559505929287e-06,
+      "loss": 0.5941,
+      "step": 7107
+    },
+    {
+      "epoch": 1.8870303995752025,
+      "grad_norm": 0.4074179156026263,
+      "learning_rate": 3.895305380938468e-06,
+      "loss": 0.5917,
+      "step": 7108
+    },
+    {
+      "epoch": 1.887295898048586,
+      "grad_norm": 0.4141455513704588,
+      "learning_rate": 3.895015675372195e-06,
+      "loss": 0.576,
+      "step": 7109
+    },
+    {
+      "epoch": 1.88756139652197,
+      "grad_norm": 0.4056926170099754,
+      "learning_rate": 3.894725942599701e-06,
+      "loss": 0.5842,
+      "step": 7110
+    },
+    {
+      "epoch": 1.8878268949953538,
+      "grad_norm": 0.3904370757020439,
+      "learning_rate": 3.894436182626636e-06,
+      "loss": 0.5772,
+      "step": 7111
+    },
+    {
+      "epoch": 1.8880923934687375,
+      "grad_norm": 0.4068417976627062,
+      "learning_rate": 3.894146395458653e-06,
+      "loss": 0.5864,
+      "step": 7112
+    },
+    {
+      "epoch": 1.8883578919421213,
+      "grad_norm": 0.42834493965508247,
+      "learning_rate": 3.893856581101399e-06,
+      "loss": 0.5723,
+      "step": 7113
+    },
+    {
+      "epoch": 1.8886233904155052,
+      "grad_norm": 0.4209024790297688,
+      "learning_rate": 3.893566739560532e-06,
+      "loss": 0.5457,
+      "step": 7114
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.4108870116465058,
+      "learning_rate": 3.893276870841699e-06,
+      "loss": 0.5806,
+      "step": 7115
+    },
+    {
+      "epoch": 1.8891543873622727,
+      "grad_norm": 0.4253025282359474,
+      "learning_rate": 3.892986974950557e-06,
+      "loss": 0.5417,
+      "step": 7116
+    },
+    {
+      "epoch": 1.8894198858356566,
+      "grad_norm": 0.41170489180390835,
+      "learning_rate": 3.892697051892758e-06,
+      "loss": 0.5709,
+      "step": 7117
+    },
+    {
+      "epoch": 1.8896853843090402,
+      "grad_norm": 0.4286715874308777,
+      "learning_rate": 3.892407101673957e-06,
+      "loss": 0.6024,
+      "step": 7118
+    },
+    {
+      "epoch": 1.8899508827824238,
+      "grad_norm": 0.4213431871891988,
+      "learning_rate": 3.8921171242998076e-06,
+      "loss": 0.5818,
+      "step": 7119
+    },
+    {
+      "epoch": 1.890216381255808,
+      "grad_norm": 0.4134217887042566,
+      "learning_rate": 3.891827119775966e-06,
+      "loss": 0.6229,
+      "step": 7120
+    },
+    {
+      "epoch": 1.8904818797291916,
+      "grad_norm": 0.39975942266424874,
+      "learning_rate": 3.891537088108088e-06,
+      "loss": 0.5957,
+      "step": 7121
+    },
+    {
+      "epoch": 1.8907473782025752,
+      "grad_norm": 0.4100904851031382,
+      "learning_rate": 3.8912470293018295e-06,
+      "loss": 0.5524,
+      "step": 7122
+    },
+    {
+      "epoch": 1.891012876675959,
+      "grad_norm": 0.41251017499776654,
+      "learning_rate": 3.890956943362848e-06,
+      "loss": 0.5927,
+      "step": 7123
+    },
+    {
+      "epoch": 1.891278375149343,
+      "grad_norm": 0.4120074103450618,
+      "learning_rate": 3.8906668302967995e-06,
+      "loss": 0.5797,
+      "step": 7124
+    },
+    {
+      "epoch": 1.8915438736227266,
+      "grad_norm": 0.4217471409968703,
+      "learning_rate": 3.8903766901093435e-06,
+      "loss": 0.625,
+      "step": 7125
+    },
+    {
+      "epoch": 1.8918093720961104,
+      "grad_norm": 0.40144946257709435,
+      "learning_rate": 3.8900865228061385e-06,
+      "loss": 0.5918,
+      "step": 7126
+    },
+    {
+      "epoch": 1.8920748705694943,
+      "grad_norm": 0.41413698412971367,
+      "learning_rate": 3.889796328392842e-06,
+      "loss": 0.5882,
+      "step": 7127
+    },
+    {
+      "epoch": 1.892340369042878,
+      "grad_norm": 0.41863856537905497,
+      "learning_rate": 3.8895061068751135e-06,
+      "loss": 0.5854,
+      "step": 7128
+    },
+    {
+      "epoch": 1.8926058675162618,
+      "grad_norm": 0.40670889262518367,
+      "learning_rate": 3.889215858258616e-06,
+      "loss": 0.5806,
+      "step": 7129
+    },
+    {
+      "epoch": 1.8928713659896457,
+      "grad_norm": 0.4153909377778881,
+      "learning_rate": 3.888925582549006e-06,
+      "loss": 0.6093,
+      "step": 7130
+    },
+    {
+      "epoch": 1.8931368644630293,
+      "grad_norm": 0.4258020703053314,
+      "learning_rate": 3.888635279751947e-06,
+      "loss": 0.5642,
+      "step": 7131
+    },
+    {
+      "epoch": 1.893402362936413,
+      "grad_norm": 0.415145681811216,
+      "learning_rate": 3.8883449498731e-06,
+      "loss": 0.5801,
+      "step": 7132
+    },
+    {
+      "epoch": 1.893667861409797,
+      "grad_norm": 0.4192543394298992,
+      "learning_rate": 3.8880545929181275e-06,
+      "loss": 0.5722,
+      "step": 7133
+    },
+    {
+      "epoch": 1.8939333598831807,
+      "grad_norm": 0.4102842553446317,
+      "learning_rate": 3.8877642088926915e-06,
+      "loss": 0.5668,
+      "step": 7134
+    },
+    {
+      "epoch": 1.8941988583565643,
+      "grad_norm": 0.408315022426324,
+      "learning_rate": 3.887473797802456e-06,
+      "loss": 0.5662,
+      "step": 7135
+    },
+    {
+      "epoch": 1.8944643568299482,
+      "grad_norm": 0.4280394709291742,
+      "learning_rate": 3.887183359653084e-06,
+      "loss": 0.5617,
+      "step": 7136
+    },
+    {
+      "epoch": 1.894729855303332,
+      "grad_norm": 0.41449501383434817,
+      "learning_rate": 3.88689289445024e-06,
+      "loss": 0.6042,
+      "step": 7137
+    },
+    {
+      "epoch": 1.8949953537767157,
+      "grad_norm": 0.4119354111836354,
+      "learning_rate": 3.88660240219959e-06,
+      "loss": 0.6229,
+      "step": 7138
+    },
+    {
+      "epoch": 1.8952608522500995,
+      "grad_norm": 0.42889587467980234,
+      "learning_rate": 3.886311882906797e-06,
+      "loss": 0.5963,
+      "step": 7139
+    },
+    {
+      "epoch": 1.8955263507234834,
+      "grad_norm": 0.42957855699604774,
+      "learning_rate": 3.886021336577528e-06,
+      "loss": 0.6243,
+      "step": 7140
+    },
+    {
+      "epoch": 1.895791849196867,
+      "grad_norm": 0.3983587462236109,
+      "learning_rate": 3.88573076321745e-06,
+      "loss": 0.5632,
+      "step": 7141
+    },
+    {
+      "epoch": 1.896057347670251,
+      "grad_norm": 0.4088288548310979,
+      "learning_rate": 3.885440162832228e-06,
+      "loss": 0.5727,
+      "step": 7142
+    },
+    {
+      "epoch": 1.8963228461436348,
+      "grad_norm": 0.40718356625853985,
+      "learning_rate": 3.885149535427532e-06,
+      "loss": 0.5832,
+      "step": 7143
+    },
+    {
+      "epoch": 1.8965883446170184,
+      "grad_norm": 0.4079566258436637,
+      "learning_rate": 3.884858881009029e-06,
+      "loss": 0.5967,
+      "step": 7144
+    },
+    {
+      "epoch": 1.8968538430904023,
+      "grad_norm": 0.4313563491976267,
+      "learning_rate": 3.8845681995823855e-06,
+      "loss": 0.5713,
+      "step": 7145
+    },
+    {
+      "epoch": 1.8971193415637861,
+      "grad_norm": 0.40342349424624524,
+      "learning_rate": 3.884277491153273e-06,
+      "loss": 0.5959,
+      "step": 7146
+    },
+    {
+      "epoch": 1.8973848400371698,
+      "grad_norm": 0.4015402096251144,
+      "learning_rate": 3.883986755727361e-06,
+      "loss": 0.5433,
+      "step": 7147
+    },
+    {
+      "epoch": 1.8976503385105534,
+      "grad_norm": 0.39872304151407295,
+      "learning_rate": 3.883695993310317e-06,
+      "loss": 0.5821,
+      "step": 7148
+    },
+    {
+      "epoch": 1.8979158369839373,
+      "grad_norm": 0.40450619407227917,
+      "learning_rate": 3.883405203907814e-06,
+      "loss": 0.5778,
+      "step": 7149
+    },
+    {
+      "epoch": 1.8981813354573212,
+      "grad_norm": 0.4206019621409669,
+      "learning_rate": 3.883114387525522e-06,
+      "loss": 0.5926,
+      "step": 7150
+    },
+    {
+      "epoch": 1.8984468339307048,
+      "grad_norm": 0.4187351264285589,
+      "learning_rate": 3.882823544169114e-06,
+      "loss": 0.563,
+      "step": 7151
+    },
+    {
+      "epoch": 1.8987123324040887,
+      "grad_norm": 0.40080680699785154,
+      "learning_rate": 3.8825326738442605e-06,
+      "loss": 0.6069,
+      "step": 7152
+    },
+    {
+      "epoch": 1.8989778308774725,
+      "grad_norm": 0.41560257005499435,
+      "learning_rate": 3.882241776556634e-06,
+      "loss": 0.5726,
+      "step": 7153
+    },
+    {
+      "epoch": 1.8992433293508562,
+      "grad_norm": 0.43318930252869076,
+      "learning_rate": 3.88195085231191e-06,
+      "loss": 0.5534,
+      "step": 7154
+    },
+    {
+      "epoch": 1.89950882782424,
+      "grad_norm": 0.4056416871609871,
+      "learning_rate": 3.8816599011157595e-06,
+      "loss": 0.6022,
+      "step": 7155
+    },
+    {
+      "epoch": 1.8997743262976239,
+      "grad_norm": 0.40568965832114234,
+      "learning_rate": 3.8813689229738595e-06,
+      "loss": 0.5757,
+      "step": 7156
+    },
+    {
+      "epoch": 1.9000398247710075,
+      "grad_norm": 0.41616048594392585,
+      "learning_rate": 3.881077917891882e-06,
+      "loss": 0.5699,
+      "step": 7157
+    },
+    {
+      "epoch": 1.9003053232443914,
+      "grad_norm": 0.42909154285676004,
+      "learning_rate": 3.880786885875505e-06,
+      "loss": 0.5895,
+      "step": 7158
+    },
+    {
+      "epoch": 1.9005708217177752,
+      "grad_norm": 0.4225142793651282,
+      "learning_rate": 3.880495826930402e-06,
+      "loss": 0.5605,
+      "step": 7159
+    },
+    {
+      "epoch": 1.900836320191159,
+      "grad_norm": 0.4052258347579196,
+      "learning_rate": 3.88020474106225e-06,
+      "loss": 0.5623,
+      "step": 7160
+    },
+    {
+      "epoch": 1.9011018186645425,
+      "grad_norm": 0.416099027524383,
+      "learning_rate": 3.879913628276727e-06,
+      "loss": 0.5802,
+      "step": 7161
+    },
+    {
+      "epoch": 1.9013673171379266,
+      "grad_norm": 0.40948217753710037,
+      "learning_rate": 3.879622488579509e-06,
+      "loss": 0.5894,
+      "step": 7162
+    },
+    {
+      "epoch": 1.9016328156113103,
+      "grad_norm": 0.41299216749806356,
+      "learning_rate": 3.879331321976275e-06,
+      "loss": 0.6044,
+      "step": 7163
+    },
+    {
+      "epoch": 1.901898314084694,
+      "grad_norm": 0.40904011412817093,
+      "learning_rate": 3.879040128472703e-06,
+      "loss": 0.5801,
+      "step": 7164
+    },
+    {
+      "epoch": 1.9021638125580778,
+      "grad_norm": 0.4023578922308481,
+      "learning_rate": 3.878748908074472e-06,
+      "loss": 0.5735,
+      "step": 7165
+    },
+    {
+      "epoch": 1.9024293110314616,
+      "grad_norm": 0.43283246343719994,
+      "learning_rate": 3.878457660787262e-06,
+      "loss": 0.5849,
+      "step": 7166
+    },
+    {
+      "epoch": 1.9026948095048453,
+      "grad_norm": 0.40883413686724884,
+      "learning_rate": 3.878166386616752e-06,
+      "loss": 0.5444,
+      "step": 7167
+    },
+    {
+      "epoch": 1.9029603079782291,
+      "grad_norm": 0.40954773542159334,
+      "learning_rate": 3.8778750855686224e-06,
+      "loss": 0.5801,
+      "step": 7168
+    },
+    {
+      "epoch": 1.903225806451613,
+      "grad_norm": 0.4047815189510977,
+      "learning_rate": 3.877583757648556e-06,
+      "loss": 0.601,
+      "step": 7169
+    },
+    {
+      "epoch": 1.9034913049249966,
+      "grad_norm": 0.4209748280684184,
+      "learning_rate": 3.877292402862233e-06,
+      "loss": 0.5653,
+      "step": 7170
+    },
+    {
+      "epoch": 1.9037568033983805,
+      "grad_norm": 0.410220557831242,
+      "learning_rate": 3.877001021215337e-06,
+      "loss": 0.5674,
+      "step": 7171
+    },
+    {
+      "epoch": 1.9040223018717644,
+      "grad_norm": 0.41436491621168636,
+      "learning_rate": 3.876709612713548e-06,
+      "loss": 0.5737,
+      "step": 7172
+    },
+    {
+      "epoch": 1.904287800345148,
+      "grad_norm": 0.4252049888717145,
+      "learning_rate": 3.876418177362552e-06,
+      "loss": 0.61,
+      "step": 7173
+    },
+    {
+      "epoch": 1.9045532988185316,
+      "grad_norm": 0.40611264795670127,
+      "learning_rate": 3.876126715168031e-06,
+      "loss": 0.5285,
+      "step": 7174
+    },
+    {
+      "epoch": 1.9048187972919157,
+      "grad_norm": 0.40313354278754837,
+      "learning_rate": 3.87583522613567e-06,
+      "loss": 0.6079,
+      "step": 7175
+    },
+    {
+      "epoch": 1.9050842957652994,
+      "grad_norm": 0.41684222460139536,
+      "learning_rate": 3.875543710271153e-06,
+      "loss": 0.6068,
+      "step": 7176
+    },
+    {
+      "epoch": 1.905349794238683,
+      "grad_norm": 0.4214278278462318,
+      "learning_rate": 3.875252167580166e-06,
+      "loss": 0.5499,
+      "step": 7177
+    },
+    {
+      "epoch": 1.9056152927120669,
+      "grad_norm": 0.4101153577327562,
+      "learning_rate": 3.874960598068394e-06,
+      "loss": 0.5739,
+      "step": 7178
+    },
+    {
+      "epoch": 1.9058807911854507,
+      "grad_norm": 0.41426093026913,
+      "learning_rate": 3.874669001741525e-06,
+      "loss": 0.5946,
+      "step": 7179
+    },
+    {
+      "epoch": 1.9061462896588344,
+      "grad_norm": 0.40268257071900093,
+      "learning_rate": 3.8743773786052444e-06,
+      "loss": 0.5602,
+      "step": 7180
+    },
+    {
+      "epoch": 1.9064117881322182,
+      "grad_norm": 0.4120395510416962,
+      "learning_rate": 3.874085728665239e-06,
+      "loss": 0.5847,
+      "step": 7181
+    },
+    {
+      "epoch": 1.906677286605602,
+      "grad_norm": 0.41739352348945763,
+      "learning_rate": 3.8737940519271985e-06,
+      "loss": 0.5723,
+      "step": 7182
+    },
+    {
+      "epoch": 1.9069427850789857,
+      "grad_norm": 0.39973271480277534,
+      "learning_rate": 3.873502348396809e-06,
+      "loss": 0.5676,
+      "step": 7183
+    },
+    {
+      "epoch": 1.9072082835523696,
+      "grad_norm": 0.4089954045043844,
+      "learning_rate": 3.873210618079762e-06,
+      "loss": 0.5866,
+      "step": 7184
+    },
+    {
+      "epoch": 1.9074737820257535,
+      "grad_norm": 0.4219370075883623,
+      "learning_rate": 3.872918860981746e-06,
+      "loss": 0.605,
+      "step": 7185
+    },
+    {
+      "epoch": 1.907739280499137,
+      "grad_norm": 0.41240840288629904,
+      "learning_rate": 3.87262707710845e-06,
+      "loss": 0.557,
+      "step": 7186
+    },
+    {
+      "epoch": 1.9080047789725207,
+      "grad_norm": 0.3991788954103238,
+      "learning_rate": 3.872335266465566e-06,
+      "loss": 0.5306,
+      "step": 7187
+    },
+    {
+      "epoch": 1.9082702774459048,
+      "grad_norm": 0.4198281710401784,
+      "learning_rate": 3.872043429058783e-06,
+      "loss": 0.5764,
+      "step": 7188
+    },
+    {
+      "epoch": 1.9085357759192885,
+      "grad_norm": 0.40761366136593047,
+      "learning_rate": 3.871751564893794e-06,
+      "loss": 0.5876,
+      "step": 7189
+    },
+    {
+      "epoch": 1.908801274392672,
+      "grad_norm": 0.4151686069673462,
+      "learning_rate": 3.8714596739762926e-06,
+      "loss": 0.6116,
+      "step": 7190
+    },
+    {
+      "epoch": 1.909066772866056,
+      "grad_norm": 0.4120251930310793,
+      "learning_rate": 3.871167756311968e-06,
+      "loss": 0.5752,
+      "step": 7191
+    },
+    {
+      "epoch": 1.9093322713394398,
+      "grad_norm": 0.3938978993848605,
+      "learning_rate": 3.870875811906515e-06,
+      "loss": 0.5936,
+      "step": 7192
+    },
+    {
+      "epoch": 1.9095977698128235,
+      "grad_norm": 0.41884482446354926,
+      "learning_rate": 3.870583840765628e-06,
+      "loss": 0.5622,
+      "step": 7193
+    },
+    {
+      "epoch": 1.9098632682862073,
+      "grad_norm": 0.4100973745888629,
+      "learning_rate": 3.8702918428949994e-06,
+      "loss": 0.5504,
+      "step": 7194
+    },
+    {
+      "epoch": 1.9101287667595912,
+      "grad_norm": 0.4072821592512897,
+      "learning_rate": 3.869999818300326e-06,
+      "loss": 0.5643,
+      "step": 7195
+    },
+    {
+      "epoch": 1.9103942652329748,
+      "grad_norm": 0.41219489722838315,
+      "learning_rate": 3.869707766987301e-06,
+      "loss": 0.6151,
+      "step": 7196
+    },
+    {
+      "epoch": 1.9106597637063587,
+      "grad_norm": 0.417313775674746,
+      "learning_rate": 3.869415688961621e-06,
+      "loss": 0.5773,
+      "step": 7197
+    },
+    {
+      "epoch": 1.9109252621797426,
+      "grad_norm": 0.4041787787380582,
+      "learning_rate": 3.869123584228982e-06,
+      "loss": 0.5854,
+      "step": 7198
+    },
+    {
+      "epoch": 1.9111907606531262,
+      "grad_norm": 0.4236120671735632,
+      "learning_rate": 3.868831452795081e-06,
+      "loss": 0.5773,
+      "step": 7199
+    },
+    {
+      "epoch": 1.91145625912651,
+      "grad_norm": 0.40693164692222916,
+      "learning_rate": 3.8685392946656145e-06,
+      "loss": 0.6105,
+      "step": 7200
+    },
+    {
+      "epoch": 1.911721757599894,
+      "grad_norm": 0.40337635707669184,
+      "learning_rate": 3.868247109846282e-06,
+      "loss": 0.566,
+      "step": 7201
+    },
+    {
+      "epoch": 1.9119872560732776,
+      "grad_norm": 0.4172642275968336,
+      "learning_rate": 3.86795489834278e-06,
+      "loss": 0.5838,
+      "step": 7202
+    },
+    {
+      "epoch": 1.9122527545466612,
+      "grad_norm": 0.40433774663092875,
+      "learning_rate": 3.867662660160809e-06,
+      "loss": 0.5072,
+      "step": 7203
+    },
+    {
+      "epoch": 1.9125182530200453,
+      "grad_norm": 0.3943275096641571,
+      "learning_rate": 3.8673703953060685e-06,
+      "loss": 0.567,
+      "step": 7204
+    },
+    {
+      "epoch": 1.912783751493429,
+      "grad_norm": 0.41790873958201535,
+      "learning_rate": 3.867078103784256e-06,
+      "loss": 0.6085,
+      "step": 7205
+    },
+    {
+      "epoch": 1.9130492499668126,
+      "grad_norm": 0.4233498400202297,
+      "learning_rate": 3.866785785601073e-06,
+      "loss": 0.5965,
+      "step": 7206
+    },
+    {
+      "epoch": 1.9133147484401964,
+      "grad_norm": 0.4231518175194749,
+      "learning_rate": 3.8664934407622205e-06,
+      "loss": 0.5811,
+      "step": 7207
+    },
+    {
+      "epoch": 1.9135802469135803,
+      "grad_norm": 0.40647426726853086,
+      "learning_rate": 3.866201069273401e-06,
+      "loss": 0.6193,
+      "step": 7208
+    },
+    {
+      "epoch": 1.913845745386964,
+      "grad_norm": 0.4248919244281186,
+      "learning_rate": 3.865908671140315e-06,
+      "loss": 0.5679,
+      "step": 7209
+    },
+    {
+      "epoch": 1.9141112438603478,
+      "grad_norm": 0.4042962506070454,
+      "learning_rate": 3.865616246368667e-06,
+      "loss": 0.5449,
+      "step": 7210
+    },
+    {
+      "epoch": 1.9143767423337317,
+      "grad_norm": 0.40072582673144136,
+      "learning_rate": 3.865323794964156e-06,
+      "loss": 0.5574,
+      "step": 7211
+    },
+    {
+      "epoch": 1.9146422408071153,
+      "grad_norm": 0.4110386395783789,
+      "learning_rate": 3.86503131693249e-06,
+      "loss": 0.5768,
+      "step": 7212
+    },
+    {
+      "epoch": 1.9149077392804992,
+      "grad_norm": 0.4209434484598474,
+      "learning_rate": 3.864738812279372e-06,
+      "loss": 0.5626,
+      "step": 7213
+    },
+    {
+      "epoch": 1.915173237753883,
+      "grad_norm": 0.4026041665023483,
+      "learning_rate": 3.8644462810105035e-06,
+      "loss": 0.6029,
+      "step": 7214
+    },
+    {
+      "epoch": 1.9154387362272667,
+      "grad_norm": 0.41299039364937057,
+      "learning_rate": 3.8641537231315935e-06,
+      "loss": 0.6005,
+      "step": 7215
+    },
+    {
+      "epoch": 1.9157042347006503,
+      "grad_norm": 0.4173446802498935,
+      "learning_rate": 3.863861138648345e-06,
+      "loss": 0.6042,
+      "step": 7216
+    },
+    {
+      "epoch": 1.9159697331740344,
+      "grad_norm": 0.41590426949559445,
+      "learning_rate": 3.863568527566465e-06,
+      "loss": 0.5762,
+      "step": 7217
+    },
+    {
+      "epoch": 1.916235231647418,
+      "grad_norm": 0.4075294254854495,
+      "learning_rate": 3.863275889891661e-06,
+      "loss": 0.5298,
+      "step": 7218
+    },
+    {
+      "epoch": 1.9165007301208017,
+      "grad_norm": 0.4076307919559676,
+      "learning_rate": 3.862983225629638e-06,
+      "loss": 0.5942,
+      "step": 7219
+    },
+    {
+      "epoch": 1.9167662285941856,
+      "grad_norm": 0.41292162118299036,
+      "learning_rate": 3.862690534786105e-06,
+      "loss": 0.5463,
+      "step": 7220
+    },
+    {
+      "epoch": 1.9170317270675694,
+      "grad_norm": 0.40924475782259645,
+      "learning_rate": 3.862397817366771e-06,
+      "loss": 0.6026,
+      "step": 7221
+    },
+    {
+      "epoch": 1.917297225540953,
+      "grad_norm": 0.41077001876865915,
+      "learning_rate": 3.862105073377344e-06,
+      "loss": 0.5265,
+      "step": 7222
+    },
+    {
+      "epoch": 1.917562724014337,
+      "grad_norm": 0.4157999763905054,
+      "learning_rate": 3.861812302823532e-06,
+      "loss": 0.5912,
+      "step": 7223
+    },
+    {
+      "epoch": 1.9178282224877208,
+      "grad_norm": 0.41209908263462247,
+      "learning_rate": 3.861519505711047e-06,
+      "loss": 0.5868,
+      "step": 7224
+    },
+    {
+      "epoch": 1.9180937209611044,
+      "grad_norm": 0.41152275665315047,
+      "learning_rate": 3.861226682045597e-06,
+      "loss": 0.5714,
+      "step": 7225
+    },
+    {
+      "epoch": 1.9183592194344883,
+      "grad_norm": 0.39744687340693086,
+      "learning_rate": 3.860933831832894e-06,
+      "loss": 0.6113,
+      "step": 7226
+    },
+    {
+      "epoch": 1.9186247179078721,
+      "grad_norm": 0.4188903427788763,
+      "learning_rate": 3.860640955078649e-06,
+      "loss": 0.6169,
+      "step": 7227
+    },
+    {
+      "epoch": 1.9188902163812558,
+      "grad_norm": 0.4197760929248185,
+      "learning_rate": 3.860348051788575e-06,
+      "loss": 0.5944,
+      "step": 7228
+    },
+    {
+      "epoch": 1.9191557148546394,
+      "grad_norm": 0.4179022404322318,
+      "learning_rate": 3.860055121968382e-06,
+      "loss": 0.5663,
+      "step": 7229
+    },
+    {
+      "epoch": 1.9194212133280235,
+      "grad_norm": 0.4126394658987617,
+      "learning_rate": 3.859762165623785e-06,
+      "loss": 0.5862,
+      "step": 7230
+    },
+    {
+      "epoch": 1.9196867118014072,
+      "grad_norm": 0.4190873015660085,
+      "learning_rate": 3.859469182760496e-06,
+      "loss": 0.6027,
+      "step": 7231
+    },
+    {
+      "epoch": 1.9199522102747908,
+      "grad_norm": 0.40684845842720996,
+      "learning_rate": 3.8591761733842295e-06,
+      "loss": 0.5854,
+      "step": 7232
+    },
+    {
+      "epoch": 1.9202177087481747,
+      "grad_norm": 0.4156621535515605,
+      "learning_rate": 3.8588831375007e-06,
+      "loss": 0.5641,
+      "step": 7233
+    },
+    {
+      "epoch": 1.9204832072215585,
+      "grad_norm": 0.4109784466585704,
+      "learning_rate": 3.858590075115623e-06,
+      "loss": 0.5998,
+      "step": 7234
+    },
+    {
+      "epoch": 1.9207487056949422,
+      "grad_norm": 0.40599372124872196,
+      "learning_rate": 3.858296986234711e-06,
+      "loss": 0.5791,
+      "step": 7235
+    },
+    {
+      "epoch": 1.921014204168326,
+      "grad_norm": 0.42788123544363277,
+      "learning_rate": 3.858003870863684e-06,
+      "loss": 0.544,
+      "step": 7236
+    },
+    {
+      "epoch": 1.9212797026417099,
+      "grad_norm": 0.4058644815726423,
+      "learning_rate": 3.857710729008256e-06,
+      "loss": 0.5603,
+      "step": 7237
+    },
+    {
+      "epoch": 1.9215452011150935,
+      "grad_norm": 0.4137391214144147,
+      "learning_rate": 3.857417560674144e-06,
+      "loss": 0.5525,
+      "step": 7238
+    },
+    {
+      "epoch": 1.9218106995884774,
+      "grad_norm": 0.4048127906917372,
+      "learning_rate": 3.857124365867066e-06,
+      "loss": 0.5757,
+      "step": 7239
+    },
+    {
+      "epoch": 1.9220761980618613,
+      "grad_norm": 0.41206914099775677,
+      "learning_rate": 3.8568311445927405e-06,
+      "loss": 0.5959,
+      "step": 7240
+    },
+    {
+      "epoch": 1.922341696535245,
+      "grad_norm": 0.423977109176925,
+      "learning_rate": 3.856537896856887e-06,
+      "loss": 0.5684,
+      "step": 7241
+    },
+    {
+      "epoch": 1.9226071950086288,
+      "grad_norm": 0.4226704449740402,
+      "learning_rate": 3.8562446226652215e-06,
+      "loss": 0.5879,
+      "step": 7242
+    },
+    {
+      "epoch": 1.9228726934820126,
+      "grad_norm": 0.41120450032532635,
+      "learning_rate": 3.855951322023465e-06,
+      "loss": 0.5457,
+      "step": 7243
+    },
+    {
+      "epoch": 1.9231381919553963,
+      "grad_norm": 0.40977560259430057,
+      "learning_rate": 3.855657994937339e-06,
+      "loss": 0.5688,
+      "step": 7244
+    },
+    {
+      "epoch": 1.92340369042878,
+      "grad_norm": 0.40903560979960996,
+      "learning_rate": 3.855364641412561e-06,
+      "loss": 0.5836,
+      "step": 7245
+    },
+    {
+      "epoch": 1.9236691889021638,
+      "grad_norm": 0.4013008960528774,
+      "learning_rate": 3.855071261454855e-06,
+      "loss": 0.5365,
+      "step": 7246
+    },
+    {
+      "epoch": 1.9239346873755476,
+      "grad_norm": 0.4381337528387058,
+      "learning_rate": 3.854777855069942e-06,
+      "loss": 0.5856,
+      "step": 7247
+    },
+    {
+      "epoch": 1.9242001858489313,
+      "grad_norm": 0.41795327366659757,
+      "learning_rate": 3.854484422263544e-06,
+      "loss": 0.5938,
+      "step": 7248
+    },
+    {
+      "epoch": 1.9244656843223151,
+      "grad_norm": 0.4187167192709015,
+      "learning_rate": 3.8541909630413825e-06,
+      "loss": 0.5443,
+      "step": 7249
+    },
+    {
+      "epoch": 1.924731182795699,
+      "grad_norm": 0.40990239990344823,
+      "learning_rate": 3.853897477409182e-06,
+      "loss": 0.5993,
+      "step": 7250
+    },
+    {
+      "epoch": 1.9249966812690826,
+      "grad_norm": 0.41379049674742924,
+      "learning_rate": 3.853603965372665e-06,
+      "loss": 0.5921,
+      "step": 7251
+    },
+    {
+      "epoch": 1.9252621797424665,
+      "grad_norm": 0.40372316307430534,
+      "learning_rate": 3.853310426937557e-06,
+      "loss": 0.5946,
+      "step": 7252
+    },
+    {
+      "epoch": 1.9255276782158504,
+      "grad_norm": 0.4219761214784973,
+      "learning_rate": 3.853016862109583e-06,
+      "loss": 0.5674,
+      "step": 7253
+    },
+    {
+      "epoch": 1.925793176689234,
+      "grad_norm": 0.407804451793417,
+      "learning_rate": 3.852723270894467e-06,
+      "loss": 0.5489,
+      "step": 7254
+    },
+    {
+      "epoch": 1.9260586751626179,
+      "grad_norm": 0.44483804256399523,
+      "learning_rate": 3.852429653297934e-06,
+      "loss": 0.5722,
+      "step": 7255
+    },
+    {
+      "epoch": 1.9263241736360017,
+      "grad_norm": 0.41375184177786095,
+      "learning_rate": 3.852136009325713e-06,
+      "loss": 0.5655,
+      "step": 7256
+    },
+    {
+      "epoch": 1.9265896721093854,
+      "grad_norm": 0.40349311891037937,
+      "learning_rate": 3.851842338983529e-06,
+      "loss": 0.5718,
+      "step": 7257
+    },
+    {
+      "epoch": 1.926855170582769,
+      "grad_norm": 0.42041434968372976,
+      "learning_rate": 3.851548642277109e-06,
+      "loss": 0.5895,
+      "step": 7258
+    },
+    {
+      "epoch": 1.927120669056153,
+      "grad_norm": 0.4037138590739769,
+      "learning_rate": 3.851254919212182e-06,
+      "loss": 0.5639,
+      "step": 7259
+    },
+    {
+      "epoch": 1.9273861675295367,
+      "grad_norm": 0.39997450320468025,
+      "learning_rate": 3.850961169794475e-06,
+      "loss": 0.5952,
+      "step": 7260
+    },
+    {
+      "epoch": 1.9276516660029204,
+      "grad_norm": 0.4084183576950089,
+      "learning_rate": 3.850667394029718e-06,
+      "loss": 0.6001,
+      "step": 7261
+    },
+    {
+      "epoch": 1.9279171644763042,
+      "grad_norm": 0.41479993199213955,
+      "learning_rate": 3.850373591923639e-06,
+      "loss": 0.5794,
+      "step": 7262
+    },
+    {
+      "epoch": 1.928182662949688,
+      "grad_norm": 0.4210915276461565,
+      "learning_rate": 3.850079763481969e-06,
+      "loss": 0.6374,
+      "step": 7263
+    },
+    {
+      "epoch": 1.9284481614230717,
+      "grad_norm": 0.41585093653916605,
+      "learning_rate": 3.849785908710438e-06,
+      "loss": 0.6002,
+      "step": 7264
+    },
+    {
+      "epoch": 1.9287136598964556,
+      "grad_norm": 0.4031317207835994,
+      "learning_rate": 3.849492027614777e-06,
+      "loss": 0.5674,
+      "step": 7265
+    },
+    {
+      "epoch": 1.9289791583698395,
+      "grad_norm": 0.41265470691605544,
+      "learning_rate": 3.8491981202007175e-06,
+      "loss": 0.5441,
+      "step": 7266
+    },
+    {
+      "epoch": 1.929244656843223,
+      "grad_norm": 0.41596383701151046,
+      "learning_rate": 3.8489041864739914e-06,
+      "loss": 0.5913,
+      "step": 7267
+    },
+    {
+      "epoch": 1.929510155316607,
+      "grad_norm": 0.4041481327039089,
+      "learning_rate": 3.84861022644033e-06,
+      "loss": 0.5745,
+      "step": 7268
+    },
+    {
+      "epoch": 1.9297756537899908,
+      "grad_norm": 0.42350082351462753,
+      "learning_rate": 3.848316240105469e-06,
+      "loss": 0.6313,
+      "step": 7269
+    },
+    {
+      "epoch": 1.9300411522633745,
+      "grad_norm": 0.40789671091642965,
+      "learning_rate": 3.8480222274751386e-06,
+      "loss": 0.5952,
+      "step": 7270
+    },
+    {
+      "epoch": 1.9303066507367581,
+      "grad_norm": 0.41692011131455387,
+      "learning_rate": 3.847728188555074e-06,
+      "loss": 0.5473,
+      "step": 7271
+    },
+    {
+      "epoch": 1.9305721492101422,
+      "grad_norm": 0.4214585485488159,
+      "learning_rate": 3.847434123351011e-06,
+      "loss": 0.574,
+      "step": 7272
+    },
+    {
+      "epoch": 1.9308376476835258,
+      "grad_norm": 0.3976340395555284,
+      "learning_rate": 3.847140031868683e-06,
+      "loss": 0.5672,
+      "step": 7273
+    },
+    {
+      "epoch": 1.9311031461569095,
+      "grad_norm": 0.4136880234069284,
+      "learning_rate": 3.846845914113826e-06,
+      "loss": 0.5907,
+      "step": 7274
+    },
+    {
+      "epoch": 1.9313686446302933,
+      "grad_norm": 0.4202107596106995,
+      "learning_rate": 3.8465517700921766e-06,
+      "loss": 0.5998,
+      "step": 7275
+    },
+    {
+      "epoch": 1.9316341431036772,
+      "grad_norm": 0.42833834639822044,
+      "learning_rate": 3.8462575998094695e-06,
+      "loss": 0.5901,
+      "step": 7276
+    },
+    {
+      "epoch": 1.9318996415770608,
+      "grad_norm": 0.41506693799550215,
+      "learning_rate": 3.845963403271444e-06,
+      "loss": 0.5528,
+      "step": 7277
+    },
+    {
+      "epoch": 1.9321651400504447,
+      "grad_norm": 0.3990362379776796,
+      "learning_rate": 3.845669180483836e-06,
+      "loss": 0.577,
+      "step": 7278
+    },
+    {
+      "epoch": 1.9324306385238286,
+      "grad_norm": 0.4219362670948583,
+      "learning_rate": 3.845374931452385e-06,
+      "loss": 0.569,
+      "step": 7279
+    },
+    {
+      "epoch": 1.9326961369972122,
+      "grad_norm": 0.4143561297914913,
+      "learning_rate": 3.845080656182829e-06,
+      "loss": 0.5832,
+      "step": 7280
+    },
+    {
+      "epoch": 1.932961635470596,
+      "grad_norm": 0.40720635082135936,
+      "learning_rate": 3.844786354680906e-06,
+      "loss": 0.5841,
+      "step": 7281
+    },
+    {
+      "epoch": 1.93322713394398,
+      "grad_norm": 0.4026420910491919,
+      "learning_rate": 3.844492026952356e-06,
+      "loss": 0.6044,
+      "step": 7282
+    },
+    {
+      "epoch": 1.9334926324173636,
+      "grad_norm": 0.40863571946330857,
+      "learning_rate": 3.8441976730029205e-06,
+      "loss": 0.5766,
+      "step": 7283
+    },
+    {
+      "epoch": 1.9337581308907472,
+      "grad_norm": 0.432909068619264,
+      "learning_rate": 3.843903292838339e-06,
+      "loss": 0.5737,
+      "step": 7284
+    },
+    {
+      "epoch": 1.9340236293641313,
+      "grad_norm": 0.42447373542135586,
+      "learning_rate": 3.843608886464353e-06,
+      "loss": 0.5932,
+      "step": 7285
+    },
+    {
+      "epoch": 1.934289127837515,
+      "grad_norm": 0.405714743475015,
+      "learning_rate": 3.843314453886704e-06,
+      "loss": 0.5928,
+      "step": 7286
+    },
+    {
+      "epoch": 1.9345546263108986,
+      "grad_norm": 0.4230486993583029,
+      "learning_rate": 3.843019995111134e-06,
+      "loss": 0.534,
+      "step": 7287
+    },
+    {
+      "epoch": 1.9348201247842824,
+      "grad_norm": 0.41842226727641196,
+      "learning_rate": 3.842725510143387e-06,
+      "loss": 0.5724,
+      "step": 7288
+    },
+    {
+      "epoch": 1.9350856232576663,
+      "grad_norm": 0.431399790817969,
+      "learning_rate": 3.842430998989203e-06,
+      "loss": 0.5919,
+      "step": 7289
+    },
+    {
+      "epoch": 1.93535112173105,
+      "grad_norm": 0.42631657849719096,
+      "learning_rate": 3.842136461654329e-06,
+      "loss": 0.6032,
+      "step": 7290
+    },
+    {
+      "epoch": 1.9356166202044338,
+      "grad_norm": 0.39359174551393095,
+      "learning_rate": 3.841841898144507e-06,
+      "loss": 0.5687,
+      "step": 7291
+    },
+    {
+      "epoch": 1.9358821186778177,
+      "grad_norm": 0.42555020533302723,
+      "learning_rate": 3.841547308465483e-06,
+      "loss": 0.5823,
+      "step": 7292
+    },
+    {
+      "epoch": 1.9361476171512013,
+      "grad_norm": 0.41936295779697286,
+      "learning_rate": 3.841252692623002e-06,
+      "loss": 0.5861,
+      "step": 7293
+    },
+    {
+      "epoch": 1.9364131156245852,
+      "grad_norm": 0.39913802978986646,
+      "learning_rate": 3.840958050622809e-06,
+      "loss": 0.5512,
+      "step": 7294
+    },
+    {
+      "epoch": 1.936678614097969,
+      "grad_norm": 0.42064337016114994,
+      "learning_rate": 3.840663382470651e-06,
+      "loss": 0.5483,
+      "step": 7295
+    },
+    {
+      "epoch": 1.9369441125713527,
+      "grad_norm": 0.39942168559428537,
+      "learning_rate": 3.840368688172276e-06,
+      "loss": 0.5832,
+      "step": 7296
+    },
+    {
+      "epoch": 1.9372096110447365,
+      "grad_norm": 0.40085220384694215,
+      "learning_rate": 3.840073967733427e-06,
+      "loss": 0.5325,
+      "step": 7297
+    },
+    {
+      "epoch": 1.9374751095181204,
+      "grad_norm": 0.41087090139393423,
+      "learning_rate": 3.839779221159856e-06,
+      "loss": 0.5977,
+      "step": 7298
+    },
+    {
+      "epoch": 1.937740607991504,
+      "grad_norm": 0.41611621392507747,
+      "learning_rate": 3.839484448457311e-06,
+      "loss": 0.5487,
+      "step": 7299
+    },
+    {
+      "epoch": 1.9380061064648877,
+      "grad_norm": 0.40748647357914997,
+      "learning_rate": 3.839189649631538e-06,
+      "loss": 0.5579,
+      "step": 7300
+    },
+    {
+      "epoch": 1.9382716049382716,
+      "grad_norm": 0.4065733952357567,
+      "learning_rate": 3.8388948246882885e-06,
+      "loss": 0.5952,
+      "step": 7301
+    },
+    {
+      "epoch": 1.9385371034116554,
+      "grad_norm": 0.4162546430091261,
+      "learning_rate": 3.838599973633312e-06,
+      "loss": 0.5625,
+      "step": 7302
+    },
+    {
+      "epoch": 1.938802601885039,
+      "grad_norm": 0.4703268687495602,
+      "learning_rate": 3.838305096472358e-06,
+      "loss": 0.5804,
+      "step": 7303
+    },
+    {
+      "epoch": 1.939068100358423,
+      "grad_norm": 0.41770832090007876,
+      "learning_rate": 3.8380101932111775e-06,
+      "loss": 0.605,
+      "step": 7304
+    },
+    {
+      "epoch": 1.9393335988318068,
+      "grad_norm": 0.41438696227764454,
+      "learning_rate": 3.837715263855524e-06,
+      "loss": 0.5796,
+      "step": 7305
+    },
+    {
+      "epoch": 1.9395990973051904,
+      "grad_norm": 0.4578819497358745,
+      "learning_rate": 3.837420308411146e-06,
+      "loss": 0.5712,
+      "step": 7306
+    },
+    {
+      "epoch": 1.9398645957785743,
+      "grad_norm": 0.4286288086263282,
+      "learning_rate": 3.837125326883797e-06,
+      "loss": 0.5593,
+      "step": 7307
+    },
+    {
+      "epoch": 1.9401300942519581,
+      "grad_norm": 0.4055658793994388,
+      "learning_rate": 3.836830319279232e-06,
+      "loss": 0.5451,
+      "step": 7308
+    },
+    {
+      "epoch": 1.9403955927253418,
+      "grad_norm": 0.4236891770860224,
+      "learning_rate": 3.836535285603201e-06,
+      "loss": 0.6062,
+      "step": 7309
+    },
+    {
+      "epoch": 1.9406610911987257,
+      "grad_norm": 0.42610792745289283,
+      "learning_rate": 3.836240225861461e-06,
+      "loss": 0.5928,
+      "step": 7310
+    },
+    {
+      "epoch": 1.9409265896721095,
+      "grad_norm": 0.41706220234604163,
+      "learning_rate": 3.835945140059764e-06,
+      "loss": 0.6066,
+      "step": 7311
+    },
+    {
+      "epoch": 1.9411920881454932,
+      "grad_norm": 0.4757271152707851,
+      "learning_rate": 3.8356500282038655e-06,
+      "loss": 0.5494,
+      "step": 7312
+    },
+    {
+      "epoch": 1.9414575866188768,
+      "grad_norm": 0.41967861085295155,
+      "learning_rate": 3.835354890299522e-06,
+      "loss": 0.5801,
+      "step": 7313
+    },
+    {
+      "epoch": 1.9417230850922609,
+      "grad_norm": 0.4215218200591324,
+      "learning_rate": 3.835059726352488e-06,
+      "loss": 0.5693,
+      "step": 7314
+    },
+    {
+      "epoch": 1.9419885835656445,
+      "grad_norm": 0.4319910953381267,
+      "learning_rate": 3.834764536368521e-06,
+      "loss": 0.5452,
+      "step": 7315
+    },
+    {
+      "epoch": 1.9422540820390282,
+      "grad_norm": 0.4790875349315884,
+      "learning_rate": 3.834469320353378e-06,
+      "loss": 0.542,
+      "step": 7316
+    },
+    {
+      "epoch": 1.942519580512412,
+      "grad_norm": 0.418557472080714,
+      "learning_rate": 3.834174078312816e-06,
+      "loss": 0.6156,
+      "step": 7317
+    },
+    {
+      "epoch": 1.9427850789857959,
+      "grad_norm": 0.42487266979654265,
+      "learning_rate": 3.833878810252592e-06,
+      "loss": 0.5927,
+      "step": 7318
+    },
+    {
+      "epoch": 1.9430505774591795,
+      "grad_norm": 0.4671579322941817,
+      "learning_rate": 3.833583516178465e-06,
+      "loss": 0.5861,
+      "step": 7319
+    },
+    {
+      "epoch": 1.9433160759325634,
+      "grad_norm": 0.4521129657543881,
+      "learning_rate": 3.833288196096194e-06,
+      "loss": 0.5552,
+      "step": 7320
+    },
+    {
+      "epoch": 1.9435815744059473,
+      "grad_norm": 0.41080678976372376,
+      "learning_rate": 3.83299285001154e-06,
+      "loss": 0.595,
+      "step": 7321
+    },
+    {
+      "epoch": 1.943847072879331,
+      "grad_norm": 0.4057780120432856,
+      "learning_rate": 3.8326974779302605e-06,
+      "loss": 0.5358,
+      "step": 7322
+    },
+    {
+      "epoch": 1.9441125713527148,
+      "grad_norm": 0.42065625410479307,
+      "learning_rate": 3.8324020798581184e-06,
+      "loss": 0.5659,
+      "step": 7323
+    },
+    {
+      "epoch": 1.9443780698260986,
+      "grad_norm": 0.43485851420317584,
+      "learning_rate": 3.832106655800872e-06,
+      "loss": 0.5927,
+      "step": 7324
+    },
+    {
+      "epoch": 1.9446435682994823,
+      "grad_norm": 0.42274232802905415,
+      "learning_rate": 3.831811205764286e-06,
+      "loss": 0.5807,
+      "step": 7325
+    },
+    {
+      "epoch": 1.944909066772866,
+      "grad_norm": 0.42485315848677196,
+      "learning_rate": 3.83151572975412e-06,
+      "loss": 0.5638,
+      "step": 7326
+    },
+    {
+      "epoch": 1.94517456524625,
+      "grad_norm": 0.4208865440074488,
+      "learning_rate": 3.831220227776136e-06,
+      "loss": 0.5513,
+      "step": 7327
+    },
+    {
+      "epoch": 1.9454400637196336,
+      "grad_norm": 0.42121111915072895,
+      "learning_rate": 3.830924699836099e-06,
+      "loss": 0.5613,
+      "step": 7328
+    },
+    {
+      "epoch": 1.9457055621930173,
+      "grad_norm": 0.4239437252746114,
+      "learning_rate": 3.830629145939772e-06,
+      "loss": 0.5748,
+      "step": 7329
+    },
+    {
+      "epoch": 1.9459710606664011,
+      "grad_norm": 0.41232702374219565,
+      "learning_rate": 3.830333566092918e-06,
+      "loss": 0.5673,
+      "step": 7330
+    },
+    {
+      "epoch": 1.946236559139785,
+      "grad_norm": 0.42393449027238267,
+      "learning_rate": 3.830037960301303e-06,
+      "loss": 0.5655,
+      "step": 7331
+    },
+    {
+      "epoch": 1.9465020576131686,
+      "grad_norm": 0.4003471598463813,
+      "learning_rate": 3.829742328570691e-06,
+      "loss": 0.5824,
+      "step": 7332
+    },
+    {
+      "epoch": 1.9467675560865525,
+      "grad_norm": 0.42178886221355294,
+      "learning_rate": 3.829446670906848e-06,
+      "loss": 0.5879,
+      "step": 7333
+    },
+    {
+      "epoch": 1.9470330545599364,
+      "grad_norm": 0.4258184891800165,
+      "learning_rate": 3.82915098731554e-06,
+      "loss": 0.5587,
+      "step": 7334
+    },
+    {
+      "epoch": 1.94729855303332,
+      "grad_norm": 0.4323379824169179,
+      "learning_rate": 3.828855277802533e-06,
+      "loss": 0.539,
+      "step": 7335
+    },
+    {
+      "epoch": 1.9475640515067039,
+      "grad_norm": 0.408000934961471,
+      "learning_rate": 3.828559542373594e-06,
+      "loss": 0.579,
+      "step": 7336
+    },
+    {
+      "epoch": 1.9478295499800877,
+      "grad_norm": 0.3996728167826378,
+      "learning_rate": 3.828263781034492e-06,
+      "loss": 0.5707,
+      "step": 7337
+    },
+    {
+      "epoch": 1.9480950484534714,
+      "grad_norm": 0.4002980116914289,
+      "learning_rate": 3.827967993790994e-06,
+      "loss": 0.5923,
+      "step": 7338
+    },
+    {
+      "epoch": 1.948360546926855,
+      "grad_norm": 0.42657760748181905,
+      "learning_rate": 3.827672180648868e-06,
+      "loss": 0.5513,
+      "step": 7339
+    },
+    {
+      "epoch": 1.948626045400239,
+      "grad_norm": 0.42031230724856017,
+      "learning_rate": 3.827376341613884e-06,
+      "loss": 0.5797,
+      "step": 7340
+    },
+    {
+      "epoch": 1.9488915438736227,
+      "grad_norm": 0.41445161550057874,
+      "learning_rate": 3.8270804766918125e-06,
+      "loss": 0.6078,
+      "step": 7341
+    },
+    {
+      "epoch": 1.9491570423470064,
+      "grad_norm": 0.4178440241879826,
+      "learning_rate": 3.826784585888421e-06,
+      "loss": 0.6078,
+      "step": 7342
+    },
+    {
+      "epoch": 1.9494225408203902,
+      "grad_norm": 0.4132298969428845,
+      "learning_rate": 3.826488669209483e-06,
+      "loss": 0.6086,
+      "step": 7343
+    },
+    {
+      "epoch": 1.949688039293774,
+      "grad_norm": 0.40184854981401397,
+      "learning_rate": 3.826192726660767e-06,
+      "loss": 0.593,
+      "step": 7344
+    },
+    {
+      "epoch": 1.9499535377671577,
+      "grad_norm": 0.4267417598297509,
+      "learning_rate": 3.825896758248046e-06,
+      "loss": 0.6028,
+      "step": 7345
+    },
+    {
+      "epoch": 1.9502190362405416,
+      "grad_norm": 0.42069510722893005,
+      "learning_rate": 3.825600763977091e-06,
+      "loss": 0.6005,
+      "step": 7346
+    },
+    {
+      "epoch": 1.9504845347139255,
+      "grad_norm": 0.40830595833856426,
+      "learning_rate": 3.825304743853676e-06,
+      "loss": 0.5894,
+      "step": 7347
+    },
+    {
+      "epoch": 1.950750033187309,
+      "grad_norm": 0.4081436091207105,
+      "learning_rate": 3.825008697883574e-06,
+      "loss": 0.5825,
+      "step": 7348
+    },
+    {
+      "epoch": 1.951015531660693,
+      "grad_norm": 0.4236863647340996,
+      "learning_rate": 3.824712626072558e-06,
+      "loss": 0.5817,
+      "step": 7349
+    },
+    {
+      "epoch": 1.9512810301340768,
+      "grad_norm": 0.4310682385563616,
+      "learning_rate": 3.824416528426403e-06,
+      "loss": 0.5522,
+      "step": 7350
+    },
+    {
+      "epoch": 1.9515465286074605,
+      "grad_norm": 0.4084756942330639,
+      "learning_rate": 3.824120404950883e-06,
+      "loss": 0.5844,
+      "step": 7351
+    },
+    {
+      "epoch": 1.9518120270808443,
+      "grad_norm": 0.40926606038917146,
+      "learning_rate": 3.8238242556517725e-06,
+      "loss": 0.5794,
+      "step": 7352
+    },
+    {
+      "epoch": 1.9520775255542282,
+      "grad_norm": 0.42301127135326394,
+      "learning_rate": 3.823528080534847e-06,
+      "loss": 0.5991,
+      "step": 7353
+    },
+    {
+      "epoch": 1.9523430240276118,
+      "grad_norm": 0.4138429565378456,
+      "learning_rate": 3.823231879605885e-06,
+      "loss": 0.5498,
+      "step": 7354
+    },
+    {
+      "epoch": 1.9526085225009955,
+      "grad_norm": 0.39313562489739623,
+      "learning_rate": 3.8229356528706605e-06,
+      "loss": 0.575,
+      "step": 7355
+    },
+    {
+      "epoch": 1.9528740209743793,
+      "grad_norm": 0.41320416235125834,
+      "learning_rate": 3.822639400334952e-06,
+      "loss": 0.586,
+      "step": 7356
+    },
+    {
+      "epoch": 1.9531395194477632,
+      "grad_norm": 0.41092052923304684,
+      "learning_rate": 3.822343122004536e-06,
+      "loss": 0.5798,
+      "step": 7357
+    },
+    {
+      "epoch": 1.9534050179211468,
+      "grad_norm": 0.4064466354432723,
+      "learning_rate": 3.8220468178851925e-06,
+      "loss": 0.552,
+      "step": 7358
+    },
+    {
+      "epoch": 1.9536705163945307,
+      "grad_norm": 0.41978745107814436,
+      "learning_rate": 3.821750487982698e-06,
+      "loss": 0.5922,
+      "step": 7359
+    },
+    {
+      "epoch": 1.9539360148679146,
+      "grad_norm": 0.4120256704044728,
+      "learning_rate": 3.821454132302834e-06,
+      "loss": 0.5852,
+      "step": 7360
+    },
+    {
+      "epoch": 1.9542015133412982,
+      "grad_norm": 0.4190237629622631,
+      "learning_rate": 3.8211577508513784e-06,
+      "loss": 0.5776,
+      "step": 7361
+    },
+    {
+      "epoch": 1.954467011814682,
+      "grad_norm": 0.40117207222015244,
+      "learning_rate": 3.820861343634113e-06,
+      "loss": 0.574,
+      "step": 7362
+    },
+    {
+      "epoch": 1.954732510288066,
+      "grad_norm": 0.42058702842285467,
+      "learning_rate": 3.820564910656815e-06,
+      "loss": 0.5945,
+      "step": 7363
+    },
+    {
+      "epoch": 1.9549980087614496,
+      "grad_norm": 0.40265179966644293,
+      "learning_rate": 3.82026845192527e-06,
+      "loss": 0.5835,
+      "step": 7364
+    },
+    {
+      "epoch": 1.9552635072348334,
+      "grad_norm": 0.42302484410861746,
+      "learning_rate": 3.819971967445256e-06,
+      "loss": 0.5984,
+      "step": 7365
+    },
+    {
+      "epoch": 1.9555290057082173,
+      "grad_norm": 0.4035889090298571,
+      "learning_rate": 3.819675457222558e-06,
+      "loss": 0.5887,
+      "step": 7366
+    },
+    {
+      "epoch": 1.955794504181601,
+      "grad_norm": 0.4194479260135337,
+      "learning_rate": 3.8193789212629575e-06,
+      "loss": 0.6092,
+      "step": 7367
+    },
+    {
+      "epoch": 1.9560600026549846,
+      "grad_norm": 0.40413352930301033,
+      "learning_rate": 3.819082359572237e-06,
+      "loss": 0.595,
+      "step": 7368
+    },
+    {
+      "epoch": 1.9563255011283687,
+      "grad_norm": 0.4184555770271347,
+      "learning_rate": 3.818785772156181e-06,
+      "loss": 0.5598,
+      "step": 7369
+    },
+    {
+      "epoch": 1.9565909996017523,
+      "grad_norm": 0.4227090579697663,
+      "learning_rate": 3.818489159020573e-06,
+      "loss": 0.5609,
+      "step": 7370
+    },
+    {
+      "epoch": 1.956856498075136,
+      "grad_norm": 0.4088025144247253,
+      "learning_rate": 3.8181925201712e-06,
+      "loss": 0.5938,
+      "step": 7371
+    },
+    {
+      "epoch": 1.9571219965485198,
+      "grad_norm": 0.4180422561093033,
+      "learning_rate": 3.817895855613843e-06,
+      "loss": 0.6161,
+      "step": 7372
+    },
+    {
+      "epoch": 1.9573874950219037,
+      "grad_norm": 0.409885794078535,
+      "learning_rate": 3.817599165354292e-06,
+      "loss": 0.5921,
+      "step": 7373
+    },
+    {
+      "epoch": 1.9576529934952873,
+      "grad_norm": 0.4098468966114667,
+      "learning_rate": 3.8173024493983305e-06,
+      "loss": 0.5138,
+      "step": 7374
+    },
+    {
+      "epoch": 1.9579184919686712,
+      "grad_norm": 0.40772967420333917,
+      "learning_rate": 3.817005707751746e-06,
+      "loss": 0.619,
+      "step": 7375
+    },
+    {
+      "epoch": 1.958183990442055,
+      "grad_norm": 0.420969508034548,
+      "learning_rate": 3.816708940420324e-06,
+      "loss": 0.5709,
+      "step": 7376
+    },
+    {
+      "epoch": 1.9584494889154387,
+      "grad_norm": 0.41034849017729425,
+      "learning_rate": 3.816412147409856e-06,
+      "loss": 0.5908,
+      "step": 7377
+    },
+    {
+      "epoch": 1.9587149873888225,
+      "grad_norm": 0.407992208602232,
+      "learning_rate": 3.816115328726127e-06,
+      "loss": 0.596,
+      "step": 7378
+    },
+    {
+      "epoch": 1.9589804858622064,
+      "grad_norm": 0.4055234760779293,
+      "learning_rate": 3.815818484374927e-06,
+      "loss": 0.605,
+      "step": 7379
+    },
+    {
+      "epoch": 1.95924598433559,
+      "grad_norm": 0.4098993209244575,
+      "learning_rate": 3.815521614362046e-06,
+      "loss": 0.5998,
+      "step": 7380
+    },
+    {
+      "epoch": 1.9595114828089737,
+      "grad_norm": 0.4103240554784892,
+      "learning_rate": 3.81522471869327e-06,
+      "loss": 0.5734,
+      "step": 7381
+    },
+    {
+      "epoch": 1.9597769812823578,
+      "grad_norm": 0.40183648352397106,
+      "learning_rate": 3.814927797374393e-06,
+      "loss": 0.5776,
+      "step": 7382
+    },
+    {
+      "epoch": 1.9600424797557414,
+      "grad_norm": 0.4269050270802489,
+      "learning_rate": 3.814630850411205e-06,
+      "loss": 0.5676,
+      "step": 7383
+    },
+    {
+      "epoch": 1.960307978229125,
+      "grad_norm": 0.42774758345991337,
+      "learning_rate": 3.8143338778094963e-06,
+      "loss": 0.586,
+      "step": 7384
+    },
+    {
+      "epoch": 1.960573476702509,
+      "grad_norm": 0.4030192793021057,
+      "learning_rate": 3.8140368795750594e-06,
+      "loss": 0.5994,
+      "step": 7385
+    },
+    {
+      "epoch": 1.9608389751758928,
+      "grad_norm": 0.42557558089397585,
+      "learning_rate": 3.8137398557136852e-06,
+      "loss": 0.5457,
+      "step": 7386
+    },
+    {
+      "epoch": 1.9611044736492764,
+      "grad_norm": 0.4345490112943575,
+      "learning_rate": 3.813442806231167e-06,
+      "loss": 0.5549,
+      "step": 7387
+    },
+    {
+      "epoch": 1.9613699721226603,
+      "grad_norm": 0.419544798137635,
+      "learning_rate": 3.813145731133299e-06,
+      "loss": 0.5727,
+      "step": 7388
+    },
+    {
+      "epoch": 1.9616354705960442,
+      "grad_norm": 0.4180544259536902,
+      "learning_rate": 3.812848630425874e-06,
+      "loss": 0.5615,
+      "step": 7389
+    },
+    {
+      "epoch": 1.9619009690694278,
+      "grad_norm": 0.4114378365322946,
+      "learning_rate": 3.812551504114686e-06,
+      "loss": 0.5964,
+      "step": 7390
+    },
+    {
+      "epoch": 1.9621664675428117,
+      "grad_norm": 0.4224188280197434,
+      "learning_rate": 3.8122543522055303e-06,
+      "loss": 0.5948,
+      "step": 7391
+    },
+    {
+      "epoch": 1.9624319660161955,
+      "grad_norm": 0.42865940324872326,
+      "learning_rate": 3.811957174704201e-06,
+      "loss": 0.6207,
+      "step": 7392
+    },
+    {
+      "epoch": 1.9626974644895792,
+      "grad_norm": 0.43074459774381124,
+      "learning_rate": 3.8116599716164947e-06,
+      "loss": 0.5657,
+      "step": 7393
+    },
+    {
+      "epoch": 1.9629629629629628,
+      "grad_norm": 0.41221153172612696,
+      "learning_rate": 3.811362742948209e-06,
+      "loss": 0.5774,
+      "step": 7394
+    },
+    {
+      "epoch": 1.9632284614363469,
+      "grad_norm": 0.41844579762903084,
+      "learning_rate": 3.811065488705138e-06,
+      "loss": 0.5979,
+      "step": 7395
+    },
+    {
+      "epoch": 1.9634939599097305,
+      "grad_norm": 0.4212079361322282,
+      "learning_rate": 3.8107682088930797e-06,
+      "loss": 0.5965,
+      "step": 7396
+    },
+    {
+      "epoch": 1.9637594583831142,
+      "grad_norm": 0.42673139201545884,
+      "learning_rate": 3.810470903517832e-06,
+      "loss": 0.5969,
+      "step": 7397
+    },
+    {
+      "epoch": 1.964024956856498,
+      "grad_norm": 0.4032462664849555,
+      "learning_rate": 3.8101735725851937e-06,
+      "loss": 0.5742,
+      "step": 7398
+    },
+    {
+      "epoch": 1.964290455329882,
+      "grad_norm": 0.41851530448704083,
+      "learning_rate": 3.8098762161009624e-06,
+      "loss": 0.5887,
+      "step": 7399
+    },
+    {
+      "epoch": 1.9645559538032655,
+      "grad_norm": 0.40742839364011607,
+      "learning_rate": 3.809578834070938e-06,
+      "loss": 0.5805,
+      "step": 7400
+    },
+    {
+      "epoch": 1.9648214522766494,
+      "grad_norm": 0.4200450734704836,
+      "learning_rate": 3.809281426500919e-06,
+      "loss": 0.5806,
+      "step": 7401
+    },
+    {
+      "epoch": 1.9650869507500333,
+      "grad_norm": 0.3931166683697361,
+      "learning_rate": 3.8089839933967076e-06,
+      "loss": 0.5923,
+      "step": 7402
+    },
+    {
+      "epoch": 1.965352449223417,
+      "grad_norm": 0.4154707503645699,
+      "learning_rate": 3.8086865347641035e-06,
+      "loss": 0.5891,
+      "step": 7403
+    },
+    {
+      "epoch": 1.9656179476968008,
+      "grad_norm": 0.41784168985414244,
+      "learning_rate": 3.808389050608907e-06,
+      "loss": 0.5571,
+      "step": 7404
+    },
+    {
+      "epoch": 1.9658834461701846,
+      "grad_norm": 0.427872826689121,
+      "learning_rate": 3.8080915409369207e-06,
+      "loss": 0.6034,
+      "step": 7405
+    },
+    {
+      "epoch": 1.9661489446435683,
+      "grad_norm": 0.4186891550456353,
+      "learning_rate": 3.807794005753947e-06,
+      "loss": 0.5597,
+      "step": 7406
+    },
+    {
+      "epoch": 1.9664144431169521,
+      "grad_norm": 0.4142455121414843,
+      "learning_rate": 3.807496445065788e-06,
+      "loss": 0.5941,
+      "step": 7407
+    },
+    {
+      "epoch": 1.966679941590336,
+      "grad_norm": 0.416150868149289,
+      "learning_rate": 3.8071988588782466e-06,
+      "loss": 0.5948,
+      "step": 7408
+    },
+    {
+      "epoch": 1.9669454400637196,
+      "grad_norm": 0.43637830656131715,
+      "learning_rate": 3.8069012471971278e-06,
+      "loss": 0.6111,
+      "step": 7409
+    },
+    {
+      "epoch": 1.9672109385371033,
+      "grad_norm": 0.42120288693073155,
+      "learning_rate": 3.806603610028234e-06,
+      "loss": 0.5805,
+      "step": 7410
+    },
+    {
+      "epoch": 1.9674764370104874,
+      "grad_norm": 0.43376020874526666,
+      "learning_rate": 3.8063059473773705e-06,
+      "loss": 0.5339,
+      "step": 7411
+    },
+    {
+      "epoch": 1.967741935483871,
+      "grad_norm": 0.4124621309741732,
+      "learning_rate": 3.806008259250343e-06,
+      "loss": 0.5867,
+      "step": 7412
+    },
+    {
+      "epoch": 1.9680074339572546,
+      "grad_norm": 0.4206963355803583,
+      "learning_rate": 3.805710545652957e-06,
+      "loss": 0.5793,
+      "step": 7413
+    },
+    {
+      "epoch": 1.9682729324306385,
+      "grad_norm": 0.41916107887977166,
+      "learning_rate": 3.8054128065910186e-06,
+      "loss": 0.6174,
+      "step": 7414
+    },
+    {
+      "epoch": 1.9685384309040224,
+      "grad_norm": 0.424104433428084,
+      "learning_rate": 3.8051150420703337e-06,
+      "loss": 0.5376,
+      "step": 7415
+    },
+    {
+      "epoch": 1.968803929377406,
+      "grad_norm": 0.4225016672918098,
+      "learning_rate": 3.8048172520967103e-06,
+      "loss": 0.5778,
+      "step": 7416
+    },
+    {
+      "epoch": 1.9690694278507899,
+      "grad_norm": 0.41054173141752887,
+      "learning_rate": 3.804519436675956e-06,
+      "loss": 0.5925,
+      "step": 7417
+    },
+    {
+      "epoch": 1.9693349263241737,
+      "grad_norm": 0.41224055467123305,
+      "learning_rate": 3.8042215958138784e-06,
+      "loss": 0.6037,
+      "step": 7418
+    },
+    {
+      "epoch": 1.9696004247975574,
+      "grad_norm": 0.43703809178912356,
+      "learning_rate": 3.8039237295162858e-06,
+      "loss": 0.6011,
+      "step": 7419
+    },
+    {
+      "epoch": 1.9698659232709412,
+      "grad_norm": 0.43038227767228815,
+      "learning_rate": 3.8036258377889882e-06,
+      "loss": 0.6118,
+      "step": 7420
+    },
+    {
+      "epoch": 1.970131421744325,
+      "grad_norm": 0.4288096208914957,
+      "learning_rate": 3.8033279206377956e-06,
+      "loss": 0.5887,
+      "step": 7421
+    },
+    {
+      "epoch": 1.9703969202177087,
+      "grad_norm": 0.40989466331715,
+      "learning_rate": 3.8030299780685164e-06,
+      "loss": 0.5657,
+      "step": 7422
+    },
+    {
+      "epoch": 1.9706624186910924,
+      "grad_norm": 0.4177392528650334,
+      "learning_rate": 3.8027320100869635e-06,
+      "loss": 0.6003,
+      "step": 7423
+    },
+    {
+      "epoch": 1.9709279171644765,
+      "grad_norm": 0.431206722063452,
+      "learning_rate": 3.802434016698946e-06,
+      "loss": 0.5818,
+      "step": 7424
+    },
+    {
+      "epoch": 1.97119341563786,
+      "grad_norm": 0.4086312185451687,
+      "learning_rate": 3.8021359979102767e-06,
+      "loss": 0.5778,
+      "step": 7425
+    },
+    {
+      "epoch": 1.9714589141112437,
+      "grad_norm": 0.4211439415890832,
+      "learning_rate": 3.801837953726767e-06,
+      "loss": 0.562,
+      "step": 7426
+    },
+    {
+      "epoch": 1.9717244125846276,
+      "grad_norm": 0.4336611730411598,
+      "learning_rate": 3.8015398841542297e-06,
+      "loss": 0.5601,
+      "step": 7427
+    },
+    {
+      "epoch": 1.9719899110580115,
+      "grad_norm": 0.4125105807222685,
+      "learning_rate": 3.8012417891984776e-06,
+      "loss": 0.5843,
+      "step": 7428
+    },
+    {
+      "epoch": 1.972255409531395,
+      "grad_norm": 0.4246092297005021,
+      "learning_rate": 3.800943668865326e-06,
+      "loss": 0.571,
+      "step": 7429
+    },
+    {
+      "epoch": 1.972520908004779,
+      "grad_norm": 0.43149052929987475,
+      "learning_rate": 3.800645523160586e-06,
+      "loss": 0.5911,
+      "step": 7430
+    },
+    {
+      "epoch": 1.9727864064781628,
+      "grad_norm": 0.41525898191498506,
+      "learning_rate": 3.800347352090074e-06,
+      "loss": 0.5449,
+      "step": 7431
+    },
+    {
+      "epoch": 1.9730519049515465,
+      "grad_norm": 0.42474713314117263,
+      "learning_rate": 3.8000491556596055e-06,
+      "loss": 0.5774,
+      "step": 7432
+    },
+    {
+      "epoch": 1.9733174034249303,
+      "grad_norm": 0.4226163744300414,
+      "learning_rate": 3.799750933874995e-06,
+      "loss": 0.5947,
+      "step": 7433
+    },
+    {
+      "epoch": 1.9735829018983142,
+      "grad_norm": 0.4190942395315396,
+      "learning_rate": 3.79945268674206e-06,
+      "loss": 0.5871,
+      "step": 7434
+    },
+    {
+      "epoch": 1.9738484003716978,
+      "grad_norm": 0.4447372453983853,
+      "learning_rate": 3.7991544142666146e-06,
+      "loss": 0.5689,
+      "step": 7435
+    },
+    {
+      "epoch": 1.9741138988450815,
+      "grad_norm": 0.41663101946232156,
+      "learning_rate": 3.798856116454478e-06,
+      "loss": 0.613,
+      "step": 7436
+    },
+    {
+      "epoch": 1.9743793973184656,
+      "grad_norm": 0.42124953480960353,
+      "learning_rate": 3.798557793311466e-06,
+      "loss": 0.577,
+      "step": 7437
+    },
+    {
+      "epoch": 1.9746448957918492,
+      "grad_norm": 0.4125730878123955,
+      "learning_rate": 3.7982594448433983e-06,
+      "loss": 0.5628,
+      "step": 7438
+    },
+    {
+      "epoch": 1.9749103942652328,
+      "grad_norm": 0.43410623056739295,
+      "learning_rate": 3.797961071056093e-06,
+      "loss": 0.6187,
+      "step": 7439
+    },
+    {
+      "epoch": 1.9751758927386167,
+      "grad_norm": 0.461363510074443,
+      "learning_rate": 3.7976626719553688e-06,
+      "loss": 0.5689,
+      "step": 7440
+    },
+    {
+      "epoch": 1.9754413912120006,
+      "grad_norm": 0.4235374205365759,
+      "learning_rate": 3.797364247547045e-06,
+      "loss": 0.5703,
+      "step": 7441
+    },
+    {
+      "epoch": 1.9757068896853842,
+      "grad_norm": 0.4126239334919761,
+      "learning_rate": 3.7970657978369418e-06,
+      "loss": 0.5861,
+      "step": 7442
+    },
+    {
+      "epoch": 1.975972388158768,
+      "grad_norm": 0.4186589837459904,
+      "learning_rate": 3.79676732283088e-06,
+      "loss": 0.5426,
+      "step": 7443
+    },
+    {
+      "epoch": 1.976237886632152,
+      "grad_norm": 0.4823851203797616,
+      "learning_rate": 3.79646882253468e-06,
+      "loss": 0.5653,
+      "step": 7444
+    },
+    {
+      "epoch": 1.9765033851055356,
+      "grad_norm": 0.41591811961301756,
+      "learning_rate": 3.796170296954164e-06,
+      "loss": 0.5903,
+      "step": 7445
+    },
+    {
+      "epoch": 1.9767688835789194,
+      "grad_norm": 0.4081284254401549,
+      "learning_rate": 3.7958717460951546e-06,
+      "loss": 0.549,
+      "step": 7446
+    },
+    {
+      "epoch": 1.9770343820523033,
+      "grad_norm": 0.41894988927461274,
+      "learning_rate": 3.7955731699634717e-06,
+      "loss": 0.6042,
+      "step": 7447
+    },
+    {
+      "epoch": 1.977299880525687,
+      "grad_norm": 0.4370634235233121,
+      "learning_rate": 3.79527456856494e-06,
+      "loss": 0.5551,
+      "step": 7448
+    },
+    {
+      "epoch": 1.9775653789990706,
+      "grad_norm": 0.4120978249653195,
+      "learning_rate": 3.7949759419053833e-06,
+      "loss": 0.5542,
+      "step": 7449
+    },
+    {
+      "epoch": 1.9778308774724547,
+      "grad_norm": 0.41561231759868,
+      "learning_rate": 3.7946772899906253e-06,
+      "loss": 0.6164,
+      "step": 7450
+    },
+    {
+      "epoch": 1.9780963759458383,
+      "grad_norm": 0.42482719222816673,
+      "learning_rate": 3.7943786128264902e-06,
+      "loss": 0.5818,
+      "step": 7451
+    },
+    {
+      "epoch": 1.978361874419222,
+      "grad_norm": 0.4144044880639814,
+      "learning_rate": 3.794079910418803e-06,
+      "loss": 0.6004,
+      "step": 7452
+    },
+    {
+      "epoch": 1.9786273728926058,
+      "grad_norm": 0.4226922735449433,
+      "learning_rate": 3.7937811827733882e-06,
+      "loss": 0.5761,
+      "step": 7453
+    },
+    {
+      "epoch": 1.9788928713659897,
+      "grad_norm": 0.4112301463198234,
+      "learning_rate": 3.793482429896074e-06,
+      "loss": 0.5876,
+      "step": 7454
+    },
+    {
+      "epoch": 1.9791583698393733,
+      "grad_norm": 0.4325794906865588,
+      "learning_rate": 3.7931836517926844e-06,
+      "loss": 0.6001,
+      "step": 7455
+    },
+    {
+      "epoch": 1.9794238683127572,
+      "grad_norm": 0.41626570417621117,
+      "learning_rate": 3.7928848484690477e-06,
+      "loss": 0.5428,
+      "step": 7456
+    },
+    {
+      "epoch": 1.979689366786141,
+      "grad_norm": 0.3996050822708342,
+      "learning_rate": 3.792586019930991e-06,
+      "loss": 0.5544,
+      "step": 7457
+    },
+    {
+      "epoch": 1.9799548652595247,
+      "grad_norm": 0.4215100294319676,
+      "learning_rate": 3.792287166184342e-06,
+      "loss": 0.5889,
+      "step": 7458
+    },
+    {
+      "epoch": 1.9802203637329086,
+      "grad_norm": 0.4110296134987471,
+      "learning_rate": 3.791988287234929e-06,
+      "loss": 0.6003,
+      "step": 7459
+    },
+    {
+      "epoch": 1.9804858622062924,
+      "grad_norm": 0.4076559851840653,
+      "learning_rate": 3.7916893830885816e-06,
+      "loss": 0.5866,
+      "step": 7460
+    },
+    {
+      "epoch": 1.980751360679676,
+      "grad_norm": 0.41650298212304376,
+      "learning_rate": 3.7913904537511283e-06,
+      "loss": 0.5948,
+      "step": 7461
+    },
+    {
+      "epoch": 1.98101685915306,
+      "grad_norm": 0.41599462841395546,
+      "learning_rate": 3.791091499228399e-06,
+      "loss": 0.5718,
+      "step": 7462
+    },
+    {
+      "epoch": 1.9812823576264438,
+      "grad_norm": 0.41144312088330043,
+      "learning_rate": 3.790792519526225e-06,
+      "loss": 0.5425,
+      "step": 7463
+    },
+    {
+      "epoch": 1.9815478560998274,
+      "grad_norm": 0.39046794691059317,
+      "learning_rate": 3.790493514650436e-06,
+      "loss": 0.5973,
+      "step": 7464
+    },
+    {
+      "epoch": 1.981813354573211,
+      "grad_norm": 0.4153129533702801,
+      "learning_rate": 3.790194484606864e-06,
+      "loss": 0.5623,
+      "step": 7465
+    },
+    {
+      "epoch": 1.9820788530465951,
+      "grad_norm": 0.39437436556230643,
+      "learning_rate": 3.78989542940134e-06,
+      "loss": 0.534,
+      "step": 7466
+    },
+    {
+      "epoch": 1.9823443515199788,
+      "grad_norm": 0.4131985715877372,
+      "learning_rate": 3.789596349039697e-06,
+      "loss": 0.5939,
+      "step": 7467
+    },
+    {
+      "epoch": 1.9826098499933624,
+      "grad_norm": 0.40769899343581045,
+      "learning_rate": 3.7892972435277686e-06,
+      "loss": 0.5719,
+      "step": 7468
+    },
+    {
+      "epoch": 1.9828753484667463,
+      "grad_norm": 0.4191367188228909,
+      "learning_rate": 3.7889981128713873e-06,
+      "loss": 0.5787,
+      "step": 7469
+    },
+    {
+      "epoch": 1.9831408469401302,
+      "grad_norm": 0.4086534301464835,
+      "learning_rate": 3.7886989570763856e-06,
+      "loss": 0.5825,
+      "step": 7470
+    },
+    {
+      "epoch": 1.9834063454135138,
+      "grad_norm": 0.4079483021682168,
+      "learning_rate": 3.7883997761485993e-06,
+      "loss": 0.5523,
+      "step": 7471
+    },
+    {
+      "epoch": 1.9836718438868977,
+      "grad_norm": 0.4221876143467679,
+      "learning_rate": 3.7881005700938635e-06,
+      "loss": 0.5938,
+      "step": 7472
+    },
+    {
+      "epoch": 1.9839373423602815,
+      "grad_norm": 0.4106825970474645,
+      "learning_rate": 3.7878013389180126e-06,
+      "loss": 0.5667,
+      "step": 7473
+    },
+    {
+      "epoch": 1.9842028408336652,
+      "grad_norm": 0.4304947278749791,
+      "learning_rate": 3.7875020826268818e-06,
+      "loss": 0.5774,
+      "step": 7474
+    },
+    {
+      "epoch": 1.984468339307049,
+      "grad_norm": 0.41587302972158063,
+      "learning_rate": 3.7872028012263084e-06,
+      "loss": 0.5689,
+      "step": 7475
+    },
+    {
+      "epoch": 1.9847338377804329,
+      "grad_norm": 0.43202715931612407,
+      "learning_rate": 3.7869034947221285e-06,
+      "loss": 0.5928,
+      "step": 7476
+    },
+    {
+      "epoch": 1.9849993362538165,
+      "grad_norm": 0.4194965374504465,
+      "learning_rate": 3.78660416312018e-06,
+      "loss": 0.55,
+      "step": 7477
+    },
+    {
+      "epoch": 1.9852648347272002,
+      "grad_norm": 0.4310260489747963,
+      "learning_rate": 3.786304806426301e-06,
+      "loss": 0.6205,
+      "step": 7478
+    },
+    {
+      "epoch": 1.9855303332005843,
+      "grad_norm": 0.4281765102576451,
+      "learning_rate": 3.7860054246463272e-06,
+      "loss": 0.5639,
+      "step": 7479
+    },
+    {
+      "epoch": 1.985795831673968,
+      "grad_norm": 0.40540803995042807,
+      "learning_rate": 3.7857060177860992e-06,
+      "loss": 0.5753,
+      "step": 7480
+    },
+    {
+      "epoch": 1.9860613301473515,
+      "grad_norm": 0.40309097399726107,
+      "learning_rate": 3.7854065858514573e-06,
+      "loss": 0.6094,
+      "step": 7481
+    },
+    {
+      "epoch": 1.9863268286207354,
+      "grad_norm": 0.40909241825909437,
+      "learning_rate": 3.7851071288482387e-06,
+      "loss": 0.5524,
+      "step": 7482
+    },
+    {
+      "epoch": 1.9865923270941193,
+      "grad_norm": 0.4100476787226209,
+      "learning_rate": 3.7848076467822846e-06,
+      "loss": 0.5629,
+      "step": 7483
+    },
+    {
+      "epoch": 1.986857825567503,
+      "grad_norm": 0.4431898317902111,
+      "learning_rate": 3.7845081396594354e-06,
+      "loss": 0.5379,
+      "step": 7484
+    },
+    {
+      "epoch": 1.9871233240408868,
+      "grad_norm": 0.4052830763386127,
+      "learning_rate": 3.7842086074855334e-06,
+      "loss": 0.5875,
+      "step": 7485
+    },
+    {
+      "epoch": 1.9873888225142706,
+      "grad_norm": 0.4146115790317509,
+      "learning_rate": 3.783909050266419e-06,
+      "loss": 0.5928,
+      "step": 7486
+    },
+    {
+      "epoch": 1.9876543209876543,
+      "grad_norm": 0.41350662762554063,
+      "learning_rate": 3.783609468007934e-06,
+      "loss": 0.55,
+      "step": 7487
+    },
+    {
+      "epoch": 1.9879198194610381,
+      "grad_norm": 0.415554981790555,
+      "learning_rate": 3.783309860715921e-06,
+      "loss": 0.5588,
+      "step": 7488
+    },
+    {
+      "epoch": 1.988185317934422,
+      "grad_norm": 0.43987157320306064,
+      "learning_rate": 3.7830102283962254e-06,
+      "loss": 0.5456,
+      "step": 7489
+    },
+    {
+      "epoch": 1.9884508164078056,
+      "grad_norm": 0.41114900283976163,
+      "learning_rate": 3.782710571054688e-06,
+      "loss": 0.6099,
+      "step": 7490
+    },
+    {
+      "epoch": 1.9887163148811893,
+      "grad_norm": 0.40792923356523525,
+      "learning_rate": 3.7824108886971535e-06,
+      "loss": 0.5866,
+      "step": 7491
+    },
+    {
+      "epoch": 1.9889818133545734,
+      "grad_norm": 0.417019160180324,
+      "learning_rate": 3.7821111813294674e-06,
+      "loss": 0.57,
+      "step": 7492
+    },
+    {
+      "epoch": 1.989247311827957,
+      "grad_norm": 0.4273576445563474,
+      "learning_rate": 3.7818114489574737e-06,
+      "loss": 0.6266,
+      "step": 7493
+    },
+    {
+      "epoch": 1.9895128103013406,
+      "grad_norm": 0.43704009096819274,
+      "learning_rate": 3.7815116915870187e-06,
+      "loss": 0.5903,
+      "step": 7494
+    },
+    {
+      "epoch": 1.9897783087747245,
+      "grad_norm": 0.4329871314136986,
+      "learning_rate": 3.781211909223948e-06,
+      "loss": 0.6059,
+      "step": 7495
+    },
+    {
+      "epoch": 1.9900438072481084,
+      "grad_norm": 0.4230724535294457,
+      "learning_rate": 3.7809121018741078e-06,
+      "loss": 0.5653,
+      "step": 7496
+    },
+    {
+      "epoch": 1.990309305721492,
+      "grad_norm": 0.41918402761396256,
+      "learning_rate": 3.7806122695433456e-06,
+      "loss": 0.5826,
+      "step": 7497
+    },
+    {
+      "epoch": 1.9905748041948759,
+      "grad_norm": 0.41636714488334237,
+      "learning_rate": 3.780312412237509e-06,
+      "loss": 0.5638,
+      "step": 7498
+    },
+    {
+      "epoch": 1.9908403026682597,
+      "grad_norm": 0.4156181713627736,
+      "learning_rate": 3.7800125299624444e-06,
+      "loss": 0.583,
+      "step": 7499
+    },
+    {
+      "epoch": 1.9911058011416434,
+      "grad_norm": 0.40901610155378626,
+      "learning_rate": 3.779712622724003e-06,
+      "loss": 0.5923,
+      "step": 7500
+    },
+    {
+      "epoch": 1.9913712996150272,
+      "grad_norm": 0.4420018060181688,
+      "learning_rate": 3.7794126905280317e-06,
+      "loss": 0.5948,
+      "step": 7501
+    },
+    {
+      "epoch": 1.991636798088411,
+      "grad_norm": 0.41691576555377097,
+      "learning_rate": 3.77911273338038e-06,
+      "loss": 0.6296,
+      "step": 7502
+    },
+    {
+      "epoch": 1.9919022965617947,
+      "grad_norm": 0.4292098675997834,
+      "learning_rate": 3.7788127512868984e-06,
+      "loss": 0.5879,
+      "step": 7503
+    },
+    {
+      "epoch": 1.9921677950351786,
+      "grad_norm": 0.4052871928629502,
+      "learning_rate": 3.778512744253437e-06,
+      "loss": 0.5857,
+      "step": 7504
+    },
+    {
+      "epoch": 1.9924332935085625,
+      "grad_norm": 0.3994916515356325,
+      "learning_rate": 3.778212712285847e-06,
+      "loss": 0.5584,
+      "step": 7505
+    },
+    {
+      "epoch": 1.992698791981946,
+      "grad_norm": 0.41964470575903884,
+      "learning_rate": 3.777912655389979e-06,
+      "loss": 0.5491,
+      "step": 7506
+    },
+    {
+      "epoch": 1.9929642904553297,
+      "grad_norm": 0.40849686894282194,
+      "learning_rate": 3.777612573571686e-06,
+      "loss": 0.5834,
+      "step": 7507
+    },
+    {
+      "epoch": 1.9932297889287136,
+      "grad_norm": 0.4076866839137105,
+      "learning_rate": 3.777312466836819e-06,
+      "loss": 0.6027,
+      "step": 7508
+    },
+    {
+      "epoch": 1.9934952874020975,
+      "grad_norm": 0.4012289472764235,
+      "learning_rate": 3.7770123351912325e-06,
+      "loss": 0.5425,
+      "step": 7509
+    },
+    {
+      "epoch": 1.9937607858754811,
+      "grad_norm": 0.42824139054135735,
+      "learning_rate": 3.776712178640778e-06,
+      "loss": 0.5956,
+      "step": 7510
+    },
+    {
+      "epoch": 1.994026284348865,
+      "grad_norm": 0.4058150445018642,
+      "learning_rate": 3.7764119971913098e-06,
+      "loss": 0.5885,
+      "step": 7511
+    },
+    {
+      "epoch": 1.9942917828222488,
+      "grad_norm": 0.42464102247810515,
+      "learning_rate": 3.7761117908486833e-06,
+      "loss": 0.5425,
+      "step": 7512
+    },
+    {
+      "epoch": 1.9945572812956325,
+      "grad_norm": 0.4153732021015508,
+      "learning_rate": 3.7758115596187517e-06,
+      "loss": 0.5942,
+      "step": 7513
+    },
+    {
+      "epoch": 1.9948227797690163,
+      "grad_norm": 0.42507348054777094,
+      "learning_rate": 3.7755113035073707e-06,
+      "loss": 0.5615,
+      "step": 7514
+    },
+    {
+      "epoch": 1.9950882782424002,
+      "grad_norm": 0.41011630486263034,
+      "learning_rate": 3.7752110225203974e-06,
+      "loss": 0.575,
+      "step": 7515
+    },
+    {
+      "epoch": 1.9953537767157838,
+      "grad_norm": 0.452965704832396,
+      "learning_rate": 3.7749107166636867e-06,
+      "loss": 0.5906,
+      "step": 7516
+    },
+    {
+      "epoch": 1.9956192751891677,
+      "grad_norm": 0.41711257779944655,
+      "learning_rate": 3.774610385943095e-06,
+      "loss": 0.5813,
+      "step": 7517
+    },
+    {
+      "epoch": 1.9958847736625516,
+      "grad_norm": 0.4103512349127069,
+      "learning_rate": 3.77431003036448e-06,
+      "loss": 0.5871,
+      "step": 7518
+    },
+    {
+      "epoch": 1.9961502721359352,
+      "grad_norm": 0.40903660395809843,
+      "learning_rate": 3.7740096499336996e-06,
+      "loss": 0.5533,
+      "step": 7519
+    },
+    {
+      "epoch": 1.9964157706093189,
+      "grad_norm": 0.39756787945628164,
+      "learning_rate": 3.7737092446566115e-06,
+      "loss": 0.5727,
+      "step": 7520
+    },
+    {
+      "epoch": 1.996681269082703,
+      "grad_norm": 0.4148081666930947,
+      "learning_rate": 3.773408814539075e-06,
+      "loss": 0.5934,
+      "step": 7521
+    },
+    {
+      "epoch": 1.9969467675560866,
+      "grad_norm": 0.4042675946497851,
+      "learning_rate": 3.773108359586948e-06,
+      "loss": 0.595,
+      "step": 7522
+    },
+    {
+      "epoch": 1.9972122660294702,
+      "grad_norm": 0.42682277902814,
+      "learning_rate": 3.772807879806091e-06,
+      "loss": 0.5483,
+      "step": 7523
+    },
+    {
+      "epoch": 1.997477764502854,
+      "grad_norm": 0.394793331758505,
+      "learning_rate": 3.772507375202365e-06,
+      "loss": 0.5506,
+      "step": 7524
+    },
+    {
+      "epoch": 1.997743262976238,
+      "grad_norm": 0.4138675275200553,
+      "learning_rate": 3.7722068457816287e-06,
+      "loss": 0.6088,
+      "step": 7525
+    },
+    {
+      "epoch": 1.9980087614496216,
+      "grad_norm": 0.4196800846173291,
+      "learning_rate": 3.771906291549744e-06,
+      "loss": 0.5843,
+      "step": 7526
+    },
+    {
+      "epoch": 1.9982742599230054,
+      "grad_norm": 0.3985408842440907,
+      "learning_rate": 3.7716057125125727e-06,
+      "loss": 0.5751,
+      "step": 7527
+    },
+    {
+      "epoch": 1.9985397583963893,
+      "grad_norm": 0.4011213842007995,
+      "learning_rate": 3.7713051086759764e-06,
+      "loss": 0.6004,
+      "step": 7528
+    },
+    {
+      "epoch": 1.998805256869773,
+      "grad_norm": 0.41196775153150056,
+      "learning_rate": 3.771004480045818e-06,
+      "loss": 0.5886,
+      "step": 7529
+    },
+    {
+      "epoch": 1.9990707553431568,
+      "grad_norm": 0.40961083270779247,
+      "learning_rate": 3.7707038266279604e-06,
+      "loss": 0.5957,
+      "step": 7530
+    },
+    {
+      "epoch": 1.9993362538165407,
+      "grad_norm": 0.41781163583119213,
+      "learning_rate": 3.7704031484282666e-06,
+      "loss": 0.5324,
+      "step": 7531
+    },
+    {
+      "epoch": 1.9996017522899243,
+      "grad_norm": 0.4136015002730909,
+      "learning_rate": 3.7701024454526015e-06,
+      "loss": 0.578,
+      "step": 7532
+    },
+    {
+      "epoch": 1.999867250763308,
+      "grad_norm": 0.4167394435791282,
+      "learning_rate": 3.769801717706828e-06,
+      "loss": 0.5867,
+      "step": 7533
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6582931606981949,
+      "learning_rate": 3.769500965196813e-06,
+      "loss": 0.5582,
+      "step": 7534
+    },
+    {
+      "epoch": 2.0002654984733836,
+      "grad_norm": 0.39849785806989524,
+      "learning_rate": 3.7692001879284203e-06,
+      "loss": 0.5269,
+      "step": 7535
+    },
+    {
+      "epoch": 2.0005309969467677,
+      "grad_norm": 0.4071731370376904,
+      "learning_rate": 3.7688993859075164e-06,
+      "loss": 0.5787,
+      "step": 7536
+    },
+    {
+      "epoch": 2.0007964954201514,
+      "grad_norm": 0.4060741465652973,
+      "learning_rate": 3.7685985591399677e-06,
+      "loss": 0.5649,
+      "step": 7537
+    },
+    {
+      "epoch": 2.001061993893535,
+      "grad_norm": 0.41035203247693974,
+      "learning_rate": 3.768297707631642e-06,
+      "loss": 0.5462,
+      "step": 7538
+    },
+    {
+      "epoch": 2.001327492366919,
+      "grad_norm": 0.40969406178316803,
+      "learning_rate": 3.7679968313884042e-06,
+      "loss": 0.5734,
+      "step": 7539
+    },
+    {
+      "epoch": 2.0015929908403027,
+      "grad_norm": 0.40768245156178695,
+      "learning_rate": 3.7676959304161238e-06,
+      "loss": 0.5512,
+      "step": 7540
+    },
+    {
+      "epoch": 2.0018584893136864,
+      "grad_norm": 0.4021573870578567,
+      "learning_rate": 3.76739500472067e-06,
+      "loss": 0.5539,
+      "step": 7541
+    },
+    {
+      "epoch": 2.00212398778707,
+      "grad_norm": 0.4124405753842156,
+      "learning_rate": 3.7670940543079093e-06,
+      "loss": 0.5433,
+      "step": 7542
+    },
+    {
+      "epoch": 2.002389486260454,
+      "grad_norm": 0.4147953159519462,
+      "learning_rate": 3.7667930791837126e-06,
+      "loss": 0.5792,
+      "step": 7543
+    },
+    {
+      "epoch": 2.0026549847338377,
+      "grad_norm": 0.4091672369166906,
+      "learning_rate": 3.7664920793539493e-06,
+      "loss": 0.5764,
+      "step": 7544
+    },
+    {
+      "epoch": 2.0029204832072214,
+      "grad_norm": 0.4199860337356155,
+      "learning_rate": 3.7661910548244896e-06,
+      "loss": 0.5781,
+      "step": 7545
+    },
+    {
+      "epoch": 2.0031859816806055,
+      "grad_norm": 0.41147385773224143,
+      "learning_rate": 3.765890005601204e-06,
+      "loss": 0.5667,
+      "step": 7546
+    },
+    {
+      "epoch": 2.003451480153989,
+      "grad_norm": 0.41587528066035606,
+      "learning_rate": 3.765588931689964e-06,
+      "loss": 0.5434,
+      "step": 7547
+    },
+    {
+      "epoch": 2.0037169786273727,
+      "grad_norm": 0.42208180926256395,
+      "learning_rate": 3.7652878330966414e-06,
+      "loss": 0.5435,
+      "step": 7548
+    },
+    {
+      "epoch": 2.003982477100757,
+      "grad_norm": 0.4387065946199643,
+      "learning_rate": 3.764986709827107e-06,
+      "loss": 0.607,
+      "step": 7549
+    },
+    {
+      "epoch": 2.0042479755741405,
+      "grad_norm": 0.4301699883843542,
+      "learning_rate": 3.764685561887236e-06,
+      "loss": 0.5479,
+      "step": 7550
+    },
+    {
+      "epoch": 2.004513474047524,
+      "grad_norm": 0.4192885969648279,
+      "learning_rate": 3.764384389282899e-06,
+      "loss": 0.6061,
+      "step": 7551
+    },
+    {
+      "epoch": 2.004778972520908,
+      "grad_norm": 0.42042741745900486,
+      "learning_rate": 3.7640831920199706e-06,
+      "loss": 0.5549,
+      "step": 7552
+    },
+    {
+      "epoch": 2.005044470994292,
+      "grad_norm": 0.4189877292495763,
+      "learning_rate": 3.7637819701043256e-06,
+      "loss": 0.5336,
+      "step": 7553
+    },
+    {
+      "epoch": 2.0053099694676755,
+      "grad_norm": 0.4283044764271871,
+      "learning_rate": 3.763480723541838e-06,
+      "loss": 0.5944,
+      "step": 7554
+    },
+    {
+      "epoch": 2.005575467941059,
+      "grad_norm": 0.41820593600901707,
+      "learning_rate": 3.7631794523383823e-06,
+      "loss": 0.5788,
+      "step": 7555
+    },
+    {
+      "epoch": 2.005840966414443,
+      "grad_norm": 0.4093590565435108,
+      "learning_rate": 3.762878156499835e-06,
+      "loss": 0.5061,
+      "step": 7556
+    },
+    {
+      "epoch": 2.006106464887827,
+      "grad_norm": 0.4257495887949887,
+      "learning_rate": 3.762576836032071e-06,
+      "loss": 0.5894,
+      "step": 7557
+    },
+    {
+      "epoch": 2.0063719633612105,
+      "grad_norm": 0.42369780156780357,
+      "learning_rate": 3.7622754909409674e-06,
+      "loss": 0.6089,
+      "step": 7558
+    },
+    {
+      "epoch": 2.0066374618345946,
+      "grad_norm": 0.4171308834477733,
+      "learning_rate": 3.761974121232401e-06,
+      "loss": 0.5569,
+      "step": 7559
+    },
+    {
+      "epoch": 2.006902960307978,
+      "grad_norm": 0.4261698926162277,
+      "learning_rate": 3.76167272691225e-06,
+      "loss": 0.5541,
+      "step": 7560
+    },
+    {
+      "epoch": 2.007168458781362,
+      "grad_norm": 0.42610925731238486,
+      "learning_rate": 3.7613713079863914e-06,
+      "loss": 0.5565,
+      "step": 7561
+    },
+    {
+      "epoch": 2.007433957254746,
+      "grad_norm": 0.4239100599712335,
+      "learning_rate": 3.7610698644607036e-06,
+      "loss": 0.6016,
+      "step": 7562
+    },
+    {
+      "epoch": 2.0076994557281296,
+      "grad_norm": 0.42451315438866494,
+      "learning_rate": 3.7607683963410665e-06,
+      "loss": 0.5724,
+      "step": 7563
+    },
+    {
+      "epoch": 2.007964954201513,
+      "grad_norm": 0.4550468384610608,
+      "learning_rate": 3.7604669036333584e-06,
+      "loss": 0.5782,
+      "step": 7564
+    },
+    {
+      "epoch": 2.0082304526748973,
+      "grad_norm": 0.42960257659456064,
+      "learning_rate": 3.7601653863434596e-06,
+      "loss": 0.5576,
+      "step": 7565
+    },
+    {
+      "epoch": 2.008495951148281,
+      "grad_norm": 0.4165642861767433,
+      "learning_rate": 3.7598638444772505e-06,
+      "loss": 0.575,
+      "step": 7566
+    },
+    {
+      "epoch": 2.0087614496216646,
+      "grad_norm": 0.41955441713771147,
+      "learning_rate": 3.7595622780406116e-06,
+      "loss": 0.5729,
+      "step": 7567
+    },
+    {
+      "epoch": 2.0090269480950487,
+      "grad_norm": 0.4142957685437747,
+      "learning_rate": 3.759260687039424e-06,
+      "loss": 0.5511,
+      "step": 7568
+    },
+    {
+      "epoch": 2.0092924465684323,
+      "grad_norm": 0.42324566322808216,
+      "learning_rate": 3.75895907147957e-06,
+      "loss": 0.569,
+      "step": 7569
+    },
+    {
+      "epoch": 2.009557945041816,
+      "grad_norm": 0.42120555280639016,
+      "learning_rate": 3.758657431366932e-06,
+      "loss": 0.5956,
+      "step": 7570
+    },
+    {
+      "epoch": 2.0098234435151996,
+      "grad_norm": 0.41309407656392527,
+      "learning_rate": 3.758355766707392e-06,
+      "loss": 0.5809,
+      "step": 7571
+    },
+    {
+      "epoch": 2.0100889419885837,
+      "grad_norm": 0.41890565553502174,
+      "learning_rate": 3.7580540775068334e-06,
+      "loss": 0.5573,
+      "step": 7572
+    },
+    {
+      "epoch": 2.0103544404619673,
+      "grad_norm": 0.42468503747452674,
+      "learning_rate": 3.757752363771141e-06,
+      "loss": 0.5866,
+      "step": 7573
+    },
+    {
+      "epoch": 2.010619938935351,
+      "grad_norm": 0.40785280191753054,
+      "learning_rate": 3.757450625506197e-06,
+      "loss": 0.5332,
+      "step": 7574
+    },
+    {
+      "epoch": 2.010885437408735,
+      "grad_norm": 0.42729196912372314,
+      "learning_rate": 3.7571488627178875e-06,
+      "loss": 0.5881,
+      "step": 7575
+    },
+    {
+      "epoch": 2.0111509358821187,
+      "grad_norm": 0.4189552892007665,
+      "learning_rate": 3.7568470754120967e-06,
+      "loss": 0.5898,
+      "step": 7576
+    },
+    {
+      "epoch": 2.0114164343555023,
+      "grad_norm": 0.4228253571150553,
+      "learning_rate": 3.7565452635947106e-06,
+      "loss": 0.6099,
+      "step": 7577
+    },
+    {
+      "epoch": 2.0116819328288864,
+      "grad_norm": 0.41617814443775253,
+      "learning_rate": 3.7562434272716157e-06,
+      "loss": 0.5865,
+      "step": 7578
+    },
+    {
+      "epoch": 2.01194743130227,
+      "grad_norm": 0.420106666536967,
+      "learning_rate": 3.7559415664486985e-06,
+      "loss": 0.5606,
+      "step": 7579
+    },
+    {
+      "epoch": 2.0122129297756537,
+      "grad_norm": 0.4164899990776468,
+      "learning_rate": 3.7556396811318453e-06,
+      "loss": 0.5773,
+      "step": 7580
+    },
+    {
+      "epoch": 2.0124784282490378,
+      "grad_norm": 0.41961464901743245,
+      "learning_rate": 3.7553377713269447e-06,
+      "loss": 0.5457,
+      "step": 7581
+    },
+    {
+      "epoch": 2.0127439267224214,
+      "grad_norm": 0.4139187721865026,
+      "learning_rate": 3.7550358370398825e-06,
+      "loss": 0.5719,
+      "step": 7582
+    },
+    {
+      "epoch": 2.013009425195805,
+      "grad_norm": 0.42218109155313555,
+      "learning_rate": 3.7547338782765496e-06,
+      "loss": 0.552,
+      "step": 7583
+    },
+    {
+      "epoch": 2.0132749236691887,
+      "grad_norm": 0.41731766090243194,
+      "learning_rate": 3.754431895042834e-06,
+      "loss": 0.5403,
+      "step": 7584
+    },
+    {
+      "epoch": 2.013540422142573,
+      "grad_norm": 0.41572247097844384,
+      "learning_rate": 3.754129887344624e-06,
+      "loss": 0.5516,
+      "step": 7585
+    },
+    {
+      "epoch": 2.0138059206159564,
+      "grad_norm": 0.41925982552092605,
+      "learning_rate": 3.753827855187811e-06,
+      "loss": 0.5741,
+      "step": 7586
+    },
+    {
+      "epoch": 2.01407141908934,
+      "grad_norm": 0.4108585298729098,
+      "learning_rate": 3.753525798578286e-06,
+      "loss": 0.5357,
+      "step": 7587
+    },
+    {
+      "epoch": 2.014336917562724,
+      "grad_norm": 0.41471959591468255,
+      "learning_rate": 3.7532237175219378e-06,
+      "loss": 0.54,
+      "step": 7588
+    },
+    {
+      "epoch": 2.014602416036108,
+      "grad_norm": 0.4281300748310313,
+      "learning_rate": 3.7529216120246585e-06,
+      "loss": 0.5714,
+      "step": 7589
+    },
+    {
+      "epoch": 2.0148679145094914,
+      "grad_norm": 0.4236491915516071,
+      "learning_rate": 3.752619482092341e-06,
+      "loss": 0.5955,
+      "step": 7590
+    },
+    {
+      "epoch": 2.0151334129828755,
+      "grad_norm": 0.41502115587310157,
+      "learning_rate": 3.7523173277308763e-06,
+      "loss": 0.5411,
+      "step": 7591
+    },
+    {
+      "epoch": 2.015398911456259,
+      "grad_norm": 0.4137020625036817,
+      "learning_rate": 3.752015148946157e-06,
+      "loss": 0.5631,
+      "step": 7592
+    },
+    {
+      "epoch": 2.015664409929643,
+      "grad_norm": 0.4325603342359071,
+      "learning_rate": 3.7517129457440772e-06,
+      "loss": 0.5581,
+      "step": 7593
+    },
+    {
+      "epoch": 2.015929908403027,
+      "grad_norm": 0.4201421730784346,
+      "learning_rate": 3.7514107181305302e-06,
+      "loss": 0.5733,
+      "step": 7594
+    },
+    {
+      "epoch": 2.0161954068764105,
+      "grad_norm": 0.40944023257344,
+      "learning_rate": 3.7511084661114094e-06,
+      "loss": 0.5389,
+      "step": 7595
+    },
+    {
+      "epoch": 2.016460905349794,
+      "grad_norm": 0.4126118131152411,
+      "learning_rate": 3.7508061896926117e-06,
+      "loss": 0.5064,
+      "step": 7596
+    },
+    {
+      "epoch": 2.016726403823178,
+      "grad_norm": 0.42269380346032576,
+      "learning_rate": 3.7505038888800304e-06,
+      "loss": 0.6093,
+      "step": 7597
+    },
+    {
+      "epoch": 2.016991902296562,
+      "grad_norm": 0.4287442720345595,
+      "learning_rate": 3.750201563679561e-06,
+      "loss": 0.5907,
+      "step": 7598
+    },
+    {
+      "epoch": 2.0172574007699455,
+      "grad_norm": 0.45242643608015465,
+      "learning_rate": 3.7498992140971007e-06,
+      "loss": 0.53,
+      "step": 7599
+    },
+    {
+      "epoch": 2.017522899243329,
+      "grad_norm": 0.42679580149521024,
+      "learning_rate": 3.749596840138545e-06,
+      "loss": 0.5527,
+      "step": 7600
+    },
+    {
+      "epoch": 2.0177883977167133,
+      "grad_norm": 0.4170698822851295,
+      "learning_rate": 3.7492944418097913e-06,
+      "loss": 0.5646,
+      "step": 7601
+    },
+    {
+      "epoch": 2.018053896190097,
+      "grad_norm": 0.4246478310886877,
+      "learning_rate": 3.7489920191167376e-06,
+      "loss": 0.5789,
+      "step": 7602
+    },
+    {
+      "epoch": 2.0183193946634805,
+      "grad_norm": 0.4403978176363262,
+      "learning_rate": 3.7486895720652818e-06,
+      "loss": 0.5738,
+      "step": 7603
+    },
+    {
+      "epoch": 2.0185848931368646,
+      "grad_norm": 0.4371435385363269,
+      "learning_rate": 3.7483871006613207e-06,
+      "loss": 0.523,
+      "step": 7604
+    },
+    {
+      "epoch": 2.0188503916102483,
+      "grad_norm": 0.3992124329816631,
+      "learning_rate": 3.7480846049107555e-06,
+      "loss": 0.5333,
+      "step": 7605
+    },
+    {
+      "epoch": 2.019115890083632,
+      "grad_norm": 0.4100451739526877,
+      "learning_rate": 3.7477820848194846e-06,
+      "loss": 0.5503,
+      "step": 7606
+    },
+    {
+      "epoch": 2.019381388557016,
+      "grad_norm": 0.4123828841803448,
+      "learning_rate": 3.7474795403934083e-06,
+      "loss": 0.6124,
+      "step": 7607
+    },
+    {
+      "epoch": 2.0196468870303996,
+      "grad_norm": 0.43135332673488136,
+      "learning_rate": 3.747176971638426e-06,
+      "loss": 0.6032,
+      "step": 7608
+    },
+    {
+      "epoch": 2.0199123855037833,
+      "grad_norm": 0.42000556467921024,
+      "learning_rate": 3.746874378560439e-06,
+      "loss": 0.5929,
+      "step": 7609
+    },
+    {
+      "epoch": 2.0201778839771674,
+      "grad_norm": 0.46319565109114114,
+      "learning_rate": 3.7465717611653495e-06,
+      "loss": 0.5828,
+      "step": 7610
+    },
+    {
+      "epoch": 2.020443382450551,
+      "grad_norm": 0.4413532638846013,
+      "learning_rate": 3.746269119459058e-06,
+      "loss": 0.5787,
+      "step": 7611
+    },
+    {
+      "epoch": 2.0207088809239346,
+      "grad_norm": 0.41573442279836215,
+      "learning_rate": 3.745966453447467e-06,
+      "loss": 0.5654,
+      "step": 7612
+    },
+    {
+      "epoch": 2.0209743793973183,
+      "grad_norm": 0.4162260772594856,
+      "learning_rate": 3.7456637631364802e-06,
+      "loss": 0.5592,
+      "step": 7613
+    },
+    {
+      "epoch": 2.0212398778707024,
+      "grad_norm": 0.4410681515808186,
+      "learning_rate": 3.7453610485319993e-06,
+      "loss": 0.55,
+      "step": 7614
+    },
+    {
+      "epoch": 2.021505376344086,
+      "grad_norm": 0.4495847194498986,
+      "learning_rate": 3.7450583096399294e-06,
+      "loss": 0.5343,
+      "step": 7615
+    },
+    {
+      "epoch": 2.0217708748174696,
+      "grad_norm": 0.42121033902836386,
+      "learning_rate": 3.744755546466174e-06,
+      "loss": 0.5929,
+      "step": 7616
+    },
+    {
+      "epoch": 2.0220363732908537,
+      "grad_norm": 0.43407158503299403,
+      "learning_rate": 3.744452759016638e-06,
+      "loss": 0.6031,
+      "step": 7617
+    },
+    {
+      "epoch": 2.0223018717642374,
+      "grad_norm": 0.4186600240145897,
+      "learning_rate": 3.7441499472972254e-06,
+      "loss": 0.5669,
+      "step": 7618
+    },
+    {
+      "epoch": 2.022567370237621,
+      "grad_norm": 0.41865460694290674,
+      "learning_rate": 3.7438471113138434e-06,
+      "loss": 0.5922,
+      "step": 7619
+    },
+    {
+      "epoch": 2.022832868711005,
+      "grad_norm": 0.42194320821898607,
+      "learning_rate": 3.7435442510723967e-06,
+      "loss": 0.5646,
+      "step": 7620
+    },
+    {
+      "epoch": 2.0230983671843887,
+      "grad_norm": 0.4317384900363479,
+      "learning_rate": 3.7432413665787924e-06,
+      "loss": 0.5969,
+      "step": 7621
+    },
+    {
+      "epoch": 2.0233638656577724,
+      "grad_norm": 0.4221989337973331,
+      "learning_rate": 3.742938457838938e-06,
+      "loss": 0.5728,
+      "step": 7622
+    },
+    {
+      "epoch": 2.0236293641311565,
+      "grad_norm": 0.4302089253587822,
+      "learning_rate": 3.74263552485874e-06,
+      "loss": 0.5962,
+      "step": 7623
+    },
+    {
+      "epoch": 2.02389486260454,
+      "grad_norm": 0.4169788331119294,
+      "learning_rate": 3.742332567644107e-06,
+      "loss": 0.5741,
+      "step": 7624
+    },
+    {
+      "epoch": 2.0241603610779237,
+      "grad_norm": 0.4127579688579369,
+      "learning_rate": 3.7420295862009474e-06,
+      "loss": 0.5491,
+      "step": 7625
+    },
+    {
+      "epoch": 2.0244258595513074,
+      "grad_norm": 0.4240942212642293,
+      "learning_rate": 3.7417265805351694e-06,
+      "loss": 0.5652,
+      "step": 7626
+    },
+    {
+      "epoch": 2.0246913580246915,
+      "grad_norm": 0.4283571800703669,
+      "learning_rate": 3.7414235506526837e-06,
+      "loss": 0.5731,
+      "step": 7627
+    },
+    {
+      "epoch": 2.024956856498075,
+      "grad_norm": 0.4460787575655464,
+      "learning_rate": 3.7411204965593982e-06,
+      "loss": 0.5805,
+      "step": 7628
+    },
+    {
+      "epoch": 2.0252223549714587,
+      "grad_norm": 0.40790318427616307,
+      "learning_rate": 3.7408174182612247e-06,
+      "loss": 0.5861,
+      "step": 7629
+    },
+    {
+      "epoch": 2.025487853444843,
+      "grad_norm": 0.41931174937974974,
+      "learning_rate": 3.740514315764073e-06,
+      "loss": 0.552,
+      "step": 7630
+    },
+    {
+      "epoch": 2.0257533519182265,
+      "grad_norm": 0.418633780463944,
+      "learning_rate": 3.7402111890738545e-06,
+      "loss": 0.5542,
+      "step": 7631
+    },
+    {
+      "epoch": 2.02601885039161,
+      "grad_norm": 0.43929183786810916,
+      "learning_rate": 3.7399080381964825e-06,
+      "loss": 0.539,
+      "step": 7632
+    },
+    {
+      "epoch": 2.026284348864994,
+      "grad_norm": 0.4331059734537798,
+      "learning_rate": 3.7396048631378674e-06,
+      "loss": 0.5622,
+      "step": 7633
+    },
+    {
+      "epoch": 2.026549847338378,
+      "grad_norm": 0.4301383493546203,
+      "learning_rate": 3.7393016639039228e-06,
+      "loss": 0.5667,
+      "step": 7634
+    },
+    {
+      "epoch": 2.0268153458117615,
+      "grad_norm": 0.4248557675957671,
+      "learning_rate": 3.7389984405005604e-06,
+      "loss": 0.5844,
+      "step": 7635
+    },
+    {
+      "epoch": 2.0270808442851456,
+      "grad_norm": 0.4203227396536002,
+      "learning_rate": 3.7386951929336957e-06,
+      "loss": 0.5519,
+      "step": 7636
+    },
+    {
+      "epoch": 2.027346342758529,
+      "grad_norm": 0.4305163204228117,
+      "learning_rate": 3.7383919212092416e-06,
+      "loss": 0.5751,
+      "step": 7637
+    },
+    {
+      "epoch": 2.027611841231913,
+      "grad_norm": 0.43216988356933955,
+      "learning_rate": 3.7380886253331126e-06,
+      "loss": 0.5831,
+      "step": 7638
+    },
+    {
+      "epoch": 2.0278773397052965,
+      "grad_norm": 0.4377616383381146,
+      "learning_rate": 3.737785305311225e-06,
+      "loss": 0.5081,
+      "step": 7639
+    },
+    {
+      "epoch": 2.0281428381786806,
+      "grad_norm": 0.41417289928853834,
+      "learning_rate": 3.7374819611494927e-06,
+      "loss": 0.554,
+      "step": 7640
+    },
+    {
+      "epoch": 2.028408336652064,
+      "grad_norm": 0.4190053409408561,
+      "learning_rate": 3.737178592853832e-06,
+      "loss": 0.6212,
+      "step": 7641
+    },
+    {
+      "epoch": 2.028673835125448,
+      "grad_norm": 0.4745876794533292,
+      "learning_rate": 3.73687520043016e-06,
+      "loss": 0.5296,
+      "step": 7642
+    },
+    {
+      "epoch": 2.028939333598832,
+      "grad_norm": 0.4146131427047848,
+      "learning_rate": 3.7365717838843928e-06,
+      "loss": 0.5037,
+      "step": 7643
+    },
+    {
+      "epoch": 2.0292048320722156,
+      "grad_norm": 0.41213155652429123,
+      "learning_rate": 3.7362683432224484e-06,
+      "loss": 0.5598,
+      "step": 7644
+    },
+    {
+      "epoch": 2.029470330545599,
+      "grad_norm": 0.4267860938556258,
+      "learning_rate": 3.7359648784502444e-06,
+      "loss": 0.5625,
+      "step": 7645
+    },
+    {
+      "epoch": 2.0297358290189833,
+      "grad_norm": 0.43792749386608554,
+      "learning_rate": 3.7356613895736993e-06,
+      "loss": 0.5766,
+      "step": 7646
+    },
+    {
+      "epoch": 2.030001327492367,
+      "grad_norm": 0.44112459213729,
+      "learning_rate": 3.7353578765987315e-06,
+      "loss": 0.5795,
+      "step": 7647
+    },
+    {
+      "epoch": 2.0302668259657506,
+      "grad_norm": 0.44261619652241374,
+      "learning_rate": 3.7350543395312604e-06,
+      "loss": 0.531,
+      "step": 7648
+    },
+    {
+      "epoch": 2.0305323244391347,
+      "grad_norm": 0.4360306095858179,
+      "learning_rate": 3.734750778377206e-06,
+      "loss": 0.5928,
+      "step": 7649
+    },
+    {
+      "epoch": 2.0307978229125183,
+      "grad_norm": 0.4333461417705298,
+      "learning_rate": 3.734447193142488e-06,
+      "loss": 0.5462,
+      "step": 7650
+    },
+    {
+      "epoch": 2.031063321385902,
+      "grad_norm": 0.4666508118346647,
+      "learning_rate": 3.734143583833027e-06,
+      "loss": 0.559,
+      "step": 7651
+    },
+    {
+      "epoch": 2.0313288198592856,
+      "grad_norm": 0.43882769553941153,
+      "learning_rate": 3.7338399504547445e-06,
+      "loss": 0.5748,
+      "step": 7652
+    },
+    {
+      "epoch": 2.0315943183326697,
+      "grad_norm": 0.4194428029826446,
+      "learning_rate": 3.733536293013563e-06,
+      "loss": 0.5978,
+      "step": 7653
+    },
+    {
+      "epoch": 2.0318598168060533,
+      "grad_norm": 0.417744108275885,
+      "learning_rate": 3.7332326115154025e-06,
+      "loss": 0.5844,
+      "step": 7654
+    },
+    {
+      "epoch": 2.032125315279437,
+      "grad_norm": 0.4354069542579492,
+      "learning_rate": 3.7329289059661868e-06,
+      "loss": 0.5549,
+      "step": 7655
+    },
+    {
+      "epoch": 2.032390813752821,
+      "grad_norm": 0.42014331447022146,
+      "learning_rate": 3.732625176371839e-06,
+      "loss": 0.5597,
+      "step": 7656
+    },
+    {
+      "epoch": 2.0326563122262047,
+      "grad_norm": 0.41969329872377203,
+      "learning_rate": 3.732321422738282e-06,
+      "loss": 0.5816,
+      "step": 7657
+    },
+    {
+      "epoch": 2.0329218106995883,
+      "grad_norm": 0.44417633924115807,
+      "learning_rate": 3.73201764507144e-06,
+      "loss": 0.5443,
+      "step": 7658
+    },
+    {
+      "epoch": 2.0331873091729724,
+      "grad_norm": 0.43760628593704454,
+      "learning_rate": 3.7317138433772377e-06,
+      "loss": 0.5481,
+      "step": 7659
+    },
+    {
+      "epoch": 2.033452807646356,
+      "grad_norm": 0.43643749682735955,
+      "learning_rate": 3.731410017661599e-06,
+      "loss": 0.562,
+      "step": 7660
+    },
+    {
+      "epoch": 2.0337183061197397,
+      "grad_norm": 0.42125702993078595,
+      "learning_rate": 3.7311061679304506e-06,
+      "loss": 0.5775,
+      "step": 7661
+    },
+    {
+      "epoch": 2.033983804593124,
+      "grad_norm": 0.412025601256085,
+      "learning_rate": 3.730802294189718e-06,
+      "loss": 0.5401,
+      "step": 7662
+    },
+    {
+      "epoch": 2.0342493030665074,
+      "grad_norm": 0.40227071163334316,
+      "learning_rate": 3.730498396445326e-06,
+      "loss": 0.5538,
+      "step": 7663
+    },
+    {
+      "epoch": 2.034514801539891,
+      "grad_norm": 0.42522090826179904,
+      "learning_rate": 3.7301944747032037e-06,
+      "loss": 0.5707,
+      "step": 7664
+    },
+    {
+      "epoch": 2.034780300013275,
+      "grad_norm": 0.4076057269084683,
+      "learning_rate": 3.7298905289692768e-06,
+      "loss": 0.5489,
+      "step": 7665
+    },
+    {
+      "epoch": 2.035045798486659,
+      "grad_norm": 0.4261798306346251,
+      "learning_rate": 3.7295865592494727e-06,
+      "loss": 0.5942,
+      "step": 7666
+    },
+    {
+      "epoch": 2.0353112969600424,
+      "grad_norm": 0.40597294431936004,
+      "learning_rate": 3.7292825655497208e-06,
+      "loss": 0.5898,
+      "step": 7667
+    },
+    {
+      "epoch": 2.035576795433426,
+      "grad_norm": 0.42170416130366817,
+      "learning_rate": 3.728978547875948e-06,
+      "loss": 0.5609,
+      "step": 7668
+    },
+    {
+      "epoch": 2.03584229390681,
+      "grad_norm": 0.42465680129814404,
+      "learning_rate": 3.7286745062340856e-06,
+      "loss": 0.5939,
+      "step": 7669
+    },
+    {
+      "epoch": 2.036107792380194,
+      "grad_norm": 0.4260096417406519,
+      "learning_rate": 3.7283704406300613e-06,
+      "loss": 0.5669,
+      "step": 7670
+    },
+    {
+      "epoch": 2.0363732908535774,
+      "grad_norm": 0.4371695031094179,
+      "learning_rate": 3.728066351069807e-06,
+      "loss": 0.6047,
+      "step": 7671
+    },
+    {
+      "epoch": 2.0366387893269615,
+      "grad_norm": 0.43228718040309333,
+      "learning_rate": 3.7277622375592504e-06,
+      "loss": 0.5872,
+      "step": 7672
+    },
+    {
+      "epoch": 2.036904287800345,
+      "grad_norm": 0.42126725579026864,
+      "learning_rate": 3.727458100104325e-06,
+      "loss": 0.5564,
+      "step": 7673
+    },
+    {
+      "epoch": 2.037169786273729,
+      "grad_norm": 0.4203582358892254,
+      "learning_rate": 3.727153938710962e-06,
+      "loss": 0.56,
+      "step": 7674
+    },
+    {
+      "epoch": 2.037435284747113,
+      "grad_norm": 0.43054509101938965,
+      "learning_rate": 3.7268497533850912e-06,
+      "loss": 0.5924,
+      "step": 7675
+    },
+    {
+      "epoch": 2.0377007832204965,
+      "grad_norm": 0.4403891041563899,
+      "learning_rate": 3.726545544132647e-06,
+      "loss": 0.561,
+      "step": 7676
+    },
+    {
+      "epoch": 2.03796628169388,
+      "grad_norm": 0.4300042180236443,
+      "learning_rate": 3.7262413109595614e-06,
+      "loss": 0.5722,
+      "step": 7677
+    },
+    {
+      "epoch": 2.0382317801672643,
+      "grad_norm": 0.41465669546975464,
+      "learning_rate": 3.725937053871768e-06,
+      "loss": 0.5469,
+      "step": 7678
+    },
+    {
+      "epoch": 2.038497278640648,
+      "grad_norm": 0.41521138930570706,
+      "learning_rate": 3.7256327728752004e-06,
+      "loss": 0.5241,
+      "step": 7679
+    },
+    {
+      "epoch": 2.0387627771140315,
+      "grad_norm": 0.4333990338459932,
+      "learning_rate": 3.7253284679757936e-06,
+      "loss": 0.5455,
+      "step": 7680
+    },
+    {
+      "epoch": 2.039028275587415,
+      "grad_norm": 0.4424437364284227,
+      "learning_rate": 3.7250241391794807e-06,
+      "loss": 0.579,
+      "step": 7681
+    },
+    {
+      "epoch": 2.0392937740607993,
+      "grad_norm": 0.4205536813721641,
+      "learning_rate": 3.724719786492198e-06,
+      "loss": 0.5407,
+      "step": 7682
+    },
+    {
+      "epoch": 2.039559272534183,
+      "grad_norm": 0.4336418021408473,
+      "learning_rate": 3.7244154099198805e-06,
+      "loss": 0.5719,
+      "step": 7683
+    },
+    {
+      "epoch": 2.0398247710075665,
+      "grad_norm": 0.4185342525626771,
+      "learning_rate": 3.7241110094684648e-06,
+      "loss": 0.5998,
+      "step": 7684
+    },
+    {
+      "epoch": 2.0400902694809506,
+      "grad_norm": 0.4092990278385108,
+      "learning_rate": 3.7238065851438875e-06,
+      "loss": 0.5307,
+      "step": 7685
+    },
+    {
+      "epoch": 2.0403557679543343,
+      "grad_norm": 0.42220281642527613,
+      "learning_rate": 3.7235021369520846e-06,
+      "loss": 0.5948,
+      "step": 7686
+    },
+    {
+      "epoch": 2.040621266427718,
+      "grad_norm": 0.41946934430187743,
+      "learning_rate": 3.7231976648989952e-06,
+      "loss": 0.555,
+      "step": 7687
+    },
+    {
+      "epoch": 2.040886764901102,
+      "grad_norm": 0.4155330687715104,
+      "learning_rate": 3.722893168990557e-06,
+      "loss": 0.5624,
+      "step": 7688
+    },
+    {
+      "epoch": 2.0411522633744856,
+      "grad_norm": 0.43694318526983805,
+      "learning_rate": 3.7225886492327066e-06,
+      "loss": 0.5874,
+      "step": 7689
+    },
+    {
+      "epoch": 2.0414177618478693,
+      "grad_norm": 0.42589338686301553,
+      "learning_rate": 3.722284105631384e-06,
+      "loss": 0.601,
+      "step": 7690
+    },
+    {
+      "epoch": 2.0416832603212534,
+      "grad_norm": 0.41972012350617455,
+      "learning_rate": 3.72197953819253e-06,
+      "loss": 0.552,
+      "step": 7691
+    },
+    {
+      "epoch": 2.041948758794637,
+      "grad_norm": 0.4254822587465102,
+      "learning_rate": 3.7216749469220824e-06,
+      "loss": 0.5359,
+      "step": 7692
+    },
+    {
+      "epoch": 2.0422142572680206,
+      "grad_norm": 0.42636959739717756,
+      "learning_rate": 3.7213703318259818e-06,
+      "loss": 0.5773,
+      "step": 7693
+    },
+    {
+      "epoch": 2.0424797557414043,
+      "grad_norm": 0.43572473475673823,
+      "learning_rate": 3.7210656929101703e-06,
+      "loss": 0.6001,
+      "step": 7694
+    },
+    {
+      "epoch": 2.0427452542147884,
+      "grad_norm": 0.4228827725201327,
+      "learning_rate": 3.7207610301805873e-06,
+      "loss": 0.5574,
+      "step": 7695
+    },
+    {
+      "epoch": 2.043010752688172,
+      "grad_norm": 0.4397596116298814,
+      "learning_rate": 3.720456343643175e-06,
+      "loss": 0.5829,
+      "step": 7696
+    },
+    {
+      "epoch": 2.0432762511615556,
+      "grad_norm": 0.4259909422937582,
+      "learning_rate": 3.720151633303876e-06,
+      "loss": 0.5973,
+      "step": 7697
+    },
+    {
+      "epoch": 2.0435417496349397,
+      "grad_norm": 0.42250652170575065,
+      "learning_rate": 3.719846899168633e-06,
+      "loss": 0.5779,
+      "step": 7698
+    },
+    {
+      "epoch": 2.0438072481083234,
+      "grad_norm": 0.42582558909710705,
+      "learning_rate": 3.7195421412433886e-06,
+      "loss": 0.5898,
+      "step": 7699
+    },
+    {
+      "epoch": 2.044072746581707,
+      "grad_norm": 0.4374999968666794,
+      "learning_rate": 3.719237359534087e-06,
+      "loss": 0.6003,
+      "step": 7700
+    },
+    {
+      "epoch": 2.044338245055091,
+      "grad_norm": 0.43402508356364894,
+      "learning_rate": 3.718932554046671e-06,
+      "loss": 0.5316,
+      "step": 7701
+    },
+    {
+      "epoch": 2.0446037435284747,
+      "grad_norm": 0.4217374193641303,
+      "learning_rate": 3.718627724787086e-06,
+      "loss": 0.5294,
+      "step": 7702
+    },
+    {
+      "epoch": 2.0448692420018584,
+      "grad_norm": 0.4357271044509355,
+      "learning_rate": 3.718322871761277e-06,
+      "loss": 0.5812,
+      "step": 7703
+    },
+    {
+      "epoch": 2.0451347404752425,
+      "grad_norm": 0.418012641235714,
+      "learning_rate": 3.7180179949751883e-06,
+      "loss": 0.5752,
+      "step": 7704
+    },
+    {
+      "epoch": 2.045400238948626,
+      "grad_norm": 0.4301091068066781,
+      "learning_rate": 3.7177130944347666e-06,
+      "loss": 0.5386,
+      "step": 7705
+    },
+    {
+      "epoch": 2.0456657374220097,
+      "grad_norm": 0.4203131046534478,
+      "learning_rate": 3.7174081701459586e-06,
+      "loss": 0.5622,
+      "step": 7706
+    },
+    {
+      "epoch": 2.0459312358953934,
+      "grad_norm": 0.4240063413130722,
+      "learning_rate": 3.7171032221147103e-06,
+      "loss": 0.5582,
+      "step": 7707
+    },
+    {
+      "epoch": 2.0461967343687775,
+      "grad_norm": 0.4060322554264638,
+      "learning_rate": 3.716798250346969e-06,
+      "loss": 0.5161,
+      "step": 7708
+    },
+    {
+      "epoch": 2.046462232842161,
+      "grad_norm": 0.43352299455194343,
+      "learning_rate": 3.716493254848683e-06,
+      "loss": 0.5981,
+      "step": 7709
+    },
+    {
+      "epoch": 2.0467277313155448,
+      "grad_norm": 0.4211014646050209,
+      "learning_rate": 3.7161882356257993e-06,
+      "loss": 0.5362,
+      "step": 7710
+    },
+    {
+      "epoch": 2.046993229788929,
+      "grad_norm": 0.4293299229493533,
+      "learning_rate": 3.7158831926842687e-06,
+      "loss": 0.6055,
+      "step": 7711
+    },
+    {
+      "epoch": 2.0472587282623125,
+      "grad_norm": 0.41760152208003454,
+      "learning_rate": 3.715578126030037e-06,
+      "loss": 0.5693,
+      "step": 7712
+    },
+    {
+      "epoch": 2.047524226735696,
+      "grad_norm": 0.4243942442997636,
+      "learning_rate": 3.715273035669057e-06,
+      "loss": 0.556,
+      "step": 7713
+    },
+    {
+      "epoch": 2.04778972520908,
+      "grad_norm": 0.4173919697997775,
+      "learning_rate": 3.7149679216072765e-06,
+      "loss": 0.5945,
+      "step": 7714
+    },
+    {
+      "epoch": 2.048055223682464,
+      "grad_norm": 0.45992829964457643,
+      "learning_rate": 3.7146627838506472e-06,
+      "loss": 0.5588,
+      "step": 7715
+    },
+    {
+      "epoch": 2.0483207221558475,
+      "grad_norm": 0.43886136896037264,
+      "learning_rate": 3.714357622405119e-06,
+      "loss": 0.5657,
+      "step": 7716
+    },
+    {
+      "epoch": 2.0485862206292316,
+      "grad_norm": 0.42856309454913505,
+      "learning_rate": 3.714052437276645e-06,
+      "loss": 0.5857,
+      "step": 7717
+    },
+    {
+      "epoch": 2.048851719102615,
+      "grad_norm": 0.4245673323923108,
+      "learning_rate": 3.713747228471175e-06,
+      "loss": 0.578,
+      "step": 7718
+    },
+    {
+      "epoch": 2.049117217575999,
+      "grad_norm": 0.42822911570742994,
+      "learning_rate": 3.7134419959946626e-06,
+      "loss": 0.6105,
+      "step": 7719
+    },
+    {
+      "epoch": 2.049382716049383,
+      "grad_norm": 0.43342355652095027,
+      "learning_rate": 3.7131367398530605e-06,
+      "loss": 0.5639,
+      "step": 7720
+    },
+    {
+      "epoch": 2.0496482145227666,
+      "grad_norm": 0.42306124914564985,
+      "learning_rate": 3.7128314600523217e-06,
+      "loss": 0.53,
+      "step": 7721
+    },
+    {
+      "epoch": 2.04991371299615,
+      "grad_norm": 0.429524960789082,
+      "learning_rate": 3.7125261565983996e-06,
+      "loss": 0.5773,
+      "step": 7722
+    },
+    {
+      "epoch": 2.050179211469534,
+      "grad_norm": 0.4165027448814866,
+      "learning_rate": 3.7122208294972484e-06,
+      "loss": 0.5814,
+      "step": 7723
+    },
+    {
+      "epoch": 2.050444709942918,
+      "grad_norm": 0.41478576095573133,
+      "learning_rate": 3.7119154787548233e-06,
+      "loss": 0.5573,
+      "step": 7724
+    },
+    {
+      "epoch": 2.0507102084163016,
+      "grad_norm": 0.42798827161094444,
+      "learning_rate": 3.711610104377079e-06,
+      "loss": 0.6165,
+      "step": 7725
+    },
+    {
+      "epoch": 2.0509757068896852,
+      "grad_norm": 0.4381756160787756,
+      "learning_rate": 3.7113047063699716e-06,
+      "loss": 0.5448,
+      "step": 7726
+    },
+    {
+      "epoch": 2.0512412053630693,
+      "grad_norm": 0.4380547496917146,
+      "learning_rate": 3.710999284739456e-06,
+      "loss": 0.5532,
+      "step": 7727
+    },
+    {
+      "epoch": 2.051506703836453,
+      "grad_norm": 0.4091692603052029,
+      "learning_rate": 3.7106938394914905e-06,
+      "loss": 0.5262,
+      "step": 7728
+    },
+    {
+      "epoch": 2.0517722023098366,
+      "grad_norm": 0.42371305527673586,
+      "learning_rate": 3.7103883706320298e-06,
+      "loss": 0.5913,
+      "step": 7729
+    },
+    {
+      "epoch": 2.0520377007832207,
+      "grad_norm": 0.44482620761765956,
+      "learning_rate": 3.710082878167032e-06,
+      "loss": 0.6009,
+      "step": 7730
+    },
+    {
+      "epoch": 2.0523031992566043,
+      "grad_norm": 0.4478590625609402,
+      "learning_rate": 3.709777362102456e-06,
+      "loss": 0.6007,
+      "step": 7731
+    },
+    {
+      "epoch": 2.052568697729988,
+      "grad_norm": 0.4200693991034797,
+      "learning_rate": 3.709471822444259e-06,
+      "loss": 0.5158,
+      "step": 7732
+    },
+    {
+      "epoch": 2.052834196203372,
+      "grad_norm": 0.4274649048330147,
+      "learning_rate": 3.7091662591984e-06,
+      "loss": 0.5679,
+      "step": 7733
+    },
+    {
+      "epoch": 2.0530996946767557,
+      "grad_norm": 0.4265328375630852,
+      "learning_rate": 3.7088606723708393e-06,
+      "loss": 0.541,
+      "step": 7734
+    },
+    {
+      "epoch": 2.0533651931501393,
+      "grad_norm": 0.43518243420586755,
+      "learning_rate": 3.7085550619675347e-06,
+      "loss": 0.5586,
+      "step": 7735
+    },
+    {
+      "epoch": 2.053630691623523,
+      "grad_norm": 0.4214372646265983,
+      "learning_rate": 3.708249427994448e-06,
+      "loss": 0.5604,
+      "step": 7736
+    },
+    {
+      "epoch": 2.053896190096907,
+      "grad_norm": 0.41798223480817026,
+      "learning_rate": 3.7079437704575395e-06,
+      "loss": 0.5592,
+      "step": 7737
+    },
+    {
+      "epoch": 2.0541616885702907,
+      "grad_norm": 0.42885288716245795,
+      "learning_rate": 3.707638089362769e-06,
+      "loss": 0.5929,
+      "step": 7738
+    },
+    {
+      "epoch": 2.0544271870436743,
+      "grad_norm": 0.4152669744642473,
+      "learning_rate": 3.7073323847160992e-06,
+      "loss": 0.6087,
+      "step": 7739
+    },
+    {
+      "epoch": 2.0546926855170584,
+      "grad_norm": 0.4269617943515965,
+      "learning_rate": 3.7070266565234917e-06,
+      "loss": 0.5771,
+      "step": 7740
+    },
+    {
+      "epoch": 2.054958183990442,
+      "grad_norm": 0.43199520898848653,
+      "learning_rate": 3.706720904790909e-06,
+      "loss": 0.6017,
+      "step": 7741
+    },
+    {
+      "epoch": 2.0552236824638257,
+      "grad_norm": 0.4200441556777171,
+      "learning_rate": 3.7064151295243146e-06,
+      "loss": 0.5731,
+      "step": 7742
+    },
+    {
+      "epoch": 2.05548918093721,
+      "grad_norm": 0.4302128136855923,
+      "learning_rate": 3.706109330729671e-06,
+      "loss": 0.566,
+      "step": 7743
+    },
+    {
+      "epoch": 2.0557546794105934,
+      "grad_norm": 0.4157932270518941,
+      "learning_rate": 3.7058035084129427e-06,
+      "loss": 0.5234,
+      "step": 7744
+    },
+    {
+      "epoch": 2.056020177883977,
+      "grad_norm": 0.42383304192797683,
+      "learning_rate": 3.705497662580093e-06,
+      "loss": 0.578,
+      "step": 7745
+    },
+    {
+      "epoch": 2.056285676357361,
+      "grad_norm": 0.418269199905343,
+      "learning_rate": 3.7051917932370886e-06,
+      "loss": 0.5404,
+      "step": 7746
+    },
+    {
+      "epoch": 2.056551174830745,
+      "grad_norm": 0.43038760469113735,
+      "learning_rate": 3.704885900389892e-06,
+      "loss": 0.5853,
+      "step": 7747
+    },
+    {
+      "epoch": 2.0568166733041284,
+      "grad_norm": 0.42837077521366473,
+      "learning_rate": 3.7045799840444712e-06,
+      "loss": 0.5683,
+      "step": 7748
+    },
+    {
+      "epoch": 2.057082171777512,
+      "grad_norm": 0.4159337958849824,
+      "learning_rate": 3.7042740442067914e-06,
+      "loss": 0.5873,
+      "step": 7749
+    },
+    {
+      "epoch": 2.057347670250896,
+      "grad_norm": 0.4155364626869827,
+      "learning_rate": 3.7039680808828187e-06,
+      "loss": 0.5045,
+      "step": 7750
+    },
+    {
+      "epoch": 2.05761316872428,
+      "grad_norm": 0.4147872071239813,
+      "learning_rate": 3.7036620940785205e-06,
+      "loss": 0.5425,
+      "step": 7751
+    },
+    {
+      "epoch": 2.0578786671976634,
+      "grad_norm": 0.43504772272129594,
+      "learning_rate": 3.703356083799865e-06,
+      "loss": 0.5861,
+      "step": 7752
+    },
+    {
+      "epoch": 2.0581441656710475,
+      "grad_norm": 0.4417888768490918,
+      "learning_rate": 3.703050050052819e-06,
+      "loss": 0.5967,
+      "step": 7753
+    },
+    {
+      "epoch": 2.058409664144431,
+      "grad_norm": 0.44110579582481657,
+      "learning_rate": 3.702743992843352e-06,
+      "loss": 0.6009,
+      "step": 7754
+    },
+    {
+      "epoch": 2.058675162617815,
+      "grad_norm": 0.44121555242124766,
+      "learning_rate": 3.7024379121774316e-06,
+      "loss": 0.5667,
+      "step": 7755
+    },
+    {
+      "epoch": 2.058940661091199,
+      "grad_norm": 0.4413542512011281,
+      "learning_rate": 3.702131808061028e-06,
+      "loss": 0.5714,
+      "step": 7756
+    },
+    {
+      "epoch": 2.0592061595645825,
+      "grad_norm": 0.41821395731660016,
+      "learning_rate": 3.701825680500112e-06,
+      "loss": 0.5289,
+      "step": 7757
+    },
+    {
+      "epoch": 2.059471658037966,
+      "grad_norm": 0.4264440248501459,
+      "learning_rate": 3.701519529500651e-06,
+      "loss": 0.5664,
+      "step": 7758
+    },
+    {
+      "epoch": 2.0597371565113503,
+      "grad_norm": 0.4348680370036617,
+      "learning_rate": 3.701213355068618e-06,
+      "loss": 0.5764,
+      "step": 7759
+    },
+    {
+      "epoch": 2.060002654984734,
+      "grad_norm": 0.4169045795485788,
+      "learning_rate": 3.7009071572099837e-06,
+      "loss": 0.5458,
+      "step": 7760
+    },
+    {
+      "epoch": 2.0602681534581175,
+      "grad_norm": 0.4212789827782027,
+      "learning_rate": 3.7006009359307193e-06,
+      "loss": 0.5645,
+      "step": 7761
+    },
+    {
+      "epoch": 2.060533651931501,
+      "grad_norm": 0.4294729359979502,
+      "learning_rate": 3.7002946912367964e-06,
+      "loss": 0.5873,
+      "step": 7762
+    },
+    {
+      "epoch": 2.0607991504048853,
+      "grad_norm": 0.42833423953440547,
+      "learning_rate": 3.6999884231341897e-06,
+      "loss": 0.5573,
+      "step": 7763
+    },
+    {
+      "epoch": 2.061064648878269,
+      "grad_norm": 0.4129329112730173,
+      "learning_rate": 3.6996821316288694e-06,
+      "loss": 0.5151,
+      "step": 7764
+    },
+    {
+      "epoch": 2.0613301473516525,
+      "grad_norm": 0.4136739284238691,
+      "learning_rate": 3.6993758167268106e-06,
+      "loss": 0.5752,
+      "step": 7765
+    },
+    {
+      "epoch": 2.0615956458250366,
+      "grad_norm": 0.4277226213962251,
+      "learning_rate": 3.6990694784339874e-06,
+      "loss": 0.5866,
+      "step": 7766
+    },
+    {
+      "epoch": 2.0618611442984203,
+      "grad_norm": 0.45117832437480265,
+      "learning_rate": 3.698763116756373e-06,
+      "loss": 0.5296,
+      "step": 7767
+    },
+    {
+      "epoch": 2.062126642771804,
+      "grad_norm": 0.4359516593196853,
+      "learning_rate": 3.6984567316999424e-06,
+      "loss": 0.5816,
+      "step": 7768
+    },
+    {
+      "epoch": 2.062392141245188,
+      "grad_norm": 0.42430459215163924,
+      "learning_rate": 3.698150323270672e-06,
+      "loss": 0.5559,
+      "step": 7769
+    },
+    {
+      "epoch": 2.0626576397185716,
+      "grad_norm": 0.4316193966556973,
+      "learning_rate": 3.697843891474536e-06,
+      "loss": 0.5435,
+      "step": 7770
+    },
+    {
+      "epoch": 2.0629231381919553,
+      "grad_norm": 0.4201709266798109,
+      "learning_rate": 3.6975374363175107e-06,
+      "loss": 0.6033,
+      "step": 7771
+    },
+    {
+      "epoch": 2.0631886366653394,
+      "grad_norm": 0.41667461001322864,
+      "learning_rate": 3.697230957805575e-06,
+      "loss": 0.6042,
+      "step": 7772
+    },
+    {
+      "epoch": 2.063454135138723,
+      "grad_norm": 0.4070286357155365,
+      "learning_rate": 3.696924455944703e-06,
+      "loss": 0.5628,
+      "step": 7773
+    },
+    {
+      "epoch": 2.0637196336121066,
+      "grad_norm": 0.4240294318916209,
+      "learning_rate": 3.6966179307408747e-06,
+      "loss": 0.6006,
+      "step": 7774
+    },
+    {
+      "epoch": 2.0639851320854907,
+      "grad_norm": 0.4333631719587378,
+      "learning_rate": 3.6963113822000663e-06,
+      "loss": 0.6029,
+      "step": 7775
+    },
+    {
+      "epoch": 2.0642506305588744,
+      "grad_norm": 0.4266711901374407,
+      "learning_rate": 3.6960048103282563e-06,
+      "loss": 0.5604,
+      "step": 7776
+    },
+    {
+      "epoch": 2.064516129032258,
+      "grad_norm": 0.4217592571285186,
+      "learning_rate": 3.6956982151314252e-06,
+      "loss": 0.5784,
+      "step": 7777
+    },
+    {
+      "epoch": 2.0647816275056416,
+      "grad_norm": 0.4178207373455119,
+      "learning_rate": 3.69539159661555e-06,
+      "loss": 0.5623,
+      "step": 7778
+    },
+    {
+      "epoch": 2.0650471259790257,
+      "grad_norm": 0.43476193943169533,
+      "learning_rate": 3.695084954786613e-06,
+      "loss": 0.5765,
+      "step": 7779
+    },
+    {
+      "epoch": 2.0653126244524094,
+      "grad_norm": 0.41752260794030094,
+      "learning_rate": 3.694778289650593e-06,
+      "loss": 0.5727,
+      "step": 7780
+    },
+    {
+      "epoch": 2.065578122925793,
+      "grad_norm": 0.42570125662112296,
+      "learning_rate": 3.6944716012134707e-06,
+      "loss": 0.5573,
+      "step": 7781
+    },
+    {
+      "epoch": 2.065843621399177,
+      "grad_norm": 0.41994692482972573,
+      "learning_rate": 3.6941648894812284e-06,
+      "loss": 0.55,
+      "step": 7782
+    },
+    {
+      "epoch": 2.0661091198725607,
+      "grad_norm": 0.42563655301362135,
+      "learning_rate": 3.693858154459846e-06,
+      "loss": 0.5887,
+      "step": 7783
+    },
+    {
+      "epoch": 2.0663746183459444,
+      "grad_norm": 0.42205018601159106,
+      "learning_rate": 3.6935513961553073e-06,
+      "loss": 0.5659,
+      "step": 7784
+    },
+    {
+      "epoch": 2.0666401168193285,
+      "grad_norm": 0.4376486616043209,
+      "learning_rate": 3.6932446145735933e-06,
+      "loss": 0.5667,
+      "step": 7785
+    },
+    {
+      "epoch": 2.066905615292712,
+      "grad_norm": 0.422678319460469,
+      "learning_rate": 3.6929378097206887e-06,
+      "loss": 0.5637,
+      "step": 7786
+    },
+    {
+      "epoch": 2.0671711137660957,
+      "grad_norm": 0.4365645980330758,
+      "learning_rate": 3.692630981602575e-06,
+      "loss": 0.5637,
+      "step": 7787
+    },
+    {
+      "epoch": 2.06743661223948,
+      "grad_norm": 0.42443119232306886,
+      "learning_rate": 3.6923241302252374e-06,
+      "loss": 0.5754,
+      "step": 7788
+    },
+    {
+      "epoch": 2.0677021107128635,
+      "grad_norm": 0.4271289335997637,
+      "learning_rate": 3.6920172555946603e-06,
+      "loss": 0.546,
+      "step": 7789
+    },
+    {
+      "epoch": 2.067967609186247,
+      "grad_norm": 0.42034105394703203,
+      "learning_rate": 3.691710357716828e-06,
+      "loss": 0.5659,
+      "step": 7790
+    },
+    {
+      "epoch": 2.0682331076596308,
+      "grad_norm": 0.43598933072187385,
+      "learning_rate": 3.691403436597726e-06,
+      "loss": 0.6016,
+      "step": 7791
+    },
+    {
+      "epoch": 2.068498606133015,
+      "grad_norm": 0.42939731646220797,
+      "learning_rate": 3.69109649224334e-06,
+      "loss": 0.576,
+      "step": 7792
+    },
+    {
+      "epoch": 2.0687641046063985,
+      "grad_norm": 0.42126264339623465,
+      "learning_rate": 3.6907895246596557e-06,
+      "loss": 0.5835,
+      "step": 7793
+    },
+    {
+      "epoch": 2.069029603079782,
+      "grad_norm": 0.41874512403417113,
+      "learning_rate": 3.69048253385266e-06,
+      "loss": 0.5481,
+      "step": 7794
+    },
+    {
+      "epoch": 2.069295101553166,
+      "grad_norm": 0.4430669140381732,
+      "learning_rate": 3.6901755198283407e-06,
+      "loss": 0.5892,
+      "step": 7795
+    },
+    {
+      "epoch": 2.06956060002655,
+      "grad_norm": 0.42390619153150105,
+      "learning_rate": 3.6898684825926845e-06,
+      "loss": 0.5641,
+      "step": 7796
+    },
+    {
+      "epoch": 2.0698260984999335,
+      "grad_norm": 0.4310311062309684,
+      "learning_rate": 3.6895614221516797e-06,
+      "loss": 0.6118,
+      "step": 7797
+    },
+    {
+      "epoch": 2.0700915969733176,
+      "grad_norm": 0.43030874812934844,
+      "learning_rate": 3.6892543385113137e-06,
+      "loss": 0.5567,
+      "step": 7798
+    },
+    {
+      "epoch": 2.070357095446701,
+      "grad_norm": 0.4218225175263902,
+      "learning_rate": 3.688947231677577e-06,
+      "loss": 0.5876,
+      "step": 7799
+    },
+    {
+      "epoch": 2.070622593920085,
+      "grad_norm": 0.4349945784479379,
+      "learning_rate": 3.688640101656459e-06,
+      "loss": 0.617,
+      "step": 7800
+    },
+    {
+      "epoch": 2.070888092393469,
+      "grad_norm": 0.4416526910478568,
+      "learning_rate": 3.688332948453948e-06,
+      "loss": 0.5852,
+      "step": 7801
+    },
+    {
+      "epoch": 2.0711535908668526,
+      "grad_norm": 0.4001023980887157,
+      "learning_rate": 3.6880257720760343e-06,
+      "loss": 0.5177,
+      "step": 7802
+    },
+    {
+      "epoch": 2.071419089340236,
+      "grad_norm": 0.4059228460822013,
+      "learning_rate": 3.6877185725287102e-06,
+      "loss": 0.5201,
+      "step": 7803
+    },
+    {
+      "epoch": 2.07168458781362,
+      "grad_norm": 0.4369884170155583,
+      "learning_rate": 3.6874113498179653e-06,
+      "loss": 0.5785,
+      "step": 7804
+    },
+    {
+      "epoch": 2.071950086287004,
+      "grad_norm": 0.4149170948361833,
+      "learning_rate": 3.6871041039497924e-06,
+      "loss": 0.5795,
+      "step": 7805
+    },
+    {
+      "epoch": 2.0722155847603876,
+      "grad_norm": 0.43479784841791536,
+      "learning_rate": 3.6867968349301825e-06,
+      "loss": 0.5827,
+      "step": 7806
+    },
+    {
+      "epoch": 2.0724810832337712,
+      "grad_norm": 0.42217766096651677,
+      "learning_rate": 3.686489542765128e-06,
+      "loss": 0.5708,
+      "step": 7807
+    },
+    {
+      "epoch": 2.0727465817071553,
+      "grad_norm": 0.441584051134115,
+      "learning_rate": 3.686182227460623e-06,
+      "loss": 0.5584,
+      "step": 7808
+    },
+    {
+      "epoch": 2.073012080180539,
+      "grad_norm": 0.4252942384765948,
+      "learning_rate": 3.6858748890226607e-06,
+      "loss": 0.5413,
+      "step": 7809
+    },
+    {
+      "epoch": 2.0732775786539226,
+      "grad_norm": 0.42717772899322654,
+      "learning_rate": 3.6855675274572343e-06,
+      "loss": 0.5658,
+      "step": 7810
+    },
+    {
+      "epoch": 2.0735430771273067,
+      "grad_norm": 0.4298944965076522,
+      "learning_rate": 3.685260142770338e-06,
+      "loss": 0.5544,
+      "step": 7811
+    },
+    {
+      "epoch": 2.0738085756006903,
+      "grad_norm": 0.43074005346617233,
+      "learning_rate": 3.6849527349679674e-06,
+      "loss": 0.6079,
+      "step": 7812
+    },
+    {
+      "epoch": 2.074074074074074,
+      "grad_norm": 0.4400285741011014,
+      "learning_rate": 3.684645304056117e-06,
+      "loss": 0.5894,
+      "step": 7813
+    },
+    {
+      "epoch": 2.074339572547458,
+      "grad_norm": 0.4306261837250573,
+      "learning_rate": 3.6843378500407824e-06,
+      "loss": 0.5645,
+      "step": 7814
+    },
+    {
+      "epoch": 2.0746050710208417,
+      "grad_norm": 0.4503920983394631,
+      "learning_rate": 3.6840303729279604e-06,
+      "loss": 0.5602,
+      "step": 7815
+    },
+    {
+      "epoch": 2.0748705694942253,
+      "grad_norm": 0.4197862097696484,
+      "learning_rate": 3.6837228727236464e-06,
+      "loss": 0.5854,
+      "step": 7816
+    },
+    {
+      "epoch": 2.075136067967609,
+      "grad_norm": 0.4202525082448422,
+      "learning_rate": 3.6834153494338387e-06,
+      "loss": 0.602,
+      "step": 7817
+    },
+    {
+      "epoch": 2.075401566440993,
+      "grad_norm": 0.43072344293661036,
+      "learning_rate": 3.6831078030645346e-06,
+      "loss": 0.5567,
+      "step": 7818
+    },
+    {
+      "epoch": 2.0756670649143767,
+      "grad_norm": 0.4122247956530686,
+      "learning_rate": 3.6828002336217307e-06,
+      "loss": 0.5537,
+      "step": 7819
+    },
+    {
+      "epoch": 2.0759325633877603,
+      "grad_norm": 0.4281749161669776,
+      "learning_rate": 3.6824926411114265e-06,
+      "loss": 0.566,
+      "step": 7820
+    },
+    {
+      "epoch": 2.0761980618611444,
+      "grad_norm": 0.4358615646343696,
+      "learning_rate": 3.682185025539621e-06,
+      "loss": 0.6005,
+      "step": 7821
+    },
+    {
+      "epoch": 2.076463560334528,
+      "grad_norm": 0.421684392003355,
+      "learning_rate": 3.6818773869123127e-06,
+      "loss": 0.5749,
+      "step": 7822
+    },
+    {
+      "epoch": 2.0767290588079117,
+      "grad_norm": 0.4264060883500459,
+      "learning_rate": 3.6815697252355016e-06,
+      "loss": 0.5729,
+      "step": 7823
+    },
+    {
+      "epoch": 2.076994557281296,
+      "grad_norm": 0.4322489423194596,
+      "learning_rate": 3.6812620405151876e-06,
+      "loss": 0.5565,
+      "step": 7824
+    },
+    {
+      "epoch": 2.0772600557546794,
+      "grad_norm": 0.4221332849925765,
+      "learning_rate": 3.6809543327573715e-06,
+      "loss": 0.5759,
+      "step": 7825
+    },
+    {
+      "epoch": 2.077525554228063,
+      "grad_norm": 0.4043112791709982,
+      "learning_rate": 3.6806466019680552e-06,
+      "loss": 0.5482,
+      "step": 7826
+    },
+    {
+      "epoch": 2.077791052701447,
+      "grad_norm": 0.44061677408904004,
+      "learning_rate": 3.6803388481532393e-06,
+      "loss": 0.5584,
+      "step": 7827
+    },
+    {
+      "epoch": 2.078056551174831,
+      "grad_norm": 0.4494377632037603,
+      "learning_rate": 3.6800310713189258e-06,
+      "loss": 0.5335,
+      "step": 7828
+    },
+    {
+      "epoch": 2.0783220496482144,
+      "grad_norm": 0.422506496395252,
+      "learning_rate": 3.679723271471117e-06,
+      "loss": 0.5445,
+      "step": 7829
+    },
+    {
+      "epoch": 2.0785875481215985,
+      "grad_norm": 0.42260268368289405,
+      "learning_rate": 3.6794154486158163e-06,
+      "loss": 0.5836,
+      "step": 7830
+    },
+    {
+      "epoch": 2.078853046594982,
+      "grad_norm": 0.4392693482202733,
+      "learning_rate": 3.6791076027590257e-06,
+      "loss": 0.617,
+      "step": 7831
+    },
+    {
+      "epoch": 2.079118545068366,
+      "grad_norm": 0.5520143619580332,
+      "learning_rate": 3.678799733906751e-06,
+      "loss": 0.5524,
+      "step": 7832
+    },
+    {
+      "epoch": 2.0793840435417494,
+      "grad_norm": 0.44238743355848575,
+      "learning_rate": 3.6784918420649952e-06,
+      "loss": 0.5948,
+      "step": 7833
+    },
+    {
+      "epoch": 2.0796495420151335,
+      "grad_norm": 0.4160650541061754,
+      "learning_rate": 3.6781839272397616e-06,
+      "loss": 0.5558,
+      "step": 7834
+    },
+    {
+      "epoch": 2.079915040488517,
+      "grad_norm": 0.4661944410342518,
+      "learning_rate": 3.677875989437058e-06,
+      "loss": 0.5924,
+      "step": 7835
+    },
+    {
+      "epoch": 2.080180538961901,
+      "grad_norm": 0.45399211395202205,
+      "learning_rate": 3.6775680286628886e-06,
+      "loss": 0.5928,
+      "step": 7836
+    },
+    {
+      "epoch": 2.080446037435285,
+      "grad_norm": 0.44154831088542723,
+      "learning_rate": 3.6772600449232594e-06,
+      "loss": 0.5499,
+      "step": 7837
+    },
+    {
+      "epoch": 2.0807115359086685,
+      "grad_norm": 0.43493627876892177,
+      "learning_rate": 3.676952038224177e-06,
+      "loss": 0.5467,
+      "step": 7838
+    },
+    {
+      "epoch": 2.080977034382052,
+      "grad_norm": 0.4406376060046193,
+      "learning_rate": 3.6766440085716478e-06,
+      "loss": 0.5637,
+      "step": 7839
+    },
+    {
+      "epoch": 2.0812425328554363,
+      "grad_norm": 0.44109584523635187,
+      "learning_rate": 3.6763359559716793e-06,
+      "loss": 0.5541,
+      "step": 7840
+    },
+    {
+      "epoch": 2.08150803132882,
+      "grad_norm": 0.4297389256959757,
+      "learning_rate": 3.676027880430281e-06,
+      "loss": 0.5512,
+      "step": 7841
+    },
+    {
+      "epoch": 2.0817735298022035,
+      "grad_norm": 0.42799420454007564,
+      "learning_rate": 3.675719781953458e-06,
+      "loss": 0.5652,
+      "step": 7842
+    },
+    {
+      "epoch": 2.0820390282755876,
+      "grad_norm": 0.4288910378779645,
+      "learning_rate": 3.6754116605472205e-06,
+      "loss": 0.5994,
+      "step": 7843
+    },
+    {
+      "epoch": 2.0823045267489713,
+      "grad_norm": 0.4278300480846039,
+      "learning_rate": 3.6751035162175784e-06,
+      "loss": 0.5502,
+      "step": 7844
+    },
+    {
+      "epoch": 2.082570025222355,
+      "grad_norm": 0.43631333428729263,
+      "learning_rate": 3.67479534897054e-06,
+      "loss": 0.5565,
+      "step": 7845
+    },
+    {
+      "epoch": 2.0828355236957385,
+      "grad_norm": 0.45569801737704285,
+      "learning_rate": 3.6744871588121173e-06,
+      "loss": 0.577,
+      "step": 7846
+    },
+    {
+      "epoch": 2.0831010221691226,
+      "grad_norm": 0.433607166444531,
+      "learning_rate": 3.674178945748318e-06,
+      "loss": 0.6072,
+      "step": 7847
+    },
+    {
+      "epoch": 2.0833665206425063,
+      "grad_norm": 0.42771059784218063,
+      "learning_rate": 3.673870709785155e-06,
+      "loss": 0.5176,
+      "step": 7848
+    },
+    {
+      "epoch": 2.08363201911589,
+      "grad_norm": 0.4327113037110931,
+      "learning_rate": 3.6735624509286393e-06,
+      "loss": 0.5737,
+      "step": 7849
+    },
+    {
+      "epoch": 2.083897517589274,
+      "grad_norm": 0.4236840089192895,
+      "learning_rate": 3.6732541691847816e-06,
+      "loss": 0.582,
+      "step": 7850
+    },
+    {
+      "epoch": 2.0841630160626576,
+      "grad_norm": 0.4187295235442086,
+      "learning_rate": 3.6729458645595957e-06,
+      "loss": 0.589,
+      "step": 7851
+    },
+    {
+      "epoch": 2.0844285145360413,
+      "grad_norm": 0.44070146895733775,
+      "learning_rate": 3.6726375370590927e-06,
+      "loss": 0.5637,
+      "step": 7852
+    },
+    {
+      "epoch": 2.0846940130094254,
+      "grad_norm": 0.4443438621289547,
+      "learning_rate": 3.6723291866892872e-06,
+      "loss": 0.5556,
+      "step": 7853
+    },
+    {
+      "epoch": 2.084959511482809,
+      "grad_norm": 0.4342890464262517,
+      "learning_rate": 3.6720208134561918e-06,
+      "loss": 0.565,
+      "step": 7854
+    },
+    {
+      "epoch": 2.0852250099561926,
+      "grad_norm": 0.4302908075942359,
+      "learning_rate": 3.671712417365822e-06,
+      "loss": 0.584,
+      "step": 7855
+    },
+    {
+      "epoch": 2.0854905084295767,
+      "grad_norm": 0.460500205522762,
+      "learning_rate": 3.67140399842419e-06,
+      "loss": 0.5568,
+      "step": 7856
+    },
+    {
+      "epoch": 2.0857560069029604,
+      "grad_norm": 0.4284636655785662,
+      "learning_rate": 3.6710955566373125e-06,
+      "loss": 0.5904,
+      "step": 7857
+    },
+    {
+      "epoch": 2.086021505376344,
+      "grad_norm": 0.4313426045883189,
+      "learning_rate": 3.670787092011205e-06,
+      "loss": 0.5881,
+      "step": 7858
+    },
+    {
+      "epoch": 2.0862870038497277,
+      "grad_norm": 0.4252411605383489,
+      "learning_rate": 3.6704786045518815e-06,
+      "loss": 0.5788,
+      "step": 7859
+    },
+    {
+      "epoch": 2.0865525023231117,
+      "grad_norm": 0.42749198497027774,
+      "learning_rate": 3.6701700942653595e-06,
+      "loss": 0.5627,
+      "step": 7860
+    },
+    {
+      "epoch": 2.0868180007964954,
+      "grad_norm": 0.43148041896821854,
+      "learning_rate": 3.669861561157656e-06,
+      "loss": 0.5496,
+      "step": 7861
+    },
+    {
+      "epoch": 2.087083499269879,
+      "grad_norm": 0.4143656108805999,
+      "learning_rate": 3.669553005234787e-06,
+      "loss": 0.5581,
+      "step": 7862
+    },
+    {
+      "epoch": 2.087348997743263,
+      "grad_norm": 0.4189094219989511,
+      "learning_rate": 3.6692444265027716e-06,
+      "loss": 0.555,
+      "step": 7863
+    },
+    {
+      "epoch": 2.0876144962166467,
+      "grad_norm": 0.42629701565663236,
+      "learning_rate": 3.6689358249676276e-06,
+      "loss": 0.5688,
+      "step": 7864
+    },
+    {
+      "epoch": 2.0878799946900304,
+      "grad_norm": 0.42472258749475905,
+      "learning_rate": 3.668627200635372e-06,
+      "loss": 0.5918,
+      "step": 7865
+    },
+    {
+      "epoch": 2.0881454931634145,
+      "grad_norm": 0.4435142083748394,
+      "learning_rate": 3.6683185535120246e-06,
+      "loss": 0.5646,
+      "step": 7866
+    },
+    {
+      "epoch": 2.088410991636798,
+      "grad_norm": 0.4251172245310108,
+      "learning_rate": 3.668009883603606e-06,
+      "loss": 0.5582,
+      "step": 7867
+    },
+    {
+      "epoch": 2.0886764901101817,
+      "grad_norm": 0.44418341786848337,
+      "learning_rate": 3.667701190916134e-06,
+      "loss": 0.5756,
+      "step": 7868
+    },
+    {
+      "epoch": 2.088941988583566,
+      "grad_norm": 0.42879137804402984,
+      "learning_rate": 3.6673924754556293e-06,
+      "loss": 0.5742,
+      "step": 7869
+    },
+    {
+      "epoch": 2.0892074870569495,
+      "grad_norm": 0.4295736665423941,
+      "learning_rate": 3.667083737228114e-06,
+      "loss": 0.5691,
+      "step": 7870
+    },
+    {
+      "epoch": 2.089472985530333,
+      "grad_norm": 0.43522413336830995,
+      "learning_rate": 3.6667749762396075e-06,
+      "loss": 0.5852,
+      "step": 7871
+    },
+    {
+      "epoch": 2.0897384840037168,
+      "grad_norm": 0.4298377639649642,
+      "learning_rate": 3.666466192496133e-06,
+      "loss": 0.5524,
+      "step": 7872
+    },
+    {
+      "epoch": 2.090003982477101,
+      "grad_norm": 0.4354748938119789,
+      "learning_rate": 3.6661573860037112e-06,
+      "loss": 0.5595,
+      "step": 7873
+    },
+    {
+      "epoch": 2.0902694809504845,
+      "grad_norm": 0.4353282255972668,
+      "learning_rate": 3.6658485567683657e-06,
+      "loss": 0.6068,
+      "step": 7874
+    },
+    {
+      "epoch": 2.090534979423868,
+      "grad_norm": 0.43772776649650164,
+      "learning_rate": 3.665539704796119e-06,
+      "loss": 0.5852,
+      "step": 7875
+    },
+    {
+      "epoch": 2.090800477897252,
+      "grad_norm": 0.4304313700711266,
+      "learning_rate": 3.6652308300929935e-06,
+      "loss": 0.5547,
+      "step": 7876
+    },
+    {
+      "epoch": 2.091065976370636,
+      "grad_norm": 0.41683603307771433,
+      "learning_rate": 3.664921932665014e-06,
+      "loss": 0.5892,
+      "step": 7877
+    },
+    {
+      "epoch": 2.0913314748440195,
+      "grad_norm": 0.41729247567345373,
+      "learning_rate": 3.664613012518205e-06,
+      "loss": 0.5591,
+      "step": 7878
+    },
+    {
+      "epoch": 2.0915969733174036,
+      "grad_norm": 0.42378015548893694,
+      "learning_rate": 3.6643040696585898e-06,
+      "loss": 0.584,
+      "step": 7879
+    },
+    {
+      "epoch": 2.091862471790787,
+      "grad_norm": 0.4197127999612347,
+      "learning_rate": 3.663995104092195e-06,
+      "loss": 0.5585,
+      "step": 7880
+    },
+    {
+      "epoch": 2.092127970264171,
+      "grad_norm": 0.42554404644342747,
+      "learning_rate": 3.663686115825047e-06,
+      "loss": 0.5608,
+      "step": 7881
+    },
+    {
+      "epoch": 2.092393468737555,
+      "grad_norm": 0.425752726289791,
+      "learning_rate": 3.663377104863169e-06,
+      "loss": 0.5621,
+      "step": 7882
+    },
+    {
+      "epoch": 2.0926589672109386,
+      "grad_norm": 0.42509746453236497,
+      "learning_rate": 3.663068071212589e-06,
+      "loss": 0.5937,
+      "step": 7883
+    },
+    {
+      "epoch": 2.092924465684322,
+      "grad_norm": 0.4357672106439374,
+      "learning_rate": 3.662759014879335e-06,
+      "loss": 0.5692,
+      "step": 7884
+    },
+    {
+      "epoch": 2.0931899641577063,
+      "grad_norm": 0.448253511697435,
+      "learning_rate": 3.662449935869433e-06,
+      "loss": 0.5504,
+      "step": 7885
+    },
+    {
+      "epoch": 2.09345546263109,
+      "grad_norm": 0.43049396137508955,
+      "learning_rate": 3.662140834188911e-06,
+      "loss": 0.56,
+      "step": 7886
+    },
+    {
+      "epoch": 2.0937209611044736,
+      "grad_norm": 0.431989678115792,
+      "learning_rate": 3.6618317098437972e-06,
+      "loss": 0.6172,
+      "step": 7887
+    },
+    {
+      "epoch": 2.0939864595778572,
+      "grad_norm": 0.4206163013659491,
+      "learning_rate": 3.6615225628401202e-06,
+      "loss": 0.5342,
+      "step": 7888
+    },
+    {
+      "epoch": 2.0942519580512413,
+      "grad_norm": 0.4249195268651649,
+      "learning_rate": 3.661213393183909e-06,
+      "loss": 0.572,
+      "step": 7889
+    },
+    {
+      "epoch": 2.094517456524625,
+      "grad_norm": 0.43142726229964284,
+      "learning_rate": 3.6609042008811946e-06,
+      "loss": 0.6235,
+      "step": 7890
+    },
+    {
+      "epoch": 2.0947829549980086,
+      "grad_norm": 0.4209923733851621,
+      "learning_rate": 3.660594985938005e-06,
+      "loss": 0.5984,
+      "step": 7891
+    },
+    {
+      "epoch": 2.0950484534713927,
+      "grad_norm": 0.4089955513512835,
+      "learning_rate": 3.6602857483603713e-06,
+      "loss": 0.5504,
+      "step": 7892
+    },
+    {
+      "epoch": 2.0953139519447763,
+      "grad_norm": 0.4153052292989263,
+      "learning_rate": 3.659976488154326e-06,
+      "loss": 0.5446,
+      "step": 7893
+    },
+    {
+      "epoch": 2.09557945041816,
+      "grad_norm": 0.42494402334825065,
+      "learning_rate": 3.659667205325898e-06,
+      "loss": 0.6031,
+      "step": 7894
+    },
+    {
+      "epoch": 2.095844948891544,
+      "grad_norm": 0.406439501645814,
+      "learning_rate": 3.659357899881121e-06,
+      "loss": 0.527,
+      "step": 7895
+    },
+    {
+      "epoch": 2.0961104473649277,
+      "grad_norm": 0.42490501779063927,
+      "learning_rate": 3.6590485718260254e-06,
+      "loss": 0.5875,
+      "step": 7896
+    },
+    {
+      "epoch": 2.0963759458383113,
+      "grad_norm": 0.4256293857511059,
+      "learning_rate": 3.658739221166645e-06,
+      "loss": 0.5102,
+      "step": 7897
+    },
+    {
+      "epoch": 2.0966414443116954,
+      "grad_norm": 0.4174293269870777,
+      "learning_rate": 3.658429847909013e-06,
+      "loss": 0.5687,
+      "step": 7898
+    },
+    {
+      "epoch": 2.096906942785079,
+      "grad_norm": 0.42167866535279896,
+      "learning_rate": 3.6581204520591625e-06,
+      "loss": 0.5511,
+      "step": 7899
+    },
+    {
+      "epoch": 2.0971724412584627,
+      "grad_norm": 0.4173528229810813,
+      "learning_rate": 3.657811033623127e-06,
+      "loss": 0.5484,
+      "step": 7900
+    },
+    {
+      "epoch": 2.0974379397318463,
+      "grad_norm": 0.4248313656996904,
+      "learning_rate": 3.6575015926069423e-06,
+      "loss": 0.5458,
+      "step": 7901
+    },
+    {
+      "epoch": 2.0977034382052304,
+      "grad_norm": 0.4293679515800337,
+      "learning_rate": 3.6571921290166425e-06,
+      "loss": 0.5815,
+      "step": 7902
+    },
+    {
+      "epoch": 2.097968936678614,
+      "grad_norm": 0.4355740075608646,
+      "learning_rate": 3.6568826428582625e-06,
+      "loss": 0.5889,
+      "step": 7903
+    },
+    {
+      "epoch": 2.0982344351519977,
+      "grad_norm": 0.4243226525417147,
+      "learning_rate": 3.656573134137839e-06,
+      "loss": 0.5705,
+      "step": 7904
+    },
+    {
+      "epoch": 2.098499933625382,
+      "grad_norm": 0.4317945401813962,
+      "learning_rate": 3.6562636028614075e-06,
+      "loss": 0.5721,
+      "step": 7905
+    },
+    {
+      "epoch": 2.0987654320987654,
+      "grad_norm": 0.42922172303927425,
+      "learning_rate": 3.6559540490350044e-06,
+      "loss": 0.559,
+      "step": 7906
+    },
+    {
+      "epoch": 2.099030930572149,
+      "grad_norm": 0.4315668031094222,
+      "learning_rate": 3.655644472664667e-06,
+      "loss": 0.5646,
+      "step": 7907
+    },
+    {
+      "epoch": 2.099296429045533,
+      "grad_norm": 0.4244453093877998,
+      "learning_rate": 3.6553348737564328e-06,
+      "loss": 0.5725,
+      "step": 7908
+    },
+    {
+      "epoch": 2.099561927518917,
+      "grad_norm": 0.41901433325931725,
+      "learning_rate": 3.65502525231634e-06,
+      "loss": 0.5548,
+      "step": 7909
+    },
+    {
+      "epoch": 2.0998274259923004,
+      "grad_norm": 0.4280858028951082,
+      "learning_rate": 3.654715608350427e-06,
+      "loss": 0.5952,
+      "step": 7910
+    },
+    {
+      "epoch": 2.1000929244656845,
+      "grad_norm": 0.41480282933417045,
+      "learning_rate": 3.654405941864732e-06,
+      "loss": 0.5894,
+      "step": 7911
+    },
+    {
+      "epoch": 2.100358422939068,
+      "grad_norm": 0.4288332772602598,
+      "learning_rate": 3.6540962528652952e-06,
+      "loss": 0.5828,
+      "step": 7912
+    },
+    {
+      "epoch": 2.100623921412452,
+      "grad_norm": 0.4151636991315554,
+      "learning_rate": 3.653786541358156e-06,
+      "loss": 0.5817,
+      "step": 7913
+    },
+    {
+      "epoch": 2.100889419885836,
+      "grad_norm": 0.43319164793869924,
+      "learning_rate": 3.6534768073493533e-06,
+      "loss": 0.5824,
+      "step": 7914
+    },
+    {
+      "epoch": 2.1011549183592195,
+      "grad_norm": 0.4258782225111984,
+      "learning_rate": 3.653167050844929e-06,
+      "loss": 0.5446,
+      "step": 7915
+    },
+    {
+      "epoch": 2.101420416832603,
+      "grad_norm": 0.4300869467993502,
+      "learning_rate": 3.6528572718509237e-06,
+      "loss": 0.5743,
+      "step": 7916
+    },
+    {
+      "epoch": 2.101685915305987,
+      "grad_norm": 0.45577949026192444,
+      "learning_rate": 3.6525474703733787e-06,
+      "loss": 0.575,
+      "step": 7917
+    },
+    {
+      "epoch": 2.101951413779371,
+      "grad_norm": 0.43087875667575537,
+      "learning_rate": 3.6522376464183372e-06,
+      "loss": 0.5694,
+      "step": 7918
+    },
+    {
+      "epoch": 2.1022169122527545,
+      "grad_norm": 0.41371958916682094,
+      "learning_rate": 3.6519277999918394e-06,
+      "loss": 0.553,
+      "step": 7919
+    },
+    {
+      "epoch": 2.102482410726138,
+      "grad_norm": 0.4267650460591571,
+      "learning_rate": 3.6516179310999294e-06,
+      "loss": 0.5441,
+      "step": 7920
+    },
+    {
+      "epoch": 2.1027479091995223,
+      "grad_norm": 0.4278498949530199,
+      "learning_rate": 3.6513080397486506e-06,
+      "loss": 0.5389,
+      "step": 7921
+    },
+    {
+      "epoch": 2.103013407672906,
+      "grad_norm": 0.4210942039430507,
+      "learning_rate": 3.650998125944045e-06,
+      "loss": 0.535,
+      "step": 7922
+    },
+    {
+      "epoch": 2.1032789061462895,
+      "grad_norm": 0.42434874349698026,
+      "learning_rate": 3.6506881896921588e-06,
+      "loss": 0.5483,
+      "step": 7923
+    },
+    {
+      "epoch": 2.1035444046196736,
+      "grad_norm": 0.4087339471219136,
+      "learning_rate": 3.6503782309990354e-06,
+      "loss": 0.5322,
+      "step": 7924
+    },
+    {
+      "epoch": 2.1038099030930573,
+      "grad_norm": 0.422093877810464,
+      "learning_rate": 3.6500682498707197e-06,
+      "loss": 0.5711,
+      "step": 7925
+    },
+    {
+      "epoch": 2.104075401566441,
+      "grad_norm": 0.4198895471968135,
+      "learning_rate": 3.649758246313257e-06,
+      "loss": 0.5393,
+      "step": 7926
+    },
+    {
+      "epoch": 2.1043409000398245,
+      "grad_norm": 0.41453907000977114,
+      "learning_rate": 3.6494482203326944e-06,
+      "loss": 0.5553,
+      "step": 7927
+    },
+    {
+      "epoch": 2.1046063985132086,
+      "grad_norm": 0.4178900416487073,
+      "learning_rate": 3.649138171935076e-06,
+      "loss": 0.568,
+      "step": 7928
+    },
+    {
+      "epoch": 2.1048718969865923,
+      "grad_norm": 0.415799286109944,
+      "learning_rate": 3.6488281011264496e-06,
+      "loss": 0.5657,
+      "step": 7929
+    },
+    {
+      "epoch": 2.105137395459976,
+      "grad_norm": 0.42752419460187796,
+      "learning_rate": 3.6485180079128636e-06,
+      "loss": 0.5697,
+      "step": 7930
+    },
+    {
+      "epoch": 2.10540289393336,
+      "grad_norm": 0.42708227175139585,
+      "learning_rate": 3.6482078923003633e-06,
+      "loss": 0.549,
+      "step": 7931
+    },
+    {
+      "epoch": 2.1056683924067436,
+      "grad_norm": 0.44545415765637775,
+      "learning_rate": 3.6478977542949982e-06,
+      "loss": 0.544,
+      "step": 7932
+    },
+    {
+      "epoch": 2.1059338908801273,
+      "grad_norm": 0.431567309040952,
+      "learning_rate": 3.6475875939028166e-06,
+      "loss": 0.5629,
+      "step": 7933
+    },
+    {
+      "epoch": 2.1061993893535114,
+      "grad_norm": 0.43517500341419263,
+      "learning_rate": 3.6472774111298668e-06,
+      "loss": 0.5719,
+      "step": 7934
+    },
+    {
+      "epoch": 2.106464887826895,
+      "grad_norm": 0.43137560609523234,
+      "learning_rate": 3.646967205982198e-06,
+      "loss": 0.5716,
+      "step": 7935
+    },
+    {
+      "epoch": 2.1067303863002786,
+      "grad_norm": 0.4190466606296365,
+      "learning_rate": 3.646656978465861e-06,
+      "loss": 0.5606,
+      "step": 7936
+    },
+    {
+      "epoch": 2.1069958847736627,
+      "grad_norm": 0.43430164920882863,
+      "learning_rate": 3.646346728586906e-06,
+      "loss": 0.5418,
+      "step": 7937
+    },
+    {
+      "epoch": 2.1072613832470464,
+      "grad_norm": 0.4279186469353687,
+      "learning_rate": 3.6460364563513823e-06,
+      "loss": 0.5984,
+      "step": 7938
+    },
+    {
+      "epoch": 2.10752688172043,
+      "grad_norm": 0.4231647221027598,
+      "learning_rate": 3.645726161765342e-06,
+      "loss": 0.5735,
+      "step": 7939
+    },
+    {
+      "epoch": 2.107792380193814,
+      "grad_norm": 0.41170892065694004,
+      "learning_rate": 3.6454158448348363e-06,
+      "loss": 0.5453,
+      "step": 7940
+    },
+    {
+      "epoch": 2.1080578786671977,
+      "grad_norm": 0.44205172045602886,
+      "learning_rate": 3.6451055055659167e-06,
+      "loss": 0.5985,
+      "step": 7941
+    },
+    {
+      "epoch": 2.1083233771405814,
+      "grad_norm": 0.4211496955186089,
+      "learning_rate": 3.6447951439646368e-06,
+      "loss": 0.5833,
+      "step": 7942
+    },
+    {
+      "epoch": 2.108588875613965,
+      "grad_norm": 0.4298877651793863,
+      "learning_rate": 3.644484760037048e-06,
+      "loss": 0.5569,
+      "step": 7943
+    },
+    {
+      "epoch": 2.108854374087349,
+      "grad_norm": 0.43779770166616805,
+      "learning_rate": 3.6441743537892045e-06,
+      "loss": 0.6026,
+      "step": 7944
+    },
+    {
+      "epoch": 2.1091198725607327,
+      "grad_norm": 0.40915734122466724,
+      "learning_rate": 3.6438639252271597e-06,
+      "loss": 0.5454,
+      "step": 7945
+    },
+    {
+      "epoch": 2.1093853710341164,
+      "grad_norm": 0.42734547561812297,
+      "learning_rate": 3.6435534743569674e-06,
+      "loss": 0.5699,
+      "step": 7946
+    },
+    {
+      "epoch": 2.1096508695075005,
+      "grad_norm": 0.43949928742694055,
+      "learning_rate": 3.643243001184683e-06,
+      "loss": 0.5908,
+      "step": 7947
+    },
+    {
+      "epoch": 2.109916367980884,
+      "grad_norm": 0.42244903702406783,
+      "learning_rate": 3.6429325057163607e-06,
+      "loss": 0.5814,
+      "step": 7948
+    },
+    {
+      "epoch": 2.1101818664542678,
+      "grad_norm": 0.42671917355588357,
+      "learning_rate": 3.642621987958056e-06,
+      "loss": 0.5301,
+      "step": 7949
+    },
+    {
+      "epoch": 2.110447364927652,
+      "grad_norm": 0.4132984600526833,
+      "learning_rate": 3.642311447915825e-06,
+      "loss": 0.535,
+      "step": 7950
+    },
+    {
+      "epoch": 2.1107128634010355,
+      "grad_norm": 0.4375408792660125,
+      "learning_rate": 3.642000885595724e-06,
+      "loss": 0.5494,
+      "step": 7951
+    },
+    {
+      "epoch": 2.110978361874419,
+      "grad_norm": 0.42541073135700963,
+      "learning_rate": 3.6416903010038096e-06,
+      "loss": 0.5951,
+      "step": 7952
+    },
+    {
+      "epoch": 2.111243860347803,
+      "grad_norm": 0.4345740682350128,
+      "learning_rate": 3.641379694146139e-06,
+      "loss": 0.5559,
+      "step": 7953
+    },
+    {
+      "epoch": 2.111509358821187,
+      "grad_norm": 0.4186693681895498,
+      "learning_rate": 3.6410690650287696e-06,
+      "loss": 0.5488,
+      "step": 7954
+    },
+    {
+      "epoch": 2.1117748572945705,
+      "grad_norm": 0.41633769862776476,
+      "learning_rate": 3.640758413657759e-06,
+      "loss": 0.5742,
+      "step": 7955
+    },
+    {
+      "epoch": 2.112040355767954,
+      "grad_norm": 0.4243956064162778,
+      "learning_rate": 3.6404477400391675e-06,
+      "loss": 0.5534,
+      "step": 7956
+    },
+    {
+      "epoch": 2.112305854241338,
+      "grad_norm": 0.4266805913086236,
+      "learning_rate": 3.640137044179052e-06,
+      "loss": 0.5407,
+      "step": 7957
+    },
+    {
+      "epoch": 2.112571352714722,
+      "grad_norm": 0.4330549608013007,
+      "learning_rate": 3.6398263260834733e-06,
+      "loss": 0.5817,
+      "step": 7958
+    },
+    {
+      "epoch": 2.1128368511881055,
+      "grad_norm": 0.4100806467189942,
+      "learning_rate": 3.6395155857584898e-06,
+      "loss": 0.5624,
+      "step": 7959
+    },
+    {
+      "epoch": 2.1131023496614896,
+      "grad_norm": 0.43435924398440234,
+      "learning_rate": 3.639204823210163e-06,
+      "loss": 0.581,
+      "step": 7960
+    },
+    {
+      "epoch": 2.113367848134873,
+      "grad_norm": 0.4329545894381265,
+      "learning_rate": 3.638894038444552e-06,
+      "loss": 0.5702,
+      "step": 7961
+    },
+    {
+      "epoch": 2.113633346608257,
+      "grad_norm": 0.45340530202681095,
+      "learning_rate": 3.6385832314677193e-06,
+      "loss": 0.5568,
+      "step": 7962
+    },
+    {
+      "epoch": 2.113898845081641,
+      "grad_norm": 0.42964666595175455,
+      "learning_rate": 3.6382724022857254e-06,
+      "loss": 0.6124,
+      "step": 7963
+    },
+    {
+      "epoch": 2.1141643435550246,
+      "grad_norm": 0.419428067900188,
+      "learning_rate": 3.6379615509046327e-06,
+      "loss": 0.5548,
+      "step": 7964
+    },
+    {
+      "epoch": 2.1144298420284082,
+      "grad_norm": 0.4516037056068796,
+      "learning_rate": 3.6376506773305047e-06,
+      "loss": 0.5831,
+      "step": 7965
+    },
+    {
+      "epoch": 2.1146953405017923,
+      "grad_norm": 0.427759583741649,
+      "learning_rate": 3.6373397815694023e-06,
+      "loss": 0.561,
+      "step": 7966
+    },
+    {
+      "epoch": 2.114960838975176,
+      "grad_norm": 0.42126228929129966,
+      "learning_rate": 3.63702886362739e-06,
+      "loss": 0.5392,
+      "step": 7967
+    },
+    {
+      "epoch": 2.1152263374485596,
+      "grad_norm": 0.42500561873963555,
+      "learning_rate": 3.63671792351053e-06,
+      "loss": 0.5771,
+      "step": 7968
+    },
+    {
+      "epoch": 2.1154918359219437,
+      "grad_norm": 0.4279648708519892,
+      "learning_rate": 3.6364069612248874e-06,
+      "loss": 0.5985,
+      "step": 7969
+    },
+    {
+      "epoch": 2.1157573343953273,
+      "grad_norm": 0.4258078527664367,
+      "learning_rate": 3.636095976776527e-06,
+      "loss": 0.5845,
+      "step": 7970
+    },
+    {
+      "epoch": 2.116022832868711,
+      "grad_norm": 0.4229238300489938,
+      "learning_rate": 3.635784970171513e-06,
+      "loss": 0.5392,
+      "step": 7971
+    },
+    {
+      "epoch": 2.1162883313420946,
+      "grad_norm": 0.42217674975822295,
+      "learning_rate": 3.6354739414159114e-06,
+      "loss": 0.5678,
+      "step": 7972
+    },
+    {
+      "epoch": 2.1165538298154787,
+      "grad_norm": 0.4244005319842895,
+      "learning_rate": 3.6351628905157886e-06,
+      "loss": 0.5554,
+      "step": 7973
+    },
+    {
+      "epoch": 2.1168193282888623,
+      "grad_norm": 0.4328728459958939,
+      "learning_rate": 3.634851817477209e-06,
+      "loss": 0.5587,
+      "step": 7974
+    },
+    {
+      "epoch": 2.117084826762246,
+      "grad_norm": 0.42755433048338387,
+      "learning_rate": 3.634540722306241e-06,
+      "loss": 0.5676,
+      "step": 7975
+    },
+    {
+      "epoch": 2.11735032523563,
+      "grad_norm": 0.41638394714778576,
+      "learning_rate": 3.634229605008951e-06,
+      "loss": 0.5536,
+      "step": 7976
+    },
+    {
+      "epoch": 2.1176158237090137,
+      "grad_norm": 0.4244115106752363,
+      "learning_rate": 3.633918465591407e-06,
+      "loss": 0.5454,
+      "step": 7977
+    },
+    {
+      "epoch": 2.1178813221823973,
+      "grad_norm": 0.43650908093444724,
+      "learning_rate": 3.6336073040596752e-06,
+      "loss": 0.5715,
+      "step": 7978
+    },
+    {
+      "epoch": 2.1181468206557814,
+      "grad_norm": 0.4223321960262128,
+      "learning_rate": 3.633296120419827e-06,
+      "loss": 0.5519,
+      "step": 7979
+    },
+    {
+      "epoch": 2.118412319129165,
+      "grad_norm": 0.43854703367933506,
+      "learning_rate": 3.632984914677929e-06,
+      "loss": 0.5834,
+      "step": 7980
+    },
+    {
+      "epoch": 2.1186778176025487,
+      "grad_norm": 0.4151364291354364,
+      "learning_rate": 3.632673686840051e-06,
+      "loss": 0.5493,
+      "step": 7981
+    },
+    {
+      "epoch": 2.1189433160759323,
+      "grad_norm": 0.441090733203994,
+      "learning_rate": 3.632362436912263e-06,
+      "loss": 0.5394,
+      "step": 7982
+    },
+    {
+      "epoch": 2.1192088145493164,
+      "grad_norm": 0.4536107233261637,
+      "learning_rate": 3.632051164900635e-06,
+      "loss": 0.5274,
+      "step": 7983
+    },
+    {
+      "epoch": 2.1194743130227,
+      "grad_norm": 0.4200671092614244,
+      "learning_rate": 3.6317398708112372e-06,
+      "loss": 0.5733,
+      "step": 7984
+    },
+    {
+      "epoch": 2.1197398114960837,
+      "grad_norm": 0.4277111804410303,
+      "learning_rate": 3.6314285546501415e-06,
+      "loss": 0.569,
+      "step": 7985
+    },
+    {
+      "epoch": 2.120005309969468,
+      "grad_norm": 0.4328333608833753,
+      "learning_rate": 3.6311172164234187e-06,
+      "loss": 0.5748,
+      "step": 7986
+    },
+    {
+      "epoch": 2.1202708084428514,
+      "grad_norm": 0.4243852529436048,
+      "learning_rate": 3.6308058561371407e-06,
+      "loss": 0.5801,
+      "step": 7987
+    },
+    {
+      "epoch": 2.120536306916235,
+      "grad_norm": 0.4265879943858205,
+      "learning_rate": 3.6304944737973794e-06,
+      "loss": 0.571,
+      "step": 7988
+    },
+    {
+      "epoch": 2.120801805389619,
+      "grad_norm": 0.4181442163237821,
+      "learning_rate": 3.6301830694102086e-06,
+      "loss": 0.5641,
+      "step": 7989
+    },
+    {
+      "epoch": 2.121067303863003,
+      "grad_norm": 0.42227138240633094,
+      "learning_rate": 3.6298716429817006e-06,
+      "loss": 0.5782,
+      "step": 7990
+    },
+    {
+      "epoch": 2.1213328023363864,
+      "grad_norm": 0.4260358429920136,
+      "learning_rate": 3.6295601945179282e-06,
+      "loss": 0.5711,
+      "step": 7991
+    },
+    {
+      "epoch": 2.1215983008097705,
+      "grad_norm": 0.43433164184629003,
+      "learning_rate": 3.6292487240249676e-06,
+      "loss": 0.5369,
+      "step": 7992
+    },
+    {
+      "epoch": 2.121863799283154,
+      "grad_norm": 0.4268611538270948,
+      "learning_rate": 3.6289372315088916e-06,
+      "loss": 0.5665,
+      "step": 7993
+    },
+    {
+      "epoch": 2.122129297756538,
+      "grad_norm": 0.4168088351296817,
+      "learning_rate": 3.6286257169757756e-06,
+      "loss": 0.5574,
+      "step": 7994
+    },
+    {
+      "epoch": 2.122394796229922,
+      "grad_norm": 0.4246541435935857,
+      "learning_rate": 3.628314180431694e-06,
+      "loss": 0.5463,
+      "step": 7995
+    },
+    {
+      "epoch": 2.1226602947033055,
+      "grad_norm": 0.4319435219741565,
+      "learning_rate": 3.6280026218827247e-06,
+      "loss": 0.5888,
+      "step": 7996
+    },
+    {
+      "epoch": 2.122925793176689,
+      "grad_norm": 0.43502368816550957,
+      "learning_rate": 3.627691041334941e-06,
+      "loss": 0.5462,
+      "step": 7997
+    },
+    {
+      "epoch": 2.123191291650073,
+      "grad_norm": 0.43009773525111383,
+      "learning_rate": 3.6273794387944216e-06,
+      "loss": 0.5422,
+      "step": 7998
+    },
+    {
+      "epoch": 2.123456790123457,
+      "grad_norm": 0.441744401430134,
+      "learning_rate": 3.6270678142672426e-06,
+      "loss": 0.6043,
+      "step": 7999
+    },
+    {
+      "epoch": 2.1237222885968405,
+      "grad_norm": 0.4414479905702202,
+      "learning_rate": 3.626756167759482e-06,
+      "loss": 0.5723,
+      "step": 8000
+    },
+    {
+      "epoch": 2.123987787070224,
+      "grad_norm": 0.43047843867502195,
+      "learning_rate": 3.626444499277217e-06,
+      "loss": 0.5541,
+      "step": 8001
+    },
+    {
+      "epoch": 2.1242532855436083,
+      "grad_norm": 0.45411837737981864,
+      "learning_rate": 3.626132808826527e-06,
+      "loss": 0.5222,
+      "step": 8002
+    },
+    {
+      "epoch": 2.124518784016992,
+      "grad_norm": 0.43146575662073927,
+      "learning_rate": 3.625821096413489e-06,
+      "loss": 0.5749,
+      "step": 8003
+    },
+    {
+      "epoch": 2.1247842824903755,
+      "grad_norm": 0.4304879493679724,
+      "learning_rate": 3.6255093620441835e-06,
+      "loss": 0.5564,
+      "step": 8004
+    },
+    {
+      "epoch": 2.1250497809637596,
+      "grad_norm": 0.4293964986404451,
+      "learning_rate": 3.62519760572469e-06,
+      "loss": 0.5852,
+      "step": 8005
+    },
+    {
+      "epoch": 2.1253152794371433,
+      "grad_norm": 0.4348393127234673,
+      "learning_rate": 3.624885827461088e-06,
+      "loss": 0.5763,
+      "step": 8006
+    },
+    {
+      "epoch": 2.125580777910527,
+      "grad_norm": 0.42904264331724723,
+      "learning_rate": 3.6245740272594584e-06,
+      "loss": 0.5952,
+      "step": 8007
+    },
+    {
+      "epoch": 2.125846276383911,
+      "grad_norm": 0.42892641580927876,
+      "learning_rate": 3.624262205125881e-06,
+      "loss": 0.5453,
+      "step": 8008
+    },
+    {
+      "epoch": 2.1261117748572946,
+      "grad_norm": 0.4337721649196399,
+      "learning_rate": 3.623950361066439e-06,
+      "loss": 0.5639,
+      "step": 8009
+    },
+    {
+      "epoch": 2.1263772733306783,
+      "grad_norm": 0.4379513763977376,
+      "learning_rate": 3.623638495087212e-06,
+      "loss": 0.5963,
+      "step": 8010
+    },
+    {
+      "epoch": 2.126642771804062,
+      "grad_norm": 0.4283183752062311,
+      "learning_rate": 3.6233266071942845e-06,
+      "loss": 0.5912,
+      "step": 8011
+    },
+    {
+      "epoch": 2.126908270277446,
+      "grad_norm": 0.42475082307663986,
+      "learning_rate": 3.623014697393737e-06,
+      "loss": 0.5317,
+      "step": 8012
+    },
+    {
+      "epoch": 2.1271737687508296,
+      "grad_norm": 0.4312996694785182,
+      "learning_rate": 3.622702765691653e-06,
+      "loss": 0.5558,
+      "step": 8013
+    },
+    {
+      "epoch": 2.1274392672242133,
+      "grad_norm": 0.4429595843632247,
+      "learning_rate": 3.622390812094117e-06,
+      "loss": 0.5431,
+      "step": 8014
+    },
+    {
+      "epoch": 2.1277047656975974,
+      "grad_norm": 0.418639786992392,
+      "learning_rate": 3.622078836607212e-06,
+      "loss": 0.5398,
+      "step": 8015
+    },
+    {
+      "epoch": 2.127970264170981,
+      "grad_norm": 0.43397650063216997,
+      "learning_rate": 3.621766839237022e-06,
+      "loss": 0.568,
+      "step": 8016
+    },
+    {
+      "epoch": 2.1282357626443646,
+      "grad_norm": 0.4283674590653182,
+      "learning_rate": 3.6214548199896315e-06,
+      "loss": 0.5535,
+      "step": 8017
+    },
+    {
+      "epoch": 2.1285012611177487,
+      "grad_norm": 0.4208320807455846,
+      "learning_rate": 3.621142778871127e-06,
+      "loss": 0.5756,
+      "step": 8018
+    },
+    {
+      "epoch": 2.1287667595911324,
+      "grad_norm": 0.4120720051953142,
+      "learning_rate": 3.6208307158875934e-06,
+      "loss": 0.5948,
+      "step": 8019
+    },
+    {
+      "epoch": 2.129032258064516,
+      "grad_norm": 0.44451895183248785,
+      "learning_rate": 3.620518631045116e-06,
+      "loss": 0.5772,
+      "step": 8020
+    },
+    {
+      "epoch": 2.1292977565379,
+      "grad_norm": 0.4189982601159866,
+      "learning_rate": 3.620206524349782e-06,
+      "loss": 0.5257,
+      "step": 8021
+    },
+    {
+      "epoch": 2.1295632550112837,
+      "grad_norm": 0.4153254071728542,
+      "learning_rate": 3.619894395807679e-06,
+      "loss": 0.5639,
+      "step": 8022
+    },
+    {
+      "epoch": 2.1298287534846674,
+      "grad_norm": 0.41547191334593847,
+      "learning_rate": 3.619582245424892e-06,
+      "loss": 0.5525,
+      "step": 8023
+    },
+    {
+      "epoch": 2.1300942519580515,
+      "grad_norm": 0.42484213287825123,
+      "learning_rate": 3.6192700732075103e-06,
+      "loss": 0.5077,
+      "step": 8024
+    },
+    {
+      "epoch": 2.130359750431435,
+      "grad_norm": 0.43510874529626953,
+      "learning_rate": 3.618957879161622e-06,
+      "loss": 0.5738,
+      "step": 8025
+    },
+    {
+      "epoch": 2.1306252489048187,
+      "grad_norm": 0.4335061024685529,
+      "learning_rate": 3.618645663293315e-06,
+      "loss": 0.6059,
+      "step": 8026
+    },
+    {
+      "epoch": 2.1308907473782024,
+      "grad_norm": 0.4303836677947689,
+      "learning_rate": 3.618333425608679e-06,
+      "loss": 0.6138,
+      "step": 8027
+    },
+    {
+      "epoch": 2.1311562458515865,
+      "grad_norm": 0.42186408094798866,
+      "learning_rate": 3.6180211661138034e-06,
+      "loss": 0.5721,
+      "step": 8028
+    },
+    {
+      "epoch": 2.13142174432497,
+      "grad_norm": 0.43452533559782536,
+      "learning_rate": 3.617708884814777e-06,
+      "loss": 0.5752,
+      "step": 8029
+    },
+    {
+      "epoch": 2.1316872427983538,
+      "grad_norm": 0.4411459749259788,
+      "learning_rate": 3.6173965817176902e-06,
+      "loss": 0.5692,
+      "step": 8030
+    },
+    {
+      "epoch": 2.131952741271738,
+      "grad_norm": 0.42028956939591017,
+      "learning_rate": 3.617084256828636e-06,
+      "loss": 0.5617,
+      "step": 8031
+    },
+    {
+      "epoch": 2.1322182397451215,
+      "grad_norm": 0.427974439534533,
+      "learning_rate": 3.6167719101537024e-06,
+      "loss": 0.5914,
+      "step": 8032
+    },
+    {
+      "epoch": 2.132483738218505,
+      "grad_norm": 0.4226976099060145,
+      "learning_rate": 3.6164595416989823e-06,
+      "loss": 0.5614,
+      "step": 8033
+    },
+    {
+      "epoch": 2.132749236691889,
+      "grad_norm": 0.4340575600979119,
+      "learning_rate": 3.6161471514705683e-06,
+      "loss": 0.5873,
+      "step": 8034
+    },
+    {
+      "epoch": 2.133014735165273,
+      "grad_norm": 0.44571354721405104,
+      "learning_rate": 3.6158347394745507e-06,
+      "loss": 0.5968,
+      "step": 8035
+    },
+    {
+      "epoch": 2.1332802336386565,
+      "grad_norm": 0.4340542730915543,
+      "learning_rate": 3.6155223057170247e-06,
+      "loss": 0.5347,
+      "step": 8036
+    },
+    {
+      "epoch": 2.13354573211204,
+      "grad_norm": 0.42381527257855994,
+      "learning_rate": 3.615209850204082e-06,
+      "loss": 0.5525,
+      "step": 8037
+    },
+    {
+      "epoch": 2.133811230585424,
+      "grad_norm": 0.41780539211667134,
+      "learning_rate": 3.614897372941817e-06,
+      "loss": 0.5564,
+      "step": 8038
+    },
+    {
+      "epoch": 2.134076729058808,
+      "grad_norm": 0.42993891244135257,
+      "learning_rate": 3.614584873936323e-06,
+      "loss": 0.5646,
+      "step": 8039
+    },
+    {
+      "epoch": 2.1343422275321915,
+      "grad_norm": 0.43428479653921503,
+      "learning_rate": 3.614272353193696e-06,
+      "loss": 0.5477,
+      "step": 8040
+    },
+    {
+      "epoch": 2.1346077260055756,
+      "grad_norm": 0.42400207189222794,
+      "learning_rate": 3.613959810720029e-06,
+      "loss": 0.5746,
+      "step": 8041
+    },
+    {
+      "epoch": 2.134873224478959,
+      "grad_norm": 0.42962978979385796,
+      "learning_rate": 3.613647246521419e-06,
+      "loss": 0.5555,
+      "step": 8042
+    },
+    {
+      "epoch": 2.135138722952343,
+      "grad_norm": 0.4480349561269434,
+      "learning_rate": 3.6133346606039605e-06,
+      "loss": 0.5532,
+      "step": 8043
+    },
+    {
+      "epoch": 2.135404221425727,
+      "grad_norm": 0.42689626530334523,
+      "learning_rate": 3.61302205297375e-06,
+      "loss": 0.5883,
+      "step": 8044
+    },
+    {
+      "epoch": 2.1356697198991106,
+      "grad_norm": 0.41723989131049216,
+      "learning_rate": 3.6127094236368842e-06,
+      "loss": 0.4984,
+      "step": 8045
+    },
+    {
+      "epoch": 2.1359352183724942,
+      "grad_norm": 0.430163185314334,
+      "learning_rate": 3.612396772599461e-06,
+      "loss": 0.5698,
+      "step": 8046
+    },
+    {
+      "epoch": 2.1362007168458783,
+      "grad_norm": 0.4322140002268539,
+      "learning_rate": 3.6120840998675766e-06,
+      "loss": 0.5881,
+      "step": 8047
+    },
+    {
+      "epoch": 2.136466215319262,
+      "grad_norm": 0.42714408247846825,
+      "learning_rate": 3.6117714054473298e-06,
+      "loss": 0.5855,
+      "step": 8048
+    },
+    {
+      "epoch": 2.1367317137926456,
+      "grad_norm": 0.416319829667135,
+      "learning_rate": 3.611458689344818e-06,
+      "loss": 0.5893,
+      "step": 8049
+    },
+    {
+      "epoch": 2.1369972122660297,
+      "grad_norm": 0.4357965771127121,
+      "learning_rate": 3.6111459515661407e-06,
+      "loss": 0.602,
+      "step": 8050
+    },
+    {
+      "epoch": 2.1372627107394133,
+      "grad_norm": 0.424766280488201,
+      "learning_rate": 3.6108331921173966e-06,
+      "loss": 0.5722,
+      "step": 8051
+    },
+    {
+      "epoch": 2.137528209212797,
+      "grad_norm": 0.42527505481123656,
+      "learning_rate": 3.610520411004686e-06,
+      "loss": 0.5733,
+      "step": 8052
+    },
+    {
+      "epoch": 2.1377937076861806,
+      "grad_norm": 0.4208184276162359,
+      "learning_rate": 3.6102076082341075e-06,
+      "loss": 0.5832,
+      "step": 8053
+    },
+    {
+      "epoch": 2.1380592061595647,
+      "grad_norm": 0.4531490526184039,
+      "learning_rate": 3.609894783811763e-06,
+      "loss": 0.5647,
+      "step": 8054
+    },
+    {
+      "epoch": 2.1383247046329483,
+      "grad_norm": 0.43306972875768146,
+      "learning_rate": 3.6095819377437523e-06,
+      "loss": 0.6015,
+      "step": 8055
+    },
+    {
+      "epoch": 2.138590203106332,
+      "grad_norm": 0.4248125265115133,
+      "learning_rate": 3.6092690700361778e-06,
+      "loss": 0.556,
+      "step": 8056
+    },
+    {
+      "epoch": 2.138855701579716,
+      "grad_norm": 0.4370564627532615,
+      "learning_rate": 3.60895618069514e-06,
+      "loss": 0.5919,
+      "step": 8057
+    },
+    {
+      "epoch": 2.1391212000530997,
+      "grad_norm": 0.41607888538439014,
+      "learning_rate": 3.6086432697267414e-06,
+      "loss": 0.5885,
+      "step": 8058
+    },
+    {
+      "epoch": 2.1393866985264833,
+      "grad_norm": 0.42706851500603765,
+      "learning_rate": 3.6083303371370847e-06,
+      "loss": 0.5934,
+      "step": 8059
+    },
+    {
+      "epoch": 2.1396521969998674,
+      "grad_norm": 0.44096229557413885,
+      "learning_rate": 3.6080173829322734e-06,
+      "loss": 0.5688,
+      "step": 8060
+    },
+    {
+      "epoch": 2.139917695473251,
+      "grad_norm": 0.4363152436603377,
+      "learning_rate": 3.60770440711841e-06,
+      "loss": 0.5972,
+      "step": 8061
+    },
+    {
+      "epoch": 2.1401831939466347,
+      "grad_norm": 0.4277806010825455,
+      "learning_rate": 3.607391409701598e-06,
+      "loss": 0.5646,
+      "step": 8062
+    },
+    {
+      "epoch": 2.140448692420019,
+      "grad_norm": 0.408768985475413,
+      "learning_rate": 3.607078390687942e-06,
+      "loss": 0.5734,
+      "step": 8063
+    },
+    {
+      "epoch": 2.1407141908934024,
+      "grad_norm": 0.4367264131501375,
+      "learning_rate": 3.606765350083547e-06,
+      "loss": 0.5662,
+      "step": 8064
+    },
+    {
+      "epoch": 2.140979689366786,
+      "grad_norm": 0.4152403541160022,
+      "learning_rate": 3.6064522878945184e-06,
+      "loss": 0.5351,
+      "step": 8065
+    },
+    {
+      "epoch": 2.1412451878401697,
+      "grad_norm": 0.42026768087018546,
+      "learning_rate": 3.6061392041269605e-06,
+      "loss": 0.5614,
+      "step": 8066
+    },
+    {
+      "epoch": 2.141510686313554,
+      "grad_norm": 0.43103548155596705,
+      "learning_rate": 3.60582609878698e-06,
+      "loss": 0.5663,
+      "step": 8067
+    },
+    {
+      "epoch": 2.1417761847869374,
+      "grad_norm": 0.41058773194358855,
+      "learning_rate": 3.6055129718806836e-06,
+      "loss": 0.5411,
+      "step": 8068
+    },
+    {
+      "epoch": 2.142041683260321,
+      "grad_norm": 0.4326831638926904,
+      "learning_rate": 3.6051998234141768e-06,
+      "loss": 0.6307,
+      "step": 8069
+    },
+    {
+      "epoch": 2.142307181733705,
+      "grad_norm": 0.42999013981109785,
+      "learning_rate": 3.6048866533935673e-06,
+      "loss": 0.5793,
+      "step": 8070
+    },
+    {
+      "epoch": 2.142572680207089,
+      "grad_norm": 0.44126473351861395,
+      "learning_rate": 3.604573461824964e-06,
+      "loss": 0.552,
+      "step": 8071
+    },
+    {
+      "epoch": 2.1428381786804724,
+      "grad_norm": 0.4611303067588401,
+      "learning_rate": 3.6042602487144716e-06,
+      "loss": 0.5756,
+      "step": 8072
+    },
+    {
+      "epoch": 2.1431036771538565,
+      "grad_norm": 0.44271024251920693,
+      "learning_rate": 3.603947014068202e-06,
+      "loss": 0.556,
+      "step": 8073
+    },
+    {
+      "epoch": 2.14336917562724,
+      "grad_norm": 0.43259245576915406,
+      "learning_rate": 3.6036337578922633e-06,
+      "loss": 0.5881,
+      "step": 8074
+    },
+    {
+      "epoch": 2.143634674100624,
+      "grad_norm": 0.4319794228696632,
+      "learning_rate": 3.6033204801927636e-06,
+      "loss": 0.5588,
+      "step": 8075
+    },
+    {
+      "epoch": 2.143900172574008,
+      "grad_norm": 0.4323547684393841,
+      "learning_rate": 3.6030071809758127e-06,
+      "loss": 0.5472,
+      "step": 8076
+    },
+    {
+      "epoch": 2.1441656710473915,
+      "grad_norm": 0.42154147532371194,
+      "learning_rate": 3.6026938602475212e-06,
+      "loss": 0.5672,
+      "step": 8077
+    },
+    {
+      "epoch": 2.144431169520775,
+      "grad_norm": 0.43110878047796575,
+      "learning_rate": 3.6023805180140005e-06,
+      "loss": 0.5643,
+      "step": 8078
+    },
+    {
+      "epoch": 2.1446966679941593,
+      "grad_norm": 0.4430635262160183,
+      "learning_rate": 3.6020671542813596e-06,
+      "loss": 0.5922,
+      "step": 8079
+    },
+    {
+      "epoch": 2.144962166467543,
+      "grad_norm": 0.42957359014746876,
+      "learning_rate": 3.6017537690557114e-06,
+      "loss": 0.5824,
+      "step": 8080
+    },
+    {
+      "epoch": 2.1452276649409265,
+      "grad_norm": 0.42740105360366026,
+      "learning_rate": 3.601440362343167e-06,
+      "loss": 0.5611,
+      "step": 8081
+    },
+    {
+      "epoch": 2.14549316341431,
+      "grad_norm": 0.46092266723123154,
+      "learning_rate": 3.601126934149838e-06,
+      "loss": 0.5469,
+      "step": 8082
+    },
+    {
+      "epoch": 2.1457586618876943,
+      "grad_norm": 0.42044305592506603,
+      "learning_rate": 3.6008134844818386e-06,
+      "loss": 0.5478,
+      "step": 8083
+    },
+    {
+      "epoch": 2.146024160361078,
+      "grad_norm": 0.41919413812783474,
+      "learning_rate": 3.600500013345281e-06,
+      "loss": 0.5518,
+      "step": 8084
+    },
+    {
+      "epoch": 2.1462896588344615,
+      "grad_norm": 0.45848238302689487,
+      "learning_rate": 3.6001865207462785e-06,
+      "loss": 0.5655,
+      "step": 8085
+    },
+    {
+      "epoch": 2.1465551573078456,
+      "grad_norm": 0.43841601883293296,
+      "learning_rate": 3.5998730066909453e-06,
+      "loss": 0.5497,
+      "step": 8086
+    },
+    {
+      "epoch": 2.1468206557812293,
+      "grad_norm": 0.438067347636522,
+      "learning_rate": 3.5995594711853955e-06,
+      "loss": 0.5965,
+      "step": 8087
+    },
+    {
+      "epoch": 2.147086154254613,
+      "grad_norm": 0.43603604180414585,
+      "learning_rate": 3.5992459142357438e-06,
+      "loss": 0.5841,
+      "step": 8088
+    },
+    {
+      "epoch": 2.147351652727997,
+      "grad_norm": 0.4308631569429699,
+      "learning_rate": 3.5989323358481053e-06,
+      "loss": 0.5906,
+      "step": 8089
+    },
+    {
+      "epoch": 2.1476171512013806,
+      "grad_norm": 0.4829050416934733,
+      "learning_rate": 3.598618736028595e-06,
+      "loss": 0.5717,
+      "step": 8090
+    },
+    {
+      "epoch": 2.1478826496747643,
+      "grad_norm": 0.4534862302323038,
+      "learning_rate": 3.59830511478333e-06,
+      "loss": 0.5893,
+      "step": 8091
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 0.41791927793458056,
+      "learning_rate": 3.5979914721184263e-06,
+      "loss": 0.5859,
+      "step": 8092
+    },
+    {
+      "epoch": 2.148413646621532,
+      "grad_norm": 0.4537914041379908,
+      "learning_rate": 3.5976778080399997e-06,
+      "loss": 0.5884,
+      "step": 8093
+    },
+    {
+      "epoch": 2.1486791450949156,
+      "grad_norm": 0.4388072513963964,
+      "learning_rate": 3.59736412255417e-06,
+      "loss": 0.5781,
+      "step": 8094
+    },
+    {
+      "epoch": 2.1489446435682993,
+      "grad_norm": 0.431744575916571,
+      "learning_rate": 3.5970504156670516e-06,
+      "loss": 0.6,
+      "step": 8095
+    },
+    {
+      "epoch": 2.1492101420416834,
+      "grad_norm": 0.42671965013016383,
+      "learning_rate": 3.5967366873847643e-06,
+      "loss": 0.6091,
+      "step": 8096
+    },
+    {
+      "epoch": 2.149475640515067,
+      "grad_norm": 0.4281110184081167,
+      "learning_rate": 3.5964229377134276e-06,
+      "loss": 0.5553,
+      "step": 8097
+    },
+    {
+      "epoch": 2.1497411389884507,
+      "grad_norm": 0.4360037237708422,
+      "learning_rate": 3.5961091666591574e-06,
+      "loss": 0.5869,
+      "step": 8098
+    },
+    {
+      "epoch": 2.1500066374618347,
+      "grad_norm": 0.4287728608608719,
+      "learning_rate": 3.5957953742280754e-06,
+      "loss": 0.6016,
+      "step": 8099
+    },
+    {
+      "epoch": 2.1502721359352184,
+      "grad_norm": 0.4220330290905117,
+      "learning_rate": 3.5954815604263004e-06,
+      "loss": 0.553,
+      "step": 8100
+    },
+    {
+      "epoch": 2.150537634408602,
+      "grad_norm": 0.432355940946539,
+      "learning_rate": 3.5951677252599527e-06,
+      "loss": 0.5819,
+      "step": 8101
+    },
+    {
+      "epoch": 2.150803132881986,
+      "grad_norm": 0.4385362896641794,
+      "learning_rate": 3.594853868735153e-06,
+      "loss": 0.5809,
+      "step": 8102
+    },
+    {
+      "epoch": 2.1510686313553697,
+      "grad_norm": 0.43228388668039286,
+      "learning_rate": 3.5945399908580233e-06,
+      "loss": 0.5941,
+      "step": 8103
+    },
+    {
+      "epoch": 2.1513341298287534,
+      "grad_norm": 0.4224342324211673,
+      "learning_rate": 3.5942260916346828e-06,
+      "loss": 0.5754,
+      "step": 8104
+    },
+    {
+      "epoch": 2.1515996283021375,
+      "grad_norm": 0.41315324887917493,
+      "learning_rate": 3.5939121710712543e-06,
+      "loss": 0.5743,
+      "step": 8105
+    },
+    {
+      "epoch": 2.151865126775521,
+      "grad_norm": 0.4297833452425808,
+      "learning_rate": 3.5935982291738603e-06,
+      "loss": 0.5863,
+      "step": 8106
+    },
+    {
+      "epoch": 2.1521306252489047,
+      "grad_norm": 0.4358073161958455,
+      "learning_rate": 3.5932842659486235e-06,
+      "loss": 0.5354,
+      "step": 8107
+    },
+    {
+      "epoch": 2.1523961237222884,
+      "grad_norm": 0.42483957253791854,
+      "learning_rate": 3.5929702814016667e-06,
+      "loss": 0.5311,
+      "step": 8108
+    },
+    {
+      "epoch": 2.1526616221956725,
+      "grad_norm": 0.423219127056296,
+      "learning_rate": 3.5926562755391126e-06,
+      "loss": 0.5983,
+      "step": 8109
+    },
+    {
+      "epoch": 2.152927120669056,
+      "grad_norm": 0.422511706840988,
+      "learning_rate": 3.592342248367086e-06,
+      "loss": 0.5679,
+      "step": 8110
+    },
+    {
+      "epoch": 2.1531926191424398,
+      "grad_norm": 0.4377553129409497,
+      "learning_rate": 3.592028199891712e-06,
+      "loss": 0.5605,
+      "step": 8111
+    },
+    {
+      "epoch": 2.153458117615824,
+      "grad_norm": 0.43475481025721574,
+      "learning_rate": 3.5917141301191137e-06,
+      "loss": 0.6075,
+      "step": 8112
+    },
+    {
+      "epoch": 2.1537236160892075,
+      "grad_norm": 0.4237312870259854,
+      "learning_rate": 3.591400039055417e-06,
+      "loss": 0.603,
+      "step": 8113
+    },
+    {
+      "epoch": 2.153989114562591,
+      "grad_norm": 0.41950635868227365,
+      "learning_rate": 3.5910859267067476e-06,
+      "loss": 0.5634,
+      "step": 8114
+    },
+    {
+      "epoch": 2.154254613035975,
+      "grad_norm": 0.4519774747924744,
+      "learning_rate": 3.5907717930792307e-06,
+      "loss": 0.5785,
+      "step": 8115
+    },
+    {
+      "epoch": 2.154520111509359,
+      "grad_norm": 0.4291079808804259,
+      "learning_rate": 3.5904576381789933e-06,
+      "loss": 0.5923,
+      "step": 8116
+    },
+    {
+      "epoch": 2.1547856099827425,
+      "grad_norm": 0.4265704040921036,
+      "learning_rate": 3.590143462012162e-06,
+      "loss": 0.5782,
+      "step": 8117
+    },
+    {
+      "epoch": 2.1550511084561266,
+      "grad_norm": 0.4334135139626096,
+      "learning_rate": 3.589829264584864e-06,
+      "loss": 0.5823,
+      "step": 8118
+    },
+    {
+      "epoch": 2.15531660692951,
+      "grad_norm": 0.43512986410951887,
+      "learning_rate": 3.5895150459032268e-06,
+      "loss": 0.5774,
+      "step": 8119
+    },
+    {
+      "epoch": 2.155582105402894,
+      "grad_norm": 0.452573659656349,
+      "learning_rate": 3.5892008059733795e-06,
+      "loss": 0.5534,
+      "step": 8120
+    },
+    {
+      "epoch": 2.1558476038762775,
+      "grad_norm": 0.41307716634649994,
+      "learning_rate": 3.5888865448014486e-06,
+      "loss": 0.5509,
+      "step": 8121
+    },
+    {
+      "epoch": 2.1561131023496616,
+      "grad_norm": 0.44650225414869477,
+      "learning_rate": 3.5885722623935648e-06,
+      "loss": 0.5652,
+      "step": 8122
+    },
+    {
+      "epoch": 2.156378600823045,
+      "grad_norm": 0.4259864728899922,
+      "learning_rate": 3.5882579587558564e-06,
+      "loss": 0.5575,
+      "step": 8123
+    },
+    {
+      "epoch": 2.156644099296429,
+      "grad_norm": 0.444593125610364,
+      "learning_rate": 3.5879436338944534e-06,
+      "loss": 0.5622,
+      "step": 8124
+    },
+    {
+      "epoch": 2.156909597769813,
+      "grad_norm": 0.4299837604372277,
+      "learning_rate": 3.5876292878154854e-06,
+      "loss": 0.535,
+      "step": 8125
+    },
+    {
+      "epoch": 2.1571750962431966,
+      "grad_norm": 0.42588043999293956,
+      "learning_rate": 3.587314920525084e-06,
+      "loss": 0.5742,
+      "step": 8126
+    },
+    {
+      "epoch": 2.1574405947165802,
+      "grad_norm": 0.4435783841822148,
+      "learning_rate": 3.587000532029379e-06,
+      "loss": 0.548,
+      "step": 8127
+    },
+    {
+      "epoch": 2.1577060931899643,
+      "grad_norm": 0.4402130338499585,
+      "learning_rate": 3.586686122334502e-06,
+      "loss": 0.5835,
+      "step": 8128
+    },
+    {
+      "epoch": 2.157971591663348,
+      "grad_norm": 0.45109904864501765,
+      "learning_rate": 3.586371691446585e-06,
+      "loss": 0.5684,
+      "step": 8129
+    },
+    {
+      "epoch": 2.1582370901367316,
+      "grad_norm": 0.4403816607230921,
+      "learning_rate": 3.58605723937176e-06,
+      "loss": 0.5946,
+      "step": 8130
+    },
+    {
+      "epoch": 2.1585025886101157,
+      "grad_norm": 0.4402194753650105,
+      "learning_rate": 3.5857427661161597e-06,
+      "loss": 0.5795,
+      "step": 8131
+    },
+    {
+      "epoch": 2.1587680870834993,
+      "grad_norm": 0.45238727725993394,
+      "learning_rate": 3.5854282716859177e-06,
+      "loss": 0.5631,
+      "step": 8132
+    },
+    {
+      "epoch": 2.159033585556883,
+      "grad_norm": 0.4550158011099502,
+      "learning_rate": 3.585113756087166e-06,
+      "loss": 0.5771,
+      "step": 8133
+    },
+    {
+      "epoch": 2.159299084030267,
+      "grad_norm": 0.44036667215678515,
+      "learning_rate": 3.58479921932604e-06,
+      "loss": 0.5737,
+      "step": 8134
+    },
+    {
+      "epoch": 2.1595645825036507,
+      "grad_norm": 0.4271960324273925,
+      "learning_rate": 3.584484661408674e-06,
+      "loss": 0.5727,
+      "step": 8135
+    },
+    {
+      "epoch": 2.1598300809770343,
+      "grad_norm": 0.4402278424947391,
+      "learning_rate": 3.5841700823412e-06,
+      "loss": 0.5649,
+      "step": 8136
+    },
+    {
+      "epoch": 2.160095579450418,
+      "grad_norm": 0.4508434051235174,
+      "learning_rate": 3.5838554821297555e-06,
+      "loss": 0.6116,
+      "step": 8137
+    },
+    {
+      "epoch": 2.160361077923802,
+      "grad_norm": 0.42963368327466434,
+      "learning_rate": 3.583540860780475e-06,
+      "loss": 0.5767,
+      "step": 8138
+    },
+    {
+      "epoch": 2.1606265763971857,
+      "grad_norm": 0.44143819305822724,
+      "learning_rate": 3.5832262182994955e-06,
+      "loss": 0.6067,
+      "step": 8139
+    },
+    {
+      "epoch": 2.1608920748705693,
+      "grad_norm": 0.42630891603874577,
+      "learning_rate": 3.582911554692953e-06,
+      "loss": 0.565,
+      "step": 8140
+    },
+    {
+      "epoch": 2.1611575733439534,
+      "grad_norm": 0.44728169741793067,
+      "learning_rate": 3.5825968699669823e-06,
+      "loss": 0.5853,
+      "step": 8141
+    },
+    {
+      "epoch": 2.161423071817337,
+      "grad_norm": 0.42220483558386435,
+      "learning_rate": 3.582282164127723e-06,
+      "loss": 0.5943,
+      "step": 8142
+    },
+    {
+      "epoch": 2.1616885702907207,
+      "grad_norm": 0.4239303312228211,
+      "learning_rate": 3.5819674371813113e-06,
+      "loss": 0.5686,
+      "step": 8143
+    },
+    {
+      "epoch": 2.161954068764105,
+      "grad_norm": 0.42560296859199037,
+      "learning_rate": 3.5816526891338856e-06,
+      "loss": 0.6004,
+      "step": 8144
+    },
+    {
+      "epoch": 2.1622195672374884,
+      "grad_norm": 0.44369941551844255,
+      "learning_rate": 3.5813379199915837e-06,
+      "loss": 0.6044,
+      "step": 8145
+    },
+    {
+      "epoch": 2.162485065710872,
+      "grad_norm": 0.4313591694719809,
+      "learning_rate": 3.5810231297605446e-06,
+      "loss": 0.5853,
+      "step": 8146
+    },
+    {
+      "epoch": 2.1627505641842557,
+      "grad_norm": 0.4124783745105141,
+      "learning_rate": 3.5807083184469083e-06,
+      "loss": 0.5293,
+      "step": 8147
+    },
+    {
+      "epoch": 2.16301606265764,
+      "grad_norm": 0.4181751115950785,
+      "learning_rate": 3.5803934860568134e-06,
+      "loss": 0.5376,
+      "step": 8148
+    },
+    {
+      "epoch": 2.1632815611310234,
+      "grad_norm": 0.43002984512747994,
+      "learning_rate": 3.580078632596401e-06,
+      "loss": 0.5607,
+      "step": 8149
+    },
+    {
+      "epoch": 2.163547059604407,
+      "grad_norm": 0.42232055231610094,
+      "learning_rate": 3.5797637580718102e-06,
+      "loss": 0.5597,
+      "step": 8150
+    },
+    {
+      "epoch": 2.163812558077791,
+      "grad_norm": 0.41709659332862226,
+      "learning_rate": 3.5794488624891823e-06,
+      "loss": 0.5562,
+      "step": 8151
+    },
+    {
+      "epoch": 2.164078056551175,
+      "grad_norm": 0.43682527666604676,
+      "learning_rate": 3.579133945854659e-06,
+      "loss": 0.5874,
+      "step": 8152
+    },
+    {
+      "epoch": 2.1643435550245584,
+      "grad_norm": 0.4286003119540955,
+      "learning_rate": 3.578819008174381e-06,
+      "loss": 0.6286,
+      "step": 8153
+    },
+    {
+      "epoch": 2.1646090534979425,
+      "grad_norm": 0.4156870132736457,
+      "learning_rate": 3.5785040494544915e-06,
+      "loss": 0.5182,
+      "step": 8154
+    },
+    {
+      "epoch": 2.164874551971326,
+      "grad_norm": 0.4285602721767473,
+      "learning_rate": 3.5781890697011323e-06,
+      "loss": 0.5518,
+      "step": 8155
+    },
+    {
+      "epoch": 2.16514005044471,
+      "grad_norm": 0.4245474171901688,
+      "learning_rate": 3.577874068920446e-06,
+      "loss": 0.5514,
+      "step": 8156
+    },
+    {
+      "epoch": 2.165405548918094,
+      "grad_norm": 0.43446645739812156,
+      "learning_rate": 3.577559047118576e-06,
+      "loss": 0.5768,
+      "step": 8157
+    },
+    {
+      "epoch": 2.1656710473914775,
+      "grad_norm": 0.4293236783877426,
+      "learning_rate": 3.5772440043016676e-06,
+      "loss": 0.5603,
+      "step": 8158
+    },
+    {
+      "epoch": 2.165936545864861,
+      "grad_norm": 0.42444648015691877,
+      "learning_rate": 3.5769289404758622e-06,
+      "loss": 0.5458,
+      "step": 8159
+    },
+    {
+      "epoch": 2.1662020443382453,
+      "grad_norm": 0.43845915890762643,
+      "learning_rate": 3.576613855647306e-06,
+      "loss": 0.5283,
+      "step": 8160
+    },
+    {
+      "epoch": 2.166467542811629,
+      "grad_norm": 0.44043928057689624,
+      "learning_rate": 3.5762987498221446e-06,
+      "loss": 0.5834,
+      "step": 8161
+    },
+    {
+      "epoch": 2.1667330412850125,
+      "grad_norm": 0.4163077062318956,
+      "learning_rate": 3.5759836230065214e-06,
+      "loss": 0.6,
+      "step": 8162
+    },
+    {
+      "epoch": 2.1669985397583966,
+      "grad_norm": 0.43435588587831675,
+      "learning_rate": 3.5756684752065833e-06,
+      "loss": 0.5668,
+      "step": 8163
+    },
+    {
+      "epoch": 2.1672640382317803,
+      "grad_norm": 0.43363716148412906,
+      "learning_rate": 3.575353306428476e-06,
+      "loss": 0.5596,
+      "step": 8164
+    },
+    {
+      "epoch": 2.167529536705164,
+      "grad_norm": 0.4257634052135782,
+      "learning_rate": 3.5750381166783466e-06,
+      "loss": 0.561,
+      "step": 8165
+    },
+    {
+      "epoch": 2.1677950351785475,
+      "grad_norm": 0.4350425327264742,
+      "learning_rate": 3.5747229059623423e-06,
+      "loss": 0.6215,
+      "step": 8166
+    },
+    {
+      "epoch": 2.1680605336519316,
+      "grad_norm": 0.44343055431724576,
+      "learning_rate": 3.574407674286609e-06,
+      "loss": 0.5518,
+      "step": 8167
+    },
+    {
+      "epoch": 2.1683260321253153,
+      "grad_norm": 0.4132255299484832,
+      "learning_rate": 3.574092421657296e-06,
+      "loss": 0.5256,
+      "step": 8168
+    },
+    {
+      "epoch": 2.168591530598699,
+      "grad_norm": 0.4185672844410977,
+      "learning_rate": 3.5737771480805512e-06,
+      "loss": 0.546,
+      "step": 8169
+    },
+    {
+      "epoch": 2.168857029072083,
+      "grad_norm": 0.4401080218773448,
+      "learning_rate": 3.573461853562522e-06,
+      "loss": 0.5384,
+      "step": 8170
+    },
+    {
+      "epoch": 2.1691225275454666,
+      "grad_norm": 0.41538005645587806,
+      "learning_rate": 3.573146538109359e-06,
+      "loss": 0.574,
+      "step": 8171
+    },
+    {
+      "epoch": 2.1693880260188503,
+      "grad_norm": 0.42820904736525567,
+      "learning_rate": 3.5728312017272116e-06,
+      "loss": 0.5972,
+      "step": 8172
+    },
+    {
+      "epoch": 2.1696535244922344,
+      "grad_norm": 0.43682301655958694,
+      "learning_rate": 3.572515844422228e-06,
+      "loss": 0.5796,
+      "step": 8173
+    },
+    {
+      "epoch": 2.169919022965618,
+      "grad_norm": 0.4404141268523503,
+      "learning_rate": 3.572200466200559e-06,
+      "loss": 0.6077,
+      "step": 8174
+    },
+    {
+      "epoch": 2.1701845214390016,
+      "grad_norm": 0.4209395464405756,
+      "learning_rate": 3.571885067068357e-06,
+      "loss": 0.5504,
+      "step": 8175
+    },
+    {
+      "epoch": 2.1704500199123853,
+      "grad_norm": 0.43985891046831765,
+      "learning_rate": 3.571569647031771e-06,
+      "loss": 0.557,
+      "step": 8176
+    },
+    {
+      "epoch": 2.1707155183857694,
+      "grad_norm": 0.4337082913883347,
+      "learning_rate": 3.5712542060969536e-06,
+      "loss": 0.6126,
+      "step": 8177
+    },
+    {
+      "epoch": 2.170981016859153,
+      "grad_norm": 0.4134877894423131,
+      "learning_rate": 3.570938744270056e-06,
+      "loss": 0.5112,
+      "step": 8178
+    },
+    {
+      "epoch": 2.1712465153325367,
+      "grad_norm": 0.4304441292411016,
+      "learning_rate": 3.570623261557231e-06,
+      "loss": 0.5728,
+      "step": 8179
+    },
+    {
+      "epoch": 2.1715120138059207,
+      "grad_norm": 0.43413905956266996,
+      "learning_rate": 3.570307757964631e-06,
+      "loss": 0.5962,
+      "step": 8180
+    },
+    {
+      "epoch": 2.1717775122793044,
+      "grad_norm": 0.4318863366540812,
+      "learning_rate": 3.5699922334984092e-06,
+      "loss": 0.584,
+      "step": 8181
+    },
+    {
+      "epoch": 2.172043010752688,
+      "grad_norm": 0.42977470607838475,
+      "learning_rate": 3.5696766881647196e-06,
+      "loss": 0.5461,
+      "step": 8182
+    },
+    {
+      "epoch": 2.172308509226072,
+      "grad_norm": 0.42740902399867675,
+      "learning_rate": 3.569361121969714e-06,
+      "loss": 0.5473,
+      "step": 8183
+    },
+    {
+      "epoch": 2.1725740076994557,
+      "grad_norm": 0.4201858059150775,
+      "learning_rate": 3.56904553491955e-06,
+      "loss": 0.6002,
+      "step": 8184
+    },
+    {
+      "epoch": 2.1728395061728394,
+      "grad_norm": 0.4214972963254856,
+      "learning_rate": 3.5687299270203797e-06,
+      "loss": 0.5679,
+      "step": 8185
+    },
+    {
+      "epoch": 2.1731050046462235,
+      "grad_norm": 0.4357809484239828,
+      "learning_rate": 3.568414298278359e-06,
+      "loss": 0.5326,
+      "step": 8186
+    },
+    {
+      "epoch": 2.173370503119607,
+      "grad_norm": 0.4182997779840323,
+      "learning_rate": 3.5680986486996437e-06,
+      "loss": 0.537,
+      "step": 8187
+    },
+    {
+      "epoch": 2.1736360015929908,
+      "grad_norm": 0.4265751365852189,
+      "learning_rate": 3.56778297829039e-06,
+      "loss": 0.5582,
+      "step": 8188
+    },
+    {
+      "epoch": 2.173901500066375,
+      "grad_norm": 0.4268933036491036,
+      "learning_rate": 3.567467287056754e-06,
+      "loss": 0.5654,
+      "step": 8189
+    },
+    {
+      "epoch": 2.1741669985397585,
+      "grad_norm": 0.4245872754260466,
+      "learning_rate": 3.5671515750048913e-06,
+      "loss": 0.5678,
+      "step": 8190
+    },
+    {
+      "epoch": 2.174432497013142,
+      "grad_norm": 0.4350204810888233,
+      "learning_rate": 3.566835842140961e-06,
+      "loss": 0.5996,
+      "step": 8191
+    },
+    {
+      "epoch": 2.1746979954865258,
+      "grad_norm": 0.4327623634826044,
+      "learning_rate": 3.5665200884711194e-06,
+      "loss": 0.5766,
+      "step": 8192
+    },
+    {
+      "epoch": 2.17496349395991,
+      "grad_norm": 0.4400935898065364,
+      "learning_rate": 3.566204314001524e-06,
+      "loss": 0.6037,
+      "step": 8193
+    },
+    {
+      "epoch": 2.1752289924332935,
+      "grad_norm": 0.4429245613639675,
+      "learning_rate": 3.5658885187383346e-06,
+      "loss": 0.5913,
+      "step": 8194
+    },
+    {
+      "epoch": 2.175494490906677,
+      "grad_norm": 0.42846851720164053,
+      "learning_rate": 3.5655727026877095e-06,
+      "loss": 0.5797,
+      "step": 8195
+    },
+    {
+      "epoch": 2.175759989380061,
+      "grad_norm": 0.4362558953744138,
+      "learning_rate": 3.565256865855808e-06,
+      "loss": 0.5229,
+      "step": 8196
+    },
+    {
+      "epoch": 2.176025487853445,
+      "grad_norm": 0.4290593696978451,
+      "learning_rate": 3.564941008248789e-06,
+      "loss": 0.5165,
+      "step": 8197
+    },
+    {
+      "epoch": 2.1762909863268285,
+      "grad_norm": 0.42812846591824266,
+      "learning_rate": 3.564625129872813e-06,
+      "loss": 0.5968,
+      "step": 8198
+    },
+    {
+      "epoch": 2.1765564848002126,
+      "grad_norm": 0.4253343475371097,
+      "learning_rate": 3.5643092307340406e-06,
+      "loss": 0.578,
+      "step": 8199
+    },
+    {
+      "epoch": 2.176821983273596,
+      "grad_norm": 0.44309630840798947,
+      "learning_rate": 3.5639933108386314e-06,
+      "loss": 0.5547,
+      "step": 8200
+    },
+    {
+      "epoch": 2.17708748174698,
+      "grad_norm": 0.44839504606183656,
+      "learning_rate": 3.5636773701927484e-06,
+      "loss": 0.5828,
+      "step": 8201
+    },
+    {
+      "epoch": 2.1773529802203635,
+      "grad_norm": 0.42413245538646405,
+      "learning_rate": 3.5633614088025513e-06,
+      "loss": 0.5709,
+      "step": 8202
+    },
+    {
+      "epoch": 2.1776184786937476,
+      "grad_norm": 0.4345874024185505,
+      "learning_rate": 3.563045426674204e-06,
+      "loss": 0.5607,
+      "step": 8203
+    },
+    {
+      "epoch": 2.1778839771671312,
+      "grad_norm": 0.4382591407266538,
+      "learning_rate": 3.562729423813869e-06,
+      "loss": 0.5688,
+      "step": 8204
+    },
+    {
+      "epoch": 2.178149475640515,
+      "grad_norm": 0.4508822147921061,
+      "learning_rate": 3.5624134002277065e-06,
+      "loss": 0.552,
+      "step": 8205
+    },
+    {
+      "epoch": 2.178414974113899,
+      "grad_norm": 0.43771362012163545,
+      "learning_rate": 3.5620973559218823e-06,
+      "loss": 0.5623,
+      "step": 8206
+    },
+    {
+      "epoch": 2.1786804725872826,
+      "grad_norm": 0.4349094077445569,
+      "learning_rate": 3.5617812909025594e-06,
+      "loss": 0.5655,
+      "step": 8207
+    },
+    {
+      "epoch": 2.1789459710606662,
+      "grad_norm": 0.419731233863387,
+      "learning_rate": 3.5614652051759012e-06,
+      "loss": 0.5582,
+      "step": 8208
+    },
+    {
+      "epoch": 2.1792114695340503,
+      "grad_norm": 0.42420289270805106,
+      "learning_rate": 3.561149098748073e-06,
+      "loss": 0.5656,
+      "step": 8209
+    },
+    {
+      "epoch": 2.179476968007434,
+      "grad_norm": 0.44416463287902547,
+      "learning_rate": 3.5608329716252388e-06,
+      "loss": 0.5858,
+      "step": 8210
+    },
+    {
+      "epoch": 2.1797424664808176,
+      "grad_norm": 0.4328970772730496,
+      "learning_rate": 3.5605168238135647e-06,
+      "loss": 0.5897,
+      "step": 8211
+    },
+    {
+      "epoch": 2.1800079649542017,
+      "grad_norm": 0.485429221181809,
+      "learning_rate": 3.560200655319216e-06,
+      "loss": 0.5677,
+      "step": 8212
+    },
+    {
+      "epoch": 2.1802734634275853,
+      "grad_norm": 0.42794070030007736,
+      "learning_rate": 3.559884466148358e-06,
+      "loss": 0.5977,
+      "step": 8213
+    },
+    {
+      "epoch": 2.180538961900969,
+      "grad_norm": 0.4303144894484324,
+      "learning_rate": 3.559568256307158e-06,
+      "loss": 0.5959,
+      "step": 8214
+    },
+    {
+      "epoch": 2.180804460374353,
+      "grad_norm": 0.420496518331875,
+      "learning_rate": 3.559252025801784e-06,
+      "loss": 0.5758,
+      "step": 8215
+    },
+    {
+      "epoch": 2.1810699588477367,
+      "grad_norm": 0.45985407294003083,
+      "learning_rate": 3.5589357746384e-06,
+      "loss": 0.569,
+      "step": 8216
+    },
+    {
+      "epoch": 2.1813354573211203,
+      "grad_norm": 0.44820774763996896,
+      "learning_rate": 3.5586195028231766e-06,
+      "loss": 0.6039,
+      "step": 8217
+    },
+    {
+      "epoch": 2.1816009557945044,
+      "grad_norm": 0.44703879405779456,
+      "learning_rate": 3.558303210362282e-06,
+      "loss": 0.5783,
+      "step": 8218
+    },
+    {
+      "epoch": 2.181866454267888,
+      "grad_norm": 0.42251292720987427,
+      "learning_rate": 3.557986897261882e-06,
+      "loss": 0.5745,
+      "step": 8219
+    },
+    {
+      "epoch": 2.1821319527412717,
+      "grad_norm": 0.4214502888738516,
+      "learning_rate": 3.5576705635281474e-06,
+      "loss": 0.5952,
+      "step": 8220
+    },
+    {
+      "epoch": 2.1823974512146553,
+      "grad_norm": 0.4299714354129137,
+      "learning_rate": 3.557354209167248e-06,
+      "loss": 0.5466,
+      "step": 8221
+    },
+    {
+      "epoch": 2.1826629496880394,
+      "grad_norm": 0.42249806672703577,
+      "learning_rate": 3.557037834185352e-06,
+      "loss": 0.5809,
+      "step": 8222
+    },
+    {
+      "epoch": 2.182928448161423,
+      "grad_norm": 0.4175285126918826,
+      "learning_rate": 3.5567214385886305e-06,
+      "loss": 0.5162,
+      "step": 8223
+    },
+    {
+      "epoch": 2.1831939466348067,
+      "grad_norm": 0.4417292348315966,
+      "learning_rate": 3.5564050223832538e-06,
+      "loss": 0.5713,
+      "step": 8224
+    },
+    {
+      "epoch": 2.183459445108191,
+      "grad_norm": 0.42572683804037714,
+      "learning_rate": 3.5560885855753926e-06,
+      "loss": 0.5477,
+      "step": 8225
+    },
+    {
+      "epoch": 2.1837249435815744,
+      "grad_norm": 0.42012695262841354,
+      "learning_rate": 3.5557721281712176e-06,
+      "loss": 0.5495,
+      "step": 8226
+    },
+    {
+      "epoch": 2.183990442054958,
+      "grad_norm": 0.43424284486386594,
+      "learning_rate": 3.5554556501769015e-06,
+      "loss": 0.5477,
+      "step": 8227
+    },
+    {
+      "epoch": 2.184255940528342,
+      "grad_norm": 0.43081846640872273,
+      "learning_rate": 3.5551391515986163e-06,
+      "loss": 0.5655,
+      "step": 8228
+    },
+    {
+      "epoch": 2.184521439001726,
+      "grad_norm": 0.43668195409678656,
+      "learning_rate": 3.554822632442534e-06,
+      "loss": 0.5812,
+      "step": 8229
+    },
+    {
+      "epoch": 2.1847869374751094,
+      "grad_norm": 0.4667848112936377,
+      "learning_rate": 3.5545060927148273e-06,
+      "loss": 0.5317,
+      "step": 8230
+    },
+    {
+      "epoch": 2.185052435948493,
+      "grad_norm": 0.4266263563969217,
+      "learning_rate": 3.5541895324216703e-06,
+      "loss": 0.5307,
+      "step": 8231
+    },
+    {
+      "epoch": 2.185317934421877,
+      "grad_norm": 0.43559739727073227,
+      "learning_rate": 3.553872951569236e-06,
+      "loss": 0.5541,
+      "step": 8232
+    },
+    {
+      "epoch": 2.185583432895261,
+      "grad_norm": 0.43064820473070414,
+      "learning_rate": 3.5535563501636993e-06,
+      "loss": 0.5627,
+      "step": 8233
+    },
+    {
+      "epoch": 2.1858489313686444,
+      "grad_norm": 0.44219672662368725,
+      "learning_rate": 3.553239728211234e-06,
+      "loss": 0.5732,
+      "step": 8234
+    },
+    {
+      "epoch": 2.1861144298420285,
+      "grad_norm": 0.42773234888704736,
+      "learning_rate": 3.552923085718016e-06,
+      "loss": 0.5234,
+      "step": 8235
+    },
+    {
+      "epoch": 2.186379928315412,
+      "grad_norm": 0.4279662449696274,
+      "learning_rate": 3.552606422690219e-06,
+      "loss": 0.5576,
+      "step": 8236
+    },
+    {
+      "epoch": 2.186645426788796,
+      "grad_norm": 0.4114495811837613,
+      "learning_rate": 3.5522897391340193e-06,
+      "loss": 0.5448,
+      "step": 8237
+    },
+    {
+      "epoch": 2.18691092526218,
+      "grad_norm": 0.43408766284319705,
+      "learning_rate": 3.5519730350555935e-06,
+      "loss": 0.5925,
+      "step": 8238
+    },
+    {
+      "epoch": 2.1871764237355635,
+      "grad_norm": 0.41787049555493355,
+      "learning_rate": 3.551656310461118e-06,
+      "loss": 0.5787,
+      "step": 8239
+    },
+    {
+      "epoch": 2.187441922208947,
+      "grad_norm": 0.42038280658004606,
+      "learning_rate": 3.551339565356769e-06,
+      "loss": 0.5781,
+      "step": 8240
+    },
+    {
+      "epoch": 2.1877074206823313,
+      "grad_norm": 0.426015512443387,
+      "learning_rate": 3.5510227997487252e-06,
+      "loss": 0.5385,
+      "step": 8241
+    },
+    {
+      "epoch": 2.187972919155715,
+      "grad_norm": 0.4208504215952971,
+      "learning_rate": 3.550706013643163e-06,
+      "loss": 0.5574,
+      "step": 8242
+    },
+    {
+      "epoch": 2.1882384176290985,
+      "grad_norm": 0.43581887705368705,
+      "learning_rate": 3.550389207046261e-06,
+      "loss": 0.5608,
+      "step": 8243
+    },
+    {
+      "epoch": 2.1885039161024826,
+      "grad_norm": 0.42153121829924495,
+      "learning_rate": 3.550072379964198e-06,
+      "loss": 0.5397,
+      "step": 8244
+    },
+    {
+      "epoch": 2.1887694145758663,
+      "grad_norm": 0.4322108869054508,
+      "learning_rate": 3.549755532403152e-06,
+      "loss": 0.5702,
+      "step": 8245
+    },
+    {
+      "epoch": 2.18903491304925,
+      "grad_norm": 0.43294191445238983,
+      "learning_rate": 3.549438664369303e-06,
+      "loss": 0.5758,
+      "step": 8246
+    },
+    {
+      "epoch": 2.1893004115226335,
+      "grad_norm": 0.4395060571566151,
+      "learning_rate": 3.549121775868831e-06,
+      "loss": 0.5859,
+      "step": 8247
+    },
+    {
+      "epoch": 2.1895659099960176,
+      "grad_norm": 0.43256943379293084,
+      "learning_rate": 3.548804866907915e-06,
+      "loss": 0.5604,
+      "step": 8248
+    },
+    {
+      "epoch": 2.1898314084694013,
+      "grad_norm": 0.46267838789413823,
+      "learning_rate": 3.548487937492736e-06,
+      "loss": 0.5406,
+      "step": 8249
+    },
+    {
+      "epoch": 2.190096906942785,
+      "grad_norm": 0.4371956688996471,
+      "learning_rate": 3.548170987629477e-06,
+      "loss": 0.5688,
+      "step": 8250
+    },
+    {
+      "epoch": 2.190362405416169,
+      "grad_norm": 0.4355967835311301,
+      "learning_rate": 3.5478540173243153e-06,
+      "loss": 0.5934,
+      "step": 8251
+    },
+    {
+      "epoch": 2.1906279038895526,
+      "grad_norm": 0.4358903697195902,
+      "learning_rate": 3.547537026583435e-06,
+      "loss": 0.5856,
+      "step": 8252
+    },
+    {
+      "epoch": 2.1908934023629363,
+      "grad_norm": 0.41563418588273116,
+      "learning_rate": 3.5472200154130187e-06,
+      "loss": 0.5655,
+      "step": 8253
+    },
+    {
+      "epoch": 2.1911589008363204,
+      "grad_norm": 0.42897289332947863,
+      "learning_rate": 3.546902983819248e-06,
+      "loss": 0.583,
+      "step": 8254
+    },
+    {
+      "epoch": 2.191424399309704,
+      "grad_norm": 0.4428316543180506,
+      "learning_rate": 3.5465859318083053e-06,
+      "loss": 0.5912,
+      "step": 8255
+    },
+    {
+      "epoch": 2.1916898977830876,
+      "grad_norm": 0.44889167917542233,
+      "learning_rate": 3.546268859386374e-06,
+      "loss": 0.5769,
+      "step": 8256
+    },
+    {
+      "epoch": 2.1919553962564717,
+      "grad_norm": 0.4196362672219529,
+      "learning_rate": 3.5459517665596387e-06,
+      "loss": 0.5845,
+      "step": 8257
+    },
+    {
+      "epoch": 2.1922208947298554,
+      "grad_norm": 0.41756354057729594,
+      "learning_rate": 3.545634653334284e-06,
+      "loss": 0.5346,
+      "step": 8258
+    },
+    {
+      "epoch": 2.192486393203239,
+      "grad_norm": 0.4314161492977687,
+      "learning_rate": 3.5453175197164923e-06,
+      "loss": 0.5666,
+      "step": 8259
+    },
+    {
+      "epoch": 2.1927518916766227,
+      "grad_norm": 0.4279655761928378,
+      "learning_rate": 3.54500036571245e-06,
+      "loss": 0.553,
+      "step": 8260
+    },
+    {
+      "epoch": 2.1930173901500067,
+      "grad_norm": 0.43685382032894915,
+      "learning_rate": 3.544683191328342e-06,
+      "loss": 0.5881,
+      "step": 8261
+    },
+    {
+      "epoch": 2.1932828886233904,
+      "grad_norm": 0.43426636251180323,
+      "learning_rate": 3.5443659965703537e-06,
+      "loss": 0.603,
+      "step": 8262
+    },
+    {
+      "epoch": 2.193548387096774,
+      "grad_norm": 0.4424976177629162,
+      "learning_rate": 3.5440487814446712e-06,
+      "loss": 0.5668,
+      "step": 8263
+    },
+    {
+      "epoch": 2.193813885570158,
+      "grad_norm": 0.4239060470627123,
+      "learning_rate": 3.5437315459574823e-06,
+      "loss": 0.5854,
+      "step": 8264
+    },
+    {
+      "epoch": 2.1940793840435417,
+      "grad_norm": 0.4184852893279085,
+      "learning_rate": 3.5434142901149714e-06,
+      "loss": 0.5392,
+      "step": 8265
+    },
+    {
+      "epoch": 2.1943448825169254,
+      "grad_norm": 0.4207446567401549,
+      "learning_rate": 3.5430970139233274e-06,
+      "loss": 0.5681,
+      "step": 8266
+    },
+    {
+      "epoch": 2.1946103809903095,
+      "grad_norm": 0.43230939022087506,
+      "learning_rate": 3.542779717388738e-06,
+      "loss": 0.5785,
+      "step": 8267
+    },
+    {
+      "epoch": 2.194875879463693,
+      "grad_norm": 0.44145054946334555,
+      "learning_rate": 3.5424624005173914e-06,
+      "loss": 0.5799,
+      "step": 8268
+    },
+    {
+      "epoch": 2.1951413779370768,
+      "grad_norm": 0.4271199181712962,
+      "learning_rate": 3.542145063315475e-06,
+      "loss": 0.5561,
+      "step": 8269
+    },
+    {
+      "epoch": 2.195406876410461,
+      "grad_norm": 0.4254452778114775,
+      "learning_rate": 3.541827705789178e-06,
+      "loss": 0.5742,
+      "step": 8270
+    },
+    {
+      "epoch": 2.1956723748838445,
+      "grad_norm": 0.42974644438478876,
+      "learning_rate": 3.5415103279446905e-06,
+      "loss": 0.5686,
+      "step": 8271
+    },
+    {
+      "epoch": 2.195937873357228,
+      "grad_norm": 0.41875794267624056,
+      "learning_rate": 3.5411929297882016e-06,
+      "loss": 0.5525,
+      "step": 8272
+    },
+    {
+      "epoch": 2.196203371830612,
+      "grad_norm": 0.41629790580721837,
+      "learning_rate": 3.5408755113259006e-06,
+      "loss": 0.5324,
+      "step": 8273
+    },
+    {
+      "epoch": 2.196468870303996,
+      "grad_norm": 0.4265955826691378,
+      "learning_rate": 3.5405580725639795e-06,
+      "loss": 0.5632,
+      "step": 8274
+    },
+    {
+      "epoch": 2.1967343687773795,
+      "grad_norm": 0.43937639606143214,
+      "learning_rate": 3.540240613508627e-06,
+      "loss": 0.573,
+      "step": 8275
+    },
+    {
+      "epoch": 2.196999867250763,
+      "grad_norm": 0.42558416950544165,
+      "learning_rate": 3.5399231341660367e-06,
+      "loss": 0.5586,
+      "step": 8276
+    },
+    {
+      "epoch": 2.197265365724147,
+      "grad_norm": 0.4502049916517734,
+      "learning_rate": 3.539605634542399e-06,
+      "loss": 0.585,
+      "step": 8277
+    },
+    {
+      "epoch": 2.197530864197531,
+      "grad_norm": 0.43075218583839087,
+      "learning_rate": 3.539288114643905e-06,
+      "loss": 0.5879,
+      "step": 8278
+    },
+    {
+      "epoch": 2.1977963626709145,
+      "grad_norm": 0.4307620764894786,
+      "learning_rate": 3.53897057447675e-06,
+      "loss": 0.5757,
+      "step": 8279
+    },
+    {
+      "epoch": 2.1980618611442986,
+      "grad_norm": 0.4065437801292464,
+      "learning_rate": 3.5386530140471235e-06,
+      "loss": 0.5488,
+      "step": 8280
+    },
+    {
+      "epoch": 2.198327359617682,
+      "grad_norm": 0.42752226628840745,
+      "learning_rate": 3.5383354333612208e-06,
+      "loss": 0.584,
+      "step": 8281
+    },
+    {
+      "epoch": 2.198592858091066,
+      "grad_norm": 0.4233632441946237,
+      "learning_rate": 3.5380178324252344e-06,
+      "loss": 0.6003,
+      "step": 8282
+    },
+    {
+      "epoch": 2.19885835656445,
+      "grad_norm": 0.41544740428974836,
+      "learning_rate": 3.5377002112453597e-06,
+      "loss": 0.5864,
+      "step": 8283
+    },
+    {
+      "epoch": 2.1991238550378336,
+      "grad_norm": 0.43009166407730703,
+      "learning_rate": 3.537382569827789e-06,
+      "loss": 0.5552,
+      "step": 8284
+    },
+    {
+      "epoch": 2.1993893535112172,
+      "grad_norm": 0.41465725294161654,
+      "learning_rate": 3.537064908178718e-06,
+      "loss": 0.6049,
+      "step": 8285
+    },
+    {
+      "epoch": 2.199654851984601,
+      "grad_norm": 0.4342671353261724,
+      "learning_rate": 3.5367472263043435e-06,
+      "loss": 0.5894,
+      "step": 8286
+    },
+    {
+      "epoch": 2.199920350457985,
+      "grad_norm": 0.43241203932344835,
+      "learning_rate": 3.536429524210859e-06,
+      "loss": 0.5864,
+      "step": 8287
+    },
+    {
+      "epoch": 2.2001858489313686,
+      "grad_norm": 0.46321167563830457,
+      "learning_rate": 3.5361118019044616e-06,
+      "loss": 0.5681,
+      "step": 8288
+    },
+    {
+      "epoch": 2.2004513474047522,
+      "grad_norm": 0.4298134654298973,
+      "learning_rate": 3.5357940593913464e-06,
+      "loss": 0.5573,
+      "step": 8289
+    },
+    {
+      "epoch": 2.2007168458781363,
+      "grad_norm": 0.43331419233183927,
+      "learning_rate": 3.5354762966777118e-06,
+      "loss": 0.581,
+      "step": 8290
+    },
+    {
+      "epoch": 2.20098234435152,
+      "grad_norm": 0.4136023374845611,
+      "learning_rate": 3.5351585137697537e-06,
+      "loss": 0.5372,
+      "step": 8291
+    },
+    {
+      "epoch": 2.2012478428249036,
+      "grad_norm": 0.44456136916550354,
+      "learning_rate": 3.53484071067367e-06,
+      "loss": 0.5666,
+      "step": 8292
+    },
+    {
+      "epoch": 2.2015133412982877,
+      "grad_norm": 0.4431851966579242,
+      "learning_rate": 3.5345228873956583e-06,
+      "loss": 0.5442,
+      "step": 8293
+    },
+    {
+      "epoch": 2.2017788397716713,
+      "grad_norm": 0.42116011574848733,
+      "learning_rate": 3.5342050439419177e-06,
+      "loss": 0.5414,
+      "step": 8294
+    },
+    {
+      "epoch": 2.202044338245055,
+      "grad_norm": 0.43025446794780403,
+      "learning_rate": 3.533887180318647e-06,
+      "loss": 0.5592,
+      "step": 8295
+    },
+    {
+      "epoch": 2.202309836718439,
+      "grad_norm": 0.4120201442272711,
+      "learning_rate": 3.533569296532045e-06,
+      "loss": 0.5608,
+      "step": 8296
+    },
+    {
+      "epoch": 2.2025753351918227,
+      "grad_norm": 0.4277289989056543,
+      "learning_rate": 3.5332513925883105e-06,
+      "loss": 0.5553,
+      "step": 8297
+    },
+    {
+      "epoch": 2.2028408336652063,
+      "grad_norm": 0.42433709926151014,
+      "learning_rate": 3.532933468493644e-06,
+      "loss": 0.5891,
+      "step": 8298
+    },
+    {
+      "epoch": 2.2031063321385904,
+      "grad_norm": 0.4414095462830388,
+      "learning_rate": 3.5326155242542465e-06,
+      "loss": 0.5707,
+      "step": 8299
+    },
+    {
+      "epoch": 2.203371830611974,
+      "grad_norm": 0.4320114733918895,
+      "learning_rate": 3.5322975598763176e-06,
+      "loss": 0.5411,
+      "step": 8300
+    },
+    {
+      "epoch": 2.2036373290853577,
+      "grad_norm": 0.4367447582100221,
+      "learning_rate": 3.5319795753660584e-06,
+      "loss": 0.545,
+      "step": 8301
+    },
+    {
+      "epoch": 2.2039028275587413,
+      "grad_norm": 0.40914417434468037,
+      "learning_rate": 3.5316615707296707e-06,
+      "loss": 0.5422,
+      "step": 8302
+    },
+    {
+      "epoch": 2.2041683260321254,
+      "grad_norm": 0.4238222869219569,
+      "learning_rate": 3.5313435459733568e-06,
+      "loss": 0.576,
+      "step": 8303
+    },
+    {
+      "epoch": 2.204433824505509,
+      "grad_norm": 0.4284090015978858,
+      "learning_rate": 3.531025501103319e-06,
+      "loss": 0.5879,
+      "step": 8304
+    },
+    {
+      "epoch": 2.2046993229788927,
+      "grad_norm": 0.41743672713568314,
+      "learning_rate": 3.530707436125759e-06,
+      "loss": 0.5287,
+      "step": 8305
+    },
+    {
+      "epoch": 2.204964821452277,
+      "grad_norm": 0.41605541793651385,
+      "learning_rate": 3.5303893510468804e-06,
+      "loss": 0.5495,
+      "step": 8306
+    },
+    {
+      "epoch": 2.2052303199256604,
+      "grad_norm": 0.44833032441811826,
+      "learning_rate": 3.5300712458728874e-06,
+      "loss": 0.5551,
+      "step": 8307
+    },
+    {
+      "epoch": 2.205495818399044,
+      "grad_norm": 0.43177408499726616,
+      "learning_rate": 3.529753120609982e-06,
+      "loss": 0.5378,
+      "step": 8308
+    },
+    {
+      "epoch": 2.205761316872428,
+      "grad_norm": 0.4299711299636745,
+      "learning_rate": 3.5294349752643695e-06,
+      "loss": 0.5416,
+      "step": 8309
+    },
+    {
+      "epoch": 2.206026815345812,
+      "grad_norm": 0.4304833589950197,
+      "learning_rate": 3.529116809842255e-06,
+      "loss": 0.5527,
+      "step": 8310
+    },
+    {
+      "epoch": 2.2062923138191954,
+      "grad_norm": 0.42530363028349183,
+      "learning_rate": 3.528798624349843e-06,
+      "loss": 0.5733,
+      "step": 8311
+    },
+    {
+      "epoch": 2.2065578122925795,
+      "grad_norm": 0.4565838053712839,
+      "learning_rate": 3.5284804187933388e-06,
+      "loss": 0.5615,
+      "step": 8312
+    },
+    {
+      "epoch": 2.206823310765963,
+      "grad_norm": 0.42473966476144603,
+      "learning_rate": 3.5281621931789483e-06,
+      "loss": 0.5518,
+      "step": 8313
+    },
+    {
+      "epoch": 2.207088809239347,
+      "grad_norm": 0.43216909661291353,
+      "learning_rate": 3.527843947512878e-06,
+      "loss": 0.5538,
+      "step": 8314
+    },
+    {
+      "epoch": 2.2073543077127304,
+      "grad_norm": 0.43320833386447716,
+      "learning_rate": 3.527525681801333e-06,
+      "loss": 0.5575,
+      "step": 8315
+    },
+    {
+      "epoch": 2.2076198061861145,
+      "grad_norm": 0.42953050913820645,
+      "learning_rate": 3.5272073960505226e-06,
+      "loss": 0.5488,
+      "step": 8316
+    },
+    {
+      "epoch": 2.207885304659498,
+      "grad_norm": 0.42303394981790604,
+      "learning_rate": 3.5268890902666524e-06,
+      "loss": 0.5829,
+      "step": 8317
+    },
+    {
+      "epoch": 2.208150803132882,
+      "grad_norm": 0.4349062050497459,
+      "learning_rate": 3.5265707644559312e-06,
+      "loss": 0.5888,
+      "step": 8318
+    },
+    {
+      "epoch": 2.208416301606266,
+      "grad_norm": 0.43152811257234547,
+      "learning_rate": 3.526252418624566e-06,
+      "loss": 0.5514,
+      "step": 8319
+    },
+    {
+      "epoch": 2.2086818000796495,
+      "grad_norm": 0.4333169027420834,
+      "learning_rate": 3.525934052778766e-06,
+      "loss": 0.5718,
+      "step": 8320
+    },
+    {
+      "epoch": 2.208947298553033,
+      "grad_norm": 0.416132330653859,
+      "learning_rate": 3.52561566692474e-06,
+      "loss": 0.5699,
+      "step": 8321
+    },
+    {
+      "epoch": 2.2092127970264173,
+      "grad_norm": 0.43472154429191506,
+      "learning_rate": 3.5252972610686974e-06,
+      "loss": 0.5688,
+      "step": 8322
+    },
+    {
+      "epoch": 2.209478295499801,
+      "grad_norm": 0.4429422051314671,
+      "learning_rate": 3.5249788352168478e-06,
+      "loss": 0.5708,
+      "step": 8323
+    },
+    {
+      "epoch": 2.2097437939731845,
+      "grad_norm": 0.4215939656471332,
+      "learning_rate": 3.5246603893754017e-06,
+      "loss": 0.5994,
+      "step": 8324
+    },
+    {
+      "epoch": 2.2100092924465686,
+      "grad_norm": 0.43953672985193265,
+      "learning_rate": 3.5243419235505693e-06,
+      "loss": 0.5646,
+      "step": 8325
+    },
+    {
+      "epoch": 2.2102747909199523,
+      "grad_norm": 0.441010459267688,
+      "learning_rate": 3.5240234377485605e-06,
+      "loss": 0.5675,
+      "step": 8326
+    },
+    {
+      "epoch": 2.210540289393336,
+      "grad_norm": 0.4297897679398152,
+      "learning_rate": 3.5237049319755884e-06,
+      "loss": 0.5809,
+      "step": 8327
+    },
+    {
+      "epoch": 2.21080578786672,
+      "grad_norm": 0.4334697137103331,
+      "learning_rate": 3.523386406237863e-06,
+      "loss": 0.558,
+      "step": 8328
+    },
+    {
+      "epoch": 2.2110712863401036,
+      "grad_norm": 0.4251408146269121,
+      "learning_rate": 3.5230678605415976e-06,
+      "loss": 0.5963,
+      "step": 8329
+    },
+    {
+      "epoch": 2.2113367848134873,
+      "grad_norm": 0.4322699901421514,
+      "learning_rate": 3.5227492948930036e-06,
+      "loss": 0.5973,
+      "step": 8330
+    },
+    {
+      "epoch": 2.211602283286871,
+      "grad_norm": 0.4332140386304007,
+      "learning_rate": 3.522430709298294e-06,
+      "loss": 0.5691,
+      "step": 8331
+    },
+    {
+      "epoch": 2.211867781760255,
+      "grad_norm": 0.4216068715173031,
+      "learning_rate": 3.5221121037636825e-06,
+      "loss": 0.5386,
+      "step": 8332
+    },
+    {
+      "epoch": 2.2121332802336386,
+      "grad_norm": 0.4385664815616991,
+      "learning_rate": 3.5217934782953826e-06,
+      "loss": 0.567,
+      "step": 8333
+    },
+    {
+      "epoch": 2.2123987787070223,
+      "grad_norm": 0.4121095691103931,
+      "learning_rate": 3.5214748328996077e-06,
+      "loss": 0.5484,
+      "step": 8334
+    },
+    {
+      "epoch": 2.2126642771804064,
+      "grad_norm": 0.4266373223965364,
+      "learning_rate": 3.521156167582573e-06,
+      "loss": 0.5729,
+      "step": 8335
+    },
+    {
+      "epoch": 2.21292977565379,
+      "grad_norm": 0.43879405509868213,
+      "learning_rate": 3.520837482350493e-06,
+      "loss": 0.5173,
+      "step": 8336
+    },
+    {
+      "epoch": 2.2131952741271737,
+      "grad_norm": 0.4481433957343789,
+      "learning_rate": 3.5205187772095816e-06,
+      "loss": 0.5656,
+      "step": 8337
+    },
+    {
+      "epoch": 2.2134607726005577,
+      "grad_norm": 0.44229196800829995,
+      "learning_rate": 3.520200052166056e-06,
+      "loss": 0.5688,
+      "step": 8338
+    },
+    {
+      "epoch": 2.2137262710739414,
+      "grad_norm": 0.4302503723254411,
+      "learning_rate": 3.5198813072261316e-06,
+      "loss": 0.5484,
+      "step": 8339
+    },
+    {
+      "epoch": 2.213991769547325,
+      "grad_norm": 0.4615235568985375,
+      "learning_rate": 3.5195625423960243e-06,
+      "loss": 0.5363,
+      "step": 8340
+    },
+    {
+      "epoch": 2.2142572680207087,
+      "grad_norm": 0.4576982532899516,
+      "learning_rate": 3.5192437576819517e-06,
+      "loss": 0.5272,
+      "step": 8341
+    },
+    {
+      "epoch": 2.2145227664940927,
+      "grad_norm": 0.4292199220014541,
+      "learning_rate": 3.5189249530901303e-06,
+      "loss": 0.5462,
+      "step": 8342
+    },
+    {
+      "epoch": 2.2147882649674764,
+      "grad_norm": 0.43028663695064484,
+      "learning_rate": 3.518606128626777e-06,
+      "loss": 0.542,
+      "step": 8343
+    },
+    {
+      "epoch": 2.21505376344086,
+      "grad_norm": 0.4417543912262415,
+      "learning_rate": 3.51828728429811e-06,
+      "loss": 0.5651,
+      "step": 8344
+    },
+    {
+      "epoch": 2.215319261914244,
+      "grad_norm": 0.44432485196135546,
+      "learning_rate": 3.5179684201103487e-06,
+      "loss": 0.5935,
+      "step": 8345
+    },
+    {
+      "epoch": 2.2155847603876277,
+      "grad_norm": 0.4388244346598191,
+      "learning_rate": 3.51764953606971e-06,
+      "loss": 0.5503,
+      "step": 8346
+    },
+    {
+      "epoch": 2.2158502588610114,
+      "grad_norm": 0.4292639269340591,
+      "learning_rate": 3.517330632182414e-06,
+      "loss": 0.5643,
+      "step": 8347
+    },
+    {
+      "epoch": 2.2161157573343955,
+      "grad_norm": 0.43544551354561456,
+      "learning_rate": 3.5170117084546797e-06,
+      "loss": 0.5493,
+      "step": 8348
+    },
+    {
+      "epoch": 2.216381255807779,
+      "grad_norm": 0.4626442194391668,
+      "learning_rate": 3.5166927648927267e-06,
+      "loss": 0.5628,
+      "step": 8349
+    },
+    {
+      "epoch": 2.2166467542811628,
+      "grad_norm": 0.44386423900366967,
+      "learning_rate": 3.516373801502776e-06,
+      "loss": 0.576,
+      "step": 8350
+    },
+    {
+      "epoch": 2.216912252754547,
+      "grad_norm": 0.4262693813730813,
+      "learning_rate": 3.516054818291048e-06,
+      "loss": 0.5561,
+      "step": 8351
+    },
+    {
+      "epoch": 2.2171777512279305,
+      "grad_norm": 0.42048196950630246,
+      "learning_rate": 3.5157358152637623e-06,
+      "loss": 0.5693,
+      "step": 8352
+    },
+    {
+      "epoch": 2.217443249701314,
+      "grad_norm": 0.4368012279693426,
+      "learning_rate": 3.515416792427142e-06,
+      "loss": 0.5329,
+      "step": 8353
+    },
+    {
+      "epoch": 2.217708748174698,
+      "grad_norm": 0.42705886435995977,
+      "learning_rate": 3.5150977497874084e-06,
+      "loss": 0.5439,
+      "step": 8354
+    },
+    {
+      "epoch": 2.217974246648082,
+      "grad_norm": 0.43666914858250333,
+      "learning_rate": 3.5147786873507823e-06,
+      "loss": 0.5806,
+      "step": 8355
+    },
+    {
+      "epoch": 2.2182397451214655,
+      "grad_norm": 0.42809537616294396,
+      "learning_rate": 3.5144596051234875e-06,
+      "loss": 0.5895,
+      "step": 8356
+    },
+    {
+      "epoch": 2.218505243594849,
+      "grad_norm": 0.42874888281055884,
+      "learning_rate": 3.514140503111747e-06,
+      "loss": 0.5556,
+      "step": 8357
+    },
+    {
+      "epoch": 2.218770742068233,
+      "grad_norm": 0.4456059422685089,
+      "learning_rate": 3.5138213813217832e-06,
+      "loss": 0.508,
+      "step": 8358
+    },
+    {
+      "epoch": 2.219036240541617,
+      "grad_norm": 0.42721250452036336,
+      "learning_rate": 3.513502239759821e-06,
+      "loss": 0.5671,
+      "step": 8359
+    },
+    {
+      "epoch": 2.2193017390150005,
+      "grad_norm": 0.41772426094133414,
+      "learning_rate": 3.5131830784320824e-06,
+      "loss": 0.5502,
+      "step": 8360
+    },
+    {
+      "epoch": 2.2195672374883846,
+      "grad_norm": 0.43000282293241876,
+      "learning_rate": 3.5128638973447937e-06,
+      "loss": 0.5762,
+      "step": 8361
+    },
+    {
+      "epoch": 2.219832735961768,
+      "grad_norm": 0.4356162598757294,
+      "learning_rate": 3.512544696504179e-06,
+      "loss": 0.5539,
+      "step": 8362
+    },
+    {
+      "epoch": 2.220098234435152,
+      "grad_norm": 0.4340345631306146,
+      "learning_rate": 3.5122254759164644e-06,
+      "loss": 0.5562,
+      "step": 8363
+    },
+    {
+      "epoch": 2.220363732908536,
+      "grad_norm": 0.44443487464485515,
+      "learning_rate": 3.511906235587873e-06,
+      "loss": 0.5678,
+      "step": 8364
+    },
+    {
+      "epoch": 2.2206292313819196,
+      "grad_norm": 0.428294238127621,
+      "learning_rate": 3.511586975524634e-06,
+      "loss": 0.5414,
+      "step": 8365
+    },
+    {
+      "epoch": 2.2208947298553032,
+      "grad_norm": 0.43479708468158773,
+      "learning_rate": 3.511267695732971e-06,
+      "loss": 0.5721,
+      "step": 8366
+    },
+    {
+      "epoch": 2.2211602283286873,
+      "grad_norm": 0.4546812245249009,
+      "learning_rate": 3.5109483962191125e-06,
+      "loss": 0.5738,
+      "step": 8367
+    },
+    {
+      "epoch": 2.221425726802071,
+      "grad_norm": 0.4437203104892749,
+      "learning_rate": 3.5106290769892853e-06,
+      "loss": 0.5647,
+      "step": 8368
+    },
+    {
+      "epoch": 2.2216912252754546,
+      "grad_norm": 0.4295385350730987,
+      "learning_rate": 3.5103097380497162e-06,
+      "loss": 0.5914,
+      "step": 8369
+    },
+    {
+      "epoch": 2.2219567237488382,
+      "grad_norm": 0.4074080012729898,
+      "learning_rate": 3.5099903794066336e-06,
+      "loss": 0.5078,
+      "step": 8370
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.4344864084178576,
+      "learning_rate": 3.509671001066266e-06,
+      "loss": 0.5773,
+      "step": 8371
+    },
+    {
+      "epoch": 2.222487720695606,
+      "grad_norm": 0.4316399878979075,
+      "learning_rate": 3.509351603034842e-06,
+      "loss": 0.5633,
+      "step": 8372
+    },
+    {
+      "epoch": 2.2227532191689896,
+      "grad_norm": 0.44313824441204963,
+      "learning_rate": 3.5090321853185894e-06,
+      "loss": 0.5839,
+      "step": 8373
+    },
+    {
+      "epoch": 2.2230187176423737,
+      "grad_norm": 0.4293862377099834,
+      "learning_rate": 3.50871274792374e-06,
+      "loss": 0.5664,
+      "step": 8374
+    },
+    {
+      "epoch": 2.2232842161157573,
+      "grad_norm": 0.4379194728789382,
+      "learning_rate": 3.508393290856521e-06,
+      "loss": 0.5422,
+      "step": 8375
+    },
+    {
+      "epoch": 2.223549714589141,
+      "grad_norm": 0.42577918884594324,
+      "learning_rate": 3.5080738141231645e-06,
+      "loss": 0.5579,
+      "step": 8376
+    },
+    {
+      "epoch": 2.223815213062525,
+      "grad_norm": 0.44540219764276706,
+      "learning_rate": 3.5077543177299e-06,
+      "loss": 0.5511,
+      "step": 8377
+    },
+    {
+      "epoch": 2.2240807115359087,
+      "grad_norm": 0.4264588727183936,
+      "learning_rate": 3.5074348016829597e-06,
+      "loss": 0.565,
+      "step": 8378
+    },
+    {
+      "epoch": 2.2243462100092923,
+      "grad_norm": 0.4316595244470432,
+      "learning_rate": 3.507115265988574e-06,
+      "loss": 0.5746,
+      "step": 8379
+    },
+    {
+      "epoch": 2.2246117084826764,
+      "grad_norm": 0.4235960961538575,
+      "learning_rate": 3.506795710652974e-06,
+      "loss": 0.5833,
+      "step": 8380
+    },
+    {
+      "epoch": 2.22487720695606,
+      "grad_norm": 0.44094632257166927,
+      "learning_rate": 3.5064761356823928e-06,
+      "loss": 0.6209,
+      "step": 8381
+    },
+    {
+      "epoch": 2.2251427054294437,
+      "grad_norm": 0.43893112092169484,
+      "learning_rate": 3.5061565410830632e-06,
+      "loss": 0.5618,
+      "step": 8382
+    },
+    {
+      "epoch": 2.225408203902828,
+      "grad_norm": 0.43910983505882245,
+      "learning_rate": 3.5058369268612177e-06,
+      "loss": 0.5783,
+      "step": 8383
+    },
+    {
+      "epoch": 2.2256737023762114,
+      "grad_norm": 0.40921972468896206,
+      "learning_rate": 3.5055172930230885e-06,
+      "loss": 0.5603,
+      "step": 8384
+    },
+    {
+      "epoch": 2.225939200849595,
+      "grad_norm": 0.41199965500912983,
+      "learning_rate": 3.5051976395749104e-06,
+      "loss": 0.5544,
+      "step": 8385
+    },
+    {
+      "epoch": 2.2262046993229787,
+      "grad_norm": 0.4363581060915745,
+      "learning_rate": 3.504877966522917e-06,
+      "loss": 0.5532,
+      "step": 8386
+    },
+    {
+      "epoch": 2.226470197796363,
+      "grad_norm": 0.4241502466672741,
+      "learning_rate": 3.5045582738733435e-06,
+      "loss": 0.5946,
+      "step": 8387
+    },
+    {
+      "epoch": 2.2267356962697464,
+      "grad_norm": 0.42268410126011297,
+      "learning_rate": 3.5042385616324243e-06,
+      "loss": 0.5882,
+      "step": 8388
+    },
+    {
+      "epoch": 2.22700119474313,
+      "grad_norm": 0.41587079027519275,
+      "learning_rate": 3.5039188298063936e-06,
+      "loss": 0.5764,
+      "step": 8389
+    },
+    {
+      "epoch": 2.227266693216514,
+      "grad_norm": 0.43309899004353225,
+      "learning_rate": 3.5035990784014877e-06,
+      "loss": 0.5546,
+      "step": 8390
+    },
+    {
+      "epoch": 2.227532191689898,
+      "grad_norm": 0.43092205022114005,
+      "learning_rate": 3.5032793074239434e-06,
+      "loss": 0.5938,
+      "step": 8391
+    },
+    {
+      "epoch": 2.2277976901632814,
+      "grad_norm": 0.4399142930259134,
+      "learning_rate": 3.5029595168799955e-06,
+      "loss": 0.6215,
+      "step": 8392
+    },
+    {
+      "epoch": 2.2280631886366655,
+      "grad_norm": 0.42700584132387315,
+      "learning_rate": 3.502639706775881e-06,
+      "loss": 0.5604,
+      "step": 8393
+    },
+    {
+      "epoch": 2.228328687110049,
+      "grad_norm": 0.4218062317008268,
+      "learning_rate": 3.502319877117838e-06,
+      "loss": 0.5358,
+      "step": 8394
+    },
+    {
+      "epoch": 2.228594185583433,
+      "grad_norm": 0.4020563061483536,
+      "learning_rate": 3.502000027912103e-06,
+      "loss": 0.5333,
+      "step": 8395
+    },
+    {
+      "epoch": 2.2288596840568164,
+      "grad_norm": 0.4173276985188629,
+      "learning_rate": 3.501680159164914e-06,
+      "loss": 0.5708,
+      "step": 8396
+    },
+    {
+      "epoch": 2.2291251825302005,
+      "grad_norm": 0.4260525630596005,
+      "learning_rate": 3.5013602708825102e-06,
+      "loss": 0.5484,
+      "step": 8397
+    },
+    {
+      "epoch": 2.229390681003584,
+      "grad_norm": 0.4279832470834054,
+      "learning_rate": 3.5010403630711288e-06,
+      "loss": 0.5622,
+      "step": 8398
+    },
+    {
+      "epoch": 2.229656179476968,
+      "grad_norm": 0.4326158327036861,
+      "learning_rate": 3.5007204357370088e-06,
+      "loss": 0.5992,
+      "step": 8399
+    },
+    {
+      "epoch": 2.229921677950352,
+      "grad_norm": 0.44446908935812274,
+      "learning_rate": 3.500400488886391e-06,
+      "loss": 0.5598,
+      "step": 8400
+    },
+    {
+      "epoch": 2.2301871764237355,
+      "grad_norm": 0.44431110876211555,
+      "learning_rate": 3.5000805225255137e-06,
+      "loss": 0.5503,
+      "step": 8401
+    },
+    {
+      "epoch": 2.230452674897119,
+      "grad_norm": 0.4267874879091476,
+      "learning_rate": 3.499760536660618e-06,
+      "loss": 0.5321,
+      "step": 8402
+    },
+    {
+      "epoch": 2.2307181733705033,
+      "grad_norm": 0.43779091362781397,
+      "learning_rate": 3.4994405312979435e-06,
+      "loss": 0.5884,
+      "step": 8403
+    },
+    {
+      "epoch": 2.230983671843887,
+      "grad_norm": 0.43100017949898217,
+      "learning_rate": 3.4991205064437315e-06,
+      "loss": 0.5664,
+      "step": 8404
+    },
+    {
+      "epoch": 2.2312491703172705,
+      "grad_norm": 0.43140197524130475,
+      "learning_rate": 3.4988004621042245e-06,
+      "loss": 0.5873,
+      "step": 8405
+    },
+    {
+      "epoch": 2.2315146687906546,
+      "grad_norm": 0.4383920729388004,
+      "learning_rate": 3.4984803982856617e-06,
+      "loss": 0.5821,
+      "step": 8406
+    },
+    {
+      "epoch": 2.2317801672640383,
+      "grad_norm": 0.43647495617701904,
+      "learning_rate": 3.4981603149942867e-06,
+      "loss": 0.5725,
+      "step": 8407
+    },
+    {
+      "epoch": 2.232045665737422,
+      "grad_norm": 0.42724851875804976,
+      "learning_rate": 3.4978402122363425e-06,
+      "loss": 0.5707,
+      "step": 8408
+    },
+    {
+      "epoch": 2.232311164210806,
+      "grad_norm": 0.4366552943288066,
+      "learning_rate": 3.49752009001807e-06,
+      "loss": 0.5336,
+      "step": 8409
+    },
+    {
+      "epoch": 2.2325766626841896,
+      "grad_norm": 0.42268126300556375,
+      "learning_rate": 3.4971999483457144e-06,
+      "loss": 0.5606,
+      "step": 8410
+    },
+    {
+      "epoch": 2.2328421611575733,
+      "grad_norm": 0.4260872511726609,
+      "learning_rate": 3.496879787225518e-06,
+      "loss": 0.6096,
+      "step": 8411
+    },
+    {
+      "epoch": 2.233107659630957,
+      "grad_norm": 0.4169713035669326,
+      "learning_rate": 3.4965596066637235e-06,
+      "loss": 0.6025,
+      "step": 8412
+    },
+    {
+      "epoch": 2.233373158104341,
+      "grad_norm": 0.43080271560299715,
+      "learning_rate": 3.4962394066665778e-06,
+      "loss": 0.5622,
+      "step": 8413
+    },
+    {
+      "epoch": 2.2336386565777246,
+      "grad_norm": 0.4316169795396076,
+      "learning_rate": 3.495919187240325e-06,
+      "loss": 0.5627,
+      "step": 8414
+    },
+    {
+      "epoch": 2.2339041550511083,
+      "grad_norm": 0.43616716114427484,
+      "learning_rate": 3.4955989483912086e-06,
+      "loss": 0.5786,
+      "step": 8415
+    },
+    {
+      "epoch": 2.2341696535244924,
+      "grad_norm": 0.4428562149461648,
+      "learning_rate": 3.4952786901254754e-06,
+      "loss": 0.5784,
+      "step": 8416
+    },
+    {
+      "epoch": 2.234435151997876,
+      "grad_norm": 0.4447422787496083,
+      "learning_rate": 3.4949584124493712e-06,
+      "loss": 0.5511,
+      "step": 8417
+    },
+    {
+      "epoch": 2.2347006504712597,
+      "grad_norm": 0.4202045889292342,
+      "learning_rate": 3.4946381153691415e-06,
+      "loss": 0.5561,
+      "step": 8418
+    },
+    {
+      "epoch": 2.2349661489446437,
+      "grad_norm": 0.44377552102135,
+      "learning_rate": 3.494317798891034e-06,
+      "loss": 0.5705,
+      "step": 8419
+    },
+    {
+      "epoch": 2.2352316474180274,
+      "grad_norm": 0.4368560959490654,
+      "learning_rate": 3.493997463021294e-06,
+      "loss": 0.5848,
+      "step": 8420
+    },
+    {
+      "epoch": 2.235497145891411,
+      "grad_norm": 0.44062910768878577,
+      "learning_rate": 3.4936771077661703e-06,
+      "loss": 0.602,
+      "step": 8421
+    },
+    {
+      "epoch": 2.235762644364795,
+      "grad_norm": 0.4334399380264875,
+      "learning_rate": 3.4933567331319092e-06,
+      "loss": 0.5582,
+      "step": 8422
+    },
+    {
+      "epoch": 2.2360281428381787,
+      "grad_norm": 0.43395549333355654,
+      "learning_rate": 3.493036339124761e-06,
+      "loss": 0.5852,
+      "step": 8423
+    },
+    {
+      "epoch": 2.2362936413115624,
+      "grad_norm": 0.4236788773462707,
+      "learning_rate": 3.4927159257509715e-06,
+      "loss": 0.5992,
+      "step": 8424
+    },
+    {
+      "epoch": 2.236559139784946,
+      "grad_norm": 0.42910042262345455,
+      "learning_rate": 3.4923954930167914e-06,
+      "loss": 0.5663,
+      "step": 8425
+    },
+    {
+      "epoch": 2.23682463825833,
+      "grad_norm": 0.43484517357838604,
+      "learning_rate": 3.4920750409284697e-06,
+      "loss": 0.5856,
+      "step": 8426
+    },
+    {
+      "epoch": 2.2370901367317138,
+      "grad_norm": 0.4392414055927907,
+      "learning_rate": 3.4917545694922558e-06,
+      "loss": 0.6267,
+      "step": 8427
+    },
+    {
+      "epoch": 2.2373556352050974,
+      "grad_norm": 0.44459950273924503,
+      "learning_rate": 3.491434078714399e-06,
+      "loss": 0.5432,
+      "step": 8428
+    },
+    {
+      "epoch": 2.2376211336784815,
+      "grad_norm": 0.4510047451879924,
+      "learning_rate": 3.4911135686011503e-06,
+      "loss": 0.5169,
+      "step": 8429
+    },
+    {
+      "epoch": 2.237886632151865,
+      "grad_norm": 0.42857302178086804,
+      "learning_rate": 3.4907930391587606e-06,
+      "loss": 0.5865,
+      "step": 8430
+    },
+    {
+      "epoch": 2.2381521306252488,
+      "grad_norm": 0.4354733485790215,
+      "learning_rate": 3.4904724903934805e-06,
+      "loss": 0.5992,
+      "step": 8431
+    },
+    {
+      "epoch": 2.238417629098633,
+      "grad_norm": 0.4808390444400206,
+      "learning_rate": 3.4901519223115615e-06,
+      "loss": 0.552,
+      "step": 8432
+    },
+    {
+      "epoch": 2.2386831275720165,
+      "grad_norm": 0.487107469131134,
+      "learning_rate": 3.489831334919256e-06,
+      "loss": 0.5911,
+      "step": 8433
+    },
+    {
+      "epoch": 2.2389486260454,
+      "grad_norm": 0.42860446638123045,
+      "learning_rate": 3.4895107282228164e-06,
+      "loss": 0.5573,
+      "step": 8434
+    },
+    {
+      "epoch": 2.239214124518784,
+      "grad_norm": 0.43268622255302314,
+      "learning_rate": 3.4891901022284946e-06,
+      "loss": 0.5883,
+      "step": 8435
+    },
+    {
+      "epoch": 2.239479622992168,
+      "grad_norm": 0.43744792851396774,
+      "learning_rate": 3.4888694569425434e-06,
+      "loss": 0.5458,
+      "step": 8436
+    },
+    {
+      "epoch": 2.2397451214655515,
+      "grad_norm": 0.4313242608517905,
+      "learning_rate": 3.4885487923712174e-06,
+      "loss": 0.5748,
+      "step": 8437
+    },
+    {
+      "epoch": 2.2400106199389356,
+      "grad_norm": 0.4447843445986176,
+      "learning_rate": 3.488228108520769e-06,
+      "loss": 0.5502,
+      "step": 8438
+    },
+    {
+      "epoch": 2.240276118412319,
+      "grad_norm": 0.4386784978531559,
+      "learning_rate": 3.4879074053974526e-06,
+      "loss": 0.5464,
+      "step": 8439
+    },
+    {
+      "epoch": 2.240541616885703,
+      "grad_norm": 0.42300779698232555,
+      "learning_rate": 3.4875866830075233e-06,
+      "loss": 0.5665,
+      "step": 8440
+    },
+    {
+      "epoch": 2.2408071153590865,
+      "grad_norm": 0.43618717888098046,
+      "learning_rate": 3.4872659413572358e-06,
+      "loss": 0.552,
+      "step": 8441
+    },
+    {
+      "epoch": 2.2410726138324706,
+      "grad_norm": 0.4421362805617268,
+      "learning_rate": 3.486945180452845e-06,
+      "loss": 0.5928,
+      "step": 8442
+    },
+    {
+      "epoch": 2.2413381123058542,
+      "grad_norm": 0.4360589976311463,
+      "learning_rate": 3.486624400300607e-06,
+      "loss": 0.5553,
+      "step": 8443
+    },
+    {
+      "epoch": 2.241603610779238,
+      "grad_norm": 0.4283297760567781,
+      "learning_rate": 3.4863036009067775e-06,
+      "loss": 0.577,
+      "step": 8444
+    },
+    {
+      "epoch": 2.241869109252622,
+      "grad_norm": 0.4304774597555228,
+      "learning_rate": 3.4859827822776127e-06,
+      "loss": 0.5724,
+      "step": 8445
+    },
+    {
+      "epoch": 2.2421346077260056,
+      "grad_norm": 0.41816395848491394,
+      "learning_rate": 3.4856619444193694e-06,
+      "loss": 0.5676,
+      "step": 8446
+    },
+    {
+      "epoch": 2.2424001061993892,
+      "grad_norm": 0.42472783781238577,
+      "learning_rate": 3.4853410873383054e-06,
+      "loss": 0.5883,
+      "step": 8447
+    },
+    {
+      "epoch": 2.2426656046727733,
+      "grad_norm": 0.4252652818311552,
+      "learning_rate": 3.4850202110406773e-06,
+      "loss": 0.5832,
+      "step": 8448
+    },
+    {
+      "epoch": 2.242931103146157,
+      "grad_norm": 0.4270459141998556,
+      "learning_rate": 3.484699315532743e-06,
+      "loss": 0.5913,
+      "step": 8449
+    },
+    {
+      "epoch": 2.2431966016195406,
+      "grad_norm": 0.4189839876235439,
+      "learning_rate": 3.4843784008207614e-06,
+      "loss": 0.5721,
+      "step": 8450
+    },
+    {
+      "epoch": 2.2434621000929242,
+      "grad_norm": 0.43433407007206165,
+      "learning_rate": 3.4840574669109906e-06,
+      "loss": 0.5918,
+      "step": 8451
+    },
+    {
+      "epoch": 2.2437275985663083,
+      "grad_norm": 0.43364892076480527,
+      "learning_rate": 3.4837365138096906e-06,
+      "loss": 0.4985,
+      "step": 8452
+    },
+    {
+      "epoch": 2.243993097039692,
+      "grad_norm": 0.43427856578928115,
+      "learning_rate": 3.48341554152312e-06,
+      "loss": 0.5889,
+      "step": 8453
+    },
+    {
+      "epoch": 2.2442585955130756,
+      "grad_norm": 0.4432576798781629,
+      "learning_rate": 3.4830945500575373e-06,
+      "loss": 0.5798,
+      "step": 8454
+    },
+    {
+      "epoch": 2.2445240939864597,
+      "grad_norm": 0.4343245018863736,
+      "learning_rate": 3.482773539419205e-06,
+      "loss": 0.5774,
+      "step": 8455
+    },
+    {
+      "epoch": 2.2447895924598433,
+      "grad_norm": 0.43857212956487013,
+      "learning_rate": 3.4824525096143815e-06,
+      "loss": 0.5666,
+      "step": 8456
+    },
+    {
+      "epoch": 2.245055090933227,
+      "grad_norm": 0.4307807419805649,
+      "learning_rate": 3.4821314606493294e-06,
+      "loss": 0.5721,
+      "step": 8457
+    },
+    {
+      "epoch": 2.245320589406611,
+      "grad_norm": 0.4305290261024878,
+      "learning_rate": 3.481810392530308e-06,
+      "loss": 0.5545,
+      "step": 8458
+    },
+    {
+      "epoch": 2.2455860878799947,
+      "grad_norm": 0.4095779568327294,
+      "learning_rate": 3.481489305263581e-06,
+      "loss": 0.5491,
+      "step": 8459
+    },
+    {
+      "epoch": 2.2458515863533783,
+      "grad_norm": 0.4282508920274349,
+      "learning_rate": 3.4811681988554095e-06,
+      "loss": 0.562,
+      "step": 8460
+    },
+    {
+      "epoch": 2.2461170848267624,
+      "grad_norm": 0.42366582840713984,
+      "learning_rate": 3.480847073312056e-06,
+      "loss": 0.567,
+      "step": 8461
+    },
+    {
+      "epoch": 2.246382583300146,
+      "grad_norm": 0.4218657841167782,
+      "learning_rate": 3.4805259286397826e-06,
+      "loss": 0.5412,
+      "step": 8462
+    },
+    {
+      "epoch": 2.2466480817735297,
+      "grad_norm": 0.44488732581491075,
+      "learning_rate": 3.480204764844853e-06,
+      "loss": 0.5924,
+      "step": 8463
+    },
+    {
+      "epoch": 2.246913580246914,
+      "grad_norm": 0.4207054773093573,
+      "learning_rate": 3.4798835819335302e-06,
+      "loss": 0.5459,
+      "step": 8464
+    },
+    {
+      "epoch": 2.2471790787202974,
+      "grad_norm": 0.43785040096124667,
+      "learning_rate": 3.479562379912079e-06,
+      "loss": 0.5719,
+      "step": 8465
+    },
+    {
+      "epoch": 2.247444577193681,
+      "grad_norm": 0.42374473645492894,
+      "learning_rate": 3.4792411587867624e-06,
+      "loss": 0.5255,
+      "step": 8466
+    },
+    {
+      "epoch": 2.2477100756670647,
+      "grad_norm": 0.42372921167139677,
+      "learning_rate": 3.478919918563846e-06,
+      "loss": 0.5785,
+      "step": 8467
+    },
+    {
+      "epoch": 2.247975574140449,
+      "grad_norm": 0.4335511307988954,
+      "learning_rate": 3.4785986592495934e-06,
+      "loss": 0.5913,
+      "step": 8468
+    },
+    {
+      "epoch": 2.2482410726138324,
+      "grad_norm": 0.4164595517929875,
+      "learning_rate": 3.4782773808502723e-06,
+      "loss": 0.541,
+      "step": 8469
+    },
+    {
+      "epoch": 2.248506571087216,
+      "grad_norm": 0.42150552189659524,
+      "learning_rate": 3.4779560833721465e-06,
+      "loss": 0.5736,
+      "step": 8470
+    },
+    {
+      "epoch": 2.2487720695606,
+      "grad_norm": 0.42220024200623407,
+      "learning_rate": 3.4776347668214827e-06,
+      "loss": 0.5525,
+      "step": 8471
+    },
+    {
+      "epoch": 2.249037568033984,
+      "grad_norm": 0.4240358801598312,
+      "learning_rate": 3.4773134312045477e-06,
+      "loss": 0.5735,
+      "step": 8472
+    },
+    {
+      "epoch": 2.2493030665073674,
+      "grad_norm": 0.419433254850018,
+      "learning_rate": 3.4769920765276066e-06,
+      "loss": 0.5596,
+      "step": 8473
+    },
+    {
+      "epoch": 2.2495685649807515,
+      "grad_norm": 0.42030087495296353,
+      "learning_rate": 3.4766707027969287e-06,
+      "loss": 0.5292,
+      "step": 8474
+    },
+    {
+      "epoch": 2.249834063454135,
+      "grad_norm": 0.4360122013527655,
+      "learning_rate": 3.4763493100187813e-06,
+      "loss": 0.5923,
+      "step": 8475
+    },
+    {
+      "epoch": 2.250099561927519,
+      "grad_norm": 0.4383475295171756,
+      "learning_rate": 3.476027898199431e-06,
+      "loss": 0.5642,
+      "step": 8476
+    },
+    {
+      "epoch": 2.2503650604009025,
+      "grad_norm": 0.43993794768516414,
+      "learning_rate": 3.4757064673451476e-06,
+      "loss": 0.5548,
+      "step": 8477
+    },
+    {
+      "epoch": 2.2506305588742865,
+      "grad_norm": 0.4317676688889174,
+      "learning_rate": 3.4753850174621983e-06,
+      "loss": 0.5622,
+      "step": 8478
+    },
+    {
+      "epoch": 2.25089605734767,
+      "grad_norm": 0.43157829445075135,
+      "learning_rate": 3.475063548556854e-06,
+      "loss": 0.5484,
+      "step": 8479
+    },
+    {
+      "epoch": 2.251161555821054,
+      "grad_norm": 0.43973720973354885,
+      "learning_rate": 3.4747420606353828e-06,
+      "loss": 0.539,
+      "step": 8480
+    },
+    {
+      "epoch": 2.251427054294438,
+      "grad_norm": 0.4373353112850105,
+      "learning_rate": 3.4744205537040547e-06,
+      "loss": 0.5857,
+      "step": 8481
+    },
+    {
+      "epoch": 2.2516925527678215,
+      "grad_norm": 0.4322246048265512,
+      "learning_rate": 3.47409902776914e-06,
+      "loss": 0.5892,
+      "step": 8482
+    },
+    {
+      "epoch": 2.251958051241205,
+      "grad_norm": 0.4308019965712213,
+      "learning_rate": 3.4737774828369096e-06,
+      "loss": 0.5533,
+      "step": 8483
+    },
+    {
+      "epoch": 2.2522235497145893,
+      "grad_norm": 0.4274043010276875,
+      "learning_rate": 3.4734559189136337e-06,
+      "loss": 0.5349,
+      "step": 8484
+    },
+    {
+      "epoch": 2.252489048187973,
+      "grad_norm": 0.44018431430350496,
+      "learning_rate": 3.4731343360055835e-06,
+      "loss": 0.5583,
+      "step": 8485
+    },
+    {
+      "epoch": 2.2527545466613565,
+      "grad_norm": 0.44283943872646336,
+      "learning_rate": 3.472812734119032e-06,
+      "loss": 0.5221,
+      "step": 8486
+    },
+    {
+      "epoch": 2.2530200451347406,
+      "grad_norm": 0.42531913118141246,
+      "learning_rate": 3.4724911132602497e-06,
+      "loss": 0.5842,
+      "step": 8487
+    },
+    {
+      "epoch": 2.2532855436081243,
+      "grad_norm": 0.431922880906971,
+      "learning_rate": 3.4721694734355097e-06,
+      "loss": 0.5612,
+      "step": 8488
+    },
+    {
+      "epoch": 2.253551042081508,
+      "grad_norm": 0.44233742161464473,
+      "learning_rate": 3.4718478146510854e-06,
+      "loss": 0.5788,
+      "step": 8489
+    },
+    {
+      "epoch": 2.253816540554892,
+      "grad_norm": 0.435427261646667,
+      "learning_rate": 3.4715261369132486e-06,
+      "loss": 0.5779,
+      "step": 8490
+    },
+    {
+      "epoch": 2.2540820390282756,
+      "grad_norm": 0.43381754174498666,
+      "learning_rate": 3.471204440228273e-06,
+      "loss": 0.5239,
+      "step": 8491
+    },
+    {
+      "epoch": 2.2543475375016593,
+      "grad_norm": 0.4282838580195437,
+      "learning_rate": 3.470882724602434e-06,
+      "loss": 0.5169,
+      "step": 8492
+    },
+    {
+      "epoch": 2.2546130359750434,
+      "grad_norm": 0.43671304507945535,
+      "learning_rate": 3.470560990042004e-06,
+      "loss": 0.5639,
+      "step": 8493
+    },
+    {
+      "epoch": 2.254878534448427,
+      "grad_norm": 0.4189419053178972,
+      "learning_rate": 3.4702392365532585e-06,
+      "loss": 0.5359,
+      "step": 8494
+    },
+    {
+      "epoch": 2.2551440329218106,
+      "grad_norm": 0.42895566325218754,
+      "learning_rate": 3.469917464142472e-06,
+      "loss": 0.5482,
+      "step": 8495
+    },
+    {
+      "epoch": 2.2554095313951943,
+      "grad_norm": 0.4488541438777685,
+      "learning_rate": 3.4695956728159208e-06,
+      "loss": 0.5922,
+      "step": 8496
+    },
+    {
+      "epoch": 2.2556750298685784,
+      "grad_norm": 0.4415913834987696,
+      "learning_rate": 3.4692738625798794e-06,
+      "loss": 0.5406,
+      "step": 8497
+    },
+    {
+      "epoch": 2.255940528341962,
+      "grad_norm": 0.4240872696874829,
+      "learning_rate": 3.468952033440625e-06,
+      "loss": 0.5512,
+      "step": 8498
+    },
+    {
+      "epoch": 2.2562060268153457,
+      "grad_norm": 0.430669657574303,
+      "learning_rate": 3.4686301854044335e-06,
+      "loss": 0.5602,
+      "step": 8499
+    },
+    {
+      "epoch": 2.2564715252887297,
+      "grad_norm": 0.4277164877732677,
+      "learning_rate": 3.468308318477582e-06,
+      "loss": 0.5646,
+      "step": 8500
+    },
+    {
+      "epoch": 2.2567370237621134,
+      "grad_norm": 0.4432112090962664,
+      "learning_rate": 3.467986432666347e-06,
+      "loss": 0.5514,
+      "step": 8501
+    },
+    {
+      "epoch": 2.257002522235497,
+      "grad_norm": 0.4293313486671919,
+      "learning_rate": 3.467664527977006e-06,
+      "loss": 0.5716,
+      "step": 8502
+    },
+    {
+      "epoch": 2.257268020708881,
+      "grad_norm": 0.4214129461685631,
+      "learning_rate": 3.4673426044158383e-06,
+      "loss": 0.5567,
+      "step": 8503
+    },
+    {
+      "epoch": 2.2575335191822647,
+      "grad_norm": 0.42392837142685924,
+      "learning_rate": 3.467020661989121e-06,
+      "loss": 0.5541,
+      "step": 8504
+    },
+    {
+      "epoch": 2.2577990176556484,
+      "grad_norm": 0.4413780549515595,
+      "learning_rate": 3.4666987007031328e-06,
+      "loss": 0.566,
+      "step": 8505
+    },
+    {
+      "epoch": 2.258064516129032,
+      "grad_norm": 0.4276301304509837,
+      "learning_rate": 3.4663767205641542e-06,
+      "loss": 0.5415,
+      "step": 8506
+    },
+    {
+      "epoch": 2.258330014602416,
+      "grad_norm": 0.4236588998327288,
+      "learning_rate": 3.4660547215784622e-06,
+      "loss": 0.5871,
+      "step": 8507
+    },
+    {
+      "epoch": 2.2585955130757998,
+      "grad_norm": 0.4341673134835704,
+      "learning_rate": 3.465732703752338e-06,
+      "loss": 0.5618,
+      "step": 8508
+    },
+    {
+      "epoch": 2.2588610115491834,
+      "grad_norm": 0.4134262277626446,
+      "learning_rate": 3.465410667092063e-06,
+      "loss": 0.5583,
+      "step": 8509
+    },
+    {
+      "epoch": 2.2591265100225675,
+      "grad_norm": 0.4382664778368452,
+      "learning_rate": 3.4650886116039145e-06,
+      "loss": 0.552,
+      "step": 8510
+    },
+    {
+      "epoch": 2.259392008495951,
+      "grad_norm": 0.41936945793274105,
+      "learning_rate": 3.4647665372941757e-06,
+      "loss": 0.5453,
+      "step": 8511
+    },
+    {
+      "epoch": 2.2596575069693348,
+      "grad_norm": 0.42806826429789735,
+      "learning_rate": 3.4644444441691276e-06,
+      "loss": 0.5861,
+      "step": 8512
+    },
+    {
+      "epoch": 2.259923005442719,
+      "grad_norm": 0.43602259223378537,
+      "learning_rate": 3.4641223322350513e-06,
+      "loss": 0.5837,
+      "step": 8513
+    },
+    {
+      "epoch": 2.2601885039161025,
+      "grad_norm": 0.4220217377826108,
+      "learning_rate": 3.4638002014982286e-06,
+      "loss": 0.5752,
+      "step": 8514
+    },
+    {
+      "epoch": 2.260454002389486,
+      "grad_norm": 0.42791868265428396,
+      "learning_rate": 3.4634780519649425e-06,
+      "loss": 0.5585,
+      "step": 8515
+    },
+    {
+      "epoch": 2.26071950086287,
+      "grad_norm": 0.4402367817049422,
+      "learning_rate": 3.4631558836414753e-06,
+      "loss": 0.5341,
+      "step": 8516
+    },
+    {
+      "epoch": 2.260984999336254,
+      "grad_norm": 0.4230092355063566,
+      "learning_rate": 3.46283369653411e-06,
+      "loss": 0.5485,
+      "step": 8517
+    },
+    {
+      "epoch": 2.2612504978096375,
+      "grad_norm": 0.4331083590954356,
+      "learning_rate": 3.4625114906491307e-06,
+      "loss": 0.5384,
+      "step": 8518
+    },
+    {
+      "epoch": 2.2615159962830216,
+      "grad_norm": 0.44429589678732834,
+      "learning_rate": 3.4621892659928204e-06,
+      "loss": 0.5528,
+      "step": 8519
+    },
+    {
+      "epoch": 2.261781494756405,
+      "grad_norm": 0.42342356479441795,
+      "learning_rate": 3.4618670225714633e-06,
+      "loss": 0.5763,
+      "step": 8520
+    },
+    {
+      "epoch": 2.262046993229789,
+      "grad_norm": 0.43625814354014386,
+      "learning_rate": 3.461544760391345e-06,
+      "loss": 0.5576,
+      "step": 8521
+    },
+    {
+      "epoch": 2.262312491703173,
+      "grad_norm": 0.4445498637903862,
+      "learning_rate": 3.461222479458749e-06,
+      "loss": 0.5693,
+      "step": 8522
+    },
+    {
+      "epoch": 2.2625779901765566,
+      "grad_norm": 0.4393516661996358,
+      "learning_rate": 3.4609001797799607e-06,
+      "loss": 0.5693,
+      "step": 8523
+    },
+    {
+      "epoch": 2.2628434886499402,
+      "grad_norm": 0.43706053978286297,
+      "learning_rate": 3.4605778613612668e-06,
+      "loss": 0.5801,
+      "step": 8524
+    },
+    {
+      "epoch": 2.263108987123324,
+      "grad_norm": 0.434759604905694,
+      "learning_rate": 3.4602555242089526e-06,
+      "loss": 0.5842,
+      "step": 8525
+    },
+    {
+      "epoch": 2.263374485596708,
+      "grad_norm": 0.43038273658022436,
+      "learning_rate": 3.4599331683293045e-06,
+      "loss": 0.5477,
+      "step": 8526
+    },
+    {
+      "epoch": 2.2636399840700916,
+      "grad_norm": 0.43737407037190174,
+      "learning_rate": 3.4596107937286092e-06,
+      "loss": 0.6033,
+      "step": 8527
+    },
+    {
+      "epoch": 2.2639054825434752,
+      "grad_norm": 0.4261202057507047,
+      "learning_rate": 3.459288400413154e-06,
+      "loss": 0.5599,
+      "step": 8528
+    },
+    {
+      "epoch": 2.2641709810168593,
+      "grad_norm": 0.4285000501431998,
+      "learning_rate": 3.458965988389227e-06,
+      "loss": 0.5613,
+      "step": 8529
+    },
+    {
+      "epoch": 2.264436479490243,
+      "grad_norm": 0.43507289319758524,
+      "learning_rate": 3.4586435576631143e-06,
+      "loss": 0.5365,
+      "step": 8530
+    },
+    {
+      "epoch": 2.2647019779636266,
+      "grad_norm": 0.4157881800965343,
+      "learning_rate": 3.4583211082411057e-06,
+      "loss": 0.5314,
+      "step": 8531
+    },
+    {
+      "epoch": 2.2649674764370102,
+      "grad_norm": 0.4263637061781523,
+      "learning_rate": 3.4579986401294884e-06,
+      "loss": 0.5594,
+      "step": 8532
+    },
+    {
+      "epoch": 2.2652329749103943,
+      "grad_norm": 0.439673133839381,
+      "learning_rate": 3.4576761533345528e-06,
+      "loss": 0.6089,
+      "step": 8533
+    },
+    {
+      "epoch": 2.265498473383778,
+      "grad_norm": 0.4199750838291158,
+      "learning_rate": 3.4573536478625864e-06,
+      "loss": 0.5481,
+      "step": 8534
+    },
+    {
+      "epoch": 2.2657639718571616,
+      "grad_norm": 0.42777124943943723,
+      "learning_rate": 3.457031123719881e-06,
+      "loss": 0.5753,
+      "step": 8535
+    },
+    {
+      "epoch": 2.2660294703305457,
+      "grad_norm": 0.4254042268643367,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.574,
+      "step": 8536
+    },
+    {
+      "epoch": 2.2662949688039293,
+      "grad_norm": 0.4287111813410994,
+      "learning_rate": 3.4563860194474086e-06,
+      "loss": 0.5487,
+      "step": 8537
+    },
+    {
+      "epoch": 2.266560467277313,
+      "grad_norm": 0.4332528353055577,
+      "learning_rate": 3.4560634393302246e-06,
+      "loss": 0.5527,
+      "step": 8538
+    },
+    {
+      "epoch": 2.266825965750697,
+      "grad_norm": 0.4381917553079303,
+      "learning_rate": 3.455740840567462e-06,
+      "loss": 0.535,
+      "step": 8539
+    },
+    {
+      "epoch": 2.2670914642240807,
+      "grad_norm": 0.44533968851103484,
+      "learning_rate": 3.455418223165412e-06,
+      "loss": 0.5692,
+      "step": 8540
+    },
+    {
+      "epoch": 2.2673569626974643,
+      "grad_norm": 0.4359696149096592,
+      "learning_rate": 3.455095587130368e-06,
+      "loss": 0.5792,
+      "step": 8541
+    },
+    {
+      "epoch": 2.2676224611708484,
+      "grad_norm": 0.4389674947402785,
+      "learning_rate": 3.454772932468622e-06,
+      "loss": 0.605,
+      "step": 8542
+    },
+    {
+      "epoch": 2.267887959644232,
+      "grad_norm": 0.43349729198072795,
+      "learning_rate": 3.454450259186466e-06,
+      "loss": 0.5919,
+      "step": 8543
+    },
+    {
+      "epoch": 2.2681534581176157,
+      "grad_norm": 0.43716179248404385,
+      "learning_rate": 3.454127567290193e-06,
+      "loss": 0.5726,
+      "step": 8544
+    },
+    {
+      "epoch": 2.268418956591,
+      "grad_norm": 0.42688664797397036,
+      "learning_rate": 3.4538048567860967e-06,
+      "loss": 0.5961,
+      "step": 8545
+    },
+    {
+      "epoch": 2.2686844550643834,
+      "grad_norm": 0.44654345742528595,
+      "learning_rate": 3.4534821276804695e-06,
+      "loss": 0.5828,
+      "step": 8546
+    },
+    {
+      "epoch": 2.268949953537767,
+      "grad_norm": 0.42852625289224566,
+      "learning_rate": 3.453159379979607e-06,
+      "loss": 0.5608,
+      "step": 8547
+    },
+    {
+      "epoch": 2.269215452011151,
+      "grad_norm": 0.4183610378943299,
+      "learning_rate": 3.452836613689803e-06,
+      "loss": 0.573,
+      "step": 8548
+    },
+    {
+      "epoch": 2.269480950484535,
+      "grad_norm": 0.4309301722327577,
+      "learning_rate": 3.452513828817351e-06,
+      "loss": 0.5735,
+      "step": 8549
+    },
+    {
+      "epoch": 2.2697464489579184,
+      "grad_norm": 0.43225726530680536,
+      "learning_rate": 3.452191025368548e-06,
+      "loss": 0.5525,
+      "step": 8550
+    },
+    {
+      "epoch": 2.270011947431302,
+      "grad_norm": 0.45698451243453,
+      "learning_rate": 3.4518682033496886e-06,
+      "loss": 0.5299,
+      "step": 8551
+    },
+    {
+      "epoch": 2.270277445904686,
+      "grad_norm": 0.4143690839663798,
+      "learning_rate": 3.4515453627670687e-06,
+      "loss": 0.578,
+      "step": 8552
+    },
+    {
+      "epoch": 2.27054294437807,
+      "grad_norm": 0.42611166357054314,
+      "learning_rate": 3.451222503626984e-06,
+      "loss": 0.4983,
+      "step": 8553
+    },
+    {
+      "epoch": 2.2708084428514534,
+      "grad_norm": 0.42408471335708403,
+      "learning_rate": 3.4508996259357318e-06,
+      "loss": 0.5485,
+      "step": 8554
+    },
+    {
+      "epoch": 2.2710739413248375,
+      "grad_norm": 0.4369268750561684,
+      "learning_rate": 3.4505767296996088e-06,
+      "loss": 0.5571,
+      "step": 8555
+    },
+    {
+      "epoch": 2.271339439798221,
+      "grad_norm": 0.4372695574096089,
+      "learning_rate": 3.450253814924912e-06,
+      "loss": 0.5493,
+      "step": 8556
+    },
+    {
+      "epoch": 2.271604938271605,
+      "grad_norm": 0.43797303587833813,
+      "learning_rate": 3.4499308816179392e-06,
+      "loss": 0.5751,
+      "step": 8557
+    },
+    {
+      "epoch": 2.271870436744989,
+      "grad_norm": 0.42989007327076006,
+      "learning_rate": 3.4496079297849893e-06,
+      "loss": 0.5453,
+      "step": 8558
+    },
+    {
+      "epoch": 2.2721359352183725,
+      "grad_norm": 0.4255473478142183,
+      "learning_rate": 3.449284959432358e-06,
+      "loss": 0.5875,
+      "step": 8559
+    },
+    {
+      "epoch": 2.272401433691756,
+      "grad_norm": 0.4372962381254977,
+      "learning_rate": 3.4489619705663464e-06,
+      "loss": 0.5452,
+      "step": 8560
+    },
+    {
+      "epoch": 2.27266693216514,
+      "grad_norm": 0.4357449121348786,
+      "learning_rate": 3.448638963193254e-06,
+      "loss": 0.5488,
+      "step": 8561
+    },
+    {
+      "epoch": 2.272932430638524,
+      "grad_norm": 0.4434780067866217,
+      "learning_rate": 3.4483159373193774e-06,
+      "loss": 0.5736,
+      "step": 8562
+    },
+    {
+      "epoch": 2.2731979291119075,
+      "grad_norm": 0.443178473120969,
+      "learning_rate": 3.4479928929510193e-06,
+      "loss": 0.5862,
+      "step": 8563
+    },
+    {
+      "epoch": 2.273463427585291,
+      "grad_norm": 0.42475222041609967,
+      "learning_rate": 3.447669830094479e-06,
+      "loss": 0.5803,
+      "step": 8564
+    },
+    {
+      "epoch": 2.2737289260586753,
+      "grad_norm": 0.4244862617112801,
+      "learning_rate": 3.4473467487560554e-06,
+      "loss": 0.5427,
+      "step": 8565
+    },
+    {
+      "epoch": 2.273994424532059,
+      "grad_norm": 0.4366460223429077,
+      "learning_rate": 3.447023648942052e-06,
+      "loss": 0.5953,
+      "step": 8566
+    },
+    {
+      "epoch": 2.2742599230054426,
+      "grad_norm": 0.43361233423863266,
+      "learning_rate": 3.446700530658768e-06,
+      "loss": 0.5684,
+      "step": 8567
+    },
+    {
+      "epoch": 2.2745254214788266,
+      "grad_norm": 0.42507294357960795,
+      "learning_rate": 3.4463773939125054e-06,
+      "loss": 0.5459,
+      "step": 8568
+    },
+    {
+      "epoch": 2.2747909199522103,
+      "grad_norm": 0.4223656834111791,
+      "learning_rate": 3.446054238709567e-06,
+      "loss": 0.5578,
+      "step": 8569
+    },
+    {
+      "epoch": 2.275056418425594,
+      "grad_norm": 0.43245622414523766,
+      "learning_rate": 3.4457310650562547e-06,
+      "loss": 0.58,
+      "step": 8570
+    },
+    {
+      "epoch": 2.275321916898978,
+      "grad_norm": 0.43190269287009964,
+      "learning_rate": 3.4454078729588708e-06,
+      "loss": 0.5553,
+      "step": 8571
+    },
+    {
+      "epoch": 2.2755874153723616,
+      "grad_norm": 0.431527965383785,
+      "learning_rate": 3.4450846624237184e-06,
+      "loss": 0.5685,
+      "step": 8572
+    },
+    {
+      "epoch": 2.2758529138457453,
+      "grad_norm": 0.45065090750922304,
+      "learning_rate": 3.4447614334571024e-06,
+      "loss": 0.5467,
+      "step": 8573
+    },
+    {
+      "epoch": 2.2761184123191294,
+      "grad_norm": 0.42685577016948234,
+      "learning_rate": 3.444438186065324e-06,
+      "loss": 0.5475,
+      "step": 8574
+    },
+    {
+      "epoch": 2.276383910792513,
+      "grad_norm": 0.45191910317104145,
+      "learning_rate": 3.4441149202546894e-06,
+      "loss": 0.579,
+      "step": 8575
+    },
+    {
+      "epoch": 2.2766494092658967,
+      "grad_norm": 0.4414055368028246,
+      "learning_rate": 3.4437916360315017e-06,
+      "loss": 0.5439,
+      "step": 8576
+    },
+    {
+      "epoch": 2.2769149077392807,
+      "grad_norm": 0.4437368098276584,
+      "learning_rate": 3.4434683334020657e-06,
+      "loss": 0.5611,
+      "step": 8577
+    },
+    {
+      "epoch": 2.2771804062126644,
+      "grad_norm": 0.43344869596613866,
+      "learning_rate": 3.4431450123726876e-06,
+      "loss": 0.565,
+      "step": 8578
+    },
+    {
+      "epoch": 2.277445904686048,
+      "grad_norm": 0.4236041179202667,
+      "learning_rate": 3.442821672949673e-06,
+      "loss": 0.5995,
+      "step": 8579
+    },
+    {
+      "epoch": 2.2777114031594317,
+      "grad_norm": 0.4388511237577311,
+      "learning_rate": 3.442498315139327e-06,
+      "loss": 0.5836,
+      "step": 8580
+    },
+    {
+      "epoch": 2.2779769016328157,
+      "grad_norm": 0.4194323193136079,
+      "learning_rate": 3.442174938947957e-06,
+      "loss": 0.5642,
+      "step": 8581
+    },
+    {
+      "epoch": 2.2782424001061994,
+      "grad_norm": 0.43122242117002885,
+      "learning_rate": 3.441851544381868e-06,
+      "loss": 0.5286,
+      "step": 8582
+    },
+    {
+      "epoch": 2.278507898579583,
+      "grad_norm": 0.42498320223986963,
+      "learning_rate": 3.441528131447368e-06,
+      "loss": 0.5555,
+      "step": 8583
+    },
+    {
+      "epoch": 2.278773397052967,
+      "grad_norm": 0.4250188446642847,
+      "learning_rate": 3.441204700150765e-06,
+      "loss": 0.5799,
+      "step": 8584
+    },
+    {
+      "epoch": 2.2790388955263507,
+      "grad_norm": 0.4285009818149595,
+      "learning_rate": 3.4408812504983645e-06,
+      "loss": 0.5889,
+      "step": 8585
+    },
+    {
+      "epoch": 2.2793043939997344,
+      "grad_norm": 0.4214319270597252,
+      "learning_rate": 3.4405577824964766e-06,
+      "loss": 0.5495,
+      "step": 8586
+    },
+    {
+      "epoch": 2.279569892473118,
+      "grad_norm": 0.43573256983697223,
+      "learning_rate": 3.440234296151409e-06,
+      "loss": 0.5928,
+      "step": 8587
+    },
+    {
+      "epoch": 2.279835390946502,
+      "grad_norm": 0.43407718567501175,
+      "learning_rate": 3.4399107914694707e-06,
+      "loss": 0.5652,
+      "step": 8588
+    },
+    {
+      "epoch": 2.2801008894198858,
+      "grad_norm": 0.4169491778406323,
+      "learning_rate": 3.4395872684569705e-06,
+      "loss": 0.5425,
+      "step": 8589
+    },
+    {
+      "epoch": 2.2803663878932694,
+      "grad_norm": 0.4521944568205286,
+      "learning_rate": 3.4392637271202178e-06,
+      "loss": 0.5803,
+      "step": 8590
+    },
+    {
+      "epoch": 2.2806318863666535,
+      "grad_norm": 0.4420948090859669,
+      "learning_rate": 3.4389401674655233e-06,
+      "loss": 0.5565,
+      "step": 8591
+    },
+    {
+      "epoch": 2.280897384840037,
+      "grad_norm": 0.4501262073979795,
+      "learning_rate": 3.4386165894991963e-06,
+      "loss": 0.5963,
+      "step": 8592
+    },
+    {
+      "epoch": 2.2811628833134208,
+      "grad_norm": 0.4249577146559445,
+      "learning_rate": 3.438292993227548e-06,
+      "loss": 0.5601,
+      "step": 8593
+    },
+    {
+      "epoch": 2.281428381786805,
+      "grad_norm": 0.43647541297813564,
+      "learning_rate": 3.4379693786568885e-06,
+      "loss": 0.5446,
+      "step": 8594
+    },
+    {
+      "epoch": 2.2816938802601885,
+      "grad_norm": 0.434658706721523,
+      "learning_rate": 3.43764574579353e-06,
+      "loss": 0.6062,
+      "step": 8595
+    },
+    {
+      "epoch": 2.281959378733572,
+      "grad_norm": 0.4639231751581402,
+      "learning_rate": 3.437322094643783e-06,
+      "loss": 0.549,
+      "step": 8596
+    },
+    {
+      "epoch": 2.282224877206956,
+      "grad_norm": 0.42841938257273166,
+      "learning_rate": 3.4369984252139605e-06,
+      "loss": 0.5652,
+      "step": 8597
+    },
+    {
+      "epoch": 2.28249037568034,
+      "grad_norm": 0.42476446939944607,
+      "learning_rate": 3.436674737510375e-06,
+      "loss": 0.547,
+      "step": 8598
+    },
+    {
+      "epoch": 2.2827558741537235,
+      "grad_norm": 0.43468818186699665,
+      "learning_rate": 3.4363510315393382e-06,
+      "loss": 0.5938,
+      "step": 8599
+    },
+    {
+      "epoch": 2.2830213726271076,
+      "grad_norm": 0.4374351682629712,
+      "learning_rate": 3.4360273073071636e-06,
+      "loss": 0.5903,
+      "step": 8600
+    },
+    {
+      "epoch": 2.2832868711004912,
+      "grad_norm": 0.4350099905812807,
+      "learning_rate": 3.4357035648201657e-06,
+      "loss": 0.6187,
+      "step": 8601
+    },
+    {
+      "epoch": 2.283552369573875,
+      "grad_norm": 0.43445281702511274,
+      "learning_rate": 3.4353798040846565e-06,
+      "loss": 0.578,
+      "step": 8602
+    },
+    {
+      "epoch": 2.283817868047259,
+      "grad_norm": 0.4281480416533931,
+      "learning_rate": 3.435056025106951e-06,
+      "loss": 0.5608,
+      "step": 8603
+    },
+    {
+      "epoch": 2.2840833665206426,
+      "grad_norm": 0.43789655630573504,
+      "learning_rate": 3.4347322278933644e-06,
+      "loss": 0.5882,
+      "step": 8604
+    },
+    {
+      "epoch": 2.2843488649940262,
+      "grad_norm": 0.4408960361267119,
+      "learning_rate": 3.434408412450209e-06,
+      "loss": 0.6032,
+      "step": 8605
+    },
+    {
+      "epoch": 2.28461436346741,
+      "grad_norm": 0.42620606987120185,
+      "learning_rate": 3.434084578783803e-06,
+      "loss": 0.5428,
+      "step": 8606
+    },
+    {
+      "epoch": 2.284879861940794,
+      "grad_norm": 0.41326492382555574,
+      "learning_rate": 3.43376072690046e-06,
+      "loss": 0.5693,
+      "step": 8607
+    },
+    {
+      "epoch": 2.2851453604141776,
+      "grad_norm": 0.44303866425971955,
+      "learning_rate": 3.4334368568064972e-06,
+      "loss": 0.6044,
+      "step": 8608
+    },
+    {
+      "epoch": 2.2854108588875612,
+      "grad_norm": 0.4211387302490756,
+      "learning_rate": 3.4331129685082306e-06,
+      "loss": 0.5569,
+      "step": 8609
+    },
+    {
+      "epoch": 2.2856763573609453,
+      "grad_norm": 0.45927611180523964,
+      "learning_rate": 3.4327890620119757e-06,
+      "loss": 0.5817,
+      "step": 8610
+    },
+    {
+      "epoch": 2.285941855834329,
+      "grad_norm": 0.4450881733344187,
+      "learning_rate": 3.4324651373240505e-06,
+      "loss": 0.5973,
+      "step": 8611
+    },
+    {
+      "epoch": 2.2862073543077126,
+      "grad_norm": 0.43432195369427223,
+      "learning_rate": 3.432141194450772e-06,
+      "loss": 0.5505,
+      "step": 8612
+    },
+    {
+      "epoch": 2.2864728527810967,
+      "grad_norm": 0.4431155234226894,
+      "learning_rate": 3.431817233398458e-06,
+      "loss": 0.5402,
+      "step": 8613
+    },
+    {
+      "epoch": 2.2867383512544803,
+      "grad_norm": 0.421946847119039,
+      "learning_rate": 3.4314932541734266e-06,
+      "loss": 0.5376,
+      "step": 8614
+    },
+    {
+      "epoch": 2.287003849727864,
+      "grad_norm": 0.44036042759908406,
+      "learning_rate": 3.431169256781995e-06,
+      "loss": 0.5905,
+      "step": 8615
+    },
+    {
+      "epoch": 2.2872693482012476,
+      "grad_norm": 0.45149330075274924,
+      "learning_rate": 3.4308452412304844e-06,
+      "loss": 0.5225,
+      "step": 8616
+    },
+    {
+      "epoch": 2.2875348466746317,
+      "grad_norm": 0.43353348919696133,
+      "learning_rate": 3.4305212075252116e-06,
+      "loss": 0.5366,
+      "step": 8617
+    },
+    {
+      "epoch": 2.2878003451480153,
+      "grad_norm": 0.4368234726549299,
+      "learning_rate": 3.4301971556724973e-06,
+      "loss": 0.6109,
+      "step": 8618
+    },
+    {
+      "epoch": 2.288065843621399,
+      "grad_norm": 0.4218767484763383,
+      "learning_rate": 3.429873085678661e-06,
+      "loss": 0.5336,
+      "step": 8619
+    },
+    {
+      "epoch": 2.288331342094783,
+      "grad_norm": 0.453800116753957,
+      "learning_rate": 3.429548997550023e-06,
+      "loss": 0.5206,
+      "step": 8620
+    },
+    {
+      "epoch": 2.2885968405681667,
+      "grad_norm": 0.43003754974788116,
+      "learning_rate": 3.429224891292904e-06,
+      "loss": 0.5879,
+      "step": 8621
+    },
+    {
+      "epoch": 2.2888623390415503,
+      "grad_norm": 0.43622186090912646,
+      "learning_rate": 3.4289007669136232e-06,
+      "loss": 0.6098,
+      "step": 8622
+    },
+    {
+      "epoch": 2.2891278375149344,
+      "grad_norm": 0.4350969956625951,
+      "learning_rate": 3.4285766244185037e-06,
+      "loss": 0.5456,
+      "step": 8623
+    },
+    {
+      "epoch": 2.289393335988318,
+      "grad_norm": 0.4302341183850153,
+      "learning_rate": 3.4282524638138664e-06,
+      "loss": 0.576,
+      "step": 8624
+    },
+    {
+      "epoch": 2.2896588344617017,
+      "grad_norm": 0.4346534193047658,
+      "learning_rate": 3.4279282851060335e-06,
+      "loss": 0.5868,
+      "step": 8625
+    },
+    {
+      "epoch": 2.289924332935086,
+      "grad_norm": 0.4346917869148051,
+      "learning_rate": 3.427604088301327e-06,
+      "loss": 0.5611,
+      "step": 8626
+    },
+    {
+      "epoch": 2.2901898314084694,
+      "grad_norm": 0.42406552900214967,
+      "learning_rate": 3.4272798734060696e-06,
+      "loss": 0.5873,
+      "step": 8627
+    },
+    {
+      "epoch": 2.290455329881853,
+      "grad_norm": 0.4669286594933157,
+      "learning_rate": 3.426955640426584e-06,
+      "loss": 0.5503,
+      "step": 8628
+    },
+    {
+      "epoch": 2.290720828355237,
+      "grad_norm": 0.45354687841586233,
+      "learning_rate": 3.4266313893691943e-06,
+      "loss": 0.5736,
+      "step": 8629
+    },
+    {
+      "epoch": 2.290986326828621,
+      "grad_norm": 0.4216700143556463,
+      "learning_rate": 3.426307120240224e-06,
+      "loss": 0.555,
+      "step": 8630
+    },
+    {
+      "epoch": 2.2912518253020044,
+      "grad_norm": 0.43172386065401885,
+      "learning_rate": 3.4259828330459966e-06,
+      "loss": 0.5085,
+      "step": 8631
+    },
+    {
+      "epoch": 2.2915173237753885,
+      "grad_norm": 0.42946655508038667,
+      "learning_rate": 3.425658527792836e-06,
+      "loss": 0.5523,
+      "step": 8632
+    },
+    {
+      "epoch": 2.291782822248772,
+      "grad_norm": 0.42809528809244657,
+      "learning_rate": 3.4253342044870685e-06,
+      "loss": 0.5778,
+      "step": 8633
+    },
+    {
+      "epoch": 2.292048320722156,
+      "grad_norm": 0.42884103648316607,
+      "learning_rate": 3.425009863135018e-06,
+      "loss": 0.5355,
+      "step": 8634
+    },
+    {
+      "epoch": 2.2923138191955394,
+      "grad_norm": 0.4282181011970201,
+      "learning_rate": 3.4246855037430106e-06,
+      "loss": 0.5502,
+      "step": 8635
+    },
+    {
+      "epoch": 2.2925793176689235,
+      "grad_norm": 0.43129714818790477,
+      "learning_rate": 3.4243611263173722e-06,
+      "loss": 0.5813,
+      "step": 8636
+    },
+    {
+      "epoch": 2.292844816142307,
+      "grad_norm": 0.4231203719941036,
+      "learning_rate": 3.424036730864428e-06,
+      "loss": 0.616,
+      "step": 8637
+    },
+    {
+      "epoch": 2.293110314615691,
+      "grad_norm": 0.4402166801138924,
+      "learning_rate": 3.423712317390505e-06,
+      "loss": 0.6032,
+      "step": 8638
+    },
+    {
+      "epoch": 2.293375813089075,
+      "grad_norm": 0.42350621013555106,
+      "learning_rate": 3.423387885901931e-06,
+      "loss": 0.5668,
+      "step": 8639
+    },
+    {
+      "epoch": 2.2936413115624585,
+      "grad_norm": 0.44008722164441844,
+      "learning_rate": 3.423063436405032e-06,
+      "loss": 0.5304,
+      "step": 8640
+    },
+    {
+      "epoch": 2.293906810035842,
+      "grad_norm": 0.4348090628063169,
+      "learning_rate": 3.4227389689061353e-06,
+      "loss": 0.5347,
+      "step": 8641
+    },
+    {
+      "epoch": 2.2941723085092263,
+      "grad_norm": 0.43435880921020203,
+      "learning_rate": 3.42241448341157e-06,
+      "loss": 0.5339,
+      "step": 8642
+    },
+    {
+      "epoch": 2.29443780698261,
+      "grad_norm": 0.4248996835375181,
+      "learning_rate": 3.4220899799276637e-06,
+      "loss": 0.5258,
+      "step": 8643
+    },
+    {
+      "epoch": 2.2947033054559935,
+      "grad_norm": 0.44058834309212397,
+      "learning_rate": 3.421765458460745e-06,
+      "loss": 0.5642,
+      "step": 8644
+    },
+    {
+      "epoch": 2.294968803929377,
+      "grad_norm": 0.41431904305329476,
+      "learning_rate": 3.421440919017144e-06,
+      "loss": 0.6193,
+      "step": 8645
+    },
+    {
+      "epoch": 2.2952343024027613,
+      "grad_norm": 0.43134524200929175,
+      "learning_rate": 3.421116361603188e-06,
+      "loss": 0.5652,
+      "step": 8646
+    },
+    {
+      "epoch": 2.295499800876145,
+      "grad_norm": 0.4338152491012483,
+      "learning_rate": 3.4207917862252083e-06,
+      "loss": 0.554,
+      "step": 8647
+    },
+    {
+      "epoch": 2.2957652993495286,
+      "grad_norm": 0.41761404718399825,
+      "learning_rate": 3.4204671928895334e-06,
+      "loss": 0.5778,
+      "step": 8648
+    },
+    {
+      "epoch": 2.2960307978229126,
+      "grad_norm": 0.43331192347395864,
+      "learning_rate": 3.420142581602495e-06,
+      "loss": 0.5746,
+      "step": 8649
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 0.45860675274287865,
+      "learning_rate": 3.4198179523704234e-06,
+      "loss": 0.5473,
+      "step": 8650
+    },
+    {
+      "epoch": 2.29656179476968,
+      "grad_norm": 0.4180147352218744,
+      "learning_rate": 3.4194933051996493e-06,
+      "loss": 0.5824,
+      "step": 8651
+    },
+    {
+      "epoch": 2.296827293243064,
+      "grad_norm": 0.44604411099370106,
+      "learning_rate": 3.4191686400965046e-06,
+      "loss": 0.5759,
+      "step": 8652
+    },
+    {
+      "epoch": 2.2970927917164476,
+      "grad_norm": 0.448373937487176,
+      "learning_rate": 3.4188439570673216e-06,
+      "loss": 0.5644,
+      "step": 8653
+    },
+    {
+      "epoch": 2.2973582901898313,
+      "grad_norm": 0.43927625938094816,
+      "learning_rate": 3.4185192561184312e-06,
+      "loss": 0.5697,
+      "step": 8654
+    },
+    {
+      "epoch": 2.2976237886632154,
+      "grad_norm": 0.43609999003164884,
+      "learning_rate": 3.4181945372561658e-06,
+      "loss": 0.5955,
+      "step": 8655
+    },
+    {
+      "epoch": 2.297889287136599,
+      "grad_norm": 0.4385595759633102,
+      "learning_rate": 3.4178698004868598e-06,
+      "loss": 0.5134,
+      "step": 8656
+    },
+    {
+      "epoch": 2.2981547856099827,
+      "grad_norm": 0.4437820845501664,
+      "learning_rate": 3.417545045816845e-06,
+      "loss": 0.5554,
+      "step": 8657
+    },
+    {
+      "epoch": 2.2984202840833667,
+      "grad_norm": 0.422961155923829,
+      "learning_rate": 3.417220273252455e-06,
+      "loss": 0.5582,
+      "step": 8658
+    },
+    {
+      "epoch": 2.2986857825567504,
+      "grad_norm": 0.43192448818067,
+      "learning_rate": 3.4168954828000245e-06,
+      "loss": 0.5735,
+      "step": 8659
+    },
+    {
+      "epoch": 2.298951281030134,
+      "grad_norm": 0.4449759670096079,
+      "learning_rate": 3.4165706744658865e-06,
+      "loss": 0.5706,
+      "step": 8660
+    },
+    {
+      "epoch": 2.299216779503518,
+      "grad_norm": 0.4405244493681199,
+      "learning_rate": 3.4162458482563756e-06,
+      "loss": 0.5407,
+      "step": 8661
+    },
+    {
+      "epoch": 2.2994822779769017,
+      "grad_norm": 0.4195382959296605,
+      "learning_rate": 3.4159210041778284e-06,
+      "loss": 0.5551,
+      "step": 8662
+    },
+    {
+      "epoch": 2.2997477764502854,
+      "grad_norm": 0.42201630037548943,
+      "learning_rate": 3.415596142236579e-06,
+      "loss": 0.568,
+      "step": 8663
+    },
+    {
+      "epoch": 2.300013274923669,
+      "grad_norm": 0.4298967016439161,
+      "learning_rate": 3.415271262438963e-06,
+      "loss": 0.5496,
+      "step": 8664
+    },
+    {
+      "epoch": 2.300278773397053,
+      "grad_norm": 0.43154018803061883,
+      "learning_rate": 3.414946364791316e-06,
+      "loss": 0.5635,
+      "step": 8665
+    },
+    {
+      "epoch": 2.3005442718704368,
+      "grad_norm": 0.4438106209256733,
+      "learning_rate": 3.414621449299975e-06,
+      "loss": 0.5941,
+      "step": 8666
+    },
+    {
+      "epoch": 2.3008097703438204,
+      "grad_norm": 0.4330238135245737,
+      "learning_rate": 3.414296515971276e-06,
+      "loss": 0.606,
+      "step": 8667
+    },
+    {
+      "epoch": 2.3010752688172045,
+      "grad_norm": 0.41537796047406034,
+      "learning_rate": 3.4139715648115575e-06,
+      "loss": 0.5432,
+      "step": 8668
+    },
+    {
+      "epoch": 2.301340767290588,
+      "grad_norm": 0.4199013477576837,
+      "learning_rate": 3.4136465958271547e-06,
+      "loss": 0.5402,
+      "step": 8669
+    },
+    {
+      "epoch": 2.3016062657639718,
+      "grad_norm": 0.42145896929941684,
+      "learning_rate": 3.4133216090244063e-06,
+      "loss": 0.5432,
+      "step": 8670
+    },
+    {
+      "epoch": 2.3018717642373554,
+      "grad_norm": 0.43515668343426206,
+      "learning_rate": 3.4129966044096503e-06,
+      "loss": 0.5988,
+      "step": 8671
+    },
+    {
+      "epoch": 2.3021372627107395,
+      "grad_norm": 0.43223548861228345,
+      "learning_rate": 3.4126715819892253e-06,
+      "loss": 0.5859,
+      "step": 8672
+    },
+    {
+      "epoch": 2.302402761184123,
+      "grad_norm": 0.43430799411907095,
+      "learning_rate": 3.4123465417694702e-06,
+      "loss": 0.5983,
+      "step": 8673
+    },
+    {
+      "epoch": 2.3026682596575068,
+      "grad_norm": 0.430194111538804,
+      "learning_rate": 3.4120214837567234e-06,
+      "loss": 0.5887,
+      "step": 8674
+    },
+    {
+      "epoch": 2.302933758130891,
+      "grad_norm": 0.42061516824386097,
+      "learning_rate": 3.4116964079573243e-06,
+      "loss": 0.5985,
+      "step": 8675
+    },
+    {
+      "epoch": 2.3031992566042745,
+      "grad_norm": 0.4296039745269631,
+      "learning_rate": 3.411371314377614e-06,
+      "loss": 0.55,
+      "step": 8676
+    },
+    {
+      "epoch": 2.303464755077658,
+      "grad_norm": 0.43966332243547074,
+      "learning_rate": 3.4110462030239304e-06,
+      "loss": 0.5702,
+      "step": 8677
+    },
+    {
+      "epoch": 2.303730253551042,
+      "grad_norm": 0.43771975287124654,
+      "learning_rate": 3.410721073902616e-06,
+      "loss": 0.5628,
+      "step": 8678
+    },
+    {
+      "epoch": 2.303995752024426,
+      "grad_norm": 0.4331116302408522,
+      "learning_rate": 3.4103959270200105e-06,
+      "loss": 0.6011,
+      "step": 8679
+    },
+    {
+      "epoch": 2.3042612504978095,
+      "grad_norm": 0.44030659592738497,
+      "learning_rate": 3.410070762382455e-06,
+      "loss": 0.592,
+      "step": 8680
+    },
+    {
+      "epoch": 2.3045267489711936,
+      "grad_norm": 0.4358715104606606,
+      "learning_rate": 3.409745579996292e-06,
+      "loss": 0.5768,
+      "step": 8681
+    },
+    {
+      "epoch": 2.3047922474445772,
+      "grad_norm": 0.4284271462370547,
+      "learning_rate": 3.4094203798678634e-06,
+      "loss": 0.5646,
+      "step": 8682
+    },
+    {
+      "epoch": 2.305057745917961,
+      "grad_norm": 0.4309236091956606,
+      "learning_rate": 3.4090951620035094e-06,
+      "loss": 0.5095,
+      "step": 8683
+    },
+    {
+      "epoch": 2.305323244391345,
+      "grad_norm": 0.42184894141151874,
+      "learning_rate": 3.4087699264095746e-06,
+      "loss": 0.5498,
+      "step": 8684
+    },
+    {
+      "epoch": 2.3055887428647286,
+      "grad_norm": 0.42655281270623524,
+      "learning_rate": 3.4084446730924016e-06,
+      "loss": 0.5355,
+      "step": 8685
+    },
+    {
+      "epoch": 2.3058542413381122,
+      "grad_norm": 0.42470084582982187,
+      "learning_rate": 3.408119402058333e-06,
+      "loss": 0.5791,
+      "step": 8686
+    },
+    {
+      "epoch": 2.3061197398114963,
+      "grad_norm": 0.4371712161975838,
+      "learning_rate": 3.407794113313712e-06,
+      "loss": 0.5354,
+      "step": 8687
+    },
+    {
+      "epoch": 2.30638523828488,
+      "grad_norm": 0.4483303382792808,
+      "learning_rate": 3.407468806864883e-06,
+      "loss": 0.527,
+      "step": 8688
+    },
+    {
+      "epoch": 2.3066507367582636,
+      "grad_norm": 0.4428717600591457,
+      "learning_rate": 3.407143482718191e-06,
+      "loss": 0.5527,
+      "step": 8689
+    },
+    {
+      "epoch": 2.3069162352316472,
+      "grad_norm": 0.4344103689174707,
+      "learning_rate": 3.40681814087998e-06,
+      "loss": 0.5927,
+      "step": 8690
+    },
+    {
+      "epoch": 2.3071817337050313,
+      "grad_norm": 0.42864000328790075,
+      "learning_rate": 3.4064927813565952e-06,
+      "loss": 0.5431,
+      "step": 8691
+    },
+    {
+      "epoch": 2.307447232178415,
+      "grad_norm": 0.4278265465540942,
+      "learning_rate": 3.406167404154381e-06,
+      "loss": 0.5556,
+      "step": 8692
+    },
+    {
+      "epoch": 2.3077127306517986,
+      "grad_norm": 0.446265498518479,
+      "learning_rate": 3.405842009279684e-06,
+      "loss": 0.6127,
+      "step": 8693
+    },
+    {
+      "epoch": 2.3079782291251827,
+      "grad_norm": 0.4281659300570047,
+      "learning_rate": 3.4055165967388502e-06,
+      "loss": 0.5349,
+      "step": 8694
+    },
+    {
+      "epoch": 2.3082437275985663,
+      "grad_norm": 0.4425767952431549,
+      "learning_rate": 3.4051911665382254e-06,
+      "loss": 0.5637,
+      "step": 8695
+    },
+    {
+      "epoch": 2.30850922607195,
+      "grad_norm": 0.4359561735033075,
+      "learning_rate": 3.4048657186841567e-06,
+      "loss": 0.5348,
+      "step": 8696
+    },
+    {
+      "epoch": 2.308774724545334,
+      "grad_norm": 0.44916543480984455,
+      "learning_rate": 3.404540253182991e-06,
+      "loss": 0.5325,
+      "step": 8697
+    },
+    {
+      "epoch": 2.3090402230187177,
+      "grad_norm": 0.4348341393286978,
+      "learning_rate": 3.4042147700410755e-06,
+      "loss": 0.5418,
+      "step": 8698
+    },
+    {
+      "epoch": 2.3093057214921013,
+      "grad_norm": 0.45393446013022204,
+      "learning_rate": 3.4038892692647584e-06,
+      "loss": 0.582,
+      "step": 8699
+    },
+    {
+      "epoch": 2.309571219965485,
+      "grad_norm": 0.4493936187700734,
+      "learning_rate": 3.4035637508603868e-06,
+      "loss": 0.571,
+      "step": 8700
+    },
+    {
+      "epoch": 2.309836718438869,
+      "grad_norm": 0.44204429344420293,
+      "learning_rate": 3.4032382148343103e-06,
+      "loss": 0.6077,
+      "step": 8701
+    },
+    {
+      "epoch": 2.3101022169122527,
+      "grad_norm": 0.44863656295787097,
+      "learning_rate": 3.4029126611928776e-06,
+      "loss": 0.5956,
+      "step": 8702
+    },
+    {
+      "epoch": 2.3103677153856363,
+      "grad_norm": 0.4893470318271801,
+      "learning_rate": 3.4025870899424363e-06,
+      "loss": 0.5306,
+      "step": 8703
+    },
+    {
+      "epoch": 2.3106332138590204,
+      "grad_norm": 0.4335001711334422,
+      "learning_rate": 3.4022615010893374e-06,
+      "loss": 0.5174,
+      "step": 8704
+    },
+    {
+      "epoch": 2.310898712332404,
+      "grad_norm": 0.431490596916666,
+      "learning_rate": 3.4019358946399304e-06,
+      "loss": 0.5549,
+      "step": 8705
+    },
+    {
+      "epoch": 2.3111642108057877,
+      "grad_norm": 0.43424166875333015,
+      "learning_rate": 3.4016102706005645e-06,
+      "loss": 0.5273,
+      "step": 8706
+    },
+    {
+      "epoch": 2.311429709279172,
+      "grad_norm": 0.4217312919572834,
+      "learning_rate": 3.4012846289775905e-06,
+      "loss": 0.5825,
+      "step": 8707
+    },
+    {
+      "epoch": 2.3116952077525554,
+      "grad_norm": 0.42817993624353834,
+      "learning_rate": 3.4009589697773605e-06,
+      "loss": 0.5657,
+      "step": 8708
+    },
+    {
+      "epoch": 2.311960706225939,
+      "grad_norm": 0.43404422005993665,
+      "learning_rate": 3.4006332930062236e-06,
+      "loss": 0.6125,
+      "step": 8709
+    },
+    {
+      "epoch": 2.312226204699323,
+      "grad_norm": 0.4373978590145837,
+      "learning_rate": 3.4003075986705326e-06,
+      "loss": 0.5926,
+      "step": 8710
+    },
+    {
+      "epoch": 2.312491703172707,
+      "grad_norm": 0.4159244910635771,
+      "learning_rate": 3.39998188677664e-06,
+      "loss": 0.5299,
+      "step": 8711
+    },
+    {
+      "epoch": 2.3127572016460904,
+      "grad_norm": 0.4172099619607329,
+      "learning_rate": 3.3996561573308963e-06,
+      "loss": 0.5286,
+      "step": 8712
+    },
+    {
+      "epoch": 2.3130227001194745,
+      "grad_norm": 0.42859753708641873,
+      "learning_rate": 3.3993304103396553e-06,
+      "loss": 0.5692,
+      "step": 8713
+    },
+    {
+      "epoch": 2.313288198592858,
+      "grad_norm": 0.44133138185157095,
+      "learning_rate": 3.399004645809269e-06,
+      "loss": 0.538,
+      "step": 8714
+    },
+    {
+      "epoch": 2.313553697066242,
+      "grad_norm": 0.434876809713245,
+      "learning_rate": 3.398678863746091e-06,
+      "loss": 0.6101,
+      "step": 8715
+    },
+    {
+      "epoch": 2.313819195539626,
+      "grad_norm": 0.4292994856603943,
+      "learning_rate": 3.3983530641564744e-06,
+      "loss": 0.5728,
+      "step": 8716
+    },
+    {
+      "epoch": 2.3140846940130095,
+      "grad_norm": 0.4307938872307361,
+      "learning_rate": 3.398027247046774e-06,
+      "loss": 0.5673,
+      "step": 8717
+    },
+    {
+      "epoch": 2.314350192486393,
+      "grad_norm": 0.4274362234983465,
+      "learning_rate": 3.397701412423343e-06,
+      "loss": 0.5613,
+      "step": 8718
+    },
+    {
+      "epoch": 2.314615690959777,
+      "grad_norm": 0.43440228854944063,
+      "learning_rate": 3.3973755602925374e-06,
+      "loss": 0.5974,
+      "step": 8719
+    },
+    {
+      "epoch": 2.314881189433161,
+      "grad_norm": 0.451912719332572,
+      "learning_rate": 3.397049690660711e-06,
+      "loss": 0.5888,
+      "step": 8720
+    },
+    {
+      "epoch": 2.3151466879065445,
+      "grad_norm": 0.42808434773507886,
+      "learning_rate": 3.3967238035342187e-06,
+      "loss": 0.5625,
+      "step": 8721
+    },
+    {
+      "epoch": 2.315412186379928,
+      "grad_norm": 0.4315311977344077,
+      "learning_rate": 3.3963978989194174e-06,
+      "loss": 0.5745,
+      "step": 8722
+    },
+    {
+      "epoch": 2.3156776848533123,
+      "grad_norm": 0.4251036422424467,
+      "learning_rate": 3.396071976822662e-06,
+      "loss": 0.5397,
+      "step": 8723
+    },
+    {
+      "epoch": 2.315943183326696,
+      "grad_norm": 0.42626404411390983,
+      "learning_rate": 3.395746037250309e-06,
+      "loss": 0.5432,
+      "step": 8724
+    },
+    {
+      "epoch": 2.3162086818000795,
+      "grad_norm": 0.4252503594702319,
+      "learning_rate": 3.3954200802087146e-06,
+      "loss": 0.5624,
+      "step": 8725
+    },
+    {
+      "epoch": 2.316474180273463,
+      "grad_norm": 0.4503302304801565,
+      "learning_rate": 3.3950941057042363e-06,
+      "loss": 0.5335,
+      "step": 8726
+    },
+    {
+      "epoch": 2.3167396787468473,
+      "grad_norm": 0.42929050395089974,
+      "learning_rate": 3.394768113743232e-06,
+      "loss": 0.566,
+      "step": 8727
+    },
+    {
+      "epoch": 2.317005177220231,
+      "grad_norm": 0.440221977185262,
+      "learning_rate": 3.3944421043320593e-06,
+      "loss": 0.5647,
+      "step": 8728
+    },
+    {
+      "epoch": 2.3172706756936146,
+      "grad_norm": 0.41752932526464925,
+      "learning_rate": 3.3941160774770744e-06,
+      "loss": 0.5475,
+      "step": 8729
+    },
+    {
+      "epoch": 2.3175361741669986,
+      "grad_norm": 0.43419345021156086,
+      "learning_rate": 3.3937900331846373e-06,
+      "loss": 0.5807,
+      "step": 8730
+    },
+    {
+      "epoch": 2.3178016726403823,
+      "grad_norm": 0.43546798131950937,
+      "learning_rate": 3.393463971461106e-06,
+      "loss": 0.5887,
+      "step": 8731
+    },
+    {
+      "epoch": 2.318067171113766,
+      "grad_norm": 0.42812839890863225,
+      "learning_rate": 3.39313789231284e-06,
+      "loss": 0.5713,
+      "step": 8732
+    },
+    {
+      "epoch": 2.31833266958715,
+      "grad_norm": 0.4351081451216406,
+      "learning_rate": 3.392811795746198e-06,
+      "loss": 0.5748,
+      "step": 8733
+    },
+    {
+      "epoch": 2.3185981680605336,
+      "grad_norm": 0.4355074277409877,
+      "learning_rate": 3.39248568176754e-06,
+      "loss": 0.5615,
+      "step": 8734
+    },
+    {
+      "epoch": 2.3188636665339173,
+      "grad_norm": 0.4289827679562849,
+      "learning_rate": 3.392159550383226e-06,
+      "loss": 0.5824,
+      "step": 8735
+    },
+    {
+      "epoch": 2.3191291650073014,
+      "grad_norm": 0.4376249382687959,
+      "learning_rate": 3.391833401599617e-06,
+      "loss": 0.5796,
+      "step": 8736
+    },
+    {
+      "epoch": 2.319394663480685,
+      "grad_norm": 0.45386129963571475,
+      "learning_rate": 3.3915072354230728e-06,
+      "loss": 0.5417,
+      "step": 8737
+    },
+    {
+      "epoch": 2.3196601619540687,
+      "grad_norm": 0.42624879960086787,
+      "learning_rate": 3.3911810518599542e-06,
+      "loss": 0.5827,
+      "step": 8738
+    },
+    {
+      "epoch": 2.3199256604274527,
+      "grad_norm": 0.4178959163332132,
+      "learning_rate": 3.3908548509166234e-06,
+      "loss": 0.5216,
+      "step": 8739
+    },
+    {
+      "epoch": 2.3201911589008364,
+      "grad_norm": 0.4458969808068912,
+      "learning_rate": 3.3905286325994423e-06,
+      "loss": 0.5902,
+      "step": 8740
+    },
+    {
+      "epoch": 2.32045665737422,
+      "grad_norm": 0.43466416442710726,
+      "learning_rate": 3.3902023969147723e-06,
+      "loss": 0.568,
+      "step": 8741
+    },
+    {
+      "epoch": 2.320722155847604,
+      "grad_norm": 0.42851178838776194,
+      "learning_rate": 3.3898761438689752e-06,
+      "loss": 0.5611,
+      "step": 8742
+    },
+    {
+      "epoch": 2.3209876543209877,
+      "grad_norm": 0.4258354349653672,
+      "learning_rate": 3.389549873468415e-06,
+      "loss": 0.5541,
+      "step": 8743
+    },
+    {
+      "epoch": 2.3212531527943714,
+      "grad_norm": 0.4342573521016042,
+      "learning_rate": 3.3892235857194545e-06,
+      "loss": 0.6051,
+      "step": 8744
+    },
+    {
+      "epoch": 2.321518651267755,
+      "grad_norm": 0.42099841562692886,
+      "learning_rate": 3.388897280628457e-06,
+      "loss": 0.5597,
+      "step": 8745
+    },
+    {
+      "epoch": 2.321784149741139,
+      "grad_norm": 0.42442378017136545,
+      "learning_rate": 3.3885709582017856e-06,
+      "loss": 0.5621,
+      "step": 8746
+    },
+    {
+      "epoch": 2.3220496482145228,
+      "grad_norm": 0.43274790826031445,
+      "learning_rate": 3.388244618445805e-06,
+      "loss": 0.5808,
+      "step": 8747
+    },
+    {
+      "epoch": 2.3223151466879064,
+      "grad_norm": 0.4229624825508171,
+      "learning_rate": 3.38791826136688e-06,
+      "loss": 0.5514,
+      "step": 8748
+    },
+    {
+      "epoch": 2.3225806451612905,
+      "grad_norm": 0.42607682743538566,
+      "learning_rate": 3.3875918869713744e-06,
+      "loss": 0.58,
+      "step": 8749
+    },
+    {
+      "epoch": 2.322846143634674,
+      "grad_norm": 0.43847846165442683,
+      "learning_rate": 3.387265495265654e-06,
+      "loss": 0.5634,
+      "step": 8750
+    },
+    {
+      "epoch": 2.3231116421080578,
+      "grad_norm": 0.4270452036960777,
+      "learning_rate": 3.3869390862560846e-06,
+      "loss": 0.547,
+      "step": 8751
+    },
+    {
+      "epoch": 2.323377140581442,
+      "grad_norm": 0.41880505997487516,
+      "learning_rate": 3.3866126599490303e-06,
+      "loss": 0.5811,
+      "step": 8752
+    },
+    {
+      "epoch": 2.3236426390548255,
+      "grad_norm": 0.40998814173754866,
+      "learning_rate": 3.3862862163508585e-06,
+      "loss": 0.5597,
+      "step": 8753
+    },
+    {
+      "epoch": 2.323908137528209,
+      "grad_norm": 0.43114239895432926,
+      "learning_rate": 3.3859597554679362e-06,
+      "loss": 0.5747,
+      "step": 8754
+    },
+    {
+      "epoch": 2.3241736360015928,
+      "grad_norm": 0.4357305441063565,
+      "learning_rate": 3.385633277306629e-06,
+      "loss": 0.5677,
+      "step": 8755
+    },
+    {
+      "epoch": 2.324439134474977,
+      "grad_norm": 0.4274764062098414,
+      "learning_rate": 3.3853067818733045e-06,
+      "loss": 0.5624,
+      "step": 8756
+    },
+    {
+      "epoch": 2.3247046329483605,
+      "grad_norm": 0.44134163501348717,
+      "learning_rate": 3.3849802691743305e-06,
+      "loss": 0.6051,
+      "step": 8757
+    },
+    {
+      "epoch": 2.324970131421744,
+      "grad_norm": 0.4168260058285323,
+      "learning_rate": 3.3846537392160743e-06,
+      "loss": 0.5667,
+      "step": 8758
+    },
+    {
+      "epoch": 2.325235629895128,
+      "grad_norm": 0.4301045316116505,
+      "learning_rate": 3.384327192004904e-06,
+      "loss": 0.579,
+      "step": 8759
+    },
+    {
+      "epoch": 2.325501128368512,
+      "grad_norm": 0.43160578163390323,
+      "learning_rate": 3.3840006275471888e-06,
+      "loss": 0.5311,
+      "step": 8760
+    },
+    {
+      "epoch": 2.3257666268418955,
+      "grad_norm": 0.41731258671713795,
+      "learning_rate": 3.3836740458492963e-06,
+      "loss": 0.548,
+      "step": 8761
+    },
+    {
+      "epoch": 2.3260321253152796,
+      "grad_norm": 0.4245560097178044,
+      "learning_rate": 3.3833474469175965e-06,
+      "loss": 0.5531,
+      "step": 8762
+    },
+    {
+      "epoch": 2.3262976237886632,
+      "grad_norm": 0.4311658019781137,
+      "learning_rate": 3.3830208307584583e-06,
+      "loss": 0.5506,
+      "step": 8763
+    },
+    {
+      "epoch": 2.326563122262047,
+      "grad_norm": 0.4350115659052762,
+      "learning_rate": 3.382694197378252e-06,
+      "loss": 0.5865,
+      "step": 8764
+    },
+    {
+      "epoch": 2.326828620735431,
+      "grad_norm": 0.42976356035906665,
+      "learning_rate": 3.3823675467833477e-06,
+      "loss": 0.5659,
+      "step": 8765
+    },
+    {
+      "epoch": 2.3270941192088146,
+      "grad_norm": 0.42217579135381045,
+      "learning_rate": 3.382040878980117e-06,
+      "loss": 0.5446,
+      "step": 8766
+    },
+    {
+      "epoch": 2.3273596176821982,
+      "grad_norm": 0.43685572696200264,
+      "learning_rate": 3.381714193974928e-06,
+      "loss": 0.5804,
+      "step": 8767
+    },
+    {
+      "epoch": 2.3276251161555823,
+      "grad_norm": 0.4215383570453018,
+      "learning_rate": 3.3813874917741546e-06,
+      "loss": 0.5214,
+      "step": 8768
+    },
+    {
+      "epoch": 2.327890614628966,
+      "grad_norm": 0.4371868338171324,
+      "learning_rate": 3.381060772384166e-06,
+      "loss": 0.5836,
+      "step": 8769
+    },
+    {
+      "epoch": 2.3281561131023496,
+      "grad_norm": 0.4320116304077911,
+      "learning_rate": 3.380734035811336e-06,
+      "loss": 0.5612,
+      "step": 8770
+    },
+    {
+      "epoch": 2.3284216115757337,
+      "grad_norm": 0.4402656182516278,
+      "learning_rate": 3.380407282062035e-06,
+      "loss": 0.5674,
+      "step": 8771
+    },
+    {
+      "epoch": 2.3286871100491173,
+      "grad_norm": 0.43072070621367076,
+      "learning_rate": 3.380080511142637e-06,
+      "loss": 0.5517,
+      "step": 8772
+    },
+    {
+      "epoch": 2.328952608522501,
+      "grad_norm": 0.4459760525571599,
+      "learning_rate": 3.379753723059514e-06,
+      "loss": 0.5788,
+      "step": 8773
+    },
+    {
+      "epoch": 2.3292181069958846,
+      "grad_norm": 0.43119698635512205,
+      "learning_rate": 3.3794269178190397e-06,
+      "loss": 0.5514,
+      "step": 8774
+    },
+    {
+      "epoch": 2.3294836054692687,
+      "grad_norm": 0.4313929811357197,
+      "learning_rate": 3.3791000954275873e-06,
+      "loss": 0.6164,
+      "step": 8775
+    },
+    {
+      "epoch": 2.3297491039426523,
+      "grad_norm": 0.42650866371809354,
+      "learning_rate": 3.3787732558915304e-06,
+      "loss": 0.5507,
+      "step": 8776
+    },
+    {
+      "epoch": 2.330014602416036,
+      "grad_norm": 0.42220623556580533,
+      "learning_rate": 3.378446399217243e-06,
+      "loss": 0.549,
+      "step": 8777
+    },
+    {
+      "epoch": 2.33028010088942,
+      "grad_norm": 0.44642883068398237,
+      "learning_rate": 3.378119525411101e-06,
+      "loss": 0.5495,
+      "step": 8778
+    },
+    {
+      "epoch": 2.3305455993628037,
+      "grad_norm": 0.42876822773267426,
+      "learning_rate": 3.3777926344794777e-06,
+      "loss": 0.5854,
+      "step": 8779
+    },
+    {
+      "epoch": 2.3308110978361873,
+      "grad_norm": 0.4248732943623475,
+      "learning_rate": 3.3774657264287484e-06,
+      "loss": 0.5159,
+      "step": 8780
+    },
+    {
+      "epoch": 2.331076596309571,
+      "grad_norm": 0.44854392927100034,
+      "learning_rate": 3.3771388012652893e-06,
+      "loss": 0.5886,
+      "step": 8781
+    },
+    {
+      "epoch": 2.331342094782955,
+      "grad_norm": 0.43426629600643435,
+      "learning_rate": 3.3768118589954764e-06,
+      "loss": 0.5529,
+      "step": 8782
+    },
+    {
+      "epoch": 2.3316075932563387,
+      "grad_norm": 0.4391509435686133,
+      "learning_rate": 3.3764848996256856e-06,
+      "loss": 0.5952,
+      "step": 8783
+    },
+    {
+      "epoch": 2.3318730917297223,
+      "grad_norm": 0.43325678444810684,
+      "learning_rate": 3.3761579231622927e-06,
+      "loss": 0.5662,
+      "step": 8784
+    },
+    {
+      "epoch": 2.3321385902031064,
+      "grad_norm": 0.41361918690565397,
+      "learning_rate": 3.375830929611675e-06,
+      "loss": 0.5371,
+      "step": 8785
+    },
+    {
+      "epoch": 2.33240408867649,
+      "grad_norm": 0.4252050249840596,
+      "learning_rate": 3.3755039189802104e-06,
+      "loss": 0.5588,
+      "step": 8786
+    },
+    {
+      "epoch": 2.3326695871498737,
+      "grad_norm": 0.42546898596264116,
+      "learning_rate": 3.375176891274275e-06,
+      "loss": 0.5665,
+      "step": 8787
+    },
+    {
+      "epoch": 2.332935085623258,
+      "grad_norm": 0.4119843155832369,
+      "learning_rate": 3.3748498465002475e-06,
+      "loss": 0.5328,
+      "step": 8788
+    },
+    {
+      "epoch": 2.3332005840966414,
+      "grad_norm": 0.42847432932985036,
+      "learning_rate": 3.3745227846645064e-06,
+      "loss": 0.547,
+      "step": 8789
+    },
+    {
+      "epoch": 2.333466082570025,
+      "grad_norm": 0.4400134953340087,
+      "learning_rate": 3.37419570577343e-06,
+      "loss": 0.5594,
+      "step": 8790
+    },
+    {
+      "epoch": 2.333731581043409,
+      "grad_norm": 0.43890434350385954,
+      "learning_rate": 3.3738686098333965e-06,
+      "loss": 0.5592,
+      "step": 8791
+    },
+    {
+      "epoch": 2.333997079516793,
+      "grad_norm": 0.43212388558860204,
+      "learning_rate": 3.373541496850786e-06,
+      "loss": 0.5636,
+      "step": 8792
+    },
+    {
+      "epoch": 2.3342625779901764,
+      "grad_norm": 0.43492001564840976,
+      "learning_rate": 3.3732143668319765e-06,
+      "loss": 0.5648,
+      "step": 8793
+    },
+    {
+      "epoch": 2.3345280764635605,
+      "grad_norm": 0.4396660771210867,
+      "learning_rate": 3.3728872197833495e-06,
+      "loss": 0.5419,
+      "step": 8794
+    },
+    {
+      "epoch": 2.334793574936944,
+      "grad_norm": 0.425635948718239,
+      "learning_rate": 3.3725600557112846e-06,
+      "loss": 0.5607,
+      "step": 8795
+    },
+    {
+      "epoch": 2.335059073410328,
+      "grad_norm": 0.43969729976873984,
+      "learning_rate": 3.3722328746221623e-06,
+      "loss": 0.5935,
+      "step": 8796
+    },
+    {
+      "epoch": 2.335324571883712,
+      "grad_norm": 0.42708600909644023,
+      "learning_rate": 3.3719056765223636e-06,
+      "loss": 0.5651,
+      "step": 8797
+    },
+    {
+      "epoch": 2.3355900703570955,
+      "grad_norm": 0.4337202269645962,
+      "learning_rate": 3.371578461418268e-06,
+      "loss": 0.5856,
+      "step": 8798
+    },
+    {
+      "epoch": 2.335855568830479,
+      "grad_norm": 0.43798683514030007,
+      "learning_rate": 3.371251229316259e-06,
+      "loss": 0.5887,
+      "step": 8799
+    },
+    {
+      "epoch": 2.336121067303863,
+      "grad_norm": 0.4336074735676317,
+      "learning_rate": 3.3709239802227188e-06,
+      "loss": 0.5837,
+      "step": 8800
+    },
+    {
+      "epoch": 2.336386565777247,
+      "grad_norm": 0.443097459964302,
+      "learning_rate": 3.3705967141440283e-06,
+      "loss": 0.5525,
+      "step": 8801
+    },
+    {
+      "epoch": 2.3366520642506305,
+      "grad_norm": 0.4334481430177968,
+      "learning_rate": 3.3702694310865696e-06,
+      "loss": 0.5812,
+      "step": 8802
+    },
+    {
+      "epoch": 2.336917562724014,
+      "grad_norm": 0.4288951040913168,
+      "learning_rate": 3.3699421310567272e-06,
+      "loss": 0.5799,
+      "step": 8803
+    },
+    {
+      "epoch": 2.3371830611973983,
+      "grad_norm": 0.4218048729184324,
+      "learning_rate": 3.3696148140608827e-06,
+      "loss": 0.5625,
+      "step": 8804
+    },
+    {
+      "epoch": 2.337448559670782,
+      "grad_norm": 0.4144412073712662,
+      "learning_rate": 3.36928748010542e-06,
+      "loss": 0.5229,
+      "step": 8805
+    },
+    {
+      "epoch": 2.3377140581441656,
+      "grad_norm": 0.4299257655900489,
+      "learning_rate": 3.3689601291967234e-06,
+      "loss": 0.5725,
+      "step": 8806
+    },
+    {
+      "epoch": 2.3379795566175496,
+      "grad_norm": 0.420646340682856,
+      "learning_rate": 3.3686327613411764e-06,
+      "loss": 0.5738,
+      "step": 8807
+    },
+    {
+      "epoch": 2.3382450550909333,
+      "grad_norm": 0.4356379548656279,
+      "learning_rate": 3.3683053765451627e-06,
+      "loss": 0.5778,
+      "step": 8808
+    },
+    {
+      "epoch": 2.338510553564317,
+      "grad_norm": 0.43375528050603035,
+      "learning_rate": 3.36797797481507e-06,
+      "loss": 0.5702,
+      "step": 8809
+    },
+    {
+      "epoch": 2.3387760520377006,
+      "grad_norm": 0.42474780495462766,
+      "learning_rate": 3.367650556157281e-06,
+      "loss": 0.584,
+      "step": 8810
+    },
+    {
+      "epoch": 2.3390415505110846,
+      "grad_norm": 0.41703617322444586,
+      "learning_rate": 3.3673231205781812e-06,
+      "loss": 0.5694,
+      "step": 8811
+    },
+    {
+      "epoch": 2.3393070489844683,
+      "grad_norm": 0.42864664088840404,
+      "learning_rate": 3.3669956680841583e-06,
+      "loss": 0.583,
+      "step": 8812
+    },
+    {
+      "epoch": 2.339572547457852,
+      "grad_norm": 0.43901959590019274,
+      "learning_rate": 3.3666681986815963e-06,
+      "loss": 0.5774,
+      "step": 8813
+    },
+    {
+      "epoch": 2.339838045931236,
+      "grad_norm": 0.4427095235203057,
+      "learning_rate": 3.366340712376882e-06,
+      "loss": 0.5683,
+      "step": 8814
+    },
+    {
+      "epoch": 2.3401035444046197,
+      "grad_norm": 0.43391177678379367,
+      "learning_rate": 3.3660132091764035e-06,
+      "loss": 0.5916,
+      "step": 8815
+    },
+    {
+      "epoch": 2.3403690428780033,
+      "grad_norm": 0.43996243402876556,
+      "learning_rate": 3.3656856890865463e-06,
+      "loss": 0.5566,
+      "step": 8816
+    },
+    {
+      "epoch": 2.3406345413513874,
+      "grad_norm": 0.42766544473265805,
+      "learning_rate": 3.365358152113699e-06,
+      "loss": 0.5585,
+      "step": 8817
+    },
+    {
+      "epoch": 2.340900039824771,
+      "grad_norm": 0.4256743187631771,
+      "learning_rate": 3.365030598264249e-06,
+      "loss": 0.5485,
+      "step": 8818
+    },
+    {
+      "epoch": 2.3411655382981547,
+      "grad_norm": 0.42607483077790975,
+      "learning_rate": 3.364703027544584e-06,
+      "loss": 0.5502,
+      "step": 8819
+    },
+    {
+      "epoch": 2.3414310367715387,
+      "grad_norm": 0.4390112592971977,
+      "learning_rate": 3.3643754399610934e-06,
+      "loss": 0.5432,
+      "step": 8820
+    },
+    {
+      "epoch": 2.3416965352449224,
+      "grad_norm": 0.43955638839497163,
+      "learning_rate": 3.364047835520165e-06,
+      "loss": 0.5666,
+      "step": 8821
+    },
+    {
+      "epoch": 2.341962033718306,
+      "grad_norm": 0.4355844466592428,
+      "learning_rate": 3.3637202142281885e-06,
+      "loss": 0.558,
+      "step": 8822
+    },
+    {
+      "epoch": 2.34222753219169,
+      "grad_norm": 0.4360865128593089,
+      "learning_rate": 3.3633925760915533e-06,
+      "loss": 0.591,
+      "step": 8823
+    },
+    {
+      "epoch": 2.3424930306650737,
+      "grad_norm": 0.43913627042112885,
+      "learning_rate": 3.3630649211166485e-06,
+      "loss": 0.5242,
+      "step": 8824
+    },
+    {
+      "epoch": 2.3427585291384574,
+      "grad_norm": 0.4449555994920618,
+      "learning_rate": 3.362737249309864e-06,
+      "loss": 0.6171,
+      "step": 8825
+    },
+    {
+      "epoch": 2.3430240276118415,
+      "grad_norm": 0.42894660818709235,
+      "learning_rate": 3.3624095606775914e-06,
+      "loss": 0.517,
+      "step": 8826
+    },
+    {
+      "epoch": 2.343289526085225,
+      "grad_norm": 0.42691282307265577,
+      "learning_rate": 3.362081855226221e-06,
+      "loss": 0.5775,
+      "step": 8827
+    },
+    {
+      "epoch": 2.3435550245586088,
+      "grad_norm": 0.429486885888311,
+      "learning_rate": 3.361754132962143e-06,
+      "loss": 0.5937,
+      "step": 8828
+    },
+    {
+      "epoch": 2.3438205230319924,
+      "grad_norm": 0.4264673928090679,
+      "learning_rate": 3.36142639389175e-06,
+      "loss": 0.5913,
+      "step": 8829
+    },
+    {
+      "epoch": 2.3440860215053765,
+      "grad_norm": 0.4591669796189079,
+      "learning_rate": 3.361098638021433e-06,
+      "loss": 0.5259,
+      "step": 8830
+    },
+    {
+      "epoch": 2.34435151997876,
+      "grad_norm": 0.434074941188246,
+      "learning_rate": 3.3607708653575844e-06,
+      "loss": 0.5774,
+      "step": 8831
+    },
+    {
+      "epoch": 2.3446170184521438,
+      "grad_norm": 0.43352346559332433,
+      "learning_rate": 3.360443075906597e-06,
+      "loss": 0.5648,
+      "step": 8832
+    },
+    {
+      "epoch": 2.344882516925528,
+      "grad_norm": 0.4489222667996664,
+      "learning_rate": 3.3601152696748617e-06,
+      "loss": 0.5534,
+      "step": 8833
+    },
+    {
+      "epoch": 2.3451480153989115,
+      "grad_norm": 0.4323721948226055,
+      "learning_rate": 3.3597874466687736e-06,
+      "loss": 0.5751,
+      "step": 8834
+    },
+    {
+      "epoch": 2.345413513872295,
+      "grad_norm": 0.43129548976884186,
+      "learning_rate": 3.359459606894725e-06,
+      "loss": 0.545,
+      "step": 8835
+    },
+    {
+      "epoch": 2.3456790123456788,
+      "grad_norm": 0.424979484629881,
+      "learning_rate": 3.3591317503591093e-06,
+      "loss": 0.5953,
+      "step": 8836
+    },
+    {
+      "epoch": 2.345944510819063,
+      "grad_norm": 0.42770430303252044,
+      "learning_rate": 3.3588038770683207e-06,
+      "loss": 0.5286,
+      "step": 8837
+    },
+    {
+      "epoch": 2.3462100092924465,
+      "grad_norm": 0.4328267637116619,
+      "learning_rate": 3.3584759870287546e-06,
+      "loss": 0.5774,
+      "step": 8838
+    },
+    {
+      "epoch": 2.34647550776583,
+      "grad_norm": 0.42777447353551135,
+      "learning_rate": 3.3581480802468047e-06,
+      "loss": 0.5749,
+      "step": 8839
+    },
+    {
+      "epoch": 2.3467410062392142,
+      "grad_norm": 0.4390972021056961,
+      "learning_rate": 3.3578201567288665e-06,
+      "loss": 0.5706,
+      "step": 8840
+    },
+    {
+      "epoch": 2.347006504712598,
+      "grad_norm": 0.4152630591066803,
+      "learning_rate": 3.357492216481334e-06,
+      "loss": 0.5676,
+      "step": 8841
+    },
+    {
+      "epoch": 2.3472720031859815,
+      "grad_norm": 0.4204784542036679,
+      "learning_rate": 3.3571642595106044e-06,
+      "loss": 0.5463,
+      "step": 8842
+    },
+    {
+      "epoch": 2.3475375016593656,
+      "grad_norm": 0.4352861529070829,
+      "learning_rate": 3.356836285823073e-06,
+      "loss": 0.5969,
+      "step": 8843
+    },
+    {
+      "epoch": 2.3478030001327492,
+      "grad_norm": 0.4125679713394544,
+      "learning_rate": 3.3565082954251352e-06,
+      "loss": 0.5675,
+      "step": 8844
+    },
+    {
+      "epoch": 2.348068498606133,
+      "grad_norm": 0.4196060397927863,
+      "learning_rate": 3.35618028832319e-06,
+      "loss": 0.5985,
+      "step": 8845
+    },
+    {
+      "epoch": 2.348333997079517,
+      "grad_norm": 0.4367278845489271,
+      "learning_rate": 3.355852264523632e-06,
+      "loss": 0.5223,
+      "step": 8846
+    },
+    {
+      "epoch": 2.3485994955529006,
+      "grad_norm": 0.4374217443141185,
+      "learning_rate": 3.3555242240328595e-06,
+      "loss": 0.5846,
+      "step": 8847
+    },
+    {
+      "epoch": 2.3488649940262842,
+      "grad_norm": 0.4435691559751568,
+      "learning_rate": 3.3551961668572697e-06,
+      "loss": 0.5712,
+      "step": 8848
+    },
+    {
+      "epoch": 2.3491304924996683,
+      "grad_norm": 0.42569185553728683,
+      "learning_rate": 3.354868093003262e-06,
+      "loss": 0.5461,
+      "step": 8849
+    },
+    {
+      "epoch": 2.349395990973052,
+      "grad_norm": 0.4348053243345388,
+      "learning_rate": 3.354540002477232e-06,
+      "loss": 0.5527,
+      "step": 8850
+    },
+    {
+      "epoch": 2.3496614894464356,
+      "grad_norm": 0.4340360346449056,
+      "learning_rate": 3.35421189528558e-06,
+      "loss": 0.5459,
+      "step": 8851
+    },
+    {
+      "epoch": 2.3499269879198197,
+      "grad_norm": 0.4349564218448799,
+      "learning_rate": 3.353883771434705e-06,
+      "loss": 0.5569,
+      "step": 8852
+    },
+    {
+      "epoch": 2.3501924863932033,
+      "grad_norm": 0.43148989000995414,
+      "learning_rate": 3.353555630931006e-06,
+      "loss": 0.5608,
+      "step": 8853
+    },
+    {
+      "epoch": 2.350457984866587,
+      "grad_norm": 0.4250984511385049,
+      "learning_rate": 3.3532274737808805e-06,
+      "loss": 0.5662,
+      "step": 8854
+    },
+    {
+      "epoch": 2.3507234833399706,
+      "grad_norm": 0.42860272349187517,
+      "learning_rate": 3.3528992999907323e-06,
+      "loss": 0.5373,
+      "step": 8855
+    },
+    {
+      "epoch": 2.3509889818133547,
+      "grad_norm": 0.42901600710336274,
+      "learning_rate": 3.352571109566958e-06,
+      "loss": 0.558,
+      "step": 8856
+    },
+    {
+      "epoch": 2.3512544802867383,
+      "grad_norm": 0.4410004963888209,
+      "learning_rate": 3.3522429025159606e-06,
+      "loss": 0.5681,
+      "step": 8857
+    },
+    {
+      "epoch": 2.351519978760122,
+      "grad_norm": 0.4402728462751001,
+      "learning_rate": 3.3519146788441405e-06,
+      "loss": 0.6142,
+      "step": 8858
+    },
+    {
+      "epoch": 2.351785477233506,
+      "grad_norm": 0.4357818477038254,
+      "learning_rate": 3.3515864385578974e-06,
+      "loss": 0.581,
+      "step": 8859
+    },
+    {
+      "epoch": 2.3520509757068897,
+      "grad_norm": 0.42294156572081576,
+      "learning_rate": 3.3512581816636336e-06,
+      "loss": 0.5458,
+      "step": 8860
+    },
+    {
+      "epoch": 2.3523164741802733,
+      "grad_norm": 0.43484318752635504,
+      "learning_rate": 3.3509299081677515e-06,
+      "loss": 0.5599,
+      "step": 8861
+    },
+    {
+      "epoch": 2.3525819726536574,
+      "grad_norm": 0.4264117234522694,
+      "learning_rate": 3.350601618076652e-06,
+      "loss": 0.5748,
+      "step": 8862
+    },
+    {
+      "epoch": 2.352847471127041,
+      "grad_norm": 0.4350566582975631,
+      "learning_rate": 3.350273311396739e-06,
+      "loss": 0.5638,
+      "step": 8863
+    },
+    {
+      "epoch": 2.3531129696004247,
+      "grad_norm": 0.4263375138965626,
+      "learning_rate": 3.3499449881344144e-06,
+      "loss": 0.558,
+      "step": 8864
+    },
+    {
+      "epoch": 2.3533784680738083,
+      "grad_norm": 0.4281857724682057,
+      "learning_rate": 3.349616648296081e-06,
+      "loss": 0.5774,
+      "step": 8865
+    },
+    {
+      "epoch": 2.3536439665471924,
+      "grad_norm": 0.4438395270856019,
+      "learning_rate": 3.3492882918881444e-06,
+      "loss": 0.5666,
+      "step": 8866
+    },
+    {
+      "epoch": 2.353909465020576,
+      "grad_norm": 0.43144998072188995,
+      "learning_rate": 3.348959918917005e-06,
+      "loss": 0.5526,
+      "step": 8867
+    },
+    {
+      "epoch": 2.3541749634939597,
+      "grad_norm": 0.42157356313686106,
+      "learning_rate": 3.3486315293890693e-06,
+      "loss": 0.5285,
+      "step": 8868
+    },
+    {
+      "epoch": 2.354440461967344,
+      "grad_norm": 0.4225050936733447,
+      "learning_rate": 3.348303123310741e-06,
+      "loss": 0.5683,
+      "step": 8869
+    },
+    {
+      "epoch": 2.3547059604407274,
+      "grad_norm": 0.42698671797859056,
+      "learning_rate": 3.3479747006884244e-06,
+      "loss": 0.5582,
+      "step": 8870
+    },
+    {
+      "epoch": 2.354971458914111,
+      "grad_norm": 0.4369811388184197,
+      "learning_rate": 3.3476462615285248e-06,
+      "loss": 0.5948,
+      "step": 8871
+    },
+    {
+      "epoch": 2.355236957387495,
+      "grad_norm": 0.43584353222099903,
+      "learning_rate": 3.3473178058374477e-06,
+      "loss": 0.5605,
+      "step": 8872
+    },
+    {
+      "epoch": 2.355502455860879,
+      "grad_norm": 0.4342988989388118,
+      "learning_rate": 3.346989333621599e-06,
+      "loss": 0.5697,
+      "step": 8873
+    },
+    {
+      "epoch": 2.3557679543342624,
+      "grad_norm": 0.4323356280350703,
+      "learning_rate": 3.346660844887385e-06,
+      "loss": 0.592,
+      "step": 8874
+    },
+    {
+      "epoch": 2.3560334528076465,
+      "grad_norm": 0.4379984258111358,
+      "learning_rate": 3.3463323396412113e-06,
+      "loss": 0.604,
+      "step": 8875
+    },
+    {
+      "epoch": 2.35629895128103,
+      "grad_norm": 0.4391101883723459,
+      "learning_rate": 3.3460038178894845e-06,
+      "loss": 0.579,
+      "step": 8876
+    },
+    {
+      "epoch": 2.356564449754414,
+      "grad_norm": 0.43034077033188173,
+      "learning_rate": 3.345675279638612e-06,
+      "loss": 0.5868,
+      "step": 8877
+    },
+    {
+      "epoch": 2.356829948227798,
+      "grad_norm": 0.4391422039112972,
+      "learning_rate": 3.3453467248950013e-06,
+      "loss": 0.5451,
+      "step": 8878
+    },
+    {
+      "epoch": 2.3570954467011815,
+      "grad_norm": 0.4388294514393079,
+      "learning_rate": 3.3450181536650596e-06,
+      "loss": 0.5736,
+      "step": 8879
+    },
+    {
+      "epoch": 2.357360945174565,
+      "grad_norm": 0.4469926633619977,
+      "learning_rate": 3.344689565955194e-06,
+      "loss": 0.5241,
+      "step": 8880
+    },
+    {
+      "epoch": 2.3576264436479493,
+      "grad_norm": 0.4306230800009585,
+      "learning_rate": 3.344360961771815e-06,
+      "loss": 0.5874,
+      "step": 8881
+    },
+    {
+      "epoch": 2.357891942121333,
+      "grad_norm": 0.4422450738544373,
+      "learning_rate": 3.344032341121329e-06,
+      "loss": 0.5298,
+      "step": 8882
+    },
+    {
+      "epoch": 2.3581574405947165,
+      "grad_norm": 0.4478856363136004,
+      "learning_rate": 3.343703704010146e-06,
+      "loss": 0.5847,
+      "step": 8883
+    },
+    {
+      "epoch": 2.3584229390681,
+      "grad_norm": 0.42473028395262435,
+      "learning_rate": 3.3433750504446753e-06,
+      "loss": 0.5447,
+      "step": 8884
+    },
+    {
+      "epoch": 2.3586884375414843,
+      "grad_norm": 0.4205272065008639,
+      "learning_rate": 3.3430463804313262e-06,
+      "loss": 0.5592,
+      "step": 8885
+    },
+    {
+      "epoch": 2.358953936014868,
+      "grad_norm": 0.4220914864315717,
+      "learning_rate": 3.342717693976508e-06,
+      "loss": 0.5557,
+      "step": 8886
+    },
+    {
+      "epoch": 2.3592194344882516,
+      "grad_norm": 0.42965589701602974,
+      "learning_rate": 3.3423889910866324e-06,
+      "loss": 0.5521,
+      "step": 8887
+    },
+    {
+      "epoch": 2.3594849329616356,
+      "grad_norm": 0.4383952630417687,
+      "learning_rate": 3.3420602717681083e-06,
+      "loss": 0.576,
+      "step": 8888
+    },
+    {
+      "epoch": 2.3597504314350193,
+      "grad_norm": 0.4502398447236067,
+      "learning_rate": 3.341731536027347e-06,
+      "loss": 0.566,
+      "step": 8889
+    },
+    {
+      "epoch": 2.360015929908403,
+      "grad_norm": 0.4341871617936367,
+      "learning_rate": 3.3414027838707604e-06,
+      "loss": 0.5773,
+      "step": 8890
+    },
+    {
+      "epoch": 2.3602814283817866,
+      "grad_norm": 0.44883184388842,
+      "learning_rate": 3.341074015304759e-06,
+      "loss": 0.5832,
+      "step": 8891
+    },
+    {
+      "epoch": 2.3605469268551706,
+      "grad_norm": 0.4241440165106155,
+      "learning_rate": 3.340745230335755e-06,
+      "loss": 0.5778,
+      "step": 8892
+    },
+    {
+      "epoch": 2.3608124253285543,
+      "grad_norm": 0.44051555281370386,
+      "learning_rate": 3.3404164289701607e-06,
+      "loss": 0.5691,
+      "step": 8893
+    },
+    {
+      "epoch": 2.361077923801938,
+      "grad_norm": 0.4245983160442818,
+      "learning_rate": 3.3400876112143883e-06,
+      "loss": 0.5985,
+      "step": 8894
+    },
+    {
+      "epoch": 2.361343422275322,
+      "grad_norm": 0.4422129724546392,
+      "learning_rate": 3.3397587770748507e-06,
+      "loss": 0.5401,
+      "step": 8895
+    },
+    {
+      "epoch": 2.3616089207487057,
+      "grad_norm": 0.4248377774710493,
+      "learning_rate": 3.3394299265579606e-06,
+      "loss": 0.5663,
+      "step": 8896
+    },
+    {
+      "epoch": 2.3618744192220893,
+      "grad_norm": 0.43193391761416816,
+      "learning_rate": 3.339101059670132e-06,
+      "loss": 0.5846,
+      "step": 8897
+    },
+    {
+      "epoch": 2.3621399176954734,
+      "grad_norm": 0.4506637390449819,
+      "learning_rate": 3.3387721764177782e-06,
+      "loss": 0.5784,
+      "step": 8898
+    },
+    {
+      "epoch": 2.362405416168857,
+      "grad_norm": 0.42713741780447695,
+      "learning_rate": 3.3384432768073126e-06,
+      "loss": 0.5747,
+      "step": 8899
+    },
+    {
+      "epoch": 2.3626709146422407,
+      "grad_norm": 0.4369320504215003,
+      "learning_rate": 3.3381143608451507e-06,
+      "loss": 0.5715,
+      "step": 8900
+    },
+    {
+      "epoch": 2.3629364131156247,
+      "grad_norm": 0.44044415273640597,
+      "learning_rate": 3.3377854285377072e-06,
+      "loss": 0.577,
+      "step": 8901
+    },
+    {
+      "epoch": 2.3632019115890084,
+      "grad_norm": 0.4285197517306018,
+      "learning_rate": 3.3374564798913955e-06,
+      "loss": 0.5864,
+      "step": 8902
+    },
+    {
+      "epoch": 2.363467410062392,
+      "grad_norm": 0.4265317839010594,
+      "learning_rate": 3.3371275149126325e-06,
+      "loss": 0.5883,
+      "step": 8903
+    },
+    {
+      "epoch": 2.363732908535776,
+      "grad_norm": 0.444399158891179,
+      "learning_rate": 3.336798533607834e-06,
+      "loss": 0.566,
+      "step": 8904
+    },
+    {
+      "epoch": 2.3639984070091598,
+      "grad_norm": 0.4353909692618772,
+      "learning_rate": 3.336469535983415e-06,
+      "loss": 0.5944,
+      "step": 8905
+    },
+    {
+      "epoch": 2.3642639054825434,
+      "grad_norm": 0.43745375585501844,
+      "learning_rate": 3.3361405220457914e-06,
+      "loss": 0.5818,
+      "step": 8906
+    },
+    {
+      "epoch": 2.3645294039559275,
+      "grad_norm": 0.4270421377911294,
+      "learning_rate": 3.335811491801381e-06,
+      "loss": 0.5801,
+      "step": 8907
+    },
+    {
+      "epoch": 2.364794902429311,
+      "grad_norm": 0.43307497995833316,
+      "learning_rate": 3.3354824452566e-06,
+      "loss": 0.5629,
+      "step": 8908
+    },
+    {
+      "epoch": 2.3650604009026948,
+      "grad_norm": 0.42055685182281116,
+      "learning_rate": 3.335153382417865e-06,
+      "loss": 0.5474,
+      "step": 8909
+    },
+    {
+      "epoch": 2.3653258993760784,
+      "grad_norm": 0.43408031922165485,
+      "learning_rate": 3.334824303291595e-06,
+      "loss": 0.5978,
+      "step": 8910
+    },
+    {
+      "epoch": 2.3655913978494625,
+      "grad_norm": 0.43897263764473665,
+      "learning_rate": 3.3344952078842063e-06,
+      "loss": 0.5654,
+      "step": 8911
+    },
+    {
+      "epoch": 2.365856896322846,
+      "grad_norm": 0.4252298461587109,
+      "learning_rate": 3.3341660962021195e-06,
+      "loss": 0.524,
+      "step": 8912
+    },
+    {
+      "epoch": 2.3661223947962298,
+      "grad_norm": 0.43166903261059497,
+      "learning_rate": 3.33383696825175e-06,
+      "loss": 0.5541,
+      "step": 8913
+    },
+    {
+      "epoch": 2.366387893269614,
+      "grad_norm": 0.4401792696403206,
+      "learning_rate": 3.333507824039518e-06,
+      "loss": 0.5506,
+      "step": 8914
+    },
+    {
+      "epoch": 2.3666533917429975,
+      "grad_norm": 0.4187047529131064,
+      "learning_rate": 3.3331786635718433e-06,
+      "loss": 0.5692,
+      "step": 8915
+    },
+    {
+      "epoch": 2.366918890216381,
+      "grad_norm": 0.43420256865427387,
+      "learning_rate": 3.3328494868551444e-06,
+      "loss": 0.5473,
+      "step": 8916
+    },
+    {
+      "epoch": 2.367184388689765,
+      "grad_norm": 0.42722773946391224,
+      "learning_rate": 3.332520293895841e-06,
+      "loss": 0.5822,
+      "step": 8917
+    },
+    {
+      "epoch": 2.367449887163149,
+      "grad_norm": 0.4412444705009898,
+      "learning_rate": 3.3321910847003543e-06,
+      "loss": 0.5729,
+      "step": 8918
+    },
+    {
+      "epoch": 2.3677153856365325,
+      "grad_norm": 0.424083896091279,
+      "learning_rate": 3.331861859275103e-06,
+      "loss": 0.5838,
+      "step": 8919
+    },
+    {
+      "epoch": 2.367980884109916,
+      "grad_norm": 0.4529796841772629,
+      "learning_rate": 3.3315326176265094e-06,
+      "loss": 0.5751,
+      "step": 8920
+    },
+    {
+      "epoch": 2.3682463825833002,
+      "grad_norm": 0.4494633823009935,
+      "learning_rate": 3.331203359760994e-06,
+      "loss": 0.588,
+      "step": 8921
+    },
+    {
+      "epoch": 2.368511881056684,
+      "grad_norm": 0.4295703481885404,
+      "learning_rate": 3.330874085684977e-06,
+      "loss": 0.5599,
+      "step": 8922
+    },
+    {
+      "epoch": 2.3687773795300675,
+      "grad_norm": 0.4273346374535615,
+      "learning_rate": 3.3305447954048814e-06,
+      "loss": 0.5675,
+      "step": 8923
+    },
+    {
+      "epoch": 2.3690428780034516,
+      "grad_norm": 0.47480264179759235,
+      "learning_rate": 3.3302154889271294e-06,
+      "loss": 0.5344,
+      "step": 8924
+    },
+    {
+      "epoch": 2.3693083764768352,
+      "grad_norm": 0.43136636991689664,
+      "learning_rate": 3.329886166258142e-06,
+      "loss": 0.5399,
+      "step": 8925
+    },
+    {
+      "epoch": 2.369573874950219,
+      "grad_norm": 0.4319479532477483,
+      "learning_rate": 3.3295568274043422e-06,
+      "loss": 0.5283,
+      "step": 8926
+    },
+    {
+      "epoch": 2.369839373423603,
+      "grad_norm": 0.438441439982824,
+      "learning_rate": 3.3292274723721536e-06,
+      "loss": 0.5275,
+      "step": 8927
+    },
+    {
+      "epoch": 2.3701048718969866,
+      "grad_norm": 0.44471836262488756,
+      "learning_rate": 3.328898101167999e-06,
+      "loss": 0.5776,
+      "step": 8928
+    },
+    {
+      "epoch": 2.3703703703703702,
+      "grad_norm": 0.4289057332067796,
+      "learning_rate": 3.3285687137983016e-06,
+      "loss": 0.5726,
+      "step": 8929
+    },
+    {
+      "epoch": 2.3706358688437543,
+      "grad_norm": 0.4470900182030393,
+      "learning_rate": 3.328239310269486e-06,
+      "loss": 0.5873,
+      "step": 8930
+    },
+    {
+      "epoch": 2.370901367317138,
+      "grad_norm": 0.4541801082005248,
+      "learning_rate": 3.327909890587976e-06,
+      "loss": 0.6118,
+      "step": 8931
+    },
+    {
+      "epoch": 2.3711668657905216,
+      "grad_norm": 0.4349202020085466,
+      "learning_rate": 3.3275804547601953e-06,
+      "loss": 0.5738,
+      "step": 8932
+    },
+    {
+      "epoch": 2.3714323642639057,
+      "grad_norm": 0.42925759601253943,
+      "learning_rate": 3.3272510027925707e-06,
+      "loss": 0.5672,
+      "step": 8933
+    },
+    {
+      "epoch": 2.3716978627372893,
+      "grad_norm": 0.4494786758818938,
+      "learning_rate": 3.326921534691525e-06,
+      "loss": 0.6012,
+      "step": 8934
+    },
+    {
+      "epoch": 2.371963361210673,
+      "grad_norm": 0.4352010167679763,
+      "learning_rate": 3.3265920504634852e-06,
+      "loss": 0.5525,
+      "step": 8935
+    },
+    {
+      "epoch": 2.372228859684057,
+      "grad_norm": 0.41406772268010444,
+      "learning_rate": 3.3262625501148764e-06,
+      "loss": 0.5448,
+      "step": 8936
+    },
+    {
+      "epoch": 2.3724943581574407,
+      "grad_norm": 0.41525166007121395,
+      "learning_rate": 3.3259330336521252e-06,
+      "loss": 0.533,
+      "step": 8937
+    },
+    {
+      "epoch": 2.3727598566308243,
+      "grad_norm": 0.42046505309234083,
+      "learning_rate": 3.3256035010816577e-06,
+      "loss": 0.592,
+      "step": 8938
+    },
+    {
+      "epoch": 2.373025355104208,
+      "grad_norm": 0.4159543881754008,
+      "learning_rate": 3.3252739524098997e-06,
+      "loss": 0.5918,
+      "step": 8939
+    },
+    {
+      "epoch": 2.373290853577592,
+      "grad_norm": 0.431125485660253,
+      "learning_rate": 3.3249443876432795e-06,
+      "loss": 0.5434,
+      "step": 8940
+    },
+    {
+      "epoch": 2.3735563520509757,
+      "grad_norm": 0.4430481936721721,
+      "learning_rate": 3.3246148067882243e-06,
+      "loss": 0.5706,
+      "step": 8941
+    },
+    {
+      "epoch": 2.3738218505243593,
+      "grad_norm": 0.42744908399488074,
+      "learning_rate": 3.3242852098511607e-06,
+      "loss": 0.563,
+      "step": 8942
+    },
+    {
+      "epoch": 2.3740873489977434,
+      "grad_norm": 0.4302090306120013,
+      "learning_rate": 3.323955596838518e-06,
+      "loss": 0.5837,
+      "step": 8943
+    },
+    {
+      "epoch": 2.374352847471127,
+      "grad_norm": 0.4278112365724571,
+      "learning_rate": 3.3236259677567232e-06,
+      "loss": 0.544,
+      "step": 8944
+    },
+    {
+      "epoch": 2.3746183459445107,
+      "grad_norm": 0.438050020826379,
+      "learning_rate": 3.323296322612205e-06,
+      "loss": 0.581,
+      "step": 8945
+    },
+    {
+      "epoch": 2.3748838444178944,
+      "grad_norm": 0.4499585228578101,
+      "learning_rate": 3.322966661411393e-06,
+      "loss": 0.5821,
+      "step": 8946
+    },
+    {
+      "epoch": 2.3751493428912784,
+      "grad_norm": 0.45782669341340915,
+      "learning_rate": 3.3226369841607174e-06,
+      "loss": 0.5765,
+      "step": 8947
+    },
+    {
+      "epoch": 2.375414841364662,
+      "grad_norm": 0.43309585800248296,
+      "learning_rate": 3.3223072908666053e-06,
+      "loss": 0.5787,
+      "step": 8948
+    },
+    {
+      "epoch": 2.3756803398380457,
+      "grad_norm": 0.43769625362966325,
+      "learning_rate": 3.3219775815354875e-06,
+      "loss": 0.5361,
+      "step": 8949
+    },
+    {
+      "epoch": 2.37594583831143,
+      "grad_norm": 0.44221921238496265,
+      "learning_rate": 3.321647856173795e-06,
+      "loss": 0.6124,
+      "step": 8950
+    },
+    {
+      "epoch": 2.3762113367848134,
+      "grad_norm": 0.43686696963561916,
+      "learning_rate": 3.3213181147879576e-06,
+      "loss": 0.5452,
+      "step": 8951
+    },
+    {
+      "epoch": 2.376476835258197,
+      "grad_norm": 0.42803595460610067,
+      "learning_rate": 3.3209883573844052e-06,
+      "loss": 0.5939,
+      "step": 8952
+    },
+    {
+      "epoch": 2.376742333731581,
+      "grad_norm": 0.42659580352178766,
+      "learning_rate": 3.3206585839695703e-06,
+      "loss": 0.6101,
+      "step": 8953
+    },
+    {
+      "epoch": 2.377007832204965,
+      "grad_norm": 0.42756629395576273,
+      "learning_rate": 3.3203287945498843e-06,
+      "loss": 0.5711,
+      "step": 8954
+    },
+    {
+      "epoch": 2.3772733306783485,
+      "grad_norm": 0.42729382465879334,
+      "learning_rate": 3.3199989891317763e-06,
+      "loss": 0.5595,
+      "step": 8955
+    },
+    {
+      "epoch": 2.3775388291517325,
+      "grad_norm": 0.4397539626090969,
+      "learning_rate": 3.319669167721682e-06,
+      "loss": 0.5205,
+      "step": 8956
+    },
+    {
+      "epoch": 2.377804327625116,
+      "grad_norm": 0.4431624990374816,
+      "learning_rate": 3.3193393303260317e-06,
+      "loss": 0.6023,
+      "step": 8957
+    },
+    {
+      "epoch": 2.3780698260985,
+      "grad_norm": 0.42947122560393286,
+      "learning_rate": 3.319009476951258e-06,
+      "loss": 0.6097,
+      "step": 8958
+    },
+    {
+      "epoch": 2.378335324571884,
+      "grad_norm": 0.42861413870650317,
+      "learning_rate": 3.318679607603795e-06,
+      "loss": 0.5385,
+      "step": 8959
+    },
+    {
+      "epoch": 2.3786008230452675,
+      "grad_norm": 0.4386827420592484,
+      "learning_rate": 3.3183497222900747e-06,
+      "loss": 0.531,
+      "step": 8960
+    },
+    {
+      "epoch": 2.378866321518651,
+      "grad_norm": 0.43612203246490566,
+      "learning_rate": 3.3180198210165316e-06,
+      "loss": 0.5716,
+      "step": 8961
+    },
+    {
+      "epoch": 2.3791318199920353,
+      "grad_norm": 0.4254797625268863,
+      "learning_rate": 3.317689903789598e-06,
+      "loss": 0.5348,
+      "step": 8962
+    },
+    {
+      "epoch": 2.379397318465419,
+      "grad_norm": 0.4392126814240156,
+      "learning_rate": 3.31735997061571e-06,
+      "loss": 0.5507,
+      "step": 8963
+    },
+    {
+      "epoch": 2.3796628169388025,
+      "grad_norm": 0.4447925197687243,
+      "learning_rate": 3.3170300215013013e-06,
+      "loss": 0.6066,
+      "step": 8964
+    },
+    {
+      "epoch": 2.379928315412186,
+      "grad_norm": 0.45156181342584906,
+      "learning_rate": 3.316700056452806e-06,
+      "loss": 0.5493,
+      "step": 8965
+    },
+    {
+      "epoch": 2.3801938138855703,
+      "grad_norm": 0.4347426934114723,
+      "learning_rate": 3.3163700754766607e-06,
+      "loss": 0.5713,
+      "step": 8966
+    },
+    {
+      "epoch": 2.380459312358954,
+      "grad_norm": 0.4295693372840132,
+      "learning_rate": 3.3160400785793007e-06,
+      "loss": 0.5383,
+      "step": 8967
+    },
+    {
+      "epoch": 2.3807248108323376,
+      "grad_norm": 0.4259879973654942,
+      "learning_rate": 3.3157100657671606e-06,
+      "loss": 0.524,
+      "step": 8968
+    },
+    {
+      "epoch": 2.3809903093057216,
+      "grad_norm": 0.4635457295863235,
+      "learning_rate": 3.3153800370466766e-06,
+      "loss": 0.5704,
+      "step": 8969
+    },
+    {
+      "epoch": 2.3812558077791053,
+      "grad_norm": 0.41704338076713193,
+      "learning_rate": 3.315049992424286e-06,
+      "loss": 0.5755,
+      "step": 8970
+    },
+    {
+      "epoch": 2.381521306252489,
+      "grad_norm": 0.4413481594447607,
+      "learning_rate": 3.314719931906425e-06,
+      "loss": 0.5714,
+      "step": 8971
+    },
+    {
+      "epoch": 2.381786804725873,
+      "grad_norm": 0.44544683772216187,
+      "learning_rate": 3.31438985549953e-06,
+      "loss": 0.5694,
+      "step": 8972
+    },
+    {
+      "epoch": 2.3820523031992566,
+      "grad_norm": 0.44183917269096257,
+      "learning_rate": 3.314059763210039e-06,
+      "loss": 0.5452,
+      "step": 8973
+    },
+    {
+      "epoch": 2.3823178016726403,
+      "grad_norm": 0.42772579390770954,
+      "learning_rate": 3.31372965504439e-06,
+      "loss": 0.5454,
+      "step": 8974
+    },
+    {
+      "epoch": 2.382583300146024,
+      "grad_norm": 0.438316228849524,
+      "learning_rate": 3.3133995310090196e-06,
+      "loss": 0.5667,
+      "step": 8975
+    },
+    {
+      "epoch": 2.382848798619408,
+      "grad_norm": 0.41846680898985295,
+      "learning_rate": 3.3130693911103672e-06,
+      "loss": 0.5691,
+      "step": 8976
+    },
+    {
+      "epoch": 2.3831142970927917,
+      "grad_norm": 0.4377084618128971,
+      "learning_rate": 3.312739235354871e-06,
+      "loss": 0.5699,
+      "step": 8977
+    },
+    {
+      "epoch": 2.3833797955661753,
+      "grad_norm": 0.436086779882674,
+      "learning_rate": 3.3124090637489696e-06,
+      "loss": 0.5748,
+      "step": 8978
+    },
+    {
+      "epoch": 2.3836452940395594,
+      "grad_norm": 0.43479769510778893,
+      "learning_rate": 3.312078876299103e-06,
+      "loss": 0.5728,
+      "step": 8979
+    },
+    {
+      "epoch": 2.383910792512943,
+      "grad_norm": 0.4354054568009158,
+      "learning_rate": 3.3117486730117092e-06,
+      "loss": 0.5927,
+      "step": 8980
+    },
+    {
+      "epoch": 2.3841762909863267,
+      "grad_norm": 0.44673025767342245,
+      "learning_rate": 3.3114184538932294e-06,
+      "loss": 0.537,
+      "step": 8981
+    },
+    {
+      "epoch": 2.3844417894597107,
+      "grad_norm": 0.430354988224778,
+      "learning_rate": 3.3110882189501022e-06,
+      "loss": 0.5895,
+      "step": 8982
+    },
+    {
+      "epoch": 2.3847072879330944,
+      "grad_norm": 0.4386886765841535,
+      "learning_rate": 3.310757968188769e-06,
+      "loss": 0.5945,
+      "step": 8983
+    },
+    {
+      "epoch": 2.384972786406478,
+      "grad_norm": 0.43263691756984424,
+      "learning_rate": 3.3104277016156707e-06,
+      "loss": 0.551,
+      "step": 8984
+    },
+    {
+      "epoch": 2.385238284879862,
+      "grad_norm": 0.45993801063373413,
+      "learning_rate": 3.3100974192372484e-06,
+      "loss": 0.5549,
+      "step": 8985
+    },
+    {
+      "epoch": 2.3855037833532458,
+      "grad_norm": 0.4321727250108383,
+      "learning_rate": 3.3097671210599425e-06,
+      "loss": 0.5553,
+      "step": 8986
+    },
+    {
+      "epoch": 2.3857692818266294,
+      "grad_norm": 0.4290232568933907,
+      "learning_rate": 3.309436807090195e-06,
+      "loss": 0.5697,
+      "step": 8987
+    },
+    {
+      "epoch": 2.3860347803000135,
+      "grad_norm": 0.4227054942908084,
+      "learning_rate": 3.309106477334448e-06,
+      "loss": 0.5497,
+      "step": 8988
+    },
+    {
+      "epoch": 2.386300278773397,
+      "grad_norm": 0.44266586522443746,
+      "learning_rate": 3.3087761317991433e-06,
+      "loss": 0.5739,
+      "step": 8989
+    },
+    {
+      "epoch": 2.3865657772467808,
+      "grad_norm": 0.418396380667841,
+      "learning_rate": 3.3084457704907245e-06,
+      "loss": 0.5674,
+      "step": 8990
+    },
+    {
+      "epoch": 2.386831275720165,
+      "grad_norm": 0.4473386595077829,
+      "learning_rate": 3.308115393415633e-06,
+      "loss": 0.5921,
+      "step": 8991
+    },
+    {
+      "epoch": 2.3870967741935485,
+      "grad_norm": 0.4335663370808527,
+      "learning_rate": 3.307785000580313e-06,
+      "loss": 0.5646,
+      "step": 8992
+    },
+    {
+      "epoch": 2.387362272666932,
+      "grad_norm": 0.4429002140683974,
+      "learning_rate": 3.3074545919912083e-06,
+      "loss": 0.5924,
+      "step": 8993
+    },
+    {
+      "epoch": 2.3876277711403158,
+      "grad_norm": 0.4359265006846511,
+      "learning_rate": 3.3071241676547617e-06,
+      "loss": 0.5409,
+      "step": 8994
+    },
+    {
+      "epoch": 2.3878932696137,
+      "grad_norm": 0.44248286040260887,
+      "learning_rate": 3.306793727577417e-06,
+      "loss": 0.5736,
+      "step": 8995
+    },
+    {
+      "epoch": 2.3881587680870835,
+      "grad_norm": 0.4344038649302684,
+      "learning_rate": 3.3064632717656203e-06,
+      "loss": 0.5784,
+      "step": 8996
+    },
+    {
+      "epoch": 2.388424266560467,
+      "grad_norm": 0.44642269304610177,
+      "learning_rate": 3.3061328002258143e-06,
+      "loss": 0.5712,
+      "step": 8997
+    },
+    {
+      "epoch": 2.388689765033851,
+      "grad_norm": 0.43630233992230993,
+      "learning_rate": 3.3058023129644458e-06,
+      "loss": 0.5527,
+      "step": 8998
+    },
+    {
+      "epoch": 2.388955263507235,
+      "grad_norm": 0.43482513745974766,
+      "learning_rate": 3.305471809987959e-06,
+      "loss": 0.5487,
+      "step": 8999
+    },
+    {
+      "epoch": 2.3892207619806185,
+      "grad_norm": 0.43528551937940396,
+      "learning_rate": 3.3051412913027993e-06,
+      "loss": 0.5508,
+      "step": 9000
+    },
+    {
+      "epoch": 2.389486260454002,
+      "grad_norm": 0.4164359446587163,
+      "learning_rate": 3.3048107569154124e-06,
+      "loss": 0.5574,
+      "step": 9001
+    },
+    {
+      "epoch": 2.3897517589273862,
+      "grad_norm": 0.43107175165329814,
+      "learning_rate": 3.304480206832247e-06,
+      "loss": 0.5529,
+      "step": 9002
+    },
+    {
+      "epoch": 2.39001725740077,
+      "grad_norm": 0.43958148467125346,
+      "learning_rate": 3.304149641059747e-06,
+      "loss": 0.5407,
+      "step": 9003
+    },
+    {
+      "epoch": 2.3902827558741535,
+      "grad_norm": 0.44814532759915765,
+      "learning_rate": 3.3038190596043595e-06,
+      "loss": 0.5925,
+      "step": 9004
+    },
+    {
+      "epoch": 2.3905482543475376,
+      "grad_norm": 0.42870170473471575,
+      "learning_rate": 3.303488462472534e-06,
+      "loss": 0.5591,
+      "step": 9005
+    },
+    {
+      "epoch": 2.3908137528209212,
+      "grad_norm": 0.4362208141538646,
+      "learning_rate": 3.3031578496707144e-06,
+      "loss": 0.5746,
+      "step": 9006
+    },
+    {
+      "epoch": 2.391079251294305,
+      "grad_norm": 0.4315727270725072,
+      "learning_rate": 3.302827221205351e-06,
+      "loss": 0.5864,
+      "step": 9007
+    },
+    {
+      "epoch": 2.391344749767689,
+      "grad_norm": 0.4252190510641299,
+      "learning_rate": 3.3024965770828914e-06,
+      "loss": 0.5889,
+      "step": 9008
+    },
+    {
+      "epoch": 2.3916102482410726,
+      "grad_norm": 0.45563921278842917,
+      "learning_rate": 3.302165917309783e-06,
+      "loss": 0.5814,
+      "step": 9009
+    },
+    {
+      "epoch": 2.3918757467144562,
+      "grad_norm": 0.4336231217139674,
+      "learning_rate": 3.301835241892476e-06,
+      "loss": 0.5592,
+      "step": 9010
+    },
+    {
+      "epoch": 2.3921412451878403,
+      "grad_norm": 0.4393491909832698,
+      "learning_rate": 3.3015045508374177e-06,
+      "loss": 0.5736,
+      "step": 9011
+    },
+    {
+      "epoch": 2.392406743661224,
+      "grad_norm": 0.4433823245956819,
+      "learning_rate": 3.301173844151059e-06,
+      "loss": 0.5749,
+      "step": 9012
+    },
+    {
+      "epoch": 2.3926722421346076,
+      "grad_norm": 0.42753774010277257,
+      "learning_rate": 3.3008431218398483e-06,
+      "loss": 0.5702,
+      "step": 9013
+    },
+    {
+      "epoch": 2.3929377406079917,
+      "grad_norm": 0.4637681095886912,
+      "learning_rate": 3.300512383910236e-06,
+      "loss": 0.5573,
+      "step": 9014
+    },
+    {
+      "epoch": 2.3932032390813753,
+      "grad_norm": 0.450509250472613,
+      "learning_rate": 3.3001816303686718e-06,
+      "loss": 0.5822,
+      "step": 9015
+    },
+    {
+      "epoch": 2.393468737554759,
+      "grad_norm": 0.46193271022882076,
+      "learning_rate": 3.299850861221608e-06,
+      "loss": 0.5783,
+      "step": 9016
+    },
+    {
+      "epoch": 2.393734236028143,
+      "grad_norm": 0.43633129242591095,
+      "learning_rate": 3.2995200764754924e-06,
+      "loss": 0.5655,
+      "step": 9017
+    },
+    {
+      "epoch": 2.3939997345015267,
+      "grad_norm": 0.4248732573455485,
+      "learning_rate": 3.299189276136778e-06,
+      "loss": 0.557,
+      "step": 9018
+    },
+    {
+      "epoch": 2.3942652329749103,
+      "grad_norm": 0.44045966667493264,
+      "learning_rate": 3.2988584602119166e-06,
+      "loss": 0.5681,
+      "step": 9019
+    },
+    {
+      "epoch": 2.394530731448294,
+      "grad_norm": 0.4373104490205752,
+      "learning_rate": 3.298527628707359e-06,
+      "loss": 0.6011,
+      "step": 9020
+    },
+    {
+      "epoch": 2.394796229921678,
+      "grad_norm": 0.43071368277808325,
+      "learning_rate": 3.2981967816295578e-06,
+      "loss": 0.5369,
+      "step": 9021
+    },
+    {
+      "epoch": 2.3950617283950617,
+      "grad_norm": 0.43101590793783506,
+      "learning_rate": 3.2978659189849647e-06,
+      "loss": 0.5302,
+      "step": 9022
+    },
+    {
+      "epoch": 2.3953272268684453,
+      "grad_norm": 0.4306973182771112,
+      "learning_rate": 3.2975350407800333e-06,
+      "loss": 0.5344,
+      "step": 9023
+    },
+    {
+      "epoch": 2.3955927253418294,
+      "grad_norm": 0.43845580504732257,
+      "learning_rate": 3.297204147021215e-06,
+      "loss": 0.592,
+      "step": 9024
+    },
+    {
+      "epoch": 2.395858223815213,
+      "grad_norm": 0.43915433873623594,
+      "learning_rate": 3.2968732377149648e-06,
+      "loss": 0.5913,
+      "step": 9025
+    },
+    {
+      "epoch": 2.3961237222885967,
+      "grad_norm": 0.4413209439489076,
+      "learning_rate": 3.2965423128677347e-06,
+      "loss": 0.5574,
+      "step": 9026
+    },
+    {
+      "epoch": 2.396389220761981,
+      "grad_norm": 0.43644401063689026,
+      "learning_rate": 3.296211372485979e-06,
+      "loss": 0.5492,
+      "step": 9027
+    },
+    {
+      "epoch": 2.3966547192353644,
+      "grad_norm": 0.4386200850754053,
+      "learning_rate": 3.295880416576153e-06,
+      "loss": 0.5719,
+      "step": 9028
+    },
+    {
+      "epoch": 2.396920217708748,
+      "grad_norm": 0.43313383845216535,
+      "learning_rate": 3.295549445144709e-06,
+      "loss": 0.5846,
+      "step": 9029
+    },
+    {
+      "epoch": 2.3971857161821317,
+      "grad_norm": 0.43967996766675765,
+      "learning_rate": 3.295218458198104e-06,
+      "loss": 0.5699,
+      "step": 9030
+    },
+    {
+      "epoch": 2.397451214655516,
+      "grad_norm": 0.45365517654404003,
+      "learning_rate": 3.2948874557427923e-06,
+      "loss": 0.5412,
+      "step": 9031
+    },
+    {
+      "epoch": 2.3977167131288994,
+      "grad_norm": 0.44637822687623785,
+      "learning_rate": 3.294556437785228e-06,
+      "loss": 0.5862,
+      "step": 9032
+    },
+    {
+      "epoch": 2.397982211602283,
+      "grad_norm": 0.43278537443010157,
+      "learning_rate": 3.2942254043318688e-06,
+      "loss": 0.5635,
+      "step": 9033
+    },
+    {
+      "epoch": 2.398247710075667,
+      "grad_norm": 0.44873148898784226,
+      "learning_rate": 3.2938943553891687e-06,
+      "loss": 0.5448,
+      "step": 9034
+    },
+    {
+      "epoch": 2.398513208549051,
+      "grad_norm": 0.4516706392984563,
+      "learning_rate": 3.2935632909635852e-06,
+      "loss": 0.5497,
+      "step": 9035
+    },
+    {
+      "epoch": 2.3987787070224345,
+      "grad_norm": 0.43999825675334375,
+      "learning_rate": 3.2932322110615744e-06,
+      "loss": 0.596,
+      "step": 9036
+    },
+    {
+      "epoch": 2.3990442054958185,
+      "grad_norm": 0.4410207508138358,
+      "learning_rate": 3.292901115689593e-06,
+      "loss": 0.5538,
+      "step": 9037
+    },
+    {
+      "epoch": 2.399309703969202,
+      "grad_norm": 0.44325866107186057,
+      "learning_rate": 3.2925700048540985e-06,
+      "loss": 0.5582,
+      "step": 9038
+    },
+    {
+      "epoch": 2.399575202442586,
+      "grad_norm": 0.44671382021661027,
+      "learning_rate": 3.29223887856155e-06,
+      "loss": 0.5958,
+      "step": 9039
+    },
+    {
+      "epoch": 2.39984070091597,
+      "grad_norm": 0.4399975235460901,
+      "learning_rate": 3.2919077368184023e-06,
+      "loss": 0.579,
+      "step": 9040
+    },
+    {
+      "epoch": 2.4001061993893535,
+      "grad_norm": 0.44199927150768964,
+      "learning_rate": 3.2915765796311143e-06,
+      "loss": 0.5921,
+      "step": 9041
+    },
+    {
+      "epoch": 2.400371697862737,
+      "grad_norm": 0.4468000911483779,
+      "learning_rate": 3.291245407006146e-06,
+      "loss": 0.5704,
+      "step": 9042
+    },
+    {
+      "epoch": 2.4006371963361213,
+      "grad_norm": 0.4283960643988952,
+      "learning_rate": 3.2909142189499545e-06,
+      "loss": 0.5535,
+      "step": 9043
+    },
+    {
+      "epoch": 2.400902694809505,
+      "grad_norm": 0.4474830594423144,
+      "learning_rate": 3.290583015468999e-06,
+      "loss": 0.5276,
+      "step": 9044
+    },
+    {
+      "epoch": 2.4011681932828886,
+      "grad_norm": 0.4387095466128506,
+      "learning_rate": 3.29025179656974e-06,
+      "loss": 0.5579,
+      "step": 9045
+    },
+    {
+      "epoch": 2.4014336917562726,
+      "grad_norm": 0.439554860831197,
+      "learning_rate": 3.289920562258635e-06,
+      "loss": 0.5366,
+      "step": 9046
+    },
+    {
+      "epoch": 2.4016991902296563,
+      "grad_norm": 0.4256204999554291,
+      "learning_rate": 3.2895893125421447e-06,
+      "loss": 0.5503,
+      "step": 9047
+    },
+    {
+      "epoch": 2.40196468870304,
+      "grad_norm": 0.4286274143110022,
+      "learning_rate": 3.2892580474267305e-06,
+      "loss": 0.5741,
+      "step": 9048
+    },
+    {
+      "epoch": 2.4022301871764236,
+      "grad_norm": 0.43799657315997353,
+      "learning_rate": 3.288926766918852e-06,
+      "loss": 0.5729,
+      "step": 9049
+    },
+    {
+      "epoch": 2.4024956856498076,
+      "grad_norm": 0.45387084500085684,
+      "learning_rate": 3.288595471024969e-06,
+      "loss": 0.5912,
+      "step": 9050
+    },
+    {
+      "epoch": 2.4027611841231913,
+      "grad_norm": 0.46089971493613646,
+      "learning_rate": 3.288264159751544e-06,
+      "loss": 0.5887,
+      "step": 9051
+    },
+    {
+      "epoch": 2.403026682596575,
+      "grad_norm": 0.45149176031626753,
+      "learning_rate": 3.287932833105038e-06,
+      "loss": 0.5636,
+      "step": 9052
+    },
+    {
+      "epoch": 2.403292181069959,
+      "grad_norm": 0.456286986382293,
+      "learning_rate": 3.2876014910919125e-06,
+      "loss": 0.5764,
+      "step": 9053
+    },
+    {
+      "epoch": 2.4035576795433427,
+      "grad_norm": 0.4312070695857693,
+      "learning_rate": 3.2872701337186298e-06,
+      "loss": 0.5504,
+      "step": 9054
+    },
+    {
+      "epoch": 2.4038231780167263,
+      "grad_norm": 0.43698954704816434,
+      "learning_rate": 3.2869387609916514e-06,
+      "loss": 0.5831,
+      "step": 9055
+    },
+    {
+      "epoch": 2.40408867649011,
+      "grad_norm": 0.45754160345467837,
+      "learning_rate": 3.286607372917441e-06,
+      "loss": 0.5648,
+      "step": 9056
+    },
+    {
+      "epoch": 2.404354174963494,
+      "grad_norm": 0.4458020898123882,
+      "learning_rate": 3.2862759695024605e-06,
+      "loss": 0.5753,
+      "step": 9057
+    },
+    {
+      "epoch": 2.4046196734368777,
+      "grad_norm": 0.4301693943844551,
+      "learning_rate": 3.285944550753173e-06,
+      "loss": 0.5454,
+      "step": 9058
+    },
+    {
+      "epoch": 2.4048851719102613,
+      "grad_norm": 0.42831206217370205,
+      "learning_rate": 3.2856131166760438e-06,
+      "loss": 0.5809,
+      "step": 9059
+    },
+    {
+      "epoch": 2.4051506703836454,
+      "grad_norm": 0.42342340460670735,
+      "learning_rate": 3.285281667277534e-06,
+      "loss": 0.569,
+      "step": 9060
+    },
+    {
+      "epoch": 2.405416168857029,
+      "grad_norm": 0.42558313523624053,
+      "learning_rate": 3.284950202564109e-06,
+      "loss": 0.5475,
+      "step": 9061
+    },
+    {
+      "epoch": 2.4056816673304127,
+      "grad_norm": 0.43462346314129197,
+      "learning_rate": 3.2846187225422343e-06,
+      "loss": 0.5627,
+      "step": 9062
+    },
+    {
+      "epoch": 2.4059471658037968,
+      "grad_norm": 0.4390428636695656,
+      "learning_rate": 3.2842872272183723e-06,
+      "loss": 0.5833,
+      "step": 9063
+    },
+    {
+      "epoch": 2.4062126642771804,
+      "grad_norm": 0.42980440260601255,
+      "learning_rate": 3.283955716598989e-06,
+      "loss": 0.544,
+      "step": 9064
+    },
+    {
+      "epoch": 2.406478162750564,
+      "grad_norm": 0.4311810242230458,
+      "learning_rate": 3.283624190690551e-06,
+      "loss": 0.5428,
+      "step": 9065
+    },
+    {
+      "epoch": 2.406743661223948,
+      "grad_norm": 0.42029966084695025,
+      "learning_rate": 3.2832926494995215e-06,
+      "loss": 0.5303,
+      "step": 9066
+    },
+    {
+      "epoch": 2.4070091596973318,
+      "grad_norm": 0.42526557041297225,
+      "learning_rate": 3.282961093032368e-06,
+      "loss": 0.5637,
+      "step": 9067
+    },
+    {
+      "epoch": 2.4072746581707154,
+      "grad_norm": 0.42468257885172217,
+      "learning_rate": 3.282629521295556e-06,
+      "loss": 0.5596,
+      "step": 9068
+    },
+    {
+      "epoch": 2.4075401566440995,
+      "grad_norm": 0.4307766994261409,
+      "learning_rate": 3.2822979342955525e-06,
+      "loss": 0.5756,
+      "step": 9069
+    },
+    {
+      "epoch": 2.407805655117483,
+      "grad_norm": 0.4338800409124982,
+      "learning_rate": 3.2819663320388233e-06,
+      "loss": 0.5513,
+      "step": 9070
+    },
+    {
+      "epoch": 2.4080711535908668,
+      "grad_norm": 0.4497810731846223,
+      "learning_rate": 3.2816347145318367e-06,
+      "loss": 0.5388,
+      "step": 9071
+    },
+    {
+      "epoch": 2.408336652064251,
+      "grad_norm": 0.4374456560631929,
+      "learning_rate": 3.2813030817810587e-06,
+      "loss": 0.5981,
+      "step": 9072
+    },
+    {
+      "epoch": 2.4086021505376345,
+      "grad_norm": 0.4301494899973307,
+      "learning_rate": 3.2809714337929584e-06,
+      "loss": 0.584,
+      "step": 9073
+    },
+    {
+      "epoch": 2.408867649011018,
+      "grad_norm": 0.42908340940462375,
+      "learning_rate": 3.280639770574002e-06,
+      "loss": 0.5518,
+      "step": 9074
+    },
+    {
+      "epoch": 2.4091331474844018,
+      "grad_norm": 0.4488992569860736,
+      "learning_rate": 3.28030809213066e-06,
+      "loss": 0.5796,
+      "step": 9075
+    },
+    {
+      "epoch": 2.409398645957786,
+      "grad_norm": 0.4387102985845898,
+      "learning_rate": 3.2799763984693982e-06,
+      "loss": 0.5325,
+      "step": 9076
+    },
+    {
+      "epoch": 2.4096641444311695,
+      "grad_norm": 0.42057295179418563,
+      "learning_rate": 3.2796446895966883e-06,
+      "loss": 0.5471,
+      "step": 9077
+    },
+    {
+      "epoch": 2.409929642904553,
+      "grad_norm": 0.4268868337030098,
+      "learning_rate": 3.279312965518997e-06,
+      "loss": 0.568,
+      "step": 9078
+    },
+    {
+      "epoch": 2.4101951413779372,
+      "grad_norm": 0.44955924659503443,
+      "learning_rate": 3.2789812262427954e-06,
+      "loss": 0.5708,
+      "step": 9079
+    },
+    {
+      "epoch": 2.410460639851321,
+      "grad_norm": 0.4440585093204899,
+      "learning_rate": 3.278649471774553e-06,
+      "loss": 0.5546,
+      "step": 9080
+    },
+    {
+      "epoch": 2.4107261383247045,
+      "grad_norm": 0.43300766950418235,
+      "learning_rate": 3.2783177021207387e-06,
+      "loss": 0.5644,
+      "step": 9081
+    },
+    {
+      "epoch": 2.4109916367980886,
+      "grad_norm": 0.43560153230019805,
+      "learning_rate": 3.2779859172878238e-06,
+      "loss": 0.5752,
+      "step": 9082
+    },
+    {
+      "epoch": 2.4112571352714722,
+      "grad_norm": 0.4378306348115646,
+      "learning_rate": 3.2776541172822783e-06,
+      "loss": 0.5631,
+      "step": 9083
+    },
+    {
+      "epoch": 2.411522633744856,
+      "grad_norm": 0.45888931326126037,
+      "learning_rate": 3.277322302110574e-06,
+      "loss": 0.5527,
+      "step": 9084
+    },
+    {
+      "epoch": 2.4117881322182395,
+      "grad_norm": 0.43992651384570824,
+      "learning_rate": 3.2769904717791818e-06,
+      "loss": 0.5663,
+      "step": 9085
+    },
+    {
+      "epoch": 2.4120536306916236,
+      "grad_norm": 0.42832766778053155,
+      "learning_rate": 3.276658626294572e-06,
+      "loss": 0.5625,
+      "step": 9086
+    },
+    {
+      "epoch": 2.4123191291650072,
+      "grad_norm": 0.43847995820519775,
+      "learning_rate": 3.276326765663218e-06,
+      "loss": 0.5542,
+      "step": 9087
+    },
+    {
+      "epoch": 2.412584627638391,
+      "grad_norm": 0.4441117513484502,
+      "learning_rate": 3.2759948898915916e-06,
+      "loss": 0.5731,
+      "step": 9088
+    },
+    {
+      "epoch": 2.412850126111775,
+      "grad_norm": 0.4343526164747089,
+      "learning_rate": 3.275662998986165e-06,
+      "loss": 0.553,
+      "step": 9089
+    },
+    {
+      "epoch": 2.4131156245851586,
+      "grad_norm": 0.4402587746978412,
+      "learning_rate": 3.27533109295341e-06,
+      "loss": 0.5685,
+      "step": 9090
+    },
+    {
+      "epoch": 2.4133811230585422,
+      "grad_norm": 0.42444058077186153,
+      "learning_rate": 3.2749991717998007e-06,
+      "loss": 0.5799,
+      "step": 9091
+    },
+    {
+      "epoch": 2.4136466215319263,
+      "grad_norm": 0.43820718119411045,
+      "learning_rate": 3.27466723553181e-06,
+      "loss": 0.5804,
+      "step": 9092
+    },
+    {
+      "epoch": 2.41391212000531,
+      "grad_norm": 0.43654068591366924,
+      "learning_rate": 3.274335284155911e-06,
+      "loss": 0.5783,
+      "step": 9093
+    },
+    {
+      "epoch": 2.4141776184786936,
+      "grad_norm": 0.43532639651992966,
+      "learning_rate": 3.2740033176785792e-06,
+      "loss": 0.5932,
+      "step": 9094
+    },
+    {
+      "epoch": 2.4144431169520777,
+      "grad_norm": 0.4348176848811108,
+      "learning_rate": 3.2736713361062873e-06,
+      "loss": 0.5648,
+      "step": 9095
+    },
+    {
+      "epoch": 2.4147086154254613,
+      "grad_norm": 0.4372382741372949,
+      "learning_rate": 3.2733393394455094e-06,
+      "loss": 0.5569,
+      "step": 9096
+    },
+    {
+      "epoch": 2.414974113898845,
+      "grad_norm": 0.41345614969053285,
+      "learning_rate": 3.273007327702722e-06,
+      "loss": 0.5488,
+      "step": 9097
+    },
+    {
+      "epoch": 2.415239612372229,
+      "grad_norm": 0.4395229362251379,
+      "learning_rate": 3.2726753008843983e-06,
+      "loss": 0.5766,
+      "step": 9098
+    },
+    {
+      "epoch": 2.4155051108456127,
+      "grad_norm": 0.43011577391031586,
+      "learning_rate": 3.2723432589970137e-06,
+      "loss": 0.5725,
+      "step": 9099
+    },
+    {
+      "epoch": 2.4157706093189963,
+      "grad_norm": 0.44600835447553777,
+      "learning_rate": 3.2720112020470463e-06,
+      "loss": 0.5344,
+      "step": 9100
+    },
+    {
+      "epoch": 2.4160361077923804,
+      "grad_norm": 0.4383011380768872,
+      "learning_rate": 3.271679130040969e-06,
+      "loss": 0.5528,
+      "step": 9101
+    },
+    {
+      "epoch": 2.416301606265764,
+      "grad_norm": 0.4385086349773373,
+      "learning_rate": 3.271347042985259e-06,
+      "loss": 0.5764,
+      "step": 9102
+    },
+    {
+      "epoch": 2.4165671047391477,
+      "grad_norm": 0.4414712358391007,
+      "learning_rate": 3.2710149408863934e-06,
+      "loss": 0.5411,
+      "step": 9103
+    },
+    {
+      "epoch": 2.4168326032125313,
+      "grad_norm": 0.43301901942809695,
+      "learning_rate": 3.2706828237508493e-06,
+      "loss": 0.5856,
+      "step": 9104
+    },
+    {
+      "epoch": 2.4170981016859154,
+      "grad_norm": 0.43166595190549506,
+      "learning_rate": 3.2703506915851024e-06,
+      "loss": 0.5793,
+      "step": 9105
+    },
+    {
+      "epoch": 2.417363600159299,
+      "grad_norm": 0.4476928142887255,
+      "learning_rate": 3.2700185443956317e-06,
+      "loss": 0.6021,
+      "step": 9106
+    },
+    {
+      "epoch": 2.4176290986326827,
+      "grad_norm": 0.44121866991769576,
+      "learning_rate": 3.2696863821889135e-06,
+      "loss": 0.582,
+      "step": 9107
+    },
+    {
+      "epoch": 2.417894597106067,
+      "grad_norm": 0.43731115327928305,
+      "learning_rate": 3.269354204971427e-06,
+      "loss": 0.5791,
+      "step": 9108
+    },
+    {
+      "epoch": 2.4181600955794504,
+      "grad_norm": 0.4259817802135003,
+      "learning_rate": 3.2690220127496485e-06,
+      "loss": 0.471,
+      "step": 9109
+    },
+    {
+      "epoch": 2.418425594052834,
+      "grad_norm": 0.4373280270444597,
+      "learning_rate": 3.2686898055300587e-06,
+      "loss": 0.5678,
+      "step": 9110
+    },
+    {
+      "epoch": 2.418691092526218,
+      "grad_norm": 0.4422824302520187,
+      "learning_rate": 3.268357583319135e-06,
+      "loss": 0.5717,
+      "step": 9111
+    },
+    {
+      "epoch": 2.418956590999602,
+      "grad_norm": 0.44709192473826037,
+      "learning_rate": 3.2680253461233578e-06,
+      "loss": 0.5924,
+      "step": 9112
+    },
+    {
+      "epoch": 2.4192220894729854,
+      "grad_norm": 0.44701150235350845,
+      "learning_rate": 3.2676930939492058e-06,
+      "loss": 0.5224,
+      "step": 9113
+    },
+    {
+      "epoch": 2.419487587946369,
+      "grad_norm": 0.4401849762627822,
+      "learning_rate": 3.2673608268031594e-06,
+      "loss": 0.5735,
+      "step": 9114
+    },
+    {
+      "epoch": 2.419753086419753,
+      "grad_norm": 0.4338187539239229,
+      "learning_rate": 3.2670285446916973e-06,
+      "loss": 0.5582,
+      "step": 9115
+    },
+    {
+      "epoch": 2.420018584893137,
+      "grad_norm": 0.4290222017807178,
+      "learning_rate": 3.2666962476213004e-06,
+      "loss": 0.5588,
+      "step": 9116
+    },
+    {
+      "epoch": 2.4202840833665205,
+      "grad_norm": 0.4338441383791346,
+      "learning_rate": 3.26636393559845e-06,
+      "loss": 0.5612,
+      "step": 9117
+    },
+    {
+      "epoch": 2.4205495818399045,
+      "grad_norm": 0.4452245653747792,
+      "learning_rate": 3.2660316086296263e-06,
+      "loss": 0.5682,
+      "step": 9118
+    },
+    {
+      "epoch": 2.420815080313288,
+      "grad_norm": 0.4337923082097538,
+      "learning_rate": 3.2656992667213104e-06,
+      "loss": 0.5374,
+      "step": 9119
+    },
+    {
+      "epoch": 2.421080578786672,
+      "grad_norm": 0.435562242885739,
+      "learning_rate": 3.265366909879984e-06,
+      "loss": 0.5827,
+      "step": 9120
+    },
+    {
+      "epoch": 2.421346077260056,
+      "grad_norm": 0.42994944474585034,
+      "learning_rate": 3.265034538112129e-06,
+      "loss": 0.5853,
+      "step": 9121
+    },
+    {
+      "epoch": 2.4216115757334395,
+      "grad_norm": 0.43472687325129333,
+      "learning_rate": 3.2647021514242277e-06,
+      "loss": 0.5361,
+      "step": 9122
+    },
+    {
+      "epoch": 2.421877074206823,
+      "grad_norm": 0.4289478879143833,
+      "learning_rate": 3.2643697498227624e-06,
+      "loss": 0.5964,
+      "step": 9123
+    },
+    {
+      "epoch": 2.4221425726802073,
+      "grad_norm": 0.4337055661130339,
+      "learning_rate": 3.264037333314215e-06,
+      "loss": 0.5838,
+      "step": 9124
+    },
+    {
+      "epoch": 2.422408071153591,
+      "grad_norm": 0.42932907827676353,
+      "learning_rate": 3.2637049019050688e-06,
+      "loss": 0.5704,
+      "step": 9125
+    },
+    {
+      "epoch": 2.4226735696269746,
+      "grad_norm": 0.435762154336256,
+      "learning_rate": 3.2633724556018077e-06,
+      "loss": 0.5502,
+      "step": 9126
+    },
+    {
+      "epoch": 2.4229390681003586,
+      "grad_norm": 0.44049891373842254,
+      "learning_rate": 3.263039994410914e-06,
+      "loss": 0.5737,
+      "step": 9127
+    },
+    {
+      "epoch": 2.4232045665737423,
+      "grad_norm": 0.43089527834065355,
+      "learning_rate": 3.2627075183388725e-06,
+      "loss": 0.5829,
+      "step": 9128
+    },
+    {
+      "epoch": 2.423470065047126,
+      "grad_norm": 0.4247600803728179,
+      "learning_rate": 3.2623750273921665e-06,
+      "loss": 0.5567,
+      "step": 9129
+    },
+    {
+      "epoch": 2.42373556352051,
+      "grad_norm": 0.4297256338559161,
+      "learning_rate": 3.262042521577281e-06,
+      "loss": 0.5603,
+      "step": 9130
+    },
+    {
+      "epoch": 2.4240010619938936,
+      "grad_norm": 0.4294093050239711,
+      "learning_rate": 3.2617100009007013e-06,
+      "loss": 0.5511,
+      "step": 9131
+    },
+    {
+      "epoch": 2.4242665604672773,
+      "grad_norm": 0.44297036894908215,
+      "learning_rate": 3.2613774653689113e-06,
+      "loss": 0.5344,
+      "step": 9132
+    },
+    {
+      "epoch": 2.424532058940661,
+      "grad_norm": 0.4279206395712597,
+      "learning_rate": 3.261044914988396e-06,
+      "loss": 0.5947,
+      "step": 9133
+    },
+    {
+      "epoch": 2.424797557414045,
+      "grad_norm": 0.44537927557467294,
+      "learning_rate": 3.260712349765642e-06,
+      "loss": 0.5689,
+      "step": 9134
+    },
+    {
+      "epoch": 2.4250630558874287,
+      "grad_norm": 0.42594071051467025,
+      "learning_rate": 3.260379769707135e-06,
+      "loss": 0.5379,
+      "step": 9135
+    },
+    {
+      "epoch": 2.4253285543608123,
+      "grad_norm": 0.4491248209290099,
+      "learning_rate": 3.2600471748193595e-06,
+      "loss": 0.6012,
+      "step": 9136
+    },
+    {
+      "epoch": 2.4255940528341964,
+      "grad_norm": 0.4431843553721626,
+      "learning_rate": 3.259714565108805e-06,
+      "loss": 0.5738,
+      "step": 9137
+    },
+    {
+      "epoch": 2.42585955130758,
+      "grad_norm": 0.45131846084871735,
+      "learning_rate": 3.2593819405819538e-06,
+      "loss": 0.5552,
+      "step": 9138
+    },
+    {
+      "epoch": 2.4261250497809637,
+      "grad_norm": 0.4511241032382652,
+      "learning_rate": 3.2590493012452974e-06,
+      "loss": 0.5733,
+      "step": 9139
+    },
+    {
+      "epoch": 2.4263905482543473,
+      "grad_norm": 0.441950605196848,
+      "learning_rate": 3.258716647105321e-06,
+      "loss": 0.5886,
+      "step": 9140
+    },
+    {
+      "epoch": 2.4266560467277314,
+      "grad_norm": 0.42335291292428895,
+      "learning_rate": 3.258383978168512e-06,
+      "loss": 0.567,
+      "step": 9141
+    },
+    {
+      "epoch": 2.426921545201115,
+      "grad_norm": 0.4368039505158051,
+      "learning_rate": 3.2580512944413583e-06,
+      "loss": 0.5435,
+      "step": 9142
+    },
+    {
+      "epoch": 2.4271870436744987,
+      "grad_norm": 0.4472270859809179,
+      "learning_rate": 3.257718595930349e-06,
+      "loss": 0.5874,
+      "step": 9143
+    },
+    {
+      "epoch": 2.4274525421478828,
+      "grad_norm": 0.4262819257228624,
+      "learning_rate": 3.257385882641971e-06,
+      "loss": 0.5089,
+      "step": 9144
+    },
+    {
+      "epoch": 2.4277180406212664,
+      "grad_norm": 0.44965705749988305,
+      "learning_rate": 3.2570531545827144e-06,
+      "loss": 0.5876,
+      "step": 9145
+    },
+    {
+      "epoch": 2.42798353909465,
+      "grad_norm": 0.42922730245834806,
+      "learning_rate": 3.2567204117590683e-06,
+      "loss": 0.5656,
+      "step": 9146
+    },
+    {
+      "epoch": 2.428249037568034,
+      "grad_norm": 0.43639259900445376,
+      "learning_rate": 3.256387654177521e-06,
+      "loss": 0.5983,
+      "step": 9147
+    },
+    {
+      "epoch": 2.4285145360414178,
+      "grad_norm": 0.4407913731239791,
+      "learning_rate": 3.2560548818445615e-06,
+      "loss": 0.5635,
+      "step": 9148
+    },
+    {
+      "epoch": 2.4287800345148014,
+      "grad_norm": 0.43014165142758554,
+      "learning_rate": 3.255722094766681e-06,
+      "loss": 0.5978,
+      "step": 9149
+    },
+    {
+      "epoch": 2.4290455329881855,
+      "grad_norm": 0.43011592406325894,
+      "learning_rate": 3.25538929295037e-06,
+      "loss": 0.5383,
+      "step": 9150
+    },
+    {
+      "epoch": 2.429311031461569,
+      "grad_norm": 0.43945246163923596,
+      "learning_rate": 3.255056476402118e-06,
+      "loss": 0.5486,
+      "step": 9151
+    },
+    {
+      "epoch": 2.4295765299349528,
+      "grad_norm": 0.44358605750138386,
+      "learning_rate": 3.2547236451284166e-06,
+      "loss": 0.599,
+      "step": 9152
+    },
+    {
+      "epoch": 2.429842028408337,
+      "grad_norm": 0.44200632162198994,
+      "learning_rate": 3.2543907991357556e-06,
+      "loss": 0.5751,
+      "step": 9153
+    },
+    {
+      "epoch": 2.4301075268817205,
+      "grad_norm": 0.43935672980888524,
+      "learning_rate": 3.254057938430627e-06,
+      "loss": 0.5903,
+      "step": 9154
+    },
+    {
+      "epoch": 2.430373025355104,
+      "grad_norm": 0.43910250531319545,
+      "learning_rate": 3.2537250630195227e-06,
+      "loss": 0.539,
+      "step": 9155
+    },
+    {
+      "epoch": 2.430638523828488,
+      "grad_norm": 0.44984755203911037,
+      "learning_rate": 3.2533921729089333e-06,
+      "loss": 0.5509,
+      "step": 9156
+    },
+    {
+      "epoch": 2.430904022301872,
+      "grad_norm": 0.42386614224297936,
+      "learning_rate": 3.253059268105353e-06,
+      "loss": 0.5763,
+      "step": 9157
+    },
+    {
+      "epoch": 2.4311695207752555,
+      "grad_norm": 0.4308931498701452,
+      "learning_rate": 3.252726348615272e-06,
+      "loss": 0.5434,
+      "step": 9158
+    },
+    {
+      "epoch": 2.431435019248639,
+      "grad_norm": 0.44955341875971067,
+      "learning_rate": 3.2523934144451853e-06,
+      "loss": 0.5342,
+      "step": 9159
+    },
+    {
+      "epoch": 2.4317005177220232,
+      "grad_norm": 0.4340219625958674,
+      "learning_rate": 3.2520604656015846e-06,
+      "loss": 0.5911,
+      "step": 9160
+    },
+    {
+      "epoch": 2.431966016195407,
+      "grad_norm": 0.4344027018014918,
+      "learning_rate": 3.2517275020909635e-06,
+      "loss": 0.5725,
+      "step": 9161
+    },
+    {
+      "epoch": 2.4322315146687905,
+      "grad_norm": 0.42813203679916667,
+      "learning_rate": 3.251394523919815e-06,
+      "loss": 0.5605,
+      "step": 9162
+    },
+    {
+      "epoch": 2.4324970131421746,
+      "grad_norm": 0.42889217798515633,
+      "learning_rate": 3.2510615310946343e-06,
+      "loss": 0.5781,
+      "step": 9163
+    },
+    {
+      "epoch": 2.4327625116155582,
+      "grad_norm": 0.43295123436598354,
+      "learning_rate": 3.2507285236219143e-06,
+      "loss": 0.546,
+      "step": 9164
+    },
+    {
+      "epoch": 2.433028010088942,
+      "grad_norm": 0.4391380261475172,
+      "learning_rate": 3.2503955015081502e-06,
+      "loss": 0.5816,
+      "step": 9165
+    },
+    {
+      "epoch": 2.433293508562326,
+      "grad_norm": 0.42916397306508125,
+      "learning_rate": 3.2500624647598362e-06,
+      "loss": 0.5743,
+      "step": 9166
+    },
+    {
+      "epoch": 2.4335590070357096,
+      "grad_norm": 0.44002645228221976,
+      "learning_rate": 3.2497294133834678e-06,
+      "loss": 0.5162,
+      "step": 9167
+    },
+    {
+      "epoch": 2.4338245055090932,
+      "grad_norm": 0.45737603065659477,
+      "learning_rate": 3.24939634738554e-06,
+      "loss": 0.5432,
+      "step": 9168
+    },
+    {
+      "epoch": 2.434090003982477,
+      "grad_norm": 0.42153568244350414,
+      "learning_rate": 3.24906326677255e-06,
+      "loss": 0.559,
+      "step": 9169
+    },
+    {
+      "epoch": 2.434355502455861,
+      "grad_norm": 0.42851174395653957,
+      "learning_rate": 3.2487301715509905e-06,
+      "loss": 0.5219,
+      "step": 9170
+    },
+    {
+      "epoch": 2.4346210009292446,
+      "grad_norm": 0.4452406195296684,
+      "learning_rate": 3.2483970617273602e-06,
+      "loss": 0.5739,
+      "step": 9171
+    },
+    {
+      "epoch": 2.4348864994026282,
+      "grad_norm": 0.44360428533414653,
+      "learning_rate": 3.248063937308155e-06,
+      "loss": 0.541,
+      "step": 9172
+    },
+    {
+      "epoch": 2.4351519978760123,
+      "grad_norm": 0.4234600100173597,
+      "learning_rate": 3.247730798299871e-06,
+      "loss": 0.5351,
+      "step": 9173
+    },
+    {
+      "epoch": 2.435417496349396,
+      "grad_norm": 0.4277662324604862,
+      "learning_rate": 3.247397644709006e-06,
+      "loss": 0.5555,
+      "step": 9174
+    },
+    {
+      "epoch": 2.4356829948227796,
+      "grad_norm": 0.421594085592504,
+      "learning_rate": 3.2470644765420567e-06,
+      "loss": 0.5322,
+      "step": 9175
+    },
+    {
+      "epoch": 2.4359484932961637,
+      "grad_norm": 0.43948652679330147,
+      "learning_rate": 3.246731293805521e-06,
+      "loss": 0.5917,
+      "step": 9176
+    },
+    {
+      "epoch": 2.4362139917695473,
+      "grad_norm": 0.4204130077193829,
+      "learning_rate": 3.2463980965058966e-06,
+      "loss": 0.554,
+      "step": 9177
+    },
+    {
+      "epoch": 2.436479490242931,
+      "grad_norm": 0.44448535970745134,
+      "learning_rate": 3.246064884649682e-06,
+      "loss": 0.5801,
+      "step": 9178
+    },
+    {
+      "epoch": 2.436744988716315,
+      "grad_norm": 0.4356196617978778,
+      "learning_rate": 3.245731658243376e-06,
+      "loss": 0.5756,
+      "step": 9179
+    },
+    {
+      "epoch": 2.4370104871896987,
+      "grad_norm": 0.42773727166948444,
+      "learning_rate": 3.2453984172934765e-06,
+      "loss": 0.5471,
+      "step": 9180
+    },
+    {
+      "epoch": 2.4372759856630823,
+      "grad_norm": 0.42462849250878887,
+      "learning_rate": 3.2450651618064827e-06,
+      "loss": 0.5616,
+      "step": 9181
+    },
+    {
+      "epoch": 2.4375414841364664,
+      "grad_norm": 0.4339715556695152,
+      "learning_rate": 3.2447318917888937e-06,
+      "loss": 0.5252,
+      "step": 9182
+    },
+    {
+      "epoch": 2.43780698260985,
+      "grad_norm": 0.4307919473090889,
+      "learning_rate": 3.24439860724721e-06,
+      "loss": 0.5767,
+      "step": 9183
+    },
+    {
+      "epoch": 2.4380724810832337,
+      "grad_norm": 0.43769399065216524,
+      "learning_rate": 3.2440653081879303e-06,
+      "loss": 0.5815,
+      "step": 9184
+    },
+    {
+      "epoch": 2.438337979556618,
+      "grad_norm": 0.4302797163901531,
+      "learning_rate": 3.243731994617555e-06,
+      "loss": 0.5465,
+      "step": 9185
+    },
+    {
+      "epoch": 2.4386034780300014,
+      "grad_norm": 0.4281632483599945,
+      "learning_rate": 3.2433986665425858e-06,
+      "loss": 0.5909,
+      "step": 9186
+    },
+    {
+      "epoch": 2.438868976503385,
+      "grad_norm": 0.44648244976525125,
+      "learning_rate": 3.243065323969522e-06,
+      "loss": 0.5673,
+      "step": 9187
+    },
+    {
+      "epoch": 2.4391344749767687,
+      "grad_norm": 0.43677578178066223,
+      "learning_rate": 3.242731966904865e-06,
+      "loss": 0.5522,
+      "step": 9188
+    },
+    {
+      "epoch": 2.439399973450153,
+      "grad_norm": 0.4337940304001195,
+      "learning_rate": 3.242398595355116e-06,
+      "loss": 0.5383,
+      "step": 9189
+    },
+    {
+      "epoch": 2.4396654719235364,
+      "grad_norm": 0.44634964527842813,
+      "learning_rate": 3.2420652093267774e-06,
+      "loss": 0.594,
+      "step": 9190
+    },
+    {
+      "epoch": 2.43993097039692,
+      "grad_norm": 0.4417383135491838,
+      "learning_rate": 3.2417318088263492e-06,
+      "loss": 0.5687,
+      "step": 9191
+    },
+    {
+      "epoch": 2.440196468870304,
+      "grad_norm": 0.4260931325446069,
+      "learning_rate": 3.241398393860336e-06,
+      "loss": 0.5581,
+      "step": 9192
+    },
+    {
+      "epoch": 2.440461967343688,
+      "grad_norm": 0.43122967616835856,
+      "learning_rate": 3.2410649644352383e-06,
+      "loss": 0.5496,
+      "step": 9193
+    },
+    {
+      "epoch": 2.4407274658170715,
+      "grad_norm": 0.42390560119565857,
+      "learning_rate": 3.2407315205575583e-06,
+      "loss": 0.5513,
+      "step": 9194
+    },
+    {
+      "epoch": 2.440992964290455,
+      "grad_norm": 0.45586334387375094,
+      "learning_rate": 3.2403980622338017e-06,
+      "loss": 0.5817,
+      "step": 9195
+    },
+    {
+      "epoch": 2.441258462763839,
+      "grad_norm": 0.43608430469160275,
+      "learning_rate": 3.2400645894704687e-06,
+      "loss": 0.5737,
+      "step": 9196
+    },
+    {
+      "epoch": 2.441523961237223,
+      "grad_norm": 0.43477080666778983,
+      "learning_rate": 3.2397311022740647e-06,
+      "loss": 0.5604,
+      "step": 9197
+    },
+    {
+      "epoch": 2.4417894597106065,
+      "grad_norm": 0.4387187335494614,
+      "learning_rate": 3.2393976006510935e-06,
+      "loss": 0.575,
+      "step": 9198
+    },
+    {
+      "epoch": 2.4420549581839905,
+      "grad_norm": 0.43094303753061264,
+      "learning_rate": 3.2390640846080585e-06,
+      "loss": 0.5702,
+      "step": 9199
+    },
+    {
+      "epoch": 2.442320456657374,
+      "grad_norm": 0.4312671545112207,
+      "learning_rate": 3.2387305541514636e-06,
+      "loss": 0.5564,
+      "step": 9200
+    },
+    {
+      "epoch": 2.442585955130758,
+      "grad_norm": 0.44552056122902023,
+      "learning_rate": 3.2383970092878153e-06,
+      "loss": 0.578,
+      "step": 9201
+    },
+    {
+      "epoch": 2.442851453604142,
+      "grad_norm": 0.4346805979038832,
+      "learning_rate": 3.2380634500236167e-06,
+      "loss": 0.5942,
+      "step": 9202
+    },
+    {
+      "epoch": 2.4431169520775255,
+      "grad_norm": 0.4269817133498969,
+      "learning_rate": 3.2377298763653735e-06,
+      "loss": 0.5552,
+      "step": 9203
+    },
+    {
+      "epoch": 2.443382450550909,
+      "grad_norm": 0.4396154861105833,
+      "learning_rate": 3.237396288319592e-06,
+      "loss": 0.5455,
+      "step": 9204
+    },
+    {
+      "epoch": 2.4436479490242933,
+      "grad_norm": 0.4438804392459179,
+      "learning_rate": 3.2370626858927773e-06,
+      "loss": 0.5639,
+      "step": 9205
+    },
+    {
+      "epoch": 2.443913447497677,
+      "grad_norm": 0.4372374444359718,
+      "learning_rate": 3.2367290690914357e-06,
+      "loss": 0.5525,
+      "step": 9206
+    },
+    {
+      "epoch": 2.4441789459710606,
+      "grad_norm": 0.44649241767222164,
+      "learning_rate": 3.236395437922073e-06,
+      "loss": 0.5788,
+      "step": 9207
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.429693238551958,
+      "learning_rate": 3.2360617923911964e-06,
+      "loss": 0.556,
+      "step": 9208
+    },
+    {
+      "epoch": 2.4447099429178283,
+      "grad_norm": 0.4143491171039682,
+      "learning_rate": 3.2357281325053138e-06,
+      "loss": 0.5287,
+      "step": 9209
+    },
+    {
+      "epoch": 2.444975441391212,
+      "grad_norm": 0.43596035684381934,
+      "learning_rate": 3.2353944582709294e-06,
+      "loss": 0.5707,
+      "step": 9210
+    },
+    {
+      "epoch": 2.445240939864596,
+      "grad_norm": 0.44351936327665104,
+      "learning_rate": 3.2350607696945535e-06,
+      "loss": 0.5761,
+      "step": 9211
+    },
+    {
+      "epoch": 2.4455064383379796,
+      "grad_norm": 0.4452840504611613,
+      "learning_rate": 3.2347270667826926e-06,
+      "loss": 0.5695,
+      "step": 9212
+    },
+    {
+      "epoch": 2.4457719368113633,
+      "grad_norm": 0.4267308160494901,
+      "learning_rate": 3.2343933495418544e-06,
+      "loss": 0.5665,
+      "step": 9213
+    },
+    {
+      "epoch": 2.446037435284747,
+      "grad_norm": 0.42609573495694336,
+      "learning_rate": 3.234059617978549e-06,
+      "loss": 0.542,
+      "step": 9214
+    },
+    {
+      "epoch": 2.446302933758131,
+      "grad_norm": 0.43217625536679605,
+      "learning_rate": 3.233725872099283e-06,
+      "loss": 0.5568,
+      "step": 9215
+    },
+    {
+      "epoch": 2.4465684322315147,
+      "grad_norm": 0.4377406039584305,
+      "learning_rate": 3.233392111910566e-06,
+      "loss": 0.546,
+      "step": 9216
+    },
+    {
+      "epoch": 2.4468339307048983,
+      "grad_norm": 0.43513108966793584,
+      "learning_rate": 3.2330583374189073e-06,
+      "loss": 0.5724,
+      "step": 9217
+    },
+    {
+      "epoch": 2.4470994291782824,
+      "grad_norm": 0.4217823896548075,
+      "learning_rate": 3.232724548630817e-06,
+      "loss": 0.5281,
+      "step": 9218
+    },
+    {
+      "epoch": 2.447364927651666,
+      "grad_norm": 0.42926041257157305,
+      "learning_rate": 3.232390745552803e-06,
+      "loss": 0.562,
+      "step": 9219
+    },
+    {
+      "epoch": 2.4476304261250497,
+      "grad_norm": 0.4319942277111498,
+      "learning_rate": 3.232056928191376e-06,
+      "loss": 0.5674,
+      "step": 9220
+    },
+    {
+      "epoch": 2.4478959245984337,
+      "grad_norm": 0.4295877169921538,
+      "learning_rate": 3.231723096553047e-06,
+      "loss": 0.5428,
+      "step": 9221
+    },
+    {
+      "epoch": 2.4481614230718174,
+      "grad_norm": 0.4440776621133491,
+      "learning_rate": 3.2313892506443264e-06,
+      "loss": 0.5902,
+      "step": 9222
+    },
+    {
+      "epoch": 2.448426921545201,
+      "grad_norm": 0.45221098831232626,
+      "learning_rate": 3.231055390471724e-06,
+      "loss": 0.5733,
+      "step": 9223
+    },
+    {
+      "epoch": 2.4486924200185847,
+      "grad_norm": 0.435474292924806,
+      "learning_rate": 3.230721516041752e-06,
+      "loss": 0.6027,
+      "step": 9224
+    },
+    {
+      "epoch": 2.4489579184919688,
+      "grad_norm": 0.4184631013785383,
+      "learning_rate": 3.2303876273609213e-06,
+      "loss": 0.5599,
+      "step": 9225
+    },
+    {
+      "epoch": 2.4492234169653524,
+      "grad_norm": 0.44857414886613467,
+      "learning_rate": 3.2300537244357435e-06,
+      "loss": 0.5273,
+      "step": 9226
+    },
+    {
+      "epoch": 2.449488915438736,
+      "grad_norm": 0.4333410469894492,
+      "learning_rate": 3.2297198072727308e-06,
+      "loss": 0.5771,
+      "step": 9227
+    },
+    {
+      "epoch": 2.44975441391212,
+      "grad_norm": 0.4375320261939352,
+      "learning_rate": 3.229385875878395e-06,
+      "loss": 0.6016,
+      "step": 9228
+    },
+    {
+      "epoch": 2.4500199123855038,
+      "grad_norm": 0.4384651629033774,
+      "learning_rate": 3.229051930259248e-06,
+      "loss": 0.5232,
+      "step": 9229
+    },
+    {
+      "epoch": 2.4502854108588874,
+      "grad_norm": 0.43918455817330104,
+      "learning_rate": 3.228717970421804e-06,
+      "loss": 0.5786,
+      "step": 9230
+    },
+    {
+      "epoch": 2.4505509093322715,
+      "grad_norm": 0.4273802015053016,
+      "learning_rate": 3.228383996372576e-06,
+      "loss": 0.5243,
+      "step": 9231
+    },
+    {
+      "epoch": 2.450816407805655,
+      "grad_norm": 0.44214970182529034,
+      "learning_rate": 3.2280500081180764e-06,
+      "loss": 0.5905,
+      "step": 9232
+    },
+    {
+      "epoch": 2.4510819062790388,
+      "grad_norm": 0.4379022413759155,
+      "learning_rate": 3.2277160056648183e-06,
+      "loss": 0.5646,
+      "step": 9233
+    },
+    {
+      "epoch": 2.451347404752423,
+      "grad_norm": 0.44213035769546577,
+      "learning_rate": 3.227381989019317e-06,
+      "loss": 0.5456,
+      "step": 9234
+    },
+    {
+      "epoch": 2.4516129032258065,
+      "grad_norm": 0.44355785006915593,
+      "learning_rate": 3.227047958188086e-06,
+      "loss": 0.5897,
+      "step": 9235
+    },
+    {
+      "epoch": 2.45187840169919,
+      "grad_norm": 0.4399791434706225,
+      "learning_rate": 3.2267139131776392e-06,
+      "loss": 0.5868,
+      "step": 9236
+    },
+    {
+      "epoch": 2.452143900172574,
+      "grad_norm": 0.4379368177073412,
+      "learning_rate": 3.2263798539944923e-06,
+      "loss": 0.5834,
+      "step": 9237
+    },
+    {
+      "epoch": 2.452409398645958,
+      "grad_norm": 0.41958616275063965,
+      "learning_rate": 3.2260457806451593e-06,
+      "loss": 0.5344,
+      "step": 9238
+    },
+    {
+      "epoch": 2.4526748971193415,
+      "grad_norm": 0.4400284390552846,
+      "learning_rate": 3.225711693136156e-06,
+      "loss": 0.5922,
+      "step": 9239
+    },
+    {
+      "epoch": 2.4529403955927256,
+      "grad_norm": 0.43673061488650344,
+      "learning_rate": 3.225377591473998e-06,
+      "loss": 0.5647,
+      "step": 9240
+    },
+    {
+      "epoch": 2.4532058940661092,
+      "grad_norm": 0.4216282446509674,
+      "learning_rate": 3.225043475665201e-06,
+      "loss": 0.566,
+      "step": 9241
+    },
+    {
+      "epoch": 2.453471392539493,
+      "grad_norm": 0.4337536443596227,
+      "learning_rate": 3.224709345716281e-06,
+      "loss": 0.5634,
+      "step": 9242
+    },
+    {
+      "epoch": 2.4537368910128765,
+      "grad_norm": 0.433214091004987,
+      "learning_rate": 3.2243752016337544e-06,
+      "loss": 0.6068,
+      "step": 9243
+    },
+    {
+      "epoch": 2.4540023894862606,
+      "grad_norm": 0.44678129969466157,
+      "learning_rate": 3.2240410434241376e-06,
+      "loss": 0.5503,
+      "step": 9244
+    },
+    {
+      "epoch": 2.4542678879596442,
+      "grad_norm": 0.43895188265329577,
+      "learning_rate": 3.223706871093948e-06,
+      "loss": 0.6122,
+      "step": 9245
+    },
+    {
+      "epoch": 2.454533386433028,
+      "grad_norm": 0.43941213832308157,
+      "learning_rate": 3.223372684649702e-06,
+      "loss": 0.5884,
+      "step": 9246
+    },
+    {
+      "epoch": 2.454798884906412,
+      "grad_norm": 0.432795591779766,
+      "learning_rate": 3.2230384840979184e-06,
+      "loss": 0.5592,
+      "step": 9247
+    },
+    {
+      "epoch": 2.4550643833797956,
+      "grad_norm": 0.4346904827910939,
+      "learning_rate": 3.2227042694451135e-06,
+      "loss": 0.5392,
+      "step": 9248
+    },
+    {
+      "epoch": 2.4553298818531792,
+      "grad_norm": 0.4301518330707813,
+      "learning_rate": 3.2223700406978052e-06,
+      "loss": 0.5844,
+      "step": 9249
+    },
+    {
+      "epoch": 2.455595380326563,
+      "grad_norm": 0.43143760152867194,
+      "learning_rate": 3.222035797862513e-06,
+      "loss": 0.5172,
+      "step": 9250
+    },
+    {
+      "epoch": 2.455860878799947,
+      "grad_norm": 0.4275449969586369,
+      "learning_rate": 3.2217015409457547e-06,
+      "loss": 0.5437,
+      "step": 9251
+    },
+    {
+      "epoch": 2.4561263772733306,
+      "grad_norm": 0.4361445645580946,
+      "learning_rate": 3.22136726995405e-06,
+      "loss": 0.5353,
+      "step": 9252
+    },
+    {
+      "epoch": 2.4563918757467142,
+      "grad_norm": 0.43985169949558883,
+      "learning_rate": 3.221032984893917e-06,
+      "loss": 0.5873,
+      "step": 9253
+    },
+    {
+      "epoch": 2.4566573742200983,
+      "grad_norm": 0.43839727355294944,
+      "learning_rate": 3.2206986857718754e-06,
+      "loss": 0.5801,
+      "step": 9254
+    },
+    {
+      "epoch": 2.456922872693482,
+      "grad_norm": 0.4298509354403553,
+      "learning_rate": 3.220364372594445e-06,
+      "loss": 0.577,
+      "step": 9255
+    },
+    {
+      "epoch": 2.4571883711668656,
+      "grad_norm": 0.4508958566467277,
+      "learning_rate": 3.2200300453681448e-06,
+      "loss": 0.5888,
+      "step": 9256
+    },
+    {
+      "epoch": 2.4574538696402497,
+      "grad_norm": 0.4401538693756923,
+      "learning_rate": 3.2196957040994962e-06,
+      "loss": 0.5808,
+      "step": 9257
+    },
+    {
+      "epoch": 2.4577193681136333,
+      "grad_norm": 0.4462528574714819,
+      "learning_rate": 3.219361348795019e-06,
+      "loss": 0.5264,
+      "step": 9258
+    },
+    {
+      "epoch": 2.457984866587017,
+      "grad_norm": 0.4311772317729529,
+      "learning_rate": 3.219026979461235e-06,
+      "loss": 0.538,
+      "step": 9259
+    },
+    {
+      "epoch": 2.458250365060401,
+      "grad_norm": 0.43117471150653824,
+      "learning_rate": 3.2186925961046632e-06,
+      "loss": 0.5352,
+      "step": 9260
+    },
+    {
+      "epoch": 2.4585158635337847,
+      "grad_norm": 0.447568904126334,
+      "learning_rate": 3.218358198731828e-06,
+      "loss": 0.5774,
+      "step": 9261
+    },
+    {
+      "epoch": 2.4587813620071683,
+      "grad_norm": 0.4380905548917899,
+      "learning_rate": 3.2180237873492477e-06,
+      "loss": 0.5632,
+      "step": 9262
+    },
+    {
+      "epoch": 2.4590468604805524,
+      "grad_norm": 0.43910650805026535,
+      "learning_rate": 3.2176893619634453e-06,
+      "loss": 0.5825,
+      "step": 9263
+    },
+    {
+      "epoch": 2.459312358953936,
+      "grad_norm": 0.43907316138183544,
+      "learning_rate": 3.2173549225809444e-06,
+      "loss": 0.5557,
+      "step": 9264
+    },
+    {
+      "epoch": 2.4595778574273197,
+      "grad_norm": 0.44960943712798707,
+      "learning_rate": 3.217020469208265e-06,
+      "loss": 0.5898,
+      "step": 9265
+    },
+    {
+      "epoch": 2.459843355900704,
+      "grad_norm": 0.4260107522996064,
+      "learning_rate": 3.216686001851931e-06,
+      "loss": 0.5554,
+      "step": 9266
+    },
+    {
+      "epoch": 2.4601088543740874,
+      "grad_norm": 0.4316678549676487,
+      "learning_rate": 3.216351520518465e-06,
+      "loss": 0.5832,
+      "step": 9267
+    },
+    {
+      "epoch": 2.460374352847471,
+      "grad_norm": 0.4345900595714598,
+      "learning_rate": 3.2160170252143913e-06,
+      "loss": 0.5495,
+      "step": 9268
+    },
+    {
+      "epoch": 2.4606398513208547,
+      "grad_norm": 0.44487367474504164,
+      "learning_rate": 3.2156825159462317e-06,
+      "loss": 0.5696,
+      "step": 9269
+    },
+    {
+      "epoch": 2.460905349794239,
+      "grad_norm": 0.44008919471229413,
+      "learning_rate": 3.2153479927205116e-06,
+      "loss": 0.5757,
+      "step": 9270
+    },
+    {
+      "epoch": 2.4611708482676224,
+      "grad_norm": 0.44487969973180636,
+      "learning_rate": 3.2150134555437536e-06,
+      "loss": 0.6114,
+      "step": 9271
+    },
+    {
+      "epoch": 2.461436346741006,
+      "grad_norm": 0.4456398133516107,
+      "learning_rate": 3.2146789044224823e-06,
+      "loss": 0.5522,
+      "step": 9272
+    },
+    {
+      "epoch": 2.46170184521439,
+      "grad_norm": 0.43313708646730037,
+      "learning_rate": 3.2143443393632233e-06,
+      "loss": 0.5587,
+      "step": 9273
+    },
+    {
+      "epoch": 2.461967343687774,
+      "grad_norm": 0.435056742891196,
+      "learning_rate": 3.2140097603725e-06,
+      "loss": 0.5755,
+      "step": 9274
+    },
+    {
+      "epoch": 2.4622328421611575,
+      "grad_norm": 0.4344219947204466,
+      "learning_rate": 3.2136751674568384e-06,
+      "loss": 0.5464,
+      "step": 9275
+    },
+    {
+      "epoch": 2.4624983406345415,
+      "grad_norm": 0.43564920120243994,
+      "learning_rate": 3.2133405606227636e-06,
+      "loss": 0.5337,
+      "step": 9276
+    },
+    {
+      "epoch": 2.462763839107925,
+      "grad_norm": 0.4148265194262022,
+      "learning_rate": 3.213005939876801e-06,
+      "loss": 0.5394,
+      "step": 9277
+    },
+    {
+      "epoch": 2.463029337581309,
+      "grad_norm": 0.4336315820065418,
+      "learning_rate": 3.212671305225478e-06,
+      "loss": 0.5534,
+      "step": 9278
+    },
+    {
+      "epoch": 2.4632948360546925,
+      "grad_norm": 0.44180895337030884,
+      "learning_rate": 3.2123366566753183e-06,
+      "loss": 0.5701,
+      "step": 9279
+    },
+    {
+      "epoch": 2.4635603345280765,
+      "grad_norm": 0.44135897544400393,
+      "learning_rate": 3.21200199423285e-06,
+      "loss": 0.5909,
+      "step": 9280
+    },
+    {
+      "epoch": 2.46382583300146,
+      "grad_norm": 0.4542709845579795,
+      "learning_rate": 3.2116673179046003e-06,
+      "loss": 0.5765,
+      "step": 9281
+    },
+    {
+      "epoch": 2.464091331474844,
+      "grad_norm": 0.4334304278213419,
+      "learning_rate": 3.2113326276970945e-06,
+      "loss": 0.5513,
+      "step": 9282
+    },
+    {
+      "epoch": 2.464356829948228,
+      "grad_norm": 0.46611268758444935,
+      "learning_rate": 3.2109979236168615e-06,
+      "loss": 0.6013,
+      "step": 9283
+    },
+    {
+      "epoch": 2.4646223284216116,
+      "grad_norm": 0.4330494564015539,
+      "learning_rate": 3.2106632056704277e-06,
+      "loss": 0.5481,
+      "step": 9284
+    },
+    {
+      "epoch": 2.464887826894995,
+      "grad_norm": 0.4314359068765981,
+      "learning_rate": 3.2103284738643216e-06,
+      "loss": 0.5728,
+      "step": 9285
+    },
+    {
+      "epoch": 2.4651533253683793,
+      "grad_norm": 0.437056312784691,
+      "learning_rate": 3.209993728205071e-06,
+      "loss": 0.5731,
+      "step": 9286
+    },
+    {
+      "epoch": 2.465418823841763,
+      "grad_norm": 0.4383553537706255,
+      "learning_rate": 3.2096589686992043e-06,
+      "loss": 0.5415,
+      "step": 9287
+    },
+    {
+      "epoch": 2.4656843223151466,
+      "grad_norm": 0.44151316454398964,
+      "learning_rate": 3.20932419535325e-06,
+      "loss": 0.5676,
+      "step": 9288
+    },
+    {
+      "epoch": 2.4659498207885306,
+      "grad_norm": 0.4336326588505913,
+      "learning_rate": 3.2089894081737373e-06,
+      "loss": 0.555,
+      "step": 9289
+    },
+    {
+      "epoch": 2.4662153192619143,
+      "grad_norm": 0.4231797704858733,
+      "learning_rate": 3.208654607167196e-06,
+      "loss": 0.5226,
+      "step": 9290
+    },
+    {
+      "epoch": 2.466480817735298,
+      "grad_norm": 0.4348513266093203,
+      "learning_rate": 3.2083197923401537e-06,
+      "loss": 0.5743,
+      "step": 9291
+    },
+    {
+      "epoch": 2.466746316208682,
+      "grad_norm": 0.4224671118271553,
+      "learning_rate": 3.2079849636991416e-06,
+      "loss": 0.5578,
+      "step": 9292
+    },
+    {
+      "epoch": 2.4670118146820657,
+      "grad_norm": 0.43936651309718844,
+      "learning_rate": 3.2076501212506896e-06,
+      "loss": 0.5783,
+      "step": 9293
+    },
+    {
+      "epoch": 2.4672773131554493,
+      "grad_norm": 0.45205520071898586,
+      "learning_rate": 3.2073152650013273e-06,
+      "loss": 0.5662,
+      "step": 9294
+    },
+    {
+      "epoch": 2.4675428116288334,
+      "grad_norm": 0.43418390292864895,
+      "learning_rate": 3.2069803949575845e-06,
+      "loss": 0.5409,
+      "step": 9295
+    },
+    {
+      "epoch": 2.467808310102217,
+      "grad_norm": 0.44506073932633294,
+      "learning_rate": 3.206645511125995e-06,
+      "loss": 0.5786,
+      "step": 9296
+    },
+    {
+      "epoch": 2.4680738085756007,
+      "grad_norm": 0.44820003310072165,
+      "learning_rate": 3.206310613513086e-06,
+      "loss": 0.5571,
+      "step": 9297
+    },
+    {
+      "epoch": 2.4683393070489843,
+      "grad_norm": 0.4663835187323188,
+      "learning_rate": 3.205975702125392e-06,
+      "loss": 0.5456,
+      "step": 9298
+    },
+    {
+      "epoch": 2.4686048055223684,
+      "grad_norm": 0.4306489245787134,
+      "learning_rate": 3.205640776969443e-06,
+      "loss": 0.557,
+      "step": 9299
+    },
+    {
+      "epoch": 2.468870303995752,
+      "grad_norm": 0.4231525121319539,
+      "learning_rate": 3.205305838051771e-06,
+      "loss": 0.5628,
+      "step": 9300
+    },
+    {
+      "epoch": 2.4691358024691357,
+      "grad_norm": 0.4418397927448869,
+      "learning_rate": 3.2049708853789087e-06,
+      "loss": 0.563,
+      "step": 9301
+    },
+    {
+      "epoch": 2.4694013009425198,
+      "grad_norm": 0.4368738420288982,
+      "learning_rate": 3.204635918957387e-06,
+      "loss": 0.5817,
+      "step": 9302
+    },
+    {
+      "epoch": 2.4696667994159034,
+      "grad_norm": 0.43947537258508945,
+      "learning_rate": 3.204300938793741e-06,
+      "loss": 0.586,
+      "step": 9303
+    },
+    {
+      "epoch": 2.469932297889287,
+      "grad_norm": 0.4315511554711722,
+      "learning_rate": 3.203965944894501e-06,
+      "loss": 0.516,
+      "step": 9304
+    },
+    {
+      "epoch": 2.4701977963626707,
+      "grad_norm": 0.44174830309953,
+      "learning_rate": 3.2036309372662017e-06,
+      "loss": 0.5764,
+      "step": 9305
+    },
+    {
+      "epoch": 2.4704632948360548,
+      "grad_norm": 0.42774219667033225,
+      "learning_rate": 3.2032959159153766e-06,
+      "loss": 0.563,
+      "step": 9306
+    },
+    {
+      "epoch": 2.4707287933094384,
+      "grad_norm": 0.43309016923940263,
+      "learning_rate": 3.2029608808485598e-06,
+      "loss": 0.5623,
+      "step": 9307
+    },
+    {
+      "epoch": 2.470994291782822,
+      "grad_norm": 0.44753072963382295,
+      "learning_rate": 3.2026258320722834e-06,
+      "loss": 0.5504,
+      "step": 9308
+    },
+    {
+      "epoch": 2.471259790256206,
+      "grad_norm": 0.4396184446589746,
+      "learning_rate": 3.2022907695930837e-06,
+      "loss": 0.5291,
+      "step": 9309
+    },
+    {
+      "epoch": 2.4715252887295898,
+      "grad_norm": 0.4304845146467045,
+      "learning_rate": 3.201955693417495e-06,
+      "loss": 0.5699,
+      "step": 9310
+    },
+    {
+      "epoch": 2.4717907872029734,
+      "grad_norm": 0.4508366669555417,
+      "learning_rate": 3.201620603552051e-06,
+      "loss": 0.5956,
+      "step": 9311
+    },
+    {
+      "epoch": 2.4720562856763575,
+      "grad_norm": 0.45239558825462195,
+      "learning_rate": 3.201285500003287e-06,
+      "loss": 0.5466,
+      "step": 9312
+    },
+    {
+      "epoch": 2.472321784149741,
+      "grad_norm": 0.4272271088422055,
+      "learning_rate": 3.200950382777739e-06,
+      "loss": 0.5467,
+      "step": 9313
+    },
+    {
+      "epoch": 2.4725872826231248,
+      "grad_norm": 0.4323833422230112,
+      "learning_rate": 3.2006152518819424e-06,
+      "loss": 0.5789,
+      "step": 9314
+    },
+    {
+      "epoch": 2.472852781096509,
+      "grad_norm": 0.43092476199328034,
+      "learning_rate": 3.200280107322433e-06,
+      "loss": 0.5515,
+      "step": 9315
+    },
+    {
+      "epoch": 2.4731182795698925,
+      "grad_norm": 0.45519324505825043,
+      "learning_rate": 3.1999449491057474e-06,
+      "loss": 0.564,
+      "step": 9316
+    },
+    {
+      "epoch": 2.473383778043276,
+      "grad_norm": 0.439487421697383,
+      "learning_rate": 3.1996097772384203e-06,
+      "loss": 0.5447,
+      "step": 9317
+    },
+    {
+      "epoch": 2.4736492765166602,
+      "grad_norm": 0.45040987974501273,
+      "learning_rate": 3.199274591726991e-06,
+      "loss": 0.587,
+      "step": 9318
+    },
+    {
+      "epoch": 2.473914774990044,
+      "grad_norm": 0.42470609183369723,
+      "learning_rate": 3.1989393925779943e-06,
+      "loss": 0.5461,
+      "step": 9319
+    },
+    {
+      "epoch": 2.4741802734634275,
+      "grad_norm": 0.44562647077999756,
+      "learning_rate": 3.1986041797979683e-06,
+      "loss": 0.5844,
+      "step": 9320
+    },
+    {
+      "epoch": 2.4744457719368116,
+      "grad_norm": 0.4532510946910471,
+      "learning_rate": 3.1982689533934496e-06,
+      "loss": 0.5506,
+      "step": 9321
+    },
+    {
+      "epoch": 2.4747112704101952,
+      "grad_norm": 0.44449700841515694,
+      "learning_rate": 3.1979337133709777e-06,
+      "loss": 0.6077,
+      "step": 9322
+    },
+    {
+      "epoch": 2.474976768883579,
+      "grad_norm": 0.44830119567768045,
+      "learning_rate": 3.197598459737089e-06,
+      "loss": 0.5515,
+      "step": 9323
+    },
+    {
+      "epoch": 2.4752422673569625,
+      "grad_norm": 0.42880806501016094,
+      "learning_rate": 3.1972631924983228e-06,
+      "loss": 0.5482,
+      "step": 9324
+    },
+    {
+      "epoch": 2.4755077658303466,
+      "grad_norm": 0.4357218750000735,
+      "learning_rate": 3.1969279116612163e-06,
+      "loss": 0.5407,
+      "step": 9325
+    },
+    {
+      "epoch": 2.4757732643037302,
+      "grad_norm": 0.43746853278861136,
+      "learning_rate": 3.1965926172323096e-06,
+      "loss": 0.5742,
+      "step": 9326
+    },
+    {
+      "epoch": 2.476038762777114,
+      "grad_norm": 0.43884137117245586,
+      "learning_rate": 3.196257309218142e-06,
+      "loss": 0.5813,
+      "step": 9327
+    },
+    {
+      "epoch": 2.476304261250498,
+      "grad_norm": 0.4501126219589327,
+      "learning_rate": 3.1959219876252513e-06,
+      "loss": 0.5757,
+      "step": 9328
+    },
+    {
+      "epoch": 2.4765697597238816,
+      "grad_norm": 0.44691863809592286,
+      "learning_rate": 3.195586652460178e-06,
+      "loss": 0.574,
+      "step": 9329
+    },
+    {
+      "epoch": 2.4768352581972652,
+      "grad_norm": 0.43685441452318735,
+      "learning_rate": 3.1952513037294624e-06,
+      "loss": 0.5855,
+      "step": 9330
+    },
+    {
+      "epoch": 2.4771007566706493,
+      "grad_norm": 0.43112116327421307,
+      "learning_rate": 3.194915941439643e-06,
+      "loss": 0.5424,
+      "step": 9331
+    },
+    {
+      "epoch": 2.477366255144033,
+      "grad_norm": 0.4361272198938356,
+      "learning_rate": 3.1945805655972624e-06,
+      "loss": 0.533,
+      "step": 9332
+    },
+    {
+      "epoch": 2.4776317536174166,
+      "grad_norm": 0.4345226729695114,
+      "learning_rate": 3.19424517620886e-06,
+      "loss": 0.5945,
+      "step": 9333
+    },
+    {
+      "epoch": 2.4778972520908003,
+      "grad_norm": 0.43966239949898617,
+      "learning_rate": 3.1939097732809763e-06,
+      "loss": 0.5375,
+      "step": 9334
+    },
+    {
+      "epoch": 2.4781627505641843,
+      "grad_norm": 0.4603663491252038,
+      "learning_rate": 3.193574356820153e-06,
+      "loss": 0.568,
+      "step": 9335
+    },
+    {
+      "epoch": 2.478428249037568,
+      "grad_norm": 0.45718711627010944,
+      "learning_rate": 3.193238926832933e-06,
+      "loss": 0.5819,
+      "step": 9336
+    },
+    {
+      "epoch": 2.4786937475109516,
+      "grad_norm": 0.4420346798540461,
+      "learning_rate": 3.1929034833258553e-06,
+      "loss": 0.5855,
+      "step": 9337
+    },
+    {
+      "epoch": 2.4789592459843357,
+      "grad_norm": 0.4708479232021837,
+      "learning_rate": 3.192568026305463e-06,
+      "loss": 0.5286,
+      "step": 9338
+    },
+    {
+      "epoch": 2.4792247444577193,
+      "grad_norm": 0.44519458042877974,
+      "learning_rate": 3.1922325557782998e-06,
+      "loss": 0.5792,
+      "step": 9339
+    },
+    {
+      "epoch": 2.479490242931103,
+      "grad_norm": 0.4757209214463942,
+      "learning_rate": 3.1918970717509057e-06,
+      "loss": 0.5602,
+      "step": 9340
+    },
+    {
+      "epoch": 2.479755741404487,
+      "grad_norm": 0.4462675605428505,
+      "learning_rate": 3.191561574229824e-06,
+      "loss": 0.5759,
+      "step": 9341
+    },
+    {
+      "epoch": 2.4800212398778707,
+      "grad_norm": 0.441487599107144,
+      "learning_rate": 3.1912260632215997e-06,
+      "loss": 0.5455,
+      "step": 9342
+    },
+    {
+      "epoch": 2.4802867383512543,
+      "grad_norm": 0.43613430422436594,
+      "learning_rate": 3.190890538732774e-06,
+      "loss": 0.5576,
+      "step": 9343
+    },
+    {
+      "epoch": 2.4805522368246384,
+      "grad_norm": 0.44867907732059414,
+      "learning_rate": 3.1905550007698914e-06,
+      "loss": 0.536,
+      "step": 9344
+    },
+    {
+      "epoch": 2.480817735298022,
+      "grad_norm": 0.43336200421984833,
+      "learning_rate": 3.190219449339496e-06,
+      "loss": 0.5843,
+      "step": 9345
+    },
+    {
+      "epoch": 2.4810832337714057,
+      "grad_norm": 0.43731371335480906,
+      "learning_rate": 3.189883884448131e-06,
+      "loss": 0.5264,
+      "step": 9346
+    },
+    {
+      "epoch": 2.48134873224479,
+      "grad_norm": 0.4270844799774336,
+      "learning_rate": 3.1895483061023413e-06,
+      "loss": 0.5391,
+      "step": 9347
+    },
+    {
+      "epoch": 2.4816142307181734,
+      "grad_norm": 0.43606513188398427,
+      "learning_rate": 3.1892127143086716e-06,
+      "loss": 0.5227,
+      "step": 9348
+    },
+    {
+      "epoch": 2.481879729191557,
+      "grad_norm": 0.42749928299191725,
+      "learning_rate": 3.188877109073666e-06,
+      "loss": 0.5679,
+      "step": 9349
+    },
+    {
+      "epoch": 2.482145227664941,
+      "grad_norm": 0.43431540937476915,
+      "learning_rate": 3.1885414904038704e-06,
+      "loss": 0.5344,
+      "step": 9350
+    },
+    {
+      "epoch": 2.482410726138325,
+      "grad_norm": 0.436061249834663,
+      "learning_rate": 3.1882058583058294e-06,
+      "loss": 0.5618,
+      "step": 9351
+    },
+    {
+      "epoch": 2.4826762246117084,
+      "grad_norm": 0.4377943226932406,
+      "learning_rate": 3.1878702127860893e-06,
+      "loss": 0.5795,
+      "step": 9352
+    },
+    {
+      "epoch": 2.482941723085092,
+      "grad_norm": 0.44184378059784135,
+      "learning_rate": 3.187534553851196e-06,
+      "loss": 0.5812,
+      "step": 9353
+    },
+    {
+      "epoch": 2.483207221558476,
+      "grad_norm": 0.43368017594018454,
+      "learning_rate": 3.1871988815076953e-06,
+      "loss": 0.5449,
+      "step": 9354
+    },
+    {
+      "epoch": 2.48347272003186,
+      "grad_norm": 0.4238069815154126,
+      "learning_rate": 3.1868631957621343e-06,
+      "loss": 0.5436,
+      "step": 9355
+    },
+    {
+      "epoch": 2.4837382185052435,
+      "grad_norm": 0.4260234172067541,
+      "learning_rate": 3.186527496621059e-06,
+      "loss": 0.5551,
+      "step": 9356
+    },
+    {
+      "epoch": 2.4840037169786275,
+      "grad_norm": 0.43921437028593535,
+      "learning_rate": 3.1861917840910166e-06,
+      "loss": 0.5564,
+      "step": 9357
+    },
+    {
+      "epoch": 2.484269215452011,
+      "grad_norm": 0.4394505298820809,
+      "learning_rate": 3.185856058178555e-06,
+      "loss": 0.5786,
+      "step": 9358
+    },
+    {
+      "epoch": 2.484534713925395,
+      "grad_norm": 0.4436985424507277,
+      "learning_rate": 3.1855203188902194e-06,
+      "loss": 0.5721,
+      "step": 9359
+    },
+    {
+      "epoch": 2.4848002123987785,
+      "grad_norm": 0.42585355772389516,
+      "learning_rate": 3.1851845662325602e-06,
+      "loss": 0.554,
+      "step": 9360
+    },
+    {
+      "epoch": 2.4850657108721625,
+      "grad_norm": 0.4383754774042248,
+      "learning_rate": 3.184848800212124e-06,
+      "loss": 0.5738,
+      "step": 9361
+    },
+    {
+      "epoch": 2.485331209345546,
+      "grad_norm": 0.43766013125967823,
+      "learning_rate": 3.1845130208354603e-06,
+      "loss": 0.5808,
+      "step": 9362
+    },
+    {
+      "epoch": 2.48559670781893,
+      "grad_norm": 0.43077167528659666,
+      "learning_rate": 3.184177228109116e-06,
+      "loss": 0.5685,
+      "step": 9363
+    },
+    {
+      "epoch": 2.485862206292314,
+      "grad_norm": 0.4407213907402458,
+      "learning_rate": 3.1838414220396407e-06,
+      "loss": 0.5962,
+      "step": 9364
+    },
+    {
+      "epoch": 2.4861277047656976,
+      "grad_norm": 0.4361684779228887,
+      "learning_rate": 3.183505602633583e-06,
+      "loss": 0.5379,
+      "step": 9365
+    },
+    {
+      "epoch": 2.486393203239081,
+      "grad_norm": 0.4307045462344444,
+      "learning_rate": 3.183169769897493e-06,
+      "loss": 0.5547,
+      "step": 9366
+    },
+    {
+      "epoch": 2.4866587017124653,
+      "grad_norm": 0.4389193712930424,
+      "learning_rate": 3.1828339238379195e-06,
+      "loss": 0.5629,
+      "step": 9367
+    },
+    {
+      "epoch": 2.486924200185849,
+      "grad_norm": 0.43812322007297244,
+      "learning_rate": 3.182498064461413e-06,
+      "loss": 0.5607,
+      "step": 9368
+    },
+    {
+      "epoch": 2.4871896986592326,
+      "grad_norm": 0.43177351334385666,
+      "learning_rate": 3.1821621917745227e-06,
+      "loss": 0.5378,
+      "step": 9369
+    },
+    {
+      "epoch": 2.4874551971326166,
+      "grad_norm": 0.42901807443409923,
+      "learning_rate": 3.1818263057837996e-06,
+      "loss": 0.5625,
+      "step": 9370
+    },
+    {
+      "epoch": 2.4877206956060003,
+      "grad_norm": 0.4212485000515425,
+      "learning_rate": 3.1814904064957943e-06,
+      "loss": 0.5417,
+      "step": 9371
+    },
+    {
+      "epoch": 2.487986194079384,
+      "grad_norm": 0.43914906144383814,
+      "learning_rate": 3.1811544939170573e-06,
+      "loss": 0.5685,
+      "step": 9372
+    },
+    {
+      "epoch": 2.488251692552768,
+      "grad_norm": 0.42594664621438766,
+      "learning_rate": 3.1808185680541404e-06,
+      "loss": 0.5591,
+      "step": 9373
+    },
+    {
+      "epoch": 2.4885171910261517,
+      "grad_norm": 0.4283460894058382,
+      "learning_rate": 3.1804826289135944e-06,
+      "loss": 0.5584,
+      "step": 9374
+    },
+    {
+      "epoch": 2.4887826894995353,
+      "grad_norm": 0.43088434320017543,
+      "learning_rate": 3.1801466765019705e-06,
+      "loss": 0.5579,
+      "step": 9375
+    },
+    {
+      "epoch": 2.4890481879729194,
+      "grad_norm": 0.4370405283436262,
+      "learning_rate": 3.1798107108258213e-06,
+      "loss": 0.5621,
+      "step": 9376
+    },
+    {
+      "epoch": 2.489313686446303,
+      "grad_norm": 0.43069731679526424,
+      "learning_rate": 3.1794747318916984e-06,
+      "loss": 0.5819,
+      "step": 9377
+    },
+    {
+      "epoch": 2.4895791849196867,
+      "grad_norm": 0.4470534406296529,
+      "learning_rate": 3.179138739706155e-06,
+      "loss": 0.5643,
+      "step": 9378
+    },
+    {
+      "epoch": 2.4898446833930703,
+      "grad_norm": 0.43352758497340826,
+      "learning_rate": 3.178802734275744e-06,
+      "loss": 0.6265,
+      "step": 9379
+    },
+    {
+      "epoch": 2.4901101818664544,
+      "grad_norm": 0.4417331957014581,
+      "learning_rate": 3.178466715607017e-06,
+      "loss": 0.5698,
+      "step": 9380
+    },
+    {
+      "epoch": 2.490375680339838,
+      "grad_norm": 0.4202582888579948,
+      "learning_rate": 3.1781306837065273e-06,
+      "loss": 0.5409,
+      "step": 9381
+    },
+    {
+      "epoch": 2.4906411788132217,
+      "grad_norm": 0.43936683233036167,
+      "learning_rate": 3.17779463858083e-06,
+      "loss": 0.5754,
+      "step": 9382
+    },
+    {
+      "epoch": 2.4909066772866058,
+      "grad_norm": 0.41849693921954273,
+      "learning_rate": 3.177458580236478e-06,
+      "loss": 0.5294,
+      "step": 9383
+    },
+    {
+      "epoch": 2.4911721757599894,
+      "grad_norm": 0.43464855887181675,
+      "learning_rate": 3.1771225086800234e-06,
+      "loss": 0.5594,
+      "step": 9384
+    },
+    {
+      "epoch": 2.491437674233373,
+      "grad_norm": 0.4292485679549083,
+      "learning_rate": 3.1767864239180234e-06,
+      "loss": 0.5773,
+      "step": 9385
+    },
+    {
+      "epoch": 2.491703172706757,
+      "grad_norm": 0.4260345270197514,
+      "learning_rate": 3.1764503259570305e-06,
+      "loss": 0.5479,
+      "step": 9386
+    },
+    {
+      "epoch": 2.4919686711801408,
+      "grad_norm": 0.43207095039306936,
+      "learning_rate": 3.1761142148035993e-06,
+      "loss": 0.5641,
+      "step": 9387
+    },
+    {
+      "epoch": 2.4922341696535244,
+      "grad_norm": 0.44440231512274675,
+      "learning_rate": 3.1757780904642864e-06,
+      "loss": 0.5622,
+      "step": 9388
+    },
+    {
+      "epoch": 2.492499668126908,
+      "grad_norm": 0.444438178675176,
+      "learning_rate": 3.175441952945646e-06,
+      "loss": 0.5544,
+      "step": 9389
+    },
+    {
+      "epoch": 2.492765166600292,
+      "grad_norm": 0.44591030742237003,
+      "learning_rate": 3.1751058022542335e-06,
+      "loss": 0.5542,
+      "step": 9390
+    },
+    {
+      "epoch": 2.4930306650736758,
+      "grad_norm": 0.4405575199835961,
+      "learning_rate": 3.174769638396606e-06,
+      "loss": 0.5532,
+      "step": 9391
+    },
+    {
+      "epoch": 2.4932961635470594,
+      "grad_norm": 0.4437465011144683,
+      "learning_rate": 3.1744334613793172e-06,
+      "loss": 0.5768,
+      "step": 9392
+    },
+    {
+      "epoch": 2.4935616620204435,
+      "grad_norm": 0.44587487694160954,
+      "learning_rate": 3.1740972712089247e-06,
+      "loss": 0.5927,
+      "step": 9393
+    },
+    {
+      "epoch": 2.493827160493827,
+      "grad_norm": 0.43148217595035476,
+      "learning_rate": 3.173761067891986e-06,
+      "loss": 0.5503,
+      "step": 9394
+    },
+    {
+      "epoch": 2.4940926589672108,
+      "grad_norm": 0.4393587490987073,
+      "learning_rate": 3.1734248514350556e-06,
+      "loss": 0.5459,
+      "step": 9395
+    },
+    {
+      "epoch": 2.494358157440595,
+      "grad_norm": 0.4347850772836811,
+      "learning_rate": 3.173088621844692e-06,
+      "loss": 0.5207,
+      "step": 9396
+    },
+    {
+      "epoch": 2.4946236559139785,
+      "grad_norm": 0.4178166944832567,
+      "learning_rate": 3.1727523791274527e-06,
+      "loss": 0.5478,
+      "step": 9397
+    },
+    {
+      "epoch": 2.494889154387362,
+      "grad_norm": 0.42730049645611723,
+      "learning_rate": 3.172416123289894e-06,
+      "loss": 0.5533,
+      "step": 9398
+    },
+    {
+      "epoch": 2.4951546528607462,
+      "grad_norm": 0.44476275157662243,
+      "learning_rate": 3.172079854338576e-06,
+      "loss": 0.5502,
+      "step": 9399
+    },
+    {
+      "epoch": 2.49542015133413,
+      "grad_norm": 0.4442515436093456,
+      "learning_rate": 3.1717435722800542e-06,
+      "loss": 0.5686,
+      "step": 9400
+    },
+    {
+      "epoch": 2.4956856498075135,
+      "grad_norm": 0.43988423286033923,
+      "learning_rate": 3.171407277120888e-06,
+      "loss": 0.5627,
+      "step": 9401
+    },
+    {
+      "epoch": 2.4959511482808976,
+      "grad_norm": 0.4437273711155727,
+      "learning_rate": 3.171070968867637e-06,
+      "loss": 0.5807,
+      "step": 9402
+    },
+    {
+      "epoch": 2.4962166467542812,
+      "grad_norm": 0.4418132440806291,
+      "learning_rate": 3.1707346475268587e-06,
+      "loss": 0.5699,
+      "step": 9403
+    },
+    {
+      "epoch": 2.496482145227665,
+      "grad_norm": 0.4244327837451305,
+      "learning_rate": 3.170398313105111e-06,
+      "loss": 0.5431,
+      "step": 9404
+    },
+    {
+      "epoch": 2.496747643701049,
+      "grad_norm": 0.43547062279590987,
+      "learning_rate": 3.1700619656089565e-06,
+      "loss": 0.5819,
+      "step": 9405
+    },
+    {
+      "epoch": 2.4970131421744326,
+      "grad_norm": 0.43619303395179776,
+      "learning_rate": 3.169725605044952e-06,
+      "loss": 0.5644,
+      "step": 9406
+    },
+    {
+      "epoch": 2.4972786406478162,
+      "grad_norm": 0.43781081050230675,
+      "learning_rate": 3.1693892314196588e-06,
+      "loss": 0.5543,
+      "step": 9407
+    },
+    {
+      "epoch": 2.4975441391212,
+      "grad_norm": 0.43690069654365465,
+      "learning_rate": 3.1690528447396375e-06,
+      "loss": 0.5752,
+      "step": 9408
+    },
+    {
+      "epoch": 2.497809637594584,
+      "grad_norm": 0.42997172892839525,
+      "learning_rate": 3.1687164450114465e-06,
+      "loss": 0.5699,
+      "step": 9409
+    },
+    {
+      "epoch": 2.4980751360679676,
+      "grad_norm": 0.4521442698318833,
+      "learning_rate": 3.168380032241648e-06,
+      "loss": 0.5338,
+      "step": 9410
+    },
+    {
+      "epoch": 2.4983406345413512,
+      "grad_norm": 0.43680720736348605,
+      "learning_rate": 3.1680436064368026e-06,
+      "loss": 0.5472,
+      "step": 9411
+    },
+    {
+      "epoch": 2.4986061330147353,
+      "grad_norm": 0.4368443393016064,
+      "learning_rate": 3.1677071676034708e-06,
+      "loss": 0.5768,
+      "step": 9412
+    },
+    {
+      "epoch": 2.498871631488119,
+      "grad_norm": 0.4418991616528072,
+      "learning_rate": 3.167370715748215e-06,
+      "loss": 0.5761,
+      "step": 9413
+    },
+    {
+      "epoch": 2.4991371299615026,
+      "grad_norm": 0.4398447504453518,
+      "learning_rate": 3.1670342508775954e-06,
+      "loss": 0.5784,
+      "step": 9414
+    },
+    {
+      "epoch": 2.4994026284348863,
+      "grad_norm": 0.4213672710563381,
+      "learning_rate": 3.1666977729981746e-06,
+      "loss": 0.5334,
+      "step": 9415
+    },
+    {
+      "epoch": 2.4996681269082703,
+      "grad_norm": 0.4322465775984401,
+      "learning_rate": 3.1663612821165156e-06,
+      "loss": 0.5459,
+      "step": 9416
+    },
+    {
+      "epoch": 2.499933625381654,
+      "grad_norm": 0.4262706490928654,
+      "learning_rate": 3.1660247782391802e-06,
+      "loss": 0.5732,
+      "step": 9417
+    },
+    {
+      "epoch": 2.5001991238550376,
+      "grad_norm": 0.4378124962515531,
+      "learning_rate": 3.16568826137273e-06,
+      "loss": 0.5788,
+      "step": 9418
+    },
+    {
+      "epoch": 2.5004646223284217,
+      "grad_norm": 0.43240347248433175,
+      "learning_rate": 3.165351731523729e-06,
+      "loss": 0.5735,
+      "step": 9419
+    },
+    {
+      "epoch": 2.5007301208018053,
+      "grad_norm": 0.43058080754786965,
+      "learning_rate": 3.1650151886987414e-06,
+      "loss": 0.5104,
+      "step": 9420
+    },
+    {
+      "epoch": 2.500995619275189,
+      "grad_norm": 0.44248267596784074,
+      "learning_rate": 3.1646786329043283e-06,
+      "loss": 0.5586,
+      "step": 9421
+    },
+    {
+      "epoch": 2.501261117748573,
+      "grad_norm": 0.4409641447977084,
+      "learning_rate": 3.1643420641470547e-06,
+      "loss": 0.5777,
+      "step": 9422
+    },
+    {
+      "epoch": 2.5015266162219567,
+      "grad_norm": 0.42398713429322904,
+      "learning_rate": 3.1640054824334842e-06,
+      "loss": 0.5365,
+      "step": 9423
+    },
+    {
+      "epoch": 2.5017921146953404,
+      "grad_norm": 0.4375260120482033,
+      "learning_rate": 3.163668887770181e-06,
+      "loss": 0.6066,
+      "step": 9424
+    },
+    {
+      "epoch": 2.5020576131687244,
+      "grad_norm": 0.42498811281789456,
+      "learning_rate": 3.16333228016371e-06,
+      "loss": 0.5733,
+      "step": 9425
+    },
+    {
+      "epoch": 2.502323111642108,
+      "grad_norm": 0.4281434155084899,
+      "learning_rate": 3.1629956596206347e-06,
+      "loss": 0.5286,
+      "step": 9426
+    },
+    {
+      "epoch": 2.5025886101154917,
+      "grad_norm": 0.4347583599855308,
+      "learning_rate": 3.162659026147521e-06,
+      "loss": 0.5728,
+      "step": 9427
+    },
+    {
+      "epoch": 2.502854108588876,
+      "grad_norm": 0.43073409845445587,
+      "learning_rate": 3.1623223797509347e-06,
+      "loss": 0.5608,
+      "step": 9428
+    },
+    {
+      "epoch": 2.5031196070622594,
+      "grad_norm": 0.4622175140503573,
+      "learning_rate": 3.161985720437439e-06,
+      "loss": 0.5839,
+      "step": 9429
+    },
+    {
+      "epoch": 2.503385105535643,
+      "grad_norm": 0.44353609136544103,
+      "learning_rate": 3.1616490482136014e-06,
+      "loss": 0.5402,
+      "step": 9430
+    },
+    {
+      "epoch": 2.503650604009027,
+      "grad_norm": 0.4214474514734924,
+      "learning_rate": 3.1613123630859876e-06,
+      "loss": 0.5502,
+      "step": 9431
+    },
+    {
+      "epoch": 2.503916102482411,
+      "grad_norm": 0.43788547739664513,
+      "learning_rate": 3.160975665061163e-06,
+      "loss": 0.5534,
+      "step": 9432
+    },
+    {
+      "epoch": 2.5041816009557945,
+      "grad_norm": 0.43323573813811256,
+      "learning_rate": 3.160638954145695e-06,
+      "loss": 0.6018,
+      "step": 9433
+    },
+    {
+      "epoch": 2.5044470994291785,
+      "grad_norm": 0.43083876935160986,
+      "learning_rate": 3.16030223034615e-06,
+      "loss": 0.5546,
+      "step": 9434
+    },
+    {
+      "epoch": 2.504712597902562,
+      "grad_norm": 0.42605749357146955,
+      "learning_rate": 3.1599654936690945e-06,
+      "loss": 0.543,
+      "step": 9435
+    },
+    {
+      "epoch": 2.504978096375946,
+      "grad_norm": 0.42765070690264945,
+      "learning_rate": 3.159628744121096e-06,
+      "loss": 0.5754,
+      "step": 9436
+    },
+    {
+      "epoch": 2.5052435948493295,
+      "grad_norm": 0.40949004682864903,
+      "learning_rate": 3.1592919817087226e-06,
+      "loss": 0.5184,
+      "step": 9437
+    },
+    {
+      "epoch": 2.5055090933227135,
+      "grad_norm": 0.4275121339098621,
+      "learning_rate": 3.1589552064385404e-06,
+      "loss": 0.5523,
+      "step": 9438
+    },
+    {
+      "epoch": 2.505774591796097,
+      "grad_norm": 0.437249294851448,
+      "learning_rate": 3.1586184183171188e-06,
+      "loss": 0.5438,
+      "step": 9439
+    },
+    {
+      "epoch": 2.506040090269481,
+      "grad_norm": 0.4401593618561723,
+      "learning_rate": 3.1582816173510257e-06,
+      "loss": 0.5576,
+      "step": 9440
+    },
+    {
+      "epoch": 2.5063055887428645,
+      "grad_norm": 0.43138843271765864,
+      "learning_rate": 3.1579448035468287e-06,
+      "loss": 0.6055,
+      "step": 9441
+    },
+    {
+      "epoch": 2.5065710872162486,
+      "grad_norm": 0.4353020564972434,
+      "learning_rate": 3.157607976911097e-06,
+      "loss": 0.5519,
+      "step": 9442
+    },
+    {
+      "epoch": 2.506836585689632,
+      "grad_norm": 0.44013769531834684,
+      "learning_rate": 3.1572711374504e-06,
+      "loss": 0.5522,
+      "step": 9443
+    },
+    {
+      "epoch": 2.507102084163016,
+      "grad_norm": 0.4323945301662739,
+      "learning_rate": 3.156934285171307e-06,
+      "loss": 0.5383,
+      "step": 9444
+    },
+    {
+      "epoch": 2.5073675826364,
+      "grad_norm": 0.4319426722078496,
+      "learning_rate": 3.1565974200803863e-06,
+      "loss": 0.5636,
+      "step": 9445
+    },
+    {
+      "epoch": 2.5076330811097836,
+      "grad_norm": 0.4263642212609335,
+      "learning_rate": 3.1562605421842085e-06,
+      "loss": 0.5248,
+      "step": 9446
+    },
+    {
+      "epoch": 2.507898579583167,
+      "grad_norm": 0.4315088290521208,
+      "learning_rate": 3.1559236514893433e-06,
+      "loss": 0.5453,
+      "step": 9447
+    },
+    {
+      "epoch": 2.5081640780565513,
+      "grad_norm": 0.44455244906088387,
+      "learning_rate": 3.1555867480023616e-06,
+      "loss": 0.5821,
+      "step": 9448
+    },
+    {
+      "epoch": 2.508429576529935,
+      "grad_norm": 0.5844649920620739,
+      "learning_rate": 3.1552498317298323e-06,
+      "loss": 0.5283,
+      "step": 9449
+    },
+    {
+      "epoch": 2.5086950750033186,
+      "grad_norm": 0.4465978020049837,
+      "learning_rate": 3.154912902678327e-06,
+      "loss": 0.5624,
+      "step": 9450
+    },
+    {
+      "epoch": 2.5089605734767026,
+      "grad_norm": 0.42463831427826604,
+      "learning_rate": 3.154575960854417e-06,
+      "loss": 0.5456,
+      "step": 9451
+    },
+    {
+      "epoch": 2.5092260719500863,
+      "grad_norm": 0.434014154650915,
+      "learning_rate": 3.154239006264672e-06,
+      "loss": 0.5618,
+      "step": 9452
+    },
+    {
+      "epoch": 2.50949157042347,
+      "grad_norm": 0.4434477667537793,
+      "learning_rate": 3.1539020389156655e-06,
+      "loss": 0.5678,
+      "step": 9453
+    },
+    {
+      "epoch": 2.509757068896854,
+      "grad_norm": 0.4271692212882504,
+      "learning_rate": 3.153565058813969e-06,
+      "loss": 0.5634,
+      "step": 9454
+    },
+    {
+      "epoch": 2.5100225673702377,
+      "grad_norm": 0.4574139663320451,
+      "learning_rate": 3.1532280659661522e-06,
+      "loss": 0.5247,
+      "step": 9455
+    },
+    {
+      "epoch": 2.5102880658436213,
+      "grad_norm": 0.4535509337897757,
+      "learning_rate": 3.1528910603787894e-06,
+      "loss": 0.5552,
+      "step": 9456
+    },
+    {
+      "epoch": 2.5105535643170054,
+      "grad_norm": 0.4466815256195863,
+      "learning_rate": 3.1525540420584525e-06,
+      "loss": 0.5486,
+      "step": 9457
+    },
+    {
+      "epoch": 2.510819062790389,
+      "grad_norm": 0.4447736236922766,
+      "learning_rate": 3.1522170110117133e-06,
+      "loss": 0.5434,
+      "step": 9458
+    },
+    {
+      "epoch": 2.5110845612637727,
+      "grad_norm": 0.4438877722799763,
+      "learning_rate": 3.1518799672451463e-06,
+      "loss": 0.5428,
+      "step": 9459
+    },
+    {
+      "epoch": 2.5113500597371567,
+      "grad_norm": 0.4449830037930682,
+      "learning_rate": 3.1515429107653233e-06,
+      "loss": 0.5532,
+      "step": 9460
+    },
+    {
+      "epoch": 2.5116155582105404,
+      "grad_norm": 0.45219101639328685,
+      "learning_rate": 3.151205841578818e-06,
+      "loss": 0.5782,
+      "step": 9461
+    },
+    {
+      "epoch": 2.511881056683924,
+      "grad_norm": 0.45266383036782876,
+      "learning_rate": 3.150868759692205e-06,
+      "loss": 0.5565,
+      "step": 9462
+    },
+    {
+      "epoch": 2.512146555157308,
+      "grad_norm": 0.44445814861076,
+      "learning_rate": 3.1505316651120584e-06,
+      "loss": 0.6087,
+      "step": 9463
+    },
+    {
+      "epoch": 2.5124120536306918,
+      "grad_norm": 0.44306589737683216,
+      "learning_rate": 3.15019455784495e-06,
+      "loss": 0.5846,
+      "step": 9464
+    },
+    {
+      "epoch": 2.5126775521040754,
+      "grad_norm": 0.4208172398550784,
+      "learning_rate": 3.1498574378974566e-06,
+      "loss": 0.5516,
+      "step": 9465
+    },
+    {
+      "epoch": 2.512943050577459,
+      "grad_norm": 0.434473317379948,
+      "learning_rate": 3.1495203052761526e-06,
+      "loss": 0.5713,
+      "step": 9466
+    },
+    {
+      "epoch": 2.513208549050843,
+      "grad_norm": 0.4388516385148562,
+      "learning_rate": 3.149183159987611e-06,
+      "loss": 0.5758,
+      "step": 9467
+    },
+    {
+      "epoch": 2.5134740475242268,
+      "grad_norm": 0.43571511140204006,
+      "learning_rate": 3.1488460020384087e-06,
+      "loss": 0.5638,
+      "step": 9468
+    },
+    {
+      "epoch": 2.5137395459976104,
+      "grad_norm": 0.43654398294349417,
+      "learning_rate": 3.1485088314351208e-06,
+      "loss": 0.5927,
+      "step": 9469
+    },
+    {
+      "epoch": 2.514005044470994,
+      "grad_norm": 0.43061235903936945,
+      "learning_rate": 3.1481716481843227e-06,
+      "loss": 0.5347,
+      "step": 9470
+    },
+    {
+      "epoch": 2.514270542944378,
+      "grad_norm": 0.4411758669881755,
+      "learning_rate": 3.1478344522925912e-06,
+      "loss": 0.5776,
+      "step": 9471
+    },
+    {
+      "epoch": 2.5145360414177618,
+      "grad_norm": 0.42823824659729354,
+      "learning_rate": 3.1474972437665015e-06,
+      "loss": 0.5126,
+      "step": 9472
+    },
+    {
+      "epoch": 2.5148015398911454,
+      "grad_norm": 0.4403791275761967,
+      "learning_rate": 3.14716002261263e-06,
+      "loss": 0.5709,
+      "step": 9473
+    },
+    {
+      "epoch": 2.5150670383645295,
+      "grad_norm": 0.44799490711890705,
+      "learning_rate": 3.1468227888375535e-06,
+      "loss": 0.5547,
+      "step": 9474
+    },
+    {
+      "epoch": 2.515332536837913,
+      "grad_norm": 0.4407152566077431,
+      "learning_rate": 3.146485542447849e-06,
+      "loss": 0.573,
+      "step": 9475
+    },
+    {
+      "epoch": 2.5155980353112968,
+      "grad_norm": 0.43329119398929544,
+      "learning_rate": 3.1461482834500936e-06,
+      "loss": 0.5587,
+      "step": 9476
+    },
+    {
+      "epoch": 2.515863533784681,
+      "grad_norm": 0.45528324970823947,
+      "learning_rate": 3.1458110118508643e-06,
+      "loss": 0.5737,
+      "step": 9477
+    },
+    {
+      "epoch": 2.5161290322580645,
+      "grad_norm": 0.4440823591559609,
+      "learning_rate": 3.145473727656739e-06,
+      "loss": 0.5683,
+      "step": 9478
+    },
+    {
+      "epoch": 2.516394530731448,
+      "grad_norm": 0.4373493703279611,
+      "learning_rate": 3.1451364308742955e-06,
+      "loss": 0.5739,
+      "step": 9479
+    },
+    {
+      "epoch": 2.5166600292048322,
+      "grad_norm": 0.4491901021562008,
+      "learning_rate": 3.1447991215101125e-06,
+      "loss": 0.5216,
+      "step": 9480
+    },
+    {
+      "epoch": 2.516925527678216,
+      "grad_norm": 0.4255890142292866,
+      "learning_rate": 3.1444617995707682e-06,
+      "loss": 0.5378,
+      "step": 9481
+    },
+    {
+      "epoch": 2.5171910261515995,
+      "grad_norm": 0.41890493629684467,
+      "learning_rate": 3.14412446506284e-06,
+      "loss": 0.5613,
+      "step": 9482
+    },
+    {
+      "epoch": 2.5174565246249836,
+      "grad_norm": 0.4313924028897693,
+      "learning_rate": 3.1437871179929086e-06,
+      "loss": 0.539,
+      "step": 9483
+    },
+    {
+      "epoch": 2.5177220230983672,
+      "grad_norm": 0.45136518996633596,
+      "learning_rate": 3.1434497583675514e-06,
+      "loss": 0.5838,
+      "step": 9484
+    },
+    {
+      "epoch": 2.517987521571751,
+      "grad_norm": 0.442987377724069,
+      "learning_rate": 3.1431123861933487e-06,
+      "loss": 0.5979,
+      "step": 9485
+    },
+    {
+      "epoch": 2.518253020045135,
+      "grad_norm": 0.45101425045741417,
+      "learning_rate": 3.1427750014768805e-06,
+      "loss": 0.5419,
+      "step": 9486
+    },
+    {
+      "epoch": 2.5185185185185186,
+      "grad_norm": 0.4351321527608984,
+      "learning_rate": 3.1424376042247253e-06,
+      "loss": 0.562,
+      "step": 9487
+    },
+    {
+      "epoch": 2.5187840169919022,
+      "grad_norm": 0.43242431047249213,
+      "learning_rate": 3.1421001944434633e-06,
+      "loss": 0.5544,
+      "step": 9488
+    },
+    {
+      "epoch": 2.5190495154652863,
+      "grad_norm": 0.4452696267555076,
+      "learning_rate": 3.141762772139676e-06,
+      "loss": 0.5489,
+      "step": 9489
+    },
+    {
+      "epoch": 2.51931501393867,
+      "grad_norm": 0.44612255944187135,
+      "learning_rate": 3.1414253373199434e-06,
+      "loss": 0.5976,
+      "step": 9490
+    },
+    {
+      "epoch": 2.5195805124120536,
+      "grad_norm": 0.47281284484367747,
+      "learning_rate": 3.141087889990846e-06,
+      "loss": 0.5886,
+      "step": 9491
+    },
+    {
+      "epoch": 2.5198460108854372,
+      "grad_norm": 0.4552362733647655,
+      "learning_rate": 3.1407504301589657e-06,
+      "loss": 0.5292,
+      "step": 9492
+    },
+    {
+      "epoch": 2.5201115093588213,
+      "grad_norm": 0.42790738361548813,
+      "learning_rate": 3.140412957830883e-06,
+      "loss": 0.5721,
+      "step": 9493
+    },
+    {
+      "epoch": 2.520377007832205,
+      "grad_norm": 0.4338820548613723,
+      "learning_rate": 3.1400754730131795e-06,
+      "loss": 0.5639,
+      "step": 9494
+    },
+    {
+      "epoch": 2.5206425063055886,
+      "grad_norm": 0.44211137226032915,
+      "learning_rate": 3.139737975712437e-06,
+      "loss": 0.5598,
+      "step": 9495
+    },
+    {
+      "epoch": 2.5209080047789723,
+      "grad_norm": 0.43897999381338676,
+      "learning_rate": 3.1394004659352373e-06,
+      "loss": 0.5646,
+      "step": 9496
+    },
+    {
+      "epoch": 2.5211735032523563,
+      "grad_norm": 0.43186772495692166,
+      "learning_rate": 3.139062943688163e-06,
+      "loss": 0.5808,
+      "step": 9497
+    },
+    {
+      "epoch": 2.52143900172574,
+      "grad_norm": 0.4225044981361485,
+      "learning_rate": 3.138725408977797e-06,
+      "loss": 0.5479,
+      "step": 9498
+    },
+    {
+      "epoch": 2.5217045001991236,
+      "grad_norm": 0.45391021063449577,
+      "learning_rate": 3.1383878618107213e-06,
+      "loss": 0.5823,
+      "step": 9499
+    },
+    {
+      "epoch": 2.5219699986725077,
+      "grad_norm": 0.4455262752039536,
+      "learning_rate": 3.13805030219352e-06,
+      "loss": 0.5808,
+      "step": 9500
+    },
+    {
+      "epoch": 2.5222354971458913,
+      "grad_norm": 0.44299883766118314,
+      "learning_rate": 3.137712730132775e-06,
+      "loss": 0.5747,
+      "step": 9501
+    },
+    {
+      "epoch": 2.522500995619275,
+      "grad_norm": 0.43684824160283026,
+      "learning_rate": 3.1373751456350703e-06,
+      "loss": 0.5588,
+      "step": 9502
+    },
+    {
+      "epoch": 2.522766494092659,
+      "grad_norm": 0.42922587695963926,
+      "learning_rate": 3.1370375487069905e-06,
+      "loss": 0.5619,
+      "step": 9503
+    },
+    {
+      "epoch": 2.5230319925660427,
+      "grad_norm": 0.43547936776945184,
+      "learning_rate": 3.136699939355118e-06,
+      "loss": 0.5809,
+      "step": 9504
+    },
+    {
+      "epoch": 2.5232974910394264,
+      "grad_norm": 0.42667776314168326,
+      "learning_rate": 3.1363623175860377e-06,
+      "loss": 0.5832,
+      "step": 9505
+    },
+    {
+      "epoch": 2.5235629895128104,
+      "grad_norm": 0.42527543983493854,
+      "learning_rate": 3.136024683406334e-06,
+      "loss": 0.533,
+      "step": 9506
+    },
+    {
+      "epoch": 2.523828487986194,
+      "grad_norm": 0.42309528273790514,
+      "learning_rate": 3.135687036822592e-06,
+      "loss": 0.5795,
+      "step": 9507
+    },
+    {
+      "epoch": 2.5240939864595777,
+      "grad_norm": 0.4497635986445369,
+      "learning_rate": 3.135349377841396e-06,
+      "loss": 0.6098,
+      "step": 9508
+    },
+    {
+      "epoch": 2.524359484932962,
+      "grad_norm": 0.42315734533823407,
+      "learning_rate": 3.1350117064693324e-06,
+      "loss": 0.5744,
+      "step": 9509
+    },
+    {
+      "epoch": 2.5246249834063454,
+      "grad_norm": 0.4355568249485254,
+      "learning_rate": 3.134674022712985e-06,
+      "loss": 0.5661,
+      "step": 9510
+    },
+    {
+      "epoch": 2.524890481879729,
+      "grad_norm": 0.43121043889104366,
+      "learning_rate": 3.1343363265789407e-06,
+      "loss": 0.564,
+      "step": 9511
+    },
+    {
+      "epoch": 2.525155980353113,
+      "grad_norm": 0.43134446245835184,
+      "learning_rate": 3.133998618073785e-06,
+      "loss": 0.5948,
+      "step": 9512
+    },
+    {
+      "epoch": 2.525421478826497,
+      "grad_norm": 0.4423339513765274,
+      "learning_rate": 3.1336608972041037e-06,
+      "loss": 0.6002,
+      "step": 9513
+    },
+    {
+      "epoch": 2.5256869772998805,
+      "grad_norm": 0.42365528100900635,
+      "learning_rate": 3.133323163976483e-06,
+      "loss": 0.5897,
+      "step": 9514
+    },
+    {
+      "epoch": 2.5259524757732645,
+      "grad_norm": 0.4389206580562018,
+      "learning_rate": 3.1329854183975106e-06,
+      "loss": 0.5714,
+      "step": 9515
+    },
+    {
+      "epoch": 2.526217974246648,
+      "grad_norm": 0.448376762853967,
+      "learning_rate": 3.132647660473772e-06,
+      "loss": 0.565,
+      "step": 9516
+    },
+    {
+      "epoch": 2.526483472720032,
+      "grad_norm": 0.4275117684605819,
+      "learning_rate": 3.1323098902118554e-06,
+      "loss": 0.5199,
+      "step": 9517
+    },
+    {
+      "epoch": 2.526748971193416,
+      "grad_norm": 0.43336623722199247,
+      "learning_rate": 3.1319721076183484e-06,
+      "loss": 0.5889,
+      "step": 9518
+    },
+    {
+      "epoch": 2.5270144696667995,
+      "grad_norm": 0.4322185586098048,
+      "learning_rate": 3.131634312699837e-06,
+      "loss": 0.5763,
+      "step": 9519
+    },
+    {
+      "epoch": 2.527279968140183,
+      "grad_norm": 0.42550780270464755,
+      "learning_rate": 3.1312965054629106e-06,
+      "loss": 0.5696,
+      "step": 9520
+    },
+    {
+      "epoch": 2.527545466613567,
+      "grad_norm": 0.43623261231775873,
+      "learning_rate": 3.1309586859141566e-06,
+      "loss": 0.5637,
+      "step": 9521
+    },
+    {
+      "epoch": 2.527810965086951,
+      "grad_norm": 0.4332095673225028,
+      "learning_rate": 3.1306208540601637e-06,
+      "loss": 0.5423,
+      "step": 9522
+    },
+    {
+      "epoch": 2.5280764635603346,
+      "grad_norm": 0.4203605509327906,
+      "learning_rate": 3.13028300990752e-06,
+      "loss": 0.5639,
+      "step": 9523
+    },
+    {
+      "epoch": 2.528341962033718,
+      "grad_norm": 0.43504759431508144,
+      "learning_rate": 3.1299451534628134e-06,
+      "loss": 0.5969,
+      "step": 9524
+    },
+    {
+      "epoch": 2.528607460507102,
+      "grad_norm": 0.42970801059919694,
+      "learning_rate": 3.129607284732634e-06,
+      "loss": 0.5752,
+      "step": 9525
+    },
+    {
+      "epoch": 2.528872958980486,
+      "grad_norm": 0.4287573954391148,
+      "learning_rate": 3.1292694037235723e-06,
+      "loss": 0.5392,
+      "step": 9526
+    },
+    {
+      "epoch": 2.5291384574538696,
+      "grad_norm": 0.41973472472414747,
+      "learning_rate": 3.128931510442216e-06,
+      "loss": 0.5369,
+      "step": 9527
+    },
+    {
+      "epoch": 2.529403955927253,
+      "grad_norm": 0.4361919066579121,
+      "learning_rate": 3.1285936048951547e-06,
+      "loss": 0.5419,
+      "step": 9528
+    },
+    {
+      "epoch": 2.5296694544006373,
+      "grad_norm": 0.4443904929303206,
+      "learning_rate": 3.12825568708898e-06,
+      "loss": 0.5876,
+      "step": 9529
+    },
+    {
+      "epoch": 2.529934952874021,
+      "grad_norm": 0.4402880808060107,
+      "learning_rate": 3.1279177570302802e-06,
+      "loss": 0.5771,
+      "step": 9530
+    },
+    {
+      "epoch": 2.5302004513474046,
+      "grad_norm": 0.45374940769583544,
+      "learning_rate": 3.1275798147256474e-06,
+      "loss": 0.5294,
+      "step": 9531
+    },
+    {
+      "epoch": 2.5304659498207887,
+      "grad_norm": 0.4363678735918915,
+      "learning_rate": 3.127241860181672e-06,
+      "loss": 0.5189,
+      "step": 9532
+    },
+    {
+      "epoch": 2.5307314482941723,
+      "grad_norm": 0.4344168100653965,
+      "learning_rate": 3.126903893404944e-06,
+      "loss": 0.5584,
+      "step": 9533
+    },
+    {
+      "epoch": 2.530996946767556,
+      "grad_norm": 0.4305850554702492,
+      "learning_rate": 3.1265659144020544e-06,
+      "loss": 0.5296,
+      "step": 9534
+    },
+    {
+      "epoch": 2.53126244524094,
+      "grad_norm": 0.44083573100602924,
+      "learning_rate": 3.126227923179597e-06,
+      "loss": 0.5621,
+      "step": 9535
+    },
+    {
+      "epoch": 2.5315279437143237,
+      "grad_norm": 0.4269563173026622,
+      "learning_rate": 3.1258899197441608e-06,
+      "loss": 0.5384,
+      "step": 9536
+    },
+    {
+      "epoch": 2.5317934421877073,
+      "grad_norm": 0.4441549610656123,
+      "learning_rate": 3.1255519041023387e-06,
+      "loss": 0.5822,
+      "step": 9537
+    },
+    {
+      "epoch": 2.5320589406610914,
+      "grad_norm": 0.4297305506469278,
+      "learning_rate": 3.125213876260723e-06,
+      "loss": 0.5348,
+      "step": 9538
+    },
+    {
+      "epoch": 2.532324439134475,
+      "grad_norm": 0.43566509214109717,
+      "learning_rate": 3.124875836225906e-06,
+      "loss": 0.5192,
+      "step": 9539
+    },
+    {
+      "epoch": 2.5325899376078587,
+      "grad_norm": 0.4416059876334432,
+      "learning_rate": 3.1245377840044805e-06,
+      "loss": 0.6153,
+      "step": 9540
+    },
+    {
+      "epoch": 2.5328554360812428,
+      "grad_norm": 0.42428828956074627,
+      "learning_rate": 3.1241997196030388e-06,
+      "loss": 0.5423,
+      "step": 9541
+    },
+    {
+      "epoch": 2.5331209345546264,
+      "grad_norm": 0.42596054185212223,
+      "learning_rate": 3.123861643028174e-06,
+      "loss": 0.5344,
+      "step": 9542
+    },
+    {
+      "epoch": 2.53338643302801,
+      "grad_norm": 0.43986906323238645,
+      "learning_rate": 3.12352355428648e-06,
+      "loss": 0.5739,
+      "step": 9543
+    },
+    {
+      "epoch": 2.533651931501394,
+      "grad_norm": 0.4317776710873684,
+      "learning_rate": 3.12318545338455e-06,
+      "loss": 0.568,
+      "step": 9544
+    },
+    {
+      "epoch": 2.5339174299747778,
+      "grad_norm": 0.42654226536784584,
+      "learning_rate": 3.1228473403289776e-06,
+      "loss": 0.5798,
+      "step": 9545
+    },
+    {
+      "epoch": 2.5341829284481614,
+      "grad_norm": 0.44675013196536045,
+      "learning_rate": 3.122509215126358e-06,
+      "loss": 0.5517,
+      "step": 9546
+    },
+    {
+      "epoch": 2.534448426921545,
+      "grad_norm": 0.42996163224144596,
+      "learning_rate": 3.1221710777832835e-06,
+      "loss": 0.5558,
+      "step": 9547
+    },
+    {
+      "epoch": 2.534713925394929,
+      "grad_norm": 0.4368409079749489,
+      "learning_rate": 3.1218329283063497e-06,
+      "loss": 0.5744,
+      "step": 9548
+    },
+    {
+      "epoch": 2.5349794238683128,
+      "grad_norm": 0.44339747108247923,
+      "learning_rate": 3.121494766702152e-06,
+      "loss": 0.5724,
+      "step": 9549
+    },
+    {
+      "epoch": 2.5352449223416964,
+      "grad_norm": 0.44277342219882204,
+      "learning_rate": 3.1211565929772845e-06,
+      "loss": 0.5916,
+      "step": 9550
+    },
+    {
+      "epoch": 2.53551042081508,
+      "grad_norm": 0.44487316385250864,
+      "learning_rate": 3.120818407138342e-06,
+      "loss": 0.582,
+      "step": 9551
+    },
+    {
+      "epoch": 2.535775919288464,
+      "grad_norm": 0.4362030697603754,
+      "learning_rate": 3.1204802091919204e-06,
+      "loss": 0.5725,
+      "step": 9552
+    },
+    {
+      "epoch": 2.5360414177618478,
+      "grad_norm": 0.435433119817375,
+      "learning_rate": 3.1201419991446165e-06,
+      "loss": 0.6164,
+      "step": 9553
+    },
+    {
+      "epoch": 2.5363069162352314,
+      "grad_norm": 0.446177343611205,
+      "learning_rate": 3.119803777003024e-06,
+      "loss": 0.563,
+      "step": 9554
+    },
+    {
+      "epoch": 2.5365724147086155,
+      "grad_norm": 0.4345535738139898,
+      "learning_rate": 3.119465542773742e-06,
+      "loss": 0.5578,
+      "step": 9555
+    },
+    {
+      "epoch": 2.536837913181999,
+      "grad_norm": 0.45093478053169606,
+      "learning_rate": 3.119127296463364e-06,
+      "loss": 0.5964,
+      "step": 9556
+    },
+    {
+      "epoch": 2.537103411655383,
+      "grad_norm": 0.4364144173436119,
+      "learning_rate": 3.118789038078488e-06,
+      "loss": 0.5248,
+      "step": 9557
+    },
+    {
+      "epoch": 2.537368910128767,
+      "grad_norm": 0.43196640147038123,
+      "learning_rate": 3.1184507676257114e-06,
+      "loss": 0.5519,
+      "step": 9558
+    },
+    {
+      "epoch": 2.5376344086021505,
+      "grad_norm": 0.43453490475121515,
+      "learning_rate": 3.1181124851116298e-06,
+      "loss": 0.5761,
+      "step": 9559
+    },
+    {
+      "epoch": 2.537899907075534,
+      "grad_norm": 0.43271268563202675,
+      "learning_rate": 3.1177741905428416e-06,
+      "loss": 0.5858,
+      "step": 9560
+    },
+    {
+      "epoch": 2.5381654055489182,
+      "grad_norm": 0.4351962821397591,
+      "learning_rate": 3.1174358839259432e-06,
+      "loss": 0.6051,
+      "step": 9561
+    },
+    {
+      "epoch": 2.538430904022302,
+      "grad_norm": 0.44725236624506076,
+      "learning_rate": 3.1170975652675347e-06,
+      "loss": 0.5602,
+      "step": 9562
+    },
+    {
+      "epoch": 2.5386964024956855,
+      "grad_norm": 0.44153102422449925,
+      "learning_rate": 3.116759234574212e-06,
+      "loss": 0.5664,
+      "step": 9563
+    },
+    {
+      "epoch": 2.5389619009690696,
+      "grad_norm": 0.4236565414140181,
+      "learning_rate": 3.1164208918525747e-06,
+      "loss": 0.569,
+      "step": 9564
+    },
+    {
+      "epoch": 2.5392273994424532,
+      "grad_norm": 0.4300754715942089,
+      "learning_rate": 3.11608253710922e-06,
+      "loss": 0.5369,
+      "step": 9565
+    },
+    {
+      "epoch": 2.539492897915837,
+      "grad_norm": 0.42609634091504456,
+      "learning_rate": 3.115744170350748e-06,
+      "loss": 0.5315,
+      "step": 9566
+    },
+    {
+      "epoch": 2.539758396389221,
+      "grad_norm": 0.42587900942237505,
+      "learning_rate": 3.115405791583757e-06,
+      "loss": 0.5801,
+      "step": 9567
+    },
+    {
+      "epoch": 2.5400238948626046,
+      "grad_norm": 0.431035010412275,
+      "learning_rate": 3.115067400814846e-06,
+      "loss": 0.5613,
+      "step": 9568
+    },
+    {
+      "epoch": 2.5402893933359882,
+      "grad_norm": 0.4339576942082634,
+      "learning_rate": 3.1147289980506153e-06,
+      "loss": 0.5588,
+      "step": 9569
+    },
+    {
+      "epoch": 2.5405548918093723,
+      "grad_norm": 0.4273639579508005,
+      "learning_rate": 3.1143905832976635e-06,
+      "loss": 0.5534,
+      "step": 9570
+    },
+    {
+      "epoch": 2.540820390282756,
+      "grad_norm": 0.44161662566454657,
+      "learning_rate": 3.114052156562591e-06,
+      "loss": 0.5517,
+      "step": 9571
+    },
+    {
+      "epoch": 2.5410858887561396,
+      "grad_norm": 0.4458514529256643,
+      "learning_rate": 3.1137137178519983e-06,
+      "loss": 0.5786,
+      "step": 9572
+    },
+    {
+      "epoch": 2.5413513872295237,
+      "grad_norm": 0.44340320434447833,
+      "learning_rate": 3.1133752671724854e-06,
+      "loss": 0.5465,
+      "step": 9573
+    },
+    {
+      "epoch": 2.5416168857029073,
+      "grad_norm": 0.4412991934226862,
+      "learning_rate": 3.1130368045306524e-06,
+      "loss": 0.584,
+      "step": 9574
+    },
+    {
+      "epoch": 2.541882384176291,
+      "grad_norm": 0.44850265188253113,
+      "learning_rate": 3.112698329933102e-06,
+      "loss": 0.5491,
+      "step": 9575
+    },
+    {
+      "epoch": 2.5421478826496746,
+      "grad_norm": 0.44833539809204115,
+      "learning_rate": 3.112359843386433e-06,
+      "loss": 0.5808,
+      "step": 9576
+    },
+    {
+      "epoch": 2.5424133811230587,
+      "grad_norm": 0.4544369119280941,
+      "learning_rate": 3.112021344897248e-06,
+      "loss": 0.5609,
+      "step": 9577
+    },
+    {
+      "epoch": 2.5426788795964423,
+      "grad_norm": 0.4313114021309775,
+      "learning_rate": 3.111682834472149e-06,
+      "loss": 0.5812,
+      "step": 9578
+    },
+    {
+      "epoch": 2.542944378069826,
+      "grad_norm": 0.4309077614383536,
+      "learning_rate": 3.1113443121177363e-06,
+      "loss": 0.5776,
+      "step": 9579
+    },
+    {
+      "epoch": 2.5432098765432096,
+      "grad_norm": 0.44483404090843465,
+      "learning_rate": 3.1110057778406116e-06,
+      "loss": 0.5782,
+      "step": 9580
+    },
+    {
+      "epoch": 2.5434753750165937,
+      "grad_norm": 0.4479458413040876,
+      "learning_rate": 3.11066723164738e-06,
+      "loss": 0.5972,
+      "step": 9581
+    },
+    {
+      "epoch": 2.5437408734899773,
+      "grad_norm": 0.42731519509149357,
+      "learning_rate": 3.1103286735446416e-06,
+      "loss": 0.556,
+      "step": 9582
+    },
+    {
+      "epoch": 2.544006371963361,
+      "grad_norm": 0.44485304815168575,
+      "learning_rate": 3.1099901035389997e-06,
+      "loss": 0.5329,
+      "step": 9583
+    },
+    {
+      "epoch": 2.544271870436745,
+      "grad_norm": 0.42894232338073646,
+      "learning_rate": 3.1096515216370576e-06,
+      "loss": 0.5779,
+      "step": 9584
+    },
+    {
+      "epoch": 2.5445373689101287,
+      "grad_norm": 0.4268356634060042,
+      "learning_rate": 3.1093129278454177e-06,
+      "loss": 0.5301,
+      "step": 9585
+    },
+    {
+      "epoch": 2.5448028673835124,
+      "grad_norm": 0.4511171018705595,
+      "learning_rate": 3.1089743221706835e-06,
+      "loss": 0.6065,
+      "step": 9586
+    },
+    {
+      "epoch": 2.5450683658568964,
+      "grad_norm": 0.4273845279605641,
+      "learning_rate": 3.1086357046194603e-06,
+      "loss": 0.5392,
+      "step": 9587
+    },
+    {
+      "epoch": 2.54533386433028,
+      "grad_norm": 0.43914860888158624,
+      "learning_rate": 3.1082970751983497e-06,
+      "loss": 0.5146,
+      "step": 9588
+    },
+    {
+      "epoch": 2.5455993628036637,
+      "grad_norm": 0.4324563904612212,
+      "learning_rate": 3.1079584339139566e-06,
+      "loss": 0.6039,
+      "step": 9589
+    },
+    {
+      "epoch": 2.545864861277048,
+      "grad_norm": 0.4271580225793852,
+      "learning_rate": 3.1076197807728857e-06,
+      "loss": 0.5515,
+      "step": 9590
+    },
+    {
+      "epoch": 2.5461303597504314,
+      "grad_norm": 0.46396719742564996,
+      "learning_rate": 3.107281115781741e-06,
+      "loss": 0.5709,
+      "step": 9591
+    },
+    {
+      "epoch": 2.546395858223815,
+      "grad_norm": 0.4349570813326812,
+      "learning_rate": 3.1069424389471292e-06,
+      "loss": 0.5636,
+      "step": 9592
+    },
+    {
+      "epoch": 2.546661356697199,
+      "grad_norm": 0.46137287062106475,
+      "learning_rate": 3.1066037502756523e-06,
+      "loss": 0.5609,
+      "step": 9593
+    },
+    {
+      "epoch": 2.546926855170583,
+      "grad_norm": 0.4500883025148973,
+      "learning_rate": 3.106265049773917e-06,
+      "loss": 0.5508,
+      "step": 9594
+    },
+    {
+      "epoch": 2.5471923536439665,
+      "grad_norm": 0.44443925103479776,
+      "learning_rate": 3.1059263374485305e-06,
+      "loss": 0.5659,
+      "step": 9595
+    },
+    {
+      "epoch": 2.5474578521173505,
+      "grad_norm": 0.4315078983705683,
+      "learning_rate": 3.105587613306095e-06,
+      "loss": 0.5667,
+      "step": 9596
+    },
+    {
+      "epoch": 2.547723350590734,
+      "grad_norm": 0.4846562068702339,
+      "learning_rate": 3.105248877353219e-06,
+      "loss": 0.5505,
+      "step": 9597
+    },
+    {
+      "epoch": 2.547988849064118,
+      "grad_norm": 0.4399541495163904,
+      "learning_rate": 3.104910129596508e-06,
+      "loss": 0.5369,
+      "step": 9598
+    },
+    {
+      "epoch": 2.548254347537502,
+      "grad_norm": 0.4236312804238317,
+      "learning_rate": 3.1045713700425685e-06,
+      "loss": 0.5206,
+      "step": 9599
+    },
+    {
+      "epoch": 2.5485198460108855,
+      "grad_norm": 0.4415132115063261,
+      "learning_rate": 3.1042325986980066e-06,
+      "loss": 0.5789,
+      "step": 9600
+    },
+    {
+      "epoch": 2.548785344484269,
+      "grad_norm": 0.43559110899813647,
+      "learning_rate": 3.1038938155694303e-06,
+      "loss": 0.5809,
+      "step": 9601
+    },
+    {
+      "epoch": 2.549050842957653,
+      "grad_norm": 0.44944716217112785,
+      "learning_rate": 3.103555020663445e-06,
+      "loss": 0.5458,
+      "step": 9602
+    },
+    {
+      "epoch": 2.549316341431037,
+      "grad_norm": 0.4462666502776378,
+      "learning_rate": 3.10321621398666e-06,
+      "loss": 0.5799,
+      "step": 9603
+    },
+    {
+      "epoch": 2.5495818399044206,
+      "grad_norm": 0.44363867021151626,
+      "learning_rate": 3.102877395545682e-06,
+      "loss": 0.5851,
+      "step": 9604
+    },
+    {
+      "epoch": 2.549847338377804,
+      "grad_norm": 0.42597884771753386,
+      "learning_rate": 3.102538565347118e-06,
+      "loss": 0.5575,
+      "step": 9605
+    },
+    {
+      "epoch": 2.550112836851188,
+      "grad_norm": 0.4290679298444189,
+      "learning_rate": 3.1021997233975766e-06,
+      "loss": 0.5285,
+      "step": 9606
+    },
+    {
+      "epoch": 2.550378335324572,
+      "grad_norm": 0.46242184259367003,
+      "learning_rate": 3.101860869703667e-06,
+      "loss": 0.5356,
+      "step": 9607
+    },
+    {
+      "epoch": 2.5506438337979556,
+      "grad_norm": 0.4294958834898439,
+      "learning_rate": 3.1015220042719955e-06,
+      "loss": 0.5771,
+      "step": 9608
+    },
+    {
+      "epoch": 2.550909332271339,
+      "grad_norm": 0.4406036185051546,
+      "learning_rate": 3.101183127109173e-06,
+      "loss": 0.5364,
+      "step": 9609
+    },
+    {
+      "epoch": 2.5511748307447233,
+      "grad_norm": 0.42664364356263557,
+      "learning_rate": 3.100844238221808e-06,
+      "loss": 0.5461,
+      "step": 9610
+    },
+    {
+      "epoch": 2.551440329218107,
+      "grad_norm": 0.4336074557591012,
+      "learning_rate": 3.100505337616509e-06,
+      "loss": 0.5513,
+      "step": 9611
+    },
+    {
+      "epoch": 2.5517058276914906,
+      "grad_norm": 0.42951183128196596,
+      "learning_rate": 3.1001664252998853e-06,
+      "loss": 0.5493,
+      "step": 9612
+    },
+    {
+      "epoch": 2.5519713261648747,
+      "grad_norm": 0.4418028266518351,
+      "learning_rate": 3.0998275012785473e-06,
+      "loss": 0.5733,
+      "step": 9613
+    },
+    {
+      "epoch": 2.5522368246382583,
+      "grad_norm": 0.4330427838676648,
+      "learning_rate": 3.099488565559104e-06,
+      "loss": 0.5734,
+      "step": 9614
+    },
+    {
+      "epoch": 2.552502323111642,
+      "grad_norm": 0.4443548976547577,
+      "learning_rate": 3.0991496181481654e-06,
+      "loss": 0.5868,
+      "step": 9615
+    },
+    {
+      "epoch": 2.552767821585026,
+      "grad_norm": 0.4411619204202222,
+      "learning_rate": 3.0988106590523425e-06,
+      "loss": 0.5515,
+      "step": 9616
+    },
+    {
+      "epoch": 2.5530333200584097,
+      "grad_norm": 0.4400094390422781,
+      "learning_rate": 3.0984716882782464e-06,
+      "loss": 0.5529,
+      "step": 9617
+    },
+    {
+      "epoch": 2.5532988185317933,
+      "grad_norm": 0.4287085676505781,
+      "learning_rate": 3.098132705832487e-06,
+      "loss": 0.578,
+      "step": 9618
+    },
+    {
+      "epoch": 2.5535643170051774,
+      "grad_norm": 0.43531895284167754,
+      "learning_rate": 3.0977937117216746e-06,
+      "loss": 0.5688,
+      "step": 9619
+    },
+    {
+      "epoch": 2.553829815478561,
+      "grad_norm": 0.4433710813700474,
+      "learning_rate": 3.097454705952421e-06,
+      "loss": 0.5429,
+      "step": 9620
+    },
+    {
+      "epoch": 2.5540953139519447,
+      "grad_norm": 0.44479554804842375,
+      "learning_rate": 3.0971156885313392e-06,
+      "loss": 0.5728,
+      "step": 9621
+    },
+    {
+      "epoch": 2.5543608124253288,
+      "grad_norm": 0.4310681895763898,
+      "learning_rate": 3.0967766594650383e-06,
+      "loss": 0.5217,
+      "step": 9622
+    },
+    {
+      "epoch": 2.5546263108987124,
+      "grad_norm": 0.43433357132671396,
+      "learning_rate": 3.096437618760131e-06,
+      "loss": 0.5698,
+      "step": 9623
+    },
+    {
+      "epoch": 2.554891809372096,
+      "grad_norm": 0.4258182024620701,
+      "learning_rate": 3.0960985664232314e-06,
+      "loss": 0.5654,
+      "step": 9624
+    },
+    {
+      "epoch": 2.55515730784548,
+      "grad_norm": 0.44314283305276586,
+      "learning_rate": 3.095759502460949e-06,
+      "loss": 0.6027,
+      "step": 9625
+    },
+    {
+      "epoch": 2.5554228063188638,
+      "grad_norm": 0.4481693364534932,
+      "learning_rate": 3.0954204268798972e-06,
+      "loss": 0.5402,
+      "step": 9626
+    },
+    {
+      "epoch": 2.5556883047922474,
+      "grad_norm": 0.44993258145260806,
+      "learning_rate": 3.095081339686691e-06,
+      "loss": 0.565,
+      "step": 9627
+    },
+    {
+      "epoch": 2.5559538032656315,
+      "grad_norm": 0.4482393847770999,
+      "learning_rate": 3.09474224088794e-06,
+      "loss": 0.6091,
+      "step": 9628
+    },
+    {
+      "epoch": 2.556219301739015,
+      "grad_norm": 0.4367650615935612,
+      "learning_rate": 3.09440313049026e-06,
+      "loss": 0.5642,
+      "step": 9629
+    },
+    {
+      "epoch": 2.5564848002123988,
+      "grad_norm": 0.44771433011254264,
+      "learning_rate": 3.0940640085002636e-06,
+      "loss": 0.5565,
+      "step": 9630
+    },
+    {
+      "epoch": 2.5567502986857824,
+      "grad_norm": 0.4376664303792177,
+      "learning_rate": 3.093724874924564e-06,
+      "loss": 0.5437,
+      "step": 9631
+    },
+    {
+      "epoch": 2.5570157971591665,
+      "grad_norm": 0.42659774084869156,
+      "learning_rate": 3.093385729769776e-06,
+      "loss": 0.5374,
+      "step": 9632
+    },
+    {
+      "epoch": 2.55728129563255,
+      "grad_norm": 0.4300593549908244,
+      "learning_rate": 3.0930465730425136e-06,
+      "loss": 0.558,
+      "step": 9633
+    },
+    {
+      "epoch": 2.5575467941059338,
+      "grad_norm": 0.4330741643839839,
+      "learning_rate": 3.092707404749391e-06,
+      "loss": 0.5566,
+      "step": 9634
+    },
+    {
+      "epoch": 2.5578122925793174,
+      "grad_norm": 0.42953398249368685,
+      "learning_rate": 3.092368224897022e-06,
+      "loss": 0.5827,
+      "step": 9635
+    },
+    {
+      "epoch": 2.5580777910527015,
+      "grad_norm": 0.419536117584414,
+      "learning_rate": 3.0920290334920228e-06,
+      "loss": 0.5379,
+      "step": 9636
+    },
+    {
+      "epoch": 2.558343289526085,
+      "grad_norm": 0.44578708584775045,
+      "learning_rate": 3.091689830541008e-06,
+      "loss": 0.5544,
+      "step": 9637
+    },
+    {
+      "epoch": 2.558608787999469,
+      "grad_norm": 0.4276303993102809,
+      "learning_rate": 3.091350616050592e-06,
+      "loss": 0.5899,
+      "step": 9638
+    },
+    {
+      "epoch": 2.558874286472853,
+      "grad_norm": 0.43882022603036214,
+      "learning_rate": 3.0910113900273924e-06,
+      "loss": 0.5519,
+      "step": 9639
+    },
+    {
+      "epoch": 2.5591397849462365,
+      "grad_norm": 0.4291659665658252,
+      "learning_rate": 3.090672152478023e-06,
+      "loss": 0.5553,
+      "step": 9640
+    },
+    {
+      "epoch": 2.55940528341962,
+      "grad_norm": 0.44232041095698077,
+      "learning_rate": 3.090332903409101e-06,
+      "loss": 0.5798,
+      "step": 9641
+    },
+    {
+      "epoch": 2.5596707818930042,
+      "grad_norm": 0.44130519804366836,
+      "learning_rate": 3.0899936428272413e-06,
+      "loss": 0.5657,
+      "step": 9642
+    },
+    {
+      "epoch": 2.559936280366388,
+      "grad_norm": 0.451146993664638,
+      "learning_rate": 3.0896543707390613e-06,
+      "loss": 0.6109,
+      "step": 9643
+    },
+    {
+      "epoch": 2.5602017788397715,
+      "grad_norm": 0.44537391289833644,
+      "learning_rate": 3.089315087151177e-06,
+      "loss": 0.5875,
+      "step": 9644
+    },
+    {
+      "epoch": 2.5604672773131556,
+      "grad_norm": 0.43163412393199746,
+      "learning_rate": 3.0889757920702057e-06,
+      "loss": 0.5899,
+      "step": 9645
+    },
+    {
+      "epoch": 2.5607327757865392,
+      "grad_norm": 0.45092081440197557,
+      "learning_rate": 3.0886364855027645e-06,
+      "loss": 0.5402,
+      "step": 9646
+    },
+    {
+      "epoch": 2.560998274259923,
+      "grad_norm": 0.4283447167542925,
+      "learning_rate": 3.0882971674554714e-06,
+      "loss": 0.5752,
+      "step": 9647
+    },
+    {
+      "epoch": 2.561263772733307,
+      "grad_norm": 0.4282776561390347,
+      "learning_rate": 3.0879578379349416e-06,
+      "loss": 0.4969,
+      "step": 9648
+    },
+    {
+      "epoch": 2.5615292712066906,
+      "grad_norm": 0.42995979364080233,
+      "learning_rate": 3.0876184969477956e-06,
+      "loss": 0.5207,
+      "step": 9649
+    },
+    {
+      "epoch": 2.5617947696800742,
+      "grad_norm": 0.4437998775448447,
+      "learning_rate": 3.08727914450065e-06,
+      "loss": 0.5301,
+      "step": 9650
+    },
+    {
+      "epoch": 2.5620602681534583,
+      "grad_norm": 0.45048063157224244,
+      "learning_rate": 3.0869397806001228e-06,
+      "loss": 0.5408,
+      "step": 9651
+    },
+    {
+      "epoch": 2.562325766626842,
+      "grad_norm": 0.4439482520063843,
+      "learning_rate": 3.0866004052528333e-06,
+      "loss": 0.5216,
+      "step": 9652
+    },
+    {
+      "epoch": 2.5625912651002256,
+      "grad_norm": 0.4518546624959561,
+      "learning_rate": 3.0862610184653987e-06,
+      "loss": 0.5493,
+      "step": 9653
+    },
+    {
+      "epoch": 2.5628567635736097,
+      "grad_norm": 0.4403500420155033,
+      "learning_rate": 3.0859216202444393e-06,
+      "loss": 0.5355,
+      "step": 9654
+    },
+    {
+      "epoch": 2.5631222620469933,
+      "grad_norm": 0.42894716317848847,
+      "learning_rate": 3.085582210596574e-06,
+      "loss": 0.567,
+      "step": 9655
+    },
+    {
+      "epoch": 2.563387760520377,
+      "grad_norm": 0.43339364611784054,
+      "learning_rate": 3.085242789528422e-06,
+      "loss": 0.5355,
+      "step": 9656
+    },
+    {
+      "epoch": 2.5636532589937606,
+      "grad_norm": 0.43156591436058017,
+      "learning_rate": 3.0849033570466017e-06,
+      "loss": 0.5758,
+      "step": 9657
+    },
+    {
+      "epoch": 2.5639187574671447,
+      "grad_norm": 0.4543316643247976,
+      "learning_rate": 3.084563913157735e-06,
+      "loss": 0.6042,
+      "step": 9658
+    },
+    {
+      "epoch": 2.5641842559405283,
+      "grad_norm": 0.44382083957879875,
+      "learning_rate": 3.0842244578684406e-06,
+      "loss": 0.6009,
+      "step": 9659
+    },
+    {
+      "epoch": 2.564449754413912,
+      "grad_norm": 0.46253761886490485,
+      "learning_rate": 3.083884991185338e-06,
+      "loss": 0.6028,
+      "step": 9660
+    },
+    {
+      "epoch": 2.5647152528872956,
+      "grad_norm": 0.43500935516070693,
+      "learning_rate": 3.0835455131150487e-06,
+      "loss": 0.5711,
+      "step": 9661
+    },
+    {
+      "epoch": 2.5649807513606797,
+      "grad_norm": 0.4594670650746794,
+      "learning_rate": 3.0832060236641936e-06,
+      "loss": 0.573,
+      "step": 9662
+    },
+    {
+      "epoch": 2.5652462498340634,
+      "grad_norm": 0.47308726852966054,
+      "learning_rate": 3.0828665228393928e-06,
+      "loss": 0.5986,
+      "step": 9663
+    },
+    {
+      "epoch": 2.565511748307447,
+      "grad_norm": 0.443394762718553,
+      "learning_rate": 3.0825270106472683e-06,
+      "loss": 0.5673,
+      "step": 9664
+    },
+    {
+      "epoch": 2.565777246780831,
+      "grad_norm": 0.4332601791833904,
+      "learning_rate": 3.0821874870944403e-06,
+      "loss": 0.5723,
+      "step": 9665
+    },
+    {
+      "epoch": 2.5660427452542147,
+      "grad_norm": 0.4371094025214304,
+      "learning_rate": 3.0818479521875307e-06,
+      "loss": 0.5766,
+      "step": 9666
+    },
+    {
+      "epoch": 2.5663082437275984,
+      "grad_norm": 0.4567428164942645,
+      "learning_rate": 3.0815084059331617e-06,
+      "loss": 0.5722,
+      "step": 9667
+    },
+    {
+      "epoch": 2.5665737422009824,
+      "grad_norm": 0.4408415113896548,
+      "learning_rate": 3.0811688483379546e-06,
+      "loss": 0.5824,
+      "step": 9668
+    },
+    {
+      "epoch": 2.566839240674366,
+      "grad_norm": 0.4325895208979403,
+      "learning_rate": 3.0808292794085327e-06,
+      "loss": 0.5682,
+      "step": 9669
+    },
+    {
+      "epoch": 2.5671047391477497,
+      "grad_norm": 0.42824672270415926,
+      "learning_rate": 3.0804896991515177e-06,
+      "loss": 0.5609,
+      "step": 9670
+    },
+    {
+      "epoch": 2.567370237621134,
+      "grad_norm": 0.4385187801553421,
+      "learning_rate": 3.080150107573531e-06,
+      "loss": 0.5439,
+      "step": 9671
+    },
+    {
+      "epoch": 2.5676357360945175,
+      "grad_norm": 0.43241807837773755,
+      "learning_rate": 3.0798105046811976e-06,
+      "loss": 0.5753,
+      "step": 9672
+    },
+    {
+      "epoch": 2.567901234567901,
+      "grad_norm": 0.45119715017116857,
+      "learning_rate": 3.07947089048114e-06,
+      "loss": 0.5653,
+      "step": 9673
+    },
+    {
+      "epoch": 2.568166733041285,
+      "grad_norm": 0.44053305754524424,
+      "learning_rate": 3.079131264979981e-06,
+      "loss": 0.5715,
+      "step": 9674
+    },
+    {
+      "epoch": 2.568432231514669,
+      "grad_norm": 0.43547949527770335,
+      "learning_rate": 3.078791628184344e-06,
+      "loss": 0.5702,
+      "step": 9675
+    },
+    {
+      "epoch": 2.5686977299880525,
+      "grad_norm": 0.4559303012637625,
+      "learning_rate": 3.0784519801008546e-06,
+      "loss": 0.5677,
+      "step": 9676
+    },
+    {
+      "epoch": 2.5689632284614365,
+      "grad_norm": 0.44548355383752986,
+      "learning_rate": 3.078112320736134e-06,
+      "loss": 0.5979,
+      "step": 9677
+    },
+    {
+      "epoch": 2.56922872693482,
+      "grad_norm": 0.43308078305629577,
+      "learning_rate": 3.0777726500968075e-06,
+      "loss": 0.5991,
+      "step": 9678
+    },
+    {
+      "epoch": 2.569494225408204,
+      "grad_norm": 0.4369511103644772,
+      "learning_rate": 3.077432968189501e-06,
+      "loss": 0.579,
+      "step": 9679
+    },
+    {
+      "epoch": 2.569759723881588,
+      "grad_norm": 0.45597989379802933,
+      "learning_rate": 3.0770932750208366e-06,
+      "loss": 0.5812,
+      "step": 9680
+    },
+    {
+      "epoch": 2.5700252223549716,
+      "grad_norm": 0.41739150056029206,
+      "learning_rate": 3.07675357059744e-06,
+      "loss": 0.5326,
+      "step": 9681
+    },
+    {
+      "epoch": 2.570290720828355,
+      "grad_norm": 0.43937655926575614,
+      "learning_rate": 3.0764138549259374e-06,
+      "loss": 0.5512,
+      "step": 9682
+    },
+    {
+      "epoch": 2.5705562193017393,
+      "grad_norm": 0.4503578024386901,
+      "learning_rate": 3.0760741280129535e-06,
+      "loss": 0.5601,
+      "step": 9683
+    },
+    {
+      "epoch": 2.570821717775123,
+      "grad_norm": 0.43065582273140096,
+      "learning_rate": 3.075734389865113e-06,
+      "loss": 0.5866,
+      "step": 9684
+    },
+    {
+      "epoch": 2.5710872162485066,
+      "grad_norm": 0.43642136230903655,
+      "learning_rate": 3.0753946404890432e-06,
+      "loss": 0.6166,
+      "step": 9685
+    },
+    {
+      "epoch": 2.57135271472189,
+      "grad_norm": 0.4387966750656058,
+      "learning_rate": 3.0750548798913687e-06,
+      "loss": 0.5479,
+      "step": 9686
+    },
+    {
+      "epoch": 2.5716182131952743,
+      "grad_norm": 0.4426165277122058,
+      "learning_rate": 3.0747151080787158e-06,
+      "loss": 0.5327,
+      "step": 9687
+    },
+    {
+      "epoch": 2.571883711668658,
+      "grad_norm": 0.4308996680887213,
+      "learning_rate": 3.0743753250577112e-06,
+      "loss": 0.5462,
+      "step": 9688
+    },
+    {
+      "epoch": 2.5721492101420416,
+      "grad_norm": 0.4329813988437253,
+      "learning_rate": 3.074035530834982e-06,
+      "loss": 0.5112,
+      "step": 9689
+    },
+    {
+      "epoch": 2.572414708615425,
+      "grad_norm": 0.42706876476087124,
+      "learning_rate": 3.0736957254171535e-06,
+      "loss": 0.5396,
+      "step": 9690
+    },
+    {
+      "epoch": 2.5726802070888093,
+      "grad_norm": 0.44080151752434044,
+      "learning_rate": 3.0733559088108543e-06,
+      "loss": 0.5276,
+      "step": 9691
+    },
+    {
+      "epoch": 2.572945705562193,
+      "grad_norm": 0.4294161573220226,
+      "learning_rate": 3.0730160810227105e-06,
+      "loss": 0.5576,
+      "step": 9692
+    },
+    {
+      "epoch": 2.5732112040355766,
+      "grad_norm": 0.4376772241817617,
+      "learning_rate": 3.0726762420593514e-06,
+      "loss": 0.5465,
+      "step": 9693
+    },
+    {
+      "epoch": 2.5734767025089607,
+      "grad_norm": 0.4398951307319967,
+      "learning_rate": 3.0723363919274026e-06,
+      "loss": 0.5903,
+      "step": 9694
+    },
+    {
+      "epoch": 2.5737422009823443,
+      "grad_norm": 0.435528661039307,
+      "learning_rate": 3.0719965306334924e-06,
+      "loss": 0.5722,
+      "step": 9695
+    },
+    {
+      "epoch": 2.574007699455728,
+      "grad_norm": 0.4390306049659235,
+      "learning_rate": 3.0716566581842507e-06,
+      "loss": 0.5426,
+      "step": 9696
+    },
+    {
+      "epoch": 2.574273197929112,
+      "grad_norm": 0.43547529734439866,
+      "learning_rate": 3.0713167745863033e-06,
+      "loss": 0.5172,
+      "step": 9697
+    },
+    {
+      "epoch": 2.5745386964024957,
+      "grad_norm": 0.43690288710044933,
+      "learning_rate": 3.0709768798462804e-06,
+      "loss": 0.5917,
+      "step": 9698
+    },
+    {
+      "epoch": 2.5748041948758793,
+      "grad_norm": 0.4297154448698437,
+      "learning_rate": 3.07063697397081e-06,
+      "loss": 0.5535,
+      "step": 9699
+    },
+    {
+      "epoch": 2.5750696933492634,
+      "grad_norm": 0.4229441284386931,
+      "learning_rate": 3.0702970569665213e-06,
+      "loss": 0.543,
+      "step": 9700
+    },
+    {
+      "epoch": 2.575335191822647,
+      "grad_norm": 0.44940251564345524,
+      "learning_rate": 3.069957128840044e-06,
+      "loss": 0.5527,
+      "step": 9701
+    },
+    {
+      "epoch": 2.5756006902960307,
+      "grad_norm": 0.45742016478394915,
+      "learning_rate": 3.0696171895980076e-06,
+      "loss": 0.5737,
+      "step": 9702
+    },
+    {
+      "epoch": 2.5758661887694148,
+      "grad_norm": 0.445479816965706,
+      "learning_rate": 3.069277239247041e-06,
+      "loss": 0.5392,
+      "step": 9703
+    },
+    {
+      "epoch": 2.5761316872427984,
+      "grad_norm": 0.440622712134827,
+      "learning_rate": 3.0689372777937745e-06,
+      "loss": 0.5898,
+      "step": 9704
+    },
+    {
+      "epoch": 2.576397185716182,
+      "grad_norm": 0.43273044340065625,
+      "learning_rate": 3.0685973052448383e-06,
+      "loss": 0.5489,
+      "step": 9705
+    },
+    {
+      "epoch": 2.576662684189566,
+      "grad_norm": 0.44018325564738436,
+      "learning_rate": 3.068257321606862e-06,
+      "loss": 0.6019,
+      "step": 9706
+    },
+    {
+      "epoch": 2.5769281826629498,
+      "grad_norm": 0.43263978548441684,
+      "learning_rate": 3.0679173268864766e-06,
+      "loss": 0.5657,
+      "step": 9707
+    },
+    {
+      "epoch": 2.5771936811363334,
+      "grad_norm": 0.44134116260728956,
+      "learning_rate": 3.0675773210903126e-06,
+      "loss": 0.5426,
+      "step": 9708
+    },
+    {
+      "epoch": 2.5774591796097175,
+      "grad_norm": 0.44403198567368973,
+      "learning_rate": 3.0672373042250014e-06,
+      "loss": 0.5841,
+      "step": 9709
+    },
+    {
+      "epoch": 2.577724678083101,
+      "grad_norm": 0.426394921965446,
+      "learning_rate": 3.066897276297174e-06,
+      "loss": 0.5746,
+      "step": 9710
+    },
+    {
+      "epoch": 2.5779901765564848,
+      "grad_norm": 0.4371630480201046,
+      "learning_rate": 3.066557237313462e-06,
+      "loss": 0.568,
+      "step": 9711
+    },
+    {
+      "epoch": 2.578255675029869,
+      "grad_norm": 0.4413669250540348,
+      "learning_rate": 3.066217187280497e-06,
+      "loss": 0.5748,
+      "step": 9712
+    },
+    {
+      "epoch": 2.5785211735032525,
+      "grad_norm": 0.440982396462472,
+      "learning_rate": 3.0658771262049104e-06,
+      "loss": 0.5373,
+      "step": 9713
+    },
+    {
+      "epoch": 2.578786671976636,
+      "grad_norm": 0.442487642415627,
+      "learning_rate": 3.0655370540933334e-06,
+      "loss": 0.6338,
+      "step": 9714
+    },
+    {
+      "epoch": 2.5790521704500198,
+      "grad_norm": 0.44948528578801356,
+      "learning_rate": 3.0651969709523997e-06,
+      "loss": 0.5582,
+      "step": 9715
+    },
+    {
+      "epoch": 2.5793176689234034,
+      "grad_norm": 0.43941225770394476,
+      "learning_rate": 3.064856876788741e-06,
+      "loss": 0.5928,
+      "step": 9716
+    },
+    {
+      "epoch": 2.5795831673967875,
+      "grad_norm": 0.43369359329392015,
+      "learning_rate": 3.06451677160899e-06,
+      "loss": 0.5789,
+      "step": 9717
+    },
+    {
+      "epoch": 2.579848665870171,
+      "grad_norm": 0.4442134752303508,
+      "learning_rate": 3.0641766554197806e-06,
+      "loss": 0.5864,
+      "step": 9718
+    },
+    {
+      "epoch": 2.580114164343555,
+      "grad_norm": 0.43899568244259546,
+      "learning_rate": 3.063836528227745e-06,
+      "loss": 0.5991,
+      "step": 9719
+    },
+    {
+      "epoch": 2.580379662816939,
+      "grad_norm": 0.4230559718199393,
+      "learning_rate": 3.063496390039516e-06,
+      "loss": 0.5607,
+      "step": 9720
+    },
+    {
+      "epoch": 2.5806451612903225,
+      "grad_norm": 0.4252969025278546,
+      "learning_rate": 3.0631562408617275e-06,
+      "loss": 0.591,
+      "step": 9721
+    },
+    {
+      "epoch": 2.580910659763706,
+      "grad_norm": 0.4560028926384697,
+      "learning_rate": 3.062816080701015e-06,
+      "loss": 0.5999,
+      "step": 9722
+    },
+    {
+      "epoch": 2.5811761582370902,
+      "grad_norm": 0.43833207953848735,
+      "learning_rate": 3.0624759095640092e-06,
+      "loss": 0.5603,
+      "step": 9723
+    },
+    {
+      "epoch": 2.581441656710474,
+      "grad_norm": 0.4363813445482862,
+      "learning_rate": 3.062135727457347e-06,
+      "loss": 0.5724,
+      "step": 9724
+    },
+    {
+      "epoch": 2.5817071551838575,
+      "grad_norm": 0.4326799520006999,
+      "learning_rate": 3.0617955343876617e-06,
+      "loss": 0.5709,
+      "step": 9725
+    },
+    {
+      "epoch": 2.5819726536572416,
+      "grad_norm": 0.4270160638351923,
+      "learning_rate": 3.0614553303615876e-06,
+      "loss": 0.5714,
+      "step": 9726
+    },
+    {
+      "epoch": 2.5822381521306252,
+      "grad_norm": 0.4422786158529615,
+      "learning_rate": 3.0611151153857592e-06,
+      "loss": 0.5579,
+      "step": 9727
+    },
+    {
+      "epoch": 2.582503650604009,
+      "grad_norm": 0.45552364147327573,
+      "learning_rate": 3.0607748894668133e-06,
+      "loss": 0.5972,
+      "step": 9728
+    },
+    {
+      "epoch": 2.582769149077393,
+      "grad_norm": 0.4452913847497938,
+      "learning_rate": 3.0604346526113836e-06,
+      "loss": 0.5267,
+      "step": 9729
+    },
+    {
+      "epoch": 2.5830346475507766,
+      "grad_norm": 0.43300755692225423,
+      "learning_rate": 3.060094404826106e-06,
+      "loss": 0.5571,
+      "step": 9730
+    },
+    {
+      "epoch": 2.5833001460241602,
+      "grad_norm": 0.4310298096754106,
+      "learning_rate": 3.0597541461176167e-06,
+      "loss": 0.5457,
+      "step": 9731
+    },
+    {
+      "epoch": 2.5835656444975443,
+      "grad_norm": 0.4445485083975476,
+      "learning_rate": 3.0594138764925507e-06,
+      "loss": 0.5489,
+      "step": 9732
+    },
+    {
+      "epoch": 2.583831142970928,
+      "grad_norm": 0.44137021353093164,
+      "learning_rate": 3.059073595957544e-06,
+      "loss": 0.5753,
+      "step": 9733
+    },
+    {
+      "epoch": 2.5840966414443116,
+      "grad_norm": 0.43509391341790743,
+      "learning_rate": 3.0587333045192343e-06,
+      "loss": 0.6055,
+      "step": 9734
+    },
+    {
+      "epoch": 2.5843621399176957,
+      "grad_norm": 0.4208525086294821,
+      "learning_rate": 3.058393002184256e-06,
+      "loss": 0.5489,
+      "step": 9735
+    },
+    {
+      "epoch": 2.5846276383910793,
+      "grad_norm": 0.4381786107640345,
+      "learning_rate": 3.0580526889592478e-06,
+      "loss": 0.5714,
+      "step": 9736
+    },
+    {
+      "epoch": 2.584893136864463,
+      "grad_norm": 0.44078837796859505,
+      "learning_rate": 3.057712364850845e-06,
+      "loss": 0.5843,
+      "step": 9737
+    },
+    {
+      "epoch": 2.585158635337847,
+      "grad_norm": 0.4356203587026822,
+      "learning_rate": 3.057372029865686e-06,
+      "loss": 0.5687,
+      "step": 9738
+    },
+    {
+      "epoch": 2.5854241338112307,
+      "grad_norm": 0.4287284696172287,
+      "learning_rate": 3.057031684010408e-06,
+      "loss": 0.5641,
+      "step": 9739
+    },
+    {
+      "epoch": 2.5856896322846143,
+      "grad_norm": 0.4347329717106706,
+      "learning_rate": 3.056691327291648e-06,
+      "loss": 0.5646,
+      "step": 9740
+    },
+    {
+      "epoch": 2.585955130757998,
+      "grad_norm": 0.4484382559259174,
+      "learning_rate": 3.056350959716044e-06,
+      "loss": 0.5585,
+      "step": 9741
+    },
+    {
+      "epoch": 2.586220629231382,
+      "grad_norm": 0.44551854589227574,
+      "learning_rate": 3.056010581290235e-06,
+      "loss": 0.5428,
+      "step": 9742
+    },
+    {
+      "epoch": 2.5864861277047657,
+      "grad_norm": 0.4485386527591479,
+      "learning_rate": 3.0556701920208574e-06,
+      "loss": 0.5826,
+      "step": 9743
+    },
+    {
+      "epoch": 2.5867516261781494,
+      "grad_norm": 0.43917854755312696,
+      "learning_rate": 3.0553297919145508e-06,
+      "loss": 0.5751,
+      "step": 9744
+    },
+    {
+      "epoch": 2.587017124651533,
+      "grad_norm": 0.4532500033425838,
+      "learning_rate": 3.0549893809779533e-06,
+      "loss": 0.5448,
+      "step": 9745
+    },
+    {
+      "epoch": 2.587282623124917,
+      "grad_norm": 0.4331426040609163,
+      "learning_rate": 3.054648959217704e-06,
+      "loss": 0.5561,
+      "step": 9746
+    },
+    {
+      "epoch": 2.5875481215983007,
+      "grad_norm": 0.42564462222038085,
+      "learning_rate": 3.054308526640442e-06,
+      "loss": 0.523,
+      "step": 9747
+    },
+    {
+      "epoch": 2.5878136200716844,
+      "grad_norm": 0.4373161538748567,
+      "learning_rate": 3.0539680832528074e-06,
+      "loss": 0.5852,
+      "step": 9748
+    },
+    {
+      "epoch": 2.5880791185450684,
+      "grad_norm": 0.43965705139518685,
+      "learning_rate": 3.053627629061438e-06,
+      "loss": 0.5554,
+      "step": 9749
+    },
+    {
+      "epoch": 2.588344617018452,
+      "grad_norm": 0.44356494700891186,
+      "learning_rate": 3.0532871640729745e-06,
+      "loss": 0.5399,
+      "step": 9750
+    },
+    {
+      "epoch": 2.5886101154918357,
+      "grad_norm": 0.44333843467667555,
+      "learning_rate": 3.052946688294057e-06,
+      "loss": 0.569,
+      "step": 9751
+    },
+    {
+      "epoch": 2.58887561396522,
+      "grad_norm": 0.4454987314272198,
+      "learning_rate": 3.052606201731325e-06,
+      "loss": 0.5295,
+      "step": 9752
+    },
+    {
+      "epoch": 2.5891411124386035,
+      "grad_norm": 0.43299005361834114,
+      "learning_rate": 3.052265704391419e-06,
+      "loss": 0.5116,
+      "step": 9753
+    },
+    {
+      "epoch": 2.589406610911987,
+      "grad_norm": 0.44992390350494116,
+      "learning_rate": 3.0519251962809793e-06,
+      "loss": 0.5933,
+      "step": 9754
+    },
+    {
+      "epoch": 2.589672109385371,
+      "grad_norm": 0.4475046198646384,
+      "learning_rate": 3.051584677406647e-06,
+      "loss": 0.5824,
+      "step": 9755
+    },
+    {
+      "epoch": 2.589937607858755,
+      "grad_norm": 0.45401987100159574,
+      "learning_rate": 3.0512441477750634e-06,
+      "loss": 0.563,
+      "step": 9756
+    },
+    {
+      "epoch": 2.5902031063321385,
+      "grad_norm": 0.44654322221378523,
+      "learning_rate": 3.0509036073928686e-06,
+      "loss": 0.5541,
+      "step": 9757
+    },
+    {
+      "epoch": 2.5904686048055225,
+      "grad_norm": 0.4563393408195813,
+      "learning_rate": 3.0505630562667056e-06,
+      "loss": 0.5402,
+      "step": 9758
+    },
+    {
+      "epoch": 2.590734103278906,
+      "grad_norm": 0.44925229949481094,
+      "learning_rate": 3.0502224944032136e-06,
+      "loss": 0.5672,
+      "step": 9759
+    },
+    {
+      "epoch": 2.59099960175229,
+      "grad_norm": 0.43030383881738954,
+      "learning_rate": 3.049881921809037e-06,
+      "loss": 0.5601,
+      "step": 9760
+    },
+    {
+      "epoch": 2.591265100225674,
+      "grad_norm": 0.44225812487070487,
+      "learning_rate": 3.0495413384908162e-06,
+      "loss": 0.5931,
+      "step": 9761
+    },
+    {
+      "epoch": 2.5915305986990576,
+      "grad_norm": 0.41567714851754306,
+      "learning_rate": 3.0492007444551935e-06,
+      "loss": 0.5593,
+      "step": 9762
+    },
+    {
+      "epoch": 2.591796097172441,
+      "grad_norm": 0.42934805119035097,
+      "learning_rate": 3.048860139708812e-06,
+      "loss": 0.585,
+      "step": 9763
+    },
+    {
+      "epoch": 2.5920615956458253,
+      "grad_norm": 0.44443860220951853,
+      "learning_rate": 3.048519524258314e-06,
+      "loss": 0.5669,
+      "step": 9764
+    },
+    {
+      "epoch": 2.592327094119209,
+      "grad_norm": 0.4370753854806174,
+      "learning_rate": 3.0481788981103423e-06,
+      "loss": 0.5085,
+      "step": 9765
+    },
+    {
+      "epoch": 2.5925925925925926,
+      "grad_norm": 0.4858166900980427,
+      "learning_rate": 3.0478382612715398e-06,
+      "loss": 0.5702,
+      "step": 9766
+    },
+    {
+      "epoch": 2.5928580910659766,
+      "grad_norm": 0.43261493530951145,
+      "learning_rate": 3.047497613748549e-06,
+      "loss": 0.5465,
+      "step": 9767
+    },
+    {
+      "epoch": 2.5931235895393603,
+      "grad_norm": 0.42241666580234893,
+      "learning_rate": 3.047156955548015e-06,
+      "loss": 0.5495,
+      "step": 9768
+    },
+    {
+      "epoch": 2.593389088012744,
+      "grad_norm": 0.4306741881923886,
+      "learning_rate": 3.046816286676581e-06,
+      "loss": 0.5557,
+      "step": 9769
+    },
+    {
+      "epoch": 2.5936545864861276,
+      "grad_norm": 0.4313434566745016,
+      "learning_rate": 3.04647560714089e-06,
+      "loss": 0.5683,
+      "step": 9770
+    },
+    {
+      "epoch": 2.593920084959511,
+      "grad_norm": 0.4396761462753513,
+      "learning_rate": 3.046134916947587e-06,
+      "loss": 0.5858,
+      "step": 9771
+    },
+    {
+      "epoch": 2.5941855834328953,
+      "grad_norm": 0.42664954869505867,
+      "learning_rate": 3.0457942161033155e-06,
+      "loss": 0.5811,
+      "step": 9772
+    },
+    {
+      "epoch": 2.594451081906279,
+      "grad_norm": 0.44056989475907066,
+      "learning_rate": 3.0454535046147204e-06,
+      "loss": 0.5482,
+      "step": 9773
+    },
+    {
+      "epoch": 2.5947165803796626,
+      "grad_norm": 0.4340837263558321,
+      "learning_rate": 3.0451127824884473e-06,
+      "loss": 0.5506,
+      "step": 9774
+    },
+    {
+      "epoch": 2.5949820788530467,
+      "grad_norm": 0.4373848539519573,
+      "learning_rate": 3.04477204973114e-06,
+      "loss": 0.5762,
+      "step": 9775
+    },
+    {
+      "epoch": 2.5952475773264303,
+      "grad_norm": 0.44704426977221406,
+      "learning_rate": 3.0444313063494433e-06,
+      "loss": 0.5532,
+      "step": 9776
+    },
+    {
+      "epoch": 2.595513075799814,
+      "grad_norm": 0.43485027470875753,
+      "learning_rate": 3.044090552350004e-06,
+      "loss": 0.5371,
+      "step": 9777
+    },
+    {
+      "epoch": 2.595778574273198,
+      "grad_norm": 0.4376157385527146,
+      "learning_rate": 3.043749787739466e-06,
+      "loss": 0.5828,
+      "step": 9778
+    },
+    {
+      "epoch": 2.5960440727465817,
+      "grad_norm": 0.4499818976277327,
+      "learning_rate": 3.0434090125244765e-06,
+      "loss": 0.5527,
+      "step": 9779
+    },
+    {
+      "epoch": 2.5963095712199653,
+      "grad_norm": 0.4281069380411903,
+      "learning_rate": 3.0430682267116808e-06,
+      "loss": 0.5657,
+      "step": 9780
+    },
+    {
+      "epoch": 2.5965750696933494,
+      "grad_norm": 0.42514294574719663,
+      "learning_rate": 3.0427274303077244e-06,
+      "loss": 0.5621,
+      "step": 9781
+    },
+    {
+      "epoch": 2.596840568166733,
+      "grad_norm": 0.4447212077327684,
+      "learning_rate": 3.0423866233192545e-06,
+      "loss": 0.5631,
+      "step": 9782
+    },
+    {
+      "epoch": 2.5971060666401167,
+      "grad_norm": 0.4460502603504364,
+      "learning_rate": 3.0420458057529173e-06,
+      "loss": 0.5577,
+      "step": 9783
+    },
+    {
+      "epoch": 2.5973715651135008,
+      "grad_norm": 0.4428713739205099,
+      "learning_rate": 3.0417049776153595e-06,
+      "loss": 0.5864,
+      "step": 9784
+    },
+    {
+      "epoch": 2.5976370635868844,
+      "grad_norm": 0.4451558243609404,
+      "learning_rate": 3.0413641389132294e-06,
+      "loss": 0.5553,
+      "step": 9785
+    },
+    {
+      "epoch": 2.597902562060268,
+      "grad_norm": 0.43309212325519936,
+      "learning_rate": 3.0410232896531724e-06,
+      "loss": 0.5461,
+      "step": 9786
+    },
+    {
+      "epoch": 2.598168060533652,
+      "grad_norm": 0.44156694273279407,
+      "learning_rate": 3.0406824298418356e-06,
+      "loss": 0.5932,
+      "step": 9787
+    },
+    {
+      "epoch": 2.5984335590070358,
+      "grad_norm": 0.4360336364953369,
+      "learning_rate": 3.040341559485869e-06,
+      "loss": 0.5198,
+      "step": 9788
+    },
+    {
+      "epoch": 2.5986990574804194,
+      "grad_norm": 0.4279731277995889,
+      "learning_rate": 3.040000678591919e-06,
+      "loss": 0.5787,
+      "step": 9789
+    },
+    {
+      "epoch": 2.5989645559538035,
+      "grad_norm": 0.4561929037477712,
+      "learning_rate": 3.039659787166633e-06,
+      "loss": 0.5824,
+      "step": 9790
+    },
+    {
+      "epoch": 2.599230054427187,
+      "grad_norm": 0.4506964886677526,
+      "learning_rate": 3.03931888521666e-06,
+      "loss": 0.581,
+      "step": 9791
+    },
+    {
+      "epoch": 2.5994955529005708,
+      "grad_norm": 0.4435214542838103,
+      "learning_rate": 3.0389779727486477e-06,
+      "loss": 0.5998,
+      "step": 9792
+    },
+    {
+      "epoch": 2.599761051373955,
+      "grad_norm": 0.45674711032145454,
+      "learning_rate": 3.038637049769246e-06,
+      "loss": 0.5544,
+      "step": 9793
+    },
+    {
+      "epoch": 2.6000265498473385,
+      "grad_norm": 0.4393354078245991,
+      "learning_rate": 3.0382961162851033e-06,
+      "loss": 0.548,
+      "step": 9794
+    },
+    {
+      "epoch": 2.600292048320722,
+      "grad_norm": 0.44694471978942,
+      "learning_rate": 3.0379551723028685e-06,
+      "loss": 0.549,
+      "step": 9795
+    },
+    {
+      "epoch": 2.600557546794106,
+      "grad_norm": 0.44017410469128426,
+      "learning_rate": 3.0376142178291895e-06,
+      "loss": 0.5687,
+      "step": 9796
+    },
+    {
+      "epoch": 2.60082304526749,
+      "grad_norm": 0.42482843852284996,
+      "learning_rate": 3.037273252870718e-06,
+      "loss": 0.5865,
+      "step": 9797
+    },
+    {
+      "epoch": 2.6010885437408735,
+      "grad_norm": 0.42901528053754867,
+      "learning_rate": 3.036932277434102e-06,
+      "loss": 0.5269,
+      "step": 9798
+    },
+    {
+      "epoch": 2.601354042214257,
+      "grad_norm": 0.44591400173005213,
+      "learning_rate": 3.036591291525992e-06,
+      "loss": 0.5873,
+      "step": 9799
+    },
+    {
+      "epoch": 2.601619540687641,
+      "grad_norm": 0.4385822464012712,
+      "learning_rate": 3.036250295153038e-06,
+      "loss": 0.5375,
+      "step": 9800
+    },
+    {
+      "epoch": 2.601885039161025,
+      "grad_norm": 0.4402534968600166,
+      "learning_rate": 3.0359092883218906e-06,
+      "loss": 0.5771,
+      "step": 9801
+    },
+    {
+      "epoch": 2.6021505376344085,
+      "grad_norm": 0.43833730244716207,
+      "learning_rate": 3.0355682710392e-06,
+      "loss": 0.585,
+      "step": 9802
+    },
+    {
+      "epoch": 2.602416036107792,
+      "grad_norm": 0.43479937801590046,
+      "learning_rate": 3.0352272433116167e-06,
+      "loss": 0.5517,
+      "step": 9803
+    },
+    {
+      "epoch": 2.6026815345811762,
+      "grad_norm": 0.41718724280708447,
+      "learning_rate": 3.0348862051457915e-06,
+      "loss": 0.545,
+      "step": 9804
+    },
+    {
+      "epoch": 2.60294703305456,
+      "grad_norm": 0.4371491257352857,
+      "learning_rate": 3.0345451565483757e-06,
+      "loss": 0.5796,
+      "step": 9805
+    },
+    {
+      "epoch": 2.6032125315279435,
+      "grad_norm": 0.44529406113816433,
+      "learning_rate": 3.034204097526021e-06,
+      "loss": 0.5476,
+      "step": 9806
+    },
+    {
+      "epoch": 2.6034780300013276,
+      "grad_norm": 0.4456608461963751,
+      "learning_rate": 3.033863028085378e-06,
+      "loss": 0.5608,
+      "step": 9807
+    },
+    {
+      "epoch": 2.6037435284747112,
+      "grad_norm": 0.4442525025244097,
+      "learning_rate": 3.0335219482330986e-06,
+      "loss": 0.5009,
+      "step": 9808
+    },
+    {
+      "epoch": 2.604009026948095,
+      "grad_norm": 0.43778616921276803,
+      "learning_rate": 3.0331808579758353e-06,
+      "loss": 0.5675,
+      "step": 9809
+    },
+    {
+      "epoch": 2.604274525421479,
+      "grad_norm": 0.43374274097977733,
+      "learning_rate": 3.0328397573202396e-06,
+      "loss": 0.5877,
+      "step": 9810
+    },
+    {
+      "epoch": 2.6045400238948626,
+      "grad_norm": 0.4412803474152078,
+      "learning_rate": 3.0324986462729646e-06,
+      "loss": 0.5445,
+      "step": 9811
+    },
+    {
+      "epoch": 2.6048055223682463,
+      "grad_norm": 0.4473201748434038,
+      "learning_rate": 3.0321575248406616e-06,
+      "loss": 0.5457,
+      "step": 9812
+    },
+    {
+      "epoch": 2.6050710208416303,
+      "grad_norm": 0.4383248526898061,
+      "learning_rate": 3.0318163930299837e-06,
+      "loss": 0.5869,
+      "step": 9813
+    },
+    {
+      "epoch": 2.605336519315014,
+      "grad_norm": 0.45294481003544385,
+      "learning_rate": 3.031475250847585e-06,
+      "loss": 0.5766,
+      "step": 9814
+    },
+    {
+      "epoch": 2.6056020177883976,
+      "grad_norm": 0.4382048548341572,
+      "learning_rate": 3.031134098300117e-06,
+      "loss": 0.5716,
+      "step": 9815
+    },
+    {
+      "epoch": 2.6058675162617817,
+      "grad_norm": 0.44171566858782185,
+      "learning_rate": 3.0307929353942334e-06,
+      "loss": 0.5787,
+      "step": 9816
+    },
+    {
+      "epoch": 2.6061330147351653,
+      "grad_norm": 0.4505579968909563,
+      "learning_rate": 3.0304517621365885e-06,
+      "loss": 0.598,
+      "step": 9817
+    },
+    {
+      "epoch": 2.606398513208549,
+      "grad_norm": 0.44601775398021676,
+      "learning_rate": 3.0301105785338345e-06,
+      "loss": 0.5794,
+      "step": 9818
+    },
+    {
+      "epoch": 2.606664011681933,
+      "grad_norm": 0.43829884704682603,
+      "learning_rate": 3.0297693845926275e-06,
+      "loss": 0.5793,
+      "step": 9819
+    },
+    {
+      "epoch": 2.6069295101553167,
+      "grad_norm": 0.42658114366263344,
+      "learning_rate": 3.02942818031962e-06,
+      "loss": 0.5501,
+      "step": 9820
+    },
+    {
+      "epoch": 2.6071950086287004,
+      "grad_norm": 0.43301045301465824,
+      "learning_rate": 3.0290869657214666e-06,
+      "loss": 0.5797,
+      "step": 9821
+    },
+    {
+      "epoch": 2.6074605071020844,
+      "grad_norm": 0.4402593412152837,
+      "learning_rate": 3.0287457408048217e-06,
+      "loss": 0.5794,
+      "step": 9822
+    },
+    {
+      "epoch": 2.607726005575468,
+      "grad_norm": 0.42525719379358584,
+      "learning_rate": 3.028404505576341e-06,
+      "loss": 0.5524,
+      "step": 9823
+    },
+    {
+      "epoch": 2.6079915040488517,
+      "grad_norm": 0.42923808583665674,
+      "learning_rate": 3.0280632600426777e-06,
+      "loss": 0.5587,
+      "step": 9824
+    },
+    {
+      "epoch": 2.6082570025222354,
+      "grad_norm": 0.42587532680181983,
+      "learning_rate": 3.027722004210489e-06,
+      "loss": 0.5202,
+      "step": 9825
+    },
+    {
+      "epoch": 2.608522500995619,
+      "grad_norm": 0.4363099031173716,
+      "learning_rate": 3.0273807380864285e-06,
+      "loss": 0.5541,
+      "step": 9826
+    },
+    {
+      "epoch": 2.608787999469003,
+      "grad_norm": 0.4372428248122534,
+      "learning_rate": 3.0270394616771527e-06,
+      "loss": 0.5341,
+      "step": 9827
+    },
+    {
+      "epoch": 2.6090534979423867,
+      "grad_norm": 0.42776404571600163,
+      "learning_rate": 3.026698174989316e-06,
+      "loss": 0.5203,
+      "step": 9828
+    },
+    {
+      "epoch": 2.6093189964157704,
+      "grad_norm": 0.44432985489600485,
+      "learning_rate": 3.026356878029576e-06,
+      "loss": 0.6067,
+      "step": 9829
+    },
+    {
+      "epoch": 2.6095844948891544,
+      "grad_norm": 0.4390615134331211,
+      "learning_rate": 3.026015570804588e-06,
+      "loss": 0.5918,
+      "step": 9830
+    },
+    {
+      "epoch": 2.609849993362538,
+      "grad_norm": 0.43020925799764337,
+      "learning_rate": 3.0256742533210087e-06,
+      "loss": 0.5758,
+      "step": 9831
+    },
+    {
+      "epoch": 2.6101154918359217,
+      "grad_norm": 0.43965596014813296,
+      "learning_rate": 3.0253329255854946e-06,
+      "loss": 0.5493,
+      "step": 9832
+    },
+    {
+      "epoch": 2.610380990309306,
+      "grad_norm": 0.44824104122635183,
+      "learning_rate": 3.024991587604702e-06,
+      "loss": 0.568,
+      "step": 9833
+    },
+    {
+      "epoch": 2.6106464887826895,
+      "grad_norm": 0.4342521722757152,
+      "learning_rate": 3.0246502393852882e-06,
+      "loss": 0.5497,
+      "step": 9834
+    },
+    {
+      "epoch": 2.610911987256073,
+      "grad_norm": 0.42379865502264136,
+      "learning_rate": 3.0243088809339093e-06,
+      "loss": 0.5263,
+      "step": 9835
+    },
+    {
+      "epoch": 2.611177485729457,
+      "grad_norm": 0.4282984012286767,
+      "learning_rate": 3.0239675122572245e-06,
+      "loss": 0.5882,
+      "step": 9836
+    },
+    {
+      "epoch": 2.611442984202841,
+      "grad_norm": 0.44381237227313886,
+      "learning_rate": 3.0236261333618897e-06,
+      "loss": 0.603,
+      "step": 9837
+    },
+    {
+      "epoch": 2.6117084826762245,
+      "grad_norm": 0.4298498609582854,
+      "learning_rate": 3.023284744254563e-06,
+      "loss": 0.5518,
+      "step": 9838
+    },
+    {
+      "epoch": 2.6119739811496085,
+      "grad_norm": 0.43593392010590215,
+      "learning_rate": 3.022943344941903e-06,
+      "loss": 0.5583,
+      "step": 9839
+    },
+    {
+      "epoch": 2.612239479622992,
+      "grad_norm": 0.4371588110601602,
+      "learning_rate": 3.022601935430568e-06,
+      "loss": 0.5702,
+      "step": 9840
+    },
+    {
+      "epoch": 2.612504978096376,
+      "grad_norm": 0.44425753489465514,
+      "learning_rate": 3.022260515727215e-06,
+      "loss": 0.525,
+      "step": 9841
+    },
+    {
+      "epoch": 2.61277047656976,
+      "grad_norm": 0.4436515239552187,
+      "learning_rate": 3.0219190858385032e-06,
+      "loss": 0.5727,
+      "step": 9842
+    },
+    {
+      "epoch": 2.6130359750431436,
+      "grad_norm": 0.4353254847410362,
+      "learning_rate": 3.0215776457710923e-06,
+      "loss": 0.5659,
+      "step": 9843
+    },
+    {
+      "epoch": 2.613301473516527,
+      "grad_norm": 0.43021916093285906,
+      "learning_rate": 3.0212361955316386e-06,
+      "loss": 0.5708,
+      "step": 9844
+    },
+    {
+      "epoch": 2.6135669719899113,
+      "grad_norm": 0.4364667903436583,
+      "learning_rate": 3.0208947351268036e-06,
+      "loss": 0.5653,
+      "step": 9845
+    },
+    {
+      "epoch": 2.613832470463295,
+      "grad_norm": 0.44241317567799704,
+      "learning_rate": 3.0205532645632462e-06,
+      "loss": 0.5937,
+      "step": 9846
+    },
+    {
+      "epoch": 2.6140979689366786,
+      "grad_norm": 0.4434230087935956,
+      "learning_rate": 3.0202117838476254e-06,
+      "loss": 0.559,
+      "step": 9847
+    },
+    {
+      "epoch": 2.6143634674100626,
+      "grad_norm": 0.44538061898001013,
+      "learning_rate": 3.019870292986601e-06,
+      "loss": 0.5901,
+      "step": 9848
+    },
+    {
+      "epoch": 2.6146289658834463,
+      "grad_norm": 0.4299221055882074,
+      "learning_rate": 3.019528791986833e-06,
+      "loss": 0.5311,
+      "step": 9849
+    },
+    {
+      "epoch": 2.61489446435683,
+      "grad_norm": 0.4356575337537641,
+      "learning_rate": 3.019187280854982e-06,
+      "loss": 0.5732,
+      "step": 9850
+    },
+    {
+      "epoch": 2.6151599628302136,
+      "grad_norm": 0.4294933456052979,
+      "learning_rate": 3.018845759597708e-06,
+      "loss": 0.5517,
+      "step": 9851
+    },
+    {
+      "epoch": 2.6154254613035977,
+      "grad_norm": 0.43726103254155024,
+      "learning_rate": 3.0185042282216707e-06,
+      "loss": 0.5451,
+      "step": 9852
+    },
+    {
+      "epoch": 2.6156909597769813,
+      "grad_norm": 0.4357481060127005,
+      "learning_rate": 3.018162686733532e-06,
+      "loss": 0.5684,
+      "step": 9853
+    },
+    {
+      "epoch": 2.615956458250365,
+      "grad_norm": 0.4550477948741028,
+      "learning_rate": 3.0178211351399513e-06,
+      "loss": 0.5524,
+      "step": 9854
+    },
+    {
+      "epoch": 2.6162219567237486,
+      "grad_norm": 0.43276500174969923,
+      "learning_rate": 3.017479573447591e-06,
+      "loss": 0.5609,
+      "step": 9855
+    },
+    {
+      "epoch": 2.6164874551971327,
+      "grad_norm": 0.4301267661100456,
+      "learning_rate": 3.017138001663112e-06,
+      "loss": 0.5535,
+      "step": 9856
+    },
+    {
+      "epoch": 2.6167529536705163,
+      "grad_norm": 0.4433054623039358,
+      "learning_rate": 3.016796419793176e-06,
+      "loss": 0.5814,
+      "step": 9857
+    },
+    {
+      "epoch": 2.6170184521439,
+      "grad_norm": 0.4455263422028563,
+      "learning_rate": 3.016454827844445e-06,
+      "loss": 0.5655,
+      "step": 9858
+    },
+    {
+      "epoch": 2.617283950617284,
+      "grad_norm": 0.44251737858832735,
+      "learning_rate": 3.0161132258235793e-06,
+      "loss": 0.569,
+      "step": 9859
+    },
+    {
+      "epoch": 2.6175494490906677,
+      "grad_norm": 0.44778610292819304,
+      "learning_rate": 3.015771613737243e-06,
+      "loss": 0.5762,
+      "step": 9860
+    },
+    {
+      "epoch": 2.6178149475640513,
+      "grad_norm": 0.4474395789662619,
+      "learning_rate": 3.0154299915920972e-06,
+      "loss": 0.5658,
+      "step": 9861
+    },
+    {
+      "epoch": 2.6180804460374354,
+      "grad_norm": 0.44436171688739035,
+      "learning_rate": 3.0150883593948037e-06,
+      "loss": 0.5503,
+      "step": 9862
+    },
+    {
+      "epoch": 2.618345944510819,
+      "grad_norm": 0.4383942255423122,
+      "learning_rate": 3.0147467171520272e-06,
+      "loss": 0.5803,
+      "step": 9863
+    },
+    {
+      "epoch": 2.6186114429842027,
+      "grad_norm": 0.4477474850048624,
+      "learning_rate": 3.014405064870428e-06,
+      "loss": 0.5573,
+      "step": 9864
+    },
+    {
+      "epoch": 2.6188769414575868,
+      "grad_norm": 0.4538311015711708,
+      "learning_rate": 3.014063402556672e-06,
+      "loss": 0.6119,
+      "step": 9865
+    },
+    {
+      "epoch": 2.6191424399309704,
+      "grad_norm": 0.434192142562641,
+      "learning_rate": 3.0137217302174205e-06,
+      "loss": 0.6023,
+      "step": 9866
+    },
+    {
+      "epoch": 2.619407938404354,
+      "grad_norm": 0.443036074110176,
+      "learning_rate": 3.0133800478593373e-06,
+      "loss": 0.6013,
+      "step": 9867
+    },
+    {
+      "epoch": 2.619673436877738,
+      "grad_norm": 0.4420928334380666,
+      "learning_rate": 3.013038355489086e-06,
+      "loss": 0.5735,
+      "step": 9868
+    },
+    {
+      "epoch": 2.6199389353511218,
+      "grad_norm": 0.4540012376124846,
+      "learning_rate": 3.0126966531133313e-06,
+      "loss": 0.5827,
+      "step": 9869
+    },
+    {
+      "epoch": 2.6202044338245054,
+      "grad_norm": 0.42553971704458887,
+      "learning_rate": 3.0123549407387357e-06,
+      "loss": 0.6071,
+      "step": 9870
+    },
+    {
+      "epoch": 2.6204699322978895,
+      "grad_norm": 0.43745212080732393,
+      "learning_rate": 3.0120132183719648e-06,
+      "loss": 0.5476,
+      "step": 9871
+    },
+    {
+      "epoch": 2.620735430771273,
+      "grad_norm": 0.44160750158320355,
+      "learning_rate": 3.011671486019683e-06,
+      "loss": 0.5545,
+      "step": 9872
+    },
+    {
+      "epoch": 2.6210009292446568,
+      "grad_norm": 0.43534641999773344,
+      "learning_rate": 3.0113297436885534e-06,
+      "loss": 0.5464,
+      "step": 9873
+    },
+    {
+      "epoch": 2.621266427718041,
+      "grad_norm": 0.42720012879130215,
+      "learning_rate": 3.010987991385242e-06,
+      "loss": 0.5709,
+      "step": 9874
+    },
+    {
+      "epoch": 2.6215319261914245,
+      "grad_norm": 0.42017140661247165,
+      "learning_rate": 3.0106462291164144e-06,
+      "loss": 0.5332,
+      "step": 9875
+    },
+    {
+      "epoch": 2.621797424664808,
+      "grad_norm": 0.43014527360916427,
+      "learning_rate": 3.0103044568887348e-06,
+      "loss": 0.5485,
+      "step": 9876
+    },
+    {
+      "epoch": 2.6220629231381922,
+      "grad_norm": 0.4502492781589114,
+      "learning_rate": 3.0099626747088682e-06,
+      "loss": 0.5359,
+      "step": 9877
+    },
+    {
+      "epoch": 2.622328421611576,
+      "grad_norm": 0.43447302548629424,
+      "learning_rate": 3.0096208825834817e-06,
+      "loss": 0.5713,
+      "step": 9878
+    },
+    {
+      "epoch": 2.6225939200849595,
+      "grad_norm": 0.43927347409848205,
+      "learning_rate": 3.00927908051924e-06,
+      "loss": 0.6004,
+      "step": 9879
+    },
+    {
+      "epoch": 2.622859418558343,
+      "grad_norm": 0.4501172378341211,
+      "learning_rate": 3.0089372685228093e-06,
+      "loss": 0.5481,
+      "step": 9880
+    },
+    {
+      "epoch": 2.6231249170317272,
+      "grad_norm": 0.4402316360607621,
+      "learning_rate": 3.008595446600856e-06,
+      "loss": 0.5646,
+      "step": 9881
+    },
+    {
+      "epoch": 2.623390415505111,
+      "grad_norm": 0.4431075994849239,
+      "learning_rate": 3.0082536147600454e-06,
+      "loss": 0.571,
+      "step": 9882
+    },
+    {
+      "epoch": 2.6236559139784945,
+      "grad_norm": 0.4293022625828644,
+      "learning_rate": 3.007911773007045e-06,
+      "loss": 0.5746,
+      "step": 9883
+    },
+    {
+      "epoch": 2.623921412451878,
+      "grad_norm": 0.44256524396095787,
+      "learning_rate": 3.007569921348522e-06,
+      "loss": 0.5678,
+      "step": 9884
+    },
+    {
+      "epoch": 2.6241869109252622,
+      "grad_norm": 0.4427629570602814,
+      "learning_rate": 3.0072280597911423e-06,
+      "loss": 0.5637,
+      "step": 9885
+    },
+    {
+      "epoch": 2.624452409398646,
+      "grad_norm": 0.46103695799888533,
+      "learning_rate": 3.0068861883415743e-06,
+      "loss": 0.5798,
+      "step": 9886
+    },
+    {
+      "epoch": 2.6247179078720295,
+      "grad_norm": 0.4471885163500397,
+      "learning_rate": 3.0065443070064843e-06,
+      "loss": 0.551,
+      "step": 9887
+    },
+    {
+      "epoch": 2.6249834063454136,
+      "grad_norm": 0.43172985904282923,
+      "learning_rate": 3.0062024157925396e-06,
+      "loss": 0.5429,
+      "step": 9888
+    },
+    {
+      "epoch": 2.6252489048187972,
+      "grad_norm": 0.4344381378988047,
+      "learning_rate": 3.005860514706409e-06,
+      "loss": 0.5442,
+      "step": 9889
+    },
+    {
+      "epoch": 2.625514403292181,
+      "grad_norm": 0.44152196416238965,
+      "learning_rate": 3.005518603754759e-06,
+      "loss": 0.539,
+      "step": 9890
+    },
+    {
+      "epoch": 2.625779901765565,
+      "grad_norm": 0.43606418948176584,
+      "learning_rate": 3.005176682944259e-06,
+      "loss": 0.5984,
+      "step": 9891
+    },
+    {
+      "epoch": 2.6260454002389486,
+      "grad_norm": 0.4551431622615924,
+      "learning_rate": 3.004834752281577e-06,
+      "loss": 0.5525,
+      "step": 9892
+    },
+    {
+      "epoch": 2.6263108987123323,
+      "grad_norm": 0.4345127376862315,
+      "learning_rate": 3.0044928117733805e-06,
+      "loss": 0.5559,
+      "step": 9893
+    },
+    {
+      "epoch": 2.6265763971857163,
+      "grad_norm": 0.43073686310602943,
+      "learning_rate": 3.004150861426339e-06,
+      "loss": 0.545,
+      "step": 9894
+    },
+    {
+      "epoch": 2.6268418956591,
+      "grad_norm": 0.4561239349515093,
+      "learning_rate": 3.0038089012471226e-06,
+      "loss": 0.6117,
+      "step": 9895
+    },
+    {
+      "epoch": 2.6271073941324836,
+      "grad_norm": 0.4543052337296326,
+      "learning_rate": 3.0034669312423985e-06,
+      "loss": 0.5521,
+      "step": 9896
+    },
+    {
+      "epoch": 2.6273728926058677,
+      "grad_norm": 0.4344483984864306,
+      "learning_rate": 3.003124951418836e-06,
+      "loss": 0.5473,
+      "step": 9897
+    },
+    {
+      "epoch": 2.6276383910792513,
+      "grad_norm": 0.437393101574257,
+      "learning_rate": 3.0027829617831063e-06,
+      "loss": 0.54,
+      "step": 9898
+    },
+    {
+      "epoch": 2.627903889552635,
+      "grad_norm": 0.42975660289869794,
+      "learning_rate": 3.0024409623418765e-06,
+      "loss": 0.5375,
+      "step": 9899
+    },
+    {
+      "epoch": 2.628169388026019,
+      "grad_norm": 0.45856132816048817,
+      "learning_rate": 3.002098953101818e-06,
+      "loss": 0.5619,
+      "step": 9900
+    },
+    {
+      "epoch": 2.6284348864994027,
+      "grad_norm": 0.4252337371629321,
+      "learning_rate": 3.0017569340696008e-06,
+      "loss": 0.5104,
+      "step": 9901
+    },
+    {
+      "epoch": 2.6287003849727864,
+      "grad_norm": 0.45033190648925175,
+      "learning_rate": 3.001414905251895e-06,
+      "loss": 0.5522,
+      "step": 9902
+    },
+    {
+      "epoch": 2.6289658834461704,
+      "grad_norm": 0.4403072129725849,
+      "learning_rate": 3.00107286665537e-06,
+      "loss": 0.5194,
+      "step": 9903
+    },
+    {
+      "epoch": 2.629231381919554,
+      "grad_norm": 0.4384666294140011,
+      "learning_rate": 3.0007308182866983e-06,
+      "loss": 0.5611,
+      "step": 9904
+    },
+    {
+      "epoch": 2.6294968803929377,
+      "grad_norm": 0.4381451154375576,
+      "learning_rate": 3.0003887601525494e-06,
+      "loss": 0.5735,
+      "step": 9905
+    },
+    {
+      "epoch": 2.6297623788663214,
+      "grad_norm": 0.4362011688325651,
+      "learning_rate": 3.000046692259595e-06,
+      "loss": 0.5649,
+      "step": 9906
+    },
+    {
+      "epoch": 2.6300278773397054,
+      "grad_norm": 0.45800339121839356,
+      "learning_rate": 2.999704614614505e-06,
+      "loss": 0.5585,
+      "step": 9907
+    },
+    {
+      "epoch": 2.630293375813089,
+      "grad_norm": 0.4407241866263677,
+      "learning_rate": 2.999362527223952e-06,
+      "loss": 0.5302,
+      "step": 9908
+    },
+    {
+      "epoch": 2.6305588742864727,
+      "grad_norm": 0.43478161686918587,
+      "learning_rate": 2.999020430094606e-06,
+      "loss": 0.5705,
+      "step": 9909
+    },
+    {
+      "epoch": 2.6308243727598564,
+      "grad_norm": 0.4353805070951268,
+      "learning_rate": 2.9986783232331407e-06,
+      "loss": 0.5492,
+      "step": 9910
+    },
+    {
+      "epoch": 2.6310898712332405,
+      "grad_norm": 0.4484856287692562,
+      "learning_rate": 2.9983362066462273e-06,
+      "loss": 0.5795,
+      "step": 9911
+    },
+    {
+      "epoch": 2.631355369706624,
+      "grad_norm": 0.4485160621100812,
+      "learning_rate": 2.997994080340538e-06,
+      "loss": 0.5957,
+      "step": 9912
+    },
+    {
+      "epoch": 2.6316208681800077,
+      "grad_norm": 0.4440385071848229,
+      "learning_rate": 2.9976519443227436e-06,
+      "loss": 0.5788,
+      "step": 9913
+    },
+    {
+      "epoch": 2.631886366653392,
+      "grad_norm": 0.4427672527353162,
+      "learning_rate": 2.9973097985995187e-06,
+      "loss": 0.5546,
+      "step": 9914
+    },
+    {
+      "epoch": 2.6321518651267755,
+      "grad_norm": 0.4283156233381252,
+      "learning_rate": 2.996967643177535e-06,
+      "loss": 0.5464,
+      "step": 9915
+    },
+    {
+      "epoch": 2.632417363600159,
+      "grad_norm": 0.4425679994489716,
+      "learning_rate": 2.9966254780634658e-06,
+      "loss": 0.5757,
+      "step": 9916
+    },
+    {
+      "epoch": 2.632682862073543,
+      "grad_norm": 0.4346918168905528,
+      "learning_rate": 2.996283303263983e-06,
+      "loss": 0.5564,
+      "step": 9917
+    },
+    {
+      "epoch": 2.632948360546927,
+      "grad_norm": 0.4349281578625761,
+      "learning_rate": 2.995941118785762e-06,
+      "loss": 0.5219,
+      "step": 9918
+    },
+    {
+      "epoch": 2.6332138590203105,
+      "grad_norm": 0.4341829099844109,
+      "learning_rate": 2.9955989246354734e-06,
+      "loss": 0.5892,
+      "step": 9919
+    },
+    {
+      "epoch": 2.6334793574936946,
+      "grad_norm": 0.4121911594040556,
+      "learning_rate": 2.9952567208197924e-06,
+      "loss": 0.5156,
+      "step": 9920
+    },
+    {
+      "epoch": 2.633744855967078,
+      "grad_norm": 0.43825855334069336,
+      "learning_rate": 2.9949145073453935e-06,
+      "loss": 0.5512,
+      "step": 9921
+    },
+    {
+      "epoch": 2.634010354440462,
+      "grad_norm": 0.4407661964039348,
+      "learning_rate": 2.9945722842189495e-06,
+      "loss": 0.5996,
+      "step": 9922
+    },
+    {
+      "epoch": 2.634275852913846,
+      "grad_norm": 0.4497807192797199,
+      "learning_rate": 2.9942300514471355e-06,
+      "loss": 0.5732,
+      "step": 9923
+    },
+    {
+      "epoch": 2.6345413513872296,
+      "grad_norm": 0.43875712577908155,
+      "learning_rate": 2.9938878090366253e-06,
+      "loss": 0.597,
+      "step": 9924
+    },
+    {
+      "epoch": 2.634806849860613,
+      "grad_norm": 0.44104759708540336,
+      "learning_rate": 2.9935455569940934e-06,
+      "loss": 0.5857,
+      "step": 9925
+    },
+    {
+      "epoch": 2.6350723483339973,
+      "grad_norm": 0.44576688880128146,
+      "learning_rate": 2.9932032953262146e-06,
+      "loss": 0.5535,
+      "step": 9926
+    },
+    {
+      "epoch": 2.635337846807381,
+      "grad_norm": 0.4306513806485075,
+      "learning_rate": 2.992861024039665e-06,
+      "loss": 0.5685,
+      "step": 9927
+    },
+    {
+      "epoch": 2.6356033452807646,
+      "grad_norm": 0.4595332475907878,
+      "learning_rate": 2.9925187431411174e-06,
+      "loss": 0.5604,
+      "step": 9928
+    },
+    {
+      "epoch": 2.6358688437541486,
+      "grad_norm": 0.429074626196928,
+      "learning_rate": 2.9921764526372488e-06,
+      "loss": 0.5458,
+      "step": 9929
+    },
+    {
+      "epoch": 2.6361343422275323,
+      "grad_norm": 0.45375906708898867,
+      "learning_rate": 2.9918341525347345e-06,
+      "loss": 0.5887,
+      "step": 9930
+    },
+    {
+      "epoch": 2.636399840700916,
+      "grad_norm": 0.43517054361955426,
+      "learning_rate": 2.9914918428402494e-06,
+      "loss": 0.4769,
+      "step": 9931
+    },
+    {
+      "epoch": 2.6366653391743,
+      "grad_norm": 0.4518620669062071,
+      "learning_rate": 2.991149523560471e-06,
+      "loss": 0.5568,
+      "step": 9932
+    },
+    {
+      "epoch": 2.6369308376476837,
+      "grad_norm": 0.433411098323447,
+      "learning_rate": 2.990807194702074e-06,
+      "loss": 0.5756,
+      "step": 9933
+    },
+    {
+      "epoch": 2.6371963361210673,
+      "grad_norm": 0.44459924396507006,
+      "learning_rate": 2.990464856271735e-06,
+      "loss": 0.5721,
+      "step": 9934
+    },
+    {
+      "epoch": 2.637461834594451,
+      "grad_norm": 0.43190456418692275,
+      "learning_rate": 2.9901225082761305e-06,
+      "loss": 0.5586,
+      "step": 9935
+    },
+    {
+      "epoch": 2.637727333067835,
+      "grad_norm": 0.4451665915362643,
+      "learning_rate": 2.9897801507219367e-06,
+      "loss": 0.5634,
+      "step": 9936
+    },
+    {
+      "epoch": 2.6379928315412187,
+      "grad_norm": 0.4344907995497725,
+      "learning_rate": 2.9894377836158307e-06,
+      "loss": 0.5739,
+      "step": 9937
+    },
+    {
+      "epoch": 2.6382583300146023,
+      "grad_norm": 0.4261535256584437,
+      "learning_rate": 2.9890954069644897e-06,
+      "loss": 0.5434,
+      "step": 9938
+    },
+    {
+      "epoch": 2.638523828487986,
+      "grad_norm": 0.43095871116470313,
+      "learning_rate": 2.988753020774591e-06,
+      "loss": 0.5514,
+      "step": 9939
+    },
+    {
+      "epoch": 2.63878932696137,
+      "grad_norm": 0.43417960813487233,
+      "learning_rate": 2.988410625052811e-06,
+      "loss": 0.5786,
+      "step": 9940
+    },
+    {
+      "epoch": 2.6390548254347537,
+      "grad_norm": 0.45532746169128974,
+      "learning_rate": 2.9880682198058284e-06,
+      "loss": 0.5786,
+      "step": 9941
+    },
+    {
+      "epoch": 2.6393203239081373,
+      "grad_norm": 0.4370798719342273,
+      "learning_rate": 2.9877258050403214e-06,
+      "loss": 0.5647,
+      "step": 9942
+    },
+    {
+      "epoch": 2.6395858223815214,
+      "grad_norm": 0.4527565676041506,
+      "learning_rate": 2.987383380762966e-06,
+      "loss": 0.5384,
+      "step": 9943
+    },
+    {
+      "epoch": 2.639851320854905,
+      "grad_norm": 0.4591415829286425,
+      "learning_rate": 2.987040946980442e-06,
+      "loss": 0.5869,
+      "step": 9944
+    },
+    {
+      "epoch": 2.6401168193282887,
+      "grad_norm": 0.4330559800873046,
+      "learning_rate": 2.9866985036994268e-06,
+      "loss": 0.5833,
+      "step": 9945
+    },
+    {
+      "epoch": 2.6403823178016728,
+      "grad_norm": 0.44154724894431163,
+      "learning_rate": 2.986356050926599e-06,
+      "loss": 0.5853,
+      "step": 9946
+    },
+    {
+      "epoch": 2.6406478162750564,
+      "grad_norm": 0.43497758344182674,
+      "learning_rate": 2.986013588668637e-06,
+      "loss": 0.563,
+      "step": 9947
+    },
+    {
+      "epoch": 2.64091331474844,
+      "grad_norm": 0.4558119909403106,
+      "learning_rate": 2.9856711169322202e-06,
+      "loss": 0.5331,
+      "step": 9948
+    },
+    {
+      "epoch": 2.641178813221824,
+      "grad_norm": 0.4297536496844275,
+      "learning_rate": 2.9853286357240275e-06,
+      "loss": 0.5675,
+      "step": 9949
+    },
+    {
+      "epoch": 2.6414443116952078,
+      "grad_norm": 0.4471329238556466,
+      "learning_rate": 2.984986145050739e-06,
+      "loss": 0.5924,
+      "step": 9950
+    },
+    {
+      "epoch": 2.6417098101685914,
+      "grad_norm": 0.4327496016628824,
+      "learning_rate": 2.9846436449190326e-06,
+      "loss": 0.5652,
+      "step": 9951
+    },
+    {
+      "epoch": 2.6419753086419755,
+      "grad_norm": 0.4347367031128844,
+      "learning_rate": 2.9843011353355885e-06,
+      "loss": 0.588,
+      "step": 9952
+    },
+    {
+      "epoch": 2.642240807115359,
+      "grad_norm": 0.4290859707748789,
+      "learning_rate": 2.9839586163070867e-06,
+      "loss": 0.5634,
+      "step": 9953
+    },
+    {
+      "epoch": 2.6425063055887428,
+      "grad_norm": 0.4340151566882483,
+      "learning_rate": 2.9836160878402064e-06,
+      "loss": 0.5268,
+      "step": 9954
+    },
+    {
+      "epoch": 2.642771804062127,
+      "grad_norm": 0.4347428623442634,
+      "learning_rate": 2.9832735499416287e-06,
+      "loss": 0.5568,
+      "step": 9955
+    },
+    {
+      "epoch": 2.6430373025355105,
+      "grad_norm": 0.44629294872974323,
+      "learning_rate": 2.9829310026180336e-06,
+      "loss": 0.5755,
+      "step": 9956
+    },
+    {
+      "epoch": 2.643302801008894,
+      "grad_norm": 0.44350117372897757,
+      "learning_rate": 2.9825884458761014e-06,
+      "loss": 0.5902,
+      "step": 9957
+    },
+    {
+      "epoch": 2.6435682994822782,
+      "grad_norm": 0.4347294438140043,
+      "learning_rate": 2.982245879722513e-06,
+      "loss": 0.5229,
+      "step": 9958
+    },
+    {
+      "epoch": 2.643833797955662,
+      "grad_norm": 0.464309377329224,
+      "learning_rate": 2.981903304163949e-06,
+      "loss": 0.5763,
+      "step": 9959
+    },
+    {
+      "epoch": 2.6440992964290455,
+      "grad_norm": 0.43347576524715087,
+      "learning_rate": 2.9815607192070905e-06,
+      "loss": 0.5316,
+      "step": 9960
+    },
+    {
+      "epoch": 2.644364794902429,
+      "grad_norm": 0.422618593162229,
+      "learning_rate": 2.9812181248586195e-06,
+      "loss": 0.5362,
+      "step": 9961
+    },
+    {
+      "epoch": 2.6446302933758132,
+      "grad_norm": 0.47056180163125516,
+      "learning_rate": 2.9808755211252166e-06,
+      "loss": 0.5674,
+      "step": 9962
+    },
+    {
+      "epoch": 2.644895791849197,
+      "grad_norm": 0.44377808454261547,
+      "learning_rate": 2.9805329080135636e-06,
+      "loss": 0.5699,
+      "step": 9963
+    },
+    {
+      "epoch": 2.6451612903225805,
+      "grad_norm": 0.4433120148103743,
+      "learning_rate": 2.9801902855303427e-06,
+      "loss": 0.5698,
+      "step": 9964
+    },
+    {
+      "epoch": 2.645426788795964,
+      "grad_norm": 0.43503419818805555,
+      "learning_rate": 2.9798476536822344e-06,
+      "loss": 0.5714,
+      "step": 9965
+    },
+    {
+      "epoch": 2.6456922872693482,
+      "grad_norm": 0.4238848177573442,
+      "learning_rate": 2.979505012475923e-06,
+      "loss": 0.5192,
+      "step": 9966
+    },
+    {
+      "epoch": 2.645957785742732,
+      "grad_norm": 0.4245364360524084,
+      "learning_rate": 2.9791623619180896e-06,
+      "loss": 0.5587,
+      "step": 9967
+    },
+    {
+      "epoch": 2.6462232842161155,
+      "grad_norm": 0.43412496439907994,
+      "learning_rate": 2.978819702015417e-06,
+      "loss": 0.5449,
+      "step": 9968
+    },
+    {
+      "epoch": 2.6464887826894996,
+      "grad_norm": 0.43815098599682767,
+      "learning_rate": 2.9784770327745875e-06,
+      "loss": 0.5539,
+      "step": 9969
+    },
+    {
+      "epoch": 2.6467542811628832,
+      "grad_norm": 0.42078824292443223,
+      "learning_rate": 2.978134354202285e-06,
+      "loss": 0.5345,
+      "step": 9970
+    },
+    {
+      "epoch": 2.647019779636267,
+      "grad_norm": 0.44171974494035376,
+      "learning_rate": 2.9777916663051916e-06,
+      "loss": 0.5511,
+      "step": 9971
+    },
+    {
+      "epoch": 2.647285278109651,
+      "grad_norm": 0.4367177314804473,
+      "learning_rate": 2.977448969089991e-06,
+      "loss": 0.5583,
+      "step": 9972
+    },
+    {
+      "epoch": 2.6475507765830346,
+      "grad_norm": 0.43275022214441033,
+      "learning_rate": 2.9771062625633666e-06,
+      "loss": 0.5498,
+      "step": 9973
+    },
+    {
+      "epoch": 2.6478162750564183,
+      "grad_norm": 0.4413254821738734,
+      "learning_rate": 2.9767635467320018e-06,
+      "loss": 0.5433,
+      "step": 9974
+    },
+    {
+      "epoch": 2.6480817735298023,
+      "grad_norm": 0.43812595901271423,
+      "learning_rate": 2.9764208216025797e-06,
+      "loss": 0.5359,
+      "step": 9975
+    },
+    {
+      "epoch": 2.648347272003186,
+      "grad_norm": 0.43561665666414756,
+      "learning_rate": 2.9760780871817856e-06,
+      "loss": 0.5391,
+      "step": 9976
+    },
+    {
+      "epoch": 2.6486127704765696,
+      "grad_norm": 0.4404188912623667,
+      "learning_rate": 2.9757353434763037e-06,
+      "loss": 0.5586,
+      "step": 9977
+    },
+    {
+      "epoch": 2.6488782689499537,
+      "grad_norm": 0.4265068887118984,
+      "learning_rate": 2.9753925904928173e-06,
+      "loss": 0.55,
+      "step": 9978
+    },
+    {
+      "epoch": 2.6491437674233373,
+      "grad_norm": 0.4334957993371011,
+      "learning_rate": 2.975049828238012e-06,
+      "loss": 0.5738,
+      "step": 9979
+    },
+    {
+      "epoch": 2.649409265896721,
+      "grad_norm": 0.42963775415310124,
+      "learning_rate": 2.974707056718571e-06,
+      "loss": 0.5375,
+      "step": 9980
+    },
+    {
+      "epoch": 2.649674764370105,
+      "grad_norm": 0.44812608556831035,
+      "learning_rate": 2.974364275941181e-06,
+      "loss": 0.5501,
+      "step": 9981
+    },
+    {
+      "epoch": 2.6499402628434887,
+      "grad_norm": 0.42247982002093054,
+      "learning_rate": 2.974021485912526e-06,
+      "loss": 0.5337,
+      "step": 9982
+    },
+    {
+      "epoch": 2.6502057613168724,
+      "grad_norm": 0.446713280995582,
+      "learning_rate": 2.973678686639291e-06,
+      "loss": 0.5725,
+      "step": 9983
+    },
+    {
+      "epoch": 2.6504712597902564,
+      "grad_norm": 0.43818268042763436,
+      "learning_rate": 2.973335878128162e-06,
+      "loss": 0.5717,
+      "step": 9984
+    },
+    {
+      "epoch": 2.65073675826364,
+      "grad_norm": 0.428673354337855,
+      "learning_rate": 2.9729930603858243e-06,
+      "loss": 0.5328,
+      "step": 9985
+    },
+    {
+      "epoch": 2.6510022567370237,
+      "grad_norm": 0.4308792338084989,
+      "learning_rate": 2.972650233418964e-06,
+      "loss": 0.545,
+      "step": 9986
+    },
+    {
+      "epoch": 2.651267755210408,
+      "grad_norm": 0.42699121503704074,
+      "learning_rate": 2.972307397234267e-06,
+      "loss": 0.5471,
+      "step": 9987
+    },
+    {
+      "epoch": 2.6515332536837914,
+      "grad_norm": 0.43465959699226836,
+      "learning_rate": 2.9719645518384194e-06,
+      "loss": 0.5461,
+      "step": 9988
+    },
+    {
+      "epoch": 2.651798752157175,
+      "grad_norm": 0.44647737583421293,
+      "learning_rate": 2.9716216972381068e-06,
+      "loss": 0.536,
+      "step": 9989
+    },
+    {
+      "epoch": 2.6520642506305587,
+      "grad_norm": 0.43951533700001766,
+      "learning_rate": 2.971278833440018e-06,
+      "loss": 0.5792,
+      "step": 9990
+    },
+    {
+      "epoch": 2.652329749103943,
+      "grad_norm": 0.4388284095506071,
+      "learning_rate": 2.9709359604508365e-06,
+      "loss": 0.594,
+      "step": 9991
+    },
+    {
+      "epoch": 2.6525952475773265,
+      "grad_norm": 0.4476194358008474,
+      "learning_rate": 2.970593078277251e-06,
+      "loss": 0.5772,
+      "step": 9992
+    },
+    {
+      "epoch": 2.65286074605071,
+      "grad_norm": 0.4346177530092667,
+      "learning_rate": 2.970250186925948e-06,
+      "loss": 0.5122,
+      "step": 9993
+    },
+    {
+      "epoch": 2.6531262445240937,
+      "grad_norm": 0.4259194812524258,
+      "learning_rate": 2.9699072864036155e-06,
+      "loss": 0.5305,
+      "step": 9994
+    },
+    {
+      "epoch": 2.653391742997478,
+      "grad_norm": 0.43401860691983807,
+      "learning_rate": 2.9695643767169404e-06,
+      "loss": 0.5495,
+      "step": 9995
+    },
+    {
+      "epoch": 2.6536572414708615,
+      "grad_norm": 0.438111486381578,
+      "learning_rate": 2.9692214578726105e-06,
+      "loss": 0.5741,
+      "step": 9996
+    },
+    {
+      "epoch": 2.653922739944245,
+      "grad_norm": 0.4447872054942858,
+      "learning_rate": 2.9688785298773127e-06,
+      "loss": 0.5716,
+      "step": 9997
+    },
+    {
+      "epoch": 2.654188238417629,
+      "grad_norm": 0.44147823236688527,
+      "learning_rate": 2.9685355927377356e-06,
+      "loss": 0.5669,
+      "step": 9998
+    },
+    {
+      "epoch": 2.654453736891013,
+      "grad_norm": 0.4543819093558324,
+      "learning_rate": 2.9681926464605683e-06,
+      "loss": 0.5789,
+      "step": 9999
+    },
+    {
+      "epoch": 2.6547192353643965,
+      "grad_norm": 0.44005221699009817,
+      "learning_rate": 2.967849691052497e-06,
+      "loss": 0.5728,
+      "step": 10000
+    },
+    {
+      "epoch": 2.6549847338377806,
+      "grad_norm": 0.4433984396748064,
+      "learning_rate": 2.9675067265202116e-06,
+      "loss": 0.5734,
+      "step": 10001
+    },
+    {
+      "epoch": 2.655250232311164,
+      "grad_norm": 0.4435068602968665,
+      "learning_rate": 2.9671637528704e-06,
+      "loss": 0.5844,
+      "step": 10002
+    },
+    {
+      "epoch": 2.655515730784548,
+      "grad_norm": 0.44560535059295125,
+      "learning_rate": 2.9668207701097516e-06,
+      "loss": 0.5588,
+      "step": 10003
+    },
+    {
+      "epoch": 2.655781229257932,
+      "grad_norm": 0.435557242600705,
+      "learning_rate": 2.966477778244956e-06,
+      "loss": 0.5109,
+      "step": 10004
+    },
+    {
+      "epoch": 2.6560467277313156,
+      "grad_norm": 0.441992460502229,
+      "learning_rate": 2.9661347772827003e-06,
+      "loss": 0.567,
+      "step": 10005
+    },
+    {
+      "epoch": 2.656312226204699,
+      "grad_norm": 0.43085623711822096,
+      "learning_rate": 2.965791767229676e-06,
+      "loss": 0.5371,
+      "step": 10006
+    },
+    {
+      "epoch": 2.6565777246780833,
+      "grad_norm": 0.4301600965929788,
+      "learning_rate": 2.965448748092572e-06,
+      "loss": 0.5421,
+      "step": 10007
+    },
+    {
+      "epoch": 2.656843223151467,
+      "grad_norm": 0.4343535198366537,
+      "learning_rate": 2.965105719878077e-06,
+      "loss": 0.5956,
+      "step": 10008
+    },
+    {
+      "epoch": 2.6571087216248506,
+      "grad_norm": 0.43443423761620603,
+      "learning_rate": 2.9647626825928823e-06,
+      "loss": 0.5669,
+      "step": 10009
+    },
+    {
+      "epoch": 2.6573742200982347,
+      "grad_norm": 0.4391693690322364,
+      "learning_rate": 2.9644196362436774e-06,
+      "loss": 0.5928,
+      "step": 10010
+    },
+    {
+      "epoch": 2.6576397185716183,
+      "grad_norm": 0.4385278233766275,
+      "learning_rate": 2.9640765808371513e-06,
+      "loss": 0.5803,
+      "step": 10011
+    },
+    {
+      "epoch": 2.657905217045002,
+      "grad_norm": 0.4487504103378222,
+      "learning_rate": 2.9637335163799965e-06,
+      "loss": 0.5444,
+      "step": 10012
+    },
+    {
+      "epoch": 2.658170715518386,
+      "grad_norm": 0.43886768781999347,
+      "learning_rate": 2.9633904428789027e-06,
+      "loss": 0.5436,
+      "step": 10013
+    },
+    {
+      "epoch": 2.6584362139917697,
+      "grad_norm": 0.44898417845346733,
+      "learning_rate": 2.9630473603405607e-06,
+      "loss": 0.5136,
+      "step": 10014
+    },
+    {
+      "epoch": 2.6587017124651533,
+      "grad_norm": 0.4325919937780632,
+      "learning_rate": 2.9627042687716613e-06,
+      "loss": 0.5362,
+      "step": 10015
+    },
+    {
+      "epoch": 2.658967210938537,
+      "grad_norm": 0.44319090233236563,
+      "learning_rate": 2.9623611681788967e-06,
+      "loss": 0.5911,
+      "step": 10016
+    },
+    {
+      "epoch": 2.659232709411921,
+      "grad_norm": 0.44158058769615516,
+      "learning_rate": 2.962018058568956e-06,
+      "loss": 0.5371,
+      "step": 10017
+    },
+    {
+      "epoch": 2.6594982078853047,
+      "grad_norm": 0.4738186887563765,
+      "learning_rate": 2.9616749399485323e-06,
+      "loss": 0.5562,
+      "step": 10018
+    },
+    {
+      "epoch": 2.6597637063586883,
+      "grad_norm": 0.44282116896878837,
+      "learning_rate": 2.961331812324317e-06,
+      "loss": 0.5845,
+      "step": 10019
+    },
+    {
+      "epoch": 2.660029204832072,
+      "grad_norm": 0.44391927108306606,
+      "learning_rate": 2.960988675703002e-06,
+      "loss": 0.5856,
+      "step": 10020
+    },
+    {
+      "epoch": 2.660294703305456,
+      "grad_norm": 0.43767251259841244,
+      "learning_rate": 2.9606455300912782e-06,
+      "loss": 0.5998,
+      "step": 10021
+    },
+    {
+      "epoch": 2.6605602017788397,
+      "grad_norm": 0.45773285112184725,
+      "learning_rate": 2.96030237549584e-06,
+      "loss": 0.5222,
+      "step": 10022
+    },
+    {
+      "epoch": 2.6608257002522233,
+      "grad_norm": 0.4418414236532392,
+      "learning_rate": 2.9599592119233777e-06,
+      "loss": 0.572,
+      "step": 10023
+    },
+    {
+      "epoch": 2.6610911987256074,
+      "grad_norm": 0.4440806721483077,
+      "learning_rate": 2.959616039380585e-06,
+      "loss": 0.5841,
+      "step": 10024
+    },
+    {
+      "epoch": 2.661356697198991,
+      "grad_norm": 0.4393298063617331,
+      "learning_rate": 2.9592728578741544e-06,
+      "loss": 0.5641,
+      "step": 10025
+    },
+    {
+      "epoch": 2.6616221956723747,
+      "grad_norm": 0.43115786132548123,
+      "learning_rate": 2.9589296674107783e-06,
+      "loss": 0.5792,
+      "step": 10026
+    },
+    {
+      "epoch": 2.6618876941457588,
+      "grad_norm": 0.438327381260109,
+      "learning_rate": 2.9585864679971497e-06,
+      "loss": 0.5261,
+      "step": 10027
+    },
+    {
+      "epoch": 2.6621531926191424,
+      "grad_norm": 0.4371737634947093,
+      "learning_rate": 2.9582432596399624e-06,
+      "loss": 0.5566,
+      "step": 10028
+    },
+    {
+      "epoch": 2.662418691092526,
+      "grad_norm": 0.42994922358797977,
+      "learning_rate": 2.9579000423459097e-06,
+      "loss": 0.5421,
+      "step": 10029
+    },
+    {
+      "epoch": 2.66268418956591,
+      "grad_norm": 0.4280656448124467,
+      "learning_rate": 2.957556816121685e-06,
+      "loss": 0.5735,
+      "step": 10030
+    },
+    {
+      "epoch": 2.6629496880392938,
+      "grad_norm": 0.435304777121573,
+      "learning_rate": 2.9572135809739815e-06,
+      "loss": 0.5599,
+      "step": 10031
+    },
+    {
+      "epoch": 2.6632151865126774,
+      "grad_norm": 0.44844620330297785,
+      "learning_rate": 2.9568703369094944e-06,
+      "loss": 0.5264,
+      "step": 10032
+    },
+    {
+      "epoch": 2.6634806849860615,
+      "grad_norm": 0.43924250591074826,
+      "learning_rate": 2.956527083934917e-06,
+      "loss": 0.5825,
+      "step": 10033
+    },
+    {
+      "epoch": 2.663746183459445,
+      "grad_norm": 0.4279164463099742,
+      "learning_rate": 2.956183822056944e-06,
+      "loss": 0.5629,
+      "step": 10034
+    },
+    {
+      "epoch": 2.664011681932829,
+      "grad_norm": 0.44262640683113946,
+      "learning_rate": 2.9558405512822686e-06,
+      "loss": 0.5473,
+      "step": 10035
+    },
+    {
+      "epoch": 2.664277180406213,
+      "grad_norm": 0.42727783753543735,
+      "learning_rate": 2.9554972716175872e-06,
+      "loss": 0.5513,
+      "step": 10036
+    },
+    {
+      "epoch": 2.6645426788795965,
+      "grad_norm": 0.4460294998299989,
+      "learning_rate": 2.9551539830695936e-06,
+      "loss": 0.5693,
+      "step": 10037
+    },
+    {
+      "epoch": 2.66480817735298,
+      "grad_norm": 0.43635902921422853,
+      "learning_rate": 2.9548106856449827e-06,
+      "loss": 0.532,
+      "step": 10038
+    },
+    {
+      "epoch": 2.6650736758263642,
+      "grad_norm": 0.4300929186771159,
+      "learning_rate": 2.9544673793504495e-06,
+      "loss": 0.5355,
+      "step": 10039
+    },
+    {
+      "epoch": 2.665339174299748,
+      "grad_norm": 0.4498479112839448,
+      "learning_rate": 2.95412406419269e-06,
+      "loss": 0.5393,
+      "step": 10040
+    },
+    {
+      "epoch": 2.6656046727731315,
+      "grad_norm": 0.4481747122242704,
+      "learning_rate": 2.9537807401783995e-06,
+      "loss": 0.49,
+      "step": 10041
+    },
+    {
+      "epoch": 2.6658701712465156,
+      "grad_norm": 0.42922945483704783,
+      "learning_rate": 2.953437407314274e-06,
+      "loss": 0.5457,
+      "step": 10042
+    },
+    {
+      "epoch": 2.6661356697198992,
+      "grad_norm": 0.45067682881804483,
+      "learning_rate": 2.9530940656070085e-06,
+      "loss": 0.5621,
+      "step": 10043
+    },
+    {
+      "epoch": 2.666401168193283,
+      "grad_norm": 0.4423377907769823,
+      "learning_rate": 2.9527507150632987e-06,
+      "loss": 0.5418,
+      "step": 10044
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.4576771676696478,
+      "learning_rate": 2.9524073556898427e-06,
+      "loss": 0.5999,
+      "step": 10045
+    },
+    {
+      "epoch": 2.6669321651400506,
+      "grad_norm": 0.41119082720464906,
+      "learning_rate": 2.9520639874933344e-06,
+      "loss": 0.5165,
+      "step": 10046
+    },
+    {
+      "epoch": 2.6671976636134342,
+      "grad_norm": 0.4407026403123097,
+      "learning_rate": 2.951720610480472e-06,
+      "loss": 0.545,
+      "step": 10047
+    },
+    {
+      "epoch": 2.667463162086818,
+      "grad_norm": 0.43943649703725896,
+      "learning_rate": 2.9513772246579513e-06,
+      "loss": 0.5383,
+      "step": 10048
+    },
+    {
+      "epoch": 2.6677286605602015,
+      "grad_norm": 0.4452665571103192,
+      "learning_rate": 2.9510338300324696e-06,
+      "loss": 0.53,
+      "step": 10049
+    },
+    {
+      "epoch": 2.6679941590335856,
+      "grad_norm": 0.45010291812331304,
+      "learning_rate": 2.950690426610724e-06,
+      "loss": 0.5865,
+      "step": 10050
+    },
+    {
+      "epoch": 2.6682596575069693,
+      "grad_norm": 0.43953229645860004,
+      "learning_rate": 2.9503470143994125e-06,
+      "loss": 0.5756,
+      "step": 10051
+    },
+    {
+      "epoch": 2.668525155980353,
+      "grad_norm": 0.43205631043090725,
+      "learning_rate": 2.9500035934052307e-06,
+      "loss": 0.546,
+      "step": 10052
+    },
+    {
+      "epoch": 2.668790654453737,
+      "grad_norm": 0.44896614006803726,
+      "learning_rate": 2.949660163634878e-06,
+      "loss": 0.5665,
+      "step": 10053
+    },
+    {
+      "epoch": 2.6690561529271206,
+      "grad_norm": 0.4448018299809349,
+      "learning_rate": 2.9493167250950492e-06,
+      "loss": 0.568,
+      "step": 10054
+    },
+    {
+      "epoch": 2.6693216514005043,
+      "grad_norm": 0.4397218621913262,
+      "learning_rate": 2.9489732777924455e-06,
+      "loss": 0.5492,
+      "step": 10055
+    },
+    {
+      "epoch": 2.6695871498738883,
+      "grad_norm": 0.4396828568097959,
+      "learning_rate": 2.948629821733764e-06,
+      "loss": 0.5266,
+      "step": 10056
+    },
+    {
+      "epoch": 2.669852648347272,
+      "grad_norm": 0.4350291166258978,
+      "learning_rate": 2.9482863569257014e-06,
+      "loss": 0.5103,
+      "step": 10057
+    },
+    {
+      "epoch": 2.6701181468206556,
+      "grad_norm": 0.4306601019035005,
+      "learning_rate": 2.9479428833749573e-06,
+      "loss": 0.5381,
+      "step": 10058
+    },
+    {
+      "epoch": 2.6703836452940397,
+      "grad_norm": 0.4562449666651449,
+      "learning_rate": 2.9475994010882314e-06,
+      "loss": 0.5645,
+      "step": 10059
+    },
+    {
+      "epoch": 2.6706491437674234,
+      "grad_norm": 0.4349976173282006,
+      "learning_rate": 2.9472559100722202e-06,
+      "loss": 0.5502,
+      "step": 10060
+    },
+    {
+      "epoch": 2.670914642240807,
+      "grad_norm": 0.4425134905784369,
+      "learning_rate": 2.946912410333624e-06,
+      "loss": 0.5698,
+      "step": 10061
+    },
+    {
+      "epoch": 2.671180140714191,
+      "grad_norm": 0.45366306352877084,
+      "learning_rate": 2.9465689018791423e-06,
+      "loss": 0.5346,
+      "step": 10062
+    },
+    {
+      "epoch": 2.6714456391875747,
+      "grad_norm": 0.4394334199558341,
+      "learning_rate": 2.9462253847154727e-06,
+      "loss": 0.5256,
+      "step": 10063
+    },
+    {
+      "epoch": 2.6717111376609584,
+      "grad_norm": 0.4420974366829381,
+      "learning_rate": 2.9458818588493154e-06,
+      "loss": 0.5544,
+      "step": 10064
+    },
+    {
+      "epoch": 2.6719766361343424,
+      "grad_norm": 0.4644453017402358,
+      "learning_rate": 2.945538324287371e-06,
+      "loss": 0.5545,
+      "step": 10065
+    },
+    {
+      "epoch": 2.672242134607726,
+      "grad_norm": 0.4452663875066514,
+      "learning_rate": 2.945194781036338e-06,
+      "loss": 0.5679,
+      "step": 10066
+    },
+    {
+      "epoch": 2.6725076330811097,
+      "grad_norm": 0.4415150797145218,
+      "learning_rate": 2.9448512291029157e-06,
+      "loss": 0.5903,
+      "step": 10067
+    },
+    {
+      "epoch": 2.672773131554494,
+      "grad_norm": 0.4408875375158782,
+      "learning_rate": 2.944507668493807e-06,
+      "loss": 0.5524,
+      "step": 10068
+    },
+    {
+      "epoch": 2.6730386300278774,
+      "grad_norm": 0.4487720661841312,
+      "learning_rate": 2.9441640992157097e-06,
+      "loss": 0.5607,
+      "step": 10069
+    },
+    {
+      "epoch": 2.673304128501261,
+      "grad_norm": 0.4436844731695859,
+      "learning_rate": 2.9438205212753246e-06,
+      "loss": 0.5467,
+      "step": 10070
+    },
+    {
+      "epoch": 2.6735696269746447,
+      "grad_norm": 0.4476656941026381,
+      "learning_rate": 2.9434769346793533e-06,
+      "loss": 0.5551,
+      "step": 10071
+    },
+    {
+      "epoch": 2.673835125448029,
+      "grad_norm": 0.4350859738000993,
+      "learning_rate": 2.943133339434495e-06,
+      "loss": 0.5613,
+      "step": 10072
+    },
+    {
+      "epoch": 2.6741006239214125,
+      "grad_norm": 0.44007298510078674,
+      "learning_rate": 2.942789735547453e-06,
+      "loss": 0.5389,
+      "step": 10073
+    },
+    {
+      "epoch": 2.674366122394796,
+      "grad_norm": 0.46238194214742406,
+      "learning_rate": 2.942446123024926e-06,
+      "loss": 0.5482,
+      "step": 10074
+    },
+    {
+      "epoch": 2.6746316208681797,
+      "grad_norm": 0.4296175389621784,
+      "learning_rate": 2.9421025018736165e-06,
+      "loss": 0.5609,
+      "step": 10075
+    },
+    {
+      "epoch": 2.674897119341564,
+      "grad_norm": 0.43174159564041675,
+      "learning_rate": 2.941758872100226e-06,
+      "loss": 0.5265,
+      "step": 10076
+    },
+    {
+      "epoch": 2.6751626178149475,
+      "grad_norm": 0.424370489314813,
+      "learning_rate": 2.9414152337114556e-06,
+      "loss": 0.5421,
+      "step": 10077
+    },
+    {
+      "epoch": 2.675428116288331,
+      "grad_norm": 0.44090883390840724,
+      "learning_rate": 2.941071586714007e-06,
+      "loss": 0.58,
+      "step": 10078
+    },
+    {
+      "epoch": 2.675693614761715,
+      "grad_norm": 0.43899599596927275,
+      "learning_rate": 2.9407279311145835e-06,
+      "loss": 0.5541,
+      "step": 10079
+    },
+    {
+      "epoch": 2.675959113235099,
+      "grad_norm": 0.4495020794366583,
+      "learning_rate": 2.9403842669198855e-06,
+      "loss": 0.5765,
+      "step": 10080
+    },
+    {
+      "epoch": 2.6762246117084825,
+      "grad_norm": 0.43188001085835753,
+      "learning_rate": 2.9400405941366166e-06,
+      "loss": 0.5382,
+      "step": 10081
+    },
+    {
+      "epoch": 2.6764901101818666,
+      "grad_norm": 0.4250489084525347,
+      "learning_rate": 2.9396969127714787e-06,
+      "loss": 0.5347,
+      "step": 10082
+    },
+    {
+      "epoch": 2.67675560865525,
+      "grad_norm": 0.4292068204209661,
+      "learning_rate": 2.939353222831174e-06,
+      "loss": 0.5336,
+      "step": 10083
+    },
+    {
+      "epoch": 2.677021107128634,
+      "grad_norm": 0.4412137586257352,
+      "learning_rate": 2.939009524322406e-06,
+      "loss": 0.5385,
+      "step": 10084
+    },
+    {
+      "epoch": 2.677286605602018,
+      "grad_norm": 0.44353663025248385,
+      "learning_rate": 2.9386658172518766e-06,
+      "loss": 0.5456,
+      "step": 10085
+    },
+    {
+      "epoch": 2.6775521040754016,
+      "grad_norm": 0.4372829510332513,
+      "learning_rate": 2.938322101626291e-06,
+      "loss": 0.5663,
+      "step": 10086
+    },
+    {
+      "epoch": 2.677817602548785,
+      "grad_norm": 0.4398102795718385,
+      "learning_rate": 2.93797837745235e-06,
+      "loss": 0.5176,
+      "step": 10087
+    },
+    {
+      "epoch": 2.6780831010221693,
+      "grad_norm": 0.433796774199086,
+      "learning_rate": 2.9376346447367598e-06,
+      "loss": 0.5778,
+      "step": 10088
+    },
+    {
+      "epoch": 2.678348599495553,
+      "grad_norm": 0.4409864604726168,
+      "learning_rate": 2.9372909034862214e-06,
+      "loss": 0.5854,
+      "step": 10089
+    },
+    {
+      "epoch": 2.6786140979689366,
+      "grad_norm": 0.4384980528467308,
+      "learning_rate": 2.9369471537074394e-06,
+      "loss": 0.5878,
+      "step": 10090
+    },
+    {
+      "epoch": 2.6788795964423207,
+      "grad_norm": 0.444255643847712,
+      "learning_rate": 2.9366033954071193e-06,
+      "loss": 0.5684,
+      "step": 10091
+    },
+    {
+      "epoch": 2.6791450949157043,
+      "grad_norm": 0.43641165987046193,
+      "learning_rate": 2.936259628591963e-06,
+      "loss": 0.5162,
+      "step": 10092
+    },
+    {
+      "epoch": 2.679410593389088,
+      "grad_norm": 0.44951006958696466,
+      "learning_rate": 2.935915853268676e-06,
+      "loss": 0.5661,
+      "step": 10093
+    },
+    {
+      "epoch": 2.679676091862472,
+      "grad_norm": 0.4313204473907715,
+      "learning_rate": 2.9355720694439627e-06,
+      "loss": 0.582,
+      "step": 10094
+    },
+    {
+      "epoch": 2.6799415903358557,
+      "grad_norm": 0.43933106602371624,
+      "learning_rate": 2.9352282771245277e-06,
+      "loss": 0.5557,
+      "step": 10095
+    },
+    {
+      "epoch": 2.6802070888092393,
+      "grad_norm": 0.43892682172970526,
+      "learning_rate": 2.9348844763170754e-06,
+      "loss": 0.6017,
+      "step": 10096
+    },
+    {
+      "epoch": 2.6804725872826234,
+      "grad_norm": 0.435635051963991,
+      "learning_rate": 2.9345406670283115e-06,
+      "loss": 0.6104,
+      "step": 10097
+    },
+    {
+      "epoch": 2.680738085756007,
+      "grad_norm": 0.45893619179706113,
+      "learning_rate": 2.9341968492649397e-06,
+      "loss": 0.5648,
+      "step": 10098
+    },
+    {
+      "epoch": 2.6810035842293907,
+      "grad_norm": 0.43116329511951146,
+      "learning_rate": 2.933853023033667e-06,
+      "loss": 0.5956,
+      "step": 10099
+    },
+    {
+      "epoch": 2.6812690827027743,
+      "grad_norm": 0.4406655863781548,
+      "learning_rate": 2.9335091883411982e-06,
+      "loss": 0.5694,
+      "step": 10100
+    },
+    {
+      "epoch": 2.6815345811761584,
+      "grad_norm": 0.4468323208880422,
+      "learning_rate": 2.9331653451942387e-06,
+      "loss": 0.5938,
+      "step": 10101
+    },
+    {
+      "epoch": 2.681800079649542,
+      "grad_norm": 0.4427978559765397,
+      "learning_rate": 2.9328214935994944e-06,
+      "loss": 0.5688,
+      "step": 10102
+    },
+    {
+      "epoch": 2.6820655781229257,
+      "grad_norm": 0.44571194839184297,
+      "learning_rate": 2.9324776335636707e-06,
+      "loss": 0.5708,
+      "step": 10103
+    },
+    {
+      "epoch": 2.6823310765963093,
+      "grad_norm": 0.44819799115975945,
+      "learning_rate": 2.9321337650934746e-06,
+      "loss": 0.5419,
+      "step": 10104
+    },
+    {
+      "epoch": 2.6825965750696934,
+      "grad_norm": 0.43864071128824067,
+      "learning_rate": 2.9317898881956123e-06,
+      "loss": 0.5473,
+      "step": 10105
+    },
+    {
+      "epoch": 2.682862073543077,
+      "grad_norm": 0.44819354612458345,
+      "learning_rate": 2.93144600287679e-06,
+      "loss": 0.5489,
+      "step": 10106
+    },
+    {
+      "epoch": 2.6831275720164607,
+      "grad_norm": 0.43484419959974174,
+      "learning_rate": 2.9311021091437137e-06,
+      "loss": 0.5816,
+      "step": 10107
+    },
+    {
+      "epoch": 2.6833930704898448,
+      "grad_norm": 0.42980750784080285,
+      "learning_rate": 2.9307582070030918e-06,
+      "loss": 0.5638,
+      "step": 10108
+    },
+    {
+      "epoch": 2.6836585689632284,
+      "grad_norm": 0.44690628571829843,
+      "learning_rate": 2.930414296461629e-06,
+      "loss": 0.5301,
+      "step": 10109
+    },
+    {
+      "epoch": 2.683924067436612,
+      "grad_norm": 0.43977425815622667,
+      "learning_rate": 2.9300703775260343e-06,
+      "loss": 0.5495,
+      "step": 10110
+    },
+    {
+      "epoch": 2.684189565909996,
+      "grad_norm": 0.43808437722194576,
+      "learning_rate": 2.9297264502030144e-06,
+      "loss": 0.5323,
+      "step": 10111
+    },
+    {
+      "epoch": 2.6844550643833798,
+      "grad_norm": 0.42324486002843337,
+      "learning_rate": 2.929382514499276e-06,
+      "loss": 0.5259,
+      "step": 10112
+    },
+    {
+      "epoch": 2.6847205628567634,
+      "grad_norm": 0.43050158565945634,
+      "learning_rate": 2.929038570421526e-06,
+      "loss": 0.5949,
+      "step": 10113
+    },
+    {
+      "epoch": 2.6849860613301475,
+      "grad_norm": 0.4549232931473547,
+      "learning_rate": 2.928694617976476e-06,
+      "loss": 0.5611,
+      "step": 10114
+    },
+    {
+      "epoch": 2.685251559803531,
+      "grad_norm": 0.45372931034691066,
+      "learning_rate": 2.92835065717083e-06,
+      "loss": 0.5259,
+      "step": 10115
+    },
+    {
+      "epoch": 2.685517058276915,
+      "grad_norm": 0.4276089160597099,
+      "learning_rate": 2.928006688011297e-06,
+      "loss": 0.5267,
+      "step": 10116
+    },
+    {
+      "epoch": 2.685782556750299,
+      "grad_norm": 0.45178187627395144,
+      "learning_rate": 2.927662710504587e-06,
+      "loss": 0.5887,
+      "step": 10117
+    },
+    {
+      "epoch": 2.6860480552236825,
+      "grad_norm": 0.4605202743122104,
+      "learning_rate": 2.9273187246574065e-06,
+      "loss": 0.5588,
+      "step": 10118
+    },
+    {
+      "epoch": 2.686313553697066,
+      "grad_norm": 0.45633522413441485,
+      "learning_rate": 2.9269747304764638e-06,
+      "loss": 0.6033,
+      "step": 10119
+    },
+    {
+      "epoch": 2.6865790521704502,
+      "grad_norm": 0.44421248833581395,
+      "learning_rate": 2.9266307279684698e-06,
+      "loss": 0.5737,
+      "step": 10120
+    },
+    {
+      "epoch": 2.686844550643834,
+      "grad_norm": 0.43618977261361214,
+      "learning_rate": 2.926286717140131e-06,
+      "loss": 0.5643,
+      "step": 10121
+    },
+    {
+      "epoch": 2.6871100491172175,
+      "grad_norm": 0.45493469690811333,
+      "learning_rate": 2.9259426979981576e-06,
+      "loss": 0.5985,
+      "step": 10122
+    },
+    {
+      "epoch": 2.6873755475906016,
+      "grad_norm": 0.4430360324782482,
+      "learning_rate": 2.9255986705492586e-06,
+      "loss": 0.5825,
+      "step": 10123
+    },
+    {
+      "epoch": 2.6876410460639852,
+      "grad_norm": 0.46115456136334293,
+      "learning_rate": 2.925254634800144e-06,
+      "loss": 0.533,
+      "step": 10124
+    },
+    {
+      "epoch": 2.687906544537369,
+      "grad_norm": 0.4569653467832881,
+      "learning_rate": 2.9249105907575226e-06,
+      "loss": 0.5621,
+      "step": 10125
+    },
+    {
+      "epoch": 2.688172043010753,
+      "grad_norm": 0.44302748387267804,
+      "learning_rate": 2.9245665384281043e-06,
+      "loss": 0.5374,
+      "step": 10126
+    },
+    {
+      "epoch": 2.6884375414841366,
+      "grad_norm": 0.435841955460323,
+      "learning_rate": 2.9242224778185985e-06,
+      "loss": 0.5525,
+      "step": 10127
+    },
+    {
+      "epoch": 2.6887030399575202,
+      "grad_norm": 0.4392626183352283,
+      "learning_rate": 2.9238784089357166e-06,
+      "loss": 0.5554,
+      "step": 10128
+    },
+    {
+      "epoch": 2.688968538430904,
+      "grad_norm": 0.4424011386025097,
+      "learning_rate": 2.9235343317861675e-06,
+      "loss": 0.5617,
+      "step": 10129
+    },
+    {
+      "epoch": 2.6892340369042875,
+      "grad_norm": 0.4411197501777723,
+      "learning_rate": 2.9231902463766614e-06,
+      "loss": 0.5555,
+      "step": 10130
+    },
+    {
+      "epoch": 2.6894995353776716,
+      "grad_norm": 0.444098152256741,
+      "learning_rate": 2.9228461527139098e-06,
+      "loss": 0.537,
+      "step": 10131
+    },
+    {
+      "epoch": 2.6897650338510553,
+      "grad_norm": 0.4445601872178917,
+      "learning_rate": 2.9225020508046233e-06,
+      "loss": 0.5836,
+      "step": 10132
+    },
+    {
+      "epoch": 2.690030532324439,
+      "grad_norm": 0.4405321199538382,
+      "learning_rate": 2.9221579406555112e-06,
+      "loss": 0.5753,
+      "step": 10133
+    },
+    {
+      "epoch": 2.690296030797823,
+      "grad_norm": 0.43234118829709495,
+      "learning_rate": 2.9218138222732872e-06,
+      "loss": 0.5522,
+      "step": 10134
+    },
+    {
+      "epoch": 2.6905615292712066,
+      "grad_norm": 0.4608281057479714,
+      "learning_rate": 2.9214696956646597e-06,
+      "loss": 0.5735,
+      "step": 10135
+    },
+    {
+      "epoch": 2.6908270277445903,
+      "grad_norm": 0.4321020388490961,
+      "learning_rate": 2.9211255608363415e-06,
+      "loss": 0.5323,
+      "step": 10136
+    },
+    {
+      "epoch": 2.6910925262179743,
+      "grad_norm": 0.44074106921283657,
+      "learning_rate": 2.9207814177950445e-06,
+      "loss": 0.5682,
+      "step": 10137
+    },
+    {
+      "epoch": 2.691358024691358,
+      "grad_norm": 0.44815829488913456,
+      "learning_rate": 2.920437266547479e-06,
+      "loss": 0.5923,
+      "step": 10138
+    },
+    {
+      "epoch": 2.6916235231647416,
+      "grad_norm": 0.45861230577868295,
+      "learning_rate": 2.9200931071003568e-06,
+      "loss": 0.5733,
+      "step": 10139
+    },
+    {
+      "epoch": 2.6918890216381257,
+      "grad_norm": 0.4534081695307583,
+      "learning_rate": 2.919748939460391e-06,
+      "loss": 0.6176,
+      "step": 10140
+    },
+    {
+      "epoch": 2.6921545201115094,
+      "grad_norm": 0.43691484881192744,
+      "learning_rate": 2.919404763634293e-06,
+      "loss": 0.5898,
+      "step": 10141
+    },
+    {
+      "epoch": 2.692420018584893,
+      "grad_norm": 0.42518703142546543,
+      "learning_rate": 2.919060579628776e-06,
+      "loss": 0.5041,
+      "step": 10142
+    },
+    {
+      "epoch": 2.692685517058277,
+      "grad_norm": 0.42544922469166396,
+      "learning_rate": 2.9187163874505507e-06,
+      "loss": 0.5289,
+      "step": 10143
+    },
+    {
+      "epoch": 2.6929510155316607,
+      "grad_norm": 0.44754361744708193,
+      "learning_rate": 2.9183721871063313e-06,
+      "loss": 0.5961,
+      "step": 10144
+    },
+    {
+      "epoch": 2.6932165140050444,
+      "grad_norm": 0.45116552148572175,
+      "learning_rate": 2.9180279786028294e-06,
+      "loss": 0.562,
+      "step": 10145
+    },
+    {
+      "epoch": 2.6934820124784284,
+      "grad_norm": 0.43796753904122016,
+      "learning_rate": 2.917683761946759e-06,
+      "loss": 0.5667,
+      "step": 10146
+    },
+    {
+      "epoch": 2.693747510951812,
+      "grad_norm": 0.441296744345199,
+      "learning_rate": 2.9173395371448322e-06,
+      "loss": 0.5666,
+      "step": 10147
+    },
+    {
+      "epoch": 2.6940130094251957,
+      "grad_norm": 0.43889977861376195,
+      "learning_rate": 2.9169953042037623e-06,
+      "loss": 0.5532,
+      "step": 10148
+    },
+    {
+      "epoch": 2.69427850789858,
+      "grad_norm": 0.43204457008165237,
+      "learning_rate": 2.9166510631302635e-06,
+      "loss": 0.5665,
+      "step": 10149
+    },
+    {
+      "epoch": 2.6945440063719635,
+      "grad_norm": 0.4287091008325839,
+      "learning_rate": 2.9163068139310484e-06,
+      "loss": 0.5744,
+      "step": 10150
+    },
+    {
+      "epoch": 2.694809504845347,
+      "grad_norm": 0.4439695107237071,
+      "learning_rate": 2.915962556612832e-06,
+      "loss": 0.5741,
+      "step": 10151
+    },
+    {
+      "epoch": 2.695075003318731,
+      "grad_norm": 0.4366243909985819,
+      "learning_rate": 2.915618291182326e-06,
+      "loss": 0.5695,
+      "step": 10152
+    },
+    {
+      "epoch": 2.695340501792115,
+      "grad_norm": 0.44477894506137927,
+      "learning_rate": 2.9152740176462463e-06,
+      "loss": 0.5965,
+      "step": 10153
+    },
+    {
+      "epoch": 2.6956060002654985,
+      "grad_norm": 0.4487188796862963,
+      "learning_rate": 2.914929736011307e-06,
+      "loss": 0.5541,
+      "step": 10154
+    },
+    {
+      "epoch": 2.695871498738882,
+      "grad_norm": 0.4437546711296801,
+      "learning_rate": 2.9145854462842214e-06,
+      "loss": 0.5646,
+      "step": 10155
+    },
+    {
+      "epoch": 2.696136997212266,
+      "grad_norm": 0.43036566495529827,
+      "learning_rate": 2.9142411484717048e-06,
+      "loss": 0.5758,
+      "step": 10156
+    },
+    {
+      "epoch": 2.69640249568565,
+      "grad_norm": 0.4365573923260459,
+      "learning_rate": 2.9138968425804716e-06,
+      "loss": 0.5576,
+      "step": 10157
+    },
+    {
+      "epoch": 2.6966679941590335,
+      "grad_norm": 0.44101860474461374,
+      "learning_rate": 2.9135525286172356e-06,
+      "loss": 0.5852,
+      "step": 10158
+    },
+    {
+      "epoch": 2.696933492632417,
+      "grad_norm": 0.4453933256532241,
+      "learning_rate": 2.9132082065887135e-06,
+      "loss": 0.566,
+      "step": 10159
+    },
+    {
+      "epoch": 2.697198991105801,
+      "grad_norm": 0.4434611364850128,
+      "learning_rate": 2.91286387650162e-06,
+      "loss": 0.604,
+      "step": 10160
+    },
+    {
+      "epoch": 2.697464489579185,
+      "grad_norm": 0.4392404423358599,
+      "learning_rate": 2.912519538362669e-06,
+      "loss": 0.5447,
+      "step": 10161
+    },
+    {
+      "epoch": 2.6977299880525685,
+      "grad_norm": 0.43303605357684627,
+      "learning_rate": 2.912175192178578e-06,
+      "loss": 0.5556,
+      "step": 10162
+    },
+    {
+      "epoch": 2.6979954865259526,
+      "grad_norm": 0.45837599748872915,
+      "learning_rate": 2.9118308379560615e-06,
+      "loss": 0.6133,
+      "step": 10163
+    },
+    {
+      "epoch": 2.698260984999336,
+      "grad_norm": 0.4526662166515674,
+      "learning_rate": 2.911486475701835e-06,
+      "loss": 0.5314,
+      "step": 10164
+    },
+    {
+      "epoch": 2.69852648347272,
+      "grad_norm": 0.43883768370901466,
+      "learning_rate": 2.9111421054226145e-06,
+      "loss": 0.5388,
+      "step": 10165
+    },
+    {
+      "epoch": 2.698791981946104,
+      "grad_norm": 0.44445997235757106,
+      "learning_rate": 2.9107977271251174e-06,
+      "loss": 0.548,
+      "step": 10166
+    },
+    {
+      "epoch": 2.6990574804194876,
+      "grad_norm": 0.4550284157712892,
+      "learning_rate": 2.910453340816058e-06,
+      "loss": 0.5913,
+      "step": 10167
+    },
+    {
+      "epoch": 2.699322978892871,
+      "grad_norm": 0.43658869262833233,
+      "learning_rate": 2.9101089465021533e-06,
+      "loss": 0.6039,
+      "step": 10168
+    },
+    {
+      "epoch": 2.6995884773662553,
+      "grad_norm": 0.44586788070347355,
+      "learning_rate": 2.90976454419012e-06,
+      "loss": 0.5529,
+      "step": 10169
+    },
+    {
+      "epoch": 2.699853975839639,
+      "grad_norm": 0.4266545587764369,
+      "learning_rate": 2.909420133886675e-06,
+      "loss": 0.5516,
+      "step": 10170
+    },
+    {
+      "epoch": 2.7001194743130226,
+      "grad_norm": 0.4431055348578981,
+      "learning_rate": 2.909075715598535e-06,
+      "loss": 0.5655,
+      "step": 10171
+    },
+    {
+      "epoch": 2.7003849727864067,
+      "grad_norm": 0.4439226564662244,
+      "learning_rate": 2.908731289332417e-06,
+      "loss": 0.5804,
+      "step": 10172
+    },
+    {
+      "epoch": 2.7006504712597903,
+      "grad_norm": 0.43741596479613637,
+      "learning_rate": 2.9083868550950378e-06,
+      "loss": 0.5411,
+      "step": 10173
+    },
+    {
+      "epoch": 2.700915969733174,
+      "grad_norm": 0.4567293071196986,
+      "learning_rate": 2.9080424128931163e-06,
+      "loss": 0.579,
+      "step": 10174
+    },
+    {
+      "epoch": 2.701181468206558,
+      "grad_norm": 0.43807146698167687,
+      "learning_rate": 2.907697962733367e-06,
+      "loss": 0.5478,
+      "step": 10175
+    },
+    {
+      "epoch": 2.7014469666799417,
+      "grad_norm": 0.4442205453998902,
+      "learning_rate": 2.9073535046225097e-06,
+      "loss": 0.5779,
+      "step": 10176
+    },
+    {
+      "epoch": 2.7017124651533253,
+      "grad_norm": 0.43460845056961783,
+      "learning_rate": 2.9070090385672625e-06,
+      "loss": 0.551,
+      "step": 10177
+    },
+    {
+      "epoch": 2.7019779636267094,
+      "grad_norm": 0.43089441786303767,
+      "learning_rate": 2.906664564574341e-06,
+      "loss": 0.5607,
+      "step": 10178
+    },
+    {
+      "epoch": 2.702243462100093,
+      "grad_norm": 0.43806138550830276,
+      "learning_rate": 2.9063200826504656e-06,
+      "loss": 0.5474,
+      "step": 10179
+    },
+    {
+      "epoch": 2.7025089605734767,
+      "grad_norm": 0.4307302636907822,
+      "learning_rate": 2.9059755928023547e-06,
+      "loss": 0.5746,
+      "step": 10180
+    },
+    {
+      "epoch": 2.7027744590468608,
+      "grad_norm": 0.4362951242567226,
+      "learning_rate": 2.9056310950367244e-06,
+      "loss": 0.532,
+      "step": 10181
+    },
+    {
+      "epoch": 2.7030399575202444,
+      "grad_norm": 0.4335504550219663,
+      "learning_rate": 2.905286589360295e-06,
+      "loss": 0.6246,
+      "step": 10182
+    },
+    {
+      "epoch": 2.703305455993628,
+      "grad_norm": 0.44551132411467825,
+      "learning_rate": 2.904942075779785e-06,
+      "loss": 0.5836,
+      "step": 10183
+    },
+    {
+      "epoch": 2.7035709544670117,
+      "grad_norm": 0.43076782821831744,
+      "learning_rate": 2.9045975543019134e-06,
+      "loss": 0.5446,
+      "step": 10184
+    },
+    {
+      "epoch": 2.7038364529403953,
+      "grad_norm": 0.4486741657587676,
+      "learning_rate": 2.904253024933398e-06,
+      "loss": 0.5476,
+      "step": 10185
+    },
+    {
+      "epoch": 2.7041019514137794,
+      "grad_norm": 0.4342935319116983,
+      "learning_rate": 2.9039084876809593e-06,
+      "loss": 0.5816,
+      "step": 10186
+    },
+    {
+      "epoch": 2.704367449887163,
+      "grad_norm": 0.4394810326791022,
+      "learning_rate": 2.903563942551316e-06,
+      "loss": 0.5811,
+      "step": 10187
+    },
+    {
+      "epoch": 2.7046329483605467,
+      "grad_norm": 0.43517064720426085,
+      "learning_rate": 2.903219389551187e-06,
+      "loss": 0.5471,
+      "step": 10188
+    },
+    {
+      "epoch": 2.7048984468339308,
+      "grad_norm": 0.4428454491543437,
+      "learning_rate": 2.9028748286872943e-06,
+      "loss": 0.5293,
+      "step": 10189
+    },
+    {
+      "epoch": 2.7051639453073144,
+      "grad_norm": 0.453359727330235,
+      "learning_rate": 2.9025302599663552e-06,
+      "loss": 0.5859,
+      "step": 10190
+    },
+    {
+      "epoch": 2.705429443780698,
+      "grad_norm": 0.4247083903059448,
+      "learning_rate": 2.9021856833950902e-06,
+      "loss": 0.5217,
+      "step": 10191
+    },
+    {
+      "epoch": 2.705694942254082,
+      "grad_norm": 0.43505388503812364,
+      "learning_rate": 2.9018410989802205e-06,
+      "loss": 0.5584,
+      "step": 10192
+    },
+    {
+      "epoch": 2.7059604407274658,
+      "grad_norm": 0.44285716758921784,
+      "learning_rate": 2.9014965067284643e-06,
+      "loss": 0.5638,
+      "step": 10193
+    },
+    {
+      "epoch": 2.7062259392008494,
+      "grad_norm": 0.45040368042488255,
+      "learning_rate": 2.9011519066465437e-06,
+      "loss": 0.554,
+      "step": 10194
+    },
+    {
+      "epoch": 2.7064914376742335,
+      "grad_norm": 0.44506375560710565,
+      "learning_rate": 2.900807298741179e-06,
+      "loss": 0.5979,
+      "step": 10195
+    },
+    {
+      "epoch": 2.706756936147617,
+      "grad_norm": 0.4400339333408414,
+      "learning_rate": 2.9004626830190903e-06,
+      "loss": 0.5794,
+      "step": 10196
+    },
+    {
+      "epoch": 2.707022434621001,
+      "grad_norm": 0.4459688244165978,
+      "learning_rate": 2.9001180594869994e-06,
+      "loss": 0.5701,
+      "step": 10197
+    },
+    {
+      "epoch": 2.707287933094385,
+      "grad_norm": 0.4436688354506341,
+      "learning_rate": 2.899773428151626e-06,
+      "loss": 0.5351,
+      "step": 10198
+    },
+    {
+      "epoch": 2.7075534315677685,
+      "grad_norm": 0.448335991666371,
+      "learning_rate": 2.8994287890196917e-06,
+      "loss": 0.612,
+      "step": 10199
+    },
+    {
+      "epoch": 2.707818930041152,
+      "grad_norm": 0.4385498421908971,
+      "learning_rate": 2.899084142097919e-06,
+      "loss": 0.5289,
+      "step": 10200
+    },
+    {
+      "epoch": 2.7080844285145362,
+      "grad_norm": 0.43042540619984915,
+      "learning_rate": 2.8987394873930276e-06,
+      "loss": 0.5506,
+      "step": 10201
+    },
+    {
+      "epoch": 2.70834992698792,
+      "grad_norm": 0.46417012457458945,
+      "learning_rate": 2.89839482491174e-06,
+      "loss": 0.5336,
+      "step": 10202
+    },
+    {
+      "epoch": 2.7086154254613035,
+      "grad_norm": 0.4321025152650331,
+      "learning_rate": 2.8980501546607785e-06,
+      "loss": 0.5557,
+      "step": 10203
+    },
+    {
+      "epoch": 2.7088809239346876,
+      "grad_norm": 0.4537749218447374,
+      "learning_rate": 2.897705476646863e-06,
+      "loss": 0.5487,
+      "step": 10204
+    },
+    {
+      "epoch": 2.7091464224080712,
+      "grad_norm": 0.4360817293315029,
+      "learning_rate": 2.8973607908767183e-06,
+      "loss": 0.5596,
+      "step": 10205
+    },
+    {
+      "epoch": 2.709411920881455,
+      "grad_norm": 0.4546763306757774,
+      "learning_rate": 2.897016097357065e-06,
+      "loss": 0.5776,
+      "step": 10206
+    },
+    {
+      "epoch": 2.709677419354839,
+      "grad_norm": 0.45476443627915464,
+      "learning_rate": 2.8966713960946256e-06,
+      "loss": 0.5843,
+      "step": 10207
+    },
+    {
+      "epoch": 2.7099429178282226,
+      "grad_norm": 0.4315199299370785,
+      "learning_rate": 2.8963266870961226e-06,
+      "loss": 0.5495,
+      "step": 10208
+    },
+    {
+      "epoch": 2.7102084163016062,
+      "grad_norm": 0.43459150804360874,
+      "learning_rate": 2.8959819703682797e-06,
+      "loss": 0.5439,
+      "step": 10209
+    },
+    {
+      "epoch": 2.71047391477499,
+      "grad_norm": 0.4412231899526384,
+      "learning_rate": 2.895637245917818e-06,
+      "loss": 0.5606,
+      "step": 10210
+    },
+    {
+      "epoch": 2.710739413248374,
+      "grad_norm": 0.4406979847113424,
+      "learning_rate": 2.895292513751461e-06,
+      "loss": 0.5633,
+      "step": 10211
+    },
+    {
+      "epoch": 2.7110049117217576,
+      "grad_norm": 0.43296962620667295,
+      "learning_rate": 2.8949477738759337e-06,
+      "loss": 0.5585,
+      "step": 10212
+    },
+    {
+      "epoch": 2.7112704101951413,
+      "grad_norm": 0.4501251361074169,
+      "learning_rate": 2.8946030262979565e-06,
+      "loss": 0.5629,
+      "step": 10213
+    },
+    {
+      "epoch": 2.711535908668525,
+      "grad_norm": 0.44126313119357236,
+      "learning_rate": 2.8942582710242534e-06,
+      "loss": 0.5576,
+      "step": 10214
+    },
+    {
+      "epoch": 2.711801407141909,
+      "grad_norm": 0.42749229612120254,
+      "learning_rate": 2.89391350806155e-06,
+      "loss": 0.5633,
+      "step": 10215
+    },
+    {
+      "epoch": 2.7120669056152926,
+      "grad_norm": 0.45957870531395834,
+      "learning_rate": 2.8935687374165684e-06,
+      "loss": 0.5579,
+      "step": 10216
+    },
+    {
+      "epoch": 2.7123324040886763,
+      "grad_norm": 0.4360893174722151,
+      "learning_rate": 2.8932239590960327e-06,
+      "loss": 0.5675,
+      "step": 10217
+    },
+    {
+      "epoch": 2.7125979025620603,
+      "grad_norm": 0.45370050701947046,
+      "learning_rate": 2.8928791731066673e-06,
+      "loss": 0.5757,
+      "step": 10218
+    },
+    {
+      "epoch": 2.712863401035444,
+      "grad_norm": 0.4403385448272651,
+      "learning_rate": 2.8925343794551956e-06,
+      "loss": 0.5485,
+      "step": 10219
+    },
+    {
+      "epoch": 2.7131288995088276,
+      "grad_norm": 0.4426351958102985,
+      "learning_rate": 2.8921895781483427e-06,
+      "loss": 0.5487,
+      "step": 10220
+    },
+    {
+      "epoch": 2.7133943979822117,
+      "grad_norm": 0.4391844449722344,
+      "learning_rate": 2.891844769192833e-06,
+      "loss": 0.5264,
+      "step": 10221
+    },
+    {
+      "epoch": 2.7136598964555954,
+      "grad_norm": 0.4304856481399934,
+      "learning_rate": 2.891499952595391e-06,
+      "loss": 0.5796,
+      "step": 10222
+    },
+    {
+      "epoch": 2.713925394928979,
+      "grad_norm": 0.4234577013632456,
+      "learning_rate": 2.891155128362741e-06,
+      "loss": 0.566,
+      "step": 10223
+    },
+    {
+      "epoch": 2.714190893402363,
+      "grad_norm": 0.4636773893498481,
+      "learning_rate": 2.890810296501608e-06,
+      "loss": 0.5482,
+      "step": 10224
+    },
+    {
+      "epoch": 2.7144563918757467,
+      "grad_norm": 0.4489502189194947,
+      "learning_rate": 2.8904654570187174e-06,
+      "loss": 0.5718,
+      "step": 10225
+    },
+    {
+      "epoch": 2.7147218903491304,
+      "grad_norm": 0.4297856085653684,
+      "learning_rate": 2.8901206099207953e-06,
+      "loss": 0.5713,
+      "step": 10226
+    },
+    {
+      "epoch": 2.7149873888225144,
+      "grad_norm": 0.4490390343748039,
+      "learning_rate": 2.889775755214565e-06,
+      "loss": 0.5253,
+      "step": 10227
+    },
+    {
+      "epoch": 2.715252887295898,
+      "grad_norm": 0.4432906699961794,
+      "learning_rate": 2.889430892906754e-06,
+      "loss": 0.5521,
+      "step": 10228
+    },
+    {
+      "epoch": 2.7155183857692817,
+      "grad_norm": 0.4269082452158409,
+      "learning_rate": 2.8890860230040867e-06,
+      "loss": 0.533,
+      "step": 10229
+    },
+    {
+      "epoch": 2.715783884242666,
+      "grad_norm": 0.45730782926041164,
+      "learning_rate": 2.8887411455132893e-06,
+      "loss": 0.5937,
+      "step": 10230
+    },
+    {
+      "epoch": 2.7160493827160495,
+      "grad_norm": 0.4333852814691258,
+      "learning_rate": 2.8883962604410875e-06,
+      "loss": 0.5924,
+      "step": 10231
+    },
+    {
+      "epoch": 2.716314881189433,
+      "grad_norm": 0.44351573462855143,
+      "learning_rate": 2.888051367794208e-06,
+      "loss": 0.5515,
+      "step": 10232
+    },
+    {
+      "epoch": 2.716580379662817,
+      "grad_norm": 0.4252288714134628,
+      "learning_rate": 2.8877064675793766e-06,
+      "loss": 0.5139,
+      "step": 10233
+    },
+    {
+      "epoch": 2.716845878136201,
+      "grad_norm": 0.44673288305101033,
+      "learning_rate": 2.8873615598033194e-06,
+      "loss": 0.5586,
+      "step": 10234
+    },
+    {
+      "epoch": 2.7171113766095845,
+      "grad_norm": 0.4285174552763303,
+      "learning_rate": 2.887016644472764e-06,
+      "loss": 0.5898,
+      "step": 10235
+    },
+    {
+      "epoch": 2.7173768750829685,
+      "grad_norm": 0.45347068066869795,
+      "learning_rate": 2.8866717215944363e-06,
+      "loss": 0.5461,
+      "step": 10236
+    },
+    {
+      "epoch": 2.717642373556352,
+      "grad_norm": 0.4440321627109502,
+      "learning_rate": 2.886326791175063e-06,
+      "loss": 0.5788,
+      "step": 10237
+    },
+    {
+      "epoch": 2.717907872029736,
+      "grad_norm": 0.4193244441322592,
+      "learning_rate": 2.885981853221372e-06,
+      "loss": 0.5083,
+      "step": 10238
+    },
+    {
+      "epoch": 2.7181733705031195,
+      "grad_norm": 0.42589757698017205,
+      "learning_rate": 2.8856369077400896e-06,
+      "loss": 0.5295,
+      "step": 10239
+    },
+    {
+      "epoch": 2.718438868976503,
+      "grad_norm": 0.43029285291308367,
+      "learning_rate": 2.885291954737943e-06,
+      "loss": 0.5315,
+      "step": 10240
+    },
+    {
+      "epoch": 2.718704367449887,
+      "grad_norm": 0.4512118297488647,
+      "learning_rate": 2.88494699422166e-06,
+      "loss": 0.5388,
+      "step": 10241
+    },
+    {
+      "epoch": 2.718969865923271,
+      "grad_norm": 0.43228604883114413,
+      "learning_rate": 2.884602026197968e-06,
+      "loss": 0.5583,
+      "step": 10242
+    },
+    {
+      "epoch": 2.7192353643966545,
+      "grad_norm": 0.4575615646582,
+      "learning_rate": 2.8842570506735952e-06,
+      "loss": 0.5681,
+      "step": 10243
+    },
+    {
+      "epoch": 2.7195008628700386,
+      "grad_norm": 0.4592349652920735,
+      "learning_rate": 2.883912067655269e-06,
+      "loss": 0.5741,
+      "step": 10244
+    },
+    {
+      "epoch": 2.719766361343422,
+      "grad_norm": 0.4269147614749827,
+      "learning_rate": 2.8835670771497174e-06,
+      "loss": 0.5757,
+      "step": 10245
+    },
+    {
+      "epoch": 2.720031859816806,
+      "grad_norm": 0.4574142720896556,
+      "learning_rate": 2.8832220791636687e-06,
+      "loss": 0.524,
+      "step": 10246
+    },
+    {
+      "epoch": 2.72029735829019,
+      "grad_norm": 0.44551944203516236,
+      "learning_rate": 2.8828770737038515e-06,
+      "loss": 0.5829,
+      "step": 10247
+    },
+    {
+      "epoch": 2.7205628567635736,
+      "grad_norm": 0.4442857150696108,
+      "learning_rate": 2.8825320607769937e-06,
+      "loss": 0.6094,
+      "step": 10248
+    },
+    {
+      "epoch": 2.720828355236957,
+      "grad_norm": 0.4556480864316161,
+      "learning_rate": 2.8821870403898233e-06,
+      "loss": 0.5496,
+      "step": 10249
+    },
+    {
+      "epoch": 2.7210938537103413,
+      "grad_norm": 0.44113528389842227,
+      "learning_rate": 2.8818420125490705e-06,
+      "loss": 0.5764,
+      "step": 10250
+    },
+    {
+      "epoch": 2.721359352183725,
+      "grad_norm": 0.4622985401187027,
+      "learning_rate": 2.881496977261463e-06,
+      "loss": 0.5835,
+      "step": 10251
+    },
+    {
+      "epoch": 2.7216248506571086,
+      "grad_norm": 0.4486770539443488,
+      "learning_rate": 2.881151934533732e-06,
+      "loss": 0.563,
+      "step": 10252
+    },
+    {
+      "epoch": 2.7218903491304927,
+      "grad_norm": 0.438055948866357,
+      "learning_rate": 2.8808068843726034e-06,
+      "loss": 0.5887,
+      "step": 10253
+    },
+    {
+      "epoch": 2.7221558476038763,
+      "grad_norm": 0.4541672053501981,
+      "learning_rate": 2.880461826784809e-06,
+      "loss": 0.5817,
+      "step": 10254
+    },
+    {
+      "epoch": 2.72242134607726,
+      "grad_norm": 0.4451859760813674,
+      "learning_rate": 2.8801167617770776e-06,
+      "loss": 0.5459,
+      "step": 10255
+    },
+    {
+      "epoch": 2.722686844550644,
+      "grad_norm": 0.44037258932811785,
+      "learning_rate": 2.879771689356138e-06,
+      "loss": 0.5893,
+      "step": 10256
+    },
+    {
+      "epoch": 2.7229523430240277,
+      "grad_norm": 0.4353396261090973,
+      "learning_rate": 2.879426609528721e-06,
+      "loss": 0.5315,
+      "step": 10257
+    },
+    {
+      "epoch": 2.7232178414974113,
+      "grad_norm": 0.4454624505272185,
+      "learning_rate": 2.879081522301556e-06,
+      "loss": 0.5844,
+      "step": 10258
+    },
+    {
+      "epoch": 2.7234833399707954,
+      "grad_norm": 0.4517189569398687,
+      "learning_rate": 2.8787364276813734e-06,
+      "loss": 0.6006,
+      "step": 10259
+    },
+    {
+      "epoch": 2.723748838444179,
+      "grad_norm": 0.43210777897024255,
+      "learning_rate": 2.878391325674902e-06,
+      "loss": 0.5587,
+      "step": 10260
+    },
+    {
+      "epoch": 2.7240143369175627,
+      "grad_norm": 0.45917523874992766,
+      "learning_rate": 2.8780462162888752e-06,
+      "loss": 0.5582,
+      "step": 10261
+    },
+    {
+      "epoch": 2.7242798353909468,
+      "grad_norm": 0.45243517432940794,
+      "learning_rate": 2.87770109953002e-06,
+      "loss": 0.598,
+      "step": 10262
+    },
+    {
+      "epoch": 2.7245453338643304,
+      "grad_norm": 0.4439983830860151,
+      "learning_rate": 2.8773559754050688e-06,
+      "loss": 0.5735,
+      "step": 10263
+    },
+    {
+      "epoch": 2.724810832337714,
+      "grad_norm": 0.433536528137935,
+      "learning_rate": 2.8770108439207534e-06,
+      "loss": 0.5274,
+      "step": 10264
+    },
+    {
+      "epoch": 2.7250763308110977,
+      "grad_norm": 0.4420608250394168,
+      "learning_rate": 2.876665705083802e-06,
+      "loss": 0.5321,
+      "step": 10265
+    },
+    {
+      "epoch": 2.7253418292844818,
+      "grad_norm": 0.447493197408292,
+      "learning_rate": 2.876320558900948e-06,
+      "loss": 0.5407,
+      "step": 10266
+    },
+    {
+      "epoch": 2.7256073277578654,
+      "grad_norm": 0.44197996759121094,
+      "learning_rate": 2.875975405378922e-06,
+      "loss": 0.5383,
+      "step": 10267
+    },
+    {
+      "epoch": 2.725872826231249,
+      "grad_norm": 0.43150406968870186,
+      "learning_rate": 2.8756302445244536e-06,
+      "loss": 0.5662,
+      "step": 10268
+    },
+    {
+      "epoch": 2.7261383247046327,
+      "grad_norm": 0.4435022757307573,
+      "learning_rate": 2.875285076344277e-06,
+      "loss": 0.5875,
+      "step": 10269
+    },
+    {
+      "epoch": 2.7264038231780168,
+      "grad_norm": 0.4630022875977805,
+      "learning_rate": 2.8749399008451217e-06,
+      "loss": 0.5283,
+      "step": 10270
+    },
+    {
+      "epoch": 2.7266693216514004,
+      "grad_norm": 0.4487880725450732,
+      "learning_rate": 2.874594718033721e-06,
+      "loss": 0.5665,
+      "step": 10271
+    },
+    {
+      "epoch": 2.726934820124784,
+      "grad_norm": 0.44693622658289794,
+      "learning_rate": 2.8742495279168057e-06,
+      "loss": 0.5402,
+      "step": 10272
+    },
+    {
+      "epoch": 2.727200318598168,
+      "grad_norm": 0.43927945919667544,
+      "learning_rate": 2.873904330501109e-06,
+      "loss": 0.5889,
+      "step": 10273
+    },
+    {
+      "epoch": 2.727465817071552,
+      "grad_norm": 0.43940486252357186,
+      "learning_rate": 2.8735591257933613e-06,
+      "loss": 0.5655,
+      "step": 10274
+    },
+    {
+      "epoch": 2.7277313155449354,
+      "grad_norm": 0.44068802549051656,
+      "learning_rate": 2.8732139138002973e-06,
+      "loss": 0.5537,
+      "step": 10275
+    },
+    {
+      "epoch": 2.7279968140183195,
+      "grad_norm": 0.43916667465849263,
+      "learning_rate": 2.872868694528647e-06,
+      "loss": 0.5255,
+      "step": 10276
+    },
+    {
+      "epoch": 2.728262312491703,
+      "grad_norm": 0.43608657115105287,
+      "learning_rate": 2.8725234679851448e-06,
+      "loss": 0.5922,
+      "step": 10277
+    },
+    {
+      "epoch": 2.728527810965087,
+      "grad_norm": 0.46293574563631634,
+      "learning_rate": 2.8721782341765223e-06,
+      "loss": 0.5445,
+      "step": 10278
+    },
+    {
+      "epoch": 2.728793309438471,
+      "grad_norm": 0.43345665379478404,
+      "learning_rate": 2.8718329931095136e-06,
+      "loss": 0.5415,
+      "step": 10279
+    },
+    {
+      "epoch": 2.7290588079118545,
+      "grad_norm": 0.43498407272427503,
+      "learning_rate": 2.8714877447908505e-06,
+      "loss": 0.5556,
+      "step": 10280
+    },
+    {
+      "epoch": 2.729324306385238,
+      "grad_norm": 0.4350387166346781,
+      "learning_rate": 2.871142489227268e-06,
+      "loss": 0.6031,
+      "step": 10281
+    },
+    {
+      "epoch": 2.7295898048586222,
+      "grad_norm": 0.44126109668365776,
+      "learning_rate": 2.8707972264254973e-06,
+      "loss": 0.5574,
+      "step": 10282
+    },
+    {
+      "epoch": 2.729855303332006,
+      "grad_norm": 0.45918024015676406,
+      "learning_rate": 2.8704519563922733e-06,
+      "loss": 0.5688,
+      "step": 10283
+    },
+    {
+      "epoch": 2.7301208018053895,
+      "grad_norm": 0.42868988141374076,
+      "learning_rate": 2.870106679134329e-06,
+      "loss": 0.5835,
+      "step": 10284
+    },
+    {
+      "epoch": 2.7303863002787736,
+      "grad_norm": 0.4535947590739218,
+      "learning_rate": 2.869761394658398e-06,
+      "loss": 0.5463,
+      "step": 10285
+    },
+    {
+      "epoch": 2.7306517987521572,
+      "grad_norm": 0.43678625128534565,
+      "learning_rate": 2.869416102971215e-06,
+      "loss": 0.5281,
+      "step": 10286
+    },
+    {
+      "epoch": 2.730917297225541,
+      "grad_norm": 0.4457301754832794,
+      "learning_rate": 2.869070804079513e-06,
+      "loss": 0.5591,
+      "step": 10287
+    },
+    {
+      "epoch": 2.731182795698925,
+      "grad_norm": 0.43669980721056834,
+      "learning_rate": 2.8687254979900263e-06,
+      "loss": 0.5901,
+      "step": 10288
+    },
+    {
+      "epoch": 2.7314482941723086,
+      "grad_norm": 0.4447984074771697,
+      "learning_rate": 2.86838018470949e-06,
+      "loss": 0.5813,
+      "step": 10289
+    },
+    {
+      "epoch": 2.7317137926456923,
+      "grad_norm": 0.44536873773597246,
+      "learning_rate": 2.8680348642446385e-06,
+      "loss": 0.5798,
+      "step": 10290
+    },
+    {
+      "epoch": 2.7319792911190763,
+      "grad_norm": 0.44187847174583095,
+      "learning_rate": 2.8676895366022064e-06,
+      "loss": 0.5468,
+      "step": 10291
+    },
+    {
+      "epoch": 2.73224478959246,
+      "grad_norm": 0.45147062982548897,
+      "learning_rate": 2.867344201788927e-06,
+      "loss": 0.5766,
+      "step": 10292
+    },
+    {
+      "epoch": 2.7325102880658436,
+      "grad_norm": 0.43411976725545276,
+      "learning_rate": 2.866998859811537e-06,
+      "loss": 0.5434,
+      "step": 10293
+    },
+    {
+      "epoch": 2.7327757865392273,
+      "grad_norm": 0.4243805492589512,
+      "learning_rate": 2.8666535106767706e-06,
+      "loss": 0.5543,
+      "step": 10294
+    },
+    {
+      "epoch": 2.7330412850126113,
+      "grad_norm": 0.4485995572979169,
+      "learning_rate": 2.8663081543913624e-06,
+      "loss": 0.56,
+      "step": 10295
+    },
+    {
+      "epoch": 2.733306783485995,
+      "grad_norm": 0.4609874392191662,
+      "learning_rate": 2.8659627909620487e-06,
+      "loss": 0.5844,
+      "step": 10296
+    },
+    {
+      "epoch": 2.7335722819593786,
+      "grad_norm": 0.4489747516785865,
+      "learning_rate": 2.8656174203955646e-06,
+      "loss": 0.5594,
+      "step": 10297
+    },
+    {
+      "epoch": 2.7338377804327623,
+      "grad_norm": 0.4285498008743859,
+      "learning_rate": 2.865272042698646e-06,
+      "loss": 0.5648,
+      "step": 10298
+    },
+    {
+      "epoch": 2.7341032789061464,
+      "grad_norm": 0.437820489377927,
+      "learning_rate": 2.864926657878028e-06,
+      "loss": 0.557,
+      "step": 10299
+    },
+    {
+      "epoch": 2.73436877737953,
+      "grad_norm": 0.4426475389305618,
+      "learning_rate": 2.864581265940446e-06,
+      "loss": 0.5898,
+      "step": 10300
+    },
+    {
+      "epoch": 2.7346342758529136,
+      "grad_norm": 0.43671278903532534,
+      "learning_rate": 2.8642358668926375e-06,
+      "loss": 0.587,
+      "step": 10301
+    },
+    {
+      "epoch": 2.7348997743262977,
+      "grad_norm": 0.4406011189568316,
+      "learning_rate": 2.8638904607413376e-06,
+      "loss": 0.5416,
+      "step": 10302
+    },
+    {
+      "epoch": 2.7351652727996814,
+      "grad_norm": 0.4400776011375223,
+      "learning_rate": 2.8635450474932826e-06,
+      "loss": 0.5575,
+      "step": 10303
+    },
+    {
+      "epoch": 2.735430771273065,
+      "grad_norm": 0.4380610064877206,
+      "learning_rate": 2.8631996271552095e-06,
+      "loss": 0.5725,
+      "step": 10304
+    },
+    {
+      "epoch": 2.735696269746449,
+      "grad_norm": 0.4367289116352328,
+      "learning_rate": 2.8628541997338534e-06,
+      "loss": 0.554,
+      "step": 10305
+    },
+    {
+      "epoch": 2.7359617682198327,
+      "grad_norm": 0.422361387702185,
+      "learning_rate": 2.8625087652359517e-06,
+      "loss": 0.5441,
+      "step": 10306
+    },
+    {
+      "epoch": 2.7362272666932164,
+      "grad_norm": 0.4308649023992815,
+      "learning_rate": 2.8621633236682433e-06,
+      "loss": 0.5378,
+      "step": 10307
+    },
+    {
+      "epoch": 2.7364927651666004,
+      "grad_norm": 0.43502383684920753,
+      "learning_rate": 2.861817875037462e-06,
+      "loss": 0.5962,
+      "step": 10308
+    },
+    {
+      "epoch": 2.736758263639984,
+      "grad_norm": 0.4450716419718206,
+      "learning_rate": 2.861472419350347e-06,
+      "loss": 0.597,
+      "step": 10309
+    },
+    {
+      "epoch": 2.7370237621133677,
+      "grad_norm": 0.4473263353340631,
+      "learning_rate": 2.8611269566136347e-06,
+      "loss": 0.6067,
+      "step": 10310
+    },
+    {
+      "epoch": 2.737289260586752,
+      "grad_norm": 0.4400704880917872,
+      "learning_rate": 2.860781486834062e-06,
+      "loss": 0.567,
+      "step": 10311
+    },
+    {
+      "epoch": 2.7375547590601355,
+      "grad_norm": 0.4552158124413648,
+      "learning_rate": 2.860436010018367e-06,
+      "loss": 0.5683,
+      "step": 10312
+    },
+    {
+      "epoch": 2.737820257533519,
+      "grad_norm": 0.4401579972712152,
+      "learning_rate": 2.860090526173288e-06,
+      "loss": 0.5725,
+      "step": 10313
+    },
+    {
+      "epoch": 2.738085756006903,
+      "grad_norm": 0.44081188723914233,
+      "learning_rate": 2.8597450353055615e-06,
+      "loss": 0.5623,
+      "step": 10314
+    },
+    {
+      "epoch": 2.738351254480287,
+      "grad_norm": 0.4294027187144235,
+      "learning_rate": 2.859399537421926e-06,
+      "loss": 0.5429,
+      "step": 10315
+    },
+    {
+      "epoch": 2.7386167529536705,
+      "grad_norm": 0.4462035420980938,
+      "learning_rate": 2.85905403252912e-06,
+      "loss": 0.5515,
+      "step": 10316
+    },
+    {
+      "epoch": 2.7388822514270545,
+      "grad_norm": 0.4503717683168987,
+      "learning_rate": 2.858708520633881e-06,
+      "loss": 0.597,
+      "step": 10317
+    },
+    {
+      "epoch": 2.739147749900438,
+      "grad_norm": 0.4467780220647689,
+      "learning_rate": 2.8583630017429474e-06,
+      "loss": 0.5665,
+      "step": 10318
+    },
+    {
+      "epoch": 2.739413248373822,
+      "grad_norm": 0.4408404904903175,
+      "learning_rate": 2.8580174758630584e-06,
+      "loss": 0.5589,
+      "step": 10319
+    },
+    {
+      "epoch": 2.7396787468472055,
+      "grad_norm": 0.4570127388806598,
+      "learning_rate": 2.8576719430009514e-06,
+      "loss": 0.5719,
+      "step": 10320
+    },
+    {
+      "epoch": 2.7399442453205896,
+      "grad_norm": 0.44217595899356277,
+      "learning_rate": 2.8573264031633663e-06,
+      "loss": 0.5466,
+      "step": 10321
+    },
+    {
+      "epoch": 2.740209743793973,
+      "grad_norm": 0.4292987868333343,
+      "learning_rate": 2.8569808563570413e-06,
+      "loss": 0.5581,
+      "step": 10322
+    },
+    {
+      "epoch": 2.740475242267357,
+      "grad_norm": 0.43339613211084893,
+      "learning_rate": 2.856635302588715e-06,
+      "loss": 0.5548,
+      "step": 10323
+    },
+    {
+      "epoch": 2.7407407407407405,
+      "grad_norm": 0.426727202307013,
+      "learning_rate": 2.8562897418651277e-06,
+      "loss": 0.5539,
+      "step": 10324
+    },
+    {
+      "epoch": 2.7410062392141246,
+      "grad_norm": 0.43100006376263783,
+      "learning_rate": 2.8559441741930176e-06,
+      "loss": 0.55,
+      "step": 10325
+    },
+    {
+      "epoch": 2.741271737687508,
+      "grad_norm": 0.4437840074908016,
+      "learning_rate": 2.8555985995791252e-06,
+      "loss": 0.5617,
+      "step": 10326
+    },
+    {
+      "epoch": 2.741537236160892,
+      "grad_norm": 0.4361272163148602,
+      "learning_rate": 2.8552530180301896e-06,
+      "loss": 0.5363,
+      "step": 10327
+    },
+    {
+      "epoch": 2.741802734634276,
+      "grad_norm": 0.45490136042657886,
+      "learning_rate": 2.8549074295529496e-06,
+      "loss": 0.5485,
+      "step": 10328
+    },
+    {
+      "epoch": 2.7420682331076596,
+      "grad_norm": 0.4411175260710266,
+      "learning_rate": 2.854561834154146e-06,
+      "loss": 0.5547,
+      "step": 10329
+    },
+    {
+      "epoch": 2.742333731581043,
+      "grad_norm": 0.44883807460656977,
+      "learning_rate": 2.854216231840519e-06,
+      "loss": 0.5543,
+      "step": 10330
+    },
+    {
+      "epoch": 2.7425992300544273,
+      "grad_norm": 0.42735714815829845,
+      "learning_rate": 2.853870622618808e-06,
+      "loss": 0.5412,
+      "step": 10331
+    },
+    {
+      "epoch": 2.742864728527811,
+      "grad_norm": 0.4365619714477265,
+      "learning_rate": 2.8535250064957524e-06,
+      "loss": 0.5703,
+      "step": 10332
+    },
+    {
+      "epoch": 2.7431302270011946,
+      "grad_norm": 0.4364700836451843,
+      "learning_rate": 2.8531793834780945e-06,
+      "loss": 0.5507,
+      "step": 10333
+    },
+    {
+      "epoch": 2.7433957254745787,
+      "grad_norm": 0.4362064039030367,
+      "learning_rate": 2.8528337535725735e-06,
+      "loss": 0.5354,
+      "step": 10334
+    },
+    {
+      "epoch": 2.7436612239479623,
+      "grad_norm": 0.4494013599318121,
+      "learning_rate": 2.852488116785931e-06,
+      "loss": 0.5315,
+      "step": 10335
+    },
+    {
+      "epoch": 2.743926722421346,
+      "grad_norm": 0.4482954919432676,
+      "learning_rate": 2.8521424731249077e-06,
+      "loss": 0.5327,
+      "step": 10336
+    },
+    {
+      "epoch": 2.74419222089473,
+      "grad_norm": 0.43838764627046833,
+      "learning_rate": 2.8517968225962427e-06,
+      "loss": 0.5162,
+      "step": 10337
+    },
+    {
+      "epoch": 2.7444577193681137,
+      "grad_norm": 0.4364985111413365,
+      "learning_rate": 2.851451165206679e-06,
+      "loss": 0.55,
+      "step": 10338
+    },
+    {
+      "epoch": 2.7447232178414973,
+      "grad_norm": 0.4333031956021111,
+      "learning_rate": 2.851105500962957e-06,
+      "loss": 0.5401,
+      "step": 10339
+    },
+    {
+      "epoch": 2.7449887163148814,
+      "grad_norm": 0.4650515288523981,
+      "learning_rate": 2.850759829871818e-06,
+      "loss": 0.5702,
+      "step": 10340
+    },
+    {
+      "epoch": 2.745254214788265,
+      "grad_norm": 0.4542592947308028,
+      "learning_rate": 2.8504141519400037e-06,
+      "loss": 0.5608,
+      "step": 10341
+    },
+    {
+      "epoch": 2.7455197132616487,
+      "grad_norm": 0.44073447454774434,
+      "learning_rate": 2.8500684671742547e-06,
+      "loss": 0.5734,
+      "step": 10342
+    },
+    {
+      "epoch": 2.7457852117350328,
+      "grad_norm": 0.431434601367619,
+      "learning_rate": 2.8497227755813138e-06,
+      "loss": 0.5419,
+      "step": 10343
+    },
+    {
+      "epoch": 2.7460507102084164,
+      "grad_norm": 0.44235496145075687,
+      "learning_rate": 2.8493770771679234e-06,
+      "loss": 0.5573,
+      "step": 10344
+    },
+    {
+      "epoch": 2.7463162086818,
+      "grad_norm": 0.4421584125154759,
+      "learning_rate": 2.8490313719408238e-06,
+      "loss": 0.5545,
+      "step": 10345
+    },
+    {
+      "epoch": 2.746581707155184,
+      "grad_norm": 0.4429390801958356,
+      "learning_rate": 2.848685659906758e-06,
+      "loss": 0.5341,
+      "step": 10346
+    },
+    {
+      "epoch": 2.7468472056285678,
+      "grad_norm": 0.43919380812696407,
+      "learning_rate": 2.848339941072468e-06,
+      "loss": 0.5553,
+      "step": 10347
+    },
+    {
+      "epoch": 2.7471127041019514,
+      "grad_norm": 0.44228658446309593,
+      "learning_rate": 2.8479942154446965e-06,
+      "loss": 0.5798,
+      "step": 10348
+    },
+    {
+      "epoch": 2.747378202575335,
+      "grad_norm": 0.4243031030067016,
+      "learning_rate": 2.8476484830301856e-06,
+      "loss": 0.5617,
+      "step": 10349
+    },
+    {
+      "epoch": 2.747643701048719,
+      "grad_norm": 0.45169389106771335,
+      "learning_rate": 2.8473027438356786e-06,
+      "loss": 0.5773,
+      "step": 10350
+    },
+    {
+      "epoch": 2.7479091995221028,
+      "grad_norm": 0.41382595798301464,
+      "learning_rate": 2.8469569978679167e-06,
+      "loss": 0.5222,
+      "step": 10351
+    },
+    {
+      "epoch": 2.7481746979954864,
+      "grad_norm": 0.434215407756758,
+      "learning_rate": 2.846611245133644e-06,
+      "loss": 0.555,
+      "step": 10352
+    },
+    {
+      "epoch": 2.74844019646887,
+      "grad_norm": 0.41105680802092653,
+      "learning_rate": 2.8462654856396045e-06,
+      "loss": 0.5177,
+      "step": 10353
+    },
+    {
+      "epoch": 2.748705694942254,
+      "grad_norm": 0.4372703123715657,
+      "learning_rate": 2.845919719392539e-06,
+      "loss": 0.5877,
+      "step": 10354
+    },
+    {
+      "epoch": 2.748971193415638,
+      "grad_norm": 0.44502346600000375,
+      "learning_rate": 2.8455739463991927e-06,
+      "loss": 0.6182,
+      "step": 10355
+    },
+    {
+      "epoch": 2.7492366918890214,
+      "grad_norm": 0.4371636584793063,
+      "learning_rate": 2.8452281666663085e-06,
+      "loss": 0.5944,
+      "step": 10356
+    },
+    {
+      "epoch": 2.7495021903624055,
+      "grad_norm": 0.44417921068643,
+      "learning_rate": 2.844882380200629e-06,
+      "loss": 0.5446,
+      "step": 10357
+    },
+    {
+      "epoch": 2.749767688835789,
+      "grad_norm": 0.4495981338486847,
+      "learning_rate": 2.8445365870088992e-06,
+      "loss": 0.57,
+      "step": 10358
+    },
+    {
+      "epoch": 2.750033187309173,
+      "grad_norm": 0.43896790265342484,
+      "learning_rate": 2.844190787097863e-06,
+      "loss": 0.5422,
+      "step": 10359
+    },
+    {
+      "epoch": 2.750298685782557,
+      "grad_norm": 0.44425507257667096,
+      "learning_rate": 2.843844980474263e-06,
+      "loss": 0.5516,
+      "step": 10360
+    },
+    {
+      "epoch": 2.7505641842559405,
+      "grad_norm": 0.4572694847231948,
+      "learning_rate": 2.843499167144843e-06,
+      "loss": 0.5897,
+      "step": 10361
+    },
+    {
+      "epoch": 2.750829682729324,
+      "grad_norm": 0.4312272602493397,
+      "learning_rate": 2.8431533471163497e-06,
+      "loss": 0.5827,
+      "step": 10362
+    },
+    {
+      "epoch": 2.7510951812027082,
+      "grad_norm": 0.44783875949204516,
+      "learning_rate": 2.8428075203955253e-06,
+      "loss": 0.5484,
+      "step": 10363
+    },
+    {
+      "epoch": 2.751360679676092,
+      "grad_norm": 0.4528468735810305,
+      "learning_rate": 2.842461686989115e-06,
+      "loss": 0.5699,
+      "step": 10364
+    },
+    {
+      "epoch": 2.7516261781494755,
+      "grad_norm": 0.44944031830805287,
+      "learning_rate": 2.842115846903864e-06,
+      "loss": 0.548,
+      "step": 10365
+    },
+    {
+      "epoch": 2.7518916766228596,
+      "grad_norm": 0.44012100969131446,
+      "learning_rate": 2.8417700001465156e-06,
+      "loss": 0.5833,
+      "step": 10366
+    },
+    {
+      "epoch": 2.7521571750962432,
+      "grad_norm": 0.44636105350907634,
+      "learning_rate": 2.8414241467238156e-06,
+      "loss": 0.5655,
+      "step": 10367
+    },
+    {
+      "epoch": 2.752422673569627,
+      "grad_norm": 0.465260730003754,
+      "learning_rate": 2.8410782866425092e-06,
+      "loss": 0.5657,
+      "step": 10368
+    },
+    {
+      "epoch": 2.752688172043011,
+      "grad_norm": 0.4375795658230993,
+      "learning_rate": 2.84073241990934e-06,
+      "loss": 0.5595,
+      "step": 10369
+    },
+    {
+      "epoch": 2.7529536705163946,
+      "grad_norm": 0.4666940037495597,
+      "learning_rate": 2.840386546531055e-06,
+      "loss": 0.5586,
+      "step": 10370
+    },
+    {
+      "epoch": 2.7532191689897783,
+      "grad_norm": 0.4342168920567903,
+      "learning_rate": 2.8400406665143987e-06,
+      "loss": 0.5654,
+      "step": 10371
+    },
+    {
+      "epoch": 2.7534846674631623,
+      "grad_norm": 0.44029261239817913,
+      "learning_rate": 2.839694779866117e-06,
+      "loss": 0.6042,
+      "step": 10372
+    },
+    {
+      "epoch": 2.753750165936546,
+      "grad_norm": 0.4406684636472852,
+      "learning_rate": 2.839348886592956e-06,
+      "loss": 0.5534,
+      "step": 10373
+    },
+    {
+      "epoch": 2.7540156644099296,
+      "grad_norm": 0.45779469119606037,
+      "learning_rate": 2.83900298670166e-06,
+      "loss": 0.5925,
+      "step": 10374
+    },
+    {
+      "epoch": 2.7542811628833133,
+      "grad_norm": 0.48758790918103284,
+      "learning_rate": 2.838657080198976e-06,
+      "loss": 0.5194,
+      "step": 10375
+    },
+    {
+      "epoch": 2.7545466613566973,
+      "grad_norm": 0.43934849146512,
+      "learning_rate": 2.83831116709165e-06,
+      "loss": 0.5638,
+      "step": 10376
+    },
+    {
+      "epoch": 2.754812159830081,
+      "grad_norm": 0.4399822948444774,
+      "learning_rate": 2.837965247386428e-06,
+      "loss": 0.5654,
+      "step": 10377
+    },
+    {
+      "epoch": 2.7550776583034646,
+      "grad_norm": 0.44247494973297535,
+      "learning_rate": 2.8376193210900555e-06,
+      "loss": 0.563,
+      "step": 10378
+    },
+    {
+      "epoch": 2.7553431567768483,
+      "grad_norm": 0.43830208665937304,
+      "learning_rate": 2.83727338820928e-06,
+      "loss": 0.5277,
+      "step": 10379
+    },
+    {
+      "epoch": 2.7556086552502324,
+      "grad_norm": 0.4765741882696403,
+      "learning_rate": 2.8369274487508476e-06,
+      "loss": 0.5568,
+      "step": 10380
+    },
+    {
+      "epoch": 2.755874153723616,
+      "grad_norm": 0.4692609222245702,
+      "learning_rate": 2.836581502721505e-06,
+      "loss": 0.5665,
+      "step": 10381
+    },
+    {
+      "epoch": 2.7561396521969996,
+      "grad_norm": 0.44860205205317094,
+      "learning_rate": 2.8362355501279994e-06,
+      "loss": 0.5525,
+      "step": 10382
+    },
+    {
+      "epoch": 2.7564051506703837,
+      "grad_norm": 0.46521192222691604,
+      "learning_rate": 2.8358895909770766e-06,
+      "loss": 0.563,
+      "step": 10383
+    },
+    {
+      "epoch": 2.7566706491437674,
+      "grad_norm": 0.4474053102130928,
+      "learning_rate": 2.8355436252754847e-06,
+      "loss": 0.5842,
+      "step": 10384
+    },
+    {
+      "epoch": 2.756936147617151,
+      "grad_norm": 0.44445799731814095,
+      "learning_rate": 2.835197653029971e-06,
+      "loss": 0.5576,
+      "step": 10385
+    },
+    {
+      "epoch": 2.757201646090535,
+      "grad_norm": 0.45401756737886223,
+      "learning_rate": 2.834851674247282e-06,
+      "loss": 0.5407,
+      "step": 10386
+    },
+    {
+      "epoch": 2.7574671445639187,
+      "grad_norm": 0.43142535199135806,
+      "learning_rate": 2.8345056889341647e-06,
+      "loss": 0.5222,
+      "step": 10387
+    },
+    {
+      "epoch": 2.7577326430373024,
+      "grad_norm": 0.4214409156777529,
+      "learning_rate": 2.8341596970973683e-06,
+      "loss": 0.5426,
+      "step": 10388
+    },
+    {
+      "epoch": 2.7579981415106865,
+      "grad_norm": 0.4658747984371521,
+      "learning_rate": 2.8338136987436397e-06,
+      "loss": 0.5382,
+      "step": 10389
+    },
+    {
+      "epoch": 2.75826363998407,
+      "grad_norm": 0.44835710242458643,
+      "learning_rate": 2.8334676938797264e-06,
+      "loss": 0.566,
+      "step": 10390
+    },
+    {
+      "epoch": 2.7585291384574537,
+      "grad_norm": 0.43426629376815395,
+      "learning_rate": 2.833121682512377e-06,
+      "loss": 0.5949,
+      "step": 10391
+    },
+    {
+      "epoch": 2.758794636930838,
+      "grad_norm": 0.43755638084860565,
+      "learning_rate": 2.832775664648339e-06,
+      "loss": 0.537,
+      "step": 10392
+    },
+    {
+      "epoch": 2.7590601354042215,
+      "grad_norm": 0.43040285557481295,
+      "learning_rate": 2.8324296402943606e-06,
+      "loss": 0.5474,
+      "step": 10393
+    },
+    {
+      "epoch": 2.759325633877605,
+      "grad_norm": 0.44265473232882074,
+      "learning_rate": 2.832083609457191e-06,
+      "loss": 0.5799,
+      "step": 10394
+    },
+    {
+      "epoch": 2.759591132350989,
+      "grad_norm": 0.43971191906185214,
+      "learning_rate": 2.831737572143577e-06,
+      "loss": 0.5434,
+      "step": 10395
+    },
+    {
+      "epoch": 2.759856630824373,
+      "grad_norm": 0.4316051737507818,
+      "learning_rate": 2.8313915283602684e-06,
+      "loss": 0.5337,
+      "step": 10396
+    },
+    {
+      "epoch": 2.7601221292977565,
+      "grad_norm": 0.441735364058747,
+      "learning_rate": 2.831045478114013e-06,
+      "loss": 0.562,
+      "step": 10397
+    },
+    {
+      "epoch": 2.7603876277711406,
+      "grad_norm": 0.4334694278868336,
+      "learning_rate": 2.830699421411561e-06,
+      "loss": 0.5434,
+      "step": 10398
+    },
+    {
+      "epoch": 2.760653126244524,
+      "grad_norm": 0.45532016165336586,
+      "learning_rate": 2.8303533582596605e-06,
+      "loss": 0.5692,
+      "step": 10399
+    },
+    {
+      "epoch": 2.760918624717908,
+      "grad_norm": 0.4310481823721104,
+      "learning_rate": 2.8300072886650604e-06,
+      "loss": 0.5336,
+      "step": 10400
+    },
+    {
+      "epoch": 2.761184123191292,
+      "grad_norm": 0.458014931421607,
+      "learning_rate": 2.8296612126345104e-06,
+      "loss": 0.5351,
+      "step": 10401
+    },
+    {
+      "epoch": 2.7614496216646756,
+      "grad_norm": 0.4547647171633738,
+      "learning_rate": 2.82931513017476e-06,
+      "loss": 0.5587,
+      "step": 10402
+    },
+    {
+      "epoch": 2.761715120138059,
+      "grad_norm": 0.4296890502659497,
+      "learning_rate": 2.8289690412925573e-06,
+      "loss": 0.5666,
+      "step": 10403
+    },
+    {
+      "epoch": 2.761980618611443,
+      "grad_norm": 0.43374033439784093,
+      "learning_rate": 2.8286229459946534e-06,
+      "loss": 0.5645,
+      "step": 10404
+    },
+    {
+      "epoch": 2.762246117084827,
+      "grad_norm": 0.4288191046132695,
+      "learning_rate": 2.8282768442877976e-06,
+      "loss": 0.5945,
+      "step": 10405
+    },
+    {
+      "epoch": 2.7625116155582106,
+      "grad_norm": 0.4433726197619747,
+      "learning_rate": 2.8279307361787394e-06,
+      "loss": 0.5722,
+      "step": 10406
+    },
+    {
+      "epoch": 2.762777114031594,
+      "grad_norm": 0.4461913777228567,
+      "learning_rate": 2.8275846216742287e-06,
+      "loss": 0.5763,
+      "step": 10407
+    },
+    {
+      "epoch": 2.763042612504978,
+      "grad_norm": 0.4394180484470262,
+      "learning_rate": 2.8272385007810166e-06,
+      "loss": 0.5343,
+      "step": 10408
+    },
+    {
+      "epoch": 2.763308110978362,
+      "grad_norm": 0.43727444056490866,
+      "learning_rate": 2.8268923735058516e-06,
+      "loss": 0.5873,
+      "step": 10409
+    },
+    {
+      "epoch": 2.7635736094517456,
+      "grad_norm": 0.4423476244037865,
+      "learning_rate": 2.8265462398554847e-06,
+      "loss": 0.5446,
+      "step": 10410
+    },
+    {
+      "epoch": 2.763839107925129,
+      "grad_norm": 0.44610067424462474,
+      "learning_rate": 2.826200099836668e-06,
+      "loss": 0.5778,
+      "step": 10411
+    },
+    {
+      "epoch": 2.7641046063985133,
+      "grad_norm": 0.4630554212385094,
+      "learning_rate": 2.8258539534561497e-06,
+      "loss": 0.5164,
+      "step": 10412
+    },
+    {
+      "epoch": 2.764370104871897,
+      "grad_norm": 0.4252987547976094,
+      "learning_rate": 2.8255078007206817e-06,
+      "loss": 0.5446,
+      "step": 10413
+    },
+    {
+      "epoch": 2.7646356033452806,
+      "grad_norm": 0.43068527404955054,
+      "learning_rate": 2.8251616416370147e-06,
+      "loss": 0.5512,
+      "step": 10414
+    },
+    {
+      "epoch": 2.7649011018186647,
+      "grad_norm": 0.4516559538378109,
+      "learning_rate": 2.8248154762118995e-06,
+      "loss": 0.5083,
+      "step": 10415
+    },
+    {
+      "epoch": 2.7651666002920483,
+      "grad_norm": 0.4511639399421076,
+      "learning_rate": 2.824469304452087e-06,
+      "loss": 0.5812,
+      "step": 10416
+    },
+    {
+      "epoch": 2.765432098765432,
+      "grad_norm": 0.43816184492083227,
+      "learning_rate": 2.824123126364329e-06,
+      "loss": 0.5845,
+      "step": 10417
+    },
+    {
+      "epoch": 2.765697597238816,
+      "grad_norm": 0.4395001660844736,
+      "learning_rate": 2.8237769419553757e-06,
+      "loss": 0.5315,
+      "step": 10418
+    },
+    {
+      "epoch": 2.7659630957121997,
+      "grad_norm": 0.43088502168696324,
+      "learning_rate": 2.82343075123198e-06,
+      "loss": 0.6021,
+      "step": 10419
+    },
+    {
+      "epoch": 2.7662285941855833,
+      "grad_norm": 0.44579347053699364,
+      "learning_rate": 2.8230845542008927e-06,
+      "loss": 0.5777,
+      "step": 10420
+    },
+    {
+      "epoch": 2.7664940926589674,
+      "grad_norm": 0.45889474200732183,
+      "learning_rate": 2.8227383508688653e-06,
+      "loss": 0.5324,
+      "step": 10421
+    },
+    {
+      "epoch": 2.766759591132351,
+      "grad_norm": 0.446563758104612,
+      "learning_rate": 2.8223921412426504e-06,
+      "loss": 0.5544,
+      "step": 10422
+    },
+    {
+      "epoch": 2.7670250896057347,
+      "grad_norm": 0.43395202679761435,
+      "learning_rate": 2.822045925328999e-06,
+      "loss": 0.5348,
+      "step": 10423
+    },
+    {
+      "epoch": 2.7672905880791188,
+      "grad_norm": 0.4362175831318424,
+      "learning_rate": 2.8216997031346633e-06,
+      "loss": 0.5746,
+      "step": 10424
+    },
+    {
+      "epoch": 2.7675560865525024,
+      "grad_norm": 0.4535493488882397,
+      "learning_rate": 2.8213534746663958e-06,
+      "loss": 0.579,
+      "step": 10425
+    },
+    {
+      "epoch": 2.767821585025886,
+      "grad_norm": 0.44663352989242516,
+      "learning_rate": 2.8210072399309484e-06,
+      "loss": 0.5618,
+      "step": 10426
+    },
+    {
+      "epoch": 2.76808708349927,
+      "grad_norm": 0.43944910461314524,
+      "learning_rate": 2.820660998935074e-06,
+      "loss": 0.5409,
+      "step": 10427
+    },
+    {
+      "epoch": 2.7683525819726538,
+      "grad_norm": 0.43732145388218513,
+      "learning_rate": 2.820314751685526e-06,
+      "loss": 0.5753,
+      "step": 10428
+    },
+    {
+      "epoch": 2.7686180804460374,
+      "grad_norm": 0.4268623368929724,
+      "learning_rate": 2.819968498189055e-06,
+      "loss": 0.5521,
+      "step": 10429
+    },
+    {
+      "epoch": 2.768883578919421,
+      "grad_norm": 0.4341207782016763,
+      "learning_rate": 2.819622238452415e-06,
+      "loss": 0.5764,
+      "step": 10430
+    },
+    {
+      "epoch": 2.769149077392805,
+      "grad_norm": 0.444066710970643,
+      "learning_rate": 2.8192759724823598e-06,
+      "loss": 0.5484,
+      "step": 10431
+    },
+    {
+      "epoch": 2.7694145758661888,
+      "grad_norm": 0.44742294242189534,
+      "learning_rate": 2.8189297002856407e-06,
+      "loss": 0.5611,
+      "step": 10432
+    },
+    {
+      "epoch": 2.7696800743395724,
+      "grad_norm": 0.4251223472526734,
+      "learning_rate": 2.8185834218690117e-06,
+      "loss": 0.5575,
+      "step": 10433
+    },
+    {
+      "epoch": 2.769945572812956,
+      "grad_norm": 0.44098748647605973,
+      "learning_rate": 2.8182371372392253e-06,
+      "loss": 0.5462,
+      "step": 10434
+    },
+    {
+      "epoch": 2.77021107128634,
+      "grad_norm": 0.4349442280508103,
+      "learning_rate": 2.8178908464030363e-06,
+      "loss": 0.6108,
+      "step": 10435
+    },
+    {
+      "epoch": 2.770476569759724,
+      "grad_norm": 0.4482186329336556,
+      "learning_rate": 2.817544549367197e-06,
+      "loss": 0.5364,
+      "step": 10436
+    },
+    {
+      "epoch": 2.7707420682331074,
+      "grad_norm": 0.4455139050163501,
+      "learning_rate": 2.8171982461384616e-06,
+      "loss": 0.5541,
+      "step": 10437
+    },
+    {
+      "epoch": 2.7710075667064915,
+      "grad_norm": 0.4376667630388307,
+      "learning_rate": 2.816851936723584e-06,
+      "loss": 0.5764,
+      "step": 10438
+    },
+    {
+      "epoch": 2.771273065179875,
+      "grad_norm": 0.4440568799011048,
+      "learning_rate": 2.8165056211293186e-06,
+      "loss": 0.5824,
+      "step": 10439
+    },
+    {
+      "epoch": 2.771538563653259,
+      "grad_norm": 0.45588536453291484,
+      "learning_rate": 2.8161592993624175e-06,
+      "loss": 0.5034,
+      "step": 10440
+    },
+    {
+      "epoch": 2.771804062126643,
+      "grad_norm": 0.4097374512846095,
+      "learning_rate": 2.815812971429636e-06,
+      "loss": 0.5279,
+      "step": 10441
+    },
+    {
+      "epoch": 2.7720695606000265,
+      "grad_norm": 0.44841972684479187,
+      "learning_rate": 2.815466637337728e-06,
+      "loss": 0.5707,
+      "step": 10442
+    },
+    {
+      "epoch": 2.77233505907341,
+      "grad_norm": 0.4338551234463223,
+      "learning_rate": 2.8151202970934487e-06,
+      "loss": 0.5716,
+      "step": 10443
+    },
+    {
+      "epoch": 2.7726005575467942,
+      "grad_norm": 0.4365821609522977,
+      "learning_rate": 2.8147739507035522e-06,
+      "loss": 0.5778,
+      "step": 10444
+    },
+    {
+      "epoch": 2.772866056020178,
+      "grad_norm": 0.4405568759647794,
+      "learning_rate": 2.8144275981747934e-06,
+      "loss": 0.5236,
+      "step": 10445
+    },
+    {
+      "epoch": 2.7731315544935615,
+      "grad_norm": 0.4514311872677338,
+      "learning_rate": 2.814081239513926e-06,
+      "loss": 0.5475,
+      "step": 10446
+    },
+    {
+      "epoch": 2.7733970529669456,
+      "grad_norm": 0.4377383858435026,
+      "learning_rate": 2.813734874727705e-06,
+      "loss": 0.5383,
+      "step": 10447
+    },
+    {
+      "epoch": 2.7736625514403292,
+      "grad_norm": 0.4357072195342958,
+      "learning_rate": 2.813388503822887e-06,
+      "loss": 0.5416,
+      "step": 10448
+    },
+    {
+      "epoch": 2.773928049913713,
+      "grad_norm": 0.4428693730865895,
+      "learning_rate": 2.8130421268062247e-06,
+      "loss": 0.5884,
+      "step": 10449
+    },
+    {
+      "epoch": 2.774193548387097,
+      "grad_norm": 0.4351616711546554,
+      "learning_rate": 2.8126957436844747e-06,
+      "loss": 0.5402,
+      "step": 10450
+    },
+    {
+      "epoch": 2.7744590468604806,
+      "grad_norm": 0.43107814665507527,
+      "learning_rate": 2.8123493544643924e-06,
+      "loss": 0.552,
+      "step": 10451
+    },
+    {
+      "epoch": 2.7747245453338643,
+      "grad_norm": 0.44684950829612546,
+      "learning_rate": 2.812002959152733e-06,
+      "loss": 0.627,
+      "step": 10452
+    },
+    {
+      "epoch": 2.7749900438072483,
+      "grad_norm": 0.4441425613276137,
+      "learning_rate": 2.811656557756251e-06,
+      "loss": 0.5809,
+      "step": 10453
+    },
+    {
+      "epoch": 2.775255542280632,
+      "grad_norm": 0.44871380809170247,
+      "learning_rate": 2.8113101502817044e-06,
+      "loss": 0.5329,
+      "step": 10454
+    },
+    {
+      "epoch": 2.7755210407540156,
+      "grad_norm": 0.453442819464806,
+      "learning_rate": 2.810963736735847e-06,
+      "loss": 0.56,
+      "step": 10455
+    },
+    {
+      "epoch": 2.7757865392273997,
+      "grad_norm": 0.440766513265747,
+      "learning_rate": 2.8106173171254353e-06,
+      "loss": 0.5357,
+      "step": 10456
+    },
+    {
+      "epoch": 2.7760520377007833,
+      "grad_norm": 0.4361895433190712,
+      "learning_rate": 2.810270891457226e-06,
+      "loss": 0.5543,
+      "step": 10457
+    },
+    {
+      "epoch": 2.776317536174167,
+      "grad_norm": 0.4292111533195813,
+      "learning_rate": 2.809924459737974e-06,
+      "loss": 0.5532,
+      "step": 10458
+    },
+    {
+      "epoch": 2.7765830346475506,
+      "grad_norm": 0.4359804552692377,
+      "learning_rate": 2.809578021974436e-06,
+      "loss": 0.59,
+      "step": 10459
+    },
+    {
+      "epoch": 2.7768485331209347,
+      "grad_norm": 0.4516275071504658,
+      "learning_rate": 2.80923157817337e-06,
+      "loss": 0.5493,
+      "step": 10460
+    },
+    {
+      "epoch": 2.7771140315943184,
+      "grad_norm": 0.44249514003272455,
+      "learning_rate": 2.8088851283415298e-06,
+      "loss": 0.5522,
+      "step": 10461
+    },
+    {
+      "epoch": 2.777379530067702,
+      "grad_norm": 0.4461739276153777,
+      "learning_rate": 2.8085386724856734e-06,
+      "loss": 0.5427,
+      "step": 10462
+    },
+    {
+      "epoch": 2.7776450285410856,
+      "grad_norm": 0.44795781004318747,
+      "learning_rate": 2.8081922106125577e-06,
+      "loss": 0.5409,
+      "step": 10463
+    },
+    {
+      "epoch": 2.7779105270144697,
+      "grad_norm": 0.43388589812294953,
+      "learning_rate": 2.807845742728939e-06,
+      "loss": 0.5413,
+      "step": 10464
+    },
+    {
+      "epoch": 2.7781760254878534,
+      "grad_norm": 0.44392046527414447,
+      "learning_rate": 2.807499268841575e-06,
+      "loss": 0.5694,
+      "step": 10465
+    },
+    {
+      "epoch": 2.778441523961237,
+      "grad_norm": 0.4491370408988263,
+      "learning_rate": 2.8071527889572224e-06,
+      "loss": 0.5827,
+      "step": 10466
+    },
+    {
+      "epoch": 2.778707022434621,
+      "grad_norm": 0.4300037785589767,
+      "learning_rate": 2.8068063030826383e-06,
+      "loss": 0.5395,
+      "step": 10467
+    },
+    {
+      "epoch": 2.7789725209080047,
+      "grad_norm": 0.45269389369088164,
+      "learning_rate": 2.80645981122458e-06,
+      "loss": 0.5593,
+      "step": 10468
+    },
+    {
+      "epoch": 2.7792380193813884,
+      "grad_norm": 0.4688278417951856,
+      "learning_rate": 2.806113313389805e-06,
+      "loss": 0.5629,
+      "step": 10469
+    },
+    {
+      "epoch": 2.7795035178547725,
+      "grad_norm": 0.43709253594876935,
+      "learning_rate": 2.805766809585071e-06,
+      "loss": 0.522,
+      "step": 10470
+    },
+    {
+      "epoch": 2.779769016328156,
+      "grad_norm": 0.4340647150768764,
+      "learning_rate": 2.8054202998171347e-06,
+      "loss": 0.5422,
+      "step": 10471
+    },
+    {
+      "epoch": 2.7800345148015397,
+      "grad_norm": 0.44013145330380476,
+      "learning_rate": 2.8050737840927554e-06,
+      "loss": 0.5842,
+      "step": 10472
+    },
+    {
+      "epoch": 2.780300013274924,
+      "grad_norm": 0.43444892789772244,
+      "learning_rate": 2.8047272624186905e-06,
+      "loss": 0.5404,
+      "step": 10473
+    },
+    {
+      "epoch": 2.7805655117483075,
+      "grad_norm": 0.460427172920911,
+      "learning_rate": 2.804380734801699e-06,
+      "loss": 0.5172,
+      "step": 10474
+    },
+    {
+      "epoch": 2.780831010221691,
+      "grad_norm": 0.44657994425161107,
+      "learning_rate": 2.804034201248536e-06,
+      "loss": 0.58,
+      "step": 10475
+    },
+    {
+      "epoch": 2.781096508695075,
+      "grad_norm": 0.4228911374224898,
+      "learning_rate": 2.8036876617659625e-06,
+      "loss": 0.5423,
+      "step": 10476
+    },
+    {
+      "epoch": 2.781362007168459,
+      "grad_norm": 0.4409461251556048,
+      "learning_rate": 2.8033411163607365e-06,
+      "loss": 0.5705,
+      "step": 10477
+    },
+    {
+      "epoch": 2.7816275056418425,
+      "grad_norm": 0.44718543232049196,
+      "learning_rate": 2.8029945650396155e-06,
+      "loss": 0.5716,
+      "step": 10478
+    },
+    {
+      "epoch": 2.7818930041152266,
+      "grad_norm": 0.4409114102438968,
+      "learning_rate": 2.802648007809359e-06,
+      "loss": 0.5636,
+      "step": 10479
+    },
+    {
+      "epoch": 2.78215850258861,
+      "grad_norm": 0.44209091517713006,
+      "learning_rate": 2.8023014446767243e-06,
+      "loss": 0.5588,
+      "step": 10480
+    },
+    {
+      "epoch": 2.782424001061994,
+      "grad_norm": 0.45289837051153686,
+      "learning_rate": 2.8019548756484722e-06,
+      "loss": 0.5811,
+      "step": 10481
+    },
+    {
+      "epoch": 2.782689499535378,
+      "grad_norm": 0.4578715155628391,
+      "learning_rate": 2.80160830073136e-06,
+      "loss": 0.5644,
+      "step": 10482
+    },
+    {
+      "epoch": 2.7829549980087616,
+      "grad_norm": 0.456940218300286,
+      "learning_rate": 2.8012617199321487e-06,
+      "loss": 0.5292,
+      "step": 10483
+    },
+    {
+      "epoch": 2.783220496482145,
+      "grad_norm": 0.4629766024596027,
+      "learning_rate": 2.8009151332575953e-06,
+      "loss": 0.5764,
+      "step": 10484
+    },
+    {
+      "epoch": 2.783485994955529,
+      "grad_norm": 0.43351524769943706,
+      "learning_rate": 2.80056854071446e-06,
+      "loss": 0.5496,
+      "step": 10485
+    },
+    {
+      "epoch": 2.783751493428913,
+      "grad_norm": 0.44332475123629084,
+      "learning_rate": 2.800221942309503e-06,
+      "loss": 0.5478,
+      "step": 10486
+    },
+    {
+      "epoch": 2.7840169919022966,
+      "grad_norm": 0.4391659746281752,
+      "learning_rate": 2.799875338049482e-06,
+      "loss": 0.5726,
+      "step": 10487
+    },
+    {
+      "epoch": 2.78428249037568,
+      "grad_norm": 0.45699278228552365,
+      "learning_rate": 2.799528727941158e-06,
+      "loss": 0.5486,
+      "step": 10488
+    },
+    {
+      "epoch": 2.784547988849064,
+      "grad_norm": 0.4527070693125884,
+      "learning_rate": 2.799182111991291e-06,
+      "loss": 0.5839,
+      "step": 10489
+    },
+    {
+      "epoch": 2.784813487322448,
+      "grad_norm": 0.43521745818840013,
+      "learning_rate": 2.7988354902066396e-06,
+      "loss": 0.5312,
+      "step": 10490
+    },
+    {
+      "epoch": 2.7850789857958316,
+      "grad_norm": 0.4334742801710779,
+      "learning_rate": 2.798488862593965e-06,
+      "loss": 0.5466,
+      "step": 10491
+    },
+    {
+      "epoch": 2.785344484269215,
+      "grad_norm": 0.44982183326516045,
+      "learning_rate": 2.7981422291600267e-06,
+      "loss": 0.5531,
+      "step": 10492
+    },
+    {
+      "epoch": 2.7856099827425993,
+      "grad_norm": 0.43731541186878287,
+      "learning_rate": 2.7977955899115846e-06,
+      "loss": 0.5351,
+      "step": 10493
+    },
+    {
+      "epoch": 2.785875481215983,
+      "grad_norm": 0.44127802375670694,
+      "learning_rate": 2.7974489448553997e-06,
+      "loss": 0.5567,
+      "step": 10494
+    },
+    {
+      "epoch": 2.7861409796893666,
+      "grad_norm": 0.4308114385907659,
+      "learning_rate": 2.797102293998232e-06,
+      "loss": 0.5212,
+      "step": 10495
+    },
+    {
+      "epoch": 2.7864064781627507,
+      "grad_norm": 0.4351691881307375,
+      "learning_rate": 2.7967556373468417e-06,
+      "loss": 0.5493,
+      "step": 10496
+    },
+    {
+      "epoch": 2.7866719766361343,
+      "grad_norm": 0.43146999624051124,
+      "learning_rate": 2.7964089749079907e-06,
+      "loss": 0.5337,
+      "step": 10497
+    },
+    {
+      "epoch": 2.786937475109518,
+      "grad_norm": 0.4306086596391168,
+      "learning_rate": 2.7960623066884384e-06,
+      "loss": 0.556,
+      "step": 10498
+    },
+    {
+      "epoch": 2.787202973582902,
+      "grad_norm": 0.44025750445403783,
+      "learning_rate": 2.7957156326949458e-06,
+      "loss": 0.5554,
+      "step": 10499
+    },
+    {
+      "epoch": 2.7874684720562857,
+      "grad_norm": 0.44842871131581036,
+      "learning_rate": 2.795368952934275e-06,
+      "loss": 0.5596,
+      "step": 10500
+    },
+    {
+      "epoch": 2.7877339705296693,
+      "grad_norm": 0.4322385909166665,
+      "learning_rate": 2.795022267413187e-06,
+      "loss": 0.577,
+      "step": 10501
+    },
+    {
+      "epoch": 2.7879994690030534,
+      "grad_norm": 0.4431076615005532,
+      "learning_rate": 2.7946755761384416e-06,
+      "loss": 0.546,
+      "step": 10502
+    },
+    {
+      "epoch": 2.788264967476437,
+      "grad_norm": 0.42339241529810967,
+      "learning_rate": 2.7943288791168015e-06,
+      "loss": 0.5182,
+      "step": 10503
+    },
+    {
+      "epoch": 2.7885304659498207,
+      "grad_norm": 0.43891955385552456,
+      "learning_rate": 2.7939821763550275e-06,
+      "loss": 0.561,
+      "step": 10504
+    },
+    {
+      "epoch": 2.7887959644232048,
+      "grad_norm": 0.442670818292544,
+      "learning_rate": 2.7936354678598806e-06,
+      "loss": 0.5751,
+      "step": 10505
+    },
+    {
+      "epoch": 2.7890614628965884,
+      "grad_norm": 0.43367325477647933,
+      "learning_rate": 2.793288753638124e-06,
+      "loss": 0.5303,
+      "step": 10506
+    },
+    {
+      "epoch": 2.789326961369972,
+      "grad_norm": 0.433386786819893,
+      "learning_rate": 2.792942033696518e-06,
+      "loss": 0.5627,
+      "step": 10507
+    },
+    {
+      "epoch": 2.789592459843356,
+      "grad_norm": 0.4498891646756554,
+      "learning_rate": 2.792595308041825e-06,
+      "loss": 0.5483,
+      "step": 10508
+    },
+    {
+      "epoch": 2.7898579583167398,
+      "grad_norm": 0.44050096668010646,
+      "learning_rate": 2.792248576680807e-06,
+      "loss": 0.5427,
+      "step": 10509
+    },
+    {
+      "epoch": 2.7901234567901234,
+      "grad_norm": 0.42376822589900326,
+      "learning_rate": 2.791901839620227e-06,
+      "loss": 0.5499,
+      "step": 10510
+    },
+    {
+      "epoch": 2.7903889552635075,
+      "grad_norm": 0.41797640623111665,
+      "learning_rate": 2.7915550968668457e-06,
+      "loss": 0.4954,
+      "step": 10511
+    },
+    {
+      "epoch": 2.790654453736891,
+      "grad_norm": 0.4543151101844229,
+      "learning_rate": 2.7912083484274266e-06,
+      "loss": 0.5929,
+      "step": 10512
+    },
+    {
+      "epoch": 2.790919952210275,
+      "grad_norm": 0.4402275738029535,
+      "learning_rate": 2.790861594308732e-06,
+      "loss": 0.5427,
+      "step": 10513
+    },
+    {
+      "epoch": 2.7911854506836584,
+      "grad_norm": 0.4425170165553875,
+      "learning_rate": 2.7905148345175236e-06,
+      "loss": 0.6049,
+      "step": 10514
+    },
+    {
+      "epoch": 2.7914509491570425,
+      "grad_norm": 0.44340689957085677,
+      "learning_rate": 2.7901680690605646e-06,
+      "loss": 0.5821,
+      "step": 10515
+    },
+    {
+      "epoch": 2.791716447630426,
+      "grad_norm": 0.4392457825631366,
+      "learning_rate": 2.7898212979446184e-06,
+      "loss": 0.6112,
+      "step": 10516
+    },
+    {
+      "epoch": 2.79198194610381,
+      "grad_norm": 0.42883478684387166,
+      "learning_rate": 2.7894745211764462e-06,
+      "loss": 0.522,
+      "step": 10517
+    },
+    {
+      "epoch": 2.7922474445771934,
+      "grad_norm": 0.42951308276688926,
+      "learning_rate": 2.789127738762813e-06,
+      "loss": 0.5389,
+      "step": 10518
+    },
+    {
+      "epoch": 2.7925129430505775,
+      "grad_norm": 0.44832622581488474,
+      "learning_rate": 2.7887809507104807e-06,
+      "loss": 0.5555,
+      "step": 10519
+    },
+    {
+      "epoch": 2.792778441523961,
+      "grad_norm": 0.4412561276020851,
+      "learning_rate": 2.788434157026213e-06,
+      "loss": 0.52,
+      "step": 10520
+    },
+    {
+      "epoch": 2.793043939997345,
+      "grad_norm": 0.4478459143617638,
+      "learning_rate": 2.7880873577167726e-06,
+      "loss": 0.5607,
+      "step": 10521
+    },
+    {
+      "epoch": 2.793309438470729,
+      "grad_norm": 0.45758509786411566,
+      "learning_rate": 2.7877405527889232e-06,
+      "loss": 0.5664,
+      "step": 10522
+    },
+    {
+      "epoch": 2.7935749369441125,
+      "grad_norm": 0.4272726399720209,
+      "learning_rate": 2.7873937422494298e-06,
+      "loss": 0.5271,
+      "step": 10523
+    },
+    {
+      "epoch": 2.793840435417496,
+      "grad_norm": 0.4487350175073668,
+      "learning_rate": 2.7870469261050536e-06,
+      "loss": 0.5741,
+      "step": 10524
+    },
+    {
+      "epoch": 2.7941059338908802,
+      "grad_norm": 0.43904134072888507,
+      "learning_rate": 2.7867001043625596e-06,
+      "loss": 0.5658,
+      "step": 10525
+    },
+    {
+      "epoch": 2.794371432364264,
+      "grad_norm": 0.4419789448856741,
+      "learning_rate": 2.786353277028712e-06,
+      "loss": 0.5265,
+      "step": 10526
+    },
+    {
+      "epoch": 2.7946369308376475,
+      "grad_norm": 0.4437510334547723,
+      "learning_rate": 2.786006444110274e-06,
+      "loss": 0.5602,
+      "step": 10527
+    },
+    {
+      "epoch": 2.7949024293110316,
+      "grad_norm": 0.43769917218369814,
+      "learning_rate": 2.78565960561401e-06,
+      "loss": 0.5493,
+      "step": 10528
+    },
+    {
+      "epoch": 2.7951679277844153,
+      "grad_norm": 0.43720577913508585,
+      "learning_rate": 2.785312761546685e-06,
+      "loss": 0.551,
+      "step": 10529
+    },
+    {
+      "epoch": 2.795433426257799,
+      "grad_norm": 0.4252455232570073,
+      "learning_rate": 2.7849659119150617e-06,
+      "loss": 0.5013,
+      "step": 10530
+    },
+    {
+      "epoch": 2.795698924731183,
+      "grad_norm": 0.42993662535985805,
+      "learning_rate": 2.7846190567259065e-06,
+      "loss": 0.5325,
+      "step": 10531
+    },
+    {
+      "epoch": 2.7959644232045666,
+      "grad_norm": 0.4510636230448123,
+      "learning_rate": 2.784272195985982e-06,
+      "loss": 0.6062,
+      "step": 10532
+    },
+    {
+      "epoch": 2.7962299216779503,
+      "grad_norm": 0.4449948562515512,
+      "learning_rate": 2.783925329702054e-06,
+      "loss": 0.6123,
+      "step": 10533
+    },
+    {
+      "epoch": 2.7964954201513343,
+      "grad_norm": 0.43664890573548476,
+      "learning_rate": 2.7835784578808867e-06,
+      "loss": 0.5426,
+      "step": 10534
+    },
+    {
+      "epoch": 2.796760918624718,
+      "grad_norm": 0.4341847820773193,
+      "learning_rate": 2.783231580529245e-06,
+      "loss": 0.5775,
+      "step": 10535
+    },
+    {
+      "epoch": 2.7970264170981016,
+      "grad_norm": 0.4466786420060746,
+      "learning_rate": 2.7828846976538937e-06,
+      "loss": 0.562,
+      "step": 10536
+    },
+    {
+      "epoch": 2.7972919155714857,
+      "grad_norm": 0.44597150859224177,
+      "learning_rate": 2.7825378092615996e-06,
+      "loss": 0.5481,
+      "step": 10537
+    },
+    {
+      "epoch": 2.7975574140448694,
+      "grad_norm": 0.4445035943142575,
+      "learning_rate": 2.7821909153591254e-06,
+      "loss": 0.5456,
+      "step": 10538
+    },
+    {
+      "epoch": 2.797822912518253,
+      "grad_norm": 0.4568422746265537,
+      "learning_rate": 2.781844015953237e-06,
+      "loss": 0.5677,
+      "step": 10539
+    },
+    {
+      "epoch": 2.7980884109916366,
+      "grad_norm": 0.44309339754799343,
+      "learning_rate": 2.781497111050701e-06,
+      "loss": 0.5617,
+      "step": 10540
+    },
+    {
+      "epoch": 2.7983539094650207,
+      "grad_norm": 0.4456142447545801,
+      "learning_rate": 2.7811502006582817e-06,
+      "loss": 0.5814,
+      "step": 10541
+    },
+    {
+      "epoch": 2.7986194079384044,
+      "grad_norm": 0.42861490071073605,
+      "learning_rate": 2.7808032847827454e-06,
+      "loss": 0.5524,
+      "step": 10542
+    },
+    {
+      "epoch": 2.798884906411788,
+      "grad_norm": 0.43350977887963693,
+      "learning_rate": 2.780456363430858e-06,
+      "loss": 0.5702,
+      "step": 10543
+    },
+    {
+      "epoch": 2.7991504048851716,
+      "grad_norm": 0.4328383772956244,
+      "learning_rate": 2.7801094366093832e-06,
+      "loss": 0.5248,
+      "step": 10544
+    },
+    {
+      "epoch": 2.7994159033585557,
+      "grad_norm": 0.43908501157336555,
+      "learning_rate": 2.77976250432509e-06,
+      "loss": 0.5667,
+      "step": 10545
+    },
+    {
+      "epoch": 2.7996814018319394,
+      "grad_norm": 0.45323643999360086,
+      "learning_rate": 2.7794155665847428e-06,
+      "loss": 0.5803,
+      "step": 10546
+    },
+    {
+      "epoch": 2.799946900305323,
+      "grad_norm": 0.44208681324142574,
+      "learning_rate": 2.7790686233951074e-06,
+      "loss": 0.589,
+      "step": 10547
+    },
+    {
+      "epoch": 2.800212398778707,
+      "grad_norm": 0.4441191195150581,
+      "learning_rate": 2.7787216747629508e-06,
+      "loss": 0.5544,
+      "step": 10548
+    },
+    {
+      "epoch": 2.8004778972520907,
+      "grad_norm": 0.45056287323084654,
+      "learning_rate": 2.7783747206950394e-06,
+      "loss": 0.55,
+      "step": 10549
+    },
+    {
+      "epoch": 2.8007433957254744,
+      "grad_norm": 0.4422943593481195,
+      "learning_rate": 2.7780277611981394e-06,
+      "loss": 0.5398,
+      "step": 10550
+    },
+    {
+      "epoch": 2.8010088941988585,
+      "grad_norm": 0.4467493248070624,
+      "learning_rate": 2.7776807962790168e-06,
+      "loss": 0.564,
+      "step": 10551
+    },
+    {
+      "epoch": 2.801274392672242,
+      "grad_norm": 0.43920425013664993,
+      "learning_rate": 2.7773338259444395e-06,
+      "loss": 0.552,
+      "step": 10552
+    },
+    {
+      "epoch": 2.8015398911456257,
+      "grad_norm": 0.443143371632332,
+      "learning_rate": 2.776986850201173e-06,
+      "loss": 0.5491,
+      "step": 10553
+    },
+    {
+      "epoch": 2.80180538961901,
+      "grad_norm": 0.427892193826595,
+      "learning_rate": 2.776639869055984e-06,
+      "loss": 0.5907,
+      "step": 10554
+    },
+    {
+      "epoch": 2.8020708880923935,
+      "grad_norm": 0.43638936817574187,
+      "learning_rate": 2.7762928825156417e-06,
+      "loss": 0.5416,
+      "step": 10555
+    },
+    {
+      "epoch": 2.802336386565777,
+      "grad_norm": 0.4264751950088093,
+      "learning_rate": 2.7759458905869106e-06,
+      "loss": 0.6014,
+      "step": 10556
+    },
+    {
+      "epoch": 2.802601885039161,
+      "grad_norm": 0.4388001403143616,
+      "learning_rate": 2.77559889327656e-06,
+      "loss": 0.5757,
+      "step": 10557
+    },
+    {
+      "epoch": 2.802867383512545,
+      "grad_norm": 0.4366498684355708,
+      "learning_rate": 2.775251890591356e-06,
+      "loss": 0.5693,
+      "step": 10558
+    },
+    {
+      "epoch": 2.8031328819859285,
+      "grad_norm": 0.4442745644298153,
+      "learning_rate": 2.774904882538066e-06,
+      "loss": 0.5953,
+      "step": 10559
+    },
+    {
+      "epoch": 2.8033983804593126,
+      "grad_norm": 0.43767105485588054,
+      "learning_rate": 2.774557869123457e-06,
+      "loss": 0.5543,
+      "step": 10560
+    },
+    {
+      "epoch": 2.803663878932696,
+      "grad_norm": 0.4331951885932399,
+      "learning_rate": 2.7742108503542984e-06,
+      "loss": 0.5588,
+      "step": 10561
+    },
+    {
+      "epoch": 2.80392937740608,
+      "grad_norm": 0.4401328264003281,
+      "learning_rate": 2.773863826237356e-06,
+      "loss": 0.5742,
+      "step": 10562
+    },
+    {
+      "epoch": 2.804194875879464,
+      "grad_norm": 0.436347563718071,
+      "learning_rate": 2.7735167967793992e-06,
+      "loss": 0.5343,
+      "step": 10563
+    },
+    {
+      "epoch": 2.8044603743528476,
+      "grad_norm": 0.4426794260742564,
+      "learning_rate": 2.7731697619871943e-06,
+      "loss": 0.5871,
+      "step": 10564
+    },
+    {
+      "epoch": 2.804725872826231,
+      "grad_norm": 0.42455976872486323,
+      "learning_rate": 2.772822721867511e-06,
+      "loss": 0.5224,
+      "step": 10565
+    },
+    {
+      "epoch": 2.8049913712996153,
+      "grad_norm": 0.4463293295311799,
+      "learning_rate": 2.7724756764271168e-06,
+      "loss": 0.5476,
+      "step": 10566
+    },
+    {
+      "epoch": 2.805256869772999,
+      "grad_norm": 0.4452776662208447,
+      "learning_rate": 2.7721286256727793e-06,
+      "loss": 0.5724,
+      "step": 10567
+    },
+    {
+      "epoch": 2.8055223682463826,
+      "grad_norm": 0.43512064655992394,
+      "learning_rate": 2.771781569611267e-06,
+      "loss": 0.5475,
+      "step": 10568
+    },
+    {
+      "epoch": 2.805787866719766,
+      "grad_norm": 0.4726925165112414,
+      "learning_rate": 2.7714345082493492e-06,
+      "loss": 0.5692,
+      "step": 10569
+    },
+    {
+      "epoch": 2.8060533651931503,
+      "grad_norm": 0.4419990293823392,
+      "learning_rate": 2.771087441593794e-06,
+      "loss": 0.5674,
+      "step": 10570
+    },
+    {
+      "epoch": 2.806318863666534,
+      "grad_norm": 0.42811353149230036,
+      "learning_rate": 2.770740369651369e-06,
+      "loss": 0.5481,
+      "step": 10571
+    },
+    {
+      "epoch": 2.8065843621399176,
+      "grad_norm": 0.43469683578923585,
+      "learning_rate": 2.7703932924288447e-06,
+      "loss": 0.5561,
+      "step": 10572
+    },
+    {
+      "epoch": 2.806849860613301,
+      "grad_norm": 0.43842701465361344,
+      "learning_rate": 2.770046209932989e-06,
+      "loss": 0.5467,
+      "step": 10573
+    },
+    {
+      "epoch": 2.8071153590866853,
+      "grad_norm": 0.4386137823938788,
+      "learning_rate": 2.7696991221705703e-06,
+      "loss": 0.5492,
+      "step": 10574
+    },
+    {
+      "epoch": 2.807380857560069,
+      "grad_norm": 0.4381966253564306,
+      "learning_rate": 2.76935202914836e-06,
+      "loss": 0.5477,
+      "step": 10575
+    },
+    {
+      "epoch": 2.8076463560334526,
+      "grad_norm": 0.44927958665691115,
+      "learning_rate": 2.769004930873125e-06,
+      "loss": 0.5964,
+      "step": 10576
+    },
+    {
+      "epoch": 2.8079118545068367,
+      "grad_norm": 0.4515440615149994,
+      "learning_rate": 2.7686578273516346e-06,
+      "loss": 0.5535,
+      "step": 10577
+    },
+    {
+      "epoch": 2.8081773529802203,
+      "grad_norm": 0.44200279878803483,
+      "learning_rate": 2.7683107185906593e-06,
+      "loss": 0.5499,
+      "step": 10578
+    },
+    {
+      "epoch": 2.808442851453604,
+      "grad_norm": 0.43924566835624673,
+      "learning_rate": 2.7679636045969678e-06,
+      "loss": 0.5548,
+      "step": 10579
+    },
+    {
+      "epoch": 2.808708349926988,
+      "grad_norm": 0.4483943066425492,
+      "learning_rate": 2.7676164853773303e-06,
+      "loss": 0.5726,
+      "step": 10580
+    },
+    {
+      "epoch": 2.8089738484003717,
+      "grad_norm": 0.4610698812710244,
+      "learning_rate": 2.7672693609385155e-06,
+      "loss": 0.5407,
+      "step": 10581
+    },
+    {
+      "epoch": 2.8092393468737553,
+      "grad_norm": 0.44800539585814536,
+      "learning_rate": 2.766922231287294e-06,
+      "loss": 0.5512,
+      "step": 10582
+    },
+    {
+      "epoch": 2.8095048453471394,
+      "grad_norm": 0.43799287921852015,
+      "learning_rate": 2.766575096430435e-06,
+      "loss": 0.5443,
+      "step": 10583
+    },
+    {
+      "epoch": 2.809770343820523,
+      "grad_norm": 0.44111341464566506,
+      "learning_rate": 2.76622795637471e-06,
+      "loss": 0.5682,
+      "step": 10584
+    },
+    {
+      "epoch": 2.8100358422939067,
+      "grad_norm": 0.4338807378969649,
+      "learning_rate": 2.7658808111268877e-06,
+      "loss": 0.5802,
+      "step": 10585
+    },
+    {
+      "epoch": 2.8103013407672908,
+      "grad_norm": 0.44332913970194315,
+      "learning_rate": 2.7655336606937384e-06,
+      "loss": 0.5451,
+      "step": 10586
+    },
+    {
+      "epoch": 2.8105668392406744,
+      "grad_norm": 0.45022207017053134,
+      "learning_rate": 2.7651865050820325e-06,
+      "loss": 0.5856,
+      "step": 10587
+    },
+    {
+      "epoch": 2.810832337714058,
+      "grad_norm": 0.4496839119812933,
+      "learning_rate": 2.7648393442985403e-06,
+      "loss": 0.5551,
+      "step": 10588
+    },
+    {
+      "epoch": 2.811097836187442,
+      "grad_norm": 0.4451370812827847,
+      "learning_rate": 2.7644921783500336e-06,
+      "loss": 0.5671,
+      "step": 10589
+    },
+    {
+      "epoch": 2.8113633346608258,
+      "grad_norm": 0.45048299410558373,
+      "learning_rate": 2.7641450072432795e-06,
+      "loss": 0.5605,
+      "step": 10590
+    },
+    {
+      "epoch": 2.8116288331342094,
+      "grad_norm": 0.4367601271746524,
+      "learning_rate": 2.763797830985053e-06,
+      "loss": 0.5944,
+      "step": 10591
+    },
+    {
+      "epoch": 2.8118943316075935,
+      "grad_norm": 0.4378956555887698,
+      "learning_rate": 2.7634506495821225e-06,
+      "loss": 0.5504,
+      "step": 10592
+    },
+    {
+      "epoch": 2.812159830080977,
+      "grad_norm": 0.444979611795338,
+      "learning_rate": 2.763103463041259e-06,
+      "loss": 0.5364,
+      "step": 10593
+    },
+    {
+      "epoch": 2.812425328554361,
+      "grad_norm": 0.44129340232543635,
+      "learning_rate": 2.7627562713692334e-06,
+      "loss": 0.5409,
+      "step": 10594
+    },
+    {
+      "epoch": 2.812690827027745,
+      "grad_norm": 0.44016923989723683,
+      "learning_rate": 2.7624090745728184e-06,
+      "loss": 0.5691,
+      "step": 10595
+    },
+    {
+      "epoch": 2.8129563255011285,
+      "grad_norm": 0.43654130244920164,
+      "learning_rate": 2.7620618726587832e-06,
+      "loss": 0.5718,
+      "step": 10596
+    },
+    {
+      "epoch": 2.813221823974512,
+      "grad_norm": 0.4273729110499703,
+      "learning_rate": 2.7617146656338995e-06,
+      "loss": 0.5603,
+      "step": 10597
+    },
+    {
+      "epoch": 2.813487322447896,
+      "grad_norm": 0.4520871551321622,
+      "learning_rate": 2.76136745350494e-06,
+      "loss": 0.5923,
+      "step": 10598
+    },
+    {
+      "epoch": 2.8137528209212794,
+      "grad_norm": 0.4451472313663982,
+      "learning_rate": 2.7610202362786743e-06,
+      "loss": 0.5816,
+      "step": 10599
+    },
+    {
+      "epoch": 2.8140183193946635,
+      "grad_norm": 0.45006732120869,
+      "learning_rate": 2.760673013961875e-06,
+      "loss": 0.5359,
+      "step": 10600
+    },
+    {
+      "epoch": 2.814283817868047,
+      "grad_norm": 0.45882437981787233,
+      "learning_rate": 2.760325786561314e-06,
+      "loss": 0.5354,
+      "step": 10601
+    },
+    {
+      "epoch": 2.814549316341431,
+      "grad_norm": 0.4410479847607938,
+      "learning_rate": 2.7599785540837624e-06,
+      "loss": 0.5213,
+      "step": 10602
+    },
+    {
+      "epoch": 2.814814814814815,
+      "grad_norm": 0.4358462493118188,
+      "learning_rate": 2.7596313165359926e-06,
+      "loss": 0.5556,
+      "step": 10603
+    },
+    {
+      "epoch": 2.8150803132881985,
+      "grad_norm": 0.4628377559891519,
+      "learning_rate": 2.759284073924777e-06,
+      "loss": 0.5763,
+      "step": 10604
+    },
+    {
+      "epoch": 2.815345811761582,
+      "grad_norm": 0.44226971996897274,
+      "learning_rate": 2.758936826256887e-06,
+      "loss": 0.5389,
+      "step": 10605
+    },
+    {
+      "epoch": 2.8156113102349662,
+      "grad_norm": 0.4557947426164022,
+      "learning_rate": 2.7585895735390945e-06,
+      "loss": 0.5746,
+      "step": 10606
+    },
+    {
+      "epoch": 2.81587680870835,
+      "grad_norm": 0.4436909232556306,
+      "learning_rate": 2.7582423157781725e-06,
+      "loss": 0.5735,
+      "step": 10607
+    },
+    {
+      "epoch": 2.8161423071817335,
+      "grad_norm": 0.4513578264659114,
+      "learning_rate": 2.7578950529808927e-06,
+      "loss": 0.5526,
+      "step": 10608
+    },
+    {
+      "epoch": 2.8164078056551176,
+      "grad_norm": 0.43611830729237494,
+      "learning_rate": 2.757547785154028e-06,
+      "loss": 0.5813,
+      "step": 10609
+    },
+    {
+      "epoch": 2.8166733041285013,
+      "grad_norm": 0.4416523485895305,
+      "learning_rate": 2.757200512304351e-06,
+      "loss": 0.5671,
+      "step": 10610
+    },
+    {
+      "epoch": 2.816938802601885,
+      "grad_norm": 0.45280427706140286,
+      "learning_rate": 2.756853234438634e-06,
+      "loss": 0.5884,
+      "step": 10611
+    },
+    {
+      "epoch": 2.817204301075269,
+      "grad_norm": 0.45190705040783713,
+      "learning_rate": 2.7565059515636503e-06,
+      "loss": 0.5268,
+      "step": 10612
+    },
+    {
+      "epoch": 2.8174697995486526,
+      "grad_norm": 0.4452826536415397,
+      "learning_rate": 2.756158663686173e-06,
+      "loss": 0.5808,
+      "step": 10613
+    },
+    {
+      "epoch": 2.8177352980220363,
+      "grad_norm": 0.440392837647871,
+      "learning_rate": 2.755811370812974e-06,
+      "loss": 0.5502,
+      "step": 10614
+    },
+    {
+      "epoch": 2.8180007964954203,
+      "grad_norm": 0.436849901388548,
+      "learning_rate": 2.7554640729508273e-06,
+      "loss": 0.5335,
+      "step": 10615
+    },
+    {
+      "epoch": 2.818266294968804,
+      "grad_norm": 0.4418294719857132,
+      "learning_rate": 2.7551167701065052e-06,
+      "loss": 0.6036,
+      "step": 10616
+    },
+    {
+      "epoch": 2.8185317934421876,
+      "grad_norm": 0.4408439808679575,
+      "learning_rate": 2.7547694622867815e-06,
+      "loss": 0.588,
+      "step": 10617
+    },
+    {
+      "epoch": 2.8187972919155717,
+      "grad_norm": 0.4221400608339185,
+      "learning_rate": 2.75442214949843e-06,
+      "loss": 0.5193,
+      "step": 10618
+    },
+    {
+      "epoch": 2.8190627903889554,
+      "grad_norm": 0.42818284074166924,
+      "learning_rate": 2.754074831748223e-06,
+      "loss": 0.5349,
+      "step": 10619
+    },
+    {
+      "epoch": 2.819328288862339,
+      "grad_norm": 0.43532512602667456,
+      "learning_rate": 2.753727509042935e-06,
+      "loss": 0.5498,
+      "step": 10620
+    },
+    {
+      "epoch": 2.819593787335723,
+      "grad_norm": 0.42561874656225274,
+      "learning_rate": 2.7533801813893397e-06,
+      "loss": 0.5417,
+      "step": 10621
+    },
+    {
+      "epoch": 2.8198592858091067,
+      "grad_norm": 0.44509111314623134,
+      "learning_rate": 2.7530328487942097e-06,
+      "loss": 0.5784,
+      "step": 10622
+    },
+    {
+      "epoch": 2.8201247842824904,
+      "grad_norm": 0.42963079375850155,
+      "learning_rate": 2.7526855112643204e-06,
+      "loss": 0.5543,
+      "step": 10623
+    },
+    {
+      "epoch": 2.820390282755874,
+      "grad_norm": 0.43517236755634625,
+      "learning_rate": 2.7523381688064447e-06,
+      "loss": 0.5578,
+      "step": 10624
+    },
+    {
+      "epoch": 2.820655781229258,
+      "grad_norm": 0.4508145134821779,
+      "learning_rate": 2.7519908214273566e-06,
+      "loss": 0.5592,
+      "step": 10625
+    },
+    {
+      "epoch": 2.8209212797026417,
+      "grad_norm": 0.4348584154929015,
+      "learning_rate": 2.7516434691338304e-06,
+      "loss": 0.5543,
+      "step": 10626
+    },
+    {
+      "epoch": 2.8211867781760254,
+      "grad_norm": 0.4421029513540858,
+      "learning_rate": 2.751296111932641e-06,
+      "loss": 0.5499,
+      "step": 10627
+    },
+    {
+      "epoch": 2.821452276649409,
+      "grad_norm": 0.4484014381364127,
+      "learning_rate": 2.7509487498305615e-06,
+      "loss": 0.5611,
+      "step": 10628
+    },
+    {
+      "epoch": 2.821717775122793,
+      "grad_norm": 0.44609023722452157,
+      "learning_rate": 2.7506013828343674e-06,
+      "loss": 0.555,
+      "step": 10629
+    },
+    {
+      "epoch": 2.8219832735961767,
+      "grad_norm": 0.4359919685918703,
+      "learning_rate": 2.750254010950833e-06,
+      "loss": 0.5219,
+      "step": 10630
+    },
+    {
+      "epoch": 2.8222487720695604,
+      "grad_norm": 0.4467371403472016,
+      "learning_rate": 2.749906634186732e-06,
+      "loss": 0.5298,
+      "step": 10631
+    },
+    {
+      "epoch": 2.8225142705429445,
+      "grad_norm": 0.4426489507273086,
+      "learning_rate": 2.74955925254884e-06,
+      "loss": 0.5474,
+      "step": 10632
+    },
+    {
+      "epoch": 2.822779769016328,
+      "grad_norm": 0.4437186878869067,
+      "learning_rate": 2.749211866043932e-06,
+      "loss": 0.5577,
+      "step": 10633
+    },
+    {
+      "epoch": 2.8230452674897117,
+      "grad_norm": 0.4447663690369507,
+      "learning_rate": 2.748864474678782e-06,
+      "loss": 0.548,
+      "step": 10634
+    },
+    {
+      "epoch": 2.823310765963096,
+      "grad_norm": 0.44371540080373734,
+      "learning_rate": 2.7485170784601654e-06,
+      "loss": 0.5129,
+      "step": 10635
+    },
+    {
+      "epoch": 2.8235762644364795,
+      "grad_norm": 0.43376039298415986,
+      "learning_rate": 2.7481696773948573e-06,
+      "loss": 0.5481,
+      "step": 10636
+    },
+    {
+      "epoch": 2.823841762909863,
+      "grad_norm": 0.43534334528229096,
+      "learning_rate": 2.747822271489633e-06,
+      "loss": 0.5398,
+      "step": 10637
+    },
+    {
+      "epoch": 2.824107261383247,
+      "grad_norm": 0.44992331876505093,
+      "learning_rate": 2.747474860751268e-06,
+      "loss": 0.5866,
+      "step": 10638
+    },
+    {
+      "epoch": 2.824372759856631,
+      "grad_norm": 0.4254682402789493,
+      "learning_rate": 2.7471274451865364e-06,
+      "loss": 0.5721,
+      "step": 10639
+    },
+    {
+      "epoch": 2.8246382583300145,
+      "grad_norm": 0.44177179167206426,
+      "learning_rate": 2.746780024802215e-06,
+      "loss": 0.5404,
+      "step": 10640
+    },
+    {
+      "epoch": 2.8249037568033986,
+      "grad_norm": 0.4329871911643497,
+      "learning_rate": 2.7464325996050795e-06,
+      "loss": 0.5837,
+      "step": 10641
+    },
+    {
+      "epoch": 2.825169255276782,
+      "grad_norm": 0.4348738002248126,
+      "learning_rate": 2.7460851696019043e-06,
+      "loss": 0.5609,
+      "step": 10642
+    },
+    {
+      "epoch": 2.825434753750166,
+      "grad_norm": 0.4483380639457842,
+      "learning_rate": 2.745737734799465e-06,
+      "loss": 0.548,
+      "step": 10643
+    },
+    {
+      "epoch": 2.82570025222355,
+      "grad_norm": 0.42716177939380223,
+      "learning_rate": 2.74539029520454e-06,
+      "loss": 0.5117,
+      "step": 10644
+    },
+    {
+      "epoch": 2.8259657506969336,
+      "grad_norm": 0.45642103432381975,
+      "learning_rate": 2.7450428508239024e-06,
+      "loss": 0.5116,
+      "step": 10645
+    },
+    {
+      "epoch": 2.826231249170317,
+      "grad_norm": 0.4396893283828446,
+      "learning_rate": 2.7446954016643283e-06,
+      "loss": 0.5614,
+      "step": 10646
+    },
+    {
+      "epoch": 2.8264967476437013,
+      "grad_norm": 0.42226040100124007,
+      "learning_rate": 2.7443479477325964e-06,
+      "loss": 0.5723,
+      "step": 10647
+    },
+    {
+      "epoch": 2.826762246117085,
+      "grad_norm": 0.4346957939485996,
+      "learning_rate": 2.7440004890354804e-06,
+      "loss": 0.5772,
+      "step": 10648
+    },
+    {
+      "epoch": 2.8270277445904686,
+      "grad_norm": 0.4236382257836881,
+      "learning_rate": 2.743653025579757e-06,
+      "loss": 0.5057,
+      "step": 10649
+    },
+    {
+      "epoch": 2.8272932430638527,
+      "grad_norm": 0.4473666579567541,
+      "learning_rate": 2.7433055573722047e-06,
+      "loss": 0.5705,
+      "step": 10650
+    },
+    {
+      "epoch": 2.8275587415372363,
+      "grad_norm": 0.42682423423317256,
+      "learning_rate": 2.742958084419597e-06,
+      "loss": 0.5693,
+      "step": 10651
+    },
+    {
+      "epoch": 2.82782424001062,
+      "grad_norm": 0.4477324596604937,
+      "learning_rate": 2.7426106067287117e-06,
+      "loss": 0.5797,
+      "step": 10652
+    },
+    {
+      "epoch": 2.8280897384840036,
+      "grad_norm": 0.44378764045073016,
+      "learning_rate": 2.7422631243063265e-06,
+      "loss": 0.5419,
+      "step": 10653
+    },
+    {
+      "epoch": 2.828355236957387,
+      "grad_norm": 0.4464112571700783,
+      "learning_rate": 2.7419156371592166e-06,
+      "loss": 0.5737,
+      "step": 10654
+    },
+    {
+      "epoch": 2.8286207354307713,
+      "grad_norm": 0.4587008636993577,
+      "learning_rate": 2.7415681452941594e-06,
+      "loss": 0.5157,
+      "step": 10655
+    },
+    {
+      "epoch": 2.828886233904155,
+      "grad_norm": 0.44779470980037994,
+      "learning_rate": 2.7412206487179316e-06,
+      "loss": 0.5667,
+      "step": 10656
+    },
+    {
+      "epoch": 2.8291517323775386,
+      "grad_norm": 0.448422829382189,
+      "learning_rate": 2.7408731474373112e-06,
+      "loss": 0.5623,
+      "step": 10657
+    },
+    {
+      "epoch": 2.8294172308509227,
+      "grad_norm": 0.4437055962953726,
+      "learning_rate": 2.740525641459075e-06,
+      "loss": 0.5845,
+      "step": 10658
+    },
+    {
+      "epoch": 2.8296827293243063,
+      "grad_norm": 0.45619250344956913,
+      "learning_rate": 2.740178130789999e-06,
+      "loss": 0.5862,
+      "step": 10659
+    },
+    {
+      "epoch": 2.82994822779769,
+      "grad_norm": 0.4539452892278317,
+      "learning_rate": 2.7398306154368613e-06,
+      "loss": 0.5831,
+      "step": 10660
+    },
+    {
+      "epoch": 2.830213726271074,
+      "grad_norm": 0.4270186235202641,
+      "learning_rate": 2.7394830954064407e-06,
+      "loss": 0.5547,
+      "step": 10661
+    },
+    {
+      "epoch": 2.8304792247444577,
+      "grad_norm": 0.44383344157764537,
+      "learning_rate": 2.7391355707055118e-06,
+      "loss": 0.5884,
+      "step": 10662
+    },
+    {
+      "epoch": 2.8307447232178413,
+      "grad_norm": 0.4409802315806133,
+      "learning_rate": 2.7387880413408546e-06,
+      "loss": 0.5531,
+      "step": 10663
+    },
+    {
+      "epoch": 2.8310102216912254,
+      "grad_norm": 0.4421780167994675,
+      "learning_rate": 2.7384405073192454e-06,
+      "loss": 0.5678,
+      "step": 10664
+    },
+    {
+      "epoch": 2.831275720164609,
+      "grad_norm": 0.43736836288697917,
+      "learning_rate": 2.7380929686474627e-06,
+      "loss": 0.5518,
+      "step": 10665
+    },
+    {
+      "epoch": 2.8315412186379927,
+      "grad_norm": 0.44415480322706846,
+      "learning_rate": 2.7377454253322842e-06,
+      "loss": 0.5953,
+      "step": 10666
+    },
+    {
+      "epoch": 2.8318067171113768,
+      "grad_norm": 0.46039082399154857,
+      "learning_rate": 2.737397877380489e-06,
+      "loss": 0.5551,
+      "step": 10667
+    },
+    {
+      "epoch": 2.8320722155847604,
+      "grad_norm": 0.4320254677483836,
+      "learning_rate": 2.7370503247988523e-06,
+      "loss": 0.5235,
+      "step": 10668
+    },
+    {
+      "epoch": 2.832337714058144,
+      "grad_norm": 0.43583894508457366,
+      "learning_rate": 2.736702767594154e-06,
+      "loss": 0.5268,
+      "step": 10669
+    },
+    {
+      "epoch": 2.832603212531528,
+      "grad_norm": 0.43627043861537007,
+      "learning_rate": 2.7363552057731736e-06,
+      "loss": 0.5681,
+      "step": 10670
+    },
+    {
+      "epoch": 2.8328687110049118,
+      "grad_norm": 0.450135259881002,
+      "learning_rate": 2.7360076393426867e-06,
+      "loss": 0.5459,
+      "step": 10671
+    },
+    {
+      "epoch": 2.8331342094782954,
+      "grad_norm": 0.47295130059901647,
+      "learning_rate": 2.735660068309473e-06,
+      "loss": 0.5697,
+      "step": 10672
+    },
+    {
+      "epoch": 2.8333997079516795,
+      "grad_norm": 0.4513715877264598,
+      "learning_rate": 2.7353124926803113e-06,
+      "loss": 0.5937,
+      "step": 10673
+    },
+    {
+      "epoch": 2.833665206425063,
+      "grad_norm": 0.44250118610240136,
+      "learning_rate": 2.7349649124619797e-06,
+      "loss": 0.5773,
+      "step": 10674
+    },
+    {
+      "epoch": 2.833930704898447,
+      "grad_norm": 0.4393591288004187,
+      "learning_rate": 2.734617327661257e-06,
+      "loss": 0.5727,
+      "step": 10675
+    },
+    {
+      "epoch": 2.834196203371831,
+      "grad_norm": 0.4464833214164496,
+      "learning_rate": 2.7342697382849228e-06,
+      "loss": 0.5437,
+      "step": 10676
+    },
+    {
+      "epoch": 2.8344617018452145,
+      "grad_norm": 0.445081832161792,
+      "learning_rate": 2.733922144339754e-06,
+      "loss": 0.5344,
+      "step": 10677
+    },
+    {
+      "epoch": 2.834727200318598,
+      "grad_norm": 0.4438103958959221,
+      "learning_rate": 2.7335745458325307e-06,
+      "loss": 0.5535,
+      "step": 10678
+    },
+    {
+      "epoch": 2.834992698791982,
+      "grad_norm": 0.4459330265843424,
+      "learning_rate": 2.733226942770033e-06,
+      "loss": 0.576,
+      "step": 10679
+    },
+    {
+      "epoch": 2.835258197265366,
+      "grad_norm": 0.4410880053471279,
+      "learning_rate": 2.732879335159038e-06,
+      "loss": 0.5907,
+      "step": 10680
+    },
+    {
+      "epoch": 2.8355236957387495,
+      "grad_norm": 0.4394805707102283,
+      "learning_rate": 2.7325317230063257e-06,
+      "loss": 0.549,
+      "step": 10681
+    },
+    {
+      "epoch": 2.835789194212133,
+      "grad_norm": 0.46052245355471666,
+      "learning_rate": 2.7321841063186754e-06,
+      "loss": 0.5839,
+      "step": 10682
+    },
+    {
+      "epoch": 2.836054692685517,
+      "grad_norm": 0.45082112542932135,
+      "learning_rate": 2.7318364851028666e-06,
+      "loss": 0.5412,
+      "step": 10683
+    },
+    {
+      "epoch": 2.836320191158901,
+      "grad_norm": 0.4446068158667599,
+      "learning_rate": 2.7314888593656797e-06,
+      "loss": 0.5673,
+      "step": 10684
+    },
+    {
+      "epoch": 2.8365856896322845,
+      "grad_norm": 0.43810989544211665,
+      "learning_rate": 2.7311412291138924e-06,
+      "loss": 0.5668,
+      "step": 10685
+    },
+    {
+      "epoch": 2.836851188105668,
+      "grad_norm": 0.4506932297462287,
+      "learning_rate": 2.730793594354285e-06,
+      "loss": 0.56,
+      "step": 10686
+    },
+    {
+      "epoch": 2.8371166865790522,
+      "grad_norm": 0.45080786409307233,
+      "learning_rate": 2.730445955093638e-06,
+      "loss": 0.5432,
+      "step": 10687
+    },
+    {
+      "epoch": 2.837382185052436,
+      "grad_norm": 0.4430069968994358,
+      "learning_rate": 2.7300983113387304e-06,
+      "loss": 0.552,
+      "step": 10688
+    },
+    {
+      "epoch": 2.8376476835258195,
+      "grad_norm": 0.4351373689010535,
+      "learning_rate": 2.7297506630963423e-06,
+      "loss": 0.5369,
+      "step": 10689
+    },
+    {
+      "epoch": 2.8379131819992036,
+      "grad_norm": 0.4317588986923279,
+      "learning_rate": 2.7294030103732537e-06,
+      "loss": 0.5732,
+      "step": 10690
+    },
+    {
+      "epoch": 2.8381786804725873,
+      "grad_norm": 0.45169959954142747,
+      "learning_rate": 2.7290553531762442e-06,
+      "loss": 0.5962,
+      "step": 10691
+    },
+    {
+      "epoch": 2.838444178945971,
+      "grad_norm": 0.4649722622860929,
+      "learning_rate": 2.7287076915120952e-06,
+      "loss": 0.5397,
+      "step": 10692
+    },
+    {
+      "epoch": 2.838709677419355,
+      "grad_norm": 0.4432483580745817,
+      "learning_rate": 2.7283600253875863e-06,
+      "loss": 0.5816,
+      "step": 10693
+    },
+    {
+      "epoch": 2.8389751758927386,
+      "grad_norm": 0.4404069548426601,
+      "learning_rate": 2.728012354809497e-06,
+      "loss": 0.5864,
+      "step": 10694
+    },
+    {
+      "epoch": 2.8392406743661223,
+      "grad_norm": 0.43347988161478124,
+      "learning_rate": 2.7276646797846096e-06,
+      "loss": 0.5278,
+      "step": 10695
+    },
+    {
+      "epoch": 2.8395061728395063,
+      "grad_norm": 0.4334464892160729,
+      "learning_rate": 2.7273170003197037e-06,
+      "loss": 0.5717,
+      "step": 10696
+    },
+    {
+      "epoch": 2.83977167131289,
+      "grad_norm": 0.4361039324888806,
+      "learning_rate": 2.726969316421558e-06,
+      "loss": 0.5205,
+      "step": 10697
+    },
+    {
+      "epoch": 2.8400371697862736,
+      "grad_norm": 0.42857044148210877,
+      "learning_rate": 2.7266216280969564e-06,
+      "loss": 0.5231,
+      "step": 10698
+    },
+    {
+      "epoch": 2.8403026682596577,
+      "grad_norm": 0.425869053781619,
+      "learning_rate": 2.726273935352678e-06,
+      "loss": 0.4966,
+      "step": 10699
+    },
+    {
+      "epoch": 2.8405681667330414,
+      "grad_norm": 0.4414217548627363,
+      "learning_rate": 2.725926238195504e-06,
+      "loss": 0.5761,
+      "step": 10700
+    },
+    {
+      "epoch": 2.840833665206425,
+      "grad_norm": 0.4397669819776329,
+      "learning_rate": 2.7255785366322134e-06,
+      "loss": 0.5731,
+      "step": 10701
+    },
+    {
+      "epoch": 2.841099163679809,
+      "grad_norm": 0.44793687205104676,
+      "learning_rate": 2.7252308306695914e-06,
+      "loss": 0.6089,
+      "step": 10702
+    },
+    {
+      "epoch": 2.8413646621531927,
+      "grad_norm": 0.44313943155049007,
+      "learning_rate": 2.7248831203144154e-06,
+      "loss": 0.561,
+      "step": 10703
+    },
+    {
+      "epoch": 2.8416301606265764,
+      "grad_norm": 0.44781255407094406,
+      "learning_rate": 2.724535405573468e-06,
+      "loss": 0.5778,
+      "step": 10704
+    },
+    {
+      "epoch": 2.8418956590999604,
+      "grad_norm": 0.4458138437268211,
+      "learning_rate": 2.724187686453531e-06,
+      "loss": 0.6014,
+      "step": 10705
+    },
+    {
+      "epoch": 2.842161157573344,
+      "grad_norm": 0.4396287078028262,
+      "learning_rate": 2.723839962961385e-06,
+      "loss": 0.5377,
+      "step": 10706
+    },
+    {
+      "epoch": 2.8424266560467277,
+      "grad_norm": 0.43263935211202564,
+      "learning_rate": 2.7234922351038117e-06,
+      "loss": 0.5612,
+      "step": 10707
+    },
+    {
+      "epoch": 2.8426921545201114,
+      "grad_norm": 0.449640752268622,
+      "learning_rate": 2.7231445028875924e-06,
+      "loss": 0.5506,
+      "step": 10708
+    },
+    {
+      "epoch": 2.842957652993495,
+      "grad_norm": 0.4425907480121384,
+      "learning_rate": 2.7227967663195092e-06,
+      "loss": 0.5848,
+      "step": 10709
+    },
+    {
+      "epoch": 2.843223151466879,
+      "grad_norm": 0.44000747206345897,
+      "learning_rate": 2.7224490254063436e-06,
+      "loss": 0.5847,
+      "step": 10710
+    },
+    {
+      "epoch": 2.8434886499402627,
+      "grad_norm": 0.44073374703708895,
+      "learning_rate": 2.7221012801548767e-06,
+      "loss": 0.5241,
+      "step": 10711
+    },
+    {
+      "epoch": 2.8437541484136464,
+      "grad_norm": 0.43607825742065365,
+      "learning_rate": 2.7217535305718914e-06,
+      "loss": 0.5619,
+      "step": 10712
+    },
+    {
+      "epoch": 2.8440196468870305,
+      "grad_norm": 0.44613009648645197,
+      "learning_rate": 2.72140577666417e-06,
+      "loss": 0.523,
+      "step": 10713
+    },
+    {
+      "epoch": 2.844285145360414,
+      "grad_norm": 0.4429421186037967,
+      "learning_rate": 2.7210580184384933e-06,
+      "loss": 0.5506,
+      "step": 10714
+    },
+    {
+      "epoch": 2.8445506438337977,
+      "grad_norm": 0.435176668911199,
+      "learning_rate": 2.7207102559016434e-06,
+      "loss": 0.579,
+      "step": 10715
+    },
+    {
+      "epoch": 2.844816142307182,
+      "grad_norm": 0.4446755129678316,
+      "learning_rate": 2.720362489060404e-06,
+      "loss": 0.5561,
+      "step": 10716
+    },
+    {
+      "epoch": 2.8450816407805655,
+      "grad_norm": 0.4681986655281571,
+      "learning_rate": 2.7200147179215563e-06,
+      "loss": 0.5488,
+      "step": 10717
+    },
+    {
+      "epoch": 2.845347139253949,
+      "grad_norm": 0.44203361895308124,
+      "learning_rate": 2.7196669424918828e-06,
+      "loss": 0.5782,
+      "step": 10718
+    },
+    {
+      "epoch": 2.845612637727333,
+      "grad_norm": 0.4502913671114277,
+      "learning_rate": 2.7193191627781656e-06,
+      "loss": 0.5982,
+      "step": 10719
+    },
+    {
+      "epoch": 2.845878136200717,
+      "grad_norm": 0.4475324941183209,
+      "learning_rate": 2.7189713787871887e-06,
+      "loss": 0.5543,
+      "step": 10720
+    },
+    {
+      "epoch": 2.8461436346741005,
+      "grad_norm": 0.4502905979729806,
+      "learning_rate": 2.718623590525733e-06,
+      "loss": 0.5787,
+      "step": 10721
+    },
+    {
+      "epoch": 2.8464091331474846,
+      "grad_norm": 0.42784602819205586,
+      "learning_rate": 2.7182757980005823e-06,
+      "loss": 0.5562,
+      "step": 10722
+    },
+    {
+      "epoch": 2.846674631620868,
+      "grad_norm": 0.44338859926273616,
+      "learning_rate": 2.7179280012185193e-06,
+      "loss": 0.5358,
+      "step": 10723
+    },
+    {
+      "epoch": 2.846940130094252,
+      "grad_norm": 0.45209777829945,
+      "learning_rate": 2.7175802001863267e-06,
+      "loss": 0.5828,
+      "step": 10724
+    },
+    {
+      "epoch": 2.847205628567636,
+      "grad_norm": 0.4532088818726932,
+      "learning_rate": 2.7172323949107874e-06,
+      "loss": 0.5591,
+      "step": 10725
+    },
+    {
+      "epoch": 2.8474711270410196,
+      "grad_norm": 0.4552617692595368,
+      "learning_rate": 2.716884585398684e-06,
+      "loss": 0.5327,
+      "step": 10726
+    },
+    {
+      "epoch": 2.847736625514403,
+      "grad_norm": 0.42108293350760584,
+      "learning_rate": 2.716536771656801e-06,
+      "loss": 0.5601,
+      "step": 10727
+    },
+    {
+      "epoch": 2.8480021239877873,
+      "grad_norm": 0.4478831595439803,
+      "learning_rate": 2.71618895369192e-06,
+      "loss": 0.5083,
+      "step": 10728
+    },
+    {
+      "epoch": 2.848267622461171,
+      "grad_norm": 0.44615603413500154,
+      "learning_rate": 2.7158411315108254e-06,
+      "loss": 0.5794,
+      "step": 10729
+    },
+    {
+      "epoch": 2.8485331209345546,
+      "grad_norm": 0.445896856924272,
+      "learning_rate": 2.715493305120301e-06,
+      "loss": 0.5812,
+      "step": 10730
+    },
+    {
+      "epoch": 2.8487986194079387,
+      "grad_norm": 0.43854884183833776,
+      "learning_rate": 2.715145474527128e-06,
+      "loss": 0.547,
+      "step": 10731
+    },
+    {
+      "epoch": 2.8490641178813223,
+      "grad_norm": 0.4458427547571963,
+      "learning_rate": 2.7147976397380925e-06,
+      "loss": 0.5703,
+      "step": 10732
+    },
+    {
+      "epoch": 2.849329616354706,
+      "grad_norm": 0.4365011106527827,
+      "learning_rate": 2.7144498007599775e-06,
+      "loss": 0.5851,
+      "step": 10733
+    },
+    {
+      "epoch": 2.8495951148280896,
+      "grad_norm": 0.43704255022741134,
+      "learning_rate": 2.7141019575995657e-06,
+      "loss": 0.552,
+      "step": 10734
+    },
+    {
+      "epoch": 2.8498606133014737,
+      "grad_norm": 0.4548942591174282,
+      "learning_rate": 2.713754110263641e-06,
+      "loss": 0.6036,
+      "step": 10735
+    },
+    {
+      "epoch": 2.8501261117748573,
+      "grad_norm": 0.44620037362776677,
+      "learning_rate": 2.713406258758989e-06,
+      "loss": 0.5559,
+      "step": 10736
+    },
+    {
+      "epoch": 2.850391610248241,
+      "grad_norm": 0.4358700255720994,
+      "learning_rate": 2.7130584030923913e-06,
+      "loss": 0.5742,
+      "step": 10737
+    },
+    {
+      "epoch": 2.8506571087216246,
+      "grad_norm": 0.4375251376284897,
+      "learning_rate": 2.7127105432706335e-06,
+      "loss": 0.5173,
+      "step": 10738
+    },
+    {
+      "epoch": 2.8509226071950087,
+      "grad_norm": 0.4354691859850981,
+      "learning_rate": 2.7123626793004997e-06,
+      "loss": 0.5564,
+      "step": 10739
+    },
+    {
+      "epoch": 2.8511881056683923,
+      "grad_norm": 0.43722452970994985,
+      "learning_rate": 2.7120148111887732e-06,
+      "loss": 0.5931,
+      "step": 10740
+    },
+    {
+      "epoch": 2.851453604141776,
+      "grad_norm": 0.4454867063980965,
+      "learning_rate": 2.711666938942239e-06,
+      "loss": 0.5217,
+      "step": 10741
+    },
+    {
+      "epoch": 2.85171910261516,
+      "grad_norm": 0.44099269311670675,
+      "learning_rate": 2.7113190625676816e-06,
+      "loss": 0.5609,
+      "step": 10742
+    },
+    {
+      "epoch": 2.8519846010885437,
+      "grad_norm": 0.43623175418248467,
+      "learning_rate": 2.710971182071884e-06,
+      "loss": 0.5685,
+      "step": 10743
+    },
+    {
+      "epoch": 2.8522500995619273,
+      "grad_norm": 0.4384356765458919,
+      "learning_rate": 2.710623297461633e-06,
+      "loss": 0.5194,
+      "step": 10744
+    },
+    {
+      "epoch": 2.8525155980353114,
+      "grad_norm": 0.4442303572050867,
+      "learning_rate": 2.7102754087437117e-06,
+      "loss": 0.5547,
+      "step": 10745
+    },
+    {
+      "epoch": 2.852781096508695,
+      "grad_norm": 0.44960201823576745,
+      "learning_rate": 2.709927515924905e-06,
+      "loss": 0.5857,
+      "step": 10746
+    },
+    {
+      "epoch": 2.8530465949820787,
+      "grad_norm": 0.45029950101068483,
+      "learning_rate": 2.7095796190119967e-06,
+      "loss": 0.5279,
+      "step": 10747
+    },
+    {
+      "epoch": 2.8533120934554628,
+      "grad_norm": 0.436445103278809,
+      "learning_rate": 2.709231718011774e-06,
+      "loss": 0.5582,
+      "step": 10748
+    },
+    {
+      "epoch": 2.8535775919288464,
+      "grad_norm": 0.44626933424831894,
+      "learning_rate": 2.7088838129310202e-06,
+      "loss": 0.5666,
+      "step": 10749
+    },
+    {
+      "epoch": 2.85384309040223,
+      "grad_norm": 0.4332365834903132,
+      "learning_rate": 2.7085359037765203e-06,
+      "loss": 0.5804,
+      "step": 10750
+    },
+    {
+      "epoch": 2.854108588875614,
+      "grad_norm": 0.45827132023683453,
+      "learning_rate": 2.7081879905550603e-06,
+      "loss": 0.5585,
+      "step": 10751
+    },
+    {
+      "epoch": 2.854374087348998,
+      "grad_norm": 0.44749313232517135,
+      "learning_rate": 2.7078400732734245e-06,
+      "loss": 0.5778,
+      "step": 10752
+    },
+    {
+      "epoch": 2.8546395858223814,
+      "grad_norm": 0.4510749864353565,
+      "learning_rate": 2.707492151938398e-06,
+      "loss": 0.5787,
+      "step": 10753
+    },
+    {
+      "epoch": 2.8549050842957655,
+      "grad_norm": 0.42752932002261185,
+      "learning_rate": 2.707144226556767e-06,
+      "loss": 0.5545,
+      "step": 10754
+    },
+    {
+      "epoch": 2.855170582769149,
+      "grad_norm": 0.43791650215467726,
+      "learning_rate": 2.706796297135316e-06,
+      "loss": 0.5458,
+      "step": 10755
+    },
+    {
+      "epoch": 2.855436081242533,
+      "grad_norm": 0.44658696973588446,
+      "learning_rate": 2.7064483636808314e-06,
+      "loss": 0.5472,
+      "step": 10756
+    },
+    {
+      "epoch": 2.855701579715917,
+      "grad_norm": 0.4688153715222747,
+      "learning_rate": 2.706100426200098e-06,
+      "loss": 0.5409,
+      "step": 10757
+    },
+    {
+      "epoch": 2.8559670781893005,
+      "grad_norm": 0.4427903144420845,
+      "learning_rate": 2.705752484699901e-06,
+      "loss": 0.5828,
+      "step": 10758
+    },
+    {
+      "epoch": 2.856232576662684,
+      "grad_norm": 0.4316400515050232,
+      "learning_rate": 2.7054045391870274e-06,
+      "loss": 0.5671,
+      "step": 10759
+    },
+    {
+      "epoch": 2.8564980751360682,
+      "grad_norm": 0.44803230039598874,
+      "learning_rate": 2.7050565896682625e-06,
+      "loss": 0.535,
+      "step": 10760
+    },
+    {
+      "epoch": 2.856763573609452,
+      "grad_norm": 0.4481997479866655,
+      "learning_rate": 2.7047086361503917e-06,
+      "loss": 0.5433,
+      "step": 10761
+    },
+    {
+      "epoch": 2.8570290720828355,
+      "grad_norm": 0.4348015755909405,
+      "learning_rate": 2.704360678640202e-06,
+      "loss": 0.5572,
+      "step": 10762
+    },
+    {
+      "epoch": 2.857294570556219,
+      "grad_norm": 0.4306697852027576,
+      "learning_rate": 2.704012717144478e-06,
+      "loss": 0.5387,
+      "step": 10763
+    },
+    {
+      "epoch": 2.8575600690296032,
+      "grad_norm": 0.4435402618330487,
+      "learning_rate": 2.7036647516700066e-06,
+      "loss": 0.5427,
+      "step": 10764
+    },
+    {
+      "epoch": 2.857825567502987,
+      "grad_norm": 0.45371648551094274,
+      "learning_rate": 2.703316782223574e-06,
+      "loss": 0.5864,
+      "step": 10765
+    },
+    {
+      "epoch": 2.8580910659763705,
+      "grad_norm": 0.44391622962162264,
+      "learning_rate": 2.702968808811966e-06,
+      "loss": 0.5602,
+      "step": 10766
+    },
+    {
+      "epoch": 2.858356564449754,
+      "grad_norm": 0.4429972812866065,
+      "learning_rate": 2.7026208314419695e-06,
+      "loss": 0.5853,
+      "step": 10767
+    },
+    {
+      "epoch": 2.8586220629231383,
+      "grad_norm": 0.43715014222390575,
+      "learning_rate": 2.702272850120371e-06,
+      "loss": 0.5555,
+      "step": 10768
+    },
+    {
+      "epoch": 2.858887561396522,
+      "grad_norm": 0.43733868211036886,
+      "learning_rate": 2.7019248648539565e-06,
+      "loss": 0.5148,
+      "step": 10769
+    },
+    {
+      "epoch": 2.8591530598699055,
+      "grad_norm": 0.4352574360301688,
+      "learning_rate": 2.7015768756495127e-06,
+      "loss": 0.5537,
+      "step": 10770
+    },
+    {
+      "epoch": 2.8594185583432896,
+      "grad_norm": 0.441906268408107,
+      "learning_rate": 2.701228882513826e-06,
+      "loss": 0.5618,
+      "step": 10771
+    },
+    {
+      "epoch": 2.8596840568166733,
+      "grad_norm": 0.43448023366429356,
+      "learning_rate": 2.700880885453684e-06,
+      "loss": 0.5545,
+      "step": 10772
+    },
+    {
+      "epoch": 2.859949555290057,
+      "grad_norm": 0.4359060938661836,
+      "learning_rate": 2.700532884475872e-06,
+      "loss": 0.5483,
+      "step": 10773
+    },
+    {
+      "epoch": 2.860215053763441,
+      "grad_norm": 0.45498619981843763,
+      "learning_rate": 2.7001848795871784e-06,
+      "loss": 0.5516,
+      "step": 10774
+    },
+    {
+      "epoch": 2.8604805522368246,
+      "grad_norm": 0.45281036419587767,
+      "learning_rate": 2.699836870794389e-06,
+      "loss": 0.6146,
+      "step": 10775
+    },
+    {
+      "epoch": 2.8607460507102083,
+      "grad_norm": 0.4556607934096583,
+      "learning_rate": 2.6994888581042915e-06,
+      "loss": 0.5602,
+      "step": 10776
+    },
+    {
+      "epoch": 2.8610115491835924,
+      "grad_norm": 0.44492451468883587,
+      "learning_rate": 2.699140841523674e-06,
+      "loss": 0.5634,
+      "step": 10777
+    },
+    {
+      "epoch": 2.861277047656976,
+      "grad_norm": 0.4278712653180521,
+      "learning_rate": 2.698792821059321e-06,
+      "loss": 0.5117,
+      "step": 10778
+    },
+    {
+      "epoch": 2.8615425461303596,
+      "grad_norm": 0.4495792879260062,
+      "learning_rate": 2.6984447967180228e-06,
+      "loss": 0.5577,
+      "step": 10779
+    },
+    {
+      "epoch": 2.8618080446037437,
+      "grad_norm": 0.4258108166799316,
+      "learning_rate": 2.698096768506564e-06,
+      "loss": 0.5296,
+      "step": 10780
+    },
+    {
+      "epoch": 2.8620735430771274,
+      "grad_norm": 0.4504538249286639,
+      "learning_rate": 2.697748736431733e-06,
+      "loss": 0.5654,
+      "step": 10781
+    },
+    {
+      "epoch": 2.862339041550511,
+      "grad_norm": 0.43014281297690643,
+      "learning_rate": 2.6974007005003176e-06,
+      "loss": 0.5518,
+      "step": 10782
+    },
+    {
+      "epoch": 2.862604540023895,
+      "grad_norm": 0.4605838312758772,
+      "learning_rate": 2.6970526607191053e-06,
+      "loss": 0.5807,
+      "step": 10783
+    },
+    {
+      "epoch": 2.8628700384972787,
+      "grad_norm": 0.431864417707121,
+      "learning_rate": 2.6967046170948843e-06,
+      "loss": 0.5726,
+      "step": 10784
+    },
+    {
+      "epoch": 2.8631355369706624,
+      "grad_norm": 0.4445798856846652,
+      "learning_rate": 2.6963565696344413e-06,
+      "loss": 0.5521,
+      "step": 10785
+    },
+    {
+      "epoch": 2.8634010354440464,
+      "grad_norm": 0.4404793748357265,
+      "learning_rate": 2.6960085183445644e-06,
+      "loss": 0.5597,
+      "step": 10786
+    },
+    {
+      "epoch": 2.86366653391743,
+      "grad_norm": 0.4525667450486386,
+      "learning_rate": 2.695660463232041e-06,
+      "loss": 0.5516,
+      "step": 10787
+    },
+    {
+      "epoch": 2.8639320323908137,
+      "grad_norm": 0.44162985586253817,
+      "learning_rate": 2.6953124043036604e-06,
+      "loss": 0.5496,
+      "step": 10788
+    },
+    {
+      "epoch": 2.8641975308641974,
+      "grad_norm": 0.45065809105279425,
+      "learning_rate": 2.694964341566209e-06,
+      "loss": 0.5594,
+      "step": 10789
+    },
+    {
+      "epoch": 2.8644630293375815,
+      "grad_norm": 0.4331856923687826,
+      "learning_rate": 2.694616275026476e-06,
+      "loss": 0.572,
+      "step": 10790
+    },
+    {
+      "epoch": 2.864728527810965,
+      "grad_norm": 0.44905390249869104,
+      "learning_rate": 2.6942682046912492e-06,
+      "loss": 0.5677,
+      "step": 10791
+    },
+    {
+      "epoch": 2.8649940262843487,
+      "grad_norm": 0.4459068620119701,
+      "learning_rate": 2.693920130567316e-06,
+      "loss": 0.5467,
+      "step": 10792
+    },
+    {
+      "epoch": 2.8652595247577324,
+      "grad_norm": 0.4447957581873441,
+      "learning_rate": 2.6935720526614656e-06,
+      "loss": 0.5658,
+      "step": 10793
+    },
+    {
+      "epoch": 2.8655250232311165,
+      "grad_norm": 0.44914334064867867,
+      "learning_rate": 2.693223970980487e-06,
+      "loss": 0.5707,
+      "step": 10794
+    },
+    {
+      "epoch": 2.8657905217045,
+      "grad_norm": 0.45208235913744277,
+      "learning_rate": 2.6928758855311678e-06,
+      "loss": 0.5969,
+      "step": 10795
+    },
+    {
+      "epoch": 2.8660560201778837,
+      "grad_norm": 0.45156122580289887,
+      "learning_rate": 2.6925277963202954e-06,
+      "loss": 0.5546,
+      "step": 10796
+    },
+    {
+      "epoch": 2.866321518651268,
+      "grad_norm": 0.4365754983809178,
+      "learning_rate": 2.6921797033546607e-06,
+      "loss": 0.5523,
+      "step": 10797
+    },
+    {
+      "epoch": 2.8665870171246515,
+      "grad_norm": 0.44211215855259733,
+      "learning_rate": 2.691831606641051e-06,
+      "loss": 0.571,
+      "step": 10798
+    },
+    {
+      "epoch": 2.866852515598035,
+      "grad_norm": 0.4531205492943909,
+      "learning_rate": 2.6914835061862554e-06,
+      "loss": 0.5866,
+      "step": 10799
+    },
+    {
+      "epoch": 2.867118014071419,
+      "grad_norm": 0.42162915272733076,
+      "learning_rate": 2.691135401997063e-06,
+      "loss": 0.5326,
+      "step": 10800
+    },
+    {
+      "epoch": 2.867383512544803,
+      "grad_norm": 0.4619213541114661,
+      "learning_rate": 2.6907872940802615e-06,
+      "loss": 0.5468,
+      "step": 10801
+    },
+    {
+      "epoch": 2.8676490110181865,
+      "grad_norm": 0.482958323474219,
+      "learning_rate": 2.6904391824426405e-06,
+      "loss": 0.5444,
+      "step": 10802
+    },
+    {
+      "epoch": 2.8679145094915706,
+      "grad_norm": 0.4347171768582086,
+      "learning_rate": 2.6900910670909897e-06,
+      "loss": 0.5312,
+      "step": 10803
+    },
+    {
+      "epoch": 2.868180007964954,
+      "grad_norm": 0.44191406865669913,
+      "learning_rate": 2.6897429480320973e-06,
+      "loss": 0.5593,
+      "step": 10804
+    },
+    {
+      "epoch": 2.868445506438338,
+      "grad_norm": 0.45942441108805737,
+      "learning_rate": 2.6893948252727537e-06,
+      "loss": 0.548,
+      "step": 10805
+    },
+    {
+      "epoch": 2.868711004911722,
+      "grad_norm": 0.43909599630631496,
+      "learning_rate": 2.689046698819746e-06,
+      "loss": 0.5415,
+      "step": 10806
+    },
+    {
+      "epoch": 2.8689765033851056,
+      "grad_norm": 0.46213125874228994,
+      "learning_rate": 2.6886985686798656e-06,
+      "loss": 0.5364,
+      "step": 10807
+    },
+    {
+      "epoch": 2.869242001858489,
+      "grad_norm": 0.44400531622663764,
+      "learning_rate": 2.6883504348599014e-06,
+      "loss": 0.5713,
+      "step": 10808
+    },
+    {
+      "epoch": 2.8695075003318733,
+      "grad_norm": 0.4521447547927035,
+      "learning_rate": 2.688002297366642e-06,
+      "loss": 0.5372,
+      "step": 10809
+    },
+    {
+      "epoch": 2.869772998805257,
+      "grad_norm": 0.44114769614711064,
+      "learning_rate": 2.6876541562068775e-06,
+      "loss": 0.5266,
+      "step": 10810
+    },
+    {
+      "epoch": 2.8700384972786406,
+      "grad_norm": 0.44883805588543824,
+      "learning_rate": 2.6873060113873973e-06,
+      "loss": 0.5416,
+      "step": 10811
+    },
+    {
+      "epoch": 2.8703039957520247,
+      "grad_norm": 0.4486930777990287,
+      "learning_rate": 2.6869578629149918e-06,
+      "loss": 0.5965,
+      "step": 10812
+    },
+    {
+      "epoch": 2.8705694942254083,
+      "grad_norm": 0.4411784366304622,
+      "learning_rate": 2.6866097107964495e-06,
+      "loss": 0.5441,
+      "step": 10813
+    },
+    {
+      "epoch": 2.870834992698792,
+      "grad_norm": 0.4376672470673291,
+      "learning_rate": 2.6862615550385616e-06,
+      "loss": 0.5773,
+      "step": 10814
+    },
+    {
+      "epoch": 2.871100491172176,
+      "grad_norm": 0.43184379446850796,
+      "learning_rate": 2.6859133956481174e-06,
+      "loss": 0.5486,
+      "step": 10815
+    },
+    {
+      "epoch": 2.8713659896455597,
+      "grad_norm": 0.4479987787890383,
+      "learning_rate": 2.6855652326319063e-06,
+      "loss": 0.5914,
+      "step": 10816
+    },
+    {
+      "epoch": 2.8716314881189433,
+      "grad_norm": 0.43305473354461277,
+      "learning_rate": 2.68521706599672e-06,
+      "loss": 0.5753,
+      "step": 10817
+    },
+    {
+      "epoch": 2.871896986592327,
+      "grad_norm": 0.44439596170296236,
+      "learning_rate": 2.684868895749346e-06,
+      "loss": 0.5784,
+      "step": 10818
+    },
+    {
+      "epoch": 2.872162485065711,
+      "grad_norm": 0.46071555032966255,
+      "learning_rate": 2.6845207218965763e-06,
+      "loss": 0.5704,
+      "step": 10819
+    },
+    {
+      "epoch": 2.8724279835390947,
+      "grad_norm": 0.43250899328799286,
+      "learning_rate": 2.6841725444452005e-06,
+      "loss": 0.5533,
+      "step": 10820
+    },
+    {
+      "epoch": 2.8726934820124783,
+      "grad_norm": 0.44290476232115306,
+      "learning_rate": 2.6838243634020095e-06,
+      "loss": 0.5478,
+      "step": 10821
+    },
+    {
+      "epoch": 2.872958980485862,
+      "grad_norm": 0.4523526018772023,
+      "learning_rate": 2.6834761787737934e-06,
+      "loss": 0.5576,
+      "step": 10822
+    },
+    {
+      "epoch": 2.873224478959246,
+      "grad_norm": 0.440979661564685,
+      "learning_rate": 2.683127990567343e-06,
+      "loss": 0.5729,
+      "step": 10823
+    },
+    {
+      "epoch": 2.8734899774326297,
+      "grad_norm": 0.43162485098917514,
+      "learning_rate": 2.6827797987894476e-06,
+      "loss": 0.5326,
+      "step": 10824
+    },
+    {
+      "epoch": 2.8737554759060133,
+      "grad_norm": 0.43149189949537026,
+      "learning_rate": 2.6824316034468987e-06,
+      "loss": 0.5563,
+      "step": 10825
+    },
+    {
+      "epoch": 2.8740209743793974,
+      "grad_norm": 0.4419009976170839,
+      "learning_rate": 2.682083404546488e-06,
+      "loss": 0.5736,
+      "step": 10826
+    },
+    {
+      "epoch": 2.874286472852781,
+      "grad_norm": 0.43802541637998726,
+      "learning_rate": 2.6817352020950037e-06,
+      "loss": 0.5569,
+      "step": 10827
+    },
+    {
+      "epoch": 2.8745519713261647,
+      "grad_norm": 0.4386676180692943,
+      "learning_rate": 2.6813869960992385e-06,
+      "loss": 0.5763,
+      "step": 10828
+    },
+    {
+      "epoch": 2.8748174697995488,
+      "grad_norm": 0.4392041135494687,
+      "learning_rate": 2.6810387865659825e-06,
+      "loss": 0.5903,
+      "step": 10829
+    },
+    {
+      "epoch": 2.8750829682729324,
+      "grad_norm": 0.4482191398101328,
+      "learning_rate": 2.6806905735020267e-06,
+      "loss": 0.5436,
+      "step": 10830
+    },
+    {
+      "epoch": 2.875348466746316,
+      "grad_norm": 0.44218635440073767,
+      "learning_rate": 2.6803423569141636e-06,
+      "loss": 0.5863,
+      "step": 10831
+    },
+    {
+      "epoch": 2.8756139652197,
+      "grad_norm": 0.43066766131922357,
+      "learning_rate": 2.679994136809182e-06,
+      "loss": 0.5648,
+      "step": 10832
+    },
+    {
+      "epoch": 2.875879463693084,
+      "grad_norm": 0.4452498218090366,
+      "learning_rate": 2.6796459131938733e-06,
+      "loss": 0.5835,
+      "step": 10833
+    },
+    {
+      "epoch": 2.8761449621664674,
+      "grad_norm": 0.45456715861688873,
+      "learning_rate": 2.6792976860750312e-06,
+      "loss": 0.5527,
+      "step": 10834
+    },
+    {
+      "epoch": 2.8764104606398515,
+      "grad_norm": 0.4380656206875285,
+      "learning_rate": 2.678949455459444e-06,
+      "loss": 0.5228,
+      "step": 10835
+    },
+    {
+      "epoch": 2.876675959113235,
+      "grad_norm": 0.46010713423228644,
+      "learning_rate": 2.678601221353905e-06,
+      "loss": 0.5588,
+      "step": 10836
+    },
+    {
+      "epoch": 2.876941457586619,
+      "grad_norm": 0.45977160600196393,
+      "learning_rate": 2.6782529837652045e-06,
+      "loss": 0.547,
+      "step": 10837
+    },
+    {
+      "epoch": 2.877206956060003,
+      "grad_norm": 0.4485596768856643,
+      "learning_rate": 2.6779047427001336e-06,
+      "loss": 0.5465,
+      "step": 10838
+    },
+    {
+      "epoch": 2.8774724545333865,
+      "grad_norm": 0.4471934281327445,
+      "learning_rate": 2.6775564981654856e-06,
+      "loss": 0.5426,
+      "step": 10839
+    },
+    {
+      "epoch": 2.87773795300677,
+      "grad_norm": 0.43891709868681367,
+      "learning_rate": 2.6772082501680514e-06,
+      "loss": 0.5622,
+      "step": 10840
+    },
+    {
+      "epoch": 2.8780034514801542,
+      "grad_norm": 0.4320520571670163,
+      "learning_rate": 2.6768599987146215e-06,
+      "loss": 0.5085,
+      "step": 10841
+    },
+    {
+      "epoch": 2.878268949953538,
+      "grad_norm": 0.4411475385879734,
+      "learning_rate": 2.676511743811989e-06,
+      "loss": 0.5256,
+      "step": 10842
+    },
+    {
+      "epoch": 2.8785344484269215,
+      "grad_norm": 0.452468711228945,
+      "learning_rate": 2.676163485466946e-06,
+      "loss": 0.5742,
+      "step": 10843
+    },
+    {
+      "epoch": 2.878799946900305,
+      "grad_norm": 0.44392858774568406,
+      "learning_rate": 2.6758152236862827e-06,
+      "loss": 0.583,
+      "step": 10844
+    },
+    {
+      "epoch": 2.8790654453736892,
+      "grad_norm": 0.48262724715708205,
+      "learning_rate": 2.675466958476793e-06,
+      "loss": 0.5067,
+      "step": 10845
+    },
+    {
+      "epoch": 2.879330943847073,
+      "grad_norm": 0.45651380054768226,
+      "learning_rate": 2.6751186898452673e-06,
+      "loss": 0.5517,
+      "step": 10846
+    },
+    {
+      "epoch": 2.8795964423204565,
+      "grad_norm": 0.4463889492644063,
+      "learning_rate": 2.6747704177984983e-06,
+      "loss": 0.58,
+      "step": 10847
+    },
+    {
+      "epoch": 2.87986194079384,
+      "grad_norm": 0.4436312994448471,
+      "learning_rate": 2.6744221423432786e-06,
+      "loss": 0.5362,
+      "step": 10848
+    },
+    {
+      "epoch": 2.8801274392672243,
+      "grad_norm": 0.4448488830640065,
+      "learning_rate": 2.6740738634864e-06,
+      "loss": 0.5565,
+      "step": 10849
+    },
+    {
+      "epoch": 2.880392937740608,
+      "grad_norm": 0.4817548404271945,
+      "learning_rate": 2.6737255812346547e-06,
+      "loss": 0.5089,
+      "step": 10850
+    },
+    {
+      "epoch": 2.8806584362139915,
+      "grad_norm": 0.4368878684261856,
+      "learning_rate": 2.6733772955948352e-06,
+      "loss": 0.5154,
+      "step": 10851
+    },
+    {
+      "epoch": 2.8809239346873756,
+      "grad_norm": 0.44767934500164847,
+      "learning_rate": 2.673029006573735e-06,
+      "loss": 0.544,
+      "step": 10852
+    },
+    {
+      "epoch": 2.8811894331607593,
+      "grad_norm": 0.44541810496343487,
+      "learning_rate": 2.672680714178144e-06,
+      "loss": 0.5747,
+      "step": 10853
+    },
+    {
+      "epoch": 2.881454931634143,
+      "grad_norm": 0.43896537345326203,
+      "learning_rate": 2.6723324184148576e-06,
+      "loss": 0.5779,
+      "step": 10854
+    },
+    {
+      "epoch": 2.881720430107527,
+      "grad_norm": 0.4310170528667718,
+      "learning_rate": 2.671984119290666e-06,
+      "loss": 0.5598,
+      "step": 10855
+    },
+    {
+      "epoch": 2.8819859285809106,
+      "grad_norm": 0.44721733651614415,
+      "learning_rate": 2.6716358168123633e-06,
+      "loss": 0.5301,
+      "step": 10856
+    },
+    {
+      "epoch": 2.8822514270542943,
+      "grad_norm": 0.42864646289597147,
+      "learning_rate": 2.671287510986742e-06,
+      "loss": 0.542,
+      "step": 10857
+    },
+    {
+      "epoch": 2.8825169255276784,
+      "grad_norm": 0.4369746610196976,
+      "learning_rate": 2.6709392018205944e-06,
+      "loss": 0.5835,
+      "step": 10858
+    },
+    {
+      "epoch": 2.882782424001062,
+      "grad_norm": 0.44602426416435603,
+      "learning_rate": 2.6705908893207144e-06,
+      "loss": 0.5564,
+      "step": 10859
+    },
+    {
+      "epoch": 2.8830479224744456,
+      "grad_norm": 0.4464058686004329,
+      "learning_rate": 2.6702425734938948e-06,
+      "loss": 0.5523,
+      "step": 10860
+    },
+    {
+      "epoch": 2.8833134209478297,
+      "grad_norm": 0.4458263398218292,
+      "learning_rate": 2.6698942543469275e-06,
+      "loss": 0.5692,
+      "step": 10861
+    },
+    {
+      "epoch": 2.8835789194212134,
+      "grad_norm": 0.44316466620704503,
+      "learning_rate": 2.669545931886606e-06,
+      "loss": 0.5786,
+      "step": 10862
+    },
+    {
+      "epoch": 2.883844417894597,
+      "grad_norm": 0.4312861516714903,
+      "learning_rate": 2.6691976061197245e-06,
+      "loss": 0.5548,
+      "step": 10863
+    },
+    {
+      "epoch": 2.884109916367981,
+      "grad_norm": 0.44255286569736574,
+      "learning_rate": 2.6688492770530748e-06,
+      "loss": 0.5859,
+      "step": 10864
+    },
+    {
+      "epoch": 2.8843754148413647,
+      "grad_norm": 0.43321625083875426,
+      "learning_rate": 2.668500944693451e-06,
+      "loss": 0.555,
+      "step": 10865
+    },
+    {
+      "epoch": 2.8846409133147484,
+      "grad_norm": 0.43531149971538413,
+      "learning_rate": 2.668152609047646e-06,
+      "loss": 0.5527,
+      "step": 10866
+    },
+    {
+      "epoch": 2.8849064117881325,
+      "grad_norm": 0.4361321588632163,
+      "learning_rate": 2.667804270122454e-06,
+      "loss": 0.5386,
+      "step": 10867
+    },
+    {
+      "epoch": 2.885171910261516,
+      "grad_norm": 0.4437577053716983,
+      "learning_rate": 2.667455927924667e-06,
+      "loss": 0.5635,
+      "step": 10868
+    },
+    {
+      "epoch": 2.8854374087348997,
+      "grad_norm": 0.44278390800727474,
+      "learning_rate": 2.6671075824610803e-06,
+      "loss": 0.5681,
+      "step": 10869
+    },
+    {
+      "epoch": 2.885702907208284,
+      "grad_norm": 0.433106111626509,
+      "learning_rate": 2.6667592337384864e-06,
+      "loss": 0.5863,
+      "step": 10870
+    },
+    {
+      "epoch": 2.8859684056816675,
+      "grad_norm": 0.4270240982678621,
+      "learning_rate": 2.6664108817636785e-06,
+      "loss": 0.5329,
+      "step": 10871
+    },
+    {
+      "epoch": 2.886233904155051,
+      "grad_norm": 0.45712822613003545,
+      "learning_rate": 2.666062526543452e-06,
+      "loss": 0.5925,
+      "step": 10872
+    },
+    {
+      "epoch": 2.8864994026284347,
+      "grad_norm": 0.4601936028118898,
+      "learning_rate": 2.6657141680845994e-06,
+      "loss": 0.5338,
+      "step": 10873
+    },
+    {
+      "epoch": 2.886764901101819,
+      "grad_norm": 0.4472989096495747,
+      "learning_rate": 2.665365806393914e-06,
+      "loss": 0.567,
+      "step": 10874
+    },
+    {
+      "epoch": 2.8870303995752025,
+      "grad_norm": 0.44955009092777526,
+      "learning_rate": 2.665017441478191e-06,
+      "loss": 0.6097,
+      "step": 10875
+    },
+    {
+      "epoch": 2.887295898048586,
+      "grad_norm": 0.43231338361898963,
+      "learning_rate": 2.6646690733442236e-06,
+      "loss": 0.5594,
+      "step": 10876
+    },
+    {
+      "epoch": 2.8875613965219697,
+      "grad_norm": 0.45075223683003657,
+      "learning_rate": 2.6643207019988065e-06,
+      "loss": 0.5877,
+      "step": 10877
+    },
+    {
+      "epoch": 2.887826894995354,
+      "grad_norm": 0.44544448527423086,
+      "learning_rate": 2.6639723274487334e-06,
+      "loss": 0.525,
+      "step": 10878
+    },
+    {
+      "epoch": 2.8880923934687375,
+      "grad_norm": 0.434334548974687,
+      "learning_rate": 2.6636239497007984e-06,
+      "loss": 0.5315,
+      "step": 10879
+    },
+    {
+      "epoch": 2.888357891942121,
+      "grad_norm": 0.44885227514256315,
+      "learning_rate": 2.6632755687617955e-06,
+      "loss": 0.5432,
+      "step": 10880
+    },
+    {
+      "epoch": 2.888623390415505,
+      "grad_norm": 0.43988230272360723,
+      "learning_rate": 2.662927184638519e-06,
+      "loss": 0.5626,
+      "step": 10881
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.45518679626984343,
+      "learning_rate": 2.662578797337764e-06,
+      "loss": 0.5822,
+      "step": 10882
+    },
+    {
+      "epoch": 2.8891543873622725,
+      "grad_norm": 0.4467505118143057,
+      "learning_rate": 2.6622304068663246e-06,
+      "loss": 0.5386,
+      "step": 10883
+    },
+    {
+      "epoch": 2.8894198858356566,
+      "grad_norm": 0.43306594349569794,
+      "learning_rate": 2.661882013230994e-06,
+      "loss": 0.5628,
+      "step": 10884
+    },
+    {
+      "epoch": 2.88968538430904,
+      "grad_norm": 0.41041798195591706,
+      "learning_rate": 2.6615336164385685e-06,
+      "loss": 0.5045,
+      "step": 10885
+    },
+    {
+      "epoch": 2.889950882782424,
+      "grad_norm": 0.45199972987522774,
+      "learning_rate": 2.6611852164958423e-06,
+      "loss": 0.5451,
+      "step": 10886
+    },
+    {
+      "epoch": 2.890216381255808,
+      "grad_norm": 0.4592370261856821,
+      "learning_rate": 2.660836813409609e-06,
+      "loss": 0.5668,
+      "step": 10887
+    },
+    {
+      "epoch": 2.8904818797291916,
+      "grad_norm": 0.4470436497488048,
+      "learning_rate": 2.660488407186664e-06,
+      "loss": 0.5981,
+      "step": 10888
+    },
+    {
+      "epoch": 2.890747378202575,
+      "grad_norm": 0.4361218504831239,
+      "learning_rate": 2.6601399978338034e-06,
+      "loss": 0.5514,
+      "step": 10889
+    },
+    {
+      "epoch": 2.8910128766759593,
+      "grad_norm": 0.44530977881106987,
+      "learning_rate": 2.6597915853578195e-06,
+      "loss": 0.5637,
+      "step": 10890
+    },
+    {
+      "epoch": 2.891278375149343,
+      "grad_norm": 0.43167158787203563,
+      "learning_rate": 2.6594431697655087e-06,
+      "loss": 0.5463,
+      "step": 10891
+    },
+    {
+      "epoch": 2.8915438736227266,
+      "grad_norm": 0.4340141491318965,
+      "learning_rate": 2.659094751063666e-06,
+      "loss": 0.5481,
+      "step": 10892
+    },
+    {
+      "epoch": 2.8918093720961107,
+      "grad_norm": 0.4442181747184613,
+      "learning_rate": 2.658746329259086e-06,
+      "loss": 0.54,
+      "step": 10893
+    },
+    {
+      "epoch": 2.8920748705694943,
+      "grad_norm": 0.44184859518917863,
+      "learning_rate": 2.6583979043585633e-06,
+      "loss": 0.5304,
+      "step": 10894
+    },
+    {
+      "epoch": 2.892340369042878,
+      "grad_norm": 0.450055242843919,
+      "learning_rate": 2.658049476368895e-06,
+      "loss": 0.5498,
+      "step": 10895
+    },
+    {
+      "epoch": 2.892605867516262,
+      "grad_norm": 0.45316290903820783,
+      "learning_rate": 2.657701045296874e-06,
+      "loss": 0.5602,
+      "step": 10896
+    },
+    {
+      "epoch": 2.8928713659896457,
+      "grad_norm": 0.4401253783443589,
+      "learning_rate": 2.657352611149296e-06,
+      "loss": 0.5591,
+      "step": 10897
+    },
+    {
+      "epoch": 2.8931368644630293,
+      "grad_norm": 0.43622151426619354,
+      "learning_rate": 2.6570041739329576e-06,
+      "loss": 0.5572,
+      "step": 10898
+    },
+    {
+      "epoch": 2.893402362936413,
+      "grad_norm": 0.42662031380810866,
+      "learning_rate": 2.6566557336546533e-06,
+      "loss": 0.5469,
+      "step": 10899
+    },
+    {
+      "epoch": 2.893667861409797,
+      "grad_norm": 0.43024760001087314,
+      "learning_rate": 2.656307290321179e-06,
+      "loss": 0.5457,
+      "step": 10900
+    },
+    {
+      "epoch": 2.8939333598831807,
+      "grad_norm": 0.4467549013135979,
+      "learning_rate": 2.65595884393933e-06,
+      "loss": 0.5519,
+      "step": 10901
+    },
+    {
+      "epoch": 2.8941988583565643,
+      "grad_norm": 0.44258297677957514,
+      "learning_rate": 2.6556103945159005e-06,
+      "loss": 0.5813,
+      "step": 10902
+    },
+    {
+      "epoch": 2.894464356829948,
+      "grad_norm": 0.45282890438368784,
+      "learning_rate": 2.6552619420576876e-06,
+      "loss": 0.5332,
+      "step": 10903
+    },
+    {
+      "epoch": 2.894729855303332,
+      "grad_norm": 0.4341136622860847,
+      "learning_rate": 2.654913486571487e-06,
+      "loss": 0.5847,
+      "step": 10904
+    },
+    {
+      "epoch": 2.8949953537767157,
+      "grad_norm": 0.44171869462126057,
+      "learning_rate": 2.654565028064094e-06,
+      "loss": 0.5655,
+      "step": 10905
+    },
+    {
+      "epoch": 2.8952608522500993,
+      "grad_norm": 0.43393078047895756,
+      "learning_rate": 2.654216566542305e-06,
+      "loss": 0.5659,
+      "step": 10906
+    },
+    {
+      "epoch": 2.8955263507234834,
+      "grad_norm": 0.43635181981188764,
+      "learning_rate": 2.653868102012915e-06,
+      "loss": 0.5548,
+      "step": 10907
+    },
+    {
+      "epoch": 2.895791849196867,
+      "grad_norm": 0.4261915598496779,
+      "learning_rate": 2.65351963448272e-06,
+      "loss": 0.5504,
+      "step": 10908
+    },
+    {
+      "epoch": 2.8960573476702507,
+      "grad_norm": 0.42926860439761766,
+      "learning_rate": 2.653171163958517e-06,
+      "loss": 0.5611,
+      "step": 10909
+    },
+    {
+      "epoch": 2.8963228461436348,
+      "grad_norm": 0.43997026515276516,
+      "learning_rate": 2.6528226904471006e-06,
+      "loss": 0.5587,
+      "step": 10910
+    },
+    {
+      "epoch": 2.8965883446170184,
+      "grad_norm": 0.42474534302224715,
+      "learning_rate": 2.652474213955267e-06,
+      "loss": 0.5479,
+      "step": 10911
+    },
+    {
+      "epoch": 2.896853843090402,
+      "grad_norm": 0.4348882855324774,
+      "learning_rate": 2.6521257344898137e-06,
+      "loss": 0.5421,
+      "step": 10912
+    },
+    {
+      "epoch": 2.897119341563786,
+      "grad_norm": 0.450593347622781,
+      "learning_rate": 2.6517772520575362e-06,
+      "loss": 0.5893,
+      "step": 10913
+    },
+    {
+      "epoch": 2.89738484003717,
+      "grad_norm": 0.44003034685439146,
+      "learning_rate": 2.6514287666652305e-06,
+      "loss": 0.5354,
+      "step": 10914
+    },
+    {
+      "epoch": 2.8976503385105534,
+      "grad_norm": 0.44192082772191166,
+      "learning_rate": 2.6510802783196933e-06,
+      "loss": 0.5334,
+      "step": 10915
+    },
+    {
+      "epoch": 2.8979158369839375,
+      "grad_norm": 0.4331852355918647,
+      "learning_rate": 2.6507317870277205e-06,
+      "loss": 0.5716,
+      "step": 10916
+    },
+    {
+      "epoch": 2.898181335457321,
+      "grad_norm": 0.4589238668487718,
+      "learning_rate": 2.650383292796109e-06,
+      "loss": 0.5968,
+      "step": 10917
+    },
+    {
+      "epoch": 2.898446833930705,
+      "grad_norm": 0.43782726818763074,
+      "learning_rate": 2.650034795631655e-06,
+      "loss": 0.5378,
+      "step": 10918
+    },
+    {
+      "epoch": 2.898712332404089,
+      "grad_norm": 0.4461233497566889,
+      "learning_rate": 2.649686295541155e-06,
+      "loss": 0.5292,
+      "step": 10919
+    },
+    {
+      "epoch": 2.8989778308774725,
+      "grad_norm": 0.45222055609815054,
+      "learning_rate": 2.6493377925314057e-06,
+      "loss": 0.5606,
+      "step": 10920
+    },
+    {
+      "epoch": 2.899243329350856,
+      "grad_norm": 0.45637331622513116,
+      "learning_rate": 2.6489892866092037e-06,
+      "loss": 0.5889,
+      "step": 10921
+    },
+    {
+      "epoch": 2.8995088278242402,
+      "grad_norm": 0.44055113014127406,
+      "learning_rate": 2.648640777781346e-06,
+      "loss": 0.5621,
+      "step": 10922
+    },
+    {
+      "epoch": 2.899774326297624,
+      "grad_norm": 0.43328442092076996,
+      "learning_rate": 2.6482922660546296e-06,
+      "loss": 0.5627,
+      "step": 10923
+    },
+    {
+      "epoch": 2.9000398247710075,
+      "grad_norm": 0.43746414192602867,
+      "learning_rate": 2.6479437514358506e-06,
+      "loss": 0.5809,
+      "step": 10924
+    },
+    {
+      "epoch": 2.9003053232443916,
+      "grad_norm": 0.44263605524024885,
+      "learning_rate": 2.6475952339318064e-06,
+      "loss": 0.5745,
+      "step": 10925
+    },
+    {
+      "epoch": 2.9005708217177752,
+      "grad_norm": 0.448103484572625,
+      "learning_rate": 2.6472467135492943e-06,
+      "loss": 0.5595,
+      "step": 10926
+    },
+    {
+      "epoch": 2.900836320191159,
+      "grad_norm": 0.4462869568646484,
+      "learning_rate": 2.6468981902951097e-06,
+      "loss": 0.5632,
+      "step": 10927
+    },
+    {
+      "epoch": 2.9011018186645425,
+      "grad_norm": 0.4384724991957383,
+      "learning_rate": 2.646549664176051e-06,
+      "loss": 0.5504,
+      "step": 10928
+    },
+    {
+      "epoch": 2.9013673171379266,
+      "grad_norm": 0.4298794056285689,
+      "learning_rate": 2.6462011351989153e-06,
+      "loss": 0.5615,
+      "step": 10929
+    },
+    {
+      "epoch": 2.9016328156113103,
+      "grad_norm": 0.44376785333814506,
+      "learning_rate": 2.6458526033704986e-06,
+      "loss": 0.5579,
+      "step": 10930
+    },
+    {
+      "epoch": 2.901898314084694,
+      "grad_norm": 0.4340368406954021,
+      "learning_rate": 2.6455040686975997e-06,
+      "loss": 0.5491,
+      "step": 10931
+    },
+    {
+      "epoch": 2.9021638125580775,
+      "grad_norm": 0.44708777893009755,
+      "learning_rate": 2.6451555311870154e-06,
+      "loss": 0.5868,
+      "step": 10932
+    },
+    {
+      "epoch": 2.9024293110314616,
+      "grad_norm": 0.44025855844739836,
+      "learning_rate": 2.6448069908455424e-06,
+      "loss": 0.5989,
+      "step": 10933
+    },
+    {
+      "epoch": 2.9026948095048453,
+      "grad_norm": 0.4462573107282268,
+      "learning_rate": 2.644458447679979e-06,
+      "loss": 0.5606,
+      "step": 10934
+    },
+    {
+      "epoch": 2.902960307978229,
+      "grad_norm": 0.4413153831115396,
+      "learning_rate": 2.644109901697122e-06,
+      "loss": 0.5816,
+      "step": 10935
+    },
+    {
+      "epoch": 2.903225806451613,
+      "grad_norm": 0.4363934289521345,
+      "learning_rate": 2.6437613529037693e-06,
+      "loss": 0.5433,
+      "step": 10936
+    },
+    {
+      "epoch": 2.9034913049249966,
+      "grad_norm": 0.4300405345290328,
+      "learning_rate": 2.643412801306718e-06,
+      "loss": 0.5549,
+      "step": 10937
+    },
+    {
+      "epoch": 2.9037568033983803,
+      "grad_norm": 0.4350433349067469,
+      "learning_rate": 2.643064246912766e-06,
+      "loss": 0.5593,
+      "step": 10938
+    },
+    {
+      "epoch": 2.9040223018717644,
+      "grad_norm": 0.4463047863693827,
+      "learning_rate": 2.642715689728711e-06,
+      "loss": 0.5603,
+      "step": 10939
+    },
+    {
+      "epoch": 2.904287800345148,
+      "grad_norm": 0.43459082214595607,
+      "learning_rate": 2.6423671297613496e-06,
+      "loss": 0.5424,
+      "step": 10940
+    },
+    {
+      "epoch": 2.9045532988185316,
+      "grad_norm": 0.4387189873759832,
+      "learning_rate": 2.642018567017482e-06,
+      "loss": 0.5567,
+      "step": 10941
+    },
+    {
+      "epoch": 2.9048187972919157,
+      "grad_norm": 0.44360597920127154,
+      "learning_rate": 2.6416700015039037e-06,
+      "loss": 0.5597,
+      "step": 10942
+    },
+    {
+      "epoch": 2.9050842957652994,
+      "grad_norm": 0.428596800162275,
+      "learning_rate": 2.641321433227414e-06,
+      "loss": 0.5441,
+      "step": 10943
+    },
+    {
+      "epoch": 2.905349794238683,
+      "grad_norm": 0.4316249702421703,
+      "learning_rate": 2.6409728621948106e-06,
+      "loss": 0.513,
+      "step": 10944
+    },
+    {
+      "epoch": 2.905615292712067,
+      "grad_norm": 0.4455273894116888,
+      "learning_rate": 2.6406242884128906e-06,
+      "loss": 0.5803,
+      "step": 10945
+    },
+    {
+      "epoch": 2.9058807911854507,
+      "grad_norm": 0.436338016312051,
+      "learning_rate": 2.6402757118884525e-06,
+      "loss": 0.5218,
+      "step": 10946
+    },
+    {
+      "epoch": 2.9061462896588344,
+      "grad_norm": 0.44455546765062054,
+      "learning_rate": 2.6399271326282957e-06,
+      "loss": 0.5517,
+      "step": 10947
+    },
+    {
+      "epoch": 2.9064117881322185,
+      "grad_norm": 0.439299393094528,
+      "learning_rate": 2.6395785506392164e-06,
+      "loss": 0.5553,
+      "step": 10948
+    },
+    {
+      "epoch": 2.906677286605602,
+      "grad_norm": 0.44216456744201266,
+      "learning_rate": 2.6392299659280134e-06,
+      "loss": 0.5533,
+      "step": 10949
+    },
+    {
+      "epoch": 2.9069427850789857,
+      "grad_norm": 0.444469068450367,
+      "learning_rate": 2.6388813785014857e-06,
+      "loss": 0.5368,
+      "step": 10950
+    },
+    {
+      "epoch": 2.90720828355237,
+      "grad_norm": 0.4359474600841476,
+      "learning_rate": 2.6385327883664304e-06,
+      "loss": 0.5514,
+      "step": 10951
+    },
+    {
+      "epoch": 2.9074737820257535,
+      "grad_norm": 0.46208991896781965,
+      "learning_rate": 2.638184195529647e-06,
+      "loss": 0.5673,
+      "step": 10952
+    },
+    {
+      "epoch": 2.907739280499137,
+      "grad_norm": 0.45757539982651235,
+      "learning_rate": 2.637835599997934e-06,
+      "loss": 0.5781,
+      "step": 10953
+    },
+    {
+      "epoch": 2.9080047789725207,
+      "grad_norm": 0.4560009721163749,
+      "learning_rate": 2.6374870017780878e-06,
+      "loss": 0.5324,
+      "step": 10954
+    },
+    {
+      "epoch": 2.908270277445905,
+      "grad_norm": 0.4498071639255698,
+      "learning_rate": 2.63713840087691e-06,
+      "loss": 0.5715,
+      "step": 10955
+    },
+    {
+      "epoch": 2.9085357759192885,
+      "grad_norm": 0.44222843965919817,
+      "learning_rate": 2.636789797301197e-06,
+      "loss": 0.5073,
+      "step": 10956
+    },
+    {
+      "epoch": 2.908801274392672,
+      "grad_norm": 0.4468151695723339,
+      "learning_rate": 2.636441191057748e-06,
+      "loss": 0.5754,
+      "step": 10957
+    },
+    {
+      "epoch": 2.9090667728660558,
+      "grad_norm": 0.424683660695068,
+      "learning_rate": 2.6360925821533617e-06,
+      "loss": 0.5249,
+      "step": 10958
+    },
+    {
+      "epoch": 2.90933227133944,
+      "grad_norm": 0.44089192749465134,
+      "learning_rate": 2.6357439705948366e-06,
+      "loss": 0.593,
+      "step": 10959
+    },
+    {
+      "epoch": 2.9095977698128235,
+      "grad_norm": 0.4552840074979933,
+      "learning_rate": 2.635395356388972e-06,
+      "loss": 0.5529,
+      "step": 10960
+    },
+    {
+      "epoch": 2.909863268286207,
+      "grad_norm": 0.45090216434136277,
+      "learning_rate": 2.6350467395425665e-06,
+      "loss": 0.5834,
+      "step": 10961
+    },
+    {
+      "epoch": 2.910128766759591,
+      "grad_norm": 0.44725579933570425,
+      "learning_rate": 2.634698120062419e-06,
+      "loss": 0.5887,
+      "step": 10962
+    },
+    {
+      "epoch": 2.910394265232975,
+      "grad_norm": 0.4397702648724235,
+      "learning_rate": 2.634349497955328e-06,
+      "loss": 0.594,
+      "step": 10963
+    },
+    {
+      "epoch": 2.9106597637063585,
+      "grad_norm": 0.4491331641583865,
+      "learning_rate": 2.6340008732280935e-06,
+      "loss": 0.593,
+      "step": 10964
+    },
+    {
+      "epoch": 2.9109252621797426,
+      "grad_norm": 0.43013541405874617,
+      "learning_rate": 2.6336522458875137e-06,
+      "loss": 0.5698,
+      "step": 10965
+    },
+    {
+      "epoch": 2.911190760653126,
+      "grad_norm": 0.4422740470419798,
+      "learning_rate": 2.6333036159403873e-06,
+      "loss": 0.5471,
+      "step": 10966
+    },
+    {
+      "epoch": 2.91145625912651,
+      "grad_norm": 0.43800276207178707,
+      "learning_rate": 2.632954983393514e-06,
+      "loss": 0.5245,
+      "step": 10967
+    },
+    {
+      "epoch": 2.911721757599894,
+      "grad_norm": 0.43542102705990954,
+      "learning_rate": 2.6326063482536933e-06,
+      "loss": 0.5134,
+      "step": 10968
+    },
+    {
+      "epoch": 2.9119872560732776,
+      "grad_norm": 0.43387884694884193,
+      "learning_rate": 2.6322577105277243e-06,
+      "loss": 0.5401,
+      "step": 10969
+    },
+    {
+      "epoch": 2.912252754546661,
+      "grad_norm": 0.4327696913375593,
+      "learning_rate": 2.631909070222406e-06,
+      "loss": 0.5175,
+      "step": 10970
+    },
+    {
+      "epoch": 2.9125182530200453,
+      "grad_norm": 0.4510040472030239,
+      "learning_rate": 2.631560427344538e-06,
+      "loss": 0.5731,
+      "step": 10971
+    },
+    {
+      "epoch": 2.912783751493429,
+      "grad_norm": 0.43554688223001164,
+      "learning_rate": 2.6312117819009192e-06,
+      "loss": 0.5489,
+      "step": 10972
+    },
+    {
+      "epoch": 2.9130492499668126,
+      "grad_norm": 0.42539966900676873,
+      "learning_rate": 2.6308631338983497e-06,
+      "loss": 0.5194,
+      "step": 10973
+    },
+    {
+      "epoch": 2.9133147484401967,
+      "grad_norm": 0.4425387934346445,
+      "learning_rate": 2.6305144833436287e-06,
+      "loss": 0.5295,
+      "step": 10974
+    },
+    {
+      "epoch": 2.9135802469135803,
+      "grad_norm": 0.44746049042270636,
+      "learning_rate": 2.630165830243555e-06,
+      "loss": 0.5549,
+      "step": 10975
+    },
+    {
+      "epoch": 2.913845745386964,
+      "grad_norm": 0.44276129815038756,
+      "learning_rate": 2.6298171746049294e-06,
+      "loss": 0.5696,
+      "step": 10976
+    },
+    {
+      "epoch": 2.914111243860348,
+      "grad_norm": 0.45662944551679474,
+      "learning_rate": 2.6294685164345513e-06,
+      "loss": 0.5262,
+      "step": 10977
+    },
+    {
+      "epoch": 2.9143767423337317,
+      "grad_norm": 0.44721510646927704,
+      "learning_rate": 2.6291198557392206e-06,
+      "loss": 0.5303,
+      "step": 10978
+    },
+    {
+      "epoch": 2.9146422408071153,
+      "grad_norm": 0.4518031286990219,
+      "learning_rate": 2.6287711925257354e-06,
+      "loss": 0.5688,
+      "step": 10979
+    },
+    {
+      "epoch": 2.9149077392804994,
+      "grad_norm": 0.483696395781235,
+      "learning_rate": 2.6284225268008975e-06,
+      "loss": 0.5444,
+      "step": 10980
+    },
+    {
+      "epoch": 2.915173237753883,
+      "grad_norm": 0.4390419868148622,
+      "learning_rate": 2.628073858571506e-06,
+      "loss": 0.5591,
+      "step": 10981
+    },
+    {
+      "epoch": 2.9154387362272667,
+      "grad_norm": 0.45552141284448083,
+      "learning_rate": 2.62772518784436e-06,
+      "loss": 0.5933,
+      "step": 10982
+    },
+    {
+      "epoch": 2.9157042347006503,
+      "grad_norm": 0.4381512500155813,
+      "learning_rate": 2.6273765146262603e-06,
+      "loss": 0.5687,
+      "step": 10983
+    },
+    {
+      "epoch": 2.9159697331740344,
+      "grad_norm": 0.4285549965447163,
+      "learning_rate": 2.627027838924007e-06,
+      "loss": 0.5388,
+      "step": 10984
+    },
+    {
+      "epoch": 2.916235231647418,
+      "grad_norm": 0.4410277820701708,
+      "learning_rate": 2.6266791607443996e-06,
+      "loss": 0.5747,
+      "step": 10985
+    },
+    {
+      "epoch": 2.9165007301208017,
+      "grad_norm": 0.42977891314856626,
+      "learning_rate": 2.626330480094238e-06,
+      "loss": 0.5452,
+      "step": 10986
+    },
+    {
+      "epoch": 2.9167662285941853,
+      "grad_norm": 0.4533809508689432,
+      "learning_rate": 2.625981796980323e-06,
+      "loss": 0.5645,
+      "step": 10987
+    },
+    {
+      "epoch": 2.9170317270675694,
+      "grad_norm": 0.44291686828721816,
+      "learning_rate": 2.6256331114094547e-06,
+      "loss": 0.5745,
+      "step": 10988
+    },
+    {
+      "epoch": 2.917297225540953,
+      "grad_norm": 0.4306466116476081,
+      "learning_rate": 2.625284423388433e-06,
+      "loss": 0.5227,
+      "step": 10989
+    },
+    {
+      "epoch": 2.9175627240143367,
+      "grad_norm": 0.45560227157112076,
+      "learning_rate": 2.624935732924059e-06,
+      "loss": 0.5711,
+      "step": 10990
+    },
+    {
+      "epoch": 2.917828222487721,
+      "grad_norm": 0.4507941930167085,
+      "learning_rate": 2.6245870400231314e-06,
+      "loss": 0.5478,
+      "step": 10991
+    },
+    {
+      "epoch": 2.9180937209611044,
+      "grad_norm": 0.44799917897270103,
+      "learning_rate": 2.624238344692451e-06,
+      "loss": 0.5883,
+      "step": 10992
+    },
+    {
+      "epoch": 2.918359219434488,
+      "grad_norm": 0.4608444081651207,
+      "learning_rate": 2.6238896469388203e-06,
+      "loss": 0.5838,
+      "step": 10993
+    },
+    {
+      "epoch": 2.918624717907872,
+      "grad_norm": 0.4445833888366233,
+      "learning_rate": 2.623540946769037e-06,
+      "loss": 0.5631,
+      "step": 10994
+    },
+    {
+      "epoch": 2.918890216381256,
+      "grad_norm": 0.4441179093401636,
+      "learning_rate": 2.623192244189903e-06,
+      "loss": 0.5746,
+      "step": 10995
+    },
+    {
+      "epoch": 2.9191557148546394,
+      "grad_norm": 0.4425067355772803,
+      "learning_rate": 2.6228435392082187e-06,
+      "loss": 0.565,
+      "step": 10996
+    },
+    {
+      "epoch": 2.9194212133280235,
+      "grad_norm": 0.45033100772278595,
+      "learning_rate": 2.6224948318307846e-06,
+      "loss": 0.5712,
+      "step": 10997
+    },
+    {
+      "epoch": 2.919686711801407,
+      "grad_norm": 0.4433845948136801,
+      "learning_rate": 2.6221461220644017e-06,
+      "loss": 0.5134,
+      "step": 10998
+    },
+    {
+      "epoch": 2.919952210274791,
+      "grad_norm": 0.4403039576170179,
+      "learning_rate": 2.62179740991587e-06,
+      "loss": 0.5766,
+      "step": 10999
+    },
+    {
+      "epoch": 2.920217708748175,
+      "grad_norm": 0.4433079939625189,
+      "learning_rate": 2.6214486953919905e-06,
+      "loss": 0.5615,
+      "step": 11000
+    },
+    {
+      "epoch": 2.9204832072215585,
+      "grad_norm": 0.42658088355665535,
+      "learning_rate": 2.6210999784995643e-06,
+      "loss": 0.5544,
+      "step": 11001
+    },
+    {
+      "epoch": 2.920748705694942,
+      "grad_norm": 0.44425514010645295,
+      "learning_rate": 2.6207512592453917e-06,
+      "loss": 0.5684,
+      "step": 11002
+    },
+    {
+      "epoch": 2.9210142041683262,
+      "grad_norm": 0.48177515772204954,
+      "learning_rate": 2.620402537636274e-06,
+      "loss": 0.5384,
+      "step": 11003
+    },
+    {
+      "epoch": 2.92127970264171,
+      "grad_norm": 0.4495142565900942,
+      "learning_rate": 2.6200538136790123e-06,
+      "loss": 0.5277,
+      "step": 11004
+    },
+    {
+      "epoch": 2.9215452011150935,
+      "grad_norm": 0.4631604728166726,
+      "learning_rate": 2.6197050873804063e-06,
+      "loss": 0.5524,
+      "step": 11005
+    },
+    {
+      "epoch": 2.9218106995884776,
+      "grad_norm": 0.4413804730273908,
+      "learning_rate": 2.6193563587472594e-06,
+      "loss": 0.5735,
+      "step": 11006
+    },
+    {
+      "epoch": 2.9220761980618613,
+      "grad_norm": 0.4542835779621849,
+      "learning_rate": 2.6190076277863706e-06,
+      "loss": 0.5928,
+      "step": 11007
+    },
+    {
+      "epoch": 2.922341696535245,
+      "grad_norm": 0.4478147334315177,
+      "learning_rate": 2.6186588945045417e-06,
+      "loss": 0.5593,
+      "step": 11008
+    },
+    {
+      "epoch": 2.922607195008629,
+      "grad_norm": 0.45343390685899143,
+      "learning_rate": 2.6183101589085735e-06,
+      "loss": 0.5522,
+      "step": 11009
+    },
+    {
+      "epoch": 2.9228726934820126,
+      "grad_norm": 0.44374661218782385,
+      "learning_rate": 2.6179614210052684e-06,
+      "loss": 0.5702,
+      "step": 11010
+    },
+    {
+      "epoch": 2.9231381919553963,
+      "grad_norm": 0.4438398912198788,
+      "learning_rate": 2.617612680801426e-06,
+      "loss": 0.5263,
+      "step": 11011
+    },
+    {
+      "epoch": 2.92340369042878,
+      "grad_norm": 0.43049070232928893,
+      "learning_rate": 2.617263938303849e-06,
+      "loss": 0.5437,
+      "step": 11012
+    },
+    {
+      "epoch": 2.9236691889021635,
+      "grad_norm": 0.4294351270321998,
+      "learning_rate": 2.616915193519337e-06,
+      "loss": 0.5328,
+      "step": 11013
+    },
+    {
+      "epoch": 2.9239346873755476,
+      "grad_norm": 0.45029589715941537,
+      "learning_rate": 2.6165664464546938e-06,
+      "loss": 0.5778,
+      "step": 11014
+    },
+    {
+      "epoch": 2.9242001858489313,
+      "grad_norm": 0.43297638513041714,
+      "learning_rate": 2.6162176971167188e-06,
+      "loss": 0.5347,
+      "step": 11015
+    },
+    {
+      "epoch": 2.924465684322315,
+      "grad_norm": 0.44810356443483884,
+      "learning_rate": 2.6158689455122145e-06,
+      "loss": 0.5747,
+      "step": 11016
+    },
+    {
+      "epoch": 2.924731182795699,
+      "grad_norm": 0.4488597834661214,
+      "learning_rate": 2.615520191647981e-06,
+      "loss": 0.5714,
+      "step": 11017
+    },
+    {
+      "epoch": 2.9249966812690826,
+      "grad_norm": 0.4461198690935552,
+      "learning_rate": 2.615171435530822e-06,
+      "loss": 0.5371,
+      "step": 11018
+    },
+    {
+      "epoch": 2.9252621797424663,
+      "grad_norm": 0.4281483751753625,
+      "learning_rate": 2.6148226771675384e-06,
+      "loss": 0.5613,
+      "step": 11019
+    },
+    {
+      "epoch": 2.9255276782158504,
+      "grad_norm": 0.44711760039536225,
+      "learning_rate": 2.6144739165649303e-06,
+      "loss": 0.5599,
+      "step": 11020
+    },
+    {
+      "epoch": 2.925793176689234,
+      "grad_norm": 0.45194636137474825,
+      "learning_rate": 2.6141251537298012e-06,
+      "loss": 0.5705,
+      "step": 11021
+    },
+    {
+      "epoch": 2.9260586751626176,
+      "grad_norm": 0.4549169775333854,
+      "learning_rate": 2.6137763886689517e-06,
+      "loss": 0.5363,
+      "step": 11022
+    },
+    {
+      "epoch": 2.9263241736360017,
+      "grad_norm": 0.4452998104532959,
+      "learning_rate": 2.613427621389184e-06,
+      "loss": 0.5263,
+      "step": 11023
+    },
+    {
+      "epoch": 2.9265896721093854,
+      "grad_norm": 0.43003348577132655,
+      "learning_rate": 2.613078851897301e-06,
+      "loss": 0.5648,
+      "step": 11024
+    },
+    {
+      "epoch": 2.926855170582769,
+      "grad_norm": 0.44389173885840344,
+      "learning_rate": 2.6127300802001025e-06,
+      "loss": 0.5466,
+      "step": 11025
+    },
+    {
+      "epoch": 2.927120669056153,
+      "grad_norm": 0.4287882558063798,
+      "learning_rate": 2.612381306304392e-06,
+      "loss": 0.528,
+      "step": 11026
+    },
+    {
+      "epoch": 2.9273861675295367,
+      "grad_norm": 0.4537860839216422,
+      "learning_rate": 2.6120325302169707e-06,
+      "loss": 0.5805,
+      "step": 11027
+    },
+    {
+      "epoch": 2.9276516660029204,
+      "grad_norm": 0.43466951533592624,
+      "learning_rate": 2.6116837519446407e-06,
+      "loss": 0.5524,
+      "step": 11028
+    },
+    {
+      "epoch": 2.9279171644763045,
+      "grad_norm": 0.43625863951260246,
+      "learning_rate": 2.611334971494204e-06,
+      "loss": 0.5735,
+      "step": 11029
+    },
+    {
+      "epoch": 2.928182662949688,
+      "grad_norm": 0.4389701497554812,
+      "learning_rate": 2.6109861888724638e-06,
+      "loss": 0.5218,
+      "step": 11030
+    },
+    {
+      "epoch": 2.9284481614230717,
+      "grad_norm": 0.4451170867562754,
+      "learning_rate": 2.6106374040862193e-06,
+      "loss": 0.5651,
+      "step": 11031
+    },
+    {
+      "epoch": 2.928713659896456,
+      "grad_norm": 0.444052952661565,
+      "learning_rate": 2.610288617142275e-06,
+      "loss": 0.5136,
+      "step": 11032
+    },
+    {
+      "epoch": 2.9289791583698395,
+      "grad_norm": 0.43545532118872726,
+      "learning_rate": 2.6099398280474337e-06,
+      "loss": 0.5682,
+      "step": 11033
+    },
+    {
+      "epoch": 2.929244656843223,
+      "grad_norm": 0.43339552867254505,
+      "learning_rate": 2.609591036808496e-06,
+      "loss": 0.5567,
+      "step": 11034
+    },
+    {
+      "epoch": 2.929510155316607,
+      "grad_norm": 0.4277448711729227,
+      "learning_rate": 2.6092422434322646e-06,
+      "loss": 0.5422,
+      "step": 11035
+    },
+    {
+      "epoch": 2.929775653789991,
+      "grad_norm": 0.43189615094205963,
+      "learning_rate": 2.6088934479255425e-06,
+      "loss": 0.5608,
+      "step": 11036
+    },
+    {
+      "epoch": 2.9300411522633745,
+      "grad_norm": 0.4349166586562321,
+      "learning_rate": 2.6085446502951313e-06,
+      "loss": 0.5619,
+      "step": 11037
+    },
+    {
+      "epoch": 2.930306650736758,
+      "grad_norm": 0.4355390012269658,
+      "learning_rate": 2.6081958505478332e-06,
+      "loss": 0.5342,
+      "step": 11038
+    },
+    {
+      "epoch": 2.930572149210142,
+      "grad_norm": 0.4441870301727794,
+      "learning_rate": 2.6078470486904515e-06,
+      "loss": 0.5557,
+      "step": 11039
+    },
+    {
+      "epoch": 2.930837647683526,
+      "grad_norm": 0.4525027666561795,
+      "learning_rate": 2.607498244729788e-06,
+      "loss": 0.5649,
+      "step": 11040
+    },
+    {
+      "epoch": 2.9311031461569095,
+      "grad_norm": 0.45828534235312707,
+      "learning_rate": 2.6071494386726458e-06,
+      "loss": 0.521,
+      "step": 11041
+    },
+    {
+      "epoch": 2.931368644630293,
+      "grad_norm": 0.4494191794458083,
+      "learning_rate": 2.606800630525827e-06,
+      "loss": 0.5422,
+      "step": 11042
+    },
+    {
+      "epoch": 2.931634143103677,
+      "grad_norm": 0.4392500616799736,
+      "learning_rate": 2.6064518202961347e-06,
+      "loss": 0.5353,
+      "step": 11043
+    },
+    {
+      "epoch": 2.931899641577061,
+      "grad_norm": 0.4341415948515654,
+      "learning_rate": 2.606103007990371e-06,
+      "loss": 0.5886,
+      "step": 11044
+    },
+    {
+      "epoch": 2.9321651400504445,
+      "grad_norm": 0.45245781210341557,
+      "learning_rate": 2.6057541936153392e-06,
+      "loss": 0.577,
+      "step": 11045
+    },
+    {
+      "epoch": 2.9324306385238286,
+      "grad_norm": 0.4784986902228421,
+      "learning_rate": 2.6054053771778413e-06,
+      "loss": 0.5643,
+      "step": 11046
+    },
+    {
+      "epoch": 2.932696136997212,
+      "grad_norm": 0.44408022124947666,
+      "learning_rate": 2.605056558684681e-06,
+      "loss": 0.5473,
+      "step": 11047
+    },
+    {
+      "epoch": 2.932961635470596,
+      "grad_norm": 0.43647939474380354,
+      "learning_rate": 2.60470773814266e-06,
+      "loss": 0.5407,
+      "step": 11048
+    },
+    {
+      "epoch": 2.93322713394398,
+      "grad_norm": 0.4519577409803698,
+      "learning_rate": 2.604358915558582e-06,
+      "loss": 0.5622,
+      "step": 11049
+    },
+    {
+      "epoch": 2.9334926324173636,
+      "grad_norm": 0.45201614717443855,
+      "learning_rate": 2.6040100909392498e-06,
+      "loss": 0.5651,
+      "step": 11050
+    },
+    {
+      "epoch": 2.933758130890747,
+      "grad_norm": 0.44193766833575415,
+      "learning_rate": 2.6036612642914656e-06,
+      "loss": 0.5925,
+      "step": 11051
+    },
+    {
+      "epoch": 2.9340236293641313,
+      "grad_norm": 0.4574321497212871,
+      "learning_rate": 2.603312435622033e-06,
+      "loss": 0.5609,
+      "step": 11052
+    },
+    {
+      "epoch": 2.934289127837515,
+      "grad_norm": 0.450262875763668,
+      "learning_rate": 2.602963604937756e-06,
+      "loss": 0.615,
+      "step": 11053
+    },
+    {
+      "epoch": 2.9345546263108986,
+      "grad_norm": 0.4518255935760746,
+      "learning_rate": 2.6026147722454355e-06,
+      "loss": 0.5779,
+      "step": 11054
+    },
+    {
+      "epoch": 2.9348201247842827,
+      "grad_norm": 0.4480873814632304,
+      "learning_rate": 2.602265937551876e-06,
+      "loss": 0.5265,
+      "step": 11055
+    },
+    {
+      "epoch": 2.9350856232576663,
+      "grad_norm": 0.4428034290581282,
+      "learning_rate": 2.601917100863881e-06,
+      "loss": 0.5916,
+      "step": 11056
+    },
+    {
+      "epoch": 2.93535112173105,
+      "grad_norm": 0.44940435041218696,
+      "learning_rate": 2.6015682621882523e-06,
+      "loss": 0.5668,
+      "step": 11057
+    },
+    {
+      "epoch": 2.935616620204434,
+      "grad_norm": 0.4399758352164834,
+      "learning_rate": 2.6012194215317938e-06,
+      "loss": 0.5799,
+      "step": 11058
+    },
+    {
+      "epoch": 2.9358821186778177,
+      "grad_norm": 0.43656769776092696,
+      "learning_rate": 2.6008705789013084e-06,
+      "loss": 0.5601,
+      "step": 11059
+    },
+    {
+      "epoch": 2.9361476171512013,
+      "grad_norm": 0.43168107847423265,
+      "learning_rate": 2.6005217343035997e-06,
+      "loss": 0.587,
+      "step": 11060
+    },
+    {
+      "epoch": 2.9364131156245854,
+      "grad_norm": 0.4325850623991144,
+      "learning_rate": 2.6001728877454717e-06,
+      "loss": 0.5576,
+      "step": 11061
+    },
+    {
+      "epoch": 2.936678614097969,
+      "grad_norm": 0.4351396388721992,
+      "learning_rate": 2.5998240392337274e-06,
+      "loss": 0.5458,
+      "step": 11062
+    },
+    {
+      "epoch": 2.9369441125713527,
+      "grad_norm": 0.4328401611566337,
+      "learning_rate": 2.5994751887751694e-06,
+      "loss": 0.5291,
+      "step": 11063
+    },
+    {
+      "epoch": 2.9372096110447368,
+      "grad_norm": 0.44967691269533355,
+      "learning_rate": 2.599126336376602e-06,
+      "loss": 0.5631,
+      "step": 11064
+    },
+    {
+      "epoch": 2.9374751095181204,
+      "grad_norm": 0.44010537129460875,
+      "learning_rate": 2.5987774820448282e-06,
+      "loss": 0.5743,
+      "step": 11065
+    },
+    {
+      "epoch": 2.937740607991504,
+      "grad_norm": 0.4550942766624417,
+      "learning_rate": 2.5984286257866513e-06,
+      "loss": 0.5237,
+      "step": 11066
+    },
+    {
+      "epoch": 2.9380061064648877,
+      "grad_norm": 0.45081598069681766,
+      "learning_rate": 2.5980797676088754e-06,
+      "loss": 0.5456,
+      "step": 11067
+    },
+    {
+      "epoch": 2.9382716049382713,
+      "grad_norm": 0.4480864310945247,
+      "learning_rate": 2.597730907518304e-06,
+      "loss": 0.5917,
+      "step": 11068
+    },
+    {
+      "epoch": 2.9385371034116554,
+      "grad_norm": 0.4550612706211758,
+      "learning_rate": 2.5973820455217404e-06,
+      "loss": 0.537,
+      "step": 11069
+    },
+    {
+      "epoch": 2.938802601885039,
+      "grad_norm": 0.44689868514503867,
+      "learning_rate": 2.5970331816259885e-06,
+      "loss": 0.562,
+      "step": 11070
+    },
+    {
+      "epoch": 2.9390681003584227,
+      "grad_norm": 0.4562716141300421,
+      "learning_rate": 2.5966843158378524e-06,
+      "loss": 0.5601,
+      "step": 11071
+    },
+    {
+      "epoch": 2.939333598831807,
+      "grad_norm": 0.4554198177235054,
+      "learning_rate": 2.596335448164135e-06,
+      "loss": 0.5914,
+      "step": 11072
+    },
+    {
+      "epoch": 2.9395990973051904,
+      "grad_norm": 0.42981824147669506,
+      "learning_rate": 2.5959865786116412e-06,
+      "loss": 0.5554,
+      "step": 11073
+    },
+    {
+      "epoch": 2.939864595778574,
+      "grad_norm": 0.44808024991165263,
+      "learning_rate": 2.595637707187173e-06,
+      "loss": 0.5871,
+      "step": 11074
+    },
+    {
+      "epoch": 2.940130094251958,
+      "grad_norm": 0.4543846402797517,
+      "learning_rate": 2.5952888338975356e-06,
+      "loss": 0.5888,
+      "step": 11075
+    },
+    {
+      "epoch": 2.940395592725342,
+      "grad_norm": 0.43603148151115634,
+      "learning_rate": 2.5949399587495333e-06,
+      "loss": 0.5759,
+      "step": 11076
+    },
+    {
+      "epoch": 2.9406610911987254,
+      "grad_norm": 0.4365660463555581,
+      "learning_rate": 2.5945910817499687e-06,
+      "loss": 0.5265,
+      "step": 11077
+    },
+    {
+      "epoch": 2.9409265896721095,
+      "grad_norm": 0.44096177315977575,
+      "learning_rate": 2.5942422029056464e-06,
+      "loss": 0.5612,
+      "step": 11078
+    },
+    {
+      "epoch": 2.941192088145493,
+      "grad_norm": 0.4436147080214331,
+      "learning_rate": 2.593893322223371e-06,
+      "loss": 0.5613,
+      "step": 11079
+    },
+    {
+      "epoch": 2.941457586618877,
+      "grad_norm": 0.4280195742595993,
+      "learning_rate": 2.593544439709945e-06,
+      "loss": 0.5553,
+      "step": 11080
+    },
+    {
+      "epoch": 2.941723085092261,
+      "grad_norm": 0.44350844930564665,
+      "learning_rate": 2.5931955553721745e-06,
+      "loss": 0.5442,
+      "step": 11081
+    },
+    {
+      "epoch": 2.9419885835656445,
+      "grad_norm": 0.43582046238552663,
+      "learning_rate": 2.592846669216862e-06,
+      "loss": 0.5755,
+      "step": 11082
+    },
+    {
+      "epoch": 2.942254082039028,
+      "grad_norm": 0.44460755027030535,
+      "learning_rate": 2.5924977812508124e-06,
+      "loss": 0.5743,
+      "step": 11083
+    },
+    {
+      "epoch": 2.9425195805124122,
+      "grad_norm": 0.4451981153873848,
+      "learning_rate": 2.592148891480829e-06,
+      "loss": 0.565,
+      "step": 11084
+    },
+    {
+      "epoch": 2.942785078985796,
+      "grad_norm": 0.43662527942919516,
+      "learning_rate": 2.5917999999137173e-06,
+      "loss": 0.5599,
+      "step": 11085
+    },
+    {
+      "epoch": 2.9430505774591795,
+      "grad_norm": 0.44888326471622875,
+      "learning_rate": 2.5914511065562804e-06,
+      "loss": 0.5437,
+      "step": 11086
+    },
+    {
+      "epoch": 2.9433160759325636,
+      "grad_norm": 0.43990252224141785,
+      "learning_rate": 2.5911022114153224e-06,
+      "loss": 0.5601,
+      "step": 11087
+    },
+    {
+      "epoch": 2.9435815744059473,
+      "grad_norm": 0.4423886506568044,
+      "learning_rate": 2.590753314497649e-06,
+      "loss": 0.5671,
+      "step": 11088
+    },
+    {
+      "epoch": 2.943847072879331,
+      "grad_norm": 0.43678266037343344,
+      "learning_rate": 2.590404415810063e-06,
+      "loss": 0.5929,
+      "step": 11089
+    },
+    {
+      "epoch": 2.944112571352715,
+      "grad_norm": 0.4418658032055756,
+      "learning_rate": 2.5900555153593703e-06,
+      "loss": 0.5356,
+      "step": 11090
+    },
+    {
+      "epoch": 2.9443780698260986,
+      "grad_norm": 0.4483786560026015,
+      "learning_rate": 2.589706613152375e-06,
+      "loss": 0.5808,
+      "step": 11091
+    },
+    {
+      "epoch": 2.9446435682994823,
+      "grad_norm": 0.4491662270321082,
+      "learning_rate": 2.5893577091958805e-06,
+      "loss": 0.5908,
+      "step": 11092
+    },
+    {
+      "epoch": 2.944909066772866,
+      "grad_norm": 0.44214251950510036,
+      "learning_rate": 2.5890088034966916e-06,
+      "loss": 0.5513,
+      "step": 11093
+    },
+    {
+      "epoch": 2.94517456524625,
+      "grad_norm": 0.44011214924044634,
+      "learning_rate": 2.5886598960616137e-06,
+      "loss": 0.5426,
+      "step": 11094
+    },
+    {
+      "epoch": 2.9454400637196336,
+      "grad_norm": 0.4407352441614994,
+      "learning_rate": 2.5883109868974506e-06,
+      "loss": 0.541,
+      "step": 11095
+    },
+    {
+      "epoch": 2.9457055621930173,
+      "grad_norm": 0.4504348396439044,
+      "learning_rate": 2.5879620760110065e-06,
+      "loss": 0.5432,
+      "step": 11096
+    },
+    {
+      "epoch": 2.945971060666401,
+      "grad_norm": 0.45832096799768707,
+      "learning_rate": 2.587613163409087e-06,
+      "loss": 0.5929,
+      "step": 11097
+    },
+    {
+      "epoch": 2.946236559139785,
+      "grad_norm": 0.4436309234137818,
+      "learning_rate": 2.5872642490984955e-06,
+      "loss": 0.5539,
+      "step": 11098
+    },
+    {
+      "epoch": 2.9465020576131686,
+      "grad_norm": 0.4512333676411674,
+      "learning_rate": 2.5869153330860393e-06,
+      "loss": 0.6047,
+      "step": 11099
+    },
+    {
+      "epoch": 2.9467675560865523,
+      "grad_norm": 0.45423819386854863,
+      "learning_rate": 2.5865664153785197e-06,
+      "loss": 0.5884,
+      "step": 11100
+    },
+    {
+      "epoch": 2.9470330545599364,
+      "grad_norm": 0.4558499481753614,
+      "learning_rate": 2.5862174959827437e-06,
+      "loss": 0.5641,
+      "step": 11101
+    },
+    {
+      "epoch": 2.94729855303332,
+      "grad_norm": 0.449877721033026,
+      "learning_rate": 2.5858685749055157e-06,
+      "loss": 0.5557,
+      "step": 11102
+    },
+    {
+      "epoch": 2.9475640515067036,
+      "grad_norm": 0.4346694747806558,
+      "learning_rate": 2.5855196521536397e-06,
+      "loss": 0.5674,
+      "step": 11103
+    },
+    {
+      "epoch": 2.9478295499800877,
+      "grad_norm": 0.4391238912896865,
+      "learning_rate": 2.585170727733921e-06,
+      "loss": 0.5483,
+      "step": 11104
+    },
+    {
+      "epoch": 2.9480950484534714,
+      "grad_norm": 0.4323549237160803,
+      "learning_rate": 2.5848218016531645e-06,
+      "loss": 0.5355,
+      "step": 11105
+    },
+    {
+      "epoch": 2.948360546926855,
+      "grad_norm": 0.43725840323667203,
+      "learning_rate": 2.584472873918176e-06,
+      "loss": 0.5308,
+      "step": 11106
+    },
+    {
+      "epoch": 2.948626045400239,
+      "grad_norm": 0.4456587492658111,
+      "learning_rate": 2.5841239445357585e-06,
+      "loss": 0.5741,
+      "step": 11107
+    },
+    {
+      "epoch": 2.9488915438736227,
+      "grad_norm": 0.4450286369348951,
+      "learning_rate": 2.5837750135127192e-06,
+      "loss": 0.5413,
+      "step": 11108
+    },
+    {
+      "epoch": 2.9491570423470064,
+      "grad_norm": 0.43196300493471296,
+      "learning_rate": 2.5834260808558614e-06,
+      "loss": 0.5783,
+      "step": 11109
+    },
+    {
+      "epoch": 2.9494225408203905,
+      "grad_norm": 0.46152166268122263,
+      "learning_rate": 2.5830771465719905e-06,
+      "loss": 0.6018,
+      "step": 11110
+    },
+    {
+      "epoch": 2.949688039293774,
+      "grad_norm": 0.44343546298115655,
+      "learning_rate": 2.582728210667913e-06,
+      "loss": 0.5831,
+      "step": 11111
+    },
+    {
+      "epoch": 2.9499535377671577,
+      "grad_norm": 0.4274778167214706,
+      "learning_rate": 2.5823792731504315e-06,
+      "loss": 0.5372,
+      "step": 11112
+    },
+    {
+      "epoch": 2.950219036240542,
+      "grad_norm": 0.4261531732814335,
+      "learning_rate": 2.582030334026353e-06,
+      "loss": 0.5071,
+      "step": 11113
+    },
+    {
+      "epoch": 2.9504845347139255,
+      "grad_norm": 0.4542437566840947,
+      "learning_rate": 2.5816813933024826e-06,
+      "loss": 0.5482,
+      "step": 11114
+    },
+    {
+      "epoch": 2.950750033187309,
+      "grad_norm": 0.45177727370652365,
+      "learning_rate": 2.581332450985624e-06,
+      "loss": 0.5661,
+      "step": 11115
+    },
+    {
+      "epoch": 2.951015531660693,
+      "grad_norm": 0.4460958136793645,
+      "learning_rate": 2.5809835070825844e-06,
+      "loss": 0.5523,
+      "step": 11116
+    },
+    {
+      "epoch": 2.951281030134077,
+      "grad_norm": 0.4475475966169349,
+      "learning_rate": 2.5806345616001684e-06,
+      "loss": 0.5379,
+      "step": 11117
+    },
+    {
+      "epoch": 2.9515465286074605,
+      "grad_norm": 0.4545354377571334,
+      "learning_rate": 2.58028561454518e-06,
+      "loss": 0.512,
+      "step": 11118
+    },
+    {
+      "epoch": 2.9518120270808446,
+      "grad_norm": 0.4411803945683953,
+      "learning_rate": 2.5799366659244263e-06,
+      "loss": 0.5623,
+      "step": 11119
+    },
+    {
+      "epoch": 2.952077525554228,
+      "grad_norm": 0.45280608820145873,
+      "learning_rate": 2.579587715744712e-06,
+      "loss": 0.5693,
+      "step": 11120
+    },
+    {
+      "epoch": 2.952343024027612,
+      "grad_norm": 0.4474194631108395,
+      "learning_rate": 2.5792387640128413e-06,
+      "loss": 0.5775,
+      "step": 11121
+    },
+    {
+      "epoch": 2.9526085225009955,
+      "grad_norm": 0.43979584780312064,
+      "learning_rate": 2.5788898107356215e-06,
+      "loss": 0.5617,
+      "step": 11122
+    },
+    {
+      "epoch": 2.952874020974379,
+      "grad_norm": 0.4509841326931812,
+      "learning_rate": 2.578540855919857e-06,
+      "loss": 0.5663,
+      "step": 11123
+    },
+    {
+      "epoch": 2.953139519447763,
+      "grad_norm": 0.4650130743622842,
+      "learning_rate": 2.578191899572353e-06,
+      "loss": 0.5489,
+      "step": 11124
+    },
+    {
+      "epoch": 2.953405017921147,
+      "grad_norm": 0.43224846161431096,
+      "learning_rate": 2.577842941699917e-06,
+      "loss": 0.5311,
+      "step": 11125
+    },
+    {
+      "epoch": 2.9536705163945305,
+      "grad_norm": 0.42911234841879764,
+      "learning_rate": 2.577493982309352e-06,
+      "loss": 0.5511,
+      "step": 11126
+    },
+    {
+      "epoch": 2.9539360148679146,
+      "grad_norm": 0.4297791775910588,
+      "learning_rate": 2.5771450214074644e-06,
+      "loss": 0.527,
+      "step": 11127
+    },
+    {
+      "epoch": 2.954201513341298,
+      "grad_norm": 0.4443164092023868,
+      "learning_rate": 2.576796059001061e-06,
+      "loss": 0.5709,
+      "step": 11128
+    },
+    {
+      "epoch": 2.954467011814682,
+      "grad_norm": 0.45503776831619025,
+      "learning_rate": 2.5764470950969457e-06,
+      "loss": 0.5533,
+      "step": 11129
+    },
+    {
+      "epoch": 2.954732510288066,
+      "grad_norm": 0.4585870546941405,
+      "learning_rate": 2.576098129701925e-06,
+      "loss": 0.5338,
+      "step": 11130
+    },
+    {
+      "epoch": 2.9549980087614496,
+      "grad_norm": 0.41715428023443785,
+      "learning_rate": 2.575749162822805e-06,
+      "loss": 0.5237,
+      "step": 11131
+    },
+    {
+      "epoch": 2.955263507234833,
+      "grad_norm": 0.4537409310070673,
+      "learning_rate": 2.57540019446639e-06,
+      "loss": 0.5458,
+      "step": 11132
+    },
+    {
+      "epoch": 2.9555290057082173,
+      "grad_norm": 0.4460184122879938,
+      "learning_rate": 2.575051224639486e-06,
+      "loss": 0.584,
+      "step": 11133
+    },
+    {
+      "epoch": 2.955794504181601,
+      "grad_norm": 0.45158707179764196,
+      "learning_rate": 2.5747022533489006e-06,
+      "loss": 0.5677,
+      "step": 11134
+    },
+    {
+      "epoch": 2.9560600026549846,
+      "grad_norm": 0.4564504884895011,
+      "learning_rate": 2.574353280601438e-06,
+      "loss": 0.5886,
+      "step": 11135
+    },
+    {
+      "epoch": 2.9563255011283687,
+      "grad_norm": 0.4499236932639771,
+      "learning_rate": 2.5740043064039043e-06,
+      "loss": 0.5735,
+      "step": 11136
+    },
+    {
+      "epoch": 2.9565909996017523,
+      "grad_norm": 0.4448097964990401,
+      "learning_rate": 2.5736553307631055e-06,
+      "loss": 0.5739,
+      "step": 11137
+    },
+    {
+      "epoch": 2.956856498075136,
+      "grad_norm": 0.44452156301113244,
+      "learning_rate": 2.5733063536858477e-06,
+      "loss": 0.5546,
+      "step": 11138
+    },
+    {
+      "epoch": 2.95712199654852,
+      "grad_norm": 0.4651078589480799,
+      "learning_rate": 2.572957375178936e-06,
+      "loss": 0.5833,
+      "step": 11139
+    },
+    {
+      "epoch": 2.9573874950219037,
+      "grad_norm": 0.44711143349524607,
+      "learning_rate": 2.5726083952491774e-06,
+      "loss": 0.5643,
+      "step": 11140
+    },
+    {
+      "epoch": 2.9576529934952873,
+      "grad_norm": 0.4360262731811912,
+      "learning_rate": 2.5722594139033767e-06,
+      "loss": 0.5563,
+      "step": 11141
+    },
+    {
+      "epoch": 2.9579184919686714,
+      "grad_norm": 0.44083417525704455,
+      "learning_rate": 2.571910431148341e-06,
+      "loss": 0.5837,
+      "step": 11142
+    },
+    {
+      "epoch": 2.958183990442055,
+      "grad_norm": 0.45805104980099365,
+      "learning_rate": 2.5715614469908754e-06,
+      "loss": 0.5638,
+      "step": 11143
+    },
+    {
+      "epoch": 2.9584494889154387,
+      "grad_norm": 0.4824903499669946,
+      "learning_rate": 2.5712124614377863e-06,
+      "loss": 0.5778,
+      "step": 11144
+    },
+    {
+      "epoch": 2.9587149873888228,
+      "grad_norm": 0.44034568417154685,
+      "learning_rate": 2.570863474495881e-06,
+      "loss": 0.61,
+      "step": 11145
+    },
+    {
+      "epoch": 2.9589804858622064,
+      "grad_norm": 0.4442593600810979,
+      "learning_rate": 2.5705144861719634e-06,
+      "loss": 0.5399,
+      "step": 11146
+    },
+    {
+      "epoch": 2.95924598433559,
+      "grad_norm": 0.44356610250779144,
+      "learning_rate": 2.5701654964728406e-06,
+      "loss": 0.5627,
+      "step": 11147
+    },
+    {
+      "epoch": 2.9595114828089737,
+      "grad_norm": 0.4370759971601249,
+      "learning_rate": 2.5698165054053197e-06,
+      "loss": 0.557,
+      "step": 11148
+    },
+    {
+      "epoch": 2.9597769812823578,
+      "grad_norm": 0.44161034389561826,
+      "learning_rate": 2.5694675129762055e-06,
+      "loss": 0.5548,
+      "step": 11149
+    },
+    {
+      "epoch": 2.9600424797557414,
+      "grad_norm": 0.4473385732591696,
+      "learning_rate": 2.569118519192304e-06,
+      "loss": 0.5403,
+      "step": 11150
+    },
+    {
+      "epoch": 2.960307978229125,
+      "grad_norm": 0.44767028642156964,
+      "learning_rate": 2.5687695240604225e-06,
+      "loss": 0.5516,
+      "step": 11151
+    },
+    {
+      "epoch": 2.9605734767025087,
+      "grad_norm": 0.4578836447155603,
+      "learning_rate": 2.5684205275873677e-06,
+      "loss": 0.5414,
+      "step": 11152
+    },
+    {
+      "epoch": 2.960838975175893,
+      "grad_norm": 0.45567472316892366,
+      "learning_rate": 2.568071529779944e-06,
+      "loss": 0.5698,
+      "step": 11153
+    },
+    {
+      "epoch": 2.9611044736492764,
+      "grad_norm": 0.44852035417122443,
+      "learning_rate": 2.56772253064496e-06,
+      "loss": 0.5387,
+      "step": 11154
+    },
+    {
+      "epoch": 2.96136997212266,
+      "grad_norm": 0.45738140434987584,
+      "learning_rate": 2.56737353018922e-06,
+      "loss": 0.5341,
+      "step": 11155
+    },
+    {
+      "epoch": 2.961635470596044,
+      "grad_norm": 0.4444677584711385,
+      "learning_rate": 2.567024528419531e-06,
+      "loss": 0.5276,
+      "step": 11156
+    },
+    {
+      "epoch": 2.961900969069428,
+      "grad_norm": 0.4585276386673583,
+      "learning_rate": 2.5666755253427006e-06,
+      "loss": 0.5658,
+      "step": 11157
+    },
+    {
+      "epoch": 2.9621664675428114,
+      "grad_norm": 0.4457923312854235,
+      "learning_rate": 2.5663265209655337e-06,
+      "loss": 0.5474,
+      "step": 11158
+    },
+    {
+      "epoch": 2.9624319660161955,
+      "grad_norm": 0.44070254998793534,
+      "learning_rate": 2.565977515294837e-06,
+      "loss": 0.5643,
+      "step": 11159
+    },
+    {
+      "epoch": 2.962697464489579,
+      "grad_norm": 0.4482683100572228,
+      "learning_rate": 2.5656285083374165e-06,
+      "loss": 0.5086,
+      "step": 11160
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 0.42554826850508143,
+      "learning_rate": 2.5652795001000807e-06,
+      "loss": 0.5073,
+      "step": 11161
+    },
+    {
+      "epoch": 2.963228461436347,
+      "grad_norm": 0.44158487755664483,
+      "learning_rate": 2.5649304905896337e-06,
+      "loss": 0.5523,
+      "step": 11162
+    },
+    {
+      "epoch": 2.9634939599097305,
+      "grad_norm": 0.4308073564463114,
+      "learning_rate": 2.564581479812884e-06,
+      "loss": 0.5206,
+      "step": 11163
+    },
+    {
+      "epoch": 2.963759458383114,
+      "grad_norm": 0.44033185425820426,
+      "learning_rate": 2.564232467776638e-06,
+      "loss": 0.5759,
+      "step": 11164
+    },
+    {
+      "epoch": 2.9640249568564982,
+      "grad_norm": 0.4433405727176709,
+      "learning_rate": 2.5638834544877e-06,
+      "loss": 0.5296,
+      "step": 11165
+    },
+    {
+      "epoch": 2.964290455329882,
+      "grad_norm": 0.43797905620138805,
+      "learning_rate": 2.5635344399528792e-06,
+      "loss": 0.5272,
+      "step": 11166
+    },
+    {
+      "epoch": 2.9645559538032655,
+      "grad_norm": 0.44396817666226107,
+      "learning_rate": 2.563185424178981e-06,
+      "loss": 0.5521,
+      "step": 11167
+    },
+    {
+      "epoch": 2.9648214522766496,
+      "grad_norm": 0.4362669506909384,
+      "learning_rate": 2.5628364071728124e-06,
+      "loss": 0.5759,
+      "step": 11168
+    },
+    {
+      "epoch": 2.9650869507500333,
+      "grad_norm": 0.4427213236912875,
+      "learning_rate": 2.56248738894118e-06,
+      "loss": 0.5669,
+      "step": 11169
+    },
+    {
+      "epoch": 2.965352449223417,
+      "grad_norm": 0.44995568915927986,
+      "learning_rate": 2.56213836949089e-06,
+      "loss": 0.5442,
+      "step": 11170
+    },
+    {
+      "epoch": 2.965617947696801,
+      "grad_norm": 0.444973114730161,
+      "learning_rate": 2.561789348828751e-06,
+      "loss": 0.5112,
+      "step": 11171
+    },
+    {
+      "epoch": 2.9658834461701846,
+      "grad_norm": 0.4452493064924916,
+      "learning_rate": 2.561440326961567e-06,
+      "loss": 0.54,
+      "step": 11172
+    },
+    {
+      "epoch": 2.9661489446435683,
+      "grad_norm": 0.4434570878155933,
+      "learning_rate": 2.561091303896147e-06,
+      "loss": 0.5669,
+      "step": 11173
+    },
+    {
+      "epoch": 2.9664144431169523,
+      "grad_norm": 0.4491126738307586,
+      "learning_rate": 2.5607422796392973e-06,
+      "loss": 0.532,
+      "step": 11174
+    },
+    {
+      "epoch": 2.966679941590336,
+      "grad_norm": 0.44919851989330994,
+      "learning_rate": 2.5603932541978232e-06,
+      "loss": 0.5703,
+      "step": 11175
+    },
+    {
+      "epoch": 2.9669454400637196,
+      "grad_norm": 0.45269339791523555,
+      "learning_rate": 2.5600442275785337e-06,
+      "loss": 0.565,
+      "step": 11176
+    },
+    {
+      "epoch": 2.9672109385371033,
+      "grad_norm": 0.4475549090862826,
+      "learning_rate": 2.5596951997882348e-06,
+      "loss": 0.5552,
+      "step": 11177
+    },
+    {
+      "epoch": 2.9674764370104874,
+      "grad_norm": 0.43623797402419323,
+      "learning_rate": 2.5593461708337326e-06,
+      "loss": 0.532,
+      "step": 11178
+    },
+    {
+      "epoch": 2.967741935483871,
+      "grad_norm": 0.4439721671894238,
+      "learning_rate": 2.558997140721835e-06,
+      "loss": 0.5377,
+      "step": 11179
+    },
+    {
+      "epoch": 2.9680074339572546,
+      "grad_norm": 0.4574268675798015,
+      "learning_rate": 2.5586481094593492e-06,
+      "loss": 0.5737,
+      "step": 11180
+    },
+    {
+      "epoch": 2.9682729324306383,
+      "grad_norm": 0.4368260741932774,
+      "learning_rate": 2.5582990770530815e-06,
+      "loss": 0.5712,
+      "step": 11181
+    },
+    {
+      "epoch": 2.9685384309040224,
+      "grad_norm": 0.44386028892392754,
+      "learning_rate": 2.5579500435098385e-06,
+      "loss": 0.5633,
+      "step": 11182
+    },
+    {
+      "epoch": 2.968803929377406,
+      "grad_norm": 0.4404921604657042,
+      "learning_rate": 2.557601008836429e-06,
+      "loss": 0.5375,
+      "step": 11183
+    },
+    {
+      "epoch": 2.9690694278507896,
+      "grad_norm": 0.4403208807612028,
+      "learning_rate": 2.557251973039658e-06,
+      "loss": 0.513,
+      "step": 11184
+    },
+    {
+      "epoch": 2.9693349263241737,
+      "grad_norm": 0.4474727524533836,
+      "learning_rate": 2.5569029361263325e-06,
+      "loss": 0.5774,
+      "step": 11185
+    },
+    {
+      "epoch": 2.9696004247975574,
+      "grad_norm": 0.4384155398363317,
+      "learning_rate": 2.556553898103262e-06,
+      "loss": 0.566,
+      "step": 11186
+    },
+    {
+      "epoch": 2.969865923270941,
+      "grad_norm": 0.4425439239664622,
+      "learning_rate": 2.556204858977251e-06,
+      "loss": 0.5788,
+      "step": 11187
+    },
+    {
+      "epoch": 2.970131421744325,
+      "grad_norm": 0.4364060757202946,
+      "learning_rate": 2.555855818755108e-06,
+      "loss": 0.5988,
+      "step": 11188
+    },
+    {
+      "epoch": 2.9703969202177087,
+      "grad_norm": 0.4404714440366446,
+      "learning_rate": 2.5555067774436393e-06,
+      "loss": 0.5453,
+      "step": 11189
+    },
+    {
+      "epoch": 2.9706624186910924,
+      "grad_norm": 0.43708389691028626,
+      "learning_rate": 2.5551577350496533e-06,
+      "loss": 0.551,
+      "step": 11190
+    },
+    {
+      "epoch": 2.9709279171644765,
+      "grad_norm": 0.452898497623725,
+      "learning_rate": 2.5548086915799562e-06,
+      "loss": 0.5864,
+      "step": 11191
+    },
+    {
+      "epoch": 2.97119341563786,
+      "grad_norm": 0.44190403868135003,
+      "learning_rate": 2.5544596470413552e-06,
+      "loss": 0.5398,
+      "step": 11192
+    },
+    {
+      "epoch": 2.9714589141112437,
+      "grad_norm": 0.4457609789472404,
+      "learning_rate": 2.5541106014406575e-06,
+      "loss": 0.5442,
+      "step": 11193
+    },
+    {
+      "epoch": 2.971724412584628,
+      "grad_norm": 0.4327364815287397,
+      "learning_rate": 2.553761554784671e-06,
+      "loss": 0.567,
+      "step": 11194
+    },
+    {
+      "epoch": 2.9719899110580115,
+      "grad_norm": 0.42877446518098794,
+      "learning_rate": 2.553412507080203e-06,
+      "loss": 0.5107,
+      "step": 11195
+    },
+    {
+      "epoch": 2.972255409531395,
+      "grad_norm": 0.4419232045802643,
+      "learning_rate": 2.553063458334059e-06,
+      "loss": 0.5847,
+      "step": 11196
+    },
+    {
+      "epoch": 2.972520908004779,
+      "grad_norm": 0.44555691797066427,
+      "learning_rate": 2.5527144085530485e-06,
+      "loss": 0.5535,
+      "step": 11197
+    },
+    {
+      "epoch": 2.972786406478163,
+      "grad_norm": 0.44789682559753,
+      "learning_rate": 2.5523653577439778e-06,
+      "loss": 0.5201,
+      "step": 11198
+    },
+    {
+      "epoch": 2.9730519049515465,
+      "grad_norm": 0.4371590915579939,
+      "learning_rate": 2.5520163059136543e-06,
+      "loss": 0.5367,
+      "step": 11199
+    },
+    {
+      "epoch": 2.9733174034249306,
+      "grad_norm": 0.4566851955363735,
+      "learning_rate": 2.5516672530688864e-06,
+      "loss": 0.5719,
+      "step": 11200
+    },
+    {
+      "epoch": 2.973582901898314,
+      "grad_norm": 0.4546836277741624,
+      "learning_rate": 2.5513181992164797e-06,
+      "loss": 0.5584,
+      "step": 11201
+    },
+    {
+      "epoch": 2.973848400371698,
+      "grad_norm": 0.4672614081173667,
+      "learning_rate": 2.550969144363242e-06,
+      "loss": 0.5675,
+      "step": 11202
+    },
+    {
+      "epoch": 2.9741138988450815,
+      "grad_norm": 0.45308862699834485,
+      "learning_rate": 2.5506200885159826e-06,
+      "loss": 0.6106,
+      "step": 11203
+    },
+    {
+      "epoch": 2.9743793973184656,
+      "grad_norm": 0.42813710358242996,
+      "learning_rate": 2.5502710316815066e-06,
+      "loss": 0.5551,
+      "step": 11204
+    },
+    {
+      "epoch": 2.974644895791849,
+      "grad_norm": 0.4573943558674008,
+      "learning_rate": 2.5499219738666226e-06,
+      "loss": 0.6021,
+      "step": 11205
+    },
+    {
+      "epoch": 2.974910394265233,
+      "grad_norm": 0.441446808633667,
+      "learning_rate": 2.5495729150781373e-06,
+      "loss": 0.5412,
+      "step": 11206
+    },
+    {
+      "epoch": 2.9751758927386165,
+      "grad_norm": 0.44655667804289145,
+      "learning_rate": 2.5492238553228592e-06,
+      "loss": 0.597,
+      "step": 11207
+    },
+    {
+      "epoch": 2.9754413912120006,
+      "grad_norm": 0.4427038334510096,
+      "learning_rate": 2.5488747946075955e-06,
+      "loss": 0.5816,
+      "step": 11208
+    },
+    {
+      "epoch": 2.975706889685384,
+      "grad_norm": 0.4373391851956795,
+      "learning_rate": 2.5485257329391537e-06,
+      "loss": 0.5514,
+      "step": 11209
+    },
+    {
+      "epoch": 2.975972388158768,
+      "grad_norm": 0.45132898161562895,
+      "learning_rate": 2.5481766703243415e-06,
+      "loss": 0.5598,
+      "step": 11210
+    },
+    {
+      "epoch": 2.976237886632152,
+      "grad_norm": 0.446054565972664,
+      "learning_rate": 2.547827606769966e-06,
+      "loss": 0.5303,
+      "step": 11211
+    },
+    {
+      "epoch": 2.9765033851055356,
+      "grad_norm": 0.44279398414427396,
+      "learning_rate": 2.5474785422828354e-06,
+      "loss": 0.5809,
+      "step": 11212
+    },
+    {
+      "epoch": 2.9767688835789192,
+      "grad_norm": 0.44780435073734165,
+      "learning_rate": 2.5471294768697562e-06,
+      "loss": 0.5702,
+      "step": 11213
+    },
+    {
+      "epoch": 2.9770343820523033,
+      "grad_norm": 0.4491173114669352,
+      "learning_rate": 2.546780410537537e-06,
+      "loss": 0.5887,
+      "step": 11214
+    },
+    {
+      "epoch": 2.977299880525687,
+      "grad_norm": 0.4363332678582455,
+      "learning_rate": 2.5464313432929855e-06,
+      "loss": 0.5657,
+      "step": 11215
+    },
+    {
+      "epoch": 2.9775653789990706,
+      "grad_norm": 0.4379776537135228,
+      "learning_rate": 2.546082275142909e-06,
+      "loss": 0.5572,
+      "step": 11216
+    },
+    {
+      "epoch": 2.9778308774724547,
+      "grad_norm": 0.44327443102852543,
+      "learning_rate": 2.545733206094116e-06,
+      "loss": 0.5362,
+      "step": 11217
+    },
+    {
+      "epoch": 2.9780963759458383,
+      "grad_norm": 0.4451125827389643,
+      "learning_rate": 2.5453841361534128e-06,
+      "loss": 0.5724,
+      "step": 11218
+    },
+    {
+      "epoch": 2.978361874419222,
+      "grad_norm": 0.4455697676567192,
+      "learning_rate": 2.545035065327608e-06,
+      "loss": 0.5762,
+      "step": 11219
+    },
+    {
+      "epoch": 2.978627372892606,
+      "grad_norm": 0.447484012520238,
+      "learning_rate": 2.5446859936235097e-06,
+      "loss": 0.5547,
+      "step": 11220
+    },
+    {
+      "epoch": 2.9788928713659897,
+      "grad_norm": 0.4446284701393264,
+      "learning_rate": 2.544336921047924e-06,
+      "loss": 0.5489,
+      "step": 11221
+    },
+    {
+      "epoch": 2.9791583698393733,
+      "grad_norm": 0.4401726567402075,
+      "learning_rate": 2.5439878476076604e-06,
+      "loss": 0.5132,
+      "step": 11222
+    },
+    {
+      "epoch": 2.9794238683127574,
+      "grad_norm": 0.4322749602899099,
+      "learning_rate": 2.543638773309526e-06,
+      "loss": 0.5397,
+      "step": 11223
+    },
+    {
+      "epoch": 2.979689366786141,
+      "grad_norm": 0.4384372150760701,
+      "learning_rate": 2.5432896981603285e-06,
+      "loss": 0.5623,
+      "step": 11224
+    },
+    {
+      "epoch": 2.9799548652595247,
+      "grad_norm": 0.4431273328655571,
+      "learning_rate": 2.5429406221668757e-06,
+      "loss": 0.5351,
+      "step": 11225
+    },
+    {
+      "epoch": 2.9802203637329088,
+      "grad_norm": 0.4422551689912942,
+      "learning_rate": 2.5425915453359766e-06,
+      "loss": 0.5305,
+      "step": 11226
+    },
+    {
+      "epoch": 2.9804858622062924,
+      "grad_norm": 0.46812144802553396,
+      "learning_rate": 2.5422424676744373e-06,
+      "loss": 0.5909,
+      "step": 11227
+    },
+    {
+      "epoch": 2.980751360679676,
+      "grad_norm": 0.45310916069138585,
+      "learning_rate": 2.5418933891890665e-06,
+      "loss": 0.5493,
+      "step": 11228
+    },
+    {
+      "epoch": 2.98101685915306,
+      "grad_norm": 0.4482978301963584,
+      "learning_rate": 2.5415443098866727e-06,
+      "loss": 0.5534,
+      "step": 11229
+    },
+    {
+      "epoch": 2.981282357626444,
+      "grad_norm": 0.44932228990485235,
+      "learning_rate": 2.541195229774062e-06,
+      "loss": 0.5749,
+      "step": 11230
+    },
+    {
+      "epoch": 2.9815478560998274,
+      "grad_norm": 0.4423315263556984,
+      "learning_rate": 2.5408461488580444e-06,
+      "loss": 0.5697,
+      "step": 11231
+    },
+    {
+      "epoch": 2.981813354573211,
+      "grad_norm": 0.44244361038119795,
+      "learning_rate": 2.5404970671454272e-06,
+      "loss": 0.5822,
+      "step": 11232
+    },
+    {
+      "epoch": 2.982078853046595,
+      "grad_norm": 0.4357755471056965,
+      "learning_rate": 2.540147984643017e-06,
+      "loss": 0.5639,
+      "step": 11233
+    },
+    {
+      "epoch": 2.982344351519979,
+      "grad_norm": 0.43880735247830027,
+      "learning_rate": 2.5397989013576223e-06,
+      "loss": 0.5741,
+      "step": 11234
+    },
+    {
+      "epoch": 2.9826098499933624,
+      "grad_norm": 0.4415388720919388,
+      "learning_rate": 2.539449817296053e-06,
+      "loss": 0.5469,
+      "step": 11235
+    },
+    {
+      "epoch": 2.982875348466746,
+      "grad_norm": 0.44034337085961955,
+      "learning_rate": 2.539100732465115e-06,
+      "loss": 0.5689,
+      "step": 11236
+    },
+    {
+      "epoch": 2.98314084694013,
+      "grad_norm": 0.44986388773439695,
+      "learning_rate": 2.538751646871617e-06,
+      "loss": 0.5684,
+      "step": 11237
+    },
+    {
+      "epoch": 2.983406345413514,
+      "grad_norm": 0.44472683084986,
+      "learning_rate": 2.538402560522368e-06,
+      "loss": 0.5702,
+      "step": 11238
+    },
+    {
+      "epoch": 2.9836718438868974,
+      "grad_norm": 0.45826683300142207,
+      "learning_rate": 2.5380534734241733e-06,
+      "loss": 0.6002,
+      "step": 11239
+    },
+    {
+      "epoch": 2.9839373423602815,
+      "grad_norm": 0.4432175779110682,
+      "learning_rate": 2.5377043855838446e-06,
+      "loss": 0.5741,
+      "step": 11240
+    },
+    {
+      "epoch": 2.984202840833665,
+      "grad_norm": 0.4399840176483594,
+      "learning_rate": 2.5373552970081866e-06,
+      "loss": 0.5601,
+      "step": 11241
+    },
+    {
+      "epoch": 2.984468339307049,
+      "grad_norm": 0.44741860754699964,
+      "learning_rate": 2.537006207704009e-06,
+      "loss": 0.5517,
+      "step": 11242
+    },
+    {
+      "epoch": 2.984733837780433,
+      "grad_norm": 0.44998287673306,
+      "learning_rate": 2.53665711767812e-06,
+      "loss": 0.5762,
+      "step": 11243
+    },
+    {
+      "epoch": 2.9849993362538165,
+      "grad_norm": 0.43809905060054016,
+      "learning_rate": 2.5363080269373274e-06,
+      "loss": 0.5629,
+      "step": 11244
+    },
+    {
+      "epoch": 2.9852648347272,
+      "grad_norm": 0.4448452922040201,
+      "learning_rate": 2.5359589354884396e-06,
+      "loss": 0.5761,
+      "step": 11245
+    },
+    {
+      "epoch": 2.9855303332005843,
+      "grad_norm": 0.449341935571204,
+      "learning_rate": 2.535609843338265e-06,
+      "loss": 0.5639,
+      "step": 11246
+    },
+    {
+      "epoch": 2.985795831673968,
+      "grad_norm": 0.4289894038335368,
+      "learning_rate": 2.53526075049361e-06,
+      "loss": 0.5384,
+      "step": 11247
+    },
+    {
+      "epoch": 2.9860613301473515,
+      "grad_norm": 0.44458250738500216,
+      "learning_rate": 2.5349116569612846e-06,
+      "loss": 0.5544,
+      "step": 11248
+    },
+    {
+      "epoch": 2.9863268286207356,
+      "grad_norm": 0.4401993868245061,
+      "learning_rate": 2.5345625627480967e-06,
+      "loss": 0.5586,
+      "step": 11249
+    },
+    {
+      "epoch": 2.9865923270941193,
+      "grad_norm": 0.4458296969510183,
+      "learning_rate": 2.5342134678608545e-06,
+      "loss": 0.5639,
+      "step": 11250
+    },
+    {
+      "epoch": 2.986857825567503,
+      "grad_norm": 0.45195500940060224,
+      "learning_rate": 2.533864372306365e-06,
+      "loss": 0.5373,
+      "step": 11251
+    },
+    {
+      "epoch": 2.987123324040887,
+      "grad_norm": 0.4365351646667504,
+      "learning_rate": 2.533515276091437e-06,
+      "loss": 0.5724,
+      "step": 11252
+    },
+    {
+      "epoch": 2.9873888225142706,
+      "grad_norm": 0.43290288455455056,
+      "learning_rate": 2.53316617922288e-06,
+      "loss": 0.5565,
+      "step": 11253
+    },
+    {
+      "epoch": 2.9876543209876543,
+      "grad_norm": 0.42887338332466673,
+      "learning_rate": 2.532817081707501e-06,
+      "loss": 0.5239,
+      "step": 11254
+    },
+    {
+      "epoch": 2.9879198194610384,
+      "grad_norm": 0.4449098984978563,
+      "learning_rate": 2.5324679835521095e-06,
+      "loss": 0.547,
+      "step": 11255
+    },
+    {
+      "epoch": 2.988185317934422,
+      "grad_norm": 0.4634421634240405,
+      "learning_rate": 2.5321188847635114e-06,
+      "loss": 0.5539,
+      "step": 11256
+    },
+    {
+      "epoch": 2.9884508164078056,
+      "grad_norm": 0.4373462119829847,
+      "learning_rate": 2.531769785348517e-06,
+      "loss": 0.5811,
+      "step": 11257
+    },
+    {
+      "epoch": 2.9887163148811893,
+      "grad_norm": 0.44354990633169195,
+      "learning_rate": 2.5314206853139345e-06,
+      "loss": 0.5576,
+      "step": 11258
+    },
+    {
+      "epoch": 2.9889818133545734,
+      "grad_norm": 0.44503574667033263,
+      "learning_rate": 2.5310715846665705e-06,
+      "loss": 0.5663,
+      "step": 11259
+    },
+    {
+      "epoch": 2.989247311827957,
+      "grad_norm": 0.4356749565603212,
+      "learning_rate": 2.5307224834132353e-06,
+      "loss": 0.5679,
+      "step": 11260
+    },
+    {
+      "epoch": 2.9895128103013406,
+      "grad_norm": 0.4430748615799877,
+      "learning_rate": 2.530373381560736e-06,
+      "loss": 0.5387,
+      "step": 11261
+    },
+    {
+      "epoch": 2.9897783087747243,
+      "grad_norm": 0.4496859478461454,
+      "learning_rate": 2.5300242791158817e-06,
+      "loss": 0.5696,
+      "step": 11262
+    },
+    {
+      "epoch": 2.9900438072481084,
+      "grad_norm": 0.4479595208313046,
+      "learning_rate": 2.5296751760854803e-06,
+      "loss": 0.58,
+      "step": 11263
+    },
+    {
+      "epoch": 2.990309305721492,
+      "grad_norm": 0.4347060236430883,
+      "learning_rate": 2.5293260724763415e-06,
+      "loss": 0.5432,
+      "step": 11264
+    },
+    {
+      "epoch": 2.9905748041948756,
+      "grad_norm": 0.4587011621086973,
+      "learning_rate": 2.5289769682952708e-06,
+      "loss": 0.5768,
+      "step": 11265
+    },
+    {
+      "epoch": 2.9908403026682597,
+      "grad_norm": 0.45993271716877443,
+      "learning_rate": 2.5286278635490796e-06,
+      "loss": 0.5617,
+      "step": 11266
+    },
+    {
+      "epoch": 2.9911058011416434,
+      "grad_norm": 0.4593435656454506,
+      "learning_rate": 2.528278758244574e-06,
+      "loss": 0.6153,
+      "step": 11267
+    },
+    {
+      "epoch": 2.991371299615027,
+      "grad_norm": 0.45793928278450274,
+      "learning_rate": 2.5279296523885636e-06,
+      "loss": 0.5955,
+      "step": 11268
+    },
+    {
+      "epoch": 2.991636798088411,
+      "grad_norm": 0.4403072305829855,
+      "learning_rate": 2.527580545987857e-06,
+      "loss": 0.5283,
+      "step": 11269
+    },
+    {
+      "epoch": 2.9919022965617947,
+      "grad_norm": 0.4429184407114556,
+      "learning_rate": 2.5272314390492613e-06,
+      "loss": 0.5672,
+      "step": 11270
+    },
+    {
+      "epoch": 2.9921677950351784,
+      "grad_norm": 0.454936784487101,
+      "learning_rate": 2.5268823315795864e-06,
+      "loss": 0.5599,
+      "step": 11271
+    },
+    {
+      "epoch": 2.9924332935085625,
+      "grad_norm": 0.463014701366042,
+      "learning_rate": 2.526533223585641e-06,
+      "loss": 0.5506,
+      "step": 11272
+    },
+    {
+      "epoch": 2.992698791981946,
+      "grad_norm": 0.4334662432865807,
+      "learning_rate": 2.5261841150742316e-06,
+      "loss": 0.5611,
+      "step": 11273
+    },
+    {
+      "epoch": 2.9929642904553297,
+      "grad_norm": 0.4365591613659824,
+      "learning_rate": 2.5258350060521685e-06,
+      "loss": 0.5455,
+      "step": 11274
+    },
+    {
+      "epoch": 2.993229788928714,
+      "grad_norm": 0.4549911110574605,
+      "learning_rate": 2.5254858965262598e-06,
+      "loss": 0.5421,
+      "step": 11275
+    },
+    {
+      "epoch": 2.9934952874020975,
+      "grad_norm": 0.45847394629016647,
+      "learning_rate": 2.525136786503313e-06,
+      "loss": 0.5363,
+      "step": 11276
+    },
+    {
+      "epoch": 2.993760785875481,
+      "grad_norm": 0.49382089052403316,
+      "learning_rate": 2.5247876759901373e-06,
+      "loss": 0.5778,
+      "step": 11277
+    },
+    {
+      "epoch": 2.994026284348865,
+      "grad_norm": 0.4338624895993122,
+      "learning_rate": 2.524438564993542e-06,
+      "loss": 0.5517,
+      "step": 11278
+    },
+    {
+      "epoch": 2.994291782822249,
+      "grad_norm": 0.4533933217734302,
+      "learning_rate": 2.5240894535203343e-06,
+      "loss": 0.5032,
+      "step": 11279
+    },
+    {
+      "epoch": 2.9945572812956325,
+      "grad_norm": 0.4562039665686122,
+      "learning_rate": 2.5237403415773226e-06,
+      "loss": 0.544,
+      "step": 11280
+    },
+    {
+      "epoch": 2.9948227797690166,
+      "grad_norm": 0.45701338523967777,
+      "learning_rate": 2.5233912291713168e-06,
+      "loss": 0.5622,
+      "step": 11281
+    },
+    {
+      "epoch": 2.9950882782424,
+      "grad_norm": 0.4599025888326689,
+      "learning_rate": 2.5230421163091252e-06,
+      "loss": 0.5566,
+      "step": 11282
+    },
+    {
+      "epoch": 2.995353776715784,
+      "grad_norm": 0.4535592936562848,
+      "learning_rate": 2.5226930029975556e-06,
+      "loss": 0.573,
+      "step": 11283
+    },
+    {
+      "epoch": 2.995619275189168,
+      "grad_norm": 0.4381908355222801,
+      "learning_rate": 2.522343889243417e-06,
+      "loss": 0.5759,
+      "step": 11284
+    },
+    {
+      "epoch": 2.9958847736625516,
+      "grad_norm": 0.42897915627450794,
+      "learning_rate": 2.521994775053517e-06,
+      "loss": 0.5615,
+      "step": 11285
+    },
+    {
+      "epoch": 2.996150272135935,
+      "grad_norm": 0.4520380469089561,
+      "learning_rate": 2.521645660434666e-06,
+      "loss": 0.5787,
+      "step": 11286
+    },
+    {
+      "epoch": 2.996415770609319,
+      "grad_norm": 0.4717392481113668,
+      "learning_rate": 2.5212965453936717e-06,
+      "loss": 0.5812,
+      "step": 11287
+    },
+    {
+      "epoch": 2.996681269082703,
+      "grad_norm": 0.4511349127266952,
+      "learning_rate": 2.520947429937342e-06,
+      "loss": 0.5641,
+      "step": 11288
+    },
+    {
+      "epoch": 2.9969467675560866,
+      "grad_norm": 0.45068160769747145,
+      "learning_rate": 2.520598314072486e-06,
+      "loss": 0.5666,
+      "step": 11289
+    },
+    {
+      "epoch": 2.99721226602947,
+      "grad_norm": 0.45313071251959997,
+      "learning_rate": 2.520249197805913e-06,
+      "loss": 0.5601,
+      "step": 11290
+    },
+    {
+      "epoch": 2.997477764502854,
+      "grad_norm": 0.44555595334324477,
+      "learning_rate": 2.5199000811444305e-06,
+      "loss": 0.5332,
+      "step": 11291
+    },
+    {
+      "epoch": 2.997743262976238,
+      "grad_norm": 0.45432961178884324,
+      "learning_rate": 2.5195509640948486e-06,
+      "loss": 0.5612,
+      "step": 11292
+    },
+    {
+      "epoch": 2.9980087614496216,
+      "grad_norm": 0.4428788659288057,
+      "learning_rate": 2.5192018466639744e-06,
+      "loss": 0.5613,
+      "step": 11293
+    },
+    {
+      "epoch": 2.9982742599230052,
+      "grad_norm": 0.45634298802191126,
+      "learning_rate": 2.518852728858617e-06,
+      "loss": 0.5894,
+      "step": 11294
+    },
+    {
+      "epoch": 2.9985397583963893,
+      "grad_norm": 0.4505354320959237,
+      "learning_rate": 2.518503610685586e-06,
+      "loss": 0.5787,
+      "step": 11295
+    },
+    {
+      "epoch": 2.998805256869773,
+      "grad_norm": 0.4419037294560243,
+      "learning_rate": 2.5181544921516887e-06,
+      "loss": 0.5518,
+      "step": 11296
+    },
+    {
+      "epoch": 2.9990707553431566,
+      "grad_norm": 0.4535601858542848,
+      "learning_rate": 2.517805373263734e-06,
+      "loss": 0.5382,
+      "step": 11297
+    },
+    {
+      "epoch": 2.9993362538165407,
+      "grad_norm": 0.43313963798149757,
+      "learning_rate": 2.5174562540285314e-06,
+      "loss": 0.5277,
+      "step": 11298
+    },
+    {
+      "epoch": 2.9996017522899243,
+      "grad_norm": 0.4428862138715205,
+      "learning_rate": 2.5171071344528893e-06,
+      "loss": 0.5325,
+      "step": 11299
+    },
+    {
+      "epoch": 2.999867250763308,
+      "grad_norm": 0.4361884270649867,
+      "learning_rate": 2.5167580145436153e-06,
+      "loss": 0.5691,
+      "step": 11300
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.4361884270649867,
+      "learning_rate": 2.5164088943075204e-06,
+      "loss": 0.5764,
+      "step": 11301
+    },
+    {
+      "epoch": 3.0002654984733836,
+      "grad_norm": 0.7088511998776494,
+      "learning_rate": 2.516059773751411e-06,
+      "loss": 0.5466,
+      "step": 11302
+    },
+    {
+      "epoch": 3.0005309969467677,
+      "grad_norm": 0.42910669148857644,
+      "learning_rate": 2.5157106528820967e-06,
+      "loss": 0.5125,
+      "step": 11303
+    },
+    {
+      "epoch": 3.0007964954201514,
+      "grad_norm": 0.4374062010008459,
+      "learning_rate": 2.515361531706387e-06,
+      "loss": 0.53,
+      "step": 11304
+    },
+    {
+      "epoch": 3.001061993893535,
+      "grad_norm": 0.4310181577069742,
+      "learning_rate": 2.5150124102310883e-06,
+      "loss": 0.5118,
+      "step": 11305
+    },
+    {
+      "epoch": 3.001327492366919,
+      "grad_norm": 0.44924336997676784,
+      "learning_rate": 2.5146632884630116e-06,
+      "loss": 0.5503,
+      "step": 11306
+    },
+    {
+      "epoch": 3.0015929908403027,
+      "grad_norm": 0.44907038022732887,
+      "learning_rate": 2.514314166408965e-06,
+      "loss": 0.5176,
+      "step": 11307
+    },
+    {
+      "epoch": 3.0018584893136864,
+      "grad_norm": 0.43672237946731995,
+      "learning_rate": 2.5139650440757567e-06,
+      "loss": 0.5216,
+      "step": 11308
+    },
+    {
+      "epoch": 3.00212398778707,
+      "grad_norm": 0.4415531017760964,
+      "learning_rate": 2.5136159214701965e-06,
+      "loss": 0.5302,
+      "step": 11309
+    },
+    {
+      "epoch": 3.002389486260454,
+      "grad_norm": 0.4447553178754593,
+      "learning_rate": 2.5132667985990926e-06,
+      "loss": 0.5441,
+      "step": 11310
+    },
+    {
+      "epoch": 3.0026549847338377,
+      "grad_norm": 0.45334867250251065,
+      "learning_rate": 2.512917675469253e-06,
+      "loss": 0.5392,
+      "step": 11311
+    },
+    {
+      "epoch": 3.0029204832072214,
+      "grad_norm": 0.45426628460225166,
+      "learning_rate": 2.5125685520874876e-06,
+      "loss": 0.5706,
+      "step": 11312
+    },
+    {
+      "epoch": 3.0031859816806055,
+      "grad_norm": 0.4397767439356563,
+      "learning_rate": 2.5122194284606043e-06,
+      "loss": 0.5522,
+      "step": 11313
+    },
+    {
+      "epoch": 3.003451480153989,
+      "grad_norm": 0.44772936916434014,
+      "learning_rate": 2.5118703045954124e-06,
+      "loss": 0.5387,
+      "step": 11314
+    },
+    {
+      "epoch": 3.0037169786273727,
+      "grad_norm": 0.441492028197051,
+      "learning_rate": 2.5115211804987205e-06,
+      "loss": 0.5922,
+      "step": 11315
+    },
+    {
+      "epoch": 3.003982477100757,
+      "grad_norm": 0.4706905287672035,
+      "learning_rate": 2.5111720561773366e-06,
+      "loss": 0.537,
+      "step": 11316
+    },
+    {
+      "epoch": 3.0042479755741405,
+      "grad_norm": 0.45222299309792324,
+      "learning_rate": 2.5108229316380712e-06,
+      "loss": 0.5596,
+      "step": 11317
+    },
+    {
+      "epoch": 3.004513474047524,
+      "grad_norm": 0.45510779905336923,
+      "learning_rate": 2.510473806887732e-06,
+      "loss": 0.5387,
+      "step": 11318
+    },
+    {
+      "epoch": 3.004778972520908,
+      "grad_norm": 0.44521453971224245,
+      "learning_rate": 2.5101246819331283e-06,
+      "loss": 0.522,
+      "step": 11319
+    },
+    {
+      "epoch": 3.005044470994292,
+      "grad_norm": 0.4497136237839376,
+      "learning_rate": 2.5097755567810677e-06,
+      "loss": 0.563,
+      "step": 11320
+    },
+    {
+      "epoch": 3.0053099694676755,
+      "grad_norm": 0.4350823791017069,
+      "learning_rate": 2.5094264314383602e-06,
+      "loss": 0.5461,
+      "step": 11321
+    },
+    {
+      "epoch": 3.005575467941059,
+      "grad_norm": 0.45808616727861845,
+      "learning_rate": 2.509077305911814e-06,
+      "loss": 0.5405,
+      "step": 11322
+    },
+    {
+      "epoch": 3.005840966414443,
+      "grad_norm": 0.4610149900937319,
+      "learning_rate": 2.508728180208238e-06,
+      "loss": 0.5188,
+      "step": 11323
+    },
+    {
+      "epoch": 3.006106464887827,
+      "grad_norm": 0.45516075150157503,
+      "learning_rate": 2.508379054334442e-06,
+      "loss": 0.5825,
+      "step": 11324
+    },
+    {
+      "epoch": 3.0063719633612105,
+      "grad_norm": 0.4459249469923601,
+      "learning_rate": 2.5080299282972333e-06,
+      "loss": 0.499,
+      "step": 11325
+    },
+    {
+      "epoch": 3.0066374618345946,
+      "grad_norm": 0.43985195152481044,
+      "learning_rate": 2.5076808021034204e-06,
+      "loss": 0.5645,
+      "step": 11326
+    },
+    {
+      "epoch": 3.006902960307978,
+      "grad_norm": 0.4498148165442387,
+      "learning_rate": 2.5073316757598147e-06,
+      "loss": 0.5106,
+      "step": 11327
+    },
+    {
+      "epoch": 3.007168458781362,
+      "grad_norm": 0.45112105530185515,
+      "learning_rate": 2.5069825492732223e-06,
+      "loss": 0.5428,
+      "step": 11328
+    },
+    {
+      "epoch": 3.007433957254746,
+      "grad_norm": 0.46333586177429326,
+      "learning_rate": 2.5066334226504537e-06,
+      "loss": 0.5325,
+      "step": 11329
+    },
+    {
+      "epoch": 3.0076994557281296,
+      "grad_norm": 0.43746005869075444,
+      "learning_rate": 2.506284295898317e-06,
+      "loss": 0.5593,
+      "step": 11330
+    },
+    {
+      "epoch": 3.007964954201513,
+      "grad_norm": 0.45077058278880017,
+      "learning_rate": 2.505935169023621e-06,
+      "loss": 0.5502,
+      "step": 11331
+    },
+    {
+      "epoch": 3.0082304526748973,
+      "grad_norm": 0.42898475421673216,
+      "learning_rate": 2.5055860420331746e-06,
+      "loss": 0.5427,
+      "step": 11332
+    },
+    {
+      "epoch": 3.008495951148281,
+      "grad_norm": 0.4527510304832329,
+      "learning_rate": 2.505236914933787e-06,
+      "loss": 0.5502,
+      "step": 11333
+    },
+    {
+      "epoch": 3.0087614496216646,
+      "grad_norm": 0.448571279366419,
+      "learning_rate": 2.5048877877322665e-06,
+      "loss": 0.5629,
+      "step": 11334
+    },
+    {
+      "epoch": 3.0090269480950487,
+      "grad_norm": 0.452246714074237,
+      "learning_rate": 2.5045386604354223e-06,
+      "loss": 0.5528,
+      "step": 11335
+    },
+    {
+      "epoch": 3.0092924465684323,
+      "grad_norm": 0.4530729964734495,
+      "learning_rate": 2.5041895330500627e-06,
+      "loss": 0.559,
+      "step": 11336
+    },
+    {
+      "epoch": 3.009557945041816,
+      "grad_norm": 0.44230561413267777,
+      "learning_rate": 2.5038404055829967e-06,
+      "loss": 0.5599,
+      "step": 11337
+    },
+    {
+      "epoch": 3.0098234435151996,
+      "grad_norm": 0.4420693238205891,
+      "learning_rate": 2.5034912780410347e-06,
+      "loss": 0.5585,
+      "step": 11338
+    },
+    {
+      "epoch": 3.0100889419885837,
+      "grad_norm": 0.45392691094090687,
+      "learning_rate": 2.503142150430983e-06,
+      "loss": 0.5341,
+      "step": 11339
+    },
+    {
+      "epoch": 3.0103544404619673,
+      "grad_norm": 0.4499986656580863,
+      "learning_rate": 2.5027930227596525e-06,
+      "loss": 0.5437,
+      "step": 11340
+    },
+    {
+      "epoch": 3.010619938935351,
+      "grad_norm": 0.4567706752591637,
+      "learning_rate": 2.502443895033851e-06,
+      "loss": 0.5411,
+      "step": 11341
+    },
+    {
+      "epoch": 3.010885437408735,
+      "grad_norm": 0.4302456838102514,
+      "learning_rate": 2.502094767260387e-06,
+      "loss": 0.5395,
+      "step": 11342
+    },
+    {
+      "epoch": 3.0111509358821187,
+      "grad_norm": 0.4436728909899451,
+      "learning_rate": 2.5017456394460706e-06,
+      "loss": 0.5393,
+      "step": 11343
+    },
+    {
+      "epoch": 3.0114164343555023,
+      "grad_norm": 0.44738266338846805,
+      "learning_rate": 2.5013965115977097e-06,
+      "loss": 0.5487,
+      "step": 11344
+    },
+    {
+      "epoch": 3.0116819328288864,
+      "grad_norm": 0.45087354958090115,
+      "learning_rate": 2.501047383722113e-06,
+      "loss": 0.5362,
+      "step": 11345
+    },
+    {
+      "epoch": 3.01194743130227,
+      "grad_norm": 0.46587253926241795,
+      "learning_rate": 2.5006982558260902e-06,
+      "loss": 0.5685,
+      "step": 11346
+    },
+    {
+      "epoch": 3.0122129297756537,
+      "grad_norm": 0.444328133728684,
+      "learning_rate": 2.50034912791645e-06,
+      "loss": 0.5585,
+      "step": 11347
+    },
+    {
+      "epoch": 3.0124784282490378,
+      "grad_norm": 0.44497849621861746,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5421,
+      "step": 11348
+    },
+    {
+      "epoch": 3.0127439267224214,
+      "grad_norm": 0.4577891992918872,
+      "learning_rate": 2.499650872083551e-06,
+      "loss": 0.5855,
+      "step": 11349
+    },
+    {
+      "epoch": 3.013009425195805,
+      "grad_norm": 0.4588169041142913,
+      "learning_rate": 2.49930174417391e-06,
+      "loss": 0.5869,
+      "step": 11350
+    },
+    {
+      "epoch": 3.0132749236691887,
+      "grad_norm": 0.44168509455356286,
+      "learning_rate": 2.4989526162778876e-06,
+      "loss": 0.5496,
+      "step": 11351
+    },
+    {
+      "epoch": 3.013540422142573,
+      "grad_norm": 0.4596326319097752,
+      "learning_rate": 2.4986034884022916e-06,
+      "loss": 0.5276,
+      "step": 11352
+    },
+    {
+      "epoch": 3.0138059206159564,
+      "grad_norm": 0.42844809859535943,
+      "learning_rate": 2.4982543605539302e-06,
+      "loss": 0.5446,
+      "step": 11353
+    },
+    {
+      "epoch": 3.01407141908934,
+      "grad_norm": 0.4422038485215486,
+      "learning_rate": 2.4979052327396133e-06,
+      "loss": 0.5294,
+      "step": 11354
+    },
+    {
+      "epoch": 3.014336917562724,
+      "grad_norm": 0.4564835458031461,
+      "learning_rate": 2.49755610496615e-06,
+      "loss": 0.5505,
+      "step": 11355
+    },
+    {
+      "epoch": 3.014602416036108,
+      "grad_norm": 0.4275397382890219,
+      "learning_rate": 2.497206977240348e-06,
+      "loss": 0.533,
+      "step": 11356
+    },
+    {
+      "epoch": 3.0148679145094914,
+      "grad_norm": 0.462186128208261,
+      "learning_rate": 2.4968578495690177e-06,
+      "loss": 0.5554,
+      "step": 11357
+    },
+    {
+      "epoch": 3.0151334129828755,
+      "grad_norm": 0.4393311992796125,
+      "learning_rate": 2.4965087219589666e-06,
+      "loss": 0.5196,
+      "step": 11358
+    },
+    {
+      "epoch": 3.015398911456259,
+      "grad_norm": 0.44654358100077113,
+      "learning_rate": 2.4961595944170037e-06,
+      "loss": 0.5613,
+      "step": 11359
+    },
+    {
+      "epoch": 3.015664409929643,
+      "grad_norm": 0.46649968442610984,
+      "learning_rate": 2.495810466949938e-06,
+      "loss": 0.554,
+      "step": 11360
+    },
+    {
+      "epoch": 3.015929908403027,
+      "grad_norm": 0.44304371775862483,
+      "learning_rate": 2.495461339564579e-06,
+      "loss": 0.552,
+      "step": 11361
+    },
+    {
+      "epoch": 3.0161954068764105,
+      "grad_norm": 0.4473936442225,
+      "learning_rate": 2.4951122122677343e-06,
+      "loss": 0.5569,
+      "step": 11362
+    },
+    {
+      "epoch": 3.016460905349794,
+      "grad_norm": 0.44715890733901936,
+      "learning_rate": 2.4947630850662143e-06,
+      "loss": 0.5516,
+      "step": 11363
+    },
+    {
+      "epoch": 3.016726403823178,
+      "grad_norm": 0.4618170394987225,
+      "learning_rate": 2.494413957966826e-06,
+      "loss": 0.5698,
+      "step": 11364
+    },
+    {
+      "epoch": 3.016991902296562,
+      "grad_norm": 0.4388459869937891,
+      "learning_rate": 2.49406483097638e-06,
+      "loss": 0.5454,
+      "step": 11365
+    },
+    {
+      "epoch": 3.0172574007699455,
+      "grad_norm": 0.4475060055813505,
+      "learning_rate": 2.493715704101684e-06,
+      "loss": 0.5447,
+      "step": 11366
+    },
+    {
+      "epoch": 3.017522899243329,
+      "grad_norm": 0.43930536026052397,
+      "learning_rate": 2.4933665773495467e-06,
+      "loss": 0.5131,
+      "step": 11367
+    },
+    {
+      "epoch": 3.0177883977167133,
+      "grad_norm": 0.449186539159897,
+      "learning_rate": 2.4930174507267776e-06,
+      "loss": 0.5791,
+      "step": 11368
+    },
+    {
+      "epoch": 3.018053896190097,
+      "grad_norm": 0.4438609294339896,
+      "learning_rate": 2.4926683242401865e-06,
+      "loss": 0.5448,
+      "step": 11369
+    },
+    {
+      "epoch": 3.0183193946634805,
+      "grad_norm": 0.46441727170252234,
+      "learning_rate": 2.49231919789658e-06,
+      "loss": 0.5662,
+      "step": 11370
+    },
+    {
+      "epoch": 3.0185848931368646,
+      "grad_norm": 0.4287865256742984,
+      "learning_rate": 2.4919700717027676e-06,
+      "loss": 0.5133,
+      "step": 11371
+    },
+    {
+      "epoch": 3.0188503916102483,
+      "grad_norm": 0.4392193479756321,
+      "learning_rate": 2.4916209456655583e-06,
+      "loss": 0.547,
+      "step": 11372
+    },
+    {
+      "epoch": 3.019115890083632,
+      "grad_norm": 0.4531874375286059,
+      "learning_rate": 2.4912718197917616e-06,
+      "loss": 0.5501,
+      "step": 11373
+    },
+    {
+      "epoch": 3.019381388557016,
+      "grad_norm": 0.44694930485001355,
+      "learning_rate": 2.490922694088187e-06,
+      "loss": 0.5645,
+      "step": 11374
+    },
+    {
+      "epoch": 3.0196468870303996,
+      "grad_norm": 0.4488028913189199,
+      "learning_rate": 2.490573568561641e-06,
+      "loss": 0.5551,
+      "step": 11375
+    },
+    {
+      "epoch": 3.0199123855037833,
+      "grad_norm": 0.458656658866499,
+      "learning_rate": 2.490224443218933e-06,
+      "loss": 0.5419,
+      "step": 11376
+    },
+    {
+      "epoch": 3.0201778839771674,
+      "grad_norm": 0.4518699617197473,
+      "learning_rate": 2.489875318066873e-06,
+      "loss": 0.5599,
+      "step": 11377
+    },
+    {
+      "epoch": 3.020443382450551,
+      "grad_norm": 0.4424438856330224,
+      "learning_rate": 2.4895261931122684e-06,
+      "loss": 0.5507,
+      "step": 11378
+    },
+    {
+      "epoch": 3.0207088809239346,
+      "grad_norm": 0.4578868787621122,
+      "learning_rate": 2.489177068361929e-06,
+      "loss": 0.5747,
+      "step": 11379
+    },
+    {
+      "epoch": 3.0209743793973183,
+      "grad_norm": 0.4556148976423923,
+      "learning_rate": 2.488827943822664e-06,
+      "loss": 0.5694,
+      "step": 11380
+    },
+    {
+      "epoch": 3.0212398778707024,
+      "grad_norm": 0.4359781928000787,
+      "learning_rate": 2.4884788195012804e-06,
+      "loss": 0.5442,
+      "step": 11381
+    },
+    {
+      "epoch": 3.021505376344086,
+      "grad_norm": 0.48363372886598793,
+      "learning_rate": 2.4881296954045884e-06,
+      "loss": 0.5346,
+      "step": 11382
+    },
+    {
+      "epoch": 3.0217708748174696,
+      "grad_norm": 0.4604823458742792,
+      "learning_rate": 2.487780571539396e-06,
+      "loss": 0.5618,
+      "step": 11383
+    },
+    {
+      "epoch": 3.0220363732908537,
+      "grad_norm": 0.44826304092119007,
+      "learning_rate": 2.4874314479125128e-06,
+      "loss": 0.5477,
+      "step": 11384
+    },
+    {
+      "epoch": 3.0223018717642374,
+      "grad_norm": 0.46236290842185895,
+      "learning_rate": 2.487082324530747e-06,
+      "loss": 0.5435,
+      "step": 11385
+    },
+    {
+      "epoch": 3.022567370237621,
+      "grad_norm": 0.45563110594913037,
+      "learning_rate": 2.4867332014009086e-06,
+      "loss": 0.5423,
+      "step": 11386
+    },
+    {
+      "epoch": 3.022832868711005,
+      "grad_norm": 0.46984263522281516,
+      "learning_rate": 2.4863840785298044e-06,
+      "loss": 0.5657,
+      "step": 11387
+    },
+    {
+      "epoch": 3.0230983671843887,
+      "grad_norm": 0.4522311850864079,
+      "learning_rate": 2.4860349559242437e-06,
+      "loss": 0.531,
+      "step": 11388
+    },
+    {
+      "epoch": 3.0233638656577724,
+      "grad_norm": 0.4538910231581137,
+      "learning_rate": 2.4856858335910354e-06,
+      "loss": 0.5383,
+      "step": 11389
+    },
+    {
+      "epoch": 3.0236293641311565,
+      "grad_norm": 0.4577613548755966,
+      "learning_rate": 2.485336711536989e-06,
+      "loss": 0.5289,
+      "step": 11390
+    },
+    {
+      "epoch": 3.02389486260454,
+      "grad_norm": 0.45878909876504076,
+      "learning_rate": 2.484987589768912e-06,
+      "loss": 0.5734,
+      "step": 11391
+    },
+    {
+      "epoch": 3.0241603610779237,
+      "grad_norm": 0.45236812046271174,
+      "learning_rate": 2.4846384682936148e-06,
+      "loss": 0.5695,
+      "step": 11392
+    },
+    {
+      "epoch": 3.0244258595513074,
+      "grad_norm": 0.4629286805437994,
+      "learning_rate": 2.484289347117904e-06,
+      "loss": 0.544,
+      "step": 11393
+    },
+    {
+      "epoch": 3.0246913580246915,
+      "grad_norm": 0.44551195360080476,
+      "learning_rate": 2.48394022624859e-06,
+      "loss": 0.5525,
+      "step": 11394
+    },
+    {
+      "epoch": 3.024956856498075,
+      "grad_norm": 0.44159570643075696,
+      "learning_rate": 2.4835911056924804e-06,
+      "loss": 0.5629,
+      "step": 11395
+    },
+    {
+      "epoch": 3.0252223549714587,
+      "grad_norm": 0.46099124349039006,
+      "learning_rate": 2.4832419854563847e-06,
+      "loss": 0.5338,
+      "step": 11396
+    },
+    {
+      "epoch": 3.025487853444843,
+      "grad_norm": 0.4608636196539273,
+      "learning_rate": 2.482892865547112e-06,
+      "loss": 0.5572,
+      "step": 11397
+    },
+    {
+      "epoch": 3.0257533519182265,
+      "grad_norm": 0.4293688698602899,
+      "learning_rate": 2.4825437459714694e-06,
+      "loss": 0.5089,
+      "step": 11398
+    },
+    {
+      "epoch": 3.02601885039161,
+      "grad_norm": 0.4740745299547139,
+      "learning_rate": 2.4821946267362664e-06,
+      "loss": 0.562,
+      "step": 11399
+    },
+    {
+      "epoch": 3.026284348864994,
+      "grad_norm": 0.4548374340272445,
+      "learning_rate": 2.481845507848312e-06,
+      "loss": 0.576,
+      "step": 11400
+    },
+    {
+      "epoch": 3.026549847338378,
+      "grad_norm": 0.460284036389579,
+      "learning_rate": 2.4814963893144146e-06,
+      "loss": 0.5576,
+      "step": 11401
+    },
+    {
+      "epoch": 3.0268153458117615,
+      "grad_norm": 0.46264279862012797,
+      "learning_rate": 2.4811472711413833e-06,
+      "loss": 0.5388,
+      "step": 11402
+    },
+    {
+      "epoch": 3.0270808442851456,
+      "grad_norm": 0.4721359342583942,
+      "learning_rate": 2.480798153336027e-06,
+      "loss": 0.5791,
+      "step": 11403
+    },
+    {
+      "epoch": 3.027346342758529,
+      "grad_norm": 0.4601367627973896,
+      "learning_rate": 2.4804490359051526e-06,
+      "loss": 0.5458,
+      "step": 11404
+    },
+    {
+      "epoch": 3.027611841231913,
+      "grad_norm": 0.44459101852355476,
+      "learning_rate": 2.48009991885557e-06,
+      "loss": 0.5747,
+      "step": 11405
+    },
+    {
+      "epoch": 3.0278773397052965,
+      "grad_norm": 0.45568392275733954,
+      "learning_rate": 2.479750802194088e-06,
+      "loss": 0.5931,
+      "step": 11406
+    },
+    {
+      "epoch": 3.0281428381786806,
+      "grad_norm": 0.4615132918076059,
+      "learning_rate": 2.479401685927514e-06,
+      "loss": 0.5743,
+      "step": 11407
+    },
+    {
+      "epoch": 3.028408336652064,
+      "grad_norm": 0.45084536070410025,
+      "learning_rate": 2.4790525700626586e-06,
+      "loss": 0.558,
+      "step": 11408
+    },
+    {
+      "epoch": 3.028673835125448,
+      "grad_norm": 0.45907923066713635,
+      "learning_rate": 2.4787034546063296e-06,
+      "loss": 0.5495,
+      "step": 11409
+    },
+    {
+      "epoch": 3.028939333598832,
+      "grad_norm": 0.4520536155943597,
+      "learning_rate": 2.4783543395653347e-06,
+      "loss": 0.5714,
+      "step": 11410
+    },
+    {
+      "epoch": 3.0292048320722156,
+      "grad_norm": 0.4461748518055214,
+      "learning_rate": 2.4780052249464832e-06,
+      "loss": 0.5776,
+      "step": 11411
+    },
+    {
+      "epoch": 3.029470330545599,
+      "grad_norm": 0.45685862231758895,
+      "learning_rate": 2.477656110756584e-06,
+      "loss": 0.5339,
+      "step": 11412
+    },
+    {
+      "epoch": 3.0297358290189833,
+      "grad_norm": 0.45441640011848605,
+      "learning_rate": 2.477306997002445e-06,
+      "loss": 0.5632,
+      "step": 11413
+    },
+    {
+      "epoch": 3.030001327492367,
+      "grad_norm": 0.4658014663903666,
+      "learning_rate": 2.476957883690875e-06,
+      "loss": 0.4945,
+      "step": 11414
+    },
+    {
+      "epoch": 3.0302668259657506,
+      "grad_norm": 0.4654265790050967,
+      "learning_rate": 2.476608770828684e-06,
+      "loss": 0.5104,
+      "step": 11415
+    },
+    {
+      "epoch": 3.0305323244391347,
+      "grad_norm": 0.4422412980407415,
+      "learning_rate": 2.4762596584226782e-06,
+      "loss": 0.4911,
+      "step": 11416
+    },
+    {
+      "epoch": 3.0307978229125183,
+      "grad_norm": 0.45044438081551114,
+      "learning_rate": 2.4759105464796665e-06,
+      "loss": 0.5421,
+      "step": 11417
+    },
+    {
+      "epoch": 3.031063321385902,
+      "grad_norm": 0.4524247600756253,
+      "learning_rate": 2.4755614350064587e-06,
+      "loss": 0.5797,
+      "step": 11418
+    },
+    {
+      "epoch": 3.0313288198592856,
+      "grad_norm": 0.46186139446113184,
+      "learning_rate": 2.4752123240098627e-06,
+      "loss": 0.5822,
+      "step": 11419
+    },
+    {
+      "epoch": 3.0315943183326697,
+      "grad_norm": 0.4824886988658856,
+      "learning_rate": 2.474863213496687e-06,
+      "loss": 0.5581,
+      "step": 11420
+    },
+    {
+      "epoch": 3.0318598168060533,
+      "grad_norm": 0.4709144087088842,
+      "learning_rate": 2.4745141034737415e-06,
+      "loss": 0.558,
+      "step": 11421
+    },
+    {
+      "epoch": 3.032125315279437,
+      "grad_norm": 0.46370638437820283,
+      "learning_rate": 2.4741649939478323e-06,
+      "loss": 0.5339,
+      "step": 11422
+    },
+    {
+      "epoch": 3.032390813752821,
+      "grad_norm": 0.4420209722156511,
+      "learning_rate": 2.473815884925769e-06,
+      "loss": 0.5648,
+      "step": 11423
+    },
+    {
+      "epoch": 3.0326563122262047,
+      "grad_norm": 0.453175139656296,
+      "learning_rate": 2.47346677641436e-06,
+      "loss": 0.5447,
+      "step": 11424
+    },
+    {
+      "epoch": 3.0329218106995883,
+      "grad_norm": 0.4503144579551726,
+      "learning_rate": 2.4731176684204136e-06,
+      "loss": 0.5451,
+      "step": 11425
+    },
+    {
+      "epoch": 3.0331873091729724,
+      "grad_norm": 0.4816394255639505,
+      "learning_rate": 2.472768560950739e-06,
+      "loss": 0.5749,
+      "step": 11426
+    },
+    {
+      "epoch": 3.033452807646356,
+      "grad_norm": 0.46793345867577285,
+      "learning_rate": 2.4724194540121442e-06,
+      "loss": 0.556,
+      "step": 11427
+    },
+    {
+      "epoch": 3.0337183061197397,
+      "grad_norm": 0.45962176416738665,
+      "learning_rate": 2.472070347611437e-06,
+      "loss": 0.5622,
+      "step": 11428
+    },
+    {
+      "epoch": 3.033983804593124,
+      "grad_norm": 0.4421272493962858,
+      "learning_rate": 2.4717212417554266e-06,
+      "loss": 0.5373,
+      "step": 11429
+    },
+    {
+      "epoch": 3.0342493030665074,
+      "grad_norm": 0.440114721170632,
+      "learning_rate": 2.471372136450921e-06,
+      "loss": 0.5419,
+      "step": 11430
+    },
+    {
+      "epoch": 3.034514801539891,
+      "grad_norm": 0.4658157923920979,
+      "learning_rate": 2.471023031704729e-06,
+      "loss": 0.5345,
+      "step": 11431
+    },
+    {
+      "epoch": 3.034780300013275,
+      "grad_norm": 0.4673664289382252,
+      "learning_rate": 2.47067392752366e-06,
+      "loss": 0.5575,
+      "step": 11432
+    },
+    {
+      "epoch": 3.035045798486659,
+      "grad_norm": 0.44119260005810235,
+      "learning_rate": 2.47032482391452e-06,
+      "loss": 0.567,
+      "step": 11433
+    },
+    {
+      "epoch": 3.0353112969600424,
+      "grad_norm": 0.4598118725679513,
+      "learning_rate": 2.4699757208841187e-06,
+      "loss": 0.5212,
+      "step": 11434
+    },
+    {
+      "epoch": 3.035576795433426,
+      "grad_norm": 0.4402465974638158,
+      "learning_rate": 2.4696266184392643e-06,
+      "loss": 0.5263,
+      "step": 11435
+    },
+    {
+      "epoch": 3.03584229390681,
+      "grad_norm": 0.46282977718900964,
+      "learning_rate": 2.469277516586765e-06,
+      "loss": 0.5728,
+      "step": 11436
+    },
+    {
+      "epoch": 3.036107792380194,
+      "grad_norm": 0.4268958013636579,
+      "learning_rate": 2.46892841533343e-06,
+      "loss": 0.5296,
+      "step": 11437
+    },
+    {
+      "epoch": 3.0363732908535774,
+      "grad_norm": 0.4512747007998144,
+      "learning_rate": 2.468579314686067e-06,
+      "loss": 0.5727,
+      "step": 11438
+    },
+    {
+      "epoch": 3.0366387893269615,
+      "grad_norm": 0.4526868229880013,
+      "learning_rate": 2.468230214651484e-06,
+      "loss": 0.5363,
+      "step": 11439
+    },
+    {
+      "epoch": 3.036904287800345,
+      "grad_norm": 0.4411217271053345,
+      "learning_rate": 2.4678811152364894e-06,
+      "loss": 0.5874,
+      "step": 11440
+    },
+    {
+      "epoch": 3.037169786273729,
+      "grad_norm": 0.4661813424573556,
+      "learning_rate": 2.4675320164478917e-06,
+      "loss": 0.5507,
+      "step": 11441
+    },
+    {
+      "epoch": 3.037435284747113,
+      "grad_norm": 0.4660355791348338,
+      "learning_rate": 2.467182918292499e-06,
+      "loss": 0.5592,
+      "step": 11442
+    },
+    {
+      "epoch": 3.0377007832204965,
+      "grad_norm": 0.452467469998572,
+      "learning_rate": 2.4668338207771202e-06,
+      "loss": 0.5491,
+      "step": 11443
+    },
+    {
+      "epoch": 3.03796628169388,
+      "grad_norm": 0.4456360523507704,
+      "learning_rate": 2.4664847239085633e-06,
+      "loss": 0.5728,
+      "step": 11444
+    },
+    {
+      "epoch": 3.0382317801672643,
+      "grad_norm": 0.4750361505643859,
+      "learning_rate": 2.466135627693636e-06,
+      "loss": 0.5541,
+      "step": 11445
+    },
+    {
+      "epoch": 3.038497278640648,
+      "grad_norm": 0.44628077167141544,
+      "learning_rate": 2.4657865321391468e-06,
+      "loss": 0.5455,
+      "step": 11446
+    },
+    {
+      "epoch": 3.0387627771140315,
+      "grad_norm": 0.4479812647464731,
+      "learning_rate": 2.4654374372519037e-06,
+      "loss": 0.5181,
+      "step": 11447
+    },
+    {
+      "epoch": 3.039028275587415,
+      "grad_norm": 0.44891300276050117,
+      "learning_rate": 2.4650883430387154e-06,
+      "loss": 0.5574,
+      "step": 11448
+    },
+    {
+      "epoch": 3.0392937740607993,
+      "grad_norm": 0.4682428612653018,
+      "learning_rate": 2.4647392495063913e-06,
+      "loss": 0.5194,
+      "step": 11449
+    },
+    {
+      "epoch": 3.039559272534183,
+      "grad_norm": 0.4589105144797848,
+      "learning_rate": 2.464390156661737e-06,
+      "loss": 0.4998,
+      "step": 11450
+    },
+    {
+      "epoch": 3.0398247710075665,
+      "grad_norm": 0.431689246109316,
+      "learning_rate": 2.4640410645115612e-06,
+      "loss": 0.5203,
+      "step": 11451
+    },
+    {
+      "epoch": 3.0400902694809506,
+      "grad_norm": 0.42166902092057346,
+      "learning_rate": 2.463691973062673e-06,
+      "loss": 0.4825,
+      "step": 11452
+    },
+    {
+      "epoch": 3.0403557679543343,
+      "grad_norm": 0.44382833823487516,
+      "learning_rate": 2.4633428823218807e-06,
+      "loss": 0.5229,
+      "step": 11453
+    },
+    {
+      "epoch": 3.040621266427718,
+      "grad_norm": 0.44025221838001893,
+      "learning_rate": 2.4629937922959913e-06,
+      "loss": 0.5533,
+      "step": 11454
+    },
+    {
+      "epoch": 3.040886764901102,
+      "grad_norm": 0.45506126405588343,
+      "learning_rate": 2.4626447029918147e-06,
+      "loss": 0.5439,
+      "step": 11455
+    },
+    {
+      "epoch": 3.0411522633744856,
+      "grad_norm": 0.46269440553358676,
+      "learning_rate": 2.462295614416157e-06,
+      "loss": 0.5532,
+      "step": 11456
+    },
+    {
+      "epoch": 3.0414177618478693,
+      "grad_norm": 0.4488400571491357,
+      "learning_rate": 2.461946526575827e-06,
+      "loss": 0.5529,
+      "step": 11457
+    },
+    {
+      "epoch": 3.0416832603212534,
+      "grad_norm": 0.45583398523540014,
+      "learning_rate": 2.461597439477633e-06,
+      "loss": 0.5316,
+      "step": 11458
+    },
+    {
+      "epoch": 3.041948758794637,
+      "grad_norm": 0.4692884638864521,
+      "learning_rate": 2.4612483531283833e-06,
+      "loss": 0.5671,
+      "step": 11459
+    },
+    {
+      "epoch": 3.0422142572680206,
+      "grad_norm": 0.4598455806394097,
+      "learning_rate": 2.460899267534885e-06,
+      "loss": 0.5229,
+      "step": 11460
+    },
+    {
+      "epoch": 3.0424797557414043,
+      "grad_norm": 0.4595506967018965,
+      "learning_rate": 2.460550182703948e-06,
+      "loss": 0.5593,
+      "step": 11461
+    },
+    {
+      "epoch": 3.0427452542147884,
+      "grad_norm": 0.45046398542995214,
+      "learning_rate": 2.460201098642378e-06,
+      "loss": 0.5641,
+      "step": 11462
+    },
+    {
+      "epoch": 3.043010752688172,
+      "grad_norm": 0.44732556270990476,
+      "learning_rate": 2.4598520153569838e-06,
+      "loss": 0.5609,
+      "step": 11463
+    },
+    {
+      "epoch": 3.0432762511615556,
+      "grad_norm": 0.4545783203477136,
+      "learning_rate": 2.459502932854574e-06,
+      "loss": 0.564,
+      "step": 11464
+    },
+    {
+      "epoch": 3.0435417496349397,
+      "grad_norm": 0.43955721065108855,
+      "learning_rate": 2.4591538511419556e-06,
+      "loss": 0.526,
+      "step": 11465
+    },
+    {
+      "epoch": 3.0438072481083234,
+      "grad_norm": 0.44704662769064885,
+      "learning_rate": 2.458804770225938e-06,
+      "loss": 0.5746,
+      "step": 11466
+    },
+    {
+      "epoch": 3.044072746581707,
+      "grad_norm": 0.4574944374101059,
+      "learning_rate": 2.4584556901133286e-06,
+      "loss": 0.5761,
+      "step": 11467
+    },
+    {
+      "epoch": 3.044338245055091,
+      "grad_norm": 0.4552512514485266,
+      "learning_rate": 2.458106610810934e-06,
+      "loss": 0.5315,
+      "step": 11468
+    },
+    {
+      "epoch": 3.0446037435284747,
+      "grad_norm": 0.45002115574984586,
+      "learning_rate": 2.4577575323255636e-06,
+      "loss": 0.5346,
+      "step": 11469
+    },
+    {
+      "epoch": 3.0448692420018584,
+      "grad_norm": 0.4635156150751205,
+      "learning_rate": 2.457408454664024e-06,
+      "loss": 0.5742,
+      "step": 11470
+    },
+    {
+      "epoch": 3.0451347404752425,
+      "grad_norm": 0.44604689242087014,
+      "learning_rate": 2.4570593778331243e-06,
+      "loss": 0.5518,
+      "step": 11471
+    },
+    {
+      "epoch": 3.045400238948626,
+      "grad_norm": 0.465332592678385,
+      "learning_rate": 2.4567103018396724e-06,
+      "loss": 0.5575,
+      "step": 11472
+    },
+    {
+      "epoch": 3.0456657374220097,
+      "grad_norm": 0.4409123989373969,
+      "learning_rate": 2.456361226690475e-06,
+      "loss": 0.5702,
+      "step": 11473
+    },
+    {
+      "epoch": 3.0459312358953934,
+      "grad_norm": 0.4528984396863483,
+      "learning_rate": 2.45601215239234e-06,
+      "loss": 0.5345,
+      "step": 11474
+    },
+    {
+      "epoch": 3.0461967343687775,
+      "grad_norm": 0.45021953590273034,
+      "learning_rate": 2.4556630789520766e-06,
+      "loss": 0.5802,
+      "step": 11475
+    },
+    {
+      "epoch": 3.046462232842161,
+      "grad_norm": 0.4456075104486512,
+      "learning_rate": 2.455314006376491e-06,
+      "loss": 0.5115,
+      "step": 11476
+    },
+    {
+      "epoch": 3.0467277313155448,
+      "grad_norm": 0.4557474169109319,
+      "learning_rate": 2.4549649346723924e-06,
+      "loss": 0.5363,
+      "step": 11477
+    },
+    {
+      "epoch": 3.046993229788929,
+      "grad_norm": 0.4598259352250728,
+      "learning_rate": 2.4546158638465885e-06,
+      "loss": 0.5433,
+      "step": 11478
+    },
+    {
+      "epoch": 3.0472587282623125,
+      "grad_norm": 0.4538640448948252,
+      "learning_rate": 2.4542667939058853e-06,
+      "loss": 0.5873,
+      "step": 11479
+    },
+    {
+      "epoch": 3.047524226735696,
+      "grad_norm": 0.4478177261734345,
+      "learning_rate": 2.4539177248570913e-06,
+      "loss": 0.5454,
+      "step": 11480
+    },
+    {
+      "epoch": 3.04778972520908,
+      "grad_norm": 0.44676755689722647,
+      "learning_rate": 2.453568656707015e-06,
+      "loss": 0.5212,
+      "step": 11481
+    },
+    {
+      "epoch": 3.048055223682464,
+      "grad_norm": 0.4482395685456643,
+      "learning_rate": 2.4532195894624634e-06,
+      "loss": 0.5576,
+      "step": 11482
+    },
+    {
+      "epoch": 3.0483207221558475,
+      "grad_norm": 0.4466585304793646,
+      "learning_rate": 2.4528705231302438e-06,
+      "loss": 0.5433,
+      "step": 11483
+    },
+    {
+      "epoch": 3.0485862206292316,
+      "grad_norm": 0.43974039635877255,
+      "learning_rate": 2.452521457717166e-06,
+      "loss": 0.5085,
+      "step": 11484
+    },
+    {
+      "epoch": 3.048851719102615,
+      "grad_norm": 0.4368953045821346,
+      "learning_rate": 2.452172393230035e-06,
+      "loss": 0.5348,
+      "step": 11485
+    },
+    {
+      "epoch": 3.049117217575999,
+      "grad_norm": 0.4488894099768197,
+      "learning_rate": 2.4518233296756593e-06,
+      "loss": 0.5718,
+      "step": 11486
+    },
+    {
+      "epoch": 3.049382716049383,
+      "grad_norm": 0.4466445759825953,
+      "learning_rate": 2.4514742670608467e-06,
+      "loss": 0.5563,
+      "step": 11487
+    },
+    {
+      "epoch": 3.0496482145227666,
+      "grad_norm": 0.4453188661397635,
+      "learning_rate": 2.451125205392405e-06,
+      "loss": 0.5522,
+      "step": 11488
+    },
+    {
+      "epoch": 3.04991371299615,
+      "grad_norm": 0.45543942893167844,
+      "learning_rate": 2.4507761446771408e-06,
+      "loss": 0.5552,
+      "step": 11489
+    },
+    {
+      "epoch": 3.050179211469534,
+      "grad_norm": 0.45348011277465744,
+      "learning_rate": 2.4504270849218635e-06,
+      "loss": 0.5706,
+      "step": 11490
+    },
+    {
+      "epoch": 3.050444709942918,
+      "grad_norm": 0.4515964587208077,
+      "learning_rate": 2.4500780261333782e-06,
+      "loss": 0.5252,
+      "step": 11491
+    },
+    {
+      "epoch": 3.0507102084163016,
+      "grad_norm": 0.4469369034678833,
+      "learning_rate": 2.449728968318494e-06,
+      "loss": 0.575,
+      "step": 11492
+    },
+    {
+      "epoch": 3.0509757068896852,
+      "grad_norm": 0.45225586994229167,
+      "learning_rate": 2.4493799114840182e-06,
+      "loss": 0.5579,
+      "step": 11493
+    },
+    {
+      "epoch": 3.0512412053630693,
+      "grad_norm": 0.446453571558936,
+      "learning_rate": 2.449030855636758e-06,
+      "loss": 0.5491,
+      "step": 11494
+    },
+    {
+      "epoch": 3.051506703836453,
+      "grad_norm": 0.4444792187149838,
+      "learning_rate": 2.4486818007835216e-06,
+      "loss": 0.5004,
+      "step": 11495
+    },
+    {
+      "epoch": 3.0517722023098366,
+      "grad_norm": 0.45411176555128274,
+      "learning_rate": 2.448332746931115e-06,
+      "loss": 0.5378,
+      "step": 11496
+    },
+    {
+      "epoch": 3.0520377007832207,
+      "grad_norm": 0.44553150626962895,
+      "learning_rate": 2.447983694086346e-06,
+      "loss": 0.5182,
+      "step": 11497
+    },
+    {
+      "epoch": 3.0523031992566043,
+      "grad_norm": 0.46040204880343016,
+      "learning_rate": 2.447634642256023e-06,
+      "loss": 0.5525,
+      "step": 11498
+    },
+    {
+      "epoch": 3.052568697729988,
+      "grad_norm": 0.4352916696340324,
+      "learning_rate": 2.447285591446952e-06,
+      "loss": 0.5566,
+      "step": 11499
+    },
+    {
+      "epoch": 3.052834196203372,
+      "grad_norm": 0.44962685288740484,
+      "learning_rate": 2.4469365416659412e-06,
+      "loss": 0.5099,
+      "step": 11500
+    },
+    {
+      "epoch": 3.0530996946767557,
+      "grad_norm": 0.4467704702648933,
+      "learning_rate": 2.4465874929197985e-06,
+      "loss": 0.5629,
+      "step": 11501
+    },
+    {
+      "epoch": 3.0533651931501393,
+      "grad_norm": 0.4671102020198274,
+      "learning_rate": 2.4462384452153297e-06,
+      "loss": 0.5634,
+      "step": 11502
+    },
+    {
+      "epoch": 3.053630691623523,
+      "grad_norm": 0.4478606016998195,
+      "learning_rate": 2.4458893985593433e-06,
+      "loss": 0.5387,
+      "step": 11503
+    },
+    {
+      "epoch": 3.053896190096907,
+      "grad_norm": 0.44633057224235834,
+      "learning_rate": 2.4455403529586456e-06,
+      "loss": 0.5572,
+      "step": 11504
+    },
+    {
+      "epoch": 3.0541616885702907,
+      "grad_norm": 0.46042069445424316,
+      "learning_rate": 2.4451913084200446e-06,
+      "loss": 0.5697,
+      "step": 11505
+    },
+    {
+      "epoch": 3.0544271870436743,
+      "grad_norm": 0.4620157840088359,
+      "learning_rate": 2.444842264950347e-06,
+      "loss": 0.5137,
+      "step": 11506
+    },
+    {
+      "epoch": 3.0546926855170584,
+      "grad_norm": 0.4537792633119077,
+      "learning_rate": 2.444493222556362e-06,
+      "loss": 0.5719,
+      "step": 11507
+    },
+    {
+      "epoch": 3.054958183990442,
+      "grad_norm": 0.45516943355134776,
+      "learning_rate": 2.444144181244893e-06,
+      "loss": 0.4969,
+      "step": 11508
+    },
+    {
+      "epoch": 3.0552236824638257,
+      "grad_norm": 0.4408842774101969,
+      "learning_rate": 2.4437951410227495e-06,
+      "loss": 0.5579,
+      "step": 11509
+    },
+    {
+      "epoch": 3.05548918093721,
+      "grad_norm": 0.4363983321233346,
+      "learning_rate": 2.4434461018967386e-06,
+      "loss": 0.491,
+      "step": 11510
+    },
+    {
+      "epoch": 3.0557546794105934,
+      "grad_norm": 0.43490518177426785,
+      "learning_rate": 2.4430970638736675e-06,
+      "loss": 0.5576,
+      "step": 11511
+    },
+    {
+      "epoch": 3.056020177883977,
+      "grad_norm": 0.4473062649657584,
+      "learning_rate": 2.4427480269603425e-06,
+      "loss": 0.5505,
+      "step": 11512
+    },
+    {
+      "epoch": 3.056285676357361,
+      "grad_norm": 0.4390747360393103,
+      "learning_rate": 2.4423989911635726e-06,
+      "loss": 0.5124,
+      "step": 11513
+    },
+    {
+      "epoch": 3.056551174830745,
+      "grad_norm": 0.44516121163425104,
+      "learning_rate": 2.4420499564901623e-06,
+      "loss": 0.5289,
+      "step": 11514
+    },
+    {
+      "epoch": 3.0568166733041284,
+      "grad_norm": 0.4552494495428004,
+      "learning_rate": 2.4417009229469193e-06,
+      "loss": 0.5275,
+      "step": 11515
+    },
+    {
+      "epoch": 3.057082171777512,
+      "grad_norm": 0.4406367144434546,
+      "learning_rate": 2.4413518905406516e-06,
+      "loss": 0.5493,
+      "step": 11516
+    },
+    {
+      "epoch": 3.057347670250896,
+      "grad_norm": 0.4515884422774925,
+      "learning_rate": 2.4410028592781655e-06,
+      "loss": 0.5156,
+      "step": 11517
+    },
+    {
+      "epoch": 3.05761316872428,
+      "grad_norm": 0.4518130833544343,
+      "learning_rate": 2.4406538291662682e-06,
+      "loss": 0.542,
+      "step": 11518
+    },
+    {
+      "epoch": 3.0578786671976634,
+      "grad_norm": 0.4440896476881966,
+      "learning_rate": 2.4403048002117665e-06,
+      "loss": 0.5021,
+      "step": 11519
+    },
+    {
+      "epoch": 3.0581441656710475,
+      "grad_norm": 0.4464073705498716,
+      "learning_rate": 2.439955772421467e-06,
+      "loss": 0.5796,
+      "step": 11520
+    },
+    {
+      "epoch": 3.058409664144431,
+      "grad_norm": 0.4624404475540543,
+      "learning_rate": 2.439606745802177e-06,
+      "loss": 0.5248,
+      "step": 11521
+    },
+    {
+      "epoch": 3.058675162617815,
+      "grad_norm": 0.4380193380351636,
+      "learning_rate": 2.439257720360704e-06,
+      "loss": 0.553,
+      "step": 11522
+    },
+    {
+      "epoch": 3.058940661091199,
+      "grad_norm": 0.4601663339204156,
+      "learning_rate": 2.4389086961038533e-06,
+      "loss": 0.5369,
+      "step": 11523
+    },
+    {
+      "epoch": 3.0592061595645825,
+      "grad_norm": 0.44994543785756863,
+      "learning_rate": 2.438559673038434e-06,
+      "loss": 0.5589,
+      "step": 11524
+    },
+    {
+      "epoch": 3.059471658037966,
+      "grad_norm": 0.46132521521794523,
+      "learning_rate": 2.43821065117125e-06,
+      "loss": 0.5522,
+      "step": 11525
+    },
+    {
+      "epoch": 3.0597371565113503,
+      "grad_norm": 0.49299844006221116,
+      "learning_rate": 2.4378616305091103e-06,
+      "loss": 0.5427,
+      "step": 11526
+    },
+    {
+      "epoch": 3.060002654984734,
+      "grad_norm": 0.45914739592182297,
+      "learning_rate": 2.4375126110588206e-06,
+      "loss": 0.5724,
+      "step": 11527
+    },
+    {
+      "epoch": 3.0602681534581175,
+      "grad_norm": 0.43819681858109005,
+      "learning_rate": 2.437163592827188e-06,
+      "loss": 0.5789,
+      "step": 11528
+    },
+    {
+      "epoch": 3.060533651931501,
+      "grad_norm": 0.4503792571486426,
+      "learning_rate": 2.4368145758210186e-06,
+      "loss": 0.5396,
+      "step": 11529
+    },
+    {
+      "epoch": 3.0607991504048853,
+      "grad_norm": 0.4594128537492662,
+      "learning_rate": 2.436465560047122e-06,
+      "loss": 0.544,
+      "step": 11530
+    },
+    {
+      "epoch": 3.061064648878269,
+      "grad_norm": 0.43268266209467476,
+      "learning_rate": 2.436116545512301e-06,
+      "loss": 0.5376,
+      "step": 11531
+    },
+    {
+      "epoch": 3.0613301473516525,
+      "grad_norm": 0.46559244937416505,
+      "learning_rate": 2.435767532223363e-06,
+      "loss": 0.5708,
+      "step": 11532
+    },
+    {
+      "epoch": 3.0615956458250366,
+      "grad_norm": 0.45138391853619286,
+      "learning_rate": 2.435418520187116e-06,
+      "loss": 0.5375,
+      "step": 11533
+    },
+    {
+      "epoch": 3.0618611442984203,
+      "grad_norm": 0.44885863464374204,
+      "learning_rate": 2.435069509410366e-06,
+      "loss": 0.5698,
+      "step": 11534
+    },
+    {
+      "epoch": 3.062126642771804,
+      "grad_norm": 0.44409449564925874,
+      "learning_rate": 2.4347204998999197e-06,
+      "loss": 0.5367,
+      "step": 11535
+    },
+    {
+      "epoch": 3.062392141245188,
+      "grad_norm": 0.45114788788934496,
+      "learning_rate": 2.434371491662584e-06,
+      "loss": 0.5324,
+      "step": 11536
+    },
+    {
+      "epoch": 3.0626576397185716,
+      "grad_norm": 0.45492169583577613,
+      "learning_rate": 2.434022484705164e-06,
+      "loss": 0.5946,
+      "step": 11537
+    },
+    {
+      "epoch": 3.0629231381919553,
+      "grad_norm": 0.44846522809331124,
+      "learning_rate": 2.433673479034467e-06,
+      "loss": 0.5552,
+      "step": 11538
+    },
+    {
+      "epoch": 3.0631886366653394,
+      "grad_norm": 0.46785575414042085,
+      "learning_rate": 2.4333244746573002e-06,
+      "loss": 0.5712,
+      "step": 11539
+    },
+    {
+      "epoch": 3.063454135138723,
+      "grad_norm": 0.46627529599984335,
+      "learning_rate": 2.432975471580469e-06,
+      "loss": 0.5668,
+      "step": 11540
+    },
+    {
+      "epoch": 3.0637196336121066,
+      "grad_norm": 0.4629146126354187,
+      "learning_rate": 2.4326264698107804e-06,
+      "loss": 0.5241,
+      "step": 11541
+    },
+    {
+      "epoch": 3.0639851320854907,
+      "grad_norm": 0.446532924828594,
+      "learning_rate": 2.432277469355042e-06,
+      "loss": 0.5326,
+      "step": 11542
+    },
+    {
+      "epoch": 3.0642506305588744,
+      "grad_norm": 0.44857166204253596,
+      "learning_rate": 2.4319284702200567e-06,
+      "loss": 0.545,
+      "step": 11543
+    },
+    {
+      "epoch": 3.064516129032258,
+      "grad_norm": 0.4684434820632688,
+      "learning_rate": 2.4315794724126336e-06,
+      "loss": 0.5306,
+      "step": 11544
+    },
+    {
+      "epoch": 3.0647816275056416,
+      "grad_norm": 0.4328742834057493,
+      "learning_rate": 2.431230475939578e-06,
+      "loss": 0.5201,
+      "step": 11545
+    },
+    {
+      "epoch": 3.0650471259790257,
+      "grad_norm": 0.4400257722068868,
+      "learning_rate": 2.4308814808076968e-06,
+      "loss": 0.5481,
+      "step": 11546
+    },
+    {
+      "epoch": 3.0653126244524094,
+      "grad_norm": 0.4578908908132894,
+      "learning_rate": 2.430532487023796e-06,
+      "loss": 0.5853,
+      "step": 11547
+    },
+    {
+      "epoch": 3.065578122925793,
+      "grad_norm": 0.45167266866409145,
+      "learning_rate": 2.4301834945946816e-06,
+      "loss": 0.5263,
+      "step": 11548
+    },
+    {
+      "epoch": 3.065843621399177,
+      "grad_norm": 0.45622879951136924,
+      "learning_rate": 2.42983450352716e-06,
+      "loss": 0.5612,
+      "step": 11549
+    },
+    {
+      "epoch": 3.0661091198725607,
+      "grad_norm": 0.4394482852577356,
+      "learning_rate": 2.4294855138280374e-06,
+      "loss": 0.5554,
+      "step": 11550
+    },
+    {
+      "epoch": 3.0663746183459444,
+      "grad_norm": 0.4488521127691896,
+      "learning_rate": 2.42913652550412e-06,
+      "loss": 0.5526,
+      "step": 11551
+    },
+    {
+      "epoch": 3.0666401168193285,
+      "grad_norm": 0.45438099409443067,
+      "learning_rate": 2.4287875385622137e-06,
+      "loss": 0.522,
+      "step": 11552
+    },
+    {
+      "epoch": 3.066905615292712,
+      "grad_norm": 0.44363928815443066,
+      "learning_rate": 2.428438553009126e-06,
+      "loss": 0.5281,
+      "step": 11553
+    },
+    {
+      "epoch": 3.0671711137660957,
+      "grad_norm": 0.4728075625948259,
+      "learning_rate": 2.42808956885166e-06,
+      "loss": 0.5897,
+      "step": 11554
+    },
+    {
+      "epoch": 3.06743661223948,
+      "grad_norm": 0.46362124702313573,
+      "learning_rate": 2.4277405860966237e-06,
+      "loss": 0.5903,
+      "step": 11555
+    },
+    {
+      "epoch": 3.0677021107128635,
+      "grad_norm": 0.45956941663315426,
+      "learning_rate": 2.427391604750823e-06,
+      "loss": 0.5339,
+      "step": 11556
+    },
+    {
+      "epoch": 3.067967609186247,
+      "grad_norm": 0.46096392490348004,
+      "learning_rate": 2.427042624821064e-06,
+      "loss": 0.5242,
+      "step": 11557
+    },
+    {
+      "epoch": 3.0682331076596308,
+      "grad_norm": 0.4506014384219963,
+      "learning_rate": 2.4266936463141523e-06,
+      "loss": 0.5642,
+      "step": 11558
+    },
+    {
+      "epoch": 3.068498606133015,
+      "grad_norm": 0.4445336789356763,
+      "learning_rate": 2.4263446692368957e-06,
+      "loss": 0.5364,
+      "step": 11559
+    },
+    {
+      "epoch": 3.0687641046063985,
+      "grad_norm": 0.4597987065181883,
+      "learning_rate": 2.4259956935960966e-06,
+      "loss": 0.5486,
+      "step": 11560
+    },
+    {
+      "epoch": 3.069029603079782,
+      "grad_norm": 0.4767547964968403,
+      "learning_rate": 2.4256467193985628e-06,
+      "loss": 0.5474,
+      "step": 11561
+    },
+    {
+      "epoch": 3.069295101553166,
+      "grad_norm": 0.4518195364294145,
+      "learning_rate": 2.4252977466511e-06,
+      "loss": 0.5378,
+      "step": 11562
+    },
+    {
+      "epoch": 3.06956060002655,
+      "grad_norm": 0.4335076815977469,
+      "learning_rate": 2.424948775360514e-06,
+      "loss": 0.5293,
+      "step": 11563
+    },
+    {
+      "epoch": 3.0698260984999335,
+      "grad_norm": 0.45401332283368523,
+      "learning_rate": 2.424599805533611e-06,
+      "loss": 0.5532,
+      "step": 11564
+    },
+    {
+      "epoch": 3.0700915969733176,
+      "grad_norm": 0.4579603467196406,
+      "learning_rate": 2.424250837177197e-06,
+      "loss": 0.5403,
+      "step": 11565
+    },
+    {
+      "epoch": 3.070357095446701,
+      "grad_norm": 0.47083952294624326,
+      "learning_rate": 2.4239018702980756e-06,
+      "loss": 0.5638,
+      "step": 11566
+    },
+    {
+      "epoch": 3.070622593920085,
+      "grad_norm": 0.44780449709131237,
+      "learning_rate": 2.423552904903055e-06,
+      "loss": 0.4998,
+      "step": 11567
+    },
+    {
+      "epoch": 3.070888092393469,
+      "grad_norm": 0.44363813725461565,
+      "learning_rate": 2.4232039409989397e-06,
+      "loss": 0.538,
+      "step": 11568
+    },
+    {
+      "epoch": 3.0711535908668526,
+      "grad_norm": 0.4614748067148487,
+      "learning_rate": 2.422854978592535e-06,
+      "loss": 0.5662,
+      "step": 11569
+    },
+    {
+      "epoch": 3.071419089340236,
+      "grad_norm": 0.44742343949730995,
+      "learning_rate": 2.4225060176906493e-06,
+      "loss": 0.5652,
+      "step": 11570
+    },
+    {
+      "epoch": 3.07168458781362,
+      "grad_norm": 0.45955378399355984,
+      "learning_rate": 2.422157058300084e-06,
+      "loss": 0.5279,
+      "step": 11571
+    },
+    {
+      "epoch": 3.071950086287004,
+      "grad_norm": 0.4379411837210556,
+      "learning_rate": 2.4218081004276474e-06,
+      "loss": 0.525,
+      "step": 11572
+    },
+    {
+      "epoch": 3.0722155847603876,
+      "grad_norm": 0.4390836364029525,
+      "learning_rate": 2.4214591440801438e-06,
+      "loss": 0.5458,
+      "step": 11573
+    },
+    {
+      "epoch": 3.0724810832337712,
+      "grad_norm": 0.4630505390553413,
+      "learning_rate": 2.4211101892643793e-06,
+      "loss": 0.5316,
+      "step": 11574
+    },
+    {
+      "epoch": 3.0727465817071553,
+      "grad_norm": 0.4421934962279115,
+      "learning_rate": 2.4207612359871587e-06,
+      "loss": 0.5079,
+      "step": 11575
+    },
+    {
+      "epoch": 3.073012080180539,
+      "grad_norm": 0.43016732646785627,
+      "learning_rate": 2.4204122842552896e-06,
+      "loss": 0.4962,
+      "step": 11576
+    },
+    {
+      "epoch": 3.0732775786539226,
+      "grad_norm": 0.4426651867660987,
+      "learning_rate": 2.4200633340755745e-06,
+      "loss": 0.5519,
+      "step": 11577
+    },
+    {
+      "epoch": 3.0735430771273067,
+      "grad_norm": 0.4498533546661004,
+      "learning_rate": 2.419714385454821e-06,
+      "loss": 0.5424,
+      "step": 11578
+    },
+    {
+      "epoch": 3.0738085756006903,
+      "grad_norm": 0.4484563663286284,
+      "learning_rate": 2.419365438399833e-06,
+      "loss": 0.5365,
+      "step": 11579
+    },
+    {
+      "epoch": 3.074074074074074,
+      "grad_norm": 0.43730048726453674,
+      "learning_rate": 2.419016492917416e-06,
+      "loss": 0.559,
+      "step": 11580
+    },
+    {
+      "epoch": 3.074339572547458,
+      "grad_norm": 0.4633450421778798,
+      "learning_rate": 2.418667549014376e-06,
+      "loss": 0.5554,
+      "step": 11581
+    },
+    {
+      "epoch": 3.0746050710208417,
+      "grad_norm": 0.45532692397865465,
+      "learning_rate": 2.4183186066975182e-06,
+      "loss": 0.5612,
+      "step": 11582
+    },
+    {
+      "epoch": 3.0748705694942253,
+      "grad_norm": 0.4566625176032235,
+      "learning_rate": 2.4179696659736474e-06,
+      "loss": 0.5285,
+      "step": 11583
+    },
+    {
+      "epoch": 3.075136067967609,
+      "grad_norm": 0.4431252456370272,
+      "learning_rate": 2.417620726849569e-06,
+      "loss": 0.5742,
+      "step": 11584
+    },
+    {
+      "epoch": 3.075401566440993,
+      "grad_norm": 0.4505934205915547,
+      "learning_rate": 2.417271789332088e-06,
+      "loss": 0.5351,
+      "step": 11585
+    },
+    {
+      "epoch": 3.0756670649143767,
+      "grad_norm": 0.4500751514309848,
+      "learning_rate": 2.4169228534280094e-06,
+      "loss": 0.5317,
+      "step": 11586
+    },
+    {
+      "epoch": 3.0759325633877603,
+      "grad_norm": 0.4659284138188159,
+      "learning_rate": 2.416573919144139e-06,
+      "loss": 0.5175,
+      "step": 11587
+    },
+    {
+      "epoch": 3.0761980618611444,
+      "grad_norm": 0.4493922908947943,
+      "learning_rate": 2.416224986487282e-06,
+      "loss": 0.58,
+      "step": 11588
+    },
+    {
+      "epoch": 3.076463560334528,
+      "grad_norm": 0.4667893481498818,
+      "learning_rate": 2.4158760554642423e-06,
+      "loss": 0.6034,
+      "step": 11589
+    },
+    {
+      "epoch": 3.0767290588079117,
+      "grad_norm": 0.44960921373662366,
+      "learning_rate": 2.4155271260818254e-06,
+      "loss": 0.5262,
+      "step": 11590
+    },
+    {
+      "epoch": 3.076994557281296,
+      "grad_norm": 0.47163116028064456,
+      "learning_rate": 2.415178198346836e-06,
+      "loss": 0.5805,
+      "step": 11591
+    },
+    {
+      "epoch": 3.0772600557546794,
+      "grad_norm": 0.45691547889692313,
+      "learning_rate": 2.4148292722660798e-06,
+      "loss": 0.5693,
+      "step": 11592
+    },
+    {
+      "epoch": 3.077525554228063,
+      "grad_norm": 0.4543799342822913,
+      "learning_rate": 2.414480347846362e-06,
+      "loss": 0.5139,
+      "step": 11593
+    },
+    {
+      "epoch": 3.077791052701447,
+      "grad_norm": 0.44919058844425375,
+      "learning_rate": 2.414131425094486e-06,
+      "loss": 0.5377,
+      "step": 11594
+    },
+    {
+      "epoch": 3.078056551174831,
+      "grad_norm": 0.45385467004272423,
+      "learning_rate": 2.413782504017257e-06,
+      "loss": 0.5333,
+      "step": 11595
+    },
+    {
+      "epoch": 3.0783220496482144,
+      "grad_norm": 0.456646600906358,
+      "learning_rate": 2.413433584621481e-06,
+      "loss": 0.5529,
+      "step": 11596
+    },
+    {
+      "epoch": 3.0785875481215985,
+      "grad_norm": 0.4628491841440177,
+      "learning_rate": 2.413084666913962e-06,
+      "loss": 0.552,
+      "step": 11597
+    },
+    {
+      "epoch": 3.078853046594982,
+      "grad_norm": 0.43762217548632976,
+      "learning_rate": 2.4127357509015045e-06,
+      "loss": 0.5567,
+      "step": 11598
+    },
+    {
+      "epoch": 3.079118545068366,
+      "grad_norm": 0.442194338312714,
+      "learning_rate": 2.412386836590914e-06,
+      "loss": 0.536,
+      "step": 11599
+    },
+    {
+      "epoch": 3.0793840435417494,
+      "grad_norm": 0.43828667231465074,
+      "learning_rate": 2.4120379239889944e-06,
+      "loss": 0.5367,
+      "step": 11600
+    },
+    {
+      "epoch": 3.0796495420151335,
+      "grad_norm": 0.44867218626389876,
+      "learning_rate": 2.4116890131025503e-06,
+      "loss": 0.5749,
+      "step": 11601
+    },
+    {
+      "epoch": 3.079915040488517,
+      "grad_norm": 0.4601512494034246,
+      "learning_rate": 2.4113401039383867e-06,
+      "loss": 0.5599,
+      "step": 11602
+    },
+    {
+      "epoch": 3.080180538961901,
+      "grad_norm": 0.43663244776695664,
+      "learning_rate": 2.410991196503309e-06,
+      "loss": 0.5661,
+      "step": 11603
+    },
+    {
+      "epoch": 3.080446037435285,
+      "grad_norm": 0.4479055628569839,
+      "learning_rate": 2.41064229080412e-06,
+      "loss": 0.6069,
+      "step": 11604
+    },
+    {
+      "epoch": 3.0807115359086685,
+      "grad_norm": 0.4575099913725818,
+      "learning_rate": 2.4102933868476263e-06,
+      "loss": 0.562,
+      "step": 11605
+    },
+    {
+      "epoch": 3.080977034382052,
+      "grad_norm": 0.45618847925682265,
+      "learning_rate": 2.4099444846406305e-06,
+      "loss": 0.5676,
+      "step": 11606
+    },
+    {
+      "epoch": 3.0812425328554363,
+      "grad_norm": 0.4532304945237393,
+      "learning_rate": 2.4095955841899372e-06,
+      "loss": 0.5424,
+      "step": 11607
+    },
+    {
+      "epoch": 3.08150803132882,
+      "grad_norm": 0.4529452313010965,
+      "learning_rate": 2.409246685502352e-06,
+      "loss": 0.544,
+      "step": 11608
+    },
+    {
+      "epoch": 3.0817735298022035,
+      "grad_norm": 0.4481273594910094,
+      "learning_rate": 2.408897788584678e-06,
+      "loss": 0.5476,
+      "step": 11609
+    },
+    {
+      "epoch": 3.0820390282755876,
+      "grad_norm": 0.4424810980465838,
+      "learning_rate": 2.408548893443721e-06,
+      "loss": 0.5383,
+      "step": 11610
+    },
+    {
+      "epoch": 3.0823045267489713,
+      "grad_norm": 0.4502165308519758,
+      "learning_rate": 2.408200000086284e-06,
+      "loss": 0.5416,
+      "step": 11611
+    },
+    {
+      "epoch": 3.082570025222355,
+      "grad_norm": 0.44556179145103497,
+      "learning_rate": 2.407851108519172e-06,
+      "loss": 0.5431,
+      "step": 11612
+    },
+    {
+      "epoch": 3.0828355236957385,
+      "grad_norm": 0.4523120791059323,
+      "learning_rate": 2.407502218749189e-06,
+      "loss": 0.5523,
+      "step": 11613
+    },
+    {
+      "epoch": 3.0831010221691226,
+      "grad_norm": 0.4489360115967645,
+      "learning_rate": 2.4071533307831386e-06,
+      "loss": 0.5901,
+      "step": 11614
+    },
+    {
+      "epoch": 3.0833665206425063,
+      "grad_norm": 0.4602028790946236,
+      "learning_rate": 2.406804444627826e-06,
+      "loss": 0.5387,
+      "step": 11615
+    },
+    {
+      "epoch": 3.08363201911589,
+      "grad_norm": 0.4352331176447486,
+      "learning_rate": 2.4064555602900557e-06,
+      "loss": 0.5346,
+      "step": 11616
+    },
+    {
+      "epoch": 3.083897517589274,
+      "grad_norm": 0.4685939320709742,
+      "learning_rate": 2.4061066777766302e-06,
+      "loss": 0.5455,
+      "step": 11617
+    },
+    {
+      "epoch": 3.0841630160626576,
+      "grad_norm": 0.4553900045941404,
+      "learning_rate": 2.4057577970943544e-06,
+      "loss": 0.5548,
+      "step": 11618
+    },
+    {
+      "epoch": 3.0844285145360413,
+      "grad_norm": 0.45309101627219217,
+      "learning_rate": 2.4054089182500325e-06,
+      "loss": 0.539,
+      "step": 11619
+    },
+    {
+      "epoch": 3.0846940130094254,
+      "grad_norm": 0.4555084307680555,
+      "learning_rate": 2.405060041250467e-06,
+      "loss": 0.5423,
+      "step": 11620
+    },
+    {
+      "epoch": 3.084959511482809,
+      "grad_norm": 0.44265387932136907,
+      "learning_rate": 2.4047111661024644e-06,
+      "loss": 0.5475,
+      "step": 11621
+    },
+    {
+      "epoch": 3.0852250099561926,
+      "grad_norm": 0.4573892909316085,
+      "learning_rate": 2.4043622928128283e-06,
+      "loss": 0.5436,
+      "step": 11622
+    },
+    {
+      "epoch": 3.0854905084295767,
+      "grad_norm": 0.46787178157115983,
+      "learning_rate": 2.4040134213883604e-06,
+      "loss": 0.6086,
+      "step": 11623
+    },
+    {
+      "epoch": 3.0857560069029604,
+      "grad_norm": 0.44104109923536905,
+      "learning_rate": 2.403664551835866e-06,
+      "loss": 0.5712,
+      "step": 11624
+    },
+    {
+      "epoch": 3.086021505376344,
+      "grad_norm": 0.45532908072829414,
+      "learning_rate": 2.4033156841621484e-06,
+      "loss": 0.559,
+      "step": 11625
+    },
+    {
+      "epoch": 3.0862870038497277,
+      "grad_norm": 0.4511078639346571,
+      "learning_rate": 2.402966818374012e-06,
+      "loss": 0.5619,
+      "step": 11626
+    },
+    {
+      "epoch": 3.0865525023231117,
+      "grad_norm": 0.45810234006875333,
+      "learning_rate": 2.40261795447826e-06,
+      "loss": 0.5606,
+      "step": 11627
+    },
+    {
+      "epoch": 3.0868180007964954,
+      "grad_norm": 0.45580471688257507,
+      "learning_rate": 2.402269092481697e-06,
+      "loss": 0.5184,
+      "step": 11628
+    },
+    {
+      "epoch": 3.087083499269879,
+      "grad_norm": 0.4471135425014241,
+      "learning_rate": 2.4019202323911255e-06,
+      "loss": 0.5235,
+      "step": 11629
+    },
+    {
+      "epoch": 3.087348997743263,
+      "grad_norm": 0.4528342444019298,
+      "learning_rate": 2.4015713742133495e-06,
+      "loss": 0.5489,
+      "step": 11630
+    },
+    {
+      "epoch": 3.0876144962166467,
+      "grad_norm": 0.47491126730620253,
+      "learning_rate": 2.4012225179551726e-06,
+      "loss": 0.5817,
+      "step": 11631
+    },
+    {
+      "epoch": 3.0878799946900304,
+      "grad_norm": 0.436825252540165,
+      "learning_rate": 2.4008736636233985e-06,
+      "loss": 0.532,
+      "step": 11632
+    },
+    {
+      "epoch": 3.0881454931634145,
+      "grad_norm": 0.4537553387947287,
+      "learning_rate": 2.400524811224831e-06,
+      "loss": 0.5413,
+      "step": 11633
+    },
+    {
+      "epoch": 3.088410991636798,
+      "grad_norm": 0.45491351212053394,
+      "learning_rate": 2.400175960766274e-06,
+      "loss": 0.5718,
+      "step": 11634
+    },
+    {
+      "epoch": 3.0886764901101817,
+      "grad_norm": 0.4725039325195927,
+      "learning_rate": 2.3998271122545287e-06,
+      "loss": 0.5829,
+      "step": 11635
+    },
+    {
+      "epoch": 3.088941988583566,
+      "grad_norm": 0.45135817280529056,
+      "learning_rate": 2.3994782656964007e-06,
+      "loss": 0.5453,
+      "step": 11636
+    },
+    {
+      "epoch": 3.0892074870569495,
+      "grad_norm": 0.44006000910393317,
+      "learning_rate": 2.3991294210986924e-06,
+      "loss": 0.5502,
+      "step": 11637
+    },
+    {
+      "epoch": 3.089472985530333,
+      "grad_norm": 0.4545602481533639,
+      "learning_rate": 2.398780578468207e-06,
+      "loss": 0.5488,
+      "step": 11638
+    },
+    {
+      "epoch": 3.0897384840037168,
+      "grad_norm": 0.44864899029729133,
+      "learning_rate": 2.398431737811749e-06,
+      "loss": 0.5697,
+      "step": 11639
+    },
+    {
+      "epoch": 3.090003982477101,
+      "grad_norm": 0.45269264897709593,
+      "learning_rate": 2.3980828991361206e-06,
+      "loss": 0.5524,
+      "step": 11640
+    },
+    {
+      "epoch": 3.0902694809504845,
+      "grad_norm": 0.4386700168310677,
+      "learning_rate": 2.3977340624481245e-06,
+      "loss": 0.5677,
+      "step": 11641
+    },
+    {
+      "epoch": 3.090534979423868,
+      "grad_norm": 0.4546090687187948,
+      "learning_rate": 2.397385227754565e-06,
+      "loss": 0.5476,
+      "step": 11642
+    },
+    {
+      "epoch": 3.090800477897252,
+      "grad_norm": 0.44710709300459756,
+      "learning_rate": 2.397036395062245e-06,
+      "loss": 0.5243,
+      "step": 11643
+    },
+    {
+      "epoch": 3.091065976370636,
+      "grad_norm": 0.4458780256927925,
+      "learning_rate": 2.396687564377967e-06,
+      "loss": 0.5855,
+      "step": 11644
+    },
+    {
+      "epoch": 3.0913314748440195,
+      "grad_norm": 0.46020534861082846,
+      "learning_rate": 2.3963387357085353e-06,
+      "loss": 0.5749,
+      "step": 11645
+    },
+    {
+      "epoch": 3.0915969733174036,
+      "grad_norm": 0.4543410836247079,
+      "learning_rate": 2.395989909060751e-06,
+      "loss": 0.5385,
+      "step": 11646
+    },
+    {
+      "epoch": 3.091862471790787,
+      "grad_norm": 0.4629066868969013,
+      "learning_rate": 2.3956410844414185e-06,
+      "loss": 0.5425,
+      "step": 11647
+    },
+    {
+      "epoch": 3.092127970264171,
+      "grad_norm": 0.44968546688436994,
+      "learning_rate": 2.3952922618573404e-06,
+      "loss": 0.5397,
+      "step": 11648
+    },
+    {
+      "epoch": 3.092393468737555,
+      "grad_norm": 0.4644782799429297,
+      "learning_rate": 2.3949434413153195e-06,
+      "loss": 0.5276,
+      "step": 11649
+    },
+    {
+      "epoch": 3.0926589672109386,
+      "grad_norm": 0.44689705541425445,
+      "learning_rate": 2.394594622822159e-06,
+      "loss": 0.542,
+      "step": 11650
+    },
+    {
+      "epoch": 3.092924465684322,
+      "grad_norm": 0.46193158441804544,
+      "learning_rate": 2.394245806384662e-06,
+      "loss": 0.544,
+      "step": 11651
+    },
+    {
+      "epoch": 3.0931899641577063,
+      "grad_norm": 0.4511477857575874,
+      "learning_rate": 2.39389699200963e-06,
+      "loss": 0.5545,
+      "step": 11652
+    },
+    {
+      "epoch": 3.09345546263109,
+      "grad_norm": 0.45913925280558493,
+      "learning_rate": 2.393548179703866e-06,
+      "loss": 0.5307,
+      "step": 11653
+    },
+    {
+      "epoch": 3.0937209611044736,
+      "grad_norm": 0.4584177238231391,
+      "learning_rate": 2.3931993694741737e-06,
+      "loss": 0.5492,
+      "step": 11654
+    },
+    {
+      "epoch": 3.0939864595778572,
+      "grad_norm": 0.44236769846587864,
+      "learning_rate": 2.3928505613273546e-06,
+      "loss": 0.5414,
+      "step": 11655
+    },
+    {
+      "epoch": 3.0942519580512413,
+      "grad_norm": 0.46533226353690077,
+      "learning_rate": 2.3925017552702125e-06,
+      "loss": 0.525,
+      "step": 11656
+    },
+    {
+      "epoch": 3.094517456524625,
+      "grad_norm": 0.44556117700918196,
+      "learning_rate": 2.3921529513095497e-06,
+      "loss": 0.5265,
+      "step": 11657
+    },
+    {
+      "epoch": 3.0947829549980086,
+      "grad_norm": 0.46380813219517747,
+      "learning_rate": 2.3918041494521676e-06,
+      "loss": 0.5134,
+      "step": 11658
+    },
+    {
+      "epoch": 3.0950484534713927,
+      "grad_norm": 0.4439846358567936,
+      "learning_rate": 2.39145534970487e-06,
+      "loss": 0.5587,
+      "step": 11659
+    },
+    {
+      "epoch": 3.0953139519447763,
+      "grad_norm": 0.4755854379493961,
+      "learning_rate": 2.3911065520744583e-06,
+      "loss": 0.5653,
+      "step": 11660
+    },
+    {
+      "epoch": 3.09557945041816,
+      "grad_norm": 0.45121226205263937,
+      "learning_rate": 2.390757756567736e-06,
+      "loss": 0.5548,
+      "step": 11661
+    },
+    {
+      "epoch": 3.095844948891544,
+      "grad_norm": 0.45490569864651714,
+      "learning_rate": 2.3904089631915044e-06,
+      "loss": 0.5551,
+      "step": 11662
+    },
+    {
+      "epoch": 3.0961104473649277,
+      "grad_norm": 0.4496392386047114,
+      "learning_rate": 2.3900601719525676e-06,
+      "loss": 0.5316,
+      "step": 11663
+    },
+    {
+      "epoch": 3.0963759458383113,
+      "grad_norm": 0.46330358960693546,
+      "learning_rate": 2.3897113828577256e-06,
+      "loss": 0.5392,
+      "step": 11664
+    },
+    {
+      "epoch": 3.0966414443116954,
+      "grad_norm": 0.46344527653492346,
+      "learning_rate": 2.3893625959137816e-06,
+      "loss": 0.5749,
+      "step": 11665
+    },
+    {
+      "epoch": 3.096906942785079,
+      "grad_norm": 0.4533480070358089,
+      "learning_rate": 2.3890138111275375e-06,
+      "loss": 0.5191,
+      "step": 11666
+    },
+    {
+      "epoch": 3.0971724412584627,
+      "grad_norm": 0.4546032152339237,
+      "learning_rate": 2.3886650285057963e-06,
+      "loss": 0.5772,
+      "step": 11667
+    },
+    {
+      "epoch": 3.0974379397318463,
+      "grad_norm": 0.4460242801885309,
+      "learning_rate": 2.3883162480553605e-06,
+      "loss": 0.5293,
+      "step": 11668
+    },
+    {
+      "epoch": 3.0977034382052304,
+      "grad_norm": 0.45026723526569723,
+      "learning_rate": 2.38796746978303e-06,
+      "loss": 0.5721,
+      "step": 11669
+    },
+    {
+      "epoch": 3.097968936678614,
+      "grad_norm": 0.44958993970359756,
+      "learning_rate": 2.387618693695609e-06,
+      "loss": 0.5762,
+      "step": 11670
+    },
+    {
+      "epoch": 3.0982344351519977,
+      "grad_norm": 0.45386416511552574,
+      "learning_rate": 2.387269919799898e-06,
+      "loss": 0.5449,
+      "step": 11671
+    },
+    {
+      "epoch": 3.098499933625382,
+      "grad_norm": 0.467705826561258,
+      "learning_rate": 2.3869211481027e-06,
+      "loss": 0.5568,
+      "step": 11672
+    },
+    {
+      "epoch": 3.0987654320987654,
+      "grad_norm": 0.45373042329394825,
+      "learning_rate": 2.386572378610816e-06,
+      "loss": 0.5748,
+      "step": 11673
+    },
+    {
+      "epoch": 3.099030930572149,
+      "grad_norm": 0.46178207720684394,
+      "learning_rate": 2.386223611331049e-06,
+      "loss": 0.56,
+      "step": 11674
+    },
+    {
+      "epoch": 3.099296429045533,
+      "grad_norm": 0.46482844879008234,
+      "learning_rate": 2.3858748462701996e-06,
+      "loss": 0.6261,
+      "step": 11675
+    },
+    {
+      "epoch": 3.099561927518917,
+      "grad_norm": 0.45341897629162803,
+      "learning_rate": 2.38552608343507e-06,
+      "loss": 0.5616,
+      "step": 11676
+    },
+    {
+      "epoch": 3.0998274259923004,
+      "grad_norm": 0.4503636359842079,
+      "learning_rate": 2.3851773228324624e-06,
+      "loss": 0.5285,
+      "step": 11677
+    },
+    {
+      "epoch": 3.1000929244656845,
+      "grad_norm": 0.4501312289403777,
+      "learning_rate": 2.384828564469178e-06,
+      "loss": 0.5199,
+      "step": 11678
+    },
+    {
+      "epoch": 3.100358422939068,
+      "grad_norm": 0.4509428598065902,
+      "learning_rate": 2.3844798083520187e-06,
+      "loss": 0.535,
+      "step": 11679
+    },
+    {
+      "epoch": 3.100623921412452,
+      "grad_norm": 0.4575445265556578,
+      "learning_rate": 2.384131054487787e-06,
+      "loss": 0.5311,
+      "step": 11680
+    },
+    {
+      "epoch": 3.100889419885836,
+      "grad_norm": 0.4549972502640145,
+      "learning_rate": 2.3837823028832825e-06,
+      "loss": 0.5376,
+      "step": 11681
+    },
+    {
+      "epoch": 3.1011549183592195,
+      "grad_norm": 0.46139631713274615,
+      "learning_rate": 2.3834335535453075e-06,
+      "loss": 0.5504,
+      "step": 11682
+    },
+    {
+      "epoch": 3.101420416832603,
+      "grad_norm": 0.4474409976447581,
+      "learning_rate": 2.3830848064806635e-06,
+      "loss": 0.5078,
+      "step": 11683
+    },
+    {
+      "epoch": 3.101685915305987,
+      "grad_norm": 0.4629258899056781,
+      "learning_rate": 2.382736061696152e-06,
+      "loss": 0.5145,
+      "step": 11684
+    },
+    {
+      "epoch": 3.101951413779371,
+      "grad_norm": 0.43832233207832816,
+      "learning_rate": 2.3823873191985743e-06,
+      "loss": 0.5674,
+      "step": 11685
+    },
+    {
+      "epoch": 3.1022169122527545,
+      "grad_norm": 0.45596779082584954,
+      "learning_rate": 2.382038578994733e-06,
+      "loss": 0.5566,
+      "step": 11686
+    },
+    {
+      "epoch": 3.102482410726138,
+      "grad_norm": 0.44157542877360395,
+      "learning_rate": 2.381689841091427e-06,
+      "loss": 0.5312,
+      "step": 11687
+    },
+    {
+      "epoch": 3.1027479091995223,
+      "grad_norm": 0.4470285227041901,
+      "learning_rate": 2.381341105495459e-06,
+      "loss": 0.5649,
+      "step": 11688
+    },
+    {
+      "epoch": 3.103013407672906,
+      "grad_norm": 0.46329320383695966,
+      "learning_rate": 2.3809923722136298e-06,
+      "loss": 0.5778,
+      "step": 11689
+    },
+    {
+      "epoch": 3.1032789061462895,
+      "grad_norm": 0.4692033651479321,
+      "learning_rate": 2.380643641252741e-06,
+      "loss": 0.5229,
+      "step": 11690
+    },
+    {
+      "epoch": 3.1035444046196736,
+      "grad_norm": 0.44616414578282565,
+      "learning_rate": 2.380294912619594e-06,
+      "loss": 0.5341,
+      "step": 11691
+    },
+    {
+      "epoch": 3.1038099030930573,
+      "grad_norm": 0.45234813987575695,
+      "learning_rate": 2.379946186320989e-06,
+      "loss": 0.5205,
+      "step": 11692
+    },
+    {
+      "epoch": 3.104075401566441,
+      "grad_norm": 0.43463918132220514,
+      "learning_rate": 2.3795974623637265e-06,
+      "loss": 0.5556,
+      "step": 11693
+    },
+    {
+      "epoch": 3.1043409000398245,
+      "grad_norm": 0.44083089330814906,
+      "learning_rate": 2.3792487407546087e-06,
+      "loss": 0.5355,
+      "step": 11694
+    },
+    {
+      "epoch": 3.1046063985132086,
+      "grad_norm": 0.4530164680787931,
+      "learning_rate": 2.3789000215004366e-06,
+      "loss": 0.562,
+      "step": 11695
+    },
+    {
+      "epoch": 3.1048718969865923,
+      "grad_norm": 0.45947491042120703,
+      "learning_rate": 2.37855130460801e-06,
+      "loss": 0.5215,
+      "step": 11696
+    },
+    {
+      "epoch": 3.105137395459976,
+      "grad_norm": 0.44575068308308474,
+      "learning_rate": 2.3782025900841314e-06,
+      "loss": 0.5885,
+      "step": 11697
+    },
+    {
+      "epoch": 3.10540289393336,
+      "grad_norm": 0.451986042189683,
+      "learning_rate": 2.3778538779355996e-06,
+      "loss": 0.5491,
+      "step": 11698
+    },
+    {
+      "epoch": 3.1056683924067436,
+      "grad_norm": 0.45402515184793113,
+      "learning_rate": 2.3775051681692162e-06,
+      "loss": 0.5935,
+      "step": 11699
+    },
+    {
+      "epoch": 3.1059338908801273,
+      "grad_norm": 0.4751543136792409,
+      "learning_rate": 2.377156460791782e-06,
+      "loss": 0.5564,
+      "step": 11700
+    },
+    {
+      "epoch": 3.1061993893535114,
+      "grad_norm": 0.45009078362303184,
+      "learning_rate": 2.3768077558100974e-06,
+      "loss": 0.5379,
+      "step": 11701
+    },
+    {
+      "epoch": 3.106464887826895,
+      "grad_norm": 0.4570427369997015,
+      "learning_rate": 2.3764590532309635e-06,
+      "loss": 0.5175,
+      "step": 11702
+    },
+    {
+      "epoch": 3.1067303863002786,
+      "grad_norm": 0.45386640858526156,
+      "learning_rate": 2.376110353061181e-06,
+      "loss": 0.5548,
+      "step": 11703
+    },
+    {
+      "epoch": 3.1069958847736627,
+      "grad_norm": 0.4518641702102767,
+      "learning_rate": 2.3757616553075493e-06,
+      "loss": 0.5569,
+      "step": 11704
+    },
+    {
+      "epoch": 3.1072613832470464,
+      "grad_norm": 0.45041484996366055,
+      "learning_rate": 2.3754129599768694e-06,
+      "loss": 0.5692,
+      "step": 11705
+    },
+    {
+      "epoch": 3.10752688172043,
+      "grad_norm": 0.4554978315785331,
+      "learning_rate": 2.375064267075942e-06,
+      "loss": 0.5716,
+      "step": 11706
+    },
+    {
+      "epoch": 3.107792380193814,
+      "grad_norm": 0.44741418343448375,
+      "learning_rate": 2.374715576611567e-06,
+      "loss": 0.4959,
+      "step": 11707
+    },
+    {
+      "epoch": 3.1080578786671977,
+      "grad_norm": 0.45503747099599895,
+      "learning_rate": 2.3743668885905457e-06,
+      "loss": 0.5518,
+      "step": 11708
+    },
+    {
+      "epoch": 3.1083233771405814,
+      "grad_norm": 0.4431368452438896,
+      "learning_rate": 2.374018203019678e-06,
+      "loss": 0.5308,
+      "step": 11709
+    },
+    {
+      "epoch": 3.108588875613965,
+      "grad_norm": 0.4458652674102593,
+      "learning_rate": 2.373669519905763e-06,
+      "loss": 0.582,
+      "step": 11710
+    },
+    {
+      "epoch": 3.108854374087349,
+      "grad_norm": 0.4692365630070526,
+      "learning_rate": 2.373320839255601e-06,
+      "loss": 0.5427,
+      "step": 11711
+    },
+    {
+      "epoch": 3.1091198725607327,
+      "grad_norm": 0.4600862992202426,
+      "learning_rate": 2.3729721610759938e-06,
+      "loss": 0.5286,
+      "step": 11712
+    },
+    {
+      "epoch": 3.1093853710341164,
+      "grad_norm": 0.4695259293918339,
+      "learning_rate": 2.37262348537374e-06,
+      "loss": 0.5427,
+      "step": 11713
+    },
+    {
+      "epoch": 3.1096508695075005,
+      "grad_norm": 0.4514257706064871,
+      "learning_rate": 2.3722748121556413e-06,
+      "loss": 0.5582,
+      "step": 11714
+    },
+    {
+      "epoch": 3.109916367980884,
+      "grad_norm": 0.4403443261707449,
+      "learning_rate": 2.3719261414284957e-06,
+      "loss": 0.5484,
+      "step": 11715
+    },
+    {
+      "epoch": 3.1101818664542678,
+      "grad_norm": 0.45837862873268126,
+      "learning_rate": 2.3715774731991034e-06,
+      "loss": 0.5788,
+      "step": 11716
+    },
+    {
+      "epoch": 3.110447364927652,
+      "grad_norm": 0.4473199203319095,
+      "learning_rate": 2.371228807474265e-06,
+      "loss": 0.502,
+      "step": 11717
+    },
+    {
+      "epoch": 3.1107128634010355,
+      "grad_norm": 0.45132572744699595,
+      "learning_rate": 2.3708801442607803e-06,
+      "loss": 0.5407,
+      "step": 11718
+    },
+    {
+      "epoch": 3.110978361874419,
+      "grad_norm": 0.44424377391294806,
+      "learning_rate": 2.3705314835654487e-06,
+      "loss": 0.5395,
+      "step": 11719
+    },
+    {
+      "epoch": 3.111243860347803,
+      "grad_norm": 0.4457680201596601,
+      "learning_rate": 2.370182825395071e-06,
+      "loss": 0.5301,
+      "step": 11720
+    },
+    {
+      "epoch": 3.111509358821187,
+      "grad_norm": 0.44869578356096657,
+      "learning_rate": 2.3698341697564452e-06,
+      "loss": 0.5489,
+      "step": 11721
+    },
+    {
+      "epoch": 3.1117748572945705,
+      "grad_norm": 0.45168581768490124,
+      "learning_rate": 2.369485516656372e-06,
+      "loss": 0.5523,
+      "step": 11722
+    },
+    {
+      "epoch": 3.112040355767954,
+      "grad_norm": 0.45681395350376636,
+      "learning_rate": 2.3691368661016507e-06,
+      "loss": 0.5215,
+      "step": 11723
+    },
+    {
+      "epoch": 3.112305854241338,
+      "grad_norm": 0.43744741076446336,
+      "learning_rate": 2.368788218099081e-06,
+      "loss": 0.5263,
+      "step": 11724
+    },
+    {
+      "epoch": 3.112571352714722,
+      "grad_norm": 0.44594978697550103,
+      "learning_rate": 2.3684395726554625e-06,
+      "loss": 0.5318,
+      "step": 11725
+    },
+    {
+      "epoch": 3.1128368511881055,
+      "grad_norm": 0.45331967052763983,
+      "learning_rate": 2.368090929777595e-06,
+      "loss": 0.5381,
+      "step": 11726
+    },
+    {
+      "epoch": 3.1131023496614896,
+      "grad_norm": 0.4694669563525516,
+      "learning_rate": 2.367742289472277e-06,
+      "loss": 0.559,
+      "step": 11727
+    },
+    {
+      "epoch": 3.113367848134873,
+      "grad_norm": 0.44937769449546033,
+      "learning_rate": 2.3673936517463075e-06,
+      "loss": 0.5295,
+      "step": 11728
+    },
+    {
+      "epoch": 3.113633346608257,
+      "grad_norm": 0.4563437266288898,
+      "learning_rate": 2.367045016606487e-06,
+      "loss": 0.5389,
+      "step": 11729
+    },
+    {
+      "epoch": 3.113898845081641,
+      "grad_norm": 0.4668787617468006,
+      "learning_rate": 2.3666963840596135e-06,
+      "loss": 0.5418,
+      "step": 11730
+    },
+    {
+      "epoch": 3.1141643435550246,
+      "grad_norm": 0.4474620015251689,
+      "learning_rate": 2.3663477541124876e-06,
+      "loss": 0.5515,
+      "step": 11731
+    },
+    {
+      "epoch": 3.1144298420284082,
+      "grad_norm": 0.45798040602451795,
+      "learning_rate": 2.3659991267719078e-06,
+      "loss": 0.5885,
+      "step": 11732
+    },
+    {
+      "epoch": 3.1146953405017923,
+      "grad_norm": 0.4669521626151058,
+      "learning_rate": 2.365650502044673e-06,
+      "loss": 0.5374,
+      "step": 11733
+    },
+    {
+      "epoch": 3.114960838975176,
+      "grad_norm": 0.45737778115290334,
+      "learning_rate": 2.365301879937582e-06,
+      "loss": 0.5623,
+      "step": 11734
+    },
+    {
+      "epoch": 3.1152263374485596,
+      "grad_norm": 0.4586176356981947,
+      "learning_rate": 2.364953260457434e-06,
+      "loss": 0.5475,
+      "step": 11735
+    },
+    {
+      "epoch": 3.1154918359219437,
+      "grad_norm": 0.4476921469239407,
+      "learning_rate": 2.364604643611028e-06,
+      "loss": 0.5353,
+      "step": 11736
+    },
+    {
+      "epoch": 3.1157573343953273,
+      "grad_norm": 0.4801789942492191,
+      "learning_rate": 2.3642560294051643e-06,
+      "loss": 0.5566,
+      "step": 11737
+    },
+    {
+      "epoch": 3.116022832868711,
+      "grad_norm": 0.4470597776203245,
+      "learning_rate": 2.363907417846639e-06,
+      "loss": 0.5745,
+      "step": 11738
+    },
+    {
+      "epoch": 3.1162883313420946,
+      "grad_norm": 0.44458031074757975,
+      "learning_rate": 2.3635588089422528e-06,
+      "loss": 0.5473,
+      "step": 11739
+    },
+    {
+      "epoch": 3.1165538298154787,
+      "grad_norm": 0.4935263470663776,
+      "learning_rate": 2.3632102026988035e-06,
+      "loss": 0.4716,
+      "step": 11740
+    },
+    {
+      "epoch": 3.1168193282888623,
+      "grad_norm": 0.45781509398117604,
+      "learning_rate": 2.3628615991230904e-06,
+      "loss": 0.5663,
+      "step": 11741
+    },
+    {
+      "epoch": 3.117084826762246,
+      "grad_norm": 0.4407026604343912,
+      "learning_rate": 2.362512998221912e-06,
+      "loss": 0.5541,
+      "step": 11742
+    },
+    {
+      "epoch": 3.11735032523563,
+      "grad_norm": 0.43370469008587437,
+      "learning_rate": 2.3621644000020677e-06,
+      "loss": 0.5501,
+      "step": 11743
+    },
+    {
+      "epoch": 3.1176158237090137,
+      "grad_norm": 0.44236715688911166,
+      "learning_rate": 2.361815804470354e-06,
+      "loss": 0.5239,
+      "step": 11744
+    },
+    {
+      "epoch": 3.1178813221823973,
+      "grad_norm": 0.4454365951175594,
+      "learning_rate": 2.3614672116335704e-06,
+      "loss": 0.5402,
+      "step": 11745
+    },
+    {
+      "epoch": 3.1181468206557814,
+      "grad_norm": 0.46649211530743734,
+      "learning_rate": 2.3611186214985156e-06,
+      "loss": 0.5819,
+      "step": 11746
+    },
+    {
+      "epoch": 3.118412319129165,
+      "grad_norm": 0.4489057380501878,
+      "learning_rate": 2.3607700340719875e-06,
+      "loss": 0.5584,
+      "step": 11747
+    },
+    {
+      "epoch": 3.1186778176025487,
+      "grad_norm": 0.4783414489956911,
+      "learning_rate": 2.3604214493607844e-06,
+      "loss": 0.5485,
+      "step": 11748
+    },
+    {
+      "epoch": 3.1189433160759323,
+      "grad_norm": 0.450927168301951,
+      "learning_rate": 2.3600728673717056e-06,
+      "loss": 0.5643,
+      "step": 11749
+    },
+    {
+      "epoch": 3.1192088145493164,
+      "grad_norm": 0.4522232063980048,
+      "learning_rate": 2.359724288111548e-06,
+      "loss": 0.5298,
+      "step": 11750
+    },
+    {
+      "epoch": 3.1194743130227,
+      "grad_norm": 0.4469001558039488,
+      "learning_rate": 2.35937571158711e-06,
+      "loss": 0.5477,
+      "step": 11751
+    },
+    {
+      "epoch": 3.1197398114960837,
+      "grad_norm": 0.44597692997654864,
+      "learning_rate": 2.3590271378051906e-06,
+      "loss": 0.5575,
+      "step": 11752
+    },
+    {
+      "epoch": 3.120005309969468,
+      "grad_norm": 0.45973556407944144,
+      "learning_rate": 2.3586785667725863e-06,
+      "loss": 0.5439,
+      "step": 11753
+    },
+    {
+      "epoch": 3.1202708084428514,
+      "grad_norm": 0.46937625765053004,
+      "learning_rate": 2.3583299984960963e-06,
+      "loss": 0.511,
+      "step": 11754
+    },
+    {
+      "epoch": 3.120536306916235,
+      "grad_norm": 0.46156198421908556,
+      "learning_rate": 2.3579814329825198e-06,
+      "loss": 0.5314,
+      "step": 11755
+    },
+    {
+      "epoch": 3.120801805389619,
+      "grad_norm": 0.4572328838983525,
+      "learning_rate": 2.3576328702386512e-06,
+      "loss": 0.5738,
+      "step": 11756
+    },
+    {
+      "epoch": 3.121067303863003,
+      "grad_norm": 0.4744917205346979,
+      "learning_rate": 2.35728431027129e-06,
+      "loss": 0.5675,
+      "step": 11757
+    },
+    {
+      "epoch": 3.1213328023363864,
+      "grad_norm": 0.4734380900001336,
+      "learning_rate": 2.356935753087235e-06,
+      "loss": 0.5861,
+      "step": 11758
+    },
+    {
+      "epoch": 3.1215983008097705,
+      "grad_norm": 0.4795260387337211,
+      "learning_rate": 2.3565871986932824e-06,
+      "loss": 0.5706,
+      "step": 11759
+    },
+    {
+      "epoch": 3.121863799283154,
+      "grad_norm": 0.48310742282946234,
+      "learning_rate": 2.356238647096231e-06,
+      "loss": 0.5081,
+      "step": 11760
+    },
+    {
+      "epoch": 3.122129297756538,
+      "grad_norm": 0.43683266659066383,
+      "learning_rate": 2.355890098302879e-06,
+      "loss": 0.5224,
+      "step": 11761
+    },
+    {
+      "epoch": 3.122394796229922,
+      "grad_norm": 0.4588944730846631,
+      "learning_rate": 2.355541552320022e-06,
+      "loss": 0.5437,
+      "step": 11762
+    },
+    {
+      "epoch": 3.1226602947033055,
+      "grad_norm": 0.4788938720972009,
+      "learning_rate": 2.355193009154458e-06,
+      "loss": 0.5722,
+      "step": 11763
+    },
+    {
+      "epoch": 3.122925793176689,
+      "grad_norm": 0.47877834442730755,
+      "learning_rate": 2.354844468812985e-06,
+      "loss": 0.5642,
+      "step": 11764
+    },
+    {
+      "epoch": 3.123191291650073,
+      "grad_norm": 0.45124919597007257,
+      "learning_rate": 2.3544959313024007e-06,
+      "loss": 0.556,
+      "step": 11765
+    },
+    {
+      "epoch": 3.123456790123457,
+      "grad_norm": 0.44213716337862324,
+      "learning_rate": 2.354147396629502e-06,
+      "loss": 0.5342,
+      "step": 11766
+    },
+    {
+      "epoch": 3.1237222885968405,
+      "grad_norm": 0.45120079869185686,
+      "learning_rate": 2.353798864801086e-06,
+      "loss": 0.5411,
+      "step": 11767
+    },
+    {
+      "epoch": 3.123987787070224,
+      "grad_norm": 0.45287146730197203,
+      "learning_rate": 2.35345033582395e-06,
+      "loss": 0.5491,
+      "step": 11768
+    },
+    {
+      "epoch": 3.1242532855436083,
+      "grad_norm": 0.44696996381347803,
+      "learning_rate": 2.353101809704891e-06,
+      "loss": 0.5317,
+      "step": 11769
+    },
+    {
+      "epoch": 3.124518784016992,
+      "grad_norm": 0.45924444509524265,
+      "learning_rate": 2.352753286450707e-06,
+      "loss": 0.5786,
+      "step": 11770
+    },
+    {
+      "epoch": 3.1247842824903755,
+      "grad_norm": 0.4626433083968456,
+      "learning_rate": 2.352404766068194e-06,
+      "loss": 0.5561,
+      "step": 11771
+    },
+    {
+      "epoch": 3.1250497809637596,
+      "grad_norm": 0.45703495688418977,
+      "learning_rate": 2.3520562485641506e-06,
+      "loss": 0.5668,
+      "step": 11772
+    },
+    {
+      "epoch": 3.1253152794371433,
+      "grad_norm": 0.4588134340702698,
+      "learning_rate": 2.3517077339453716e-06,
+      "loss": 0.5526,
+      "step": 11773
+    },
+    {
+      "epoch": 3.125580777910527,
+      "grad_norm": 0.4583810498174728,
+      "learning_rate": 2.3513592222186545e-06,
+      "loss": 0.5599,
+      "step": 11774
+    },
+    {
+      "epoch": 3.125846276383911,
+      "grad_norm": 0.45117006327058934,
+      "learning_rate": 2.3510107133907967e-06,
+      "loss": 0.5376,
+      "step": 11775
+    },
+    {
+      "epoch": 3.1261117748572946,
+      "grad_norm": 0.44717385572374685,
+      "learning_rate": 2.3506622074685947e-06,
+      "loss": 0.5809,
+      "step": 11776
+    },
+    {
+      "epoch": 3.1263772733306783,
+      "grad_norm": 0.45564658491177,
+      "learning_rate": 2.3503137044588457e-06,
+      "loss": 0.5615,
+      "step": 11777
+    },
+    {
+      "epoch": 3.126642771804062,
+      "grad_norm": 0.4578593928636026,
+      "learning_rate": 2.349965204368346e-06,
+      "loss": 0.5679,
+      "step": 11778
+    },
+    {
+      "epoch": 3.126908270277446,
+      "grad_norm": 0.4496146029151605,
+      "learning_rate": 2.349616707203892e-06,
+      "loss": 0.5511,
+      "step": 11779
+    },
+    {
+      "epoch": 3.1271737687508296,
+      "grad_norm": 0.449285539507093,
+      "learning_rate": 2.3492682129722803e-06,
+      "loss": 0.5444,
+      "step": 11780
+    },
+    {
+      "epoch": 3.1274392672242133,
+      "grad_norm": 0.44869662148393685,
+      "learning_rate": 2.348919721680307e-06,
+      "loss": 0.5083,
+      "step": 11781
+    },
+    {
+      "epoch": 3.1277047656975974,
+      "grad_norm": 0.43883607143646797,
+      "learning_rate": 2.34857123333477e-06,
+      "loss": 0.501,
+      "step": 11782
+    },
+    {
+      "epoch": 3.127970264170981,
+      "grad_norm": 0.4514988200857329,
+      "learning_rate": 2.348222747942464e-06,
+      "loss": 0.5332,
+      "step": 11783
+    },
+    {
+      "epoch": 3.1282357626443646,
+      "grad_norm": 0.4532026232902584,
+      "learning_rate": 2.3478742655101867e-06,
+      "loss": 0.5571,
+      "step": 11784
+    },
+    {
+      "epoch": 3.1285012611177487,
+      "grad_norm": 0.46522718198984775,
+      "learning_rate": 2.347525786044733e-06,
+      "loss": 0.5219,
+      "step": 11785
+    },
+    {
+      "epoch": 3.1287667595911324,
+      "grad_norm": 0.4618598216065151,
+      "learning_rate": 2.3471773095529002e-06,
+      "loss": 0.5596,
+      "step": 11786
+    },
+    {
+      "epoch": 3.129032258064516,
+      "grad_norm": 0.4479095073007572,
+      "learning_rate": 2.346828836041484e-06,
+      "loss": 0.5762,
+      "step": 11787
+    },
+    {
+      "epoch": 3.1292977565379,
+      "grad_norm": 0.4710935356662311,
+      "learning_rate": 2.34648036551728e-06,
+      "loss": 0.5192,
+      "step": 11788
+    },
+    {
+      "epoch": 3.1295632550112837,
+      "grad_norm": 0.44054884582373033,
+      "learning_rate": 2.3461318979870863e-06,
+      "loss": 0.5481,
+      "step": 11789
+    },
+    {
+      "epoch": 3.1298287534846674,
+      "grad_norm": 0.48327267587012185,
+      "learning_rate": 2.3457834334576964e-06,
+      "loss": 0.5776,
+      "step": 11790
+    },
+    {
+      "epoch": 3.1300942519580515,
+      "grad_norm": 0.46704537426248677,
+      "learning_rate": 2.345434971935907e-06,
+      "loss": 0.5239,
+      "step": 11791
+    },
+    {
+      "epoch": 3.130359750431435,
+      "grad_norm": 0.4510464443700734,
+      "learning_rate": 2.345086513428514e-06,
+      "loss": 0.5439,
+      "step": 11792
+    },
+    {
+      "epoch": 3.1306252489048187,
+      "grad_norm": 0.4427174169988707,
+      "learning_rate": 2.344738057942313e-06,
+      "loss": 0.5515,
+      "step": 11793
+    },
+    {
+      "epoch": 3.1308907473782024,
+      "grad_norm": 0.45321403456527054,
+      "learning_rate": 2.3443896054841003e-06,
+      "loss": 0.5542,
+      "step": 11794
+    },
+    {
+      "epoch": 3.1311562458515865,
+      "grad_norm": 0.45340279915072434,
+      "learning_rate": 2.344041156060672e-06,
+      "loss": 0.5313,
+      "step": 11795
+    },
+    {
+      "epoch": 3.13142174432497,
+      "grad_norm": 0.4680933198061559,
+      "learning_rate": 2.343692709678822e-06,
+      "loss": 0.5099,
+      "step": 11796
+    },
+    {
+      "epoch": 3.1316872427983538,
+      "grad_norm": 0.4481888361268329,
+      "learning_rate": 2.3433442663453475e-06,
+      "loss": 0.4944,
+      "step": 11797
+    },
+    {
+      "epoch": 3.131952741271738,
+      "grad_norm": 0.445128295739164,
+      "learning_rate": 2.342995826067043e-06,
+      "loss": 0.5205,
+      "step": 11798
+    },
+    {
+      "epoch": 3.1322182397451215,
+      "grad_norm": 0.45366238247168494,
+      "learning_rate": 2.342647388850704e-06,
+      "loss": 0.565,
+      "step": 11799
+    },
+    {
+      "epoch": 3.132483738218505,
+      "grad_norm": 0.45470504689544455,
+      "learning_rate": 2.3422989547031265e-06,
+      "loss": 0.5202,
+      "step": 11800
+    },
+    {
+      "epoch": 3.132749236691889,
+      "grad_norm": 0.4470652866631795,
+      "learning_rate": 2.3419505236311064e-06,
+      "loss": 0.5642,
+      "step": 11801
+    },
+    {
+      "epoch": 3.133014735165273,
+      "grad_norm": 0.4642963466438548,
+      "learning_rate": 2.341602095641437e-06,
+      "loss": 0.5258,
+      "step": 11802
+    },
+    {
+      "epoch": 3.1332802336386565,
+      "grad_norm": 0.4485459058918683,
+      "learning_rate": 2.3412536707409143e-06,
+      "loss": 0.5121,
+      "step": 11803
+    },
+    {
+      "epoch": 3.13354573211204,
+      "grad_norm": 0.44115598330684735,
+      "learning_rate": 2.3409052489363342e-06,
+      "loss": 0.5578,
+      "step": 11804
+    },
+    {
+      "epoch": 3.133811230585424,
+      "grad_norm": 0.4703540551522825,
+      "learning_rate": 2.3405568302344913e-06,
+      "loss": 0.5415,
+      "step": 11805
+    },
+    {
+      "epoch": 3.134076729058808,
+      "grad_norm": 0.4509277986609646,
+      "learning_rate": 2.3402084146421804e-06,
+      "loss": 0.5525,
+      "step": 11806
+    },
+    {
+      "epoch": 3.1343422275321915,
+      "grad_norm": 0.4514718874244348,
+      "learning_rate": 2.339860002166198e-06,
+      "loss": 0.5354,
+      "step": 11807
+    },
+    {
+      "epoch": 3.1346077260055756,
+      "grad_norm": 0.4600614937927662,
+      "learning_rate": 2.339511592813336e-06,
+      "loss": 0.5406,
+      "step": 11808
+    },
+    {
+      "epoch": 3.134873224478959,
+      "grad_norm": 0.45846880663391903,
+      "learning_rate": 2.3391631865903918e-06,
+      "loss": 0.5323,
+      "step": 11809
+    },
+    {
+      "epoch": 3.135138722952343,
+      "grad_norm": 0.46783416479276835,
+      "learning_rate": 2.3388147835041586e-06,
+      "loss": 0.5707,
+      "step": 11810
+    },
+    {
+      "epoch": 3.135404221425727,
+      "grad_norm": 0.45068343291927604,
+      "learning_rate": 2.338466383561432e-06,
+      "loss": 0.5625,
+      "step": 11811
+    },
+    {
+      "epoch": 3.1356697198991106,
+      "grad_norm": 0.4591908076813028,
+      "learning_rate": 2.3381179867690067e-06,
+      "loss": 0.5471,
+      "step": 11812
+    },
+    {
+      "epoch": 3.1359352183724942,
+      "grad_norm": 0.4599059088040292,
+      "learning_rate": 2.3377695931336766e-06,
+      "loss": 0.5612,
+      "step": 11813
+    },
+    {
+      "epoch": 3.1362007168458783,
+      "grad_norm": 0.45690702686973267,
+      "learning_rate": 2.3374212026622365e-06,
+      "loss": 0.5755,
+      "step": 11814
+    },
+    {
+      "epoch": 3.136466215319262,
+      "grad_norm": 0.4621955302218684,
+      "learning_rate": 2.3370728153614813e-06,
+      "loss": 0.5441,
+      "step": 11815
+    },
+    {
+      "epoch": 3.1367317137926456,
+      "grad_norm": 0.45919592802179554,
+      "learning_rate": 2.336724431238205e-06,
+      "loss": 0.5474,
+      "step": 11816
+    },
+    {
+      "epoch": 3.1369972122660297,
+      "grad_norm": 0.45810490096345674,
+      "learning_rate": 2.336376050299202e-06,
+      "loss": 0.5599,
+      "step": 11817
+    },
+    {
+      "epoch": 3.1372627107394133,
+      "grad_norm": 0.44726843871709215,
+      "learning_rate": 2.3360276725512683e-06,
+      "loss": 0.564,
+      "step": 11818
+    },
+    {
+      "epoch": 3.137528209212797,
+      "grad_norm": 0.4566162827304913,
+      "learning_rate": 2.3356792980011947e-06,
+      "loss": 0.5357,
+      "step": 11819
+    },
+    {
+      "epoch": 3.1377937076861806,
+      "grad_norm": 0.4556921610849757,
+      "learning_rate": 2.335330926655777e-06,
+      "loss": 0.5206,
+      "step": 11820
+    },
+    {
+      "epoch": 3.1380592061595647,
+      "grad_norm": 0.46243462376451944,
+      "learning_rate": 2.33498255852181e-06,
+      "loss": 0.5623,
+      "step": 11821
+    },
+    {
+      "epoch": 3.1383247046329483,
+      "grad_norm": 0.4519184831768744,
+      "learning_rate": 2.3346341936060867e-06,
+      "loss": 0.5405,
+      "step": 11822
+    },
+    {
+      "epoch": 3.138590203106332,
+      "grad_norm": 0.4619948634004656,
+      "learning_rate": 2.334285831915401e-06,
+      "loss": 0.5318,
+      "step": 11823
+    },
+    {
+      "epoch": 3.138855701579716,
+      "grad_norm": 0.4575298450110613,
+      "learning_rate": 2.333937473456549e-06,
+      "loss": 0.5668,
+      "step": 11824
+    },
+    {
+      "epoch": 3.1391212000530997,
+      "grad_norm": 0.4496015943208066,
+      "learning_rate": 2.333589118236322e-06,
+      "loss": 0.552,
+      "step": 11825
+    },
+    {
+      "epoch": 3.1393866985264833,
+      "grad_norm": 0.44877558260701894,
+      "learning_rate": 2.3332407662615145e-06,
+      "loss": 0.5255,
+      "step": 11826
+    },
+    {
+      "epoch": 3.1396521969998674,
+      "grad_norm": 0.453504290556328,
+      "learning_rate": 2.33289241753892e-06,
+      "loss": 0.5534,
+      "step": 11827
+    },
+    {
+      "epoch": 3.139917695473251,
+      "grad_norm": 0.4506235391189544,
+      "learning_rate": 2.332544072075333e-06,
+      "loss": 0.5512,
+      "step": 11828
+    },
+    {
+      "epoch": 3.1401831939466347,
+      "grad_norm": 0.45145279435136004,
+      "learning_rate": 2.3321957298775464e-06,
+      "loss": 0.5583,
+      "step": 11829
+    },
+    {
+      "epoch": 3.140448692420019,
+      "grad_norm": 0.452387072937862,
+      "learning_rate": 2.3318473909523543e-06,
+      "loss": 0.5604,
+      "step": 11830
+    },
+    {
+      "epoch": 3.1407141908934024,
+      "grad_norm": 0.4607830426546223,
+      "learning_rate": 2.3314990553065494e-06,
+      "loss": 0.5708,
+      "step": 11831
+    },
+    {
+      "epoch": 3.140979689366786,
+      "grad_norm": 0.4568216243191098,
+      "learning_rate": 2.3311507229469256e-06,
+      "loss": 0.5591,
+      "step": 11832
+    },
+    {
+      "epoch": 3.1412451878401697,
+      "grad_norm": 0.45178472430567024,
+      "learning_rate": 2.330802393880276e-06,
+      "loss": 0.5148,
+      "step": 11833
+    },
+    {
+      "epoch": 3.141510686313554,
+      "grad_norm": 0.44573715501904415,
+      "learning_rate": 2.330454068113394e-06,
+      "loss": 0.5067,
+      "step": 11834
+    },
+    {
+      "epoch": 3.1417761847869374,
+      "grad_norm": 0.4438082651575303,
+      "learning_rate": 2.330105745653074e-06,
+      "loss": 0.5616,
+      "step": 11835
+    },
+    {
+      "epoch": 3.142041683260321,
+      "grad_norm": 0.4572405008364656,
+      "learning_rate": 2.329757426506107e-06,
+      "loss": 0.5788,
+      "step": 11836
+    },
+    {
+      "epoch": 3.142307181733705,
+      "grad_norm": 0.4464227669642761,
+      "learning_rate": 2.3294091106792864e-06,
+      "loss": 0.5119,
+      "step": 11837
+    },
+    {
+      "epoch": 3.142572680207089,
+      "grad_norm": 0.43968926664032654,
+      "learning_rate": 2.329060798179406e-06,
+      "loss": 0.5545,
+      "step": 11838
+    },
+    {
+      "epoch": 3.1428381786804724,
+      "grad_norm": 0.4587224890881845,
+      "learning_rate": 2.328712489013259e-06,
+      "loss": 0.528,
+      "step": 11839
+    },
+    {
+      "epoch": 3.1431036771538565,
+      "grad_norm": 0.463554552097399,
+      "learning_rate": 2.3283641831876375e-06,
+      "loss": 0.5584,
+      "step": 11840
+    },
+    {
+      "epoch": 3.14336917562724,
+      "grad_norm": 0.44545713975558804,
+      "learning_rate": 2.3280158807093355e-06,
+      "loss": 0.5176,
+      "step": 11841
+    },
+    {
+      "epoch": 3.143634674100624,
+      "grad_norm": 0.4662438381352694,
+      "learning_rate": 2.327667581585144e-06,
+      "loss": 0.5348,
+      "step": 11842
+    },
+    {
+      "epoch": 3.143900172574008,
+      "grad_norm": 0.45100909553468216,
+      "learning_rate": 2.3273192858218567e-06,
+      "loss": 0.551,
+      "step": 11843
+    },
+    {
+      "epoch": 3.1441656710473915,
+      "grad_norm": 0.4476004829756491,
+      "learning_rate": 2.3269709934262664e-06,
+      "loss": 0.5576,
+      "step": 11844
+    },
+    {
+      "epoch": 3.144431169520775,
+      "grad_norm": 0.4647501023873188,
+      "learning_rate": 2.3266227044051647e-06,
+      "loss": 0.5554,
+      "step": 11845
+    },
+    {
+      "epoch": 3.1446966679941593,
+      "grad_norm": 0.4546178945372103,
+      "learning_rate": 2.3262744187653453e-06,
+      "loss": 0.5793,
+      "step": 11846
+    },
+    {
+      "epoch": 3.144962166467543,
+      "grad_norm": 0.4634833018000255,
+      "learning_rate": 2.3259261365136014e-06,
+      "loss": 0.554,
+      "step": 11847
+    },
+    {
+      "epoch": 3.1452276649409265,
+      "grad_norm": 0.45095087246713256,
+      "learning_rate": 2.325577857656722e-06,
+      "loss": 0.5694,
+      "step": 11848
+    },
+    {
+      "epoch": 3.14549316341431,
+      "grad_norm": 0.4532292387845422,
+      "learning_rate": 2.325229582201502e-06,
+      "loss": 0.5149,
+      "step": 11849
+    },
+    {
+      "epoch": 3.1457586618876943,
+      "grad_norm": 0.4460049402496013,
+      "learning_rate": 2.324881310154733e-06,
+      "loss": 0.546,
+      "step": 11850
+    },
+    {
+      "epoch": 3.146024160361078,
+      "grad_norm": 0.45422293174408374,
+      "learning_rate": 2.3245330415232074e-06,
+      "loss": 0.5702,
+      "step": 11851
+    },
+    {
+      "epoch": 3.1462896588344615,
+      "grad_norm": 0.43826609261206567,
+      "learning_rate": 2.3241847763137172e-06,
+      "loss": 0.5775,
+      "step": 11852
+    },
+    {
+      "epoch": 3.1465551573078456,
+      "grad_norm": 0.465867842377378,
+      "learning_rate": 2.3238365145330554e-06,
+      "loss": 0.5611,
+      "step": 11853
+    },
+    {
+      "epoch": 3.1468206557812293,
+      "grad_norm": 0.45578392589909705,
+      "learning_rate": 2.3234882561880117e-06,
+      "loss": 0.5068,
+      "step": 11854
+    },
+    {
+      "epoch": 3.147086154254613,
+      "grad_norm": 0.4512131423831718,
+      "learning_rate": 2.323140001285379e-06,
+      "loss": 0.5383,
+      "step": 11855
+    },
+    {
+      "epoch": 3.147351652727997,
+      "grad_norm": 0.4536569472563547,
+      "learning_rate": 2.3227917498319495e-06,
+      "loss": 0.5741,
+      "step": 11856
+    },
+    {
+      "epoch": 3.1476171512013806,
+      "grad_norm": 0.4592691910167352,
+      "learning_rate": 2.3224435018345148e-06,
+      "loss": 0.5286,
+      "step": 11857
+    },
+    {
+      "epoch": 3.1478826496747643,
+      "grad_norm": 0.44847520456716194,
+      "learning_rate": 2.322095257299867e-06,
+      "loss": 0.5921,
+      "step": 11858
+    },
+    {
+      "epoch": 3.148148148148148,
+      "grad_norm": 0.4615011128530479,
+      "learning_rate": 2.3217470162347967e-06,
+      "loss": 0.5438,
+      "step": 11859
+    },
+    {
+      "epoch": 3.148413646621532,
+      "grad_norm": 0.45434002064169093,
+      "learning_rate": 2.3213987786460963e-06,
+      "loss": 0.5749,
+      "step": 11860
+    },
+    {
+      "epoch": 3.1486791450949156,
+      "grad_norm": 0.4574653987830054,
+      "learning_rate": 2.3210505445405564e-06,
+      "loss": 0.5549,
+      "step": 11861
+    },
+    {
+      "epoch": 3.1489446435682993,
+      "grad_norm": 0.458277267074342,
+      "learning_rate": 2.3207023139249696e-06,
+      "loss": 0.5327,
+      "step": 11862
+    },
+    {
+      "epoch": 3.1492101420416834,
+      "grad_norm": 0.4553064141982108,
+      "learning_rate": 2.3203540868061263e-06,
+      "loss": 0.5675,
+      "step": 11863
+    },
+    {
+      "epoch": 3.149475640515067,
+      "grad_norm": 0.46381709315182235,
+      "learning_rate": 2.3200058631908196e-06,
+      "loss": 0.6012,
+      "step": 11864
+    },
+    {
+      "epoch": 3.1497411389884507,
+      "grad_norm": 0.45828934432132007,
+      "learning_rate": 2.319657643085838e-06,
+      "loss": 0.5772,
+      "step": 11865
+    },
+    {
+      "epoch": 3.1500066374618347,
+      "grad_norm": 0.45868509950179476,
+      "learning_rate": 2.3193094264979737e-06,
+      "loss": 0.5499,
+      "step": 11866
+    },
+    {
+      "epoch": 3.1502721359352184,
+      "grad_norm": 0.47561676424653343,
+      "learning_rate": 2.3189612134340184e-06,
+      "loss": 0.5544,
+      "step": 11867
+    },
+    {
+      "epoch": 3.150537634408602,
+      "grad_norm": 0.44259374677351376,
+      "learning_rate": 2.3186130039007628e-06,
+      "loss": 0.5387,
+      "step": 11868
+    },
+    {
+      "epoch": 3.150803132881986,
+      "grad_norm": 0.46992140656105286,
+      "learning_rate": 2.3182647979049967e-06,
+      "loss": 0.5647,
+      "step": 11869
+    },
+    {
+      "epoch": 3.1510686313553697,
+      "grad_norm": 0.4384444378142557,
+      "learning_rate": 2.3179165954535138e-06,
+      "loss": 0.543,
+      "step": 11870
+    },
+    {
+      "epoch": 3.1513341298287534,
+      "grad_norm": 0.447322539832735,
+      "learning_rate": 2.317568396553102e-06,
+      "loss": 0.5354,
+      "step": 11871
+    },
+    {
+      "epoch": 3.1515996283021375,
+      "grad_norm": 0.45522410185966244,
+      "learning_rate": 2.3172202012105533e-06,
+      "loss": 0.553,
+      "step": 11872
+    },
+    {
+      "epoch": 3.151865126775521,
+      "grad_norm": 0.47185336038782544,
+      "learning_rate": 2.316872009432658e-06,
+      "loss": 0.5499,
+      "step": 11873
+    },
+    {
+      "epoch": 3.1521306252489047,
+      "grad_norm": 0.46740893731854743,
+      "learning_rate": 2.316523821226207e-06,
+      "loss": 0.5442,
+      "step": 11874
+    },
+    {
+      "epoch": 3.1523961237222884,
+      "grad_norm": 0.4538978349820548,
+      "learning_rate": 2.3161756365979905e-06,
+      "loss": 0.5106,
+      "step": 11875
+    },
+    {
+      "epoch": 3.1526616221956725,
+      "grad_norm": 0.4383725636056833,
+      "learning_rate": 2.3158274555548e-06,
+      "loss": 0.5369,
+      "step": 11876
+    },
+    {
+      "epoch": 3.152927120669056,
+      "grad_norm": 0.458289464802042,
+      "learning_rate": 2.3154792781034245e-06,
+      "loss": 0.5317,
+      "step": 11877
+    },
+    {
+      "epoch": 3.1531926191424398,
+      "grad_norm": 0.4485180957545061,
+      "learning_rate": 2.3151311042506547e-06,
+      "loss": 0.5455,
+      "step": 11878
+    },
+    {
+      "epoch": 3.153458117615824,
+      "grad_norm": 0.4424529529518882,
+      "learning_rate": 2.314782934003281e-06,
+      "loss": 0.577,
+      "step": 11879
+    },
+    {
+      "epoch": 3.1537236160892075,
+      "grad_norm": 0.48368952294590534,
+      "learning_rate": 2.3144347673680937e-06,
+      "loss": 0.5554,
+      "step": 11880
+    },
+    {
+      "epoch": 3.153989114562591,
+      "grad_norm": 0.4468828892820441,
+      "learning_rate": 2.314086604351883e-06,
+      "loss": 0.5401,
+      "step": 11881
+    },
+    {
+      "epoch": 3.154254613035975,
+      "grad_norm": 0.4534676232194772,
+      "learning_rate": 2.3137384449614392e-06,
+      "loss": 0.5468,
+      "step": 11882
+    },
+    {
+      "epoch": 3.154520111509359,
+      "grad_norm": 0.4617975977696021,
+      "learning_rate": 2.313390289203551e-06,
+      "loss": 0.5633,
+      "step": 11883
+    },
+    {
+      "epoch": 3.1547856099827425,
+      "grad_norm": 0.45961329025233716,
+      "learning_rate": 2.313042137085009e-06,
+      "loss": 0.5466,
+      "step": 11884
+    },
+    {
+      "epoch": 3.1550511084561266,
+      "grad_norm": 0.46140884465785875,
+      "learning_rate": 2.312693988612603e-06,
+      "loss": 0.5235,
+      "step": 11885
+    },
+    {
+      "epoch": 3.15531660692951,
+      "grad_norm": 0.46017972046761896,
+      "learning_rate": 2.3123458437931234e-06,
+      "loss": 0.5257,
+      "step": 11886
+    },
+    {
+      "epoch": 3.155582105402894,
+      "grad_norm": 0.45782551115668935,
+      "learning_rate": 2.311997702633359e-06,
+      "loss": 0.5486,
+      "step": 11887
+    },
+    {
+      "epoch": 3.1558476038762775,
+      "grad_norm": 0.4556464378648574,
+      "learning_rate": 2.3116495651401e-06,
+      "loss": 0.5691,
+      "step": 11888
+    },
+    {
+      "epoch": 3.1561131023496616,
+      "grad_norm": 0.4527791128096492,
+      "learning_rate": 2.3113014313201353e-06,
+      "loss": 0.589,
+      "step": 11889
+    },
+    {
+      "epoch": 3.156378600823045,
+      "grad_norm": 0.4694596305845738,
+      "learning_rate": 2.3109533011802544e-06,
+      "loss": 0.5637,
+      "step": 11890
+    },
+    {
+      "epoch": 3.156644099296429,
+      "grad_norm": 0.44273057471544064,
+      "learning_rate": 2.310605174727247e-06,
+      "loss": 0.5154,
+      "step": 11891
+    },
+    {
+      "epoch": 3.156909597769813,
+      "grad_norm": 0.4619852478624283,
+      "learning_rate": 2.3102570519679027e-06,
+      "loss": 0.5636,
+      "step": 11892
+    },
+    {
+      "epoch": 3.1571750962431966,
+      "grad_norm": 0.4646091603101284,
+      "learning_rate": 2.3099089329090115e-06,
+      "loss": 0.5556,
+      "step": 11893
+    },
+    {
+      "epoch": 3.1574405947165802,
+      "grad_norm": 0.4668777010606396,
+      "learning_rate": 2.30956081755736e-06,
+      "loss": 0.5325,
+      "step": 11894
+    },
+    {
+      "epoch": 3.1577060931899643,
+      "grad_norm": 0.45927350078000767,
+      "learning_rate": 2.3092127059197394e-06,
+      "loss": 0.5736,
+      "step": 11895
+    },
+    {
+      "epoch": 3.157971591663348,
+      "grad_norm": 0.46308138195763815,
+      "learning_rate": 2.3088645980029377e-06,
+      "loss": 0.5999,
+      "step": 11896
+    },
+    {
+      "epoch": 3.1582370901367316,
+      "grad_norm": 0.46950608585504244,
+      "learning_rate": 2.3085164938137446e-06,
+      "loss": 0.5709,
+      "step": 11897
+    },
+    {
+      "epoch": 3.1585025886101157,
+      "grad_norm": 0.45039099700353363,
+      "learning_rate": 2.3081683933589486e-06,
+      "loss": 0.5408,
+      "step": 11898
+    },
+    {
+      "epoch": 3.1587680870834993,
+      "grad_norm": 0.4364745188609442,
+      "learning_rate": 2.30782029664534e-06,
+      "loss": 0.5558,
+      "step": 11899
+    },
+    {
+      "epoch": 3.159033585556883,
+      "grad_norm": 0.45798365668270336,
+      "learning_rate": 2.307472203679705e-06,
+      "loss": 0.5464,
+      "step": 11900
+    },
+    {
+      "epoch": 3.159299084030267,
+      "grad_norm": 0.4462187590629579,
+      "learning_rate": 2.3071241144688335e-06,
+      "loss": 0.5425,
+      "step": 11901
+    },
+    {
+      "epoch": 3.1595645825036507,
+      "grad_norm": 0.4475251677886612,
+      "learning_rate": 2.306776029019514e-06,
+      "loss": 0.5647,
+      "step": 11902
+    },
+    {
+      "epoch": 3.1598300809770343,
+      "grad_norm": 0.44981184194941526,
+      "learning_rate": 2.3064279473385344e-06,
+      "loss": 0.5285,
+      "step": 11903
+    },
+    {
+      "epoch": 3.160095579450418,
+      "grad_norm": 0.4682206800698678,
+      "learning_rate": 2.3060798694326843e-06,
+      "loss": 0.5745,
+      "step": 11904
+    },
+    {
+      "epoch": 3.160361077923802,
+      "grad_norm": 0.44524995744002105,
+      "learning_rate": 2.305731795308752e-06,
+      "loss": 0.5622,
+      "step": 11905
+    },
+    {
+      "epoch": 3.1606265763971857,
+      "grad_norm": 0.4558035053055899,
+      "learning_rate": 2.305383724973525e-06,
+      "loss": 0.5555,
+      "step": 11906
+    },
+    {
+      "epoch": 3.1608920748705693,
+      "grad_norm": 0.4721828951515266,
+      "learning_rate": 2.3050356584337917e-06,
+      "loss": 0.5416,
+      "step": 11907
+    },
+    {
+      "epoch": 3.1611575733439534,
+      "grad_norm": 0.4446206220869704,
+      "learning_rate": 2.30468759569634e-06,
+      "loss": 0.5602,
+      "step": 11908
+    },
+    {
+      "epoch": 3.161423071817337,
+      "grad_norm": 0.46952691845740646,
+      "learning_rate": 2.304339536767959e-06,
+      "loss": 0.5496,
+      "step": 11909
+    },
+    {
+      "epoch": 3.1616885702907207,
+      "grad_norm": 0.46082072633497934,
+      "learning_rate": 2.3039914816554373e-06,
+      "loss": 0.5308,
+      "step": 11910
+    },
+    {
+      "epoch": 3.161954068764105,
+      "grad_norm": 0.45829078660720957,
+      "learning_rate": 2.30364343036556e-06,
+      "loss": 0.5559,
+      "step": 11911
+    },
+    {
+      "epoch": 3.1622195672374884,
+      "grad_norm": 0.46516840904427664,
+      "learning_rate": 2.3032953829051166e-06,
+      "loss": 0.5579,
+      "step": 11912
+    },
+    {
+      "epoch": 3.162485065710872,
+      "grad_norm": 0.4456581336622198,
+      "learning_rate": 2.302947339280895e-06,
+      "loss": 0.5396,
+      "step": 11913
+    },
+    {
+      "epoch": 3.1627505641842557,
+      "grad_norm": 0.4596621465567285,
+      "learning_rate": 2.302599299499683e-06,
+      "loss": 0.5486,
+      "step": 11914
+    },
+    {
+      "epoch": 3.16301606265764,
+      "grad_norm": 0.44672630278292447,
+      "learning_rate": 2.302251263568267e-06,
+      "loss": 0.5362,
+      "step": 11915
+    },
+    {
+      "epoch": 3.1632815611310234,
+      "grad_norm": 0.4548573150208644,
+      "learning_rate": 2.3019032314934377e-06,
+      "loss": 0.5362,
+      "step": 11916
+    },
+    {
+      "epoch": 3.163547059604407,
+      "grad_norm": 0.4757949569485824,
+      "learning_rate": 2.301555203281979e-06,
+      "loss": 0.5432,
+      "step": 11917
+    },
+    {
+      "epoch": 3.163812558077791,
+      "grad_norm": 0.450711984322392,
+      "learning_rate": 2.3012071789406797e-06,
+      "loss": 0.5365,
+      "step": 11918
+    },
+    {
+      "epoch": 3.164078056551175,
+      "grad_norm": 0.4711714557958494,
+      "learning_rate": 2.3008591584763266e-06,
+      "loss": 0.5494,
+      "step": 11919
+    },
+    {
+      "epoch": 3.1643435550245584,
+      "grad_norm": 0.45993654962098,
+      "learning_rate": 2.3005111418957084e-06,
+      "loss": 0.5151,
+      "step": 11920
+    },
+    {
+      "epoch": 3.1646090534979425,
+      "grad_norm": 0.44092050017653484,
+      "learning_rate": 2.300163129205611e-06,
+      "loss": 0.5517,
+      "step": 11921
+    },
+    {
+      "epoch": 3.164874551971326,
+      "grad_norm": 0.4577916406260054,
+      "learning_rate": 2.2998151204128225e-06,
+      "loss": 0.5346,
+      "step": 11922
+    },
+    {
+      "epoch": 3.16514005044471,
+      "grad_norm": 0.4566487614485163,
+      "learning_rate": 2.2994671155241287e-06,
+      "loss": 0.5512,
+      "step": 11923
+    },
+    {
+      "epoch": 3.165405548918094,
+      "grad_norm": 0.4586303785217844,
+      "learning_rate": 2.299119114546317e-06,
+      "loss": 0.5074,
+      "step": 11924
+    },
+    {
+      "epoch": 3.1656710473914775,
+      "grad_norm": 0.4431922571512932,
+      "learning_rate": 2.2987711174861745e-06,
+      "loss": 0.5307,
+      "step": 11925
+    },
+    {
+      "epoch": 3.165936545864861,
+      "grad_norm": 0.4465445116872388,
+      "learning_rate": 2.2984231243504877e-06,
+      "loss": 0.5395,
+      "step": 11926
+    },
+    {
+      "epoch": 3.1662020443382453,
+      "grad_norm": 0.47166576253612597,
+      "learning_rate": 2.298075135146044e-06,
+      "loss": 0.567,
+      "step": 11927
+    },
+    {
+      "epoch": 3.166467542811629,
+      "grad_norm": 0.45857533208259,
+      "learning_rate": 2.2977271498796303e-06,
+      "loss": 0.5617,
+      "step": 11928
+    },
+    {
+      "epoch": 3.1667330412850125,
+      "grad_norm": 0.44057361651831567,
+      "learning_rate": 2.2973791685580313e-06,
+      "loss": 0.5344,
+      "step": 11929
+    },
+    {
+      "epoch": 3.1669985397583966,
+      "grad_norm": 0.46064935425687686,
+      "learning_rate": 2.2970311911880345e-06,
+      "loss": 0.5892,
+      "step": 11930
+    },
+    {
+      "epoch": 3.1672640382317803,
+      "grad_norm": 0.45689291662741055,
+      "learning_rate": 2.2966832177764268e-06,
+      "loss": 0.579,
+      "step": 11931
+    },
+    {
+      "epoch": 3.167529536705164,
+      "grad_norm": 0.45590590724984836,
+      "learning_rate": 2.296335248329994e-06,
+      "loss": 0.5589,
+      "step": 11932
+    },
+    {
+      "epoch": 3.1677950351785475,
+      "grad_norm": 0.4432010792154784,
+      "learning_rate": 2.295987282855523e-06,
+      "loss": 0.5517,
+      "step": 11933
+    },
+    {
+      "epoch": 3.1680605336519316,
+      "grad_norm": 0.4535460304201819,
+      "learning_rate": 2.2956393213597996e-06,
+      "loss": 0.5575,
+      "step": 11934
+    },
+    {
+      "epoch": 3.1683260321253153,
+      "grad_norm": 0.4472429511601056,
+      "learning_rate": 2.2952913638496088e-06,
+      "loss": 0.5666,
+      "step": 11935
+    },
+    {
+      "epoch": 3.168591530598699,
+      "grad_norm": 0.45519272895729573,
+      "learning_rate": 2.294943410331738e-06,
+      "loss": 0.5492,
+      "step": 11936
+    },
+    {
+      "epoch": 3.168857029072083,
+      "grad_norm": 0.4496165437786142,
+      "learning_rate": 2.294595460812973e-06,
+      "loss": 0.5577,
+      "step": 11937
+    },
+    {
+      "epoch": 3.1691225275454666,
+      "grad_norm": 0.4567582927424457,
+      "learning_rate": 2.294247515300099e-06,
+      "loss": 0.5308,
+      "step": 11938
+    },
+    {
+      "epoch": 3.1693880260188503,
+      "grad_norm": 0.4559478049438848,
+      "learning_rate": 2.293899573799904e-06,
+      "loss": 0.5481,
+      "step": 11939
+    },
+    {
+      "epoch": 3.1696535244922344,
+      "grad_norm": 0.4580583642804652,
+      "learning_rate": 2.2935516363191695e-06,
+      "loss": 0.5803,
+      "step": 11940
+    },
+    {
+      "epoch": 3.169919022965618,
+      "grad_norm": 0.4512815847150975,
+      "learning_rate": 2.2932037028646843e-06,
+      "loss": 0.5643,
+      "step": 11941
+    },
+    {
+      "epoch": 3.1701845214390016,
+      "grad_norm": 0.4395173673345484,
+      "learning_rate": 2.2928557734432334e-06,
+      "loss": 0.5134,
+      "step": 11942
+    },
+    {
+      "epoch": 3.1704500199123853,
+      "grad_norm": 0.45927137400986795,
+      "learning_rate": 2.2925078480616023e-06,
+      "loss": 0.5538,
+      "step": 11943
+    },
+    {
+      "epoch": 3.1707155183857694,
+      "grad_norm": 0.4648944791227291,
+      "learning_rate": 2.292159926726576e-06,
+      "loss": 0.5378,
+      "step": 11944
+    },
+    {
+      "epoch": 3.170981016859153,
+      "grad_norm": 0.4488916888498538,
+      "learning_rate": 2.291812009444941e-06,
+      "loss": 0.5304,
+      "step": 11945
+    },
+    {
+      "epoch": 3.1712465153325367,
+      "grad_norm": 0.44352560345299413,
+      "learning_rate": 2.2914640962234805e-06,
+      "loss": 0.5136,
+      "step": 11946
+    },
+    {
+      "epoch": 3.1715120138059207,
+      "grad_norm": 0.4708600371657463,
+      "learning_rate": 2.2911161870689806e-06,
+      "loss": 0.5465,
+      "step": 11947
+    },
+    {
+      "epoch": 3.1717775122793044,
+      "grad_norm": 0.45775028611192936,
+      "learning_rate": 2.2907682819882263e-06,
+      "loss": 0.5737,
+      "step": 11948
+    },
+    {
+      "epoch": 3.172043010752688,
+      "grad_norm": 0.46017534142323596,
+      "learning_rate": 2.2904203809880033e-06,
+      "loss": 0.55,
+      "step": 11949
+    },
+    {
+      "epoch": 3.172308509226072,
+      "grad_norm": 0.448756471154226,
+      "learning_rate": 2.290072484075096e-06,
+      "loss": 0.5707,
+      "step": 11950
+    },
+    {
+      "epoch": 3.1725740076994557,
+      "grad_norm": 0.4634043266146457,
+      "learning_rate": 2.2897245912562895e-06,
+      "loss": 0.5513,
+      "step": 11951
+    },
+    {
+      "epoch": 3.1728395061728394,
+      "grad_norm": 0.44983267361748785,
+      "learning_rate": 2.289376702538368e-06,
+      "loss": 0.5596,
+      "step": 11952
+    },
+    {
+      "epoch": 3.1731050046462235,
+      "grad_norm": 0.4472643215390135,
+      "learning_rate": 2.2890288179281163e-06,
+      "loss": 0.527,
+      "step": 11953
+    },
+    {
+      "epoch": 3.173370503119607,
+      "grad_norm": 0.45415175539460245,
+      "learning_rate": 2.2886809374323193e-06,
+      "loss": 0.5296,
+      "step": 11954
+    },
+    {
+      "epoch": 3.1736360015929908,
+      "grad_norm": 0.45180212643048523,
+      "learning_rate": 2.288333061057761e-06,
+      "loss": 0.5674,
+      "step": 11955
+    },
+    {
+      "epoch": 3.173901500066375,
+      "grad_norm": 0.482653627030646,
+      "learning_rate": 2.287985188811228e-06,
+      "loss": 0.5568,
+      "step": 11956
+    },
+    {
+      "epoch": 3.1741669985397585,
+      "grad_norm": 0.4682150997849175,
+      "learning_rate": 2.2876373206995016e-06,
+      "loss": 0.5556,
+      "step": 11957
+    },
+    {
+      "epoch": 3.174432497013142,
+      "grad_norm": 0.46590560439215517,
+      "learning_rate": 2.2872894567293673e-06,
+      "loss": 0.6059,
+      "step": 11958
+    },
+    {
+      "epoch": 3.1746979954865258,
+      "grad_norm": 0.46556796630195274,
+      "learning_rate": 2.2869415969076095e-06,
+      "loss": 0.5665,
+      "step": 11959
+    },
+    {
+      "epoch": 3.17496349395991,
+      "grad_norm": 0.4424825426775279,
+      "learning_rate": 2.2865937412410117e-06,
+      "loss": 0.5079,
+      "step": 11960
+    },
+    {
+      "epoch": 3.1752289924332935,
+      "grad_norm": 0.45887644809287786,
+      "learning_rate": 2.286245889736359e-06,
+      "loss": 0.5498,
+      "step": 11961
+    },
+    {
+      "epoch": 3.175494490906677,
+      "grad_norm": 0.4409957756866072,
+      "learning_rate": 2.2858980424004356e-06,
+      "loss": 0.5537,
+      "step": 11962
+    },
+    {
+      "epoch": 3.175759989380061,
+      "grad_norm": 0.46018503733655214,
+      "learning_rate": 2.2855501992400238e-06,
+      "loss": 0.5522,
+      "step": 11963
+    },
+    {
+      "epoch": 3.176025487853445,
+      "grad_norm": 0.4478207696342093,
+      "learning_rate": 2.285202360261908e-06,
+      "loss": 0.5561,
+      "step": 11964
+    },
+    {
+      "epoch": 3.1762909863268285,
+      "grad_norm": 0.4509626756394883,
+      "learning_rate": 2.2848545254728723e-06,
+      "loss": 0.5937,
+      "step": 11965
+    },
+    {
+      "epoch": 3.1765564848002126,
+      "grad_norm": 0.4735172319303348,
+      "learning_rate": 2.2845066948797003e-06,
+      "loss": 0.532,
+      "step": 11966
+    },
+    {
+      "epoch": 3.176821983273596,
+      "grad_norm": 0.45510264104461606,
+      "learning_rate": 2.2841588684891746e-06,
+      "loss": 0.5762,
+      "step": 11967
+    },
+    {
+      "epoch": 3.17708748174698,
+      "grad_norm": 0.45940511443012727,
+      "learning_rate": 2.283811046308081e-06,
+      "loss": 0.5882,
+      "step": 11968
+    },
+    {
+      "epoch": 3.1773529802203635,
+      "grad_norm": 0.4593209916670848,
+      "learning_rate": 2.2834632283432e-06,
+      "loss": 0.561,
+      "step": 11969
+    },
+    {
+      "epoch": 3.1776184786937476,
+      "grad_norm": 0.4677320326489966,
+      "learning_rate": 2.2831154146013163e-06,
+      "loss": 0.5745,
+      "step": 11970
+    },
+    {
+      "epoch": 3.1778839771671312,
+      "grad_norm": 0.44025127110192513,
+      "learning_rate": 2.282767605089213e-06,
+      "loss": 0.5404,
+      "step": 11971
+    },
+    {
+      "epoch": 3.178149475640515,
+      "grad_norm": 0.459600201265043,
+      "learning_rate": 2.2824197998136737e-06,
+      "loss": 0.5345,
+      "step": 11972
+    },
+    {
+      "epoch": 3.178414974113899,
+      "grad_norm": 0.4510852089507729,
+      "learning_rate": 2.2820719987814806e-06,
+      "loss": 0.5756,
+      "step": 11973
+    },
+    {
+      "epoch": 3.1786804725872826,
+      "grad_norm": 0.45612439516926667,
+      "learning_rate": 2.2817242019994185e-06,
+      "loss": 0.5756,
+      "step": 11974
+    },
+    {
+      "epoch": 3.1789459710606662,
+      "grad_norm": 0.47673808738057594,
+      "learning_rate": 2.281376409474268e-06,
+      "loss": 0.562,
+      "step": 11975
+    },
+    {
+      "epoch": 3.1792114695340503,
+      "grad_norm": 0.46393072132264174,
+      "learning_rate": 2.281028621212812e-06,
+      "loss": 0.5119,
+      "step": 11976
+    },
+    {
+      "epoch": 3.179476968007434,
+      "grad_norm": 0.474091952795938,
+      "learning_rate": 2.280680837221835e-06,
+      "loss": 0.5173,
+      "step": 11977
+    },
+    {
+      "epoch": 3.1797424664808176,
+      "grad_norm": 0.4626248482084982,
+      "learning_rate": 2.2803330575081177e-06,
+      "loss": 0.5518,
+      "step": 11978
+    },
+    {
+      "epoch": 3.1800079649542017,
+      "grad_norm": 0.46396505374532665,
+      "learning_rate": 2.279985282078445e-06,
+      "loss": 0.4993,
+      "step": 11979
+    },
+    {
+      "epoch": 3.1802734634275853,
+      "grad_norm": 0.4577769150827353,
+      "learning_rate": 2.2796375109395968e-06,
+      "loss": 0.5705,
+      "step": 11980
+    },
+    {
+      "epoch": 3.180538961900969,
+      "grad_norm": 0.4560043970137087,
+      "learning_rate": 2.279289744098357e-06,
+      "loss": 0.5355,
+      "step": 11981
+    },
+    {
+      "epoch": 3.180804460374353,
+      "grad_norm": 0.4401055290037387,
+      "learning_rate": 2.2789419815615075e-06,
+      "loss": 0.5225,
+      "step": 11982
+    },
+    {
+      "epoch": 3.1810699588477367,
+      "grad_norm": 0.4482367031330957,
+      "learning_rate": 2.2785942233358308e-06,
+      "loss": 0.5499,
+      "step": 11983
+    },
+    {
+      "epoch": 3.1813354573211203,
+      "grad_norm": 0.45572598590679947,
+      "learning_rate": 2.2782464694281086e-06,
+      "loss": 0.5251,
+      "step": 11984
+    },
+    {
+      "epoch": 3.1816009557945044,
+      "grad_norm": 0.4667275212883273,
+      "learning_rate": 2.2778987198451237e-06,
+      "loss": 0.5705,
+      "step": 11985
+    },
+    {
+      "epoch": 3.181866454267888,
+      "grad_norm": 0.44958010361997885,
+      "learning_rate": 2.2775509745936573e-06,
+      "loss": 0.5558,
+      "step": 11986
+    },
+    {
+      "epoch": 3.1821319527412717,
+      "grad_norm": 0.46626592902557124,
+      "learning_rate": 2.277203233680491e-06,
+      "loss": 0.5574,
+      "step": 11987
+    },
+    {
+      "epoch": 3.1823974512146553,
+      "grad_norm": 0.4635906252070708,
+      "learning_rate": 2.276855497112408e-06,
+      "loss": 0.5213,
+      "step": 11988
+    },
+    {
+      "epoch": 3.1826629496880394,
+      "grad_norm": 0.45264797469297147,
+      "learning_rate": 2.2765077648961887e-06,
+      "loss": 0.5401,
+      "step": 11989
+    },
+    {
+      "epoch": 3.182928448161423,
+      "grad_norm": 0.4399546623244452,
+      "learning_rate": 2.276160037038615e-06,
+      "loss": 0.5237,
+      "step": 11990
+    },
+    {
+      "epoch": 3.1831939466348067,
+      "grad_norm": 0.4487082587186952,
+      "learning_rate": 2.27581231354647e-06,
+      "loss": 0.5294,
+      "step": 11991
+    },
+    {
+      "epoch": 3.183459445108191,
+      "grad_norm": 0.47512218392202166,
+      "learning_rate": 2.275464594426533e-06,
+      "loss": 0.5897,
+      "step": 11992
+    },
+    {
+      "epoch": 3.1837249435815744,
+      "grad_norm": 0.468212533986608,
+      "learning_rate": 2.2751168796855855e-06,
+      "loss": 0.5088,
+      "step": 11993
+    },
+    {
+      "epoch": 3.183990442054958,
+      "grad_norm": 0.46094414792627186,
+      "learning_rate": 2.27476916933041e-06,
+      "loss": 0.5116,
+      "step": 11994
+    },
+    {
+      "epoch": 3.184255940528342,
+      "grad_norm": 0.44229883939908693,
+      "learning_rate": 2.274421463367786e-06,
+      "loss": 0.547,
+      "step": 11995
+    },
+    {
+      "epoch": 3.184521439001726,
+      "grad_norm": 0.46093167367687804,
+      "learning_rate": 2.274073761804497e-06,
+      "loss": 0.543,
+      "step": 11996
+    },
+    {
+      "epoch": 3.1847869374751094,
+      "grad_norm": 0.4487065111767353,
+      "learning_rate": 2.273726064647323e-06,
+      "loss": 0.5426,
+      "step": 11997
+    },
+    {
+      "epoch": 3.185052435948493,
+      "grad_norm": 0.45691013144107645,
+      "learning_rate": 2.2733783719030444e-06,
+      "loss": 0.5303,
+      "step": 11998
+    },
+    {
+      "epoch": 3.185317934421877,
+      "grad_norm": 0.45419276777520756,
+      "learning_rate": 2.2730306835784423e-06,
+      "loss": 0.5156,
+      "step": 11999
+    },
+    {
+      "epoch": 3.185583432895261,
+      "grad_norm": 0.45450884051682744,
+      "learning_rate": 2.272682999680297e-06,
+      "loss": 0.5085,
+      "step": 12000
+    },
+    {
+      "epoch": 3.1858489313686444,
+      "grad_norm": 0.463686702003172,
+      "learning_rate": 2.2723353202153908e-06,
+      "loss": 0.5728,
+      "step": 12001
+    },
+    {
+      "epoch": 3.1861144298420285,
+      "grad_norm": 0.4612983885435976,
+      "learning_rate": 2.2719876451905028e-06,
+      "loss": 0.5512,
+      "step": 12002
+    },
+    {
+      "epoch": 3.186379928315412,
+      "grad_norm": 0.4450596182726371,
+      "learning_rate": 2.271639974612415e-06,
+      "loss": 0.5655,
+      "step": 12003
+    },
+    {
+      "epoch": 3.186645426788796,
+      "grad_norm": 0.466999257292668,
+      "learning_rate": 2.2712923084879056e-06,
+      "loss": 0.5431,
+      "step": 12004
+    },
+    {
+      "epoch": 3.18691092526218,
+      "grad_norm": 0.4555153611578263,
+      "learning_rate": 2.2709446468237566e-06,
+      "loss": 0.5498,
+      "step": 12005
+    },
+    {
+      "epoch": 3.1871764237355635,
+      "grad_norm": 0.4629705146813975,
+      "learning_rate": 2.2705969896267467e-06,
+      "loss": 0.5468,
+      "step": 12006
+    },
+    {
+      "epoch": 3.187441922208947,
+      "grad_norm": 0.4509704544541768,
+      "learning_rate": 2.270249336903658e-06,
+      "loss": 0.5719,
+      "step": 12007
+    },
+    {
+      "epoch": 3.1877074206823313,
+      "grad_norm": 0.46999624999682965,
+      "learning_rate": 2.269901688661271e-06,
+      "loss": 0.5387,
+      "step": 12008
+    },
+    {
+      "epoch": 3.187972919155715,
+      "grad_norm": 0.44971157740260614,
+      "learning_rate": 2.2695540449063634e-06,
+      "loss": 0.5787,
+      "step": 12009
+    },
+    {
+      "epoch": 3.1882384176290985,
+      "grad_norm": 0.45668394990435407,
+      "learning_rate": 2.2692064056457157e-06,
+      "loss": 0.558,
+      "step": 12010
+    },
+    {
+      "epoch": 3.1885039161024826,
+      "grad_norm": 0.4496143677955816,
+      "learning_rate": 2.2688587708861084e-06,
+      "loss": 0.5545,
+      "step": 12011
+    },
+    {
+      "epoch": 3.1887694145758663,
+      "grad_norm": 0.440637091969325,
+      "learning_rate": 2.268511140634321e-06,
+      "loss": 0.5616,
+      "step": 12012
+    },
+    {
+      "epoch": 3.18903491304925,
+      "grad_norm": 0.4660644754839236,
+      "learning_rate": 2.2681635148971334e-06,
+      "loss": 0.5393,
+      "step": 12013
+    },
+    {
+      "epoch": 3.1893004115226335,
+      "grad_norm": 0.44462407578478047,
+      "learning_rate": 2.267815893681325e-06,
+      "loss": 0.552,
+      "step": 12014
+    },
+    {
+      "epoch": 3.1895659099960176,
+      "grad_norm": 0.46355666669317747,
+      "learning_rate": 2.267468276993675e-06,
+      "loss": 0.5576,
+      "step": 12015
+    },
+    {
+      "epoch": 3.1898314084694013,
+      "grad_norm": 0.4581048296855899,
+      "learning_rate": 2.2671206648409625e-06,
+      "loss": 0.5872,
+      "step": 12016
+    },
+    {
+      "epoch": 3.190096906942785,
+      "grad_norm": 0.46121080803605136,
+      "learning_rate": 2.266773057229968e-06,
+      "loss": 0.5531,
+      "step": 12017
+    },
+    {
+      "epoch": 3.190362405416169,
+      "grad_norm": 0.4608905259243993,
+      "learning_rate": 2.2664254541674692e-06,
+      "loss": 0.5465,
+      "step": 12018
+    },
+    {
+      "epoch": 3.1906279038895526,
+      "grad_norm": 0.4618892315378303,
+      "learning_rate": 2.266077855660246e-06,
+      "loss": 0.5309,
+      "step": 12019
+    },
+    {
+      "epoch": 3.1908934023629363,
+      "grad_norm": 0.4577611693930053,
+      "learning_rate": 2.2657302617150785e-06,
+      "loss": 0.5607,
+      "step": 12020
+    },
+    {
+      "epoch": 3.1911589008363204,
+      "grad_norm": 0.4388856593475079,
+      "learning_rate": 2.265382672338744e-06,
+      "loss": 0.4986,
+      "step": 12021
+    },
+    {
+      "epoch": 3.191424399309704,
+      "grad_norm": 0.4535276687953131,
+      "learning_rate": 2.265035087538021e-06,
+      "loss": 0.5137,
+      "step": 12022
+    },
+    {
+      "epoch": 3.1916898977830876,
+      "grad_norm": 0.43577116403381694,
+      "learning_rate": 2.2646875073196895e-06,
+      "loss": 0.5543,
+      "step": 12023
+    },
+    {
+      "epoch": 3.1919553962564717,
+      "grad_norm": 0.4639696857394536,
+      "learning_rate": 2.2643399316905277e-06,
+      "loss": 0.5468,
+      "step": 12024
+    },
+    {
+      "epoch": 3.1922208947298554,
+      "grad_norm": 0.4434211391947143,
+      "learning_rate": 2.263992360657314e-06,
+      "loss": 0.5343,
+      "step": 12025
+    },
+    {
+      "epoch": 3.192486393203239,
+      "grad_norm": 0.4590767927260959,
+      "learning_rate": 2.263644794226828e-06,
+      "loss": 0.5257,
+      "step": 12026
+    },
+    {
+      "epoch": 3.1927518916766227,
+      "grad_norm": 0.45200571658006,
+      "learning_rate": 2.2632972324058462e-06,
+      "loss": 0.5569,
+      "step": 12027
+    },
+    {
+      "epoch": 3.1930173901500067,
+      "grad_norm": 0.4707935313823779,
+      "learning_rate": 2.2629496752011486e-06,
+      "loss": 0.5768,
+      "step": 12028
+    },
+    {
+      "epoch": 3.1932828886233904,
+      "grad_norm": 0.4614747272850468,
+      "learning_rate": 2.2626021226195123e-06,
+      "loss": 0.5615,
+      "step": 12029
+    },
+    {
+      "epoch": 3.193548387096774,
+      "grad_norm": 0.4548660363541643,
+      "learning_rate": 2.2622545746677157e-06,
+      "loss": 0.5441,
+      "step": 12030
+    },
+    {
+      "epoch": 3.193813885570158,
+      "grad_norm": 0.44517187875612163,
+      "learning_rate": 2.2619070313525377e-06,
+      "loss": 0.4936,
+      "step": 12031
+    },
+    {
+      "epoch": 3.1940793840435417,
+      "grad_norm": 0.44847273092111056,
+      "learning_rate": 2.2615594926807554e-06,
+      "loss": 0.5438,
+      "step": 12032
+    },
+    {
+      "epoch": 3.1943448825169254,
+      "grad_norm": 0.4582845204066536,
+      "learning_rate": 2.2612119586591462e-06,
+      "loss": 0.5158,
+      "step": 12033
+    },
+    {
+      "epoch": 3.1946103809903095,
+      "grad_norm": 0.4691001153907748,
+      "learning_rate": 2.2608644292944886e-06,
+      "loss": 0.5714,
+      "step": 12034
+    },
+    {
+      "epoch": 3.194875879463693,
+      "grad_norm": 0.45743358215071744,
+      "learning_rate": 2.2605169045935606e-06,
+      "loss": 0.5613,
+      "step": 12035
+    },
+    {
+      "epoch": 3.1951413779370768,
+      "grad_norm": 0.4341302331623997,
+      "learning_rate": 2.2601693845631387e-06,
+      "loss": 0.5389,
+      "step": 12036
+    },
+    {
+      "epoch": 3.195406876410461,
+      "grad_norm": 0.4727388039673577,
+      "learning_rate": 2.2598218692100023e-06,
+      "loss": 0.565,
+      "step": 12037
+    },
+    {
+      "epoch": 3.1956723748838445,
+      "grad_norm": 0.4592950393171834,
+      "learning_rate": 2.2594743585409263e-06,
+      "loss": 0.5123,
+      "step": 12038
+    },
+    {
+      "epoch": 3.195937873357228,
+      "grad_norm": 0.4360063807145816,
+      "learning_rate": 2.2591268525626896e-06,
+      "loss": 0.5354,
+      "step": 12039
+    },
+    {
+      "epoch": 3.196203371830612,
+      "grad_norm": 0.44070428224772124,
+      "learning_rate": 2.2587793512820688e-06,
+      "loss": 0.5511,
+      "step": 12040
+    },
+    {
+      "epoch": 3.196468870303996,
+      "grad_norm": 0.4527837138948097,
+      "learning_rate": 2.2584318547058414e-06,
+      "loss": 0.5429,
+      "step": 12041
+    },
+    {
+      "epoch": 3.1967343687773795,
+      "grad_norm": 0.46156954611492307,
+      "learning_rate": 2.2580843628407842e-06,
+      "loss": 0.5968,
+      "step": 12042
+    },
+    {
+      "epoch": 3.196999867250763,
+      "grad_norm": 0.4600964713946351,
+      "learning_rate": 2.2577368756936747e-06,
+      "loss": 0.5519,
+      "step": 12043
+    },
+    {
+      "epoch": 3.197265365724147,
+      "grad_norm": 0.4588229331576669,
+      "learning_rate": 2.257389393271289e-06,
+      "loss": 0.5397,
+      "step": 12044
+    },
+    {
+      "epoch": 3.197530864197531,
+      "grad_norm": 0.4508866314953257,
+      "learning_rate": 2.2570419155804038e-06,
+      "loss": 0.5408,
+      "step": 12045
+    },
+    {
+      "epoch": 3.1977963626709145,
+      "grad_norm": 0.4645064039383363,
+      "learning_rate": 2.256694442627796e-06,
+      "loss": 0.5216,
+      "step": 12046
+    },
+    {
+      "epoch": 3.1980618611442986,
+      "grad_norm": 0.4564802153939025,
+      "learning_rate": 2.256346974420243e-06,
+      "loss": 0.5128,
+      "step": 12047
+    },
+    {
+      "epoch": 3.198327359617682,
+      "grad_norm": 0.44992453751388645,
+      "learning_rate": 2.25599951096452e-06,
+      "loss": 0.5775,
+      "step": 12048
+    },
+    {
+      "epoch": 3.198592858091066,
+      "grad_norm": 0.45272866581134596,
+      "learning_rate": 2.255652052267405e-06,
+      "loss": 0.5654,
+      "step": 12049
+    },
+    {
+      "epoch": 3.19885835656445,
+      "grad_norm": 0.46225153529222474,
+      "learning_rate": 2.2553045983356726e-06,
+      "loss": 0.5701,
+      "step": 12050
+    },
+    {
+      "epoch": 3.1991238550378336,
+      "grad_norm": 0.4535628174109866,
+      "learning_rate": 2.2549571491760985e-06,
+      "loss": 0.5458,
+      "step": 12051
+    },
+    {
+      "epoch": 3.1993893535112172,
+      "grad_norm": 0.4607797671906974,
+      "learning_rate": 2.254609704795461e-06,
+      "loss": 0.5467,
+      "step": 12052
+    },
+    {
+      "epoch": 3.199654851984601,
+      "grad_norm": 0.470816052538922,
+      "learning_rate": 2.2542622652005346e-06,
+      "loss": 0.5469,
+      "step": 12053
+    },
+    {
+      "epoch": 3.199920350457985,
+      "grad_norm": 0.45565485148084167,
+      "learning_rate": 2.2539148303980974e-06,
+      "loss": 0.5171,
+      "step": 12054
+    },
+    {
+      "epoch": 3.2001858489313686,
+      "grad_norm": 0.4479978505775746,
+      "learning_rate": 2.253567400394922e-06,
+      "loss": 0.5481,
+      "step": 12055
+    },
+    {
+      "epoch": 3.2004513474047522,
+      "grad_norm": 0.44181825140960695,
+      "learning_rate": 2.2532199751977856e-06,
+      "loss": 0.5608,
+      "step": 12056
+    },
+    {
+      "epoch": 3.2007168458781363,
+      "grad_norm": 0.4479570571438855,
+      "learning_rate": 2.252872554813464e-06,
+      "loss": 0.5152,
+      "step": 12057
+    },
+    {
+      "epoch": 3.20098234435152,
+      "grad_norm": 0.47002670294432924,
+      "learning_rate": 2.2525251392487328e-06,
+      "loss": 0.5261,
+      "step": 12058
+    },
+    {
+      "epoch": 3.2012478428249036,
+      "grad_norm": 0.45402092226230917,
+      "learning_rate": 2.2521777285103673e-06,
+      "loss": 0.5485,
+      "step": 12059
+    },
+    {
+      "epoch": 3.2015133412982877,
+      "grad_norm": 0.4690168015304434,
+      "learning_rate": 2.2518303226051435e-06,
+      "loss": 0.5532,
+      "step": 12060
+    },
+    {
+      "epoch": 3.2017788397716713,
+      "grad_norm": 0.4498716109355263,
+      "learning_rate": 2.2514829215398354e-06,
+      "loss": 0.5327,
+      "step": 12061
+    },
+    {
+      "epoch": 3.202044338245055,
+      "grad_norm": 0.45858012458963276,
+      "learning_rate": 2.2511355253212186e-06,
+      "loss": 0.5606,
+      "step": 12062
+    },
+    {
+      "epoch": 3.202309836718439,
+      "grad_norm": 0.4548005249086935,
+      "learning_rate": 2.2507881339560687e-06,
+      "loss": 0.5403,
+      "step": 12063
+    },
+    {
+      "epoch": 3.2025753351918227,
+      "grad_norm": 0.49008197469923637,
+      "learning_rate": 2.25044074745116e-06,
+      "loss": 0.5416,
+      "step": 12064
+    },
+    {
+      "epoch": 3.2028408336652063,
+      "grad_norm": 0.4551498377877728,
+      "learning_rate": 2.2500933658132683e-06,
+      "loss": 0.5738,
+      "step": 12065
+    },
+    {
+      "epoch": 3.2031063321385904,
+      "grad_norm": 0.4618772963686445,
+      "learning_rate": 2.2497459890491682e-06,
+      "loss": 0.5457,
+      "step": 12066
+    },
+    {
+      "epoch": 3.203371830611974,
+      "grad_norm": 0.45878121212282486,
+      "learning_rate": 2.2493986171656335e-06,
+      "loss": 0.5835,
+      "step": 12067
+    },
+    {
+      "epoch": 3.2036373290853577,
+      "grad_norm": 0.4752351285463351,
+      "learning_rate": 2.2490512501694394e-06,
+      "loss": 0.5295,
+      "step": 12068
+    },
+    {
+      "epoch": 3.2039028275587413,
+      "grad_norm": 0.44611285960299535,
+      "learning_rate": 2.2487038880673597e-06,
+      "loss": 0.5754,
+      "step": 12069
+    },
+    {
+      "epoch": 3.2041683260321254,
+      "grad_norm": 0.4709357865274885,
+      "learning_rate": 2.24835653086617e-06,
+      "loss": 0.5422,
+      "step": 12070
+    },
+    {
+      "epoch": 3.204433824505509,
+      "grad_norm": 0.4610817227608981,
+      "learning_rate": 2.248009178572644e-06,
+      "loss": 0.5233,
+      "step": 12071
+    },
+    {
+      "epoch": 3.2046993229788927,
+      "grad_norm": 0.4528930371804233,
+      "learning_rate": 2.2476618311935566e-06,
+      "loss": 0.508,
+      "step": 12072
+    },
+    {
+      "epoch": 3.204964821452277,
+      "grad_norm": 0.4481102857538323,
+      "learning_rate": 2.247314488735681e-06,
+      "loss": 0.5432,
+      "step": 12073
+    },
+    {
+      "epoch": 3.2052303199256604,
+      "grad_norm": 0.4671262988786407,
+      "learning_rate": 2.2469671512057907e-06,
+      "loss": 0.5736,
+      "step": 12074
+    },
+    {
+      "epoch": 3.205495818399044,
+      "grad_norm": 0.45117701472613003,
+      "learning_rate": 2.246619818610661e-06,
+      "loss": 0.5501,
+      "step": 12075
+    },
+    {
+      "epoch": 3.205761316872428,
+      "grad_norm": 0.4890338961575178,
+      "learning_rate": 2.2462724909570654e-06,
+      "loss": 0.4999,
+      "step": 12076
+    },
+    {
+      "epoch": 3.206026815345812,
+      "grad_norm": 0.4547605036692314,
+      "learning_rate": 2.2459251682517778e-06,
+      "loss": 0.5612,
+      "step": 12077
+    },
+    {
+      "epoch": 3.2062923138191954,
+      "grad_norm": 0.44653337837380486,
+      "learning_rate": 2.245577850501571e-06,
+      "loss": 0.5576,
+      "step": 12078
+    },
+    {
+      "epoch": 3.2065578122925795,
+      "grad_norm": 0.4652059954447466,
+      "learning_rate": 2.245230537713219e-06,
+      "loss": 0.5774,
+      "step": 12079
+    },
+    {
+      "epoch": 3.206823310765963,
+      "grad_norm": 0.4501068914291192,
+      "learning_rate": 2.244883229893495e-06,
+      "loss": 0.5703,
+      "step": 12080
+    },
+    {
+      "epoch": 3.207088809239347,
+      "grad_norm": 0.46878727246919116,
+      "learning_rate": 2.2445359270491735e-06,
+      "loss": 0.5747,
+      "step": 12081
+    },
+    {
+      "epoch": 3.2073543077127304,
+      "grad_norm": 0.4729406387605314,
+      "learning_rate": 2.2441886291870263e-06,
+      "loss": 0.5677,
+      "step": 12082
+    },
+    {
+      "epoch": 3.2076198061861145,
+      "grad_norm": 0.45966063067065105,
+      "learning_rate": 2.2438413363138283e-06,
+      "loss": 0.5436,
+      "step": 12083
+    },
+    {
+      "epoch": 3.207885304659498,
+      "grad_norm": 0.45546426349227226,
+      "learning_rate": 2.2434940484363505e-06,
+      "loss": 0.5318,
+      "step": 12084
+    },
+    {
+      "epoch": 3.208150803132882,
+      "grad_norm": 0.4549677523565809,
+      "learning_rate": 2.2431467655613667e-06,
+      "loss": 0.5372,
+      "step": 12085
+    },
+    {
+      "epoch": 3.208416301606266,
+      "grad_norm": 0.45277210155815134,
+      "learning_rate": 2.24279948769565e-06,
+      "loss": 0.5494,
+      "step": 12086
+    },
+    {
+      "epoch": 3.2086818000796495,
+      "grad_norm": 0.4627747034900131,
+      "learning_rate": 2.2424522148459727e-06,
+      "loss": 0.5635,
+      "step": 12087
+    },
+    {
+      "epoch": 3.208947298553033,
+      "grad_norm": 0.4510628916255789,
+      "learning_rate": 2.2421049470191077e-06,
+      "loss": 0.557,
+      "step": 12088
+    },
+    {
+      "epoch": 3.2092127970264173,
+      "grad_norm": 0.4533828423471941,
+      "learning_rate": 2.2417576842218287e-06,
+      "loss": 0.5344,
+      "step": 12089
+    },
+    {
+      "epoch": 3.209478295499801,
+      "grad_norm": 0.4669724652502168,
+      "learning_rate": 2.2414104264609064e-06,
+      "loss": 0.5167,
+      "step": 12090
+    },
+    {
+      "epoch": 3.2097437939731845,
+      "grad_norm": 0.45594452700812144,
+      "learning_rate": 2.2410631737431136e-06,
+      "loss": 0.5792,
+      "step": 12091
+    },
+    {
+      "epoch": 3.2100092924465686,
+      "grad_norm": 0.46395962574191985,
+      "learning_rate": 2.2407159260752233e-06,
+      "loss": 0.554,
+      "step": 12092
+    },
+    {
+      "epoch": 3.2102747909199523,
+      "grad_norm": 0.46931571697257185,
+      "learning_rate": 2.2403686834640074e-06,
+      "loss": 0.5514,
+      "step": 12093
+    },
+    {
+      "epoch": 3.210540289393336,
+      "grad_norm": 0.4848476102512686,
+      "learning_rate": 2.2400214459162376e-06,
+      "loss": 0.5647,
+      "step": 12094
+    },
+    {
+      "epoch": 3.21080578786672,
+      "grad_norm": 0.46571891452389336,
+      "learning_rate": 2.239674213438687e-06,
+      "loss": 0.5742,
+      "step": 12095
+    },
+    {
+      "epoch": 3.2110712863401036,
+      "grad_norm": 0.4656596644272797,
+      "learning_rate": 2.239326986038126e-06,
+      "loss": 0.5357,
+      "step": 12096
+    },
+    {
+      "epoch": 3.2113367848134873,
+      "grad_norm": 0.4763818268153712,
+      "learning_rate": 2.238979763721326e-06,
+      "loss": 0.5257,
+      "step": 12097
+    },
+    {
+      "epoch": 3.211602283286871,
+      "grad_norm": 0.46889928409729764,
+      "learning_rate": 2.2386325464950604e-06,
+      "loss": 0.5586,
+      "step": 12098
+    },
+    {
+      "epoch": 3.211867781760255,
+      "grad_norm": 0.446437680670925,
+      "learning_rate": 2.2382853343661005e-06,
+      "loss": 0.5754,
+      "step": 12099
+    },
+    {
+      "epoch": 3.2121332802336386,
+      "grad_norm": 0.4754970797630485,
+      "learning_rate": 2.237938127341218e-06,
+      "loss": 0.5612,
+      "step": 12100
+    },
+    {
+      "epoch": 3.2123987787070223,
+      "grad_norm": 0.4677208564964637,
+      "learning_rate": 2.237590925427183e-06,
+      "loss": 0.5391,
+      "step": 12101
+    },
+    {
+      "epoch": 3.2126642771804064,
+      "grad_norm": 0.4695931236951604,
+      "learning_rate": 2.237243728630767e-06,
+      "loss": 0.5193,
+      "step": 12102
+    },
+    {
+      "epoch": 3.21292977565379,
+      "grad_norm": 0.4497338233579656,
+      "learning_rate": 2.2368965369587416e-06,
+      "loss": 0.5175,
+      "step": 12103
+    },
+    {
+      "epoch": 3.2131952741271737,
+      "grad_norm": 0.451723849156023,
+      "learning_rate": 2.2365493504178784e-06,
+      "loss": 0.5267,
+      "step": 12104
+    },
+    {
+      "epoch": 3.2134607726005577,
+      "grad_norm": 0.45168296666022734,
+      "learning_rate": 2.2362021690149476e-06,
+      "loss": 0.52,
+      "step": 12105
+    },
+    {
+      "epoch": 3.2137262710739414,
+      "grad_norm": 0.45246909625047804,
+      "learning_rate": 2.235854992756721e-06,
+      "loss": 0.559,
+      "step": 12106
+    },
+    {
+      "epoch": 3.213991769547325,
+      "grad_norm": 0.4573653635617974,
+      "learning_rate": 2.235507821649968e-06,
+      "loss": 0.5096,
+      "step": 12107
+    },
+    {
+      "epoch": 3.2142572680207087,
+      "grad_norm": 0.45368240343468075,
+      "learning_rate": 2.23516065570146e-06,
+      "loss": 0.5518,
+      "step": 12108
+    },
+    {
+      "epoch": 3.2145227664940927,
+      "grad_norm": 0.46948971125596406,
+      "learning_rate": 2.234813494917968e-06,
+      "loss": 0.5315,
+      "step": 12109
+    },
+    {
+      "epoch": 3.2147882649674764,
+      "grad_norm": 0.4585606700071447,
+      "learning_rate": 2.234466339306262e-06,
+      "loss": 0.4933,
+      "step": 12110
+    },
+    {
+      "epoch": 3.21505376344086,
+      "grad_norm": 0.4703784538853916,
+      "learning_rate": 2.2341191888731127e-06,
+      "loss": 0.5552,
+      "step": 12111
+    },
+    {
+      "epoch": 3.215319261914244,
+      "grad_norm": 0.44464686912513046,
+      "learning_rate": 2.233772043625291e-06,
+      "loss": 0.5875,
+      "step": 12112
+    },
+    {
+      "epoch": 3.2155847603876277,
+      "grad_norm": 0.46274246587299905,
+      "learning_rate": 2.2334249035695656e-06,
+      "loss": 0.5572,
+      "step": 12113
+    },
+    {
+      "epoch": 3.2158502588610114,
+      "grad_norm": 0.44512400348257325,
+      "learning_rate": 2.233077768712707e-06,
+      "loss": 0.5539,
+      "step": 12114
+    },
+    {
+      "epoch": 3.2161157573343955,
+      "grad_norm": 0.44128150992307696,
+      "learning_rate": 2.2327306390614854e-06,
+      "loss": 0.5681,
+      "step": 12115
+    },
+    {
+      "epoch": 3.216381255807779,
+      "grad_norm": 0.4529521159573287,
+      "learning_rate": 2.2323835146226706e-06,
+      "loss": 0.5112,
+      "step": 12116
+    },
+    {
+      "epoch": 3.2166467542811628,
+      "grad_norm": 0.47760805388944544,
+      "learning_rate": 2.2320363954030326e-06,
+      "loss": 0.5559,
+      "step": 12117
+    },
+    {
+      "epoch": 3.216912252754547,
+      "grad_norm": 0.46417438150451695,
+      "learning_rate": 2.231689281409342e-06,
+      "loss": 0.5112,
+      "step": 12118
+    },
+    {
+      "epoch": 3.2171777512279305,
+      "grad_norm": 0.4699795409075221,
+      "learning_rate": 2.2313421726483663e-06,
+      "loss": 0.5391,
+      "step": 12119
+    },
+    {
+      "epoch": 3.217443249701314,
+      "grad_norm": 0.45310331280150884,
+      "learning_rate": 2.2309950691268763e-06,
+      "loss": 0.5514,
+      "step": 12120
+    },
+    {
+      "epoch": 3.217708748174698,
+      "grad_norm": 0.46829186434105485,
+      "learning_rate": 2.230647970851641e-06,
+      "loss": 0.508,
+      "step": 12121
+    },
+    {
+      "epoch": 3.217974246648082,
+      "grad_norm": 0.4718360330198163,
+      "learning_rate": 2.2303008778294292e-06,
+      "loss": 0.5316,
+      "step": 12122
+    },
+    {
+      "epoch": 3.2182397451214655,
+      "grad_norm": 0.4647252295897281,
+      "learning_rate": 2.2299537900670114e-06,
+      "loss": 0.5433,
+      "step": 12123
+    },
+    {
+      "epoch": 3.218505243594849,
+      "grad_norm": 0.4578173347754867,
+      "learning_rate": 2.229606707571156e-06,
+      "loss": 0.5778,
+      "step": 12124
+    },
+    {
+      "epoch": 3.218770742068233,
+      "grad_norm": 0.4511035043832162,
+      "learning_rate": 2.2292596303486314e-06,
+      "loss": 0.5274,
+      "step": 12125
+    },
+    {
+      "epoch": 3.219036240541617,
+      "grad_norm": 0.47448842514494627,
+      "learning_rate": 2.2289125584062073e-06,
+      "loss": 0.53,
+      "step": 12126
+    },
+    {
+      "epoch": 3.2193017390150005,
+      "grad_norm": 0.4639869128370032,
+      "learning_rate": 2.228565491750651e-06,
+      "loss": 0.5299,
+      "step": 12127
+    },
+    {
+      "epoch": 3.2195672374883846,
+      "grad_norm": 0.4584662461537314,
+      "learning_rate": 2.2282184303887333e-06,
+      "loss": 0.5007,
+      "step": 12128
+    },
+    {
+      "epoch": 3.219832735961768,
+      "grad_norm": 0.4663110405042132,
+      "learning_rate": 2.2278713743272224e-06,
+      "loss": 0.564,
+      "step": 12129
+    },
+    {
+      "epoch": 3.220098234435152,
+      "grad_norm": 0.4673406841538499,
+      "learning_rate": 2.2275243235728845e-06,
+      "loss": 0.5731,
+      "step": 12130
+    },
+    {
+      "epoch": 3.220363732908536,
+      "grad_norm": 0.47030636110289986,
+      "learning_rate": 2.22717727813249e-06,
+      "loss": 0.5801,
+      "step": 12131
+    },
+    {
+      "epoch": 3.2206292313819196,
+      "grad_norm": 0.46900073011499105,
+      "learning_rate": 2.226830238012806e-06,
+      "loss": 0.5217,
+      "step": 12132
+    },
+    {
+      "epoch": 3.2208947298553032,
+      "grad_norm": 0.4674387077100141,
+      "learning_rate": 2.2264832032206016e-06,
+      "loss": 0.5449,
+      "step": 12133
+    },
+    {
+      "epoch": 3.2211602283286873,
+      "grad_norm": 0.46153847590659663,
+      "learning_rate": 2.2261361737626443e-06,
+      "loss": 0.5246,
+      "step": 12134
+    },
+    {
+      "epoch": 3.221425726802071,
+      "grad_norm": 0.4336889807400292,
+      "learning_rate": 2.225789149645703e-06,
+      "loss": 0.4993,
+      "step": 12135
+    },
+    {
+      "epoch": 3.2216912252754546,
+      "grad_norm": 0.4697712684557118,
+      "learning_rate": 2.2254421308765436e-06,
+      "loss": 0.552,
+      "step": 12136
+    },
+    {
+      "epoch": 3.2219567237488382,
+      "grad_norm": 0.440115187230645,
+      "learning_rate": 2.225095117461935e-06,
+      "loss": 0.5416,
+      "step": 12137
+    },
+    {
+      "epoch": 3.2222222222222223,
+      "grad_norm": 0.4806838936062625,
+      "learning_rate": 2.2247481094086447e-06,
+      "loss": 0.5607,
+      "step": 12138
+    },
+    {
+      "epoch": 3.222487720695606,
+      "grad_norm": 0.4579858849906208,
+      "learning_rate": 2.2244011067234405e-06,
+      "loss": 0.577,
+      "step": 12139
+    },
+    {
+      "epoch": 3.2227532191689896,
+      "grad_norm": 0.4432900058184422,
+      "learning_rate": 2.224054109413089e-06,
+      "loss": 0.5323,
+      "step": 12140
+    },
+    {
+      "epoch": 3.2230187176423737,
+      "grad_norm": 0.45549097321417753,
+      "learning_rate": 2.2237071174843596e-06,
+      "loss": 0.5833,
+      "step": 12141
+    },
+    {
+      "epoch": 3.2232842161157573,
+      "grad_norm": 0.45396196674628586,
+      "learning_rate": 2.2233601309440163e-06,
+      "loss": 0.5568,
+      "step": 12142
+    },
+    {
+      "epoch": 3.223549714589141,
+      "grad_norm": 0.4614920207590504,
+      "learning_rate": 2.2230131497988278e-06,
+      "loss": 0.5603,
+      "step": 12143
+    },
+    {
+      "epoch": 3.223815213062525,
+      "grad_norm": 0.44942017586443134,
+      "learning_rate": 2.2226661740555614e-06,
+      "loss": 0.539,
+      "step": 12144
+    },
+    {
+      "epoch": 3.2240807115359087,
+      "grad_norm": 0.44521133851088185,
+      "learning_rate": 2.2223192037209836e-06,
+      "loss": 0.522,
+      "step": 12145
+    },
+    {
+      "epoch": 3.2243462100092923,
+      "grad_norm": 0.4521787414758281,
+      "learning_rate": 2.221972238801861e-06,
+      "loss": 0.5386,
+      "step": 12146
+    },
+    {
+      "epoch": 3.2246117084826764,
+      "grad_norm": 0.4477524357178312,
+      "learning_rate": 2.221625279304962e-06,
+      "loss": 0.5244,
+      "step": 12147
+    },
+    {
+      "epoch": 3.22487720695606,
+      "grad_norm": 0.4592824521022462,
+      "learning_rate": 2.2212783252370496e-06,
+      "loss": 0.5676,
+      "step": 12148
+    },
+    {
+      "epoch": 3.2251427054294437,
+      "grad_norm": 0.4417146126638557,
+      "learning_rate": 2.2209313766048934e-06,
+      "loss": 0.5403,
+      "step": 12149
+    },
+    {
+      "epoch": 3.225408203902828,
+      "grad_norm": 0.4366688082949209,
+      "learning_rate": 2.220584433415258e-06,
+      "loss": 0.5253,
+      "step": 12150
+    },
+    {
+      "epoch": 3.2256737023762114,
+      "grad_norm": 0.4560865805538416,
+      "learning_rate": 2.2202374956749105e-06,
+      "loss": 0.5396,
+      "step": 12151
+    },
+    {
+      "epoch": 3.225939200849595,
+      "grad_norm": 0.45554731995836184,
+      "learning_rate": 2.219890563390617e-06,
+      "loss": 0.5596,
+      "step": 12152
+    },
+    {
+      "epoch": 3.2262046993229787,
+      "grad_norm": 0.45145979357005994,
+      "learning_rate": 2.2195436365691434e-06,
+      "loss": 0.5193,
+      "step": 12153
+    },
+    {
+      "epoch": 3.226470197796363,
+      "grad_norm": 0.4580316692631737,
+      "learning_rate": 2.2191967152172554e-06,
+      "loss": 0.5433,
+      "step": 12154
+    },
+    {
+      "epoch": 3.2267356962697464,
+      "grad_norm": 0.4748272990049663,
+      "learning_rate": 2.2188497993417187e-06,
+      "loss": 0.5716,
+      "step": 12155
+    },
+    {
+      "epoch": 3.22700119474313,
+      "grad_norm": 0.46202573047756074,
+      "learning_rate": 2.2185028889493e-06,
+      "loss": 0.5847,
+      "step": 12156
+    },
+    {
+      "epoch": 3.227266693216514,
+      "grad_norm": 0.44798805376898804,
+      "learning_rate": 2.218155984046763e-06,
+      "loss": 0.538,
+      "step": 12157
+    },
+    {
+      "epoch": 3.227532191689898,
+      "grad_norm": 0.4660034709747994,
+      "learning_rate": 2.2178090846408762e-06,
+      "loss": 0.587,
+      "step": 12158
+    },
+    {
+      "epoch": 3.2277976901632814,
+      "grad_norm": 0.48147163212559496,
+      "learning_rate": 2.217462190738402e-06,
+      "loss": 0.5597,
+      "step": 12159
+    },
+    {
+      "epoch": 3.2280631886366655,
+      "grad_norm": 0.4586045285283479,
+      "learning_rate": 2.2171153023461067e-06,
+      "loss": 0.5515,
+      "step": 12160
+    },
+    {
+      "epoch": 3.228328687110049,
+      "grad_norm": 0.46502968493999525,
+      "learning_rate": 2.216768419470756e-06,
+      "loss": 0.5858,
+      "step": 12161
+    },
+    {
+      "epoch": 3.228594185583433,
+      "grad_norm": 0.4591106631592454,
+      "learning_rate": 2.2164215421191145e-06,
+      "loss": 0.5454,
+      "step": 12162
+    },
+    {
+      "epoch": 3.2288596840568164,
+      "grad_norm": 0.4958019888662801,
+      "learning_rate": 2.2160746702979465e-06,
+      "loss": 0.5402,
+      "step": 12163
+    },
+    {
+      "epoch": 3.2291251825302005,
+      "grad_norm": 0.4849185985974037,
+      "learning_rate": 2.215727804014019e-06,
+      "loss": 0.5369,
+      "step": 12164
+    },
+    {
+      "epoch": 3.229390681003584,
+      "grad_norm": 0.47333501783178134,
+      "learning_rate": 2.2153809432740948e-06,
+      "loss": 0.5616,
+      "step": 12165
+    },
+    {
+      "epoch": 3.229656179476968,
+      "grad_norm": 0.4592387052920761,
+      "learning_rate": 2.2150340880849387e-06,
+      "loss": 0.5512,
+      "step": 12166
+    },
+    {
+      "epoch": 3.229921677950352,
+      "grad_norm": 0.467101843241378,
+      "learning_rate": 2.2146872384533153e-06,
+      "loss": 0.4867,
+      "step": 12167
+    },
+    {
+      "epoch": 3.2301871764237355,
+      "grad_norm": 0.4529533947825373,
+      "learning_rate": 2.2143403943859903e-06,
+      "loss": 0.5482,
+      "step": 12168
+    },
+    {
+      "epoch": 3.230452674897119,
+      "grad_norm": 0.47249892257166426,
+      "learning_rate": 2.2139935558897263e-06,
+      "loss": 0.5489,
+      "step": 12169
+    },
+    {
+      "epoch": 3.2307181733705033,
+      "grad_norm": 0.44970666156245037,
+      "learning_rate": 2.2136467229712886e-06,
+      "loss": 0.5353,
+      "step": 12170
+    },
+    {
+      "epoch": 3.230983671843887,
+      "grad_norm": 0.46488357509370176,
+      "learning_rate": 2.213299895637441e-06,
+      "loss": 0.5431,
+      "step": 12171
+    },
+    {
+      "epoch": 3.2312491703172705,
+      "grad_norm": 0.4534898071054505,
+      "learning_rate": 2.2129530738949472e-06,
+      "loss": 0.5546,
+      "step": 12172
+    },
+    {
+      "epoch": 3.2315146687906546,
+      "grad_norm": 0.44740152800151645,
+      "learning_rate": 2.212606257750571e-06,
+      "loss": 0.5242,
+      "step": 12173
+    },
+    {
+      "epoch": 3.2317801672640383,
+      "grad_norm": 0.4649213048231864,
+      "learning_rate": 2.2122594472110763e-06,
+      "loss": 0.5471,
+      "step": 12174
+    },
+    {
+      "epoch": 3.232045665737422,
+      "grad_norm": 0.4820295704321057,
+      "learning_rate": 2.2119126422832286e-06,
+      "loss": 0.5454,
+      "step": 12175
+    },
+    {
+      "epoch": 3.232311164210806,
+      "grad_norm": 0.46720009037869287,
+      "learning_rate": 2.211565842973788e-06,
+      "loss": 0.5599,
+      "step": 12176
+    },
+    {
+      "epoch": 3.2325766626841896,
+      "grad_norm": 0.45121648978633006,
+      "learning_rate": 2.21121904928952e-06,
+      "loss": 0.558,
+      "step": 12177
+    },
+    {
+      "epoch": 3.2328421611575733,
+      "grad_norm": 0.46940860421368497,
+      "learning_rate": 2.210872261237188e-06,
+      "loss": 0.5634,
+      "step": 12178
+    },
+    {
+      "epoch": 3.233107659630957,
+      "grad_norm": 0.44614122509916604,
+      "learning_rate": 2.210525478823554e-06,
+      "loss": 0.5404,
+      "step": 12179
+    },
+    {
+      "epoch": 3.233373158104341,
+      "grad_norm": 0.4528417382414176,
+      "learning_rate": 2.2101787020553825e-06,
+      "loss": 0.5412,
+      "step": 12180
+    },
+    {
+      "epoch": 3.2336386565777246,
+      "grad_norm": 0.45937324115672057,
+      "learning_rate": 2.2098319309394366e-06,
+      "loss": 0.5466,
+      "step": 12181
+    },
+    {
+      "epoch": 3.2339041550511083,
+      "grad_norm": 0.4569451644713486,
+      "learning_rate": 2.2094851654824772e-06,
+      "loss": 0.5546,
+      "step": 12182
+    },
+    {
+      "epoch": 3.2341696535244924,
+      "grad_norm": 0.4481938655920306,
+      "learning_rate": 2.209138405691269e-06,
+      "loss": 0.5378,
+      "step": 12183
+    },
+    {
+      "epoch": 3.234435151997876,
+      "grad_norm": 0.4512656372869046,
+      "learning_rate": 2.2087916515725738e-06,
+      "loss": 0.557,
+      "step": 12184
+    },
+    {
+      "epoch": 3.2347006504712597,
+      "grad_norm": 0.46281960152815793,
+      "learning_rate": 2.2084449031331547e-06,
+      "loss": 0.559,
+      "step": 12185
+    },
+    {
+      "epoch": 3.2349661489446437,
+      "grad_norm": 0.44736310895191644,
+      "learning_rate": 2.2080981603797733e-06,
+      "loss": 0.5648,
+      "step": 12186
+    },
+    {
+      "epoch": 3.2352316474180274,
+      "grad_norm": 0.4543565784586833,
+      "learning_rate": 2.2077514233191937e-06,
+      "loss": 0.5266,
+      "step": 12187
+    },
+    {
+      "epoch": 3.235497145891411,
+      "grad_norm": 0.46002096479386184,
+      "learning_rate": 2.2074046919581755e-06,
+      "loss": 0.5332,
+      "step": 12188
+    },
+    {
+      "epoch": 3.235762644364795,
+      "grad_norm": 0.4538934417673362,
+      "learning_rate": 2.207057966303483e-06,
+      "loss": 0.5735,
+      "step": 12189
+    },
+    {
+      "epoch": 3.2360281428381787,
+      "grad_norm": 0.45619657442827855,
+      "learning_rate": 2.2067112463618765e-06,
+      "loss": 0.5644,
+      "step": 12190
+    },
+    {
+      "epoch": 3.2362936413115624,
+      "grad_norm": 0.44448051857426224,
+      "learning_rate": 2.20636453214012e-06,
+      "loss": 0.5687,
+      "step": 12191
+    },
+    {
+      "epoch": 3.236559139784946,
+      "grad_norm": 0.444621750598178,
+      "learning_rate": 2.206017823644973e-06,
+      "loss": 0.5211,
+      "step": 12192
+    },
+    {
+      "epoch": 3.23682463825833,
+      "grad_norm": 0.4485008157651884,
+      "learning_rate": 2.2056711208831997e-06,
+      "loss": 0.5677,
+      "step": 12193
+    },
+    {
+      "epoch": 3.2370901367317138,
+      "grad_norm": 0.4528398790508969,
+      "learning_rate": 2.205324423861559e-06,
+      "loss": 0.5088,
+      "step": 12194
+    },
+    {
+      "epoch": 3.2373556352050974,
+      "grad_norm": 0.4471990174442139,
+      "learning_rate": 2.2049777325868136e-06,
+      "loss": 0.5256,
+      "step": 12195
+    },
+    {
+      "epoch": 3.2376211336784815,
+      "grad_norm": 0.445900984045634,
+      "learning_rate": 2.204631047065725e-06,
+      "loss": 0.5817,
+      "step": 12196
+    },
+    {
+      "epoch": 3.237886632151865,
+      "grad_norm": 0.45361077452901394,
+      "learning_rate": 2.204284367305054e-06,
+      "loss": 0.5159,
+      "step": 12197
+    },
+    {
+      "epoch": 3.2381521306252488,
+      "grad_norm": 0.452310084875701,
+      "learning_rate": 2.2039376933115625e-06,
+      "loss": 0.5282,
+      "step": 12198
+    },
+    {
+      "epoch": 3.238417629098633,
+      "grad_norm": 0.4488876666616326,
+      "learning_rate": 2.20359102509201e-06,
+      "loss": 0.5388,
+      "step": 12199
+    },
+    {
+      "epoch": 3.2386831275720165,
+      "grad_norm": 0.45576514812790636,
+      "learning_rate": 2.2032443626531587e-06,
+      "loss": 0.535,
+      "step": 12200
+    },
+    {
+      "epoch": 3.2389486260454,
+      "grad_norm": 0.462596653433408,
+      "learning_rate": 2.2028977060017687e-06,
+      "loss": 0.5768,
+      "step": 12201
+    },
+    {
+      "epoch": 3.239214124518784,
+      "grad_norm": 0.45109602400534715,
+      "learning_rate": 2.2025510551446007e-06,
+      "loss": 0.5857,
+      "step": 12202
+    },
+    {
+      "epoch": 3.239479622992168,
+      "grad_norm": 0.4595647937941383,
+      "learning_rate": 2.2022044100884154e-06,
+      "loss": 0.5234,
+      "step": 12203
+    },
+    {
+      "epoch": 3.2397451214655515,
+      "grad_norm": 0.4771367523638718,
+      "learning_rate": 2.2018577708399746e-06,
+      "loss": 0.5432,
+      "step": 12204
+    },
+    {
+      "epoch": 3.2400106199389356,
+      "grad_norm": 0.4652568836582108,
+      "learning_rate": 2.201511137406036e-06,
+      "loss": 0.5487,
+      "step": 12205
+    },
+    {
+      "epoch": 3.240276118412319,
+      "grad_norm": 0.45308743888268965,
+      "learning_rate": 2.2011645097933612e-06,
+      "loss": 0.5494,
+      "step": 12206
+    },
+    {
+      "epoch": 3.240541616885703,
+      "grad_norm": 0.4763589233900814,
+      "learning_rate": 2.20081788800871e-06,
+      "loss": 0.5645,
+      "step": 12207
+    },
+    {
+      "epoch": 3.2408071153590865,
+      "grad_norm": 0.46797291432006216,
+      "learning_rate": 2.2004712720588424e-06,
+      "loss": 0.5599,
+      "step": 12208
+    },
+    {
+      "epoch": 3.2410726138324706,
+      "grad_norm": 0.4523293511370984,
+      "learning_rate": 2.200124661950518e-06,
+      "loss": 0.5515,
+      "step": 12209
+    },
+    {
+      "epoch": 3.2413381123058542,
+      "grad_norm": 0.4784021315369009,
+      "learning_rate": 2.1997780576904983e-06,
+      "loss": 0.5567,
+      "step": 12210
+    },
+    {
+      "epoch": 3.241603610779238,
+      "grad_norm": 0.4712859001401952,
+      "learning_rate": 2.199431459285541e-06,
+      "loss": 0.5517,
+      "step": 12211
+    },
+    {
+      "epoch": 3.241869109252622,
+      "grad_norm": 0.4466948521886155,
+      "learning_rate": 2.1990848667424055e-06,
+      "loss": 0.5073,
+      "step": 12212
+    },
+    {
+      "epoch": 3.2421346077260056,
+      "grad_norm": 0.46764161057134546,
+      "learning_rate": 2.198738280067852e-06,
+      "loss": 0.5625,
+      "step": 12213
+    },
+    {
+      "epoch": 3.2424001061993892,
+      "grad_norm": 0.4732155112232539,
+      "learning_rate": 2.1983916992686404e-06,
+      "loss": 0.5553,
+      "step": 12214
+    },
+    {
+      "epoch": 3.2426656046727733,
+      "grad_norm": 0.46146798220541685,
+      "learning_rate": 2.198045124351528e-06,
+      "loss": 0.5472,
+      "step": 12215
+    },
+    {
+      "epoch": 3.242931103146157,
+      "grad_norm": 0.4473519016083044,
+      "learning_rate": 2.197698555323276e-06,
+      "loss": 0.5522,
+      "step": 12216
+    },
+    {
+      "epoch": 3.2431966016195406,
+      "grad_norm": 0.4684066269483547,
+      "learning_rate": 2.1973519921906424e-06,
+      "loss": 0.5246,
+      "step": 12217
+    },
+    {
+      "epoch": 3.2434621000929242,
+      "grad_norm": 0.4537425994399478,
+      "learning_rate": 2.1970054349603853e-06,
+      "loss": 0.548,
+      "step": 12218
+    },
+    {
+      "epoch": 3.2437275985663083,
+      "grad_norm": 0.45478074346668496,
+      "learning_rate": 2.196658883639264e-06,
+      "loss": 0.5296,
+      "step": 12219
+    },
+    {
+      "epoch": 3.243993097039692,
+      "grad_norm": 0.4513207887052805,
+      "learning_rate": 2.1963123382340374e-06,
+      "loss": 0.5484,
+      "step": 12220
+    },
+    {
+      "epoch": 3.2442585955130756,
+      "grad_norm": 0.4514299286004895,
+      "learning_rate": 2.195965798751465e-06,
+      "loss": 0.5486,
+      "step": 12221
+    },
+    {
+      "epoch": 3.2445240939864597,
+      "grad_norm": 0.4659791010550979,
+      "learning_rate": 2.195619265198303e-06,
+      "loss": 0.5422,
+      "step": 12222
+    },
+    {
+      "epoch": 3.2447895924598433,
+      "grad_norm": 0.45580487147978704,
+      "learning_rate": 2.19527273758131e-06,
+      "loss": 0.5557,
+      "step": 12223
+    },
+    {
+      "epoch": 3.245055090933227,
+      "grad_norm": 0.4468500408575031,
+      "learning_rate": 2.194926215907245e-06,
+      "loss": 0.5601,
+      "step": 12224
+    },
+    {
+      "epoch": 3.245320589406611,
+      "grad_norm": 0.4554152796638522,
+      "learning_rate": 2.1945797001828657e-06,
+      "loss": 0.5285,
+      "step": 12225
+    },
+    {
+      "epoch": 3.2455860878799947,
+      "grad_norm": 0.4523583981292327,
+      "learning_rate": 2.19423319041493e-06,
+      "loss": 0.5264,
+      "step": 12226
+    },
+    {
+      "epoch": 3.2458515863533783,
+      "grad_norm": 0.43941840350684247,
+      "learning_rate": 2.1938866866101962e-06,
+      "loss": 0.4943,
+      "step": 12227
+    },
+    {
+      "epoch": 3.2461170848267624,
+      "grad_norm": 0.4736205118896015,
+      "learning_rate": 2.1935401887754213e-06,
+      "loss": 0.5576,
+      "step": 12228
+    },
+    {
+      "epoch": 3.246382583300146,
+      "grad_norm": 0.45033360717758586,
+      "learning_rate": 2.193193696917363e-06,
+      "loss": 0.5956,
+      "step": 12229
+    },
+    {
+      "epoch": 3.2466480817735297,
+      "grad_norm": 0.4639290296005261,
+      "learning_rate": 2.1928472110427784e-06,
+      "loss": 0.5287,
+      "step": 12230
+    },
+    {
+      "epoch": 3.246913580246914,
+      "grad_norm": 0.4481895081183422,
+      "learning_rate": 2.1925007311584256e-06,
+      "loss": 0.5571,
+      "step": 12231
+    },
+    {
+      "epoch": 3.2471790787202974,
+      "grad_norm": 0.46467440781894104,
+      "learning_rate": 2.1921542572710615e-06,
+      "loss": 0.5407,
+      "step": 12232
+    },
+    {
+      "epoch": 3.247444577193681,
+      "grad_norm": 0.4626729522932544,
+      "learning_rate": 2.1918077893874436e-06,
+      "loss": 0.5786,
+      "step": 12233
+    },
+    {
+      "epoch": 3.2477100756670647,
+      "grad_norm": 0.4650606058278087,
+      "learning_rate": 2.1914613275143274e-06,
+      "loss": 0.5364,
+      "step": 12234
+    },
+    {
+      "epoch": 3.247975574140449,
+      "grad_norm": 0.45336035894428495,
+      "learning_rate": 2.191114871658471e-06,
+      "loss": 0.5641,
+      "step": 12235
+    },
+    {
+      "epoch": 3.2482410726138324,
+      "grad_norm": 0.47239233246393225,
+      "learning_rate": 2.190768421826631e-06,
+      "loss": 0.5477,
+      "step": 12236
+    },
+    {
+      "epoch": 3.248506571087216,
+      "grad_norm": 0.44701614909347615,
+      "learning_rate": 2.190421978025564e-06,
+      "loss": 0.5373,
+      "step": 12237
+    },
+    {
+      "epoch": 3.2487720695606,
+      "grad_norm": 0.4609332627294107,
+      "learning_rate": 2.190075540262026e-06,
+      "loss": 0.5384,
+      "step": 12238
+    },
+    {
+      "epoch": 3.249037568033984,
+      "grad_norm": 0.4458144543497252,
+      "learning_rate": 2.189729108542775e-06,
+      "loss": 0.5225,
+      "step": 12239
+    },
+    {
+      "epoch": 3.2493030665073674,
+      "grad_norm": 0.4413670248541004,
+      "learning_rate": 2.1893826828745656e-06,
+      "loss": 0.5428,
+      "step": 12240
+    },
+    {
+      "epoch": 3.2495685649807515,
+      "grad_norm": 0.45306350599003875,
+      "learning_rate": 2.1890362632641536e-06,
+      "loss": 0.5455,
+      "step": 12241
+    },
+    {
+      "epoch": 3.249834063454135,
+      "grad_norm": 0.45364293517903237,
+      "learning_rate": 2.188689849718296e-06,
+      "loss": 0.5685,
+      "step": 12242
+    },
+    {
+      "epoch": 3.250099561927519,
+      "grad_norm": 0.47282242529865876,
+      "learning_rate": 2.1883434422437487e-06,
+      "loss": 0.573,
+      "step": 12243
+    },
+    {
+      "epoch": 3.2503650604009025,
+      "grad_norm": 0.46429110361282744,
+      "learning_rate": 2.1879970408472675e-06,
+      "loss": 0.5483,
+      "step": 12244
+    },
+    {
+      "epoch": 3.2506305588742865,
+      "grad_norm": 0.4559969392553884,
+      "learning_rate": 2.1876506455356084e-06,
+      "loss": 0.5682,
+      "step": 12245
+    },
+    {
+      "epoch": 3.25089605734767,
+      "grad_norm": 0.45674876607865655,
+      "learning_rate": 2.1873042563155257e-06,
+      "loss": 0.5144,
+      "step": 12246
+    },
+    {
+      "epoch": 3.251161555821054,
+      "grad_norm": 0.4620448442801122,
+      "learning_rate": 2.186957873193776e-06,
+      "loss": 0.5343,
+      "step": 12247
+    },
+    {
+      "epoch": 3.251427054294438,
+      "grad_norm": 0.4640661206716123,
+      "learning_rate": 2.186611496177114e-06,
+      "loss": 0.5521,
+      "step": 12248
+    },
+    {
+      "epoch": 3.2516925527678215,
+      "grad_norm": 0.4673762151763436,
+      "learning_rate": 2.186265125272295e-06,
+      "loss": 0.5567,
+      "step": 12249
+    },
+    {
+      "epoch": 3.251958051241205,
+      "grad_norm": 0.4540230118986643,
+      "learning_rate": 2.1859187604860753e-06,
+      "loss": 0.5536,
+      "step": 12250
+    },
+    {
+      "epoch": 3.2522235497145893,
+      "grad_norm": 0.448016804956048,
+      "learning_rate": 2.185572401825208e-06,
+      "loss": 0.5524,
+      "step": 12251
+    },
+    {
+      "epoch": 3.252489048187973,
+      "grad_norm": 0.4504075288766789,
+      "learning_rate": 2.1852260492964486e-06,
+      "loss": 0.5649,
+      "step": 12252
+    },
+    {
+      "epoch": 3.2527545466613565,
+      "grad_norm": 0.45060298785596997,
+      "learning_rate": 2.1848797029065517e-06,
+      "loss": 0.5317,
+      "step": 12253
+    },
+    {
+      "epoch": 3.2530200451347406,
+      "grad_norm": 0.46116606816857464,
+      "learning_rate": 2.1845333626622723e-06,
+      "loss": 0.5562,
+      "step": 12254
+    },
+    {
+      "epoch": 3.2532855436081243,
+      "grad_norm": 0.4502536140958218,
+      "learning_rate": 2.184187028570364e-06,
+      "loss": 0.5406,
+      "step": 12255
+    },
+    {
+      "epoch": 3.253551042081508,
+      "grad_norm": 0.46030513987158267,
+      "learning_rate": 2.1838407006375837e-06,
+      "loss": 0.5609,
+      "step": 12256
+    },
+    {
+      "epoch": 3.253816540554892,
+      "grad_norm": 0.46397187384573835,
+      "learning_rate": 2.183494378870683e-06,
+      "loss": 0.5662,
+      "step": 12257
+    },
+    {
+      "epoch": 3.2540820390282756,
+      "grad_norm": 0.4581781020205727,
+      "learning_rate": 2.1831480632764164e-06,
+      "loss": 0.5493,
+      "step": 12258
+    },
+    {
+      "epoch": 3.2543475375016593,
+      "grad_norm": 0.44988742127413267,
+      "learning_rate": 2.182801753861539e-06,
+      "loss": 0.5263,
+      "step": 12259
+    },
+    {
+      "epoch": 3.2546130359750434,
+      "grad_norm": 0.4618088019976701,
+      "learning_rate": 2.1824554506328033e-06,
+      "loss": 0.5532,
+      "step": 12260
+    },
+    {
+      "epoch": 3.254878534448427,
+      "grad_norm": 0.4435286655850674,
+      "learning_rate": 2.182109153596964e-06,
+      "loss": 0.5122,
+      "step": 12261
+    },
+    {
+      "epoch": 3.2551440329218106,
+      "grad_norm": 0.45841993963219524,
+      "learning_rate": 2.181762862760775e-06,
+      "loss": 0.5734,
+      "step": 12262
+    },
+    {
+      "epoch": 3.2554095313951943,
+      "grad_norm": 0.46582912516943265,
+      "learning_rate": 2.181416578130989e-06,
+      "loss": 0.5516,
+      "step": 12263
+    },
+    {
+      "epoch": 3.2556750298685784,
+      "grad_norm": 0.4535072663127884,
+      "learning_rate": 2.1810702997143597e-06,
+      "loss": 0.5419,
+      "step": 12264
+    },
+    {
+      "epoch": 3.255940528341962,
+      "grad_norm": 0.45088452958914055,
+      "learning_rate": 2.1807240275176406e-06,
+      "loss": 0.5242,
+      "step": 12265
+    },
+    {
+      "epoch": 3.2562060268153457,
+      "grad_norm": 0.4464104589626324,
+      "learning_rate": 2.180377761547585e-06,
+      "loss": 0.5415,
+      "step": 12266
+    },
+    {
+      "epoch": 3.2564715252887297,
+      "grad_norm": 0.4489483961090341,
+      "learning_rate": 2.180031501810945e-06,
+      "loss": 0.5359,
+      "step": 12267
+    },
+    {
+      "epoch": 3.2567370237621134,
+      "grad_norm": 0.46077893288079735,
+      "learning_rate": 2.179685248314475e-06,
+      "loss": 0.5317,
+      "step": 12268
+    },
+    {
+      "epoch": 3.257002522235497,
+      "grad_norm": 0.4604300868922512,
+      "learning_rate": 2.1793390010649263e-06,
+      "loss": 0.5467,
+      "step": 12269
+    },
+    {
+      "epoch": 3.257268020708881,
+      "grad_norm": 0.44047003662387385,
+      "learning_rate": 2.178992760069052e-06,
+      "loss": 0.531,
+      "step": 12270
+    },
+    {
+      "epoch": 3.2575335191822647,
+      "grad_norm": 0.4659524432466133,
+      "learning_rate": 2.178646525333605e-06,
+      "loss": 0.5319,
+      "step": 12271
+    },
+    {
+      "epoch": 3.2577990176556484,
+      "grad_norm": 0.45961894884958276,
+      "learning_rate": 2.1783002968653375e-06,
+      "loss": 0.5161,
+      "step": 12272
+    },
+    {
+      "epoch": 3.258064516129032,
+      "grad_norm": 0.4550542980535424,
+      "learning_rate": 2.1779540746710028e-06,
+      "loss": 0.5581,
+      "step": 12273
+    },
+    {
+      "epoch": 3.258330014602416,
+      "grad_norm": 0.46943214935068683,
+      "learning_rate": 2.177607858757351e-06,
+      "loss": 0.5542,
+      "step": 12274
+    },
+    {
+      "epoch": 3.2585955130757998,
+      "grad_norm": 0.4693414891660907,
+      "learning_rate": 2.1772616491311355e-06,
+      "loss": 0.595,
+      "step": 12275
+    },
+    {
+      "epoch": 3.2588610115491834,
+      "grad_norm": 0.45634352290625546,
+      "learning_rate": 2.176915445799108e-06,
+      "loss": 0.5547,
+      "step": 12276
+    },
+    {
+      "epoch": 3.2591265100225675,
+      "grad_norm": 0.4505007710819265,
+      "learning_rate": 2.1765692487680207e-06,
+      "loss": 0.5502,
+      "step": 12277
+    },
+    {
+      "epoch": 3.259392008495951,
+      "grad_norm": 0.45089164788539504,
+      "learning_rate": 2.1762230580446247e-06,
+      "loss": 0.5391,
+      "step": 12278
+    },
+    {
+      "epoch": 3.2596575069693348,
+      "grad_norm": 0.4756201383844795,
+      "learning_rate": 2.175876873635673e-06,
+      "loss": 0.5554,
+      "step": 12279
+    },
+    {
+      "epoch": 3.259923005442719,
+      "grad_norm": 0.46986271506127986,
+      "learning_rate": 2.1755306955479137e-06,
+      "loss": 0.5679,
+      "step": 12280
+    },
+    {
+      "epoch": 3.2601885039161025,
+      "grad_norm": 0.45375843892770507,
+      "learning_rate": 2.1751845237881013e-06,
+      "loss": 0.5328,
+      "step": 12281
+    },
+    {
+      "epoch": 3.260454002389486,
+      "grad_norm": 0.45079587443096764,
+      "learning_rate": 2.1748383583629857e-06,
+      "loss": 0.5547,
+      "step": 12282
+    },
+    {
+      "epoch": 3.26071950086287,
+      "grad_norm": 0.47380479799000436,
+      "learning_rate": 2.1744921992793187e-06,
+      "loss": 0.5519,
+      "step": 12283
+    },
+    {
+      "epoch": 3.260984999336254,
+      "grad_norm": 0.4541047389290832,
+      "learning_rate": 2.1741460465438507e-06,
+      "loss": 0.5281,
+      "step": 12284
+    },
+    {
+      "epoch": 3.2612504978096375,
+      "grad_norm": 0.46714790443204585,
+      "learning_rate": 2.1737999001633333e-06,
+      "loss": 0.542,
+      "step": 12285
+    },
+    {
+      "epoch": 3.2615159962830216,
+      "grad_norm": 0.4499956143798917,
+      "learning_rate": 2.1734537601445157e-06,
+      "loss": 0.5015,
+      "step": 12286
+    },
+    {
+      "epoch": 3.261781494756405,
+      "grad_norm": 0.44120477205898767,
+      "learning_rate": 2.173107626494149e-06,
+      "loss": 0.5094,
+      "step": 12287
+    },
+    {
+      "epoch": 3.262046993229789,
+      "grad_norm": 0.4657718455642984,
+      "learning_rate": 2.1727614992189842e-06,
+      "loss": 0.5339,
+      "step": 12288
+    },
+    {
+      "epoch": 3.262312491703173,
+      "grad_norm": 0.45233398986833173,
+      "learning_rate": 2.1724153783257717e-06,
+      "loss": 0.5179,
+      "step": 12289
+    },
+    {
+      "epoch": 3.2625779901765566,
+      "grad_norm": 0.46576182783896775,
+      "learning_rate": 2.172069263821261e-06,
+      "loss": 0.5213,
+      "step": 12290
+    },
+    {
+      "epoch": 3.2628434886499402,
+      "grad_norm": 0.44764117450619784,
+      "learning_rate": 2.1717231557122032e-06,
+      "loss": 0.5296,
+      "step": 12291
+    },
+    {
+      "epoch": 3.263108987123324,
+      "grad_norm": 0.46302309626466576,
+      "learning_rate": 2.1713770540053474e-06,
+      "loss": 0.5757,
+      "step": 12292
+    },
+    {
+      "epoch": 3.263374485596708,
+      "grad_norm": 0.45631612419447926,
+      "learning_rate": 2.171030958707443e-06,
+      "loss": 0.5762,
+      "step": 12293
+    },
+    {
+      "epoch": 3.2636399840700916,
+      "grad_norm": 0.4492634451632467,
+      "learning_rate": 2.1706848698252408e-06,
+      "loss": 0.5664,
+      "step": 12294
+    },
+    {
+      "epoch": 3.2639054825434752,
+      "grad_norm": 0.4575033662401832,
+      "learning_rate": 2.1703387873654896e-06,
+      "loss": 0.5582,
+      "step": 12295
+    },
+    {
+      "epoch": 3.2641709810168593,
+      "grad_norm": 0.46534394944435503,
+      "learning_rate": 2.1699927113349404e-06,
+      "loss": 0.5315,
+      "step": 12296
+    },
+    {
+      "epoch": 3.264436479490243,
+      "grad_norm": 0.4556667307815348,
+      "learning_rate": 2.1696466417403407e-06,
+      "loss": 0.5602,
+      "step": 12297
+    },
+    {
+      "epoch": 3.2647019779636266,
+      "grad_norm": 0.4717002788198049,
+      "learning_rate": 2.16930057858844e-06,
+      "loss": 0.5417,
+      "step": 12298
+    },
+    {
+      "epoch": 3.2649674764370102,
+      "grad_norm": 0.45028413547115853,
+      "learning_rate": 2.1689545218859876e-06,
+      "loss": 0.5638,
+      "step": 12299
+    },
+    {
+      "epoch": 3.2652329749103943,
+      "grad_norm": 0.453856312188405,
+      "learning_rate": 2.168608471639732e-06,
+      "loss": 0.5384,
+      "step": 12300
+    },
+    {
+      "epoch": 3.265498473383778,
+      "grad_norm": 0.4551687062613768,
+      "learning_rate": 2.1682624278564234e-06,
+      "loss": 0.5525,
+      "step": 12301
+    },
+    {
+      "epoch": 3.2657639718571616,
+      "grad_norm": 0.4496812130929585,
+      "learning_rate": 2.167916390542811e-06,
+      "loss": 0.5314,
+      "step": 12302
+    },
+    {
+      "epoch": 3.2660294703305457,
+      "grad_norm": 0.4478350802724008,
+      "learning_rate": 2.1675703597056402e-06,
+      "loss": 0.568,
+      "step": 12303
+    },
+    {
+      "epoch": 3.2662949688039293,
+      "grad_norm": 0.46553783889200445,
+      "learning_rate": 2.167224335351662e-06,
+      "loss": 0.5301,
+      "step": 12304
+    },
+    {
+      "epoch": 3.266560467277313,
+      "grad_norm": 0.45729653131524617,
+      "learning_rate": 2.1668783174876233e-06,
+      "loss": 0.5335,
+      "step": 12305
+    },
+    {
+      "epoch": 3.266825965750697,
+      "grad_norm": 0.4561360195630808,
+      "learning_rate": 2.1665323061202736e-06,
+      "loss": 0.5668,
+      "step": 12306
+    },
+    {
+      "epoch": 3.2670914642240807,
+      "grad_norm": 0.44572420469491425,
+      "learning_rate": 2.1661863012563607e-06,
+      "loss": 0.5385,
+      "step": 12307
+    },
+    {
+      "epoch": 3.2673569626974643,
+      "grad_norm": 0.4574547912527172,
+      "learning_rate": 2.165840302902632e-06,
+      "loss": 0.5604,
+      "step": 12308
+    },
+    {
+      "epoch": 3.2676224611708484,
+      "grad_norm": 0.4789080110011661,
+      "learning_rate": 2.1654943110658357e-06,
+      "loss": 0.6085,
+      "step": 12309
+    },
+    {
+      "epoch": 3.267887959644232,
+      "grad_norm": 0.4543267047960963,
+      "learning_rate": 2.165148325752719e-06,
+      "loss": 0.5521,
+      "step": 12310
+    },
+    {
+      "epoch": 3.2681534581176157,
+      "grad_norm": 0.4496362006975845,
+      "learning_rate": 2.1648023469700296e-06,
+      "loss": 0.5577,
+      "step": 12311
+    },
+    {
+      "epoch": 3.268418956591,
+      "grad_norm": 0.4734540988237096,
+      "learning_rate": 2.1644563747245157e-06,
+      "loss": 0.575,
+      "step": 12312
+    },
+    {
+      "epoch": 3.2686844550643834,
+      "grad_norm": 0.4523182378340119,
+      "learning_rate": 2.164110409022924e-06,
+      "loss": 0.5378,
+      "step": 12313
+    },
+    {
+      "epoch": 3.268949953537767,
+      "grad_norm": 0.44522234767187285,
+      "learning_rate": 2.163764449872002e-06,
+      "loss": 0.5616,
+      "step": 12314
+    },
+    {
+      "epoch": 3.269215452011151,
+      "grad_norm": 0.457241552116792,
+      "learning_rate": 2.163418497278496e-06,
+      "loss": 0.5447,
+      "step": 12315
+    },
+    {
+      "epoch": 3.269480950484535,
+      "grad_norm": 0.4783272653819701,
+      "learning_rate": 2.163072551249153e-06,
+      "loss": 0.571,
+      "step": 12316
+    },
+    {
+      "epoch": 3.2697464489579184,
+      "grad_norm": 0.47933894285094675,
+      "learning_rate": 2.162726611790721e-06,
+      "loss": 0.5803,
+      "step": 12317
+    },
+    {
+      "epoch": 3.270011947431302,
+      "grad_norm": 0.45006221733087765,
+      "learning_rate": 2.1623806789099453e-06,
+      "loss": 0.5411,
+      "step": 12318
+    },
+    {
+      "epoch": 3.270277445904686,
+      "grad_norm": 0.4674729523320809,
+      "learning_rate": 2.1620347526135734e-06,
+      "loss": 0.5736,
+      "step": 12319
+    },
+    {
+      "epoch": 3.27054294437807,
+      "grad_norm": 0.4630387602480421,
+      "learning_rate": 2.161688832908351e-06,
+      "loss": 0.5539,
+      "step": 12320
+    },
+    {
+      "epoch": 3.2708084428514534,
+      "grad_norm": 0.46411430443083534,
+      "learning_rate": 2.1613429198010246e-06,
+      "loss": 0.5236,
+      "step": 12321
+    },
+    {
+      "epoch": 3.2710739413248375,
+      "grad_norm": 0.46004419225889487,
+      "learning_rate": 2.160997013298341e-06,
+      "loss": 0.5593,
+      "step": 12322
+    },
+    {
+      "epoch": 3.271339439798221,
+      "grad_norm": 0.4752256263664418,
+      "learning_rate": 2.160651113407045e-06,
+      "loss": 0.5583,
+      "step": 12323
+    },
+    {
+      "epoch": 3.271604938271605,
+      "grad_norm": 0.46458881658982487,
+      "learning_rate": 2.160305220133883e-06,
+      "loss": 0.5604,
+      "step": 12324
+    },
+    {
+      "epoch": 3.271870436744989,
+      "grad_norm": 0.46593997070643833,
+      "learning_rate": 2.159959333485602e-06,
+      "loss": 0.5425,
+      "step": 12325
+    },
+    {
+      "epoch": 3.2721359352183725,
+      "grad_norm": 0.46545380918963536,
+      "learning_rate": 2.1596134534689454e-06,
+      "loss": 0.545,
+      "step": 12326
+    },
+    {
+      "epoch": 3.272401433691756,
+      "grad_norm": 0.46336840608494917,
+      "learning_rate": 2.159267580090661e-06,
+      "loss": 0.5467,
+      "step": 12327
+    },
+    {
+      "epoch": 3.27266693216514,
+      "grad_norm": 0.4640023370949904,
+      "learning_rate": 2.158921713357492e-06,
+      "loss": 0.533,
+      "step": 12328
+    },
+    {
+      "epoch": 3.272932430638524,
+      "grad_norm": 0.44217133281709387,
+      "learning_rate": 2.1585758532761848e-06,
+      "loss": 0.56,
+      "step": 12329
+    },
+    {
+      "epoch": 3.2731979291119075,
+      "grad_norm": 0.45820796240795947,
+      "learning_rate": 2.158229999853485e-06,
+      "loss": 0.5412,
+      "step": 12330
+    },
+    {
+      "epoch": 3.273463427585291,
+      "grad_norm": 0.45728993190774014,
+      "learning_rate": 2.1578841530961375e-06,
+      "loss": 0.5027,
+      "step": 12331
+    },
+    {
+      "epoch": 3.2737289260586753,
+      "grad_norm": 0.45754961032599484,
+      "learning_rate": 2.1575383130108857e-06,
+      "loss": 0.5496,
+      "step": 12332
+    },
+    {
+      "epoch": 3.273994424532059,
+      "grad_norm": 0.4486056707893225,
+      "learning_rate": 2.1571924796044755e-06,
+      "loss": 0.5375,
+      "step": 12333
+    },
+    {
+      "epoch": 3.2742599230054426,
+      "grad_norm": 0.44641523693761453,
+      "learning_rate": 2.156846652883651e-06,
+      "loss": 0.5778,
+      "step": 12334
+    },
+    {
+      "epoch": 3.2745254214788266,
+      "grad_norm": 0.45246039508430874,
+      "learning_rate": 2.1565008328551573e-06,
+      "loss": 0.5506,
+      "step": 12335
+    },
+    {
+      "epoch": 3.2747909199522103,
+      "grad_norm": 0.44492815619577564,
+      "learning_rate": 2.156155019525738e-06,
+      "loss": 0.4808,
+      "step": 12336
+    },
+    {
+      "epoch": 3.275056418425594,
+      "grad_norm": 0.4732053512559555,
+      "learning_rate": 2.1558092129021385e-06,
+      "loss": 0.5433,
+      "step": 12337
+    },
+    {
+      "epoch": 3.275321916898978,
+      "grad_norm": 0.43960027974845506,
+      "learning_rate": 2.1554634129911016e-06,
+      "loss": 0.5592,
+      "step": 12338
+    },
+    {
+      "epoch": 3.2755874153723616,
+      "grad_norm": 0.46746031254095655,
+      "learning_rate": 2.1551176197993717e-06,
+      "loss": 0.5444,
+      "step": 12339
+    },
+    {
+      "epoch": 3.2758529138457453,
+      "grad_norm": 0.45197983128030766,
+      "learning_rate": 2.1547718333336923e-06,
+      "loss": 0.5453,
+      "step": 12340
+    },
+    {
+      "epoch": 3.2761184123191294,
+      "grad_norm": 0.4415563336430487,
+      "learning_rate": 2.1544260536008077e-06,
+      "loss": 0.5269,
+      "step": 12341
+    },
+    {
+      "epoch": 3.276383910792513,
+      "grad_norm": 0.4593169665518242,
+      "learning_rate": 2.154080280607462e-06,
+      "loss": 0.5413,
+      "step": 12342
+    },
+    {
+      "epoch": 3.2766494092658967,
+      "grad_norm": 0.4484731636262464,
+      "learning_rate": 2.1537345143603967e-06,
+      "loss": 0.4824,
+      "step": 12343
+    },
+    {
+      "epoch": 3.2769149077392807,
+      "grad_norm": 0.455571153332722,
+      "learning_rate": 2.1533887548663564e-06,
+      "loss": 0.5691,
+      "step": 12344
+    },
+    {
+      "epoch": 3.2771804062126644,
+      "grad_norm": 0.4404313283931506,
+      "learning_rate": 2.153043002132084e-06,
+      "loss": 0.5478,
+      "step": 12345
+    },
+    {
+      "epoch": 3.277445904686048,
+      "grad_norm": 0.48317361280282894,
+      "learning_rate": 2.1526972561643222e-06,
+      "loss": 0.5583,
+      "step": 12346
+    },
+    {
+      "epoch": 3.2777114031594317,
+      "grad_norm": 0.4410181208477213,
+      "learning_rate": 2.1523515169698144e-06,
+      "loss": 0.5816,
+      "step": 12347
+    },
+    {
+      "epoch": 3.2779769016328157,
+      "grad_norm": 0.4576468337877864,
+      "learning_rate": 2.1520057845553048e-06,
+      "loss": 0.5535,
+      "step": 12348
+    },
+    {
+      "epoch": 3.2782424001061994,
+      "grad_norm": 0.4434151125089868,
+      "learning_rate": 2.151660058927533e-06,
+      "loss": 0.5479,
+      "step": 12349
+    },
+    {
+      "epoch": 3.278507898579583,
+      "grad_norm": 0.44108424075470404,
+      "learning_rate": 2.151314340093243e-06,
+      "loss": 0.566,
+      "step": 12350
+    },
+    {
+      "epoch": 3.278773397052967,
+      "grad_norm": 0.4632574225230179,
+      "learning_rate": 2.1509686280591766e-06,
+      "loss": 0.5614,
+      "step": 12351
+    },
+    {
+      "epoch": 3.2790388955263507,
+      "grad_norm": 0.46406939028470656,
+      "learning_rate": 2.1506229228320775e-06,
+      "loss": 0.55,
+      "step": 12352
+    },
+    {
+      "epoch": 3.2793043939997344,
+      "grad_norm": 0.46178228423300816,
+      "learning_rate": 2.1502772244186862e-06,
+      "loss": 0.5917,
+      "step": 12353
+    },
+    {
+      "epoch": 3.279569892473118,
+      "grad_norm": 0.47012841625942065,
+      "learning_rate": 2.1499315328257457e-06,
+      "loss": 0.564,
+      "step": 12354
+    },
+    {
+      "epoch": 3.279835390946502,
+      "grad_norm": 0.47144097405157837,
+      "learning_rate": 2.1495858480599975e-06,
+      "loss": 0.5553,
+      "step": 12355
+    },
+    {
+      "epoch": 3.2801008894198858,
+      "grad_norm": 0.46739273924940306,
+      "learning_rate": 2.1492401701281827e-06,
+      "loss": 0.5712,
+      "step": 12356
+    },
+    {
+      "epoch": 3.2803663878932694,
+      "grad_norm": 0.4535229787918185,
+      "learning_rate": 2.1488944990370436e-06,
+      "loss": 0.509,
+      "step": 12357
+    },
+    {
+      "epoch": 3.2806318863666535,
+      "grad_norm": 0.4518493623382024,
+      "learning_rate": 2.1485488347933215e-06,
+      "loss": 0.5353,
+      "step": 12358
+    },
+    {
+      "epoch": 3.280897384840037,
+      "grad_norm": 0.468178091379063,
+      "learning_rate": 2.1482031774037573e-06,
+      "loss": 0.5526,
+      "step": 12359
+    },
+    {
+      "epoch": 3.2811628833134208,
+      "grad_norm": 0.4765956333059461,
+      "learning_rate": 2.147857526875094e-06,
+      "loss": 0.5643,
+      "step": 12360
+    },
+    {
+      "epoch": 3.281428381786805,
+      "grad_norm": 0.4496352903577852,
+      "learning_rate": 2.1475118832140696e-06,
+      "loss": 0.5364,
+      "step": 12361
+    },
+    {
+      "epoch": 3.2816938802601885,
+      "grad_norm": 0.44899248191256785,
+      "learning_rate": 2.147166246427427e-06,
+      "loss": 0.5069,
+      "step": 12362
+    },
+    {
+      "epoch": 3.281959378733572,
+      "grad_norm": 0.44507611563600624,
+      "learning_rate": 2.146820616521906e-06,
+      "loss": 0.5579,
+      "step": 12363
+    },
+    {
+      "epoch": 3.282224877206956,
+      "grad_norm": 0.45456431956357113,
+      "learning_rate": 2.146474993504248e-06,
+      "loss": 0.5426,
+      "step": 12364
+    },
+    {
+      "epoch": 3.28249037568034,
+      "grad_norm": 0.4635750203014093,
+      "learning_rate": 2.146129377381193e-06,
+      "loss": 0.5378,
+      "step": 12365
+    },
+    {
+      "epoch": 3.2827558741537235,
+      "grad_norm": 0.46148378318825967,
+      "learning_rate": 2.145783768159482e-06,
+      "loss": 0.5935,
+      "step": 12366
+    },
+    {
+      "epoch": 3.2830213726271076,
+      "grad_norm": 0.45021015544327986,
+      "learning_rate": 2.145438165845855e-06,
+      "loss": 0.4956,
+      "step": 12367
+    },
+    {
+      "epoch": 3.2832868711004912,
+      "grad_norm": 0.4580421117606796,
+      "learning_rate": 2.145092570447051e-06,
+      "loss": 0.5642,
+      "step": 12368
+    },
+    {
+      "epoch": 3.283552369573875,
+      "grad_norm": 0.46712077048895817,
+      "learning_rate": 2.1447469819698113e-06,
+      "loss": 0.5637,
+      "step": 12369
+    },
+    {
+      "epoch": 3.283817868047259,
+      "grad_norm": 0.4573299099565559,
+      "learning_rate": 2.144401400420875e-06,
+      "loss": 0.5648,
+      "step": 12370
+    },
+    {
+      "epoch": 3.2840833665206426,
+      "grad_norm": 0.46309959404780165,
+      "learning_rate": 2.1440558258069828e-06,
+      "loss": 0.5246,
+      "step": 12371
+    },
+    {
+      "epoch": 3.2843488649940262,
+      "grad_norm": 0.4608769272595653,
+      "learning_rate": 2.1437102581348727e-06,
+      "loss": 0.5532,
+      "step": 12372
+    },
+    {
+      "epoch": 3.28461436346741,
+      "grad_norm": 0.4266641255972511,
+      "learning_rate": 2.1433646974112856e-06,
+      "loss": 0.5374,
+      "step": 12373
+    },
+    {
+      "epoch": 3.284879861940794,
+      "grad_norm": 0.4676102793528825,
+      "learning_rate": 2.1430191436429596e-06,
+      "loss": 0.5509,
+      "step": 12374
+    },
+    {
+      "epoch": 3.2851453604141776,
+      "grad_norm": 0.4521447846602556,
+      "learning_rate": 2.142673596836634e-06,
+      "loss": 0.5623,
+      "step": 12375
+    },
+    {
+      "epoch": 3.2854108588875612,
+      "grad_norm": 0.4696830646397409,
+      "learning_rate": 2.1423280569990486e-06,
+      "loss": 0.5313,
+      "step": 12376
+    },
+    {
+      "epoch": 3.2856763573609453,
+      "grad_norm": 0.46817297306686456,
+      "learning_rate": 2.141982524136943e-06,
+      "loss": 0.5511,
+      "step": 12377
+    },
+    {
+      "epoch": 3.285941855834329,
+      "grad_norm": 0.46625064409427197,
+      "learning_rate": 2.1416369982570535e-06,
+      "loss": 0.5161,
+      "step": 12378
+    },
+    {
+      "epoch": 3.2862073543077126,
+      "grad_norm": 0.4470850947765853,
+      "learning_rate": 2.14129147936612e-06,
+      "loss": 0.5119,
+      "step": 12379
+    },
+    {
+      "epoch": 3.2864728527810967,
+      "grad_norm": 0.45845483365353,
+      "learning_rate": 2.1409459674708805e-06,
+      "loss": 0.5653,
+      "step": 12380
+    },
+    {
+      "epoch": 3.2867383512544803,
+      "grad_norm": 0.463464445182446,
+      "learning_rate": 2.1406004625780742e-06,
+      "loss": 0.5326,
+      "step": 12381
+    },
+    {
+      "epoch": 3.287003849727864,
+      "grad_norm": 0.47213776118015705,
+      "learning_rate": 2.140254964694439e-06,
+      "loss": 0.556,
+      "step": 12382
+    },
+    {
+      "epoch": 3.2872693482012476,
+      "grad_norm": 0.4557688333004908,
+      "learning_rate": 2.139909473826713e-06,
+      "loss": 0.5626,
+      "step": 12383
+    },
+    {
+      "epoch": 3.2875348466746317,
+      "grad_norm": 0.4658287046462975,
+      "learning_rate": 2.139563989981633e-06,
+      "loss": 0.5272,
+      "step": 12384
+    },
+    {
+      "epoch": 3.2878003451480153,
+      "grad_norm": 0.45081826173860234,
+      "learning_rate": 2.1392185131659388e-06,
+      "loss": 0.5702,
+      "step": 12385
+    },
+    {
+      "epoch": 3.288065843621399,
+      "grad_norm": 0.4639014895647045,
+      "learning_rate": 2.138873043386366e-06,
+      "loss": 0.5292,
+      "step": 12386
+    },
+    {
+      "epoch": 3.288331342094783,
+      "grad_norm": 0.45025599929965054,
+      "learning_rate": 2.1385275806496536e-06,
+      "loss": 0.5231,
+      "step": 12387
+    },
+    {
+      "epoch": 3.2885968405681667,
+      "grad_norm": 0.4619551742318092,
+      "learning_rate": 2.1381821249625383e-06,
+      "loss": 0.5505,
+      "step": 12388
+    },
+    {
+      "epoch": 3.2888623390415503,
+      "grad_norm": 0.47476434652582916,
+      "learning_rate": 2.137836676331758e-06,
+      "loss": 0.5299,
+      "step": 12389
+    },
+    {
+      "epoch": 3.2891278375149344,
+      "grad_norm": 0.45406518245314403,
+      "learning_rate": 2.1374912347640487e-06,
+      "loss": 0.5534,
+      "step": 12390
+    },
+    {
+      "epoch": 3.289393335988318,
+      "grad_norm": 0.46029059240639586,
+      "learning_rate": 2.137145800266148e-06,
+      "loss": 0.5172,
+      "step": 12391
+    },
+    {
+      "epoch": 3.2896588344617017,
+      "grad_norm": 0.47365373366549757,
+      "learning_rate": 2.136800372844792e-06,
+      "loss": 0.5642,
+      "step": 12392
+    },
+    {
+      "epoch": 3.289924332935086,
+      "grad_norm": 0.4468839273599769,
+      "learning_rate": 2.136454952506718e-06,
+      "loss": 0.5498,
+      "step": 12393
+    },
+    {
+      "epoch": 3.2901898314084694,
+      "grad_norm": 0.44770564727568374,
+      "learning_rate": 2.136109539258664e-06,
+      "loss": 0.5434,
+      "step": 12394
+    },
+    {
+      "epoch": 3.290455329881853,
+      "grad_norm": 0.47233595705624315,
+      "learning_rate": 2.1357641331073637e-06,
+      "loss": 0.5676,
+      "step": 12395
+    },
+    {
+      "epoch": 3.290720828355237,
+      "grad_norm": 0.475650800074712,
+      "learning_rate": 2.1354187340595548e-06,
+      "loss": 0.5907,
+      "step": 12396
+    },
+    {
+      "epoch": 3.290986326828621,
+      "grad_norm": 0.46681056752671507,
+      "learning_rate": 2.1350733421219733e-06,
+      "loss": 0.5241,
+      "step": 12397
+    },
+    {
+      "epoch": 3.2912518253020044,
+      "grad_norm": 0.4481327864639219,
+      "learning_rate": 2.134727957301355e-06,
+      "loss": 0.5539,
+      "step": 12398
+    },
+    {
+      "epoch": 3.2915173237753885,
+      "grad_norm": 0.4669374564227783,
+      "learning_rate": 2.134382579604436e-06,
+      "loss": 0.5578,
+      "step": 12399
+    },
+    {
+      "epoch": 3.291782822248772,
+      "grad_norm": 0.4504344209920806,
+      "learning_rate": 2.1340372090379517e-06,
+      "loss": 0.5469,
+      "step": 12400
+    },
+    {
+      "epoch": 3.292048320722156,
+      "grad_norm": 0.4589140818451969,
+      "learning_rate": 2.133691845608638e-06,
+      "loss": 0.5457,
+      "step": 12401
+    },
+    {
+      "epoch": 3.2923138191955394,
+      "grad_norm": 0.4624496461005644,
+      "learning_rate": 2.1333464893232303e-06,
+      "loss": 0.5519,
+      "step": 12402
+    },
+    {
+      "epoch": 3.2925793176689235,
+      "grad_norm": 0.44704220602180317,
+      "learning_rate": 2.1330011401884636e-06,
+      "loss": 0.5732,
+      "step": 12403
+    },
+    {
+      "epoch": 3.292844816142307,
+      "grad_norm": 0.46443935285081467,
+      "learning_rate": 2.1326557982110733e-06,
+      "loss": 0.5525,
+      "step": 12404
+    },
+    {
+      "epoch": 3.293110314615691,
+      "grad_norm": 0.4535438882521344,
+      "learning_rate": 2.132310463397794e-06,
+      "loss": 0.5578,
+      "step": 12405
+    },
+    {
+      "epoch": 3.293375813089075,
+      "grad_norm": 0.4606368317492769,
+      "learning_rate": 2.1319651357553623e-06,
+      "loss": 0.5499,
+      "step": 12406
+    },
+    {
+      "epoch": 3.2936413115624585,
+      "grad_norm": 0.4572589082842307,
+      "learning_rate": 2.1316198152905103e-06,
+      "loss": 0.5517,
+      "step": 12407
+    },
+    {
+      "epoch": 3.293906810035842,
+      "grad_norm": 0.46150367572548606,
+      "learning_rate": 2.131274502009974e-06,
+      "loss": 0.5591,
+      "step": 12408
+    },
+    {
+      "epoch": 3.2941723085092263,
+      "grad_norm": 0.4576877179941248,
+      "learning_rate": 2.130929195920488e-06,
+      "loss": 0.5428,
+      "step": 12409
+    },
+    {
+      "epoch": 3.29443780698261,
+      "grad_norm": 0.4465836768067627,
+      "learning_rate": 2.130583897028786e-06,
+      "loss": 0.5316,
+      "step": 12410
+    },
+    {
+      "epoch": 3.2947033054559935,
+      "grad_norm": 0.46352521000381525,
+      "learning_rate": 2.1302386053416026e-06,
+      "loss": 0.5043,
+      "step": 12411
+    },
+    {
+      "epoch": 3.294968803929377,
+      "grad_norm": 0.45032463938605893,
+      "learning_rate": 2.129893320865672e-06,
+      "loss": 0.5519,
+      "step": 12412
+    },
+    {
+      "epoch": 3.2952343024027613,
+      "grad_norm": 0.4709078747874101,
+      "learning_rate": 2.1295480436077275e-06,
+      "loss": 0.5451,
+      "step": 12413
+    },
+    {
+      "epoch": 3.295499800876145,
+      "grad_norm": 0.4457063021855977,
+      "learning_rate": 2.129202773574503e-06,
+      "loss": 0.5635,
+      "step": 12414
+    },
+    {
+      "epoch": 3.2957652993495286,
+      "grad_norm": 0.4616322554486769,
+      "learning_rate": 2.128857510772733e-06,
+      "loss": 0.5392,
+      "step": 12415
+    },
+    {
+      "epoch": 3.2960307978229126,
+      "grad_norm": 0.45572101045975255,
+      "learning_rate": 2.128512255209149e-06,
+      "loss": 0.5647,
+      "step": 12416
+    },
+    {
+      "epoch": 3.2962962962962963,
+      "grad_norm": 0.4604240898712047,
+      "learning_rate": 2.1281670068904868e-06,
+      "loss": 0.5801,
+      "step": 12417
+    },
+    {
+      "epoch": 3.29656179476968,
+      "grad_norm": 0.4743002024590024,
+      "learning_rate": 2.127821765823478e-06,
+      "loss": 0.529,
+      "step": 12418
+    },
+    {
+      "epoch": 3.296827293243064,
+      "grad_norm": 0.4813014068520655,
+      "learning_rate": 2.127476532014856e-06,
+      "loss": 0.5521,
+      "step": 12419
+    },
+    {
+      "epoch": 3.2970927917164476,
+      "grad_norm": 0.4561602439983453,
+      "learning_rate": 2.1271313054713534e-06,
+      "loss": 0.5148,
+      "step": 12420
+    },
+    {
+      "epoch": 3.2973582901898313,
+      "grad_norm": 0.4517895152320192,
+      "learning_rate": 2.1267860861997035e-06,
+      "loss": 0.5633,
+      "step": 12421
+    },
+    {
+      "epoch": 3.2976237886632154,
+      "grad_norm": 0.4658096977668192,
+      "learning_rate": 2.1264408742066387e-06,
+      "loss": 0.5595,
+      "step": 12422
+    },
+    {
+      "epoch": 3.297889287136599,
+      "grad_norm": 0.44924424446352795,
+      "learning_rate": 2.1260956694988924e-06,
+      "loss": 0.5867,
+      "step": 12423
+    },
+    {
+      "epoch": 3.2981547856099827,
+      "grad_norm": 0.4650390134535376,
+      "learning_rate": 2.125750472083195e-06,
+      "loss": 0.5902,
+      "step": 12424
+    },
+    {
+      "epoch": 3.2984202840833667,
+      "grad_norm": 0.47204397022329564,
+      "learning_rate": 2.12540528196628e-06,
+      "loss": 0.5762,
+      "step": 12425
+    },
+    {
+      "epoch": 3.2986857825567504,
+      "grad_norm": 0.45965819358481436,
+      "learning_rate": 2.125060099154879e-06,
+      "loss": 0.5714,
+      "step": 12426
+    },
+    {
+      "epoch": 3.298951281030134,
+      "grad_norm": 0.4755868604587048,
+      "learning_rate": 2.124714923655724e-06,
+      "loss": 0.5714,
+      "step": 12427
+    },
+    {
+      "epoch": 3.299216779503518,
+      "grad_norm": 0.451780964745843,
+      "learning_rate": 2.1243697554755468e-06,
+      "loss": 0.5461,
+      "step": 12428
+    },
+    {
+      "epoch": 3.2994822779769017,
+      "grad_norm": 0.46427767702418216,
+      "learning_rate": 2.1240245946210797e-06,
+      "loss": 0.5728,
+      "step": 12429
+    },
+    {
+      "epoch": 3.2997477764502854,
+      "grad_norm": 0.4610754211575944,
+      "learning_rate": 2.123679441099053e-06,
+      "loss": 0.592,
+      "step": 12430
+    },
+    {
+      "epoch": 3.300013274923669,
+      "grad_norm": 0.4417926175054234,
+      "learning_rate": 2.1233342949161984e-06,
+      "loss": 0.522,
+      "step": 12431
+    },
+    {
+      "epoch": 3.300278773397053,
+      "grad_norm": 0.4680247748595976,
+      "learning_rate": 2.1229891560792474e-06,
+      "loss": 0.5727,
+      "step": 12432
+    },
+    {
+      "epoch": 3.3005442718704368,
+      "grad_norm": 0.46310851191953706,
+      "learning_rate": 2.122644024594931e-06,
+      "loss": 0.5525,
+      "step": 12433
+    },
+    {
+      "epoch": 3.3008097703438204,
+      "grad_norm": 0.4560786977925917,
+      "learning_rate": 2.1222989004699803e-06,
+      "loss": 0.5463,
+      "step": 12434
+    },
+    {
+      "epoch": 3.3010752688172045,
+      "grad_norm": 0.4661549219068018,
+      "learning_rate": 2.1219537837111264e-06,
+      "loss": 0.5522,
+      "step": 12435
+    },
+    {
+      "epoch": 3.301340767290588,
+      "grad_norm": 0.4619028587425063,
+      "learning_rate": 2.1216086743250985e-06,
+      "loss": 0.5666,
+      "step": 12436
+    },
+    {
+      "epoch": 3.3016062657639718,
+      "grad_norm": 0.4516607422828077,
+      "learning_rate": 2.1212635723186274e-06,
+      "loss": 0.5122,
+      "step": 12437
+    },
+    {
+      "epoch": 3.3018717642373554,
+      "grad_norm": 0.46857206034300036,
+      "learning_rate": 2.120918477698444e-06,
+      "loss": 0.584,
+      "step": 12438
+    },
+    {
+      "epoch": 3.3021372627107395,
+      "grad_norm": 0.44527405967549133,
+      "learning_rate": 2.1205733904712793e-06,
+      "loss": 0.4977,
+      "step": 12439
+    },
+    {
+      "epoch": 3.302402761184123,
+      "grad_norm": 0.4507144902397387,
+      "learning_rate": 2.1202283106438632e-06,
+      "loss": 0.51,
+      "step": 12440
+    },
+    {
+      "epoch": 3.3026682596575068,
+      "grad_norm": 0.46250077713026594,
+      "learning_rate": 2.1198832382229236e-06,
+      "loss": 0.5389,
+      "step": 12441
+    },
+    {
+      "epoch": 3.302933758130891,
+      "grad_norm": 0.4359955068355555,
+      "learning_rate": 2.1195381732151918e-06,
+      "loss": 0.5535,
+      "step": 12442
+    },
+    {
+      "epoch": 3.3031992566042745,
+      "grad_norm": 0.4554354739491032,
+      "learning_rate": 2.119193115627397e-06,
+      "loss": 0.522,
+      "step": 12443
+    },
+    {
+      "epoch": 3.303464755077658,
+      "grad_norm": 0.47096946804051365,
+      "learning_rate": 2.118848065466269e-06,
+      "loss": 0.5244,
+      "step": 12444
+    },
+    {
+      "epoch": 3.303730253551042,
+      "grad_norm": 0.4601930551412257,
+      "learning_rate": 2.118503022738537e-06,
+      "loss": 0.558,
+      "step": 12445
+    },
+    {
+      "epoch": 3.303995752024426,
+      "grad_norm": 0.4680848282067587,
+      "learning_rate": 2.11815798745093e-06,
+      "loss": 0.5338,
+      "step": 12446
+    },
+    {
+      "epoch": 3.3042612504978095,
+      "grad_norm": 0.4636880869855386,
+      "learning_rate": 2.1178129596101775e-06,
+      "loss": 0.54,
+      "step": 12447
+    },
+    {
+      "epoch": 3.3045267489711936,
+      "grad_norm": 0.4571575295334612,
+      "learning_rate": 2.1174679392230076e-06,
+      "loss": 0.5426,
+      "step": 12448
+    },
+    {
+      "epoch": 3.3047922474445772,
+      "grad_norm": 0.45945410776255485,
+      "learning_rate": 2.1171229262961494e-06,
+      "loss": 0.567,
+      "step": 12449
+    },
+    {
+      "epoch": 3.305057745917961,
+      "grad_norm": 0.46038683212836085,
+      "learning_rate": 2.1167779208363317e-06,
+      "loss": 0.5372,
+      "step": 12450
+    },
+    {
+      "epoch": 3.305323244391345,
+      "grad_norm": 0.4410129222152262,
+      "learning_rate": 2.116432922850283e-06,
+      "loss": 0.5161,
+      "step": 12451
+    },
+    {
+      "epoch": 3.3055887428647286,
+      "grad_norm": 0.4562754830833976,
+      "learning_rate": 2.116087932344732e-06,
+      "loss": 0.5437,
+      "step": 12452
+    },
+    {
+      "epoch": 3.3058542413381122,
+      "grad_norm": 0.4714121654480041,
+      "learning_rate": 2.1157429493264056e-06,
+      "loss": 0.5461,
+      "step": 12453
+    },
+    {
+      "epoch": 3.3061197398114963,
+      "grad_norm": 0.46305007700456285,
+      "learning_rate": 2.1153979738020325e-06,
+      "loss": 0.5773,
+      "step": 12454
+    },
+    {
+      "epoch": 3.30638523828488,
+      "grad_norm": 0.4636551694201425,
+      "learning_rate": 2.1150530057783407e-06,
+      "loss": 0.5503,
+      "step": 12455
+    },
+    {
+      "epoch": 3.3066507367582636,
+      "grad_norm": 0.4600401005916624,
+      "learning_rate": 2.1147080452620573e-06,
+      "loss": 0.5429,
+      "step": 12456
+    },
+    {
+      "epoch": 3.3069162352316472,
+      "grad_norm": 0.44521919634042084,
+      "learning_rate": 2.1143630922599108e-06,
+      "loss": 0.5447,
+      "step": 12457
+    },
+    {
+      "epoch": 3.3071817337050313,
+      "grad_norm": 0.44575622568736406,
+      "learning_rate": 2.114018146778629e-06,
+      "loss": 0.5122,
+      "step": 12458
+    },
+    {
+      "epoch": 3.307447232178415,
+      "grad_norm": 0.448285534633307,
+      "learning_rate": 2.1136732088249372e-06,
+      "loss": 0.5424,
+      "step": 12459
+    },
+    {
+      "epoch": 3.3077127306517986,
+      "grad_norm": 0.46119896375268676,
+      "learning_rate": 2.113328278405564e-06,
+      "loss": 0.5752,
+      "step": 12460
+    },
+    {
+      "epoch": 3.3079782291251827,
+      "grad_norm": 0.47090997878443963,
+      "learning_rate": 2.1129833555272363e-06,
+      "loss": 0.5579,
+      "step": 12461
+    },
+    {
+      "epoch": 3.3082437275985663,
+      "grad_norm": 0.4596055827363926,
+      "learning_rate": 2.1126384401966806e-06,
+      "loss": 0.5604,
+      "step": 12462
+    },
+    {
+      "epoch": 3.30850922607195,
+      "grad_norm": 0.44433855933873495,
+      "learning_rate": 2.1122935324206238e-06,
+      "loss": 0.4976,
+      "step": 12463
+    },
+    {
+      "epoch": 3.308774724545334,
+      "grad_norm": 0.4598679614786918,
+      "learning_rate": 2.111948632205793e-06,
+      "loss": 0.5276,
+      "step": 12464
+    },
+    {
+      "epoch": 3.3090402230187177,
+      "grad_norm": 0.45450933980777275,
+      "learning_rate": 2.111603739558913e-06,
+      "loss": 0.5389,
+      "step": 12465
+    },
+    {
+      "epoch": 3.3093057214921013,
+      "grad_norm": 0.45739907359601684,
+      "learning_rate": 2.1112588544867115e-06,
+      "loss": 0.5303,
+      "step": 12466
+    },
+    {
+      "epoch": 3.309571219965485,
+      "grad_norm": 0.459642658666993,
+      "learning_rate": 2.110913976995914e-06,
+      "loss": 0.4873,
+      "step": 12467
+    },
+    {
+      "epoch": 3.309836718438869,
+      "grad_norm": 0.46011984312522497,
+      "learning_rate": 2.1105691070932465e-06,
+      "loss": 0.5357,
+      "step": 12468
+    },
+    {
+      "epoch": 3.3101022169122527,
+      "grad_norm": 0.46828927507083296,
+      "learning_rate": 2.110224244785436e-06,
+      "loss": 0.5498,
+      "step": 12469
+    },
+    {
+      "epoch": 3.3103677153856363,
+      "grad_norm": 0.45531443583831677,
+      "learning_rate": 2.109879390079206e-06,
+      "loss": 0.5362,
+      "step": 12470
+    },
+    {
+      "epoch": 3.3106332138590204,
+      "grad_norm": 0.4410326958464635,
+      "learning_rate": 2.1095345429812835e-06,
+      "loss": 0.5203,
+      "step": 12471
+    },
+    {
+      "epoch": 3.310898712332404,
+      "grad_norm": 0.4405471853967866,
+      "learning_rate": 2.1091897034983926e-06,
+      "loss": 0.5727,
+      "step": 12472
+    },
+    {
+      "epoch": 3.3111642108057877,
+      "grad_norm": 0.46935547032205094,
+      "learning_rate": 2.10884487163726e-06,
+      "loss": 0.5242,
+      "step": 12473
+    },
+    {
+      "epoch": 3.311429709279172,
+      "grad_norm": 0.44328272263231056,
+      "learning_rate": 2.10850004740461e-06,
+      "loss": 0.5545,
+      "step": 12474
+    },
+    {
+      "epoch": 3.3116952077525554,
+      "grad_norm": 0.45478070066380205,
+      "learning_rate": 2.1081552308071678e-06,
+      "loss": 0.5245,
+      "step": 12475
+    },
+    {
+      "epoch": 3.311960706225939,
+      "grad_norm": 0.45699553552185684,
+      "learning_rate": 2.1078104218516577e-06,
+      "loss": 0.552,
+      "step": 12476
+    },
+    {
+      "epoch": 3.312226204699323,
+      "grad_norm": 0.4458374547274046,
+      "learning_rate": 2.1074656205448048e-06,
+      "loss": 0.5542,
+      "step": 12477
+    },
+    {
+      "epoch": 3.312491703172707,
+      "grad_norm": 0.4638519657956804,
+      "learning_rate": 2.1071208268933335e-06,
+      "loss": 0.5587,
+      "step": 12478
+    },
+    {
+      "epoch": 3.3127572016460904,
+      "grad_norm": 0.4642410948923572,
+      "learning_rate": 2.1067760409039678e-06,
+      "loss": 0.5879,
+      "step": 12479
+    },
+    {
+      "epoch": 3.3130227001194745,
+      "grad_norm": 0.45855673658755025,
+      "learning_rate": 2.106431262583432e-06,
+      "loss": 0.5178,
+      "step": 12480
+    },
+    {
+      "epoch": 3.313288198592858,
+      "grad_norm": 0.4560886432338761,
+      "learning_rate": 2.1060864919384514e-06,
+      "loss": 0.5511,
+      "step": 12481
+    },
+    {
+      "epoch": 3.313553697066242,
+      "grad_norm": 0.4479522484387603,
+      "learning_rate": 2.1057417289757474e-06,
+      "loss": 0.538,
+      "step": 12482
+    },
+    {
+      "epoch": 3.313819195539626,
+      "grad_norm": 0.46573541513302824,
+      "learning_rate": 2.1053969737020448e-06,
+      "loss": 0.5702,
+      "step": 12483
+    },
+    {
+      "epoch": 3.3140846940130095,
+      "grad_norm": 0.4611913376015197,
+      "learning_rate": 2.1050522261240676e-06,
+      "loss": 0.5117,
+      "step": 12484
+    },
+    {
+      "epoch": 3.314350192486393,
+      "grad_norm": 0.4489390341304483,
+      "learning_rate": 2.1047074862485387e-06,
+      "loss": 0.5296,
+      "step": 12485
+    },
+    {
+      "epoch": 3.314615690959777,
+      "grad_norm": 0.4607158986637834,
+      "learning_rate": 2.1043627540821824e-06,
+      "loss": 0.5705,
+      "step": 12486
+    },
+    {
+      "epoch": 3.314881189433161,
+      "grad_norm": 0.4631355032749449,
+      "learning_rate": 2.1040180296317216e-06,
+      "loss": 0.5244,
+      "step": 12487
+    },
+    {
+      "epoch": 3.3151466879065445,
+      "grad_norm": 0.4545705844970469,
+      "learning_rate": 2.1036733129038782e-06,
+      "loss": 0.5517,
+      "step": 12488
+    },
+    {
+      "epoch": 3.315412186379928,
+      "grad_norm": 0.4590455565072322,
+      "learning_rate": 2.1033286039053752e-06,
+      "loss": 0.5125,
+      "step": 12489
+    },
+    {
+      "epoch": 3.3156776848533123,
+      "grad_norm": 0.4598888840084843,
+      "learning_rate": 2.1029839026429356e-06,
+      "loss": 0.5424,
+      "step": 12490
+    },
+    {
+      "epoch": 3.315943183326696,
+      "grad_norm": 0.45589539434615006,
+      "learning_rate": 2.102639209123282e-06,
+      "loss": 0.5493,
+      "step": 12491
+    },
+    {
+      "epoch": 3.3162086818000795,
+      "grad_norm": 0.46140689487330533,
+      "learning_rate": 2.1022945233531373e-06,
+      "loss": 0.5411,
+      "step": 12492
+    },
+    {
+      "epoch": 3.316474180273463,
+      "grad_norm": 0.4649259303458875,
+      "learning_rate": 2.1019498453392227e-06,
+      "loss": 0.587,
+      "step": 12493
+    },
+    {
+      "epoch": 3.3167396787468473,
+      "grad_norm": 0.45352260855505916,
+      "learning_rate": 2.1016051750882606e-06,
+      "loss": 0.5435,
+      "step": 12494
+    },
+    {
+      "epoch": 3.317005177220231,
+      "grad_norm": 0.45600065120720623,
+      "learning_rate": 2.101260512606973e-06,
+      "loss": 0.5267,
+      "step": 12495
+    },
+    {
+      "epoch": 3.3172706756936146,
+      "grad_norm": 0.4521577010053124,
+      "learning_rate": 2.1009158579020815e-06,
+      "loss": 0.5766,
+      "step": 12496
+    },
+    {
+      "epoch": 3.3175361741669986,
+      "grad_norm": 0.4579730024523372,
+      "learning_rate": 2.1005712109803083e-06,
+      "loss": 0.5831,
+      "step": 12497
+    },
+    {
+      "epoch": 3.3178016726403823,
+      "grad_norm": 0.45822623507432075,
+      "learning_rate": 2.1002265718483757e-06,
+      "loss": 0.5538,
+      "step": 12498
+    },
+    {
+      "epoch": 3.318067171113766,
+      "grad_norm": 0.44890247009933903,
+      "learning_rate": 2.0998819405130023e-06,
+      "loss": 0.5535,
+      "step": 12499
+    },
+    {
+      "epoch": 3.31833266958715,
+      "grad_norm": 0.4619831177424958,
+      "learning_rate": 2.099537316980911e-06,
+      "loss": 0.5656,
+      "step": 12500
+    },
+    {
+      "epoch": 3.3185981680605336,
+      "grad_norm": 0.4490152863751281,
+      "learning_rate": 2.099192701258822e-06,
+      "loss": 0.5364,
+      "step": 12501
+    },
+    {
+      "epoch": 3.3188636665339173,
+      "grad_norm": 0.47080043520137294,
+      "learning_rate": 2.098848093353457e-06,
+      "loss": 0.5707,
+      "step": 12502
+    },
+    {
+      "epoch": 3.3191291650073014,
+      "grad_norm": 0.457352177522771,
+      "learning_rate": 2.098503493271536e-06,
+      "loss": 0.5744,
+      "step": 12503
+    },
+    {
+      "epoch": 3.319394663480685,
+      "grad_norm": 0.4577390034326609,
+      "learning_rate": 2.098158901019781e-06,
+      "loss": 0.5146,
+      "step": 12504
+    },
+    {
+      "epoch": 3.3196601619540687,
+      "grad_norm": 0.46068076476183023,
+      "learning_rate": 2.0978143166049106e-06,
+      "loss": 0.5452,
+      "step": 12505
+    },
+    {
+      "epoch": 3.3199256604274527,
+      "grad_norm": 0.46622527022841176,
+      "learning_rate": 2.0974697400336456e-06,
+      "loss": 0.5522,
+      "step": 12506
+    },
+    {
+      "epoch": 3.3201911589008364,
+      "grad_norm": 0.45328441817231063,
+      "learning_rate": 2.0971251713127065e-06,
+      "loss": 0.5093,
+      "step": 12507
+    },
+    {
+      "epoch": 3.32045665737422,
+      "grad_norm": 0.43375240043693214,
+      "learning_rate": 2.096780610448813e-06,
+      "loss": 0.5379,
+      "step": 12508
+    },
+    {
+      "epoch": 3.320722155847604,
+      "grad_norm": 0.4563664377394157,
+      "learning_rate": 2.0964360574486846e-06,
+      "loss": 0.5857,
+      "step": 12509
+    },
+    {
+      "epoch": 3.3209876543209877,
+      "grad_norm": 0.46269890887163606,
+      "learning_rate": 2.0960915123190416e-06,
+      "loss": 0.5665,
+      "step": 12510
+    },
+    {
+      "epoch": 3.3212531527943714,
+      "grad_norm": 0.4686226873913855,
+      "learning_rate": 2.0957469750666026e-06,
+      "loss": 0.5694,
+      "step": 12511
+    },
+    {
+      "epoch": 3.321518651267755,
+      "grad_norm": 0.4667222560680663,
+      "learning_rate": 2.0954024456980875e-06,
+      "loss": 0.5363,
+      "step": 12512
+    },
+    {
+      "epoch": 3.321784149741139,
+      "grad_norm": 0.45605912239926993,
+      "learning_rate": 2.0950579242202153e-06,
+      "loss": 0.5235,
+      "step": 12513
+    },
+    {
+      "epoch": 3.3220496482145228,
+      "grad_norm": 0.4573541374565224,
+      "learning_rate": 2.0947134106397053e-06,
+      "loss": 0.5697,
+      "step": 12514
+    },
+    {
+      "epoch": 3.3223151466879064,
+      "grad_norm": 0.45532022688939805,
+      "learning_rate": 2.094368904963277e-06,
+      "loss": 0.5369,
+      "step": 12515
+    },
+    {
+      "epoch": 3.3225806451612905,
+      "grad_norm": 0.45840015082414337,
+      "learning_rate": 2.0940244071976466e-06,
+      "loss": 0.5725,
+      "step": 12516
+    },
+    {
+      "epoch": 3.322846143634674,
+      "grad_norm": 0.45857147291040484,
+      "learning_rate": 2.0936799173495348e-06,
+      "loss": 0.5534,
+      "step": 12517
+    },
+    {
+      "epoch": 3.3231116421080578,
+      "grad_norm": 0.46554016390189934,
+      "learning_rate": 2.0933354354256593e-06,
+      "loss": 0.5241,
+      "step": 12518
+    },
+    {
+      "epoch": 3.323377140581442,
+      "grad_norm": 0.447877572767853,
+      "learning_rate": 2.092990961432739e-06,
+      "loss": 0.5598,
+      "step": 12519
+    },
+    {
+      "epoch": 3.3236426390548255,
+      "grad_norm": 0.4553507984115381,
+      "learning_rate": 2.0926464953774907e-06,
+      "loss": 0.5391,
+      "step": 12520
+    },
+    {
+      "epoch": 3.323908137528209,
+      "grad_norm": 0.45434112405971894,
+      "learning_rate": 2.092302037266634e-06,
+      "loss": 0.5417,
+      "step": 12521
+    },
+    {
+      "epoch": 3.3241736360015928,
+      "grad_norm": 0.461418455961987,
+      "learning_rate": 2.0919575871068854e-06,
+      "loss": 0.5692,
+      "step": 12522
+    },
+    {
+      "epoch": 3.324439134474977,
+      "grad_norm": 0.4511552933044096,
+      "learning_rate": 2.0916131449049626e-06,
+      "loss": 0.5579,
+      "step": 12523
+    },
+    {
+      "epoch": 3.3247046329483605,
+      "grad_norm": 0.4610784963791445,
+      "learning_rate": 2.0912687106675835e-06,
+      "loss": 0.5428,
+      "step": 12524
+    },
+    {
+      "epoch": 3.324970131421744,
+      "grad_norm": 0.433307369833326,
+      "learning_rate": 2.0909242844014654e-06,
+      "loss": 0.4747,
+      "step": 12525
+    },
+    {
+      "epoch": 3.325235629895128,
+      "grad_norm": 0.43909895900227575,
+      "learning_rate": 2.0905798661133253e-06,
+      "loss": 0.5294,
+      "step": 12526
+    },
+    {
+      "epoch": 3.325501128368512,
+      "grad_norm": 0.4649584388562537,
+      "learning_rate": 2.0902354558098815e-06,
+      "loss": 0.5803,
+      "step": 12527
+    },
+    {
+      "epoch": 3.3257666268418955,
+      "grad_norm": 0.46429632614289235,
+      "learning_rate": 2.0898910534978476e-06,
+      "loss": 0.5618,
+      "step": 12528
+    },
+    {
+      "epoch": 3.3260321253152796,
+      "grad_norm": 0.47489956001679057,
+      "learning_rate": 2.089546659183943e-06,
+      "loss": 0.5304,
+      "step": 12529
+    },
+    {
+      "epoch": 3.3262976237886632,
+      "grad_norm": 0.45379297064326735,
+      "learning_rate": 2.0892022728748834e-06,
+      "loss": 0.5572,
+      "step": 12530
+    },
+    {
+      "epoch": 3.326563122262047,
+      "grad_norm": 0.46281803927806087,
+      "learning_rate": 2.088857894577385e-06,
+      "loss": 0.5619,
+      "step": 12531
+    },
+    {
+      "epoch": 3.326828620735431,
+      "grad_norm": 0.4532160774894718,
+      "learning_rate": 2.088513524298165e-06,
+      "loss": 0.5149,
+      "step": 12532
+    },
+    {
+      "epoch": 3.3270941192088146,
+      "grad_norm": 0.47249267368046766,
+      "learning_rate": 2.0881691620439397e-06,
+      "loss": 0.5494,
+      "step": 12533
+    },
+    {
+      "epoch": 3.3273596176821982,
+      "grad_norm": 0.45049153749463267,
+      "learning_rate": 2.0878248078214225e-06,
+      "loss": 0.5449,
+      "step": 12534
+    },
+    {
+      "epoch": 3.3276251161555823,
+      "grad_norm": 0.46322255618295577,
+      "learning_rate": 2.0874804616373314e-06,
+      "loss": 0.5394,
+      "step": 12535
+    },
+    {
+      "epoch": 3.327890614628966,
+      "grad_norm": 0.44444655621863416,
+      "learning_rate": 2.087136123498381e-06,
+      "loss": 0.5783,
+      "step": 12536
+    },
+    {
+      "epoch": 3.3281561131023496,
+      "grad_norm": 0.47398517757180264,
+      "learning_rate": 2.086791793411287e-06,
+      "loss": 0.5352,
+      "step": 12537
+    },
+    {
+      "epoch": 3.3284216115757337,
+      "grad_norm": 0.4790311055902139,
+      "learning_rate": 2.0864474713827653e-06,
+      "loss": 0.518,
+      "step": 12538
+    },
+    {
+      "epoch": 3.3286871100491173,
+      "grad_norm": 0.46572904239698093,
+      "learning_rate": 2.0861031574195296e-06,
+      "loss": 0.5186,
+      "step": 12539
+    },
+    {
+      "epoch": 3.328952608522501,
+      "grad_norm": 0.4519932846922998,
+      "learning_rate": 2.085758851528296e-06,
+      "loss": 0.5623,
+      "step": 12540
+    },
+    {
+      "epoch": 3.3292181069958846,
+      "grad_norm": 0.46093688790464066,
+      "learning_rate": 2.085414553715779e-06,
+      "loss": 0.5302,
+      "step": 12541
+    },
+    {
+      "epoch": 3.3294836054692687,
+      "grad_norm": 0.4639913452167748,
+      "learning_rate": 2.0850702639886937e-06,
+      "loss": 0.5798,
+      "step": 12542
+    },
+    {
+      "epoch": 3.3297491039426523,
+      "grad_norm": 0.458542296916341,
+      "learning_rate": 2.0847259823537537e-06,
+      "loss": 0.5601,
+      "step": 12543
+    },
+    {
+      "epoch": 3.330014602416036,
+      "grad_norm": 0.4611919718961669,
+      "learning_rate": 2.084381708817675e-06,
+      "loss": 0.5498,
+      "step": 12544
+    },
+    {
+      "epoch": 3.33028010088942,
+      "grad_norm": 0.46957024733982206,
+      "learning_rate": 2.0840374433871697e-06,
+      "loss": 0.5104,
+      "step": 12545
+    },
+    {
+      "epoch": 3.3305455993628037,
+      "grad_norm": 0.4814538441969792,
+      "learning_rate": 2.0836931860689524e-06,
+      "loss": 0.5516,
+      "step": 12546
+    },
+    {
+      "epoch": 3.3308110978361873,
+      "grad_norm": 0.4684573872955521,
+      "learning_rate": 2.0833489368697373e-06,
+      "loss": 0.5508,
+      "step": 12547
+    },
+    {
+      "epoch": 3.331076596309571,
+      "grad_norm": 0.46992997876637577,
+      "learning_rate": 2.083004695796238e-06,
+      "loss": 0.5246,
+      "step": 12548
+    },
+    {
+      "epoch": 3.331342094782955,
+      "grad_norm": 0.4802963158317988,
+      "learning_rate": 2.082660462855168e-06,
+      "loss": 0.5452,
+      "step": 12549
+    },
+    {
+      "epoch": 3.3316075932563387,
+      "grad_norm": 0.46271809428080457,
+      "learning_rate": 2.082316238053242e-06,
+      "loss": 0.5644,
+      "step": 12550
+    },
+    {
+      "epoch": 3.3318730917297223,
+      "grad_norm": 0.45403915028829606,
+      "learning_rate": 2.081972021397171e-06,
+      "loss": 0.5604,
+      "step": 12551
+    },
+    {
+      "epoch": 3.3321385902031064,
+      "grad_norm": 0.45091164578003795,
+      "learning_rate": 2.0816278128936695e-06,
+      "loss": 0.5552,
+      "step": 12552
+    },
+    {
+      "epoch": 3.33240408867649,
+      "grad_norm": 0.4600388462010232,
+      "learning_rate": 2.0812836125494497e-06,
+      "loss": 0.5875,
+      "step": 12553
+    },
+    {
+      "epoch": 3.3326695871498737,
+      "grad_norm": 0.4623470080786759,
+      "learning_rate": 2.0809394203712246e-06,
+      "loss": 0.5367,
+      "step": 12554
+    },
+    {
+      "epoch": 3.332935085623258,
+      "grad_norm": 0.44988545638785415,
+      "learning_rate": 2.080595236365707e-06,
+      "loss": 0.5684,
+      "step": 12555
+    },
+    {
+      "epoch": 3.3332005840966414,
+      "grad_norm": 0.47218987534434853,
+      "learning_rate": 2.0802510605396097e-06,
+      "loss": 0.5667,
+      "step": 12556
+    },
+    {
+      "epoch": 3.333466082570025,
+      "grad_norm": 0.46985412824531037,
+      "learning_rate": 2.0799068928996436e-06,
+      "loss": 0.6013,
+      "step": 12557
+    },
+    {
+      "epoch": 3.333731581043409,
+      "grad_norm": 0.4635316492686305,
+      "learning_rate": 2.079562733452522e-06,
+      "loss": 0.5173,
+      "step": 12558
+    },
+    {
+      "epoch": 3.333997079516793,
+      "grad_norm": 0.4555583645910946,
+      "learning_rate": 2.0792185822049563e-06,
+      "loss": 0.5313,
+      "step": 12559
+    },
+    {
+      "epoch": 3.3342625779901764,
+      "grad_norm": 0.4703939992715734,
+      "learning_rate": 2.0788744391636585e-06,
+      "loss": 0.5399,
+      "step": 12560
+    },
+    {
+      "epoch": 3.3345280764635605,
+      "grad_norm": 0.46322704782253765,
+      "learning_rate": 2.0785303043353415e-06,
+      "loss": 0.5564,
+      "step": 12561
+    },
+    {
+      "epoch": 3.334793574936944,
+      "grad_norm": 0.4601696510479565,
+      "learning_rate": 2.0781861777267145e-06,
+      "loss": 0.553,
+      "step": 12562
+    },
+    {
+      "epoch": 3.335059073410328,
+      "grad_norm": 0.45087305751246654,
+      "learning_rate": 2.077842059344489e-06,
+      "loss": 0.5249,
+      "step": 12563
+    },
+    {
+      "epoch": 3.335324571883712,
+      "grad_norm": 0.4653875480917446,
+      "learning_rate": 2.077497949195378e-06,
+      "loss": 0.5416,
+      "step": 12564
+    },
+    {
+      "epoch": 3.3355900703570955,
+      "grad_norm": 0.4693265173218801,
+      "learning_rate": 2.0771538472860906e-06,
+      "loss": 0.5374,
+      "step": 12565
+    },
+    {
+      "epoch": 3.335855568830479,
+      "grad_norm": 0.4612467910279887,
+      "learning_rate": 2.076809753623339e-06,
+      "loss": 0.5907,
+      "step": 12566
+    },
+    {
+      "epoch": 3.336121067303863,
+      "grad_norm": 0.4643558252177538,
+      "learning_rate": 2.076465668213834e-06,
+      "loss": 0.5667,
+      "step": 12567
+    },
+    {
+      "epoch": 3.336386565777247,
+      "grad_norm": 0.4700801785951776,
+      "learning_rate": 2.0761215910642847e-06,
+      "loss": 0.5284,
+      "step": 12568
+    },
+    {
+      "epoch": 3.3366520642506305,
+      "grad_norm": 0.47276837622266266,
+      "learning_rate": 2.0757775221814024e-06,
+      "loss": 0.5406,
+      "step": 12569
+    },
+    {
+      "epoch": 3.336917562724014,
+      "grad_norm": 0.4653828001244092,
+      "learning_rate": 2.0754334615718965e-06,
+      "loss": 0.5761,
+      "step": 12570
+    },
+    {
+      "epoch": 3.3371830611973983,
+      "grad_norm": 0.4640501182710139,
+      "learning_rate": 2.0750894092424783e-06,
+      "loss": 0.5676,
+      "step": 12571
+    },
+    {
+      "epoch": 3.337448559670782,
+      "grad_norm": 0.4551360914591177,
+      "learning_rate": 2.0747453651998565e-06,
+      "loss": 0.5314,
+      "step": 12572
+    },
+    {
+      "epoch": 3.3377140581441656,
+      "grad_norm": 0.45851887900778243,
+      "learning_rate": 2.0744013294507427e-06,
+      "loss": 0.524,
+      "step": 12573
+    },
+    {
+      "epoch": 3.3379795566175496,
+      "grad_norm": 0.45644004033825764,
+      "learning_rate": 2.0740573020018432e-06,
+      "loss": 0.5496,
+      "step": 12574
+    },
+    {
+      "epoch": 3.3382450550909333,
+      "grad_norm": 0.47957405640277395,
+      "learning_rate": 2.07371328285987e-06,
+      "loss": 0.5531,
+      "step": 12575
+    },
+    {
+      "epoch": 3.338510553564317,
+      "grad_norm": 0.46352746089664215,
+      "learning_rate": 2.073369272031531e-06,
+      "loss": 0.5762,
+      "step": 12576
+    },
+    {
+      "epoch": 3.3387760520377006,
+      "grad_norm": 0.46957871511109023,
+      "learning_rate": 2.0730252695235366e-06,
+      "loss": 0.5829,
+      "step": 12577
+    },
+    {
+      "epoch": 3.3390415505110846,
+      "grad_norm": 0.45905547577794376,
+      "learning_rate": 2.072681275342594e-06,
+      "loss": 0.519,
+      "step": 12578
+    },
+    {
+      "epoch": 3.3393070489844683,
+      "grad_norm": 0.4705870867385918,
+      "learning_rate": 2.072337289495414e-06,
+      "loss": 0.5849,
+      "step": 12579
+    },
+    {
+      "epoch": 3.339572547457852,
+      "grad_norm": 0.4662620438832262,
+      "learning_rate": 2.0719933119887032e-06,
+      "loss": 0.5554,
+      "step": 12580
+    },
+    {
+      "epoch": 3.339838045931236,
+      "grad_norm": 0.4743917837331077,
+      "learning_rate": 2.071649342829171e-06,
+      "loss": 0.5838,
+      "step": 12581
+    },
+    {
+      "epoch": 3.3401035444046197,
+      "grad_norm": 0.4621100641668938,
+      "learning_rate": 2.071305382023525e-06,
+      "loss": 0.5826,
+      "step": 12582
+    },
+    {
+      "epoch": 3.3403690428780033,
+      "grad_norm": 0.46659278903280366,
+      "learning_rate": 2.0709614295784734e-06,
+      "loss": 0.5747,
+      "step": 12583
+    },
+    {
+      "epoch": 3.3406345413513874,
+      "grad_norm": 0.4695482153466633,
+      "learning_rate": 2.070617485500725e-06,
+      "loss": 0.5534,
+      "step": 12584
+    },
+    {
+      "epoch": 3.340900039824771,
+      "grad_norm": 0.4665102211763585,
+      "learning_rate": 2.0702735497969873e-06,
+      "loss": 0.54,
+      "step": 12585
+    },
+    {
+      "epoch": 3.3411655382981547,
+      "grad_norm": 0.447311988563106,
+      "learning_rate": 2.069929622473967e-06,
+      "loss": 0.5523,
+      "step": 12586
+    },
+    {
+      "epoch": 3.3414310367715387,
+      "grad_norm": 0.47468945482066466,
+      "learning_rate": 2.0695857035383716e-06,
+      "loss": 0.4991,
+      "step": 12587
+    },
+    {
+      "epoch": 3.3416965352449224,
+      "grad_norm": 0.4685305937007719,
+      "learning_rate": 2.069241792996909e-06,
+      "loss": 0.5696,
+      "step": 12588
+    },
+    {
+      "epoch": 3.341962033718306,
+      "grad_norm": 0.468934319344743,
+      "learning_rate": 2.0688978908562863e-06,
+      "loss": 0.5474,
+      "step": 12589
+    },
+    {
+      "epoch": 3.34222753219169,
+      "grad_norm": 0.45826752054056613,
+      "learning_rate": 2.0685539971232114e-06,
+      "loss": 0.5327,
+      "step": 12590
+    },
+    {
+      "epoch": 3.3424930306650737,
+      "grad_norm": 0.4519640110125526,
+      "learning_rate": 2.068210111804389e-06,
+      "loss": 0.5342,
+      "step": 12591
+    },
+    {
+      "epoch": 3.3427585291384574,
+      "grad_norm": 0.45240682141334926,
+      "learning_rate": 2.0678662349065263e-06,
+      "loss": 0.5556,
+      "step": 12592
+    },
+    {
+      "epoch": 3.3430240276118415,
+      "grad_norm": 0.4690611911684847,
+      "learning_rate": 2.06752236643633e-06,
+      "loss": 0.5498,
+      "step": 12593
+    },
+    {
+      "epoch": 3.343289526085225,
+      "grad_norm": 0.45762953819575786,
+      "learning_rate": 2.067178506400507e-06,
+      "loss": 0.5522,
+      "step": 12594
+    },
+    {
+      "epoch": 3.3435550245586088,
+      "grad_norm": 0.4599788801632664,
+      "learning_rate": 2.0668346548057617e-06,
+      "loss": 0.5209,
+      "step": 12595
+    },
+    {
+      "epoch": 3.3438205230319924,
+      "grad_norm": 0.4647232001682794,
+      "learning_rate": 2.066490811658803e-06,
+      "loss": 0.5506,
+      "step": 12596
+    },
+    {
+      "epoch": 3.3440860215053765,
+      "grad_norm": 0.4708194205547255,
+      "learning_rate": 2.0661469769663335e-06,
+      "loss": 0.5883,
+      "step": 12597
+    },
+    {
+      "epoch": 3.34435151997876,
+      "grad_norm": 0.4552194989903483,
+      "learning_rate": 2.0658031507350607e-06,
+      "loss": 0.5836,
+      "step": 12598
+    },
+    {
+      "epoch": 3.3446170184521438,
+      "grad_norm": 0.482861944006749,
+      "learning_rate": 2.0654593329716893e-06,
+      "loss": 0.5638,
+      "step": 12599
+    },
+    {
+      "epoch": 3.344882516925528,
+      "grad_norm": 0.4570799635466538,
+      "learning_rate": 2.065115523682925e-06,
+      "loss": 0.5326,
+      "step": 12600
+    },
+    {
+      "epoch": 3.3451480153989115,
+      "grad_norm": 0.45793396524811736,
+      "learning_rate": 2.0647717228754727e-06,
+      "loss": 0.5468,
+      "step": 12601
+    },
+    {
+      "epoch": 3.345413513872295,
+      "grad_norm": 0.4644400156413009,
+      "learning_rate": 2.064427930556038e-06,
+      "loss": 0.5302,
+      "step": 12602
+    },
+    {
+      "epoch": 3.3456790123456788,
+      "grad_norm": 0.45037040767798153,
+      "learning_rate": 2.0640841467313246e-06,
+      "loss": 0.5462,
+      "step": 12603
+    },
+    {
+      "epoch": 3.345944510819063,
+      "grad_norm": 0.4709461828124922,
+      "learning_rate": 2.0637403714080376e-06,
+      "loss": 0.5572,
+      "step": 12604
+    },
+    {
+      "epoch": 3.3462100092924465,
+      "grad_norm": 0.4508927605673675,
+      "learning_rate": 2.0633966045928815e-06,
+      "loss": 0.5316,
+      "step": 12605
+    },
+    {
+      "epoch": 3.34647550776583,
+      "grad_norm": 0.4518100503192149,
+      "learning_rate": 2.0630528462925606e-06,
+      "loss": 0.5002,
+      "step": 12606
+    },
+    {
+      "epoch": 3.3467410062392142,
+      "grad_norm": 0.4650078283910292,
+      "learning_rate": 2.062709096513779e-06,
+      "loss": 0.5247,
+      "step": 12607
+    },
+    {
+      "epoch": 3.347006504712598,
+      "grad_norm": 0.462860954048532,
+      "learning_rate": 2.0623653552632415e-06,
+      "loss": 0.5685,
+      "step": 12608
+    },
+    {
+      "epoch": 3.3472720031859815,
+      "grad_norm": 0.4581316154958781,
+      "learning_rate": 2.0620216225476506e-06,
+      "loss": 0.5382,
+      "step": 12609
+    },
+    {
+      "epoch": 3.3475375016593656,
+      "grad_norm": 0.44994723903079986,
+      "learning_rate": 2.0616778983737103e-06,
+      "loss": 0.538,
+      "step": 12610
+    },
+    {
+      "epoch": 3.3478030001327492,
+      "grad_norm": 0.47737660026916773,
+      "learning_rate": 2.061334182748124e-06,
+      "loss": 0.5579,
+      "step": 12611
+    },
+    {
+      "epoch": 3.348068498606133,
+      "grad_norm": 0.4623047795624245,
+      "learning_rate": 2.060990475677595e-06,
+      "loss": 0.5317,
+      "step": 12612
+    },
+    {
+      "epoch": 3.348333997079517,
+      "grad_norm": 0.4587593415806032,
+      "learning_rate": 2.060646777168827e-06,
+      "loss": 0.4956,
+      "step": 12613
+    },
+    {
+      "epoch": 3.3485994955529006,
+      "grad_norm": 0.43871413611112003,
+      "learning_rate": 2.0603030872285226e-06,
+      "loss": 0.5138,
+      "step": 12614
+    },
+    {
+      "epoch": 3.3488649940262842,
+      "grad_norm": 0.4621758748062588,
+      "learning_rate": 2.0599594058633843e-06,
+      "loss": 0.5273,
+      "step": 12615
+    },
+    {
+      "epoch": 3.3491304924996683,
+      "grad_norm": 0.43588792995830405,
+      "learning_rate": 2.059615733080115e-06,
+      "loss": 0.5635,
+      "step": 12616
+    },
+    {
+      "epoch": 3.349395990973052,
+      "grad_norm": 0.47872700303162763,
+      "learning_rate": 2.059272068885417e-06,
+      "loss": 0.5635,
+      "step": 12617
+    },
+    {
+      "epoch": 3.3496614894464356,
+      "grad_norm": 0.4609291878356045,
+      "learning_rate": 2.058928413285993e-06,
+      "loss": 0.5556,
+      "step": 12618
+    },
+    {
+      "epoch": 3.3499269879198197,
+      "grad_norm": 0.45478630011007515,
+      "learning_rate": 2.0585847662885457e-06,
+      "loss": 0.5721,
+      "step": 12619
+    },
+    {
+      "epoch": 3.3501924863932033,
+      "grad_norm": 0.47566391685826487,
+      "learning_rate": 2.058241127899775e-06,
+      "loss": 0.5442,
+      "step": 12620
+    },
+    {
+      "epoch": 3.350457984866587,
+      "grad_norm": 0.4753381187272507,
+      "learning_rate": 2.057897498126384e-06,
+      "loss": 0.5623,
+      "step": 12621
+    },
+    {
+      "epoch": 3.3507234833399706,
+      "grad_norm": 0.4488691617329402,
+      "learning_rate": 2.0575538769750742e-06,
+      "loss": 0.5465,
+      "step": 12622
+    },
+    {
+      "epoch": 3.3509889818133547,
+      "grad_norm": 0.45356363308536435,
+      "learning_rate": 2.0572102644525476e-06,
+      "loss": 0.5437,
+      "step": 12623
+    },
+    {
+      "epoch": 3.3512544802867383,
+      "grad_norm": 0.4501010548049581,
+      "learning_rate": 2.0568666605655045e-06,
+      "loss": 0.5594,
+      "step": 12624
+    },
+    {
+      "epoch": 3.351519978760122,
+      "grad_norm": 0.45113569038960444,
+      "learning_rate": 2.056523065320648e-06,
+      "loss": 0.5408,
+      "step": 12625
+    },
+    {
+      "epoch": 3.351785477233506,
+      "grad_norm": 0.47599124266137316,
+      "learning_rate": 2.0561794787246763e-06,
+      "loss": 0.5177,
+      "step": 12626
+    },
+    {
+      "epoch": 3.3520509757068897,
+      "grad_norm": 0.456514770114625,
+      "learning_rate": 2.0558359007842916e-06,
+      "loss": 0.5449,
+      "step": 12627
+    },
+    {
+      "epoch": 3.3523164741802733,
+      "grad_norm": 0.4715250900398764,
+      "learning_rate": 2.055492331506194e-06,
+      "loss": 0.5657,
+      "step": 12628
+    },
+    {
+      "epoch": 3.3525819726536574,
+      "grad_norm": 0.45875132714516287,
+      "learning_rate": 2.0551487708970842e-06,
+      "loss": 0.5303,
+      "step": 12629
+    },
+    {
+      "epoch": 3.352847471127041,
+      "grad_norm": 0.49091743875106414,
+      "learning_rate": 2.054805218963663e-06,
+      "loss": 0.5169,
+      "step": 12630
+    },
+    {
+      "epoch": 3.3531129696004247,
+      "grad_norm": 0.47080179680134926,
+      "learning_rate": 2.05446167571263e-06,
+      "loss": 0.5511,
+      "step": 12631
+    },
+    {
+      "epoch": 3.3533784680738083,
+      "grad_norm": 0.4697817611520423,
+      "learning_rate": 2.054118141150685e-06,
+      "loss": 0.5312,
+      "step": 12632
+    },
+    {
+      "epoch": 3.3536439665471924,
+      "grad_norm": 0.4379885173106567,
+      "learning_rate": 2.053774615284528e-06,
+      "loss": 0.5413,
+      "step": 12633
+    },
+    {
+      "epoch": 3.353909465020576,
+      "grad_norm": 0.45811651613746984,
+      "learning_rate": 2.053431098120859e-06,
+      "loss": 0.5277,
+      "step": 12634
+    },
+    {
+      "epoch": 3.3541749634939597,
+      "grad_norm": 0.4519560773256558,
+      "learning_rate": 2.053087589666376e-06,
+      "loss": 0.5568,
+      "step": 12635
+    },
+    {
+      "epoch": 3.354440461967344,
+      "grad_norm": 0.45257613853947937,
+      "learning_rate": 2.052744089927781e-06,
+      "loss": 0.5529,
+      "step": 12636
+    },
+    {
+      "epoch": 3.3547059604407274,
+      "grad_norm": 0.4505801029398686,
+      "learning_rate": 2.0524005989117703e-06,
+      "loss": 0.5526,
+      "step": 12637
+    },
+    {
+      "epoch": 3.354971458914111,
+      "grad_norm": 0.47661685066395887,
+      "learning_rate": 2.0520571166250435e-06,
+      "loss": 0.53,
+      "step": 12638
+    },
+    {
+      "epoch": 3.355236957387495,
+      "grad_norm": 0.46645999381889575,
+      "learning_rate": 2.0517136430742994e-06,
+      "loss": 0.5629,
+      "step": 12639
+    },
+    {
+      "epoch": 3.355502455860879,
+      "grad_norm": 0.4667808182590336,
+      "learning_rate": 2.051370178266237e-06,
+      "loss": 0.5686,
+      "step": 12640
+    },
+    {
+      "epoch": 3.3557679543342624,
+      "grad_norm": 0.45526729876677424,
+      "learning_rate": 2.0510267222075545e-06,
+      "loss": 0.5579,
+      "step": 12641
+    },
+    {
+      "epoch": 3.3560334528076465,
+      "grad_norm": 0.4555619777967573,
+      "learning_rate": 2.0506832749049516e-06,
+      "loss": 0.5639,
+      "step": 12642
+    },
+    {
+      "epoch": 3.35629895128103,
+      "grad_norm": 0.46278641999002956,
+      "learning_rate": 2.050339836365124e-06,
+      "loss": 0.511,
+      "step": 12643
+    },
+    {
+      "epoch": 3.356564449754414,
+      "grad_norm": 0.45812681937415933,
+      "learning_rate": 2.04999640659477e-06,
+      "loss": 0.5345,
+      "step": 12644
+    },
+    {
+      "epoch": 3.356829948227798,
+      "grad_norm": 0.46301954097363396,
+      "learning_rate": 2.0496529856005884e-06,
+      "loss": 0.5303,
+      "step": 12645
+    },
+    {
+      "epoch": 3.3570954467011815,
+      "grad_norm": 0.4555211957160118,
+      "learning_rate": 2.049309573389276e-06,
+      "loss": 0.5484,
+      "step": 12646
+    },
+    {
+      "epoch": 3.357360945174565,
+      "grad_norm": 0.4602332130468546,
+      "learning_rate": 2.0489661699675304e-06,
+      "loss": 0.5513,
+      "step": 12647
+    },
+    {
+      "epoch": 3.3576264436479493,
+      "grad_norm": 0.46493012312978776,
+      "learning_rate": 2.0486227753420496e-06,
+      "loss": 0.5313,
+      "step": 12648
+    },
+    {
+      "epoch": 3.357891942121333,
+      "grad_norm": 0.47455733903475056,
+      "learning_rate": 2.048279389519529e-06,
+      "loss": 0.5778,
+      "step": 12649
+    },
+    {
+      "epoch": 3.3581574405947165,
+      "grad_norm": 0.4653231034782884,
+      "learning_rate": 2.0479360125066664e-06,
+      "loss": 0.553,
+      "step": 12650
+    },
+    {
+      "epoch": 3.3584229390681,
+      "grad_norm": 0.4707784105566937,
+      "learning_rate": 2.047592644310158e-06,
+      "loss": 0.5125,
+      "step": 12651
+    },
+    {
+      "epoch": 3.3586884375414843,
+      "grad_norm": 0.46782438989601033,
+      "learning_rate": 2.0472492849367017e-06,
+      "loss": 0.5552,
+      "step": 12652
+    },
+    {
+      "epoch": 3.358953936014868,
+      "grad_norm": 0.47938251783339464,
+      "learning_rate": 2.046905934392992e-06,
+      "loss": 0.572,
+      "step": 12653
+    },
+    {
+      "epoch": 3.3592194344882516,
+      "grad_norm": 0.46808079929520907,
+      "learning_rate": 2.0465625926857272e-06,
+      "loss": 0.5283,
+      "step": 12654
+    },
+    {
+      "epoch": 3.3594849329616356,
+      "grad_norm": 0.456763664122745,
+      "learning_rate": 2.0462192598216013e-06,
+      "loss": 0.5387,
+      "step": 12655
+    },
+    {
+      "epoch": 3.3597504314350193,
+      "grad_norm": 0.47898908520285965,
+      "learning_rate": 2.0458759358073103e-06,
+      "loss": 0.5386,
+      "step": 12656
+    },
+    {
+      "epoch": 3.360015929908403,
+      "grad_norm": 0.4572871051451324,
+      "learning_rate": 2.045532620649551e-06,
+      "loss": 0.5593,
+      "step": 12657
+    },
+    {
+      "epoch": 3.3602814283817866,
+      "grad_norm": 0.46092080626684123,
+      "learning_rate": 2.0451893143550185e-06,
+      "loss": 0.5111,
+      "step": 12658
+    },
+    {
+      "epoch": 3.3605469268551706,
+      "grad_norm": 0.45457161365711435,
+      "learning_rate": 2.0448460169304076e-06,
+      "loss": 0.543,
+      "step": 12659
+    },
+    {
+      "epoch": 3.3608124253285543,
+      "grad_norm": 0.4577665914247044,
+      "learning_rate": 2.044502728382414e-06,
+      "loss": 0.5103,
+      "step": 12660
+    },
+    {
+      "epoch": 3.361077923801938,
+      "grad_norm": 0.44672302320097573,
+      "learning_rate": 2.0441594487177322e-06,
+      "loss": 0.5203,
+      "step": 12661
+    },
+    {
+      "epoch": 3.361343422275322,
+      "grad_norm": 0.4652005993774999,
+      "learning_rate": 2.043816177943057e-06,
+      "loss": 0.6096,
+      "step": 12662
+    },
+    {
+      "epoch": 3.3616089207487057,
+      "grad_norm": 0.4709952848217192,
+      "learning_rate": 2.0434729160650835e-06,
+      "loss": 0.5161,
+      "step": 12663
+    },
+    {
+      "epoch": 3.3618744192220893,
+      "grad_norm": 0.4558251077918749,
+      "learning_rate": 2.043129663090506e-06,
+      "loss": 0.5787,
+      "step": 12664
+    },
+    {
+      "epoch": 3.3621399176954734,
+      "grad_norm": 0.4574013331667824,
+      "learning_rate": 2.0427864190260193e-06,
+      "loss": 0.5739,
+      "step": 12665
+    },
+    {
+      "epoch": 3.362405416168857,
+      "grad_norm": 0.4552401248377977,
+      "learning_rate": 2.042443183878316e-06,
+      "loss": 0.511,
+      "step": 12666
+    },
+    {
+      "epoch": 3.3626709146422407,
+      "grad_norm": 0.4342146830621328,
+      "learning_rate": 2.042099957654091e-06,
+      "loss": 0.5048,
+      "step": 12667
+    },
+    {
+      "epoch": 3.3629364131156247,
+      "grad_norm": 0.4675184512527628,
+      "learning_rate": 2.041756740360038e-06,
+      "loss": 0.5704,
+      "step": 12668
+    },
+    {
+      "epoch": 3.3632019115890084,
+      "grad_norm": 0.46599059307563045,
+      "learning_rate": 2.0414135320028507e-06,
+      "loss": 0.5539,
+      "step": 12669
+    },
+    {
+      "epoch": 3.363467410062392,
+      "grad_norm": 0.4545448916893775,
+      "learning_rate": 2.041070332589222e-06,
+      "loss": 0.559,
+      "step": 12670
+    },
+    {
+      "epoch": 3.363732908535776,
+      "grad_norm": 0.46387004846292806,
+      "learning_rate": 2.040727142125847e-06,
+      "loss": 0.547,
+      "step": 12671
+    },
+    {
+      "epoch": 3.3639984070091598,
+      "grad_norm": 0.46405523679722044,
+      "learning_rate": 2.040383960619416e-06,
+      "loss": 0.546,
+      "step": 12672
+    },
+    {
+      "epoch": 3.3642639054825434,
+      "grad_norm": 0.45406283070999814,
+      "learning_rate": 2.0400407880766227e-06,
+      "loss": 0.5355,
+      "step": 12673
+    },
+    {
+      "epoch": 3.3645294039559275,
+      "grad_norm": 0.4553928011584363,
+      "learning_rate": 2.0396976245041604e-06,
+      "loss": 0.5368,
+      "step": 12674
+    },
+    {
+      "epoch": 3.364794902429311,
+      "grad_norm": 0.46421618356922373,
+      "learning_rate": 2.0393544699087217e-06,
+      "loss": 0.592,
+      "step": 12675
+    },
+    {
+      "epoch": 3.3650604009026948,
+      "grad_norm": 0.4699202017390662,
+      "learning_rate": 2.0390113242969984e-06,
+      "loss": 0.5088,
+      "step": 12676
+    },
+    {
+      "epoch": 3.3653258993760784,
+      "grad_norm": 0.4523131168077975,
+      "learning_rate": 2.0386681876756837e-06,
+      "loss": 0.5257,
+      "step": 12677
+    },
+    {
+      "epoch": 3.3655913978494625,
+      "grad_norm": 0.4613211736289155,
+      "learning_rate": 2.0383250600514686e-06,
+      "loss": 0.5161,
+      "step": 12678
+    },
+    {
+      "epoch": 3.365856896322846,
+      "grad_norm": 0.46259182936929705,
+      "learning_rate": 2.0379819414310447e-06,
+      "loss": 0.5228,
+      "step": 12679
+    },
+    {
+      "epoch": 3.3661223947962298,
+      "grad_norm": 0.4657565528463685,
+      "learning_rate": 2.037638831821104e-06,
+      "loss": 0.5343,
+      "step": 12680
+    },
+    {
+      "epoch": 3.366387893269614,
+      "grad_norm": 0.4671720927539018,
+      "learning_rate": 2.0372957312283387e-06,
+      "loss": 0.5561,
+      "step": 12681
+    },
+    {
+      "epoch": 3.3666533917429975,
+      "grad_norm": 0.45716590645948557,
+      "learning_rate": 2.0369526396594405e-06,
+      "loss": 0.5658,
+      "step": 12682
+    },
+    {
+      "epoch": 3.366918890216381,
+      "grad_norm": 0.4521434708036079,
+      "learning_rate": 2.0366095571210986e-06,
+      "loss": 0.5668,
+      "step": 12683
+    },
+    {
+      "epoch": 3.367184388689765,
+      "grad_norm": 0.47364572366767593,
+      "learning_rate": 2.0362664836200044e-06,
+      "loss": 0.561,
+      "step": 12684
+    },
+    {
+      "epoch": 3.367449887163149,
+      "grad_norm": 0.4682103062912439,
+      "learning_rate": 2.0359234191628495e-06,
+      "loss": 0.5766,
+      "step": 12685
+    },
+    {
+      "epoch": 3.3677153856365325,
+      "grad_norm": 0.46041100668933177,
+      "learning_rate": 2.035580363756324e-06,
+      "loss": 0.5719,
+      "step": 12686
+    },
+    {
+      "epoch": 3.367980884109916,
+      "grad_norm": 0.466542172915299,
+      "learning_rate": 2.035237317407118e-06,
+      "loss": 0.5345,
+      "step": 12687
+    },
+    {
+      "epoch": 3.3682463825833002,
+      "grad_norm": 0.4708751311317681,
+      "learning_rate": 2.034894280121924e-06,
+      "loss": 0.5268,
+      "step": 12688
+    },
+    {
+      "epoch": 3.368511881056684,
+      "grad_norm": 0.4533753567975711,
+      "learning_rate": 2.0345512519074296e-06,
+      "loss": 0.5605,
+      "step": 12689
+    },
+    {
+      "epoch": 3.3687773795300675,
+      "grad_norm": 0.4615146989112833,
+      "learning_rate": 2.034208232770325e-06,
+      "loss": 0.5503,
+      "step": 12690
+    },
+    {
+      "epoch": 3.3690428780034516,
+      "grad_norm": 0.45615675051746146,
+      "learning_rate": 2.0338652227173e-06,
+      "loss": 0.5485,
+      "step": 12691
+    },
+    {
+      "epoch": 3.3693083764768352,
+      "grad_norm": 0.45547259338486745,
+      "learning_rate": 2.033522221755045e-06,
+      "loss": 0.5412,
+      "step": 12692
+    },
+    {
+      "epoch": 3.369573874950219,
+      "grad_norm": 0.4652872244193801,
+      "learning_rate": 2.0331792298902488e-06,
+      "loss": 0.5727,
+      "step": 12693
+    },
+    {
+      "epoch": 3.369839373423603,
+      "grad_norm": 0.45665773321693587,
+      "learning_rate": 2.032836247129601e-06,
+      "loss": 0.4726,
+      "step": 12694
+    },
+    {
+      "epoch": 3.3701048718969866,
+      "grad_norm": 0.46254817676681753,
+      "learning_rate": 2.032493273479789e-06,
+      "loss": 0.5431,
+      "step": 12695
+    },
+    {
+      "epoch": 3.3703703703703702,
+      "grad_norm": 0.43649196613551833,
+      "learning_rate": 2.0321503089475037e-06,
+      "loss": 0.5207,
+      "step": 12696
+    },
+    {
+      "epoch": 3.3706358688437543,
+      "grad_norm": 0.4513365164719909,
+      "learning_rate": 2.0318073535394326e-06,
+      "loss": 0.5042,
+      "step": 12697
+    },
+    {
+      "epoch": 3.370901367317138,
+      "grad_norm": 0.4541239707611342,
+      "learning_rate": 2.031464407262264e-06,
+      "loss": 0.5562,
+      "step": 12698
+    },
+    {
+      "epoch": 3.3711668657905216,
+      "grad_norm": 0.46062478475997776,
+      "learning_rate": 2.0311214701226873e-06,
+      "loss": 0.5481,
+      "step": 12699
+    },
+    {
+      "epoch": 3.3714323642639057,
+      "grad_norm": 0.4763213131042151,
+      "learning_rate": 2.0307785421273908e-06,
+      "loss": 0.5234,
+      "step": 12700
+    },
+    {
+      "epoch": 3.3716978627372893,
+      "grad_norm": 0.44533687605477784,
+      "learning_rate": 2.0304356232830604e-06,
+      "loss": 0.5436,
+      "step": 12701
+    },
+    {
+      "epoch": 3.371963361210673,
+      "grad_norm": 0.4540547204679213,
+      "learning_rate": 2.0300927135963853e-06,
+      "loss": 0.5645,
+      "step": 12702
+    },
+    {
+      "epoch": 3.372228859684057,
+      "grad_norm": 0.4632642732790654,
+      "learning_rate": 2.0297498130740523e-06,
+      "loss": 0.5833,
+      "step": 12703
+    },
+    {
+      "epoch": 3.3724943581574407,
+      "grad_norm": 0.4563815304405746,
+      "learning_rate": 2.0294069217227495e-06,
+      "loss": 0.544,
+      "step": 12704
+    },
+    {
+      "epoch": 3.3727598566308243,
+      "grad_norm": 0.4543192415455526,
+      "learning_rate": 2.0290640395491644e-06,
+      "loss": 0.5484,
+      "step": 12705
+    },
+    {
+      "epoch": 3.373025355104208,
+      "grad_norm": 0.4583226164977842,
+      "learning_rate": 2.0287211665599837e-06,
+      "loss": 0.565,
+      "step": 12706
+    },
+    {
+      "epoch": 3.373290853577592,
+      "grad_norm": 0.47179552237435135,
+      "learning_rate": 2.0283783027618936e-06,
+      "loss": 0.5475,
+      "step": 12707
+    },
+    {
+      "epoch": 3.3735563520509757,
+      "grad_norm": 0.462000262766302,
+      "learning_rate": 2.0280354481615814e-06,
+      "loss": 0.5676,
+      "step": 12708
+    },
+    {
+      "epoch": 3.3738218505243593,
+      "grad_norm": 0.45905549356288383,
+      "learning_rate": 2.0276926027657336e-06,
+      "loss": 0.549,
+      "step": 12709
+    },
+    {
+      "epoch": 3.3740873489977434,
+      "grad_norm": 0.4676680465265721,
+      "learning_rate": 2.0273497665810366e-06,
+      "loss": 0.5845,
+      "step": 12710
+    },
+    {
+      "epoch": 3.374352847471127,
+      "grad_norm": 0.445421986051473,
+      "learning_rate": 2.027006939614177e-06,
+      "loss": 0.5281,
+      "step": 12711
+    },
+    {
+      "epoch": 3.3746183459445107,
+      "grad_norm": 0.4502513787802284,
+      "learning_rate": 2.0266641218718394e-06,
+      "loss": 0.5023,
+      "step": 12712
+    },
+    {
+      "epoch": 3.3748838444178944,
+      "grad_norm": 0.468636401900213,
+      "learning_rate": 2.02632131336071e-06,
+      "loss": 0.5382,
+      "step": 12713
+    },
+    {
+      "epoch": 3.3751493428912784,
+      "grad_norm": 0.45251351388224287,
+      "learning_rate": 2.0259785140874747e-06,
+      "loss": 0.5582,
+      "step": 12714
+    },
+    {
+      "epoch": 3.375414841364662,
+      "grad_norm": 0.46789445242809474,
+      "learning_rate": 2.0256357240588194e-06,
+      "loss": 0.5398,
+      "step": 12715
+    },
+    {
+      "epoch": 3.3756803398380457,
+      "grad_norm": 0.450838309436373,
+      "learning_rate": 2.025292943281429e-06,
+      "loss": 0.5535,
+      "step": 12716
+    },
+    {
+      "epoch": 3.37594583831143,
+      "grad_norm": 0.45656990226275346,
+      "learning_rate": 2.0249501717619894e-06,
+      "loss": 0.5375,
+      "step": 12717
+    },
+    {
+      "epoch": 3.3762113367848134,
+      "grad_norm": 0.4682129890930129,
+      "learning_rate": 2.0246074095071836e-06,
+      "loss": 0.5473,
+      "step": 12718
+    },
+    {
+      "epoch": 3.376476835258197,
+      "grad_norm": 0.46941676804578764,
+      "learning_rate": 2.024264656523697e-06,
+      "loss": 0.5514,
+      "step": 12719
+    },
+    {
+      "epoch": 3.376742333731581,
+      "grad_norm": 0.45003823293324147,
+      "learning_rate": 2.023921912818215e-06,
+      "loss": 0.5221,
+      "step": 12720
+    },
+    {
+      "epoch": 3.377007832204965,
+      "grad_norm": 0.46461030662741104,
+      "learning_rate": 2.0235791783974207e-06,
+      "loss": 0.5739,
+      "step": 12721
+    },
+    {
+      "epoch": 3.3772733306783485,
+      "grad_norm": 0.45386025062765284,
+      "learning_rate": 2.0232364532679995e-06,
+      "loss": 0.5746,
+      "step": 12722
+    },
+    {
+      "epoch": 3.3775388291517325,
+      "grad_norm": 0.4789801949002753,
+      "learning_rate": 2.0228937374366347e-06,
+      "loss": 0.5576,
+      "step": 12723
+    },
+    {
+      "epoch": 3.377804327625116,
+      "grad_norm": 0.4702227259538851,
+      "learning_rate": 2.0225510309100098e-06,
+      "loss": 0.545,
+      "step": 12724
+    },
+    {
+      "epoch": 3.3780698260985,
+      "grad_norm": 0.447527878863501,
+      "learning_rate": 2.0222083336948088e-06,
+      "loss": 0.5546,
+      "step": 12725
+    },
+    {
+      "epoch": 3.378335324571884,
+      "grad_norm": 0.4563851337757835,
+      "learning_rate": 2.0218656457977156e-06,
+      "loss": 0.5307,
+      "step": 12726
+    },
+    {
+      "epoch": 3.3786008230452675,
+      "grad_norm": 0.4524985408492862,
+      "learning_rate": 2.0215229672254124e-06,
+      "loss": 0.5521,
+      "step": 12727
+    },
+    {
+      "epoch": 3.378866321518651,
+      "grad_norm": 0.4686207721898212,
+      "learning_rate": 2.0211802979845833e-06,
+      "loss": 0.5308,
+      "step": 12728
+    },
+    {
+      "epoch": 3.3791318199920353,
+      "grad_norm": 0.47140076214538823,
+      "learning_rate": 2.0208376380819116e-06,
+      "loss": 0.5636,
+      "step": 12729
+    },
+    {
+      "epoch": 3.379397318465419,
+      "grad_norm": 0.5092787635550697,
+      "learning_rate": 2.020494987524078e-06,
+      "loss": 0.5522,
+      "step": 12730
+    },
+    {
+      "epoch": 3.3796628169388025,
+      "grad_norm": 0.4540468814921841,
+      "learning_rate": 2.020152346317766e-06,
+      "loss": 0.5034,
+      "step": 12731
+    },
+    {
+      "epoch": 3.379928315412186,
+      "grad_norm": 0.451221713730231,
+      "learning_rate": 2.0198097144696577e-06,
+      "loss": 0.5932,
+      "step": 12732
+    },
+    {
+      "epoch": 3.3801938138855703,
+      "grad_norm": 0.4770972343878027,
+      "learning_rate": 2.0194670919864364e-06,
+      "loss": 0.5505,
+      "step": 12733
+    },
+    {
+      "epoch": 3.380459312358954,
+      "grad_norm": 0.45090680707980824,
+      "learning_rate": 2.0191244788747842e-06,
+      "loss": 0.5514,
+      "step": 12734
+    },
+    {
+      "epoch": 3.3807248108323376,
+      "grad_norm": 0.4529283941542656,
+      "learning_rate": 2.0187818751413813e-06,
+      "loss": 0.5554,
+      "step": 12735
+    },
+    {
+      "epoch": 3.3809903093057216,
+      "grad_norm": 0.493989687412165,
+      "learning_rate": 2.01843928079291e-06,
+      "loss": 0.5319,
+      "step": 12736
+    },
+    {
+      "epoch": 3.3812558077791053,
+      "grad_norm": 0.4522283620390422,
+      "learning_rate": 2.0180966958360514e-06,
+      "loss": 0.512,
+      "step": 12737
+    },
+    {
+      "epoch": 3.381521306252489,
+      "grad_norm": 0.4479473287332283,
+      "learning_rate": 2.0177541202774876e-06,
+      "loss": 0.5212,
+      "step": 12738
+    },
+    {
+      "epoch": 3.381786804725873,
+      "grad_norm": 0.45242665165070406,
+      "learning_rate": 2.017411554123899e-06,
+      "loss": 0.5232,
+      "step": 12739
+    },
+    {
+      "epoch": 3.3820523031992566,
+      "grad_norm": 0.45451541064152395,
+      "learning_rate": 2.017068997381967e-06,
+      "loss": 0.5362,
+      "step": 12740
+    },
+    {
+      "epoch": 3.3823178016726403,
+      "grad_norm": 0.4460571725812792,
+      "learning_rate": 2.016726450058372e-06,
+      "loss": 0.5329,
+      "step": 12741
+    },
+    {
+      "epoch": 3.382583300146024,
+      "grad_norm": 0.46601782009726894,
+      "learning_rate": 2.016383912159794e-06,
+      "loss": 0.5566,
+      "step": 12742
+    },
+    {
+      "epoch": 3.382848798619408,
+      "grad_norm": 0.4674835143262182,
+      "learning_rate": 2.016041383692914e-06,
+      "loss": 0.5326,
+      "step": 12743
+    },
+    {
+      "epoch": 3.3831142970927917,
+      "grad_norm": 0.4883092000371992,
+      "learning_rate": 2.015698864664412e-06,
+      "loss": 0.5201,
+      "step": 12744
+    },
+    {
+      "epoch": 3.3833797955661753,
+      "grad_norm": 0.45165210404600376,
+      "learning_rate": 2.015356355080968e-06,
+      "loss": 0.5413,
+      "step": 12745
+    },
+    {
+      "epoch": 3.3836452940395594,
+      "grad_norm": 0.4646914973137166,
+      "learning_rate": 2.0150138549492625e-06,
+      "loss": 0.5855,
+      "step": 12746
+    },
+    {
+      "epoch": 3.383910792512943,
+      "grad_norm": 0.4716558753972731,
+      "learning_rate": 2.014671364275973e-06,
+      "loss": 0.5386,
+      "step": 12747
+    },
+    {
+      "epoch": 3.3841762909863267,
+      "grad_norm": 0.46898554050258334,
+      "learning_rate": 2.01432888306778e-06,
+      "loss": 0.5506,
+      "step": 12748
+    },
+    {
+      "epoch": 3.3844417894597107,
+      "grad_norm": 0.4638898678583906,
+      "learning_rate": 2.013986411331364e-06,
+      "loss": 0.5352,
+      "step": 12749
+    },
+    {
+      "epoch": 3.3847072879330944,
+      "grad_norm": 0.4696325181739473,
+      "learning_rate": 2.0136439490734023e-06,
+      "loss": 0.5275,
+      "step": 12750
+    },
+    {
+      "epoch": 3.384972786406478,
+      "grad_norm": 0.4554793897498558,
+      "learning_rate": 2.013301496300574e-06,
+      "loss": 0.5815,
+      "step": 12751
+    },
+    {
+      "epoch": 3.385238284879862,
+      "grad_norm": 0.470266714263882,
+      "learning_rate": 2.0129590530195593e-06,
+      "loss": 0.5662,
+      "step": 12752
+    },
+    {
+      "epoch": 3.3855037833532458,
+      "grad_norm": 0.45596083357900896,
+      "learning_rate": 2.012616619237035e-06,
+      "loss": 0.5521,
+      "step": 12753
+    },
+    {
+      "epoch": 3.3857692818266294,
+      "grad_norm": 0.47509611512895283,
+      "learning_rate": 2.01227419495968e-06,
+      "loss": 0.5094,
+      "step": 12754
+    },
+    {
+      "epoch": 3.3860347803000135,
+      "grad_norm": 0.45012087086726,
+      "learning_rate": 2.011931780194172e-06,
+      "loss": 0.5612,
+      "step": 12755
+    },
+    {
+      "epoch": 3.386300278773397,
+      "grad_norm": 0.4643096228219624,
+      "learning_rate": 2.0115893749471892e-06,
+      "loss": 0.5589,
+      "step": 12756
+    },
+    {
+      "epoch": 3.3865657772467808,
+      "grad_norm": 0.4614065325401749,
+      "learning_rate": 2.01124697922541e-06,
+      "loss": 0.5475,
+      "step": 12757
+    },
+    {
+      "epoch": 3.386831275720165,
+      "grad_norm": 0.4499457641025427,
+      "learning_rate": 2.010904593035511e-06,
+      "loss": 0.5424,
+      "step": 12758
+    },
+    {
+      "epoch": 3.3870967741935485,
+      "grad_norm": 0.4625461139829754,
+      "learning_rate": 2.01056221638417e-06,
+      "loss": 0.5426,
+      "step": 12759
+    },
+    {
+      "epoch": 3.387362272666932,
+      "grad_norm": 0.4594882605342509,
+      "learning_rate": 2.010219849278064e-06,
+      "loss": 0.566,
+      "step": 12760
+    },
+    {
+      "epoch": 3.3876277711403158,
+      "grad_norm": 0.4737230055503393,
+      "learning_rate": 2.0098774917238708e-06,
+      "loss": 0.5632,
+      "step": 12761
+    },
+    {
+      "epoch": 3.3878932696137,
+      "grad_norm": 0.46136477287257904,
+      "learning_rate": 2.0095351437282655e-06,
+      "loss": 0.5266,
+      "step": 12762
+    },
+    {
+      "epoch": 3.3881587680870835,
+      "grad_norm": 0.47064068519465296,
+      "learning_rate": 2.009192805297927e-06,
+      "loss": 0.5458,
+      "step": 12763
+    },
+    {
+      "epoch": 3.388424266560467,
+      "grad_norm": 0.45879778255256043,
+      "learning_rate": 2.00885047643953e-06,
+      "loss": 0.5607,
+      "step": 12764
+    },
+    {
+      "epoch": 3.388689765033851,
+      "grad_norm": 0.4744306629926466,
+      "learning_rate": 2.008508157159751e-06,
+      "loss": 0.5084,
+      "step": 12765
+    },
+    {
+      "epoch": 3.388955263507235,
+      "grad_norm": 0.4688473828309847,
+      "learning_rate": 2.0081658474652663e-06,
+      "loss": 0.5483,
+      "step": 12766
+    },
+    {
+      "epoch": 3.3892207619806185,
+      "grad_norm": 0.44937018408371066,
+      "learning_rate": 2.007823547362752e-06,
+      "loss": 0.5586,
+      "step": 12767
+    },
+    {
+      "epoch": 3.389486260454002,
+      "grad_norm": 0.47422754117108673,
+      "learning_rate": 2.007481256858883e-06,
+      "loss": 0.5158,
+      "step": 12768
+    },
+    {
+      "epoch": 3.3897517589273862,
+      "grad_norm": 0.45521373201687426,
+      "learning_rate": 2.0071389759603366e-06,
+      "loss": 0.5453,
+      "step": 12769
+    },
+    {
+      "epoch": 3.39001725740077,
+      "grad_norm": 0.46347686684427203,
+      "learning_rate": 2.006796704673786e-06,
+      "loss": 0.5724,
+      "step": 12770
+    },
+    {
+      "epoch": 3.3902827558741535,
+      "grad_norm": 0.4615950752991866,
+      "learning_rate": 2.006454443005907e-06,
+      "loss": 0.5555,
+      "step": 12771
+    },
+    {
+      "epoch": 3.3905482543475376,
+      "grad_norm": 0.46261355357107664,
+      "learning_rate": 2.006112190963375e-06,
+      "loss": 0.5579,
+      "step": 12772
+    },
+    {
+      "epoch": 3.3908137528209212,
+      "grad_norm": 0.45761945006015803,
+      "learning_rate": 2.005769948552865e-06,
+      "loss": 0.5504,
+      "step": 12773
+    },
+    {
+      "epoch": 3.391079251294305,
+      "grad_norm": 0.45213388161587553,
+      "learning_rate": 2.0054277157810505e-06,
+      "loss": 0.5432,
+      "step": 12774
+    },
+    {
+      "epoch": 3.391344749767689,
+      "grad_norm": 0.4751857821962484,
+      "learning_rate": 2.0050854926546078e-06,
+      "loss": 0.5867,
+      "step": 12775
+    },
+    {
+      "epoch": 3.3916102482410726,
+      "grad_norm": 0.4623675886448786,
+      "learning_rate": 2.0047432791802084e-06,
+      "loss": 0.5486,
+      "step": 12776
+    },
+    {
+      "epoch": 3.3918757467144562,
+      "grad_norm": 0.45858716547291906,
+      "learning_rate": 2.004401075364527e-06,
+      "loss": 0.5366,
+      "step": 12777
+    },
+    {
+      "epoch": 3.3921412451878403,
+      "grad_norm": 0.4756204065848985,
+      "learning_rate": 2.004058881214239e-06,
+      "loss": 0.5279,
+      "step": 12778
+    },
+    {
+      "epoch": 3.392406743661224,
+      "grad_norm": 0.47153198509465694,
+      "learning_rate": 2.0037166967360174e-06,
+      "loss": 0.5312,
+      "step": 12779
+    },
+    {
+      "epoch": 3.3926722421346076,
+      "grad_norm": 0.4589978915018234,
+      "learning_rate": 2.003374521936536e-06,
+      "loss": 0.5728,
+      "step": 12780
+    },
+    {
+      "epoch": 3.3929377406079917,
+      "grad_norm": 0.4814911934067407,
+      "learning_rate": 2.0030323568224662e-06,
+      "loss": 0.5221,
+      "step": 12781
+    },
+    {
+      "epoch": 3.3932032390813753,
+      "grad_norm": 0.44575684633522467,
+      "learning_rate": 2.002690201400482e-06,
+      "loss": 0.4994,
+      "step": 12782
+    },
+    {
+      "epoch": 3.393468737554759,
+      "grad_norm": 0.4414049810759403,
+      "learning_rate": 2.002348055677257e-06,
+      "loss": 0.5841,
+      "step": 12783
+    },
+    {
+      "epoch": 3.393734236028143,
+      "grad_norm": 0.46338737339316516,
+      "learning_rate": 2.002005919659463e-06,
+      "loss": 0.5258,
+      "step": 12784
+    },
+    {
+      "epoch": 3.3939997345015267,
+      "grad_norm": 0.44667358661694545,
+      "learning_rate": 2.001663793353773e-06,
+      "loss": 0.5495,
+      "step": 12785
+    },
+    {
+      "epoch": 3.3942652329749103,
+      "grad_norm": 0.4560120278022204,
+      "learning_rate": 2.0013216767668597e-06,
+      "loss": 0.5649,
+      "step": 12786
+    },
+    {
+      "epoch": 3.394530731448294,
+      "grad_norm": 0.4622811354421234,
+      "learning_rate": 2.0009795699053942e-06,
+      "loss": 0.5595,
+      "step": 12787
+    },
+    {
+      "epoch": 3.394796229921678,
+      "grad_norm": 0.4496964910300942,
+      "learning_rate": 2.000637472776049e-06,
+      "loss": 0.5578,
+      "step": 12788
+    },
+    {
+      "epoch": 3.3950617283950617,
+      "grad_norm": 0.4568951263216965,
+      "learning_rate": 2.0002953853854955e-06,
+      "loss": 0.5776,
+      "step": 12789
+    },
+    {
+      "epoch": 3.3953272268684453,
+      "grad_norm": 0.45282303515490324,
+      "learning_rate": 1.9999533077404055e-06,
+      "loss": 0.5275,
+      "step": 12790
+    },
+    {
+      "epoch": 3.3955927253418294,
+      "grad_norm": 0.44016764025507116,
+      "learning_rate": 1.9996112398474506e-06,
+      "loss": 0.5463,
+      "step": 12791
+    },
+    {
+      "epoch": 3.395858223815213,
+      "grad_norm": 0.46519713526303824,
+      "learning_rate": 1.9992691817133025e-06,
+      "loss": 0.5531,
+      "step": 12792
+    },
+    {
+      "epoch": 3.3961237222885967,
+      "grad_norm": 0.45163241962601036,
+      "learning_rate": 1.9989271333446303e-06,
+      "loss": 0.5233,
+      "step": 12793
+    },
+    {
+      "epoch": 3.396389220761981,
+      "grad_norm": 0.46807815723645896,
+      "learning_rate": 1.998585094748106e-06,
+      "loss": 0.5694,
+      "step": 12794
+    },
+    {
+      "epoch": 3.3966547192353644,
+      "grad_norm": 0.4521198123980459,
+      "learning_rate": 1.9982430659304e-06,
+      "loss": 0.5744,
+      "step": 12795
+    },
+    {
+      "epoch": 3.396920217708748,
+      "grad_norm": 0.48158550107251835,
+      "learning_rate": 1.997901046898183e-06,
+      "loss": 0.5159,
+      "step": 12796
+    },
+    {
+      "epoch": 3.3971857161821317,
+      "grad_norm": 0.46534561546315134,
+      "learning_rate": 1.997559037658124e-06,
+      "loss": 0.5612,
+      "step": 12797
+    },
+    {
+      "epoch": 3.397451214655516,
+      "grad_norm": 0.46056984890636204,
+      "learning_rate": 1.9972170382168954e-06,
+      "loss": 0.573,
+      "step": 12798
+    },
+    {
+      "epoch": 3.3977167131288994,
+      "grad_norm": 0.47148293572806876,
+      "learning_rate": 1.9968750485811646e-06,
+      "loss": 0.5287,
+      "step": 12799
+    },
+    {
+      "epoch": 3.397982211602283,
+      "grad_norm": 0.48154856205884095,
+      "learning_rate": 1.9965330687576023e-06,
+      "loss": 0.5334,
+      "step": 12800
+    },
+    {
+      "epoch": 3.398247710075667,
+      "grad_norm": 0.47307298459509983,
+      "learning_rate": 1.9961910987528778e-06,
+      "loss": 0.5813,
+      "step": 12801
+    },
+    {
+      "epoch": 3.398513208549051,
+      "grad_norm": 0.43470676471261355,
+      "learning_rate": 1.9958491385736604e-06,
+      "loss": 0.5156,
+      "step": 12802
+    },
+    {
+      "epoch": 3.3987787070224345,
+      "grad_norm": 0.4686682255237146,
+      "learning_rate": 1.99550718822662e-06,
+      "loss": 0.5589,
+      "step": 12803
+    },
+    {
+      "epoch": 3.3990442054958185,
+      "grad_norm": 0.45497913772036674,
+      "learning_rate": 1.995165247718424e-06,
+      "loss": 0.5435,
+      "step": 12804
+    },
+    {
+      "epoch": 3.399309703969202,
+      "grad_norm": 0.4732442319763158,
+      "learning_rate": 1.9948233170557417e-06,
+      "loss": 0.5388,
+      "step": 12805
+    },
+    {
+      "epoch": 3.399575202442586,
+      "grad_norm": 0.45095023470538853,
+      "learning_rate": 1.9944813962452414e-06,
+      "loss": 0.5729,
+      "step": 12806
+    },
+    {
+      "epoch": 3.39984070091597,
+      "grad_norm": 0.47412774582553213,
+      "learning_rate": 1.9941394852935914e-06,
+      "loss": 0.5553,
+      "step": 12807
+    },
+    {
+      "epoch": 3.4001061993893535,
+      "grad_norm": 0.47183938419715915,
+      "learning_rate": 1.9937975842074604e-06,
+      "loss": 0.5535,
+      "step": 12808
+    },
+    {
+      "epoch": 3.400371697862737,
+      "grad_norm": 0.4680973187737429,
+      "learning_rate": 1.9934556929935174e-06,
+      "loss": 0.5124,
+      "step": 12809
+    },
+    {
+      "epoch": 3.4006371963361213,
+      "grad_norm": 0.4621455439858813,
+      "learning_rate": 1.993113811658427e-06,
+      "loss": 0.553,
+      "step": 12810
+    },
+    {
+      "epoch": 3.400902694809505,
+      "grad_norm": 0.4599085011440135,
+      "learning_rate": 1.9927719402088585e-06,
+      "loss": 0.5355,
+      "step": 12811
+    },
+    {
+      "epoch": 3.4011681932828886,
+      "grad_norm": 0.49478933307566275,
+      "learning_rate": 1.9924300786514787e-06,
+      "loss": 0.5294,
+      "step": 12812
+    },
+    {
+      "epoch": 3.4014336917562726,
+      "grad_norm": 0.4607392388405931,
+      "learning_rate": 1.9920882269929554e-06,
+      "loss": 0.5538,
+      "step": 12813
+    },
+    {
+      "epoch": 3.4016991902296563,
+      "grad_norm": 0.451448263256135,
+      "learning_rate": 1.991746385239955e-06,
+      "loss": 0.5277,
+      "step": 12814
+    },
+    {
+      "epoch": 3.40196468870304,
+      "grad_norm": 0.44729227497842544,
+      "learning_rate": 1.9914045533991456e-06,
+      "loss": 0.548,
+      "step": 12815
+    },
+    {
+      "epoch": 3.4022301871764236,
+      "grad_norm": 0.4564741795468326,
+      "learning_rate": 1.991062731477192e-06,
+      "loss": 0.5302,
+      "step": 12816
+    },
+    {
+      "epoch": 3.4024956856498076,
+      "grad_norm": 0.4610699670531906,
+      "learning_rate": 1.9907209194807607e-06,
+      "loss": 0.5071,
+      "step": 12817
+    },
+    {
+      "epoch": 3.4027611841231913,
+      "grad_norm": 0.4690518790762982,
+      "learning_rate": 1.9903791174165187e-06,
+      "loss": 0.5551,
+      "step": 12818
+    },
+    {
+      "epoch": 3.403026682596575,
+      "grad_norm": 0.4593606997471958,
+      "learning_rate": 1.9900373252911318e-06,
+      "loss": 0.5211,
+      "step": 12819
+    },
+    {
+      "epoch": 3.403292181069959,
+      "grad_norm": 0.4532941863254376,
+      "learning_rate": 1.9896955431112656e-06,
+      "loss": 0.5791,
+      "step": 12820
+    },
+    {
+      "epoch": 3.4035576795433427,
+      "grad_norm": 0.4720703055954754,
+      "learning_rate": 1.989353770883587e-06,
+      "loss": 0.5906,
+      "step": 12821
+    },
+    {
+      "epoch": 3.4038231780167263,
+      "grad_norm": 0.4624228327745245,
+      "learning_rate": 1.9890120086147586e-06,
+      "loss": 0.5592,
+      "step": 12822
+    },
+    {
+      "epoch": 3.40408867649011,
+      "grad_norm": 0.4717813246024548,
+      "learning_rate": 1.988670256311447e-06,
+      "loss": 0.5621,
+      "step": 12823
+    },
+    {
+      "epoch": 3.404354174963494,
+      "grad_norm": 0.4823129694044606,
+      "learning_rate": 1.9883285139803175e-06,
+      "loss": 0.5399,
+      "step": 12824
+    },
+    {
+      "epoch": 3.4046196734368777,
+      "grad_norm": 0.4545405613069321,
+      "learning_rate": 1.987986781628035e-06,
+      "loss": 0.5336,
+      "step": 12825
+    },
+    {
+      "epoch": 3.4048851719102613,
+      "grad_norm": 0.46806772701864774,
+      "learning_rate": 1.9876450592612643e-06,
+      "loss": 0.5493,
+      "step": 12826
+    },
+    {
+      "epoch": 3.4051506703836454,
+      "grad_norm": 0.46297188021680963,
+      "learning_rate": 1.98730334688667e-06,
+      "loss": 0.547,
+      "step": 12827
+    },
+    {
+      "epoch": 3.405416168857029,
+      "grad_norm": 0.46214696435760505,
+      "learning_rate": 1.9869616445109146e-06,
+      "loss": 0.5257,
+      "step": 12828
+    },
+    {
+      "epoch": 3.4056816673304127,
+      "grad_norm": 0.45961305683228143,
+      "learning_rate": 1.986619952140664e-06,
+      "loss": 0.5374,
+      "step": 12829
+    },
+    {
+      "epoch": 3.4059471658037968,
+      "grad_norm": 0.4539132220365548,
+      "learning_rate": 1.9862782697825807e-06,
+      "loss": 0.5046,
+      "step": 12830
+    },
+    {
+      "epoch": 3.4062126642771804,
+      "grad_norm": 0.45665071879455377,
+      "learning_rate": 1.9859365974433286e-06,
+      "loss": 0.5807,
+      "step": 12831
+    },
+    {
+      "epoch": 3.406478162750564,
+      "grad_norm": 0.47370034791747617,
+      "learning_rate": 1.9855949351295724e-06,
+      "loss": 0.522,
+      "step": 12832
+    },
+    {
+      "epoch": 3.406743661223948,
+      "grad_norm": 0.4650789210759116,
+      "learning_rate": 1.985253282847974e-06,
+      "loss": 0.575,
+      "step": 12833
+    },
+    {
+      "epoch": 3.4070091596973318,
+      "grad_norm": 0.45900766862452075,
+      "learning_rate": 1.9849116406051967e-06,
+      "loss": 0.5447,
+      "step": 12834
+    },
+    {
+      "epoch": 3.4072746581707154,
+      "grad_norm": 0.4765245161450273,
+      "learning_rate": 1.984570008407904e-06,
+      "loss": 0.5527,
+      "step": 12835
+    },
+    {
+      "epoch": 3.4075401566440995,
+      "grad_norm": 0.4503584108780535,
+      "learning_rate": 1.9842283862627575e-06,
+      "loss": 0.5766,
+      "step": 12836
+    },
+    {
+      "epoch": 3.407805655117483,
+      "grad_norm": 0.4500123549928934,
+      "learning_rate": 1.9838867741764207e-06,
+      "loss": 0.5438,
+      "step": 12837
+    },
+    {
+      "epoch": 3.4080711535908668,
+      "grad_norm": 0.44972832557463593,
+      "learning_rate": 1.9835451721555564e-06,
+      "loss": 0.5684,
+      "step": 12838
+    },
+    {
+      "epoch": 3.408336652064251,
+      "grad_norm": 0.4705877893456991,
+      "learning_rate": 1.983203580206825e-06,
+      "loss": 0.5453,
+      "step": 12839
+    },
+    {
+      "epoch": 3.4086021505376345,
+      "grad_norm": 0.46942107608376754,
+      "learning_rate": 1.9828619983368884e-06,
+      "loss": 0.5519,
+      "step": 12840
+    },
+    {
+      "epoch": 3.408867649011018,
+      "grad_norm": 0.4598823261174286,
+      "learning_rate": 1.9825204265524093e-06,
+      "loss": 0.5294,
+      "step": 12841
+    },
+    {
+      "epoch": 3.4091331474844018,
+      "grad_norm": 0.4518402010744265,
+      "learning_rate": 1.982178864860049e-06,
+      "loss": 0.5091,
+      "step": 12842
+    },
+    {
+      "epoch": 3.409398645957786,
+      "grad_norm": 0.4418022717976794,
+      "learning_rate": 1.9818373132664693e-06,
+      "loss": 0.5701,
+      "step": 12843
+    },
+    {
+      "epoch": 3.4096641444311695,
+      "grad_norm": 0.45969587376638854,
+      "learning_rate": 1.9814957717783305e-06,
+      "loss": 0.5976,
+      "step": 12844
+    },
+    {
+      "epoch": 3.409929642904553,
+      "grad_norm": 0.4529674799169325,
+      "learning_rate": 1.981154240402293e-06,
+      "loss": 0.5543,
+      "step": 12845
+    },
+    {
+      "epoch": 3.4101951413779372,
+      "grad_norm": 0.4615682052425449,
+      "learning_rate": 1.9808127191450186e-06,
+      "loss": 0.5519,
+      "step": 12846
+    },
+    {
+      "epoch": 3.410460639851321,
+      "grad_norm": 0.47708133755211074,
+      "learning_rate": 1.9804712080131672e-06,
+      "loss": 0.566,
+      "step": 12847
+    },
+    {
+      "epoch": 3.4107261383247045,
+      "grad_norm": 0.46859110158502626,
+      "learning_rate": 1.9801297070133994e-06,
+      "loss": 0.5742,
+      "step": 12848
+    },
+    {
+      "epoch": 3.4109916367980886,
+      "grad_norm": 0.4632276221339485,
+      "learning_rate": 1.979788216152375e-06,
+      "loss": 0.5521,
+      "step": 12849
+    },
+    {
+      "epoch": 3.4112571352714722,
+      "grad_norm": 0.4580371246156873,
+      "learning_rate": 1.9794467354367546e-06,
+      "loss": 0.5754,
+      "step": 12850
+    },
+    {
+      "epoch": 3.411522633744856,
+      "grad_norm": 0.4762097544937925,
+      "learning_rate": 1.979105264873197e-06,
+      "loss": 0.5409,
+      "step": 12851
+    },
+    {
+      "epoch": 3.4117881322182395,
+      "grad_norm": 0.4572323394054003,
+      "learning_rate": 1.9787638044683618e-06,
+      "loss": 0.5906,
+      "step": 12852
+    },
+    {
+      "epoch": 3.4120536306916236,
+      "grad_norm": 0.4591241225194067,
+      "learning_rate": 1.978422354228909e-06,
+      "loss": 0.5308,
+      "step": 12853
+    },
+    {
+      "epoch": 3.4123191291650072,
+      "grad_norm": 0.4731521598331077,
+      "learning_rate": 1.978080914161497e-06,
+      "loss": 0.5366,
+      "step": 12854
+    },
+    {
+      "epoch": 3.412584627638391,
+      "grad_norm": 0.47882979229162537,
+      "learning_rate": 1.9777394842727863e-06,
+      "loss": 0.4872,
+      "step": 12855
+    },
+    {
+      "epoch": 3.412850126111775,
+      "grad_norm": 0.46590549077603866,
+      "learning_rate": 1.9773980645694332e-06,
+      "loss": 0.5266,
+      "step": 12856
+    },
+    {
+      "epoch": 3.4131156245851586,
+      "grad_norm": 0.4614104090952774,
+      "learning_rate": 1.9770566550580977e-06,
+      "loss": 0.5015,
+      "step": 12857
+    },
+    {
+      "epoch": 3.4133811230585422,
+      "grad_norm": 0.45889354545369804,
+      "learning_rate": 1.9767152557454373e-06,
+      "loss": 0.5583,
+      "step": 12858
+    },
+    {
+      "epoch": 3.4136466215319263,
+      "grad_norm": 0.46555821475775977,
+      "learning_rate": 1.976373866638111e-06,
+      "loss": 0.5592,
+      "step": 12859
+    },
+    {
+      "epoch": 3.41391212000531,
+      "grad_norm": 0.45068964719599014,
+      "learning_rate": 1.9760324877427763e-06,
+      "loss": 0.539,
+      "step": 12860
+    },
+    {
+      "epoch": 3.4141776184786936,
+      "grad_norm": 0.45570304991383503,
+      "learning_rate": 1.9756911190660915e-06,
+      "loss": 0.5334,
+      "step": 12861
+    },
+    {
+      "epoch": 3.4144431169520777,
+      "grad_norm": 0.4501857328043723,
+      "learning_rate": 1.975349760614713e-06,
+      "loss": 0.5464,
+      "step": 12862
+    },
+    {
+      "epoch": 3.4147086154254613,
+      "grad_norm": 0.46148786629963523,
+      "learning_rate": 1.975008412395299e-06,
+      "loss": 0.5751,
+      "step": 12863
+    },
+    {
+      "epoch": 3.414974113898845,
+      "grad_norm": 0.4585975356195316,
+      "learning_rate": 1.974667074414506e-06,
+      "loss": 0.5191,
+      "step": 12864
+    },
+    {
+      "epoch": 3.415239612372229,
+      "grad_norm": 0.4510225928457239,
+      "learning_rate": 1.9743257466789917e-06,
+      "loss": 0.525,
+      "step": 12865
+    },
+    {
+      "epoch": 3.4155051108456127,
+      "grad_norm": 0.4654253341380489,
+      "learning_rate": 1.973984429195412e-06,
+      "loss": 0.5664,
+      "step": 12866
+    },
+    {
+      "epoch": 3.4157706093189963,
+      "grad_norm": 0.4694476702514629,
+      "learning_rate": 1.973643121970425e-06,
+      "loss": 0.5712,
+      "step": 12867
+    },
+    {
+      "epoch": 3.4160361077923804,
+      "grad_norm": 0.4722267490423635,
+      "learning_rate": 1.973301825010685e-06,
+      "loss": 0.5667,
+      "step": 12868
+    },
+    {
+      "epoch": 3.416301606265764,
+      "grad_norm": 0.47336771971303976,
+      "learning_rate": 1.9729605383228485e-06,
+      "loss": 0.5635,
+      "step": 12869
+    },
+    {
+      "epoch": 3.4165671047391477,
+      "grad_norm": 0.4704264060272286,
+      "learning_rate": 1.972619261913572e-06,
+      "loss": 0.5388,
+      "step": 12870
+    },
+    {
+      "epoch": 3.4168326032125313,
+      "grad_norm": 0.46361806422929314,
+      "learning_rate": 1.9722779957895114e-06,
+      "loss": 0.5522,
+      "step": 12871
+    },
+    {
+      "epoch": 3.4170981016859154,
+      "grad_norm": 0.4410717204039472,
+      "learning_rate": 1.971936739957322e-06,
+      "loss": 0.5249,
+      "step": 12872
+    },
+    {
+      "epoch": 3.417363600159299,
+      "grad_norm": 0.45280642979272157,
+      "learning_rate": 1.9715954944236603e-06,
+      "loss": 0.555,
+      "step": 12873
+    },
+    {
+      "epoch": 3.4176290986326827,
+      "grad_norm": 0.4587876387956169,
+      "learning_rate": 1.9712542591951787e-06,
+      "loss": 0.548,
+      "step": 12874
+    },
+    {
+      "epoch": 3.417894597106067,
+      "grad_norm": 0.44137031716794967,
+      "learning_rate": 1.970913034278534e-06,
+      "loss": 0.5423,
+      "step": 12875
+    },
+    {
+      "epoch": 3.4181600955794504,
+      "grad_norm": 0.45230503039464187,
+      "learning_rate": 1.9705718196803807e-06,
+      "loss": 0.5187,
+      "step": 12876
+    },
+    {
+      "epoch": 3.418425594052834,
+      "grad_norm": 0.4706104204206226,
+      "learning_rate": 1.970230615407373e-06,
+      "loss": 0.5585,
+      "step": 12877
+    },
+    {
+      "epoch": 3.418691092526218,
+      "grad_norm": 0.44932892089176757,
+      "learning_rate": 1.969889421466166e-06,
+      "loss": 0.5303,
+      "step": 12878
+    },
+    {
+      "epoch": 3.418956590999602,
+      "grad_norm": 0.4603762722031276,
+      "learning_rate": 1.9695482378634124e-06,
+      "loss": 0.5377,
+      "step": 12879
+    },
+    {
+      "epoch": 3.4192220894729854,
+      "grad_norm": 0.47030706162554686,
+      "learning_rate": 1.969207064605767e-06,
+      "loss": 0.5482,
+      "step": 12880
+    },
+    {
+      "epoch": 3.419487587946369,
+      "grad_norm": 0.46159296769742875,
+      "learning_rate": 1.9688659016998836e-06,
+      "loss": 0.5831,
+      "step": 12881
+    },
+    {
+      "epoch": 3.419753086419753,
+      "grad_norm": 0.4910339424156637,
+      "learning_rate": 1.9685247491524157e-06,
+      "loss": 0.5573,
+      "step": 12882
+    },
+    {
+      "epoch": 3.420018584893137,
+      "grad_norm": 0.44733412723024585,
+      "learning_rate": 1.9681836069700163e-06,
+      "loss": 0.5749,
+      "step": 12883
+    },
+    {
+      "epoch": 3.4202840833665205,
+      "grad_norm": 0.4471894912424108,
+      "learning_rate": 1.9678424751593397e-06,
+      "loss": 0.5334,
+      "step": 12884
+    },
+    {
+      "epoch": 3.4205495818399045,
+      "grad_norm": 0.4644009215745443,
+      "learning_rate": 1.9675013537270367e-06,
+      "loss": 0.5209,
+      "step": 12885
+    },
+    {
+      "epoch": 3.420815080313288,
+      "grad_norm": 0.45148650407953794,
+      "learning_rate": 1.9671602426797613e-06,
+      "loss": 0.5392,
+      "step": 12886
+    },
+    {
+      "epoch": 3.421080578786672,
+      "grad_norm": 0.4612184389492654,
+      "learning_rate": 1.9668191420241656e-06,
+      "loss": 0.5413,
+      "step": 12887
+    },
+    {
+      "epoch": 3.421346077260056,
+      "grad_norm": 0.45381529045092095,
+      "learning_rate": 1.9664780517669018e-06,
+      "loss": 0.5318,
+      "step": 12888
+    },
+    {
+      "epoch": 3.4216115757334395,
+      "grad_norm": 0.46233706778645767,
+      "learning_rate": 1.9661369719146224e-06,
+      "loss": 0.6002,
+      "step": 12889
+    },
+    {
+      "epoch": 3.421877074206823,
+      "grad_norm": 0.44544983775776853,
+      "learning_rate": 1.96579590247398e-06,
+      "loss": 0.5177,
+      "step": 12890
+    },
+    {
+      "epoch": 3.4221425726802073,
+      "grad_norm": 0.4582796123374127,
+      "learning_rate": 1.965454843451625e-06,
+      "loss": 0.5341,
+      "step": 12891
+    },
+    {
+      "epoch": 3.422408071153591,
+      "grad_norm": 0.4605727241122683,
+      "learning_rate": 1.9651137948542093e-06,
+      "loss": 0.5346,
+      "step": 12892
+    },
+    {
+      "epoch": 3.4226735696269746,
+      "grad_norm": 0.4624770345395566,
+      "learning_rate": 1.9647727566883838e-06,
+      "loss": 0.5203,
+      "step": 12893
+    },
+    {
+      "epoch": 3.4229390681003586,
+      "grad_norm": 0.45980284378791414,
+      "learning_rate": 1.9644317289608006e-06,
+      "loss": 0.5474,
+      "step": 12894
+    },
+    {
+      "epoch": 3.4232045665737423,
+      "grad_norm": 0.46819796216641063,
+      "learning_rate": 1.96409071167811e-06,
+      "loss": 0.5601,
+      "step": 12895
+    },
+    {
+      "epoch": 3.423470065047126,
+      "grad_norm": 0.4523077971838003,
+      "learning_rate": 1.9637497048469625e-06,
+      "loss": 0.5344,
+      "step": 12896
+    },
+    {
+      "epoch": 3.42373556352051,
+      "grad_norm": 0.45830761716803503,
+      "learning_rate": 1.9634087084740085e-06,
+      "loss": 0.545,
+      "step": 12897
+    },
+    {
+      "epoch": 3.4240010619938936,
+      "grad_norm": 0.45830484494469326,
+      "learning_rate": 1.9630677225658984e-06,
+      "loss": 0.5344,
+      "step": 12898
+    },
+    {
+      "epoch": 3.4242665604672773,
+      "grad_norm": 0.46134814845585176,
+      "learning_rate": 1.9627267471292826e-06,
+      "loss": 0.5385,
+      "step": 12899
+    },
+    {
+      "epoch": 3.424532058940661,
+      "grad_norm": 0.45729474976246104,
+      "learning_rate": 1.9623857821708105e-06,
+      "loss": 0.5267,
+      "step": 12900
+    },
+    {
+      "epoch": 3.424797557414045,
+      "grad_norm": 0.46117831968806905,
+      "learning_rate": 1.962044827697133e-06,
+      "loss": 0.5662,
+      "step": 12901
+    },
+    {
+      "epoch": 3.4250630558874287,
+      "grad_norm": 0.4641718812991612,
+      "learning_rate": 1.961703883714898e-06,
+      "loss": 0.5484,
+      "step": 12902
+    },
+    {
+      "epoch": 3.4253285543608123,
+      "grad_norm": 0.45924887664451797,
+      "learning_rate": 1.9613629502307547e-06,
+      "loss": 0.5623,
+      "step": 12903
+    },
+    {
+      "epoch": 3.4255940528341964,
+      "grad_norm": 0.47499072665285175,
+      "learning_rate": 1.9610220272513527e-06,
+      "loss": 0.5635,
+      "step": 12904
+    },
+    {
+      "epoch": 3.42585955130758,
+      "grad_norm": 0.4596436615054621,
+      "learning_rate": 1.9606811147833408e-06,
+      "loss": 0.5324,
+      "step": 12905
+    },
+    {
+      "epoch": 3.4261250497809637,
+      "grad_norm": 0.4579141422195126,
+      "learning_rate": 1.9603402128333676e-06,
+      "loss": 0.5477,
+      "step": 12906
+    },
+    {
+      "epoch": 3.4263905482543473,
+      "grad_norm": 0.4663383945138551,
+      "learning_rate": 1.959999321408082e-06,
+      "loss": 0.5543,
+      "step": 12907
+    },
+    {
+      "epoch": 3.4266560467277314,
+      "grad_norm": 0.46575955643248895,
+      "learning_rate": 1.9596584405141317e-06,
+      "loss": 0.5628,
+      "step": 12908
+    },
+    {
+      "epoch": 3.426921545201115,
+      "grad_norm": 0.46297050723337135,
+      "learning_rate": 1.959317570158165e-06,
+      "loss": 0.5413,
+      "step": 12909
+    },
+    {
+      "epoch": 3.4271870436744987,
+      "grad_norm": 0.46136492395227163,
+      "learning_rate": 1.958976710346829e-06,
+      "loss": 0.5624,
+      "step": 12910
+    },
+    {
+      "epoch": 3.4274525421478828,
+      "grad_norm": 0.47160562951935325,
+      "learning_rate": 1.9586358610867715e-06,
+      "loss": 0.5293,
+      "step": 12911
+    },
+    {
+      "epoch": 3.4277180406212664,
+      "grad_norm": 0.4485740460042685,
+      "learning_rate": 1.9582950223846405e-06,
+      "loss": 0.5745,
+      "step": 12912
+    },
+    {
+      "epoch": 3.42798353909465,
+      "grad_norm": 0.4658432484885228,
+      "learning_rate": 1.9579541942470844e-06,
+      "loss": 0.5131,
+      "step": 12913
+    },
+    {
+      "epoch": 3.428249037568034,
+      "grad_norm": 0.45407041267032655,
+      "learning_rate": 1.9576133766807463e-06,
+      "loss": 0.529,
+      "step": 12914
+    },
+    {
+      "epoch": 3.4285145360414178,
+      "grad_norm": 0.4725532226115244,
+      "learning_rate": 1.957272569692276e-06,
+      "loss": 0.57,
+      "step": 12915
+    },
+    {
+      "epoch": 3.4287800345148014,
+      "grad_norm": 0.46011816212871426,
+      "learning_rate": 1.95693177328832e-06,
+      "loss": 0.5403,
+      "step": 12916
+    },
+    {
+      "epoch": 3.4290455329881855,
+      "grad_norm": 0.4468210798597445,
+      "learning_rate": 1.956590987475524e-06,
+      "loss": 0.557,
+      "step": 12917
+    },
+    {
+      "epoch": 3.429311031461569,
+      "grad_norm": 0.4625799592730683,
+      "learning_rate": 1.956250212260534e-06,
+      "loss": 0.5736,
+      "step": 12918
+    },
+    {
+      "epoch": 3.4295765299349528,
+      "grad_norm": 0.4549201024386158,
+      "learning_rate": 1.9559094476499973e-06,
+      "loss": 0.5381,
+      "step": 12919
+    },
+    {
+      "epoch": 3.429842028408337,
+      "grad_norm": 0.4553946652009198,
+      "learning_rate": 1.955568693650557e-06,
+      "loss": 0.5448,
+      "step": 12920
+    },
+    {
+      "epoch": 3.4301075268817205,
+      "grad_norm": 0.45028598328021646,
+      "learning_rate": 1.955227950268861e-06,
+      "loss": 0.5606,
+      "step": 12921
+    },
+    {
+      "epoch": 3.430373025355104,
+      "grad_norm": 0.4461221682248535,
+      "learning_rate": 1.954887217511553e-06,
+      "loss": 0.5507,
+      "step": 12922
+    },
+    {
+      "epoch": 3.430638523828488,
+      "grad_norm": 0.4574590538434611,
+      "learning_rate": 1.9545464953852796e-06,
+      "loss": 0.5421,
+      "step": 12923
+    },
+    {
+      "epoch": 3.430904022301872,
+      "grad_norm": 0.46589754237031117,
+      "learning_rate": 1.9542057838966854e-06,
+      "loss": 0.5535,
+      "step": 12924
+    },
+    {
+      "epoch": 3.4311695207752555,
+      "grad_norm": 0.44298247358947884,
+      "learning_rate": 1.953865083052414e-06,
+      "loss": 0.4863,
+      "step": 12925
+    },
+    {
+      "epoch": 3.431435019248639,
+      "grad_norm": 0.46628421302932604,
+      "learning_rate": 1.9535243928591108e-06,
+      "loss": 0.5141,
+      "step": 12926
+    },
+    {
+      "epoch": 3.4317005177220232,
+      "grad_norm": 0.45992650915481853,
+      "learning_rate": 1.9531837133234195e-06,
+      "loss": 0.5544,
+      "step": 12927
+    },
+    {
+      "epoch": 3.431966016195407,
+      "grad_norm": 0.4592860348279172,
+      "learning_rate": 1.9528430444519853e-06,
+      "loss": 0.5605,
+      "step": 12928
+    },
+    {
+      "epoch": 3.4322315146687905,
+      "grad_norm": 0.4607819633495851,
+      "learning_rate": 1.952502386251451e-06,
+      "loss": 0.5341,
+      "step": 12929
+    },
+    {
+      "epoch": 3.4324970131421746,
+      "grad_norm": 0.461864596051747,
+      "learning_rate": 1.952161738728462e-06,
+      "loss": 0.5465,
+      "step": 12930
+    },
+    {
+      "epoch": 3.4327625116155582,
+      "grad_norm": 0.45195735747783655,
+      "learning_rate": 1.9518211018896593e-06,
+      "loss": 0.5769,
+      "step": 12931
+    },
+    {
+      "epoch": 3.433028010088942,
+      "grad_norm": 0.479496426249268,
+      "learning_rate": 1.9514804757416873e-06,
+      "loss": 0.5449,
+      "step": 12932
+    },
+    {
+      "epoch": 3.433293508562326,
+      "grad_norm": 0.4517342348341856,
+      "learning_rate": 1.9511398602911888e-06,
+      "loss": 0.5433,
+      "step": 12933
+    },
+    {
+      "epoch": 3.4335590070357096,
+      "grad_norm": 0.46119555680016766,
+      "learning_rate": 1.950799255544807e-06,
+      "loss": 0.5689,
+      "step": 12934
+    },
+    {
+      "epoch": 3.4338245055090932,
+      "grad_norm": 0.4691819288154536,
+      "learning_rate": 1.9504586615091838e-06,
+      "loss": 0.5724,
+      "step": 12935
+    },
+    {
+      "epoch": 3.434090003982477,
+      "grad_norm": 0.4742284134038077,
+      "learning_rate": 1.950118078190964e-06,
+      "loss": 0.5507,
+      "step": 12936
+    },
+    {
+      "epoch": 3.434355502455861,
+      "grad_norm": 0.4628072947168689,
+      "learning_rate": 1.9497775055967868e-06,
+      "loss": 0.5636,
+      "step": 12937
+    },
+    {
+      "epoch": 3.4346210009292446,
+      "grad_norm": 0.44871096409521455,
+      "learning_rate": 1.9494369437332957e-06,
+      "loss": 0.6042,
+      "step": 12938
+    },
+    {
+      "epoch": 3.4348864994026282,
+      "grad_norm": 0.45240624036476185,
+      "learning_rate": 1.9490963926071318e-06,
+      "loss": 0.5242,
+      "step": 12939
+    },
+    {
+      "epoch": 3.4351519978760123,
+      "grad_norm": 0.45733864780703987,
+      "learning_rate": 1.948755852224937e-06,
+      "loss": 0.5379,
+      "step": 12940
+    },
+    {
+      "epoch": 3.435417496349396,
+      "grad_norm": 0.44943786161537186,
+      "learning_rate": 1.9484153225933535e-06,
+      "loss": 0.5392,
+      "step": 12941
+    },
+    {
+      "epoch": 3.4356829948227796,
+      "grad_norm": 0.4613805516585124,
+      "learning_rate": 1.9480748037190215e-06,
+      "loss": 0.5409,
+      "step": 12942
+    },
+    {
+      "epoch": 3.4359484932961637,
+      "grad_norm": 0.4488877160524104,
+      "learning_rate": 1.947734295608582e-06,
+      "loss": 0.5633,
+      "step": 12943
+    },
+    {
+      "epoch": 3.4362139917695473,
+      "grad_norm": 0.4856772261979697,
+      "learning_rate": 1.947393798268676e-06,
+      "loss": 0.5423,
+      "step": 12944
+    },
+    {
+      "epoch": 3.436479490242931,
+      "grad_norm": 0.4620520066414312,
+      "learning_rate": 1.9470533117059436e-06,
+      "loss": 0.5693,
+      "step": 12945
+    },
+    {
+      "epoch": 3.436744988716315,
+      "grad_norm": 0.4622929628487836,
+      "learning_rate": 1.9467128359270255e-06,
+      "loss": 0.5579,
+      "step": 12946
+    },
+    {
+      "epoch": 3.4370104871896987,
+      "grad_norm": 0.4635115627694077,
+      "learning_rate": 1.9463723709385617e-06,
+      "loss": 0.5498,
+      "step": 12947
+    },
+    {
+      "epoch": 3.4372759856630823,
+      "grad_norm": 0.46035499387710876,
+      "learning_rate": 1.9460319167471934e-06,
+      "loss": 0.5891,
+      "step": 12948
+    },
+    {
+      "epoch": 3.4375414841364664,
+      "grad_norm": 0.46650348115065166,
+      "learning_rate": 1.9456914733595584e-06,
+      "loss": 0.5179,
+      "step": 12949
+    },
+    {
+      "epoch": 3.43780698260985,
+      "grad_norm": 0.45488514103637406,
+      "learning_rate": 1.9453510407822967e-06,
+      "loss": 0.5342,
+      "step": 12950
+    },
+    {
+      "epoch": 3.4380724810832337,
+      "grad_norm": 0.46553115481654656,
+      "learning_rate": 1.9450106190220475e-06,
+      "loss": 0.5726,
+      "step": 12951
+    },
+    {
+      "epoch": 3.438337979556618,
+      "grad_norm": 0.4700238232533625,
+      "learning_rate": 1.94467020808545e-06,
+      "loss": 0.564,
+      "step": 12952
+    },
+    {
+      "epoch": 3.4386034780300014,
+      "grad_norm": 0.45134818803427124,
+      "learning_rate": 1.944329807979144e-06,
+      "loss": 0.5195,
+      "step": 12953
+    },
+    {
+      "epoch": 3.438868976503385,
+      "grad_norm": 0.46680281992986916,
+      "learning_rate": 1.9439894187097665e-06,
+      "loss": 0.558,
+      "step": 12954
+    },
+    {
+      "epoch": 3.4391344749767687,
+      "grad_norm": 0.469956051572363,
+      "learning_rate": 1.9436490402839563e-06,
+      "loss": 0.5668,
+      "step": 12955
+    },
+    {
+      "epoch": 3.439399973450153,
+      "grad_norm": 0.4542521471498369,
+      "learning_rate": 1.9433086727083527e-06,
+      "loss": 0.5394,
+      "step": 12956
+    },
+    {
+      "epoch": 3.4396654719235364,
+      "grad_norm": 0.45752943744675856,
+      "learning_rate": 1.9429683159895926e-06,
+      "loss": 0.5587,
+      "step": 12957
+    },
+    {
+      "epoch": 3.43993097039692,
+      "grad_norm": 0.4741053283748696,
+      "learning_rate": 1.9426279701343143e-06,
+      "loss": 0.5798,
+      "step": 12958
+    },
+    {
+      "epoch": 3.440196468870304,
+      "grad_norm": 0.455475030055455,
+      "learning_rate": 1.9422876351491563e-06,
+      "loss": 0.5558,
+      "step": 12959
+    },
+    {
+      "epoch": 3.440461967343688,
+      "grad_norm": 0.46591850052394224,
+      "learning_rate": 1.9419473110407535e-06,
+      "loss": 0.554,
+      "step": 12960
+    },
+    {
+      "epoch": 3.4407274658170715,
+      "grad_norm": 0.45612183076371154,
+      "learning_rate": 1.9416069978157444e-06,
+      "loss": 0.5206,
+      "step": 12961
+    },
+    {
+      "epoch": 3.440992964290455,
+      "grad_norm": 0.47201723229335385,
+      "learning_rate": 1.941266695480767e-06,
+      "loss": 0.5328,
+      "step": 12962
+    },
+    {
+      "epoch": 3.441258462763839,
+      "grad_norm": 0.44387207177367005,
+      "learning_rate": 1.9409264040424562e-06,
+      "loss": 0.5684,
+      "step": 12963
+    },
+    {
+      "epoch": 3.441523961237223,
+      "grad_norm": 0.44364764654564254,
+      "learning_rate": 1.9405861235074497e-06,
+      "loss": 0.5585,
+      "step": 12964
+    },
+    {
+      "epoch": 3.4417894597106065,
+      "grad_norm": 0.4776868517754954,
+      "learning_rate": 1.9402458538823846e-06,
+      "loss": 0.5169,
+      "step": 12965
+    },
+    {
+      "epoch": 3.4420549581839905,
+      "grad_norm": 0.44619152084240254,
+      "learning_rate": 1.9399055951738945e-06,
+      "loss": 0.5228,
+      "step": 12966
+    },
+    {
+      "epoch": 3.442320456657374,
+      "grad_norm": 0.4524314091409978,
+      "learning_rate": 1.939565347388617e-06,
+      "loss": 0.5239,
+      "step": 12967
+    },
+    {
+      "epoch": 3.442585955130758,
+      "grad_norm": 0.45549877485236334,
+      "learning_rate": 1.939225110533187e-06,
+      "loss": 0.5657,
+      "step": 12968
+    },
+    {
+      "epoch": 3.442851453604142,
+      "grad_norm": 0.46692350590442394,
+      "learning_rate": 1.938884884614241e-06,
+      "loss": 0.5609,
+      "step": 12969
+    },
+    {
+      "epoch": 3.4431169520775255,
+      "grad_norm": 0.4802941260261063,
+      "learning_rate": 1.938544669638413e-06,
+      "loss": 0.5422,
+      "step": 12970
+    },
+    {
+      "epoch": 3.443382450550909,
+      "grad_norm": 0.45933349881384117,
+      "learning_rate": 1.9382044656123395e-06,
+      "loss": 0.5698,
+      "step": 12971
+    },
+    {
+      "epoch": 3.4436479490242933,
+      "grad_norm": 0.4785034673748016,
+      "learning_rate": 1.937864272542654e-06,
+      "loss": 0.5752,
+      "step": 12972
+    },
+    {
+      "epoch": 3.443913447497677,
+      "grad_norm": 0.47396198014458224,
+      "learning_rate": 1.937524090435991e-06,
+      "loss": 0.5533,
+      "step": 12973
+    },
+    {
+      "epoch": 3.4441789459710606,
+      "grad_norm": 0.46184131634312225,
+      "learning_rate": 1.937183919298986e-06,
+      "loss": 0.5911,
+      "step": 12974
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 0.4636070941821404,
+      "learning_rate": 1.936843759138272e-06,
+      "loss": 0.5669,
+      "step": 12975
+    },
+    {
+      "epoch": 3.4447099429178283,
+      "grad_norm": 0.4625025838055727,
+      "learning_rate": 1.9365036099604853e-06,
+      "loss": 0.555,
+      "step": 12976
+    },
+    {
+      "epoch": 3.444975441391212,
+      "grad_norm": 0.4637125271824205,
+      "learning_rate": 1.9361634717722564e-06,
+      "loss": 0.5129,
+      "step": 12977
+    },
+    {
+      "epoch": 3.445240939864596,
+      "grad_norm": 0.45672742797316584,
+      "learning_rate": 1.9358233445802207e-06,
+      "loss": 0.5504,
+      "step": 12978
+    },
+    {
+      "epoch": 3.4455064383379796,
+      "grad_norm": 0.47064989311627237,
+      "learning_rate": 1.9354832283910104e-06,
+      "loss": 0.5462,
+      "step": 12979
+    },
+    {
+      "epoch": 3.4457719368113633,
+      "grad_norm": 0.4597680371772616,
+      "learning_rate": 1.9351431232112595e-06,
+      "loss": 0.5383,
+      "step": 12980
+    },
+    {
+      "epoch": 3.446037435284747,
+      "grad_norm": 0.4489850745812362,
+      "learning_rate": 1.9348030290476007e-06,
+      "loss": 0.5017,
+      "step": 12981
+    },
+    {
+      "epoch": 3.446302933758131,
+      "grad_norm": 0.4747827448856693,
+      "learning_rate": 1.934462945906668e-06,
+      "loss": 0.5641,
+      "step": 12982
+    },
+    {
+      "epoch": 3.4465684322315147,
+      "grad_norm": 0.45141052289917666,
+      "learning_rate": 1.9341228737950913e-06,
+      "loss": 0.5332,
+      "step": 12983
+    },
+    {
+      "epoch": 3.4468339307048983,
+      "grad_norm": 0.47190703453491,
+      "learning_rate": 1.933782812719504e-06,
+      "loss": 0.5269,
+      "step": 12984
+    },
+    {
+      "epoch": 3.4470994291782824,
+      "grad_norm": 0.4565627979082249,
+      "learning_rate": 1.9334427626865383e-06,
+      "loss": 0.5484,
+      "step": 12985
+    },
+    {
+      "epoch": 3.447364927651666,
+      "grad_norm": 0.46595123566203817,
+      "learning_rate": 1.933102723702826e-06,
+      "loss": 0.5344,
+      "step": 12986
+    },
+    {
+      "epoch": 3.4476304261250497,
+      "grad_norm": 0.4510514834444204,
+      "learning_rate": 1.9327626957749986e-06,
+      "loss": 0.5501,
+      "step": 12987
+    },
+    {
+      "epoch": 3.4478959245984337,
+      "grad_norm": 0.45489740136255896,
+      "learning_rate": 1.932422678909688e-06,
+      "loss": 0.5627,
+      "step": 12988
+    },
+    {
+      "epoch": 3.4481614230718174,
+      "grad_norm": 0.45673515331617426,
+      "learning_rate": 1.9320826731135243e-06,
+      "loss": 0.543,
+      "step": 12989
+    },
+    {
+      "epoch": 3.448426921545201,
+      "grad_norm": 0.4640846262798768,
+      "learning_rate": 1.931742678393139e-06,
+      "loss": 0.5384,
+      "step": 12990
+    },
+    {
+      "epoch": 3.4486924200185847,
+      "grad_norm": 0.4542670857082774,
+      "learning_rate": 1.9314026947551626e-06,
+      "loss": 0.5359,
+      "step": 12991
+    },
+    {
+      "epoch": 3.4489579184919688,
+      "grad_norm": 0.46515483064203755,
+      "learning_rate": 1.931062722206226e-06,
+      "loss": 0.566,
+      "step": 12992
+    },
+    {
+      "epoch": 3.4492234169653524,
+      "grad_norm": 0.4604447888085944,
+      "learning_rate": 1.9307227607529593e-06,
+      "loss": 0.5551,
+      "step": 12993
+    },
+    {
+      "epoch": 3.449488915438736,
+      "grad_norm": 0.4446424993801546,
+      "learning_rate": 1.9303828104019937e-06,
+      "loss": 0.5346,
+      "step": 12994
+    },
+    {
+      "epoch": 3.44975441391212,
+      "grad_norm": 0.4484489012863962,
+      "learning_rate": 1.9300428711599566e-06,
+      "loss": 0.5352,
+      "step": 12995
+    },
+    {
+      "epoch": 3.4500199123855038,
+      "grad_norm": 0.46231579352536706,
+      "learning_rate": 1.9297029430334795e-06,
+      "loss": 0.532,
+      "step": 12996
+    },
+    {
+      "epoch": 3.4502854108588874,
+      "grad_norm": 0.4559535463049683,
+      "learning_rate": 1.929363026029191e-06,
+      "loss": 0.5567,
+      "step": 12997
+    },
+    {
+      "epoch": 3.4505509093322715,
+      "grad_norm": 0.46000416486733225,
+      "learning_rate": 1.929023120153721e-06,
+      "loss": 0.5744,
+      "step": 12998
+    },
+    {
+      "epoch": 3.450816407805655,
+      "grad_norm": 0.47586689951210737,
+      "learning_rate": 1.928683225413698e-06,
+      "loss": 0.5478,
+      "step": 12999
+    },
+    {
+      "epoch": 3.4510819062790388,
+      "grad_norm": 0.46517710338815815,
+      "learning_rate": 1.928343341815751e-06,
+      "loss": 0.5794,
+      "step": 13000
+    },
+    {
+      "epoch": 3.451347404752423,
+      "grad_norm": 0.46423245458865164,
+      "learning_rate": 1.928003469366508e-06,
+      "loss": 0.5449,
+      "step": 13001
+    },
+    {
+      "epoch": 3.4516129032258065,
+      "grad_norm": 0.4565970700937282,
+      "learning_rate": 1.9276636080725982e-06,
+      "loss": 0.535,
+      "step": 13002
+    },
+    {
+      "epoch": 3.45187840169919,
+      "grad_norm": 0.4743534714770601,
+      "learning_rate": 1.9273237579406494e-06,
+      "loss": 0.5566,
+      "step": 13003
+    },
+    {
+      "epoch": 3.452143900172574,
+      "grad_norm": 0.4611510372062144,
+      "learning_rate": 1.9269839189772895e-06,
+      "loss": 0.5505,
+      "step": 13004
+    },
+    {
+      "epoch": 3.452409398645958,
+      "grad_norm": 0.4615038078690836,
+      "learning_rate": 1.9266440911891465e-06,
+      "loss": 0.5423,
+      "step": 13005
+    },
+    {
+      "epoch": 3.4526748971193415,
+      "grad_norm": 0.45557747074812305,
+      "learning_rate": 1.926304274582847e-06,
+      "loss": 0.5415,
+      "step": 13006
+    },
+    {
+      "epoch": 3.4529403955927256,
+      "grad_norm": 0.4528829310353177,
+      "learning_rate": 1.925964469165019e-06,
+      "loss": 0.5423,
+      "step": 13007
+    },
+    {
+      "epoch": 3.4532058940661092,
+      "grad_norm": 0.47067752584222516,
+      "learning_rate": 1.925624674942289e-06,
+      "loss": 0.5628,
+      "step": 13008
+    },
+    {
+      "epoch": 3.453471392539493,
+      "grad_norm": 0.4607514602768546,
+      "learning_rate": 1.9252848919212846e-06,
+      "loss": 0.5355,
+      "step": 13009
+    },
+    {
+      "epoch": 3.4537368910128765,
+      "grad_norm": 0.44985792204100933,
+      "learning_rate": 1.9249451201086317e-06,
+      "loss": 0.5654,
+      "step": 13010
+    },
+    {
+      "epoch": 3.4540023894862606,
+      "grad_norm": 0.45449182303826563,
+      "learning_rate": 1.924605359510958e-06,
+      "loss": 0.5616,
+      "step": 13011
+    },
+    {
+      "epoch": 3.4542678879596442,
+      "grad_norm": 0.47226212212236696,
+      "learning_rate": 1.9242656101348874e-06,
+      "loss": 0.586,
+      "step": 13012
+    },
+    {
+      "epoch": 3.454533386433028,
+      "grad_norm": 0.4691818821114095,
+      "learning_rate": 1.9239258719870473e-06,
+      "loss": 0.5587,
+      "step": 13013
+    },
+    {
+      "epoch": 3.454798884906412,
+      "grad_norm": 0.45479814401412116,
+      "learning_rate": 1.923586145074063e-06,
+      "loss": 0.5378,
+      "step": 13014
+    },
+    {
+      "epoch": 3.4550643833797956,
+      "grad_norm": 0.4621556722674822,
+      "learning_rate": 1.9232464294025603e-06,
+      "loss": 0.5463,
+      "step": 13015
+    },
+    {
+      "epoch": 3.4553298818531792,
+      "grad_norm": 0.46983334038367863,
+      "learning_rate": 1.9229067249791642e-06,
+      "loss": 0.5186,
+      "step": 13016
+    },
+    {
+      "epoch": 3.455595380326563,
+      "grad_norm": 0.46574689452112356,
+      "learning_rate": 1.9225670318105004e-06,
+      "loss": 0.4987,
+      "step": 13017
+    },
+    {
+      "epoch": 3.455860878799947,
+      "grad_norm": 0.4493221731157662,
+      "learning_rate": 1.922227349903193e-06,
+      "loss": 0.5164,
+      "step": 13018
+    },
+    {
+      "epoch": 3.4561263772733306,
+      "grad_norm": 0.45585204320790923,
+      "learning_rate": 1.921887679263867e-06,
+      "loss": 0.5781,
+      "step": 13019
+    },
+    {
+      "epoch": 3.4563918757467142,
+      "grad_norm": 0.46515726418237735,
+      "learning_rate": 1.9215480198991466e-06,
+      "loss": 0.5308,
+      "step": 13020
+    },
+    {
+      "epoch": 3.4566573742200983,
+      "grad_norm": 0.45750436723102567,
+      "learning_rate": 1.921208371815656e-06,
+      "loss": 0.5863,
+      "step": 13021
+    },
+    {
+      "epoch": 3.456922872693482,
+      "grad_norm": 0.46999797936050514,
+      "learning_rate": 1.9208687350200203e-06,
+      "loss": 0.525,
+      "step": 13022
+    },
+    {
+      "epoch": 3.4571883711668656,
+      "grad_norm": 0.4576964641477319,
+      "learning_rate": 1.920529109518861e-06,
+      "loss": 0.518,
+      "step": 13023
+    },
+    {
+      "epoch": 3.4574538696402497,
+      "grad_norm": 0.4600663580915891,
+      "learning_rate": 1.9201894953188033e-06,
+      "loss": 0.5569,
+      "step": 13024
+    },
+    {
+      "epoch": 3.4577193681136333,
+      "grad_norm": 0.4544749581659812,
+      "learning_rate": 1.9198498924264696e-06,
+      "loss": 0.5049,
+      "step": 13025
+    },
+    {
+      "epoch": 3.457984866587017,
+      "grad_norm": 0.469239365371383,
+      "learning_rate": 1.919510300848483e-06,
+      "loss": 0.5541,
+      "step": 13026
+    },
+    {
+      "epoch": 3.458250365060401,
+      "grad_norm": 0.45208879213401876,
+      "learning_rate": 1.9191707205914677e-06,
+      "loss": 0.5682,
+      "step": 13027
+    },
+    {
+      "epoch": 3.4585158635337847,
+      "grad_norm": 0.44926342839868555,
+      "learning_rate": 1.9188311516620466e-06,
+      "loss": 0.5358,
+      "step": 13028
+    },
+    {
+      "epoch": 3.4587813620071683,
+      "grad_norm": 0.46910612332612955,
+      "learning_rate": 1.9184915940668395e-06,
+      "loss": 0.5737,
+      "step": 13029
+    },
+    {
+      "epoch": 3.4590468604805524,
+      "grad_norm": 0.4606913784856473,
+      "learning_rate": 1.9181520478124705e-06,
+      "loss": 0.5358,
+      "step": 13030
+    },
+    {
+      "epoch": 3.459312358953936,
+      "grad_norm": 0.4548509454852839,
+      "learning_rate": 1.917812512905561e-06,
+      "loss": 0.5558,
+      "step": 13031
+    },
+    {
+      "epoch": 3.4595778574273197,
+      "grad_norm": 0.4642261328111714,
+      "learning_rate": 1.917472989352733e-06,
+      "loss": 0.5515,
+      "step": 13032
+    },
+    {
+      "epoch": 3.459843355900704,
+      "grad_norm": 0.45112194009618994,
+      "learning_rate": 1.9171334771606076e-06,
+      "loss": 0.5609,
+      "step": 13033
+    },
+    {
+      "epoch": 3.4601088543740874,
+      "grad_norm": 0.4477191705133069,
+      "learning_rate": 1.9167939763358072e-06,
+      "loss": 0.5354,
+      "step": 13034
+    },
+    {
+      "epoch": 3.460374352847471,
+      "grad_norm": 0.4524173428885542,
+      "learning_rate": 1.9164544868849517e-06,
+      "loss": 0.5604,
+      "step": 13035
+    },
+    {
+      "epoch": 3.4606398513208547,
+      "grad_norm": 0.4625634275331173,
+      "learning_rate": 1.9161150088146627e-06,
+      "loss": 0.5377,
+      "step": 13036
+    },
+    {
+      "epoch": 3.460905349794239,
+      "grad_norm": 0.4739278956968761,
+      "learning_rate": 1.9157755421315607e-06,
+      "loss": 0.5264,
+      "step": 13037
+    },
+    {
+      "epoch": 3.4611708482676224,
+      "grad_norm": 0.45844106438913584,
+      "learning_rate": 1.9154360868422656e-06,
+      "loss": 0.5721,
+      "step": 13038
+    },
+    {
+      "epoch": 3.461436346741006,
+      "grad_norm": 0.4700908487606907,
+      "learning_rate": 1.9150966429533982e-06,
+      "loss": 0.5785,
+      "step": 13039
+    },
+    {
+      "epoch": 3.46170184521439,
+      "grad_norm": 0.47715054490431075,
+      "learning_rate": 1.9147572104715795e-06,
+      "loss": 0.5617,
+      "step": 13040
+    },
+    {
+      "epoch": 3.461967343687774,
+      "grad_norm": 0.45784774837914427,
+      "learning_rate": 1.9144177894034273e-06,
+      "loss": 0.5082,
+      "step": 13041
+    },
+    {
+      "epoch": 3.4622328421611575,
+      "grad_norm": 0.4518804508076868,
+      "learning_rate": 1.9140783797555615e-06,
+      "loss": 0.5167,
+      "step": 13042
+    },
+    {
+      "epoch": 3.4624983406345415,
+      "grad_norm": 0.47300834412284143,
+      "learning_rate": 1.9137389815346017e-06,
+      "loss": 0.5326,
+      "step": 13043
+    },
+    {
+      "epoch": 3.462763839107925,
+      "grad_norm": 0.4616618705479794,
+      "learning_rate": 1.913399594747168e-06,
+      "loss": 0.5402,
+      "step": 13044
+    },
+    {
+      "epoch": 3.463029337581309,
+      "grad_norm": 0.4551845044573789,
+      "learning_rate": 1.9130602193998785e-06,
+      "loss": 0.5754,
+      "step": 13045
+    },
+    {
+      "epoch": 3.4632948360546925,
+      "grad_norm": 0.46970169612768575,
+      "learning_rate": 1.912720855499351e-06,
+      "loss": 0.5335,
+      "step": 13046
+    },
+    {
+      "epoch": 3.4635603345280765,
+      "grad_norm": 0.4655763088520913,
+      "learning_rate": 1.912381503052205e-06,
+      "loss": 0.543,
+      "step": 13047
+    },
+    {
+      "epoch": 3.46382583300146,
+      "grad_norm": 0.5140019203188393,
+      "learning_rate": 1.912042162065059e-06,
+      "loss": 0.5473,
+      "step": 13048
+    },
+    {
+      "epoch": 3.464091331474844,
+      "grad_norm": 0.46056213612832314,
+      "learning_rate": 1.9117028325445295e-06,
+      "loss": 0.5518,
+      "step": 13049
+    },
+    {
+      "epoch": 3.464356829948228,
+      "grad_norm": 0.4616426820011735,
+      "learning_rate": 1.9113635144972355e-06,
+      "loss": 0.5824,
+      "step": 13050
+    },
+    {
+      "epoch": 3.4646223284216116,
+      "grad_norm": 0.45991151898802507,
+      "learning_rate": 1.9110242079297947e-06,
+      "loss": 0.5245,
+      "step": 13051
+    },
+    {
+      "epoch": 3.464887826894995,
+      "grad_norm": 0.4620963646500981,
+      "learning_rate": 1.9106849128488237e-06,
+      "loss": 0.5492,
+      "step": 13052
+    },
+    {
+      "epoch": 3.4651533253683793,
+      "grad_norm": 0.4534765755682093,
+      "learning_rate": 1.9103456292609395e-06,
+      "loss": 0.4942,
+      "step": 13053
+    },
+    {
+      "epoch": 3.465418823841763,
+      "grad_norm": 0.48193355425528095,
+      "learning_rate": 1.910006357172759e-06,
+      "loss": 0.5458,
+      "step": 13054
+    },
+    {
+      "epoch": 3.4656843223151466,
+      "grad_norm": 0.4687853146649047,
+      "learning_rate": 1.9096670965908997e-06,
+      "loss": 0.5549,
+      "step": 13055
+    },
+    {
+      "epoch": 3.4659498207885306,
+      "grad_norm": 0.4586536230777629,
+      "learning_rate": 1.909327847521977e-06,
+      "loss": 0.5435,
+      "step": 13056
+    },
+    {
+      "epoch": 3.4662153192619143,
+      "grad_norm": 0.461099145124684,
+      "learning_rate": 1.9089886099726085e-06,
+      "loss": 0.5491,
+      "step": 13057
+    },
+    {
+      "epoch": 3.466480817735298,
+      "grad_norm": 0.4747190918901796,
+      "learning_rate": 1.9086493839494082e-06,
+      "loss": 0.5778,
+      "step": 13058
+    },
+    {
+      "epoch": 3.466746316208682,
+      "grad_norm": 0.4496562357210341,
+      "learning_rate": 1.9083101694589927e-06,
+      "loss": 0.5205,
+      "step": 13059
+    },
+    {
+      "epoch": 3.4670118146820657,
+      "grad_norm": 0.4628152852119069,
+      "learning_rate": 1.907970966507978e-06,
+      "loss": 0.549,
+      "step": 13060
+    },
+    {
+      "epoch": 3.4672773131554493,
+      "grad_norm": 0.4632163628687318,
+      "learning_rate": 1.9076317751029784e-06,
+      "loss": 0.5389,
+      "step": 13061
+    },
+    {
+      "epoch": 3.4675428116288334,
+      "grad_norm": 0.4619241150626122,
+      "learning_rate": 1.90729259525061e-06,
+      "loss": 0.5224,
+      "step": 13062
+    },
+    {
+      "epoch": 3.467808310102217,
+      "grad_norm": 0.44826178100420744,
+      "learning_rate": 1.9069534269574875e-06,
+      "loss": 0.5668,
+      "step": 13063
+    },
+    {
+      "epoch": 3.4680738085756007,
+      "grad_norm": 0.45579877008943154,
+      "learning_rate": 1.9066142702302248e-06,
+      "loss": 0.5458,
+      "step": 13064
+    },
+    {
+      "epoch": 3.4683393070489843,
+      "grad_norm": 0.4617797553771512,
+      "learning_rate": 1.9062751250754367e-06,
+      "loss": 0.5638,
+      "step": 13065
+    },
+    {
+      "epoch": 3.4686048055223684,
+      "grad_norm": 0.4699196368062382,
+      "learning_rate": 1.9059359914997372e-06,
+      "loss": 0.5572,
+      "step": 13066
+    },
+    {
+      "epoch": 3.468870303995752,
+      "grad_norm": 0.45506100296512775,
+      "learning_rate": 1.9055968695097404e-06,
+      "loss": 0.5786,
+      "step": 13067
+    },
+    {
+      "epoch": 3.4691358024691357,
+      "grad_norm": 0.46458393426558037,
+      "learning_rate": 1.9052577591120603e-06,
+      "loss": 0.5197,
+      "step": 13068
+    },
+    {
+      "epoch": 3.4694013009425198,
+      "grad_norm": 0.4633086425999223,
+      "learning_rate": 1.9049186603133108e-06,
+      "loss": 0.5708,
+      "step": 13069
+    },
+    {
+      "epoch": 3.4696667994159034,
+      "grad_norm": 0.4554800080836122,
+      "learning_rate": 1.9045795731201034e-06,
+      "loss": 0.5383,
+      "step": 13070
+    },
+    {
+      "epoch": 3.469932297889287,
+      "grad_norm": 0.45265343120360074,
+      "learning_rate": 1.9042404975390522e-06,
+      "loss": 0.5342,
+      "step": 13071
+    },
+    {
+      "epoch": 3.4701977963626707,
+      "grad_norm": 0.4601467132721148,
+      "learning_rate": 1.9039014335767694e-06,
+      "loss": 0.5106,
+      "step": 13072
+    },
+    {
+      "epoch": 3.4704632948360548,
+      "grad_norm": 0.4726864455980346,
+      "learning_rate": 1.9035623812398688e-06,
+      "loss": 0.5233,
+      "step": 13073
+    },
+    {
+      "epoch": 3.4707287933094384,
+      "grad_norm": 0.4746648009140776,
+      "learning_rate": 1.9032233405349632e-06,
+      "loss": 0.5532,
+      "step": 13074
+    },
+    {
+      "epoch": 3.470994291782822,
+      "grad_norm": 0.4548252819721672,
+      "learning_rate": 1.9028843114686625e-06,
+      "loss": 0.5431,
+      "step": 13075
+    },
+    {
+      "epoch": 3.471259790256206,
+      "grad_norm": 0.4611773211640143,
+      "learning_rate": 1.9025452940475795e-06,
+      "loss": 0.5221,
+      "step": 13076
+    },
+    {
+      "epoch": 3.4715252887295898,
+      "grad_norm": 0.4482303071975979,
+      "learning_rate": 1.9022062882783263e-06,
+      "loss": 0.5704,
+      "step": 13077
+    },
+    {
+      "epoch": 3.4717907872029734,
+      "grad_norm": 0.4576110816729715,
+      "learning_rate": 1.9018672941675142e-06,
+      "loss": 0.5407,
+      "step": 13078
+    },
+    {
+      "epoch": 3.4720562856763575,
+      "grad_norm": 0.48543926866284937,
+      "learning_rate": 1.9015283117217543e-06,
+      "loss": 0.5391,
+      "step": 13079
+    },
+    {
+      "epoch": 3.472321784149741,
+      "grad_norm": 0.4781050667958576,
+      "learning_rate": 1.901189340947658e-06,
+      "loss": 0.5377,
+      "step": 13080
+    },
+    {
+      "epoch": 3.4725872826231248,
+      "grad_norm": 0.4586390122398653,
+      "learning_rate": 1.9008503818518354e-06,
+      "loss": 0.5141,
+      "step": 13081
+    },
+    {
+      "epoch": 3.472852781096509,
+      "grad_norm": 0.4669321724121264,
+      "learning_rate": 1.900511434440897e-06,
+      "loss": 0.5449,
+      "step": 13082
+    },
+    {
+      "epoch": 3.4731182795698925,
+      "grad_norm": 0.47109310017189765,
+      "learning_rate": 1.9001724987214538e-06,
+      "loss": 0.5487,
+      "step": 13083
+    },
+    {
+      "epoch": 3.473383778043276,
+      "grad_norm": 0.4677290850715841,
+      "learning_rate": 1.8998335747001151e-06,
+      "loss": 0.5061,
+      "step": 13084
+    },
+    {
+      "epoch": 3.4736492765166602,
+      "grad_norm": 0.4688399397970539,
+      "learning_rate": 1.8994946623834915e-06,
+      "loss": 0.5545,
+      "step": 13085
+    },
+    {
+      "epoch": 3.473914774990044,
+      "grad_norm": 0.4750485003142622,
+      "learning_rate": 1.899155761778193e-06,
+      "loss": 0.5892,
+      "step": 13086
+    },
+    {
+      "epoch": 3.4741802734634275,
+      "grad_norm": 0.4636533595233714,
+      "learning_rate": 1.8988168728908277e-06,
+      "loss": 0.5482,
+      "step": 13087
+    },
+    {
+      "epoch": 3.4744457719368116,
+      "grad_norm": 0.4851767125423052,
+      "learning_rate": 1.898477995728005e-06,
+      "loss": 0.5363,
+      "step": 13088
+    },
+    {
+      "epoch": 3.4747112704101952,
+      "grad_norm": 0.47499587715252584,
+      "learning_rate": 1.8981391302963342e-06,
+      "loss": 0.5709,
+      "step": 13089
+    },
+    {
+      "epoch": 3.474976768883579,
+      "grad_norm": 0.45779698420299514,
+      "learning_rate": 1.897800276602424e-06,
+      "loss": 0.5754,
+      "step": 13090
+    },
+    {
+      "epoch": 3.4752422673569625,
+      "grad_norm": 0.4649477856952448,
+      "learning_rate": 1.8974614346528827e-06,
+      "loss": 0.5193,
+      "step": 13091
+    },
+    {
+      "epoch": 3.4755077658303466,
+      "grad_norm": 0.4672334249412886,
+      "learning_rate": 1.8971226044543195e-06,
+      "loss": 0.5794,
+      "step": 13092
+    },
+    {
+      "epoch": 3.4757732643037302,
+      "grad_norm": 0.4561299731054907,
+      "learning_rate": 1.8967837860133408e-06,
+      "loss": 0.531,
+      "step": 13093
+    },
+    {
+      "epoch": 3.476038762777114,
+      "grad_norm": 0.46007594291993337,
+      "learning_rate": 1.8964449793365552e-06,
+      "loss": 0.544,
+      "step": 13094
+    },
+    {
+      "epoch": 3.476304261250498,
+      "grad_norm": 0.44921333712032363,
+      "learning_rate": 1.8961061844305707e-06,
+      "loss": 0.5076,
+      "step": 13095
+    },
+    {
+      "epoch": 3.4765697597238816,
+      "grad_norm": 0.4424093869623847,
+      "learning_rate": 1.8957674013019939e-06,
+      "loss": 0.5549,
+      "step": 13096
+    },
+    {
+      "epoch": 3.4768352581972652,
+      "grad_norm": 0.4656226841477639,
+      "learning_rate": 1.8954286299574326e-06,
+      "loss": 0.5394,
+      "step": 13097
+    },
+    {
+      "epoch": 3.4771007566706493,
+      "grad_norm": 0.49083088494877003,
+      "learning_rate": 1.8950898704034928e-06,
+      "loss": 0.5352,
+      "step": 13098
+    },
+    {
+      "epoch": 3.477366255144033,
+      "grad_norm": 0.43954675686839756,
+      "learning_rate": 1.8947511226467816e-06,
+      "loss": 0.5623,
+      "step": 13099
+    },
+    {
+      "epoch": 3.4776317536174166,
+      "grad_norm": 0.4614653643503416,
+      "learning_rate": 1.8944123866939053e-06,
+      "loss": 0.5407,
+      "step": 13100
+    },
+    {
+      "epoch": 3.4778972520908003,
+      "grad_norm": 0.4609393066520792,
+      "learning_rate": 1.8940736625514705e-06,
+      "loss": 0.593,
+      "step": 13101
+    },
+    {
+      "epoch": 3.4781627505641843,
+      "grad_norm": 0.4606841016586335,
+      "learning_rate": 1.8937349502260827e-06,
+      "loss": 0.4745,
+      "step": 13102
+    },
+    {
+      "epoch": 3.478428249037568,
+      "grad_norm": 0.47236562273268456,
+      "learning_rate": 1.893396249724349e-06,
+      "loss": 0.5428,
+      "step": 13103
+    },
+    {
+      "epoch": 3.4786937475109516,
+      "grad_norm": 0.46115447466717713,
+      "learning_rate": 1.8930575610528723e-06,
+      "loss": 0.53,
+      "step": 13104
+    },
+    {
+      "epoch": 3.4789592459843357,
+      "grad_norm": 0.4510127108505656,
+      "learning_rate": 1.8927188842182592e-06,
+      "loss": 0.5666,
+      "step": 13105
+    },
+    {
+      "epoch": 3.4792247444577193,
+      "grad_norm": 0.46337362742117966,
+      "learning_rate": 1.8923802192271147e-06,
+      "loss": 0.5573,
+      "step": 13106
+    },
+    {
+      "epoch": 3.479490242931103,
+      "grad_norm": 0.4516553698833458,
+      "learning_rate": 1.8920415660860442e-06,
+      "loss": 0.564,
+      "step": 13107
+    },
+    {
+      "epoch": 3.479755741404487,
+      "grad_norm": 0.45462409865497344,
+      "learning_rate": 1.891702924801651e-06,
+      "loss": 0.5397,
+      "step": 13108
+    },
+    {
+      "epoch": 3.4800212398778707,
+      "grad_norm": 0.44521164333003743,
+      "learning_rate": 1.8913642953805412e-06,
+      "loss": 0.548,
+      "step": 13109
+    },
+    {
+      "epoch": 3.4802867383512543,
+      "grad_norm": 0.47200593064224217,
+      "learning_rate": 1.8910256778293169e-06,
+      "loss": 0.5094,
+      "step": 13110
+    },
+    {
+      "epoch": 3.4805522368246384,
+      "grad_norm": 0.4741676227031946,
+      "learning_rate": 1.8906870721545832e-06,
+      "loss": 0.5818,
+      "step": 13111
+    },
+    {
+      "epoch": 3.480817735298022,
+      "grad_norm": 0.45919879621874155,
+      "learning_rate": 1.8903484783629433e-06,
+      "loss": 0.5626,
+      "step": 13112
+    },
+    {
+      "epoch": 3.4810832337714057,
+      "grad_norm": 0.4667153948976328,
+      "learning_rate": 1.8900098964610009e-06,
+      "loss": 0.5697,
+      "step": 13113
+    },
+    {
+      "epoch": 3.48134873224479,
+      "grad_norm": 0.4818615202166082,
+      "learning_rate": 1.8896713264553588e-06,
+      "loss": 0.5132,
+      "step": 13114
+    },
+    {
+      "epoch": 3.4816142307181734,
+      "grad_norm": 0.4571252144887379,
+      "learning_rate": 1.8893327683526214e-06,
+      "loss": 0.4993,
+      "step": 13115
+    },
+    {
+      "epoch": 3.481879729191557,
+      "grad_norm": 0.4508327623192309,
+      "learning_rate": 1.888994222159389e-06,
+      "loss": 0.506,
+      "step": 13116
+    },
+    {
+      "epoch": 3.482145227664941,
+      "grad_norm": 0.4604709533209294,
+      "learning_rate": 1.8886556878822648e-06,
+      "loss": 0.548,
+      "step": 13117
+    },
+    {
+      "epoch": 3.482410726138325,
+      "grad_norm": 0.4602635516792217,
+      "learning_rate": 1.888317165527852e-06,
+      "loss": 0.5701,
+      "step": 13118
+    },
+    {
+      "epoch": 3.4826762246117084,
+      "grad_norm": 0.46044951665657796,
+      "learning_rate": 1.8879786551027524e-06,
+      "loss": 0.5653,
+      "step": 13119
+    },
+    {
+      "epoch": 3.482941723085092,
+      "grad_norm": 0.45350328146382823,
+      "learning_rate": 1.8876401566135683e-06,
+      "loss": 0.5705,
+      "step": 13120
+    },
+    {
+      "epoch": 3.483207221558476,
+      "grad_norm": 0.46948995429357676,
+      "learning_rate": 1.8873016700668996e-06,
+      "loss": 0.5526,
+      "step": 13121
+    },
+    {
+      "epoch": 3.48347272003186,
+      "grad_norm": 0.4654642207694232,
+      "learning_rate": 1.8869631954693478e-06,
+      "loss": 0.5666,
+      "step": 13122
+    },
+    {
+      "epoch": 3.4837382185052435,
+      "grad_norm": 0.4556160800819771,
+      "learning_rate": 1.8866247328275156e-06,
+      "loss": 0.5538,
+      "step": 13123
+    },
+    {
+      "epoch": 3.4840037169786275,
+      "grad_norm": 0.45978900889141083,
+      "learning_rate": 1.8862862821480023e-06,
+      "loss": 0.536,
+      "step": 13124
+    },
+    {
+      "epoch": 3.484269215452011,
+      "grad_norm": 0.45080616960655173,
+      "learning_rate": 1.8859478434374095e-06,
+      "loss": 0.5598,
+      "step": 13125
+    },
+    {
+      "epoch": 3.484534713925395,
+      "grad_norm": 0.4648588181649765,
+      "learning_rate": 1.8856094167023375e-06,
+      "loss": 0.5736,
+      "step": 13126
+    },
+    {
+      "epoch": 3.4848002123987785,
+      "grad_norm": 0.45755327070054214,
+      "learning_rate": 1.8852710019493858e-06,
+      "loss": 0.5467,
+      "step": 13127
+    },
+    {
+      "epoch": 3.4850657108721625,
+      "grad_norm": 0.45627472745113706,
+      "learning_rate": 1.8849325991851543e-06,
+      "loss": 0.5125,
+      "step": 13128
+    },
+    {
+      "epoch": 3.485331209345546,
+      "grad_norm": 0.4505009862628162,
+      "learning_rate": 1.8845942084162435e-06,
+      "loss": 0.5431,
+      "step": 13129
+    },
+    {
+      "epoch": 3.48559670781893,
+      "grad_norm": 0.46495264564422034,
+      "learning_rate": 1.8842558296492525e-06,
+      "loss": 0.5709,
+      "step": 13130
+    },
+    {
+      "epoch": 3.485862206292314,
+      "grad_norm": 0.4562322356932985,
+      "learning_rate": 1.88391746289078e-06,
+      "loss": 0.5348,
+      "step": 13131
+    },
+    {
+      "epoch": 3.4861277047656976,
+      "grad_norm": 0.4534614736578336,
+      "learning_rate": 1.8835791081474266e-06,
+      "loss": 0.5564,
+      "step": 13132
+    },
+    {
+      "epoch": 3.486393203239081,
+      "grad_norm": 0.46696695532369054,
+      "learning_rate": 1.8832407654257888e-06,
+      "loss": 0.5597,
+      "step": 13133
+    },
+    {
+      "epoch": 3.4866587017124653,
+      "grad_norm": 0.46089847027854514,
+      "learning_rate": 1.8829024347324662e-06,
+      "loss": 0.5627,
+      "step": 13134
+    },
+    {
+      "epoch": 3.486924200185849,
+      "grad_norm": 0.4833492203102998,
+      "learning_rate": 1.8825641160740572e-06,
+      "loss": 0.6206,
+      "step": 13135
+    },
+    {
+      "epoch": 3.4871896986592326,
+      "grad_norm": 0.4697515245265211,
+      "learning_rate": 1.8822258094571594e-06,
+      "loss": 0.5412,
+      "step": 13136
+    },
+    {
+      "epoch": 3.4874551971326166,
+      "grad_norm": 0.4426142784437776,
+      "learning_rate": 1.881887514888371e-06,
+      "loss": 0.5224,
+      "step": 13137
+    },
+    {
+      "epoch": 3.4877206956060003,
+      "grad_norm": 0.47129788669639094,
+      "learning_rate": 1.88154923237429e-06,
+      "loss": 0.5188,
+      "step": 13138
+    },
+    {
+      "epoch": 3.487986194079384,
+      "grad_norm": 0.4369495855962656,
+      "learning_rate": 1.8812109619215127e-06,
+      "loss": 0.5014,
+      "step": 13139
+    },
+    {
+      "epoch": 3.488251692552768,
+      "grad_norm": 0.47178333216586293,
+      "learning_rate": 1.8808727035366367e-06,
+      "loss": 0.5642,
+      "step": 13140
+    },
+    {
+      "epoch": 3.4885171910261517,
+      "grad_norm": 0.4800236894260246,
+      "learning_rate": 1.8805344572262588e-06,
+      "loss": 0.5151,
+      "step": 13141
+    },
+    {
+      "epoch": 3.4887826894995353,
+      "grad_norm": 0.46405485935384677,
+      "learning_rate": 1.8801962229969756e-06,
+      "loss": 0.5248,
+      "step": 13142
+    },
+    {
+      "epoch": 3.4890481879729194,
+      "grad_norm": 0.4642600858831924,
+      "learning_rate": 1.8798580008553844e-06,
+      "loss": 0.5311,
+      "step": 13143
+    },
+    {
+      "epoch": 3.489313686446303,
+      "grad_norm": 0.45091809800597543,
+      "learning_rate": 1.8795197908080798e-06,
+      "loss": 0.5763,
+      "step": 13144
+    },
+    {
+      "epoch": 3.4895791849196867,
+      "grad_norm": 0.47113371351316113,
+      "learning_rate": 1.8791815928616585e-06,
+      "loss": 0.536,
+      "step": 13145
+    },
+    {
+      "epoch": 3.4898446833930703,
+      "grad_norm": 0.46543894029201194,
+      "learning_rate": 1.8788434070227163e-06,
+      "loss": 0.5632,
+      "step": 13146
+    },
+    {
+      "epoch": 3.4901101818664544,
+      "grad_norm": 0.4474759154916583,
+      "learning_rate": 1.8785052332978485e-06,
+      "loss": 0.5287,
+      "step": 13147
+    },
+    {
+      "epoch": 3.490375680339838,
+      "grad_norm": 0.44304691706684995,
+      "learning_rate": 1.87816707169365e-06,
+      "loss": 0.5583,
+      "step": 13148
+    },
+    {
+      "epoch": 3.4906411788132217,
+      "grad_norm": 0.4587295735368923,
+      "learning_rate": 1.8778289222167175e-06,
+      "loss": 0.5431,
+      "step": 13149
+    },
+    {
+      "epoch": 3.4909066772866058,
+      "grad_norm": 0.45910701903488704,
+      "learning_rate": 1.8774907848736434e-06,
+      "loss": 0.5565,
+      "step": 13150
+    },
+    {
+      "epoch": 3.4911721757599894,
+      "grad_norm": 0.45228110901400725,
+      "learning_rate": 1.877152659671023e-06,
+      "loss": 0.5687,
+      "step": 13151
+    },
+    {
+      "epoch": 3.491437674233373,
+      "grad_norm": 0.4577606175365332,
+      "learning_rate": 1.8768145466154505e-06,
+      "loss": 0.5327,
+      "step": 13152
+    },
+    {
+      "epoch": 3.491703172706757,
+      "grad_norm": 0.48210337548405086,
+      "learning_rate": 1.8764764457135204e-06,
+      "loss": 0.5609,
+      "step": 13153
+    },
+    {
+      "epoch": 3.4919686711801408,
+      "grad_norm": 0.43958827660965805,
+      "learning_rate": 1.8761383569718265e-06,
+      "loss": 0.5586,
+      "step": 13154
+    },
+    {
+      "epoch": 3.4922341696535244,
+      "grad_norm": 0.4688682857359886,
+      "learning_rate": 1.875800280396962e-06,
+      "loss": 0.5621,
+      "step": 13155
+    },
+    {
+      "epoch": 3.492499668126908,
+      "grad_norm": 0.4592577645187981,
+      "learning_rate": 1.8754622159955205e-06,
+      "loss": 0.5482,
+      "step": 13156
+    },
+    {
+      "epoch": 3.492765166600292,
+      "grad_norm": 0.46793632483283865,
+      "learning_rate": 1.8751241637740947e-06,
+      "loss": 0.5267,
+      "step": 13157
+    },
+    {
+      "epoch": 3.4930306650736758,
+      "grad_norm": 0.4627307671019581,
+      "learning_rate": 1.8747861237392772e-06,
+      "loss": 0.5212,
+      "step": 13158
+    },
+    {
+      "epoch": 3.4932961635470594,
+      "grad_norm": 0.46754340511612763,
+      "learning_rate": 1.8744480958976619e-06,
+      "loss": 0.572,
+      "step": 13159
+    },
+    {
+      "epoch": 3.4935616620204435,
+      "grad_norm": 0.4473161739180905,
+      "learning_rate": 1.8741100802558399e-06,
+      "loss": 0.5654,
+      "step": 13160
+    },
+    {
+      "epoch": 3.493827160493827,
+      "grad_norm": 0.47247497464222454,
+      "learning_rate": 1.8737720768204047e-06,
+      "loss": 0.5145,
+      "step": 13161
+    },
+    {
+      "epoch": 3.4940926589672108,
+      "grad_norm": 0.4584942994941058,
+      "learning_rate": 1.873434085597946e-06,
+      "loss": 0.5269,
+      "step": 13162
+    },
+    {
+      "epoch": 3.494358157440595,
+      "grad_norm": 0.44534585119093867,
+      "learning_rate": 1.8730961065950567e-06,
+      "loss": 0.5443,
+      "step": 13163
+    },
+    {
+      "epoch": 3.4946236559139785,
+      "grad_norm": 0.4577392355231514,
+      "learning_rate": 1.8727581398183287e-06,
+      "loss": 0.5473,
+      "step": 13164
+    },
+    {
+      "epoch": 3.494889154387362,
+      "grad_norm": 0.46469570974731267,
+      "learning_rate": 1.8724201852743526e-06,
+      "loss": 0.545,
+      "step": 13165
+    },
+    {
+      "epoch": 3.4951546528607462,
+      "grad_norm": 0.4401786987559962,
+      "learning_rate": 1.8720822429697193e-06,
+      "loss": 0.5353,
+      "step": 13166
+    },
+    {
+      "epoch": 3.49542015133413,
+      "grad_norm": 0.4542033800032534,
+      "learning_rate": 1.8717443129110212e-06,
+      "loss": 0.5262,
+      "step": 13167
+    },
+    {
+      "epoch": 3.4956856498075135,
+      "grad_norm": 0.46175017544644553,
+      "learning_rate": 1.8714063951048462e-06,
+      "loss": 0.5101,
+      "step": 13168
+    },
+    {
+      "epoch": 3.4959511482808976,
+      "grad_norm": 0.4600274337983664,
+      "learning_rate": 1.8710684895577853e-06,
+      "loss": 0.5822,
+      "step": 13169
+    },
+    {
+      "epoch": 3.4962166467542812,
+      "grad_norm": 0.4845600631862799,
+      "learning_rate": 1.8707305962764283e-06,
+      "loss": 0.5585,
+      "step": 13170
+    },
+    {
+      "epoch": 3.496482145227665,
+      "grad_norm": 0.4637113917385523,
+      "learning_rate": 1.870392715267366e-06,
+      "loss": 0.5427,
+      "step": 13171
+    },
+    {
+      "epoch": 3.496747643701049,
+      "grad_norm": 0.4651673444293219,
+      "learning_rate": 1.8700548465371877e-06,
+      "loss": 0.5845,
+      "step": 13172
+    },
+    {
+      "epoch": 3.4970131421744326,
+      "grad_norm": 0.460838142013931,
+      "learning_rate": 1.8697169900924817e-06,
+      "loss": 0.5315,
+      "step": 13173
+    },
+    {
+      "epoch": 3.4972786406478162,
+      "grad_norm": 0.45938638451473546,
+      "learning_rate": 1.8693791459398376e-06,
+      "loss": 0.5612,
+      "step": 13174
+    },
+    {
+      "epoch": 3.4975441391212,
+      "grad_norm": 0.46938756741124216,
+      "learning_rate": 1.8690413140858438e-06,
+      "loss": 0.566,
+      "step": 13175
+    },
+    {
+      "epoch": 3.497809637594584,
+      "grad_norm": 0.44987329047846875,
+      "learning_rate": 1.8687034945370896e-06,
+      "loss": 0.5032,
+      "step": 13176
+    },
+    {
+      "epoch": 3.4980751360679676,
+      "grad_norm": 0.46084191858840934,
+      "learning_rate": 1.868365687300163e-06,
+      "loss": 0.5133,
+      "step": 13177
+    },
+    {
+      "epoch": 3.4983406345413512,
+      "grad_norm": 0.4457448430336164,
+      "learning_rate": 1.868027892381653e-06,
+      "loss": 0.5457,
+      "step": 13178
+    },
+    {
+      "epoch": 3.4986061330147353,
+      "grad_norm": 0.45071788882914327,
+      "learning_rate": 1.8676901097881452e-06,
+      "loss": 0.4888,
+      "step": 13179
+    },
+    {
+      "epoch": 3.498871631488119,
+      "grad_norm": 0.439683004064656,
+      "learning_rate": 1.8673523395262289e-06,
+      "loss": 0.5582,
+      "step": 13180
+    },
+    {
+      "epoch": 3.4991371299615026,
+      "grad_norm": 0.4723019366949332,
+      "learning_rate": 1.8670145816024904e-06,
+      "loss": 0.5814,
+      "step": 13181
+    },
+    {
+      "epoch": 3.4994026284348863,
+      "grad_norm": 0.444985374888047,
+      "learning_rate": 1.8666768360235178e-06,
+      "loss": 0.5524,
+      "step": 13182
+    },
+    {
+      "epoch": 3.4996681269082703,
+      "grad_norm": 0.4520293573667949,
+      "learning_rate": 1.8663391027958971e-06,
+      "loss": 0.5306,
+      "step": 13183
+    },
+    {
+      "epoch": 3.499933625381654,
+      "grad_norm": 0.4513887029299811,
+      "learning_rate": 1.8660013819262162e-06,
+      "loss": 0.5699,
+      "step": 13184
+    },
+    {
+      "epoch": 3.5001991238550376,
+      "grad_norm": 0.4734639150604951,
+      "learning_rate": 1.86566367342106e-06,
+      "loss": 0.537,
+      "step": 13185
+    },
+    {
+      "epoch": 3.5004646223284217,
+      "grad_norm": 0.4483989773127945,
+      "learning_rate": 1.8653259772870153e-06,
+      "loss": 0.5544,
+      "step": 13186
+    },
+    {
+      "epoch": 3.5007301208018053,
+      "grad_norm": 0.4646666182030601,
+      "learning_rate": 1.8649882935306682e-06,
+      "loss": 0.5561,
+      "step": 13187
+    },
+    {
+      "epoch": 3.500995619275189,
+      "grad_norm": 0.4661581327853897,
+      "learning_rate": 1.864650622158604e-06,
+      "loss": 0.5582,
+      "step": 13188
+    },
+    {
+      "epoch": 3.501261117748573,
+      "grad_norm": 0.45406504418374377,
+      "learning_rate": 1.864312963177408e-06,
+      "loss": 0.5521,
+      "step": 13189
+    },
+    {
+      "epoch": 3.5015266162219567,
+      "grad_norm": 0.46087495382865173,
+      "learning_rate": 1.8639753165936666e-06,
+      "loss": 0.5447,
+      "step": 13190
+    },
+    {
+      "epoch": 3.5017921146953404,
+      "grad_norm": 0.4535285068202552,
+      "learning_rate": 1.8636376824139631e-06,
+      "loss": 0.5587,
+      "step": 13191
+    },
+    {
+      "epoch": 3.5020576131687244,
+      "grad_norm": 0.45915532879592114,
+      "learning_rate": 1.863300060644883e-06,
+      "loss": 0.5646,
+      "step": 13192
+    },
+    {
+      "epoch": 3.502323111642108,
+      "grad_norm": 0.4645994524977131,
+      "learning_rate": 1.8629624512930106e-06,
+      "loss": 0.5365,
+      "step": 13193
+    },
+    {
+      "epoch": 3.5025886101154917,
+      "grad_norm": 0.4513661305357719,
+      "learning_rate": 1.86262485436493e-06,
+      "loss": 0.535,
+      "step": 13194
+    },
+    {
+      "epoch": 3.502854108588876,
+      "grad_norm": 0.43827157958341256,
+      "learning_rate": 1.8622872698672262e-06,
+      "loss": 0.5154,
+      "step": 13195
+    },
+    {
+      "epoch": 3.5031196070622594,
+      "grad_norm": 0.4506053522555669,
+      "learning_rate": 1.8619496978064813e-06,
+      "loss": 0.5467,
+      "step": 13196
+    },
+    {
+      "epoch": 3.503385105535643,
+      "grad_norm": 0.46505737547395676,
+      "learning_rate": 1.8616121381892793e-06,
+      "loss": 0.5495,
+      "step": 13197
+    },
+    {
+      "epoch": 3.503650604009027,
+      "grad_norm": 0.4510448468374083,
+      "learning_rate": 1.8612745910222038e-06,
+      "loss": 0.5717,
+      "step": 13198
+    },
+    {
+      "epoch": 3.503916102482411,
+      "grad_norm": 0.47066050125560743,
+      "learning_rate": 1.8609370563118375e-06,
+      "loss": 0.5596,
+      "step": 13199
+    },
+    {
+      "epoch": 3.5041816009557945,
+      "grad_norm": 0.466524672061839,
+      "learning_rate": 1.8605995340647637e-06,
+      "loss": 0.5591,
+      "step": 13200
+    },
+    {
+      "epoch": 3.5044470994291785,
+      "grad_norm": 0.44756731624600676,
+      "learning_rate": 1.8602620242875646e-06,
+      "loss": 0.5313,
+      "step": 13201
+    },
+    {
+      "epoch": 3.504712597902562,
+      "grad_norm": 0.47791408719069856,
+      "learning_rate": 1.8599245269868218e-06,
+      "loss": 0.5553,
+      "step": 13202
+    },
+    {
+      "epoch": 3.504978096375946,
+      "grad_norm": 0.46507393385105034,
+      "learning_rate": 1.8595870421691181e-06,
+      "loss": 0.5799,
+      "step": 13203
+    },
+    {
+      "epoch": 3.5052435948493295,
+      "grad_norm": 0.467800533648253,
+      "learning_rate": 1.8592495698410352e-06,
+      "loss": 0.5519,
+      "step": 13204
+    },
+    {
+      "epoch": 3.5055090933227135,
+      "grad_norm": 0.4532616912575588,
+      "learning_rate": 1.8589121100091543e-06,
+      "loss": 0.5192,
+      "step": 13205
+    },
+    {
+      "epoch": 3.505774591796097,
+      "grad_norm": 0.4532209423553535,
+      "learning_rate": 1.8585746626800568e-06,
+      "loss": 0.5155,
+      "step": 13206
+    },
+    {
+      "epoch": 3.506040090269481,
+      "grad_norm": 0.46289448648902665,
+      "learning_rate": 1.8582372278603251e-06,
+      "loss": 0.5157,
+      "step": 13207
+    },
+    {
+      "epoch": 3.5063055887428645,
+      "grad_norm": 0.4561396995106719,
+      "learning_rate": 1.8578998055565378e-06,
+      "loss": 0.5636,
+      "step": 13208
+    },
+    {
+      "epoch": 3.5065710872162486,
+      "grad_norm": 0.47240421547746986,
+      "learning_rate": 1.8575623957752758e-06,
+      "loss": 0.5351,
+      "step": 13209
+    },
+    {
+      "epoch": 3.506836585689632,
+      "grad_norm": 0.47129968397973104,
+      "learning_rate": 1.8572249985231206e-06,
+      "loss": 0.548,
+      "step": 13210
+    },
+    {
+      "epoch": 3.507102084163016,
+      "grad_norm": 0.46654157997260776,
+      "learning_rate": 1.8568876138066515e-06,
+      "loss": 0.544,
+      "step": 13211
+    },
+    {
+      "epoch": 3.5073675826364,
+      "grad_norm": 0.449067104004744,
+      "learning_rate": 1.8565502416324488e-06,
+      "loss": 0.5486,
+      "step": 13212
+    },
+    {
+      "epoch": 3.5076330811097836,
+      "grad_norm": 0.4837915043261843,
+      "learning_rate": 1.8562128820070927e-06,
+      "loss": 0.5792,
+      "step": 13213
+    },
+    {
+      "epoch": 3.507898579583167,
+      "grad_norm": 0.47048910569028435,
+      "learning_rate": 1.8558755349371607e-06,
+      "loss": 0.5497,
+      "step": 13214
+    },
+    {
+      "epoch": 3.5081640780565513,
+      "grad_norm": 0.4502862274042531,
+      "learning_rate": 1.8555382004292328e-06,
+      "loss": 0.5461,
+      "step": 13215
+    },
+    {
+      "epoch": 3.508429576529935,
+      "grad_norm": 0.48896827945236027,
+      "learning_rate": 1.855200878489888e-06,
+      "loss": 0.5508,
+      "step": 13216
+    },
+    {
+      "epoch": 3.5086950750033186,
+      "grad_norm": 0.47388248050262494,
+      "learning_rate": 1.8548635691257047e-06,
+      "loss": 0.5559,
+      "step": 13217
+    },
+    {
+      "epoch": 3.5089605734767026,
+      "grad_norm": 0.4491553182116303,
+      "learning_rate": 1.854526272343262e-06,
+      "loss": 0.55,
+      "step": 13218
+    },
+    {
+      "epoch": 3.5092260719500863,
+      "grad_norm": 0.4756736157273259,
+      "learning_rate": 1.8541889881491367e-06,
+      "loss": 0.5712,
+      "step": 13219
+    },
+    {
+      "epoch": 3.50949157042347,
+      "grad_norm": 0.5005277079223958,
+      "learning_rate": 1.8538517165499077e-06,
+      "loss": 0.5456,
+      "step": 13220
+    },
+    {
+      "epoch": 3.509757068896854,
+      "grad_norm": 0.4475895248385321,
+      "learning_rate": 1.853514457552152e-06,
+      "loss": 0.5684,
+      "step": 13221
+    },
+    {
+      "epoch": 3.5100225673702377,
+      "grad_norm": 0.44870713024964237,
+      "learning_rate": 1.853177211162447e-06,
+      "loss": 0.5213,
+      "step": 13222
+    },
+    {
+      "epoch": 3.5102880658436213,
+      "grad_norm": 0.45681337885406564,
+      "learning_rate": 1.8528399773873705e-06,
+      "loss": 0.5413,
+      "step": 13223
+    },
+    {
+      "epoch": 3.5105535643170054,
+      "grad_norm": 0.4712305929267227,
+      "learning_rate": 1.8525027562335e-06,
+      "loss": 0.5611,
+      "step": 13224
+    },
+    {
+      "epoch": 3.510819062790389,
+      "grad_norm": 0.45428853962533106,
+      "learning_rate": 1.8521655477074096e-06,
+      "loss": 0.5669,
+      "step": 13225
+    },
+    {
+      "epoch": 3.5110845612637727,
+      "grad_norm": 0.46873811593732023,
+      "learning_rate": 1.8518283518156777e-06,
+      "loss": 0.5712,
+      "step": 13226
+    },
+    {
+      "epoch": 3.5113500597371567,
+      "grad_norm": 0.46113602968033557,
+      "learning_rate": 1.8514911685648798e-06,
+      "loss": 0.5457,
+      "step": 13227
+    },
+    {
+      "epoch": 3.5116155582105404,
+      "grad_norm": 0.4729629403413077,
+      "learning_rate": 1.8511539979615917e-06,
+      "loss": 0.5614,
+      "step": 13228
+    },
+    {
+      "epoch": 3.511881056683924,
+      "grad_norm": 0.4514113331773916,
+      "learning_rate": 1.8508168400123892e-06,
+      "loss": 0.5333,
+      "step": 13229
+    },
+    {
+      "epoch": 3.512146555157308,
+      "grad_norm": 0.4529695813427385,
+      "learning_rate": 1.850479694723849e-06,
+      "loss": 0.5571,
+      "step": 13230
+    },
+    {
+      "epoch": 3.5124120536306918,
+      "grad_norm": 0.4783513167964535,
+      "learning_rate": 1.8501425621025442e-06,
+      "loss": 0.5584,
+      "step": 13231
+    },
+    {
+      "epoch": 3.5126775521040754,
+      "grad_norm": 0.4628705015491312,
+      "learning_rate": 1.8498054421550503e-06,
+      "loss": 0.5116,
+      "step": 13232
+    },
+    {
+      "epoch": 3.512943050577459,
+      "grad_norm": 0.4273953725805247,
+      "learning_rate": 1.8494683348879425e-06,
+      "loss": 0.5605,
+      "step": 13233
+    },
+    {
+      "epoch": 3.513208549050843,
+      "grad_norm": 0.4632686395473244,
+      "learning_rate": 1.8491312403077949e-06,
+      "loss": 0.5283,
+      "step": 13234
+    },
+    {
+      "epoch": 3.5134740475242268,
+      "grad_norm": 0.44693044747784266,
+      "learning_rate": 1.8487941584211816e-06,
+      "loss": 0.5589,
+      "step": 13235
+    },
+    {
+      "epoch": 3.5137395459976104,
+      "grad_norm": 0.47825926918762085,
+      "learning_rate": 1.8484570892346773e-06,
+      "loss": 0.5395,
+      "step": 13236
+    },
+    {
+      "epoch": 3.514005044470994,
+      "grad_norm": 0.46580654070141414,
+      "learning_rate": 1.8481200327548543e-06,
+      "loss": 0.5442,
+      "step": 13237
+    },
+    {
+      "epoch": 3.514270542944378,
+      "grad_norm": 0.47061693879046246,
+      "learning_rate": 1.847782988988287e-06,
+      "loss": 0.5436,
+      "step": 13238
+    },
+    {
+      "epoch": 3.5145360414177618,
+      "grad_norm": 0.4424925894558598,
+      "learning_rate": 1.8474459579415483e-06,
+      "loss": 0.5084,
+      "step": 13239
+    },
+    {
+      "epoch": 3.5148015398911454,
+      "grad_norm": 0.46315892406069886,
+      "learning_rate": 1.847108939621211e-06,
+      "loss": 0.5303,
+      "step": 13240
+    },
+    {
+      "epoch": 3.5150670383645295,
+      "grad_norm": 0.4674993342925463,
+      "learning_rate": 1.8467719340338488e-06,
+      "loss": 0.5426,
+      "step": 13241
+    },
+    {
+      "epoch": 3.515332536837913,
+      "grad_norm": 0.4602981208284301,
+      "learning_rate": 1.8464349411860328e-06,
+      "loss": 0.5852,
+      "step": 13242
+    },
+    {
+      "epoch": 3.5155980353112968,
+      "grad_norm": 0.46143714845301903,
+      "learning_rate": 1.846097961084335e-06,
+      "loss": 0.5694,
+      "step": 13243
+    },
+    {
+      "epoch": 3.515863533784681,
+      "grad_norm": 0.4715205101564161,
+      "learning_rate": 1.8457609937353282e-06,
+      "loss": 0.5562,
+      "step": 13244
+    },
+    {
+      "epoch": 3.5161290322580645,
+      "grad_norm": 0.45765554411315496,
+      "learning_rate": 1.845424039145584e-06,
+      "loss": 0.5587,
+      "step": 13245
+    },
+    {
+      "epoch": 3.516394530731448,
+      "grad_norm": 0.4709391265367314,
+      "learning_rate": 1.8450870973216737e-06,
+      "loss": 0.5535,
+      "step": 13246
+    },
+    {
+      "epoch": 3.5166600292048322,
+      "grad_norm": 0.47008690742043996,
+      "learning_rate": 1.844750168270169e-06,
+      "loss": 0.5601,
+      "step": 13247
+    },
+    {
+      "epoch": 3.516925527678216,
+      "grad_norm": 0.47340117550938954,
+      "learning_rate": 1.8444132519976398e-06,
+      "loss": 0.5372,
+      "step": 13248
+    },
+    {
+      "epoch": 3.5171910261515995,
+      "grad_norm": 0.4457831933419617,
+      "learning_rate": 1.8440763485106571e-06,
+      "loss": 0.5237,
+      "step": 13249
+    },
+    {
+      "epoch": 3.5174565246249836,
+      "grad_norm": 0.45639209819098053,
+      "learning_rate": 1.843739457815792e-06,
+      "loss": 0.5234,
+      "step": 13250
+    },
+    {
+      "epoch": 3.5177220230983672,
+      "grad_norm": 0.45561845301128556,
+      "learning_rate": 1.8434025799196143e-06,
+      "loss": 0.5538,
+      "step": 13251
+    },
+    {
+      "epoch": 3.517987521571751,
+      "grad_norm": 0.4709719895481183,
+      "learning_rate": 1.8430657148286935e-06,
+      "loss": 0.5438,
+      "step": 13252
+    },
+    {
+      "epoch": 3.518253020045135,
+      "grad_norm": 0.45973882588537845,
+      "learning_rate": 1.842728862549601e-06,
+      "loss": 0.5702,
+      "step": 13253
+    },
+    {
+      "epoch": 3.5185185185185186,
+      "grad_norm": 0.45834895771424367,
+      "learning_rate": 1.8423920230889034e-06,
+      "loss": 0.5649,
+      "step": 13254
+    },
+    {
+      "epoch": 3.5187840169919022,
+      "grad_norm": 0.46665031091715986,
+      "learning_rate": 1.842055196453172e-06,
+      "loss": 0.5558,
+      "step": 13255
+    },
+    {
+      "epoch": 3.5190495154652863,
+      "grad_norm": 0.45325246521907353,
+      "learning_rate": 1.841718382648975e-06,
+      "loss": 0.5408,
+      "step": 13256
+    },
+    {
+      "epoch": 3.51931501393867,
+      "grad_norm": 0.470805498621211,
+      "learning_rate": 1.8413815816828814e-06,
+      "loss": 0.574,
+      "step": 13257
+    },
+    {
+      "epoch": 3.5195805124120536,
+      "grad_norm": 0.45198319067684056,
+      "learning_rate": 1.8410447935614594e-06,
+      "loss": 0.5724,
+      "step": 13258
+    },
+    {
+      "epoch": 3.5198460108854372,
+      "grad_norm": 0.4902640714583506,
+      "learning_rate": 1.840708018291279e-06,
+      "loss": 0.5593,
+      "step": 13259
+    },
+    {
+      "epoch": 3.5201115093588213,
+      "grad_norm": 0.4641633326195899,
+      "learning_rate": 1.8403712558789047e-06,
+      "loss": 0.5484,
+      "step": 13260
+    },
+    {
+      "epoch": 3.520377007832205,
+      "grad_norm": 0.4629416724239757,
+      "learning_rate": 1.8400345063309061e-06,
+      "loss": 0.5422,
+      "step": 13261
+    },
+    {
+      "epoch": 3.5206425063055886,
+      "grad_norm": 0.48338434642896316,
+      "learning_rate": 1.839697769653851e-06,
+      "loss": 0.56,
+      "step": 13262
+    },
+    {
+      "epoch": 3.5209080047789723,
+      "grad_norm": 0.47800442909365365,
+      "learning_rate": 1.8393610458543053e-06,
+      "loss": 0.572,
+      "step": 13263
+    },
+    {
+      "epoch": 3.5211735032523563,
+      "grad_norm": 0.46909204054396975,
+      "learning_rate": 1.8390243349388378e-06,
+      "loss": 0.5188,
+      "step": 13264
+    },
+    {
+      "epoch": 3.52143900172574,
+      "grad_norm": 0.4599590445302613,
+      "learning_rate": 1.8386876369140133e-06,
+      "loss": 0.536,
+      "step": 13265
+    },
+    {
+      "epoch": 3.5217045001991236,
+      "grad_norm": 0.4583910969891272,
+      "learning_rate": 1.8383509517863995e-06,
+      "loss": 0.5417,
+      "step": 13266
+    },
+    {
+      "epoch": 3.5219699986725077,
+      "grad_norm": 0.45847752569190753,
+      "learning_rate": 1.8380142795625616e-06,
+      "loss": 0.5387,
+      "step": 13267
+    },
+    {
+      "epoch": 3.5222354971458913,
+      "grad_norm": 0.4483954538654492,
+      "learning_rate": 1.8376776202490666e-06,
+      "loss": 0.5008,
+      "step": 13268
+    },
+    {
+      "epoch": 3.522500995619275,
+      "grad_norm": 0.484081330855067,
+      "learning_rate": 1.8373409738524792e-06,
+      "loss": 0.5649,
+      "step": 13269
+    },
+    {
+      "epoch": 3.522766494092659,
+      "grad_norm": 0.46437315742621915,
+      "learning_rate": 1.8370043403793664e-06,
+      "loss": 0.5676,
+      "step": 13270
+    },
+    {
+      "epoch": 3.5230319925660427,
+      "grad_norm": 0.47277666110222477,
+      "learning_rate": 1.8366677198362915e-06,
+      "loss": 0.5514,
+      "step": 13271
+    },
+    {
+      "epoch": 3.5232974910394264,
+      "grad_norm": 0.44436856196423663,
+      "learning_rate": 1.83633111222982e-06,
+      "loss": 0.5391,
+      "step": 13272
+    },
+    {
+      "epoch": 3.5235629895128104,
+      "grad_norm": 0.4668212535172757,
+      "learning_rate": 1.8359945175665168e-06,
+      "loss": 0.5115,
+      "step": 13273
+    },
+    {
+      "epoch": 3.523828487986194,
+      "grad_norm": 0.4670259967674747,
+      "learning_rate": 1.835657935852946e-06,
+      "loss": 0.5693,
+      "step": 13274
+    },
+    {
+      "epoch": 3.5240939864595777,
+      "grad_norm": 0.4613363608669837,
+      "learning_rate": 1.8353213670956715e-06,
+      "loss": 0.5382,
+      "step": 13275
+    },
+    {
+      "epoch": 3.524359484932962,
+      "grad_norm": 0.4713032662397352,
+      "learning_rate": 1.8349848113012596e-06,
+      "loss": 0.5622,
+      "step": 13276
+    },
+    {
+      "epoch": 3.5246249834063454,
+      "grad_norm": 0.470170149949293,
+      "learning_rate": 1.8346482684762711e-06,
+      "loss": 0.5837,
+      "step": 13277
+    },
+    {
+      "epoch": 3.524890481879729,
+      "grad_norm": 0.47276090716932484,
+      "learning_rate": 1.8343117386272704e-06,
+      "loss": 0.5413,
+      "step": 13278
+    },
+    {
+      "epoch": 3.525155980353113,
+      "grad_norm": 0.46894148530938673,
+      "learning_rate": 1.8339752217608206e-06,
+      "loss": 0.5307,
+      "step": 13279
+    },
+    {
+      "epoch": 3.525421478826497,
+      "grad_norm": 0.47187297321402716,
+      "learning_rate": 1.8336387178834846e-06,
+      "loss": 0.5284,
+      "step": 13280
+    },
+    {
+      "epoch": 3.5256869772998805,
+      "grad_norm": 0.4543840406280816,
+      "learning_rate": 1.8333022270018252e-06,
+      "loss": 0.5367,
+      "step": 13281
+    },
+    {
+      "epoch": 3.5259524757732645,
+      "grad_norm": 0.47936599613911385,
+      "learning_rate": 1.8329657491224056e-06,
+      "loss": 0.5662,
+      "step": 13282
+    },
+    {
+      "epoch": 3.526217974246648,
+      "grad_norm": 0.4625249162753347,
+      "learning_rate": 1.8326292842517862e-06,
+      "loss": 0.5406,
+      "step": 13283
+    },
+    {
+      "epoch": 3.526483472720032,
+      "grad_norm": 0.46943946690036775,
+      "learning_rate": 1.8322928323965296e-06,
+      "loss": 0.5277,
+      "step": 13284
+    },
+    {
+      "epoch": 3.526748971193416,
+      "grad_norm": 0.46599423073980717,
+      "learning_rate": 1.8319563935631978e-06,
+      "loss": 0.5462,
+      "step": 13285
+    },
+    {
+      "epoch": 3.5270144696667995,
+      "grad_norm": 0.4634110607645398,
+      "learning_rate": 1.831619967758352e-06,
+      "loss": 0.562,
+      "step": 13286
+    },
+    {
+      "epoch": 3.527279968140183,
+      "grad_norm": 0.463352001772562,
+      "learning_rate": 1.8312835549885533e-06,
+      "loss": 0.5208,
+      "step": 13287
+    },
+    {
+      "epoch": 3.527545466613567,
+      "grad_norm": 0.46889574647607046,
+      "learning_rate": 1.8309471552603635e-06,
+      "loss": 0.5265,
+      "step": 13288
+    },
+    {
+      "epoch": 3.527810965086951,
+      "grad_norm": 0.4777500723757507,
+      "learning_rate": 1.8306107685803414e-06,
+      "loss": 0.526,
+      "step": 13289
+    },
+    {
+      "epoch": 3.5280764635603346,
+      "grad_norm": 0.46716339708865007,
+      "learning_rate": 1.8302743949550482e-06,
+      "loss": 0.5448,
+      "step": 13290
+    },
+    {
+      "epoch": 3.528341962033718,
+      "grad_norm": 0.4665450385486929,
+      "learning_rate": 1.8299380343910444e-06,
+      "loss": 0.5611,
+      "step": 13291
+    },
+    {
+      "epoch": 3.528607460507102,
+      "grad_norm": 0.47304557238745437,
+      "learning_rate": 1.829601686894889e-06,
+      "loss": 0.572,
+      "step": 13292
+    },
+    {
+      "epoch": 3.528872958980486,
+      "grad_norm": 0.4527224953404673,
+      "learning_rate": 1.829265352473143e-06,
+      "loss": 0.5777,
+      "step": 13293
+    },
+    {
+      "epoch": 3.5291384574538696,
+      "grad_norm": 0.47273454086721106,
+      "learning_rate": 1.8289290311323643e-06,
+      "loss": 0.5353,
+      "step": 13294
+    },
+    {
+      "epoch": 3.529403955927253,
+      "grad_norm": 0.4766050228456209,
+      "learning_rate": 1.8285927228791122e-06,
+      "loss": 0.5744,
+      "step": 13295
+    },
+    {
+      "epoch": 3.5296694544006373,
+      "grad_norm": 0.4677897275648548,
+      "learning_rate": 1.8282564277199464e-06,
+      "loss": 0.5861,
+      "step": 13296
+    },
+    {
+      "epoch": 3.529934952874021,
+      "grad_norm": 0.4598820405046237,
+      "learning_rate": 1.8279201456614248e-06,
+      "loss": 0.5292,
+      "step": 13297
+    },
+    {
+      "epoch": 3.5302004513474046,
+      "grad_norm": 0.4598023766553167,
+      "learning_rate": 1.8275838767101057e-06,
+      "loss": 0.569,
+      "step": 13298
+    },
+    {
+      "epoch": 3.5304659498207887,
+      "grad_norm": 0.4746521254565789,
+      "learning_rate": 1.8272476208725488e-06,
+      "loss": 0.5104,
+      "step": 13299
+    },
+    {
+      "epoch": 3.5307314482941723,
+      "grad_norm": 0.44422174385585755,
+      "learning_rate": 1.8269113781553084e-06,
+      "loss": 0.5079,
+      "step": 13300
+    },
+    {
+      "epoch": 3.530996946767556,
+      "grad_norm": 0.4478682705457898,
+      "learning_rate": 1.826575148564945e-06,
+      "loss": 0.5537,
+      "step": 13301
+    },
+    {
+      "epoch": 3.53126244524094,
+      "grad_norm": 0.4483312850326237,
+      "learning_rate": 1.8262389321080149e-06,
+      "loss": 0.5324,
+      "step": 13302
+    },
+    {
+      "epoch": 3.5315279437143237,
+      "grad_norm": 0.4769387268999272,
+      "learning_rate": 1.8259027287910753e-06,
+      "loss": 0.5995,
+      "step": 13303
+    },
+    {
+      "epoch": 3.5317934421877073,
+      "grad_norm": 0.4615593927356842,
+      "learning_rate": 1.8255665386206828e-06,
+      "loss": 0.5621,
+      "step": 13304
+    },
+    {
+      "epoch": 3.5320589406610914,
+      "grad_norm": 0.4528036825045676,
+      "learning_rate": 1.8252303616033956e-06,
+      "loss": 0.5389,
+      "step": 13305
+    },
+    {
+      "epoch": 3.532324439134475,
+      "grad_norm": 0.44333208725803025,
+      "learning_rate": 1.8248941977457669e-06,
+      "loss": 0.5225,
+      "step": 13306
+    },
+    {
+      "epoch": 3.5325899376078587,
+      "grad_norm": 0.4667592701124061,
+      "learning_rate": 1.8245580470543546e-06,
+      "loss": 0.5749,
+      "step": 13307
+    },
+    {
+      "epoch": 3.5328554360812428,
+      "grad_norm": 0.46159898879308614,
+      "learning_rate": 1.8242219095357138e-06,
+      "loss": 0.5513,
+      "step": 13308
+    },
+    {
+      "epoch": 3.5331209345546264,
+      "grad_norm": 0.45633665558466274,
+      "learning_rate": 1.823885785196401e-06,
+      "loss": 0.5144,
+      "step": 13309
+    },
+    {
+      "epoch": 3.53338643302801,
+      "grad_norm": 0.46867442588239344,
+      "learning_rate": 1.8235496740429704e-06,
+      "loss": 0.541,
+      "step": 13310
+    },
+    {
+      "epoch": 3.533651931501394,
+      "grad_norm": 0.4573686548094638,
+      "learning_rate": 1.823213576081978e-06,
+      "loss": 0.5883,
+      "step": 13311
+    },
+    {
+      "epoch": 3.5339174299747778,
+      "grad_norm": 0.47876951728198947,
+      "learning_rate": 1.822877491319977e-06,
+      "loss": 0.546,
+      "step": 13312
+    },
+    {
+      "epoch": 3.5341829284481614,
+      "grad_norm": 0.46581391849534404,
+      "learning_rate": 1.8225414197635233e-06,
+      "loss": 0.5482,
+      "step": 13313
+    },
+    {
+      "epoch": 3.534448426921545,
+      "grad_norm": 0.45135367091489537,
+      "learning_rate": 1.8222053614191703e-06,
+      "loss": 0.5238,
+      "step": 13314
+    },
+    {
+      "epoch": 3.534713925394929,
+      "grad_norm": 0.47063411666491073,
+      "learning_rate": 1.8218693162934725e-06,
+      "loss": 0.6163,
+      "step": 13315
+    },
+    {
+      "epoch": 3.5349794238683128,
+      "grad_norm": 0.45800671457918946,
+      "learning_rate": 1.8215332843929844e-06,
+      "loss": 0.5251,
+      "step": 13316
+    },
+    {
+      "epoch": 3.5352449223416964,
+      "grad_norm": 0.46450362297552295,
+      "learning_rate": 1.8211972657242574e-06,
+      "loss": 0.5337,
+      "step": 13317
+    },
+    {
+      "epoch": 3.53551042081508,
+      "grad_norm": 0.46501257474805646,
+      "learning_rate": 1.8208612602938458e-06,
+      "loss": 0.5403,
+      "step": 13318
+    },
+    {
+      "epoch": 3.535775919288464,
+      "grad_norm": 0.4728168066816079,
+      "learning_rate": 1.8205252681083022e-06,
+      "loss": 0.5622,
+      "step": 13319
+    },
+    {
+      "epoch": 3.5360414177618478,
+      "grad_norm": 0.46987943080612654,
+      "learning_rate": 1.8201892891741796e-06,
+      "loss": 0.5315,
+      "step": 13320
+    },
+    {
+      "epoch": 3.5363069162352314,
+      "grad_norm": 0.46578639058516846,
+      "learning_rate": 1.8198533234980299e-06,
+      "loss": 0.5766,
+      "step": 13321
+    },
+    {
+      "epoch": 3.5365724147086155,
+      "grad_norm": 0.4612838261516672,
+      "learning_rate": 1.8195173710864072e-06,
+      "loss": 0.5404,
+      "step": 13322
+    },
+    {
+      "epoch": 3.536837913181999,
+      "grad_norm": 0.464509874657138,
+      "learning_rate": 1.8191814319458607e-06,
+      "loss": 0.5232,
+      "step": 13323
+    },
+    {
+      "epoch": 3.537103411655383,
+      "grad_norm": 0.46551828232540565,
+      "learning_rate": 1.8188455060829433e-06,
+      "loss": 0.5487,
+      "step": 13324
+    },
+    {
+      "epoch": 3.537368910128767,
+      "grad_norm": 0.4704121344797628,
+      "learning_rate": 1.8185095935042063e-06,
+      "loss": 0.5288,
+      "step": 13325
+    },
+    {
+      "epoch": 3.5376344086021505,
+      "grad_norm": 0.458157532290727,
+      "learning_rate": 1.818173694216201e-06,
+      "loss": 0.5308,
+      "step": 13326
+    },
+    {
+      "epoch": 3.537899907075534,
+      "grad_norm": 0.463476029786354,
+      "learning_rate": 1.8178378082254778e-06,
+      "loss": 0.5279,
+      "step": 13327
+    },
+    {
+      "epoch": 3.5381654055489182,
+      "grad_norm": 0.46049158692446607,
+      "learning_rate": 1.8175019355385881e-06,
+      "loss": 0.51,
+      "step": 13328
+    },
+    {
+      "epoch": 3.538430904022302,
+      "grad_norm": 0.46050540381666566,
+      "learning_rate": 1.8171660761620814e-06,
+      "loss": 0.5399,
+      "step": 13329
+    },
+    {
+      "epoch": 3.5386964024956855,
+      "grad_norm": 0.46728068136518575,
+      "learning_rate": 1.8168302301025076e-06,
+      "loss": 0.5721,
+      "step": 13330
+    },
+    {
+      "epoch": 3.5389619009690696,
+      "grad_norm": 0.45899014975898283,
+      "learning_rate": 1.8164943973664173e-06,
+      "loss": 0.5815,
+      "step": 13331
+    },
+    {
+      "epoch": 3.5392273994424532,
+      "grad_norm": 0.46072481237734786,
+      "learning_rate": 1.8161585779603597e-06,
+      "loss": 0.5524,
+      "step": 13332
+    },
+    {
+      "epoch": 3.539492897915837,
+      "grad_norm": 0.4624546498758775,
+      "learning_rate": 1.8158227718908844e-06,
+      "loss": 0.5309,
+      "step": 13333
+    },
+    {
+      "epoch": 3.539758396389221,
+      "grad_norm": 0.466793258928801,
+      "learning_rate": 1.8154869791645407e-06,
+      "loss": 0.5571,
+      "step": 13334
+    },
+    {
+      "epoch": 3.5400238948626046,
+      "grad_norm": 0.47937328580533517,
+      "learning_rate": 1.8151511997878765e-06,
+      "loss": 0.5315,
+      "step": 13335
+    },
+    {
+      "epoch": 3.5402893933359882,
+      "grad_norm": 0.4645636547339201,
+      "learning_rate": 1.8148154337674404e-06,
+      "loss": 0.4943,
+      "step": 13336
+    },
+    {
+      "epoch": 3.5405548918093723,
+      "grad_norm": 0.47437267430919217,
+      "learning_rate": 1.8144796811097808e-06,
+      "loss": 0.5421,
+      "step": 13337
+    },
+    {
+      "epoch": 3.540820390282756,
+      "grad_norm": 0.4572086120215907,
+      "learning_rate": 1.8141439418214462e-06,
+      "loss": 0.5388,
+      "step": 13338
+    },
+    {
+      "epoch": 3.5410858887561396,
+      "grad_norm": 0.46296689253857026,
+      "learning_rate": 1.8138082159089842e-06,
+      "loss": 0.5641,
+      "step": 13339
+    },
+    {
+      "epoch": 3.5413513872295237,
+      "grad_norm": 0.48060611012956767,
+      "learning_rate": 1.813472503378942e-06,
+      "loss": 0.5231,
+      "step": 13340
+    },
+    {
+      "epoch": 3.5416168857029073,
+      "grad_norm": 0.45447153441145993,
+      "learning_rate": 1.8131368042378667e-06,
+      "loss": 0.5404,
+      "step": 13341
+    },
+    {
+      "epoch": 3.541882384176291,
+      "grad_norm": 0.4619167152753581,
+      "learning_rate": 1.812801118492305e-06,
+      "loss": 0.5541,
+      "step": 13342
+    },
+    {
+      "epoch": 3.5421478826496746,
+      "grad_norm": 0.46565886796446704,
+      "learning_rate": 1.8124654461488044e-06,
+      "loss": 0.5262,
+      "step": 13343
+    },
+    {
+      "epoch": 3.5424133811230587,
+      "grad_norm": 0.454052094043848,
+      "learning_rate": 1.812129787213911e-06,
+      "loss": 0.5651,
+      "step": 13344
+    },
+    {
+      "epoch": 3.5426788795964423,
+      "grad_norm": 0.46296804636010425,
+      "learning_rate": 1.811794141694172e-06,
+      "loss": 0.5417,
+      "step": 13345
+    },
+    {
+      "epoch": 3.542944378069826,
+      "grad_norm": 0.4624527419520514,
+      "learning_rate": 1.8114585095961307e-06,
+      "loss": 0.5738,
+      "step": 13346
+    },
+    {
+      "epoch": 3.5432098765432096,
+      "grad_norm": 0.45415210932607414,
+      "learning_rate": 1.8111228909263349e-06,
+      "loss": 0.5565,
+      "step": 13347
+    },
+    {
+      "epoch": 3.5434753750165937,
+      "grad_norm": 0.458967212229915,
+      "learning_rate": 1.8107872856913293e-06,
+      "loss": 0.5361,
+      "step": 13348
+    },
+    {
+      "epoch": 3.5437408734899773,
+      "grad_norm": 0.4754983463032878,
+      "learning_rate": 1.8104516938976591e-06,
+      "loss": 0.5695,
+      "step": 13349
+    },
+    {
+      "epoch": 3.544006371963361,
+      "grad_norm": 0.44949523539603625,
+      "learning_rate": 1.8101161155518688e-06,
+      "loss": 0.5524,
+      "step": 13350
+    },
+    {
+      "epoch": 3.544271870436745,
+      "grad_norm": 0.4619064471165001,
+      "learning_rate": 1.8097805506605048e-06,
+      "loss": 0.5784,
+      "step": 13351
+    },
+    {
+      "epoch": 3.5445373689101287,
+      "grad_norm": 0.465932710885596,
+      "learning_rate": 1.8094449992301088e-06,
+      "loss": 0.5349,
+      "step": 13352
+    },
+    {
+      "epoch": 3.5448028673835124,
+      "grad_norm": 0.4430064312820136,
+      "learning_rate": 1.8091094612672267e-06,
+      "loss": 0.5495,
+      "step": 13353
+    },
+    {
+      "epoch": 3.5450683658568964,
+      "grad_norm": 0.4574567695152756,
+      "learning_rate": 1.808773936778401e-06,
+      "loss": 0.5439,
+      "step": 13354
+    },
+    {
+      "epoch": 3.54533386433028,
+      "grad_norm": 0.44565611113583403,
+      "learning_rate": 1.8084384257701762e-06,
+      "loss": 0.5222,
+      "step": 13355
+    },
+    {
+      "epoch": 3.5455993628036637,
+      "grad_norm": 0.4747693722507459,
+      "learning_rate": 1.8081029282490954e-06,
+      "loss": 0.5768,
+      "step": 13356
+    },
+    {
+      "epoch": 3.545864861277048,
+      "grad_norm": 0.4605451950163074,
+      "learning_rate": 1.807767444221702e-06,
+      "loss": 0.5553,
+      "step": 13357
+    },
+    {
+      "epoch": 3.5461303597504314,
+      "grad_norm": 0.47668760621679146,
+      "learning_rate": 1.8074319736945375e-06,
+      "loss": 0.55,
+      "step": 13358
+    },
+    {
+      "epoch": 3.546395858223815,
+      "grad_norm": 0.45765602951094014,
+      "learning_rate": 1.8070965166741456e-06,
+      "loss": 0.5563,
+      "step": 13359
+    },
+    {
+      "epoch": 3.546661356697199,
+      "grad_norm": 0.4426848554524709,
+      "learning_rate": 1.806761073167068e-06,
+      "loss": 0.5485,
+      "step": 13360
+    },
+    {
+      "epoch": 3.546926855170583,
+      "grad_norm": 0.4614411963566644,
+      "learning_rate": 1.8064256431798468e-06,
+      "loss": 0.4962,
+      "step": 13361
+    },
+    {
+      "epoch": 3.5471923536439665,
+      "grad_norm": 0.46302872444446624,
+      "learning_rate": 1.806090226719025e-06,
+      "loss": 0.573,
+      "step": 13362
+    },
+    {
+      "epoch": 3.5474578521173505,
+      "grad_norm": 0.4654944901248699,
+      "learning_rate": 1.8057548237911415e-06,
+      "loss": 0.5523,
+      "step": 13363
+    },
+    {
+      "epoch": 3.547723350590734,
+      "grad_norm": 0.46221678040871367,
+      "learning_rate": 1.8054194344027386e-06,
+      "loss": 0.5587,
+      "step": 13364
+    },
+    {
+      "epoch": 3.547988849064118,
+      "grad_norm": 0.48212872324430195,
+      "learning_rate": 1.8050840585603575e-06,
+      "loss": 0.5835,
+      "step": 13365
+    },
+    {
+      "epoch": 3.548254347537502,
+      "grad_norm": 0.4677183548544849,
+      "learning_rate": 1.8047486962705382e-06,
+      "loss": 0.5494,
+      "step": 13366
+    },
+    {
+      "epoch": 3.5485198460108855,
+      "grad_norm": 0.4734338976072318,
+      "learning_rate": 1.8044133475398224e-06,
+      "loss": 0.5296,
+      "step": 13367
+    },
+    {
+      "epoch": 3.548785344484269,
+      "grad_norm": 0.47970326485162174,
+      "learning_rate": 1.8040780123747497e-06,
+      "loss": 0.5268,
+      "step": 13368
+    },
+    {
+      "epoch": 3.549050842957653,
+      "grad_norm": 0.45692631691054686,
+      "learning_rate": 1.8037426907818595e-06,
+      "loss": 0.5367,
+      "step": 13369
+    },
+    {
+      "epoch": 3.549316341431037,
+      "grad_norm": 0.45819299834256083,
+      "learning_rate": 1.803407382767691e-06,
+      "loss": 0.5437,
+      "step": 13370
+    },
+    {
+      "epoch": 3.5495818399044206,
+      "grad_norm": 0.4606304405507358,
+      "learning_rate": 1.8030720883387843e-06,
+      "loss": 0.5561,
+      "step": 13371
+    },
+    {
+      "epoch": 3.549847338377804,
+      "grad_norm": 0.4512011432969114,
+      "learning_rate": 1.8027368075016783e-06,
+      "loss": 0.5257,
+      "step": 13372
+    },
+    {
+      "epoch": 3.550112836851188,
+      "grad_norm": 0.4446618314979587,
+      "learning_rate": 1.8024015402629114e-06,
+      "loss": 0.5242,
+      "step": 13373
+    },
+    {
+      "epoch": 3.550378335324572,
+      "grad_norm": 0.4584723772905301,
+      "learning_rate": 1.8020662866290234e-06,
+      "loss": 0.5497,
+      "step": 13374
+    },
+    {
+      "epoch": 3.5506438337979556,
+      "grad_norm": 0.45836627776091277,
+      "learning_rate": 1.8017310466065508e-06,
+      "loss": 0.5582,
+      "step": 13375
+    },
+    {
+      "epoch": 3.550909332271339,
+      "grad_norm": 0.4586827143790453,
+      "learning_rate": 1.8013958202020327e-06,
+      "loss": 0.5259,
+      "step": 13376
+    },
+    {
+      "epoch": 3.5511748307447233,
+      "grad_norm": 0.4822886002524828,
+      "learning_rate": 1.8010606074220066e-06,
+      "loss": 0.5386,
+      "step": 13377
+    },
+    {
+      "epoch": 3.551440329218107,
+      "grad_norm": 0.4491399056760416,
+      "learning_rate": 1.8007254082730098e-06,
+      "loss": 0.568,
+      "step": 13378
+    },
+    {
+      "epoch": 3.5517058276914906,
+      "grad_norm": 0.4590620565459339,
+      "learning_rate": 1.8003902227615797e-06,
+      "loss": 0.4901,
+      "step": 13379
+    },
+    {
+      "epoch": 3.5519713261648747,
+      "grad_norm": 0.4658584849920062,
+      "learning_rate": 1.8000550508942543e-06,
+      "loss": 0.5742,
+      "step": 13380
+    },
+    {
+      "epoch": 3.5522368246382583,
+      "grad_norm": 0.4791880490889075,
+      "learning_rate": 1.799719892677568e-06,
+      "loss": 0.5765,
+      "step": 13381
+    },
+    {
+      "epoch": 3.552502323111642,
+      "grad_norm": 0.45790078740687573,
+      "learning_rate": 1.7993847481180584e-06,
+      "loss": 0.5211,
+      "step": 13382
+    },
+    {
+      "epoch": 3.552767821585026,
+      "grad_norm": 0.4603405802422378,
+      "learning_rate": 1.7990496172222618e-06,
+      "loss": 0.5659,
+      "step": 13383
+    },
+    {
+      "epoch": 3.5530333200584097,
+      "grad_norm": 0.47657050938825063,
+      "learning_rate": 1.7987144999967138e-06,
+      "loss": 0.588,
+      "step": 13384
+    },
+    {
+      "epoch": 3.5532988185317933,
+      "grad_norm": 0.4820428828671815,
+      "learning_rate": 1.7983793964479506e-06,
+      "loss": 0.5881,
+      "step": 13385
+    },
+    {
+      "epoch": 3.5535643170051774,
+      "grad_norm": 0.47247994579564584,
+      "learning_rate": 1.7980443065825064e-06,
+      "loss": 0.5497,
+      "step": 13386
+    },
+    {
+      "epoch": 3.553829815478561,
+      "grad_norm": 0.4628977112928438,
+      "learning_rate": 1.7977092304069172e-06,
+      "loss": 0.5633,
+      "step": 13387
+    },
+    {
+      "epoch": 3.5540953139519447,
+      "grad_norm": 0.4603827851613823,
+      "learning_rate": 1.797374167927717e-06,
+      "loss": 0.5617,
+      "step": 13388
+    },
+    {
+      "epoch": 3.5543608124253288,
+      "grad_norm": 0.4685640936464193,
+      "learning_rate": 1.7970391191514413e-06,
+      "loss": 0.5288,
+      "step": 13389
+    },
+    {
+      "epoch": 3.5546263108987124,
+      "grad_norm": 0.49156489185667773,
+      "learning_rate": 1.7967040840846238e-06,
+      "loss": 0.5546,
+      "step": 13390
+    },
+    {
+      "epoch": 3.554891809372096,
+      "grad_norm": 0.48724063683434726,
+      "learning_rate": 1.7963690627337991e-06,
+      "loss": 0.5267,
+      "step": 13391
+    },
+    {
+      "epoch": 3.55515730784548,
+      "grad_norm": 0.4621055018985786,
+      "learning_rate": 1.7960340551055e-06,
+      "loss": 0.5234,
+      "step": 13392
+    },
+    {
+      "epoch": 3.5554228063188638,
+      "grad_norm": 0.4572332916654144,
+      "learning_rate": 1.7956990612062603e-06,
+      "loss": 0.5884,
+      "step": 13393
+    },
+    {
+      "epoch": 3.5556883047922474,
+      "grad_norm": 0.45258758492541956,
+      "learning_rate": 1.7953640810426134e-06,
+      "loss": 0.5837,
+      "step": 13394
+    },
+    {
+      "epoch": 3.5559538032656315,
+      "grad_norm": 0.47264031210983315,
+      "learning_rate": 1.7950291146210921e-06,
+      "loss": 0.5471,
+      "step": 13395
+    },
+    {
+      "epoch": 3.556219301739015,
+      "grad_norm": 0.4571966160127003,
+      "learning_rate": 1.7946941619482295e-06,
+      "loss": 0.5404,
+      "step": 13396
+    },
+    {
+      "epoch": 3.5564848002123988,
+      "grad_norm": 0.4525961944234777,
+      "learning_rate": 1.7943592230305583e-06,
+      "loss": 0.5436,
+      "step": 13397
+    },
+    {
+      "epoch": 3.5567502986857824,
+      "grad_norm": 0.47068183889238013,
+      "learning_rate": 1.794024297874609e-06,
+      "loss": 0.5781,
+      "step": 13398
+    },
+    {
+      "epoch": 3.5570157971591665,
+      "grad_norm": 0.4685688698136096,
+      "learning_rate": 1.7936893864869143e-06,
+      "loss": 0.5266,
+      "step": 13399
+    },
+    {
+      "epoch": 3.55728129563255,
+      "grad_norm": 0.46791732086375687,
+      "learning_rate": 1.7933544888740062e-06,
+      "loss": 0.5597,
+      "step": 13400
+    },
+    {
+      "epoch": 3.5575467941059338,
+      "grad_norm": 0.46104317132614264,
+      "learning_rate": 1.7930196050424153e-06,
+      "loss": 0.5126,
+      "step": 13401
+    },
+    {
+      "epoch": 3.5578122925793174,
+      "grad_norm": 0.4510859060957634,
+      "learning_rate": 1.7926847349986738e-06,
+      "loss": 0.5766,
+      "step": 13402
+    },
+    {
+      "epoch": 3.5580777910527015,
+      "grad_norm": 0.4663932058041748,
+      "learning_rate": 1.7923498787493115e-06,
+      "loss": 0.6024,
+      "step": 13403
+    },
+    {
+      "epoch": 3.558343289526085,
+      "grad_norm": 0.46559606932309877,
+      "learning_rate": 1.7920150363008592e-06,
+      "loss": 0.5404,
+      "step": 13404
+    },
+    {
+      "epoch": 3.558608787999469,
+      "grad_norm": 0.46039578516232343,
+      "learning_rate": 1.791680207659847e-06,
+      "loss": 0.552,
+      "step": 13405
+    },
+    {
+      "epoch": 3.558874286472853,
+      "grad_norm": 0.4595817941004434,
+      "learning_rate": 1.7913453928328047e-06,
+      "loss": 0.5589,
+      "step": 13406
+    },
+    {
+      "epoch": 3.5591397849462365,
+      "grad_norm": 0.4494892316142684,
+      "learning_rate": 1.7910105918262627e-06,
+      "loss": 0.5478,
+      "step": 13407
+    },
+    {
+      "epoch": 3.55940528341962,
+      "grad_norm": 0.47063165969361465,
+      "learning_rate": 1.7906758046467499e-06,
+      "loss": 0.5527,
+      "step": 13408
+    },
+    {
+      "epoch": 3.5596707818930042,
+      "grad_norm": 0.4580673824266431,
+      "learning_rate": 1.7903410313007968e-06,
+      "loss": 0.4956,
+      "step": 13409
+    },
+    {
+      "epoch": 3.559936280366388,
+      "grad_norm": 0.43336700747446555,
+      "learning_rate": 1.79000627179493e-06,
+      "loss": 0.558,
+      "step": 13410
+    },
+    {
+      "epoch": 3.5602017788397715,
+      "grad_norm": 0.46408108982588064,
+      "learning_rate": 1.7896715261356795e-06,
+      "loss": 0.5732,
+      "step": 13411
+    },
+    {
+      "epoch": 3.5604672773131556,
+      "grad_norm": 0.4632723039607594,
+      "learning_rate": 1.7893367943295725e-06,
+      "loss": 0.5676,
+      "step": 13412
+    },
+    {
+      "epoch": 3.5607327757865392,
+      "grad_norm": 0.4647143154743714,
+      "learning_rate": 1.789002076383139e-06,
+      "loss": 0.5426,
+      "step": 13413
+    },
+    {
+      "epoch": 3.560998274259923,
+      "grad_norm": 0.4691977312057472,
+      "learning_rate": 1.7886673723029063e-06,
+      "loss": 0.5492,
+      "step": 13414
+    },
+    {
+      "epoch": 3.561263772733307,
+      "grad_norm": 0.47854810178648244,
+      "learning_rate": 1.7883326820954008e-06,
+      "loss": 0.5756,
+      "step": 13415
+    },
+    {
+      "epoch": 3.5615292712066906,
+      "grad_norm": 0.4653652352121288,
+      "learning_rate": 1.7879980057671503e-06,
+      "loss": 0.546,
+      "step": 13416
+    },
+    {
+      "epoch": 3.5617947696800742,
+      "grad_norm": 0.46469167744999185,
+      "learning_rate": 1.7876633433246823e-06,
+      "loss": 0.5435,
+      "step": 13417
+    },
+    {
+      "epoch": 3.5620602681534583,
+      "grad_norm": 0.4747169375924761,
+      "learning_rate": 1.787328694774523e-06,
+      "loss": 0.5287,
+      "step": 13418
+    },
+    {
+      "epoch": 3.562325766626842,
+      "grad_norm": 0.4712499618288174,
+      "learning_rate": 1.7869940601231989e-06,
+      "loss": 0.5247,
+      "step": 13419
+    },
+    {
+      "epoch": 3.5625912651002256,
+      "grad_norm": 0.4685053149937222,
+      "learning_rate": 1.7866594393772375e-06,
+      "loss": 0.5649,
+      "step": 13420
+    },
+    {
+      "epoch": 3.5628567635736097,
+      "grad_norm": 0.46585725039815645,
+      "learning_rate": 1.7863248325431622e-06,
+      "loss": 0.5578,
+      "step": 13421
+    },
+    {
+      "epoch": 3.5631222620469933,
+      "grad_norm": 0.47506420689991613,
+      "learning_rate": 1.7859902396275006e-06,
+      "loss": 0.5072,
+      "step": 13422
+    },
+    {
+      "epoch": 3.563387760520377,
+      "grad_norm": 0.45989845537698193,
+      "learning_rate": 1.7856556606367776e-06,
+      "loss": 0.5701,
+      "step": 13423
+    },
+    {
+      "epoch": 3.5636532589937606,
+      "grad_norm": 0.4545099109375478,
+      "learning_rate": 1.785321095577518e-06,
+      "loss": 0.5377,
+      "step": 13424
+    },
+    {
+      "epoch": 3.5639187574671447,
+      "grad_norm": 0.4671811360140433,
+      "learning_rate": 1.7849865444562466e-06,
+      "loss": 0.5417,
+      "step": 13425
+    },
+    {
+      "epoch": 3.5641842559405283,
+      "grad_norm": 0.4551352377572186,
+      "learning_rate": 1.7846520072794899e-06,
+      "loss": 0.5207,
+      "step": 13426
+    },
+    {
+      "epoch": 3.564449754413912,
+      "grad_norm": 0.4392007843676805,
+      "learning_rate": 1.7843174840537691e-06,
+      "loss": 0.5423,
+      "step": 13427
+    },
+    {
+      "epoch": 3.5647152528872956,
+      "grad_norm": 0.4632468758790232,
+      "learning_rate": 1.7839829747856096e-06,
+      "loss": 0.5118,
+      "step": 13428
+    },
+    {
+      "epoch": 3.5649807513606797,
+      "grad_norm": 0.452307858788729,
+      "learning_rate": 1.7836484794815354e-06,
+      "loss": 0.5903,
+      "step": 13429
+    },
+    {
+      "epoch": 3.5652462498340634,
+      "grad_norm": 0.47923715578627024,
+      "learning_rate": 1.7833139981480695e-06,
+      "loss": 0.5838,
+      "step": 13430
+    },
+    {
+      "epoch": 3.565511748307447,
+      "grad_norm": 0.4543391655277271,
+      "learning_rate": 1.7829795307917359e-06,
+      "loss": 0.5345,
+      "step": 13431
+    },
+    {
+      "epoch": 3.565777246780831,
+      "grad_norm": 0.46217600378670654,
+      "learning_rate": 1.782645077419057e-06,
+      "loss": 0.5208,
+      "step": 13432
+    },
+    {
+      "epoch": 3.5660427452542147,
+      "grad_norm": 0.4553002923613811,
+      "learning_rate": 1.782310638036555e-06,
+      "loss": 0.5692,
+      "step": 13433
+    },
+    {
+      "epoch": 3.5663082437275984,
+      "grad_norm": 0.46997920258942016,
+      "learning_rate": 1.7819762126507531e-06,
+      "loss": 0.5726,
+      "step": 13434
+    },
+    {
+      "epoch": 3.5665737422009824,
+      "grad_norm": 0.45042019117233795,
+      "learning_rate": 1.7816418012681729e-06,
+      "loss": 0.5507,
+      "step": 13435
+    },
+    {
+      "epoch": 3.566839240674366,
+      "grad_norm": 0.4646775304034689,
+      "learning_rate": 1.7813074038953365e-06,
+      "loss": 0.5105,
+      "step": 13436
+    },
+    {
+      "epoch": 3.5671047391477497,
+      "grad_norm": 0.46897989614462854,
+      "learning_rate": 1.780973020538766e-06,
+      "loss": 0.5315,
+      "step": 13437
+    },
+    {
+      "epoch": 3.567370237621134,
+      "grad_norm": 0.45409252558744917,
+      "learning_rate": 1.7806386512049812e-06,
+      "loss": 0.531,
+      "step": 13438
+    },
+    {
+      "epoch": 3.5676357360945175,
+      "grad_norm": 0.4485736420246746,
+      "learning_rate": 1.7803042959005044e-06,
+      "loss": 0.5159,
+      "step": 13439
+    },
+    {
+      "epoch": 3.567901234567901,
+      "grad_norm": 0.44362931415859147,
+      "learning_rate": 1.7799699546318554e-06,
+      "loss": 0.5182,
+      "step": 13440
+    },
+    {
+      "epoch": 3.568166733041285,
+      "grad_norm": 0.4738063188948455,
+      "learning_rate": 1.779635627405556e-06,
+      "loss": 0.5762,
+      "step": 13441
+    },
+    {
+      "epoch": 3.568432231514669,
+      "grad_norm": 0.49126397181579917,
+      "learning_rate": 1.7793013142281252e-06,
+      "loss": 0.5744,
+      "step": 13442
+    },
+    {
+      "epoch": 3.5686977299880525,
+      "grad_norm": 0.48246570501670905,
+      "learning_rate": 1.7789670151060844e-06,
+      "loss": 0.5727,
+      "step": 13443
+    },
+    {
+      "epoch": 3.5689632284614365,
+      "grad_norm": 0.4698829369249473,
+      "learning_rate": 1.778632730045951e-06,
+      "loss": 0.5378,
+      "step": 13444
+    },
+    {
+      "epoch": 3.56922872693482,
+      "grad_norm": 0.4462295579486581,
+      "learning_rate": 1.7782984590542457e-06,
+      "loss": 0.5196,
+      "step": 13445
+    },
+    {
+      "epoch": 3.569494225408204,
+      "grad_norm": 0.4586111211394395,
+      "learning_rate": 1.7779642021374878e-06,
+      "loss": 0.5242,
+      "step": 13446
+    },
+    {
+      "epoch": 3.569759723881588,
+      "grad_norm": 0.46778783325941115,
+      "learning_rate": 1.7776299593021954e-06,
+      "loss": 0.5402,
+      "step": 13447
+    },
+    {
+      "epoch": 3.5700252223549716,
+      "grad_norm": 0.46404866019170077,
+      "learning_rate": 1.7772957305548878e-06,
+      "loss": 0.5167,
+      "step": 13448
+    },
+    {
+      "epoch": 3.570290720828355,
+      "grad_norm": 0.4689455599244113,
+      "learning_rate": 1.776961515902083e-06,
+      "loss": 0.5509,
+      "step": 13449
+    },
+    {
+      "epoch": 3.5705562193017393,
+      "grad_norm": 0.46995535060411964,
+      "learning_rate": 1.7766273153502987e-06,
+      "loss": 0.5289,
+      "step": 13450
+    },
+    {
+      "epoch": 3.570821717775123,
+      "grad_norm": 0.46512019045487785,
+      "learning_rate": 1.7762931289060526e-06,
+      "loss": 0.5431,
+      "step": 13451
+    },
+    {
+      "epoch": 3.5710872162485066,
+      "grad_norm": 0.4533239612693231,
+      "learning_rate": 1.7759589565758628e-06,
+      "loss": 0.5519,
+      "step": 13452
+    },
+    {
+      "epoch": 3.57135271472189,
+      "grad_norm": 0.47385085007076516,
+      "learning_rate": 1.7756247983662462e-06,
+      "loss": 0.5228,
+      "step": 13453
+    },
+    {
+      "epoch": 3.5716182131952743,
+      "grad_norm": 0.4715054638091034,
+      "learning_rate": 1.775290654283719e-06,
+      "loss": 0.5533,
+      "step": 13454
+    },
+    {
+      "epoch": 3.571883711668658,
+      "grad_norm": 0.46916100755297774,
+      "learning_rate": 1.7749565243348e-06,
+      "loss": 0.545,
+      "step": 13455
+    },
+    {
+      "epoch": 3.5721492101420416,
+      "grad_norm": 0.49002467578488784,
+      "learning_rate": 1.7746224085260027e-06,
+      "loss": 0.5574,
+      "step": 13456
+    },
+    {
+      "epoch": 3.572414708615425,
+      "grad_norm": 0.4573229806166725,
+      "learning_rate": 1.7742883068638447e-06,
+      "loss": 0.5359,
+      "step": 13457
+    },
+    {
+      "epoch": 3.5726802070888093,
+      "grad_norm": 0.47149043313575806,
+      "learning_rate": 1.7739542193548409e-06,
+      "loss": 0.5477,
+      "step": 13458
+    },
+    {
+      "epoch": 3.572945705562193,
+      "grad_norm": 0.4713092628733232,
+      "learning_rate": 1.7736201460055081e-06,
+      "loss": 0.5601,
+      "step": 13459
+    },
+    {
+      "epoch": 3.5732112040355766,
+      "grad_norm": 0.4632081871651939,
+      "learning_rate": 1.7732860868223618e-06,
+      "loss": 0.5214,
+      "step": 13460
+    },
+    {
+      "epoch": 3.5734767025089607,
+      "grad_norm": 0.46795326791708436,
+      "learning_rate": 1.772952041811915e-06,
+      "loss": 0.5181,
+      "step": 13461
+    },
+    {
+      "epoch": 3.5737422009823443,
+      "grad_norm": 0.4642618365293363,
+      "learning_rate": 1.772618010980684e-06,
+      "loss": 0.5059,
+      "step": 13462
+    },
+    {
+      "epoch": 3.574007699455728,
+      "grad_norm": 0.45538823820263696,
+      "learning_rate": 1.772283994335182e-06,
+      "loss": 0.544,
+      "step": 13463
+    },
+    {
+      "epoch": 3.574273197929112,
+      "grad_norm": 0.4497995411658983,
+      "learning_rate": 1.7719499918819245e-06,
+      "loss": 0.5492,
+      "step": 13464
+    },
+    {
+      "epoch": 3.5745386964024957,
+      "grad_norm": 0.44930793691615756,
+      "learning_rate": 1.7716160036274244e-06,
+      "loss": 0.5199,
+      "step": 13465
+    },
+    {
+      "epoch": 3.5748041948758793,
+      "grad_norm": 0.4656445415602291,
+      "learning_rate": 1.7712820295781964e-06,
+      "loss": 0.5162,
+      "step": 13466
+    },
+    {
+      "epoch": 3.5750696933492634,
+      "grad_norm": 0.45028278674883476,
+      "learning_rate": 1.7709480697407522e-06,
+      "loss": 0.5173,
+      "step": 13467
+    },
+    {
+      "epoch": 3.575335191822647,
+      "grad_norm": 0.46591458074512626,
+      "learning_rate": 1.7706141241216058e-06,
+      "loss": 0.5449,
+      "step": 13468
+    },
+    {
+      "epoch": 3.5756006902960307,
+      "grad_norm": 0.44660273366065795,
+      "learning_rate": 1.77028019272727e-06,
+      "loss": 0.533,
+      "step": 13469
+    },
+    {
+      "epoch": 3.5758661887694148,
+      "grad_norm": 0.4772883622271824,
+      "learning_rate": 1.769946275564257e-06,
+      "loss": 0.5429,
+      "step": 13470
+    },
+    {
+      "epoch": 3.5761316872427984,
+      "grad_norm": 0.46090785423914027,
+      "learning_rate": 1.7696123726390791e-06,
+      "loss": 0.5291,
+      "step": 13471
+    },
+    {
+      "epoch": 3.576397185716182,
+      "grad_norm": 0.45780964091687737,
+      "learning_rate": 1.769278483958249e-06,
+      "loss": 0.5564,
+      "step": 13472
+    },
+    {
+      "epoch": 3.576662684189566,
+      "grad_norm": 0.48494254412986126,
+      "learning_rate": 1.7689446095282767e-06,
+      "loss": 0.5765,
+      "step": 13473
+    },
+    {
+      "epoch": 3.5769281826629498,
+      "grad_norm": 0.46426098328621784,
+      "learning_rate": 1.7686107493556745e-06,
+      "loss": 0.5181,
+      "step": 13474
+    },
+    {
+      "epoch": 3.5771936811363334,
+      "grad_norm": 0.4532630970030804,
+      "learning_rate": 1.7682769034469535e-06,
+      "loss": 0.561,
+      "step": 13475
+    },
+    {
+      "epoch": 3.5774591796097175,
+      "grad_norm": 0.4650211973398225,
+      "learning_rate": 1.7679430718086244e-06,
+      "loss": 0.5805,
+      "step": 13476
+    },
+    {
+      "epoch": 3.577724678083101,
+      "grad_norm": 0.470134843266266,
+      "learning_rate": 1.767609254447198e-06,
+      "loss": 0.5227,
+      "step": 13477
+    },
+    {
+      "epoch": 3.5779901765564848,
+      "grad_norm": 0.45509087622096434,
+      "learning_rate": 1.7672754513691842e-06,
+      "loss": 0.5653,
+      "step": 13478
+    },
+    {
+      "epoch": 3.578255675029869,
+      "grad_norm": 0.4501670867602088,
+      "learning_rate": 1.766941662581093e-06,
+      "loss": 0.5434,
+      "step": 13479
+    },
+    {
+      "epoch": 3.5785211735032525,
+      "grad_norm": 0.47383790659574726,
+      "learning_rate": 1.7666078880894344e-06,
+      "loss": 0.5778,
+      "step": 13480
+    },
+    {
+      "epoch": 3.578786671976636,
+      "grad_norm": 0.47332026242927877,
+      "learning_rate": 1.7662741279007175e-06,
+      "loss": 0.5381,
+      "step": 13481
+    },
+    {
+      "epoch": 3.5790521704500198,
+      "grad_norm": 0.4630851023373439,
+      "learning_rate": 1.7659403820214516e-06,
+      "loss": 0.5707,
+      "step": 13482
+    },
+    {
+      "epoch": 3.5793176689234034,
+      "grad_norm": 0.4569584486873443,
+      "learning_rate": 1.765606650458146e-06,
+      "loss": 0.5771,
+      "step": 13483
+    },
+    {
+      "epoch": 3.5795831673967875,
+      "grad_norm": 0.45945231131832176,
+      "learning_rate": 1.7652729332173082e-06,
+      "loss": 0.5387,
+      "step": 13484
+    },
+    {
+      "epoch": 3.579848665870171,
+      "grad_norm": 0.46590589802264126,
+      "learning_rate": 1.7649392303054475e-06,
+      "loss": 0.5704,
+      "step": 13485
+    },
+    {
+      "epoch": 3.580114164343555,
+      "grad_norm": 0.4705960375163898,
+      "learning_rate": 1.7646055417290712e-06,
+      "loss": 0.5646,
+      "step": 13486
+    },
+    {
+      "epoch": 3.580379662816939,
+      "grad_norm": 0.4708874203168277,
+      "learning_rate": 1.7642718674946875e-06,
+      "loss": 0.5523,
+      "step": 13487
+    },
+    {
+      "epoch": 3.5806451612903225,
+      "grad_norm": 0.45605608095120925,
+      "learning_rate": 1.7639382076088036e-06,
+      "loss": 0.4931,
+      "step": 13488
+    },
+    {
+      "epoch": 3.580910659763706,
+      "grad_norm": 0.4669148709602922,
+      "learning_rate": 1.763604562077928e-06,
+      "loss": 0.5468,
+      "step": 13489
+    },
+    {
+      "epoch": 3.5811761582370902,
+      "grad_norm": 0.46088999239476847,
+      "learning_rate": 1.7632709309085656e-06,
+      "loss": 0.5757,
+      "step": 13490
+    },
+    {
+      "epoch": 3.581441656710474,
+      "grad_norm": 0.4836082446183638,
+      "learning_rate": 1.7629373141072237e-06,
+      "loss": 0.5408,
+      "step": 13491
+    },
+    {
+      "epoch": 3.5817071551838575,
+      "grad_norm": 0.4306765177362808,
+      "learning_rate": 1.7626037116804087e-06,
+      "loss": 0.5186,
+      "step": 13492
+    },
+    {
+      "epoch": 3.5819726536572416,
+      "grad_norm": 0.4629337087409011,
+      "learning_rate": 1.7622701236346267e-06,
+      "loss": 0.555,
+      "step": 13493
+    },
+    {
+      "epoch": 3.5822381521306252,
+      "grad_norm": 0.4728193109940487,
+      "learning_rate": 1.7619365499763841e-06,
+      "loss": 0.5235,
+      "step": 13494
+    },
+    {
+      "epoch": 3.582503650604009,
+      "grad_norm": 0.4584089852387107,
+      "learning_rate": 1.761602990712186e-06,
+      "loss": 0.5403,
+      "step": 13495
+    },
+    {
+      "epoch": 3.582769149077393,
+      "grad_norm": 0.4696528660413869,
+      "learning_rate": 1.7612694458485368e-06,
+      "loss": 0.5293,
+      "step": 13496
+    },
+    {
+      "epoch": 3.5830346475507766,
+      "grad_norm": 0.45349439629781474,
+      "learning_rate": 1.7609359153919426e-06,
+      "loss": 0.5565,
+      "step": 13497
+    },
+    {
+      "epoch": 3.5833001460241602,
+      "grad_norm": 0.46409306959296853,
+      "learning_rate": 1.7606023993489071e-06,
+      "loss": 0.5087,
+      "step": 13498
+    },
+    {
+      "epoch": 3.5835656444975443,
+      "grad_norm": 0.45655254362140885,
+      "learning_rate": 1.7602688977259354e-06,
+      "loss": 0.5362,
+      "step": 13499
+    },
+    {
+      "epoch": 3.583831142970928,
+      "grad_norm": 0.4609754259998067,
+      "learning_rate": 1.7599354105295313e-06,
+      "loss": 0.5014,
+      "step": 13500
+    },
+    {
+      "epoch": 3.5840966414443116,
+      "grad_norm": 0.4585782267669953,
+      "learning_rate": 1.7596019377662e-06,
+      "loss": 0.5489,
+      "step": 13501
+    },
+    {
+      "epoch": 3.5843621399176957,
+      "grad_norm": 0.45110713391992563,
+      "learning_rate": 1.7592684794424425e-06,
+      "loss": 0.5418,
+      "step": 13502
+    },
+    {
+      "epoch": 3.5846276383910793,
+      "grad_norm": 0.47947962867186145,
+      "learning_rate": 1.7589350355647628e-06,
+      "loss": 0.5563,
+      "step": 13503
+    },
+    {
+      "epoch": 3.584893136864463,
+      "grad_norm": 0.4614398257400887,
+      "learning_rate": 1.7586016061396646e-06,
+      "loss": 0.5509,
+      "step": 13504
+    },
+    {
+      "epoch": 3.585158635337847,
+      "grad_norm": 0.4494460046415863,
+      "learning_rate": 1.7582681911736505e-06,
+      "loss": 0.5392,
+      "step": 13505
+    },
+    {
+      "epoch": 3.5854241338112307,
+      "grad_norm": 0.4501420768803699,
+      "learning_rate": 1.7579347906732238e-06,
+      "loss": 0.5527,
+      "step": 13506
+    },
+    {
+      "epoch": 3.5856896322846143,
+      "grad_norm": 0.4569535939121678,
+      "learning_rate": 1.7576014046448846e-06,
+      "loss": 0.5068,
+      "step": 13507
+    },
+    {
+      "epoch": 3.585955130757998,
+      "grad_norm": 0.44355688328757864,
+      "learning_rate": 1.7572680330951359e-06,
+      "loss": 0.5525,
+      "step": 13508
+    },
+    {
+      "epoch": 3.586220629231382,
+      "grad_norm": 0.4720060511626888,
+      "learning_rate": 1.7569346760304786e-06,
+      "loss": 0.5423,
+      "step": 13509
+    },
+    {
+      "epoch": 3.5864861277047657,
+      "grad_norm": 0.4536943308848015,
+      "learning_rate": 1.7566013334574146e-06,
+      "loss": 0.5576,
+      "step": 13510
+    },
+    {
+      "epoch": 3.5867516261781494,
+      "grad_norm": 0.4584524395324461,
+      "learning_rate": 1.7562680053824448e-06,
+      "loss": 0.5554,
+      "step": 13511
+    },
+    {
+      "epoch": 3.587017124651533,
+      "grad_norm": 0.45760156151223275,
+      "learning_rate": 1.7559346918120708e-06,
+      "loss": 0.5545,
+      "step": 13512
+    },
+    {
+      "epoch": 3.587282623124917,
+      "grad_norm": 0.45162586451809417,
+      "learning_rate": 1.755601392752791e-06,
+      "loss": 0.5416,
+      "step": 13513
+    },
+    {
+      "epoch": 3.5875481215983007,
+      "grad_norm": 0.47282761243564664,
+      "learning_rate": 1.7552681082111067e-06,
+      "loss": 0.559,
+      "step": 13514
+    },
+    {
+      "epoch": 3.5878136200716844,
+      "grad_norm": 0.46998835058176913,
+      "learning_rate": 1.754934838193518e-06,
+      "loss": 0.5829,
+      "step": 13515
+    },
+    {
+      "epoch": 3.5880791185450684,
+      "grad_norm": 0.47521146468018266,
+      "learning_rate": 1.754601582706524e-06,
+      "loss": 0.5337,
+      "step": 13516
+    },
+    {
+      "epoch": 3.588344617018452,
+      "grad_norm": 0.45311333421120553,
+      "learning_rate": 1.7542683417566242e-06,
+      "loss": 0.5303,
+      "step": 13517
+    },
+    {
+      "epoch": 3.5886101154918357,
+      "grad_norm": 0.45546536908517565,
+      "learning_rate": 1.7539351153503187e-06,
+      "loss": 0.5385,
+      "step": 13518
+    },
+    {
+      "epoch": 3.58887561396522,
+      "grad_norm": 0.4690156912406919,
+      "learning_rate": 1.7536019034941042e-06,
+      "loss": 0.5606,
+      "step": 13519
+    },
+    {
+      "epoch": 3.5891411124386035,
+      "grad_norm": 0.45947773561372895,
+      "learning_rate": 1.7532687061944798e-06,
+      "loss": 0.5275,
+      "step": 13520
+    },
+    {
+      "epoch": 3.589406610911987,
+      "grad_norm": 0.46085606263172146,
+      "learning_rate": 1.7529355234579443e-06,
+      "loss": 0.5482,
+      "step": 13521
+    },
+    {
+      "epoch": 3.589672109385371,
+      "grad_norm": 0.461940512163368,
+      "learning_rate": 1.752602355290995e-06,
+      "loss": 0.5789,
+      "step": 13522
+    },
+    {
+      "epoch": 3.589937607858755,
+      "grad_norm": 0.4730792736893672,
+      "learning_rate": 1.7522692017001296e-06,
+      "loss": 0.5216,
+      "step": 13523
+    },
+    {
+      "epoch": 3.5902031063321385,
+      "grad_norm": 0.4844521368790733,
+      "learning_rate": 1.7519360626918464e-06,
+      "loss": 0.5008,
+      "step": 13524
+    },
+    {
+      "epoch": 3.5904686048055225,
+      "grad_norm": 0.45045897881455704,
+      "learning_rate": 1.7516029382726408e-06,
+      "loss": 0.542,
+      "step": 13525
+    },
+    {
+      "epoch": 3.590734103278906,
+      "grad_norm": 0.4676984019742277,
+      "learning_rate": 1.7512698284490101e-06,
+      "loss": 0.5667,
+      "step": 13526
+    },
+    {
+      "epoch": 3.59099960175229,
+      "grad_norm": 0.4616785801193104,
+      "learning_rate": 1.7509367332274512e-06,
+      "loss": 0.5572,
+      "step": 13527
+    },
+    {
+      "epoch": 3.591265100225674,
+      "grad_norm": 0.47411424307148375,
+      "learning_rate": 1.7506036526144599e-06,
+      "loss": 0.5511,
+      "step": 13528
+    },
+    {
+      "epoch": 3.5915305986990576,
+      "grad_norm": 0.5009161702096905,
+      "learning_rate": 1.7502705866165326e-06,
+      "loss": 0.5535,
+      "step": 13529
+    },
+    {
+      "epoch": 3.591796097172441,
+      "grad_norm": 0.48698262732125325,
+      "learning_rate": 1.7499375352401646e-06,
+      "loss": 0.5026,
+      "step": 13530
+    },
+    {
+      "epoch": 3.5920615956458253,
+      "grad_norm": 0.4773655651478285,
+      "learning_rate": 1.7496044984918508e-06,
+      "loss": 0.5447,
+      "step": 13531
+    },
+    {
+      "epoch": 3.592327094119209,
+      "grad_norm": 0.44994684746788105,
+      "learning_rate": 1.7492714763780865e-06,
+      "loss": 0.5118,
+      "step": 13532
+    },
+    {
+      "epoch": 3.5925925925925926,
+      "grad_norm": 0.47142237175539886,
+      "learning_rate": 1.7489384689053663e-06,
+      "loss": 0.5763,
+      "step": 13533
+    },
+    {
+      "epoch": 3.5928580910659766,
+      "grad_norm": 0.49761640814276,
+      "learning_rate": 1.7486054760801853e-06,
+      "loss": 0.5431,
+      "step": 13534
+    },
+    {
+      "epoch": 3.5931235895393603,
+      "grad_norm": 0.523945669984737,
+      "learning_rate": 1.7482724979090382e-06,
+      "loss": 0.5603,
+      "step": 13535
+    },
+    {
+      "epoch": 3.593389088012744,
+      "grad_norm": 0.48456622887973827,
+      "learning_rate": 1.7479395343984164e-06,
+      "loss": 0.5684,
+      "step": 13536
+    },
+    {
+      "epoch": 3.5936545864861276,
+      "grad_norm": 0.4783161901373854,
+      "learning_rate": 1.7476065855548158e-06,
+      "loss": 0.5755,
+      "step": 13537
+    },
+    {
+      "epoch": 3.593920084959511,
+      "grad_norm": 0.46748155178915257,
+      "learning_rate": 1.7472736513847283e-06,
+      "loss": 0.4945,
+      "step": 13538
+    },
+    {
+      "epoch": 3.5941855834328953,
+      "grad_norm": 0.44811314369627037,
+      "learning_rate": 1.746940731894648e-06,
+      "loss": 0.5287,
+      "step": 13539
+    },
+    {
+      "epoch": 3.594451081906279,
+      "grad_norm": 0.4682869275336895,
+      "learning_rate": 1.746607827091067e-06,
+      "loss": 0.5801,
+      "step": 13540
+    },
+    {
+      "epoch": 3.5947165803796626,
+      "grad_norm": 0.45917664176170214,
+      "learning_rate": 1.7462749369804788e-06,
+      "loss": 0.5722,
+      "step": 13541
+    },
+    {
+      "epoch": 3.5949820788530467,
+      "grad_norm": 0.4728410025339026,
+      "learning_rate": 1.745942061569374e-06,
+      "loss": 0.5576,
+      "step": 13542
+    },
+    {
+      "epoch": 3.5952475773264303,
+      "grad_norm": 0.47159865805610757,
+      "learning_rate": 1.745609200864245e-06,
+      "loss": 0.569,
+      "step": 13543
+    },
+    {
+      "epoch": 3.595513075799814,
+      "grad_norm": 0.48419035298459817,
+      "learning_rate": 1.745276354871584e-06,
+      "loss": 0.5319,
+      "step": 13544
+    },
+    {
+      "epoch": 3.595778574273198,
+      "grad_norm": 0.46495236107469884,
+      "learning_rate": 1.744943523597882e-06,
+      "loss": 0.5301,
+      "step": 13545
+    },
+    {
+      "epoch": 3.5960440727465817,
+      "grad_norm": 0.46122650207334315,
+      "learning_rate": 1.74461070704963e-06,
+      "loss": 0.5544,
+      "step": 13546
+    },
+    {
+      "epoch": 3.5963095712199653,
+      "grad_norm": 0.4499103909485904,
+      "learning_rate": 1.7442779052333196e-06,
+      "loss": 0.5775,
+      "step": 13547
+    },
+    {
+      "epoch": 3.5965750696933494,
+      "grad_norm": 0.467187799874779,
+      "learning_rate": 1.7439451181554396e-06,
+      "loss": 0.5074,
+      "step": 13548
+    },
+    {
+      "epoch": 3.596840568166733,
+      "grad_norm": 0.47675999681071635,
+      "learning_rate": 1.7436123458224801e-06,
+      "loss": 0.559,
+      "step": 13549
+    },
+    {
+      "epoch": 3.5971060666401167,
+      "grad_norm": 0.4640558966115894,
+      "learning_rate": 1.7432795882409326e-06,
+      "loss": 0.5238,
+      "step": 13550
+    },
+    {
+      "epoch": 3.5973715651135008,
+      "grad_norm": 0.46998406412200033,
+      "learning_rate": 1.7429468454172858e-06,
+      "loss": 0.5296,
+      "step": 13551
+    },
+    {
+      "epoch": 3.5976370635868844,
+      "grad_norm": 0.45108524878875705,
+      "learning_rate": 1.742614117358029e-06,
+      "loss": 0.5474,
+      "step": 13552
+    },
+    {
+      "epoch": 3.597902562060268,
+      "grad_norm": 0.4698085632406593,
+      "learning_rate": 1.7422814040696523e-06,
+      "loss": 0.5632,
+      "step": 13553
+    },
+    {
+      "epoch": 3.598168060533652,
+      "grad_norm": 0.47612071347764656,
+      "learning_rate": 1.7419487055586425e-06,
+      "loss": 0.5588,
+      "step": 13554
+    },
+    {
+      "epoch": 3.5984335590070358,
+      "grad_norm": 0.47702669160319555,
+      "learning_rate": 1.741616021831489e-06,
+      "loss": 0.5638,
+      "step": 13555
+    },
+    {
+      "epoch": 3.5986990574804194,
+      "grad_norm": 0.4699379446796062,
+      "learning_rate": 1.74128335289468e-06,
+      "loss": 0.562,
+      "step": 13556
+    },
+    {
+      "epoch": 3.5989645559538035,
+      "grad_norm": 0.4602961472409376,
+      "learning_rate": 1.7409506987547032e-06,
+      "loss": 0.5674,
+      "step": 13557
+    },
+    {
+      "epoch": 3.599230054427187,
+      "grad_norm": 0.46559632964917874,
+      "learning_rate": 1.7406180594180466e-06,
+      "loss": 0.543,
+      "step": 13558
+    },
+    {
+      "epoch": 3.5994955529005708,
+      "grad_norm": 0.46674183003026837,
+      "learning_rate": 1.7402854348911968e-06,
+      "loss": 0.5176,
+      "step": 13559
+    },
+    {
+      "epoch": 3.599761051373955,
+      "grad_norm": 0.4522972304929055,
+      "learning_rate": 1.7399528251806411e-06,
+      "loss": 0.5379,
+      "step": 13560
+    },
+    {
+      "epoch": 3.6000265498473385,
+      "grad_norm": 0.47912756529151496,
+      "learning_rate": 1.739620230292866e-06,
+      "loss": 0.5332,
+      "step": 13561
+    },
+    {
+      "epoch": 3.600292048320722,
+      "grad_norm": 0.4716449912045407,
+      "learning_rate": 1.7392876502343587e-06,
+      "loss": 0.5616,
+      "step": 13562
+    },
+    {
+      "epoch": 3.600557546794106,
+      "grad_norm": 0.4594146892071589,
+      "learning_rate": 1.7389550850116044e-06,
+      "loss": 0.5463,
+      "step": 13563
+    },
+    {
+      "epoch": 3.60082304526749,
+      "grad_norm": 0.470518691819596,
+      "learning_rate": 1.7386225346310903e-06,
+      "loss": 0.5214,
+      "step": 13564
+    },
+    {
+      "epoch": 3.6010885437408735,
+      "grad_norm": 0.4709061899337295,
+      "learning_rate": 1.7382899990992996e-06,
+      "loss": 0.5699,
+      "step": 13565
+    },
+    {
+      "epoch": 3.601354042214257,
+      "grad_norm": 0.46598305534954343,
+      "learning_rate": 1.7379574784227192e-06,
+      "loss": 0.5676,
+      "step": 13566
+    },
+    {
+      "epoch": 3.601619540687641,
+      "grad_norm": 0.46249379887871217,
+      "learning_rate": 1.7376249726078343e-06,
+      "loss": 0.547,
+      "step": 13567
+    },
+    {
+      "epoch": 3.601885039161025,
+      "grad_norm": 0.48189805557502413,
+      "learning_rate": 1.7372924816611283e-06,
+      "loss": 0.5471,
+      "step": 13568
+    },
+    {
+      "epoch": 3.6021505376344085,
+      "grad_norm": 0.4681194709696662,
+      "learning_rate": 1.7369600055890861e-06,
+      "loss": 0.5513,
+      "step": 13569
+    },
+    {
+      "epoch": 3.602416036107792,
+      "grad_norm": 0.4683496453924496,
+      "learning_rate": 1.7366275443981936e-06,
+      "loss": 0.5272,
+      "step": 13570
+    },
+    {
+      "epoch": 3.6026815345811762,
+      "grad_norm": 0.4682665521946284,
+      "learning_rate": 1.7362950980949323e-06,
+      "loss": 0.5342,
+      "step": 13571
+    },
+    {
+      "epoch": 3.60294703305456,
+      "grad_norm": 0.45808760624460554,
+      "learning_rate": 1.735962666685786e-06,
+      "loss": 0.5522,
+      "step": 13572
+    },
+    {
+      "epoch": 3.6032125315279435,
+      "grad_norm": 0.4647097304585912,
+      "learning_rate": 1.7356302501772385e-06,
+      "loss": 0.5391,
+      "step": 13573
+    },
+    {
+      "epoch": 3.6034780300013276,
+      "grad_norm": 0.47034793355342025,
+      "learning_rate": 1.7352978485757727e-06,
+      "loss": 0.5546,
+      "step": 13574
+    },
+    {
+      "epoch": 3.6037435284747112,
+      "grad_norm": 0.4448627507602738,
+      "learning_rate": 1.7349654618878708e-06,
+      "loss": 0.5413,
+      "step": 13575
+    },
+    {
+      "epoch": 3.604009026948095,
+      "grad_norm": 0.46162949965669164,
+      "learning_rate": 1.7346330901200164e-06,
+      "loss": 0.5372,
+      "step": 13576
+    },
+    {
+      "epoch": 3.604274525421479,
+      "grad_norm": 0.4518532704969727,
+      "learning_rate": 1.7343007332786904e-06,
+      "loss": 0.5031,
+      "step": 13577
+    },
+    {
+      "epoch": 3.6045400238948626,
+      "grad_norm": 0.4700331515662211,
+      "learning_rate": 1.7339683913703745e-06,
+      "loss": 0.5317,
+      "step": 13578
+    },
+    {
+      "epoch": 3.6048055223682463,
+      "grad_norm": 0.4678001322655851,
+      "learning_rate": 1.7336360644015508e-06,
+      "loss": 0.5826,
+      "step": 13579
+    },
+    {
+      "epoch": 3.6050710208416303,
+      "grad_norm": 0.4575028624396153,
+      "learning_rate": 1.7333037523787e-06,
+      "loss": 0.5311,
+      "step": 13580
+    },
+    {
+      "epoch": 3.605336519315014,
+      "grad_norm": 0.46315715560371845,
+      "learning_rate": 1.7329714553083044e-06,
+      "loss": 0.5426,
+      "step": 13581
+    },
+    {
+      "epoch": 3.6056020177883976,
+      "grad_norm": 0.47113471415086106,
+      "learning_rate": 1.7326391731968423e-06,
+      "loss": 0.5136,
+      "step": 13582
+    },
+    {
+      "epoch": 3.6058675162617817,
+      "grad_norm": 0.4567439978450365,
+      "learning_rate": 1.7323069060507953e-06,
+      "loss": 0.5151,
+      "step": 13583
+    },
+    {
+      "epoch": 3.6061330147351653,
+      "grad_norm": 0.4491878621633084,
+      "learning_rate": 1.731974653876643e-06,
+      "loss": 0.5308,
+      "step": 13584
+    },
+    {
+      "epoch": 3.606398513208549,
+      "grad_norm": 0.459542964572663,
+      "learning_rate": 1.7316424166808654e-06,
+      "loss": 0.5644,
+      "step": 13585
+    },
+    {
+      "epoch": 3.606664011681933,
+      "grad_norm": 0.48029558159350694,
+      "learning_rate": 1.7313101944699421e-06,
+      "loss": 0.5318,
+      "step": 13586
+    },
+    {
+      "epoch": 3.6069295101553167,
+      "grad_norm": 0.458615514035395,
+      "learning_rate": 1.7309779872503523e-06,
+      "loss": 0.5713,
+      "step": 13587
+    },
+    {
+      "epoch": 3.6071950086287004,
+      "grad_norm": 0.4664399596060654,
+      "learning_rate": 1.7306457950285747e-06,
+      "loss": 0.5036,
+      "step": 13588
+    },
+    {
+      "epoch": 3.6074605071020844,
+      "grad_norm": 0.4692681958041445,
+      "learning_rate": 1.7303136178110873e-06,
+      "loss": 0.5017,
+      "step": 13589
+    },
+    {
+      "epoch": 3.607726005575468,
+      "grad_norm": 0.46142858668685666,
+      "learning_rate": 1.7299814556043694e-06,
+      "loss": 0.5442,
+      "step": 13590
+    },
+    {
+      "epoch": 3.6079915040488517,
+      "grad_norm": 0.45118892028036384,
+      "learning_rate": 1.729649308414898e-06,
+      "loss": 0.5671,
+      "step": 13591
+    },
+    {
+      "epoch": 3.6082570025222354,
+      "grad_norm": 0.4598542880949478,
+      "learning_rate": 1.729317176249151e-06,
+      "loss": 0.525,
+      "step": 13592
+    },
+    {
+      "epoch": 3.608522500995619,
+      "grad_norm": 0.46819627471808545,
+      "learning_rate": 1.7289850591136077e-06,
+      "loss": 0.5709,
+      "step": 13593
+    },
+    {
+      "epoch": 3.608787999469003,
+      "grad_norm": 0.45636397953153746,
+      "learning_rate": 1.7286529570147415e-06,
+      "loss": 0.5717,
+      "step": 13594
+    },
+    {
+      "epoch": 3.6090534979423867,
+      "grad_norm": 0.47415080865320663,
+      "learning_rate": 1.7283208699590321e-06,
+      "loss": 0.5535,
+      "step": 13595
+    },
+    {
+      "epoch": 3.6093189964157704,
+      "grad_norm": 0.44461879784687897,
+      "learning_rate": 1.7279887979529548e-06,
+      "loss": 0.5514,
+      "step": 13596
+    },
+    {
+      "epoch": 3.6095844948891544,
+      "grad_norm": 0.46604672820334747,
+      "learning_rate": 1.7276567410029863e-06,
+      "loss": 0.5693,
+      "step": 13597
+    },
+    {
+      "epoch": 3.609849993362538,
+      "grad_norm": 0.46406717274669723,
+      "learning_rate": 1.7273246991156023e-06,
+      "loss": 0.5396,
+      "step": 13598
+    },
+    {
+      "epoch": 3.6101154918359217,
+      "grad_norm": 0.4663992982029978,
+      "learning_rate": 1.7269926722972796e-06,
+      "loss": 0.5132,
+      "step": 13599
+    },
+    {
+      "epoch": 3.610380990309306,
+      "grad_norm": 0.4774319859445451,
+      "learning_rate": 1.7266606605544914e-06,
+      "loss": 0.5384,
+      "step": 13600
+    },
+    {
+      "epoch": 3.6106464887826895,
+      "grad_norm": 0.46548168841747417,
+      "learning_rate": 1.7263286638937138e-06,
+      "loss": 0.5458,
+      "step": 13601
+    },
+    {
+      "epoch": 3.610911987256073,
+      "grad_norm": 0.4484765598962136,
+      "learning_rate": 1.7259966823214214e-06,
+      "loss": 0.5101,
+      "step": 13602
+    },
+    {
+      "epoch": 3.611177485729457,
+      "grad_norm": 0.47103830221202686,
+      "learning_rate": 1.7256647158440887e-06,
+      "loss": 0.5355,
+      "step": 13603
+    },
+    {
+      "epoch": 3.611442984202841,
+      "grad_norm": 0.458278323284456,
+      "learning_rate": 1.7253327644681908e-06,
+      "loss": 0.5425,
+      "step": 13604
+    },
+    {
+      "epoch": 3.6117084826762245,
+      "grad_norm": 0.4612397541487162,
+      "learning_rate": 1.7250008282002e-06,
+      "loss": 0.5212,
+      "step": 13605
+    },
+    {
+      "epoch": 3.6119739811496085,
+      "grad_norm": 0.4688759358010234,
+      "learning_rate": 1.7246689070465906e-06,
+      "loss": 0.5613,
+      "step": 13606
+    },
+    {
+      "epoch": 3.612239479622992,
+      "grad_norm": 0.4574925144570196,
+      "learning_rate": 1.724337001013836e-06,
+      "loss": 0.5323,
+      "step": 13607
+    },
+    {
+      "epoch": 3.612504978096376,
+      "grad_norm": 0.45692543219069714,
+      "learning_rate": 1.7240051101084088e-06,
+      "loss": 0.5521,
+      "step": 13608
+    },
+    {
+      "epoch": 3.61277047656976,
+      "grad_norm": 0.4570651800360103,
+      "learning_rate": 1.7236732343367818e-06,
+      "loss": 0.5974,
+      "step": 13609
+    },
+    {
+      "epoch": 3.6130359750431436,
+      "grad_norm": 0.4714511045563157,
+      "learning_rate": 1.7233413737054289e-06,
+      "loss": 0.5411,
+      "step": 13610
+    },
+    {
+      "epoch": 3.613301473516527,
+      "grad_norm": 0.46681311007707244,
+      "learning_rate": 1.7230095282208197e-06,
+      "loss": 0.5138,
+      "step": 13611
+    },
+    {
+      "epoch": 3.6135669719899113,
+      "grad_norm": 0.4374633771891128,
+      "learning_rate": 1.7226776978894269e-06,
+      "loss": 0.5124,
+      "step": 13612
+    },
+    {
+      "epoch": 3.613832470463295,
+      "grad_norm": 0.46014361574268586,
+      "learning_rate": 1.7223458827177226e-06,
+      "loss": 0.5388,
+      "step": 13613
+    },
+    {
+      "epoch": 3.6140979689366786,
+      "grad_norm": 0.4526472867290781,
+      "learning_rate": 1.722014082712177e-06,
+      "loss": 0.5508,
+      "step": 13614
+    },
+    {
+      "epoch": 3.6143634674100626,
+      "grad_norm": 0.45206992035680404,
+      "learning_rate": 1.7216822978792613e-06,
+      "loss": 0.5165,
+      "step": 13615
+    },
+    {
+      "epoch": 3.6146289658834463,
+      "grad_norm": 0.4801129442343729,
+      "learning_rate": 1.7213505282254482e-06,
+      "loss": 0.5697,
+      "step": 13616
+    },
+    {
+      "epoch": 3.61489446435683,
+      "grad_norm": 0.48423488502181944,
+      "learning_rate": 1.7210187737572055e-06,
+      "loss": 0.549,
+      "step": 13617
+    },
+    {
+      "epoch": 3.6151599628302136,
+      "grad_norm": 0.4598635727984862,
+      "learning_rate": 1.7206870344810034e-06,
+      "loss": 0.5412,
+      "step": 13618
+    },
+    {
+      "epoch": 3.6154254613035977,
+      "grad_norm": 0.45683239145922466,
+      "learning_rate": 1.7203553104033124e-06,
+      "loss": 0.5426,
+      "step": 13619
+    },
+    {
+      "epoch": 3.6156909597769813,
+      "grad_norm": 0.4714143392031203,
+      "learning_rate": 1.7200236015306018e-06,
+      "loss": 0.564,
+      "step": 13620
+    },
+    {
+      "epoch": 3.615956458250365,
+      "grad_norm": 0.45942074409938555,
+      "learning_rate": 1.7196919078693408e-06,
+      "loss": 0.548,
+      "step": 13621
+    },
+    {
+      "epoch": 3.6162219567237486,
+      "grad_norm": 0.4672479852630923,
+      "learning_rate": 1.7193602294259983e-06,
+      "loss": 0.5596,
+      "step": 13622
+    },
+    {
+      "epoch": 3.6164874551971327,
+      "grad_norm": 0.4596295563065969,
+      "learning_rate": 1.7190285662070424e-06,
+      "loss": 0.5783,
+      "step": 13623
+    },
+    {
+      "epoch": 3.6167529536705163,
+      "grad_norm": 0.47054608348118204,
+      "learning_rate": 1.7186969182189417e-06,
+      "loss": 0.505,
+      "step": 13624
+    },
+    {
+      "epoch": 3.6170184521439,
+      "grad_norm": 0.4824321631269175,
+      "learning_rate": 1.7183652854681637e-06,
+      "loss": 0.568,
+      "step": 13625
+    },
+    {
+      "epoch": 3.617283950617284,
+      "grad_norm": 0.4606630924047428,
+      "learning_rate": 1.718033667961177e-06,
+      "loss": 0.5517,
+      "step": 13626
+    },
+    {
+      "epoch": 3.6175494490906677,
+      "grad_norm": 0.4719866749369406,
+      "learning_rate": 1.7177020657044488e-06,
+      "loss": 0.5631,
+      "step": 13627
+    },
+    {
+      "epoch": 3.6178149475640513,
+      "grad_norm": 0.46929153589945016,
+      "learning_rate": 1.7173704787044448e-06,
+      "loss": 0.5485,
+      "step": 13628
+    },
+    {
+      "epoch": 3.6180804460374354,
+      "grad_norm": 0.4541068694132094,
+      "learning_rate": 1.717038906967633e-06,
+      "loss": 0.5394,
+      "step": 13629
+    },
+    {
+      "epoch": 3.618345944510819,
+      "grad_norm": 0.45401441817806126,
+      "learning_rate": 1.7167073505004792e-06,
+      "loss": 0.5579,
+      "step": 13630
+    },
+    {
+      "epoch": 3.6186114429842027,
+      "grad_norm": 0.47340886187189046,
+      "learning_rate": 1.7163758093094502e-06,
+      "loss": 0.5495,
+      "step": 13631
+    },
+    {
+      "epoch": 3.6188769414575868,
+      "grad_norm": 0.4757218638647428,
+      "learning_rate": 1.716044283401011e-06,
+      "loss": 0.5393,
+      "step": 13632
+    },
+    {
+      "epoch": 3.6191424399309704,
+      "grad_norm": 0.4562527725970025,
+      "learning_rate": 1.7157127727816286e-06,
+      "loss": 0.5283,
+      "step": 13633
+    },
+    {
+      "epoch": 3.619407938404354,
+      "grad_norm": 0.458041298347764,
+      "learning_rate": 1.7153812774577672e-06,
+      "loss": 0.5403,
+      "step": 13634
+    },
+    {
+      "epoch": 3.619673436877738,
+      "grad_norm": 0.47282952261417815,
+      "learning_rate": 1.7150497974358914e-06,
+      "loss": 0.5216,
+      "step": 13635
+    },
+    {
+      "epoch": 3.6199389353511218,
+      "grad_norm": 0.46716491856811143,
+      "learning_rate": 1.714718332722467e-06,
+      "loss": 0.5533,
+      "step": 13636
+    },
+    {
+      "epoch": 3.6202044338245054,
+      "grad_norm": 0.467141076609319,
+      "learning_rate": 1.7143868833239575e-06,
+      "loss": 0.5371,
+      "step": 13637
+    },
+    {
+      "epoch": 3.6204699322978895,
+      "grad_norm": 0.4654792678552746,
+      "learning_rate": 1.714055449246827e-06,
+      "loss": 0.5118,
+      "step": 13638
+    },
+    {
+      "epoch": 3.620735430771273,
+      "grad_norm": 0.45384211363342897,
+      "learning_rate": 1.7137240304975411e-06,
+      "loss": 0.5409,
+      "step": 13639
+    },
+    {
+      "epoch": 3.6210009292446568,
+      "grad_norm": 0.4745162033476175,
+      "learning_rate": 1.71339262708256e-06,
+      "loss": 0.5743,
+      "step": 13640
+    },
+    {
+      "epoch": 3.621266427718041,
+      "grad_norm": 0.45193187242856186,
+      "learning_rate": 1.713061239008349e-06,
+      "loss": 0.5471,
+      "step": 13641
+    },
+    {
+      "epoch": 3.6215319261914245,
+      "grad_norm": 0.4705267753487941,
+      "learning_rate": 1.7127298662813706e-06,
+      "loss": 0.5616,
+      "step": 13642
+    },
+    {
+      "epoch": 3.621797424664808,
+      "grad_norm": 0.46118296825061666,
+      "learning_rate": 1.7123985089080875e-06,
+      "loss": 0.5462,
+      "step": 13643
+    },
+    {
+      "epoch": 3.6220629231381922,
+      "grad_norm": 0.4468382816640049,
+      "learning_rate": 1.712067166894962e-06,
+      "loss": 0.5182,
+      "step": 13644
+    },
+    {
+      "epoch": 3.622328421611576,
+      "grad_norm": 0.4644414446232178,
+      "learning_rate": 1.7117358402484568e-06,
+      "loss": 0.5514,
+      "step": 13645
+    },
+    {
+      "epoch": 3.6225939200849595,
+      "grad_norm": 0.4728811353813963,
+      "learning_rate": 1.7114045289750319e-06,
+      "loss": 0.5995,
+      "step": 13646
+    },
+    {
+      "epoch": 3.622859418558343,
+      "grad_norm": 0.4547843982903181,
+      "learning_rate": 1.7110732330811492e-06,
+      "loss": 0.5478,
+      "step": 13647
+    },
+    {
+      "epoch": 3.6231249170317272,
+      "grad_norm": 0.4695241271381432,
+      "learning_rate": 1.7107419525732701e-06,
+      "loss": 0.5576,
+      "step": 13648
+    },
+    {
+      "epoch": 3.623390415505111,
+      "grad_norm": 0.458804988875267,
+      "learning_rate": 1.7104106874578555e-06,
+      "loss": 0.5976,
+      "step": 13649
+    },
+    {
+      "epoch": 3.6236559139784945,
+      "grad_norm": 0.46468763330087354,
+      "learning_rate": 1.7100794377413658e-06,
+      "loss": 0.5451,
+      "step": 13650
+    },
+    {
+      "epoch": 3.623921412451878,
+      "grad_norm": 0.4791723989770146,
+      "learning_rate": 1.7097482034302616e-06,
+      "loss": 0.5474,
+      "step": 13651
+    },
+    {
+      "epoch": 3.6241869109252622,
+      "grad_norm": 0.4614585352340324,
+      "learning_rate": 1.7094169845310017e-06,
+      "loss": 0.5245,
+      "step": 13652
+    },
+    {
+      "epoch": 3.624452409398646,
+      "grad_norm": 0.45746898951882653,
+      "learning_rate": 1.7090857810500462e-06,
+      "loss": 0.5282,
+      "step": 13653
+    },
+    {
+      "epoch": 3.6247179078720295,
+      "grad_norm": 0.4741656899525762,
+      "learning_rate": 1.7087545929938545e-06,
+      "loss": 0.5738,
+      "step": 13654
+    },
+    {
+      "epoch": 3.6249834063454136,
+      "grad_norm": 0.46060643980717986,
+      "learning_rate": 1.7084234203688855e-06,
+      "loss": 0.5407,
+      "step": 13655
+    },
+    {
+      "epoch": 3.6252489048187972,
+      "grad_norm": 0.4647203963659346,
+      "learning_rate": 1.7080922631815994e-06,
+      "loss": 0.4934,
+      "step": 13656
+    },
+    {
+      "epoch": 3.625514403292181,
+      "grad_norm": 0.45171062586580485,
+      "learning_rate": 1.7077611214384514e-06,
+      "loss": 0.5155,
+      "step": 13657
+    },
+    {
+      "epoch": 3.625779901765565,
+      "grad_norm": 0.45525417035870663,
+      "learning_rate": 1.7074299951459017e-06,
+      "loss": 0.5234,
+      "step": 13658
+    },
+    {
+      "epoch": 3.6260454002389486,
+      "grad_norm": 0.4639623621962392,
+      "learning_rate": 1.7070988843104074e-06,
+      "loss": 0.5318,
+      "step": 13659
+    },
+    {
+      "epoch": 3.6263108987123323,
+      "grad_norm": 0.46311921372434306,
+      "learning_rate": 1.7067677889384265e-06,
+      "loss": 0.5005,
+      "step": 13660
+    },
+    {
+      "epoch": 3.6265763971857163,
+      "grad_norm": 0.45319119470606894,
+      "learning_rate": 1.7064367090364152e-06,
+      "loss": 0.5307,
+      "step": 13661
+    },
+    {
+      "epoch": 3.6268418956591,
+      "grad_norm": 0.46147456132470915,
+      "learning_rate": 1.7061056446108325e-06,
+      "loss": 0.5396,
+      "step": 13662
+    },
+    {
+      "epoch": 3.6271073941324836,
+      "grad_norm": 0.46380341042187784,
+      "learning_rate": 1.7057745956681327e-06,
+      "loss": 0.55,
+      "step": 13663
+    },
+    {
+      "epoch": 3.6273728926058677,
+      "grad_norm": 0.46266186746131904,
+      "learning_rate": 1.7054435622147727e-06,
+      "loss": 0.5642,
+      "step": 13664
+    },
+    {
+      "epoch": 3.6276383910792513,
+      "grad_norm": 0.45589186004984195,
+      "learning_rate": 1.7051125442572087e-06,
+      "loss": 0.5301,
+      "step": 13665
+    },
+    {
+      "epoch": 3.627903889552635,
+      "grad_norm": 0.453659468490755,
+      "learning_rate": 1.7047815418018965e-06,
+      "loss": 0.543,
+      "step": 13666
+    },
+    {
+      "epoch": 3.628169388026019,
+      "grad_norm": 0.4709650239203018,
+      "learning_rate": 1.7044505548552908e-06,
+      "loss": 0.5346,
+      "step": 13667
+    },
+    {
+      "epoch": 3.6284348864994027,
+      "grad_norm": 0.45100151231476326,
+      "learning_rate": 1.704119583423848e-06,
+      "loss": 0.5309,
+      "step": 13668
+    },
+    {
+      "epoch": 3.6287003849727864,
+      "grad_norm": 0.45782276589120025,
+      "learning_rate": 1.7037886275140213e-06,
+      "loss": 0.5082,
+      "step": 13669
+    },
+    {
+      "epoch": 3.6289658834461704,
+      "grad_norm": 0.4617019052182944,
+      "learning_rate": 1.7034576871322661e-06,
+      "loss": 0.5723,
+      "step": 13670
+    },
+    {
+      "epoch": 3.629231381919554,
+      "grad_norm": 0.4705898177021215,
+      "learning_rate": 1.703126762285036e-06,
+      "loss": 0.5408,
+      "step": 13671
+    },
+    {
+      "epoch": 3.6294968803929377,
+      "grad_norm": 0.46026739951165246,
+      "learning_rate": 1.7027958529787853e-06,
+      "loss": 0.5753,
+      "step": 13672
+    },
+    {
+      "epoch": 3.6297623788663214,
+      "grad_norm": 0.4774369464407052,
+      "learning_rate": 1.702464959219967e-06,
+      "loss": 0.5792,
+      "step": 13673
+    },
+    {
+      "epoch": 3.6300278773397054,
+      "grad_norm": 0.4770219619640301,
+      "learning_rate": 1.7021340810150361e-06,
+      "loss": 0.5241,
+      "step": 13674
+    },
+    {
+      "epoch": 3.630293375813089,
+      "grad_norm": 0.4650085159622131,
+      "learning_rate": 1.701803218370443e-06,
+      "loss": 0.512,
+      "step": 13675
+    },
+    {
+      "epoch": 3.6305588742864727,
+      "grad_norm": 0.4618618539422062,
+      "learning_rate": 1.7014723712926414e-06,
+      "loss": 0.5709,
+      "step": 13676
+    },
+    {
+      "epoch": 3.6308243727598564,
+      "grad_norm": 0.4608173417979016,
+      "learning_rate": 1.7011415397880838e-06,
+      "loss": 0.549,
+      "step": 13677
+    },
+    {
+      "epoch": 3.6310898712332405,
+      "grad_norm": 0.48138934099222763,
+      "learning_rate": 1.7008107238632222e-06,
+      "loss": 0.5653,
+      "step": 13678
+    },
+    {
+      "epoch": 3.631355369706624,
+      "grad_norm": 0.47405279024724833,
+      "learning_rate": 1.7004799235245084e-06,
+      "loss": 0.5298,
+      "step": 13679
+    },
+    {
+      "epoch": 3.6316208681800077,
+      "grad_norm": 0.475434812157762,
+      "learning_rate": 1.7001491387783936e-06,
+      "loss": 0.5243,
+      "step": 13680
+    },
+    {
+      "epoch": 3.631886366653392,
+      "grad_norm": 0.4555221980389128,
+      "learning_rate": 1.6998183696313286e-06,
+      "loss": 0.5524,
+      "step": 13681
+    },
+    {
+      "epoch": 3.6321518651267755,
+      "grad_norm": 0.46018521990203676,
+      "learning_rate": 1.6994876160897647e-06,
+      "loss": 0.543,
+      "step": 13682
+    },
+    {
+      "epoch": 3.632417363600159,
+      "grad_norm": 0.48111701125486367,
+      "learning_rate": 1.6991568781601522e-06,
+      "loss": 0.5634,
+      "step": 13683
+    },
+    {
+      "epoch": 3.632682862073543,
+      "grad_norm": 0.4683753571383182,
+      "learning_rate": 1.6988261558489416e-06,
+      "loss": 0.5445,
+      "step": 13684
+    },
+    {
+      "epoch": 3.632948360546927,
+      "grad_norm": 0.47775647821815037,
+      "learning_rate": 1.6984954491625833e-06,
+      "loss": 0.5367,
+      "step": 13685
+    },
+    {
+      "epoch": 3.6332138590203105,
+      "grad_norm": 0.4790413181753658,
+      "learning_rate": 1.698164758107525e-06,
+      "loss": 0.5289,
+      "step": 13686
+    },
+    {
+      "epoch": 3.6334793574936946,
+      "grad_norm": 0.46344828824385936,
+      "learning_rate": 1.6978340826902173e-06,
+      "loss": 0.5064,
+      "step": 13687
+    },
+    {
+      "epoch": 3.633744855967078,
+      "grad_norm": 0.46296307530320996,
+      "learning_rate": 1.6975034229171092e-06,
+      "loss": 0.5263,
+      "step": 13688
+    },
+    {
+      "epoch": 3.634010354440462,
+      "grad_norm": 0.4703805306281193,
+      "learning_rate": 1.697172778794649e-06,
+      "loss": 0.5636,
+      "step": 13689
+    },
+    {
+      "epoch": 3.634275852913846,
+      "grad_norm": 0.48484719367374407,
+      "learning_rate": 1.6968421503292856e-06,
+      "loss": 0.5531,
+      "step": 13690
+    },
+    {
+      "epoch": 3.6345413513872296,
+      "grad_norm": 0.47379127579867886,
+      "learning_rate": 1.6965115375274678e-06,
+      "loss": 0.5528,
+      "step": 13691
+    },
+    {
+      "epoch": 3.634806849860613,
+      "grad_norm": 0.4544197893375784,
+      "learning_rate": 1.6961809403956409e-06,
+      "loss": 0.5349,
+      "step": 13692
+    },
+    {
+      "epoch": 3.6350723483339973,
+      "grad_norm": 0.4713125982530026,
+      "learning_rate": 1.6958503589402539e-06,
+      "loss": 0.5843,
+      "step": 13693
+    },
+    {
+      "epoch": 3.635337846807381,
+      "grad_norm": 0.4667524768297605,
+      "learning_rate": 1.6955197931677538e-06,
+      "loss": 0.5291,
+      "step": 13694
+    },
+    {
+      "epoch": 3.6356033452807646,
+      "grad_norm": 0.4512196666018313,
+      "learning_rate": 1.6951892430845874e-06,
+      "loss": 0.5313,
+      "step": 13695
+    },
+    {
+      "epoch": 3.6358688437541486,
+      "grad_norm": 0.4693406989880126,
+      "learning_rate": 1.6948587086972013e-06,
+      "loss": 0.5597,
+      "step": 13696
+    },
+    {
+      "epoch": 3.6361343422275323,
+      "grad_norm": 0.46678191467790486,
+      "learning_rate": 1.6945281900120425e-06,
+      "loss": 0.549,
+      "step": 13697
+    },
+    {
+      "epoch": 3.636399840700916,
+      "grad_norm": 0.4645748361508512,
+      "learning_rate": 1.6941976870355553e-06,
+      "loss": 0.5307,
+      "step": 13698
+    },
+    {
+      "epoch": 3.6366653391743,
+      "grad_norm": 0.45548918114258136,
+      "learning_rate": 1.6938671997741863e-06,
+      "loss": 0.5291,
+      "step": 13699
+    },
+    {
+      "epoch": 3.6369308376476837,
+      "grad_norm": 0.4608563291416018,
+      "learning_rate": 1.6935367282343806e-06,
+      "loss": 0.5471,
+      "step": 13700
+    },
+    {
+      "epoch": 3.6371963361210673,
+      "grad_norm": 0.46696330066728287,
+      "learning_rate": 1.693206272422583e-06,
+      "loss": 0.5347,
+      "step": 13701
+    },
+    {
+      "epoch": 3.637461834594451,
+      "grad_norm": 0.46854927397595214,
+      "learning_rate": 1.69287583234524e-06,
+      "loss": 0.5366,
+      "step": 13702
+    },
+    {
+      "epoch": 3.637727333067835,
+      "grad_norm": 0.4610060401679918,
+      "learning_rate": 1.6925454080087931e-06,
+      "loss": 0.5241,
+      "step": 13703
+    },
+    {
+      "epoch": 3.6379928315412187,
+      "grad_norm": 0.4816351182302213,
+      "learning_rate": 1.6922149994196878e-06,
+      "loss": 0.5504,
+      "step": 13704
+    },
+    {
+      "epoch": 3.6382583300146023,
+      "grad_norm": 0.4629181425049328,
+      "learning_rate": 1.6918846065843677e-06,
+      "loss": 0.5331,
+      "step": 13705
+    },
+    {
+      "epoch": 3.638523828487986,
+      "grad_norm": 0.45681701236609173,
+      "learning_rate": 1.691554229509276e-06,
+      "loss": 0.5288,
+      "step": 13706
+    },
+    {
+      "epoch": 3.63878932696137,
+      "grad_norm": 0.4687411694448322,
+      "learning_rate": 1.6912238682008567e-06,
+      "loss": 0.5009,
+      "step": 13707
+    },
+    {
+      "epoch": 3.6390548254347537,
+      "grad_norm": 0.460669153949547,
+      "learning_rate": 1.6908935226655532e-06,
+      "loss": 0.5276,
+      "step": 13708
+    },
+    {
+      "epoch": 3.6393203239081373,
+      "grad_norm": 0.45111500154666356,
+      "learning_rate": 1.690563192909806e-06,
+      "loss": 0.5322,
+      "step": 13709
+    },
+    {
+      "epoch": 3.6395858223815214,
+      "grad_norm": 0.46686859367134764,
+      "learning_rate": 1.6902328789400585e-06,
+      "loss": 0.5512,
+      "step": 13710
+    },
+    {
+      "epoch": 3.639851320854905,
+      "grad_norm": 0.46572273536612585,
+      "learning_rate": 1.6899025807627525e-06,
+      "loss": 0.554,
+      "step": 13711
+    },
+    {
+      "epoch": 3.6401168193282887,
+      "grad_norm": 0.45905155799806036,
+      "learning_rate": 1.6895722983843295e-06,
+      "loss": 0.5275,
+      "step": 13712
+    },
+    {
+      "epoch": 3.6403823178016728,
+      "grad_norm": 0.4593243862512204,
+      "learning_rate": 1.6892420318112312e-06,
+      "loss": 0.5599,
+      "step": 13713
+    },
+    {
+      "epoch": 3.6406478162750564,
+      "grad_norm": 0.4532097012078511,
+      "learning_rate": 1.6889117810498986e-06,
+      "loss": 0.5468,
+      "step": 13714
+    },
+    {
+      "epoch": 3.64091331474844,
+      "grad_norm": 0.46951339482759724,
+      "learning_rate": 1.6885815461067717e-06,
+      "loss": 0.5498,
+      "step": 13715
+    },
+    {
+      "epoch": 3.641178813221824,
+      "grad_norm": 0.46255136715925005,
+      "learning_rate": 1.6882513269882916e-06,
+      "loss": 0.502,
+      "step": 13716
+    },
+    {
+      "epoch": 3.6414443116952078,
+      "grad_norm": 0.45581375133165497,
+      "learning_rate": 1.687921123700898e-06,
+      "loss": 0.5559,
+      "step": 13717
+    },
+    {
+      "epoch": 3.6417098101685914,
+      "grad_norm": 0.4563335031048507,
+      "learning_rate": 1.6875909362510306e-06,
+      "loss": 0.546,
+      "step": 13718
+    },
+    {
+      "epoch": 3.6419753086419755,
+      "grad_norm": 0.4606115766645989,
+      "learning_rate": 1.6872607646451292e-06,
+      "loss": 0.5359,
+      "step": 13719
+    },
+    {
+      "epoch": 3.642240807115359,
+      "grad_norm": 0.4678123251846704,
+      "learning_rate": 1.6869306088896336e-06,
+      "loss": 0.5001,
+      "step": 13720
+    },
+    {
+      "epoch": 3.6425063055887428,
+      "grad_norm": 0.46661629944344796,
+      "learning_rate": 1.6866004689909812e-06,
+      "loss": 0.5617,
+      "step": 13721
+    },
+    {
+      "epoch": 3.642771804062127,
+      "grad_norm": 0.46385917180605735,
+      "learning_rate": 1.6862703449556112e-06,
+      "loss": 0.5664,
+      "step": 13722
+    },
+    {
+      "epoch": 3.6430373025355105,
+      "grad_norm": 0.473229430186421,
+      "learning_rate": 1.6859402367899616e-06,
+      "loss": 0.5335,
+      "step": 13723
+    },
+    {
+      "epoch": 3.643302801008894,
+      "grad_norm": 0.467403493787694,
+      "learning_rate": 1.6856101445004709e-06,
+      "loss": 0.562,
+      "step": 13724
+    },
+    {
+      "epoch": 3.6435682994822782,
+      "grad_norm": 0.5369963572586336,
+      "learning_rate": 1.6852800680935763e-06,
+      "loss": 0.5232,
+      "step": 13725
+    },
+    {
+      "epoch": 3.643833797955662,
+      "grad_norm": 0.4624955681115691,
+      "learning_rate": 1.684950007575715e-06,
+      "loss": 0.5451,
+      "step": 13726
+    },
+    {
+      "epoch": 3.6440992964290455,
+      "grad_norm": 0.4731734794609201,
+      "learning_rate": 1.684619962953324e-06,
+      "loss": 0.6037,
+      "step": 13727
+    },
+    {
+      "epoch": 3.644364794902429,
+      "grad_norm": 0.4777244040321825,
+      "learning_rate": 1.6842899342328403e-06,
+      "loss": 0.5751,
+      "step": 13728
+    },
+    {
+      "epoch": 3.6446302933758132,
+      "grad_norm": 0.47354660154344547,
+      "learning_rate": 1.6839599214206997e-06,
+      "loss": 0.5518,
+      "step": 13729
+    },
+    {
+      "epoch": 3.644895791849197,
+      "grad_norm": 0.4646577938387865,
+      "learning_rate": 1.6836299245233393e-06,
+      "loss": 0.5504,
+      "step": 13730
+    },
+    {
+      "epoch": 3.6451612903225805,
+      "grad_norm": 0.45674061939232097,
+      "learning_rate": 1.6832999435471942e-06,
+      "loss": 0.5501,
+      "step": 13731
+    },
+    {
+      "epoch": 3.645426788795964,
+      "grad_norm": 0.46075901989042545,
+      "learning_rate": 1.6829699784986995e-06,
+      "loss": 0.5415,
+      "step": 13732
+    },
+    {
+      "epoch": 3.6456922872693482,
+      "grad_norm": 0.47252930868368015,
+      "learning_rate": 1.6826400293842904e-06,
+      "loss": 0.5573,
+      "step": 13733
+    },
+    {
+      "epoch": 3.645957785742732,
+      "grad_norm": 0.4642419027528526,
+      "learning_rate": 1.682310096210402e-06,
+      "loss": 0.5426,
+      "step": 13734
+    },
+    {
+      "epoch": 3.6462232842161155,
+      "grad_norm": 0.4422289085002464,
+      "learning_rate": 1.6819801789834695e-06,
+      "loss": 0.5207,
+      "step": 13735
+    },
+    {
+      "epoch": 3.6464887826894996,
+      "grad_norm": 0.45187185258765883,
+      "learning_rate": 1.6816502777099255e-06,
+      "loss": 0.5632,
+      "step": 13736
+    },
+    {
+      "epoch": 3.6467542811628832,
+      "grad_norm": 0.4543059246818115,
+      "learning_rate": 1.6813203923962062e-06,
+      "loss": 0.5148,
+      "step": 13737
+    },
+    {
+      "epoch": 3.647019779636267,
+      "grad_norm": 0.45540447985108073,
+      "learning_rate": 1.6809905230487424e-06,
+      "loss": 0.5285,
+      "step": 13738
+    },
+    {
+      "epoch": 3.647285278109651,
+      "grad_norm": 0.44950988158098654,
+      "learning_rate": 1.680660669673969e-06,
+      "loss": 0.5581,
+      "step": 13739
+    },
+    {
+      "epoch": 3.6475507765830346,
+      "grad_norm": 0.44480987406676287,
+      "learning_rate": 1.6803308322783186e-06,
+      "loss": 0.4945,
+      "step": 13740
+    },
+    {
+      "epoch": 3.6478162750564183,
+      "grad_norm": 0.4437292871157782,
+      "learning_rate": 1.6800010108682235e-06,
+      "loss": 0.5191,
+      "step": 13741
+    },
+    {
+      "epoch": 3.6480817735298023,
+      "grad_norm": 0.46783100596878574,
+      "learning_rate": 1.679671205450117e-06,
+      "loss": 0.5154,
+      "step": 13742
+    },
+    {
+      "epoch": 3.648347272003186,
+      "grad_norm": 0.4662660159818644,
+      "learning_rate": 1.6793414160304306e-06,
+      "loss": 0.5106,
+      "step": 13743
+    },
+    {
+      "epoch": 3.6486127704765696,
+      "grad_norm": 0.45063723228620567,
+      "learning_rate": 1.6790116426155952e-06,
+      "loss": 0.5191,
+      "step": 13744
+    },
+    {
+      "epoch": 3.6488782689499537,
+      "grad_norm": 0.47530364731471897,
+      "learning_rate": 1.6786818852120434e-06,
+      "loss": 0.5571,
+      "step": 13745
+    },
+    {
+      "epoch": 3.6491437674233373,
+      "grad_norm": 0.45783278452546355,
+      "learning_rate": 1.6783521438262055e-06,
+      "loss": 0.5163,
+      "step": 13746
+    },
+    {
+      "epoch": 3.649409265896721,
+      "grad_norm": 0.4658339520731718,
+      "learning_rate": 1.6780224184645125e-06,
+      "loss": 0.5324,
+      "step": 13747
+    },
+    {
+      "epoch": 3.649674764370105,
+      "grad_norm": 0.46906537898002826,
+      "learning_rate": 1.677692709133396e-06,
+      "loss": 0.5195,
+      "step": 13748
+    },
+    {
+      "epoch": 3.6499402628434887,
+      "grad_norm": 0.4796470923598252,
+      "learning_rate": 1.6773630158392839e-06,
+      "loss": 0.5642,
+      "step": 13749
+    },
+    {
+      "epoch": 3.6502057613168724,
+      "grad_norm": 0.4624525968299761,
+      "learning_rate": 1.6770333385886072e-06,
+      "loss": 0.5394,
+      "step": 13750
+    },
+    {
+      "epoch": 3.6504712597902564,
+      "grad_norm": 0.46916695987056584,
+      "learning_rate": 1.6767036773877955e-06,
+      "loss": 0.533,
+      "step": 13751
+    },
+    {
+      "epoch": 3.65073675826364,
+      "grad_norm": 0.4601153019971286,
+      "learning_rate": 1.6763740322432774e-06,
+      "loss": 0.5526,
+      "step": 13752
+    },
+    {
+      "epoch": 3.6510022567370237,
+      "grad_norm": 0.4666390969867862,
+      "learning_rate": 1.6760444031614825e-06,
+      "loss": 0.5158,
+      "step": 13753
+    },
+    {
+      "epoch": 3.651267755210408,
+      "grad_norm": 0.48048271925305885,
+      "learning_rate": 1.6757147901488402e-06,
+      "loss": 0.5817,
+      "step": 13754
+    },
+    {
+      "epoch": 3.6515332536837914,
+      "grad_norm": 0.4568907473787726,
+      "learning_rate": 1.6753851932117767e-06,
+      "loss": 0.5532,
+      "step": 13755
+    },
+    {
+      "epoch": 3.651798752157175,
+      "grad_norm": 0.4656320036619044,
+      "learning_rate": 1.6750556123567213e-06,
+      "loss": 0.5276,
+      "step": 13756
+    },
+    {
+      "epoch": 3.6520642506305587,
+      "grad_norm": 0.48038360636326377,
+      "learning_rate": 1.6747260475901005e-06,
+      "loss": 0.5521,
+      "step": 13757
+    },
+    {
+      "epoch": 3.652329749103943,
+      "grad_norm": 0.4591990968948632,
+      "learning_rate": 1.6743964989183432e-06,
+      "loss": 0.5403,
+      "step": 13758
+    },
+    {
+      "epoch": 3.6525952475773265,
+      "grad_norm": 0.4497268732977616,
+      "learning_rate": 1.6740669663478752e-06,
+      "loss": 0.5411,
+      "step": 13759
+    },
+    {
+      "epoch": 3.65286074605071,
+      "grad_norm": 0.4695250918736584,
+      "learning_rate": 1.6737374498851242e-06,
+      "loss": 0.5799,
+      "step": 13760
+    },
+    {
+      "epoch": 3.6531262445240937,
+      "grad_norm": 0.47880899544657424,
+      "learning_rate": 1.6734079495365152e-06,
+      "loss": 0.509,
+      "step": 13761
+    },
+    {
+      "epoch": 3.653391742997478,
+      "grad_norm": 0.45150262748154263,
+      "learning_rate": 1.6730784653084753e-06,
+      "loss": 0.5422,
+      "step": 13762
+    },
+    {
+      "epoch": 3.6536572414708615,
+      "grad_norm": 0.46926536260451784,
+      "learning_rate": 1.6727489972074302e-06,
+      "loss": 0.5458,
+      "step": 13763
+    },
+    {
+      "epoch": 3.653922739944245,
+      "grad_norm": 0.46755137129013313,
+      "learning_rate": 1.6724195452398045e-06,
+      "loss": 0.5474,
+      "step": 13764
+    },
+    {
+      "epoch": 3.654188238417629,
+      "grad_norm": 0.475678818659334,
+      "learning_rate": 1.6720901094120242e-06,
+      "loss": 0.5475,
+      "step": 13765
+    },
+    {
+      "epoch": 3.654453736891013,
+      "grad_norm": 0.46702358092713714,
+      "learning_rate": 1.671760689730515e-06,
+      "loss": 0.5482,
+      "step": 13766
+    },
+    {
+      "epoch": 3.6547192353643965,
+      "grad_norm": 0.45767050178879953,
+      "learning_rate": 1.671431286201699e-06,
+      "loss": 0.564,
+      "step": 13767
+    },
+    {
+      "epoch": 3.6549847338377806,
+      "grad_norm": 0.46927537560672444,
+      "learning_rate": 1.6711018988320018e-06,
+      "loss": 0.5275,
+      "step": 13768
+    },
+    {
+      "epoch": 3.655250232311164,
+      "grad_norm": 0.47007928774544794,
+      "learning_rate": 1.670772527627847e-06,
+      "loss": 0.5544,
+      "step": 13769
+    },
+    {
+      "epoch": 3.655515730784548,
+      "grad_norm": 0.4621877853463103,
+      "learning_rate": 1.6704431725956582e-06,
+      "loss": 0.5267,
+      "step": 13770
+    },
+    {
+      "epoch": 3.655781229257932,
+      "grad_norm": 0.4719922100923226,
+      "learning_rate": 1.6701138337418588e-06,
+      "loss": 0.5462,
+      "step": 13771
+    },
+    {
+      "epoch": 3.6560467277313156,
+      "grad_norm": 0.4856879016586914,
+      "learning_rate": 1.6697845110728716e-06,
+      "loss": 0.5934,
+      "step": 13772
+    },
+    {
+      "epoch": 3.656312226204699,
+      "grad_norm": 0.46600035176760796,
+      "learning_rate": 1.669455204595119e-06,
+      "loss": 0.5879,
+      "step": 13773
+    },
+    {
+      "epoch": 3.6565777246780833,
+      "grad_norm": 0.473317566841543,
+      "learning_rate": 1.6691259143150234e-06,
+      "loss": 0.5468,
+      "step": 13774
+    },
+    {
+      "epoch": 3.656843223151467,
+      "grad_norm": 0.4700812442788767,
+      "learning_rate": 1.6687966402390072e-06,
+      "loss": 0.5505,
+      "step": 13775
+    },
+    {
+      "epoch": 3.6571087216248506,
+      "grad_norm": 0.4617394420070805,
+      "learning_rate": 1.668467382373491e-06,
+      "loss": 0.5372,
+      "step": 13776
+    },
+    {
+      "epoch": 3.6573742200982347,
+      "grad_norm": 0.46302034078613996,
+      "learning_rate": 1.6681381407248976e-06,
+      "loss": 0.5342,
+      "step": 13777
+    },
+    {
+      "epoch": 3.6576397185716183,
+      "grad_norm": 0.47138132180944026,
+      "learning_rate": 1.6678089152996466e-06,
+      "loss": 0.5757,
+      "step": 13778
+    },
+    {
+      "epoch": 3.657905217045002,
+      "grad_norm": 0.4743354747005439,
+      "learning_rate": 1.6674797061041592e-06,
+      "loss": 0.5375,
+      "step": 13779
+    },
+    {
+      "epoch": 3.658170715518386,
+      "grad_norm": 0.4672946190054541,
+      "learning_rate": 1.6671505131448562e-06,
+      "loss": 0.538,
+      "step": 13780
+    },
+    {
+      "epoch": 3.6584362139917697,
+      "grad_norm": 0.4672790473691408,
+      "learning_rate": 1.6668213364281571e-06,
+      "loss": 0.5306,
+      "step": 13781
+    },
+    {
+      "epoch": 3.6587017124651533,
+      "grad_norm": 0.47824961447080494,
+      "learning_rate": 1.666492175960482e-06,
+      "loss": 0.4846,
+      "step": 13782
+    },
+    {
+      "epoch": 3.658967210938537,
+      "grad_norm": 0.46018164315105453,
+      "learning_rate": 1.6661630317482514e-06,
+      "loss": 0.5169,
+      "step": 13783
+    },
+    {
+      "epoch": 3.659232709411921,
+      "grad_norm": 0.4672544641386213,
+      "learning_rate": 1.6658339037978822e-06,
+      "loss": 0.5224,
+      "step": 13784
+    },
+    {
+      "epoch": 3.6594982078853047,
+      "grad_norm": 0.4617404304902396,
+      "learning_rate": 1.665504792115794e-06,
+      "loss": 0.5311,
+      "step": 13785
+    },
+    {
+      "epoch": 3.6597637063586883,
+      "grad_norm": 0.46531970146958807,
+      "learning_rate": 1.665175696708406e-06,
+      "loss": 0.5393,
+      "step": 13786
+    },
+    {
+      "epoch": 3.660029204832072,
+      "grad_norm": 0.4728687915400536,
+      "learning_rate": 1.6648466175821354e-06,
+      "loss": 0.5558,
+      "step": 13787
+    },
+    {
+      "epoch": 3.660294703305456,
+      "grad_norm": 0.47411357878230637,
+      "learning_rate": 1.6645175547434008e-06,
+      "loss": 0.5703,
+      "step": 13788
+    },
+    {
+      "epoch": 3.6605602017788397,
+      "grad_norm": 0.4691443142671677,
+      "learning_rate": 1.66418850819862e-06,
+      "loss": 0.5641,
+      "step": 13789
+    },
+    {
+      "epoch": 3.6608257002522233,
+      "grad_norm": 0.46475373500301737,
+      "learning_rate": 1.6638594779542094e-06,
+      "loss": 0.5819,
+      "step": 13790
+    },
+    {
+      "epoch": 3.6610911987256074,
+      "grad_norm": 0.44387761698007133,
+      "learning_rate": 1.6635304640165858e-06,
+      "loss": 0.5553,
+      "step": 13791
+    },
+    {
+      "epoch": 3.661356697198991,
+      "grad_norm": 0.4810550724624136,
+      "learning_rate": 1.6632014663921663e-06,
+      "loss": 0.5656,
+      "step": 13792
+    },
+    {
+      "epoch": 3.6616221956723747,
+      "grad_norm": 0.4689449555235533,
+      "learning_rate": 1.6628724850873672e-06,
+      "loss": 0.5521,
+      "step": 13793
+    },
+    {
+      "epoch": 3.6618876941457588,
+      "grad_norm": 0.45270153457013607,
+      "learning_rate": 1.6625435201086043e-06,
+      "loss": 0.527,
+      "step": 13794
+    },
+    {
+      "epoch": 3.6621531926191424,
+      "grad_norm": 0.4640409986992459,
+      "learning_rate": 1.6622145714622945e-06,
+      "loss": 0.5276,
+      "step": 13795
+    },
+    {
+      "epoch": 3.662418691092526,
+      "grad_norm": 0.46032140187441206,
+      "learning_rate": 1.6618856391548501e-06,
+      "loss": 0.5382,
+      "step": 13796
+    },
+    {
+      "epoch": 3.66268418956591,
+      "grad_norm": 0.4593336761700628,
+      "learning_rate": 1.6615567231926883e-06,
+      "loss": 0.5715,
+      "step": 13797
+    },
+    {
+      "epoch": 3.6629496880392938,
+      "grad_norm": 0.4650367964551919,
+      "learning_rate": 1.6612278235822226e-06,
+      "loss": 0.5569,
+      "step": 13798
+    },
+    {
+      "epoch": 3.6632151865126774,
+      "grad_norm": 0.4759064884383785,
+      "learning_rate": 1.6608989403298687e-06,
+      "loss": 0.5341,
+      "step": 13799
+    },
+    {
+      "epoch": 3.6634806849860615,
+      "grad_norm": 0.45127955764380284,
+      "learning_rate": 1.6605700734420404e-06,
+      "loss": 0.5545,
+      "step": 13800
+    },
+    {
+      "epoch": 3.663746183459445,
+      "grad_norm": 0.47173306366153084,
+      "learning_rate": 1.6602412229251506e-06,
+      "loss": 0.5589,
+      "step": 13801
+    },
+    {
+      "epoch": 3.664011681932829,
+      "grad_norm": 0.45609750597281096,
+      "learning_rate": 1.6599123887856125e-06,
+      "loss": 0.5559,
+      "step": 13802
+    },
+    {
+      "epoch": 3.664277180406213,
+      "grad_norm": 0.44657638749684,
+      "learning_rate": 1.6595835710298399e-06,
+      "loss": 0.5483,
+      "step": 13803
+    },
+    {
+      "epoch": 3.6645426788795965,
+      "grad_norm": 0.4693042614205532,
+      "learning_rate": 1.6592547696642454e-06,
+      "loss": 0.5263,
+      "step": 13804
+    },
+    {
+      "epoch": 3.66480817735298,
+      "grad_norm": 0.4673434022835937,
+      "learning_rate": 1.6589259846952416e-06,
+      "loss": 0.5281,
+      "step": 13805
+    },
+    {
+      "epoch": 3.6650736758263642,
+      "grad_norm": 0.44618661743463467,
+      "learning_rate": 1.6585972161292407e-06,
+      "loss": 0.5651,
+      "step": 13806
+    },
+    {
+      "epoch": 3.665339174299748,
+      "grad_norm": 0.47838916677146576,
+      "learning_rate": 1.6582684639726538e-06,
+      "loss": 0.5778,
+      "step": 13807
+    },
+    {
+      "epoch": 3.6656046727731315,
+      "grad_norm": 0.44572004513080526,
+      "learning_rate": 1.6579397282318924e-06,
+      "loss": 0.5453,
+      "step": 13808
+    },
+    {
+      "epoch": 3.6658701712465156,
+      "grad_norm": 0.46079325263267695,
+      "learning_rate": 1.6576110089133685e-06,
+      "loss": 0.5757,
+      "step": 13809
+    },
+    {
+      "epoch": 3.6661356697198992,
+      "grad_norm": 0.47767259239488713,
+      "learning_rate": 1.6572823060234922e-06,
+      "loss": 0.5318,
+      "step": 13810
+    },
+    {
+      "epoch": 3.666401168193283,
+      "grad_norm": 0.44752203724949646,
+      "learning_rate": 1.6569536195686742e-06,
+      "loss": 0.5669,
+      "step": 13811
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 0.4481408048638044,
+      "learning_rate": 1.6566249495553257e-06,
+      "loss": 0.5674,
+      "step": 13812
+    },
+    {
+      "epoch": 3.6669321651400506,
+      "grad_norm": 0.4583911188615994,
+      "learning_rate": 1.656296295989855e-06,
+      "loss": 0.5085,
+      "step": 13813
+    },
+    {
+      "epoch": 3.6671976636134342,
+      "grad_norm": 0.46124650894906316,
+      "learning_rate": 1.6559676588786719e-06,
+      "loss": 0.564,
+      "step": 13814
+    },
+    {
+      "epoch": 3.667463162086818,
+      "grad_norm": 0.4636484665243648,
+      "learning_rate": 1.6556390382281863e-06,
+      "loss": 0.5378,
+      "step": 13815
+    },
+    {
+      "epoch": 3.6677286605602015,
+      "grad_norm": 0.46431658553635075,
+      "learning_rate": 1.6553104340448064e-06,
+      "loss": 0.5128,
+      "step": 13816
+    },
+    {
+      "epoch": 3.6679941590335856,
+      "grad_norm": 0.4595474444114003,
+      "learning_rate": 1.6549818463349414e-06,
+      "loss": 0.5736,
+      "step": 13817
+    },
+    {
+      "epoch": 3.6682596575069693,
+      "grad_norm": 0.47041724140920504,
+      "learning_rate": 1.654653275105e-06,
+      "loss": 0.5793,
+      "step": 13818
+    },
+    {
+      "epoch": 3.668525155980353,
+      "grad_norm": 0.46345124598520104,
+      "learning_rate": 1.6543247203613888e-06,
+      "loss": 0.5654,
+      "step": 13819
+    },
+    {
+      "epoch": 3.668790654453737,
+      "grad_norm": 0.4588294040718161,
+      "learning_rate": 1.653996182110516e-06,
+      "loss": 0.5401,
+      "step": 13820
+    },
+    {
+      "epoch": 3.6690561529271206,
+      "grad_norm": 0.45203161847891504,
+      "learning_rate": 1.6536676603587895e-06,
+      "loss": 0.5767,
+      "step": 13821
+    },
+    {
+      "epoch": 3.6693216514005043,
+      "grad_norm": 0.47595138219758726,
+      "learning_rate": 1.6533391551126155e-06,
+      "loss": 0.5463,
+      "step": 13822
+    },
+    {
+      "epoch": 3.6695871498738883,
+      "grad_norm": 0.46340827814896357,
+      "learning_rate": 1.6530106663784014e-06,
+      "loss": 0.5431,
+      "step": 13823
+    },
+    {
+      "epoch": 3.669852648347272,
+      "grad_norm": 0.47845663224381507,
+      "learning_rate": 1.6526821941625527e-06,
+      "loss": 0.5534,
+      "step": 13824
+    },
+    {
+      "epoch": 3.6701181468206556,
+      "grad_norm": 0.4640173704946256,
+      "learning_rate": 1.652353738471476e-06,
+      "loss": 0.5181,
+      "step": 13825
+    },
+    {
+      "epoch": 3.6703836452940397,
+      "grad_norm": 0.4463265522923325,
+      "learning_rate": 1.6520252993115764e-06,
+      "loss": 0.527,
+      "step": 13826
+    },
+    {
+      "epoch": 3.6706491437674234,
+      "grad_norm": 0.4769353909832081,
+      "learning_rate": 1.65169687668926e-06,
+      "loss": 0.5346,
+      "step": 13827
+    },
+    {
+      "epoch": 3.670914642240807,
+      "grad_norm": 0.47438968445297885,
+      "learning_rate": 1.6513684706109311e-06,
+      "loss": 0.5638,
+      "step": 13828
+    },
+    {
+      "epoch": 3.671180140714191,
+      "grad_norm": 0.4815962358437599,
+      "learning_rate": 1.6510400810829963e-06,
+      "loss": 0.5452,
+      "step": 13829
+    },
+    {
+      "epoch": 3.6714456391875747,
+      "grad_norm": 0.45755692206997206,
+      "learning_rate": 1.6507117081118573e-06,
+      "loss": 0.5519,
+      "step": 13830
+    },
+    {
+      "epoch": 3.6717111376609584,
+      "grad_norm": 0.4712517847688879,
+      "learning_rate": 1.6503833517039195e-06,
+      "loss": 0.5438,
+      "step": 13831
+    },
+    {
+      "epoch": 3.6719766361343424,
+      "grad_norm": 0.47004399864078256,
+      "learning_rate": 1.6500550118655864e-06,
+      "loss": 0.5115,
+      "step": 13832
+    },
+    {
+      "epoch": 3.672242134607726,
+      "grad_norm": 0.46609737019115044,
+      "learning_rate": 1.6497266886032618e-06,
+      "loss": 0.5258,
+      "step": 13833
+    },
+    {
+      "epoch": 3.6725076330811097,
+      "grad_norm": 0.47863831508876836,
+      "learning_rate": 1.6493983819233483e-06,
+      "loss": 0.5779,
+      "step": 13834
+    },
+    {
+      "epoch": 3.672773131554494,
+      "grad_norm": 0.45920612412700157,
+      "learning_rate": 1.64907009183225e-06,
+      "loss": 0.5617,
+      "step": 13835
+    },
+    {
+      "epoch": 3.6730386300278774,
+      "grad_norm": 0.4675138351061314,
+      "learning_rate": 1.6487418183363674e-06,
+      "loss": 0.5699,
+      "step": 13836
+    },
+    {
+      "epoch": 3.673304128501261,
+      "grad_norm": 0.4458090676072325,
+      "learning_rate": 1.6484135614421037e-06,
+      "loss": 0.551,
+      "step": 13837
+    },
+    {
+      "epoch": 3.6735696269746447,
+      "grad_norm": 0.46268100456788447,
+      "learning_rate": 1.6480853211558606e-06,
+      "loss": 0.5222,
+      "step": 13838
+    },
+    {
+      "epoch": 3.673835125448029,
+      "grad_norm": 0.45726517388185506,
+      "learning_rate": 1.6477570974840396e-06,
+      "loss": 0.5345,
+      "step": 13839
+    },
+    {
+      "epoch": 3.6741006239214125,
+      "grad_norm": 0.46255457686740953,
+      "learning_rate": 1.6474288904330416e-06,
+      "loss": 0.5962,
+      "step": 13840
+    },
+    {
+      "epoch": 3.674366122394796,
+      "grad_norm": 0.46181795465539005,
+      "learning_rate": 1.6471007000092692e-06,
+      "loss": 0.5185,
+      "step": 13841
+    },
+    {
+      "epoch": 3.6746316208681797,
+      "grad_norm": 0.4549028654235298,
+      "learning_rate": 1.6467725262191197e-06,
+      "loss": 0.5227,
+      "step": 13842
+    },
+    {
+      "epoch": 3.674897119341564,
+      "grad_norm": 0.4816546919531103,
+      "learning_rate": 1.6464443690689952e-06,
+      "loss": 0.5642,
+      "step": 13843
+    },
+    {
+      "epoch": 3.6751626178149475,
+      "grad_norm": 0.4546686745562164,
+      "learning_rate": 1.6461162285652954e-06,
+      "loss": 0.5461,
+      "step": 13844
+    },
+    {
+      "epoch": 3.675428116288331,
+      "grad_norm": 0.45597902204896146,
+      "learning_rate": 1.6457881047144202e-06,
+      "loss": 0.5377,
+      "step": 13845
+    },
+    {
+      "epoch": 3.675693614761715,
+      "grad_norm": 0.4751503934380114,
+      "learning_rate": 1.6454599975227693e-06,
+      "loss": 0.5724,
+      "step": 13846
+    },
+    {
+      "epoch": 3.675959113235099,
+      "grad_norm": 0.45704125522101935,
+      "learning_rate": 1.6451319069967397e-06,
+      "loss": 0.54,
+      "step": 13847
+    },
+    {
+      "epoch": 3.6762246117084825,
+      "grad_norm": 0.4761212291809819,
+      "learning_rate": 1.6448038331427307e-06,
+      "loss": 0.5376,
+      "step": 13848
+    },
+    {
+      "epoch": 3.6764901101818666,
+      "grad_norm": 0.47287030099643695,
+      "learning_rate": 1.6444757759671412e-06,
+      "loss": 0.5593,
+      "step": 13849
+    },
+    {
+      "epoch": 3.67675560865525,
+      "grad_norm": 0.4606940443573965,
+      "learning_rate": 1.6441477354763685e-06,
+      "loss": 0.5558,
+      "step": 13850
+    },
+    {
+      "epoch": 3.677021107128634,
+      "grad_norm": 0.458971606590005,
+      "learning_rate": 1.6438197116768107e-06,
+      "loss": 0.5474,
+      "step": 13851
+    },
+    {
+      "epoch": 3.677286605602018,
+      "grad_norm": 0.48013030574761045,
+      "learning_rate": 1.6434917045748652e-06,
+      "loss": 0.5691,
+      "step": 13852
+    },
+    {
+      "epoch": 3.6775521040754016,
+      "grad_norm": 0.4891441748302624,
+      "learning_rate": 1.643163714176928e-06,
+      "loss": 0.5661,
+      "step": 13853
+    },
+    {
+      "epoch": 3.677817602548785,
+      "grad_norm": 0.4632434781297104,
+      "learning_rate": 1.6428357404893964e-06,
+      "loss": 0.5204,
+      "step": 13854
+    },
+    {
+      "epoch": 3.6780831010221693,
+      "grad_norm": 0.44504795363461613,
+      "learning_rate": 1.6425077835186666e-06,
+      "loss": 0.5123,
+      "step": 13855
+    },
+    {
+      "epoch": 3.678348599495553,
+      "grad_norm": 0.4612803336897518,
+      "learning_rate": 1.6421798432711344e-06,
+      "loss": 0.572,
+      "step": 13856
+    },
+    {
+      "epoch": 3.6786140979689366,
+      "grad_norm": 0.4955079775505199,
+      "learning_rate": 1.6418519197531953e-06,
+      "loss": 0.5494,
+      "step": 13857
+    },
+    {
+      "epoch": 3.6788795964423207,
+      "grad_norm": 0.468319232696857,
+      "learning_rate": 1.6415240129712462e-06,
+      "loss": 0.5343,
+      "step": 13858
+    },
+    {
+      "epoch": 3.6791450949157043,
+      "grad_norm": 0.4570726011425613,
+      "learning_rate": 1.6411961229316797e-06,
+      "loss": 0.5401,
+      "step": 13859
+    },
+    {
+      "epoch": 3.679410593389088,
+      "grad_norm": 0.48574608820282317,
+      "learning_rate": 1.6408682496408918e-06,
+      "loss": 0.5688,
+      "step": 13860
+    },
+    {
+      "epoch": 3.679676091862472,
+      "grad_norm": 0.45720660188781775,
+      "learning_rate": 1.640540393105276e-06,
+      "loss": 0.5418,
+      "step": 13861
+    },
+    {
+      "epoch": 3.6799415903358557,
+      "grad_norm": 0.4530380208607555,
+      "learning_rate": 1.640212553331227e-06,
+      "loss": 0.5411,
+      "step": 13862
+    },
+    {
+      "epoch": 3.6802070888092393,
+      "grad_norm": 0.46057470830283,
+      "learning_rate": 1.6398847303251385e-06,
+      "loss": 0.5231,
+      "step": 13863
+    },
+    {
+      "epoch": 3.6804725872826234,
+      "grad_norm": 0.45520904342865004,
+      "learning_rate": 1.6395569240934042e-06,
+      "loss": 0.5154,
+      "step": 13864
+    },
+    {
+      "epoch": 3.680738085756007,
+      "grad_norm": 0.4727417815262151,
+      "learning_rate": 1.639229134642416e-06,
+      "loss": 0.5682,
+      "step": 13865
+    },
+    {
+      "epoch": 3.6810035842293907,
+      "grad_norm": 0.4504506264367193,
+      "learning_rate": 1.6389013619785673e-06,
+      "loss": 0.5716,
+      "step": 13866
+    },
+    {
+      "epoch": 3.6812690827027743,
+      "grad_norm": 0.46328484284373306,
+      "learning_rate": 1.6385736061082502e-06,
+      "loss": 0.514,
+      "step": 13867
+    },
+    {
+      "epoch": 3.6815345811761584,
+      "grad_norm": 0.47518937869733086,
+      "learning_rate": 1.638245867037857e-06,
+      "loss": 0.5499,
+      "step": 13868
+    },
+    {
+      "epoch": 3.681800079649542,
+      "grad_norm": 0.43831306084815297,
+      "learning_rate": 1.6379181447737796e-06,
+      "loss": 0.556,
+      "step": 13869
+    },
+    {
+      "epoch": 3.6820655781229257,
+      "grad_norm": 0.46597164407659364,
+      "learning_rate": 1.6375904393224095e-06,
+      "loss": 0.5626,
+      "step": 13870
+    },
+    {
+      "epoch": 3.6823310765963093,
+      "grad_norm": 0.4724242796507134,
+      "learning_rate": 1.6372627506901368e-06,
+      "loss": 0.5574,
+      "step": 13871
+    },
+    {
+      "epoch": 3.6825965750696934,
+      "grad_norm": 0.44406006482576116,
+      "learning_rate": 1.6369350788833526e-06,
+      "loss": 0.5441,
+      "step": 13872
+    },
+    {
+      "epoch": 3.682862073543077,
+      "grad_norm": 0.4746675405411556,
+      "learning_rate": 1.6366074239084478e-06,
+      "loss": 0.5405,
+      "step": 13873
+    },
+    {
+      "epoch": 3.6831275720164607,
+      "grad_norm": 0.45067043633564874,
+      "learning_rate": 1.6362797857718121e-06,
+      "loss": 0.5311,
+      "step": 13874
+    },
+    {
+      "epoch": 3.6833930704898448,
+      "grad_norm": 0.4715372294439943,
+      "learning_rate": 1.6359521644798363e-06,
+      "loss": 0.5624,
+      "step": 13875
+    },
+    {
+      "epoch": 3.6836585689632284,
+      "grad_norm": 0.4641625166813518,
+      "learning_rate": 1.6356245600389076e-06,
+      "loss": 0.5546,
+      "step": 13876
+    },
+    {
+      "epoch": 3.683924067436612,
+      "grad_norm": 0.4841146630617197,
+      "learning_rate": 1.6352969724554169e-06,
+      "loss": 0.5753,
+      "step": 13877
+    },
+    {
+      "epoch": 3.684189565909996,
+      "grad_norm": 0.47600627256485334,
+      "learning_rate": 1.634969401735752e-06,
+      "loss": 0.5089,
+      "step": 13878
+    },
+    {
+      "epoch": 3.6844550643833798,
+      "grad_norm": 0.4388927895339281,
+      "learning_rate": 1.6346418478863017e-06,
+      "loss": 0.5042,
+      "step": 13879
+    },
+    {
+      "epoch": 3.6847205628567634,
+      "grad_norm": 0.45746641196419136,
+      "learning_rate": 1.634314310913454e-06,
+      "loss": 0.5546,
+      "step": 13880
+    },
+    {
+      "epoch": 3.6849860613301475,
+      "grad_norm": 0.4771462424705586,
+      "learning_rate": 1.6339867908235975e-06,
+      "loss": 0.5552,
+      "step": 13881
+    },
+    {
+      "epoch": 3.685251559803531,
+      "grad_norm": 0.4712223924104759,
+      "learning_rate": 1.6336592876231187e-06,
+      "loss": 0.5445,
+      "step": 13882
+    },
+    {
+      "epoch": 3.685517058276915,
+      "grad_norm": 0.44977422498067143,
+      "learning_rate": 1.6333318013184047e-06,
+      "loss": 0.5497,
+      "step": 13883
+    },
+    {
+      "epoch": 3.685782556750299,
+      "grad_norm": 0.46828108350791414,
+      "learning_rate": 1.6330043319158425e-06,
+      "loss": 0.5554,
+      "step": 13884
+    },
+    {
+      "epoch": 3.6860480552236825,
+      "grad_norm": 0.46552111808760377,
+      "learning_rate": 1.6326768794218187e-06,
+      "loss": 0.5647,
+      "step": 13885
+    },
+    {
+      "epoch": 3.686313553697066,
+      "grad_norm": 0.4737033665491398,
+      "learning_rate": 1.6323494438427193e-06,
+      "loss": 0.5423,
+      "step": 13886
+    },
+    {
+      "epoch": 3.6865790521704502,
+      "grad_norm": 0.4724461744141565,
+      "learning_rate": 1.6320220251849312e-06,
+      "loss": 0.5748,
+      "step": 13887
+    },
+    {
+      "epoch": 3.686844550643834,
+      "grad_norm": 0.43996545078195703,
+      "learning_rate": 1.6316946234548375e-06,
+      "loss": 0.5245,
+      "step": 13888
+    },
+    {
+      "epoch": 3.6871100491172175,
+      "grad_norm": 0.4638149328901893,
+      "learning_rate": 1.6313672386588245e-06,
+      "loss": 0.5296,
+      "step": 13889
+    },
+    {
+      "epoch": 3.6873755475906016,
+      "grad_norm": 0.4601278630077606,
+      "learning_rate": 1.6310398708032773e-06,
+      "loss": 0.5623,
+      "step": 13890
+    },
+    {
+      "epoch": 3.6876410460639852,
+      "grad_norm": 0.4769117955376731,
+      "learning_rate": 1.6307125198945806e-06,
+      "loss": 0.5181,
+      "step": 13891
+    },
+    {
+      "epoch": 3.687906544537369,
+      "grad_norm": 0.46463664186935766,
+      "learning_rate": 1.630385185939118e-06,
+      "loss": 0.5392,
+      "step": 13892
+    },
+    {
+      "epoch": 3.688172043010753,
+      "grad_norm": 0.466850451238197,
+      "learning_rate": 1.6300578689432742e-06,
+      "loss": 0.5566,
+      "step": 13893
+    },
+    {
+      "epoch": 3.6884375414841366,
+      "grad_norm": 0.48184318201135246,
+      "learning_rate": 1.629730568913431e-06,
+      "loss": 0.5546,
+      "step": 13894
+    },
+    {
+      "epoch": 3.6887030399575202,
+      "grad_norm": 0.4577660396965125,
+      "learning_rate": 1.6294032858559727e-06,
+      "loss": 0.5497,
+      "step": 13895
+    },
+    {
+      "epoch": 3.688968538430904,
+      "grad_norm": 0.45783928710687316,
+      "learning_rate": 1.6290760197772818e-06,
+      "loss": 0.5568,
+      "step": 13896
+    },
+    {
+      "epoch": 3.6892340369042875,
+      "grad_norm": 0.4628803253871579,
+      "learning_rate": 1.6287487706837407e-06,
+      "loss": 0.5456,
+      "step": 13897
+    },
+    {
+      "epoch": 3.6894995353776716,
+      "grad_norm": 0.4635932855947962,
+      "learning_rate": 1.6284215385817326e-06,
+      "loss": 0.537,
+      "step": 13898
+    },
+    {
+      "epoch": 3.6897650338510553,
+      "grad_norm": 0.47520761872809547,
+      "learning_rate": 1.6280943234776378e-06,
+      "loss": 0.5488,
+      "step": 13899
+    },
+    {
+      "epoch": 3.690030532324439,
+      "grad_norm": 0.4754631243992757,
+      "learning_rate": 1.6277671253778387e-06,
+      "loss": 0.5216,
+      "step": 13900
+    },
+    {
+      "epoch": 3.690296030797823,
+      "grad_norm": 0.46259647229771034,
+      "learning_rate": 1.627439944288716e-06,
+      "loss": 0.5298,
+      "step": 13901
+    },
+    {
+      "epoch": 3.6905615292712066,
+      "grad_norm": 0.4585881187975157,
+      "learning_rate": 1.627112780216651e-06,
+      "loss": 0.5369,
+      "step": 13902
+    },
+    {
+      "epoch": 3.6908270277445903,
+      "grad_norm": 0.45425879226152227,
+      "learning_rate": 1.6267856331680237e-06,
+      "loss": 0.5279,
+      "step": 13903
+    },
+    {
+      "epoch": 3.6910925262179743,
+      "grad_norm": 0.46521762478199624,
+      "learning_rate": 1.6264585031492158e-06,
+      "loss": 0.5847,
+      "step": 13904
+    },
+    {
+      "epoch": 3.691358024691358,
+      "grad_norm": 0.4660984073751585,
+      "learning_rate": 1.6261313901666047e-06,
+      "loss": 0.5332,
+      "step": 13905
+    },
+    {
+      "epoch": 3.6916235231647416,
+      "grad_norm": 0.48665427669182987,
+      "learning_rate": 1.6258042942265712e-06,
+      "loss": 0.5345,
+      "step": 13906
+    },
+    {
+      "epoch": 3.6918890216381257,
+      "grad_norm": 0.45131391718563035,
+      "learning_rate": 1.6254772153354943e-06,
+      "loss": 0.5276,
+      "step": 13907
+    },
+    {
+      "epoch": 3.6921545201115094,
+      "grad_norm": 0.4649475060263303,
+      "learning_rate": 1.6251501534997529e-06,
+      "loss": 0.5486,
+      "step": 13908
+    },
+    {
+      "epoch": 3.692420018584893,
+      "grad_norm": 0.46623062789875286,
+      "learning_rate": 1.6248231087257254e-06,
+      "loss": 0.5944,
+      "step": 13909
+    },
+    {
+      "epoch": 3.692685517058277,
+      "grad_norm": 0.4682689806776586,
+      "learning_rate": 1.624496081019791e-06,
+      "loss": 0.5826,
+      "step": 13910
+    },
+    {
+      "epoch": 3.6929510155316607,
+      "grad_norm": 0.45466535491127313,
+      "learning_rate": 1.624169070388326e-06,
+      "loss": 0.5555,
+      "step": 13911
+    },
+    {
+      "epoch": 3.6932165140050444,
+      "grad_norm": 0.4609226118457142,
+      "learning_rate": 1.6238420768377084e-06,
+      "loss": 0.5804,
+      "step": 13912
+    },
+    {
+      "epoch": 3.6934820124784284,
+      "grad_norm": 0.4618726155874749,
+      "learning_rate": 1.6235151003743154e-06,
+      "loss": 0.5144,
+      "step": 13913
+    },
+    {
+      "epoch": 3.693747510951812,
+      "grad_norm": 0.4747378174786819,
+      "learning_rate": 1.623188141004524e-06,
+      "loss": 0.5367,
+      "step": 13914
+    },
+    {
+      "epoch": 3.6940130094251957,
+      "grad_norm": 0.4656448617720923,
+      "learning_rate": 1.6228611987347107e-06,
+      "loss": 0.577,
+      "step": 13915
+    },
+    {
+      "epoch": 3.69427850789858,
+      "grad_norm": 0.4821937949863833,
+      "learning_rate": 1.622534273571252e-06,
+      "loss": 0.5548,
+      "step": 13916
+    },
+    {
+      "epoch": 3.6945440063719635,
+      "grad_norm": 0.47272017190138055,
+      "learning_rate": 1.6222073655205232e-06,
+      "loss": 0.5346,
+      "step": 13917
+    },
+    {
+      "epoch": 3.694809504845347,
+      "grad_norm": 0.48947600039813927,
+      "learning_rate": 1.6218804745888996e-06,
+      "loss": 0.5496,
+      "step": 13918
+    },
+    {
+      "epoch": 3.695075003318731,
+      "grad_norm": 0.4628365735585339,
+      "learning_rate": 1.621553600782757e-06,
+      "loss": 0.5553,
+      "step": 13919
+    },
+    {
+      "epoch": 3.695340501792115,
+      "grad_norm": 0.4671841927704189,
+      "learning_rate": 1.62122674410847e-06,
+      "loss": 0.5132,
+      "step": 13920
+    },
+    {
+      "epoch": 3.6956060002654985,
+      "grad_norm": 0.46874744754035863,
+      "learning_rate": 1.620899904572414e-06,
+      "loss": 0.5518,
+      "step": 13921
+    },
+    {
+      "epoch": 3.695871498738882,
+      "grad_norm": 0.4611338121045755,
+      "learning_rate": 1.6205730821809611e-06,
+      "loss": 0.546,
+      "step": 13922
+    },
+    {
+      "epoch": 3.696136997212266,
+      "grad_norm": 0.46299008145604237,
+      "learning_rate": 1.6202462769404866e-06,
+      "loss": 0.5435,
+      "step": 13923
+    },
+    {
+      "epoch": 3.69640249568565,
+      "grad_norm": 0.4674528864076,
+      "learning_rate": 1.619919488857364e-06,
+      "loss": 0.5446,
+      "step": 13924
+    },
+    {
+      "epoch": 3.6966679941590335,
+      "grad_norm": 0.4573710094174378,
+      "learning_rate": 1.6195927179379656e-06,
+      "loss": 0.5627,
+      "step": 13925
+    },
+    {
+      "epoch": 3.696933492632417,
+      "grad_norm": 0.45716538728410083,
+      "learning_rate": 1.6192659641886648e-06,
+      "loss": 0.5618,
+      "step": 13926
+    },
+    {
+      "epoch": 3.697198991105801,
+      "grad_norm": 0.4749133491026972,
+      "learning_rate": 1.618939227615835e-06,
+      "loss": 0.5494,
+      "step": 13927
+    },
+    {
+      "epoch": 3.697464489579185,
+      "grad_norm": 0.4580850014047163,
+      "learning_rate": 1.6186125082258466e-06,
+      "loss": 0.5794,
+      "step": 13928
+    },
+    {
+      "epoch": 3.6977299880525685,
+      "grad_norm": 0.49375317023769816,
+      "learning_rate": 1.6182858060250726e-06,
+      "loss": 0.5403,
+      "step": 13929
+    },
+    {
+      "epoch": 3.6979954865259526,
+      "grad_norm": 0.45791496821629646,
+      "learning_rate": 1.6179591210198842e-06,
+      "loss": 0.5713,
+      "step": 13930
+    },
+    {
+      "epoch": 3.698260984999336,
+      "grad_norm": 0.46493461749910653,
+      "learning_rate": 1.617632453216652e-06,
+      "loss": 0.5328,
+      "step": 13931
+    },
+    {
+      "epoch": 3.69852648347272,
+      "grad_norm": 0.46741895580898485,
+      "learning_rate": 1.617305802621748e-06,
+      "loss": 0.5322,
+      "step": 13932
+    },
+    {
+      "epoch": 3.698791981946104,
+      "grad_norm": 0.48606468392420066,
+      "learning_rate": 1.616979169241543e-06,
+      "loss": 0.5633,
+      "step": 13933
+    },
+    {
+      "epoch": 3.6990574804194876,
+      "grad_norm": 0.46635827365048776,
+      "learning_rate": 1.6166525530824045e-06,
+      "loss": 0.4823,
+      "step": 13934
+    },
+    {
+      "epoch": 3.699322978892871,
+      "grad_norm": 0.483742766973483,
+      "learning_rate": 1.6163259541507045e-06,
+      "loss": 0.5252,
+      "step": 13935
+    },
+    {
+      "epoch": 3.6995884773662553,
+      "grad_norm": 0.48002758390466216,
+      "learning_rate": 1.6159993724528123e-06,
+      "loss": 0.548,
+      "step": 13936
+    },
+    {
+      "epoch": 3.699853975839639,
+      "grad_norm": 0.4803884467536995,
+      "learning_rate": 1.6156728079950962e-06,
+      "loss": 0.5465,
+      "step": 13937
+    },
+    {
+      "epoch": 3.7001194743130226,
+      "grad_norm": 0.4569002973412173,
+      "learning_rate": 1.615346260783926e-06,
+      "loss": 0.5542,
+      "step": 13938
+    },
+    {
+      "epoch": 3.7003849727864067,
+      "grad_norm": 0.46924788082767865,
+      "learning_rate": 1.6150197308256703e-06,
+      "loss": 0.557,
+      "step": 13939
+    },
+    {
+      "epoch": 3.7006504712597903,
+      "grad_norm": 0.47314279083053973,
+      "learning_rate": 1.6146932181266962e-06,
+      "loss": 0.5436,
+      "step": 13940
+    },
+    {
+      "epoch": 3.700915969733174,
+      "grad_norm": 0.5313183201914511,
+      "learning_rate": 1.6143667226933718e-06,
+      "loss": 0.5339,
+      "step": 13941
+    },
+    {
+      "epoch": 3.701181468206558,
+      "grad_norm": 0.47921918196469343,
+      "learning_rate": 1.6140402445320646e-06,
+      "loss": 0.512,
+      "step": 13942
+    },
+    {
+      "epoch": 3.7014469666799417,
+      "grad_norm": 0.46459005172820533,
+      "learning_rate": 1.6137137836491417e-06,
+      "loss": 0.54,
+      "step": 13943
+    },
+    {
+      "epoch": 3.7017124651533253,
+      "grad_norm": 0.47263746969667975,
+      "learning_rate": 1.613387340050971e-06,
+      "loss": 0.5305,
+      "step": 13944
+    },
+    {
+      "epoch": 3.7019779636267094,
+      "grad_norm": 0.47009001472941886,
+      "learning_rate": 1.613060913743917e-06,
+      "loss": 0.5388,
+      "step": 13945
+    },
+    {
+      "epoch": 3.702243462100093,
+      "grad_norm": 0.47743816368702907,
+      "learning_rate": 1.6127345047343467e-06,
+      "loss": 0.5085,
+      "step": 13946
+    },
+    {
+      "epoch": 3.7025089605734767,
+      "grad_norm": 0.47882869565384395,
+      "learning_rate": 1.6124081130286262e-06,
+      "loss": 0.5483,
+      "step": 13947
+    },
+    {
+      "epoch": 3.7027744590468608,
+      "grad_norm": 0.4846219203354417,
+      "learning_rate": 1.6120817386331205e-06,
+      "loss": 0.5357,
+      "step": 13948
+    },
+    {
+      "epoch": 3.7030399575202444,
+      "grad_norm": 0.4611814051502773,
+      "learning_rate": 1.6117553815541952e-06,
+      "loss": 0.5603,
+      "step": 13949
+    },
+    {
+      "epoch": 3.703305455993628,
+      "grad_norm": 0.4741688231407391,
+      "learning_rate": 1.6114290417982156e-06,
+      "loss": 0.5532,
+      "step": 13950
+    },
+    {
+      "epoch": 3.7035709544670117,
+      "grad_norm": 0.4624028702997076,
+      "learning_rate": 1.6111027193715446e-06,
+      "loss": 0.5571,
+      "step": 13951
+    },
+    {
+      "epoch": 3.7038364529403953,
+      "grad_norm": 0.4792280960525569,
+      "learning_rate": 1.6107764142805465e-06,
+      "loss": 0.5495,
+      "step": 13952
+    },
+    {
+      "epoch": 3.7041019514137794,
+      "grad_norm": 0.46827108651071253,
+      "learning_rate": 1.6104501265315856e-06,
+      "loss": 0.526,
+      "step": 13953
+    },
+    {
+      "epoch": 3.704367449887163,
+      "grad_norm": 0.4513240760024143,
+      "learning_rate": 1.6101238561310254e-06,
+      "loss": 0.5258,
+      "step": 13954
+    },
+    {
+      "epoch": 3.7046329483605467,
+      "grad_norm": 0.4817332644741952,
+      "learning_rate": 1.609797603085228e-06,
+      "loss": 0.5643,
+      "step": 13955
+    },
+    {
+      "epoch": 3.7048984468339308,
+      "grad_norm": 0.4706657004875287,
+      "learning_rate": 1.6094713674005587e-06,
+      "loss": 0.5345,
+      "step": 13956
+    },
+    {
+      "epoch": 3.7051639453073144,
+      "grad_norm": 0.47702912757364696,
+      "learning_rate": 1.609145149083377e-06,
+      "loss": 0.5401,
+      "step": 13957
+    },
+    {
+      "epoch": 3.705429443780698,
+      "grad_norm": 0.45719426570333266,
+      "learning_rate": 1.6088189481400462e-06,
+      "loss": 0.5641,
+      "step": 13958
+    },
+    {
+      "epoch": 3.705694942254082,
+      "grad_norm": 0.4717978058671165,
+      "learning_rate": 1.608492764576928e-06,
+      "loss": 0.5848,
+      "step": 13959
+    },
+    {
+      "epoch": 3.7059604407274658,
+      "grad_norm": 0.4558463698394017,
+      "learning_rate": 1.6081665984003837e-06,
+      "loss": 0.5389,
+      "step": 13960
+    },
+    {
+      "epoch": 3.7062259392008494,
+      "grad_norm": 0.45039616958114526,
+      "learning_rate": 1.6078404496167738e-06,
+      "loss": 0.5638,
+      "step": 13961
+    },
+    {
+      "epoch": 3.7064914376742335,
+      "grad_norm": 0.4726891741244302,
+      "learning_rate": 1.6075143182324605e-06,
+      "loss": 0.5647,
+      "step": 13962
+    },
+    {
+      "epoch": 3.706756936147617,
+      "grad_norm": 0.45377895141054436,
+      "learning_rate": 1.6071882042538027e-06,
+      "loss": 0.5183,
+      "step": 13963
+    },
+    {
+      "epoch": 3.707022434621001,
+      "grad_norm": 0.46712960183505264,
+      "learning_rate": 1.6068621076871605e-06,
+      "loss": 0.5696,
+      "step": 13964
+    },
+    {
+      "epoch": 3.707287933094385,
+      "grad_norm": 0.4827404860993805,
+      "learning_rate": 1.6065360285388942e-06,
+      "loss": 0.5841,
+      "step": 13965
+    },
+    {
+      "epoch": 3.7075534315677685,
+      "grad_norm": 0.4908811403818706,
+      "learning_rate": 1.606209966815363e-06,
+      "loss": 0.5721,
+      "step": 13966
+    },
+    {
+      "epoch": 3.707818930041152,
+      "grad_norm": 0.4788514392622545,
+      "learning_rate": 1.6058839225229267e-06,
+      "loss": 0.515,
+      "step": 13967
+    },
+    {
+      "epoch": 3.7080844285145362,
+      "grad_norm": 0.445737067469096,
+      "learning_rate": 1.6055578956679422e-06,
+      "loss": 0.5304,
+      "step": 13968
+    },
+    {
+      "epoch": 3.70834992698792,
+      "grad_norm": 0.45283892177050783,
+      "learning_rate": 1.6052318862567687e-06,
+      "loss": 0.5325,
+      "step": 13969
+    },
+    {
+      "epoch": 3.7086154254613035,
+      "grad_norm": 0.46370054133355104,
+      "learning_rate": 1.604905894295764e-06,
+      "loss": 0.5729,
+      "step": 13970
+    },
+    {
+      "epoch": 3.7088809239346876,
+      "grad_norm": 0.4555017536285113,
+      "learning_rate": 1.6045799197912859e-06,
+      "loss": 0.5112,
+      "step": 13971
+    },
+    {
+      "epoch": 3.7091464224080712,
+      "grad_norm": 0.45157386613356704,
+      "learning_rate": 1.6042539627496922e-06,
+      "loss": 0.5247,
+      "step": 13972
+    },
+    {
+      "epoch": 3.709411920881455,
+      "grad_norm": 0.4798367542277795,
+      "learning_rate": 1.6039280231773396e-06,
+      "loss": 0.5661,
+      "step": 13973
+    },
+    {
+      "epoch": 3.709677419354839,
+      "grad_norm": 0.4689371554252817,
+      "learning_rate": 1.6036021010805838e-06,
+      "loss": 0.5376,
+      "step": 13974
+    },
+    {
+      "epoch": 3.7099429178282226,
+      "grad_norm": 0.46109600438568477,
+      "learning_rate": 1.603276196465782e-06,
+      "loss": 0.5063,
+      "step": 13975
+    },
+    {
+      "epoch": 3.7102084163016062,
+      "grad_norm": 0.4686636423379695,
+      "learning_rate": 1.60295030933929e-06,
+      "loss": 0.5404,
+      "step": 13976
+    },
+    {
+      "epoch": 3.71047391477499,
+      "grad_norm": 0.45903305570379066,
+      "learning_rate": 1.602624439707463e-06,
+      "loss": 0.5235,
+      "step": 13977
+    },
+    {
+      "epoch": 3.710739413248374,
+      "grad_norm": 0.4503795922045036,
+      "learning_rate": 1.602298587576657e-06,
+      "loss": 0.5551,
+      "step": 13978
+    },
+    {
+      "epoch": 3.7110049117217576,
+      "grad_norm": 0.4683078318820563,
+      "learning_rate": 1.6019727529532275e-06,
+      "loss": 0.5351,
+      "step": 13979
+    },
+    {
+      "epoch": 3.7112704101951413,
+      "grad_norm": 0.47236374610325926,
+      "learning_rate": 1.601646935843526e-06,
+      "loss": 0.529,
+      "step": 13980
+    },
+    {
+      "epoch": 3.711535908668525,
+      "grad_norm": 0.45045476106818827,
+      "learning_rate": 1.60132113625391e-06,
+      "loss": 0.5361,
+      "step": 13981
+    },
+    {
+      "epoch": 3.711801407141909,
+      "grad_norm": 0.4818288768367668,
+      "learning_rate": 1.6009953541907316e-06,
+      "loss": 0.5337,
+      "step": 13982
+    },
+    {
+      "epoch": 3.7120669056152926,
+      "grad_norm": 0.4592896302201238,
+      "learning_rate": 1.6006695896603453e-06,
+      "loss": 0.5509,
+      "step": 13983
+    },
+    {
+      "epoch": 3.7123324040886763,
+      "grad_norm": 0.4578212135807717,
+      "learning_rate": 1.6003438426691037e-06,
+      "loss": 0.5238,
+      "step": 13984
+    },
+    {
+      "epoch": 3.7125979025620603,
+      "grad_norm": 0.4521174056743562,
+      "learning_rate": 1.600018113223361e-06,
+      "loss": 0.5155,
+      "step": 13985
+    },
+    {
+      "epoch": 3.712863401035444,
+      "grad_norm": 0.4596232729534214,
+      "learning_rate": 1.5996924013294678e-06,
+      "loss": 0.5349,
+      "step": 13986
+    },
+    {
+      "epoch": 3.7131288995088276,
+      "grad_norm": 0.4810597563683559,
+      "learning_rate": 1.5993667069937772e-06,
+      "loss": 0.5497,
+      "step": 13987
+    },
+    {
+      "epoch": 3.7133943979822117,
+      "grad_norm": 0.45672500074153605,
+      "learning_rate": 1.5990410302226405e-06,
+      "loss": 0.5394,
+      "step": 13988
+    },
+    {
+      "epoch": 3.7136598964555954,
+      "grad_norm": 0.48144456966661453,
+      "learning_rate": 1.5987153710224102e-06,
+      "loss": 0.5259,
+      "step": 13989
+    },
+    {
+      "epoch": 3.713925394928979,
+      "grad_norm": 0.4708938282933909,
+      "learning_rate": 1.5983897293994363e-06,
+      "loss": 0.5657,
+      "step": 13990
+    },
+    {
+      "epoch": 3.714190893402363,
+      "grad_norm": 0.4630481578459902,
+      "learning_rate": 1.5980641053600709e-06,
+      "loss": 0.5235,
+      "step": 13991
+    },
+    {
+      "epoch": 3.7144563918757467,
+      "grad_norm": 0.4662343235015243,
+      "learning_rate": 1.5977384989106634e-06,
+      "loss": 0.5381,
+      "step": 13992
+    },
+    {
+      "epoch": 3.7147218903491304,
+      "grad_norm": 0.4815318169614546,
+      "learning_rate": 1.5974129100575641e-06,
+      "loss": 0.5529,
+      "step": 13993
+    },
+    {
+      "epoch": 3.7149873888225144,
+      "grad_norm": 0.47144939258861923,
+      "learning_rate": 1.5970873388071232e-06,
+      "loss": 0.5261,
+      "step": 13994
+    },
+    {
+      "epoch": 3.715252887295898,
+      "grad_norm": 0.4754920297639651,
+      "learning_rate": 1.5967617851656897e-06,
+      "loss": 0.5585,
+      "step": 13995
+    },
+    {
+      "epoch": 3.7155183857692817,
+      "grad_norm": 0.4708984872778865,
+      "learning_rate": 1.596436249139614e-06,
+      "loss": 0.5071,
+      "step": 13996
+    },
+    {
+      "epoch": 3.715783884242666,
+      "grad_norm": 0.4775123442750793,
+      "learning_rate": 1.5961107307352428e-06,
+      "loss": 0.4857,
+      "step": 13997
+    },
+    {
+      "epoch": 3.7160493827160495,
+      "grad_norm": 0.47555086418049525,
+      "learning_rate": 1.5957852299589255e-06,
+      "loss": 0.5529,
+      "step": 13998
+    },
+    {
+      "epoch": 3.716314881189433,
+      "grad_norm": 0.4664510654874107,
+      "learning_rate": 1.5954597468170098e-06,
+      "loss": 0.5266,
+      "step": 13999
+    },
+    {
+      "epoch": 3.716580379662817,
+      "grad_norm": 0.4794408476235134,
+      "learning_rate": 1.595134281315844e-06,
+      "loss": 0.5527,
+      "step": 14000
+    },
+    {
+      "epoch": 3.716845878136201,
+      "grad_norm": 0.47434262143827643,
+      "learning_rate": 1.5948088334617745e-06,
+      "loss": 0.5611,
+      "step": 14001
+    },
+    {
+      "epoch": 3.7171113766095845,
+      "grad_norm": 0.4697606501569821,
+      "learning_rate": 1.5944834032611506e-06,
+      "loss": 0.5039,
+      "step": 14002
+    },
+    {
+      "epoch": 3.7173768750829685,
+      "grad_norm": 0.438691579646571,
+      "learning_rate": 1.5941579907203166e-06,
+      "loss": 0.5582,
+      "step": 14003
+    },
+    {
+      "epoch": 3.717642373556352,
+      "grad_norm": 0.47460392170986415,
+      "learning_rate": 1.5938325958456197e-06,
+      "loss": 0.5634,
+      "step": 14004
+    },
+    {
+      "epoch": 3.717907872029736,
+      "grad_norm": 0.45733436648954034,
+      "learning_rate": 1.5935072186434056e-06,
+      "loss": 0.5567,
+      "step": 14005
+    },
+    {
+      "epoch": 3.7181733705031195,
+      "grad_norm": 0.4653273471991448,
+      "learning_rate": 1.5931818591200203e-06,
+      "loss": 0.5398,
+      "step": 14006
+    },
+    {
+      "epoch": 3.718438868976503,
+      "grad_norm": 0.4624898457928785,
+      "learning_rate": 1.592856517281809e-06,
+      "loss": 0.5296,
+      "step": 14007
+    },
+    {
+      "epoch": 3.718704367449887,
+      "grad_norm": 0.47841349143065914,
+      "learning_rate": 1.5925311931351172e-06,
+      "loss": 0.542,
+      "step": 14008
+    },
+    {
+      "epoch": 3.718969865923271,
+      "grad_norm": 0.46830228555091596,
+      "learning_rate": 1.592205886686289e-06,
+      "loss": 0.5441,
+      "step": 14009
+    },
+    {
+      "epoch": 3.7192353643966545,
+      "grad_norm": 0.46408189201490874,
+      "learning_rate": 1.591880597941668e-06,
+      "loss": 0.523,
+      "step": 14010
+    },
+    {
+      "epoch": 3.7195008628700386,
+      "grad_norm": 0.4568198058342598,
+      "learning_rate": 1.5915553269075989e-06,
+      "loss": 0.5391,
+      "step": 14011
+    },
+    {
+      "epoch": 3.719766361343422,
+      "grad_norm": 0.47465780068653185,
+      "learning_rate": 1.5912300735904252e-06,
+      "loss": 0.5498,
+      "step": 14012
+    },
+    {
+      "epoch": 3.720031859816806,
+      "grad_norm": 0.464038828101832,
+      "learning_rate": 1.5909048379964904e-06,
+      "loss": 0.5524,
+      "step": 14013
+    },
+    {
+      "epoch": 3.72029735829019,
+      "grad_norm": 0.4713679855204704,
+      "learning_rate": 1.5905796201321378e-06,
+      "loss": 0.547,
+      "step": 14014
+    },
+    {
+      "epoch": 3.7205628567635736,
+      "grad_norm": 0.46960477873354917,
+      "learning_rate": 1.5902544200037084e-06,
+      "loss": 0.5594,
+      "step": 14015
+    },
+    {
+      "epoch": 3.720828355236957,
+      "grad_norm": 0.4630010464039946,
+      "learning_rate": 1.5899292376175451e-06,
+      "loss": 0.547,
+      "step": 14016
+    },
+    {
+      "epoch": 3.7210938537103413,
+      "grad_norm": 0.46599825242366705,
+      "learning_rate": 1.58960407297999e-06,
+      "loss": 0.5158,
+      "step": 14017
+    },
+    {
+      "epoch": 3.721359352183725,
+      "grad_norm": 0.464129095162835,
+      "learning_rate": 1.5892789260973846e-06,
+      "loss": 0.5323,
+      "step": 14018
+    },
+    {
+      "epoch": 3.7216248506571086,
+      "grad_norm": 0.45242721019960425,
+      "learning_rate": 1.5889537969760704e-06,
+      "loss": 0.5267,
+      "step": 14019
+    },
+    {
+      "epoch": 3.7218903491304927,
+      "grad_norm": 0.46224502955299124,
+      "learning_rate": 1.5886286856223876e-06,
+      "loss": 0.5948,
+      "step": 14020
+    },
+    {
+      "epoch": 3.7221558476038763,
+      "grad_norm": 0.47642409515405704,
+      "learning_rate": 1.5883035920426765e-06,
+      "loss": 0.537,
+      "step": 14021
+    },
+    {
+      "epoch": 3.72242134607726,
+      "grad_norm": 0.48979447676449406,
+      "learning_rate": 1.5879785162432777e-06,
+      "loss": 0.5458,
+      "step": 14022
+    },
+    {
+      "epoch": 3.722686844550644,
+      "grad_norm": 0.4716302142955136,
+      "learning_rate": 1.5876534582305306e-06,
+      "loss": 0.5297,
+      "step": 14023
+    },
+    {
+      "epoch": 3.7229523430240277,
+      "grad_norm": 0.45977962044863896,
+      "learning_rate": 1.587328418010775e-06,
+      "loss": 0.5173,
+      "step": 14024
+    },
+    {
+      "epoch": 3.7232178414974113,
+      "grad_norm": 0.4757333385130725,
+      "learning_rate": 1.5870033955903508e-06,
+      "loss": 0.5558,
+      "step": 14025
+    },
+    {
+      "epoch": 3.7234833399707954,
+      "grad_norm": 0.4647279888470987,
+      "learning_rate": 1.5866783909755945e-06,
+      "loss": 0.5472,
+      "step": 14026
+    },
+    {
+      "epoch": 3.723748838444179,
+      "grad_norm": 0.4595301093408153,
+      "learning_rate": 1.5863534041728463e-06,
+      "loss": 0.5745,
+      "step": 14027
+    },
+    {
+      "epoch": 3.7240143369175627,
+      "grad_norm": 0.4557547095475187,
+      "learning_rate": 1.5860284351884433e-06,
+      "loss": 0.5694,
+      "step": 14028
+    },
+    {
+      "epoch": 3.7242798353909468,
+      "grad_norm": 0.46598219163401927,
+      "learning_rate": 1.5857034840287238e-06,
+      "loss": 0.5221,
+      "step": 14029
+    },
+    {
+      "epoch": 3.7245453338643304,
+      "grad_norm": 0.46745037414065777,
+      "learning_rate": 1.5853785507000253e-06,
+      "loss": 0.5507,
+      "step": 14030
+    },
+    {
+      "epoch": 3.724810832337714,
+      "grad_norm": 0.4643586203810378,
+      "learning_rate": 1.585053635208685e-06,
+      "loss": 0.5454,
+      "step": 14031
+    },
+    {
+      "epoch": 3.7250763308110977,
+      "grad_norm": 0.4490529417547004,
+      "learning_rate": 1.5847287375610382e-06,
+      "loss": 0.5336,
+      "step": 14032
+    },
+    {
+      "epoch": 3.7253418292844818,
+      "grad_norm": 0.46334427494015706,
+      "learning_rate": 1.5844038577634218e-06,
+      "loss": 0.561,
+      "step": 14033
+    },
+    {
+      "epoch": 3.7256073277578654,
+      "grad_norm": 0.4773383910244735,
+      "learning_rate": 1.584078995822172e-06,
+      "loss": 0.591,
+      "step": 14034
+    },
+    {
+      "epoch": 3.725872826231249,
+      "grad_norm": 0.4521197587887517,
+      "learning_rate": 1.5837541517436244e-06,
+      "loss": 0.5427,
+      "step": 14035
+    },
+    {
+      "epoch": 3.7261383247046327,
+      "grad_norm": 0.47851141726519236,
+      "learning_rate": 1.5834293255341143e-06,
+      "loss": 0.5332,
+      "step": 14036
+    },
+    {
+      "epoch": 3.7264038231780168,
+      "grad_norm": 0.45555362619218864,
+      "learning_rate": 1.583104517199977e-06,
+      "loss": 0.5523,
+      "step": 14037
+    },
+    {
+      "epoch": 3.7266693216514004,
+      "grad_norm": 0.46394069286110046,
+      "learning_rate": 1.5827797267475458e-06,
+      "loss": 0.5507,
+      "step": 14038
+    },
+    {
+      "epoch": 3.726934820124784,
+      "grad_norm": 0.46670052645269156,
+      "learning_rate": 1.582454954183156e-06,
+      "loss": 0.5386,
+      "step": 14039
+    },
+    {
+      "epoch": 3.727200318598168,
+      "grad_norm": 0.4602634805367968,
+      "learning_rate": 1.582130199513141e-06,
+      "loss": 0.4983,
+      "step": 14040
+    },
+    {
+      "epoch": 3.727465817071552,
+      "grad_norm": 0.45986926124912514,
+      "learning_rate": 1.5818054627438344e-06,
+      "loss": 0.56,
+      "step": 14041
+    },
+    {
+      "epoch": 3.7277313155449354,
+      "grad_norm": 0.4807354143039927,
+      "learning_rate": 1.5814807438815704e-06,
+      "loss": 0.5327,
+      "step": 14042
+    },
+    {
+      "epoch": 3.7279968140183195,
+      "grad_norm": 0.44852251511983077,
+      "learning_rate": 1.58115604293268e-06,
+      "loss": 0.545,
+      "step": 14043
+    },
+    {
+      "epoch": 3.728262312491703,
+      "grad_norm": 0.45572005634133306,
+      "learning_rate": 1.5808313599034963e-06,
+      "loss": 0.5121,
+      "step": 14044
+    },
+    {
+      "epoch": 3.728527810965087,
+      "grad_norm": 0.4561595856380745,
+      "learning_rate": 1.5805066948003518e-06,
+      "loss": 0.5592,
+      "step": 14045
+    },
+    {
+      "epoch": 3.728793309438471,
+      "grad_norm": 0.46463527285217193,
+      "learning_rate": 1.580182047629577e-06,
+      "loss": 0.5478,
+      "step": 14046
+    },
+    {
+      "epoch": 3.7290588079118545,
+      "grad_norm": 0.46442425642510066,
+      "learning_rate": 1.5798574183975052e-06,
+      "loss": 0.5474,
+      "step": 14047
+    },
+    {
+      "epoch": 3.729324306385238,
+      "grad_norm": 0.458550221409625,
+      "learning_rate": 1.5795328071104677e-06,
+      "loss": 0.5377,
+      "step": 14048
+    },
+    {
+      "epoch": 3.7295898048586222,
+      "grad_norm": 0.47089280908586767,
+      "learning_rate": 1.5792082137747932e-06,
+      "loss": 0.5466,
+      "step": 14049
+    },
+    {
+      "epoch": 3.729855303332006,
+      "grad_norm": 0.4665835127020925,
+      "learning_rate": 1.578883638396813e-06,
+      "loss": 0.5381,
+      "step": 14050
+    },
+    {
+      "epoch": 3.7301208018053895,
+      "grad_norm": 0.47175018564770754,
+      "learning_rate": 1.5785590809828572e-06,
+      "loss": 0.5707,
+      "step": 14051
+    },
+    {
+      "epoch": 3.7303863002787736,
+      "grad_norm": 0.48065741563771125,
+      "learning_rate": 1.5782345415392553e-06,
+      "loss": 0.6159,
+      "step": 14052
+    },
+    {
+      "epoch": 3.7306517987521572,
+      "grad_norm": 0.4671378933997786,
+      "learning_rate": 1.5779100200723365e-06,
+      "loss": 0.5237,
+      "step": 14053
+    },
+    {
+      "epoch": 3.730917297225541,
+      "grad_norm": 0.46555770384453865,
+      "learning_rate": 1.5775855165884307e-06,
+      "loss": 0.5615,
+      "step": 14054
+    },
+    {
+      "epoch": 3.731182795698925,
+      "grad_norm": 0.4705426754631287,
+      "learning_rate": 1.5772610310938651e-06,
+      "loss": 0.5757,
+      "step": 14055
+    },
+    {
+      "epoch": 3.7314482941723086,
+      "grad_norm": 0.48111052999150883,
+      "learning_rate": 1.5769365635949691e-06,
+      "loss": 0.5716,
+      "step": 14056
+    },
+    {
+      "epoch": 3.7317137926456923,
+      "grad_norm": 0.4783846596911704,
+      "learning_rate": 1.5766121140980697e-06,
+      "loss": 0.5615,
+      "step": 14057
+    },
+    {
+      "epoch": 3.7319792911190763,
+      "grad_norm": 0.45921066186769616,
+      "learning_rate": 1.5762876826094953e-06,
+      "loss": 0.5436,
+      "step": 14058
+    },
+    {
+      "epoch": 3.73224478959246,
+      "grad_norm": 0.4707470663063092,
+      "learning_rate": 1.5759632691355726e-06,
+      "loss": 0.5522,
+      "step": 14059
+    },
+    {
+      "epoch": 3.7325102880658436,
+      "grad_norm": 0.46844028418797196,
+      "learning_rate": 1.5756388736826294e-06,
+      "loss": 0.5465,
+      "step": 14060
+    },
+    {
+      "epoch": 3.7327757865392273,
+      "grad_norm": 0.4761099712401972,
+      "learning_rate": 1.5753144962569905e-06,
+      "loss": 0.5535,
+      "step": 14061
+    },
+    {
+      "epoch": 3.7330412850126113,
+      "grad_norm": 0.46316910287909924,
+      "learning_rate": 1.5749901368649828e-06,
+      "loss": 0.5265,
+      "step": 14062
+    },
+    {
+      "epoch": 3.733306783485995,
+      "grad_norm": 0.4668734526674542,
+      "learning_rate": 1.5746657955129325e-06,
+      "loss": 0.5283,
+      "step": 14063
+    },
+    {
+      "epoch": 3.7335722819593786,
+      "grad_norm": 0.4556384463968969,
+      "learning_rate": 1.5743414722071646e-06,
+      "loss": 0.5342,
+      "step": 14064
+    },
+    {
+      "epoch": 3.7338377804327623,
+      "grad_norm": 0.4632358335100543,
+      "learning_rate": 1.574017166954005e-06,
+      "loss": 0.5638,
+      "step": 14065
+    },
+    {
+      "epoch": 3.7341032789061464,
+      "grad_norm": 0.4545067266750769,
+      "learning_rate": 1.5736928797597773e-06,
+      "loss": 0.575,
+      "step": 14066
+    },
+    {
+      "epoch": 3.73436877737953,
+      "grad_norm": 0.46235884499032825,
+      "learning_rate": 1.5733686106308063e-06,
+      "loss": 0.4791,
+      "step": 14067
+    },
+    {
+      "epoch": 3.7346342758529136,
+      "grad_norm": 0.45747485624648193,
+      "learning_rate": 1.5730443595734162e-06,
+      "loss": 0.5536,
+      "step": 14068
+    },
+    {
+      "epoch": 3.7348997743262977,
+      "grad_norm": 0.4788272081449488,
+      "learning_rate": 1.572720126593931e-06,
+      "loss": 0.5459,
+      "step": 14069
+    },
+    {
+      "epoch": 3.7351652727996814,
+      "grad_norm": 0.4606531577542309,
+      "learning_rate": 1.5723959116986733e-06,
+      "loss": 0.5177,
+      "step": 14070
+    },
+    {
+      "epoch": 3.735430771273065,
+      "grad_norm": 0.44896799254195546,
+      "learning_rate": 1.5720717148939673e-06,
+      "loss": 0.5459,
+      "step": 14071
+    },
+    {
+      "epoch": 3.735696269746449,
+      "grad_norm": 0.47194539036748584,
+      "learning_rate": 1.5717475361861342e-06,
+      "loss": 0.5309,
+      "step": 14072
+    },
+    {
+      "epoch": 3.7359617682198327,
+      "grad_norm": 0.4666282782255182,
+      "learning_rate": 1.5714233755814969e-06,
+      "loss": 0.5458,
+      "step": 14073
+    },
+    {
+      "epoch": 3.7362272666932164,
+      "grad_norm": 0.4514419452370618,
+      "learning_rate": 1.5710992330863772e-06,
+      "loss": 0.5094,
+      "step": 14074
+    },
+    {
+      "epoch": 3.7364927651666004,
+      "grad_norm": 0.47802713437696404,
+      "learning_rate": 1.5707751087070972e-06,
+      "loss": 0.5643,
+      "step": 14075
+    },
+    {
+      "epoch": 3.736758263639984,
+      "grad_norm": 0.4621559787055226,
+      "learning_rate": 1.5704510024499775e-06,
+      "loss": 0.5368,
+      "step": 14076
+    },
+    {
+      "epoch": 3.7370237621133677,
+      "grad_norm": 0.46975279722094376,
+      "learning_rate": 1.57012691432134e-06,
+      "loss": 0.5349,
+      "step": 14077
+    },
+    {
+      "epoch": 3.737289260586752,
+      "grad_norm": 0.462153972407686,
+      "learning_rate": 1.5698028443275035e-06,
+      "loss": 0.5185,
+      "step": 14078
+    },
+    {
+      "epoch": 3.7375547590601355,
+      "grad_norm": 0.45544824401053713,
+      "learning_rate": 1.569478792474789e-06,
+      "loss": 0.5101,
+      "step": 14079
+    },
+    {
+      "epoch": 3.737820257533519,
+      "grad_norm": 0.4614501022541727,
+      "learning_rate": 1.5691547587695167e-06,
+      "loss": 0.5149,
+      "step": 14080
+    },
+    {
+      "epoch": 3.738085756006903,
+      "grad_norm": 0.43899634509229857,
+      "learning_rate": 1.568830743218005e-06,
+      "loss": 0.5423,
+      "step": 14081
+    },
+    {
+      "epoch": 3.738351254480287,
+      "grad_norm": 0.466974910594938,
+      "learning_rate": 1.5685067458265742e-06,
+      "loss": 0.5586,
+      "step": 14082
+    },
+    {
+      "epoch": 3.7386167529536705,
+      "grad_norm": 0.48080311004045906,
+      "learning_rate": 1.5681827666015431e-06,
+      "loss": 0.5323,
+      "step": 14083
+    },
+    {
+      "epoch": 3.7388822514270545,
+      "grad_norm": 0.4610910508519491,
+      "learning_rate": 1.5678588055492289e-06,
+      "loss": 0.5819,
+      "step": 14084
+    },
+    {
+      "epoch": 3.739147749900438,
+      "grad_norm": 0.46625495393892213,
+      "learning_rate": 1.5675348626759501e-06,
+      "loss": 0.5376,
+      "step": 14085
+    },
+    {
+      "epoch": 3.739413248373822,
+      "grad_norm": 0.4692032275781226,
+      "learning_rate": 1.567210937988025e-06,
+      "loss": 0.5418,
+      "step": 14086
+    },
+    {
+      "epoch": 3.7396787468472055,
+      "grad_norm": 0.4626499037264671,
+      "learning_rate": 1.56688703149177e-06,
+      "loss": 0.4998,
+      "step": 14087
+    },
+    {
+      "epoch": 3.7399442453205896,
+      "grad_norm": 0.473183185442196,
+      "learning_rate": 1.5665631431935036e-06,
+      "loss": 0.5507,
+      "step": 14088
+    },
+    {
+      "epoch": 3.740209743793973,
+      "grad_norm": 0.472429269896874,
+      "learning_rate": 1.5662392730995408e-06,
+      "loss": 0.4906,
+      "step": 14089
+    },
+    {
+      "epoch": 3.740475242267357,
+      "grad_norm": 0.46444205147050116,
+      "learning_rate": 1.565915421216198e-06,
+      "loss": 0.5397,
+      "step": 14090
+    },
+    {
+      "epoch": 3.7407407407407405,
+      "grad_norm": 0.4724857395032314,
+      "learning_rate": 1.5655915875497915e-06,
+      "loss": 0.5586,
+      "step": 14091
+    },
+    {
+      "epoch": 3.7410062392141246,
+      "grad_norm": 0.4594644289338305,
+      "learning_rate": 1.5652677721066364e-06,
+      "loss": 0.5826,
+      "step": 14092
+    },
+    {
+      "epoch": 3.741271737687508,
+      "grad_norm": 0.4565426786648952,
+      "learning_rate": 1.564943974893049e-06,
+      "loss": 0.5655,
+      "step": 14093
+    },
+    {
+      "epoch": 3.741537236160892,
+      "grad_norm": 0.4613102449586906,
+      "learning_rate": 1.5646201959153446e-06,
+      "loss": 0.5567,
+      "step": 14094
+    },
+    {
+      "epoch": 3.741802734634276,
+      "grad_norm": 0.4656469888339646,
+      "learning_rate": 1.5642964351798353e-06,
+      "loss": 0.5233,
+      "step": 14095
+    },
+    {
+      "epoch": 3.7420682331076596,
+      "grad_norm": 0.4605834463327008,
+      "learning_rate": 1.5639726926928368e-06,
+      "loss": 0.5554,
+      "step": 14096
+    },
+    {
+      "epoch": 3.742333731581043,
+      "grad_norm": 0.46272121324459853,
+      "learning_rate": 1.5636489684606624e-06,
+      "loss": 0.5233,
+      "step": 14097
+    },
+    {
+      "epoch": 3.7425992300544273,
+      "grad_norm": 0.47381366440702893,
+      "learning_rate": 1.5633252624896256e-06,
+      "loss": 0.559,
+      "step": 14098
+    },
+    {
+      "epoch": 3.742864728527811,
+      "grad_norm": 0.49302717878618785,
+      "learning_rate": 1.5630015747860397e-06,
+      "loss": 0.5547,
+      "step": 14099
+    },
+    {
+      "epoch": 3.7431302270011946,
+      "grad_norm": 0.4692119237585427,
+      "learning_rate": 1.5626779053562177e-06,
+      "loss": 0.5499,
+      "step": 14100
+    },
+    {
+      "epoch": 3.7433957254745787,
+      "grad_norm": 0.4635284235067521,
+      "learning_rate": 1.5623542542064712e-06,
+      "loss": 0.5576,
+      "step": 14101
+    },
+    {
+      "epoch": 3.7436612239479623,
+      "grad_norm": 0.47322606476546286,
+      "learning_rate": 1.5620306213431121e-06,
+      "loss": 0.5679,
+      "step": 14102
+    },
+    {
+      "epoch": 3.743926722421346,
+      "grad_norm": 0.4593292619626133,
+      "learning_rate": 1.5617070067724528e-06,
+      "loss": 0.5554,
+      "step": 14103
+    },
+    {
+      "epoch": 3.74419222089473,
+      "grad_norm": 0.4570012072944531,
+      "learning_rate": 1.5613834105008042e-06,
+      "loss": 0.5316,
+      "step": 14104
+    },
+    {
+      "epoch": 3.7444577193681137,
+      "grad_norm": 0.4836065398740782,
+      "learning_rate": 1.5610598325344769e-06,
+      "loss": 0.54,
+      "step": 14105
+    },
+    {
+      "epoch": 3.7447232178414973,
+      "grad_norm": 0.45267301603514387,
+      "learning_rate": 1.560736272879783e-06,
+      "loss": 0.5344,
+      "step": 14106
+    },
+    {
+      "epoch": 3.7449887163148814,
+      "grad_norm": 0.46849438334900084,
+      "learning_rate": 1.5604127315430305e-06,
+      "loss": 0.5504,
+      "step": 14107
+    },
+    {
+      "epoch": 3.745254214788265,
+      "grad_norm": 0.4598448626115179,
+      "learning_rate": 1.5600892085305303e-06,
+      "loss": 0.5826,
+      "step": 14108
+    },
+    {
+      "epoch": 3.7455197132616487,
+      "grad_norm": 0.47265533526469505,
+      "learning_rate": 1.5597657038485916e-06,
+      "loss": 0.567,
+      "step": 14109
+    },
+    {
+      "epoch": 3.7457852117350328,
+      "grad_norm": 0.4606289040756109,
+      "learning_rate": 1.559442217503524e-06,
+      "loss": 0.5088,
+      "step": 14110
+    },
+    {
+      "epoch": 3.7460507102084164,
+      "grad_norm": 0.45830198382405063,
+      "learning_rate": 1.559118749501636e-06,
+      "loss": 0.5757,
+      "step": 14111
+    },
+    {
+      "epoch": 3.7463162086818,
+      "grad_norm": 0.47692293144030207,
+      "learning_rate": 1.5587952998492367e-06,
+      "loss": 0.5799,
+      "step": 14112
+    },
+    {
+      "epoch": 3.746581707155184,
+      "grad_norm": 0.46547077227637007,
+      "learning_rate": 1.5584718685526327e-06,
+      "loss": 0.5538,
+      "step": 14113
+    },
+    {
+      "epoch": 3.7468472056285678,
+      "grad_norm": 0.4755419270336829,
+      "learning_rate": 1.5581484556181325e-06,
+      "loss": 0.5356,
+      "step": 14114
+    },
+    {
+      "epoch": 3.7471127041019514,
+      "grad_norm": 0.47574451654945993,
+      "learning_rate": 1.5578250610520437e-06,
+      "loss": 0.5348,
+      "step": 14115
+    },
+    {
+      "epoch": 3.747378202575335,
+      "grad_norm": 0.46424667495492333,
+      "learning_rate": 1.557501684860673e-06,
+      "loss": 0.5316,
+      "step": 14116
+    },
+    {
+      "epoch": 3.747643701048719,
+      "grad_norm": 0.4711814640324193,
+      "learning_rate": 1.5571783270503277e-06,
+      "loss": 0.5273,
+      "step": 14117
+    },
+    {
+      "epoch": 3.7479091995221028,
+      "grad_norm": 0.44395035674952754,
+      "learning_rate": 1.5568549876273126e-06,
+      "loss": 0.5322,
+      "step": 14118
+    },
+    {
+      "epoch": 3.7481746979954864,
+      "grad_norm": 0.4692596407197698,
+      "learning_rate": 1.5565316665979347e-06,
+      "loss": 0.5172,
+      "step": 14119
+    },
+    {
+      "epoch": 3.74844019646887,
+      "grad_norm": 0.46053713594499845,
+      "learning_rate": 1.5562083639684994e-06,
+      "loss": 0.5385,
+      "step": 14120
+    },
+    {
+      "epoch": 3.748705694942254,
+      "grad_norm": 0.46262925446653674,
+      "learning_rate": 1.5558850797453117e-06,
+      "loss": 0.5616,
+      "step": 14121
+    },
+    {
+      "epoch": 3.748971193415638,
+      "grad_norm": 0.4699826910517659,
+      "learning_rate": 1.5555618139346766e-06,
+      "loss": 0.5239,
+      "step": 14122
+    },
+    {
+      "epoch": 3.7492366918890214,
+      "grad_norm": 0.4422348623783503,
+      "learning_rate": 1.5552385665428993e-06,
+      "loss": 0.5381,
+      "step": 14123
+    },
+    {
+      "epoch": 3.7495021903624055,
+      "grad_norm": 0.4753881555299518,
+      "learning_rate": 1.554915337576282e-06,
+      "loss": 0.5618,
+      "step": 14124
+    },
+    {
+      "epoch": 3.749767688835789,
+      "grad_norm": 0.4593523001586029,
+      "learning_rate": 1.5545921270411303e-06,
+      "loss": 0.5576,
+      "step": 14125
+    },
+    {
+      "epoch": 3.750033187309173,
+      "grad_norm": 0.4441459213475431,
+      "learning_rate": 1.5542689349437462e-06,
+      "loss": 0.5307,
+      "step": 14126
+    },
+    {
+      "epoch": 3.750298685782557,
+      "grad_norm": 0.4688868136805969,
+      "learning_rate": 1.5539457612904336e-06,
+      "loss": 0.5405,
+      "step": 14127
+    },
+    {
+      "epoch": 3.7505641842559405,
+      "grad_norm": 0.45672135263786906,
+      "learning_rate": 1.5536226060874948e-06,
+      "loss": 0.5368,
+      "step": 14128
+    },
+    {
+      "epoch": 3.750829682729324,
+      "grad_norm": 0.44788237743165515,
+      "learning_rate": 1.5532994693412335e-06,
+      "loss": 0.5419,
+      "step": 14129
+    },
+    {
+      "epoch": 3.7510951812027082,
+      "grad_norm": 0.48443930831327164,
+      "learning_rate": 1.5529763510579492e-06,
+      "loss": 0.5385,
+      "step": 14130
+    },
+    {
+      "epoch": 3.751360679676092,
+      "grad_norm": 0.4491891747531123,
+      "learning_rate": 1.5526532512439452e-06,
+      "loss": 0.5346,
+      "step": 14131
+    },
+    {
+      "epoch": 3.7516261781494755,
+      "grad_norm": 0.4722058708203047,
+      "learning_rate": 1.552330169905522e-06,
+      "loss": 0.5445,
+      "step": 14132
+    },
+    {
+      "epoch": 3.7518916766228596,
+      "grad_norm": 0.45254181154495215,
+      "learning_rate": 1.5520071070489811e-06,
+      "loss": 0.5603,
+      "step": 14133
+    },
+    {
+      "epoch": 3.7521571750962432,
+      "grad_norm": 0.4542584920689708,
+      "learning_rate": 1.5516840626806226e-06,
+      "loss": 0.5314,
+      "step": 14134
+    },
+    {
+      "epoch": 3.752422673569627,
+      "grad_norm": 0.4656317828637685,
+      "learning_rate": 1.5513610368067478e-06,
+      "loss": 0.5688,
+      "step": 14135
+    },
+    {
+      "epoch": 3.752688172043011,
+      "grad_norm": 0.47768548199775884,
+      "learning_rate": 1.551038029433654e-06,
+      "loss": 0.5624,
+      "step": 14136
+    },
+    {
+      "epoch": 3.7529536705163946,
+      "grad_norm": 0.4649648373231188,
+      "learning_rate": 1.5507150405676427e-06,
+      "loss": 0.519,
+      "step": 14137
+    },
+    {
+      "epoch": 3.7532191689897783,
+      "grad_norm": 0.46361827345434214,
+      "learning_rate": 1.5503920702150115e-06,
+      "loss": 0.5524,
+      "step": 14138
+    },
+    {
+      "epoch": 3.7534846674631623,
+      "grad_norm": 0.45863664716041846,
+      "learning_rate": 1.5500691183820608e-06,
+      "loss": 0.5646,
+      "step": 14139
+    },
+    {
+      "epoch": 3.753750165936546,
+      "grad_norm": 0.46648489603091786,
+      "learning_rate": 1.549746185075089e-06,
+      "loss": 0.5347,
+      "step": 14140
+    },
+    {
+      "epoch": 3.7540156644099296,
+      "grad_norm": 0.46623498511021055,
+      "learning_rate": 1.549423270300392e-06,
+      "loss": 0.5569,
+      "step": 14141
+    },
+    {
+      "epoch": 3.7542811628833133,
+      "grad_norm": 0.4563435477628405,
+      "learning_rate": 1.5491003740642688e-06,
+      "loss": 0.536,
+      "step": 14142
+    },
+    {
+      "epoch": 3.7545466613566973,
+      "grad_norm": 0.46742232530729466,
+      "learning_rate": 1.5487774963730166e-06,
+      "loss": 0.5531,
+      "step": 14143
+    },
+    {
+      "epoch": 3.754812159830081,
+      "grad_norm": 0.4687147083898061,
+      "learning_rate": 1.5484546372329317e-06,
+      "loss": 0.5503,
+      "step": 14144
+    },
+    {
+      "epoch": 3.7550776583034646,
+      "grad_norm": 0.4611813122899733,
+      "learning_rate": 1.5481317966503119e-06,
+      "loss": 0.5319,
+      "step": 14145
+    },
+    {
+      "epoch": 3.7553431567768483,
+      "grad_norm": 0.45901687651977074,
+      "learning_rate": 1.5478089746314529e-06,
+      "loss": 0.5676,
+      "step": 14146
+    },
+    {
+      "epoch": 3.7556086552502324,
+      "grad_norm": 0.47304115708617367,
+      "learning_rate": 1.5474861711826494e-06,
+      "loss": 0.5584,
+      "step": 14147
+    },
+    {
+      "epoch": 3.755874153723616,
+      "grad_norm": 0.4492820497942645,
+      "learning_rate": 1.5471633863101982e-06,
+      "loss": 0.5333,
+      "step": 14148
+    },
+    {
+      "epoch": 3.7561396521969996,
+      "grad_norm": 0.4716639682607868,
+      "learning_rate": 1.5468406200203933e-06,
+      "loss": 0.5565,
+      "step": 14149
+    },
+    {
+      "epoch": 3.7564051506703837,
+      "grad_norm": 0.4585976654031539,
+      "learning_rate": 1.5465178723195306e-06,
+      "loss": 0.5614,
+      "step": 14150
+    },
+    {
+      "epoch": 3.7566706491437674,
+      "grad_norm": 0.46439300898643665,
+      "learning_rate": 1.5461951432139037e-06,
+      "loss": 0.5398,
+      "step": 14151
+    },
+    {
+      "epoch": 3.756936147617151,
+      "grad_norm": 0.4691646075945546,
+      "learning_rate": 1.545872432709808e-06,
+      "loss": 0.5467,
+      "step": 14152
+    },
+    {
+      "epoch": 3.757201646090535,
+      "grad_norm": 0.44993226764445804,
+      "learning_rate": 1.5455497408135348e-06,
+      "loss": 0.5089,
+      "step": 14153
+    },
+    {
+      "epoch": 3.7574671445639187,
+      "grad_norm": 0.46664183547218313,
+      "learning_rate": 1.5452270675313786e-06,
+      "loss": 0.5428,
+      "step": 14154
+    },
+    {
+      "epoch": 3.7577326430373024,
+      "grad_norm": 0.46316653873858293,
+      "learning_rate": 1.5449044128696321e-06,
+      "loss": 0.5479,
+      "step": 14155
+    },
+    {
+      "epoch": 3.7579981415106865,
+      "grad_norm": 0.4549140776418115,
+      "learning_rate": 1.5445817768345883e-06,
+      "loss": 0.5423,
+      "step": 14156
+    },
+    {
+      "epoch": 3.75826363998407,
+      "grad_norm": 0.47043199261157453,
+      "learning_rate": 1.544259159432539e-06,
+      "loss": 0.5401,
+      "step": 14157
+    },
+    {
+      "epoch": 3.7585291384574537,
+      "grad_norm": 0.4626591132457924,
+      "learning_rate": 1.5439365606697767e-06,
+      "loss": 0.5584,
+      "step": 14158
+    },
+    {
+      "epoch": 3.758794636930838,
+      "grad_norm": 0.47208030798589096,
+      "learning_rate": 1.5436139805525918e-06,
+      "loss": 0.5439,
+      "step": 14159
+    },
+    {
+      "epoch": 3.7590601354042215,
+      "grad_norm": 0.4614881069001578,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.5792,
+      "step": 14160
+    },
+    {
+      "epoch": 3.759325633877605,
+      "grad_norm": 0.4641126748050402,
+      "learning_rate": 1.5429688762801198e-06,
+      "loss": 0.5078,
+      "step": 14161
+    },
+    {
+      "epoch": 3.759591132350989,
+      "grad_norm": 0.47313052671372646,
+      "learning_rate": 1.5426463521374136e-06,
+      "loss": 0.5379,
+      "step": 14162
+    },
+    {
+      "epoch": 3.759856630824373,
+      "grad_norm": 0.46400997890945683,
+      "learning_rate": 1.5423238466654483e-06,
+      "loss": 0.545,
+      "step": 14163
+    },
+    {
+      "epoch": 3.7601221292977565,
+      "grad_norm": 0.47074210383803505,
+      "learning_rate": 1.5420013598705122e-06,
+      "loss": 0.5482,
+      "step": 14164
+    },
+    {
+      "epoch": 3.7603876277711406,
+      "grad_norm": 0.4649314673898206,
+      "learning_rate": 1.5416788917588951e-06,
+      "loss": 0.5504,
+      "step": 14165
+    },
+    {
+      "epoch": 3.760653126244524,
+      "grad_norm": 0.4711338780606955,
+      "learning_rate": 1.541356442336886e-06,
+      "loss": 0.5329,
+      "step": 14166
+    },
+    {
+      "epoch": 3.760918624717908,
+      "grad_norm": 0.45930638280344877,
+      "learning_rate": 1.5410340116107736e-06,
+      "loss": 0.5123,
+      "step": 14167
+    },
+    {
+      "epoch": 3.761184123191292,
+      "grad_norm": 0.4383401983935441,
+      "learning_rate": 1.5407115995868458e-06,
+      "loss": 0.5594,
+      "step": 14168
+    },
+    {
+      "epoch": 3.7614496216646756,
+      "grad_norm": 0.46732753969976504,
+      "learning_rate": 1.5403892062713916e-06,
+      "loss": 0.5638,
+      "step": 14169
+    },
+    {
+      "epoch": 3.761715120138059,
+      "grad_norm": 0.4683283767847534,
+      "learning_rate": 1.5400668316706963e-06,
+      "loss": 0.5176,
+      "step": 14170
+    },
+    {
+      "epoch": 3.761980618611443,
+      "grad_norm": 0.47326179173337873,
+      "learning_rate": 1.5397444757910482e-06,
+      "loss": 0.518,
+      "step": 14171
+    },
+    {
+      "epoch": 3.762246117084827,
+      "grad_norm": 0.46933222046778494,
+      "learning_rate": 1.5394221386387339e-06,
+      "loss": 0.5051,
+      "step": 14172
+    },
+    {
+      "epoch": 3.7625116155582106,
+      "grad_norm": 0.5041568566719479,
+      "learning_rate": 1.5390998202200397e-06,
+      "loss": 0.5654,
+      "step": 14173
+    },
+    {
+      "epoch": 3.762777114031594,
+      "grad_norm": 0.455140207780505,
+      "learning_rate": 1.538777520541252e-06,
+      "loss": 0.5405,
+      "step": 14174
+    },
+    {
+      "epoch": 3.763042612504978,
+      "grad_norm": 0.4606004466048283,
+      "learning_rate": 1.5384552396086564e-06,
+      "loss": 0.5607,
+      "step": 14175
+    },
+    {
+      "epoch": 3.763308110978362,
+      "grad_norm": 0.4574048347688159,
+      "learning_rate": 1.5381329774285375e-06,
+      "loss": 0.5365,
+      "step": 14176
+    },
+    {
+      "epoch": 3.7635736094517456,
+      "grad_norm": 0.46712493614989464,
+      "learning_rate": 1.5378107340071804e-06,
+      "loss": 0.5343,
+      "step": 14177
+    },
+    {
+      "epoch": 3.763839107925129,
+      "grad_norm": 0.47263627279934206,
+      "learning_rate": 1.53748850935087e-06,
+      "loss": 0.5325,
+      "step": 14178
+    },
+    {
+      "epoch": 3.7641046063985133,
+      "grad_norm": 0.46263170172867796,
+      "learning_rate": 1.53716630346589e-06,
+      "loss": 0.5259,
+      "step": 14179
+    },
+    {
+      "epoch": 3.764370104871897,
+      "grad_norm": 0.46947621929996913,
+      "learning_rate": 1.5368441163585251e-06,
+      "loss": 0.5297,
+      "step": 14180
+    },
+    {
+      "epoch": 3.7646356033452806,
+      "grad_norm": 0.4754590092064743,
+      "learning_rate": 1.5365219480350587e-06,
+      "loss": 0.5163,
+      "step": 14181
+    },
+    {
+      "epoch": 3.7649011018186647,
+      "grad_norm": 0.45825797284702485,
+      "learning_rate": 1.5361997985017723e-06,
+      "loss": 0.5415,
+      "step": 14182
+    },
+    {
+      "epoch": 3.7651666002920483,
+      "grad_norm": 0.476110609699576,
+      "learning_rate": 1.5358776677649495e-06,
+      "loss": 0.5501,
+      "step": 14183
+    },
+    {
+      "epoch": 3.765432098765432,
+      "grad_norm": 0.4696895496839855,
+      "learning_rate": 1.5355555558308726e-06,
+      "loss": 0.5521,
+      "step": 14184
+    },
+    {
+      "epoch": 3.765697597238816,
+      "grad_norm": 0.47632564194477317,
+      "learning_rate": 1.5352334627058243e-06,
+      "loss": 0.5679,
+      "step": 14185
+    },
+    {
+      "epoch": 3.7659630957121997,
+      "grad_norm": 0.46612802089194644,
+      "learning_rate": 1.5349113883960865e-06,
+      "loss": 0.5735,
+      "step": 14186
+    },
+    {
+      "epoch": 3.7662285941855833,
+      "grad_norm": 0.4490705760323217,
+      "learning_rate": 1.5345893329079386e-06,
+      "loss": 0.537,
+      "step": 14187
+    },
+    {
+      "epoch": 3.7664940926589674,
+      "grad_norm": 0.47574549701559665,
+      "learning_rate": 1.5342672962476623e-06,
+      "loss": 0.5508,
+      "step": 14188
+    },
+    {
+      "epoch": 3.766759591132351,
+      "grad_norm": 0.45127674119658506,
+      "learning_rate": 1.5339452784215384e-06,
+      "loss": 0.5109,
+      "step": 14189
+    },
+    {
+      "epoch": 3.7670250896057347,
+      "grad_norm": 0.4739759112568161,
+      "learning_rate": 1.5336232794358468e-06,
+      "loss": 0.5072,
+      "step": 14190
+    },
+    {
+      "epoch": 3.7672905880791188,
+      "grad_norm": 0.471017294318739,
+      "learning_rate": 1.5333012992968672e-06,
+      "loss": 0.536,
+      "step": 14191
+    },
+    {
+      "epoch": 3.7675560865525024,
+      "grad_norm": 0.47965777557191386,
+      "learning_rate": 1.53297933801088e-06,
+      "loss": 0.5463,
+      "step": 14192
+    },
+    {
+      "epoch": 3.767821585025886,
+      "grad_norm": 0.4730400259524348,
+      "learning_rate": 1.5326573955841625e-06,
+      "loss": 0.5848,
+      "step": 14193
+    },
+    {
+      "epoch": 3.76808708349927,
+      "grad_norm": 0.47093039427389355,
+      "learning_rate": 1.5323354720229945e-06,
+      "loss": 0.5341,
+      "step": 14194
+    },
+    {
+      "epoch": 3.7683525819726538,
+      "grad_norm": 0.46697219926777084,
+      "learning_rate": 1.532013567333654e-06,
+      "loss": 0.5548,
+      "step": 14195
+    },
+    {
+      "epoch": 3.7686180804460374,
+      "grad_norm": 0.46259618436138317,
+      "learning_rate": 1.5316916815224189e-06,
+      "loss": 0.5255,
+      "step": 14196
+    },
+    {
+      "epoch": 3.768883578919421,
+      "grad_norm": 0.4512855912663067,
+      "learning_rate": 1.531369814595567e-06,
+      "loss": 0.5383,
+      "step": 14197
+    },
+    {
+      "epoch": 3.769149077392805,
+      "grad_norm": 0.479558634468816,
+      "learning_rate": 1.531047966559376e-06,
+      "loss": 0.5159,
+      "step": 14198
+    },
+    {
+      "epoch": 3.7694145758661888,
+      "grad_norm": 0.4549524141193485,
+      "learning_rate": 1.5307261374201216e-06,
+      "loss": 0.5831,
+      "step": 14199
+    },
+    {
+      "epoch": 3.7696800743395724,
+      "grad_norm": 0.4701022812441394,
+      "learning_rate": 1.53040432718408e-06,
+      "loss": 0.5454,
+      "step": 14200
+    },
+    {
+      "epoch": 3.769945572812956,
+      "grad_norm": 0.46919677287507405,
+      "learning_rate": 1.5300825358575286e-06,
+      "loss": 0.519,
+      "step": 14201
+    },
+    {
+      "epoch": 3.77021107128634,
+      "grad_norm": 0.45721756471685293,
+      "learning_rate": 1.5297607634467424e-06,
+      "loss": 0.5183,
+      "step": 14202
+    },
+    {
+      "epoch": 3.770476569759724,
+      "grad_norm": 0.45926350050062215,
+      "learning_rate": 1.5294390099579967e-06,
+      "loss": 0.5485,
+      "step": 14203
+    },
+    {
+      "epoch": 3.7707420682331074,
+      "grad_norm": 0.4678462504936208,
+      "learning_rate": 1.5291172753975672e-06,
+      "loss": 0.4939,
+      "step": 14204
+    },
+    {
+      "epoch": 3.7710075667064915,
+      "grad_norm": 0.448983046629253,
+      "learning_rate": 1.5287955597717274e-06,
+      "loss": 0.5188,
+      "step": 14205
+    },
+    {
+      "epoch": 3.771273065179875,
+      "grad_norm": 0.4659836263418916,
+      "learning_rate": 1.5284738630867525e-06,
+      "loss": 0.5434,
+      "step": 14206
+    },
+    {
+      "epoch": 3.771538563653259,
+      "grad_norm": 0.4765383989395894,
+      "learning_rate": 1.5281521853489154e-06,
+      "loss": 0.5726,
+      "step": 14207
+    },
+    {
+      "epoch": 3.771804062126643,
+      "grad_norm": 0.4544630757379573,
+      "learning_rate": 1.5278305265644905e-06,
+      "loss": 0.5239,
+      "step": 14208
+    },
+    {
+      "epoch": 3.7720695606000265,
+      "grad_norm": 0.47659364852386443,
+      "learning_rate": 1.5275088867397508e-06,
+      "loss": 0.5611,
+      "step": 14209
+    },
+    {
+      "epoch": 3.77233505907341,
+      "grad_norm": 0.4650810045165806,
+      "learning_rate": 1.527187265880969e-06,
+      "loss": 0.5862,
+      "step": 14210
+    },
+    {
+      "epoch": 3.7726005575467942,
+      "grad_norm": 0.46150674179319245,
+      "learning_rate": 1.526865663994417e-06,
+      "loss": 0.5277,
+      "step": 14211
+    },
+    {
+      "epoch": 3.772866056020178,
+      "grad_norm": 0.478079467895871,
+      "learning_rate": 1.526544081086367e-06,
+      "loss": 0.5698,
+      "step": 14212
+    },
+    {
+      "epoch": 3.7731315544935615,
+      "grad_norm": 0.45511240665043146,
+      "learning_rate": 1.5262225171630913e-06,
+      "loss": 0.5101,
+      "step": 14213
+    },
+    {
+      "epoch": 3.7733970529669456,
+      "grad_norm": 0.46177868951407436,
+      "learning_rate": 1.5259009722308605e-06,
+      "loss": 0.5638,
+      "step": 14214
+    },
+    {
+      "epoch": 3.7736625514403292,
+      "grad_norm": 0.46710375414889743,
+      "learning_rate": 1.5255794462959466e-06,
+      "loss": 0.5647,
+      "step": 14215
+    },
+    {
+      "epoch": 3.773928049913713,
+      "grad_norm": 0.47510377273736837,
+      "learning_rate": 1.5252579393646183e-06,
+      "loss": 0.5521,
+      "step": 14216
+    },
+    {
+      "epoch": 3.774193548387097,
+      "grad_norm": 0.4732149253766579,
+      "learning_rate": 1.5249364514431469e-06,
+      "loss": 0.5121,
+      "step": 14217
+    },
+    {
+      "epoch": 3.7744590468604806,
+      "grad_norm": 0.44246816831853997,
+      "learning_rate": 1.524614982537802e-06,
+      "loss": 0.5275,
+      "step": 14218
+    },
+    {
+      "epoch": 3.7747245453338643,
+      "grad_norm": 0.4864985393189019,
+      "learning_rate": 1.524293532654853e-06,
+      "loss": 0.5385,
+      "step": 14219
+    },
+    {
+      "epoch": 3.7749900438072483,
+      "grad_norm": 0.45719647036488437,
+      "learning_rate": 1.5239721018005692e-06,
+      "loss": 0.5657,
+      "step": 14220
+    },
+    {
+      "epoch": 3.775255542280632,
+      "grad_norm": 0.46570500618607347,
+      "learning_rate": 1.5236506899812198e-06,
+      "loss": 0.5028,
+      "step": 14221
+    },
+    {
+      "epoch": 3.7755210407540156,
+      "grad_norm": 0.4785090858172073,
+      "learning_rate": 1.5233292972030717e-06,
+      "loss": 0.6005,
+      "step": 14222
+    },
+    {
+      "epoch": 3.7757865392273997,
+      "grad_norm": 0.48188933767877945,
+      "learning_rate": 1.5230079234723938e-06,
+      "loss": 0.5435,
+      "step": 14223
+    },
+    {
+      "epoch": 3.7760520377007833,
+      "grad_norm": 0.48281571172493926,
+      "learning_rate": 1.5226865687954533e-06,
+      "loss": 0.5475,
+      "step": 14224
+    },
+    {
+      "epoch": 3.776317536174167,
+      "grad_norm": 0.4565732545235267,
+      "learning_rate": 1.5223652331785177e-06,
+      "loss": 0.5531,
+      "step": 14225
+    },
+    {
+      "epoch": 3.7765830346475506,
+      "grad_norm": 0.45469069433779674,
+      "learning_rate": 1.522043916627854e-06,
+      "loss": 0.5635,
+      "step": 14226
+    },
+    {
+      "epoch": 3.7768485331209347,
+      "grad_norm": 0.4673266070009971,
+      "learning_rate": 1.521722619149729e-06,
+      "loss": 0.554,
+      "step": 14227
+    },
+    {
+      "epoch": 3.7771140315943184,
+      "grad_norm": 0.4659146112129088,
+      "learning_rate": 1.521401340750407e-06,
+      "loss": 0.5406,
+      "step": 14228
+    },
+    {
+      "epoch": 3.777379530067702,
+      "grad_norm": 0.47131386373090756,
+      "learning_rate": 1.5210800814361548e-06,
+      "loss": 0.5622,
+      "step": 14229
+    },
+    {
+      "epoch": 3.7776450285410856,
+      "grad_norm": 0.4698670457443824,
+      "learning_rate": 1.520758841213238e-06,
+      "loss": 0.5275,
+      "step": 14230
+    },
+    {
+      "epoch": 3.7779105270144697,
+      "grad_norm": 0.461861436313446,
+      "learning_rate": 1.5204376200879214e-06,
+      "loss": 0.5373,
+      "step": 14231
+    },
+    {
+      "epoch": 3.7781760254878534,
+      "grad_norm": 0.4809959313284668,
+      "learning_rate": 1.52011641806647e-06,
+      "loss": 0.5679,
+      "step": 14232
+    },
+    {
+      "epoch": 3.778441523961237,
+      "grad_norm": 0.4645756254962713,
+      "learning_rate": 1.519795235155148e-06,
+      "loss": 0.5285,
+      "step": 14233
+    },
+    {
+      "epoch": 3.778707022434621,
+      "grad_norm": 0.4595264123217648,
+      "learning_rate": 1.5194740713602184e-06,
+      "loss": 0.5598,
+      "step": 14234
+    },
+    {
+      "epoch": 3.7789725209080047,
+      "grad_norm": 0.47212455714652074,
+      "learning_rate": 1.5191529266879448e-06,
+      "loss": 0.5596,
+      "step": 14235
+    },
+    {
+      "epoch": 3.7792380193813884,
+      "grad_norm": 0.48195499968682903,
+      "learning_rate": 1.5188318011445907e-06,
+      "loss": 0.5185,
+      "step": 14236
+    },
+    {
+      "epoch": 3.7795035178547725,
+      "grad_norm": 0.4594241349240103,
+      "learning_rate": 1.5185106947364191e-06,
+      "loss": 0.573,
+      "step": 14237
+    },
+    {
+      "epoch": 3.779769016328156,
+      "grad_norm": 0.4922396420602789,
+      "learning_rate": 1.5181896074696923e-06,
+      "loss": 0.5693,
+      "step": 14238
+    },
+    {
+      "epoch": 3.7800345148015397,
+      "grad_norm": 0.4773497931957487,
+      "learning_rate": 1.5178685393506714e-06,
+      "loss": 0.5439,
+      "step": 14239
+    },
+    {
+      "epoch": 3.780300013274924,
+      "grad_norm": 0.46411952210108837,
+      "learning_rate": 1.5175474903856191e-06,
+      "loss": 0.5416,
+      "step": 14240
+    },
+    {
+      "epoch": 3.7805655117483075,
+      "grad_norm": 0.46802209086818164,
+      "learning_rate": 1.5172264605807958e-06,
+      "loss": 0.5345,
+      "step": 14241
+    },
+    {
+      "epoch": 3.780831010221691,
+      "grad_norm": 0.4620134979726957,
+      "learning_rate": 1.5169054499424629e-06,
+      "loss": 0.5326,
+      "step": 14242
+    },
+    {
+      "epoch": 3.781096508695075,
+      "grad_norm": 0.45728762087002545,
+      "learning_rate": 1.5165844584768808e-06,
+      "loss": 0.5618,
+      "step": 14243
+    },
+    {
+      "epoch": 3.781362007168459,
+      "grad_norm": 0.4631897804995381,
+      "learning_rate": 1.5162634861903106e-06,
+      "loss": 0.5408,
+      "step": 14244
+    },
+    {
+      "epoch": 3.7816275056418425,
+      "grad_norm": 0.45919151059566493,
+      "learning_rate": 1.5159425330890098e-06,
+      "loss": 0.5417,
+      "step": 14245
+    },
+    {
+      "epoch": 3.7818930041152266,
+      "grad_norm": 0.4502948640783779,
+      "learning_rate": 1.5156215991792394e-06,
+      "loss": 0.5272,
+      "step": 14246
+    },
+    {
+      "epoch": 3.78215850258861,
+      "grad_norm": 0.4614817440882876,
+      "learning_rate": 1.5153006844672573e-06,
+      "loss": 0.5382,
+      "step": 14247
+    },
+    {
+      "epoch": 3.782424001061994,
+      "grad_norm": 0.4734694506670941,
+      "learning_rate": 1.5149797889593237e-06,
+      "loss": 0.5528,
+      "step": 14248
+    },
+    {
+      "epoch": 3.782689499535378,
+      "grad_norm": 0.4774470773048459,
+      "learning_rate": 1.5146589126616953e-06,
+      "loss": 0.5631,
+      "step": 14249
+    },
+    {
+      "epoch": 3.7829549980087616,
+      "grad_norm": 0.46553325165739917,
+      "learning_rate": 1.5143380555806315e-06,
+      "loss": 0.5799,
+      "step": 14250
+    },
+    {
+      "epoch": 3.783220496482145,
+      "grad_norm": 0.4704822328945651,
+      "learning_rate": 1.5140172177223884e-06,
+      "loss": 0.5686,
+      "step": 14251
+    },
+    {
+      "epoch": 3.783485994955529,
+      "grad_norm": 0.48043657987176347,
+      "learning_rate": 1.5136963990932236e-06,
+      "loss": 0.5594,
+      "step": 14252
+    },
+    {
+      "epoch": 3.783751493428913,
+      "grad_norm": 0.46528168845864426,
+      "learning_rate": 1.5133755996993937e-06,
+      "loss": 0.5512,
+      "step": 14253
+    },
+    {
+      "epoch": 3.7840169919022966,
+      "grad_norm": 0.4502793248338295,
+      "learning_rate": 1.5130548195471554e-06,
+      "loss": 0.5337,
+      "step": 14254
+    },
+    {
+      "epoch": 3.78428249037568,
+      "grad_norm": 0.4692016784601357,
+      "learning_rate": 1.5127340586427646e-06,
+      "loss": 0.5611,
+      "step": 14255
+    },
+    {
+      "epoch": 3.784547988849064,
+      "grad_norm": 0.4559760124667694,
+      "learning_rate": 1.5124133169924773e-06,
+      "loss": 0.5305,
+      "step": 14256
+    },
+    {
+      "epoch": 3.784813487322448,
+      "grad_norm": 0.4505917176982701,
+      "learning_rate": 1.5120925946025478e-06,
+      "loss": 0.5618,
+      "step": 14257
+    },
+    {
+      "epoch": 3.7850789857958316,
+      "grad_norm": 0.47546895568641556,
+      "learning_rate": 1.5117718914792318e-06,
+      "loss": 0.5382,
+      "step": 14258
+    },
+    {
+      "epoch": 3.785344484269215,
+      "grad_norm": 0.4639670482399162,
+      "learning_rate": 1.5114512076287835e-06,
+      "loss": 0.525,
+      "step": 14259
+    },
+    {
+      "epoch": 3.7856099827425993,
+      "grad_norm": 0.4726281398809493,
+      "learning_rate": 1.511130543057457e-06,
+      "loss": 0.5523,
+      "step": 14260
+    },
+    {
+      "epoch": 3.785875481215983,
+      "grad_norm": 0.4556047756476211,
+      "learning_rate": 1.5108098977715068e-06,
+      "loss": 0.5294,
+      "step": 14261
+    },
+    {
+      "epoch": 3.7861409796893666,
+      "grad_norm": 0.4751286302268576,
+      "learning_rate": 1.5104892717771846e-06,
+      "loss": 0.53,
+      "step": 14262
+    },
+    {
+      "epoch": 3.7864064781627507,
+      "grad_norm": 0.45137645035804386,
+      "learning_rate": 1.5101686650807446e-06,
+      "loss": 0.5219,
+      "step": 14263
+    },
+    {
+      "epoch": 3.7866719766361343,
+      "grad_norm": 0.4679210685989948,
+      "learning_rate": 1.5098480776884389e-06,
+      "loss": 0.568,
+      "step": 14264
+    },
+    {
+      "epoch": 3.786937475109518,
+      "grad_norm": 0.47379446436026784,
+      "learning_rate": 1.5095275096065203e-06,
+      "loss": 0.5511,
+      "step": 14265
+    },
+    {
+      "epoch": 3.787202973582902,
+      "grad_norm": 0.4766575379702056,
+      "learning_rate": 1.5092069608412402e-06,
+      "loss": 0.5401,
+      "step": 14266
+    },
+    {
+      "epoch": 3.7874684720562857,
+      "grad_norm": 0.45344540195607325,
+      "learning_rate": 1.5088864313988507e-06,
+      "loss": 0.5734,
+      "step": 14267
+    },
+    {
+      "epoch": 3.7877339705296693,
+      "grad_norm": 0.47757820917918764,
+      "learning_rate": 1.508565921285602e-06,
+      "loss": 0.5169,
+      "step": 14268
+    },
+    {
+      "epoch": 3.7879994690030534,
+      "grad_norm": 0.4508937528219623,
+      "learning_rate": 1.5082454305077455e-06,
+      "loss": 0.5458,
+      "step": 14269
+    },
+    {
+      "epoch": 3.788264967476437,
+      "grad_norm": 0.46597878656759767,
+      "learning_rate": 1.5079249590715309e-06,
+      "loss": 0.5492,
+      "step": 14270
+    },
+    {
+      "epoch": 3.7885304659498207,
+      "grad_norm": 0.47184756971392877,
+      "learning_rate": 1.5076045069832088e-06,
+      "loss": 0.5662,
+      "step": 14271
+    },
+    {
+      "epoch": 3.7887959644232048,
+      "grad_norm": 0.46757912090755344,
+      "learning_rate": 1.5072840742490285e-06,
+      "loss": 0.5609,
+      "step": 14272
+    },
+    {
+      "epoch": 3.7890614628965884,
+      "grad_norm": 0.4593729915670754,
+      "learning_rate": 1.5069636608752409e-06,
+      "loss": 0.5378,
+      "step": 14273
+    },
+    {
+      "epoch": 3.789326961369972,
+      "grad_norm": 0.4629796176203357,
+      "learning_rate": 1.5066432668680916e-06,
+      "loss": 0.5123,
+      "step": 14274
+    },
+    {
+      "epoch": 3.789592459843356,
+      "grad_norm": 0.4496752974954881,
+      "learning_rate": 1.5063228922338306e-06,
+      "loss": 0.5452,
+      "step": 14275
+    },
+    {
+      "epoch": 3.7898579583167398,
+      "grad_norm": 0.4660184669343393,
+      "learning_rate": 1.5060025369787065e-06,
+      "loss": 0.5097,
+      "step": 14276
+    },
+    {
+      "epoch": 3.7901234567901234,
+      "grad_norm": 0.46449246547749035,
+      "learning_rate": 1.5056822011089667e-06,
+      "loss": 0.5428,
+      "step": 14277
+    },
+    {
+      "epoch": 3.7903889552635075,
+      "grad_norm": 0.44628606804148463,
+      "learning_rate": 1.5053618846308585e-06,
+      "loss": 0.5116,
+      "step": 14278
+    },
+    {
+      "epoch": 3.790654453736891,
+      "grad_norm": 0.4713317079471035,
+      "learning_rate": 1.5050415875506296e-06,
+      "loss": 0.5328,
+      "step": 14279
+    },
+    {
+      "epoch": 3.790919952210275,
+      "grad_norm": 0.47065117455100447,
+      "learning_rate": 1.5047213098745254e-06,
+      "loss": 0.5277,
+      "step": 14280
+    },
+    {
+      "epoch": 3.7911854506836584,
+      "grad_norm": 0.469574895305161,
+      "learning_rate": 1.504401051608792e-06,
+      "loss": 0.535,
+      "step": 14281
+    },
+    {
+      "epoch": 3.7914509491570425,
+      "grad_norm": 0.47241031745461826,
+      "learning_rate": 1.5040808127596763e-06,
+      "loss": 0.5509,
+      "step": 14282
+    },
+    {
+      "epoch": 3.791716447630426,
+      "grad_norm": 0.46563778512404824,
+      "learning_rate": 1.5037605933334227e-06,
+      "loss": 0.5386,
+      "step": 14283
+    },
+    {
+      "epoch": 3.79198194610381,
+      "grad_norm": 0.46433832434161454,
+      "learning_rate": 1.5034403933362773e-06,
+      "loss": 0.5645,
+      "step": 14284
+    },
+    {
+      "epoch": 3.7922474445771934,
+      "grad_norm": 0.469463197434822,
+      "learning_rate": 1.5031202127744836e-06,
+      "loss": 0.5554,
+      "step": 14285
+    },
+    {
+      "epoch": 3.7925129430505775,
+      "grad_norm": 0.4769221326208225,
+      "learning_rate": 1.5028000516542867e-06,
+      "loss": 0.5352,
+      "step": 14286
+    },
+    {
+      "epoch": 3.792778441523961,
+      "grad_norm": 0.4546887213542327,
+      "learning_rate": 1.5024799099819304e-06,
+      "loss": 0.5251,
+      "step": 14287
+    },
+    {
+      "epoch": 3.793043939997345,
+      "grad_norm": 0.46961964453831107,
+      "learning_rate": 1.5021597877636584e-06,
+      "loss": 0.5511,
+      "step": 14288
+    },
+    {
+      "epoch": 3.793309438470729,
+      "grad_norm": 0.46794875897737165,
+      "learning_rate": 1.5018396850057133e-06,
+      "loss": 0.4994,
+      "step": 14289
+    },
+    {
+      "epoch": 3.7935749369441125,
+      "grad_norm": 0.45679025757187486,
+      "learning_rate": 1.5015196017143397e-06,
+      "loss": 0.579,
+      "step": 14290
+    },
+    {
+      "epoch": 3.793840435417496,
+      "grad_norm": 0.4625444253415987,
+      "learning_rate": 1.5011995378957772e-06,
+      "loss": 0.5379,
+      "step": 14291
+    },
+    {
+      "epoch": 3.7941059338908802,
+      "grad_norm": 0.4613853189765263,
+      "learning_rate": 1.5008794935562693e-06,
+      "loss": 0.5735,
+      "step": 14292
+    },
+    {
+      "epoch": 3.794371432364264,
+      "grad_norm": 0.47584492283496405,
+      "learning_rate": 1.5005594687020576e-06,
+      "loss": 0.5369,
+      "step": 14293
+    },
+    {
+      "epoch": 3.7946369308376475,
+      "grad_norm": 0.47583246786974676,
+      "learning_rate": 1.500239463339383e-06,
+      "loss": 0.5634,
+      "step": 14294
+    },
+    {
+      "epoch": 3.7949024293110316,
+      "grad_norm": 0.46178991889172755,
+      "learning_rate": 1.4999194774744863e-06,
+      "loss": 0.5425,
+      "step": 14295
+    },
+    {
+      "epoch": 3.7951679277844153,
+      "grad_norm": 0.47049713903604845,
+      "learning_rate": 1.4995995111136102e-06,
+      "loss": 0.5781,
+      "step": 14296
+    },
+    {
+      "epoch": 3.795433426257799,
+      "grad_norm": 0.46479861695990177,
+      "learning_rate": 1.499279564262992e-06,
+      "loss": 0.5482,
+      "step": 14297
+    },
+    {
+      "epoch": 3.795698924731183,
+      "grad_norm": 0.46861104556202055,
+      "learning_rate": 1.4989596369288723e-06,
+      "loss": 0.5763,
+      "step": 14298
+    },
+    {
+      "epoch": 3.7959644232045666,
+      "grad_norm": 0.4655584797180522,
+      "learning_rate": 1.4986397291174906e-06,
+      "loss": 0.5314,
+      "step": 14299
+    },
+    {
+      "epoch": 3.7962299216779503,
+      "grad_norm": 0.46196263811205207,
+      "learning_rate": 1.4983198408350863e-06,
+      "loss": 0.5317,
+      "step": 14300
+    },
+    {
+      "epoch": 3.7964954201513343,
+      "grad_norm": 0.47418777778015764,
+      "learning_rate": 1.4979999720878975e-06,
+      "loss": 0.5712,
+      "step": 14301
+    },
+    {
+      "epoch": 3.796760918624718,
+      "grad_norm": 0.4710991704508042,
+      "learning_rate": 1.4976801228821625e-06,
+      "loss": 0.5443,
+      "step": 14302
+    },
+    {
+      "epoch": 3.7970264170981016,
+      "grad_norm": 0.4778669367536355,
+      "learning_rate": 1.4973602932241193e-06,
+      "loss": 0.5588,
+      "step": 14303
+    },
+    {
+      "epoch": 3.7972919155714857,
+      "grad_norm": 0.46518698161276734,
+      "learning_rate": 1.4970404831200053e-06,
+      "loss": 0.5568,
+      "step": 14304
+    },
+    {
+      "epoch": 3.7975574140448694,
+      "grad_norm": 0.4664777751508425,
+      "learning_rate": 1.4967206925760573e-06,
+      "loss": 0.5429,
+      "step": 14305
+    },
+    {
+      "epoch": 3.797822912518253,
+      "grad_norm": 0.4754504041047351,
+      "learning_rate": 1.4964009215985122e-06,
+      "loss": 0.529,
+      "step": 14306
+    },
+    {
+      "epoch": 3.7980884109916366,
+      "grad_norm": 0.4626791191768184,
+      "learning_rate": 1.4960811701936076e-06,
+      "loss": 0.5474,
+      "step": 14307
+    },
+    {
+      "epoch": 3.7983539094650207,
+      "grad_norm": 0.46972224717961636,
+      "learning_rate": 1.495761438367577e-06,
+      "loss": 0.5648,
+      "step": 14308
+    },
+    {
+      "epoch": 3.7986194079384044,
+      "grad_norm": 0.459739570489443,
+      "learning_rate": 1.4954417261266573e-06,
+      "loss": 0.5488,
+      "step": 14309
+    },
+    {
+      "epoch": 3.798884906411788,
+      "grad_norm": 0.45959208686586267,
+      "learning_rate": 1.4951220334770833e-06,
+      "loss": 0.5212,
+      "step": 14310
+    },
+    {
+      "epoch": 3.7991504048851716,
+      "grad_norm": 0.461988454472999,
+      "learning_rate": 1.4948023604250902e-06,
+      "loss": 0.5728,
+      "step": 14311
+    },
+    {
+      "epoch": 3.7994159033585557,
+      "grad_norm": 0.4872759434077837,
+      "learning_rate": 1.4944827069769125e-06,
+      "loss": 0.553,
+      "step": 14312
+    },
+    {
+      "epoch": 3.7996814018319394,
+      "grad_norm": 0.46677084580224454,
+      "learning_rate": 1.494163073138784e-06,
+      "loss": 0.5814,
+      "step": 14313
+    },
+    {
+      "epoch": 3.799946900305323,
+      "grad_norm": 0.47477082042944535,
+      "learning_rate": 1.4938434589169378e-06,
+      "loss": 0.5039,
+      "step": 14314
+    },
+    {
+      "epoch": 3.800212398778707,
+      "grad_norm": 0.4519686286655203,
+      "learning_rate": 1.4935238643176079e-06,
+      "loss": 0.5228,
+      "step": 14315
+    },
+    {
+      "epoch": 3.8004778972520907,
+      "grad_norm": 0.4650749605615722,
+      "learning_rate": 1.4932042893470266e-06,
+      "loss": 0.5507,
+      "step": 14316
+    },
+    {
+      "epoch": 3.8007433957254744,
+      "grad_norm": 0.4752879449841521,
+      "learning_rate": 1.4928847340114271e-06,
+      "loss": 0.5445,
+      "step": 14317
+    },
+    {
+      "epoch": 3.8010088941988585,
+      "grad_norm": 0.46376848105078305,
+      "learning_rate": 1.492565198317041e-06,
+      "loss": 0.5319,
+      "step": 14318
+    },
+    {
+      "epoch": 3.801274392672242,
+      "grad_norm": 0.4694493607772262,
+      "learning_rate": 1.492245682270101e-06,
+      "loss": 0.5114,
+      "step": 14319
+    },
+    {
+      "epoch": 3.8015398911456257,
+      "grad_norm": 0.4798675830000061,
+      "learning_rate": 1.4919261858768363e-06,
+      "loss": 0.5209,
+      "step": 14320
+    },
+    {
+      "epoch": 3.80180538961901,
+      "grad_norm": 0.4640277711755632,
+      "learning_rate": 1.4916067091434795e-06,
+      "loss": 0.5211,
+      "step": 14321
+    },
+    {
+      "epoch": 3.8020708880923935,
+      "grad_norm": 0.4529010554991586,
+      "learning_rate": 1.4912872520762609e-06,
+      "loss": 0.5402,
+      "step": 14322
+    },
+    {
+      "epoch": 3.802336386565777,
+      "grad_norm": 0.4575938855109411,
+      "learning_rate": 1.4909678146814104e-06,
+      "loss": 0.5277,
+      "step": 14323
+    },
+    {
+      "epoch": 3.802601885039161,
+      "grad_norm": 0.46197337962599033,
+      "learning_rate": 1.4906483969651587e-06,
+      "loss": 0.5639,
+      "step": 14324
+    },
+    {
+      "epoch": 3.802867383512545,
+      "grad_norm": 0.4823898392660827,
+      "learning_rate": 1.490328998933735e-06,
+      "loss": 0.5518,
+      "step": 14325
+    },
+    {
+      "epoch": 3.8031328819859285,
+      "grad_norm": 0.4675881464292851,
+      "learning_rate": 1.4900096205933672e-06,
+      "loss": 0.508,
+      "step": 14326
+    },
+    {
+      "epoch": 3.8033983804593126,
+      "grad_norm": 0.4625001234238671,
+      "learning_rate": 1.4896902619502846e-06,
+      "loss": 0.5519,
+      "step": 14327
+    },
+    {
+      "epoch": 3.803663878932696,
+      "grad_norm": 0.46875681257116325,
+      "learning_rate": 1.4893709230107156e-06,
+      "loss": 0.5654,
+      "step": 14328
+    },
+    {
+      "epoch": 3.80392937740608,
+      "grad_norm": 0.4633042663645761,
+      "learning_rate": 1.4890516037808877e-06,
+      "loss": 0.5288,
+      "step": 14329
+    },
+    {
+      "epoch": 3.804194875879464,
+      "grad_norm": 0.48152113101093835,
+      "learning_rate": 1.4887323042670298e-06,
+      "loss": 0.5548,
+      "step": 14330
+    },
+    {
+      "epoch": 3.8044603743528476,
+      "grad_norm": 0.4852801034845774,
+      "learning_rate": 1.488413024475367e-06,
+      "loss": 0.5584,
+      "step": 14331
+    },
+    {
+      "epoch": 3.804725872826231,
+      "grad_norm": 0.4689323902391257,
+      "learning_rate": 1.4880937644121273e-06,
+      "loss": 0.5496,
+      "step": 14332
+    },
+    {
+      "epoch": 3.8049913712996153,
+      "grad_norm": 0.4608974987803858,
+      "learning_rate": 1.4877745240835368e-06,
+      "loss": 0.561,
+      "step": 14333
+    },
+    {
+      "epoch": 3.805256869772999,
+      "grad_norm": 0.4780259210240867,
+      "learning_rate": 1.487455303495821e-06,
+      "loss": 0.5248,
+      "step": 14334
+    },
+    {
+      "epoch": 3.8055223682463826,
+      "grad_norm": 0.492456924304905,
+      "learning_rate": 1.4871361026552067e-06,
+      "loss": 0.5625,
+      "step": 14335
+    },
+    {
+      "epoch": 3.805787866719766,
+      "grad_norm": 0.4809124671828694,
+      "learning_rate": 1.4868169215679188e-06,
+      "loss": 0.571,
+      "step": 14336
+    },
+    {
+      "epoch": 3.8060533651931503,
+      "grad_norm": 0.4714769115685962,
+      "learning_rate": 1.4864977602401806e-06,
+      "loss": 0.5533,
+      "step": 14337
+    },
+    {
+      "epoch": 3.806318863666534,
+      "grad_norm": 0.45694315668196317,
+      "learning_rate": 1.4861786186782176e-06,
+      "loss": 0.5458,
+      "step": 14338
+    },
+    {
+      "epoch": 3.8065843621399176,
+      "grad_norm": 0.47207225486480636,
+      "learning_rate": 1.4858594968882539e-06,
+      "loss": 0.563,
+      "step": 14339
+    },
+    {
+      "epoch": 3.806849860613301,
+      "grad_norm": 0.467871781759021,
+      "learning_rate": 1.4855403948765129e-06,
+      "loss": 0.5473,
+      "step": 14340
+    },
+    {
+      "epoch": 3.8071153590866853,
+      "grad_norm": 0.4654050096973252,
+      "learning_rate": 1.4852213126492177e-06,
+      "loss": 0.5298,
+      "step": 14341
+    },
+    {
+      "epoch": 3.807380857560069,
+      "grad_norm": 0.4660229739535231,
+      "learning_rate": 1.484902250212593e-06,
+      "loss": 0.5391,
+      "step": 14342
+    },
+    {
+      "epoch": 3.8076463560334526,
+      "grad_norm": 0.4685184569408039,
+      "learning_rate": 1.4845832075728589e-06,
+      "loss": 0.5599,
+      "step": 14343
+    },
+    {
+      "epoch": 3.8079118545068367,
+      "grad_norm": 0.4748748085920868,
+      "learning_rate": 1.4842641847362383e-06,
+      "loss": 0.5434,
+      "step": 14344
+    },
+    {
+      "epoch": 3.8081773529802203,
+      "grad_norm": 0.4719656503493941,
+      "learning_rate": 1.4839451817089528e-06,
+      "loss": 0.4853,
+      "step": 14345
+    },
+    {
+      "epoch": 3.808442851453604,
+      "grad_norm": 0.456788903507843,
+      "learning_rate": 1.4836261984972245e-06,
+      "loss": 0.5652,
+      "step": 14346
+    },
+    {
+      "epoch": 3.808708349926988,
+      "grad_norm": 0.46988639329071397,
+      "learning_rate": 1.4833072351072733e-06,
+      "loss": 0.5671,
+      "step": 14347
+    },
+    {
+      "epoch": 3.8089738484003717,
+      "grad_norm": 0.4534394757121887,
+      "learning_rate": 1.482988291545321e-06,
+      "loss": 0.5445,
+      "step": 14348
+    },
+    {
+      "epoch": 3.8092393468737553,
+      "grad_norm": 0.47365051182009255,
+      "learning_rate": 1.4826693678175869e-06,
+      "loss": 0.587,
+      "step": 14349
+    },
+    {
+      "epoch": 3.8095048453471394,
+      "grad_norm": 0.47773049958899966,
+      "learning_rate": 1.4823504639302905e-06,
+      "loss": 0.5439,
+      "step": 14350
+    },
+    {
+      "epoch": 3.809770343820523,
+      "grad_norm": 0.4679571382142268,
+      "learning_rate": 1.482031579889652e-06,
+      "loss": 0.5477,
+      "step": 14351
+    },
+    {
+      "epoch": 3.8100358422939067,
+      "grad_norm": 0.46938896335987407,
+      "learning_rate": 1.48171271570189e-06,
+      "loss": 0.5648,
+      "step": 14352
+    },
+    {
+      "epoch": 3.8103013407672908,
+      "grad_norm": 0.4779634706566664,
+      "learning_rate": 1.4813938713732235e-06,
+      "loss": 0.5096,
+      "step": 14353
+    },
+    {
+      "epoch": 3.8105668392406744,
+      "grad_norm": 0.47918547125911526,
+      "learning_rate": 1.4810750469098711e-06,
+      "loss": 0.5557,
+      "step": 14354
+    },
+    {
+      "epoch": 3.810832337714058,
+      "grad_norm": 0.47045561143239895,
+      "learning_rate": 1.4807562423180494e-06,
+      "loss": 0.5271,
+      "step": 14355
+    },
+    {
+      "epoch": 3.811097836187442,
+      "grad_norm": 0.46673636899865867,
+      "learning_rate": 1.4804374576039763e-06,
+      "loss": 0.5103,
+      "step": 14356
+    },
+    {
+      "epoch": 3.8113633346608258,
+      "grad_norm": 0.4624595970867824,
+      "learning_rate": 1.4801186927738689e-06,
+      "loss": 0.5065,
+      "step": 14357
+    },
+    {
+      "epoch": 3.8116288331342094,
+      "grad_norm": 0.46663858806130387,
+      "learning_rate": 1.4797999478339443e-06,
+      "loss": 0.553,
+      "step": 14358
+    },
+    {
+      "epoch": 3.8118943316075935,
+      "grad_norm": 0.48630159158713454,
+      "learning_rate": 1.4794812227904192e-06,
+      "loss": 0.5297,
+      "step": 14359
+    },
+    {
+      "epoch": 3.812159830080977,
+      "grad_norm": 0.4835106063040936,
+      "learning_rate": 1.4791625176495083e-06,
+      "loss": 0.5481,
+      "step": 14360
+    },
+    {
+      "epoch": 3.812425328554361,
+      "grad_norm": 0.4753636835452331,
+      "learning_rate": 1.478843832417428e-06,
+      "loss": 0.5501,
+      "step": 14361
+    },
+    {
+      "epoch": 3.812690827027745,
+      "grad_norm": 0.4651875084848534,
+      "learning_rate": 1.4785251671003927e-06,
+      "loss": 0.5615,
+      "step": 14362
+    },
+    {
+      "epoch": 3.8129563255011285,
+      "grad_norm": 0.46656368627726663,
+      "learning_rate": 1.4782065217046178e-06,
+      "loss": 0.5746,
+      "step": 14363
+    },
+    {
+      "epoch": 3.813221823974512,
+      "grad_norm": 0.47365670200672455,
+      "learning_rate": 1.4778878962363177e-06,
+      "loss": 0.542,
+      "step": 14364
+    },
+    {
+      "epoch": 3.813487322447896,
+      "grad_norm": 0.45628129341498064,
+      "learning_rate": 1.477569290701707e-06,
+      "loss": 0.5219,
+      "step": 14365
+    },
+    {
+      "epoch": 3.8137528209212794,
+      "grad_norm": 0.4669487538842543,
+      "learning_rate": 1.4772507051069972e-06,
+      "loss": 0.5857,
+      "step": 14366
+    },
+    {
+      "epoch": 3.8140183193946635,
+      "grad_norm": 0.47181475862756933,
+      "learning_rate": 1.476932139458403e-06,
+      "loss": 0.5461,
+      "step": 14367
+    },
+    {
+      "epoch": 3.814283817868047,
+      "grad_norm": 0.4630794460422562,
+      "learning_rate": 1.4766135937621372e-06,
+      "loss": 0.5225,
+      "step": 14368
+    },
+    {
+      "epoch": 3.814549316341431,
+      "grad_norm": 0.46823294179306313,
+      "learning_rate": 1.476295068024412e-06,
+      "loss": 0.5619,
+      "step": 14369
+    },
+    {
+      "epoch": 3.814814814814815,
+      "grad_norm": 0.4605088479269549,
+      "learning_rate": 1.4759765622514394e-06,
+      "loss": 0.5369,
+      "step": 14370
+    },
+    {
+      "epoch": 3.8150803132881985,
+      "grad_norm": 0.4821744351837662,
+      "learning_rate": 1.4756580764494322e-06,
+      "loss": 0.5418,
+      "step": 14371
+    },
+    {
+      "epoch": 3.815345811761582,
+      "grad_norm": 0.46880476449760783,
+      "learning_rate": 1.4753396106245993e-06,
+      "loss": 0.5569,
+      "step": 14372
+    },
+    {
+      "epoch": 3.8156113102349662,
+      "grad_norm": 0.4702449836733929,
+      "learning_rate": 1.4750211647831529e-06,
+      "loss": 0.5601,
+      "step": 14373
+    },
+    {
+      "epoch": 3.81587680870835,
+      "grad_norm": 0.465968062237587,
+      "learning_rate": 1.4747027389313032e-06,
+      "loss": 0.5523,
+      "step": 14374
+    },
+    {
+      "epoch": 3.8161423071817335,
+      "grad_norm": 0.46233518110951977,
+      "learning_rate": 1.4743843330752606e-06,
+      "loss": 0.5477,
+      "step": 14375
+    },
+    {
+      "epoch": 3.8164078056551176,
+      "grad_norm": 0.45895099695681163,
+      "learning_rate": 1.4740659472212347e-06,
+      "loss": 0.5694,
+      "step": 14376
+    },
+    {
+      "epoch": 3.8166733041285013,
+      "grad_norm": 0.4638001212321143,
+      "learning_rate": 1.4737475813754352e-06,
+      "loss": 0.4955,
+      "step": 14377
+    },
+    {
+      "epoch": 3.816938802601885,
+      "grad_norm": 0.46641427785805883,
+      "learning_rate": 1.4734292355440698e-06,
+      "loss": 0.5025,
+      "step": 14378
+    },
+    {
+      "epoch": 3.817204301075269,
+      "grad_norm": 0.48551446748419047,
+      "learning_rate": 1.473110909733348e-06,
+      "loss": 0.5509,
+      "step": 14379
+    },
+    {
+      "epoch": 3.8174697995486526,
+      "grad_norm": 0.45394307393484445,
+      "learning_rate": 1.4727926039494778e-06,
+      "loss": 0.5308,
+      "step": 14380
+    },
+    {
+      "epoch": 3.8177352980220363,
+      "grad_norm": 0.46843074556273445,
+      "learning_rate": 1.4724743181986668e-06,
+      "loss": 0.5472,
+      "step": 14381
+    },
+    {
+      "epoch": 3.8180007964954203,
+      "grad_norm": 0.44974838616594165,
+      "learning_rate": 1.4721560524871237e-06,
+      "loss": 0.5258,
+      "step": 14382
+    },
+    {
+      "epoch": 3.818266294968804,
+      "grad_norm": 0.4735508573254619,
+      "learning_rate": 1.4718378068210525e-06,
+      "loss": 0.5566,
+      "step": 14383
+    },
+    {
+      "epoch": 3.8185317934421876,
+      "grad_norm": 0.4700249550404407,
+      "learning_rate": 1.4715195812066619e-06,
+      "loss": 0.5414,
+      "step": 14384
+    },
+    {
+      "epoch": 3.8187972919155717,
+      "grad_norm": 0.4730080691362576,
+      "learning_rate": 1.4712013756501575e-06,
+      "loss": 0.5231,
+      "step": 14385
+    },
+    {
+      "epoch": 3.8190627903889554,
+      "grad_norm": 0.47169148682180734,
+      "learning_rate": 1.4708831901577453e-06,
+      "loss": 0.5193,
+      "step": 14386
+    },
+    {
+      "epoch": 3.819328288862339,
+      "grad_norm": 0.4763434968252582,
+      "learning_rate": 1.47056502473563e-06,
+      "loss": 0.5632,
+      "step": 14387
+    },
+    {
+      "epoch": 3.819593787335723,
+      "grad_norm": 0.4668472113874937,
+      "learning_rate": 1.4702468793900187e-06,
+      "loss": 0.5171,
+      "step": 14388
+    },
+    {
+      "epoch": 3.8198592858091067,
+      "grad_norm": 0.4560642203070396,
+      "learning_rate": 1.469928754127114e-06,
+      "loss": 0.5373,
+      "step": 14389
+    },
+    {
+      "epoch": 3.8201247842824904,
+      "grad_norm": 0.4661632993884505,
+      "learning_rate": 1.46961064895312e-06,
+      "loss": 0.523,
+      "step": 14390
+    },
+    {
+      "epoch": 3.820390282755874,
+      "grad_norm": 0.4498891809085212,
+      "learning_rate": 1.4692925638742416e-06,
+      "loss": 0.5621,
+      "step": 14391
+    },
+    {
+      "epoch": 3.820655781229258,
+      "grad_norm": 0.45782544834191014,
+      "learning_rate": 1.4689744988966814e-06,
+      "loss": 0.54,
+      "step": 14392
+    },
+    {
+      "epoch": 3.8209212797026417,
+      "grad_norm": 0.46138547897797916,
+      "learning_rate": 1.4686564540266432e-06,
+      "loss": 0.562,
+      "step": 14393
+    },
+    {
+      "epoch": 3.8211867781760254,
+      "grad_norm": 0.46067454050847145,
+      "learning_rate": 1.4683384292703295e-06,
+      "loss": 0.5297,
+      "step": 14394
+    },
+    {
+      "epoch": 3.821452276649409,
+      "grad_norm": 0.4772049945599968,
+      "learning_rate": 1.4680204246339425e-06,
+      "loss": 0.5597,
+      "step": 14395
+    },
+    {
+      "epoch": 3.821717775122793,
+      "grad_norm": 0.47401754810777014,
+      "learning_rate": 1.4677024401236834e-06,
+      "loss": 0.6031,
+      "step": 14396
+    },
+    {
+      "epoch": 3.8219832735961767,
+      "grad_norm": 0.4691731186206942,
+      "learning_rate": 1.4673844757457543e-06,
+      "loss": 0.527,
+      "step": 14397
+    },
+    {
+      "epoch": 3.8222487720695604,
+      "grad_norm": 0.4725447736989617,
+      "learning_rate": 1.467066531506356e-06,
+      "loss": 0.5449,
+      "step": 14398
+    },
+    {
+      "epoch": 3.8225142705429445,
+      "grad_norm": 0.4461468753699611,
+      "learning_rate": 1.4667486074116899e-06,
+      "loss": 0.5432,
+      "step": 14399
+    },
+    {
+      "epoch": 3.822779769016328,
+      "grad_norm": 0.4680426203236417,
+      "learning_rate": 1.4664307034679564e-06,
+      "loss": 0.558,
+      "step": 14400
+    },
+    {
+      "epoch": 3.8230452674897117,
+      "grad_norm": 0.4699348262991877,
+      "learning_rate": 1.466112819681354e-06,
+      "loss": 0.5149,
+      "step": 14401
+    },
+    {
+      "epoch": 3.823310765963096,
+      "grad_norm": 0.4559604435265895,
+      "learning_rate": 1.4657949560580829e-06,
+      "loss": 0.5156,
+      "step": 14402
+    },
+    {
+      "epoch": 3.8235762644364795,
+      "grad_norm": 0.45339592668133893,
+      "learning_rate": 1.465477112604342e-06,
+      "loss": 0.5373,
+      "step": 14403
+    },
+    {
+      "epoch": 3.823841762909863,
+      "grad_norm": 0.4636450337425917,
+      "learning_rate": 1.4651592893263309e-06,
+      "loss": 0.5356,
+      "step": 14404
+    },
+    {
+      "epoch": 3.824107261383247,
+      "grad_norm": 0.46785954943298463,
+      "learning_rate": 1.4648414862302476e-06,
+      "loss": 0.5522,
+      "step": 14405
+    },
+    {
+      "epoch": 3.824372759856631,
+      "grad_norm": 0.4645724093299341,
+      "learning_rate": 1.4645237033222895e-06,
+      "loss": 0.5796,
+      "step": 14406
+    },
+    {
+      "epoch": 3.8246382583300145,
+      "grad_norm": 0.4670454227207539,
+      "learning_rate": 1.4642059406086544e-06,
+      "loss": 0.5609,
+      "step": 14407
+    },
+    {
+      "epoch": 3.8249037568033986,
+      "grad_norm": 0.4716641920710068,
+      "learning_rate": 1.4638881980955394e-06,
+      "loss": 0.5897,
+      "step": 14408
+    },
+    {
+      "epoch": 3.825169255276782,
+      "grad_norm": 0.479027020079382,
+      "learning_rate": 1.4635704757891416e-06,
+      "loss": 0.592,
+      "step": 14409
+    },
+    {
+      "epoch": 3.825434753750166,
+      "grad_norm": 0.47218402186689384,
+      "learning_rate": 1.463252773695657e-06,
+      "loss": 0.576,
+      "step": 14410
+    },
+    {
+      "epoch": 3.82570025222355,
+      "grad_norm": 0.4897964541648506,
+      "learning_rate": 1.462935091821282e-06,
+      "loss": 0.5628,
+      "step": 14411
+    },
+    {
+      "epoch": 3.8259657506969336,
+      "grad_norm": 0.47220511644334695,
+      "learning_rate": 1.4626174301722119e-06,
+      "loss": 0.539,
+      "step": 14412
+    },
+    {
+      "epoch": 3.826231249170317,
+      "grad_norm": 0.4679090359651499,
+      "learning_rate": 1.4622997887546414e-06,
+      "loss": 0.5251,
+      "step": 14413
+    },
+    {
+      "epoch": 3.8264967476437013,
+      "grad_norm": 0.46316218365467776,
+      "learning_rate": 1.4619821675747658e-06,
+      "loss": 0.5381,
+      "step": 14414
+    },
+    {
+      "epoch": 3.826762246117085,
+      "grad_norm": 0.4700058145615277,
+      "learning_rate": 1.4616645666387796e-06,
+      "loss": 0.5419,
+      "step": 14415
+    },
+    {
+      "epoch": 3.8270277445904686,
+      "grad_norm": 0.48648563306576964,
+      "learning_rate": 1.4613469859528767e-06,
+      "loss": 0.5528,
+      "step": 14416
+    },
+    {
+      "epoch": 3.8272932430638527,
+      "grad_norm": 0.47850112833147723,
+      "learning_rate": 1.4610294255232516e-06,
+      "loss": 0.5497,
+      "step": 14417
+    },
+    {
+      "epoch": 3.8275587415372363,
+      "grad_norm": 0.4580334723020406,
+      "learning_rate": 1.4607118853560953e-06,
+      "loss": 0.5616,
+      "step": 14418
+    },
+    {
+      "epoch": 3.82782424001062,
+      "grad_norm": 0.48575969235574473,
+      "learning_rate": 1.460394365457602e-06,
+      "loss": 0.5387,
+      "step": 14419
+    },
+    {
+      "epoch": 3.8280897384840036,
+      "grad_norm": 0.463895146831036,
+      "learning_rate": 1.460076865833964e-06,
+      "loss": 0.5431,
+      "step": 14420
+    },
+    {
+      "epoch": 3.828355236957387,
+      "grad_norm": 0.48271634012227266,
+      "learning_rate": 1.4597593864913734e-06,
+      "loss": 0.5337,
+      "step": 14421
+    },
+    {
+      "epoch": 3.8286207354307713,
+      "grad_norm": 0.4651240594002985,
+      "learning_rate": 1.4594419274360213e-06,
+      "loss": 0.5602,
+      "step": 14422
+    },
+    {
+      "epoch": 3.828886233904155,
+      "grad_norm": 0.4602898879836279,
+      "learning_rate": 1.4591244886741002e-06,
+      "loss": 0.5161,
+      "step": 14423
+    },
+    {
+      "epoch": 3.8291517323775386,
+      "grad_norm": 0.47538170474468056,
+      "learning_rate": 1.4588070702118e-06,
+      "loss": 0.5979,
+      "step": 14424
+    },
+    {
+      "epoch": 3.8294172308509227,
+      "grad_norm": 0.4642100688108346,
+      "learning_rate": 1.45848967205531e-06,
+      "loss": 0.5545,
+      "step": 14425
+    },
+    {
+      "epoch": 3.8296827293243063,
+      "grad_norm": 0.5203533782236369,
+      "learning_rate": 1.4581722942108228e-06,
+      "loss": 0.545,
+      "step": 14426
+    },
+    {
+      "epoch": 3.82994822779769,
+      "grad_norm": 0.46207216440036447,
+      "learning_rate": 1.4578549366845257e-06,
+      "loss": 0.4999,
+      "step": 14427
+    },
+    {
+      "epoch": 3.830213726271074,
+      "grad_norm": 0.45854272060469004,
+      "learning_rate": 1.4575375994826097e-06,
+      "loss": 0.5281,
+      "step": 14428
+    },
+    {
+      "epoch": 3.8304792247444577,
+      "grad_norm": 0.45925303277682006,
+      "learning_rate": 1.4572202826112631e-06,
+      "loss": 0.5696,
+      "step": 14429
+    },
+    {
+      "epoch": 3.8307447232178413,
+      "grad_norm": 0.4638681669388345,
+      "learning_rate": 1.4569029860766728e-06,
+      "loss": 0.5766,
+      "step": 14430
+    },
+    {
+      "epoch": 3.8310102216912254,
+      "grad_norm": 0.4749060810405108,
+      "learning_rate": 1.4565857098850295e-06,
+      "loss": 0.5734,
+      "step": 14431
+    },
+    {
+      "epoch": 3.831275720164609,
+      "grad_norm": 0.4651022907169953,
+      "learning_rate": 1.4562684540425188e-06,
+      "loss": 0.5415,
+      "step": 14432
+    },
+    {
+      "epoch": 3.8315412186379927,
+      "grad_norm": 0.46227844972563475,
+      "learning_rate": 1.4559512185553292e-06,
+      "loss": 0.5603,
+      "step": 14433
+    },
+    {
+      "epoch": 3.8318067171113768,
+      "grad_norm": 0.4608024626035467,
+      "learning_rate": 1.4556340034296474e-06,
+      "loss": 0.5522,
+      "step": 14434
+    },
+    {
+      "epoch": 3.8320722155847604,
+      "grad_norm": 0.47195819865298355,
+      "learning_rate": 1.4553168086716584e-06,
+      "loss": 0.5318,
+      "step": 14435
+    },
+    {
+      "epoch": 3.832337714058144,
+      "grad_norm": 0.46974536143754225,
+      "learning_rate": 1.4549996342875507e-06,
+      "loss": 0.5412,
+      "step": 14436
+    },
+    {
+      "epoch": 3.832603212531528,
+      "grad_norm": 0.4698426666976415,
+      "learning_rate": 1.4546824802835076e-06,
+      "loss": 0.5366,
+      "step": 14437
+    },
+    {
+      "epoch": 3.8328687110049118,
+      "grad_norm": 0.4689159660350755,
+      "learning_rate": 1.4543653466657167e-06,
+      "loss": 0.535,
+      "step": 14438
+    },
+    {
+      "epoch": 3.8331342094782954,
+      "grad_norm": 0.4695055979222109,
+      "learning_rate": 1.4540482334403605e-06,
+      "loss": 0.5208,
+      "step": 14439
+    },
+    {
+      "epoch": 3.8333997079516795,
+      "grad_norm": 0.4739325010572114,
+      "learning_rate": 1.4537311406136267e-06,
+      "loss": 0.5551,
+      "step": 14440
+    },
+    {
+      "epoch": 3.833665206425063,
+      "grad_norm": 0.44968479892602603,
+      "learning_rate": 1.4534140681916955e-06,
+      "loss": 0.5396,
+      "step": 14441
+    },
+    {
+      "epoch": 3.833930704898447,
+      "grad_norm": 0.460112160600991,
+      "learning_rate": 1.4530970161807533e-06,
+      "loss": 0.5643,
+      "step": 14442
+    },
+    {
+      "epoch": 3.834196203371831,
+      "grad_norm": 0.4780980853671893,
+      "learning_rate": 1.4527799845869817e-06,
+      "loss": 0.5784,
+      "step": 14443
+    },
+    {
+      "epoch": 3.8344617018452145,
+      "grad_norm": 0.4718784366523411,
+      "learning_rate": 1.4524629734165652e-06,
+      "loss": 0.5318,
+      "step": 14444
+    },
+    {
+      "epoch": 3.834727200318598,
+      "grad_norm": 0.46814092938972407,
+      "learning_rate": 1.4521459826756847e-06,
+      "loss": 0.5792,
+      "step": 14445
+    },
+    {
+      "epoch": 3.834992698791982,
+      "grad_norm": 0.4747068394867822,
+      "learning_rate": 1.4518290123705242e-06,
+      "loss": 0.5382,
+      "step": 14446
+    },
+    {
+      "epoch": 3.835258197265366,
+      "grad_norm": 0.4610826260007812,
+      "learning_rate": 1.4515120625072642e-06,
+      "loss": 0.5262,
+      "step": 14447
+    },
+    {
+      "epoch": 3.8355236957387495,
+      "grad_norm": 0.46857610022842733,
+      "learning_rate": 1.4511951330920854e-06,
+      "loss": 0.5127,
+      "step": 14448
+    },
+    {
+      "epoch": 3.835789194212133,
+      "grad_norm": 0.4536533350890589,
+      "learning_rate": 1.45087822413117e-06,
+      "loss": 0.5144,
+      "step": 14449
+    },
+    {
+      "epoch": 3.836054692685517,
+      "grad_norm": 0.4603118458162056,
+      "learning_rate": 1.4505613356306969e-06,
+      "loss": 0.5569,
+      "step": 14450
+    },
+    {
+      "epoch": 3.836320191158901,
+      "grad_norm": 0.46594922010807804,
+      "learning_rate": 1.4502444675968496e-06,
+      "loss": 0.531,
+      "step": 14451
+    },
+    {
+      "epoch": 3.8365856896322845,
+      "grad_norm": 0.47209478582647557,
+      "learning_rate": 1.4499276200358031e-06,
+      "loss": 0.5442,
+      "step": 14452
+    },
+    {
+      "epoch": 3.836851188105668,
+      "grad_norm": 0.45790910803150287,
+      "learning_rate": 1.4496107929537403e-06,
+      "loss": 0.534,
+      "step": 14453
+    },
+    {
+      "epoch": 3.8371166865790522,
+      "grad_norm": 0.47481131408810984,
+      "learning_rate": 1.4492939863568379e-06,
+      "loss": 0.5524,
+      "step": 14454
+    },
+    {
+      "epoch": 3.837382185052436,
+      "grad_norm": 0.4807311968012792,
+      "learning_rate": 1.4489772002512748e-06,
+      "loss": 0.5186,
+      "step": 14455
+    },
+    {
+      "epoch": 3.8376476835258195,
+      "grad_norm": 0.463560076994564,
+      "learning_rate": 1.4486604346432311e-06,
+      "loss": 0.5713,
+      "step": 14456
+    },
+    {
+      "epoch": 3.8379131819992036,
+      "grad_norm": 0.4739190493513621,
+      "learning_rate": 1.448343689538883e-06,
+      "loss": 0.5577,
+      "step": 14457
+    },
+    {
+      "epoch": 3.8381786804725873,
+      "grad_norm": 0.46897797591620927,
+      "learning_rate": 1.4480269649444078e-06,
+      "loss": 0.5099,
+      "step": 14458
+    },
+    {
+      "epoch": 3.838444178945971,
+      "grad_norm": 0.46093583781014796,
+      "learning_rate": 1.4477102608659816e-06,
+      "loss": 0.5237,
+      "step": 14459
+    },
+    {
+      "epoch": 3.838709677419355,
+      "grad_norm": 0.47881392450239413,
+      "learning_rate": 1.4473935773097824e-06,
+      "loss": 0.5502,
+      "step": 14460
+    },
+    {
+      "epoch": 3.8389751758927386,
+      "grad_norm": 0.47346760678605226,
+      "learning_rate": 1.4470769142819849e-06,
+      "loss": 0.5349,
+      "step": 14461
+    },
+    {
+      "epoch": 3.8392406743661223,
+      "grad_norm": 0.45337099630836036,
+      "learning_rate": 1.4467602717887667e-06,
+      "loss": 0.5477,
+      "step": 14462
+    },
+    {
+      "epoch": 3.8395061728395063,
+      "grad_norm": 0.4790657068986263,
+      "learning_rate": 1.4464436498363015e-06,
+      "loss": 0.5595,
+      "step": 14463
+    },
+    {
+      "epoch": 3.83977167131289,
+      "grad_norm": 0.47294628174352765,
+      "learning_rate": 1.4461270484307644e-06,
+      "loss": 0.5556,
+      "step": 14464
+    },
+    {
+      "epoch": 3.8400371697862736,
+      "grad_norm": 0.47324777868539475,
+      "learning_rate": 1.4458104675783308e-06,
+      "loss": 0.5304,
+      "step": 14465
+    },
+    {
+      "epoch": 3.8403026682596577,
+      "grad_norm": 0.47758335482577563,
+      "learning_rate": 1.445493907285173e-06,
+      "loss": 0.5674,
+      "step": 14466
+    },
+    {
+      "epoch": 3.8405681667330414,
+      "grad_norm": 0.47106589828473805,
+      "learning_rate": 1.445177367557467e-06,
+      "loss": 0.5449,
+      "step": 14467
+    },
+    {
+      "epoch": 3.840833665206425,
+      "grad_norm": 0.48431801935508106,
+      "learning_rate": 1.444860848401384e-06,
+      "loss": 0.5494,
+      "step": 14468
+    },
+    {
+      "epoch": 3.841099163679809,
+      "grad_norm": 0.4636042267192663,
+      "learning_rate": 1.4445443498231e-06,
+      "loss": 0.5735,
+      "step": 14469
+    },
+    {
+      "epoch": 3.8413646621531927,
+      "grad_norm": 0.47590359672857124,
+      "learning_rate": 1.4442278718287833e-06,
+      "loss": 0.589,
+      "step": 14470
+    },
+    {
+      "epoch": 3.8416301606265764,
+      "grad_norm": 0.4802790373688087,
+      "learning_rate": 1.443911414424609e-06,
+      "loss": 0.5282,
+      "step": 14471
+    },
+    {
+      "epoch": 3.8418956590999604,
+      "grad_norm": 0.46361325091453537,
+      "learning_rate": 1.4435949776167468e-06,
+      "loss": 0.5439,
+      "step": 14472
+    },
+    {
+      "epoch": 3.842161157573344,
+      "grad_norm": 0.46770859599294196,
+      "learning_rate": 1.4432785614113703e-06,
+      "loss": 0.5363,
+      "step": 14473
+    },
+    {
+      "epoch": 3.8424266560467277,
+      "grad_norm": 0.4585079705100281,
+      "learning_rate": 1.4429621658146482e-06,
+      "loss": 0.5165,
+      "step": 14474
+    },
+    {
+      "epoch": 3.8426921545201114,
+      "grad_norm": 0.46219808819634706,
+      "learning_rate": 1.4426457908327527e-06,
+      "loss": 0.5386,
+      "step": 14475
+    },
+    {
+      "epoch": 3.842957652993495,
+      "grad_norm": 0.4623280415737434,
+      "learning_rate": 1.4423294364718532e-06,
+      "loss": 0.5155,
+      "step": 14476
+    },
+    {
+      "epoch": 3.843223151466879,
+      "grad_norm": 0.46750998376957675,
+      "learning_rate": 1.4420131027381185e-06,
+      "loss": 0.5372,
+      "step": 14477
+    },
+    {
+      "epoch": 3.8434886499402627,
+      "grad_norm": 0.4801609342687644,
+      "learning_rate": 1.4416967896377192e-06,
+      "loss": 0.5085,
+      "step": 14478
+    },
+    {
+      "epoch": 3.8437541484136464,
+      "grad_norm": 0.451489283015974,
+      "learning_rate": 1.441380497176823e-06,
+      "loss": 0.56,
+      "step": 14479
+    },
+    {
+      "epoch": 3.8440196468870305,
+      "grad_norm": 0.45443244011911504,
+      "learning_rate": 1.4410642253616003e-06,
+      "loss": 0.5495,
+      "step": 14480
+    },
+    {
+      "epoch": 3.844285145360414,
+      "grad_norm": 0.48046781862994653,
+      "learning_rate": 1.4407479741982178e-06,
+      "loss": 0.5564,
+      "step": 14481
+    },
+    {
+      "epoch": 3.8445506438337977,
+      "grad_norm": 0.48152291067862224,
+      "learning_rate": 1.4404317436928423e-06,
+      "loss": 0.5564,
+      "step": 14482
+    },
+    {
+      "epoch": 3.844816142307182,
+      "grad_norm": 0.46260963055464,
+      "learning_rate": 1.4401155338516426e-06,
+      "loss": 0.5201,
+      "step": 14483
+    },
+    {
+      "epoch": 3.8450816407805655,
+      "grad_norm": 0.4682342478224062,
+      "learning_rate": 1.4397993446807844e-06,
+      "loss": 0.5263,
+      "step": 14484
+    },
+    {
+      "epoch": 3.845347139253949,
+      "grad_norm": 0.47557628930474133,
+      "learning_rate": 1.4394831761864359e-06,
+      "loss": 0.5634,
+      "step": 14485
+    },
+    {
+      "epoch": 3.845612637727333,
+      "grad_norm": 0.4670738646762263,
+      "learning_rate": 1.4391670283747618e-06,
+      "loss": 0.538,
+      "step": 14486
+    },
+    {
+      "epoch": 3.845878136200717,
+      "grad_norm": 0.4715891982322148,
+      "learning_rate": 1.4388509012519283e-06,
+      "loss": 0.554,
+      "step": 14487
+    },
+    {
+      "epoch": 3.8461436346741005,
+      "grad_norm": 0.45574556400426836,
+      "learning_rate": 1.438534794824099e-06,
+      "loss": 0.5299,
+      "step": 14488
+    },
+    {
+      "epoch": 3.8464091331474846,
+      "grad_norm": 0.4861468355702461,
+      "learning_rate": 1.4382187090974415e-06,
+      "loss": 0.5455,
+      "step": 14489
+    },
+    {
+      "epoch": 3.846674631620868,
+      "grad_norm": 0.45165950114669745,
+      "learning_rate": 1.437902644078118e-06,
+      "loss": 0.5726,
+      "step": 14490
+    },
+    {
+      "epoch": 3.846940130094252,
+      "grad_norm": 0.48619679649492226,
+      "learning_rate": 1.4375865997722937e-06,
+      "loss": 0.5475,
+      "step": 14491
+    },
+    {
+      "epoch": 3.847205628567636,
+      "grad_norm": 0.4780104130389211,
+      "learning_rate": 1.4372705761861325e-06,
+      "loss": 0.4853,
+      "step": 14492
+    },
+    {
+      "epoch": 3.8474711270410196,
+      "grad_norm": 0.4660728869490942,
+      "learning_rate": 1.4369545733257962e-06,
+      "loss": 0.5354,
+      "step": 14493
+    },
+    {
+      "epoch": 3.847736625514403,
+      "grad_norm": 0.4690404047889851,
+      "learning_rate": 1.436638591197449e-06,
+      "loss": 0.5351,
+      "step": 14494
+    },
+    {
+      "epoch": 3.8480021239877873,
+      "grad_norm": 0.455333183407981,
+      "learning_rate": 1.436322629807252e-06,
+      "loss": 0.5556,
+      "step": 14495
+    },
+    {
+      "epoch": 3.848267622461171,
+      "grad_norm": 0.4628174230390237,
+      "learning_rate": 1.4360066891613692e-06,
+      "loss": 0.5633,
+      "step": 14496
+    },
+    {
+      "epoch": 3.8485331209345546,
+      "grad_norm": 0.49102227684238586,
+      "learning_rate": 1.43569076926596e-06,
+      "loss": 0.5201,
+      "step": 14497
+    },
+    {
+      "epoch": 3.8487986194079387,
+      "grad_norm": 0.46614854090669466,
+      "learning_rate": 1.4353748701271885e-06,
+      "loss": 0.5479,
+      "step": 14498
+    },
+    {
+      "epoch": 3.8490641178813223,
+      "grad_norm": 0.491352301238081,
+      "learning_rate": 1.435058991751212e-06,
+      "loss": 0.5702,
+      "step": 14499
+    },
+    {
+      "epoch": 3.849329616354706,
+      "grad_norm": 0.4706510491831853,
+      "learning_rate": 1.4347431341441934e-06,
+      "loss": 0.4972,
+      "step": 14500
+    },
+    {
+      "epoch": 3.8495951148280896,
+      "grad_norm": 0.4646121401158364,
+      "learning_rate": 1.4344272973122909e-06,
+      "loss": 0.5461,
+      "step": 14501
+    },
+    {
+      "epoch": 3.8498606133014737,
+      "grad_norm": 0.46455496969584625,
+      "learning_rate": 1.434111481261665e-06,
+      "loss": 0.5474,
+      "step": 14502
+    },
+    {
+      "epoch": 3.8501261117748573,
+      "grad_norm": 0.4709938083723083,
+      "learning_rate": 1.433795685998477e-06,
+      "loss": 0.5204,
+      "step": 14503
+    },
+    {
+      "epoch": 3.850391610248241,
+      "grad_norm": 0.4970623959685443,
+      "learning_rate": 1.4334799115288816e-06,
+      "loss": 0.537,
+      "step": 14504
+    },
+    {
+      "epoch": 3.8506571087216246,
+      "grad_norm": 0.45683009922022233,
+      "learning_rate": 1.4331641578590405e-06,
+      "loss": 0.5064,
+      "step": 14505
+    },
+    {
+      "epoch": 3.8509226071950087,
+      "grad_norm": 0.47260997375881325,
+      "learning_rate": 1.432848424995109e-06,
+      "loss": 0.5559,
+      "step": 14506
+    },
+    {
+      "epoch": 3.8511881056683923,
+      "grad_norm": 0.47397194433524387,
+      "learning_rate": 1.4325327129432473e-06,
+      "loss": 0.5455,
+      "step": 14507
+    },
+    {
+      "epoch": 3.851453604141776,
+      "grad_norm": 0.48146172107955476,
+      "learning_rate": 1.4322170217096104e-06,
+      "loss": 0.5526,
+      "step": 14508
+    },
+    {
+      "epoch": 3.85171910261516,
+      "grad_norm": 0.452216970845305,
+      "learning_rate": 1.4319013513003565e-06,
+      "loss": 0.5486,
+      "step": 14509
+    },
+    {
+      "epoch": 3.8519846010885437,
+      "grad_norm": 0.48138527483949667,
+      "learning_rate": 1.4315857017216417e-06,
+      "loss": 0.558,
+      "step": 14510
+    },
+    {
+      "epoch": 3.8522500995619273,
+      "grad_norm": 0.47406008339555655,
+      "learning_rate": 1.4312700729796207e-06,
+      "loss": 0.5268,
+      "step": 14511
+    },
+    {
+      "epoch": 3.8525155980353114,
+      "grad_norm": 0.4519140120162252,
+      "learning_rate": 1.430954465080451e-06,
+      "loss": 0.5437,
+      "step": 14512
+    },
+    {
+      "epoch": 3.852781096508695,
+      "grad_norm": 0.46678061348580907,
+      "learning_rate": 1.4306388780302858e-06,
+      "loss": 0.5261,
+      "step": 14513
+    },
+    {
+      "epoch": 3.8530465949820787,
+      "grad_norm": 0.457884201921897,
+      "learning_rate": 1.4303233118352817e-06,
+      "loss": 0.5604,
+      "step": 14514
+    },
+    {
+      "epoch": 3.8533120934554628,
+      "grad_norm": 0.469108932427907,
+      "learning_rate": 1.4300077665015916e-06,
+      "loss": 0.5744,
+      "step": 14515
+    },
+    {
+      "epoch": 3.8535775919288464,
+      "grad_norm": 0.4762336786578173,
+      "learning_rate": 1.4296922420353705e-06,
+      "loss": 0.58,
+      "step": 14516
+    },
+    {
+      "epoch": 3.85384309040223,
+      "grad_norm": 0.47588247918152304,
+      "learning_rate": 1.4293767384427695e-06,
+      "loss": 0.5452,
+      "step": 14517
+    },
+    {
+      "epoch": 3.854108588875614,
+      "grad_norm": 0.4676911749308556,
+      "learning_rate": 1.429061255729945e-06,
+      "loss": 0.5294,
+      "step": 14518
+    },
+    {
+      "epoch": 3.854374087348998,
+      "grad_norm": 0.4799964495116778,
+      "learning_rate": 1.428745793903047e-06,
+      "loss": 0.5888,
+      "step": 14519
+    },
+    {
+      "epoch": 3.8546395858223814,
+      "grad_norm": 0.47179741497921146,
+      "learning_rate": 1.4284303529682297e-06,
+      "loss": 0.556,
+      "step": 14520
+    },
+    {
+      "epoch": 3.8549050842957655,
+      "grad_norm": 0.4669240938625809,
+      "learning_rate": 1.4281149329316442e-06,
+      "loss": 0.5177,
+      "step": 14521
+    },
+    {
+      "epoch": 3.855170582769149,
+      "grad_norm": 0.4641420336555434,
+      "learning_rate": 1.427799533799441e-06,
+      "loss": 0.5428,
+      "step": 14522
+    },
+    {
+      "epoch": 3.855436081242533,
+      "grad_norm": 0.4555473662121771,
+      "learning_rate": 1.4274841555777736e-06,
+      "loss": 0.5606,
+      "step": 14523
+    },
+    {
+      "epoch": 3.855701579715917,
+      "grad_norm": 0.47527342394282035,
+      "learning_rate": 1.4271687982727895e-06,
+      "loss": 0.5639,
+      "step": 14524
+    },
+    {
+      "epoch": 3.8559670781893005,
+      "grad_norm": 0.4533830929855282,
+      "learning_rate": 1.4268534618906415e-06,
+      "loss": 0.557,
+      "step": 14525
+    },
+    {
+      "epoch": 3.856232576662684,
+      "grad_norm": 0.4658551494045167,
+      "learning_rate": 1.4265381464374789e-06,
+      "loss": 0.5647,
+      "step": 14526
+    },
+    {
+      "epoch": 3.8564980751360682,
+      "grad_norm": 0.47112906264934346,
+      "learning_rate": 1.4262228519194498e-06,
+      "loss": 0.5317,
+      "step": 14527
+    },
+    {
+      "epoch": 3.856763573609452,
+      "grad_norm": 0.4556243613220827,
+      "learning_rate": 1.4259075783427048e-06,
+      "loss": 0.5391,
+      "step": 14528
+    },
+    {
+      "epoch": 3.8570290720828355,
+      "grad_norm": 0.4713431254775322,
+      "learning_rate": 1.425592325713391e-06,
+      "loss": 0.5436,
+      "step": 14529
+    },
+    {
+      "epoch": 3.857294570556219,
+      "grad_norm": 0.46917277435205174,
+      "learning_rate": 1.4252770940376587e-06,
+      "loss": 0.5559,
+      "step": 14530
+    },
+    {
+      "epoch": 3.8575600690296032,
+      "grad_norm": 0.460098556294013,
+      "learning_rate": 1.424961883321653e-06,
+      "loss": 0.5274,
+      "step": 14531
+    },
+    {
+      "epoch": 3.857825567502987,
+      "grad_norm": 0.4604312877004099,
+      "learning_rate": 1.424646693571525e-06,
+      "loss": 0.567,
+      "step": 14532
+    },
+    {
+      "epoch": 3.8580910659763705,
+      "grad_norm": 0.4749245116778108,
+      "learning_rate": 1.4243315247934173e-06,
+      "loss": 0.5543,
+      "step": 14533
+    },
+    {
+      "epoch": 3.858356564449754,
+      "grad_norm": 0.4753517947610403,
+      "learning_rate": 1.4240163769934796e-06,
+      "loss": 0.5548,
+      "step": 14534
+    },
+    {
+      "epoch": 3.8586220629231383,
+      "grad_norm": 0.46057543412474194,
+      "learning_rate": 1.4237012501778559e-06,
+      "loss": 0.5679,
+      "step": 14535
+    },
+    {
+      "epoch": 3.858887561396522,
+      "grad_norm": 0.46831632774078913,
+      "learning_rate": 1.4233861443526943e-06,
+      "loss": 0.5124,
+      "step": 14536
+    },
+    {
+      "epoch": 3.8591530598699055,
+      "grad_norm": 0.4625166646183859,
+      "learning_rate": 1.4230710595241376e-06,
+      "loss": 0.528,
+      "step": 14537
+    },
+    {
+      "epoch": 3.8594185583432896,
+      "grad_norm": 0.4579865720526916,
+      "learning_rate": 1.4227559956983332e-06,
+      "loss": 0.5123,
+      "step": 14538
+    },
+    {
+      "epoch": 3.8596840568166733,
+      "grad_norm": 0.4577274472522476,
+      "learning_rate": 1.4224409528814242e-06,
+      "loss": 0.58,
+      "step": 14539
+    },
+    {
+      "epoch": 3.859949555290057,
+      "grad_norm": 0.4635938974773529,
+      "learning_rate": 1.4221259310795542e-06,
+      "loss": 0.5502,
+      "step": 14540
+    },
+    {
+      "epoch": 3.860215053763441,
+      "grad_norm": 0.4669266858180894,
+      "learning_rate": 1.4218109302988687e-06,
+      "loss": 0.5685,
+      "step": 14541
+    },
+    {
+      "epoch": 3.8604805522368246,
+      "grad_norm": 0.4815379442617971,
+      "learning_rate": 1.421495950545509e-06,
+      "loss": 0.5319,
+      "step": 14542
+    },
+    {
+      "epoch": 3.8607460507102083,
+      "grad_norm": 0.46579120182554096,
+      "learning_rate": 1.4211809918256196e-06,
+      "loss": 0.5804,
+      "step": 14543
+    },
+    {
+      "epoch": 3.8610115491835924,
+      "grad_norm": 0.467292638195511,
+      "learning_rate": 1.4208660541453423e-06,
+      "loss": 0.569,
+      "step": 14544
+    },
+    {
+      "epoch": 3.861277047656976,
+      "grad_norm": 0.48179160825088924,
+      "learning_rate": 1.4205511375108188e-06,
+      "loss": 0.5557,
+      "step": 14545
+    },
+    {
+      "epoch": 3.8615425461303596,
+      "grad_norm": 0.4779179731973171,
+      "learning_rate": 1.4202362419281906e-06,
+      "loss": 0.569,
+      "step": 14546
+    },
+    {
+      "epoch": 3.8618080446037437,
+      "grad_norm": 0.457104363064252,
+      "learning_rate": 1.4199213674035994e-06,
+      "loss": 0.5423,
+      "step": 14547
+    },
+    {
+      "epoch": 3.8620735430771274,
+      "grad_norm": 0.46092670192163904,
+      "learning_rate": 1.4196065139431866e-06,
+      "loss": 0.5332,
+      "step": 14548
+    },
+    {
+      "epoch": 3.862339041550511,
+      "grad_norm": 0.4585835145816026,
+      "learning_rate": 1.4192916815530925e-06,
+      "loss": 0.5568,
+      "step": 14549
+    },
+    {
+      "epoch": 3.862604540023895,
+      "grad_norm": 0.4620136357378521,
+      "learning_rate": 1.4189768702394563e-06,
+      "loss": 0.519,
+      "step": 14550
+    },
+    {
+      "epoch": 3.8628700384972787,
+      "grad_norm": 0.45864056388057356,
+      "learning_rate": 1.4186620800084171e-06,
+      "loss": 0.5414,
+      "step": 14551
+    },
+    {
+      "epoch": 3.8631355369706624,
+      "grad_norm": 0.44807336374252454,
+      "learning_rate": 1.4183473108661159e-06,
+      "loss": 0.5161,
+      "step": 14552
+    },
+    {
+      "epoch": 3.8634010354440464,
+      "grad_norm": 0.48061621217202666,
+      "learning_rate": 1.4180325628186895e-06,
+      "loss": 0.5386,
+      "step": 14553
+    },
+    {
+      "epoch": 3.86366653391743,
+      "grad_norm": 0.4469776231650078,
+      "learning_rate": 1.4177178358722781e-06,
+      "loss": 0.5439,
+      "step": 14554
+    },
+    {
+      "epoch": 3.8639320323908137,
+      "grad_norm": 0.4700133305226297,
+      "learning_rate": 1.4174031300330187e-06,
+      "loss": 0.5669,
+      "step": 14555
+    },
+    {
+      "epoch": 3.8641975308641974,
+      "grad_norm": 0.469667653125786,
+      "learning_rate": 1.4170884453070482e-06,
+      "loss": 0.535,
+      "step": 14556
+    },
+    {
+      "epoch": 3.8644630293375815,
+      "grad_norm": 0.46710397011748295,
+      "learning_rate": 1.4167737817005054e-06,
+      "loss": 0.5358,
+      "step": 14557
+    },
+    {
+      "epoch": 3.864728527810965,
+      "grad_norm": 0.4688581128704027,
+      "learning_rate": 1.416459139219525e-06,
+      "loss": 0.5406,
+      "step": 14558
+    },
+    {
+      "epoch": 3.8649940262843487,
+      "grad_norm": 0.44771648381199225,
+      "learning_rate": 1.4161445178702454e-06,
+      "loss": 0.5444,
+      "step": 14559
+    },
+    {
+      "epoch": 3.8652595247577324,
+      "grad_norm": 0.49076397048980513,
+      "learning_rate": 1.4158299176588003e-06,
+      "loss": 0.5365,
+      "step": 14560
+    },
+    {
+      "epoch": 3.8655250232311165,
+      "grad_norm": 0.457903973364067,
+      "learning_rate": 1.4155153385913283e-06,
+      "loss": 0.5712,
+      "step": 14561
+    },
+    {
+      "epoch": 3.8657905217045,
+      "grad_norm": 0.46776419291963567,
+      "learning_rate": 1.4152007806739607e-06,
+      "loss": 0.525,
+      "step": 14562
+    },
+    {
+      "epoch": 3.8660560201778837,
+      "grad_norm": 0.4759045900853393,
+      "learning_rate": 1.4148862439128347e-06,
+      "loss": 0.5245,
+      "step": 14563
+    },
+    {
+      "epoch": 3.866321518651268,
+      "grad_norm": 0.46393932652777425,
+      "learning_rate": 1.4145717283140827e-06,
+      "loss": 0.5413,
+      "step": 14564
+    },
+    {
+      "epoch": 3.8665870171246515,
+      "grad_norm": 0.4684092891614069,
+      "learning_rate": 1.414257233883841e-06,
+      "loss": 0.5894,
+      "step": 14565
+    },
+    {
+      "epoch": 3.866852515598035,
+      "grad_norm": 0.48128400002742555,
+      "learning_rate": 1.4139427606282401e-06,
+      "loss": 0.5629,
+      "step": 14566
+    },
+    {
+      "epoch": 3.867118014071419,
+      "grad_norm": 0.4720521679192385,
+      "learning_rate": 1.4136283085534158e-06,
+      "loss": 0.586,
+      "step": 14567
+    },
+    {
+      "epoch": 3.867383512544803,
+      "grad_norm": 0.4698173089740331,
+      "learning_rate": 1.4133138776654993e-06,
+      "loss": 0.5533,
+      "step": 14568
+    },
+    {
+      "epoch": 3.8676490110181865,
+      "grad_norm": 0.4584665536373183,
+      "learning_rate": 1.412999467970622e-06,
+      "loss": 0.5095,
+      "step": 14569
+    },
+    {
+      "epoch": 3.8679145094915706,
+      "grad_norm": 0.46109944385299584,
+      "learning_rate": 1.4126850794749172e-06,
+      "loss": 0.55,
+      "step": 14570
+    },
+    {
+      "epoch": 3.868180007964954,
+      "grad_norm": 0.4626015167052272,
+      "learning_rate": 1.4123707121845148e-06,
+      "loss": 0.5653,
+      "step": 14571
+    },
+    {
+      "epoch": 3.868445506438338,
+      "grad_norm": 0.45827894692358045,
+      "learning_rate": 1.4120563661055472e-06,
+      "loss": 0.5439,
+      "step": 14572
+    },
+    {
+      "epoch": 3.868711004911722,
+      "grad_norm": 0.47474894652575855,
+      "learning_rate": 1.4117420412441447e-06,
+      "loss": 0.566,
+      "step": 14573
+    },
+    {
+      "epoch": 3.8689765033851056,
+      "grad_norm": 0.4647975689649319,
+      "learning_rate": 1.4114277376064357e-06,
+      "loss": 0.5378,
+      "step": 14574
+    },
+    {
+      "epoch": 3.869242001858489,
+      "grad_norm": 0.46221238287730937,
+      "learning_rate": 1.4111134551985516e-06,
+      "loss": 0.5716,
+      "step": 14575
+    },
+    {
+      "epoch": 3.8695075003318733,
+      "grad_norm": 0.47946702977695727,
+      "learning_rate": 1.4107991940266207e-06,
+      "loss": 0.5619,
+      "step": 14576
+    },
+    {
+      "epoch": 3.869772998805257,
+      "grad_norm": 0.4725087688956933,
+      "learning_rate": 1.4104849540967732e-06,
+      "loss": 0.5568,
+      "step": 14577
+    },
+    {
+      "epoch": 3.8700384972786406,
+      "grad_norm": 0.46783068147796747,
+      "learning_rate": 1.4101707354151367e-06,
+      "loss": 0.556,
+      "step": 14578
+    },
+    {
+      "epoch": 3.8703039957520247,
+      "grad_norm": 0.48317973303036454,
+      "learning_rate": 1.4098565379878393e-06,
+      "loss": 0.5337,
+      "step": 14579
+    },
+    {
+      "epoch": 3.8705694942254083,
+      "grad_norm": 0.47190204917526046,
+      "learning_rate": 1.4095423618210075e-06,
+      "loss": 0.5494,
+      "step": 14580
+    },
+    {
+      "epoch": 3.870834992698792,
+      "grad_norm": 0.4710787389006093,
+      "learning_rate": 1.4092282069207706e-06,
+      "loss": 0.5438,
+      "step": 14581
+    },
+    {
+      "epoch": 3.871100491172176,
+      "grad_norm": 0.45138610046654265,
+      "learning_rate": 1.4089140732932533e-06,
+      "loss": 0.5271,
+      "step": 14582
+    },
+    {
+      "epoch": 3.8713659896455597,
+      "grad_norm": 0.4514330753956673,
+      "learning_rate": 1.4085999609445838e-06,
+      "loss": 0.5455,
+      "step": 14583
+    },
+    {
+      "epoch": 3.8716314881189433,
+      "grad_norm": 0.4695342622796812,
+      "learning_rate": 1.4082858698808876e-06,
+      "loss": 0.5377,
+      "step": 14584
+    },
+    {
+      "epoch": 3.871896986592327,
+      "grad_norm": 0.4630429783151699,
+      "learning_rate": 1.4079718001082888e-06,
+      "loss": 0.5425,
+      "step": 14585
+    },
+    {
+      "epoch": 3.872162485065711,
+      "grad_norm": 0.482889101243248,
+      "learning_rate": 1.4076577516329143e-06,
+      "loss": 0.5619,
+      "step": 14586
+    },
+    {
+      "epoch": 3.8724279835390947,
+      "grad_norm": 0.4606513131518334,
+      "learning_rate": 1.4073437244608878e-06,
+      "loss": 0.5299,
+      "step": 14587
+    },
+    {
+      "epoch": 3.8726934820124783,
+      "grad_norm": 0.4626473748109309,
+      "learning_rate": 1.4070297185983344e-06,
+      "loss": 0.5556,
+      "step": 14588
+    },
+    {
+      "epoch": 3.872958980485862,
+      "grad_norm": 0.45200956060469094,
+      "learning_rate": 1.4067157340513767e-06,
+      "loss": 0.5275,
+      "step": 14589
+    },
+    {
+      "epoch": 3.873224478959246,
+      "grad_norm": 0.4723308718625095,
+      "learning_rate": 1.4064017708261414e-06,
+      "loss": 0.5583,
+      "step": 14590
+    },
+    {
+      "epoch": 3.8734899774326297,
+      "grad_norm": 0.47588599151662264,
+      "learning_rate": 1.4060878289287467e-06,
+      "loss": 0.5638,
+      "step": 14591
+    },
+    {
+      "epoch": 3.8737554759060133,
+      "grad_norm": 0.47446160133502857,
+      "learning_rate": 1.4057739083653176e-06,
+      "loss": 0.5717,
+      "step": 14592
+    },
+    {
+      "epoch": 3.8740209743793974,
+      "grad_norm": 0.48387548471050923,
+      "learning_rate": 1.405460009141978e-06,
+      "loss": 0.5557,
+      "step": 14593
+    },
+    {
+      "epoch": 3.874286472852781,
+      "grad_norm": 0.45927624963912456,
+      "learning_rate": 1.4051461312648465e-06,
+      "loss": 0.5574,
+      "step": 14594
+    },
+    {
+      "epoch": 3.8745519713261647,
+      "grad_norm": 0.45610238706752826,
+      "learning_rate": 1.4048322747400472e-06,
+      "loss": 0.5347,
+      "step": 14595
+    },
+    {
+      "epoch": 3.8748174697995488,
+      "grad_norm": 0.4680973971337468,
+      "learning_rate": 1.4045184395737004e-06,
+      "loss": 0.5308,
+      "step": 14596
+    },
+    {
+      "epoch": 3.8750829682729324,
+      "grad_norm": 0.4571717387097576,
+      "learning_rate": 1.404204625771926e-06,
+      "loss": 0.573,
+      "step": 14597
+    },
+    {
+      "epoch": 3.875348466746316,
+      "grad_norm": 0.4759557863354639,
+      "learning_rate": 1.4038908333408432e-06,
+      "loss": 0.5791,
+      "step": 14598
+    },
+    {
+      "epoch": 3.8756139652197,
+      "grad_norm": 0.46506543551314217,
+      "learning_rate": 1.4035770622865741e-06,
+      "loss": 0.5415,
+      "step": 14599
+    },
+    {
+      "epoch": 3.875879463693084,
+      "grad_norm": 0.4679457645941911,
+      "learning_rate": 1.4032633126152357e-06,
+      "loss": 0.5566,
+      "step": 14600
+    },
+    {
+      "epoch": 3.8761449621664674,
+      "grad_norm": 0.44938418925165097,
+      "learning_rate": 1.4029495843329488e-06,
+      "loss": 0.549,
+      "step": 14601
+    },
+    {
+      "epoch": 3.8764104606398515,
+      "grad_norm": 0.45220553531263163,
+      "learning_rate": 1.4026358774458315e-06,
+      "loss": 0.5227,
+      "step": 14602
+    },
+    {
+      "epoch": 3.876675959113235,
+      "grad_norm": 0.46582176291676886,
+      "learning_rate": 1.40232219196e-06,
+      "loss": 0.5303,
+      "step": 14603
+    },
+    {
+      "epoch": 3.876941457586619,
+      "grad_norm": 0.4619228670997488,
+      "learning_rate": 1.4020085278815745e-06,
+      "loss": 0.5704,
+      "step": 14604
+    },
+    {
+      "epoch": 3.877206956060003,
+      "grad_norm": 0.47270917262846524,
+      "learning_rate": 1.40169488521667e-06,
+      "loss": 0.5606,
+      "step": 14605
+    },
+    {
+      "epoch": 3.8774724545333865,
+      "grad_norm": 0.45156577527593345,
+      "learning_rate": 1.4013812639714053e-06,
+      "loss": 0.5514,
+      "step": 14606
+    },
+    {
+      "epoch": 3.87773795300677,
+      "grad_norm": 0.4591387232748298,
+      "learning_rate": 1.4010676641518961e-06,
+      "loss": 0.5178,
+      "step": 14607
+    },
+    {
+      "epoch": 3.8780034514801542,
+      "grad_norm": 0.4598402067380847,
+      "learning_rate": 1.4007540857642577e-06,
+      "loss": 0.526,
+      "step": 14608
+    },
+    {
+      "epoch": 3.878268949953538,
+      "grad_norm": 0.4567263766755188,
+      "learning_rate": 1.4004405288146056e-06,
+      "loss": 0.5452,
+      "step": 14609
+    },
+    {
+      "epoch": 3.8785344484269215,
+      "grad_norm": 0.47399802519519074,
+      "learning_rate": 1.4001269933090558e-06,
+      "loss": 0.574,
+      "step": 14610
+    },
+    {
+      "epoch": 3.878799946900305,
+      "grad_norm": 0.4832410721172336,
+      "learning_rate": 1.399813479253722e-06,
+      "loss": 0.5753,
+      "step": 14611
+    },
+    {
+      "epoch": 3.8790654453736892,
+      "grad_norm": 0.4605716009281784,
+      "learning_rate": 1.3994999866547199e-06,
+      "loss": 0.5551,
+      "step": 14612
+    },
+    {
+      "epoch": 3.879330943847073,
+      "grad_norm": 0.47817868653202517,
+      "learning_rate": 1.3991865155181622e-06,
+      "loss": 0.5477,
+      "step": 14613
+    },
+    {
+      "epoch": 3.8795964423204565,
+      "grad_norm": 0.4807410134645125,
+      "learning_rate": 1.398873065850162e-06,
+      "loss": 0.5419,
+      "step": 14614
+    },
+    {
+      "epoch": 3.87986194079384,
+      "grad_norm": 0.4635422867732951,
+      "learning_rate": 1.3985596376568345e-06,
+      "loss": 0.5799,
+      "step": 14615
+    },
+    {
+      "epoch": 3.8801274392672243,
+      "grad_norm": 0.49090881298409933,
+      "learning_rate": 1.398246230944289e-06,
+      "loss": 0.5194,
+      "step": 14616
+    },
+    {
+      "epoch": 3.880392937740608,
+      "grad_norm": 0.47559051214039766,
+      "learning_rate": 1.3979328457186414e-06,
+      "loss": 0.5652,
+      "step": 14617
+    },
+    {
+      "epoch": 3.8806584362139915,
+      "grad_norm": 0.46196078691913994,
+      "learning_rate": 1.3976194819860002e-06,
+      "loss": 0.5202,
+      "step": 14618
+    },
+    {
+      "epoch": 3.8809239346873756,
+      "grad_norm": 0.46768709326986047,
+      "learning_rate": 1.397306139752479e-06,
+      "loss": 0.5331,
+      "step": 14619
+    },
+    {
+      "epoch": 3.8811894331607593,
+      "grad_norm": 0.4955741192236411,
+      "learning_rate": 1.3969928190241883e-06,
+      "loss": 0.5208,
+      "step": 14620
+    },
+    {
+      "epoch": 3.881454931634143,
+      "grad_norm": 0.5233763873355358,
+      "learning_rate": 1.3966795198072368e-06,
+      "loss": 0.5526,
+      "step": 14621
+    },
+    {
+      "epoch": 3.881720430107527,
+      "grad_norm": 0.4970407519276715,
+      "learning_rate": 1.3963662421077375e-06,
+      "loss": 0.5559,
+      "step": 14622
+    },
+    {
+      "epoch": 3.8819859285809106,
+      "grad_norm": 0.4418444968684636,
+      "learning_rate": 1.3960529859317975e-06,
+      "loss": 0.4978,
+      "step": 14623
+    },
+    {
+      "epoch": 3.8822514270542943,
+      "grad_norm": 0.4728863184738252,
+      "learning_rate": 1.3957397512855292e-06,
+      "loss": 0.5227,
+      "step": 14624
+    },
+    {
+      "epoch": 3.8825169255276784,
+      "grad_norm": 0.49615760807246234,
+      "learning_rate": 1.3954265381750376e-06,
+      "loss": 0.537,
+      "step": 14625
+    },
+    {
+      "epoch": 3.882782424001062,
+      "grad_norm": 0.47594930561159454,
+      "learning_rate": 1.3951133466064338e-06,
+      "loss": 0.5344,
+      "step": 14626
+    },
+    {
+      "epoch": 3.8830479224744456,
+      "grad_norm": 0.43243043685319377,
+      "learning_rate": 1.394800176585824e-06,
+      "loss": 0.529,
+      "step": 14627
+    },
+    {
+      "epoch": 3.8833134209478297,
+      "grad_norm": 0.4769097540956097,
+      "learning_rate": 1.3944870281193178e-06,
+      "loss": 0.5297,
+      "step": 14628
+    },
+    {
+      "epoch": 3.8835789194212134,
+      "grad_norm": 0.46903828762799354,
+      "learning_rate": 1.3941739012130201e-06,
+      "loss": 0.5988,
+      "step": 14629
+    },
+    {
+      "epoch": 3.883844417894597,
+      "grad_norm": 0.48002657148983124,
+      "learning_rate": 1.3938607958730401e-06,
+      "loss": 0.5218,
+      "step": 14630
+    },
+    {
+      "epoch": 3.884109916367981,
+      "grad_norm": 0.45161501279156036,
+      "learning_rate": 1.393547712105483e-06,
+      "loss": 0.5094,
+      "step": 14631
+    },
+    {
+      "epoch": 3.8843754148413647,
+      "grad_norm": 0.4631617045904441,
+      "learning_rate": 1.3932346499164535e-06,
+      "loss": 0.5152,
+      "step": 14632
+    },
+    {
+      "epoch": 3.8846409133147484,
+      "grad_norm": 0.47897668056463677,
+      "learning_rate": 1.3929216093120588e-06,
+      "loss": 0.5548,
+      "step": 14633
+    },
+    {
+      "epoch": 3.8849064117881325,
+      "grad_norm": 0.4564437826538093,
+      "learning_rate": 1.3926085902984028e-06,
+      "loss": 0.5471,
+      "step": 14634
+    },
+    {
+      "epoch": 3.885171910261516,
+      "grad_norm": 0.47695954296621607,
+      "learning_rate": 1.3922955928815911e-06,
+      "loss": 0.5581,
+      "step": 14635
+    },
+    {
+      "epoch": 3.8854374087348997,
+      "grad_norm": 0.46393842216661885,
+      "learning_rate": 1.3919826170677278e-06,
+      "loss": 0.5384,
+      "step": 14636
+    },
+    {
+      "epoch": 3.885702907208284,
+      "grad_norm": 0.4725223226359163,
+      "learning_rate": 1.3916696628629153e-06,
+      "loss": 0.5502,
+      "step": 14637
+    },
+    {
+      "epoch": 3.8859684056816675,
+      "grad_norm": 0.46169768282044543,
+      "learning_rate": 1.3913567302732588e-06,
+      "loss": 0.5725,
+      "step": 14638
+    },
+    {
+      "epoch": 3.886233904155051,
+      "grad_norm": 0.48043285542458675,
+      "learning_rate": 1.3910438193048598e-06,
+      "loss": 0.5453,
+      "step": 14639
+    },
+    {
+      "epoch": 3.8864994026284347,
+      "grad_norm": 0.460069084328353,
+      "learning_rate": 1.3907309299638229e-06,
+      "loss": 0.5177,
+      "step": 14640
+    },
+    {
+      "epoch": 3.886764901101819,
+      "grad_norm": 0.4806198299168726,
+      "learning_rate": 1.390418062256247e-06,
+      "loss": 0.5744,
+      "step": 14641
+    },
+    {
+      "epoch": 3.8870303995752025,
+      "grad_norm": 0.4625539970677221,
+      "learning_rate": 1.3901052161882382e-06,
+      "loss": 0.5472,
+      "step": 14642
+    },
+    {
+      "epoch": 3.887295898048586,
+      "grad_norm": 0.4662855328497534,
+      "learning_rate": 1.389792391765893e-06,
+      "loss": 0.562,
+      "step": 14643
+    },
+    {
+      "epoch": 3.8875613965219697,
+      "grad_norm": 0.46775846636456314,
+      "learning_rate": 1.3894795889953152e-06,
+      "loss": 0.5752,
+      "step": 14644
+    },
+    {
+      "epoch": 3.887826894995354,
+      "grad_norm": 0.47299443247000866,
+      "learning_rate": 1.3891668078826036e-06,
+      "loss": 0.5658,
+      "step": 14645
+    },
+    {
+      "epoch": 3.8880923934687375,
+      "grad_norm": 0.46865238246738633,
+      "learning_rate": 1.3888540484338604e-06,
+      "loss": 0.5484,
+      "step": 14646
+    },
+    {
+      "epoch": 3.888357891942121,
+      "grad_norm": 0.45843471154084525,
+      "learning_rate": 1.3885413106551832e-06,
+      "loss": 0.5454,
+      "step": 14647
+    },
+    {
+      "epoch": 3.888623390415505,
+      "grad_norm": 0.4532309740362265,
+      "learning_rate": 1.388228594552671e-06,
+      "loss": 0.5506,
+      "step": 14648
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 0.4717129902575639,
+      "learning_rate": 1.3879159001324244e-06,
+      "loss": 0.5379,
+      "step": 14649
+    },
+    {
+      "epoch": 3.8891543873622725,
+      "grad_norm": 0.4661341063159216,
+      "learning_rate": 1.3876032274005397e-06,
+      "loss": 0.5183,
+      "step": 14650
+    },
+    {
+      "epoch": 3.8894198858356566,
+      "grad_norm": 0.46499082008360926,
+      "learning_rate": 1.3872905763631162e-06,
+      "loss": 0.5495,
+      "step": 14651
+    },
+    {
+      "epoch": 3.88968538430904,
+      "grad_norm": 0.4603289353528457,
+      "learning_rate": 1.3869779470262501e-06,
+      "loss": 0.5331,
+      "step": 14652
+    },
+    {
+      "epoch": 3.889950882782424,
+      "grad_norm": 0.47228495258726216,
+      "learning_rate": 1.3866653393960414e-06,
+      "loss": 0.5774,
+      "step": 14653
+    },
+    {
+      "epoch": 3.890216381255808,
+      "grad_norm": 0.45824429091559105,
+      "learning_rate": 1.3863527534785823e-06,
+      "loss": 0.5507,
+      "step": 14654
+    },
+    {
+      "epoch": 3.8904818797291916,
+      "grad_norm": 0.47484075317806473,
+      "learning_rate": 1.386040189279972e-06,
+      "loss": 0.5446,
+      "step": 14655
+    },
+    {
+      "epoch": 3.890747378202575,
+      "grad_norm": 0.45598208266014634,
+      "learning_rate": 1.3857276468063047e-06,
+      "loss": 0.5555,
+      "step": 14656
+    },
+    {
+      "epoch": 3.8910128766759593,
+      "grad_norm": 0.47959725027240613,
+      "learning_rate": 1.3854151260636777e-06,
+      "loss": 0.5282,
+      "step": 14657
+    },
+    {
+      "epoch": 3.891278375149343,
+      "grad_norm": 0.46013834574993473,
+      "learning_rate": 1.3851026270581835e-06,
+      "loss": 0.5549,
+      "step": 14658
+    },
+    {
+      "epoch": 3.8915438736227266,
+      "grad_norm": 0.47313477684873456,
+      "learning_rate": 1.3847901497959188e-06,
+      "loss": 0.5429,
+      "step": 14659
+    },
+    {
+      "epoch": 3.8918093720961107,
+      "grad_norm": 0.476838494579563,
+      "learning_rate": 1.3844776942829766e-06,
+      "loss": 0.5231,
+      "step": 14660
+    },
+    {
+      "epoch": 3.8920748705694943,
+      "grad_norm": 0.47064964081203203,
+      "learning_rate": 1.3841652605254497e-06,
+      "loss": 0.5802,
+      "step": 14661
+    },
+    {
+      "epoch": 3.892340369042878,
+      "grad_norm": 0.46724405130390656,
+      "learning_rate": 1.3838528485294333e-06,
+      "loss": 0.5878,
+      "step": 14662
+    },
+    {
+      "epoch": 3.892605867516262,
+      "grad_norm": 0.469479644671558,
+      "learning_rate": 1.3835404583010181e-06,
+      "loss": 0.5687,
+      "step": 14663
+    },
+    {
+      "epoch": 3.8928713659896457,
+      "grad_norm": 0.48722486070806914,
+      "learning_rate": 1.3832280898462984e-06,
+      "loss": 0.5779,
+      "step": 14664
+    },
+    {
+      "epoch": 3.8931368644630293,
+      "grad_norm": 0.4824443942501269,
+      "learning_rate": 1.3829157431713655e-06,
+      "loss": 0.5495,
+      "step": 14665
+    },
+    {
+      "epoch": 3.893402362936413,
+      "grad_norm": 0.46940619450233945,
+      "learning_rate": 1.3826034182823095e-06,
+      "loss": 0.5599,
+      "step": 14666
+    },
+    {
+      "epoch": 3.893667861409797,
+      "grad_norm": 0.47881562288195634,
+      "learning_rate": 1.3822911151852238e-06,
+      "loss": 0.5768,
+      "step": 14667
+    },
+    {
+      "epoch": 3.8939333598831807,
+      "grad_norm": 0.47096488523874175,
+      "learning_rate": 1.3819788338861973e-06,
+      "loss": 0.5698,
+      "step": 14668
+    },
+    {
+      "epoch": 3.8941988583565643,
+      "grad_norm": 0.48081977540503884,
+      "learning_rate": 1.3816665743913215e-06,
+      "loss": 0.5291,
+      "step": 14669
+    },
+    {
+      "epoch": 3.894464356829948,
+      "grad_norm": 0.47604222578883354,
+      "learning_rate": 1.3813543367066857e-06,
+      "loss": 0.5644,
+      "step": 14670
+    },
+    {
+      "epoch": 3.894729855303332,
+      "grad_norm": 0.4716669152894754,
+      "learning_rate": 1.3810421208383796e-06,
+      "loss": 0.5548,
+      "step": 14671
+    },
+    {
+      "epoch": 3.8949953537767157,
+      "grad_norm": 0.459527186825024,
+      "learning_rate": 1.3807299267924903e-06,
+      "loss": 0.558,
+      "step": 14672
+    },
+    {
+      "epoch": 3.8952608522500993,
+      "grad_norm": 0.47570077369854546,
+      "learning_rate": 1.380417754575109e-06,
+      "loss": 0.5201,
+      "step": 14673
+    },
+    {
+      "epoch": 3.8955263507234834,
+      "grad_norm": 0.47964365079636573,
+      "learning_rate": 1.380105604192322e-06,
+      "loss": 0.5423,
+      "step": 14674
+    },
+    {
+      "epoch": 3.895791849196867,
+      "grad_norm": 0.47107753609713254,
+      "learning_rate": 1.3797934756502185e-06,
+      "loss": 0.5801,
+      "step": 14675
+    },
+    {
+      "epoch": 3.8960573476702507,
+      "grad_norm": 0.4734145644408046,
+      "learning_rate": 1.3794813689548846e-06,
+      "loss": 0.5708,
+      "step": 14676
+    },
+    {
+      "epoch": 3.8963228461436348,
+      "grad_norm": 0.45209016135240243,
+      "learning_rate": 1.3791692841124072e-06,
+      "loss": 0.4716,
+      "step": 14677
+    },
+    {
+      "epoch": 3.8965883446170184,
+      "grad_norm": 0.4763351749587214,
+      "learning_rate": 1.3788572211288737e-06,
+      "loss": 0.5699,
+      "step": 14678
+    },
+    {
+      "epoch": 3.896853843090402,
+      "grad_norm": 0.4622684200439051,
+      "learning_rate": 1.3785451800103683e-06,
+      "loss": 0.5181,
+      "step": 14679
+    },
+    {
+      "epoch": 3.897119341563786,
+      "grad_norm": 0.45861001528541717,
+      "learning_rate": 1.378233160762979e-06,
+      "loss": 0.5743,
+      "step": 14680
+    },
+    {
+      "epoch": 3.89738484003717,
+      "grad_norm": 0.4876186019673994,
+      "learning_rate": 1.3779211633927886e-06,
+      "loss": 0.5922,
+      "step": 14681
+    },
+    {
+      "epoch": 3.8976503385105534,
+      "grad_norm": 0.46860008132698133,
+      "learning_rate": 1.3776091879058845e-06,
+      "loss": 0.5393,
+      "step": 14682
+    },
+    {
+      "epoch": 3.8979158369839375,
+      "grad_norm": 0.459396805439591,
+      "learning_rate": 1.3772972343083474e-06,
+      "loss": 0.5375,
+      "step": 14683
+    },
+    {
+      "epoch": 3.898181335457321,
+      "grad_norm": 0.4767621731472709,
+      "learning_rate": 1.3769853026062631e-06,
+      "loss": 0.5705,
+      "step": 14684
+    },
+    {
+      "epoch": 3.898446833930705,
+      "grad_norm": 0.47236867147316836,
+      "learning_rate": 1.3766733928057163e-06,
+      "loss": 0.5424,
+      "step": 14685
+    },
+    {
+      "epoch": 3.898712332404089,
+      "grad_norm": 0.4659832277113816,
+      "learning_rate": 1.3763615049127877e-06,
+      "loss": 0.5331,
+      "step": 14686
+    },
+    {
+      "epoch": 3.8989778308774725,
+      "grad_norm": 0.47957665241764624,
+      "learning_rate": 1.3760496389335615e-06,
+      "loss": 0.5672,
+      "step": 14687
+    },
+    {
+      "epoch": 3.899243329350856,
+      "grad_norm": 0.4621717170143549,
+      "learning_rate": 1.3757377948741193e-06,
+      "loss": 0.5577,
+      "step": 14688
+    },
+    {
+      "epoch": 3.8995088278242402,
+      "grad_norm": 0.4773324531027125,
+      "learning_rate": 1.3754259727405429e-06,
+      "loss": 0.5652,
+      "step": 14689
+    },
+    {
+      "epoch": 3.899774326297624,
+      "grad_norm": 0.4391500050693001,
+      "learning_rate": 1.3751141725389128e-06,
+      "loss": 0.5528,
+      "step": 14690
+    },
+    {
+      "epoch": 3.9000398247710075,
+      "grad_norm": 0.4710193575485356,
+      "learning_rate": 1.3748023942753109e-06,
+      "loss": 0.5366,
+      "step": 14691
+    },
+    {
+      "epoch": 3.9003053232443916,
+      "grad_norm": 0.4924782222550652,
+      "learning_rate": 1.3744906379558165e-06,
+      "loss": 0.526,
+      "step": 14692
+    },
+    {
+      "epoch": 3.9005708217177752,
+      "grad_norm": 0.48019451054364637,
+      "learning_rate": 1.3741789035865116e-06,
+      "loss": 0.5331,
+      "step": 14693
+    },
+    {
+      "epoch": 3.900836320191159,
+      "grad_norm": 0.4737861642442203,
+      "learning_rate": 1.3738671911734747e-06,
+      "loss": 0.5509,
+      "step": 14694
+    },
+    {
+      "epoch": 3.9011018186645425,
+      "grad_norm": 0.4821552896576482,
+      "learning_rate": 1.3735555007227832e-06,
+      "loss": 0.5284,
+      "step": 14695
+    },
+    {
+      "epoch": 3.9013673171379266,
+      "grad_norm": 0.46493916851530914,
+      "learning_rate": 1.373243832240519e-06,
+      "loss": 0.5074,
+      "step": 14696
+    },
+    {
+      "epoch": 3.9016328156113103,
+      "grad_norm": 0.48142185346293015,
+      "learning_rate": 1.3729321857327576e-06,
+      "loss": 0.5471,
+      "step": 14697
+    },
+    {
+      "epoch": 3.901898314084694,
+      "grad_norm": 0.4679003214755217,
+      "learning_rate": 1.3726205612055793e-06,
+      "loss": 0.5292,
+      "step": 14698
+    },
+    {
+      "epoch": 3.9021638125580775,
+      "grad_norm": 0.46199303878001674,
+      "learning_rate": 1.37230895866506e-06,
+      "loss": 0.5425,
+      "step": 14699
+    },
+    {
+      "epoch": 3.9024293110314616,
+      "grad_norm": 0.4650254977365548,
+      "learning_rate": 1.3719973781172774e-06,
+      "loss": 0.5266,
+      "step": 14700
+    },
+    {
+      "epoch": 3.9026948095048453,
+      "grad_norm": 0.4599529816970895,
+      "learning_rate": 1.3716858195683064e-06,
+      "loss": 0.5501,
+      "step": 14701
+    },
+    {
+      "epoch": 3.902960307978229,
+      "grad_norm": 0.47174524987245536,
+      "learning_rate": 1.371374283024226e-06,
+      "loss": 0.5196,
+      "step": 14702
+    },
+    {
+      "epoch": 3.903225806451613,
+      "grad_norm": 0.4860061819626319,
+      "learning_rate": 1.3710627684911092e-06,
+      "loss": 0.5455,
+      "step": 14703
+    },
+    {
+      "epoch": 3.9034913049249966,
+      "grad_norm": 0.4647455081266664,
+      "learning_rate": 1.3707512759750324e-06,
+      "loss": 0.5011,
+      "step": 14704
+    },
+    {
+      "epoch": 3.9037568033983803,
+      "grad_norm": 0.4632936640194412,
+      "learning_rate": 1.3704398054820728e-06,
+      "loss": 0.5645,
+      "step": 14705
+    },
+    {
+      "epoch": 3.9040223018717644,
+      "grad_norm": 0.47138748801159297,
+      "learning_rate": 1.3701283570183004e-06,
+      "loss": 0.5351,
+      "step": 14706
+    },
+    {
+      "epoch": 3.904287800345148,
+      "grad_norm": 0.4748584263480536,
+      "learning_rate": 1.3698169305897926e-06,
+      "loss": 0.5416,
+      "step": 14707
+    },
+    {
+      "epoch": 3.9045532988185316,
+      "grad_norm": 0.4731263686579328,
+      "learning_rate": 1.3695055262026208e-06,
+      "loss": 0.5676,
+      "step": 14708
+    },
+    {
+      "epoch": 3.9048187972919157,
+      "grad_norm": 0.45485223735229724,
+      "learning_rate": 1.3691941438628604e-06,
+      "loss": 0.5306,
+      "step": 14709
+    },
+    {
+      "epoch": 3.9050842957652994,
+      "grad_norm": 0.46877002053759415,
+      "learning_rate": 1.3688827835765817e-06,
+      "loss": 0.531,
+      "step": 14710
+    },
+    {
+      "epoch": 3.905349794238683,
+      "grad_norm": 0.46882009389509993,
+      "learning_rate": 1.3685714453498589e-06,
+      "loss": 0.5317,
+      "step": 14711
+    },
+    {
+      "epoch": 3.905615292712067,
+      "grad_norm": 0.4927628202076177,
+      "learning_rate": 1.3682601291887632e-06,
+      "loss": 0.5696,
+      "step": 14712
+    },
+    {
+      "epoch": 3.9058807911854507,
+      "grad_norm": 0.46806782904094213,
+      "learning_rate": 1.3679488350993653e-06,
+      "loss": 0.5307,
+      "step": 14713
+    },
+    {
+      "epoch": 3.9061462896588344,
+      "grad_norm": 0.4659189979546424,
+      "learning_rate": 1.3676375630877376e-06,
+      "loss": 0.5101,
+      "step": 14714
+    },
+    {
+      "epoch": 3.9064117881322185,
+      "grad_norm": 0.46741460357093256,
+      "learning_rate": 1.3673263131599491e-06,
+      "loss": 0.5593,
+      "step": 14715
+    },
+    {
+      "epoch": 3.906677286605602,
+      "grad_norm": 0.4683662134335134,
+      "learning_rate": 1.3670150853220716e-06,
+      "loss": 0.5499,
+      "step": 14716
+    },
+    {
+      "epoch": 3.9069427850789857,
+      "grad_norm": 0.48825242576534766,
+      "learning_rate": 1.3667038795801741e-06,
+      "loss": 0.5333,
+      "step": 14717
+    },
+    {
+      "epoch": 3.90720828355237,
+      "grad_norm": 0.467194376278856,
+      "learning_rate": 1.3663926959403256e-06,
+      "loss": 0.551,
+      "step": 14718
+    },
+    {
+      "epoch": 3.9074737820257535,
+      "grad_norm": 0.4707464622147138,
+      "learning_rate": 1.366081534408594e-06,
+      "loss": 0.5506,
+      "step": 14719
+    },
+    {
+      "epoch": 3.907739280499137,
+      "grad_norm": 0.4633981052191705,
+      "learning_rate": 1.36577039499105e-06,
+      "loss": 0.5324,
+      "step": 14720
+    },
+    {
+      "epoch": 3.9080047789725207,
+      "grad_norm": 0.49972504157934,
+      "learning_rate": 1.3654592776937592e-06,
+      "loss": 0.5666,
+      "step": 14721
+    },
+    {
+      "epoch": 3.908270277445905,
+      "grad_norm": 0.48421220256866543,
+      "learning_rate": 1.3651481825227914e-06,
+      "loss": 0.5377,
+      "step": 14722
+    },
+    {
+      "epoch": 3.9085357759192885,
+      "grad_norm": 0.46252709166324507,
+      "learning_rate": 1.3648371094842128e-06,
+      "loss": 0.5494,
+      "step": 14723
+    },
+    {
+      "epoch": 3.908801274392672,
+      "grad_norm": 0.4655571775981846,
+      "learning_rate": 1.3645260585840886e-06,
+      "loss": 0.5442,
+      "step": 14724
+    },
+    {
+      "epoch": 3.9090667728660558,
+      "grad_norm": 0.45749525360659066,
+      "learning_rate": 1.3642150298284876e-06,
+      "loss": 0.5408,
+      "step": 14725
+    },
+    {
+      "epoch": 3.90933227133944,
+      "grad_norm": 0.47346948334622846,
+      "learning_rate": 1.3639040232234734e-06,
+      "loss": 0.547,
+      "step": 14726
+    },
+    {
+      "epoch": 3.9095977698128235,
+      "grad_norm": 0.45563439821141155,
+      "learning_rate": 1.3635930387751134e-06,
+      "loss": 0.5444,
+      "step": 14727
+    },
+    {
+      "epoch": 3.909863268286207,
+      "grad_norm": 0.47627641554470057,
+      "learning_rate": 1.3632820764894715e-06,
+      "loss": 0.5614,
+      "step": 14728
+    },
+    {
+      "epoch": 3.910128766759591,
+      "grad_norm": 0.46449916185390694,
+      "learning_rate": 1.362971136372611e-06,
+      "loss": 0.5546,
+      "step": 14729
+    },
+    {
+      "epoch": 3.910394265232975,
+      "grad_norm": 0.47105299584319793,
+      "learning_rate": 1.3626602184305987e-06,
+      "loss": 0.5421,
+      "step": 14730
+    },
+    {
+      "epoch": 3.9106597637063585,
+      "grad_norm": 0.46411543357985235,
+      "learning_rate": 1.3623493226694955e-06,
+      "loss": 0.542,
+      "step": 14731
+    },
+    {
+      "epoch": 3.9109252621797426,
+      "grad_norm": 0.4534898392466617,
+      "learning_rate": 1.3620384490953669e-06,
+      "loss": 0.5155,
+      "step": 14732
+    },
+    {
+      "epoch": 3.911190760653126,
+      "grad_norm": 0.4691154059876752,
+      "learning_rate": 1.3617275977142742e-06,
+      "loss": 0.5414,
+      "step": 14733
+    },
+    {
+      "epoch": 3.91145625912651,
+      "grad_norm": 0.47389648575515175,
+      "learning_rate": 1.3614167685322821e-06,
+      "loss": 0.5665,
+      "step": 14734
+    },
+    {
+      "epoch": 3.911721757599894,
+      "grad_norm": 0.47715441810130527,
+      "learning_rate": 1.3611059615554487e-06,
+      "loss": 0.565,
+      "step": 14735
+    },
+    {
+      "epoch": 3.9119872560732776,
+      "grad_norm": 0.470702503721107,
+      "learning_rate": 1.3607951767898386e-06,
+      "loss": 0.5513,
+      "step": 14736
+    },
+    {
+      "epoch": 3.912252754546661,
+      "grad_norm": 0.4606711704956551,
+      "learning_rate": 1.3604844142415108e-06,
+      "loss": 0.5188,
+      "step": 14737
+    },
+    {
+      "epoch": 3.9125182530200453,
+      "grad_norm": 0.4805378115092284,
+      "learning_rate": 1.3601736739165278e-06,
+      "loss": 0.5436,
+      "step": 14738
+    },
+    {
+      "epoch": 3.912783751493429,
+      "grad_norm": 0.4682016375021728,
+      "learning_rate": 1.359862955820948e-06,
+      "loss": 0.5516,
+      "step": 14739
+    },
+    {
+      "epoch": 3.9130492499668126,
+      "grad_norm": 0.46870657749539757,
+      "learning_rate": 1.3595522599608333e-06,
+      "loss": 0.5714,
+      "step": 14740
+    },
+    {
+      "epoch": 3.9133147484401967,
+      "grad_norm": 0.47012056161416405,
+      "learning_rate": 1.3592415863422415e-06,
+      "loss": 0.5523,
+      "step": 14741
+    },
+    {
+      "epoch": 3.9135802469135803,
+      "grad_norm": 0.4606365008269365,
+      "learning_rate": 1.358930934971231e-06,
+      "loss": 0.5331,
+      "step": 14742
+    },
+    {
+      "epoch": 3.913845745386964,
+      "grad_norm": 0.45945994973378745,
+      "learning_rate": 1.3586203058538622e-06,
+      "loss": 0.5457,
+      "step": 14743
+    },
+    {
+      "epoch": 3.914111243860348,
+      "grad_norm": 0.4517789628173039,
+      "learning_rate": 1.3583096989961908e-06,
+      "loss": 0.5674,
+      "step": 14744
+    },
+    {
+      "epoch": 3.9143767423337317,
+      "grad_norm": 0.4793254762824073,
+      "learning_rate": 1.3579991144042775e-06,
+      "loss": 0.5808,
+      "step": 14745
+    },
+    {
+      "epoch": 3.9146422408071153,
+      "grad_norm": 0.4649389067770687,
+      "learning_rate": 1.357688552084176e-06,
+      "loss": 0.5673,
+      "step": 14746
+    },
+    {
+      "epoch": 3.9149077392804994,
+      "grad_norm": 0.45690113761182305,
+      "learning_rate": 1.3573780120419455e-06,
+      "loss": 0.5459,
+      "step": 14747
+    },
+    {
+      "epoch": 3.915173237753883,
+      "grad_norm": 0.4545940354472747,
+      "learning_rate": 1.3570674942836401e-06,
+      "loss": 0.5472,
+      "step": 14748
+    },
+    {
+      "epoch": 3.9154387362272667,
+      "grad_norm": 0.4878460082851874,
+      "learning_rate": 1.3567569988153172e-06,
+      "loss": 0.5089,
+      "step": 14749
+    },
+    {
+      "epoch": 3.9157042347006503,
+      "grad_norm": 0.44989938921248396,
+      "learning_rate": 1.3564465256430328e-06,
+      "loss": 0.5473,
+      "step": 14750
+    },
+    {
+      "epoch": 3.9159697331740344,
+      "grad_norm": 0.4681212086887664,
+      "learning_rate": 1.3561360747728411e-06,
+      "loss": 0.5552,
+      "step": 14751
+    },
+    {
+      "epoch": 3.916235231647418,
+      "grad_norm": 0.47618175085230485,
+      "learning_rate": 1.3558256462107965e-06,
+      "loss": 0.5645,
+      "step": 14752
+    },
+    {
+      "epoch": 3.9165007301208017,
+      "grad_norm": 0.46681277787031755,
+      "learning_rate": 1.3555152399629523e-06,
+      "loss": 0.5805,
+      "step": 14753
+    },
+    {
+      "epoch": 3.9167662285941853,
+      "grad_norm": 0.475974934513959,
+      "learning_rate": 1.3552048560353643e-06,
+      "loss": 0.5596,
+      "step": 14754
+    },
+    {
+      "epoch": 3.9170317270675694,
+      "grad_norm": 0.46687103599044416,
+      "learning_rate": 1.3548944944340833e-06,
+      "loss": 0.5534,
+      "step": 14755
+    },
+    {
+      "epoch": 3.917297225540953,
+      "grad_norm": 0.4638734034082637,
+      "learning_rate": 1.3545841551651646e-06,
+      "loss": 0.5303,
+      "step": 14756
+    },
+    {
+      "epoch": 3.9175627240143367,
+      "grad_norm": 0.463631788461443,
+      "learning_rate": 1.3542738382346593e-06,
+      "loss": 0.5287,
+      "step": 14757
+    },
+    {
+      "epoch": 3.917828222487721,
+      "grad_norm": 0.45519936138630906,
+      "learning_rate": 1.353963543648618e-06,
+      "loss": 0.5065,
+      "step": 14758
+    },
+    {
+      "epoch": 3.9180937209611044,
+      "grad_norm": 0.4549791289533537,
+      "learning_rate": 1.353653271413095e-06,
+      "loss": 0.5528,
+      "step": 14759
+    },
+    {
+      "epoch": 3.918359219434488,
+      "grad_norm": 0.47460011697691484,
+      "learning_rate": 1.3533430215341387e-06,
+      "loss": 0.5686,
+      "step": 14760
+    },
+    {
+      "epoch": 3.918624717907872,
+      "grad_norm": 0.4656560956477604,
+      "learning_rate": 1.353032794017802e-06,
+      "loss": 0.5851,
+      "step": 14761
+    },
+    {
+      "epoch": 3.918890216381256,
+      "grad_norm": 0.47292410819616754,
+      "learning_rate": 1.3527225888701334e-06,
+      "loss": 0.5917,
+      "step": 14762
+    },
+    {
+      "epoch": 3.9191557148546394,
+      "grad_norm": 0.47564106618319935,
+      "learning_rate": 1.352412406097185e-06,
+      "loss": 0.5342,
+      "step": 14763
+    },
+    {
+      "epoch": 3.9194212133280235,
+      "grad_norm": 0.4676455558627603,
+      "learning_rate": 1.3521022457050026e-06,
+      "loss": 0.5223,
+      "step": 14764
+    },
+    {
+      "epoch": 3.919686711801407,
+      "grad_norm": 0.4618649659818899,
+      "learning_rate": 1.3517921076996375e-06,
+      "loss": 0.5328,
+      "step": 14765
+    },
+    {
+      "epoch": 3.919952210274791,
+      "grad_norm": 0.47475299638691565,
+      "learning_rate": 1.3514819920871372e-06,
+      "loss": 0.5189,
+      "step": 14766
+    },
+    {
+      "epoch": 3.920217708748175,
+      "grad_norm": 0.4695573020525525,
+      "learning_rate": 1.3511718988735508e-06,
+      "loss": 0.5367,
+      "step": 14767
+    },
+    {
+      "epoch": 3.9204832072215585,
+      "grad_norm": 0.47557159012312633,
+      "learning_rate": 1.3508618280649256e-06,
+      "loss": 0.5404,
+      "step": 14768
+    },
+    {
+      "epoch": 3.920748705694942,
+      "grad_norm": 0.45567423353409175,
+      "learning_rate": 1.3505517796673069e-06,
+      "loss": 0.5524,
+      "step": 14769
+    },
+    {
+      "epoch": 3.9210142041683262,
+      "grad_norm": 0.4498274454675176,
+      "learning_rate": 1.350241753686744e-06,
+      "loss": 0.5504,
+      "step": 14770
+    },
+    {
+      "epoch": 3.92127970264171,
+      "grad_norm": 0.4709544892520749,
+      "learning_rate": 1.3499317501292807e-06,
+      "loss": 0.5184,
+      "step": 14771
+    },
+    {
+      "epoch": 3.9215452011150935,
+      "grad_norm": 0.46477447907187447,
+      "learning_rate": 1.3496217690009655e-06,
+      "loss": 0.5624,
+      "step": 14772
+    },
+    {
+      "epoch": 3.9218106995884776,
+      "grad_norm": 0.4641581477946725,
+      "learning_rate": 1.3493118103078412e-06,
+      "loss": 0.5492,
+      "step": 14773
+    },
+    {
+      "epoch": 3.9220761980618613,
+      "grad_norm": 0.4661662264609392,
+      "learning_rate": 1.3490018740559551e-06,
+      "loss": 0.5399,
+      "step": 14774
+    },
+    {
+      "epoch": 3.922341696535245,
+      "grad_norm": 0.4507233777564372,
+      "learning_rate": 1.3486919602513504e-06,
+      "loss": 0.5526,
+      "step": 14775
+    },
+    {
+      "epoch": 3.922607195008629,
+      "grad_norm": 0.45844797168259865,
+      "learning_rate": 1.3483820689000706e-06,
+      "loss": 0.5573,
+      "step": 14776
+    },
+    {
+      "epoch": 3.9228726934820126,
+      "grad_norm": 0.46032682770825356,
+      "learning_rate": 1.3480722000081612e-06,
+      "loss": 0.5215,
+      "step": 14777
+    },
+    {
+      "epoch": 3.9231381919553963,
+      "grad_norm": 0.47644789528767656,
+      "learning_rate": 1.347762353581663e-06,
+      "loss": 0.5343,
+      "step": 14778
+    },
+    {
+      "epoch": 3.92340369042878,
+      "grad_norm": 0.4741474173426853,
+      "learning_rate": 1.347452529626621e-06,
+      "loss": 0.5337,
+      "step": 14779
+    },
+    {
+      "epoch": 3.9236691889021635,
+      "grad_norm": 0.4470769954143039,
+      "learning_rate": 1.3471427281490767e-06,
+      "loss": 0.5311,
+      "step": 14780
+    },
+    {
+      "epoch": 3.9239346873755476,
+      "grad_norm": 0.47366138218918785,
+      "learning_rate": 1.3468329491550722e-06,
+      "loss": 0.556,
+      "step": 14781
+    },
+    {
+      "epoch": 3.9242001858489313,
+      "grad_norm": 0.4716855584503836,
+      "learning_rate": 1.3465231926506473e-06,
+      "loss": 0.5453,
+      "step": 14782
+    },
+    {
+      "epoch": 3.924465684322315,
+      "grad_norm": 0.45839694715983687,
+      "learning_rate": 1.3462134586418454e-06,
+      "loss": 0.5579,
+      "step": 14783
+    },
+    {
+      "epoch": 3.924731182795699,
+      "grad_norm": 0.4626636308750909,
+      "learning_rate": 1.3459037471347054e-06,
+      "loss": 0.5474,
+      "step": 14784
+    },
+    {
+      "epoch": 3.9249966812690826,
+      "grad_norm": 0.46704971921087174,
+      "learning_rate": 1.3455940581352683e-06,
+      "loss": 0.5748,
+      "step": 14785
+    },
+    {
+      "epoch": 3.9252621797424663,
+      "grad_norm": 0.4697042045144857,
+      "learning_rate": 1.345284391649574e-06,
+      "loss": 0.5486,
+      "step": 14786
+    },
+    {
+      "epoch": 3.9255276782158504,
+      "grad_norm": 0.46126506591276045,
+      "learning_rate": 1.3449747476836602e-06,
+      "loss": 0.5347,
+      "step": 14787
+    },
+    {
+      "epoch": 3.925793176689234,
+      "grad_norm": 0.4645857555559977,
+      "learning_rate": 1.3446651262435679e-06,
+      "loss": 0.5461,
+      "step": 14788
+    },
+    {
+      "epoch": 3.9260586751626176,
+      "grad_norm": 0.45783179615857134,
+      "learning_rate": 1.344355527335333e-06,
+      "loss": 0.5377,
+      "step": 14789
+    },
+    {
+      "epoch": 3.9263241736360017,
+      "grad_norm": 0.4597795593447878,
+      "learning_rate": 1.3440459509649967e-06,
+      "loss": 0.5544,
+      "step": 14790
+    },
+    {
+      "epoch": 3.9265896721093854,
+      "grad_norm": 0.49152681822810573,
+      "learning_rate": 1.3437363971385938e-06,
+      "loss": 0.5651,
+      "step": 14791
+    },
+    {
+      "epoch": 3.926855170582769,
+      "grad_norm": 0.4806215672257861,
+      "learning_rate": 1.3434268658621625e-06,
+      "loss": 0.5487,
+      "step": 14792
+    },
+    {
+      "epoch": 3.927120669056153,
+      "grad_norm": 0.47156812211848087,
+      "learning_rate": 1.3431173571417381e-06,
+      "loss": 0.5393,
+      "step": 14793
+    },
+    {
+      "epoch": 3.9273861675295367,
+      "grad_norm": 0.4699178651649868,
+      "learning_rate": 1.3428078709833587e-06,
+      "loss": 0.5408,
+      "step": 14794
+    },
+    {
+      "epoch": 3.9276516660029204,
+      "grad_norm": 0.45337841094996384,
+      "learning_rate": 1.3424984073930581e-06,
+      "loss": 0.5187,
+      "step": 14795
+    },
+    {
+      "epoch": 3.9279171644763045,
+      "grad_norm": 0.4676407777429846,
+      "learning_rate": 1.3421889663768725e-06,
+      "loss": 0.5569,
+      "step": 14796
+    },
+    {
+      "epoch": 3.928182662949688,
+      "grad_norm": 0.47774782040438807,
+      "learning_rate": 1.3418795479408391e-06,
+      "loss": 0.538,
+      "step": 14797
+    },
+    {
+      "epoch": 3.9284481614230717,
+      "grad_norm": 0.4696848613453608,
+      "learning_rate": 1.341570152090988e-06,
+      "loss": 0.5259,
+      "step": 14798
+    },
+    {
+      "epoch": 3.928713659896456,
+      "grad_norm": 0.47483248781232507,
+      "learning_rate": 1.3412607788333561e-06,
+      "loss": 0.5616,
+      "step": 14799
+    },
+    {
+      "epoch": 3.9289791583698395,
+      "grad_norm": 0.47468798862305933,
+      "learning_rate": 1.3409514281739752e-06,
+      "loss": 0.5367,
+      "step": 14800
+    },
+    {
+      "epoch": 3.929244656843223,
+      "grad_norm": 0.4693627597834553,
+      "learning_rate": 1.3406421001188804e-06,
+      "loss": 0.5768,
+      "step": 14801
+    },
+    {
+      "epoch": 3.929510155316607,
+      "grad_norm": 0.45419436746230885,
+      "learning_rate": 1.3403327946741023e-06,
+      "loss": 0.5147,
+      "step": 14802
+    },
+    {
+      "epoch": 3.929775653789991,
+      "grad_norm": 0.4703471378904216,
+      "learning_rate": 1.3400235118456747e-06,
+      "loss": 0.5283,
+      "step": 14803
+    },
+    {
+      "epoch": 3.9300411522633745,
+      "grad_norm": 0.47465765235788776,
+      "learning_rate": 1.3397142516396289e-06,
+      "loss": 0.5827,
+      "step": 14804
+    },
+    {
+      "epoch": 3.930306650736758,
+      "grad_norm": 0.45880215706656197,
+      "learning_rate": 1.3394050140619952e-06,
+      "loss": 0.5509,
+      "step": 14805
+    },
+    {
+      "epoch": 3.930572149210142,
+      "grad_norm": 0.47061397682799333,
+      "learning_rate": 1.3390957991188065e-06,
+      "loss": 0.5093,
+      "step": 14806
+    },
+    {
+      "epoch": 3.930837647683526,
+      "grad_norm": 0.45233954420363803,
+      "learning_rate": 1.3387866068160907e-06,
+      "loss": 0.5772,
+      "step": 14807
+    },
+    {
+      "epoch": 3.9311031461569095,
+      "grad_norm": 0.4668690771195901,
+      "learning_rate": 1.3384774371598802e-06,
+      "loss": 0.5532,
+      "step": 14808
+    },
+    {
+      "epoch": 3.931368644630293,
+      "grad_norm": 0.4758249101545327,
+      "learning_rate": 1.3381682901562038e-06,
+      "loss": 0.5659,
+      "step": 14809
+    },
+    {
+      "epoch": 3.931634143103677,
+      "grad_norm": 0.4665533378575868,
+      "learning_rate": 1.3378591658110907e-06,
+      "loss": 0.5389,
+      "step": 14810
+    },
+    {
+      "epoch": 3.931899641577061,
+      "grad_norm": 0.48120507322810896,
+      "learning_rate": 1.3375500641305678e-06,
+      "loss": 0.5306,
+      "step": 14811
+    },
+    {
+      "epoch": 3.9321651400504445,
+      "grad_norm": 0.46340656197819297,
+      "learning_rate": 1.3372409851206658e-06,
+      "loss": 0.5434,
+      "step": 14812
+    },
+    {
+      "epoch": 3.9324306385238286,
+      "grad_norm": 0.46759445951567835,
+      "learning_rate": 1.3369319287874106e-06,
+      "loss": 0.5323,
+      "step": 14813
+    },
+    {
+      "epoch": 3.932696136997212,
+      "grad_norm": 0.4651530615615993,
+      "learning_rate": 1.3366228951368316e-06,
+      "loss": 0.5463,
+      "step": 14814
+    },
+    {
+      "epoch": 3.932961635470596,
+      "grad_norm": 0.4687370250929926,
+      "learning_rate": 1.3363138841749546e-06,
+      "loss": 0.5649,
+      "step": 14815
+    },
+    {
+      "epoch": 3.93322713394398,
+      "grad_norm": 0.46066991619179465,
+      "learning_rate": 1.3360048959078049e-06,
+      "loss": 0.5304,
+      "step": 14816
+    },
+    {
+      "epoch": 3.9334926324173636,
+      "grad_norm": 0.47275272449684125,
+      "learning_rate": 1.3356959303414109e-06,
+      "loss": 0.5398,
+      "step": 14817
+    },
+    {
+      "epoch": 3.933758130890747,
+      "grad_norm": 0.47817314307800457,
+      "learning_rate": 1.3353869874817956e-06,
+      "loss": 0.5679,
+      "step": 14818
+    },
+    {
+      "epoch": 3.9340236293641313,
+      "grad_norm": 0.4664786438407297,
+      "learning_rate": 1.3350780673349866e-06,
+      "loss": 0.5856,
+      "step": 14819
+    },
+    {
+      "epoch": 3.934289127837515,
+      "grad_norm": 0.4720284762677622,
+      "learning_rate": 1.3347691699070077e-06,
+      "loss": 0.5802,
+      "step": 14820
+    },
+    {
+      "epoch": 3.9345546263108986,
+      "grad_norm": 0.48540750881622635,
+      "learning_rate": 1.334460295203882e-06,
+      "loss": 0.5127,
+      "step": 14821
+    },
+    {
+      "epoch": 3.9348201247842827,
+      "grad_norm": 0.4601365605090677,
+      "learning_rate": 1.3341514432316351e-06,
+      "loss": 0.5423,
+      "step": 14822
+    },
+    {
+      "epoch": 3.9350856232576663,
+      "grad_norm": 0.4896311309951293,
+      "learning_rate": 1.3338426139962888e-06,
+      "loss": 0.5681,
+      "step": 14823
+    },
+    {
+      "epoch": 3.93535112173105,
+      "grad_norm": 0.4595978496339517,
+      "learning_rate": 1.3335338075038672e-06,
+      "loss": 0.5158,
+      "step": 14824
+    },
+    {
+      "epoch": 3.935616620204434,
+      "grad_norm": 0.45037210659350246,
+      "learning_rate": 1.333225023760392e-06,
+      "loss": 0.5377,
+      "step": 14825
+    },
+    {
+      "epoch": 3.9358821186778177,
+      "grad_norm": 0.4667936830525237,
+      "learning_rate": 1.3329162627718872e-06,
+      "loss": 0.564,
+      "step": 14826
+    },
+    {
+      "epoch": 3.9361476171512013,
+      "grad_norm": 0.47828288652236456,
+      "learning_rate": 1.3326075245443709e-06,
+      "loss": 0.542,
+      "step": 14827
+    },
+    {
+      "epoch": 3.9364131156245854,
+      "grad_norm": 0.4642667444604403,
+      "learning_rate": 1.3322988090838672e-06,
+      "loss": 0.5373,
+      "step": 14828
+    },
+    {
+      "epoch": 3.936678614097969,
+      "grad_norm": 0.47312198338014666,
+      "learning_rate": 1.331990116396395e-06,
+      "loss": 0.518,
+      "step": 14829
+    },
+    {
+      "epoch": 3.9369441125713527,
+      "grad_norm": 0.4551803703513252,
+      "learning_rate": 1.3316814464879756e-06,
+      "loss": 0.5149,
+      "step": 14830
+    },
+    {
+      "epoch": 3.9372096110447368,
+      "grad_norm": 0.4794636626720055,
+      "learning_rate": 1.331372799364628e-06,
+      "loss": 0.5473,
+      "step": 14831
+    },
+    {
+      "epoch": 3.9374751095181204,
+      "grad_norm": 0.4654946039303813,
+      "learning_rate": 1.3310641750323732e-06,
+      "loss": 0.5383,
+      "step": 14832
+    },
+    {
+      "epoch": 3.937740607991504,
+      "grad_norm": 0.4664094002493494,
+      "learning_rate": 1.3307555734972288e-06,
+      "loss": 0.5724,
+      "step": 14833
+    },
+    {
+      "epoch": 3.9380061064648877,
+      "grad_norm": 0.4765294250005343,
+      "learning_rate": 1.3304469947652127e-06,
+      "loss": 0.5212,
+      "step": 14834
+    },
+    {
+      "epoch": 3.9382716049382713,
+      "grad_norm": 0.45930543403932333,
+      "learning_rate": 1.330138438842345e-06,
+      "loss": 0.5675,
+      "step": 14835
+    },
+    {
+      "epoch": 3.9385371034116554,
+      "grad_norm": 0.4677371969994075,
+      "learning_rate": 1.3298299057346407e-06,
+      "loss": 0.5308,
+      "step": 14836
+    },
+    {
+      "epoch": 3.938802601885039,
+      "grad_norm": 0.48130753586316255,
+      "learning_rate": 1.3295213954481193e-06,
+      "loss": 0.5784,
+      "step": 14837
+    },
+    {
+      "epoch": 3.9390681003584227,
+      "grad_norm": 0.4631457997059472,
+      "learning_rate": 1.3292129079887967e-06,
+      "loss": 0.5505,
+      "step": 14838
+    },
+    {
+      "epoch": 3.939333598831807,
+      "grad_norm": 0.46810729642748544,
+      "learning_rate": 1.3289044433626885e-06,
+      "loss": 0.549,
+      "step": 14839
+    },
+    {
+      "epoch": 3.9395990973051904,
+      "grad_norm": 0.497236890113771,
+      "learning_rate": 1.3285960015758103e-06,
+      "loss": 0.5839,
+      "step": 14840
+    },
+    {
+      "epoch": 3.939864595778574,
+      "grad_norm": 0.47780628487422905,
+      "learning_rate": 1.3282875826341786e-06,
+      "loss": 0.5642,
+      "step": 14841
+    },
+    {
+      "epoch": 3.940130094251958,
+      "grad_norm": 0.47180685368458763,
+      "learning_rate": 1.3279791865438082e-06,
+      "loss": 0.5336,
+      "step": 14842
+    },
+    {
+      "epoch": 3.940395592725342,
+      "grad_norm": 0.46569956576017574,
+      "learning_rate": 1.3276708133107136e-06,
+      "loss": 0.5616,
+      "step": 14843
+    },
+    {
+      "epoch": 3.9406610911987254,
+      "grad_norm": 0.48243148254923435,
+      "learning_rate": 1.3273624629409081e-06,
+      "loss": 0.5652,
+      "step": 14844
+    },
+    {
+      "epoch": 3.9409265896721095,
+      "grad_norm": 0.4682108082491541,
+      "learning_rate": 1.3270541354404054e-06,
+      "loss": 0.5018,
+      "step": 14845
+    },
+    {
+      "epoch": 3.941192088145493,
+      "grad_norm": 0.4657765859010105,
+      "learning_rate": 1.3267458308152192e-06,
+      "loss": 0.5737,
+      "step": 14846
+    },
+    {
+      "epoch": 3.941457586618877,
+      "grad_norm": 0.47517881991646244,
+      "learning_rate": 1.3264375490713615e-06,
+      "loss": 0.5452,
+      "step": 14847
+    },
+    {
+      "epoch": 3.941723085092261,
+      "grad_norm": 0.4720445690260556,
+      "learning_rate": 1.3261292902148457e-06,
+      "loss": 0.5613,
+      "step": 14848
+    },
+    {
+      "epoch": 3.9419885835656445,
+      "grad_norm": 0.46282798434916744,
+      "learning_rate": 1.3258210542516831e-06,
+      "loss": 0.529,
+      "step": 14849
+    },
+    {
+      "epoch": 3.942254082039028,
+      "grad_norm": 0.45637987901777416,
+      "learning_rate": 1.3255128411878835e-06,
+      "loss": 0.5737,
+      "step": 14850
+    },
+    {
+      "epoch": 3.9425195805124122,
+      "grad_norm": 0.46718630484378726,
+      "learning_rate": 1.3252046510294603e-06,
+      "loss": 0.5242,
+      "step": 14851
+    },
+    {
+      "epoch": 3.942785078985796,
+      "grad_norm": 0.45520849313014605,
+      "learning_rate": 1.3248964837824218e-06,
+      "loss": 0.5079,
+      "step": 14852
+    },
+    {
+      "epoch": 3.9430505774591795,
+      "grad_norm": 0.46631529817604855,
+      "learning_rate": 1.3245883394527797e-06,
+      "loss": 0.5518,
+      "step": 14853
+    },
+    {
+      "epoch": 3.9433160759325636,
+      "grad_norm": 0.468082249143158,
+      "learning_rate": 1.3242802180465425e-06,
+      "loss": 0.5505,
+      "step": 14854
+    },
+    {
+      "epoch": 3.9435815744059473,
+      "grad_norm": 0.4736161409528284,
+      "learning_rate": 1.3239721195697215e-06,
+      "loss": 0.5504,
+      "step": 14855
+    },
+    {
+      "epoch": 3.943847072879331,
+      "grad_norm": 0.46790978106992953,
+      "learning_rate": 1.3236640440283211e-06,
+      "loss": 0.5733,
+      "step": 14856
+    },
+    {
+      "epoch": 3.944112571352715,
+      "grad_norm": 0.4571746779654453,
+      "learning_rate": 1.3233559914283533e-06,
+      "loss": 0.5464,
+      "step": 14857
+    },
+    {
+      "epoch": 3.9443780698260986,
+      "grad_norm": 0.47775425692843776,
+      "learning_rate": 1.3230479617758238e-06,
+      "loss": 0.591,
+      "step": 14858
+    },
+    {
+      "epoch": 3.9446435682994823,
+      "grad_norm": 0.46886693288058695,
+      "learning_rate": 1.3227399550767417e-06,
+      "loss": 0.5693,
+      "step": 14859
+    },
+    {
+      "epoch": 3.944909066772866,
+      "grad_norm": 0.46013332659341644,
+      "learning_rate": 1.3224319713371118e-06,
+      "loss": 0.5235,
+      "step": 14860
+    },
+    {
+      "epoch": 3.94517456524625,
+      "grad_norm": 0.47029919247162005,
+      "learning_rate": 1.3221240105629424e-06,
+      "loss": 0.5676,
+      "step": 14861
+    },
+    {
+      "epoch": 3.9454400637196336,
+      "grad_norm": 0.44522102222662774,
+      "learning_rate": 1.3218160727602392e-06,
+      "loss": 0.5397,
+      "step": 14862
+    },
+    {
+      "epoch": 3.9457055621930173,
+      "grad_norm": 0.47232099266320243,
+      "learning_rate": 1.3215081579350058e-06,
+      "loss": 0.5517,
+      "step": 14863
+    },
+    {
+      "epoch": 3.945971060666401,
+      "grad_norm": 0.4769886166124681,
+      "learning_rate": 1.32120026609325e-06,
+      "loss": 0.5228,
+      "step": 14864
+    },
+    {
+      "epoch": 3.946236559139785,
+      "grad_norm": 0.45065780731769306,
+      "learning_rate": 1.3208923972409743e-06,
+      "loss": 0.5311,
+      "step": 14865
+    },
+    {
+      "epoch": 3.9465020576131686,
+      "grad_norm": 0.46535236450105755,
+      "learning_rate": 1.3205845513841848e-06,
+      "loss": 0.5565,
+      "step": 14866
+    },
+    {
+      "epoch": 3.9467675560865523,
+      "grad_norm": 0.4617883519259185,
+      "learning_rate": 1.3202767285288841e-06,
+      "loss": 0.5249,
+      "step": 14867
+    },
+    {
+      "epoch": 3.9470330545599364,
+      "grad_norm": 0.4476884485491734,
+      "learning_rate": 1.3199689286810746e-06,
+      "loss": 0.5244,
+      "step": 14868
+    },
+    {
+      "epoch": 3.94729855303332,
+      "grad_norm": 0.4544271027766419,
+      "learning_rate": 1.3196611518467613e-06,
+      "loss": 0.5102,
+      "step": 14869
+    },
+    {
+      "epoch": 3.9475640515067036,
+      "grad_norm": 0.4662108009250808,
+      "learning_rate": 1.3193533980319445e-06,
+      "loss": 0.5111,
+      "step": 14870
+    },
+    {
+      "epoch": 3.9478295499800877,
+      "grad_norm": 0.4597279757957401,
+      "learning_rate": 1.3190456672426283e-06,
+      "loss": 0.5281,
+      "step": 14871
+    },
+    {
+      "epoch": 3.9480950484534714,
+      "grad_norm": 0.48514579912030736,
+      "learning_rate": 1.3187379594848132e-06,
+      "loss": 0.547,
+      "step": 14872
+    },
+    {
+      "epoch": 3.948360546926855,
+      "grad_norm": 0.46871236579097586,
+      "learning_rate": 1.3184302747644998e-06,
+      "loss": 0.5188,
+      "step": 14873
+    },
+    {
+      "epoch": 3.948626045400239,
+      "grad_norm": 0.4611095522184463,
+      "learning_rate": 1.318122613087688e-06,
+      "loss": 0.5167,
+      "step": 14874
+    },
+    {
+      "epoch": 3.9488915438736227,
+      "grad_norm": 0.4490969768247677,
+      "learning_rate": 1.3178149744603802e-06,
+      "loss": 0.5404,
+      "step": 14875
+    },
+    {
+      "epoch": 3.9491570423470064,
+      "grad_norm": 0.4764462021797949,
+      "learning_rate": 1.3175073588885737e-06,
+      "loss": 0.5569,
+      "step": 14876
+    },
+    {
+      "epoch": 3.9494225408203905,
+      "grad_norm": 0.4723204586935236,
+      "learning_rate": 1.31719976637827e-06,
+      "loss": 0.5439,
+      "step": 14877
+    },
+    {
+      "epoch": 3.949688039293774,
+      "grad_norm": 0.45915051903743664,
+      "learning_rate": 1.316892196935467e-06,
+      "loss": 0.5253,
+      "step": 14878
+    },
+    {
+      "epoch": 3.9499535377671577,
+      "grad_norm": 0.4696321657211413,
+      "learning_rate": 1.3165846505661617e-06,
+      "loss": 0.5399,
+      "step": 14879
+    },
+    {
+      "epoch": 3.950219036240542,
+      "grad_norm": 0.46455405135823113,
+      "learning_rate": 1.3162771272763542e-06,
+      "loss": 0.5546,
+      "step": 14880
+    },
+    {
+      "epoch": 3.9504845347139255,
+      "grad_norm": 0.48124207104601524,
+      "learning_rate": 1.3159696270720402e-06,
+      "loss": 0.5493,
+      "step": 14881
+    },
+    {
+      "epoch": 3.950750033187309,
+      "grad_norm": 0.46823800725655346,
+      "learning_rate": 1.3156621499592182e-06,
+      "loss": 0.5516,
+      "step": 14882
+    },
+    {
+      "epoch": 3.951015531660693,
+      "grad_norm": 0.4669344561071254,
+      "learning_rate": 1.3153546959438831e-06,
+      "loss": 0.5164,
+      "step": 14883
+    },
+    {
+      "epoch": 3.951281030134077,
+      "grad_norm": 0.46787232763637854,
+      "learning_rate": 1.3150472650320339e-06,
+      "loss": 0.5513,
+      "step": 14884
+    },
+    {
+      "epoch": 3.9515465286074605,
+      "grad_norm": 0.47087851483888055,
+      "learning_rate": 1.3147398572296628e-06,
+      "loss": 0.5198,
+      "step": 14885
+    },
+    {
+      "epoch": 3.9518120270808446,
+      "grad_norm": 0.4630743249748499,
+      "learning_rate": 1.314432472542766e-06,
+      "loss": 0.5308,
+      "step": 14886
+    },
+    {
+      "epoch": 3.952077525554228,
+      "grad_norm": 0.4823119321843204,
+      "learning_rate": 1.3141251109773395e-06,
+      "loss": 0.5175,
+      "step": 14887
+    },
+    {
+      "epoch": 3.952343024027612,
+      "grad_norm": 0.473496364292403,
+      "learning_rate": 1.3138177725393763e-06,
+      "loss": 0.568,
+      "step": 14888
+    },
+    {
+      "epoch": 3.9526085225009955,
+      "grad_norm": 0.46097811089796303,
+      "learning_rate": 1.3135104572348728e-06,
+      "loss": 0.5209,
+      "step": 14889
+    },
+    {
+      "epoch": 3.952874020974379,
+      "grad_norm": 0.47341156683093305,
+      "learning_rate": 1.3132031650698185e-06,
+      "loss": 0.5668,
+      "step": 14890
+    },
+    {
+      "epoch": 3.953139519447763,
+      "grad_norm": 0.4823496185058247,
+      "learning_rate": 1.3128958960502093e-06,
+      "loss": 0.5578,
+      "step": 14891
+    },
+    {
+      "epoch": 3.953405017921147,
+      "grad_norm": 0.46678230106189345,
+      "learning_rate": 1.312588650182035e-06,
+      "loss": 0.5444,
+      "step": 14892
+    },
+    {
+      "epoch": 3.9536705163945305,
+      "grad_norm": 0.47024215035505973,
+      "learning_rate": 1.3122814274712908e-06,
+      "loss": 0.541,
+      "step": 14893
+    },
+    {
+      "epoch": 3.9539360148679146,
+      "grad_norm": 0.47079368322296195,
+      "learning_rate": 1.3119742279239659e-06,
+      "loss": 0.5531,
+      "step": 14894
+    },
+    {
+      "epoch": 3.954201513341298,
+      "grad_norm": 0.4667568399265366,
+      "learning_rate": 1.3116670515460534e-06,
+      "loss": 0.5166,
+      "step": 14895
+    },
+    {
+      "epoch": 3.954467011814682,
+      "grad_norm": 0.4520028673063646,
+      "learning_rate": 1.3113598983435427e-06,
+      "loss": 0.5665,
+      "step": 14896
+    },
+    {
+      "epoch": 3.954732510288066,
+      "grad_norm": 0.4652620205820074,
+      "learning_rate": 1.3110527683224233e-06,
+      "loss": 0.5506,
+      "step": 14897
+    },
+    {
+      "epoch": 3.9549980087614496,
+      "grad_norm": 0.4703129809452641,
+      "learning_rate": 1.310745661488687e-06,
+      "loss": 0.5401,
+      "step": 14898
+    },
+    {
+      "epoch": 3.955263507234833,
+      "grad_norm": 0.4537487879806315,
+      "learning_rate": 1.310438577848321e-06,
+      "loss": 0.5349,
+      "step": 14899
+    },
+    {
+      "epoch": 3.9555290057082173,
+      "grad_norm": 0.47308000784751225,
+      "learning_rate": 1.3101315174073162e-06,
+      "loss": 0.552,
+      "step": 14900
+    },
+    {
+      "epoch": 3.955794504181601,
+      "grad_norm": 0.4599804241241482,
+      "learning_rate": 1.3098244801716603e-06,
+      "loss": 0.5603,
+      "step": 14901
+    },
+    {
+      "epoch": 3.9560600026549846,
+      "grad_norm": 0.4780432011214476,
+      "learning_rate": 1.309517466147341e-06,
+      "loss": 0.5734,
+      "step": 14902
+    },
+    {
+      "epoch": 3.9563255011283687,
+      "grad_norm": 0.468186307602429,
+      "learning_rate": 1.3092104753403451e-06,
+      "loss": 0.5691,
+      "step": 14903
+    },
+    {
+      "epoch": 3.9565909996017523,
+      "grad_norm": 0.4796090794452198,
+      "learning_rate": 1.3089035077566615e-06,
+      "loss": 0.5492,
+      "step": 14904
+    },
+    {
+      "epoch": 3.956856498075136,
+      "grad_norm": 0.47630735178455125,
+      "learning_rate": 1.3085965634022746e-06,
+      "loss": 0.5589,
+      "step": 14905
+    },
+    {
+      "epoch": 3.95712199654852,
+      "grad_norm": 0.49052019406847824,
+      "learning_rate": 1.3082896422831726e-06,
+      "loss": 0.541,
+      "step": 14906
+    },
+    {
+      "epoch": 3.9573874950219037,
+      "grad_norm": 0.4753799756949209,
+      "learning_rate": 1.3079827444053405e-06,
+      "loss": 0.5609,
+      "step": 14907
+    },
+    {
+      "epoch": 3.9576529934952873,
+      "grad_norm": 0.46578493267809273,
+      "learning_rate": 1.3076758697747626e-06,
+      "loss": 0.5229,
+      "step": 14908
+    },
+    {
+      "epoch": 3.9579184919686714,
+      "grad_norm": 0.4879936706424716,
+      "learning_rate": 1.3073690183974258e-06,
+      "loss": 0.5238,
+      "step": 14909
+    },
+    {
+      "epoch": 3.958183990442055,
+      "grad_norm": 0.45275372421366017,
+      "learning_rate": 1.3070621902793119e-06,
+      "loss": 0.5357,
+      "step": 14910
+    },
+    {
+      "epoch": 3.9584494889154387,
+      "grad_norm": 0.47478716552696326,
+      "learning_rate": 1.306755385426407e-06,
+      "loss": 0.5709,
+      "step": 14911
+    },
+    {
+      "epoch": 3.9587149873888228,
+      "grad_norm": 0.4625179402739653,
+      "learning_rate": 1.3064486038446938e-06,
+      "loss": 0.5445,
+      "step": 14912
+    },
+    {
+      "epoch": 3.9589804858622064,
+      "grad_norm": 0.46130592873113957,
+      "learning_rate": 1.3061418455401541e-06,
+      "loss": 0.5689,
+      "step": 14913
+    },
+    {
+      "epoch": 3.95924598433559,
+      "grad_norm": 0.4688052922858833,
+      "learning_rate": 1.3058351105187727e-06,
+      "loss": 0.5602,
+      "step": 14914
+    },
+    {
+      "epoch": 3.9595114828089737,
+      "grad_norm": 0.4755720544845919,
+      "learning_rate": 1.305528398786529e-06,
+      "loss": 0.5277,
+      "step": 14915
+    },
+    {
+      "epoch": 3.9597769812823578,
+      "grad_norm": 0.4722655047576312,
+      "learning_rate": 1.3052217103494074e-06,
+      "loss": 0.5287,
+      "step": 14916
+    },
+    {
+      "epoch": 3.9600424797557414,
+      "grad_norm": 0.4686952058619256,
+      "learning_rate": 1.304915045213387e-06,
+      "loss": 0.5177,
+      "step": 14917
+    },
+    {
+      "epoch": 3.960307978229125,
+      "grad_norm": 0.4668946603789969,
+      "learning_rate": 1.3046084033844507e-06,
+      "loss": 0.5323,
+      "step": 14918
+    },
+    {
+      "epoch": 3.9605734767025087,
+      "grad_norm": 0.47850541577127204,
+      "learning_rate": 1.3043017848685758e-06,
+      "loss": 0.5769,
+      "step": 14919
+    },
+    {
+      "epoch": 3.960838975175893,
+      "grad_norm": 0.4711221477497711,
+      "learning_rate": 1.3039951896717445e-06,
+      "loss": 0.5267,
+      "step": 14920
+    },
+    {
+      "epoch": 3.9611044736492764,
+      "grad_norm": 0.46058899888001475,
+      "learning_rate": 1.3036886177999347e-06,
+      "loss": 0.5534,
+      "step": 14921
+    },
+    {
+      "epoch": 3.96136997212266,
+      "grad_norm": 0.45203158357220796,
+      "learning_rate": 1.3033820692591264e-06,
+      "loss": 0.5408,
+      "step": 14922
+    },
+    {
+      "epoch": 3.961635470596044,
+      "grad_norm": 0.46265349759709323,
+      "learning_rate": 1.3030755440552967e-06,
+      "loss": 0.5162,
+      "step": 14923
+    },
+    {
+      "epoch": 3.961900969069428,
+      "grad_norm": 0.47067238410677087,
+      "learning_rate": 1.3027690421944256e-06,
+      "loss": 0.5659,
+      "step": 14924
+    },
+    {
+      "epoch": 3.9621664675428114,
+      "grad_norm": 0.46148835914390723,
+      "learning_rate": 1.3024625636824894e-06,
+      "loss": 0.4911,
+      "step": 14925
+    },
+    {
+      "epoch": 3.9624319660161955,
+      "grad_norm": 0.46504065583258425,
+      "learning_rate": 1.3021561085254643e-06,
+      "loss": 0.5517,
+      "step": 14926
+    },
+    {
+      "epoch": 3.962697464489579,
+      "grad_norm": 0.46869581892077433,
+      "learning_rate": 1.3018496767293292e-06,
+      "loss": 0.5475,
+      "step": 14927
+    },
+    {
+      "epoch": 3.962962962962963,
+      "grad_norm": 0.46919287656944525,
+      "learning_rate": 1.3015432683000578e-06,
+      "loss": 0.5306,
+      "step": 14928
+    },
+    {
+      "epoch": 3.963228461436347,
+      "grad_norm": 0.485851738415879,
+      "learning_rate": 1.3012368832436278e-06,
+      "loss": 0.5858,
+      "step": 14929
+    },
+    {
+      "epoch": 3.9634939599097305,
+      "grad_norm": 0.4732987072397958,
+      "learning_rate": 1.3009305215660139e-06,
+      "loss": 0.5535,
+      "step": 14930
+    },
+    {
+      "epoch": 3.963759458383114,
+      "grad_norm": 0.450174190873984,
+      "learning_rate": 1.3006241832731902e-06,
+      "loss": 0.5323,
+      "step": 14931
+    },
+    {
+      "epoch": 3.9640249568564982,
+      "grad_norm": 0.46949930360077097,
+      "learning_rate": 1.300317868371131e-06,
+      "loss": 0.5209,
+      "step": 14932
+    },
+    {
+      "epoch": 3.964290455329882,
+      "grad_norm": 0.4676639960211,
+      "learning_rate": 1.3000115768658105e-06,
+      "loss": 0.5845,
+      "step": 14933
+    },
+    {
+      "epoch": 3.9645559538032655,
+      "grad_norm": 0.4719369244599207,
+      "learning_rate": 1.2997053087632034e-06,
+      "loss": 0.5388,
+      "step": 14934
+    },
+    {
+      "epoch": 3.9648214522766496,
+      "grad_norm": 0.470898942578198,
+      "learning_rate": 1.2993990640692805e-06,
+      "loss": 0.526,
+      "step": 14935
+    },
+    {
+      "epoch": 3.9650869507500333,
+      "grad_norm": 0.466199010049318,
+      "learning_rate": 1.2990928427900175e-06,
+      "loss": 0.5942,
+      "step": 14936
+    },
+    {
+      "epoch": 3.965352449223417,
+      "grad_norm": 0.47872100441615095,
+      "learning_rate": 1.2987866449313824e-06,
+      "loss": 0.5379,
+      "step": 14937
+    },
+    {
+      "epoch": 3.965617947696801,
+      "grad_norm": 0.4752506107634398,
+      "learning_rate": 1.2984804704993499e-06,
+      "loss": 0.5446,
+      "step": 14938
+    },
+    {
+      "epoch": 3.9658834461701846,
+      "grad_norm": 0.4659055309987489,
+      "learning_rate": 1.2981743194998892e-06,
+      "loss": 0.5641,
+      "step": 14939
+    },
+    {
+      "epoch": 3.9661489446435683,
+      "grad_norm": 0.48334892369584376,
+      "learning_rate": 1.2978681919389725e-06,
+      "loss": 0.5439,
+      "step": 14940
+    },
+    {
+      "epoch": 3.9664144431169523,
+      "grad_norm": 0.46241064877041416,
+      "learning_rate": 1.2975620878225692e-06,
+      "loss": 0.5573,
+      "step": 14941
+    },
+    {
+      "epoch": 3.966679941590336,
+      "grad_norm": 0.46948984537769817,
+      "learning_rate": 1.2972560071566486e-06,
+      "loss": 0.5261,
+      "step": 14942
+    },
+    {
+      "epoch": 3.9669454400637196,
+      "grad_norm": 0.4523405533912353,
+      "learning_rate": 1.2969499499471815e-06,
+      "loss": 0.5463,
+      "step": 14943
+    },
+    {
+      "epoch": 3.9672109385371033,
+      "grad_norm": 0.47618601543911376,
+      "learning_rate": 1.2966439162001353e-06,
+      "loss": 0.5717,
+      "step": 14944
+    },
+    {
+      "epoch": 3.9674764370104874,
+      "grad_norm": 0.46217312074765254,
+      "learning_rate": 1.2963379059214799e-06,
+      "loss": 0.5502,
+      "step": 14945
+    },
+    {
+      "epoch": 3.967741935483871,
+      "grad_norm": 0.4691137261536149,
+      "learning_rate": 1.2960319191171813e-06,
+      "loss": 0.5652,
+      "step": 14946
+    },
+    {
+      "epoch": 3.9680074339572546,
+      "grad_norm": 0.4852501491985933,
+      "learning_rate": 1.29572595579321e-06,
+      "loss": 0.5465,
+      "step": 14947
+    },
+    {
+      "epoch": 3.9682729324306383,
+      "grad_norm": 0.46379310370950805,
+      "learning_rate": 1.2954200159555294e-06,
+      "loss": 0.5306,
+      "step": 14948
+    },
+    {
+      "epoch": 3.9685384309040224,
+      "grad_norm": 0.47226271747624765,
+      "learning_rate": 1.2951140996101086e-06,
+      "loss": 0.5799,
+      "step": 14949
+    },
+    {
+      "epoch": 3.968803929377406,
+      "grad_norm": 0.47750893448024595,
+      "learning_rate": 1.294808206762912e-06,
+      "loss": 0.5478,
+      "step": 14950
+    },
+    {
+      "epoch": 3.9690694278507896,
+      "grad_norm": 0.47414409582408085,
+      "learning_rate": 1.2945023374199072e-06,
+      "loss": 0.5465,
+      "step": 14951
+    },
+    {
+      "epoch": 3.9693349263241737,
+      "grad_norm": 0.477569394229894,
+      "learning_rate": 1.2941964915870575e-06,
+      "loss": 0.5238,
+      "step": 14952
+    },
+    {
+      "epoch": 3.9696004247975574,
+      "grad_norm": 0.464624804336812,
+      "learning_rate": 1.2938906692703297e-06,
+      "loss": 0.5759,
+      "step": 14953
+    },
+    {
+      "epoch": 3.969865923270941,
+      "grad_norm": 0.4703693306690422,
+      "learning_rate": 1.2935848704756865e-06,
+      "loss": 0.5454,
+      "step": 14954
+    },
+    {
+      "epoch": 3.970131421744325,
+      "grad_norm": 0.454985436313049,
+      "learning_rate": 1.2932790952090912e-06,
+      "loss": 0.563,
+      "step": 14955
+    },
+    {
+      "epoch": 3.9703969202177087,
+      "grad_norm": 0.46422809659587905,
+      "learning_rate": 1.292973343476509e-06,
+      "loss": 0.5229,
+      "step": 14956
+    },
+    {
+      "epoch": 3.9706624186910924,
+      "grad_norm": 0.4841446626267229,
+      "learning_rate": 1.2926676152839014e-06,
+      "loss": 0.5696,
+      "step": 14957
+    },
+    {
+      "epoch": 3.9709279171644765,
+      "grad_norm": 0.4814343609958354,
+      "learning_rate": 1.292361910637232e-06,
+      "loss": 0.5497,
+      "step": 14958
+    },
+    {
+      "epoch": 3.97119341563786,
+      "grad_norm": 0.45941633575722013,
+      "learning_rate": 1.2920562295424622e-06,
+      "loss": 0.561,
+      "step": 14959
+    },
+    {
+      "epoch": 3.9714589141112437,
+      "grad_norm": 0.5050150602135764,
+      "learning_rate": 1.2917505720055522e-06,
+      "loss": 0.531,
+      "step": 14960
+    },
+    {
+      "epoch": 3.971724412584628,
+      "grad_norm": 0.47499322069921385,
+      "learning_rate": 1.2914449380324657e-06,
+      "loss": 0.5448,
+      "step": 14961
+    },
+    {
+      "epoch": 3.9719899110580115,
+      "grad_norm": 0.4837694118754429,
+      "learning_rate": 1.291139327629161e-06,
+      "loss": 0.5768,
+      "step": 14962
+    },
+    {
+      "epoch": 3.972255409531395,
+      "grad_norm": 0.4642834733446551,
+      "learning_rate": 1.2908337408016002e-06,
+      "loss": 0.5143,
+      "step": 14963
+    },
+    {
+      "epoch": 3.972520908004779,
+      "grad_norm": 0.45152121745984114,
+      "learning_rate": 1.290528177555742e-06,
+      "loss": 0.5475,
+      "step": 14964
+    },
+    {
+      "epoch": 3.972786406478163,
+      "grad_norm": 0.47164458270950205,
+      "learning_rate": 1.2902226378975451e-06,
+      "loss": 0.5346,
+      "step": 14965
+    },
+    {
+      "epoch": 3.9730519049515465,
+      "grad_norm": 0.49289374882653286,
+      "learning_rate": 1.2899171218329687e-06,
+      "loss": 0.5592,
+      "step": 14966
+    },
+    {
+      "epoch": 3.9733174034249306,
+      "grad_norm": 0.4620583919626006,
+      "learning_rate": 1.2896116293679717e-06,
+      "loss": 0.5454,
+      "step": 14967
+    },
+    {
+      "epoch": 3.973582901898314,
+      "grad_norm": 0.4750846373561971,
+      "learning_rate": 1.2893061605085105e-06,
+      "loss": 0.516,
+      "step": 14968
+    },
+    {
+      "epoch": 3.973848400371698,
+      "grad_norm": 0.4451672458331521,
+      "learning_rate": 1.2890007152605444e-06,
+      "loss": 0.5264,
+      "step": 14969
+    },
+    {
+      "epoch": 3.9741138988450815,
+      "grad_norm": 0.46290862552706813,
+      "learning_rate": 1.2886952936300295e-06,
+      "loss": 0.5601,
+      "step": 14970
+    },
+    {
+      "epoch": 3.9743793973184656,
+      "grad_norm": 0.47511836888988485,
+      "learning_rate": 1.2883898956229213e-06,
+      "loss": 0.6266,
+      "step": 14971
+    },
+    {
+      "epoch": 3.974644895791849,
+      "grad_norm": 0.46549697888156516,
+      "learning_rate": 1.2880845212451776e-06,
+      "loss": 0.547,
+      "step": 14972
+    },
+    {
+      "epoch": 3.974910394265233,
+      "grad_norm": 0.47350507962105876,
+      "learning_rate": 1.2877791705027518e-06,
+      "loss": 0.5252,
+      "step": 14973
+    },
+    {
+      "epoch": 3.9751758927386165,
+      "grad_norm": 0.47653616983064395,
+      "learning_rate": 1.2874738434016012e-06,
+      "loss": 0.5274,
+      "step": 14974
+    },
+    {
+      "epoch": 3.9754413912120006,
+      "grad_norm": 0.4801727463890366,
+      "learning_rate": 1.2871685399476785e-06,
+      "loss": 0.5597,
+      "step": 14975
+    },
+    {
+      "epoch": 3.975706889685384,
+      "grad_norm": 0.4916498242613426,
+      "learning_rate": 1.286863260146941e-06,
+      "loss": 0.5626,
+      "step": 14976
+    },
+    {
+      "epoch": 3.975972388158768,
+      "grad_norm": 0.47674732949084403,
+      "learning_rate": 1.2865580040053382e-06,
+      "loss": 0.5693,
+      "step": 14977
+    },
+    {
+      "epoch": 3.976237886632152,
+      "grad_norm": 0.47814701520184716,
+      "learning_rate": 1.286252771528825e-06,
+      "loss": 0.5456,
+      "step": 14978
+    },
+    {
+      "epoch": 3.9765033851055356,
+      "grad_norm": 0.46373599533236476,
+      "learning_rate": 1.2859475627233558e-06,
+      "loss": 0.5501,
+      "step": 14979
+    },
+    {
+      "epoch": 3.9767688835789192,
+      "grad_norm": 0.4740841867568828,
+      "learning_rate": 1.2856423775948807e-06,
+      "loss": 0.5269,
+      "step": 14980
+    },
+    {
+      "epoch": 3.9770343820523033,
+      "grad_norm": 0.47173562719011025,
+      "learning_rate": 1.2853372161493532e-06,
+      "loss": 0.5736,
+      "step": 14981
+    },
+    {
+      "epoch": 3.977299880525687,
+      "grad_norm": 0.4696597120038557,
+      "learning_rate": 1.2850320783927242e-06,
+      "loss": 0.5306,
+      "step": 14982
+    },
+    {
+      "epoch": 3.9775653789990706,
+      "grad_norm": 0.47349760083705583,
+      "learning_rate": 1.2847269643309444e-06,
+      "loss": 0.5907,
+      "step": 14983
+    },
+    {
+      "epoch": 3.9778308774724547,
+      "grad_norm": 0.4584450553247524,
+      "learning_rate": 1.2844218739699632e-06,
+      "loss": 0.5323,
+      "step": 14984
+    },
+    {
+      "epoch": 3.9780963759458383,
+      "grad_norm": 0.46733540990008776,
+      "learning_rate": 1.2841168073157328e-06,
+      "loss": 0.5579,
+      "step": 14985
+    },
+    {
+      "epoch": 3.978361874419222,
+      "grad_norm": 0.4630328162399076,
+      "learning_rate": 1.2838117643742007e-06,
+      "loss": 0.5377,
+      "step": 14986
+    },
+    {
+      "epoch": 3.978627372892606,
+      "grad_norm": 0.45743296090341,
+      "learning_rate": 1.2835067451513179e-06,
+      "loss": 0.5568,
+      "step": 14987
+    },
+    {
+      "epoch": 3.9788928713659897,
+      "grad_norm": 0.4764792511234045,
+      "learning_rate": 1.2832017496530319e-06,
+      "loss": 0.5369,
+      "step": 14988
+    },
+    {
+      "epoch": 3.9791583698393733,
+      "grad_norm": 0.4631299068274557,
+      "learning_rate": 1.2828967778852901e-06,
+      "loss": 0.5591,
+      "step": 14989
+    },
+    {
+      "epoch": 3.9794238683127574,
+      "grad_norm": 0.46736424303170726,
+      "learning_rate": 1.2825918298540425e-06,
+      "loss": 0.5734,
+      "step": 14990
+    },
+    {
+      "epoch": 3.979689366786141,
+      "grad_norm": 0.4596673588404188,
+      "learning_rate": 1.2822869055652332e-06,
+      "loss": 0.5685,
+      "step": 14991
+    },
+    {
+      "epoch": 3.9799548652595247,
+      "grad_norm": 0.4608041842215645,
+      "learning_rate": 1.2819820050248122e-06,
+      "loss": 0.5536,
+      "step": 14992
+    },
+    {
+      "epoch": 3.9802203637329088,
+      "grad_norm": 0.45844389242061234,
+      "learning_rate": 1.2816771282387241e-06,
+      "loss": 0.5664,
+      "step": 14993
+    },
+    {
+      "epoch": 3.9804858622062924,
+      "grad_norm": 0.46642773693792194,
+      "learning_rate": 1.2813722752129153e-06,
+      "loss": 0.5438,
+      "step": 14994
+    },
+    {
+      "epoch": 3.980751360679676,
+      "grad_norm": 0.47945142635099863,
+      "learning_rate": 1.2810674459533295e-06,
+      "loss": 0.6066,
+      "step": 14995
+    },
+    {
+      "epoch": 3.98101685915306,
+      "grad_norm": 0.47017956493759977,
+      "learning_rate": 1.2807626404659144e-06,
+      "loss": 0.5233,
+      "step": 14996
+    },
+    {
+      "epoch": 3.981282357626444,
+      "grad_norm": 0.46503565794306523,
+      "learning_rate": 1.2804578587566118e-06,
+      "loss": 0.5322,
+      "step": 14997
+    },
+    {
+      "epoch": 3.9815478560998274,
+      "grad_norm": 0.46604491644306034,
+      "learning_rate": 1.2801531008313666e-06,
+      "loss": 0.5379,
+      "step": 14998
+    },
+    {
+      "epoch": 3.981813354573211,
+      "grad_norm": 0.4813322584832262,
+      "learning_rate": 1.279848366696125e-06,
+      "loss": 0.5515,
+      "step": 14999
+    },
+    {
+      "epoch": 3.982078853046595,
+      "grad_norm": 0.47208275209553235,
+      "learning_rate": 1.2795436563568255e-06,
+      "loss": 0.5398,
+      "step": 15000
+    },
+    {
+      "epoch": 3.982344351519979,
+      "grad_norm": 0.46783170853735734,
+      "learning_rate": 1.2792389698194141e-06,
+      "loss": 0.5711,
+      "step": 15001
+    },
+    {
+      "epoch": 3.9826098499933624,
+      "grad_norm": 0.48102809369045024,
+      "learning_rate": 1.2789343070898308e-06,
+      "loss": 0.55,
+      "step": 15002
+    },
+    {
+      "epoch": 3.982875348466746,
+      "grad_norm": 0.44590997140997696,
+      "learning_rate": 1.2786296681740187e-06,
+      "loss": 0.4976,
+      "step": 15003
+    },
+    {
+      "epoch": 3.98314084694013,
+      "grad_norm": 0.46793353801810605,
+      "learning_rate": 1.278325053077918e-06,
+      "loss": 0.5645,
+      "step": 15004
+    },
+    {
+      "epoch": 3.983406345413514,
+      "grad_norm": 0.46773767470305805,
+      "learning_rate": 1.2780204618074707e-06,
+      "loss": 0.5299,
+      "step": 15005
+    },
+    {
+      "epoch": 3.9836718438868974,
+      "grad_norm": 0.476223994819842,
+      "learning_rate": 1.2777158943686164e-06,
+      "loss": 0.5807,
+      "step": 15006
+    },
+    {
+      "epoch": 3.9839373423602815,
+      "grad_norm": 0.46343285836038395,
+      "learning_rate": 1.2774113507672936e-06,
+      "loss": 0.5488,
+      "step": 15007
+    },
+    {
+      "epoch": 3.984202840833665,
+      "grad_norm": 0.4797275722706083,
+      "learning_rate": 1.2771068310094441e-06,
+      "loss": 0.5641,
+      "step": 15008
+    },
+    {
+      "epoch": 3.984468339307049,
+      "grad_norm": 0.48851460862904245,
+      "learning_rate": 1.2768023351010045e-06,
+      "loss": 0.5685,
+      "step": 15009
+    },
+    {
+      "epoch": 3.984733837780433,
+      "grad_norm": 0.4870580072755374,
+      "learning_rate": 1.2764978630479164e-06,
+      "loss": 0.5024,
+      "step": 15010
+    },
+    {
+      "epoch": 3.9849993362538165,
+      "grad_norm": 0.47268997662237905,
+      "learning_rate": 1.2761934148561133e-06,
+      "loss": 0.546,
+      "step": 15011
+    },
+    {
+      "epoch": 3.9852648347272,
+      "grad_norm": 0.4644151461883268,
+      "learning_rate": 1.2758889905315363e-06,
+      "loss": 0.5472,
+      "step": 15012
+    },
+    {
+      "epoch": 3.9855303332005843,
+      "grad_norm": 0.4591858538447491,
+      "learning_rate": 1.2755845900801201e-06,
+      "loss": 0.5828,
+      "step": 15013
+    },
+    {
+      "epoch": 3.985795831673968,
+      "grad_norm": 0.4801824961065117,
+      "learning_rate": 1.2752802135078032e-06,
+      "loss": 0.5231,
+      "step": 15014
+    },
+    {
+      "epoch": 3.9860613301473515,
+      "grad_norm": 0.4550210477955791,
+      "learning_rate": 1.27497586082052e-06,
+      "loss": 0.5197,
+      "step": 15015
+    },
+    {
+      "epoch": 3.9863268286207356,
+      "grad_norm": 0.4669918544578851,
+      "learning_rate": 1.2746715320242075e-06,
+      "loss": 0.5582,
+      "step": 15016
+    },
+    {
+      "epoch": 3.9865923270941193,
+      "grad_norm": 0.46599598017282245,
+      "learning_rate": 1.2743672271248004e-06,
+      "loss": 0.5379,
+      "step": 15017
+    },
+    {
+      "epoch": 3.986857825567503,
+      "grad_norm": 0.463843518734699,
+      "learning_rate": 1.2740629461282322e-06,
+      "loss": 0.554,
+      "step": 15018
+    },
+    {
+      "epoch": 3.987123324040887,
+      "grad_norm": 0.46760077274773915,
+      "learning_rate": 1.273758689040439e-06,
+      "loss": 0.5255,
+      "step": 15019
+    },
+    {
+      "epoch": 3.9873888225142706,
+      "grad_norm": 0.4580637302811964,
+      "learning_rate": 1.2734544558673533e-06,
+      "loss": 0.5602,
+      "step": 15020
+    },
+    {
+      "epoch": 3.9876543209876543,
+      "grad_norm": 0.456662076845957,
+      "learning_rate": 1.2731502466149094e-06,
+      "loss": 0.5572,
+      "step": 15021
+    },
+    {
+      "epoch": 3.9879198194610384,
+      "grad_norm": 0.47140501833254544,
+      "learning_rate": 1.2728460612890399e-06,
+      "loss": 0.5251,
+      "step": 15022
+    },
+    {
+      "epoch": 3.988185317934422,
+      "grad_norm": 0.4619975230201168,
+      "learning_rate": 1.272541899895675e-06,
+      "loss": 0.5672,
+      "step": 15023
+    },
+    {
+      "epoch": 3.9884508164078056,
+      "grad_norm": 0.47328048651979565,
+      "learning_rate": 1.27223776244075e-06,
+      "loss": 0.5725,
+      "step": 15024
+    },
+    {
+      "epoch": 3.9887163148811893,
+      "grad_norm": 0.4639850826698106,
+      "learning_rate": 1.2719336489301937e-06,
+      "loss": 0.5766,
+      "step": 15025
+    },
+    {
+      "epoch": 3.9889818133545734,
+      "grad_norm": 0.4516415805619876,
+      "learning_rate": 1.271629559369939e-06,
+      "loss": 0.5498,
+      "step": 15026
+    },
+    {
+      "epoch": 3.989247311827957,
+      "grad_norm": 0.4725439799571295,
+      "learning_rate": 1.2713254937659141e-06,
+      "loss": 0.5433,
+      "step": 15027
+    },
+    {
+      "epoch": 3.9895128103013406,
+      "grad_norm": 0.4716593087307472,
+      "learning_rate": 1.2710214521240527e-06,
+      "loss": 0.5836,
+      "step": 15028
+    },
+    {
+      "epoch": 3.9897783087747243,
+      "grad_norm": 0.47480872975186994,
+      "learning_rate": 1.2707174344502805e-06,
+      "loss": 0.5435,
+      "step": 15029
+    },
+    {
+      "epoch": 3.9900438072481084,
+      "grad_norm": 0.4723830975557132,
+      "learning_rate": 1.2704134407505284e-06,
+      "loss": 0.5226,
+      "step": 15030
+    },
+    {
+      "epoch": 3.990309305721492,
+      "grad_norm": 0.4740560231250832,
+      "learning_rate": 1.270109471030724e-06,
+      "loss": 0.5487,
+      "step": 15031
+    },
+    {
+      "epoch": 3.9905748041948756,
+      "grad_norm": 0.46974667786352736,
+      "learning_rate": 1.2698055252967973e-06,
+      "loss": 0.5332,
+      "step": 15032
+    },
+    {
+      "epoch": 3.9908403026682597,
+      "grad_norm": 0.47602224874885063,
+      "learning_rate": 1.2695016035546747e-06,
+      "loss": 0.5454,
+      "step": 15033
+    },
+    {
+      "epoch": 3.9911058011416434,
+      "grad_norm": 0.4649963303079685,
+      "learning_rate": 1.2691977058102828e-06,
+      "loss": 0.5499,
+      "step": 15034
+    },
+    {
+      "epoch": 3.991371299615027,
+      "grad_norm": 0.470827609554469,
+      "learning_rate": 1.2688938320695502e-06,
+      "loss": 0.5511,
+      "step": 15035
+    },
+    {
+      "epoch": 3.991636798088411,
+      "grad_norm": 0.4825681075893088,
+      "learning_rate": 1.268589982338401e-06,
+      "loss": 0.501,
+      "step": 15036
+    },
+    {
+      "epoch": 3.9919022965617947,
+      "grad_norm": 0.4691903735719901,
+      "learning_rate": 1.2682861566227634e-06,
+      "loss": 0.5257,
+      "step": 15037
+    },
+    {
+      "epoch": 3.9921677950351784,
+      "grad_norm": 0.4551943054152704,
+      "learning_rate": 1.2679823549285602e-06,
+      "loss": 0.5284,
+      "step": 15038
+    },
+    {
+      "epoch": 3.9924332935085625,
+      "grad_norm": 0.4912882594930225,
+      "learning_rate": 1.2676785772617195e-06,
+      "loss": 0.5639,
+      "step": 15039
+    },
+    {
+      "epoch": 3.992698791981946,
+      "grad_norm": 0.46242044827996776,
+      "learning_rate": 1.267374823628162e-06,
+      "loss": 0.5562,
+      "step": 15040
+    },
+    {
+      "epoch": 3.9929642904553297,
+      "grad_norm": 0.46519884382487803,
+      "learning_rate": 1.2670710940338143e-06,
+      "loss": 0.5559,
+      "step": 15041
+    },
+    {
+      "epoch": 3.993229788928714,
+      "grad_norm": 0.48517177385218785,
+      "learning_rate": 1.2667673884845983e-06,
+      "loss": 0.5084,
+      "step": 15042
+    },
+    {
+      "epoch": 3.9934952874020975,
+      "grad_norm": 0.4565881485044334,
+      "learning_rate": 1.2664637069864382e-06,
+      "loss": 0.5008,
+      "step": 15043
+    },
+    {
+      "epoch": 3.993760785875481,
+      "grad_norm": 0.46227386453857205,
+      "learning_rate": 1.2661600495452553e-06,
+      "loss": 0.5433,
+      "step": 15044
+    },
+    {
+      "epoch": 3.994026284348865,
+      "grad_norm": 0.47903784432868496,
+      "learning_rate": 1.2658564161669734e-06,
+      "loss": 0.5318,
+      "step": 15045
+    },
+    {
+      "epoch": 3.994291782822249,
+      "grad_norm": 0.4802638886121846,
+      "learning_rate": 1.2655528068575134e-06,
+      "loss": 0.585,
+      "step": 15046
+    },
+    {
+      "epoch": 3.9945572812956325,
+      "grad_norm": 0.48037249518518654,
+      "learning_rate": 1.2652492216227946e-06,
+      "loss": 0.5503,
+      "step": 15047
+    },
+    {
+      "epoch": 3.9948227797690166,
+      "grad_norm": 0.465752790651287,
+      "learning_rate": 1.2649456604687404e-06,
+      "loss": 0.5192,
+      "step": 15048
+    },
+    {
+      "epoch": 3.9950882782424,
+      "grad_norm": 0.47675762816693384,
+      "learning_rate": 1.264642123401269e-06,
+      "loss": 0.5089,
+      "step": 15049
+    },
+    {
+      "epoch": 3.995353776715784,
+      "grad_norm": 0.47698388151398013,
+      "learning_rate": 1.2643386104263015e-06,
+      "loss": 0.5611,
+      "step": 15050
+    },
+    {
+      "epoch": 3.995619275189168,
+      "grad_norm": 0.457349939900204,
+      "learning_rate": 1.2640351215497565e-06,
+      "loss": 0.5281,
+      "step": 15051
+    },
+    {
+      "epoch": 3.9958847736625516,
+      "grad_norm": 0.47369061902029114,
+      "learning_rate": 1.263731656777552e-06,
+      "loss": 0.5078,
+      "step": 15052
+    },
+    {
+      "epoch": 3.996150272135935,
+      "grad_norm": 0.46872906783928264,
+      "learning_rate": 1.263428216115608e-06,
+      "loss": 0.589,
+      "step": 15053
+    },
+    {
+      "epoch": 3.996415770609319,
+      "grad_norm": 0.4674385960335175,
+      "learning_rate": 1.2631247995698405e-06,
+      "loss": 0.5777,
+      "step": 15054
+    },
+    {
+      "epoch": 3.996681269082703,
+      "grad_norm": 0.4657049807108756,
+      "learning_rate": 1.2628214071461687e-06,
+      "loss": 0.5381,
+      "step": 15055
+    },
+    {
+      "epoch": 3.9969467675560866,
+      "grad_norm": 0.46820083907544646,
+      "learning_rate": 1.2625180388505077e-06,
+      "loss": 0.5603,
+      "step": 15056
+    },
+    {
+      "epoch": 3.99721226602947,
+      "grad_norm": 0.48812374655270796,
+      "learning_rate": 1.2622146946887768e-06,
+      "loss": 0.5025,
+      "step": 15057
+    },
+    {
+      "epoch": 3.997477764502854,
+      "grad_norm": 0.4760098176879443,
+      "learning_rate": 1.261911374666888e-06,
+      "loss": 0.508,
+      "step": 15058
+    },
+    {
+      "epoch": 3.997743262976238,
+      "grad_norm": 0.47084740892680127,
+      "learning_rate": 1.2616080787907597e-06,
+      "loss": 0.5239,
+      "step": 15059
+    },
+    {
+      "epoch": 3.9980087614496216,
+      "grad_norm": 0.4669291354264763,
+      "learning_rate": 1.2613048070663048e-06,
+      "loss": 0.5397,
+      "step": 15060
+    },
+    {
+      "epoch": 3.9982742599230052,
+      "grad_norm": 0.4676787332570303,
+      "learning_rate": 1.2610015594994402e-06,
+      "loss": 0.5899,
+      "step": 15061
+    },
+    {
+      "epoch": 3.9985397583963893,
+      "grad_norm": 0.4840774397232102,
+      "learning_rate": 1.260698336096079e-06,
+      "loss": 0.55,
+      "step": 15062
+    },
+    {
+      "epoch": 3.998805256869773,
+      "grad_norm": 0.4788547395594691,
+      "learning_rate": 1.2603951368621332e-06,
+      "loss": 0.5546,
+      "step": 15063
+    },
+    {
+      "epoch": 3.9990707553431566,
+      "grad_norm": 0.47469291867159735,
+      "learning_rate": 1.2600919618035185e-06,
+      "loss": 0.5581,
+      "step": 15064
+    },
+    {
+      "epoch": 3.9993362538165407,
+      "grad_norm": 0.4670355853769135,
+      "learning_rate": 1.2597888109261452e-06,
+      "loss": 0.4904,
+      "step": 15065
+    },
+    {
+      "epoch": 3.9996017522899243,
+      "grad_norm": 0.4522452935609553,
+      "learning_rate": 1.2594856842359276e-06,
+      "loss": 0.5373,
+      "step": 15066
+    },
+    {
+      "epoch": 3.999867250763308,
+      "grad_norm": 0.48647357772196476,
+      "learning_rate": 1.259182581738776e-06,
+      "loss": 0.5413,
+      "step": 15067
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.7339422262195487,
+      "learning_rate": 1.2588795034406037e-06,
+      "loss": 0.524,
+      "step": 15068
+    },
+    {
+      "epoch": 4.000265498473384,
+      "grad_norm": 0.4820985888355415,
+      "learning_rate": 1.258576449347318e-06,
+      "loss": 0.544,
+      "step": 15069
+    },
+    {
+      "epoch": 4.000530996946767,
+      "grad_norm": 0.4891848268480109,
+      "learning_rate": 1.2582734194648308e-06,
+      "loss": 0.5128,
+      "step": 15070
+    },
+    {
+      "epoch": 4.000796495420151,
+      "grad_norm": 0.4645609964529606,
+      "learning_rate": 1.2579704137990534e-06,
+      "loss": 0.5411,
+      "step": 15071
+    },
+    {
+      "epoch": 4.0010619938935355,
+      "grad_norm": 0.46660115952006626,
+      "learning_rate": 1.257667432355893e-06,
+      "loss": 0.5325,
+      "step": 15072
+    },
+    {
+      "epoch": 4.001327492366919,
+      "grad_norm": 0.47860768609232207,
+      "learning_rate": 1.2573644751412601e-06,
+      "loss": 0.5332,
+      "step": 15073
+    },
+    {
+      "epoch": 4.001592990840303,
+      "grad_norm": 0.4596484348263277,
+      "learning_rate": 1.2570615421610627e-06,
+      "loss": 0.538,
+      "step": 15074
+    },
+    {
+      "epoch": 4.001858489313687,
+      "grad_norm": 0.4729811199921599,
+      "learning_rate": 1.2567586334212085e-06,
+      "loss": 0.5362,
+      "step": 15075
+    },
+    {
+      "epoch": 4.00212398778707,
+      "grad_norm": 0.45889927094779487,
+      "learning_rate": 1.2564557489276035e-06,
+      "loss": 0.5362,
+      "step": 15076
+    },
+    {
+      "epoch": 4.002389486260454,
+      "grad_norm": 0.47692005688045225,
+      "learning_rate": 1.2561528886861576e-06,
+      "loss": 0.5034,
+      "step": 15077
+    },
+    {
+      "epoch": 4.002654984733838,
+      "grad_norm": 0.4838636670985966,
+      "learning_rate": 1.2558500527027748e-06,
+      "loss": 0.5512,
+      "step": 15078
+    },
+    {
+      "epoch": 4.002920483207221,
+      "grad_norm": 0.469017416043095,
+      "learning_rate": 1.255547240983363e-06,
+      "loss": 0.5283,
+      "step": 15079
+    },
+    {
+      "epoch": 4.0031859816806055,
+      "grad_norm": 0.4784606122633153,
+      "learning_rate": 1.2552444535338271e-06,
+      "loss": 0.5405,
+      "step": 15080
+    },
+    {
+      "epoch": 4.0034514801539896,
+      "grad_norm": 0.4615472487565943,
+      "learning_rate": 1.2549416903600708e-06,
+      "loss": 0.562,
+      "step": 15081
+    },
+    {
+      "epoch": 4.003716978627373,
+      "grad_norm": 0.47166247930936844,
+      "learning_rate": 1.2546389514680013e-06,
+      "loss": 0.514,
+      "step": 15082
+    },
+    {
+      "epoch": 4.003982477100757,
+      "grad_norm": 0.46931310789938635,
+      "learning_rate": 1.2543362368635204e-06,
+      "loss": 0.5201,
+      "step": 15083
+    },
+    {
+      "epoch": 4.00424797557414,
+      "grad_norm": 0.4601523589558971,
+      "learning_rate": 1.2540335465525333e-06,
+      "loss": 0.56,
+      "step": 15084
+    },
+    {
+      "epoch": 4.004513474047524,
+      "grad_norm": 0.4647278613682582,
+      "learning_rate": 1.2537308805409431e-06,
+      "loss": 0.5088,
+      "step": 15085
+    },
+    {
+      "epoch": 4.004778972520908,
+      "grad_norm": 0.46579187163982266,
+      "learning_rate": 1.253428238834652e-06,
+      "loss": 0.5327,
+      "step": 15086
+    },
+    {
+      "epoch": 4.005044470994291,
+      "grad_norm": 0.4699192785317288,
+      "learning_rate": 1.2531256214395616e-06,
+      "loss": 0.5437,
+      "step": 15087
+    },
+    {
+      "epoch": 4.0053099694676755,
+      "grad_norm": 0.46590889105783856,
+      "learning_rate": 1.2528230283615753e-06,
+      "loss": 0.5602,
+      "step": 15088
+    },
+    {
+      "epoch": 4.00557546794106,
+      "grad_norm": 0.4694549150317479,
+      "learning_rate": 1.2525204596065926e-06,
+      "loss": 0.5317,
+      "step": 15089
+    },
+    {
+      "epoch": 4.005840966414443,
+      "grad_norm": 0.4659588151797014,
+      "learning_rate": 1.2522179151805154e-06,
+      "loss": 0.5257,
+      "step": 15090
+    },
+    {
+      "epoch": 4.006106464887827,
+      "grad_norm": 0.4769916576213578,
+      "learning_rate": 1.2519153950892454e-06,
+      "loss": 0.5546,
+      "step": 15091
+    },
+    {
+      "epoch": 4.006371963361211,
+      "grad_norm": 0.47532170357266684,
+      "learning_rate": 1.2516128993386795e-06,
+      "loss": 0.5268,
+      "step": 15092
+    },
+    {
+      "epoch": 4.006637461834594,
+      "grad_norm": 0.47856842474950184,
+      "learning_rate": 1.2513104279347197e-06,
+      "loss": 0.5472,
+      "step": 15093
+    },
+    {
+      "epoch": 4.006902960307978,
+      "grad_norm": 0.4747897205521734,
+      "learning_rate": 1.2510079808832626e-06,
+      "loss": 0.5702,
+      "step": 15094
+    },
+    {
+      "epoch": 4.007168458781362,
+      "grad_norm": 0.466727683258946,
+      "learning_rate": 1.250705558190209e-06,
+      "loss": 0.5452,
+      "step": 15095
+    },
+    {
+      "epoch": 4.0074339572547455,
+      "grad_norm": 0.4694928059112087,
+      "learning_rate": 1.2504031598614552e-06,
+      "loss": 0.522,
+      "step": 15096
+    },
+    {
+      "epoch": 4.00769945572813,
+      "grad_norm": 0.47282426766526553,
+      "learning_rate": 1.2501007859029e-06,
+      "loss": 0.5473,
+      "step": 15097
+    },
+    {
+      "epoch": 4.007964954201514,
+      "grad_norm": 0.4700057955716974,
+      "learning_rate": 1.2497984363204397e-06,
+      "loss": 0.5521,
+      "step": 15098
+    },
+    {
+      "epoch": 4.008230452674897,
+      "grad_norm": 0.4748939460167583,
+      "learning_rate": 1.24949611111997e-06,
+      "loss": 0.5592,
+      "step": 15099
+    },
+    {
+      "epoch": 4.008495951148281,
+      "grad_norm": 0.4658720456268867,
+      "learning_rate": 1.249193810307389e-06,
+      "loss": 0.5385,
+      "step": 15100
+    },
+    {
+      "epoch": 4.008761449621665,
+      "grad_norm": 0.46925426369022755,
+      "learning_rate": 1.2488915338885898e-06,
+      "loss": 0.5587,
+      "step": 15101
+    },
+    {
+      "epoch": 4.009026948095048,
+      "grad_norm": 0.4739052126587744,
+      "learning_rate": 1.2485892818694704e-06,
+      "loss": 0.5338,
+      "step": 15102
+    },
+    {
+      "epoch": 4.009292446568432,
+      "grad_norm": 0.46871425812520534,
+      "learning_rate": 1.2482870542559236e-06,
+      "loss": 0.505,
+      "step": 15103
+    },
+    {
+      "epoch": 4.009557945041816,
+      "grad_norm": 0.48031954924964076,
+      "learning_rate": 1.2479848510538441e-06,
+      "loss": 0.5658,
+      "step": 15104
+    },
+    {
+      "epoch": 4.0098234435152,
+      "grad_norm": 0.4663280610617128,
+      "learning_rate": 1.2476826722691246e-06,
+      "loss": 0.5169,
+      "step": 15105
+    },
+    {
+      "epoch": 4.010088941988584,
+      "grad_norm": 0.48017454406500715,
+      "learning_rate": 1.2473805179076602e-06,
+      "loss": 0.5135,
+      "step": 15106
+    },
+    {
+      "epoch": 4.010354440461968,
+      "grad_norm": 0.4582409422880802,
+      "learning_rate": 1.2470783879753414e-06,
+      "loss": 0.4932,
+      "step": 15107
+    },
+    {
+      "epoch": 4.010619938935351,
+      "grad_norm": 0.4767150972651666,
+      "learning_rate": 1.246776282478063e-06,
+      "loss": 0.5328,
+      "step": 15108
+    },
+    {
+      "epoch": 4.010885437408735,
+      "grad_norm": 0.4710208021070399,
+      "learning_rate": 1.2464742014217154e-06,
+      "loss": 0.5288,
+      "step": 15109
+    },
+    {
+      "epoch": 4.011150935882118,
+      "grad_norm": 0.4479582410856575,
+      "learning_rate": 1.2461721448121892e-06,
+      "loss": 0.497,
+      "step": 15110
+    },
+    {
+      "epoch": 4.011416434355502,
+      "grad_norm": 0.4666377704642966,
+      "learning_rate": 1.245870112655377e-06,
+      "loss": 0.5223,
+      "step": 15111
+    },
+    {
+      "epoch": 4.011681932828886,
+      "grad_norm": 0.4743029599968229,
+      "learning_rate": 1.2455681049571668e-06,
+      "loss": 0.5302,
+      "step": 15112
+    },
+    {
+      "epoch": 4.01194743130227,
+      "grad_norm": 0.46113639576866716,
+      "learning_rate": 1.2452661217234513e-06,
+      "loss": 0.5272,
+      "step": 15113
+    },
+    {
+      "epoch": 4.012212929775654,
+      "grad_norm": 0.45353049843610055,
+      "learning_rate": 1.2449641629601186e-06,
+      "loss": 0.5247,
+      "step": 15114
+    },
+    {
+      "epoch": 4.012478428249038,
+      "grad_norm": 0.482724336445895,
+      "learning_rate": 1.2446622286730565e-06,
+      "loss": 0.5622,
+      "step": 15115
+    },
+    {
+      "epoch": 4.012743926722421,
+      "grad_norm": 0.4593623862960477,
+      "learning_rate": 1.2443603188681556e-06,
+      "loss": 0.5417,
+      "step": 15116
+    },
+    {
+      "epoch": 4.013009425195805,
+      "grad_norm": 0.48565089234772596,
+      "learning_rate": 1.2440584335513017e-06,
+      "loss": 0.5334,
+      "step": 15117
+    },
+    {
+      "epoch": 4.013274923669189,
+      "grad_norm": 0.4694511639189563,
+      "learning_rate": 1.2437565727283847e-06,
+      "loss": 0.5215,
+      "step": 15118
+    },
+    {
+      "epoch": 4.013540422142572,
+      "grad_norm": 0.4795164205350878,
+      "learning_rate": 1.243454736405289e-06,
+      "loss": 0.525,
+      "step": 15119
+    },
+    {
+      "epoch": 4.013805920615956,
+      "grad_norm": 0.46760623786188343,
+      "learning_rate": 1.2431529245879045e-06,
+      "loss": 0.5537,
+      "step": 15120
+    },
+    {
+      "epoch": 4.0140714190893405,
+      "grad_norm": 0.4650283214255255,
+      "learning_rate": 1.2428511372821133e-06,
+      "loss": 0.5345,
+      "step": 15121
+    },
+    {
+      "epoch": 4.014336917562724,
+      "grad_norm": 0.4582910710954033,
+      "learning_rate": 1.242549374493804e-06,
+      "loss": 0.5226,
+      "step": 15122
+    },
+    {
+      "epoch": 4.014602416036108,
+      "grad_norm": 0.45872486497099324,
+      "learning_rate": 1.24224763622886e-06,
+      "loss": 0.5354,
+      "step": 15123
+    },
+    {
+      "epoch": 4.014867914509492,
+      "grad_norm": 0.47317562291897564,
+      "learning_rate": 1.241945922493167e-06,
+      "loss": 0.5488,
+      "step": 15124
+    },
+    {
+      "epoch": 4.015133412982875,
+      "grad_norm": 0.4632405809174931,
+      "learning_rate": 1.2416442332926081e-06,
+      "loss": 0.5217,
+      "step": 15125
+    },
+    {
+      "epoch": 4.015398911456259,
+      "grad_norm": 0.4828355395608374,
+      "learning_rate": 1.2413425686330687e-06,
+      "loss": 0.5224,
+      "step": 15126
+    },
+    {
+      "epoch": 4.015664409929643,
+      "grad_norm": 0.46521468550106915,
+      "learning_rate": 1.2410409285204309e-06,
+      "loss": 0.5824,
+      "step": 15127
+    },
+    {
+      "epoch": 4.015929908403026,
+      "grad_norm": 0.4609770207623812,
+      "learning_rate": 1.2407393129605764e-06,
+      "loss": 0.5386,
+      "step": 15128
+    },
+    {
+      "epoch": 4.0161954068764105,
+      "grad_norm": 0.4727625267966508,
+      "learning_rate": 1.2404377219593895e-06,
+      "loss": 0.5392,
+      "step": 15129
+    },
+    {
+      "epoch": 4.016460905349795,
+      "grad_norm": 0.4572189493513875,
+      "learning_rate": 1.2401361555227497e-06,
+      "loss": 0.5321,
+      "step": 15130
+    },
+    {
+      "epoch": 4.016726403823178,
+      "grad_norm": 0.4690048434692796,
+      "learning_rate": 1.239834613656542e-06,
+      "loss": 0.5074,
+      "step": 15131
+    },
+    {
+      "epoch": 4.016991902296562,
+      "grad_norm": 0.4808546755740111,
+      "learning_rate": 1.2395330963666424e-06,
+      "loss": 0.5854,
+      "step": 15132
+    },
+    {
+      "epoch": 4.017257400769946,
+      "grad_norm": 0.46795790230096174,
+      "learning_rate": 1.2392316036589348e-06,
+      "loss": 0.4958,
+      "step": 15133
+    },
+    {
+      "epoch": 4.017522899243329,
+      "grad_norm": 0.4572411191246579,
+      "learning_rate": 1.2389301355392968e-06,
+      "loss": 0.5234,
+      "step": 15134
+    },
+    {
+      "epoch": 4.017788397716713,
+      "grad_norm": 0.4737647827584667,
+      "learning_rate": 1.2386286920136086e-06,
+      "loss": 0.5432,
+      "step": 15135
+    },
+    {
+      "epoch": 4.018053896190097,
+      "grad_norm": 0.47081635468680044,
+      "learning_rate": 1.2383272730877505e-06,
+      "loss": 0.5435,
+      "step": 15136
+    },
+    {
+      "epoch": 4.0183193946634805,
+      "grad_norm": 0.47403332627906536,
+      "learning_rate": 1.2380258787675994e-06,
+      "loss": 0.5403,
+      "step": 15137
+    },
+    {
+      "epoch": 4.018584893136865,
+      "grad_norm": 0.4787354795810244,
+      "learning_rate": 1.237724509059034e-06,
+      "loss": 0.5667,
+      "step": 15138
+    },
+    {
+      "epoch": 4.018850391610248,
+      "grad_norm": 0.4631558901475824,
+      "learning_rate": 1.2374231639679298e-06,
+      "loss": 0.5407,
+      "step": 15139
+    },
+    {
+      "epoch": 4.019115890083632,
+      "grad_norm": 0.4698161161490426,
+      "learning_rate": 1.2371218435001663e-06,
+      "loss": 0.5459,
+      "step": 15140
+    },
+    {
+      "epoch": 4.019381388557016,
+      "grad_norm": 0.4717195378694964,
+      "learning_rate": 1.2368205476616183e-06,
+      "loss": 0.5237,
+      "step": 15141
+    },
+    {
+      "epoch": 4.019646887030399,
+      "grad_norm": 0.4537909073316982,
+      "learning_rate": 1.2365192764581629e-06,
+      "loss": 0.4669,
+      "step": 15142
+    },
+    {
+      "epoch": 4.019912385503783,
+      "grad_norm": 0.4721326414633325,
+      "learning_rate": 1.2362180298956753e-06,
+      "loss": 0.5559,
+      "step": 15143
+    },
+    {
+      "epoch": 4.020177883977167,
+      "grad_norm": 0.46193782272043576,
+      "learning_rate": 1.2359168079800296e-06,
+      "loss": 0.5116,
+      "step": 15144
+    },
+    {
+      "epoch": 4.0204433824505506,
+      "grad_norm": 0.48091718268308714,
+      "learning_rate": 1.2356156107171019e-06,
+      "loss": 0.5592,
+      "step": 15145
+    },
+    {
+      "epoch": 4.020708880923935,
+      "grad_norm": 0.470809424251723,
+      "learning_rate": 1.2353144381127646e-06,
+      "loss": 0.5357,
+      "step": 15146
+    },
+    {
+      "epoch": 4.020974379397319,
+      "grad_norm": 0.4566358104233879,
+      "learning_rate": 1.2350132901728934e-06,
+      "loss": 0.4914,
+      "step": 15147
+    },
+    {
+      "epoch": 4.021239877870702,
+      "grad_norm": 0.47403974634995716,
+      "learning_rate": 1.234712166903359e-06,
+      "loss": 0.4885,
+      "step": 15148
+    },
+    {
+      "epoch": 4.021505376344086,
+      "grad_norm": 0.4812993924959656,
+      "learning_rate": 1.2344110683100375e-06,
+      "loss": 0.5493,
+      "step": 15149
+    },
+    {
+      "epoch": 4.02177087481747,
+      "grad_norm": 0.4767209232024363,
+      "learning_rate": 1.2341099943987969e-06,
+      "loss": 0.5397,
+      "step": 15150
+    },
+    {
+      "epoch": 4.022036373290853,
+      "grad_norm": 0.4911823060173162,
+      "learning_rate": 1.2338089451755117e-06,
+      "loss": 0.5044,
+      "step": 15151
+    },
+    {
+      "epoch": 4.022301871764237,
+      "grad_norm": 0.49424609422786137,
+      "learning_rate": 1.2335079206460512e-06,
+      "loss": 0.5445,
+      "step": 15152
+    },
+    {
+      "epoch": 4.0225673702376215,
+      "grad_norm": 0.4806444626602993,
+      "learning_rate": 1.2332069208162884e-06,
+      "loss": 0.5049,
+      "step": 15153
+    },
+    {
+      "epoch": 4.022832868711005,
+      "grad_norm": 0.47078534670571326,
+      "learning_rate": 1.232905945692092e-06,
+      "loss": 0.501,
+      "step": 15154
+    },
+    {
+      "epoch": 4.023098367184389,
+      "grad_norm": 0.4784174344647439,
+      "learning_rate": 1.2326049952793312e-06,
+      "loss": 0.5632,
+      "step": 15155
+    },
+    {
+      "epoch": 4.023363865657773,
+      "grad_norm": 0.4615727309926617,
+      "learning_rate": 1.232304069583877e-06,
+      "loss": 0.5103,
+      "step": 15156
+    },
+    {
+      "epoch": 4.023629364131156,
+      "grad_norm": 0.4862371864053715,
+      "learning_rate": 1.2320031686115966e-06,
+      "loss": 0.56,
+      "step": 15157
+    },
+    {
+      "epoch": 4.02389486260454,
+      "grad_norm": 0.47448233343347956,
+      "learning_rate": 1.2317022923683596e-06,
+      "loss": 0.5233,
+      "step": 15158
+    },
+    {
+      "epoch": 4.024160361077924,
+      "grad_norm": 0.4706810586303399,
+      "learning_rate": 1.2314014408600325e-06,
+      "loss": 0.5419,
+      "step": 15159
+    },
+    {
+      "epoch": 4.024425859551307,
+      "grad_norm": 0.4662326618062684,
+      "learning_rate": 1.2311006140924842e-06,
+      "loss": 0.5183,
+      "step": 15160
+    },
+    {
+      "epoch": 4.0246913580246915,
+      "grad_norm": 0.4875483424826858,
+      "learning_rate": 1.2307998120715807e-06,
+      "loss": 0.5652,
+      "step": 15161
+    },
+    {
+      "epoch": 4.0249568564980756,
+      "grad_norm": 0.46283564302730323,
+      "learning_rate": 1.2304990348031876e-06,
+      "loss": 0.5147,
+      "step": 15162
+    },
+    {
+      "epoch": 4.025222354971459,
+      "grad_norm": 0.47172967620604617,
+      "learning_rate": 1.2301982822931723e-06,
+      "loss": 0.5333,
+      "step": 15163
+    },
+    {
+      "epoch": 4.025487853444843,
+      "grad_norm": 0.4653283732356548,
+      "learning_rate": 1.229897554547399e-06,
+      "loss": 0.5439,
+      "step": 15164
+    },
+    {
+      "epoch": 4.025753351918226,
+      "grad_norm": 0.47772153127868505,
+      "learning_rate": 1.2295968515717339e-06,
+      "loss": 0.5441,
+      "step": 15165
+    },
+    {
+      "epoch": 4.02601885039161,
+      "grad_norm": 0.4552695518886499,
+      "learning_rate": 1.2292961733720404e-06,
+      "loss": 0.5113,
+      "step": 15166
+    },
+    {
+      "epoch": 4.026284348864994,
+      "grad_norm": 0.487452889726155,
+      "learning_rate": 1.228995519954183e-06,
+      "loss": 0.5379,
+      "step": 15167
+    },
+    {
+      "epoch": 4.026549847338377,
+      "grad_norm": 0.4852489657333027,
+      "learning_rate": 1.228694891324024e-06,
+      "loss": 0.5366,
+      "step": 15168
+    },
+    {
+      "epoch": 4.0268153458117615,
+      "grad_norm": 0.4725465835954684,
+      "learning_rate": 1.228394287487428e-06,
+      "loss": 0.5572,
+      "step": 15169
+    },
+    {
+      "epoch": 4.027080844285146,
+      "grad_norm": 0.4799833013088205,
+      "learning_rate": 1.2280937084502562e-06,
+      "loss": 0.5192,
+      "step": 15170
+    },
+    {
+      "epoch": 4.027346342758529,
+      "grad_norm": 0.46122959950651277,
+      "learning_rate": 1.227793154218372e-06,
+      "loss": 0.5245,
+      "step": 15171
+    },
+    {
+      "epoch": 4.027611841231913,
+      "grad_norm": 0.4681215278308655,
+      "learning_rate": 1.2274926247976363e-06,
+      "loss": 0.5274,
+      "step": 15172
+    },
+    {
+      "epoch": 4.027877339705297,
+      "grad_norm": 0.4803107173108429,
+      "learning_rate": 1.2271921201939088e-06,
+      "loss": 0.4897,
+      "step": 15173
+    },
+    {
+      "epoch": 4.02814283817868,
+      "grad_norm": 0.48259089077755596,
+      "learning_rate": 1.2268916404130527e-06,
+      "loss": 0.5689,
+      "step": 15174
+    },
+    {
+      "epoch": 4.028408336652064,
+      "grad_norm": 0.47767060409598194,
+      "learning_rate": 1.2265911854609256e-06,
+      "loss": 0.5282,
+      "step": 15175
+    },
+    {
+      "epoch": 4.028673835125448,
+      "grad_norm": 0.463270902948231,
+      "learning_rate": 1.226290755343389e-06,
+      "loss": 0.5192,
+      "step": 15176
+    },
+    {
+      "epoch": 4.0289393335988315,
+      "grad_norm": 0.47500790849470903,
+      "learning_rate": 1.2259903500663004e-06,
+      "loss": 0.5794,
+      "step": 15177
+    },
+    {
+      "epoch": 4.029204832072216,
+      "grad_norm": 0.473154625041754,
+      "learning_rate": 1.2256899696355213e-06,
+      "loss": 0.571,
+      "step": 15178
+    },
+    {
+      "epoch": 4.0294703305456,
+      "grad_norm": 0.47431211324842987,
+      "learning_rate": 1.225389614056906e-06,
+      "loss": 0.5478,
+      "step": 15179
+    },
+    {
+      "epoch": 4.029735829018983,
+      "grad_norm": 0.47081352177226965,
+      "learning_rate": 1.2250892833363148e-06,
+      "loss": 0.5453,
+      "step": 15180
+    },
+    {
+      "epoch": 4.030001327492367,
+      "grad_norm": 0.4856017007307107,
+      "learning_rate": 1.2247889774796032e-06,
+      "loss": 0.5483,
+      "step": 15181
+    },
+    {
+      "epoch": 4.030266825965751,
+      "grad_norm": 0.45904397838686417,
+      "learning_rate": 1.2244886964926286e-06,
+      "loss": 0.533,
+      "step": 15182
+    },
+    {
+      "epoch": 4.030532324439134,
+      "grad_norm": 0.48175309887582607,
+      "learning_rate": 1.2241884403812496e-06,
+      "loss": 0.5575,
+      "step": 15183
+    },
+    {
+      "epoch": 4.030797822912518,
+      "grad_norm": 0.46019218586563576,
+      "learning_rate": 1.2238882091513177e-06,
+      "loss": 0.4939,
+      "step": 15184
+    },
+    {
+      "epoch": 4.031063321385902,
+      "grad_norm": 0.4772008950455075,
+      "learning_rate": 1.223588002808691e-06,
+      "loss": 0.5359,
+      "step": 15185
+    },
+    {
+      "epoch": 4.031328819859286,
+      "grad_norm": 0.46075610778062087,
+      "learning_rate": 1.2232878213592228e-06,
+      "loss": 0.5074,
+      "step": 15186
+    },
+    {
+      "epoch": 4.03159431833267,
+      "grad_norm": 0.44992203894422067,
+      "learning_rate": 1.2229876648087688e-06,
+      "loss": 0.5137,
+      "step": 15187
+    },
+    {
+      "epoch": 4.031859816806054,
+      "grad_norm": 0.47564738011028035,
+      "learning_rate": 1.222687533163181e-06,
+      "loss": 0.5565,
+      "step": 15188
+    },
+    {
+      "epoch": 4.032125315279437,
+      "grad_norm": 0.4712632496105434,
+      "learning_rate": 1.2223874264283148e-06,
+      "loss": 0.5017,
+      "step": 15189
+    },
+    {
+      "epoch": 4.032390813752821,
+      "grad_norm": 0.4786331251356405,
+      "learning_rate": 1.2220873446100219e-06,
+      "loss": 0.5111,
+      "step": 15190
+    },
+    {
+      "epoch": 4.032656312226205,
+      "grad_norm": 0.44860954709131745,
+      "learning_rate": 1.2217872877141534e-06,
+      "loss": 0.498,
+      "step": 15191
+    },
+    {
+      "epoch": 4.032921810699588,
+      "grad_norm": 0.46175062564963604,
+      "learning_rate": 1.2214872557465637e-06,
+      "loss": 0.5144,
+      "step": 15192
+    },
+    {
+      "epoch": 4.033187309172972,
+      "grad_norm": 0.47049623638302096,
+      "learning_rate": 1.2211872487131018e-06,
+      "loss": 0.5324,
+      "step": 15193
+    },
+    {
+      "epoch": 4.033452807646356,
+      "grad_norm": 0.4812756516586452,
+      "learning_rate": 1.2208872666196205e-06,
+      "loss": 0.5196,
+      "step": 15194
+    },
+    {
+      "epoch": 4.03371830611974,
+      "grad_norm": 0.4635024827044325,
+      "learning_rate": 1.2205873094719695e-06,
+      "loss": 0.5659,
+      "step": 15195
+    },
+    {
+      "epoch": 4.033983804593124,
+      "grad_norm": 0.47225619901555776,
+      "learning_rate": 1.2202873772759983e-06,
+      "loss": 0.5476,
+      "step": 15196
+    },
+    {
+      "epoch": 4.034249303066507,
+      "grad_norm": 0.4682788135016163,
+      "learning_rate": 1.2199874700375558e-06,
+      "loss": 0.5342,
+      "step": 15197
+    },
+    {
+      "epoch": 4.034514801539891,
+      "grad_norm": 0.47748398198053876,
+      "learning_rate": 1.2196875877624922e-06,
+      "loss": 0.5418,
+      "step": 15198
+    },
+    {
+      "epoch": 4.034780300013275,
+      "grad_norm": 0.4800900381401272,
+      "learning_rate": 1.2193877304566548e-06,
+      "loss": 0.5446,
+      "step": 15199
+    },
+    {
+      "epoch": 4.035045798486658,
+      "grad_norm": 0.46530528291875894,
+      "learning_rate": 1.219087898125893e-06,
+      "loss": 0.533,
+      "step": 15200
+    },
+    {
+      "epoch": 4.035311296960042,
+      "grad_norm": 0.4676566056172326,
+      "learning_rate": 1.2187880907760532e-06,
+      "loss": 0.5186,
+      "step": 15201
+    },
+    {
+      "epoch": 4.0355767954334265,
+      "grad_norm": 0.46970752411790573,
+      "learning_rate": 1.218488308412982e-06,
+      "loss": 0.5582,
+      "step": 15202
+    },
+    {
+      "epoch": 4.03584229390681,
+      "grad_norm": 0.4831752456650113,
+      "learning_rate": 1.218188551042527e-06,
+      "loss": 0.549,
+      "step": 15203
+    },
+    {
+      "epoch": 4.036107792380194,
+      "grad_norm": 0.46515161664200344,
+      "learning_rate": 1.2178888186705328e-06,
+      "loss": 0.5062,
+      "step": 15204
+    },
+    {
+      "epoch": 4.036373290853578,
+      "grad_norm": 0.4737898749217837,
+      "learning_rate": 1.217589111302847e-06,
+      "loss": 0.5219,
+      "step": 15205
+    },
+    {
+      "epoch": 4.036638789326961,
+      "grad_norm": 0.4707188988826055,
+      "learning_rate": 1.2172894289453133e-06,
+      "loss": 0.5617,
+      "step": 15206
+    },
+    {
+      "epoch": 4.036904287800345,
+      "grad_norm": 0.4717745918345947,
+      "learning_rate": 1.2169897716037755e-06,
+      "loss": 0.5021,
+      "step": 15207
+    },
+    {
+      "epoch": 4.037169786273729,
+      "grad_norm": 0.47983557361231594,
+      "learning_rate": 1.216690139284079e-06,
+      "loss": 0.483,
+      "step": 15208
+    },
+    {
+      "epoch": 4.037435284747112,
+      "grad_norm": 0.46951526672174815,
+      "learning_rate": 1.2163905319920666e-06,
+      "loss": 0.5508,
+      "step": 15209
+    },
+    {
+      "epoch": 4.0377007832204965,
+      "grad_norm": 0.47225471633764826,
+      "learning_rate": 1.2160909497335822e-06,
+      "loss": 0.5614,
+      "step": 15210
+    },
+    {
+      "epoch": 4.037966281693881,
+      "grad_norm": 0.4677476702573962,
+      "learning_rate": 1.2157913925144668e-06,
+      "loss": 0.5165,
+      "step": 15211
+    },
+    {
+      "epoch": 4.038231780167264,
+      "grad_norm": 0.47124823412181144,
+      "learning_rate": 1.2154918603405654e-06,
+      "loss": 0.5284,
+      "step": 15212
+    },
+    {
+      "epoch": 4.038497278640648,
+      "grad_norm": 0.4683950895334416,
+      "learning_rate": 1.2151923532177162e-06,
+      "loss": 0.5627,
+      "step": 15213
+    },
+    {
+      "epoch": 4.038762777114032,
+      "grad_norm": 0.4810815170116101,
+      "learning_rate": 1.2148928711517628e-06,
+      "loss": 0.5454,
+      "step": 15214
+    },
+    {
+      "epoch": 4.039028275587415,
+      "grad_norm": 0.46453213325088966,
+      "learning_rate": 1.2145934141485436e-06,
+      "loss": 0.5199,
+      "step": 15215
+    },
+    {
+      "epoch": 4.039293774060799,
+      "grad_norm": 0.46572784396581357,
+      "learning_rate": 1.214293982213901e-06,
+      "loss": 0.5085,
+      "step": 15216
+    },
+    {
+      "epoch": 4.039559272534183,
+      "grad_norm": 0.45327669140427373,
+      "learning_rate": 1.213994575353673e-06,
+      "loss": 0.5039,
+      "step": 15217
+    },
+    {
+      "epoch": 4.0398247710075665,
+      "grad_norm": 0.46761248484343776,
+      "learning_rate": 1.2136951935737002e-06,
+      "loss": 0.5478,
+      "step": 15218
+    },
+    {
+      "epoch": 4.040090269480951,
+      "grad_norm": 0.4948793322285906,
+      "learning_rate": 1.2133958368798208e-06,
+      "loss": 0.5321,
+      "step": 15219
+    },
+    {
+      "epoch": 4.040355767954335,
+      "grad_norm": 0.47296897417976336,
+      "learning_rate": 1.2130965052778715e-06,
+      "loss": 0.5476,
+      "step": 15220
+    },
+    {
+      "epoch": 4.040621266427718,
+      "grad_norm": 0.4773097338499646,
+      "learning_rate": 1.2127971987736922e-06,
+      "loss": 0.5189,
+      "step": 15221
+    },
+    {
+      "epoch": 4.040886764901102,
+      "grad_norm": 0.4825265851453496,
+      "learning_rate": 1.2124979173731186e-06,
+      "loss": 0.5337,
+      "step": 15222
+    },
+    {
+      "epoch": 4.041152263374485,
+      "grad_norm": 0.4783762641194874,
+      "learning_rate": 1.2121986610819885e-06,
+      "loss": 0.5392,
+      "step": 15223
+    },
+    {
+      "epoch": 4.041417761847869,
+      "grad_norm": 0.48612510042936663,
+      "learning_rate": 1.2118994299061376e-06,
+      "loss": 0.5446,
+      "step": 15224
+    },
+    {
+      "epoch": 4.041683260321253,
+      "grad_norm": 0.468994732260362,
+      "learning_rate": 1.2116002238514018e-06,
+      "loss": 0.5396,
+      "step": 15225
+    },
+    {
+      "epoch": 4.0419487587946366,
+      "grad_norm": 0.48934230686631847,
+      "learning_rate": 1.2113010429236152e-06,
+      "loss": 0.5394,
+      "step": 15226
+    },
+    {
+      "epoch": 4.042214257268021,
+      "grad_norm": 0.4882510660968333,
+      "learning_rate": 1.2110018871286136e-06,
+      "loss": 0.539,
+      "step": 15227
+    },
+    {
+      "epoch": 4.042479755741405,
+      "grad_norm": 0.47741061682579605,
+      "learning_rate": 1.210702756472232e-06,
+      "loss": 0.5056,
+      "step": 15228
+    },
+    {
+      "epoch": 4.042745254214788,
+      "grad_norm": 0.45013491291370716,
+      "learning_rate": 1.2104036509603035e-06,
+      "loss": 0.5027,
+      "step": 15229
+    },
+    {
+      "epoch": 4.043010752688172,
+      "grad_norm": 0.47532277911891085,
+      "learning_rate": 1.210104570598661e-06,
+      "loss": 0.4995,
+      "step": 15230
+    },
+    {
+      "epoch": 4.043276251161556,
+      "grad_norm": 0.49375967662526105,
+      "learning_rate": 1.209805515393137e-06,
+      "loss": 0.4771,
+      "step": 15231
+    },
+    {
+      "epoch": 4.043541749634939,
+      "grad_norm": 0.4692949273710238,
+      "learning_rate": 1.2095064853495654e-06,
+      "loss": 0.4713,
+      "step": 15232
+    },
+    {
+      "epoch": 4.043807248108323,
+      "grad_norm": 0.4715172167833221,
+      "learning_rate": 1.2092074804737758e-06,
+      "loss": 0.5732,
+      "step": 15233
+    },
+    {
+      "epoch": 4.0440727465817075,
+      "grad_norm": 0.46584287593035656,
+      "learning_rate": 1.2089085007716018e-06,
+      "loss": 0.5369,
+      "step": 15234
+    },
+    {
+      "epoch": 4.044338245055091,
+      "grad_norm": 0.47600923575858844,
+      "learning_rate": 1.2086095462488732e-06,
+      "loss": 0.5409,
+      "step": 15235
+    },
+    {
+      "epoch": 4.044603743528475,
+      "grad_norm": 0.4597264401641963,
+      "learning_rate": 1.2083106169114192e-06,
+      "loss": 0.537,
+      "step": 15236
+    },
+    {
+      "epoch": 4.044869242001859,
+      "grad_norm": 0.46813744482225095,
+      "learning_rate": 1.2080117127650718e-06,
+      "loss": 0.5358,
+      "step": 15237
+    },
+    {
+      "epoch": 4.045134740475242,
+      "grad_norm": 0.4608256052459822,
+      "learning_rate": 1.2077128338156582e-06,
+      "loss": 0.5394,
+      "step": 15238
+    },
+    {
+      "epoch": 4.045400238948626,
+      "grad_norm": 0.4626668247201617,
+      "learning_rate": 1.2074139800690096e-06,
+      "loss": 0.5302,
+      "step": 15239
+    },
+    {
+      "epoch": 4.04566573742201,
+      "grad_norm": 0.46870202786166637,
+      "learning_rate": 1.2071151515309523e-06,
+      "loss": 0.534,
+      "step": 15240
+    },
+    {
+      "epoch": 4.045931235895393,
+      "grad_norm": 0.4541865240216533,
+      "learning_rate": 1.2068163482073169e-06,
+      "loss": 0.5177,
+      "step": 15241
+    },
+    {
+      "epoch": 4.0461967343687775,
+      "grad_norm": 0.49038073174270946,
+      "learning_rate": 1.206517570103927e-06,
+      "loss": 0.5575,
+      "step": 15242
+    },
+    {
+      "epoch": 4.0464622328421616,
+      "grad_norm": 0.4684684004724672,
+      "learning_rate": 1.2062188172266124e-06,
+      "loss": 0.5337,
+      "step": 15243
+    },
+    {
+      "epoch": 4.046727731315545,
+      "grad_norm": 0.4667188547874859,
+      "learning_rate": 1.205920089581198e-06,
+      "loss": 0.5475,
+      "step": 15244
+    },
+    {
+      "epoch": 4.046993229788929,
+      "grad_norm": 0.46382710253369824,
+      "learning_rate": 1.2056213871735106e-06,
+      "loss": 0.527,
+      "step": 15245
+    },
+    {
+      "epoch": 4.047258728262313,
+      "grad_norm": 0.48021942292750724,
+      "learning_rate": 1.2053227100093749e-06,
+      "loss": 0.5203,
+      "step": 15246
+    },
+    {
+      "epoch": 4.047524226735696,
+      "grad_norm": 0.48472603190774366,
+      "learning_rate": 1.205024058094617e-06,
+      "loss": 0.5366,
+      "step": 15247
+    },
+    {
+      "epoch": 4.04778972520908,
+      "grad_norm": 0.46430059336920937,
+      "learning_rate": 1.2047254314350606e-06,
+      "loss": 0.5308,
+      "step": 15248
+    },
+    {
+      "epoch": 4.048055223682463,
+      "grad_norm": 0.46567218920106784,
+      "learning_rate": 1.204426830036529e-06,
+      "loss": 0.5113,
+      "step": 15249
+    },
+    {
+      "epoch": 4.0483207221558475,
+      "grad_norm": 0.4807050830318009,
+      "learning_rate": 1.204128253904847e-06,
+      "loss": 0.5574,
+      "step": 15250
+    },
+    {
+      "epoch": 4.048586220629232,
+      "grad_norm": 0.4743521584489147,
+      "learning_rate": 1.203829703045836e-06,
+      "loss": 0.5378,
+      "step": 15251
+    },
+    {
+      "epoch": 4.048851719102615,
+      "grad_norm": 0.4604919292489176,
+      "learning_rate": 1.2035311774653202e-06,
+      "loss": 0.5314,
+      "step": 15252
+    },
+    {
+      "epoch": 4.049117217575999,
+      "grad_norm": 0.47683426397981876,
+      "learning_rate": 1.2032326771691212e-06,
+      "loss": 0.5327,
+      "step": 15253
+    },
+    {
+      "epoch": 4.049382716049383,
+      "grad_norm": 0.4732988959960685,
+      "learning_rate": 1.2029342021630586e-06,
+      "loss": 0.5446,
+      "step": 15254
+    },
+    {
+      "epoch": 4.049648214522766,
+      "grad_norm": 0.45847991352379697,
+      "learning_rate": 1.2026357524529558e-06,
+      "loss": 0.536,
+      "step": 15255
+    },
+    {
+      "epoch": 4.04991371299615,
+      "grad_norm": 0.49095872625856923,
+      "learning_rate": 1.2023373280446316e-06,
+      "loss": 0.56,
+      "step": 15256
+    },
+    {
+      "epoch": 4.050179211469534,
+      "grad_norm": 0.48524195458709124,
+      "learning_rate": 1.2020389289439075e-06,
+      "loss": 0.5649,
+      "step": 15257
+    },
+    {
+      "epoch": 4.0504447099429175,
+      "grad_norm": 0.4820878099628291,
+      "learning_rate": 1.2017405551566021e-06,
+      "loss": 0.5712,
+      "step": 15258
+    },
+    {
+      "epoch": 4.050710208416302,
+      "grad_norm": 0.46590308238229877,
+      "learning_rate": 1.2014422066885351e-06,
+      "loss": 0.5015,
+      "step": 15259
+    },
+    {
+      "epoch": 4.050975706889686,
+      "grad_norm": 0.4759747425476968,
+      "learning_rate": 1.201143883545523e-06,
+      "loss": 0.5319,
+      "step": 15260
+    },
+    {
+      "epoch": 4.051241205363069,
+      "grad_norm": 0.4613691980873959,
+      "learning_rate": 1.2008455857333864e-06,
+      "loss": 0.5009,
+      "step": 15261
+    },
+    {
+      "epoch": 4.051506703836453,
+      "grad_norm": 0.4858590990687217,
+      "learning_rate": 1.200547313257941e-06,
+      "loss": 0.5648,
+      "step": 15262
+    },
+    {
+      "epoch": 4.051772202309837,
+      "grad_norm": 0.4626307351952384,
+      "learning_rate": 1.2002490661250054e-06,
+      "loss": 0.5422,
+      "step": 15263
+    },
+    {
+      "epoch": 4.05203770078322,
+      "grad_norm": 0.4822199844852934,
+      "learning_rate": 1.1999508443403954e-06,
+      "loss": 0.5568,
+      "step": 15264
+    },
+    {
+      "epoch": 4.052303199256604,
+      "grad_norm": 0.4775558199724062,
+      "learning_rate": 1.199652647909926e-06,
+      "loss": 0.5282,
+      "step": 15265
+    },
+    {
+      "epoch": 4.052568697729988,
+      "grad_norm": 0.4724474403137747,
+      "learning_rate": 1.1993544768394147e-06,
+      "loss": 0.5232,
+      "step": 15266
+    },
+    {
+      "epoch": 4.052834196203372,
+      "grad_norm": 0.47124186596810175,
+      "learning_rate": 1.1990563311346748e-06,
+      "loss": 0.5595,
+      "step": 15267
+    },
+    {
+      "epoch": 4.053099694676756,
+      "grad_norm": 0.4699839949251924,
+      "learning_rate": 1.1987582108015228e-06,
+      "loss": 0.5351,
+      "step": 15268
+    },
+    {
+      "epoch": 4.05336519315014,
+      "grad_norm": 0.4753557568539712,
+      "learning_rate": 1.1984601158457703e-06,
+      "loss": 0.5396,
+      "step": 15269
+    },
+    {
+      "epoch": 4.053630691623523,
+      "grad_norm": 0.4763047255567218,
+      "learning_rate": 1.1981620462732344e-06,
+      "loss": 0.5414,
+      "step": 15270
+    },
+    {
+      "epoch": 4.053896190096907,
+      "grad_norm": 0.4767971089431187,
+      "learning_rate": 1.1978640020897242e-06,
+      "loss": 0.5444,
+      "step": 15271
+    },
+    {
+      "epoch": 4.054161688570291,
+      "grad_norm": 0.4681479133829081,
+      "learning_rate": 1.1975659833010539e-06,
+      "loss": 0.5483,
+      "step": 15272
+    },
+    {
+      "epoch": 4.054427187043674,
+      "grad_norm": 0.47090736788337534,
+      "learning_rate": 1.1972679899130371e-06,
+      "loss": 0.5622,
+      "step": 15273
+    },
+    {
+      "epoch": 4.054692685517058,
+      "grad_norm": 0.46396486852501095,
+      "learning_rate": 1.1969700219314831e-06,
+      "loss": 0.5122,
+      "step": 15274
+    },
+    {
+      "epoch": 4.054958183990442,
+      "grad_norm": 0.4745448359019339,
+      "learning_rate": 1.1966720793622048e-06,
+      "loss": 0.5469,
+      "step": 15275
+    },
+    {
+      "epoch": 4.055223682463826,
+      "grad_norm": 0.4626784809823271,
+      "learning_rate": 1.1963741622110122e-06,
+      "loss": 0.5214,
+      "step": 15276
+    },
+    {
+      "epoch": 4.05548918093721,
+      "grad_norm": 0.46653088299621404,
+      "learning_rate": 1.196076270483715e-06,
+      "loss": 0.5549,
+      "step": 15277
+    },
+    {
+      "epoch": 4.055754679410593,
+      "grad_norm": 0.4707670261389961,
+      "learning_rate": 1.1957784041861226e-06,
+      "loss": 0.5596,
+      "step": 15278
+    },
+    {
+      "epoch": 4.056020177883977,
+      "grad_norm": 0.48157785941842957,
+      "learning_rate": 1.195480563324045e-06,
+      "loss": 0.5513,
+      "step": 15279
+    },
+    {
+      "epoch": 4.056285676357361,
+      "grad_norm": 0.46439429571247676,
+      "learning_rate": 1.1951827479032901e-06,
+      "loss": 0.5364,
+      "step": 15280
+    },
+    {
+      "epoch": 4.056551174830744,
+      "grad_norm": 0.4735868234163533,
+      "learning_rate": 1.194884957929667e-06,
+      "loss": 0.5159,
+      "step": 15281
+    },
+    {
+      "epoch": 4.056816673304128,
+      "grad_norm": 0.48221284524780833,
+      "learning_rate": 1.1945871934089826e-06,
+      "loss": 0.566,
+      "step": 15282
+    },
+    {
+      "epoch": 4.0570821717775125,
+      "grad_norm": 0.4886720479988301,
+      "learning_rate": 1.1942894543470434e-06,
+      "loss": 0.505,
+      "step": 15283
+    },
+    {
+      "epoch": 4.057347670250896,
+      "grad_norm": 0.46084531172381,
+      "learning_rate": 1.1939917407496576e-06,
+      "loss": 0.5152,
+      "step": 15284
+    },
+    {
+      "epoch": 4.05761316872428,
+      "grad_norm": 0.48300248616089253,
+      "learning_rate": 1.1936940526226295e-06,
+      "loss": 0.5528,
+      "step": 15285
+    },
+    {
+      "epoch": 4.057878667197664,
+      "grad_norm": 0.47918006766306404,
+      "learning_rate": 1.1933963899717668e-06,
+      "loss": 0.562,
+      "step": 15286
+    },
+    {
+      "epoch": 4.058144165671047,
+      "grad_norm": 0.46790617448347743,
+      "learning_rate": 1.1930987528028737e-06,
+      "loss": 0.5491,
+      "step": 15287
+    },
+    {
+      "epoch": 4.058409664144431,
+      "grad_norm": 0.4756146151115185,
+      "learning_rate": 1.1928011411217545e-06,
+      "loss": 0.5009,
+      "step": 15288
+    },
+    {
+      "epoch": 4.058675162617815,
+      "grad_norm": 0.4689201213308079,
+      "learning_rate": 1.192503554934213e-06,
+      "loss": 0.5393,
+      "step": 15289
+    },
+    {
+      "epoch": 4.058940661091198,
+      "grad_norm": 0.4767789691691604,
+      "learning_rate": 1.1922059942460539e-06,
+      "loss": 0.5716,
+      "step": 15290
+    },
+    {
+      "epoch": 4.0592061595645825,
+      "grad_norm": 0.4782801756543797,
+      "learning_rate": 1.1919084590630794e-06,
+      "loss": 0.5178,
+      "step": 15291
+    },
+    {
+      "epoch": 4.059471658037967,
+      "grad_norm": 0.46368782840948475,
+      "learning_rate": 1.1916109493910937e-06,
+      "loss": 0.482,
+      "step": 15292
+    },
+    {
+      "epoch": 4.05973715651135,
+      "grad_norm": 0.4631731760114486,
+      "learning_rate": 1.191313465235898e-06,
+      "loss": 0.5461,
+      "step": 15293
+    },
+    {
+      "epoch": 4.060002654984734,
+      "grad_norm": 0.4811916201362852,
+      "learning_rate": 1.1910160066032928e-06,
+      "loss": 0.537,
+      "step": 15294
+    },
+    {
+      "epoch": 4.060268153458118,
+      "grad_norm": 0.4648490441782714,
+      "learning_rate": 1.1907185734990814e-06,
+      "loss": 0.5052,
+      "step": 15295
+    },
+    {
+      "epoch": 4.060533651931501,
+      "grad_norm": 0.4848048504538911,
+      "learning_rate": 1.1904211659290627e-06,
+      "loss": 0.5461,
+      "step": 15296
+    },
+    {
+      "epoch": 4.060799150404885,
+      "grad_norm": 0.4651667451206629,
+      "learning_rate": 1.1901237838990385e-06,
+      "loss": 0.5357,
+      "step": 15297
+    },
+    {
+      "epoch": 4.061064648878269,
+      "grad_norm": 0.477377319462052,
+      "learning_rate": 1.1898264274148067e-06,
+      "loss": 0.5214,
+      "step": 15298
+    },
+    {
+      "epoch": 4.0613301473516525,
+      "grad_norm": 0.4872921338439011,
+      "learning_rate": 1.1895290964821685e-06,
+      "loss": 0.5331,
+      "step": 15299
+    },
+    {
+      "epoch": 4.061595645825037,
+      "grad_norm": 0.46819902409357517,
+      "learning_rate": 1.1892317911069212e-06,
+      "loss": 0.5272,
+      "step": 15300
+    },
+    {
+      "epoch": 4.061861144298421,
+      "grad_norm": 0.4630313282628995,
+      "learning_rate": 1.1889345112948624e-06,
+      "loss": 0.546,
+      "step": 15301
+    },
+    {
+      "epoch": 4.062126642771804,
+      "grad_norm": 0.4821826806296449,
+      "learning_rate": 1.1886372570517917e-06,
+      "loss": 0.518,
+      "step": 15302
+    },
+    {
+      "epoch": 4.062392141245188,
+      "grad_norm": 0.47594826621935443,
+      "learning_rate": 1.1883400283835044e-06,
+      "loss": 0.5247,
+      "step": 15303
+    },
+    {
+      "epoch": 4.062657639718571,
+      "grad_norm": 0.4833855245415668,
+      "learning_rate": 1.1880428252958001e-06,
+      "loss": 0.5878,
+      "step": 15304
+    },
+    {
+      "epoch": 4.062923138191955,
+      "grad_norm": 0.45649196946309656,
+      "learning_rate": 1.1877456477944707e-06,
+      "loss": 0.5236,
+      "step": 15305
+    },
+    {
+      "epoch": 4.063188636665339,
+      "grad_norm": 0.4557553307661307,
+      "learning_rate": 1.1874484958853153e-06,
+      "loss": 0.5304,
+      "step": 15306
+    },
+    {
+      "epoch": 4.063454135138723,
+      "grad_norm": 0.4711309149466802,
+      "learning_rate": 1.187151369574127e-06,
+      "loss": 0.5337,
+      "step": 15307
+    },
+    {
+      "epoch": 4.063719633612107,
+      "grad_norm": 0.47590406561081133,
+      "learning_rate": 1.186854268866702e-06,
+      "loss": 0.503,
+      "step": 15308
+    },
+    {
+      "epoch": 4.063985132085491,
+      "grad_norm": 0.48984722589565893,
+      "learning_rate": 1.186557193768833e-06,
+      "loss": 0.54,
+      "step": 15309
+    },
+    {
+      "epoch": 4.064250630558874,
+      "grad_norm": 0.48796745471143543,
+      "learning_rate": 1.1862601442863158e-06,
+      "loss": 0.553,
+      "step": 15310
+    },
+    {
+      "epoch": 4.064516129032258,
+      "grad_norm": 0.4765259659436347,
+      "learning_rate": 1.1859631204249422e-06,
+      "loss": 0.5454,
+      "step": 15311
+    },
+    {
+      "epoch": 4.064781627505642,
+      "grad_norm": 0.47654842225866384,
+      "learning_rate": 1.1856661221905041e-06,
+      "loss": 0.5581,
+      "step": 15312
+    },
+    {
+      "epoch": 4.065047125979025,
+      "grad_norm": 0.48410421429554673,
+      "learning_rate": 1.1853691495887958e-06,
+      "loss": 0.5462,
+      "step": 15313
+    },
+    {
+      "epoch": 4.065312624452409,
+      "grad_norm": 0.47497710593538495,
+      "learning_rate": 1.1850722026256067e-06,
+      "loss": 0.5135,
+      "step": 15314
+    },
+    {
+      "epoch": 4.0655781229257935,
+      "grad_norm": 0.47271319237251674,
+      "learning_rate": 1.1847752813067304e-06,
+      "loss": 0.5448,
+      "step": 15315
+    },
+    {
+      "epoch": 4.065843621399177,
+      "grad_norm": 0.4861264911121132,
+      "learning_rate": 1.184478385637956e-06,
+      "loss": 0.5708,
+      "step": 15316
+    },
+    {
+      "epoch": 4.066109119872561,
+      "grad_norm": 0.4816718011082714,
+      "learning_rate": 1.184181515625074e-06,
+      "loss": 0.5158,
+      "step": 15317
+    },
+    {
+      "epoch": 4.066374618345945,
+      "grad_norm": 0.4826982993719964,
+      "learning_rate": 1.1838846712738735e-06,
+      "loss": 0.5214,
+      "step": 15318
+    },
+    {
+      "epoch": 4.066640116819328,
+      "grad_norm": 0.46762958161854784,
+      "learning_rate": 1.1835878525901443e-06,
+      "loss": 0.5266,
+      "step": 15319
+    },
+    {
+      "epoch": 4.066905615292712,
+      "grad_norm": 0.4791146882970028,
+      "learning_rate": 1.1832910595796757e-06,
+      "loss": 0.5227,
+      "step": 15320
+    },
+    {
+      "epoch": 4.067171113766096,
+      "grad_norm": 0.46856974905413734,
+      "learning_rate": 1.1829942922482545e-06,
+      "loss": 0.5296,
+      "step": 15321
+    },
+    {
+      "epoch": 4.067436612239479,
+      "grad_norm": 0.4640277367142785,
+      "learning_rate": 1.1826975506016714e-06,
+      "loss": 0.5401,
+      "step": 15322
+    },
+    {
+      "epoch": 4.0677021107128635,
+      "grad_norm": 0.4743643068830104,
+      "learning_rate": 1.182400834645709e-06,
+      "loss": 0.581,
+      "step": 15323
+    },
+    {
+      "epoch": 4.067967609186248,
+      "grad_norm": 0.4632200314572056,
+      "learning_rate": 1.1821041443861575e-06,
+      "loss": 0.5132,
+      "step": 15324
+    },
+    {
+      "epoch": 4.068233107659631,
+      "grad_norm": 0.46129276437946515,
+      "learning_rate": 1.1818074798288012e-06,
+      "loss": 0.5575,
+      "step": 15325
+    },
+    {
+      "epoch": 4.068498606133015,
+      "grad_norm": 0.4613027380576767,
+      "learning_rate": 1.1815108409794273e-06,
+      "loss": 0.5064,
+      "step": 15326
+    },
+    {
+      "epoch": 4.068764104606399,
+      "grad_norm": 0.4753145277392267,
+      "learning_rate": 1.1812142278438202e-06,
+      "loss": 0.5625,
+      "step": 15327
+    },
+    {
+      "epoch": 4.069029603079782,
+      "grad_norm": 0.4805050311831935,
+      "learning_rate": 1.1809176404277637e-06,
+      "loss": 0.5367,
+      "step": 15328
+    },
+    {
+      "epoch": 4.069295101553166,
+      "grad_norm": 0.47317748272457183,
+      "learning_rate": 1.1806210787370436e-06,
+      "loss": 0.5285,
+      "step": 15329
+    },
+    {
+      "epoch": 4.06956060002655,
+      "grad_norm": 0.46825559705124414,
+      "learning_rate": 1.1803245427774424e-06,
+      "loss": 0.5169,
+      "step": 15330
+    },
+    {
+      "epoch": 4.0698260984999335,
+      "grad_norm": 0.476465886374575,
+      "learning_rate": 1.1800280325547445e-06,
+      "loss": 0.5323,
+      "step": 15331
+    },
+    {
+      "epoch": 4.070091596973318,
+      "grad_norm": 0.4532279973895195,
+      "learning_rate": 1.1797315480747307e-06,
+      "loss": 0.4969,
+      "step": 15332
+    },
+    {
+      "epoch": 4.070357095446701,
+      "grad_norm": 0.4756226834657503,
+      "learning_rate": 1.179435089343186e-06,
+      "loss": 0.5649,
+      "step": 15333
+    },
+    {
+      "epoch": 4.070622593920085,
+      "grad_norm": 0.4495892052217501,
+      "learning_rate": 1.179138656365889e-06,
+      "loss": 0.5048,
+      "step": 15334
+    },
+    {
+      "epoch": 4.070888092393469,
+      "grad_norm": 0.4783500987244886,
+      "learning_rate": 1.1788422491486228e-06,
+      "loss": 0.4976,
+      "step": 15335
+    },
+    {
+      "epoch": 4.071153590866852,
+      "grad_norm": 0.4814531287656348,
+      "learning_rate": 1.1785458676971666e-06,
+      "loss": 0.5444,
+      "step": 15336
+    },
+    {
+      "epoch": 4.071419089340236,
+      "grad_norm": 0.4726424582963027,
+      "learning_rate": 1.1782495120173024e-06,
+      "loss": 0.57,
+      "step": 15337
+    },
+    {
+      "epoch": 4.07168458781362,
+      "grad_norm": 0.47045057779885807,
+      "learning_rate": 1.1779531821148083e-06,
+      "loss": 0.5419,
+      "step": 15338
+    },
+    {
+      "epoch": 4.0719500862870035,
+      "grad_norm": 0.4995612227957978,
+      "learning_rate": 1.1776568779954644e-06,
+      "loss": 0.579,
+      "step": 15339
+    },
+    {
+      "epoch": 4.072215584760388,
+      "grad_norm": 0.47865369747694886,
+      "learning_rate": 1.1773605996650495e-06,
+      "loss": 0.5494,
+      "step": 15340
+    },
+    {
+      "epoch": 4.072481083233772,
+      "grad_norm": 0.4756500319571732,
+      "learning_rate": 1.1770643471293403e-06,
+      "loss": 0.551,
+      "step": 15341
+    },
+    {
+      "epoch": 4.072746581707155,
+      "grad_norm": 0.47400874728396725,
+      "learning_rate": 1.1767681203941165e-06,
+      "loss": 0.5416,
+      "step": 15342
+    },
+    {
+      "epoch": 4.073012080180539,
+      "grad_norm": 0.45946739185190155,
+      "learning_rate": 1.1764719194651528e-06,
+      "loss": 0.5346,
+      "step": 15343
+    },
+    {
+      "epoch": 4.073277578653923,
+      "grad_norm": 0.47532188945572756,
+      "learning_rate": 1.1761757443482285e-06,
+      "loss": 0.5285,
+      "step": 15344
+    },
+    {
+      "epoch": 4.073543077127306,
+      "grad_norm": 0.4797388114666441,
+      "learning_rate": 1.1758795950491184e-06,
+      "loss": 0.5798,
+      "step": 15345
+    },
+    {
+      "epoch": 4.07380857560069,
+      "grad_norm": 0.485271505679119,
+      "learning_rate": 1.1755834715735973e-06,
+      "loss": 0.528,
+      "step": 15346
+    },
+    {
+      "epoch": 4.074074074074074,
+      "grad_norm": 0.47970389572301153,
+      "learning_rate": 1.175287373927442e-06,
+      "loss": 0.5382,
+      "step": 15347
+    },
+    {
+      "epoch": 4.074339572547458,
+      "grad_norm": 0.46593528882108654,
+      "learning_rate": 1.1749913021164255e-06,
+      "loss": 0.5276,
+      "step": 15348
+    },
+    {
+      "epoch": 4.074605071020842,
+      "grad_norm": 0.48487399286365107,
+      "learning_rate": 1.1746952561463238e-06,
+      "loss": 0.4979,
+      "step": 15349
+    },
+    {
+      "epoch": 4.074870569494226,
+      "grad_norm": 0.45727755486644306,
+      "learning_rate": 1.1743992360229096e-06,
+      "loss": 0.5512,
+      "step": 15350
+    },
+    {
+      "epoch": 4.075136067967609,
+      "grad_norm": 0.48028764181032724,
+      "learning_rate": 1.1741032417519557e-06,
+      "loss": 0.5581,
+      "step": 15351
+    },
+    {
+      "epoch": 4.075401566440993,
+      "grad_norm": 0.47768840036637783,
+      "learning_rate": 1.1738072733392343e-06,
+      "loss": 0.5156,
+      "step": 15352
+    },
+    {
+      "epoch": 4.075667064914377,
+      "grad_norm": 0.4601302356588837,
+      "learning_rate": 1.1735113307905186e-06,
+      "loss": 0.5285,
+      "step": 15353
+    },
+    {
+      "epoch": 4.07593256338776,
+      "grad_norm": 0.46846694396161453,
+      "learning_rate": 1.1732154141115793e-06,
+      "loss": 0.5157,
+      "step": 15354
+    },
+    {
+      "epoch": 4.076198061861144,
+      "grad_norm": 0.4812337278251216,
+      "learning_rate": 1.1729195233081885e-06,
+      "loss": 0.5711,
+      "step": 15355
+    },
+    {
+      "epoch": 4.0764635603345285,
+      "grad_norm": 0.4692316621318943,
+      "learning_rate": 1.1726236583861164e-06,
+      "loss": 0.5394,
+      "step": 15356
+    },
+    {
+      "epoch": 4.076729058807912,
+      "grad_norm": 0.49621358990959286,
+      "learning_rate": 1.1723278193511322e-06,
+      "loss": 0.5581,
+      "step": 15357
+    },
+    {
+      "epoch": 4.076994557281296,
+      "grad_norm": 0.46629862975972525,
+      "learning_rate": 1.172032006209007e-06,
+      "loss": 0.5088,
+      "step": 15358
+    },
+    {
+      "epoch": 4.077260055754679,
+      "grad_norm": 0.4678915136677538,
+      "learning_rate": 1.1717362189655081e-06,
+      "loss": 0.513,
+      "step": 15359
+    },
+    {
+      "epoch": 4.077525554228063,
+      "grad_norm": 0.48223867474177495,
+      "learning_rate": 1.1714404576264063e-06,
+      "loss": 0.5577,
+      "step": 15360
+    },
+    {
+      "epoch": 4.077791052701447,
+      "grad_norm": 0.48019797143928405,
+      "learning_rate": 1.1711447221974672e-06,
+      "loss": 0.5247,
+      "step": 15361
+    },
+    {
+      "epoch": 4.07805655117483,
+      "grad_norm": 0.4772406195961187,
+      "learning_rate": 1.1708490126844615e-06,
+      "loss": 0.5424,
+      "step": 15362
+    },
+    {
+      "epoch": 4.078322049648214,
+      "grad_norm": 0.4769514324409852,
+      "learning_rate": 1.170553329093153e-06,
+      "loss": 0.5464,
+      "step": 15363
+    },
+    {
+      "epoch": 4.0785875481215985,
+      "grad_norm": 0.49150032153037826,
+      "learning_rate": 1.170257671429309e-06,
+      "loss": 0.5736,
+      "step": 15364
+    },
+    {
+      "epoch": 4.078853046594982,
+      "grad_norm": 0.4828193422297218,
+      "learning_rate": 1.1699620396986974e-06,
+      "loss": 0.545,
+      "step": 15365
+    },
+    {
+      "epoch": 4.079118545068366,
+      "grad_norm": 0.4845799788006039,
+      "learning_rate": 1.1696664339070815e-06,
+      "loss": 0.5521,
+      "step": 15366
+    },
+    {
+      "epoch": 4.07938404354175,
+      "grad_norm": 0.4830359249173579,
+      "learning_rate": 1.1693708540602284e-06,
+      "loss": 0.5666,
+      "step": 15367
+    },
+    {
+      "epoch": 4.079649542015133,
+      "grad_norm": 0.4733848034091678,
+      "learning_rate": 1.1690753001639012e-06,
+      "loss": 0.5385,
+      "step": 15368
+    },
+    {
+      "epoch": 4.079915040488517,
+      "grad_norm": 0.4853737838345869,
+      "learning_rate": 1.1687797722238647e-06,
+      "loss": 0.5176,
+      "step": 15369
+    },
+    {
+      "epoch": 4.080180538961901,
+      "grad_norm": 0.4721494892442286,
+      "learning_rate": 1.1684842702458812e-06,
+      "loss": 0.5039,
+      "step": 15370
+    },
+    {
+      "epoch": 4.0804460374352844,
+      "grad_norm": 0.46281745321023837,
+      "learning_rate": 1.1681887942357154e-06,
+      "loss": 0.5414,
+      "step": 15371
+    },
+    {
+      "epoch": 4.0807115359086685,
+      "grad_norm": 0.46314902099768956,
+      "learning_rate": 1.1678933441991278e-06,
+      "loss": 0.5339,
+      "step": 15372
+    },
+    {
+      "epoch": 4.080977034382053,
+      "grad_norm": 0.47762926617506124,
+      "learning_rate": 1.1675979201418824e-06,
+      "loss": 0.5588,
+      "step": 15373
+    },
+    {
+      "epoch": 4.081242532855436,
+      "grad_norm": 0.4804263269469661,
+      "learning_rate": 1.1673025220697401e-06,
+      "loss": 0.5445,
+      "step": 15374
+    },
+    {
+      "epoch": 4.08150803132882,
+      "grad_norm": 0.4587883281413736,
+      "learning_rate": 1.1670071499884608e-06,
+      "loss": 0.4721,
+      "step": 15375
+    },
+    {
+      "epoch": 4.081773529802204,
+      "grad_norm": 0.4885394960529969,
+      "learning_rate": 1.1667118039038063e-06,
+      "loss": 0.5525,
+      "step": 15376
+    },
+    {
+      "epoch": 4.082039028275587,
+      "grad_norm": 0.4819447851660282,
+      "learning_rate": 1.1664164838215355e-06,
+      "loss": 0.5116,
+      "step": 15377
+    },
+    {
+      "epoch": 4.082304526748971,
+      "grad_norm": 0.46866352339357137,
+      "learning_rate": 1.1661211897474092e-06,
+      "loss": 0.5282,
+      "step": 15378
+    },
+    {
+      "epoch": 4.082570025222355,
+      "grad_norm": 0.4686368275561569,
+      "learning_rate": 1.1658259216871858e-06,
+      "loss": 0.5611,
+      "step": 15379
+    },
+    {
+      "epoch": 4.0828355236957385,
+      "grad_norm": 0.46414397678079544,
+      "learning_rate": 1.1655306796466234e-06,
+      "loss": 0.507,
+      "step": 15380
+    },
+    {
+      "epoch": 4.083101022169123,
+      "grad_norm": 0.4731892407955562,
+      "learning_rate": 1.1652354636314792e-06,
+      "loss": 0.5453,
+      "step": 15381
+    },
+    {
+      "epoch": 4.083366520642507,
+      "grad_norm": 0.4929337264327907,
+      "learning_rate": 1.1649402736475124e-06,
+      "loss": 0.5067,
+      "step": 15382
+    },
+    {
+      "epoch": 4.08363201911589,
+      "grad_norm": 0.4660932981669949,
+      "learning_rate": 1.1646451097004781e-06,
+      "loss": 0.5056,
+      "step": 15383
+    },
+    {
+      "epoch": 4.083897517589274,
+      "grad_norm": 0.4707144092539894,
+      "learning_rate": 1.164349971796134e-06,
+      "loss": 0.5582,
+      "step": 15384
+    },
+    {
+      "epoch": 4.084163016062657,
+      "grad_norm": 0.48288355973386526,
+      "learning_rate": 1.1640548599402371e-06,
+      "loss": 0.5329,
+      "step": 15385
+    },
+    {
+      "epoch": 4.084428514536041,
+      "grad_norm": 0.4935139811592437,
+      "learning_rate": 1.1637597741385398e-06,
+      "loss": 0.5176,
+      "step": 15386
+    },
+    {
+      "epoch": 4.084694013009425,
+      "grad_norm": 0.47923897780261815,
+      "learning_rate": 1.1634647143967994e-06,
+      "loss": 0.5724,
+      "step": 15387
+    },
+    {
+      "epoch": 4.084959511482809,
+      "grad_norm": 0.4790410497393249,
+      "learning_rate": 1.1631696807207688e-06,
+      "loss": 0.4873,
+      "step": 15388
+    },
+    {
+      "epoch": 4.085225009956193,
+      "grad_norm": 0.49323182560277307,
+      "learning_rate": 1.162874673116203e-06,
+      "loss": 0.5239,
+      "step": 15389
+    },
+    {
+      "epoch": 4.085490508429577,
+      "grad_norm": 0.47724023629664386,
+      "learning_rate": 1.1625796915888544e-06,
+      "loss": 0.5806,
+      "step": 15390
+    },
+    {
+      "epoch": 4.08575600690296,
+      "grad_norm": 0.47681147842768906,
+      "learning_rate": 1.1622847361444773e-06,
+      "loss": 0.5379,
+      "step": 15391
+    },
+    {
+      "epoch": 4.086021505376344,
+      "grad_norm": 0.47231315409210184,
+      "learning_rate": 1.1619898067888226e-06,
+      "loss": 0.5381,
+      "step": 15392
+    },
+    {
+      "epoch": 4.086287003849728,
+      "grad_norm": 0.4695759483539134,
+      "learning_rate": 1.1616949035276423e-06,
+      "loss": 0.5401,
+      "step": 15393
+    },
+    {
+      "epoch": 4.086552502323111,
+      "grad_norm": 0.4789459177210307,
+      "learning_rate": 1.161400026366689e-06,
+      "loss": 0.5251,
+      "step": 15394
+    },
+    {
+      "epoch": 4.086818000796495,
+      "grad_norm": 0.4634008031719571,
+      "learning_rate": 1.1611051753117117e-06,
+      "loss": 0.5399,
+      "step": 15395
+    },
+    {
+      "epoch": 4.0870834992698795,
+      "grad_norm": 0.46131747788778044,
+      "learning_rate": 1.1608103503684623e-06,
+      "loss": 0.5137,
+      "step": 15396
+    },
+    {
+      "epoch": 4.087348997743263,
+      "grad_norm": 0.45453070236140375,
+      "learning_rate": 1.1605155515426903e-06,
+      "loss": 0.5037,
+      "step": 15397
+    },
+    {
+      "epoch": 4.087614496216647,
+      "grad_norm": 0.46897052288012575,
+      "learning_rate": 1.1602207788401446e-06,
+      "loss": 0.5551,
+      "step": 15398
+    },
+    {
+      "epoch": 4.087879994690031,
+      "grad_norm": 0.4785642935837407,
+      "learning_rate": 1.1599260322665728e-06,
+      "loss": 0.548,
+      "step": 15399
+    },
+    {
+      "epoch": 4.088145493163414,
+      "grad_norm": 0.4698628269754639,
+      "learning_rate": 1.1596313118277257e-06,
+      "loss": 0.5273,
+      "step": 15400
+    },
+    {
+      "epoch": 4.088410991636798,
+      "grad_norm": 0.47769471394607665,
+      "learning_rate": 1.1593366175293488e-06,
+      "loss": 0.5739,
+      "step": 15401
+    },
+    {
+      "epoch": 4.088676490110182,
+      "grad_norm": 0.4815844365115442,
+      "learning_rate": 1.1590419493771915e-06,
+      "loss": 0.5472,
+      "step": 15402
+    },
+    {
+      "epoch": 4.088941988583565,
+      "grad_norm": 0.4744064643503215,
+      "learning_rate": 1.1587473073769991e-06,
+      "loss": 0.5597,
+      "step": 15403
+    },
+    {
+      "epoch": 4.0892074870569495,
+      "grad_norm": 0.4709593160176362,
+      "learning_rate": 1.1584526915345174e-06,
+      "loss": 0.5319,
+      "step": 15404
+    },
+    {
+      "epoch": 4.089472985530334,
+      "grad_norm": 0.47509094804297247,
+      "learning_rate": 1.1581581018554939e-06,
+      "loss": 0.5517,
+      "step": 15405
+    },
+    {
+      "epoch": 4.089738484003717,
+      "grad_norm": 0.46634770021070465,
+      "learning_rate": 1.1578635383456716e-06,
+      "loss": 0.5461,
+      "step": 15406
+    },
+    {
+      "epoch": 4.090003982477101,
+      "grad_norm": 0.46550451517465524,
+      "learning_rate": 1.1575690010107976e-06,
+      "loss": 0.5517,
+      "step": 15407
+    },
+    {
+      "epoch": 4.090269480950485,
+      "grad_norm": 0.489951664934729,
+      "learning_rate": 1.1572744898566146e-06,
+      "loss": 0.5859,
+      "step": 15408
+    },
+    {
+      "epoch": 4.090534979423868,
+      "grad_norm": 0.48053420500710897,
+      "learning_rate": 1.156980004888866e-06,
+      "loss": 0.5589,
+      "step": 15409
+    },
+    {
+      "epoch": 4.090800477897252,
+      "grad_norm": 0.46893462199055064,
+      "learning_rate": 1.1566855461132963e-06,
+      "loss": 0.5376,
+      "step": 15410
+    },
+    {
+      "epoch": 4.091065976370636,
+      "grad_norm": 0.4588324628884706,
+      "learning_rate": 1.1563911135356468e-06,
+      "loss": 0.5344,
+      "step": 15411
+    },
+    {
+      "epoch": 4.0913314748440195,
+      "grad_norm": 0.4822858221906309,
+      "learning_rate": 1.1560967071616612e-06,
+      "loss": 0.5305,
+      "step": 15412
+    },
+    {
+      "epoch": 4.091596973317404,
+      "grad_norm": 0.4788976470478269,
+      "learning_rate": 1.155802326997079e-06,
+      "loss": 0.5284,
+      "step": 15413
+    },
+    {
+      "epoch": 4.091862471790787,
+      "grad_norm": 0.48463632455514644,
+      "learning_rate": 1.1555079730476448e-06,
+      "loss": 0.5681,
+      "step": 15414
+    },
+    {
+      "epoch": 4.092127970264171,
+      "grad_norm": 0.47671625881163676,
+      "learning_rate": 1.155213645319095e-06,
+      "loss": 0.5436,
+      "step": 15415
+    },
+    {
+      "epoch": 4.092393468737555,
+      "grad_norm": 0.47351882971634107,
+      "learning_rate": 1.1549193438171724e-06,
+      "loss": 0.532,
+      "step": 15416
+    },
+    {
+      "epoch": 4.092658967210938,
+      "grad_norm": 0.4833109004186374,
+      "learning_rate": 1.1546250685476155e-06,
+      "loss": 0.5336,
+      "step": 15417
+    },
+    {
+      "epoch": 4.092924465684322,
+      "grad_norm": 0.47087944813485966,
+      "learning_rate": 1.1543308195161643e-06,
+      "loss": 0.5163,
+      "step": 15418
+    },
+    {
+      "epoch": 4.093189964157706,
+      "grad_norm": 0.4647098002558098,
+      "learning_rate": 1.154036596728556e-06,
+      "loss": 0.5154,
+      "step": 15419
+    },
+    {
+      "epoch": 4.0934554626310895,
+      "grad_norm": 0.4640195698522609,
+      "learning_rate": 1.1537424001905309e-06,
+      "loss": 0.5534,
+      "step": 15420
+    },
+    {
+      "epoch": 4.093720961104474,
+      "grad_norm": 0.4722462043037923,
+      "learning_rate": 1.1534482299078247e-06,
+      "loss": 0.4885,
+      "step": 15421
+    },
+    {
+      "epoch": 4.093986459577858,
+      "grad_norm": 0.47256089414829705,
+      "learning_rate": 1.1531540858861741e-06,
+      "loss": 0.5402,
+      "step": 15422
+    },
+    {
+      "epoch": 4.094251958051241,
+      "grad_norm": 0.47321317654208594,
+      "learning_rate": 1.1528599681313177e-06,
+      "loss": 0.5385,
+      "step": 15423
+    },
+    {
+      "epoch": 4.094517456524625,
+      "grad_norm": 0.4844058743497153,
+      "learning_rate": 1.152565876648989e-06,
+      "loss": 0.5435,
+      "step": 15424
+    },
+    {
+      "epoch": 4.094782954998009,
+      "grad_norm": 0.46486715437271514,
+      "learning_rate": 1.1522718114449272e-06,
+      "loss": 0.518,
+      "step": 15425
+    },
+    {
+      "epoch": 4.095048453471392,
+      "grad_norm": 0.4624993544497032,
+      "learning_rate": 1.1519777725248627e-06,
+      "loss": 0.5102,
+      "step": 15426
+    },
+    {
+      "epoch": 4.095313951944776,
+      "grad_norm": 0.4732863416922178,
+      "learning_rate": 1.151683759894533e-06,
+      "loss": 0.5273,
+      "step": 15427
+    },
+    {
+      "epoch": 4.09557945041816,
+      "grad_norm": 0.46464623280533895,
+      "learning_rate": 1.1513897735596702e-06,
+      "loss": 0.5183,
+      "step": 15428
+    },
+    {
+      "epoch": 4.095844948891544,
+      "grad_norm": 0.4663027851831423,
+      "learning_rate": 1.15109581352601e-06,
+      "loss": 0.5569,
+      "step": 15429
+    },
+    {
+      "epoch": 4.096110447364928,
+      "grad_norm": 0.4902728855169866,
+      "learning_rate": 1.1508018797992832e-06,
+      "loss": 0.5314,
+      "step": 15430
+    },
+    {
+      "epoch": 4.096375945838312,
+      "grad_norm": 0.4821755261176922,
+      "learning_rate": 1.150507972385224e-06,
+      "loss": 0.5778,
+      "step": 15431
+    },
+    {
+      "epoch": 4.096641444311695,
+      "grad_norm": 0.4573758376930445,
+      "learning_rate": 1.1502140912895633e-06,
+      "loss": 0.5022,
+      "step": 15432
+    },
+    {
+      "epoch": 4.096906942785079,
+      "grad_norm": 0.4871239544339889,
+      "learning_rate": 1.1499202365180317e-06,
+      "loss": 0.5216,
+      "step": 15433
+    },
+    {
+      "epoch": 4.097172441258463,
+      "grad_norm": 0.47534498859128965,
+      "learning_rate": 1.1496264080763622e-06,
+      "loss": 0.5349,
+      "step": 15434
+    },
+    {
+      "epoch": 4.097437939731846,
+      "grad_norm": 0.48286790749003305,
+      "learning_rate": 1.149332605970283e-06,
+      "loss": 0.5383,
+      "step": 15435
+    },
+    {
+      "epoch": 4.09770343820523,
+      "grad_norm": 0.4751447745762735,
+      "learning_rate": 1.1490388302055257e-06,
+      "loss": 0.5044,
+      "step": 15436
+    },
+    {
+      "epoch": 4.0979689366786145,
+      "grad_norm": 0.4627625048492895,
+      "learning_rate": 1.1487450807878196e-06,
+      "loss": 0.5248,
+      "step": 15437
+    },
+    {
+      "epoch": 4.098234435151998,
+      "grad_norm": 0.47613938667242905,
+      "learning_rate": 1.148451357722891e-06,
+      "loss": 0.5557,
+      "step": 15438
+    },
+    {
+      "epoch": 4.098499933625382,
+      "grad_norm": 0.48247944211328836,
+      "learning_rate": 1.148157661016472e-06,
+      "loss": 0.5588,
+      "step": 15439
+    },
+    {
+      "epoch": 4.098765432098766,
+      "grad_norm": 0.4799147552815529,
+      "learning_rate": 1.147863990674287e-06,
+      "loss": 0.541,
+      "step": 15440
+    },
+    {
+      "epoch": 4.099030930572149,
+      "grad_norm": 0.4694703630859369,
+      "learning_rate": 1.1475703467020658e-06,
+      "loss": 0.5573,
+      "step": 15441
+    },
+    {
+      "epoch": 4.099296429045533,
+      "grad_norm": 0.48546992650724763,
+      "learning_rate": 1.147276729105533e-06,
+      "loss": 0.5859,
+      "step": 15442
+    },
+    {
+      "epoch": 4.099561927518916,
+      "grad_norm": 0.47419531800443093,
+      "learning_rate": 1.1469831378904183e-06,
+      "loss": 0.4804,
+      "step": 15443
+    },
+    {
+      "epoch": 4.0998274259923,
+      "grad_norm": 0.48278851249749727,
+      "learning_rate": 1.146689573062443e-06,
+      "loss": 0.5685,
+      "step": 15444
+    },
+    {
+      "epoch": 4.1000929244656845,
+      "grad_norm": 0.4649842514134834,
+      "learning_rate": 1.1463960346273356e-06,
+      "loss": 0.5503,
+      "step": 15445
+    },
+    {
+      "epoch": 4.100358422939068,
+      "grad_norm": 0.4728270775971377,
+      "learning_rate": 1.1461025225908188e-06,
+      "loss": 0.5665,
+      "step": 15446
+    },
+    {
+      "epoch": 4.100623921412452,
+      "grad_norm": 0.47134001836755457,
+      "learning_rate": 1.1458090369586186e-06,
+      "loss": 0.5334,
+      "step": 15447
+    },
+    {
+      "epoch": 4.100889419885836,
+      "grad_norm": 0.4735592696378864,
+      "learning_rate": 1.1455155777364576e-06,
+      "loss": 0.5448,
+      "step": 15448
+    },
+    {
+      "epoch": 4.101154918359219,
+      "grad_norm": 0.4833817732500202,
+      "learning_rate": 1.1452221449300588e-06,
+      "loss": 0.561,
+      "step": 15449
+    },
+    {
+      "epoch": 4.101420416832603,
+      "grad_norm": 0.4774546010657383,
+      "learning_rate": 1.1449287385451458e-06,
+      "loss": 0.5265,
+      "step": 15450
+    },
+    {
+      "epoch": 4.101685915305987,
+      "grad_norm": 0.4756031557966933,
+      "learning_rate": 1.1446353585874393e-06,
+      "loss": 0.5602,
+      "step": 15451
+    },
+    {
+      "epoch": 4.1019514137793704,
+      "grad_norm": 0.4735285706441791,
+      "learning_rate": 1.1443420050626624e-06,
+      "loss": 0.5511,
+      "step": 15452
+    },
+    {
+      "epoch": 4.1022169122527545,
+      "grad_norm": 0.4823422993637103,
+      "learning_rate": 1.1440486779765352e-06,
+      "loss": 0.5355,
+      "step": 15453
+    },
+    {
+      "epoch": 4.102482410726139,
+      "grad_norm": 0.47948960013614494,
+      "learning_rate": 1.1437553773347804e-06,
+      "loss": 0.547,
+      "step": 15454
+    },
+    {
+      "epoch": 4.102747909199522,
+      "grad_norm": 0.48302260518832413,
+      "learning_rate": 1.1434621031431145e-06,
+      "loss": 0.561,
+      "step": 15455
+    },
+    {
+      "epoch": 4.103013407672906,
+      "grad_norm": 0.480244665964944,
+      "learning_rate": 1.143168855407259e-06,
+      "loss": 0.5575,
+      "step": 15456
+    },
+    {
+      "epoch": 4.10327890614629,
+      "grad_norm": 0.47675986978699214,
+      "learning_rate": 1.1428756341329338e-06,
+      "loss": 0.5334,
+      "step": 15457
+    },
+    {
+      "epoch": 4.103544404619673,
+      "grad_norm": 0.4821504789013667,
+      "learning_rate": 1.142582439325856e-06,
+      "loss": 0.5332,
+      "step": 15458
+    },
+    {
+      "epoch": 4.103809903093057,
+      "grad_norm": 0.47611330885118996,
+      "learning_rate": 1.1422892709917446e-06,
+      "loss": 0.5315,
+      "step": 15459
+    },
+    {
+      "epoch": 4.104075401566441,
+      "grad_norm": 0.48452878116476633,
+      "learning_rate": 1.141996129136317e-06,
+      "loss": 0.5499,
+      "step": 15460
+    },
+    {
+      "epoch": 4.1043409000398245,
+      "grad_norm": 0.48164522330123033,
+      "learning_rate": 1.1417030137652896e-06,
+      "loss": 0.5777,
+      "step": 15461
+    },
+    {
+      "epoch": 4.104606398513209,
+      "grad_norm": 0.48619550810479,
+      "learning_rate": 1.1414099248843782e-06,
+      "loss": 0.5501,
+      "step": 15462
+    },
+    {
+      "epoch": 4.104871896986593,
+      "grad_norm": 0.4848212799772456,
+      "learning_rate": 1.1411168624993004e-06,
+      "loss": 0.5339,
+      "step": 15463
+    },
+    {
+      "epoch": 4.105137395459976,
+      "grad_norm": 0.4887231954192791,
+      "learning_rate": 1.1408238266157703e-06,
+      "loss": 0.5489,
+      "step": 15464
+    },
+    {
+      "epoch": 4.10540289393336,
+      "grad_norm": 0.49186598726095804,
+      "learning_rate": 1.1405308172395043e-06,
+      "loss": 0.5421,
+      "step": 15465
+    },
+    {
+      "epoch": 4.105668392406744,
+      "grad_norm": 0.4833677954385405,
+      "learning_rate": 1.140237834376216e-06,
+      "loss": 0.5036,
+      "step": 15466
+    },
+    {
+      "epoch": 4.105933890880127,
+      "grad_norm": 0.4787778057518926,
+      "learning_rate": 1.139944878031618e-06,
+      "loss": 0.5486,
+      "step": 15467
+    },
+    {
+      "epoch": 4.106199389353511,
+      "grad_norm": 0.4835492936274251,
+      "learning_rate": 1.139651948211426e-06,
+      "loss": 0.5459,
+      "step": 15468
+    },
+    {
+      "epoch": 4.106464887826895,
+      "grad_norm": 0.47353895102262933,
+      "learning_rate": 1.139359044921351e-06,
+      "loss": 0.5282,
+      "step": 15469
+    },
+    {
+      "epoch": 4.106730386300279,
+      "grad_norm": 0.46089384567683056,
+      "learning_rate": 1.1390661681671066e-06,
+      "loss": 0.5328,
+      "step": 15470
+    },
+    {
+      "epoch": 4.106995884773663,
+      "grad_norm": 0.48443599195542786,
+      "learning_rate": 1.1387733179544042e-06,
+      "loss": 0.5242,
+      "step": 15471
+    },
+    {
+      "epoch": 4.107261383247046,
+      "grad_norm": 0.49848100963334163,
+      "learning_rate": 1.1384804942889549e-06,
+      "loss": 0.5665,
+      "step": 15472
+    },
+    {
+      "epoch": 4.10752688172043,
+      "grad_norm": 0.4722537934171106,
+      "learning_rate": 1.1381876971764688e-06,
+      "loss": 0.558,
+      "step": 15473
+    },
+    {
+      "epoch": 4.107792380193814,
+      "grad_norm": 0.45797572626807087,
+      "learning_rate": 1.1378949266226577e-06,
+      "loss": 0.5132,
+      "step": 15474
+    },
+    {
+      "epoch": 4.108057878667197,
+      "grad_norm": 0.4733250590764185,
+      "learning_rate": 1.1376021826332295e-06,
+      "loss": 0.5505,
+      "step": 15475
+    },
+    {
+      "epoch": 4.108323377140581,
+      "grad_norm": 0.5381160805824715,
+      "learning_rate": 1.1373094652138945e-06,
+      "loss": 0.5351,
+      "step": 15476
+    },
+    {
+      "epoch": 4.1085888756139655,
+      "grad_norm": 0.48283512177860494,
+      "learning_rate": 1.1370167743703634e-06,
+      "loss": 0.5365,
+      "step": 15477
+    },
+    {
+      "epoch": 4.108854374087349,
+      "grad_norm": 0.49282344315052784,
+      "learning_rate": 1.1367241101083404e-06,
+      "loss": 0.555,
+      "step": 15478
+    },
+    {
+      "epoch": 4.109119872560733,
+      "grad_norm": 0.478607039300628,
+      "learning_rate": 1.1364314724335359e-06,
+      "loss": 0.5047,
+      "step": 15479
+    },
+    {
+      "epoch": 4.109385371034117,
+      "grad_norm": 0.4850599282933851,
+      "learning_rate": 1.1361388613516555e-06,
+      "loss": 0.5764,
+      "step": 15480
+    },
+    {
+      "epoch": 4.1096508695075,
+      "grad_norm": 0.4824368961899814,
+      "learning_rate": 1.1358462768684075e-06,
+      "loss": 0.5073,
+      "step": 15481
+    },
+    {
+      "epoch": 4.109916367980884,
+      "grad_norm": 0.4950245034499109,
+      "learning_rate": 1.1355537189894963e-06,
+      "loss": 0.5355,
+      "step": 15482
+    },
+    {
+      "epoch": 4.110181866454268,
+      "grad_norm": 0.47658368926283523,
+      "learning_rate": 1.135261187720629e-06,
+      "loss": 0.5754,
+      "step": 15483
+    },
+    {
+      "epoch": 4.110447364927651,
+      "grad_norm": 0.46280756800415745,
+      "learning_rate": 1.1349686830675102e-06,
+      "loss": 0.5353,
+      "step": 15484
+    },
+    {
+      "epoch": 4.1107128634010355,
+      "grad_norm": 0.47251089894170517,
+      "learning_rate": 1.1346762050358434e-06,
+      "loss": 0.5322,
+      "step": 15485
+    },
+    {
+      "epoch": 4.11097836187442,
+      "grad_norm": 0.48628681443600474,
+      "learning_rate": 1.1343837536313342e-06,
+      "loss": 0.5064,
+      "step": 15486
+    },
+    {
+      "epoch": 4.111243860347803,
+      "grad_norm": 0.47553115784868455,
+      "learning_rate": 1.1340913288596846e-06,
+      "loss": 0.5134,
+      "step": 15487
+    },
+    {
+      "epoch": 4.111509358821187,
+      "grad_norm": 0.48641577675168407,
+      "learning_rate": 1.1337989307265995e-06,
+      "loss": 0.5705,
+      "step": 15488
+    },
+    {
+      "epoch": 4.111774857294571,
+      "grad_norm": 0.47735331959611227,
+      "learning_rate": 1.13350655923778e-06,
+      "loss": 0.5261,
+      "step": 15489
+    },
+    {
+      "epoch": 4.112040355767954,
+      "grad_norm": 0.4660302700507674,
+      "learning_rate": 1.1332142143989285e-06,
+      "loss": 0.4889,
+      "step": 15490
+    },
+    {
+      "epoch": 4.112305854241338,
+      "grad_norm": 0.47030674494640967,
+      "learning_rate": 1.1329218962157454e-06,
+      "loss": 0.5025,
+      "step": 15491
+    },
+    {
+      "epoch": 4.112571352714722,
+      "grad_norm": 0.46511406413352496,
+      "learning_rate": 1.1326296046939334e-06,
+      "loss": 0.5302,
+      "step": 15492
+    },
+    {
+      "epoch": 4.1128368511881055,
+      "grad_norm": 0.4734691688143414,
+      "learning_rate": 1.1323373398391909e-06,
+      "loss": 0.5662,
+      "step": 15493
+    },
+    {
+      "epoch": 4.11310234966149,
+      "grad_norm": 0.46993128516122984,
+      "learning_rate": 1.13204510165722e-06,
+      "loss": 0.5855,
+      "step": 15494
+    },
+    {
+      "epoch": 4.113367848134874,
+      "grad_norm": 0.46978789182989583,
+      "learning_rate": 1.131752890153719e-06,
+      "loss": 0.5553,
+      "step": 15495
+    },
+    {
+      "epoch": 4.113633346608257,
+      "grad_norm": 0.46397060441926696,
+      "learning_rate": 1.1314607053343857e-06,
+      "loss": 0.5083,
+      "step": 15496
+    },
+    {
+      "epoch": 4.113898845081641,
+      "grad_norm": 0.4836015882237248,
+      "learning_rate": 1.1311685472049204e-06,
+      "loss": 0.5447,
+      "step": 15497
+    },
+    {
+      "epoch": 4.114164343555024,
+      "grad_norm": 0.473901865326143,
+      "learning_rate": 1.1308764157710187e-06,
+      "loss": 0.5489,
+      "step": 15498
+    },
+    {
+      "epoch": 4.114429842028408,
+      "grad_norm": 0.4794470876563034,
+      "learning_rate": 1.13058431103838e-06,
+      "loss": 0.5278,
+      "step": 15499
+    },
+    {
+      "epoch": 4.114695340501792,
+      "grad_norm": 0.49734911428279194,
+      "learning_rate": 1.1302922330127005e-06,
+      "loss": 0.5242,
+      "step": 15500
+    },
+    {
+      "epoch": 4.1149608389751755,
+      "grad_norm": 0.4805064052929783,
+      "learning_rate": 1.130000181699675e-06,
+      "loss": 0.4972,
+      "step": 15501
+    },
+    {
+      "epoch": 4.11522633744856,
+      "grad_norm": 0.4627186404709467,
+      "learning_rate": 1.1297081571050014e-06,
+      "loss": 0.5646,
+      "step": 15502
+    },
+    {
+      "epoch": 4.115491835921944,
+      "grad_norm": 0.47185857201879167,
+      "learning_rate": 1.1294161592343725e-06,
+      "loss": 0.5281,
+      "step": 15503
+    },
+    {
+      "epoch": 4.115757334395327,
+      "grad_norm": 0.47484701618902164,
+      "learning_rate": 1.1291241880934855e-06,
+      "loss": 0.5537,
+      "step": 15504
+    },
+    {
+      "epoch": 4.116022832868711,
+      "grad_norm": 0.5006905644164343,
+      "learning_rate": 1.1288322436880322e-06,
+      "loss": 0.5576,
+      "step": 15505
+    },
+    {
+      "epoch": 4.116288331342095,
+      "grad_norm": 0.4940666638969997,
+      "learning_rate": 1.1285403260237093e-06,
+      "loss": 0.5328,
+      "step": 15506
+    },
+    {
+      "epoch": 4.116553829815478,
+      "grad_norm": 0.4760851919295786,
+      "learning_rate": 1.128248435106206e-06,
+      "loss": 0.5304,
+      "step": 15507
+    },
+    {
+      "epoch": 4.116819328288862,
+      "grad_norm": 0.47790399710346,
+      "learning_rate": 1.127956570941218e-06,
+      "loss": 0.5468,
+      "step": 15508
+    },
+    {
+      "epoch": 4.117084826762246,
+      "grad_norm": 0.477314963493014,
+      "learning_rate": 1.127664733534435e-06,
+      "loss": 0.5259,
+      "step": 15509
+    },
+    {
+      "epoch": 4.11735032523563,
+      "grad_norm": 0.4794506393572603,
+      "learning_rate": 1.127372922891551e-06,
+      "loss": 0.5374,
+      "step": 15510
+    },
+    {
+      "epoch": 4.117615823709014,
+      "grad_norm": 0.4655015605684314,
+      "learning_rate": 1.1270811390182543e-06,
+      "loss": 0.5493,
+      "step": 15511
+    },
+    {
+      "epoch": 4.117881322182398,
+      "grad_norm": 0.4795395110755202,
+      "learning_rate": 1.1267893819202383e-06,
+      "loss": 0.5553,
+      "step": 15512
+    },
+    {
+      "epoch": 4.118146820655781,
+      "grad_norm": 0.4745873644949828,
+      "learning_rate": 1.1264976516031914e-06,
+      "loss": 0.5142,
+      "step": 15513
+    },
+    {
+      "epoch": 4.118412319129165,
+      "grad_norm": 0.4829250288103117,
+      "learning_rate": 1.1262059480728022e-06,
+      "loss": 0.5453,
+      "step": 15514
+    },
+    {
+      "epoch": 4.118677817602549,
+      "grad_norm": 0.48749227015319374,
+      "learning_rate": 1.1259142713347615e-06,
+      "loss": 0.5745,
+      "step": 15515
+    },
+    {
+      "epoch": 4.118943316075932,
+      "grad_norm": 0.4949231274059334,
+      "learning_rate": 1.125622621394756e-06,
+      "loss": 0.5656,
+      "step": 15516
+    },
+    {
+      "epoch": 4.119208814549316,
+      "grad_norm": 0.459439133885228,
+      "learning_rate": 1.1253309982584755e-06,
+      "loss": 0.5203,
+      "step": 15517
+    },
+    {
+      "epoch": 4.1194743130227005,
+      "grad_norm": 0.48280030971257576,
+      "learning_rate": 1.1250394019316063e-06,
+      "loss": 0.5287,
+      "step": 15518
+    },
+    {
+      "epoch": 4.119739811496084,
+      "grad_norm": 0.47449002357217707,
+      "learning_rate": 1.1247478324198351e-06,
+      "loss": 0.5292,
+      "step": 15519
+    },
+    {
+      "epoch": 4.120005309969468,
+      "grad_norm": 0.4992125424553765,
+      "learning_rate": 1.1244562897288474e-06,
+      "loss": 0.5392,
+      "step": 15520
+    },
+    {
+      "epoch": 4.120270808442852,
+      "grad_norm": 0.4899011082425855,
+      "learning_rate": 1.1241647738643302e-06,
+      "loss": 0.5622,
+      "step": 15521
+    },
+    {
+      "epoch": 4.120536306916235,
+      "grad_norm": 0.4669537961698866,
+      "learning_rate": 1.1238732848319695e-06,
+      "loss": 0.5127,
+      "step": 15522
+    },
+    {
+      "epoch": 4.120801805389619,
+      "grad_norm": 0.46826171538247113,
+      "learning_rate": 1.1235818226374489e-06,
+      "loss": 0.5527,
+      "step": 15523
+    },
+    {
+      "epoch": 4.121067303863002,
+      "grad_norm": 0.4835485196809336,
+      "learning_rate": 1.1232903872864529e-06,
+      "loss": 0.5356,
+      "step": 15524
+    },
+    {
+      "epoch": 4.121332802336386,
+      "grad_norm": 0.4672756468223092,
+      "learning_rate": 1.122998978784664e-06,
+      "loss": 0.4988,
+      "step": 15525
+    },
+    {
+      "epoch": 4.1215983008097705,
+      "grad_norm": 0.4684412976558153,
+      "learning_rate": 1.1227075971377676e-06,
+      "loss": 0.5164,
+      "step": 15526
+    },
+    {
+      "epoch": 4.121863799283154,
+      "grad_norm": 0.4796260289739082,
+      "learning_rate": 1.1224162423514442e-06,
+      "loss": 0.5812,
+      "step": 15527
+    },
+    {
+      "epoch": 4.122129297756538,
+      "grad_norm": 0.46535898132102727,
+      "learning_rate": 1.1221249144313778e-06,
+      "loss": 0.5258,
+      "step": 15528
+    },
+    {
+      "epoch": 4.122394796229922,
+      "grad_norm": 0.47595586486401287,
+      "learning_rate": 1.1218336133832494e-06,
+      "loss": 0.5144,
+      "step": 15529
+    },
+    {
+      "epoch": 4.122660294703305,
+      "grad_norm": 0.47088727439423417,
+      "learning_rate": 1.1215423392127388e-06,
+      "loss": 0.5357,
+      "step": 15530
+    },
+    {
+      "epoch": 4.122925793176689,
+      "grad_norm": 0.46291951163652123,
+      "learning_rate": 1.121251091925529e-06,
+      "loss": 0.5143,
+      "step": 15531
+    },
+    {
+      "epoch": 4.123191291650073,
+      "grad_norm": 0.484197933142213,
+      "learning_rate": 1.1209598715272974e-06,
+      "loss": 0.5142,
+      "step": 15532
+    },
+    {
+      "epoch": 4.1234567901234565,
+      "grad_norm": 0.48458272870296576,
+      "learning_rate": 1.1206686780237256e-06,
+      "loss": 0.531,
+      "step": 15533
+    },
+    {
+      "epoch": 4.1237222885968405,
+      "grad_norm": 0.4882887384324555,
+      "learning_rate": 1.120377511420491e-06,
+      "loss": 0.5236,
+      "step": 15534
+    },
+    {
+      "epoch": 4.123987787070225,
+      "grad_norm": 0.477147298130285,
+      "learning_rate": 1.1200863717232745e-06,
+      "loss": 0.5489,
+      "step": 15535
+    },
+    {
+      "epoch": 4.124253285543608,
+      "grad_norm": 0.47210302805973253,
+      "learning_rate": 1.1197952589377506e-06,
+      "loss": 0.5258,
+      "step": 15536
+    },
+    {
+      "epoch": 4.124518784016992,
+      "grad_norm": 0.47528758191717485,
+      "learning_rate": 1.1195041730695997e-06,
+      "loss": 0.5454,
+      "step": 15537
+    },
+    {
+      "epoch": 4.124784282490376,
+      "grad_norm": 0.5012021287384323,
+      "learning_rate": 1.119213114124496e-06,
+      "loss": 0.53,
+      "step": 15538
+    },
+    {
+      "epoch": 4.125049780963759,
+      "grad_norm": 0.4653418544208738,
+      "learning_rate": 1.1189220821081187e-06,
+      "loss": 0.5306,
+      "step": 15539
+    },
+    {
+      "epoch": 4.125315279437143,
+      "grad_norm": 0.47880937642808624,
+      "learning_rate": 1.118631077026141e-06,
+      "loss": 0.5477,
+      "step": 15540
+    },
+    {
+      "epoch": 4.125580777910527,
+      "grad_norm": 0.49837154386921517,
+      "learning_rate": 1.1183400988842405e-06,
+      "loss": 0.5784,
+      "step": 15541
+    },
+    {
+      "epoch": 4.1258462763839105,
+      "grad_norm": 0.46201191978439693,
+      "learning_rate": 1.118049147688091e-06,
+      "loss": 0.5208,
+      "step": 15542
+    },
+    {
+      "epoch": 4.126111774857295,
+      "grad_norm": 0.4766994454286879,
+      "learning_rate": 1.117758223443366e-06,
+      "loss": 0.5472,
+      "step": 15543
+    },
+    {
+      "epoch": 4.126377273330679,
+      "grad_norm": 0.4833781782210179,
+      "learning_rate": 1.1174673261557405e-06,
+      "loss": 0.5325,
+      "step": 15544
+    },
+    {
+      "epoch": 4.126642771804062,
+      "grad_norm": 0.49259926823461087,
+      "learning_rate": 1.1171764558308865e-06,
+      "loss": 0.5383,
+      "step": 15545
+    },
+    {
+      "epoch": 4.126908270277446,
+      "grad_norm": 0.45673830618872674,
+      "learning_rate": 1.1168856124744782e-06,
+      "loss": 0.5226,
+      "step": 15546
+    },
+    {
+      "epoch": 4.12717376875083,
+      "grad_norm": 0.46550724569612123,
+      "learning_rate": 1.1165947960921868e-06,
+      "loss": 0.5354,
+      "step": 15547
+    },
+    {
+      "epoch": 4.127439267224213,
+      "grad_norm": 0.480245422161727,
+      "learning_rate": 1.1163040066896832e-06,
+      "loss": 0.5631,
+      "step": 15548
+    },
+    {
+      "epoch": 4.127704765697597,
+      "grad_norm": 0.4834019673595126,
+      "learning_rate": 1.1160132442726404e-06,
+      "loss": 0.5221,
+      "step": 15549
+    },
+    {
+      "epoch": 4.1279702641709815,
+      "grad_norm": 0.4924064340013863,
+      "learning_rate": 1.1157225088467268e-06,
+      "loss": 0.5264,
+      "step": 15550
+    },
+    {
+      "epoch": 4.128235762644365,
+      "grad_norm": 0.4868811244069276,
+      "learning_rate": 1.1154318004176147e-06,
+      "loss": 0.5383,
+      "step": 15551
+    },
+    {
+      "epoch": 4.128501261117749,
+      "grad_norm": 0.4694884383800759,
+      "learning_rate": 1.1151411189909724e-06,
+      "loss": 0.5262,
+      "step": 15552
+    },
+    {
+      "epoch": 4.128766759591132,
+      "grad_norm": 0.48940522990978125,
+      "learning_rate": 1.114850464572469e-06,
+      "loss": 0.5779,
+      "step": 15553
+    },
+    {
+      "epoch": 4.129032258064516,
+      "grad_norm": 0.4781934060803823,
+      "learning_rate": 1.1145598371677718e-06,
+      "loss": 0.5369,
+      "step": 15554
+    },
+    {
+      "epoch": 4.1292977565379,
+      "grad_norm": 0.48811123907939447,
+      "learning_rate": 1.1142692367825512e-06,
+      "loss": 0.5169,
+      "step": 15555
+    },
+    {
+      "epoch": 4.129563255011283,
+      "grad_norm": 0.4582092128417319,
+      "learning_rate": 1.1139786634224723e-06,
+      "loss": 0.5323,
+      "step": 15556
+    },
+    {
+      "epoch": 4.129828753484667,
+      "grad_norm": 0.4805331425137868,
+      "learning_rate": 1.113688117093204e-06,
+      "loss": 0.5354,
+      "step": 15557
+    },
+    {
+      "epoch": 4.1300942519580515,
+      "grad_norm": 0.4653630522683209,
+      "learning_rate": 1.1133975978004117e-06,
+      "loss": 0.5451,
+      "step": 15558
+    },
+    {
+      "epoch": 4.130359750431435,
+      "grad_norm": 0.4814286033916301,
+      "learning_rate": 1.11310710554976e-06,
+      "loss": 0.5495,
+      "step": 15559
+    },
+    {
+      "epoch": 4.130625248904819,
+      "grad_norm": 0.4780070513290123,
+      "learning_rate": 1.1128166403469165e-06,
+      "loss": 0.5303,
+      "step": 15560
+    },
+    {
+      "epoch": 4.130890747378203,
+      "grad_norm": 0.48993472337058136,
+      "learning_rate": 1.112526202197544e-06,
+      "loss": 0.582,
+      "step": 15561
+    },
+    {
+      "epoch": 4.131156245851586,
+      "grad_norm": 0.4633367386425925,
+      "learning_rate": 1.112235791107309e-06,
+      "loss": 0.5472,
+      "step": 15562
+    },
+    {
+      "epoch": 4.13142174432497,
+      "grad_norm": 0.4897531974201303,
+      "learning_rate": 1.1119454070818725e-06,
+      "loss": 0.5481,
+      "step": 15563
+    },
+    {
+      "epoch": 4.131687242798354,
+      "grad_norm": 0.4658986328199154,
+      "learning_rate": 1.1116550501269013e-06,
+      "loss": 0.5316,
+      "step": 15564
+    },
+    {
+      "epoch": 4.131952741271737,
+      "grad_norm": 0.47410322638262953,
+      "learning_rate": 1.1113647202480538e-06,
+      "loss": 0.5229,
+      "step": 15565
+    },
+    {
+      "epoch": 4.1322182397451215,
+      "grad_norm": 0.48771309885865977,
+      "learning_rate": 1.1110744174509952e-06,
+      "loss": 0.5463,
+      "step": 15566
+    },
+    {
+      "epoch": 4.132483738218506,
+      "grad_norm": 0.47224637346761367,
+      "learning_rate": 1.110784141741385e-06,
+      "loss": 0.5318,
+      "step": 15567
+    },
+    {
+      "epoch": 4.132749236691889,
+      "grad_norm": 0.4707038842073985,
+      "learning_rate": 1.1104938931248857e-06,
+      "loss": 0.5524,
+      "step": 15568
+    },
+    {
+      "epoch": 4.133014735165273,
+      "grad_norm": 0.4607022075219156,
+      "learning_rate": 1.1102036716071596e-06,
+      "loss": 0.4783,
+      "step": 15569
+    },
+    {
+      "epoch": 4.133280233638657,
+      "grad_norm": 0.47113615676178994,
+      "learning_rate": 1.1099134771938624e-06,
+      "loss": 0.5294,
+      "step": 15570
+    },
+    {
+      "epoch": 4.13354573211204,
+      "grad_norm": 0.4902044192166875,
+      "learning_rate": 1.1096233098906571e-06,
+      "loss": 0.5146,
+      "step": 15571
+    },
+    {
+      "epoch": 4.133811230585424,
+      "grad_norm": 0.4859085061662544,
+      "learning_rate": 1.1093331697032007e-06,
+      "loss": 0.5225,
+      "step": 15572
+    },
+    {
+      "epoch": 4.134076729058808,
+      "grad_norm": 0.46643061471805203,
+      "learning_rate": 1.109043056637153e-06,
+      "loss": 0.5204,
+      "step": 15573
+    },
+    {
+      "epoch": 4.1343422275321915,
+      "grad_norm": 0.4897898641260012,
+      "learning_rate": 1.1087529706981707e-06,
+      "loss": 0.5528,
+      "step": 15574
+    },
+    {
+      "epoch": 4.134607726005576,
+      "grad_norm": 0.4797901952282531,
+      "learning_rate": 1.1084629118919127e-06,
+      "loss": 0.5352,
+      "step": 15575
+    },
+    {
+      "epoch": 4.13487322447896,
+      "grad_norm": 0.46364227762620613,
+      "learning_rate": 1.1081728802240348e-06,
+      "loss": 0.5197,
+      "step": 15576
+    },
+    {
+      "epoch": 4.135138722952343,
+      "grad_norm": 0.47871785662735733,
+      "learning_rate": 1.1078828757001928e-06,
+      "loss": 0.5537,
+      "step": 15577
+    },
+    {
+      "epoch": 4.135404221425727,
+      "grad_norm": 0.4731982000786994,
+      "learning_rate": 1.1075928983260442e-06,
+      "loss": 0.5459,
+      "step": 15578
+    },
+    {
+      "epoch": 4.135669719899111,
+      "grad_norm": 0.49052294271584956,
+      "learning_rate": 1.1073029481072423e-06,
+      "loss": 0.5589,
+      "step": 15579
+    },
+    {
+      "epoch": 4.135935218372494,
+      "grad_norm": 0.47539561775462214,
+      "learning_rate": 1.1070130250494435e-06,
+      "loss": 0.5182,
+      "step": 15580
+    },
+    {
+      "epoch": 4.136200716845878,
+      "grad_norm": 0.48104889508605886,
+      "learning_rate": 1.1067231291583016e-06,
+      "loss": 0.5189,
+      "step": 15581
+    },
+    {
+      "epoch": 4.1364662153192615,
+      "grad_norm": 0.47762855462445847,
+      "learning_rate": 1.1064332604394698e-06,
+      "loss": 0.5432,
+      "step": 15582
+    },
+    {
+      "epoch": 4.136731713792646,
+      "grad_norm": 0.4813338715069982,
+      "learning_rate": 1.106143418898601e-06,
+      "loss": 0.5443,
+      "step": 15583
+    },
+    {
+      "epoch": 4.13699721226603,
+      "grad_norm": 0.48179542270873427,
+      "learning_rate": 1.1058536045413486e-06,
+      "loss": 0.5599,
+      "step": 15584
+    },
+    {
+      "epoch": 4.137262710739413,
+      "grad_norm": 0.4682756838872559,
+      "learning_rate": 1.1055638173733638e-06,
+      "loss": 0.4961,
+      "step": 15585
+    },
+    {
+      "epoch": 4.137528209212797,
+      "grad_norm": 0.4685254421693967,
+      "learning_rate": 1.1052740574002996e-06,
+      "loss": 0.5201,
+      "step": 15586
+    },
+    {
+      "epoch": 4.137793707686181,
+      "grad_norm": 0.4668272981890748,
+      "learning_rate": 1.104984324627806e-06,
+      "loss": 0.5485,
+      "step": 15587
+    },
+    {
+      "epoch": 4.138059206159564,
+      "grad_norm": 0.4843672061429562,
+      "learning_rate": 1.104694619061533e-06,
+      "loss": 0.5365,
+      "step": 15588
+    },
+    {
+      "epoch": 4.138324704632948,
+      "grad_norm": 0.4754774843218086,
+      "learning_rate": 1.104404940707132e-06,
+      "loss": 0.5689,
+      "step": 15589
+    },
+    {
+      "epoch": 4.138590203106332,
+      "grad_norm": 0.4648329191717306,
+      "learning_rate": 1.1041152895702506e-06,
+      "loss": 0.5304,
+      "step": 15590
+    },
+    {
+      "epoch": 4.138855701579716,
+      "grad_norm": 0.4762410385576523,
+      "learning_rate": 1.1038256656565398e-06,
+      "loss": 0.5321,
+      "step": 15591
+    },
+    {
+      "epoch": 4.1391212000531,
+      "grad_norm": 0.47261792211761755,
+      "learning_rate": 1.103536068971647e-06,
+      "loss": 0.5535,
+      "step": 15592
+    },
+    {
+      "epoch": 4.139386698526484,
+      "grad_norm": 0.4659213927497224,
+      "learning_rate": 1.103246499521219e-06,
+      "loss": 0.5441,
+      "step": 15593
+    },
+    {
+      "epoch": 4.139652196999867,
+      "grad_norm": 0.475025068989063,
+      "learning_rate": 1.1029569573109052e-06,
+      "loss": 0.5567,
+      "step": 15594
+    },
+    {
+      "epoch": 4.139917695473251,
+      "grad_norm": 0.46770203524811976,
+      "learning_rate": 1.1026674423463502e-06,
+      "loss": 0.5518,
+      "step": 15595
+    },
+    {
+      "epoch": 4.140183193946635,
+      "grad_norm": 0.48339099623767884,
+      "learning_rate": 1.1023779546332023e-06,
+      "loss": 0.5722,
+      "step": 15596
+    },
+    {
+      "epoch": 4.140448692420018,
+      "grad_norm": 0.46940204914652384,
+      "learning_rate": 1.1020884941771055e-06,
+      "loss": 0.5508,
+      "step": 15597
+    },
+    {
+      "epoch": 4.140714190893402,
+      "grad_norm": 0.4700794376705288,
+      "learning_rate": 1.1017990609837074e-06,
+      "loss": 0.4877,
+      "step": 15598
+    },
+    {
+      "epoch": 4.1409796893667865,
+      "grad_norm": 0.4771422175118768,
+      "learning_rate": 1.1015096550586491e-06,
+      "loss": 0.5329,
+      "step": 15599
+    },
+    {
+      "epoch": 4.14124518784017,
+      "grad_norm": 0.48015149765604603,
+      "learning_rate": 1.1012202764075774e-06,
+      "loss": 0.5223,
+      "step": 15600
+    },
+    {
+      "epoch": 4.141510686313554,
+      "grad_norm": 0.45896049722337495,
+      "learning_rate": 1.1009309250361344e-06,
+      "loss": 0.5307,
+      "step": 15601
+    },
+    {
+      "epoch": 4.141776184786938,
+      "grad_norm": 0.5018457356281955,
+      "learning_rate": 1.1006416009499648e-06,
+      "loss": 0.5299,
+      "step": 15602
+    },
+    {
+      "epoch": 4.142041683260321,
+      "grad_norm": 0.47232860626027773,
+      "learning_rate": 1.1003523041547094e-06,
+      "loss": 0.5501,
+      "step": 15603
+    },
+    {
+      "epoch": 4.142307181733705,
+      "grad_norm": 0.47563475155287643,
+      "learning_rate": 1.1000630346560118e-06,
+      "loss": 0.5312,
+      "step": 15604
+    },
+    {
+      "epoch": 4.142572680207088,
+      "grad_norm": 0.47294516911440865,
+      "learning_rate": 1.0997737924595125e-06,
+      "loss": 0.552,
+      "step": 15605
+    },
+    {
+      "epoch": 4.142838178680472,
+      "grad_norm": 0.4670639174234273,
+      "learning_rate": 1.0994845775708518e-06,
+      "loss": 0.5366,
+      "step": 15606
+    },
+    {
+      "epoch": 4.1431036771538565,
+      "grad_norm": 0.4896518973428236,
+      "learning_rate": 1.099195389995672e-06,
+      "loss": 0.5614,
+      "step": 15607
+    },
+    {
+      "epoch": 4.14336917562724,
+      "grad_norm": 0.4704579177338021,
+      "learning_rate": 1.0989062297396108e-06,
+      "loss": 0.5654,
+      "step": 15608
+    },
+    {
+      "epoch": 4.143634674100624,
+      "grad_norm": 0.47688782471548824,
+      "learning_rate": 1.0986170968083093e-06,
+      "loss": 0.5553,
+      "step": 15609
+    },
+    {
+      "epoch": 4.143900172574008,
+      "grad_norm": 0.4867397031128611,
+      "learning_rate": 1.0983279912074061e-06,
+      "loss": 0.5716,
+      "step": 15610
+    },
+    {
+      "epoch": 4.144165671047391,
+      "grad_norm": 0.4708918646391312,
+      "learning_rate": 1.0980389129425385e-06,
+      "loss": 0.5278,
+      "step": 15611
+    },
+    {
+      "epoch": 4.144431169520775,
+      "grad_norm": 0.5014593503860955,
+      "learning_rate": 1.0977498620193438e-06,
+      "loss": 0.5496,
+      "step": 15612
+    },
+    {
+      "epoch": 4.144696667994159,
+      "grad_norm": 0.4793916859520966,
+      "learning_rate": 1.0974608384434602e-06,
+      "loss": 0.5044,
+      "step": 15613
+    },
+    {
+      "epoch": 4.1449621664675425,
+      "grad_norm": 0.4751833558243376,
+      "learning_rate": 1.0971718422205252e-06,
+      "loss": 0.5303,
+      "step": 15614
+    },
+    {
+      "epoch": 4.1452276649409265,
+      "grad_norm": 0.4917791479577007,
+      "learning_rate": 1.096882873356174e-06,
+      "loss": 0.5471,
+      "step": 15615
+    },
+    {
+      "epoch": 4.145493163414311,
+      "grad_norm": 0.4801873104571178,
+      "learning_rate": 1.0965939318560421e-06,
+      "loss": 0.5841,
+      "step": 15616
+    },
+    {
+      "epoch": 4.145758661887694,
+      "grad_norm": 0.4535522819737755,
+      "learning_rate": 1.0963050177257637e-06,
+      "loss": 0.5517,
+      "step": 15617
+    },
+    {
+      "epoch": 4.146024160361078,
+      "grad_norm": 0.48636694625259147,
+      "learning_rate": 1.0960161309709753e-06,
+      "loss": 0.5527,
+      "step": 15618
+    },
+    {
+      "epoch": 4.146289658834462,
+      "grad_norm": 0.49205196097159415,
+      "learning_rate": 1.095727271597309e-06,
+      "loss": 0.5587,
+      "step": 15619
+    },
+    {
+      "epoch": 4.146555157307845,
+      "grad_norm": 0.46930758742504225,
+      "learning_rate": 1.0954384396104e-06,
+      "loss": 0.5339,
+      "step": 15620
+    },
+    {
+      "epoch": 4.146820655781229,
+      "grad_norm": 0.47073698252177665,
+      "learning_rate": 1.0951496350158804e-06,
+      "loss": 0.5539,
+      "step": 15621
+    },
+    {
+      "epoch": 4.147086154254613,
+      "grad_norm": 0.47749835500701826,
+      "learning_rate": 1.0948608578193816e-06,
+      "loss": 0.5145,
+      "step": 15622
+    },
+    {
+      "epoch": 4.1473516527279966,
+      "grad_norm": 0.4863400294798167,
+      "learning_rate": 1.0945721080265377e-06,
+      "loss": 0.584,
+      "step": 15623
+    },
+    {
+      "epoch": 4.147617151201381,
+      "grad_norm": 0.4505560926588953,
+      "learning_rate": 1.0942833856429777e-06,
+      "loss": 0.5021,
+      "step": 15624
+    },
+    {
+      "epoch": 4.147882649674765,
+      "grad_norm": 0.4783220294329117,
+      "learning_rate": 1.0939946906743342e-06,
+      "loss": 0.5612,
+      "step": 15625
+    },
+    {
+      "epoch": 4.148148148148148,
+      "grad_norm": 0.4760714907730766,
+      "learning_rate": 1.0937060231262358e-06,
+      "loss": 0.5268,
+      "step": 15626
+    },
+    {
+      "epoch": 4.148413646621532,
+      "grad_norm": 0.47409249678504867,
+      "learning_rate": 1.0934173830043153e-06,
+      "loss": 0.5107,
+      "step": 15627
+    },
+    {
+      "epoch": 4.148679145094916,
+      "grad_norm": 0.46554352700927415,
+      "learning_rate": 1.0931287703141977e-06,
+      "loss": 0.5291,
+      "step": 15628
+    },
+    {
+      "epoch": 4.148944643568299,
+      "grad_norm": 0.48432925912608943,
+      "learning_rate": 1.0928401850615144e-06,
+      "loss": 0.5564,
+      "step": 15629
+    },
+    {
+      "epoch": 4.149210142041683,
+      "grad_norm": 0.49315448153297736,
+      "learning_rate": 1.092551627251892e-06,
+      "loss": 0.5746,
+      "step": 15630
+    },
+    {
+      "epoch": 4.1494756405150675,
+      "grad_norm": 0.496241098270634,
+      "learning_rate": 1.0922630968909602e-06,
+      "loss": 0.5292,
+      "step": 15631
+    },
+    {
+      "epoch": 4.149741138988451,
+      "grad_norm": 0.4756442440758925,
+      "learning_rate": 1.0919745939843435e-06,
+      "loss": 0.4961,
+      "step": 15632
+    },
+    {
+      "epoch": 4.150006637461835,
+      "grad_norm": 0.48058388657686724,
+      "learning_rate": 1.0916861185376706e-06,
+      "loss": 0.5408,
+      "step": 15633
+    },
+    {
+      "epoch": 4.150272135935218,
+      "grad_norm": 0.47799591834016036,
+      "learning_rate": 1.0913976705565666e-06,
+      "loss": 0.5549,
+      "step": 15634
+    },
+    {
+      "epoch": 4.150537634408602,
+      "grad_norm": 0.4895221315472336,
+      "learning_rate": 1.0911092500466558e-06,
+      "loss": 0.5553,
+      "step": 15635
+    },
+    {
+      "epoch": 4.150803132881986,
+      "grad_norm": 0.4709301438419086,
+      "learning_rate": 1.0908208570135654e-06,
+      "loss": 0.5532,
+      "step": 15636
+    },
+    {
+      "epoch": 4.151068631355369,
+      "grad_norm": 0.48130540949252976,
+      "learning_rate": 1.0905324914629173e-06,
+      "loss": 0.5636,
+      "step": 15637
+    },
+    {
+      "epoch": 4.151334129828753,
+      "grad_norm": 0.47873598813665563,
+      "learning_rate": 1.0902441534003378e-06,
+      "loss": 0.5285,
+      "step": 15638
+    },
+    {
+      "epoch": 4.1515996283021375,
+      "grad_norm": 0.4834338954471441,
+      "learning_rate": 1.0899558428314489e-06,
+      "loss": 0.5205,
+      "step": 15639
+    },
+    {
+      "epoch": 4.151865126775521,
+      "grad_norm": 0.46395662207142596,
+      "learning_rate": 1.0896675597618725e-06,
+      "loss": 0.5026,
+      "step": 15640
+    },
+    {
+      "epoch": 4.152130625248905,
+      "grad_norm": 0.5040384996878225,
+      "learning_rate": 1.0893793041972325e-06,
+      "loss": 0.5773,
+      "step": 15641
+    },
+    {
+      "epoch": 4.152396123722289,
+      "grad_norm": 0.4701412011660318,
+      "learning_rate": 1.089091076143149e-06,
+      "loss": 0.5309,
+      "step": 15642
+    },
+    {
+      "epoch": 4.152661622195672,
+      "grad_norm": 0.49043270850002174,
+      "learning_rate": 1.088802875605245e-06,
+      "loss": 0.5106,
+      "step": 15643
+    },
+    {
+      "epoch": 4.152927120669056,
+      "grad_norm": 0.4913715841290042,
+      "learning_rate": 1.0885147025891404e-06,
+      "loss": 0.5447,
+      "step": 15644
+    },
+    {
+      "epoch": 4.15319261914244,
+      "grad_norm": 0.4734505134804812,
+      "learning_rate": 1.088226557100455e-06,
+      "loss": 0.5556,
+      "step": 15645
+    },
+    {
+      "epoch": 4.153458117615823,
+      "grad_norm": 0.4798403915641649,
+      "learning_rate": 1.087938439144807e-06,
+      "loss": 0.5185,
+      "step": 15646
+    },
+    {
+      "epoch": 4.1537236160892075,
+      "grad_norm": 0.48184063571663016,
+      "learning_rate": 1.0876503487278178e-06,
+      "loss": 0.5673,
+      "step": 15647
+    },
+    {
+      "epoch": 4.153989114562592,
+      "grad_norm": 0.47539130489602216,
+      "learning_rate": 1.0873622858551042e-06,
+      "loss": 0.5385,
+      "step": 15648
+    },
+    {
+      "epoch": 4.154254613035975,
+      "grad_norm": 0.4882432993942981,
+      "learning_rate": 1.0870742505322854e-06,
+      "loss": 0.5741,
+      "step": 15649
+    },
+    {
+      "epoch": 4.154520111509359,
+      "grad_norm": 0.4733264755012237,
+      "learning_rate": 1.0867862427649783e-06,
+      "loss": 0.5457,
+      "step": 15650
+    },
+    {
+      "epoch": 4.154785609982743,
+      "grad_norm": 0.49683287541177046,
+      "learning_rate": 1.0864982625587986e-06,
+      "loss": 0.5629,
+      "step": 15651
+    },
+    {
+      "epoch": 4.155051108456126,
+      "grad_norm": 0.47911576978441145,
+      "learning_rate": 1.0862103099193646e-06,
+      "loss": 0.5491,
+      "step": 15652
+    },
+    {
+      "epoch": 4.15531660692951,
+      "grad_norm": 0.4833308084409787,
+      "learning_rate": 1.0859223848522902e-06,
+      "loss": 0.5118,
+      "step": 15653
+    },
+    {
+      "epoch": 4.155582105402894,
+      "grad_norm": 0.4784609363065905,
+      "learning_rate": 1.0856344873631924e-06,
+      "loss": 0.5098,
+      "step": 15654
+    },
+    {
+      "epoch": 4.1558476038762775,
+      "grad_norm": 0.4668208354231387,
+      "learning_rate": 1.0853466174576843e-06,
+      "loss": 0.522,
+      "step": 15655
+    },
+    {
+      "epoch": 4.156113102349662,
+      "grad_norm": 0.46346170465603853,
+      "learning_rate": 1.0850587751413824e-06,
+      "loss": 0.5001,
+      "step": 15656
+    },
+    {
+      "epoch": 4.156378600823046,
+      "grad_norm": 0.47095824474909026,
+      "learning_rate": 1.0847709604198972e-06,
+      "loss": 0.5668,
+      "step": 15657
+    },
+    {
+      "epoch": 4.156644099296429,
+      "grad_norm": 0.4763162224508801,
+      "learning_rate": 1.0844831732988434e-06,
+      "loss": 0.5577,
+      "step": 15658
+    },
+    {
+      "epoch": 4.156909597769813,
+      "grad_norm": 0.4759551912536526,
+      "learning_rate": 1.0841954137838342e-06,
+      "loss": 0.5198,
+      "step": 15659
+    },
+    {
+      "epoch": 4.157175096243197,
+      "grad_norm": 0.47322081924139175,
+      "learning_rate": 1.08390768188048e-06,
+      "loss": 0.543,
+      "step": 15660
+    },
+    {
+      "epoch": 4.15744059471658,
+      "grad_norm": 0.4709030150311447,
+      "learning_rate": 1.0836199775943942e-06,
+      "loss": 0.5118,
+      "step": 15661
+    },
+    {
+      "epoch": 4.157706093189964,
+      "grad_norm": 0.47927412791656393,
+      "learning_rate": 1.0833323009311867e-06,
+      "loss": 0.5384,
+      "step": 15662
+    },
+    {
+      "epoch": 4.1579715916633475,
+      "grad_norm": 0.47047171960373263,
+      "learning_rate": 1.083044651896468e-06,
+      "loss": 0.5348,
+      "step": 15663
+    },
+    {
+      "epoch": 4.158237090136732,
+      "grad_norm": 0.47454374726311693,
+      "learning_rate": 1.0827570304958467e-06,
+      "loss": 0.5494,
+      "step": 15664
+    },
+    {
+      "epoch": 4.158502588610116,
+      "grad_norm": 0.47845325854692045,
+      "learning_rate": 1.0824694367349346e-06,
+      "loss": 0.5094,
+      "step": 15665
+    },
+    {
+      "epoch": 4.158768087083499,
+      "grad_norm": 0.4692408311694858,
+      "learning_rate": 1.0821818706193379e-06,
+      "loss": 0.5521,
+      "step": 15666
+    },
+    {
+      "epoch": 4.159033585556883,
+      "grad_norm": 0.4910113783130177,
+      "learning_rate": 1.0818943321546673e-06,
+      "loss": 0.5675,
+      "step": 15667
+    },
+    {
+      "epoch": 4.159299084030267,
+      "grad_norm": 0.48358433429505404,
+      "learning_rate": 1.0816068213465295e-06,
+      "loss": 0.5642,
+      "step": 15668
+    },
+    {
+      "epoch": 4.15956458250365,
+      "grad_norm": 0.46768123465324485,
+      "learning_rate": 1.0813193382005306e-06,
+      "loss": 0.5214,
+      "step": 15669
+    },
+    {
+      "epoch": 4.159830080977034,
+      "grad_norm": 0.47353513836109445,
+      "learning_rate": 1.0810318827222785e-06,
+      "loss": 0.5631,
+      "step": 15670
+    },
+    {
+      "epoch": 4.160095579450418,
+      "grad_norm": 0.4892440492161654,
+      "learning_rate": 1.0807444549173784e-06,
+      "loss": 0.531,
+      "step": 15671
+    },
+    {
+      "epoch": 4.160361077923802,
+      "grad_norm": 0.4776180910061011,
+      "learning_rate": 1.0804570547914372e-06,
+      "loss": 0.5186,
+      "step": 15672
+    },
+    {
+      "epoch": 4.160626576397186,
+      "grad_norm": 0.4827916115411352,
+      "learning_rate": 1.0801696823500592e-06,
+      "loss": 0.5726,
+      "step": 15673
+    },
+    {
+      "epoch": 4.16089207487057,
+      "grad_norm": 0.48963619137701425,
+      "learning_rate": 1.0798823375988487e-06,
+      "loss": 0.538,
+      "step": 15674
+    },
+    {
+      "epoch": 4.161157573343953,
+      "grad_norm": 0.47304734721066966,
+      "learning_rate": 1.0795950205434088e-06,
+      "loss": 0.4914,
+      "step": 15675
+    },
+    {
+      "epoch": 4.161423071817337,
+      "grad_norm": 0.4752259889217294,
+      "learning_rate": 1.0793077311893447e-06,
+      "loss": 0.5196,
+      "step": 15676
+    },
+    {
+      "epoch": 4.161688570290721,
+      "grad_norm": 0.4869680743126387,
+      "learning_rate": 1.0790204695422571e-06,
+      "loss": 0.5811,
+      "step": 15677
+    },
+    {
+      "epoch": 4.161954068764104,
+      "grad_norm": 0.4625635226420737,
+      "learning_rate": 1.0787332356077496e-06,
+      "loss": 0.5709,
+      "step": 15678
+    },
+    {
+      "epoch": 4.162219567237488,
+      "grad_norm": 0.4762615401200068,
+      "learning_rate": 1.078446029391426e-06,
+      "loss": 0.5143,
+      "step": 15679
+    },
+    {
+      "epoch": 4.1624850657108725,
+      "grad_norm": 0.4788627478488273,
+      "learning_rate": 1.078158850898883e-06,
+      "loss": 0.5638,
+      "step": 15680
+    },
+    {
+      "epoch": 4.162750564184256,
+      "grad_norm": 0.47660238093927376,
+      "learning_rate": 1.077871700135725e-06,
+      "loss": 0.5461,
+      "step": 15681
+    },
+    {
+      "epoch": 4.16301606265764,
+      "grad_norm": 0.4855898410808992,
+      "learning_rate": 1.07758457710755e-06,
+      "loss": 0.5553,
+      "step": 15682
+    },
+    {
+      "epoch": 4.163281561131024,
+      "grad_norm": 0.46956226094612036,
+      "learning_rate": 1.077297481819959e-06,
+      "loss": 0.5247,
+      "step": 15683
+    },
+    {
+      "epoch": 4.163547059604407,
+      "grad_norm": 0.45767437513522463,
+      "learning_rate": 1.0770104142785499e-06,
+      "loss": 0.5088,
+      "step": 15684
+    },
+    {
+      "epoch": 4.163812558077791,
+      "grad_norm": 0.48756026668722924,
+      "learning_rate": 1.0767233744889221e-06,
+      "loss": 0.6005,
+      "step": 15685
+    },
+    {
+      "epoch": 4.164078056551175,
+      "grad_norm": 0.48423336578759374,
+      "learning_rate": 1.076436362456674e-06,
+      "loss": 0.5685,
+      "step": 15686
+    },
+    {
+      "epoch": 4.164343555024558,
+      "grad_norm": 0.48627331479466046,
+      "learning_rate": 1.076149378187401e-06,
+      "loss": 0.5421,
+      "step": 15687
+    },
+    {
+      "epoch": 4.1646090534979425,
+      "grad_norm": 0.48211017665897593,
+      "learning_rate": 1.0758624216867022e-06,
+      "loss": 0.5436,
+      "step": 15688
+    },
+    {
+      "epoch": 4.164874551971327,
+      "grad_norm": 0.480244637142,
+      "learning_rate": 1.0755754929601726e-06,
+      "loss": 0.5339,
+      "step": 15689
+    },
+    {
+      "epoch": 4.16514005044471,
+      "grad_norm": 0.4781502935866322,
+      "learning_rate": 1.0752885920134098e-06,
+      "loss": 0.5803,
+      "step": 15690
+    },
+    {
+      "epoch": 4.165405548918094,
+      "grad_norm": 0.4743428694996764,
+      "learning_rate": 1.0750017188520063e-06,
+      "loss": 0.5548,
+      "step": 15691
+    },
+    {
+      "epoch": 4.165671047391477,
+      "grad_norm": 0.47841753093213657,
+      "learning_rate": 1.0747148734815587e-06,
+      "loss": 0.54,
+      "step": 15692
+    },
+    {
+      "epoch": 4.165936545864861,
+      "grad_norm": 0.46059801769971387,
+      "learning_rate": 1.0744280559076602e-06,
+      "loss": 0.5043,
+      "step": 15693
+    },
+    {
+      "epoch": 4.166202044338245,
+      "grad_norm": 0.4619070769801546,
+      "learning_rate": 1.074141266135906e-06,
+      "loss": 0.5172,
+      "step": 15694
+    },
+    {
+      "epoch": 4.1664675428116285,
+      "grad_norm": 0.48613543849746815,
+      "learning_rate": 1.0738545041718873e-06,
+      "loss": 0.5706,
+      "step": 15695
+    },
+    {
+      "epoch": 4.1667330412850125,
+      "grad_norm": 0.4814975462576854,
+      "learning_rate": 1.0735677700211982e-06,
+      "loss": 0.5604,
+      "step": 15696
+    },
+    {
+      "epoch": 4.166998539758397,
+      "grad_norm": 0.45904673752790504,
+      "learning_rate": 1.0732810636894304e-06,
+      "loss": 0.5168,
+      "step": 15697
+    },
+    {
+      "epoch": 4.16726403823178,
+      "grad_norm": 0.4657578420663952,
+      "learning_rate": 1.072994385182174e-06,
+      "loss": 0.5272,
+      "step": 15698
+    },
+    {
+      "epoch": 4.167529536705164,
+      "grad_norm": 0.4706186582753637,
+      "learning_rate": 1.0727077345050218e-06,
+      "loss": 0.5439,
+      "step": 15699
+    },
+    {
+      "epoch": 4.167795035178548,
+      "grad_norm": 0.47812731891053234,
+      "learning_rate": 1.0724211116635628e-06,
+      "loss": 0.5661,
+      "step": 15700
+    },
+    {
+      "epoch": 4.168060533651931,
+      "grad_norm": 0.4773364919210583,
+      "learning_rate": 1.072134516663388e-06,
+      "loss": 0.558,
+      "step": 15701
+    },
+    {
+      "epoch": 4.168326032125315,
+      "grad_norm": 0.48999880360261217,
+      "learning_rate": 1.0718479495100864e-06,
+      "loss": 0.5459,
+      "step": 15702
+    },
+    {
+      "epoch": 4.168591530598699,
+      "grad_norm": 0.48169021501988535,
+      "learning_rate": 1.0715614102092455e-06,
+      "loss": 0.5333,
+      "step": 15703
+    },
+    {
+      "epoch": 4.1688570290720826,
+      "grad_norm": 0.4782256139721503,
+      "learning_rate": 1.0712748987664554e-06,
+      "loss": 0.5581,
+      "step": 15704
+    },
+    {
+      "epoch": 4.169122527545467,
+      "grad_norm": 0.4835018960535134,
+      "learning_rate": 1.070988415187302e-06,
+      "loss": 0.5354,
+      "step": 15705
+    },
+    {
+      "epoch": 4.169388026018851,
+      "grad_norm": 0.48371136889518485,
+      "learning_rate": 1.0707019594773744e-06,
+      "loss": 0.5516,
+      "step": 15706
+    },
+    {
+      "epoch": 4.169653524492234,
+      "grad_norm": 0.4835419597659858,
+      "learning_rate": 1.070415531642257e-06,
+      "loss": 0.5417,
+      "step": 15707
+    },
+    {
+      "epoch": 4.169919022965618,
+      "grad_norm": 0.48807210873478907,
+      "learning_rate": 1.070129131687539e-06,
+      "loss": 0.5417,
+      "step": 15708
+    },
+    {
+      "epoch": 4.170184521439002,
+      "grad_norm": 0.4581948774842411,
+      "learning_rate": 1.069842759618802e-06,
+      "loss": 0.5309,
+      "step": 15709
+    },
+    {
+      "epoch": 4.170450019912385,
+      "grad_norm": 0.47472127779898404,
+      "learning_rate": 1.0695564154416333e-06,
+      "loss": 0.5038,
+      "step": 15710
+    },
+    {
+      "epoch": 4.170715518385769,
+      "grad_norm": 0.47985532794409486,
+      "learning_rate": 1.0692700991616161e-06,
+      "loss": 0.5542,
+      "step": 15711
+    },
+    {
+      "epoch": 4.1709810168591535,
+      "grad_norm": 0.47393132832617674,
+      "learning_rate": 1.0689838107843362e-06,
+      "loss": 0.5151,
+      "step": 15712
+    },
+    {
+      "epoch": 4.171246515332537,
+      "grad_norm": 0.46779888435621175,
+      "learning_rate": 1.0686975503153756e-06,
+      "loss": 0.4988,
+      "step": 15713
+    },
+    {
+      "epoch": 4.171512013805921,
+      "grad_norm": 0.4575252250191547,
+      "learning_rate": 1.0684113177603161e-06,
+      "loss": 0.5465,
+      "step": 15714
+    },
+    {
+      "epoch": 4.171777512279305,
+      "grad_norm": 0.4705538585063664,
+      "learning_rate": 1.0681251131247423e-06,
+      "loss": 0.5779,
+      "step": 15715
+    },
+    {
+      "epoch": 4.172043010752688,
+      "grad_norm": 0.481342298589126,
+      "learning_rate": 1.0678389364142336e-06,
+      "loss": 0.5302,
+      "step": 15716
+    },
+    {
+      "epoch": 4.172308509226072,
+      "grad_norm": 0.48902391370159254,
+      "learning_rate": 1.0675527876343727e-06,
+      "loss": 0.5563,
+      "step": 15717
+    },
+    {
+      "epoch": 4.172574007699455,
+      "grad_norm": 0.47558219684957925,
+      "learning_rate": 1.067266666790739e-06,
+      "loss": 0.4971,
+      "step": 15718
+    },
+    {
+      "epoch": 4.172839506172839,
+      "grad_norm": 0.4784512161688012,
+      "learning_rate": 1.0669805738889152e-06,
+      "loss": 0.5537,
+      "step": 15719
+    },
+    {
+      "epoch": 4.1731050046462235,
+      "grad_norm": 0.489921125433436,
+      "learning_rate": 1.0666945089344771e-06,
+      "loss": 0.5427,
+      "step": 15720
+    },
+    {
+      "epoch": 4.173370503119607,
+      "grad_norm": 0.47990864313583836,
+      "learning_rate": 1.0664084719330061e-06,
+      "loss": 0.5633,
+      "step": 15721
+    },
+    {
+      "epoch": 4.173636001592991,
+      "grad_norm": 0.47401647350908505,
+      "learning_rate": 1.066122462890079e-06,
+      "loss": 0.5538,
+      "step": 15722
+    },
+    {
+      "epoch": 4.173901500066375,
+      "grad_norm": 0.47581908227713804,
+      "learning_rate": 1.0658364818112758e-06,
+      "loss": 0.5312,
+      "step": 15723
+    },
+    {
+      "epoch": 4.174166998539758,
+      "grad_norm": 0.4855315803699059,
+      "learning_rate": 1.0655505287021714e-06,
+      "loss": 0.5612,
+      "step": 15724
+    },
+    {
+      "epoch": 4.174432497013142,
+      "grad_norm": 0.4826275883005352,
+      "learning_rate": 1.0652646035683452e-06,
+      "loss": 0.5263,
+      "step": 15725
+    },
+    {
+      "epoch": 4.174697995486526,
+      "grad_norm": 0.4734590465662751,
+      "learning_rate": 1.0649787064153716e-06,
+      "loss": 0.5258,
+      "step": 15726
+    },
+    {
+      "epoch": 4.174963493959909,
+      "grad_norm": 0.4760709919088463,
+      "learning_rate": 1.064692837248826e-06,
+      "loss": 0.5475,
+      "step": 15727
+    },
+    {
+      "epoch": 4.1752289924332935,
+      "grad_norm": 0.48253965061722287,
+      "learning_rate": 1.0644069960742853e-06,
+      "loss": 0.5265,
+      "step": 15728
+    },
+    {
+      "epoch": 4.175494490906678,
+      "grad_norm": 0.4810148237581286,
+      "learning_rate": 1.064121182897322e-06,
+      "loss": 0.5261,
+      "step": 15729
+    },
+    {
+      "epoch": 4.175759989380061,
+      "grad_norm": 0.47672167900209766,
+      "learning_rate": 1.0638353977235125e-06,
+      "loss": 0.5465,
+      "step": 15730
+    },
+    {
+      "epoch": 4.176025487853445,
+      "grad_norm": 0.4833559058161596,
+      "learning_rate": 1.0635496405584286e-06,
+      "loss": 0.5644,
+      "step": 15731
+    },
+    {
+      "epoch": 4.176290986326829,
+      "grad_norm": 0.47290001327260356,
+      "learning_rate": 1.0632639114076432e-06,
+      "loss": 0.5409,
+      "step": 15732
+    },
+    {
+      "epoch": 4.176556484800212,
+      "grad_norm": 0.4757374099398108,
+      "learning_rate": 1.0629782102767303e-06,
+      "loss": 0.572,
+      "step": 15733
+    },
+    {
+      "epoch": 4.176821983273596,
+      "grad_norm": 0.4865106661101705,
+      "learning_rate": 1.0626925371712596e-06,
+      "loss": 0.5657,
+      "step": 15734
+    },
+    {
+      "epoch": 4.17708748174698,
+      "grad_norm": 0.47053084093472325,
+      "learning_rate": 1.0624068920968044e-06,
+      "loss": 0.5395,
+      "step": 15735
+    },
+    {
+      "epoch": 4.1773529802203635,
+      "grad_norm": 0.46430694205337447,
+      "learning_rate": 1.0621212750589346e-06,
+      "loss": 0.52,
+      "step": 15736
+    },
+    {
+      "epoch": 4.177618478693748,
+      "grad_norm": 0.4747787480069449,
+      "learning_rate": 1.0618356860632209e-06,
+      "loss": 0.5115,
+      "step": 15737
+    },
+    {
+      "epoch": 4.177883977167132,
+      "grad_norm": 0.471792817491463,
+      "learning_rate": 1.0615501251152314e-06,
+      "loss": 0.5148,
+      "step": 15738
+    },
+    {
+      "epoch": 4.178149475640515,
+      "grad_norm": 0.4767625350760104,
+      "learning_rate": 1.0612645922205373e-06,
+      "loss": 0.5184,
+      "step": 15739
+    },
+    {
+      "epoch": 4.178414974113899,
+      "grad_norm": 0.4779025911445821,
+      "learning_rate": 1.0609790873847051e-06,
+      "loss": 0.5479,
+      "step": 15740
+    },
+    {
+      "epoch": 4.178680472587283,
+      "grad_norm": 0.47172344489076556,
+      "learning_rate": 1.0606936106133054e-06,
+      "loss": 0.539,
+      "step": 15741
+    },
+    {
+      "epoch": 4.178945971060666,
+      "grad_norm": 0.46882636633974095,
+      "learning_rate": 1.060408161911904e-06,
+      "loss": 0.5214,
+      "step": 15742
+    },
+    {
+      "epoch": 4.17921146953405,
+      "grad_norm": 0.47229362713666023,
+      "learning_rate": 1.0601227412860674e-06,
+      "loss": 0.5509,
+      "step": 15743
+    },
+    {
+      "epoch": 4.1794769680074335,
+      "grad_norm": 0.474879630717398,
+      "learning_rate": 1.0598373487413634e-06,
+      "loss": 0.5154,
+      "step": 15744
+    },
+    {
+      "epoch": 4.179742466480818,
+      "grad_norm": 0.45010427264510255,
+      "learning_rate": 1.0595519842833565e-06,
+      "loss": 0.4937,
+      "step": 15745
+    },
+    {
+      "epoch": 4.180007964954202,
+      "grad_norm": 0.48301703582448147,
+      "learning_rate": 1.0592666479176139e-06,
+      "loss": 0.521,
+      "step": 15746
+    },
+    {
+      "epoch": 4.180273463427585,
+      "grad_norm": 0.4785468044719761,
+      "learning_rate": 1.058981339649698e-06,
+      "loss": 0.5472,
+      "step": 15747
+    },
+    {
+      "epoch": 4.180538961900969,
+      "grad_norm": 0.47447337999250233,
+      "learning_rate": 1.0586960594851762e-06,
+      "loss": 0.5204,
+      "step": 15748
+    },
+    {
+      "epoch": 4.180804460374353,
+      "grad_norm": 0.4778942286194906,
+      "learning_rate": 1.0584108074296082e-06,
+      "loss": 0.5366,
+      "step": 15749
+    },
+    {
+      "epoch": 4.181069958847736,
+      "grad_norm": 0.4945115402202853,
+      "learning_rate": 1.0581255834885589e-06,
+      "loss": 0.5668,
+      "step": 15750
+    },
+    {
+      "epoch": 4.18133545732112,
+      "grad_norm": 0.4694389901768765,
+      "learning_rate": 1.0578403876675921e-06,
+      "loss": 0.5296,
+      "step": 15751
+    },
+    {
+      "epoch": 4.181600955794504,
+      "grad_norm": 0.46976216932579046,
+      "learning_rate": 1.057555219972268e-06,
+      "loss": 0.5557,
+      "step": 15752
+    },
+    {
+      "epoch": 4.181866454267888,
+      "grad_norm": 0.4794819469668981,
+      "learning_rate": 1.0572700804081497e-06,
+      "loss": 0.5259,
+      "step": 15753
+    },
+    {
+      "epoch": 4.182131952741272,
+      "grad_norm": 0.4603142029328358,
+      "learning_rate": 1.0569849689807973e-06,
+      "loss": 0.5087,
+      "step": 15754
+    },
+    {
+      "epoch": 4.182397451214656,
+      "grad_norm": 0.4912490196602543,
+      "learning_rate": 1.0566998856957707e-06,
+      "loss": 0.5443,
+      "step": 15755
+    },
+    {
+      "epoch": 4.182662949688039,
+      "grad_norm": 0.4925570321937711,
+      "learning_rate": 1.0564148305586296e-06,
+      "loss": 0.5252,
+      "step": 15756
+    },
+    {
+      "epoch": 4.182928448161423,
+      "grad_norm": 0.4795003420522764,
+      "learning_rate": 1.0561298035749348e-06,
+      "loss": 0.518,
+      "step": 15757
+    },
+    {
+      "epoch": 4.183193946634807,
+      "grad_norm": 0.4668182639786524,
+      "learning_rate": 1.055844804750243e-06,
+      "loss": 0.5411,
+      "step": 15758
+    },
+    {
+      "epoch": 4.18345944510819,
+      "grad_norm": 0.47684510322568757,
+      "learning_rate": 1.055559834090114e-06,
+      "loss": 0.5435,
+      "step": 15759
+    },
+    {
+      "epoch": 4.183724943581574,
+      "grad_norm": 0.4684372087409417,
+      "learning_rate": 1.0552748916001054e-06,
+      "loss": 0.5509,
+      "step": 15760
+    },
+    {
+      "epoch": 4.1839904420549585,
+      "grad_norm": 0.46994319082365404,
+      "learning_rate": 1.0549899772857724e-06,
+      "loss": 0.5363,
+      "step": 15761
+    },
+    {
+      "epoch": 4.184255940528342,
+      "grad_norm": 0.46786250486815784,
+      "learning_rate": 1.054705091152674e-06,
+      "loss": 0.5478,
+      "step": 15762
+    },
+    {
+      "epoch": 4.184521439001726,
+      "grad_norm": 0.4779272570984364,
+      "learning_rate": 1.0544202332063637e-06,
+      "loss": 0.5051,
+      "step": 15763
+    },
+    {
+      "epoch": 4.18478693747511,
+      "grad_norm": 0.4652851216827191,
+      "learning_rate": 1.0541354034523997e-06,
+      "loss": 0.5172,
+      "step": 15764
+    },
+    {
+      "epoch": 4.185052435948493,
+      "grad_norm": 0.4785416639510209,
+      "learning_rate": 1.0538506018963352e-06,
+      "loss": 0.5518,
+      "step": 15765
+    },
+    {
+      "epoch": 4.185317934421877,
+      "grad_norm": 0.4694615231733207,
+      "learning_rate": 1.0535658285437245e-06,
+      "loss": 0.5258,
+      "step": 15766
+    },
+    {
+      "epoch": 4.185583432895261,
+      "grad_norm": 0.47578635822774545,
+      "learning_rate": 1.0532810834001208e-06,
+      "loss": 0.5335,
+      "step": 15767
+    },
+    {
+      "epoch": 4.185848931368644,
+      "grad_norm": 0.4698792494983958,
+      "learning_rate": 1.0529963664710794e-06,
+      "loss": 0.5464,
+      "step": 15768
+    },
+    {
+      "epoch": 4.1861144298420285,
+      "grad_norm": 0.4748498706339072,
+      "learning_rate": 1.0527116777621505e-06,
+      "loss": 0.5495,
+      "step": 15769
+    },
+    {
+      "epoch": 4.186379928315413,
+      "grad_norm": 0.46611474543374687,
+      "learning_rate": 1.0524270172788874e-06,
+      "loss": 0.541,
+      "step": 15770
+    },
+    {
+      "epoch": 4.186645426788796,
+      "grad_norm": 0.4886192620314173,
+      "learning_rate": 1.052142385026844e-06,
+      "loss": 0.5931,
+      "step": 15771
+    },
+    {
+      "epoch": 4.18691092526218,
+      "grad_norm": 0.47441555552000675,
+      "learning_rate": 1.0518577810115668e-06,
+      "loss": 0.5759,
+      "step": 15772
+    },
+    {
+      "epoch": 4.187176423735563,
+      "grad_norm": 0.46505979551740484,
+      "learning_rate": 1.0515732052386095e-06,
+      "loss": 0.5104,
+      "step": 15773
+    },
+    {
+      "epoch": 4.187441922208947,
+      "grad_norm": 0.4556154464288007,
+      "learning_rate": 1.0512886577135203e-06,
+      "loss": 0.5502,
+      "step": 15774
+    },
+    {
+      "epoch": 4.187707420682331,
+      "grad_norm": 0.4809501149952131,
+      "learning_rate": 1.05100413844185e-06,
+      "loss": 0.56,
+      "step": 15775
+    },
+    {
+      "epoch": 4.1879729191557145,
+      "grad_norm": 0.4781242533682232,
+      "learning_rate": 1.0507196474291462e-06,
+      "loss": 0.5522,
+      "step": 15776
+    },
+    {
+      "epoch": 4.1882384176290985,
+      "grad_norm": 0.46930550074800953,
+      "learning_rate": 1.0504351846809585e-06,
+      "loss": 0.536,
+      "step": 15777
+    },
+    {
+      "epoch": 4.188503916102483,
+      "grad_norm": 0.4791523379790535,
+      "learning_rate": 1.0501507502028339e-06,
+      "loss": 0.5464,
+      "step": 15778
+    },
+    {
+      "epoch": 4.188769414575866,
+      "grad_norm": 0.4781422789645589,
+      "learning_rate": 1.0498663440003185e-06,
+      "loss": 0.5413,
+      "step": 15779
+    },
+    {
+      "epoch": 4.18903491304925,
+      "grad_norm": 0.46019783224004224,
+      "learning_rate": 1.0495819660789607e-06,
+      "loss": 0.5148,
+      "step": 15780
+    },
+    {
+      "epoch": 4.189300411522634,
+      "grad_norm": 0.4773654089268839,
+      "learning_rate": 1.049297616444305e-06,
+      "loss": 0.543,
+      "step": 15781
+    },
+    {
+      "epoch": 4.189565909996017,
+      "grad_norm": 0.4825802363934642,
+      "learning_rate": 1.0490132951018986e-06,
+      "loss": 0.5119,
+      "step": 15782
+    },
+    {
+      "epoch": 4.189831408469401,
+      "grad_norm": 0.48507517733597844,
+      "learning_rate": 1.0487290020572857e-06,
+      "loss": 0.5426,
+      "step": 15783
+    },
+    {
+      "epoch": 4.190096906942785,
+      "grad_norm": 0.4897126798205443,
+      "learning_rate": 1.0484447373160106e-06,
+      "loss": 0.5554,
+      "step": 15784
+    },
+    {
+      "epoch": 4.190362405416169,
+      "grad_norm": 0.49361586397445456,
+      "learning_rate": 1.048160500883616e-06,
+      "loss": 0.5396,
+      "step": 15785
+    },
+    {
+      "epoch": 4.190627903889553,
+      "grad_norm": 0.4632751690889525,
+      "learning_rate": 1.0478762927656474e-06,
+      "loss": 0.5256,
+      "step": 15786
+    },
+    {
+      "epoch": 4.190893402362937,
+      "grad_norm": 0.4715623999435698,
+      "learning_rate": 1.0475921129676455e-06,
+      "loss": 0.5667,
+      "step": 15787
+    },
+    {
+      "epoch": 4.19115890083632,
+      "grad_norm": 0.46614743397444725,
+      "learning_rate": 1.0473079614951546e-06,
+      "loss": 0.5153,
+      "step": 15788
+    },
+    {
+      "epoch": 4.191424399309704,
+      "grad_norm": 0.5017620379302506,
+      "learning_rate": 1.0470238383537152e-06,
+      "loss": 0.5663,
+      "step": 15789
+    },
+    {
+      "epoch": 4.191689897783088,
+      "grad_norm": 0.48458098953727874,
+      "learning_rate": 1.0467397435488673e-06,
+      "loss": 0.5731,
+      "step": 15790
+    },
+    {
+      "epoch": 4.191955396256471,
+      "grad_norm": 0.4768514341780122,
+      "learning_rate": 1.046455677086154e-06,
+      "loss": 0.5482,
+      "step": 15791
+    },
+    {
+      "epoch": 4.192220894729855,
+      "grad_norm": 0.4841196781653079,
+      "learning_rate": 1.0461716389711128e-06,
+      "loss": 0.5392,
+      "step": 15792
+    },
+    {
+      "epoch": 4.1924863932032395,
+      "grad_norm": 0.4684832148464701,
+      "learning_rate": 1.0458876292092848e-06,
+      "loss": 0.5369,
+      "step": 15793
+    },
+    {
+      "epoch": 4.192751891676623,
+      "grad_norm": 0.47899904644737745,
+      "learning_rate": 1.0456036478062085e-06,
+      "loss": 0.5414,
+      "step": 15794
+    },
+    {
+      "epoch": 4.193017390150007,
+      "grad_norm": 0.4766981606065527,
+      "learning_rate": 1.0453196947674212e-06,
+      "loss": 0.5394,
+      "step": 15795
+    },
+    {
+      "epoch": 4.193282888623391,
+      "grad_norm": 0.4794784140554601,
+      "learning_rate": 1.0450357700984624e-06,
+      "loss": 0.5286,
+      "step": 15796
+    },
+    {
+      "epoch": 4.193548387096774,
+      "grad_norm": 0.4696173030359798,
+      "learning_rate": 1.0447518738048676e-06,
+      "loss": 0.5337,
+      "step": 15797
+    },
+    {
+      "epoch": 4.193813885570158,
+      "grad_norm": 0.4754889404917988,
+      "learning_rate": 1.044468005892175e-06,
+      "loss": 0.5589,
+      "step": 15798
+    },
+    {
+      "epoch": 4.194079384043542,
+      "grad_norm": 0.48365008091434597,
+      "learning_rate": 1.0441841663659193e-06,
+      "loss": 0.5751,
+      "step": 15799
+    },
+    {
+      "epoch": 4.194344882516925,
+      "grad_norm": 0.4739409824701588,
+      "learning_rate": 1.0439003552316387e-06,
+      "loss": 0.5473,
+      "step": 15800
+    },
+    {
+      "epoch": 4.1946103809903095,
+      "grad_norm": 0.4740853854824402,
+      "learning_rate": 1.0436165724948644e-06,
+      "loss": 0.5344,
+      "step": 15801
+    },
+    {
+      "epoch": 4.194875879463693,
+      "grad_norm": 0.48335579784914723,
+      "learning_rate": 1.043332818161134e-06,
+      "loss": 0.587,
+      "step": 15802
+    },
+    {
+      "epoch": 4.195141377937077,
+      "grad_norm": 0.4628608439969883,
+      "learning_rate": 1.0430490922359787e-06,
+      "loss": 0.5156,
+      "step": 15803
+    },
+    {
+      "epoch": 4.195406876410461,
+      "grad_norm": 0.46762070763234215,
+      "learning_rate": 1.0427653947249345e-06,
+      "loss": 0.5044,
+      "step": 15804
+    },
+    {
+      "epoch": 4.195672374883844,
+      "grad_norm": 0.4660179700702556,
+      "learning_rate": 1.0424817256335324e-06,
+      "loss": 0.5377,
+      "step": 15805
+    },
+    {
+      "epoch": 4.195937873357228,
+      "grad_norm": 0.47089826122905587,
+      "learning_rate": 1.042198084967306e-06,
+      "loss": 0.5221,
+      "step": 15806
+    },
+    {
+      "epoch": 4.196203371830612,
+      "grad_norm": 0.47687851371220574,
+      "learning_rate": 1.041914472731786e-06,
+      "loss": 0.5234,
+      "step": 15807
+    },
+    {
+      "epoch": 4.196468870303995,
+      "grad_norm": 0.48087517686309145,
+      "learning_rate": 1.041630888932503e-06,
+      "loss": 0.5373,
+      "step": 15808
+    },
+    {
+      "epoch": 4.1967343687773795,
+      "grad_norm": 0.48255578804077875,
+      "learning_rate": 1.0413473335749894e-06,
+      "loss": 0.5427,
+      "step": 15809
+    },
+    {
+      "epoch": 4.196999867250764,
+      "grad_norm": 0.4791727927412968,
+      "learning_rate": 1.0410638066647733e-06,
+      "loss": 0.517,
+      "step": 15810
+    },
+    {
+      "epoch": 4.197265365724147,
+      "grad_norm": 0.4876404795385016,
+      "learning_rate": 1.0407803082073867e-06,
+      "loss": 0.5307,
+      "step": 15811
+    },
+    {
+      "epoch": 4.197530864197531,
+      "grad_norm": 0.48055114090260526,
+      "learning_rate": 1.0404968382083553e-06,
+      "loss": 0.5251,
+      "step": 15812
+    },
+    {
+      "epoch": 4.197796362670915,
+      "grad_norm": 0.4796792245686959,
+      "learning_rate": 1.0402133966732098e-06,
+      "loss": 0.559,
+      "step": 15813
+    },
+    {
+      "epoch": 4.198061861144298,
+      "grad_norm": 0.4833996014113798,
+      "learning_rate": 1.0399299836074762e-06,
+      "loss": 0.5615,
+      "step": 15814
+    },
+    {
+      "epoch": 4.198327359617682,
+      "grad_norm": 0.4633363803989168,
+      "learning_rate": 1.0396465990166829e-06,
+      "loss": 0.5267,
+      "step": 15815
+    },
+    {
+      "epoch": 4.198592858091066,
+      "grad_norm": 0.4777288993868714,
+      "learning_rate": 1.0393632429063575e-06,
+      "loss": 0.5735,
+      "step": 15816
+    },
+    {
+      "epoch": 4.1988583565644495,
+      "grad_norm": 0.47420527543503405,
+      "learning_rate": 1.039079915282025e-06,
+      "loss": 0.5141,
+      "step": 15817
+    },
+    {
+      "epoch": 4.199123855037834,
+      "grad_norm": 0.47716209695746187,
+      "learning_rate": 1.0387966161492111e-06,
+      "loss": 0.5505,
+      "step": 15818
+    },
+    {
+      "epoch": 4.199389353511218,
+      "grad_norm": 0.4755603780171045,
+      "learning_rate": 1.03851334551344e-06,
+      "loss": 0.5675,
+      "step": 15819
+    },
+    {
+      "epoch": 4.199654851984601,
+      "grad_norm": 0.47678007924957305,
+      "learning_rate": 1.0382301033802379e-06,
+      "loss": 0.5327,
+      "step": 15820
+    },
+    {
+      "epoch": 4.199920350457985,
+      "grad_norm": 0.48171762282732833,
+      "learning_rate": 1.0379468897551267e-06,
+      "loss": 0.5704,
+      "step": 15821
+    },
+    {
+      "epoch": 4.200185848931369,
+      "grad_norm": 0.490028259144393,
+      "learning_rate": 1.0376637046436323e-06,
+      "loss": 0.4882,
+      "step": 15822
+    },
+    {
+      "epoch": 4.200451347404752,
+      "grad_norm": 0.476688594985123,
+      "learning_rate": 1.0373805480512755e-06,
+      "loss": 0.559,
+      "step": 15823
+    },
+    {
+      "epoch": 4.200716845878136,
+      "grad_norm": 0.48776632700402484,
+      "learning_rate": 1.0370974199835785e-06,
+      "loss": 0.5645,
+      "step": 15824
+    },
+    {
+      "epoch": 4.20098234435152,
+      "grad_norm": 0.48072199869614013,
+      "learning_rate": 1.0368143204460643e-06,
+      "loss": 0.5433,
+      "step": 15825
+    },
+    {
+      "epoch": 4.201247842824904,
+      "grad_norm": 0.4720978912349653,
+      "learning_rate": 1.0365312494442528e-06,
+      "loss": 0.5315,
+      "step": 15826
+    },
+    {
+      "epoch": 4.201513341298288,
+      "grad_norm": 0.4503657636212197,
+      "learning_rate": 1.0362482069836658e-06,
+      "loss": 0.5061,
+      "step": 15827
+    },
+    {
+      "epoch": 4.201778839771672,
+      "grad_norm": 0.4818982649339037,
+      "learning_rate": 1.0359651930698217e-06,
+      "loss": 0.5542,
+      "step": 15828
+    },
+    {
+      "epoch": 4.202044338245055,
+      "grad_norm": 0.4762736540411629,
+      "learning_rate": 1.0356822077082428e-06,
+      "loss": 0.5412,
+      "step": 15829
+    },
+    {
+      "epoch": 4.202309836718439,
+      "grad_norm": 0.47059643406848206,
+      "learning_rate": 1.0353992509044445e-06,
+      "loss": 0.5229,
+      "step": 15830
+    },
+    {
+      "epoch": 4.202575335191822,
+      "grad_norm": 0.46810946096065087,
+      "learning_rate": 1.0351163226639474e-06,
+      "loss": 0.5461,
+      "step": 15831
+    },
+    {
+      "epoch": 4.202840833665206,
+      "grad_norm": 0.47334207261048367,
+      "learning_rate": 1.0348334229922677e-06,
+      "loss": 0.5565,
+      "step": 15832
+    },
+    {
+      "epoch": 4.20310633213859,
+      "grad_norm": 0.4577515337816793,
+      "learning_rate": 1.0345505518949247e-06,
+      "loss": 0.5332,
+      "step": 15833
+    },
+    {
+      "epoch": 4.203371830611974,
+      "grad_norm": 0.4644843714654492,
+      "learning_rate": 1.034267709377434e-06,
+      "loss": 0.5295,
+      "step": 15834
+    },
+    {
+      "epoch": 4.203637329085358,
+      "grad_norm": 0.4786982362174368,
+      "learning_rate": 1.0339848954453105e-06,
+      "loss": 0.538,
+      "step": 15835
+    },
+    {
+      "epoch": 4.203902827558742,
+      "grad_norm": 0.4913151811164922,
+      "learning_rate": 1.0337021101040724e-06,
+      "loss": 0.5287,
+      "step": 15836
+    },
+    {
+      "epoch": 4.204168326032125,
+      "grad_norm": 0.4954858732638189,
+      "learning_rate": 1.033419353359232e-06,
+      "loss": 0.5735,
+      "step": 15837
+    },
+    {
+      "epoch": 4.204433824505509,
+      "grad_norm": 0.4817342773849108,
+      "learning_rate": 1.0331366252163058e-06,
+      "loss": 0.5358,
+      "step": 15838
+    },
+    {
+      "epoch": 4.204699322978893,
+      "grad_norm": 0.47222194218079133,
+      "learning_rate": 1.0328539256808063e-06,
+      "loss": 0.5672,
+      "step": 15839
+    },
+    {
+      "epoch": 4.204964821452276,
+      "grad_norm": 0.4759953838767612,
+      "learning_rate": 1.032571254758248e-06,
+      "loss": 0.5564,
+      "step": 15840
+    },
+    {
+      "epoch": 4.20523031992566,
+      "grad_norm": 0.48462847967122563,
+      "learning_rate": 1.0322886124541433e-06,
+      "loss": 0.5759,
+      "step": 15841
+    },
+    {
+      "epoch": 4.2054958183990445,
+      "grad_norm": 0.47342175752153015,
+      "learning_rate": 1.0320059987740036e-06,
+      "loss": 0.5712,
+      "step": 15842
+    },
+    {
+      "epoch": 4.205761316872428,
+      "grad_norm": 0.48556260003642543,
+      "learning_rate": 1.0317234137233417e-06,
+      "loss": 0.5791,
+      "step": 15843
+    },
+    {
+      "epoch": 4.206026815345812,
+      "grad_norm": 0.4800324194060265,
+      "learning_rate": 1.0314408573076676e-06,
+      "loss": 0.5102,
+      "step": 15844
+    },
+    {
+      "epoch": 4.206292313819196,
+      "grad_norm": 0.47934191640752666,
+      "learning_rate": 1.031158329532493e-06,
+      "loss": 0.5344,
+      "step": 15845
+    },
+    {
+      "epoch": 4.206557812292579,
+      "grad_norm": 0.46728851077621186,
+      "learning_rate": 1.0308758304033276e-06,
+      "loss": 0.5172,
+      "step": 15846
+    },
+    {
+      "epoch": 4.206823310765963,
+      "grad_norm": 0.46783268879461065,
+      "learning_rate": 1.0305933599256805e-06,
+      "loss": 0.5473,
+      "step": 15847
+    },
+    {
+      "epoch": 4.207088809239347,
+      "grad_norm": 0.469213312191495,
+      "learning_rate": 1.0303109181050597e-06,
+      "loss": 0.5156,
+      "step": 15848
+    },
+    {
+      "epoch": 4.2073543077127304,
+      "grad_norm": 0.47580852028287096,
+      "learning_rate": 1.0300285049469752e-06,
+      "loss": 0.5303,
+      "step": 15849
+    },
+    {
+      "epoch": 4.2076198061861145,
+      "grad_norm": 0.4649832905144798,
+      "learning_rate": 1.029746120456933e-06,
+      "loss": 0.5637,
+      "step": 15850
+    },
+    {
+      "epoch": 4.207885304659499,
+      "grad_norm": 0.4639440059770558,
+      "learning_rate": 1.0294637646404424e-06,
+      "loss": 0.5388,
+      "step": 15851
+    },
+    {
+      "epoch": 4.208150803132882,
+      "grad_norm": 0.4743992927242743,
+      "learning_rate": 1.0291814375030087e-06,
+      "loss": 0.5252,
+      "step": 15852
+    },
+    {
+      "epoch": 4.208416301606266,
+      "grad_norm": 0.4753260657613669,
+      "learning_rate": 1.0288991390501373e-06,
+      "loss": 0.5135,
+      "step": 15853
+    },
+    {
+      "epoch": 4.208681800079649,
+      "grad_norm": 0.47742979299859034,
+      "learning_rate": 1.0286168692873355e-06,
+      "loss": 0.5805,
+      "step": 15854
+    },
+    {
+      "epoch": 4.208947298553033,
+      "grad_norm": 0.4778318080148841,
+      "learning_rate": 1.0283346282201064e-06,
+      "loss": 0.5714,
+      "step": 15855
+    },
+    {
+      "epoch": 4.209212797026417,
+      "grad_norm": 0.4827808863498797,
+      "learning_rate": 1.0280524158539562e-06,
+      "loss": 0.5412,
+      "step": 15856
+    },
+    {
+      "epoch": 4.2094782954998005,
+      "grad_norm": 0.46258722103823363,
+      "learning_rate": 1.027770232194388e-06,
+      "loss": 0.5117,
+      "step": 15857
+    },
+    {
+      "epoch": 4.2097437939731845,
+      "grad_norm": 0.47981473072896585,
+      "learning_rate": 1.0274880772469048e-06,
+      "loss": 0.5073,
+      "step": 15858
+    },
+    {
+      "epoch": 4.210009292446569,
+      "grad_norm": 0.4778023426602347,
+      "learning_rate": 1.0272059510170084e-06,
+      "loss": 0.5305,
+      "step": 15859
+    },
+    {
+      "epoch": 4.210274790919952,
+      "grad_norm": 0.47170898785160725,
+      "learning_rate": 1.026923853510203e-06,
+      "loss": 0.5163,
+      "step": 15860
+    },
+    {
+      "epoch": 4.210540289393336,
+      "grad_norm": 0.4723815154284675,
+      "learning_rate": 1.026641784731988e-06,
+      "loss": 0.5339,
+      "step": 15861
+    },
+    {
+      "epoch": 4.21080578786672,
+      "grad_norm": 0.4776143822266705,
+      "learning_rate": 1.0263597446878663e-06,
+      "loss": 0.5325,
+      "step": 15862
+    },
+    {
+      "epoch": 4.211071286340103,
+      "grad_norm": 0.48134166642271997,
+      "learning_rate": 1.0260777333833388e-06,
+      "loss": 0.5405,
+      "step": 15863
+    },
+    {
+      "epoch": 4.211336784813487,
+      "grad_norm": 0.4821551303003591,
+      "learning_rate": 1.0257957508239028e-06,
+      "loss": 0.5126,
+      "step": 15864
+    },
+    {
+      "epoch": 4.211602283286871,
+      "grad_norm": 0.4823589925371034,
+      "learning_rate": 1.0255137970150598e-06,
+      "loss": 0.5052,
+      "step": 15865
+    },
+    {
+      "epoch": 4.211867781760255,
+      "grad_norm": 0.47531023868110134,
+      "learning_rate": 1.0252318719623072e-06,
+      "loss": 0.5448,
+      "step": 15866
+    },
+    {
+      "epoch": 4.212133280233639,
+      "grad_norm": 0.47731256335508954,
+      "learning_rate": 1.0249499756711447e-06,
+      "loss": 0.5587,
+      "step": 15867
+    },
+    {
+      "epoch": 4.212398778707023,
+      "grad_norm": 0.47722080867640504,
+      "learning_rate": 1.0246681081470685e-06,
+      "loss": 0.5212,
+      "step": 15868
+    },
+    {
+      "epoch": 4.212664277180406,
+      "grad_norm": 0.47918871013826214,
+      "learning_rate": 1.0243862693955772e-06,
+      "loss": 0.5489,
+      "step": 15869
+    },
+    {
+      "epoch": 4.21292977565379,
+      "grad_norm": 0.4783584564069484,
+      "learning_rate": 1.0241044594221666e-06,
+      "loss": 0.5728,
+      "step": 15870
+    },
+    {
+      "epoch": 4.213195274127174,
+      "grad_norm": 0.4932118852095281,
+      "learning_rate": 1.0238226782323318e-06,
+      "loss": 0.5336,
+      "step": 15871
+    },
+    {
+      "epoch": 4.213460772600557,
+      "grad_norm": 0.4645038030882401,
+      "learning_rate": 1.02354092583157e-06,
+      "loss": 0.524,
+      "step": 15872
+    },
+    {
+      "epoch": 4.213726271073941,
+      "grad_norm": 0.4704331305131458,
+      "learning_rate": 1.0232592022253743e-06,
+      "loss": 0.542,
+      "step": 15873
+    },
+    {
+      "epoch": 4.2139917695473255,
+      "grad_norm": 0.4748354462780193,
+      "learning_rate": 1.0229775074192405e-06,
+      "loss": 0.5219,
+      "step": 15874
+    },
+    {
+      "epoch": 4.214257268020709,
+      "grad_norm": 0.4808335730089421,
+      "learning_rate": 1.022695841418662e-06,
+      "loss": 0.5567,
+      "step": 15875
+    },
+    {
+      "epoch": 4.214522766494093,
+      "grad_norm": 0.48397537639404015,
+      "learning_rate": 1.0224142042291313e-06,
+      "loss": 0.531,
+      "step": 15876
+    },
+    {
+      "epoch": 4.214788264967477,
+      "grad_norm": 0.4913913781668344,
+      "learning_rate": 1.0221325958561406e-06,
+      "loss": 0.5878,
+      "step": 15877
+    },
+    {
+      "epoch": 4.21505376344086,
+      "grad_norm": 0.4871073066019187,
+      "learning_rate": 1.0218510163051837e-06,
+      "loss": 0.5359,
+      "step": 15878
+    },
+    {
+      "epoch": 4.215319261914244,
+      "grad_norm": 0.46763669227248683,
+      "learning_rate": 1.02156946558175e-06,
+      "loss": 0.5034,
+      "step": 15879
+    },
+    {
+      "epoch": 4.215584760387628,
+      "grad_norm": 0.4879447462348556,
+      "learning_rate": 1.021287943691333e-06,
+      "loss": 0.5373,
+      "step": 15880
+    },
+    {
+      "epoch": 4.215850258861011,
+      "grad_norm": 0.47483084478082516,
+      "learning_rate": 1.021006450639421e-06,
+      "loss": 0.5077,
+      "step": 15881
+    },
+    {
+      "epoch": 4.2161157573343955,
+      "grad_norm": 0.493133603760487,
+      "learning_rate": 1.0207249864315038e-06,
+      "loss": 0.5385,
+      "step": 15882
+    },
+    {
+      "epoch": 4.216381255807779,
+      "grad_norm": 0.4748920336912919,
+      "learning_rate": 1.020443551073072e-06,
+      "loss": 0.5767,
+      "step": 15883
+    },
+    {
+      "epoch": 4.216646754281163,
+      "grad_norm": 0.4707026778562516,
+      "learning_rate": 1.0201621445696128e-06,
+      "loss": 0.5492,
+      "step": 15884
+    },
+    {
+      "epoch": 4.216912252754547,
+      "grad_norm": 0.48389348468213805,
+      "learning_rate": 1.0198807669266158e-06,
+      "loss": 0.544,
+      "step": 15885
+    },
+    {
+      "epoch": 4.21717775122793,
+      "grad_norm": 0.4750360799136636,
+      "learning_rate": 1.019599418149568e-06,
+      "loss": 0.5662,
+      "step": 15886
+    },
+    {
+      "epoch": 4.217443249701314,
+      "grad_norm": 0.4705706520969915,
+      "learning_rate": 1.0193180982439552e-06,
+      "loss": 0.5387,
+      "step": 15887
+    },
+    {
+      "epoch": 4.217708748174698,
+      "grad_norm": 0.47822659720091076,
+      "learning_rate": 1.0190368072152657e-06,
+      "loss": 0.5658,
+      "step": 15888
+    },
+    {
+      "epoch": 4.217974246648081,
+      "grad_norm": 0.4675187228909948,
+      "learning_rate": 1.0187555450689837e-06,
+      "loss": 0.5441,
+      "step": 15889
+    },
+    {
+      "epoch": 4.2182397451214655,
+      "grad_norm": 0.4697110696289429,
+      "learning_rate": 1.018474311810596e-06,
+      "loss": 0.5252,
+      "step": 15890
+    },
+    {
+      "epoch": 4.21850524359485,
+      "grad_norm": 0.48491602632538877,
+      "learning_rate": 1.0181931074455858e-06,
+      "loss": 0.5532,
+      "step": 15891
+    },
+    {
+      "epoch": 4.218770742068233,
+      "grad_norm": 0.4644217257885918,
+      "learning_rate": 1.0179119319794399e-06,
+      "loss": 0.5143,
+      "step": 15892
+    },
+    {
+      "epoch": 4.219036240541617,
+      "grad_norm": 0.4675934969381847,
+      "learning_rate": 1.0176307854176383e-06,
+      "loss": 0.5121,
+      "step": 15893
+    },
+    {
+      "epoch": 4.219301739015001,
+      "grad_norm": 0.4746763685315101,
+      "learning_rate": 1.0173496677656666e-06,
+      "loss": 0.5564,
+      "step": 15894
+    },
+    {
+      "epoch": 4.219567237488384,
+      "grad_norm": 0.47304827154072404,
+      "learning_rate": 1.0170685790290058e-06,
+      "loss": 0.5236,
+      "step": 15895
+    },
+    {
+      "epoch": 4.219832735961768,
+      "grad_norm": 0.4653343826433345,
+      "learning_rate": 1.0167875192131394e-06,
+      "loss": 0.5064,
+      "step": 15896
+    },
+    {
+      "epoch": 4.220098234435152,
+      "grad_norm": 0.4740909098740407,
+      "learning_rate": 1.016506488323547e-06,
+      "loss": 0.5213,
+      "step": 15897
+    },
+    {
+      "epoch": 4.2203637329085355,
+      "grad_norm": 0.47913438400532027,
+      "learning_rate": 1.0162254863657112e-06,
+      "loss": 0.5668,
+      "step": 15898
+    },
+    {
+      "epoch": 4.22062923138192,
+      "grad_norm": 0.48515452996387065,
+      "learning_rate": 1.0159445133451116e-06,
+      "loss": 0.5291,
+      "step": 15899
+    },
+    {
+      "epoch": 4.220894729855304,
+      "grad_norm": 0.4814378623759041,
+      "learning_rate": 1.0156635692672261e-06,
+      "loss": 0.5503,
+      "step": 15900
+    },
+    {
+      "epoch": 4.221160228328687,
+      "grad_norm": 0.4644732072739585,
+      "learning_rate": 1.0153826541375366e-06,
+      "loss": 0.5109,
+      "step": 15901
+    },
+    {
+      "epoch": 4.221425726802071,
+      "grad_norm": 0.4739078262498862,
+      "learning_rate": 1.0151017679615194e-06,
+      "loss": 0.5133,
+      "step": 15902
+    },
+    {
+      "epoch": 4.221691225275455,
+      "grad_norm": 0.4775277760577075,
+      "learning_rate": 1.014820910744654e-06,
+      "loss": 0.5482,
+      "step": 15903
+    },
+    {
+      "epoch": 4.221956723748838,
+      "grad_norm": 0.4771427804362491,
+      "learning_rate": 1.0145400824924176e-06,
+      "loss": 0.5437,
+      "step": 15904
+    },
+    {
+      "epoch": 4.222222222222222,
+      "grad_norm": 0.47618627437789246,
+      "learning_rate": 1.0142592832102863e-06,
+      "loss": 0.512,
+      "step": 15905
+    },
+    {
+      "epoch": 4.222487720695606,
+      "grad_norm": 0.47847148916228677,
+      "learning_rate": 1.013978512903736e-06,
+      "loss": 0.5426,
+      "step": 15906
+    },
+    {
+      "epoch": 4.22275321916899,
+      "grad_norm": 0.4852771486454056,
+      "learning_rate": 1.0136977715782431e-06,
+      "loss": 0.5464,
+      "step": 15907
+    },
+    {
+      "epoch": 4.223018717642374,
+      "grad_norm": 0.4692580065780568,
+      "learning_rate": 1.0134170592392837e-06,
+      "loss": 0.5396,
+      "step": 15908
+    },
+    {
+      "epoch": 4.223284216115758,
+      "grad_norm": 0.4688993432918405,
+      "learning_rate": 1.013136375892331e-06,
+      "loss": 0.5182,
+      "step": 15909
+    },
+    {
+      "epoch": 4.223549714589141,
+      "grad_norm": 0.4800633816061315,
+      "learning_rate": 1.0128557215428597e-06,
+      "loss": 0.5347,
+      "step": 15910
+    },
+    {
+      "epoch": 4.223815213062525,
+      "grad_norm": 0.4645576836122541,
+      "learning_rate": 1.0125750961963424e-06,
+      "loss": 0.5181,
+      "step": 15911
+    },
+    {
+      "epoch": 4.224080711535908,
+      "grad_norm": 0.4769913040956293,
+      "learning_rate": 1.0122944998582528e-06,
+      "loss": 0.5413,
+      "step": 15912
+    },
+    {
+      "epoch": 4.224346210009292,
+      "grad_norm": 0.4757370444146832,
+      "learning_rate": 1.0120139325340627e-06,
+      "loss": 0.5454,
+      "step": 15913
+    },
+    {
+      "epoch": 4.224611708482676,
+      "grad_norm": 0.47129215624269116,
+      "learning_rate": 1.0117333942292448e-06,
+      "loss": 0.5925,
+      "step": 15914
+    },
+    {
+      "epoch": 4.22487720695606,
+      "grad_norm": 0.4818781799857873,
+      "learning_rate": 1.0114528849492698e-06,
+      "loss": 0.5233,
+      "step": 15915
+    },
+    {
+      "epoch": 4.225142705429444,
+      "grad_norm": 0.48251590852443565,
+      "learning_rate": 1.0111724046996069e-06,
+      "loss": 0.5283,
+      "step": 15916
+    },
+    {
+      "epoch": 4.225408203902828,
+      "grad_norm": 0.4816599107045372,
+      "learning_rate": 1.0108919534857284e-06,
+      "loss": 0.5568,
+      "step": 15917
+    },
+    {
+      "epoch": 4.225673702376211,
+      "grad_norm": 0.5001858710165548,
+      "learning_rate": 1.010611531313102e-06,
+      "loss": 0.5933,
+      "step": 15918
+    },
+    {
+      "epoch": 4.225939200849595,
+      "grad_norm": 0.4863384905923578,
+      "learning_rate": 1.0103311381871983e-06,
+      "loss": 0.5811,
+      "step": 15919
+    },
+    {
+      "epoch": 4.226204699322979,
+      "grad_norm": 0.49967567532830104,
+      "learning_rate": 1.0100507741134836e-06,
+      "loss": 0.562,
+      "step": 15920
+    },
+    {
+      "epoch": 4.226470197796362,
+      "grad_norm": 0.48503391419084313,
+      "learning_rate": 1.009770439097429e-06,
+      "loss": 0.527,
+      "step": 15921
+    },
+    {
+      "epoch": 4.226735696269746,
+      "grad_norm": 0.47518587412633456,
+      "learning_rate": 1.0094901331444974e-06,
+      "loss": 0.5169,
+      "step": 15922
+    },
+    {
+      "epoch": 4.2270011947431305,
+      "grad_norm": 0.48262071059950545,
+      "learning_rate": 1.0092098562601585e-06,
+      "loss": 0.535,
+      "step": 15923
+    },
+    {
+      "epoch": 4.227266693216514,
+      "grad_norm": 0.47111494732321507,
+      "learning_rate": 1.0089296084498768e-06,
+      "loss": 0.5162,
+      "step": 15924
+    },
+    {
+      "epoch": 4.227532191689898,
+      "grad_norm": 0.4630830746118903,
+      "learning_rate": 1.0086493897191194e-06,
+      "loss": 0.5276,
+      "step": 15925
+    },
+    {
+      "epoch": 4.227797690163282,
+      "grad_norm": 0.4808478967248429,
+      "learning_rate": 1.0083692000733495e-06,
+      "loss": 0.5257,
+      "step": 15926
+    },
+    {
+      "epoch": 4.228063188636665,
+      "grad_norm": 0.46938887313113176,
+      "learning_rate": 1.008089039518033e-06,
+      "loss": 0.5148,
+      "step": 15927
+    },
+    {
+      "epoch": 4.228328687110049,
+      "grad_norm": 0.4708743551013019,
+      "learning_rate": 1.0078089080586331e-06,
+      "loss": 0.5197,
+      "step": 15928
+    },
+    {
+      "epoch": 4.228594185583433,
+      "grad_norm": 0.47646650794977047,
+      "learning_rate": 1.0075288057006122e-06,
+      "loss": 0.5469,
+      "step": 15929
+    },
+    {
+      "epoch": 4.2288596840568164,
+      "grad_norm": 0.4741249961849285,
+      "learning_rate": 1.0072487324494347e-06,
+      "loss": 0.5355,
+      "step": 15930
+    },
+    {
+      "epoch": 4.2291251825302005,
+      "grad_norm": 0.4804513957740347,
+      "learning_rate": 1.0069686883105608e-06,
+      "loss": 0.5556,
+      "step": 15931
+    },
+    {
+      "epoch": 4.229390681003585,
+      "grad_norm": 0.48234061686951074,
+      "learning_rate": 1.006688673289454e-06,
+      "loss": 0.5022,
+      "step": 15932
+    },
+    {
+      "epoch": 4.229656179476968,
+      "grad_norm": 0.47688139674142555,
+      "learning_rate": 1.0064086873915743e-06,
+      "loss": 0.5431,
+      "step": 15933
+    },
+    {
+      "epoch": 4.229921677950352,
+      "grad_norm": 0.47863507090959995,
+      "learning_rate": 1.0061287306223814e-06,
+      "loss": 0.5518,
+      "step": 15934
+    },
+    {
+      "epoch": 4.230187176423736,
+      "grad_norm": 0.4910866207723701,
+      "learning_rate": 1.0058488029873367e-06,
+      "loss": 0.5474,
+      "step": 15935
+    },
+    {
+      "epoch": 4.230452674897119,
+      "grad_norm": 0.474248423892993,
+      "learning_rate": 1.0055689044918979e-06,
+      "loss": 0.5491,
+      "step": 15936
+    },
+    {
+      "epoch": 4.230718173370503,
+      "grad_norm": 0.48336703789497126,
+      "learning_rate": 1.005289035141525e-06,
+      "loss": 0.502,
+      "step": 15937
+    },
+    {
+      "epoch": 4.230983671843887,
+      "grad_norm": 0.4838438719219541,
+      "learning_rate": 1.0050091949416758e-06,
+      "loss": 0.5465,
+      "step": 15938
+    },
+    {
+      "epoch": 4.2312491703172705,
+      "grad_norm": 0.48221486111858924,
+      "learning_rate": 1.0047293838978076e-06,
+      "loss": 0.5725,
+      "step": 15939
+    },
+    {
+      "epoch": 4.231514668790655,
+      "grad_norm": 0.4783826217859718,
+      "learning_rate": 1.0044496020153766e-06,
+      "loss": 0.5718,
+      "step": 15940
+    },
+    {
+      "epoch": 4.231780167264038,
+      "grad_norm": 0.48866707437572793,
+      "learning_rate": 1.004169849299841e-06,
+      "loss": 0.5707,
+      "step": 15941
+    },
+    {
+      "epoch": 4.232045665737422,
+      "grad_norm": 0.4886087644318049,
+      "learning_rate": 1.0038901257566549e-06,
+      "loss": 0.5392,
+      "step": 15942
+    },
+    {
+      "epoch": 4.232311164210806,
+      "grad_norm": 0.46473920359812265,
+      "learning_rate": 1.0036104313912753e-06,
+      "loss": 0.5145,
+      "step": 15943
+    },
+    {
+      "epoch": 4.232576662684189,
+      "grad_norm": 0.47626016489057527,
+      "learning_rate": 1.0033307662091564e-06,
+      "loss": 0.5312,
+      "step": 15944
+    },
+    {
+      "epoch": 4.232842161157573,
+      "grad_norm": 0.47641897969300506,
+      "learning_rate": 1.0030511302157507e-06,
+      "loss": 0.5343,
+      "step": 15945
+    },
+    {
+      "epoch": 4.233107659630957,
+      "grad_norm": 0.4804188349795335,
+      "learning_rate": 1.0027715234165141e-06,
+      "loss": 0.5571,
+      "step": 15946
+    },
+    {
+      "epoch": 4.233373158104341,
+      "grad_norm": 0.4809974956275609,
+      "learning_rate": 1.002491945816898e-06,
+      "loss": 0.5447,
+      "step": 15947
+    },
+    {
+      "epoch": 4.233638656577725,
+      "grad_norm": 0.48272081441051834,
+      "learning_rate": 1.0022123974223561e-06,
+      "loss": 0.5083,
+      "step": 15948
+    },
+    {
+      "epoch": 4.233904155051109,
+      "grad_norm": 0.4799969637914395,
+      "learning_rate": 1.001932878238339e-06,
+      "loss": 0.5793,
+      "step": 15949
+    },
+    {
+      "epoch": 4.234169653524492,
+      "grad_norm": 0.48157006782917455,
+      "learning_rate": 1.0016533882703003e-06,
+      "loss": 0.5167,
+      "step": 15950
+    },
+    {
+      "epoch": 4.234435151997876,
+      "grad_norm": 0.4745047614306491,
+      "learning_rate": 1.0013739275236873e-06,
+      "loss": 0.5536,
+      "step": 15951
+    },
+    {
+      "epoch": 4.23470065047126,
+      "grad_norm": 0.47684744179159017,
+      "learning_rate": 1.001094496003952e-06,
+      "loss": 0.5862,
+      "step": 15952
+    },
+    {
+      "epoch": 4.234966148944643,
+      "grad_norm": 0.4743926638992742,
+      "learning_rate": 1.000815093716545e-06,
+      "loss": 0.5328,
+      "step": 15953
+    },
+    {
+      "epoch": 4.235231647418027,
+      "grad_norm": 0.4726109594210413,
+      "learning_rate": 1.0005357206669133e-06,
+      "loss": 0.558,
+      "step": 15954
+    },
+    {
+      "epoch": 4.2354971458914115,
+      "grad_norm": 0.4884984905044093,
+      "learning_rate": 1.000256376860508e-06,
+      "loss": 0.5364,
+      "step": 15955
+    },
+    {
+      "epoch": 4.235762644364795,
+      "grad_norm": 0.4719704050174625,
+      "learning_rate": 9.99977062302774e-07,
+      "loss": 0.5442,
+      "step": 15956
+    },
+    {
+      "epoch": 4.236028142838179,
+      "grad_norm": 0.46870701889438227,
+      "learning_rate": 9.996977769991606e-07,
+      "loss": 0.5382,
+      "step": 15957
+    },
+    {
+      "epoch": 4.236293641311563,
+      "grad_norm": 0.46666157490804416,
+      "learning_rate": 9.99418520955113e-07,
+      "loss": 0.54,
+      "step": 15958
+    },
+    {
+      "epoch": 4.236559139784946,
+      "grad_norm": 0.46596040893020063,
+      "learning_rate": 9.991392941760794e-07,
+      "loss": 0.5205,
+      "step": 15959
+    },
+    {
+      "epoch": 4.23682463825833,
+      "grad_norm": 0.47799301608879863,
+      "learning_rate": 9.988600966675032e-07,
+      "loss": 0.5393,
+      "step": 15960
+    },
+    {
+      "epoch": 4.237090136731714,
+      "grad_norm": 0.46859255423847684,
+      "learning_rate": 9.985809284348317e-07,
+      "loss": 0.5061,
+      "step": 15961
+    },
+    {
+      "epoch": 4.237355635205097,
+      "grad_norm": 0.46207119164410576,
+      "learning_rate": 9.983017894835079e-07,
+      "loss": 0.54,
+      "step": 15962
+    },
+    {
+      "epoch": 4.2376211336784815,
+      "grad_norm": 0.47510397793530307,
+      "learning_rate": 9.980226798189752e-07,
+      "loss": 0.539,
+      "step": 15963
+    },
+    {
+      "epoch": 4.237886632151865,
+      "grad_norm": 0.49287419808481087,
+      "learning_rate": 9.977435994466787e-07,
+      "loss": 0.5415,
+      "step": 15964
+    },
+    {
+      "epoch": 4.238152130625249,
+      "grad_norm": 0.47409514978860545,
+      "learning_rate": 9.974645483720591e-07,
+      "loss": 0.5116,
+      "step": 15965
+    },
+    {
+      "epoch": 4.238417629098633,
+      "grad_norm": 0.47710044379054645,
+      "learning_rate": 9.971855266005605e-07,
+      "loss": 0.5231,
+      "step": 15966
+    },
+    {
+      "epoch": 4.238683127572016,
+      "grad_norm": 0.47780619371812816,
+      "learning_rate": 9.96906534137624e-07,
+      "loss": 0.5509,
+      "step": 15967
+    },
+    {
+      "epoch": 4.2389486260454,
+      "grad_norm": 0.4779700258715163,
+      "learning_rate": 9.966275709886899e-07,
+      "loss": 0.5296,
+      "step": 15968
+    },
+    {
+      "epoch": 4.239214124518784,
+      "grad_norm": 0.4875739553432791,
+      "learning_rate": 9.963486371591984e-07,
+      "loss": 0.5316,
+      "step": 15969
+    },
+    {
+      "epoch": 4.239479622992167,
+      "grad_norm": 0.47674675590545296,
+      "learning_rate": 9.960697326545909e-07,
+      "loss": 0.5646,
+      "step": 15970
+    },
+    {
+      "epoch": 4.2397451214655515,
+      "grad_norm": 0.47229727053842213,
+      "learning_rate": 9.95790857480305e-07,
+      "loss": 0.517,
+      "step": 15971
+    },
+    {
+      "epoch": 4.240010619938936,
+      "grad_norm": 0.47586131703811907,
+      "learning_rate": 9.95512011641781e-07,
+      "loss": 0.5261,
+      "step": 15972
+    },
+    {
+      "epoch": 4.240276118412319,
+      "grad_norm": 0.47888792352102183,
+      "learning_rate": 9.952331951444566e-07,
+      "loss": 0.5759,
+      "step": 15973
+    },
+    {
+      "epoch": 4.240541616885703,
+      "grad_norm": 0.46001179587519053,
+      "learning_rate": 9.949544079937684e-07,
+      "loss": 0.5085,
+      "step": 15974
+    },
+    {
+      "epoch": 4.240807115359087,
+      "grad_norm": 0.4757442630119836,
+      "learning_rate": 9.946756501951549e-07,
+      "loss": 0.5233,
+      "step": 15975
+    },
+    {
+      "epoch": 4.24107261383247,
+      "grad_norm": 0.4845871760270505,
+      "learning_rate": 9.94396921754051e-07,
+      "loss": 0.5206,
+      "step": 15976
+    },
+    {
+      "epoch": 4.241338112305854,
+      "grad_norm": 0.47762528619974337,
+      "learning_rate": 9.941182226758948e-07,
+      "loss": 0.5697,
+      "step": 15977
+    },
+    {
+      "epoch": 4.241603610779238,
+      "grad_norm": 0.4761888890384406,
+      "learning_rate": 9.938395529661188e-07,
+      "loss": 0.5209,
+      "step": 15978
+    },
+    {
+      "epoch": 4.2418691092526215,
+      "grad_norm": 0.4635347399061296,
+      "learning_rate": 9.935609126301603e-07,
+      "loss": 0.5024,
+      "step": 15979
+    },
+    {
+      "epoch": 4.242134607726006,
+      "grad_norm": 0.4744707139161358,
+      "learning_rate": 9.932823016734528e-07,
+      "loss": 0.5531,
+      "step": 15980
+    },
+    {
+      "epoch": 4.24240010619939,
+      "grad_norm": 0.48150205727966855,
+      "learning_rate": 9.930037201014282e-07,
+      "loss": 0.5269,
+      "step": 15981
+    },
+    {
+      "epoch": 4.242665604672773,
+      "grad_norm": 0.482923144486339,
+      "learning_rate": 9.927251679195218e-07,
+      "loss": 0.5324,
+      "step": 15982
+    },
+    {
+      "epoch": 4.242931103146157,
+      "grad_norm": 0.47463125858726435,
+      "learning_rate": 9.92446645133164e-07,
+      "loss": 0.5342,
+      "step": 15983
+    },
+    {
+      "epoch": 4.243196601619541,
+      "grad_norm": 0.476528507446879,
+      "learning_rate": 9.9216815174779e-07,
+      "loss": 0.5265,
+      "step": 15984
+    },
+    {
+      "epoch": 4.243462100092924,
+      "grad_norm": 0.4830035630828717,
+      "learning_rate": 9.918896877688266e-07,
+      "loss": 0.5248,
+      "step": 15985
+    },
+    {
+      "epoch": 4.243727598566308,
+      "grad_norm": 0.4869457089488714,
+      "learning_rate": 9.91611253201708e-07,
+      "loss": 0.5499,
+      "step": 15986
+    },
+    {
+      "epoch": 4.243993097039692,
+      "grad_norm": 0.47789855810616927,
+      "learning_rate": 9.91332848051862e-07,
+      "loss": 0.5478,
+      "step": 15987
+    },
+    {
+      "epoch": 4.244258595513076,
+      "grad_norm": 0.48049778461741777,
+      "learning_rate": 9.910544723247204e-07,
+      "loss": 0.5003,
+      "step": 15988
+    },
+    {
+      "epoch": 4.24452409398646,
+      "grad_norm": 0.48802707897228115,
+      "learning_rate": 9.907761260257101e-07,
+      "loss": 0.5394,
+      "step": 15989
+    },
+    {
+      "epoch": 4.244789592459844,
+      "grad_norm": 0.4784880012741722,
+      "learning_rate": 9.904978091602615e-07,
+      "loss": 0.5395,
+      "step": 15990
+    },
+    {
+      "epoch": 4.245055090933227,
+      "grad_norm": 0.46522718711992217,
+      "learning_rate": 9.902195217338014e-07,
+      "loss": 0.5172,
+      "step": 15991
+    },
+    {
+      "epoch": 4.245320589406611,
+      "grad_norm": 0.4783737405844053,
+      "learning_rate": 9.899412637517567e-07,
+      "loss": 0.561,
+      "step": 15992
+    },
+    {
+      "epoch": 4.245586087879994,
+      "grad_norm": 0.47611464593511765,
+      "learning_rate": 9.896630352195553e-07,
+      "loss": 0.5338,
+      "step": 15993
+    },
+    {
+      "epoch": 4.245851586353378,
+      "grad_norm": 0.47763745446218864,
+      "learning_rate": 9.893848361426213e-07,
+      "loss": 0.5112,
+      "step": 15994
+    },
+    {
+      "epoch": 4.246117084826762,
+      "grad_norm": 0.4690369275846122,
+      "learning_rate": 9.891066665263831e-07,
+      "loss": 0.5231,
+      "step": 15995
+    },
+    {
+      "epoch": 4.246382583300146,
+      "grad_norm": 0.47004676018678654,
+      "learning_rate": 9.88828526376264e-07,
+      "loss": 0.527,
+      "step": 15996
+    },
+    {
+      "epoch": 4.24664808177353,
+      "grad_norm": 0.47394702047689996,
+      "learning_rate": 9.88550415697689e-07,
+      "loss": 0.5603,
+      "step": 15997
+    },
+    {
+      "epoch": 4.246913580246914,
+      "grad_norm": 0.4695180362922194,
+      "learning_rate": 9.8827233449608e-07,
+      "loss": 0.5024,
+      "step": 15998
+    },
+    {
+      "epoch": 4.247179078720297,
+      "grad_norm": 0.48834802311827824,
+      "learning_rate": 9.879942827768621e-07,
+      "loss": 0.4921,
+      "step": 15999
+    },
+    {
+      "epoch": 4.247444577193681,
+      "grad_norm": 0.49290687926842247,
+      "learning_rate": 9.877162605454587e-07,
+      "loss": 0.5302,
+      "step": 16000
+    },
+    {
+      "epoch": 4.247710075667065,
+      "grad_norm": 0.4769398085062818,
+      "learning_rate": 9.8743826780729e-07,
+      "loss": 0.5455,
+      "step": 16001
+    },
+    {
+      "epoch": 4.247975574140448,
+      "grad_norm": 0.49579909742191913,
+      "learning_rate": 9.871603045677802e-07,
+      "loss": 0.5499,
+      "step": 16002
+    },
+    {
+      "epoch": 4.248241072613832,
+      "grad_norm": 0.47489923116737565,
+      "learning_rate": 9.86882370832347e-07,
+      "loss": 0.5246,
+      "step": 16003
+    },
+    {
+      "epoch": 4.2485065710872165,
+      "grad_norm": 0.46381284918068677,
+      "learning_rate": 9.866044666064128e-07,
+      "loss": 0.5444,
+      "step": 16004
+    },
+    {
+      "epoch": 4.2487720695606,
+      "grad_norm": 0.4795434403804182,
+      "learning_rate": 9.863265918953962e-07,
+      "loss": 0.5286,
+      "step": 16005
+    },
+    {
+      "epoch": 4.249037568033984,
+      "grad_norm": 0.48327450098696323,
+      "learning_rate": 9.86048746704718e-07,
+      "loss": 0.5807,
+      "step": 16006
+    },
+    {
+      "epoch": 4.249303066507368,
+      "grad_norm": 0.47626729010875246,
+      "learning_rate": 9.857709310397962e-07,
+      "loss": 0.5285,
+      "step": 16007
+    },
+    {
+      "epoch": 4.249568564980751,
+      "grad_norm": 0.4843302235506611,
+      "learning_rate": 9.854931449060479e-07,
+      "loss": 0.5509,
+      "step": 16008
+    },
+    {
+      "epoch": 4.249834063454135,
+      "grad_norm": 0.48323191514862807,
+      "learning_rate": 9.852153883088922e-07,
+      "loss": 0.5635,
+      "step": 16009
+    },
+    {
+      "epoch": 4.250099561927519,
+      "grad_norm": 0.46662611076328203,
+      "learning_rate": 9.849376612537442e-07,
+      "loss": 0.5314,
+      "step": 16010
+    },
+    {
+      "epoch": 4.2503650604009025,
+      "grad_norm": 0.47660802677753505,
+      "learning_rate": 9.846599637460224e-07,
+      "loss": 0.5433,
+      "step": 16011
+    },
+    {
+      "epoch": 4.2506305588742865,
+      "grad_norm": 0.4935621095105905,
+      "learning_rate": 9.843822957911405e-07,
+      "loss": 0.5442,
+      "step": 16012
+    },
+    {
+      "epoch": 4.250896057347671,
+      "grad_norm": 0.48849525065551946,
+      "learning_rate": 9.841046573945165e-07,
+      "loss": 0.5341,
+      "step": 16013
+    },
+    {
+      "epoch": 4.251161555821054,
+      "grad_norm": 0.4777556110114682,
+      "learning_rate": 9.838270485615613e-07,
+      "loss": 0.5513,
+      "step": 16014
+    },
+    {
+      "epoch": 4.251427054294438,
+      "grad_norm": 0.4736586790615974,
+      "learning_rate": 9.83549469297692e-07,
+      "loss": 0.5341,
+      "step": 16015
+    },
+    {
+      "epoch": 4.251692552767822,
+      "grad_norm": 0.4783070597224486,
+      "learning_rate": 9.8327191960832e-07,
+      "loss": 0.5588,
+      "step": 16016
+    },
+    {
+      "epoch": 4.251958051241205,
+      "grad_norm": 0.4835368521055435,
+      "learning_rate": 9.829943994988603e-07,
+      "loss": 0.5481,
+      "step": 16017
+    },
+    {
+      "epoch": 4.252223549714589,
+      "grad_norm": 0.4870136685130338,
+      "learning_rate": 9.82716908974723e-07,
+      "loss": 0.5815,
+      "step": 16018
+    },
+    {
+      "epoch": 4.252489048187973,
+      "grad_norm": 0.46232329302794406,
+      "learning_rate": 9.82439448041322e-07,
+      "loss": 0.5575,
+      "step": 16019
+    },
+    {
+      "epoch": 4.2527545466613565,
+      "grad_norm": 0.46938662065875303,
+      "learning_rate": 9.821620167040669e-07,
+      "loss": 0.5462,
+      "step": 16020
+    },
+    {
+      "epoch": 4.253020045134741,
+      "grad_norm": 0.48660290109037174,
+      "learning_rate": 9.818846149683683e-07,
+      "loss": 0.5134,
+      "step": 16021
+    },
+    {
+      "epoch": 4.253285543608124,
+      "grad_norm": 0.47521846723545547,
+      "learning_rate": 9.816072428396376e-07,
+      "loss": 0.5336,
+      "step": 16022
+    },
+    {
+      "epoch": 4.253551042081508,
+      "grad_norm": 0.47149676881478686,
+      "learning_rate": 9.81329900323282e-07,
+      "loss": 0.5338,
+      "step": 16023
+    },
+    {
+      "epoch": 4.253816540554892,
+      "grad_norm": 0.4763948392473039,
+      "learning_rate": 9.810525874247128e-07,
+      "loss": 0.5734,
+      "step": 16024
+    },
+    {
+      "epoch": 4.254082039028275,
+      "grad_norm": 0.472106145233036,
+      "learning_rate": 9.807753041493374e-07,
+      "loss": 0.5572,
+      "step": 16025
+    },
+    {
+      "epoch": 4.254347537501659,
+      "grad_norm": 0.48380889344319267,
+      "learning_rate": 9.804980505025621e-07,
+      "loss": 0.5487,
+      "step": 16026
+    },
+    {
+      "epoch": 4.254613035975043,
+      "grad_norm": 0.4725136356423027,
+      "learning_rate": 9.802208264897965e-07,
+      "loss": 0.5697,
+      "step": 16027
+    },
+    {
+      "epoch": 4.254878534448427,
+      "grad_norm": 0.4738780429007621,
+      "learning_rate": 9.799436321164446e-07,
+      "loss": 0.5272,
+      "step": 16028
+    },
+    {
+      "epoch": 4.255144032921811,
+      "grad_norm": 0.4728705996349,
+      "learning_rate": 9.796664673879143e-07,
+      "loss": 0.524,
+      "step": 16029
+    },
+    {
+      "epoch": 4.255409531395195,
+      "grad_norm": 0.4842737139152287,
+      "learning_rate": 9.793893323096106e-07,
+      "loss": 0.5855,
+      "step": 16030
+    },
+    {
+      "epoch": 4.255675029868578,
+      "grad_norm": 0.4880141373677577,
+      "learning_rate": 9.791122268869383e-07,
+      "loss": 0.5411,
+      "step": 16031
+    },
+    {
+      "epoch": 4.255940528341962,
+      "grad_norm": 0.4805442288193504,
+      "learning_rate": 9.788351511252999e-07,
+      "loss": 0.5225,
+      "step": 16032
+    },
+    {
+      "epoch": 4.256206026815346,
+      "grad_norm": 0.4814089510712732,
+      "learning_rate": 9.785581050301019e-07,
+      "loss": 0.5534,
+      "step": 16033
+    },
+    {
+      "epoch": 4.256471525288729,
+      "grad_norm": 0.4794097831717837,
+      "learning_rate": 9.782810886067448e-07,
+      "loss": 0.569,
+      "step": 16034
+    },
+    {
+      "epoch": 4.256737023762113,
+      "grad_norm": 0.482781456838931,
+      "learning_rate": 9.78004101860633e-07,
+      "loss": 0.5611,
+      "step": 16035
+    },
+    {
+      "epoch": 4.2570025222354975,
+      "grad_norm": 0.4715717009632437,
+      "learning_rate": 9.77727144797168e-07,
+      "loss": 0.5392,
+      "step": 16036
+    },
+    {
+      "epoch": 4.257268020708881,
+      "grad_norm": 0.49242737961084665,
+      "learning_rate": 9.7745021742175e-07,
+      "loss": 0.5435,
+      "step": 16037
+    },
+    {
+      "epoch": 4.257533519182265,
+      "grad_norm": 0.4857115041572971,
+      "learning_rate": 9.771733197397815e-07,
+      "loss": 0.5821,
+      "step": 16038
+    },
+    {
+      "epoch": 4.257799017655649,
+      "grad_norm": 0.47833497809488323,
+      "learning_rate": 9.768964517566607e-07,
+      "loss": 0.5398,
+      "step": 16039
+    },
+    {
+      "epoch": 4.258064516129032,
+      "grad_norm": 0.48023255944996,
+      "learning_rate": 9.766196134777893e-07,
+      "loss": 0.5264,
+      "step": 16040
+    },
+    {
+      "epoch": 4.258330014602416,
+      "grad_norm": 0.4724451195861857,
+      "learning_rate": 9.763428049085643e-07,
+      "loss": 0.5303,
+      "step": 16041
+    },
+    {
+      "epoch": 4.2585955130758,
+      "grad_norm": 0.47375952723417264,
+      "learning_rate": 9.76066026054387e-07,
+      "loss": 0.5233,
+      "step": 16042
+    },
+    {
+      "epoch": 4.258861011549183,
+      "grad_norm": 0.4686268518570557,
+      "learning_rate": 9.757892769206517e-07,
+      "loss": 0.5441,
+      "step": 16043
+    },
+    {
+      "epoch": 4.2591265100225675,
+      "grad_norm": 0.4880967754837364,
+      "learning_rate": 9.755125575127575e-07,
+      "loss": 0.5569,
+      "step": 16044
+    },
+    {
+      "epoch": 4.259392008495951,
+      "grad_norm": 0.48539376577705884,
+      "learning_rate": 9.752358678361015e-07,
+      "loss": 0.5407,
+      "step": 16045
+    },
+    {
+      "epoch": 4.259657506969335,
+      "grad_norm": 0.48876764730175115,
+      "learning_rate": 9.74959207896079e-07,
+      "loss": 0.5375,
+      "step": 16046
+    },
+    {
+      "epoch": 4.259923005442719,
+      "grad_norm": 0.48391938008695684,
+      "learning_rate": 9.746825776980864e-07,
+      "loss": 0.5329,
+      "step": 16047
+    },
+    {
+      "epoch": 4.260188503916103,
+      "grad_norm": 0.4970687844112541,
+      "learning_rate": 9.744059772475182e-07,
+      "loss": 0.5384,
+      "step": 16048
+    },
+    {
+      "epoch": 4.260454002389486,
+      "grad_norm": 0.4779618739741265,
+      "learning_rate": 9.74129406549769e-07,
+      "loss": 0.5572,
+      "step": 16049
+    },
+    {
+      "epoch": 4.26071950086287,
+      "grad_norm": 0.48276551268483087,
+      "learning_rate": 9.73852865610231e-07,
+      "loss": 0.5465,
+      "step": 16050
+    },
+    {
+      "epoch": 4.260984999336253,
+      "grad_norm": 0.46703669058627545,
+      "learning_rate": 9.735763544343e-07,
+      "loss": 0.514,
+      "step": 16051
+    },
+    {
+      "epoch": 4.2612504978096375,
+      "grad_norm": 0.488214021625802,
+      "learning_rate": 9.732998730273665e-07,
+      "loss": 0.5528,
+      "step": 16052
+    },
+    {
+      "epoch": 4.261515996283022,
+      "grad_norm": 0.4751823210750789,
+      "learning_rate": 9.730234213948245e-07,
+      "loss": 0.5636,
+      "step": 16053
+    },
+    {
+      "epoch": 4.261781494756405,
+      "grad_norm": 0.4834320507378837,
+      "learning_rate": 9.727469995420642e-07,
+      "loss": 0.5209,
+      "step": 16054
+    },
+    {
+      "epoch": 4.262046993229789,
+      "grad_norm": 0.48519488910479647,
+      "learning_rate": 9.72470607474476e-07,
+      "loss": 0.5236,
+      "step": 16055
+    },
+    {
+      "epoch": 4.262312491703173,
+      "grad_norm": 0.48297816464771204,
+      "learning_rate": 9.721942451974516e-07,
+      "loss": 0.5618,
+      "step": 16056
+    },
+    {
+      "epoch": 4.262577990176556,
+      "grad_norm": 0.48494023212641846,
+      "learning_rate": 9.719179127163797e-07,
+      "loss": 0.5463,
+      "step": 16057
+    },
+    {
+      "epoch": 4.26284348864994,
+      "grad_norm": 0.4561690057664008,
+      "learning_rate": 9.716416100366504e-07,
+      "loss": 0.4978,
+      "step": 16058
+    },
+    {
+      "epoch": 4.263108987123324,
+      "grad_norm": 0.478148824142442,
+      "learning_rate": 9.713653371636521e-07,
+      "loss": 0.5297,
+      "step": 16059
+    },
+    {
+      "epoch": 4.2633744855967075,
+      "grad_norm": 0.493625423472556,
+      "learning_rate": 9.710890941027724e-07,
+      "loss": 0.564,
+      "step": 16060
+    },
+    {
+      "epoch": 4.263639984070092,
+      "grad_norm": 0.4789588176883131,
+      "learning_rate": 9.708128808593977e-07,
+      "loss": 0.55,
+      "step": 16061
+    },
+    {
+      "epoch": 4.263905482543476,
+      "grad_norm": 0.4743936301301481,
+      "learning_rate": 9.70536697438917e-07,
+      "loss": 0.5339,
+      "step": 16062
+    },
+    {
+      "epoch": 4.264170981016859,
+      "grad_norm": 0.4759998978829584,
+      "learning_rate": 9.702605438467144e-07,
+      "loss": 0.5452,
+      "step": 16063
+    },
+    {
+      "epoch": 4.264436479490243,
+      "grad_norm": 0.4792951823034538,
+      "learning_rate": 9.69984420088177e-07,
+      "loss": 0.5686,
+      "step": 16064
+    },
+    {
+      "epoch": 4.264701977963627,
+      "grad_norm": 0.48056652033675523,
+      "learning_rate": 9.69708326168691e-07,
+      "loss": 0.5648,
+      "step": 16065
+    },
+    {
+      "epoch": 4.26496747643701,
+      "grad_norm": 0.46039414951950647,
+      "learning_rate": 9.694322620936377e-07,
+      "loss": 0.5079,
+      "step": 16066
+    },
+    {
+      "epoch": 4.265232974910394,
+      "grad_norm": 0.4840736501554974,
+      "learning_rate": 9.691562278684035e-07,
+      "loss": 0.5642,
+      "step": 16067
+    },
+    {
+      "epoch": 4.265498473383778,
+      "grad_norm": 0.4819643709749258,
+      "learning_rate": 9.688802234983706e-07,
+      "loss": 0.5777,
+      "step": 16068
+    },
+    {
+      "epoch": 4.265763971857162,
+      "grad_norm": 0.4860825727021681,
+      "learning_rate": 9.686042489889225e-07,
+      "loss": 0.5386,
+      "step": 16069
+    },
+    {
+      "epoch": 4.266029470330546,
+      "grad_norm": 0.4744512794430315,
+      "learning_rate": 9.683283043454407e-07,
+      "loss": 0.5569,
+      "step": 16070
+    },
+    {
+      "epoch": 4.26629496880393,
+      "grad_norm": 0.4885789344864933,
+      "learning_rate": 9.680523895733078e-07,
+      "loss": 0.546,
+      "step": 16071
+    },
+    {
+      "epoch": 4.266560467277313,
+      "grad_norm": 0.48135840635621785,
+      "learning_rate": 9.67776504677904e-07,
+      "loss": 0.5148,
+      "step": 16072
+    },
+    {
+      "epoch": 4.266825965750697,
+      "grad_norm": 0.4920760143132962,
+      "learning_rate": 9.67500649664609e-07,
+      "loss": 0.5483,
+      "step": 16073
+    },
+    {
+      "epoch": 4.26709146422408,
+      "grad_norm": 0.48226587383837854,
+      "learning_rate": 9.672248245388043e-07,
+      "loss": 0.5712,
+      "step": 16074
+    },
+    {
+      "epoch": 4.267356962697464,
+      "grad_norm": 0.48995922230162264,
+      "learning_rate": 9.669490293058681e-07,
+      "loss": 0.538,
+      "step": 16075
+    },
+    {
+      "epoch": 4.267622461170848,
+      "grad_norm": 0.5042151762882503,
+      "learning_rate": 9.666732639711808e-07,
+      "loss": 0.5459,
+      "step": 16076
+    },
+    {
+      "epoch": 4.2678879596442325,
+      "grad_norm": 0.4655224328954768,
+      "learning_rate": 9.663975285401172e-07,
+      "loss": 0.5518,
+      "step": 16077
+    },
+    {
+      "epoch": 4.268153458117616,
+      "grad_norm": 0.4664643310146895,
+      "learning_rate": 9.661218230180577e-07,
+      "loss": 0.5099,
+      "step": 16078
+    },
+    {
+      "epoch": 4.268418956591,
+      "grad_norm": 0.4913369467815235,
+      "learning_rate": 9.658461474103772e-07,
+      "loss": 0.5336,
+      "step": 16079
+    },
+    {
+      "epoch": 4.268684455064383,
+      "grad_norm": 0.4894777850480446,
+      "learning_rate": 9.655705017224543e-07,
+      "loss": 0.5361,
+      "step": 16080
+    },
+    {
+      "epoch": 4.268949953537767,
+      "grad_norm": 0.4619893316056146,
+      "learning_rate": 9.652948859596626e-07,
+      "loss": 0.5077,
+      "step": 16081
+    },
+    {
+      "epoch": 4.269215452011151,
+      "grad_norm": 0.4850913789551262,
+      "learning_rate": 9.65019300127379e-07,
+      "loss": 0.5317,
+      "step": 16082
+    },
+    {
+      "epoch": 4.269480950484534,
+      "grad_norm": 0.4764435888125086,
+      "learning_rate": 9.647437442309771e-07,
+      "loss": 0.5237,
+      "step": 16083
+    },
+    {
+      "epoch": 4.269746448957918,
+      "grad_norm": 0.48223306966055324,
+      "learning_rate": 9.644682182758305e-07,
+      "loss": 0.5139,
+      "step": 16084
+    },
+    {
+      "epoch": 4.2700119474313025,
+      "grad_norm": 0.473251847050034,
+      "learning_rate": 9.641927222673142e-07,
+      "loss": 0.5125,
+      "step": 16085
+    },
+    {
+      "epoch": 4.270277445904686,
+      "grad_norm": 0.46133355324013003,
+      "learning_rate": 9.63917256210799e-07,
+      "loss": 0.5168,
+      "step": 16086
+    },
+    {
+      "epoch": 4.27054294437807,
+      "grad_norm": 0.47156208699791874,
+      "learning_rate": 9.636418201116595e-07,
+      "loss": 0.4969,
+      "step": 16087
+    },
+    {
+      "epoch": 4.270808442851454,
+      "grad_norm": 0.4847854205290721,
+      "learning_rate": 9.63366413975266e-07,
+      "loss": 0.5243,
+      "step": 16088
+    },
+    {
+      "epoch": 4.271073941324837,
+      "grad_norm": 0.47570561452831767,
+      "learning_rate": 9.630910378069888e-07,
+      "loss": 0.5529,
+      "step": 16089
+    },
+    {
+      "epoch": 4.271339439798221,
+      "grad_norm": 0.4757927710267138,
+      "learning_rate": 9.628156916122002e-07,
+      "loss": 0.5448,
+      "step": 16090
+    },
+    {
+      "epoch": 4.271604938271605,
+      "grad_norm": 0.4858879579251946,
+      "learning_rate": 9.625403753962686e-07,
+      "loss": 0.545,
+      "step": 16091
+    },
+    {
+      "epoch": 4.2718704367449885,
+      "grad_norm": 0.4777561412323649,
+      "learning_rate": 9.62265089164565e-07,
+      "loss": 0.542,
+      "step": 16092
+    },
+    {
+      "epoch": 4.2721359352183725,
+      "grad_norm": 0.47979147112681003,
+      "learning_rate": 9.619898329224562e-07,
+      "loss": 0.5824,
+      "step": 16093
+    },
+    {
+      "epoch": 4.272401433691757,
+      "grad_norm": 0.48484741406015935,
+      "learning_rate": 9.61714606675313e-07,
+      "loss": 0.5017,
+      "step": 16094
+    },
+    {
+      "epoch": 4.27266693216514,
+      "grad_norm": 0.4818843720111582,
+      "learning_rate": 9.614394104284996e-07,
+      "loss": 0.5388,
+      "step": 16095
+    },
+    {
+      "epoch": 4.272932430638524,
+      "grad_norm": 0.4742567143723967,
+      "learning_rate": 9.611642441873856e-07,
+      "loss": 0.573,
+      "step": 16096
+    },
+    {
+      "epoch": 4.273197929111908,
+      "grad_norm": 0.4929283846730227,
+      "learning_rate": 9.608891079573357e-07,
+      "loss": 0.5519,
+      "step": 16097
+    },
+    {
+      "epoch": 4.273463427585291,
+      "grad_norm": 0.4824255007576174,
+      "learning_rate": 9.606140017437176e-07,
+      "loss": 0.5496,
+      "step": 16098
+    },
+    {
+      "epoch": 4.273728926058675,
+      "grad_norm": 0.4706925454237134,
+      "learning_rate": 9.603389255518947e-07,
+      "loss": 0.5233,
+      "step": 16099
+    },
+    {
+      "epoch": 4.273994424532059,
+      "grad_norm": 0.4788694614300321,
+      "learning_rate": 9.60063879387233e-07,
+      "loss": 0.5714,
+      "step": 16100
+    },
+    {
+      "epoch": 4.2742599230054426,
+      "grad_norm": 0.46395638356149377,
+      "learning_rate": 9.597888632550963e-07,
+      "loss": 0.5447,
+      "step": 16101
+    },
+    {
+      "epoch": 4.274525421478827,
+      "grad_norm": 0.4649147910818229,
+      "learning_rate": 9.59513877160847e-07,
+      "loss": 0.4857,
+      "step": 16102
+    },
+    {
+      "epoch": 4.27479091995221,
+      "grad_norm": 0.4752046417424867,
+      "learning_rate": 9.5923892110985e-07,
+      "loss": 0.5016,
+      "step": 16103
+    },
+    {
+      "epoch": 4.275056418425594,
+      "grad_norm": 0.4816590262517745,
+      "learning_rate": 9.58963995107465e-07,
+      "loss": 0.5823,
+      "step": 16104
+    },
+    {
+      "epoch": 4.275321916898978,
+      "grad_norm": 0.48660476014912474,
+      "learning_rate": 9.586890991590575e-07,
+      "loss": 0.5667,
+      "step": 16105
+    },
+    {
+      "epoch": 4.275587415372361,
+      "grad_norm": 0.47604121262834925,
+      "learning_rate": 9.584142332699842e-07,
+      "loss": 0.4962,
+      "step": 16106
+    },
+    {
+      "epoch": 4.275852913845745,
+      "grad_norm": 0.4849036248325664,
+      "learning_rate": 9.581393974456089e-07,
+      "loss": 0.5424,
+      "step": 16107
+    },
+    {
+      "epoch": 4.276118412319129,
+      "grad_norm": 0.4847917749237604,
+      "learning_rate": 9.578645916912896e-07,
+      "loss": 0.5605,
+      "step": 16108
+    },
+    {
+      "epoch": 4.276383910792513,
+      "grad_norm": 0.4755094529760263,
+      "learning_rate": 9.575898160123873e-07,
+      "loss": 0.5151,
+      "step": 16109
+    },
+    {
+      "epoch": 4.276649409265897,
+      "grad_norm": 0.47892046251268966,
+      "learning_rate": 9.573150704142592e-07,
+      "loss": 0.54,
+      "step": 16110
+    },
+    {
+      "epoch": 4.276914907739281,
+      "grad_norm": 0.4797209134977321,
+      "learning_rate": 9.570403549022652e-07,
+      "loss": 0.5764,
+      "step": 16111
+    },
+    {
+      "epoch": 4.277180406212664,
+      "grad_norm": 0.4623075018790371,
+      "learning_rate": 9.567656694817622e-07,
+      "loss": 0.5255,
+      "step": 16112
+    },
+    {
+      "epoch": 4.277445904686048,
+      "grad_norm": 0.48044970077507604,
+      "learning_rate": 9.564910141581062e-07,
+      "loss": 0.5571,
+      "step": 16113
+    },
+    {
+      "epoch": 4.277711403159432,
+      "grad_norm": 0.4665407751577506,
+      "learning_rate": 9.562163889366558e-07,
+      "loss": 0.5368,
+      "step": 16114
+    },
+    {
+      "epoch": 4.277976901632815,
+      "grad_norm": 0.4670286381747807,
+      "learning_rate": 9.559417938227645e-07,
+      "loss": 0.4958,
+      "step": 16115
+    },
+    {
+      "epoch": 4.278242400106199,
+      "grad_norm": 0.4961140339950053,
+      "learning_rate": 9.556672288217894e-07,
+      "loss": 0.5506,
+      "step": 16116
+    },
+    {
+      "epoch": 4.2785078985795835,
+      "grad_norm": 0.4703959930478149,
+      "learning_rate": 9.553926939390847e-07,
+      "loss": 0.546,
+      "step": 16117
+    },
+    {
+      "epoch": 4.278773397052967,
+      "grad_norm": 0.47164671502162875,
+      "learning_rate": 9.551181891800034e-07,
+      "loss": 0.5652,
+      "step": 16118
+    },
+    {
+      "epoch": 4.279038895526351,
+      "grad_norm": 0.4759257335530126,
+      "learning_rate": 9.54843714549901e-07,
+      "loss": 0.5451,
+      "step": 16119
+    },
+    {
+      "epoch": 4.279304393999735,
+      "grad_norm": 0.4942306060661886,
+      "learning_rate": 9.545692700541284e-07,
+      "loss": 0.558,
+      "step": 16120
+    },
+    {
+      "epoch": 4.279569892473118,
+      "grad_norm": 0.48477733508936244,
+      "learning_rate": 9.542948556980396e-07,
+      "loss": 0.5201,
+      "step": 16121
+    },
+    {
+      "epoch": 4.279835390946502,
+      "grad_norm": 0.46041307978856433,
+      "learning_rate": 9.54020471486985e-07,
+      "loss": 0.463,
+      "step": 16122
+    },
+    {
+      "epoch": 4.280100889419886,
+      "grad_norm": 0.4813174865330322,
+      "learning_rate": 9.53746117426318e-07,
+      "loss": 0.5663,
+      "step": 16123
+    },
+    {
+      "epoch": 4.280366387893269,
+      "grad_norm": 0.4877195222172295,
+      "learning_rate": 9.534717935213861e-07,
+      "loss": 0.5312,
+      "step": 16124
+    },
+    {
+      "epoch": 4.2806318863666535,
+      "grad_norm": 0.48403467329840777,
+      "learning_rate": 9.531974997775415e-07,
+      "loss": 0.5543,
+      "step": 16125
+    },
+    {
+      "epoch": 4.280897384840038,
+      "grad_norm": 0.4683525674011287,
+      "learning_rate": 9.529232362001317e-07,
+      "loss": 0.5512,
+      "step": 16126
+    },
+    {
+      "epoch": 4.281162883313421,
+      "grad_norm": 0.4998225066379177,
+      "learning_rate": 9.526490027945079e-07,
+      "loss": 0.5535,
+      "step": 16127
+    },
+    {
+      "epoch": 4.281428381786805,
+      "grad_norm": 0.47963758682379953,
+      "learning_rate": 9.523747995660171e-07,
+      "loss": 0.512,
+      "step": 16128
+    },
+    {
+      "epoch": 4.281693880260189,
+      "grad_norm": 0.4881294811680851,
+      "learning_rate": 9.521006265200061e-07,
+      "loss": 0.5617,
+      "step": 16129
+    },
+    {
+      "epoch": 4.281959378733572,
+      "grad_norm": 0.45100918106953597,
+      "learning_rate": 9.518264836618235e-07,
+      "loss": 0.503,
+      "step": 16130
+    },
+    {
+      "epoch": 4.282224877206956,
+      "grad_norm": 0.5006723882399844,
+      "learning_rate": 9.515523709968145e-07,
+      "loss": 0.5348,
+      "step": 16131
+    },
+    {
+      "epoch": 4.282490375680339,
+      "grad_norm": 0.5146927350541792,
+      "learning_rate": 9.51278288530326e-07,
+      "loss": 0.5141,
+      "step": 16132
+    },
+    {
+      "epoch": 4.2827558741537235,
+      "grad_norm": 0.48732172393054124,
+      "learning_rate": 9.510042362677022e-07,
+      "loss": 0.5527,
+      "step": 16133
+    },
+    {
+      "epoch": 4.283021372627108,
+      "grad_norm": 0.46928567272209737,
+      "learning_rate": 9.507302142142899e-07,
+      "loss": 0.5466,
+      "step": 16134
+    },
+    {
+      "epoch": 4.283286871100491,
+      "grad_norm": 0.4717044534597236,
+      "learning_rate": 9.504562223754299e-07,
+      "loss": 0.5338,
+      "step": 16135
+    },
+    {
+      "epoch": 4.283552369573875,
+      "grad_norm": 0.47831350543841095,
+      "learning_rate": 9.501822607564678e-07,
+      "loss": 0.5575,
+      "step": 16136
+    },
+    {
+      "epoch": 4.283817868047259,
+      "grad_norm": 0.47793444674497115,
+      "learning_rate": 9.49908329362747e-07,
+      "loss": 0.547,
+      "step": 16137
+    },
+    {
+      "epoch": 4.284083366520642,
+      "grad_norm": 0.47907698575589613,
+      "learning_rate": 9.496344281996081e-07,
+      "loss": 0.549,
+      "step": 16138
+    },
+    {
+      "epoch": 4.284348864994026,
+      "grad_norm": 0.47739258316581606,
+      "learning_rate": 9.493605572723946e-07,
+      "loss": 0.5353,
+      "step": 16139
+    },
+    {
+      "epoch": 4.28461436346741,
+      "grad_norm": 0.48723796176825107,
+      "learning_rate": 9.49086716586447e-07,
+      "loss": 0.5607,
+      "step": 16140
+    },
+    {
+      "epoch": 4.2848798619407935,
+      "grad_norm": 0.4769773182167794,
+      "learning_rate": 9.488129061471057e-07,
+      "loss": 0.5003,
+      "step": 16141
+    },
+    {
+      "epoch": 4.285145360414178,
+      "grad_norm": 0.4747009446944575,
+      "learning_rate": 9.485391259597099e-07,
+      "loss": 0.5131,
+      "step": 16142
+    },
+    {
+      "epoch": 4.285410858887562,
+      "grad_norm": 0.47806353165227333,
+      "learning_rate": 9.482653760296007e-07,
+      "loss": 0.5058,
+      "step": 16143
+    },
+    {
+      "epoch": 4.285676357360945,
+      "grad_norm": 0.4798065459503907,
+      "learning_rate": 9.479916563621147e-07,
+      "loss": 0.5438,
+      "step": 16144
+    },
+    {
+      "epoch": 4.285941855834329,
+      "grad_norm": 0.4780843780240689,
+      "learning_rate": 9.477179669625927e-07,
+      "loss": 0.5478,
+      "step": 16145
+    },
+    {
+      "epoch": 4.286207354307713,
+      "grad_norm": 0.46730210238748343,
+      "learning_rate": 9.474443078363707e-07,
+      "loss": 0.549,
+      "step": 16146
+    },
+    {
+      "epoch": 4.286472852781096,
+      "grad_norm": 0.47109015247716934,
+      "learning_rate": 9.471706789887852e-07,
+      "loss": 0.5329,
+      "step": 16147
+    },
+    {
+      "epoch": 4.28673835125448,
+      "grad_norm": 0.49357752495817003,
+      "learning_rate": 9.468970804251742e-07,
+      "loss": 0.5431,
+      "step": 16148
+    },
+    {
+      "epoch": 4.287003849727864,
+      "grad_norm": 0.4739079515181912,
+      "learning_rate": 9.466235121508722e-07,
+      "loss": 0.5608,
+      "step": 16149
+    },
+    {
+      "epoch": 4.287269348201248,
+      "grad_norm": 0.4720426246815597,
+      "learning_rate": 9.463499741712157e-07,
+      "loss": 0.5396,
+      "step": 16150
+    },
+    {
+      "epoch": 4.287534846674632,
+      "grad_norm": 0.49223986858574126,
+      "learning_rate": 9.460764664915386e-07,
+      "loss": 0.5415,
+      "step": 16151
+    },
+    {
+      "epoch": 4.287800345148016,
+      "grad_norm": 0.4945450951568442,
+      "learning_rate": 9.458029891171755e-07,
+      "loss": 0.5519,
+      "step": 16152
+    },
+    {
+      "epoch": 4.288065843621399,
+      "grad_norm": 0.49565694318361314,
+      "learning_rate": 9.455295420534583e-07,
+      "loss": 0.531,
+      "step": 16153
+    },
+    {
+      "epoch": 4.288331342094783,
+      "grad_norm": 0.48152222448362336,
+      "learning_rate": 9.452561253057219e-07,
+      "loss": 0.5617,
+      "step": 16154
+    },
+    {
+      "epoch": 4.288596840568167,
+      "grad_norm": 0.4857186548844074,
+      "learning_rate": 9.449827388792968e-07,
+      "loss": 0.5314,
+      "step": 16155
+    },
+    {
+      "epoch": 4.28886233904155,
+      "grad_norm": 0.49285837061140425,
+      "learning_rate": 9.447093827795156e-07,
+      "loss": 0.5426,
+      "step": 16156
+    },
+    {
+      "epoch": 4.289127837514934,
+      "grad_norm": 0.478490371523138,
+      "learning_rate": 9.444360570117111e-07,
+      "loss": 0.5193,
+      "step": 16157
+    },
+    {
+      "epoch": 4.2893933359883185,
+      "grad_norm": 0.4802187041363442,
+      "learning_rate": 9.441627615812107e-07,
+      "loss": 0.5642,
+      "step": 16158
+    },
+    {
+      "epoch": 4.289658834461702,
+      "grad_norm": 0.4769516921480737,
+      "learning_rate": 9.438894964933465e-07,
+      "loss": 0.5493,
+      "step": 16159
+    },
+    {
+      "epoch": 4.289924332935086,
+      "grad_norm": 0.4771215505453742,
+      "learning_rate": 9.436162617534464e-07,
+      "loss": 0.5269,
+      "step": 16160
+    },
+    {
+      "epoch": 4.290189831408469,
+      "grad_norm": 0.49507969056374723,
+      "learning_rate": 9.433430573668406e-07,
+      "loss": 0.5169,
+      "step": 16161
+    },
+    {
+      "epoch": 4.290455329881853,
+      "grad_norm": 0.48236974891098067,
+      "learning_rate": 9.430698833388558e-07,
+      "loss": 0.5811,
+      "step": 16162
+    },
+    {
+      "epoch": 4.290720828355237,
+      "grad_norm": 0.4782656939701205,
+      "learning_rate": 9.427967396748211e-07,
+      "loss": 0.5435,
+      "step": 16163
+    },
+    {
+      "epoch": 4.29098632682862,
+      "grad_norm": 0.47722168296213874,
+      "learning_rate": 9.425236263800624e-07,
+      "loss": 0.5508,
+      "step": 16164
+    },
+    {
+      "epoch": 4.291251825302004,
+      "grad_norm": 0.48304854026727073,
+      "learning_rate": 9.422505434599058e-07,
+      "loss": 0.5704,
+      "step": 16165
+    },
+    {
+      "epoch": 4.2915173237753885,
+      "grad_norm": 0.4732693905455067,
+      "learning_rate": 9.419774909196785e-07,
+      "loss": 0.5453,
+      "step": 16166
+    },
+    {
+      "epoch": 4.291782822248772,
+      "grad_norm": 0.47712885680913475,
+      "learning_rate": 9.417044687647039e-07,
+      "loss": 0.551,
+      "step": 16167
+    },
+    {
+      "epoch": 4.292048320722156,
+      "grad_norm": 0.4727841774069548,
+      "learning_rate": 9.414314770003083e-07,
+      "loss": 0.512,
+      "step": 16168
+    },
+    {
+      "epoch": 4.29231381919554,
+      "grad_norm": 0.4700497144419276,
+      "learning_rate": 9.411585156318151e-07,
+      "loss": 0.5449,
+      "step": 16169
+    },
+    {
+      "epoch": 4.292579317668923,
+      "grad_norm": 0.48989582100232487,
+      "learning_rate": 9.408855846645476e-07,
+      "loss": 0.5438,
+      "step": 16170
+    },
+    {
+      "epoch": 4.292844816142307,
+      "grad_norm": 0.4722147662197645,
+      "learning_rate": 9.406126841038274e-07,
+      "loss": 0.5433,
+      "step": 16171
+    },
+    {
+      "epoch": 4.293110314615691,
+      "grad_norm": 0.4896317156584069,
+      "learning_rate": 9.403398139549791e-07,
+      "loss": 0.5575,
+      "step": 16172
+    },
+    {
+      "epoch": 4.2933758130890745,
+      "grad_norm": 0.482051978527843,
+      "learning_rate": 9.400669742233221e-07,
+      "loss": 0.5202,
+      "step": 16173
+    },
+    {
+      "epoch": 4.2936413115624585,
+      "grad_norm": 0.4802108574272704,
+      "learning_rate": 9.397941649141793e-07,
+      "loss": 0.537,
+      "step": 16174
+    },
+    {
+      "epoch": 4.293906810035843,
+      "grad_norm": 0.48280442652870736,
+      "learning_rate": 9.395213860328703e-07,
+      "loss": 0.5154,
+      "step": 16175
+    },
+    {
+      "epoch": 4.294172308509226,
+      "grad_norm": 0.47299456323769246,
+      "learning_rate": 9.39248637584714e-07,
+      "loss": 0.5286,
+      "step": 16176
+    },
+    {
+      "epoch": 4.29443780698261,
+      "grad_norm": 0.48648646953501956,
+      "learning_rate": 9.389759195750317e-07,
+      "loss": 0.5006,
+      "step": 16177
+    },
+    {
+      "epoch": 4.294703305455994,
+      "grad_norm": 0.4833304883808922,
+      "learning_rate": 9.387032320091402e-07,
+      "loss": 0.5376,
+      "step": 16178
+    },
+    {
+      "epoch": 4.294968803929377,
+      "grad_norm": 0.483472522186496,
+      "learning_rate": 9.384305748923592e-07,
+      "loss": 0.5081,
+      "step": 16179
+    },
+    {
+      "epoch": 4.295234302402761,
+      "grad_norm": 0.453136929126105,
+      "learning_rate": 9.381579482300054e-07,
+      "loss": 0.5205,
+      "step": 16180
+    },
+    {
+      "epoch": 4.295499800876145,
+      "grad_norm": 0.49560667540920283,
+      "learning_rate": 9.378853520273948e-07,
+      "loss": 0.5323,
+      "step": 16181
+    },
+    {
+      "epoch": 4.2957652993495286,
+      "grad_norm": 0.49499776283313485,
+      "learning_rate": 9.376127862898454e-07,
+      "loss": 0.5726,
+      "step": 16182
+    },
+    {
+      "epoch": 4.296030797822913,
+      "grad_norm": 0.47334489340833974,
+      "learning_rate": 9.373402510226712e-07,
+      "loss": 0.5101,
+      "step": 16183
+    },
+    {
+      "epoch": 4.296296296296296,
+      "grad_norm": 0.47208004492161953,
+      "learning_rate": 9.37067746231189e-07,
+      "loss": 0.5227,
+      "step": 16184
+    },
+    {
+      "epoch": 4.29656179476968,
+      "grad_norm": 0.48304629466859617,
+      "learning_rate": 9.367952719207115e-07,
+      "loss": 0.5678,
+      "step": 16185
+    },
+    {
+      "epoch": 4.296827293243064,
+      "grad_norm": 0.48358985269045435,
+      "learning_rate": 9.365228280965555e-07,
+      "loss": 0.5831,
+      "step": 16186
+    },
+    {
+      "epoch": 4.297092791716448,
+      "grad_norm": 0.47454324764997263,
+      "learning_rate": 9.362504147640305e-07,
+      "loss": 0.5726,
+      "step": 16187
+    },
+    {
+      "epoch": 4.297358290189831,
+      "grad_norm": 0.4873600374620913,
+      "learning_rate": 9.35978031928452e-07,
+      "loss": 0.5416,
+      "step": 16188
+    },
+    {
+      "epoch": 4.297623788663215,
+      "grad_norm": 0.4880037734697343,
+      "learning_rate": 9.357056795951305e-07,
+      "loss": 0.5467,
+      "step": 16189
+    },
+    {
+      "epoch": 4.297889287136599,
+      "grad_norm": 0.4849809209648875,
+      "learning_rate": 9.354333577693791e-07,
+      "loss": 0.5482,
+      "step": 16190
+    },
+    {
+      "epoch": 4.298154785609983,
+      "grad_norm": 0.4676136087894692,
+      "learning_rate": 9.351610664565072e-07,
+      "loss": 0.5404,
+      "step": 16191
+    },
+    {
+      "epoch": 4.298420284083367,
+      "grad_norm": 0.47110654767898347,
+      "learning_rate": 9.348888056618266e-07,
+      "loss": 0.5283,
+      "step": 16192
+    },
+    {
+      "epoch": 4.29868578255675,
+      "grad_norm": 0.48995997897553945,
+      "learning_rate": 9.346165753906464e-07,
+      "loss": 0.5952,
+      "step": 16193
+    },
+    {
+      "epoch": 4.298951281030134,
+      "grad_norm": 0.48660949210062937,
+      "learning_rate": 9.343443756482748e-07,
+      "loss": 0.555,
+      "step": 16194
+    },
+    {
+      "epoch": 4.299216779503518,
+      "grad_norm": 0.4963571687802391,
+      "learning_rate": 9.340722064400223e-07,
+      "loss": 0.5525,
+      "step": 16195
+    },
+    {
+      "epoch": 4.299482277976901,
+      "grad_norm": 0.487496856989421,
+      "learning_rate": 9.338000677711948e-07,
+      "loss": 0.5742,
+      "step": 16196
+    },
+    {
+      "epoch": 4.299747776450285,
+      "grad_norm": 0.4953732818832387,
+      "learning_rate": 9.335279596471025e-07,
+      "loss": 0.5519,
+      "step": 16197
+    },
+    {
+      "epoch": 4.3000132749236695,
+      "grad_norm": 0.4743435817553822,
+      "learning_rate": 9.332558820730486e-07,
+      "loss": 0.5233,
+      "step": 16198
+    },
+    {
+      "epoch": 4.300278773397053,
+      "grad_norm": 0.4698061009828211,
+      "learning_rate": 9.329838350543418e-07,
+      "loss": 0.5097,
+      "step": 16199
+    },
+    {
+      "epoch": 4.300544271870437,
+      "grad_norm": 0.47745370230663964,
+      "learning_rate": 9.327118185962864e-07,
+      "loss": 0.5574,
+      "step": 16200
+    },
+    {
+      "epoch": 4.300809770343821,
+      "grad_norm": 0.47717845958316396,
+      "learning_rate": 9.324398327041875e-07,
+      "loss": 0.5233,
+      "step": 16201
+    },
+    {
+      "epoch": 4.301075268817204,
+      "grad_norm": 0.49697791197964475,
+      "learning_rate": 9.321678773833509e-07,
+      "loss": 0.5452,
+      "step": 16202
+    },
+    {
+      "epoch": 4.301340767290588,
+      "grad_norm": 0.48272440510497366,
+      "learning_rate": 9.318959526390797e-07,
+      "loss": 0.5468,
+      "step": 16203
+    },
+    {
+      "epoch": 4.301606265763972,
+      "grad_norm": 0.46774505865339017,
+      "learning_rate": 9.316240584766764e-07,
+      "loss": 0.4921,
+      "step": 16204
+    },
+    {
+      "epoch": 4.301871764237355,
+      "grad_norm": 0.4893266242632504,
+      "learning_rate": 9.313521949014434e-07,
+      "loss": 0.5566,
+      "step": 16205
+    },
+    {
+      "epoch": 4.3021372627107395,
+      "grad_norm": 0.46944538895016286,
+      "learning_rate": 9.310803619186842e-07,
+      "loss": 0.5561,
+      "step": 16206
+    },
+    {
+      "epoch": 4.302402761184124,
+      "grad_norm": 0.46029536551563216,
+      "learning_rate": 9.308085595336983e-07,
+      "loss": 0.5186,
+      "step": 16207
+    },
+    {
+      "epoch": 4.302668259657507,
+      "grad_norm": 0.45728366771854756,
+      "learning_rate": 9.305367877517885e-07,
+      "loss": 0.5004,
+      "step": 16208
+    },
+    {
+      "epoch": 4.302933758130891,
+      "grad_norm": 0.4701312054524912,
+      "learning_rate": 9.30265046578254e-07,
+      "loss": 0.5281,
+      "step": 16209
+    },
+    {
+      "epoch": 4.303199256604275,
+      "grad_norm": 0.4716828165770199,
+      "learning_rate": 9.299933360183935e-07,
+      "loss": 0.514,
+      "step": 16210
+    },
+    {
+      "epoch": 4.303464755077658,
+      "grad_norm": 0.4733186995820755,
+      "learning_rate": 9.297216560775079e-07,
+      "loss": 0.565,
+      "step": 16211
+    },
+    {
+      "epoch": 4.303730253551042,
+      "grad_norm": 0.4969800792379079,
+      "learning_rate": 9.294500067608941e-07,
+      "loss": 0.5261,
+      "step": 16212
+    },
+    {
+      "epoch": 4.303995752024425,
+      "grad_norm": 0.46699768906856187,
+      "learning_rate": 9.291783880738511e-07,
+      "loss": 0.5571,
+      "step": 16213
+    },
+    {
+      "epoch": 4.3042612504978095,
+      "grad_norm": 0.4708692890400055,
+      "learning_rate": 9.28906800021675e-07,
+      "loss": 0.5296,
+      "step": 16214
+    },
+    {
+      "epoch": 4.304526748971194,
+      "grad_norm": 0.4689040155195078,
+      "learning_rate": 9.286352426096645e-07,
+      "loss": 0.5469,
+      "step": 16215
+    },
+    {
+      "epoch": 4.304792247444577,
+      "grad_norm": 0.4672868363691084,
+      "learning_rate": 9.283637158431124e-07,
+      "loss": 0.5475,
+      "step": 16216
+    },
+    {
+      "epoch": 4.305057745917961,
+      "grad_norm": 0.491190810246371,
+      "learning_rate": 9.280922197273168e-07,
+      "loss": 0.5937,
+      "step": 16217
+    },
+    {
+      "epoch": 4.305323244391345,
+      "grad_norm": 0.51341845180117,
+      "learning_rate": 9.278207542675708e-07,
+      "loss": 0.5231,
+      "step": 16218
+    },
+    {
+      "epoch": 4.305588742864728,
+      "grad_norm": 0.48648144639881646,
+      "learning_rate": 9.2754931946917e-07,
+      "loss": 0.5327,
+      "step": 16219
+    },
+    {
+      "epoch": 4.305854241338112,
+      "grad_norm": 0.47554248197372156,
+      "learning_rate": 9.27277915337407e-07,
+      "loss": 0.549,
+      "step": 16220
+    },
+    {
+      "epoch": 4.306119739811496,
+      "grad_norm": 0.48650848049535017,
+      "learning_rate": 9.270065418775762e-07,
+      "loss": 0.5455,
+      "step": 16221
+    },
+    {
+      "epoch": 4.3063852382848795,
+      "grad_norm": 0.47652047841134165,
+      "learning_rate": 9.267351990949689e-07,
+      "loss": 0.5701,
+      "step": 16222
+    },
+    {
+      "epoch": 4.306650736758264,
+      "grad_norm": 0.4708329438607657,
+      "learning_rate": 9.264638869948766e-07,
+      "loss": 0.5039,
+      "step": 16223
+    },
+    {
+      "epoch": 4.306916235231648,
+      "grad_norm": 0.4885416853075303,
+      "learning_rate": 9.261926055825918e-07,
+      "loss": 0.5444,
+      "step": 16224
+    },
+    {
+      "epoch": 4.307181733705031,
+      "grad_norm": 0.47224592581410313,
+      "learning_rate": 9.259213548634038e-07,
+      "loss": 0.523,
+      "step": 16225
+    },
+    {
+      "epoch": 4.307447232178415,
+      "grad_norm": 0.4701605892166753,
+      "learning_rate": 9.256501348426045e-07,
+      "loss": 0.551,
+      "step": 16226
+    },
+    {
+      "epoch": 4.307712730651799,
+      "grad_norm": 0.4739326924814448,
+      "learning_rate": 9.253789455254819e-07,
+      "loss": 0.5143,
+      "step": 16227
+    },
+    {
+      "epoch": 4.307978229125182,
+      "grad_norm": 0.4660940473803378,
+      "learning_rate": 9.251077869173244e-07,
+      "loss": 0.5206,
+      "step": 16228
+    },
+    {
+      "epoch": 4.308243727598566,
+      "grad_norm": 0.4862391187333547,
+      "learning_rate": 9.248366590234223e-07,
+      "loss": 0.5449,
+      "step": 16229
+    },
+    {
+      "epoch": 4.30850922607195,
+      "grad_norm": 0.4728453801821844,
+      "learning_rate": 9.245655618490607e-07,
+      "loss": 0.5705,
+      "step": 16230
+    },
+    {
+      "epoch": 4.308774724545334,
+      "grad_norm": 0.4777446418873602,
+      "learning_rate": 9.24294495399529e-07,
+      "loss": 0.5578,
+      "step": 16231
+    },
+    {
+      "epoch": 4.309040223018718,
+      "grad_norm": 0.47592370135748197,
+      "learning_rate": 9.240234596801125e-07,
+      "loss": 0.5007,
+      "step": 16232
+    },
+    {
+      "epoch": 4.309305721492102,
+      "grad_norm": 0.4849336033241753,
+      "learning_rate": 9.237524546960975e-07,
+      "loss": 0.5817,
+      "step": 16233
+    },
+    {
+      "epoch": 4.309571219965485,
+      "grad_norm": 0.4841216031843534,
+      "learning_rate": 9.234814804527678e-07,
+      "loss": 0.5858,
+      "step": 16234
+    },
+    {
+      "epoch": 4.309836718438869,
+      "grad_norm": 0.47981510459941557,
+      "learning_rate": 9.232105369554101e-07,
+      "loss": 0.5123,
+      "step": 16235
+    },
+    {
+      "epoch": 4.310102216912253,
+      "grad_norm": 0.4827896995813429,
+      "learning_rate": 9.22939624209307e-07,
+      "loss": 0.5339,
+      "step": 16236
+    },
+    {
+      "epoch": 4.310367715385636,
+      "grad_norm": 0.4801186329778531,
+      "learning_rate": 9.226687422197431e-07,
+      "loss": 0.5608,
+      "step": 16237
+    },
+    {
+      "epoch": 4.31063321385902,
+      "grad_norm": 0.4572255553183137,
+      "learning_rate": 9.223978909920009e-07,
+      "loss": 0.5,
+      "step": 16238
+    },
+    {
+      "epoch": 4.3108987123324045,
+      "grad_norm": 0.474467992914293,
+      "learning_rate": 9.221270705313614e-07,
+      "loss": 0.5243,
+      "step": 16239
+    },
+    {
+      "epoch": 4.311164210805788,
+      "grad_norm": 0.4635335030024976,
+      "learning_rate": 9.218562808431084e-07,
+      "loss": 0.5031,
+      "step": 16240
+    },
+    {
+      "epoch": 4.311429709279172,
+      "grad_norm": 0.4819448188718896,
+      "learning_rate": 9.215855219325207e-07,
+      "loss": 0.552,
+      "step": 16241
+    },
+    {
+      "epoch": 4.311695207752555,
+      "grad_norm": 0.4811299703677316,
+      "learning_rate": 9.21314793804881e-07,
+      "loss": 0.5628,
+      "step": 16242
+    },
+    {
+      "epoch": 4.311960706225939,
+      "grad_norm": 0.4835694712294409,
+      "learning_rate": 9.210440964654674e-07,
+      "loss": 0.5476,
+      "step": 16243
+    },
+    {
+      "epoch": 4.312226204699323,
+      "grad_norm": 0.4815989460394889,
+      "learning_rate": 9.207734299195614e-07,
+      "loss": 0.5203,
+      "step": 16244
+    },
+    {
+      "epoch": 4.312491703172706,
+      "grad_norm": 0.48566166185474274,
+      "learning_rate": 9.205027941724384e-07,
+      "loss": 0.544,
+      "step": 16245
+    },
+    {
+      "epoch": 4.31275720164609,
+      "grad_norm": 0.4665034270706131,
+      "learning_rate": 9.202321892293791e-07,
+      "loss": 0.5468,
+      "step": 16246
+    },
+    {
+      "epoch": 4.3130227001194745,
+      "grad_norm": 0.46694843431482913,
+      "learning_rate": 9.199616150956592e-07,
+      "loss": 0.5468,
+      "step": 16247
+    },
+    {
+      "epoch": 4.313288198592858,
+      "grad_norm": 0.45966111206574933,
+      "learning_rate": 9.196910717765564e-07,
+      "loss": 0.5253,
+      "step": 16248
+    },
+    {
+      "epoch": 4.313553697066242,
+      "grad_norm": 0.4783669474696989,
+      "learning_rate": 9.194205592773486e-07,
+      "loss": 0.5397,
+      "step": 16249
+    },
+    {
+      "epoch": 4.313819195539626,
+      "grad_norm": 0.46669517610391437,
+      "learning_rate": 9.191500776033083e-07,
+      "loss": 0.5003,
+      "step": 16250
+    },
+    {
+      "epoch": 4.314084694013009,
+      "grad_norm": 0.4700783277246493,
+      "learning_rate": 9.188796267597127e-07,
+      "loss": 0.5443,
+      "step": 16251
+    },
+    {
+      "epoch": 4.314350192486393,
+      "grad_norm": 0.4793370698317174,
+      "learning_rate": 9.186092067518348e-07,
+      "loss": 0.541,
+      "step": 16252
+    },
+    {
+      "epoch": 4.314615690959777,
+      "grad_norm": 0.4904006860925553,
+      "learning_rate": 9.183388175849503e-07,
+      "loss": 0.5604,
+      "step": 16253
+    },
+    {
+      "epoch": 4.3148811894331605,
+      "grad_norm": 0.4570081260307148,
+      "learning_rate": 9.180684592643305e-07,
+      "loss": 0.5371,
+      "step": 16254
+    },
+    {
+      "epoch": 4.3151466879065445,
+      "grad_norm": 0.4742716717005486,
+      "learning_rate": 9.1779813179525e-07,
+      "loss": 0.5013,
+      "step": 16255
+    },
+    {
+      "epoch": 4.315412186379929,
+      "grad_norm": 0.4832400161382249,
+      "learning_rate": 9.175278351829794e-07,
+      "loss": 0.5232,
+      "step": 16256
+    },
+    {
+      "epoch": 4.315677684853312,
+      "grad_norm": 0.4629710767602226,
+      "learning_rate": 9.172575694327901e-07,
+      "loss": 0.5264,
+      "step": 16257
+    },
+    {
+      "epoch": 4.315943183326696,
+      "grad_norm": 0.4990161627033716,
+      "learning_rate": 9.169873345499541e-07,
+      "loss": 0.5386,
+      "step": 16258
+    },
+    {
+      "epoch": 4.31620868180008,
+      "grad_norm": 0.4807878482598025,
+      "learning_rate": 9.167171305397401e-07,
+      "loss": 0.5326,
+      "step": 16259
+    },
+    {
+      "epoch": 4.316474180273463,
+      "grad_norm": 0.47089063093857714,
+      "learning_rate": 9.164469574074198e-07,
+      "loss": 0.5493,
+      "step": 16260
+    },
+    {
+      "epoch": 4.316739678746847,
+      "grad_norm": 0.4674818559717405,
+      "learning_rate": 9.161768151582607e-07,
+      "loss": 0.5324,
+      "step": 16261
+    },
+    {
+      "epoch": 4.317005177220231,
+      "grad_norm": 0.4805347364712472,
+      "learning_rate": 9.159067037975317e-07,
+      "loss": 0.5444,
+      "step": 16262
+    },
+    {
+      "epoch": 4.317270675693615,
+      "grad_norm": 0.4842855701445545,
+      "learning_rate": 9.156366233304995e-07,
+      "loss": 0.5624,
+      "step": 16263
+    },
+    {
+      "epoch": 4.317536174166999,
+      "grad_norm": 0.47862276976868473,
+      "learning_rate": 9.153665737624331e-07,
+      "loss": 0.5361,
+      "step": 16264
+    },
+    {
+      "epoch": 4.317801672640383,
+      "grad_norm": 0.4684594369400062,
+      "learning_rate": 9.150965550985976e-07,
+      "loss": 0.5257,
+      "step": 16265
+    },
+    {
+      "epoch": 4.318067171113766,
+      "grad_norm": 0.48064983757426455,
+      "learning_rate": 9.148265673442608e-07,
+      "loss": 0.5327,
+      "step": 16266
+    },
+    {
+      "epoch": 4.31833266958715,
+      "grad_norm": 0.4741107566999684,
+      "learning_rate": 9.145566105046872e-07,
+      "loss": 0.5693,
+      "step": 16267
+    },
+    {
+      "epoch": 4.318598168060534,
+      "grad_norm": 0.472037677150413,
+      "learning_rate": 9.142866845851403e-07,
+      "loss": 0.5152,
+      "step": 16268
+    },
+    {
+      "epoch": 4.318863666533917,
+      "grad_norm": 0.4886084695344888,
+      "learning_rate": 9.140167895908867e-07,
+      "loss": 0.5755,
+      "step": 16269
+    },
+    {
+      "epoch": 4.319129165007301,
+      "grad_norm": 0.5135276331933102,
+      "learning_rate": 9.13746925527188e-07,
+      "loss": 0.5618,
+      "step": 16270
+    },
+    {
+      "epoch": 4.319394663480685,
+      "grad_norm": 0.47746310522230434,
+      "learning_rate": 9.134770923993088e-07,
+      "loss": 0.5891,
+      "step": 16271
+    },
+    {
+      "epoch": 4.319660161954069,
+      "grad_norm": 0.4814733504421007,
+      "learning_rate": 9.132072902125108e-07,
+      "loss": 0.5243,
+      "step": 16272
+    },
+    {
+      "epoch": 4.319925660427453,
+      "grad_norm": 0.4634371537090962,
+      "learning_rate": 9.129375189720549e-07,
+      "loss": 0.52,
+      "step": 16273
+    },
+    {
+      "epoch": 4.320191158900836,
+      "grad_norm": 0.47701958698664676,
+      "learning_rate": 9.126677786832041e-07,
+      "loss": 0.522,
+      "step": 16274
+    },
+    {
+      "epoch": 4.32045665737422,
+      "grad_norm": 0.47326203536861716,
+      "learning_rate": 9.123980693512172e-07,
+      "loss": 0.5742,
+      "step": 16275
+    },
+    {
+      "epoch": 4.320722155847604,
+      "grad_norm": 0.47469105458997823,
+      "learning_rate": 9.121283909813558e-07,
+      "loss": 0.5273,
+      "step": 16276
+    },
+    {
+      "epoch": 4.320987654320987,
+      "grad_norm": 0.4691269667741205,
+      "learning_rate": 9.118587435788778e-07,
+      "loss": 0.5454,
+      "step": 16277
+    },
+    {
+      "epoch": 4.321253152794371,
+      "grad_norm": 0.48678897507555063,
+      "learning_rate": 9.115891271490446e-07,
+      "loss": 0.5457,
+      "step": 16278
+    },
+    {
+      "epoch": 4.3215186512677555,
+      "grad_norm": 0.4774631618420226,
+      "learning_rate": 9.113195416971105e-07,
+      "loss": 0.5422,
+      "step": 16279
+    },
+    {
+      "epoch": 4.321784149741139,
+      "grad_norm": 0.48343449183843334,
+      "learning_rate": 9.110499872283363e-07,
+      "loss": 0.5725,
+      "step": 16280
+    },
+    {
+      "epoch": 4.322049648214523,
+      "grad_norm": 0.4828506292153186,
+      "learning_rate": 9.107804637479767e-07,
+      "loss": 0.5412,
+      "step": 16281
+    },
+    {
+      "epoch": 4.322315146687907,
+      "grad_norm": 0.4833603996511266,
+      "learning_rate": 9.1051097126129e-07,
+      "loss": 0.5636,
+      "step": 16282
+    },
+    {
+      "epoch": 4.32258064516129,
+      "grad_norm": 0.4876591786379884,
+      "learning_rate": 9.102415097735304e-07,
+      "loss": 0.5523,
+      "step": 16283
+    },
+    {
+      "epoch": 4.322846143634674,
+      "grad_norm": 0.485920593324433,
+      "learning_rate": 9.099720792899544e-07,
+      "loss": 0.5428,
+      "step": 16284
+    },
+    {
+      "epoch": 4.323111642108058,
+      "grad_norm": 0.46044418943972143,
+      "learning_rate": 9.097026798158159e-07,
+      "loss": 0.5554,
+      "step": 16285
+    },
+    {
+      "epoch": 4.323377140581441,
+      "grad_norm": 0.4759835195925352,
+      "learning_rate": 9.09433311356368e-07,
+      "loss": 0.511,
+      "step": 16286
+    },
+    {
+      "epoch": 4.3236426390548255,
+      "grad_norm": 0.48286241322937196,
+      "learning_rate": 9.091639739168661e-07,
+      "loss": 0.5516,
+      "step": 16287
+    },
+    {
+      "epoch": 4.32390813752821,
+      "grad_norm": 0.48296999689141584,
+      "learning_rate": 9.088946675025607e-07,
+      "loss": 0.537,
+      "step": 16288
+    },
+    {
+      "epoch": 4.324173636001593,
+      "grad_norm": 0.47129293045010023,
+      "learning_rate": 9.086253921187058e-07,
+      "loss": 0.5322,
+      "step": 16289
+    },
+    {
+      "epoch": 4.324439134474977,
+      "grad_norm": 0.45400931223834307,
+      "learning_rate": 9.083561477705521e-07,
+      "loss": 0.5276,
+      "step": 16290
+    },
+    {
+      "epoch": 4.324704632948361,
+      "grad_norm": 0.4717739181427923,
+      "learning_rate": 9.080869344633506e-07,
+      "loss": 0.4833,
+      "step": 16291
+    },
+    {
+      "epoch": 4.324970131421744,
+      "grad_norm": 0.48248620035767176,
+      "learning_rate": 9.078177522023504e-07,
+      "loss": 0.5309,
+      "step": 16292
+    },
+    {
+      "epoch": 4.325235629895128,
+      "grad_norm": 0.47006097161126564,
+      "learning_rate": 9.075486009928028e-07,
+      "loss": 0.5138,
+      "step": 16293
+    },
+    {
+      "epoch": 4.325501128368511,
+      "grad_norm": 0.46906622991409325,
+      "learning_rate": 9.072794808399571e-07,
+      "loss": 0.5577,
+      "step": 16294
+    },
+    {
+      "epoch": 4.3257666268418955,
+      "grad_norm": 0.46613551172219764,
+      "learning_rate": 9.07010391749061e-07,
+      "loss": 0.5274,
+      "step": 16295
+    },
+    {
+      "epoch": 4.32603212531528,
+      "grad_norm": 0.48324233240297965,
+      "learning_rate": 9.067413337253631e-07,
+      "loss": 0.5776,
+      "step": 16296
+    },
+    {
+      "epoch": 4.326297623788664,
+      "grad_norm": 0.46050292650406666,
+      "learning_rate": 9.064723067741089e-07,
+      "loss": 0.4998,
+      "step": 16297
+    },
+    {
+      "epoch": 4.326563122262047,
+      "grad_norm": 0.48131217816698757,
+      "learning_rate": 9.062033109005472e-07,
+      "loss": 0.5396,
+      "step": 16298
+    },
+    {
+      "epoch": 4.326828620735431,
+      "grad_norm": 0.48024506138220074,
+      "learning_rate": 9.059343461099227e-07,
+      "loss": 0.5089,
+      "step": 16299
+    },
+    {
+      "epoch": 4.327094119208814,
+      "grad_norm": 0.48212067758377564,
+      "learning_rate": 9.056654124074818e-07,
+      "loss": 0.5881,
+      "step": 16300
+    },
+    {
+      "epoch": 4.327359617682198,
+      "grad_norm": 0.4706352870017432,
+      "learning_rate": 9.053965097984694e-07,
+      "loss": 0.5136,
+      "step": 16301
+    },
+    {
+      "epoch": 4.327625116155582,
+      "grad_norm": 0.47980685297609904,
+      "learning_rate": 9.05127638288128e-07,
+      "loss": 0.5582,
+      "step": 16302
+    },
+    {
+      "epoch": 4.3278906146289655,
+      "grad_norm": 0.48576016251790705,
+      "learning_rate": 9.048587978817036e-07,
+      "loss": 0.5578,
+      "step": 16303
+    },
+    {
+      "epoch": 4.32815611310235,
+      "grad_norm": 0.48721438187293226,
+      "learning_rate": 9.045899885844376e-07,
+      "loss": 0.5272,
+      "step": 16304
+    },
+    {
+      "epoch": 4.328421611575734,
+      "grad_norm": 0.4775734136990346,
+      "learning_rate": 9.043212104015736e-07,
+      "loss": 0.5199,
+      "step": 16305
+    },
+    {
+      "epoch": 4.328687110049117,
+      "grad_norm": 0.4852569658872731,
+      "learning_rate": 9.040524633383524e-07,
+      "loss": 0.5407,
+      "step": 16306
+    },
+    {
+      "epoch": 4.328952608522501,
+      "grad_norm": 0.48198080052217207,
+      "learning_rate": 9.037837474000172e-07,
+      "loss": 0.5407,
+      "step": 16307
+    },
+    {
+      "epoch": 4.329218106995885,
+      "grad_norm": 0.48369817900686396,
+      "learning_rate": 9.035150625918054e-07,
+      "loss": 0.5316,
+      "step": 16308
+    },
+    {
+      "epoch": 4.329483605469268,
+      "grad_norm": 0.4828136312453083,
+      "learning_rate": 9.032464089189597e-07,
+      "loss": 0.559,
+      "step": 16309
+    },
+    {
+      "epoch": 4.329749103942652,
+      "grad_norm": 0.4793198651925156,
+      "learning_rate": 9.029777863867176e-07,
+      "loss": 0.5287,
+      "step": 16310
+    },
+    {
+      "epoch": 4.330014602416036,
+      "grad_norm": 0.4825451280497991,
+      "learning_rate": 9.027091950003197e-07,
+      "loss": 0.5421,
+      "step": 16311
+    },
+    {
+      "epoch": 4.33028010088942,
+      "grad_norm": 0.47649580950533116,
+      "learning_rate": 9.024406347650025e-07,
+      "loss": 0.5209,
+      "step": 16312
+    },
+    {
+      "epoch": 4.330545599362804,
+      "grad_norm": 0.46668818047879146,
+      "learning_rate": 9.021721056860056e-07,
+      "loss": 0.5625,
+      "step": 16313
+    },
+    {
+      "epoch": 4.330811097836188,
+      "grad_norm": 0.49407683489090953,
+      "learning_rate": 9.019036077685644e-07,
+      "loss": 0.5702,
+      "step": 16314
+    },
+    {
+      "epoch": 4.331076596309571,
+      "grad_norm": 0.463319271207252,
+      "learning_rate": 9.01635141017915e-07,
+      "loss": 0.5101,
+      "step": 16315
+    },
+    {
+      "epoch": 4.331342094782955,
+      "grad_norm": 0.47507833204575983,
+      "learning_rate": 9.013667054392944e-07,
+      "loss": 0.5342,
+      "step": 16316
+    },
+    {
+      "epoch": 4.331607593256339,
+      "grad_norm": 0.4820185338383231,
+      "learning_rate": 9.010983010379365e-07,
+      "loss": 0.5499,
+      "step": 16317
+    },
+    {
+      "epoch": 4.331873091729722,
+      "grad_norm": 0.47464099235441276,
+      "learning_rate": 9.008299278190774e-07,
+      "loss": 0.5247,
+      "step": 16318
+    },
+    {
+      "epoch": 4.332138590203106,
+      "grad_norm": 0.4818044907209706,
+      "learning_rate": 9.005615857879502e-07,
+      "loss": 0.5424,
+      "step": 16319
+    },
+    {
+      "epoch": 4.3324040886764905,
+      "grad_norm": 0.4751429660100645,
+      "learning_rate": 9.002932749497873e-07,
+      "loss": 0.5366,
+      "step": 16320
+    },
+    {
+      "epoch": 4.332669587149874,
+      "grad_norm": 0.49614938126701463,
+      "learning_rate": 9.00024995309823e-07,
+      "loss": 0.5442,
+      "step": 16321
+    },
+    {
+      "epoch": 4.332935085623258,
+      "grad_norm": 0.491677585627819,
+      "learning_rate": 8.997567468732882e-07,
+      "loss": 0.5462,
+      "step": 16322
+    },
+    {
+      "epoch": 4.333200584096641,
+      "grad_norm": 0.4820238971739135,
+      "learning_rate": 8.994885296454153e-07,
+      "loss": 0.5332,
+      "step": 16323
+    },
+    {
+      "epoch": 4.333466082570025,
+      "grad_norm": 0.4973643342920117,
+      "learning_rate": 8.992203436314351e-07,
+      "loss": 0.5778,
+      "step": 16324
+    },
+    {
+      "epoch": 4.333731581043409,
+      "grad_norm": 0.47249798507687596,
+      "learning_rate": 8.989521888365774e-07,
+      "loss": 0.5392,
+      "step": 16325
+    },
+    {
+      "epoch": 4.333997079516793,
+      "grad_norm": 0.4866948396650183,
+      "learning_rate": 8.986840652660713e-07,
+      "loss": 0.5289,
+      "step": 16326
+    },
+    {
+      "epoch": 4.3342625779901764,
+      "grad_norm": 0.47275864954586744,
+      "learning_rate": 8.984159729251477e-07,
+      "loss": 0.5343,
+      "step": 16327
+    },
+    {
+      "epoch": 4.3345280764635605,
+      "grad_norm": 0.47582861257715003,
+      "learning_rate": 8.981479118190329e-07,
+      "loss": 0.5527,
+      "step": 16328
+    },
+    {
+      "epoch": 4.334793574936944,
+      "grad_norm": 0.4676699656079308,
+      "learning_rate": 8.978798819529566e-07,
+      "loss": 0.4909,
+      "step": 16329
+    },
+    {
+      "epoch": 4.335059073410328,
+      "grad_norm": 0.4617238203609274,
+      "learning_rate": 8.976118833321453e-07,
+      "loss": 0.5054,
+      "step": 16330
+    },
+    {
+      "epoch": 4.335324571883712,
+      "grad_norm": 0.47904408964188494,
+      "learning_rate": 8.97343915961825e-07,
+      "loss": 0.5339,
+      "step": 16331
+    },
+    {
+      "epoch": 4.335590070357095,
+      "grad_norm": 0.47422561303894234,
+      "learning_rate": 8.970759798472229e-07,
+      "loss": 0.543,
+      "step": 16332
+    },
+    {
+      "epoch": 4.335855568830479,
+      "grad_norm": 0.46660585915131697,
+      "learning_rate": 8.968080749935632e-07,
+      "loss": 0.5601,
+      "step": 16333
+    },
+    {
+      "epoch": 4.336121067303863,
+      "grad_norm": 0.47113454487936157,
+      "learning_rate": 8.96540201406072e-07,
+      "loss": 0.5458,
+      "step": 16334
+    },
+    {
+      "epoch": 4.3363865657772465,
+      "grad_norm": 0.48941246030759455,
+      "learning_rate": 8.962723590899719e-07,
+      "loss": 0.5434,
+      "step": 16335
+    },
+    {
+      "epoch": 4.3366520642506305,
+      "grad_norm": 0.4762056612955769,
+      "learning_rate": 8.960045480504892e-07,
+      "loss": 0.5483,
+      "step": 16336
+    },
+    {
+      "epoch": 4.336917562724015,
+      "grad_norm": 0.46372563334411826,
+      "learning_rate": 8.957367682928434e-07,
+      "loss": 0.4953,
+      "step": 16337
+    },
+    {
+      "epoch": 4.337183061197398,
+      "grad_norm": 0.48682278493161013,
+      "learning_rate": 8.954690198222585e-07,
+      "loss": 0.5359,
+      "step": 16338
+    },
+    {
+      "epoch": 4.337448559670782,
+      "grad_norm": 0.4849276778574229,
+      "learning_rate": 8.952013026439571e-07,
+      "loss": 0.5301,
+      "step": 16339
+    },
+    {
+      "epoch": 4.337714058144166,
+      "grad_norm": 0.4840970923871177,
+      "learning_rate": 8.949336167631587e-07,
+      "loss": 0.5411,
+      "step": 16340
+    },
+    {
+      "epoch": 4.337979556617549,
+      "grad_norm": 0.47605207428791535,
+      "learning_rate": 8.946659621850856e-07,
+      "loss": 0.5192,
+      "step": 16341
+    },
+    {
+      "epoch": 4.338245055090933,
+      "grad_norm": 0.48124455871737276,
+      "learning_rate": 8.943983389149566e-07,
+      "loss": 0.5336,
+      "step": 16342
+    },
+    {
+      "epoch": 4.338510553564317,
+      "grad_norm": 0.465209246371635,
+      "learning_rate": 8.941307469579913e-07,
+      "loss": 0.5267,
+      "step": 16343
+    },
+    {
+      "epoch": 4.338776052037701,
+      "grad_norm": 0.4757634415782952,
+      "learning_rate": 8.938631863194074e-07,
+      "loss": 0.5577,
+      "step": 16344
+    },
+    {
+      "epoch": 4.339041550511085,
+      "grad_norm": 0.47984485563879087,
+      "learning_rate": 8.935956570044249e-07,
+      "loss": 0.5571,
+      "step": 16345
+    },
+    {
+      "epoch": 4.339307048984469,
+      "grad_norm": 0.48335074589637267,
+      "learning_rate": 8.933281590182591e-07,
+      "loss": 0.557,
+      "step": 16346
+    },
+    {
+      "epoch": 4.339572547457852,
+      "grad_norm": 0.4724573534049644,
+      "learning_rate": 8.930606923661289e-07,
+      "loss": 0.5136,
+      "step": 16347
+    },
+    {
+      "epoch": 4.339838045931236,
+      "grad_norm": 0.4588307856771073,
+      "learning_rate": 8.927932570532499e-07,
+      "loss": 0.5066,
+      "step": 16348
+    },
+    {
+      "epoch": 4.34010354440462,
+      "grad_norm": 0.4743879300495682,
+      "learning_rate": 8.925258530848366e-07,
+      "loss": 0.5422,
+      "step": 16349
+    },
+    {
+      "epoch": 4.340369042878003,
+      "grad_norm": 0.47978505944828254,
+      "learning_rate": 8.922584804661056e-07,
+      "loss": 0.5313,
+      "step": 16350
+    },
+    {
+      "epoch": 4.340634541351387,
+      "grad_norm": 0.4615665841172697,
+      "learning_rate": 8.919911392022698e-07,
+      "loss": 0.535,
+      "step": 16351
+    },
+    {
+      "epoch": 4.340900039824771,
+      "grad_norm": 0.45884887646469985,
+      "learning_rate": 8.91723829298545e-07,
+      "loss": 0.4936,
+      "step": 16352
+    },
+    {
+      "epoch": 4.341165538298155,
+      "grad_norm": 0.4819228592108612,
+      "learning_rate": 8.91456550760143e-07,
+      "loss": 0.5672,
+      "step": 16353
+    },
+    {
+      "epoch": 4.341431036771539,
+      "grad_norm": 0.48558608985376944,
+      "learning_rate": 8.911893035922767e-07,
+      "loss": 0.5183,
+      "step": 16354
+    },
+    {
+      "epoch": 4.341696535244922,
+      "grad_norm": 0.49858477790193595,
+      "learning_rate": 8.909220878001573e-07,
+      "loss": 0.5512,
+      "step": 16355
+    },
+    {
+      "epoch": 4.341962033718306,
+      "grad_norm": 0.49048724191280935,
+      "learning_rate": 8.906549033889974e-07,
+      "loss": 0.5554,
+      "step": 16356
+    },
+    {
+      "epoch": 4.34222753219169,
+      "grad_norm": 0.4855794160800867,
+      "learning_rate": 8.903877503640068e-07,
+      "loss": 0.534,
+      "step": 16357
+    },
+    {
+      "epoch": 4.342493030665073,
+      "grad_norm": 0.4861068765618246,
+      "learning_rate": 8.901206287303968e-07,
+      "loss": 0.5637,
+      "step": 16358
+    },
+    {
+      "epoch": 4.342758529138457,
+      "grad_norm": 0.48771376240422293,
+      "learning_rate": 8.898535384933762e-07,
+      "loss": 0.5435,
+      "step": 16359
+    },
+    {
+      "epoch": 4.3430240276118415,
+      "grad_norm": 0.48448852480354004,
+      "learning_rate": 8.89586479658153e-07,
+      "loss": 0.5329,
+      "step": 16360
+    },
+    {
+      "epoch": 4.343289526085225,
+      "grad_norm": 0.4612957097296225,
+      "learning_rate": 8.893194522299373e-07,
+      "loss": 0.5477,
+      "step": 16361
+    },
+    {
+      "epoch": 4.343555024558609,
+      "grad_norm": 0.4783091768413817,
+      "learning_rate": 8.890524562139352e-07,
+      "loss": 0.5305,
+      "step": 16362
+    },
+    {
+      "epoch": 4.343820523031993,
+      "grad_norm": 0.4846329953759779,
+      "learning_rate": 8.887854916153552e-07,
+      "loss": 0.5129,
+      "step": 16363
+    },
+    {
+      "epoch": 4.344086021505376,
+      "grad_norm": 0.4834315532007541,
+      "learning_rate": 8.885185584394021e-07,
+      "loss": 0.5688,
+      "step": 16364
+    },
+    {
+      "epoch": 4.34435151997876,
+      "grad_norm": 0.4853017459557938,
+      "learning_rate": 8.882516566912838e-07,
+      "loss": 0.5309,
+      "step": 16365
+    },
+    {
+      "epoch": 4.344617018452144,
+      "grad_norm": 0.48305038628455726,
+      "learning_rate": 8.879847863762043e-07,
+      "loss": 0.5502,
+      "step": 16366
+    },
+    {
+      "epoch": 4.344882516925527,
+      "grad_norm": 0.46782007025759254,
+      "learning_rate": 8.877179474993675e-07,
+      "loss": 0.4994,
+      "step": 16367
+    },
+    {
+      "epoch": 4.3451480153989115,
+      "grad_norm": 0.4772463568664131,
+      "learning_rate": 8.874511400659791e-07,
+      "loss": 0.5521,
+      "step": 16368
+    },
+    {
+      "epoch": 4.345413513872296,
+      "grad_norm": 0.47463804725150294,
+      "learning_rate": 8.871843640812408e-07,
+      "loss": 0.5403,
+      "step": 16369
+    },
+    {
+      "epoch": 4.345679012345679,
+      "grad_norm": 0.4806707999051861,
+      "learning_rate": 8.869176195503579e-07,
+      "loss": 0.5648,
+      "step": 16370
+    },
+    {
+      "epoch": 4.345944510819063,
+      "grad_norm": 0.46998330578487396,
+      "learning_rate": 8.866509064785295e-07,
+      "loss": 0.5529,
+      "step": 16371
+    },
+    {
+      "epoch": 4.346210009292447,
+      "grad_norm": 0.4859555482762523,
+      "learning_rate": 8.863842248709592e-07,
+      "loss": 0.5485,
+      "step": 16372
+    },
+    {
+      "epoch": 4.34647550776583,
+      "grad_norm": 0.4774210319585395,
+      "learning_rate": 8.861175747328465e-07,
+      "loss": 0.5291,
+      "step": 16373
+    },
+    {
+      "epoch": 4.346741006239214,
+      "grad_norm": 0.46802633008101124,
+      "learning_rate": 8.858509560693934e-07,
+      "loss": 0.5493,
+      "step": 16374
+    },
+    {
+      "epoch": 4.347006504712598,
+      "grad_norm": 0.4690032759470632,
+      "learning_rate": 8.855843688857979e-07,
+      "loss": 0.5346,
+      "step": 16375
+    },
+    {
+      "epoch": 4.3472720031859815,
+      "grad_norm": 0.47376021796352863,
+      "learning_rate": 8.853178131872611e-07,
+      "loss": 0.5493,
+      "step": 16376
+    },
+    {
+      "epoch": 4.347537501659366,
+      "grad_norm": 0.4783184504129662,
+      "learning_rate": 8.8505128897898e-07,
+      "loss": 0.5591,
+      "step": 16377
+    },
+    {
+      "epoch": 4.34780300013275,
+      "grad_norm": 0.48061145571289315,
+      "learning_rate": 8.847847962661521e-07,
+      "loss": 0.5392,
+      "step": 16378
+    },
+    {
+      "epoch": 4.348068498606133,
+      "grad_norm": 0.4638283397117864,
+      "learning_rate": 8.845183350539763e-07,
+      "loss": 0.5021,
+      "step": 16379
+    },
+    {
+      "epoch": 4.348333997079517,
+      "grad_norm": 0.474443415607321,
+      "learning_rate": 8.842519053476476e-07,
+      "loss": 0.5289,
+      "step": 16380
+    },
+    {
+      "epoch": 4.3485994955529,
+      "grad_norm": 0.46414824985298847,
+      "learning_rate": 8.839855071523634e-07,
+      "loss": 0.5583,
+      "step": 16381
+    },
+    {
+      "epoch": 4.348864994026284,
+      "grad_norm": 0.48232483875074494,
+      "learning_rate": 8.837191404733186e-07,
+      "loss": 0.5557,
+      "step": 16382
+    },
+    {
+      "epoch": 4.349130492499668,
+      "grad_norm": 0.4792298494066698,
+      "learning_rate": 8.834528053157082e-07,
+      "loss": 0.5375,
+      "step": 16383
+    },
+    {
+      "epoch": 4.3493959909730515,
+      "grad_norm": 0.4868181308119856,
+      "learning_rate": 8.831865016847249e-07,
+      "loss": 0.5772,
+      "step": 16384
+    },
+    {
+      "epoch": 4.349661489446436,
+      "grad_norm": 0.4969230199257655,
+      "learning_rate": 8.829202295855635e-07,
+      "loss": 0.5143,
+      "step": 16385
+    },
+    {
+      "epoch": 4.34992698791982,
+      "grad_norm": 0.4786057844181593,
+      "learning_rate": 8.826539890234179e-07,
+      "loss": 0.5434,
+      "step": 16386
+    },
+    {
+      "epoch": 4.350192486393203,
+      "grad_norm": 0.4705197284882441,
+      "learning_rate": 8.823877800034783e-07,
+      "loss": 0.5202,
+      "step": 16387
+    },
+    {
+      "epoch": 4.350457984866587,
+      "grad_norm": 0.48086710614907097,
+      "learning_rate": 8.821216025309395e-07,
+      "loss": 0.5242,
+      "step": 16388
+    },
+    {
+      "epoch": 4.350723483339971,
+      "grad_norm": 0.4802249339539081,
+      "learning_rate": 8.818554566109891e-07,
+      "loss": 0.5378,
+      "step": 16389
+    },
+    {
+      "epoch": 4.350988981813354,
+      "grad_norm": 0.47031188344181596,
+      "learning_rate": 8.815893422488198e-07,
+      "loss": 0.5012,
+      "step": 16390
+    },
+    {
+      "epoch": 4.351254480286738,
+      "grad_norm": 0.47344812198782077,
+      "learning_rate": 8.813232594496201e-07,
+      "loss": 0.5211,
+      "step": 16391
+    },
+    {
+      "epoch": 4.351519978760122,
+      "grad_norm": 0.49847537240202067,
+      "learning_rate": 8.810572082185811e-07,
+      "loss": 0.5289,
+      "step": 16392
+    },
+    {
+      "epoch": 4.351785477233506,
+      "grad_norm": 0.48524962963409896,
+      "learning_rate": 8.807911885608902e-07,
+      "loss": 0.5372,
+      "step": 16393
+    },
+    {
+      "epoch": 4.35205097570689,
+      "grad_norm": 0.48752323248758206,
+      "learning_rate": 8.80525200481735e-07,
+      "loss": 0.5601,
+      "step": 16394
+    },
+    {
+      "epoch": 4.352316474180274,
+      "grad_norm": 0.4653201634358049,
+      "learning_rate": 8.802592439863041e-07,
+      "loss": 0.5575,
+      "step": 16395
+    },
+    {
+      "epoch": 4.352581972653657,
+      "grad_norm": 0.4710851386131509,
+      "learning_rate": 8.799933190797829e-07,
+      "loss": 0.5223,
+      "step": 16396
+    },
+    {
+      "epoch": 4.352847471127041,
+      "grad_norm": 0.4833994723769361,
+      "learning_rate": 8.797274257673596e-07,
+      "loss": 0.5325,
+      "step": 16397
+    },
+    {
+      "epoch": 4.353112969600425,
+      "grad_norm": 0.48754080848517856,
+      "learning_rate": 8.794615640542175e-07,
+      "loss": 0.5629,
+      "step": 16398
+    },
+    {
+      "epoch": 4.353378468073808,
+      "grad_norm": 0.48194049280322226,
+      "learning_rate": 8.791957339455445e-07,
+      "loss": 0.5473,
+      "step": 16399
+    },
+    {
+      "epoch": 4.353643966547192,
+      "grad_norm": 0.4893975002802356,
+      "learning_rate": 8.789299354465214e-07,
+      "loss": 0.5387,
+      "step": 16400
+    },
+    {
+      "epoch": 4.3539094650205765,
+      "grad_norm": 0.4787728565403449,
+      "learning_rate": 8.786641685623343e-07,
+      "loss": 0.5213,
+      "step": 16401
+    },
+    {
+      "epoch": 4.35417496349396,
+      "grad_norm": 0.47519259222751964,
+      "learning_rate": 8.783984332981649e-07,
+      "loss": 0.5224,
+      "step": 16402
+    },
+    {
+      "epoch": 4.354440461967344,
+      "grad_norm": 0.48565591029409966,
+      "learning_rate": 8.781327296591971e-07,
+      "loss": 0.5608,
+      "step": 16403
+    },
+    {
+      "epoch": 4.354705960440727,
+      "grad_norm": 0.471917558807958,
+      "learning_rate": 8.778670576506115e-07,
+      "loss": 0.5243,
+      "step": 16404
+    },
+    {
+      "epoch": 4.354971458914111,
+      "grad_norm": 0.47524778000913703,
+      "learning_rate": 8.776014172775907e-07,
+      "loss": 0.5198,
+      "step": 16405
+    },
+    {
+      "epoch": 4.355236957387495,
+      "grad_norm": 0.48780914634974865,
+      "learning_rate": 8.773358085453146e-07,
+      "loss": 0.5625,
+      "step": 16406
+    },
+    {
+      "epoch": 4.355502455860879,
+      "grad_norm": 0.4768870064046254,
+      "learning_rate": 8.770702314589624e-07,
+      "loss": 0.5034,
+      "step": 16407
+    },
+    {
+      "epoch": 4.3557679543342624,
+      "grad_norm": 0.4481681818768645,
+      "learning_rate": 8.768046860237151e-07,
+      "loss": 0.5147,
+      "step": 16408
+    },
+    {
+      "epoch": 4.3560334528076465,
+      "grad_norm": 0.4840034071226703,
+      "learning_rate": 8.765391722447497e-07,
+      "loss": 0.538,
+      "step": 16409
+    },
+    {
+      "epoch": 4.35629895128103,
+      "grad_norm": 0.4701280455740269,
+      "learning_rate": 8.762736901272462e-07,
+      "loss": 0.5079,
+      "step": 16410
+    },
+    {
+      "epoch": 4.356564449754414,
+      "grad_norm": 0.48662864316860477,
+      "learning_rate": 8.760082396763814e-07,
+      "loss": 0.5521,
+      "step": 16411
+    },
+    {
+      "epoch": 4.356829948227798,
+      "grad_norm": 0.4740126672256831,
+      "learning_rate": 8.757428208973315e-07,
+      "loss": 0.5712,
+      "step": 16412
+    },
+    {
+      "epoch": 4.357095446701181,
+      "grad_norm": 0.47819925643180794,
+      "learning_rate": 8.754774337952741e-07,
+      "loss": 0.5481,
+      "step": 16413
+    },
+    {
+      "epoch": 4.357360945174565,
+      "grad_norm": 0.47208656122415316,
+      "learning_rate": 8.752120783753834e-07,
+      "loss": 0.5171,
+      "step": 16414
+    },
+    {
+      "epoch": 4.357626443647949,
+      "grad_norm": 0.4730238657390892,
+      "learning_rate": 8.749467546428361e-07,
+      "loss": 0.5196,
+      "step": 16415
+    },
+    {
+      "epoch": 4.3578919421213325,
+      "grad_norm": 0.47602421248567667,
+      "learning_rate": 8.746814626028061e-07,
+      "loss": 0.5282,
+      "step": 16416
+    },
+    {
+      "epoch": 4.3581574405947165,
+      "grad_norm": 0.48629965541526915,
+      "learning_rate": 8.744162022604671e-07,
+      "loss": 0.5147,
+      "step": 16417
+    },
+    {
+      "epoch": 4.358422939068101,
+      "grad_norm": 0.4767117987420647,
+      "learning_rate": 8.741509736209914e-07,
+      "loss": 0.537,
+      "step": 16418
+    },
+    {
+      "epoch": 4.358688437541484,
+      "grad_norm": 0.5089207286591754,
+      "learning_rate": 8.738857766895531e-07,
+      "loss": 0.5862,
+      "step": 16419
+    },
+    {
+      "epoch": 4.358953936014868,
+      "grad_norm": 0.4650629990211462,
+      "learning_rate": 8.736206114713228e-07,
+      "loss": 0.5253,
+      "step": 16420
+    },
+    {
+      "epoch": 4.359219434488252,
+      "grad_norm": 0.46228703190055936,
+      "learning_rate": 8.733554779714735e-07,
+      "loss": 0.5157,
+      "step": 16421
+    },
+    {
+      "epoch": 4.359484932961635,
+      "grad_norm": 0.4856301585054266,
+      "learning_rate": 8.730903761951751e-07,
+      "loss": 0.5256,
+      "step": 16422
+    },
+    {
+      "epoch": 4.359750431435019,
+      "grad_norm": 0.4879196366426895,
+      "learning_rate": 8.728253061475969e-07,
+      "loss": 0.5563,
+      "step": 16423
+    },
+    {
+      "epoch": 4.360015929908403,
+      "grad_norm": 0.4934569647011793,
+      "learning_rate": 8.725602678339102e-07,
+      "loss": 0.5337,
+      "step": 16424
+    },
+    {
+      "epoch": 4.360281428381787,
+      "grad_norm": 0.4813837401848103,
+      "learning_rate": 8.722952612592819e-07,
+      "loss": 0.5497,
+      "step": 16425
+    },
+    {
+      "epoch": 4.360546926855171,
+      "grad_norm": 0.46357170336120795,
+      "learning_rate": 8.720302864288824e-07,
+      "loss": 0.4824,
+      "step": 16426
+    },
+    {
+      "epoch": 4.360812425328555,
+      "grad_norm": 0.4845317083872418,
+      "learning_rate": 8.717653433478771e-07,
+      "loss": 0.5295,
+      "step": 16427
+    },
+    {
+      "epoch": 4.361077923801938,
+      "grad_norm": 0.4831719751238828,
+      "learning_rate": 8.715004320214362e-07,
+      "loss": 0.5494,
+      "step": 16428
+    },
+    {
+      "epoch": 4.361343422275322,
+      "grad_norm": 0.47314500074569443,
+      "learning_rate": 8.712355524547222e-07,
+      "loss": 0.4967,
+      "step": 16429
+    },
+    {
+      "epoch": 4.361608920748706,
+      "grad_norm": 0.486665266853582,
+      "learning_rate": 8.709707046529029e-07,
+      "loss": 0.5471,
+      "step": 16430
+    },
+    {
+      "epoch": 4.361874419222089,
+      "grad_norm": 0.47753467377033604,
+      "learning_rate": 8.707058886211445e-07,
+      "loss": 0.5524,
+      "step": 16431
+    },
+    {
+      "epoch": 4.362139917695473,
+      "grad_norm": 0.47214797699595473,
+      "learning_rate": 8.704411043646093e-07,
+      "loss": 0.5028,
+      "step": 16432
+    },
+    {
+      "epoch": 4.362405416168857,
+      "grad_norm": 0.49130154937513226,
+      "learning_rate": 8.701763518884635e-07,
+      "loss": 0.5456,
+      "step": 16433
+    },
+    {
+      "epoch": 4.362670914642241,
+      "grad_norm": 0.4804179284476958,
+      "learning_rate": 8.699116311978692e-07,
+      "loss": 0.5318,
+      "step": 16434
+    },
+    {
+      "epoch": 4.362936413115625,
+      "grad_norm": 0.4859676460896525,
+      "learning_rate": 8.696469422979892e-07,
+      "loss": 0.541,
+      "step": 16435
+    },
+    {
+      "epoch": 4.363201911589009,
+      "grad_norm": 0.4686204562038585,
+      "learning_rate": 8.693822851939848e-07,
+      "loss": 0.5258,
+      "step": 16436
+    },
+    {
+      "epoch": 4.363467410062392,
+      "grad_norm": 0.5048712585747511,
+      "learning_rate": 8.691176598910192e-07,
+      "loss": 0.5633,
+      "step": 16437
+    },
+    {
+      "epoch": 4.363732908535776,
+      "grad_norm": 0.4900047480155294,
+      "learning_rate": 8.688530663942512e-07,
+      "loss": 0.5316,
+      "step": 16438
+    },
+    {
+      "epoch": 4.363998407009159,
+      "grad_norm": 0.4902297600612942,
+      "learning_rate": 8.685885047088433e-07,
+      "loss": 0.5542,
+      "step": 16439
+    },
+    {
+      "epoch": 4.364263905482543,
+      "grad_norm": 0.4688927481096436,
+      "learning_rate": 8.683239748399538e-07,
+      "loss": 0.521,
+      "step": 16440
+    },
+    {
+      "epoch": 4.3645294039559275,
+      "grad_norm": 0.4768434337273866,
+      "learning_rate": 8.680594767927411e-07,
+      "loss": 0.491,
+      "step": 16441
+    },
+    {
+      "epoch": 4.364794902429311,
+      "grad_norm": 0.47306849264282297,
+      "learning_rate": 8.677950105723651e-07,
+      "loss": 0.5528,
+      "step": 16442
+    },
+    {
+      "epoch": 4.365060400902695,
+      "grad_norm": 0.4836012128268889,
+      "learning_rate": 8.675305761839817e-07,
+      "loss": 0.56,
+      "step": 16443
+    },
+    {
+      "epoch": 4.365325899376079,
+      "grad_norm": 0.47115084782918304,
+      "learning_rate": 8.672661736327501e-07,
+      "loss": 0.4955,
+      "step": 16444
+    },
+    {
+      "epoch": 4.365591397849462,
+      "grad_norm": 0.47490498342574056,
+      "learning_rate": 8.670018029238256e-07,
+      "loss": 0.5381,
+      "step": 16445
+    },
+    {
+      "epoch": 4.365856896322846,
+      "grad_norm": 0.5028330251847454,
+      "learning_rate": 8.667374640623643e-07,
+      "loss": 0.5486,
+      "step": 16446
+    },
+    {
+      "epoch": 4.36612239479623,
+      "grad_norm": 0.45846120956587877,
+      "learning_rate": 8.664731570535203e-07,
+      "loss": 0.525,
+      "step": 16447
+    },
+    {
+      "epoch": 4.366387893269613,
+      "grad_norm": 0.48595478647348195,
+      "learning_rate": 8.662088819024506e-07,
+      "loss": 0.5381,
+      "step": 16448
+    },
+    {
+      "epoch": 4.3666533917429975,
+      "grad_norm": 0.48037713207581445,
+      "learning_rate": 8.659446386143067e-07,
+      "loss": 0.506,
+      "step": 16449
+    },
+    {
+      "epoch": 4.366918890216382,
+      "grad_norm": 0.4952378953379151,
+      "learning_rate": 8.656804271942435e-07,
+      "loss": 0.5437,
+      "step": 16450
+    },
+    {
+      "epoch": 4.367184388689765,
+      "grad_norm": 0.48205468594592643,
+      "learning_rate": 8.654162476474148e-07,
+      "loss": 0.4611,
+      "step": 16451
+    },
+    {
+      "epoch": 4.367449887163149,
+      "grad_norm": 0.4805859708436209,
+      "learning_rate": 8.651520999789697e-07,
+      "loss": 0.5111,
+      "step": 16452
+    },
+    {
+      "epoch": 4.367715385636533,
+      "grad_norm": 0.4713638114590739,
+      "learning_rate": 8.648879841940628e-07,
+      "loss": 0.5422,
+      "step": 16453
+    },
+    {
+      "epoch": 4.367980884109916,
+      "grad_norm": 0.4855901660460948,
+      "learning_rate": 8.646239002978423e-07,
+      "loss": 0.5352,
+      "step": 16454
+    },
+    {
+      "epoch": 4.3682463825833,
+      "grad_norm": 0.47968039391209205,
+      "learning_rate": 8.643598482954609e-07,
+      "loss": 0.5645,
+      "step": 16455
+    },
+    {
+      "epoch": 4.368511881056684,
+      "grad_norm": 0.47576115075520664,
+      "learning_rate": 8.640958281920664e-07,
+      "loss": 0.5002,
+      "step": 16456
+    },
+    {
+      "epoch": 4.3687773795300675,
+      "grad_norm": 0.4847981982072536,
+      "learning_rate": 8.638318399928094e-07,
+      "loss": 0.5331,
+      "step": 16457
+    },
+    {
+      "epoch": 4.369042878003452,
+      "grad_norm": 0.47499641056420255,
+      "learning_rate": 8.635678837028375e-07,
+      "loss": 0.5336,
+      "step": 16458
+    },
+    {
+      "epoch": 4.369308376476836,
+      "grad_norm": 0.46533369416006304,
+      "learning_rate": 8.633039593272979e-07,
+      "loss": 0.5132,
+      "step": 16459
+    },
+    {
+      "epoch": 4.369573874950219,
+      "grad_norm": 0.4493736123311235,
+      "learning_rate": 8.63040066871339e-07,
+      "loss": 0.4804,
+      "step": 16460
+    },
+    {
+      "epoch": 4.369839373423603,
+      "grad_norm": 0.4867906347780406,
+      "learning_rate": 8.62776206340106e-07,
+      "loss": 0.5451,
+      "step": 16461
+    },
+    {
+      "epoch": 4.370104871896986,
+      "grad_norm": 0.47350326379858393,
+      "learning_rate": 8.625123777387465e-07,
+      "loss": 0.576,
+      "step": 16462
+    },
+    {
+      "epoch": 4.37037037037037,
+      "grad_norm": 0.48038342508389703,
+      "learning_rate": 8.622485810724049e-07,
+      "loss": 0.5212,
+      "step": 16463
+    },
+    {
+      "epoch": 4.370635868843754,
+      "grad_norm": 0.4786057351633928,
+      "learning_rate": 8.619848163462258e-07,
+      "loss": 0.5585,
+      "step": 16464
+    },
+    {
+      "epoch": 4.3709013673171375,
+      "grad_norm": 0.47812931114272944,
+      "learning_rate": 8.617210835653525e-07,
+      "loss": 0.525,
+      "step": 16465
+    },
+    {
+      "epoch": 4.371166865790522,
+      "grad_norm": 0.47572591113148127,
+      "learning_rate": 8.614573827349304e-07,
+      "loss": 0.5579,
+      "step": 16466
+    },
+    {
+      "epoch": 4.371432364263906,
+      "grad_norm": 0.49049259906487147,
+      "learning_rate": 8.611937138601001e-07,
+      "loss": 0.5838,
+      "step": 16467
+    },
+    {
+      "epoch": 4.371697862737289,
+      "grad_norm": 0.49229789461599016,
+      "learning_rate": 8.609300769460055e-07,
+      "loss": 0.5412,
+      "step": 16468
+    },
+    {
+      "epoch": 4.371963361210673,
+      "grad_norm": 0.46771799767543853,
+      "learning_rate": 8.606664719977878e-07,
+      "loss": 0.5295,
+      "step": 16469
+    },
+    {
+      "epoch": 4.372228859684057,
+      "grad_norm": 0.47877790215944677,
+      "learning_rate": 8.604028990205868e-07,
+      "loss": 0.528,
+      "step": 16470
+    },
+    {
+      "epoch": 4.37249435815744,
+      "grad_norm": 0.4811853678439923,
+      "learning_rate": 8.601393580195444e-07,
+      "loss": 0.5264,
+      "step": 16471
+    },
+    {
+      "epoch": 4.372759856630824,
+      "grad_norm": 0.49228722926254653,
+      "learning_rate": 8.598758489997988e-07,
+      "loss": 0.5267,
+      "step": 16472
+    },
+    {
+      "epoch": 4.373025355104208,
+      "grad_norm": 0.4838105018200669,
+      "learning_rate": 8.596123719664909e-07,
+      "loss": 0.5644,
+      "step": 16473
+    },
+    {
+      "epoch": 4.373290853577592,
+      "grad_norm": 0.4842143312238847,
+      "learning_rate": 8.59348926924758e-07,
+      "loss": 0.5426,
+      "step": 16474
+    },
+    {
+      "epoch": 4.373556352050976,
+      "grad_norm": 0.4877996631690246,
+      "learning_rate": 8.590855138797371e-07,
+      "loss": 0.5584,
+      "step": 16475
+    },
+    {
+      "epoch": 4.37382185052436,
+      "grad_norm": 0.48886061495515487,
+      "learning_rate": 8.588221328365673e-07,
+      "loss": 0.5191,
+      "step": 16476
+    },
+    {
+      "epoch": 4.374087348997743,
+      "grad_norm": 0.4729444029136176,
+      "learning_rate": 8.585587838003834e-07,
+      "loss": 0.5702,
+      "step": 16477
+    },
+    {
+      "epoch": 4.374352847471127,
+      "grad_norm": 0.4728799075768128,
+      "learning_rate": 8.582954667763227e-07,
+      "loss": 0.5123,
+      "step": 16478
+    },
+    {
+      "epoch": 4.374618345944511,
+      "grad_norm": 0.4849476488238954,
+      "learning_rate": 8.580321817695195e-07,
+      "loss": 0.5636,
+      "step": 16479
+    },
+    {
+      "epoch": 4.374883844417894,
+      "grad_norm": 0.49145738686351426,
+      "learning_rate": 8.577689287851104e-07,
+      "loss": 0.5487,
+      "step": 16480
+    },
+    {
+      "epoch": 4.375149342891278,
+      "grad_norm": 0.4899937451678765,
+      "learning_rate": 8.575057078282265e-07,
+      "loss": 0.5299,
+      "step": 16481
+    },
+    {
+      "epoch": 4.3754148413646625,
+      "grad_norm": 0.4918975105621015,
+      "learning_rate": 8.572425189040037e-07,
+      "loss": 0.5626,
+      "step": 16482
+    },
+    {
+      "epoch": 4.375680339838046,
+      "grad_norm": 0.46645229417290207,
+      "learning_rate": 8.569793620175729e-07,
+      "loss": 0.5324,
+      "step": 16483
+    },
+    {
+      "epoch": 4.37594583831143,
+      "grad_norm": 0.4836505718989422,
+      "learning_rate": 8.567162371740684e-07,
+      "loss": 0.5528,
+      "step": 16484
+    },
+    {
+      "epoch": 4.376211336784814,
+      "grad_norm": 0.4657065342430494,
+      "learning_rate": 8.564531443786197e-07,
+      "loss": 0.5316,
+      "step": 16485
+    },
+    {
+      "epoch": 4.376476835258197,
+      "grad_norm": 0.4857349698654789,
+      "learning_rate": 8.561900836363598e-07,
+      "loss": 0.5705,
+      "step": 16486
+    },
+    {
+      "epoch": 4.376742333731581,
+      "grad_norm": 0.4630604878641226,
+      "learning_rate": 8.55927054952418e-07,
+      "loss": 0.5029,
+      "step": 16487
+    },
+    {
+      "epoch": 4.377007832204965,
+      "grad_norm": 0.4731217490369237,
+      "learning_rate": 8.556640583319231e-07,
+      "loss": 0.4904,
+      "step": 16488
+    },
+    {
+      "epoch": 4.3772733306783485,
+      "grad_norm": 0.4826008371106639,
+      "learning_rate": 8.554010937800059e-07,
+      "loss": 0.5454,
+      "step": 16489
+    },
+    {
+      "epoch": 4.3775388291517325,
+      "grad_norm": 0.5005054669993394,
+      "learning_rate": 8.551381613017934e-07,
+      "loss": 0.5451,
+      "step": 16490
+    },
+    {
+      "epoch": 4.377804327625116,
+      "grad_norm": 0.47129794505716416,
+      "learning_rate": 8.548752609024156e-07,
+      "loss": 0.5071,
+      "step": 16491
+    },
+    {
+      "epoch": 4.3780698260985,
+      "grad_norm": 0.4878527156392193,
+      "learning_rate": 8.546123925869965e-07,
+      "loss": 0.5903,
+      "step": 16492
+    },
+    {
+      "epoch": 4.378335324571884,
+      "grad_norm": 0.49120077183024247,
+      "learning_rate": 8.543495563606651e-07,
+      "loss": 0.5436,
+      "step": 16493
+    },
+    {
+      "epoch": 4.378600823045267,
+      "grad_norm": 0.477550403604748,
+      "learning_rate": 8.540867522285459e-07,
+      "loss": 0.5384,
+      "step": 16494
+    },
+    {
+      "epoch": 4.378866321518651,
+      "grad_norm": 0.47572948529315984,
+      "learning_rate": 8.538239801957657e-07,
+      "loss": 0.5424,
+      "step": 16495
+    },
+    {
+      "epoch": 4.379131819992035,
+      "grad_norm": 0.47630281357863713,
+      "learning_rate": 8.535612402674476e-07,
+      "loss": 0.5294,
+      "step": 16496
+    },
+    {
+      "epoch": 4.3793973184654185,
+      "grad_norm": 0.4920222106821019,
+      "learning_rate": 8.532985324487172e-07,
+      "loss": 0.5663,
+      "step": 16497
+    },
+    {
+      "epoch": 4.3796628169388025,
+      "grad_norm": 0.47665864220000415,
+      "learning_rate": 8.530358567446976e-07,
+      "loss": 0.5249,
+      "step": 16498
+    },
+    {
+      "epoch": 4.379928315412187,
+      "grad_norm": 0.4791771372540893,
+      "learning_rate": 8.527732131605101e-07,
+      "loss": 0.5543,
+      "step": 16499
+    },
+    {
+      "epoch": 4.38019381388557,
+      "grad_norm": 0.49087745991054277,
+      "learning_rate": 8.525106017012788e-07,
+      "loss": 0.5504,
+      "step": 16500
+    },
+    {
+      "epoch": 4.380459312358954,
+      "grad_norm": 0.4653373786361846,
+      "learning_rate": 8.522480223721241e-07,
+      "loss": 0.5215,
+      "step": 16501
+    },
+    {
+      "epoch": 4.380724810832338,
+      "grad_norm": 0.4730824873341873,
+      "learning_rate": 8.519854751781679e-07,
+      "loss": 0.5544,
+      "step": 16502
+    },
+    {
+      "epoch": 4.380990309305721,
+      "grad_norm": 0.4865970197321445,
+      "learning_rate": 8.517229601245299e-07,
+      "loss": 0.5477,
+      "step": 16503
+    },
+    {
+      "epoch": 4.381255807779105,
+      "grad_norm": 0.48863799887270964,
+      "learning_rate": 8.514604772163293e-07,
+      "loss": 0.5691,
+      "step": 16504
+    },
+    {
+      "epoch": 4.381521306252489,
+      "grad_norm": 0.4861374617992733,
+      "learning_rate": 8.511980264586861e-07,
+      "loss": 0.545,
+      "step": 16505
+    },
+    {
+      "epoch": 4.381786804725873,
+      "grad_norm": 0.4774383052610317,
+      "learning_rate": 8.50935607856718e-07,
+      "loss": 0.5583,
+      "step": 16506
+    },
+    {
+      "epoch": 4.382052303199257,
+      "grad_norm": 0.47449155817775995,
+      "learning_rate": 8.506732214155436e-07,
+      "loss": 0.5116,
+      "step": 16507
+    },
+    {
+      "epoch": 4.382317801672641,
+      "grad_norm": 0.4670936394096365,
+      "learning_rate": 8.504108671402791e-07,
+      "loss": 0.5557,
+      "step": 16508
+    },
+    {
+      "epoch": 4.382583300146024,
+      "grad_norm": 0.46826203165877606,
+      "learning_rate": 8.501485450360431e-07,
+      "loss": 0.5524,
+      "step": 16509
+    },
+    {
+      "epoch": 4.382848798619408,
+      "grad_norm": 0.47470717086633746,
+      "learning_rate": 8.498862551079484e-07,
+      "loss": 0.5171,
+      "step": 16510
+    },
+    {
+      "epoch": 4.383114297092792,
+      "grad_norm": 0.4863571357843977,
+      "learning_rate": 8.496239973611126e-07,
+      "loss": 0.5554,
+      "step": 16511
+    },
+    {
+      "epoch": 4.383379795566175,
+      "grad_norm": 0.46605684082027554,
+      "learning_rate": 8.493617718006489e-07,
+      "loss": 0.5274,
+      "step": 16512
+    },
+    {
+      "epoch": 4.383645294039559,
+      "grad_norm": 0.49102223487760194,
+      "learning_rate": 8.49099578431673e-07,
+      "loss": 0.5538,
+      "step": 16513
+    },
+    {
+      "epoch": 4.3839107925129435,
+      "grad_norm": 0.4824003622620835,
+      "learning_rate": 8.488374172592976e-07,
+      "loss": 0.5646,
+      "step": 16514
+    },
+    {
+      "epoch": 4.384176290986327,
+      "grad_norm": 0.480698107916931,
+      "learning_rate": 8.485752882886342e-07,
+      "loss": 0.5682,
+      "step": 16515
+    },
+    {
+      "epoch": 4.384441789459711,
+      "grad_norm": 0.47950961080315,
+      "learning_rate": 8.483131915247969e-07,
+      "loss": 0.5509,
+      "step": 16516
+    },
+    {
+      "epoch": 4.384707287933095,
+      "grad_norm": 0.4908380095054776,
+      "learning_rate": 8.480511269728955e-07,
+      "loss": 0.5466,
+      "step": 16517
+    },
+    {
+      "epoch": 4.384972786406478,
+      "grad_norm": 0.4869741368537473,
+      "learning_rate": 8.477890946380429e-07,
+      "loss": 0.5515,
+      "step": 16518
+    },
+    {
+      "epoch": 4.385238284879862,
+      "grad_norm": 0.4801131785219321,
+      "learning_rate": 8.475270945253472e-07,
+      "loss": 0.5232,
+      "step": 16519
+    },
+    {
+      "epoch": 4.385503783353245,
+      "grad_norm": 0.4697789475375817,
+      "learning_rate": 8.472651266399201e-07,
+      "loss": 0.5413,
+      "step": 16520
+    },
+    {
+      "epoch": 4.385769281826629,
+      "grad_norm": 0.47088206796034004,
+      "learning_rate": 8.470031909868695e-07,
+      "loss": 0.5385,
+      "step": 16521
+    },
+    {
+      "epoch": 4.3860347803000135,
+      "grad_norm": 0.4595517236539239,
+      "learning_rate": 8.467412875713032e-07,
+      "loss": 0.4956,
+      "step": 16522
+    },
+    {
+      "epoch": 4.386300278773397,
+      "grad_norm": 0.49018867475150263,
+      "learning_rate": 8.464794163983303e-07,
+      "loss": 0.5412,
+      "step": 16523
+    },
+    {
+      "epoch": 4.386565777246781,
+      "grad_norm": 0.47797847688737366,
+      "learning_rate": 8.462175774730566e-07,
+      "loss": 0.5342,
+      "step": 16524
+    },
+    {
+      "epoch": 4.386831275720165,
+      "grad_norm": 0.4720384625925957,
+      "learning_rate": 8.459557708005902e-07,
+      "loss": 0.5512,
+      "step": 16525
+    },
+    {
+      "epoch": 4.387096774193548,
+      "grad_norm": 0.47004813559713393,
+      "learning_rate": 8.456939963860358e-07,
+      "loss": 0.5348,
+      "step": 16526
+    },
+    {
+      "epoch": 4.387362272666932,
+      "grad_norm": 0.47210250729405256,
+      "learning_rate": 8.454322542344992e-07,
+      "loss": 0.548,
+      "step": 16527
+    },
+    {
+      "epoch": 4.387627771140316,
+      "grad_norm": 0.5069301901037149,
+      "learning_rate": 8.451705443510838e-07,
+      "loss": 0.555,
+      "step": 16528
+    },
+    {
+      "epoch": 4.387893269613699,
+      "grad_norm": 0.4795219091120522,
+      "learning_rate": 8.449088667408953e-07,
+      "loss": 0.5341,
+      "step": 16529
+    },
+    {
+      "epoch": 4.3881587680870835,
+      "grad_norm": 0.4545990003373758,
+      "learning_rate": 8.446472214090354e-07,
+      "loss": 0.5031,
+      "step": 16530
+    },
+    {
+      "epoch": 4.388424266560468,
+      "grad_norm": 0.47664165031644634,
+      "learning_rate": 8.443856083606086e-07,
+      "loss": 0.538,
+      "step": 16531
+    },
+    {
+      "epoch": 4.388689765033851,
+      "grad_norm": 0.47883239273265965,
+      "learning_rate": 8.441240276007162e-07,
+      "loss": 0.5349,
+      "step": 16532
+    },
+    {
+      "epoch": 4.388955263507235,
+      "grad_norm": 0.483734129951063,
+      "learning_rate": 8.438624791344585e-07,
+      "loss": 0.5265,
+      "step": 16533
+    },
+    {
+      "epoch": 4.389220761980619,
+      "grad_norm": 0.4827238771762319,
+      "learning_rate": 8.436009629669384e-07,
+      "loss": 0.5664,
+      "step": 16534
+    },
+    {
+      "epoch": 4.389486260454002,
+      "grad_norm": 0.49356122586816153,
+      "learning_rate": 8.43339479103254e-07,
+      "loss": 0.4943,
+      "step": 16535
+    },
+    {
+      "epoch": 4.389751758927386,
+      "grad_norm": 0.479835733205734,
+      "learning_rate": 8.43078027548507e-07,
+      "loss": 0.5453,
+      "step": 16536
+    },
+    {
+      "epoch": 4.39001725740077,
+      "grad_norm": 0.46247650845404736,
+      "learning_rate": 8.428166083077952e-07,
+      "loss": 0.5061,
+      "step": 16537
+    },
+    {
+      "epoch": 4.3902827558741535,
+      "grad_norm": 0.46944378309625523,
+      "learning_rate": 8.42555221386217e-07,
+      "loss": 0.5194,
+      "step": 16538
+    },
+    {
+      "epoch": 4.390548254347538,
+      "grad_norm": 0.49150570669490645,
+      "learning_rate": 8.422938667888695e-07,
+      "loss": 0.5339,
+      "step": 16539
+    },
+    {
+      "epoch": 4.390813752820922,
+      "grad_norm": 0.4788798543853999,
+      "learning_rate": 8.420325445208508e-07,
+      "loss": 0.5458,
+      "step": 16540
+    },
+    {
+      "epoch": 4.391079251294305,
+      "grad_norm": 0.4843552123459841,
+      "learning_rate": 8.417712545872564e-07,
+      "loss": 0.5637,
+      "step": 16541
+    },
+    {
+      "epoch": 4.391344749767689,
+      "grad_norm": 0.4811327189883475,
+      "learning_rate": 8.415099969931823e-07,
+      "loss": 0.551,
+      "step": 16542
+    },
+    {
+      "epoch": 4.391610248241072,
+      "grad_norm": 0.48341769739296103,
+      "learning_rate": 8.412487717437259e-07,
+      "loss": 0.5119,
+      "step": 16543
+    },
+    {
+      "epoch": 4.391875746714456,
+      "grad_norm": 0.46833997994923415,
+      "learning_rate": 8.409875788439778e-07,
+      "loss": 0.5251,
+      "step": 16544
+    },
+    {
+      "epoch": 4.39214124518784,
+      "grad_norm": 0.4823732669104166,
+      "learning_rate": 8.407264182990349e-07,
+      "loss": 0.5425,
+      "step": 16545
+    },
+    {
+      "epoch": 4.392406743661224,
+      "grad_norm": 0.48815187719077285,
+      "learning_rate": 8.404652901139887e-07,
+      "loss": 0.5665,
+      "step": 16546
+    },
+    {
+      "epoch": 4.392672242134608,
+      "grad_norm": 0.47635689641019785,
+      "learning_rate": 8.402041942939332e-07,
+      "loss": 0.5238,
+      "step": 16547
+    },
+    {
+      "epoch": 4.392937740607992,
+      "grad_norm": 0.4778722834037993,
+      "learning_rate": 8.399431308439592e-07,
+      "loss": 0.5448,
+      "step": 16548
+    },
+    {
+      "epoch": 4.393203239081375,
+      "grad_norm": 0.4730014547639294,
+      "learning_rate": 8.396820997691593e-07,
+      "loss": 0.5332,
+      "step": 16549
+    },
+    {
+      "epoch": 4.393468737554759,
+      "grad_norm": 0.48587683855389313,
+      "learning_rate": 8.394211010746237e-07,
+      "loss": 0.56,
+      "step": 16550
+    },
+    {
+      "epoch": 4.393734236028143,
+      "grad_norm": 0.4900715009798407,
+      "learning_rate": 8.391601347654415e-07,
+      "loss": 0.5293,
+      "step": 16551
+    },
+    {
+      "epoch": 4.393999734501526,
+      "grad_norm": 0.48409093114950374,
+      "learning_rate": 8.388992008467043e-07,
+      "loss": 0.5858,
+      "step": 16552
+    },
+    {
+      "epoch": 4.39426523297491,
+      "grad_norm": 0.4716282235252544,
+      "learning_rate": 8.386382993234984e-07,
+      "loss": 0.518,
+      "step": 16553
+    },
+    {
+      "epoch": 4.394530731448294,
+      "grad_norm": 0.47956117951526955,
+      "learning_rate": 8.383774302009146e-07,
+      "loss": 0.549,
+      "step": 16554
+    },
+    {
+      "epoch": 4.394796229921678,
+      "grad_norm": 0.4848508500680948,
+      "learning_rate": 8.381165934840391e-07,
+      "loss": 0.5596,
+      "step": 16555
+    },
+    {
+      "epoch": 4.395061728395062,
+      "grad_norm": 0.4712953426031584,
+      "learning_rate": 8.37855789177959e-07,
+      "loss": 0.5455,
+      "step": 16556
+    },
+    {
+      "epoch": 4.395327226868446,
+      "grad_norm": 0.49910339687729405,
+      "learning_rate": 8.375950172877598e-07,
+      "loss": 0.543,
+      "step": 16557
+    },
+    {
+      "epoch": 4.395592725341829,
+      "grad_norm": 0.4689768542396897,
+      "learning_rate": 8.37334277818529e-07,
+      "loss": 0.5414,
+      "step": 16558
+    },
+    {
+      "epoch": 4.395858223815213,
+      "grad_norm": 0.4790907031745384,
+      "learning_rate": 8.370735707753497e-07,
+      "loss": 0.5292,
+      "step": 16559
+    },
+    {
+      "epoch": 4.396123722288597,
+      "grad_norm": 0.4684591991960239,
+      "learning_rate": 8.368128961633082e-07,
+      "loss": 0.5045,
+      "step": 16560
+    },
+    {
+      "epoch": 4.39638922076198,
+      "grad_norm": 0.4843910607353818,
+      "learning_rate": 8.365522539874874e-07,
+      "loss": 0.5345,
+      "step": 16561
+    },
+    {
+      "epoch": 4.396654719235364,
+      "grad_norm": 0.4814716658497471,
+      "learning_rate": 8.362916442529698e-07,
+      "loss": 0.5484,
+      "step": 16562
+    },
+    {
+      "epoch": 4.3969202177087485,
+      "grad_norm": 0.4808893516028332,
+      "learning_rate": 8.360310669648392e-07,
+      "loss": 0.5268,
+      "step": 16563
+    },
+    {
+      "epoch": 4.397185716182132,
+      "grad_norm": 0.488372798884931,
+      "learning_rate": 8.357705221281762e-07,
+      "loss": 0.506,
+      "step": 16564
+    },
+    {
+      "epoch": 4.397451214655516,
+      "grad_norm": 0.4540464216553004,
+      "learning_rate": 8.355100097480634e-07,
+      "loss": 0.5205,
+      "step": 16565
+    },
+    {
+      "epoch": 4.3977167131289,
+      "grad_norm": 0.48227418497272384,
+      "learning_rate": 8.35249529829581e-07,
+      "loss": 0.5328,
+      "step": 16566
+    },
+    {
+      "epoch": 4.397982211602283,
+      "grad_norm": 0.4991728818605788,
+      "learning_rate": 8.349890823778079e-07,
+      "loss": 0.5416,
+      "step": 16567
+    },
+    {
+      "epoch": 4.398247710075667,
+      "grad_norm": 0.4836881787915766,
+      "learning_rate": 8.347286673978253e-07,
+      "loss": 0.5436,
+      "step": 16568
+    },
+    {
+      "epoch": 4.398513208549051,
+      "grad_norm": 0.47149152568632324,
+      "learning_rate": 8.344682848947099e-07,
+      "loss": 0.5045,
+      "step": 16569
+    },
+    {
+      "epoch": 4.3987787070224345,
+      "grad_norm": 0.47823935343197504,
+      "learning_rate": 8.342079348735419e-07,
+      "loss": 0.5632,
+      "step": 16570
+    },
+    {
+      "epoch": 4.3990442054958185,
+      "grad_norm": 0.49255780728054716,
+      "learning_rate": 8.339476173393968e-07,
+      "loss": 0.5487,
+      "step": 16571
+    },
+    {
+      "epoch": 4.399309703969202,
+      "grad_norm": 0.48310629352880186,
+      "learning_rate": 8.336873322973537e-07,
+      "loss": 0.5452,
+      "step": 16572
+    },
+    {
+      "epoch": 4.399575202442586,
+      "grad_norm": 0.47366107896474235,
+      "learning_rate": 8.334270797524863e-07,
+      "loss": 0.5097,
+      "step": 16573
+    },
+    {
+      "epoch": 4.39984070091597,
+      "grad_norm": 0.49603665753956616,
+      "learning_rate": 8.331668597098719e-07,
+      "loss": 0.5408,
+      "step": 16574
+    },
+    {
+      "epoch": 4.400106199389353,
+      "grad_norm": 0.4861467766323368,
+      "learning_rate": 8.32906672174584e-07,
+      "loss": 0.5325,
+      "step": 16575
+    },
+    {
+      "epoch": 4.400371697862737,
+      "grad_norm": 0.49185672269455355,
+      "learning_rate": 8.326465171516987e-07,
+      "loss": 0.5412,
+      "step": 16576
+    },
+    {
+      "epoch": 4.400637196336121,
+      "grad_norm": 0.5036226909742733,
+      "learning_rate": 8.323863946462879e-07,
+      "loss": 0.5699,
+      "step": 16577
+    },
+    {
+      "epoch": 4.4009026948095045,
+      "grad_norm": 0.4934988122715888,
+      "learning_rate": 8.321263046634262e-07,
+      "loss": 0.5379,
+      "step": 16578
+    },
+    {
+      "epoch": 4.4011681932828886,
+      "grad_norm": 0.48140185505303684,
+      "learning_rate": 8.318662472081851e-07,
+      "loss": 0.5558,
+      "step": 16579
+    },
+    {
+      "epoch": 4.401433691756273,
+      "grad_norm": 0.4901830703409975,
+      "learning_rate": 8.316062222856358e-07,
+      "loss": 0.5225,
+      "step": 16580
+    },
+    {
+      "epoch": 4.401699190229656,
+      "grad_norm": 0.4847336852715075,
+      "learning_rate": 8.31346229900851e-07,
+      "loss": 0.5577,
+      "step": 16581
+    },
+    {
+      "epoch": 4.40196468870304,
+      "grad_norm": 0.47053959647474014,
+      "learning_rate": 8.310862700588992e-07,
+      "loss": 0.5241,
+      "step": 16582
+    },
+    {
+      "epoch": 4.402230187176424,
+      "grad_norm": 0.476964339308851,
+      "learning_rate": 8.308263427648522e-07,
+      "loss": 0.5296,
+      "step": 16583
+    },
+    {
+      "epoch": 4.402495685649807,
+      "grad_norm": 0.47501458555709297,
+      "learning_rate": 8.305664480237785e-07,
+      "loss": 0.5609,
+      "step": 16584
+    },
+    {
+      "epoch": 4.402761184123191,
+      "grad_norm": 0.4771999506840208,
+      "learning_rate": 8.303065858407467e-07,
+      "loss": 0.5334,
+      "step": 16585
+    },
+    {
+      "epoch": 4.403026682596575,
+      "grad_norm": 0.48438442130647985,
+      "learning_rate": 8.300467562208234e-07,
+      "loss": 0.5355,
+      "step": 16586
+    },
+    {
+      "epoch": 4.403292181069959,
+      "grad_norm": 0.4725408041333438,
+      "learning_rate": 8.297869591690771e-07,
+      "loss": 0.5349,
+      "step": 16587
+    },
+    {
+      "epoch": 4.403557679543343,
+      "grad_norm": 0.4832178313289328,
+      "learning_rate": 8.295271946905756e-07,
+      "loss": 0.5496,
+      "step": 16588
+    },
+    {
+      "epoch": 4.403823178016727,
+      "grad_norm": 0.48884472571661386,
+      "learning_rate": 8.292674627903838e-07,
+      "loss": 0.5517,
+      "step": 16589
+    },
+    {
+      "epoch": 4.40408867649011,
+      "grad_norm": 0.48661902335888485,
+      "learning_rate": 8.290077634735668e-07,
+      "loss": 0.5703,
+      "step": 16590
+    },
+    {
+      "epoch": 4.404354174963494,
+      "grad_norm": 0.4989595638043249,
+      "learning_rate": 8.287480967451894e-07,
+      "loss": 0.5703,
+      "step": 16591
+    },
+    {
+      "epoch": 4.404619673436878,
+      "grad_norm": 0.48283408291535,
+      "learning_rate": 8.284884626103165e-07,
+      "loss": 0.5549,
+      "step": 16592
+    },
+    {
+      "epoch": 4.404885171910261,
+      "grad_norm": 0.4769709345382824,
+      "learning_rate": 8.282288610740102e-07,
+      "loss": 0.5384,
+      "step": 16593
+    },
+    {
+      "epoch": 4.405150670383645,
+      "grad_norm": 0.4903974098635621,
+      "learning_rate": 8.279692921413354e-07,
+      "loss": 0.5535,
+      "step": 16594
+    },
+    {
+      "epoch": 4.4054161688570295,
+      "grad_norm": 0.4836010580018324,
+      "learning_rate": 8.277097558173533e-07,
+      "loss": 0.5474,
+      "step": 16595
+    },
+    {
+      "epoch": 4.405681667330413,
+      "grad_norm": 0.46643252334484436,
+      "learning_rate": 8.274502521071246e-07,
+      "loss": 0.5325,
+      "step": 16596
+    },
+    {
+      "epoch": 4.405947165803797,
+      "grad_norm": 0.48157133937959873,
+      "learning_rate": 8.271907810157117e-07,
+      "loss": 0.5588,
+      "step": 16597
+    },
+    {
+      "epoch": 4.406212664277181,
+      "grad_norm": 0.4774852041615962,
+      "learning_rate": 8.269313425481737e-07,
+      "loss": 0.5346,
+      "step": 16598
+    },
+    {
+      "epoch": 4.406478162750564,
+      "grad_norm": 0.47200433828200533,
+      "learning_rate": 8.266719367095716e-07,
+      "loss": 0.5246,
+      "step": 16599
+    },
+    {
+      "epoch": 4.406743661223948,
+      "grad_norm": 0.48085196306466316,
+      "learning_rate": 8.264125635049627e-07,
+      "loss": 0.5426,
+      "step": 16600
+    },
+    {
+      "epoch": 4.407009159697331,
+      "grad_norm": 0.48506682186569927,
+      "learning_rate": 8.261532229394084e-07,
+      "loss": 0.5335,
+      "step": 16601
+    },
+    {
+      "epoch": 4.407274658170715,
+      "grad_norm": 0.47262651255568006,
+      "learning_rate": 8.258939150179629e-07,
+      "loss": 0.5319,
+      "step": 16602
+    },
+    {
+      "epoch": 4.4075401566440995,
+      "grad_norm": 0.4838610920771168,
+      "learning_rate": 8.256346397456857e-07,
+      "loss": 0.5578,
+      "step": 16603
+    },
+    {
+      "epoch": 4.407805655117483,
+      "grad_norm": 0.47495246358248394,
+      "learning_rate": 8.253753971276319e-07,
+      "loss": 0.5452,
+      "step": 16604
+    },
+    {
+      "epoch": 4.408071153590867,
+      "grad_norm": 0.4620917749744837,
+      "learning_rate": 8.251161871688587e-07,
+      "loss": 0.4971,
+      "step": 16605
+    },
+    {
+      "epoch": 4.408336652064251,
+      "grad_norm": 0.46823420460255805,
+      "learning_rate": 8.248570098744199e-07,
+      "loss": 0.5385,
+      "step": 16606
+    },
+    {
+      "epoch": 4.408602150537634,
+      "grad_norm": 0.49352096457911426,
+      "learning_rate": 8.245978652493717e-07,
+      "loss": 0.5333,
+      "step": 16607
+    },
+    {
+      "epoch": 4.408867649011018,
+      "grad_norm": 0.4806888410227712,
+      "learning_rate": 8.24338753298767e-07,
+      "loss": 0.5265,
+      "step": 16608
+    },
+    {
+      "epoch": 4.409133147484402,
+      "grad_norm": 0.4795033100910115,
+      "learning_rate": 8.240796740276588e-07,
+      "loss": 0.5471,
+      "step": 16609
+    },
+    {
+      "epoch": 4.409398645957785,
+      "grad_norm": 0.4866614643725927,
+      "learning_rate": 8.238206274411012e-07,
+      "loss": 0.5157,
+      "step": 16610
+    },
+    {
+      "epoch": 4.4096641444311695,
+      "grad_norm": 0.4877475479925792,
+      "learning_rate": 8.235616135441444e-07,
+      "loss": 0.5654,
+      "step": 16611
+    },
+    {
+      "epoch": 4.409929642904554,
+      "grad_norm": 0.48497645434175274,
+      "learning_rate": 8.233026323418417e-07,
+      "loss": 0.5435,
+      "step": 16612
+    },
+    {
+      "epoch": 4.410195141377937,
+      "grad_norm": 0.4791618025070629,
+      "learning_rate": 8.230436838392427e-07,
+      "loss": 0.5564,
+      "step": 16613
+    },
+    {
+      "epoch": 4.410460639851321,
+      "grad_norm": 0.4753206556623597,
+      "learning_rate": 8.227847680413969e-07,
+      "loss": 0.5151,
+      "step": 16614
+    },
+    {
+      "epoch": 4.410726138324705,
+      "grad_norm": 0.49408224235282816,
+      "learning_rate": 8.225258849533557e-07,
+      "loss": 0.5316,
+      "step": 16615
+    },
+    {
+      "epoch": 4.410991636798088,
+      "grad_norm": 0.4809189150117858,
+      "learning_rate": 8.222670345801661e-07,
+      "loss": 0.5672,
+      "step": 16616
+    },
+    {
+      "epoch": 4.411257135271472,
+      "grad_norm": 0.47512747077281287,
+      "learning_rate": 8.220082169268778e-07,
+      "loss": 0.576,
+      "step": 16617
+    },
+    {
+      "epoch": 4.411522633744856,
+      "grad_norm": 0.47325840946973086,
+      "learning_rate": 8.217494319985378e-07,
+      "loss": 0.53,
+      "step": 16618
+    },
+    {
+      "epoch": 4.4117881322182395,
+      "grad_norm": 0.4634513071347273,
+      "learning_rate": 8.214906798001928e-07,
+      "loss": 0.5648,
+      "step": 16619
+    },
+    {
+      "epoch": 4.412053630691624,
+      "grad_norm": 0.47887318760265474,
+      "learning_rate": 8.212319603368887e-07,
+      "loss": 0.5055,
+      "step": 16620
+    },
+    {
+      "epoch": 4.412319129165008,
+      "grad_norm": 0.4818422253452254,
+      "learning_rate": 8.209732736136725e-07,
+      "loss": 0.5506,
+      "step": 16621
+    },
+    {
+      "epoch": 4.412584627638391,
+      "grad_norm": 0.4933637411606336,
+      "learning_rate": 8.207146196355873e-07,
+      "loss": 0.5655,
+      "step": 16622
+    },
+    {
+      "epoch": 4.412850126111775,
+      "grad_norm": 0.47212837137853614,
+      "learning_rate": 8.204559984076796e-07,
+      "loss": 0.5542,
+      "step": 16623
+    },
+    {
+      "epoch": 4.413115624585159,
+      "grad_norm": 0.4770536731793999,
+      "learning_rate": 8.201974099349922e-07,
+      "loss": 0.5435,
+      "step": 16624
+    },
+    {
+      "epoch": 4.413381123058542,
+      "grad_norm": 0.47370131246906244,
+      "learning_rate": 8.199388542225672e-07,
+      "loss": 0.4914,
+      "step": 16625
+    },
+    {
+      "epoch": 4.413646621531926,
+      "grad_norm": 0.4788157878171548,
+      "learning_rate": 8.196803312754493e-07,
+      "loss": 0.5617,
+      "step": 16626
+    },
+    {
+      "epoch": 4.41391212000531,
+      "grad_norm": 0.47935543589093,
+      "learning_rate": 8.194218410986779e-07,
+      "loss": 0.5757,
+      "step": 16627
+    },
+    {
+      "epoch": 4.414177618478694,
+      "grad_norm": 0.49523343750168153,
+      "learning_rate": 8.191633836972962e-07,
+      "loss": 0.5256,
+      "step": 16628
+    },
+    {
+      "epoch": 4.414443116952078,
+      "grad_norm": 0.47880038529065844,
+      "learning_rate": 8.189049590763432e-07,
+      "loss": 0.5124,
+      "step": 16629
+    },
+    {
+      "epoch": 4.414708615425461,
+      "grad_norm": 0.47265548806566515,
+      "learning_rate": 8.186465672408608e-07,
+      "loss": 0.5396,
+      "step": 16630
+    },
+    {
+      "epoch": 4.414974113898845,
+      "grad_norm": 0.48925554747587374,
+      "learning_rate": 8.183882081958858e-07,
+      "loss": 0.5599,
+      "step": 16631
+    },
+    {
+      "epoch": 4.415239612372229,
+      "grad_norm": 0.4904556805467935,
+      "learning_rate": 8.181298819464578e-07,
+      "loss": 0.527,
+      "step": 16632
+    },
+    {
+      "epoch": 4.415505110845612,
+      "grad_norm": 0.48415443548359194,
+      "learning_rate": 8.17871588497616e-07,
+      "loss": 0.5239,
+      "step": 16633
+    },
+    {
+      "epoch": 4.415770609318996,
+      "grad_norm": 0.49475954637904473,
+      "learning_rate": 8.17613327854396e-07,
+      "loss": 0.5327,
+      "step": 16634
+    },
+    {
+      "epoch": 4.41603610779238,
+      "grad_norm": 0.4679431339657498,
+      "learning_rate": 8.173551000218369e-07,
+      "loss": 0.5072,
+      "step": 16635
+    },
+    {
+      "epoch": 4.416301606265764,
+      "grad_norm": 0.4980948548298773,
+      "learning_rate": 8.170969050049712e-07,
+      "loss": 0.5219,
+      "step": 16636
+    },
+    {
+      "epoch": 4.416567104739148,
+      "grad_norm": 0.4929770917999851,
+      "learning_rate": 8.168387428088376e-07,
+      "loss": 0.5539,
+      "step": 16637
+    },
+    {
+      "epoch": 4.416832603212532,
+      "grad_norm": 0.48826704303946095,
+      "learning_rate": 8.165806134384685e-07,
+      "loss": 0.5229,
+      "step": 16638
+    },
+    {
+      "epoch": 4.417098101685915,
+      "grad_norm": 0.4948866613781495,
+      "learning_rate": 8.163225168989e-07,
+      "loss": 0.5654,
+      "step": 16639
+    },
+    {
+      "epoch": 4.417363600159299,
+      "grad_norm": 0.48333305550936245,
+      "learning_rate": 8.160644531951642e-07,
+      "loss": 0.5702,
+      "step": 16640
+    },
+    {
+      "epoch": 4.417629098632683,
+      "grad_norm": 0.4922566496502325,
+      "learning_rate": 8.15806422332295e-07,
+      "loss": 0.5148,
+      "step": 16641
+    },
+    {
+      "epoch": 4.417894597106066,
+      "grad_norm": 0.488723838977445,
+      "learning_rate": 8.155484243153244e-07,
+      "loss": 0.51,
+      "step": 16642
+    },
+    {
+      "epoch": 4.41816009557945,
+      "grad_norm": 0.4838047773076337,
+      "learning_rate": 8.152904591492831e-07,
+      "loss": 0.5429,
+      "step": 16643
+    },
+    {
+      "epoch": 4.4184255940528345,
+      "grad_norm": 0.49117705174130283,
+      "learning_rate": 8.150325268392035e-07,
+      "loss": 0.5272,
+      "step": 16644
+    },
+    {
+      "epoch": 4.418691092526218,
+      "grad_norm": 0.49278922379736834,
+      "learning_rate": 8.147746273901138e-07,
+      "loss": 0.5288,
+      "step": 16645
+    },
+    {
+      "epoch": 4.418956590999602,
+      "grad_norm": 0.4987614344238964,
+      "learning_rate": 8.145167608070464e-07,
+      "loss": 0.5095,
+      "step": 16646
+    },
+    {
+      "epoch": 4.419222089472986,
+      "grad_norm": 0.48822431453689114,
+      "learning_rate": 8.142589270950288e-07,
+      "loss": 0.5324,
+      "step": 16647
+    },
+    {
+      "epoch": 4.419487587946369,
+      "grad_norm": 0.4881151593755466,
+      "learning_rate": 8.140011262590894e-07,
+      "loss": 0.5671,
+      "step": 16648
+    },
+    {
+      "epoch": 4.419753086419753,
+      "grad_norm": 0.4958922765345893,
+      "learning_rate": 8.137433583042553e-07,
+      "loss": 0.5493,
+      "step": 16649
+    },
+    {
+      "epoch": 4.420018584893137,
+      "grad_norm": 0.4828907017588553,
+      "learning_rate": 8.134856232355554e-07,
+      "loss": 0.5523,
+      "step": 16650
+    },
+    {
+      "epoch": 4.4202840833665205,
+      "grad_norm": 0.5013118522835209,
+      "learning_rate": 8.132279210580141e-07,
+      "loss": 0.5141,
+      "step": 16651
+    },
+    {
+      "epoch": 4.4205495818399045,
+      "grad_norm": 0.4750123617068865,
+      "learning_rate": 8.129702517766591e-07,
+      "loss": 0.5629,
+      "step": 16652
+    },
+    {
+      "epoch": 4.420815080313288,
+      "grad_norm": 0.4824018506989978,
+      "learning_rate": 8.127126153965148e-07,
+      "loss": 0.5649,
+      "step": 16653
+    },
+    {
+      "epoch": 4.421080578786672,
+      "grad_norm": 0.4719677230021761,
+      "learning_rate": 8.124550119226049e-07,
+      "loss": 0.5629,
+      "step": 16654
+    },
+    {
+      "epoch": 4.421346077260056,
+      "grad_norm": 0.4742841990041128,
+      "learning_rate": 8.121974413599549e-07,
+      "loss": 0.5243,
+      "step": 16655
+    },
+    {
+      "epoch": 4.42161157573344,
+      "grad_norm": 0.4846558428319592,
+      "learning_rate": 8.119399037135864e-07,
+      "loss": 0.5375,
+      "step": 16656
+    },
+    {
+      "epoch": 4.421877074206823,
+      "grad_norm": 0.48948565814441936,
+      "learning_rate": 8.116823989885236e-07,
+      "loss": 0.5732,
+      "step": 16657
+    },
+    {
+      "epoch": 4.422142572680207,
+      "grad_norm": 0.479782947037128,
+      "learning_rate": 8.114249271897881e-07,
+      "loss": 0.5447,
+      "step": 16658
+    },
+    {
+      "epoch": 4.4224080711535905,
+      "grad_norm": 0.48622118740659476,
+      "learning_rate": 8.111674883223997e-07,
+      "loss": 0.562,
+      "step": 16659
+    },
+    {
+      "epoch": 4.4226735696269746,
+      "grad_norm": 0.48554013848007177,
+      "learning_rate": 8.109100823913813e-07,
+      "loss": 0.5512,
+      "step": 16660
+    },
+    {
+      "epoch": 4.422939068100359,
+      "grad_norm": 0.4786866890688573,
+      "learning_rate": 8.106527094017508e-07,
+      "loss": 0.5319,
+      "step": 16661
+    },
+    {
+      "epoch": 4.423204566573742,
+      "grad_norm": 0.4831168918515697,
+      "learning_rate": 8.103953693585295e-07,
+      "loss": 0.5585,
+      "step": 16662
+    },
+    {
+      "epoch": 4.423470065047126,
+      "grad_norm": 0.4701209835145499,
+      "learning_rate": 8.101380622667349e-07,
+      "loss": 0.5144,
+      "step": 16663
+    },
+    {
+      "epoch": 4.42373556352051,
+      "grad_norm": 0.4697152135209769,
+      "learning_rate": 8.098807881313869e-07,
+      "loss": 0.5394,
+      "step": 16664
+    },
+    {
+      "epoch": 4.424001061993893,
+      "grad_norm": 0.4776201155560845,
+      "learning_rate": 8.096235469575e-07,
+      "loss": 0.5445,
+      "step": 16665
+    },
+    {
+      "epoch": 4.424266560467277,
+      "grad_norm": 0.48887263534609116,
+      "learning_rate": 8.093663387500938e-07,
+      "loss": 0.5518,
+      "step": 16666
+    },
+    {
+      "epoch": 4.424532058940661,
+      "grad_norm": 0.4665235798745896,
+      "learning_rate": 8.091091635141824e-07,
+      "loss": 0.5822,
+      "step": 16667
+    },
+    {
+      "epoch": 4.424797557414045,
+      "grad_norm": 0.48493862822043754,
+      "learning_rate": 8.088520212547832e-07,
+      "loss": 0.5015,
+      "step": 16668
+    },
+    {
+      "epoch": 4.425063055887429,
+      "grad_norm": 0.4718393227144154,
+      "learning_rate": 8.085949119769093e-07,
+      "loss": 0.531,
+      "step": 16669
+    },
+    {
+      "epoch": 4.425328554360813,
+      "grad_norm": 0.480798850924045,
+      "learning_rate": 8.083378356855767e-07,
+      "loss": 0.5689,
+      "step": 16670
+    },
+    {
+      "epoch": 4.425594052834196,
+      "grad_norm": 0.4698678135095887,
+      "learning_rate": 8.080807923857983e-07,
+      "loss": 0.5409,
+      "step": 16671
+    },
+    {
+      "epoch": 4.42585955130758,
+      "grad_norm": 0.47302478007096793,
+      "learning_rate": 8.078237820825862e-07,
+      "loss": 0.5207,
+      "step": 16672
+    },
+    {
+      "epoch": 4.426125049780964,
+      "grad_norm": 0.47353423074426676,
+      "learning_rate": 8.075668047809543e-07,
+      "loss": 0.5701,
+      "step": 16673
+    },
+    {
+      "epoch": 4.426390548254347,
+      "grad_norm": 0.476212131602487,
+      "learning_rate": 8.073098604859125e-07,
+      "loss": 0.5246,
+      "step": 16674
+    },
+    {
+      "epoch": 4.426656046727731,
+      "grad_norm": 0.4897021972643576,
+      "learning_rate": 8.07052949202474e-07,
+      "loss": 0.5205,
+      "step": 16675
+    },
+    {
+      "epoch": 4.4269215452011155,
+      "grad_norm": 0.4837969089029706,
+      "learning_rate": 8.067960709356479e-07,
+      "loss": 0.5441,
+      "step": 16676
+    },
+    {
+      "epoch": 4.427187043674499,
+      "grad_norm": 0.4788561982980135,
+      "learning_rate": 8.065392256904439e-07,
+      "loss": 0.5579,
+      "step": 16677
+    },
+    {
+      "epoch": 4.427452542147883,
+      "grad_norm": 0.4891693595594505,
+      "learning_rate": 8.062824134718706e-07,
+      "loss": 0.5647,
+      "step": 16678
+    },
+    {
+      "epoch": 4.427718040621267,
+      "grad_norm": 0.46813081890134806,
+      "learning_rate": 8.060256342849373e-07,
+      "loss": 0.5412,
+      "step": 16679
+    },
+    {
+      "epoch": 4.42798353909465,
+      "grad_norm": 0.4792719926285562,
+      "learning_rate": 8.057688881346523e-07,
+      "loss": 0.5505,
+      "step": 16680
+    },
+    {
+      "epoch": 4.428249037568034,
+      "grad_norm": 0.4661606059578263,
+      "learning_rate": 8.055121750260215e-07,
+      "loss": 0.5292,
+      "step": 16681
+    },
+    {
+      "epoch": 4.428514536041417,
+      "grad_norm": 0.48008116629997893,
+      "learning_rate": 8.052554949640537e-07,
+      "loss": 0.5196,
+      "step": 16682
+    },
+    {
+      "epoch": 4.428780034514801,
+      "grad_norm": 0.4827863889923207,
+      "learning_rate": 8.049988479537515e-07,
+      "loss": 0.536,
+      "step": 16683
+    },
+    {
+      "epoch": 4.4290455329881855,
+      "grad_norm": 0.4652426058007033,
+      "learning_rate": 8.047422340001229e-07,
+      "loss": 0.5072,
+      "step": 16684
+    },
+    {
+      "epoch": 4.429311031461569,
+      "grad_norm": 0.48168034844975,
+      "learning_rate": 8.044856531081705e-07,
+      "loss": 0.5345,
+      "step": 16685
+    },
+    {
+      "epoch": 4.429576529934953,
+      "grad_norm": 0.47899488626405107,
+      "learning_rate": 8.042291052828999e-07,
+      "loss": 0.5617,
+      "step": 16686
+    },
+    {
+      "epoch": 4.429842028408337,
+      "grad_norm": 0.46937329015362367,
+      "learning_rate": 8.039725905293139e-07,
+      "loss": 0.5276,
+      "step": 16687
+    },
+    {
+      "epoch": 4.43010752688172,
+      "grad_norm": 0.4943864339158835,
+      "learning_rate": 8.03716108852414e-07,
+      "loss": 0.5466,
+      "step": 16688
+    },
+    {
+      "epoch": 4.430373025355104,
+      "grad_norm": 0.4662196098037641,
+      "learning_rate": 8.03459660257204e-07,
+      "loss": 0.5432,
+      "step": 16689
+    },
+    {
+      "epoch": 4.430638523828488,
+      "grad_norm": 0.4612466141456489,
+      "learning_rate": 8.032032447486837e-07,
+      "loss": 0.4958,
+      "step": 16690
+    },
+    {
+      "epoch": 4.430904022301871,
+      "grad_norm": 0.4816510280112654,
+      "learning_rate": 8.029468623318554e-07,
+      "loss": 0.5306,
+      "step": 16691
+    },
+    {
+      "epoch": 4.4311695207752555,
+      "grad_norm": 0.4786703416356911,
+      "learning_rate": 8.026905130117177e-07,
+      "loss": 0.553,
+      "step": 16692
+    },
+    {
+      "epoch": 4.43143501924864,
+      "grad_norm": 0.48149719945164426,
+      "learning_rate": 8.024341967932722e-07,
+      "loss": 0.5351,
+      "step": 16693
+    },
+    {
+      "epoch": 4.431700517722023,
+      "grad_norm": 0.47399154768588453,
+      "learning_rate": 8.021779136815147e-07,
+      "loss": 0.5149,
+      "step": 16694
+    },
+    {
+      "epoch": 4.431966016195407,
+      "grad_norm": 0.47216508910420213,
+      "learning_rate": 8.019216636814455e-07,
+      "loss": 0.5892,
+      "step": 16695
+    },
+    {
+      "epoch": 4.432231514668791,
+      "grad_norm": 0.4796382068302887,
+      "learning_rate": 8.016654467980609e-07,
+      "loss": 0.5366,
+      "step": 16696
+    },
+    {
+      "epoch": 4.432497013142174,
+      "grad_norm": 0.49496431736136587,
+      "learning_rate": 8.01409263036359e-07,
+      "loss": 0.5495,
+      "step": 16697
+    },
+    {
+      "epoch": 4.432762511615558,
+      "grad_norm": 0.45821421861579115,
+      "learning_rate": 8.011531124013347e-07,
+      "loss": 0.5199,
+      "step": 16698
+    },
+    {
+      "epoch": 4.433028010088942,
+      "grad_norm": 0.4802740697578364,
+      "learning_rate": 8.008969948979848e-07,
+      "loss": 0.5339,
+      "step": 16699
+    },
+    {
+      "epoch": 4.4332935085623255,
+      "grad_norm": 0.48833097301451095,
+      "learning_rate": 8.006409105313037e-07,
+      "loss": 0.5564,
+      "step": 16700
+    },
+    {
+      "epoch": 4.43355900703571,
+      "grad_norm": 0.48503996782502184,
+      "learning_rate": 8.003848593062849e-07,
+      "loss": 0.5462,
+      "step": 16701
+    },
+    {
+      "epoch": 4.433824505509094,
+      "grad_norm": 0.47691588345034797,
+      "learning_rate": 8.001288412279232e-07,
+      "loss": 0.5645,
+      "step": 16702
+    },
+    {
+      "epoch": 4.434090003982477,
+      "grad_norm": 0.471904522744315,
+      "learning_rate": 7.998728563012106e-07,
+      "loss": 0.5323,
+      "step": 16703
+    },
+    {
+      "epoch": 4.434355502455861,
+      "grad_norm": 0.47149038026698187,
+      "learning_rate": 7.996169045311405e-07,
+      "loss": 0.5263,
+      "step": 16704
+    },
+    {
+      "epoch": 4.434621000929245,
+      "grad_norm": 0.47541902670267966,
+      "learning_rate": 7.993609859227042e-07,
+      "loss": 0.5024,
+      "step": 16705
+    },
+    {
+      "epoch": 4.434886499402628,
+      "grad_norm": 0.47867409564014474,
+      "learning_rate": 7.991051004808917e-07,
+      "loss": 0.5366,
+      "step": 16706
+    },
+    {
+      "epoch": 4.435151997876012,
+      "grad_norm": 0.4865322128941707,
+      "learning_rate": 7.988492482106952e-07,
+      "loss": 0.5545,
+      "step": 16707
+    },
+    {
+      "epoch": 4.435417496349396,
+      "grad_norm": 0.47569377851924544,
+      "learning_rate": 7.985934291171024e-07,
+      "loss": 0.5352,
+      "step": 16708
+    },
+    {
+      "epoch": 4.43568299482278,
+      "grad_norm": 0.48054140606946316,
+      "learning_rate": 7.983376432051046e-07,
+      "loss": 0.5352,
+      "step": 16709
+    },
+    {
+      "epoch": 4.435948493296164,
+      "grad_norm": 0.4725020440233165,
+      "learning_rate": 7.980818904796888e-07,
+      "loss": 0.5325,
+      "step": 16710
+    },
+    {
+      "epoch": 4.436213991769547,
+      "grad_norm": 0.48126112781960134,
+      "learning_rate": 7.978261709458434e-07,
+      "loss": 0.5626,
+      "step": 16711
+    },
+    {
+      "epoch": 4.436479490242931,
+      "grad_norm": 0.4946834364495555,
+      "learning_rate": 7.975704846085544e-07,
+      "loss": 0.503,
+      "step": 16712
+    },
+    {
+      "epoch": 4.436744988716315,
+      "grad_norm": 0.4640553607220983,
+      "learning_rate": 7.973148314728102e-07,
+      "loss": 0.5366,
+      "step": 16713
+    },
+    {
+      "epoch": 4.437010487189698,
+      "grad_norm": 0.4640623552025175,
+      "learning_rate": 7.970592115435946e-07,
+      "loss": 0.5516,
+      "step": 16714
+    },
+    {
+      "epoch": 4.437275985663082,
+      "grad_norm": 0.475925820198665,
+      "learning_rate": 7.968036248258951e-07,
+      "loss": 0.5401,
+      "step": 16715
+    },
+    {
+      "epoch": 4.437541484136466,
+      "grad_norm": 0.4816646501219935,
+      "learning_rate": 7.965480713246948e-07,
+      "loss": 0.5485,
+      "step": 16716
+    },
+    {
+      "epoch": 4.43780698260985,
+      "grad_norm": 0.4750407822401424,
+      "learning_rate": 7.96292551044977e-07,
+      "loss": 0.5401,
+      "step": 16717
+    },
+    {
+      "epoch": 4.438072481083234,
+      "grad_norm": 0.4897962754862172,
+      "learning_rate": 7.960370639917267e-07,
+      "loss": 0.617,
+      "step": 16718
+    },
+    {
+      "epoch": 4.438337979556618,
+      "grad_norm": 0.48956763703120704,
+      "learning_rate": 7.957816101699248e-07,
+      "loss": 0.5249,
+      "step": 16719
+    },
+    {
+      "epoch": 4.438603478030001,
+      "grad_norm": 0.4700803075201632,
+      "learning_rate": 7.955261895845551e-07,
+      "loss": 0.5505,
+      "step": 16720
+    },
+    {
+      "epoch": 4.438868976503385,
+      "grad_norm": 0.4703157179682438,
+      "learning_rate": 7.952708022405972e-07,
+      "loss": 0.4998,
+      "step": 16721
+    },
+    {
+      "epoch": 4.439134474976769,
+      "grad_norm": 0.4676187978465773,
+      "learning_rate": 7.950154481430338e-07,
+      "loss": 0.5437,
+      "step": 16722
+    },
+    {
+      "epoch": 4.439399973450152,
+      "grad_norm": 0.476401499372817,
+      "learning_rate": 7.947601272968425e-07,
+      "loss": 0.5374,
+      "step": 16723
+    },
+    {
+      "epoch": 4.439665471923536,
+      "grad_norm": 0.4856438162776569,
+      "learning_rate": 7.945048397070037e-07,
+      "loss": 0.5533,
+      "step": 16724
+    },
+    {
+      "epoch": 4.4399309703969205,
+      "grad_norm": 0.45674573049263667,
+      "learning_rate": 7.942495853784973e-07,
+      "loss": 0.5158,
+      "step": 16725
+    },
+    {
+      "epoch": 4.440196468870304,
+      "grad_norm": 0.4884248879681289,
+      "learning_rate": 7.939943643162992e-07,
+      "loss": 0.5281,
+      "step": 16726
+    },
+    {
+      "epoch": 4.440461967343688,
+      "grad_norm": 0.4846921576477788,
+      "learning_rate": 7.937391765253891e-07,
+      "loss": 0.5115,
+      "step": 16727
+    },
+    {
+      "epoch": 4.440727465817072,
+      "grad_norm": 0.47855556948809413,
+      "learning_rate": 7.934840220107426e-07,
+      "loss": 0.5596,
+      "step": 16728
+    },
+    {
+      "epoch": 4.440992964290455,
+      "grad_norm": 0.4753505047030627,
+      "learning_rate": 7.932289007773361e-07,
+      "loss": 0.5071,
+      "step": 16729
+    },
+    {
+      "epoch": 4.441258462763839,
+      "grad_norm": 0.4663231667013425,
+      "learning_rate": 7.92973812830144e-07,
+      "loss": 0.5468,
+      "step": 16730
+    },
+    {
+      "epoch": 4.441523961237223,
+      "grad_norm": 0.47037980895030174,
+      "learning_rate": 7.927187581741427e-07,
+      "loss": 0.5007,
+      "step": 16731
+    },
+    {
+      "epoch": 4.4417894597106065,
+      "grad_norm": 0.47351158245135905,
+      "learning_rate": 7.924637368143051e-07,
+      "loss": 0.5349,
+      "step": 16732
+    },
+    {
+      "epoch": 4.4420549581839905,
+      "grad_norm": 0.48662427446942946,
+      "learning_rate": 7.922087487556063e-07,
+      "loss": 0.5484,
+      "step": 16733
+    },
+    {
+      "epoch": 4.442320456657375,
+      "grad_norm": 0.48256038950280983,
+      "learning_rate": 7.919537940030184e-07,
+      "loss": 0.5159,
+      "step": 16734
+    },
+    {
+      "epoch": 4.442585955130758,
+      "grad_norm": 0.4785808617281227,
+      "learning_rate": 7.916988725615124e-07,
+      "loss": 0.5431,
+      "step": 16735
+    },
+    {
+      "epoch": 4.442851453604142,
+      "grad_norm": 0.47371839271109195,
+      "learning_rate": 7.914439844360619e-07,
+      "loss": 0.527,
+      "step": 16736
+    },
+    {
+      "epoch": 4.443116952077526,
+      "grad_norm": 0.4765259850200113,
+      "learning_rate": 7.911891296316363e-07,
+      "loss": 0.5377,
+      "step": 16737
+    },
+    {
+      "epoch": 4.443382450550909,
+      "grad_norm": 0.5016182844340449,
+      "learning_rate": 7.909343081532073e-07,
+      "loss": 0.5236,
+      "step": 16738
+    },
+    {
+      "epoch": 4.443647949024293,
+      "grad_norm": 0.4852329858543557,
+      "learning_rate": 7.906795200057438e-07,
+      "loss": 0.5366,
+      "step": 16739
+    },
+    {
+      "epoch": 4.4439134474976765,
+      "grad_norm": 0.48338999111474706,
+      "learning_rate": 7.904247651942148e-07,
+      "loss": 0.5606,
+      "step": 16740
+    },
+    {
+      "epoch": 4.444178945971061,
+      "grad_norm": 0.49035578233200916,
+      "learning_rate": 7.901700437235879e-07,
+      "loss": 0.5422,
+      "step": 16741
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.4831257543465122,
+      "learning_rate": 7.899153555988323e-07,
+      "loss": 0.544,
+      "step": 16742
+    },
+    {
+      "epoch": 4.444709942917828,
+      "grad_norm": 0.4836291559174445,
+      "learning_rate": 7.896607008249135e-07,
+      "loss": 0.4983,
+      "step": 16743
+    },
+    {
+      "epoch": 4.444975441391212,
+      "grad_norm": 0.47972689334546736,
+      "learning_rate": 7.894060794067987e-07,
+      "loss": 0.5574,
+      "step": 16744
+    },
+    {
+      "epoch": 4.445240939864596,
+      "grad_norm": 0.46919579797531896,
+      "learning_rate": 7.891514913494552e-07,
+      "loss": 0.5123,
+      "step": 16745
+    },
+    {
+      "epoch": 4.445506438337979,
+      "grad_norm": 0.4958050831216418,
+      "learning_rate": 7.888969366578447e-07,
+      "loss": 0.5154,
+      "step": 16746
+    },
+    {
+      "epoch": 4.445771936811363,
+      "grad_norm": 0.46484540848926803,
+      "learning_rate": 7.886424153369345e-07,
+      "loss": 0.5173,
+      "step": 16747
+    },
+    {
+      "epoch": 4.446037435284747,
+      "grad_norm": 0.49628724274766695,
+      "learning_rate": 7.883879273916864e-07,
+      "loss": 0.5447,
+      "step": 16748
+    },
+    {
+      "epoch": 4.446302933758131,
+      "grad_norm": 0.47494283027982415,
+      "learning_rate": 7.881334728270654e-07,
+      "loss": 0.556,
+      "step": 16749
+    },
+    {
+      "epoch": 4.446568432231515,
+      "grad_norm": 0.47685395495933697,
+      "learning_rate": 7.87879051648032e-07,
+      "loss": 0.572,
+      "step": 16750
+    },
+    {
+      "epoch": 4.446833930704899,
+      "grad_norm": 0.4910985040099849,
+      "learning_rate": 7.8762466385955e-07,
+      "loss": 0.5568,
+      "step": 16751
+    },
+    {
+      "epoch": 4.447099429178282,
+      "grad_norm": 0.48524782649668274,
+      "learning_rate": 7.873703094665797e-07,
+      "loss": 0.533,
+      "step": 16752
+    },
+    {
+      "epoch": 4.447364927651666,
+      "grad_norm": 0.48305599840260044,
+      "learning_rate": 7.871159884740809e-07,
+      "loss": 0.5494,
+      "step": 16753
+    },
+    {
+      "epoch": 4.44763042612505,
+      "grad_norm": 0.48820265328632984,
+      "learning_rate": 7.868617008870147e-07,
+      "loss": 0.5337,
+      "step": 16754
+    },
+    {
+      "epoch": 4.447895924598433,
+      "grad_norm": 0.4697167839572464,
+      "learning_rate": 7.866074467103393e-07,
+      "loss": 0.5623,
+      "step": 16755
+    },
+    {
+      "epoch": 4.448161423071817,
+      "grad_norm": 0.4782113379668213,
+      "learning_rate": 7.863532259490153e-07,
+      "loss": 0.5286,
+      "step": 16756
+    },
+    {
+      "epoch": 4.4484269215452015,
+      "grad_norm": 0.46709937070414953,
+      "learning_rate": 7.860990386079975e-07,
+      "loss": 0.5164,
+      "step": 16757
+    },
+    {
+      "epoch": 4.448692420018585,
+      "grad_norm": 0.48349875174302726,
+      "learning_rate": 7.858448846922457e-07,
+      "loss": 0.5542,
+      "step": 16758
+    },
+    {
+      "epoch": 4.448957918491969,
+      "grad_norm": 0.48189910242283673,
+      "learning_rate": 7.855907642067146e-07,
+      "loss": 0.5576,
+      "step": 16759
+    },
+    {
+      "epoch": 4.449223416965353,
+      "grad_norm": 0.46975698146348355,
+      "learning_rate": 7.853366771563623e-07,
+      "loss": 0.5366,
+      "step": 16760
+    },
+    {
+      "epoch": 4.449488915438736,
+      "grad_norm": 0.4729750889908494,
+      "learning_rate": 7.85082623546142e-07,
+      "loss": 0.5426,
+      "step": 16761
+    },
+    {
+      "epoch": 4.44975441391212,
+      "grad_norm": 0.4812866589759176,
+      "learning_rate": 7.848286033810101e-07,
+      "loss": 0.5735,
+      "step": 16762
+    },
+    {
+      "epoch": 4.450019912385503,
+      "grad_norm": 0.4846423364481977,
+      "learning_rate": 7.845746166659202e-07,
+      "loss": 0.5574,
+      "step": 16763
+    },
+    {
+      "epoch": 4.450285410858887,
+      "grad_norm": 0.4875330358884606,
+      "learning_rate": 7.843206634058242e-07,
+      "loss": 0.5522,
+      "step": 16764
+    },
+    {
+      "epoch": 4.4505509093322715,
+      "grad_norm": 0.47797265466770783,
+      "learning_rate": 7.84066743605677e-07,
+      "loss": 0.5076,
+      "step": 16765
+    },
+    {
+      "epoch": 4.450816407805656,
+      "grad_norm": 0.47567480459066036,
+      "learning_rate": 7.838128572704287e-07,
+      "loss": 0.5368,
+      "step": 16766
+    },
+    {
+      "epoch": 4.451081906279039,
+      "grad_norm": 0.47934673595619226,
+      "learning_rate": 7.835590044050326e-07,
+      "loss": 0.5058,
+      "step": 16767
+    },
+    {
+      "epoch": 4.451347404752423,
+      "grad_norm": 0.4761895208501416,
+      "learning_rate": 7.833051850144386e-07,
+      "loss": 0.5351,
+      "step": 16768
+    },
+    {
+      "epoch": 4.451612903225806,
+      "grad_norm": 0.4750377818317134,
+      "learning_rate": 7.830513991035957e-07,
+      "loss": 0.5341,
+      "step": 16769
+    },
+    {
+      "epoch": 4.45187840169919,
+      "grad_norm": 0.459227752282165,
+      "learning_rate": 7.82797646677455e-07,
+      "loss": 0.5131,
+      "step": 16770
+    },
+    {
+      "epoch": 4.452143900172574,
+      "grad_norm": 0.48329382337217675,
+      "learning_rate": 7.825439277409638e-07,
+      "loss": 0.5756,
+      "step": 16771
+    },
+    {
+      "epoch": 4.452409398645957,
+      "grad_norm": 0.4844951852258455,
+      "learning_rate": 7.822902422990722e-07,
+      "loss": 0.5441,
+      "step": 16772
+    },
+    {
+      "epoch": 4.4526748971193415,
+      "grad_norm": 0.48052774196571807,
+      "learning_rate": 7.820365903567256e-07,
+      "loss": 0.5024,
+      "step": 16773
+    },
+    {
+      "epoch": 4.452940395592726,
+      "grad_norm": 0.4930231968862931,
+      "learning_rate": 7.817829719188732e-07,
+      "loss": 0.5396,
+      "step": 16774
+    },
+    {
+      "epoch": 4.453205894066109,
+      "grad_norm": 0.476994970560152,
+      "learning_rate": 7.815293869904586e-07,
+      "loss": 0.5516,
+      "step": 16775
+    },
+    {
+      "epoch": 4.453471392539493,
+      "grad_norm": 0.46195527064165515,
+      "learning_rate": 7.812758355764288e-07,
+      "loss": 0.5298,
+      "step": 16776
+    },
+    {
+      "epoch": 4.453736891012877,
+      "grad_norm": 0.48098791457085927,
+      "learning_rate": 7.810223176817278e-07,
+      "loss": 0.5331,
+      "step": 16777
+    },
+    {
+      "epoch": 4.45400238948626,
+      "grad_norm": 0.4881889105233967,
+      "learning_rate": 7.807688333113011e-07,
+      "loss": 0.5557,
+      "step": 16778
+    },
+    {
+      "epoch": 4.454267887959644,
+      "grad_norm": 0.4783860724493161,
+      "learning_rate": 7.805153824700917e-07,
+      "loss": 0.5772,
+      "step": 16779
+    },
+    {
+      "epoch": 4.454533386433028,
+      "grad_norm": 0.46467341343593643,
+      "learning_rate": 7.802619651630416e-07,
+      "loss": 0.5289,
+      "step": 16780
+    },
+    {
+      "epoch": 4.4547988849064115,
+      "grad_norm": 0.4729765867223776,
+      "learning_rate": 7.800085813950944e-07,
+      "loss": 0.5198,
+      "step": 16781
+    },
+    {
+      "epoch": 4.455064383379796,
+      "grad_norm": 0.4894763189619003,
+      "learning_rate": 7.797552311711906e-07,
+      "loss": 0.5591,
+      "step": 16782
+    },
+    {
+      "epoch": 4.45532988185318,
+      "grad_norm": 0.4795746775098918,
+      "learning_rate": 7.795019144962723e-07,
+      "loss": 0.5494,
+      "step": 16783
+    },
+    {
+      "epoch": 4.455595380326563,
+      "grad_norm": 0.4783304874710241,
+      "learning_rate": 7.792486313752784e-07,
+      "loss": 0.5073,
+      "step": 16784
+    },
+    {
+      "epoch": 4.455860878799947,
+      "grad_norm": 0.47836348382014443,
+      "learning_rate": 7.789953818131509e-07,
+      "loss": 0.5359,
+      "step": 16785
+    },
+    {
+      "epoch": 4.456126377273331,
+      "grad_norm": 0.4636302201351266,
+      "learning_rate": 7.787421658148256e-07,
+      "loss": 0.5106,
+      "step": 16786
+    },
+    {
+      "epoch": 4.456391875746714,
+      "grad_norm": 0.46698177063766205,
+      "learning_rate": 7.784889833852433e-07,
+      "loss": 0.5116,
+      "step": 16787
+    },
+    {
+      "epoch": 4.456657374220098,
+      "grad_norm": 0.47998549117174644,
+      "learning_rate": 7.7823583452934e-07,
+      "loss": 0.512,
+      "step": 16788
+    },
+    {
+      "epoch": 4.456922872693482,
+      "grad_norm": 0.47774672511202276,
+      "learning_rate": 7.779827192520545e-07,
+      "loss": 0.585,
+      "step": 16789
+    },
+    {
+      "epoch": 4.457188371166866,
+      "grad_norm": 0.487583737099386,
+      "learning_rate": 7.777296375583209e-07,
+      "loss": 0.5421,
+      "step": 16790
+    },
+    {
+      "epoch": 4.45745386964025,
+      "grad_norm": 0.48931293033171347,
+      "learning_rate": 7.774765894530776e-07,
+      "loss": 0.5577,
+      "step": 16791
+    },
+    {
+      "epoch": 4.457719368113633,
+      "grad_norm": 0.476947046330505,
+      "learning_rate": 7.77223574941258e-07,
+      "loss": 0.5575,
+      "step": 16792
+    },
+    {
+      "epoch": 4.457984866587017,
+      "grad_norm": 0.47321848869325484,
+      "learning_rate": 7.769705940277958e-07,
+      "loss": 0.5156,
+      "step": 16793
+    },
+    {
+      "epoch": 4.458250365060401,
+      "grad_norm": 0.4688683850512754,
+      "learning_rate": 7.767176467176266e-07,
+      "loss": 0.5339,
+      "step": 16794
+    },
+    {
+      "epoch": 4.458515863533785,
+      "grad_norm": 0.47543291376002594,
+      "learning_rate": 7.764647330156818e-07,
+      "loss": 0.5194,
+      "step": 16795
+    },
+    {
+      "epoch": 4.458781362007168,
+      "grad_norm": 0.4795077400845625,
+      "learning_rate": 7.762118529268959e-07,
+      "loss": 0.5392,
+      "step": 16796
+    },
+    {
+      "epoch": 4.459046860480552,
+      "grad_norm": 0.47132659691120776,
+      "learning_rate": 7.759590064561989e-07,
+      "loss": 0.5416,
+      "step": 16797
+    },
+    {
+      "epoch": 4.459312358953936,
+      "grad_norm": 0.4700233051810075,
+      "learning_rate": 7.757061936085217e-07,
+      "loss": 0.496,
+      "step": 16798
+    },
+    {
+      "epoch": 4.45957785742732,
+      "grad_norm": 0.45174579096216716,
+      "learning_rate": 7.754534143887965e-07,
+      "loss": 0.469,
+      "step": 16799
+    },
+    {
+      "epoch": 4.459843355900704,
+      "grad_norm": 0.4868248325826904,
+      "learning_rate": 7.752006688019512e-07,
+      "loss": 0.512,
+      "step": 16800
+    },
+    {
+      "epoch": 4.460108854374087,
+      "grad_norm": 0.48859601130345104,
+      "learning_rate": 7.74947956852917e-07,
+      "loss": 0.535,
+      "step": 16801
+    },
+    {
+      "epoch": 4.460374352847471,
+      "grad_norm": 0.4742936560337386,
+      "learning_rate": 7.746952785466202e-07,
+      "loss": 0.5205,
+      "step": 16802
+    },
+    {
+      "epoch": 4.460639851320855,
+      "grad_norm": 0.4721852507099897,
+      "learning_rate": 7.744426338879912e-07,
+      "loss": 0.5219,
+      "step": 16803
+    },
+    {
+      "epoch": 4.460905349794238,
+      "grad_norm": 0.4817985646773516,
+      "learning_rate": 7.741900228819546e-07,
+      "loss": 0.5437,
+      "step": 16804
+    },
+    {
+      "epoch": 4.4611708482676224,
+      "grad_norm": 0.4740503672211233,
+      "learning_rate": 7.739374455334386e-07,
+      "loss": 0.5485,
+      "step": 16805
+    },
+    {
+      "epoch": 4.4614363467410065,
+      "grad_norm": 0.47462231297439617,
+      "learning_rate": 7.736849018473677e-07,
+      "loss": 0.5233,
+      "step": 16806
+    },
+    {
+      "epoch": 4.46170184521439,
+      "grad_norm": 0.4698423276837917,
+      "learning_rate": 7.734323918286688e-07,
+      "loss": 0.5409,
+      "step": 16807
+    },
+    {
+      "epoch": 4.461967343687774,
+      "grad_norm": 0.47837117070306,
+      "learning_rate": 7.731799154822656e-07,
+      "loss": 0.5568,
+      "step": 16808
+    },
+    {
+      "epoch": 4.462232842161158,
+      "grad_norm": 0.4790102385519283,
+      "learning_rate": 7.729274728130812e-07,
+      "loss": 0.5444,
+      "step": 16809
+    },
+    {
+      "epoch": 4.462498340634541,
+      "grad_norm": 0.48203130885808015,
+      "learning_rate": 7.726750638260408e-07,
+      "loss": 0.5724,
+      "step": 16810
+    },
+    {
+      "epoch": 4.462763839107925,
+      "grad_norm": 0.463640351984473,
+      "learning_rate": 7.724226885260647e-07,
+      "loss": 0.5309,
+      "step": 16811
+    },
+    {
+      "epoch": 4.463029337581309,
+      "grad_norm": 0.48554503779120334,
+      "learning_rate": 7.721703469180769e-07,
+      "loss": 0.5745,
+      "step": 16812
+    },
+    {
+      "epoch": 4.4632948360546925,
+      "grad_norm": 0.47484646678052106,
+      "learning_rate": 7.719180390069972e-07,
+      "loss": 0.561,
+      "step": 16813
+    },
+    {
+      "epoch": 4.4635603345280765,
+      "grad_norm": 0.49526564017900865,
+      "learning_rate": 7.716657647977482e-07,
+      "loss": 0.5312,
+      "step": 16814
+    },
+    {
+      "epoch": 4.463825833001461,
+      "grad_norm": 0.4944937048713317,
+      "learning_rate": 7.71413524295247e-07,
+      "loss": 0.5724,
+      "step": 16815
+    },
+    {
+      "epoch": 4.464091331474844,
+      "grad_norm": 0.4830335153746338,
+      "learning_rate": 7.711613175044141e-07,
+      "loss": 0.5407,
+      "step": 16816
+    },
+    {
+      "epoch": 4.464356829948228,
+      "grad_norm": 0.4914262325565297,
+      "learning_rate": 7.709091444301694e-07,
+      "loss": 0.5295,
+      "step": 16817
+    },
+    {
+      "epoch": 4.464622328421612,
+      "grad_norm": 0.49815447029919036,
+      "learning_rate": 7.706570050774293e-07,
+      "loss": 0.5585,
+      "step": 16818
+    },
+    {
+      "epoch": 4.464887826894995,
+      "grad_norm": 0.481570914138703,
+      "learning_rate": 7.704048994511123e-07,
+      "loss": 0.5237,
+      "step": 16819
+    },
+    {
+      "epoch": 4.465153325368379,
+      "grad_norm": 0.4772179254194095,
+      "learning_rate": 7.701528275561349e-07,
+      "loss": 0.5626,
+      "step": 16820
+    },
+    {
+      "epoch": 4.4654188238417625,
+      "grad_norm": 0.4745463224914997,
+      "learning_rate": 7.699007893974125e-07,
+      "loss": 0.5313,
+      "step": 16821
+    },
+    {
+      "epoch": 4.465684322315147,
+      "grad_norm": 0.48371416579121224,
+      "learning_rate": 7.696487849798598e-07,
+      "loss": 0.5491,
+      "step": 16822
+    },
+    {
+      "epoch": 4.465949820788531,
+      "grad_norm": 0.4783245824257991,
+      "learning_rate": 7.693968143083936e-07,
+      "loss": 0.5469,
+      "step": 16823
+    },
+    {
+      "epoch": 4.466215319261914,
+      "grad_norm": 0.47661642220234274,
+      "learning_rate": 7.691448773879257e-07,
+      "loss": 0.5239,
+      "step": 16824
+    },
+    {
+      "epoch": 4.466480817735298,
+      "grad_norm": 0.47825627645453545,
+      "learning_rate": 7.688929742233714e-07,
+      "loss": 0.537,
+      "step": 16825
+    },
+    {
+      "epoch": 4.466746316208682,
+      "grad_norm": 0.48049150338784724,
+      "learning_rate": 7.686411048196426e-07,
+      "loss": 0.5156,
+      "step": 16826
+    },
+    {
+      "epoch": 4.467011814682065,
+      "grad_norm": 0.4853666868482445,
+      "learning_rate": 7.683892691816505e-07,
+      "loss": 0.5582,
+      "step": 16827
+    },
+    {
+      "epoch": 4.467277313155449,
+      "grad_norm": 0.46326720083328293,
+      "learning_rate": 7.681374673143083e-07,
+      "loss": 0.4836,
+      "step": 16828
+    },
+    {
+      "epoch": 4.467542811628833,
+      "grad_norm": 0.4708498491803565,
+      "learning_rate": 7.678856992225248e-07,
+      "loss": 0.5533,
+      "step": 16829
+    },
+    {
+      "epoch": 4.467808310102217,
+      "grad_norm": 0.4763351351571721,
+      "learning_rate": 7.676339649112121e-07,
+      "loss": 0.5732,
+      "step": 16830
+    },
+    {
+      "epoch": 4.468073808575601,
+      "grad_norm": 0.4827704324084567,
+      "learning_rate": 7.673822643852785e-07,
+      "loss": 0.5531,
+      "step": 16831
+    },
+    {
+      "epoch": 4.468339307048985,
+      "grad_norm": 0.4811741930674635,
+      "learning_rate": 7.671305976496329e-07,
+      "loss": 0.5488,
+      "step": 16832
+    },
+    {
+      "epoch": 4.468604805522368,
+      "grad_norm": 0.4974384865027843,
+      "learning_rate": 7.668789647091828e-07,
+      "loss": 0.5393,
+      "step": 16833
+    },
+    {
+      "epoch": 4.468870303995752,
+      "grad_norm": 0.4801160540271777,
+      "learning_rate": 7.666273655688372e-07,
+      "loss": 0.571,
+      "step": 16834
+    },
+    {
+      "epoch": 4.469135802469136,
+      "grad_norm": 0.47421844996759044,
+      "learning_rate": 7.663758002335009e-07,
+      "loss": 0.5245,
+      "step": 16835
+    },
+    {
+      "epoch": 4.469401300942519,
+      "grad_norm": 0.4746300183106361,
+      "learning_rate": 7.661242687080813e-07,
+      "loss": 0.5273,
+      "step": 16836
+    },
+    {
+      "epoch": 4.469666799415903,
+      "grad_norm": 0.47818752052738495,
+      "learning_rate": 7.658727709974853e-07,
+      "loss": 0.555,
+      "step": 16837
+    },
+    {
+      "epoch": 4.4699322978892875,
+      "grad_norm": 0.47902922010848037,
+      "learning_rate": 7.656213071066146e-07,
+      "loss": 0.5239,
+      "step": 16838
+    },
+    {
+      "epoch": 4.470197796362671,
+      "grad_norm": 0.4710875706464958,
+      "learning_rate": 7.653698770403755e-07,
+      "loss": 0.5467,
+      "step": 16839
+    },
+    {
+      "epoch": 4.470463294836055,
+      "grad_norm": 0.49564338962342525,
+      "learning_rate": 7.651184808036705e-07,
+      "loss": 0.5475,
+      "step": 16840
+    },
+    {
+      "epoch": 4.470728793309439,
+      "grad_norm": 0.47452879767399997,
+      "learning_rate": 7.648671184014033e-07,
+      "loss": 0.5133,
+      "step": 16841
+    },
+    {
+      "epoch": 4.470994291782822,
+      "grad_norm": 0.484640531195572,
+      "learning_rate": 7.646157898384751e-07,
+      "loss": 0.5247,
+      "step": 16842
+    },
+    {
+      "epoch": 4.471259790256206,
+      "grad_norm": 0.48931215975181286,
+      "learning_rate": 7.643644951197885e-07,
+      "loss": 0.557,
+      "step": 16843
+    },
+    {
+      "epoch": 4.47152528872959,
+      "grad_norm": 0.48300915134314404,
+      "learning_rate": 7.641132342502439e-07,
+      "loss": 0.5065,
+      "step": 16844
+    },
+    {
+      "epoch": 4.471790787202973,
+      "grad_norm": 0.46721172452743115,
+      "learning_rate": 7.638620072347405e-07,
+      "loss": 0.5473,
+      "step": 16845
+    },
+    {
+      "epoch": 4.4720562856763575,
+      "grad_norm": 0.48094695221407097,
+      "learning_rate": 7.636108140781798e-07,
+      "loss": 0.5526,
+      "step": 16846
+    },
+    {
+      "epoch": 4.472321784149742,
+      "grad_norm": 0.4794989833368633,
+      "learning_rate": 7.633596547854588e-07,
+      "loss": 0.5486,
+      "step": 16847
+    },
+    {
+      "epoch": 4.472587282623125,
+      "grad_norm": 0.48316894742047806,
+      "learning_rate": 7.631085293614771e-07,
+      "loss": 0.5339,
+      "step": 16848
+    },
+    {
+      "epoch": 4.472852781096509,
+      "grad_norm": 0.49101837193917114,
+      "learning_rate": 7.62857437811132e-07,
+      "loss": 0.55,
+      "step": 16849
+    },
+    {
+      "epoch": 4.473118279569892,
+      "grad_norm": 0.4785061978352384,
+      "learning_rate": 7.6260638013932e-07,
+      "loss": 0.4978,
+      "step": 16850
+    },
+    {
+      "epoch": 4.473383778043276,
+      "grad_norm": 0.4683036301392794,
+      "learning_rate": 7.623553563509367e-07,
+      "loss": 0.5402,
+      "step": 16851
+    },
+    {
+      "epoch": 4.47364927651666,
+      "grad_norm": 0.47321788405586923,
+      "learning_rate": 7.621043664508793e-07,
+      "loss": 0.5127,
+      "step": 16852
+    },
+    {
+      "epoch": 4.473914774990043,
+      "grad_norm": 0.467525947748451,
+      "learning_rate": 7.618534104440409e-07,
+      "loss": 0.5448,
+      "step": 16853
+    },
+    {
+      "epoch": 4.4741802734634275,
+      "grad_norm": 0.46238664931981793,
+      "learning_rate": 7.616024883353176e-07,
+      "loss": 0.5053,
+      "step": 16854
+    },
+    {
+      "epoch": 4.474445771936812,
+      "grad_norm": 0.4821784597825707,
+      "learning_rate": 7.61351600129602e-07,
+      "loss": 0.5339,
+      "step": 16855
+    },
+    {
+      "epoch": 4.474711270410195,
+      "grad_norm": 0.49618910802445,
+      "learning_rate": 7.611007458317865e-07,
+      "loss": 0.5516,
+      "step": 16856
+    },
+    {
+      "epoch": 4.474976768883579,
+      "grad_norm": 0.4871873461800691,
+      "learning_rate": 7.608499254467647e-07,
+      "loss": 0.5262,
+      "step": 16857
+    },
+    {
+      "epoch": 4.475242267356963,
+      "grad_norm": 0.48274045317548486,
+      "learning_rate": 7.605991389794268e-07,
+      "loss": 0.5557,
+      "step": 16858
+    },
+    {
+      "epoch": 4.475507765830346,
+      "grad_norm": 0.4767275049196292,
+      "learning_rate": 7.603483864346653e-07,
+      "loss": 0.5166,
+      "step": 16859
+    },
+    {
+      "epoch": 4.47577326430373,
+      "grad_norm": 0.4755331375435532,
+      "learning_rate": 7.600976678173697e-07,
+      "loss": 0.5488,
+      "step": 16860
+    },
+    {
+      "epoch": 4.476038762777114,
+      "grad_norm": 0.4719080735202083,
+      "learning_rate": 7.598469831324284e-07,
+      "loss": 0.5176,
+      "step": 16861
+    },
+    {
+      "epoch": 4.4763042612504975,
+      "grad_norm": 0.46911830415695305,
+      "learning_rate": 7.595963323847328e-07,
+      "loss": 0.5193,
+      "step": 16862
+    },
+    {
+      "epoch": 4.476569759723882,
+      "grad_norm": 0.4933431117284641,
+      "learning_rate": 7.593457155791692e-07,
+      "loss": 0.5242,
+      "step": 16863
+    },
+    {
+      "epoch": 4.476835258197266,
+      "grad_norm": 0.48850423620668215,
+      "learning_rate": 7.590951327206264e-07,
+      "loss": 0.5049,
+      "step": 16864
+    },
+    {
+      "epoch": 4.477100756670649,
+      "grad_norm": 0.475543322350099,
+      "learning_rate": 7.588445838139905e-07,
+      "loss": 0.5362,
+      "step": 16865
+    },
+    {
+      "epoch": 4.477366255144033,
+      "grad_norm": 0.482123911051005,
+      "learning_rate": 7.585940688641499e-07,
+      "loss": 0.5541,
+      "step": 16866
+    },
+    {
+      "epoch": 4.477631753617417,
+      "grad_norm": 0.4771534579196199,
+      "learning_rate": 7.58343587875987e-07,
+      "loss": 0.525,
+      "step": 16867
+    },
+    {
+      "epoch": 4.4778972520908,
+      "grad_norm": 0.477193930247143,
+      "learning_rate": 7.58093140854389e-07,
+      "loss": 0.5495,
+      "step": 16868
+    },
+    {
+      "epoch": 4.478162750564184,
+      "grad_norm": 0.48754187436692237,
+      "learning_rate": 7.578427278042391e-07,
+      "loss": 0.5234,
+      "step": 16869
+    },
+    {
+      "epoch": 4.478428249037568,
+      "grad_norm": 0.4765260618184064,
+      "learning_rate": 7.575923487304224e-07,
+      "loss": 0.5305,
+      "step": 16870
+    },
+    {
+      "epoch": 4.478693747510952,
+      "grad_norm": 0.47742572936281885,
+      "learning_rate": 7.573420036378204e-07,
+      "loss": 0.528,
+      "step": 16871
+    },
+    {
+      "epoch": 4.478959245984336,
+      "grad_norm": 0.4835757187076809,
+      "learning_rate": 7.570916925313165e-07,
+      "loss": 0.532,
+      "step": 16872
+    },
+    {
+      "epoch": 4.479224744457719,
+      "grad_norm": 0.47941219447669864,
+      "learning_rate": 7.568414154157922e-07,
+      "loss": 0.5516,
+      "step": 16873
+    },
+    {
+      "epoch": 4.479490242931103,
+      "grad_norm": 0.48778509592229713,
+      "learning_rate": 7.565911722961275e-07,
+      "loss": 0.5141,
+      "step": 16874
+    },
+    {
+      "epoch": 4.479755741404487,
+      "grad_norm": 0.47460255721038447,
+      "learning_rate": 7.563409631772043e-07,
+      "loss": 0.5233,
+      "step": 16875
+    },
+    {
+      "epoch": 4.480021239877871,
+      "grad_norm": 0.48913858413621,
+      "learning_rate": 7.560907880639007e-07,
+      "loss": 0.5188,
+      "step": 16876
+    },
+    {
+      "epoch": 4.480286738351254,
+      "grad_norm": 0.49581404032252385,
+      "learning_rate": 7.558406469610982e-07,
+      "loss": 0.5298,
+      "step": 16877
+    },
+    {
+      "epoch": 4.480552236824638,
+      "grad_norm": 0.47283289376814264,
+      "learning_rate": 7.555905398736718e-07,
+      "loss": 0.5023,
+      "step": 16878
+    },
+    {
+      "epoch": 4.480817735298022,
+      "grad_norm": 0.4904549769963252,
+      "learning_rate": 7.553404668065017e-07,
+      "loss": 0.5682,
+      "step": 16879
+    },
+    {
+      "epoch": 4.481083233771406,
+      "grad_norm": 0.48689180341885197,
+      "learning_rate": 7.550904277644635e-07,
+      "loss": 0.5339,
+      "step": 16880
+    },
+    {
+      "epoch": 4.48134873224479,
+      "grad_norm": 0.4835409913859052,
+      "learning_rate": 7.548404227524342e-07,
+      "loss": 0.5399,
+      "step": 16881
+    },
+    {
+      "epoch": 4.481614230718173,
+      "grad_norm": 0.4938640234370828,
+      "learning_rate": 7.545904517752906e-07,
+      "loss": 0.5721,
+      "step": 16882
+    },
+    {
+      "epoch": 4.481879729191557,
+      "grad_norm": 0.4849340218789454,
+      "learning_rate": 7.543405148379063e-07,
+      "loss": 0.5473,
+      "step": 16883
+    },
+    {
+      "epoch": 4.482145227664941,
+      "grad_norm": 0.48938650005707524,
+      "learning_rate": 7.540906119451564e-07,
+      "loss": 0.544,
+      "step": 16884
+    },
+    {
+      "epoch": 4.482410726138324,
+      "grad_norm": 0.4666671839274467,
+      "learning_rate": 7.538407431019132e-07,
+      "loss": 0.5366,
+      "step": 16885
+    },
+    {
+      "epoch": 4.4826762246117084,
+      "grad_norm": 0.47559293955184323,
+      "learning_rate": 7.535909083130515e-07,
+      "loss": 0.5472,
+      "step": 16886
+    },
+    {
+      "epoch": 4.4829417230850925,
+      "grad_norm": 0.5006126537964072,
+      "learning_rate": 7.533411075834426e-07,
+      "loss": 0.5593,
+      "step": 16887
+    },
+    {
+      "epoch": 4.483207221558476,
+      "grad_norm": 0.48276985867423183,
+      "learning_rate": 7.530913409179594e-07,
+      "loss": 0.536,
+      "step": 16888
+    },
+    {
+      "epoch": 4.48347272003186,
+      "grad_norm": 0.5027431727573257,
+      "learning_rate": 7.52841608321472e-07,
+      "loss": 0.5491,
+      "step": 16889
+    },
+    {
+      "epoch": 4.483738218505244,
+      "grad_norm": 0.48647038534828857,
+      "learning_rate": 7.525919097988504e-07,
+      "loss": 0.5572,
+      "step": 16890
+    },
+    {
+      "epoch": 4.484003716978627,
+      "grad_norm": 0.4755863022857314,
+      "learning_rate": 7.523422453549659e-07,
+      "loss": 0.5291,
+      "step": 16891
+    },
+    {
+      "epoch": 4.484269215452011,
+      "grad_norm": 0.47451792876456994,
+      "learning_rate": 7.520926149946856e-07,
+      "loss": 0.5298,
+      "step": 16892
+    },
+    {
+      "epoch": 4.484534713925395,
+      "grad_norm": 0.4817184723231692,
+      "learning_rate": 7.518430187228798e-07,
+      "loss": 0.5457,
+      "step": 16893
+    },
+    {
+      "epoch": 4.4848002123987785,
+      "grad_norm": 0.5005842622885278,
+      "learning_rate": 7.515934565444147e-07,
+      "loss": 0.5484,
+      "step": 16894
+    },
+    {
+      "epoch": 4.4850657108721625,
+      "grad_norm": 0.47156579819148675,
+      "learning_rate": 7.513439284641593e-07,
+      "loss": 0.5462,
+      "step": 16895
+    },
+    {
+      "epoch": 4.485331209345547,
+      "grad_norm": 0.4816356064933116,
+      "learning_rate": 7.510944344869775e-07,
+      "loss": 0.5437,
+      "step": 16896
+    },
+    {
+      "epoch": 4.48559670781893,
+      "grad_norm": 0.4697736734255272,
+      "learning_rate": 7.508449746177368e-07,
+      "loss": 0.5451,
+      "step": 16897
+    },
+    {
+      "epoch": 4.485862206292314,
+      "grad_norm": 0.48860648227893994,
+      "learning_rate": 7.505955488613009e-07,
+      "loss": 0.5647,
+      "step": 16898
+    },
+    {
+      "epoch": 4.486127704765698,
+      "grad_norm": 0.48804592987491996,
+      "learning_rate": 7.503461572225363e-07,
+      "loss": 0.515,
+      "step": 16899
+    },
+    {
+      "epoch": 4.486393203239081,
+      "grad_norm": 0.4837839544995788,
+      "learning_rate": 7.500967997063053e-07,
+      "loss": 0.5405,
+      "step": 16900
+    },
+    {
+      "epoch": 4.486658701712465,
+      "grad_norm": 0.46677871479160515,
+      "learning_rate": 7.498474763174701e-07,
+      "loss": 0.5161,
+      "step": 16901
+    },
+    {
+      "epoch": 4.4869242001858485,
+      "grad_norm": 0.48240247205356773,
+      "learning_rate": 7.495981870608954e-07,
+      "loss": 0.512,
+      "step": 16902
+    },
+    {
+      "epoch": 4.487189698659233,
+      "grad_norm": 0.4790539763797526,
+      "learning_rate": 7.49348931941441e-07,
+      "loss": 0.5577,
+      "step": 16903
+    },
+    {
+      "epoch": 4.487455197132617,
+      "grad_norm": 0.4792779104803753,
+      "learning_rate": 7.490997109639694e-07,
+      "loss": 0.4978,
+      "step": 16904
+    },
+    {
+      "epoch": 4.487720695606001,
+      "grad_norm": 0.483206801909153,
+      "learning_rate": 7.488505241333396e-07,
+      "loss": 0.5409,
+      "step": 16905
+    },
+    {
+      "epoch": 4.487986194079384,
+      "grad_norm": 0.4801602877983675,
+      "learning_rate": 7.48601371454413e-07,
+      "loss": 0.515,
+      "step": 16906
+    },
+    {
+      "epoch": 4.488251692552768,
+      "grad_norm": 0.4676238656834283,
+      "learning_rate": 7.483522529320478e-07,
+      "loss": 0.5146,
+      "step": 16907
+    },
+    {
+      "epoch": 4.488517191026151,
+      "grad_norm": 0.4592553050851243,
+      "learning_rate": 7.481031685711015e-07,
+      "loss": 0.503,
+      "step": 16908
+    },
+    {
+      "epoch": 4.488782689499535,
+      "grad_norm": 0.47310982065033996,
+      "learning_rate": 7.478541183764337e-07,
+      "loss": 0.5453,
+      "step": 16909
+    },
+    {
+      "epoch": 4.489048187972919,
+      "grad_norm": 0.4791007215648069,
+      "learning_rate": 7.476051023528999e-07,
+      "loss": 0.5303,
+      "step": 16910
+    },
+    {
+      "epoch": 4.489313686446303,
+      "grad_norm": 0.48350580083832817,
+      "learning_rate": 7.473561205053579e-07,
+      "loss": 0.5429,
+      "step": 16911
+    },
+    {
+      "epoch": 4.489579184919687,
+      "grad_norm": 0.48356158255585346,
+      "learning_rate": 7.471071728386628e-07,
+      "loss": 0.5611,
+      "step": 16912
+    },
+    {
+      "epoch": 4.489844683393071,
+      "grad_norm": 0.48434905581566023,
+      "learning_rate": 7.468582593576695e-07,
+      "loss": 0.5205,
+      "step": 16913
+    },
+    {
+      "epoch": 4.490110181866454,
+      "grad_norm": 0.4863816130482888,
+      "learning_rate": 7.46609380067232e-07,
+      "loss": 0.5643,
+      "step": 16914
+    },
+    {
+      "epoch": 4.490375680339838,
+      "grad_norm": 0.49023488038667695,
+      "learning_rate": 7.463605349722053e-07,
+      "loss": 0.5129,
+      "step": 16915
+    },
+    {
+      "epoch": 4.490641178813222,
+      "grad_norm": 0.5013896789594667,
+      "learning_rate": 7.46111724077441e-07,
+      "loss": 0.5697,
+      "step": 16916
+    },
+    {
+      "epoch": 4.490906677286605,
+      "grad_norm": 0.47410695460108165,
+      "learning_rate": 7.458629473877932e-07,
+      "loss": 0.5164,
+      "step": 16917
+    },
+    {
+      "epoch": 4.491172175759989,
+      "grad_norm": 0.46937514587447743,
+      "learning_rate": 7.456142049081128e-07,
+      "loss": 0.5098,
+      "step": 16918
+    },
+    {
+      "epoch": 4.4914376742333735,
+      "grad_norm": 0.48969793829860647,
+      "learning_rate": 7.453654966432499e-07,
+      "loss": 0.5548,
+      "step": 16919
+    },
+    {
+      "epoch": 4.491703172706757,
+      "grad_norm": 0.4824671567736626,
+      "learning_rate": 7.451168225980568e-07,
+      "loss": 0.5427,
+      "step": 16920
+    },
+    {
+      "epoch": 4.491968671180141,
+      "grad_norm": 0.4797311162807997,
+      "learning_rate": 7.448681827773818e-07,
+      "loss": 0.5186,
+      "step": 16921
+    },
+    {
+      "epoch": 4.492234169653525,
+      "grad_norm": 0.48393732244421994,
+      "learning_rate": 7.44619577186075e-07,
+      "loss": 0.5331,
+      "step": 16922
+    },
+    {
+      "epoch": 4.492499668126908,
+      "grad_norm": 0.5381283689486273,
+      "learning_rate": 7.443710058289838e-07,
+      "loss": 0.5393,
+      "step": 16923
+    },
+    {
+      "epoch": 4.492765166600292,
+      "grad_norm": 0.49030020167986993,
+      "learning_rate": 7.44122468710958e-07,
+      "loss": 0.5396,
+      "step": 16924
+    },
+    {
+      "epoch": 4.493030665073676,
+      "grad_norm": 0.4930901571664935,
+      "learning_rate": 7.438739658368415e-07,
+      "loss": 0.5472,
+      "step": 16925
+    },
+    {
+      "epoch": 4.493296163547059,
+      "grad_norm": 0.4826504580637794,
+      "learning_rate": 7.436254972114834e-07,
+      "loss": 0.5316,
+      "step": 16926
+    },
+    {
+      "epoch": 4.4935616620204435,
+      "grad_norm": 0.4717327993254758,
+      "learning_rate": 7.433770628397275e-07,
+      "loss": 0.5626,
+      "step": 16927
+    },
+    {
+      "epoch": 4.493827160493828,
+      "grad_norm": 0.49770538511141765,
+      "learning_rate": 7.4312866272642e-07,
+      "loss": 0.5296,
+      "step": 16928
+    },
+    {
+      "epoch": 4.494092658967211,
+      "grad_norm": 0.4963797481207869,
+      "learning_rate": 7.428802968764065e-07,
+      "loss": 0.5442,
+      "step": 16929
+    },
+    {
+      "epoch": 4.494358157440595,
+      "grad_norm": 0.4953809504470916,
+      "learning_rate": 7.42631965294528e-07,
+      "loss": 0.5343,
+      "step": 16930
+    },
+    {
+      "epoch": 4.494623655913978,
+      "grad_norm": 0.47500050298483265,
+      "learning_rate": 7.423836679856297e-07,
+      "loss": 0.5487,
+      "step": 16931
+    },
+    {
+      "epoch": 4.494889154387362,
+      "grad_norm": 0.4770482723518948,
+      "learning_rate": 7.421354049545523e-07,
+      "loss": 0.5268,
+      "step": 16932
+    },
+    {
+      "epoch": 4.495154652860746,
+      "grad_norm": 0.4908095563182328,
+      "learning_rate": 7.418871762061395e-07,
+      "loss": 0.5634,
+      "step": 16933
+    },
+    {
+      "epoch": 4.495420151334129,
+      "grad_norm": 0.4614574327736395,
+      "learning_rate": 7.416389817452305e-07,
+      "loss": 0.5092,
+      "step": 16934
+    },
+    {
+      "epoch": 4.4956856498075135,
+      "grad_norm": 0.47582695627563304,
+      "learning_rate": 7.413908215766674e-07,
+      "loss": 0.5297,
+      "step": 16935
+    },
+    {
+      "epoch": 4.495951148280898,
+      "grad_norm": 0.4783684967687841,
+      "learning_rate": 7.411426957052889e-07,
+      "loss": 0.5128,
+      "step": 16936
+    },
+    {
+      "epoch": 4.496216646754281,
+      "grad_norm": 0.49209127875958075,
+      "learning_rate": 7.408946041359335e-07,
+      "loss": 0.5599,
+      "step": 16937
+    },
+    {
+      "epoch": 4.496482145227665,
+      "grad_norm": 0.4655614755894159,
+      "learning_rate": 7.406465468734414e-07,
+      "loss": 0.5116,
+      "step": 16938
+    },
+    {
+      "epoch": 4.496747643701049,
+      "grad_norm": 0.4805901012572312,
+      "learning_rate": 7.403985239226483e-07,
+      "loss": 0.5537,
+      "step": 16939
+    },
+    {
+      "epoch": 4.497013142174432,
+      "grad_norm": 0.46882041325265494,
+      "learning_rate": 7.401505352883931e-07,
+      "loss": 0.542,
+      "step": 16940
+    },
+    {
+      "epoch": 4.497278640647816,
+      "grad_norm": 0.4761215044096941,
+      "learning_rate": 7.399025809755115e-07,
+      "loss": 0.5193,
+      "step": 16941
+    },
+    {
+      "epoch": 4.4975441391212,
+      "grad_norm": 0.4700648182285545,
+      "learning_rate": 7.396546609888386e-07,
+      "loss": 0.5467,
+      "step": 16942
+    },
+    {
+      "epoch": 4.4978096375945835,
+      "grad_norm": 0.49771163063530444,
+      "learning_rate": 7.394067753332093e-07,
+      "loss": 0.5503,
+      "step": 16943
+    },
+    {
+      "epoch": 4.498075136067968,
+      "grad_norm": 0.47450447993969835,
+      "learning_rate": 7.391589240134595e-07,
+      "loss": 0.524,
+      "step": 16944
+    },
+    {
+      "epoch": 4.498340634541352,
+      "grad_norm": 0.4813077247300755,
+      "learning_rate": 7.38911107034421e-07,
+      "loss": 0.5,
+      "step": 16945
+    },
+    {
+      "epoch": 4.498606133014735,
+      "grad_norm": 0.48201271872923684,
+      "learning_rate": 7.386633244009286e-07,
+      "loss": 0.5687,
+      "step": 16946
+    },
+    {
+      "epoch": 4.498871631488119,
+      "grad_norm": 0.49633406856176554,
+      "learning_rate": 7.384155761178139e-07,
+      "loss": 0.5351,
+      "step": 16947
+    },
+    {
+      "epoch": 4.499137129961503,
+      "grad_norm": 0.47938631256457137,
+      "learning_rate": 7.381678621899077e-07,
+      "loss": 0.5347,
+      "step": 16948
+    },
+    {
+      "epoch": 4.499402628434886,
+      "grad_norm": 0.4896837414055922,
+      "learning_rate": 7.379201826220428e-07,
+      "loss": 0.5548,
+      "step": 16949
+    },
+    {
+      "epoch": 4.49966812690827,
+      "grad_norm": 0.487473603836185,
+      "learning_rate": 7.376725374190477e-07,
+      "loss": 0.5342,
+      "step": 16950
+    },
+    {
+      "epoch": 4.499933625381654,
+      "grad_norm": 0.4811244634122249,
+      "learning_rate": 7.374249265857539e-07,
+      "loss": 0.5362,
+      "step": 16951
+    },
+    {
+      "epoch": 4.500199123855038,
+      "grad_norm": 0.4753147625143773,
+      "learning_rate": 7.371773501269896e-07,
+      "loss": 0.4855,
+      "step": 16952
+    },
+    {
+      "epoch": 4.500464622328422,
+      "grad_norm": 0.4926499510261048,
+      "learning_rate": 7.369298080475823e-07,
+      "loss": 0.528,
+      "step": 16953
+    },
+    {
+      "epoch": 4.500730120801805,
+      "grad_norm": 0.4680023665475965,
+      "learning_rate": 7.366823003523613e-07,
+      "loss": 0.528,
+      "step": 16954
+    },
+    {
+      "epoch": 4.500995619275189,
+      "grad_norm": 0.4735317292590741,
+      "learning_rate": 7.364348270461521e-07,
+      "loss": 0.5233,
+      "step": 16955
+    },
+    {
+      "epoch": 4.501261117748573,
+      "grad_norm": 0.489973539259352,
+      "learning_rate": 7.361873881337825e-07,
+      "loss": 0.5659,
+      "step": 16956
+    },
+    {
+      "epoch": 4.501526616221957,
+      "grad_norm": 0.4924248897348761,
+      "learning_rate": 7.359399836200765e-07,
+      "loss": 0.5374,
+      "step": 16957
+    },
+    {
+      "epoch": 4.50179211469534,
+      "grad_norm": 0.49282710857938555,
+      "learning_rate": 7.356926135098613e-07,
+      "loss": 0.5487,
+      "step": 16958
+    },
+    {
+      "epoch": 4.502057613168724,
+      "grad_norm": 0.49443093333366245,
+      "learning_rate": 7.354452778079588e-07,
+      "loss": 0.5323,
+      "step": 16959
+    },
+    {
+      "epoch": 4.502323111642108,
+      "grad_norm": 0.489493308546048,
+      "learning_rate": 7.351979765191944e-07,
+      "loss": 0.5599,
+      "step": 16960
+    },
+    {
+      "epoch": 4.502588610115492,
+      "grad_norm": 0.4951842396036879,
+      "learning_rate": 7.349507096483896e-07,
+      "loss": 0.5773,
+      "step": 16961
+    },
+    {
+      "epoch": 4.502854108588876,
+      "grad_norm": 0.46474732981387473,
+      "learning_rate": 7.347034772003683e-07,
+      "loss": 0.5555,
+      "step": 16962
+    },
+    {
+      "epoch": 4.50311960706226,
+      "grad_norm": 0.4797496398541281,
+      "learning_rate": 7.344562791799506e-07,
+      "loss": 0.4795,
+      "step": 16963
+    },
+    {
+      "epoch": 4.503385105535643,
+      "grad_norm": 0.4602019662658021,
+      "learning_rate": 7.34209115591959e-07,
+      "loss": 0.4744,
+      "step": 16964
+    },
+    {
+      "epoch": 4.503650604009027,
+      "grad_norm": 0.48097318012081397,
+      "learning_rate": 7.339619864412132e-07,
+      "loss": 0.5273,
+      "step": 16965
+    },
+    {
+      "epoch": 4.50391610248241,
+      "grad_norm": 0.4829890226546275,
+      "learning_rate": 7.337148917325318e-07,
+      "loss": 0.5171,
+      "step": 16966
+    },
+    {
+      "epoch": 4.5041816009557945,
+      "grad_norm": 0.4758960958572048,
+      "learning_rate": 7.334678314707352e-07,
+      "loss": 0.522,
+      "step": 16967
+    },
+    {
+      "epoch": 4.5044470994291785,
+      "grad_norm": 0.48494206463868483,
+      "learning_rate": 7.332208056606407e-07,
+      "loss": 0.5436,
+      "step": 16968
+    },
+    {
+      "epoch": 4.504712597902562,
+      "grad_norm": 0.46903458364130224,
+      "learning_rate": 7.329738143070669e-07,
+      "loss": 0.5391,
+      "step": 16969
+    },
+    {
+      "epoch": 4.504978096375946,
+      "grad_norm": 0.4727224555518883,
+      "learning_rate": 7.327268574148303e-07,
+      "loss": 0.5441,
+      "step": 16970
+    },
+    {
+      "epoch": 4.50524359484933,
+      "grad_norm": 0.4681686313130701,
+      "learning_rate": 7.324799349887468e-07,
+      "loss": 0.5221,
+      "step": 16971
+    },
+    {
+      "epoch": 4.505509093322713,
+      "grad_norm": 0.461541023757282,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.5193,
+      "step": 16972
+    },
+    {
+      "epoch": 4.505774591796097,
+      "grad_norm": 0.4709099460401255,
+      "learning_rate": 7.319861935542999e-07,
+      "loss": 0.5255,
+      "step": 16973
+    },
+    {
+      "epoch": 4.506040090269481,
+      "grad_norm": 0.48224645745410866,
+      "learning_rate": 7.317393745555673e-07,
+      "loss": 0.5185,
+      "step": 16974
+    },
+    {
+      "epoch": 4.5063055887428645,
+      "grad_norm": 0.46845469518806615,
+      "learning_rate": 7.314925900422462e-07,
+      "loss": 0.5175,
+      "step": 16975
+    },
+    {
+      "epoch": 4.5065710872162486,
+      "grad_norm": 0.48838061835994057,
+      "learning_rate": 7.3124584001915e-07,
+      "loss": 0.5107,
+      "step": 16976
+    },
+    {
+      "epoch": 4.506836585689633,
+      "grad_norm": 0.48566651021651097,
+      "learning_rate": 7.309991244910894e-07,
+      "loss": 0.5332,
+      "step": 16977
+    },
+    {
+      "epoch": 4.507102084163016,
+      "grad_norm": 0.4756539894724518,
+      "learning_rate": 7.307524434628782e-07,
+      "loss": 0.5055,
+      "step": 16978
+    },
+    {
+      "epoch": 4.5073675826364,
+      "grad_norm": 0.49749365224412573,
+      "learning_rate": 7.305057969393253e-07,
+      "loss": 0.5689,
+      "step": 16979
+    },
+    {
+      "epoch": 4.507633081109784,
+      "grad_norm": 0.4995769357074368,
+      "learning_rate": 7.302591849252424e-07,
+      "loss": 0.559,
+      "step": 16980
+    },
+    {
+      "epoch": 4.507898579583167,
+      "grad_norm": 0.4917647366377767,
+      "learning_rate": 7.300126074254388e-07,
+      "loss": 0.5296,
+      "step": 16981
+    },
+    {
+      "epoch": 4.508164078056551,
+      "grad_norm": 0.47123873377533154,
+      "learning_rate": 7.297660644447222e-07,
+      "loss": 0.5372,
+      "step": 16982
+    },
+    {
+      "epoch": 4.5084295765299345,
+      "grad_norm": 0.4895773710019,
+      "learning_rate": 7.295195559879023e-07,
+      "loss": 0.5179,
+      "step": 16983
+    },
+    {
+      "epoch": 4.508695075003319,
+      "grad_norm": 0.48024350512628794,
+      "learning_rate": 7.292730820597851e-07,
+      "loss": 0.54,
+      "step": 16984
+    },
+    {
+      "epoch": 4.508960573476703,
+      "grad_norm": 0.4904819133172295,
+      "learning_rate": 7.290266426651793e-07,
+      "loss": 0.5753,
+      "step": 16985
+    },
+    {
+      "epoch": 4.509226071950087,
+      "grad_norm": 0.4815809186989495,
+      "learning_rate": 7.28780237808889e-07,
+      "loss": 0.5578,
+      "step": 16986
+    },
+    {
+      "epoch": 4.50949157042347,
+      "grad_norm": 0.49296871762187355,
+      "learning_rate": 7.285338674957227e-07,
+      "loss": 0.5094,
+      "step": 16987
+    },
+    {
+      "epoch": 4.509757068896854,
+      "grad_norm": 0.479858091462161,
+      "learning_rate": 7.282875317304814e-07,
+      "loss": 0.5331,
+      "step": 16988
+    },
+    {
+      "epoch": 4.510022567370237,
+      "grad_norm": 0.47403651661129304,
+      "learning_rate": 7.280412305179721e-07,
+      "loss": 0.5272,
+      "step": 16989
+    },
+    {
+      "epoch": 4.510288065843621,
+      "grad_norm": 0.4794431122574388,
+      "learning_rate": 7.277949638629963e-07,
+      "loss": 0.5348,
+      "step": 16990
+    },
+    {
+      "epoch": 4.510553564317005,
+      "grad_norm": 0.480067006900869,
+      "learning_rate": 7.275487317703587e-07,
+      "loss": 0.5241,
+      "step": 16991
+    },
+    {
+      "epoch": 4.510819062790389,
+      "grad_norm": 0.4782274469247924,
+      "learning_rate": 7.273025342448597e-07,
+      "loss": 0.531,
+      "step": 16992
+    },
+    {
+      "epoch": 4.511084561263773,
+      "grad_norm": 0.4826277832994021,
+      "learning_rate": 7.270563712913026e-07,
+      "loss": 0.5217,
+      "step": 16993
+    },
+    {
+      "epoch": 4.511350059737157,
+      "grad_norm": 0.47191955247493816,
+      "learning_rate": 7.268102429144869e-07,
+      "loss": 0.5001,
+      "step": 16994
+    },
+    {
+      "epoch": 4.51161555821054,
+      "grad_norm": 0.4883096277948197,
+      "learning_rate": 7.265641491192124e-07,
+      "loss": 0.5446,
+      "step": 16995
+    },
+    {
+      "epoch": 4.511881056683924,
+      "grad_norm": 0.49203822214061277,
+      "learning_rate": 7.263180899102798e-07,
+      "loss": 0.5356,
+      "step": 16996
+    },
+    {
+      "epoch": 4.512146555157308,
+      "grad_norm": 0.47464331347661587,
+      "learning_rate": 7.260720652924863e-07,
+      "loss": 0.5382,
+      "step": 16997
+    },
+    {
+      "epoch": 4.512412053630691,
+      "grad_norm": 0.4838217958530632,
+      "learning_rate": 7.258260752706319e-07,
+      "loss": 0.5745,
+      "step": 16998
+    },
+    {
+      "epoch": 4.512677552104075,
+      "grad_norm": 0.4959804291619121,
+      "learning_rate": 7.255801198495127e-07,
+      "loss": 0.5249,
+      "step": 16999
+    },
+    {
+      "epoch": 4.5129430505774595,
+      "grad_norm": 0.4773349809650804,
+      "learning_rate": 7.253341990339252e-07,
+      "loss": 0.5132,
+      "step": 17000
+    },
+    {
+      "epoch": 4.513208549050843,
+      "grad_norm": 0.49146660998110225,
+      "learning_rate": 7.250883128286665e-07,
+      "loss": 0.5209,
+      "step": 17001
+    },
+    {
+      "epoch": 4.513474047524227,
+      "grad_norm": 0.4884867978572633,
+      "learning_rate": 7.248424612385305e-07,
+      "loss": 0.5415,
+      "step": 17002
+    },
+    {
+      "epoch": 4.513739545997611,
+      "grad_norm": 0.4915365845990471,
+      "learning_rate": 7.245966442683139e-07,
+      "loss": 0.5302,
+      "step": 17003
+    },
+    {
+      "epoch": 4.514005044470994,
+      "grad_norm": 0.48480243727730016,
+      "learning_rate": 7.243508619228095e-07,
+      "loss": 0.5269,
+      "step": 17004
+    },
+    {
+      "epoch": 4.514270542944378,
+      "grad_norm": 0.4810746972375595,
+      "learning_rate": 7.241051142068109e-07,
+      "loss": 0.5291,
+      "step": 17005
+    },
+    {
+      "epoch": 4.514536041417762,
+      "grad_norm": 0.4878136577160576,
+      "learning_rate": 7.238594011251101e-07,
+      "loss": 0.5066,
+      "step": 17006
+    },
+    {
+      "epoch": 4.514801539891145,
+      "grad_norm": 0.4767093077299178,
+      "learning_rate": 7.236137226825001e-07,
+      "loss": 0.5697,
+      "step": 17007
+    },
+    {
+      "epoch": 4.5150670383645295,
+      "grad_norm": 0.47739927779179636,
+      "learning_rate": 7.233680788837716e-07,
+      "loss": 0.5302,
+      "step": 17008
+    },
+    {
+      "epoch": 4.515332536837914,
+      "grad_norm": 0.485306985148506,
+      "learning_rate": 7.231224697337161e-07,
+      "loss": 0.559,
+      "step": 17009
+    },
+    {
+      "epoch": 4.515598035311297,
+      "grad_norm": 0.49644838930471236,
+      "learning_rate": 7.228768952371226e-07,
+      "loss": 0.5697,
+      "step": 17010
+    },
+    {
+      "epoch": 4.515863533784681,
+      "grad_norm": 0.4815247460210243,
+      "learning_rate": 7.226313553987805e-07,
+      "loss": 0.5488,
+      "step": 17011
+    },
+    {
+      "epoch": 4.516129032258064,
+      "grad_norm": 0.47162127896877176,
+      "learning_rate": 7.223858502234792e-07,
+      "loss": 0.4941,
+      "step": 17012
+    },
+    {
+      "epoch": 4.516394530731448,
+      "grad_norm": 0.46595440916367403,
+      "learning_rate": 7.221403797160056e-07,
+      "loss": 0.4967,
+      "step": 17013
+    },
+    {
+      "epoch": 4.516660029204832,
+      "grad_norm": 0.48137482716616214,
+      "learning_rate": 7.218949438811482e-07,
+      "loss": 0.5563,
+      "step": 17014
+    },
+    {
+      "epoch": 4.516925527678216,
+      "grad_norm": 0.4711263560496944,
+      "learning_rate": 7.216495427236922e-07,
+      "loss": 0.519,
+      "step": 17015
+    },
+    {
+      "epoch": 4.5171910261515995,
+      "grad_norm": 0.49741487686336977,
+      "learning_rate": 7.214041762484259e-07,
+      "loss": 0.5356,
+      "step": 17016
+    },
+    {
+      "epoch": 4.517456524624984,
+      "grad_norm": 0.4713400802878359,
+      "learning_rate": 7.211588444601311e-07,
+      "loss": 0.5363,
+      "step": 17017
+    },
+    {
+      "epoch": 4.517722023098367,
+      "grad_norm": 0.4728650057350629,
+      "learning_rate": 7.209135473635945e-07,
+      "loss": 0.4862,
+      "step": 17018
+    },
+    {
+      "epoch": 4.517987521571751,
+      "grad_norm": 0.48674330140410627,
+      "learning_rate": 7.206682849636001e-07,
+      "loss": 0.5257,
+      "step": 17019
+    },
+    {
+      "epoch": 4.518253020045135,
+      "grad_norm": 0.4776414610147334,
+      "learning_rate": 7.204230572649304e-07,
+      "loss": 0.5427,
+      "step": 17020
+    },
+    {
+      "epoch": 4.518518518518518,
+      "grad_norm": 0.48051128268866045,
+      "learning_rate": 7.201778642723694e-07,
+      "loss": 0.5468,
+      "step": 17021
+    },
+    {
+      "epoch": 4.518784016991902,
+      "grad_norm": 0.4813400816146693,
+      "learning_rate": 7.199327059906963e-07,
+      "loss": 0.5598,
+      "step": 17022
+    },
+    {
+      "epoch": 4.519049515465286,
+      "grad_norm": 0.48261664746660016,
+      "learning_rate": 7.196875824246946e-07,
+      "loss": 0.5372,
+      "step": 17023
+    },
+    {
+      "epoch": 4.5193150139386695,
+      "grad_norm": 0.47203144796706337,
+      "learning_rate": 7.194424935791434e-07,
+      "loss": 0.5166,
+      "step": 17024
+    },
+    {
+      "epoch": 4.519580512412054,
+      "grad_norm": 0.4971673053623054,
+      "learning_rate": 7.191974394588238e-07,
+      "loss": 0.5361,
+      "step": 17025
+    },
+    {
+      "epoch": 4.519846010885438,
+      "grad_norm": 0.47271344208986416,
+      "learning_rate": 7.189524200685135e-07,
+      "loss": 0.514,
+      "step": 17026
+    },
+    {
+      "epoch": 4.520111509358821,
+      "grad_norm": 0.4901810751028744,
+      "learning_rate": 7.187074354129925e-07,
+      "loss": 0.5562,
+      "step": 17027
+    },
+    {
+      "epoch": 4.520377007832205,
+      "grad_norm": 0.4668910811460749,
+      "learning_rate": 7.184624854970379e-07,
+      "loss": 0.5202,
+      "step": 17028
+    },
+    {
+      "epoch": 4.520642506305589,
+      "grad_norm": 0.4803585779873794,
+      "learning_rate": 7.18217570325426e-07,
+      "loss": 0.5273,
+      "step": 17029
+    },
+    {
+      "epoch": 4.520908004778972,
+      "grad_norm": 0.4665387721606301,
+      "learning_rate": 7.179726899029349e-07,
+      "loss": 0.5292,
+      "step": 17030
+    },
+    {
+      "epoch": 4.521173503252356,
+      "grad_norm": 0.48034290046347355,
+      "learning_rate": 7.177278442343386e-07,
+      "loss": 0.547,
+      "step": 17031
+    },
+    {
+      "epoch": 4.52143900172574,
+      "grad_norm": 0.4872394285217787,
+      "learning_rate": 7.17483033324414e-07,
+      "loss": 0.5685,
+      "step": 17032
+    },
+    {
+      "epoch": 4.521704500199124,
+      "grad_norm": 0.49518086525454763,
+      "learning_rate": 7.172382571779344e-07,
+      "loss": 0.5485,
+      "step": 17033
+    },
+    {
+      "epoch": 4.521969998672508,
+      "grad_norm": 0.4903048099403242,
+      "learning_rate": 7.169935157996738e-07,
+      "loss": 0.5282,
+      "step": 17034
+    },
+    {
+      "epoch": 4.522235497145892,
+      "grad_norm": 0.4977164194726281,
+      "learning_rate": 7.167488091944044e-07,
+      "loss": 0.5286,
+      "step": 17035
+    },
+    {
+      "epoch": 4.522500995619275,
+      "grad_norm": 0.4860800785834062,
+      "learning_rate": 7.165041373669005e-07,
+      "loss": 0.5449,
+      "step": 17036
+    },
+    {
+      "epoch": 4.522766494092659,
+      "grad_norm": 0.48135710568921836,
+      "learning_rate": 7.162595003219314e-07,
+      "loss": 0.5034,
+      "step": 17037
+    },
+    {
+      "epoch": 4.523031992566043,
+      "grad_norm": 0.48533795912033045,
+      "learning_rate": 7.160148980642706e-07,
+      "loss": 0.542,
+      "step": 17038
+    },
+    {
+      "epoch": 4.523297491039426,
+      "grad_norm": 0.4777820063167383,
+      "learning_rate": 7.157703305986868e-07,
+      "loss": 0.5359,
+      "step": 17039
+    },
+    {
+      "epoch": 4.52356298951281,
+      "grad_norm": 0.492223486819021,
+      "learning_rate": 7.155257979299496e-07,
+      "loss": 0.5068,
+      "step": 17040
+    },
+    {
+      "epoch": 4.523828487986194,
+      "grad_norm": 0.4876367816458037,
+      "learning_rate": 7.152813000628292e-07,
+      "loss": 0.5352,
+      "step": 17041
+    },
+    {
+      "epoch": 4.524093986459578,
+      "grad_norm": 0.46748608766400057,
+      "learning_rate": 7.150368370020924e-07,
+      "loss": 0.5236,
+      "step": 17042
+    },
+    {
+      "epoch": 4.524359484932962,
+      "grad_norm": 0.4858141671373916,
+      "learning_rate": 7.147924087525085e-07,
+      "loss": 0.5531,
+      "step": 17043
+    },
+    {
+      "epoch": 4.524624983406346,
+      "grad_norm": 0.4766927055743187,
+      "learning_rate": 7.145480153188428e-07,
+      "loss": 0.514,
+      "step": 17044
+    },
+    {
+      "epoch": 4.524890481879729,
+      "grad_norm": 0.480040758558741,
+      "learning_rate": 7.143036567058631e-07,
+      "loss": 0.5225,
+      "step": 17045
+    },
+    {
+      "epoch": 4.525155980353113,
+      "grad_norm": 0.4814218686804199,
+      "learning_rate": 7.14059332918334e-07,
+      "loss": 0.5629,
+      "step": 17046
+    },
+    {
+      "epoch": 4.525421478826496,
+      "grad_norm": 0.4749218711190256,
+      "learning_rate": 7.138150439610201e-07,
+      "loss": 0.5262,
+      "step": 17047
+    },
+    {
+      "epoch": 4.5256869772998805,
+      "grad_norm": 0.48041544151665233,
+      "learning_rate": 7.135707898386871e-07,
+      "loss": 0.5318,
+      "step": 17048
+    },
+    {
+      "epoch": 4.5259524757732645,
+      "grad_norm": 0.47582984384942173,
+      "learning_rate": 7.133265705560966e-07,
+      "loss": 0.5567,
+      "step": 17049
+    },
+    {
+      "epoch": 4.526217974246648,
+      "grad_norm": 0.47671511484036394,
+      "learning_rate": 7.130823861180139e-07,
+      "loss": 0.5374,
+      "step": 17050
+    },
+    {
+      "epoch": 4.526483472720032,
+      "grad_norm": 0.4609036756337673,
+      "learning_rate": 7.128382365291986e-07,
+      "loss": 0.5277,
+      "step": 17051
+    },
+    {
+      "epoch": 4.526748971193416,
+      "grad_norm": 0.4839139772650828,
+      "learning_rate": 7.12594121794414e-07,
+      "loss": 0.5269,
+      "step": 17052
+    },
+    {
+      "epoch": 4.527014469666799,
+      "grad_norm": 0.4833401243595286,
+      "learning_rate": 7.123500419184196e-07,
+      "loss": 0.5344,
+      "step": 17053
+    },
+    {
+      "epoch": 4.527279968140183,
+      "grad_norm": 0.482129850937298,
+      "learning_rate": 7.121059969059771e-07,
+      "loss": 0.5562,
+      "step": 17054
+    },
+    {
+      "epoch": 4.527545466613567,
+      "grad_norm": 0.47768283695240804,
+      "learning_rate": 7.118619867618445e-07,
+      "loss": 0.4703,
+      "step": 17055
+    },
+    {
+      "epoch": 4.5278109650869505,
+      "grad_norm": 0.4880094429733977,
+      "learning_rate": 7.11618011490782e-07,
+      "loss": 0.5612,
+      "step": 17056
+    },
+    {
+      "epoch": 4.5280764635603346,
+      "grad_norm": 0.48673600528819844,
+      "learning_rate": 7.113740710975467e-07,
+      "loss": 0.5453,
+      "step": 17057
+    },
+    {
+      "epoch": 4.528341962033719,
+      "grad_norm": 0.4924967704691144,
+      "learning_rate": 7.11130165586896e-07,
+      "loss": 0.5359,
+      "step": 17058
+    },
+    {
+      "epoch": 4.528607460507102,
+      "grad_norm": 0.47139587850695874,
+      "learning_rate": 7.108862949635875e-07,
+      "loss": 0.5141,
+      "step": 17059
+    },
+    {
+      "epoch": 4.528872958980486,
+      "grad_norm": 0.4915602756778128,
+      "learning_rate": 7.106424592323761e-07,
+      "loss": 0.5446,
+      "step": 17060
+    },
+    {
+      "epoch": 4.52913845745387,
+      "grad_norm": 0.4658447919188393,
+      "learning_rate": 7.103986583980186e-07,
+      "loss": 0.4881,
+      "step": 17061
+    },
+    {
+      "epoch": 4.529403955927253,
+      "grad_norm": 0.48436827616519074,
+      "learning_rate": 7.101548924652688e-07,
+      "loss": 0.5206,
+      "step": 17062
+    },
+    {
+      "epoch": 4.529669454400637,
+      "grad_norm": 0.47446860303135163,
+      "learning_rate": 7.099111614388809e-07,
+      "loss": 0.5444,
+      "step": 17063
+    },
+    {
+      "epoch": 4.5299349528740205,
+      "grad_norm": 0.47647728961568964,
+      "learning_rate": 7.096674653236077e-07,
+      "loss": 0.5382,
+      "step": 17064
+    },
+    {
+      "epoch": 4.530200451347405,
+      "grad_norm": 0.47482278007781303,
+      "learning_rate": 7.094238041242021e-07,
+      "loss": 0.5428,
+      "step": 17065
+    },
+    {
+      "epoch": 4.530465949820789,
+      "grad_norm": 0.479764788278246,
+      "learning_rate": 7.091801778454175e-07,
+      "loss": 0.4763,
+      "step": 17066
+    },
+    {
+      "epoch": 4.530731448294173,
+      "grad_norm": 0.47913954189425495,
+      "learning_rate": 7.089365864920031e-07,
+      "loss": 0.5285,
+      "step": 17067
+    },
+    {
+      "epoch": 4.530996946767556,
+      "grad_norm": 0.48287101917509945,
+      "learning_rate": 7.086930300687123e-07,
+      "loss": 0.5479,
+      "step": 17068
+    },
+    {
+      "epoch": 4.53126244524094,
+      "grad_norm": 0.4791178092018074,
+      "learning_rate": 7.084495085802917e-07,
+      "loss": 0.5309,
+      "step": 17069
+    },
+    {
+      "epoch": 4.531527943714323,
+      "grad_norm": 0.4712089037534123,
+      "learning_rate": 7.082060220314929e-07,
+      "loss": 0.5465,
+      "step": 17070
+    },
+    {
+      "epoch": 4.531793442187707,
+      "grad_norm": 0.4861988046991871,
+      "learning_rate": 7.079625704270629e-07,
+      "loss": 0.5649,
+      "step": 17071
+    },
+    {
+      "epoch": 4.532058940661091,
+      "grad_norm": 0.48825746295085554,
+      "learning_rate": 7.077191537717512e-07,
+      "loss": 0.5585,
+      "step": 17072
+    },
+    {
+      "epoch": 4.5323244391344755,
+      "grad_norm": 0.47734541682794585,
+      "learning_rate": 7.074757720703044e-07,
+      "loss": 0.5352,
+      "step": 17073
+    },
+    {
+      "epoch": 4.532589937607859,
+      "grad_norm": 0.48806459323741086,
+      "learning_rate": 7.072324253274679e-07,
+      "loss": 0.5256,
+      "step": 17074
+    },
+    {
+      "epoch": 4.532855436081243,
+      "grad_norm": 0.49030320511559183,
+      "learning_rate": 7.069891135479895e-07,
+      "loss": 0.553,
+      "step": 17075
+    },
+    {
+      "epoch": 4.533120934554626,
+      "grad_norm": 0.4836595261590822,
+      "learning_rate": 7.067458367366125e-07,
+      "loss": 0.5393,
+      "step": 17076
+    },
+    {
+      "epoch": 4.53338643302801,
+      "grad_norm": 0.4902939498173901,
+      "learning_rate": 7.065025948980833e-07,
+      "loss": 0.5689,
+      "step": 17077
+    },
+    {
+      "epoch": 4.533651931501394,
+      "grad_norm": 0.5036341596647592,
+      "learning_rate": 7.062593880371441e-07,
+      "loss": 0.5557,
+      "step": 17078
+    },
+    {
+      "epoch": 4.533917429974777,
+      "grad_norm": 0.47971191567358085,
+      "learning_rate": 7.060162161585398e-07,
+      "loss": 0.5065,
+      "step": 17079
+    },
+    {
+      "epoch": 4.534182928448161,
+      "grad_norm": 0.49103041316689,
+      "learning_rate": 7.057730792670106e-07,
+      "loss": 0.5719,
+      "step": 17080
+    },
+    {
+      "epoch": 4.5344484269215455,
+      "grad_norm": 0.48625113524707025,
+      "learning_rate": 7.055299773672999e-07,
+      "loss": 0.5408,
+      "step": 17081
+    },
+    {
+      "epoch": 4.534713925394929,
+      "grad_norm": 0.46214254661361515,
+      "learning_rate": 7.052869104641475e-07,
+      "loss": 0.5159,
+      "step": 17082
+    },
+    {
+      "epoch": 4.534979423868313,
+      "grad_norm": 0.4862982622162461,
+      "learning_rate": 7.050438785622954e-07,
+      "loss": 0.539,
+      "step": 17083
+    },
+    {
+      "epoch": 4.535244922341697,
+      "grad_norm": 0.47196227991581247,
+      "learning_rate": 7.048008816664817e-07,
+      "loss": 0.5266,
+      "step": 17084
+    },
+    {
+      "epoch": 4.53551042081508,
+      "grad_norm": 0.4783948465096514,
+      "learning_rate": 7.045579197814467e-07,
+      "loss": 0.5951,
+      "step": 17085
+    },
+    {
+      "epoch": 4.535775919288464,
+      "grad_norm": 0.4845641896880734,
+      "learning_rate": 7.043149929119286e-07,
+      "loss": 0.4968,
+      "step": 17086
+    },
+    {
+      "epoch": 4.536041417761848,
+      "grad_norm": 0.48431994506276194,
+      "learning_rate": 7.040721010626639e-07,
+      "loss": 0.5326,
+      "step": 17087
+    },
+    {
+      "epoch": 4.536306916235231,
+      "grad_norm": 0.49026023733211976,
+      "learning_rate": 7.038292442383912e-07,
+      "loss": 0.5431,
+      "step": 17088
+    },
+    {
+      "epoch": 4.5365724147086155,
+      "grad_norm": 0.4765346801577704,
+      "learning_rate": 7.035864224438452e-07,
+      "loss": 0.54,
+      "step": 17089
+    },
+    {
+      "epoch": 4.536837913182,
+      "grad_norm": 0.4779254366833416,
+      "learning_rate": 7.033436356837631e-07,
+      "loss": 0.5125,
+      "step": 17090
+    },
+    {
+      "epoch": 4.537103411655383,
+      "grad_norm": 0.48547530478247264,
+      "learning_rate": 7.031008839628792e-07,
+      "loss": 0.5396,
+      "step": 17091
+    },
+    {
+      "epoch": 4.537368910128767,
+      "grad_norm": 0.4954785495080932,
+      "learning_rate": 7.028581672859266e-07,
+      "loss": 0.5416,
+      "step": 17092
+    },
+    {
+      "epoch": 4.53763440860215,
+      "grad_norm": 0.47137601000978313,
+      "learning_rate": 7.026154856576408e-07,
+      "loss": 0.5232,
+      "step": 17093
+    },
+    {
+      "epoch": 4.537899907075534,
+      "grad_norm": 0.4948690525051935,
+      "learning_rate": 7.023728390827531e-07,
+      "loss": 0.5751,
+      "step": 17094
+    },
+    {
+      "epoch": 4.538165405548918,
+      "grad_norm": 0.486729437993439,
+      "learning_rate": 7.021302275659972e-07,
+      "loss": 0.5492,
+      "step": 17095
+    },
+    {
+      "epoch": 4.538430904022302,
+      "grad_norm": 0.4717125563238074,
+      "learning_rate": 7.018876511121036e-07,
+      "loss": 0.5011,
+      "step": 17096
+    },
+    {
+      "epoch": 4.5386964024956855,
+      "grad_norm": 0.482055994808619,
+      "learning_rate": 7.016451097258034e-07,
+      "loss": 0.5629,
+      "step": 17097
+    },
+    {
+      "epoch": 4.53896190096907,
+      "grad_norm": 0.4711325570521054,
+      "learning_rate": 7.01402603411826e-07,
+      "loss": 0.5102,
+      "step": 17098
+    },
+    {
+      "epoch": 4.539227399442453,
+      "grad_norm": 0.4665266406850192,
+      "learning_rate": 7.011601321749023e-07,
+      "loss": 0.5211,
+      "step": 17099
+    },
+    {
+      "epoch": 4.539492897915837,
+      "grad_norm": 0.47348041357416615,
+      "learning_rate": 7.009176960197598e-07,
+      "loss": 0.5372,
+      "step": 17100
+    },
+    {
+      "epoch": 4.539758396389221,
+      "grad_norm": 0.48600806257825885,
+      "learning_rate": 7.006752949511276e-07,
+      "loss": 0.5094,
+      "step": 17101
+    },
+    {
+      "epoch": 4.540023894862604,
+      "grad_norm": 0.4787955309567304,
+      "learning_rate": 7.004329289737327e-07,
+      "loss": 0.5479,
+      "step": 17102
+    },
+    {
+      "epoch": 4.540289393335988,
+      "grad_norm": 0.48966247462132756,
+      "learning_rate": 7.00190598092301e-07,
+      "loss": 0.5503,
+      "step": 17103
+    },
+    {
+      "epoch": 4.540554891809372,
+      "grad_norm": 0.4630849057938057,
+      "learning_rate": 6.999483023115603e-07,
+      "loss": 0.5174,
+      "step": 17104
+    },
+    {
+      "epoch": 4.5408203902827555,
+      "grad_norm": 0.47101241465481675,
+      "learning_rate": 6.997060416362339e-07,
+      "loss": 0.5232,
+      "step": 17105
+    },
+    {
+      "epoch": 4.54108588875614,
+      "grad_norm": 0.4783609951345164,
+      "learning_rate": 6.994638160710488e-07,
+      "loss": 0.5467,
+      "step": 17106
+    },
+    {
+      "epoch": 4.541351387229524,
+      "grad_norm": 0.49704181747810544,
+      "learning_rate": 6.992216256207265e-07,
+      "loss": 0.5433,
+      "step": 17107
+    },
+    {
+      "epoch": 4.541616885702907,
+      "grad_norm": 0.4855917038479662,
+      "learning_rate": 6.989794702899932e-07,
+      "loss": 0.5552,
+      "step": 17108
+    },
+    {
+      "epoch": 4.541882384176291,
+      "grad_norm": 0.47130637007583176,
+      "learning_rate": 6.987373500835687e-07,
+      "loss": 0.5371,
+      "step": 17109
+    },
+    {
+      "epoch": 4.542147882649675,
+      "grad_norm": 0.4814510693213958,
+      "learning_rate": 6.984952650061758e-07,
+      "loss": 0.5333,
+      "step": 17110
+    },
+    {
+      "epoch": 4.542413381123058,
+      "grad_norm": 0.4806272355497954,
+      "learning_rate": 6.98253215062537e-07,
+      "loss": 0.5126,
+      "step": 17111
+    },
+    {
+      "epoch": 4.542678879596442,
+      "grad_norm": 0.48588513037576586,
+      "learning_rate": 6.980112002573711e-07,
+      "loss": 0.5507,
+      "step": 17112
+    },
+    {
+      "epoch": 4.542944378069826,
+      "grad_norm": 0.4791730005141744,
+      "learning_rate": 6.977692205953995e-07,
+      "loss": 0.5295,
+      "step": 17113
+    },
+    {
+      "epoch": 4.54320987654321,
+      "grad_norm": 0.5072940205877899,
+      "learning_rate": 6.975272760813409e-07,
+      "loss": 0.5609,
+      "step": 17114
+    },
+    {
+      "epoch": 4.543475375016594,
+      "grad_norm": 0.48647322768105367,
+      "learning_rate": 6.972853667199137e-07,
+      "loss": 0.5542,
+      "step": 17115
+    },
+    {
+      "epoch": 4.543740873489978,
+      "grad_norm": 0.4934932090540298,
+      "learning_rate": 6.970434925158346e-07,
+      "loss": 0.5385,
+      "step": 17116
+    },
+    {
+      "epoch": 4.544006371963361,
+      "grad_norm": 0.4908918802037143,
+      "learning_rate": 6.968016534738225e-07,
+      "loss": 0.5405,
+      "step": 17117
+    },
+    {
+      "epoch": 4.544271870436745,
+      "grad_norm": 0.4806260677079348,
+      "learning_rate": 6.965598495985929e-07,
+      "loss": 0.5518,
+      "step": 17118
+    },
+    {
+      "epoch": 4.544537368910129,
+      "grad_norm": 0.4652201748869212,
+      "learning_rate": 6.963180808948622e-07,
+      "loss": 0.5197,
+      "step": 17119
+    },
+    {
+      "epoch": 4.544802867383512,
+      "grad_norm": 0.485869148445642,
+      "learning_rate": 6.960763473673451e-07,
+      "loss": 0.5676,
+      "step": 17120
+    },
+    {
+      "epoch": 4.545068365856896,
+      "grad_norm": 0.4866266573608137,
+      "learning_rate": 6.958346490207551e-07,
+      "loss": 0.5612,
+      "step": 17121
+    },
+    {
+      "epoch": 4.54533386433028,
+      "grad_norm": 0.4756666381204711,
+      "learning_rate": 6.955929858598079e-07,
+      "loss": 0.533,
+      "step": 17122
+    },
+    {
+      "epoch": 4.545599362803664,
+      "grad_norm": 0.48016387869433635,
+      "learning_rate": 6.953513578892146e-07,
+      "loss": 0.5311,
+      "step": 17123
+    },
+    {
+      "epoch": 4.545864861277048,
+      "grad_norm": 0.4687052844037668,
+      "learning_rate": 6.95109765113689e-07,
+      "loss": 0.5315,
+      "step": 17124
+    },
+    {
+      "epoch": 4.546130359750432,
+      "grad_norm": 0.5048588431508544,
+      "learning_rate": 6.94868207537942e-07,
+      "loss": 0.5447,
+      "step": 17125
+    },
+    {
+      "epoch": 4.546395858223815,
+      "grad_norm": 0.4958975491939647,
+      "learning_rate": 6.94626685166685e-07,
+      "loss": 0.5262,
+      "step": 17126
+    },
+    {
+      "epoch": 4.546661356697199,
+      "grad_norm": 0.4879210798133318,
+      "learning_rate": 6.943851980046271e-07,
+      "loss": 0.5432,
+      "step": 17127
+    },
+    {
+      "epoch": 4.546926855170582,
+      "grad_norm": 0.4885303984486785,
+      "learning_rate": 6.941437460564792e-07,
+      "loss": 0.5509,
+      "step": 17128
+    },
+    {
+      "epoch": 4.5471923536439665,
+      "grad_norm": 0.483111641145223,
+      "learning_rate": 6.939023293269492e-07,
+      "loss": 0.5314,
+      "step": 17129
+    },
+    {
+      "epoch": 4.5474578521173505,
+      "grad_norm": 0.49041160091805946,
+      "learning_rate": 6.936609478207459e-07,
+      "loss": 0.5061,
+      "step": 17130
+    },
+    {
+      "epoch": 4.547723350590734,
+      "grad_norm": 0.46582950451098615,
+      "learning_rate": 6.934196015425784e-07,
+      "loss": 0.5455,
+      "step": 17131
+    },
+    {
+      "epoch": 4.547988849064118,
+      "grad_norm": 0.4791485881221155,
+      "learning_rate": 6.931782904971504e-07,
+      "loss": 0.5526,
+      "step": 17132
+    },
+    {
+      "epoch": 4.548254347537502,
+      "grad_norm": 0.4882908665359969,
+      "learning_rate": 6.929370146891703e-07,
+      "loss": 0.5401,
+      "step": 17133
+    },
+    {
+      "epoch": 4.548519846010885,
+      "grad_norm": 0.4786581692068963,
+      "learning_rate": 6.926957741233419e-07,
+      "loss": 0.5199,
+      "step": 17134
+    },
+    {
+      "epoch": 4.548785344484269,
+      "grad_norm": 0.4903894802705,
+      "learning_rate": 6.924545688043721e-07,
+      "loss": 0.5448,
+      "step": 17135
+    },
+    {
+      "epoch": 4.549050842957653,
+      "grad_norm": 0.4859267182188451,
+      "learning_rate": 6.922133987369628e-07,
+      "loss": 0.5675,
+      "step": 17136
+    },
+    {
+      "epoch": 4.5493163414310365,
+      "grad_norm": 0.45543675142027945,
+      "learning_rate": 6.919722639258195e-07,
+      "loss": 0.5333,
+      "step": 17137
+    },
+    {
+      "epoch": 4.5495818399044206,
+      "grad_norm": 0.4870302225979478,
+      "learning_rate": 6.917311643756436e-07,
+      "loss": 0.5299,
+      "step": 17138
+    },
+    {
+      "epoch": 4.549847338377805,
+      "grad_norm": 0.48684877157998746,
+      "learning_rate": 6.914901000911367e-07,
+      "loss": 0.5197,
+      "step": 17139
+    },
+    {
+      "epoch": 4.550112836851188,
+      "grad_norm": 0.4757733777122875,
+      "learning_rate": 6.912490710770015e-07,
+      "loss": 0.5224,
+      "step": 17140
+    },
+    {
+      "epoch": 4.550378335324572,
+      "grad_norm": 0.4875327884697538,
+      "learning_rate": 6.910080773379374e-07,
+      "loss": 0.5312,
+      "step": 17141
+    },
+    {
+      "epoch": 4.550643833797956,
+      "grad_norm": 0.4790581017913816,
+      "learning_rate": 6.907671188786463e-07,
+      "loss": 0.5534,
+      "step": 17142
+    },
+    {
+      "epoch": 4.550909332271339,
+      "grad_norm": 0.48079907762830476,
+      "learning_rate": 6.90526195703825e-07,
+      "loss": 0.5699,
+      "step": 17143
+    },
+    {
+      "epoch": 4.551174830744723,
+      "grad_norm": 0.46956144029583297,
+      "learning_rate": 6.902853078181737e-07,
+      "loss": 0.4964,
+      "step": 17144
+    },
+    {
+      "epoch": 4.551440329218107,
+      "grad_norm": 0.48387744386522435,
+      "learning_rate": 6.900444552263891e-07,
+      "loss": 0.5134,
+      "step": 17145
+    },
+    {
+      "epoch": 4.551705827691491,
+      "grad_norm": 0.47149620046592317,
+      "learning_rate": 6.898036379331699e-07,
+      "loss": 0.5504,
+      "step": 17146
+    },
+    {
+      "epoch": 4.551971326164875,
+      "grad_norm": 0.4929477458935296,
+      "learning_rate": 6.895628559432113e-07,
+      "loss": 0.5555,
+      "step": 17147
+    },
+    {
+      "epoch": 4.552236824638259,
+      "grad_norm": 0.4899617239122328,
+      "learning_rate": 6.893221092612104e-07,
+      "loss": 0.5259,
+      "step": 17148
+    },
+    {
+      "epoch": 4.552502323111642,
+      "grad_norm": 0.46162926133999677,
+      "learning_rate": 6.890813978918617e-07,
+      "loss": 0.5035,
+      "step": 17149
+    },
+    {
+      "epoch": 4.552767821585026,
+      "grad_norm": 0.485307788365755,
+      "learning_rate": 6.888407218398588e-07,
+      "loss": 0.5616,
+      "step": 17150
+    },
+    {
+      "epoch": 4.553033320058409,
+      "grad_norm": 0.4773210174683122,
+      "learning_rate": 6.88600081109897e-07,
+      "loss": 0.532,
+      "step": 17151
+    },
+    {
+      "epoch": 4.553298818531793,
+      "grad_norm": 0.48455310720350026,
+      "learning_rate": 6.88359475706668e-07,
+      "loss": 0.5554,
+      "step": 17152
+    },
+    {
+      "epoch": 4.553564317005177,
+      "grad_norm": 0.48360112787341364,
+      "learning_rate": 6.881189056348658e-07,
+      "loss": 0.5647,
+      "step": 17153
+    },
+    {
+      "epoch": 4.5538298154785615,
+      "grad_norm": 0.4778923441340996,
+      "learning_rate": 6.87878370899181e-07,
+      "loss": 0.5382,
+      "step": 17154
+    },
+    {
+      "epoch": 4.554095313951945,
+      "grad_norm": 0.4792811632668715,
+      "learning_rate": 6.876378715043042e-07,
+      "loss": 0.5378,
+      "step": 17155
+    },
+    {
+      "epoch": 4.554360812425329,
+      "grad_norm": 0.4806240874642617,
+      "learning_rate": 6.87397407454927e-07,
+      "loss": 0.5803,
+      "step": 17156
+    },
+    {
+      "epoch": 4.554626310898712,
+      "grad_norm": 0.4776272868560844,
+      "learning_rate": 6.871569787557375e-07,
+      "loss": 0.5322,
+      "step": 17157
+    },
+    {
+      "epoch": 4.554891809372096,
+      "grad_norm": 0.49056719442931385,
+      "learning_rate": 6.869165854114262e-07,
+      "loss": 0.5451,
+      "step": 17158
+    },
+    {
+      "epoch": 4.55515730784548,
+      "grad_norm": 0.46707632864129345,
+      "learning_rate": 6.866762274266802e-07,
+      "loss": 0.5107,
+      "step": 17159
+    },
+    {
+      "epoch": 4.555422806318863,
+      "grad_norm": 0.4951714741395873,
+      "learning_rate": 6.864359048061891e-07,
+      "loss": 0.5536,
+      "step": 17160
+    },
+    {
+      "epoch": 4.555688304792247,
+      "grad_norm": 0.49489527732218297,
+      "learning_rate": 6.861956175546366e-07,
+      "loss": 0.5205,
+      "step": 17161
+    },
+    {
+      "epoch": 4.5559538032656315,
+      "grad_norm": 0.4920662916684385,
+      "learning_rate": 6.859553656767112e-07,
+      "loss": 0.5466,
+      "step": 17162
+    },
+    {
+      "epoch": 4.556219301739015,
+      "grad_norm": 0.4882950732383972,
+      "learning_rate": 6.85715149177097e-07,
+      "loss": 0.5187,
+      "step": 17163
+    },
+    {
+      "epoch": 4.556484800212399,
+      "grad_norm": 0.4772643547634087,
+      "learning_rate": 6.854749680604802e-07,
+      "loss": 0.5604,
+      "step": 17164
+    },
+    {
+      "epoch": 4.556750298685783,
+      "grad_norm": 0.47708757447293043,
+      "learning_rate": 6.852348223315435e-07,
+      "loss": 0.5099,
+      "step": 17165
+    },
+    {
+      "epoch": 4.557015797159166,
+      "grad_norm": 0.5020379710488587,
+      "learning_rate": 6.849947119949721e-07,
+      "loss": 0.5669,
+      "step": 17166
+    },
+    {
+      "epoch": 4.55728129563255,
+      "grad_norm": 0.489642298055821,
+      "learning_rate": 6.847546370554475e-07,
+      "loss": 0.5308,
+      "step": 17167
+    },
+    {
+      "epoch": 4.557546794105934,
+      "grad_norm": 0.475728110611278,
+      "learning_rate": 6.845145975176512e-07,
+      "loss": 0.5442,
+      "step": 17168
+    },
+    {
+      "epoch": 4.557812292579317,
+      "grad_norm": 0.47753854951884805,
+      "learning_rate": 6.842745933862663e-07,
+      "loss": 0.5527,
+      "step": 17169
+    },
+    {
+      "epoch": 4.5580777910527015,
+      "grad_norm": 0.49509551128834955,
+      "learning_rate": 6.840346246659716e-07,
+      "loss": 0.5646,
+      "step": 17170
+    },
+    {
+      "epoch": 4.558343289526086,
+      "grad_norm": 0.46540404747565056,
+      "learning_rate": 6.837946913614493e-07,
+      "loss": 0.4778,
+      "step": 17171
+    },
+    {
+      "epoch": 4.558608787999469,
+      "grad_norm": 0.49200152924063617,
+      "learning_rate": 6.835547934773759e-07,
+      "loss": 0.5237,
+      "step": 17172
+    },
+    {
+      "epoch": 4.558874286472853,
+      "grad_norm": 0.48048909654545585,
+      "learning_rate": 6.833149310184322e-07,
+      "loss": 0.5028,
+      "step": 17173
+    },
+    {
+      "epoch": 4.559139784946236,
+      "grad_norm": 0.48079049309936467,
+      "learning_rate": 6.83075103989295e-07,
+      "loss": 0.5282,
+      "step": 17174
+    },
+    {
+      "epoch": 4.55940528341962,
+      "grad_norm": 0.4980994638297322,
+      "learning_rate": 6.828353123946421e-07,
+      "loss": 0.5544,
+      "step": 17175
+    },
+    {
+      "epoch": 4.559670781893004,
+      "grad_norm": 0.48429804693509954,
+      "learning_rate": 6.825955562391493e-07,
+      "loss": 0.5268,
+      "step": 17176
+    },
+    {
+      "epoch": 4.559936280366388,
+      "grad_norm": 0.4834066983432881,
+      "learning_rate": 6.823558355274937e-07,
+      "loss": 0.5279,
+      "step": 17177
+    },
+    {
+      "epoch": 4.5602017788397715,
+      "grad_norm": 0.49291989688726173,
+      "learning_rate": 6.821161502643495e-07,
+      "loss": 0.5168,
+      "step": 17178
+    },
+    {
+      "epoch": 4.560467277313156,
+      "grad_norm": 0.4770239503189813,
+      "learning_rate": 6.818765004543906e-07,
+      "loss": 0.5342,
+      "step": 17179
+    },
+    {
+      "epoch": 4.560732775786539,
+      "grad_norm": 0.47066236248491955,
+      "learning_rate": 6.816368861022923e-07,
+      "loss": 0.5558,
+      "step": 17180
+    },
+    {
+      "epoch": 4.560998274259923,
+      "grad_norm": 0.469710280253976,
+      "learning_rate": 6.813973072127261e-07,
+      "loss": 0.5305,
+      "step": 17181
+    },
+    {
+      "epoch": 4.561263772733307,
+      "grad_norm": 0.4759236660418632,
+      "learning_rate": 6.81157763790366e-07,
+      "loss": 0.5329,
+      "step": 17182
+    },
+    {
+      "epoch": 4.561529271206691,
+      "grad_norm": 0.4821425532973299,
+      "learning_rate": 6.809182558398828e-07,
+      "loss": 0.5425,
+      "step": 17183
+    },
+    {
+      "epoch": 4.561794769680074,
+      "grad_norm": 0.476824950258736,
+      "learning_rate": 6.806787833659464e-07,
+      "loss": 0.5053,
+      "step": 17184
+    },
+    {
+      "epoch": 4.562060268153458,
+      "grad_norm": 0.47586108360500484,
+      "learning_rate": 6.804393463732293e-07,
+      "loss": 0.5445,
+      "step": 17185
+    },
+    {
+      "epoch": 4.5623257666268415,
+      "grad_norm": 0.4878188895390884,
+      "learning_rate": 6.801999448663991e-07,
+      "loss": 0.5805,
+      "step": 17186
+    },
+    {
+      "epoch": 4.562591265100226,
+      "grad_norm": 0.46892668516340896,
+      "learning_rate": 6.799605788501262e-07,
+      "loss": 0.5261,
+      "step": 17187
+    },
+    {
+      "epoch": 4.56285676357361,
+      "grad_norm": 0.4784872825847286,
+      "learning_rate": 6.797212483290777e-07,
+      "loss": 0.5451,
+      "step": 17188
+    },
+    {
+      "epoch": 4.563122262046993,
+      "grad_norm": 0.4940402384855071,
+      "learning_rate": 6.79481953307923e-07,
+      "loss": 0.5622,
+      "step": 17189
+    },
+    {
+      "epoch": 4.563387760520377,
+      "grad_norm": 0.4982410041945445,
+      "learning_rate": 6.792426937913263e-07,
+      "loss": 0.5183,
+      "step": 17190
+    },
+    {
+      "epoch": 4.563653258993761,
+      "grad_norm": 0.48748470306447933,
+      "learning_rate": 6.790034697839556e-07,
+      "loss": 0.5485,
+      "step": 17191
+    },
+    {
+      "epoch": 4.563918757467144,
+      "grad_norm": 0.48581950585643846,
+      "learning_rate": 6.787642812904752e-07,
+      "loss": 0.5714,
+      "step": 17192
+    },
+    {
+      "epoch": 4.564184255940528,
+      "grad_norm": 0.4974989678766032,
+      "learning_rate": 6.785251283155511e-07,
+      "loss": 0.5569,
+      "step": 17193
+    },
+    {
+      "epoch": 4.564449754413912,
+      "grad_norm": 0.4854672056003726,
+      "learning_rate": 6.78286010863847e-07,
+      "loss": 0.5555,
+      "step": 17194
+    },
+    {
+      "epoch": 4.564715252887296,
+      "grad_norm": 0.4798503326164381,
+      "learning_rate": 6.780469289400249e-07,
+      "loss": 0.5337,
+      "step": 17195
+    },
+    {
+      "epoch": 4.56498075136068,
+      "grad_norm": 0.48818149924633414,
+      "learning_rate": 6.778078825487497e-07,
+      "loss": 0.5463,
+      "step": 17196
+    },
+    {
+      "epoch": 4.565246249834064,
+      "grad_norm": 0.48280400619828245,
+      "learning_rate": 6.775688716946813e-07,
+      "loss": 0.5456,
+      "step": 17197
+    },
+    {
+      "epoch": 4.565511748307447,
+      "grad_norm": 0.476226632930575,
+      "learning_rate": 6.773298963824829e-07,
+      "loss": 0.5359,
+      "step": 17198
+    },
+    {
+      "epoch": 4.565777246780831,
+      "grad_norm": 0.4911757587209145,
+      "learning_rate": 6.770909566168133e-07,
+      "loss": 0.515,
+      "step": 17199
+    },
+    {
+      "epoch": 4.566042745254215,
+      "grad_norm": 0.48265399089289346,
+      "learning_rate": 6.768520524023348e-07,
+      "loss": 0.5246,
+      "step": 17200
+    },
+    {
+      "epoch": 4.566308243727598,
+      "grad_norm": 0.47824632329777716,
+      "learning_rate": 6.766131837437037e-07,
+      "loss": 0.5128,
+      "step": 17201
+    },
+    {
+      "epoch": 4.5665737422009824,
+      "grad_norm": 0.48136169839817466,
+      "learning_rate": 6.763743506455797e-07,
+      "loss": 0.5002,
+      "step": 17202
+    },
+    {
+      "epoch": 4.566839240674366,
+      "grad_norm": 0.47642750138075096,
+      "learning_rate": 6.761355531126221e-07,
+      "loss": 0.5354,
+      "step": 17203
+    },
+    {
+      "epoch": 4.56710473914775,
+      "grad_norm": 0.4792102793522238,
+      "learning_rate": 6.758967911494857e-07,
+      "loss": 0.5131,
+      "step": 17204
+    },
+    {
+      "epoch": 4.567370237621134,
+      "grad_norm": 0.4696769252991467,
+      "learning_rate": 6.756580647608288e-07,
+      "loss": 0.5304,
+      "step": 17205
+    },
+    {
+      "epoch": 4.567635736094518,
+      "grad_norm": 0.4695128355675209,
+      "learning_rate": 6.754193739513065e-07,
+      "loss": 0.5215,
+      "step": 17206
+    },
+    {
+      "epoch": 4.567901234567901,
+      "grad_norm": 0.488938932940753,
+      "learning_rate": 6.751807187255738e-07,
+      "loss": 0.5202,
+      "step": 17207
+    },
+    {
+      "epoch": 4.568166733041285,
+      "grad_norm": 0.4969564664600508,
+      "learning_rate": 6.74942099088284e-07,
+      "loss": 0.5336,
+      "step": 17208
+    },
+    {
+      "epoch": 4.568432231514668,
+      "grad_norm": 0.46748869405619303,
+      "learning_rate": 6.747035150440928e-07,
+      "loss": 0.5387,
+      "step": 17209
+    },
+    {
+      "epoch": 4.5686977299880525,
+      "grad_norm": 0.47279814089595606,
+      "learning_rate": 6.744649665976513e-07,
+      "loss": 0.5513,
+      "step": 17210
+    },
+    {
+      "epoch": 4.5689632284614365,
+      "grad_norm": 0.47653588690922505,
+      "learning_rate": 6.742264537536133e-07,
+      "loss": 0.5028,
+      "step": 17211
+    },
+    {
+      "epoch": 4.56922872693482,
+      "grad_norm": 0.47406683259452637,
+      "learning_rate": 6.739879765166299e-07,
+      "loss": 0.5234,
+      "step": 17212
+    },
+    {
+      "epoch": 4.569494225408204,
+      "grad_norm": 0.48962323254468404,
+      "learning_rate": 6.737495348913509e-07,
+      "loss": 0.5264,
+      "step": 17213
+    },
+    {
+      "epoch": 4.569759723881588,
+      "grad_norm": 0.4843586101828353,
+      "learning_rate": 6.735111288824281e-07,
+      "loss": 0.5477,
+      "step": 17214
+    },
+    {
+      "epoch": 4.570025222354971,
+      "grad_norm": 0.47734474171207525,
+      "learning_rate": 6.732727584945098e-07,
+      "loss": 0.5233,
+      "step": 17215
+    },
+    {
+      "epoch": 4.570290720828355,
+      "grad_norm": 0.48214337715167,
+      "learning_rate": 6.730344237322459e-07,
+      "loss": 0.5396,
+      "step": 17216
+    },
+    {
+      "epoch": 4.570556219301739,
+      "grad_norm": 0.5021900943159227,
+      "learning_rate": 6.72796124600284e-07,
+      "loss": 0.5167,
+      "step": 17217
+    },
+    {
+      "epoch": 4.5708217177751225,
+      "grad_norm": 0.48158096792201477,
+      "learning_rate": 6.725578611032713e-07,
+      "loss": 0.5285,
+      "step": 17218
+    },
+    {
+      "epoch": 4.571087216248507,
+      "grad_norm": 0.4724053983274434,
+      "learning_rate": 6.723196332458539e-07,
+      "loss": 0.5405,
+      "step": 17219
+    },
+    {
+      "epoch": 4.571352714721891,
+      "grad_norm": 0.48787416209390355,
+      "learning_rate": 6.720814410326795e-07,
+      "loss": 0.5425,
+      "step": 17220
+    },
+    {
+      "epoch": 4.571618213195274,
+      "grad_norm": 0.4641381056761872,
+      "learning_rate": 6.718432844683917e-07,
+      "loss": 0.5223,
+      "step": 17221
+    },
+    {
+      "epoch": 4.571883711668658,
+      "grad_norm": 0.48918285892535657,
+      "learning_rate": 6.716051635576359e-07,
+      "loss": 0.547,
+      "step": 17222
+    },
+    {
+      "epoch": 4.572149210142042,
+      "grad_norm": 0.4727924425579615,
+      "learning_rate": 6.713670783050577e-07,
+      "loss": 0.5502,
+      "step": 17223
+    },
+    {
+      "epoch": 4.572414708615425,
+      "grad_norm": 0.47997722969102286,
+      "learning_rate": 6.711290287152972e-07,
+      "loss": 0.534,
+      "step": 17224
+    },
+    {
+      "epoch": 4.572680207088809,
+      "grad_norm": 0.4791669935808119,
+      "learning_rate": 6.708910147929993e-07,
+      "loss": 0.5237,
+      "step": 17225
+    },
+    {
+      "epoch": 4.572945705562193,
+      "grad_norm": 0.48252595732700376,
+      "learning_rate": 6.70653036542804e-07,
+      "loss": 0.5501,
+      "step": 17226
+    },
+    {
+      "epoch": 4.573211204035577,
+      "grad_norm": 0.4980275857003703,
+      "learning_rate": 6.704150939693546e-07,
+      "loss": 0.5489,
+      "step": 17227
+    },
+    {
+      "epoch": 4.573476702508961,
+      "grad_norm": 0.491811048352953,
+      "learning_rate": 6.701771870772894e-07,
+      "loss": 0.5256,
+      "step": 17228
+    },
+    {
+      "epoch": 4.573742200982345,
+      "grad_norm": 0.47983251729763715,
+      "learning_rate": 6.699393158712503e-07,
+      "loss": 0.581,
+      "step": 17229
+    },
+    {
+      "epoch": 4.574007699455728,
+      "grad_norm": 0.4890747790824589,
+      "learning_rate": 6.697014803558752e-07,
+      "loss": 0.5765,
+      "step": 17230
+    },
+    {
+      "epoch": 4.574273197929112,
+      "grad_norm": 0.47633728958575655,
+      "learning_rate": 6.694636805358016e-07,
+      "loss": 0.5458,
+      "step": 17231
+    },
+    {
+      "epoch": 4.574538696402495,
+      "grad_norm": 0.47726055600972683,
+      "learning_rate": 6.692259164156692e-07,
+      "loss": 0.5028,
+      "step": 17232
+    },
+    {
+      "epoch": 4.574804194875879,
+      "grad_norm": 0.4740314086061241,
+      "learning_rate": 6.689881880001131e-07,
+      "loss": 0.5448,
+      "step": 17233
+    },
+    {
+      "epoch": 4.575069693349263,
+      "grad_norm": 0.48065267457395444,
+      "learning_rate": 6.68750495293771e-07,
+      "loss": 0.5214,
+      "step": 17234
+    },
+    {
+      "epoch": 4.5753351918226475,
+      "grad_norm": 0.47090751322597807,
+      "learning_rate": 6.685128383012782e-07,
+      "loss": 0.5404,
+      "step": 17235
+    },
+    {
+      "epoch": 4.575600690296031,
+      "grad_norm": 0.47849260927998805,
+      "learning_rate": 6.682752170272691e-07,
+      "loss": 0.5468,
+      "step": 17236
+    },
+    {
+      "epoch": 4.575866188769415,
+      "grad_norm": 0.5016774457200815,
+      "learning_rate": 6.680376314763773e-07,
+      "loss": 0.5396,
+      "step": 17237
+    },
+    {
+      "epoch": 4.576131687242798,
+      "grad_norm": 0.47640312142178937,
+      "learning_rate": 6.678000816532381e-07,
+      "loss": 0.5244,
+      "step": 17238
+    },
+    {
+      "epoch": 4.576397185716182,
+      "grad_norm": 0.469484645253419,
+      "learning_rate": 6.675625675624825e-07,
+      "loss": 0.5296,
+      "step": 17239
+    },
+    {
+      "epoch": 4.576662684189566,
+      "grad_norm": 0.4802714649381744,
+      "learning_rate": 6.673250892087441e-07,
+      "loss": 0.5119,
+      "step": 17240
+    },
+    {
+      "epoch": 4.576928182662949,
+      "grad_norm": 0.484651396021477,
+      "learning_rate": 6.670876465966536e-07,
+      "loss": 0.5356,
+      "step": 17241
+    },
+    {
+      "epoch": 4.577193681136333,
+      "grad_norm": 0.474452397273157,
+      "learning_rate": 6.66850239730841e-07,
+      "loss": 0.5159,
+      "step": 17242
+    },
+    {
+      "epoch": 4.5774591796097175,
+      "grad_norm": 0.48715717450839036,
+      "learning_rate": 6.666128686159376e-07,
+      "loss": 0.5124,
+      "step": 17243
+    },
+    {
+      "epoch": 4.577724678083101,
+      "grad_norm": 0.4847015381863419,
+      "learning_rate": 6.663755332565716e-07,
+      "loss": 0.548,
+      "step": 17244
+    },
+    {
+      "epoch": 4.577990176556485,
+      "grad_norm": 0.4620611511065152,
+      "learning_rate": 6.66138233657373e-07,
+      "loss": 0.4913,
+      "step": 17245
+    },
+    {
+      "epoch": 4.578255675029869,
+      "grad_norm": 0.4708939455040296,
+      "learning_rate": 6.65900969822969e-07,
+      "loss": 0.5596,
+      "step": 17246
+    },
+    {
+      "epoch": 4.578521173503252,
+      "grad_norm": 0.4918910957013495,
+      "learning_rate": 6.656637417579859e-07,
+      "loss": 0.5236,
+      "step": 17247
+    },
+    {
+      "epoch": 4.578786671976636,
+      "grad_norm": 0.4931152916154499,
+      "learning_rate": 6.654265494670517e-07,
+      "loss": 0.5262,
+      "step": 17248
+    },
+    {
+      "epoch": 4.57905217045002,
+      "grad_norm": 0.4707283422372326,
+      "learning_rate": 6.65189392954791e-07,
+      "loss": 0.5443,
+      "step": 17249
+    },
+    {
+      "epoch": 4.579317668923403,
+      "grad_norm": 0.48242748625921567,
+      "learning_rate": 6.649522722258301e-07,
+      "loss": 0.5632,
+      "step": 17250
+    },
+    {
+      "epoch": 4.5795831673967875,
+      "grad_norm": 0.4786895828320936,
+      "learning_rate": 6.647151872847921e-07,
+      "loss": 0.4868,
+      "step": 17251
+    },
+    {
+      "epoch": 4.579848665870172,
+      "grad_norm": 0.4693656573111654,
+      "learning_rate": 6.64478138136303e-07,
+      "loss": 0.5279,
+      "step": 17252
+    },
+    {
+      "epoch": 4.580114164343555,
+      "grad_norm": 0.48385856252776116,
+      "learning_rate": 6.64241124784983e-07,
+      "loss": 0.5094,
+      "step": 17253
+    },
+    {
+      "epoch": 4.580379662816939,
+      "grad_norm": 0.506119600576592,
+      "learning_rate": 6.640041472354567e-07,
+      "loss": 0.5558,
+      "step": 17254
+    },
+    {
+      "epoch": 4.580645161290323,
+      "grad_norm": 0.4915980632534202,
+      "learning_rate": 6.637672054923436e-07,
+      "loss": 0.5483,
+      "step": 17255
+    },
+    {
+      "epoch": 4.580910659763706,
+      "grad_norm": 0.46420933717999996,
+      "learning_rate": 6.635302995602669e-07,
+      "loss": 0.5418,
+      "step": 17256
+    },
+    {
+      "epoch": 4.58117615823709,
+      "grad_norm": 0.48700086116791647,
+      "learning_rate": 6.632934294438451e-07,
+      "loss": 0.5278,
+      "step": 17257
+    },
+    {
+      "epoch": 4.581441656710474,
+      "grad_norm": 0.49054521279438595,
+      "learning_rate": 6.630565951476992e-07,
+      "loss": 0.5319,
+      "step": 17258
+    },
+    {
+      "epoch": 4.5817071551838575,
+      "grad_norm": 0.47721463339418885,
+      "learning_rate": 6.628197966764471e-07,
+      "loss": 0.56,
+      "step": 17259
+    },
+    {
+      "epoch": 4.581972653657242,
+      "grad_norm": 0.47392959116876726,
+      "learning_rate": 6.625830340347067e-07,
+      "loss": 0.5258,
+      "step": 17260
+    },
+    {
+      "epoch": 4.582238152130625,
+      "grad_norm": 0.4862036638340638,
+      "learning_rate": 6.623463072270964e-07,
+      "loss": 0.559,
+      "step": 17261
+    },
+    {
+      "epoch": 4.582503650604009,
+      "grad_norm": 0.4720959035807394,
+      "learning_rate": 6.621096162582319e-07,
+      "loss": 0.5418,
+      "step": 17262
+    },
+    {
+      "epoch": 4.582769149077393,
+      "grad_norm": 0.4718342250451868,
+      "learning_rate": 6.618729611327313e-07,
+      "loss": 0.5203,
+      "step": 17263
+    },
+    {
+      "epoch": 4.583034647550777,
+      "grad_norm": 0.48642522012845996,
+      "learning_rate": 6.616363418552069e-07,
+      "loss": 0.5039,
+      "step": 17264
+    },
+    {
+      "epoch": 4.58330014602416,
+      "grad_norm": 0.4959256711853409,
+      "learning_rate": 6.613997584302756e-07,
+      "loss": 0.5538,
+      "step": 17265
+    },
+    {
+      "epoch": 4.583565644497544,
+      "grad_norm": 0.46902689892776284,
+      "learning_rate": 6.611632108625498e-07,
+      "loss": 0.5227,
+      "step": 17266
+    },
+    {
+      "epoch": 4.5838311429709275,
+      "grad_norm": 0.4773103957908522,
+      "learning_rate": 6.60926699156644e-07,
+      "loss": 0.5257,
+      "step": 17267
+    },
+    {
+      "epoch": 4.584096641444312,
+      "grad_norm": 0.47231769178954197,
+      "learning_rate": 6.60690223317171e-07,
+      "loss": 0.5147,
+      "step": 17268
+    },
+    {
+      "epoch": 4.584362139917696,
+      "grad_norm": 0.49492769996366154,
+      "learning_rate": 6.60453783348742e-07,
+      "loss": 0.5471,
+      "step": 17269
+    },
+    {
+      "epoch": 4.584627638391079,
+      "grad_norm": 0.4868640983641583,
+      "learning_rate": 6.602173792559683e-07,
+      "loss": 0.5608,
+      "step": 17270
+    },
+    {
+      "epoch": 4.584893136864463,
+      "grad_norm": 0.48078771009235843,
+      "learning_rate": 6.599810110434593e-07,
+      "loss": 0.5589,
+      "step": 17271
+    },
+    {
+      "epoch": 4.585158635337847,
+      "grad_norm": 0.48171629121128,
+      "learning_rate": 6.597446787158268e-07,
+      "loss": 0.5667,
+      "step": 17272
+    },
+    {
+      "epoch": 4.58542413381123,
+      "grad_norm": 0.4795312123798523,
+      "learning_rate": 6.59508382277678e-07,
+      "loss": 0.5357,
+      "step": 17273
+    },
+    {
+      "epoch": 4.585689632284614,
+      "grad_norm": 0.4826982129620007,
+      "learning_rate": 6.592721217336226e-07,
+      "loss": 0.5011,
+      "step": 17274
+    },
+    {
+      "epoch": 4.585955130757998,
+      "grad_norm": 0.49521054719034063,
+      "learning_rate": 6.590358970882677e-07,
+      "loss": 0.5479,
+      "step": 17275
+    },
+    {
+      "epoch": 4.586220629231382,
+      "grad_norm": 0.47797629804790837,
+      "learning_rate": 6.587997083462197e-07,
+      "loss": 0.5431,
+      "step": 17276
+    },
+    {
+      "epoch": 4.586486127704766,
+      "grad_norm": 0.4961224433377459,
+      "learning_rate": 6.585635555120859e-07,
+      "loss": 0.5472,
+      "step": 17277
+    },
+    {
+      "epoch": 4.58675162617815,
+      "grad_norm": 0.4961438713426632,
+      "learning_rate": 6.583274385904709e-07,
+      "loss": 0.5558,
+      "step": 17278
+    },
+    {
+      "epoch": 4.587017124651533,
+      "grad_norm": 0.4788743574320205,
+      "learning_rate": 6.580913575859807e-07,
+      "loss": 0.5429,
+      "step": 17279
+    },
+    {
+      "epoch": 4.587282623124917,
+      "grad_norm": 0.4779864084731545,
+      "learning_rate": 6.57855312503218e-07,
+      "loss": 0.5132,
+      "step": 17280
+    },
+    {
+      "epoch": 4.587548121598301,
+      "grad_norm": 0.46735563780690403,
+      "learning_rate": 6.576193033467887e-07,
+      "loss": 0.4697,
+      "step": 17281
+    },
+    {
+      "epoch": 4.587813620071684,
+      "grad_norm": 0.47583677429816956,
+      "learning_rate": 6.573833301212923e-07,
+      "loss": 0.5369,
+      "step": 17282
+    },
+    {
+      "epoch": 4.5880791185450684,
+      "grad_norm": 0.4725832828491452,
+      "learning_rate": 6.571473928313332e-07,
+      "loss": 0.5733,
+      "step": 17283
+    },
+    {
+      "epoch": 4.5883446170184525,
+      "grad_norm": 0.4964485779259073,
+      "learning_rate": 6.569114914815111e-07,
+      "loss": 0.5249,
+      "step": 17284
+    },
+    {
+      "epoch": 4.588610115491836,
+      "grad_norm": 0.4642813379363941,
+      "learning_rate": 6.566756260764287e-07,
+      "loss": 0.5364,
+      "step": 17285
+    },
+    {
+      "epoch": 4.58887561396522,
+      "grad_norm": 0.48672996767569293,
+      "learning_rate": 6.564397966206836e-07,
+      "loss": 0.5557,
+      "step": 17286
+    },
+    {
+      "epoch": 4.589141112438604,
+      "grad_norm": 0.48797537442724864,
+      "learning_rate": 6.562040031188773e-07,
+      "loss": 0.5737,
+      "step": 17287
+    },
+    {
+      "epoch": 4.589406610911987,
+      "grad_norm": 0.47369139712473873,
+      "learning_rate": 6.559682455756075e-07,
+      "loss": 0.5257,
+      "step": 17288
+    },
+    {
+      "epoch": 4.589672109385371,
+      "grad_norm": 0.4800866137753457,
+      "learning_rate": 6.55732523995471e-07,
+      "loss": 0.5493,
+      "step": 17289
+    },
+    {
+      "epoch": 4.589937607858754,
+      "grad_norm": 0.4790438024171878,
+      "learning_rate": 6.554968383830665e-07,
+      "loss": 0.5484,
+      "step": 17290
+    },
+    {
+      "epoch": 4.5902031063321385,
+      "grad_norm": 0.4926238312728841,
+      "learning_rate": 6.552611887429891e-07,
+      "loss": 0.5345,
+      "step": 17291
+    },
+    {
+      "epoch": 4.5904686048055225,
+      "grad_norm": 0.48702839882089777,
+      "learning_rate": 6.550255750798359e-07,
+      "loss": 0.5344,
+      "step": 17292
+    },
+    {
+      "epoch": 4.590734103278907,
+      "grad_norm": 0.47881925044588713,
+      "learning_rate": 6.547899973982014e-07,
+      "loss": 0.4999,
+      "step": 17293
+    },
+    {
+      "epoch": 4.59099960175229,
+      "grad_norm": 0.4831463568891131,
+      "learning_rate": 6.54554455702679e-07,
+      "loss": 0.5359,
+      "step": 17294
+    },
+    {
+      "epoch": 4.591265100225674,
+      "grad_norm": 0.49014652527471375,
+      "learning_rate": 6.543189499978638e-07,
+      "loss": 0.5213,
+      "step": 17295
+    },
+    {
+      "epoch": 4.591530598699057,
+      "grad_norm": 0.48506302121277745,
+      "learning_rate": 6.540834802883475e-07,
+      "loss": 0.555,
+      "step": 17296
+    },
+    {
+      "epoch": 4.591796097172441,
+      "grad_norm": 0.47845680846038047,
+      "learning_rate": 6.538480465787236e-07,
+      "loss": 0.5512,
+      "step": 17297
+    },
+    {
+      "epoch": 4.592061595645825,
+      "grad_norm": 0.4814313943474611,
+      "learning_rate": 6.536126488735828e-07,
+      "loss": 0.5227,
+      "step": 17298
+    },
+    {
+      "epoch": 4.5923270941192085,
+      "grad_norm": 0.470977617637183,
+      "learning_rate": 6.533772871775162e-07,
+      "loss": 0.5061,
+      "step": 17299
+    },
+    {
+      "epoch": 4.592592592592593,
+      "grad_norm": 0.4894117386603172,
+      "learning_rate": 6.531419614951132e-07,
+      "loss": 0.535,
+      "step": 17300
+    },
+    {
+      "epoch": 4.592858091065977,
+      "grad_norm": 0.4887360920426576,
+      "learning_rate": 6.529066718309643e-07,
+      "loss": 0.5095,
+      "step": 17301
+    },
+    {
+      "epoch": 4.59312358953936,
+      "grad_norm": 0.4869726230841305,
+      "learning_rate": 6.526714181896573e-07,
+      "loss": 0.5433,
+      "step": 17302
+    },
+    {
+      "epoch": 4.593389088012744,
+      "grad_norm": 0.47769366664393115,
+      "learning_rate": 6.524362005757811e-07,
+      "loss": 0.527,
+      "step": 17303
+    },
+    {
+      "epoch": 4.593654586486128,
+      "grad_norm": 0.4870302974104966,
+      "learning_rate": 6.522010189939229e-07,
+      "loss": 0.5449,
+      "step": 17304
+    },
+    {
+      "epoch": 4.593920084959511,
+      "grad_norm": 0.5024733537923884,
+      "learning_rate": 6.519658734486681e-07,
+      "loss": 0.5718,
+      "step": 17305
+    },
+    {
+      "epoch": 4.594185583432895,
+      "grad_norm": 0.4811477113574121,
+      "learning_rate": 6.517307639446044e-07,
+      "loss": 0.561,
+      "step": 17306
+    },
+    {
+      "epoch": 4.594451081906279,
+      "grad_norm": 0.468422224431071,
+      "learning_rate": 6.514956904863154e-07,
+      "loss": 0.5239,
+      "step": 17307
+    },
+    {
+      "epoch": 4.594716580379663,
+      "grad_norm": 0.4722604426185916,
+      "learning_rate": 6.512606530783872e-07,
+      "loss": 0.5588,
+      "step": 17308
+    },
+    {
+      "epoch": 4.594982078853047,
+      "grad_norm": 0.494579951202955,
+      "learning_rate": 6.510256517254018e-07,
+      "loss": 0.549,
+      "step": 17309
+    },
+    {
+      "epoch": 4.595247577326431,
+      "grad_norm": 0.47287098344723394,
+      "learning_rate": 6.507906864319446e-07,
+      "loss": 0.5334,
+      "step": 17310
+    },
+    {
+      "epoch": 4.595513075799814,
+      "grad_norm": 0.466100249257651,
+      "learning_rate": 6.505557572025955e-07,
+      "loss": 0.4973,
+      "step": 17311
+    },
+    {
+      "epoch": 4.595778574273198,
+      "grad_norm": 0.47361141344074614,
+      "learning_rate": 6.503208640419381e-07,
+      "loss": 0.5384,
+      "step": 17312
+    },
+    {
+      "epoch": 4.596044072746581,
+      "grad_norm": 0.4817522343316887,
+      "learning_rate": 6.500860069545517e-07,
+      "loss": 0.5496,
+      "step": 17313
+    },
+    {
+      "epoch": 4.596309571219965,
+      "grad_norm": 0.4872812452313049,
+      "learning_rate": 6.498511859450177e-07,
+      "loss": 0.538,
+      "step": 17314
+    },
+    {
+      "epoch": 4.596575069693349,
+      "grad_norm": 0.48697952867984573,
+      "learning_rate": 6.496164010179167e-07,
+      "loss": 0.5386,
+      "step": 17315
+    },
+    {
+      "epoch": 4.5968405681667335,
+      "grad_norm": 0.475896187012161,
+      "learning_rate": 6.493816521778254e-07,
+      "loss": 0.5393,
+      "step": 17316
+    },
+    {
+      "epoch": 4.597106066640117,
+      "grad_norm": 0.4734418847840232,
+      "learning_rate": 6.491469394293235e-07,
+      "loss": 0.512,
+      "step": 17317
+    },
+    {
+      "epoch": 4.597371565113501,
+      "grad_norm": 0.487635026913717,
+      "learning_rate": 6.48912262776987e-07,
+      "loss": 0.5337,
+      "step": 17318
+    },
+    {
+      "epoch": 4.597637063586884,
+      "grad_norm": 0.4828028027958323,
+      "learning_rate": 6.486776222253946e-07,
+      "loss": 0.5107,
+      "step": 17319
+    },
+    {
+      "epoch": 4.597902562060268,
+      "grad_norm": 0.4698875026378944,
+      "learning_rate": 6.484430177791207e-07,
+      "loss": 0.525,
+      "step": 17320
+    },
+    {
+      "epoch": 4.598168060533652,
+      "grad_norm": 0.4808326233541619,
+      "learning_rate": 6.482084494427418e-07,
+      "loss": 0.5515,
+      "step": 17321
+    },
+    {
+      "epoch": 4.598433559007036,
+      "grad_norm": 0.4858764275157084,
+      "learning_rate": 6.479739172208324e-07,
+      "loss": 0.5185,
+      "step": 17322
+    },
+    {
+      "epoch": 4.598699057480419,
+      "grad_norm": 0.5107945480221684,
+      "learning_rate": 6.477394211179652e-07,
+      "loss": 0.5338,
+      "step": 17323
+    },
+    {
+      "epoch": 4.5989645559538035,
+      "grad_norm": 0.49059586222909546,
+      "learning_rate": 6.475049611387152e-07,
+      "loss": 0.5416,
+      "step": 17324
+    },
+    {
+      "epoch": 4.599230054427187,
+      "grad_norm": 0.47927003748091607,
+      "learning_rate": 6.472705372876533e-07,
+      "loss": 0.5237,
+      "step": 17325
+    },
+    {
+      "epoch": 4.599495552900571,
+      "grad_norm": 0.4674364960090774,
+      "learning_rate": 6.470361495693534e-07,
+      "loss": 0.4785,
+      "step": 17326
+    },
+    {
+      "epoch": 4.599761051373955,
+      "grad_norm": 0.4757338269593397,
+      "learning_rate": 6.46801797988385e-07,
+      "loss": 0.5208,
+      "step": 17327
+    },
+    {
+      "epoch": 4.600026549847338,
+      "grad_norm": 0.4851689833037894,
+      "learning_rate": 6.46567482549319e-07,
+      "loss": 0.5725,
+      "step": 17328
+    },
+    {
+      "epoch": 4.600292048320722,
+      "grad_norm": 0.49755397757279407,
+      "learning_rate": 6.463332032567244e-07,
+      "loss": 0.5332,
+      "step": 17329
+    },
+    {
+      "epoch": 4.600557546794106,
+      "grad_norm": 0.4804176864232585,
+      "learning_rate": 6.460989601151716e-07,
+      "loss": 0.5792,
+      "step": 17330
+    },
+    {
+      "epoch": 4.600823045267489,
+      "grad_norm": 0.4799247096087188,
+      "learning_rate": 6.458647531292276e-07,
+      "loss": 0.5782,
+      "step": 17331
+    },
+    {
+      "epoch": 4.6010885437408735,
+      "grad_norm": 0.4678981572949605,
+      "learning_rate": 6.456305823034612e-07,
+      "loss": 0.5216,
+      "step": 17332
+    },
+    {
+      "epoch": 4.601354042214258,
+      "grad_norm": 0.4959774940702711,
+      "learning_rate": 6.453964476424387e-07,
+      "loss": 0.5194,
+      "step": 17333
+    },
+    {
+      "epoch": 4.601619540687641,
+      "grad_norm": 0.47432003621312463,
+      "learning_rate": 6.451623491507256e-07,
+      "loss": 0.5559,
+      "step": 17334
+    },
+    {
+      "epoch": 4.601885039161025,
+      "grad_norm": 0.48193639714086517,
+      "learning_rate": 6.449282868328891e-07,
+      "loss": 0.5384,
+      "step": 17335
+    },
+    {
+      "epoch": 4.602150537634409,
+      "grad_norm": 0.4951029700848952,
+      "learning_rate": 6.446942606934919e-07,
+      "loss": 0.5146,
+      "step": 17336
+    },
+    {
+      "epoch": 4.602416036107792,
+      "grad_norm": 0.49995458114157404,
+      "learning_rate": 6.444602707371e-07,
+      "loss": 0.5782,
+      "step": 17337
+    },
+    {
+      "epoch": 4.602681534581176,
+      "grad_norm": 0.46960299613210243,
+      "learning_rate": 6.44226316968276e-07,
+      "loss": 0.521,
+      "step": 17338
+    },
+    {
+      "epoch": 4.60294703305456,
+      "grad_norm": 0.4840516926504308,
+      "learning_rate": 6.43992399391582e-07,
+      "loss": 0.5278,
+      "step": 17339
+    },
+    {
+      "epoch": 4.6032125315279435,
+      "grad_norm": 0.4852931836358397,
+      "learning_rate": 6.43758518011581e-07,
+      "loss": 0.5178,
+      "step": 17340
+    },
+    {
+      "epoch": 4.603478030001328,
+      "grad_norm": 0.4688824781239619,
+      "learning_rate": 6.435246728328332e-07,
+      "loss": 0.5425,
+      "step": 17341
+    },
+    {
+      "epoch": 4.603743528474711,
+      "grad_norm": 0.48388384248791577,
+      "learning_rate": 6.432908638599003e-07,
+      "loss": 0.5414,
+      "step": 17342
+    },
+    {
+      "epoch": 4.604009026948095,
+      "grad_norm": 0.4732440774880686,
+      "learning_rate": 6.43057091097341e-07,
+      "loss": 0.5036,
+      "step": 17343
+    },
+    {
+      "epoch": 4.604274525421479,
+      "grad_norm": 0.4784110938977977,
+      "learning_rate": 6.428233545497165e-07,
+      "loss": 0.53,
+      "step": 17344
+    },
+    {
+      "epoch": 4.604540023894863,
+      "grad_norm": 0.4766481092781776,
+      "learning_rate": 6.425896542215821e-07,
+      "loss": 0.5566,
+      "step": 17345
+    },
+    {
+      "epoch": 4.604805522368246,
+      "grad_norm": 0.4900214150419999,
+      "learning_rate": 6.423559901174981e-07,
+      "loss": 0.5586,
+      "step": 17346
+    },
+    {
+      "epoch": 4.60507102084163,
+      "grad_norm": 0.48743031055794483,
+      "learning_rate": 6.421223622420194e-07,
+      "loss": 0.5359,
+      "step": 17347
+    },
+    {
+      "epoch": 4.6053365193150135,
+      "grad_norm": 0.48164934780152713,
+      "learning_rate": 6.418887705997046e-07,
+      "loss": 0.5355,
+      "step": 17348
+    },
+    {
+      "epoch": 4.605602017788398,
+      "grad_norm": 0.4830949060834166,
+      "learning_rate": 6.416552151951075e-07,
+      "loss": 0.5683,
+      "step": 17349
+    },
+    {
+      "epoch": 4.605867516261782,
+      "grad_norm": 0.4836943040325858,
+      "learning_rate": 6.414216960327844e-07,
+      "loss": 0.5631,
+      "step": 17350
+    },
+    {
+      "epoch": 4.606133014735165,
+      "grad_norm": 0.47832669251993887,
+      "learning_rate": 6.411882131172887e-07,
+      "loss": 0.5528,
+      "step": 17351
+    },
+    {
+      "epoch": 4.606398513208549,
+      "grad_norm": 0.47680765733040154,
+      "learning_rate": 6.409547664531734e-07,
+      "loss": 0.5348,
+      "step": 17352
+    },
+    {
+      "epoch": 4.606664011681933,
+      "grad_norm": 0.4804330388960981,
+      "learning_rate": 6.407213560449926e-07,
+      "loss": 0.5887,
+      "step": 17353
+    },
+    {
+      "epoch": 4.606929510155316,
+      "grad_norm": 0.4839095738969038,
+      "learning_rate": 6.40487981897297e-07,
+      "loss": 0.5231,
+      "step": 17354
+    },
+    {
+      "epoch": 4.6071950086287,
+      "grad_norm": 0.49127087479993997,
+      "learning_rate": 6.40254644014639e-07,
+      "loss": 0.5592,
+      "step": 17355
+    },
+    {
+      "epoch": 4.607460507102084,
+      "grad_norm": 0.4876203261880093,
+      "learning_rate": 6.400213424015691e-07,
+      "loss": 0.5591,
+      "step": 17356
+    },
+    {
+      "epoch": 4.607726005575468,
+      "grad_norm": 0.471595103459599,
+      "learning_rate": 6.397880770626372e-07,
+      "loss": 0.5157,
+      "step": 17357
+    },
+    {
+      "epoch": 4.607991504048852,
+      "grad_norm": 0.4847718370299789,
+      "learning_rate": 6.395548480023917e-07,
+      "loss": 0.5516,
+      "step": 17358
+    },
+    {
+      "epoch": 4.608257002522236,
+      "grad_norm": 0.4820107632343782,
+      "learning_rate": 6.393216552253814e-07,
+      "loss": 0.58,
+      "step": 17359
+    },
+    {
+      "epoch": 4.608522500995619,
+      "grad_norm": 0.47716768056181674,
+      "learning_rate": 6.390884987361556e-07,
+      "loss": 0.528,
+      "step": 17360
+    },
+    {
+      "epoch": 4.608787999469003,
+      "grad_norm": 0.486268430343264,
+      "learning_rate": 6.388553785392604e-07,
+      "loss": 0.5023,
+      "step": 17361
+    },
+    {
+      "epoch": 4.609053497942387,
+      "grad_norm": 0.4804585050846759,
+      "learning_rate": 6.38622294639242e-07,
+      "loss": 0.5558,
+      "step": 17362
+    },
+    {
+      "epoch": 4.60931899641577,
+      "grad_norm": 0.47449720872194084,
+      "learning_rate": 6.383892470406456e-07,
+      "loss": 0.5282,
+      "step": 17363
+    },
+    {
+      "epoch": 4.6095844948891544,
+      "grad_norm": 0.46913942144166004,
+      "learning_rate": 6.381562357480176e-07,
+      "loss": 0.523,
+      "step": 17364
+    },
+    {
+      "epoch": 4.6098499933625385,
+      "grad_norm": 0.48453931208963535,
+      "learning_rate": 6.379232607659008e-07,
+      "loss": 0.5327,
+      "step": 17365
+    },
+    {
+      "epoch": 4.610115491835922,
+      "grad_norm": 0.4923217881141727,
+      "learning_rate": 6.376903220988406e-07,
+      "loss": 0.5433,
+      "step": 17366
+    },
+    {
+      "epoch": 4.610380990309306,
+      "grad_norm": 0.48994490808075924,
+      "learning_rate": 6.374574197513786e-07,
+      "loss": 0.5417,
+      "step": 17367
+    },
+    {
+      "epoch": 4.61064648878269,
+      "grad_norm": 0.48034568735882427,
+      "learning_rate": 6.372245537280566e-07,
+      "loss": 0.5509,
+      "step": 17368
+    },
+    {
+      "epoch": 4.610911987256073,
+      "grad_norm": 0.49149677299488925,
+      "learning_rate": 6.369917240334173e-07,
+      "loss": 0.5294,
+      "step": 17369
+    },
+    {
+      "epoch": 4.611177485729457,
+      "grad_norm": 0.4788799625973278,
+      "learning_rate": 6.367589306720001e-07,
+      "loss": 0.5272,
+      "step": 17370
+    },
+    {
+      "epoch": 4.61144298420284,
+      "grad_norm": 0.487062945198114,
+      "learning_rate": 6.365261736483464e-07,
+      "loss": 0.5384,
+      "step": 17371
+    },
+    {
+      "epoch": 4.6117084826762245,
+      "grad_norm": 0.49938578574907366,
+      "learning_rate": 6.362934529669942e-07,
+      "loss": 0.5655,
+      "step": 17372
+    },
+    {
+      "epoch": 4.6119739811496085,
+      "grad_norm": 0.47647180187905896,
+      "learning_rate": 6.360607686324843e-07,
+      "loss": 0.5174,
+      "step": 17373
+    },
+    {
+      "epoch": 4.612239479622993,
+      "grad_norm": 0.48676080379822495,
+      "learning_rate": 6.358281206493516e-07,
+      "loss": 0.5675,
+      "step": 17374
+    },
+    {
+      "epoch": 4.612504978096376,
+      "grad_norm": 0.4935539926240541,
+      "learning_rate": 6.355955090221358e-07,
+      "loss": 0.5271,
+      "step": 17375
+    },
+    {
+      "epoch": 4.61277047656976,
+      "grad_norm": 0.47258890903380474,
+      "learning_rate": 6.353629337553716e-07,
+      "loss": 0.5137,
+      "step": 17376
+    },
+    {
+      "epoch": 4.613035975043143,
+      "grad_norm": 0.4885127829573645,
+      "learning_rate": 6.351303948535961e-07,
+      "loss": 0.5436,
+      "step": 17377
+    },
+    {
+      "epoch": 4.613301473516527,
+      "grad_norm": 0.4943823281167259,
+      "learning_rate": 6.348978923213434e-07,
+      "loss": 0.575,
+      "step": 17378
+    },
+    {
+      "epoch": 4.613566971989911,
+      "grad_norm": 0.4662464502894001,
+      "learning_rate": 6.346654261631488e-07,
+      "loss": 0.5433,
+      "step": 17379
+    },
+    {
+      "epoch": 4.6138324704632945,
+      "grad_norm": 0.4794442306213752,
+      "learning_rate": 6.344329963835458e-07,
+      "loss": 0.5635,
+      "step": 17380
+    },
+    {
+      "epoch": 4.614097968936679,
+      "grad_norm": 0.4715141785878349,
+      "learning_rate": 6.342006029870662e-07,
+      "loss": 0.505,
+      "step": 17381
+    },
+    {
+      "epoch": 4.614363467410063,
+      "grad_norm": 0.48262661391284667,
+      "learning_rate": 6.33968245978244e-07,
+      "loss": 0.5249,
+      "step": 17382
+    },
+    {
+      "epoch": 4.614628965883446,
+      "grad_norm": 0.4734583988441969,
+      "learning_rate": 6.33735925361609e-07,
+      "loss": 0.524,
+      "step": 17383
+    },
+    {
+      "epoch": 4.61489446435683,
+      "grad_norm": 0.4800495271630748,
+      "learning_rate": 6.335036411416937e-07,
+      "loss": 0.5558,
+      "step": 17384
+    },
+    {
+      "epoch": 4.615159962830214,
+      "grad_norm": 0.47829998197458673,
+      "learning_rate": 6.332713933230273e-07,
+      "loss": 0.536,
+      "step": 17385
+    },
+    {
+      "epoch": 4.615425461303597,
+      "grad_norm": 0.478371725356284,
+      "learning_rate": 6.330391819101383e-07,
+      "loss": 0.563,
+      "step": 17386
+    },
+    {
+      "epoch": 4.615690959776981,
+      "grad_norm": 0.4890444885993846,
+      "learning_rate": 6.328070069075573e-07,
+      "loss": 0.568,
+      "step": 17387
+    },
+    {
+      "epoch": 4.615956458250365,
+      "grad_norm": 0.47233389043203333,
+      "learning_rate": 6.325748683198105e-07,
+      "loss": 0.5428,
+      "step": 17388
+    },
+    {
+      "epoch": 4.616221956723749,
+      "grad_norm": 0.4953522092553053,
+      "learning_rate": 6.32342766151427e-07,
+      "loss": 0.5258,
+      "step": 17389
+    },
+    {
+      "epoch": 4.616487455197133,
+      "grad_norm": 0.47744473881326605,
+      "learning_rate": 6.321107004069321e-07,
+      "loss": 0.4984,
+      "step": 17390
+    },
+    {
+      "epoch": 4.616752953670517,
+      "grad_norm": 0.4678089315618051,
+      "learning_rate": 6.31878671090852e-07,
+      "loss": 0.5021,
+      "step": 17391
+    },
+    {
+      "epoch": 4.6170184521439,
+      "grad_norm": 0.4812693702339518,
+      "learning_rate": 6.31646678207711e-07,
+      "loss": 0.5154,
+      "step": 17392
+    },
+    {
+      "epoch": 4.617283950617284,
+      "grad_norm": 0.48341641424434456,
+      "learning_rate": 6.314147217620353e-07,
+      "loss": 0.523,
+      "step": 17393
+    },
+    {
+      "epoch": 4.617549449090668,
+      "grad_norm": 0.4760495685496969,
+      "learning_rate": 6.311828017583465e-07,
+      "loss": 0.5374,
+      "step": 17394
+    },
+    {
+      "epoch": 4.617814947564051,
+      "grad_norm": 0.4710700820505262,
+      "learning_rate": 6.309509182011697e-07,
+      "loss": 0.5217,
+      "step": 17395
+    },
+    {
+      "epoch": 4.618080446037435,
+      "grad_norm": 0.4706078408148169,
+      "learning_rate": 6.307190710950262e-07,
+      "loss": 0.5,
+      "step": 17396
+    },
+    {
+      "epoch": 4.6183459445108195,
+      "grad_norm": 0.4761084428167169,
+      "learning_rate": 6.304872604444368e-07,
+      "loss": 0.5032,
+      "step": 17397
+    },
+    {
+      "epoch": 4.618611442984203,
+      "grad_norm": 0.4809530096122509,
+      "learning_rate": 6.30255486253924e-07,
+      "loss": 0.5053,
+      "step": 17398
+    },
+    {
+      "epoch": 4.618876941457587,
+      "grad_norm": 0.48062221779797143,
+      "learning_rate": 6.300237485280061e-07,
+      "loss": 0.5306,
+      "step": 17399
+    },
+    {
+      "epoch": 4.61914243993097,
+      "grad_norm": 0.47712197772264653,
+      "learning_rate": 6.297920472712046e-07,
+      "loss": 0.5115,
+      "step": 17400
+    },
+    {
+      "epoch": 4.619407938404354,
+      "grad_norm": 0.4766162981676967,
+      "learning_rate": 6.295603824880364e-07,
+      "loss": 0.5269,
+      "step": 17401
+    },
+    {
+      "epoch": 4.619673436877738,
+      "grad_norm": 0.480338787414521,
+      "learning_rate": 6.293287541830221e-07,
+      "loss": 0.5034,
+      "step": 17402
+    },
+    {
+      "epoch": 4.619938935351122,
+      "grad_norm": 0.49319928129128415,
+      "learning_rate": 6.290971623606756e-07,
+      "loss": 0.5282,
+      "step": 17403
+    },
+    {
+      "epoch": 4.620204433824505,
+      "grad_norm": 0.49239175771260724,
+      "learning_rate": 6.288656070255151e-07,
+      "loss": 0.556,
+      "step": 17404
+    },
+    {
+      "epoch": 4.6204699322978895,
+      "grad_norm": 0.4860605010334084,
+      "learning_rate": 6.286340881820574e-07,
+      "loss": 0.54,
+      "step": 17405
+    },
+    {
+      "epoch": 4.620735430771273,
+      "grad_norm": 0.48913162410872707,
+      "learning_rate": 6.284026058348161e-07,
+      "loss": 0.5188,
+      "step": 17406
+    },
+    {
+      "epoch": 4.621000929244657,
+      "grad_norm": 0.5068455248672303,
+      "learning_rate": 6.281711599883075e-07,
+      "loss": 0.5666,
+      "step": 17407
+    },
+    {
+      "epoch": 4.621266427718041,
+      "grad_norm": 0.4877095348112145,
+      "learning_rate": 6.279397506470439e-07,
+      "loss": 0.5421,
+      "step": 17408
+    },
+    {
+      "epoch": 4.621531926191424,
+      "grad_norm": 0.47402155753709624,
+      "learning_rate": 6.27708377815539e-07,
+      "loss": 0.5685,
+      "step": 17409
+    },
+    {
+      "epoch": 4.621797424664808,
+      "grad_norm": 0.48390233346666245,
+      "learning_rate": 6.274770414983039e-07,
+      "loss": 0.5185,
+      "step": 17410
+    },
+    {
+      "epoch": 4.622062923138192,
+      "grad_norm": 0.48774895120884687,
+      "learning_rate": 6.272457416998523e-07,
+      "loss": 0.5138,
+      "step": 17411
+    },
+    {
+      "epoch": 4.622328421611575,
+      "grad_norm": 0.4851430959718481,
+      "learning_rate": 6.270144784246931e-07,
+      "loss": 0.5521,
+      "step": 17412
+    },
+    {
+      "epoch": 4.6225939200849595,
+      "grad_norm": 0.48529250735105295,
+      "learning_rate": 6.267832516773381e-07,
+      "loss": 0.5275,
+      "step": 17413
+    },
+    {
+      "epoch": 4.622859418558344,
+      "grad_norm": 0.476380753878276,
+      "learning_rate": 6.265520614622963e-07,
+      "loss": 0.5014,
+      "step": 17414
+    },
+    {
+      "epoch": 4.623124917031727,
+      "grad_norm": 0.4794223010468181,
+      "learning_rate": 6.263209077840754e-07,
+      "loss": 0.5295,
+      "step": 17415
+    },
+    {
+      "epoch": 4.623390415505111,
+      "grad_norm": 0.48075648986178093,
+      "learning_rate": 6.260897906471852e-07,
+      "loss": 0.5582,
+      "step": 17416
+    },
+    {
+      "epoch": 4.623655913978495,
+      "grad_norm": 0.469727595806367,
+      "learning_rate": 6.258587100561314e-07,
+      "loss": 0.515,
+      "step": 17417
+    },
+    {
+      "epoch": 4.623921412451878,
+      "grad_norm": 0.4619622467732009,
+      "learning_rate": 6.256276660154223e-07,
+      "loss": 0.521,
+      "step": 17418
+    },
+    {
+      "epoch": 4.624186910925262,
+      "grad_norm": 0.48091931716641556,
+      "learning_rate": 6.253966585295629e-07,
+      "loss": 0.5242,
+      "step": 17419
+    },
+    {
+      "epoch": 4.624452409398646,
+      "grad_norm": 0.46125218228401604,
+      "learning_rate": 6.251656876030585e-07,
+      "loss": 0.5219,
+      "step": 17420
+    },
+    {
+      "epoch": 4.6247179078720295,
+      "grad_norm": 0.4817523300192529,
+      "learning_rate": 6.249347532404126e-07,
+      "loss": 0.5312,
+      "step": 17421
+    },
+    {
+      "epoch": 4.624983406345414,
+      "grad_norm": 0.4762771483367275,
+      "learning_rate": 6.24703855446131e-07,
+      "loss": 0.5079,
+      "step": 17422
+    },
+    {
+      "epoch": 4.625248904818797,
+      "grad_norm": 0.44350279705194995,
+      "learning_rate": 6.244729942247149e-07,
+      "loss": 0.4568,
+      "step": 17423
+    },
+    {
+      "epoch": 4.625514403292181,
+      "grad_norm": 0.4818949817427838,
+      "learning_rate": 6.24242169580668e-07,
+      "loss": 0.5562,
+      "step": 17424
+    },
+    {
+      "epoch": 4.625779901765565,
+      "grad_norm": 0.452745423213986,
+      "learning_rate": 6.240113815184917e-07,
+      "loss": 0.542,
+      "step": 17425
+    },
+    {
+      "epoch": 4.626045400238949,
+      "grad_norm": 0.4824915490620668,
+      "learning_rate": 6.23780630042686e-07,
+      "loss": 0.5394,
+      "step": 17426
+    },
+    {
+      "epoch": 4.626310898712332,
+      "grad_norm": 0.48891760215921237,
+      "learning_rate": 6.235499151577523e-07,
+      "loss": 0.4947,
+      "step": 17427
+    },
+    {
+      "epoch": 4.626576397185716,
+      "grad_norm": 0.4944088870161946,
+      "learning_rate": 6.23319236868189e-07,
+      "loss": 0.5138,
+      "step": 17428
+    },
+    {
+      "epoch": 4.6268418956590995,
+      "grad_norm": 0.48849470466160577,
+      "learning_rate": 6.230885951784962e-07,
+      "loss": 0.5355,
+      "step": 17429
+    },
+    {
+      "epoch": 4.627107394132484,
+      "grad_norm": 0.4898893242228827,
+      "learning_rate": 6.228579900931705e-07,
+      "loss": 0.5185,
+      "step": 17430
+    },
+    {
+      "epoch": 4.627372892605868,
+      "grad_norm": 0.48601509631922596,
+      "learning_rate": 6.226274216167108e-07,
+      "loss": 0.5363,
+      "step": 17431
+    },
+    {
+      "epoch": 4.627638391079252,
+      "grad_norm": 0.49998110713082267,
+      "learning_rate": 6.223968897536132e-07,
+      "loss": 0.5733,
+      "step": 17432
+    },
+    {
+      "epoch": 4.627903889552635,
+      "grad_norm": 0.4759075321254341,
+      "learning_rate": 6.221663945083722e-07,
+      "loss": 0.5518,
+      "step": 17433
+    },
+    {
+      "epoch": 4.628169388026019,
+      "grad_norm": 0.460289053919721,
+      "learning_rate": 6.219359358854854e-07,
+      "loss": 0.5181,
+      "step": 17434
+    },
+    {
+      "epoch": 4.628434886499402,
+      "grad_norm": 0.4746934852634388,
+      "learning_rate": 6.217055138894452e-07,
+      "loss": 0.5158,
+      "step": 17435
+    },
+    {
+      "epoch": 4.628700384972786,
+      "grad_norm": 0.4862767813331973,
+      "learning_rate": 6.214751285247478e-07,
+      "loss": 0.5499,
+      "step": 17436
+    },
+    {
+      "epoch": 4.62896588344617,
+      "grad_norm": 0.4934686294367308,
+      "learning_rate": 6.212447797958835e-07,
+      "loss": 0.5463,
+      "step": 17437
+    },
+    {
+      "epoch": 4.629231381919554,
+      "grad_norm": 0.47940395996941554,
+      "learning_rate": 6.210144677073466e-07,
+      "loss": 0.5395,
+      "step": 17438
+    },
+    {
+      "epoch": 4.629496880392938,
+      "grad_norm": 0.47877706630815836,
+      "learning_rate": 6.207841922636273e-07,
+      "loss": 0.5907,
+      "step": 17439
+    },
+    {
+      "epoch": 4.629762378866322,
+      "grad_norm": 0.4773088501757365,
+      "learning_rate": 6.205539534692184e-07,
+      "loss": 0.5017,
+      "step": 17440
+    },
+    {
+      "epoch": 4.630027877339705,
+      "grad_norm": 0.48735265289989993,
+      "learning_rate": 6.203237513286081e-07,
+      "loss": 0.5603,
+      "step": 17441
+    },
+    {
+      "epoch": 4.630293375813089,
+      "grad_norm": 0.4860241075083344,
+      "learning_rate": 6.200935858462876e-07,
+      "loss": 0.5646,
+      "step": 17442
+    },
+    {
+      "epoch": 4.630558874286473,
+      "grad_norm": 0.47325123215489806,
+      "learning_rate": 6.19863457026745e-07,
+      "loss": 0.5362,
+      "step": 17443
+    },
+    {
+      "epoch": 4.630824372759856,
+      "grad_norm": 0.48808316553876985,
+      "learning_rate": 6.196333648744679e-07,
+      "loss": 0.5678,
+      "step": 17444
+    },
+    {
+      "epoch": 4.6310898712332405,
+      "grad_norm": 0.489241009603056,
+      "learning_rate": 6.194033093939444e-07,
+      "loss": 0.5231,
+      "step": 17445
+    },
+    {
+      "epoch": 4.6313553697066245,
+      "grad_norm": 0.48524170259881627,
+      "learning_rate": 6.191732905896605e-07,
+      "loss": 0.545,
+      "step": 17446
+    },
+    {
+      "epoch": 4.631620868180008,
+      "grad_norm": 0.48817508160181505,
+      "learning_rate": 6.189433084661032e-07,
+      "loss": 0.5665,
+      "step": 17447
+    },
+    {
+      "epoch": 4.631886366653392,
+      "grad_norm": 0.4853968632618056,
+      "learning_rate": 6.187133630277567e-07,
+      "loss": 0.5283,
+      "step": 17448
+    },
+    {
+      "epoch": 4.632151865126776,
+      "grad_norm": 0.49036848600591465,
+      "learning_rate": 6.184834542791052e-07,
+      "loss": 0.5415,
+      "step": 17449
+    },
+    {
+      "epoch": 4.632417363600159,
+      "grad_norm": 0.502976468837631,
+      "learning_rate": 6.182535822246341e-07,
+      "loss": 0.5731,
+      "step": 17450
+    },
+    {
+      "epoch": 4.632682862073543,
+      "grad_norm": 0.48508347639929883,
+      "learning_rate": 6.180237468688244e-07,
+      "loss": 0.5412,
+      "step": 17451
+    },
+    {
+      "epoch": 4.632948360546926,
+      "grad_norm": 0.4816568495650498,
+      "learning_rate": 6.177939482161605e-07,
+      "loss": 0.5283,
+      "step": 17452
+    },
+    {
+      "epoch": 4.6332138590203105,
+      "grad_norm": 0.46404991222252395,
+      "learning_rate": 6.17564186271122e-07,
+      "loss": 0.5133,
+      "step": 17453
+    },
+    {
+      "epoch": 4.6334793574936946,
+      "grad_norm": 0.47244615237967086,
+      "learning_rate": 6.173344610381924e-07,
+      "loss": 0.502,
+      "step": 17454
+    },
+    {
+      "epoch": 4.633744855967079,
+      "grad_norm": 0.47972845434142186,
+      "learning_rate": 6.17104772521849e-07,
+      "loss": 0.5492,
+      "step": 17455
+    },
+    {
+      "epoch": 4.634010354440462,
+      "grad_norm": 0.49119879894851753,
+      "learning_rate": 6.168751207265736e-07,
+      "loss": 0.562,
+      "step": 17456
+    },
+    {
+      "epoch": 4.634275852913846,
+      "grad_norm": 0.47486576763100935,
+      "learning_rate": 6.16645505656843e-07,
+      "loss": 0.5259,
+      "step": 17457
+    },
+    {
+      "epoch": 4.634541351387229,
+      "grad_norm": 0.4877641438497815,
+      "learning_rate": 6.164159273171369e-07,
+      "loss": 0.5323,
+      "step": 17458
+    },
+    {
+      "epoch": 4.634806849860613,
+      "grad_norm": 0.4961935495312679,
+      "learning_rate": 6.161863857119324e-07,
+      "loss": 0.5311,
+      "step": 17459
+    },
+    {
+      "epoch": 4.635072348333997,
+      "grad_norm": 0.481101578654917,
+      "learning_rate": 6.159568808457048e-07,
+      "loss": 0.5003,
+      "step": 17460
+    },
+    {
+      "epoch": 4.6353378468073805,
+      "grad_norm": 0.4839613214708843,
+      "learning_rate": 6.157274127229318e-07,
+      "loss": 0.555,
+      "step": 17461
+    },
+    {
+      "epoch": 4.635603345280765,
+      "grad_norm": 0.4812460667614524,
+      "learning_rate": 6.154979813480869e-07,
+      "loss": 0.5484,
+      "step": 17462
+    },
+    {
+      "epoch": 4.635868843754149,
+      "grad_norm": 0.4794861512602386,
+      "learning_rate": 6.152685867256461e-07,
+      "loss": 0.5363,
+      "step": 17463
+    },
+    {
+      "epoch": 4.636134342227532,
+      "grad_norm": 0.4834313531542578,
+      "learning_rate": 6.150392288600821e-07,
+      "loss": 0.516,
+      "step": 17464
+    },
+    {
+      "epoch": 4.636399840700916,
+      "grad_norm": 0.47850102674842715,
+      "learning_rate": 6.148099077558695e-07,
+      "loss": 0.5573,
+      "step": 17465
+    },
+    {
+      "epoch": 4.6366653391743,
+      "grad_norm": 0.47814185892697764,
+      "learning_rate": 6.145806234174778e-07,
+      "loss": 0.5377,
+      "step": 17466
+    },
+    {
+      "epoch": 4.636930837647683,
+      "grad_norm": 0.4851323906919248,
+      "learning_rate": 6.143513758493813e-07,
+      "loss": 0.5475,
+      "step": 17467
+    },
+    {
+      "epoch": 4.637196336121067,
+      "grad_norm": 0.470426608263812,
+      "learning_rate": 6.141221650560489e-07,
+      "loss": 0.5094,
+      "step": 17468
+    },
+    {
+      "epoch": 4.637461834594451,
+      "grad_norm": 0.4838404876221839,
+      "learning_rate": 6.138929910419525e-07,
+      "loss": 0.526,
+      "step": 17469
+    },
+    {
+      "epoch": 4.637727333067835,
+      "grad_norm": 0.4907255134779473,
+      "learning_rate": 6.136638538115603e-07,
+      "loss": 0.5207,
+      "step": 17470
+    },
+    {
+      "epoch": 4.637992831541219,
+      "grad_norm": 0.4741284719927159,
+      "learning_rate": 6.134347533693419e-07,
+      "loss": 0.5308,
+      "step": 17471
+    },
+    {
+      "epoch": 4.638258330014603,
+      "grad_norm": 0.4996884731539871,
+      "learning_rate": 6.13205689719765e-07,
+      "loss": 0.5452,
+      "step": 17472
+    },
+    {
+      "epoch": 4.638523828487986,
+      "grad_norm": 0.4956992712917525,
+      "learning_rate": 6.129766628672959e-07,
+      "loss": 0.5545,
+      "step": 17473
+    },
+    {
+      "epoch": 4.63878932696137,
+      "grad_norm": 0.4794383293491906,
+      "learning_rate": 6.127476728164031e-07,
+      "loss": 0.5232,
+      "step": 17474
+    },
+    {
+      "epoch": 4.639054825434754,
+      "grad_norm": 0.48137553056288507,
+      "learning_rate": 6.125187195715507e-07,
+      "loss": 0.5121,
+      "step": 17475
+    },
+    {
+      "epoch": 4.639320323908137,
+      "grad_norm": 0.4772370957135108,
+      "learning_rate": 6.122898031372052e-07,
+      "loss": 0.5319,
+      "step": 17476
+    },
+    {
+      "epoch": 4.639585822381521,
+      "grad_norm": 0.48624608121262863,
+      "learning_rate": 6.120609235178307e-07,
+      "loss": 0.5427,
+      "step": 17477
+    },
+    {
+      "epoch": 4.6398513208549055,
+      "grad_norm": 0.48644441560115587,
+      "learning_rate": 6.118320807178899e-07,
+      "loss": 0.5517,
+      "step": 17478
+    },
+    {
+      "epoch": 4.640116819328289,
+      "grad_norm": 0.4844592690779896,
+      "learning_rate": 6.116032747418471e-07,
+      "loss": 0.5303,
+      "step": 17479
+    },
+    {
+      "epoch": 4.640382317801673,
+      "grad_norm": 0.4701282192906114,
+      "learning_rate": 6.113745055941634e-07,
+      "loss": 0.5403,
+      "step": 17480
+    },
+    {
+      "epoch": 4.640647816275056,
+      "grad_norm": 0.4779649823021076,
+      "learning_rate": 6.111457732793017e-07,
+      "loss": 0.5288,
+      "step": 17481
+    },
+    {
+      "epoch": 4.64091331474844,
+      "grad_norm": 0.4885092674320545,
+      "learning_rate": 6.10917077801722e-07,
+      "loss": 0.5502,
+      "step": 17482
+    },
+    {
+      "epoch": 4.641178813221824,
+      "grad_norm": 0.4975695693618214,
+      "learning_rate": 6.106884191658846e-07,
+      "loss": 0.5174,
+      "step": 17483
+    },
+    {
+      "epoch": 4.641444311695208,
+      "grad_norm": 0.4849399329062805,
+      "learning_rate": 6.104597973762483e-07,
+      "loss": 0.5424,
+      "step": 17484
+    },
+    {
+      "epoch": 4.641709810168591,
+      "grad_norm": 0.4826121018508602,
+      "learning_rate": 6.102312124372731e-07,
+      "loss": 0.5203,
+      "step": 17485
+    },
+    {
+      "epoch": 4.6419753086419755,
+      "grad_norm": 0.4932130839163423,
+      "learning_rate": 6.100026643534154e-07,
+      "loss": 0.5564,
+      "step": 17486
+    },
+    {
+      "epoch": 4.642240807115359,
+      "grad_norm": 0.4724307312405337,
+      "learning_rate": 6.09774153129134e-07,
+      "loss": 0.5029,
+      "step": 17487
+    },
+    {
+      "epoch": 4.642506305588743,
+      "grad_norm": 0.47417809804209765,
+      "learning_rate": 6.095456787688847e-07,
+      "loss": 0.5157,
+      "step": 17488
+    },
+    {
+      "epoch": 4.642771804062127,
+      "grad_norm": 0.49435612741861995,
+      "learning_rate": 6.093172412771223e-07,
+      "loss": 0.5528,
+      "step": 17489
+    },
+    {
+      "epoch": 4.64303730253551,
+      "grad_norm": 0.4842382068137344,
+      "learning_rate": 6.090888406583039e-07,
+      "loss": 0.5226,
+      "step": 17490
+    },
+    {
+      "epoch": 4.643302801008894,
+      "grad_norm": 0.4918516301146902,
+      "learning_rate": 6.088604769168821e-07,
+      "loss": 0.5585,
+      "step": 17491
+    },
+    {
+      "epoch": 4.643568299482278,
+      "grad_norm": 0.4801535752771731,
+      "learning_rate": 6.08632150057312e-07,
+      "loss": 0.5415,
+      "step": 17492
+    },
+    {
+      "epoch": 4.643833797955661,
+      "grad_norm": 0.48523799197822826,
+      "learning_rate": 6.08403860084045e-07,
+      "loss": 0.52,
+      "step": 17493
+    },
+    {
+      "epoch": 4.6440992964290455,
+      "grad_norm": 0.4632014856315313,
+      "learning_rate": 6.081756070015358e-07,
+      "loss": 0.5296,
+      "step": 17494
+    },
+    {
+      "epoch": 4.64436479490243,
+      "grad_norm": 0.4837804512705638,
+      "learning_rate": 6.079473908142328e-07,
+      "loss": 0.5391,
+      "step": 17495
+    },
+    {
+      "epoch": 4.644630293375813,
+      "grad_norm": 0.4789590340781613,
+      "learning_rate": 6.077192115265881e-07,
+      "loss": 0.5223,
+      "step": 17496
+    },
+    {
+      "epoch": 4.644895791849197,
+      "grad_norm": 0.4735435715764255,
+      "learning_rate": 6.074910691430528e-07,
+      "loss": 0.5224,
+      "step": 17497
+    },
+    {
+      "epoch": 4.645161290322581,
+      "grad_norm": 0.48915196614027545,
+      "learning_rate": 6.072629636680744e-07,
+      "loss": 0.5418,
+      "step": 17498
+    },
+    {
+      "epoch": 4.645426788795964,
+      "grad_norm": 0.46584572037357247,
+      "learning_rate": 6.070348951061031e-07,
+      "loss": 0.5428,
+      "step": 17499
+    },
+    {
+      "epoch": 4.645692287269348,
+      "grad_norm": 0.48995319752052197,
+      "learning_rate": 6.068068634615862e-07,
+      "loss": 0.5644,
+      "step": 17500
+    },
+    {
+      "epoch": 4.645957785742732,
+      "grad_norm": 0.5090369309397824,
+      "learning_rate": 6.065788687389706e-07,
+      "loss": 0.5739,
+      "step": 17501
+    },
+    {
+      "epoch": 4.6462232842161155,
+      "grad_norm": 0.48019478309496105,
+      "learning_rate": 6.063509109427026e-07,
+      "loss": 0.5028,
+      "step": 17502
+    },
+    {
+      "epoch": 4.6464887826895,
+      "grad_norm": 0.4939007719504089,
+      "learning_rate": 6.061229900772286e-07,
+      "loss": 0.5528,
+      "step": 17503
+    },
+    {
+      "epoch": 4.646754281162884,
+      "grad_norm": 0.480556049844145,
+      "learning_rate": 6.058951061469931e-07,
+      "loss": 0.5177,
+      "step": 17504
+    },
+    {
+      "epoch": 4.647019779636267,
+      "grad_norm": 0.482218024270548,
+      "learning_rate": 6.056672591564409e-07,
+      "loss": 0.5476,
+      "step": 17505
+    },
+    {
+      "epoch": 4.647285278109651,
+      "grad_norm": 0.4849464296642132,
+      "learning_rate": 6.054394491100154e-07,
+      "loss": 0.5715,
+      "step": 17506
+    },
+    {
+      "epoch": 4.647550776583035,
+      "grad_norm": 0.4753320244116484,
+      "learning_rate": 6.052116760121588e-07,
+      "loss": 0.5483,
+      "step": 17507
+    },
+    {
+      "epoch": 4.647816275056418,
+      "grad_norm": 0.47972073926217157,
+      "learning_rate": 6.049839398673141e-07,
+      "loss": 0.5705,
+      "step": 17508
+    },
+    {
+      "epoch": 4.648081773529802,
+      "grad_norm": 0.4762911679638246,
+      "learning_rate": 6.04756240679922e-07,
+      "loss": 0.5069,
+      "step": 17509
+    },
+    {
+      "epoch": 4.6483472720031855,
+      "grad_norm": 0.4792631696037987,
+      "learning_rate": 6.045285784544242e-07,
+      "loss": 0.5519,
+      "step": 17510
+    },
+    {
+      "epoch": 4.64861277047657,
+      "grad_norm": 0.4641960891860845,
+      "learning_rate": 6.043009531952599e-07,
+      "loss": 0.5268,
+      "step": 17511
+    },
+    {
+      "epoch": 4.648878268949954,
+      "grad_norm": 0.4771457169299319,
+      "learning_rate": 6.040733649068686e-07,
+      "loss": 0.5365,
+      "step": 17512
+    },
+    {
+      "epoch": 4.649143767423338,
+      "grad_norm": 0.48621309220103204,
+      "learning_rate": 6.038458135936878e-07,
+      "loss": 0.5445,
+      "step": 17513
+    },
+    {
+      "epoch": 4.649409265896721,
+      "grad_norm": 0.49196159482123314,
+      "learning_rate": 6.036182992601572e-07,
+      "loss": 0.549,
+      "step": 17514
+    },
+    {
+      "epoch": 4.649674764370105,
+      "grad_norm": 0.4869467828458674,
+      "learning_rate": 6.033908219107121e-07,
+      "loss": 0.5067,
+      "step": 17515
+    },
+    {
+      "epoch": 4.649940262843488,
+      "grad_norm": 0.4963452041268275,
+      "learning_rate": 6.031633815497894e-07,
+      "loss": 0.5486,
+      "step": 17516
+    },
+    {
+      "epoch": 4.650205761316872,
+      "grad_norm": 0.47970259841092555,
+      "learning_rate": 6.02935978181827e-07,
+      "loss": 0.5442,
+      "step": 17517
+    },
+    {
+      "epoch": 4.650471259790256,
+      "grad_norm": 0.47584059691761377,
+      "learning_rate": 6.02708611811256e-07,
+      "loss": 0.5265,
+      "step": 17518
+    },
+    {
+      "epoch": 4.65073675826364,
+      "grad_norm": 0.48528805935740743,
+      "learning_rate": 6.024812824425133e-07,
+      "loss": 0.5506,
+      "step": 17519
+    },
+    {
+      "epoch": 4.651002256737024,
+      "grad_norm": 0.4871816146120627,
+      "learning_rate": 6.022539900800306e-07,
+      "loss": 0.5518,
+      "step": 17520
+    },
+    {
+      "epoch": 4.651267755210408,
+      "grad_norm": 0.48405214628733373,
+      "learning_rate": 6.020267347282427e-07,
+      "loss": 0.5479,
+      "step": 17521
+    },
+    {
+      "epoch": 4.651533253683791,
+      "grad_norm": 0.47814266671981204,
+      "learning_rate": 6.017995163915794e-07,
+      "loss": 0.5528,
+      "step": 17522
+    },
+    {
+      "epoch": 4.651798752157175,
+      "grad_norm": 0.4774122536340053,
+      "learning_rate": 6.01572335074474e-07,
+      "loss": 0.5079,
+      "step": 17523
+    },
+    {
+      "epoch": 4.652064250630559,
+      "grad_norm": 0.47598598565462225,
+      "learning_rate": 6.013451907813561e-07,
+      "loss": 0.5323,
+      "step": 17524
+    },
+    {
+      "epoch": 4.652329749103942,
+      "grad_norm": 0.47714727082369907,
+      "learning_rate": 6.011180835166553e-07,
+      "loss": 0.5472,
+      "step": 17525
+    },
+    {
+      "epoch": 4.6525952475773265,
+      "grad_norm": 0.4914648203897081,
+      "learning_rate": 6.008910132848018e-07,
+      "loss": 0.5788,
+      "step": 17526
+    },
+    {
+      "epoch": 4.6528607460507105,
+      "grad_norm": 0.48297420672682345,
+      "learning_rate": 6.006639800902223e-07,
+      "loss": 0.5333,
+      "step": 17527
+    },
+    {
+      "epoch": 4.653126244524094,
+      "grad_norm": 0.5029774032039211,
+      "learning_rate": 6.004369839373464e-07,
+      "loss": 0.5255,
+      "step": 17528
+    },
+    {
+      "epoch": 4.653391742997478,
+      "grad_norm": 0.4906322123643664,
+      "learning_rate": 6.002100248306003e-07,
+      "loss": 0.5043,
+      "step": 17529
+    },
+    {
+      "epoch": 4.653657241470862,
+      "grad_norm": 0.47776693799916914,
+      "learning_rate": 5.999831027744102e-07,
+      "loss": 0.5192,
+      "step": 17530
+    },
+    {
+      "epoch": 4.653922739944245,
+      "grad_norm": 0.4761885990384588,
+      "learning_rate": 5.997562177732011e-07,
+      "loss": 0.5277,
+      "step": 17531
+    },
+    {
+      "epoch": 4.654188238417629,
+      "grad_norm": 0.48960847557217146,
+      "learning_rate": 5.995293698313989e-07,
+      "loss": 0.541,
+      "step": 17532
+    },
+    {
+      "epoch": 4.654453736891012,
+      "grad_norm": 0.46827458238421205,
+      "learning_rate": 5.993025589534263e-07,
+      "loss": 0.5272,
+      "step": 17533
+    },
+    {
+      "epoch": 4.6547192353643965,
+      "grad_norm": 0.48098516453084234,
+      "learning_rate": 5.990757851437084e-07,
+      "loss": 0.5146,
+      "step": 17534
+    },
+    {
+      "epoch": 4.6549847338377806,
+      "grad_norm": 0.48522723591487577,
+      "learning_rate": 5.98849048406667e-07,
+      "loss": 0.5346,
+      "step": 17535
+    },
+    {
+      "epoch": 4.655250232311165,
+      "grad_norm": 0.4885720496346273,
+      "learning_rate": 5.98622348746723e-07,
+      "loss": 0.6009,
+      "step": 17536
+    },
+    {
+      "epoch": 4.655515730784548,
+      "grad_norm": 0.4724206205717385,
+      "learning_rate": 5.983956861682996e-07,
+      "loss": 0.521,
+      "step": 17537
+    },
+    {
+      "epoch": 4.655781229257932,
+      "grad_norm": 0.48080486246124465,
+      "learning_rate": 5.981690606758153e-07,
+      "loss": 0.5242,
+      "step": 17538
+    },
+    {
+      "epoch": 4.656046727731315,
+      "grad_norm": 0.48382632734128944,
+      "learning_rate": 5.979424722736917e-07,
+      "loss": 0.561,
+      "step": 17539
+    },
+    {
+      "epoch": 4.656312226204699,
+      "grad_norm": 0.47459040019245186,
+      "learning_rate": 5.977159209663469e-07,
+      "loss": 0.522,
+      "step": 17540
+    },
+    {
+      "epoch": 4.656577724678083,
+      "grad_norm": 0.4960771640876124,
+      "learning_rate": 5.974894067581985e-07,
+      "loss": 0.5528,
+      "step": 17541
+    },
+    {
+      "epoch": 4.656843223151467,
+      "grad_norm": 0.4946943687706496,
+      "learning_rate": 5.972629296536656e-07,
+      "loss": 0.5504,
+      "step": 17542
+    },
+    {
+      "epoch": 4.657108721624851,
+      "grad_norm": 0.47413147783826876,
+      "learning_rate": 5.970364896571632e-07,
+      "loss": 0.5345,
+      "step": 17543
+    },
+    {
+      "epoch": 4.657374220098235,
+      "grad_norm": 0.48884385933972324,
+      "learning_rate": 5.968100867731097e-07,
+      "loss": 0.5576,
+      "step": 17544
+    },
+    {
+      "epoch": 4.657639718571618,
+      "grad_norm": 0.4723688093824866,
+      "learning_rate": 5.965837210059183e-07,
+      "loss": 0.5052,
+      "step": 17545
+    },
+    {
+      "epoch": 4.657905217045002,
+      "grad_norm": 0.5053242426606632,
+      "learning_rate": 5.963573923600061e-07,
+      "loss": 0.5286,
+      "step": 17546
+    },
+    {
+      "epoch": 4.658170715518386,
+      "grad_norm": 0.49443720557402704,
+      "learning_rate": 5.961311008397844e-07,
+      "loss": 0.5357,
+      "step": 17547
+    },
+    {
+      "epoch": 4.658436213991769,
+      "grad_norm": 0.4882227340649235,
+      "learning_rate": 5.959048464496684e-07,
+      "loss": 0.5503,
+      "step": 17548
+    },
+    {
+      "epoch": 4.658701712465153,
+      "grad_norm": 0.48024454238455566,
+      "learning_rate": 5.956786291940692e-07,
+      "loss": 0.5528,
+      "step": 17549
+    },
+    {
+      "epoch": 4.658967210938537,
+      "grad_norm": 0.4870678964927204,
+      "learning_rate": 5.954524490774e-07,
+      "loss": 0.5456,
+      "step": 17550
+    },
+    {
+      "epoch": 4.659232709411921,
+      "grad_norm": 0.4905170549489484,
+      "learning_rate": 5.952263061040706e-07,
+      "loss": 0.5258,
+      "step": 17551
+    },
+    {
+      "epoch": 4.659498207885305,
+      "grad_norm": 0.4659922391824595,
+      "learning_rate": 5.950002002784925e-07,
+      "loss": 0.5226,
+      "step": 17552
+    },
+    {
+      "epoch": 4.659763706358689,
+      "grad_norm": 0.49058509734621847,
+      "learning_rate": 5.947741316050748e-07,
+      "loss": 0.5563,
+      "step": 17553
+    },
+    {
+      "epoch": 4.660029204832072,
+      "grad_norm": 0.49105648665873314,
+      "learning_rate": 5.945481000882256e-07,
+      "loss": 0.485,
+      "step": 17554
+    },
+    {
+      "epoch": 4.660294703305456,
+      "grad_norm": 0.46140717033832296,
+      "learning_rate": 5.943221057323545e-07,
+      "loss": 0.5257,
+      "step": 17555
+    },
+    {
+      "epoch": 4.66056020177884,
+      "grad_norm": 0.47002253033801006,
+      "learning_rate": 5.940961485418676e-07,
+      "loss": 0.5219,
+      "step": 17556
+    },
+    {
+      "epoch": 4.660825700252223,
+      "grad_norm": 0.4683450032792175,
+      "learning_rate": 5.938702285211734e-07,
+      "loss": 0.5507,
+      "step": 17557
+    },
+    {
+      "epoch": 4.661091198725607,
+      "grad_norm": 0.4821275873845253,
+      "learning_rate": 5.936443456746757e-07,
+      "loss": 0.5077,
+      "step": 17558
+    },
+    {
+      "epoch": 4.6613566971989915,
+      "grad_norm": 0.47015878379510917,
+      "learning_rate": 5.934185000067813e-07,
+      "loss": 0.5359,
+      "step": 17559
+    },
+    {
+      "epoch": 4.661622195672375,
+      "grad_norm": 0.4819174436246306,
+      "learning_rate": 5.931926915218936e-07,
+      "loss": 0.5516,
+      "step": 17560
+    },
+    {
+      "epoch": 4.661887694145759,
+      "grad_norm": 0.49564587563414625,
+      "learning_rate": 5.929669202244171e-07,
+      "loss": 0.5257,
+      "step": 17561
+    },
+    {
+      "epoch": 4.662153192619142,
+      "grad_norm": 0.4722876064451484,
+      "learning_rate": 5.927411861187557e-07,
+      "loss": 0.5582,
+      "step": 17562
+    },
+    {
+      "epoch": 4.662418691092526,
+      "grad_norm": 0.4713146127236207,
+      "learning_rate": 5.925154892093107e-07,
+      "loss": 0.5088,
+      "step": 17563
+    },
+    {
+      "epoch": 4.66268418956591,
+      "grad_norm": 0.4978120518295372,
+      "learning_rate": 5.922898295004842e-07,
+      "loss": 0.5377,
+      "step": 17564
+    },
+    {
+      "epoch": 4.662949688039294,
+      "grad_norm": 0.4840002431160575,
+      "learning_rate": 5.920642069966761e-07,
+      "loss": 0.5594,
+      "step": 17565
+    },
+    {
+      "epoch": 4.663215186512677,
+      "grad_norm": 0.48583530444246537,
+      "learning_rate": 5.918386217022882e-07,
+      "loss": 0.5554,
+      "step": 17566
+    },
+    {
+      "epoch": 4.6634806849860615,
+      "grad_norm": 0.49066548228280116,
+      "learning_rate": 5.916130736217184e-07,
+      "loss": 0.5542,
+      "step": 17567
+    },
+    {
+      "epoch": 4.663746183459445,
+      "grad_norm": 0.48258149995797395,
+      "learning_rate": 5.913875627593668e-07,
+      "loss": 0.5564,
+      "step": 17568
+    },
+    {
+      "epoch": 4.664011681932829,
+      "grad_norm": 0.47329799252740956,
+      "learning_rate": 5.91162089119631e-07,
+      "loss": 0.5197,
+      "step": 17569
+    },
+    {
+      "epoch": 4.664277180406213,
+      "grad_norm": 0.4713087173524675,
+      "learning_rate": 5.909366527069074e-07,
+      "loss": 0.5379,
+      "step": 17570
+    },
+    {
+      "epoch": 4.664542678879596,
+      "grad_norm": 0.5033998052036267,
+      "learning_rate": 5.907112535255938e-07,
+      "loss": 0.5525,
+      "step": 17571
+    },
+    {
+      "epoch": 4.66480817735298,
+      "grad_norm": 0.4879632228198571,
+      "learning_rate": 5.904858915800849e-07,
+      "loss": 0.5625,
+      "step": 17572
+    },
+    {
+      "epoch": 4.665073675826364,
+      "grad_norm": 0.4731921613301923,
+      "learning_rate": 5.902605668747771e-07,
+      "loss": 0.5282,
+      "step": 17573
+    },
+    {
+      "epoch": 4.665339174299747,
+      "grad_norm": 0.504304310238166,
+      "learning_rate": 5.900352794140635e-07,
+      "loss": 0.5786,
+      "step": 17574
+    },
+    {
+      "epoch": 4.6656046727731315,
+      "grad_norm": 0.48854622053171703,
+      "learning_rate": 5.898100292023396e-07,
+      "loss": 0.5413,
+      "step": 17575
+    },
+    {
+      "epoch": 4.665870171246516,
+      "grad_norm": 0.46770579152568525,
+      "learning_rate": 5.895848162439957e-07,
+      "loss": 0.5137,
+      "step": 17576
+    },
+    {
+      "epoch": 4.666135669719899,
+      "grad_norm": 0.4809217873879484,
+      "learning_rate": 5.89359640543426e-07,
+      "loss": 0.5576,
+      "step": 17577
+    },
+    {
+      "epoch": 4.666401168193283,
+      "grad_norm": 0.4846342404448908,
+      "learning_rate": 5.89134502105021e-07,
+      "loss": 0.5294,
+      "step": 17578
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 0.46090536410155997,
+      "learning_rate": 5.889094009331723e-07,
+      "loss": 0.5338,
+      "step": 17579
+    },
+    {
+      "epoch": 4.66693216514005,
+      "grad_norm": 0.4896591115933786,
+      "learning_rate": 5.886843370322692e-07,
+      "loss": 0.5715,
+      "step": 17580
+    },
+    {
+      "epoch": 4.667197663613434,
+      "grad_norm": 0.48910987785018056,
+      "learning_rate": 5.884593104067007e-07,
+      "loss": 0.532,
+      "step": 17581
+    },
+    {
+      "epoch": 4.667463162086818,
+      "grad_norm": 0.4765665283913625,
+      "learning_rate": 5.882343210608568e-07,
+      "loss": 0.5524,
+      "step": 17582
+    },
+    {
+      "epoch": 4.6677286605602015,
+      "grad_norm": 0.48390681919362444,
+      "learning_rate": 5.880093689991237e-07,
+      "loss": 0.5548,
+      "step": 17583
+    },
+    {
+      "epoch": 4.667994159033586,
+      "grad_norm": 0.47642255429498725,
+      "learning_rate": 5.877844542258898e-07,
+      "loss": 0.5641,
+      "step": 17584
+    },
+    {
+      "epoch": 4.66825965750697,
+      "grad_norm": 0.48236550095023334,
+      "learning_rate": 5.875595767455403e-07,
+      "loss": 0.5566,
+      "step": 17585
+    },
+    {
+      "epoch": 4.668525155980353,
+      "grad_norm": 0.46751842960521256,
+      "learning_rate": 5.873347365624624e-07,
+      "loss": 0.53,
+      "step": 17586
+    },
+    {
+      "epoch": 4.668790654453737,
+      "grad_norm": 0.4742034038382578,
+      "learning_rate": 5.871099336810399e-07,
+      "loss": 0.5416,
+      "step": 17587
+    },
+    {
+      "epoch": 4.669056152927121,
+      "grad_norm": 0.4814675753361654,
+      "learning_rate": 5.868851681056567e-07,
+      "loss": 0.5396,
+      "step": 17588
+    },
+    {
+      "epoch": 4.669321651400504,
+      "grad_norm": 0.4977385163735499,
+      "learning_rate": 5.866604398406974e-07,
+      "loss": 0.5444,
+      "step": 17589
+    },
+    {
+      "epoch": 4.669587149873888,
+      "grad_norm": 0.4851284381527935,
+      "learning_rate": 5.864357488905436e-07,
+      "loss": 0.5472,
+      "step": 17590
+    },
+    {
+      "epoch": 4.6698526483472715,
+      "grad_norm": 0.48440975359483257,
+      "learning_rate": 5.862110952595784e-07,
+      "loss": 0.5273,
+      "step": 17591
+    },
+    {
+      "epoch": 4.670118146820656,
+      "grad_norm": 0.4836313881015596,
+      "learning_rate": 5.859864789521829e-07,
+      "loss": 0.5526,
+      "step": 17592
+    },
+    {
+      "epoch": 4.67038364529404,
+      "grad_norm": 0.48868472492330045,
+      "learning_rate": 5.857618999727374e-07,
+      "loss": 0.5491,
+      "step": 17593
+    },
+    {
+      "epoch": 4.670649143767424,
+      "grad_norm": 0.4733038777661721,
+      "learning_rate": 5.855373583256208e-07,
+      "loss": 0.5218,
+      "step": 17594
+    },
+    {
+      "epoch": 4.670914642240807,
+      "grad_norm": 0.49228433959135315,
+      "learning_rate": 5.85312854015214e-07,
+      "loss": 0.5149,
+      "step": 17595
+    },
+    {
+      "epoch": 4.671180140714191,
+      "grad_norm": 0.4849395865814369,
+      "learning_rate": 5.850883870458938e-07,
+      "loss": 0.5666,
+      "step": 17596
+    },
+    {
+      "epoch": 4.671445639187574,
+      "grad_norm": 0.49142280099286195,
+      "learning_rate": 5.848639574220388e-07,
+      "loss": 0.5419,
+      "step": 17597
+    },
+    {
+      "epoch": 4.671711137660958,
+      "grad_norm": 0.47582298862092975,
+      "learning_rate": 5.846395651480263e-07,
+      "loss": 0.549,
+      "step": 17598
+    },
+    {
+      "epoch": 4.671976636134342,
+      "grad_norm": 0.4969085833713176,
+      "learning_rate": 5.844152102282308e-07,
+      "loss": 0.5531,
+      "step": 17599
+    },
+    {
+      "epoch": 4.672242134607726,
+      "grad_norm": 0.48560548166621376,
+      "learning_rate": 5.841908926670298e-07,
+      "loss": 0.5412,
+      "step": 17600
+    },
+    {
+      "epoch": 4.67250763308111,
+      "grad_norm": 0.4881200512521312,
+      "learning_rate": 5.839666124687963e-07,
+      "loss": 0.5385,
+      "step": 17601
+    },
+    {
+      "epoch": 4.672773131554494,
+      "grad_norm": 0.4848123636599459,
+      "learning_rate": 5.837423696379058e-07,
+      "loss": 0.5515,
+      "step": 17602
+    },
+    {
+      "epoch": 4.673038630027877,
+      "grad_norm": 0.48706127766612134,
+      "learning_rate": 5.83518164178731e-07,
+      "loss": 0.5335,
+      "step": 17603
+    },
+    {
+      "epoch": 4.673304128501261,
+      "grad_norm": 0.47707888814198224,
+      "learning_rate": 5.832939960956443e-07,
+      "loss": 0.5456,
+      "step": 17604
+    },
+    {
+      "epoch": 4.673569626974645,
+      "grad_norm": 0.48954553381221605,
+      "learning_rate": 5.83069865393017e-07,
+      "loss": 0.5707,
+      "step": 17605
+    },
+    {
+      "epoch": 4.673835125448028,
+      "grad_norm": 0.4723074869203257,
+      "learning_rate": 5.828457720752212e-07,
+      "loss": 0.525,
+      "step": 17606
+    },
+    {
+      "epoch": 4.6741006239214125,
+      "grad_norm": 0.47291595883586074,
+      "learning_rate": 5.826217161466263e-07,
+      "loss": 0.5376,
+      "step": 17607
+    },
+    {
+      "epoch": 4.6743661223947965,
+      "grad_norm": 0.49306722599386016,
+      "learning_rate": 5.823976976116025e-07,
+      "loss": 0.529,
+      "step": 17608
+    },
+    {
+      "epoch": 4.67463162086818,
+      "grad_norm": 0.4845012409211947,
+      "learning_rate": 5.8217371647452e-07,
+      "loss": 0.5562,
+      "step": 17609
+    },
+    {
+      "epoch": 4.674897119341564,
+      "grad_norm": 0.48527092002464906,
+      "learning_rate": 5.819497727397444e-07,
+      "loss": 0.5487,
+      "step": 17610
+    },
+    {
+      "epoch": 4.675162617814948,
+      "grad_norm": 0.48576677735608037,
+      "learning_rate": 5.817258664116451e-07,
+      "loss": 0.5898,
+      "step": 17611
+    },
+    {
+      "epoch": 4.675428116288331,
+      "grad_norm": 0.47639349612336684,
+      "learning_rate": 5.815019974945873e-07,
+      "loss": 0.5193,
+      "step": 17612
+    },
+    {
+      "epoch": 4.675693614761715,
+      "grad_norm": 0.4838384117617082,
+      "learning_rate": 5.81278165992939e-07,
+      "loss": 0.5582,
+      "step": 17613
+    },
+    {
+      "epoch": 4.675959113235099,
+      "grad_norm": 0.4623622340569345,
+      "learning_rate": 5.810543719110631e-07,
+      "loss": 0.5408,
+      "step": 17614
+    },
+    {
+      "epoch": 4.6762246117084825,
+      "grad_norm": 0.4900859380467966,
+      "learning_rate": 5.808306152533264e-07,
+      "loss": 0.5425,
+      "step": 17615
+    },
+    {
+      "epoch": 4.6764901101818666,
+      "grad_norm": 0.4803258612960572,
+      "learning_rate": 5.806068960240912e-07,
+      "loss": 0.5525,
+      "step": 17616
+    },
+    {
+      "epoch": 4.676755608655251,
+      "grad_norm": 0.4684219371412779,
+      "learning_rate": 5.803832142277208e-07,
+      "loss": 0.4709,
+      "step": 17617
+    },
+    {
+      "epoch": 4.677021107128634,
+      "grad_norm": 0.4851800992748377,
+      "learning_rate": 5.80159569868578e-07,
+      "loss": 0.5329,
+      "step": 17618
+    },
+    {
+      "epoch": 4.677286605602018,
+      "grad_norm": 0.4838513057364973,
+      "learning_rate": 5.799359629510238e-07,
+      "loss": 0.5374,
+      "step": 17619
+    },
+    {
+      "epoch": 4.677552104075401,
+      "grad_norm": 0.4764960125443222,
+      "learning_rate": 5.797123934794199e-07,
+      "loss": 0.549,
+      "step": 17620
+    },
+    {
+      "epoch": 4.677817602548785,
+      "grad_norm": 0.47855145978086844,
+      "learning_rate": 5.794888614581262e-07,
+      "loss": 0.5184,
+      "step": 17621
+    },
+    {
+      "epoch": 4.678083101022169,
+      "grad_norm": 0.4755579249056458,
+      "learning_rate": 5.792653668915019e-07,
+      "loss": 0.5589,
+      "step": 17622
+    },
+    {
+      "epoch": 4.678348599495553,
+      "grad_norm": 0.4753189801687665,
+      "learning_rate": 5.79041909783905e-07,
+      "loss": 0.5257,
+      "step": 17623
+    },
+    {
+      "epoch": 4.678614097968937,
+      "grad_norm": 0.48193284402381986,
+      "learning_rate": 5.788184901396948e-07,
+      "loss": 0.5382,
+      "step": 17624
+    },
+    {
+      "epoch": 4.678879596442321,
+      "grad_norm": 0.4890202180691973,
+      "learning_rate": 5.785951079632273e-07,
+      "loss": 0.5439,
+      "step": 17625
+    },
+    {
+      "epoch": 4.679145094915704,
+      "grad_norm": 0.48442205472478506,
+      "learning_rate": 5.783717632588604e-07,
+      "loss": 0.5433,
+      "step": 17626
+    },
+    {
+      "epoch": 4.679410593389088,
+      "grad_norm": 0.4760361706132955,
+      "learning_rate": 5.781484560309489e-07,
+      "loss": 0.512,
+      "step": 17627
+    },
+    {
+      "epoch": 4.679676091862472,
+      "grad_norm": 0.46979925260422123,
+      "learning_rate": 5.779251862838473e-07,
+      "loss": 0.5184,
+      "step": 17628
+    },
+    {
+      "epoch": 4.679941590335855,
+      "grad_norm": 0.48876903818091666,
+      "learning_rate": 5.777019540219115e-07,
+      "loss": 0.5271,
+      "step": 17629
+    },
+    {
+      "epoch": 4.680207088809239,
+      "grad_norm": 0.4919130465922979,
+      "learning_rate": 5.774787592494932e-07,
+      "loss": 0.5367,
+      "step": 17630
+    },
+    {
+      "epoch": 4.680472587282623,
+      "grad_norm": 0.4912653966397032,
+      "learning_rate": 5.772556019709471e-07,
+      "loss": 0.5471,
+      "step": 17631
+    },
+    {
+      "epoch": 4.680738085756007,
+      "grad_norm": 0.4677955252629279,
+      "learning_rate": 5.770324821906245e-07,
+      "loss": 0.5156,
+      "step": 17632
+    },
+    {
+      "epoch": 4.681003584229391,
+      "grad_norm": 0.4782225828547463,
+      "learning_rate": 5.76809399912876e-07,
+      "loss": 0.548,
+      "step": 17633
+    },
+    {
+      "epoch": 4.681269082702775,
+      "grad_norm": 0.5007467722677937,
+      "learning_rate": 5.765863551420536e-07,
+      "loss": 0.5408,
+      "step": 17634
+    },
+    {
+      "epoch": 4.681534581176158,
+      "grad_norm": 0.4791252201571909,
+      "learning_rate": 5.763633478825062e-07,
+      "loss": 0.5242,
+      "step": 17635
+    },
+    {
+      "epoch": 4.681800079649542,
+      "grad_norm": 0.48141369199550105,
+      "learning_rate": 5.761403781385838e-07,
+      "loss": 0.5128,
+      "step": 17636
+    },
+    {
+      "epoch": 4.682065578122926,
+      "grad_norm": 0.4851972073438632,
+      "learning_rate": 5.759174459146338e-07,
+      "loss": 0.5166,
+      "step": 17637
+    },
+    {
+      "epoch": 4.682331076596309,
+      "grad_norm": 0.4688306452182194,
+      "learning_rate": 5.756945512150062e-07,
+      "loss": 0.4894,
+      "step": 17638
+    },
+    {
+      "epoch": 4.682596575069693,
+      "grad_norm": 0.4763806003390963,
+      "learning_rate": 5.754716940440444e-07,
+      "loss": 0.5412,
+      "step": 17639
+    },
+    {
+      "epoch": 4.6828620735430775,
+      "grad_norm": 0.467762019792464,
+      "learning_rate": 5.752488744060977e-07,
+      "loss": 0.5163,
+      "step": 17640
+    },
+    {
+      "epoch": 4.683127572016461,
+      "grad_norm": 0.48043048877006794,
+      "learning_rate": 5.750260923055098e-07,
+      "loss": 0.5481,
+      "step": 17641
+    },
+    {
+      "epoch": 4.683393070489845,
+      "grad_norm": 0.4782161327375079,
+      "learning_rate": 5.748033477466272e-07,
+      "loss": 0.534,
+      "step": 17642
+    },
+    {
+      "epoch": 4.683658568963228,
+      "grad_norm": 0.48181523038820157,
+      "learning_rate": 5.74580640733792e-07,
+      "loss": 0.4998,
+      "step": 17643
+    },
+    {
+      "epoch": 4.683924067436612,
+      "grad_norm": 0.48400657392850793,
+      "learning_rate": 5.743579712713493e-07,
+      "loss": 0.534,
+      "step": 17644
+    },
+    {
+      "epoch": 4.684189565909996,
+      "grad_norm": 0.48203313466833236,
+      "learning_rate": 5.741353393636409e-07,
+      "loss": 0.5444,
+      "step": 17645
+    },
+    {
+      "epoch": 4.68445506438338,
+      "grad_norm": 0.46303026024735244,
+      "learning_rate": 5.739127450150081e-07,
+      "loss": 0.5056,
+      "step": 17646
+    },
+    {
+      "epoch": 4.684720562856763,
+      "grad_norm": 0.49002623863633915,
+      "learning_rate": 5.736901882297935e-07,
+      "loss": 0.5391,
+      "step": 17647
+    },
+    {
+      "epoch": 4.6849860613301475,
+      "grad_norm": 0.48493375159790675,
+      "learning_rate": 5.734676690123359e-07,
+      "loss": 0.5643,
+      "step": 17648
+    },
+    {
+      "epoch": 4.685251559803531,
+      "grad_norm": 0.47225961223539814,
+      "learning_rate": 5.732451873669762e-07,
+      "loss": 0.588,
+      "step": 17649
+    },
+    {
+      "epoch": 4.685517058276915,
+      "grad_norm": 0.4825259799409351,
+      "learning_rate": 5.730227432980529e-07,
+      "loss": 0.5431,
+      "step": 17650
+    },
+    {
+      "epoch": 4.685782556750299,
+      "grad_norm": 0.47259817650436947,
+      "learning_rate": 5.728003368099042e-07,
+      "loss": 0.5239,
+      "step": 17651
+    },
+    {
+      "epoch": 4.686048055223683,
+      "grad_norm": 0.48974897267404094,
+      "learning_rate": 5.725779679068669e-07,
+      "loss": 0.5031,
+      "step": 17652
+    },
+    {
+      "epoch": 4.686313553697066,
+      "grad_norm": 0.4779099885894124,
+      "learning_rate": 5.723556365932783e-07,
+      "loss": 0.5467,
+      "step": 17653
+    },
+    {
+      "epoch": 4.68657905217045,
+      "grad_norm": 0.4725783187954063,
+      "learning_rate": 5.721333428734751e-07,
+      "loss": 0.5717,
+      "step": 17654
+    },
+    {
+      "epoch": 4.686844550643833,
+      "grad_norm": 0.4718272669169045,
+      "learning_rate": 5.719110867517921e-07,
+      "loss": 0.4784,
+      "step": 17655
+    },
+    {
+      "epoch": 4.6871100491172175,
+      "grad_norm": 0.4787208466826315,
+      "learning_rate": 5.716888682325636e-07,
+      "loss": 0.5243,
+      "step": 17656
+    },
+    {
+      "epoch": 4.687375547590602,
+      "grad_norm": 0.48185092113592476,
+      "learning_rate": 5.714666873201227e-07,
+      "loss": 0.5194,
+      "step": 17657
+    },
+    {
+      "epoch": 4.687641046063985,
+      "grad_norm": 0.46988242520776813,
+      "learning_rate": 5.712445440188042e-07,
+      "loss": 0.5385,
+      "step": 17658
+    },
+    {
+      "epoch": 4.687906544537369,
+      "grad_norm": 0.4992162803197245,
+      "learning_rate": 5.710224383329388e-07,
+      "loss": 0.5161,
+      "step": 17659
+    },
+    {
+      "epoch": 4.688172043010753,
+      "grad_norm": 0.4853268441272868,
+      "learning_rate": 5.708003702668594e-07,
+      "loss": 0.5066,
+      "step": 17660
+    },
+    {
+      "epoch": 4.688437541484136,
+      "grad_norm": 0.49216437185841266,
+      "learning_rate": 5.705783398248962e-07,
+      "loss": 0.5309,
+      "step": 17661
+    },
+    {
+      "epoch": 4.68870303995752,
+      "grad_norm": 0.48741835680902423,
+      "learning_rate": 5.703563470113787e-07,
+      "loss": 0.5506,
+      "step": 17662
+    },
+    {
+      "epoch": 4.688968538430904,
+      "grad_norm": 0.49334566346511916,
+      "learning_rate": 5.701343918306379e-07,
+      "loss": 0.5507,
+      "step": 17663
+    },
+    {
+      "epoch": 4.6892340369042875,
+      "grad_norm": 0.4863879418295176,
+      "learning_rate": 5.699124742870008e-07,
+      "loss": 0.5418,
+      "step": 17664
+    },
+    {
+      "epoch": 4.689499535377672,
+      "grad_norm": 0.4850993576671298,
+      "learning_rate": 5.696905943847966e-07,
+      "loss": 0.5512,
+      "step": 17665
+    },
+    {
+      "epoch": 4.689765033851056,
+      "grad_norm": 0.4745206470230527,
+      "learning_rate": 5.694687521283515e-07,
+      "loss": 0.5532,
+      "step": 17666
+    },
+    {
+      "epoch": 4.690030532324439,
+      "grad_norm": 0.48718978757060233,
+      "learning_rate": 5.692469475219939e-07,
+      "loss": 0.5715,
+      "step": 17667
+    },
+    {
+      "epoch": 4.690296030797823,
+      "grad_norm": 0.4935803637370086,
+      "learning_rate": 5.690251805700467e-07,
+      "loss": 0.5268,
+      "step": 17668
+    },
+    {
+      "epoch": 4.690561529271207,
+      "grad_norm": 0.49716031661918836,
+      "learning_rate": 5.68803451276837e-07,
+      "loss": 0.5605,
+      "step": 17669
+    },
+    {
+      "epoch": 4.69082702774459,
+      "grad_norm": 0.47096071214419916,
+      "learning_rate": 5.685817596466878e-07,
+      "loss": 0.542,
+      "step": 17670
+    },
+    {
+      "epoch": 4.691092526217974,
+      "grad_norm": 0.4782796887933522,
+      "learning_rate": 5.683601056839239e-07,
+      "loss": 0.4998,
+      "step": 17671
+    },
+    {
+      "epoch": 4.6913580246913575,
+      "grad_norm": 0.47083572006107266,
+      "learning_rate": 5.681384893928666e-07,
+      "loss": 0.5197,
+      "step": 17672
+    },
+    {
+      "epoch": 4.691623523164742,
+      "grad_norm": 0.4687143729760963,
+      "learning_rate": 5.679169107778396e-07,
+      "loss": 0.531,
+      "step": 17673
+    },
+    {
+      "epoch": 4.691889021638126,
+      "grad_norm": 0.48754782802963226,
+      "learning_rate": 5.676953698431631e-07,
+      "loss": 0.5154,
+      "step": 17674
+    },
+    {
+      "epoch": 4.69215452011151,
+      "grad_norm": 0.4826093035524664,
+      "learning_rate": 5.674738665931575e-07,
+      "loss": 0.5472,
+      "step": 17675
+    },
+    {
+      "epoch": 4.692420018584893,
+      "grad_norm": 0.4725475949278952,
+      "learning_rate": 5.672524010321439e-07,
+      "loss": 0.5239,
+      "step": 17676
+    },
+    {
+      "epoch": 4.692685517058277,
+      "grad_norm": 0.4802382598068678,
+      "learning_rate": 5.6703097316444e-07,
+      "loss": 0.5338,
+      "step": 17677
+    },
+    {
+      "epoch": 4.69295101553166,
+      "grad_norm": 0.47618119521432156,
+      "learning_rate": 5.668095829943654e-07,
+      "loss": 0.5257,
+      "step": 17678
+    },
+    {
+      "epoch": 4.693216514005044,
+      "grad_norm": 0.4829501546062373,
+      "learning_rate": 5.665882305262371e-07,
+      "loss": 0.5273,
+      "step": 17679
+    },
+    {
+      "epoch": 4.6934820124784284,
+      "grad_norm": 0.497360091971884,
+      "learning_rate": 5.663669157643717e-07,
+      "loss": 0.5423,
+      "step": 17680
+    },
+    {
+      "epoch": 4.693747510951812,
+      "grad_norm": 0.48470220784625606,
+      "learning_rate": 5.661456387130865e-07,
+      "loss": 0.5227,
+      "step": 17681
+    },
+    {
+      "epoch": 4.694013009425196,
+      "grad_norm": 0.48083656551025106,
+      "learning_rate": 5.659243993766953e-07,
+      "loss": 0.5273,
+      "step": 17682
+    },
+    {
+      "epoch": 4.69427850789858,
+      "grad_norm": 0.46991881641280187,
+      "learning_rate": 5.657031977595145e-07,
+      "loss": 0.5206,
+      "step": 17683
+    },
+    {
+      "epoch": 4.694544006371963,
+      "grad_norm": 0.47428118763504223,
+      "learning_rate": 5.654820338658573e-07,
+      "loss": 0.5159,
+      "step": 17684
+    },
+    {
+      "epoch": 4.694809504845347,
+      "grad_norm": 0.48141874363334025,
+      "learning_rate": 5.65260907700037e-07,
+      "loss": 0.529,
+      "step": 17685
+    },
+    {
+      "epoch": 4.695075003318731,
+      "grad_norm": 0.4807693940853336,
+      "learning_rate": 5.650398192663653e-07,
+      "loss": 0.5187,
+      "step": 17686
+    },
+    {
+      "epoch": 4.695340501792114,
+      "grad_norm": 0.47242237498816697,
+      "learning_rate": 5.648187685691556e-07,
+      "loss": 0.5219,
+      "step": 17687
+    },
+    {
+      "epoch": 4.6956060002654985,
+      "grad_norm": 0.4660428159676864,
+      "learning_rate": 5.645977556127171e-07,
+      "loss": 0.5459,
+      "step": 17688
+    },
+    {
+      "epoch": 4.6958714987388825,
+      "grad_norm": 0.47842588385539825,
+      "learning_rate": 5.64376780401362e-07,
+      "loss": 0.5063,
+      "step": 17689
+    },
+    {
+      "epoch": 4.696136997212266,
+      "grad_norm": 0.49167894992186506,
+      "learning_rate": 5.64155842939399e-07,
+      "loss": 0.524,
+      "step": 17690
+    },
+    {
+      "epoch": 4.69640249568565,
+      "grad_norm": 0.4832866071866792,
+      "learning_rate": 5.639349432311358e-07,
+      "loss": 0.5551,
+      "step": 17691
+    },
+    {
+      "epoch": 4.696667994159034,
+      "grad_norm": 0.4719939196201807,
+      "learning_rate": 5.637140812808822e-07,
+      "loss": 0.5306,
+      "step": 17692
+    },
+    {
+      "epoch": 4.696933492632417,
+      "grad_norm": 0.4825039411499423,
+      "learning_rate": 5.634932570929444e-07,
+      "loss": 0.5563,
+      "step": 17693
+    },
+    {
+      "epoch": 4.697198991105801,
+      "grad_norm": 0.46505992805491425,
+      "learning_rate": 5.6327247067163e-07,
+      "loss": 0.5036,
+      "step": 17694
+    },
+    {
+      "epoch": 4.697464489579185,
+      "grad_norm": 0.48955913827148284,
+      "learning_rate": 5.630517220212436e-07,
+      "loss": 0.5732,
+      "step": 17695
+    },
+    {
+      "epoch": 4.6977299880525685,
+      "grad_norm": 0.4790493167250704,
+      "learning_rate": 5.628310111460927e-07,
+      "loss": 0.5124,
+      "step": 17696
+    },
+    {
+      "epoch": 4.697995486525953,
+      "grad_norm": 0.47657144084990877,
+      "learning_rate": 5.626103380504785e-07,
+      "loss": 0.5246,
+      "step": 17697
+    },
+    {
+      "epoch": 4.698260984999337,
+      "grad_norm": 0.47579745159895026,
+      "learning_rate": 5.623897027387066e-07,
+      "loss": 0.5492,
+      "step": 17698
+    },
+    {
+      "epoch": 4.69852648347272,
+      "grad_norm": 0.49439699695857486,
+      "learning_rate": 5.621691052150799e-07,
+      "loss": 0.5499,
+      "step": 17699
+    },
+    {
+      "epoch": 4.698791981946104,
+      "grad_norm": 0.4856148155891063,
+      "learning_rate": 5.619485454838996e-07,
+      "loss": 0.5149,
+      "step": 17700
+    },
+    {
+      "epoch": 4.699057480419487,
+      "grad_norm": 0.4711075584651752,
+      "learning_rate": 5.617280235494696e-07,
+      "loss": 0.5551,
+      "step": 17701
+    },
+    {
+      "epoch": 4.699322978892871,
+      "grad_norm": 0.4888132691610238,
+      "learning_rate": 5.615075394160871e-07,
+      "loss": 0.5263,
+      "step": 17702
+    },
+    {
+      "epoch": 4.699588477366255,
+      "grad_norm": 0.48477841209254213,
+      "learning_rate": 5.612870930880546e-07,
+      "loss": 0.5275,
+      "step": 17703
+    },
+    {
+      "epoch": 4.699853975839639,
+      "grad_norm": 0.500526299657142,
+      "learning_rate": 5.6106668456967e-07,
+      "loss": 0.5413,
+      "step": 17704
+    },
+    {
+      "epoch": 4.700119474313023,
+      "grad_norm": 0.4891374348174846,
+      "learning_rate": 5.608463138652329e-07,
+      "loss": 0.538,
+      "step": 17705
+    },
+    {
+      "epoch": 4.700384972786407,
+      "grad_norm": 0.4928144589744349,
+      "learning_rate": 5.606259809790398e-07,
+      "loss": 0.5328,
+      "step": 17706
+    },
+    {
+      "epoch": 4.70065047125979,
+      "grad_norm": 0.46935866802592463,
+      "learning_rate": 5.604056859153895e-07,
+      "loss": 0.5232,
+      "step": 17707
+    },
+    {
+      "epoch": 4.700915969733174,
+      "grad_norm": 0.4775826680745534,
+      "learning_rate": 5.601854286785769e-07,
+      "loss": 0.5004,
+      "step": 17708
+    },
+    {
+      "epoch": 4.701181468206558,
+      "grad_norm": 0.47012894265697863,
+      "learning_rate": 5.599652092728974e-07,
+      "loss": 0.5242,
+      "step": 17709
+    },
+    {
+      "epoch": 4.701446966679941,
+      "grad_norm": 0.482002968086078,
+      "learning_rate": 5.59745027702647e-07,
+      "loss": 0.5216,
+      "step": 17710
+    },
+    {
+      "epoch": 4.701712465153325,
+      "grad_norm": 0.48505688665332697,
+      "learning_rate": 5.595248839721185e-07,
+      "loss": 0.5591,
+      "step": 17711
+    },
+    {
+      "epoch": 4.701977963626709,
+      "grad_norm": 0.48418410369329307,
+      "learning_rate": 5.593047780856065e-07,
+      "loss": 0.5401,
+      "step": 17712
+    },
+    {
+      "epoch": 4.702243462100093,
+      "grad_norm": 0.4705404933901663,
+      "learning_rate": 5.590847100474031e-07,
+      "loss": 0.5462,
+      "step": 17713
+    },
+    {
+      "epoch": 4.702508960573477,
+      "grad_norm": 0.47372703100535973,
+      "learning_rate": 5.588646798618e-07,
+      "loss": 0.5053,
+      "step": 17714
+    },
+    {
+      "epoch": 4.702774459046861,
+      "grad_norm": 0.5028926597338859,
+      "learning_rate": 5.586446875330875e-07,
+      "loss": 0.5546,
+      "step": 17715
+    },
+    {
+      "epoch": 4.703039957520244,
+      "grad_norm": 0.4951186410103098,
+      "learning_rate": 5.584247330655577e-07,
+      "loss": 0.5375,
+      "step": 17716
+    },
+    {
+      "epoch": 4.703305455993628,
+      "grad_norm": 0.4742472055461465,
+      "learning_rate": 5.582048164634984e-07,
+      "loss": 0.582,
+      "step": 17717
+    },
+    {
+      "epoch": 4.703570954467012,
+      "grad_norm": 0.4692746048232161,
+      "learning_rate": 5.579849377312005e-07,
+      "loss": 0.5112,
+      "step": 17718
+    },
+    {
+      "epoch": 4.703836452940395,
+      "grad_norm": 0.48377818946665885,
+      "learning_rate": 5.577650968729512e-07,
+      "loss": 0.5491,
+      "step": 17719
+    },
+    {
+      "epoch": 4.704101951413779,
+      "grad_norm": 0.4886841292710266,
+      "learning_rate": 5.575452938930373e-07,
+      "loss": 0.5322,
+      "step": 17720
+    },
+    {
+      "epoch": 4.7043674498871635,
+      "grad_norm": 0.4736128380926772,
+      "learning_rate": 5.573255287957466e-07,
+      "loss": 0.5256,
+      "step": 17721
+    },
+    {
+      "epoch": 4.704632948360547,
+      "grad_norm": 0.5512949986442008,
+      "learning_rate": 5.571058015853639e-07,
+      "loss": 0.548,
+      "step": 17722
+    },
+    {
+      "epoch": 4.704898446833931,
+      "grad_norm": 0.48024427215948706,
+      "learning_rate": 5.568861122661759e-07,
+      "loss": 0.5531,
+      "step": 17723
+    },
+    {
+      "epoch": 4.705163945307315,
+      "grad_norm": 0.4790075735255522,
+      "learning_rate": 5.566664608424663e-07,
+      "loss": 0.5321,
+      "step": 17724
+    },
+    {
+      "epoch": 4.705429443780698,
+      "grad_norm": 0.4773187956376124,
+      "learning_rate": 5.56446847318518e-07,
+      "loss": 0.5383,
+      "step": 17725
+    },
+    {
+      "epoch": 4.705694942254082,
+      "grad_norm": 0.47673629409431423,
+      "learning_rate": 5.562272716986155e-07,
+      "loss": 0.5098,
+      "step": 17726
+    },
+    {
+      "epoch": 4.705960440727466,
+      "grad_norm": 0.4690923263898515,
+      "learning_rate": 5.560077339870398e-07,
+      "loss": 0.5068,
+      "step": 17727
+    },
+    {
+      "epoch": 4.706225939200849,
+      "grad_norm": 0.4777111756311077,
+      "learning_rate": 5.557882341880735e-07,
+      "loss": 0.5127,
+      "step": 17728
+    },
+    {
+      "epoch": 4.7064914376742335,
+      "grad_norm": 0.4769588377876074,
+      "learning_rate": 5.555687723059963e-07,
+      "loss": 0.5436,
+      "step": 17729
+    },
+    {
+      "epoch": 4.706756936147617,
+      "grad_norm": 0.4838361199272956,
+      "learning_rate": 5.553493483450903e-07,
+      "loss": 0.5429,
+      "step": 17730
+    },
+    {
+      "epoch": 4.707022434621001,
+      "grad_norm": 0.48380839188958163,
+      "learning_rate": 5.551299623096318e-07,
+      "loss": 0.5523,
+      "step": 17731
+    },
+    {
+      "epoch": 4.707287933094385,
+      "grad_norm": 0.4704060853857825,
+      "learning_rate": 5.549106142039018e-07,
+      "loss": 0.5343,
+      "step": 17732
+    },
+    {
+      "epoch": 4.707553431567769,
+      "grad_norm": 0.49027113289736296,
+      "learning_rate": 5.546913040321763e-07,
+      "loss": 0.4773,
+      "step": 17733
+    },
+    {
+      "epoch": 4.707818930041152,
+      "grad_norm": 0.4808552880080877,
+      "learning_rate": 5.544720317987343e-07,
+      "loss": 0.5647,
+      "step": 17734
+    },
+    {
+      "epoch": 4.708084428514536,
+      "grad_norm": 0.47396429879709573,
+      "learning_rate": 5.542527975078502e-07,
+      "loss": 0.5153,
+      "step": 17735
+    },
+    {
+      "epoch": 4.708349926987919,
+      "grad_norm": 0.46603318972425906,
+      "learning_rate": 5.540336011638012e-07,
+      "loss": 0.505,
+      "step": 17736
+    },
+    {
+      "epoch": 4.7086154254613035,
+      "grad_norm": 0.48374807699342665,
+      "learning_rate": 5.538144427708617e-07,
+      "loss": 0.5579,
+      "step": 17737
+    },
+    {
+      "epoch": 4.708880923934688,
+      "grad_norm": 0.4839115586845375,
+      "learning_rate": 5.535953223333049e-07,
+      "loss": 0.523,
+      "step": 17738
+    },
+    {
+      "epoch": 4.709146422408071,
+      "grad_norm": 0.4651048068699564,
+      "learning_rate": 5.533762398554057e-07,
+      "loss": 0.5244,
+      "step": 17739
+    },
+    {
+      "epoch": 4.709411920881455,
+      "grad_norm": 0.4781862500942812,
+      "learning_rate": 5.531571953414353e-07,
+      "loss": 0.5055,
+      "step": 17740
+    },
+    {
+      "epoch": 4.709677419354839,
+      "grad_norm": 0.4818592551247389,
+      "learning_rate": 5.529381887956667e-07,
+      "loss": 0.5417,
+      "step": 17741
+    },
+    {
+      "epoch": 4.709942917828222,
+      "grad_norm": 0.47474156603044654,
+      "learning_rate": 5.527192202223708e-07,
+      "loss": 0.5122,
+      "step": 17742
+    },
+    {
+      "epoch": 4.710208416301606,
+      "grad_norm": 0.46940917969442003,
+      "learning_rate": 5.52500289625818e-07,
+      "loss": 0.5162,
+      "step": 17743
+    },
+    {
+      "epoch": 4.71047391477499,
+      "grad_norm": 0.4720338835065068,
+      "learning_rate": 5.522813970102769e-07,
+      "loss": 0.506,
+      "step": 17744
+    },
+    {
+      "epoch": 4.7107394132483735,
+      "grad_norm": 0.48044926928744863,
+      "learning_rate": 5.520625423800177e-07,
+      "loss": 0.5672,
+      "step": 17745
+    },
+    {
+      "epoch": 4.711004911721758,
+      "grad_norm": 0.48480859747769245,
+      "learning_rate": 5.518437257393088e-07,
+      "loss": 0.5384,
+      "step": 17746
+    },
+    {
+      "epoch": 4.711270410195142,
+      "grad_norm": 0.4765361350458473,
+      "learning_rate": 5.516249470924165e-07,
+      "loss": 0.5426,
+      "step": 17747
+    },
+    {
+      "epoch": 4.711535908668525,
+      "grad_norm": 0.48704799483949296,
+      "learning_rate": 5.514062064436096e-07,
+      "loss": 0.5326,
+      "step": 17748
+    },
+    {
+      "epoch": 4.711801407141909,
+      "grad_norm": 0.4713546300070442,
+      "learning_rate": 5.511875037971512e-07,
+      "loss": 0.5312,
+      "step": 17749
+    },
+    {
+      "epoch": 4.712066905615293,
+      "grad_norm": 0.4696325781583951,
+      "learning_rate": 5.509688391573087e-07,
+      "loss": 0.5604,
+      "step": 17750
+    },
+    {
+      "epoch": 4.712332404088676,
+      "grad_norm": 0.4797799861549737,
+      "learning_rate": 5.507502125283454e-07,
+      "loss": 0.5668,
+      "step": 17751
+    },
+    {
+      "epoch": 4.71259790256206,
+      "grad_norm": 0.48349585165362075,
+      "learning_rate": 5.505316239145261e-07,
+      "loss": 0.5257,
+      "step": 17752
+    },
+    {
+      "epoch": 4.712863401035444,
+      "grad_norm": 0.4810519624117421,
+      "learning_rate": 5.503130733201132e-07,
+      "loss": 0.5166,
+      "step": 17753
+    },
+    {
+      "epoch": 4.713128899508828,
+      "grad_norm": 0.4785625845715155,
+      "learning_rate": 5.500945607493682e-07,
+      "loss": 0.5432,
+      "step": 17754
+    },
+    {
+      "epoch": 4.713394397982212,
+      "grad_norm": 0.49089174685542164,
+      "learning_rate": 5.498760862065544e-07,
+      "loss": 0.4986,
+      "step": 17755
+    },
+    {
+      "epoch": 4.713659896455596,
+      "grad_norm": 0.4976340495275194,
+      "learning_rate": 5.496576496959307e-07,
+      "loss": 0.5134,
+      "step": 17756
+    },
+    {
+      "epoch": 4.713925394928979,
+      "grad_norm": 0.4749378062246209,
+      "learning_rate": 5.494392512217592e-07,
+      "loss": 0.5364,
+      "step": 17757
+    },
+    {
+      "epoch": 4.714190893402363,
+      "grad_norm": 0.48884412787440684,
+      "learning_rate": 5.492208907882971e-07,
+      "loss": 0.5663,
+      "step": 17758
+    },
+    {
+      "epoch": 4.714456391875746,
+      "grad_norm": 0.4829621297573895,
+      "learning_rate": 5.490025683998054e-07,
+      "loss": 0.5563,
+      "step": 17759
+    },
+    {
+      "epoch": 4.71472189034913,
+      "grad_norm": 0.47666660247528025,
+      "learning_rate": 5.487842840605389e-07,
+      "loss": 0.5531,
+      "step": 17760
+    },
+    {
+      "epoch": 4.7149873888225144,
+      "grad_norm": 0.4831572082266196,
+      "learning_rate": 5.485660377747573e-07,
+      "loss": 0.5508,
+      "step": 17761
+    },
+    {
+      "epoch": 4.7152528872958985,
+      "grad_norm": 0.493285742808171,
+      "learning_rate": 5.483478295467152e-07,
+      "loss": 0.5773,
+      "step": 17762
+    },
+    {
+      "epoch": 4.715518385769282,
+      "grad_norm": 0.46995877336638525,
+      "learning_rate": 5.481296593806692e-07,
+      "loss": 0.5004,
+      "step": 17763
+    },
+    {
+      "epoch": 4.715783884242666,
+      "grad_norm": 0.47369676905795444,
+      "learning_rate": 5.479115272808736e-07,
+      "loss": 0.5135,
+      "step": 17764
+    },
+    {
+      "epoch": 4.716049382716049,
+      "grad_norm": 0.4766463206880944,
+      "learning_rate": 5.476934332515831e-07,
+      "loss": 0.523,
+      "step": 17765
+    },
+    {
+      "epoch": 4.716314881189433,
+      "grad_norm": 0.479706849844324,
+      "learning_rate": 5.47475377297051e-07,
+      "loss": 0.5604,
+      "step": 17766
+    },
+    {
+      "epoch": 4.716580379662817,
+      "grad_norm": 0.4685820976551383,
+      "learning_rate": 5.472573594215285e-07,
+      "loss": 0.5251,
+      "step": 17767
+    },
+    {
+      "epoch": 4.7168458781362,
+      "grad_norm": 0.4857829123935467,
+      "learning_rate": 5.470393796292697e-07,
+      "loss": 0.5744,
+      "step": 17768
+    },
+    {
+      "epoch": 4.7171113766095845,
+      "grad_norm": 0.477079145233189,
+      "learning_rate": 5.46821437924524e-07,
+      "loss": 0.5172,
+      "step": 17769
+    },
+    {
+      "epoch": 4.7173768750829685,
+      "grad_norm": 0.47500316034050705,
+      "learning_rate": 5.46603534311543e-07,
+      "loss": 0.5361,
+      "step": 17770
+    },
+    {
+      "epoch": 4.717642373556352,
+      "grad_norm": 0.47286348970400893,
+      "learning_rate": 5.46385668794576e-07,
+      "loss": 0.5388,
+      "step": 17771
+    },
+    {
+      "epoch": 4.717907872029736,
+      "grad_norm": 0.4923784286365227,
+      "learning_rate": 5.461678413778709e-07,
+      "loss": 0.5214,
+      "step": 17772
+    },
+    {
+      "epoch": 4.71817337050312,
+      "grad_norm": 0.4665442428930693,
+      "learning_rate": 5.459500520656774e-07,
+      "loss": 0.5391,
+      "step": 17773
+    },
+    {
+      "epoch": 4.718438868976503,
+      "grad_norm": 0.47962067629857,
+      "learning_rate": 5.457323008622414e-07,
+      "loss": 0.5351,
+      "step": 17774
+    },
+    {
+      "epoch": 4.718704367449887,
+      "grad_norm": 0.45918478839971744,
+      "learning_rate": 5.455145877718112e-07,
+      "loss": 0.4914,
+      "step": 17775
+    },
+    {
+      "epoch": 4.718969865923271,
+      "grad_norm": 0.48571893380996733,
+      "learning_rate": 5.452969127986321e-07,
+      "loss": 0.4886,
+      "step": 17776
+    },
+    {
+      "epoch": 4.7192353643966545,
+      "grad_norm": 0.486727468431806,
+      "learning_rate": 5.450792759469492e-07,
+      "loss": 0.5146,
+      "step": 17777
+    },
+    {
+      "epoch": 4.719500862870039,
+      "grad_norm": 0.47601929007289173,
+      "learning_rate": 5.448616772210058e-07,
+      "loss": 0.5303,
+      "step": 17778
+    },
+    {
+      "epoch": 4.719766361343423,
+      "grad_norm": 0.48436796609739574,
+      "learning_rate": 5.446441166250477e-07,
+      "loss": 0.5056,
+      "step": 17779
+    },
+    {
+      "epoch": 4.720031859816806,
+      "grad_norm": 0.48107112526008555,
+      "learning_rate": 5.444265941633161e-07,
+      "loss": 0.5272,
+      "step": 17780
+    },
+    {
+      "epoch": 4.72029735829019,
+      "grad_norm": 0.4843916083388342,
+      "learning_rate": 5.442091098400548e-07,
+      "loss": 0.5341,
+      "step": 17781
+    },
+    {
+      "epoch": 4.720562856763573,
+      "grad_norm": 0.47748784773561725,
+      "learning_rate": 5.439916636595041e-07,
+      "loss": 0.5292,
+      "step": 17782
+    },
+    {
+      "epoch": 4.720828355236957,
+      "grad_norm": 0.4814756067568349,
+      "learning_rate": 5.437742556259048e-07,
+      "loss": 0.5404,
+      "step": 17783
+    },
+    {
+      "epoch": 4.721093853710341,
+      "grad_norm": 0.476424819551458,
+      "learning_rate": 5.435568857434975e-07,
+      "loss": 0.5592,
+      "step": 17784
+    },
+    {
+      "epoch": 4.721359352183725,
+      "grad_norm": 0.4690113284547608,
+      "learning_rate": 5.433395540165207e-07,
+      "loss": 0.5043,
+      "step": 17785
+    },
+    {
+      "epoch": 4.721624850657109,
+      "grad_norm": 0.4968395550800811,
+      "learning_rate": 5.43122260449214e-07,
+      "loss": 0.5509,
+      "step": 17786
+    },
+    {
+      "epoch": 4.721890349130493,
+      "grad_norm": 0.47507585521685436,
+      "learning_rate": 5.429050050458138e-07,
+      "loss": 0.5479,
+      "step": 17787
+    },
+    {
+      "epoch": 4.722155847603876,
+      "grad_norm": 0.4878486108722755,
+      "learning_rate": 5.42687787810559e-07,
+      "loss": 0.5391,
+      "step": 17788
+    },
+    {
+      "epoch": 4.72242134607726,
+      "grad_norm": 0.48011797870712347,
+      "learning_rate": 5.424706087476836e-07,
+      "loss": 0.56,
+      "step": 17789
+    },
+    {
+      "epoch": 4.722686844550644,
+      "grad_norm": 0.47253407556489935,
+      "learning_rate": 5.422534678614239e-07,
+      "loss": 0.5375,
+      "step": 17790
+    },
+    {
+      "epoch": 4.722952343024028,
+      "grad_norm": 0.4789269619830572,
+      "learning_rate": 5.420363651560157e-07,
+      "loss": 0.5411,
+      "step": 17791
+    },
+    {
+      "epoch": 4.723217841497411,
+      "grad_norm": 0.49035580299854287,
+      "learning_rate": 5.418193006356917e-07,
+      "loss": 0.528,
+      "step": 17792
+    },
+    {
+      "epoch": 4.723483339970795,
+      "grad_norm": 0.47868331057485014,
+      "learning_rate": 5.416022743046865e-07,
+      "loss": 0.5016,
+      "step": 17793
+    },
+    {
+      "epoch": 4.723748838444179,
+      "grad_norm": 0.4936565249839819,
+      "learning_rate": 5.413852861672317e-07,
+      "loss": 0.5465,
+      "step": 17794
+    },
+    {
+      "epoch": 4.724014336917563,
+      "grad_norm": 0.4857426582701661,
+      "learning_rate": 5.411683362275597e-07,
+      "loss": 0.5569,
+      "step": 17795
+    },
+    {
+      "epoch": 4.724279835390947,
+      "grad_norm": 0.4938315807209008,
+      "learning_rate": 5.409514244899003e-07,
+      "loss": 0.5191,
+      "step": 17796
+    },
+    {
+      "epoch": 4.72454533386433,
+      "grad_norm": 0.48188217638278147,
+      "learning_rate": 5.407345509584852e-07,
+      "loss": 0.5523,
+      "step": 17797
+    },
+    {
+      "epoch": 4.724810832337714,
+      "grad_norm": 0.46958625126377557,
+      "learning_rate": 5.405177156375432e-07,
+      "loss": 0.5203,
+      "step": 17798
+    },
+    {
+      "epoch": 4.725076330811098,
+      "grad_norm": 0.46543687305075016,
+      "learning_rate": 5.403009185313038e-07,
+      "loss": 0.5705,
+      "step": 17799
+    },
+    {
+      "epoch": 4.725341829284481,
+      "grad_norm": 0.4716544000804069,
+      "learning_rate": 5.400841596439945e-07,
+      "loss": 0.559,
+      "step": 17800
+    },
+    {
+      "epoch": 4.725607327757865,
+      "grad_norm": 0.4845506043626805,
+      "learning_rate": 5.398674389798425e-07,
+      "loss": 0.5217,
+      "step": 17801
+    },
+    {
+      "epoch": 4.7258728262312495,
+      "grad_norm": 0.4928690571091297,
+      "learning_rate": 5.39650756543075e-07,
+      "loss": 0.539,
+      "step": 17802
+    },
+    {
+      "epoch": 4.726138324704633,
+      "grad_norm": 0.49063428928526887,
+      "learning_rate": 5.394341123379171e-07,
+      "loss": 0.5832,
+      "step": 17803
+    },
+    {
+      "epoch": 4.726403823178017,
+      "grad_norm": 0.4725043437707859,
+      "learning_rate": 5.39217506368595e-07,
+      "loss": 0.5202,
+      "step": 17804
+    },
+    {
+      "epoch": 4.726669321651401,
+      "grad_norm": 0.493025484427121,
+      "learning_rate": 5.390009386393322e-07,
+      "loss": 0.5703,
+      "step": 17805
+    },
+    {
+      "epoch": 4.726934820124784,
+      "grad_norm": 0.47783903156507096,
+      "learning_rate": 5.387844091543526e-07,
+      "loss": 0.5286,
+      "step": 17806
+    },
+    {
+      "epoch": 4.727200318598168,
+      "grad_norm": 0.4769408680306065,
+      "learning_rate": 5.385679179178779e-07,
+      "loss": 0.5522,
+      "step": 17807
+    },
+    {
+      "epoch": 4.727465817071552,
+      "grad_norm": 0.48960243540671167,
+      "learning_rate": 5.383514649341321e-07,
+      "loss": 0.5426,
+      "step": 17808
+    },
+    {
+      "epoch": 4.727731315544935,
+      "grad_norm": 0.491492774226586,
+      "learning_rate": 5.381350502073351e-07,
+      "loss": 0.5855,
+      "step": 17809
+    },
+    {
+      "epoch": 4.7279968140183195,
+      "grad_norm": 0.48396600345049673,
+      "learning_rate": 5.37918673741708e-07,
+      "loss": 0.5389,
+      "step": 17810
+    },
+    {
+      "epoch": 4.728262312491703,
+      "grad_norm": 0.4748045389359221,
+      "learning_rate": 5.37702335541472e-07,
+      "loss": 0.5152,
+      "step": 17811
+    },
+    {
+      "epoch": 4.728527810965087,
+      "grad_norm": 0.517192239055643,
+      "learning_rate": 5.374860356108438e-07,
+      "loss": 0.5714,
+      "step": 17812
+    },
+    {
+      "epoch": 4.728793309438471,
+      "grad_norm": 0.4709989980691845,
+      "learning_rate": 5.372697739540437e-07,
+      "loss": 0.507,
+      "step": 17813
+    },
+    {
+      "epoch": 4.729058807911855,
+      "grad_norm": 0.49378919968521184,
+      "learning_rate": 5.370535505752877e-07,
+      "loss": 0.5643,
+      "step": 17814
+    },
+    {
+      "epoch": 4.729324306385238,
+      "grad_norm": 0.4988201036825255,
+      "learning_rate": 5.368373654787945e-07,
+      "loss": 0.5399,
+      "step": 17815
+    },
+    {
+      "epoch": 4.729589804858622,
+      "grad_norm": 0.4900107737672029,
+      "learning_rate": 5.366212186687783e-07,
+      "loss": 0.5465,
+      "step": 17816
+    },
+    {
+      "epoch": 4.729855303332005,
+      "grad_norm": 0.4832615405479388,
+      "learning_rate": 5.364051101494566e-07,
+      "loss": 0.5223,
+      "step": 17817
+    },
+    {
+      "epoch": 4.7301208018053895,
+      "grad_norm": 0.49560682784076504,
+      "learning_rate": 5.361890399250428e-07,
+      "loss": 0.499,
+      "step": 17818
+    },
+    {
+      "epoch": 4.730386300278774,
+      "grad_norm": 0.48039025017433135,
+      "learning_rate": 5.359730079997502e-07,
+      "loss": 0.4923,
+      "step": 17819
+    },
+    {
+      "epoch": 4.730651798752157,
+      "grad_norm": 0.49508421605896696,
+      "learning_rate": 5.357570143777935e-07,
+      "loss": 0.5417,
+      "step": 17820
+    },
+    {
+      "epoch": 4.730917297225541,
+      "grad_norm": 0.48283838790412154,
+      "learning_rate": 5.355410590633837e-07,
+      "loss": 0.5635,
+      "step": 17821
+    },
+    {
+      "epoch": 4.731182795698925,
+      "grad_norm": 0.48369023724818666,
+      "learning_rate": 5.353251420607342e-07,
+      "loss": 0.555,
+      "step": 17822
+    },
+    {
+      "epoch": 4.731448294172308,
+      "grad_norm": 0.47993500126706407,
+      "learning_rate": 5.351092633740535e-07,
+      "loss": 0.4962,
+      "step": 17823
+    },
+    {
+      "epoch": 4.731713792645692,
+      "grad_norm": 0.4876845465463659,
+      "learning_rate": 5.34893423007554e-07,
+      "loss": 0.6008,
+      "step": 17824
+    },
+    {
+      "epoch": 4.731979291119076,
+      "grad_norm": 0.4995686420789376,
+      "learning_rate": 5.34677620965443e-07,
+      "loss": 0.5504,
+      "step": 17825
+    },
+    {
+      "epoch": 4.7322447895924595,
+      "grad_norm": 0.47472462982823826,
+      "learning_rate": 5.344618572519314e-07,
+      "loss": 0.4944,
+      "step": 17826
+    },
+    {
+      "epoch": 4.732510288065844,
+      "grad_norm": 0.47977697904143607,
+      "learning_rate": 5.342461318712252e-07,
+      "loss": 0.5267,
+      "step": 17827
+    },
+    {
+      "epoch": 4.732775786539228,
+      "grad_norm": 0.4891405012587452,
+      "learning_rate": 5.34030444827533e-07,
+      "loss": 0.5325,
+      "step": 17828
+    },
+    {
+      "epoch": 4.733041285012611,
+      "grad_norm": 0.48416261435235364,
+      "learning_rate": 5.338147961250609e-07,
+      "loss": 0.5726,
+      "step": 17829
+    },
+    {
+      "epoch": 4.733306783485995,
+      "grad_norm": 0.4822392410141068,
+      "learning_rate": 5.335991857680134e-07,
+      "loss": 0.5511,
+      "step": 17830
+    },
+    {
+      "epoch": 4.733572281959379,
+      "grad_norm": 0.49142617049579085,
+      "learning_rate": 5.333836137605972e-07,
+      "loss": 0.5821,
+      "step": 17831
+    },
+    {
+      "epoch": 4.733837780432762,
+      "grad_norm": 0.49579224423848245,
+      "learning_rate": 5.33168080107015e-07,
+      "loss": 0.5409,
+      "step": 17832
+    },
+    {
+      "epoch": 4.734103278906146,
+      "grad_norm": 0.4784210969400949,
+      "learning_rate": 5.329525848114714e-07,
+      "loss": 0.5127,
+      "step": 17833
+    },
+    {
+      "epoch": 4.73436877737953,
+      "grad_norm": 0.4821335956651352,
+      "learning_rate": 5.327371278781687e-07,
+      "loss": 0.5141,
+      "step": 17834
+    },
+    {
+      "epoch": 4.734634275852914,
+      "grad_norm": 0.6219567450843645,
+      "learning_rate": 5.325217093113081e-07,
+      "loss": 0.5731,
+      "step": 17835
+    },
+    {
+      "epoch": 4.734899774326298,
+      "grad_norm": 0.48768407851887385,
+      "learning_rate": 5.323063291150918e-07,
+      "loss": 0.5426,
+      "step": 17836
+    },
+    {
+      "epoch": 4.735165272799682,
+      "grad_norm": 0.48931731965500336,
+      "learning_rate": 5.320909872937193e-07,
+      "loss": 0.5429,
+      "step": 17837
+    },
+    {
+      "epoch": 4.735430771273065,
+      "grad_norm": 0.47214192820631357,
+      "learning_rate": 5.318756838513914e-07,
+      "loss": 0.521,
+      "step": 17838
+    },
+    {
+      "epoch": 4.735696269746449,
+      "grad_norm": 0.4755773624768587,
+      "learning_rate": 5.316604187923061e-07,
+      "loss": 0.5318,
+      "step": 17839
+    },
+    {
+      "epoch": 4.735961768219832,
+      "grad_norm": 0.4871698385010478,
+      "learning_rate": 5.314451921206628e-07,
+      "loss": 0.5373,
+      "step": 17840
+    },
+    {
+      "epoch": 4.736227266693216,
+      "grad_norm": 0.48149180897571026,
+      "learning_rate": 5.31230003840657e-07,
+      "loss": 0.5333,
+      "step": 17841
+    },
+    {
+      "epoch": 4.7364927651666004,
+      "grad_norm": 0.4946073261261152,
+      "learning_rate": 5.310148539564875e-07,
+      "loss": 0.5285,
+      "step": 17842
+    },
+    {
+      "epoch": 4.7367582636399845,
+      "grad_norm": 0.4821929696750198,
+      "learning_rate": 5.307997424723482e-07,
+      "loss": 0.5535,
+      "step": 17843
+    },
+    {
+      "epoch": 4.737023762113368,
+      "grad_norm": 0.46948646453571613,
+      "learning_rate": 5.30584669392436e-07,
+      "loss": 0.516,
+      "step": 17844
+    },
+    {
+      "epoch": 4.737289260586752,
+      "grad_norm": 0.47290285940746773,
+      "learning_rate": 5.30369634720945e-07,
+      "loss": 0.5437,
+      "step": 17845
+    },
+    {
+      "epoch": 4.737554759060135,
+      "grad_norm": 0.49611573199031583,
+      "learning_rate": 5.301546384620676e-07,
+      "loss": 0.5525,
+      "step": 17846
+    },
+    {
+      "epoch": 4.737820257533519,
+      "grad_norm": 0.4781772101680418,
+      "learning_rate": 5.299396806199985e-07,
+      "loss": 0.5749,
+      "step": 17847
+    },
+    {
+      "epoch": 4.738085756006903,
+      "grad_norm": 0.4785786865365697,
+      "learning_rate": 5.297247611989287e-07,
+      "loss": 0.4968,
+      "step": 17848
+    },
+    {
+      "epoch": 4.738351254480286,
+      "grad_norm": 0.4839685349661115,
+      "learning_rate": 5.295098802030508e-07,
+      "loss": 0.528,
+      "step": 17849
+    },
+    {
+      "epoch": 4.7386167529536705,
+      "grad_norm": 0.4992330422642702,
+      "learning_rate": 5.292950376365538e-07,
+      "loss": 0.5844,
+      "step": 17850
+    },
+    {
+      "epoch": 4.7388822514270545,
+      "grad_norm": 0.47452235010199206,
+      "learning_rate": 5.290802335036305e-07,
+      "loss": 0.5581,
+      "step": 17851
+    },
+    {
+      "epoch": 4.739147749900438,
+      "grad_norm": 0.5012832851396919,
+      "learning_rate": 5.288654678084667e-07,
+      "loss": 0.5165,
+      "step": 17852
+    },
+    {
+      "epoch": 4.739413248373822,
+      "grad_norm": 0.4900019888244732,
+      "learning_rate": 5.286507405552529e-07,
+      "loss": 0.5535,
+      "step": 17853
+    },
+    {
+      "epoch": 4.739678746847206,
+      "grad_norm": 0.48990715308594884,
+      "learning_rate": 5.28436051748176e-07,
+      "loss": 0.5499,
+      "step": 17854
+    },
+    {
+      "epoch": 4.739944245320589,
+      "grad_norm": 0.47780676195317995,
+      "learning_rate": 5.282214013914239e-07,
+      "loss": 0.5378,
+      "step": 17855
+    },
+    {
+      "epoch": 4.740209743793973,
+      "grad_norm": 0.48041583958724254,
+      "learning_rate": 5.280067894891816e-07,
+      "loss": 0.577,
+      "step": 17856
+    },
+    {
+      "epoch": 4.740475242267357,
+      "grad_norm": 0.48920429200985316,
+      "learning_rate": 5.277922160456359e-07,
+      "loss": 0.5383,
+      "step": 17857
+    },
+    {
+      "epoch": 4.7407407407407405,
+      "grad_norm": 0.5036382516933882,
+      "learning_rate": 5.275776810649705e-07,
+      "loss": 0.5378,
+      "step": 17858
+    },
+    {
+      "epoch": 4.741006239214125,
+      "grad_norm": 0.4817731354875225,
+      "learning_rate": 5.273631845513694e-07,
+      "loss": 0.5443,
+      "step": 17859
+    },
+    {
+      "epoch": 4.741271737687509,
+      "grad_norm": 0.4855748358380614,
+      "learning_rate": 5.271487265090163e-07,
+      "loss": 0.5129,
+      "step": 17860
+    },
+    {
+      "epoch": 4.741537236160892,
+      "grad_norm": 0.4873658075620909,
+      "learning_rate": 5.269343069420929e-07,
+      "loss": 0.522,
+      "step": 17861
+    },
+    {
+      "epoch": 4.741802734634276,
+      "grad_norm": 0.49449433311161883,
+      "learning_rate": 5.267199258547823e-07,
+      "loss": 0.5392,
+      "step": 17862
+    },
+    {
+      "epoch": 4.74206823310766,
+      "grad_norm": 0.4897701980181507,
+      "learning_rate": 5.265055832512642e-07,
+      "loss": 0.532,
+      "step": 17863
+    },
+    {
+      "epoch": 4.742333731581043,
+      "grad_norm": 0.48009249532659065,
+      "learning_rate": 5.262912791357186e-07,
+      "loss": 0.5152,
+      "step": 17864
+    },
+    {
+      "epoch": 4.742599230054427,
+      "grad_norm": 0.49299664221327266,
+      "learning_rate": 5.260770135123264e-07,
+      "loss": 0.5607,
+      "step": 17865
+    },
+    {
+      "epoch": 4.742864728527811,
+      "grad_norm": 0.48056243567319673,
+      "learning_rate": 5.258627863852645e-07,
+      "loss": 0.5243,
+      "step": 17866
+    },
+    {
+      "epoch": 4.743130227001195,
+      "grad_norm": 0.48029086009398775,
+      "learning_rate": 5.256485977587125e-07,
+      "loss": 0.5286,
+      "step": 17867
+    },
+    {
+      "epoch": 4.743395725474579,
+      "grad_norm": 0.48929674048932,
+      "learning_rate": 5.254344476368464e-07,
+      "loss": 0.5217,
+      "step": 17868
+    },
+    {
+      "epoch": 4.743661223947962,
+      "grad_norm": 0.4858147899327403,
+      "learning_rate": 5.252203360238444e-07,
+      "loss": 0.5682,
+      "step": 17869
+    },
+    {
+      "epoch": 4.743926722421346,
+      "grad_norm": 0.48458490904277157,
+      "learning_rate": 5.250062629238795e-07,
+      "loss": 0.509,
+      "step": 17870
+    },
+    {
+      "epoch": 4.74419222089473,
+      "grad_norm": 0.4731475084205726,
+      "learning_rate": 5.247922283411289e-07,
+      "loss": 0.5156,
+      "step": 17871
+    },
+    {
+      "epoch": 4.744457719368114,
+      "grad_norm": 0.4787421572516064,
+      "learning_rate": 5.245782322797652e-07,
+      "loss": 0.5484,
+      "step": 17872
+    },
+    {
+      "epoch": 4.744723217841497,
+      "grad_norm": 0.4874932232981817,
+      "learning_rate": 5.243642747439634e-07,
+      "loss": 0.541,
+      "step": 17873
+    },
+    {
+      "epoch": 4.744988716314881,
+      "grad_norm": 0.48364104917233885,
+      "learning_rate": 5.241503557378955e-07,
+      "loss": 0.5073,
+      "step": 17874
+    },
+    {
+      "epoch": 4.745254214788265,
+      "grad_norm": 0.4658280581470388,
+      "learning_rate": 5.239364752657325e-07,
+      "loss": 0.5369,
+      "step": 17875
+    },
+    {
+      "epoch": 4.745519713261649,
+      "grad_norm": 0.48171165780406,
+      "learning_rate": 5.237226333316473e-07,
+      "loss": 0.5575,
+      "step": 17876
+    },
+    {
+      "epoch": 4.745785211735033,
+      "grad_norm": 0.48208409461387125,
+      "learning_rate": 5.235088299398089e-07,
+      "loss": 0.5318,
+      "step": 17877
+    },
+    {
+      "epoch": 4.746050710208416,
+      "grad_norm": 0.48627812758945066,
+      "learning_rate": 5.23295065094388e-07,
+      "loss": 0.5159,
+      "step": 17878
+    },
+    {
+      "epoch": 4.7463162086818,
+      "grad_norm": 0.47543509111298726,
+      "learning_rate": 5.230813387995526e-07,
+      "loss": 0.5214,
+      "step": 17879
+    },
+    {
+      "epoch": 4.746581707155184,
+      "grad_norm": 0.4900454127980515,
+      "learning_rate": 5.228676510594729e-07,
+      "loss": 0.5547,
+      "step": 17880
+    },
+    {
+      "epoch": 4.746847205628567,
+      "grad_norm": 0.4872030856030643,
+      "learning_rate": 5.226540018783135e-07,
+      "loss": 0.4989,
+      "step": 17881
+    },
+    {
+      "epoch": 4.747112704101951,
+      "grad_norm": 0.48074368310599463,
+      "learning_rate": 5.224403912602421e-07,
+      "loss": 0.5458,
+      "step": 17882
+    },
+    {
+      "epoch": 4.7473782025753355,
+      "grad_norm": 0.47451362721899926,
+      "learning_rate": 5.22226819209426e-07,
+      "loss": 0.5228,
+      "step": 17883
+    },
+    {
+      "epoch": 4.747643701048719,
+      "grad_norm": 0.47336968920937994,
+      "learning_rate": 5.220132857300286e-07,
+      "loss": 0.5181,
+      "step": 17884
+    },
+    {
+      "epoch": 4.747909199522103,
+      "grad_norm": 0.488278870143053,
+      "learning_rate": 5.217997908262156e-07,
+      "loss": 0.5774,
+      "step": 17885
+    },
+    {
+      "epoch": 4.748174697995487,
+      "grad_norm": 0.48861088360431393,
+      "learning_rate": 5.215863345021504e-07,
+      "loss": 0.5394,
+      "step": 17886
+    },
+    {
+      "epoch": 4.74844019646887,
+      "grad_norm": 0.48481392186195105,
+      "learning_rate": 5.213729167619955e-07,
+      "loss": 0.5623,
+      "step": 17887
+    },
+    {
+      "epoch": 4.748705694942254,
+      "grad_norm": 0.48985043599973693,
+      "learning_rate": 5.211595376099127e-07,
+      "loss": 0.5523,
+      "step": 17888
+    },
+    {
+      "epoch": 4.748971193415638,
+      "grad_norm": 0.4855192542574241,
+      "learning_rate": 5.209461970500646e-07,
+      "loss": 0.532,
+      "step": 17889
+    },
+    {
+      "epoch": 4.749236691889021,
+      "grad_norm": 0.48531222067919866,
+      "learning_rate": 5.207328950866103e-07,
+      "loss": 0.5532,
+      "step": 17890
+    },
+    {
+      "epoch": 4.7495021903624055,
+      "grad_norm": 0.48986217041125296,
+      "learning_rate": 5.205196317237116e-07,
+      "loss": 0.5203,
+      "step": 17891
+    },
+    {
+      "epoch": 4.749767688835789,
+      "grad_norm": 0.4845208375274857,
+      "learning_rate": 5.203064069655267e-07,
+      "loss": 0.5108,
+      "step": 17892
+    },
+    {
+      "epoch": 4.750033187309173,
+      "grad_norm": 0.47293316343275177,
+      "learning_rate": 5.20093220816213e-07,
+      "loss": 0.5232,
+      "step": 17893
+    },
+    {
+      "epoch": 4.750298685782557,
+      "grad_norm": 0.4685382210727578,
+      "learning_rate": 5.1988007327993e-07,
+      "loss": 0.5087,
+      "step": 17894
+    },
+    {
+      "epoch": 4.750564184255941,
+      "grad_norm": 0.48188921501061877,
+      "learning_rate": 5.196669643608332e-07,
+      "loss": 0.5495,
+      "step": 17895
+    },
+    {
+      "epoch": 4.750829682729324,
+      "grad_norm": 0.48190939784719933,
+      "learning_rate": 5.194538940630797e-07,
+      "loss": 0.5408,
+      "step": 17896
+    },
+    {
+      "epoch": 4.751095181202708,
+      "grad_norm": 0.4917213646389072,
+      "learning_rate": 5.192408623908246e-07,
+      "loss": 0.5307,
+      "step": 17897
+    },
+    {
+      "epoch": 4.751360679676091,
+      "grad_norm": 0.48580259334528425,
+      "learning_rate": 5.190278693482223e-07,
+      "loss": 0.5548,
+      "step": 17898
+    },
+    {
+      "epoch": 4.7516261781494755,
+      "grad_norm": 0.4834763923441684,
+      "learning_rate": 5.188149149394265e-07,
+      "loss": 0.5401,
+      "step": 17899
+    },
+    {
+      "epoch": 4.75189167662286,
+      "grad_norm": 0.5045726428295548,
+      "learning_rate": 5.186019991685908e-07,
+      "loss": 0.5429,
+      "step": 17900
+    },
+    {
+      "epoch": 4.752157175096244,
+      "grad_norm": 0.48783341637129474,
+      "learning_rate": 5.183891220398671e-07,
+      "loss": 0.5474,
+      "step": 17901
+    },
+    {
+      "epoch": 4.752422673569627,
+      "grad_norm": 0.4811210208871398,
+      "learning_rate": 5.181762835574072e-07,
+      "loss": 0.5225,
+      "step": 17902
+    },
+    {
+      "epoch": 4.752688172043011,
+      "grad_norm": 0.49374317360592773,
+      "learning_rate": 5.179634837253636e-07,
+      "loss": 0.5629,
+      "step": 17903
+    },
+    {
+      "epoch": 4.752953670516394,
+      "grad_norm": 0.48719439799501607,
+      "learning_rate": 5.177507225478834e-07,
+      "loss": 0.5543,
+      "step": 17904
+    },
+    {
+      "epoch": 4.753219168989778,
+      "grad_norm": 0.474326964396191,
+      "learning_rate": 5.175380000291183e-07,
+      "loss": 0.5086,
+      "step": 17905
+    },
+    {
+      "epoch": 4.753484667463162,
+      "grad_norm": 0.4827963566019136,
+      "learning_rate": 5.173253161732153e-07,
+      "loss": 0.5628,
+      "step": 17906
+    },
+    {
+      "epoch": 4.7537501659365455,
+      "grad_norm": 0.4927884638320415,
+      "learning_rate": 5.171126709843239e-07,
+      "loss": 0.5439,
+      "step": 17907
+    },
+    {
+      "epoch": 4.75401566440993,
+      "grad_norm": 0.4743941885771043,
+      "learning_rate": 5.169000644665895e-07,
+      "loss": 0.5254,
+      "step": 17908
+    },
+    {
+      "epoch": 4.754281162883314,
+      "grad_norm": 0.47208689482823535,
+      "learning_rate": 5.166874966241603e-07,
+      "loss": 0.4969,
+      "step": 17909
+    },
+    {
+      "epoch": 4.754546661356697,
+      "grad_norm": 0.47163712092322796,
+      "learning_rate": 5.164749674611805e-07,
+      "loss": 0.5217,
+      "step": 17910
+    },
+    {
+      "epoch": 4.754812159830081,
+      "grad_norm": 0.4847964076644604,
+      "learning_rate": 5.162624769817948e-07,
+      "loss": 0.5193,
+      "step": 17911
+    },
+    {
+      "epoch": 4.755077658303465,
+      "grad_norm": 0.48028106159435885,
+      "learning_rate": 5.160500251901487e-07,
+      "loss": 0.5359,
+      "step": 17912
+    },
+    {
+      "epoch": 4.755343156776848,
+      "grad_norm": 0.48025008009487075,
+      "learning_rate": 5.158376120903838e-07,
+      "loss": 0.5489,
+      "step": 17913
+    },
+    {
+      "epoch": 4.755608655250232,
+      "grad_norm": 0.4806269640401778,
+      "learning_rate": 5.156252376866441e-07,
+      "loss": 0.5338,
+      "step": 17914
+    },
+    {
+      "epoch": 4.755874153723616,
+      "grad_norm": 0.47999443219255883,
+      "learning_rate": 5.154129019830712e-07,
+      "loss": 0.5123,
+      "step": 17915
+    },
+    {
+      "epoch": 4.756139652197,
+      "grad_norm": 0.468921481054419,
+      "learning_rate": 5.152006049838054e-07,
+      "loss": 0.4917,
+      "step": 17916
+    },
+    {
+      "epoch": 4.756405150670384,
+      "grad_norm": 0.4835779455628183,
+      "learning_rate": 5.14988346692987e-07,
+      "loss": 0.513,
+      "step": 17917
+    },
+    {
+      "epoch": 4.756670649143768,
+      "grad_norm": 0.4777725284622087,
+      "learning_rate": 5.147761271147567e-07,
+      "loss": 0.5573,
+      "step": 17918
+    },
+    {
+      "epoch": 4.756936147617151,
+      "grad_norm": 0.48552926948335057,
+      "learning_rate": 5.145639462532517e-07,
+      "loss": 0.5583,
+      "step": 17919
+    },
+    {
+      "epoch": 4.757201646090535,
+      "grad_norm": 0.4803955917499256,
+      "learning_rate": 5.14351804112612e-07,
+      "loss": 0.5363,
+      "step": 17920
+    },
+    {
+      "epoch": 4.757467144563918,
+      "grad_norm": 0.49788221277017647,
+      "learning_rate": 5.141397006969734e-07,
+      "loss": 0.579,
+      "step": 17921
+    },
+    {
+      "epoch": 4.757732643037302,
+      "grad_norm": 0.4881799136032062,
+      "learning_rate": 5.139276360104726e-07,
+      "loss": 0.5635,
+      "step": 17922
+    },
+    {
+      "epoch": 4.7579981415106865,
+      "grad_norm": 0.501212464325793,
+      "learning_rate": 5.137156100572457e-07,
+      "loss": 0.6115,
+      "step": 17923
+    },
+    {
+      "epoch": 4.7582636399840705,
+      "grad_norm": 0.492331187666112,
+      "learning_rate": 5.135036228414275e-07,
+      "loss": 0.5231,
+      "step": 17924
+    },
+    {
+      "epoch": 4.758529138457454,
+      "grad_norm": 0.48549881108151655,
+      "learning_rate": 5.132916743671529e-07,
+      "loss": 0.5176,
+      "step": 17925
+    },
+    {
+      "epoch": 4.758794636930838,
+      "grad_norm": 0.4883286352545921,
+      "learning_rate": 5.130797646385551e-07,
+      "loss": 0.5486,
+      "step": 17926
+    },
+    {
+      "epoch": 4.759060135404221,
+      "grad_norm": 0.48783689371453287,
+      "learning_rate": 5.128678936597658e-07,
+      "loss": 0.5487,
+      "step": 17927
+    },
+    {
+      "epoch": 4.759325633877605,
+      "grad_norm": 0.48037008516744284,
+      "learning_rate": 5.126560614349189e-07,
+      "loss": 0.5559,
+      "step": 17928
+    },
+    {
+      "epoch": 4.759591132350989,
+      "grad_norm": 0.48421568964901096,
+      "learning_rate": 5.124442679681438e-07,
+      "loss": 0.5551,
+      "step": 17929
+    },
+    {
+      "epoch": 4.759856630824372,
+      "grad_norm": 0.47347003584851377,
+      "learning_rate": 5.122325132635724e-07,
+      "loss": 0.5428,
+      "step": 17930
+    },
+    {
+      "epoch": 4.7601221292977565,
+      "grad_norm": 0.4901966133754933,
+      "learning_rate": 5.120207973253336e-07,
+      "loss": 0.5613,
+      "step": 17931
+    },
+    {
+      "epoch": 4.7603876277711406,
+      "grad_norm": 0.4815469858212861,
+      "learning_rate": 5.118091201575578e-07,
+      "loss": 0.5341,
+      "step": 17932
+    },
+    {
+      "epoch": 4.760653126244524,
+      "grad_norm": 0.4799333996442336,
+      "learning_rate": 5.115974817643707e-07,
+      "loss": 0.5417,
+      "step": 17933
+    },
+    {
+      "epoch": 4.760918624717908,
+      "grad_norm": 0.47202794477798515,
+      "learning_rate": 5.113858821499018e-07,
+      "loss": 0.5338,
+      "step": 17934
+    },
+    {
+      "epoch": 4.761184123191292,
+      "grad_norm": 0.4854419638023756,
+      "learning_rate": 5.111743213182765e-07,
+      "loss": 0.5336,
+      "step": 17935
+    },
+    {
+      "epoch": 4.761449621664675,
+      "grad_norm": 0.47736216886216437,
+      "learning_rate": 5.109627992736221e-07,
+      "loss": 0.5623,
+      "step": 17936
+    },
+    {
+      "epoch": 4.761715120138059,
+      "grad_norm": 0.4848890508828434,
+      "learning_rate": 5.107513160200625e-07,
+      "loss": 0.5119,
+      "step": 17937
+    },
+    {
+      "epoch": 4.761980618611443,
+      "grad_norm": 0.4732577720970517,
+      "learning_rate": 5.105398715617235e-07,
+      "loss": 0.5323,
+      "step": 17938
+    },
+    {
+      "epoch": 4.7622461170848265,
+      "grad_norm": 0.4723088264776877,
+      "learning_rate": 5.10328465902728e-07,
+      "loss": 0.527,
+      "step": 17939
+    },
+    {
+      "epoch": 4.762511615558211,
+      "grad_norm": 0.4811508142267022,
+      "learning_rate": 5.101170990471982e-07,
+      "loss": 0.5315,
+      "step": 17940
+    },
+    {
+      "epoch": 4.762777114031595,
+      "grad_norm": 0.48242907486658976,
+      "learning_rate": 5.099057709992577e-07,
+      "loss": 0.5612,
+      "step": 17941
+    },
+    {
+      "epoch": 4.763042612504978,
+      "grad_norm": 0.4859614605068404,
+      "learning_rate": 5.096944817630267e-07,
+      "loss": 0.5199,
+      "step": 17942
+    },
+    {
+      "epoch": 4.763308110978362,
+      "grad_norm": 0.4749249901315378,
+      "learning_rate": 5.094832313426276e-07,
+      "loss": 0.5383,
+      "step": 17943
+    },
+    {
+      "epoch": 4.763573609451746,
+      "grad_norm": 0.5045171490318809,
+      "learning_rate": 5.09272019742178e-07,
+      "loss": 0.5452,
+      "step": 17944
+    },
+    {
+      "epoch": 4.763839107925129,
+      "grad_norm": 0.47788859236726694,
+      "learning_rate": 5.090608469657987e-07,
+      "loss": 0.5369,
+      "step": 17945
+    },
+    {
+      "epoch": 4.764104606398513,
+      "grad_norm": 0.46960584792995125,
+      "learning_rate": 5.088497130176068e-07,
+      "loss": 0.5102,
+      "step": 17946
+    },
+    {
+      "epoch": 4.764370104871897,
+      "grad_norm": 0.47424828535665375,
+      "learning_rate": 5.086386179017208e-07,
+      "loss": 0.5224,
+      "step": 17947
+    },
+    {
+      "epoch": 4.764635603345281,
+      "grad_norm": 0.49090557505809523,
+      "learning_rate": 5.084275616222581e-07,
+      "loss": 0.5196,
+      "step": 17948
+    },
+    {
+      "epoch": 4.764901101818665,
+      "grad_norm": 0.4739687791773348,
+      "learning_rate": 5.08216544183334e-07,
+      "loss": 0.4848,
+      "step": 17949
+    },
+    {
+      "epoch": 4.765166600292048,
+      "grad_norm": 0.4958530307022878,
+      "learning_rate": 5.080055655890642e-07,
+      "loss": 0.5962,
+      "step": 17950
+    },
+    {
+      "epoch": 4.765432098765432,
+      "grad_norm": 0.48015002213866975,
+      "learning_rate": 5.077946258435624e-07,
+      "loss": 0.5181,
+      "step": 17951
+    },
+    {
+      "epoch": 4.765697597238816,
+      "grad_norm": 0.4751311976269479,
+      "learning_rate": 5.075837249509438e-07,
+      "loss": 0.5365,
+      "step": 17952
+    },
+    {
+      "epoch": 4.7659630957122,
+      "grad_norm": 0.4895303183181923,
+      "learning_rate": 5.073728629153201e-07,
+      "loss": 0.5657,
+      "step": 17953
+    },
+    {
+      "epoch": 4.766228594185583,
+      "grad_norm": 0.4794424901830242,
+      "learning_rate": 5.071620397408053e-07,
+      "loss": 0.5614,
+      "step": 17954
+    },
+    {
+      "epoch": 4.766494092658967,
+      "grad_norm": 0.4815297694787281,
+      "learning_rate": 5.069512554315098e-07,
+      "loss": 0.5188,
+      "step": 17955
+    },
+    {
+      "epoch": 4.766759591132351,
+      "grad_norm": 0.49009559761974075,
+      "learning_rate": 5.06740509991544e-07,
+      "loss": 0.5427,
+      "step": 17956
+    },
+    {
+      "epoch": 4.767025089605735,
+      "grad_norm": 0.4650005958166525,
+      "learning_rate": 5.065298034250194e-07,
+      "loss": 0.5458,
+      "step": 17957
+    },
+    {
+      "epoch": 4.767290588079119,
+      "grad_norm": 0.4741141592963785,
+      "learning_rate": 5.063191357360437e-07,
+      "loss": 0.5598,
+      "step": 17958
+    },
+    {
+      "epoch": 4.767556086552502,
+      "grad_norm": 0.4920850418606148,
+      "learning_rate": 5.061085069287267e-07,
+      "loss": 0.553,
+      "step": 17959
+    },
+    {
+      "epoch": 4.767821585025886,
+      "grad_norm": 0.49322030854640186,
+      "learning_rate": 5.058979170071754e-07,
+      "loss": 0.4915,
+      "step": 17960
+    },
+    {
+      "epoch": 4.76808708349927,
+      "grad_norm": 0.48341125843990435,
+      "learning_rate": 5.056873659754982e-07,
+      "loss": 0.5876,
+      "step": 17961
+    },
+    {
+      "epoch": 4.768352581972653,
+      "grad_norm": 0.4776790773986275,
+      "learning_rate": 5.054768538377991e-07,
+      "loss": 0.5565,
+      "step": 17962
+    },
+    {
+      "epoch": 4.768618080446037,
+      "grad_norm": 0.485103299948917,
+      "learning_rate": 5.052663805981853e-07,
+      "loss": 0.5518,
+      "step": 17963
+    },
+    {
+      "epoch": 4.7688835789194215,
+      "grad_norm": 0.4754670898136354,
+      "learning_rate": 5.050559462607605e-07,
+      "loss": 0.5337,
+      "step": 17964
+    },
+    {
+      "epoch": 4.769149077392805,
+      "grad_norm": 0.47136126517508237,
+      "learning_rate": 5.048455508296299e-07,
+      "loss": 0.5122,
+      "step": 17965
+    },
+    {
+      "epoch": 4.769414575866189,
+      "grad_norm": 0.482062471416318,
+      "learning_rate": 5.04635194308896e-07,
+      "loss": 0.5131,
+      "step": 17966
+    },
+    {
+      "epoch": 4.769680074339573,
+      "grad_norm": 0.4796535054689214,
+      "learning_rate": 5.044248767026605e-07,
+      "loss": 0.5483,
+      "step": 17967
+    },
+    {
+      "epoch": 4.769945572812956,
+      "grad_norm": 0.4797795483510902,
+      "learning_rate": 5.042145980150268e-07,
+      "loss": 0.5826,
+      "step": 17968
+    },
+    {
+      "epoch": 4.77021107128634,
+      "grad_norm": 0.48733148488387107,
+      "learning_rate": 5.040043582500939e-07,
+      "loss": 0.5648,
+      "step": 17969
+    },
+    {
+      "epoch": 4.770476569759724,
+      "grad_norm": 0.49188546172471137,
+      "learning_rate": 5.037941574119642e-07,
+      "loss": 0.5266,
+      "step": 17970
+    },
+    {
+      "epoch": 4.770742068233107,
+      "grad_norm": 0.5203461590322918,
+      "learning_rate": 5.035839955047351e-07,
+      "loss": 0.5219,
+      "step": 17971
+    },
+    {
+      "epoch": 4.7710075667064915,
+      "grad_norm": 0.4825189718460147,
+      "learning_rate": 5.033738725325069e-07,
+      "loss": 0.5524,
+      "step": 17972
+    },
+    {
+      "epoch": 4.771273065179876,
+      "grad_norm": 0.47650800032065554,
+      "learning_rate": 5.031637884993768e-07,
+      "loss": 0.5275,
+      "step": 17973
+    },
+    {
+      "epoch": 4.771538563653259,
+      "grad_norm": 0.49027195208704977,
+      "learning_rate": 5.029537434094411e-07,
+      "loss": 0.5415,
+      "step": 17974
+    },
+    {
+      "epoch": 4.771804062126643,
+      "grad_norm": 0.47321546031407014,
+      "learning_rate": 5.027437372667976e-07,
+      "loss": 0.5426,
+      "step": 17975
+    },
+    {
+      "epoch": 4.772069560600027,
+      "grad_norm": 0.47694543927081745,
+      "learning_rate": 5.025337700755409e-07,
+      "loss": 0.5453,
+      "step": 17976
+    },
+    {
+      "epoch": 4.77233505907341,
+      "grad_norm": 0.4944207033297632,
+      "learning_rate": 5.023238418397669e-07,
+      "loss": 0.5184,
+      "step": 17977
+    },
+    {
+      "epoch": 4.772600557546794,
+      "grad_norm": 0.4935000509331629,
+      "learning_rate": 5.021139525635693e-07,
+      "loss": 0.5657,
+      "step": 17978
+    },
+    {
+      "epoch": 4.772866056020177,
+      "grad_norm": 0.46909103904099425,
+      "learning_rate": 5.019041022510413e-07,
+      "loss": 0.5307,
+      "step": 17979
+    },
+    {
+      "epoch": 4.7731315544935615,
+      "grad_norm": 0.4845856763415569,
+      "learning_rate": 5.016942909062744e-07,
+      "loss": 0.5372,
+      "step": 17980
+    },
+    {
+      "epoch": 4.773397052966946,
+      "grad_norm": 0.4969572340541802,
+      "learning_rate": 5.014845185333625e-07,
+      "loss": 0.5176,
+      "step": 17981
+    },
+    {
+      "epoch": 4.77366255144033,
+      "grad_norm": 0.4704764211053215,
+      "learning_rate": 5.012747851363949e-07,
+      "loss": 0.4902,
+      "step": 17982
+    },
+    {
+      "epoch": 4.773928049913713,
+      "grad_norm": 0.49391565351521594,
+      "learning_rate": 5.010650907194636e-07,
+      "loss": 0.5146,
+      "step": 17983
+    },
+    {
+      "epoch": 4.774193548387097,
+      "grad_norm": 0.48607878756327755,
+      "learning_rate": 5.008554352866574e-07,
+      "loss": 0.548,
+      "step": 17984
+    },
+    {
+      "epoch": 4.77445904686048,
+      "grad_norm": 0.4726870812051951,
+      "learning_rate": 5.00645818842064e-07,
+      "loss": 0.5405,
+      "step": 17985
+    },
+    {
+      "epoch": 4.774724545333864,
+      "grad_norm": 0.472521350403769,
+      "learning_rate": 5.004362413897732e-07,
+      "loss": 0.521,
+      "step": 17986
+    },
+    {
+      "epoch": 4.774990043807248,
+      "grad_norm": 0.4920755405573564,
+      "learning_rate": 5.002267029338706e-07,
+      "loss": 0.5647,
+      "step": 17987
+    },
+    {
+      "epoch": 4.7752555422806315,
+      "grad_norm": 0.4679958859970393,
+      "learning_rate": 5.000172034784442e-07,
+      "loss": 0.4773,
+      "step": 17988
+    },
+    {
+      "epoch": 4.775521040754016,
+      "grad_norm": 0.47033909448224076,
+      "learning_rate": 4.998077430275789e-07,
+      "loss": 0.5527,
+      "step": 17989
+    },
+    {
+      "epoch": 4.7757865392274,
+      "grad_norm": 0.49211092173185367,
+      "learning_rate": 4.995983215853609e-07,
+      "loss": 0.5383,
+      "step": 17990
+    },
+    {
+      "epoch": 4.776052037700783,
+      "grad_norm": 0.478128156626304,
+      "learning_rate": 4.993889391558721e-07,
+      "loss": 0.5451,
+      "step": 17991
+    },
+    {
+      "epoch": 4.776317536174167,
+      "grad_norm": 0.4907558027169939,
+      "learning_rate": 4.99179595743198e-07,
+      "loss": 0.5618,
+      "step": 17992
+    },
+    {
+      "epoch": 4.776583034647551,
+      "grad_norm": 0.48639768900763636,
+      "learning_rate": 4.989702913514202e-07,
+      "loss": 0.564,
+      "step": 17993
+    },
+    {
+      "epoch": 4.776848533120934,
+      "grad_norm": 0.4998120079017491,
+      "learning_rate": 4.987610259846209e-07,
+      "loss": 0.5238,
+      "step": 17994
+    },
+    {
+      "epoch": 4.777114031594318,
+      "grad_norm": 0.4697520678111949,
+      "learning_rate": 4.98551799646883e-07,
+      "loss": 0.5339,
+      "step": 17995
+    },
+    {
+      "epoch": 4.777379530067702,
+      "grad_norm": 0.4875955869912905,
+      "learning_rate": 4.983426123422838e-07,
+      "loss": 0.5385,
+      "step": 17996
+    },
+    {
+      "epoch": 4.777645028541086,
+      "grad_norm": 0.479003098587137,
+      "learning_rate": 4.981334640749056e-07,
+      "loss": 0.5599,
+      "step": 17997
+    },
+    {
+      "epoch": 4.77791052701447,
+      "grad_norm": 0.48278670573028626,
+      "learning_rate": 4.979243548488252e-07,
+      "loss": 0.5318,
+      "step": 17998
+    },
+    {
+      "epoch": 4.778176025487854,
+      "grad_norm": 0.48341103935871743,
+      "learning_rate": 4.977152846681227e-07,
+      "loss": 0.5272,
+      "step": 17999
+    },
+    {
+      "epoch": 4.778441523961237,
+      "grad_norm": 0.4756196890187936,
+      "learning_rate": 4.975062535368739e-07,
+      "loss": 0.5433,
+      "step": 18000
+    },
+    {
+      "epoch": 4.778707022434621,
+      "grad_norm": 0.48160775073401735,
+      "learning_rate": 4.972972614591568e-07,
+      "loss": 0.5452,
+      "step": 18001
+    },
+    {
+      "epoch": 4.778972520908004,
+      "grad_norm": 0.49010757349740214,
+      "learning_rate": 4.970883084390463e-07,
+      "loss": 0.5536,
+      "step": 18002
+    },
+    {
+      "epoch": 4.779238019381388,
+      "grad_norm": 0.4648630186166757,
+      "learning_rate": 4.968793944806171e-07,
+      "loss": 0.5223,
+      "step": 18003
+    },
+    {
+      "epoch": 4.7795035178547725,
+      "grad_norm": 0.49366987902299253,
+      "learning_rate": 4.966705195879454e-07,
+      "loss": 0.5901,
+      "step": 18004
+    },
+    {
+      "epoch": 4.7797690163281565,
+      "grad_norm": 0.48845875544002404,
+      "learning_rate": 4.964616837651024e-07,
+      "loss": 0.5119,
+      "step": 18005
+    },
+    {
+      "epoch": 4.78003451480154,
+      "grad_norm": 0.47370209877336367,
+      "learning_rate": 4.962528870161629e-07,
+      "loss": 0.5367,
+      "step": 18006
+    },
+    {
+      "epoch": 4.780300013274924,
+      "grad_norm": 0.47661059646892523,
+      "learning_rate": 4.960441293451981e-07,
+      "loss": 0.5134,
+      "step": 18007
+    },
+    {
+      "epoch": 4.780565511748307,
+      "grad_norm": 0.47190686787763464,
+      "learning_rate": 4.958354107562793e-07,
+      "loss": 0.4963,
+      "step": 18008
+    },
+    {
+      "epoch": 4.780831010221691,
+      "grad_norm": 0.48666748430530027,
+      "learning_rate": 4.956267312534762e-07,
+      "loss": 0.5306,
+      "step": 18009
+    },
+    {
+      "epoch": 4.781096508695075,
+      "grad_norm": 0.48077044730429325,
+      "learning_rate": 4.954180908408604e-07,
+      "loss": 0.5131,
+      "step": 18010
+    },
+    {
+      "epoch": 4.781362007168459,
+      "grad_norm": 0.4926430032018436,
+      "learning_rate": 4.95209489522499e-07,
+      "loss": 0.5455,
+      "step": 18011
+    },
+    {
+      "epoch": 4.7816275056418425,
+      "grad_norm": 0.47520193951153944,
+      "learning_rate": 4.950009273024619e-07,
+      "loss": 0.4892,
+      "step": 18012
+    },
+    {
+      "epoch": 4.7818930041152266,
+      "grad_norm": 0.47987384381434717,
+      "learning_rate": 4.947924041848157e-07,
+      "loss": 0.5358,
+      "step": 18013
+    },
+    {
+      "epoch": 4.78215850258861,
+      "grad_norm": 0.47823294223916746,
+      "learning_rate": 4.945839201736266e-07,
+      "loss": 0.553,
+      "step": 18014
+    },
+    {
+      "epoch": 4.782424001061994,
+      "grad_norm": 0.4812049530182868,
+      "learning_rate": 4.943754752729618e-07,
+      "loss": 0.5464,
+      "step": 18015
+    },
+    {
+      "epoch": 4.782689499535378,
+      "grad_norm": 0.46993030851701145,
+      "learning_rate": 4.941670694868853e-07,
+      "loss": 0.5216,
+      "step": 18016
+    },
+    {
+      "epoch": 4.782954998008761,
+      "grad_norm": 0.47952954602096765,
+      "learning_rate": 4.939587028194625e-07,
+      "loss": 0.5597,
+      "step": 18017
+    },
+    {
+      "epoch": 4.783220496482145,
+      "grad_norm": 0.47811568430812457,
+      "learning_rate": 4.937503752747569e-07,
+      "loss": 0.5307,
+      "step": 18018
+    },
+    {
+      "epoch": 4.783485994955529,
+      "grad_norm": 0.48591690381905783,
+      "learning_rate": 4.935420868568303e-07,
+      "loss": 0.5542,
+      "step": 18019
+    },
+    {
+      "epoch": 4.7837514934289125,
+      "grad_norm": 0.4806279745635947,
+      "learning_rate": 4.933338375697464e-07,
+      "loss": 0.5541,
+      "step": 18020
+    },
+    {
+      "epoch": 4.784016991902297,
+      "grad_norm": 0.48046943643012774,
+      "learning_rate": 4.931256274175652e-07,
+      "loss": 0.5778,
+      "step": 18021
+    },
+    {
+      "epoch": 4.784282490375681,
+      "grad_norm": 0.4849189602190248,
+      "learning_rate": 4.929174564043484e-07,
+      "loss": 0.5465,
+      "step": 18022
+    },
+    {
+      "epoch": 4.784547988849064,
+      "grad_norm": 0.47786896744000656,
+      "learning_rate": 4.92709324534155e-07,
+      "loss": 0.5326,
+      "step": 18023
+    },
+    {
+      "epoch": 4.784813487322448,
+      "grad_norm": 0.4760970245939848,
+      "learning_rate": 4.925012318110455e-07,
+      "loss": 0.517,
+      "step": 18024
+    },
+    {
+      "epoch": 4.785078985795832,
+      "grad_norm": 0.49509001498487903,
+      "learning_rate": 4.922931782390761e-07,
+      "loss": 0.5306,
+      "step": 18025
+    },
+    {
+      "epoch": 4.785344484269215,
+      "grad_norm": 0.48295788476680324,
+      "learning_rate": 4.920851638223062e-07,
+      "loss": 0.5618,
+      "step": 18026
+    },
+    {
+      "epoch": 4.785609982742599,
+      "grad_norm": 0.4710697182910148,
+      "learning_rate": 4.918771885647908e-07,
+      "loss": 0.5683,
+      "step": 18027
+    },
+    {
+      "epoch": 4.785875481215983,
+      "grad_norm": 0.4741252788130125,
+      "learning_rate": 4.916692524705879e-07,
+      "loss": 0.5287,
+      "step": 18028
+    },
+    {
+      "epoch": 4.786140979689367,
+      "grad_norm": 0.4707627291170045,
+      "learning_rate": 4.914613555437511e-07,
+      "loss": 0.5354,
+      "step": 18029
+    },
+    {
+      "epoch": 4.786406478162751,
+      "grad_norm": 0.478199321523249,
+      "learning_rate": 4.912534977883363e-07,
+      "loss": 0.5541,
+      "step": 18030
+    },
+    {
+      "epoch": 4.786671976636134,
+      "grad_norm": 0.4673310226376173,
+      "learning_rate": 4.910456792083965e-07,
+      "loss": 0.5102,
+      "step": 18031
+    },
+    {
+      "epoch": 4.786937475109518,
+      "grad_norm": 0.48598384448173665,
+      "learning_rate": 4.908378998079839e-07,
+      "loss": 0.5095,
+      "step": 18032
+    },
+    {
+      "epoch": 4.787202973582902,
+      "grad_norm": 0.46688879767938213,
+      "learning_rate": 4.906301595911523e-07,
+      "loss": 0.514,
+      "step": 18033
+    },
+    {
+      "epoch": 4.787468472056286,
+      "grad_norm": 0.4759717991044667,
+      "learning_rate": 4.904224585619519e-07,
+      "loss": 0.5368,
+      "step": 18034
+    },
+    {
+      "epoch": 4.787733970529669,
+      "grad_norm": 0.486806039995556,
+      "learning_rate": 4.902147967244342e-07,
+      "loss": 0.5312,
+      "step": 18035
+    },
+    {
+      "epoch": 4.787999469003053,
+      "grad_norm": 0.4751445623996909,
+      "learning_rate": 4.90007174082649e-07,
+      "loss": 0.5307,
+      "step": 18036
+    },
+    {
+      "epoch": 4.788264967476437,
+      "grad_norm": 0.46700504255277153,
+      "learning_rate": 4.897995906406453e-07,
+      "loss": 0.5372,
+      "step": 18037
+    },
+    {
+      "epoch": 4.788530465949821,
+      "grad_norm": 0.49595746518074574,
+      "learning_rate": 4.895920464024703e-07,
+      "loss": 0.5552,
+      "step": 18038
+    },
+    {
+      "epoch": 4.788795964423205,
+      "grad_norm": 0.48252636881136807,
+      "learning_rate": 4.893845413721732e-07,
+      "loss": 0.4963,
+      "step": 18039
+    },
+    {
+      "epoch": 4.789061462896588,
+      "grad_norm": 0.4817440671839129,
+      "learning_rate": 4.891770755538008e-07,
+      "loss": 0.5383,
+      "step": 18040
+    },
+    {
+      "epoch": 4.789326961369972,
+      "grad_norm": 0.48808918807872487,
+      "learning_rate": 4.889696489513987e-07,
+      "loss": 0.5328,
+      "step": 18041
+    },
+    {
+      "epoch": 4.789592459843356,
+      "grad_norm": 0.48080614391787435,
+      "learning_rate": 4.887622615690124e-07,
+      "loss": 0.5383,
+      "step": 18042
+    },
+    {
+      "epoch": 4.789857958316739,
+      "grad_norm": 0.48891648476418476,
+      "learning_rate": 4.885549134106859e-07,
+      "loss": 0.534,
+      "step": 18043
+    },
+    {
+      "epoch": 4.790123456790123,
+      "grad_norm": 0.48198921729106275,
+      "learning_rate": 4.883476044804638e-07,
+      "loss": 0.5147,
+      "step": 18044
+    },
+    {
+      "epoch": 4.7903889552635075,
+      "grad_norm": 0.4780619129473246,
+      "learning_rate": 4.881403347823885e-07,
+      "loss": 0.5287,
+      "step": 18045
+    },
+    {
+      "epoch": 4.790654453736891,
+      "grad_norm": 0.47193975189744247,
+      "learning_rate": 4.879331043205029e-07,
+      "loss": 0.5189,
+      "step": 18046
+    },
+    {
+      "epoch": 4.790919952210275,
+      "grad_norm": 0.49309846216577247,
+      "learning_rate": 4.877259130988482e-07,
+      "loss": 0.5391,
+      "step": 18047
+    },
+    {
+      "epoch": 4.791185450683659,
+      "grad_norm": 0.47887047798048377,
+      "learning_rate": 4.875187611214644e-07,
+      "loss": 0.5262,
+      "step": 18048
+    },
+    {
+      "epoch": 4.791450949157042,
+      "grad_norm": 0.49074135708434174,
+      "learning_rate": 4.873116483923929e-07,
+      "loss": 0.5474,
+      "step": 18049
+    },
+    {
+      "epoch": 4.791716447630426,
+      "grad_norm": 0.4752380108267149,
+      "learning_rate": 4.871045749156713e-07,
+      "loss": 0.5517,
+      "step": 18050
+    },
+    {
+      "epoch": 4.79198194610381,
+      "grad_norm": 0.4847861712212924,
+      "learning_rate": 4.8689754069534e-07,
+      "loss": 0.5216,
+      "step": 18051
+    },
+    {
+      "epoch": 4.792247444577193,
+      "grad_norm": 0.47164673399218426,
+      "learning_rate": 4.866905457354346e-07,
+      "loss": 0.4983,
+      "step": 18052
+    },
+    {
+      "epoch": 4.7925129430505775,
+      "grad_norm": 0.486669600177474,
+      "learning_rate": 4.864835900399945e-07,
+      "loss": 0.558,
+      "step": 18053
+    },
+    {
+      "epoch": 4.792778441523962,
+      "grad_norm": 0.4703553359373661,
+      "learning_rate": 4.862766736130528e-07,
+      "loss": 0.4913,
+      "step": 18054
+    },
+    {
+      "epoch": 4.793043939997345,
+      "grad_norm": 0.47885172037638307,
+      "learning_rate": 4.860697964586472e-07,
+      "loss": 0.5278,
+      "step": 18055
+    },
+    {
+      "epoch": 4.793309438470729,
+      "grad_norm": 0.47643238443238634,
+      "learning_rate": 4.858629585808109e-07,
+      "loss": 0.5497,
+      "step": 18056
+    },
+    {
+      "epoch": 4.793574936944113,
+      "grad_norm": 0.4805325425023738,
+      "learning_rate": 4.856561599835787e-07,
+      "loss": 0.5053,
+      "step": 18057
+    },
+    {
+      "epoch": 4.793840435417496,
+      "grad_norm": 0.493304177954982,
+      "learning_rate": 4.854494006709831e-07,
+      "loss": 0.5622,
+      "step": 18058
+    },
+    {
+      "epoch": 4.79410593389088,
+      "grad_norm": 0.48153951469383405,
+      "learning_rate": 4.852426806470573e-07,
+      "loss": 0.5379,
+      "step": 18059
+    },
+    {
+      "epoch": 4.794371432364263,
+      "grad_norm": 0.47106527057176684,
+      "learning_rate": 4.850359999158319e-07,
+      "loss": 0.5132,
+      "step": 18060
+    },
+    {
+      "epoch": 4.7946369308376475,
+      "grad_norm": 0.49692382849406064,
+      "learning_rate": 4.848293584813377e-07,
+      "loss": 0.561,
+      "step": 18061
+    },
+    {
+      "epoch": 4.794902429311032,
+      "grad_norm": 0.48447572619569007,
+      "learning_rate": 4.846227563476052e-07,
+      "loss": 0.5436,
+      "step": 18062
+    },
+    {
+      "epoch": 4.795167927784416,
+      "grad_norm": 0.4788628276328648,
+      "learning_rate": 4.844161935186631e-07,
+      "loss": 0.5347,
+      "step": 18063
+    },
+    {
+      "epoch": 4.795433426257799,
+      "grad_norm": 0.49826638870186746,
+      "learning_rate": 4.842096699985408e-07,
+      "loss": 0.5487,
+      "step": 18064
+    },
+    {
+      "epoch": 4.795698924731183,
+      "grad_norm": 0.4830054635995838,
+      "learning_rate": 4.840031857912655e-07,
+      "loss": 0.5362,
+      "step": 18065
+    },
+    {
+      "epoch": 4.795964423204566,
+      "grad_norm": 0.4803891413626469,
+      "learning_rate": 4.837967409008634e-07,
+      "loss": 0.5377,
+      "step": 18066
+    },
+    {
+      "epoch": 4.79622992167795,
+      "grad_norm": 0.4811342475177499,
+      "learning_rate": 4.83590335331362e-07,
+      "loss": 0.546,
+      "step": 18067
+    },
+    {
+      "epoch": 4.796495420151334,
+      "grad_norm": 0.4941977940912998,
+      "learning_rate": 4.833839690867853e-07,
+      "loss": 0.56,
+      "step": 18068
+    },
+    {
+      "epoch": 4.7967609186247175,
+      "grad_norm": 0.4926254723278097,
+      "learning_rate": 4.831776421711595e-07,
+      "loss": 0.5667,
+      "step": 18069
+    },
+    {
+      "epoch": 4.797026417098102,
+      "grad_norm": 0.4862035652341599,
+      "learning_rate": 4.829713545885078e-07,
+      "loss": 0.5446,
+      "step": 18070
+    },
+    {
+      "epoch": 4.797291915571486,
+      "grad_norm": 0.47240278887666426,
+      "learning_rate": 4.827651063428532e-07,
+      "loss": 0.5546,
+      "step": 18071
+    },
+    {
+      "epoch": 4.797557414044869,
+      "grad_norm": 0.47790974443906487,
+      "learning_rate": 4.825588974382175e-07,
+      "loss": 0.54,
+      "step": 18072
+    },
+    {
+      "epoch": 4.797822912518253,
+      "grad_norm": 0.4864142006364324,
+      "learning_rate": 4.823527278786233e-07,
+      "loss": 0.5436,
+      "step": 18073
+    },
+    {
+      "epoch": 4.798088410991637,
+      "grad_norm": 0.4857802115007357,
+      "learning_rate": 4.821465976680903e-07,
+      "loss": 0.5475,
+      "step": 18074
+    },
+    {
+      "epoch": 4.79835390946502,
+      "grad_norm": 0.47750188603211785,
+      "learning_rate": 4.8194050681064e-07,
+      "loss": 0.5019,
+      "step": 18075
+    },
+    {
+      "epoch": 4.798619407938404,
+      "grad_norm": 0.49657485060990497,
+      "learning_rate": 4.81734455310291e-07,
+      "loss": 0.5543,
+      "step": 18076
+    },
+    {
+      "epoch": 4.798884906411788,
+      "grad_norm": 0.4702482105271751,
+      "learning_rate": 4.815284431710606e-07,
+      "loss": 0.5354,
+      "step": 18077
+    },
+    {
+      "epoch": 4.799150404885172,
+      "grad_norm": 0.485621242238804,
+      "learning_rate": 4.813224703969685e-07,
+      "loss": 0.5384,
+      "step": 18078
+    },
+    {
+      "epoch": 4.799415903358556,
+      "grad_norm": 0.4777676597048149,
+      "learning_rate": 4.811165369920304e-07,
+      "loss": 0.507,
+      "step": 18079
+    },
+    {
+      "epoch": 4.79968140183194,
+      "grad_norm": 0.4695559184168845,
+      "learning_rate": 4.809106429602633e-07,
+      "loss": 0.5221,
+      "step": 18080
+    },
+    {
+      "epoch": 4.799946900305323,
+      "grad_norm": 0.480461962549017,
+      "learning_rate": 4.807047883056815e-07,
+      "loss": 0.5271,
+      "step": 18081
+    },
+    {
+      "epoch": 4.800212398778707,
+      "grad_norm": 0.4806052795296357,
+      "learning_rate": 4.80498973032302e-07,
+      "loss": 0.5353,
+      "step": 18082
+    },
+    {
+      "epoch": 4.800477897252091,
+      "grad_norm": 0.4830690033753034,
+      "learning_rate": 4.802931971441354e-07,
+      "loss": 0.5561,
+      "step": 18083
+    },
+    {
+      "epoch": 4.800743395725474,
+      "grad_norm": 0.47219807663387287,
+      "learning_rate": 4.800874606451969e-07,
+      "loss": 0.5156,
+      "step": 18084
+    },
+    {
+      "epoch": 4.8010088941988585,
+      "grad_norm": 0.47760549244543227,
+      "learning_rate": 4.798817635394992e-07,
+      "loss": 0.5117,
+      "step": 18085
+    },
+    {
+      "epoch": 4.8012743926722425,
+      "grad_norm": 0.4711090619099747,
+      "learning_rate": 4.796761058310526e-07,
+      "loss": 0.5038,
+      "step": 18086
+    },
+    {
+      "epoch": 4.801539891145626,
+      "grad_norm": 0.4746281417910265,
+      "learning_rate": 4.79470487523869e-07,
+      "loss": 0.5375,
+      "step": 18087
+    },
+    {
+      "epoch": 4.80180538961901,
+      "grad_norm": 0.4950114929714865,
+      "learning_rate": 4.792649086219577e-07,
+      "loss": 0.5309,
+      "step": 18088
+    },
+    {
+      "epoch": 4.802070888092393,
+      "grad_norm": 0.4679886885819684,
+      "learning_rate": 4.790593691293288e-07,
+      "loss": 0.5022,
+      "step": 18089
+    },
+    {
+      "epoch": 4.802336386565777,
+      "grad_norm": 0.486960213970011,
+      "learning_rate": 4.788538690499894e-07,
+      "loss": 0.5343,
+      "step": 18090
+    },
+    {
+      "epoch": 4.802601885039161,
+      "grad_norm": 0.48447050951531645,
+      "learning_rate": 4.786484083879491e-07,
+      "loss": 0.5126,
+      "step": 18091
+    },
+    {
+      "epoch": 4.802867383512545,
+      "grad_norm": 0.482871808561654,
+      "learning_rate": 4.784429871472132e-07,
+      "loss": 0.5421,
+      "step": 18092
+    },
+    {
+      "epoch": 4.8031328819859285,
+      "grad_norm": 0.4792006832898944,
+      "learning_rate": 4.782376053317891e-07,
+      "loss": 0.5278,
+      "step": 18093
+    },
+    {
+      "epoch": 4.8033983804593126,
+      "grad_norm": 0.4970194550934078,
+      "learning_rate": 4.780322629456822e-07,
+      "loss": 0.5523,
+      "step": 18094
+    },
+    {
+      "epoch": 4.803663878932696,
+      "grad_norm": 0.49029906821320396,
+      "learning_rate": 4.778269599928959e-07,
+      "loss": 0.5303,
+      "step": 18095
+    },
+    {
+      "epoch": 4.80392937740608,
+      "grad_norm": 0.4867039217964488,
+      "learning_rate": 4.776216964774355e-07,
+      "loss": 0.5106,
+      "step": 18096
+    },
+    {
+      "epoch": 4.804194875879464,
+      "grad_norm": 0.4779262534036763,
+      "learning_rate": 4.774164724033034e-07,
+      "loss": 0.5185,
+      "step": 18097
+    },
+    {
+      "epoch": 4.804460374352847,
+      "grad_norm": 0.4814860079155979,
+      "learning_rate": 4.772112877745025e-07,
+      "loss": 0.5403,
+      "step": 18098
+    },
+    {
+      "epoch": 4.804725872826231,
+      "grad_norm": 0.4716925891695341,
+      "learning_rate": 4.770061425950345e-07,
+      "loss": 0.5453,
+      "step": 18099
+    },
+    {
+      "epoch": 4.804991371299615,
+      "grad_norm": 0.4882553779601527,
+      "learning_rate": 4.7680103686889967e-07,
+      "loss": 0.5207,
+      "step": 18100
+    },
+    {
+      "epoch": 4.8052568697729985,
+      "grad_norm": 0.4881965131752467,
+      "learning_rate": 4.765959706000975e-07,
+      "loss": 0.5573,
+      "step": 18101
+    },
+    {
+      "epoch": 4.805522368246383,
+      "grad_norm": 0.47582457175955745,
+      "learning_rate": 4.7639094379262895e-07,
+      "loss": 0.5185,
+      "step": 18102
+    },
+    {
+      "epoch": 4.805787866719767,
+      "grad_norm": 0.4735654749075319,
+      "learning_rate": 4.7618595645049085e-07,
+      "loss": 0.5081,
+      "step": 18103
+    },
+    {
+      "epoch": 4.80605336519315,
+      "grad_norm": 0.47415381038675497,
+      "learning_rate": 4.7598100857768225e-07,
+      "loss": 0.5401,
+      "step": 18104
+    },
+    {
+      "epoch": 4.806318863666534,
+      "grad_norm": 0.4973999258339989,
+      "learning_rate": 4.7577610017819993e-07,
+      "loss": 0.5381,
+      "step": 18105
+    },
+    {
+      "epoch": 4.806584362139918,
+      "grad_norm": 0.48847274027990056,
+      "learning_rate": 4.75571231256039e-07,
+      "loss": 0.559,
+      "step": 18106
+    },
+    {
+      "epoch": 4.806849860613301,
+      "grad_norm": 0.4836283631469511,
+      "learning_rate": 4.753664018151963e-07,
+      "loss": 0.5516,
+      "step": 18107
+    },
+    {
+      "epoch": 4.807115359086685,
+      "grad_norm": 0.47401409668649647,
+      "learning_rate": 4.7516161185966554e-07,
+      "loss": 0.5243,
+      "step": 18108
+    },
+    {
+      "epoch": 4.807380857560069,
+      "grad_norm": 0.47440283071387174,
+      "learning_rate": 4.749568613934413e-07,
+      "loss": 0.5457,
+      "step": 18109
+    },
+    {
+      "epoch": 4.807646356033453,
+      "grad_norm": 0.48993423402398245,
+      "learning_rate": 4.74752150420516e-07,
+      "loss": 0.5681,
+      "step": 18110
+    },
+    {
+      "epoch": 4.807911854506837,
+      "grad_norm": 0.48264029038372586,
+      "learning_rate": 4.7454747894488313e-07,
+      "loss": 0.5447,
+      "step": 18111
+    },
+    {
+      "epoch": 4.80817735298022,
+      "grad_norm": 0.4778735567774687,
+      "learning_rate": 4.743428469705336e-07,
+      "loss": 0.5359,
+      "step": 18112
+    },
+    {
+      "epoch": 4.808442851453604,
+      "grad_norm": 0.5071405805350742,
+      "learning_rate": 4.741382545014572e-07,
+      "loss": 0.5273,
+      "step": 18113
+    },
+    {
+      "epoch": 4.808708349926988,
+      "grad_norm": 0.4716033665844714,
+      "learning_rate": 4.739337015416462e-07,
+      "loss": 0.5478,
+      "step": 18114
+    },
+    {
+      "epoch": 4.808973848400372,
+      "grad_norm": 0.47590538175302516,
+      "learning_rate": 4.7372918809508756e-07,
+      "loss": 0.5211,
+      "step": 18115
+    },
+    {
+      "epoch": 4.809239346873755,
+      "grad_norm": 0.49302474112585337,
+      "learning_rate": 4.735247141657726e-07,
+      "loss": 0.532,
+      "step": 18116
+    },
+    {
+      "epoch": 4.809504845347139,
+      "grad_norm": 0.48734258933003977,
+      "learning_rate": 4.733202797576858e-07,
+      "loss": 0.567,
+      "step": 18117
+    },
+    {
+      "epoch": 4.809770343820523,
+      "grad_norm": 0.48143797998846577,
+      "learning_rate": 4.731158848748163e-07,
+      "loss": 0.5396,
+      "step": 18118
+    },
+    {
+      "epoch": 4.810035842293907,
+      "grad_norm": 0.4824228712274929,
+      "learning_rate": 4.729115295211492e-07,
+      "loss": 0.541,
+      "step": 18119
+    },
+    {
+      "epoch": 4.810301340767291,
+      "grad_norm": 0.4867457179269042,
+      "learning_rate": 4.727072137006714e-07,
+      "loss": 0.5172,
+      "step": 18120
+    },
+    {
+      "epoch": 4.810566839240675,
+      "grad_norm": 0.48854723419454854,
+      "learning_rate": 4.7250293741736563e-07,
+      "loss": 0.5345,
+      "step": 18121
+    },
+    {
+      "epoch": 4.810832337714058,
+      "grad_norm": 0.4768223803871091,
+      "learning_rate": 4.722987006752172e-07,
+      "loss": 0.5511,
+      "step": 18122
+    },
+    {
+      "epoch": 4.811097836187442,
+      "grad_norm": 0.49535996872987526,
+      "learning_rate": 4.72094503478209e-07,
+      "loss": 0.562,
+      "step": 18123
+    },
+    {
+      "epoch": 4.811363334660825,
+      "grad_norm": 0.48268802670533223,
+      "learning_rate": 4.718903458303226e-07,
+      "loss": 0.5443,
+      "step": 18124
+    },
+    {
+      "epoch": 4.811628833134209,
+      "grad_norm": 0.48018821427385866,
+      "learning_rate": 4.716862277355408e-07,
+      "loss": 0.5274,
+      "step": 18125
+    },
+    {
+      "epoch": 4.8118943316075935,
+      "grad_norm": 0.48088665841504524,
+      "learning_rate": 4.7148214919784275e-07,
+      "loss": 0.5124,
+      "step": 18126
+    },
+    {
+      "epoch": 4.812159830080977,
+      "grad_norm": 0.48176295696429045,
+      "learning_rate": 4.712781102212105e-07,
+      "loss": 0.543,
+      "step": 18127
+    },
+    {
+      "epoch": 4.812425328554361,
+      "grad_norm": 0.4982674844525797,
+      "learning_rate": 4.71074110809622e-07,
+      "loss": 0.513,
+      "step": 18128
+    },
+    {
+      "epoch": 4.812690827027745,
+      "grad_norm": 0.4677446549873634,
+      "learning_rate": 4.7087015096705603e-07,
+      "loss": 0.543,
+      "step": 18129
+    },
+    {
+      "epoch": 4.812956325501128,
+      "grad_norm": 0.4939002897779738,
+      "learning_rate": 4.706662306974899e-07,
+      "loss": 0.5991,
+      "step": 18130
+    },
+    {
+      "epoch": 4.813221823974512,
+      "grad_norm": 0.4853412190363089,
+      "learning_rate": 4.704623500049005e-07,
+      "loss": 0.5319,
+      "step": 18131
+    },
+    {
+      "epoch": 4.813487322447896,
+      "grad_norm": 0.49305192430154066,
+      "learning_rate": 4.702585088932654e-07,
+      "loss": 0.5475,
+      "step": 18132
+    },
+    {
+      "epoch": 4.813752820921279,
+      "grad_norm": 0.4815837904848243,
+      "learning_rate": 4.7005470736655846e-07,
+      "loss": 0.4847,
+      "step": 18133
+    },
+    {
+      "epoch": 4.8140183193946635,
+      "grad_norm": 0.4799775425045277,
+      "learning_rate": 4.698509454287559e-07,
+      "loss": 0.5442,
+      "step": 18134
+    },
+    {
+      "epoch": 4.814283817868048,
+      "grad_norm": 0.47966064458161495,
+      "learning_rate": 4.696472230838295e-07,
+      "loss": 0.5331,
+      "step": 18135
+    },
+    {
+      "epoch": 4.814549316341431,
+      "grad_norm": 0.4737009868692425,
+      "learning_rate": 4.694435403357539e-07,
+      "loss": 0.484,
+      "step": 18136
+    },
+    {
+      "epoch": 4.814814814814815,
+      "grad_norm": 0.4791416821208662,
+      "learning_rate": 4.692398971885004e-07,
+      "loss": 0.5481,
+      "step": 18137
+    },
+    {
+      "epoch": 4.815080313288199,
+      "grad_norm": 0.4732011119990352,
+      "learning_rate": 4.690362936460416e-07,
+      "loss": 0.5393,
+      "step": 18138
+    },
+    {
+      "epoch": 4.815345811761582,
+      "grad_norm": 0.4827608154763677,
+      "learning_rate": 4.6883272971234766e-07,
+      "loss": 0.5269,
+      "step": 18139
+    },
+    {
+      "epoch": 4.815611310234966,
+      "grad_norm": 0.47441625978358054,
+      "learning_rate": 4.686292053913882e-07,
+      "loss": 0.5322,
+      "step": 18140
+    },
+    {
+      "epoch": 4.815876808708349,
+      "grad_norm": 0.47989421335478794,
+      "learning_rate": 4.684257206871334e-07,
+      "loss": 0.4774,
+      "step": 18141
+    },
+    {
+      "epoch": 4.8161423071817335,
+      "grad_norm": 0.48865268856326216,
+      "learning_rate": 4.6822227560355067e-07,
+      "loss": 0.5489,
+      "step": 18142
+    },
+    {
+      "epoch": 4.816407805655118,
+      "grad_norm": 0.47886811139050633,
+      "learning_rate": 4.6801887014460897e-07,
+      "loss": 0.5264,
+      "step": 18143
+    },
+    {
+      "epoch": 4.816673304128502,
+      "grad_norm": 0.4683115566017144,
+      "learning_rate": 4.678155043142735e-07,
+      "loss": 0.4857,
+      "step": 18144
+    },
+    {
+      "epoch": 4.816938802601885,
+      "grad_norm": 0.48007986453529333,
+      "learning_rate": 4.676121781165127e-07,
+      "loss": 0.5118,
+      "step": 18145
+    },
+    {
+      "epoch": 4.817204301075269,
+      "grad_norm": 0.47362498224390864,
+      "learning_rate": 4.6740889155528936e-07,
+      "loss": 0.5381,
+      "step": 18146
+    },
+    {
+      "epoch": 4.817469799548652,
+      "grad_norm": 0.4824805111585414,
+      "learning_rate": 4.672056446345699e-07,
+      "loss": 0.5145,
+      "step": 18147
+    },
+    {
+      "epoch": 4.817735298022036,
+      "grad_norm": 0.5045695610769937,
+      "learning_rate": 4.6700243735831705e-07,
+      "loss": 0.5677,
+      "step": 18148
+    },
+    {
+      "epoch": 4.81800079649542,
+      "grad_norm": 0.4862918137323932,
+      "learning_rate": 4.6679926973049463e-07,
+      "loss": 0.5316,
+      "step": 18149
+    },
+    {
+      "epoch": 4.8182662949688035,
+      "grad_norm": 0.47055863785523566,
+      "learning_rate": 4.6659614175506406e-07,
+      "loss": 0.525,
+      "step": 18150
+    },
+    {
+      "epoch": 4.818531793442188,
+      "grad_norm": 0.4855345887166264,
+      "learning_rate": 4.6639305343598813e-07,
+      "loss": 0.5414,
+      "step": 18151
+    },
+    {
+      "epoch": 4.818797291915572,
+      "grad_norm": 0.47851235252749896,
+      "learning_rate": 4.6619000477722667e-07,
+      "loss": 0.5441,
+      "step": 18152
+    },
+    {
+      "epoch": 4.819062790388955,
+      "grad_norm": 0.4824769438137112,
+      "learning_rate": 4.65986995782739e-07,
+      "loss": 0.5338,
+      "step": 18153
+    },
+    {
+      "epoch": 4.819328288862339,
+      "grad_norm": 0.4837672136707721,
+      "learning_rate": 4.6578402645648583e-07,
+      "loss": 0.5603,
+      "step": 18154
+    },
+    {
+      "epoch": 4.819593787335723,
+      "grad_norm": 0.4899059287755638,
+      "learning_rate": 4.655810968024241e-07,
+      "loss": 0.5138,
+      "step": 18155
+    },
+    {
+      "epoch": 4.819859285809106,
+      "grad_norm": 0.46850423650454737,
+      "learning_rate": 4.6537820682451273e-07,
+      "loss": 0.4995,
+      "step": 18156
+    },
+    {
+      "epoch": 4.82012478428249,
+      "grad_norm": 0.48479285382385445,
+      "learning_rate": 4.6517535652670783e-07,
+      "loss": 0.5392,
+      "step": 18157
+    },
+    {
+      "epoch": 4.8203902827558744,
+      "grad_norm": 0.4790650207220103,
+      "learning_rate": 4.64972545912965e-07,
+      "loss": 0.5581,
+      "step": 18158
+    },
+    {
+      "epoch": 4.820655781229258,
+      "grad_norm": 0.4836268539881204,
+      "learning_rate": 4.6476977498724096e-07,
+      "loss": 0.5591,
+      "step": 18159
+    },
+    {
+      "epoch": 4.820921279702642,
+      "grad_norm": 0.47822459089724667,
+      "learning_rate": 4.6456704375348854e-07,
+      "loss": 0.5117,
+      "step": 18160
+    },
+    {
+      "epoch": 4.821186778176026,
+      "grad_norm": 0.4699591622068747,
+      "learning_rate": 4.6436435221566324e-07,
+      "loss": 0.5097,
+      "step": 18161
+    },
+    {
+      "epoch": 4.821452276649409,
+      "grad_norm": 0.48198029505544215,
+      "learning_rate": 4.6416170037771684e-07,
+      "loss": 0.512,
+      "step": 18162
+    },
+    {
+      "epoch": 4.821717775122793,
+      "grad_norm": 0.47489210279638394,
+      "learning_rate": 4.639590882436021e-07,
+      "loss": 0.5431,
+      "step": 18163
+    },
+    {
+      "epoch": 4.821983273596177,
+      "grad_norm": 0.4809508372540027,
+      "learning_rate": 4.6375651581726967e-07,
+      "loss": 0.5129,
+      "step": 18164
+    },
+    {
+      "epoch": 4.82224877206956,
+      "grad_norm": 0.4776208445584323,
+      "learning_rate": 4.6355398310267113e-07,
+      "loss": 0.543,
+      "step": 18165
+    },
+    {
+      "epoch": 4.8225142705429445,
+      "grad_norm": 0.4843297757247386,
+      "learning_rate": 4.633514901037556e-07,
+      "loss": 0.5362,
+      "step": 18166
+    },
+    {
+      "epoch": 4.8227797690163285,
+      "grad_norm": 0.4824303225651672,
+      "learning_rate": 4.6314903682447314e-07,
+      "loss": 0.5366,
+      "step": 18167
+    },
+    {
+      "epoch": 4.823045267489712,
+      "grad_norm": 0.48605894939820843,
+      "learning_rate": 4.6294662326877153e-07,
+      "loss": 0.5469,
+      "step": 18168
+    },
+    {
+      "epoch": 4.823310765963096,
+      "grad_norm": 0.48139663113332626,
+      "learning_rate": 4.627442494405976e-07,
+      "loss": 0.5571,
+      "step": 18169
+    },
+    {
+      "epoch": 4.823576264436479,
+      "grad_norm": 0.4927290232874083,
+      "learning_rate": 4.625419153438998e-07,
+      "loss": 0.5409,
+      "step": 18170
+    },
+    {
+      "epoch": 4.823841762909863,
+      "grad_norm": 0.4734861802890927,
+      "learning_rate": 4.6233962098262227e-07,
+      "loss": 0.53,
+      "step": 18171
+    },
+    {
+      "epoch": 4.824107261383247,
+      "grad_norm": 0.47913699765607326,
+      "learning_rate": 4.6213736636071206e-07,
+      "loss": 0.5605,
+      "step": 18172
+    },
+    {
+      "epoch": 4.824372759856631,
+      "grad_norm": 0.4847538285866058,
+      "learning_rate": 4.619351514821119e-07,
+      "loss": 0.5456,
+      "step": 18173
+    },
+    {
+      "epoch": 4.8246382583300145,
+      "grad_norm": 0.47324927929082805,
+      "learning_rate": 4.6173297635076777e-07,
+      "loss": 0.5267,
+      "step": 18174
+    },
+    {
+      "epoch": 4.824903756803399,
+      "grad_norm": 0.4770241092552351,
+      "learning_rate": 4.6153084097061956e-07,
+      "loss": 0.546,
+      "step": 18175
+    },
+    {
+      "epoch": 4.825169255276782,
+      "grad_norm": 0.48818549571127556,
+      "learning_rate": 4.613287453456111e-07,
+      "loss": 0.5433,
+      "step": 18176
+    },
+    {
+      "epoch": 4.825434753750166,
+      "grad_norm": 0.48018850705115834,
+      "learning_rate": 4.611266894796845e-07,
+      "loss": 0.5441,
+      "step": 18177
+    },
+    {
+      "epoch": 4.82570025222355,
+      "grad_norm": 0.4684016441834493,
+      "learning_rate": 4.609246733767786e-07,
+      "loss": 0.5234,
+      "step": 18178
+    },
+    {
+      "epoch": 4.825965750696933,
+      "grad_norm": 0.47033956474640726,
+      "learning_rate": 4.607226970408349e-07,
+      "loss": 0.5131,
+      "step": 18179
+    },
+    {
+      "epoch": 4.826231249170317,
+      "grad_norm": 0.48658010291152065,
+      "learning_rate": 4.6052076047579164e-07,
+      "loss": 0.5303,
+      "step": 18180
+    },
+    {
+      "epoch": 4.826496747643701,
+      "grad_norm": 0.48793584579626553,
+      "learning_rate": 4.603188636855871e-07,
+      "loss": 0.5437,
+      "step": 18181
+    },
+    {
+      "epoch": 4.8267622461170845,
+      "grad_norm": 0.47445806267992957,
+      "learning_rate": 4.601170066741578e-07,
+      "loss": 0.5219,
+      "step": 18182
+    },
+    {
+      "epoch": 4.827027744590469,
+      "grad_norm": 0.4749005446564915,
+      "learning_rate": 4.5991518944544205e-07,
+      "loss": 0.5219,
+      "step": 18183
+    },
+    {
+      "epoch": 4.827293243063853,
+      "grad_norm": 0.4817087833539402,
+      "learning_rate": 4.597134120033747e-07,
+      "loss": 0.5262,
+      "step": 18184
+    },
+    {
+      "epoch": 4.827558741537236,
+      "grad_norm": 0.4760468240128116,
+      "learning_rate": 4.595116743518918e-07,
+      "loss": 0.5211,
+      "step": 18185
+    },
+    {
+      "epoch": 4.82782424001062,
+      "grad_norm": 0.4866302088437665,
+      "learning_rate": 4.5930997649492714e-07,
+      "loss": 0.5233,
+      "step": 18186
+    },
+    {
+      "epoch": 4.828089738484004,
+      "grad_norm": 0.49661268674400105,
+      "learning_rate": 4.5910831843641395e-07,
+      "loss": 0.5282,
+      "step": 18187
+    },
+    {
+      "epoch": 4.828355236957387,
+      "grad_norm": 0.48557770498767167,
+      "learning_rate": 4.5890670018028606e-07,
+      "loss": 0.5426,
+      "step": 18188
+    },
+    {
+      "epoch": 4.828620735430771,
+      "grad_norm": 0.4830864070788451,
+      "learning_rate": 4.5870512173047443e-07,
+      "loss": 0.5402,
+      "step": 18189
+    },
+    {
+      "epoch": 4.828886233904155,
+      "grad_norm": 0.49150802696870766,
+      "learning_rate": 4.5850358309091126e-07,
+      "loss": 0.5529,
+      "step": 18190
+    },
+    {
+      "epoch": 4.829151732377539,
+      "grad_norm": 0.4957062608003524,
+      "learning_rate": 4.5830208426552695e-07,
+      "loss": 0.5399,
+      "step": 18191
+    },
+    {
+      "epoch": 4.829417230850923,
+      "grad_norm": 0.48304110841302,
+      "learning_rate": 4.5810062525825065e-07,
+      "loss": 0.5359,
+      "step": 18192
+    },
+    {
+      "epoch": 4.829682729324307,
+      "grad_norm": 0.4891686879486978,
+      "learning_rate": 4.5789920607301114e-07,
+      "loss": 0.5316,
+      "step": 18193
+    },
+    {
+      "epoch": 4.82994822779769,
+      "grad_norm": 0.47356423400101555,
+      "learning_rate": 4.5769782671373773e-07,
+      "loss": 0.5076,
+      "step": 18194
+    },
+    {
+      "epoch": 4.830213726271074,
+      "grad_norm": 0.47515981423606446,
+      "learning_rate": 4.574964871843562e-07,
+      "loss": 0.5132,
+      "step": 18195
+    },
+    {
+      "epoch": 4.830479224744458,
+      "grad_norm": 0.4750746730522271,
+      "learning_rate": 4.572951874887943e-07,
+      "loss": 0.5486,
+      "step": 18196
+    },
+    {
+      "epoch": 4.830744723217841,
+      "grad_norm": 0.49365837543658897,
+      "learning_rate": 4.570939276309791e-07,
+      "loss": 0.5379,
+      "step": 18197
+    },
+    {
+      "epoch": 4.831010221691225,
+      "grad_norm": 0.4878811430545162,
+      "learning_rate": 4.568927076148327e-07,
+      "loss": 0.5298,
+      "step": 18198
+    },
+    {
+      "epoch": 4.831275720164609,
+      "grad_norm": 0.4853627979123752,
+      "learning_rate": 4.566915274442815e-07,
+      "loss": 0.53,
+      "step": 18199
+    },
+    {
+      "epoch": 4.831541218637993,
+      "grad_norm": 0.46725238031499355,
+      "learning_rate": 4.564903871232479e-07,
+      "loss": 0.5162,
+      "step": 18200
+    },
+    {
+      "epoch": 4.831806717111377,
+      "grad_norm": 0.48373388527294836,
+      "learning_rate": 4.562892866556556e-07,
+      "loss": 0.5325,
+      "step": 18201
+    },
+    {
+      "epoch": 4.832072215584761,
+      "grad_norm": 0.4724450341628017,
+      "learning_rate": 4.560882260454255e-07,
+      "loss": 0.5185,
+      "step": 18202
+    },
+    {
+      "epoch": 4.832337714058144,
+      "grad_norm": 0.4849813928971904,
+      "learning_rate": 4.5588720529647997e-07,
+      "loss": 0.5363,
+      "step": 18203
+    },
+    {
+      "epoch": 4.832603212531528,
+      "grad_norm": 0.4913263352483113,
+      "learning_rate": 4.5568622441273887e-07,
+      "loss": 0.5062,
+      "step": 18204
+    },
+    {
+      "epoch": 4.832868711004911,
+      "grad_norm": 0.4870198405355169,
+      "learning_rate": 4.5548528339812105e-07,
+      "loss": 0.5378,
+      "step": 18205
+    },
+    {
+      "epoch": 4.833134209478295,
+      "grad_norm": 0.48662564170345235,
+      "learning_rate": 4.552843822565467e-07,
+      "loss": 0.5668,
+      "step": 18206
+    },
+    {
+      "epoch": 4.8333997079516795,
+      "grad_norm": 0.46700308405032365,
+      "learning_rate": 4.550835209919327e-07,
+      "loss": 0.5194,
+      "step": 18207
+    },
+    {
+      "epoch": 4.833665206425063,
+      "grad_norm": 0.48875253438239225,
+      "learning_rate": 4.548826996081973e-07,
+      "loss": 0.544,
+      "step": 18208
+    },
+    {
+      "epoch": 4.833930704898447,
+      "grad_norm": 0.49350496050109544,
+      "learning_rate": 4.5468191810925657e-07,
+      "loss": 0.4957,
+      "step": 18209
+    },
+    {
+      "epoch": 4.834196203371831,
+      "grad_norm": 0.4853233057802422,
+      "learning_rate": 4.5448117649902645e-07,
+      "loss": 0.5202,
+      "step": 18210
+    },
+    {
+      "epoch": 4.834461701845214,
+      "grad_norm": 0.4788494891193707,
+      "learning_rate": 4.5428047478142086e-07,
+      "loss": 0.5495,
+      "step": 18211
+    },
+    {
+      "epoch": 4.834727200318598,
+      "grad_norm": 0.4712102044167938,
+      "learning_rate": 4.540798129603555e-07,
+      "loss": 0.4849,
+      "step": 18212
+    },
+    {
+      "epoch": 4.834992698791982,
+      "grad_norm": 0.49614897070249664,
+      "learning_rate": 4.538791910397422e-07,
+      "loss": 0.5137,
+      "step": 18213
+    },
+    {
+      "epoch": 4.835258197265365,
+      "grad_norm": 0.4921753917728957,
+      "learning_rate": 4.5367860902349543e-07,
+      "loss": 0.5471,
+      "step": 18214
+    },
+    {
+      "epoch": 4.8355236957387495,
+      "grad_norm": 0.48425578510103867,
+      "learning_rate": 4.5347806691552586e-07,
+      "loss": 0.5448,
+      "step": 18215
+    },
+    {
+      "epoch": 4.835789194212134,
+      "grad_norm": 0.49339433617640477,
+      "learning_rate": 4.5327756471974406e-07,
+      "loss": 0.5368,
+      "step": 18216
+    },
+    {
+      "epoch": 4.836054692685517,
+      "grad_norm": 0.49204854645679075,
+      "learning_rate": 4.5307710244006183e-07,
+      "loss": 0.5206,
+      "step": 18217
+    },
+    {
+      "epoch": 4.836320191158901,
+      "grad_norm": 0.48904457464597656,
+      "learning_rate": 4.5287668008038686e-07,
+      "loss": 0.5722,
+      "step": 18218
+    },
+    {
+      "epoch": 4.836585689632285,
+      "grad_norm": 0.4871505466359667,
+      "learning_rate": 4.5267629764462973e-07,
+      "loss": 0.5415,
+      "step": 18219
+    },
+    {
+      "epoch": 4.836851188105668,
+      "grad_norm": 0.4803969649868923,
+      "learning_rate": 4.524759551366975e-07,
+      "loss": 0.5432,
+      "step": 18220
+    },
+    {
+      "epoch": 4.837116686579052,
+      "grad_norm": 0.46756671668052974,
+      "learning_rate": 4.522756525604968e-07,
+      "loss": 0.5456,
+      "step": 18221
+    },
+    {
+      "epoch": 4.837382185052436,
+      "grad_norm": 0.4832044148201414,
+      "learning_rate": 4.520753899199351e-07,
+      "loss": 0.5396,
+      "step": 18222
+    },
+    {
+      "epoch": 4.8376476835258195,
+      "grad_norm": 0.4796197062455762,
+      "learning_rate": 4.518751672189167e-07,
+      "loss": 0.5299,
+      "step": 18223
+    },
+    {
+      "epoch": 4.837913181999204,
+      "grad_norm": 0.48019675896819974,
+      "learning_rate": 4.51674984461348e-07,
+      "loss": 0.5643,
+      "step": 18224
+    },
+    {
+      "epoch": 4.838178680472588,
+      "grad_norm": 0.4947748731846147,
+      "learning_rate": 4.5147484165113163e-07,
+      "loss": 0.5403,
+      "step": 18225
+    },
+    {
+      "epoch": 4.838444178945971,
+      "grad_norm": 0.4802050819031008,
+      "learning_rate": 4.5127473879217287e-07,
+      "loss": 0.5214,
+      "step": 18226
+    },
+    {
+      "epoch": 4.838709677419355,
+      "grad_norm": 0.4899538452008818,
+      "learning_rate": 4.510746758883713e-07,
+      "loss": 0.521,
+      "step": 18227
+    },
+    {
+      "epoch": 4.838975175892738,
+      "grad_norm": 0.49285298244474074,
+      "learning_rate": 4.508746529436311e-07,
+      "loss": 0.5003,
+      "step": 18228
+    },
+    {
+      "epoch": 4.839240674366122,
+      "grad_norm": 0.4929044803721831,
+      "learning_rate": 4.506746699618514e-07,
+      "loss": 0.5715,
+      "step": 18229
+    },
+    {
+      "epoch": 4.839506172839506,
+      "grad_norm": 0.4745155841416195,
+      "learning_rate": 4.504747269469342e-07,
+      "loss": 0.5688,
+      "step": 18230
+    },
+    {
+      "epoch": 4.83977167131289,
+      "grad_norm": 0.46442873115919686,
+      "learning_rate": 4.502748239027771e-07,
+      "loss": 0.5297,
+      "step": 18231
+    },
+    {
+      "epoch": 4.840037169786274,
+      "grad_norm": 0.48352127910570664,
+      "learning_rate": 4.500749608332802e-07,
+      "loss": 0.5173,
+      "step": 18232
+    },
+    {
+      "epoch": 4.840302668259658,
+      "grad_norm": 0.49404479596608086,
+      "learning_rate": 4.4987513774234046e-07,
+      "loss": 0.5419,
+      "step": 18233
+    },
+    {
+      "epoch": 4.840568166733041,
+      "grad_norm": 0.47963627401077424,
+      "learning_rate": 4.4967535463385473e-07,
+      "loss": 0.5253,
+      "step": 18234
+    },
+    {
+      "epoch": 4.840833665206425,
+      "grad_norm": 0.4885439922256354,
+      "learning_rate": 4.494756115117202e-07,
+      "loss": 0.5473,
+      "step": 18235
+    },
+    {
+      "epoch": 4.841099163679809,
+      "grad_norm": 0.4849336342702325,
+      "learning_rate": 4.492759083798312e-07,
+      "loss": 0.5374,
+      "step": 18236
+    },
+    {
+      "epoch": 4.841364662153192,
+      "grad_norm": 0.4916676254833159,
+      "learning_rate": 4.490762452420841e-07,
+      "loss": 0.518,
+      "step": 18237
+    },
+    {
+      "epoch": 4.841630160626576,
+      "grad_norm": 0.4886490040434697,
+      "learning_rate": 4.4887662210237077e-07,
+      "loss": 0.5308,
+      "step": 18238
+    },
+    {
+      "epoch": 4.8418956590999604,
+      "grad_norm": 0.49726314273208444,
+      "learning_rate": 4.486770389645856e-07,
+      "loss": 0.5349,
+      "step": 18239
+    },
+    {
+      "epoch": 4.842161157573344,
+      "grad_norm": 0.47590246318515445,
+      "learning_rate": 4.484774958326202e-07,
+      "loss": 0.5099,
+      "step": 18240
+    },
+    {
+      "epoch": 4.842426656046728,
+      "grad_norm": 0.49860990709074327,
+      "learning_rate": 4.4827799271036726e-07,
+      "loss": 0.5487,
+      "step": 18241
+    },
+    {
+      "epoch": 4.842692154520112,
+      "grad_norm": 0.5076029422708265,
+      "learning_rate": 4.480785296017165e-07,
+      "loss": 0.5698,
+      "step": 18242
+    },
+    {
+      "epoch": 4.842957652993495,
+      "grad_norm": 0.4898359385744992,
+      "learning_rate": 4.4787910651055895e-07,
+      "loss": 0.5417,
+      "step": 18243
+    },
+    {
+      "epoch": 4.843223151466879,
+      "grad_norm": 0.47424842287744257,
+      "learning_rate": 4.4767972344078313e-07,
+      "loss": 0.501,
+      "step": 18244
+    },
+    {
+      "epoch": 4.843488649940263,
+      "grad_norm": 0.4814491146193889,
+      "learning_rate": 4.474803803962771e-07,
+      "loss": 0.5405,
+      "step": 18245
+    },
+    {
+      "epoch": 4.843754148413646,
+      "grad_norm": 0.4996965243954745,
+      "learning_rate": 4.472810773809297e-07,
+      "loss": 0.542,
+      "step": 18246
+    },
+    {
+      "epoch": 4.8440196468870305,
+      "grad_norm": 0.48664894160352984,
+      "learning_rate": 4.4708181439862664e-07,
+      "loss": 0.5078,
+      "step": 18247
+    },
+    {
+      "epoch": 4.8442851453604145,
+      "grad_norm": 0.49326884127997844,
+      "learning_rate": 4.4688259145325513e-07,
+      "loss": 0.5544,
+      "step": 18248
+    },
+    {
+      "epoch": 4.844550643833798,
+      "grad_norm": 0.4839033103331838,
+      "learning_rate": 4.466834085487001e-07,
+      "loss": 0.5566,
+      "step": 18249
+    },
+    {
+      "epoch": 4.844816142307182,
+      "grad_norm": 0.4892763481700645,
+      "learning_rate": 4.464842656888452e-07,
+      "loss": 0.5392,
+      "step": 18250
+    },
+    {
+      "epoch": 4.845081640780565,
+      "grad_norm": 0.4912769114930382,
+      "learning_rate": 4.4628516287757577e-07,
+      "loss": 0.5673,
+      "step": 18251
+    },
+    {
+      "epoch": 4.845347139253949,
+      "grad_norm": 0.48215085914139333,
+      "learning_rate": 4.4608610011877327e-07,
+      "loss": 0.5502,
+      "step": 18252
+    },
+    {
+      "epoch": 4.845612637727333,
+      "grad_norm": 0.46597944576406264,
+      "learning_rate": 4.4588707741632153e-07,
+      "loss": 0.544,
+      "step": 18253
+    },
+    {
+      "epoch": 4.845878136200717,
+      "grad_norm": 0.4766857587248487,
+      "learning_rate": 4.4568809477410017e-07,
+      "loss": 0.5166,
+      "step": 18254
+    },
+    {
+      "epoch": 4.8461436346741005,
+      "grad_norm": 0.4690858602459936,
+      "learning_rate": 4.4548915219599225e-07,
+      "loss": 0.5288,
+      "step": 18255
+    },
+    {
+      "epoch": 4.846409133147485,
+      "grad_norm": 0.4774485275070768,
+      "learning_rate": 4.452902496858749e-07,
+      "loss": 0.525,
+      "step": 18256
+    },
+    {
+      "epoch": 4.846674631620868,
+      "grad_norm": 0.4879781403870331,
+      "learning_rate": 4.4509138724762894e-07,
+      "loss": 0.5589,
+      "step": 18257
+    },
+    {
+      "epoch": 4.846940130094252,
+      "grad_norm": 0.5055840146162429,
+      "learning_rate": 4.4489256488513185e-07,
+      "loss": 0.5181,
+      "step": 18258
+    },
+    {
+      "epoch": 4.847205628567636,
+      "grad_norm": 0.4784207536330953,
+      "learning_rate": 4.446937826022621e-07,
+      "loss": 0.5294,
+      "step": 18259
+    },
+    {
+      "epoch": 4.84747112704102,
+      "grad_norm": 0.46506180789815005,
+      "learning_rate": 4.4449504040289556e-07,
+      "loss": 0.5426,
+      "step": 18260
+    },
+    {
+      "epoch": 4.847736625514403,
+      "grad_norm": 0.49062190296753627,
+      "learning_rate": 4.4429633829090826e-07,
+      "loss": 0.55,
+      "step": 18261
+    },
+    {
+      "epoch": 4.848002123987787,
+      "grad_norm": 0.4867807625462896,
+      "learning_rate": 4.44097676270176e-07,
+      "loss": 0.5539,
+      "step": 18262
+    },
+    {
+      "epoch": 4.8482676224611705,
+      "grad_norm": 0.4784792603668485,
+      "learning_rate": 4.4389905434457206e-07,
+      "loss": 0.5113,
+      "step": 18263
+    },
+    {
+      "epoch": 4.848533120934555,
+      "grad_norm": 0.48150939191993825,
+      "learning_rate": 4.437004725179714e-07,
+      "loss": 0.5505,
+      "step": 18264
+    },
+    {
+      "epoch": 4.848798619407939,
+      "grad_norm": 0.4807576879116201,
+      "learning_rate": 4.435019307942456e-07,
+      "loss": 0.535,
+      "step": 18265
+    },
+    {
+      "epoch": 4.849064117881322,
+      "grad_norm": 0.47467386767008,
+      "learning_rate": 4.4330342917726865e-07,
+      "loss": 0.5425,
+      "step": 18266
+    },
+    {
+      "epoch": 4.849329616354706,
+      "grad_norm": 0.47139180136038833,
+      "learning_rate": 4.431049676709093e-07,
+      "loss": 0.5323,
+      "step": 18267
+    },
+    {
+      "epoch": 4.84959511482809,
+      "grad_norm": 0.47395333030386544,
+      "learning_rate": 4.4290654627903913e-07,
+      "loss": 0.5135,
+      "step": 18268
+    },
+    {
+      "epoch": 4.849860613301473,
+      "grad_norm": 0.48729254602856314,
+      "learning_rate": 4.4270816500552873e-07,
+      "loss": 0.5611,
+      "step": 18269
+    },
+    {
+      "epoch": 4.850126111774857,
+      "grad_norm": 0.4905670327092857,
+      "learning_rate": 4.425098238542455e-07,
+      "loss": 0.5468,
+      "step": 18270
+    },
+    {
+      "epoch": 4.850391610248241,
+      "grad_norm": 0.4942800627082329,
+      "learning_rate": 4.423115228290589e-07,
+      "loss": 0.5341,
+      "step": 18271
+    },
+    {
+      "epoch": 4.850657108721625,
+      "grad_norm": 0.48776186386997894,
+      "learning_rate": 4.4211326193383556e-07,
+      "loss": 0.5501,
+      "step": 18272
+    },
+    {
+      "epoch": 4.850922607195009,
+      "grad_norm": 0.46830835397990495,
+      "learning_rate": 4.419150411724424e-07,
+      "loss": 0.5149,
+      "step": 18273
+    },
+    {
+      "epoch": 4.851188105668393,
+      "grad_norm": 0.4781424274544481,
+      "learning_rate": 4.417168605487443e-07,
+      "loss": 0.558,
+      "step": 18274
+    },
+    {
+      "epoch": 4.851453604141776,
+      "grad_norm": 0.47903613914742793,
+      "learning_rate": 4.4151872006660767e-07,
+      "loss": 0.5262,
+      "step": 18275
+    },
+    {
+      "epoch": 4.85171910261516,
+      "grad_norm": 0.46715317502719883,
+      "learning_rate": 4.413206197298953e-07,
+      "loss": 0.5515,
+      "step": 18276
+    },
+    {
+      "epoch": 4.851984601088544,
+      "grad_norm": 0.48665928929664454,
+      "learning_rate": 4.41122559542472e-07,
+      "loss": 0.5343,
+      "step": 18277
+    },
+    {
+      "epoch": 4.852250099561927,
+      "grad_norm": 0.49341714990139246,
+      "learning_rate": 4.4092453950819987e-07,
+      "loss": 0.5337,
+      "step": 18278
+    },
+    {
+      "epoch": 4.852515598035311,
+      "grad_norm": 0.48156720776399903,
+      "learning_rate": 4.407265596309399e-07,
+      "loss": 0.5337,
+      "step": 18279
+    },
+    {
+      "epoch": 4.852781096508695,
+      "grad_norm": 0.49431267943017476,
+      "learning_rate": 4.4052861991455477e-07,
+      "loss": 0.5211,
+      "step": 18280
+    },
+    {
+      "epoch": 4.853046594982079,
+      "grad_norm": 0.47494032367250116,
+      "learning_rate": 4.403307203629034e-07,
+      "loss": 0.5305,
+      "step": 18281
+    },
+    {
+      "epoch": 4.853312093455463,
+      "grad_norm": 0.4669439687539629,
+      "learning_rate": 4.4013286097984635e-07,
+      "loss": 0.4886,
+      "step": 18282
+    },
+    {
+      "epoch": 4.853577591928847,
+      "grad_norm": 0.4659243942496253,
+      "learning_rate": 4.399350417692422e-07,
+      "loss": 0.5417,
+      "step": 18283
+    },
+    {
+      "epoch": 4.85384309040223,
+      "grad_norm": 0.4699833696960174,
+      "learning_rate": 4.397372627349486e-07,
+      "loss": 0.5608,
+      "step": 18284
+    },
+    {
+      "epoch": 4.854108588875614,
+      "grad_norm": 0.48611024280735016,
+      "learning_rate": 4.39539523880822e-07,
+      "loss": 0.5893,
+      "step": 18285
+    },
+    {
+      "epoch": 4.854374087348997,
+      "grad_norm": 0.4727216185859443,
+      "learning_rate": 4.393418252107201e-07,
+      "loss": 0.5746,
+      "step": 18286
+    },
+    {
+      "epoch": 4.854639585822381,
+      "grad_norm": 0.4825396894258424,
+      "learning_rate": 4.3914416672849765e-07,
+      "loss": 0.5305,
+      "step": 18287
+    },
+    {
+      "epoch": 4.8549050842957655,
+      "grad_norm": 0.4631118347073883,
+      "learning_rate": 4.3894654843800955e-07,
+      "loss": 0.5195,
+      "step": 18288
+    },
+    {
+      "epoch": 4.855170582769149,
+      "grad_norm": 0.4815487419035392,
+      "learning_rate": 4.3874897034311135e-07,
+      "loss": 0.5249,
+      "step": 18289
+    },
+    {
+      "epoch": 4.855436081242533,
+      "grad_norm": 0.47901427907153116,
+      "learning_rate": 4.385514324476539e-07,
+      "loss": 0.5306,
+      "step": 18290
+    },
+    {
+      "epoch": 4.855701579715917,
+      "grad_norm": 0.4743473026236478,
+      "learning_rate": 4.383539347554916e-07,
+      "loss": 0.5406,
+      "step": 18291
+    },
+    {
+      "epoch": 4.8559670781893,
+      "grad_norm": 0.47672892382941806,
+      "learning_rate": 4.381564772704744e-07,
+      "loss": 0.5109,
+      "step": 18292
+    },
+    {
+      "epoch": 4.856232576662684,
+      "grad_norm": 0.4703273562898987,
+      "learning_rate": 4.37959059996455e-07,
+      "loss": 0.5247,
+      "step": 18293
+    },
+    {
+      "epoch": 4.856498075136068,
+      "grad_norm": 0.5000011467933769,
+      "learning_rate": 4.377616829372819e-07,
+      "loss": 0.5286,
+      "step": 18294
+    },
+    {
+      "epoch": 4.856763573609451,
+      "grad_norm": 0.47903690460561604,
+      "learning_rate": 4.3756434609680596e-07,
+      "loss": 0.5193,
+      "step": 18295
+    },
+    {
+      "epoch": 4.8570290720828355,
+      "grad_norm": 0.4900974222353822,
+      "learning_rate": 4.3736704947887484e-07,
+      "loss": 0.5354,
+      "step": 18296
+    },
+    {
+      "epoch": 4.85729457055622,
+      "grad_norm": 0.48754374634206304,
+      "learning_rate": 4.3716979308733585e-07,
+      "loss": 0.547,
+      "step": 18297
+    },
+    {
+      "epoch": 4.857560069029603,
+      "grad_norm": 0.49183816952895903,
+      "learning_rate": 4.3697257692603746e-07,
+      "loss": 0.5823,
+      "step": 18298
+    },
+    {
+      "epoch": 4.857825567502987,
+      "grad_norm": 0.4935093666861926,
+      "learning_rate": 4.3677540099882396e-07,
+      "loss": 0.574,
+      "step": 18299
+    },
+    {
+      "epoch": 4.858091065976371,
+      "grad_norm": 0.4753492178349335,
+      "learning_rate": 4.3657826530954265e-07,
+      "loss": 0.5241,
+      "step": 18300
+    },
+    {
+      "epoch": 4.858356564449754,
+      "grad_norm": 0.4762913084357427,
+      "learning_rate": 4.363811698620374e-07,
+      "loss": 0.508,
+      "step": 18301
+    },
+    {
+      "epoch": 4.858622062923138,
+      "grad_norm": 0.48661581907854756,
+      "learning_rate": 4.361841146601517e-07,
+      "loss": 0.5338,
+      "step": 18302
+    },
+    {
+      "epoch": 4.858887561396522,
+      "grad_norm": 0.49258432449991685,
+      "learning_rate": 4.359870997077284e-07,
+      "loss": 0.5125,
+      "step": 18303
+    },
+    {
+      "epoch": 4.8591530598699055,
+      "grad_norm": 0.47835515930088757,
+      "learning_rate": 4.3579012500861077e-07,
+      "loss": 0.516,
+      "step": 18304
+    },
+    {
+      "epoch": 4.85941855834329,
+      "grad_norm": 0.46401972363365607,
+      "learning_rate": 4.3559319056663913e-07,
+      "loss": 0.5002,
+      "step": 18305
+    },
+    {
+      "epoch": 4.859684056816674,
+      "grad_norm": 0.4660094628200897,
+      "learning_rate": 4.3539629638565564e-07,
+      "loss": 0.4862,
+      "step": 18306
+    },
+    {
+      "epoch": 4.859949555290057,
+      "grad_norm": 0.4744740975242065,
+      "learning_rate": 4.3519944246949916e-07,
+      "loss": 0.5436,
+      "step": 18307
+    },
+    {
+      "epoch": 4.860215053763441,
+      "grad_norm": 0.49298669598264994,
+      "learning_rate": 4.350026288220083e-07,
+      "loss": 0.5731,
+      "step": 18308
+    },
+    {
+      "epoch": 4.860480552236824,
+      "grad_norm": 0.4859497351426526,
+      "learning_rate": 4.3480585544702303e-07,
+      "loss": 0.5504,
+      "step": 18309
+    },
+    {
+      "epoch": 4.860746050710208,
+      "grad_norm": 0.48817212132266324,
+      "learning_rate": 4.346091223483795e-07,
+      "loss": 0.5367,
+      "step": 18310
+    },
+    {
+      "epoch": 4.861011549183592,
+      "grad_norm": 0.4874449910658859,
+      "learning_rate": 4.344124295299154e-07,
+      "loss": 0.5873,
+      "step": 18311
+    },
+    {
+      "epoch": 4.861277047656976,
+      "grad_norm": 0.46369317067882787,
+      "learning_rate": 4.3421577699546664e-07,
+      "loss": 0.4988,
+      "step": 18312
+    },
+    {
+      "epoch": 4.86154254613036,
+      "grad_norm": 0.47484947910381026,
+      "learning_rate": 4.340191647488673e-07,
+      "loss": 0.5415,
+      "step": 18313
+    },
+    {
+      "epoch": 4.861808044603744,
+      "grad_norm": 0.48757078264101217,
+      "learning_rate": 4.3382259279395324e-07,
+      "loss": 0.5462,
+      "step": 18314
+    },
+    {
+      "epoch": 4.862073543077127,
+      "grad_norm": 0.46874855865464987,
+      "learning_rate": 4.336260611345569e-07,
+      "loss": 0.538,
+      "step": 18315
+    },
+    {
+      "epoch": 4.862339041550511,
+      "grad_norm": 0.48526604921869676,
+      "learning_rate": 4.334295697745125e-07,
+      "loss": 0.5396,
+      "step": 18316
+    },
+    {
+      "epoch": 4.862604540023895,
+      "grad_norm": 0.47850374365919873,
+      "learning_rate": 4.332331187176506e-07,
+      "loss": 0.5512,
+      "step": 18317
+    },
+    {
+      "epoch": 4.862870038497278,
+      "grad_norm": 0.483379625297864,
+      "learning_rate": 4.3303670796780437e-07,
+      "loss": 0.5556,
+      "step": 18318
+    },
+    {
+      "epoch": 4.863135536970662,
+      "grad_norm": 0.4948248268211596,
+      "learning_rate": 4.3284033752880204e-07,
+      "loss": 0.5168,
+      "step": 18319
+    },
+    {
+      "epoch": 4.8634010354440464,
+      "grad_norm": 0.4716475426914833,
+      "learning_rate": 4.3264400740447493e-07,
+      "loss": 0.556,
+      "step": 18320
+    },
+    {
+      "epoch": 4.86366653391743,
+      "grad_norm": 0.48275748309020167,
+      "learning_rate": 4.3244771759865077e-07,
+      "loss": 0.5268,
+      "step": 18321
+    },
+    {
+      "epoch": 4.863932032390814,
+      "grad_norm": 0.49213430238850403,
+      "learning_rate": 4.3225146811515905e-07,
+      "loss": 0.5434,
+      "step": 18322
+    },
+    {
+      "epoch": 4.864197530864198,
+      "grad_norm": 0.4845106807170754,
+      "learning_rate": 4.3205525895782583e-07,
+      "loss": 0.5464,
+      "step": 18323
+    },
+    {
+      "epoch": 4.864463029337581,
+      "grad_norm": 0.493687376700083,
+      "learning_rate": 4.3185909013047895e-07,
+      "loss": 0.5344,
+      "step": 18324
+    },
+    {
+      "epoch": 4.864728527810965,
+      "grad_norm": 0.48835114409912345,
+      "learning_rate": 4.3166296163694335e-07,
+      "loss": 0.5487,
+      "step": 18325
+    },
+    {
+      "epoch": 4.864994026284349,
+      "grad_norm": 0.4855202072451562,
+      "learning_rate": 4.3146687348104343e-07,
+      "loss": 0.5389,
+      "step": 18326
+    },
+    {
+      "epoch": 4.865259524757732,
+      "grad_norm": 0.4713399823454973,
+      "learning_rate": 4.3127082566660513e-07,
+      "loss": 0.4949,
+      "step": 18327
+    },
+    {
+      "epoch": 4.8655250232311165,
+      "grad_norm": 0.47773640536140555,
+      "learning_rate": 4.3107481819745e-07,
+      "loss": 0.5354,
+      "step": 18328
+    },
+    {
+      "epoch": 4.8657905217045005,
+      "grad_norm": 0.4859872363105243,
+      "learning_rate": 4.30878851077402e-07,
+      "loss": 0.5689,
+      "step": 18329
+    },
+    {
+      "epoch": 4.866056020177884,
+      "grad_norm": 0.48302689644037183,
+      "learning_rate": 4.306829243102828e-07,
+      "loss": 0.5221,
+      "step": 18330
+    },
+    {
+      "epoch": 4.866321518651268,
+      "grad_norm": 0.46060417865132963,
+      "learning_rate": 4.3048703789991293e-07,
+      "loss": 0.5086,
+      "step": 18331
+    },
+    {
+      "epoch": 4.866587017124652,
+      "grad_norm": 0.48343277839427656,
+      "learning_rate": 4.302911918501121e-07,
+      "loss": 0.518,
+      "step": 18332
+    },
+    {
+      "epoch": 4.866852515598035,
+      "grad_norm": 0.48172000263431325,
+      "learning_rate": 4.3009538616470086e-07,
+      "loss": 0.5344,
+      "step": 18333
+    },
+    {
+      "epoch": 4.867118014071419,
+      "grad_norm": 0.46326566281550374,
+      "learning_rate": 4.2989962084749813e-07,
+      "loss": 0.5124,
+      "step": 18334
+    },
+    {
+      "epoch": 4.867383512544803,
+      "grad_norm": 0.47931751993884397,
+      "learning_rate": 4.2970389590232115e-07,
+      "loss": 0.5275,
+      "step": 18335
+    },
+    {
+      "epoch": 4.8676490110181865,
+      "grad_norm": 0.4949485801208798,
+      "learning_rate": 4.295082113329871e-07,
+      "loss": 0.5216,
+      "step": 18336
+    },
+    {
+      "epoch": 4.867914509491571,
+      "grad_norm": 0.4811969202303654,
+      "learning_rate": 4.293125671433118e-07,
+      "loss": 0.5298,
+      "step": 18337
+    },
+    {
+      "epoch": 4.868180007964954,
+      "grad_norm": 0.4790510325048018,
+      "learning_rate": 4.291169633371123e-07,
+      "loss": 0.5409,
+      "step": 18338
+    },
+    {
+      "epoch": 4.868445506438338,
+      "grad_norm": 0.4982732152367065,
+      "learning_rate": 4.2892139991820155e-07,
+      "loss": 0.5286,
+      "step": 18339
+    },
+    {
+      "epoch": 4.868711004911722,
+      "grad_norm": 0.4833922763599852,
+      "learning_rate": 4.2872587689039486e-07,
+      "loss": 0.5614,
+      "step": 18340
+    },
+    {
+      "epoch": 4.868976503385106,
+      "grad_norm": 0.4864055240320268,
+      "learning_rate": 4.2853039425750504e-07,
+      "loss": 0.5351,
+      "step": 18341
+    },
+    {
+      "epoch": 4.869242001858489,
+      "grad_norm": 0.48218077225771605,
+      "learning_rate": 4.2833495202334374e-07,
+      "loss": 0.5249,
+      "step": 18342
+    },
+    {
+      "epoch": 4.869507500331873,
+      "grad_norm": 0.4751794345863071,
+      "learning_rate": 4.2813955019172377e-07,
+      "loss": 0.5584,
+      "step": 18343
+    },
+    {
+      "epoch": 4.8697729988052565,
+      "grad_norm": 0.4813275314226737,
+      "learning_rate": 4.279441887664548e-07,
+      "loss": 0.5306,
+      "step": 18344
+    },
+    {
+      "epoch": 4.870038497278641,
+      "grad_norm": 0.4756377729291008,
+      "learning_rate": 4.2774886775134805e-07,
+      "loss": 0.5034,
+      "step": 18345
+    },
+    {
+      "epoch": 4.870303995752025,
+      "grad_norm": 0.49312856886810574,
+      "learning_rate": 4.2755358715021153e-07,
+      "loss": 0.5422,
+      "step": 18346
+    },
+    {
+      "epoch": 4.870569494225408,
+      "grad_norm": 0.4908499628146212,
+      "learning_rate": 4.2735834696685546e-07,
+      "loss": 0.5613,
+      "step": 18347
+    },
+    {
+      "epoch": 4.870834992698792,
+      "grad_norm": 0.47977338055579594,
+      "learning_rate": 4.271631472050852e-07,
+      "loss": 0.4989,
+      "step": 18348
+    },
+    {
+      "epoch": 4.871100491172176,
+      "grad_norm": 0.49012474114926224,
+      "learning_rate": 4.269679878687091e-07,
+      "loss": 0.5464,
+      "step": 18349
+    },
+    {
+      "epoch": 4.871365989645559,
+      "grad_norm": 0.4856052040788197,
+      "learning_rate": 4.2677286896153246e-07,
+      "loss": 0.5466,
+      "step": 18350
+    },
+    {
+      "epoch": 4.871631488118943,
+      "grad_norm": 0.48966160842269185,
+      "learning_rate": 4.2657779048736166e-07,
+      "loss": 0.5371,
+      "step": 18351
+    },
+    {
+      "epoch": 4.871896986592327,
+      "grad_norm": 0.4980364903622559,
+      "learning_rate": 4.2638275245e-07,
+      "loss": 0.5495,
+      "step": 18352
+    },
+    {
+      "epoch": 4.872162485065711,
+      "grad_norm": 0.5163094063901065,
+      "learning_rate": 4.2618775485325227e-07,
+      "loss": 0.5562,
+      "step": 18353
+    },
+    {
+      "epoch": 4.872427983539095,
+      "grad_norm": 0.5009413326110693,
+      "learning_rate": 4.2599279770092093e-07,
+      "loss": 0.5114,
+      "step": 18354
+    },
+    {
+      "epoch": 4.872693482012479,
+      "grad_norm": 0.5107163430498799,
+      "learning_rate": 4.257978809968075e-07,
+      "loss": 0.5381,
+      "step": 18355
+    },
+    {
+      "epoch": 4.872958980485862,
+      "grad_norm": 0.4860251587138234,
+      "learning_rate": 4.256030047447143e-07,
+      "loss": 0.5421,
+      "step": 18356
+    },
+    {
+      "epoch": 4.873224478959246,
+      "grad_norm": 0.47779130887325744,
+      "learning_rate": 4.25408168948441e-07,
+      "loss": 0.533,
+      "step": 18357
+    },
+    {
+      "epoch": 4.87348997743263,
+      "grad_norm": 0.48679416287399835,
+      "learning_rate": 4.252133736117886e-07,
+      "loss": 0.5419,
+      "step": 18358
+    },
+    {
+      "epoch": 4.873755475906013,
+      "grad_norm": 0.4708893080108311,
+      "learning_rate": 4.2501861873855523e-07,
+      "loss": 0.5016,
+      "step": 18359
+    },
+    {
+      "epoch": 4.874020974379397,
+      "grad_norm": 0.4721462520961156,
+      "learning_rate": 4.248239043325389e-07,
+      "loss": 0.5457,
+      "step": 18360
+    },
+    {
+      "epoch": 4.874286472852781,
+      "grad_norm": 0.48281378544516595,
+      "learning_rate": 4.2462923039753773e-07,
+      "loss": 0.5179,
+      "step": 18361
+    },
+    {
+      "epoch": 4.874551971326165,
+      "grad_norm": 0.4796895380351003,
+      "learning_rate": 4.244345969373473e-07,
+      "loss": 0.539,
+      "step": 18362
+    },
+    {
+      "epoch": 4.874817469799549,
+      "grad_norm": 0.4807227578314435,
+      "learning_rate": 4.242400039557648e-07,
+      "loss": 0.5027,
+      "step": 18363
+    },
+    {
+      "epoch": 4.875082968272933,
+      "grad_norm": 0.4897859449715257,
+      "learning_rate": 4.240454514565845e-07,
+      "loss": 0.5629,
+      "step": 18364
+    },
+    {
+      "epoch": 4.875348466746316,
+      "grad_norm": 0.48739320303971895,
+      "learning_rate": 4.238509394436008e-07,
+      "loss": 0.5615,
+      "step": 18365
+    },
+    {
+      "epoch": 4.8756139652197,
+      "grad_norm": 0.4912821025482378,
+      "learning_rate": 4.236564679206062e-07,
+      "loss": 0.5437,
+      "step": 18366
+    },
+    {
+      "epoch": 4.875879463693083,
+      "grad_norm": 0.4673104812641789,
+      "learning_rate": 4.2346203689139496e-07,
+      "loss": 0.5158,
+      "step": 18367
+    },
+    {
+      "epoch": 4.876144962166467,
+      "grad_norm": 0.4937501096770388,
+      "learning_rate": 4.232676463597579e-07,
+      "loss": 0.5203,
+      "step": 18368
+    },
+    {
+      "epoch": 4.8764104606398515,
+      "grad_norm": 0.4801489823516729,
+      "learning_rate": 4.2307329632948674e-07,
+      "loss": 0.5479,
+      "step": 18369
+    },
+    {
+      "epoch": 4.876675959113236,
+      "grad_norm": 0.48233570381407764,
+      "learning_rate": 4.228789868043717e-07,
+      "loss": 0.5398,
+      "step": 18370
+    },
+    {
+      "epoch": 4.876941457586619,
+      "grad_norm": 0.4747119792356855,
+      "learning_rate": 4.226847177882015e-07,
+      "loss": 0.5475,
+      "step": 18371
+    },
+    {
+      "epoch": 4.877206956060003,
+      "grad_norm": 0.4732960639484342,
+      "learning_rate": 4.224904892847659e-07,
+      "loss": 0.5262,
+      "step": 18372
+    },
+    {
+      "epoch": 4.877472454533386,
+      "grad_norm": 0.5022349703582525,
+      "learning_rate": 4.2229630129785177e-07,
+      "loss": 0.5487,
+      "step": 18373
+    },
+    {
+      "epoch": 4.87773795300677,
+      "grad_norm": 0.4858987405689483,
+      "learning_rate": 4.221021538312475e-07,
+      "loss": 0.5502,
+      "step": 18374
+    },
+    {
+      "epoch": 4.878003451480154,
+      "grad_norm": 0.48850277880214665,
+      "learning_rate": 4.219080468887382e-07,
+      "loss": 0.5874,
+      "step": 18375
+    },
+    {
+      "epoch": 4.878268949953537,
+      "grad_norm": 0.4819318916630262,
+      "learning_rate": 4.217139804741113e-07,
+      "loss": 0.5141,
+      "step": 18376
+    },
+    {
+      "epoch": 4.8785344484269215,
+      "grad_norm": 0.48319873590506474,
+      "learning_rate": 4.2151995459114914e-07,
+      "loss": 0.5316,
+      "step": 18377
+    },
+    {
+      "epoch": 4.878799946900306,
+      "grad_norm": 0.4751816904790201,
+      "learning_rate": 4.2132596924363666e-07,
+      "loss": 0.5546,
+      "step": 18378
+    },
+    {
+      "epoch": 4.879065445373689,
+      "grad_norm": 0.4911216679462683,
+      "learning_rate": 4.211320244353581e-07,
+      "loss": 0.5476,
+      "step": 18379
+    },
+    {
+      "epoch": 4.879330943847073,
+      "grad_norm": 0.49113664844913535,
+      "learning_rate": 4.2093812017009435e-07,
+      "loss": 0.5606,
+      "step": 18380
+    },
+    {
+      "epoch": 4.879596442320457,
+      "grad_norm": 0.508419431758893,
+      "learning_rate": 4.207442564516287e-07,
+      "loss": 0.5364,
+      "step": 18381
+    },
+    {
+      "epoch": 4.87986194079384,
+      "grad_norm": 0.4990457146422748,
+      "learning_rate": 4.2055043328373986e-07,
+      "loss": 0.5978,
+      "step": 18382
+    },
+    {
+      "epoch": 4.880127439267224,
+      "grad_norm": 0.4776995102640195,
+      "learning_rate": 4.203566506702095e-07,
+      "loss": 0.5442,
+      "step": 18383
+    },
+    {
+      "epoch": 4.880392937740608,
+      "grad_norm": 0.4989231219101574,
+      "learning_rate": 4.2016290861481594e-07,
+      "loss": 0.5463,
+      "step": 18384
+    },
+    {
+      "epoch": 4.8806584362139915,
+      "grad_norm": 0.4929326685793675,
+      "learning_rate": 4.199692071213382e-07,
+      "loss": 0.5211,
+      "step": 18385
+    },
+    {
+      "epoch": 4.880923934687376,
+      "grad_norm": 0.4733748801751327,
+      "learning_rate": 4.197755461935535e-07,
+      "loss": 0.5181,
+      "step": 18386
+    },
+    {
+      "epoch": 4.88118943316076,
+      "grad_norm": 0.47189202901291055,
+      "learning_rate": 4.1958192583523933e-07,
+      "loss": 0.5096,
+      "step": 18387
+    },
+    {
+      "epoch": 4.881454931634143,
+      "grad_norm": 0.4701019391446451,
+      "learning_rate": 4.1938834605017133e-07,
+      "loss": 0.5012,
+      "step": 18388
+    },
+    {
+      "epoch": 4.881720430107527,
+      "grad_norm": 0.48673683063968365,
+      "learning_rate": 4.191948068421239e-07,
+      "loss": 0.5746,
+      "step": 18389
+    },
+    {
+      "epoch": 4.88198592858091,
+      "grad_norm": 0.4764424312405619,
+      "learning_rate": 4.1900130821487364e-07,
+      "loss": 0.541,
+      "step": 18390
+    },
+    {
+      "epoch": 4.882251427054294,
+      "grad_norm": 0.49890411321842876,
+      "learning_rate": 4.1880785017219207e-07,
+      "loss": 0.5323,
+      "step": 18391
+    },
+    {
+      "epoch": 4.882516925527678,
+      "grad_norm": 0.49590494527481394,
+      "learning_rate": 4.1861443271785375e-07,
+      "loss": 0.5566,
+      "step": 18392
+    },
+    {
+      "epoch": 4.882782424001062,
+      "grad_norm": 0.4654112185478426,
+      "learning_rate": 4.1842105585562983e-07,
+      "loss": 0.5113,
+      "step": 18393
+    },
+    {
+      "epoch": 4.883047922474446,
+      "grad_norm": 0.5148089946837342,
+      "learning_rate": 4.1822771958929225e-07,
+      "loss": 0.5435,
+      "step": 18394
+    },
+    {
+      "epoch": 4.88331342094783,
+      "grad_norm": 0.4998546571759412,
+      "learning_rate": 4.1803442392261057e-07,
+      "loss": 0.5344,
+      "step": 18395
+    },
+    {
+      "epoch": 4.883578919421213,
+      "grad_norm": 0.4850552841319684,
+      "learning_rate": 4.178411688593556e-07,
+      "loss": 0.5111,
+      "step": 18396
+    },
+    {
+      "epoch": 4.883844417894597,
+      "grad_norm": 0.48460109747212365,
+      "learning_rate": 4.176479544032952e-07,
+      "loss": 0.5341,
+      "step": 18397
+    },
+    {
+      "epoch": 4.884109916367981,
+      "grad_norm": 0.49271501971659776,
+      "learning_rate": 4.1745478055819884e-07,
+      "loss": 0.5114,
+      "step": 18398
+    },
+    {
+      "epoch": 4.884375414841364,
+      "grad_norm": 0.4954059306983671,
+      "learning_rate": 4.172616473278332e-07,
+      "loss": 0.5178,
+      "step": 18399
+    },
+    {
+      "epoch": 4.884640913314748,
+      "grad_norm": 0.48306772868386066,
+      "learning_rate": 4.1706855471596446e-07,
+      "loss": 0.5298,
+      "step": 18400
+    },
+    {
+      "epoch": 4.8849064117881325,
+      "grad_norm": 0.4786021070958391,
+      "learning_rate": 4.168755027263591e-07,
+      "loss": 0.5519,
+      "step": 18401
+    },
+    {
+      "epoch": 4.885171910261516,
+      "grad_norm": 0.48412201472714805,
+      "learning_rate": 4.166824913627815e-07,
+      "loss": 0.5187,
+      "step": 18402
+    },
+    {
+      "epoch": 4.8854374087349,
+      "grad_norm": 0.4843226980062491,
+      "learning_rate": 4.164895206289965e-07,
+      "loss": 0.5294,
+      "step": 18403
+    },
+    {
+      "epoch": 4.885702907208284,
+      "grad_norm": 0.48542181014244334,
+      "learning_rate": 4.162965905287672e-07,
+      "loss": 0.5517,
+      "step": 18404
+    },
+    {
+      "epoch": 4.885968405681667,
+      "grad_norm": 0.48028580718071917,
+      "learning_rate": 4.1610370106585586e-07,
+      "loss": 0.5372,
+      "step": 18405
+    },
+    {
+      "epoch": 4.886233904155051,
+      "grad_norm": 0.49130598904825407,
+      "learning_rate": 4.15910852244025e-07,
+      "loss": 0.5485,
+      "step": 18406
+    },
+    {
+      "epoch": 4.886499402628435,
+      "grad_norm": 0.4888697734491091,
+      "learning_rate": 4.157180440670344e-07,
+      "loss": 0.5954,
+      "step": 18407
+    },
+    {
+      "epoch": 4.886764901101818,
+      "grad_norm": 0.4846521356232636,
+      "learning_rate": 4.155252765386461e-07,
+      "loss": 0.5953,
+      "step": 18408
+    },
+    {
+      "epoch": 4.8870303995752025,
+      "grad_norm": 0.4911611927125826,
+      "learning_rate": 4.1533254966261787e-07,
+      "loss": 0.5706,
+      "step": 18409
+    },
+    {
+      "epoch": 4.8872958980485866,
+      "grad_norm": 0.4879318565105792,
+      "learning_rate": 4.151398634427103e-07,
+      "loss": 0.5224,
+      "step": 18410
+    },
+    {
+      "epoch": 4.88756139652197,
+      "grad_norm": 0.4725602566725544,
+      "learning_rate": 4.1494721788267854e-07,
+      "loss": 0.5417,
+      "step": 18411
+    },
+    {
+      "epoch": 4.887826894995354,
+      "grad_norm": 0.48994874100731417,
+      "learning_rate": 4.14754612986282e-07,
+      "loss": 0.5515,
+      "step": 18412
+    },
+    {
+      "epoch": 4.888092393468738,
+      "grad_norm": 0.49380180248989286,
+      "learning_rate": 4.1456204875727545e-07,
+      "loss": 0.5512,
+      "step": 18413
+    },
+    {
+      "epoch": 4.888357891942121,
+      "grad_norm": 0.47075381119159593,
+      "learning_rate": 4.1436952519941564e-07,
+      "loss": 0.5211,
+      "step": 18414
+    },
+    {
+      "epoch": 4.888623390415505,
+      "grad_norm": 0.4963607221965845,
+      "learning_rate": 4.1417704231645594e-07,
+      "loss": 0.5316,
+      "step": 18415
+    },
+    {
+      "epoch": 4.888888888888889,
+      "grad_norm": 0.48921921958476,
+      "learning_rate": 4.139846001121514e-07,
+      "loss": 0.5414,
+      "step": 18416
+    },
+    {
+      "epoch": 4.8891543873622725,
+      "grad_norm": 0.48685124256258455,
+      "learning_rate": 4.137921985902546e-07,
+      "loss": 0.5224,
+      "step": 18417
+    },
+    {
+      "epoch": 4.889419885835657,
+      "grad_norm": 0.4846439398388234,
+      "learning_rate": 4.1359983775451724e-07,
+      "loss": 0.5429,
+      "step": 18418
+    },
+    {
+      "epoch": 4.88968538430904,
+      "grad_norm": 0.47937026845934433,
+      "learning_rate": 4.1340751760869216e-07,
+      "loss": 0.548,
+      "step": 18419
+    },
+    {
+      "epoch": 4.889950882782424,
+      "grad_norm": 0.471580656178394,
+      "learning_rate": 4.132152381565288e-07,
+      "loss": 0.5351,
+      "step": 18420
+    },
+    {
+      "epoch": 4.890216381255808,
+      "grad_norm": 0.49544373258000335,
+      "learning_rate": 4.1302299940177785e-07,
+      "loss": 0.553,
+      "step": 18421
+    },
+    {
+      "epoch": 4.890481879729192,
+      "grad_norm": 0.4737998174151155,
+      "learning_rate": 4.1283080134818853e-07,
+      "loss": 0.5274,
+      "step": 18422
+    },
+    {
+      "epoch": 4.890747378202575,
+      "grad_norm": 0.4867390890113528,
+      "learning_rate": 4.126386439995089e-07,
+      "loss": 0.5467,
+      "step": 18423
+    },
+    {
+      "epoch": 4.891012876675959,
+      "grad_norm": 0.4762126250908237,
+      "learning_rate": 4.124465273594858e-07,
+      "loss": 0.5484,
+      "step": 18424
+    },
+    {
+      "epoch": 4.8912783751493425,
+      "grad_norm": 0.4718947677489696,
+      "learning_rate": 4.122544514318666e-07,
+      "loss": 0.5497,
+      "step": 18425
+    },
+    {
+      "epoch": 4.891543873622727,
+      "grad_norm": 0.4780129714432393,
+      "learning_rate": 4.1206241622039763e-07,
+      "loss": 0.5646,
+      "step": 18426
+    },
+    {
+      "epoch": 4.891809372096111,
+      "grad_norm": 0.482147985436244,
+      "learning_rate": 4.1187042172882414e-07,
+      "loss": 0.529,
+      "step": 18427
+    },
+    {
+      "epoch": 4.892074870569494,
+      "grad_norm": 0.46716268000097677,
+      "learning_rate": 4.1167846796088983e-07,
+      "loss": 0.5335,
+      "step": 18428
+    },
+    {
+      "epoch": 4.892340369042878,
+      "grad_norm": 0.49129417012713433,
+      "learning_rate": 4.1148655492033786e-07,
+      "loss": 0.5309,
+      "step": 18429
+    },
+    {
+      "epoch": 4.892605867516262,
+      "grad_norm": 0.47774999971490945,
+      "learning_rate": 4.112946826109121e-07,
+      "loss": 0.5484,
+      "step": 18430
+    },
+    {
+      "epoch": 4.892871365989645,
+      "grad_norm": 0.4757632842027513,
+      "learning_rate": 4.111028510363538e-07,
+      "loss": 0.513,
+      "step": 18431
+    },
+    {
+      "epoch": 4.893136864463029,
+      "grad_norm": 0.4933324270655687,
+      "learning_rate": 4.109110602004046e-07,
+      "loss": 0.521,
+      "step": 18432
+    },
+    {
+      "epoch": 4.893402362936413,
+      "grad_norm": 0.4814300463409591,
+      "learning_rate": 4.107193101068049e-07,
+      "loss": 0.5413,
+      "step": 18433
+    },
+    {
+      "epoch": 4.893667861409797,
+      "grad_norm": 0.4834698435760068,
+      "learning_rate": 4.105276007592937e-07,
+      "loss": 0.5323,
+      "step": 18434
+    },
+    {
+      "epoch": 4.893933359883181,
+      "grad_norm": 0.4752344866048308,
+      "learning_rate": 4.1033593216161047e-07,
+      "loss": 0.5272,
+      "step": 18435
+    },
+    {
+      "epoch": 4.894198858356565,
+      "grad_norm": 0.48028625811257475,
+      "learning_rate": 4.101443043174924e-07,
+      "loss": 0.5421,
+      "step": 18436
+    },
+    {
+      "epoch": 4.894464356829948,
+      "grad_norm": 0.4834073871047219,
+      "learning_rate": 4.0995271723067804e-07,
+      "loss": 0.5546,
+      "step": 18437
+    },
+    {
+      "epoch": 4.894729855303332,
+      "grad_norm": 0.48160827285640334,
+      "learning_rate": 4.0976117090490234e-07,
+      "loss": 0.5168,
+      "step": 18438
+    },
+    {
+      "epoch": 4.894995353776716,
+      "grad_norm": 0.46428187831004203,
+      "learning_rate": 4.0956966534390267e-07,
+      "loss": 0.5585,
+      "step": 18439
+    },
+    {
+      "epoch": 4.895260852250099,
+      "grad_norm": 0.47083026619909496,
+      "learning_rate": 4.0937820055141184e-07,
+      "loss": 0.5343,
+      "step": 18440
+    },
+    {
+      "epoch": 4.895526350723483,
+      "grad_norm": 0.4608205803943905,
+      "learning_rate": 4.0918677653116523e-07,
+      "loss": 0.5073,
+      "step": 18441
+    },
+    {
+      "epoch": 4.8957918491968675,
+      "grad_norm": 0.48486301344327903,
+      "learning_rate": 4.089953932868951e-07,
+      "loss": 0.5234,
+      "step": 18442
+    },
+    {
+      "epoch": 4.896057347670251,
+      "grad_norm": 0.49201647852666075,
+      "learning_rate": 4.088040508223351e-07,
+      "loss": 0.5529,
+      "step": 18443
+    },
+    {
+      "epoch": 4.896322846143635,
+      "grad_norm": 0.4800721186511747,
+      "learning_rate": 4.0861274914121534e-07,
+      "loss": 0.5555,
+      "step": 18444
+    },
+    {
+      "epoch": 4.896588344617019,
+      "grad_norm": 0.49060033990975155,
+      "learning_rate": 4.0842148824726866e-07,
+      "loss": 0.5084,
+      "step": 18445
+    },
+    {
+      "epoch": 4.896853843090402,
+      "grad_norm": 0.4877058000778876,
+      "learning_rate": 4.082302681442238e-07,
+      "loss": 0.5614,
+      "step": 18446
+    },
+    {
+      "epoch": 4.897119341563786,
+      "grad_norm": 0.4630546679428516,
+      "learning_rate": 4.080390888358096e-07,
+      "loss": 0.5191,
+      "step": 18447
+    },
+    {
+      "epoch": 4.897384840037169,
+      "grad_norm": 0.4958745600192032,
+      "learning_rate": 4.078479503257557e-07,
+      "loss": 0.5295,
+      "step": 18448
+    },
+    {
+      "epoch": 4.897650338510553,
+      "grad_norm": 0.4982075995240193,
+      "learning_rate": 4.076568526177888e-07,
+      "loss": 0.5335,
+      "step": 18449
+    },
+    {
+      "epoch": 4.8979158369839375,
+      "grad_norm": 0.4868892885918251,
+      "learning_rate": 4.0746579571563667e-07,
+      "loss": 0.5499,
+      "step": 18450
+    },
+    {
+      "epoch": 4.898181335457322,
+      "grad_norm": 0.4857059544818654,
+      "learning_rate": 4.07274779623025e-07,
+      "loss": 0.5398,
+      "step": 18451
+    },
+    {
+      "epoch": 4.898446833930705,
+      "grad_norm": 0.4819152727112816,
+      "learning_rate": 4.070838043436787e-07,
+      "loss": 0.5596,
+      "step": 18452
+    },
+    {
+      "epoch": 4.898712332404089,
+      "grad_norm": 0.4768515191830725,
+      "learning_rate": 4.0689286988132263e-07,
+      "loss": 0.5113,
+      "step": 18453
+    },
+    {
+      "epoch": 4.898977830877472,
+      "grad_norm": 0.4714311116391779,
+      "learning_rate": 4.067019762396801e-07,
+      "loss": 0.5262,
+      "step": 18454
+    },
+    {
+      "epoch": 4.899243329350856,
+      "grad_norm": 0.4645356801345613,
+      "learning_rate": 4.06511123422475e-07,
+      "loss": 0.5425,
+      "step": 18455
+    },
+    {
+      "epoch": 4.89950882782424,
+      "grad_norm": 0.4786660574805506,
+      "learning_rate": 4.063203114334285e-07,
+      "loss": 0.5207,
+      "step": 18456
+    },
+    {
+      "epoch": 4.899774326297623,
+      "grad_norm": 0.48979638518458085,
+      "learning_rate": 4.0612954027626215e-07,
+      "loss": 0.5082,
+      "step": 18457
+    },
+    {
+      "epoch": 4.9000398247710075,
+      "grad_norm": 0.4782625960353114,
+      "learning_rate": 4.05938809954696e-07,
+      "loss": 0.5252,
+      "step": 18458
+    },
+    {
+      "epoch": 4.900305323244392,
+      "grad_norm": 0.4966983409946847,
+      "learning_rate": 4.057481204724509e-07,
+      "loss": 0.5009,
+      "step": 18459
+    },
+    {
+      "epoch": 4.900570821717775,
+      "grad_norm": 0.480135260760648,
+      "learning_rate": 4.055574718332442e-07,
+      "loss": 0.5413,
+      "step": 18460
+    },
+    {
+      "epoch": 4.900836320191159,
+      "grad_norm": 0.49228855592047943,
+      "learning_rate": 4.053668640407957e-07,
+      "loss": 0.5293,
+      "step": 18461
+    },
+    {
+      "epoch": 4.901101818664543,
+      "grad_norm": 0.4991104434773967,
+      "learning_rate": 4.0517629709882185e-07,
+      "loss": 0.5537,
+      "step": 18462
+    },
+    {
+      "epoch": 4.901367317137926,
+      "grad_norm": 0.4848441092343395,
+      "learning_rate": 4.049857710110383e-07,
+      "loss": 0.5453,
+      "step": 18463
+    },
+    {
+      "epoch": 4.90163281561131,
+      "grad_norm": 0.4802812102801752,
+      "learning_rate": 4.0479528578116266e-07,
+      "loss": 0.5517,
+      "step": 18464
+    },
+    {
+      "epoch": 4.901898314084694,
+      "grad_norm": 0.5010223849110018,
+      "learning_rate": 4.0460484141290807e-07,
+      "loss": 0.5383,
+      "step": 18465
+    },
+    {
+      "epoch": 4.9021638125580775,
+      "grad_norm": 0.49163946234859396,
+      "learning_rate": 4.044144379099904e-07,
+      "loss": 0.5478,
+      "step": 18466
+    },
+    {
+      "epoch": 4.902429311031462,
+      "grad_norm": 0.48166823573972256,
+      "learning_rate": 4.0422407527612116e-07,
+      "loss": 0.5604,
+      "step": 18467
+    },
+    {
+      "epoch": 4.902694809504846,
+      "grad_norm": 0.4866767661300027,
+      "learning_rate": 4.0403375351501515e-07,
+      "loss": 0.5114,
+      "step": 18468
+    },
+    {
+      "epoch": 4.902960307978229,
+      "grad_norm": 0.4805071361612573,
+      "learning_rate": 4.038434726303814e-07,
+      "loss": 0.5385,
+      "step": 18469
+    },
+    {
+      "epoch": 4.903225806451613,
+      "grad_norm": 0.4791933171802118,
+      "learning_rate": 4.036532326259324e-07,
+      "loss": 0.5205,
+      "step": 18470
+    },
+    {
+      "epoch": 4.903491304924996,
+      "grad_norm": 0.4955759113248641,
+      "learning_rate": 4.0346303350537853e-07,
+      "loss": 0.5555,
+      "step": 18471
+    },
+    {
+      "epoch": 4.90375680339838,
+      "grad_norm": 0.48804101614789513,
+      "learning_rate": 4.032728752724283e-07,
+      "loss": 0.5497,
+      "step": 18472
+    },
+    {
+      "epoch": 4.904022301871764,
+      "grad_norm": 0.48754085262225133,
+      "learning_rate": 4.030827579307911e-07,
+      "loss": 0.5209,
+      "step": 18473
+    },
+    {
+      "epoch": 4.904287800345148,
+      "grad_norm": 0.46968906619538753,
+      "learning_rate": 4.0289268148417437e-07,
+      "loss": 0.523,
+      "step": 18474
+    },
+    {
+      "epoch": 4.904553298818532,
+      "grad_norm": 0.47739486215422067,
+      "learning_rate": 4.0270264593628485e-07,
+      "loss": 0.5357,
+      "step": 18475
+    },
+    {
+      "epoch": 4.904818797291916,
+      "grad_norm": 0.4661677278280224,
+      "learning_rate": 4.025126512908284e-07,
+      "loss": 0.4996,
+      "step": 18476
+    },
+    {
+      "epoch": 4.905084295765299,
+      "grad_norm": 0.4821901112508252,
+      "learning_rate": 4.0232269755151126e-07,
+      "loss": 0.5124,
+      "step": 18477
+    },
+    {
+      "epoch": 4.905349794238683,
+      "grad_norm": 0.4843198988973941,
+      "learning_rate": 4.021327847220369e-07,
+      "loss": 0.5358,
+      "step": 18478
+    },
+    {
+      "epoch": 4.905615292712067,
+      "grad_norm": 0.48987834295660043,
+      "learning_rate": 4.019429128061103e-07,
+      "loss": 0.5241,
+      "step": 18479
+    },
+    {
+      "epoch": 4.905880791185451,
+      "grad_norm": 0.4866309815177567,
+      "learning_rate": 4.017530818074339e-07,
+      "loss": 0.5261,
+      "step": 18480
+    },
+    {
+      "epoch": 4.906146289658834,
+      "grad_norm": 0.49597631488660615,
+      "learning_rate": 4.0156329172970907e-07,
+      "loss": 0.5613,
+      "step": 18481
+    },
+    {
+      "epoch": 4.9064117881322185,
+      "grad_norm": 0.48317651561446956,
+      "learning_rate": 4.013735425766388e-07,
+      "loss": 0.5256,
+      "step": 18482
+    },
+    {
+      "epoch": 4.906677286605602,
+      "grad_norm": 0.47907128565238893,
+      "learning_rate": 4.011838343519217e-07,
+      "loss": 0.5168,
+      "step": 18483
+    },
+    {
+      "epoch": 4.906942785078986,
+      "grad_norm": 0.48552897716575216,
+      "learning_rate": 4.009941670592596e-07,
+      "loss": 0.508,
+      "step": 18484
+    },
+    {
+      "epoch": 4.90720828355237,
+      "grad_norm": 0.4925522713194775,
+      "learning_rate": 4.008045407023503e-07,
+      "loss": 0.5306,
+      "step": 18485
+    },
+    {
+      "epoch": 4.907473782025753,
+      "grad_norm": 0.47141868739228854,
+      "learning_rate": 4.006149552848923e-07,
+      "loss": 0.5495,
+      "step": 18486
+    },
+    {
+      "epoch": 4.907739280499137,
+      "grad_norm": 0.5010911304282613,
+      "learning_rate": 4.0042541081058233e-07,
+      "loss": 0.5548,
+      "step": 18487
+    },
+    {
+      "epoch": 4.908004778972521,
+      "grad_norm": 0.48496202925962095,
+      "learning_rate": 4.0023590728311803e-07,
+      "loss": 0.6015,
+      "step": 18488
+    },
+    {
+      "epoch": 4.908270277445904,
+      "grad_norm": 0.4783687116046605,
+      "learning_rate": 4.0004644470619387e-07,
+      "loss": 0.5148,
+      "step": 18489
+    },
+    {
+      "epoch": 4.9085357759192885,
+      "grad_norm": 0.47008191416413353,
+      "learning_rate": 3.9985702308350584e-07,
+      "loss": 0.5305,
+      "step": 18490
+    },
+    {
+      "epoch": 4.9088012743926726,
+      "grad_norm": 0.478511553942559,
+      "learning_rate": 3.9966764241874874e-07,
+      "loss": 0.5145,
+      "step": 18491
+    },
+    {
+      "epoch": 4.909066772866056,
+      "grad_norm": 0.4995702624103274,
+      "learning_rate": 3.994783027156143e-07,
+      "loss": 0.5736,
+      "step": 18492
+    },
+    {
+      "epoch": 4.90933227133944,
+      "grad_norm": 0.4781811665389976,
+      "learning_rate": 3.9928900397779637e-07,
+      "loss": 0.5309,
+      "step": 18493
+    },
+    {
+      "epoch": 4.909597769812824,
+      "grad_norm": 0.482656080505446,
+      "learning_rate": 3.9909974620898574e-07,
+      "loss": 0.5345,
+      "step": 18494
+    },
+    {
+      "epoch": 4.909863268286207,
+      "grad_norm": 0.5063366573195083,
+      "learning_rate": 3.989105294128745e-07,
+      "loss": 0.546,
+      "step": 18495
+    },
+    {
+      "epoch": 4.910128766759591,
+      "grad_norm": 0.4753574593879358,
+      "learning_rate": 3.9872135359315165e-07,
+      "loss": 0.5381,
+      "step": 18496
+    },
+    {
+      "epoch": 4.910394265232975,
+      "grad_norm": 0.5057387420964584,
+      "learning_rate": 3.985322187535079e-07,
+      "loss": 0.5764,
+      "step": 18497
+    },
+    {
+      "epoch": 4.9106597637063585,
+      "grad_norm": 0.4754794799989428,
+      "learning_rate": 3.9834312489763105e-07,
+      "loss": 0.5507,
+      "step": 18498
+    },
+    {
+      "epoch": 4.910925262179743,
+      "grad_norm": 0.47542382266471056,
+      "learning_rate": 3.9815407202920845e-07,
+      "loss": 0.515,
+      "step": 18499
+    },
+    {
+      "epoch": 4.911190760653126,
+      "grad_norm": 0.4914924354150179,
+      "learning_rate": 3.979650601519283e-07,
+      "loss": 0.5519,
+      "step": 18500
+    },
+    {
+      "epoch": 4.91145625912651,
+      "grad_norm": 0.4793312098374365,
+      "learning_rate": 3.977760892694757e-07,
+      "loss": 0.5543,
+      "step": 18501
+    },
+    {
+      "epoch": 4.911721757599894,
+      "grad_norm": 0.4996713772854791,
+      "learning_rate": 3.975871593855374e-07,
+      "loss": 0.5426,
+      "step": 18502
+    },
+    {
+      "epoch": 4.911987256073278,
+      "grad_norm": 0.4911547718273215,
+      "learning_rate": 3.97398270503796e-07,
+      "loss": 0.5713,
+      "step": 18503
+    },
+    {
+      "epoch": 4.912252754546661,
+      "grad_norm": 0.4791871555851505,
+      "learning_rate": 3.9720942262793714e-07,
+      "loss": 0.5428,
+      "step": 18504
+    },
+    {
+      "epoch": 4.912518253020045,
+      "grad_norm": 0.4776306721020352,
+      "learning_rate": 3.970206157616421e-07,
+      "loss": 0.5118,
+      "step": 18505
+    },
+    {
+      "epoch": 4.9127837514934285,
+      "grad_norm": 0.4671634164674486,
+      "learning_rate": 3.968318499085949e-07,
+      "loss": 0.5152,
+      "step": 18506
+    },
+    {
+      "epoch": 4.913049249966813,
+      "grad_norm": 0.4857562801231992,
+      "learning_rate": 3.966431250724756e-07,
+      "loss": 0.493,
+      "step": 18507
+    },
+    {
+      "epoch": 4.913314748440197,
+      "grad_norm": 0.478472167706853,
+      "learning_rate": 3.964544412569657e-07,
+      "loss": 0.5213,
+      "step": 18508
+    },
+    {
+      "epoch": 4.91358024691358,
+      "grad_norm": 0.48059785723801673,
+      "learning_rate": 3.962657984657445e-07,
+      "loss": 0.5216,
+      "step": 18509
+    },
+    {
+      "epoch": 4.913845745386964,
+      "grad_norm": 0.48665086487252285,
+      "learning_rate": 3.960771967024904e-07,
+      "loss": 0.5525,
+      "step": 18510
+    },
+    {
+      "epoch": 4.914111243860348,
+      "grad_norm": 0.7240881027538407,
+      "learning_rate": 3.95888635970883e-07,
+      "loss": 0.5419,
+      "step": 18511
+    },
+    {
+      "epoch": 4.914376742333731,
+      "grad_norm": 0.48476410048770946,
+      "learning_rate": 3.9570011627459826e-07,
+      "loss": 0.5297,
+      "step": 18512
+    },
+    {
+      "epoch": 4.914642240807115,
+      "grad_norm": 0.4824206757330293,
+      "learning_rate": 3.955116376173143e-07,
+      "loss": 0.516,
+      "step": 18513
+    },
+    {
+      "epoch": 4.914907739280499,
+      "grad_norm": 0.479489862742604,
+      "learning_rate": 3.953232000027057e-07,
+      "loss": 0.512,
+      "step": 18514
+    },
+    {
+      "epoch": 4.915173237753883,
+      "grad_norm": 0.48042652914045975,
+      "learning_rate": 3.9513480343444767e-07,
+      "loss": 0.5282,
+      "step": 18515
+    },
+    {
+      "epoch": 4.915438736227267,
+      "grad_norm": 0.476208601326996,
+      "learning_rate": 3.949464479162149e-07,
+      "loss": 0.5158,
+      "step": 18516
+    },
+    {
+      "epoch": 4.915704234700651,
+      "grad_norm": 0.4783775570938376,
+      "learning_rate": 3.947581334516798e-07,
+      "loss": 0.5315,
+      "step": 18517
+    },
+    {
+      "epoch": 4.915969733174034,
+      "grad_norm": 0.4825384710768808,
+      "learning_rate": 3.945698600445164e-07,
+      "loss": 0.5226,
+      "step": 18518
+    },
+    {
+      "epoch": 4.916235231647418,
+      "grad_norm": 0.517856000659449,
+      "learning_rate": 3.94381627698395e-07,
+      "loss": 0.5261,
+      "step": 18519
+    },
+    {
+      "epoch": 4.916500730120802,
+      "grad_norm": 0.46861481281274486,
+      "learning_rate": 3.941934364169886e-07,
+      "loss": 0.5395,
+      "step": 18520
+    },
+    {
+      "epoch": 4.916766228594185,
+      "grad_norm": 0.4792002069101839,
+      "learning_rate": 3.9400528620396476e-07,
+      "loss": 0.5536,
+      "step": 18521
+    },
+    {
+      "epoch": 4.917031727067569,
+      "grad_norm": 0.49079393897979434,
+      "learning_rate": 3.938171770629948e-07,
+      "loss": 0.525,
+      "step": 18522
+    },
+    {
+      "epoch": 4.9172972255409535,
+      "grad_norm": 0.47827088443541654,
+      "learning_rate": 3.936291089977462e-07,
+      "loss": 0.5375,
+      "step": 18523
+    },
+    {
+      "epoch": 4.917562724014337,
+      "grad_norm": 0.47453747008458924,
+      "learning_rate": 3.934410820118878e-07,
+      "loss": 0.5285,
+      "step": 18524
+    },
+    {
+      "epoch": 4.917828222487721,
+      "grad_norm": 0.495049605967072,
+      "learning_rate": 3.9325309610908605e-07,
+      "loss": 0.5319,
+      "step": 18525
+    },
+    {
+      "epoch": 4.918093720961105,
+      "grad_norm": 0.49075001426179393,
+      "learning_rate": 3.930651512930067e-07,
+      "loss": 0.5129,
+      "step": 18526
+    },
+    {
+      "epoch": 4.918359219434488,
+      "grad_norm": 0.49585670143567556,
+      "learning_rate": 3.928772475673159e-07,
+      "loss": 0.5748,
+      "step": 18527
+    },
+    {
+      "epoch": 4.918624717907872,
+      "grad_norm": 0.4867543719554076,
+      "learning_rate": 3.926893849356775e-07,
+      "loss": 0.4963,
+      "step": 18528
+    },
+    {
+      "epoch": 4.918890216381255,
+      "grad_norm": 0.4963868164907265,
+      "learning_rate": 3.925015634017562e-07,
+      "loss": 0.5401,
+      "step": 18529
+    },
+    {
+      "epoch": 4.919155714854639,
+      "grad_norm": 0.5000253852292696,
+      "learning_rate": 3.923137829692139e-07,
+      "loss": 0.5501,
+      "step": 18530
+    },
+    {
+      "epoch": 4.9194212133280235,
+      "grad_norm": 0.4868866511394403,
+      "learning_rate": 3.921260436417146e-07,
+      "loss": 0.5304,
+      "step": 18531
+    },
+    {
+      "epoch": 4.919686711801408,
+      "grad_norm": 0.47810845377949884,
+      "learning_rate": 3.9193834542291697e-07,
+      "loss": 0.5669,
+      "step": 18532
+    },
+    {
+      "epoch": 4.919952210274791,
+      "grad_norm": 0.4711032835708588,
+      "learning_rate": 3.91750688316484e-07,
+      "loss": 0.5353,
+      "step": 18533
+    },
+    {
+      "epoch": 4.920217708748175,
+      "grad_norm": 0.4676897210146569,
+      "learning_rate": 3.9156307232607356e-07,
+      "loss": 0.5191,
+      "step": 18534
+    },
+    {
+      "epoch": 4.920483207221558,
+      "grad_norm": 0.47762298468185405,
+      "learning_rate": 3.913754974553463e-07,
+      "loss": 0.5189,
+      "step": 18535
+    },
+    {
+      "epoch": 4.920748705694942,
+      "grad_norm": 0.4825479158136567,
+      "learning_rate": 3.911879637079591e-07,
+      "loss": 0.5777,
+      "step": 18536
+    },
+    {
+      "epoch": 4.921014204168326,
+      "grad_norm": 0.4930226277792407,
+      "learning_rate": 3.9100047108757035e-07,
+      "loss": 0.5164,
+      "step": 18537
+    },
+    {
+      "epoch": 4.921279702641709,
+      "grad_norm": 0.48990775286020416,
+      "learning_rate": 3.90813019597836e-07,
+      "loss": 0.5486,
+      "step": 18538
+    },
+    {
+      "epoch": 4.9215452011150935,
+      "grad_norm": 0.47574640914097976,
+      "learning_rate": 3.9062560924241157e-07,
+      "loss": 0.5236,
+      "step": 18539
+    },
+    {
+      "epoch": 4.921810699588478,
+      "grad_norm": 0.49012084744826817,
+      "learning_rate": 3.904382400249529e-07,
+      "loss": 0.4853,
+      "step": 18540
+    },
+    {
+      "epoch": 4.922076198061861,
+      "grad_norm": 0.48410955923556176,
+      "learning_rate": 3.9025091194911324e-07,
+      "loss": 0.5738,
+      "step": 18541
+    },
+    {
+      "epoch": 4.922341696535245,
+      "grad_norm": 0.4840025452121903,
+      "learning_rate": 3.900636250185469e-07,
+      "loss": 0.5732,
+      "step": 18542
+    },
+    {
+      "epoch": 4.922607195008629,
+      "grad_norm": 0.490437790818542,
+      "learning_rate": 3.8987637923690563e-07,
+      "loss": 0.5089,
+      "step": 18543
+    },
+    {
+      "epoch": 4.922872693482012,
+      "grad_norm": 0.48291252182453726,
+      "learning_rate": 3.89689174607841e-07,
+      "loss": 0.493,
+      "step": 18544
+    },
+    {
+      "epoch": 4.923138191955396,
+      "grad_norm": 0.47536743612066074,
+      "learning_rate": 3.8950201113500506e-07,
+      "loss": 0.5348,
+      "step": 18545
+    },
+    {
+      "epoch": 4.92340369042878,
+      "grad_norm": 0.4962383988934327,
+      "learning_rate": 3.893148888220466e-07,
+      "loss": 0.5904,
+      "step": 18546
+    },
+    {
+      "epoch": 4.9236691889021635,
+      "grad_norm": 0.4785118027013009,
+      "learning_rate": 3.8912780767261624e-07,
+      "loss": 0.5614,
+      "step": 18547
+    },
+    {
+      "epoch": 4.923934687375548,
+      "grad_norm": 0.4830810397040389,
+      "learning_rate": 3.88940767690362e-07,
+      "loss": 0.4981,
+      "step": 18548
+    },
+    {
+      "epoch": 4.924200185848932,
+      "grad_norm": 0.4821007701102928,
+      "learning_rate": 3.8875376887893167e-07,
+      "loss": 0.5406,
+      "step": 18549
+    },
+    {
+      "epoch": 4.924465684322315,
+      "grad_norm": 0.48888873471141336,
+      "learning_rate": 3.8856681124197134e-07,
+      "loss": 0.5424,
+      "step": 18550
+    },
+    {
+      "epoch": 4.924731182795699,
+      "grad_norm": 0.4805356512392803,
+      "learning_rate": 3.8837989478312826e-07,
+      "loss": 0.523,
+      "step": 18551
+    },
+    {
+      "epoch": 4.924996681269083,
+      "grad_norm": 0.47657229307970145,
+      "learning_rate": 3.881930195060471e-07,
+      "loss": 0.5525,
+      "step": 18552
+    },
+    {
+      "epoch": 4.925262179742466,
+      "grad_norm": 0.47415844691002984,
+      "learning_rate": 3.8800618541437325e-07,
+      "loss": 0.5011,
+      "step": 18553
+    },
+    {
+      "epoch": 4.92552767821585,
+      "grad_norm": 0.4828400353913829,
+      "learning_rate": 3.878193925117496e-07,
+      "loss": 0.5505,
+      "step": 18554
+    },
+    {
+      "epoch": 4.925793176689234,
+      "grad_norm": 0.47792941207804485,
+      "learning_rate": 3.8763264080181884e-07,
+      "loss": 0.5538,
+      "step": 18555
+    },
+    {
+      "epoch": 4.926058675162618,
+      "grad_norm": 0.48792268823134527,
+      "learning_rate": 3.8744593028822413e-07,
+      "loss": 0.5405,
+      "step": 18556
+    },
+    {
+      "epoch": 4.926324173636002,
+      "grad_norm": 0.4916759277318384,
+      "learning_rate": 3.8725926097460564e-07,
+      "loss": 0.5662,
+      "step": 18557
+    },
+    {
+      "epoch": 4.926589672109385,
+      "grad_norm": 0.48987155003560207,
+      "learning_rate": 3.870726328646049e-07,
+      "loss": 0.5165,
+      "step": 18558
+    },
+    {
+      "epoch": 4.926855170582769,
+      "grad_norm": 0.492260618175659,
+      "learning_rate": 3.8688604596186064e-07,
+      "loss": 0.587,
+      "step": 18559
+    },
+    {
+      "epoch": 4.927120669056153,
+      "grad_norm": 0.47788339944142905,
+      "learning_rate": 3.866995002700136e-07,
+      "loss": 0.5732,
+      "step": 18560
+    },
+    {
+      "epoch": 4.927386167529537,
+      "grad_norm": 0.5048354955337484,
+      "learning_rate": 3.865129957926991e-07,
+      "loss": 0.5508,
+      "step": 18561
+    },
+    {
+      "epoch": 4.92765166600292,
+      "grad_norm": 0.49001145767624876,
+      "learning_rate": 3.8632653253355634e-07,
+      "loss": 0.5358,
+      "step": 18562
+    },
+    {
+      "epoch": 4.9279171644763045,
+      "grad_norm": 0.48819152785113407,
+      "learning_rate": 3.8614011049622146e-07,
+      "loss": 0.5632,
+      "step": 18563
+    },
+    {
+      "epoch": 4.928182662949688,
+      "grad_norm": 0.49074566907092343,
+      "learning_rate": 3.8595372968432963e-07,
+      "loss": 0.5539,
+      "step": 18564
+    },
+    {
+      "epoch": 4.928448161423072,
+      "grad_norm": 0.4828841185215907,
+      "learning_rate": 3.8576739010151685e-07,
+      "loss": 0.5372,
+      "step": 18565
+    },
+    {
+      "epoch": 4.928713659896456,
+      "grad_norm": 0.47576791153936066,
+      "learning_rate": 3.8558109175141635e-07,
+      "loss": 0.5221,
+      "step": 18566
+    },
+    {
+      "epoch": 4.928979158369839,
+      "grad_norm": 0.4974036727332416,
+      "learning_rate": 3.853948346376618e-07,
+      "loss": 0.5419,
+      "step": 18567
+    },
+    {
+      "epoch": 4.929244656843223,
+      "grad_norm": 0.490460630744477,
+      "learning_rate": 3.8520861876388455e-07,
+      "loss": 0.541,
+      "step": 18568
+    },
+    {
+      "epoch": 4.929510155316607,
+      "grad_norm": 0.4822493102783241,
+      "learning_rate": 3.8502244413371777e-07,
+      "loss": 0.539,
+      "step": 18569
+    },
+    {
+      "epoch": 4.92977565378999,
+      "grad_norm": 0.4800024751663519,
+      "learning_rate": 3.848363107507913e-07,
+      "loss": 0.5533,
+      "step": 18570
+    },
+    {
+      "epoch": 4.9300411522633745,
+      "grad_norm": 0.48878815364254047,
+      "learning_rate": 3.846502186187359e-07,
+      "loss": 0.5371,
+      "step": 18571
+    },
+    {
+      "epoch": 4.9303066507367586,
+      "grad_norm": 0.4834995899126379,
+      "learning_rate": 3.8446416774118063e-07,
+      "loss": 0.5385,
+      "step": 18572
+    },
+    {
+      "epoch": 4.930572149210142,
+      "grad_norm": 0.49032832515294433,
+      "learning_rate": 3.8427815812175335e-07,
+      "loss": 0.5391,
+      "step": 18573
+    },
+    {
+      "epoch": 4.930837647683526,
+      "grad_norm": 0.482556768370718,
+      "learning_rate": 3.840921897640826e-07,
+      "loss": 0.534,
+      "step": 18574
+    },
+    {
+      "epoch": 4.93110314615691,
+      "grad_norm": 0.46460504353809134,
+      "learning_rate": 3.839062626717943e-07,
+      "loss": 0.513,
+      "step": 18575
+    },
+    {
+      "epoch": 4.931368644630293,
+      "grad_norm": 0.4790184507684636,
+      "learning_rate": 3.837203768485154e-07,
+      "loss": 0.5346,
+      "step": 18576
+    },
+    {
+      "epoch": 4.931634143103677,
+      "grad_norm": 0.4907319441185008,
+      "learning_rate": 3.835345322978709e-07,
+      "loss": 0.5443,
+      "step": 18577
+    },
+    {
+      "epoch": 4.931899641577061,
+      "grad_norm": 0.4806731454409003,
+      "learning_rate": 3.83348729023485e-07,
+      "loss": 0.5383,
+      "step": 18578
+    },
+    {
+      "epoch": 4.9321651400504445,
+      "grad_norm": 0.48863168197168405,
+      "learning_rate": 3.831629670289805e-07,
+      "loss": 0.5333,
+      "step": 18579
+    },
+    {
+      "epoch": 4.932430638523829,
+      "grad_norm": 0.47635623359569684,
+      "learning_rate": 3.829772463179818e-07,
+      "loss": 0.552,
+      "step": 18580
+    },
+    {
+      "epoch": 4.932696136997212,
+      "grad_norm": 0.4831409769858091,
+      "learning_rate": 3.827915668941096e-07,
+      "loss": 0.5185,
+      "step": 18581
+    },
+    {
+      "epoch": 4.932961635470596,
+      "grad_norm": 0.4998529380478762,
+      "learning_rate": 3.826059287609857e-07,
+      "loss": 0.538,
+      "step": 18582
+    },
+    {
+      "epoch": 4.93322713394398,
+      "grad_norm": 0.4773249131333924,
+      "learning_rate": 3.824203319222317e-07,
+      "loss": 0.5022,
+      "step": 18583
+    },
+    {
+      "epoch": 4.933492632417364,
+      "grad_norm": 0.49026793531777124,
+      "learning_rate": 3.822347763814646e-07,
+      "loss": 0.54,
+      "step": 18584
+    },
+    {
+      "epoch": 4.933758130890747,
+      "grad_norm": 0.4965747908967882,
+      "learning_rate": 3.8204926214230527e-07,
+      "loss": 0.5576,
+      "step": 18585
+    },
+    {
+      "epoch": 4.934023629364131,
+      "grad_norm": 0.475106955268532,
+      "learning_rate": 3.8186378920837014e-07,
+      "loss": 0.5404,
+      "step": 18586
+    },
+    {
+      "epoch": 4.9342891278375145,
+      "grad_norm": 0.48533335843139475,
+      "learning_rate": 3.8167835758327804e-07,
+      "loss": 0.5138,
+      "step": 18587
+    },
+    {
+      "epoch": 4.934554626310899,
+      "grad_norm": 0.47911340497081223,
+      "learning_rate": 3.8149296727064383e-07,
+      "loss": 0.5052,
+      "step": 18588
+    },
+    {
+      "epoch": 4.934820124784283,
+      "grad_norm": 0.4767558354845794,
+      "learning_rate": 3.8130761827408436e-07,
+      "loss": 0.5366,
+      "step": 18589
+    },
+    {
+      "epoch": 4.935085623257667,
+      "grad_norm": 0.4726929127109575,
+      "learning_rate": 3.811223105972137e-07,
+      "loss": 0.5295,
+      "step": 18590
+    },
+    {
+      "epoch": 4.93535112173105,
+      "grad_norm": 0.4894448944838308,
+      "learning_rate": 3.809370442436455e-07,
+      "loss": 0.5422,
+      "step": 18591
+    },
+    {
+      "epoch": 4.935616620204434,
+      "grad_norm": 0.49040063434492476,
+      "learning_rate": 3.8075181921699375e-07,
+      "loss": 0.5279,
+      "step": 18592
+    },
+    {
+      "epoch": 4.935882118677817,
+      "grad_norm": 0.4835052449953252,
+      "learning_rate": 3.8056663552086983e-07,
+      "loss": 0.5618,
+      "step": 18593
+    },
+    {
+      "epoch": 4.936147617151201,
+      "grad_norm": 0.46555451023887073,
+      "learning_rate": 3.8038149315888616e-07,
+      "loss": 0.4958,
+      "step": 18594
+    },
+    {
+      "epoch": 4.936413115624585,
+      "grad_norm": 0.4740254715261996,
+      "learning_rate": 3.8019639213465327e-07,
+      "loss": 0.5473,
+      "step": 18595
+    },
+    {
+      "epoch": 4.936678614097969,
+      "grad_norm": 0.46919658406170817,
+      "learning_rate": 3.8001133245178094e-07,
+      "loss": 0.5321,
+      "step": 18596
+    },
+    {
+      "epoch": 4.936944112571353,
+      "grad_norm": 0.4899852953691373,
+      "learning_rate": 3.7982631411387776e-07,
+      "loss": 0.5583,
+      "step": 18597
+    },
+    {
+      "epoch": 4.937209611044737,
+      "grad_norm": 0.4749674759234017,
+      "learning_rate": 3.7964133712455303e-07,
+      "loss": 0.4933,
+      "step": 18598
+    },
+    {
+      "epoch": 4.93747510951812,
+      "grad_norm": 0.47994563383834343,
+      "learning_rate": 3.7945640148741304e-07,
+      "loss": 0.5268,
+      "step": 18599
+    },
+    {
+      "epoch": 4.937740607991504,
+      "grad_norm": 0.49265929455979257,
+      "learning_rate": 3.7927150720606596e-07,
+      "loss": 0.5311,
+      "step": 18600
+    },
+    {
+      "epoch": 4.938006106464888,
+      "grad_norm": 0.4852296548860318,
+      "learning_rate": 3.790866542841168e-07,
+      "loss": 0.5131,
+      "step": 18601
+    },
+    {
+      "epoch": 4.938271604938271,
+      "grad_norm": 0.4742500964562554,
+      "learning_rate": 3.7890184272517005e-07,
+      "loss": 0.5448,
+      "step": 18602
+    },
+    {
+      "epoch": 4.938537103411655,
+      "grad_norm": 0.4582623340860457,
+      "learning_rate": 3.7871707253283154e-07,
+      "loss": 0.5277,
+      "step": 18603
+    },
+    {
+      "epoch": 4.9388026018850395,
+      "grad_norm": 0.48440467348833105,
+      "learning_rate": 3.785323437107033e-07,
+      "loss": 0.534,
+      "step": 18604
+    },
+    {
+      "epoch": 4.939068100358423,
+      "grad_norm": 0.48861207440362897,
+      "learning_rate": 3.7834765626238917e-07,
+      "loss": 0.5338,
+      "step": 18605
+    },
+    {
+      "epoch": 4.939333598831807,
+      "grad_norm": 0.4961318281342257,
+      "learning_rate": 3.7816301019149044e-07,
+      "loss": 0.5413,
+      "step": 18606
+    },
+    {
+      "epoch": 4.939599097305191,
+      "grad_norm": 0.4819360356898252,
+      "learning_rate": 3.7797840550160755e-07,
+      "loss": 0.5234,
+      "step": 18607
+    },
+    {
+      "epoch": 4.939864595778574,
+      "grad_norm": 0.4788953899918431,
+      "learning_rate": 3.7779384219634204e-07,
+      "loss": 0.4852,
+      "step": 18608
+    },
+    {
+      "epoch": 4.940130094251958,
+      "grad_norm": 0.4826652755359662,
+      "learning_rate": 3.776093202792919e-07,
+      "loss": 0.4755,
+      "step": 18609
+    },
+    {
+      "epoch": 4.940395592725341,
+      "grad_norm": 0.4876500860098827,
+      "learning_rate": 3.7742483975405724e-07,
+      "loss": 0.5183,
+      "step": 18610
+    },
+    {
+      "epoch": 4.940661091198725,
+      "grad_norm": 0.49113042006583135,
+      "learning_rate": 3.772404006242347e-07,
+      "loss": 0.5059,
+      "step": 18611
+    },
+    {
+      "epoch": 4.9409265896721095,
+      "grad_norm": 0.4856748783687735,
+      "learning_rate": 3.770560028934228e-07,
+      "loss": 0.5551,
+      "step": 18612
+    },
+    {
+      "epoch": 4.941192088145494,
+      "grad_norm": 0.4821366001985778,
+      "learning_rate": 3.768716465652153e-07,
+      "loss": 0.5774,
+      "step": 18613
+    },
+    {
+      "epoch": 4.941457586618877,
+      "grad_norm": 0.4722988821796132,
+      "learning_rate": 3.7668733164320985e-07,
+      "loss": 0.5304,
+      "step": 18614
+    },
+    {
+      "epoch": 4.941723085092261,
+      "grad_norm": 0.47421245594591194,
+      "learning_rate": 3.765030581309997e-07,
+      "loss": 0.4971,
+      "step": 18615
+    },
+    {
+      "epoch": 4.941988583565644,
+      "grad_norm": 0.46324693720813953,
+      "learning_rate": 3.763188260321796e-07,
+      "loss": 0.528,
+      "step": 18616
+    },
+    {
+      "epoch": 4.942254082039028,
+      "grad_norm": 0.4819911822591266,
+      "learning_rate": 3.7613463535034146e-07,
+      "loss": 0.5153,
+      "step": 18617
+    },
+    {
+      "epoch": 4.942519580512412,
+      "grad_norm": 0.4673841807184212,
+      "learning_rate": 3.7595048608907863e-07,
+      "loss": 0.5112,
+      "step": 18618
+    },
+    {
+      "epoch": 4.942785078985795,
+      "grad_norm": 0.47966255228515575,
+      "learning_rate": 3.7576637825198185e-07,
+      "loss": 0.5409,
+      "step": 18619
+    },
+    {
+      "epoch": 4.9430505774591795,
+      "grad_norm": 0.47642925619490206,
+      "learning_rate": 3.7558231184264134e-07,
+      "loss": 0.5504,
+      "step": 18620
+    },
+    {
+      "epoch": 4.943316075932564,
+      "grad_norm": 0.4889325007237682,
+      "learning_rate": 3.753982868646475e-07,
+      "loss": 0.5407,
+      "step": 18621
+    },
+    {
+      "epoch": 4.943581574405947,
+      "grad_norm": 0.4861365265331885,
+      "learning_rate": 3.7521430332158853e-07,
+      "loss": 0.555,
+      "step": 18622
+    },
+    {
+      "epoch": 4.943847072879331,
+      "grad_norm": 0.48502231027701137,
+      "learning_rate": 3.750303612170542e-07,
+      "loss": 0.5213,
+      "step": 18623
+    },
+    {
+      "epoch": 4.944112571352715,
+      "grad_norm": 0.49282846907661104,
+      "learning_rate": 3.748464605546295e-07,
+      "loss": 0.5166,
+      "step": 18624
+    },
+    {
+      "epoch": 4.944378069826098,
+      "grad_norm": 0.47715250605183523,
+      "learning_rate": 3.746626013379026e-07,
+      "loss": 0.5516,
+      "step": 18625
+    },
+    {
+      "epoch": 4.944643568299482,
+      "grad_norm": 0.4771917755850158,
+      "learning_rate": 3.744787835704583e-07,
+      "loss": 0.5525,
+      "step": 18626
+    },
+    {
+      "epoch": 4.944909066772866,
+      "grad_norm": 0.47555438851240833,
+      "learning_rate": 3.7429500725588183e-07,
+      "loss": 0.5251,
+      "step": 18627
+    },
+    {
+      "epoch": 4.9451745652462495,
+      "grad_norm": 0.4779196693292834,
+      "learning_rate": 3.7411127239775774e-07,
+      "loss": 0.5109,
+      "step": 18628
+    },
+    {
+      "epoch": 4.945440063719634,
+      "grad_norm": 0.49224439961885713,
+      "learning_rate": 3.739275789996691e-07,
+      "loss": 0.5273,
+      "step": 18629
+    },
+    {
+      "epoch": 4.945705562193018,
+      "grad_norm": 0.4852469797287295,
+      "learning_rate": 3.7374392706519826e-07,
+      "loss": 0.4969,
+      "step": 18630
+    },
+    {
+      "epoch": 4.945971060666401,
+      "grad_norm": 0.4809848470611654,
+      "learning_rate": 3.7356031659792596e-07,
+      "loss": 0.5173,
+      "step": 18631
+    },
+    {
+      "epoch": 4.946236559139785,
+      "grad_norm": 0.48931889078185314,
+      "learning_rate": 3.7337674760143466e-07,
+      "loss": 0.5271,
+      "step": 18632
+    },
+    {
+      "epoch": 4.946502057613169,
+      "grad_norm": 0.48401441867161266,
+      "learning_rate": 3.7319322007930314e-07,
+      "loss": 0.5388,
+      "step": 18633
+    },
+    {
+      "epoch": 4.946767556086552,
+      "grad_norm": 0.4849443653856644,
+      "learning_rate": 3.730097340351116e-07,
+      "loss": 0.5544,
+      "step": 18634
+    },
+    {
+      "epoch": 4.947033054559936,
+      "grad_norm": 0.4867101828063559,
+      "learning_rate": 3.72826289472438e-07,
+      "loss": 0.533,
+      "step": 18635
+    },
+    {
+      "epoch": 4.9472985530333204,
+      "grad_norm": 0.49454880277379515,
+      "learning_rate": 3.7264288639485946e-07,
+      "loss": 0.543,
+      "step": 18636
+    },
+    {
+      "epoch": 4.947564051506704,
+      "grad_norm": 0.4758699611430989,
+      "learning_rate": 3.72459524805954e-07,
+      "loss": 0.5226,
+      "step": 18637
+    },
+    {
+      "epoch": 4.947829549980088,
+      "grad_norm": 0.4687696815904308,
+      "learning_rate": 3.722762047092962e-07,
+      "loss": 0.5188,
+      "step": 18638
+    },
+    {
+      "epoch": 4.948095048453471,
+      "grad_norm": 0.48853154596075177,
+      "learning_rate": 3.7209292610846243e-07,
+      "loss": 0.5516,
+      "step": 18639
+    },
+    {
+      "epoch": 4.948360546926855,
+      "grad_norm": 0.5109172374636374,
+      "learning_rate": 3.719096890070262e-07,
+      "loss": 0.5386,
+      "step": 18640
+    },
+    {
+      "epoch": 4.948626045400239,
+      "grad_norm": 0.4851181904669391,
+      "learning_rate": 3.7172649340856263e-07,
+      "loss": 0.527,
+      "step": 18641
+    },
+    {
+      "epoch": 4.948891543873623,
+      "grad_norm": 0.4737321112101461,
+      "learning_rate": 3.7154333931664235e-07,
+      "loss": 0.5092,
+      "step": 18642
+    },
+    {
+      "epoch": 4.949157042347006,
+      "grad_norm": 0.4798326051425066,
+      "learning_rate": 3.7136022673483876e-07,
+      "loss": 0.5163,
+      "step": 18643
+    },
+    {
+      "epoch": 4.9494225408203905,
+      "grad_norm": 0.4850624259823333,
+      "learning_rate": 3.711771556667218e-07,
+      "loss": 0.5029,
+      "step": 18644
+    },
+    {
+      "epoch": 4.949688039293774,
+      "grad_norm": 0.4783642148622547,
+      "learning_rate": 3.7099412611586336e-07,
+      "loss": 0.5191,
+      "step": 18645
+    },
+    {
+      "epoch": 4.949953537767158,
+      "grad_norm": 0.49106768459088873,
+      "learning_rate": 3.7081113808583204e-07,
+      "loss": 0.5413,
+      "step": 18646
+    },
+    {
+      "epoch": 4.950219036240542,
+      "grad_norm": 0.49034099586984053,
+      "learning_rate": 3.7062819158019627e-07,
+      "loss": 0.5311,
+      "step": 18647
+    },
+    {
+      "epoch": 4.950484534713925,
+      "grad_norm": 0.4915936222637951,
+      "learning_rate": 3.704452866025246e-07,
+      "loss": 0.5321,
+      "step": 18648
+    },
+    {
+      "epoch": 4.950750033187309,
+      "grad_norm": 0.4774199709177357,
+      "learning_rate": 3.7026242315638365e-07,
+      "loss": 0.4973,
+      "step": 18649
+    },
+    {
+      "epoch": 4.951015531660693,
+      "grad_norm": 0.4786519037140113,
+      "learning_rate": 3.700796012453406e-07,
+      "loss": 0.5724,
+      "step": 18650
+    },
+    {
+      "epoch": 4.951281030134076,
+      "grad_norm": 0.47887300831780266,
+      "learning_rate": 3.6989682087295946e-07,
+      "loss": 0.5099,
+      "step": 18651
+    },
+    {
+      "epoch": 4.9515465286074605,
+      "grad_norm": 0.4847848513893285,
+      "learning_rate": 3.697140820428063e-07,
+      "loss": 0.5332,
+      "step": 18652
+    },
+    {
+      "epoch": 4.951812027080845,
+      "grad_norm": 0.4807975944993597,
+      "learning_rate": 3.6953138475844446e-07,
+      "loss": 0.5629,
+      "step": 18653
+    },
+    {
+      "epoch": 4.952077525554228,
+      "grad_norm": 0.4904485424891691,
+      "learning_rate": 3.693487290234363e-07,
+      "loss": 0.5351,
+      "step": 18654
+    },
+    {
+      "epoch": 4.952343024027612,
+      "grad_norm": 0.47240343450083344,
+      "learning_rate": 3.69166114841345e-07,
+      "loss": 0.5528,
+      "step": 18655
+    },
+    {
+      "epoch": 4.952608522500996,
+      "grad_norm": 0.4836608213456476,
+      "learning_rate": 3.6898354221573145e-07,
+      "loss": 0.5528,
+      "step": 18656
+    },
+    {
+      "epoch": 4.952874020974379,
+      "grad_norm": 0.4888590013267215,
+      "learning_rate": 3.6880101115015695e-07,
+      "loss": 0.5611,
+      "step": 18657
+    },
+    {
+      "epoch": 4.953139519447763,
+      "grad_norm": 0.4756328682507688,
+      "learning_rate": 3.6861852164818054e-07,
+      "loss": 0.4875,
+      "step": 18658
+    },
+    {
+      "epoch": 4.953405017921147,
+      "grad_norm": 0.4729403962985496,
+      "learning_rate": 3.684360737133616e-07,
+      "loss": 0.5167,
+      "step": 18659
+    },
+    {
+      "epoch": 4.9536705163945305,
+      "grad_norm": 0.48789278788718127,
+      "learning_rate": 3.682536673492579e-07,
+      "loss": 0.5412,
+      "step": 18660
+    },
+    {
+      "epoch": 4.953936014867915,
+      "grad_norm": 0.4959424916247629,
+      "learning_rate": 3.6807130255942735e-07,
+      "loss": 0.5515,
+      "step": 18661
+    },
+    {
+      "epoch": 4.954201513341299,
+      "grad_norm": 0.4838180021637111,
+      "learning_rate": 3.678889793474258e-07,
+      "loss": 0.534,
+      "step": 18662
+    },
+    {
+      "epoch": 4.954467011814682,
+      "grad_norm": 0.48986983561832637,
+      "learning_rate": 3.6770669771681003e-07,
+      "loss": 0.4888,
+      "step": 18663
+    },
+    {
+      "epoch": 4.954732510288066,
+      "grad_norm": 0.4821046560728604,
+      "learning_rate": 3.6752445767113424e-07,
+      "loss": 0.5283,
+      "step": 18664
+    },
+    {
+      "epoch": 4.95499800876145,
+      "grad_norm": 0.4821905608386961,
+      "learning_rate": 3.673422592139525e-07,
+      "loss": 0.535,
+      "step": 18665
+    },
+    {
+      "epoch": 4.955263507234833,
+      "grad_norm": 0.4758972218483108,
+      "learning_rate": 3.671601023488186e-07,
+      "loss": 0.5463,
+      "step": 18666
+    },
+    {
+      "epoch": 4.955529005708217,
+      "grad_norm": 0.48950219014966956,
+      "learning_rate": 3.669779870792844e-07,
+      "loss": 0.5524,
+      "step": 18667
+    },
+    {
+      "epoch": 4.9557945041816005,
+      "grad_norm": 0.5017984247641869,
+      "learning_rate": 3.667959134089025e-07,
+      "loss": 0.5392,
+      "step": 18668
+    },
+    {
+      "epoch": 4.956060002654985,
+      "grad_norm": 0.4885380615647711,
+      "learning_rate": 3.6661388134122343e-07,
+      "loss": 0.549,
+      "step": 18669
+    },
+    {
+      "epoch": 4.956325501128369,
+      "grad_norm": 0.48157185421652493,
+      "learning_rate": 3.6643189087979704e-07,
+      "loss": 0.4585,
+      "step": 18670
+    },
+    {
+      "epoch": 4.956590999601753,
+      "grad_norm": 0.495493526089683,
+      "learning_rate": 3.6624994202817204e-07,
+      "loss": 0.5584,
+      "step": 18671
+    },
+    {
+      "epoch": 4.956856498075136,
+      "grad_norm": 0.4886310663845349,
+      "learning_rate": 3.660680347898979e-07,
+      "loss": 0.5748,
+      "step": 18672
+    },
+    {
+      "epoch": 4.95712199654852,
+      "grad_norm": 0.4818810479469065,
+      "learning_rate": 3.658861691685214e-07,
+      "loss": 0.5519,
+      "step": 18673
+    },
+    {
+      "epoch": 4.957387495021903,
+      "grad_norm": 0.47457472355625385,
+      "learning_rate": 3.6570434516758985e-07,
+      "loss": 0.5357,
+      "step": 18674
+    },
+    {
+      "epoch": 4.957652993495287,
+      "grad_norm": 0.49376838512270477,
+      "learning_rate": 3.655225627906503e-07,
+      "loss": 0.5243,
+      "step": 18675
+    },
+    {
+      "epoch": 4.957918491968671,
+      "grad_norm": 0.4825921600878026,
+      "learning_rate": 3.6534082204124577e-07,
+      "loss": 0.5052,
+      "step": 18676
+    },
+    {
+      "epoch": 4.958183990442055,
+      "grad_norm": 0.4937926286626435,
+      "learning_rate": 3.6515912292292235e-07,
+      "loss": 0.5628,
+      "step": 18677
+    },
+    {
+      "epoch": 4.958449488915439,
+      "grad_norm": 0.4842924655380837,
+      "learning_rate": 3.649774654392224e-07,
+      "loss": 0.5988,
+      "step": 18678
+    },
+    {
+      "epoch": 4.958714987388823,
+      "grad_norm": 0.47252435383699554,
+      "learning_rate": 3.647958495936901e-07,
+      "loss": 0.5293,
+      "step": 18679
+    },
+    {
+      "epoch": 4.958980485862206,
+      "grad_norm": 0.4937646149317888,
+      "learning_rate": 3.6461427538986574e-07,
+      "loss": 0.5307,
+      "step": 18680
+    },
+    {
+      "epoch": 4.95924598433559,
+      "grad_norm": 0.4859138426198157,
+      "learning_rate": 3.6443274283129215e-07,
+      "loss": 0.4955,
+      "step": 18681
+    },
+    {
+      "epoch": 4.959511482808974,
+      "grad_norm": 0.4823319912277644,
+      "learning_rate": 3.6425125192150856e-07,
+      "loss": 0.5136,
+      "step": 18682
+    },
+    {
+      "epoch": 4.959776981282357,
+      "grad_norm": 0.4923013154480677,
+      "learning_rate": 3.6406980266405464e-07,
+      "loss": 0.5313,
+      "step": 18683
+    },
+    {
+      "epoch": 4.960042479755741,
+      "grad_norm": 0.4968554047916396,
+      "learning_rate": 3.6388839506246943e-07,
+      "loss": 0.5593,
+      "step": 18684
+    },
+    {
+      "epoch": 4.9603079782291255,
+      "grad_norm": 0.4721044134151305,
+      "learning_rate": 3.6370702912029015e-07,
+      "loss": 0.495,
+      "step": 18685
+    },
+    {
+      "epoch": 4.960573476702509,
+      "grad_norm": 0.49744630719236416,
+      "learning_rate": 3.635257048410548e-07,
+      "loss": 0.5475,
+      "step": 18686
+    },
+    {
+      "epoch": 4.960838975175893,
+      "grad_norm": 0.47521184458999854,
+      "learning_rate": 3.633444222282995e-07,
+      "loss": 0.5199,
+      "step": 18687
+    },
+    {
+      "epoch": 4.961104473649277,
+      "grad_norm": 0.4783360711237551,
+      "learning_rate": 3.6316318128555903e-07,
+      "loss": 0.4891,
+      "step": 18688
+    },
+    {
+      "epoch": 4.96136997212266,
+      "grad_norm": 0.464569261198877,
+      "learning_rate": 3.62981982016368e-07,
+      "loss": 0.5092,
+      "step": 18689
+    },
+    {
+      "epoch": 4.961635470596044,
+      "grad_norm": 0.4909053327965359,
+      "learning_rate": 3.6280082442426146e-07,
+      "loss": 0.5281,
+      "step": 18690
+    },
+    {
+      "epoch": 4.961900969069428,
+      "grad_norm": 0.48548796997054855,
+      "learning_rate": 3.6261970851277056e-07,
+      "loss": 0.5389,
+      "step": 18691
+    },
+    {
+      "epoch": 4.962166467542811,
+      "grad_norm": 0.46895375443214554,
+      "learning_rate": 3.6243863428542943e-07,
+      "loss": 0.5035,
+      "step": 18692
+    },
+    {
+      "epoch": 4.9624319660161955,
+      "grad_norm": 0.4868794431415527,
+      "learning_rate": 3.622576017457685e-07,
+      "loss": 0.5326,
+      "step": 18693
+    },
+    {
+      "epoch": 4.96269746448958,
+      "grad_norm": 0.4847308208503863,
+      "learning_rate": 3.620766108973181e-07,
+      "loss": 0.5409,
+      "step": 18694
+    },
+    {
+      "epoch": 4.962962962962963,
+      "grad_norm": 0.4945754614439959,
+      "learning_rate": 3.618956617436087e-07,
+      "loss": 0.5375,
+      "step": 18695
+    },
+    {
+      "epoch": 4.963228461436347,
+      "grad_norm": 0.48248365827032014,
+      "learning_rate": 3.6171475428816825e-07,
+      "loss": 0.5582,
+      "step": 18696
+    },
+    {
+      "epoch": 4.96349395990973,
+      "grad_norm": 0.48600160698443134,
+      "learning_rate": 3.6153388853452597e-07,
+      "loss": 0.5374,
+      "step": 18697
+    },
+    {
+      "epoch": 4.963759458383114,
+      "grad_norm": 0.46848027781391877,
+      "learning_rate": 3.6135306448620897e-07,
+      "loss": 0.5311,
+      "step": 18698
+    },
+    {
+      "epoch": 4.964024956856498,
+      "grad_norm": 0.4877257568714426,
+      "learning_rate": 3.611722821467428e-07,
+      "loss": 0.5559,
+      "step": 18699
+    },
+    {
+      "epoch": 4.964290455329882,
+      "grad_norm": 0.48128282509655346,
+      "learning_rate": 3.6099154151965464e-07,
+      "loss": 0.5357,
+      "step": 18700
+    },
+    {
+      "epoch": 4.9645559538032655,
+      "grad_norm": 0.4846458357851168,
+      "learning_rate": 3.6081084260846807e-07,
+      "loss": 0.5463,
+      "step": 18701
+    },
+    {
+      "epoch": 4.96482145227665,
+      "grad_norm": 0.4789873477100868,
+      "learning_rate": 3.6063018541670797e-07,
+      "loss": 0.5443,
+      "step": 18702
+    },
+    {
+      "epoch": 4.965086950750033,
+      "grad_norm": 0.48242243812768804,
+      "learning_rate": 3.604495699478969e-07,
+      "loss": 0.5223,
+      "step": 18703
+    },
+    {
+      "epoch": 4.965352449223417,
+      "grad_norm": 0.4728823077510297,
+      "learning_rate": 3.602689962055589e-07,
+      "loss": 0.5364,
+      "step": 18704
+    },
+    {
+      "epoch": 4.965617947696801,
+      "grad_norm": 0.47092999274712277,
+      "learning_rate": 3.6008846419321343e-07,
+      "loss": 0.5207,
+      "step": 18705
+    },
+    {
+      "epoch": 4.965883446170184,
+      "grad_norm": 0.4803962799714533,
+      "learning_rate": 3.5990797391438266e-07,
+      "loss": 0.5783,
+      "step": 18706
+    },
+    {
+      "epoch": 4.966148944643568,
+      "grad_norm": 0.48301418984580624,
+      "learning_rate": 3.59727525372586e-07,
+      "loss": 0.5375,
+      "step": 18707
+    },
+    {
+      "epoch": 4.966414443116952,
+      "grad_norm": 0.4881872282437468,
+      "learning_rate": 3.595471185713431e-07,
+      "loss": 0.5293,
+      "step": 18708
+    },
+    {
+      "epoch": 4.9666799415903355,
+      "grad_norm": 0.4819097289209797,
+      "learning_rate": 3.5936675351417176e-07,
+      "loss": 0.5505,
+      "step": 18709
+    },
+    {
+      "epoch": 4.96694544006372,
+      "grad_norm": 0.49414411746553144,
+      "learning_rate": 3.5918643020459027e-07,
+      "loss": 0.545,
+      "step": 18710
+    },
+    {
+      "epoch": 4.967210938537104,
+      "grad_norm": 0.4750403270674448,
+      "learning_rate": 3.5900614864611524e-07,
+      "loss": 0.5613,
+      "step": 18711
+    },
+    {
+      "epoch": 4.967476437010487,
+      "grad_norm": 0.4747875007621468,
+      "learning_rate": 3.588259088422616e-07,
+      "loss": 0.513,
+      "step": 18712
+    },
+    {
+      "epoch": 4.967741935483871,
+      "grad_norm": 0.47660142663826566,
+      "learning_rate": 3.58645710796546e-07,
+      "loss": 0.5598,
+      "step": 18713
+    },
+    {
+      "epoch": 4.968007433957255,
+      "grad_norm": 0.4800378624514367,
+      "learning_rate": 3.5846555451248127e-07,
+      "loss": 0.5207,
+      "step": 18714
+    },
+    {
+      "epoch": 4.968272932430638,
+      "grad_norm": 0.4835767165732828,
+      "learning_rate": 3.582854399935823e-07,
+      "loss": 0.5502,
+      "step": 18715
+    },
+    {
+      "epoch": 4.968538430904022,
+      "grad_norm": 0.48918076458837106,
+      "learning_rate": 3.58105367243361e-07,
+      "loss": 0.5182,
+      "step": 18716
+    },
+    {
+      "epoch": 4.9688039293774064,
+      "grad_norm": 0.4796889758540998,
+      "learning_rate": 3.5792533626532926e-07,
+      "loss": 0.514,
+      "step": 18717
+    },
+    {
+      "epoch": 4.96906942785079,
+      "grad_norm": 0.48838156494881524,
+      "learning_rate": 3.577453470629977e-07,
+      "loss": 0.5412,
+      "step": 18718
+    },
+    {
+      "epoch": 4.969334926324174,
+      "grad_norm": 0.4854446964625709,
+      "learning_rate": 3.575653996398767e-07,
+      "loss": 0.5525,
+      "step": 18719
+    },
+    {
+      "epoch": 4.969600424797557,
+      "grad_norm": 0.4679947403581326,
+      "learning_rate": 3.5738549399947697e-07,
+      "loss": 0.5183,
+      "step": 18720
+    },
+    {
+      "epoch": 4.969865923270941,
+      "grad_norm": 0.4810974962719777,
+      "learning_rate": 3.5720563014530584e-07,
+      "loss": 0.5326,
+      "step": 18721
+    },
+    {
+      "epoch": 4.970131421744325,
+      "grad_norm": 0.48156671568334447,
+      "learning_rate": 3.570258080808717e-07,
+      "loss": 0.5424,
+      "step": 18722
+    },
+    {
+      "epoch": 4.970396920217709,
+      "grad_norm": 0.4814013087643834,
+      "learning_rate": 3.5684602780968023e-07,
+      "loss": 0.5131,
+      "step": 18723
+    },
+    {
+      "epoch": 4.970662418691092,
+      "grad_norm": 0.48730841724383445,
+      "learning_rate": 3.566662893352396e-07,
+      "loss": 0.5178,
+      "step": 18724
+    },
+    {
+      "epoch": 4.9709279171644765,
+      "grad_norm": 0.4903040381872622,
+      "learning_rate": 3.5648659266105306e-07,
+      "loss": 0.5462,
+      "step": 18725
+    },
+    {
+      "epoch": 4.97119341563786,
+      "grad_norm": 0.498251460336798,
+      "learning_rate": 3.563069377906273e-07,
+      "loss": 0.5628,
+      "step": 18726
+    },
+    {
+      "epoch": 4.971458914111244,
+      "grad_norm": 0.468467496122541,
+      "learning_rate": 3.561273247274644e-07,
+      "loss": 0.5489,
+      "step": 18727
+    },
+    {
+      "epoch": 4.971724412584628,
+      "grad_norm": 0.4743843567876023,
+      "learning_rate": 3.5594775347506744e-07,
+      "loss": 0.5056,
+      "step": 18728
+    },
+    {
+      "epoch": 4.971989911058012,
+      "grad_norm": 0.4728286225134879,
+      "learning_rate": 3.557682240369395e-07,
+      "loss": 0.5471,
+      "step": 18729
+    },
+    {
+      "epoch": 4.972255409531395,
+      "grad_norm": 0.47712501737487595,
+      "learning_rate": 3.5558873641658053e-07,
+      "loss": 0.535,
+      "step": 18730
+    },
+    {
+      "epoch": 4.972520908004779,
+      "grad_norm": 0.4880995126518145,
+      "learning_rate": 3.554092906174919e-07,
+      "loss": 0.5248,
+      "step": 18731
+    },
+    {
+      "epoch": 4.972786406478162,
+      "grad_norm": 0.4802526080105834,
+      "learning_rate": 3.5522988664317276e-07,
+      "loss": 0.5172,
+      "step": 18732
+    },
+    {
+      "epoch": 4.9730519049515465,
+      "grad_norm": 0.48035825422141615,
+      "learning_rate": 3.550505244971231e-07,
+      "loss": 0.5422,
+      "step": 18733
+    },
+    {
+      "epoch": 4.973317403424931,
+      "grad_norm": 0.49948058736344414,
+      "learning_rate": 3.548712041828389e-07,
+      "loss": 0.5496,
+      "step": 18734
+    },
+    {
+      "epoch": 4.973582901898314,
+      "grad_norm": 0.4810691356135354,
+      "learning_rate": 3.5469192570381895e-07,
+      "loss": 0.5364,
+      "step": 18735
+    },
+    {
+      "epoch": 4.973848400371698,
+      "grad_norm": 0.4919905711942463,
+      "learning_rate": 3.5451268906355863e-07,
+      "loss": 0.532,
+      "step": 18736
+    },
+    {
+      "epoch": 4.974113898845082,
+      "grad_norm": 0.4745660901866937,
+      "learning_rate": 3.5433349426555435e-07,
+      "loss": 0.5435,
+      "step": 18737
+    },
+    {
+      "epoch": 4.974379397318465,
+      "grad_norm": 0.46926626011927774,
+      "learning_rate": 3.5415434131329975e-07,
+      "loss": 0.5377,
+      "step": 18738
+    },
+    {
+      "epoch": 4.974644895791849,
+      "grad_norm": 0.48352491841440093,
+      "learning_rate": 3.539752302102903e-07,
+      "loss": 0.5355,
+      "step": 18739
+    },
+    {
+      "epoch": 4.974910394265233,
+      "grad_norm": 0.4892084017691704,
+      "learning_rate": 3.5379616096001793e-07,
+      "loss": 0.553,
+      "step": 18740
+    },
+    {
+      "epoch": 4.9751758927386165,
+      "grad_norm": 0.4709332902635177,
+      "learning_rate": 3.536171335659749e-07,
+      "loss": 0.5386,
+      "step": 18741
+    },
+    {
+      "epoch": 4.975441391212001,
+      "grad_norm": 0.49728074117137816,
+      "learning_rate": 3.5343814803165343e-07,
+      "loss": 0.5642,
+      "step": 18742
+    },
+    {
+      "epoch": 4.975706889685385,
+      "grad_norm": 0.47353901253014874,
+      "learning_rate": 3.5325920436054317e-07,
+      "loss": 0.5057,
+      "step": 18743
+    },
+    {
+      "epoch": 4.975972388158768,
+      "grad_norm": 0.48343660607836414,
+      "learning_rate": 3.530803025561352e-07,
+      "loss": 0.5089,
+      "step": 18744
+    },
+    {
+      "epoch": 4.976237886632152,
+      "grad_norm": 0.47010537631113863,
+      "learning_rate": 3.5290144262191793e-07,
+      "loss": 0.5004,
+      "step": 18745
+    },
+    {
+      "epoch": 4.976503385105536,
+      "grad_norm": 0.4854216007989131,
+      "learning_rate": 3.5272262456137907e-07,
+      "loss": 0.5064,
+      "step": 18746
+    },
+    {
+      "epoch": 4.976768883578919,
+      "grad_norm": 0.486344566893522,
+      "learning_rate": 3.5254384837800665e-07,
+      "loss": 0.5166,
+      "step": 18747
+    },
+    {
+      "epoch": 4.977034382052303,
+      "grad_norm": 0.47866713880850925,
+      "learning_rate": 3.5236511407528676e-07,
+      "loss": 0.5254,
+      "step": 18748
+    },
+    {
+      "epoch": 4.9772998805256865,
+      "grad_norm": 0.48422343487697445,
+      "learning_rate": 3.5218642165670615e-07,
+      "loss": 0.5174,
+      "step": 18749
+    },
+    {
+      "epoch": 4.977565378999071,
+      "grad_norm": 0.4775095961397552,
+      "learning_rate": 3.5200777112574887e-07,
+      "loss": 0.5485,
+      "step": 18750
+    },
+    {
+      "epoch": 4.977830877472455,
+      "grad_norm": 0.47522313546915,
+      "learning_rate": 3.5182916248589936e-07,
+      "loss": 0.5086,
+      "step": 18751
+    },
+    {
+      "epoch": 4.978096375945839,
+      "grad_norm": 0.47745446120793833,
+      "learning_rate": 3.516505957406402e-07,
+      "loss": 0.5324,
+      "step": 18752
+    },
+    {
+      "epoch": 4.978361874419222,
+      "grad_norm": 0.4768804011263821,
+      "learning_rate": 3.5147207089345517e-07,
+      "loss": 0.5602,
+      "step": 18753
+    },
+    {
+      "epoch": 4.978627372892606,
+      "grad_norm": 0.4837562991864166,
+      "learning_rate": 3.5129358794782437e-07,
+      "loss": 0.5799,
+      "step": 18754
+    },
+    {
+      "epoch": 4.978892871365989,
+      "grad_norm": 0.48729229971609395,
+      "learning_rate": 3.5111514690723047e-07,
+      "loss": 0.5647,
+      "step": 18755
+    },
+    {
+      "epoch": 4.979158369839373,
+      "grad_norm": 0.48636862146841775,
+      "learning_rate": 3.509367477751524e-07,
+      "loss": 0.5475,
+      "step": 18756
+    },
+    {
+      "epoch": 4.979423868312757,
+      "grad_norm": 0.480310178091797,
+      "learning_rate": 3.5075839055506904e-07,
+      "loss": 0.5531,
+      "step": 18757
+    },
+    {
+      "epoch": 4.979689366786141,
+      "grad_norm": 0.47727235095674175,
+      "learning_rate": 3.505800752504598e-07,
+      "loss": 0.4967,
+      "step": 18758
+    },
+    {
+      "epoch": 4.979954865259525,
+      "grad_norm": 0.48298450880407545,
+      "learning_rate": 3.5040180186480146e-07,
+      "loss": 0.5443,
+      "step": 18759
+    },
+    {
+      "epoch": 4.980220363732909,
+      "grad_norm": 0.48515029827444145,
+      "learning_rate": 3.5022357040157133e-07,
+      "loss": 0.5296,
+      "step": 18760
+    },
+    {
+      "epoch": 4.980485862206292,
+      "grad_norm": 0.48182284254107854,
+      "learning_rate": 3.500453808642448e-07,
+      "loss": 0.5453,
+      "step": 18761
+    },
+    {
+      "epoch": 4.980751360679676,
+      "grad_norm": 0.49290036154168454,
+      "learning_rate": 3.4986723325629826e-07,
+      "loss": 0.5473,
+      "step": 18762
+    },
+    {
+      "epoch": 4.98101685915306,
+      "grad_norm": 0.4689320893074932,
+      "learning_rate": 3.496891275812042e-07,
+      "loss": 0.537,
+      "step": 18763
+    },
+    {
+      "epoch": 4.981282357626443,
+      "grad_norm": 0.4864500505214823,
+      "learning_rate": 3.4951106384243705e-07,
+      "loss": 0.5539,
+      "step": 18764
+    },
+    {
+      "epoch": 4.981547856099827,
+      "grad_norm": 0.4788829032691235,
+      "learning_rate": 3.493330420434698e-07,
+      "loss": 0.5402,
+      "step": 18765
+    },
+    {
+      "epoch": 4.9818133545732115,
+      "grad_norm": 0.49873468843221863,
+      "learning_rate": 3.491550621877737e-07,
+      "loss": 0.5356,
+      "step": 18766
+    },
+    {
+      "epoch": 4.982078853046595,
+      "grad_norm": 0.47589641311019887,
+      "learning_rate": 3.489771242788209e-07,
+      "loss": 0.4985,
+      "step": 18767
+    },
+    {
+      "epoch": 4.982344351519979,
+      "grad_norm": 0.47754316253841833,
+      "learning_rate": 3.4879922832007973e-07,
+      "loss": 0.514,
+      "step": 18768
+    },
+    {
+      "epoch": 4.982609849993363,
+      "grad_norm": 0.49102389800488977,
+      "learning_rate": 3.486213743150213e-07,
+      "loss": 0.5825,
+      "step": 18769
+    },
+    {
+      "epoch": 4.982875348466746,
+      "grad_norm": 0.47979077965055483,
+      "learning_rate": 3.4844356226711313e-07,
+      "loss": 0.5403,
+      "step": 18770
+    },
+    {
+      "epoch": 4.98314084694013,
+      "grad_norm": 0.4891051746517009,
+      "learning_rate": 3.482657921798235e-07,
+      "loss": 0.5178,
+      "step": 18771
+    },
+    {
+      "epoch": 4.983406345413514,
+      "grad_norm": 0.4716265052757659,
+      "learning_rate": 3.4808806405661924e-07,
+      "loss": 0.4943,
+      "step": 18772
+    },
+    {
+      "epoch": 4.983671843886897,
+      "grad_norm": 0.492127212378581,
+      "learning_rate": 3.479103779009668e-07,
+      "loss": 0.5242,
+      "step": 18773
+    },
+    {
+      "epoch": 4.9839373423602815,
+      "grad_norm": 0.4723678385591992,
+      "learning_rate": 3.477327337163314e-07,
+      "loss": 0.5233,
+      "step": 18774
+    },
+    {
+      "epoch": 4.984202840833666,
+      "grad_norm": 0.48689285178041564,
+      "learning_rate": 3.475551315061765e-07,
+      "loss": 0.5419,
+      "step": 18775
+    },
+    {
+      "epoch": 4.984468339307049,
+      "grad_norm": 0.4942897862841746,
+      "learning_rate": 3.473775712739674e-07,
+      "loss": 0.5717,
+      "step": 18776
+    },
+    {
+      "epoch": 4.984733837780433,
+      "grad_norm": 0.47805759576703044,
+      "learning_rate": 3.4720005302316563e-07,
+      "loss": 0.529,
+      "step": 18777
+    },
+    {
+      "epoch": 4.984999336253816,
+      "grad_norm": 0.4976608598096262,
+      "learning_rate": 3.470225767572344e-07,
+      "loss": 0.5601,
+      "step": 18778
+    },
+    {
+      "epoch": 4.9852648347272,
+      "grad_norm": 0.470813742974377,
+      "learning_rate": 3.46845142479634e-07,
+      "loss": 0.5129,
+      "step": 18779
+    },
+    {
+      "epoch": 4.985530333200584,
+      "grad_norm": 0.4828222664911521,
+      "learning_rate": 3.466677501938254e-07,
+      "loss": 0.5034,
+      "step": 18780
+    },
+    {
+      "epoch": 4.985795831673968,
+      "grad_norm": 0.4783921888168491,
+      "learning_rate": 3.4649039990326735e-07,
+      "loss": 0.5326,
+      "step": 18781
+    },
+    {
+      "epoch": 4.9860613301473515,
+      "grad_norm": 0.468635628011706,
+      "learning_rate": 3.463130916114199e-07,
+      "loss": 0.5364,
+      "step": 18782
+    },
+    {
+      "epoch": 4.986326828620736,
+      "grad_norm": 0.47604844788297507,
+      "learning_rate": 3.461358253217395e-07,
+      "loss": 0.512,
+      "step": 18783
+    },
+    {
+      "epoch": 4.986592327094119,
+      "grad_norm": 0.47740841774481096,
+      "learning_rate": 3.459586010376847e-07,
+      "loss": 0.5148,
+      "step": 18784
+    },
+    {
+      "epoch": 4.986857825567503,
+      "grad_norm": 0.48359404501110004,
+      "learning_rate": 3.457814187627112e-07,
+      "loss": 0.4995,
+      "step": 18785
+    },
+    {
+      "epoch": 4.987123324040887,
+      "grad_norm": 0.48335223197119337,
+      "learning_rate": 3.4560427850027415e-07,
+      "loss": 0.5391,
+      "step": 18786
+    },
+    {
+      "epoch": 4.98738882251427,
+      "grad_norm": 0.470493980271917,
+      "learning_rate": 3.454271802538289e-07,
+      "loss": 0.5346,
+      "step": 18787
+    },
+    {
+      "epoch": 4.987654320987654,
+      "grad_norm": 0.48460474033786055,
+      "learning_rate": 3.4525012402682826e-07,
+      "loss": 0.5463,
+      "step": 18788
+    },
+    {
+      "epoch": 4.987919819461038,
+      "grad_norm": 0.4784347771818127,
+      "learning_rate": 3.4507310982272666e-07,
+      "loss": 0.5172,
+      "step": 18789
+    },
+    {
+      "epoch": 4.9881853179344215,
+      "grad_norm": 0.49480991019427667,
+      "learning_rate": 3.4489613764497497e-07,
+      "loss": 0.5693,
+      "step": 18790
+    },
+    {
+      "epoch": 4.988450816407806,
+      "grad_norm": 0.4810477804170264,
+      "learning_rate": 3.447192074970257e-07,
+      "loss": 0.5571,
+      "step": 18791
+    },
+    {
+      "epoch": 4.98871631488119,
+      "grad_norm": 0.4822363258022917,
+      "learning_rate": 3.445423193823291e-07,
+      "loss": 0.5498,
+      "step": 18792
+    },
+    {
+      "epoch": 4.988981813354573,
+      "grad_norm": 0.49220510717359184,
+      "learning_rate": 3.443654733043339e-07,
+      "loss": 0.5543,
+      "step": 18793
+    },
+    {
+      "epoch": 4.989247311827957,
+      "grad_norm": 0.48231534486548044,
+      "learning_rate": 3.441886692664906e-07,
+      "loss": 0.5184,
+      "step": 18794
+    },
+    {
+      "epoch": 4.989512810301341,
+      "grad_norm": 0.47782981980740347,
+      "learning_rate": 3.4401190727224613e-07,
+      "loss": 0.5309,
+      "step": 18795
+    },
+    {
+      "epoch": 4.989778308774724,
+      "grad_norm": 0.48630838268734555,
+      "learning_rate": 3.438351873250492e-07,
+      "loss": 0.4796,
+      "step": 18796
+    },
+    {
+      "epoch": 4.990043807248108,
+      "grad_norm": 0.4934706032721086,
+      "learning_rate": 3.436585094283443e-07,
+      "loss": 0.5266,
+      "step": 18797
+    },
+    {
+      "epoch": 4.9903093057214924,
+      "grad_norm": 0.4869275946686896,
+      "learning_rate": 3.4348187358557857e-07,
+      "loss": 0.542,
+      "step": 18798
+    },
+    {
+      "epoch": 4.990574804194876,
+      "grad_norm": 0.4805067624099405,
+      "learning_rate": 3.4330527980019594e-07,
+      "loss": 0.5601,
+      "step": 18799
+    },
+    {
+      "epoch": 4.99084030266826,
+      "grad_norm": 0.48307008859608574,
+      "learning_rate": 3.4312872807564123e-07,
+      "loss": 0.5455,
+      "step": 18800
+    },
+    {
+      "epoch": 4.991105801141644,
+      "grad_norm": 0.4823304534151557,
+      "learning_rate": 3.4295221841535694e-07,
+      "loss": 0.5639,
+      "step": 18801
+    },
+    {
+      "epoch": 4.991371299615027,
+      "grad_norm": 0.48811325987571563,
+      "learning_rate": 3.427757508227861e-07,
+      "loss": 0.5719,
+      "step": 18802
+    },
+    {
+      "epoch": 4.991636798088411,
+      "grad_norm": 0.4833029648962473,
+      "learning_rate": 3.4259932530136987e-07,
+      "loss": 0.5711,
+      "step": 18803
+    },
+    {
+      "epoch": 4.991902296561795,
+      "grad_norm": 0.4933835588482694,
+      "learning_rate": 3.4242294185454856e-07,
+      "loss": 0.5501,
+      "step": 18804
+    },
+    {
+      "epoch": 4.992167795035178,
+      "grad_norm": 0.463343344995082,
+      "learning_rate": 3.422466004857633e-07,
+      "loss": 0.4921,
+      "step": 18805
+    },
+    {
+      "epoch": 4.9924332935085625,
+      "grad_norm": 0.48564498302134157,
+      "learning_rate": 3.4207030119845164e-07,
+      "loss": 0.5308,
+      "step": 18806
+    },
+    {
+      "epoch": 4.992698791981946,
+      "grad_norm": 0.4970449323807783,
+      "learning_rate": 3.4189404399605304e-07,
+      "loss": 0.568,
+      "step": 18807
+    },
+    {
+      "epoch": 4.99296429045533,
+      "grad_norm": 0.48632847286141584,
+      "learning_rate": 3.4171782888200474e-07,
+      "loss": 0.5541,
+      "step": 18808
+    },
+    {
+      "epoch": 4.993229788928714,
+      "grad_norm": 0.47516016234793,
+      "learning_rate": 3.4154165585974315e-07,
+      "loss": 0.523,
+      "step": 18809
+    },
+    {
+      "epoch": 4.993495287402098,
+      "grad_norm": 0.48970214863871936,
+      "learning_rate": 3.4136552493270356e-07,
+      "loss": 0.5328,
+      "step": 18810
+    },
+    {
+      "epoch": 4.993760785875481,
+      "grad_norm": 0.4909447820809886,
+      "learning_rate": 3.4118943610432166e-07,
+      "loss": 0.5708,
+      "step": 18811
+    },
+    {
+      "epoch": 4.994026284348865,
+      "grad_norm": 0.4745921013786922,
+      "learning_rate": 3.410133893780315e-07,
+      "loss": 0.5025,
+      "step": 18812
+    },
+    {
+      "epoch": 4.994291782822248,
+      "grad_norm": 0.47901306157195167,
+      "learning_rate": 3.408373847572663e-07,
+      "loss": 0.5119,
+      "step": 18813
+    },
+    {
+      "epoch": 4.9945572812956325,
+      "grad_norm": 0.48667431313675213,
+      "learning_rate": 3.4066142224545935e-07,
+      "loss": 0.5368,
+      "step": 18814
+    },
+    {
+      "epoch": 4.994822779769017,
+      "grad_norm": 0.4831763553290102,
+      "learning_rate": 3.4048550184604095e-07,
+      "loss": 0.5381,
+      "step": 18815
+    },
+    {
+      "epoch": 4.9950882782424,
+      "grad_norm": 0.4883589707403283,
+      "learning_rate": 3.4030962356244286e-07,
+      "loss": 0.5636,
+      "step": 18816
+    },
+    {
+      "epoch": 4.995353776715784,
+      "grad_norm": 0.47951061930060734,
+      "learning_rate": 3.401337873980945e-07,
+      "loss": 0.5282,
+      "step": 18817
+    },
+    {
+      "epoch": 4.995619275189168,
+      "grad_norm": 0.5011478772790252,
+      "learning_rate": 3.3995799335642623e-07,
+      "loss": 0.5594,
+      "step": 18818
+    },
+    {
+      "epoch": 4.995884773662551,
+      "grad_norm": 0.4898536737671288,
+      "learning_rate": 3.3978224144086553e-07,
+      "loss": 0.5723,
+      "step": 18819
+    },
+    {
+      "epoch": 4.996150272135935,
+      "grad_norm": 0.47160743428203444,
+      "learning_rate": 3.3960653165484e-07,
+      "loss": 0.5444,
+      "step": 18820
+    },
+    {
+      "epoch": 4.996415770609319,
+      "grad_norm": 0.4799995902215757,
+      "learning_rate": 3.3943086400177684e-07,
+      "loss": 0.5698,
+      "step": 18821
+    },
+    {
+      "epoch": 4.9966812690827025,
+      "grad_norm": 0.4886629222639214,
+      "learning_rate": 3.3925523848510166e-07,
+      "loss": 0.5366,
+      "step": 18822
+    },
+    {
+      "epoch": 4.996946767556087,
+      "grad_norm": 0.4796562879291048,
+      "learning_rate": 3.3907965510824007e-07,
+      "loss": 0.534,
+      "step": 18823
+    },
+    {
+      "epoch": 4.997212266029471,
+      "grad_norm": 0.4979055413256308,
+      "learning_rate": 3.3890411387461545e-07,
+      "loss": 0.5084,
+      "step": 18824
+    },
+    {
+      "epoch": 4.997477764502854,
+      "grad_norm": 0.4765304279572943,
+      "learning_rate": 3.387286147876531e-07,
+      "loss": 0.5226,
+      "step": 18825
+    },
+    {
+      "epoch": 4.997743262976238,
+      "grad_norm": 0.4794582167661844,
+      "learning_rate": 3.385531578507734e-07,
+      "loss": 0.5682,
+      "step": 18826
+    },
+    {
+      "epoch": 4.998008761449622,
+      "grad_norm": 0.47680896308803433,
+      "learning_rate": 3.3837774306739985e-07,
+      "loss": 0.521,
+      "step": 18827
+    },
+    {
+      "epoch": 4.998274259923005,
+      "grad_norm": 0.4890485813167451,
+      "learning_rate": 3.3820237044095203e-07,
+      "loss": 0.5207,
+      "step": 18828
+    },
+    {
+      "epoch": 4.998539758396389,
+      "grad_norm": 0.4664643056741581,
+      "learning_rate": 3.380270399748517e-07,
+      "loss": 0.5292,
+      "step": 18829
+    },
+    {
+      "epoch": 4.9988052568697725,
+      "grad_norm": 0.6349181820465865,
+      "learning_rate": 3.378517516725172e-07,
+      "loss": 0.5385,
+      "step": 18830
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 22596,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 3766,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2429043891240960.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}