diff --git "a/checkpoint-1005/trainer_state.json" "b/checkpoint-1005/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1005/trainer_state.json"
@@ -0,0 +1,7068 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.978476821192053,
+  "eval_steps": 500,
+  "global_step": 1005,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004966887417218543,
+      "grad_norm": 35.1089973449707,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 5.1044,
+      "step": 1
+    },
+    {
+      "epoch": 0.009933774834437087,
+      "grad_norm": 35.22032165527344,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 5.0735,
+      "step": 2
+    },
+    {
+      "epoch": 0.014900662251655629,
+      "grad_norm": 35.83143997192383,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 5.1028,
+      "step": 3
+    },
+    {
+      "epoch": 0.019867549668874173,
+      "grad_norm": 35.02216339111328,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 5.0132,
+      "step": 4
+    },
+    {
+      "epoch": 0.024834437086092714,
+      "grad_norm": 35.219600677490234,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 4.9707,
+      "step": 5
+    },
+    {
+      "epoch": 0.029801324503311258,
+      "grad_norm": 36.146759033203125,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 5.0997,
+      "step": 6
+    },
+    {
+      "epoch": 0.0347682119205298,
+      "grad_norm": 35.079566955566406,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 5.0102,
+      "step": 7
+    },
+    {
+      "epoch": 0.039735099337748346,
+      "grad_norm": 33.08433532714844,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 4.7433,
+      "step": 8
+    },
+    {
+      "epoch": 0.04470198675496689,
+      "grad_norm": 33.799949645996094,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 4.8223,
+      "step": 9
+    },
+    {
+      "epoch": 0.04966887417218543,
+      "grad_norm": 35.39978790283203,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 5.1291,
+      "step": 10
+    },
+    {
+      "epoch": 0.054635761589403975,
+      "grad_norm": 33.871028900146484,
+      "learning_rate": 5.5e-07,
+      "loss": 4.8411,
+      "step": 11
+    },
+    {
+      "epoch": 0.059602649006622516,
+      "grad_norm": 33.94557571411133,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 4.8888,
+      "step": 12
+    },
+    {
+      "epoch": 0.06456953642384106,
+      "grad_norm": 35.7645263671875,
+      "learning_rate": 6.5e-07,
+      "loss": 4.9706,
+      "step": 13
+    },
+    {
+      "epoch": 0.0695364238410596,
+      "grad_norm": 33.34524917602539,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 4.8249,
+      "step": 14
+    },
+    {
+      "epoch": 0.07450331125827815,
+      "grad_norm": 31.388317108154297,
+      "learning_rate": 7.5e-07,
+      "loss": 4.4837,
+      "step": 15
+    },
+    {
+      "epoch": 0.07947019867549669,
+      "grad_norm": 30.788236618041992,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 4.4459,
+      "step": 16
+    },
+    {
+      "epoch": 0.08443708609271523,
+      "grad_norm": 30.05523109436035,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 4.4497,
+      "step": 17
+    },
+    {
+      "epoch": 0.08940397350993377,
+      "grad_norm": 27.767133712768555,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 4.3426,
+      "step": 18
+    },
+    {
+      "epoch": 0.09437086092715231,
+      "grad_norm": 26.096942901611328,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 4.1846,
+      "step": 19
+    },
+    {
+      "epoch": 0.09933774834437085,
+      "grad_norm": 24.659503936767578,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 4.1404,
+      "step": 20
+    },
+    {
+      "epoch": 0.10430463576158941,
+      "grad_norm": 22.810558319091797,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 3.9254,
+      "step": 21
+    },
+    {
+      "epoch": 0.10927152317880795,
+      "grad_norm": 21.360898971557617,
+      "learning_rate": 1.1e-06,
+      "loss": 3.7266,
+      "step": 22
+    },
+    {
+      "epoch": 0.11423841059602649,
+      "grad_norm": 22.106414794921875,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 3.6787,
+      "step": 23
+    },
+    {
+      "epoch": 0.11920529801324503,
+      "grad_norm": 19.86927032470703,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.3947,
+      "step": 24
+    },
+    {
+      "epoch": 0.12417218543046357,
+      "grad_norm": 20.86886215209961,
+      "learning_rate": 1.25e-06,
+      "loss": 3.3619,
+      "step": 25
+    },
+    {
+      "epoch": 0.1291390728476821,
+      "grad_norm": 20.39168357849121,
+      "learning_rate": 1.3e-06,
+      "loss": 3.1309,
+      "step": 26
+    },
+    {
+      "epoch": 0.13410596026490065,
+      "grad_norm": 20.989742279052734,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 3.0103,
+      "step": 27
+    },
+    {
+      "epoch": 0.1390728476821192,
+      "grad_norm": 19.428770065307617,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.8232,
+      "step": 28
+    },
+    {
+      "epoch": 0.14403973509933773,
+      "grad_norm": 16.829797744750977,
+      "learning_rate": 1.45e-06,
+      "loss": 2.6573,
+      "step": 29
+    },
+    {
+      "epoch": 0.1490066225165563,
+      "grad_norm": 15.379148483276367,
+      "learning_rate": 1.5e-06,
+      "loss": 2.5938,
+      "step": 30
+    },
+    {
+      "epoch": 0.15397350993377484,
+      "grad_norm": 14.703971862792969,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 2.4588,
+      "step": 31
+    },
+    {
+      "epoch": 0.15894039735099338,
+      "grad_norm": 14.567538261413574,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.211,
+      "step": 32
+    },
+    {
+      "epoch": 0.16390728476821192,
+      "grad_norm": 14.908529281616211,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 2.0326,
+      "step": 33
+    },
+    {
+      "epoch": 0.16887417218543047,
+      "grad_norm": 15.71338939666748,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.9728,
+      "step": 34
+    },
+    {
+      "epoch": 0.173841059602649,
+      "grad_norm": 15.91982364654541,
+      "learning_rate": 1.75e-06,
+      "loss": 1.8389,
+      "step": 35
+    },
+    {
+      "epoch": 0.17880794701986755,
+      "grad_norm": 15.001914024353027,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.6801,
+      "step": 36
+    },
+    {
+      "epoch": 0.1837748344370861,
+      "grad_norm": 13.963605880737305,
+      "learning_rate": 1.85e-06,
+      "loss": 1.4723,
+      "step": 37
+    },
+    {
+      "epoch": 0.18874172185430463,
+      "grad_norm": 13.326650619506836,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.3989,
+      "step": 38
+    },
+    {
+      "epoch": 0.19370860927152317,
+      "grad_norm": 13.237666130065918,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.2334,
+      "step": 39
+    },
+    {
+      "epoch": 0.1986754966887417,
+      "grad_norm": 12.756895065307617,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.0125,
+      "step": 40
+    },
+    {
+      "epoch": 0.20364238410596028,
+      "grad_norm": 12.65739917755127,
+      "learning_rate": 2.05e-06,
+      "loss": 0.9345,
+      "step": 41
+    },
+    {
+      "epoch": 0.20860927152317882,
+      "grad_norm": 12.268824577331543,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.7775,
+      "step": 42
+    },
+    {
+      "epoch": 0.21357615894039736,
+      "grad_norm": 11.216493606567383,
+      "learning_rate": 2.15e-06,
+      "loss": 0.6453,
+      "step": 43
+    },
+    {
+      "epoch": 0.2185430463576159,
+      "grad_norm": 10.217277526855469,
+      "learning_rate": 2.2e-06,
+      "loss": 0.5069,
+      "step": 44
+    },
+    {
+      "epoch": 0.22350993377483444,
+      "grad_norm": 8.963619232177734,
+      "learning_rate": 2.25e-06,
+      "loss": 0.4146,
+      "step": 45
+    },
+    {
+      "epoch": 0.22847682119205298,
+      "grad_norm": 7.5903706550598145,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.3154,
+      "step": 46
+    },
+    {
+      "epoch": 0.23344370860927152,
+      "grad_norm": 5.7549147605896,
+      "learning_rate": 2.35e-06,
+      "loss": 0.2397,
+      "step": 47
+    },
+    {
+      "epoch": 0.23841059602649006,
+      "grad_norm": 3.816410779953003,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.1793,
+      "step": 48
+    },
+    {
+      "epoch": 0.2433774834437086,
+      "grad_norm": 2.867004156112671,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.1624,
+      "step": 49
+    },
+    {
+      "epoch": 0.24834437086092714,
+      "grad_norm": 2.462583303451538,
+      "learning_rate": 2.5e-06,
+      "loss": 0.135,
+      "step": 50
+    },
+    {
+      "epoch": 0.2533112582781457,
+      "grad_norm": 2.168409824371338,
+      "learning_rate": 2.55e-06,
+      "loss": 0.1176,
+      "step": 51
+    },
+    {
+      "epoch": 0.2582781456953642,
+      "grad_norm": 1.5076923370361328,
+      "learning_rate": 2.6e-06,
+      "loss": 0.1294,
+      "step": 52
+    },
+    {
+      "epoch": 0.2632450331125828,
+      "grad_norm": 1.422659158706665,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.0929,
+      "step": 53
+    },
+    {
+      "epoch": 0.2682119205298013,
+      "grad_norm": 1.2882518768310547,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.0872,
+      "step": 54
+    },
+    {
+      "epoch": 0.2731788079470199,
+      "grad_norm": 0.9768906235694885,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.1038,
+      "step": 55
+    },
+    {
+      "epoch": 0.2781456953642384,
+      "grad_norm": 0.6768905520439148,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.0839,
+      "step": 56
+    },
+    {
+      "epoch": 0.28311258278145696,
+      "grad_norm": 0.8317804932594299,
+      "learning_rate": 2.85e-06,
+      "loss": 0.0793,
+      "step": 57
+    },
+    {
+      "epoch": 0.28807947019867547,
+      "grad_norm": 1.0260225534439087,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0907,
+      "step": 58
+    },
+    {
+      "epoch": 0.29304635761589404,
+      "grad_norm": 0.69094318151474,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0748,
+      "step": 59
+    },
+    {
+      "epoch": 0.2980132450331126,
+      "grad_norm": 0.6794005036354065,
+      "learning_rate": 3e-06,
+      "loss": 0.0784,
+      "step": 60
+    },
+    {
+      "epoch": 0.3029801324503311,
+      "grad_norm": 0.6374951004981995,
+      "learning_rate": 3.05e-06,
+      "loss": 0.0739,
+      "step": 61
+    },
+    {
+      "epoch": 0.3079470198675497,
+      "grad_norm": 0.7064375281333923,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.0729,
+      "step": 62
+    },
+    {
+      "epoch": 0.3129139072847682,
+      "grad_norm": 0.4373936951160431,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0682,
+      "step": 63
+    },
+    {
+      "epoch": 0.31788079470198677,
+      "grad_norm": 0.5880410075187683,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0666,
+      "step": 64
+    },
+    {
+      "epoch": 0.3228476821192053,
+      "grad_norm": 0.4951966106891632,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.0709,
+      "step": 65
+    },
+    {
+      "epoch": 0.32781456953642385,
+      "grad_norm": 0.46179434657096863,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.0669,
+      "step": 66
+    },
+    {
+      "epoch": 0.33278145695364236,
+      "grad_norm": 0.543757438659668,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.0716,
+      "step": 67
+    },
+    {
+      "epoch": 0.33774834437086093,
+      "grad_norm": 0.5536766052246094,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0751,
+      "step": 68
+    },
+    {
+      "epoch": 0.34271523178807944,
+      "grad_norm": 0.4959389865398407,
+      "learning_rate": 3.45e-06,
+      "loss": 0.0725,
+      "step": 69
+    },
+    {
+      "epoch": 0.347682119205298,
+      "grad_norm": 0.40888911485671997,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0696,
+      "step": 70
+    },
+    {
+      "epoch": 0.3526490066225166,
+      "grad_norm": 0.41407281160354614,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.0577,
+      "step": 71
+    },
+    {
+      "epoch": 0.3576158940397351,
+      "grad_norm": 0.45143064856529236,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.0666,
+      "step": 72
+    },
+    {
+      "epoch": 0.36258278145695366,
+      "grad_norm": 0.4229239523410797,
+      "learning_rate": 3.65e-06,
+      "loss": 0.0657,
+      "step": 73
+    },
+    {
+      "epoch": 0.3675496688741722,
+      "grad_norm": 0.32228904962539673,
+      "learning_rate": 3.7e-06,
+      "loss": 0.0604,
+      "step": 74
+    },
+    {
+      "epoch": 0.37251655629139074,
+      "grad_norm": 0.48797911405563354,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.0658,
+      "step": 75
+    },
+    {
+      "epoch": 0.37748344370860926,
+      "grad_norm": 0.5396533012390137,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.0652,
+      "step": 76
+    },
+    {
+      "epoch": 0.3824503311258278,
+      "grad_norm": 0.3672020733356476,
+      "learning_rate": 3.85e-06,
+      "loss": 0.0559,
+      "step": 77
+    },
+    {
+      "epoch": 0.38741721854304634,
+      "grad_norm": 0.4659889340400696,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0596,
+      "step": 78
+    },
+    {
+      "epoch": 0.3923841059602649,
+      "grad_norm": 0.40139755606651306,
+      "learning_rate": 3.95e-06,
+      "loss": 0.0594,
+      "step": 79
+    },
+    {
+      "epoch": 0.3973509933774834,
+      "grad_norm": 0.32445284724235535,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0574,
+      "step": 80
+    },
+    {
+      "epoch": 0.402317880794702,
+      "grad_norm": 0.5123882293701172,
+      "learning_rate": 4.05e-06,
+      "loss": 0.0694,
+      "step": 81
+    },
+    {
+      "epoch": 0.40728476821192056,
+      "grad_norm": 0.3744450509548187,
+      "learning_rate": 4.1e-06,
+      "loss": 0.0663,
+      "step": 82
+    },
+    {
+      "epoch": 0.41225165562913907,
+      "grad_norm": 0.4288279414176941,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0575,
+      "step": 83
+    },
+    {
+      "epoch": 0.41721854304635764,
+      "grad_norm": 0.3742220997810364,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0643,
+      "step": 84
+    },
+    {
+      "epoch": 0.42218543046357615,
+      "grad_norm": 0.3413388133049011,
+      "learning_rate": 4.25e-06,
+      "loss": 0.0599,
+      "step": 85
+    },
+    {
+      "epoch": 0.4271523178807947,
+      "grad_norm": 0.42702731490135193,
+      "learning_rate": 4.3e-06,
+      "loss": 0.0607,
+      "step": 86
+    },
+    {
+      "epoch": 0.43211920529801323,
+      "grad_norm": 0.28603485226631165,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.0539,
+      "step": 87
+    },
+    {
+      "epoch": 0.4370860927152318,
+      "grad_norm": 0.39420223236083984,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0536,
+      "step": 88
+    },
+    {
+      "epoch": 0.4420529801324503,
+      "grad_norm": 0.3824305534362793,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.0545,
+      "step": 89
+    },
+    {
+      "epoch": 0.4470198675496689,
+      "grad_norm": 0.7128148674964905,
+      "learning_rate": 4.5e-06,
+      "loss": 0.066,
+      "step": 90
+    },
+    {
+      "epoch": 0.4519867549668874,
+      "grad_norm": 0.43943652510643005,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.0551,
+      "step": 91
+    },
+    {
+      "epoch": 0.45695364238410596,
+      "grad_norm": 0.4158875048160553,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.0597,
+      "step": 92
+    },
+    {
+      "epoch": 0.46192052980132453,
+      "grad_norm": 0.4504539668560028,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0651,
+      "step": 93
+    },
+    {
+      "epoch": 0.46688741721854304,
+      "grad_norm": 0.2968044579029083,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0601,
+      "step": 94
+    },
+    {
+      "epoch": 0.4718543046357616,
+      "grad_norm": 0.3136419951915741,
+      "learning_rate": 4.75e-06,
+      "loss": 0.0537,
+      "step": 95
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 0.3263697922229767,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0571,
+      "step": 96
+    },
+    {
+      "epoch": 0.4817880794701987,
+      "grad_norm": 0.4513280391693115,
+      "learning_rate": 4.85e-06,
+      "loss": 0.0578,
+      "step": 97
+    },
+    {
+      "epoch": 0.4867549668874172,
+      "grad_norm": 0.32309144735336304,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0532,
+      "step": 98
+    },
+    {
+      "epoch": 0.4917218543046358,
+      "grad_norm": 0.3191864788532257,
+      "learning_rate": 4.95e-06,
+      "loss": 0.0563,
+      "step": 99
+    },
+    {
+      "epoch": 0.4966887417218543,
+      "grad_norm": 0.3069349229335785,
+      "learning_rate": 5e-06,
+      "loss": 0.0492,
+      "step": 100
+    },
+    {
+      "epoch": 0.5016556291390728,
+      "grad_norm": 0.592820942401886,
+      "learning_rate": 4.999989914458693e-06,
+      "loss": 0.051,
+      "step": 101
+    },
+    {
+      "epoch": 0.5066225165562914,
+      "grad_norm": 0.6325397491455078,
+      "learning_rate": 4.999959657916147e-06,
+      "loss": 0.0621,
+      "step": 102
+    },
+    {
+      "epoch": 0.5115894039735099,
+      "grad_norm": 0.43557360768318176,
+      "learning_rate": 4.999909230616483e-06,
+      "loss": 0.0579,
+      "step": 103
+    },
+    {
+      "epoch": 0.5165562913907285,
+      "grad_norm": 0.5436730980873108,
+      "learning_rate": 4.999838632966572e-06,
+      "loss": 0.0688,
+      "step": 104
+    },
+    {
+      "epoch": 0.5215231788079471,
+      "grad_norm": 0.4149836301803589,
+      "learning_rate": 4.999747865536025e-06,
+      "loss": 0.0504,
+      "step": 105
+    },
+    {
+      "epoch": 0.5264900662251656,
+      "grad_norm": 0.3056333661079407,
+      "learning_rate": 4.999636929057196e-06,
+      "loss": 0.0582,
+      "step": 106
+    },
+    {
+      "epoch": 0.5314569536423841,
+      "grad_norm": 0.3057827651500702,
+      "learning_rate": 4.999505824425164e-06,
+      "loss": 0.0583,
+      "step": 107
+    },
+    {
+      "epoch": 0.5364238410596026,
+      "grad_norm": 0.29118627309799194,
+      "learning_rate": 4.999354552697742e-06,
+      "loss": 0.0545,
+      "step": 108
+    },
+    {
+      "epoch": 0.5413907284768212,
+      "grad_norm": 0.4180416762828827,
+      "learning_rate": 4.999183115095453e-06,
+      "loss": 0.0635,
+      "step": 109
+    },
+    {
+      "epoch": 0.5463576158940397,
+      "grad_norm": 0.36227947473526,
+      "learning_rate": 4.998991513001532e-06,
+      "loss": 0.0581,
+      "step": 110
+    },
+    {
+      "epoch": 0.5513245033112583,
+      "grad_norm": 0.4391280710697174,
+      "learning_rate": 4.998779747961906e-06,
+      "loss": 0.0569,
+      "step": 111
+    },
+    {
+      "epoch": 0.5562913907284768,
+      "grad_norm": 0.40176400542259216,
+      "learning_rate": 4.998547821685188e-06,
+      "loss": 0.0479,
+      "step": 112
+    },
+    {
+      "epoch": 0.5612582781456954,
+      "grad_norm": 0.28267550468444824,
+      "learning_rate": 4.998295736042659e-06,
+      "loss": 0.0562,
+      "step": 113
+    },
+    {
+      "epoch": 0.5662251655629139,
+      "grad_norm": 0.29833951592445374,
+      "learning_rate": 4.998023493068255e-06,
+      "loss": 0.0551,
+      "step": 114
+    },
+    {
+      "epoch": 0.5711920529801324,
+      "grad_norm": 0.4476202130317688,
+      "learning_rate": 4.997731094958551e-06,
+      "loss": 0.0493,
+      "step": 115
+    },
+    {
+      "epoch": 0.5761589403973509,
+      "grad_norm": 0.3935531973838806,
+      "learning_rate": 4.997418544072742e-06,
+      "loss": 0.0602,
+      "step": 116
+    },
+    {
+      "epoch": 0.5811258278145696,
+      "grad_norm": 0.3693472743034363,
+      "learning_rate": 4.9970858429326215e-06,
+      "loss": 0.0567,
+      "step": 117
+    },
+    {
+      "epoch": 0.5860927152317881,
+      "grad_norm": 0.5814310312271118,
+      "learning_rate": 4.99673299422257e-06,
+      "loss": 0.0696,
+      "step": 118
+    },
+    {
+      "epoch": 0.5910596026490066,
+      "grad_norm": 0.32046371698379517,
+      "learning_rate": 4.996360000789519e-06,
+      "loss": 0.0498,
+      "step": 119
+    },
+    {
+      "epoch": 0.5960264900662252,
+      "grad_norm": 0.45766863226890564,
+      "learning_rate": 4.995966865642946e-06,
+      "loss": 0.0501,
+      "step": 120
+    },
+    {
+      "epoch": 0.6009933774834437,
+      "grad_norm": 0.2858210504055023,
+      "learning_rate": 4.995553591954832e-06,
+      "loss": 0.0516,
+      "step": 121
+    },
+    {
+      "epoch": 0.6059602649006622,
+      "grad_norm": 0.2883959412574768,
+      "learning_rate": 4.9951201830596505e-06,
+      "loss": 0.0565,
+      "step": 122
+    },
+    {
+      "epoch": 0.6109271523178808,
+      "grad_norm": 0.2570095360279083,
+      "learning_rate": 4.994666642454331e-06,
+      "loss": 0.0517,
+      "step": 123
+    },
+    {
+      "epoch": 0.6158940397350994,
+      "grad_norm": 0.34462130069732666,
+      "learning_rate": 4.994192973798236e-06,
+      "loss": 0.0589,
+      "step": 124
+    },
+    {
+      "epoch": 0.6208609271523179,
+      "grad_norm": 0.2416890263557434,
+      "learning_rate": 4.993699180913127e-06,
+      "loss": 0.0544,
+      "step": 125
+    },
+    {
+      "epoch": 0.6258278145695364,
+      "grad_norm": 0.2676381468772888,
+      "learning_rate": 4.993185267783142e-06,
+      "loss": 0.0506,
+      "step": 126
+    },
+    {
+      "epoch": 0.6307947019867549,
+      "grad_norm": 0.2828262150287628,
+      "learning_rate": 4.992651238554753e-06,
+      "loss": 0.0504,
+      "step": 127
+    },
+    {
+      "epoch": 0.6357615894039735,
+      "grad_norm": 0.2980761229991913,
+      "learning_rate": 4.99209709753674e-06,
+      "loss": 0.053,
+      "step": 128
+    },
+    {
+      "epoch": 0.640728476821192,
+      "grad_norm": 0.30427825450897217,
+      "learning_rate": 4.991522849200152e-06,
+      "loss": 0.0517,
+      "step": 129
+    },
+    {
+      "epoch": 0.6456953642384106,
+      "grad_norm": 0.32718971371650696,
+      "learning_rate": 4.990928498178274e-06,
+      "loss": 0.0543,
+      "step": 130
+    },
+    {
+      "epoch": 0.6506622516556292,
+      "grad_norm": 0.26971936225891113,
+      "learning_rate": 4.990314049266586e-06,
+      "loss": 0.0498,
+      "step": 131
+    },
+    {
+      "epoch": 0.6556291390728477,
+      "grad_norm": 0.2683403193950653,
+      "learning_rate": 4.989679507422728e-06,
+      "loss": 0.0456,
+      "step": 132
+    },
+    {
+      "epoch": 0.6605960264900662,
+      "grad_norm": 0.3088124990463257,
+      "learning_rate": 4.989024877766461e-06,
+      "loss": 0.0549,
+      "step": 133
+    },
+    {
+      "epoch": 0.6655629139072847,
+      "grad_norm": 0.2979089021682739,
+      "learning_rate": 4.988350165579618e-06,
+      "loss": 0.0493,
+      "step": 134
+    },
+    {
+      "epoch": 0.6705298013245033,
+      "grad_norm": 0.3119763731956482,
+      "learning_rate": 4.987655376306069e-06,
+      "loss": 0.0494,
+      "step": 135
+    },
+    {
+      "epoch": 0.6754966887417219,
+      "grad_norm": 0.26640820503234863,
+      "learning_rate": 4.986940515551676e-06,
+      "loss": 0.0466,
+      "step": 136
+    },
+    {
+      "epoch": 0.6804635761589404,
+      "grad_norm": 0.34190040826797485,
+      "learning_rate": 4.9862055890842455e-06,
+      "loss": 0.052,
+      "step": 137
+    },
+    {
+      "epoch": 0.6854304635761589,
+      "grad_norm": 0.3251188099384308,
+      "learning_rate": 4.9854506028334805e-06,
+      "loss": 0.0475,
+      "step": 138
+    },
+    {
+      "epoch": 0.6903973509933775,
+      "grad_norm": 0.31759873032569885,
+      "learning_rate": 4.984675562890939e-06,
+      "loss": 0.0473,
+      "step": 139
+    },
+    {
+      "epoch": 0.695364238410596,
+      "grad_norm": 0.382521390914917,
+      "learning_rate": 4.983880475509978e-06,
+      "loss": 0.0569,
+      "step": 140
+    },
+    {
+      "epoch": 0.7003311258278145,
+      "grad_norm": 0.2814515233039856,
+      "learning_rate": 4.983065347105707e-06,
+      "loss": 0.0512,
+      "step": 141
+    },
+    {
+      "epoch": 0.7052980132450332,
+      "grad_norm": 0.3223007321357727,
+      "learning_rate": 4.982230184254934e-06,
+      "loss": 0.0563,
+      "step": 142
+    },
+    {
+      "epoch": 0.7102649006622517,
+      "grad_norm": 0.30451443791389465,
+      "learning_rate": 4.981374993696116e-06,
+      "loss": 0.0412,
+      "step": 143
+    },
+    {
+      "epoch": 0.7152317880794702,
+      "grad_norm": 0.39687198400497437,
+      "learning_rate": 4.9804997823292996e-06,
+      "loss": 0.0507,
+      "step": 144
+    },
+    {
+      "epoch": 0.7201986754966887,
+      "grad_norm": 0.311040997505188,
+      "learning_rate": 4.97960455721607e-06,
+      "loss": 0.0483,
+      "step": 145
+    },
+    {
+      "epoch": 0.7251655629139073,
+      "grad_norm": 0.33765530586242676,
+      "learning_rate": 4.978689325579491e-06,
+      "loss": 0.0431,
+      "step": 146
+    },
+    {
+      "epoch": 0.7301324503311258,
+      "grad_norm": 0.321216344833374,
+      "learning_rate": 4.9777540948040474e-06,
+      "loss": 0.0537,
+      "step": 147
+    },
+    {
+      "epoch": 0.7350993377483444,
+      "grad_norm": 0.38110417127609253,
+      "learning_rate": 4.976798872435586e-06,
+      "loss": 0.0535,
+      "step": 148
+    },
+    {
+      "epoch": 0.7400662251655629,
+      "grad_norm": 0.35567203164100647,
+      "learning_rate": 4.975823666181256e-06,
+      "loss": 0.0481,
+      "step": 149
+    },
+    {
+      "epoch": 0.7450331125827815,
+      "grad_norm": 0.27548110485076904,
+      "learning_rate": 4.974828483909441e-06,
+      "loss": 0.0467,
+      "step": 150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33838632702827454,
+      "learning_rate": 4.9738133336497045e-06,
+      "loss": 0.0551,
+      "step": 151
+    },
+    {
+      "epoch": 0.7549668874172185,
+      "grad_norm": 0.29228779673576355,
+      "learning_rate": 4.972778223592717e-06,
+      "loss": 0.0483,
+      "step": 152
+    },
+    {
+      "epoch": 0.7599337748344371,
+      "grad_norm": 0.28962090611457825,
+      "learning_rate": 4.9717231620901964e-06,
+      "loss": 0.0509,
+      "step": 153
+    },
+    {
+      "epoch": 0.7649006622516556,
+      "grad_norm": 0.2481975555419922,
+      "learning_rate": 4.970648157654836e-06,
+      "loss": 0.0473,
+      "step": 154
+    },
+    {
+      "epoch": 0.7698675496688742,
+      "grad_norm": 0.2947401702404022,
+      "learning_rate": 4.969553218960235e-06,
+      "loss": 0.0449,
+      "step": 155
+    },
+    {
+      "epoch": 0.7748344370860927,
+      "grad_norm": 0.291072815656662,
+      "learning_rate": 4.968438354840834e-06,
+      "loss": 0.0547,
+      "step": 156
+    },
+    {
+      "epoch": 0.7798013245033113,
+      "grad_norm": 0.2935287356376648,
+      "learning_rate": 4.96730357429184e-06,
+      "loss": 0.0517,
+      "step": 157
+    },
+    {
+      "epoch": 0.7847682119205298,
+      "grad_norm": 0.3540295958518982,
+      "learning_rate": 4.966148886469153e-06,
+      "loss": 0.0468,
+      "step": 158
+    },
+    {
+      "epoch": 0.7897350993377483,
+      "grad_norm": 0.2809925973415375,
+      "learning_rate": 4.964974300689295e-06,
+      "loss": 0.0474,
+      "step": 159
+    },
+    {
+      "epoch": 0.7947019867549668,
+      "grad_norm": 0.3492342531681061,
+      "learning_rate": 4.963779826429333e-06,
+      "loss": 0.056,
+      "step": 160
+    },
+    {
+      "epoch": 0.7996688741721855,
+      "grad_norm": 0.39458170533180237,
+      "learning_rate": 4.9625654733268016e-06,
+      "loss": 0.0496,
+      "step": 161
+    },
+    {
+      "epoch": 0.804635761589404,
+      "grad_norm": 0.30106887221336365,
+      "learning_rate": 4.961331251179629e-06,
+      "loss": 0.0401,
+      "step": 162
+    },
+    {
+      "epoch": 0.8096026490066225,
+      "grad_norm": 0.297493577003479,
+      "learning_rate": 4.960077169946052e-06,
+      "loss": 0.0488,
+      "step": 163
+    },
+    {
+      "epoch": 0.8145695364238411,
+      "grad_norm": 0.30854183435440063,
+      "learning_rate": 4.958803239744542e-06,
+      "loss": 0.0452,
+      "step": 164
+    },
+    {
+      "epoch": 0.8195364238410596,
+      "grad_norm": 0.3040302097797394,
+      "learning_rate": 4.9575094708537205e-06,
+      "loss": 0.0497,
+      "step": 165
+    },
+    {
+      "epoch": 0.8245033112582781,
+      "grad_norm": 0.3189365267753601,
+      "learning_rate": 4.956195873712274e-06,
+      "loss": 0.0404,
+      "step": 166
+    },
+    {
+      "epoch": 0.8294701986754967,
+      "grad_norm": 0.27114391326904297,
+      "learning_rate": 4.954862458918873e-06,
+      "loss": 0.0376,
+      "step": 167
+    },
+    {
+      "epoch": 0.8344370860927153,
+      "grad_norm": 0.37876346707344055,
+      "learning_rate": 4.953509237232086e-06,
+      "loss": 0.0536,
+      "step": 168
+    },
+    {
+      "epoch": 0.8394039735099338,
+      "grad_norm": 0.3651692867279053,
+      "learning_rate": 4.952136219570292e-06,
+      "loss": 0.0513,
+      "step": 169
+    },
+    {
+      "epoch": 0.8443708609271523,
+      "grad_norm": 0.2958744764328003,
+      "learning_rate": 4.950743417011591e-06,
+      "loss": 0.0448,
+      "step": 170
+    },
+    {
+      "epoch": 0.8493377483443708,
+      "grad_norm": 0.2732123136520386,
+      "learning_rate": 4.9493308407937176e-06,
+      "loss": 0.0441,
+      "step": 171
+    },
+    {
+      "epoch": 0.8543046357615894,
+      "grad_norm": 0.2853662073612213,
+      "learning_rate": 4.947898502313949e-06,
+      "loss": 0.039,
+      "step": 172
+    },
+    {
+      "epoch": 0.859271523178808,
+      "grad_norm": 0.3872867822647095,
+      "learning_rate": 4.946446413129011e-06,
+      "loss": 0.0561,
+      "step": 173
+    },
+    {
+      "epoch": 0.8642384105960265,
+      "grad_norm": 0.33438095450401306,
+      "learning_rate": 4.944974584954989e-06,
+      "loss": 0.0427,
+      "step": 174
+    },
+    {
+      "epoch": 0.8692052980132451,
+      "grad_norm": 0.3008469045162201,
+      "learning_rate": 4.943483029667231e-06,
+      "loss": 0.0417,
+      "step": 175
+    },
+    {
+      "epoch": 0.8741721854304636,
+      "grad_norm": 0.298586368560791,
+      "learning_rate": 4.941971759300249e-06,
+      "loss": 0.0508,
+      "step": 176
+    },
+    {
+      "epoch": 0.8791390728476821,
+      "grad_norm": 0.2677428126335144,
+      "learning_rate": 4.9404407860476275e-06,
+      "loss": 0.0468,
+      "step": 177
+    },
+    {
+      "epoch": 0.8841059602649006,
+      "grad_norm": 0.2635156214237213,
+      "learning_rate": 4.938890122261923e-06,
+      "loss": 0.0467,
+      "step": 178
+    },
+    {
+      "epoch": 0.8890728476821192,
+      "grad_norm": 0.29530155658721924,
+      "learning_rate": 4.937319780454559e-06,
+      "loss": 0.0408,
+      "step": 179
+    },
+    {
+      "epoch": 0.8940397350993378,
+      "grad_norm": 0.2985950708389282,
+      "learning_rate": 4.935729773295738e-06,
+      "loss": 0.0397,
+      "step": 180
+    },
+    {
+      "epoch": 0.8990066225165563,
+      "grad_norm": 0.3108009696006775,
+      "learning_rate": 4.934120113614322e-06,
+      "loss": 0.0405,
+      "step": 181
+    },
+    {
+      "epoch": 0.9039735099337748,
+      "grad_norm": 0.32336103916168213,
+      "learning_rate": 4.932490814397744e-06,
+      "loss": 0.0435,
+      "step": 182
+    },
+    {
+      "epoch": 0.9089403973509934,
+      "grad_norm": 0.3786524534225464,
+      "learning_rate": 4.930841888791898e-06,
+      "loss": 0.0455,
+      "step": 183
+    },
+    {
+      "epoch": 0.9139072847682119,
+      "grad_norm": 0.32045990228652954,
+      "learning_rate": 4.929173350101025e-06,
+      "loss": 0.0424,
+      "step": 184
+    },
+    {
+      "epoch": 0.9188741721854304,
+      "grad_norm": 0.4801805913448334,
+      "learning_rate": 4.927485211787622e-06,
+      "loss": 0.0558,
+      "step": 185
+    },
+    {
+      "epoch": 0.9238410596026491,
+      "grad_norm": 0.28597143292427063,
+      "learning_rate": 4.925777487472318e-06,
+      "loss": 0.0378,
+      "step": 186
+    },
+    {
+      "epoch": 0.9288079470198676,
+      "grad_norm": 0.2773878872394562,
+      "learning_rate": 4.9240501909337725e-06,
+      "loss": 0.0439,
+      "step": 187
+    },
+    {
+      "epoch": 0.9337748344370861,
+      "grad_norm": 0.2605089545249939,
+      "learning_rate": 4.922303336108562e-06,
+      "loss": 0.0361,
+      "step": 188
+    },
+    {
+      "epoch": 0.9387417218543046,
+      "grad_norm": 0.2805444598197937,
+      "learning_rate": 4.920536937091068e-06,
+      "loss": 0.0415,
+      "step": 189
+    },
+    {
+      "epoch": 0.9437086092715232,
+      "grad_norm": 0.33180198073387146,
+      "learning_rate": 4.918751008133362e-06,
+      "loss": 0.0427,
+      "step": 190
+    },
+    {
+      "epoch": 0.9486754966887417,
+      "grad_norm": 0.28023046255111694,
+      "learning_rate": 4.916945563645093e-06,
+      "loss": 0.0419,
+      "step": 191
+    },
+    {
+      "epoch": 0.9536423841059603,
+      "grad_norm": 0.32969868183135986,
+      "learning_rate": 4.915120618193369e-06,
+      "loss": 0.0486,
+      "step": 192
+    },
+    {
+      "epoch": 0.9586092715231788,
+      "grad_norm": 0.3446713984012604,
+      "learning_rate": 4.913276186502639e-06,
+      "loss": 0.0479,
+      "step": 193
+    },
+    {
+      "epoch": 0.9635761589403974,
+      "grad_norm": 0.2720557749271393,
+      "learning_rate": 4.911412283454579e-06,
+      "loss": 0.0388,
+      "step": 194
+    },
+    {
+      "epoch": 0.9685430463576159,
+      "grad_norm": 0.28475743532180786,
+      "learning_rate": 4.909528924087963e-06,
+      "loss": 0.0402,
+      "step": 195
+    },
+    {
+      "epoch": 0.9735099337748344,
+      "grad_norm": 0.3087317943572998,
+      "learning_rate": 4.907626123598552e-06,
+      "loss": 0.0421,
+      "step": 196
+    },
+    {
+      "epoch": 0.9784768211920529,
+      "grad_norm": 0.3794298470020294,
+      "learning_rate": 4.9057038973389635e-06,
+      "loss": 0.0489,
+      "step": 197
+    },
+    {
+      "epoch": 0.9834437086092715,
+      "grad_norm": 0.31470081210136414,
+      "learning_rate": 4.903762260818552e-06,
+      "loss": 0.0419,
+      "step": 198
+    },
+    {
+      "epoch": 0.9884105960264901,
+      "grad_norm": 0.3073600232601166,
+      "learning_rate": 4.90180122970328e-06,
+      "loss": 0.0373,
+      "step": 199
+    },
+    {
+      "epoch": 0.9933774834437086,
+      "grad_norm": 0.35700473189353943,
+      "learning_rate": 4.899820819815598e-06,
+      "loss": 0.0382,
+      "step": 200
+    },
+    {
+      "epoch": 0.9983443708609272,
+      "grad_norm": 0.38011306524276733,
+      "learning_rate": 4.89782104713431e-06,
+      "loss": 0.0437,
+      "step": 201
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38011306524276733,
+      "learning_rate": 4.895801927794448e-06,
+      "loss": 0.0379,
+      "step": 202
+    },
+    {
+      "epoch": 1.0049668874172186,
+      "grad_norm": 0.5858293175697327,
+      "learning_rate": 4.8937634780871406e-06,
+      "loss": 0.0381,
+      "step": 203
+    },
+    {
+      "epoch": 1.009933774834437,
+      "grad_norm": 0.27857479453086853,
+      "learning_rate": 4.891705714459483e-06,
+      "loss": 0.0357,
+      "step": 204
+    },
+    {
+      "epoch": 1.0149006622516556,
+      "grad_norm": 0.331392377614975,
+      "learning_rate": 4.8896286535144025e-06,
+      "loss": 0.0327,
+      "step": 205
+    },
+    {
+      "epoch": 1.0198675496688743,
+      "grad_norm": 0.31564489006996155,
+      "learning_rate": 4.8875323120105275e-06,
+      "loss": 0.0329,
+      "step": 206
+    },
+    {
+      "epoch": 1.0248344370860927,
+      "grad_norm": 0.3151126801967621,
+      "learning_rate": 4.885416706862049e-06,
+      "loss": 0.0361,
+      "step": 207
+    },
+    {
+      "epoch": 1.0298013245033113,
+      "grad_norm": 0.3161706030368805,
+      "learning_rate": 4.883281855138585e-06,
+      "loss": 0.0363,
+      "step": 208
+    },
+    {
+      "epoch": 1.0347682119205297,
+      "grad_norm": 0.3697150945663452,
+      "learning_rate": 4.8811277740650436e-06,
+      "loss": 0.0402,
+      "step": 209
+    },
+    {
+      "epoch": 1.0397350993377483,
+      "grad_norm": 0.3180171549320221,
+      "learning_rate": 4.878954481021484e-06,
+      "loss": 0.0305,
+      "step": 210
+    },
+    {
+      "epoch": 1.044701986754967,
+      "grad_norm": 0.3020327389240265,
+      "learning_rate": 4.876761993542975e-06,
+      "loss": 0.0322,
+      "step": 211
+    },
+    {
+      "epoch": 1.0496688741721854,
+      "grad_norm": 0.44095057249069214,
+      "learning_rate": 4.874550329319457e-06,
+      "loss": 0.033,
+      "step": 212
+    },
+    {
+      "epoch": 1.054635761589404,
+      "grad_norm": 0.36998042464256287,
+      "learning_rate": 4.872319506195593e-06,
+      "loss": 0.0296,
+      "step": 213
+    },
+    {
+      "epoch": 1.0596026490066226,
+      "grad_norm": 0.3310064971446991,
+      "learning_rate": 4.87006954217063e-06,
+      "loss": 0.0309,
+      "step": 214
+    },
+    {
+      "epoch": 1.064569536423841,
+      "grad_norm": 0.3527117669582367,
+      "learning_rate": 4.867800455398252e-06,
+      "loss": 0.0367,
+      "step": 215
+    },
+    {
+      "epoch": 1.0695364238410596,
+      "grad_norm": 0.36431634426116943,
+      "learning_rate": 4.8655122641864335e-06,
+      "loss": 0.0323,
+      "step": 216
+    },
+    {
+      "epoch": 1.0745033112582782,
+      "grad_norm": 0.3809157609939575,
+      "learning_rate": 4.8632049869972944e-06,
+      "loss": 0.033,
+      "step": 217
+    },
+    {
+      "epoch": 1.0794701986754967,
+      "grad_norm": 0.27598831057548523,
+      "learning_rate": 4.8608786424469434e-06,
+      "loss": 0.0302,
+      "step": 218
+    },
+    {
+      "epoch": 1.0844370860927153,
+      "grad_norm": 0.36997970938682556,
+      "learning_rate": 4.858533249305337e-06,
+      "loss": 0.0317,
+      "step": 219
+    },
+    {
+      "epoch": 1.0894039735099337,
+      "grad_norm": 0.30946972966194153,
+      "learning_rate": 4.856168826496123e-06,
+      "loss": 0.0342,
+      "step": 220
+    },
+    {
+      "epoch": 1.0943708609271523,
+      "grad_norm": 0.3649260699748993,
+      "learning_rate": 4.853785393096487e-06,
+      "loss": 0.0363,
+      "step": 221
+    },
+    {
+      "epoch": 1.099337748344371,
+      "grad_norm": 0.3411741554737091,
+      "learning_rate": 4.851382968337004e-06,
+      "loss": 0.0297,
+      "step": 222
+    },
+    {
+      "epoch": 1.1043046357615893,
+      "grad_norm": 0.34998854994773865,
+      "learning_rate": 4.848961571601476e-06,
+      "loss": 0.0406,
+      "step": 223
+    },
+    {
+      "epoch": 1.109271523178808,
+      "grad_norm": 0.3024022579193115,
+      "learning_rate": 4.84652122242678e-06,
+      "loss": 0.0311,
+      "step": 224
+    },
+    {
+      "epoch": 1.1142384105960266,
+      "grad_norm": 0.302640825510025,
+      "learning_rate": 4.844061940502711e-06,
+      "loss": 0.0354,
+      "step": 225
+    },
+    {
+      "epoch": 1.119205298013245,
+      "grad_norm": 0.3342653512954712,
+      "learning_rate": 4.84158374567182e-06,
+      "loss": 0.037,
+      "step": 226
+    },
+    {
+      "epoch": 1.1241721854304636,
+      "grad_norm": 0.3428351581096649,
+      "learning_rate": 4.839086657929256e-06,
+      "loss": 0.0299,
+      "step": 227
+    },
+    {
+      "epoch": 1.129139072847682,
+      "grad_norm": 0.2952878177165985,
+      "learning_rate": 4.836570697422605e-06,
+      "loss": 0.0331,
+      "step": 228
+    },
+    {
+      "epoch": 1.1341059602649006,
+      "grad_norm": 0.3805794417858124,
+      "learning_rate": 4.834035884451725e-06,
+      "loss": 0.0429,
+      "step": 229
+    },
+    {
+      "epoch": 1.1390728476821192,
+      "grad_norm": 0.33509016036987305,
+      "learning_rate": 4.831482239468585e-06,
+      "loss": 0.0302,
+      "step": 230
+    },
+    {
+      "epoch": 1.1440397350993377,
+      "grad_norm": 0.3166174590587616,
+      "learning_rate": 4.8289097830770995e-06,
+      "loss": 0.0306,
+      "step": 231
+    },
+    {
+      "epoch": 1.1490066225165563,
+      "grad_norm": 0.3474961817264557,
+      "learning_rate": 4.826318536032959e-06,
+      "loss": 0.0355,
+      "step": 232
+    },
+    {
+      "epoch": 1.153973509933775,
+      "grad_norm": 0.3208262026309967,
+      "learning_rate": 4.823708519243468e-06,
+      "loss": 0.0298,
+      "step": 233
+    },
+    {
+      "epoch": 1.1589403973509933,
+      "grad_norm": 0.3397170901298523,
+      "learning_rate": 4.821079753767371e-06,
+      "loss": 0.032,
+      "step": 234
+    },
+    {
+      "epoch": 1.163907284768212,
+      "grad_norm": 0.3516407012939453,
+      "learning_rate": 4.818432260814688e-06,
+      "loss": 0.0323,
+      "step": 235
+    },
+    {
+      "epoch": 1.1688741721854305,
+      "grad_norm": 0.313174307346344,
+      "learning_rate": 4.815766061746538e-06,
+      "loss": 0.03,
+      "step": 236
+    },
+    {
+      "epoch": 1.173841059602649,
+      "grad_norm": 0.3192870020866394,
+      "learning_rate": 4.813081178074969e-06,
+      "loss": 0.0256,
+      "step": 237
+    },
+    {
+      "epoch": 1.1788079470198676,
+      "grad_norm": 0.3418073356151581,
+      "learning_rate": 4.8103776314627845e-06,
+      "loss": 0.0249,
+      "step": 238
+    },
+    {
+      "epoch": 1.1837748344370862,
+      "grad_norm": 0.38076093792915344,
+      "learning_rate": 4.807655443723371e-06,
+      "loss": 0.0355,
+      "step": 239
+    },
+    {
+      "epoch": 1.1887417218543046,
+      "grad_norm": 0.4882449805736542,
+      "learning_rate": 4.804914636820517e-06,
+      "loss": 0.0332,
+      "step": 240
+    },
+    {
+      "epoch": 1.1937086092715232,
+      "grad_norm": 0.40234386920928955,
+      "learning_rate": 4.80215523286824e-06,
+      "loss": 0.0272,
+      "step": 241
+    },
+    {
+      "epoch": 1.1986754966887416,
+      "grad_norm": 0.4203873872756958,
+      "learning_rate": 4.799377254130606e-06,
+      "loss": 0.0357,
+      "step": 242
+    },
+    {
+      "epoch": 1.2036423841059603,
+      "grad_norm": 0.41364455223083496,
+      "learning_rate": 4.79658072302155e-06,
+      "loss": 0.0291,
+      "step": 243
+    },
+    {
+      "epoch": 1.2086092715231789,
+      "grad_norm": 0.3474646508693695,
+      "learning_rate": 4.7937656621046966e-06,
+      "loss": 0.026,
+      "step": 244
+    },
+    {
+      "epoch": 1.2135761589403973,
+      "grad_norm": 0.36813393235206604,
+      "learning_rate": 4.790932094093176e-06,
+      "loss": 0.0262,
+      "step": 245
+    },
+    {
+      "epoch": 1.218543046357616,
+      "grad_norm": 0.3225281238555908,
+      "learning_rate": 4.788080041849443e-06,
+      "loss": 0.0334,
+      "step": 246
+    },
+    {
+      "epoch": 1.2235099337748345,
+      "grad_norm": 0.3596038222312927,
+      "learning_rate": 4.785209528385087e-06,
+      "loss": 0.0312,
+      "step": 247
+    },
+    {
+      "epoch": 1.228476821192053,
+      "grad_norm": 0.3386133313179016,
+      "learning_rate": 4.7823205768606575e-06,
+      "loss": 0.0239,
+      "step": 248
+    },
+    {
+      "epoch": 1.2334437086092715,
+      "grad_norm": 0.34161025285720825,
+      "learning_rate": 4.779413210585464e-06,
+      "loss": 0.0296,
+      "step": 249
+    },
+    {
+      "epoch": 1.23841059602649,
+      "grad_norm": 0.4003826975822449,
+      "learning_rate": 4.776487453017398e-06,
+      "loss": 0.0328,
+      "step": 250
+    },
+    {
+      "epoch": 1.2433774834437086,
+      "grad_norm": 0.3628286123275757,
+      "learning_rate": 4.773543327762737e-06,
+      "loss": 0.0245,
+      "step": 251
+    },
+    {
+      "epoch": 1.2483443708609272,
+      "grad_norm": 0.34538671374320984,
+      "learning_rate": 4.770580858575961e-06,
+      "loss": 0.0267,
+      "step": 252
+    },
+    {
+      "epoch": 1.2533112582781456,
+      "grad_norm": 0.3221345543861389,
+      "learning_rate": 4.767600069359551e-06,
+      "loss": 0.0202,
+      "step": 253
+    },
+    {
+      "epoch": 1.2582781456953642,
+      "grad_norm": 0.3598504066467285,
+      "learning_rate": 4.764600984163809e-06,
+      "loss": 0.0284,
+      "step": 254
+    },
+    {
+      "epoch": 1.2632450331125828,
+      "grad_norm": 0.3942805826663971,
+      "learning_rate": 4.761583627186649e-06,
+      "loss": 0.0242,
+      "step": 255
+    },
+    {
+      "epoch": 1.2682119205298013,
+      "grad_norm": 0.3926999270915985,
+      "learning_rate": 4.7585480227734175e-06,
+      "loss": 0.0272,
+      "step": 256
+    },
+    {
+      "epoch": 1.2731788079470199,
+      "grad_norm": 0.43967002630233765,
+      "learning_rate": 4.755494195416683e-06,
+      "loss": 0.0197,
+      "step": 257
+    },
+    {
+      "epoch": 1.2781456953642385,
+      "grad_norm": 0.37213134765625,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 0.0304,
+      "step": 258
+    },
+    {
+      "epoch": 1.283112582781457,
+      "grad_norm": 0.3521837592124939,
+      "learning_rate": 4.749331970577946e-06,
+      "loss": 0.0192,
+      "step": 259
+    },
+    {
+      "epoch": 1.2880794701986755,
+      "grad_norm": 0.3851149082183838,
+      "learning_rate": 4.746223622815441e-06,
+      "loss": 0.0281,
+      "step": 260
+    },
+    {
+      "epoch": 1.2930463576158941,
+      "grad_norm": 0.3802141547203064,
+      "learning_rate": 4.743097151548031e-06,
+      "loss": 0.0234,
+      "step": 261
+    },
+    {
+      "epoch": 1.2980132450331126,
+      "grad_norm": 0.38271185755729675,
+      "learning_rate": 4.739952582001438e-06,
+      "loss": 0.0299,
+      "step": 262
+    },
+    {
+      "epoch": 1.3029801324503312,
+      "grad_norm": 0.4407588839530945,
+      "learning_rate": 4.736789939547411e-06,
+      "loss": 0.0269,
+      "step": 263
+    },
+    {
+      "epoch": 1.3079470198675498,
+      "grad_norm": 0.36599260568618774,
+      "learning_rate": 4.733609249703522e-06,
+      "loss": 0.0328,
+      "step": 264
+    },
+    {
+      "epoch": 1.3129139072847682,
+      "grad_norm": 0.440237820148468,
+      "learning_rate": 4.730410538132949e-06,
+      "loss": 0.0276,
+      "step": 265
+    },
+    {
+      "epoch": 1.3178807947019868,
+      "grad_norm": 0.4273228943347931,
+      "learning_rate": 4.7271938306442855e-06,
+      "loss": 0.0256,
+      "step": 266
+    },
+    {
+      "epoch": 1.3228476821192052,
+      "grad_norm": 0.36070936918258667,
+      "learning_rate": 4.72395915319132e-06,
+      "loss": 0.0261,
+      "step": 267
+    },
+    {
+      "epoch": 1.3278145695364238,
+      "grad_norm": 0.37429606914520264,
+      "learning_rate": 4.72070653187283e-06,
+      "loss": 0.0232,
+      "step": 268
+    },
+    {
+      "epoch": 1.3327814569536423,
+      "grad_norm": 0.3114779591560364,
+      "learning_rate": 4.717435992932374e-06,
+      "loss": 0.0255,
+      "step": 269
+    },
+    {
+      "epoch": 1.3377483443708609,
+      "grad_norm": 0.435793936252594,
+      "learning_rate": 4.714147562758076e-06,
+      "loss": 0.0246,
+      "step": 270
+    },
+    {
+      "epoch": 1.3427152317880795,
+      "grad_norm": 0.35718899965286255,
+      "learning_rate": 4.710841267882414e-06,
+      "loss": 0.0203,
+      "step": 271
+    },
+    {
+      "epoch": 1.347682119205298,
+      "grad_norm": 0.5373225808143616,
+      "learning_rate": 4.7075171349820084e-06,
+      "loss": 0.0309,
+      "step": 272
+    },
+    {
+      "epoch": 1.3526490066225165,
+      "grad_norm": 0.4737128019332886,
+      "learning_rate": 4.704175190877401e-06,
+      "loss": 0.0246,
+      "step": 273
+    },
+    {
+      "epoch": 1.3576158940397351,
+      "grad_norm": 0.4406234323978424,
+      "learning_rate": 4.700815462532846e-06,
+      "loss": 0.0285,
+      "step": 274
+    },
+    {
+      "epoch": 1.3625827814569536,
+      "grad_norm": 0.4181138873100281,
+      "learning_rate": 4.697437977056085e-06,
+      "loss": 0.0227,
+      "step": 275
+    },
+    {
+      "epoch": 1.3675496688741722,
+      "grad_norm": 0.3848871886730194,
+      "learning_rate": 4.694042761698135e-06,
+      "loss": 0.0274,
+      "step": 276
+    },
+    {
+      "epoch": 1.3725165562913908,
+      "grad_norm": 0.3615216016769409,
+      "learning_rate": 4.690629843853061e-06,
+      "loss": 0.0274,
+      "step": 277
+    },
+    {
+      "epoch": 1.3774834437086092,
+      "grad_norm": 0.34359046816825867,
+      "learning_rate": 4.687199251057765e-06,
+      "loss": 0.0197,
+      "step": 278
+    },
+    {
+      "epoch": 1.3824503311258278,
+      "grad_norm": 0.3309035301208496,
+      "learning_rate": 4.683751010991755e-06,
+      "loss": 0.0187,
+      "step": 279
+    },
+    {
+      "epoch": 1.3874172185430464,
+      "grad_norm": 0.33149197697639465,
+      "learning_rate": 4.6802851514769235e-06,
+      "loss": 0.0297,
+      "step": 280
+    },
+    {
+      "epoch": 1.3923841059602649,
+      "grad_norm": 0.3412141799926758,
+      "learning_rate": 4.676801700477327e-06,
+      "loss": 0.0235,
+      "step": 281
+    },
+    {
+      "epoch": 1.3973509933774835,
+      "grad_norm": 0.3106585144996643,
+      "learning_rate": 4.673300686098957e-06,
+      "loss": 0.0218,
+      "step": 282
+    },
+    {
+      "epoch": 1.402317880794702,
+      "grad_norm": 0.32849153876304626,
+      "learning_rate": 4.669782136589512e-06,
+      "loss": 0.0213,
+      "step": 283
+    },
+    {
+      "epoch": 1.4072847682119205,
+      "grad_norm": 0.43728697299957275,
+      "learning_rate": 4.666246080338175e-06,
+      "loss": 0.0212,
+      "step": 284
+    },
+    {
+      "epoch": 1.4122516556291391,
+      "grad_norm": 0.42004719376564026,
+      "learning_rate": 4.662692545875379e-06,
+      "loss": 0.024,
+      "step": 285
+    },
+    {
+      "epoch": 1.4172185430463577,
+      "grad_norm": 0.5100045800209045,
+      "learning_rate": 4.6591215618725775e-06,
+      "loss": 0.0259,
+      "step": 286
+    },
+    {
+      "epoch": 1.4221854304635762,
+      "grad_norm": 0.4625544250011444,
+      "learning_rate": 4.655533157142016e-06,
+      "loss": 0.0184,
+      "step": 287
+    },
+    {
+      "epoch": 1.4271523178807948,
+      "grad_norm": 0.4128156900405884,
+      "learning_rate": 4.651927360636499e-06,
+      "loss": 0.0154,
+      "step": 288
+    },
+    {
+      "epoch": 1.4321192052980132,
+      "grad_norm": 0.43833258748054504,
+      "learning_rate": 4.648304201449153e-06,
+      "loss": 0.0203,
+      "step": 289
+    },
+    {
+      "epoch": 1.4370860927152318,
+      "grad_norm": 0.4920101463794708,
+      "learning_rate": 4.644663708813196e-06,
+      "loss": 0.0262,
+      "step": 290
+    },
+    {
+      "epoch": 1.4420529801324502,
+      "grad_norm": 0.5934157371520996,
+      "learning_rate": 4.641005912101699e-06,
+      "loss": 0.0293,
+      "step": 291
+    },
+    {
+      "epoch": 1.4470198675496688,
+      "grad_norm": 0.4239577651023865,
+      "learning_rate": 4.63733084082735e-06,
+      "loss": 0.014,
+      "step": 292
+    },
+    {
+      "epoch": 1.4519867549668874,
+      "grad_norm": 0.34310004115104675,
+      "learning_rate": 4.633638524642216e-06,
+      "loss": 0.019,
+      "step": 293
+    },
+    {
+      "epoch": 1.4569536423841059,
+      "grad_norm": 0.37336331605911255,
+      "learning_rate": 4.629928993337501e-06,
+      "loss": 0.023,
+      "step": 294
+    },
+    {
+      "epoch": 1.4619205298013245,
+      "grad_norm": 0.5502914786338806,
+      "learning_rate": 4.626202276843312e-06,
+      "loss": 0.0256,
+      "step": 295
+    },
+    {
+      "epoch": 1.466887417218543,
+      "grad_norm": 0.4025366008281708,
+      "learning_rate": 4.622458405228411e-06,
+      "loss": 0.0212,
+      "step": 296
+    },
+    {
+      "epoch": 1.4718543046357615,
+      "grad_norm": 0.35220882296562195,
+      "learning_rate": 4.618697408699974e-06,
+      "loss": 0.0193,
+      "step": 297
+    },
+    {
+      "epoch": 1.4768211920529801,
+      "grad_norm": 0.42906683683395386,
+      "learning_rate": 4.614919317603351e-06,
+      "loss": 0.0183,
+      "step": 298
+    },
+    {
+      "epoch": 1.4817880794701987,
+      "grad_norm": 0.41879740357398987,
+      "learning_rate": 4.611124162421817e-06,
+      "loss": 0.024,
+      "step": 299
+    },
+    {
+      "epoch": 1.4867549668874172,
+      "grad_norm": 0.3032141923904419,
+      "learning_rate": 4.607311973776329e-06,
+      "loss": 0.0161,
+      "step": 300
+    },
+    {
+      "epoch": 1.4917218543046358,
+      "grad_norm": 0.4052025377750397,
+      "learning_rate": 4.603482782425272e-06,
+      "loss": 0.0185,
+      "step": 301
+    },
+    {
+      "epoch": 1.4966887417218544,
+      "grad_norm": 0.3545592129230499,
+      "learning_rate": 4.599636619264223e-06,
+      "loss": 0.023,
+      "step": 302
+    },
+    {
+      "epoch": 1.5016556291390728,
+      "grad_norm": 0.4538367986679077,
+      "learning_rate": 4.595773515325692e-06,
+      "loss": 0.0142,
+      "step": 303
+    },
+    {
+      "epoch": 1.5066225165562914,
+      "grad_norm": 0.6581038236618042,
+      "learning_rate": 4.591893501778873e-06,
+      "loss": 0.0173,
+      "step": 304
+    },
+    {
+      "epoch": 1.51158940397351,
+      "grad_norm": 0.5108897089958191,
+      "learning_rate": 4.587996609929396e-06,
+      "loss": 0.0176,
+      "step": 305
+    },
+    {
+      "epoch": 1.5165562913907285,
+      "grad_norm": 0.3592700660228729,
+      "learning_rate": 4.5840828712190725e-06,
+      "loss": 0.0142,
+      "step": 306
+    },
+    {
+      "epoch": 1.521523178807947,
+      "grad_norm": 0.6172234416007996,
+      "learning_rate": 4.5801523172256415e-06,
+      "loss": 0.018,
+      "step": 307
+    },
+    {
+      "epoch": 1.5264900662251657,
+      "grad_norm": 0.3826143741607666,
+      "learning_rate": 4.576204979662513e-06,
+      "loss": 0.0158,
+      "step": 308
+    },
+    {
+      "epoch": 1.531456953642384,
+      "grad_norm": 0.5595983862876892,
+      "learning_rate": 4.572240890378518e-06,
+      "loss": 0.0224,
+      "step": 309
+    },
+    {
+      "epoch": 1.5364238410596025,
+      "grad_norm": 0.4325421452522278,
+      "learning_rate": 4.568260081357644e-06,
+      "loss": 0.014,
+      "step": 310
+    },
+    {
+      "epoch": 1.5413907284768213,
+      "grad_norm": 0.4050692617893219,
+      "learning_rate": 4.564262584718782e-06,
+      "loss": 0.0137,
+      "step": 311
+    },
+    {
+      "epoch": 1.5463576158940397,
+      "grad_norm": 0.4626636803150177,
+      "learning_rate": 4.560248432715467e-06,
+      "loss": 0.0158,
+      "step": 312
+    },
+    {
+      "epoch": 1.5513245033112582,
+      "grad_norm": 0.5102334022521973,
+      "learning_rate": 4.556217657735616e-06,
+      "loss": 0.0184,
+      "step": 313
+    },
+    {
+      "epoch": 1.5562913907284768,
+      "grad_norm": 0.3994642496109009,
+      "learning_rate": 4.552170292301265e-06,
+      "loss": 0.0114,
+      "step": 314
+    },
+    {
+      "epoch": 1.5612582781456954,
+      "grad_norm": 0.4770815968513489,
+      "learning_rate": 4.548106369068312e-06,
+      "loss": 0.0148,
+      "step": 315
+    },
+    {
+      "epoch": 1.5662251655629138,
+      "grad_norm": 0.33928975462913513,
+      "learning_rate": 4.54402592082625e-06,
+      "loss": 0.0175,
+      "step": 316
+    },
+    {
+      "epoch": 1.5711920529801324,
+      "grad_norm": 0.4363187849521637,
+      "learning_rate": 4.539928980497903e-06,
+      "loss": 0.0072,
+      "step": 317
+    },
+    {
+      "epoch": 1.576158940397351,
+      "grad_norm": 0.4603089392185211,
+      "learning_rate": 4.535815581139158e-06,
+      "loss": 0.0104,
+      "step": 318
+    },
+    {
+      "epoch": 1.5811258278145695,
+      "grad_norm": 0.41406941413879395,
+      "learning_rate": 4.531685755938704e-06,
+      "loss": 0.009,
+      "step": 319
+    },
+    {
+      "epoch": 1.586092715231788,
+      "grad_norm": 0.5116543173789978,
+      "learning_rate": 4.527539538217758e-06,
+      "loss": 0.0245,
+      "step": 320
+    },
+    {
+      "epoch": 1.5910596026490067,
+      "grad_norm": 0.47456589341163635,
+      "learning_rate": 4.523376961429801e-06,
+      "loss": 0.0133,
+      "step": 321
+    },
+    {
+      "epoch": 1.596026490066225,
+      "grad_norm": 0.39687836170196533,
+      "learning_rate": 4.519198059160303e-06,
+      "loss": 0.0123,
+      "step": 322
+    },
+    {
+      "epoch": 1.6009933774834437,
+      "grad_norm": 0.5981389880180359,
+      "learning_rate": 4.51500286512646e-06,
+      "loss": 0.0157,
+      "step": 323
+    },
+    {
+      "epoch": 1.6059602649006623,
+      "grad_norm": 0.49498969316482544,
+      "learning_rate": 4.510791413176912e-06,
+      "loss": 0.0207,
+      "step": 324
+    },
+    {
+      "epoch": 1.6109271523178808,
+      "grad_norm": 0.4387689530849457,
+      "learning_rate": 4.506563737291479e-06,
+      "loss": 0.0116,
+      "step": 325
+    },
+    {
+      "epoch": 1.6158940397350994,
+      "grad_norm": 0.31551048159599304,
+      "learning_rate": 4.502319871580879e-06,
+      "loss": 0.0126,
+      "step": 326
+    },
+    {
+      "epoch": 1.620860927152318,
+      "grad_norm": 0.4342552423477173,
+      "learning_rate": 4.498059850286459e-06,
+      "loss": 0.0137,
+      "step": 327
+    },
+    {
+      "epoch": 1.6258278145695364,
+      "grad_norm": 0.43873131275177,
+      "learning_rate": 4.493783707779916e-06,
+      "loss": 0.0144,
+      "step": 328
+    },
+    {
+      "epoch": 1.6307947019867548,
+      "grad_norm": 0.3671911358833313,
+      "learning_rate": 4.489491478563019e-06,
+      "loss": 0.0171,
+      "step": 329
+    },
+    {
+      "epoch": 1.6357615894039736,
+      "grad_norm": 0.37722423672676086,
+      "learning_rate": 4.4851831972673335e-06,
+      "loss": 0.0173,
+      "step": 330
+    },
+    {
+      "epoch": 1.640728476821192,
+      "grad_norm": 0.3447704613208771,
+      "learning_rate": 4.480858898653936e-06,
+      "loss": 0.009,
+      "step": 331
+    },
+    {
+      "epoch": 1.6456953642384105,
+      "grad_norm": 0.325804740190506,
+      "learning_rate": 4.476518617613143e-06,
+      "loss": 0.015,
+      "step": 332
+    },
+    {
+      "epoch": 1.6506622516556293,
+      "grad_norm": 0.3346543312072754,
+      "learning_rate": 4.472162389164219e-06,
+      "loss": 0.0164,
+      "step": 333
+    },
+    {
+      "epoch": 1.6556291390728477,
+      "grad_norm": 0.35681554675102234,
+      "learning_rate": 4.467790248455103e-06,
+      "loss": 0.012,
+      "step": 334
+    },
+    {
+      "epoch": 1.660596026490066,
+      "grad_norm": 0.2275272011756897,
+      "learning_rate": 4.46340223076212e-06,
+      "loss": 0.0074,
+      "step": 335
+    },
+    {
+      "epoch": 1.6655629139072847,
+      "grad_norm": 0.44045180082321167,
+      "learning_rate": 4.458998371489695e-06,
+      "loss": 0.0085,
+      "step": 336
+    },
+    {
+      "epoch": 1.6705298013245033,
+      "grad_norm": 0.325270414352417,
+      "learning_rate": 4.454578706170075e-06,
+      "loss": 0.0141,
+      "step": 337
+    },
+    {
+      "epoch": 1.6754966887417218,
+      "grad_norm": 0.5011297464370728,
+      "learning_rate": 4.450143270463031e-06,
+      "loss": 0.012,
+      "step": 338
+    },
+    {
+      "epoch": 1.6804635761589404,
+      "grad_norm": 0.3271932303905487,
+      "learning_rate": 4.445692100155579e-06,
+      "loss": 0.0119,
+      "step": 339
+    },
+    {
+      "epoch": 1.685430463576159,
+      "grad_norm": 0.4148896634578705,
+      "learning_rate": 4.441225231161691e-06,
+      "loss": 0.0141,
+      "step": 340
+    },
+    {
+      "epoch": 1.6903973509933774,
+      "grad_norm": 0.2449064701795578,
+      "learning_rate": 4.436742699521998e-06,
+      "loss": 0.0154,
+      "step": 341
+    },
+    {
+      "epoch": 1.695364238410596,
+      "grad_norm": 0.46324676275253296,
+      "learning_rate": 4.432244541403506e-06,
+      "loss": 0.0102,
+      "step": 342
+    },
+    {
+      "epoch": 1.7003311258278146,
+      "grad_norm": 0.3034791946411133,
+      "learning_rate": 4.4277307930993055e-06,
+      "loss": 0.0132,
+      "step": 343
+    },
+    {
+      "epoch": 1.705298013245033,
+      "grad_norm": 0.502938985824585,
+      "learning_rate": 4.42320149102827e-06,
+      "loss": 0.0122,
+      "step": 344
+    },
+    {
+      "epoch": 1.7102649006622517,
+      "grad_norm": 0.35072168707847595,
+      "learning_rate": 4.41865667173477e-06,
+      "loss": 0.0081,
+      "step": 345
+    },
+    {
+      "epoch": 1.7152317880794703,
+      "grad_norm": 0.4683709442615509,
+      "learning_rate": 4.414096371888378e-06,
+      "loss": 0.0152,
+      "step": 346
+    },
+    {
+      "epoch": 1.7201986754966887,
+      "grad_norm": 0.22903448343276978,
+      "learning_rate": 4.409520628283566e-06,
+      "loss": 0.0074,
+      "step": 347
+    },
+    {
+      "epoch": 1.7251655629139073,
+      "grad_norm": 0.4040380120277405,
+      "learning_rate": 4.404929477839415e-06,
+      "loss": 0.0088,
+      "step": 348
+    },
+    {
+      "epoch": 1.730132450331126,
+      "grad_norm": 0.3377884328365326,
+      "learning_rate": 4.400322957599314e-06,
+      "loss": 0.0156,
+      "step": 349
+    },
+    {
+      "epoch": 1.7350993377483444,
+      "grad_norm": 0.34837478399276733,
+      "learning_rate": 4.395701104730667e-06,
+      "loss": 0.011,
+      "step": 350
+    },
+    {
+      "epoch": 1.7400662251655628,
+      "grad_norm": 0.3430151343345642,
+      "learning_rate": 4.3910639565245805e-06,
+      "loss": 0.012,
+      "step": 351
+    },
+    {
+      "epoch": 1.7450331125827816,
+      "grad_norm": 0.3015480041503906,
+      "learning_rate": 4.386411550395576e-06,
+      "loss": 0.011,
+      "step": 352
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3030356466770172,
+      "learning_rate": 4.3817439238812794e-06,
+      "loss": 0.0072,
+      "step": 353
+    },
+    {
+      "epoch": 1.7549668874172184,
+      "grad_norm": 0.2574770748615265,
+      "learning_rate": 4.377061114642126e-06,
+      "loss": 0.0058,
+      "step": 354
+    },
+    {
+      "epoch": 1.7599337748344372,
+      "grad_norm": 0.32613250613212585,
+      "learning_rate": 4.372363160461045e-06,
+      "loss": 0.01,
+      "step": 355
+    },
+    {
+      "epoch": 1.7649006622516556,
+      "grad_norm": 0.31548556685447693,
+      "learning_rate": 4.367650099243167e-06,
+      "loss": 0.0106,
+      "step": 356
+    },
+    {
+      "epoch": 1.769867549668874,
+      "grad_norm": 0.3480343520641327,
+      "learning_rate": 4.36292196901551e-06,
+      "loss": 0.0142,
+      "step": 357
+    },
+    {
+      "epoch": 1.7748344370860927,
+      "grad_norm": 0.4790445566177368,
+      "learning_rate": 4.358178807926678e-06,
+      "loss": 0.0112,
+      "step": 358
+    },
+    {
+      "epoch": 1.7798013245033113,
+      "grad_norm": 0.38480401039123535,
+      "learning_rate": 4.353420654246546e-06,
+      "loss": 0.0134,
+      "step": 359
+    },
+    {
+      "epoch": 1.7847682119205297,
+      "grad_norm": 0.355960488319397,
+      "learning_rate": 4.34864754636596e-06,
+      "loss": 0.0059,
+      "step": 360
+    },
+    {
+      "epoch": 1.7897350993377483,
+      "grad_norm": 0.3221181035041809,
+      "learning_rate": 4.3438595227964206e-06,
+      "loss": 0.011,
+      "step": 361
+    },
+    {
+      "epoch": 1.794701986754967,
+      "grad_norm": 0.44287869334220886,
+      "learning_rate": 4.339056622169777e-06,
+      "loss": 0.01,
+      "step": 362
+    },
+    {
+      "epoch": 1.7996688741721854,
+      "grad_norm": 0.30105292797088623,
+      "learning_rate": 4.33423888323791e-06,
+      "loss": 0.0078,
+      "step": 363
+    },
+    {
+      "epoch": 1.804635761589404,
+      "grad_norm": 0.43845006823539734,
+      "learning_rate": 4.329406344872423e-06,
+      "loss": 0.0109,
+      "step": 364
+    },
+    {
+      "epoch": 1.8096026490066226,
+      "grad_norm": 0.4705260694026947,
+      "learning_rate": 4.3245590460643296e-06,
+      "loss": 0.0091,
+      "step": 365
+    },
+    {
+      "epoch": 1.814569536423841,
+      "grad_norm": 0.2653980255126953,
+      "learning_rate": 4.319697025923736e-06,
+      "loss": 0.0058,
+      "step": 366
+    },
+    {
+      "epoch": 1.8195364238410596,
+      "grad_norm": 0.3086546063423157,
+      "learning_rate": 4.314820323679524e-06,
+      "loss": 0.0091,
+      "step": 367
+    },
+    {
+      "epoch": 1.8245033112582782,
+      "grad_norm": 0.43733829259872437,
+      "learning_rate": 4.309928978679041e-06,
+      "loss": 0.0089,
+      "step": 368
+    },
+    {
+      "epoch": 1.8294701986754967,
+      "grad_norm": 0.3844042718410492,
+      "learning_rate": 4.305023030387776e-06,
+      "loss": 0.0078,
+      "step": 369
+    },
+    {
+      "epoch": 1.8344370860927153,
+      "grad_norm": 0.3441635072231293,
+      "learning_rate": 4.3001025183890445e-06,
+      "loss": 0.01,
+      "step": 370
+    },
+    {
+      "epoch": 1.839403973509934,
+      "grad_norm": 0.44443604350090027,
+      "learning_rate": 4.295167482383667e-06,
+      "loss": 0.0121,
+      "step": 371
+    },
+    {
+      "epoch": 1.8443708609271523,
+      "grad_norm": 0.5100785493850708,
+      "learning_rate": 4.290217962189654e-06,
+      "loss": 0.0122,
+      "step": 372
+    },
+    {
+      "epoch": 1.8493377483443707,
+      "grad_norm": 0.44218710064888,
+      "learning_rate": 4.285253997741875e-06,
+      "loss": 0.0038,
+      "step": 373
+    },
+    {
+      "epoch": 1.8543046357615895,
+      "grad_norm": 0.4703184962272644,
+      "learning_rate": 4.2802756290917454e-06,
+      "loss": 0.0081,
+      "step": 374
+    },
+    {
+      "epoch": 1.859271523178808,
+      "grad_norm": 0.4065220057964325,
+      "learning_rate": 4.2752828964068996e-06,
+      "loss": 0.0087,
+      "step": 375
+    },
+    {
+      "epoch": 1.8642384105960264,
+      "grad_norm": 0.45641881227493286,
+      "learning_rate": 4.270275839970868e-06,
+      "loss": 0.0067,
+      "step": 376
+    },
+    {
+      "epoch": 1.8692052980132452,
+      "grad_norm": 0.28219515085220337,
+      "learning_rate": 4.2652545001827475e-06,
+      "loss": 0.0076,
+      "step": 377
+    },
+    {
+      "epoch": 1.8741721854304636,
+      "grad_norm": 0.35546135902404785,
+      "learning_rate": 4.260218917556885e-06,
+      "loss": 0.0104,
+      "step": 378
+    },
+    {
+      "epoch": 1.879139072847682,
+      "grad_norm": 0.33450189232826233,
+      "learning_rate": 4.25516913272254e-06,
+      "loss": 0.0048,
+      "step": 379
+    },
+    {
+      "epoch": 1.8841059602649006,
+      "grad_norm": 0.3185473084449768,
+      "learning_rate": 4.250105186423564e-06,
+      "loss": 0.0106,
+      "step": 380
+    },
+    {
+      "epoch": 1.8890728476821192,
+      "grad_norm": 0.31713488698005676,
+      "learning_rate": 4.245027119518068e-06,
+      "loss": 0.0066,
+      "step": 381
+    },
+    {
+      "epoch": 1.8940397350993377,
+      "grad_norm": 0.5118482708930969,
+      "learning_rate": 4.239934972978096e-06,
+      "loss": 0.0056,
+      "step": 382
+    },
+    {
+      "epoch": 1.8990066225165563,
+      "grad_norm": 0.5072982311248779,
+      "learning_rate": 4.234828787889291e-06,
+      "loss": 0.0051,
+      "step": 383
+    },
+    {
+      "epoch": 1.903973509933775,
+      "grad_norm": 0.33334094285964966,
+      "learning_rate": 4.229708605450563e-06,
+      "loss": 0.0104,
+      "step": 384
+    },
+    {
+      "epoch": 1.9089403973509933,
+      "grad_norm": 0.33032891154289246,
+      "learning_rate": 4.224574466973766e-06,
+      "loss": 0.006,
+      "step": 385
+    },
+    {
+      "epoch": 1.913907284768212,
+      "grad_norm": 0.36170345544815063,
+      "learning_rate": 4.219426413883348e-06,
+      "loss": 0.0068,
+      "step": 386
+    },
+    {
+      "epoch": 1.9188741721854305,
+      "grad_norm": 0.2647745609283447,
+      "learning_rate": 4.2142644877160334e-06,
+      "loss": 0.0041,
+      "step": 387
+    },
+    {
+      "epoch": 1.923841059602649,
+      "grad_norm": 0.35909488797187805,
+      "learning_rate": 4.2090887301204765e-06,
+      "loss": 0.0108,
+      "step": 388
+    },
+    {
+      "epoch": 1.9288079470198676,
+      "grad_norm": 2.257672071456909,
+      "learning_rate": 4.2038991828569325e-06,
+      "loss": 0.0067,
+      "step": 389
+    },
+    {
+      "epoch": 1.9337748344370862,
+      "grad_norm": 0.21206845343112946,
+      "learning_rate": 4.198695887796915e-06,
+      "loss": 0.0029,
+      "step": 390
+    },
+    {
+      "epoch": 1.9387417218543046,
+      "grad_norm": 0.3053521513938904,
+      "learning_rate": 4.193478886922861e-06,
+      "loss": 0.0076,
+      "step": 391
+    },
+    {
+      "epoch": 1.9437086092715232,
+      "grad_norm": 0.28233954310417175,
+      "learning_rate": 4.188248222327794e-06,
+      "loss": 0.007,
+      "step": 392
+    },
+    {
+      "epoch": 1.9486754966887418,
+      "grad_norm": 0.23707984387874603,
+      "learning_rate": 4.183003936214981e-06,
+      "loss": 0.0035,
+      "step": 393
+    },
+    {
+      "epoch": 1.9536423841059603,
+      "grad_norm": 0.3283425569534302,
+      "learning_rate": 4.177746070897593e-06,
+      "loss": 0.007,
+      "step": 394
+    },
+    {
+      "epoch": 1.9586092715231787,
+      "grad_norm": 0.43405845761299133,
+      "learning_rate": 4.172474668798363e-06,
+      "loss": 0.0062,
+      "step": 395
+    },
+    {
+      "epoch": 1.9635761589403975,
+      "grad_norm": 0.4817218482494354,
+      "learning_rate": 4.167189772449248e-06,
+      "loss": 0.0081,
+      "step": 396
+    },
+    {
+      "epoch": 1.968543046357616,
+      "grad_norm": 0.4744330048561096,
+      "learning_rate": 4.16189142449108e-06,
+      "loss": 0.0031,
+      "step": 397
+    },
+    {
+      "epoch": 1.9735099337748343,
+      "grad_norm": 0.302048921585083,
+      "learning_rate": 4.156579667673224e-06,
+      "loss": 0.0024,
+      "step": 398
+    },
+    {
+      "epoch": 1.978476821192053,
+      "grad_norm": 0.24172364175319672,
+      "learning_rate": 4.151254544853235e-06,
+      "loss": 0.0029,
+      "step": 399
+    },
+    {
+      "epoch": 1.9834437086092715,
+      "grad_norm": 0.26656797528266907,
+      "learning_rate": 4.145916098996509e-06,
+      "loss": 0.0066,
+      "step": 400
+    },
+    {
+      "epoch": 1.98841059602649,
+      "grad_norm": 0.35597696900367737,
+      "learning_rate": 4.140564373175939e-06,
+      "loss": 0.0045,
+      "step": 401
+    },
+    {
+      "epoch": 1.9933774834437086,
+      "grad_norm": 0.4242151081562042,
+      "learning_rate": 4.135199410571568e-06,
+      "loss": 0.0057,
+      "step": 402
+    },
+    {
+      "epoch": 1.9983443708609272,
+      "grad_norm": 0.26455965638160706,
+      "learning_rate": 4.129821254470236e-06,
+      "loss": 0.0031,
+      "step": 403
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.26455965638160706,
+      "learning_rate": 4.124429948265235e-06,
+      "loss": 0.0042,
+      "step": 404
+    },
+    {
+      "epoch": 2.0049668874172184,
+      "grad_norm": 1.3334167003631592,
+      "learning_rate": 4.11902553545596e-06,
+      "loss": 0.0027,
+      "step": 405
+    },
+    {
+      "epoch": 2.0099337748344372,
+      "grad_norm": 0.2760920226573944,
+      "learning_rate": 4.113608059647553e-06,
+      "loss": 0.0028,
+      "step": 406
+    },
+    {
+      "epoch": 2.0149006622516556,
+      "grad_norm": 0.17451119422912598,
+      "learning_rate": 4.108177564550555e-06,
+      "loss": 0.0041,
+      "step": 407
+    },
+    {
+      "epoch": 2.019867549668874,
+      "grad_norm": 0.1261451691389084,
+      "learning_rate": 4.10273409398055e-06,
+      "loss": 0.001,
+      "step": 408
+    },
+    {
+      "epoch": 2.024834437086093,
+      "grad_norm": 0.1620579957962036,
+      "learning_rate": 4.097277691857819e-06,
+      "loss": 0.0029,
+      "step": 409
+    },
+    {
+      "epoch": 2.0298013245033113,
+      "grad_norm": 0.17437221109867096,
+      "learning_rate": 4.091808402206977e-06,
+      "loss": 0.0034,
+      "step": 410
+    },
+    {
+      "epoch": 2.0347682119205297,
+      "grad_norm": 0.17286483943462372,
+      "learning_rate": 4.0863262691566186e-06,
+      "loss": 0.0023,
+      "step": 411
+    },
+    {
+      "epoch": 2.0397350993377485,
+      "grad_norm": 0.22569458186626434,
+      "learning_rate": 4.080831336938969e-06,
+      "loss": 0.0024,
+      "step": 412
+    },
+    {
+      "epoch": 2.044701986754967,
+      "grad_norm": 0.2184755802154541,
+      "learning_rate": 4.075323649889522e-06,
+      "loss": 0.0013,
+      "step": 413
+    },
+    {
+      "epoch": 2.0496688741721854,
+      "grad_norm": 0.23581546545028687,
+      "learning_rate": 4.069803252446679e-06,
+      "loss": 0.0021,
+      "step": 414
+    },
+    {
+      "epoch": 2.054635761589404,
+      "grad_norm": 0.20207545161247253,
+      "learning_rate": 4.0642701891514e-06,
+      "loss": 0.0022,
+      "step": 415
+    },
+    {
+      "epoch": 2.0596026490066226,
+      "grad_norm": 0.18072167038917542,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 0.0029,
+      "step": 416
+    },
+    {
+      "epoch": 2.064569536423841,
+      "grad_norm": 0.163478285074234,
+      "learning_rate": 4.053166243677966e-06,
+      "loss": 0.0018,
+      "step": 417
+    },
+    {
+      "epoch": 2.0695364238410594,
+      "grad_norm": 0.22167491912841797,
+      "learning_rate": 4.047595451091252e-06,
+      "loss": 0.0019,
+      "step": 418
+    },
+    {
+      "epoch": 2.0745033112582782,
+      "grad_norm": 0.14667968451976776,
+      "learning_rate": 4.042012171834259e-06,
+      "loss": 0.0018,
+      "step": 419
+    },
+    {
+      "epoch": 2.0794701986754967,
+      "grad_norm": 0.12372825294733047,
+      "learning_rate": 4.036416450955301e-06,
+      "loss": 0.0017,
+      "step": 420
+    },
+    {
+      "epoch": 2.084437086092715,
+      "grad_norm": 0.23118308186531067,
+      "learning_rate": 4.030808333603079e-06,
+      "loss": 0.0013,
+      "step": 421
+    },
+    {
+      "epoch": 2.089403973509934,
+      "grad_norm": 0.33420321345329285,
+      "learning_rate": 4.025187865026311e-06,
+      "loss": 0.002,
+      "step": 422
+    },
+    {
+      "epoch": 2.0943708609271523,
+      "grad_norm": 0.2636815011501312,
+      "learning_rate": 4.019555090573373e-06,
+      "loss": 0.001,
+      "step": 423
+    },
+    {
+      "epoch": 2.0993377483443707,
+      "grad_norm": 0.3790287375450134,
+      "learning_rate": 4.013910055691927e-06,
+      "loss": 0.0054,
+      "step": 424
+    },
+    {
+      "epoch": 2.1043046357615895,
+      "grad_norm": 0.29654383659362793,
+      "learning_rate": 4.008252805928559e-06,
+      "loss": 0.0017,
+      "step": 425
+    },
+    {
+      "epoch": 2.109271523178808,
+      "grad_norm": 0.19992245733737946,
+      "learning_rate": 4.00258338692841e-06,
+      "loss": 0.0013,
+      "step": 426
+    },
+    {
+      "epoch": 2.1142384105960264,
+      "grad_norm": 0.13655278086662292,
+      "learning_rate": 3.996901844434809e-06,
+      "loss": 0.0009,
+      "step": 427
+    },
+    {
+      "epoch": 2.119205298013245,
+      "grad_norm": 0.3091754615306854,
+      "learning_rate": 3.9912082242889e-06,
+      "loss": 0.002,
+      "step": 428
+    },
+    {
+      "epoch": 2.1241721854304636,
+      "grad_norm": 0.15480048954486847,
+      "learning_rate": 3.985502572429277e-06,
+      "loss": 0.0014,
+      "step": 429
+    },
+    {
+      "epoch": 2.129139072847682,
+      "grad_norm": 0.11412292718887329,
+      "learning_rate": 3.979784934891608e-06,
+      "loss": 0.0006,
+      "step": 430
+    },
+    {
+      "epoch": 2.134105960264901,
+      "grad_norm": 0.3870677053928375,
+      "learning_rate": 3.97405535780827e-06,
+      "loss": 0.0028,
+      "step": 431
+    },
+    {
+      "epoch": 2.1390728476821192,
+      "grad_norm": 0.23800161480903625,
+      "learning_rate": 3.9683138874079716e-06,
+      "loss": 0.0003,
+      "step": 432
+    },
+    {
+      "epoch": 2.1440397350993377,
+      "grad_norm": 0.18240606784820557,
+      "learning_rate": 3.962560570015384e-06,
+      "loss": 0.0014,
+      "step": 433
+    },
+    {
+      "epoch": 2.1490066225165565,
+      "grad_norm": 0.46892914175987244,
+      "learning_rate": 3.95679545205076e-06,
+      "loss": 0.0005,
+      "step": 434
+    },
+    {
+      "epoch": 2.153973509933775,
+      "grad_norm": 0.5709494948387146,
+      "learning_rate": 3.951018580029572e-06,
+      "loss": 0.0046,
+      "step": 435
+    },
+    {
+      "epoch": 2.1589403973509933,
+      "grad_norm": 0.1728259176015854,
+      "learning_rate": 3.945230000562121e-06,
+      "loss": 0.0005,
+      "step": 436
+    },
+    {
+      "epoch": 2.1639072847682117,
+      "grad_norm": 0.12012696266174316,
+      "learning_rate": 3.939429760353176e-06,
+      "loss": 0.0004,
+      "step": 437
+    },
+    {
+      "epoch": 2.1688741721854305,
+      "grad_norm": 0.07688380032777786,
+      "learning_rate": 3.933617906201586e-06,
+      "loss": 0.0032,
+      "step": 438
+    },
+    {
+      "epoch": 2.173841059602649,
+      "grad_norm": 0.28859803080558777,
+      "learning_rate": 3.927794484999905e-06,
+      "loss": 0.0002,
+      "step": 439
+    },
+    {
+      "epoch": 2.1788079470198674,
+      "grad_norm": 0.747829020023346,
+      "learning_rate": 3.921959543734021e-06,
+      "loss": 0.006,
+      "step": 440
+    },
+    {
+      "epoch": 2.183774834437086,
+      "grad_norm": 0.23982951045036316,
+      "learning_rate": 3.916113129482762e-06,
+      "loss": 0.0021,
+      "step": 441
+    },
+    {
+      "epoch": 2.1887417218543046,
+      "grad_norm": 0.25354939699172974,
+      "learning_rate": 3.910255289417535e-06,
+      "loss": 0.0029,
+      "step": 442
+    },
+    {
+      "epoch": 2.193708609271523,
+      "grad_norm": 0.5069767832756042,
+      "learning_rate": 3.9043860708019275e-06,
+      "loss": 0.0004,
+      "step": 443
+    },
+    {
+      "epoch": 2.198675496688742,
+      "grad_norm": 0.060515280812978745,
+      "learning_rate": 3.898505520991337e-06,
+      "loss": 0.001,
+      "step": 444
+    },
+    {
+      "epoch": 2.2036423841059603,
+      "grad_norm": 0.3152879476547241,
+      "learning_rate": 3.892613687432587e-06,
+      "loss": 0.0045,
+      "step": 445
+    },
+    {
+      "epoch": 2.2086092715231787,
+      "grad_norm": 0.12870974838733673,
+      "learning_rate": 3.8867106176635405e-06,
+      "loss": 0.0019,
+      "step": 446
+    },
+    {
+      "epoch": 2.2135761589403975,
+      "grad_norm": 0.2350538969039917,
+      "learning_rate": 3.880796359312722e-06,
+      "loss": 0.0015,
+      "step": 447
+    },
+    {
+      "epoch": 2.218543046357616,
+      "grad_norm": 0.15666623413562775,
+      "learning_rate": 3.87487096009893e-06,
+      "loss": 0.0054,
+      "step": 448
+    },
+    {
+      "epoch": 2.2235099337748343,
+      "grad_norm": 0.2582405209541321,
+      "learning_rate": 3.868934467830848e-06,
+      "loss": 0.0004,
+      "step": 449
+    },
+    {
+      "epoch": 2.228476821192053,
+      "grad_norm": 0.10142242908477783,
+      "learning_rate": 3.862986930406669e-06,
+      "loss": 0.0046,
+      "step": 450
+    },
+    {
+      "epoch": 2.2334437086092715,
+      "grad_norm": 0.3277742266654968,
+      "learning_rate": 3.857028395813701e-06,
+      "loss": 0.0022,
+      "step": 451
+    },
+    {
+      "epoch": 2.23841059602649,
+      "grad_norm": 0.060168925672769547,
+      "learning_rate": 3.851058912127979e-06,
+      "loss": 0.0012,
+      "step": 452
+    },
+    {
+      "epoch": 2.243377483443709,
+      "grad_norm": 0.29440394043922424,
+      "learning_rate": 3.845078527513883e-06,
+      "loss": 0.0005,
+      "step": 453
+    },
+    {
+      "epoch": 2.248344370860927,
+      "grad_norm": 0.06611655652523041,
+      "learning_rate": 3.8390872902237485e-06,
+      "loss": 0.0021,
+      "step": 454
+    },
+    {
+      "epoch": 2.2533112582781456,
+      "grad_norm": 0.12377471476793289,
+      "learning_rate": 3.83308524859747e-06,
+      "loss": 0.0029,
+      "step": 455
+    },
+    {
+      "epoch": 2.258278145695364,
+      "grad_norm": 1.3894189596176147,
+      "learning_rate": 3.827072451062118e-06,
+      "loss": 0.0008,
+      "step": 456
+    },
+    {
+      "epoch": 2.263245033112583,
+      "grad_norm": 0.13014522194862366,
+      "learning_rate": 3.821048946131549e-06,
+      "loss": 0.0019,
+      "step": 457
+    },
+    {
+      "epoch": 2.2682119205298013,
+      "grad_norm": 0.21175627410411835,
+      "learning_rate": 3.815014782406008e-06,
+      "loss": 0.0021,
+      "step": 458
+    },
+    {
+      "epoch": 2.27317880794702,
+      "grad_norm": 0.292000412940979,
+      "learning_rate": 3.808970008571741e-06,
+      "loss": 0.0015,
+      "step": 459
+    },
+    {
+      "epoch": 2.2781456953642385,
+      "grad_norm": 0.500977635383606,
+      "learning_rate": 3.8029146734005994e-06,
+      "loss": 0.001,
+      "step": 460
+    },
+    {
+      "epoch": 2.283112582781457,
+      "grad_norm": 0.26836568117141724,
+      "learning_rate": 3.7968488257496527e-06,
+      "loss": 0.0023,
+      "step": 461
+    },
+    {
+      "epoch": 2.2880794701986753,
+      "grad_norm": 0.17154644429683685,
+      "learning_rate": 3.7907725145607856e-06,
+      "loss": 0.003,
+      "step": 462
+    },
+    {
+      "epoch": 2.293046357615894,
+      "grad_norm": 0.3577005863189697,
+      "learning_rate": 3.784685788860306e-06,
+      "loss": 0.0009,
+      "step": 463
+    },
+    {
+      "epoch": 2.2980132450331126,
+      "grad_norm": 0.1332630217075348,
+      "learning_rate": 3.7785886977585562e-06,
+      "loss": 0.0017,
+      "step": 464
+    },
+    {
+      "epoch": 2.302980132450331,
+      "grad_norm": 0.23743191361427307,
+      "learning_rate": 3.772481290449504e-06,
+      "loss": 0.0032,
+      "step": 465
+    },
+    {
+      "epoch": 2.30794701986755,
+      "grad_norm": 0.47357678413391113,
+      "learning_rate": 3.7663636162103583e-06,
+      "loss": 0.0052,
+      "step": 466
+    },
+    {
+      "epoch": 2.312913907284768,
+      "grad_norm": 0.5023639798164368,
+      "learning_rate": 3.7602357244011646e-06,
+      "loss": 0.0018,
+      "step": 467
+    },
+    {
+      "epoch": 2.3178807947019866,
+      "grad_norm": 0.3237062096595764,
+      "learning_rate": 3.7540976644644055e-06,
+      "loss": 0.0006,
+      "step": 468
+    },
+    {
+      "epoch": 2.3228476821192054,
+      "grad_norm": 0.04664800316095352,
+      "learning_rate": 3.747949485924608e-06,
+      "loss": 0.0006,
+      "step": 469
+    },
+    {
+      "epoch": 2.327814569536424,
+      "grad_norm": 0.1138332411646843,
+      "learning_rate": 3.74179123838794e-06,
+      "loss": 0.0002,
+      "step": 470
+    },
+    {
+      "epoch": 2.3327814569536423,
+      "grad_norm": 0.050861991941928864,
+      "learning_rate": 3.7356229715418074e-06,
+      "loss": 0.002,
+      "step": 471
+    },
+    {
+      "epoch": 2.337748344370861,
+      "grad_norm": 0.16194187104701996,
+      "learning_rate": 3.72944473515446e-06,
+      "loss": 0.0002,
+      "step": 472
+    },
+    {
+      "epoch": 2.3427152317880795,
+      "grad_norm": 0.03834187611937523,
+      "learning_rate": 3.7232565790745835e-06,
+      "loss": 0.0008,
+      "step": 473
+    },
+    {
+      "epoch": 2.347682119205298,
+      "grad_norm": 0.32071059942245483,
+      "learning_rate": 3.7170585532309e-06,
+      "loss": 0.0028,
+      "step": 474
+    },
+    {
+      "epoch": 2.3526490066225167,
+      "grad_norm": 0.18589355051517487,
+      "learning_rate": 3.710850707631768e-06,
+      "loss": 0.0009,
+      "step": 475
+    },
+    {
+      "epoch": 2.357615894039735,
+      "grad_norm": 0.13891851902008057,
+      "learning_rate": 3.7046330923647733e-06,
+      "loss": 0.0011,
+      "step": 476
+    },
+    {
+      "epoch": 2.3625827814569536,
+      "grad_norm": 0.23633414506912231,
+      "learning_rate": 3.6984057575963273e-06,
+      "loss": 0.0002,
+      "step": 477
+    },
+    {
+      "epoch": 2.3675496688741724,
+      "grad_norm": 0.041672997176647186,
+      "learning_rate": 3.6921687535712657e-06,
+      "loss": 0.002,
+      "step": 478
+    },
+    {
+      "epoch": 2.372516556291391,
+      "grad_norm": 0.2581554651260376,
+      "learning_rate": 3.685922130612436e-06,
+      "loss": 0.001,
+      "step": 479
+    },
+    {
+      "epoch": 2.377483443708609,
+      "grad_norm": 0.1872328668832779,
+      "learning_rate": 3.679665939120298e-06,
+      "loss": 0.0045,
+      "step": 480
+    },
+    {
+      "epoch": 2.3824503311258276,
+      "grad_norm": 1.451582431793213,
+      "learning_rate": 3.673400229572515e-06,
+      "loss": 0.0003,
+      "step": 481
+    },
+    {
+      "epoch": 2.3874172185430464,
+      "grad_norm": 0.08090563863515854,
+      "learning_rate": 3.6671250525235424e-06,
+      "loss": 0.0016,
+      "step": 482
+    },
+    {
+      "epoch": 2.392384105960265,
+      "grad_norm": 0.2399580329656601,
+      "learning_rate": 3.6608404586042286e-06,
+      "loss": 0.0047,
+      "step": 483
+    },
+    {
+      "epoch": 2.3973509933774833,
+      "grad_norm": 0.2617029547691345,
+      "learning_rate": 3.6545464985213973e-06,
+      "loss": 0.0008,
+      "step": 484
+    },
+    {
+      "epoch": 2.402317880794702,
+      "grad_norm": 0.24318858981132507,
+      "learning_rate": 3.6482432230574448e-06,
+      "loss": 0.0026,
+      "step": 485
+    },
+    {
+      "epoch": 2.4072847682119205,
+      "grad_norm": 0.18111035227775574,
+      "learning_rate": 3.6419306830699275e-06,
+      "loss": 0.0009,
+      "step": 486
+    },
+    {
+      "epoch": 2.412251655629139,
+      "grad_norm": 0.13809381425380707,
+      "learning_rate": 3.63560892949115e-06,
+      "loss": 0.0032,
+      "step": 487
+    },
+    {
+      "epoch": 2.4172185430463577,
+      "grad_norm": 0.17829202115535736,
+      "learning_rate": 3.6292780133277593e-06,
+      "loss": 0.002,
+      "step": 488
+    },
+    {
+      "epoch": 2.422185430463576,
+      "grad_norm": 0.374436616897583,
+      "learning_rate": 3.622937985660329e-06,
+      "loss": 0.0019,
+      "step": 489
+    },
+    {
+      "epoch": 2.4271523178807946,
+      "grad_norm": 0.07355311512947083,
+      "learning_rate": 3.616588897642945e-06,
+      "loss": 0.0008,
+      "step": 490
+    },
+    {
+      "epoch": 2.4321192052980134,
+      "grad_norm": 0.0755995437502861,
+      "learning_rate": 3.6102308005028027e-06,
+      "loss": 0.0004,
+      "step": 491
+    },
+    {
+      "epoch": 2.437086092715232,
+      "grad_norm": 0.09817372262477875,
+      "learning_rate": 3.6038637455397802e-06,
+      "loss": 0.0013,
+      "step": 492
+    },
+    {
+      "epoch": 2.44205298013245,
+      "grad_norm": 0.11779039353132248,
+      "learning_rate": 3.597487784126035e-06,
+      "loss": 0.0014,
+      "step": 493
+    },
+    {
+      "epoch": 2.447019867549669,
+      "grad_norm": 0.21497368812561035,
+      "learning_rate": 3.5911029677055866e-06,
+      "loss": 0.0007,
+      "step": 494
+    },
+    {
+      "epoch": 2.4519867549668874,
+      "grad_norm": 0.06322400271892548,
+      "learning_rate": 3.5847093477938955e-06,
+      "loss": 0.0008,
+      "step": 495
+    },
+    {
+      "epoch": 2.456953642384106,
+      "grad_norm": 0.2135360687971115,
+      "learning_rate": 3.578306975977459e-06,
+      "loss": 0.0044,
+      "step": 496
+    },
+    {
+      "epoch": 2.4619205298013247,
+      "grad_norm": 0.2878398597240448,
+      "learning_rate": 3.5718959039133837e-06,
+      "loss": 0.0016,
+      "step": 497
+    },
+    {
+      "epoch": 2.466887417218543,
+      "grad_norm": 0.15305981040000916,
+      "learning_rate": 3.5654761833289754e-06,
+      "loss": 0.0023,
+      "step": 498
+    },
+    {
+      "epoch": 2.4718543046357615,
+      "grad_norm": 0.23793183267116547,
+      "learning_rate": 3.5590478660213214e-06,
+      "loss": 0.0009,
+      "step": 499
+    },
+    {
+      "epoch": 2.47682119205298,
+      "grad_norm": 0.09096159040927887,
+      "learning_rate": 3.552611003856867e-06,
+      "loss": 0.0007,
+      "step": 500
+    },
+    {
+      "epoch": 2.4817880794701987,
+      "grad_norm": 0.08077141642570496,
+      "learning_rate": 3.5461656487710045e-06,
+      "loss": 0.0003,
+      "step": 501
+    },
+    {
+      "epoch": 2.486754966887417,
+      "grad_norm": 0.04294414445757866,
+      "learning_rate": 3.539711852767651e-06,
+      "loss": 0.0022,
+      "step": 502
+    },
+    {
+      "epoch": 2.491721854304636,
+      "grad_norm": 0.25107285380363464,
+      "learning_rate": 3.533249667918827e-06,
+      "loss": 0.0001,
+      "step": 503
+    },
+    {
+      "epoch": 2.4966887417218544,
+      "grad_norm": 0.04402798041701317,
+      "learning_rate": 3.526779146364237e-06,
+      "loss": 0.001,
+      "step": 504
+    },
+    {
+      "epoch": 2.501655629139073,
+      "grad_norm": 0.18737560510635376,
+      "learning_rate": 3.520300340310852e-06,
+      "loss": 0.0013,
+      "step": 505
+    },
+    {
+      "epoch": 2.506622516556291,
+      "grad_norm": 0.13711388409137726,
+      "learning_rate": 3.513813302032485e-06,
+      "loss": 0.0002,
+      "step": 506
+    },
+    {
+      "epoch": 2.51158940397351,
+      "grad_norm": 0.04720345884561539,
+      "learning_rate": 3.50731808386937e-06,
+      "loss": 0.0002,
+      "step": 507
+    },
+    {
+      "epoch": 2.5165562913907285,
+      "grad_norm": 0.09136200696229935,
+      "learning_rate": 3.500814738227739e-06,
+      "loss": 0.0015,
+      "step": 508
+    },
+    {
+      "epoch": 2.5215231788079473,
+      "grad_norm": 0.15046486258506775,
+      "learning_rate": 3.4943033175794017e-06,
+      "loss": 0.0002,
+      "step": 509
+    },
+    {
+      "epoch": 2.5264900662251657,
+      "grad_norm": 0.05822240933775902,
+      "learning_rate": 3.4877838744613196e-06,
+      "loss": 0.0002,
+      "step": 510
+    },
+    {
+      "epoch": 2.531456953642384,
+      "grad_norm": 0.017863569781184196,
+      "learning_rate": 3.4812564614751825e-06,
+      "loss": 0.001,
+      "step": 511
+    },
+    {
+      "epoch": 2.5364238410596025,
+      "grad_norm": 0.16180965304374695,
+      "learning_rate": 3.4747211312869854e-06,
+      "loss": 0.0011,
+      "step": 512
+    },
+    {
+      "epoch": 2.5413907284768213,
+      "grad_norm": 0.5808257460594177,
+      "learning_rate": 3.4681779366266033e-06,
+      "loss": 0.0019,
+      "step": 513
+    },
+    {
+      "epoch": 2.5463576158940397,
+      "grad_norm": 0.04765927046537399,
+      "learning_rate": 3.461626930287362e-06,
+      "loss": 0.0008,
+      "step": 514
+    },
+    {
+      "epoch": 2.551324503311258,
+      "grad_norm": 0.2293013632297516,
+      "learning_rate": 3.4550681651256194e-06,
+      "loss": 0.0011,
+      "step": 515
+    },
+    {
+      "epoch": 2.556291390728477,
+      "grad_norm": 0.23624388873577118,
+      "learning_rate": 3.448501694060332e-06,
+      "loss": 0.0049,
+      "step": 516
+    },
+    {
+      "epoch": 2.5612582781456954,
+      "grad_norm": 0.41582730412483215,
+      "learning_rate": 3.4419275700726325e-06,
+      "loss": 0.0003,
+      "step": 517
+    },
+    {
+      "epoch": 2.566225165562914,
+      "grad_norm": 0.047652896493673325,
+      "learning_rate": 3.4353458462054e-06,
+      "loss": 0.0002,
+      "step": 518
+    },
+    {
+      "epoch": 2.571192052980132,
+      "grad_norm": 0.035407904535532,
+      "learning_rate": 3.428756575562833e-06,
+      "loss": 0.0002,
+      "step": 519
+    },
+    {
+      "epoch": 2.576158940397351,
+      "grad_norm": 0.034042540937662125,
+      "learning_rate": 3.4221598113100196e-06,
+      "loss": 0.0029,
+      "step": 520
+    },
+    {
+      "epoch": 2.5811258278145695,
+      "grad_norm": 0.15980283915996552,
+      "learning_rate": 3.4155556066725125e-06,
+      "loss": 0.001,
+      "step": 521
+    },
+    {
+      "epoch": 2.5860927152317883,
+      "grad_norm": 0.18785570561885834,
+      "learning_rate": 3.408944014935892e-06,
+      "loss": 0.0008,
+      "step": 522
+    },
+    {
+      "epoch": 2.5910596026490067,
+      "grad_norm": 0.33988872170448303,
+      "learning_rate": 3.4023250894453465e-06,
+      "loss": 0.001,
+      "step": 523
+    },
+    {
+      "epoch": 2.596026490066225,
+      "grad_norm": 0.2074756771326065,
+      "learning_rate": 3.3956988836052294e-06,
+      "loss": 0.0003,
+      "step": 524
+    },
+    {
+      "epoch": 2.6009933774834435,
+      "grad_norm": 0.06671739369630814,
+      "learning_rate": 3.389065450878642e-06,
+      "loss": 0.0003,
+      "step": 525
+    },
+    {
+      "epoch": 2.6059602649006623,
+      "grad_norm": 0.029861386865377426,
+      "learning_rate": 3.38242484478699e-06,
+      "loss": 0.0026,
+      "step": 526
+    },
+    {
+      "epoch": 2.6109271523178808,
+      "grad_norm": 0.4710339605808258,
+      "learning_rate": 3.375777118909561e-06,
+      "loss": 0.0005,
+      "step": 527
+    },
+    {
+      "epoch": 2.6158940397350996,
+      "grad_norm": 0.23238687217235565,
+      "learning_rate": 3.369122326883085e-06,
+      "loss": 0.0008,
+      "step": 528
+    },
+    {
+      "epoch": 2.620860927152318,
+      "grad_norm": 0.02251332439482212,
+      "learning_rate": 3.3624605224013057e-06,
+      "loss": 0.0018,
+      "step": 529
+    },
+    {
+      "epoch": 2.6258278145695364,
+      "grad_norm": 0.3648238182067871,
+      "learning_rate": 3.3557917592145467e-06,
+      "loss": 0.0003,
+      "step": 530
+    },
+    {
+      "epoch": 2.630794701986755,
+      "grad_norm": 0.3544326424598694,
+      "learning_rate": 3.349116091129278e-06,
+      "loss": 0.0002,
+      "step": 531
+    },
+    {
+      "epoch": 2.6357615894039736,
+      "grad_norm": 0.06514076888561249,
+      "learning_rate": 3.3424335720076794e-06,
+      "loss": 0.0008,
+      "step": 532
+    },
+    {
+      "epoch": 2.640728476821192,
+      "grad_norm": 0.520621657371521,
+      "learning_rate": 3.3357442557672096e-06,
+      "loss": 0.0002,
+      "step": 533
+    },
+    {
+      "epoch": 2.6456953642384105,
+      "grad_norm": 0.033616483211517334,
+      "learning_rate": 3.32904819638017e-06,
+      "loss": 0.0004,
+      "step": 534
+    },
+    {
+      "epoch": 2.6506622516556293,
+      "grad_norm": 0.08390014618635178,
+      "learning_rate": 3.3223454478732653e-06,
+      "loss": 0.0001,
+      "step": 535
+    },
+    {
+      "epoch": 2.6556291390728477,
+      "grad_norm": 0.07171134650707245,
+      "learning_rate": 3.315636064327174e-06,
+      "loss": 0.0003,
+      "step": 536
+    },
+    {
+      "epoch": 2.660596026490066,
+      "grad_norm": 0.05604681745171547,
+      "learning_rate": 3.3089200998761086e-06,
+      "loss": 0.0022,
+      "step": 537
+    },
+    {
+      "epoch": 2.6655629139072845,
+      "grad_norm": 0.10902980715036392,
+      "learning_rate": 3.3021976087073772e-06,
+      "loss": 0.0002,
+      "step": 538
+    },
+    {
+      "epoch": 2.6705298013245033,
+      "grad_norm": 0.2084953933954239,
+      "learning_rate": 3.2954686450609513e-06,
+      "loss": 0.0014,
+      "step": 539
+    },
+    {
+      "epoch": 2.6754966887417218,
+      "grad_norm": 0.3191259205341339,
+      "learning_rate": 3.2887332632290225e-06,
+      "loss": 0.0016,
+      "step": 540
+    },
+    {
+      "epoch": 2.6804635761589406,
+      "grad_norm": 0.12397249788045883,
+      "learning_rate": 3.281991517555568e-06,
+      "loss": 0.0025,
+      "step": 541
+    },
+    {
+      "epoch": 2.685430463576159,
+      "grad_norm": 0.3077525198459625,
+      "learning_rate": 3.275243462435913e-06,
+      "loss": 0.0004,
+      "step": 542
+    },
+    {
+      "epoch": 2.6903973509933774,
+      "grad_norm": 0.11203525960445404,
+      "learning_rate": 3.268489152316286e-06,
+      "loss": 0.0001,
+      "step": 543
+    },
+    {
+      "epoch": 2.695364238410596,
+      "grad_norm": 0.016934402287006378,
+      "learning_rate": 3.2617286416933876e-06,
+      "loss": 0.0002,
+      "step": 544
+    },
+    {
+      "epoch": 2.7003311258278146,
+      "grad_norm": 0.09976313263177872,
+      "learning_rate": 3.254961985113945e-06,
+      "loss": 0.0009,
+      "step": 545
+    },
+    {
+      "epoch": 2.705298013245033,
+      "grad_norm": 0.03238361328840256,
+      "learning_rate": 3.248189237174273e-06,
+      "loss": 0.0006,
+      "step": 546
+    },
+    {
+      "epoch": 2.710264900662252,
+      "grad_norm": 0.06707025319337845,
+      "learning_rate": 3.2414104525198354e-06,
+      "loss": 0.0046,
+      "step": 547
+    },
+    {
+      "epoch": 2.7152317880794703,
+      "grad_norm": 0.6409168243408203,
+      "learning_rate": 3.234625685844803e-06,
+      "loss": 0.0014,
+      "step": 548
+    },
+    {
+      "epoch": 2.7201986754966887,
+      "grad_norm": 0.16995637118816376,
+      "learning_rate": 3.2278349918916096e-06,
+      "loss": 0.0008,
+      "step": 549
+    },
+    {
+      "epoch": 2.725165562913907,
+      "grad_norm": 0.11772691458463669,
+      "learning_rate": 3.2210384254505166e-06,
+      "loss": 0.0003,
+      "step": 550
+    },
+    {
+      "epoch": 2.730132450331126,
+      "grad_norm": 0.04289670288562775,
+      "learning_rate": 3.2142360413591643e-06,
+      "loss": 0.0004,
+      "step": 551
+    },
+    {
+      "epoch": 2.7350993377483444,
+      "grad_norm": 0.18225699663162231,
+      "learning_rate": 3.207427894502133e-06,
+      "loss": 0.0002,
+      "step": 552
+    },
+    {
+      "epoch": 2.7400662251655628,
+      "grad_norm": 0.05074049532413483,
+      "learning_rate": 3.2006140398105014e-06,
+      "loss": 0.0002,
+      "step": 553
+    },
+    {
+      "epoch": 2.7450331125827816,
+      "grad_norm": 0.10272341966629028,
+      "learning_rate": 3.1937945322613976e-06,
+      "loss": 0.0011,
+      "step": 554
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.16255953907966614,
+      "learning_rate": 3.186969426877563e-06,
+      "loss": 0.0003,
+      "step": 555
+    },
+    {
+      "epoch": 2.7549668874172184,
+      "grad_norm": 0.06192003935575485,
+      "learning_rate": 3.1801387787269043e-06,
+      "loss": 0.0002,
+      "step": 556
+    },
+    {
+      "epoch": 2.7599337748344372,
+      "grad_norm": 0.032033104449510574,
+      "learning_rate": 3.1733026429220464e-06,
+      "loss": 0.0001,
+      "step": 557
+    },
+    {
+      "epoch": 2.7649006622516556,
+      "grad_norm": 0.12052201479673386,
+      "learning_rate": 3.1664610746198953e-06,
+      "loss": 0.0009,
+      "step": 558
+    },
+    {
+      "epoch": 2.769867549668874,
+      "grad_norm": 0.020031720399856567,
+      "learning_rate": 3.1596141290211857e-06,
+      "loss": 0.0013,
+      "step": 559
+    },
+    {
+      "epoch": 2.774834437086093,
+      "grad_norm": 0.16622111201286316,
+      "learning_rate": 3.15276186137004e-06,
+      "loss": 0.0002,
+      "step": 560
+    },
+    {
+      "epoch": 2.7798013245033113,
+      "grad_norm": 0.09004916995763779,
+      "learning_rate": 3.1459043269535216e-06,
+      "loss": 0.0007,
+      "step": 561
+    },
+    {
+      "epoch": 2.7847682119205297,
+      "grad_norm": 0.26586073637008667,
+      "learning_rate": 3.139041581101187e-06,
+      "loss": 0.0017,
+      "step": 562
+    },
+    {
+      "epoch": 2.789735099337748,
+      "grad_norm": 0.10252036154270172,
+      "learning_rate": 3.132173679184642e-06,
+      "loss": 0.0006,
+      "step": 563
+    },
+    {
+      "epoch": 2.794701986754967,
+      "grad_norm": 0.18708603084087372,
+      "learning_rate": 3.1253006766170934e-06,
+      "loss": 0.0006,
+      "step": 564
+    },
+    {
+      "epoch": 2.7996688741721854,
+      "grad_norm": 0.08964866399765015,
+      "learning_rate": 3.118422628852901e-06,
+      "loss": 0.0007,
+      "step": 565
+    },
+    {
+      "epoch": 2.804635761589404,
+      "grad_norm": 0.11937995254993439,
+      "learning_rate": 3.1115395913871355e-06,
+      "loss": 0.0002,
+      "step": 566
+    },
+    {
+      "epoch": 2.8096026490066226,
+      "grad_norm": 0.020172245800495148,
+      "learning_rate": 3.1046516197551206e-06,
+      "loss": 0.0002,
+      "step": 567
+    },
+    {
+      "epoch": 2.814569536423841,
+      "grad_norm": 0.24419821798801422,
+      "learning_rate": 3.0977587695319965e-06,
+      "loss": 0.0012,
+      "step": 568
+    },
+    {
+      "epoch": 2.8195364238410594,
+      "grad_norm": 0.05208774283528328,
+      "learning_rate": 3.090861096332263e-06,
+      "loss": 0.0001,
+      "step": 569
+    },
+    {
+      "epoch": 2.8245033112582782,
+      "grad_norm": 0.023159924894571304,
+      "learning_rate": 3.0839586558093337e-06,
+      "loss": 0.0002,
+      "step": 570
+    },
+    {
+      "epoch": 2.8294701986754967,
+      "grad_norm": 0.054164085537195206,
+      "learning_rate": 3.077051503655089e-06,
+      "loss": 0.0002,
+      "step": 571
+    },
+    {
+      "epoch": 2.8344370860927155,
+      "grad_norm": 0.039680492132902145,
+      "learning_rate": 3.0701396955994233e-06,
+      "loss": 0.0001,
+      "step": 572
+    },
+    {
+      "epoch": 2.839403973509934,
+      "grad_norm": 0.03068345971405506,
+      "learning_rate": 3.063223287409797e-06,
+      "loss": 0.0002,
+      "step": 573
+    },
+    {
+      "epoch": 2.8443708609271523,
+      "grad_norm": 0.02883596159517765,
+      "learning_rate": 3.056302334890786e-06,
+      "loss": 0.0018,
+      "step": 574
+    },
+    {
+      "epoch": 2.8493377483443707,
+      "grad_norm": 0.27271297574043274,
+      "learning_rate": 3.049376893883633e-06,
+      "loss": 0.0002,
+      "step": 575
+    },
+    {
+      "epoch": 2.8543046357615895,
+      "grad_norm": 0.01868012733757496,
+      "learning_rate": 3.0424470202657953e-06,
+      "loss": 0.0003,
+      "step": 576
+    },
+    {
+      "epoch": 2.859271523178808,
+      "grad_norm": 0.07278701663017273,
+      "learning_rate": 3.035512769950494e-06,
+      "loss": 0.0004,
+      "step": 577
+    },
+    {
+      "epoch": 2.8642384105960264,
+      "grad_norm": 0.1979231983423233,
+      "learning_rate": 3.0285741988862623e-06,
+      "loss": 0.0003,
+      "step": 578
+    },
+    {
+      "epoch": 2.869205298013245,
+      "grad_norm": 0.0352243073284626,
+      "learning_rate": 3.0216313630564975e-06,
+      "loss": 0.0003,
+      "step": 579
+    },
+    {
+      "epoch": 2.8741721854304636,
+      "grad_norm": 0.15890070796012878,
+      "learning_rate": 3.014684318479006e-06,
+      "loss": 0.0003,
+      "step": 580
+    },
+    {
+      "epoch": 2.879139072847682,
+      "grad_norm": 0.034408748149871826,
+      "learning_rate": 3.00773312120555e-06,
+      "loss": 0.0002,
+      "step": 581
+    },
+    {
+      "epoch": 2.8841059602649004,
+      "grad_norm": 0.035290759056806564,
+      "learning_rate": 3.000777827321402e-06,
+      "loss": 0.0001,
+      "step": 582
+    },
+    {
+      "epoch": 2.8890728476821192,
+      "grad_norm": 0.01860523782670498,
+      "learning_rate": 2.993818492944882e-06,
+      "loss": 0.0001,
+      "step": 583
+    },
+    {
+      "epoch": 2.8940397350993377,
+      "grad_norm": 0.01943597011268139,
+      "learning_rate": 2.9868551742269154e-06,
+      "loss": 0.0001,
+      "step": 584
+    },
+    {
+      "epoch": 2.8990066225165565,
+      "grad_norm": 0.03670092299580574,
+      "learning_rate": 2.9798879273505734e-06,
+      "loss": 0.0002,
+      "step": 585
+    },
+    {
+      "epoch": 2.903973509933775,
+      "grad_norm": 0.01346322800964117,
+      "learning_rate": 2.972916808530619e-06,
+      "loss": 0.0001,
+      "step": 586
+    },
+    {
+      "epoch": 2.9089403973509933,
+      "grad_norm": 0.017677849158644676,
+      "learning_rate": 2.965941874013059e-06,
+      "loss": 0.0001,
+      "step": 587
+    },
+    {
+      "epoch": 2.9139072847682117,
+      "grad_norm": 0.01772226393222809,
+      "learning_rate": 2.9589631800746865e-06,
+      "loss": 0.0001,
+      "step": 588
+    },
+    {
+      "epoch": 2.9188741721854305,
+      "grad_norm": 0.02305142767727375,
+      "learning_rate": 2.951980783022624e-06,
+      "loss": 0.0001,
+      "step": 589
+    },
+    {
+      "epoch": 2.923841059602649,
+      "grad_norm": 0.004405085928738117,
+      "learning_rate": 2.9449947391938768e-06,
+      "loss": 0.0001,
+      "step": 590
+    },
+    {
+      "epoch": 2.928807947019868,
+      "grad_norm": 0.027060767635703087,
+      "learning_rate": 2.9380051049548695e-06,
+      "loss": 0.0002,
+      "step": 591
+    },
+    {
+      "epoch": 2.933774834437086,
+      "grad_norm": 0.28232133388519287,
+      "learning_rate": 2.9310119367009995e-06,
+      "loss": 0.0001,
+      "step": 592
+    },
+    {
+      "epoch": 2.9387417218543046,
+      "grad_norm": 0.1976020634174347,
+      "learning_rate": 2.924015290856177e-06,
+      "loss": 0.0038,
+      "step": 593
+    },
+    {
+      "epoch": 2.943708609271523,
+      "grad_norm": 0.3534559905529022,
+      "learning_rate": 2.9170152238723694e-06,
+      "loss": 0.0001,
+      "step": 594
+    },
+    {
+      "epoch": 2.948675496688742,
+      "grad_norm": 0.005455069709569216,
+      "learning_rate": 2.9100117922291483e-06,
+      "loss": 0.0,
+      "step": 595
+    },
+    {
+      "epoch": 2.9536423841059603,
+      "grad_norm": 0.01252190675586462,
+      "learning_rate": 2.903005052433234e-06,
+      "loss": 0.0002,
+      "step": 596
+    },
+    {
+      "epoch": 2.9586092715231787,
+      "grad_norm": 0.14540104568004608,
+      "learning_rate": 2.8959950610180376e-06,
+      "loss": 0.0001,
+      "step": 597
+    },
+    {
+      "epoch": 2.9635761589403975,
+      "grad_norm": 0.00811131950467825,
+      "learning_rate": 2.888981874543205e-06,
+      "loss": 0.0001,
+      "step": 598
+    },
+    {
+      "epoch": 2.968543046357616,
+      "grad_norm": 0.005490665789693594,
+      "learning_rate": 2.8819655495941614e-06,
+      "loss": 0.0001,
+      "step": 599
+    },
+    {
+      "epoch": 2.9735099337748343,
+      "grad_norm": 0.08258894085884094,
+      "learning_rate": 2.874946142781655e-06,
+      "loss": 0.0002,
+      "step": 600
+    },
+    {
+      "epoch": 2.9784768211920527,
+      "grad_norm": 0.06661806255578995,
+      "learning_rate": 2.8679237107413003e-06,
+      "loss": 0.0002,
+      "step": 601
+    },
+    {
+      "epoch": 2.9834437086092715,
+      "grad_norm": 0.05473367124795914,
+      "learning_rate": 2.860898310133119e-06,
+      "loss": 0.0001,
+      "step": 602
+    },
+    {
+      "epoch": 2.98841059602649,
+      "grad_norm": 0.16531488299369812,
+      "learning_rate": 2.853869997641086e-06,
+      "loss": 0.0014,
+      "step": 603
+    },
+    {
+      "epoch": 2.993377483443709,
+      "grad_norm": 0.28194862604141235,
+      "learning_rate": 2.8468388299726714e-06,
+      "loss": 0.0001,
+      "step": 604
+    },
+    {
+      "epoch": 2.998344370860927,
+      "grad_norm": 0.08649928122758865,
+      "learning_rate": 2.839804863858378e-06,
+      "loss": 0.0007,
+      "step": 605
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.006507672369480133,
+      "learning_rate": 2.832768156051293e-06,
+      "loss": 0.0,
+      "step": 606
+    },
+    {
+      "epoch": 3.0049668874172184,
+      "grad_norm": 0.004866187926381826,
+      "learning_rate": 2.8257287633266207e-06,
+      "loss": 0.0001,
+      "step": 607
+    },
+    {
+      "epoch": 3.0099337748344372,
+      "grad_norm": 0.009720404632389545,
+      "learning_rate": 2.81868674248123e-06,
+      "loss": 0.0001,
+      "step": 608
+    },
+    {
+      "epoch": 3.0149006622516556,
+      "grad_norm": 0.009803484193980694,
+      "learning_rate": 2.8116421503331958e-06,
+      "loss": 0.0001,
+      "step": 609
+    },
+    {
+      "epoch": 3.019867549668874,
+      "grad_norm": 0.033285852521657944,
+      "learning_rate": 2.8045950437213376e-06,
+      "loss": 0.0002,
+      "step": 610
+    },
+    {
+      "epoch": 3.024834437086093,
+      "grad_norm": 0.006313662976026535,
+      "learning_rate": 2.7975454795047624e-06,
+      "loss": 0.0001,
+      "step": 611
+    },
+    {
+      "epoch": 3.0298013245033113,
+      "grad_norm": 0.14627031981945038,
+      "learning_rate": 2.7904935145624085e-06,
+      "loss": 0.0011,
+      "step": 612
+    },
+    {
+      "epoch": 3.0347682119205297,
+      "grad_norm": 0.06508054584264755,
+      "learning_rate": 2.783439205792581e-06,
+      "loss": 0.0004,
+      "step": 613
+    },
+    {
+      "epoch": 3.0397350993377485,
+      "grad_norm": 0.010734508745372295,
+      "learning_rate": 2.7763826101124996e-06,
+      "loss": 0.0001,
+      "step": 614
+    },
+    {
+      "epoch": 3.044701986754967,
+      "grad_norm": 0.011842168867588043,
+      "learning_rate": 2.7693237844578342e-06,
+      "loss": 0.0001,
+      "step": 615
+    },
+    {
+      "epoch": 3.0496688741721854,
+      "grad_norm": 0.008452030830085278,
+      "learning_rate": 2.7622627857822453e-06,
+      "loss": 0.0001,
+      "step": 616
+    },
+    {
+      "epoch": 3.054635761589404,
+      "grad_norm": 0.007807641755789518,
+      "learning_rate": 2.7551996710569296e-06,
+      "loss": 0.0001,
+      "step": 617
+    },
+    {
+      "epoch": 3.0596026490066226,
+      "grad_norm": 0.008009343408048153,
+      "learning_rate": 2.7481344972701545e-06,
+      "loss": 0.0001,
+      "step": 618
+    },
+    {
+      "epoch": 3.064569536423841,
+      "grad_norm": 0.014504185877740383,
+      "learning_rate": 2.741067321426802e-06,
+      "loss": 0.0001,
+      "step": 619
+    },
+    {
+      "epoch": 3.0695364238410594,
+      "grad_norm": 0.014211749657988548,
+      "learning_rate": 2.7339982005479065e-06,
+      "loss": 0.0001,
+      "step": 620
+    },
+    {
+      "epoch": 3.0745033112582782,
+      "grad_norm": 0.01568361185491085,
+      "learning_rate": 2.726927191670197e-06,
+      "loss": 0.0001,
+      "step": 621
+    },
+    {
+      "epoch": 3.0794701986754967,
+      "grad_norm": 0.011463419534265995,
+      "learning_rate": 2.719854351845636e-06,
+      "loss": 0.0001,
+      "step": 622
+    },
+    {
+      "epoch": 3.084437086092715,
+      "grad_norm": 0.015127774327993393,
+      "learning_rate": 2.712779738140957e-06,
+      "loss": 0.0001,
+      "step": 623
+    },
+    {
+      "epoch": 3.089403973509934,
+      "grad_norm": 0.013217778876423836,
+      "learning_rate": 2.7057034076372076e-06,
+      "loss": 0.0001,
+      "step": 624
+    },
+    {
+      "epoch": 3.0943708609271523,
+      "grad_norm": 0.0072747692465782166,
+      "learning_rate": 2.6986254174292865e-06,
+      "loss": 0.0,
+      "step": 625
+    },
+    {
+      "epoch": 3.0993377483443707,
+      "grad_norm": 0.017146291211247444,
+      "learning_rate": 2.691545824625483e-06,
+      "loss": 0.0001,
+      "step": 626
+    },
+    {
+      "epoch": 3.1043046357615895,
+      "grad_norm": 0.05981040373444557,
+      "learning_rate": 2.6844646863470187e-06,
+      "loss": 0.0002,
+      "step": 627
+    },
+    {
+      "epoch": 3.109271523178808,
+      "grad_norm": 0.009139766916632652,
+      "learning_rate": 2.6773820597275833e-06,
+      "loss": 0.0001,
+      "step": 628
+    },
+    {
+      "epoch": 3.1142384105960264,
+      "grad_norm": 0.009842149913311005,
+      "learning_rate": 2.6702980019128753e-06,
+      "loss": 0.0001,
+      "step": 629
+    },
+    {
+      "epoch": 3.119205298013245,
+      "grad_norm": 0.006375189870595932,
+      "learning_rate": 2.6632125700601413e-06,
+      "loss": 0.0001,
+      "step": 630
+    },
+    {
+      "epoch": 3.1241721854304636,
+      "grad_norm": 0.012160423211753368,
+      "learning_rate": 2.656125821337714e-06,
+      "loss": 0.0001,
+      "step": 631
+    },
+    {
+      "epoch": 3.129139072847682,
+      "grad_norm": 0.00759340962395072,
+      "learning_rate": 2.64903781292455e-06,
+      "loss": 0.0001,
+      "step": 632
+    },
+    {
+      "epoch": 3.134105960264901,
+      "grad_norm": 0.02468971349298954,
+      "learning_rate": 2.6419486020097718e-06,
+      "loss": 0.0001,
+      "step": 633
+    },
+    {
+      "epoch": 3.1390728476821192,
+      "grad_norm": 0.025819886475801468,
+      "learning_rate": 2.634858245792201e-06,
+      "loss": 0.0001,
+      "step": 634
+    },
+    {
+      "epoch": 3.1440397350993377,
+      "grad_norm": 0.007346616592258215,
+      "learning_rate": 2.6277668014799045e-06,
+      "loss": 0.0001,
+      "step": 635
+    },
+    {
+      "epoch": 3.1490066225165565,
+      "grad_norm": 0.0058665769174695015,
+      "learning_rate": 2.620674326289725e-06,
+      "loss": 0.0001,
+      "step": 636
+    },
+    {
+      "epoch": 3.153973509933775,
+      "grad_norm": 0.10270977020263672,
+      "learning_rate": 2.6135808774468224e-06,
+      "loss": 0.0001,
+      "step": 637
+    },
+    {
+      "epoch": 3.1589403973509933,
+      "grad_norm": 0.0043976991437375546,
+      "learning_rate": 2.6064865121842154e-06,
+      "loss": 0.0,
+      "step": 638
+    },
+    {
+      "epoch": 3.1639072847682117,
+      "grad_norm": 0.007021864876151085,
+      "learning_rate": 2.599391287742315e-06,
+      "loss": 0.0,
+      "step": 639
+    },
+    {
+      "epoch": 3.1688741721854305,
+      "grad_norm": 0.005025348160415888,
+      "learning_rate": 2.5922952613684634e-06,
+      "loss": 0.0001,
+      "step": 640
+    },
+    {
+      "epoch": 3.173841059602649,
+      "grad_norm": 0.006400761194527149,
+      "learning_rate": 2.585198490316475e-06,
+      "loss": 0.0001,
+      "step": 641
+    },
+    {
+      "epoch": 3.1788079470198674,
+      "grad_norm": 0.011485670693218708,
+      "learning_rate": 2.578101031846172e-06,
+      "loss": 0.0001,
+      "step": 642
+    },
+    {
+      "epoch": 3.183774834437086,
+      "grad_norm": 0.004010629840195179,
+      "learning_rate": 2.5710029432229228e-06,
+      "loss": 0.0,
+      "step": 643
+    },
+    {
+      "epoch": 3.1887417218543046,
+      "grad_norm": 0.024299191311001778,
+      "learning_rate": 2.5639042817171806e-06,
+      "loss": 0.0001,
+      "step": 644
+    },
+    {
+      "epoch": 3.193708609271523,
+      "grad_norm": 0.023588698357343674,
+      "learning_rate": 2.5568051046040197e-06,
+      "loss": 0.0001,
+      "step": 645
+    },
+    {
+      "epoch": 3.198675496688742,
+      "grad_norm": 0.009379717521369457,
+      "learning_rate": 2.5497054691626754e-06,
+      "loss": 0.0001,
+      "step": 646
+    },
+    {
+      "epoch": 3.2036423841059603,
+      "grad_norm": 0.0047790114767849445,
+      "learning_rate": 2.542605432676082e-06,
+      "loss": 0.0001,
+      "step": 647
+    },
+    {
+      "epoch": 3.2086092715231787,
+      "grad_norm": 0.1327018141746521,
+      "learning_rate": 2.535505052430407e-06,
+      "loss": 0.0003,
+      "step": 648
+    },
+    {
+      "epoch": 3.2135761589403975,
+      "grad_norm": 0.0033648740500211716,
+      "learning_rate": 2.528404385714594e-06,
+      "loss": 0.0,
+      "step": 649
+    },
+    {
+      "epoch": 3.218543046357616,
+      "grad_norm": 0.035580579191446304,
+      "learning_rate": 2.5213034898198963e-06,
+      "loss": 0.0002,
+      "step": 650
+    },
+    {
+      "epoch": 3.2235099337748343,
+      "grad_norm": 0.004227795172482729,
+      "learning_rate": 2.514202422039417e-06,
+      "loss": 0.0,
+      "step": 651
+    },
+    {
+      "epoch": 3.228476821192053,
+      "grad_norm": 0.003661267925053835,
+      "learning_rate": 2.5071012396676475e-06,
+      "loss": 0.0,
+      "step": 652
+    },
+    {
+      "epoch": 3.2334437086092715,
+      "grad_norm": 0.00439853873103857,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0,
+      "step": 653
+    },
+    {
+      "epoch": 3.23841059602649,
+      "grad_norm": 0.16040471196174622,
+      "learning_rate": 2.4928987603323533e-06,
+      "loss": 0.0004,
+      "step": 654
+    },
+    {
+      "epoch": 3.243377483443709,
+      "grad_norm": 0.020739218220114708,
+      "learning_rate": 2.485797577960583e-06,
+      "loss": 0.0001,
+      "step": 655
+    },
+    {
+      "epoch": 3.248344370860927,
+      "grad_norm": 0.13917802274227142,
+      "learning_rate": 2.478696510180105e-06,
+      "loss": 0.0004,
+      "step": 656
+    },
+    {
+      "epoch": 3.2533112582781456,
+      "grad_norm": 0.007261666003614664,
+      "learning_rate": 2.471595614285407e-06,
+      "loss": 0.0001,
+      "step": 657
+    },
+    {
+      "epoch": 3.258278145695364,
+      "grad_norm": 0.005395522806793451,
+      "learning_rate": 2.464494947569594e-06,
+      "loss": 0.0001,
+      "step": 658
+    },
+    {
+      "epoch": 3.263245033112583,
+      "grad_norm": 0.002714567119255662,
+      "learning_rate": 2.457394567323919e-06,
+      "loss": 0.0,
+      "step": 659
+    },
+    {
+      "epoch": 3.2682119205298013,
+      "grad_norm": 0.005617187824100256,
+      "learning_rate": 2.4502945308373246e-06,
+      "loss": 0.0001,
+      "step": 660
+    },
+    {
+      "epoch": 3.27317880794702,
+      "grad_norm": 0.21615061163902283,
+      "learning_rate": 2.4431948953959807e-06,
+      "loss": 0.0007,
+      "step": 661
+    },
+    {
+      "epoch": 3.2781456953642385,
+      "grad_norm": 0.01344997901469469,
+      "learning_rate": 2.4360957182828203e-06,
+      "loss": 0.0001,
+      "step": 662
+    },
+    {
+      "epoch": 3.283112582781457,
+      "grad_norm": 0.006362255197018385,
+      "learning_rate": 2.428997056777078e-06,
+      "loss": 0.0001,
+      "step": 663
+    },
+    {
+      "epoch": 3.2880794701986753,
+      "grad_norm": 0.0031819718424230814,
+      "learning_rate": 2.421898968153829e-06,
+      "loss": 0.0,
+      "step": 664
+    },
+    {
+      "epoch": 3.293046357615894,
+      "grad_norm": 0.005469321273267269,
+      "learning_rate": 2.4148015096835255e-06,
+      "loss": 0.0001,
+      "step": 665
+    },
+    {
+      "epoch": 3.2980132450331126,
+      "grad_norm": 0.004727331921458244,
+      "learning_rate": 2.4077047386315375e-06,
+      "loss": 0.0001,
+      "step": 666
+    },
+    {
+      "epoch": 3.302980132450331,
+      "grad_norm": 0.033079843968153,
+      "learning_rate": 2.4006087122576867e-06,
+      "loss": 0.0001,
+      "step": 667
+    },
+    {
+      "epoch": 3.30794701986755,
+      "grad_norm": 0.018462467938661575,
+      "learning_rate": 2.3935134878157854e-06,
+      "loss": 0.0001,
+      "step": 668
+    },
+    {
+      "epoch": 3.312913907284768,
+      "grad_norm": 0.004247081000357866,
+      "learning_rate": 2.3864191225531784e-06,
+      "loss": 0.0,
+      "step": 669
+    },
+    {
+      "epoch": 3.3178807947019866,
+      "grad_norm": 0.2113010585308075,
+      "learning_rate": 2.379325673710276e-06,
+      "loss": 0.001,
+      "step": 670
+    },
+    {
+      "epoch": 3.3228476821192054,
+      "grad_norm": 0.007492613047361374,
+      "learning_rate": 2.372233198520096e-06,
+      "loss": 0.0001,
+      "step": 671
+    },
+    {
+      "epoch": 3.327814569536424,
+      "grad_norm": 0.007846361957490444,
+      "learning_rate": 2.3651417542077997e-06,
+      "loss": 0.0001,
+      "step": 672
+    },
+    {
+      "epoch": 3.3327814569536423,
+      "grad_norm": 0.01878543011844158,
+      "learning_rate": 2.35805139799023e-06,
+      "loss": 0.0001,
+      "step": 673
+    },
+    {
+      "epoch": 3.337748344370861,
+      "grad_norm": 0.005067222751677036,
+      "learning_rate": 2.3509621870754505e-06,
+      "loss": 0.0001,
+      "step": 674
+    },
+    {
+      "epoch": 3.3427152317880795,
+      "grad_norm": 0.009619520045816898,
+      "learning_rate": 2.3438741786622864e-06,
+      "loss": 0.0001,
+      "step": 675
+    },
+    {
+      "epoch": 3.347682119205298,
+      "grad_norm": 0.005869759246706963,
+      "learning_rate": 2.3367874299398587e-06,
+      "loss": 0.0001,
+      "step": 676
+    },
+    {
+      "epoch": 3.3526490066225167,
+      "grad_norm": 0.045028045773506165,
+      "learning_rate": 2.3297019980871243e-06,
+      "loss": 0.0002,
+      "step": 677
+    },
+    {
+      "epoch": 3.357615894039735,
+      "grad_norm": 0.010993806645274162,
+      "learning_rate": 2.3226179402724175e-06,
+      "loss": 0.0001,
+      "step": 678
+    },
+    {
+      "epoch": 3.3625827814569536,
+      "grad_norm": 0.023547396063804626,
+      "learning_rate": 2.315535313652982e-06,
+      "loss": 0.0001,
+      "step": 679
+    },
+    {
+      "epoch": 3.3675496688741724,
+      "grad_norm": 0.02556329220533371,
+      "learning_rate": 2.3084541753745176e-06,
+      "loss": 0.0001,
+      "step": 680
+    },
+    {
+      "epoch": 3.372516556291391,
+      "grad_norm": 0.013620496727526188,
+      "learning_rate": 2.3013745825707143e-06,
+      "loss": 0.0001,
+      "step": 681
+    },
+    {
+      "epoch": 3.377483443708609,
+      "grad_norm": 0.03706781566143036,
+      "learning_rate": 2.294296592362793e-06,
+      "loss": 0.0001,
+      "step": 682
+    },
+    {
+      "epoch": 3.3824503311258276,
+      "grad_norm": 0.024646388366818428,
+      "learning_rate": 2.287220261859044e-06,
+      "loss": 0.0001,
+      "step": 683
+    },
+    {
+      "epoch": 3.3874172185430464,
+      "grad_norm": 0.006639148574322462,
+      "learning_rate": 2.280145648154365e-06,
+      "loss": 0.0001,
+      "step": 684
+    },
+    {
+      "epoch": 3.392384105960265,
+      "grad_norm": 0.02329806238412857,
+      "learning_rate": 2.2730728083298032e-06,
+      "loss": 0.0001,
+      "step": 685
+    },
+    {
+      "epoch": 3.3973509933774833,
+      "grad_norm": 0.08801553398370743,
+      "learning_rate": 2.266001799452094e-06,
+      "loss": 0.0002,
+      "step": 686
+    },
+    {
+      "epoch": 3.402317880794702,
+      "grad_norm": 0.00732615077868104,
+      "learning_rate": 2.258932678573199e-06,
+      "loss": 0.0001,
+      "step": 687
+    },
+    {
+      "epoch": 3.4072847682119205,
+      "grad_norm": 0.0041299196891486645,
+      "learning_rate": 2.2518655027298468e-06,
+      "loss": 0.0,
+      "step": 688
+    },
+    {
+      "epoch": 3.412251655629139,
+      "grad_norm": 0.17387643456459045,
+      "learning_rate": 2.2448003289430712e-06,
+      "loss": 0.0003,
+      "step": 689
+    },
+    {
+      "epoch": 3.4172185430463577,
+      "grad_norm": 1.4390313625335693,
+      "learning_rate": 2.237737214217755e-06,
+      "loss": 0.0036,
+      "step": 690
+    },
+    {
+      "epoch": 3.422185430463576,
+      "grad_norm": 0.004580878186970949,
+      "learning_rate": 2.2306762155421666e-06,
+      "loss": 0.0001,
+      "step": 691
+    },
+    {
+      "epoch": 3.4271523178807946,
+      "grad_norm": 0.0038564889691770077,
+      "learning_rate": 2.2236173898875004e-06,
+      "loss": 0.0,
+      "step": 692
+    },
+    {
+      "epoch": 3.4321192052980134,
+      "grad_norm": 0.011446455493569374,
+      "learning_rate": 2.2165607942074193e-06,
+      "loss": 0.0001,
+      "step": 693
+    },
+    {
+      "epoch": 3.437086092715232,
+      "grad_norm": 0.006604525726288557,
+      "learning_rate": 2.209506485437593e-06,
+      "loss": 0.0001,
+      "step": 694
+    },
+    {
+      "epoch": 3.44205298013245,
+      "grad_norm": 0.006285684648901224,
+      "learning_rate": 2.2024545204952384e-06,
+      "loss": 0.0,
+      "step": 695
+    },
+    {
+      "epoch": 3.447019867549669,
+      "grad_norm": 0.005863998085260391,
+      "learning_rate": 2.195404956278663e-06,
+      "loss": 0.0001,
+      "step": 696
+    },
+    {
+      "epoch": 3.4519867549668874,
+      "grad_norm": 0.009609216824173927,
+      "learning_rate": 2.1883578496668046e-06,
+      "loss": 0.0001,
+      "step": 697
+    },
+    {
+      "epoch": 3.456953642384106,
+      "grad_norm": 0.005348480772227049,
+      "learning_rate": 2.1813132575187703e-06,
+      "loss": 0.0001,
+      "step": 698
+    },
+    {
+      "epoch": 3.4619205298013247,
+      "grad_norm": 0.1555260270833969,
+      "learning_rate": 2.1742712366733805e-06,
+      "loss": 0.0007,
+      "step": 699
+    },
+    {
+      "epoch": 3.466887417218543,
+      "grad_norm": 0.002812250517308712,
+      "learning_rate": 2.1672318439487084e-06,
+      "loss": 0.0,
+      "step": 700
+    },
+    {
+      "epoch": 3.4718543046357615,
+      "grad_norm": 0.005335136782377958,
+      "learning_rate": 2.1601951361416224e-06,
+      "loss": 0.0001,
+      "step": 701
+    },
+    {
+      "epoch": 3.47682119205298,
+      "grad_norm": 0.0035038047935813665,
+      "learning_rate": 2.15316117002733e-06,
+      "loss": 0.0,
+      "step": 702
+    },
+    {
+      "epoch": 3.4817880794701987,
+      "grad_norm": 0.003302652155980468,
+      "learning_rate": 2.146130002358914e-06,
+      "loss": 0.0,
+      "step": 703
+    },
+    {
+      "epoch": 3.486754966887417,
+      "grad_norm": 0.013222907669842243,
+      "learning_rate": 2.139101689866881e-06,
+      "loss": 0.0,
+      "step": 704
+    },
+    {
+      "epoch": 3.491721854304636,
+      "grad_norm": 0.002825352596119046,
+      "learning_rate": 2.132076289258701e-06,
+      "loss": 0.0,
+      "step": 705
+    },
+    {
+      "epoch": 3.4966887417218544,
+      "grad_norm": 0.005810688249766827,
+      "learning_rate": 2.125053857218346e-06,
+      "loss": 0.0001,
+      "step": 706
+    },
+    {
+      "epoch": 3.501655629139073,
+      "grad_norm": 0.005243207328021526,
+      "learning_rate": 2.1180344504058394e-06,
+      "loss": 0.0,
+      "step": 707
+    },
+    {
+      "epoch": 3.506622516556291,
+      "grad_norm": 0.005770714953541756,
+      "learning_rate": 2.1110181254567958e-06,
+      "loss": 0.0001,
+      "step": 708
+    },
+    {
+      "epoch": 3.51158940397351,
+      "grad_norm": 0.011787684634327888,
+      "learning_rate": 2.1040049389819628e-06,
+      "loss": 0.0001,
+      "step": 709
+    },
+    {
+      "epoch": 3.5165562913907285,
+      "grad_norm": 0.007399479392915964,
+      "learning_rate": 2.0969949475667662e-06,
+      "loss": 0.0001,
+      "step": 710
+    },
+    {
+      "epoch": 3.5215231788079473,
+      "grad_norm": 0.004310545511543751,
+      "learning_rate": 2.089988207770852e-06,
+      "loss": 0.0001,
+      "step": 711
+    },
+    {
+      "epoch": 3.5264900662251657,
+      "grad_norm": 0.008017109706997871,
+      "learning_rate": 2.082984776127632e-06,
+      "loss": 0.0001,
+      "step": 712
+    },
+    {
+      "epoch": 3.531456953642384,
+      "grad_norm": 0.006948905996978283,
+      "learning_rate": 2.075984709143824e-06,
+      "loss": 0.0001,
+      "step": 713
+    },
+    {
+      "epoch": 3.5364238410596025,
+      "grad_norm": 0.011265809647738934,
+      "learning_rate": 2.068988063299001e-06,
+      "loss": 0.0001,
+      "step": 714
+    },
+    {
+      "epoch": 3.5413907284768213,
+      "grad_norm": 0.022066382691264153,
+      "learning_rate": 2.0619948950451317e-06,
+      "loss": 0.0001,
+      "step": 715
+    },
+    {
+      "epoch": 3.5463576158940397,
+      "grad_norm": 0.0048875873908400536,
+      "learning_rate": 2.055005260806125e-06,
+      "loss": 0.0,
+      "step": 716
+    },
+    {
+      "epoch": 3.551324503311258,
+      "grad_norm": 0.11520480364561081,
+      "learning_rate": 2.0480192169773764e-06,
+      "loss": 0.0015,
+      "step": 717
+    },
+    {
+      "epoch": 3.556291390728477,
+      "grad_norm": 0.04643610492348671,
+      "learning_rate": 2.0410368199253143e-06,
+      "loss": 0.0001,
+      "step": 718
+    },
+    {
+      "epoch": 3.5612582781456954,
+      "grad_norm": 0.005898310337215662,
+      "learning_rate": 2.034058125986941e-06,
+      "loss": 0.0001,
+      "step": 719
+    },
+    {
+      "epoch": 3.566225165562914,
+      "grad_norm": 0.009210104122757912,
+      "learning_rate": 2.0270831914693813e-06,
+      "loss": 0.0001,
+      "step": 720
+    },
+    {
+      "epoch": 3.571192052980132,
+      "grad_norm": 0.009020411409437656,
+      "learning_rate": 2.020112072649428e-06,
+      "loss": 0.0001,
+      "step": 721
+    },
+    {
+      "epoch": 3.576158940397351,
+      "grad_norm": 0.20272137224674225,
+      "learning_rate": 2.013144825773085e-06,
+      "loss": 0.0008,
+      "step": 722
+    },
+    {
+      "epoch": 3.5811258278145695,
+      "grad_norm": 0.009092798456549644,
+      "learning_rate": 2.0061815070551184e-06,
+      "loss": 0.0001,
+      "step": 723
+    },
+    {
+      "epoch": 3.5860927152317883,
+      "grad_norm": 0.002982528181746602,
+      "learning_rate": 1.999222172678599e-06,
+      "loss": 0.0,
+      "step": 724
+    },
+    {
+      "epoch": 3.5910596026490067,
+      "grad_norm": 0.013882700353860855,
+      "learning_rate": 1.99226687879445e-06,
+      "loss": 0.0001,
+      "step": 725
+    },
+    {
+      "epoch": 3.596026490066225,
+      "grad_norm": 0.008199700154364109,
+      "learning_rate": 1.9853156815209955e-06,
+      "loss": 0.0001,
+      "step": 726
+    },
+    {
+      "epoch": 3.6009933774834435,
+      "grad_norm": 0.009564278647303581,
+      "learning_rate": 1.9783686369435034e-06,
+      "loss": 0.0001,
+      "step": 727
+    },
+    {
+      "epoch": 3.6059602649006623,
+      "grad_norm": 0.004816321190446615,
+      "learning_rate": 1.9714258011137385e-06,
+      "loss": 0.0001,
+      "step": 728
+    },
+    {
+      "epoch": 3.6109271523178808,
+      "grad_norm": 0.08846727013587952,
+      "learning_rate": 1.964487230049507e-06,
+      "loss": 0.0011,
+      "step": 729
+    },
+    {
+      "epoch": 3.6158940397350996,
+      "grad_norm": 0.010213276371359825,
+      "learning_rate": 1.957552979734205e-06,
+      "loss": 0.0001,
+      "step": 730
+    },
+    {
+      "epoch": 3.620860927152318,
+      "grad_norm": 0.11782242357730865,
+      "learning_rate": 1.9506231061163673e-06,
+      "loss": 0.0003,
+      "step": 731
+    },
+    {
+      "epoch": 3.6258278145695364,
+      "grad_norm": 0.010244274511933327,
+      "learning_rate": 1.9436976651092143e-06,
+      "loss": 0.0,
+      "step": 732
+    },
+    {
+      "epoch": 3.630794701986755,
+      "grad_norm": 0.03418213129043579,
+      "learning_rate": 1.9367767125902033e-06,
+      "loss": 0.0001,
+      "step": 733
+    },
+    {
+      "epoch": 3.6357615894039736,
+      "grad_norm": 0.007177441380918026,
+      "learning_rate": 1.9298603044005775e-06,
+      "loss": 0.0001,
+      "step": 734
+    },
+    {
+      "epoch": 3.640728476821192,
+      "grad_norm": 0.006836403626948595,
+      "learning_rate": 1.9229484963449114e-06,
+      "loss": 0.0001,
+      "step": 735
+    },
+    {
+      "epoch": 3.6456953642384105,
+      "grad_norm": 0.07157158851623535,
+      "learning_rate": 1.9160413441906667e-06,
+      "loss": 0.0002,
+      "step": 736
+    },
+    {
+      "epoch": 3.6506622516556293,
+      "grad_norm": 0.012204978615045547,
+      "learning_rate": 1.9091389036677384e-06,
+      "loss": 0.0001,
+      "step": 737
+    },
+    {
+      "epoch": 3.6556291390728477,
+      "grad_norm": 0.011574111878871918,
+      "learning_rate": 1.9022412304680044e-06,
+      "loss": 0.0001,
+      "step": 738
+    },
+    {
+      "epoch": 3.660596026490066,
+      "grad_norm": 0.003680022433400154,
+      "learning_rate": 1.89534838024488e-06,
+      "loss": 0.0001,
+      "step": 739
+    },
+    {
+      "epoch": 3.6655629139072845,
+      "grad_norm": 0.011809752322733402,
+      "learning_rate": 1.8884604086128656e-06,
+      "loss": 0.0001,
+      "step": 740
+    },
+    {
+      "epoch": 3.6705298013245033,
+      "grad_norm": 0.01406245306134224,
+      "learning_rate": 1.8815773711470988e-06,
+      "loss": 0.0001,
+      "step": 741
+    },
+    {
+      "epoch": 3.6754966887417218,
+      "grad_norm": 0.005696302279829979,
+      "learning_rate": 1.874699323382908e-06,
+      "loss": 0.0,
+      "step": 742
+    },
+    {
+      "epoch": 3.6804635761589406,
+      "grad_norm": 0.007031074725091457,
+      "learning_rate": 1.8678263208153591e-06,
+      "loss": 0.0001,
+      "step": 743
+    },
+    {
+      "epoch": 3.685430463576159,
+      "grad_norm": 0.054343827068805695,
+      "learning_rate": 1.8609584188988135e-06,
+      "loss": 0.0002,
+      "step": 744
+    },
+    {
+      "epoch": 3.6903973509933774,
+      "grad_norm": 0.008045045658946037,
+      "learning_rate": 1.8540956730464789e-06,
+      "loss": 0.0001,
+      "step": 745
+    },
+    {
+      "epoch": 3.695364238410596,
+      "grad_norm": 0.22399218380451202,
+      "learning_rate": 1.8472381386299597e-06,
+      "loss": 0.0008,
+      "step": 746
+    },
+    {
+      "epoch": 3.7003311258278146,
+      "grad_norm": 0.013838261365890503,
+      "learning_rate": 1.8403858709788154e-06,
+      "loss": 0.0001,
+      "step": 747
+    },
+    {
+      "epoch": 3.705298013245033,
+      "grad_norm": 0.008240136317908764,
+      "learning_rate": 1.8335389253801056e-06,
+      "loss": 0.0001,
+      "step": 748
+    },
+    {
+      "epoch": 3.710264900662252,
+      "grad_norm": 0.0033084205351769924,
+      "learning_rate": 1.8266973570779542e-06,
+      "loss": 0.0,
+      "step": 749
+    },
+    {
+      "epoch": 3.7152317880794703,
+      "grad_norm": 0.0045149358920753,
+      "learning_rate": 1.8198612212730965e-06,
+      "loss": 0.0,
+      "step": 750
+    },
+    {
+      "epoch": 3.7201986754966887,
+      "grad_norm": 0.0026892581954598427,
+      "learning_rate": 1.8130305731224368e-06,
+      "loss": 0.0,
+      "step": 751
+    },
+    {
+      "epoch": 3.725165562913907,
+      "grad_norm": 0.005407802294939756,
+      "learning_rate": 1.8062054677386023e-06,
+      "loss": 0.0001,
+      "step": 752
+    },
+    {
+      "epoch": 3.730132450331126,
+      "grad_norm": 0.004870439879596233,
+      "learning_rate": 1.7993859601894995e-06,
+      "loss": 0.0001,
+      "step": 753
+    },
+    {
+      "epoch": 3.7350993377483444,
+      "grad_norm": 0.00942834373563528,
+      "learning_rate": 1.7925721054978677e-06,
+      "loss": 0.0001,
+      "step": 754
+    },
+    {
+      "epoch": 3.7400662251655628,
+      "grad_norm": 0.009291539900004864,
+      "learning_rate": 1.7857639586408366e-06,
+      "loss": 0.0,
+      "step": 755
+    },
+    {
+      "epoch": 3.7450331125827816,
+      "grad_norm": 0.006173161324113607,
+      "learning_rate": 1.7789615745494842e-06,
+      "loss": 0.0001,
+      "step": 756
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.02726677805185318,
+      "learning_rate": 1.7721650081083909e-06,
+      "loss": 0.0001,
+      "step": 757
+    },
+    {
+      "epoch": 3.7549668874172184,
+      "grad_norm": 0.011498328298330307,
+      "learning_rate": 1.7653743141551983e-06,
+      "loss": 0.0001,
+      "step": 758
+    },
+    {
+      "epoch": 3.7599337748344372,
+      "grad_norm": 0.017378000542521477,
+      "learning_rate": 1.758589547480165e-06,
+      "loss": 0.0001,
+      "step": 759
+    },
+    {
+      "epoch": 3.7649006622516556,
+      "grad_norm": 0.11044502258300781,
+      "learning_rate": 1.7518107628257275e-06,
+      "loss": 0.0004,
+      "step": 760
+    },
+    {
+      "epoch": 3.769867549668874,
+      "grad_norm": 0.007891656830906868,
+      "learning_rate": 1.7450380148860558e-06,
+      "loss": 0.0,
+      "step": 761
+    },
+    {
+      "epoch": 3.774834437086093,
+      "grad_norm": 0.004707221873104572,
+      "learning_rate": 1.7382713583066126e-06,
+      "loss": 0.0,
+      "step": 762
+    },
+    {
+      "epoch": 3.7798013245033113,
+      "grad_norm": 0.003836105577647686,
+      "learning_rate": 1.7315108476837141e-06,
+      "loss": 0.0001,
+      "step": 763
+    },
+    {
+      "epoch": 3.7847682119205297,
+      "grad_norm": 0.011506680399179459,
+      "learning_rate": 1.7247565375640883e-06,
+      "loss": 0.0001,
+      "step": 764
+    },
+    {
+      "epoch": 3.789735099337748,
+      "grad_norm": 0.006728639826178551,
+      "learning_rate": 1.7180084824444326e-06,
+      "loss": 0.0001,
+      "step": 765
+    },
+    {
+      "epoch": 3.794701986754967,
+      "grad_norm": 0.018557390198111534,
+      "learning_rate": 1.7112667367709784e-06,
+      "loss": 0.0001,
+      "step": 766
+    },
+    {
+      "epoch": 3.7996688741721854,
+      "grad_norm": 0.010772901587188244,
+      "learning_rate": 1.704531354939049e-06,
+      "loss": 0.0001,
+      "step": 767
+    },
+    {
+      "epoch": 3.804635761589404,
+      "grad_norm": 0.10443193465471268,
+      "learning_rate": 1.6978023912926227e-06,
+      "loss": 0.0004,
+      "step": 768
+    },
+    {
+      "epoch": 3.8096026490066226,
+      "grad_norm": 0.004351992625743151,
+      "learning_rate": 1.6910799001238925e-06,
+      "loss": 0.0,
+      "step": 769
+    },
+    {
+      "epoch": 3.814569536423841,
+      "grad_norm": 0.01436906773597002,
+      "learning_rate": 1.684363935672827e-06,
+      "loss": 0.0001,
+      "step": 770
+    },
+    {
+      "epoch": 3.8195364238410594,
+      "grad_norm": 0.0039618974551558495,
+      "learning_rate": 1.6776545521267358e-06,
+      "loss": 0.0,
+      "step": 771
+    },
+    {
+      "epoch": 3.8245033112582782,
+      "grad_norm": 0.00797906145453453,
+      "learning_rate": 1.6709518036198307e-06,
+      "loss": 0.0001,
+      "step": 772
+    },
+    {
+      "epoch": 3.8294701986754967,
+      "grad_norm": 0.01255651842802763,
+      "learning_rate": 1.66425574423279e-06,
+      "loss": 0.0001,
+      "step": 773
+    },
+    {
+      "epoch": 3.8344370860927155,
+      "grad_norm": 0.004594859667122364,
+      "learning_rate": 1.6575664279923214e-06,
+      "loss": 0.0,
+      "step": 774
+    },
+    {
+      "epoch": 3.839403973509934,
+      "grad_norm": 0.004516359884291887,
+      "learning_rate": 1.6508839088707232e-06,
+      "loss": 0.0001,
+      "step": 775
+    },
+    {
+      "epoch": 3.8443708609271523,
+      "grad_norm": 0.004302950110286474,
+      "learning_rate": 1.644208240785454e-06,
+      "loss": 0.0001,
+      "step": 776
+    },
+    {
+      "epoch": 3.8493377483443707,
+      "grad_norm": 0.006869391072541475,
+      "learning_rate": 1.6375394775986953e-06,
+      "loss": 0.0001,
+      "step": 777
+    },
+    {
+      "epoch": 3.8543046357615895,
+      "grad_norm": 0.01430842000991106,
+      "learning_rate": 1.6308776731169157e-06,
+      "loss": 0.0001,
+      "step": 778
+    },
+    {
+      "epoch": 3.859271523178808,
+      "grad_norm": 0.014300678856670856,
+      "learning_rate": 1.6242228810904393e-06,
+      "loss": 0.0001,
+      "step": 779
+    },
+    {
+      "epoch": 3.8642384105960264,
+      "grad_norm": 0.0071014598943293095,
+      "learning_rate": 1.6175751552130103e-06,
+      "loss": 0.0001,
+      "step": 780
+    },
+    {
+      "epoch": 3.869205298013245,
+      "grad_norm": 0.004982541315257549,
+      "learning_rate": 1.6109345491213588e-06,
+      "loss": 0.0,
+      "step": 781
+    },
+    {
+      "epoch": 3.8741721854304636,
+      "grad_norm": 0.014259067364037037,
+      "learning_rate": 1.6043011163947708e-06,
+      "loss": 0.0001,
+      "step": 782
+    },
+    {
+      "epoch": 3.879139072847682,
+      "grad_norm": 0.0021392295602709055,
+      "learning_rate": 1.5976749105546545e-06,
+      "loss": 0.0,
+      "step": 783
+    },
+    {
+      "epoch": 3.8841059602649004,
+      "grad_norm": 0.005686462391167879,
+      "learning_rate": 1.5910559850641077e-06,
+      "loss": 0.0,
+      "step": 784
+    },
+    {
+      "epoch": 3.8890728476821192,
+      "grad_norm": 0.007352852262556553,
+      "learning_rate": 1.5844443933274888e-06,
+      "loss": 0.0,
+      "step": 785
+    },
+    {
+      "epoch": 3.8940397350993377,
+      "grad_norm": 0.009414524771273136,
+      "learning_rate": 1.5778401886899808e-06,
+      "loss": 0.0,
+      "step": 786
+    },
+    {
+      "epoch": 3.8990066225165565,
+      "grad_norm": 0.009815333411097527,
+      "learning_rate": 1.5712434244371677e-06,
+      "loss": 0.0001,
+      "step": 787
+    },
+    {
+      "epoch": 3.903973509933775,
+      "grad_norm": 0.20881228148937225,
+      "learning_rate": 1.5646541537946006e-06,
+      "loss": 0.0001,
+      "step": 788
+    },
+    {
+      "epoch": 3.9089403973509933,
+      "grad_norm": 0.006103521678596735,
+      "learning_rate": 1.5580724299273679e-06,
+      "loss": 0.0001,
+      "step": 789
+    },
+    {
+      "epoch": 3.9139072847682117,
+      "grad_norm": 0.006266051437705755,
+      "learning_rate": 1.5514983059396693e-06,
+      "loss": 0.0,
+      "step": 790
+    },
+    {
+      "epoch": 3.9188741721854305,
+      "grad_norm": 0.00585981598123908,
+      "learning_rate": 1.5449318348743818e-06,
+      "loss": 0.0,
+      "step": 791
+    },
+    {
+      "epoch": 3.923841059602649,
+      "grad_norm": 0.00204283045604825,
+      "learning_rate": 1.5383730697126387e-06,
+      "loss": 0.0,
+      "step": 792
+    },
+    {
+      "epoch": 3.928807947019868,
+      "grad_norm": 0.005272861570119858,
+      "learning_rate": 1.5318220633733977e-06,
+      "loss": 0.0001,
+      "step": 793
+    },
+    {
+      "epoch": 3.933774834437086,
+      "grad_norm": 0.004095260985195637,
+      "learning_rate": 1.5252788687130146e-06,
+      "loss": 0.0,
+      "step": 794
+    },
+    {
+      "epoch": 3.9387417218543046,
+      "grad_norm": 0.006596855353564024,
+      "learning_rate": 1.5187435385248173e-06,
+      "loss": 0.0001,
+      "step": 795
+    },
+    {
+      "epoch": 3.943708609271523,
+      "grad_norm": 0.007137851323932409,
+      "learning_rate": 1.5122161255386815e-06,
+      "loss": 0.0001,
+      "step": 796
+    },
+    {
+      "epoch": 3.948675496688742,
+      "grad_norm": 0.0031777587719261646,
+      "learning_rate": 1.505696682420599e-06,
+      "loss": 0.0,
+      "step": 797
+    },
+    {
+      "epoch": 3.9536423841059603,
+      "grad_norm": 0.007171051576733589,
+      "learning_rate": 1.499185261772262e-06,
+      "loss": 0.0001,
+      "step": 798
+    },
+    {
+      "epoch": 3.9586092715231787,
+      "grad_norm": 0.004072366748005152,
+      "learning_rate": 1.4926819161306309e-06,
+      "loss": 0.0,
+      "step": 799
+    },
+    {
+      "epoch": 3.9635761589403975,
+      "grad_norm": 0.003464906942099333,
+      "learning_rate": 1.4861866979675155e-06,
+      "loss": 0.0,
+      "step": 800
+    },
+    {
+      "epoch": 3.968543046357616,
+      "grad_norm": 0.0020631717052310705,
+      "learning_rate": 1.4796996596891489e-06,
+      "loss": 0.0,
+      "step": 801
+    },
+    {
+      "epoch": 3.9735099337748343,
+      "grad_norm": 0.005078312940895557,
+      "learning_rate": 1.473220853635764e-06,
+      "loss": 0.0001,
+      "step": 802
+    },
+    {
+      "epoch": 3.9784768211920527,
+      "grad_norm": 0.004130552057176828,
+      "learning_rate": 1.466750332081174e-06,
+      "loss": 0.0,
+      "step": 803
+    },
+    {
+      "epoch": 3.9834437086092715,
+      "grad_norm": 0.004617821425199509,
+      "learning_rate": 1.46028814723235e-06,
+      "loss": 0.0,
+      "step": 804
+    },
+    {
+      "epoch": 3.98841059602649,
+      "grad_norm": 0.004183633252978325,
+      "learning_rate": 1.453834351228996e-06,
+      "loss": 0.0001,
+      "step": 805
+    },
+    {
+      "epoch": 3.993377483443709,
+      "grad_norm": 0.012064768001437187,
+      "learning_rate": 1.4473889961431342e-06,
+      "loss": 0.0001,
+      "step": 806
+    },
+    {
+      "epoch": 3.998344370860927,
+      "grad_norm": 0.008102312684059143,
+      "learning_rate": 1.4409521339786809e-06,
+      "loss": 0.0001,
+      "step": 807
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.008102312684059143,
+      "learning_rate": 1.4345238166710256e-06,
+      "loss": 0.0001,
+      "step": 808
+    },
+    {
+      "epoch": 4.004966887417218,
+      "grad_norm": 0.008242284879088402,
+      "learning_rate": 1.4281040960866178e-06,
+      "loss": 0.0,
+      "step": 809
+    },
+    {
+      "epoch": 4.009933774834437,
+      "grad_norm": 0.002578236162662506,
+      "learning_rate": 1.421693024022542e-06,
+      "loss": 0.0,
+      "step": 810
+    },
+    {
+      "epoch": 4.014900662251655,
+      "grad_norm": 0.0038229613564908504,
+      "learning_rate": 1.415290652206105e-06,
+      "loss": 0.0001,
+      "step": 811
+    },
+    {
+      "epoch": 4.0198675496688745,
+      "grad_norm": 0.00330353737808764,
+      "learning_rate": 1.4088970322944145e-06,
+      "loss": 0.0001,
+      "step": 812
+    },
+    {
+      "epoch": 4.024834437086093,
+      "grad_norm": 0.0158391073346138,
+      "learning_rate": 1.402512215873965e-06,
+      "loss": 0.0,
+      "step": 813
+    },
+    {
+      "epoch": 4.029801324503311,
+      "grad_norm": 0.005667423363775015,
+      "learning_rate": 1.3961362544602215e-06,
+      "loss": 0.0001,
+      "step": 814
+    },
+    {
+      "epoch": 4.03476821192053,
+      "grad_norm": 0.009708857163786888,
+      "learning_rate": 1.3897691994971988e-06,
+      "loss": 0.0001,
+      "step": 815
+    },
+    {
+      "epoch": 4.039735099337748,
+      "grad_norm": 0.003954901825636625,
+      "learning_rate": 1.3834111023570557e-06,
+      "loss": 0.0,
+      "step": 816
+    },
+    {
+      "epoch": 4.0447019867549665,
+      "grad_norm": 0.010039578191936016,
+      "learning_rate": 1.3770620143396723e-06,
+      "loss": 0.0001,
+      "step": 817
+    },
+    {
+      "epoch": 4.049668874172186,
+      "grad_norm": 0.0023660482838749886,
+      "learning_rate": 1.370721986672241e-06,
+      "loss": 0.0001,
+      "step": 818
+    },
+    {
+      "epoch": 4.054635761589404,
+      "grad_norm": 0.005724453367292881,
+      "learning_rate": 1.3643910705088504e-06,
+      "loss": 0.0,
+      "step": 819
+    },
+    {
+      "epoch": 4.059602649006623,
+      "grad_norm": 0.0023897697683423758,
+      "learning_rate": 1.358069316930073e-06,
+      "loss": 0.0,
+      "step": 820
+    },
+    {
+      "epoch": 4.064569536423841,
+      "grad_norm": 0.004279536195099354,
+      "learning_rate": 1.351756776942555e-06,
+      "loss": 0.0,
+      "step": 821
+    },
+    {
+      "epoch": 4.069536423841059,
+      "grad_norm": 0.004504794254899025,
+      "learning_rate": 1.3454535014786025e-06,
+      "loss": 0.0,
+      "step": 822
+    },
+    {
+      "epoch": 4.074503311258278,
+      "grad_norm": 0.003042777767404914,
+      "learning_rate": 1.339159541395772e-06,
+      "loss": 0.0001,
+      "step": 823
+    },
+    {
+      "epoch": 4.079470198675497,
+      "grad_norm": 0.010188202373683453,
+      "learning_rate": 1.332874947476458e-06,
+      "loss": 0.0001,
+      "step": 824
+    },
+    {
+      "epoch": 4.0844370860927155,
+      "grad_norm": 0.0021820038091391325,
+      "learning_rate": 1.3265997704274858e-06,
+      "loss": 0.0,
+      "step": 825
+    },
+    {
+      "epoch": 4.089403973509934,
+      "grad_norm": 0.003920790273696184,
+      "learning_rate": 1.320334060879702e-06,
+      "loss": 0.0,
+      "step": 826
+    },
+    {
+      "epoch": 4.094370860927152,
+      "grad_norm": 0.0024695182219147682,
+      "learning_rate": 1.3140778693875639e-06,
+      "loss": 0.0,
+      "step": 827
+    },
+    {
+      "epoch": 4.099337748344371,
+      "grad_norm": 0.003500124206766486,
+      "learning_rate": 1.3078312464287355e-06,
+      "loss": 0.0,
+      "step": 828
+    },
+    {
+      "epoch": 4.104304635761589,
+      "grad_norm": 0.005404484458267689,
+      "learning_rate": 1.301594242403673e-06,
+      "loss": 0.0001,
+      "step": 829
+    },
+    {
+      "epoch": 4.109271523178808,
+      "grad_norm": 0.0021421818528324366,
+      "learning_rate": 1.2953669076352277e-06,
+      "loss": 0.0,
+      "step": 830
+    },
+    {
+      "epoch": 4.114238410596027,
+      "grad_norm": 0.0024207893293350935,
+      "learning_rate": 1.2891492923682325e-06,
+      "loss": 0.0,
+      "step": 831
+    },
+    {
+      "epoch": 4.119205298013245,
+      "grad_norm": 0.0036852953489869833,
+      "learning_rate": 1.2829414467691e-06,
+      "loss": 0.0001,
+      "step": 832
+    },
+    {
+      "epoch": 4.124172185430464,
+      "grad_norm": 0.002881386550143361,
+      "learning_rate": 1.2767434209254182e-06,
+      "loss": 0.0,
+      "step": 833
+    },
+    {
+      "epoch": 4.129139072847682,
+      "grad_norm": 0.004867165815085173,
+      "learning_rate": 1.270555264845541e-06,
+      "loss": 0.0,
+      "step": 834
+    },
+    {
+      "epoch": 4.1341059602649,
+      "grad_norm": 0.0034252949990332127,
+      "learning_rate": 1.264377028458193e-06,
+      "loss": 0.0,
+      "step": 835
+    },
+    {
+      "epoch": 4.139072847682119,
+      "grad_norm": 0.04827704280614853,
+      "learning_rate": 1.258208761612061e-06,
+      "loss": 0.0001,
+      "step": 836
+    },
+    {
+      "epoch": 4.144039735099338,
+      "grad_norm": 0.0032711862586438656,
+      "learning_rate": 1.252050514075392e-06,
+      "loss": 0.0,
+      "step": 837
+    },
+    {
+      "epoch": 4.1490066225165565,
+      "grad_norm": 0.0023121500853449106,
+      "learning_rate": 1.2459023355355947e-06,
+      "loss": 0.0,
+      "step": 838
+    },
+    {
+      "epoch": 4.153973509933775,
+      "grad_norm": 0.0022882502526044846,
+      "learning_rate": 1.2397642755988371e-06,
+      "loss": 0.0,
+      "step": 839
+    },
+    {
+      "epoch": 4.158940397350993,
+      "grad_norm": 0.003316567512229085,
+      "learning_rate": 1.2336363837896423e-06,
+      "loss": 0.0001,
+      "step": 840
+    },
+    {
+      "epoch": 4.163907284768212,
+      "grad_norm": 0.004417945630848408,
+      "learning_rate": 1.2275187095504966e-06,
+      "loss": 0.0,
+      "step": 841
+    },
+    {
+      "epoch": 4.16887417218543,
+      "grad_norm": 0.003385403659194708,
+      "learning_rate": 1.2214113022414448e-06,
+      "loss": 0.0,
+      "step": 842
+    },
+    {
+      "epoch": 4.173841059602649,
+      "grad_norm": 0.0020309528335928917,
+      "learning_rate": 1.2153142111396938e-06,
+      "loss": 0.0,
+      "step": 843
+    },
+    {
+      "epoch": 4.178807947019868,
+      "grad_norm": 0.003971997648477554,
+      "learning_rate": 1.2092274854392159e-06,
+      "loss": 0.0001,
+      "step": 844
+    },
+    {
+      "epoch": 4.183774834437086,
+      "grad_norm": 0.006885501556098461,
+      "learning_rate": 1.2031511742503479e-06,
+      "loss": 0.0001,
+      "step": 845
+    },
+    {
+      "epoch": 4.188741721854305,
+      "grad_norm": 0.003782738232985139,
+      "learning_rate": 1.1970853265994008e-06,
+      "loss": 0.0,
+      "step": 846
+    },
+    {
+      "epoch": 4.193708609271523,
+      "grad_norm": 0.0034930650144815445,
+      "learning_rate": 1.1910299914282602e-06,
+      "loss": 0.0,
+      "step": 847
+    },
+    {
+      "epoch": 4.198675496688741,
+      "grad_norm": 0.0049952068366110325,
+      "learning_rate": 1.184985217593993e-06,
+      "loss": 0.0,
+      "step": 848
+    },
+    {
+      "epoch": 4.203642384105961,
+      "grad_norm": 0.002861351240426302,
+      "learning_rate": 1.1789510538684524e-06,
+      "loss": 0.0,
+      "step": 849
+    },
+    {
+      "epoch": 4.208609271523179,
+      "grad_norm": 0.0035430407151579857,
+      "learning_rate": 1.1729275489378827e-06,
+      "loss": 0.0,
+      "step": 850
+    },
+    {
+      "epoch": 4.2135761589403975,
+      "grad_norm": 0.00294266315177083,
+      "learning_rate": 1.166914751402531e-06,
+      "loss": 0.0,
+      "step": 851
+    },
+    {
+      "epoch": 4.218543046357616,
+      "grad_norm": 0.001805895590223372,
+      "learning_rate": 1.1609127097762521e-06,
+      "loss": 0.0,
+      "step": 852
+    },
+    {
+      "epoch": 4.223509933774834,
+      "grad_norm": 0.003681682515889406,
+      "learning_rate": 1.1549214724861168e-06,
+      "loss": 0.0,
+      "step": 853
+    },
+    {
+      "epoch": 4.228476821192053,
+      "grad_norm": 0.004073623102158308,
+      "learning_rate": 1.1489410878720217e-06,
+      "loss": 0.0001,
+      "step": 854
+    },
+    {
+      "epoch": 4.233443708609271,
+      "grad_norm": 0.004671689122915268,
+      "learning_rate": 1.1429716041863009e-06,
+      "loss": 0.0,
+      "step": 855
+    },
+    {
+      "epoch": 4.23841059602649,
+      "grad_norm": 0.002555918414145708,
+      "learning_rate": 1.1370130695933317e-06,
+      "loss": 0.0001,
+      "step": 856
+    },
+    {
+      "epoch": 4.243377483443709,
+      "grad_norm": 0.009303344413638115,
+      "learning_rate": 1.1310655321691527e-06,
+      "loss": 0.0001,
+      "step": 857
+    },
+    {
+      "epoch": 4.248344370860927,
+      "grad_norm": 0.004324618261307478,
+      "learning_rate": 1.1251290399010713e-06,
+      "loss": 0.0,
+      "step": 858
+    },
+    {
+      "epoch": 4.253311258278146,
+      "grad_norm": 0.004484944976866245,
+      "learning_rate": 1.1192036406872784e-06,
+      "loss": 0.0001,
+      "step": 859
+    },
+    {
+      "epoch": 4.258278145695364,
+      "grad_norm": 0.002815540647134185,
+      "learning_rate": 1.1132893823364597e-06,
+      "loss": 0.0001,
+      "step": 860
+    },
+    {
+      "epoch": 4.263245033112582,
+      "grad_norm": 0.0048917559906840324,
+      "learning_rate": 1.1073863125674137e-06,
+      "loss": 0.0001,
+      "step": 861
+    },
+    {
+      "epoch": 4.268211920529802,
+      "grad_norm": 0.007124399300664663,
+      "learning_rate": 1.1014944790086632e-06,
+      "loss": 0.0001,
+      "step": 862
+    },
+    {
+      "epoch": 4.27317880794702,
+      "grad_norm": 0.00369991990737617,
+      "learning_rate": 1.0956139291980727e-06,
+      "loss": 0.0,
+      "step": 863
+    },
+    {
+      "epoch": 4.2781456953642385,
+      "grad_norm": 0.003621634328737855,
+      "learning_rate": 1.0897447105824646e-06,
+      "loss": 0.0001,
+      "step": 864
+    },
+    {
+      "epoch": 4.283112582781457,
+      "grad_norm": 0.0018358862726017833,
+      "learning_rate": 1.0838868705172378e-06,
+      "loss": 0.0,
+      "step": 865
+    },
+    {
+      "epoch": 4.288079470198675,
+      "grad_norm": 0.0026274514384567738,
+      "learning_rate": 1.07804045626598e-06,
+      "loss": 0.0,
+      "step": 866
+    },
+    {
+      "epoch": 4.293046357615894,
+      "grad_norm": 0.00418589124456048,
+      "learning_rate": 1.072205515000095e-06,
+      "loss": 0.0,
+      "step": 867
+    },
+    {
+      "epoch": 4.298013245033113,
+      "grad_norm": 0.007845349609851837,
+      "learning_rate": 1.0663820937984148e-06,
+      "loss": 0.0001,
+      "step": 868
+    },
+    {
+      "epoch": 4.302980132450331,
+      "grad_norm": 0.0014837852213531733,
+      "learning_rate": 1.060570239646824e-06,
+      "loss": 0.0,
+      "step": 869
+    },
+    {
+      "epoch": 4.30794701986755,
+      "grad_norm": 0.0073472606018185616,
+      "learning_rate": 1.0547699994378787e-06,
+      "loss": 0.0,
+      "step": 870
+    },
+    {
+      "epoch": 4.312913907284768,
+      "grad_norm": 0.003900103038176894,
+      "learning_rate": 1.0489814199704292e-06,
+      "loss": 0.0,
+      "step": 871
+    },
+    {
+      "epoch": 4.317880794701987,
+      "grad_norm": 0.001788733177818358,
+      "learning_rate": 1.04320454794924e-06,
+      "loss": 0.0,
+      "step": 872
+    },
+    {
+      "epoch": 4.322847682119205,
+      "grad_norm": 0.002956331241875887,
+      "learning_rate": 1.037439429984617e-06,
+      "loss": 0.0,
+      "step": 873
+    },
+    {
+      "epoch": 4.327814569536423,
+      "grad_norm": 0.003076766850426793,
+      "learning_rate": 1.0316861125920282e-06,
+      "loss": 0.0,
+      "step": 874
+    },
+    {
+      "epoch": 4.332781456953643,
+      "grad_norm": 0.0033401965629309416,
+      "learning_rate": 1.02594464219173e-06,
+      "loss": 0.0,
+      "step": 875
+    },
+    {
+      "epoch": 4.337748344370861,
+      "grad_norm": 0.0028227621223777533,
+      "learning_rate": 1.020215065108393e-06,
+      "loss": 0.0,
+      "step": 876
+    },
+    {
+      "epoch": 4.3427152317880795,
+      "grad_norm": 0.004530815407633781,
+      "learning_rate": 1.0144974275707243e-06,
+      "loss": 0.0,
+      "step": 877
+    },
+    {
+      "epoch": 4.347682119205298,
+      "grad_norm": 0.002803055103868246,
+      "learning_rate": 1.0087917757111001e-06,
+      "loss": 0.0,
+      "step": 878
+    },
+    {
+      "epoch": 4.352649006622516,
+      "grad_norm": 0.006075066048651934,
+      "learning_rate": 1.003098155565191e-06,
+      "loss": 0.0,
+      "step": 879
+    },
+    {
+      "epoch": 4.357615894039735,
+      "grad_norm": 0.001356339082121849,
+      "learning_rate": 9.974166130715895e-07,
+      "loss": 0.0,
+      "step": 880
+    },
+    {
+      "epoch": 4.362582781456954,
+      "grad_norm": 0.002172323176637292,
+      "learning_rate": 9.917471940714414e-07,
+      "loss": 0.0,
+      "step": 881
+    },
+    {
+      "epoch": 4.367549668874172,
+      "grad_norm": 0.0025884981732815504,
+      "learning_rate": 9.860899443080743e-07,
+      "loss": 0.0001,
+      "step": 882
+    },
+    {
+      "epoch": 4.372516556291391,
+      "grad_norm": 0.003944643307477236,
+      "learning_rate": 9.80444909426628e-07,
+      "loss": 0.0,
+      "step": 883
+    },
+    {
+      "epoch": 4.377483443708609,
+      "grad_norm": 0.002183829201385379,
+      "learning_rate": 9.74812134973689e-07,
+      "loss": 0.0,
+      "step": 884
+    },
+    {
+      "epoch": 4.382450331125828,
+      "grad_norm": 0.005370482802391052,
+      "learning_rate": 9.691916663969214e-07,
+      "loss": 0.0001,
+      "step": 885
+    },
+    {
+      "epoch": 4.387417218543046,
+      "grad_norm": 0.004634561482816935,
+      "learning_rate": 9.635835490446993e-07,
+      "loss": 0.0,
+      "step": 886
+    },
+    {
+      "epoch": 4.392384105960265,
+      "grad_norm": 0.0025161257945001125,
+      "learning_rate": 9.579878281657428e-07,
+      "loss": 0.0,
+      "step": 887
+    },
+    {
+      "epoch": 4.397350993377484,
+      "grad_norm": 0.004764308221638203,
+      "learning_rate": 9.524045489087494e-07,
+      "loss": 0.0,
+      "step": 888
+    },
+    {
+      "epoch": 4.402317880794702,
+      "grad_norm": 0.0024163629859685898,
+      "learning_rate": 9.468337563220351e-07,
+      "loss": 0.0,
+      "step": 889
+    },
+    {
+      "epoch": 4.4072847682119205,
+      "grad_norm": 0.005767002701759338,
+      "learning_rate": 9.412754953531664e-07,
+      "loss": 0.0001,
+      "step": 890
+    },
+    {
+      "epoch": 4.412251655629139,
+      "grad_norm": 0.0024256010074168444,
+      "learning_rate": 9.357298108486004e-07,
+      "loss": 0.0,
+      "step": 891
+    },
+    {
+      "epoch": 4.417218543046357,
+      "grad_norm": 0.004122421145439148,
+      "learning_rate": 9.301967475533216e-07,
+      "loss": 0.0001,
+      "step": 892
+    },
+    {
+      "epoch": 4.422185430463577,
+      "grad_norm": 0.003680233843624592,
+      "learning_rate": 9.246763501104794e-07,
+      "loss": 0.0,
+      "step": 893
+    },
+    {
+      "epoch": 4.427152317880795,
+      "grad_norm": 0.003066811477765441,
+      "learning_rate": 9.191686630610313e-07,
+      "loss": 0.0,
+      "step": 894
+    },
+    {
+      "epoch": 4.432119205298013,
+      "grad_norm": 0.003086198354139924,
+      "learning_rate": 9.136737308433822e-07,
+      "loss": 0.0,
+      "step": 895
+    },
+    {
+      "epoch": 4.437086092715232,
+      "grad_norm": 0.0018603306962177157,
+      "learning_rate": 9.081915977930242e-07,
+      "loss": 0.0,
+      "step": 896
+    },
+    {
+      "epoch": 4.44205298013245,
+      "grad_norm": 0.008020590990781784,
+      "learning_rate": 9.027223081421807e-07,
+      "loss": 0.0001,
+      "step": 897
+    },
+    {
+      "epoch": 4.447019867549669,
+      "grad_norm": 0.0016992095625028014,
+      "learning_rate": 8.972659060194505e-07,
+      "loss": 0.0,
+      "step": 898
+    },
+    {
+      "epoch": 4.451986754966887,
+      "grad_norm": 0.003738005878403783,
+      "learning_rate": 8.918224354494467e-07,
+      "loss": 0.0,
+      "step": 899
+    },
+    {
+      "epoch": 4.456953642384106,
+      "grad_norm": 0.0064679947681725025,
+      "learning_rate": 8.86391940352448e-07,
+      "loss": 0.0001,
+      "step": 900
+    },
+    {
+      "epoch": 4.461920529801325,
+      "grad_norm": 0.0033405704889446497,
+      "learning_rate": 8.809744645440405e-07,
+      "loss": 0.0001,
+      "step": 901
+    },
+    {
+      "epoch": 4.466887417218543,
+      "grad_norm": 0.0026824744418263435,
+      "learning_rate": 8.755700517347651e-07,
+      "loss": 0.0001,
+      "step": 902
+    },
+    {
+      "epoch": 4.4718543046357615,
+      "grad_norm": 0.003286386374384165,
+      "learning_rate": 8.701787455297647e-07,
+      "loss": 0.0,
+      "step": 903
+    },
+    {
+      "epoch": 4.47682119205298,
+      "grad_norm": 0.0038229452911764383,
+      "learning_rate": 8.648005894284325e-07,
+      "loss": 0.0,
+      "step": 904
+    },
+    {
+      "epoch": 4.481788079470198,
+      "grad_norm": 0.002128089778125286,
+      "learning_rate": 8.594356268240617e-07,
+      "loss": 0.0,
+      "step": 905
+    },
+    {
+      "epoch": 4.486754966887418,
+      "grad_norm": 0.0020676383282989264,
+      "learning_rate": 8.54083901003492e-07,
+      "loss": 0.0,
+      "step": 906
+    },
+    {
+      "epoch": 4.491721854304636,
+      "grad_norm": 0.0024742605164647102,
+      "learning_rate": 8.487454551467658e-07,
+      "loss": 0.0,
+      "step": 907
+    },
+    {
+      "epoch": 4.496688741721854,
+      "grad_norm": 0.0029265685006976128,
+      "learning_rate": 8.434203323267764e-07,
+      "loss": 0.0001,
+      "step": 908
+    },
+    {
+      "epoch": 4.501655629139073,
+      "grad_norm": 0.021650822833180428,
+      "learning_rate": 8.381085755089202e-07,
+      "loss": 0.0,
+      "step": 909
+    },
+    {
+      "epoch": 4.506622516556291,
+      "grad_norm": 0.0027908131014555693,
+      "learning_rate": 8.328102275507519e-07,
+      "loss": 0.0,
+      "step": 910
+    },
+    {
+      "epoch": 4.51158940397351,
+      "grad_norm": 0.0025155984330922365,
+      "learning_rate": 8.27525331201637e-07,
+      "loss": 0.0,
+      "step": 911
+    },
+    {
+      "epoch": 4.516556291390728,
+      "grad_norm": 0.004211023915559053,
+      "learning_rate": 8.222539291024079e-07,
+      "loss": 0.0,
+      "step": 912
+    },
+    {
+      "epoch": 4.521523178807947,
+      "grad_norm": 0.0038350329268723726,
+      "learning_rate": 8.169960637850194e-07,
+      "loss": 0.0,
+      "step": 913
+    },
+    {
+      "epoch": 4.526490066225166,
+      "grad_norm": 0.0038504060357809067,
+      "learning_rate": 8.117517776722067e-07,
+      "loss": 0.0,
+      "step": 914
+    },
+    {
+      "epoch": 4.531456953642384,
+      "grad_norm": 0.005449031945317984,
+      "learning_rate": 8.065211130771395e-07,
+      "loss": 0.0001,
+      "step": 915
+    },
+    {
+      "epoch": 4.5364238410596025,
+      "grad_norm": 0.0023090175818651915,
+      "learning_rate": 8.013041122030857e-07,
+      "loss": 0.0001,
+      "step": 916
+    },
+    {
+      "epoch": 4.541390728476821,
+      "grad_norm": 0.004053482785820961,
+      "learning_rate": 7.961008171430679e-07,
+      "loss": 0.0,
+      "step": 917
+    },
+    {
+      "epoch": 4.54635761589404,
+      "grad_norm": 0.011484500952064991,
+      "learning_rate": 7.909112698795233e-07,
+      "loss": 0.0,
+      "step": 918
+    },
+    {
+      "epoch": 4.551324503311259,
+      "grad_norm": 0.003258092561736703,
+      "learning_rate": 7.857355122839674e-07,
+      "loss": 0.0001,
+      "step": 919
+    },
+    {
+      "epoch": 4.556291390728477,
+      "grad_norm": 0.00284211547113955,
+      "learning_rate": 7.805735861166527e-07,
+      "loss": 0.0,
+      "step": 920
+    },
+    {
+      "epoch": 4.561258278145695,
+      "grad_norm": 0.0024764398112893105,
+      "learning_rate": 7.754255330262353e-07,
+      "loss": 0.0,
+      "step": 921
+    },
+    {
+      "epoch": 4.566225165562914,
+      "grad_norm": 0.0025240599643439054,
+      "learning_rate": 7.70291394549437e-07,
+      "loss": 0.0,
+      "step": 922
+    },
+    {
+      "epoch": 4.571192052980132,
+      "grad_norm": 0.0017671928508207202,
+      "learning_rate": 7.651712121107102e-07,
+      "loss": 0.0,
+      "step": 923
+    },
+    {
+      "epoch": 4.576158940397351,
+      "grad_norm": 0.003009581472724676,
+      "learning_rate": 7.600650270219045e-07,
+      "loss": 0.0,
+      "step": 924
+    },
+    {
+      "epoch": 4.58112582781457,
+      "grad_norm": 0.0028741969726979733,
+      "learning_rate": 7.549728804819326e-07,
+      "loss": 0.0,
+      "step": 925
+    },
+    {
+      "epoch": 4.586092715231788,
+      "grad_norm": 0.0025166133418679237,
+      "learning_rate": 7.49894813576437e-07,
+      "loss": 0.0,
+      "step": 926
+    },
+    {
+      "epoch": 4.591059602649007,
+      "grad_norm": 0.00866135023534298,
+      "learning_rate": 7.448308672774607e-07,
+      "loss": 0.0001,
+      "step": 927
+    },
+    {
+      "epoch": 4.596026490066225,
+      "grad_norm": 0.0022679597605019808,
+      "learning_rate": 7.397810824431156e-07,
+      "loss": 0.0,
+      "step": 928
+    },
+    {
+      "epoch": 4.6009933774834435,
+      "grad_norm": 0.004445367027074099,
+      "learning_rate": 7.347454998172523e-07,
+      "loss": 0.0,
+      "step": 929
+    },
+    {
+      "epoch": 4.605960264900662,
+      "grad_norm": 0.0038761026225984097,
+      "learning_rate": 7.297241600291335e-07,
+      "loss": 0.0,
+      "step": 930
+    },
+    {
+      "epoch": 4.610927152317881,
+      "grad_norm": 0.020167728886008263,
+      "learning_rate": 7.247171035931011e-07,
+      "loss": 0.0001,
+      "step": 931
+    },
+    {
+      "epoch": 4.6158940397351,
+      "grad_norm": 0.0027330240700393915,
+      "learning_rate": 7.197243709082555e-07,
+      "loss": 0.0,
+      "step": 932
+    },
+    {
+      "epoch": 4.620860927152318,
+      "grad_norm": 0.0024728626012802124,
+      "learning_rate": 7.147460022581257e-07,
+      "loss": 0.0,
+      "step": 933
+    },
+    {
+      "epoch": 4.625827814569536,
+      "grad_norm": 0.0028774477541446686,
+      "learning_rate": 7.097820378103465e-07,
+      "loss": 0.0,
+      "step": 934
+    },
+    {
+      "epoch": 4.630794701986755,
+      "grad_norm": 0.005338730756193399,
+      "learning_rate": 7.048325176163332e-07,
+      "loss": 0.0001,
+      "step": 935
+    },
+    {
+      "epoch": 4.635761589403973,
+      "grad_norm": 0.004418341908603907,
+      "learning_rate": 6.998974816109566e-07,
+      "loss": 0.0001,
+      "step": 936
+    },
+    {
+      "epoch": 4.640728476821192,
+      "grad_norm": 0.0033981436863541603,
+      "learning_rate": 6.949769696122249e-07,
+      "loss": 0.0,
+      "step": 937
+    },
+    {
+      "epoch": 4.645695364238411,
+      "grad_norm": 0.0035744442138820887,
+      "learning_rate": 6.900710213209596e-07,
+      "loss": 0.0001,
+      "step": 938
+    },
+    {
+      "epoch": 4.650662251655629,
+      "grad_norm": 0.005834420211613178,
+      "learning_rate": 6.851796763204766e-07,
+      "loss": 0.0,
+      "step": 939
+    },
+    {
+      "epoch": 4.655629139072848,
+      "grad_norm": 0.0030793931800872087,
+      "learning_rate": 6.803029740762648e-07,
+      "loss": 0.0,
+      "step": 940
+    },
+    {
+      "epoch": 4.660596026490066,
+      "grad_norm": 0.003223895560950041,
+      "learning_rate": 6.75440953935671e-07,
+      "loss": 0.0001,
+      "step": 941
+    },
+    {
+      "epoch": 4.6655629139072845,
+      "grad_norm": 0.004697455558925867,
+      "learning_rate": 6.705936551275774e-07,
+      "loss": 0.0,
+      "step": 942
+    },
+    {
+      "epoch": 4.670529801324504,
+      "grad_norm": 0.0021367729641497135,
+      "learning_rate": 6.657611167620909e-07,
+      "loss": 0.0001,
+      "step": 943
+    },
+    {
+      "epoch": 4.675496688741722,
+      "grad_norm": 0.008959624916315079,
+      "learning_rate": 6.609433778302235e-07,
+      "loss": 0.0,
+      "step": 944
+    },
+    {
+      "epoch": 4.680463576158941,
+      "grad_norm": 0.003460271516814828,
+      "learning_rate": 6.561404772035792e-07,
+      "loss": 0.0,
+      "step": 945
+    },
+    {
+      "epoch": 4.685430463576159,
+      "grad_norm": 0.002240000758320093,
+      "learning_rate": 6.513524536340412e-07,
+      "loss": 0.0,
+      "step": 946
+    },
+    {
+      "epoch": 4.690397350993377,
+      "grad_norm": 0.0028772966470569372,
+      "learning_rate": 6.465793457534553e-07,
+      "loss": 0.0,
+      "step": 947
+    },
+    {
+      "epoch": 4.695364238410596,
+      "grad_norm": 0.002292924327775836,
+      "learning_rate": 6.418211920733236e-07,
+      "loss": 0.0,
+      "step": 948
+    },
+    {
+      "epoch": 4.700331125827814,
+      "grad_norm": 0.002371082780882716,
+      "learning_rate": 6.370780309844907e-07,
+      "loss": 0.0,
+      "step": 949
+    },
+    {
+      "epoch": 4.7052980132450335,
+      "grad_norm": 0.0032526294235140085,
+      "learning_rate": 6.323499007568337e-07,
+      "loss": 0.0,
+      "step": 950
+    },
+    {
+      "epoch": 4.710264900662252,
+      "grad_norm": 0.0034900642931461334,
+      "learning_rate": 6.276368395389557e-07,
+      "loss": 0.0001,
+      "step": 951
+    },
+    {
+      "epoch": 4.71523178807947,
+      "grad_norm": 0.013080969452857971,
+      "learning_rate": 6.229388853578749e-07,
+      "loss": 0.0,
+      "step": 952
+    },
+    {
+      "epoch": 4.720198675496689,
+      "grad_norm": 0.0029847905971109867,
+      "learning_rate": 6.182560761187201e-07,
+      "loss": 0.0,
+      "step": 953
+    },
+    {
+      "epoch": 4.725165562913907,
+      "grad_norm": 0.0031260282266885042,
+      "learning_rate": 6.135884496044245e-07,
+      "loss": 0.0,
+      "step": 954
+    },
+    {
+      "epoch": 4.7301324503311255,
+      "grad_norm": 0.006553035695105791,
+      "learning_rate": 6.089360434754204e-07,
+      "loss": 0.0001,
+      "step": 955
+    },
+    {
+      "epoch": 4.735099337748345,
+      "grad_norm": 0.00812547281384468,
+      "learning_rate": 6.04298895269334e-07,
+      "loss": 0.0,
+      "step": 956
+    },
+    {
+      "epoch": 4.740066225165563,
+      "grad_norm": 0.002327099209651351,
+      "learning_rate": 5.996770424006857e-07,
+      "loss": 0.0,
+      "step": 957
+    },
+    {
+      "epoch": 4.745033112582782,
+      "grad_norm": 0.00504528172314167,
+      "learning_rate": 5.950705221605859e-07,
+      "loss": 0.0001,
+      "step": 958
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0031369777861982584,
+      "learning_rate": 5.90479371716435e-07,
+      "loss": 0.0,
+      "step": 959
+    },
+    {
+      "epoch": 4.754966887417218,
+      "grad_norm": 0.011925214901566505,
+      "learning_rate": 5.859036281116226e-07,
+      "loss": 0.0001,
+      "step": 960
+    },
+    {
+      "epoch": 4.759933774834437,
+      "grad_norm": 0.003020922187715769,
+      "learning_rate": 5.813433282652297e-07,
+      "loss": 0.0,
+      "step": 961
+    },
+    {
+      "epoch": 4.764900662251655,
+      "grad_norm": 0.002923258114606142,
+      "learning_rate": 5.767985089717312e-07,
+      "loss": 0.0,
+      "step": 962
+    },
+    {
+      "epoch": 4.7698675496688745,
+      "grad_norm": 0.001969816628843546,
+      "learning_rate": 5.722692069006958e-07,
+      "loss": 0.0,
+      "step": 963
+    },
+    {
+      "epoch": 4.774834437086093,
+      "grad_norm": 0.0044808341190218925,
+      "learning_rate": 5.677554585964945e-07,
+      "loss": 0.0,
+      "step": 964
+    },
+    {
+      "epoch": 4.779801324503311,
+      "grad_norm": 0.002164355479180813,
+      "learning_rate": 5.632573004780031e-07,
+      "loss": 0.0,
+      "step": 965
+    },
+    {
+      "epoch": 4.78476821192053,
+      "grad_norm": 0.0020656068809330463,
+      "learning_rate": 5.5877476883831e-07,
+      "loss": 0.0,
+      "step": 966
+    },
+    {
+      "epoch": 4.789735099337748,
+      "grad_norm": 0.007483918219804764,
+      "learning_rate": 5.543078998444207e-07,
+      "loss": 0.0,
+      "step": 967
+    },
+    {
+      "epoch": 4.7947019867549665,
+      "grad_norm": 0.0021352292969822884,
+      "learning_rate": 5.4985672953697e-07,
+      "loss": 0.0,
+      "step": 968
+    },
+    {
+      "epoch": 4.799668874172186,
+      "grad_norm": 0.0021575619466602802,
+      "learning_rate": 5.454212938299256e-07,
+      "loss": 0.0,
+      "step": 969
+    },
+    {
+      "epoch": 4.804635761589404,
+      "grad_norm": 0.0029056204948574305,
+      "learning_rate": 5.410016285103045e-07,
+      "loss": 0.0,
+      "step": 970
+    },
+    {
+      "epoch": 4.809602649006623,
+      "grad_norm": 0.0027450064662843943,
+      "learning_rate": 5.365977692378804e-07,
+      "loss": 0.0,
+      "step": 971
+    },
+    {
+      "epoch": 4.814569536423841,
+      "grad_norm": 0.002298718085512519,
+      "learning_rate": 5.322097515448971e-07,
+      "loss": 0.0,
+      "step": 972
+    },
+    {
+      "epoch": 4.819536423841059,
+      "grad_norm": 0.0021699871867895126,
+      "learning_rate": 5.278376108357819e-07,
+      "loss": 0.0,
+      "step": 973
+    },
+    {
+      "epoch": 4.824503311258278,
+      "grad_norm": 0.00849721860140562,
+      "learning_rate": 5.234813823868584e-07,
+      "loss": 0.0,
+      "step": 974
+    },
+    {
+      "epoch": 4.829470198675496,
+      "grad_norm": 0.002524444367736578,
+      "learning_rate": 5.191411013460646e-07,
+      "loss": 0.0,
+      "step": 975
+    },
+    {
+      "epoch": 4.8344370860927155,
+      "grad_norm": 0.003216218203306198,
+      "learning_rate": 5.148168027326672e-07,
+      "loss": 0.0,
+      "step": 976
+    },
+    {
+      "epoch": 4.839403973509934,
+      "grad_norm": 0.0029136831872165203,
+      "learning_rate": 5.105085214369806e-07,
+      "loss": 0.0001,
+      "step": 977
+    },
+    {
+      "epoch": 4.844370860927152,
+      "grad_norm": 0.003092824248597026,
+      "learning_rate": 5.062162922200844e-07,
+      "loss": 0.0,
+      "step": 978
+    },
+    {
+      "epoch": 4.849337748344371,
+      "grad_norm": 0.0023300074972212315,
+      "learning_rate": 5.019401497135415e-07,
+      "loss": 0.0,
+      "step": 979
+    },
+    {
+      "epoch": 4.854304635761589,
+      "grad_norm": 0.006137893535196781,
+      "learning_rate": 4.976801284191216e-07,
+      "loss": 0.0,
+      "step": 980
+    },
+    {
+      "epoch": 4.859271523178808,
+      "grad_norm": 0.005596006754785776,
+      "learning_rate": 4.934362627085218e-07,
+      "loss": 0.0001,
+      "step": 981
+    },
+    {
+      "epoch": 4.864238410596027,
+      "grad_norm": 0.007335844915360212,
+      "learning_rate": 4.892085868230881e-07,
+      "loss": 0.0,
+      "step": 982
+    },
+    {
+      "epoch": 4.869205298013245,
+      "grad_norm": 0.0026359714102,
+      "learning_rate": 4.849971348735405e-07,
+      "loss": 0.0,
+      "step": 983
+    },
+    {
+      "epoch": 4.874172185430464,
+      "grad_norm": 0.0036263810470700264,
+      "learning_rate": 4.808019408396977e-07,
+      "loss": 0.0,
+      "step": 984
+    },
+    {
+      "epoch": 4.879139072847682,
+      "grad_norm": 0.006067920010536909,
+      "learning_rate": 4.766230385702006e-07,
+      "loss": 0.0,
+      "step": 985
+    },
+    {
+      "epoch": 4.8841059602649,
+      "grad_norm": 0.0036801102105528116,
+      "learning_rate": 4.724604617822429e-07,
+      "loss": 0.0,
+      "step": 986
+    },
+    {
+      "epoch": 4.889072847682119,
+      "grad_norm": 0.0035613749641925097,
+      "learning_rate": 4.6831424406129674e-07,
+      "loss": 0.0,
+      "step": 987
+    },
+    {
+      "epoch": 4.894039735099338,
+      "grad_norm": 0.0014954039361327887,
+      "learning_rate": 4.6418441886084213e-07,
+      "loss": 0.0,
+      "step": 988
+    },
+    {
+      "epoch": 4.8990066225165565,
+      "grad_norm": 0.0019667819142341614,
+      "learning_rate": 4.6007101950209827e-07,
+      "loss": 0.0,
+      "step": 989
+    },
+    {
+      "epoch": 4.903973509933775,
+      "grad_norm": 0.0027786430437117815,
+      "learning_rate": 4.559740791737505e-07,
+      "loss": 0.0,
+      "step": 990
+    },
+    {
+      "epoch": 4.908940397350993,
+      "grad_norm": 0.0034396268893033266,
+      "learning_rate": 4.5189363093168875e-07,
+      "loss": 0.0001,
+      "step": 991
+    },
+    {
+      "epoch": 4.913907284768212,
+      "grad_norm": 0.006226430647075176,
+      "learning_rate": 4.4782970769873617e-07,
+      "loss": 0.0001,
+      "step": 992
+    },
+    {
+      "epoch": 4.91887417218543,
+      "grad_norm": 0.001311521977186203,
+      "learning_rate": 4.4378234226438555e-07,
+      "loss": 0.0,
+      "step": 993
+    },
+    {
+      "epoch": 4.923841059602649,
+      "grad_norm": 0.0034123172517865896,
+      "learning_rate": 4.397515672845334e-07,
+      "loss": 0.0,
+      "step": 994
+    },
+    {
+      "epoch": 4.928807947019868,
+      "grad_norm": 0.0030540761072188616,
+      "learning_rate": 4.357374152812191e-07,
+      "loss": 0.0,
+      "step": 995
+    },
+    {
+      "epoch": 4.933774834437086,
+      "grad_norm": 0.002518687630072236,
+      "learning_rate": 4.317399186423574e-07,
+      "loss": 0.0,
+      "step": 996
+    },
+    {
+      "epoch": 4.938741721854305,
+      "grad_norm": 0.003642142517492175,
+      "learning_rate": 4.277591096214834e-07,
+      "loss": 0.0,
+      "step": 997
+    },
+    {
+      "epoch": 4.943708609271523,
+      "grad_norm": 0.0037643522955477238,
+      "learning_rate": 4.237950203374877e-07,
+      "loss": 0.0001,
+      "step": 998
+    },
+    {
+      "epoch": 4.948675496688741,
+      "grad_norm": 0.00907442718744278,
+      "learning_rate": 4.1984768277435974e-07,
+      "loss": 0.0,
+      "step": 999
+    },
+    {
+      "epoch": 4.95364238410596,
+      "grad_norm": 0.004632457625120878,
+      "learning_rate": 4.15917128780928e-07,
+      "loss": 0.0001,
+      "step": 1000
+    },
+    {
+      "epoch": 4.958609271523179,
+      "grad_norm": 0.005262897349894047,
+      "learning_rate": 4.120033900706041e-07,
+      "loss": 0.0,
+      "step": 1001
+    },
+    {
+      "epoch": 4.9635761589403975,
+      "grad_norm": 0.002897941740229726,
+      "learning_rate": 4.0810649822112745e-07,
+      "loss": 0.0001,
+      "step": 1002
+    },
+    {
+      "epoch": 4.968543046357616,
+      "grad_norm": 0.016272269189357758,
+      "learning_rate": 4.0422648467430855e-07,
+      "loss": 0.0001,
+      "step": 1003
+    },
+    {
+      "epoch": 4.973509933774834,
+      "grad_norm": 0.004333376884460449,
+      "learning_rate": 4.003633807357768e-07,
+      "loss": 0.0,
+      "step": 1004
+    },
+    {
+      "epoch": 4.978476821192053,
+      "grad_norm": 0.0036270914133638144,
+      "learning_rate": 3.965172175747284e-07,
+      "loss": 0.0,
+      "step": 1005
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1206,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 201,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2453700044406129e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}