diff --git "a/checkpoint-5696/trainer_state.json" "b/checkpoint-5696/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-5696/trainer_state.json"
@@ -0,0 +1,39905 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 5696,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00035112359550561797,
+      "grad_norm": 2.8951518535614014,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 1.0953,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007022471910112359,
+      "grad_norm": 2.683199167251587,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.0614,
+      "step": 2
+    },
+    {
+      "epoch": 0.001053370786516854,
+      "grad_norm": 2.72992205619812,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.0674,
+      "step": 3
+    },
+    {
+      "epoch": 0.0014044943820224719,
+      "grad_norm": 2.6386663913726807,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.058,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017556179775280898,
+      "grad_norm": 2.672447443008423,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.0523,
+      "step": 5
+    },
+    {
+      "epoch": 0.002106741573033708,
+      "grad_norm": 2.5482404232025146,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.0454,
+      "step": 6
+    },
+    {
+      "epoch": 0.0024578651685393258,
+      "grad_norm": 2.651028633117676,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.0646,
+      "step": 7
+    },
+    {
+      "epoch": 0.0028089887640449437,
+      "grad_norm": 2.76011323928833,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.061,
+      "step": 8
+    },
+    {
+      "epoch": 0.0031601123595505617,
+      "grad_norm": 2.4421677589416504,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 1.0827,
+      "step": 9
+    },
+    {
+      "epoch": 0.0035112359550561797,
+      "grad_norm": 2.460603713989258,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.0925,
+      "step": 10
+    },
+    {
+      "epoch": 0.0038623595505617976,
+      "grad_norm": 2.3482112884521484,
+      "learning_rate": 5.5e-07,
+      "loss": 1.0679,
+      "step": 11
+    },
+    {
+      "epoch": 0.004213483146067416,
+      "grad_norm": 2.4199249744415283,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.0685,
+      "step": 12
+    },
+    {
+      "epoch": 0.004564606741573034,
+      "grad_norm": 2.171464443206787,
+      "learning_rate": 6.5e-07,
+      "loss": 1.0455,
+      "step": 13
+    },
+    {
+      "epoch": 0.0049157303370786515,
+      "grad_norm": 2.212657928466797,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.0718,
+      "step": 14
+    },
+    {
+      "epoch": 0.0052668539325842695,
+      "grad_norm": 2.0458426475524902,
+      "learning_rate": 7.5e-07,
+      "loss": 1.0213,
+      "step": 15
+    },
+    {
+      "epoch": 0.0056179775280898875,
+      "grad_norm": 1.8622595071792603,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.0266,
+      "step": 16
+    },
+    {
+      "epoch": 0.0059691011235955055,
+      "grad_norm": 1.7484514713287354,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 1.0439,
+      "step": 17
+    },
+    {
+      "epoch": 0.006320224719101123,
+      "grad_norm": 1.626994252204895,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.0274,
+      "step": 18
+    },
+    {
+      "epoch": 0.006671348314606741,
+      "grad_norm": 1.462096095085144,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 1.036,
+      "step": 19
+    },
+    {
+      "epoch": 0.007022471910112359,
+      "grad_norm": 1.4871413707733154,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.033,
+      "step": 20
+    },
+    {
+      "epoch": 0.007373595505617977,
+      "grad_norm": 1.4984582662582397,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.0279,
+      "step": 21
+    },
+    {
+      "epoch": 0.007724719101123595,
+      "grad_norm": 1.5213954448699951,
+      "learning_rate": 1.1e-06,
+      "loss": 1.0468,
+      "step": 22
+    },
+    {
+      "epoch": 0.008075842696629214,
+      "grad_norm": 1.3945717811584473,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 1.0252,
+      "step": 23
+    },
+    {
+      "epoch": 0.008426966292134831,
+      "grad_norm": 1.2632803916931152,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.9637,
+      "step": 24
+    },
+    {
+      "epoch": 0.00877808988764045,
+      "grad_norm": 1.2537329196929932,
+      "learning_rate": 1.25e-06,
+      "loss": 1.0194,
+      "step": 25
+    },
+    {
+      "epoch": 0.009129213483146067,
+      "grad_norm": 1.287577748298645,
+      "learning_rate": 1.3e-06,
+      "loss": 0.9491,
+      "step": 26
+    },
+    {
+      "epoch": 0.009480337078651686,
+      "grad_norm": 1.1792410612106323,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 1.0,
+      "step": 27
+    },
+    {
+      "epoch": 0.009831460674157303,
+      "grad_norm": 1.1308109760284424,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.002,
+      "step": 28
+    },
+    {
+      "epoch": 0.010182584269662922,
+      "grad_norm": 1.1292983293533325,
+      "learning_rate": 1.45e-06,
+      "loss": 1.0114,
+      "step": 29
+    },
+    {
+      "epoch": 0.010533707865168539,
+      "grad_norm": 1.0696202516555786,
+      "learning_rate": 1.5e-06,
+      "loss": 0.9371,
+      "step": 30
+    },
+    {
+      "epoch": 0.010884831460674158,
+      "grad_norm": 1.0532927513122559,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.9805,
+      "step": 31
+    },
+    {
+      "epoch": 0.011235955056179775,
+      "grad_norm": 1.039662480354309,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.9591,
+      "step": 32
+    },
+    {
+      "epoch": 0.011587078651685394,
+      "grad_norm": 0.9709979891777039,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.9599,
+      "step": 33
+    },
+    {
+      "epoch": 0.011938202247191011,
+      "grad_norm": 0.9826909303665161,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.9492,
+      "step": 34
+    },
+    {
+      "epoch": 0.01228932584269663,
+      "grad_norm": 1.01341712474823,
+      "learning_rate": 1.75e-06,
+      "loss": 1.0066,
+      "step": 35
+    },
+    {
+      "epoch": 0.012640449438202247,
+      "grad_norm": 1.0003409385681152,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.9641,
+      "step": 36
+    },
+    {
+      "epoch": 0.012991573033707866,
+      "grad_norm": 0.9853388667106628,
+      "learning_rate": 1.85e-06,
+      "loss": 0.9681,
+      "step": 37
+    },
+    {
+      "epoch": 0.013342696629213483,
+      "grad_norm": 0.9508251547813416,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.9242,
+      "step": 38
+    },
+    {
+      "epoch": 0.013693820224719102,
+      "grad_norm": 0.9218304753303528,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.8917,
+      "step": 39
+    },
+    {
+      "epoch": 0.014044943820224719,
+      "grad_norm": 0.9143016934394836,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8715,
+      "step": 40
+    },
+    {
+      "epoch": 0.014396067415730338,
+      "grad_norm": 0.9491222500801086,
+      "learning_rate": 2.05e-06,
+      "loss": 0.9157,
+      "step": 41
+    },
+    {
+      "epoch": 0.014747191011235955,
+      "grad_norm": 0.9453691244125366,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.9537,
+      "step": 42
+    },
+    {
+      "epoch": 0.015098314606741573,
+      "grad_norm": 0.9372655153274536,
+      "learning_rate": 2.15e-06,
+      "loss": 0.9288,
+      "step": 43
+    },
+    {
+      "epoch": 0.01544943820224719,
+      "grad_norm": 0.9183003306388855,
+      "learning_rate": 2.2e-06,
+      "loss": 0.8841,
+      "step": 44
+    },
+    {
+      "epoch": 0.01580056179775281,
+      "grad_norm": 0.924336314201355,
+      "learning_rate": 2.25e-06,
+      "loss": 0.9305,
+      "step": 45
+    },
+    {
+      "epoch": 0.016151685393258428,
+      "grad_norm": 0.9038196206092834,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.9376,
+      "step": 46
+    },
+    {
+      "epoch": 0.016502808988764044,
+      "grad_norm": 0.9091538786888123,
+      "learning_rate": 2.35e-06,
+      "loss": 0.9243,
+      "step": 47
+    },
+    {
+      "epoch": 0.016853932584269662,
+      "grad_norm": 0.9212149381637573,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.9515,
+      "step": 48
+    },
+    {
+      "epoch": 0.01720505617977528,
+      "grad_norm": 0.8974660038948059,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.893,
+      "step": 49
+    },
+    {
+      "epoch": 0.0175561797752809,
+      "grad_norm": 0.8716650009155273,
+      "learning_rate": 2.5e-06,
+      "loss": 0.9485,
+      "step": 50
+    },
+    {
+      "epoch": 0.017907303370786515,
+      "grad_norm": 0.8729946613311768,
+      "learning_rate": 2.55e-06,
+      "loss": 0.912,
+      "step": 51
+    },
+    {
+      "epoch": 0.018258426966292134,
+      "grad_norm": 0.8486078977584839,
+      "learning_rate": 2.6e-06,
+      "loss": 0.9479,
+      "step": 52
+    },
+    {
+      "epoch": 0.018609550561797753,
+      "grad_norm": 0.8713150024414062,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.9049,
+      "step": 53
+    },
+    {
+      "epoch": 0.018960674157303372,
+      "grad_norm": 0.855265200138092,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.921,
+      "step": 54
+    },
+    {
+      "epoch": 0.019311797752808987,
+      "grad_norm": 0.8279822468757629,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.8507,
+      "step": 55
+    },
+    {
+      "epoch": 0.019662921348314606,
+      "grad_norm": 0.8624402284622192,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.9292,
+      "step": 56
+    },
+    {
+      "epoch": 0.020014044943820225,
+      "grad_norm": 0.8482896089553833,
+      "learning_rate": 2.85e-06,
+      "loss": 0.881,
+      "step": 57
+    },
+    {
+      "epoch": 0.020365168539325844,
+      "grad_norm": 0.8510951399803162,
+      "learning_rate": 2.9e-06,
+      "loss": 0.8702,
+      "step": 58
+    },
+    {
+      "epoch": 0.02071629213483146,
+      "grad_norm": 0.8771702647209167,
+      "learning_rate": 2.95e-06,
+      "loss": 0.9329,
+      "step": 59
+    },
+    {
+      "epoch": 0.021067415730337078,
+      "grad_norm": 0.8551446199417114,
+      "learning_rate": 3e-06,
+      "loss": 0.8908,
+      "step": 60
+    },
+    {
+      "epoch": 0.021418539325842697,
+      "grad_norm": 0.8884496688842773,
+      "learning_rate": 3.05e-06,
+      "loss": 0.8791,
+      "step": 61
+    },
+    {
+      "epoch": 0.021769662921348316,
+      "grad_norm": 0.8115338087081909,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.8544,
+      "step": 62
+    },
+    {
+      "epoch": 0.02212078651685393,
+      "grad_norm": 0.8549612760543823,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.8963,
+      "step": 63
+    },
+    {
+      "epoch": 0.02247191011235955,
+      "grad_norm": 0.8259643316268921,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.8299,
+      "step": 64
+    },
+    {
+      "epoch": 0.02282303370786517,
+      "grad_norm": 0.8829634785652161,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.8873,
+      "step": 65
+    },
+    {
+      "epoch": 0.023174157303370788,
+      "grad_norm": 0.8536106944084167,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.9009,
+      "step": 66
+    },
+    {
+      "epoch": 0.023525280898876403,
+      "grad_norm": 0.8468846082687378,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.8921,
+      "step": 67
+    },
+    {
+      "epoch": 0.023876404494382022,
+      "grad_norm": 0.8253934383392334,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.848,
+      "step": 68
+    },
+    {
+      "epoch": 0.02422752808988764,
+      "grad_norm": 0.8778877854347229,
+      "learning_rate": 3.45e-06,
+      "loss": 0.8887,
+      "step": 69
+    },
+    {
+      "epoch": 0.02457865168539326,
+      "grad_norm": 0.8790880441665649,
+      "learning_rate": 3.5e-06,
+      "loss": 0.8849,
+      "step": 70
+    },
+    {
+      "epoch": 0.024929775280898875,
+      "grad_norm": 0.843599259853363,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.8972,
+      "step": 71
+    },
+    {
+      "epoch": 0.025280898876404494,
+      "grad_norm": 0.8428463339805603,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.9027,
+      "step": 72
+    },
+    {
+      "epoch": 0.025632022471910113,
+      "grad_norm": 0.8088714480400085,
+      "learning_rate": 3.65e-06,
+      "loss": 0.8562,
+      "step": 73
+    },
+    {
+      "epoch": 0.02598314606741573,
+      "grad_norm": 0.857280969619751,
+      "learning_rate": 3.7e-06,
+      "loss": 0.8725,
+      "step": 74
+    },
+    {
+      "epoch": 0.026334269662921347,
+      "grad_norm": 0.8321940898895264,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8656,
+      "step": 75
+    },
+    {
+      "epoch": 0.026685393258426966,
+      "grad_norm": 0.8289700150489807,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.8211,
+      "step": 76
+    },
+    {
+      "epoch": 0.027036516853932584,
+      "grad_norm": 0.8454659581184387,
+      "learning_rate": 3.85e-06,
+      "loss": 0.9003,
+      "step": 77
+    },
+    {
+      "epoch": 0.027387640449438203,
+      "grad_norm": 0.8450126647949219,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.8595,
+      "step": 78
+    },
+    {
+      "epoch": 0.02773876404494382,
+      "grad_norm": 0.8216248750686646,
+      "learning_rate": 3.95e-06,
+      "loss": 0.8503,
+      "step": 79
+    },
+    {
+      "epoch": 0.028089887640449437,
+      "grad_norm": 0.8196912407875061,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8965,
+      "step": 80
+    },
+    {
+      "epoch": 0.028441011235955056,
+      "grad_norm": 0.8264586329460144,
+      "learning_rate": 4.05e-06,
+      "loss": 0.879,
+      "step": 81
+    },
+    {
+      "epoch": 0.028792134831460675,
+      "grad_norm": 0.8712279200553894,
+      "learning_rate": 4.1e-06,
+      "loss": 0.8825,
+      "step": 82
+    },
+    {
+      "epoch": 0.02914325842696629,
+      "grad_norm": 0.8609493970870972,
+      "learning_rate": 4.15e-06,
+      "loss": 0.8901,
+      "step": 83
+    },
+    {
+      "epoch": 0.02949438202247191,
+      "grad_norm": 0.8273038864135742,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.8589,
+      "step": 84
+    },
+    {
+      "epoch": 0.029845505617977528,
+      "grad_norm": 0.8164820075035095,
+      "learning_rate": 4.25e-06,
+      "loss": 0.8099,
+      "step": 85
+    },
+    {
+      "epoch": 0.030196629213483147,
+      "grad_norm": 0.8666368722915649,
+      "learning_rate": 4.3e-06,
+      "loss": 0.8535,
+      "step": 86
+    },
+    {
+      "epoch": 0.030547752808988762,
+      "grad_norm": 0.8173773288726807,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.806,
+      "step": 87
+    },
+    {
+      "epoch": 0.03089887640449438,
+      "grad_norm": 0.8484762907028198,
+      "learning_rate": 4.4e-06,
+      "loss": 0.8678,
+      "step": 88
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 0.8572366833686829,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.8939,
+      "step": 89
+    },
+    {
+      "epoch": 0.03160112359550562,
+      "grad_norm": 0.8265636563301086,
+      "learning_rate": 4.5e-06,
+      "loss": 0.8373,
+      "step": 90
+    },
+    {
+      "epoch": 0.03195224719101124,
+      "grad_norm": 0.8756334185600281,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.8322,
+      "step": 91
+    },
+    {
+      "epoch": 0.032303370786516857,
+      "grad_norm": 0.8236197829246521,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.8425,
+      "step": 92
+    },
+    {
+      "epoch": 0.032654494382022475,
+      "grad_norm": 0.855594277381897,
+      "learning_rate": 4.65e-06,
+      "loss": 0.8779,
+      "step": 93
+    },
+    {
+      "epoch": 0.03300561797752809,
+      "grad_norm": 0.8004137873649597,
+      "learning_rate": 4.7e-06,
+      "loss": 0.8078,
+      "step": 94
+    },
+    {
+      "epoch": 0.033356741573033706,
+      "grad_norm": 0.7953411340713501,
+      "learning_rate": 4.75e-06,
+      "loss": 0.8363,
+      "step": 95
+    },
+    {
+      "epoch": 0.033707865168539325,
+      "grad_norm": 0.8278580904006958,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.8301,
+      "step": 96
+    },
+    {
+      "epoch": 0.034058988764044944,
+      "grad_norm": 0.8473982810974121,
+      "learning_rate": 4.85e-06,
+      "loss": 0.8408,
+      "step": 97
+    },
+    {
+      "epoch": 0.03441011235955056,
+      "grad_norm": 0.8142527341842651,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.8295,
+      "step": 98
+    },
+    {
+      "epoch": 0.03476123595505618,
+      "grad_norm": 0.8610883355140686,
+      "learning_rate": 4.95e-06,
+      "loss": 0.8405,
+      "step": 99
+    },
+    {
+      "epoch": 0.0351123595505618,
+      "grad_norm": 0.8903759717941284,
+      "learning_rate": 5e-06,
+      "loss": 0.8793,
+      "step": 100
+    },
+    {
+      "epoch": 0.03546348314606742,
+      "grad_norm": 0.8775374293327332,
+      "learning_rate": 4.999999957251071e-06,
+      "loss": 0.8516,
+      "step": 101
+    },
+    {
+      "epoch": 0.03581460674157303,
+      "grad_norm": 0.8515768647193909,
+      "learning_rate": 4.999999829004281e-06,
+      "loss": 0.8045,
+      "step": 102
+    },
+    {
+      "epoch": 0.03616573033707865,
+      "grad_norm": 1.0148895978927612,
+      "learning_rate": 4.9999996152596355e-06,
+      "loss": 0.8247,
+      "step": 103
+    },
+    {
+      "epoch": 0.03651685393258427,
+      "grad_norm": 0.829903244972229,
+      "learning_rate": 4.999999316017144e-06,
+      "loss": 0.8282,
+      "step": 104
+    },
+    {
+      "epoch": 0.03686797752808989,
+      "grad_norm": 0.8467276096343994,
+      "learning_rate": 4.999998931276815e-06,
+      "loss": 0.8325,
+      "step": 105
+    },
+    {
+      "epoch": 0.037219101123595506,
+      "grad_norm": 0.8355855941772461,
+      "learning_rate": 4.999998461038661e-06,
+      "loss": 0.813,
+      "step": 106
+    },
+    {
+      "epoch": 0.037570224719101125,
+      "grad_norm": 0.8558890223503113,
+      "learning_rate": 4.999997905302699e-06,
+      "loss": 0.8592,
+      "step": 107
+    },
+    {
+      "epoch": 0.037921348314606744,
+      "grad_norm": 0.9446994662284851,
+      "learning_rate": 4.999997264068948e-06,
+      "loss": 0.8323,
+      "step": 108
+    },
+    {
+      "epoch": 0.03827247191011236,
+      "grad_norm": 0.8493000268936157,
+      "learning_rate": 4.99999653733743e-06,
+      "loss": 0.8178,
+      "step": 109
+    },
+    {
+      "epoch": 0.038623595505617975,
+      "grad_norm": 0.8132368922233582,
+      "learning_rate": 4.99999572510817e-06,
+      "loss": 0.7977,
+      "step": 110
+    },
+    {
+      "epoch": 0.038974719101123594,
+      "grad_norm": 0.8313632607460022,
+      "learning_rate": 4.999994827381196e-06,
+      "loss": 0.8407,
+      "step": 111
+    },
+    {
+      "epoch": 0.03932584269662921,
+      "grad_norm": 0.8673725724220276,
+      "learning_rate": 4.999993844156537e-06,
+      "loss": 0.862,
+      "step": 112
+    },
+    {
+      "epoch": 0.03967696629213483,
+      "grad_norm": 0.8794373273849487,
+      "learning_rate": 4.999992775434229e-06,
+      "loss": 0.8599,
+      "step": 113
+    },
+    {
+      "epoch": 0.04002808988764045,
+      "grad_norm": 0.8612070083618164,
+      "learning_rate": 4.9999916212143065e-06,
+      "loss": 0.8013,
+      "step": 114
+    },
+    {
+      "epoch": 0.04037921348314607,
+      "grad_norm": 0.8925568461418152,
+      "learning_rate": 4.9999903814968095e-06,
+      "loss": 0.7529,
+      "step": 115
+    },
+    {
+      "epoch": 0.04073033707865169,
+      "grad_norm": 0.8791881799697876,
+      "learning_rate": 4.999989056281781e-06,
+      "loss": 0.8756,
+      "step": 116
+    },
+    {
+      "epoch": 0.04108146067415731,
+      "grad_norm": 0.8983411192893982,
+      "learning_rate": 4.999987645569267e-06,
+      "loss": 0.8329,
+      "step": 117
+    },
+    {
+      "epoch": 0.04143258426966292,
+      "grad_norm": 0.8358251452445984,
+      "learning_rate": 4.999986149359314e-06,
+      "loss": 0.7792,
+      "step": 118
+    },
+    {
+      "epoch": 0.04178370786516854,
+      "grad_norm": 0.8848925828933716,
+      "learning_rate": 4.999984567651974e-06,
+      "loss": 0.8242,
+      "step": 119
+    },
+    {
+      "epoch": 0.042134831460674156,
+      "grad_norm": 0.9365871548652649,
+      "learning_rate": 4.999982900447301e-06,
+      "loss": 0.7875,
+      "step": 120
+    },
+    {
+      "epoch": 0.042485955056179775,
+      "grad_norm": 0.8330186605453491,
+      "learning_rate": 4.999981147745352e-06,
+      "loss": 0.8387,
+      "step": 121
+    },
+    {
+      "epoch": 0.042837078651685394,
+      "grad_norm": 0.8438044190406799,
+      "learning_rate": 4.999979309546187e-06,
+      "loss": 0.8547,
+      "step": 122
+    },
+    {
+      "epoch": 0.04318820224719101,
+      "grad_norm": 0.8812524676322937,
+      "learning_rate": 4.999977385849869e-06,
+      "loss": 0.7891,
+      "step": 123
+    },
+    {
+      "epoch": 0.04353932584269663,
+      "grad_norm": 0.8498764634132385,
+      "learning_rate": 4.999975376656464e-06,
+      "loss": 0.8259,
+      "step": 124
+    },
+    {
+      "epoch": 0.04389044943820225,
+      "grad_norm": 0.8646953105926514,
+      "learning_rate": 4.99997328196604e-06,
+      "loss": 0.8269,
+      "step": 125
+    },
+    {
+      "epoch": 0.04424157303370786,
+      "grad_norm": 0.8175535798072815,
+      "learning_rate": 4.9999711017786686e-06,
+      "loss": 0.7943,
+      "step": 126
+    },
+    {
+      "epoch": 0.04459269662921348,
+      "grad_norm": 0.8761390447616577,
+      "learning_rate": 4.999968836094425e-06,
+      "loss": 0.8668,
+      "step": 127
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 0.8925103545188904,
+      "learning_rate": 4.999966484913387e-06,
+      "loss": 0.8562,
+      "step": 128
+    },
+    {
+      "epoch": 0.04529494382022472,
+      "grad_norm": 0.8750078678131104,
+      "learning_rate": 4.999964048235634e-06,
+      "loss": 0.8058,
+      "step": 129
+    },
+    {
+      "epoch": 0.04564606741573034,
+      "grad_norm": 0.8446925282478333,
+      "learning_rate": 4.999961526061251e-06,
+      "loss": 0.8197,
+      "step": 130
+    },
+    {
+      "epoch": 0.045997191011235956,
+      "grad_norm": 0.8947305679321289,
+      "learning_rate": 4.999958918390321e-06,
+      "loss": 0.8347,
+      "step": 131
+    },
+    {
+      "epoch": 0.046348314606741575,
+      "grad_norm": 0.8591403961181641,
+      "learning_rate": 4.999956225222937e-06,
+      "loss": 0.8117,
+      "step": 132
+    },
+    {
+      "epoch": 0.046699438202247194,
+      "grad_norm": 0.8698204755783081,
+      "learning_rate": 4.999953446559188e-06,
+      "loss": 0.8378,
+      "step": 133
+    },
+    {
+      "epoch": 0.047050561797752806,
+      "grad_norm": 0.8451970219612122,
+      "learning_rate": 4.999950582399171e-06,
+      "loss": 0.8324,
+      "step": 134
+    },
+    {
+      "epoch": 0.047401685393258425,
+      "grad_norm": 0.8550758957862854,
+      "learning_rate": 4.999947632742984e-06,
+      "loss": 0.8291,
+      "step": 135
+    },
+    {
+      "epoch": 0.047752808988764044,
+      "grad_norm": 0.8345015645027161,
+      "learning_rate": 4.999944597590726e-06,
+      "loss": 0.8315,
+      "step": 136
+    },
+    {
+      "epoch": 0.04810393258426966,
+      "grad_norm": 0.9103791117668152,
+      "learning_rate": 4.999941476942502e-06,
+      "loss": 0.8009,
+      "step": 137
+    },
+    {
+      "epoch": 0.04845505617977528,
+      "grad_norm": 0.9898645877838135,
+      "learning_rate": 4.99993827079842e-06,
+      "loss": 0.8574,
+      "step": 138
+    },
+    {
+      "epoch": 0.0488061797752809,
+      "grad_norm": 0.8579599261283875,
+      "learning_rate": 4.9999349791585876e-06,
+      "loss": 0.7747,
+      "step": 139
+    },
+    {
+      "epoch": 0.04915730337078652,
+      "grad_norm": 0.8869126439094543,
+      "learning_rate": 4.999931602023118e-06,
+      "loss": 0.8472,
+      "step": 140
+    },
+    {
+      "epoch": 0.04950842696629214,
+      "grad_norm": 0.9130380749702454,
+      "learning_rate": 4.999928139392127e-06,
+      "loss": 0.8247,
+      "step": 141
+    },
+    {
+      "epoch": 0.04985955056179775,
+      "grad_norm": 0.8806836605072021,
+      "learning_rate": 4.999924591265732e-06,
+      "loss": 0.8658,
+      "step": 142
+    },
+    {
+      "epoch": 0.05021067415730337,
+      "grad_norm": 0.9065203666687012,
+      "learning_rate": 4.999920957644057e-06,
+      "loss": 0.8097,
+      "step": 143
+    },
+    {
+      "epoch": 0.05056179775280899,
+      "grad_norm": 0.8982106447219849,
+      "learning_rate": 4.999917238527223e-06,
+      "loss": 0.8207,
+      "step": 144
+    },
+    {
+      "epoch": 0.050912921348314606,
+      "grad_norm": 0.9152344465255737,
+      "learning_rate": 4.999913433915359e-06,
+      "loss": 0.845,
+      "step": 145
+    },
+    {
+      "epoch": 0.051264044943820225,
+      "grad_norm": 0.8733872771263123,
+      "learning_rate": 4.999909543808595e-06,
+      "loss": 0.8337,
+      "step": 146
+    },
+    {
+      "epoch": 0.051615168539325844,
+      "grad_norm": 0.8482681512832642,
+      "learning_rate": 4.999905568207064e-06,
+      "loss": 0.8226,
+      "step": 147
+    },
+    {
+      "epoch": 0.05196629213483146,
+      "grad_norm": 0.8690842986106873,
+      "learning_rate": 4.999901507110901e-06,
+      "loss": 0.8677,
+      "step": 148
+    },
+    {
+      "epoch": 0.05231741573033708,
+      "grad_norm": 0.8329693675041199,
+      "learning_rate": 4.999897360520246e-06,
+      "loss": 0.7628,
+      "step": 149
+    },
+    {
+      "epoch": 0.05266853932584269,
+      "grad_norm": 0.913270890712738,
+      "learning_rate": 4.999893128435241e-06,
+      "loss": 0.8374,
+      "step": 150
+    },
+    {
+      "epoch": 0.05301966292134831,
+      "grad_norm": 0.843010425567627,
+      "learning_rate": 4.999888810856029e-06,
+      "loss": 0.8343,
+      "step": 151
+    },
+    {
+      "epoch": 0.05337078651685393,
+      "grad_norm": 0.8601850867271423,
+      "learning_rate": 4.99988440778276e-06,
+      "loss": 0.8254,
+      "step": 152
+    },
+    {
+      "epoch": 0.05372191011235955,
+      "grad_norm": 0.8458811640739441,
+      "learning_rate": 4.999879919215583e-06,
+      "loss": 0.8238,
+      "step": 153
+    },
+    {
+      "epoch": 0.05407303370786517,
+      "grad_norm": 0.8649472594261169,
+      "learning_rate": 4.999875345154652e-06,
+      "loss": 0.8545,
+      "step": 154
+    },
+    {
+      "epoch": 0.05442415730337079,
+      "grad_norm": 0.8692646622657776,
+      "learning_rate": 4.999870685600123e-06,
+      "loss": 0.769,
+      "step": 155
+    },
+    {
+      "epoch": 0.054775280898876406,
+      "grad_norm": 0.8382206559181213,
+      "learning_rate": 4.999865940552157e-06,
+      "loss": 0.8076,
+      "step": 156
+    },
+    {
+      "epoch": 0.055126404494382025,
+      "grad_norm": 0.8909122943878174,
+      "learning_rate": 4.999861110010914e-06,
+      "loss": 0.8118,
+      "step": 157
+    },
+    {
+      "epoch": 0.05547752808988764,
+      "grad_norm": 0.8760329484939575,
+      "learning_rate": 4.99985619397656e-06,
+      "loss": 0.8625,
+      "step": 158
+    },
+    {
+      "epoch": 0.055828651685393256,
+      "grad_norm": 1.0767908096313477,
+      "learning_rate": 4.999851192449263e-06,
+      "loss": 0.8038,
+      "step": 159
+    },
+    {
+      "epoch": 0.056179775280898875,
+      "grad_norm": 0.8571433424949646,
+      "learning_rate": 4.999846105429196e-06,
+      "loss": 0.8123,
+      "step": 160
+    },
+    {
+      "epoch": 0.056530898876404494,
+      "grad_norm": 0.8204314112663269,
+      "learning_rate": 4.99984093291653e-06,
+      "loss": 0.7713,
+      "step": 161
+    },
+    {
+      "epoch": 0.05688202247191011,
+      "grad_norm": 0.8622918128967285,
+      "learning_rate": 4.9998356749114434e-06,
+      "loss": 0.7816,
+      "step": 162
+    },
+    {
+      "epoch": 0.05723314606741573,
+      "grad_norm": 0.991470456123352,
+      "learning_rate": 4.999830331414116e-06,
+      "loss": 0.8171,
+      "step": 163
+    },
+    {
+      "epoch": 0.05758426966292135,
+      "grad_norm": 0.9167494177818298,
+      "learning_rate": 4.99982490242473e-06,
+      "loss": 0.8096,
+      "step": 164
+    },
+    {
+      "epoch": 0.05793539325842697,
+      "grad_norm": 0.8794699311256409,
+      "learning_rate": 4.999819387943472e-06,
+      "loss": 0.8284,
+      "step": 165
+    },
+    {
+      "epoch": 0.05828651685393258,
+      "grad_norm": 0.9374570250511169,
+      "learning_rate": 4.99981378797053e-06,
+      "loss": 0.7887,
+      "step": 166
+    },
+    {
+      "epoch": 0.0586376404494382,
+      "grad_norm": 0.8803961277008057,
+      "learning_rate": 4.999808102506095e-06,
+      "loss": 0.8331,
+      "step": 167
+    },
+    {
+      "epoch": 0.05898876404494382,
+      "grad_norm": 0.903806209564209,
+      "learning_rate": 4.999802331550363e-06,
+      "loss": 0.8002,
+      "step": 168
+    },
+    {
+      "epoch": 0.05933988764044944,
+      "grad_norm": 0.8599410057067871,
+      "learning_rate": 4.99979647510353e-06,
+      "loss": 0.8342,
+      "step": 169
+    },
+    {
+      "epoch": 0.059691011235955056,
+      "grad_norm": 0.9048354625701904,
+      "learning_rate": 4.999790533165797e-06,
+      "loss": 0.8197,
+      "step": 170
+    },
+    {
+      "epoch": 0.060042134831460675,
+      "grad_norm": 0.8594435453414917,
+      "learning_rate": 4.999784505737366e-06,
+      "loss": 0.7994,
+      "step": 171
+    },
+    {
+      "epoch": 0.060393258426966294,
+      "grad_norm": 0.8999492526054382,
+      "learning_rate": 4.999778392818444e-06,
+      "loss": 0.8062,
+      "step": 172
+    },
+    {
+      "epoch": 0.06074438202247191,
+      "grad_norm": 0.8730378150939941,
+      "learning_rate": 4.999772194409241e-06,
+      "loss": 0.8348,
+      "step": 173
+    },
+    {
+      "epoch": 0.061095505617977525,
+      "grad_norm": 0.8853862881660461,
+      "learning_rate": 4.999765910509967e-06,
+      "loss": 0.8235,
+      "step": 174
+    },
+    {
+      "epoch": 0.061446629213483143,
+      "grad_norm": 0.8512648344039917,
+      "learning_rate": 4.999759541120839e-06,
+      "loss": 0.8254,
+      "step": 175
+    },
+    {
+      "epoch": 0.06179775280898876,
+      "grad_norm": 0.925890326499939,
+      "learning_rate": 4.999753086242073e-06,
+      "loss": 0.8498,
+      "step": 176
+    },
+    {
+      "epoch": 0.06214887640449438,
+      "grad_norm": 0.9359338283538818,
+      "learning_rate": 4.999746545873891e-06,
+      "loss": 0.8006,
+      "step": 177
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.8511932492256165,
+      "learning_rate": 4.999739920016516e-06,
+      "loss": 0.7843,
+      "step": 178
+    },
+    {
+      "epoch": 0.06285112359550561,
+      "grad_norm": 0.920088529586792,
+      "learning_rate": 4.999733208670174e-06,
+      "loss": 0.8173,
+      "step": 179
+    },
+    {
+      "epoch": 0.06320224719101124,
+      "grad_norm": 0.8991772532463074,
+      "learning_rate": 4.9997264118350965e-06,
+      "loss": 0.8227,
+      "step": 180
+    },
+    {
+      "epoch": 0.06355337078651685,
+      "grad_norm": 0.9123396873474121,
+      "learning_rate": 4.9997195295115145e-06,
+      "loss": 0.7767,
+      "step": 181
+    },
+    {
+      "epoch": 0.06390449438202248,
+      "grad_norm": 0.888533353805542,
+      "learning_rate": 4.9997125616996635e-06,
+      "loss": 0.8093,
+      "step": 182
+    },
+    {
+      "epoch": 0.06425561797752809,
+      "grad_norm": 0.9099261164665222,
+      "learning_rate": 4.999705508399782e-06,
+      "loss": 0.8142,
+      "step": 183
+    },
+    {
+      "epoch": 0.06460674157303371,
+      "grad_norm": 0.9250622987747192,
+      "learning_rate": 4.999698369612111e-06,
+      "loss": 0.7913,
+      "step": 184
+    },
+    {
+      "epoch": 0.06495786516853932,
+      "grad_norm": 0.8754624128341675,
+      "learning_rate": 4.999691145336895e-06,
+      "loss": 0.7755,
+      "step": 185
+    },
+    {
+      "epoch": 0.06530898876404495,
+      "grad_norm": 0.8488731384277344,
+      "learning_rate": 4.99968383557438e-06,
+      "loss": 0.8209,
+      "step": 186
+    },
+    {
+      "epoch": 0.06566011235955056,
+      "grad_norm": 0.8827488422393799,
+      "learning_rate": 4.9996764403248175e-06,
+      "loss": 0.8121,
+      "step": 187
+    },
+    {
+      "epoch": 0.06601123595505617,
+      "grad_norm": 0.8604486584663391,
+      "learning_rate": 4.99966895958846e-06,
+      "loss": 0.8046,
+      "step": 188
+    },
+    {
+      "epoch": 0.0663623595505618,
+      "grad_norm": 0.855309247970581,
+      "learning_rate": 4.999661393365562e-06,
+      "loss": 0.7345,
+      "step": 189
+    },
+    {
+      "epoch": 0.06671348314606741,
+      "grad_norm": 0.8943816423416138,
+      "learning_rate": 4.999653741656385e-06,
+      "loss": 0.7937,
+      "step": 190
+    },
+    {
+      "epoch": 0.06706460674157304,
+      "grad_norm": 0.9106631278991699,
+      "learning_rate": 4.999646004461188e-06,
+      "loss": 0.8157,
+      "step": 191
+    },
+    {
+      "epoch": 0.06741573033707865,
+      "grad_norm": 0.9037092924118042,
+      "learning_rate": 4.999638181780237e-06,
+      "loss": 0.8091,
+      "step": 192
+    },
+    {
+      "epoch": 0.06776685393258428,
+      "grad_norm": 0.854783833026886,
+      "learning_rate": 4.9996302736137994e-06,
+      "loss": 0.7784,
+      "step": 193
+    },
+    {
+      "epoch": 0.06811797752808989,
+      "grad_norm": 0.8792631030082703,
+      "learning_rate": 4.999622279962144e-06,
+      "loss": 0.7737,
+      "step": 194
+    },
+    {
+      "epoch": 0.0684691011235955,
+      "grad_norm": 0.8999345898628235,
+      "learning_rate": 4.9996142008255465e-06,
+      "loss": 0.7858,
+      "step": 195
+    },
+    {
+      "epoch": 0.06882022471910113,
+      "grad_norm": 0.8954832553863525,
+      "learning_rate": 4.999606036204282e-06,
+      "loss": 0.8045,
+      "step": 196
+    },
+    {
+      "epoch": 0.06917134831460674,
+      "grad_norm": 0.8854857087135315,
+      "learning_rate": 4.999597786098631e-06,
+      "loss": 0.7842,
+      "step": 197
+    },
+    {
+      "epoch": 0.06952247191011236,
+      "grad_norm": 0.8617668747901917,
+      "learning_rate": 4.999589450508874e-06,
+      "loss": 0.807,
+      "step": 198
+    },
+    {
+      "epoch": 0.06987359550561797,
+      "grad_norm": 0.8737844824790955,
+      "learning_rate": 4.999581029435296e-06,
+      "loss": 0.8268,
+      "step": 199
+    },
+    {
+      "epoch": 0.0702247191011236,
+      "grad_norm": 0.8421310782432556,
+      "learning_rate": 4.999572522878186e-06,
+      "loss": 0.7701,
+      "step": 200
+    },
+    {
+      "epoch": 0.07057584269662921,
+      "grad_norm": 0.8911473155021667,
+      "learning_rate": 4.9995639308378365e-06,
+      "loss": 0.756,
+      "step": 201
+    },
+    {
+      "epoch": 0.07092696629213484,
+      "grad_norm": 0.879984974861145,
+      "learning_rate": 4.999555253314538e-06,
+      "loss": 0.8023,
+      "step": 202
+    },
+    {
+      "epoch": 0.07127808988764045,
+      "grad_norm": 0.9726935625076294,
+      "learning_rate": 4.9995464903085885e-06,
+      "loss": 0.8017,
+      "step": 203
+    },
+    {
+      "epoch": 0.07162921348314606,
+      "grad_norm": 0.9257283210754395,
+      "learning_rate": 4.999537641820288e-06,
+      "loss": 0.8336,
+      "step": 204
+    },
+    {
+      "epoch": 0.07198033707865169,
+      "grad_norm": 0.9612401723861694,
+      "learning_rate": 4.9995287078499385e-06,
+      "loss": 0.8014,
+      "step": 205
+    },
+    {
+      "epoch": 0.0723314606741573,
+      "grad_norm": 0.8790804743766785,
+      "learning_rate": 4.999519688397847e-06,
+      "loss": 0.7592,
+      "step": 206
+    },
+    {
+      "epoch": 0.07268258426966293,
+      "grad_norm": 0.9160227179527283,
+      "learning_rate": 4.999510583464321e-06,
+      "loss": 0.7905,
+      "step": 207
+    },
+    {
+      "epoch": 0.07303370786516854,
+      "grad_norm": 0.9160066246986389,
+      "learning_rate": 4.999501393049671e-06,
+      "loss": 0.8382,
+      "step": 208
+    },
+    {
+      "epoch": 0.07338483146067416,
+      "grad_norm": 0.8971685767173767,
+      "learning_rate": 4.999492117154214e-06,
+      "loss": 0.7881,
+      "step": 209
+    },
+    {
+      "epoch": 0.07373595505617977,
+      "grad_norm": 0.9171693921089172,
+      "learning_rate": 4.9994827557782644e-06,
+      "loss": 0.8101,
+      "step": 210
+    },
+    {
+      "epoch": 0.07408707865168539,
+      "grad_norm": 0.8446071743965149,
+      "learning_rate": 4.999473308922143e-06,
+      "loss": 0.7749,
+      "step": 211
+    },
+    {
+      "epoch": 0.07443820224719101,
+      "grad_norm": 0.8867083787918091,
+      "learning_rate": 4.999463776586174e-06,
+      "loss": 0.7656,
+      "step": 212
+    },
+    {
+      "epoch": 0.07478932584269662,
+      "grad_norm": 0.8746408820152283,
+      "learning_rate": 4.9994541587706815e-06,
+      "loss": 0.7778,
+      "step": 213
+    },
+    {
+      "epoch": 0.07514044943820225,
+      "grad_norm": 0.884986162185669,
+      "learning_rate": 4.999444455475997e-06,
+      "loss": 0.7975,
+      "step": 214
+    },
+    {
+      "epoch": 0.07549157303370786,
+      "grad_norm": 0.8919047713279724,
+      "learning_rate": 4.9994346667024505e-06,
+      "loss": 0.7552,
+      "step": 215
+    },
+    {
+      "epoch": 0.07584269662921349,
+      "grad_norm": 0.8840435147285461,
+      "learning_rate": 4.999424792450377e-06,
+      "loss": 0.7876,
+      "step": 216
+    },
+    {
+      "epoch": 0.0761938202247191,
+      "grad_norm": 0.9024021625518799,
+      "learning_rate": 4.999414832720114e-06,
+      "loss": 0.7921,
+      "step": 217
+    },
+    {
+      "epoch": 0.07654494382022473,
+      "grad_norm": 0.8887689113616943,
+      "learning_rate": 4.999404787512003e-06,
+      "loss": 0.7843,
+      "step": 218
+    },
+    {
+      "epoch": 0.07689606741573034,
+      "grad_norm": 0.903285026550293,
+      "learning_rate": 4.9993946568263866e-06,
+      "loss": 0.8164,
+      "step": 219
+    },
+    {
+      "epoch": 0.07724719101123595,
+      "grad_norm": 0.8626529574394226,
+      "learning_rate": 4.9993844406636124e-06,
+      "loss": 0.7941,
+      "step": 220
+    },
+    {
+      "epoch": 0.07759831460674158,
+      "grad_norm": 0.8919528126716614,
+      "learning_rate": 4.999374139024028e-06,
+      "loss": 0.8054,
+      "step": 221
+    },
+    {
+      "epoch": 0.07794943820224719,
+      "grad_norm": 0.9159940481185913,
+      "learning_rate": 4.999363751907988e-06,
+      "loss": 0.8071,
+      "step": 222
+    },
+    {
+      "epoch": 0.07830056179775281,
+      "grad_norm": 0.8792368173599243,
+      "learning_rate": 4.999353279315846e-06,
+      "loss": 0.7949,
+      "step": 223
+    },
+    {
+      "epoch": 0.07865168539325842,
+      "grad_norm": 0.9184228181838989,
+      "learning_rate": 4.9993427212479604e-06,
+      "loss": 0.8013,
+      "step": 224
+    },
+    {
+      "epoch": 0.07900280898876405,
+      "grad_norm": 0.8925946354866028,
+      "learning_rate": 4.999332077704692e-06,
+      "loss": 0.813,
+      "step": 225
+    },
+    {
+      "epoch": 0.07935393258426966,
+      "grad_norm": 0.9273242354393005,
+      "learning_rate": 4.999321348686406e-06,
+      "loss": 0.8107,
+      "step": 226
+    },
+    {
+      "epoch": 0.07970505617977527,
+      "grad_norm": 0.8914328813552856,
+      "learning_rate": 4.999310534193468e-06,
+      "loss": 0.8229,
+      "step": 227
+    },
+    {
+      "epoch": 0.0800561797752809,
+      "grad_norm": 0.9217993021011353,
+      "learning_rate": 4.999299634226249e-06,
+      "loss": 0.7722,
+      "step": 228
+    },
+    {
+      "epoch": 0.08040730337078651,
+      "grad_norm": 0.9615241289138794,
+      "learning_rate": 4.999288648785121e-06,
+      "loss": 0.8105,
+      "step": 229
+    },
+    {
+      "epoch": 0.08075842696629214,
+      "grad_norm": 0.9098490476608276,
+      "learning_rate": 4.99927757787046e-06,
+      "loss": 0.814,
+      "step": 230
+    },
+    {
+      "epoch": 0.08110955056179775,
+      "grad_norm": 0.9211421012878418,
+      "learning_rate": 4.999266421482645e-06,
+      "loss": 0.7542,
+      "step": 231
+    },
+    {
+      "epoch": 0.08146067415730338,
+      "grad_norm": 0.841416597366333,
+      "learning_rate": 4.999255179622056e-06,
+      "loss": 0.7961,
+      "step": 232
+    },
+    {
+      "epoch": 0.08181179775280899,
+      "grad_norm": 0.8942486643791199,
+      "learning_rate": 4.9992438522890785e-06,
+      "loss": 0.8095,
+      "step": 233
+    },
+    {
+      "epoch": 0.08216292134831461,
+      "grad_norm": 0.8535089492797852,
+      "learning_rate": 4.9992324394841005e-06,
+      "loss": 0.801,
+      "step": 234
+    },
+    {
+      "epoch": 0.08251404494382023,
+      "grad_norm": 1.02827787399292,
+      "learning_rate": 4.999220941207511e-06,
+      "loss": 0.8057,
+      "step": 235
+    },
+    {
+      "epoch": 0.08286516853932584,
+      "grad_norm": 0.9189361929893494,
+      "learning_rate": 4.9992093574597046e-06,
+      "loss": 0.7682,
+      "step": 236
+    },
+    {
+      "epoch": 0.08321629213483146,
+      "grad_norm": 0.9160380959510803,
+      "learning_rate": 4.999197688241076e-06,
+      "loss": 0.8218,
+      "step": 237
+    },
+    {
+      "epoch": 0.08356741573033707,
+      "grad_norm": 0.8860989809036255,
+      "learning_rate": 4.999185933552027e-06,
+      "loss": 0.7907,
+      "step": 238
+    },
+    {
+      "epoch": 0.0839185393258427,
+      "grad_norm": 0.8413179516792297,
+      "learning_rate": 4.999174093392956e-06,
+      "loss": 0.7711,
+      "step": 239
+    },
+    {
+      "epoch": 0.08426966292134831,
+      "grad_norm": 0.8843936324119568,
+      "learning_rate": 4.99916216776427e-06,
+      "loss": 0.7356,
+      "step": 240
+    },
+    {
+      "epoch": 0.08462078651685394,
+      "grad_norm": 0.9339604377746582,
+      "learning_rate": 4.999150156666376e-06,
+      "loss": 0.8017,
+      "step": 241
+    },
+    {
+      "epoch": 0.08497191011235955,
+      "grad_norm": 0.8932393789291382,
+      "learning_rate": 4.999138060099685e-06,
+      "loss": 0.7773,
+      "step": 242
+    },
+    {
+      "epoch": 0.08532303370786516,
+      "grad_norm": 0.9117473363876343,
+      "learning_rate": 4.999125878064611e-06,
+      "loss": 0.7828,
+      "step": 243
+    },
+    {
+      "epoch": 0.08567415730337079,
+      "grad_norm": 0.8696773052215576,
+      "learning_rate": 4.999113610561571e-06,
+      "loss": 0.7681,
+      "step": 244
+    },
+    {
+      "epoch": 0.0860252808988764,
+      "grad_norm": 0.9062797427177429,
+      "learning_rate": 4.999101257590985e-06,
+      "loss": 0.7766,
+      "step": 245
+    },
+    {
+      "epoch": 0.08637640449438203,
+      "grad_norm": 0.893018364906311,
+      "learning_rate": 4.999088819153273e-06,
+      "loss": 0.735,
+      "step": 246
+    },
+    {
+      "epoch": 0.08672752808988764,
+      "grad_norm": 0.9077386856079102,
+      "learning_rate": 4.999076295248863e-06,
+      "loss": 0.8092,
+      "step": 247
+    },
+    {
+      "epoch": 0.08707865168539326,
+      "grad_norm": 0.9474624395370483,
+      "learning_rate": 4.9990636858781805e-06,
+      "loss": 0.7841,
+      "step": 248
+    },
+    {
+      "epoch": 0.08742977528089887,
+      "grad_norm": 0.8823204040527344,
+      "learning_rate": 4.9990509910416595e-06,
+      "loss": 0.7827,
+      "step": 249
+    },
+    {
+      "epoch": 0.0877808988764045,
+      "grad_norm": 0.9361265897750854,
+      "learning_rate": 4.999038210739733e-06,
+      "loss": 0.8274,
+      "step": 250
+    },
+    {
+      "epoch": 0.08813202247191011,
+      "grad_norm": 0.9128723740577698,
+      "learning_rate": 4.999025344972838e-06,
+      "loss": 0.7681,
+      "step": 251
+    },
+    {
+      "epoch": 0.08848314606741572,
+      "grad_norm": 0.8730833530426025,
+      "learning_rate": 4.999012393741415e-06,
+      "loss": 0.8086,
+      "step": 252
+    },
+    {
+      "epoch": 0.08883426966292135,
+      "grad_norm": 0.8662939071655273,
+      "learning_rate": 4.998999357045906e-06,
+      "loss": 0.7587,
+      "step": 253
+    },
+    {
+      "epoch": 0.08918539325842696,
+      "grad_norm": 0.8748407959938049,
+      "learning_rate": 4.998986234886758e-06,
+      "loss": 0.7937,
+      "step": 254
+    },
+    {
+      "epoch": 0.08953651685393259,
+      "grad_norm": 0.9770292639732361,
+      "learning_rate": 4.998973027264419e-06,
+      "loss": 0.7815,
+      "step": 255
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 0.8764484524726868,
+      "learning_rate": 4.99895973417934e-06,
+      "loss": 0.766,
+      "step": 256
+    },
+    {
+      "epoch": 0.09023876404494383,
+      "grad_norm": 0.8940339684486389,
+      "learning_rate": 4.998946355631978e-06,
+      "loss": 0.8166,
+      "step": 257
+    },
+    {
+      "epoch": 0.09058988764044944,
+      "grad_norm": 0.903728723526001,
+      "learning_rate": 4.998932891622788e-06,
+      "loss": 0.7801,
+      "step": 258
+    },
+    {
+      "epoch": 0.09094101123595505,
+      "grad_norm": 0.9147090315818787,
+      "learning_rate": 4.998919342152232e-06,
+      "loss": 0.814,
+      "step": 259
+    },
+    {
+      "epoch": 0.09129213483146068,
+      "grad_norm": 0.9121170043945312,
+      "learning_rate": 4.9989057072207725e-06,
+      "loss": 0.7718,
+      "step": 260
+    },
+    {
+      "epoch": 0.09164325842696629,
+      "grad_norm": 0.8610017895698547,
+      "learning_rate": 4.998891986828877e-06,
+      "loss": 0.7539,
+      "step": 261
+    },
+    {
+      "epoch": 0.09199438202247191,
+      "grad_norm": 0.9325037598609924,
+      "learning_rate": 4.9988781809770124e-06,
+      "loss": 0.7699,
+      "step": 262
+    },
+    {
+      "epoch": 0.09234550561797752,
+      "grad_norm": 0.92354416847229,
+      "learning_rate": 4.998864289665654e-06,
+      "loss": 0.8224,
+      "step": 263
+    },
+    {
+      "epoch": 0.09269662921348315,
+      "grad_norm": 0.9121785759925842,
+      "learning_rate": 4.998850312895274e-06,
+      "loss": 0.7881,
+      "step": 264
+    },
+    {
+      "epoch": 0.09304775280898876,
+      "grad_norm": 0.8718385696411133,
+      "learning_rate": 4.998836250666352e-06,
+      "loss": 0.7865,
+      "step": 265
+    },
+    {
+      "epoch": 0.09339887640449439,
+      "grad_norm": 0.8944951295852661,
+      "learning_rate": 4.9988221029793685e-06,
+      "loss": 0.8315,
+      "step": 266
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 0.9362013339996338,
+      "learning_rate": 4.998807869834807e-06,
+      "loss": 0.7572,
+      "step": 267
+    },
+    {
+      "epoch": 0.09410112359550561,
+      "grad_norm": 0.8949217796325684,
+      "learning_rate": 4.998793551233155e-06,
+      "loss": 0.7614,
+      "step": 268
+    },
+    {
+      "epoch": 0.09445224719101124,
+      "grad_norm": 0.9898912310600281,
+      "learning_rate": 4.9987791471749015e-06,
+      "loss": 0.8111,
+      "step": 269
+    },
+    {
+      "epoch": 0.09480337078651685,
+      "grad_norm": 0.8814307451248169,
+      "learning_rate": 4.9987646576605395e-06,
+      "loss": 0.7774,
+      "step": 270
+    },
+    {
+      "epoch": 0.09515449438202248,
+      "grad_norm": 0.9214469790458679,
+      "learning_rate": 4.998750082690564e-06,
+      "loss": 0.7944,
+      "step": 271
+    },
+    {
+      "epoch": 0.09550561797752809,
+      "grad_norm": 0.8990268707275391,
+      "learning_rate": 4.998735422265475e-06,
+      "loss": 0.7661,
+      "step": 272
+    },
+    {
+      "epoch": 0.09585674157303371,
+      "grad_norm": 0.8504504561424255,
+      "learning_rate": 4.998720676385772e-06,
+      "loss": 0.7643,
+      "step": 273
+    },
+    {
+      "epoch": 0.09620786516853932,
+      "grad_norm": 0.9197403788566589,
+      "learning_rate": 4.99870584505196e-06,
+      "loss": 0.7763,
+      "step": 274
+    },
+    {
+      "epoch": 0.09655898876404495,
+      "grad_norm": 0.9017580151557922,
+      "learning_rate": 4.998690928264547e-06,
+      "loss": 0.8137,
+      "step": 275
+    },
+    {
+      "epoch": 0.09691011235955056,
+      "grad_norm": 0.880264937877655,
+      "learning_rate": 4.9986759260240416e-06,
+      "loss": 0.8088,
+      "step": 276
+    },
+    {
+      "epoch": 0.09726123595505617,
+      "grad_norm": 0.8751420378684998,
+      "learning_rate": 4.998660838330958e-06,
+      "loss": 0.7802,
+      "step": 277
+    },
+    {
+      "epoch": 0.0976123595505618,
+      "grad_norm": 0.8619885444641113,
+      "learning_rate": 4.998645665185812e-06,
+      "loss": 0.7584,
+      "step": 278
+    },
+    {
+      "epoch": 0.09796348314606741,
+      "grad_norm": 0.8772746324539185,
+      "learning_rate": 4.998630406589122e-06,
+      "loss": 0.7739,
+      "step": 279
+    },
+    {
+      "epoch": 0.09831460674157304,
+      "grad_norm": 0.9019191265106201,
+      "learning_rate": 4.998615062541411e-06,
+      "loss": 0.806,
+      "step": 280
+    },
+    {
+      "epoch": 0.09866573033707865,
+      "grad_norm": 0.9600672721862793,
+      "learning_rate": 4.998599633043202e-06,
+      "loss": 0.7902,
+      "step": 281
+    },
+    {
+      "epoch": 0.09901685393258428,
+      "grad_norm": 0.9067400097846985,
+      "learning_rate": 4.9985841180950245e-06,
+      "loss": 0.7779,
+      "step": 282
+    },
+    {
+      "epoch": 0.09936797752808989,
+      "grad_norm": 0.8911014795303345,
+      "learning_rate": 4.998568517697408e-06,
+      "loss": 0.7605,
+      "step": 283
+    },
+    {
+      "epoch": 0.0997191011235955,
+      "grad_norm": 0.8919551968574524,
+      "learning_rate": 4.998552831850886e-06,
+      "loss": 0.7416,
+      "step": 284
+    },
+    {
+      "epoch": 0.10007022471910113,
+      "grad_norm": 0.8880231380462646,
+      "learning_rate": 4.998537060555995e-06,
+      "loss": 0.7757,
+      "step": 285
+    },
+    {
+      "epoch": 0.10042134831460674,
+      "grad_norm": 0.9274663329124451,
+      "learning_rate": 4.998521203813275e-06,
+      "loss": 0.7925,
+      "step": 286
+    },
+    {
+      "epoch": 0.10077247191011236,
+      "grad_norm": 0.9023240804672241,
+      "learning_rate": 4.998505261623266e-06,
+      "loss": 0.8081,
+      "step": 287
+    },
+    {
+      "epoch": 0.10112359550561797,
+      "grad_norm": 0.8825272917747498,
+      "learning_rate": 4.998489233986518e-06,
+      "loss": 0.7864,
+      "step": 288
+    },
+    {
+      "epoch": 0.1014747191011236,
+      "grad_norm": 0.885826826095581,
+      "learning_rate": 4.9984731209035745e-06,
+      "loss": 0.7759,
+      "step": 289
+    },
+    {
+      "epoch": 0.10182584269662921,
+      "grad_norm": 0.8802645802497864,
+      "learning_rate": 4.998456922374988e-06,
+      "loss": 0.769,
+      "step": 290
+    },
+    {
+      "epoch": 0.10217696629213484,
+      "grad_norm": 0.9090613126754761,
+      "learning_rate": 4.998440638401314e-06,
+      "loss": 0.7931,
+      "step": 291
+    },
+    {
+      "epoch": 0.10252808988764045,
+      "grad_norm": 0.9373996257781982,
+      "learning_rate": 4.998424268983107e-06,
+      "loss": 0.7443,
+      "step": 292
+    },
+    {
+      "epoch": 0.10287921348314606,
+      "grad_norm": 0.9003362655639648,
+      "learning_rate": 4.998407814120928e-06,
+      "loss": 0.7983,
+      "step": 293
+    },
+    {
+      "epoch": 0.10323033707865169,
+      "grad_norm": 0.9281851053237915,
+      "learning_rate": 4.998391273815341e-06,
+      "loss": 0.8026,
+      "step": 294
+    },
+    {
+      "epoch": 0.1035814606741573,
+      "grad_norm": 0.8824150562286377,
+      "learning_rate": 4.998374648066909e-06,
+      "loss": 0.7435,
+      "step": 295
+    },
+    {
+      "epoch": 0.10393258426966293,
+      "grad_norm": 0.9006941318511963,
+      "learning_rate": 4.998357936876202e-06,
+      "loss": 0.7567,
+      "step": 296
+    },
+    {
+      "epoch": 0.10428370786516854,
+      "grad_norm": 0.8622567057609558,
+      "learning_rate": 4.998341140243792e-06,
+      "loss": 0.7532,
+      "step": 297
+    },
+    {
+      "epoch": 0.10463483146067416,
+      "grad_norm": 0.8983448147773743,
+      "learning_rate": 4.9983242581702525e-06,
+      "loss": 0.7686,
+      "step": 298
+    },
+    {
+      "epoch": 0.10498595505617977,
+      "grad_norm": 0.8893060684204102,
+      "learning_rate": 4.998307290656162e-06,
+      "loss": 0.7697,
+      "step": 299
+    },
+    {
+      "epoch": 0.10533707865168539,
+      "grad_norm": 0.8906247615814209,
+      "learning_rate": 4.998290237702098e-06,
+      "loss": 0.7628,
+      "step": 300
+    },
+    {
+      "epoch": 0.10568820224719101,
+      "grad_norm": 0.8763146996498108,
+      "learning_rate": 4.998273099308647e-06,
+      "loss": 0.7742,
+      "step": 301
+    },
+    {
+      "epoch": 0.10603932584269662,
+      "grad_norm": 0.8954668045043945,
+      "learning_rate": 4.998255875476394e-06,
+      "loss": 0.7666,
+      "step": 302
+    },
+    {
+      "epoch": 0.10639044943820225,
+      "grad_norm": 0.9178535342216492,
+      "learning_rate": 4.9982385662059275e-06,
+      "loss": 0.7679,
+      "step": 303
+    },
+    {
+      "epoch": 0.10674157303370786,
+      "grad_norm": 0.9487611651420593,
+      "learning_rate": 4.99822117149784e-06,
+      "loss": 0.7744,
+      "step": 304
+    },
+    {
+      "epoch": 0.10709269662921349,
+      "grad_norm": 0.9466339945793152,
+      "learning_rate": 4.998203691352726e-06,
+      "loss": 0.7608,
+      "step": 305
+    },
+    {
+      "epoch": 0.1074438202247191,
+      "grad_norm": 0.9768984913825989,
+      "learning_rate": 4.998186125771184e-06,
+      "loss": 0.8028,
+      "step": 306
+    },
+    {
+      "epoch": 0.10779494382022473,
+      "grad_norm": 0.9045122861862183,
+      "learning_rate": 4.998168474753814e-06,
+      "loss": 0.7365,
+      "step": 307
+    },
+    {
+      "epoch": 0.10814606741573034,
+      "grad_norm": 0.9124319553375244,
+      "learning_rate": 4.99815073830122e-06,
+      "loss": 0.7686,
+      "step": 308
+    },
+    {
+      "epoch": 0.10849719101123595,
+      "grad_norm": 0.9172065854072571,
+      "learning_rate": 4.998132916414008e-06,
+      "loss": 0.778,
+      "step": 309
+    },
+    {
+      "epoch": 0.10884831460674158,
+      "grad_norm": 0.9291918873786926,
+      "learning_rate": 4.9981150090927876e-06,
+      "loss": 0.7828,
+      "step": 310
+    },
+    {
+      "epoch": 0.10919943820224719,
+      "grad_norm": 0.9258780479431152,
+      "learning_rate": 4.998097016338172e-06,
+      "loss": 0.7822,
+      "step": 311
+    },
+    {
+      "epoch": 0.10955056179775281,
+      "grad_norm": 0.9169499278068542,
+      "learning_rate": 4.998078938150776e-06,
+      "loss": 0.7735,
+      "step": 312
+    },
+    {
+      "epoch": 0.10990168539325842,
+      "grad_norm": 0.8983737826347351,
+      "learning_rate": 4.998060774531218e-06,
+      "loss": 0.7729,
+      "step": 313
+    },
+    {
+      "epoch": 0.11025280898876405,
+      "grad_norm": 0.949192225933075,
+      "learning_rate": 4.99804252548012e-06,
+      "loss": 0.8153,
+      "step": 314
+    },
+    {
+      "epoch": 0.11060393258426966,
+      "grad_norm": 0.9175098538398743,
+      "learning_rate": 4.998024190998104e-06,
+      "loss": 0.7784,
+      "step": 315
+    },
+    {
+      "epoch": 0.11095505617977527,
+      "grad_norm": 0.9033719897270203,
+      "learning_rate": 4.998005771085799e-06,
+      "loss": 0.748,
+      "step": 316
+    },
+    {
+      "epoch": 0.1113061797752809,
+      "grad_norm": 0.9692884683609009,
+      "learning_rate": 4.997987265743834e-06,
+      "loss": 0.7795,
+      "step": 317
+    },
+    {
+      "epoch": 0.11165730337078651,
+      "grad_norm": 0.8999868035316467,
+      "learning_rate": 4.997968674972842e-06,
+      "loss": 0.7564,
+      "step": 318
+    },
+    {
+      "epoch": 0.11200842696629214,
+      "grad_norm": 0.9252747893333435,
+      "learning_rate": 4.997949998773458e-06,
+      "loss": 0.8054,
+      "step": 319
+    },
+    {
+      "epoch": 0.11235955056179775,
+      "grad_norm": 0.8910780549049377,
+      "learning_rate": 4.997931237146323e-06,
+      "loss": 0.7671,
+      "step": 320
+    },
+    {
+      "epoch": 0.11271067415730338,
+      "grad_norm": 0.9301937818527222,
+      "learning_rate": 4.997912390092077e-06,
+      "loss": 0.7906,
+      "step": 321
+    },
+    {
+      "epoch": 0.11306179775280899,
+      "grad_norm": 0.9235373735427856,
+      "learning_rate": 4.997893457611364e-06,
+      "loss": 0.7626,
+      "step": 322
+    },
+    {
+      "epoch": 0.11341292134831461,
+      "grad_norm": 0.9015495777130127,
+      "learning_rate": 4.997874439704833e-06,
+      "loss": 0.7634,
+      "step": 323
+    },
+    {
+      "epoch": 0.11376404494382023,
+      "grad_norm": 0.8859757781028748,
+      "learning_rate": 4.997855336373133e-06,
+      "loss": 0.7435,
+      "step": 324
+    },
+    {
+      "epoch": 0.11411516853932584,
+      "grad_norm": 0.9129098057746887,
+      "learning_rate": 4.9978361476169176e-06,
+      "loss": 0.7505,
+      "step": 325
+    },
+    {
+      "epoch": 0.11446629213483146,
+      "grad_norm": 0.9024680256843567,
+      "learning_rate": 4.997816873436845e-06,
+      "loss": 0.7402,
+      "step": 326
+    },
+    {
+      "epoch": 0.11481741573033707,
+      "grad_norm": 0.9123002290725708,
+      "learning_rate": 4.997797513833571e-06,
+      "loss": 0.7751,
+      "step": 327
+    },
+    {
+      "epoch": 0.1151685393258427,
+      "grad_norm": 0.8788150548934937,
+      "learning_rate": 4.9977780688077606e-06,
+      "loss": 0.7882,
+      "step": 328
+    },
+    {
+      "epoch": 0.11551966292134831,
+      "grad_norm": 0.9502024054527283,
+      "learning_rate": 4.997758538360077e-06,
+      "loss": 0.7777,
+      "step": 329
+    },
+    {
+      "epoch": 0.11587078651685394,
+      "grad_norm": 0.9458324909210205,
+      "learning_rate": 4.9977389224911884e-06,
+      "loss": 0.7666,
+      "step": 330
+    },
+    {
+      "epoch": 0.11622191011235955,
+      "grad_norm": 0.9191945195198059,
+      "learning_rate": 4.997719221201766e-06,
+      "loss": 0.7678,
+      "step": 331
+    },
+    {
+      "epoch": 0.11657303370786516,
+      "grad_norm": 0.9368025660514832,
+      "learning_rate": 4.997699434492485e-06,
+      "loss": 0.7689,
+      "step": 332
+    },
+    {
+      "epoch": 0.11692415730337079,
+      "grad_norm": 0.9120011925697327,
+      "learning_rate": 4.99767956236402e-06,
+      "loss": 0.7489,
+      "step": 333
+    },
+    {
+      "epoch": 0.1172752808988764,
+      "grad_norm": 0.9374188780784607,
+      "learning_rate": 4.997659604817051e-06,
+      "loss": 0.7358,
+      "step": 334
+    },
+    {
+      "epoch": 0.11762640449438203,
+      "grad_norm": 0.9129819273948669,
+      "learning_rate": 4.997639561852261e-06,
+      "loss": 0.7591,
+      "step": 335
+    },
+    {
+      "epoch": 0.11797752808988764,
+      "grad_norm": 0.9249037504196167,
+      "learning_rate": 4.997619433470334e-06,
+      "loss": 0.782,
+      "step": 336
+    },
+    {
+      "epoch": 0.11832865168539326,
+      "grad_norm": 0.8872626423835754,
+      "learning_rate": 4.9975992196719614e-06,
+      "loss": 0.7477,
+      "step": 337
+    },
+    {
+      "epoch": 0.11867977528089887,
+      "grad_norm": 0.9343289732933044,
+      "learning_rate": 4.9975789204578315e-06,
+      "loss": 0.7702,
+      "step": 338
+    },
+    {
+      "epoch": 0.1190308988764045,
+      "grad_norm": 0.9379141926765442,
+      "learning_rate": 4.997558535828641e-06,
+      "loss": 0.7847,
+      "step": 339
+    },
+    {
+      "epoch": 0.11938202247191011,
+      "grad_norm": 0.9160316586494446,
+      "learning_rate": 4.997538065785085e-06,
+      "loss": 0.7658,
+      "step": 340
+    },
+    {
+      "epoch": 0.11973314606741572,
+      "grad_norm": 0.9021292328834534,
+      "learning_rate": 4.997517510327864e-06,
+      "loss": 0.8114,
+      "step": 341
+    },
+    {
+      "epoch": 0.12008426966292135,
+      "grad_norm": 0.9155046343803406,
+      "learning_rate": 4.997496869457681e-06,
+      "loss": 0.7579,
+      "step": 342
+    },
+    {
+      "epoch": 0.12043539325842696,
+      "grad_norm": 0.8777235746383667,
+      "learning_rate": 4.997476143175243e-06,
+      "loss": 0.7832,
+      "step": 343
+    },
+    {
+      "epoch": 0.12078651685393259,
+      "grad_norm": 0.8728359341621399,
+      "learning_rate": 4.997455331481258e-06,
+      "loss": 0.7647,
+      "step": 344
+    },
+    {
+      "epoch": 0.1211376404494382,
+      "grad_norm": 0.9153422713279724,
+      "learning_rate": 4.997434434376437e-06,
+      "loss": 0.7396,
+      "step": 345
+    },
+    {
+      "epoch": 0.12148876404494383,
+      "grad_norm": 0.900204598903656,
+      "learning_rate": 4.9974134518614964e-06,
+      "loss": 0.7619,
+      "step": 346
+    },
+    {
+      "epoch": 0.12183988764044944,
+      "grad_norm": 0.9725254774093628,
+      "learning_rate": 4.997392383937153e-06,
+      "loss": 0.7588,
+      "step": 347
+    },
+    {
+      "epoch": 0.12219101123595505,
+      "grad_norm": 0.9041215777397156,
+      "learning_rate": 4.997371230604126e-06,
+      "loss": 0.7712,
+      "step": 348
+    },
+    {
+      "epoch": 0.12254213483146068,
+      "grad_norm": 0.9517892599105835,
+      "learning_rate": 4.997349991863141e-06,
+      "loss": 0.7758,
+      "step": 349
+    },
+    {
+      "epoch": 0.12289325842696629,
+      "grad_norm": 0.9502705335617065,
+      "learning_rate": 4.997328667714922e-06,
+      "loss": 0.8003,
+      "step": 350
+    },
+    {
+      "epoch": 0.12324438202247191,
+      "grad_norm": 0.9965865612030029,
+      "learning_rate": 4.9973072581602005e-06,
+      "loss": 0.751,
+      "step": 351
+    },
+    {
+      "epoch": 0.12359550561797752,
+      "grad_norm": 0.9221112132072449,
+      "learning_rate": 4.997285763199707e-06,
+      "loss": 0.7679,
+      "step": 352
+    },
+    {
+      "epoch": 0.12394662921348315,
+      "grad_norm": 0.940747857093811,
+      "learning_rate": 4.997264182834179e-06,
+      "loss": 0.7863,
+      "step": 353
+    },
+    {
+      "epoch": 0.12429775280898876,
+      "grad_norm": 0.925940215587616,
+      "learning_rate": 4.997242517064351e-06,
+      "loss": 0.7711,
+      "step": 354
+    },
+    {
+      "epoch": 0.12464887640449439,
+      "grad_norm": 0.8775484561920166,
+      "learning_rate": 4.997220765890967e-06,
+      "loss": 0.7504,
+      "step": 355
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.9700306057929993,
+      "learning_rate": 4.997198929314769e-06,
+      "loss": 0.7385,
+      "step": 356
+    },
+    {
+      "epoch": 0.12535112359550563,
+      "grad_norm": 0.9070924520492554,
+      "learning_rate": 4.997177007336505e-06,
+      "loss": 0.7538,
+      "step": 357
+    },
+    {
+      "epoch": 0.12570224719101122,
+      "grad_norm": 0.8817124962806702,
+      "learning_rate": 4.997154999956924e-06,
+      "loss": 0.7549,
+      "step": 358
+    },
+    {
+      "epoch": 0.12605337078651685,
+      "grad_norm": 0.9093615412712097,
+      "learning_rate": 4.997132907176779e-06,
+      "loss": 0.7621,
+      "step": 359
+    },
+    {
+      "epoch": 0.12640449438202248,
+      "grad_norm": 0.8783479332923889,
+      "learning_rate": 4.9971107289968256e-06,
+      "loss": 0.7369,
+      "step": 360
+    },
+    {
+      "epoch": 0.1267556179775281,
+      "grad_norm": 0.9039416313171387,
+      "learning_rate": 4.997088465417822e-06,
+      "loss": 0.7664,
+      "step": 361
+    },
+    {
+      "epoch": 0.1271067415730337,
+      "grad_norm": 0.9135423302650452,
+      "learning_rate": 4.997066116440529e-06,
+      "loss": 0.7887,
+      "step": 362
+    },
+    {
+      "epoch": 0.12745786516853932,
+      "grad_norm": 0.8956524133682251,
+      "learning_rate": 4.997043682065712e-06,
+      "loss": 0.7303,
+      "step": 363
+    },
+    {
+      "epoch": 0.12780898876404495,
+      "grad_norm": 0.9221348166465759,
+      "learning_rate": 4.997021162294138e-06,
+      "loss": 0.7707,
+      "step": 364
+    },
+    {
+      "epoch": 0.12816011235955055,
+      "grad_norm": 0.920876681804657,
+      "learning_rate": 4.996998557126577e-06,
+      "loss": 0.7825,
+      "step": 365
+    },
+    {
+      "epoch": 0.12851123595505617,
+      "grad_norm": 0.9243044853210449,
+      "learning_rate": 4.996975866563802e-06,
+      "loss": 0.7362,
+      "step": 366
+    },
+    {
+      "epoch": 0.1288623595505618,
+      "grad_norm": 0.89571613073349,
+      "learning_rate": 4.996953090606589e-06,
+      "loss": 0.7645,
+      "step": 367
+    },
+    {
+      "epoch": 0.12921348314606743,
+      "grad_norm": 0.9237456917762756,
+      "learning_rate": 4.9969302292557165e-06,
+      "loss": 0.766,
+      "step": 368
+    },
+    {
+      "epoch": 0.12956460674157302,
+      "grad_norm": 0.9664248824119568,
+      "learning_rate": 4.996907282511968e-06,
+      "loss": 0.7801,
+      "step": 369
+    },
+    {
+      "epoch": 0.12991573033707865,
+      "grad_norm": 0.9590004682540894,
+      "learning_rate": 4.9968842503761265e-06,
+      "loss": 0.7544,
+      "step": 370
+    },
+    {
+      "epoch": 0.13026685393258428,
+      "grad_norm": 0.9026381969451904,
+      "learning_rate": 4.99686113284898e-06,
+      "loss": 0.7756,
+      "step": 371
+    },
+    {
+      "epoch": 0.1306179775280899,
+      "grad_norm": 0.9528935551643372,
+      "learning_rate": 4.9968379299313195e-06,
+      "loss": 0.7673,
+      "step": 372
+    },
+    {
+      "epoch": 0.1309691011235955,
+      "grad_norm": 0.9700410962104797,
+      "learning_rate": 4.996814641623937e-06,
+      "loss": 0.7746,
+      "step": 373
+    },
+    {
+      "epoch": 0.13132022471910113,
+      "grad_norm": 0.931694746017456,
+      "learning_rate": 4.996791267927632e-06,
+      "loss": 0.8078,
+      "step": 374
+    },
+    {
+      "epoch": 0.13167134831460675,
+      "grad_norm": 0.9812837839126587,
+      "learning_rate": 4.996767808843203e-06,
+      "loss": 0.7662,
+      "step": 375
+    },
+    {
+      "epoch": 0.13202247191011235,
+      "grad_norm": 0.9266939163208008,
+      "learning_rate": 4.996744264371449e-06,
+      "loss": 0.7468,
+      "step": 376
+    },
+    {
+      "epoch": 0.13237359550561797,
+      "grad_norm": 0.8732329607009888,
+      "learning_rate": 4.996720634513179e-06,
+      "loss": 0.7774,
+      "step": 377
+    },
+    {
+      "epoch": 0.1327247191011236,
+      "grad_norm": 0.9104029536247253,
+      "learning_rate": 4.9966969192692e-06,
+      "loss": 0.7423,
+      "step": 378
+    },
+    {
+      "epoch": 0.13307584269662923,
+      "grad_norm": 0.8799782991409302,
+      "learning_rate": 4.996673118640323e-06,
+      "loss": 0.7086,
+      "step": 379
+    },
+    {
+      "epoch": 0.13342696629213482,
+      "grad_norm": 0.9318313598632812,
+      "learning_rate": 4.9966492326273605e-06,
+      "loss": 0.7492,
+      "step": 380
+    },
+    {
+      "epoch": 0.13377808988764045,
+      "grad_norm": 0.9577075839042664,
+      "learning_rate": 4.996625261231131e-06,
+      "loss": 0.7935,
+      "step": 381
+    },
+    {
+      "epoch": 0.13412921348314608,
+      "grad_norm": 0.9912057518959045,
+      "learning_rate": 4.996601204452455e-06,
+      "loss": 0.7515,
+      "step": 382
+    },
+    {
+      "epoch": 0.13448033707865167,
+      "grad_norm": 0.9166970252990723,
+      "learning_rate": 4.996577062292154e-06,
+      "loss": 0.7364,
+      "step": 383
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 0.9245953559875488,
+      "learning_rate": 4.996552834751053e-06,
+      "loss": 0.7895,
+      "step": 384
+    },
+    {
+      "epoch": 0.13518258426966293,
+      "grad_norm": 0.9169378876686096,
+      "learning_rate": 4.996528521829982e-06,
+      "loss": 0.7632,
+      "step": 385
+    },
+    {
+      "epoch": 0.13553370786516855,
+      "grad_norm": 0.8847085237503052,
+      "learning_rate": 4.996504123529772e-06,
+      "loss": 0.7607,
+      "step": 386
+    },
+    {
+      "epoch": 0.13588483146067415,
+      "grad_norm": 0.8744500875473022,
+      "learning_rate": 4.996479639851256e-06,
+      "loss": 0.7146,
+      "step": 387
+    },
+    {
+      "epoch": 0.13623595505617977,
+      "grad_norm": 0.9195184707641602,
+      "learning_rate": 4.996455070795274e-06,
+      "loss": 0.771,
+      "step": 388
+    },
+    {
+      "epoch": 0.1365870786516854,
+      "grad_norm": 0.8896746039390564,
+      "learning_rate": 4.996430416362664e-06,
+      "loss": 0.7239,
+      "step": 389
+    },
+    {
+      "epoch": 0.136938202247191,
+      "grad_norm": 0.9265782237052917,
+      "learning_rate": 4.99640567655427e-06,
+      "loss": 0.7607,
+      "step": 390
+    },
+    {
+      "epoch": 0.13728932584269662,
+      "grad_norm": 0.9531838893890381,
+      "learning_rate": 4.996380851370939e-06,
+      "loss": 0.7487,
+      "step": 391
+    },
+    {
+      "epoch": 0.13764044943820225,
+      "grad_norm": 0.9503312706947327,
+      "learning_rate": 4.996355940813518e-06,
+      "loss": 0.7889,
+      "step": 392
+    },
+    {
+      "epoch": 0.13799157303370788,
+      "grad_norm": 0.8666608929634094,
+      "learning_rate": 4.99633094488286e-06,
+      "loss": 0.7539,
+      "step": 393
+    },
+    {
+      "epoch": 0.13834269662921347,
+      "grad_norm": 0.9313355684280396,
+      "learning_rate": 4.996305863579821e-06,
+      "loss": 0.7674,
+      "step": 394
+    },
+    {
+      "epoch": 0.1386938202247191,
+      "grad_norm": 0.9345743656158447,
+      "learning_rate": 4.9962806969052565e-06,
+      "loss": 0.8057,
+      "step": 395
+    },
+    {
+      "epoch": 0.13904494382022473,
+      "grad_norm": 0.922758936882019,
+      "learning_rate": 4.996255444860029e-06,
+      "loss": 0.7404,
+      "step": 396
+    },
+    {
+      "epoch": 0.13939606741573032,
+      "grad_norm": 0.8888351321220398,
+      "learning_rate": 4.996230107445001e-06,
+      "loss": 0.753,
+      "step": 397
+    },
+    {
+      "epoch": 0.13974719101123595,
+      "grad_norm": 0.9118607640266418,
+      "learning_rate": 4.996204684661039e-06,
+      "loss": 0.7562,
+      "step": 398
+    },
+    {
+      "epoch": 0.14009831460674158,
+      "grad_norm": 0.9427488446235657,
+      "learning_rate": 4.996179176509013e-06,
+      "loss": 0.7085,
+      "step": 399
+    },
+    {
+      "epoch": 0.1404494382022472,
+      "grad_norm": 0.9284588098526001,
+      "learning_rate": 4.996153582989795e-06,
+      "loss": 0.7674,
+      "step": 400
+    },
+    {
+      "epoch": 0.1408005617977528,
+      "grad_norm": 0.8856985569000244,
+      "learning_rate": 4.996127904104261e-06,
+      "loss": 0.7357,
+      "step": 401
+    },
+    {
+      "epoch": 0.14115168539325842,
+      "grad_norm": 0.9342722296714783,
+      "learning_rate": 4.996102139853289e-06,
+      "loss": 0.7387,
+      "step": 402
+    },
+    {
+      "epoch": 0.14150280898876405,
+      "grad_norm": 0.9672777056694031,
+      "learning_rate": 4.996076290237759e-06,
+      "loss": 0.7681,
+      "step": 403
+    },
+    {
+      "epoch": 0.14185393258426968,
+      "grad_norm": 0.929104745388031,
+      "learning_rate": 4.996050355258556e-06,
+      "loss": 0.7266,
+      "step": 404
+    },
+    {
+      "epoch": 0.14220505617977527,
+      "grad_norm": 0.9371368288993835,
+      "learning_rate": 4.996024334916567e-06,
+      "loss": 0.7841,
+      "step": 405
+    },
+    {
+      "epoch": 0.1425561797752809,
+      "grad_norm": 0.8848541975021362,
+      "learning_rate": 4.995998229212681e-06,
+      "loss": 0.7659,
+      "step": 406
+    },
+    {
+      "epoch": 0.14290730337078653,
+      "grad_norm": 0.9632852673530579,
+      "learning_rate": 4.995972038147792e-06,
+      "loss": 0.7871,
+      "step": 407
+    },
+    {
+      "epoch": 0.14325842696629212,
+      "grad_norm": 0.9157865047454834,
+      "learning_rate": 4.9959457617227946e-06,
+      "loss": 0.7529,
+      "step": 408
+    },
+    {
+      "epoch": 0.14360955056179775,
+      "grad_norm": 0.9337014555931091,
+      "learning_rate": 4.995919399938588e-06,
+      "loss": 0.7427,
+      "step": 409
+    },
+    {
+      "epoch": 0.14396067415730338,
+      "grad_norm": 0.9515252113342285,
+      "learning_rate": 4.995892952796074e-06,
+      "loss": 0.7577,
+      "step": 410
+    },
+    {
+      "epoch": 0.144311797752809,
+      "grad_norm": 0.9301478266716003,
+      "learning_rate": 4.995866420296157e-06,
+      "loss": 0.7825,
+      "step": 411
+    },
+    {
+      "epoch": 0.1446629213483146,
+      "grad_norm": 1.4020230770111084,
+      "learning_rate": 4.995839802439745e-06,
+      "loss": 0.7434,
+      "step": 412
+    },
+    {
+      "epoch": 0.14501404494382023,
+      "grad_norm": 0.9150766730308533,
+      "learning_rate": 4.995813099227745e-06,
+      "loss": 0.7038,
+      "step": 413
+    },
+    {
+      "epoch": 0.14536516853932585,
+      "grad_norm": 0.9467854499816895,
+      "learning_rate": 4.995786310661075e-06,
+      "loss": 0.7721,
+      "step": 414
+    },
+    {
+      "epoch": 0.14571629213483145,
+      "grad_norm": 0.8949883580207825,
+      "learning_rate": 4.995759436740648e-06,
+      "loss": 0.7174,
+      "step": 415
+    },
+    {
+      "epoch": 0.14606741573033707,
+      "grad_norm": 0.9632487893104553,
+      "learning_rate": 4.995732477467383e-06,
+      "loss": 0.8002,
+      "step": 416
+    },
+    {
+      "epoch": 0.1464185393258427,
+      "grad_norm": 0.8924590945243835,
+      "learning_rate": 4.995705432842204e-06,
+      "loss": 0.7575,
+      "step": 417
+    },
+    {
+      "epoch": 0.14676966292134833,
+      "grad_norm": 0.8902977705001831,
+      "learning_rate": 4.995678302866035e-06,
+      "loss": 0.73,
+      "step": 418
+    },
+    {
+      "epoch": 0.14712078651685392,
+      "grad_norm": 0.9031338095664978,
+      "learning_rate": 4.9956510875398035e-06,
+      "loss": 0.7697,
+      "step": 419
+    },
+    {
+      "epoch": 0.14747191011235955,
+      "grad_norm": 0.9241436719894409,
+      "learning_rate": 4.99562378686444e-06,
+      "loss": 0.7594,
+      "step": 420
+    },
+    {
+      "epoch": 0.14782303370786518,
+      "grad_norm": 1.0044499635696411,
+      "learning_rate": 4.995596400840879e-06,
+      "loss": 0.7938,
+      "step": 421
+    },
+    {
+      "epoch": 0.14817415730337077,
+      "grad_norm": 0.9393342733383179,
+      "learning_rate": 4.995568929470055e-06,
+      "loss": 0.8142,
+      "step": 422
+    },
+    {
+      "epoch": 0.1485252808988764,
+      "grad_norm": 0.9527561664581299,
+      "learning_rate": 4.99554137275291e-06,
+      "loss": 0.7825,
+      "step": 423
+    },
+    {
+      "epoch": 0.14887640449438203,
+      "grad_norm": 0.9195716381072998,
+      "learning_rate": 4.995513730690386e-06,
+      "loss": 0.7577,
+      "step": 424
+    },
+    {
+      "epoch": 0.14922752808988765,
+      "grad_norm": 0.9349954128265381,
+      "learning_rate": 4.995486003283428e-06,
+      "loss": 0.7159,
+      "step": 425
+    },
+    {
+      "epoch": 0.14957865168539325,
+      "grad_norm": 0.922205924987793,
+      "learning_rate": 4.9954581905329845e-06,
+      "loss": 0.8017,
+      "step": 426
+    },
+    {
+      "epoch": 0.14992977528089887,
+      "grad_norm": 0.955886721611023,
+      "learning_rate": 4.995430292440005e-06,
+      "loss": 0.7813,
+      "step": 427
+    },
+    {
+      "epoch": 0.1502808988764045,
+      "grad_norm": 0.9729349613189697,
+      "learning_rate": 4.995402309005445e-06,
+      "loss": 0.7744,
+      "step": 428
+    },
+    {
+      "epoch": 0.1506320224719101,
+      "grad_norm": 0.8943726420402527,
+      "learning_rate": 4.995374240230262e-06,
+      "loss": 0.775,
+      "step": 429
+    },
+    {
+      "epoch": 0.15098314606741572,
+      "grad_norm": 0.9235265254974365,
+      "learning_rate": 4.995346086115416e-06,
+      "loss": 0.759,
+      "step": 430
+    },
+    {
+      "epoch": 0.15133426966292135,
+      "grad_norm": 0.9518915414810181,
+      "learning_rate": 4.995317846661868e-06,
+      "loss": 0.7563,
+      "step": 431
+    },
+    {
+      "epoch": 0.15168539325842698,
+      "grad_norm": 0.9186070561408997,
+      "learning_rate": 4.9952895218705845e-06,
+      "loss": 0.7551,
+      "step": 432
+    },
+    {
+      "epoch": 0.15203651685393257,
+      "grad_norm": 0.9712167978286743,
+      "learning_rate": 4.995261111742536e-06,
+      "loss": 0.7765,
+      "step": 433
+    },
+    {
+      "epoch": 0.1523876404494382,
+      "grad_norm": 0.9043035507202148,
+      "learning_rate": 4.995232616278691e-06,
+      "loss": 0.7224,
+      "step": 434
+    },
+    {
+      "epoch": 0.15273876404494383,
+      "grad_norm": 1.357176661491394,
+      "learning_rate": 4.995204035480027e-06,
+      "loss": 0.7538,
+      "step": 435
+    },
+    {
+      "epoch": 0.15308988764044945,
+      "grad_norm": 0.9199062585830688,
+      "learning_rate": 4.995175369347521e-06,
+      "loss": 0.7396,
+      "step": 436
+    },
+    {
+      "epoch": 0.15344101123595505,
+      "grad_norm": 0.9276919960975647,
+      "learning_rate": 4.995146617882151e-06,
+      "loss": 0.7626,
+      "step": 437
+    },
+    {
+      "epoch": 0.15379213483146068,
+      "grad_norm": 0.8831456303596497,
+      "learning_rate": 4.995117781084901e-06,
+      "loss": 0.7692,
+      "step": 438
+    },
+    {
+      "epoch": 0.1541432584269663,
+      "grad_norm": 0.9020641446113586,
+      "learning_rate": 4.99508885895676e-06,
+      "loss": 0.7578,
+      "step": 439
+    },
+    {
+      "epoch": 0.1544943820224719,
+      "grad_norm": 0.9140202403068542,
+      "learning_rate": 4.995059851498714e-06,
+      "loss": 0.7728,
+      "step": 440
+    },
+    {
+      "epoch": 0.15484550561797752,
+      "grad_norm": 0.9247411489486694,
+      "learning_rate": 4.995030758711756e-06,
+      "loss": 0.7397,
+      "step": 441
+    },
+    {
+      "epoch": 0.15519662921348315,
+      "grad_norm": 0.8848091959953308,
+      "learning_rate": 4.995001580596882e-06,
+      "loss": 0.6819,
+      "step": 442
+    },
+    {
+      "epoch": 0.15554775280898878,
+      "grad_norm": 1.0068150758743286,
+      "learning_rate": 4.994972317155088e-06,
+      "loss": 0.7437,
+      "step": 443
+    },
+    {
+      "epoch": 0.15589887640449437,
+      "grad_norm": 0.9932318329811096,
+      "learning_rate": 4.994942968387376e-06,
+      "loss": 0.7693,
+      "step": 444
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 0.920421838760376,
+      "learning_rate": 4.99491353429475e-06,
+      "loss": 0.7649,
+      "step": 445
+    },
+    {
+      "epoch": 0.15660112359550563,
+      "grad_norm": 0.9176994562149048,
+      "learning_rate": 4.9948840148782165e-06,
+      "loss": 0.7391,
+      "step": 446
+    },
+    {
+      "epoch": 0.15695224719101122,
+      "grad_norm": 0.9588678479194641,
+      "learning_rate": 4.994854410138783e-06,
+      "loss": 0.7615,
+      "step": 447
+    },
+    {
+      "epoch": 0.15730337078651685,
+      "grad_norm": 0.9149226546287537,
+      "learning_rate": 4.9948247200774645e-06,
+      "loss": 0.75,
+      "step": 448
+    },
+    {
+      "epoch": 0.15765449438202248,
+      "grad_norm": 0.8968827724456787,
+      "learning_rate": 4.994794944695275e-06,
+      "loss": 0.7582,
+      "step": 449
+    },
+    {
+      "epoch": 0.1580056179775281,
+      "grad_norm": 0.8770960569381714,
+      "learning_rate": 4.994765083993234e-06,
+      "loss": 0.7121,
+      "step": 450
+    },
+    {
+      "epoch": 0.1583567415730337,
+      "grad_norm": 0.9019699692726135,
+      "learning_rate": 4.9947351379723605e-06,
+      "loss": 0.7131,
+      "step": 451
+    },
+    {
+      "epoch": 0.15870786516853932,
+      "grad_norm": 0.8884302973747253,
+      "learning_rate": 4.994705106633682e-06,
+      "loss": 0.7292,
+      "step": 452
+    },
+    {
+      "epoch": 0.15905898876404495,
+      "grad_norm": 0.9590429663658142,
+      "learning_rate": 4.994674989978222e-06,
+      "loss": 0.7454,
+      "step": 453
+    },
+    {
+      "epoch": 0.15941011235955055,
+      "grad_norm": 0.8948072195053101,
+      "learning_rate": 4.9946447880070124e-06,
+      "loss": 0.7667,
+      "step": 454
+    },
+    {
+      "epoch": 0.15976123595505617,
+      "grad_norm": 0.9360845685005188,
+      "learning_rate": 4.994614500721086e-06,
+      "loss": 0.7633,
+      "step": 455
+    },
+    {
+      "epoch": 0.1601123595505618,
+      "grad_norm": 0.9061880111694336,
+      "learning_rate": 4.9945841281214785e-06,
+      "loss": 0.7026,
+      "step": 456
+    },
+    {
+      "epoch": 0.16046348314606743,
+      "grad_norm": 0.918328583240509,
+      "learning_rate": 4.994553670209228e-06,
+      "loss": 0.7367,
+      "step": 457
+    },
+    {
+      "epoch": 0.16081460674157302,
+      "grad_norm": 0.9201732277870178,
+      "learning_rate": 4.994523126985377e-06,
+      "loss": 0.7326,
+      "step": 458
+    },
+    {
+      "epoch": 0.16116573033707865,
+      "grad_norm": 0.9703920483589172,
+      "learning_rate": 4.9944924984509695e-06,
+      "loss": 0.7638,
+      "step": 459
+    },
+    {
+      "epoch": 0.16151685393258428,
+      "grad_norm": 0.8904596567153931,
+      "learning_rate": 4.994461784607053e-06,
+      "loss": 0.7717,
+      "step": 460
+    },
+    {
+      "epoch": 0.1618679775280899,
+      "grad_norm": 0.9822039604187012,
+      "learning_rate": 4.994430985454678e-06,
+      "loss": 0.7674,
+      "step": 461
+    },
+    {
+      "epoch": 0.1622191011235955,
+      "grad_norm": 0.9268808364868164,
+      "learning_rate": 4.9944001009948986e-06,
+      "loss": 0.7725,
+      "step": 462
+    },
+    {
+      "epoch": 0.16257022471910113,
+      "grad_norm": 0.8941001296043396,
+      "learning_rate": 4.994369131228769e-06,
+      "loss": 0.7361,
+      "step": 463
+    },
+    {
+      "epoch": 0.16292134831460675,
+      "grad_norm": 0.8854597210884094,
+      "learning_rate": 4.99433807615735e-06,
+      "loss": 0.7426,
+      "step": 464
+    },
+    {
+      "epoch": 0.16327247191011235,
+      "grad_norm": 1.0173115730285645,
+      "learning_rate": 4.994306935781704e-06,
+      "loss": 0.7724,
+      "step": 465
+    },
+    {
+      "epoch": 0.16362359550561797,
+      "grad_norm": 0.9098212122917175,
+      "learning_rate": 4.994275710102894e-06,
+      "loss": 0.6989,
+      "step": 466
+    },
+    {
+      "epoch": 0.1639747191011236,
+      "grad_norm": 0.9020840525627136,
+      "learning_rate": 4.99424439912199e-06,
+      "loss": 0.7312,
+      "step": 467
+    },
+    {
+      "epoch": 0.16432584269662923,
+      "grad_norm": 0.9145157337188721,
+      "learning_rate": 4.994213002840062e-06,
+      "loss": 0.7334,
+      "step": 468
+    },
+    {
+      "epoch": 0.16467696629213482,
+      "grad_norm": 0.9347027540206909,
+      "learning_rate": 4.994181521258183e-06,
+      "loss": 0.7508,
+      "step": 469
+    },
+    {
+      "epoch": 0.16502808988764045,
+      "grad_norm": 0.9202942252159119,
+      "learning_rate": 4.994149954377431e-06,
+      "loss": 0.7351,
+      "step": 470
+    },
+    {
+      "epoch": 0.16537921348314608,
+      "grad_norm": 0.9013196229934692,
+      "learning_rate": 4.994118302198884e-06,
+      "loss": 0.7618,
+      "step": 471
+    },
+    {
+      "epoch": 0.16573033707865167,
+      "grad_norm": 0.9281770586967468,
+      "learning_rate": 4.9940865647236255e-06,
+      "loss": 0.7661,
+      "step": 472
+    },
+    {
+      "epoch": 0.1660814606741573,
+      "grad_norm": 0.9201061129570007,
+      "learning_rate": 4.99405474195274e-06,
+      "loss": 0.7203,
+      "step": 473
+    },
+    {
+      "epoch": 0.16643258426966293,
+      "grad_norm": 0.9211746454238892,
+      "learning_rate": 4.994022833887318e-06,
+      "loss": 0.724,
+      "step": 474
+    },
+    {
+      "epoch": 0.16678370786516855,
+      "grad_norm": 0.9289520978927612,
+      "learning_rate": 4.993990840528448e-06,
+      "loss": 0.7907,
+      "step": 475
+    },
+    {
+      "epoch": 0.16713483146067415,
+      "grad_norm": 0.9343749284744263,
+      "learning_rate": 4.993958761877227e-06,
+      "loss": 0.7707,
+      "step": 476
+    },
+    {
+      "epoch": 0.16748595505617977,
+      "grad_norm": 0.9140953421592712,
+      "learning_rate": 4.9939265979347495e-06,
+      "loss": 0.7558,
+      "step": 477
+    },
+    {
+      "epoch": 0.1678370786516854,
+      "grad_norm": 0.8889614939689636,
+      "learning_rate": 4.993894348702117e-06,
+      "loss": 0.7615,
+      "step": 478
+    },
+    {
+      "epoch": 0.168188202247191,
+      "grad_norm": 0.8894888162612915,
+      "learning_rate": 4.993862014180431e-06,
+      "loss": 0.7536,
+      "step": 479
+    },
+    {
+      "epoch": 0.16853932584269662,
+      "grad_norm": 0.8904688954353333,
+      "learning_rate": 4.993829594370798e-06,
+      "loss": 0.7471,
+      "step": 480
+    },
+    {
+      "epoch": 0.16889044943820225,
+      "grad_norm": 0.9064297080039978,
+      "learning_rate": 4.993797089274327e-06,
+      "loss": 0.7524,
+      "step": 481
+    },
+    {
+      "epoch": 0.16924157303370788,
+      "grad_norm": 0.8890877962112427,
+      "learning_rate": 4.99376449889213e-06,
+      "loss": 0.7727,
+      "step": 482
+    },
+    {
+      "epoch": 0.16959269662921347,
+      "grad_norm": 0.9089619517326355,
+      "learning_rate": 4.993731823225321e-06,
+      "loss": 0.7661,
+      "step": 483
+    },
+    {
+      "epoch": 0.1699438202247191,
+      "grad_norm": 0.9284274578094482,
+      "learning_rate": 4.993699062275017e-06,
+      "loss": 0.7337,
+      "step": 484
+    },
+    {
+      "epoch": 0.17029494382022473,
+      "grad_norm": 0.9275715947151184,
+      "learning_rate": 4.99366621604234e-06,
+      "loss": 0.7417,
+      "step": 485
+    },
+    {
+      "epoch": 0.17064606741573032,
+      "grad_norm": 0.9078076481819153,
+      "learning_rate": 4.9936332845284105e-06,
+      "loss": 0.7366,
+      "step": 486
+    },
+    {
+      "epoch": 0.17099719101123595,
+      "grad_norm": 0.9106430411338806,
+      "learning_rate": 4.993600267734359e-06,
+      "loss": 0.7738,
+      "step": 487
+    },
+    {
+      "epoch": 0.17134831460674158,
+      "grad_norm": 1.0632219314575195,
+      "learning_rate": 4.99356716566131e-06,
+      "loss": 0.724,
+      "step": 488
+    },
+    {
+      "epoch": 0.1716994382022472,
+      "grad_norm": 0.9347037076950073,
+      "learning_rate": 4.9935339783103985e-06,
+      "loss": 0.7273,
+      "step": 489
+    },
+    {
+      "epoch": 0.1720505617977528,
+      "grad_norm": 0.9145859479904175,
+      "learning_rate": 4.9935007056827586e-06,
+      "loss": 0.7603,
+      "step": 490
+    },
+    {
+      "epoch": 0.17240168539325842,
+      "grad_norm": 0.924089252948761,
+      "learning_rate": 4.993467347779529e-06,
+      "loss": 0.7289,
+      "step": 491
+    },
+    {
+      "epoch": 0.17275280898876405,
+      "grad_norm": 0.9254995584487915,
+      "learning_rate": 4.993433904601849e-06,
+      "loss": 0.7614,
+      "step": 492
+    },
+    {
+      "epoch": 0.17310393258426968,
+      "grad_norm": 0.9493343234062195,
+      "learning_rate": 4.993400376150863e-06,
+      "loss": 0.7889,
+      "step": 493
+    },
+    {
+      "epoch": 0.17345505617977527,
+      "grad_norm": 0.8984596729278564,
+      "learning_rate": 4.9933667624277184e-06,
+      "loss": 0.7304,
+      "step": 494
+    },
+    {
+      "epoch": 0.1738061797752809,
+      "grad_norm": 0.9199086427688599,
+      "learning_rate": 4.9933330634335634e-06,
+      "loss": 0.7386,
+      "step": 495
+    },
+    {
+      "epoch": 0.17415730337078653,
+      "grad_norm": 0.9047686457633972,
+      "learning_rate": 4.993299279169552e-06,
+      "loss": 0.7469,
+      "step": 496
+    },
+    {
+      "epoch": 0.17450842696629212,
+      "grad_norm": 0.9360693097114563,
+      "learning_rate": 4.993265409636839e-06,
+      "loss": 0.7595,
+      "step": 497
+    },
+    {
+      "epoch": 0.17485955056179775,
+      "grad_norm": 0.9289848804473877,
+      "learning_rate": 4.993231454836581e-06,
+      "loss": 0.7701,
+      "step": 498
+    },
+    {
+      "epoch": 0.17521067415730338,
+      "grad_norm": 0.8970044255256653,
+      "learning_rate": 4.993197414769942e-06,
+      "loss": 0.7191,
+      "step": 499
+    },
+    {
+      "epoch": 0.175561797752809,
+      "grad_norm": 0.92835932970047,
+      "learning_rate": 4.993163289438085e-06,
+      "loss": 0.7395,
+      "step": 500
+    },
+    {
+      "epoch": 0.1759129213483146,
+      "grad_norm": 0.9345782995223999,
+      "learning_rate": 4.993129078842176e-06,
+      "loss": 0.8146,
+      "step": 501
+    },
+    {
+      "epoch": 0.17626404494382023,
+      "grad_norm": 0.9650789499282837,
+      "learning_rate": 4.993094782983386e-06,
+      "loss": 0.7634,
+      "step": 502
+    },
+    {
+      "epoch": 0.17661516853932585,
+      "grad_norm": 0.9201986193656921,
+      "learning_rate": 4.993060401862888e-06,
+      "loss": 0.7476,
+      "step": 503
+    },
+    {
+      "epoch": 0.17696629213483145,
+      "grad_norm": 0.964146077632904,
+      "learning_rate": 4.993025935481858e-06,
+      "loss": 0.7822,
+      "step": 504
+    },
+    {
+      "epoch": 0.17731741573033707,
+      "grad_norm": 0.8705427646636963,
+      "learning_rate": 4.992991383841475e-06,
+      "loss": 0.6773,
+      "step": 505
+    },
+    {
+      "epoch": 0.1776685393258427,
+      "grad_norm": 0.9599820971488953,
+      "learning_rate": 4.992956746942919e-06,
+      "loss": 0.7516,
+      "step": 506
+    },
+    {
+      "epoch": 0.17801966292134833,
+      "grad_norm": 0.9984298944473267,
+      "learning_rate": 4.992922024787374e-06,
+      "loss": 0.7714,
+      "step": 507
+    },
+    {
+      "epoch": 0.17837078651685392,
+      "grad_norm": 0.9363481402397156,
+      "learning_rate": 4.992887217376032e-06,
+      "loss": 0.784,
+      "step": 508
+    },
+    {
+      "epoch": 0.17872191011235955,
+      "grad_norm": 0.8900370597839355,
+      "learning_rate": 4.992852324710078e-06,
+      "loss": 0.7749,
+      "step": 509
+    },
+    {
+      "epoch": 0.17907303370786518,
+      "grad_norm": 0.9664030075073242,
+      "learning_rate": 4.9928173467907085e-06,
+      "loss": 0.7635,
+      "step": 510
+    },
+    {
+      "epoch": 0.17942415730337077,
+      "grad_norm": 0.9362422227859497,
+      "learning_rate": 4.9927822836191185e-06,
+      "loss": 0.7497,
+      "step": 511
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 0.8915205001831055,
+      "learning_rate": 4.992747135196508e-06,
+      "loss": 0.7257,
+      "step": 512
+    },
+    {
+      "epoch": 0.18012640449438203,
+      "grad_norm": 0.9136397838592529,
+      "learning_rate": 4.992711901524079e-06,
+      "loss": 0.733,
+      "step": 513
+    },
+    {
+      "epoch": 0.18047752808988765,
+      "grad_norm": 0.9412914514541626,
+      "learning_rate": 4.992676582603035e-06,
+      "loss": 0.7757,
+      "step": 514
+    },
+    {
+      "epoch": 0.18082865168539325,
+      "grad_norm": 0.8885975480079651,
+      "learning_rate": 4.992641178434586e-06,
+      "loss": 0.7219,
+      "step": 515
+    },
+    {
+      "epoch": 0.18117977528089887,
+      "grad_norm": 0.950423538684845,
+      "learning_rate": 4.992605689019941e-06,
+      "loss": 0.7596,
+      "step": 516
+    },
+    {
+      "epoch": 0.1815308988764045,
+      "grad_norm": 0.9234071969985962,
+      "learning_rate": 4.992570114360314e-06,
+      "loss": 0.7624,
+      "step": 517
+    },
+    {
+      "epoch": 0.1818820224719101,
+      "grad_norm": 0.9259780049324036,
+      "learning_rate": 4.9925344544569225e-06,
+      "loss": 0.7755,
+      "step": 518
+    },
+    {
+      "epoch": 0.18223314606741572,
+      "grad_norm": 0.9503250122070312,
+      "learning_rate": 4.992498709310986e-06,
+      "loss": 0.7051,
+      "step": 519
+    },
+    {
+      "epoch": 0.18258426966292135,
+      "grad_norm": 0.9408488273620605,
+      "learning_rate": 4.992462878923725e-06,
+      "loss": 0.7281,
+      "step": 520
+    },
+    {
+      "epoch": 0.18293539325842698,
+      "grad_norm": 0.9426605105400085,
+      "learning_rate": 4.992426963296369e-06,
+      "loss": 0.7666,
+      "step": 521
+    },
+    {
+      "epoch": 0.18328651685393257,
+      "grad_norm": 0.9386591911315918,
+      "learning_rate": 4.992390962430142e-06,
+      "loss": 0.7352,
+      "step": 522
+    },
+    {
+      "epoch": 0.1836376404494382,
+      "grad_norm": 0.9154810309410095,
+      "learning_rate": 4.992354876326276e-06,
+      "loss": 0.7597,
+      "step": 523
+    },
+    {
+      "epoch": 0.18398876404494383,
+      "grad_norm": 0.936019241809845,
+      "learning_rate": 4.992318704986007e-06,
+      "loss": 0.7497,
+      "step": 524
+    },
+    {
+      "epoch": 0.18433988764044945,
+      "grad_norm": 0.9355201721191406,
+      "learning_rate": 4.99228244841057e-06,
+      "loss": 0.7868,
+      "step": 525
+    },
+    {
+      "epoch": 0.18469101123595505,
+      "grad_norm": 0.9210653305053711,
+      "learning_rate": 4.9922461066012075e-06,
+      "loss": 0.6998,
+      "step": 526
+    },
+    {
+      "epoch": 0.18504213483146068,
+      "grad_norm": 0.9298537969589233,
+      "learning_rate": 4.99220967955916e-06,
+      "loss": 0.7253,
+      "step": 527
+    },
+    {
+      "epoch": 0.1853932584269663,
+      "grad_norm": 0.9049412608146667,
+      "learning_rate": 4.9921731672856746e-06,
+      "loss": 0.7621,
+      "step": 528
+    },
+    {
+      "epoch": 0.1857443820224719,
+      "grad_norm": 0.9317775964736938,
+      "learning_rate": 4.992136569781999e-06,
+      "loss": 0.7393,
+      "step": 529
+    },
+    {
+      "epoch": 0.18609550561797752,
+      "grad_norm": 0.923850417137146,
+      "learning_rate": 4.9920998870493856e-06,
+      "loss": 0.7384,
+      "step": 530
+    },
+    {
+      "epoch": 0.18644662921348315,
+      "grad_norm": 0.9435550570487976,
+      "learning_rate": 4.992063119089088e-06,
+      "loss": 0.7811,
+      "step": 531
+    },
+    {
+      "epoch": 0.18679775280898878,
+      "grad_norm": 0.9249152541160583,
+      "learning_rate": 4.992026265902364e-06,
+      "loss": 0.7451,
+      "step": 532
+    },
+    {
+      "epoch": 0.18714887640449437,
+      "grad_norm": 0.9371122717857361,
+      "learning_rate": 4.991989327490475e-06,
+      "loss": 0.7507,
+      "step": 533
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.8902071714401245,
+      "learning_rate": 4.991952303854683e-06,
+      "loss": 0.7445,
+      "step": 534
+    },
+    {
+      "epoch": 0.18785112359550563,
+      "grad_norm": 0.8958204984664917,
+      "learning_rate": 4.991915194996254e-06,
+      "loss": 0.7029,
+      "step": 535
+    },
+    {
+      "epoch": 0.18820224719101122,
+      "grad_norm": 0.9265782237052917,
+      "learning_rate": 4.991878000916459e-06,
+      "loss": 0.7421,
+      "step": 536
+    },
+    {
+      "epoch": 0.18855337078651685,
+      "grad_norm": 0.8842232823371887,
+      "learning_rate": 4.991840721616568e-06,
+      "loss": 0.7678,
+      "step": 537
+    },
+    {
+      "epoch": 0.18890449438202248,
+      "grad_norm": 0.9105944037437439,
+      "learning_rate": 4.9918033570978565e-06,
+      "loss": 0.7257,
+      "step": 538
+    },
+    {
+      "epoch": 0.1892556179775281,
+      "grad_norm": 0.875582754611969,
+      "learning_rate": 4.991765907361603e-06,
+      "loss": 0.739,
+      "step": 539
+    },
+    {
+      "epoch": 0.1896067415730337,
+      "grad_norm": 0.90901780128479,
+      "learning_rate": 4.991728372409087e-06,
+      "loss": 0.7467,
+      "step": 540
+    },
+    {
+      "epoch": 0.18995786516853932,
+      "grad_norm": 0.9725927710533142,
+      "learning_rate": 4.991690752241594e-06,
+      "loss": 0.7779,
+      "step": 541
+    },
+    {
+      "epoch": 0.19030898876404495,
+      "grad_norm": 0.9324540495872498,
+      "learning_rate": 4.991653046860408e-06,
+      "loss": 0.7337,
+      "step": 542
+    },
+    {
+      "epoch": 0.19066011235955055,
+      "grad_norm": 0.929425060749054,
+      "learning_rate": 4.9916152562668205e-06,
+      "loss": 0.7432,
+      "step": 543
+    },
+    {
+      "epoch": 0.19101123595505617,
+      "grad_norm": 0.9117162227630615,
+      "learning_rate": 4.991577380462124e-06,
+      "loss": 0.7556,
+      "step": 544
+    },
+    {
+      "epoch": 0.1913623595505618,
+      "grad_norm": 0.9356319904327393,
+      "learning_rate": 4.9915394194476115e-06,
+      "loss": 0.783,
+      "step": 545
+    },
+    {
+      "epoch": 0.19171348314606743,
+      "grad_norm": 0.9007960557937622,
+      "learning_rate": 4.991501373224584e-06,
+      "loss": 0.7485,
+      "step": 546
+    },
+    {
+      "epoch": 0.19206460674157302,
+      "grad_norm": 0.9083287715911865,
+      "learning_rate": 4.991463241794342e-06,
+      "loss": 0.7497,
+      "step": 547
+    },
+    {
+      "epoch": 0.19241573033707865,
+      "grad_norm": 0.9226256608963013,
+      "learning_rate": 4.991425025158188e-06,
+      "loss": 0.7687,
+      "step": 548
+    },
+    {
+      "epoch": 0.19276685393258428,
+      "grad_norm": 0.8791014552116394,
+      "learning_rate": 4.991386723317432e-06,
+      "loss": 0.7542,
+      "step": 549
+    },
+    {
+      "epoch": 0.1931179775280899,
+      "grad_norm": 0.8970210552215576,
+      "learning_rate": 4.991348336273381e-06,
+      "loss": 0.7169,
+      "step": 550
+    },
+    {
+      "epoch": 0.1934691011235955,
+      "grad_norm": 0.9191828966140747,
+      "learning_rate": 4.9913098640273475e-06,
+      "loss": 0.7261,
+      "step": 551
+    },
+    {
+      "epoch": 0.19382022471910113,
+      "grad_norm": 0.9275765419006348,
+      "learning_rate": 4.991271306580649e-06,
+      "loss": 0.7469,
+      "step": 552
+    },
+    {
+      "epoch": 0.19417134831460675,
+      "grad_norm": 0.9179903864860535,
+      "learning_rate": 4.9912326639346045e-06,
+      "loss": 0.7385,
+      "step": 553
+    },
+    {
+      "epoch": 0.19452247191011235,
+      "grad_norm": 0.9225823283195496,
+      "learning_rate": 4.991193936090535e-06,
+      "loss": 0.6986,
+      "step": 554
+    },
+    {
+      "epoch": 0.19487359550561797,
+      "grad_norm": 0.9241051077842712,
+      "learning_rate": 4.991155123049764e-06,
+      "loss": 0.783,
+      "step": 555
+    },
+    {
+      "epoch": 0.1952247191011236,
+      "grad_norm": 0.9368206858634949,
+      "learning_rate": 4.991116224813619e-06,
+      "loss": 0.7712,
+      "step": 556
+    },
+    {
+      "epoch": 0.19557584269662923,
+      "grad_norm": 0.9148329496383667,
+      "learning_rate": 4.991077241383431e-06,
+      "loss": 0.8179,
+      "step": 557
+    },
+    {
+      "epoch": 0.19592696629213482,
+      "grad_norm": 0.8920343518257141,
+      "learning_rate": 4.991038172760533e-06,
+      "loss": 0.7271,
+      "step": 558
+    },
+    {
+      "epoch": 0.19627808988764045,
+      "grad_norm": 0.9275470972061157,
+      "learning_rate": 4.990999018946262e-06,
+      "loss": 0.7704,
+      "step": 559
+    },
+    {
+      "epoch": 0.19662921348314608,
+      "grad_norm": 0.9231950640678406,
+      "learning_rate": 4.990959779941955e-06,
+      "loss": 0.6871,
+      "step": 560
+    },
+    {
+      "epoch": 0.19698033707865167,
+      "grad_norm": 0.9179824590682983,
+      "learning_rate": 4.990920455748955e-06,
+      "loss": 0.7573,
+      "step": 561
+    },
+    {
+      "epoch": 0.1973314606741573,
+      "grad_norm": 0.9076899290084839,
+      "learning_rate": 4.9908810463686076e-06,
+      "loss": 0.7463,
+      "step": 562
+    },
+    {
+      "epoch": 0.19768258426966293,
+      "grad_norm": 0.9685149788856506,
+      "learning_rate": 4.99084155180226e-06,
+      "loss": 0.7207,
+      "step": 563
+    },
+    {
+      "epoch": 0.19803370786516855,
+      "grad_norm": 1.3696738481521606,
+      "learning_rate": 4.990801972051262e-06,
+      "loss": 0.768,
+      "step": 564
+    },
+    {
+      "epoch": 0.19838483146067415,
+      "grad_norm": 0.9174782633781433,
+      "learning_rate": 4.990762307116969e-06,
+      "loss": 0.7711,
+      "step": 565
+    },
+    {
+      "epoch": 0.19873595505617977,
+      "grad_norm": 0.9449765086174011,
+      "learning_rate": 4.990722557000736e-06,
+      "loss": 0.7267,
+      "step": 566
+    },
+    {
+      "epoch": 0.1990870786516854,
+      "grad_norm": 0.9291553497314453,
+      "learning_rate": 4.990682721703922e-06,
+      "loss": 0.7017,
+      "step": 567
+    },
+    {
+      "epoch": 0.199438202247191,
+      "grad_norm": 0.920346736907959,
+      "learning_rate": 4.9906428012278915e-06,
+      "loss": 0.7486,
+      "step": 568
+    },
+    {
+      "epoch": 0.19978932584269662,
+      "grad_norm": 0.9206670522689819,
+      "learning_rate": 4.990602795574007e-06,
+      "loss": 0.7833,
+      "step": 569
+    },
+    {
+      "epoch": 0.20014044943820225,
+      "grad_norm": 0.9318204522132874,
+      "learning_rate": 4.99056270474364e-06,
+      "loss": 0.7467,
+      "step": 570
+    },
+    {
+      "epoch": 0.20049157303370788,
+      "grad_norm": 0.9259383082389832,
+      "learning_rate": 4.990522528738159e-06,
+      "loss": 0.7196,
+      "step": 571
+    },
+    {
+      "epoch": 0.20084269662921347,
+      "grad_norm": 0.9336618185043335,
+      "learning_rate": 4.9904822675589385e-06,
+      "loss": 0.7498,
+      "step": 572
+    },
+    {
+      "epoch": 0.2011938202247191,
+      "grad_norm": 0.9120561480522156,
+      "learning_rate": 4.990441921207356e-06,
+      "loss": 0.7621,
+      "step": 573
+    },
+    {
+      "epoch": 0.20154494382022473,
+      "grad_norm": 0.9047495126724243,
+      "learning_rate": 4.99040148968479e-06,
+      "loss": 0.7103,
+      "step": 574
+    },
+    {
+      "epoch": 0.20189606741573032,
+      "grad_norm": 0.9809283018112183,
+      "learning_rate": 4.990360972992625e-06,
+      "loss": 0.7341,
+      "step": 575
+    },
+    {
+      "epoch": 0.20224719101123595,
+      "grad_norm": 1.0163681507110596,
+      "learning_rate": 4.990320371132245e-06,
+      "loss": 0.7219,
+      "step": 576
+    },
+    {
+      "epoch": 0.20259831460674158,
+      "grad_norm": 0.8804516792297363,
+      "learning_rate": 4.99027968410504e-06,
+      "loss": 0.718,
+      "step": 577
+    },
+    {
+      "epoch": 0.2029494382022472,
+      "grad_norm": 0.9172428846359253,
+      "learning_rate": 4.9902389119124005e-06,
+      "loss": 0.7361,
+      "step": 578
+    },
+    {
+      "epoch": 0.2033005617977528,
+      "grad_norm": 0.9390194416046143,
+      "learning_rate": 4.990198054555721e-06,
+      "loss": 0.7195,
+      "step": 579
+    },
+    {
+      "epoch": 0.20365168539325842,
+      "grad_norm": 0.9496874213218689,
+      "learning_rate": 4.990157112036399e-06,
+      "loss": 0.7305,
+      "step": 580
+    },
+    {
+      "epoch": 0.20400280898876405,
+      "grad_norm": 0.9417531490325928,
+      "learning_rate": 4.990116084355835e-06,
+      "loss": 0.7647,
+      "step": 581
+    },
+    {
+      "epoch": 0.20435393258426968,
+      "grad_norm": 0.8786060810089111,
+      "learning_rate": 4.99007497151543e-06,
+      "loss": 0.7654,
+      "step": 582
+    },
+    {
+      "epoch": 0.20470505617977527,
+      "grad_norm": 0.8952715992927551,
+      "learning_rate": 4.990033773516594e-06,
+      "loss": 0.7149,
+      "step": 583
+    },
+    {
+      "epoch": 0.2050561797752809,
+      "grad_norm": 0.9071550369262695,
+      "learning_rate": 4.989992490360734e-06,
+      "loss": 0.7368,
+      "step": 584
+    },
+    {
+      "epoch": 0.20540730337078653,
+      "grad_norm": 0.8937875628471375,
+      "learning_rate": 4.989951122049259e-06,
+      "loss": 0.7469,
+      "step": 585
+    },
+    {
+      "epoch": 0.20575842696629212,
+      "grad_norm": 0.9116646647453308,
+      "learning_rate": 4.989909668583588e-06,
+      "loss": 0.7348,
+      "step": 586
+    },
+    {
+      "epoch": 0.20610955056179775,
+      "grad_norm": 0.9020063877105713,
+      "learning_rate": 4.989868129965137e-06,
+      "loss": 0.6871,
+      "step": 587
+    },
+    {
+      "epoch": 0.20646067415730338,
+      "grad_norm": 0.8991970419883728,
+      "learning_rate": 4.989826506195326e-06,
+      "loss": 0.7136,
+      "step": 588
+    },
+    {
+      "epoch": 0.206811797752809,
+      "grad_norm": 0.9156064987182617,
+      "learning_rate": 4.9897847972755795e-06,
+      "loss": 0.7082,
+      "step": 589
+    },
+    {
+      "epoch": 0.2071629213483146,
+      "grad_norm": 0.9545648097991943,
+      "learning_rate": 4.989743003207323e-06,
+      "loss": 0.7628,
+      "step": 590
+    },
+    {
+      "epoch": 0.20751404494382023,
+      "grad_norm": 0.9077872633934021,
+      "learning_rate": 4.989701123991987e-06,
+      "loss": 0.7595,
+      "step": 591
+    },
+    {
+      "epoch": 0.20786516853932585,
+      "grad_norm": 0.9414958953857422,
+      "learning_rate": 4.989659159631003e-06,
+      "loss": 0.7604,
+      "step": 592
+    },
+    {
+      "epoch": 0.20821629213483145,
+      "grad_norm": 0.9313225150108337,
+      "learning_rate": 4.989617110125805e-06,
+      "loss": 0.7175,
+      "step": 593
+    },
+    {
+      "epoch": 0.20856741573033707,
+      "grad_norm": 0.9247497320175171,
+      "learning_rate": 4.989574975477833e-06,
+      "loss": 0.7497,
+      "step": 594
+    },
+    {
+      "epoch": 0.2089185393258427,
+      "grad_norm": 0.8851250410079956,
+      "learning_rate": 4.989532755688527e-06,
+      "loss": 0.665,
+      "step": 595
+    },
+    {
+      "epoch": 0.20926966292134833,
+      "grad_norm": 0.9276209473609924,
+      "learning_rate": 4.989490450759331e-06,
+      "loss": 0.6919,
+      "step": 596
+    },
+    {
+      "epoch": 0.20962078651685392,
+      "grad_norm": 0.9292905330657959,
+      "learning_rate": 4.9894480606916925e-06,
+      "loss": 0.7545,
+      "step": 597
+    },
+    {
+      "epoch": 0.20997191011235955,
+      "grad_norm": 0.9364456534385681,
+      "learning_rate": 4.98940558548706e-06,
+      "loss": 0.8026,
+      "step": 598
+    },
+    {
+      "epoch": 0.21032303370786518,
+      "grad_norm": 0.9156911969184875,
+      "learning_rate": 4.9893630251468874e-06,
+      "loss": 0.7279,
+      "step": 599
+    },
+    {
+      "epoch": 0.21067415730337077,
+      "grad_norm": 0.9316357374191284,
+      "learning_rate": 4.989320379672629e-06,
+      "loss": 0.7664,
+      "step": 600
+    },
+    {
+      "epoch": 0.2110252808988764,
+      "grad_norm": 0.945326566696167,
+      "learning_rate": 4.989277649065744e-06,
+      "loss": 0.7456,
+      "step": 601
+    },
+    {
+      "epoch": 0.21137640449438203,
+      "grad_norm": 0.9577007293701172,
+      "learning_rate": 4.989234833327693e-06,
+      "loss": 0.7097,
+      "step": 602
+    },
+    {
+      "epoch": 0.21172752808988765,
+      "grad_norm": 0.898679792881012,
+      "learning_rate": 4.989191932459941e-06,
+      "loss": 0.7085,
+      "step": 603
+    },
+    {
+      "epoch": 0.21207865168539325,
+      "grad_norm": 0.9334394931793213,
+      "learning_rate": 4.989148946463955e-06,
+      "loss": 0.7441,
+      "step": 604
+    },
+    {
+      "epoch": 0.21242977528089887,
+      "grad_norm": 0.9490911364555359,
+      "learning_rate": 4.989105875341206e-06,
+      "loss": 0.7571,
+      "step": 605
+    },
+    {
+      "epoch": 0.2127808988764045,
+      "grad_norm": 0.9073877930641174,
+      "learning_rate": 4.989062719093164e-06,
+      "loss": 0.748,
+      "step": 606
+    },
+    {
+      "epoch": 0.2131320224719101,
+      "grad_norm": 0.9007843136787415,
+      "learning_rate": 4.989019477721309e-06,
+      "loss": 0.7099,
+      "step": 607
+    },
+    {
+      "epoch": 0.21348314606741572,
+      "grad_norm": 0.9324343204498291,
+      "learning_rate": 4.988976151227116e-06,
+      "loss": 0.7688,
+      "step": 608
+    },
+    {
+      "epoch": 0.21383426966292135,
+      "grad_norm": 0.9213019013404846,
+      "learning_rate": 4.988932739612069e-06,
+      "loss": 0.7302,
+      "step": 609
+    },
+    {
+      "epoch": 0.21418539325842698,
+      "grad_norm": 0.9618836641311646,
+      "learning_rate": 4.988889242877653e-06,
+      "loss": 0.7484,
+      "step": 610
+    },
+    {
+      "epoch": 0.21453651685393257,
+      "grad_norm": 0.9354860186576843,
+      "learning_rate": 4.988845661025353e-06,
+      "loss": 0.7192,
+      "step": 611
+    },
+    {
+      "epoch": 0.2148876404494382,
+      "grad_norm": 0.9579604268074036,
+      "learning_rate": 4.988801994056663e-06,
+      "loss": 0.7727,
+      "step": 612
+    },
+    {
+      "epoch": 0.21523876404494383,
+      "grad_norm": 0.8972249627113342,
+      "learning_rate": 4.988758241973073e-06,
+      "loss": 0.7816,
+      "step": 613
+    },
+    {
+      "epoch": 0.21558988764044945,
+      "grad_norm": 0.9025394916534424,
+      "learning_rate": 4.988714404776082e-06,
+      "loss": 0.6994,
+      "step": 614
+    },
+    {
+      "epoch": 0.21594101123595505,
+      "grad_norm": 0.9159374833106995,
+      "learning_rate": 4.9886704824671874e-06,
+      "loss": 0.7319,
+      "step": 615
+    },
+    {
+      "epoch": 0.21629213483146068,
+      "grad_norm": 0.9703385829925537,
+      "learning_rate": 4.988626475047892e-06,
+      "loss": 0.7739,
+      "step": 616
+    },
+    {
+      "epoch": 0.2166432584269663,
+      "grad_norm": 0.9332022070884705,
+      "learning_rate": 4.9885823825197e-06,
+      "loss": 0.7209,
+      "step": 617
+    },
+    {
+      "epoch": 0.2169943820224719,
+      "grad_norm": 0.9919439554214478,
+      "learning_rate": 4.988538204884121e-06,
+      "loss": 0.7631,
+      "step": 618
+    },
+    {
+      "epoch": 0.21734550561797752,
+      "grad_norm": 0.9095845818519592,
+      "learning_rate": 4.988493942142664e-06,
+      "loss": 0.7311,
+      "step": 619
+    },
+    {
+      "epoch": 0.21769662921348315,
+      "grad_norm": 0.9250304698944092,
+      "learning_rate": 4.988449594296845e-06,
+      "loss": 0.7616,
+      "step": 620
+    },
+    {
+      "epoch": 0.21804775280898878,
+      "grad_norm": 0.9156744480133057,
+      "learning_rate": 4.988405161348178e-06,
+      "loss": 0.7446,
+      "step": 621
+    },
+    {
+      "epoch": 0.21839887640449437,
+      "grad_norm": 0.9540274143218994,
+      "learning_rate": 4.988360643298184e-06,
+      "loss": 0.745,
+      "step": 622
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 0.9059514999389648,
+      "learning_rate": 4.988316040148386e-06,
+      "loss": 0.7169,
+      "step": 623
+    },
+    {
+      "epoch": 0.21910112359550563,
+      "grad_norm": 0.8970401287078857,
+      "learning_rate": 4.988271351900308e-06,
+      "loss": 0.7551,
+      "step": 624
+    },
+    {
+      "epoch": 0.21945224719101122,
+      "grad_norm": 0.8862085938453674,
+      "learning_rate": 4.988226578555479e-06,
+      "loss": 0.7141,
+      "step": 625
+    },
+    {
+      "epoch": 0.21980337078651685,
+      "grad_norm": 0.9373227953910828,
+      "learning_rate": 4.98818172011543e-06,
+      "loss": 0.7637,
+      "step": 626
+    },
+    {
+      "epoch": 0.22015449438202248,
+      "grad_norm": 0.9211307168006897,
+      "learning_rate": 4.988136776581697e-06,
+      "loss": 0.7428,
+      "step": 627
+    },
+    {
+      "epoch": 0.2205056179775281,
+      "grad_norm": 0.9092309474945068,
+      "learning_rate": 4.9880917479558135e-06,
+      "loss": 0.7341,
+      "step": 628
+    },
+    {
+      "epoch": 0.2208567415730337,
+      "grad_norm": 0.9342793822288513,
+      "learning_rate": 4.988046634239322e-06,
+      "loss": 0.7397,
+      "step": 629
+    },
+    {
+      "epoch": 0.22120786516853932,
+      "grad_norm": 0.9277600646018982,
+      "learning_rate": 4.9880014354337645e-06,
+      "loss": 0.718,
+      "step": 630
+    },
+    {
+      "epoch": 0.22155898876404495,
+      "grad_norm": 1.1010304689407349,
+      "learning_rate": 4.987956151540687e-06,
+      "loss": 0.7199,
+      "step": 631
+    },
+    {
+      "epoch": 0.22191011235955055,
+      "grad_norm": 0.9733719229698181,
+      "learning_rate": 4.9879107825616375e-06,
+      "loss": 0.7518,
+      "step": 632
+    },
+    {
+      "epoch": 0.22226123595505617,
+      "grad_norm": 0.9306866526603699,
+      "learning_rate": 4.9878653284981694e-06,
+      "loss": 0.7388,
+      "step": 633
+    },
+    {
+      "epoch": 0.2226123595505618,
+      "grad_norm": 0.9351629614830017,
+      "learning_rate": 4.987819789351835e-06,
+      "loss": 0.7326,
+      "step": 634
+    },
+    {
+      "epoch": 0.22296348314606743,
+      "grad_norm": 0.9454215168952942,
+      "learning_rate": 4.987774165124194e-06,
+      "loss": 0.7929,
+      "step": 635
+    },
+    {
+      "epoch": 0.22331460674157302,
+      "grad_norm": 0.8752790093421936,
+      "learning_rate": 4.987728455816804e-06,
+      "loss": 0.7282,
+      "step": 636
+    },
+    {
+      "epoch": 0.22366573033707865,
+      "grad_norm": 0.9457923173904419,
+      "learning_rate": 4.98768266143123e-06,
+      "loss": 0.7297,
+      "step": 637
+    },
+    {
+      "epoch": 0.22401685393258428,
+      "grad_norm": 0.9447070360183716,
+      "learning_rate": 4.987636781969038e-06,
+      "loss": 0.7502,
+      "step": 638
+    },
+    {
+      "epoch": 0.2243679775280899,
+      "grad_norm": 1.4757858514785767,
+      "learning_rate": 4.987590817431797e-06,
+      "loss": 0.754,
+      "step": 639
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.9355869889259338,
+      "learning_rate": 4.987544767821078e-06,
+      "loss": 0.7468,
+      "step": 640
+    },
+    {
+      "epoch": 0.22507022471910113,
+      "grad_norm": 0.9021878242492676,
+      "learning_rate": 4.9874986331384566e-06,
+      "loss": 0.6914,
+      "step": 641
+    },
+    {
+      "epoch": 0.22542134831460675,
+      "grad_norm": 0.9627546072006226,
+      "learning_rate": 4.98745241338551e-06,
+      "loss": 0.6998,
+      "step": 642
+    },
+    {
+      "epoch": 0.22577247191011235,
+      "grad_norm": 0.9815264940261841,
+      "learning_rate": 4.98740610856382e-06,
+      "loss": 0.7392,
+      "step": 643
+    },
+    {
+      "epoch": 0.22612359550561797,
+      "grad_norm": 0.9428890943527222,
+      "learning_rate": 4.98735971867497e-06,
+      "loss": 0.737,
+      "step": 644
+    },
+    {
+      "epoch": 0.2264747191011236,
+      "grad_norm": 0.9130154252052307,
+      "learning_rate": 4.987313243720546e-06,
+      "loss": 0.7363,
+      "step": 645
+    },
+    {
+      "epoch": 0.22682584269662923,
+      "grad_norm": 0.9495055079460144,
+      "learning_rate": 4.987266683702137e-06,
+      "loss": 0.7104,
+      "step": 646
+    },
+    {
+      "epoch": 0.22717696629213482,
+      "grad_norm": 0.9111932516098022,
+      "learning_rate": 4.987220038621336e-06,
+      "loss": 0.7107,
+      "step": 647
+    },
+    {
+      "epoch": 0.22752808988764045,
+      "grad_norm": 0.9328299760818481,
+      "learning_rate": 4.987173308479738e-06,
+      "loss": 0.7294,
+      "step": 648
+    },
+    {
+      "epoch": 0.22787921348314608,
+      "grad_norm": 0.9401968717575073,
+      "learning_rate": 4.987126493278942e-06,
+      "loss": 0.7089,
+      "step": 649
+    },
+    {
+      "epoch": 0.22823033707865167,
+      "grad_norm": 0.9041705131530762,
+      "learning_rate": 4.987079593020547e-06,
+      "loss": 0.7249,
+      "step": 650
+    },
+    {
+      "epoch": 0.2285814606741573,
+      "grad_norm": 0.8919897675514221,
+      "learning_rate": 4.987032607706158e-06,
+      "loss": 0.6662,
+      "step": 651
+    },
+    {
+      "epoch": 0.22893258426966293,
+      "grad_norm": 0.9634642004966736,
+      "learning_rate": 4.986985537337384e-06,
+      "loss": 0.7547,
+      "step": 652
+    },
+    {
+      "epoch": 0.22928370786516855,
+      "grad_norm": 0.9561633467674255,
+      "learning_rate": 4.9869383819158305e-06,
+      "loss": 0.7448,
+      "step": 653
+    },
+    {
+      "epoch": 0.22963483146067415,
+      "grad_norm": 0.9113203287124634,
+      "learning_rate": 4.986891141443113e-06,
+      "loss": 0.7064,
+      "step": 654
+    },
+    {
+      "epoch": 0.22998595505617977,
+      "grad_norm": 0.8830083012580872,
+      "learning_rate": 4.986843815920847e-06,
+      "loss": 0.6909,
+      "step": 655
+    },
+    {
+      "epoch": 0.2303370786516854,
+      "grad_norm": 0.9515541791915894,
+      "learning_rate": 4.98679640535065e-06,
+      "loss": 0.7262,
+      "step": 656
+    },
+    {
+      "epoch": 0.230688202247191,
+      "grad_norm": 0.9469254612922668,
+      "learning_rate": 4.986748909734145e-06,
+      "loss": 0.7177,
+      "step": 657
+    },
+    {
+      "epoch": 0.23103932584269662,
+      "grad_norm": 0.9044115543365479,
+      "learning_rate": 4.986701329072954e-06,
+      "loss": 0.7371,
+      "step": 658
+    },
+    {
+      "epoch": 0.23139044943820225,
+      "grad_norm": 0.9210987687110901,
+      "learning_rate": 4.986653663368706e-06,
+      "loss": 0.7327,
+      "step": 659
+    },
+    {
+      "epoch": 0.23174157303370788,
+      "grad_norm": 0.8826014995574951,
+      "learning_rate": 4.98660591262303e-06,
+      "loss": 0.7455,
+      "step": 660
+    },
+    {
+      "epoch": 0.23209269662921347,
+      "grad_norm": 0.8898245096206665,
+      "learning_rate": 4.986558076837561e-06,
+      "loss": 0.7242,
+      "step": 661
+    },
+    {
+      "epoch": 0.2324438202247191,
+      "grad_norm": 0.9125939607620239,
+      "learning_rate": 4.9865101560139324e-06,
+      "loss": 0.7587,
+      "step": 662
+    },
+    {
+      "epoch": 0.23279494382022473,
+      "grad_norm": 0.9783604741096497,
+      "learning_rate": 4.9864621501537845e-06,
+      "loss": 0.7356,
+      "step": 663
+    },
+    {
+      "epoch": 0.23314606741573032,
+      "grad_norm": 0.9296882152557373,
+      "learning_rate": 4.986414059258758e-06,
+      "loss": 0.7059,
+      "step": 664
+    },
+    {
+      "epoch": 0.23349719101123595,
+      "grad_norm": 0.9341956973075867,
+      "learning_rate": 4.9863658833305e-06,
+      "loss": 0.7342,
+      "step": 665
+    },
+    {
+      "epoch": 0.23384831460674158,
+      "grad_norm": 0.9274594783782959,
+      "learning_rate": 4.986317622370655e-06,
+      "loss": 0.7254,
+      "step": 666
+    },
+    {
+      "epoch": 0.2341994382022472,
+      "grad_norm": 0.8959493637084961,
+      "learning_rate": 4.986269276380875e-06,
+      "loss": 0.7138,
+      "step": 667
+    },
+    {
+      "epoch": 0.2345505617977528,
+      "grad_norm": 0.9621686935424805,
+      "learning_rate": 4.986220845362814e-06,
+      "loss": 0.7301,
+      "step": 668
+    },
+    {
+      "epoch": 0.23490168539325842,
+      "grad_norm": 0.9666121602058411,
+      "learning_rate": 4.9861723293181275e-06,
+      "loss": 0.754,
+      "step": 669
+    },
+    {
+      "epoch": 0.23525280898876405,
+      "grad_norm": 0.9423872828483582,
+      "learning_rate": 4.986123728248474e-06,
+      "loss": 0.7464,
+      "step": 670
+    },
+    {
+      "epoch": 0.23560393258426968,
+      "grad_norm": 0.9208926558494568,
+      "learning_rate": 4.986075042155517e-06,
+      "loss": 0.7092,
+      "step": 671
+    },
+    {
+      "epoch": 0.23595505617977527,
+      "grad_norm": 0.880066454410553,
+      "learning_rate": 4.986026271040921e-06,
+      "loss": 0.6935,
+      "step": 672
+    },
+    {
+      "epoch": 0.2363061797752809,
+      "grad_norm": 0.9050423502922058,
+      "learning_rate": 4.985977414906353e-06,
+      "loss": 0.742,
+      "step": 673
+    },
+    {
+      "epoch": 0.23665730337078653,
+      "grad_norm": 0.9089376330375671,
+      "learning_rate": 4.985928473753486e-06,
+      "loss": 0.7366,
+      "step": 674
+    },
+    {
+      "epoch": 0.23700842696629212,
+      "grad_norm": 0.9702712893486023,
+      "learning_rate": 4.985879447583992e-06,
+      "loss": 0.7314,
+      "step": 675
+    },
+    {
+      "epoch": 0.23735955056179775,
+      "grad_norm": 0.9272444248199463,
+      "learning_rate": 4.985830336399548e-06,
+      "loss": 0.7156,
+      "step": 676
+    },
+    {
+      "epoch": 0.23771067415730338,
+      "grad_norm": 0.9623541831970215,
+      "learning_rate": 4.985781140201833e-06,
+      "loss": 0.705,
+      "step": 677
+    },
+    {
+      "epoch": 0.238061797752809,
+      "grad_norm": 0.9228547215461731,
+      "learning_rate": 4.985731858992531e-06,
+      "loss": 0.7254,
+      "step": 678
+    },
+    {
+      "epoch": 0.2384129213483146,
+      "grad_norm": 0.8912785649299622,
+      "learning_rate": 4.985682492773326e-06,
+      "loss": 0.7254,
+      "step": 679
+    },
+    {
+      "epoch": 0.23876404494382023,
+      "grad_norm": 0.9913116693496704,
+      "learning_rate": 4.985633041545906e-06,
+      "loss": 0.7172,
+      "step": 680
+    },
+    {
+      "epoch": 0.23911516853932585,
+      "grad_norm": 0.9851749539375305,
+      "learning_rate": 4.985583505311965e-06,
+      "loss": 0.7711,
+      "step": 681
+    },
+    {
+      "epoch": 0.23946629213483145,
+      "grad_norm": 0.9252983331680298,
+      "learning_rate": 4.9855338840731934e-06,
+      "loss": 0.7356,
+      "step": 682
+    },
+    {
+      "epoch": 0.23981741573033707,
+      "grad_norm": 0.9541698694229126,
+      "learning_rate": 4.9854841778312905e-06,
+      "loss": 0.7111,
+      "step": 683
+    },
+    {
+      "epoch": 0.2401685393258427,
+      "grad_norm": 0.9459705948829651,
+      "learning_rate": 4.985434386587956e-06,
+      "loss": 0.7254,
+      "step": 684
+    },
+    {
+      "epoch": 0.24051966292134833,
+      "grad_norm": 1.0094040632247925,
+      "learning_rate": 4.985384510344892e-06,
+      "loss": 0.7638,
+      "step": 685
+    },
+    {
+      "epoch": 0.24087078651685392,
+      "grad_norm": 0.9937487244606018,
+      "learning_rate": 4.985334549103804e-06,
+      "loss": 0.7373,
+      "step": 686
+    },
+    {
+      "epoch": 0.24122191011235955,
+      "grad_norm": 0.9358577132225037,
+      "learning_rate": 4.985284502866401e-06,
+      "loss": 0.697,
+      "step": 687
+    },
+    {
+      "epoch": 0.24157303370786518,
+      "grad_norm": 0.9751777052879333,
+      "learning_rate": 4.985234371634395e-06,
+      "loss": 0.7626,
+      "step": 688
+    },
+    {
+      "epoch": 0.24192415730337077,
+      "grad_norm": 0.9775291085243225,
+      "learning_rate": 4.9851841554095e-06,
+      "loss": 0.7031,
+      "step": 689
+    },
+    {
+      "epoch": 0.2422752808988764,
+      "grad_norm": 0.9366273880004883,
+      "learning_rate": 4.9851338541934345e-06,
+      "loss": 0.7268,
+      "step": 690
+    },
+    {
+      "epoch": 0.24262640449438203,
+      "grad_norm": 0.8989854454994202,
+      "learning_rate": 4.985083467987917e-06,
+      "loss": 0.7102,
+      "step": 691
+    },
+    {
+      "epoch": 0.24297752808988765,
+      "grad_norm": 0.9326997399330139,
+      "learning_rate": 4.9850329967946715e-06,
+      "loss": 0.6963,
+      "step": 692
+    },
+    {
+      "epoch": 0.24332865168539325,
+      "grad_norm": 0.9004776477813721,
+      "learning_rate": 4.984982440615423e-06,
+      "loss": 0.7133,
+      "step": 693
+    },
+    {
+      "epoch": 0.24367977528089887,
+      "grad_norm": 0.915814220905304,
+      "learning_rate": 4.984931799451903e-06,
+      "loss": 0.7197,
+      "step": 694
+    },
+    {
+      "epoch": 0.2440308988764045,
+      "grad_norm": 0.9562146067619324,
+      "learning_rate": 4.984881073305841e-06,
+      "loss": 0.7175,
+      "step": 695
+    },
+    {
+      "epoch": 0.2443820224719101,
+      "grad_norm": 0.9829107522964478,
+      "learning_rate": 4.984830262178974e-06,
+      "loss": 0.7343,
+      "step": 696
+    },
+    {
+      "epoch": 0.24473314606741572,
+      "grad_norm": 0.9095830917358398,
+      "learning_rate": 4.984779366073036e-06,
+      "loss": 0.7474,
+      "step": 697
+    },
+    {
+      "epoch": 0.24508426966292135,
+      "grad_norm": 0.9279952049255371,
+      "learning_rate": 4.984728384989772e-06,
+      "loss": 0.7574,
+      "step": 698
+    },
+    {
+      "epoch": 0.24543539325842698,
+      "grad_norm": 1.07504141330719,
+      "learning_rate": 4.984677318930922e-06,
+      "loss": 0.7309,
+      "step": 699
+    },
+    {
+      "epoch": 0.24578651685393257,
+      "grad_norm": 0.9406894445419312,
+      "learning_rate": 4.984626167898235e-06,
+      "loss": 0.7187,
+      "step": 700
+    },
+    {
+      "epoch": 0.2461376404494382,
+      "grad_norm": 0.9229064583778381,
+      "learning_rate": 4.9845749318934575e-06,
+      "loss": 0.7379,
+      "step": 701
+    },
+    {
+      "epoch": 0.24648876404494383,
+      "grad_norm": 0.9469361901283264,
+      "learning_rate": 4.984523610918345e-06,
+      "loss": 0.7585,
+      "step": 702
+    },
+    {
+      "epoch": 0.24683988764044945,
+      "grad_norm": 0.9020796418190002,
+      "learning_rate": 4.984472204974651e-06,
+      "loss": 0.714,
+      "step": 703
+    },
+    {
+      "epoch": 0.24719101123595505,
+      "grad_norm": 0.9876411557197571,
+      "learning_rate": 4.9844207140641336e-06,
+      "loss": 0.7013,
+      "step": 704
+    },
+    {
+      "epoch": 0.24754213483146068,
+      "grad_norm": 0.9139506816864014,
+      "learning_rate": 4.984369138188553e-06,
+      "loss": 0.7252,
+      "step": 705
+    },
+    {
+      "epoch": 0.2478932584269663,
+      "grad_norm": 0.9223290681838989,
+      "learning_rate": 4.9843174773496735e-06,
+      "loss": 0.7603,
+      "step": 706
+    },
+    {
+      "epoch": 0.2482443820224719,
+      "grad_norm": 0.9541411995887756,
+      "learning_rate": 4.984265731549263e-06,
+      "loss": 0.6997,
+      "step": 707
+    },
+    {
+      "epoch": 0.24859550561797752,
+      "grad_norm": 0.9302672147750854,
+      "learning_rate": 4.98421390078909e-06,
+      "loss": 0.7252,
+      "step": 708
+    },
+    {
+      "epoch": 0.24894662921348315,
+      "grad_norm": 0.9554900527000427,
+      "learning_rate": 4.984161985070927e-06,
+      "loss": 0.7305,
+      "step": 709
+    },
+    {
+      "epoch": 0.24929775280898878,
+      "grad_norm": 0.9729623794555664,
+      "learning_rate": 4.98410998439655e-06,
+      "loss": 0.7318,
+      "step": 710
+    },
+    {
+      "epoch": 0.24964887640449437,
+      "grad_norm": 0.9298332929611206,
+      "learning_rate": 4.984057898767738e-06,
+      "loss": 0.7348,
+      "step": 711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9653823375701904,
+      "learning_rate": 4.984005728186271e-06,
+      "loss": 0.7566,
+      "step": 712
+    },
+    {
+      "epoch": 0.2503511235955056,
+      "grad_norm": 0.9283461570739746,
+      "learning_rate": 4.983953472653933e-06,
+      "loss": 0.6999,
+      "step": 713
+    },
+    {
+      "epoch": 0.25070224719101125,
+      "grad_norm": 0.9600170254707336,
+      "learning_rate": 4.983901132172513e-06,
+      "loss": 0.7378,
+      "step": 714
+    },
+    {
+      "epoch": 0.25105337078651685,
+      "grad_norm": 0.9261863827705383,
+      "learning_rate": 4.983848706743799e-06,
+      "loss": 0.7245,
+      "step": 715
+    },
+    {
+      "epoch": 0.25140449438202245,
+      "grad_norm": 0.9267969727516174,
+      "learning_rate": 4.983796196369585e-06,
+      "loss": 0.6921,
+      "step": 716
+    },
+    {
+      "epoch": 0.2517556179775281,
+      "grad_norm": 0.9277015328407288,
+      "learning_rate": 4.983743601051666e-06,
+      "loss": 0.715,
+      "step": 717
+    },
+    {
+      "epoch": 0.2521067415730337,
+      "grad_norm": 0.9029396176338196,
+      "learning_rate": 4.9836909207918415e-06,
+      "loss": 0.7739,
+      "step": 718
+    },
+    {
+      "epoch": 0.25245786516853935,
+      "grad_norm": 0.9498785138130188,
+      "learning_rate": 4.983638155591913e-06,
+      "loss": 0.7071,
+      "step": 719
+    },
+    {
+      "epoch": 0.25280898876404495,
+      "grad_norm": 0.934562087059021,
+      "learning_rate": 4.983585305453685e-06,
+      "loss": 0.7183,
+      "step": 720
+    },
+    {
+      "epoch": 0.25316011235955055,
+      "grad_norm": 0.9572330713272095,
+      "learning_rate": 4.983532370378964e-06,
+      "loss": 0.7535,
+      "step": 721
+    },
+    {
+      "epoch": 0.2535112359550562,
+      "grad_norm": 0.9443022608757019,
+      "learning_rate": 4.983479350369562e-06,
+      "loss": 0.7127,
+      "step": 722
+    },
+    {
+      "epoch": 0.2538623595505618,
+      "grad_norm": 0.9258742928504944,
+      "learning_rate": 4.983426245427291e-06,
+      "loss": 0.7002,
+      "step": 723
+    },
+    {
+      "epoch": 0.2542134831460674,
+      "grad_norm": 0.939900279045105,
+      "learning_rate": 4.983373055553968e-06,
+      "loss": 0.7234,
+      "step": 724
+    },
+    {
+      "epoch": 0.25456460674157305,
+      "grad_norm": 0.9580402374267578,
+      "learning_rate": 4.983319780751411e-06,
+      "loss": 0.7519,
+      "step": 725
+    },
+    {
+      "epoch": 0.25491573033707865,
+      "grad_norm": 0.9556792974472046,
+      "learning_rate": 4.983266421021442e-06,
+      "loss": 0.7701,
+      "step": 726
+    },
+    {
+      "epoch": 0.25526685393258425,
+      "grad_norm": 0.9831913709640503,
+      "learning_rate": 4.983212976365887e-06,
+      "loss": 0.7189,
+      "step": 727
+    },
+    {
+      "epoch": 0.2556179775280899,
+      "grad_norm": 0.9363849759101868,
+      "learning_rate": 4.983159446786573e-06,
+      "loss": 0.7061,
+      "step": 728
+    },
+    {
+      "epoch": 0.2559691011235955,
+      "grad_norm": 0.9291892647743225,
+      "learning_rate": 4.983105832285331e-06,
+      "loss": 0.6942,
+      "step": 729
+    },
+    {
+      "epoch": 0.2563202247191011,
+      "grad_norm": 0.9837853908538818,
+      "learning_rate": 4.983052132863995e-06,
+      "loss": 0.7251,
+      "step": 730
+    },
+    {
+      "epoch": 0.25667134831460675,
+      "grad_norm": 0.9630160331726074,
+      "learning_rate": 4.9829983485244e-06,
+      "loss": 0.694,
+      "step": 731
+    },
+    {
+      "epoch": 0.25702247191011235,
+      "grad_norm": 0.972981333732605,
+      "learning_rate": 4.982944479268386e-06,
+      "loss": 0.7232,
+      "step": 732
+    },
+    {
+      "epoch": 0.257373595505618,
+      "grad_norm": 0.9112554788589478,
+      "learning_rate": 4.9828905250977955e-06,
+      "loss": 0.7553,
+      "step": 733
+    },
+    {
+      "epoch": 0.2577247191011236,
+      "grad_norm": 0.9407148361206055,
+      "learning_rate": 4.982836486014474e-06,
+      "loss": 0.7931,
+      "step": 734
+    },
+    {
+      "epoch": 0.2580758426966292,
+      "grad_norm": 0.9683741927146912,
+      "learning_rate": 4.98278236202027e-06,
+      "loss": 0.76,
+      "step": 735
+    },
+    {
+      "epoch": 0.25842696629213485,
+      "grad_norm": 0.9538892507553101,
+      "learning_rate": 4.9827281531170325e-06,
+      "loss": 0.7491,
+      "step": 736
+    },
+    {
+      "epoch": 0.25877808988764045,
+      "grad_norm": 0.9792529344558716,
+      "learning_rate": 4.982673859306617e-06,
+      "loss": 0.752,
+      "step": 737
+    },
+    {
+      "epoch": 0.25912921348314605,
+      "grad_norm": 0.9384421110153198,
+      "learning_rate": 4.98261948059088e-06,
+      "loss": 0.7587,
+      "step": 738
+    },
+    {
+      "epoch": 0.2594803370786517,
+      "grad_norm": 0.9048983454704285,
+      "learning_rate": 4.982565016971682e-06,
+      "loss": 0.7429,
+      "step": 739
+    },
+    {
+      "epoch": 0.2598314606741573,
+      "grad_norm": 0.893500804901123,
+      "learning_rate": 4.982510468450884e-06,
+      "loss": 0.6985,
+      "step": 740
+    },
+    {
+      "epoch": 0.2601825842696629,
+      "grad_norm": 0.9900831580162048,
+      "learning_rate": 4.982455835030352e-06,
+      "loss": 0.7336,
+      "step": 741
+    },
+    {
+      "epoch": 0.26053370786516855,
+      "grad_norm": 0.9558880925178528,
+      "learning_rate": 4.982401116711955e-06,
+      "loss": 0.7561,
+      "step": 742
+    },
+    {
+      "epoch": 0.26088483146067415,
+      "grad_norm": 0.8868340849876404,
+      "learning_rate": 4.982346313497564e-06,
+      "loss": 0.6967,
+      "step": 743
+    },
+    {
+      "epoch": 0.2612359550561798,
+      "grad_norm": 0.9625622034072876,
+      "learning_rate": 4.9822914253890535e-06,
+      "loss": 0.7382,
+      "step": 744
+    },
+    {
+      "epoch": 0.2615870786516854,
+      "grad_norm": 0.9891008138656616,
+      "learning_rate": 4.9822364523883e-06,
+      "loss": 0.7316,
+      "step": 745
+    },
+    {
+      "epoch": 0.261938202247191,
+      "grad_norm": 0.9161337018013,
+      "learning_rate": 4.982181394497184e-06,
+      "loss": 0.7225,
+      "step": 746
+    },
+    {
+      "epoch": 0.26228932584269665,
+      "grad_norm": 0.9532581567764282,
+      "learning_rate": 4.982126251717589e-06,
+      "loss": 0.7412,
+      "step": 747
+    },
+    {
+      "epoch": 0.26264044943820225,
+      "grad_norm": 0.90705406665802,
+      "learning_rate": 4.982071024051399e-06,
+      "loss": 0.6938,
+      "step": 748
+    },
+    {
+      "epoch": 0.26299157303370785,
+      "grad_norm": 0.9263889789581299,
+      "learning_rate": 4.982015711500505e-06,
+      "loss": 0.7575,
+      "step": 749
+    },
+    {
+      "epoch": 0.2633426966292135,
+      "grad_norm": 0.9247738718986511,
+      "learning_rate": 4.981960314066797e-06,
+      "loss": 0.7184,
+      "step": 750
+    },
+    {
+      "epoch": 0.2636938202247191,
+      "grad_norm": 0.9485296607017517,
+      "learning_rate": 4.98190483175217e-06,
+      "loss": 0.7297,
+      "step": 751
+    },
+    {
+      "epoch": 0.2640449438202247,
+      "grad_norm": 0.9109437465667725,
+      "learning_rate": 4.981849264558523e-06,
+      "loss": 0.7422,
+      "step": 752
+    },
+    {
+      "epoch": 0.26439606741573035,
+      "grad_norm": 0.9430893063545227,
+      "learning_rate": 4.981793612487753e-06,
+      "loss": 0.7634,
+      "step": 753
+    },
+    {
+      "epoch": 0.26474719101123595,
+      "grad_norm": 0.9232211709022522,
+      "learning_rate": 4.981737875541765e-06,
+      "loss": 0.7587,
+      "step": 754
+    },
+    {
+      "epoch": 0.26509831460674155,
+      "grad_norm": 0.8920214176177979,
+      "learning_rate": 4.9816820537224675e-06,
+      "loss": 0.7222,
+      "step": 755
+    },
+    {
+      "epoch": 0.2654494382022472,
+      "grad_norm": 0.9339509010314941,
+      "learning_rate": 4.981626147031765e-06,
+      "loss": 0.7253,
+      "step": 756
+    },
+    {
+      "epoch": 0.2658005617977528,
+      "grad_norm": 0.890581488609314,
+      "learning_rate": 4.981570155471573e-06,
+      "loss": 0.7166,
+      "step": 757
+    },
+    {
+      "epoch": 0.26615168539325845,
+      "grad_norm": 0.9297827482223511,
+      "learning_rate": 4.981514079043806e-06,
+      "loss": 0.713,
+      "step": 758
+    },
+    {
+      "epoch": 0.26650280898876405,
+      "grad_norm": 0.9213199019432068,
+      "learning_rate": 4.98145791775038e-06,
+      "loss": 0.7312,
+      "step": 759
+    },
+    {
+      "epoch": 0.26685393258426965,
+      "grad_norm": 0.9185107946395874,
+      "learning_rate": 4.981401671593216e-06,
+      "loss": 0.7215,
+      "step": 760
+    },
+    {
+      "epoch": 0.2672050561797753,
+      "grad_norm": 0.8874849081039429,
+      "learning_rate": 4.981345340574239e-06,
+      "loss": 0.7082,
+      "step": 761
+    },
+    {
+      "epoch": 0.2675561797752809,
+      "grad_norm": 0.9179455637931824,
+      "learning_rate": 4.981288924695374e-06,
+      "loss": 0.7243,
+      "step": 762
+    },
+    {
+      "epoch": 0.2679073033707865,
+      "grad_norm": 0.8956063985824585,
+      "learning_rate": 4.981232423958552e-06,
+      "loss": 0.704,
+      "step": 763
+    },
+    {
+      "epoch": 0.26825842696629215,
+      "grad_norm": 0.9451838731765747,
+      "learning_rate": 4.981175838365704e-06,
+      "loss": 0.7518,
+      "step": 764
+    },
+    {
+      "epoch": 0.26860955056179775,
+      "grad_norm": 0.9265857338905334,
+      "learning_rate": 4.981119167918765e-06,
+      "loss": 0.7143,
+      "step": 765
+    },
+    {
+      "epoch": 0.26896067415730335,
+      "grad_norm": 0.9493303298950195,
+      "learning_rate": 4.981062412619674e-06,
+      "loss": 0.7172,
+      "step": 766
+    },
+    {
+      "epoch": 0.269311797752809,
+      "grad_norm": 0.9223463535308838,
+      "learning_rate": 4.981005572470372e-06,
+      "loss": 0.7383,
+      "step": 767
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 0.9502999782562256,
+      "learning_rate": 4.980948647472802e-06,
+      "loss": 0.7137,
+      "step": 768
+    },
+    {
+      "epoch": 0.2700140449438202,
+      "grad_norm": 0.9198105335235596,
+      "learning_rate": 4.980891637628911e-06,
+      "loss": 0.706,
+      "step": 769
+    },
+    {
+      "epoch": 0.27036516853932585,
+      "grad_norm": 0.9435210227966309,
+      "learning_rate": 4.980834542940649e-06,
+      "loss": 0.6921,
+      "step": 770
+    },
+    {
+      "epoch": 0.27071629213483145,
+      "grad_norm": 0.9600005745887756,
+      "learning_rate": 4.980777363409969e-06,
+      "loss": 0.72,
+      "step": 771
+    },
+    {
+      "epoch": 0.2710674157303371,
+      "grad_norm": 0.9591335654258728,
+      "learning_rate": 4.980720099038825e-06,
+      "loss": 0.7386,
+      "step": 772
+    },
+    {
+      "epoch": 0.2714185393258427,
+      "grad_norm": 0.9055672883987427,
+      "learning_rate": 4.980662749829177e-06,
+      "loss": 0.7568,
+      "step": 773
+    },
+    {
+      "epoch": 0.2717696629213483,
+      "grad_norm": 0.9077644348144531,
+      "learning_rate": 4.9806053157829855e-06,
+      "loss": 0.7122,
+      "step": 774
+    },
+    {
+      "epoch": 0.27212078651685395,
+      "grad_norm": 0.9248702526092529,
+      "learning_rate": 4.980547796902216e-06,
+      "loss": 0.6866,
+      "step": 775
+    },
+    {
+      "epoch": 0.27247191011235955,
+      "grad_norm": 0.9374347925186157,
+      "learning_rate": 4.980490193188834e-06,
+      "loss": 0.7283,
+      "step": 776
+    },
+    {
+      "epoch": 0.27282303370786515,
+      "grad_norm": 0.9280509352684021,
+      "learning_rate": 4.980432504644809e-06,
+      "loss": 0.6896,
+      "step": 777
+    },
+    {
+      "epoch": 0.2731741573033708,
+      "grad_norm": 0.9158201813697815,
+      "learning_rate": 4.980374731272115e-06,
+      "loss": 0.7222,
+      "step": 778
+    },
+    {
+      "epoch": 0.2735252808988764,
+      "grad_norm": 0.918114960193634,
+      "learning_rate": 4.980316873072729e-06,
+      "loss": 0.7528,
+      "step": 779
+    },
+    {
+      "epoch": 0.273876404494382,
+      "grad_norm": 0.891121506690979,
+      "learning_rate": 4.980258930048627e-06,
+      "loss": 0.7265,
+      "step": 780
+    },
+    {
+      "epoch": 0.27422752808988765,
+      "grad_norm": 0.9570785760879517,
+      "learning_rate": 4.980200902201792e-06,
+      "loss": 0.6957,
+      "step": 781
+    },
+    {
+      "epoch": 0.27457865168539325,
+      "grad_norm": 0.8888586759567261,
+      "learning_rate": 4.980142789534209e-06,
+      "loss": 0.7258,
+      "step": 782
+    },
+    {
+      "epoch": 0.2749297752808989,
+      "grad_norm": 0.9602726101875305,
+      "learning_rate": 4.980084592047866e-06,
+      "loss": 0.741,
+      "step": 783
+    },
+    {
+      "epoch": 0.2752808988764045,
+      "grad_norm": 0.9402101635932922,
+      "learning_rate": 4.9800263097447505e-06,
+      "loss": 0.7789,
+      "step": 784
+    },
+    {
+      "epoch": 0.2756320224719101,
+      "grad_norm": 0.916023850440979,
+      "learning_rate": 4.9799679426268575e-06,
+      "loss": 0.6746,
+      "step": 785
+    },
+    {
+      "epoch": 0.27598314606741575,
+      "grad_norm": 1.0306038856506348,
+      "learning_rate": 4.979909490696184e-06,
+      "loss": 0.7396,
+      "step": 786
+    },
+    {
+      "epoch": 0.27633426966292135,
+      "grad_norm": 0.9112997651100159,
+      "learning_rate": 4.979850953954727e-06,
+      "loss": 0.7225,
+      "step": 787
+    },
+    {
+      "epoch": 0.27668539325842695,
+      "grad_norm": 0.9124306440353394,
+      "learning_rate": 4.97979233240449e-06,
+      "loss": 0.7244,
+      "step": 788
+    },
+    {
+      "epoch": 0.2770365168539326,
+      "grad_norm": 0.8682541847229004,
+      "learning_rate": 4.979733626047477e-06,
+      "loss": 0.6983,
+      "step": 789
+    },
+    {
+      "epoch": 0.2773876404494382,
+      "grad_norm": 0.9099353551864624,
+      "learning_rate": 4.979674834885696e-06,
+      "loss": 0.723,
+      "step": 790
+    },
+    {
+      "epoch": 0.2777387640449438,
+      "grad_norm": 0.9057777523994446,
+      "learning_rate": 4.979615958921158e-06,
+      "loss": 0.7063,
+      "step": 791
+    },
+    {
+      "epoch": 0.27808988764044945,
+      "grad_norm": 0.9777807593345642,
+      "learning_rate": 4.979556998155876e-06,
+      "loss": 0.7047,
+      "step": 792
+    },
+    {
+      "epoch": 0.27844101123595505,
+      "grad_norm": 0.952314019203186,
+      "learning_rate": 4.979497952591865e-06,
+      "loss": 0.7589,
+      "step": 793
+    },
+    {
+      "epoch": 0.27879213483146065,
+      "grad_norm": 0.8948432207107544,
+      "learning_rate": 4.979438822231147e-06,
+      "loss": 0.6893,
+      "step": 794
+    },
+    {
+      "epoch": 0.2791432584269663,
+      "grad_norm": 0.9998204708099365,
+      "learning_rate": 4.979379607075743e-06,
+      "loss": 0.7176,
+      "step": 795
+    },
+    {
+      "epoch": 0.2794943820224719,
+      "grad_norm": 0.9392699003219604,
+      "learning_rate": 4.979320307127678e-06,
+      "loss": 0.7607,
+      "step": 796
+    },
+    {
+      "epoch": 0.27984550561797755,
+      "grad_norm": 0.916340708732605,
+      "learning_rate": 4.97926092238898e-06,
+      "loss": 0.6797,
+      "step": 797
+    },
+    {
+      "epoch": 0.28019662921348315,
+      "grad_norm": 0.9007791876792908,
+      "learning_rate": 4.979201452861679e-06,
+      "loss": 0.7015,
+      "step": 798
+    },
+    {
+      "epoch": 0.28054775280898875,
+      "grad_norm": 0.9042453765869141,
+      "learning_rate": 4.979141898547811e-06,
+      "loss": 0.7113,
+      "step": 799
+    },
+    {
+      "epoch": 0.2808988764044944,
+      "grad_norm": 0.9455578327178955,
+      "learning_rate": 4.9790822594494116e-06,
+      "loss": 0.7031,
+      "step": 800
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 0.9787691235542297,
+      "learning_rate": 4.979022535568519e-06,
+      "loss": 0.7328,
+      "step": 801
+    },
+    {
+      "epoch": 0.2816011235955056,
+      "grad_norm": 0.9504345059394836,
+      "learning_rate": 4.978962726907179e-06,
+      "loss": 0.7301,
+      "step": 802
+    },
+    {
+      "epoch": 0.28195224719101125,
+      "grad_norm": 0.9304360747337341,
+      "learning_rate": 4.978902833467434e-06,
+      "loss": 0.7195,
+      "step": 803
+    },
+    {
+      "epoch": 0.28230337078651685,
+      "grad_norm": 0.9509701132774353,
+      "learning_rate": 4.978842855251333e-06,
+      "loss": 0.7532,
+      "step": 804
+    },
+    {
+      "epoch": 0.28265449438202245,
+      "grad_norm": 0.9462277889251709,
+      "learning_rate": 4.9787827922609276e-06,
+      "loss": 0.708,
+      "step": 805
+    },
+    {
+      "epoch": 0.2830056179775281,
+      "grad_norm": 0.9095754027366638,
+      "learning_rate": 4.978722644498272e-06,
+      "loss": 0.7484,
+      "step": 806
+    },
+    {
+      "epoch": 0.2833567415730337,
+      "grad_norm": 0.8992451429367065,
+      "learning_rate": 4.978662411965424e-06,
+      "loss": 0.7033,
+      "step": 807
+    },
+    {
+      "epoch": 0.28370786516853935,
+      "grad_norm": 0.913910984992981,
+      "learning_rate": 4.978602094664441e-06,
+      "loss": 0.7333,
+      "step": 808
+    },
+    {
+      "epoch": 0.28405898876404495,
+      "grad_norm": 0.9041484594345093,
+      "learning_rate": 4.978541692597388e-06,
+      "loss": 0.7196,
+      "step": 809
+    },
+    {
+      "epoch": 0.28441011235955055,
+      "grad_norm": 0.9570571184158325,
+      "learning_rate": 4.97848120576633e-06,
+      "loss": 0.7199,
+      "step": 810
+    },
+    {
+      "epoch": 0.2847612359550562,
+      "grad_norm": 0.9184895753860474,
+      "learning_rate": 4.978420634173336e-06,
+      "loss": 0.6982,
+      "step": 811
+    },
+    {
+      "epoch": 0.2851123595505618,
+      "grad_norm": 0.9278954267501831,
+      "learning_rate": 4.978359977820476e-06,
+      "loss": 0.7475,
+      "step": 812
+    },
+    {
+      "epoch": 0.2854634831460674,
+      "grad_norm": 0.9236921668052673,
+      "learning_rate": 4.978299236709826e-06,
+      "loss": 0.7343,
+      "step": 813
+    },
+    {
+      "epoch": 0.28581460674157305,
+      "grad_norm": 0.9183304309844971,
+      "learning_rate": 4.978238410843464e-06,
+      "loss": 0.7111,
+      "step": 814
+    },
+    {
+      "epoch": 0.28616573033707865,
+      "grad_norm": 0.895050585269928,
+      "learning_rate": 4.978177500223468e-06,
+      "loss": 0.6746,
+      "step": 815
+    },
+    {
+      "epoch": 0.28651685393258425,
+      "grad_norm": 0.9193557500839233,
+      "learning_rate": 4.9781165048519216e-06,
+      "loss": 0.7025,
+      "step": 816
+    },
+    {
+      "epoch": 0.2868679775280899,
+      "grad_norm": 0.9046199917793274,
+      "learning_rate": 4.978055424730912e-06,
+      "loss": 0.701,
+      "step": 817
+    },
+    {
+      "epoch": 0.2872191011235955,
+      "grad_norm": 0.9200829267501831,
+      "learning_rate": 4.977994259862527e-06,
+      "loss": 0.755,
+      "step": 818
+    },
+    {
+      "epoch": 0.2875702247191011,
+      "grad_norm": 0.929964542388916,
+      "learning_rate": 4.977933010248859e-06,
+      "loss": 0.7274,
+      "step": 819
+    },
+    {
+      "epoch": 0.28792134831460675,
+      "grad_norm": 0.9332923293113708,
+      "learning_rate": 4.977871675892003e-06,
+      "loss": 0.7503,
+      "step": 820
+    },
+    {
+      "epoch": 0.28827247191011235,
+      "grad_norm": 0.8993156552314758,
+      "learning_rate": 4.977810256794055e-06,
+      "loss": 0.6954,
+      "step": 821
+    },
+    {
+      "epoch": 0.288623595505618,
+      "grad_norm": 0.8985475301742554,
+      "learning_rate": 4.977748752957116e-06,
+      "loss": 0.7287,
+      "step": 822
+    },
+    {
+      "epoch": 0.2889747191011236,
+      "grad_norm": 0.9083667397499084,
+      "learning_rate": 4.977687164383291e-06,
+      "loss": 0.7352,
+      "step": 823
+    },
+    {
+      "epoch": 0.2893258426966292,
+      "grad_norm": 0.9311912059783936,
+      "learning_rate": 4.977625491074686e-06,
+      "loss": 0.7199,
+      "step": 824
+    },
+    {
+      "epoch": 0.28967696629213485,
+      "grad_norm": 0.8959439396858215,
+      "learning_rate": 4.977563733033408e-06,
+      "loss": 0.6992,
+      "step": 825
+    },
+    {
+      "epoch": 0.29002808988764045,
+      "grad_norm": 1.018567681312561,
+      "learning_rate": 4.9775018902615715e-06,
+      "loss": 0.7278,
+      "step": 826
+    },
+    {
+      "epoch": 0.29037921348314605,
+      "grad_norm": 0.9925201535224915,
+      "learning_rate": 4.977439962761289e-06,
+      "loss": 0.7423,
+      "step": 827
+    },
+    {
+      "epoch": 0.2907303370786517,
+      "grad_norm": 1.0001310110092163,
+      "learning_rate": 4.977377950534681e-06,
+      "loss": 0.7395,
+      "step": 828
+    },
+    {
+      "epoch": 0.2910814606741573,
+      "grad_norm": 0.9196740984916687,
+      "learning_rate": 4.977315853583866e-06,
+      "loss": 0.7588,
+      "step": 829
+    },
+    {
+      "epoch": 0.2914325842696629,
+      "grad_norm": 0.9672794342041016,
+      "learning_rate": 4.977253671910969e-06,
+      "loss": 0.7378,
+      "step": 830
+    },
+    {
+      "epoch": 0.29178370786516855,
+      "grad_norm": 1.0301952362060547,
+      "learning_rate": 4.977191405518116e-06,
+      "loss": 0.7757,
+      "step": 831
+    },
+    {
+      "epoch": 0.29213483146067415,
+      "grad_norm": 0.9453630447387695,
+      "learning_rate": 4.977129054407437e-06,
+      "loss": 0.7335,
+      "step": 832
+    },
+    {
+      "epoch": 0.2924859550561798,
+      "grad_norm": 0.9876764416694641,
+      "learning_rate": 4.977066618581065e-06,
+      "loss": 0.7437,
+      "step": 833
+    },
+    {
+      "epoch": 0.2928370786516854,
+      "grad_norm": 0.9211217164993286,
+      "learning_rate": 4.977004098041133e-06,
+      "loss": 0.7333,
+      "step": 834
+    },
+    {
+      "epoch": 0.293188202247191,
+      "grad_norm": 0.9148357510566711,
+      "learning_rate": 4.976941492789781e-06,
+      "loss": 0.677,
+      "step": 835
+    },
+    {
+      "epoch": 0.29353932584269665,
+      "grad_norm": 0.9743471741676331,
+      "learning_rate": 4.976878802829149e-06,
+      "loss": 0.7586,
+      "step": 836
+    },
+    {
+      "epoch": 0.29389044943820225,
+      "grad_norm": 0.9516198039054871,
+      "learning_rate": 4.976816028161382e-06,
+      "loss": 0.7364,
+      "step": 837
+    },
+    {
+      "epoch": 0.29424157303370785,
+      "grad_norm": 0.8972299098968506,
+      "learning_rate": 4.9767531687886265e-06,
+      "loss": 0.6912,
+      "step": 838
+    },
+    {
+      "epoch": 0.2945926966292135,
+      "grad_norm": 0.8950793147087097,
+      "learning_rate": 4.976690224713031e-06,
+      "loss": 0.6969,
+      "step": 839
+    },
+    {
+      "epoch": 0.2949438202247191,
+      "grad_norm": 0.9539402723312378,
+      "learning_rate": 4.976627195936749e-06,
+      "loss": 0.7348,
+      "step": 840
+    },
+    {
+      "epoch": 0.2952949438202247,
+      "grad_norm": 1.32026207447052,
+      "learning_rate": 4.976564082461938e-06,
+      "loss": 0.7192,
+      "step": 841
+    },
+    {
+      "epoch": 0.29564606741573035,
+      "grad_norm": 0.9319935441017151,
+      "learning_rate": 4.976500884290753e-06,
+      "loss": 0.6957,
+      "step": 842
+    },
+    {
+      "epoch": 0.29599719101123595,
+      "grad_norm": 0.9559399485588074,
+      "learning_rate": 4.976437601425358e-06,
+      "loss": 0.7394,
+      "step": 843
+    },
+    {
+      "epoch": 0.29634831460674155,
+      "grad_norm": 0.9115216135978699,
+      "learning_rate": 4.976374233867915e-06,
+      "loss": 0.6955,
+      "step": 844
+    },
+    {
+      "epoch": 0.2966994382022472,
+      "grad_norm": 0.8987361788749695,
+      "learning_rate": 4.976310781620592e-06,
+      "loss": 0.7057,
+      "step": 845
+    },
+    {
+      "epoch": 0.2970505617977528,
+      "grad_norm": 0.9885380268096924,
+      "learning_rate": 4.976247244685561e-06,
+      "loss": 0.7542,
+      "step": 846
+    },
+    {
+      "epoch": 0.29740168539325845,
+      "grad_norm": 0.9176011681556702,
+      "learning_rate": 4.9761836230649914e-06,
+      "loss": 0.6836,
+      "step": 847
+    },
+    {
+      "epoch": 0.29775280898876405,
+      "grad_norm": 0.9235329031944275,
+      "learning_rate": 4.976119916761063e-06,
+      "loss": 0.7541,
+      "step": 848
+    },
+    {
+      "epoch": 0.29810393258426965,
+      "grad_norm": 0.9174650311470032,
+      "learning_rate": 4.9760561257759504e-06,
+      "loss": 0.7149,
+      "step": 849
+    },
+    {
+      "epoch": 0.2984550561797753,
+      "grad_norm": 0.9441576600074768,
+      "learning_rate": 4.975992250111837e-06,
+      "loss": 0.7222,
+      "step": 850
+    },
+    {
+      "epoch": 0.2988061797752809,
+      "grad_norm": 0.9702540636062622,
+      "learning_rate": 4.975928289770909e-06,
+      "loss": 0.7193,
+      "step": 851
+    },
+    {
+      "epoch": 0.2991573033707865,
+      "grad_norm": 0.9554623961448669,
+      "learning_rate": 4.9758642447553505e-06,
+      "loss": 0.7537,
+      "step": 852
+    },
+    {
+      "epoch": 0.29950842696629215,
+      "grad_norm": 0.9210932850837708,
+      "learning_rate": 4.975800115067355e-06,
+      "loss": 0.7159,
+      "step": 853
+    },
+    {
+      "epoch": 0.29985955056179775,
+      "grad_norm": 0.902009904384613,
+      "learning_rate": 4.975735900709113e-06,
+      "loss": 0.6952,
+      "step": 854
+    },
+    {
+      "epoch": 0.30021067415730335,
+      "grad_norm": 0.9207450151443481,
+      "learning_rate": 4.975671601682822e-06,
+      "loss": 0.6705,
+      "step": 855
+    },
+    {
+      "epoch": 0.300561797752809,
+      "grad_norm": 0.9092718362808228,
+      "learning_rate": 4.975607217990681e-06,
+      "loss": 0.723,
+      "step": 856
+    },
+    {
+      "epoch": 0.3009129213483146,
+      "grad_norm": 0.955410897731781,
+      "learning_rate": 4.975542749634891e-06,
+      "loss": 0.728,
+      "step": 857
+    },
+    {
+      "epoch": 0.3012640449438202,
+      "grad_norm": 0.9404045343399048,
+      "learning_rate": 4.975478196617658e-06,
+      "loss": 0.7381,
+      "step": 858
+    },
+    {
+      "epoch": 0.30161516853932585,
+      "grad_norm": 0.9133435487747192,
+      "learning_rate": 4.9754135589411875e-06,
+      "loss": 0.7306,
+      "step": 859
+    },
+    {
+      "epoch": 0.30196629213483145,
+      "grad_norm": 0.9013397693634033,
+      "learning_rate": 4.975348836607693e-06,
+      "loss": 0.7189,
+      "step": 860
+    },
+    {
+      "epoch": 0.3023174157303371,
+      "grad_norm": 0.9037486910820007,
+      "learning_rate": 4.975284029619386e-06,
+      "loss": 0.7152,
+      "step": 861
+    },
+    {
+      "epoch": 0.3026685393258427,
+      "grad_norm": 0.9141895771026611,
+      "learning_rate": 4.975219137978484e-06,
+      "loss": 0.71,
+      "step": 862
+    },
+    {
+      "epoch": 0.3030196629213483,
+      "grad_norm": 0.942010223865509,
+      "learning_rate": 4.975154161687205e-06,
+      "loss": 0.7255,
+      "step": 863
+    },
+    {
+      "epoch": 0.30337078651685395,
+      "grad_norm": 0.9464311599731445,
+      "learning_rate": 4.9750891007477716e-06,
+      "loss": 0.7012,
+      "step": 864
+    },
+    {
+      "epoch": 0.30372191011235955,
+      "grad_norm": 0.9062106609344482,
+      "learning_rate": 4.975023955162409e-06,
+      "loss": 0.6909,
+      "step": 865
+    },
+    {
+      "epoch": 0.30407303370786515,
+      "grad_norm": 0.921981155872345,
+      "learning_rate": 4.974958724933344e-06,
+      "loss": 0.7312,
+      "step": 866
+    },
+    {
+      "epoch": 0.3044241573033708,
+      "grad_norm": 0.902531087398529,
+      "learning_rate": 4.97489341006281e-06,
+      "loss": 0.674,
+      "step": 867
+    },
+    {
+      "epoch": 0.3047752808988764,
+      "grad_norm": 0.9487978219985962,
+      "learning_rate": 4.974828010553039e-06,
+      "loss": 0.7054,
+      "step": 868
+    },
+    {
+      "epoch": 0.305126404494382,
+      "grad_norm": 0.9971505999565125,
+      "learning_rate": 4.9747625264062675e-06,
+      "loss": 0.7676,
+      "step": 869
+    },
+    {
+      "epoch": 0.30547752808988765,
+      "grad_norm": 0.9108356237411499,
+      "learning_rate": 4.974696957624735e-06,
+      "loss": 0.7199,
+      "step": 870
+    },
+    {
+      "epoch": 0.30582865168539325,
+      "grad_norm": 0.9270719885826111,
+      "learning_rate": 4.974631304210684e-06,
+      "loss": 0.748,
+      "step": 871
+    },
+    {
+      "epoch": 0.3061797752808989,
+      "grad_norm": 0.9216623306274414,
+      "learning_rate": 4.97456556616636e-06,
+      "loss": 0.7333,
+      "step": 872
+    },
+    {
+      "epoch": 0.3065308988764045,
+      "grad_norm": 0.980196475982666,
+      "learning_rate": 4.9744997434940125e-06,
+      "loss": 0.7175,
+      "step": 873
+    },
+    {
+      "epoch": 0.3068820224719101,
+      "grad_norm": 0.8955034613609314,
+      "learning_rate": 4.97443383619589e-06,
+      "loss": 0.7133,
+      "step": 874
+    },
+    {
+      "epoch": 0.30723314606741575,
+      "grad_norm": 0.9280632734298706,
+      "learning_rate": 4.974367844274248e-06,
+      "loss": 0.6953,
+      "step": 875
+    },
+    {
+      "epoch": 0.30758426966292135,
+      "grad_norm": 0.9301517605781555,
+      "learning_rate": 4.974301767731343e-06,
+      "loss": 0.7057,
+      "step": 876
+    },
+    {
+      "epoch": 0.30793539325842695,
+      "grad_norm": 0.9015063047409058,
+      "learning_rate": 4.974235606569434e-06,
+      "loss": 0.699,
+      "step": 877
+    },
+    {
+      "epoch": 0.3082865168539326,
+      "grad_norm": 0.9064421057701111,
+      "learning_rate": 4.974169360790786e-06,
+      "loss": 0.6937,
+      "step": 878
+    },
+    {
+      "epoch": 0.3086376404494382,
+      "grad_norm": 0.9030662775039673,
+      "learning_rate": 4.974103030397662e-06,
+      "loss": 0.7455,
+      "step": 879
+    },
+    {
+      "epoch": 0.3089887640449438,
+      "grad_norm": 0.9049381613731384,
+      "learning_rate": 4.974036615392332e-06,
+      "loss": 0.7203,
+      "step": 880
+    },
+    {
+      "epoch": 0.30933988764044945,
+      "grad_norm": 0.9038983583450317,
+      "learning_rate": 4.973970115777067e-06,
+      "loss": 0.7122,
+      "step": 881
+    },
+    {
+      "epoch": 0.30969101123595505,
+      "grad_norm": 0.96694415807724,
+      "learning_rate": 4.973903531554141e-06,
+      "loss": 0.7292,
+      "step": 882
+    },
+    {
+      "epoch": 0.31004213483146065,
+      "grad_norm": 0.9090932011604309,
+      "learning_rate": 4.973836862725831e-06,
+      "loss": 0.7463,
+      "step": 883
+    },
+    {
+      "epoch": 0.3103932584269663,
+      "grad_norm": 0.9024604558944702,
+      "learning_rate": 4.973770109294416e-06,
+      "loss": 0.7488,
+      "step": 884
+    },
+    {
+      "epoch": 0.3107443820224719,
+      "grad_norm": 0.9796269536018372,
+      "learning_rate": 4.973703271262181e-06,
+      "loss": 0.7398,
+      "step": 885
+    },
+    {
+      "epoch": 0.31109550561797755,
+      "grad_norm": 0.9187832474708557,
+      "learning_rate": 4.973636348631412e-06,
+      "loss": 0.7162,
+      "step": 886
+    },
+    {
+      "epoch": 0.31144662921348315,
+      "grad_norm": 0.8963474631309509,
+      "learning_rate": 4.973569341404395e-06,
+      "loss": 0.6812,
+      "step": 887
+    },
+    {
+      "epoch": 0.31179775280898875,
+      "grad_norm": 0.934047281742096,
+      "learning_rate": 4.973502249583424e-06,
+      "loss": 0.7176,
+      "step": 888
+    },
+    {
+      "epoch": 0.3121488764044944,
+      "grad_norm": 0.9109435677528381,
+      "learning_rate": 4.9734350731707925e-06,
+      "loss": 0.684,
+      "step": 889
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.9629603028297424,
+      "learning_rate": 4.973367812168798e-06,
+      "loss": 0.756,
+      "step": 890
+    },
+    {
+      "epoch": 0.3128511235955056,
+      "grad_norm": 0.8999819159507751,
+      "learning_rate": 4.973300466579741e-06,
+      "loss": 0.7211,
+      "step": 891
+    },
+    {
+      "epoch": 0.31320224719101125,
+      "grad_norm": 0.9093006253242493,
+      "learning_rate": 4.9732330364059245e-06,
+      "loss": 0.6893,
+      "step": 892
+    },
+    {
+      "epoch": 0.31355337078651685,
+      "grad_norm": 0.925269603729248,
+      "learning_rate": 4.973165521649655e-06,
+      "loss": 0.6631,
+      "step": 893
+    },
+    {
+      "epoch": 0.31390449438202245,
+      "grad_norm": 0.9515416622161865,
+      "learning_rate": 4.973097922313239e-06,
+      "loss": 0.7578,
+      "step": 894
+    },
+    {
+      "epoch": 0.3142556179775281,
+      "grad_norm": 0.9192783236503601,
+      "learning_rate": 4.973030238398992e-06,
+      "loss": 0.6884,
+      "step": 895
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 0.905805230140686,
+      "learning_rate": 4.972962469909227e-06,
+      "loss": 0.7037,
+      "step": 896
+    },
+    {
+      "epoch": 0.31495786516853935,
+      "grad_norm": 0.9037303328514099,
+      "learning_rate": 4.972894616846261e-06,
+      "loss": 0.6989,
+      "step": 897
+    },
+    {
+      "epoch": 0.31530898876404495,
+      "grad_norm": 0.8707857728004456,
+      "learning_rate": 4.9728266792124155e-06,
+      "loss": 0.6865,
+      "step": 898
+    },
+    {
+      "epoch": 0.31566011235955055,
+      "grad_norm": 0.9623267650604248,
+      "learning_rate": 4.972758657010014e-06,
+      "loss": 0.7028,
+      "step": 899
+    },
+    {
+      "epoch": 0.3160112359550562,
+      "grad_norm": 0.9062806367874146,
+      "learning_rate": 4.972690550241383e-06,
+      "loss": 0.6803,
+      "step": 900
+    },
+    {
+      "epoch": 0.3163623595505618,
+      "grad_norm": 0.9363368153572083,
+      "learning_rate": 4.97262235890885e-06,
+      "loss": 0.7359,
+      "step": 901
+    },
+    {
+      "epoch": 0.3167134831460674,
+      "grad_norm": 0.9362224340438843,
+      "learning_rate": 4.972554083014748e-06,
+      "loss": 0.7432,
+      "step": 902
+    },
+    {
+      "epoch": 0.31706460674157305,
+      "grad_norm": 0.9579148292541504,
+      "learning_rate": 4.972485722561412e-06,
+      "loss": 0.7348,
+      "step": 903
+    },
+    {
+      "epoch": 0.31741573033707865,
+      "grad_norm": 0.9559216499328613,
+      "learning_rate": 4.97241727755118e-06,
+      "loss": 0.7095,
+      "step": 904
+    },
+    {
+      "epoch": 0.31776685393258425,
+      "grad_norm": 0.9043166637420654,
+      "learning_rate": 4.972348747986394e-06,
+      "loss": 0.7289,
+      "step": 905
+    },
+    {
+      "epoch": 0.3181179775280899,
+      "grad_norm": 0.925491988658905,
+      "learning_rate": 4.972280133869396e-06,
+      "loss": 0.6921,
+      "step": 906
+    },
+    {
+      "epoch": 0.3184691011235955,
+      "grad_norm": 0.9442777037620544,
+      "learning_rate": 4.972211435202533e-06,
+      "loss": 0.7104,
+      "step": 907
+    },
+    {
+      "epoch": 0.3188202247191011,
+      "grad_norm": 0.943748950958252,
+      "learning_rate": 4.972142651988153e-06,
+      "loss": 0.7688,
+      "step": 908
+    },
+    {
+      "epoch": 0.31917134831460675,
+      "grad_norm": 0.8944805860519409,
+      "learning_rate": 4.972073784228612e-06,
+      "loss": 0.7321,
+      "step": 909
+    },
+    {
+      "epoch": 0.31952247191011235,
+      "grad_norm": 0.9088547825813293,
+      "learning_rate": 4.972004831926261e-06,
+      "loss": 0.6973,
+      "step": 910
+    },
+    {
+      "epoch": 0.319873595505618,
+      "grad_norm": 0.9270226359367371,
+      "learning_rate": 4.971935795083461e-06,
+      "loss": 0.6956,
+      "step": 911
+    },
+    {
+      "epoch": 0.3202247191011236,
+      "grad_norm": 0.910492479801178,
+      "learning_rate": 4.971866673702573e-06,
+      "loss": 0.681,
+      "step": 912
+    },
+    {
+      "epoch": 0.3205758426966292,
+      "grad_norm": 0.9591485857963562,
+      "learning_rate": 4.971797467785958e-06,
+      "loss": 0.7763,
+      "step": 913
+    },
+    {
+      "epoch": 0.32092696629213485,
+      "grad_norm": 0.9132441878318787,
+      "learning_rate": 4.971728177335986e-06,
+      "loss": 0.7294,
+      "step": 914
+    },
+    {
+      "epoch": 0.32127808988764045,
+      "grad_norm": 0.9387941956520081,
+      "learning_rate": 4.971658802355025e-06,
+      "loss": 0.755,
+      "step": 915
+    },
+    {
+      "epoch": 0.32162921348314605,
+      "grad_norm": 0.930505096912384,
+      "learning_rate": 4.971589342845449e-06,
+      "loss": 0.7337,
+      "step": 916
+    },
+    {
+      "epoch": 0.3219803370786517,
+      "grad_norm": 0.8818848133087158,
+      "learning_rate": 4.971519798809631e-06,
+      "loss": 0.7114,
+      "step": 917
+    },
+    {
+      "epoch": 0.3223314606741573,
+      "grad_norm": 0.9763118028640747,
+      "learning_rate": 4.971450170249952e-06,
+      "loss": 0.73,
+      "step": 918
+    },
+    {
+      "epoch": 0.3226825842696629,
+      "grad_norm": 0.865503191947937,
+      "learning_rate": 4.971380457168791e-06,
+      "loss": 0.6733,
+      "step": 919
+    },
+    {
+      "epoch": 0.32303370786516855,
+      "grad_norm": 0.9227094650268555,
+      "learning_rate": 4.9713106595685336e-06,
+      "loss": 0.6975,
+      "step": 920
+    },
+    {
+      "epoch": 0.32338483146067415,
+      "grad_norm": 0.9164808988571167,
+      "learning_rate": 4.971240777451566e-06,
+      "loss": 0.7266,
+      "step": 921
+    },
+    {
+      "epoch": 0.3237359550561798,
+      "grad_norm": 0.9115430116653442,
+      "learning_rate": 4.971170810820279e-06,
+      "loss": 0.6775,
+      "step": 922
+    },
+    {
+      "epoch": 0.3240870786516854,
+      "grad_norm": 0.9133177399635315,
+      "learning_rate": 4.971100759677064e-06,
+      "loss": 0.7163,
+      "step": 923
+    },
+    {
+      "epoch": 0.324438202247191,
+      "grad_norm": 0.9289098381996155,
+      "learning_rate": 4.971030624024319e-06,
+      "loss": 0.7448,
+      "step": 924
+    },
+    {
+      "epoch": 0.32478932584269665,
+      "grad_norm": 0.872006356716156,
+      "learning_rate": 4.97096040386444e-06,
+      "loss": 0.6492,
+      "step": 925
+    },
+    {
+      "epoch": 0.32514044943820225,
+      "grad_norm": 0.9258365035057068,
+      "learning_rate": 4.97089009919983e-06,
+      "loss": 0.6931,
+      "step": 926
+    },
+    {
+      "epoch": 0.32549157303370785,
+      "grad_norm": 0.9044914841651917,
+      "learning_rate": 4.970819710032893e-06,
+      "loss": 0.7002,
+      "step": 927
+    },
+    {
+      "epoch": 0.3258426966292135,
+      "grad_norm": 0.9293612837791443,
+      "learning_rate": 4.970749236366037e-06,
+      "loss": 0.7471,
+      "step": 928
+    },
+    {
+      "epoch": 0.3261938202247191,
+      "grad_norm": 0.8999515771865845,
+      "learning_rate": 4.9706786782016706e-06,
+      "loss": 0.6992,
+      "step": 929
+    },
+    {
+      "epoch": 0.3265449438202247,
+      "grad_norm": 0.9285963177680969,
+      "learning_rate": 4.970608035542207e-06,
+      "loss": 0.7229,
+      "step": 930
+    },
+    {
+      "epoch": 0.32689606741573035,
+      "grad_norm": 0.9883245229721069,
+      "learning_rate": 4.970537308390063e-06,
+      "loss": 0.7343,
+      "step": 931
+    },
+    {
+      "epoch": 0.32724719101123595,
+      "grad_norm": 0.9650030732154846,
+      "learning_rate": 4.970466496747658e-06,
+      "loss": 0.7206,
+      "step": 932
+    },
+    {
+      "epoch": 0.32759831460674155,
+      "grad_norm": 0.9341951012611389,
+      "learning_rate": 4.9703956006174125e-06,
+      "loss": 0.7029,
+      "step": 933
+    },
+    {
+      "epoch": 0.3279494382022472,
+      "grad_norm": 0.9105697274208069,
+      "learning_rate": 4.9703246200017504e-06,
+      "loss": 0.7127,
+      "step": 934
+    },
+    {
+      "epoch": 0.3283005617977528,
+      "grad_norm": 0.8891416192054749,
+      "learning_rate": 4.970253554903102e-06,
+      "loss": 0.6952,
+      "step": 935
+    },
+    {
+      "epoch": 0.32865168539325845,
+      "grad_norm": 0.9152025580406189,
+      "learning_rate": 4.970182405323894e-06,
+      "loss": 0.7126,
+      "step": 936
+    },
+    {
+      "epoch": 0.32900280898876405,
+      "grad_norm": 0.9279428720474243,
+      "learning_rate": 4.970111171266563e-06,
+      "loss": 0.7066,
+      "step": 937
+    },
+    {
+      "epoch": 0.32935393258426965,
+      "grad_norm": 0.9324163198471069,
+      "learning_rate": 4.970039852733542e-06,
+      "loss": 0.6796,
+      "step": 938
+    },
+    {
+      "epoch": 0.3297050561797753,
+      "grad_norm": 0.9290011525154114,
+      "learning_rate": 4.9699684497272736e-06,
+      "loss": 0.7192,
+      "step": 939
+    },
+    {
+      "epoch": 0.3300561797752809,
+      "grad_norm": 0.8959920406341553,
+      "learning_rate": 4.969896962250196e-06,
+      "loss": 0.7132,
+      "step": 940
+    },
+    {
+      "epoch": 0.3304073033707865,
+      "grad_norm": 0.8849387168884277,
+      "learning_rate": 4.969825390304757e-06,
+      "loss": 0.6618,
+      "step": 941
+    },
+    {
+      "epoch": 0.33075842696629215,
+      "grad_norm": 0.8804863691329956,
+      "learning_rate": 4.969753733893403e-06,
+      "loss": 0.7021,
+      "step": 942
+    },
+    {
+      "epoch": 0.33110955056179775,
+      "grad_norm": 0.9022026062011719,
+      "learning_rate": 4.969681993018584e-06,
+      "loss": 0.6842,
+      "step": 943
+    },
+    {
+      "epoch": 0.33146067415730335,
+      "grad_norm": 0.9250811338424683,
+      "learning_rate": 4.969610167682754e-06,
+      "loss": 0.6977,
+      "step": 944
+    },
+    {
+      "epoch": 0.331811797752809,
+      "grad_norm": 0.9264232516288757,
+      "learning_rate": 4.969538257888371e-06,
+      "loss": 0.7135,
+      "step": 945
+    },
+    {
+      "epoch": 0.3321629213483146,
+      "grad_norm": 0.9282289147377014,
+      "learning_rate": 4.969466263637892e-06,
+      "loss": 0.7304,
+      "step": 946
+    },
+    {
+      "epoch": 0.3325140449438202,
+      "grad_norm": 0.8798340559005737,
+      "learning_rate": 4.9693941849337804e-06,
+      "loss": 0.6996,
+      "step": 947
+    },
+    {
+      "epoch": 0.33286516853932585,
+      "grad_norm": 0.9778987169265747,
+      "learning_rate": 4.9693220217785e-06,
+      "loss": 0.7132,
+      "step": 948
+    },
+    {
+      "epoch": 0.33321629213483145,
+      "grad_norm": 0.9029282331466675,
+      "learning_rate": 4.969249774174519e-06,
+      "loss": 0.6951,
+      "step": 949
+    },
+    {
+      "epoch": 0.3335674157303371,
+      "grad_norm": 0.9430015087127686,
+      "learning_rate": 4.96917744212431e-06,
+      "loss": 0.6943,
+      "step": 950
+    },
+    {
+      "epoch": 0.3339185393258427,
+      "grad_norm": 0.9794940948486328,
+      "learning_rate": 4.9691050256303445e-06,
+      "loss": 0.7054,
+      "step": 951
+    },
+    {
+      "epoch": 0.3342696629213483,
+      "grad_norm": 0.9288370013237,
+      "learning_rate": 4.969032524695101e-06,
+      "loss": 0.7122,
+      "step": 952
+    },
+    {
+      "epoch": 0.33462078651685395,
+      "grad_norm": 0.9557552933692932,
+      "learning_rate": 4.968959939321057e-06,
+      "loss": 0.7598,
+      "step": 953
+    },
+    {
+      "epoch": 0.33497191011235955,
+      "grad_norm": 0.9375066161155701,
+      "learning_rate": 4.968887269510696e-06,
+      "loss": 0.6924,
+      "step": 954
+    },
+    {
+      "epoch": 0.33532303370786515,
+      "grad_norm": 0.9338163733482361,
+      "learning_rate": 4.968814515266503e-06,
+      "loss": 0.7267,
+      "step": 955
+    },
+    {
+      "epoch": 0.3356741573033708,
+      "grad_norm": 0.9154375791549683,
+      "learning_rate": 4.968741676590966e-06,
+      "loss": 0.7321,
+      "step": 956
+    },
+    {
+      "epoch": 0.3360252808988764,
+      "grad_norm": 0.9233065843582153,
+      "learning_rate": 4.968668753486576e-06,
+      "loss": 0.7245,
+      "step": 957
+    },
+    {
+      "epoch": 0.336376404494382,
+      "grad_norm": 0.9419113993644714,
+      "learning_rate": 4.968595745955828e-06,
+      "loss": 0.7007,
+      "step": 958
+    },
+    {
+      "epoch": 0.33672752808988765,
+      "grad_norm": 0.961121678352356,
+      "learning_rate": 4.968522654001217e-06,
+      "loss": 0.7308,
+      "step": 959
+    },
+    {
+      "epoch": 0.33707865168539325,
+      "grad_norm": 0.934877336025238,
+      "learning_rate": 4.9684494776252445e-06,
+      "loss": 0.7388,
+      "step": 960
+    },
+    {
+      "epoch": 0.3374297752808989,
+      "grad_norm": 0.912398636341095,
+      "learning_rate": 4.9683762168304115e-06,
+      "loss": 0.7088,
+      "step": 961
+    },
+    {
+      "epoch": 0.3377808988764045,
+      "grad_norm": 0.948833167552948,
+      "learning_rate": 4.9683028716192246e-06,
+      "loss": 0.7278,
+      "step": 962
+    },
+    {
+      "epoch": 0.3381320224719101,
+      "grad_norm": 0.9257057905197144,
+      "learning_rate": 4.968229441994191e-06,
+      "loss": 0.7338,
+      "step": 963
+    },
+    {
+      "epoch": 0.33848314606741575,
+      "grad_norm": 0.9129422903060913,
+      "learning_rate": 4.968155927957823e-06,
+      "loss": 0.7058,
+      "step": 964
+    },
+    {
+      "epoch": 0.33883426966292135,
+      "grad_norm": 0.9381750822067261,
+      "learning_rate": 4.968082329512634e-06,
+      "loss": 0.7326,
+      "step": 965
+    },
+    {
+      "epoch": 0.33918539325842695,
+      "grad_norm": 0.9108783602714539,
+      "learning_rate": 4.968008646661141e-06,
+      "loss": 0.66,
+      "step": 966
+    },
+    {
+      "epoch": 0.3395365168539326,
+      "grad_norm": 0.9329748153686523,
+      "learning_rate": 4.9679348794058655e-06,
+      "loss": 0.7117,
+      "step": 967
+    },
+    {
+      "epoch": 0.3398876404494382,
+      "grad_norm": 0.9643568992614746,
+      "learning_rate": 4.967861027749327e-06,
+      "loss": 0.7555,
+      "step": 968
+    },
+    {
+      "epoch": 0.3402387640449438,
+      "grad_norm": 0.9551818370819092,
+      "learning_rate": 4.967787091694055e-06,
+      "loss": 0.7474,
+      "step": 969
+    },
+    {
+      "epoch": 0.34058988764044945,
+      "grad_norm": 0.9381354451179504,
+      "learning_rate": 4.967713071242576e-06,
+      "loss": 0.7155,
+      "step": 970
+    },
+    {
+      "epoch": 0.34094101123595505,
+      "grad_norm": 0.9513474702835083,
+      "learning_rate": 4.967638966397421e-06,
+      "loss": 0.765,
+      "step": 971
+    },
+    {
+      "epoch": 0.34129213483146065,
+      "grad_norm": 0.9263571500778198,
+      "learning_rate": 4.967564777161124e-06,
+      "loss": 0.7112,
+      "step": 972
+    },
+    {
+      "epoch": 0.3416432584269663,
+      "grad_norm": 0.9278532862663269,
+      "learning_rate": 4.967490503536225e-06,
+      "loss": 0.7183,
+      "step": 973
+    },
+    {
+      "epoch": 0.3419943820224719,
+      "grad_norm": 0.9487691521644592,
+      "learning_rate": 4.967416145525261e-06,
+      "loss": 0.7312,
+      "step": 974
+    },
+    {
+      "epoch": 0.34234550561797755,
+      "grad_norm": 0.8866797089576721,
+      "learning_rate": 4.967341703130777e-06,
+      "loss": 0.704,
+      "step": 975
+    },
+    {
+      "epoch": 0.34269662921348315,
+      "grad_norm": 0.9202519655227661,
+      "learning_rate": 4.967267176355318e-06,
+      "loss": 0.7434,
+      "step": 976
+    },
+    {
+      "epoch": 0.34304775280898875,
+      "grad_norm": 0.9155771732330322,
+      "learning_rate": 4.967192565201432e-06,
+      "loss": 0.7166,
+      "step": 977
+    },
+    {
+      "epoch": 0.3433988764044944,
+      "grad_norm": 0.9582788944244385,
+      "learning_rate": 4.967117869671672e-06,
+      "loss": 0.7536,
+      "step": 978
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 0.9524344801902771,
+      "learning_rate": 4.967043089768592e-06,
+      "loss": 0.7739,
+      "step": 979
+    },
+    {
+      "epoch": 0.3441011235955056,
+      "grad_norm": 0.9230315685272217,
+      "learning_rate": 4.96696822549475e-06,
+      "loss": 0.7181,
+      "step": 980
+    },
+    {
+      "epoch": 0.34445224719101125,
+      "grad_norm": 0.9080458283424377,
+      "learning_rate": 4.966893276852704e-06,
+      "loss": 0.7107,
+      "step": 981
+    },
+    {
+      "epoch": 0.34480337078651685,
+      "grad_norm": 0.9102587699890137,
+      "learning_rate": 4.966818243845021e-06,
+      "loss": 0.7051,
+      "step": 982
+    },
+    {
+      "epoch": 0.34515449438202245,
+      "grad_norm": 0.8925254940986633,
+      "learning_rate": 4.966743126474264e-06,
+      "loss": 0.7106,
+      "step": 983
+    },
+    {
+      "epoch": 0.3455056179775281,
+      "grad_norm": 0.9102403521537781,
+      "learning_rate": 4.966667924743002e-06,
+      "loss": 0.6849,
+      "step": 984
+    },
+    {
+      "epoch": 0.3458567415730337,
+      "grad_norm": 0.9164739847183228,
+      "learning_rate": 4.966592638653808e-06,
+      "loss": 0.7027,
+      "step": 985
+    },
+    {
+      "epoch": 0.34620786516853935,
+      "grad_norm": 0.89588463306427,
+      "learning_rate": 4.966517268209258e-06,
+      "loss": 0.7086,
+      "step": 986
+    },
+    {
+      "epoch": 0.34655898876404495,
+      "grad_norm": 0.8812805414199829,
+      "learning_rate": 4.966441813411928e-06,
+      "loss": 0.7073,
+      "step": 987
+    },
+    {
+      "epoch": 0.34691011235955055,
+      "grad_norm": 0.8935489654541016,
+      "learning_rate": 4.966366274264397e-06,
+      "loss": 0.6749,
+      "step": 988
+    },
+    {
+      "epoch": 0.3472612359550562,
+      "grad_norm": 0.935444712638855,
+      "learning_rate": 4.966290650769251e-06,
+      "loss": 0.7283,
+      "step": 989
+    },
+    {
+      "epoch": 0.3476123595505618,
+      "grad_norm": 0.9046710729598999,
+      "learning_rate": 4.966214942929074e-06,
+      "loss": 0.7138,
+      "step": 990
+    },
+    {
+      "epoch": 0.3479634831460674,
+      "grad_norm": 0.9328753352165222,
+      "learning_rate": 4.966139150746458e-06,
+      "loss": 0.7022,
+      "step": 991
+    },
+    {
+      "epoch": 0.34831460674157305,
+      "grad_norm": 0.9144490361213684,
+      "learning_rate": 4.966063274223992e-06,
+      "loss": 0.738,
+      "step": 992
+    },
+    {
+      "epoch": 0.34866573033707865,
+      "grad_norm": 0.9606062173843384,
+      "learning_rate": 4.965987313364272e-06,
+      "loss": 0.6937,
+      "step": 993
+    },
+    {
+      "epoch": 0.34901685393258425,
+      "grad_norm": 0.94340980052948,
+      "learning_rate": 4.965911268169897e-06,
+      "loss": 0.6811,
+      "step": 994
+    },
+    {
+      "epoch": 0.3493679775280899,
+      "grad_norm": 0.8755514025688171,
+      "learning_rate": 4.9658351386434665e-06,
+      "loss": 0.6707,
+      "step": 995
+    },
+    {
+      "epoch": 0.3497191011235955,
+      "grad_norm": 0.8640532493591309,
+      "learning_rate": 4.965758924787584e-06,
+      "loss": 0.6378,
+      "step": 996
+    },
+    {
+      "epoch": 0.3500702247191011,
+      "grad_norm": 0.9261877536773682,
+      "learning_rate": 4.965682626604856e-06,
+      "loss": 0.6953,
+      "step": 997
+    },
+    {
+      "epoch": 0.35042134831460675,
+      "grad_norm": 0.8998399376869202,
+      "learning_rate": 4.965606244097894e-06,
+      "loss": 0.7154,
+      "step": 998
+    },
+    {
+      "epoch": 0.35077247191011235,
+      "grad_norm": 0.9440179467201233,
+      "learning_rate": 4.9655297772693066e-06,
+      "loss": 0.6992,
+      "step": 999
+    },
+    {
+      "epoch": 0.351123595505618,
+      "grad_norm": 0.9105573892593384,
+      "learning_rate": 4.965453226121711e-06,
+      "loss": 0.7033,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3514747191011236,
+      "grad_norm": 0.9109785556793213,
+      "learning_rate": 4.965376590657724e-06,
+      "loss": 0.7057,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3518258426966292,
+      "grad_norm": 0.9462383985519409,
+      "learning_rate": 4.9652998708799684e-06,
+      "loss": 0.6979,
+      "step": 1002
+    },
+    {
+      "epoch": 0.35217696629213485,
+      "grad_norm": 0.9546105861663818,
+      "learning_rate": 4.965223066791066e-06,
+      "loss": 0.7227,
+      "step": 1003
+    },
+    {
+      "epoch": 0.35252808988764045,
+      "grad_norm": 1.0831300020217896,
+      "learning_rate": 4.965146178393645e-06,
+      "loss": 0.6849,
+      "step": 1004
+    },
+    {
+      "epoch": 0.35287921348314605,
+      "grad_norm": 0.9323068261146545,
+      "learning_rate": 4.965069205690334e-06,
+      "loss": 0.739,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3532303370786517,
+      "grad_norm": 0.9407396912574768,
+      "learning_rate": 4.964992148683765e-06,
+      "loss": 0.7386,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3535814606741573,
+      "grad_norm": 0.9513671398162842,
+      "learning_rate": 4.964915007376574e-06,
+      "loss": 0.666,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3539325842696629,
+      "grad_norm": 0.9194923043251038,
+      "learning_rate": 4.964837781771399e-06,
+      "loss": 0.7601,
+      "step": 1008
+    },
+    {
+      "epoch": 0.35428370786516855,
+      "grad_norm": 0.9639756083488464,
+      "learning_rate": 4.964760471870882e-06,
+      "loss": 0.7124,
+      "step": 1009
+    },
+    {
+      "epoch": 0.35463483146067415,
+      "grad_norm": 0.9628319144248962,
+      "learning_rate": 4.964683077677664e-06,
+      "loss": 0.7003,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3549859550561798,
+      "grad_norm": 0.9003061652183533,
+      "learning_rate": 4.964605599194395e-06,
+      "loss": 0.7199,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3553370786516854,
+      "grad_norm": 0.9365134835243225,
+      "learning_rate": 4.964528036423723e-06,
+      "loss": 0.6651,
+      "step": 1012
+    },
+    {
+      "epoch": 0.355688202247191,
+      "grad_norm": 0.9309297204017639,
+      "learning_rate": 4.964450389368303e-06,
+      "loss": 0.7354,
+      "step": 1013
+    },
+    {
+      "epoch": 0.35603932584269665,
+      "grad_norm": 0.9624804258346558,
+      "learning_rate": 4.964372658030786e-06,
+      "loss": 0.6955,
+      "step": 1014
+    },
+    {
+      "epoch": 0.35639044943820225,
+      "grad_norm": 0.9497263431549072,
+      "learning_rate": 4.964294842413835e-06,
+      "loss": 0.7374,
+      "step": 1015
+    },
+    {
+      "epoch": 0.35674157303370785,
+      "grad_norm": 0.9235740900039673,
+      "learning_rate": 4.964216942520108e-06,
+      "loss": 0.6906,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3570926966292135,
+      "grad_norm": 0.9985111951828003,
+      "learning_rate": 4.9641389583522705e-06,
+      "loss": 0.7131,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3574438202247191,
+      "grad_norm": 0.9399976134300232,
+      "learning_rate": 4.964060889912988e-06,
+      "loss": 0.6935,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3577949438202247,
+      "grad_norm": 0.9676353931427002,
+      "learning_rate": 4.9639827372049335e-06,
+      "loss": 0.7513,
+      "step": 1019
+    },
+    {
+      "epoch": 0.35814606741573035,
+      "grad_norm": 0.9015031456947327,
+      "learning_rate": 4.9639045002307765e-06,
+      "loss": 0.687,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35849719101123595,
+      "grad_norm": 0.9401846528053284,
+      "learning_rate": 4.963826178993195e-06,
+      "loss": 0.6963,
+      "step": 1021
+    },
+    {
+      "epoch": 0.35884831460674155,
+      "grad_norm": 1.00020170211792,
+      "learning_rate": 4.9637477734948654e-06,
+      "loss": 0.6747,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3591994382022472,
+      "grad_norm": 0.8997200727462769,
+      "learning_rate": 4.963669283738471e-06,
+      "loss": 0.7253,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 0.8761312365531921,
+      "learning_rate": 4.963590709726695e-06,
+      "loss": 0.6832,
+      "step": 1024
+    },
+    {
+      "epoch": 0.35990168539325845,
+      "grad_norm": 0.9415677189826965,
+      "learning_rate": 4.963512051462226e-06,
+      "loss": 0.6998,
+      "step": 1025
+    },
+    {
+      "epoch": 0.36025280898876405,
+      "grad_norm": 0.9139187932014465,
+      "learning_rate": 4.963433308947751e-06,
+      "loss": 0.7292,
+      "step": 1026
+    },
+    {
+      "epoch": 0.36060393258426965,
+      "grad_norm": 0.9207375049591064,
+      "learning_rate": 4.963354482185966e-06,
+      "loss": 0.6779,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3609550561797753,
+      "grad_norm": 0.9517902135848999,
+      "learning_rate": 4.963275571179565e-06,
+      "loss": 0.7525,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3613061797752809,
+      "grad_norm": 0.9088241457939148,
+      "learning_rate": 4.963196575931248e-06,
+      "loss": 0.7196,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3616573033707865,
+      "grad_norm": 0.8898507356643677,
+      "learning_rate": 4.963117496443715e-06,
+      "loss": 0.7205,
+      "step": 1030
+    },
+    {
+      "epoch": 0.36200842696629215,
+      "grad_norm": 0.9045488834381104,
+      "learning_rate": 4.963038332719672e-06,
+      "loss": 0.6937,
+      "step": 1031
+    },
+    {
+      "epoch": 0.36235955056179775,
+      "grad_norm": 0.9505866169929504,
+      "learning_rate": 4.962959084761826e-06,
+      "loss": 0.7006,
+      "step": 1032
+    },
+    {
+      "epoch": 0.36271067415730335,
+      "grad_norm": 0.8908640742301941,
+      "learning_rate": 4.962879752572885e-06,
+      "loss": 0.7191,
+      "step": 1033
+    },
+    {
+      "epoch": 0.363061797752809,
+      "grad_norm": 0.925959050655365,
+      "learning_rate": 4.962800336155567e-06,
+      "loss": 0.6963,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3634129213483146,
+      "grad_norm": 0.9082480072975159,
+      "learning_rate": 4.962720835512583e-06,
+      "loss": 0.7084,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3637640449438202,
+      "grad_norm": 0.9313839077949524,
+      "learning_rate": 4.962641250646654e-06,
+      "loss": 0.7425,
+      "step": 1036
+    },
+    {
+      "epoch": 0.36411516853932585,
+      "grad_norm": 0.9484413266181946,
+      "learning_rate": 4.962561581560501e-06,
+      "loss": 0.6814,
+      "step": 1037
+    },
+    {
+      "epoch": 0.36446629213483145,
+      "grad_norm": 0.9550908207893372,
+      "learning_rate": 4.96248182825685e-06,
+      "loss": 0.6747,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3648174157303371,
+      "grad_norm": 0.9393624663352966,
+      "learning_rate": 4.962401990738428e-06,
+      "loss": 0.7172,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3651685393258427,
+      "grad_norm": 0.961174726486206,
+      "learning_rate": 4.962322069007964e-06,
+      "loss": 0.6967,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3655196629213483,
+      "grad_norm": 0.9074354767799377,
+      "learning_rate": 4.962242063068193e-06,
+      "loss": 0.7019,
+      "step": 1041
+    },
+    {
+      "epoch": 0.36587078651685395,
+      "grad_norm": 0.9572650194168091,
+      "learning_rate": 4.962161972921851e-06,
+      "loss": 0.717,
+      "step": 1042
+    },
+    {
+      "epoch": 0.36622191011235955,
+      "grad_norm": 0.9262009263038635,
+      "learning_rate": 4.962081798571675e-06,
+      "loss": 0.7332,
+      "step": 1043
+    },
+    {
+      "epoch": 0.36657303370786515,
+      "grad_norm": 0.8846391439437866,
+      "learning_rate": 4.9620015400204094e-06,
+      "loss": 0.7359,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3669241573033708,
+      "grad_norm": 0.9351332187652588,
+      "learning_rate": 4.961921197270797e-06,
+      "loss": 0.7809,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3672752808988764,
+      "grad_norm": 0.9081857204437256,
+      "learning_rate": 4.961840770325587e-06,
+      "loss": 0.7169,
+      "step": 1046
+    },
+    {
+      "epoch": 0.367626404494382,
+      "grad_norm": 0.9414200782775879,
+      "learning_rate": 4.961760259187529e-06,
+      "loss": 0.7154,
+      "step": 1047
+    },
+    {
+      "epoch": 0.36797752808988765,
+      "grad_norm": 0.9009916186332703,
+      "learning_rate": 4.961679663859377e-06,
+      "loss": 0.7013,
+      "step": 1048
+    },
+    {
+      "epoch": 0.36832865168539325,
+      "grad_norm": 0.9196215271949768,
+      "learning_rate": 4.961598984343886e-06,
+      "loss": 0.706,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3686797752808989,
+      "grad_norm": 0.947223424911499,
+      "learning_rate": 4.961518220643817e-06,
+      "loss": 0.712,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3690308988764045,
+      "grad_norm": 0.8798839449882507,
+      "learning_rate": 4.961437372761932e-06,
+      "loss": 0.6821,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3693820224719101,
+      "grad_norm": 0.9458513259887695,
+      "learning_rate": 4.961356440700994e-06,
+      "loss": 0.7503,
+      "step": 1052
+    },
+    {
+      "epoch": 0.36973314606741575,
+      "grad_norm": 0.9466217160224915,
+      "learning_rate": 4.961275424463772e-06,
+      "loss": 0.7086,
+      "step": 1053
+    },
+    {
+      "epoch": 0.37008426966292135,
+      "grad_norm": 0.9832736849784851,
+      "learning_rate": 4.961194324053037e-06,
+      "loss": 0.7332,
+      "step": 1054
+    },
+    {
+      "epoch": 0.37043539325842695,
+      "grad_norm": 0.9869272708892822,
+      "learning_rate": 4.961113139471562e-06,
+      "loss": 0.7595,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3707865168539326,
+      "grad_norm": 0.9590395092964172,
+      "learning_rate": 4.961031870722123e-06,
+      "loss": 0.7471,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3711376404494382,
+      "grad_norm": 0.9194419980049133,
+      "learning_rate": 4.9609505178075015e-06,
+      "loss": 0.7147,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3714887640449438,
+      "grad_norm": 0.8812661170959473,
+      "learning_rate": 4.960869080730477e-06,
+      "loss": 0.6723,
+      "step": 1058
+    },
+    {
+      "epoch": 0.37183988764044945,
+      "grad_norm": 0.9303286671638489,
+      "learning_rate": 4.960787559493836e-06,
+      "loss": 0.7253,
+      "step": 1059
+    },
+    {
+      "epoch": 0.37219101123595505,
+      "grad_norm": 0.9536207318305969,
+      "learning_rate": 4.960705954100366e-06,
+      "loss": 0.7325,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37254213483146065,
+      "grad_norm": 0.9165772795677185,
+      "learning_rate": 4.960624264552858e-06,
+      "loss": 0.6949,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3728932584269663,
+      "grad_norm": 0.9144523739814758,
+      "learning_rate": 4.9605424908541065e-06,
+      "loss": 0.7355,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3732443820224719,
+      "grad_norm": 0.8873782753944397,
+      "learning_rate": 4.960460633006906e-06,
+      "loss": 0.6819,
+      "step": 1063
+    },
+    {
+      "epoch": 0.37359550561797755,
+      "grad_norm": 0.9444409012794495,
+      "learning_rate": 4.960378691014058e-06,
+      "loss": 0.6851,
+      "step": 1064
+    },
+    {
+      "epoch": 0.37394662921348315,
+      "grad_norm": 0.959851861000061,
+      "learning_rate": 4.960296664878364e-06,
+      "loss": 0.7127,
+      "step": 1065
+    },
+    {
+      "epoch": 0.37429775280898875,
+      "grad_norm": 0.9353641867637634,
+      "learning_rate": 4.9602145546026305e-06,
+      "loss": 0.7208,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3746488764044944,
+      "grad_norm": 0.9105174541473389,
+      "learning_rate": 4.960132360189663e-06,
+      "loss": 0.6717,
+      "step": 1067
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.9150325655937195,
+      "learning_rate": 4.960050081642274e-06,
+      "loss": 0.7233,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3753511235955056,
+      "grad_norm": 0.9387450814247131,
+      "learning_rate": 4.959967718963279e-06,
+      "loss": 0.7104,
+      "step": 1069
+    },
+    {
+      "epoch": 0.37570224719101125,
+      "grad_norm": 0.9268593788146973,
+      "learning_rate": 4.959885272155491e-06,
+      "loss": 0.7209,
+      "step": 1070
+    },
+    {
+      "epoch": 0.37605337078651685,
+      "grad_norm": 0.9124265909194946,
+      "learning_rate": 4.959802741221733e-06,
+      "loss": 0.6982,
+      "step": 1071
+    },
+    {
+      "epoch": 0.37640449438202245,
+      "grad_norm": 0.9302802681922913,
+      "learning_rate": 4.9597201261648255e-06,
+      "loss": 0.7105,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3767556179775281,
+      "grad_norm": 0.9560679793357849,
+      "learning_rate": 4.959637426987595e-06,
+      "loss": 0.7115,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3771067415730337,
+      "grad_norm": 0.9246500730514526,
+      "learning_rate": 4.9595546436928685e-06,
+      "loss": 0.7319,
+      "step": 1074
+    },
+    {
+      "epoch": 0.37745786516853935,
+      "grad_norm": 0.9188322424888611,
+      "learning_rate": 4.959471776283479e-06,
+      "loss": 0.6829,
+      "step": 1075
+    },
+    {
+      "epoch": 0.37780898876404495,
+      "grad_norm": 0.9240451455116272,
+      "learning_rate": 4.959388824762258e-06,
+      "loss": 0.7183,
+      "step": 1076
+    },
+    {
+      "epoch": 0.37816011235955055,
+      "grad_norm": 0.8934212923049927,
+      "learning_rate": 4.959305789132044e-06,
+      "loss": 0.6982,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3785112359550562,
+      "grad_norm": 0.9386299848556519,
+      "learning_rate": 4.959222669395677e-06,
+      "loss": 0.6942,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3788623595505618,
+      "grad_norm": 0.9435991644859314,
+      "learning_rate": 4.959139465555999e-06,
+      "loss": 0.7382,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3792134831460674,
+      "grad_norm": 0.9004935026168823,
+      "learning_rate": 4.959056177615855e-06,
+      "loss": 0.6944,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37956460674157305,
+      "grad_norm": 0.9269630312919617,
+      "learning_rate": 4.958972805578095e-06,
+      "loss": 0.7037,
+      "step": 1081
+    },
+    {
+      "epoch": 0.37991573033707865,
+      "grad_norm": 0.9682616591453552,
+      "learning_rate": 4.9588893494455684e-06,
+      "loss": 0.6874,
+      "step": 1082
+    },
+    {
+      "epoch": 0.38026685393258425,
+      "grad_norm": 0.8845170140266418,
+      "learning_rate": 4.958805809221131e-06,
+      "loss": 0.6945,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3806179775280899,
+      "grad_norm": 0.940234363079071,
+      "learning_rate": 4.958722184907638e-06,
+      "loss": 0.6813,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3809691011235955,
+      "grad_norm": 0.9013071656227112,
+      "learning_rate": 4.95863847650795e-06,
+      "loss": 0.694,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3813202247191011,
+      "grad_norm": 0.8996568322181702,
+      "learning_rate": 4.958554684024931e-06,
+      "loss": 0.6558,
+      "step": 1086
+    },
+    {
+      "epoch": 0.38167134831460675,
+      "grad_norm": 0.9193875193595886,
+      "learning_rate": 4.9584708074614455e-06,
+      "loss": 0.6996,
+      "step": 1087
+    },
+    {
+      "epoch": 0.38202247191011235,
+      "grad_norm": 0.9180407524108887,
+      "learning_rate": 4.958386846820361e-06,
+      "loss": 0.7147,
+      "step": 1088
+    },
+    {
+      "epoch": 0.382373595505618,
+      "grad_norm": 1.0345251560211182,
+      "learning_rate": 4.958302802104551e-06,
+      "loss": 0.7393,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3827247191011236,
+      "grad_norm": 0.9022161364555359,
+      "learning_rate": 4.958218673316889e-06,
+      "loss": 0.6949,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3830758426966292,
+      "grad_norm": 0.9014359712600708,
+      "learning_rate": 4.958134460460251e-06,
+      "loss": 0.6947,
+      "step": 1091
+    },
+    {
+      "epoch": 0.38342696629213485,
+      "grad_norm": 0.9596666693687439,
+      "learning_rate": 4.958050163537519e-06,
+      "loss": 0.7763,
+      "step": 1092
+    },
+    {
+      "epoch": 0.38377808988764045,
+      "grad_norm": 0.8888493180274963,
+      "learning_rate": 4.957965782551574e-06,
+      "loss": 0.7093,
+      "step": 1093
+    },
+    {
+      "epoch": 0.38412921348314605,
+      "grad_norm": 0.9068475365638733,
+      "learning_rate": 4.957881317505303e-06,
+      "loss": 0.6893,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3844803370786517,
+      "grad_norm": 0.8918416500091553,
+      "learning_rate": 4.957796768401594e-06,
+      "loss": 0.6926,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3848314606741573,
+      "grad_norm": 0.8772871494293213,
+      "learning_rate": 4.957712135243339e-06,
+      "loss": 0.6714,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3851825842696629,
+      "grad_norm": 0.9466787576675415,
+      "learning_rate": 4.9576274180334315e-06,
+      "loss": 0.693,
+      "step": 1097
+    },
+    {
+      "epoch": 0.38553370786516855,
+      "grad_norm": 0.8940508961677551,
+      "learning_rate": 4.957542616774768e-06,
+      "loss": 0.6727,
+      "step": 1098
+    },
+    {
+      "epoch": 0.38588483146067415,
+      "grad_norm": 0.92746901512146,
+      "learning_rate": 4.957457731470253e-06,
+      "loss": 0.6832,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3862359550561798,
+      "grad_norm": 0.9406936168670654,
+      "learning_rate": 4.9573727621227845e-06,
+      "loss": 0.7121,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3865870786516854,
+      "grad_norm": 0.9280776381492615,
+      "learning_rate": 4.957287708735271e-06,
+      "loss": 0.7149,
+      "step": 1101
+    },
+    {
+      "epoch": 0.386938202247191,
+      "grad_norm": 0.9443319439888,
+      "learning_rate": 4.9572025713106195e-06,
+      "loss": 0.7205,
+      "step": 1102
+    },
+    {
+      "epoch": 0.38728932584269665,
+      "grad_norm": 0.9101063013076782,
+      "learning_rate": 4.957117349851745e-06,
+      "loss": 0.703,
+      "step": 1103
+    },
+    {
+      "epoch": 0.38764044943820225,
+      "grad_norm": 0.8856722116470337,
+      "learning_rate": 4.957032044361557e-06,
+      "loss": 0.7079,
+      "step": 1104
+    },
+    {
+      "epoch": 0.38799157303370785,
+      "grad_norm": 0.928821325302124,
+      "learning_rate": 4.956946654842977e-06,
+      "loss": 0.6935,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3883426966292135,
+      "grad_norm": 0.9314661026000977,
+      "learning_rate": 4.956861181298924e-06,
+      "loss": 0.7062,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3886938202247191,
+      "grad_norm": 0.9111589193344116,
+      "learning_rate": 4.956775623732321e-06,
+      "loss": 0.7343,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3890449438202247,
+      "grad_norm": 0.8977031707763672,
+      "learning_rate": 4.9566899821460935e-06,
+      "loss": 0.6702,
+      "step": 1108
+    },
+    {
+      "epoch": 0.38939606741573035,
+      "grad_norm": 0.9246605634689331,
+      "learning_rate": 4.956604256543171e-06,
+      "loss": 0.7062,
+      "step": 1109
+    },
+    {
+      "epoch": 0.38974719101123595,
+      "grad_norm": 0.9252159595489502,
+      "learning_rate": 4.956518446926485e-06,
+      "loss": 0.6981,
+      "step": 1110
+    },
+    {
+      "epoch": 0.39009831460674155,
+      "grad_norm": 0.9128794074058533,
+      "learning_rate": 4.956432553298971e-06,
+      "loss": 0.6802,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3904494382022472,
+      "grad_norm": 0.9178376197814941,
+      "learning_rate": 4.956346575663564e-06,
+      "loss": 0.6753,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3908005617977528,
+      "grad_norm": 0.8804501891136169,
+      "learning_rate": 4.956260514023207e-06,
+      "loss": 0.7378,
+      "step": 1113
+    },
+    {
+      "epoch": 0.39115168539325845,
+      "grad_norm": 0.9001629948616028,
+      "learning_rate": 4.956174368380842e-06,
+      "loss": 0.7061,
+      "step": 1114
+    },
+    {
+      "epoch": 0.39150280898876405,
+      "grad_norm": 0.8731476664543152,
+      "learning_rate": 4.956088138739416e-06,
+      "loss": 0.6749,
+      "step": 1115
+    },
+    {
+      "epoch": 0.39185393258426965,
+      "grad_norm": 0.9055764079093933,
+      "learning_rate": 4.956001825101877e-06,
+      "loss": 0.7172,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3922050561797753,
+      "grad_norm": 0.9398018717765808,
+      "learning_rate": 4.955915427471178e-06,
+      "loss": 0.711,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3925561797752809,
+      "grad_norm": 0.9070323705673218,
+      "learning_rate": 4.9558289458502715e-06,
+      "loss": 0.7467,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3929073033707865,
+      "grad_norm": 0.916070818901062,
+      "learning_rate": 4.9557423802421166e-06,
+      "loss": 0.6929,
+      "step": 1119
+    },
+    {
+      "epoch": 0.39325842696629215,
+      "grad_norm": 0.9212610721588135,
+      "learning_rate": 4.955655730649674e-06,
+      "loss": 0.7098,
+      "step": 1120
+    },
+    {
+      "epoch": 0.39360955056179775,
+      "grad_norm": 0.8813148736953735,
+      "learning_rate": 4.955568997075908e-06,
+      "loss": 0.7241,
+      "step": 1121
+    },
+    {
+      "epoch": 0.39396067415730335,
+      "grad_norm": 0.960132360458374,
+      "learning_rate": 4.955482179523782e-06,
+      "loss": 0.7531,
+      "step": 1122
+    },
+    {
+      "epoch": 0.394311797752809,
+      "grad_norm": 0.8893491625785828,
+      "learning_rate": 4.955395277996268e-06,
+      "loss": 0.708,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3946629213483146,
+      "grad_norm": 0.9316296577453613,
+      "learning_rate": 4.955308292496336e-06,
+      "loss": 0.7129,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3950140449438202,
+      "grad_norm": 0.8866797685623169,
+      "learning_rate": 4.955221223026961e-06,
+      "loss": 0.6986,
+      "step": 1125
+    },
+    {
+      "epoch": 0.39536516853932585,
+      "grad_norm": 0.9073642492294312,
+      "learning_rate": 4.955134069591121e-06,
+      "loss": 0.7192,
+      "step": 1126
+    },
+    {
+      "epoch": 0.39571629213483145,
+      "grad_norm": 0.8996924757957458,
+      "learning_rate": 4.9550468321917975e-06,
+      "loss": 0.6902,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3960674157303371,
+      "grad_norm": 0.9362848401069641,
+      "learning_rate": 4.9549595108319726e-06,
+      "loss": 0.7331,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3964185393258427,
+      "grad_norm": 0.9147123098373413,
+      "learning_rate": 4.954872105514633e-06,
+      "loss": 0.731,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3967696629213483,
+      "grad_norm": 0.9537502527236938,
+      "learning_rate": 4.954784616242769e-06,
+      "loss": 0.7347,
+      "step": 1130
+    },
+    {
+      "epoch": 0.39712078651685395,
+      "grad_norm": 0.9200554490089417,
+      "learning_rate": 4.9546970430193705e-06,
+      "loss": 0.7205,
+      "step": 1131
+    },
+    {
+      "epoch": 0.39747191011235955,
+      "grad_norm": 0.929223358631134,
+      "learning_rate": 4.954609385847434e-06,
+      "loss": 0.7108,
+      "step": 1132
+    },
+    {
+      "epoch": 0.39782303370786515,
+      "grad_norm": 0.9166401624679565,
+      "learning_rate": 4.954521644729958e-06,
+      "loss": 0.732,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3981741573033708,
+      "grad_norm": 0.9084526300430298,
+      "learning_rate": 4.954433819669941e-06,
+      "loss": 0.7078,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3985252808988764,
+      "grad_norm": 0.9210569858551025,
+      "learning_rate": 4.954345910670388e-06,
+      "loss": 0.6905,
+      "step": 1135
+    },
+    {
+      "epoch": 0.398876404494382,
+      "grad_norm": 0.890738308429718,
+      "learning_rate": 4.9542579177343054e-06,
+      "loss": 0.6975,
+      "step": 1136
+    },
+    {
+      "epoch": 0.39922752808988765,
+      "grad_norm": 0.8724343776702881,
+      "learning_rate": 4.954169840864701e-06,
+      "loss": 0.6323,
+      "step": 1137
+    },
+    {
+      "epoch": 0.39957865168539325,
+      "grad_norm": 0.9185892343521118,
+      "learning_rate": 4.95408168006459e-06,
+      "loss": 0.7106,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3999297752808989,
+      "grad_norm": 0.9605926871299744,
+      "learning_rate": 4.953993435336983e-06,
+      "loss": 0.7208,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4002808988764045,
+      "grad_norm": 0.9934019446372986,
+      "learning_rate": 4.953905106684902e-06,
+      "loss": 0.7394,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4006320224719101,
+      "grad_norm": 0.8674790263175964,
+      "learning_rate": 4.953816694111367e-06,
+      "loss": 0.6588,
+      "step": 1141
+    },
+    {
+      "epoch": 0.40098314606741575,
+      "grad_norm": 0.9420093894004822,
+      "learning_rate": 4.9537281976194e-06,
+      "loss": 0.6934,
+      "step": 1142
+    },
+    {
+      "epoch": 0.40133426966292135,
+      "grad_norm": 0.8882573246955872,
+      "learning_rate": 4.953639617212028e-06,
+      "loss": 0.7244,
+      "step": 1143
+    },
+    {
+      "epoch": 0.40168539325842695,
+      "grad_norm": 1.0398857593536377,
+      "learning_rate": 4.95355095289228e-06,
+      "loss": 0.7402,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4020365168539326,
+      "grad_norm": 0.9207865595817566,
+      "learning_rate": 4.95346220466319e-06,
+      "loss": 0.6942,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4023876404494382,
+      "grad_norm": 0.9174979329109192,
+      "learning_rate": 4.9533733725277924e-06,
+      "loss": 0.7241,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4027387640449438,
+      "grad_norm": 0.9054567217826843,
+      "learning_rate": 4.953284456489124e-06,
+      "loss": 0.6666,
+      "step": 1147
+    },
+    {
+      "epoch": 0.40308988764044945,
+      "grad_norm": 0.9376466870307922,
+      "learning_rate": 4.953195456550227e-06,
+      "loss": 0.718,
+      "step": 1148
+    },
+    {
+      "epoch": 0.40344101123595505,
+      "grad_norm": 0.9617171883583069,
+      "learning_rate": 4.953106372714144e-06,
+      "loss": 0.7226,
+      "step": 1149
+    },
+    {
+      "epoch": 0.40379213483146065,
+      "grad_norm": 0.8789179921150208,
+      "learning_rate": 4.953017204983923e-06,
+      "loss": 0.6981,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4041432584269663,
+      "grad_norm": 0.8860397934913635,
+      "learning_rate": 4.952927953362612e-06,
+      "loss": 0.6894,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 0.9390692114830017,
+      "learning_rate": 4.952838617853265e-06,
+      "loss": 0.6788,
+      "step": 1152
+    },
+    {
+      "epoch": 0.40484550561797755,
+      "grad_norm": 0.9032955765724182,
+      "learning_rate": 4.952749198458935e-06,
+      "loss": 0.7417,
+      "step": 1153
+    },
+    {
+      "epoch": 0.40519662921348315,
+      "grad_norm": 0.9163577556610107,
+      "learning_rate": 4.952659695182682e-06,
+      "loss": 0.7195,
+      "step": 1154
+    },
+    {
+      "epoch": 0.40554775280898875,
+      "grad_norm": 0.8781930804252625,
+      "learning_rate": 4.952570108027566e-06,
+      "loss": 0.6691,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4058988764044944,
+      "grad_norm": 0.887157678604126,
+      "learning_rate": 4.9524804369966505e-06,
+      "loss": 0.6993,
+      "step": 1156
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 0.9891164302825928,
+      "learning_rate": 4.9523906820930035e-06,
+      "loss": 0.7357,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4066011235955056,
+      "grad_norm": 0.9419492483139038,
+      "learning_rate": 4.952300843319693e-06,
+      "loss": 0.674,
+      "step": 1158
+    },
+    {
+      "epoch": 0.40695224719101125,
+      "grad_norm": 0.9248577356338501,
+      "learning_rate": 4.952210920679793e-06,
+      "loss": 0.724,
+      "step": 1159
+    },
+    {
+      "epoch": 0.40730337078651685,
+      "grad_norm": 0.9277257919311523,
+      "learning_rate": 4.9521209141763775e-06,
+      "loss": 0.7179,
+      "step": 1160
+    },
+    {
+      "epoch": 0.40765449438202245,
+      "grad_norm": 0.9654591083526611,
+      "learning_rate": 4.952030823812525e-06,
+      "loss": 0.6796,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4080056179775281,
+      "grad_norm": 0.9730082154273987,
+      "learning_rate": 4.951940649591316e-06,
+      "loss": 0.715,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4083567415730337,
+      "grad_norm": 0.9230988621711731,
+      "learning_rate": 4.9518503915158356e-06,
+      "loss": 0.7218,
+      "step": 1163
+    },
+    {
+      "epoch": 0.40870786516853935,
+      "grad_norm": 0.9274692535400391,
+      "learning_rate": 4.951760049589169e-06,
+      "loss": 0.6773,
+      "step": 1164
+    },
+    {
+      "epoch": 0.40905898876404495,
+      "grad_norm": 0.894001841545105,
+      "learning_rate": 4.951669623814408e-06,
+      "loss": 0.7118,
+      "step": 1165
+    },
+    {
+      "epoch": 0.40941011235955055,
+      "grad_norm": 0.924102783203125,
+      "learning_rate": 4.951579114194643e-06,
+      "loss": 0.6648,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4097612359550562,
+      "grad_norm": 0.9237919449806213,
+      "learning_rate": 4.95148852073297e-06,
+      "loss": 0.697,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4101123595505618,
+      "grad_norm": 0.9063787460327148,
+      "learning_rate": 4.951397843432489e-06,
+      "loss": 0.699,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4104634831460674,
+      "grad_norm": 0.8865030407905579,
+      "learning_rate": 4.951307082296297e-06,
+      "loss": 0.6768,
+      "step": 1169
+    },
+    {
+      "epoch": 0.41081460674157305,
+      "grad_norm": 0.9309532046318054,
+      "learning_rate": 4.9512162373275015e-06,
+      "loss": 0.7168,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41116573033707865,
+      "grad_norm": 0.9091811776161194,
+      "learning_rate": 4.951125308529208e-06,
+      "loss": 0.6962,
+      "step": 1171
+    },
+    {
+      "epoch": 0.41151685393258425,
+      "grad_norm": 0.8983654975891113,
+      "learning_rate": 4.951034295904527e-06,
+      "loss": 0.7193,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4118679775280899,
+      "grad_norm": 0.8983482718467712,
+      "learning_rate": 4.95094319945657e-06,
+      "loss": 0.7234,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4122191011235955,
+      "grad_norm": 0.9178565740585327,
+      "learning_rate": 4.950852019188454e-06,
+      "loss": 0.7133,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4125702247191011,
+      "grad_norm": 0.9206750988960266,
+      "learning_rate": 4.950760755103294e-06,
+      "loss": 0.7417,
+      "step": 1175
+    },
+    {
+      "epoch": 0.41292134831460675,
+      "grad_norm": 0.9141495227813721,
+      "learning_rate": 4.950669407204215e-06,
+      "loss": 0.7238,
+      "step": 1176
+    },
+    {
+      "epoch": 0.41327247191011235,
+      "grad_norm": 0.9251378178596497,
+      "learning_rate": 4.9505779754943385e-06,
+      "loss": 0.6944,
+      "step": 1177
+    },
+    {
+      "epoch": 0.413623595505618,
+      "grad_norm": 0.9240232706069946,
+      "learning_rate": 4.950486459976793e-06,
+      "loss": 0.7241,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4139747191011236,
+      "grad_norm": 0.9050787687301636,
+      "learning_rate": 4.950394860654707e-06,
+      "loss": 0.7044,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4143258426966292,
+      "grad_norm": 0.9517340064048767,
+      "learning_rate": 4.950303177531213e-06,
+      "loss": 0.7363,
+      "step": 1180
+    },
+    {
+      "epoch": 0.41467696629213485,
+      "grad_norm": 0.9461115598678589,
+      "learning_rate": 4.950211410609448e-06,
+      "loss": 0.6786,
+      "step": 1181
+    },
+    {
+      "epoch": 0.41502808988764045,
+      "grad_norm": 0.9148724675178528,
+      "learning_rate": 4.950119559892549e-06,
+      "loss": 0.6873,
+      "step": 1182
+    },
+    {
+      "epoch": 0.41537921348314605,
+      "grad_norm": 0.9085513353347778,
+      "learning_rate": 4.950027625383657e-06,
+      "loss": 0.7209,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4157303370786517,
+      "grad_norm": 0.9327560067176819,
+      "learning_rate": 4.949935607085919e-06,
+      "loss": 0.7086,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4160814606741573,
+      "grad_norm": 0.936128556728363,
+      "learning_rate": 4.949843505002478e-06,
+      "loss": 0.6894,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4164325842696629,
+      "grad_norm": 0.931978702545166,
+      "learning_rate": 4.9497513191364855e-06,
+      "loss": 0.7209,
+      "step": 1186
+    },
+    {
+      "epoch": 0.41678370786516855,
+      "grad_norm": 0.8966789245605469,
+      "learning_rate": 4.949659049491095e-06,
+      "loss": 0.7211,
+      "step": 1187
+    },
+    {
+      "epoch": 0.41713483146067415,
+      "grad_norm": 0.9003615975379944,
+      "learning_rate": 4.9495666960694596e-06,
+      "loss": 0.6835,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4174859550561798,
+      "grad_norm": 0.933268666267395,
+      "learning_rate": 4.9494742588747405e-06,
+      "loss": 0.7335,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4178370786516854,
+      "grad_norm": 0.8876200318336487,
+      "learning_rate": 4.949381737910098e-06,
+      "loss": 0.7132,
+      "step": 1190
+    },
+    {
+      "epoch": 0.418188202247191,
+      "grad_norm": 0.9531140923500061,
+      "learning_rate": 4.949289133178695e-06,
+      "loss": 0.746,
+      "step": 1191
+    },
+    {
+      "epoch": 0.41853932584269665,
+      "grad_norm": 0.9262682795524597,
+      "learning_rate": 4.949196444683701e-06,
+      "loss": 0.7155,
+      "step": 1192
+    },
+    {
+      "epoch": 0.41889044943820225,
+      "grad_norm": 0.9318521022796631,
+      "learning_rate": 4.949103672428282e-06,
+      "loss": 0.7234,
+      "step": 1193
+    },
+    {
+      "epoch": 0.41924157303370785,
+      "grad_norm": 0.922127366065979,
+      "learning_rate": 4.949010816415616e-06,
+      "loss": 0.7116,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4195926966292135,
+      "grad_norm": 0.9445079565048218,
+      "learning_rate": 4.948917876648874e-06,
+      "loss": 0.6993,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4199438202247191,
+      "grad_norm": 0.9283545613288879,
+      "learning_rate": 4.948824853131237e-06,
+      "loss": 0.7078,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4202949438202247,
+      "grad_norm": 0.8923372626304626,
+      "learning_rate": 4.9487317458658845e-06,
+      "loss": 0.6818,
+      "step": 1197
+    },
+    {
+      "epoch": 0.42064606741573035,
+      "grad_norm": 0.907924234867096,
+      "learning_rate": 4.948638554856002e-06,
+      "loss": 0.6799,
+      "step": 1198
+    },
+    {
+      "epoch": 0.42099719101123595,
+      "grad_norm": 0.8753466606140137,
+      "learning_rate": 4.948545280104777e-06,
+      "loss": 0.6789,
+      "step": 1199
+    },
+    {
+      "epoch": 0.42134831460674155,
+      "grad_norm": 0.965340256690979,
+      "learning_rate": 4.948451921615398e-06,
+      "loss": 0.6911,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4216994382022472,
+      "grad_norm": 0.9519873261451721,
+      "learning_rate": 4.948358479391059e-06,
+      "loss": 0.6899,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4220505617977528,
+      "grad_norm": 0.907811164855957,
+      "learning_rate": 4.948264953434955e-06,
+      "loss": 0.7401,
+      "step": 1202
+    },
+    {
+      "epoch": 0.42240168539325845,
+      "grad_norm": 0.9398930072784424,
+      "learning_rate": 4.948171343750284e-06,
+      "loss": 0.7036,
+      "step": 1203
+    },
+    {
+      "epoch": 0.42275280898876405,
+      "grad_norm": 0.9245709776878357,
+      "learning_rate": 4.948077650340248e-06,
+      "loss": 0.7042,
+      "step": 1204
+    },
+    {
+      "epoch": 0.42310393258426965,
+      "grad_norm": 0.8742227554321289,
+      "learning_rate": 4.947983873208052e-06,
+      "loss": 0.6941,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4234550561797753,
+      "grad_norm": 0.9183447957038879,
+      "learning_rate": 4.947890012356902e-06,
+      "loss": 0.7432,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4238061797752809,
+      "grad_norm": 0.9594230055809021,
+      "learning_rate": 4.947796067790009e-06,
+      "loss": 0.7486,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4241573033707865,
+      "grad_norm": 0.9002628326416016,
+      "learning_rate": 4.947702039510583e-06,
+      "loss": 0.6652,
+      "step": 1208
+    },
+    {
+      "epoch": 0.42450842696629215,
+      "grad_norm": 0.9353854060173035,
+      "learning_rate": 4.947607927521844e-06,
+      "loss": 0.7501,
+      "step": 1209
+    },
+    {
+      "epoch": 0.42485955056179775,
+      "grad_norm": 0.8836169242858887,
+      "learning_rate": 4.947513731827007e-06,
+      "loss": 0.6904,
+      "step": 1210
+    },
+    {
+      "epoch": 0.42521067415730335,
+      "grad_norm": 0.9366084337234497,
+      "learning_rate": 4.947419452429295e-06,
+      "loss": 0.7013,
+      "step": 1211
+    },
+    {
+      "epoch": 0.425561797752809,
+      "grad_norm": 0.9252647161483765,
+      "learning_rate": 4.947325089331932e-06,
+      "loss": 0.7343,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4259129213483146,
+      "grad_norm": 0.9203290343284607,
+      "learning_rate": 4.947230642538145e-06,
+      "loss": 0.7129,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4262640449438202,
+      "grad_norm": 0.8908094763755798,
+      "learning_rate": 4.947136112051164e-06,
+      "loss": 0.6785,
+      "step": 1214
+    },
+    {
+      "epoch": 0.42661516853932585,
+      "grad_norm": 0.9249038696289062,
+      "learning_rate": 4.947041497874223e-06,
+      "loss": 0.6664,
+      "step": 1215
+    },
+    {
+      "epoch": 0.42696629213483145,
+      "grad_norm": 0.9546428918838501,
+      "learning_rate": 4.946946800010556e-06,
+      "loss": 0.7276,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4273174157303371,
+      "grad_norm": 0.9032455682754517,
+      "learning_rate": 4.946852018463403e-06,
+      "loss": 0.6757,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4276685393258427,
+      "grad_norm": 0.9056017994880676,
+      "learning_rate": 4.946757153236003e-06,
+      "loss": 0.6878,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4280196629213483,
+      "grad_norm": 0.9265672564506531,
+      "learning_rate": 4.946662204331603e-06,
+      "loss": 0.7043,
+      "step": 1219
+    },
+    {
+      "epoch": 0.42837078651685395,
+      "grad_norm": 0.8721897006034851,
+      "learning_rate": 4.94656717175345e-06,
+      "loss": 0.627,
+      "step": 1220
+    },
+    {
+      "epoch": 0.42872191011235955,
+      "grad_norm": 0.9220138192176819,
+      "learning_rate": 4.9464720555047935e-06,
+      "loss": 0.7233,
+      "step": 1221
+    },
+    {
+      "epoch": 0.42907303370786515,
+      "grad_norm": 0.9333030581474304,
+      "learning_rate": 4.9463768555888855e-06,
+      "loss": 0.7065,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4294241573033708,
+      "grad_norm": 0.9611704349517822,
+      "learning_rate": 4.946281572008982e-06,
+      "loss": 0.7044,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4297752808988764,
+      "grad_norm": 0.9443504810333252,
+      "learning_rate": 4.946186204768343e-06,
+      "loss": 0.7322,
+      "step": 1224
+    },
+    {
+      "epoch": 0.430126404494382,
+      "grad_norm": 0.9180960059165955,
+      "learning_rate": 4.946090753870229e-06,
+      "loss": 0.7435,
+      "step": 1225
+    },
+    {
+      "epoch": 0.43047752808988765,
+      "grad_norm": 0.9297470450401306,
+      "learning_rate": 4.945995219317903e-06,
+      "loss": 0.7211,
+      "step": 1226
+    },
+    {
+      "epoch": 0.43082865168539325,
+      "grad_norm": 0.9243416786193848,
+      "learning_rate": 4.945899601114634e-06,
+      "loss": 0.7255,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4311797752808989,
+      "grad_norm": 0.9293515086174011,
+      "learning_rate": 4.945803899263692e-06,
+      "loss": 0.7466,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4315308988764045,
+      "grad_norm": 0.8804196715354919,
+      "learning_rate": 4.94570811376835e-06,
+      "loss": 0.6495,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4318820224719101,
+      "grad_norm": 0.8709548115730286,
+      "learning_rate": 4.945612244631884e-06,
+      "loss": 0.6573,
+      "step": 1230
+    },
+    {
+      "epoch": 0.43223314606741575,
+      "grad_norm": 0.9267548322677612,
+      "learning_rate": 4.94551629185757e-06,
+      "loss": 0.6949,
+      "step": 1231
+    },
+    {
+      "epoch": 0.43258426966292135,
+      "grad_norm": 0.9049427509307861,
+      "learning_rate": 4.945420255448693e-06,
+      "loss": 0.6724,
+      "step": 1232
+    },
+    {
+      "epoch": 0.43293539325842695,
+      "grad_norm": 0.9540070295333862,
+      "learning_rate": 4.945324135408535e-06,
+      "loss": 0.7162,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4332865168539326,
+      "grad_norm": 0.9069761633872986,
+      "learning_rate": 4.945227931740384e-06,
+      "loss": 0.6851,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4336376404494382,
+      "grad_norm": 0.9255860447883606,
+      "learning_rate": 4.94513164444753e-06,
+      "loss": 0.7108,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4339887640449438,
+      "grad_norm": 0.9611523151397705,
+      "learning_rate": 4.945035273533267e-06,
+      "loss": 0.7156,
+      "step": 1236
+    },
+    {
+      "epoch": 0.43433988764044945,
+      "grad_norm": 0.9171714782714844,
+      "learning_rate": 4.9449388190008896e-06,
+      "loss": 0.6744,
+      "step": 1237
+    },
+    {
+      "epoch": 0.43469101123595505,
+      "grad_norm": 0.8625194430351257,
+      "learning_rate": 4.944842280853696e-06,
+      "loss": 0.6792,
+      "step": 1238
+    },
+    {
+      "epoch": 0.43504213483146065,
+      "grad_norm": 0.945565938949585,
+      "learning_rate": 4.94474565909499e-06,
+      "loss": 0.7015,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4353932584269663,
+      "grad_norm": 0.8865808248519897,
+      "learning_rate": 4.944648953728073e-06,
+      "loss": 0.6664,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4357443820224719,
+      "grad_norm": 0.8900501728057861,
+      "learning_rate": 4.9445521647562545e-06,
+      "loss": 0.6831,
+      "step": 1241
+    },
+    {
+      "epoch": 0.43609550561797755,
+      "grad_norm": 0.9169394373893738,
+      "learning_rate": 4.944455292182844e-06,
+      "loss": 0.6377,
+      "step": 1242
+    },
+    {
+      "epoch": 0.43644662921348315,
+      "grad_norm": 0.9147849082946777,
+      "learning_rate": 4.944358336011154e-06,
+      "loss": 0.735,
+      "step": 1243
+    },
+    {
+      "epoch": 0.43679775280898875,
+      "grad_norm": 0.9505800604820251,
+      "learning_rate": 4.9442612962445e-06,
+      "loss": 0.7309,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4371488764044944,
+      "grad_norm": 0.8858315348625183,
+      "learning_rate": 4.9441641728862015e-06,
+      "loss": 0.6904,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.9335656762123108,
+      "learning_rate": 4.94406696593958e-06,
+      "loss": 0.7219,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4378511235955056,
+      "grad_norm": 0.8928138613700867,
+      "learning_rate": 4.94396967540796e-06,
+      "loss": 0.6521,
+      "step": 1247
+    },
+    {
+      "epoch": 0.43820224719101125,
+      "grad_norm": 0.9080971479415894,
+      "learning_rate": 4.943872301294668e-06,
+      "loss": 0.6925,
+      "step": 1248
+    },
+    {
+      "epoch": 0.43855337078651685,
+      "grad_norm": 0.9383649826049805,
+      "learning_rate": 4.943774843603034e-06,
+      "loss": 0.708,
+      "step": 1249
+    },
+    {
+      "epoch": 0.43890449438202245,
+      "grad_norm": 0.9101510047912598,
+      "learning_rate": 4.943677302336393e-06,
+      "loss": 0.6901,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4392556179775281,
+      "grad_norm": 0.9402358531951904,
+      "learning_rate": 4.943579677498078e-06,
+      "loss": 0.7169,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4396067415730337,
+      "grad_norm": 0.8998861908912659,
+      "learning_rate": 4.943481969091429e-06,
+      "loss": 0.6962,
+      "step": 1252
+    },
+    {
+      "epoch": 0.43995786516853935,
+      "grad_norm": 0.9605402946472168,
+      "learning_rate": 4.943384177119788e-06,
+      "loss": 0.6783,
+      "step": 1253
+    },
+    {
+      "epoch": 0.44030898876404495,
+      "grad_norm": 0.9242511987686157,
+      "learning_rate": 4.9432863015865e-06,
+      "loss": 0.71,
+      "step": 1254
+    },
+    {
+      "epoch": 0.44066011235955055,
+      "grad_norm": 0.9048755168914795,
+      "learning_rate": 4.943188342494909e-06,
+      "loss": 0.7143,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4410112359550562,
+      "grad_norm": 0.883234977722168,
+      "learning_rate": 4.9430902998483685e-06,
+      "loss": 0.6922,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4413623595505618,
+      "grad_norm": 0.9546087980270386,
+      "learning_rate": 4.942992173650231e-06,
+      "loss": 0.725,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4417134831460674,
+      "grad_norm": 0.9286230206489563,
+      "learning_rate": 4.942893963903851e-06,
+      "loss": 0.7263,
+      "step": 1258
+    },
+    {
+      "epoch": 0.44206460674157305,
+      "grad_norm": 0.901672899723053,
+      "learning_rate": 4.942795670612588e-06,
+      "loss": 0.6826,
+      "step": 1259
+    },
+    {
+      "epoch": 0.44241573033707865,
+      "grad_norm": 0.9127205610275269,
+      "learning_rate": 4.9426972937798035e-06,
+      "loss": 0.6727,
+      "step": 1260
+    },
+    {
+      "epoch": 0.44276685393258425,
+      "grad_norm": 0.9450622797012329,
+      "learning_rate": 4.942598833408861e-06,
+      "loss": 0.726,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4431179775280899,
+      "grad_norm": 0.9145720601081848,
+      "learning_rate": 4.94250028950313e-06,
+      "loss": 0.702,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4434691011235955,
+      "grad_norm": 0.9019625782966614,
+      "learning_rate": 4.9424016620659786e-06,
+      "loss": 0.6682,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4438202247191011,
+      "grad_norm": 0.9257623553276062,
+      "learning_rate": 4.9423029511007795e-06,
+      "loss": 0.6981,
+      "step": 1264
+    },
+    {
+      "epoch": 0.44417134831460675,
+      "grad_norm": 0.9080988764762878,
+      "learning_rate": 4.94220415661091e-06,
+      "loss": 0.708,
+      "step": 1265
+    },
+    {
+      "epoch": 0.44452247191011235,
+      "grad_norm": 0.9560781121253967,
+      "learning_rate": 4.942105278599748e-06,
+      "loss": 0.7021,
+      "step": 1266
+    },
+    {
+      "epoch": 0.444873595505618,
+      "grad_norm": 0.9422893524169922,
+      "learning_rate": 4.942006317070676e-06,
+      "loss": 0.6861,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4452247191011236,
+      "grad_norm": 0.8909929990768433,
+      "learning_rate": 4.9419072720270764e-06,
+      "loss": 0.7143,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4455758426966292,
+      "grad_norm": 0.9080603718757629,
+      "learning_rate": 4.941808143472339e-06,
+      "loss": 0.7073,
+      "step": 1269
+    },
+    {
+      "epoch": 0.44592696629213485,
+      "grad_norm": 0.8691872358322144,
+      "learning_rate": 4.941708931409851e-06,
+      "loss": 0.6425,
+      "step": 1270
+    },
+    {
+      "epoch": 0.44627808988764045,
+      "grad_norm": 0.9223096966743469,
+      "learning_rate": 4.9416096358430084e-06,
+      "loss": 0.6876,
+      "step": 1271
+    },
+    {
+      "epoch": 0.44662921348314605,
+      "grad_norm": 0.9030603766441345,
+      "learning_rate": 4.9415102567752045e-06,
+      "loss": 0.6893,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4469803370786517,
+      "grad_norm": 0.9421838521957397,
+      "learning_rate": 4.94141079420984e-06,
+      "loss": 0.7267,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4473314606741573,
+      "grad_norm": 0.8724151253700256,
+      "learning_rate": 4.941311248150314e-06,
+      "loss": 0.6587,
+      "step": 1274
+    },
+    {
+      "epoch": 0.4476825842696629,
+      "grad_norm": 0.8822656869888306,
+      "learning_rate": 4.9412116186000346e-06,
+      "loss": 0.6969,
+      "step": 1275
+    },
+    {
+      "epoch": 0.44803370786516855,
+      "grad_norm": 0.8846755623817444,
+      "learning_rate": 4.941111905562405e-06,
+      "loss": 0.6848,
+      "step": 1276
+    },
+    {
+      "epoch": 0.44838483146067415,
+      "grad_norm": 0.8999640941619873,
+      "learning_rate": 4.941012109040838e-06,
+      "loss": 0.6465,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4487359550561798,
+      "grad_norm": 0.9194796681404114,
+      "learning_rate": 4.940912229038746e-06,
+      "loss": 0.6995,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4490870786516854,
+      "grad_norm": 0.9296586513519287,
+      "learning_rate": 4.940812265559543e-06,
+      "loss": 0.6399,
+      "step": 1279
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.9012690782546997,
+      "learning_rate": 4.94071221860665e-06,
+      "loss": 0.7109,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44978932584269665,
+      "grad_norm": 0.9507624506950378,
+      "learning_rate": 4.940612088183488e-06,
+      "loss": 0.7498,
+      "step": 1281
+    },
+    {
+      "epoch": 0.45014044943820225,
+      "grad_norm": 0.8964364528656006,
+      "learning_rate": 4.940511874293481e-06,
+      "loss": 0.674,
+      "step": 1282
+    },
+    {
+      "epoch": 0.45049157303370785,
+      "grad_norm": 0.9396047592163086,
+      "learning_rate": 4.940411576940056e-06,
+      "loss": 0.699,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4508426966292135,
+      "grad_norm": 0.9038118720054626,
+      "learning_rate": 4.940311196126643e-06,
+      "loss": 0.6598,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4511938202247191,
+      "grad_norm": 0.9001778960227966,
+      "learning_rate": 4.940210731856675e-06,
+      "loss": 0.7129,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4515449438202247,
+      "grad_norm": 0.9313659071922302,
+      "learning_rate": 4.94011018413359e-06,
+      "loss": 0.7138,
+      "step": 1286
+    },
+    {
+      "epoch": 0.45189606741573035,
+      "grad_norm": 0.9225174784660339,
+      "learning_rate": 4.9400095529608226e-06,
+      "loss": 0.7192,
+      "step": 1287
+    },
+    {
+      "epoch": 0.45224719101123595,
+      "grad_norm": 0.9100403189659119,
+      "learning_rate": 4.939908838341818e-06,
+      "loss": 0.6593,
+      "step": 1288
+    },
+    {
+      "epoch": 0.45259831460674155,
+      "grad_norm": 0.8900116086006165,
+      "learning_rate": 4.939808040280017e-06,
+      "loss": 0.6799,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4529494382022472,
+      "grad_norm": 0.885908305644989,
+      "learning_rate": 4.939707158778872e-06,
+      "loss": 0.6889,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4533005617977528,
+      "grad_norm": 0.9219464659690857,
+      "learning_rate": 4.939606193841827e-06,
+      "loss": 0.7198,
+      "step": 1291
+    },
+    {
+      "epoch": 0.45365168539325845,
+      "grad_norm": 0.9270339012145996,
+      "learning_rate": 4.939505145472339e-06,
+      "loss": 0.7121,
+      "step": 1292
+    },
+    {
+      "epoch": 0.45400280898876405,
+      "grad_norm": 0.9297556281089783,
+      "learning_rate": 4.939404013673862e-06,
+      "loss": 0.7119,
+      "step": 1293
+    },
+    {
+      "epoch": 0.45435393258426965,
+      "grad_norm": 0.9266655445098877,
+      "learning_rate": 4.9393027984498555e-06,
+      "loss": 0.6994,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4547050561797753,
+      "grad_norm": 0.9152787923812866,
+      "learning_rate": 4.939201499803781e-06,
+      "loss": 0.6618,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4550561797752809,
+      "grad_norm": 0.9240756034851074,
+      "learning_rate": 4.939100117739102e-06,
+      "loss": 0.7359,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4554073033707865,
+      "grad_norm": 0.9413486123085022,
+      "learning_rate": 4.9389986522592855e-06,
+      "loss": 0.7266,
+      "step": 1297
+    },
+    {
+      "epoch": 0.45575842696629215,
+      "grad_norm": 0.8976860046386719,
+      "learning_rate": 4.938897103367803e-06,
+      "loss": 0.7063,
+      "step": 1298
+    },
+    {
+      "epoch": 0.45610955056179775,
+      "grad_norm": 0.9674167633056641,
+      "learning_rate": 4.938795471068126e-06,
+      "loss": 0.7191,
+      "step": 1299
+    },
+    {
+      "epoch": 0.45646067415730335,
+      "grad_norm": 0.9469913840293884,
+      "learning_rate": 4.938693755363731e-06,
+      "loss": 0.7263,
+      "step": 1300
+    },
+    {
+      "epoch": 0.456811797752809,
+      "grad_norm": 0.9106466174125671,
+      "learning_rate": 4.938591956258097e-06,
+      "loss": 0.7324,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4571629213483146,
+      "grad_norm": 0.9906820058822632,
+      "learning_rate": 4.938490073754705e-06,
+      "loss": 0.7273,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4575140449438202,
+      "grad_norm": 0.9152560234069824,
+      "learning_rate": 4.938388107857038e-06,
+      "loss": 0.6648,
+      "step": 1303
+    },
+    {
+      "epoch": 0.45786516853932585,
+      "grad_norm": 0.9459682703018188,
+      "learning_rate": 4.938286058568585e-06,
+      "loss": 0.6869,
+      "step": 1304
+    },
+    {
+      "epoch": 0.45821629213483145,
+      "grad_norm": 0.9387467503547668,
+      "learning_rate": 4.938183925892835e-06,
+      "loss": 0.7028,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4585674157303371,
+      "grad_norm": 0.9118907451629639,
+      "learning_rate": 4.938081709833281e-06,
+      "loss": 0.6875,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4589185393258427,
+      "grad_norm": 0.9032610654830933,
+      "learning_rate": 4.937979410393419e-06,
+      "loss": 0.7157,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4592696629213483,
+      "grad_norm": 0.9495104551315308,
+      "learning_rate": 4.937877027576748e-06,
+      "loss": 0.6914,
+      "step": 1308
+    },
+    {
+      "epoch": 0.45962078651685395,
+      "grad_norm": 0.8904275894165039,
+      "learning_rate": 4.937774561386768e-06,
+      "loss": 0.7351,
+      "step": 1309
+    },
+    {
+      "epoch": 0.45997191011235955,
+      "grad_norm": 0.8993387222290039,
+      "learning_rate": 4.937672011826984e-06,
+      "loss": 0.7004,
+      "step": 1310
+    },
+    {
+      "epoch": 0.46032303370786515,
+      "grad_norm": 0.9243264198303223,
+      "learning_rate": 4.937569378900903e-06,
+      "loss": 0.7059,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4606741573033708,
+      "grad_norm": 0.8777183294296265,
+      "learning_rate": 4.937466662612036e-06,
+      "loss": 0.6795,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4610252808988764,
+      "grad_norm": 0.9348961114883423,
+      "learning_rate": 4.937363862963895e-06,
+      "loss": 0.6726,
+      "step": 1313
+    },
+    {
+      "epoch": 0.461376404494382,
+      "grad_norm": 0.9088345766067505,
+      "learning_rate": 4.937260979959994e-06,
+      "loss": 0.7114,
+      "step": 1314
+    },
+    {
+      "epoch": 0.46172752808988765,
+      "grad_norm": 0.9143536686897278,
+      "learning_rate": 4.937158013603854e-06,
+      "loss": 0.6819,
+      "step": 1315
+    },
+    {
+      "epoch": 0.46207865168539325,
+      "grad_norm": 0.9480457901954651,
+      "learning_rate": 4.937054963898996e-06,
+      "loss": 0.7335,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4624297752808989,
+      "grad_norm": 0.9207438230514526,
+      "learning_rate": 4.9369518308489425e-06,
+      "loss": 0.7091,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4627808988764045,
+      "grad_norm": 0.9547027349472046,
+      "learning_rate": 4.936848614457223e-06,
+      "loss": 0.7388,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4631320224719101,
+      "grad_norm": 0.9442094564437866,
+      "learning_rate": 4.9367453147273645e-06,
+      "loss": 0.709,
+      "step": 1319
+    },
+    {
+      "epoch": 0.46348314606741575,
+      "grad_norm": 0.9056798815727234,
+      "learning_rate": 4.9366419316629024e-06,
+      "loss": 0.6919,
+      "step": 1320
+    },
+    {
+      "epoch": 0.46383426966292135,
+      "grad_norm": 0.8800448179244995,
+      "learning_rate": 4.936538465267371e-06,
+      "loss": 0.7141,
+      "step": 1321
+    },
+    {
+      "epoch": 0.46418539325842695,
+      "grad_norm": 0.9331648349761963,
+      "learning_rate": 4.93643491554431e-06,
+      "loss": 0.6871,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4645365168539326,
+      "grad_norm": 0.9699183702468872,
+      "learning_rate": 4.9363312824972586e-06,
+      "loss": 0.729,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4648876404494382,
+      "grad_norm": 0.9243112206459045,
+      "learning_rate": 4.9362275661297634e-06,
+      "loss": 0.6751,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4652387640449438,
+      "grad_norm": 0.9310767650604248,
+      "learning_rate": 4.936123766445368e-06,
+      "loss": 0.7263,
+      "step": 1325
+    },
+    {
+      "epoch": 0.46558988764044945,
+      "grad_norm": 0.9163933992385864,
+      "learning_rate": 4.936019883447627e-06,
+      "loss": 0.715,
+      "step": 1326
+    },
+    {
+      "epoch": 0.46594101123595505,
+      "grad_norm": 0.9764875173568726,
+      "learning_rate": 4.935915917140089e-06,
+      "loss": 0.7139,
+      "step": 1327
+    },
+    {
+      "epoch": 0.46629213483146065,
+      "grad_norm": 0.9456989169120789,
+      "learning_rate": 4.935811867526311e-06,
+      "loss": 0.6667,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4666432584269663,
+      "grad_norm": 0.9617751836776733,
+      "learning_rate": 4.935707734609852e-06,
+      "loss": 0.691,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4669943820224719,
+      "grad_norm": 0.8945748209953308,
+      "learning_rate": 4.935603518394273e-06,
+      "loss": 0.711,
+      "step": 1330
+    },
+    {
+      "epoch": 0.46734550561797755,
+      "grad_norm": 0.9066562652587891,
+      "learning_rate": 4.935499218883137e-06,
+      "loss": 0.685,
+      "step": 1331
+    },
+    {
+      "epoch": 0.46769662921348315,
+      "grad_norm": 0.8701803684234619,
+      "learning_rate": 4.935394836080012e-06,
+      "loss": 0.6735,
+      "step": 1332
+    },
+    {
+      "epoch": 0.46804775280898875,
+      "grad_norm": 0.9355669021606445,
+      "learning_rate": 4.935290369988468e-06,
+      "loss": 0.7335,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4683988764044944,
+      "grad_norm": 0.882905900478363,
+      "learning_rate": 4.9351858206120774e-06,
+      "loss": 0.7196,
+      "step": 1334
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 0.9127156734466553,
+      "learning_rate": 4.935081187954415e-06,
+      "loss": 0.714,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4691011235955056,
+      "grad_norm": 1.284045934677124,
+      "learning_rate": 4.9349764720190595e-06,
+      "loss": 0.7167,
+      "step": 1336
+    },
+    {
+      "epoch": 0.46945224719101125,
+      "grad_norm": 0.919476330280304,
+      "learning_rate": 4.934871672809594e-06,
+      "loss": 0.6804,
+      "step": 1337
+    },
+    {
+      "epoch": 0.46980337078651685,
+      "grad_norm": 0.9066704511642456,
+      "learning_rate": 4.934766790329599e-06,
+      "loss": 0.706,
+      "step": 1338
+    },
+    {
+      "epoch": 0.47015449438202245,
+      "grad_norm": 0.9020872116088867,
+      "learning_rate": 4.934661824582664e-06,
+      "loss": 0.7277,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4705056179775281,
+      "grad_norm": 0.9096556901931763,
+      "learning_rate": 4.934556775572377e-06,
+      "loss": 0.7184,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4708567415730337,
+      "grad_norm": 0.8852599859237671,
+      "learning_rate": 4.934451643302332e-06,
+      "loss": 0.6952,
+      "step": 1341
+    },
+    {
+      "epoch": 0.47120786516853935,
+      "grad_norm": 0.9317033290863037,
+      "learning_rate": 4.934346427776125e-06,
+      "loss": 0.7087,
+      "step": 1342
+    },
+    {
+      "epoch": 0.47155898876404495,
+      "grad_norm": 0.9349064230918884,
+      "learning_rate": 4.934241128997353e-06,
+      "loss": 0.7295,
+      "step": 1343
+    },
+    {
+      "epoch": 0.47191011235955055,
+      "grad_norm": 0.8978156447410583,
+      "learning_rate": 4.934135746969617e-06,
+      "loss": 0.7129,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4722612359550562,
+      "grad_norm": 0.8900869488716125,
+      "learning_rate": 4.934030281696521e-06,
+      "loss": 0.6737,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4726123595505618,
+      "grad_norm": 0.9193541407585144,
+      "learning_rate": 4.933924733181673e-06,
+      "loss": 0.7189,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4729634831460674,
+      "grad_norm": 0.8839302659034729,
+      "learning_rate": 4.933819101428681e-06,
+      "loss": 0.6883,
+      "step": 1347
+    },
+    {
+      "epoch": 0.47331460674157305,
+      "grad_norm": 0.8768748641014099,
+      "learning_rate": 4.933713386441159e-06,
+      "loss": 0.6825,
+      "step": 1348
+    },
+    {
+      "epoch": 0.47366573033707865,
+      "grad_norm": 0.9012611508369446,
+      "learning_rate": 4.933607588222721e-06,
+      "loss": 0.7477,
+      "step": 1349
+    },
+    {
+      "epoch": 0.47401685393258425,
+      "grad_norm": 0.9348760843276978,
+      "learning_rate": 4.933501706776986e-06,
+      "loss": 0.7071,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4743679775280899,
+      "grad_norm": 0.9126866459846497,
+      "learning_rate": 4.933395742107576e-06,
+      "loss": 0.7259,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4747191011235955,
+      "grad_norm": 1.1260316371917725,
+      "learning_rate": 4.9332896942181135e-06,
+      "loss": 0.6647,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4750702247191011,
+      "grad_norm": 0.9544892907142639,
+      "learning_rate": 4.933183563112226e-06,
+      "loss": 0.7228,
+      "step": 1353
+    },
+    {
+      "epoch": 0.47542134831460675,
+      "grad_norm": 0.9179894924163818,
+      "learning_rate": 4.933077348793542e-06,
+      "loss": 0.7115,
+      "step": 1354
+    },
+    {
+      "epoch": 0.47577247191011235,
+      "grad_norm": 0.8785374164581299,
+      "learning_rate": 4.932971051265695e-06,
+      "loss": 0.6651,
+      "step": 1355
+    },
+    {
+      "epoch": 0.476123595505618,
+      "grad_norm": 0.9006591439247131,
+      "learning_rate": 4.93286467053232e-06,
+      "loss": 0.7166,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4764747191011236,
+      "grad_norm": 0.9105557799339294,
+      "learning_rate": 4.932758206597055e-06,
+      "loss": 0.6876,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4768258426966292,
+      "grad_norm": 0.8968350887298584,
+      "learning_rate": 4.932651659463541e-06,
+      "loss": 0.6873,
+      "step": 1358
+    },
+    {
+      "epoch": 0.47717696629213485,
+      "grad_norm": 0.9052225947380066,
+      "learning_rate": 4.932545029135422e-06,
+      "loss": 0.7134,
+      "step": 1359
+    },
+    {
+      "epoch": 0.47752808988764045,
+      "grad_norm": 0.946305513381958,
+      "learning_rate": 4.932438315616345e-06,
+      "loss": 0.7061,
+      "step": 1360
+    },
+    {
+      "epoch": 0.47787921348314605,
+      "grad_norm": 0.9128836393356323,
+      "learning_rate": 4.9323315189099586e-06,
+      "loss": 0.7101,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4782303370786517,
+      "grad_norm": 0.9199392199516296,
+      "learning_rate": 4.9322246390199155e-06,
+      "loss": 0.7073,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4785814606741573,
+      "grad_norm": 1.077110767364502,
+      "learning_rate": 4.932117675949871e-06,
+      "loss": 0.6828,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4789325842696629,
+      "grad_norm": 0.9046890735626221,
+      "learning_rate": 4.932010629703484e-06,
+      "loss": 0.6991,
+      "step": 1364
+    },
+    {
+      "epoch": 0.47928370786516855,
+      "grad_norm": 0.8711754083633423,
+      "learning_rate": 4.931903500284414e-06,
+      "loss": 0.6788,
+      "step": 1365
+    },
+    {
+      "epoch": 0.47963483146067415,
+      "grad_norm": 0.8938144445419312,
+      "learning_rate": 4.931796287696325e-06,
+      "loss": 0.7219,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4799859550561798,
+      "grad_norm": 0.9003678560256958,
+      "learning_rate": 4.931688991942884e-06,
+      "loss": 0.6908,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4803370786516854,
+      "grad_norm": 0.9107855558395386,
+      "learning_rate": 4.93158161302776e-06,
+      "loss": 0.6821,
+      "step": 1368
+    },
+    {
+      "epoch": 0.480688202247191,
+      "grad_norm": 0.9014522433280945,
+      "learning_rate": 4.931474150954626e-06,
+      "loss": 0.7071,
+      "step": 1369
+    },
+    {
+      "epoch": 0.48103932584269665,
+      "grad_norm": 0.9273253083229065,
+      "learning_rate": 4.931366605727157e-06,
+      "loss": 0.7027,
+      "step": 1370
+    },
+    {
+      "epoch": 0.48139044943820225,
+      "grad_norm": 0.9555942416191101,
+      "learning_rate": 4.93125897734903e-06,
+      "loss": 0.6952,
+      "step": 1371
+    },
+    {
+      "epoch": 0.48174157303370785,
+      "grad_norm": 0.9457711577415466,
+      "learning_rate": 4.931151265823927e-06,
+      "loss": 0.6893,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4820926966292135,
+      "grad_norm": 0.9268205761909485,
+      "learning_rate": 4.931043471155531e-06,
+      "loss": 0.7169,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4824438202247191,
+      "grad_norm": 0.9136787056922913,
+      "learning_rate": 4.930935593347529e-06,
+      "loss": 0.7326,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4827949438202247,
+      "grad_norm": 0.925116777420044,
+      "learning_rate": 4.930827632403608e-06,
+      "loss": 0.7037,
+      "step": 1375
+    },
+    {
+      "epoch": 0.48314606741573035,
+      "grad_norm": 0.9161933064460754,
+      "learning_rate": 4.930719588327464e-06,
+      "loss": 0.7222,
+      "step": 1376
+    },
+    {
+      "epoch": 0.48349719101123595,
+      "grad_norm": 0.9238081574440002,
+      "learning_rate": 4.93061146112279e-06,
+      "loss": 0.7186,
+      "step": 1377
+    },
+    {
+      "epoch": 0.48384831460674155,
+      "grad_norm": 0.9430643320083618,
+      "learning_rate": 4.930503250793283e-06,
+      "loss": 0.7622,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4841994382022472,
+      "grad_norm": 0.9295469522476196,
+      "learning_rate": 4.930394957342645e-06,
+      "loss": 0.7354,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4845505617977528,
+      "grad_norm": 0.9064199924468994,
+      "learning_rate": 4.9302865807745795e-06,
+      "loss": 0.6487,
+      "step": 1380
+    },
+    {
+      "epoch": 0.48490168539325845,
+      "grad_norm": 0.9455371499061584,
+      "learning_rate": 4.930178121092792e-06,
+      "loss": 0.7468,
+      "step": 1381
+    },
+    {
+      "epoch": 0.48525280898876405,
+      "grad_norm": 0.9263501763343811,
+      "learning_rate": 4.930069578300992e-06,
+      "loss": 0.7099,
+      "step": 1382
+    },
+    {
+      "epoch": 0.48560393258426965,
+      "grad_norm": 0.8904842734336853,
+      "learning_rate": 4.929960952402892e-06,
+      "loss": 0.6519,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4859550561797753,
+      "grad_norm": 0.8946258425712585,
+      "learning_rate": 4.929852243402208e-06,
+      "loss": 0.7011,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4863061797752809,
+      "grad_norm": 0.9584969282150269,
+      "learning_rate": 4.929743451302656e-06,
+      "loss": 0.724,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4866573033707865,
+      "grad_norm": 0.9212526082992554,
+      "learning_rate": 4.929634576107956e-06,
+      "loss": 0.6821,
+      "step": 1386
+    },
+    {
+      "epoch": 0.48700842696629215,
+      "grad_norm": 0.9416231513023376,
+      "learning_rate": 4.929525617821834e-06,
+      "loss": 0.7457,
+      "step": 1387
+    },
+    {
+      "epoch": 0.48735955056179775,
+      "grad_norm": 0.9377028346061707,
+      "learning_rate": 4.929416576448014e-06,
+      "loss": 0.6871,
+      "step": 1388
+    },
+    {
+      "epoch": 0.48771067415730335,
+      "grad_norm": 0.9065041542053223,
+      "learning_rate": 4.929307451990226e-06,
+      "loss": 0.7261,
+      "step": 1389
+    },
+    {
+      "epoch": 0.488061797752809,
+      "grad_norm": 0.844350278377533,
+      "learning_rate": 4.929198244452202e-06,
+      "loss": 0.7099,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4884129213483146,
+      "grad_norm": 0.9054086804389954,
+      "learning_rate": 4.9290889538376775e-06,
+      "loss": 0.669,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4887640449438202,
+      "grad_norm": 0.8729206919670105,
+      "learning_rate": 4.928979580150388e-06,
+      "loss": 0.6816,
+      "step": 1392
+    },
+    {
+      "epoch": 0.48911516853932585,
+      "grad_norm": 0.9076043367385864,
+      "learning_rate": 4.928870123394077e-06,
+      "loss": 0.7044,
+      "step": 1393
+    },
+    {
+      "epoch": 0.48946629213483145,
+      "grad_norm": 0.9312844276428223,
+      "learning_rate": 4.928760583572485e-06,
+      "loss": 0.7107,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4898174157303371,
+      "grad_norm": 0.9900015592575073,
+      "learning_rate": 4.928650960689359e-06,
+      "loss": 0.703,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4901685393258427,
+      "grad_norm": 0.9289678931236267,
+      "learning_rate": 4.92854125474845e-06,
+      "loss": 0.714,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4905196629213483,
+      "grad_norm": 0.9454885125160217,
+      "learning_rate": 4.928431465753507e-06,
+      "loss": 0.7312,
+      "step": 1397
+    },
+    {
+      "epoch": 0.49087078651685395,
+      "grad_norm": 0.9117372035980225,
+      "learning_rate": 4.928321593708286e-06,
+      "loss": 0.6984,
+      "step": 1398
+    },
+    {
+      "epoch": 0.49122191011235955,
+      "grad_norm": 0.9010490775108337,
+      "learning_rate": 4.928211638616545e-06,
+      "loss": 0.672,
+      "step": 1399
+    },
+    {
+      "epoch": 0.49157303370786515,
+      "grad_norm": 0.9090813994407654,
+      "learning_rate": 4.928101600482043e-06,
+      "loss": 0.7132,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4919241573033708,
+      "grad_norm": 0.905497133731842,
+      "learning_rate": 4.927991479308545e-06,
+      "loss": 0.677,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4922752808988764,
+      "grad_norm": 0.902889609336853,
+      "learning_rate": 4.927881275099815e-06,
+      "loss": 0.6601,
+      "step": 1402
+    },
+    {
+      "epoch": 0.492626404494382,
+      "grad_norm": 0.9104518294334412,
+      "learning_rate": 4.927770987859624e-06,
+      "loss": 0.7192,
+      "step": 1403
+    },
+    {
+      "epoch": 0.49297752808988765,
+      "grad_norm": 0.8784206509590149,
+      "learning_rate": 4.9276606175917415e-06,
+      "loss": 0.6556,
+      "step": 1404
+    },
+    {
+      "epoch": 0.49332865168539325,
+      "grad_norm": 0.9406437277793884,
+      "learning_rate": 4.927550164299944e-06,
+      "loss": 0.679,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4936797752808989,
+      "grad_norm": 0.8956720232963562,
+      "learning_rate": 4.9274396279880075e-06,
+      "loss": 0.662,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4940308988764045,
+      "grad_norm": 0.9246423840522766,
+      "learning_rate": 4.927329008659714e-06,
+      "loss": 0.7155,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 0.9392451047897339,
+      "learning_rate": 4.927218306318845e-06,
+      "loss": 0.6819,
+      "step": 1408
+    },
+    {
+      "epoch": 0.49473314606741575,
+      "grad_norm": 0.9309847950935364,
+      "learning_rate": 4.927107520969187e-06,
+      "loss": 0.7195,
+      "step": 1409
+    },
+    {
+      "epoch": 0.49508426966292135,
+      "grad_norm": 0.890052080154419,
+      "learning_rate": 4.926996652614529e-06,
+      "loss": 0.6703,
+      "step": 1410
+    },
+    {
+      "epoch": 0.49543539325842695,
+      "grad_norm": 0.9011236429214478,
+      "learning_rate": 4.926885701258661e-06,
+      "loss": 0.6822,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4957865168539326,
+      "grad_norm": 0.915877103805542,
+      "learning_rate": 4.92677466690538e-06,
+      "loss": 0.7156,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4961376404494382,
+      "grad_norm": 0.8800505995750427,
+      "learning_rate": 4.9266635495584814e-06,
+      "loss": 0.6817,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4964887640449438,
+      "grad_norm": 0.9167606234550476,
+      "learning_rate": 4.926552349221766e-06,
+      "loss": 0.6982,
+      "step": 1414
+    },
+    {
+      "epoch": 0.49683988764044945,
+      "grad_norm": 0.9250441193580627,
+      "learning_rate": 4.926441065899037e-06,
+      "loss": 0.6559,
+      "step": 1415
+    },
+    {
+      "epoch": 0.49719101123595505,
+      "grad_norm": 0.910899817943573,
+      "learning_rate": 4.9263296995941004e-06,
+      "loss": 0.6654,
+      "step": 1416
+    },
+    {
+      "epoch": 0.49754213483146065,
+      "grad_norm": 0.8955602049827576,
+      "learning_rate": 4.926218250310764e-06,
+      "loss": 0.7022,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4978932584269663,
+      "grad_norm": 0.9195224642753601,
+      "learning_rate": 4.926106718052839e-06,
+      "loss": 0.7058,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4982443820224719,
+      "grad_norm": 0.9715495705604553,
+      "learning_rate": 4.92599510282414e-06,
+      "loss": 0.6917,
+      "step": 1419
+    },
+    {
+      "epoch": 0.49859550561797755,
+      "grad_norm": 0.8591254353523254,
+      "learning_rate": 4.925883404628485e-06,
+      "loss": 0.6415,
+      "step": 1420
+    },
+    {
+      "epoch": 0.49894662921348315,
+      "grad_norm": 0.903109610080719,
+      "learning_rate": 4.925771623469694e-06,
+      "loss": 0.702,
+      "step": 1421
+    },
+    {
+      "epoch": 0.49929775280898875,
+      "grad_norm": 0.9081088900566101,
+      "learning_rate": 4.9256597593515885e-06,
+      "loss": 0.7098,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4996488764044944,
+      "grad_norm": 0.9099250435829163,
+      "learning_rate": 4.925547812277994e-06,
+      "loss": 0.6752,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9163993000984192,
+      "learning_rate": 4.925435782252742e-06,
+      "loss": 0.7098,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5003511235955056,
+      "grad_norm": 0.9282369017601013,
+      "learning_rate": 4.92532366927966e-06,
+      "loss": 0.7003,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5007022471910112,
+      "grad_norm": 0.9375564455986023,
+      "learning_rate": 4.925211473362585e-06,
+      "loss": 0.6981,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5010533707865169,
+      "grad_norm": 0.9100632071495056,
+      "learning_rate": 4.925099194505352e-06,
+      "loss": 0.6794,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5014044943820225,
+      "grad_norm": 0.8592130541801453,
+      "learning_rate": 4.924986832711802e-06,
+      "loss": 0.6547,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5017556179775281,
+      "grad_norm": 0.9379070401191711,
+      "learning_rate": 4.924874387985777e-06,
+      "loss": 0.6829,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5021067415730337,
+      "grad_norm": 0.9097589254379272,
+      "learning_rate": 4.924761860331125e-06,
+      "loss": 0.682,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5024578651685393,
+      "grad_norm": 0.925594687461853,
+      "learning_rate": 4.92464924975169e-06,
+      "loss": 0.7312,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5028089887640449,
+      "grad_norm": 0.8913887739181519,
+      "learning_rate": 4.924536556251326e-06,
+      "loss": 0.6613,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5031601123595506,
+      "grad_norm": 0.8830445408821106,
+      "learning_rate": 4.924423779833887e-06,
+      "loss": 0.689,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5035112359550562,
+      "grad_norm": 0.9521420001983643,
+      "learning_rate": 4.924310920503229e-06,
+      "loss": 0.678,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5038623595505618,
+      "grad_norm": 0.918493390083313,
+      "learning_rate": 4.924197978263212e-06,
+      "loss": 0.7188,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5042134831460674,
+      "grad_norm": 0.9049700498580933,
+      "learning_rate": 4.9240849531176986e-06,
+      "loss": 0.6846,
+      "step": 1436
+    },
+    {
+      "epoch": 0.504564606741573,
+      "grad_norm": 0.8993126153945923,
+      "learning_rate": 4.9239718450705545e-06,
+      "loss": 0.7298,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5049157303370787,
+      "grad_norm": 0.9134989380836487,
+      "learning_rate": 4.923858654125647e-06,
+      "loss": 0.6945,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5052668539325843,
+      "grad_norm": 0.9160113334655762,
+      "learning_rate": 4.9237453802868475e-06,
+      "loss": 0.7074,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5056179775280899,
+      "grad_norm": 0.9044978022575378,
+      "learning_rate": 4.923632023558031e-06,
+      "loss": 0.6814,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5059691011235955,
+      "grad_norm": 0.9281947612762451,
+      "learning_rate": 4.923518583943072e-06,
+      "loss": 0.6806,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5063202247191011,
+      "grad_norm": 0.8679924607276917,
+      "learning_rate": 4.923405061445852e-06,
+      "loss": 0.6654,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5066713483146067,
+      "grad_norm": 0.905013918876648,
+      "learning_rate": 4.923291456070252e-06,
+      "loss": 0.6938,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5070224719101124,
+      "grad_norm": 0.8952536582946777,
+      "learning_rate": 4.923177767820158e-06,
+      "loss": 0.6802,
+      "step": 1444
+    },
+    {
+      "epoch": 0.507373595505618,
+      "grad_norm": 0.9175031185150146,
+      "learning_rate": 4.923063996699458e-06,
+      "loss": 0.6984,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5077247191011236,
+      "grad_norm": 0.8895412087440491,
+      "learning_rate": 4.922950142712044e-06,
+      "loss": 0.7255,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5080758426966292,
+      "grad_norm": 0.874991238117218,
+      "learning_rate": 4.922836205861807e-06,
+      "loss": 0.6709,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5084269662921348,
+      "grad_norm": 0.8750325441360474,
+      "learning_rate": 4.922722186152646e-06,
+      "loss": 0.6414,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5087780898876404,
+      "grad_norm": 0.8989003300666809,
+      "learning_rate": 4.922608083588459e-06,
+      "loss": 0.6928,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5091292134831461,
+      "grad_norm": 0.9451534152030945,
+      "learning_rate": 4.922493898173148e-06,
+      "loss": 0.7143,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5094803370786517,
+      "grad_norm": 0.9093509912490845,
+      "learning_rate": 4.922379629910619e-06,
+      "loss": 0.7027,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5098314606741573,
+      "grad_norm": 0.932784914970398,
+      "learning_rate": 4.9222652788047795e-06,
+      "loss": 0.7024,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5101825842696629,
+      "grad_norm": 0.9492952227592468,
+      "learning_rate": 4.92215084485954e-06,
+      "loss": 0.6856,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5105337078651685,
+      "grad_norm": 0.8805875182151794,
+      "learning_rate": 4.922036328078815e-06,
+      "loss": 0.7114,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5108848314606742,
+      "grad_norm": 0.8962169289588928,
+      "learning_rate": 4.92192172846652e-06,
+      "loss": 0.6962,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5112359550561798,
+      "grad_norm": 0.923744797706604,
+      "learning_rate": 4.921807046026574e-06,
+      "loss": 0.7325,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5115870786516854,
+      "grad_norm": 0.9285497069358826,
+      "learning_rate": 4.9216922807628995e-06,
+      "loss": 0.7208,
+      "step": 1457
+    },
+    {
+      "epoch": 0.511938202247191,
+      "grad_norm": 0.879386842250824,
+      "learning_rate": 4.921577432679421e-06,
+      "loss": 0.6404,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5122893258426966,
+      "grad_norm": 0.9378132820129395,
+      "learning_rate": 4.921462501780067e-06,
+      "loss": 0.7286,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5126404494382022,
+      "grad_norm": 0.9210785031318665,
+      "learning_rate": 4.921347488068767e-06,
+      "loss": 0.7344,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5129915730337079,
+      "grad_norm": 0.9228889346122742,
+      "learning_rate": 4.921232391549456e-06,
+      "loss": 0.7108,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5133426966292135,
+      "grad_norm": 0.8554604649543762,
+      "learning_rate": 4.9211172122260675e-06,
+      "loss": 0.6272,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5136938202247191,
+      "grad_norm": 0.8758804202079773,
+      "learning_rate": 4.921001950102543e-06,
+      "loss": 0.6463,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5140449438202247,
+      "grad_norm": 0.9541190266609192,
+      "learning_rate": 4.920886605182823e-06,
+      "loss": 0.7274,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5143960674157303,
+      "grad_norm": 0.9359861016273499,
+      "learning_rate": 4.920771177470853e-06,
+      "loss": 0.6961,
+      "step": 1465
+    },
+    {
+      "epoch": 0.514747191011236,
+      "grad_norm": 0.8696016073226929,
+      "learning_rate": 4.92065566697058e-06,
+      "loss": 0.7263,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5150983146067416,
+      "grad_norm": 0.8890820741653442,
+      "learning_rate": 4.920540073685956e-06,
+      "loss": 0.7063,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5154494382022472,
+      "grad_norm": 0.90772545337677,
+      "learning_rate": 4.920424397620931e-06,
+      "loss": 0.6858,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5158005617977528,
+      "grad_norm": 0.9410818219184875,
+      "learning_rate": 4.920308638779463e-06,
+      "loss": 0.6655,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5161516853932584,
+      "grad_norm": 0.8993090987205505,
+      "learning_rate": 4.920192797165511e-06,
+      "loss": 0.7215,
+      "step": 1470
+    },
+    {
+      "epoch": 0.516502808988764,
+      "grad_norm": 0.8889127969741821,
+      "learning_rate": 4.920076872783036e-06,
+      "loss": 0.7214,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5168539325842697,
+      "grad_norm": 0.9573481678962708,
+      "learning_rate": 4.919960865636004e-06,
+      "loss": 0.6754,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5172050561797753,
+      "grad_norm": 0.9702395796775818,
+      "learning_rate": 4.91984477572838e-06,
+      "loss": 0.6986,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5175561797752809,
+      "grad_norm": 0.9364273548126221,
+      "learning_rate": 4.919728603064136e-06,
+      "loss": 0.7098,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5179073033707865,
+      "grad_norm": 1.0006465911865234,
+      "learning_rate": 4.919612347647244e-06,
+      "loss": 0.7143,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5182584269662921,
+      "grad_norm": 0.9335517287254333,
+      "learning_rate": 4.91949600948168e-06,
+      "loss": 0.6837,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5186095505617978,
+      "grad_norm": 0.9131624698638916,
+      "learning_rate": 4.919379588571424e-06,
+      "loss": 0.7254,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5189606741573034,
+      "grad_norm": 0.9354121685028076,
+      "learning_rate": 4.919263084920455e-06,
+      "loss": 0.707,
+      "step": 1478
+    },
+    {
+      "epoch": 0.519311797752809,
+      "grad_norm": 0.8909806609153748,
+      "learning_rate": 4.919146498532758e-06,
+      "loss": 0.6655,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5196629213483146,
+      "grad_norm": 0.9220510721206665,
+      "learning_rate": 4.919029829412323e-06,
+      "loss": 0.6859,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5200140449438202,
+      "grad_norm": 0.9528585076332092,
+      "learning_rate": 4.918913077563137e-06,
+      "loss": 0.694,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5203651685393258,
+      "grad_norm": 0.9287347793579102,
+      "learning_rate": 4.918796242989193e-06,
+      "loss": 0.6775,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5207162921348315,
+      "grad_norm": 0.9118848443031311,
+      "learning_rate": 4.9186793256944885e-06,
+      "loss": 0.6869,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5210674157303371,
+      "grad_norm": 0.8987846374511719,
+      "learning_rate": 4.91856232568302e-06,
+      "loss": 0.6883,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5214185393258427,
+      "grad_norm": 0.8864469528198242,
+      "learning_rate": 4.918445242958789e-06,
+      "loss": 0.7144,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5217696629213483,
+      "grad_norm": 0.8598546385765076,
+      "learning_rate": 4.918328077525802e-06,
+      "loss": 0.6752,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5221207865168539,
+      "grad_norm": 0.8941078782081604,
+      "learning_rate": 4.918210829388061e-06,
+      "loss": 0.6733,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5224719101123596,
+      "grad_norm": 0.914069414138794,
+      "learning_rate": 4.918093498549582e-06,
+      "loss": 0.7032,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5228230337078652,
+      "grad_norm": 0.9472395181655884,
+      "learning_rate": 4.917976085014373e-06,
+      "loss": 0.6666,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5231741573033708,
+      "grad_norm": 0.9576021432876587,
+      "learning_rate": 4.9178585887864515e-06,
+      "loss": 0.7056,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5235252808988764,
+      "grad_norm": 0.9019284248352051,
+      "learning_rate": 4.9177410098698354e-06,
+      "loss": 0.6841,
+      "step": 1491
+    },
+    {
+      "epoch": 0.523876404494382,
+      "grad_norm": 0.8970656394958496,
+      "learning_rate": 4.917623348268546e-06,
+      "loss": 0.7169,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5242275280898876,
+      "grad_norm": 0.8948590159416199,
+      "learning_rate": 4.917505603986605e-06,
+      "loss": 0.6666,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5245786516853933,
+      "grad_norm": 0.9220554828643799,
+      "learning_rate": 4.9173877770280425e-06,
+      "loss": 0.7249,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5249297752808989,
+      "grad_norm": 0.9420086741447449,
+      "learning_rate": 4.917269867396887e-06,
+      "loss": 0.7205,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5252808988764045,
+      "grad_norm": 0.9278038144111633,
+      "learning_rate": 4.917151875097169e-06,
+      "loss": 0.7032,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5256320224719101,
+      "grad_norm": 0.8682548999786377,
+      "learning_rate": 4.917033800132927e-06,
+      "loss": 0.7,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5259831460674157,
+      "grad_norm": 0.8908578753471375,
+      "learning_rate": 4.916915642508196e-06,
+      "loss": 0.6891,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5263342696629213,
+      "grad_norm": 0.8928547501564026,
+      "learning_rate": 4.916797402227019e-06,
+      "loss": 0.7247,
+      "step": 1499
+    },
+    {
+      "epoch": 0.526685393258427,
+      "grad_norm": 0.9188054203987122,
+      "learning_rate": 4.916679079293438e-06,
+      "loss": 0.7001,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5270365168539326,
+      "grad_norm": 0.8950523138046265,
+      "learning_rate": 4.916560673711502e-06,
+      "loss": 0.6473,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5273876404494382,
+      "grad_norm": 0.9325989484786987,
+      "learning_rate": 4.916442185485258e-06,
+      "loss": 0.7096,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5277387640449438,
+      "grad_norm": 0.9268350005149841,
+      "learning_rate": 4.9163236146187586e-06,
+      "loss": 0.7089,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5280898876404494,
+      "grad_norm": 1.009692907333374,
+      "learning_rate": 4.91620496111606e-06,
+      "loss": 0.6721,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5284410112359551,
+      "grad_norm": 0.8952979445457458,
+      "learning_rate": 4.916086224981219e-06,
+      "loss": 0.7074,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5287921348314607,
+      "grad_norm": 0.896232008934021,
+      "learning_rate": 4.915967406218296e-06,
+      "loss": 0.6765,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5291432584269663,
+      "grad_norm": 0.9074482917785645,
+      "learning_rate": 4.915848504831356e-06,
+      "loss": 0.7296,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5294943820224719,
+      "grad_norm": 0.8722943067550659,
+      "learning_rate": 4.9157295208244635e-06,
+      "loss": 0.636,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5298455056179775,
+      "grad_norm": 0.917375922203064,
+      "learning_rate": 4.915610454201689e-06,
+      "loss": 0.6844,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5301966292134831,
+      "grad_norm": 0.8942883610725403,
+      "learning_rate": 4.9154913049671045e-06,
+      "loss": 0.6589,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5305477528089888,
+      "grad_norm": 0.894601583480835,
+      "learning_rate": 4.915372073124784e-06,
+      "loss": 0.6919,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5308988764044944,
+      "grad_norm": 1.0230928659439087,
+      "learning_rate": 4.9152527586788056e-06,
+      "loss": 0.6949,
+      "step": 1512
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 1.0016642808914185,
+      "learning_rate": 4.915133361633249e-06,
+      "loss": 0.6792,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5316011235955056,
+      "grad_norm": 0.8965452909469604,
+      "learning_rate": 4.9150138819921985e-06,
+      "loss": 0.6993,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5319522471910112,
+      "grad_norm": 0.9045464396476746,
+      "learning_rate": 4.91489431975974e-06,
+      "loss": 0.6905,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5323033707865169,
+      "grad_norm": 0.9282147288322449,
+      "learning_rate": 4.914774674939962e-06,
+      "loss": 0.7121,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5326544943820225,
+      "grad_norm": 0.9581758379936218,
+      "learning_rate": 4.914654947536956e-06,
+      "loss": 0.7156,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5330056179775281,
+      "grad_norm": 0.8986783623695374,
+      "learning_rate": 4.9145351375548184e-06,
+      "loss": 0.6827,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5333567415730337,
+      "grad_norm": 1.0093053579330444,
+      "learning_rate": 4.9144152449976444e-06,
+      "loss": 0.6642,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5337078651685393,
+      "grad_norm": 1.0274914503097534,
+      "learning_rate": 4.914295269869535e-06,
+      "loss": 0.6667,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5340589887640449,
+      "grad_norm": 0.9218011498451233,
+      "learning_rate": 4.914175212174594e-06,
+      "loss": 0.6902,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5344101123595506,
+      "grad_norm": 0.902390718460083,
+      "learning_rate": 4.914055071916925e-06,
+      "loss": 0.7009,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5347612359550562,
+      "grad_norm": 0.9239642024040222,
+      "learning_rate": 4.9139348491006396e-06,
+      "loss": 0.7038,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5351123595505618,
+      "grad_norm": 0.8971140384674072,
+      "learning_rate": 4.913814543729847e-06,
+      "loss": 0.7187,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5354634831460674,
+      "grad_norm": 0.9525965452194214,
+      "learning_rate": 4.913694155808663e-06,
+      "loss": 0.6694,
+      "step": 1525
+    },
+    {
+      "epoch": 0.535814606741573,
+      "grad_norm": 0.915665864944458,
+      "learning_rate": 4.913573685341205e-06,
+      "loss": 0.7181,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5361657303370787,
+      "grad_norm": 0.913693368434906,
+      "learning_rate": 4.913453132331592e-06,
+      "loss": 0.7122,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5365168539325843,
+      "grad_norm": 0.91572505235672,
+      "learning_rate": 4.913332496783947e-06,
+      "loss": 0.7143,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5368679775280899,
+      "grad_norm": 0.947823166847229,
+      "learning_rate": 4.913211778702395e-06,
+      "loss": 0.7239,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5372191011235955,
+      "grad_norm": 0.9384491443634033,
+      "learning_rate": 4.913090978091066e-06,
+      "loss": 0.7147,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5375702247191011,
+      "grad_norm": 0.882896363735199,
+      "learning_rate": 4.912970094954091e-06,
+      "loss": 0.6548,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5379213483146067,
+      "grad_norm": 0.8977818489074707,
+      "learning_rate": 4.912849129295601e-06,
+      "loss": 0.6898,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5382724719101124,
+      "grad_norm": 0.8661186099052429,
+      "learning_rate": 4.912728081119738e-06,
+      "loss": 0.6605,
+      "step": 1533
+    },
+    {
+      "epoch": 0.538623595505618,
+      "grad_norm": 0.8856679797172546,
+      "learning_rate": 4.9126069504306385e-06,
+      "loss": 0.6891,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5389747191011236,
+      "grad_norm": 0.9184430241584778,
+      "learning_rate": 4.9124857372324455e-06,
+      "loss": 0.7221,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 0.9198493957519531,
+      "learning_rate": 4.912364441529306e-06,
+      "loss": 0.7011,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5396769662921348,
+      "grad_norm": 0.9443540573120117,
+      "learning_rate": 4.912243063325365e-06,
+      "loss": 0.6523,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5400280898876404,
+      "grad_norm": 0.9004515409469604,
+      "learning_rate": 4.912121602624777e-06,
+      "loss": 0.6674,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5403792134831461,
+      "grad_norm": 0.9147352576255798,
+      "learning_rate": 4.912000059431693e-06,
+      "loss": 0.684,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5407303370786517,
+      "grad_norm": 0.9233749508857727,
+      "learning_rate": 4.911878433750271e-06,
+      "loss": 0.702,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5410814606741573,
+      "grad_norm": 0.8945138454437256,
+      "learning_rate": 4.911756725584672e-06,
+      "loss": 0.6694,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5414325842696629,
+      "grad_norm": 0.9332498908042908,
+      "learning_rate": 4.911634934939055e-06,
+      "loss": 0.703,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5417837078651685,
+      "grad_norm": 0.8984339237213135,
+      "learning_rate": 4.911513061817588e-06,
+      "loss": 0.6607,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5421348314606742,
+      "grad_norm": 0.9096390604972839,
+      "learning_rate": 4.9113911062244364e-06,
+      "loss": 0.6786,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5424859550561798,
+      "grad_norm": 0.8916593790054321,
+      "learning_rate": 4.911269068163775e-06,
+      "loss": 0.6644,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5428370786516854,
+      "grad_norm": 0.9222617149353027,
+      "learning_rate": 4.911146947639772e-06,
+      "loss": 0.6983,
+      "step": 1546
+    },
+    {
+      "epoch": 0.543188202247191,
+      "grad_norm": 0.8953000903129578,
+      "learning_rate": 4.911024744656608e-06,
+      "loss": 0.7026,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5435393258426966,
+      "grad_norm": 0.9042890071868896,
+      "learning_rate": 4.910902459218461e-06,
+      "loss": 0.7196,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5438904494382022,
+      "grad_norm": 0.9504362940788269,
+      "learning_rate": 4.910780091329513e-06,
+      "loss": 0.6707,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5442415730337079,
+      "grad_norm": 0.9307934641838074,
+      "learning_rate": 4.910657640993948e-06,
+      "loss": 0.683,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5445926966292135,
+      "grad_norm": 0.9294170141220093,
+      "learning_rate": 4.910535108215955e-06,
+      "loss": 0.6969,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5449438202247191,
+      "grad_norm": 0.9138461351394653,
+      "learning_rate": 4.910412492999725e-06,
+      "loss": 0.7347,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5452949438202247,
+      "grad_norm": 0.8926948308944702,
+      "learning_rate": 4.910289795349449e-06,
+      "loss": 0.7217,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5456460674157303,
+      "grad_norm": 0.8452489376068115,
+      "learning_rate": 4.910167015269325e-06,
+      "loss": 0.6478,
+      "step": 1554
+    },
+    {
+      "epoch": 0.545997191011236,
+      "grad_norm": 0.8928716778755188,
+      "learning_rate": 4.9100441527635516e-06,
+      "loss": 0.6246,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5463483146067416,
+      "grad_norm": 0.8877882361412048,
+      "learning_rate": 4.90992120783633e-06,
+      "loss": 0.6771,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5466994382022472,
+      "grad_norm": 0.8971658945083618,
+      "learning_rate": 4.909798180491865e-06,
+      "loss": 0.6664,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5470505617977528,
+      "grad_norm": 0.9224132895469666,
+      "learning_rate": 4.909675070734365e-06,
+      "loss": 0.6958,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5474016853932584,
+      "grad_norm": 0.9234408140182495,
+      "learning_rate": 4.9095518785680395e-06,
+      "loss": 0.6944,
+      "step": 1559
+    },
+    {
+      "epoch": 0.547752808988764,
+      "grad_norm": 0.9604536294937134,
+      "learning_rate": 4.909428603997101e-06,
+      "loss": 0.6597,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5481039325842697,
+      "grad_norm": 0.9268457889556885,
+      "learning_rate": 4.909305247025767e-06,
+      "loss": 0.6942,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5484550561797753,
+      "grad_norm": 0.9222556352615356,
+      "learning_rate": 4.909181807658256e-06,
+      "loss": 0.6747,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5488061797752809,
+      "grad_norm": 0.8885440826416016,
+      "learning_rate": 4.909058285898787e-06,
+      "loss": 0.6555,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5491573033707865,
+      "grad_norm": 0.9378191232681274,
+      "learning_rate": 4.908934681751586e-06,
+      "loss": 0.7272,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5495084269662921,
+      "grad_norm": 0.9337999820709229,
+      "learning_rate": 4.908810995220882e-06,
+      "loss": 0.6721,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5498595505617978,
+      "grad_norm": 0.8756289482116699,
+      "learning_rate": 4.908687226310902e-06,
+      "loss": 0.6733,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5502106741573034,
+      "grad_norm": 0.8794893026351929,
+      "learning_rate": 4.90856337502588e-06,
+      "loss": 0.6225,
+      "step": 1567
+    },
+    {
+      "epoch": 0.550561797752809,
+      "grad_norm": 0.8720929026603699,
+      "learning_rate": 4.9084394413700515e-06,
+      "loss": 0.7074,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5509129213483146,
+      "grad_norm": 0.8621073365211487,
+      "learning_rate": 4.908315425347655e-06,
+      "loss": 0.6439,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5512640449438202,
+      "grad_norm": 0.937415361404419,
+      "learning_rate": 4.908191326962932e-06,
+      "loss": 0.7121,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5516151685393258,
+      "grad_norm": 0.908561646938324,
+      "learning_rate": 4.908067146220127e-06,
+      "loss": 0.699,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5519662921348315,
+      "grad_norm": 0.8949597477912903,
+      "learning_rate": 4.907942883123485e-06,
+      "loss": 0.7148,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5523174157303371,
+      "grad_norm": 0.9320929646492004,
+      "learning_rate": 4.907818537677258e-06,
+      "loss": 0.6894,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5526685393258427,
+      "grad_norm": 0.8810732960700989,
+      "learning_rate": 4.907694109885697e-06,
+      "loss": 0.6814,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5530196629213483,
+      "grad_norm": 0.9485206007957458,
+      "learning_rate": 4.907569599753058e-06,
+      "loss": 0.7125,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5533707865168539,
+      "grad_norm": 0.8823009729385376,
+      "learning_rate": 4.907445007283598e-06,
+      "loss": 0.7239,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5537219101123596,
+      "grad_norm": 0.9091922640800476,
+      "learning_rate": 4.907320332481579e-06,
+      "loss": 0.6949,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5540730337078652,
+      "grad_norm": 0.9078152775764465,
+      "learning_rate": 4.907195575351265e-06,
+      "loss": 0.701,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5544241573033708,
+      "grad_norm": 0.9011525511741638,
+      "learning_rate": 4.907070735896923e-06,
+      "loss": 0.7039,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5547752808988764,
+      "grad_norm": 0.8794327974319458,
+      "learning_rate": 4.90694581412282e-06,
+      "loss": 0.6543,
+      "step": 1580
+    },
+    {
+      "epoch": 0.555126404494382,
+      "grad_norm": 0.921090841293335,
+      "learning_rate": 4.90682081003323e-06,
+      "loss": 0.7031,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5554775280898876,
+      "grad_norm": 0.8931931853294373,
+      "learning_rate": 4.906695723632429e-06,
+      "loss": 0.6938,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5558286516853933,
+      "grad_norm": 0.9168527722358704,
+      "learning_rate": 4.9065705549246935e-06,
+      "loss": 0.7472,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5561797752808989,
+      "grad_norm": 0.9369787573814392,
+      "learning_rate": 4.906445303914303e-06,
+      "loss": 0.6945,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5565308988764045,
+      "grad_norm": 0.9574870467185974,
+      "learning_rate": 4.906319970605544e-06,
+      "loss": 0.7117,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5568820224719101,
+      "grad_norm": 0.9056572318077087,
+      "learning_rate": 4.9061945550027e-06,
+      "loss": 0.7117,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5572331460674157,
+      "grad_norm": 0.9152799844741821,
+      "learning_rate": 4.906069057110062e-06,
+      "loss": 0.7262,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5575842696629213,
+      "grad_norm": 0.9061770439147949,
+      "learning_rate": 4.9059434769319205e-06,
+      "loss": 0.6746,
+      "step": 1588
+    },
+    {
+      "epoch": 0.557935393258427,
+      "grad_norm": 0.8756861090660095,
+      "learning_rate": 4.905817814472572e-06,
+      "loss": 0.6547,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5582865168539326,
+      "grad_norm": 0.8801287412643433,
+      "learning_rate": 4.905692069736312e-06,
+      "loss": 0.7026,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5586376404494382,
+      "grad_norm": 0.871894359588623,
+      "learning_rate": 4.9055662427274415e-06,
+      "loss": 0.6967,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5589887640449438,
+      "grad_norm": 0.8951716423034668,
+      "learning_rate": 4.905440333450264e-06,
+      "loss": 0.6769,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5593398876404494,
+      "grad_norm": 0.8958552479743958,
+      "learning_rate": 4.905314341909086e-06,
+      "loss": 0.6462,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5596910112359551,
+      "grad_norm": 0.8950044512748718,
+      "learning_rate": 4.9051882681082155e-06,
+      "loss": 0.6943,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5600421348314607,
+      "grad_norm": 0.8931154608726501,
+      "learning_rate": 4.905062112051965e-06,
+      "loss": 0.6712,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5603932584269663,
+      "grad_norm": 0.9284601211547852,
+      "learning_rate": 4.904935873744647e-06,
+      "loss": 0.7057,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5607443820224719,
+      "grad_norm": 0.8948965072631836,
+      "learning_rate": 4.904809553190582e-06,
+      "loss": 0.6923,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5610955056179775,
+      "grad_norm": 0.9073783755302429,
+      "learning_rate": 4.904683150394086e-06,
+      "loss": 0.7183,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5614466292134831,
+      "grad_norm": 0.9169145226478577,
+      "learning_rate": 4.904556665359485e-06,
+      "loss": 0.71,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5617977528089888,
+      "grad_norm": 1.0053675174713135,
+      "learning_rate": 4.904430098091104e-06,
+      "loss": 0.7078,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5621488764044944,
+      "grad_norm": 0.893157422542572,
+      "learning_rate": 4.904303448593271e-06,
+      "loss": 0.6814,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.8943068981170654,
+      "learning_rate": 4.904176716870317e-06,
+      "loss": 0.6915,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5628511235955056,
+      "grad_norm": 0.8920650482177734,
+      "learning_rate": 4.904049902926576e-06,
+      "loss": 0.6953,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5632022471910112,
+      "grad_norm": 0.8903055787086487,
+      "learning_rate": 4.9039230067663865e-06,
+      "loss": 0.6708,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5635533707865169,
+      "grad_norm": 0.9348129034042358,
+      "learning_rate": 4.903796028394086e-06,
+      "loss": 0.6983,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5639044943820225,
+      "grad_norm": 0.946652889251709,
+      "learning_rate": 4.903668967814021e-06,
+      "loss": 0.6997,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5642556179775281,
+      "grad_norm": 0.9114198684692383,
+      "learning_rate": 4.9035418250305314e-06,
+      "loss": 0.6741,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5646067415730337,
+      "grad_norm": 0.8928046226501465,
+      "learning_rate": 4.903414600047969e-06,
+      "loss": 0.6866,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5649578651685393,
+      "grad_norm": 0.9179702997207642,
+      "learning_rate": 4.903287292870684e-06,
+      "loss": 0.696,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5653089887640449,
+      "grad_norm": 0.9323627948760986,
+      "learning_rate": 4.903159903503031e-06,
+      "loss": 0.6939,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5656601123595506,
+      "grad_norm": 0.8928480744361877,
+      "learning_rate": 4.903032431949365e-06,
+      "loss": 0.6819,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5660112359550562,
+      "grad_norm": 0.8628221154212952,
+      "learning_rate": 4.9029048782140454e-06,
+      "loss": 0.687,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5663623595505618,
+      "grad_norm": 0.8917481303215027,
+      "learning_rate": 4.9027772423014354e-06,
+      "loss": 0.7087,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5667134831460674,
+      "grad_norm": 0.8859778642654419,
+      "learning_rate": 4.9026495242159e-06,
+      "loss": 0.6619,
+      "step": 1614
+    },
+    {
+      "epoch": 0.567064606741573,
+      "grad_norm": 0.9722530841827393,
+      "learning_rate": 4.902521723961807e-06,
+      "loss": 0.7133,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5674157303370787,
+      "grad_norm": 0.8985425233840942,
+      "learning_rate": 4.9023938415435265e-06,
+      "loss": 0.6599,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5677668539325843,
+      "grad_norm": 0.9105679392814636,
+      "learning_rate": 4.902265876965434e-06,
+      "loss": 0.6985,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5681179775280899,
+      "grad_norm": 0.8673763871192932,
+      "learning_rate": 4.902137830231903e-06,
+      "loss": 0.692,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5684691011235955,
+      "grad_norm": 0.8669525384902954,
+      "learning_rate": 4.902009701347313e-06,
+      "loss": 0.6515,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5688202247191011,
+      "grad_norm": 0.9088237881660461,
+      "learning_rate": 4.9018814903160475e-06,
+      "loss": 0.6847,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5691713483146067,
+      "grad_norm": 0.9107149243354797,
+      "learning_rate": 4.90175319714249e-06,
+      "loss": 0.6769,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5695224719101124,
+      "grad_norm": 0.8820178508758545,
+      "learning_rate": 4.901624821831028e-06,
+      "loss": 0.7169,
+      "step": 1622
+    },
+    {
+      "epoch": 0.569873595505618,
+      "grad_norm": 0.8633939623832703,
+      "learning_rate": 4.901496364386053e-06,
+      "loss": 0.6395,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5702247191011236,
+      "grad_norm": 0.9044683575630188,
+      "learning_rate": 4.901367824811958e-06,
+      "loss": 0.7011,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5705758426966292,
+      "grad_norm": 0.9092311263084412,
+      "learning_rate": 4.901239203113137e-06,
+      "loss": 0.6643,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5709269662921348,
+      "grad_norm": 0.9932267665863037,
+      "learning_rate": 4.90111049929399e-06,
+      "loss": 0.7203,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5712780898876404,
+      "grad_norm": 0.8947046995162964,
+      "learning_rate": 4.9009817133589186e-06,
+      "loss": 0.683,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5716292134831461,
+      "grad_norm": 0.9035671353340149,
+      "learning_rate": 4.900852845312328e-06,
+      "loss": 0.6761,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5719803370786517,
+      "grad_norm": 0.905048668384552,
+      "learning_rate": 4.900723895158624e-06,
+      "loss": 0.7018,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5723314606741573,
+      "grad_norm": 0.9121686220169067,
+      "learning_rate": 4.9005948629022175e-06,
+      "loss": 0.7506,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5726825842696629,
+      "grad_norm": 0.9041679501533508,
+      "learning_rate": 4.90046574854752e-06,
+      "loss": 0.7058,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5730337078651685,
+      "grad_norm": 0.8760393857955933,
+      "learning_rate": 4.900336552098949e-06,
+      "loss": 0.7067,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5733848314606742,
+      "grad_norm": 0.9228975772857666,
+      "learning_rate": 4.900207273560921e-06,
+      "loss": 0.6959,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5737359550561798,
+      "grad_norm": 0.8803319334983826,
+      "learning_rate": 4.900077912937858e-06,
+      "loss": 0.634,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5740870786516854,
+      "grad_norm": 0.8787641525268555,
+      "learning_rate": 4.899948470234185e-06,
+      "loss": 0.7098,
+      "step": 1635
+    },
+    {
+      "epoch": 0.574438202247191,
+      "grad_norm": 0.8598030209541321,
+      "learning_rate": 4.899818945454328e-06,
+      "loss": 0.7062,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5747893258426966,
+      "grad_norm": 0.920392632484436,
+      "learning_rate": 4.899689338602716e-06,
+      "loss": 0.6514,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5751404494382022,
+      "grad_norm": 0.9178962111473083,
+      "learning_rate": 4.899559649683783e-06,
+      "loss": 0.6626,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5754915730337079,
+      "grad_norm": 0.8671183586120605,
+      "learning_rate": 4.899429878701963e-06,
+      "loss": 0.704,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5758426966292135,
+      "grad_norm": 0.8965294361114502,
+      "learning_rate": 4.899300025661694e-06,
+      "loss": 0.7058,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5761938202247191,
+      "grad_norm": 0.8801050186157227,
+      "learning_rate": 4.8991700905674165e-06,
+      "loss": 0.695,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5765449438202247,
+      "grad_norm": 0.9138525724411011,
+      "learning_rate": 4.899040073423577e-06,
+      "loss": 0.7061,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5768960674157303,
+      "grad_norm": 0.8814200162887573,
+      "learning_rate": 4.898909974234618e-06,
+      "loss": 0.6659,
+      "step": 1643
+    },
+    {
+      "epoch": 0.577247191011236,
+      "grad_norm": 0.927841305732727,
+      "learning_rate": 4.898779793004991e-06,
+      "loss": 0.6864,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5775983146067416,
+      "grad_norm": 0.9304884076118469,
+      "learning_rate": 4.898649529739148e-06,
+      "loss": 0.66,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5779494382022472,
+      "grad_norm": 0.8991161584854126,
+      "learning_rate": 4.898519184441544e-06,
+      "loss": 0.6901,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5783005617977528,
+      "grad_norm": 0.8874373435974121,
+      "learning_rate": 4.898388757116636e-06,
+      "loss": 0.6866,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5786516853932584,
+      "grad_norm": 0.8752896785736084,
+      "learning_rate": 4.898258247768885e-06,
+      "loss": 0.6966,
+      "step": 1648
+    },
+    {
+      "epoch": 0.579002808988764,
+      "grad_norm": 0.9005473852157593,
+      "learning_rate": 4.898127656402753e-06,
+      "loss": 0.7192,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5793539325842697,
+      "grad_norm": 0.8968993425369263,
+      "learning_rate": 4.897996983022709e-06,
+      "loss": 0.7156,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5797050561797753,
+      "grad_norm": 0.881846010684967,
+      "learning_rate": 4.897866227633219e-06,
+      "loss": 0.6848,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5800561797752809,
+      "grad_norm": 0.9290207028388977,
+      "learning_rate": 4.897735390238756e-06,
+      "loss": 0.7263,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5804073033707865,
+      "grad_norm": 0.8702242374420166,
+      "learning_rate": 4.897604470843795e-06,
+      "loss": 0.6796,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5807584269662921,
+      "grad_norm": 0.9286425709724426,
+      "learning_rate": 4.897473469452812e-06,
+      "loss": 0.7203,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5811095505617978,
+      "grad_norm": 0.9703604578971863,
+      "learning_rate": 4.897342386070289e-06,
+      "loss": 0.6574,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5814606741573034,
+      "grad_norm": 0.9450727701187134,
+      "learning_rate": 4.897211220700707e-06,
+      "loss": 0.7004,
+      "step": 1656
+    },
+    {
+      "epoch": 0.581811797752809,
+      "grad_norm": 0.8609772324562073,
+      "learning_rate": 4.897079973348553e-06,
+      "loss": 0.658,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5821629213483146,
+      "grad_norm": 0.8902559876441956,
+      "learning_rate": 4.896948644018315e-06,
+      "loss": 0.6995,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5825140449438202,
+      "grad_norm": 0.8805756568908691,
+      "learning_rate": 4.896817232714484e-06,
+      "loss": 0.6919,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5828651685393258,
+      "grad_norm": 0.8647357225418091,
+      "learning_rate": 4.896685739441555e-06,
+      "loss": 0.7053,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5832162921348315,
+      "grad_norm": 0.8912749290466309,
+      "learning_rate": 4.896554164204025e-06,
+      "loss": 0.6848,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5835674157303371,
+      "grad_norm": 0.8912277221679688,
+      "learning_rate": 4.896422507006392e-06,
+      "loss": 0.687,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5839185393258427,
+      "grad_norm": 0.8887133598327637,
+      "learning_rate": 4.896290767853161e-06,
+      "loss": 0.6703,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 0.9464722871780396,
+      "learning_rate": 4.896158946748837e-06,
+      "loss": 0.6834,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5846207865168539,
+      "grad_norm": 0.8864126801490784,
+      "learning_rate": 4.8960270436979265e-06,
+      "loss": 0.6983,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5849719101123596,
+      "grad_norm": 0.9276545643806458,
+      "learning_rate": 4.895895058704941e-06,
+      "loss": 0.6989,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5853230337078652,
+      "grad_norm": 0.9041825532913208,
+      "learning_rate": 4.8957629917743955e-06,
+      "loss": 0.7188,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5856741573033708,
+      "grad_norm": 0.9835352301597595,
+      "learning_rate": 4.895630842910805e-06,
+      "loss": 0.6981,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5860252808988764,
+      "grad_norm": 0.8841162323951721,
+      "learning_rate": 4.89549861211869e-06,
+      "loss": 0.6686,
+      "step": 1669
+    },
+    {
+      "epoch": 0.586376404494382,
+      "grad_norm": 0.9026679992675781,
+      "learning_rate": 4.895366299402572e-06,
+      "loss": 0.6782,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5867275280898876,
+      "grad_norm": 0.8805919885635376,
+      "learning_rate": 4.895233904766978e-06,
+      "loss": 0.7267,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5870786516853933,
+      "grad_norm": 0.8851740956306458,
+      "learning_rate": 4.895101428216432e-06,
+      "loss": 0.6841,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5874297752808989,
+      "grad_norm": 0.911161482334137,
+      "learning_rate": 4.894968869755467e-06,
+      "loss": 0.6938,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5877808988764045,
+      "grad_norm": 0.900994598865509,
+      "learning_rate": 4.894836229388617e-06,
+      "loss": 0.6809,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5881320224719101,
+      "grad_norm": 0.9121220111846924,
+      "learning_rate": 4.894703507120417e-06,
+      "loss": 0.6941,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5884831460674157,
+      "grad_norm": 0.8991403579711914,
+      "learning_rate": 4.894570702955406e-06,
+      "loss": 0.6925,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5888342696629213,
+      "grad_norm": 0.8850685358047485,
+      "learning_rate": 4.894437816898125e-06,
+      "loss": 0.6904,
+      "step": 1677
+    },
+    {
+      "epoch": 0.589185393258427,
+      "grad_norm": 0.8999640345573425,
+      "learning_rate": 4.894304848953122e-06,
+      "loss": 0.6959,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5895365168539326,
+      "grad_norm": 0.9228804707527161,
+      "learning_rate": 4.89417179912494e-06,
+      "loss": 0.6663,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5898876404494382,
+      "grad_norm": 0.8978350162506104,
+      "learning_rate": 4.894038667418132e-06,
+      "loss": 0.6948,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5902387640449438,
+      "grad_norm": 0.9258869290351868,
+      "learning_rate": 4.89390545383725e-06,
+      "loss": 0.7268,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5905898876404494,
+      "grad_norm": 0.9010106325149536,
+      "learning_rate": 4.893772158386851e-06,
+      "loss": 0.6819,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5909410112359551,
+      "grad_norm": 0.9146981835365295,
+      "learning_rate": 4.8936387810714904e-06,
+      "loss": 0.6942,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5912921348314607,
+      "grad_norm": 0.959784746170044,
+      "learning_rate": 4.893505321895733e-06,
+      "loss": 0.7312,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5916432584269663,
+      "grad_norm": 0.9104639887809753,
+      "learning_rate": 4.8933717808641415e-06,
+      "loss": 0.6842,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5919943820224719,
+      "grad_norm": 0.8836262822151184,
+      "learning_rate": 4.893238157981283e-06,
+      "loss": 0.6613,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5923455056179775,
+      "grad_norm": 0.903002142906189,
+      "learning_rate": 4.893104453251728e-06,
+      "loss": 0.6558,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5926966292134831,
+      "grad_norm": 0.9008224010467529,
+      "learning_rate": 4.892970666680048e-06,
+      "loss": 0.7283,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5930477528089888,
+      "grad_norm": 0.9016857743263245,
+      "learning_rate": 4.892836798270819e-06,
+      "loss": 0.704,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5933988764044944,
+      "grad_norm": 0.9839923977851868,
+      "learning_rate": 4.8927028480286195e-06,
+      "loss": 0.6804,
+      "step": 1690
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 0.9112361669540405,
+      "learning_rate": 4.8925688159580294e-06,
+      "loss": 0.6885,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5941011235955056,
+      "grad_norm": 0.9427493214607239,
+      "learning_rate": 4.892434702063634e-06,
+      "loss": 0.6816,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5944522471910112,
+      "grad_norm": 0.9189691543579102,
+      "learning_rate": 4.892300506350018e-06,
+      "loss": 0.7258,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5948033707865169,
+      "grad_norm": 0.9497784972190857,
+      "learning_rate": 4.892166228821772e-06,
+      "loss": 0.6805,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5951544943820225,
+      "grad_norm": 0.867682933807373,
+      "learning_rate": 4.892031869483489e-06,
+      "loss": 0.6952,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5955056179775281,
+      "grad_norm": 0.9247297644615173,
+      "learning_rate": 4.891897428339763e-06,
+      "loss": 0.6892,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5958567415730337,
+      "grad_norm": 0.9473358392715454,
+      "learning_rate": 4.89176290539519e-06,
+      "loss": 0.704,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5962078651685393,
+      "grad_norm": 0.9231151938438416,
+      "learning_rate": 4.891628300654373e-06,
+      "loss": 0.725,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5965589887640449,
+      "grad_norm": 0.9092977046966553,
+      "learning_rate": 4.891493614121915e-06,
+      "loss": 0.7209,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5969101123595506,
+      "grad_norm": 0.9357640147209167,
+      "learning_rate": 4.891358845802422e-06,
+      "loss": 0.6577,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5972612359550562,
+      "grad_norm": 0.9201985001564026,
+      "learning_rate": 4.891223995700503e-06,
+      "loss": 0.7037,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5976123595505618,
+      "grad_norm": 0.8873964548110962,
+      "learning_rate": 4.8910890638207685e-06,
+      "loss": 0.6655,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5979634831460674,
+      "grad_norm": 0.8884997367858887,
+      "learning_rate": 4.890954050167835e-06,
+      "loss": 0.6927,
+      "step": 1703
+    },
+    {
+      "epoch": 0.598314606741573,
+      "grad_norm": 0.9006896018981934,
+      "learning_rate": 4.890818954746318e-06,
+      "loss": 0.6879,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5986657303370787,
+      "grad_norm": 0.8921645283699036,
+      "learning_rate": 4.8906837775608395e-06,
+      "loss": 0.7091,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5990168539325843,
+      "grad_norm": 0.8784016370773315,
+      "learning_rate": 4.890548518616022e-06,
+      "loss": 0.6611,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5993679775280899,
+      "grad_norm": 0.8995918035507202,
+      "learning_rate": 4.89041317791649e-06,
+      "loss": 0.699,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5997191011235955,
+      "grad_norm": 0.9034433364868164,
+      "learning_rate": 4.890277755466872e-06,
+      "loss": 0.6861,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6000702247191011,
+      "grad_norm": 0.896384596824646,
+      "learning_rate": 4.890142251271801e-06,
+      "loss": 0.6658,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6004213483146067,
+      "grad_norm": 0.8753630518913269,
+      "learning_rate": 4.890006665335911e-06,
+      "loss": 0.6886,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6007724719101124,
+      "grad_norm": 0.9117717146873474,
+      "learning_rate": 4.889870997663837e-06,
+      "loss": 0.6885,
+      "step": 1711
+    },
+    {
+      "epoch": 0.601123595505618,
+      "grad_norm": 0.9002839922904968,
+      "learning_rate": 4.889735248260221e-06,
+      "loss": 0.6707,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6014747191011236,
+      "grad_norm": 0.849831223487854,
+      "learning_rate": 4.889599417129703e-06,
+      "loss": 0.6458,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6018258426966292,
+      "grad_norm": 0.9035419225692749,
+      "learning_rate": 4.889463504276931e-06,
+      "loss": 0.6465,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6021769662921348,
+      "grad_norm": 0.8928185701370239,
+      "learning_rate": 4.889327509706552e-06,
+      "loss": 0.6596,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6025280898876404,
+      "grad_norm": 0.9307925701141357,
+      "learning_rate": 4.889191433423216e-06,
+      "loss": 0.6855,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6028792134831461,
+      "grad_norm": 0.8939083814620972,
+      "learning_rate": 4.889055275431578e-06,
+      "loss": 0.7251,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6032303370786517,
+      "grad_norm": 0.8944793343544006,
+      "learning_rate": 4.888919035736293e-06,
+      "loss": 0.6952,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6035814606741573,
+      "grad_norm": 0.92487633228302,
+      "learning_rate": 4.888782714342022e-06,
+      "loss": 0.685,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6039325842696629,
+      "grad_norm": 0.9212489724159241,
+      "learning_rate": 4.888646311253426e-06,
+      "loss": 0.6487,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6042837078651685,
+      "grad_norm": 0.9022998213768005,
+      "learning_rate": 4.888509826475171e-06,
+      "loss": 0.6699,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6046348314606742,
+      "grad_norm": 0.868482232093811,
+      "learning_rate": 4.8883732600119225e-06,
+      "loss": 0.6771,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6049859550561798,
+      "grad_norm": 0.8934330344200134,
+      "learning_rate": 4.888236611868353e-06,
+      "loss": 0.7027,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6053370786516854,
+      "grad_norm": 0.9520792961120605,
+      "learning_rate": 4.888099882049135e-06,
+      "loss": 0.701,
+      "step": 1724
+    },
+    {
+      "epoch": 0.605688202247191,
+      "grad_norm": 0.921665370464325,
+      "learning_rate": 4.887963070558944e-06,
+      "loss": 0.7058,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6060393258426966,
+      "grad_norm": 0.8855586647987366,
+      "learning_rate": 4.88782617740246e-06,
+      "loss": 0.6789,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6063904494382022,
+      "grad_norm": 0.9030165672302246,
+      "learning_rate": 4.887689202584364e-06,
+      "loss": 0.6883,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6067415730337079,
+      "grad_norm": 0.8876542448997498,
+      "learning_rate": 4.88755214610934e-06,
+      "loss": 0.713,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6070926966292135,
+      "grad_norm": 0.9762006998062134,
+      "learning_rate": 4.887415007982076e-06,
+      "loss": 0.7003,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6074438202247191,
+      "grad_norm": 0.9035171270370483,
+      "learning_rate": 4.887277788207261e-06,
+      "loss": 0.7089,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6077949438202247,
+      "grad_norm": 0.8891435265541077,
+      "learning_rate": 4.887140486789589e-06,
+      "loss": 0.6846,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6081460674157303,
+      "grad_norm": 0.899974524974823,
+      "learning_rate": 4.887003103733755e-06,
+      "loss": 0.6662,
+      "step": 1732
+    },
+    {
+      "epoch": 0.608497191011236,
+      "grad_norm": 0.958362340927124,
+      "learning_rate": 4.886865639044458e-06,
+      "loss": 0.6732,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6088483146067416,
+      "grad_norm": 0.9114941954612732,
+      "learning_rate": 4.8867280927263985e-06,
+      "loss": 0.6872,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6091994382022472,
+      "grad_norm": 0.9042050242424011,
+      "learning_rate": 4.88659046478428e-06,
+      "loss": 0.6654,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6095505617977528,
+      "grad_norm": 0.8774390816688538,
+      "learning_rate": 4.88645275522281e-06,
+      "loss": 0.666,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6099016853932584,
+      "grad_norm": 0.8930482268333435,
+      "learning_rate": 4.8863149640466975e-06,
+      "loss": 0.6796,
+      "step": 1737
+    },
+    {
+      "epoch": 0.610252808988764,
+      "grad_norm": 0.9068736433982849,
+      "learning_rate": 4.886177091260656e-06,
+      "loss": 0.6599,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6106039325842697,
+      "grad_norm": 0.8848494291305542,
+      "learning_rate": 4.8860391368694e-06,
+      "loss": 0.7131,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6109550561797753,
+      "grad_norm": 0.9204060435295105,
+      "learning_rate": 4.885901100877647e-06,
+      "loss": 0.7236,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6113061797752809,
+      "grad_norm": 0.912101149559021,
+      "learning_rate": 4.885762983290117e-06,
+      "loss": 0.7401,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6116573033707865,
+      "grad_norm": 0.8925957083702087,
+      "learning_rate": 4.885624784111536e-06,
+      "loss": 0.6837,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6120084269662921,
+      "grad_norm": 0.9438378214836121,
+      "learning_rate": 4.885486503346628e-06,
+      "loss": 0.6398,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6123595505617978,
+      "grad_norm": 0.885036826133728,
+      "learning_rate": 4.8853481410001225e-06,
+      "loss": 0.6341,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6127106741573034,
+      "grad_norm": 0.8819115161895752,
+      "learning_rate": 4.885209697076752e-06,
+      "loss": 0.673,
+      "step": 1745
+    },
+    {
+      "epoch": 0.613061797752809,
+      "grad_norm": 0.9269865155220032,
+      "learning_rate": 4.8850711715812515e-06,
+      "loss": 0.7188,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6134129213483146,
+      "grad_norm": 0.9204302430152893,
+      "learning_rate": 4.8849325645183575e-06,
+      "loss": 0.7151,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6137640449438202,
+      "grad_norm": 0.9573183655738831,
+      "learning_rate": 4.88479387589281e-06,
+      "loss": 0.7048,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6141151685393258,
+      "grad_norm": 1.0055536031723022,
+      "learning_rate": 4.884655105709354e-06,
+      "loss": 0.6865,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6144662921348315,
+      "grad_norm": 0.9006555080413818,
+      "learning_rate": 4.8845162539727335e-06,
+      "loss": 0.6931,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6148174157303371,
+      "grad_norm": 0.8895266056060791,
+      "learning_rate": 4.884377320687697e-06,
+      "loss": 0.6788,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6151685393258427,
+      "grad_norm": 0.8907827138900757,
+      "learning_rate": 4.884238305858997e-06,
+      "loss": 0.6559,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6155196629213483,
+      "grad_norm": 0.8671404719352722,
+      "learning_rate": 4.884099209491386e-06,
+      "loss": 0.692,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6158707865168539,
+      "grad_norm": 0.8965876698493958,
+      "learning_rate": 4.883960031589624e-06,
+      "loss": 0.6662,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6162219101123596,
+      "grad_norm": 0.8783154487609863,
+      "learning_rate": 4.883820772158467e-06,
+      "loss": 0.6802,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6165730337078652,
+      "grad_norm": 0.9085580706596375,
+      "learning_rate": 4.883681431202682e-06,
+      "loss": 0.7184,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6169241573033708,
+      "grad_norm": 0.9406449794769287,
+      "learning_rate": 4.883542008727029e-06,
+      "loss": 0.6998,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6172752808988764,
+      "grad_norm": 0.9234215021133423,
+      "learning_rate": 4.88340250473628e-06,
+      "loss": 0.7069,
+      "step": 1758
+    },
+    {
+      "epoch": 0.617626404494382,
+      "grad_norm": 0.8770806789398193,
+      "learning_rate": 4.883262919235205e-06,
+      "loss": 0.738,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6179775280898876,
+      "grad_norm": 0.9186333417892456,
+      "learning_rate": 4.8831232522285786e-06,
+      "loss": 0.7019,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6183286516853933,
+      "grad_norm": 0.9067515730857849,
+      "learning_rate": 4.8829835037211745e-06,
+      "loss": 0.6724,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6186797752808989,
+      "grad_norm": 0.962090790271759,
+      "learning_rate": 4.8828436737177744e-06,
+      "loss": 0.6521,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6190308988764045,
+      "grad_norm": 0.8877134323120117,
+      "learning_rate": 4.88270376222316e-06,
+      "loss": 0.7029,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6193820224719101,
+      "grad_norm": 0.9146241545677185,
+      "learning_rate": 4.882563769242116e-06,
+      "loss": 0.6999,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6197331460674157,
+      "grad_norm": 0.8948286175727844,
+      "learning_rate": 4.882423694779429e-06,
+      "loss": 0.6677,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6200842696629213,
+      "grad_norm": 0.958892285823822,
+      "learning_rate": 4.88228353883989e-06,
+      "loss": 0.7261,
+      "step": 1766
+    },
+    {
+      "epoch": 0.620435393258427,
+      "grad_norm": 0.906204104423523,
+      "learning_rate": 4.882143301428293e-06,
+      "loss": 0.671,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6207865168539326,
+      "grad_norm": 0.9508899450302124,
+      "learning_rate": 4.882002982549435e-06,
+      "loss": 0.6718,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6211376404494382,
+      "grad_norm": 0.8806086778640747,
+      "learning_rate": 4.8818625822081116e-06,
+      "loss": 0.6835,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6214887640449438,
+      "grad_norm": 0.9053204655647278,
+      "learning_rate": 4.8817221004091266e-06,
+      "loss": 0.7018,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6218398876404494,
+      "grad_norm": 0.8768627047538757,
+      "learning_rate": 4.881581537157284e-06,
+      "loss": 0.6819,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6221910112359551,
+      "grad_norm": 0.8701360821723938,
+      "learning_rate": 4.881440892457391e-06,
+      "loss": 0.6766,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6225421348314607,
+      "grad_norm": 0.9027068614959717,
+      "learning_rate": 4.881300166314257e-06,
+      "loss": 0.6819,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6228932584269663,
+      "grad_norm": 0.9427441954612732,
+      "learning_rate": 4.881159358732695e-06,
+      "loss": 0.6904,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6232443820224719,
+      "grad_norm": 0.9292178750038147,
+      "learning_rate": 4.88101846971752e-06,
+      "loss": 0.7045,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6235955056179775,
+      "grad_norm": 0.864090085029602,
+      "learning_rate": 4.880877499273551e-06,
+      "loss": 0.6496,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6239466292134831,
+      "grad_norm": 0.9168904423713684,
+      "learning_rate": 4.880736447405609e-06,
+      "loss": 0.6805,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6242977528089888,
+      "grad_norm": 0.9072659015655518,
+      "learning_rate": 4.880595314118518e-06,
+      "loss": 0.6699,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6246488764044944,
+      "grad_norm": 0.9015219807624817,
+      "learning_rate": 4.880454099417104e-06,
+      "loss": 0.6781,
+      "step": 1779
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.9149530529975891,
+      "learning_rate": 4.880312803306197e-06,
+      "loss": 0.7069,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6253511235955056,
+      "grad_norm": 0.9412938952445984,
+      "learning_rate": 4.880171425790629e-06,
+      "loss": 0.6851,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6257022471910112,
+      "grad_norm": 0.9359396696090698,
+      "learning_rate": 4.880029966875235e-06,
+      "loss": 0.6738,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6260533707865169,
+      "grad_norm": 0.9232779741287231,
+      "learning_rate": 4.879888426564853e-06,
+      "loss": 0.6688,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6264044943820225,
+      "grad_norm": 0.9085204005241394,
+      "learning_rate": 4.879746804864323e-06,
+      "loss": 0.6738,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6267556179775281,
+      "grad_norm": 0.916195809841156,
+      "learning_rate": 4.879605101778488e-06,
+      "loss": 0.6691,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6271067415730337,
+      "grad_norm": 0.891173243522644,
+      "learning_rate": 4.8794633173121954e-06,
+      "loss": 0.6941,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6274578651685393,
+      "grad_norm": 0.8935718536376953,
+      "learning_rate": 4.879321451470294e-06,
+      "loss": 0.65,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6278089887640449,
+      "grad_norm": 0.9326649308204651,
+      "learning_rate": 4.879179504257634e-06,
+      "loss": 0.7116,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6281601123595506,
+      "grad_norm": 0.9544984102249146,
+      "learning_rate": 4.879037475679071e-06,
+      "loss": 0.6531,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6285112359550562,
+      "grad_norm": 0.8751811981201172,
+      "learning_rate": 4.878895365739463e-06,
+      "loss": 0.7009,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6288623595505618,
+      "grad_norm": 0.860333263874054,
+      "learning_rate": 4.878753174443669e-06,
+      "loss": 0.7003,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 0.9002732038497925,
+      "learning_rate": 4.878610901796551e-06,
+      "loss": 0.6768,
+      "step": 1792
+    },
+    {
+      "epoch": 0.629564606741573,
+      "grad_norm": 0.8901023864746094,
+      "learning_rate": 4.8784685478029765e-06,
+      "loss": 0.6572,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6299157303370787,
+      "grad_norm": 0.8898550271987915,
+      "learning_rate": 4.878326112467813e-06,
+      "loss": 0.7141,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6302668539325843,
+      "grad_norm": 0.9085826277732849,
+      "learning_rate": 4.878183595795931e-06,
+      "loss": 0.7033,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6306179775280899,
+      "grad_norm": 0.9423730969429016,
+      "learning_rate": 4.878040997792205e-06,
+      "loss": 0.6763,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6309691011235955,
+      "grad_norm": 0.9046456217765808,
+      "learning_rate": 4.877898318461512e-06,
+      "loss": 0.6438,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6313202247191011,
+      "grad_norm": 0.9388726353645325,
+      "learning_rate": 4.877755557808731e-06,
+      "loss": 0.6921,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6316713483146067,
+      "grad_norm": 0.8939900398254395,
+      "learning_rate": 4.877612715838746e-06,
+      "loss": 0.6729,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6320224719101124,
+      "grad_norm": 0.919826328754425,
+      "learning_rate": 4.877469792556439e-06,
+      "loss": 0.7064,
+      "step": 1800
+    },
+    {
+      "epoch": 0.632373595505618,
+      "grad_norm": 0.8846567273139954,
+      "learning_rate": 4.8773267879667e-06,
+      "loss": 0.6463,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6327247191011236,
+      "grad_norm": 0.9120215177536011,
+      "learning_rate": 4.8771837020744195e-06,
+      "loss": 0.7055,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6330758426966292,
+      "grad_norm": 0.9156368970870972,
+      "learning_rate": 4.87704053488449e-06,
+      "loss": 0.7379,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6334269662921348,
+      "grad_norm": 0.9091489315032959,
+      "learning_rate": 4.876897286401808e-06,
+      "loss": 0.6618,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6337780898876404,
+      "grad_norm": 0.8682206273078918,
+      "learning_rate": 4.876753956631274e-06,
+      "loss": 0.6773,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6341292134831461,
+      "grad_norm": 0.9142380952835083,
+      "learning_rate": 4.876610545577788e-06,
+      "loss": 0.7102,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6344803370786517,
+      "grad_norm": 0.938529908657074,
+      "learning_rate": 4.876467053246255e-06,
+      "loss": 0.6967,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6348314606741573,
+      "grad_norm": 0.9012477397918701,
+      "learning_rate": 4.876323479641582e-06,
+      "loss": 0.6548,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6351825842696629,
+      "grad_norm": 0.8724120855331421,
+      "learning_rate": 4.8761798247686795e-06,
+      "loss": 0.6205,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6355337078651685,
+      "grad_norm": 0.9118915796279907,
+      "learning_rate": 4.876036088632461e-06,
+      "loss": 0.6931,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6358848314606742,
+      "grad_norm": 0.8679929375648499,
+      "learning_rate": 4.875892271237841e-06,
+      "loss": 0.6808,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6362359550561798,
+      "grad_norm": 0.8951993584632874,
+      "learning_rate": 4.875748372589738e-06,
+      "loss": 0.6665,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6365870786516854,
+      "grad_norm": 0.9559075832366943,
+      "learning_rate": 4.875604392693074e-06,
+      "loss": 0.6634,
+      "step": 1813
+    },
+    {
+      "epoch": 0.636938202247191,
+      "grad_norm": 0.8831387758255005,
+      "learning_rate": 4.875460331552772e-06,
+      "loss": 0.7054,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6372893258426966,
+      "grad_norm": 0.8540201783180237,
+      "learning_rate": 4.875316189173761e-06,
+      "loss": 0.6676,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6376404494382022,
+      "grad_norm": 0.8935166001319885,
+      "learning_rate": 4.875171965560967e-06,
+      "loss": 0.6789,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6379915730337079,
+      "grad_norm": 0.9121273159980774,
+      "learning_rate": 4.875027660719326e-06,
+      "loss": 0.6645,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6383426966292135,
+      "grad_norm": 0.9133923053741455,
+      "learning_rate": 4.87488327465377e-06,
+      "loss": 0.7053,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6386938202247191,
+      "grad_norm": 0.9285811185836792,
+      "learning_rate": 4.874738807369239e-06,
+      "loss": 0.7172,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6390449438202247,
+      "grad_norm": 0.9026995301246643,
+      "learning_rate": 4.874594258870673e-06,
+      "loss": 0.6848,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6393960674157303,
+      "grad_norm": 0.9294858574867249,
+      "learning_rate": 4.874449629163015e-06,
+      "loss": 0.7073,
+      "step": 1821
+    },
+    {
+      "epoch": 0.639747191011236,
+      "grad_norm": 0.9029496908187866,
+      "learning_rate": 4.874304918251212e-06,
+      "loss": 0.7049,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6400983146067416,
+      "grad_norm": 0.8920079469680786,
+      "learning_rate": 4.874160126140211e-06,
+      "loss": 0.6649,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6404494382022472,
+      "grad_norm": 0.8667442798614502,
+      "learning_rate": 4.874015252834966e-06,
+      "loss": 0.6327,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6408005617977528,
+      "grad_norm": 0.8910001516342163,
+      "learning_rate": 4.8738702983404315e-06,
+      "loss": 0.6964,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6411516853932584,
+      "grad_norm": 0.8773640394210815,
+      "learning_rate": 4.873725262661564e-06,
+      "loss": 0.6763,
+      "step": 1826
+    },
+    {
+      "epoch": 0.641502808988764,
+      "grad_norm": 0.9394330978393555,
+      "learning_rate": 4.873580145803323e-06,
+      "loss": 0.7796,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6418539325842697,
+      "grad_norm": 0.8596976399421692,
+      "learning_rate": 4.8734349477706725e-06,
+      "loss": 0.6229,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6422050561797753,
+      "grad_norm": 0.9220896363258362,
+      "learning_rate": 4.873289668568578e-06,
+      "loss": 0.6801,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6425561797752809,
+      "grad_norm": 0.8582680821418762,
+      "learning_rate": 4.873144308202008e-06,
+      "loss": 0.6475,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6429073033707865,
+      "grad_norm": 0.916015088558197,
+      "learning_rate": 4.872998866675933e-06,
+      "loss": 0.6296,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6432584269662921,
+      "grad_norm": 0.8732319474220276,
+      "learning_rate": 4.872853343995327e-06,
+      "loss": 0.6294,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6436095505617978,
+      "grad_norm": 0.8985139727592468,
+      "learning_rate": 4.872707740165168e-06,
+      "loss": 0.6647,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6439606741573034,
+      "grad_norm": 0.9100089073181152,
+      "learning_rate": 4.872562055190433e-06,
+      "loss": 0.66,
+      "step": 1834
+    },
+    {
+      "epoch": 0.644311797752809,
+      "grad_norm": 0.8807663917541504,
+      "learning_rate": 4.872416289076107e-06,
+      "loss": 0.714,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6446629213483146,
+      "grad_norm": 0.924670934677124,
+      "learning_rate": 4.872270441827174e-06,
+      "loss": 0.6851,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6450140449438202,
+      "grad_norm": 0.9182237982749939,
+      "learning_rate": 4.872124513448622e-06,
+      "loss": 0.6728,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6453651685393258,
+      "grad_norm": 0.8587349057197571,
+      "learning_rate": 4.871978503945442e-06,
+      "loss": 0.6632,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6457162921348315,
+      "grad_norm": 0.8839679956436157,
+      "learning_rate": 4.871832413322626e-06,
+      "loss": 0.6531,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6460674157303371,
+      "grad_norm": 0.9729541540145874,
+      "learning_rate": 4.871686241585171e-06,
+      "loss": 0.6855,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6464185393258427,
+      "grad_norm": 0.895285964012146,
+      "learning_rate": 4.871539988738076e-06,
+      "loss": 0.6738,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6467696629213483,
+      "grad_norm": 0.8678845167160034,
+      "learning_rate": 4.871393654786344e-06,
+      "loss": 0.6494,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6471207865168539,
+      "grad_norm": 0.8896480798721313,
+      "learning_rate": 4.871247239734977e-06,
+      "loss": 0.6823,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6474719101123596,
+      "grad_norm": 0.8339593410491943,
+      "learning_rate": 4.871100743588983e-06,
+      "loss": 0.6193,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6478230337078652,
+      "grad_norm": 0.8970949053764343,
+      "learning_rate": 4.870954166353374e-06,
+      "loss": 0.6944,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6481741573033708,
+      "grad_norm": 0.8810001015663147,
+      "learning_rate": 4.870807508033161e-06,
+      "loss": 0.7222,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6485252808988764,
+      "grad_norm": 0.8821990489959717,
+      "learning_rate": 4.870660768633359e-06,
+      "loss": 0.6881,
+      "step": 1847
+    },
+    {
+      "epoch": 0.648876404494382,
+      "grad_norm": 0.9234027862548828,
+      "learning_rate": 4.8705139481589885e-06,
+      "loss": 0.7074,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6492275280898876,
+      "grad_norm": 0.9003174901008606,
+      "learning_rate": 4.870367046615069e-06,
+      "loss": 0.6754,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6495786516853933,
+      "grad_norm": 0.9838910698890686,
+      "learning_rate": 4.870220064006625e-06,
+      "loss": 0.716,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6499297752808989,
+      "grad_norm": 0.9417749047279358,
+      "learning_rate": 4.870073000338683e-06,
+      "loss": 0.6826,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6502808988764045,
+      "grad_norm": 0.9156447052955627,
+      "learning_rate": 4.869925855616272e-06,
+      "loss": 0.6472,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6506320224719101,
+      "grad_norm": 0.8980277180671692,
+      "learning_rate": 4.869778629844425e-06,
+      "loss": 0.6536,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6509831460674157,
+      "grad_norm": 0.9398206472396851,
+      "learning_rate": 4.8696313230281776e-06,
+      "loss": 0.6618,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6513342696629213,
+      "grad_norm": 0.9095361232757568,
+      "learning_rate": 4.869483935172566e-06,
+      "loss": 0.6758,
+      "step": 1855
+    },
+    {
+      "epoch": 0.651685393258427,
+      "grad_norm": 0.9065535664558411,
+      "learning_rate": 4.86933646628263e-06,
+      "loss": 0.6785,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6520365168539326,
+      "grad_norm": 0.8876499533653259,
+      "learning_rate": 4.869188916363417e-06,
+      "loss": 0.642,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6523876404494382,
+      "grad_norm": 0.8867465853691101,
+      "learning_rate": 4.869041285419968e-06,
+      "loss": 0.6926,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6527387640449438,
+      "grad_norm": 0.9016279578208923,
+      "learning_rate": 4.868893573457335e-06,
+      "loss": 0.7007,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6530898876404494,
+      "grad_norm": 0.9355380535125732,
+      "learning_rate": 4.868745780480569e-06,
+      "loss": 0.7025,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6534410112359551,
+      "grad_norm": 0.8738915324211121,
+      "learning_rate": 4.868597906494725e-06,
+      "loss": 0.6633,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6537921348314607,
+      "grad_norm": 0.9167248010635376,
+      "learning_rate": 4.868449951504859e-06,
+      "loss": 0.6669,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6541432584269663,
+      "grad_norm": 0.9220420122146606,
+      "learning_rate": 4.868301915516031e-06,
+      "loss": 0.6419,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6544943820224719,
+      "grad_norm": 0.9482148885726929,
+      "learning_rate": 4.868153798533304e-06,
+      "loss": 0.6934,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6548455056179775,
+      "grad_norm": 0.928871214389801,
+      "learning_rate": 4.8680056005617436e-06,
+      "loss": 0.6506,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6551966292134831,
+      "grad_norm": 0.9240955114364624,
+      "learning_rate": 4.867857321606418e-06,
+      "loss": 0.6745,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6555477528089888,
+      "grad_norm": 0.8898888230323792,
+      "learning_rate": 4.867708961672399e-06,
+      "loss": 0.6925,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6558988764044944,
+      "grad_norm": 0.9390647411346436,
+      "learning_rate": 4.8675605207647595e-06,
+      "loss": 0.7212,
+      "step": 1868
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 0.9547630548477173,
+      "learning_rate": 4.8674119988885756e-06,
+      "loss": 0.7168,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6566011235955056,
+      "grad_norm": 0.8770663738250732,
+      "learning_rate": 4.867263396048927e-06,
+      "loss": 0.6373,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6569522471910112,
+      "grad_norm": 0.899808406829834,
+      "learning_rate": 4.867114712250897e-06,
+      "loss": 0.708,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6573033707865169,
+      "grad_norm": 0.9099377989768982,
+      "learning_rate": 4.866965947499568e-06,
+      "loss": 0.6888,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6576544943820225,
+      "grad_norm": 0.896460235118866,
+      "learning_rate": 4.866817101800029e-06,
+      "loss": 0.6995,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6580056179775281,
+      "grad_norm": 0.8833721280097961,
+      "learning_rate": 4.866668175157372e-06,
+      "loss": 0.659,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6583567415730337,
+      "grad_norm": 0.830697774887085,
+      "learning_rate": 4.866519167576687e-06,
+      "loss": 0.6501,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6587078651685393,
+      "grad_norm": 0.8897443413734436,
+      "learning_rate": 4.866370079063073e-06,
+      "loss": 0.6769,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6590589887640449,
+      "grad_norm": 0.8772610425949097,
+      "learning_rate": 4.866220909621625e-06,
+      "loss": 0.6474,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6594101123595506,
+      "grad_norm": 0.9042930006980896,
+      "learning_rate": 4.8660716592574486e-06,
+      "loss": 0.6827,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6597612359550562,
+      "grad_norm": 0.9102811217308044,
+      "learning_rate": 4.865922327975646e-06,
+      "loss": 0.7126,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6601123595505618,
+      "grad_norm": 0.9031997323036194,
+      "learning_rate": 4.865772915781324e-06,
+      "loss": 0.6963,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6604634831460674,
+      "grad_norm": 0.89446622133255,
+      "learning_rate": 4.865623422679593e-06,
+      "loss": 0.6718,
+      "step": 1881
+    },
+    {
+      "epoch": 0.660814606741573,
+      "grad_norm": 0.879215657711029,
+      "learning_rate": 4.865473848675565e-06,
+      "loss": 0.6711,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6611657303370787,
+      "grad_norm": 0.916566789150238,
+      "learning_rate": 4.865324193774355e-06,
+      "loss": 0.66,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6615168539325843,
+      "grad_norm": 0.8872631192207336,
+      "learning_rate": 4.865174457981082e-06,
+      "loss": 0.6994,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6618679775280899,
+      "grad_norm": 0.9287483096122742,
+      "learning_rate": 4.865024641300866e-06,
+      "loss": 0.7071,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6622191011235955,
+      "grad_norm": 0.8984839916229248,
+      "learning_rate": 4.864874743738832e-06,
+      "loss": 0.6512,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6625702247191011,
+      "grad_norm": 0.8624014258384705,
+      "learning_rate": 4.864724765300105e-06,
+      "loss": 0.6649,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6629213483146067,
+      "grad_norm": 0.8853806257247925,
+      "learning_rate": 4.864574705989814e-06,
+      "loss": 0.6562,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6632724719101124,
+      "grad_norm": 0.9202688336372375,
+      "learning_rate": 4.864424565813092e-06,
+      "loss": 0.7005,
+      "step": 1889
+    },
+    {
+      "epoch": 0.663623595505618,
+      "grad_norm": 0.9426562786102295,
+      "learning_rate": 4.8642743447750735e-06,
+      "loss": 0.6802,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6639747191011236,
+      "grad_norm": 0.9027799367904663,
+      "learning_rate": 4.864124042880895e-06,
+      "loss": 0.6718,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6643258426966292,
+      "grad_norm": 0.9138725399971008,
+      "learning_rate": 4.863973660135697e-06,
+      "loss": 0.7225,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6646769662921348,
+      "grad_norm": 0.9003073573112488,
+      "learning_rate": 4.863823196544623e-06,
+      "loss": 0.6603,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6650280898876404,
+      "grad_norm": 0.9345358610153198,
+      "learning_rate": 4.863672652112819e-06,
+      "loss": 0.6648,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6653792134831461,
+      "grad_norm": 0.9087908864021301,
+      "learning_rate": 4.863522026845432e-06,
+      "loss": 0.701,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6657303370786517,
+      "grad_norm": 0.889224112033844,
+      "learning_rate": 4.863371320747615e-06,
+      "loss": 0.7157,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6660814606741573,
+      "grad_norm": 0.8854907751083374,
+      "learning_rate": 4.863220533824521e-06,
+      "loss": 0.6965,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6664325842696629,
+      "grad_norm": 0.9223704934120178,
+      "learning_rate": 4.863069666081307e-06,
+      "loss": 0.6962,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6667837078651685,
+      "grad_norm": 0.9045940637588501,
+      "learning_rate": 4.8629187175231314e-06,
+      "loss": 0.6394,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6671348314606742,
+      "grad_norm": 0.8898254036903381,
+      "learning_rate": 4.862767688155159e-06,
+      "loss": 0.6781,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6674859550561798,
+      "grad_norm": 0.882418692111969,
+      "learning_rate": 4.862616577982553e-06,
+      "loss": 0.6654,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6678370786516854,
+      "grad_norm": 0.8898456692695618,
+      "learning_rate": 4.862465387010482e-06,
+      "loss": 0.6733,
+      "step": 1902
+    },
+    {
+      "epoch": 0.668188202247191,
+      "grad_norm": 0.9240460395812988,
+      "learning_rate": 4.862314115244116e-06,
+      "loss": 0.6661,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6685393258426966,
+      "grad_norm": 0.8646409511566162,
+      "learning_rate": 4.862162762688629e-06,
+      "loss": 0.6751,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6688904494382022,
+      "grad_norm": 0.9161674976348877,
+      "learning_rate": 4.862011329349196e-06,
+      "loss": 0.6762,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6692415730337079,
+      "grad_norm": 0.8989123702049255,
+      "learning_rate": 4.861859815230998e-06,
+      "loss": 0.6948,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6695926966292135,
+      "grad_norm": 0.9007300138473511,
+      "learning_rate": 4.861708220339215e-06,
+      "loss": 0.6954,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6699438202247191,
+      "grad_norm": 0.9009419679641724,
+      "learning_rate": 4.861556544679032e-06,
+      "loss": 0.6968,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6702949438202247,
+      "grad_norm": 0.8684524297714233,
+      "learning_rate": 4.861404788255636e-06,
+      "loss": 0.6548,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6706460674157303,
+      "grad_norm": 0.8866965174674988,
+      "learning_rate": 4.861252951074216e-06,
+      "loss": 0.6867,
+      "step": 1910
+    },
+    {
+      "epoch": 0.670997191011236,
+      "grad_norm": 0.8514602184295654,
+      "learning_rate": 4.861101033139967e-06,
+      "loss": 0.6677,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6713483146067416,
+      "grad_norm": 0.9115993976593018,
+      "learning_rate": 4.8609490344580815e-06,
+      "loss": 0.7163,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6716994382022472,
+      "grad_norm": 0.9031485319137573,
+      "learning_rate": 4.86079695503376e-06,
+      "loss": 0.6557,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6720505617977528,
+      "grad_norm": 0.9129601716995239,
+      "learning_rate": 4.8606447948722036e-06,
+      "loss": 0.6857,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6724016853932584,
+      "grad_norm": 0.8930435180664062,
+      "learning_rate": 4.860492553978614e-06,
+      "loss": 0.6888,
+      "step": 1915
+    },
+    {
+      "epoch": 0.672752808988764,
+      "grad_norm": 0.8881438374519348,
+      "learning_rate": 4.860340232358199e-06,
+      "loss": 0.6979,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6731039325842697,
+      "grad_norm": 0.8837933540344238,
+      "learning_rate": 4.860187830016169e-06,
+      "loss": 0.6729,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6734550561797753,
+      "grad_norm": 0.8822264671325684,
+      "learning_rate": 4.860035346957734e-06,
+      "loss": 0.6883,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6738061797752809,
+      "grad_norm": 0.8592543005943298,
+      "learning_rate": 4.8598827831881105e-06,
+      "loss": 0.6579,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.8520956039428711,
+      "learning_rate": 4.859730138712514e-06,
+      "loss": 0.6547,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6745084269662921,
+      "grad_norm": 0.8868786096572876,
+      "learning_rate": 4.859577413536167e-06,
+      "loss": 0.6834,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6748595505617978,
+      "grad_norm": 0.894353449344635,
+      "learning_rate": 4.859424607664291e-06,
+      "loss": 0.6622,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6752106741573034,
+      "grad_norm": 0.8527143597602844,
+      "learning_rate": 4.859271721102112e-06,
+      "loss": 0.6557,
+      "step": 1923
+    },
+    {
+      "epoch": 0.675561797752809,
+      "grad_norm": 0.840531587600708,
+      "learning_rate": 4.85911875385486e-06,
+      "loss": 0.6392,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6759129213483146,
+      "grad_norm": 0.8790932297706604,
+      "learning_rate": 4.858965705927765e-06,
+      "loss": 0.7031,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6762640449438202,
+      "grad_norm": 0.8507084250450134,
+      "learning_rate": 4.85881257732606e-06,
+      "loss": 0.6459,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6766151685393258,
+      "grad_norm": 0.894951581954956,
+      "learning_rate": 4.858659368054986e-06,
+      "loss": 0.6885,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6769662921348315,
+      "grad_norm": 0.9011364579200745,
+      "learning_rate": 4.858506078119778e-06,
+      "loss": 0.7135,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6773174157303371,
+      "grad_norm": 0.8751960396766663,
+      "learning_rate": 4.858352707525681e-06,
+      "loss": 0.6964,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6776685393258427,
+      "grad_norm": 0.8583717942237854,
+      "learning_rate": 4.85819925627794e-06,
+      "loss": 0.6495,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6780196629213483,
+      "grad_norm": 0.9073567390441895,
+      "learning_rate": 4.858045724381801e-06,
+      "loss": 0.6693,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6783707865168539,
+      "grad_norm": 0.875085711479187,
+      "learning_rate": 4.857892111842517e-06,
+      "loss": 0.6552,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6787219101123596,
+      "grad_norm": 0.890121340751648,
+      "learning_rate": 4.8577384186653405e-06,
+      "loss": 0.6816,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6790730337078652,
+      "grad_norm": 0.9079762697219849,
+      "learning_rate": 4.8575846448555275e-06,
+      "loss": 0.7012,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6794241573033708,
+      "grad_norm": 0.9067320823669434,
+      "learning_rate": 4.857430790418337e-06,
+      "loss": 0.6707,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6797752808988764,
+      "grad_norm": 0.863487184047699,
+      "learning_rate": 4.857276855359031e-06,
+      "loss": 0.6532,
+      "step": 1936
+    },
+    {
+      "epoch": 0.680126404494382,
+      "grad_norm": 0.887187659740448,
+      "learning_rate": 4.857122839682873e-06,
+      "loss": 0.68,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6804775280898876,
+      "grad_norm": 0.9105770587921143,
+      "learning_rate": 4.856968743395132e-06,
+      "loss": 0.7263,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6808286516853933,
+      "grad_norm": 0.859266459941864,
+      "learning_rate": 4.856814566501075e-06,
+      "loss": 0.6612,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6811797752808989,
+      "grad_norm": 0.8715479969978333,
+      "learning_rate": 4.856660309005979e-06,
+      "loss": 0.7078,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6815308988764045,
+      "grad_norm": 0.8754439949989319,
+      "learning_rate": 4.856505970915115e-06,
+      "loss": 0.7088,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6818820224719101,
+      "grad_norm": 0.8877387642860413,
+      "learning_rate": 4.856351552233764e-06,
+      "loss": 0.6944,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6822331460674157,
+      "grad_norm": 0.9069793224334717,
+      "learning_rate": 4.856197052967207e-06,
+      "loss": 0.6567,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6825842696629213,
+      "grad_norm": 0.9112107157707214,
+      "learning_rate": 4.856042473120726e-06,
+      "loss": 0.6617,
+      "step": 1944
+    },
+    {
+      "epoch": 0.682935393258427,
+      "grad_norm": 0.9331689476966858,
+      "learning_rate": 4.855887812699609e-06,
+      "loss": 0.7051,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6832865168539326,
+      "grad_norm": 0.909070611000061,
+      "learning_rate": 4.855733071709145e-06,
+      "loss": 0.6827,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6836376404494382,
+      "grad_norm": 0.9068970084190369,
+      "learning_rate": 4.855578250154624e-06,
+      "loss": 0.658,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6839887640449438,
+      "grad_norm": 0.9052385687828064,
+      "learning_rate": 4.855423348041345e-06,
+      "loss": 0.6991,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6843398876404494,
+      "grad_norm": 0.8857588768005371,
+      "learning_rate": 4.855268365374601e-06,
+      "loss": 0.6604,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6846910112359551,
+      "grad_norm": 0.8484088778495789,
+      "learning_rate": 4.855113302159697e-06,
+      "loss": 0.6591,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6850421348314607,
+      "grad_norm": 0.9859545826911926,
+      "learning_rate": 4.854958158401931e-06,
+      "loss": 0.6652,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6853932584269663,
+      "grad_norm": 0.8704702854156494,
+      "learning_rate": 4.854802934106611e-06,
+      "loss": 0.6792,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6857443820224719,
+      "grad_norm": 0.8603319525718689,
+      "learning_rate": 4.854647629279047e-06,
+      "loss": 0.6768,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6860955056179775,
+      "grad_norm": 0.9137224555015564,
+      "learning_rate": 4.854492243924549e-06,
+      "loss": 0.6865,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6864466292134831,
+      "grad_norm": 0.8924287557601929,
+      "learning_rate": 4.8543367780484304e-06,
+      "loss": 0.685,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6867977528089888,
+      "grad_norm": 0.8736245632171631,
+      "learning_rate": 4.854181231656009e-06,
+      "loss": 0.6747,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6871488764044944,
+      "grad_norm": 0.8849700093269348,
+      "learning_rate": 4.854025604752604e-06,
+      "loss": 0.679,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.9219119548797607,
+      "learning_rate": 4.853869897343537e-06,
+      "loss": 0.7069,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6878511235955056,
+      "grad_norm": 0.8864405751228333,
+      "learning_rate": 4.853714109434134e-06,
+      "loss": 0.648,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6882022471910112,
+      "grad_norm": 0.9157871603965759,
+      "learning_rate": 4.853558241029723e-06,
+      "loss": 0.6838,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6885533707865169,
+      "grad_norm": 0.8845330476760864,
+      "learning_rate": 4.853402292135634e-06,
+      "loss": 0.67,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6889044943820225,
+      "grad_norm": 0.8602401614189148,
+      "learning_rate": 4.8532462627572e-06,
+      "loss": 0.6661,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6892556179775281,
+      "grad_norm": 0.9007517695426941,
+      "learning_rate": 4.853090152899758e-06,
+      "loss": 0.6717,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6896067415730337,
+      "grad_norm": 0.8941755890846252,
+      "learning_rate": 4.852933962568647e-06,
+      "loss": 0.6824,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6899578651685393,
+      "grad_norm": 0.9077978134155273,
+      "learning_rate": 4.852777691769207e-06,
+      "loss": 0.695,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6903089887640449,
+      "grad_norm": 0.9053329229354858,
+      "learning_rate": 4.852621340506782e-06,
+      "loss": 0.6781,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6906601123595506,
+      "grad_norm": 0.885201096534729,
+      "learning_rate": 4.852464908786722e-06,
+      "loss": 0.6307,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6910112359550562,
+      "grad_norm": 0.9176447987556458,
+      "learning_rate": 4.8523083966143755e-06,
+      "loss": 0.6797,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6913623595505618,
+      "grad_norm": 0.8907782435417175,
+      "learning_rate": 4.852151803995094e-06,
+      "loss": 0.661,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6917134831460674,
+      "grad_norm": 0.9267505407333374,
+      "learning_rate": 4.8519951309342335e-06,
+      "loss": 0.6711,
+      "step": 1970
+    },
+    {
+      "epoch": 0.692064606741573,
+      "grad_norm": 0.9260626435279846,
+      "learning_rate": 4.8518383774371525e-06,
+      "loss": 0.6667,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6924157303370787,
+      "grad_norm": 0.947289228439331,
+      "learning_rate": 4.851681543509211e-06,
+      "loss": 0.6802,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6927668539325843,
+      "grad_norm": 0.9016532897949219,
+      "learning_rate": 4.851524629155773e-06,
+      "loss": 0.6788,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6931179775280899,
+      "grad_norm": 0.926960825920105,
+      "learning_rate": 4.851367634382206e-06,
+      "loss": 0.6868,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6934691011235955,
+      "grad_norm": 0.886313259601593,
+      "learning_rate": 4.851210559193877e-06,
+      "loss": 0.664,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6938202247191011,
+      "grad_norm": 0.8867488503456116,
+      "learning_rate": 4.851053403596158e-06,
+      "loss": 0.6903,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6941713483146067,
+      "grad_norm": 0.8759608864784241,
+      "learning_rate": 4.850896167594426e-06,
+      "loss": 0.6542,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6945224719101124,
+      "grad_norm": 0.8738287091255188,
+      "learning_rate": 4.8507388511940565e-06,
+      "loss": 0.6844,
+      "step": 1978
+    },
+    {
+      "epoch": 0.694873595505618,
+      "grad_norm": 0.875167191028595,
+      "learning_rate": 4.85058145440043e-06,
+      "loss": 0.6515,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6952247191011236,
+      "grad_norm": 0.8686906099319458,
+      "learning_rate": 4.850423977218929e-06,
+      "loss": 0.6558,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6955758426966292,
+      "grad_norm": 0.8697443604469299,
+      "learning_rate": 4.850266419654939e-06,
+      "loss": 0.6431,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6959269662921348,
+      "grad_norm": 0.9007163047790527,
+      "learning_rate": 4.850108781713848e-06,
+      "loss": 0.6996,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6962780898876404,
+      "grad_norm": 0.9064307808876038,
+      "learning_rate": 4.849951063401049e-06,
+      "loss": 0.6625,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6966292134831461,
+      "grad_norm": 0.9589636921882629,
+      "learning_rate": 4.849793264721934e-06,
+      "loss": 0.7276,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6969803370786517,
+      "grad_norm": 0.9150680899620056,
+      "learning_rate": 4.8496353856819e-06,
+      "loss": 0.6808,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6973314606741573,
+      "grad_norm": 0.9379653334617615,
+      "learning_rate": 4.849477426286346e-06,
+      "loss": 0.7148,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6976825842696629,
+      "grad_norm": 0.8566982746124268,
+      "learning_rate": 4.849319386540675e-06,
+      "loss": 0.676,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6980337078651685,
+      "grad_norm": 0.9016757607460022,
+      "learning_rate": 4.84916126645029e-06,
+      "loss": 0.6672,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6983848314606742,
+      "grad_norm": 0.9286409020423889,
+      "learning_rate": 4.849003066020602e-06,
+      "loss": 0.7028,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6987359550561798,
+      "grad_norm": 0.8700134754180908,
+      "learning_rate": 4.848844785257019e-06,
+      "loss": 0.6919,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6990870786516854,
+      "grad_norm": 0.8518657088279724,
+      "learning_rate": 4.848686424164953e-06,
+      "loss": 0.663,
+      "step": 1991
+    },
+    {
+      "epoch": 0.699438202247191,
+      "grad_norm": 0.8883944749832153,
+      "learning_rate": 4.848527982749822e-06,
+      "loss": 0.6249,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6997893258426966,
+      "grad_norm": 0.9609950184822083,
+      "learning_rate": 4.848369461017043e-06,
+      "loss": 0.7114,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7001404494382022,
+      "grad_norm": 0.8920437097549438,
+      "learning_rate": 4.848210858972039e-06,
+      "loss": 0.7093,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7004915730337079,
+      "grad_norm": 0.9256541132926941,
+      "learning_rate": 4.848052176620232e-06,
+      "loss": 0.6713,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7008426966292135,
+      "grad_norm": 0.8984493613243103,
+      "learning_rate": 4.847893413967049e-06,
+      "loss": 0.6921,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7011938202247191,
+      "grad_norm": 0.889815628528595,
+      "learning_rate": 4.847734571017923e-06,
+      "loss": 0.6539,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7015449438202247,
+      "grad_norm": 0.856476366519928,
+      "learning_rate": 4.847575647778281e-06,
+      "loss": 0.6538,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7018960674157303,
+      "grad_norm": 0.9042733311653137,
+      "learning_rate": 4.847416644253562e-06,
+      "loss": 0.672,
+      "step": 1999
+    },
+    {
+      "epoch": 0.702247191011236,
+      "grad_norm": 0.847193717956543,
+      "learning_rate": 4.847257560449202e-06,
+      "loss": 0.6296,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7025983146067416,
+      "grad_norm": 0.922281801700592,
+      "learning_rate": 4.847098396370643e-06,
+      "loss": 0.6306,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7029494382022472,
+      "grad_norm": 0.9229936599731445,
+      "learning_rate": 4.846939152023326e-06,
+      "loss": 0.7127,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7033005617977528,
+      "grad_norm": 0.8789034485816956,
+      "learning_rate": 4.8467798274126995e-06,
+      "loss": 0.6477,
+      "step": 2003
+    },
+    {
+      "epoch": 0.7036516853932584,
+      "grad_norm": 0.87799471616745,
+      "learning_rate": 4.84662042254421e-06,
+      "loss": 0.646,
+      "step": 2004
+    },
+    {
+      "epoch": 0.704002808988764,
+      "grad_norm": 0.8978516459465027,
+      "learning_rate": 4.846460937423311e-06,
+      "loss": 0.6327,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7043539325842697,
+      "grad_norm": 0.9151999354362488,
+      "learning_rate": 4.846301372055455e-06,
+      "loss": 0.6726,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7047050561797753,
+      "grad_norm": 0.9256817698478699,
+      "learning_rate": 4.846141726446101e-06,
+      "loss": 0.6842,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7050561797752809,
+      "grad_norm": 0.89638352394104,
+      "learning_rate": 4.845982000600706e-06,
+      "loss": 0.6792,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7054073033707865,
+      "grad_norm": 0.8957139849662781,
+      "learning_rate": 4.845822194524735e-06,
+      "loss": 0.7032,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7057584269662921,
+      "grad_norm": 0.9059911370277405,
+      "learning_rate": 4.845662308223652e-06,
+      "loss": 0.7272,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7061095505617978,
+      "grad_norm": 0.8813321590423584,
+      "learning_rate": 4.845502341702926e-06,
+      "loss": 0.6918,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7064606741573034,
+      "grad_norm": 0.9412849545478821,
+      "learning_rate": 4.845342294968026e-06,
+      "loss": 0.7301,
+      "step": 2012
+    },
+    {
+      "epoch": 0.706811797752809,
+      "grad_norm": 0.9104598164558411,
+      "learning_rate": 4.845182168024427e-06,
+      "loss": 0.6655,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7071629213483146,
+      "grad_norm": 0.8984184861183167,
+      "learning_rate": 4.845021960877605e-06,
+      "loss": 0.6816,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7075140449438202,
+      "grad_norm": 0.9031736850738525,
+      "learning_rate": 4.844861673533038e-06,
+      "loss": 0.6941,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7078651685393258,
+      "grad_norm": 0.9205414652824402,
+      "learning_rate": 4.8447013059962085e-06,
+      "loss": 0.6976,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7082162921348315,
+      "grad_norm": 0.9202782511711121,
+      "learning_rate": 4.8445408582726e-06,
+      "loss": 0.6859,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7085674157303371,
+      "grad_norm": 0.9281151294708252,
+      "learning_rate": 4.844380330367701e-06,
+      "loss": 0.6827,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7089185393258427,
+      "grad_norm": 0.8505613207817078,
+      "learning_rate": 4.844219722287001e-06,
+      "loss": 0.6198,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7092696629213483,
+      "grad_norm": 0.8852872252464294,
+      "learning_rate": 4.844059034035993e-06,
+      "loss": 0.7026,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7096207865168539,
+      "grad_norm": 0.8684709072113037,
+      "learning_rate": 4.843898265620171e-06,
+      "loss": 0.6602,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7099719101123596,
+      "grad_norm": 0.9983372688293457,
+      "learning_rate": 4.843737417045035e-06,
+      "loss": 0.6646,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7103230337078652,
+      "grad_norm": 0.8986789584159851,
+      "learning_rate": 4.843576488316084e-06,
+      "loss": 0.7246,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7106741573033708,
+      "grad_norm": 0.898418664932251,
+      "learning_rate": 4.843415479438822e-06,
+      "loss": 0.7011,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7110252808988764,
+      "grad_norm": 0.9410746097564697,
+      "learning_rate": 4.843254390418757e-06,
+      "loss": 0.6753,
+      "step": 2025
+    },
+    {
+      "epoch": 0.711376404494382,
+      "grad_norm": 0.9196873307228088,
+      "learning_rate": 4.843093221261396e-06,
+      "loss": 0.7028,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7117275280898876,
+      "grad_norm": 0.8739240169525146,
+      "learning_rate": 4.842931971972252e-06,
+      "loss": 0.6477,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7120786516853933,
+      "grad_norm": 0.863125205039978,
+      "learning_rate": 4.842770642556841e-06,
+      "loss": 0.6855,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7124297752808989,
+      "grad_norm": 0.8746199607849121,
+      "learning_rate": 4.842609233020677e-06,
+      "loss": 0.6368,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7127808988764045,
+      "grad_norm": 0.9002125859260559,
+      "learning_rate": 4.842447743369283e-06,
+      "loss": 0.666,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7131320224719101,
+      "grad_norm": 0.8976905345916748,
+      "learning_rate": 4.84228617360818e-06,
+      "loss": 0.6807,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7134831460674157,
+      "grad_norm": 0.9237125515937805,
+      "learning_rate": 4.8421245237428935e-06,
+      "loss": 0.6749,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7138342696629213,
+      "grad_norm": 0.8990357518196106,
+      "learning_rate": 4.841962793778954e-06,
+      "loss": 0.7136,
+      "step": 2033
+    },
+    {
+      "epoch": 0.714185393258427,
+      "grad_norm": 0.9369171857833862,
+      "learning_rate": 4.84180098372189e-06,
+      "loss": 0.7157,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7145365168539326,
+      "grad_norm": 0.9113399386405945,
+      "learning_rate": 4.841639093577236e-06,
+      "loss": 0.6733,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7148876404494382,
+      "grad_norm": 0.9038696885108948,
+      "learning_rate": 4.84147712335053e-06,
+      "loss": 0.7065,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7152387640449438,
+      "grad_norm": 0.8818309903144836,
+      "learning_rate": 4.841315073047309e-06,
+      "loss": 0.6633,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7155898876404494,
+      "grad_norm": 0.8951486349105835,
+      "learning_rate": 4.841152942673116e-06,
+      "loss": 0.6965,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7159410112359551,
+      "grad_norm": 0.8958115577697754,
+      "learning_rate": 4.840990732233496e-06,
+      "loss": 0.6536,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7162921348314607,
+      "grad_norm": 0.8991239070892334,
+      "learning_rate": 4.840828441733996e-06,
+      "loss": 0.7002,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7166432584269663,
+      "grad_norm": 0.8847957849502563,
+      "learning_rate": 4.840666071180167e-06,
+      "loss": 0.7011,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7169943820224719,
+      "grad_norm": 0.8697375655174255,
+      "learning_rate": 4.840503620577561e-06,
+      "loss": 0.6869,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7173455056179775,
+      "grad_norm": 0.8789458274841309,
+      "learning_rate": 4.840341089931734e-06,
+      "loss": 0.6624,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7176966292134831,
+      "grad_norm": 0.868335485458374,
+      "learning_rate": 4.8401784792482436e-06,
+      "loss": 0.6624,
+      "step": 2044
+    },
+    {
+      "epoch": 0.7180477528089888,
+      "grad_norm": 0.8949484825134277,
+      "learning_rate": 4.840015788532653e-06,
+      "loss": 0.6621,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7183988764044944,
+      "grad_norm": 0.8744524717330933,
+      "learning_rate": 4.839853017790523e-06,
+      "loss": 0.6055,
+      "step": 2046
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 0.8852532505989075,
+      "learning_rate": 4.8396901670274236e-06,
+      "loss": 0.7299,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7191011235955056,
+      "grad_norm": 0.8787648677825928,
+      "learning_rate": 4.839527236248922e-06,
+      "loss": 0.6687,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7194522471910112,
+      "grad_norm": 0.8882533311843872,
+      "learning_rate": 4.839364225460591e-06,
+      "loss": 0.6488,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7198033707865169,
+      "grad_norm": 0.8676881194114685,
+      "learning_rate": 4.839201134668006e-06,
+      "loss": 0.7084,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7201544943820225,
+      "grad_norm": 0.9025247097015381,
+      "learning_rate": 4.839037963876743e-06,
+      "loss": 0.6805,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7205056179775281,
+      "grad_norm": 0.8956056833267212,
+      "learning_rate": 4.838874713092383e-06,
+      "loss": 0.6905,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7208567415730337,
+      "grad_norm": 0.8418490290641785,
+      "learning_rate": 4.8387113823205104e-06,
+      "loss": 0.6224,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7212078651685393,
+      "grad_norm": 0.897898256778717,
+      "learning_rate": 4.838547971566708e-06,
+      "loss": 0.6505,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7215589887640449,
+      "grad_norm": 0.9430592060089111,
+      "learning_rate": 4.838384480836568e-06,
+      "loss": 0.7031,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7219101123595506,
+      "grad_norm": 0.9099498987197876,
+      "learning_rate": 4.838220910135678e-06,
+      "loss": 0.7372,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7222612359550562,
+      "grad_norm": 0.8936906456947327,
+      "learning_rate": 4.838057259469635e-06,
+      "loss": 0.6782,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7226123595505618,
+      "grad_norm": 0.8931390643119812,
+      "learning_rate": 4.8378935288440346e-06,
+      "loss": 0.6552,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7229634831460674,
+      "grad_norm": 0.893640398979187,
+      "learning_rate": 4.837729718264475e-06,
+      "loss": 0.6971,
+      "step": 2059
+    },
+    {
+      "epoch": 0.723314606741573,
+      "grad_norm": 0.9098536968231201,
+      "learning_rate": 4.83756582773656e-06,
+      "loss": 0.6816,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7236657303370787,
+      "grad_norm": 0.8403578400611877,
+      "learning_rate": 4.837401857265894e-06,
+      "loss": 0.6233,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7240168539325843,
+      "grad_norm": 0.894605815410614,
+      "learning_rate": 4.8372378068580835e-06,
+      "loss": 0.6747,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7243679775280899,
+      "grad_norm": 0.8893696665763855,
+      "learning_rate": 4.837073676518742e-06,
+      "loss": 0.6769,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7247191011235955,
+      "grad_norm": 0.9210422039031982,
+      "learning_rate": 4.836909466253479e-06,
+      "loss": 0.7177,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7250702247191011,
+      "grad_norm": 0.9131370782852173,
+      "learning_rate": 4.836745176067913e-06,
+      "loss": 0.6899,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7254213483146067,
+      "grad_norm": 0.8986336588859558,
+      "learning_rate": 4.836580805967661e-06,
+      "loss": 0.68,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7257724719101124,
+      "grad_norm": 0.8927890658378601,
+      "learning_rate": 4.836416355958345e-06,
+      "loss": 0.6971,
+      "step": 2067
+    },
+    {
+      "epoch": 0.726123595505618,
+      "grad_norm": 0.9166001677513123,
+      "learning_rate": 4.836251826045589e-06,
+      "loss": 0.7008,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7264747191011236,
+      "grad_norm": 0.8931081891059875,
+      "learning_rate": 4.836087216235021e-06,
+      "loss": 0.6816,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7268258426966292,
+      "grad_norm": 0.8896619081497192,
+      "learning_rate": 4.835922526532267e-06,
+      "loss": 0.7011,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7271769662921348,
+      "grad_norm": 0.8900657296180725,
+      "learning_rate": 4.835757756942963e-06,
+      "loss": 0.6754,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7275280898876404,
+      "grad_norm": 0.860923707485199,
+      "learning_rate": 4.8355929074727415e-06,
+      "loss": 0.6407,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7278792134831461,
+      "grad_norm": 0.8783923983573914,
+      "learning_rate": 4.835427978127241e-06,
+      "loss": 0.7052,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7282303370786517,
+      "grad_norm": 0.9070163369178772,
+      "learning_rate": 4.835262968912103e-06,
+      "loss": 0.7003,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7285814606741573,
+      "grad_norm": 1.025547742843628,
+      "learning_rate": 4.83509787983297e-06,
+      "loss": 0.6829,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7289325842696629,
+      "grad_norm": 0.9013466835021973,
+      "learning_rate": 4.834932710895486e-06,
+      "loss": 0.6475,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7292837078651685,
+      "grad_norm": 0.894381582736969,
+      "learning_rate": 4.834767462105303e-06,
+      "loss": 0.6414,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7296348314606742,
+      "grad_norm": 0.8694182634353638,
+      "learning_rate": 4.83460213346807e-06,
+      "loss": 0.6442,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7299859550561798,
+      "grad_norm": 0.8485740423202515,
+      "learning_rate": 4.834436724989442e-06,
+      "loss": 0.6105,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7303370786516854,
+      "grad_norm": 0.8988609910011292,
+      "learning_rate": 4.834271236675076e-06,
+      "loss": 0.6615,
+      "step": 2080
+    },
+    {
+      "epoch": 0.730688202247191,
+      "grad_norm": 0.9217643141746521,
+      "learning_rate": 4.8341056685306295e-06,
+      "loss": 0.7194,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7310393258426966,
+      "grad_norm": 0.8651581406593323,
+      "learning_rate": 4.833940020561768e-06,
+      "loss": 0.6831,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7313904494382022,
+      "grad_norm": 0.9073237776756287,
+      "learning_rate": 4.833774292774154e-06,
+      "loss": 0.7034,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7317415730337079,
+      "grad_norm": 0.9331315159797668,
+      "learning_rate": 4.833608485173458e-06,
+      "loss": 0.7328,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7320926966292135,
+      "grad_norm": 0.9014981985092163,
+      "learning_rate": 4.833442597765346e-06,
+      "loss": 0.6776,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7324438202247191,
+      "grad_norm": 0.8719879388809204,
+      "learning_rate": 4.833276630555496e-06,
+      "loss": 0.6672,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7327949438202247,
+      "grad_norm": 0.8750829100608826,
+      "learning_rate": 4.8331105835495816e-06,
+      "loss": 0.6344,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7331460674157303,
+      "grad_norm": 0.8622833490371704,
+      "learning_rate": 4.832944456753281e-06,
+      "loss": 0.6453,
+      "step": 2088
+    },
+    {
+      "epoch": 0.733497191011236,
+      "grad_norm": 0.8901452422142029,
+      "learning_rate": 4.8327782501722766e-06,
+      "loss": 0.6809,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7338483146067416,
+      "grad_norm": 0.9039089679718018,
+      "learning_rate": 4.8326119638122525e-06,
+      "loss": 0.7185,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7341994382022472,
+      "grad_norm": 0.9443939328193665,
+      "learning_rate": 4.832445597678895e-06,
+      "loss": 0.6647,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7345505617977528,
+      "grad_norm": 0.9074206352233887,
+      "learning_rate": 4.8322791517778945e-06,
+      "loss": 0.6696,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7349016853932584,
+      "grad_norm": 0.9349908232688904,
+      "learning_rate": 4.832112626114942e-06,
+      "loss": 0.6735,
+      "step": 2093
+    },
+    {
+      "epoch": 0.735252808988764,
+      "grad_norm": 0.8699564933776855,
+      "learning_rate": 4.831946020695733e-06,
+      "loss": 0.7143,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7356039325842697,
+      "grad_norm": 0.8945705890655518,
+      "learning_rate": 4.831779335525967e-06,
+      "loss": 0.6748,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7359550561797753,
+      "grad_norm": 0.8274276256561279,
+      "learning_rate": 4.831612570611343e-06,
+      "loss": 0.647,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7363061797752809,
+      "grad_norm": 0.8812181949615479,
+      "learning_rate": 4.831445725957563e-06,
+      "loss": 0.6644,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7366573033707865,
+      "grad_norm": 0.8417778015136719,
+      "learning_rate": 4.831278801570335e-06,
+      "loss": 0.613,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7370084269662921,
+      "grad_norm": 0.862417459487915,
+      "learning_rate": 4.831111797455366e-06,
+      "loss": 0.6592,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7373595505617978,
+      "grad_norm": 0.8994525074958801,
+      "learning_rate": 4.830944713618369e-06,
+      "loss": 0.7038,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7377106741573034,
+      "grad_norm": 0.8670651316642761,
+      "learning_rate": 4.830777550065058e-06,
+      "loss": 0.6344,
+      "step": 2101
+    },
+    {
+      "epoch": 0.738061797752809,
+      "grad_norm": 0.9172879457473755,
+      "learning_rate": 4.830610306801148e-06,
+      "loss": 0.657,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7384129213483146,
+      "grad_norm": 0.8903508186340332,
+      "learning_rate": 4.830442983832361e-06,
+      "loss": 0.6419,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7387640449438202,
+      "grad_norm": 0.913711428642273,
+      "learning_rate": 4.830275581164417e-06,
+      "loss": 0.6551,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7391151685393258,
+      "grad_norm": 0.8964325189590454,
+      "learning_rate": 4.830108098803043e-06,
+      "loss": 0.6839,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7394662921348315,
+      "grad_norm": 0.8451694846153259,
+      "learning_rate": 4.829940536753965e-06,
+      "loss": 0.6846,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7398174157303371,
+      "grad_norm": 0.8537585735321045,
+      "learning_rate": 4.829772895022916e-06,
+      "loss": 0.6632,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7401685393258427,
+      "grad_norm": 0.882422924041748,
+      "learning_rate": 4.829605173615627e-06,
+      "loss": 0.6827,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7405196629213483,
+      "grad_norm": 0.9608381986618042,
+      "learning_rate": 4.829437372537834e-06,
+      "loss": 0.6698,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7408707865168539,
+      "grad_norm": 0.8602509498596191,
+      "learning_rate": 4.829269491795277e-06,
+      "loss": 0.6842,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7412219101123596,
+      "grad_norm": 0.9101216197013855,
+      "learning_rate": 4.829101531393695e-06,
+      "loss": 0.6909,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7415730337078652,
+      "grad_norm": 0.9232078194618225,
+      "learning_rate": 4.8289334913388344e-06,
+      "loss": 0.6544,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7419241573033708,
+      "grad_norm": 0.9027913212776184,
+      "learning_rate": 4.828765371636442e-06,
+      "loss": 0.7055,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7422752808988764,
+      "grad_norm": 0.8847595453262329,
+      "learning_rate": 4.828597172292266e-06,
+      "loss": 0.6585,
+      "step": 2114
+    },
+    {
+      "epoch": 0.742626404494382,
+      "grad_norm": 0.9057716131210327,
+      "learning_rate": 4.8284288933120595e-06,
+      "loss": 0.6815,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7429775280898876,
+      "grad_norm": 0.833390474319458,
+      "learning_rate": 4.828260534701577e-06,
+      "loss": 0.6113,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7433286516853933,
+      "grad_norm": 0.9086408615112305,
+      "learning_rate": 4.828092096466578e-06,
+      "loss": 0.6537,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7436797752808989,
+      "grad_norm": 0.9064264297485352,
+      "learning_rate": 4.82792357861282e-06,
+      "loss": 0.71,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7440308988764045,
+      "grad_norm": 0.9102289080619812,
+      "learning_rate": 4.827754981146068e-06,
+      "loss": 0.6711,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7443820224719101,
+      "grad_norm": 0.9276719093322754,
+      "learning_rate": 4.827586304072087e-06,
+      "loss": 0.6667,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7447331460674157,
+      "grad_norm": 0.9276329278945923,
+      "learning_rate": 4.827417547396647e-06,
+      "loss": 0.7072,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7450842696629213,
+      "grad_norm": 0.8715528249740601,
+      "learning_rate": 4.827248711125518e-06,
+      "loss": 0.6921,
+      "step": 2122
+    },
+    {
+      "epoch": 0.745435393258427,
+      "grad_norm": 0.8835285902023315,
+      "learning_rate": 4.827079795264475e-06,
+      "loss": 0.6557,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7457865168539326,
+      "grad_norm": 0.9126173853874207,
+      "learning_rate": 4.826910799819293e-06,
+      "loss": 0.6965,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7461376404494382,
+      "grad_norm": 0.8943221569061279,
+      "learning_rate": 4.8267417247957545e-06,
+      "loss": 0.6532,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7464887640449438,
+      "grad_norm": 0.8731197714805603,
+      "learning_rate": 4.826572570199639e-06,
+      "loss": 0.6462,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7468398876404494,
+      "grad_norm": 0.8800104856491089,
+      "learning_rate": 4.826403336036733e-06,
+      "loss": 0.6862,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7471910112359551,
+      "grad_norm": 0.8647268414497375,
+      "learning_rate": 4.826234022312823e-06,
+      "loss": 0.6901,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7475421348314607,
+      "grad_norm": 0.8699238896369934,
+      "learning_rate": 4.826064629033701e-06,
+      "loss": 0.6848,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7478932584269663,
+      "grad_norm": 0.881797730922699,
+      "learning_rate": 4.825895156205158e-06,
+      "loss": 0.7022,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7482443820224719,
+      "grad_norm": 0.8925601840019226,
+      "learning_rate": 4.825725603832992e-06,
+      "loss": 0.6668,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7485955056179775,
+      "grad_norm": 0.874530553817749,
+      "learning_rate": 4.825555971923e-06,
+      "loss": 0.6914,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7489466292134831,
+      "grad_norm": 0.8640453815460205,
+      "learning_rate": 4.8253862604809836e-06,
+      "loss": 0.6301,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7492977528089888,
+      "grad_norm": 0.9012243151664734,
+      "learning_rate": 4.8252164695127475e-06,
+      "loss": 0.6876,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7496488764044944,
+      "grad_norm": 0.8837961554527283,
+      "learning_rate": 4.825046599024098e-06,
+      "loss": 0.6898,
+      "step": 2135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8629834651947021,
+      "learning_rate": 4.8248766490208434e-06,
+      "loss": 0.6709,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7503511235955056,
+      "grad_norm": 0.8934418559074402,
+      "learning_rate": 4.824706619508797e-06,
+      "loss": 0.6406,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7507022471910112,
+      "grad_norm": 0.8818619251251221,
+      "learning_rate": 4.8245365104937734e-06,
+      "loss": 0.6751,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7510533707865169,
+      "grad_norm": 0.871660590171814,
+      "learning_rate": 4.824366321981591e-06,
+      "loss": 0.6628,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7514044943820225,
+      "grad_norm": 0.8862018585205078,
+      "learning_rate": 4.824196053978069e-06,
+      "loss": 0.655,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7517556179775281,
+      "grad_norm": 0.8964344263076782,
+      "learning_rate": 4.824025706489031e-06,
+      "loss": 0.6784,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7521067415730337,
+      "grad_norm": 0.8510351181030273,
+      "learning_rate": 4.8238552795203025e-06,
+      "loss": 0.6521,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7524578651685393,
+      "grad_norm": 0.8770909905433655,
+      "learning_rate": 4.823684773077712e-06,
+      "loss": 0.6828,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7528089887640449,
+      "grad_norm": 0.9084598422050476,
+      "learning_rate": 4.823514187167091e-06,
+      "loss": 0.7081,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7531601123595506,
+      "grad_norm": 0.8897272944450378,
+      "learning_rate": 4.823343521794273e-06,
+      "loss": 0.6812,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7535112359550562,
+      "grad_norm": 0.8769894242286682,
+      "learning_rate": 4.8231727769650934e-06,
+      "loss": 0.69,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7538623595505618,
+      "grad_norm": 0.8911151885986328,
+      "learning_rate": 4.823001952685394e-06,
+      "loss": 0.6573,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7542134831460674,
+      "grad_norm": 0.8679448366165161,
+      "learning_rate": 4.822831048961015e-06,
+      "loss": 0.7047,
+      "step": 2148
+    },
+    {
+      "epoch": 0.754564606741573,
+      "grad_norm": 0.8955689072608948,
+      "learning_rate": 4.8226600657978024e-06,
+      "loss": 0.6775,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7549157303370787,
+      "grad_norm": 0.8726634383201599,
+      "learning_rate": 4.822489003201602e-06,
+      "loss": 0.6744,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7552668539325843,
+      "grad_norm": 0.8858023881912231,
+      "learning_rate": 4.822317861178266e-06,
+      "loss": 0.6975,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7556179775280899,
+      "grad_norm": 0.8668063879013062,
+      "learning_rate": 4.822146639733647e-06,
+      "loss": 0.6866,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7559691011235955,
+      "grad_norm": 0.8951655626296997,
+      "learning_rate": 4.821975338873599e-06,
+      "loss": 0.7165,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7563202247191011,
+      "grad_norm": 0.9189617037773132,
+      "learning_rate": 4.821803958603982e-06,
+      "loss": 0.6899,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7566713483146067,
+      "grad_norm": 0.8782833814620972,
+      "learning_rate": 4.821632498930656e-06,
+      "loss": 0.6613,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7570224719101124,
+      "grad_norm": 0.882354736328125,
+      "learning_rate": 4.821460959859485e-06,
+      "loss": 0.7047,
+      "step": 2156
+    },
+    {
+      "epoch": 0.757373595505618,
+      "grad_norm": 0.8961043357849121,
+      "learning_rate": 4.821289341396336e-06,
+      "loss": 0.6699,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7577247191011236,
+      "grad_norm": 0.9026042819023132,
+      "learning_rate": 4.8211176435470774e-06,
+      "loss": 0.6874,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7580758426966292,
+      "grad_norm": 0.89781653881073,
+      "learning_rate": 4.820945866317582e-06,
+      "loss": 0.6764,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7584269662921348,
+      "grad_norm": 0.9029053449630737,
+      "learning_rate": 4.820774009713725e-06,
+      "loss": 0.671,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7587780898876404,
+      "grad_norm": 0.8790138959884644,
+      "learning_rate": 4.820602073741381e-06,
+      "loss": 0.6855,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7591292134831461,
+      "grad_norm": 0.8737415671348572,
+      "learning_rate": 4.820430058406433e-06,
+      "loss": 0.6746,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7594803370786517,
+      "grad_norm": 0.8937375545501709,
+      "learning_rate": 4.8202579637147615e-06,
+      "loss": 0.6742,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7598314606741573,
+      "grad_norm": 0.8693381547927856,
+      "learning_rate": 4.820085789672254e-06,
+      "loss": 0.643,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7601825842696629,
+      "grad_norm": 0.8404312133789062,
+      "learning_rate": 4.819913536284798e-06,
+      "loss": 0.6247,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7605337078651685,
+      "grad_norm": 0.9133322238922119,
+      "learning_rate": 4.819741203558284e-06,
+      "loss": 0.6999,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7608848314606742,
+      "grad_norm": 0.878828763961792,
+      "learning_rate": 4.8195687914986055e-06,
+      "loss": 0.642,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7612359550561798,
+      "grad_norm": 0.8653201460838318,
+      "learning_rate": 4.81939630011166e-06,
+      "loss": 0.6668,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7615870786516854,
+      "grad_norm": 0.8635656833648682,
+      "learning_rate": 4.819223729403345e-06,
+      "loss": 0.6749,
+      "step": 2169
+    },
+    {
+      "epoch": 0.761938202247191,
+      "grad_norm": 0.9288333654403687,
+      "learning_rate": 4.819051079379563e-06,
+      "loss": 0.6687,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7622893258426966,
+      "grad_norm": 0.8941252827644348,
+      "learning_rate": 4.818878350046219e-06,
+      "loss": 0.6795,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7626404494382022,
+      "grad_norm": 0.8719800114631653,
+      "learning_rate": 4.81870554140922e-06,
+      "loss": 0.6273,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7629915730337079,
+      "grad_norm": 0.9125568866729736,
+      "learning_rate": 4.818532653474475e-06,
+      "loss": 0.6992,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7633426966292135,
+      "grad_norm": 0.9224469661712646,
+      "learning_rate": 4.818359686247898e-06,
+      "loss": 0.6984,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7636938202247191,
+      "grad_norm": 0.9007505178451538,
+      "learning_rate": 4.818186639735403e-06,
+      "loss": 0.6994,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7640449438202247,
+      "grad_norm": 0.9869738817214966,
+      "learning_rate": 4.818013513942908e-06,
+      "loss": 0.7248,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7643960674157303,
+      "grad_norm": 0.9010109305381775,
+      "learning_rate": 4.817840308876335e-06,
+      "loss": 0.6987,
+      "step": 2177
+    },
+    {
+      "epoch": 0.764747191011236,
+      "grad_norm": 0.8850472569465637,
+      "learning_rate": 4.8176670245416076e-06,
+      "loss": 0.6561,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7650983146067416,
+      "grad_norm": 0.9040166139602661,
+      "learning_rate": 4.81749366094465e-06,
+      "loss": 0.6789,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7654494382022472,
+      "grad_norm": 0.879569411277771,
+      "learning_rate": 4.817320218091393e-06,
+      "loss": 0.6728,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7658005617977528,
+      "grad_norm": 0.8991327285766602,
+      "learning_rate": 4.817146695987768e-06,
+      "loss": 0.6919,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7661516853932584,
+      "grad_norm": 0.8922879695892334,
+      "learning_rate": 4.816973094639708e-06,
+      "loss": 0.6696,
+      "step": 2182
+    },
+    {
+      "epoch": 0.766502808988764,
+      "grad_norm": 0.9196280837059021,
+      "learning_rate": 4.8167994140531504e-06,
+      "loss": 0.6928,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7668539325842697,
+      "grad_norm": 0.8855655193328857,
+      "learning_rate": 4.816625654234036e-06,
+      "loss": 0.67,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7672050561797753,
+      "grad_norm": 0.9374614953994751,
+      "learning_rate": 4.816451815188306e-06,
+      "loss": 0.6514,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7675561797752809,
+      "grad_norm": 0.8807061910629272,
+      "learning_rate": 4.816277896921907e-06,
+      "loss": 0.666,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7679073033707865,
+      "grad_norm": 0.8804175853729248,
+      "learning_rate": 4.816103899440785e-06,
+      "loss": 0.6687,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7682584269662921,
+      "grad_norm": 0.8722162842750549,
+      "learning_rate": 4.815929822750893e-06,
+      "loss": 0.6924,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7686095505617978,
+      "grad_norm": 0.8696702718734741,
+      "learning_rate": 4.815755666858182e-06,
+      "loss": 0.6883,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7689606741573034,
+      "grad_norm": 0.8709911704063416,
+      "learning_rate": 4.815581431768609e-06,
+      "loss": 0.6533,
+      "step": 2190
+    },
+    {
+      "epoch": 0.769311797752809,
+      "grad_norm": 0.9215819835662842,
+      "learning_rate": 4.815407117488132e-06,
+      "loss": 0.6519,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7696629213483146,
+      "grad_norm": 0.9619512557983398,
+      "learning_rate": 4.815232724022713e-06,
+      "loss": 0.6523,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7700140449438202,
+      "grad_norm": 0.8896926045417786,
+      "learning_rate": 4.815058251378316e-06,
+      "loss": 0.6772,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7703651685393258,
+      "grad_norm": 0.9150664806365967,
+      "learning_rate": 4.814883699560908e-06,
+      "loss": 0.7005,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7707162921348315,
+      "grad_norm": 0.8808259963989258,
+      "learning_rate": 4.8147090685764585e-06,
+      "loss": 0.683,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7710674157303371,
+      "grad_norm": 0.8727423548698425,
+      "learning_rate": 4.8145343584309404e-06,
+      "loss": 0.6499,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7714185393258427,
+      "grad_norm": 0.8704618811607361,
+      "learning_rate": 4.814359569130326e-06,
+      "loss": 0.6678,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7717696629213483,
+      "grad_norm": 0.8853801488876343,
+      "learning_rate": 4.814184700680596e-06,
+      "loss": 0.6788,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7721207865168539,
+      "grad_norm": 0.8947190046310425,
+      "learning_rate": 4.81400975308773e-06,
+      "loss": 0.6919,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7724719101123596,
+      "grad_norm": 0.8671098351478577,
+      "learning_rate": 4.813834726357709e-06,
+      "loss": 0.6344,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7728230337078652,
+      "grad_norm": 0.8669817447662354,
+      "learning_rate": 4.813659620496521e-06,
+      "loss": 0.6655,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7731741573033708,
+      "grad_norm": 0.8825884461402893,
+      "learning_rate": 4.813484435510154e-06,
+      "loss": 0.687,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7735252808988764,
+      "grad_norm": 0.9186983704566956,
+      "learning_rate": 4.8133091714045985e-06,
+      "loss": 0.6807,
+      "step": 2203
+    },
+    {
+      "epoch": 0.773876404494382,
+      "grad_norm": 0.8848013877868652,
+      "learning_rate": 4.81313382818585e-06,
+      "loss": 0.6974,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7742275280898876,
+      "grad_norm": 0.8753053545951843,
+      "learning_rate": 4.812958405859903e-06,
+      "loss": 0.7025,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7745786516853933,
+      "grad_norm": 0.8476365208625793,
+      "learning_rate": 4.812782904432758e-06,
+      "loss": 0.6466,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7749297752808989,
+      "grad_norm": 0.8795397281646729,
+      "learning_rate": 4.812607323910416e-06,
+      "loss": 0.6669,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7752808988764045,
+      "grad_norm": 0.8774227499961853,
+      "learning_rate": 4.812431664298884e-06,
+      "loss": 0.6488,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7756320224719101,
+      "grad_norm": 0.888466477394104,
+      "learning_rate": 4.812255925604167e-06,
+      "loss": 0.6367,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7759831460674157,
+      "grad_norm": 0.8345751166343689,
+      "learning_rate": 4.812080107832275e-06,
+      "loss": 0.6204,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7763342696629213,
+      "grad_norm": 0.8765496611595154,
+      "learning_rate": 4.811904210989223e-06,
+      "loss": 0.697,
+      "step": 2211
+    },
+    {
+      "epoch": 0.776685393258427,
+      "grad_norm": 0.893101692199707,
+      "learning_rate": 4.8117282350810255e-06,
+      "loss": 0.6673,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7770365168539326,
+      "grad_norm": 0.909160852432251,
+      "learning_rate": 4.8115521801137e-06,
+      "loss": 0.6973,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7773876404494382,
+      "grad_norm": 0.8834896087646484,
+      "learning_rate": 4.811376046093268e-06,
+      "loss": 0.6878,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7777387640449438,
+      "grad_norm": 0.8608538508415222,
+      "learning_rate": 4.8111998330257535e-06,
+      "loss": 0.6688,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7780898876404494,
+      "grad_norm": 0.9701086282730103,
+      "learning_rate": 4.811023540917181e-06,
+      "loss": 0.6721,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7784410112359551,
+      "grad_norm": 0.8717615008354187,
+      "learning_rate": 4.8108471697735826e-06,
+      "loss": 0.6834,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7787921348314607,
+      "grad_norm": 0.9188772439956665,
+      "learning_rate": 4.810670719600987e-06,
+      "loss": 0.7135,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7791432584269663,
+      "grad_norm": 0.8877284526824951,
+      "learning_rate": 4.810494190405432e-06,
+      "loss": 0.7184,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7794943820224719,
+      "grad_norm": 0.8861910104751587,
+      "learning_rate": 4.8103175821929515e-06,
+      "loss": 0.6545,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7798455056179775,
+      "grad_norm": 0.9048117995262146,
+      "learning_rate": 4.810140894969587e-06,
+      "loss": 0.6995,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7801966292134831,
+      "grad_norm": 0.8994266390800476,
+      "learning_rate": 4.809964128741381e-06,
+      "loss": 0.6837,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7805477528089888,
+      "grad_norm": 0.9110033512115479,
+      "learning_rate": 4.8097872835143775e-06,
+      "loss": 0.6575,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7808988764044944,
+      "grad_norm": 0.8434987664222717,
+      "learning_rate": 4.8096103592946265e-06,
+      "loss": 0.6535,
+      "step": 2224
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.91326504945755,
+      "learning_rate": 4.8094333560881765e-06,
+      "loss": 0.6714,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7816011235955056,
+      "grad_norm": 0.8931390643119812,
+      "learning_rate": 4.809256273901083e-06,
+      "loss": 0.667,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7819522471910112,
+      "grad_norm": 0.8923385143280029,
+      "learning_rate": 4.8090791127393995e-06,
+      "loss": 0.6792,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7823033707865169,
+      "grad_norm": 0.8582352995872498,
+      "learning_rate": 4.808901872609188e-06,
+      "loss": 0.674,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7826544943820225,
+      "grad_norm": 1.004380702972412,
+      "learning_rate": 4.808724553516507e-06,
+      "loss": 0.6736,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7830056179775281,
+      "grad_norm": 0.9153397083282471,
+      "learning_rate": 4.808547155467422e-06,
+      "loss": 0.6842,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7833567415730337,
+      "grad_norm": 0.8752025961875916,
+      "learning_rate": 4.808369678468001e-06,
+      "loss": 0.702,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7837078651685393,
+      "grad_norm": 0.9121509790420532,
+      "learning_rate": 4.808192122524311e-06,
+      "loss": 0.6794,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7840589887640449,
+      "grad_norm": 0.893732488155365,
+      "learning_rate": 4.808014487642426e-06,
+      "loss": 0.6783,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7844101123595506,
+      "grad_norm": 0.8836367726325989,
+      "learning_rate": 4.8078367738284205e-06,
+      "loss": 0.7095,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7847612359550562,
+      "grad_norm": 0.8630282878875732,
+      "learning_rate": 4.8076589810883725e-06,
+      "loss": 0.6262,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7851123595505618,
+      "grad_norm": 0.8813517093658447,
+      "learning_rate": 4.807481109428362e-06,
+      "loss": 0.6728,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7854634831460674,
+      "grad_norm": 0.8723793625831604,
+      "learning_rate": 4.80730315885447e-06,
+      "loss": 0.6554,
+      "step": 2237
+    },
+    {
+      "epoch": 0.785814606741573,
+      "grad_norm": 0.8989778757095337,
+      "learning_rate": 4.807125129372787e-06,
+      "loss": 0.6587,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7861657303370787,
+      "grad_norm": 0.8957408666610718,
+      "learning_rate": 4.8069470209893976e-06,
+      "loss": 0.6604,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7865168539325843,
+      "grad_norm": 0.9145181775093079,
+      "learning_rate": 4.806768833710395e-06,
+      "loss": 0.6794,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7868679775280899,
+      "grad_norm": 0.8940600752830505,
+      "learning_rate": 4.806590567541871e-06,
+      "loss": 0.686,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7872191011235955,
+      "grad_norm": 0.8759947419166565,
+      "learning_rate": 4.806412222489925e-06,
+      "loss": 0.7002,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7875702247191011,
+      "grad_norm": 0.8928653001785278,
+      "learning_rate": 4.8062337985606535e-06,
+      "loss": 0.6343,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7879213483146067,
+      "grad_norm": 0.9009339809417725,
+      "learning_rate": 4.806055295760159e-06,
+      "loss": 0.7096,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7882724719101124,
+      "grad_norm": 0.8576456308364868,
+      "learning_rate": 4.8058767140945475e-06,
+      "loss": 0.6449,
+      "step": 2245
+    },
+    {
+      "epoch": 0.788623595505618,
+      "grad_norm": 0.8837461471557617,
+      "learning_rate": 4.805698053569926e-06,
+      "loss": 0.6565,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7889747191011236,
+      "grad_norm": 1.0404770374298096,
+      "learning_rate": 4.805519314192403e-06,
+      "loss": 0.6726,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7893258426966292,
+      "grad_norm": 0.9084009528160095,
+      "learning_rate": 4.805340495968093e-06,
+      "loss": 0.6518,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7896769662921348,
+      "grad_norm": 0.9170441031455994,
+      "learning_rate": 4.805161598903111e-06,
+      "loss": 0.6733,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7900280898876404,
+      "grad_norm": 0.9394940733909607,
+      "learning_rate": 4.804982623003574e-06,
+      "loss": 0.6898,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7903792134831461,
+      "grad_norm": 0.9983392953872681,
+      "learning_rate": 4.8048035682756046e-06,
+      "loss": 0.6739,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7907303370786517,
+      "grad_norm": 0.9523327946662903,
+      "learning_rate": 4.8046244347253245e-06,
+      "loss": 0.6999,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7910814606741573,
+      "grad_norm": 0.8619253635406494,
+      "learning_rate": 4.804445222358862e-06,
+      "loss": 0.6699,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7914325842696629,
+      "grad_norm": 0.9346986413002014,
+      "learning_rate": 4.804265931182343e-06,
+      "loss": 0.6971,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7917837078651685,
+      "grad_norm": 0.912071943283081,
+      "learning_rate": 4.804086561201903e-06,
+      "loss": 0.6913,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7921348314606742,
+      "grad_norm": 0.9198614358901978,
+      "learning_rate": 4.8039071124236734e-06,
+      "loss": 0.6787,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7924859550561798,
+      "grad_norm": 0.8504485487937927,
+      "learning_rate": 4.803727584853792e-06,
+      "loss": 0.6574,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7928370786516854,
+      "grad_norm": 0.860133707523346,
+      "learning_rate": 4.8035479784983984e-06,
+      "loss": 0.7023,
+      "step": 2258
+    },
+    {
+      "epoch": 0.793188202247191,
+      "grad_norm": 0.8657998442649841,
+      "learning_rate": 4.803368293363635e-06,
+      "loss": 0.667,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7935393258426966,
+      "grad_norm": 0.8903993368148804,
+      "learning_rate": 4.803188529455647e-06,
+      "loss": 0.6504,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7938904494382022,
+      "grad_norm": 0.9583306908607483,
+      "learning_rate": 4.803008686780582e-06,
+      "loss": 0.6624,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7942415730337079,
+      "grad_norm": 0.9079137444496155,
+      "learning_rate": 4.802828765344592e-06,
+      "loss": 0.6845,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7945926966292135,
+      "grad_norm": 0.8480777740478516,
+      "learning_rate": 4.802648765153828e-06,
+      "loss": 0.6416,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7949438202247191,
+      "grad_norm": 0.9052068591117859,
+      "learning_rate": 4.802468686214446e-06,
+      "loss": 0.676,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7952949438202247,
+      "grad_norm": 0.8713710904121399,
+      "learning_rate": 4.802288528532606e-06,
+      "loss": 0.6823,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7956460674157303,
+      "grad_norm": 0.8888510465621948,
+      "learning_rate": 4.802108292114468e-06,
+      "loss": 0.6932,
+      "step": 2266
+    },
+    {
+      "epoch": 0.795997191011236,
+      "grad_norm": 0.9380448460578918,
+      "learning_rate": 4.8019279769661965e-06,
+      "loss": 0.7137,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7963483146067416,
+      "grad_norm": 0.8584917783737183,
+      "learning_rate": 4.801747583093958e-06,
+      "loss": 0.66,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7966994382022472,
+      "grad_norm": 0.8634847402572632,
+      "learning_rate": 4.8015671105039205e-06,
+      "loss": 0.6553,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7970505617977528,
+      "grad_norm": 1.1672545671463013,
+      "learning_rate": 4.801386559202259e-06,
+      "loss": 0.7002,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7974016853932584,
+      "grad_norm": 0.8866782188415527,
+      "learning_rate": 4.801205929195146e-06,
+      "loss": 0.6791,
+      "step": 2271
+    },
+    {
+      "epoch": 0.797752808988764,
+      "grad_norm": 0.92203289270401,
+      "learning_rate": 4.801025220488759e-06,
+      "loss": 0.6586,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7981039325842697,
+      "grad_norm": 0.9017732739448547,
+      "learning_rate": 4.8008444330892785e-06,
+      "loss": 0.6559,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7984550561797753,
+      "grad_norm": 0.9230211973190308,
+      "learning_rate": 4.800663567002888e-06,
+      "loss": 0.6968,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7988061797752809,
+      "grad_norm": 0.9016463160514832,
+      "learning_rate": 4.8004826222357715e-06,
+      "loss": 0.6624,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7991573033707865,
+      "grad_norm": 0.851717472076416,
+      "learning_rate": 4.8003015987941185e-06,
+      "loss": 0.6454,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7995084269662921,
+      "grad_norm": 0.8657079935073853,
+      "learning_rate": 4.800120496684119e-06,
+      "loss": 0.6537,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7998595505617978,
+      "grad_norm": 0.8756804466247559,
+      "learning_rate": 4.7999393159119665e-06,
+      "loss": 0.6777,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8002106741573034,
+      "grad_norm": 0.8858351111412048,
+      "learning_rate": 4.799758056483858e-06,
+      "loss": 0.673,
+      "step": 2279
+    },
+    {
+      "epoch": 0.800561797752809,
+      "grad_norm": 1.0514605045318604,
+      "learning_rate": 4.799576718405991e-06,
+      "loss": 0.6417,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8009129213483146,
+      "grad_norm": 0.8649700284004211,
+      "learning_rate": 4.799395301684569e-06,
+      "loss": 0.6753,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8012640449438202,
+      "grad_norm": 0.893130362033844,
+      "learning_rate": 4.799213806325794e-06,
+      "loss": 0.7144,
+      "step": 2282
+    },
+    {
+      "epoch": 0.8016151685393258,
+      "grad_norm": 0.8683095574378967,
+      "learning_rate": 4.799032232335875e-06,
+      "loss": 0.6565,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8019662921348315,
+      "grad_norm": 0.8900540471076965,
+      "learning_rate": 4.798850579721022e-06,
+      "loss": 0.7065,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8023174157303371,
+      "grad_norm": 0.9185386300086975,
+      "learning_rate": 4.7986688484874446e-06,
+      "loss": 0.6371,
+      "step": 2285
+    },
+    {
+      "epoch": 0.8026685393258427,
+      "grad_norm": 0.8507232666015625,
+      "learning_rate": 4.79848703864136e-06,
+      "loss": 0.66,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8030196629213483,
+      "grad_norm": 0.8661529421806335,
+      "learning_rate": 4.798305150188986e-06,
+      "loss": 0.6643,
+      "step": 2287
+    },
+    {
+      "epoch": 0.8033707865168539,
+      "grad_norm": 0.8859586119651794,
+      "learning_rate": 4.798123183136542e-06,
+      "loss": 0.652,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8037219101123596,
+      "grad_norm": 0.9179227948188782,
+      "learning_rate": 4.797941137490253e-06,
+      "loss": 0.7033,
+      "step": 2289
+    },
+    {
+      "epoch": 0.8040730337078652,
+      "grad_norm": 0.8715042471885681,
+      "learning_rate": 4.797759013256343e-06,
+      "loss": 0.6482,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8044241573033708,
+      "grad_norm": 0.8953105211257935,
+      "learning_rate": 4.797576810441041e-06,
+      "loss": 0.6986,
+      "step": 2291
+    },
+    {
+      "epoch": 0.8047752808988764,
+      "grad_norm": 0.9580156803131104,
+      "learning_rate": 4.797394529050577e-06,
+      "loss": 0.6512,
+      "step": 2292
+    },
+    {
+      "epoch": 0.805126404494382,
+      "grad_norm": 0.9238266348838806,
+      "learning_rate": 4.7972121690911865e-06,
+      "loss": 0.6796,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8054775280898876,
+      "grad_norm": 0.8915815949440002,
+      "learning_rate": 4.797029730569106e-06,
+      "loss": 0.707,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8058286516853933,
+      "grad_norm": 0.8978826999664307,
+      "learning_rate": 4.796847213490574e-06,
+      "loss": 0.7162,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8061797752808989,
+      "grad_norm": 0.8888537287712097,
+      "learning_rate": 4.796664617861832e-06,
+      "loss": 0.6832,
+      "step": 2296
+    },
+    {
+      "epoch": 0.8065308988764045,
+      "grad_norm": 0.8834527730941772,
+      "learning_rate": 4.796481943689127e-06,
+      "loss": 0.6863,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8068820224719101,
+      "grad_norm": 0.8643321394920349,
+      "learning_rate": 4.796299190978704e-06,
+      "loss": 0.6739,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8072331460674157,
+      "grad_norm": 0.8717115521430969,
+      "learning_rate": 4.796116359736813e-06,
+      "loss": 0.6502,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8075842696629213,
+      "grad_norm": 0.8730166554450989,
+      "learning_rate": 4.7959334499697085e-06,
+      "loss": 0.6604,
+      "step": 2300
+    },
+    {
+      "epoch": 0.807935393258427,
+      "grad_norm": 0.8617185950279236,
+      "learning_rate": 4.795750461683643e-06,
+      "loss": 0.667,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8082865168539326,
+      "grad_norm": 0.826866626739502,
+      "learning_rate": 4.795567394884878e-06,
+      "loss": 0.6273,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8086376404494382,
+      "grad_norm": 0.9160782098770142,
+      "learning_rate": 4.795384249579672e-06,
+      "loss": 0.6841,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8089887640449438,
+      "grad_norm": 0.8521285653114319,
+      "learning_rate": 4.795201025774287e-06,
+      "loss": 0.6311,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8093398876404494,
+      "grad_norm": 0.9138549566268921,
+      "learning_rate": 4.795017723474993e-06,
+      "loss": 0.6848,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8096910112359551,
+      "grad_norm": 0.8857465386390686,
+      "learning_rate": 4.794834342688056e-06,
+      "loss": 0.6457,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8100421348314607,
+      "grad_norm": 0.934290885925293,
+      "learning_rate": 4.7946508834197495e-06,
+      "loss": 0.6853,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8103932584269663,
+      "grad_norm": 0.8928509950637817,
+      "learning_rate": 4.794467345676345e-06,
+      "loss": 0.6884,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8107443820224719,
+      "grad_norm": 0.9139081835746765,
+      "learning_rate": 4.79428372946412e-06,
+      "loss": 0.7047,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8110955056179775,
+      "grad_norm": 0.8674109578132629,
+      "learning_rate": 4.794100034789356e-06,
+      "loss": 0.6397,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8114466292134831,
+      "grad_norm": 0.9009836912155151,
+      "learning_rate": 4.793916261658333e-06,
+      "loss": 0.6954,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8117977528089888,
+      "grad_norm": 0.8968256115913391,
+      "learning_rate": 4.7937324100773375e-06,
+      "loss": 0.6983,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8121488764044944,
+      "grad_norm": 0.9040331244468689,
+      "learning_rate": 4.793548480052656e-06,
+      "loss": 0.6932,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.9160118699073792,
+      "learning_rate": 4.7933644715905794e-06,
+      "loss": 0.702,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8128511235955056,
+      "grad_norm": 0.8957852721214294,
+      "learning_rate": 4.7931803846974005e-06,
+      "loss": 0.6737,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8132022471910112,
+      "grad_norm": 0.9111645817756653,
+      "learning_rate": 4.7929962193794145e-06,
+      "loss": 0.6716,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8135533707865169,
+      "grad_norm": 0.9393349885940552,
+      "learning_rate": 4.792811975642919e-06,
+      "loss": 0.7166,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8139044943820225,
+      "grad_norm": 0.8632532358169556,
+      "learning_rate": 4.792627653494217e-06,
+      "loss": 0.6545,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8142556179775281,
+      "grad_norm": 0.8554111123085022,
+      "learning_rate": 4.792443252939612e-06,
+      "loss": 0.6421,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8146067415730337,
+      "grad_norm": 0.8907461166381836,
+      "learning_rate": 4.792258773985408e-06,
+      "loss": 0.6874,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8149578651685393,
+      "grad_norm": 0.8739638924598694,
+      "learning_rate": 4.792074216637916e-06,
+      "loss": 0.6662,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8153089887640449,
+      "grad_norm": 0.8694685101509094,
+      "learning_rate": 4.791889580903446e-06,
+      "loss": 0.6536,
+      "step": 2322
+    },
+    {
+      "epoch": 0.8156601123595506,
+      "grad_norm": 0.9057086706161499,
+      "learning_rate": 4.791704866788314e-06,
+      "loss": 0.6655,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8160112359550562,
+      "grad_norm": 0.9522676467895508,
+      "learning_rate": 4.791520074298837e-06,
+      "loss": 0.7235,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8163623595505618,
+      "grad_norm": 0.8711917996406555,
+      "learning_rate": 4.791335203441335e-06,
+      "loss": 0.7015,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8167134831460674,
+      "grad_norm": 0.8643499612808228,
+      "learning_rate": 4.791150254222129e-06,
+      "loss": 0.6229,
+      "step": 2326
+    },
+    {
+      "epoch": 0.817064606741573,
+      "grad_norm": 0.8863927721977234,
+      "learning_rate": 4.790965226647545e-06,
+      "loss": 0.7017,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8174157303370787,
+      "grad_norm": 0.8969694375991821,
+      "learning_rate": 4.790780120723911e-06,
+      "loss": 0.6926,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8177668539325843,
+      "grad_norm": 0.923423171043396,
+      "learning_rate": 4.790594936457556e-06,
+      "loss": 0.707,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8181179775280899,
+      "grad_norm": 0.8579330444335938,
+      "learning_rate": 4.790409673854815e-06,
+      "loss": 0.6224,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8184691011235955,
+      "grad_norm": 0.8940487504005432,
+      "learning_rate": 4.790224332922023e-06,
+      "loss": 0.6859,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8188202247191011,
+      "grad_norm": 0.8670837879180908,
+      "learning_rate": 4.790038913665519e-06,
+      "loss": 0.6669,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8191713483146067,
+      "grad_norm": 0.8588036298751831,
+      "learning_rate": 4.789853416091643e-06,
+      "loss": 0.6951,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8195224719101124,
+      "grad_norm": 0.8730223774909973,
+      "learning_rate": 4.78966784020674e-06,
+      "loss": 0.6751,
+      "step": 2334
+    },
+    {
+      "epoch": 0.819873595505618,
+      "grad_norm": 0.8985008001327515,
+      "learning_rate": 4.789482186017157e-06,
+      "loss": 0.6927,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8202247191011236,
+      "grad_norm": 0.8898709416389465,
+      "learning_rate": 4.78929645352924e-06,
+      "loss": 0.6307,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8205758426966292,
+      "grad_norm": 0.8730369806289673,
+      "learning_rate": 4.789110642749346e-06,
+      "loss": 0.6548,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8209269662921348,
+      "grad_norm": 0.8780831694602966,
+      "learning_rate": 4.788924753683826e-06,
+      "loss": 0.6451,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8212780898876404,
+      "grad_norm": 0.8665900230407715,
+      "learning_rate": 4.788738786339037e-06,
+      "loss": 0.6656,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8216292134831461,
+      "grad_norm": 0.9199172854423523,
+      "learning_rate": 4.788552740721341e-06,
+      "loss": 0.6629,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8219803370786517,
+      "grad_norm": 0.8883637189865112,
+      "learning_rate": 4.7883666168370996e-06,
+      "loss": 0.6789,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8223314606741573,
+      "grad_norm": 0.8635849356651306,
+      "learning_rate": 4.788180414692678e-06,
+      "loss": 0.6823,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8226825842696629,
+      "grad_norm": 0.9055063128471375,
+      "learning_rate": 4.787994134294444e-06,
+      "loss": 0.6667,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8230337078651685,
+      "grad_norm": 0.8952557444572449,
+      "learning_rate": 4.787807775648768e-06,
+      "loss": 0.6694,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8233848314606742,
+      "grad_norm": 0.9171862006187439,
+      "learning_rate": 4.787621338762025e-06,
+      "loss": 0.6938,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8237359550561798,
+      "grad_norm": 0.9036487936973572,
+      "learning_rate": 4.787434823640589e-06,
+      "loss": 0.6565,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8240870786516854,
+      "grad_norm": 0.8997727036476135,
+      "learning_rate": 4.787248230290839e-06,
+      "loss": 0.6586,
+      "step": 2347
+    },
+    {
+      "epoch": 0.824438202247191,
+      "grad_norm": 0.9203174114227295,
+      "learning_rate": 4.787061558719158e-06,
+      "loss": 0.6684,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8247893258426966,
+      "grad_norm": 0.8566135168075562,
+      "learning_rate": 4.786874808931928e-06,
+      "loss": 0.6366,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8251404494382022,
+      "grad_norm": 0.8947523832321167,
+      "learning_rate": 4.786687980935536e-06,
+      "loss": 0.6954,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8254915730337079,
+      "grad_norm": 0.8714900016784668,
+      "learning_rate": 4.7865010747363725e-06,
+      "loss": 0.6678,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8258426966292135,
+      "grad_norm": 0.8742727041244507,
+      "learning_rate": 4.786314090340829e-06,
+      "loss": 0.6356,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8261938202247191,
+      "grad_norm": 0.8638887405395508,
+      "learning_rate": 4.7861270277553e-06,
+      "loss": 0.6643,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8265449438202247,
+      "grad_norm": 0.8582507967948914,
+      "learning_rate": 4.785939886986182e-06,
+      "loss": 0.6943,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8268960674157303,
+      "grad_norm": 0.8985755443572998,
+      "learning_rate": 4.785752668039877e-06,
+      "loss": 0.6899,
+      "step": 2355
+    },
+    {
+      "epoch": 0.827247191011236,
+      "grad_norm": 0.889373242855072,
+      "learning_rate": 4.7855653709227856e-06,
+      "loss": 0.6687,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8275983146067416,
+      "grad_norm": 0.8748716711997986,
+      "learning_rate": 4.785377995641315e-06,
+      "loss": 0.6813,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8279494382022472,
+      "grad_norm": 0.912598729133606,
+      "learning_rate": 4.785190542201873e-06,
+      "loss": 0.7037,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8283005617977528,
+      "grad_norm": 0.8579078912734985,
+      "learning_rate": 4.785003010610869e-06,
+      "loss": 0.651,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8286516853932584,
+      "grad_norm": 0.8841166496276855,
+      "learning_rate": 4.7848154008747185e-06,
+      "loss": 0.6527,
+      "step": 2360
+    },
+    {
+      "epoch": 0.829002808988764,
+      "grad_norm": 0.9118697047233582,
+      "learning_rate": 4.784627712999836e-06,
+      "loss": 0.7196,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8293539325842697,
+      "grad_norm": 0.8655892610549927,
+      "learning_rate": 4.78443994699264e-06,
+      "loss": 0.6813,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8297050561797753,
+      "grad_norm": 0.8933113217353821,
+      "learning_rate": 4.784252102859553e-06,
+      "loss": 0.6276,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8300561797752809,
+      "grad_norm": 0.8977401852607727,
+      "learning_rate": 4.784064180606999e-06,
+      "loss": 0.7107,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8304073033707865,
+      "grad_norm": 0.8428375720977783,
+      "learning_rate": 4.783876180241404e-06,
+      "loss": 0.6187,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8307584269662921,
+      "grad_norm": 0.8943214416503906,
+      "learning_rate": 4.783688101769199e-06,
+      "loss": 0.71,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8311095505617978,
+      "grad_norm": 0.8977285027503967,
+      "learning_rate": 4.783499945196814e-06,
+      "loss": 0.6895,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8314606741573034,
+      "grad_norm": 0.8746622800827026,
+      "learning_rate": 4.783311710530685e-06,
+      "loss": 0.656,
+      "step": 2368
+    },
+    {
+      "epoch": 0.831811797752809,
+      "grad_norm": 0.9107025265693665,
+      "learning_rate": 4.783123397777249e-06,
+      "loss": 0.6869,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8321629213483146,
+      "grad_norm": 0.8555697202682495,
+      "learning_rate": 4.782935006942948e-06,
+      "loss": 0.6607,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8325140449438202,
+      "grad_norm": 0.9654397368431091,
+      "learning_rate": 4.782746538034222e-06,
+      "loss": 0.7033,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8328651685393258,
+      "grad_norm": 0.8786224722862244,
+      "learning_rate": 4.782557991057519e-06,
+      "loss": 0.6855,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8332162921348315,
+      "grad_norm": 0.8756742477416992,
+      "learning_rate": 4.782369366019285e-06,
+      "loss": 0.6665,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8335674157303371,
+      "grad_norm": 0.8499640226364136,
+      "learning_rate": 4.782180662925971e-06,
+      "loss": 0.6311,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8339185393258427,
+      "grad_norm": 0.8776542544364929,
+      "learning_rate": 4.781991881784033e-06,
+      "loss": 0.6894,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8342696629213483,
+      "grad_norm": 0.8762704730033875,
+      "learning_rate": 4.7818030225999245e-06,
+      "loss": 0.6637,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8346207865168539,
+      "grad_norm": 0.8762661218643188,
+      "learning_rate": 4.7816140853801054e-06,
+      "loss": 0.6672,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8349719101123596,
+      "grad_norm": 0.8843420743942261,
+      "learning_rate": 4.781425070131036e-06,
+      "loss": 0.6614,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8353230337078652,
+      "grad_norm": 0.8788619041442871,
+      "learning_rate": 4.781235976859183e-06,
+      "loss": 0.6853,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8356741573033708,
+      "grad_norm": 0.9134128093719482,
+      "learning_rate": 4.78104680557101e-06,
+      "loss": 0.6904,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8360252808988764,
+      "grad_norm": 0.8329752683639526,
+      "learning_rate": 4.7808575562729894e-06,
+      "loss": 0.6318,
+      "step": 2381
+    },
+    {
+      "epoch": 0.836376404494382,
+      "grad_norm": 0.8506373763084412,
+      "learning_rate": 4.780668228971592e-06,
+      "loss": 0.6757,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8367275280898876,
+      "grad_norm": 0.8934277892112732,
+      "learning_rate": 4.780478823673292e-06,
+      "loss": 0.6961,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8370786516853933,
+      "grad_norm": 0.90306556224823,
+      "learning_rate": 4.780289340384569e-06,
+      "loss": 0.729,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8374297752808989,
+      "grad_norm": 0.8692042231559753,
+      "learning_rate": 4.780099779111901e-06,
+      "loss": 0.6695,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8377808988764045,
+      "grad_norm": 0.8926328420639038,
+      "learning_rate": 4.779910139861771e-06,
+      "loss": 0.6895,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8381320224719101,
+      "grad_norm": 0.8823839426040649,
+      "learning_rate": 4.779720422640666e-06,
+      "loss": 0.6587,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8384831460674157,
+      "grad_norm": 0.8437162637710571,
+      "learning_rate": 4.7795306274550735e-06,
+      "loss": 0.6577,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8388342696629213,
+      "grad_norm": 0.8754181265830994,
+      "learning_rate": 4.779340754311483e-06,
+      "loss": 0.6751,
+      "step": 2389
+    },
+    {
+      "epoch": 0.839185393258427,
+      "grad_norm": 0.8797729015350342,
+      "learning_rate": 4.77915080321639e-06,
+      "loss": 0.675,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8395365168539326,
+      "grad_norm": 0.8657658100128174,
+      "learning_rate": 4.778960774176291e-06,
+      "loss": 0.6504,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8398876404494382,
+      "grad_norm": 0.8967252969741821,
+      "learning_rate": 4.778770667197682e-06,
+      "loss": 0.6623,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8402387640449438,
+      "grad_norm": 0.881457507610321,
+      "learning_rate": 4.778580482287067e-06,
+      "loss": 0.6514,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8405898876404494,
+      "grad_norm": 0.8704044222831726,
+      "learning_rate": 4.778390219450949e-06,
+      "loss": 0.6695,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8409410112359551,
+      "grad_norm": 0.9062277674674988,
+      "learning_rate": 4.778199878695835e-06,
+      "loss": 0.6394,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8412921348314607,
+      "grad_norm": 0.873767614364624,
+      "learning_rate": 4.778009460028235e-06,
+      "loss": 0.6059,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8416432584269663,
+      "grad_norm": 0.9467976689338684,
+      "learning_rate": 4.77781896345466e-06,
+      "loss": 0.6834,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8419943820224719,
+      "grad_norm": 0.8977733254432678,
+      "learning_rate": 4.777628388981626e-06,
+      "loss": 0.703,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8423455056179775,
+      "grad_norm": 0.8968080282211304,
+      "learning_rate": 4.7774377366156505e-06,
+      "loss": 0.678,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8426966292134831,
+      "grad_norm": 0.8892068862915039,
+      "learning_rate": 4.777247006363253e-06,
+      "loss": 0.7203,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8430477528089888,
+      "grad_norm": 0.8844308853149414,
+      "learning_rate": 4.777056198230956e-06,
+      "loss": 0.6694,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8433988764044944,
+      "grad_norm": 0.8951604962348938,
+      "learning_rate": 4.7768653122252865e-06,
+      "loss": 0.6701,
+      "step": 2402
+    },
+    {
+      "epoch": 0.84375,
+      "grad_norm": 0.8704971075057983,
+      "learning_rate": 4.7766743483527715e-06,
+      "loss": 0.6616,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8441011235955056,
+      "grad_norm": 0.908189058303833,
+      "learning_rate": 4.776483306619941e-06,
+      "loss": 0.6668,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8444522471910112,
+      "grad_norm": 0.8829559087753296,
+      "learning_rate": 4.77629218703333e-06,
+      "loss": 0.6891,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8448033707865169,
+      "grad_norm": 0.912239134311676,
+      "learning_rate": 4.776100989599475e-06,
+      "loss": 0.7016,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8451544943820225,
+      "grad_norm": 0.8914303779602051,
+      "learning_rate": 4.775909714324912e-06,
+      "loss": 0.6593,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8455056179775281,
+      "grad_norm": 0.9197940826416016,
+      "learning_rate": 4.775718361216186e-06,
+      "loss": 0.677,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8458567415730337,
+      "grad_norm": 0.9217466115951538,
+      "learning_rate": 4.775526930279839e-06,
+      "loss": 0.7026,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8462078651685393,
+      "grad_norm": 0.8749791383743286,
+      "learning_rate": 4.775335421522418e-06,
+      "loss": 0.6502,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8465589887640449,
+      "grad_norm": 0.8961182236671448,
+      "learning_rate": 4.7751438349504716e-06,
+      "loss": 0.651,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8469101123595506,
+      "grad_norm": 0.8679896593093872,
+      "learning_rate": 4.7749521705705546e-06,
+      "loss": 0.6445,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8472612359550562,
+      "grad_norm": 0.8874735236167908,
+      "learning_rate": 4.774760428389219e-06,
+      "loss": 0.685,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8476123595505618,
+      "grad_norm": 0.8967884182929993,
+      "learning_rate": 4.774568608413024e-06,
+      "loss": 0.6812,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8479634831460674,
+      "grad_norm": 0.8984326720237732,
+      "learning_rate": 4.774376710648528e-06,
+      "loss": 0.6847,
+      "step": 2415
+    },
+    {
+      "epoch": 0.848314606741573,
+      "grad_norm": 0.861842155456543,
+      "learning_rate": 4.7741847351022955e-06,
+      "loss": 0.6643,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8486657303370787,
+      "grad_norm": 0.8478919267654419,
+      "learning_rate": 4.773992681780891e-06,
+      "loss": 0.6309,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8490168539325843,
+      "grad_norm": 0.8753276467323303,
+      "learning_rate": 4.773800550690882e-06,
+      "loss": 0.6906,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8493679775280899,
+      "grad_norm": 0.9123186469078064,
+      "learning_rate": 4.7736083418388404e-06,
+      "loss": 0.6698,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8497191011235955,
+      "grad_norm": 0.8448005318641663,
+      "learning_rate": 4.773416055231339e-06,
+      "loss": 0.6671,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8500702247191011,
+      "grad_norm": 0.886045515537262,
+      "learning_rate": 4.773223690874954e-06,
+      "loss": 0.6925,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8504213483146067,
+      "grad_norm": 0.8761569857597351,
+      "learning_rate": 4.773031248776264e-06,
+      "loss": 0.6988,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8507724719101124,
+      "grad_norm": 0.8618139028549194,
+      "learning_rate": 4.772838728941851e-06,
+      "loss": 0.6345,
+      "step": 2423
+    },
+    {
+      "epoch": 0.851123595505618,
+      "grad_norm": 0.8982943296432495,
+      "learning_rate": 4.772646131378297e-06,
+      "loss": 0.6775,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8514747191011236,
+      "grad_norm": 0.9190080761909485,
+      "learning_rate": 4.772453456092191e-06,
+      "loss": 0.6683,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8518258426966292,
+      "grad_norm": 0.8613485097885132,
+      "learning_rate": 4.772260703090121e-06,
+      "loss": 0.6378,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8521769662921348,
+      "grad_norm": 0.8885965943336487,
+      "learning_rate": 4.7720678723786796e-06,
+      "loss": 0.6971,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8525280898876404,
+      "grad_norm": 0.8576337099075317,
+      "learning_rate": 4.771874963964461e-06,
+      "loss": 0.6417,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8528792134831461,
+      "grad_norm": 0.8758781552314758,
+      "learning_rate": 4.771681977854062e-06,
+      "loss": 0.668,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8532303370786517,
+      "grad_norm": 0.881989061832428,
+      "learning_rate": 4.771488914054085e-06,
+      "loss": 0.655,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8535814606741573,
+      "grad_norm": 0.8963753581047058,
+      "learning_rate": 4.77129577257113e-06,
+      "loss": 0.647,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8539325842696629,
+      "grad_norm": 0.8885678648948669,
+      "learning_rate": 4.771102553411803e-06,
+      "loss": 0.6832,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8542837078651685,
+      "grad_norm": 0.8854667544364929,
+      "learning_rate": 4.770909256582712e-06,
+      "loss": 0.6387,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8546348314606742,
+      "grad_norm": 0.9203517436981201,
+      "learning_rate": 4.770715882090468e-06,
+      "loss": 0.6606,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8549859550561798,
+      "grad_norm": 0.9271900653839111,
+      "learning_rate": 4.770522429941683e-06,
+      "loss": 0.6787,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8553370786516854,
+      "grad_norm": 0.8828155994415283,
+      "learning_rate": 4.770328900142975e-06,
+      "loss": 0.6742,
+      "step": 2436
+    },
+    {
+      "epoch": 0.855688202247191,
+      "grad_norm": 0.8759010434150696,
+      "learning_rate": 4.770135292700961e-06,
+      "loss": 0.6742,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8560393258426966,
+      "grad_norm": 0.9363099932670593,
+      "learning_rate": 4.769941607622262e-06,
+      "loss": 0.6997,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8563904494382022,
+      "grad_norm": 0.89815753698349,
+      "learning_rate": 4.769747844913503e-06,
+      "loss": 0.6643,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8567415730337079,
+      "grad_norm": 0.8725033402442932,
+      "learning_rate": 4.76955400458131e-06,
+      "loss": 0.6433,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8570926966292135,
+      "grad_norm": 0.8719140291213989,
+      "learning_rate": 4.769360086632312e-06,
+      "loss": 0.6215,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8574438202247191,
+      "grad_norm": 0.8736863732337952,
+      "learning_rate": 4.76916609107314e-06,
+      "loss": 0.6854,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8577949438202247,
+      "grad_norm": 0.8882306814193726,
+      "learning_rate": 4.768972017910432e-06,
+      "loss": 0.678,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8581460674157303,
+      "grad_norm": 0.8788784146308899,
+      "learning_rate": 4.76877786715082e-06,
+      "loss": 0.6413,
+      "step": 2444
+    },
+    {
+      "epoch": 0.858497191011236,
+      "grad_norm": 0.8777339458465576,
+      "learning_rate": 4.768583638800947e-06,
+      "loss": 0.703,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8588483146067416,
+      "grad_norm": 0.8905528783798218,
+      "learning_rate": 4.7683893328674556e-06,
+      "loss": 0.6653,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8591994382022472,
+      "grad_norm": 0.8809261322021484,
+      "learning_rate": 4.76819494935699e-06,
+      "loss": 0.6357,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8595505617977528,
+      "grad_norm": 0.9078806638717651,
+      "learning_rate": 4.768000488276198e-06,
+      "loss": 0.6924,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8599016853932584,
+      "grad_norm": 0.8684207201004028,
+      "learning_rate": 4.767805949631729e-06,
+      "loss": 0.6753,
+      "step": 2449
+    },
+    {
+      "epoch": 0.860252808988764,
+      "grad_norm": 0.9299638867378235,
+      "learning_rate": 4.767611333430238e-06,
+      "loss": 0.6986,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8606039325842697,
+      "grad_norm": 0.9025764465332031,
+      "learning_rate": 4.7674166396783796e-06,
+      "loss": 0.6832,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8609550561797753,
+      "grad_norm": 0.8947761058807373,
+      "learning_rate": 4.767221868382812e-06,
+      "loss": 0.6927,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8613061797752809,
+      "grad_norm": 0.9044678807258606,
+      "learning_rate": 4.767027019550198e-06,
+      "loss": 0.6803,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8616573033707865,
+      "grad_norm": 0.8838576078414917,
+      "learning_rate": 4.766832093187199e-06,
+      "loss": 0.7069,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8620084269662921,
+      "grad_norm": 0.8987936973571777,
+      "learning_rate": 4.766637089300481e-06,
+      "loss": 0.6986,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8623595505617978,
+      "grad_norm": 0.9236810207366943,
+      "learning_rate": 4.766442007896716e-06,
+      "loss": 0.6912,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8627106741573034,
+      "grad_norm": 0.9139038920402527,
+      "learning_rate": 4.766246848982573e-06,
+      "loss": 0.6647,
+      "step": 2457
+    },
+    {
+      "epoch": 0.863061797752809,
+      "grad_norm": 0.8569420576095581,
+      "learning_rate": 4.766051612564726e-06,
+      "loss": 0.6092,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8634129213483146,
+      "grad_norm": 0.895006537437439,
+      "learning_rate": 4.765856298649855e-06,
+      "loss": 0.7071,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8637640449438202,
+      "grad_norm": 0.8703930377960205,
+      "learning_rate": 4.7656609072446355e-06,
+      "loss": 0.6839,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8641151685393258,
+      "grad_norm": 0.8899871706962585,
+      "learning_rate": 4.765465438355753e-06,
+      "loss": 0.6543,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8644662921348315,
+      "grad_norm": 0.929097592830658,
+      "learning_rate": 4.765269891989891e-06,
+      "loss": 0.6848,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8648174157303371,
+      "grad_norm": 0.8808541297912598,
+      "learning_rate": 4.765074268153736e-06,
+      "loss": 0.6625,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8651685393258427,
+      "grad_norm": 0.9321450591087341,
+      "learning_rate": 4.764878566853981e-06,
+      "loss": 0.6483,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8655196629213483,
+      "grad_norm": 0.8869593143463135,
+      "learning_rate": 4.764682788097317e-06,
+      "loss": 0.6546,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8658707865168539,
+      "grad_norm": 0.9124717116355896,
+      "learning_rate": 4.764486931890438e-06,
+      "loss": 0.6637,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8662219101123596,
+      "grad_norm": 0.9037688374519348,
+      "learning_rate": 4.764290998240044e-06,
+      "loss": 0.6656,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8665730337078652,
+      "grad_norm": 0.9029697179794312,
+      "learning_rate": 4.7640949871528365e-06,
+      "loss": 0.6837,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8669241573033708,
+      "grad_norm": 0.9010921716690063,
+      "learning_rate": 4.763898898635517e-06,
+      "loss": 0.6527,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8672752808988764,
+      "grad_norm": 0.8654772043228149,
+      "learning_rate": 4.763702732694792e-06,
+      "loss": 0.6983,
+      "step": 2470
+    },
+    {
+      "epoch": 0.867626404494382,
+      "grad_norm": 0.8883764147758484,
+      "learning_rate": 4.763506489337372e-06,
+      "loss": 0.7035,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8679775280898876,
+      "grad_norm": 0.902722179889679,
+      "learning_rate": 4.763310168569966e-06,
+      "loss": 0.6928,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8683286516853933,
+      "grad_norm": 0.900360643863678,
+      "learning_rate": 4.76311377039929e-06,
+      "loss": 0.653,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8686797752808989,
+      "grad_norm": 0.8905735015869141,
+      "learning_rate": 4.762917294832058e-06,
+      "loss": 0.6644,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8690308988764045,
+      "grad_norm": 0.8971735835075378,
+      "learning_rate": 4.762720741874992e-06,
+      "loss": 0.6751,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8693820224719101,
+      "grad_norm": 0.9062357544898987,
+      "learning_rate": 4.762524111534813e-06,
+      "loss": 0.6207,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8697331460674157,
+      "grad_norm": 0.8765476942062378,
+      "learning_rate": 4.762327403818245e-06,
+      "loss": 0.6972,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8700842696629213,
+      "grad_norm": 0.8708880543708801,
+      "learning_rate": 4.762130618732016e-06,
+      "loss": 0.6704,
+      "step": 2478
+    },
+    {
+      "epoch": 0.870435393258427,
+      "grad_norm": 0.8837103247642517,
+      "learning_rate": 4.761933756282856e-06,
+      "loss": 0.6314,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8707865168539326,
+      "grad_norm": 0.8385997414588928,
+      "learning_rate": 4.7617368164774956e-06,
+      "loss": 0.6511,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8711376404494382,
+      "grad_norm": 0.8760436773300171,
+      "learning_rate": 4.761539799322672e-06,
+      "loss": 0.6927,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8714887640449438,
+      "grad_norm": 0.8821952939033508,
+      "learning_rate": 4.761342704825124e-06,
+      "loss": 0.6682,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8718398876404494,
+      "grad_norm": 0.8891538381576538,
+      "learning_rate": 4.761145532991589e-06,
+      "loss": 0.6796,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8721910112359551,
+      "grad_norm": 0.9022207260131836,
+      "learning_rate": 4.760948283828812e-06,
+      "loss": 0.7124,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8725421348314607,
+      "grad_norm": 0.8623314499855042,
+      "learning_rate": 4.760750957343538e-06,
+      "loss": 0.6677,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8728932584269663,
+      "grad_norm": 0.8784610033035278,
+      "learning_rate": 4.760553553542517e-06,
+      "loss": 0.6731,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8732443820224719,
+      "grad_norm": 0.8761183619499207,
+      "learning_rate": 4.7603560724324985e-06,
+      "loss": 0.6246,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8735955056179775,
+      "grad_norm": 0.8424711227416992,
+      "learning_rate": 4.7601585140202354e-06,
+      "loss": 0.6453,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8739466292134831,
+      "grad_norm": 0.8858649730682373,
+      "learning_rate": 4.7599608783124865e-06,
+      "loss": 0.6869,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8742977528089888,
+      "grad_norm": 0.8897294402122498,
+      "learning_rate": 4.759763165316008e-06,
+      "loss": 0.673,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8746488764044944,
+      "grad_norm": 0.8692971467971802,
+      "learning_rate": 4.759565375037564e-06,
+      "loss": 0.6691,
+      "step": 2491
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.9153478741645813,
+      "learning_rate": 4.759367507483918e-06,
+      "loss": 0.7012,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8753511235955056,
+      "grad_norm": 0.9016615748405457,
+      "learning_rate": 4.759169562661836e-06,
+      "loss": 0.6758,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8757022471910112,
+      "grad_norm": 0.8769146203994751,
+      "learning_rate": 4.758971540578088e-06,
+      "loss": 0.6989,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8760533707865169,
+      "grad_norm": 0.8640443086624146,
+      "learning_rate": 4.758773441239447e-06,
+      "loss": 0.6355,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8764044943820225,
+      "grad_norm": 0.9241236448287964,
+      "learning_rate": 4.758575264652687e-06,
+      "loss": 0.642,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8767556179775281,
+      "grad_norm": 0.8472598791122437,
+      "learning_rate": 4.758377010824585e-06,
+      "loss": 0.6229,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8771067415730337,
+      "grad_norm": 0.8446848392486572,
+      "learning_rate": 4.758178679761922e-06,
+      "loss": 0.6234,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8774578651685393,
+      "grad_norm": 0.8826449513435364,
+      "learning_rate": 4.7579802714714815e-06,
+      "loss": 0.6799,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8778089887640449,
+      "grad_norm": 0.8934155106544495,
+      "learning_rate": 4.757781785960047e-06,
+      "loss": 0.655,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8781601123595506,
+      "grad_norm": 0.8733129501342773,
+      "learning_rate": 4.757583223234408e-06,
+      "loss": 0.6754,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8785112359550562,
+      "grad_norm": 0.8651498556137085,
+      "learning_rate": 4.757384583301353e-06,
+      "loss": 0.6839,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8788623595505618,
+      "grad_norm": 0.9135331511497498,
+      "learning_rate": 4.757185866167679e-06,
+      "loss": 0.6593,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8792134831460674,
+      "grad_norm": 0.86785888671875,
+      "learning_rate": 4.756987071840179e-06,
+      "loss": 0.6568,
+      "step": 2504
+    },
+    {
+      "epoch": 0.879564606741573,
+      "grad_norm": 0.9271560311317444,
+      "learning_rate": 4.756788200325652e-06,
+      "loss": 0.6364,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8799157303370787,
+      "grad_norm": 0.911849319934845,
+      "learning_rate": 4.7565892516309e-06,
+      "loss": 0.675,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8802668539325843,
+      "grad_norm": 0.8646685481071472,
+      "learning_rate": 4.756390225762727e-06,
+      "loss": 0.6056,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8806179775280899,
+      "grad_norm": 0.8820457458496094,
+      "learning_rate": 4.756191122727939e-06,
+      "loss": 0.6632,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8809691011235955,
+      "grad_norm": 0.8895183801651001,
+      "learning_rate": 4.755991942533344e-06,
+      "loss": 0.6819,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8813202247191011,
+      "grad_norm": 0.8657563924789429,
+      "learning_rate": 4.755792685185758e-06,
+      "loss": 0.649,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8816713483146067,
+      "grad_norm": 0.9095283150672913,
+      "learning_rate": 4.755593350691989e-06,
+      "loss": 0.718,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8820224719101124,
+      "grad_norm": 0.8777075409889221,
+      "learning_rate": 4.75539393905886e-06,
+      "loss": 0.6915,
+      "step": 2512
+    },
+    {
+      "epoch": 0.882373595505618,
+      "grad_norm": 0.8759570121765137,
+      "learning_rate": 4.755194450293187e-06,
+      "loss": 0.6749,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8827247191011236,
+      "grad_norm": 0.8764015436172485,
+      "learning_rate": 4.754994884401794e-06,
+      "loss": 0.6652,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8830758426966292,
+      "grad_norm": 0.9175446033477783,
+      "learning_rate": 4.754795241391505e-06,
+      "loss": 0.6626,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8834269662921348,
+      "grad_norm": 0.871394693851471,
+      "learning_rate": 4.754595521269149e-06,
+      "loss": 0.694,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8837780898876404,
+      "grad_norm": 0.9084363579750061,
+      "learning_rate": 4.754395724041556e-06,
+      "loss": 0.6725,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8841292134831461,
+      "grad_norm": 0.8901112079620361,
+      "learning_rate": 4.754195849715557e-06,
+      "loss": 0.6614,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8844803370786517,
+      "grad_norm": 0.8882237672805786,
+      "learning_rate": 4.7539958982979894e-06,
+      "loss": 0.6605,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8848314606741573,
+      "grad_norm": 0.899763286113739,
+      "learning_rate": 4.753795869795691e-06,
+      "loss": 0.6423,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8851825842696629,
+      "grad_norm": 0.9296014904975891,
+      "learning_rate": 4.753595764215503e-06,
+      "loss": 0.7166,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8855337078651685,
+      "grad_norm": 0.883310854434967,
+      "learning_rate": 4.753395581564267e-06,
+      "loss": 0.6404,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8858848314606742,
+      "grad_norm": 0.8889510035514832,
+      "learning_rate": 4.753195321848831e-06,
+      "loss": 0.649,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8862359550561798,
+      "grad_norm": 0.9097694754600525,
+      "learning_rate": 4.752994985076044e-06,
+      "loss": 0.6604,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8865870786516854,
+      "grad_norm": 0.8716392517089844,
+      "learning_rate": 4.752794571252755e-06,
+      "loss": 0.6989,
+      "step": 2525
+    },
+    {
+      "epoch": 0.886938202247191,
+      "grad_norm": 0.8778386116027832,
+      "learning_rate": 4.75259408038582e-06,
+      "loss": 0.638,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8872893258426966,
+      "grad_norm": 0.8789075016975403,
+      "learning_rate": 4.752393512482095e-06,
+      "loss": 0.6713,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8876404494382022,
+      "grad_norm": 0.8854068517684937,
+      "learning_rate": 4.752192867548439e-06,
+      "loss": 0.6534,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8879915730337079,
+      "grad_norm": 0.8878395557403564,
+      "learning_rate": 4.7519921455917145e-06,
+      "loss": 0.684,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8883426966292135,
+      "grad_norm": 0.8584818840026855,
+      "learning_rate": 4.751791346618786e-06,
+      "loss": 0.6259,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8886938202247191,
+      "grad_norm": 0.8890157341957092,
+      "learning_rate": 4.75159047063652e-06,
+      "loss": 0.6607,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8890449438202247,
+      "grad_norm": 0.851600706577301,
+      "learning_rate": 4.751389517651787e-06,
+      "loss": 0.66,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8893960674157303,
+      "grad_norm": 0.8763266801834106,
+      "learning_rate": 4.751188487671457e-06,
+      "loss": 0.6427,
+      "step": 2533
+    },
+    {
+      "epoch": 0.889747191011236,
+      "grad_norm": 0.9055187702178955,
+      "learning_rate": 4.75098738070241e-06,
+      "loss": 0.6804,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8900983146067416,
+      "grad_norm": 0.8860245943069458,
+      "learning_rate": 4.750786196751519e-06,
+      "loss": 0.6534,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8904494382022472,
+      "grad_norm": 0.8951229453086853,
+      "learning_rate": 4.7505849358256665e-06,
+      "loss": 0.6637,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8908005617977528,
+      "grad_norm": 0.9056205153465271,
+      "learning_rate": 4.7503835979317345e-06,
+      "loss": 0.6742,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8911516853932584,
+      "grad_norm": 0.8903935551643372,
+      "learning_rate": 4.75018218307661e-06,
+      "loss": 0.6889,
+      "step": 2538
+    },
+    {
+      "epoch": 0.891502808988764,
+      "grad_norm": 0.9006560444831848,
+      "learning_rate": 4.74998069126718e-06,
+      "loss": 0.7042,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8918539325842697,
+      "grad_norm": 0.8635199069976807,
+      "learning_rate": 4.749779122510337e-06,
+      "loss": 0.6692,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8922050561797753,
+      "grad_norm": 0.8788262009620667,
+      "learning_rate": 4.749577476812971e-06,
+      "loss": 0.6652,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8925561797752809,
+      "grad_norm": 0.8723106384277344,
+      "learning_rate": 4.749375754181981e-06,
+      "loss": 0.6373,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8929073033707865,
+      "grad_norm": 0.8863840103149414,
+      "learning_rate": 4.749173954624265e-06,
+      "loss": 0.6701,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8932584269662921,
+      "grad_norm": 0.9265256524085999,
+      "learning_rate": 4.748972078146725e-06,
+      "loss": 0.6591,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8936095505617978,
+      "grad_norm": 0.9035506844520569,
+      "learning_rate": 4.748770124756265e-06,
+      "loss": 0.6585,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8939606741573034,
+      "grad_norm": 0.8865588307380676,
+      "learning_rate": 4.748568094459789e-06,
+      "loss": 0.6903,
+      "step": 2546
+    },
+    {
+      "epoch": 0.894311797752809,
+      "grad_norm": 0.8840475678443909,
+      "learning_rate": 4.74836598726421e-06,
+      "loss": 0.6704,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8946629213483146,
+      "grad_norm": 0.8937922120094299,
+      "learning_rate": 4.748163803176437e-06,
+      "loss": 0.6649,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8950140449438202,
+      "grad_norm": 0.897407054901123,
+      "learning_rate": 4.747961542203387e-06,
+      "loss": 0.6705,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8953651685393258,
+      "grad_norm": 0.8580518364906311,
+      "learning_rate": 4.747759204351975e-06,
+      "loss": 0.6217,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8957162921348315,
+      "grad_norm": 0.8982558250427246,
+      "learning_rate": 4.747556789629121e-06,
+      "loss": 0.6873,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8960674157303371,
+      "grad_norm": 0.922828197479248,
+      "learning_rate": 4.747354298041749e-06,
+      "loss": 0.6833,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8964185393258427,
+      "grad_norm": 0.9047195315361023,
+      "learning_rate": 4.747151729596783e-06,
+      "loss": 0.6731,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8967696629213483,
+      "grad_norm": 0.9132267236709595,
+      "learning_rate": 4.7469490843011515e-06,
+      "loss": 0.6752,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8971207865168539,
+      "grad_norm": 0.8471392393112183,
+      "learning_rate": 4.7467463621617825e-06,
+      "loss": 0.6552,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8974719101123596,
+      "grad_norm": 0.8814915418624878,
+      "learning_rate": 4.746543563185612e-06,
+      "loss": 0.676,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8978230337078652,
+      "grad_norm": 0.8573102355003357,
+      "learning_rate": 4.746340687379573e-06,
+      "loss": 0.6397,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8981741573033708,
+      "grad_norm": 0.8620328903198242,
+      "learning_rate": 4.746137734750604e-06,
+      "loss": 0.625,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8985252808988764,
+      "grad_norm": 0.8743171691894531,
+      "learning_rate": 4.745934705305648e-06,
+      "loss": 0.6729,
+      "step": 2559
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.838996410369873,
+      "learning_rate": 4.745731599051646e-06,
+      "loss": 0.6507,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8992275280898876,
+      "grad_norm": 0.9150429964065552,
+      "learning_rate": 4.745528415995546e-06,
+      "loss": 0.6869,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8995786516853933,
+      "grad_norm": 0.8738797903060913,
+      "learning_rate": 4.745325156144295e-06,
+      "loss": 0.6843,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8999297752808989,
+      "grad_norm": 0.871146559715271,
+      "learning_rate": 4.745121819504845e-06,
+      "loss": 0.6773,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9002808988764045,
+      "grad_norm": 0.8615120649337769,
+      "learning_rate": 4.74491840608415e-06,
+      "loss": 0.6383,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9006320224719101,
+      "grad_norm": 0.8572788834571838,
+      "learning_rate": 4.744714915889167e-06,
+      "loss": 0.6488,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9009831460674157,
+      "grad_norm": 0.8589732050895691,
+      "learning_rate": 4.744511348926855e-06,
+      "loss": 0.6464,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9013342696629213,
+      "grad_norm": 0.8711311221122742,
+      "learning_rate": 4.744307705204174e-06,
+      "loss": 0.6809,
+      "step": 2567
+    },
+    {
+      "epoch": 0.901685393258427,
+      "grad_norm": 0.9025472402572632,
+      "learning_rate": 4.744103984728091e-06,
+      "loss": 0.6973,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9020365168539326,
+      "grad_norm": 0.8648887872695923,
+      "learning_rate": 4.743900187505572e-06,
+      "loss": 0.6666,
+      "step": 2569
+    },
+    {
+      "epoch": 0.9023876404494382,
+      "grad_norm": 0.8922040462493896,
+      "learning_rate": 4.7436963135435864e-06,
+      "loss": 0.6875,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9027387640449438,
+      "grad_norm": 0.8693889379501343,
+      "learning_rate": 4.743492362849107e-06,
+      "loss": 0.6926,
+      "step": 2571
+    },
+    {
+      "epoch": 0.9030898876404494,
+      "grad_norm": 0.8809205889701843,
+      "learning_rate": 4.743288335429109e-06,
+      "loss": 0.6841,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9034410112359551,
+      "grad_norm": 0.8538480401039124,
+      "learning_rate": 4.743084231290569e-06,
+      "loss": 0.6416,
+      "step": 2573
+    },
+    {
+      "epoch": 0.9037921348314607,
+      "grad_norm": 0.8628817200660706,
+      "learning_rate": 4.742880050440468e-06,
+      "loss": 0.649,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9041432584269663,
+      "grad_norm": 0.9055478572845459,
+      "learning_rate": 4.742675792885788e-06,
+      "loss": 0.6819,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9044943820224719,
+      "grad_norm": 0.8601617217063904,
+      "learning_rate": 4.742471458633516e-06,
+      "loss": 0.6675,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9048455056179775,
+      "grad_norm": 0.8635382056236267,
+      "learning_rate": 4.742267047690638e-06,
+      "loss": 0.6537,
+      "step": 2577
+    },
+    {
+      "epoch": 0.9051966292134831,
+      "grad_norm": 0.892162561416626,
+      "learning_rate": 4.742062560064146e-06,
+      "loss": 0.6627,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9055477528089888,
+      "grad_norm": 0.9016808867454529,
+      "learning_rate": 4.741857995761033e-06,
+      "loss": 0.666,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9058988764044944,
+      "grad_norm": 0.8793184161186218,
+      "learning_rate": 4.741653354788295e-06,
+      "loss": 0.638,
+      "step": 2580
+    },
+    {
+      "epoch": 0.90625,
+      "grad_norm": 0.8721537590026855,
+      "learning_rate": 4.741448637152931e-06,
+      "loss": 0.6816,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9066011235955056,
+      "grad_norm": 0.9141138792037964,
+      "learning_rate": 4.741243842861941e-06,
+      "loss": 0.6959,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9069522471910112,
+      "grad_norm": 0.8552782535552979,
+      "learning_rate": 4.741038971922329e-06,
+      "loss": 0.6868,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9073033707865169,
+      "grad_norm": 0.8622416257858276,
+      "learning_rate": 4.740834024341103e-06,
+      "loss": 0.6861,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9076544943820225,
+      "grad_norm": 0.8589463829994202,
+      "learning_rate": 4.74062900012527e-06,
+      "loss": 0.6525,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9080056179775281,
+      "grad_norm": 0.9055203795433044,
+      "learning_rate": 4.740423899281843e-06,
+      "loss": 0.6724,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9083567415730337,
+      "grad_norm": 0.9066650867462158,
+      "learning_rate": 4.740218721817836e-06,
+      "loss": 0.6929,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9087078651685393,
+      "grad_norm": 0.8654912114143372,
+      "learning_rate": 4.7400134677402655e-06,
+      "loss": 0.6704,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9090589887640449,
+      "grad_norm": 0.9301373958587646,
+      "learning_rate": 4.739808137056151e-06,
+      "loss": 0.6797,
+      "step": 2589
+    },
+    {
+      "epoch": 0.9094101123595506,
+      "grad_norm": 0.8802602291107178,
+      "learning_rate": 4.739602729772515e-06,
+      "loss": 0.6657,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9097612359550562,
+      "grad_norm": 0.873859167098999,
+      "learning_rate": 4.739397245896382e-06,
+      "loss": 0.6364,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9101123595505618,
+      "grad_norm": 0.8515645861625671,
+      "learning_rate": 4.739191685434779e-06,
+      "loss": 0.6472,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9104634831460674,
+      "grad_norm": 0.8645114898681641,
+      "learning_rate": 4.738986048394737e-06,
+      "loss": 0.6792,
+      "step": 2593
+    },
+    {
+      "epoch": 0.910814606741573,
+      "grad_norm": 0.862397313117981,
+      "learning_rate": 4.738780334783288e-06,
+      "loss": 0.6522,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9111657303370787,
+      "grad_norm": 0.8489444255828857,
+      "learning_rate": 4.738574544607467e-06,
+      "loss": 0.6757,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9115168539325843,
+      "grad_norm": 0.8827484250068665,
+      "learning_rate": 4.738368677874313e-06,
+      "loss": 0.6842,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9118679775280899,
+      "grad_norm": 0.8966019749641418,
+      "learning_rate": 4.738162734590865e-06,
+      "loss": 0.6658,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9122191011235955,
+      "grad_norm": 0.8434582948684692,
+      "learning_rate": 4.7379567147641665e-06,
+      "loss": 0.6195,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9125702247191011,
+      "grad_norm": 0.9089177250862122,
+      "learning_rate": 4.737750618401264e-06,
+      "loss": 0.7168,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9129213483146067,
+      "grad_norm": 0.9108870625495911,
+      "learning_rate": 4.737544445509204e-06,
+      "loss": 0.6995,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9132724719101124,
+      "grad_norm": 0.8936570882797241,
+      "learning_rate": 4.73733819609504e-06,
+      "loss": 0.6726,
+      "step": 2601
+    },
+    {
+      "epoch": 0.913623595505618,
+      "grad_norm": 0.8761941194534302,
+      "learning_rate": 4.737131870165824e-06,
+      "loss": 0.6642,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9139747191011236,
+      "grad_norm": 0.870694100856781,
+      "learning_rate": 4.736925467728612e-06,
+      "loss": 0.6945,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9143258426966292,
+      "grad_norm": 0.8931704759597778,
+      "learning_rate": 4.736718988790464e-06,
+      "loss": 0.6777,
+      "step": 2604
+    },
+    {
+      "epoch": 0.9146769662921348,
+      "grad_norm": 0.8875551223754883,
+      "learning_rate": 4.73651243335844e-06,
+      "loss": 0.6664,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9150280898876404,
+      "grad_norm": 0.8898186087608337,
+      "learning_rate": 4.736305801439603e-06,
+      "loss": 0.6652,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9153792134831461,
+      "grad_norm": 0.8893354535102844,
+      "learning_rate": 4.736099093041023e-06,
+      "loss": 0.6587,
+      "step": 2607
+    },
+    {
+      "epoch": 0.9157303370786517,
+      "grad_norm": 0.9278960824012756,
+      "learning_rate": 4.735892308169768e-06,
+      "loss": 0.6674,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9160814606741573,
+      "grad_norm": 0.9112112522125244,
+      "learning_rate": 4.735685446832908e-06,
+      "loss": 0.6435,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9164325842696629,
+      "grad_norm": 0.902930736541748,
+      "learning_rate": 4.735478509037519e-06,
+      "loss": 0.6402,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9167837078651685,
+      "grad_norm": 0.8808385729789734,
+      "learning_rate": 4.7352714947906785e-06,
+      "loss": 0.707,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9171348314606742,
+      "grad_norm": 0.8619163036346436,
+      "learning_rate": 4.735064404099465e-06,
+      "loss": 0.6728,
+      "step": 2612
+    },
+    {
+      "epoch": 0.9174859550561798,
+      "grad_norm": 0.8795300126075745,
+      "learning_rate": 4.734857236970962e-06,
+      "loss": 0.6543,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9178370786516854,
+      "grad_norm": 0.9281381964683533,
+      "learning_rate": 4.734649993412254e-06,
+      "loss": 0.6594,
+      "step": 2614
+    },
+    {
+      "epoch": 0.918188202247191,
+      "grad_norm": 0.8637987971305847,
+      "learning_rate": 4.734442673430428e-06,
+      "loss": 0.6664,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9185393258426966,
+      "grad_norm": 0.8663271069526672,
+      "learning_rate": 4.734235277032575e-06,
+      "loss": 0.6658,
+      "step": 2616
+    },
+    {
+      "epoch": 0.9188904494382022,
+      "grad_norm": 0.9028787016868591,
+      "learning_rate": 4.734027804225787e-06,
+      "loss": 0.6473,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9192415730337079,
+      "grad_norm": 0.8826063275337219,
+      "learning_rate": 4.733820255017161e-06,
+      "loss": 0.626,
+      "step": 2618
+    },
+    {
+      "epoch": 0.9195926966292135,
+      "grad_norm": 0.8579033613204956,
+      "learning_rate": 4.7336126294137915e-06,
+      "loss": 0.6527,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9199438202247191,
+      "grad_norm": 0.8992764353752136,
+      "learning_rate": 4.733404927422783e-06,
+      "loss": 0.6841,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9202949438202247,
+      "grad_norm": 0.9202234148979187,
+      "learning_rate": 4.733197149051237e-06,
+      "loss": 0.6871,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9206460674157303,
+      "grad_norm": 0.8881893754005432,
+      "learning_rate": 4.73298929430626e-06,
+      "loss": 0.6851,
+      "step": 2622
+    },
+    {
+      "epoch": 0.920997191011236,
+      "grad_norm": 0.9517401456832886,
+      "learning_rate": 4.732781363194958e-06,
+      "loss": 0.6748,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9213483146067416,
+      "grad_norm": 0.9222083687782288,
+      "learning_rate": 4.7325733557244455e-06,
+      "loss": 0.6939,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9216994382022472,
+      "grad_norm": 0.8639568090438843,
+      "learning_rate": 4.732365271901834e-06,
+      "loss": 0.664,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9220505617977528,
+      "grad_norm": 0.9032568335533142,
+      "learning_rate": 4.732157111734241e-06,
+      "loss": 0.6779,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9224016853932584,
+      "grad_norm": 0.9278046488761902,
+      "learning_rate": 4.731948875228784e-06,
+      "loss": 0.682,
+      "step": 2627
+    },
+    {
+      "epoch": 0.922752808988764,
+      "grad_norm": 0.9298899173736572,
+      "learning_rate": 4.731740562392587e-06,
+      "loss": 0.687,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9231039325842697,
+      "grad_norm": 0.8798907995223999,
+      "learning_rate": 4.7315321732327715e-06,
+      "loss": 0.6722,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9234550561797753,
+      "grad_norm": 0.8544273376464844,
+      "learning_rate": 4.731323707756465e-06,
+      "loss": 0.6801,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9238061797752809,
+      "grad_norm": 0.8940860033035278,
+      "learning_rate": 4.731115165970797e-06,
+      "loss": 0.6548,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9241573033707865,
+      "grad_norm": 0.9600421190261841,
+      "learning_rate": 4.730906547882901e-06,
+      "loss": 0.6935,
+      "step": 2632
+    },
+    {
+      "epoch": 0.9245084269662921,
+      "grad_norm": 0.9062276482582092,
+      "learning_rate": 4.730697853499908e-06,
+      "loss": 0.7043,
+      "step": 2633
+    },
+    {
+      "epoch": 0.9248595505617978,
+      "grad_norm": 0.8218638300895691,
+      "learning_rate": 4.730489082828959e-06,
+      "loss": 0.5993,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9252106741573034,
+      "grad_norm": 0.8894577622413635,
+      "learning_rate": 4.7302802358771905e-06,
+      "loss": 0.6411,
+      "step": 2635
+    },
+    {
+      "epoch": 0.925561797752809,
+      "grad_norm": 0.881628155708313,
+      "learning_rate": 4.730071312651747e-06,
+      "loss": 0.6686,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9259129213483146,
+      "grad_norm": 0.925878643989563,
+      "learning_rate": 4.7298623131597735e-06,
+      "loss": 0.6674,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9262640449438202,
+      "grad_norm": 0.8464183211326599,
+      "learning_rate": 4.729653237408416e-06,
+      "loss": 0.6902,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9266151685393258,
+      "grad_norm": 0.873918890953064,
+      "learning_rate": 4.729444085404826e-06,
+      "loss": 0.6737,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9269662921348315,
+      "grad_norm": 0.8666852116584778,
+      "learning_rate": 4.7292348571561566e-06,
+      "loss": 0.6601,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9273174157303371,
+      "grad_norm": 1.0832674503326416,
+      "learning_rate": 4.729025552669563e-06,
+      "loss": 0.654,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9276685393258427,
+      "grad_norm": 0.8796826004981995,
+      "learning_rate": 4.728816171952202e-06,
+      "loss": 0.6365,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9280196629213483,
+      "grad_norm": 0.8894056081771851,
+      "learning_rate": 4.728606715011235e-06,
+      "loss": 0.6955,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9283707865168539,
+      "grad_norm": 0.9130917191505432,
+      "learning_rate": 4.728397181853826e-06,
+      "loss": 0.6807,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9287219101123596,
+      "grad_norm": 0.8895679116249084,
+      "learning_rate": 4.72818757248714e-06,
+      "loss": 0.71,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9290730337078652,
+      "grad_norm": 0.8796029090881348,
+      "learning_rate": 4.727977886918345e-06,
+      "loss": 0.666,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9294241573033708,
+      "grad_norm": 0.8870053887367249,
+      "learning_rate": 4.727768125154614e-06,
+      "loss": 0.6724,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9297752808988764,
+      "grad_norm": 0.8785473108291626,
+      "learning_rate": 4.727558287203119e-06,
+      "loss": 0.6735,
+      "step": 2648
+    },
+    {
+      "epoch": 0.930126404494382,
+      "grad_norm": 0.8722987174987793,
+      "learning_rate": 4.727348373071037e-06,
+      "loss": 0.6816,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9304775280898876,
+      "grad_norm": 0.8995219469070435,
+      "learning_rate": 4.7271383827655455e-06,
+      "loss": 0.6657,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9308286516853933,
+      "grad_norm": 0.8835882544517517,
+      "learning_rate": 4.726928316293828e-06,
+      "loss": 0.6973,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9311797752808989,
+      "grad_norm": 0.8654088973999023,
+      "learning_rate": 4.726718173663068e-06,
+      "loss": 0.6083,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9315308988764045,
+      "grad_norm": 0.832149088382721,
+      "learning_rate": 4.726507954880451e-06,
+      "loss": 0.5915,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9318820224719101,
+      "grad_norm": 0.8820677995681763,
+      "learning_rate": 4.726297659953169e-06,
+      "loss": 0.641,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9322331460674157,
+      "grad_norm": 0.9029566049575806,
+      "learning_rate": 4.7260872888884105e-06,
+      "loss": 0.6845,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9325842696629213,
+      "grad_norm": 0.8395662307739258,
+      "learning_rate": 4.725876841693372e-06,
+      "loss": 0.6566,
+      "step": 2656
+    },
+    {
+      "epoch": 0.932935393258427,
+      "grad_norm": 0.894307017326355,
+      "learning_rate": 4.7256663183752495e-06,
+      "loss": 0.6985,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9332865168539326,
+      "grad_norm": 0.9100427031517029,
+      "learning_rate": 4.725455718941245e-06,
+      "loss": 0.6687,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9336376404494382,
+      "grad_norm": 0.874563455581665,
+      "learning_rate": 4.725245043398558e-06,
+      "loss": 0.6692,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9339887640449438,
+      "grad_norm": 0.8920844793319702,
+      "learning_rate": 4.725034291754394e-06,
+      "loss": 0.6334,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9343398876404494,
+      "grad_norm": 0.8333306908607483,
+      "learning_rate": 4.724823464015963e-06,
+      "loss": 0.6322,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9346910112359551,
+      "grad_norm": 0.8971813917160034,
+      "learning_rate": 4.724612560190473e-06,
+      "loss": 0.6823,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9350421348314607,
+      "grad_norm": 0.8798514008522034,
+      "learning_rate": 4.724401580285136e-06,
+      "loss": 0.6742,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9353932584269663,
+      "grad_norm": 0.827732503414154,
+      "learning_rate": 4.724190524307169e-06,
+      "loss": 0.6574,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9357443820224719,
+      "grad_norm": 0.8697452545166016,
+      "learning_rate": 4.7239793922637896e-06,
+      "loss": 0.637,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9360955056179775,
+      "grad_norm": 0.887448251247406,
+      "learning_rate": 4.723768184162217e-06,
+      "loss": 0.6762,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9364466292134831,
+      "grad_norm": 0.9078186750411987,
+      "learning_rate": 4.723556900009677e-06,
+      "loss": 0.6832,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9367977528089888,
+      "grad_norm": 0.8617350459098816,
+      "learning_rate": 4.723345539813393e-06,
+      "loss": 0.6591,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9371488764044944,
+      "grad_norm": 0.8676525354385376,
+      "learning_rate": 4.723134103580594e-06,
+      "loss": 0.6941,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.8551449179649353,
+      "learning_rate": 4.722922591318511e-06,
+      "loss": 0.6142,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9378511235955056,
+      "grad_norm": 0.9027611017227173,
+      "learning_rate": 4.722711003034378e-06,
+      "loss": 0.6893,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9382022471910112,
+      "grad_norm": 0.8602604269981384,
+      "learning_rate": 4.7224993387354305e-06,
+      "loss": 0.644,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9385533707865169,
+      "grad_norm": 0.9275681376457214,
+      "learning_rate": 4.722287598428907e-06,
+      "loss": 0.676,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9389044943820225,
+      "grad_norm": 0.880574107170105,
+      "learning_rate": 4.7220757821220495e-06,
+      "loss": 0.6068,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9392556179775281,
+      "grad_norm": 0.8658181428909302,
+      "learning_rate": 4.721863889822102e-06,
+      "loss": 0.6714,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9396067415730337,
+      "grad_norm": 0.890250563621521,
+      "learning_rate": 4.7216519215363095e-06,
+      "loss": 0.6624,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9399578651685393,
+      "grad_norm": 0.870218813419342,
+      "learning_rate": 4.721439877271924e-06,
+      "loss": 0.6692,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9403089887640449,
+      "grad_norm": 0.8716661334037781,
+      "learning_rate": 4.721227757036195e-06,
+      "loss": 0.6513,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9406601123595506,
+      "grad_norm": 0.8979821801185608,
+      "learning_rate": 4.721015560836377e-06,
+      "loss": 0.6658,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9410112359550562,
+      "grad_norm": 1.3192576169967651,
+      "learning_rate": 4.7208032886797285e-06,
+      "loss": 0.6726,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9413623595505618,
+      "grad_norm": 0.9058758020401001,
+      "learning_rate": 4.720590940573507e-06,
+      "loss": 0.6764,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9417134831460674,
+      "grad_norm": 0.8641902208328247,
+      "learning_rate": 4.720378516524976e-06,
+      "loss": 0.7075,
+      "step": 2682
+    },
+    {
+      "epoch": 0.942064606741573,
+      "grad_norm": 0.8593077659606934,
+      "learning_rate": 4.7201660165413994e-06,
+      "loss": 0.6575,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9424157303370787,
+      "grad_norm": 0.8111205697059631,
+      "learning_rate": 4.719953440630045e-06,
+      "loss": 0.5865,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9427668539325843,
+      "grad_norm": 0.8845921754837036,
+      "learning_rate": 4.719740788798182e-06,
+      "loss": 0.6787,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9431179775280899,
+      "grad_norm": 0.893483579158783,
+      "learning_rate": 4.719528061053084e-06,
+      "loss": 0.657,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9434691011235955,
+      "grad_norm": 0.882802426815033,
+      "learning_rate": 4.719315257402026e-06,
+      "loss": 0.6678,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9438202247191011,
+      "grad_norm": 0.8567876815795898,
+      "learning_rate": 4.7191023778522844e-06,
+      "loss": 0.6167,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9441713483146067,
+      "grad_norm": 0.8992242217063904,
+      "learning_rate": 4.718889422411141e-06,
+      "loss": 0.6691,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9445224719101124,
+      "grad_norm": 0.8735215663909912,
+      "learning_rate": 4.718676391085878e-06,
+      "loss": 0.63,
+      "step": 2690
+    },
+    {
+      "epoch": 0.944873595505618,
+      "grad_norm": 0.8774729371070862,
+      "learning_rate": 4.71846328388378e-06,
+      "loss": 0.6702,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9452247191011236,
+      "grad_norm": 0.9470273852348328,
+      "learning_rate": 4.718250100812138e-06,
+      "loss": 0.6997,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9455758426966292,
+      "grad_norm": 0.9050356149673462,
+      "learning_rate": 4.71803684187824e-06,
+      "loss": 0.6451,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9459269662921348,
+      "grad_norm": 0.9217947125434875,
+      "learning_rate": 4.7178235070893796e-06,
+      "loss": 0.6556,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9462780898876404,
+      "grad_norm": 0.909965991973877,
+      "learning_rate": 4.717610096452852e-06,
+      "loss": 0.6772,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9466292134831461,
+      "grad_norm": 0.8807794451713562,
+      "learning_rate": 4.717396609975959e-06,
+      "loss": 0.6805,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9469803370786517,
+      "grad_norm": 0.9008699059486389,
+      "learning_rate": 4.717183047665998e-06,
+      "loss": 0.6718,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9473314606741573,
+      "grad_norm": 0.9289125204086304,
+      "learning_rate": 4.716969409530274e-06,
+      "loss": 0.6705,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9476825842696629,
+      "grad_norm": 0.9074335098266602,
+      "learning_rate": 4.716755695576094e-06,
+      "loss": 0.6897,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9480337078651685,
+      "grad_norm": 0.9094159603118896,
+      "learning_rate": 4.716541905810766e-06,
+      "loss": 0.6625,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9483848314606742,
+      "grad_norm": 0.8951701521873474,
+      "learning_rate": 4.716328040241601e-06,
+      "loss": 0.6234,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9487359550561798,
+      "grad_norm": 0.8777403831481934,
+      "learning_rate": 4.716114098875914e-06,
+      "loss": 0.6418,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9490870786516854,
+      "grad_norm": 0.9449620842933655,
+      "learning_rate": 4.715900081721021e-06,
+      "loss": 0.6814,
+      "step": 2703
+    },
+    {
+      "epoch": 0.949438202247191,
+      "grad_norm": 0.8588343262672424,
+      "learning_rate": 4.7156859887842416e-06,
+      "loss": 0.6042,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9497893258426966,
+      "grad_norm": 0.8675066232681274,
+      "learning_rate": 4.715471820072896e-06,
+      "loss": 0.6591,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9501404494382022,
+      "grad_norm": 0.8843618631362915,
+      "learning_rate": 4.715257575594312e-06,
+      "loss": 0.6594,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9504915730337079,
+      "grad_norm": 0.8740993142127991,
+      "learning_rate": 4.715043255355814e-06,
+      "loss": 0.6578,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9508426966292135,
+      "grad_norm": 0.8732353448867798,
+      "learning_rate": 4.7148288593647315e-06,
+      "loss": 0.6399,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9511938202247191,
+      "grad_norm": 0.8841140866279602,
+      "learning_rate": 4.714614387628398e-06,
+      "loss": 0.6879,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9515449438202247,
+      "grad_norm": 0.9016351103782654,
+      "learning_rate": 4.714399840154147e-06,
+      "loss": 0.6618,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9518960674157303,
+      "grad_norm": 0.8737030625343323,
+      "learning_rate": 4.714185216949317e-06,
+      "loss": 0.6616,
+      "step": 2711
+    },
+    {
+      "epoch": 0.952247191011236,
+      "grad_norm": 0.8543287515640259,
+      "learning_rate": 4.713970518021246e-06,
+      "loss": 0.627,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9525983146067416,
+      "grad_norm": 0.8959774374961853,
+      "learning_rate": 4.713755743377278e-06,
+      "loss": 0.6936,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9529494382022472,
+      "grad_norm": 0.9039850831031799,
+      "learning_rate": 4.713540893024759e-06,
+      "loss": 0.6465,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9533005617977528,
+      "grad_norm": 0.8787552118301392,
+      "learning_rate": 4.713325966971035e-06,
+      "loss": 0.6127,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9536516853932584,
+      "grad_norm": 0.8473005294799805,
+      "learning_rate": 4.713110965223457e-06,
+      "loss": 0.6648,
+      "step": 2716
+    },
+    {
+      "epoch": 0.954002808988764,
+      "grad_norm": 0.937426745891571,
+      "learning_rate": 4.712895887789377e-06,
+      "loss": 0.6594,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9543539325842697,
+      "grad_norm": 0.8719070553779602,
+      "learning_rate": 4.712680734676153e-06,
+      "loss": 0.6652,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9547050561797753,
+      "grad_norm": 0.8872880935668945,
+      "learning_rate": 4.7124655058911395e-06,
+      "loss": 0.6987,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9550561797752809,
+      "grad_norm": 0.8662359714508057,
+      "learning_rate": 4.7122502014417e-06,
+      "loss": 0.6723,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9554073033707865,
+      "grad_norm": 0.8790738582611084,
+      "learning_rate": 4.712034821335196e-06,
+      "loss": 0.6421,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9557584269662921,
+      "grad_norm": 0.8946492671966553,
+      "learning_rate": 4.711819365578995e-06,
+      "loss": 0.6439,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9561095505617978,
+      "grad_norm": 0.9047174453735352,
+      "learning_rate": 4.711603834180464e-06,
+      "loss": 0.6859,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9564606741573034,
+      "grad_norm": 0.9128047823905945,
+      "learning_rate": 4.711388227146974e-06,
+      "loss": 0.671,
+      "step": 2724
+    },
+    {
+      "epoch": 0.956811797752809,
+      "grad_norm": 0.8675251007080078,
+      "learning_rate": 4.711172544485898e-06,
+      "loss": 0.672,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9571629213483146,
+      "grad_norm": 0.8741472363471985,
+      "learning_rate": 4.710956786204614e-06,
+      "loss": 0.6614,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9575140449438202,
+      "grad_norm": 0.851613461971283,
+      "learning_rate": 4.710740952310501e-06,
+      "loss": 0.645,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9578651685393258,
+      "grad_norm": 0.9370160698890686,
+      "learning_rate": 4.7105250428109375e-06,
+      "loss": 0.6743,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9582162921348315,
+      "grad_norm": 0.9051042199134827,
+      "learning_rate": 4.71030905771331e-06,
+      "loss": 0.6166,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9585674157303371,
+      "grad_norm": 0.8878385424613953,
+      "learning_rate": 4.7100929970250035e-06,
+      "loss": 0.6447,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9589185393258427,
+      "grad_norm": 0.857430100440979,
+      "learning_rate": 4.709876860753409e-06,
+      "loss": 0.6557,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9592696629213483,
+      "grad_norm": 0.8934675455093384,
+      "learning_rate": 4.709660648905916e-06,
+      "loss": 0.675,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9596207865168539,
+      "grad_norm": 0.8855241537094116,
+      "learning_rate": 4.709444361489918e-06,
+      "loss": 0.6705,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9599719101123596,
+      "grad_norm": 0.8473418354988098,
+      "learning_rate": 4.7092279985128165e-06,
+      "loss": 0.6398,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9603230337078652,
+      "grad_norm": 0.9071454405784607,
+      "learning_rate": 4.709011559982006e-06,
+      "loss": 0.6824,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9606741573033708,
+      "grad_norm": 0.8947198390960693,
+      "learning_rate": 4.708795045904891e-06,
+      "loss": 0.6794,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9610252808988764,
+      "grad_norm": 0.8824878931045532,
+      "learning_rate": 4.708578456288875e-06,
+      "loss": 0.6453,
+      "step": 2737
+    },
+    {
+      "epoch": 0.961376404494382,
+      "grad_norm": 0.8897850513458252,
+      "learning_rate": 4.708361791141367e-06,
+      "loss": 0.6708,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9617275280898876,
+      "grad_norm": 0.8753004670143127,
+      "learning_rate": 4.708145050469774e-06,
+      "loss": 0.6574,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9620786516853933,
+      "grad_norm": 0.8702830672264099,
+      "learning_rate": 4.7079282342815105e-06,
+      "loss": 0.6363,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9624297752808989,
+      "grad_norm": 0.8557823896408081,
+      "learning_rate": 4.707711342583991e-06,
+      "loss": 0.6607,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9627808988764045,
+      "grad_norm": 0.9189360737800598,
+      "learning_rate": 4.7074943753846324e-06,
+      "loss": 0.6732,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9631320224719101,
+      "grad_norm": 0.8712623715400696,
+      "learning_rate": 4.707277332690855e-06,
+      "loss": 0.6962,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9634831460674157,
+      "grad_norm": 0.8758993744850159,
+      "learning_rate": 4.707060214510082e-06,
+      "loss": 0.688,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9638342696629213,
+      "grad_norm": 0.9074702262878418,
+      "learning_rate": 4.706843020849738e-06,
+      "loss": 0.6668,
+      "step": 2745
+    },
+    {
+      "epoch": 0.964185393258427,
+      "grad_norm": 0.8798408508300781,
+      "learning_rate": 4.706625751717251e-06,
+      "loss": 0.6728,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9645365168539326,
+      "grad_norm": 0.8691076636314392,
+      "learning_rate": 4.706408407120053e-06,
+      "loss": 0.6217,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9648876404494382,
+      "grad_norm": 0.8737530708312988,
+      "learning_rate": 4.7061909870655734e-06,
+      "loss": 0.6618,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9652387640449438,
+      "grad_norm": 0.893994152545929,
+      "learning_rate": 4.70597349156125e-06,
+      "loss": 0.6899,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9655898876404494,
+      "grad_norm": 0.8533192276954651,
+      "learning_rate": 4.705755920614522e-06,
+      "loss": 0.601,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9659410112359551,
+      "grad_norm": 0.8612004518508911,
+      "learning_rate": 4.705538274232829e-06,
+      "loss": 0.6407,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9662921348314607,
+      "grad_norm": 0.8780084848403931,
+      "learning_rate": 4.7053205524236136e-06,
+      "loss": 0.6689,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9666432584269663,
+      "grad_norm": 0.8959993124008179,
+      "learning_rate": 4.705102755194323e-06,
+      "loss": 0.6588,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9669943820224719,
+      "grad_norm": 0.9221241474151611,
+      "learning_rate": 4.704884882552405e-06,
+      "loss": 0.6454,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9673455056179775,
+      "grad_norm": 0.8571012020111084,
+      "learning_rate": 4.70466693450531e-06,
+      "loss": 0.6101,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9676966292134831,
+      "grad_norm": 0.9270880222320557,
+      "learning_rate": 4.704448911060493e-06,
+      "loss": 0.6561,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9680477528089888,
+      "grad_norm": 0.8872186541557312,
+      "learning_rate": 4.70423081222541e-06,
+      "loss": 0.6704,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9683988764044944,
+      "grad_norm": 0.8734565377235413,
+      "learning_rate": 4.704012638007519e-06,
+      "loss": 0.6626,
+      "step": 2758
+    },
+    {
+      "epoch": 0.96875,
+      "grad_norm": 0.9009286761283875,
+      "learning_rate": 4.703794388414281e-06,
+      "loss": 0.6727,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9691011235955056,
+      "grad_norm": 0.8499953746795654,
+      "learning_rate": 4.703576063453162e-06,
+      "loss": 0.6574,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9694522471910112,
+      "grad_norm": 0.8824490308761597,
+      "learning_rate": 4.703357663131627e-06,
+      "loss": 0.6618,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9698033707865169,
+      "grad_norm": 0.8579809665679932,
+      "learning_rate": 4.703139187457145e-06,
+      "loss": 0.6677,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9701544943820225,
+      "grad_norm": 0.8739168643951416,
+      "learning_rate": 4.702920636437188e-06,
+      "loss": 0.6648,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9705056179775281,
+      "grad_norm": 0.8557081818580627,
+      "learning_rate": 4.702702010079231e-06,
+      "loss": 0.6588,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9708567415730337,
+      "grad_norm": 0.8667842149734497,
+      "learning_rate": 4.702483308390749e-06,
+      "loss": 0.6432,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9712078651685393,
+      "grad_norm": 0.8367511034011841,
+      "learning_rate": 4.7022645313792235e-06,
+      "loss": 0.6428,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9715589887640449,
+      "grad_norm": 0.8607786297798157,
+      "learning_rate": 4.702045679052136e-06,
+      "loss": 0.628,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9719101123595506,
+      "grad_norm": 0.8698323369026184,
+      "learning_rate": 4.701826751416969e-06,
+      "loss": 0.6241,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9722612359550562,
+      "grad_norm": 0.8620527386665344,
+      "learning_rate": 4.701607748481213e-06,
+      "loss": 0.6698,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9726123595505618,
+      "grad_norm": 0.8832960724830627,
+      "learning_rate": 4.701388670252355e-06,
+      "loss": 0.6332,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9729634831460674,
+      "grad_norm": 0.8846637606620789,
+      "learning_rate": 4.701169516737888e-06,
+      "loss": 0.6968,
+      "step": 2771
+    },
+    {
+      "epoch": 0.973314606741573,
+      "grad_norm": 0.8862726092338562,
+      "learning_rate": 4.7009502879453076e-06,
+      "loss": 0.7282,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9736657303370787,
+      "grad_norm": 0.8758227229118347,
+      "learning_rate": 4.70073098388211e-06,
+      "loss": 0.658,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9740168539325843,
+      "grad_norm": 0.8702247738838196,
+      "learning_rate": 4.700511604555796e-06,
+      "loss": 0.65,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9743679775280899,
+      "grad_norm": 0.8507943153381348,
+      "learning_rate": 4.700292149973869e-06,
+      "loss": 0.6625,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9747191011235955,
+      "grad_norm": 0.8701155185699463,
+      "learning_rate": 4.700072620143833e-06,
+      "loss": 0.6742,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9750702247191011,
+      "grad_norm": 0.8407930135726929,
+      "learning_rate": 4.6998530150731955e-06,
+      "loss": 0.642,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9754213483146067,
+      "grad_norm": 0.858070969581604,
+      "learning_rate": 4.6996333347694675e-06,
+      "loss": 0.6752,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9757724719101124,
+      "grad_norm": 0.8673107624053955,
+      "learning_rate": 4.6994135792401615e-06,
+      "loss": 0.6826,
+      "step": 2779
+    },
+    {
+      "epoch": 0.976123595505618,
+      "grad_norm": 0.8591476082801819,
+      "learning_rate": 4.699193748492794e-06,
+      "loss": 0.6972,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9764747191011236,
+      "grad_norm": 0.8738571405410767,
+      "learning_rate": 4.698973842534882e-06,
+      "loss": 0.6559,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9768258426966292,
+      "grad_norm": 0.8602177500724792,
+      "learning_rate": 4.698753861373946e-06,
+      "loss": 0.6433,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9771769662921348,
+      "grad_norm": 0.8604179620742798,
+      "learning_rate": 4.6985338050175096e-06,
+      "loss": 0.6499,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9775280898876404,
+      "grad_norm": 0.8430880904197693,
+      "learning_rate": 4.698313673473098e-06,
+      "loss": 0.6008,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9778792134831461,
+      "grad_norm": 0.8904216885566711,
+      "learning_rate": 4.69809346674824e-06,
+      "loss": 0.6869,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9782303370786517,
+      "grad_norm": 0.8424625396728516,
+      "learning_rate": 4.6978731848504665e-06,
+      "loss": 0.6519,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9785814606741573,
+      "grad_norm": 0.8511930704116821,
+      "learning_rate": 4.697652827787311e-06,
+      "loss": 0.6511,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9789325842696629,
+      "grad_norm": 0.8731825351715088,
+      "learning_rate": 4.697432395566309e-06,
+      "loss": 0.6756,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9792837078651685,
+      "grad_norm": 0.8926572203636169,
+      "learning_rate": 4.697211888194999e-06,
+      "loss": 0.6457,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9796348314606742,
+      "grad_norm": 0.8796569108963013,
+      "learning_rate": 4.696991305680923e-06,
+      "loss": 0.6435,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9799859550561798,
+      "grad_norm": 0.89739990234375,
+      "learning_rate": 4.6967706480316256e-06,
+      "loss": 0.666,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9803370786516854,
+      "grad_norm": 0.8666273951530457,
+      "learning_rate": 4.69654991525465e-06,
+      "loss": 0.6658,
+      "step": 2792
+    },
+    {
+      "epoch": 0.980688202247191,
+      "grad_norm": 0.8648870587348938,
+      "learning_rate": 4.696329107357548e-06,
+      "loss": 0.6541,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9810393258426966,
+      "grad_norm": 0.8609452247619629,
+      "learning_rate": 4.69610822434787e-06,
+      "loss": 0.6646,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9813904494382022,
+      "grad_norm": 0.8994244933128357,
+      "learning_rate": 4.695887266233169e-06,
+      "loss": 0.6778,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9817415730337079,
+      "grad_norm": 0.8689225316047668,
+      "learning_rate": 4.6956662330210035e-06,
+      "loss": 0.6581,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9820926966292135,
+      "grad_norm": 0.8597514629364014,
+      "learning_rate": 4.695445124718932e-06,
+      "loss": 0.6322,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9824438202247191,
+      "grad_norm": 0.8898642063140869,
+      "learning_rate": 4.695223941334514e-06,
+      "loss": 0.6866,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9827949438202247,
+      "grad_norm": 0.9043296575546265,
+      "learning_rate": 4.695002682875317e-06,
+      "loss": 0.686,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9831460674157303,
+      "grad_norm": 0.8787916898727417,
+      "learning_rate": 4.694781349348907e-06,
+      "loss": 0.6819,
+      "step": 2800
+    },
+    {
+      "epoch": 0.983497191011236,
+      "grad_norm": 0.86040198802948,
+      "learning_rate": 4.694559940762853e-06,
+      "loss": 0.6591,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9838483146067416,
+      "grad_norm": 0.8670945167541504,
+      "learning_rate": 4.694338457124726e-06,
+      "loss": 0.6828,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9841994382022472,
+      "grad_norm": 0.8818464875221252,
+      "learning_rate": 4.694116898442102e-06,
+      "loss": 0.6556,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9845505617977528,
+      "grad_norm": 0.850309431552887,
+      "learning_rate": 4.693895264722557e-06,
+      "loss": 0.6803,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9849016853932584,
+      "grad_norm": 0.8555431365966797,
+      "learning_rate": 4.6936735559736715e-06,
+      "loss": 0.6484,
+      "step": 2805
+    },
+    {
+      "epoch": 0.985252808988764,
+      "grad_norm": 0.8949291110038757,
+      "learning_rate": 4.693451772203028e-06,
+      "loss": 0.6889,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9856039325842697,
+      "grad_norm": 0.8468831181526184,
+      "learning_rate": 4.6932299134182104e-06,
+      "loss": 0.6398,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9859550561797753,
+      "grad_norm": 0.8642711043357849,
+      "learning_rate": 4.693007979626807e-06,
+      "loss": 0.6789,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9863061797752809,
+      "grad_norm": 0.8652757406234741,
+      "learning_rate": 4.692785970836407e-06,
+      "loss": 0.6862,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9866573033707865,
+      "grad_norm": 0.8652501106262207,
+      "learning_rate": 4.692563887054603e-06,
+      "loss": 0.647,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9870084269662921,
+      "grad_norm": 0.8379038572311401,
+      "learning_rate": 4.69234172828899e-06,
+      "loss": 0.616,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9873595505617978,
+      "grad_norm": 0.8986461758613586,
+      "learning_rate": 4.692119494547167e-06,
+      "loss": 0.6746,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9877106741573034,
+      "grad_norm": 0.8548451662063599,
+      "learning_rate": 4.691897185836733e-06,
+      "loss": 0.6607,
+      "step": 2813
+    },
+    {
+      "epoch": 0.988061797752809,
+      "grad_norm": 0.8519273400306702,
+      "learning_rate": 4.69167480216529e-06,
+      "loss": 0.6341,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9884129213483146,
+      "grad_norm": 0.8934301137924194,
+      "learning_rate": 4.691452343540445e-06,
+      "loss": 0.6601,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "grad_norm": 0.8649933338165283,
+      "learning_rate": 4.6912298099698046e-06,
+      "loss": 0.6406,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9891151685393258,
+      "grad_norm": 0.8976009488105774,
+      "learning_rate": 4.69100720146098e-06,
+      "loss": 0.6391,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9894662921348315,
+      "grad_norm": 0.8619791865348816,
+      "learning_rate": 4.690784518021584e-06,
+      "loss": 0.6932,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9898174157303371,
+      "grad_norm": 0.8594460487365723,
+      "learning_rate": 4.690561759659232e-06,
+      "loss": 0.6666,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9901685393258427,
+      "grad_norm": 0.8755744099617004,
+      "learning_rate": 4.690338926381542e-06,
+      "loss": 0.6458,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9905196629213483,
+      "grad_norm": 0.8844261765480042,
+      "learning_rate": 4.690116018196135e-06,
+      "loss": 0.6344,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9908707865168539,
+      "grad_norm": 0.871732234954834,
+      "learning_rate": 4.689893035110635e-06,
+      "loss": 0.6316,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9912219101123596,
+      "grad_norm": 0.8930003046989441,
+      "learning_rate": 4.689669977132667e-06,
+      "loss": 0.6887,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9915730337078652,
+      "grad_norm": 0.9592486619949341,
+      "learning_rate": 4.689446844269859e-06,
+      "loss": 0.653,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9919241573033708,
+      "grad_norm": 0.8865694999694824,
+      "learning_rate": 4.6892236365298425e-06,
+      "loss": 0.648,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9922752808988764,
+      "grad_norm": 0.8786531686782837,
+      "learning_rate": 4.689000353920251e-06,
+      "loss": 0.647,
+      "step": 2826
+    },
+    {
+      "epoch": 0.992626404494382,
+      "grad_norm": 0.8980904817581177,
+      "learning_rate": 4.6887769964487215e-06,
+      "loss": 0.6556,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9929775280898876,
+      "grad_norm": 0.8920478224754333,
+      "learning_rate": 4.688553564122891e-06,
+      "loss": 0.6802,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9933286516853933,
+      "grad_norm": 0.8804338574409485,
+      "learning_rate": 4.688330056950401e-06,
+      "loss": 0.6695,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9936797752808989,
+      "grad_norm": 0.8846019506454468,
+      "learning_rate": 4.688106474938896e-06,
+      "loss": 0.6484,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9940308988764045,
+      "grad_norm": 0.8795493841171265,
+      "learning_rate": 4.687882818096022e-06,
+      "loss": 0.6465,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9943820224719101,
+      "grad_norm": 0.858067512512207,
+      "learning_rate": 4.687659086429427e-06,
+      "loss": 0.6471,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9947331460674157,
+      "grad_norm": 0.8714587092399597,
+      "learning_rate": 4.687435279946765e-06,
+      "loss": 0.6446,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9950842696629213,
+      "grad_norm": 0.8674041628837585,
+      "learning_rate": 4.687211398655686e-06,
+      "loss": 0.6419,
+      "step": 2834
+    },
+    {
+      "epoch": 0.995435393258427,
+      "grad_norm": 0.8507161736488342,
+      "learning_rate": 4.6869874425638505e-06,
+      "loss": 0.6202,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9957865168539326,
+      "grad_norm": 0.8718125224113464,
+      "learning_rate": 4.686763411678915e-06,
+      "loss": 0.6807,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9961376404494382,
+      "grad_norm": 0.8641071319580078,
+      "learning_rate": 4.686539306008543e-06,
+      "loss": 0.6492,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9964887640449438,
+      "grad_norm": 0.8599060773849487,
+      "learning_rate": 4.686315125560397e-06,
+      "loss": 0.6779,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9968398876404494,
+      "grad_norm": 0.8444222211837769,
+      "learning_rate": 4.686090870342145e-06,
+      "loss": 0.6841,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9971910112359551,
+      "grad_norm": 0.86264568567276,
+      "learning_rate": 4.685866540361456e-06,
+      "loss": 0.6627,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9975421348314607,
+      "grad_norm": 0.9181270599365234,
+      "learning_rate": 4.685642135626001e-06,
+      "loss": 0.6537,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9978932584269663,
+      "grad_norm": 0.8701741695404053,
+      "learning_rate": 4.685417656143456e-06,
+      "loss": 0.6545,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9982443820224719,
+      "grad_norm": 0.8788411021232605,
+      "learning_rate": 4.685193101921497e-06,
+      "loss": 0.6431,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9985955056179775,
+      "grad_norm": 0.8748607635498047,
+      "learning_rate": 4.684968472967804e-06,
+      "loss": 0.6698,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9989466292134831,
+      "grad_norm": 0.9054082632064819,
+      "learning_rate": 4.684743769290059e-06,
+      "loss": 0.6801,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9992977528089888,
+      "grad_norm": 0.8799982666969299,
+      "learning_rate": 4.684518990895945e-06,
+      "loss": 0.6764,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9996488764044944,
+      "grad_norm": 0.9407274127006531,
+      "learning_rate": 4.684294137793153e-06,
+      "loss": 0.7098,
+      "step": 2847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.91474848985672,
+      "learning_rate": 4.684069209989369e-06,
+      "loss": 0.6617,
+      "step": 2848
+    },
+    {
+      "epoch": 1.0003511235955056,
+      "grad_norm": 0.845862090587616,
+      "learning_rate": 4.683844207492289e-06,
+      "loss": 0.6516,
+      "step": 2849
+    },
+    {
+      "epoch": 1.0007022471910112,
+      "grad_norm": 0.8708998560905457,
+      "learning_rate": 4.683619130309604e-06,
+      "loss": 0.6668,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0010533707865168,
+      "grad_norm": 0.8879108428955078,
+      "learning_rate": 4.683393978449014e-06,
+      "loss": 0.6155,
+      "step": 2851
+    },
+    {
+      "epoch": 1.0014044943820224,
+      "grad_norm": 0.8889144062995911,
+      "learning_rate": 4.683168751918218e-06,
+      "loss": 0.6625,
+      "step": 2852
+    },
+    {
+      "epoch": 1.001755617977528,
+      "grad_norm": 0.8704389929771423,
+      "learning_rate": 4.682943450724919e-06,
+      "loss": 0.6769,
+      "step": 2853
+    },
+    {
+      "epoch": 1.0021067415730338,
+      "grad_norm": 0.8852386474609375,
+      "learning_rate": 4.682718074876822e-06,
+      "loss": 0.6532,
+      "step": 2854
+    },
+    {
+      "epoch": 1.0024578651685394,
+      "grad_norm": 0.9099629521369934,
+      "learning_rate": 4.682492624381635e-06,
+      "loss": 0.6217,
+      "step": 2855
+    },
+    {
+      "epoch": 1.002808988764045,
+      "grad_norm": 0.9146444201469421,
+      "learning_rate": 4.682267099247066e-06,
+      "loss": 0.6483,
+      "step": 2856
+    },
+    {
+      "epoch": 1.0031601123595506,
+      "grad_norm": 0.9506009221076965,
+      "learning_rate": 4.682041499480832e-06,
+      "loss": 0.6958,
+      "step": 2857
+    },
+    {
+      "epoch": 1.0035112359550562,
+      "grad_norm": 0.9254932999610901,
+      "learning_rate": 4.681815825090645e-06,
+      "loss": 0.6637,
+      "step": 2858
+    },
+    {
+      "epoch": 1.0038623595505618,
+      "grad_norm": 0.9010646939277649,
+      "learning_rate": 4.681590076084224e-06,
+      "loss": 0.6594,
+      "step": 2859
+    },
+    {
+      "epoch": 1.0042134831460674,
+      "grad_norm": 0.8690000772476196,
+      "learning_rate": 4.6813642524692886e-06,
+      "loss": 0.6238,
+      "step": 2860
+    },
+    {
+      "epoch": 1.004564606741573,
+      "grad_norm": 0.8916565179824829,
+      "learning_rate": 4.681138354253563e-06,
+      "loss": 0.6279,
+      "step": 2861
+    },
+    {
+      "epoch": 1.0049157303370786,
+      "grad_norm": 0.9105271697044373,
+      "learning_rate": 4.680912381444771e-06,
+      "loss": 0.6636,
+      "step": 2862
+    },
+    {
+      "epoch": 1.0052668539325842,
+      "grad_norm": 0.9027879238128662,
+      "learning_rate": 4.680686334050643e-06,
+      "loss": 0.6818,
+      "step": 2863
+    },
+    {
+      "epoch": 1.0056179775280898,
+      "grad_norm": 0.8352002501487732,
+      "learning_rate": 4.6804602120789075e-06,
+      "loss": 0.6071,
+      "step": 2864
+    },
+    {
+      "epoch": 1.0059691011235956,
+      "grad_norm": 0.9015585780143738,
+      "learning_rate": 4.680234015537299e-06,
+      "loss": 0.6091,
+      "step": 2865
+    },
+    {
+      "epoch": 1.0063202247191012,
+      "grad_norm": 0.8762530088424683,
+      "learning_rate": 4.680007744433553e-06,
+      "loss": 0.6307,
+      "step": 2866
+    },
+    {
+      "epoch": 1.0066713483146068,
+      "grad_norm": 0.9097511768341064,
+      "learning_rate": 4.679781398775408e-06,
+      "loss": 0.6567,
+      "step": 2867
+    },
+    {
+      "epoch": 1.0070224719101124,
+      "grad_norm": 0.9116111993789673,
+      "learning_rate": 4.679554978570604e-06,
+      "loss": 0.6421,
+      "step": 2868
+    },
+    {
+      "epoch": 1.007373595505618,
+      "grad_norm": 0.8672003746032715,
+      "learning_rate": 4.679328483826885e-06,
+      "loss": 0.6465,
+      "step": 2869
+    },
+    {
+      "epoch": 1.0077247191011236,
+      "grad_norm": 0.9330756068229675,
+      "learning_rate": 4.679101914551997e-06,
+      "loss": 0.6187,
+      "step": 2870
+    },
+    {
+      "epoch": 1.0080758426966292,
+      "grad_norm": 0.8700326085090637,
+      "learning_rate": 4.678875270753687e-06,
+      "loss": 0.6089,
+      "step": 2871
+    },
+    {
+      "epoch": 1.0084269662921348,
+      "grad_norm": 0.9126592874526978,
+      "learning_rate": 4.678648552439708e-06,
+      "loss": 0.6386,
+      "step": 2872
+    },
+    {
+      "epoch": 1.0087780898876404,
+      "grad_norm": 0.9096201658248901,
+      "learning_rate": 4.678421759617813e-06,
+      "loss": 0.6059,
+      "step": 2873
+    },
+    {
+      "epoch": 1.009129213483146,
+      "grad_norm": 0.8730313777923584,
+      "learning_rate": 4.678194892295758e-06,
+      "loss": 0.6228,
+      "step": 2874
+    },
+    {
+      "epoch": 1.0094803370786516,
+      "grad_norm": 0.9104415774345398,
+      "learning_rate": 4.6779679504813e-06,
+      "loss": 0.6251,
+      "step": 2875
+    },
+    {
+      "epoch": 1.0098314606741574,
+      "grad_norm": 0.8940228819847107,
+      "learning_rate": 4.677740934182204e-06,
+      "loss": 0.6705,
+      "step": 2876
+    },
+    {
+      "epoch": 1.010182584269663,
+      "grad_norm": 0.8782382607460022,
+      "learning_rate": 4.67751384340623e-06,
+      "loss": 0.6203,
+      "step": 2877
+    },
+    {
+      "epoch": 1.0105337078651686,
+      "grad_norm": 0.8854807615280151,
+      "learning_rate": 4.6772866781611474e-06,
+      "loss": 0.6414,
+      "step": 2878
+    },
+    {
+      "epoch": 1.0108848314606742,
+      "grad_norm": 0.8533816933631897,
+      "learning_rate": 4.677059438454722e-06,
+      "loss": 0.6019,
+      "step": 2879
+    },
+    {
+      "epoch": 1.0112359550561798,
+      "grad_norm": 0.8798636198043823,
+      "learning_rate": 4.6768321242947276e-06,
+      "loss": 0.643,
+      "step": 2880
+    },
+    {
+      "epoch": 1.0115870786516854,
+      "grad_norm": 0.8717870712280273,
+      "learning_rate": 4.676604735688937e-06,
+      "loss": 0.6125,
+      "step": 2881
+    },
+    {
+      "epoch": 1.011938202247191,
+      "grad_norm": 0.8863524198532104,
+      "learning_rate": 4.676377272645127e-06,
+      "loss": 0.6475,
+      "step": 2882
+    },
+    {
+      "epoch": 1.0122893258426966,
+      "grad_norm": 0.9243369698524475,
+      "learning_rate": 4.6761497351710775e-06,
+      "loss": 0.654,
+      "step": 2883
+    },
+    {
+      "epoch": 1.0126404494382022,
+      "grad_norm": 0.9152374267578125,
+      "learning_rate": 4.675922123274568e-06,
+      "loss": 0.685,
+      "step": 2884
+    },
+    {
+      "epoch": 1.0129915730337078,
+      "grad_norm": 0.8751019835472107,
+      "learning_rate": 4.675694436963384e-06,
+      "loss": 0.6177,
+      "step": 2885
+    },
+    {
+      "epoch": 1.0133426966292134,
+      "grad_norm": 0.9058424830436707,
+      "learning_rate": 4.675466676245313e-06,
+      "loss": 0.6462,
+      "step": 2886
+    },
+    {
+      "epoch": 1.0136938202247192,
+      "grad_norm": 0.8993322849273682,
+      "learning_rate": 4.675238841128142e-06,
+      "loss": 0.6424,
+      "step": 2887
+    },
+    {
+      "epoch": 1.0140449438202248,
+      "grad_norm": 0.9268403053283691,
+      "learning_rate": 4.675010931619665e-06,
+      "loss": 0.6514,
+      "step": 2888
+    },
+    {
+      "epoch": 1.0143960674157304,
+      "grad_norm": 0.9016578197479248,
+      "learning_rate": 4.674782947727674e-06,
+      "loss": 0.6512,
+      "step": 2889
+    },
+    {
+      "epoch": 1.014747191011236,
+      "grad_norm": 0.8681126236915588,
+      "learning_rate": 4.674554889459969e-06,
+      "loss": 0.6703,
+      "step": 2890
+    },
+    {
+      "epoch": 1.0150983146067416,
+      "grad_norm": 0.8943293690681458,
+      "learning_rate": 4.6743267568243455e-06,
+      "loss": 0.6363,
+      "step": 2891
+    },
+    {
+      "epoch": 1.0154494382022472,
+      "grad_norm": 0.9124616384506226,
+      "learning_rate": 4.674098549828609e-06,
+      "loss": 0.6177,
+      "step": 2892
+    },
+    {
+      "epoch": 1.0158005617977528,
+      "grad_norm": 0.8788502216339111,
+      "learning_rate": 4.673870268480563e-06,
+      "loss": 0.6459,
+      "step": 2893
+    },
+    {
+      "epoch": 1.0161516853932584,
+      "grad_norm": 0.8843396306037903,
+      "learning_rate": 4.6736419127880126e-06,
+      "loss": 0.6513,
+      "step": 2894
+    },
+    {
+      "epoch": 1.016502808988764,
+      "grad_norm": 0.8527891039848328,
+      "learning_rate": 4.673413482758769e-06,
+      "loss": 0.6349,
+      "step": 2895
+    },
+    {
+      "epoch": 1.0168539325842696,
+      "grad_norm": 0.8894069194793701,
+      "learning_rate": 4.673184978400645e-06,
+      "loss": 0.627,
+      "step": 2896
+    },
+    {
+      "epoch": 1.0172050561797752,
+      "grad_norm": 0.913414478302002,
+      "learning_rate": 4.6729563997214534e-06,
+      "loss": 0.6433,
+      "step": 2897
+    },
+    {
+      "epoch": 1.0175561797752808,
+      "grad_norm": 0.9203279614448547,
+      "learning_rate": 4.672727746729013e-06,
+      "loss": 0.642,
+      "step": 2898
+    },
+    {
+      "epoch": 1.0179073033707866,
+      "grad_norm": 0.9143649935722351,
+      "learning_rate": 4.672499019431143e-06,
+      "loss": 0.6529,
+      "step": 2899
+    },
+    {
+      "epoch": 1.0182584269662922,
+      "grad_norm": 0.9485595226287842,
+      "learning_rate": 4.672270217835665e-06,
+      "loss": 0.6509,
+      "step": 2900
+    },
+    {
+      "epoch": 1.0186095505617978,
+      "grad_norm": 0.9203194379806519,
+      "learning_rate": 4.672041341950404e-06,
+      "loss": 0.69,
+      "step": 2901
+    },
+    {
+      "epoch": 1.0189606741573034,
+      "grad_norm": 0.9102271199226379,
+      "learning_rate": 4.671812391783188e-06,
+      "loss": 0.6387,
+      "step": 2902
+    },
+    {
+      "epoch": 1.019311797752809,
+      "grad_norm": 0.8916968107223511,
+      "learning_rate": 4.671583367341848e-06,
+      "loss": 0.6411,
+      "step": 2903
+    },
+    {
+      "epoch": 1.0196629213483146,
+      "grad_norm": 0.9091634750366211,
+      "learning_rate": 4.671354268634214e-06,
+      "loss": 0.6601,
+      "step": 2904
+    },
+    {
+      "epoch": 1.0200140449438202,
+      "grad_norm": 0.8721140027046204,
+      "learning_rate": 4.671125095668122e-06,
+      "loss": 0.6299,
+      "step": 2905
+    },
+    {
+      "epoch": 1.0203651685393258,
+      "grad_norm": 0.8712757229804993,
+      "learning_rate": 4.67089584845141e-06,
+      "loss": 0.6649,
+      "step": 2906
+    },
+    {
+      "epoch": 1.0207162921348314,
+      "grad_norm": 0.9056061506271362,
+      "learning_rate": 4.670666526991917e-06,
+      "loss": 0.5834,
+      "step": 2907
+    },
+    {
+      "epoch": 1.021067415730337,
+      "grad_norm": 0.9042145013809204,
+      "learning_rate": 4.670437131297487e-06,
+      "loss": 0.6487,
+      "step": 2908
+    },
+    {
+      "epoch": 1.0214185393258426,
+      "grad_norm": 0.914989173412323,
+      "learning_rate": 4.6702076613759635e-06,
+      "loss": 0.6595,
+      "step": 2909
+    },
+    {
+      "epoch": 1.0217696629213484,
+      "grad_norm": 0.850035548210144,
+      "learning_rate": 4.669978117235195e-06,
+      "loss": 0.6253,
+      "step": 2910
+    },
+    {
+      "epoch": 1.022120786516854,
+      "grad_norm": 0.9134849309921265,
+      "learning_rate": 4.6697484988830334e-06,
+      "loss": 0.6486,
+      "step": 2911
+    },
+    {
+      "epoch": 1.0224719101123596,
+      "grad_norm": 0.8706740736961365,
+      "learning_rate": 4.669518806327328e-06,
+      "loss": 0.6734,
+      "step": 2912
+    },
+    {
+      "epoch": 1.0228230337078652,
+      "grad_norm": 0.8819549083709717,
+      "learning_rate": 4.669289039575937e-06,
+      "loss": 0.6034,
+      "step": 2913
+    },
+    {
+      "epoch": 1.0231741573033708,
+      "grad_norm": 0.8862223625183105,
+      "learning_rate": 4.6690591986367175e-06,
+      "loss": 0.637,
+      "step": 2914
+    },
+    {
+      "epoch": 1.0235252808988764,
+      "grad_norm": 0.895828127861023,
+      "learning_rate": 4.668829283517528e-06,
+      "loss": 0.6243,
+      "step": 2915
+    },
+    {
+      "epoch": 1.023876404494382,
+      "grad_norm": 0.9082605838775635,
+      "learning_rate": 4.668599294226234e-06,
+      "loss": 0.6333,
+      "step": 2916
+    },
+    {
+      "epoch": 1.0242275280898876,
+      "grad_norm": 0.9978623390197754,
+      "learning_rate": 4.6683692307707e-06,
+      "loss": 0.6772,
+      "step": 2917
+    },
+    {
+      "epoch": 1.0245786516853932,
+      "grad_norm": 0.8527755737304688,
+      "learning_rate": 4.668139093158793e-06,
+      "loss": 0.6118,
+      "step": 2918
+    },
+    {
+      "epoch": 1.0249297752808988,
+      "grad_norm": 0.8660287857055664,
+      "learning_rate": 4.667908881398384e-06,
+      "loss": 0.6458,
+      "step": 2919
+    },
+    {
+      "epoch": 1.0252808988764044,
+      "grad_norm": 0.8769094347953796,
+      "learning_rate": 4.667678595497348e-06,
+      "loss": 0.6458,
+      "step": 2920
+    },
+    {
+      "epoch": 1.0256320224719102,
+      "grad_norm": 0.8812835812568665,
+      "learning_rate": 4.6674482354635576e-06,
+      "loss": 0.6761,
+      "step": 2921
+    },
+    {
+      "epoch": 1.0259831460674158,
+      "grad_norm": 0.9102697372436523,
+      "learning_rate": 4.6672178013048925e-06,
+      "loss": 0.62,
+      "step": 2922
+    },
+    {
+      "epoch": 1.0263342696629214,
+      "grad_norm": 0.8357862234115601,
+      "learning_rate": 4.6669872930292336e-06,
+      "loss": 0.5875,
+      "step": 2923
+    },
+    {
+      "epoch": 1.026685393258427,
+      "grad_norm": 0.8696123361587524,
+      "learning_rate": 4.666756710644463e-06,
+      "loss": 0.6224,
+      "step": 2924
+    },
+    {
+      "epoch": 1.0270365168539326,
+      "grad_norm": 0.9066466689109802,
+      "learning_rate": 4.666526054158467e-06,
+      "loss": 0.6789,
+      "step": 2925
+    },
+    {
+      "epoch": 1.0273876404494382,
+      "grad_norm": 0.8718449473381042,
+      "learning_rate": 4.666295323579134e-06,
+      "loss": 0.633,
+      "step": 2926
+    },
+    {
+      "epoch": 1.0277387640449438,
+      "grad_norm": 0.9176349639892578,
+      "learning_rate": 4.666064518914355e-06,
+      "loss": 0.6343,
+      "step": 2927
+    },
+    {
+      "epoch": 1.0280898876404494,
+      "grad_norm": 0.8917323350906372,
+      "learning_rate": 4.6658336401720226e-06,
+      "loss": 0.6122,
+      "step": 2928
+    },
+    {
+      "epoch": 1.028441011235955,
+      "grad_norm": 0.884037971496582,
+      "learning_rate": 4.665602687360034e-06,
+      "loss": 0.6387,
+      "step": 2929
+    },
+    {
+      "epoch": 1.0287921348314606,
+      "grad_norm": 0.901550829410553,
+      "learning_rate": 4.665371660486285e-06,
+      "loss": 0.6496,
+      "step": 2930
+    },
+    {
+      "epoch": 1.0291432584269662,
+      "grad_norm": 0.8866931200027466,
+      "learning_rate": 4.66514055955868e-06,
+      "loss": 0.6266,
+      "step": 2931
+    },
+    {
+      "epoch": 1.029494382022472,
+      "grad_norm": 0.8960362672805786,
+      "learning_rate": 4.664909384585119e-06,
+      "loss": 0.6186,
+      "step": 2932
+    },
+    {
+      "epoch": 1.0298455056179776,
+      "grad_norm": 0.8991187810897827,
+      "learning_rate": 4.66467813557351e-06,
+      "loss": 0.6298,
+      "step": 2933
+    },
+    {
+      "epoch": 1.0301966292134832,
+      "grad_norm": 0.869250476360321,
+      "learning_rate": 4.664446812531761e-06,
+      "loss": 0.5864,
+      "step": 2934
+    },
+    {
+      "epoch": 1.0305477528089888,
+      "grad_norm": 0.8768630027770996,
+      "learning_rate": 4.664215415467784e-06,
+      "loss": 0.6434,
+      "step": 2935
+    },
+    {
+      "epoch": 1.0308988764044944,
+      "grad_norm": 0.9055302143096924,
+      "learning_rate": 4.663983944389491e-06,
+      "loss": 0.6467,
+      "step": 2936
+    },
+    {
+      "epoch": 1.03125,
+      "grad_norm": 0.9194045066833496,
+      "learning_rate": 4.663752399304799e-06,
+      "loss": 0.6136,
+      "step": 2937
+    },
+    {
+      "epoch": 1.0316011235955056,
+      "grad_norm": 0.8887422680854797,
+      "learning_rate": 4.663520780221626e-06,
+      "loss": 0.6614,
+      "step": 2938
+    },
+    {
+      "epoch": 1.0319522471910112,
+      "grad_norm": 0.9037759304046631,
+      "learning_rate": 4.663289087147894e-06,
+      "loss": 0.6892,
+      "step": 2939
+    },
+    {
+      "epoch": 1.0323033707865168,
+      "grad_norm": 0.8857134580612183,
+      "learning_rate": 4.6630573200915275e-06,
+      "loss": 0.6227,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0326544943820224,
+      "grad_norm": 0.8977053165435791,
+      "learning_rate": 4.66282547906045e-06,
+      "loss": 0.6445,
+      "step": 2941
+    },
+    {
+      "epoch": 1.033005617977528,
+      "grad_norm": 0.9049233794212341,
+      "learning_rate": 4.662593564062593e-06,
+      "loss": 0.6139,
+      "step": 2942
+    },
+    {
+      "epoch": 1.0333567415730338,
+      "grad_norm": 0.894765317440033,
+      "learning_rate": 4.6623615751058856e-06,
+      "loss": 0.6186,
+      "step": 2943
+    },
+    {
+      "epoch": 1.0337078651685394,
+      "grad_norm": 0.9289073944091797,
+      "learning_rate": 4.6621295121982634e-06,
+      "loss": 0.6496,
+      "step": 2944
+    },
+    {
+      "epoch": 1.034058988764045,
+      "grad_norm": 0.8969010710716248,
+      "learning_rate": 4.661897375347662e-06,
+      "loss": 0.6634,
+      "step": 2945
+    },
+    {
+      "epoch": 1.0344101123595506,
+      "grad_norm": 0.9372804164886475,
+      "learning_rate": 4.66166516456202e-06,
+      "loss": 0.6445,
+      "step": 2946
+    },
+    {
+      "epoch": 1.0347612359550562,
+      "grad_norm": 0.9191506505012512,
+      "learning_rate": 4.66143287984928e-06,
+      "loss": 0.6105,
+      "step": 2947
+    },
+    {
+      "epoch": 1.0351123595505618,
+      "grad_norm": 0.9287760853767395,
+      "learning_rate": 4.661200521217384e-06,
+      "loss": 0.6174,
+      "step": 2948
+    },
+    {
+      "epoch": 1.0354634831460674,
+      "grad_norm": 0.8911864757537842,
+      "learning_rate": 4.66096808867428e-06,
+      "loss": 0.6857,
+      "step": 2949
+    },
+    {
+      "epoch": 1.035814606741573,
+      "grad_norm": 0.8997723460197449,
+      "learning_rate": 4.660735582227917e-06,
+      "loss": 0.6199,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0361657303370786,
+      "grad_norm": 0.9067447185516357,
+      "learning_rate": 4.660503001886245e-06,
+      "loss": 0.684,
+      "step": 2951
+    },
+    {
+      "epoch": 1.0365168539325842,
+      "grad_norm": 0.9156118631362915,
+      "learning_rate": 4.66027034765722e-06,
+      "loss": 0.6705,
+      "step": 2952
+    },
+    {
+      "epoch": 1.0368679775280898,
+      "grad_norm": 0.8432906866073608,
+      "learning_rate": 4.660037619548797e-06,
+      "loss": 0.6146,
+      "step": 2953
+    },
+    {
+      "epoch": 1.0372191011235956,
+      "grad_norm": 0.8450725674629211,
+      "learning_rate": 4.659804817568937e-06,
+      "loss": 0.6166,
+      "step": 2954
+    },
+    {
+      "epoch": 1.0375702247191012,
+      "grad_norm": 0.854310929775238,
+      "learning_rate": 4.659571941725599e-06,
+      "loss": 0.6527,
+      "step": 2955
+    },
+    {
+      "epoch": 1.0379213483146068,
+      "grad_norm": 0.8777969479560852,
+      "learning_rate": 4.659338992026749e-06,
+      "loss": 0.6587,
+      "step": 2956
+    },
+    {
+      "epoch": 1.0382724719101124,
+      "grad_norm": 0.8768917918205261,
+      "learning_rate": 4.659105968480353e-06,
+      "loss": 0.6239,
+      "step": 2957
+    },
+    {
+      "epoch": 1.038623595505618,
+      "grad_norm": 0.8562551736831665,
+      "learning_rate": 4.658872871094381e-06,
+      "loss": 0.6188,
+      "step": 2958
+    },
+    {
+      "epoch": 1.0389747191011236,
+      "grad_norm": 0.8880004286766052,
+      "learning_rate": 4.658639699876804e-06,
+      "loss": 0.6379,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0393258426966292,
+      "grad_norm": 0.8730093240737915,
+      "learning_rate": 4.658406454835596e-06,
+      "loss": 0.6575,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0396769662921348,
+      "grad_norm": 0.8812423944473267,
+      "learning_rate": 4.658173135978735e-06,
+      "loss": 0.6451,
+      "step": 2961
+    },
+    {
+      "epoch": 1.0400280898876404,
+      "grad_norm": 0.8783746957778931,
+      "learning_rate": 4.657939743314199e-06,
+      "loss": 0.6775,
+      "step": 2962
+    },
+    {
+      "epoch": 1.040379213483146,
+      "grad_norm": 0.8458737134933472,
+      "learning_rate": 4.657706276849971e-06,
+      "loss": 0.644,
+      "step": 2963
+    },
+    {
+      "epoch": 1.0407303370786516,
+      "grad_norm": 0.8714883923530579,
+      "learning_rate": 4.6574727365940344e-06,
+      "loss": 0.6193,
+      "step": 2964
+    },
+    {
+      "epoch": 1.0410814606741574,
+      "grad_norm": 0.8596876859664917,
+      "learning_rate": 4.657239122554376e-06,
+      "loss": 0.6392,
+      "step": 2965
+    },
+    {
+      "epoch": 1.041432584269663,
+      "grad_norm": 0.831123948097229,
+      "learning_rate": 4.657005434738986e-06,
+      "loss": 0.6212,
+      "step": 2966
+    },
+    {
+      "epoch": 1.0417837078651686,
+      "grad_norm": 0.8762812614440918,
+      "learning_rate": 4.656771673155856e-06,
+      "loss": 0.6422,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0421348314606742,
+      "grad_norm": 0.8940248489379883,
+      "learning_rate": 4.65653783781298e-06,
+      "loss": 0.5996,
+      "step": 2968
+    },
+    {
+      "epoch": 1.0424859550561798,
+      "grad_norm": 0.9237715601921082,
+      "learning_rate": 4.6563039287183556e-06,
+      "loss": 0.6476,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0428370786516854,
+      "grad_norm": 0.8804140686988831,
+      "learning_rate": 4.656069945879982e-06,
+      "loss": 0.6311,
+      "step": 2970
+    },
+    {
+      "epoch": 1.043188202247191,
+      "grad_norm": 0.8878976702690125,
+      "learning_rate": 4.655835889305861e-06,
+      "loss": 0.637,
+      "step": 2971
+    },
+    {
+      "epoch": 1.0435393258426966,
+      "grad_norm": 0.9103527665138245,
+      "learning_rate": 4.655601759003997e-06,
+      "loss": 0.611,
+      "step": 2972
+    },
+    {
+      "epoch": 1.0438904494382022,
+      "grad_norm": 0.8697876334190369,
+      "learning_rate": 4.6553675549823975e-06,
+      "loss": 0.6484,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0442415730337078,
+      "grad_norm": 0.8702844381332397,
+      "learning_rate": 4.655133277249073e-06,
+      "loss": 0.6206,
+      "step": 2974
+    },
+    {
+      "epoch": 1.0445926966292134,
+      "grad_norm": 0.8851619362831116,
+      "learning_rate": 4.654898925812034e-06,
+      "loss": 0.6368,
+      "step": 2975
+    },
+    {
+      "epoch": 1.0449438202247192,
+      "grad_norm": 0.9111335873603821,
+      "learning_rate": 4.654664500679295e-06,
+      "loss": 0.6178,
+      "step": 2976
+    },
+    {
+      "epoch": 1.0452949438202248,
+      "grad_norm": 0.8853289484977722,
+      "learning_rate": 4.654430001858874e-06,
+      "loss": 0.6461,
+      "step": 2977
+    },
+    {
+      "epoch": 1.0456460674157304,
+      "grad_norm": 0.8976671099662781,
+      "learning_rate": 4.654195429358792e-06,
+      "loss": 0.6459,
+      "step": 2978
+    },
+    {
+      "epoch": 1.045997191011236,
+      "grad_norm": 0.8783121705055237,
+      "learning_rate": 4.6539607831870685e-06,
+      "loss": 0.695,
+      "step": 2979
+    },
+    {
+      "epoch": 1.0463483146067416,
+      "grad_norm": 0.8905385732650757,
+      "learning_rate": 4.653726063351729e-06,
+      "loss": 0.6458,
+      "step": 2980
+    },
+    {
+      "epoch": 1.0466994382022472,
+      "grad_norm": 0.89134681224823,
+      "learning_rate": 4.6534912698608015e-06,
+      "loss": 0.6405,
+      "step": 2981
+    },
+    {
+      "epoch": 1.0470505617977528,
+      "grad_norm": 0.8397007584571838,
+      "learning_rate": 4.653256402722316e-06,
+      "loss": 0.6003,
+      "step": 2982
+    },
+    {
+      "epoch": 1.0474016853932584,
+      "grad_norm": 0.8812448978424072,
+      "learning_rate": 4.653021461944304e-06,
+      "loss": 0.6249,
+      "step": 2983
+    },
+    {
+      "epoch": 1.047752808988764,
+      "grad_norm": 0.8805206418037415,
+      "learning_rate": 4.652786447534799e-06,
+      "loss": 0.6545,
+      "step": 2984
+    },
+    {
+      "epoch": 1.0481039325842696,
+      "grad_norm": 0.8674848675727844,
+      "learning_rate": 4.652551359501841e-06,
+      "loss": 0.615,
+      "step": 2985
+    },
+    {
+      "epoch": 1.0484550561797752,
+      "grad_norm": 0.8602030277252197,
+      "learning_rate": 4.652316197853468e-06,
+      "loss": 0.6166,
+      "step": 2986
+    },
+    {
+      "epoch": 1.0488061797752808,
+      "grad_norm": 0.866513192653656,
+      "learning_rate": 4.6520809625977224e-06,
+      "loss": 0.6312,
+      "step": 2987
+    },
+    {
+      "epoch": 1.0491573033707866,
+      "grad_norm": 0.8837951421737671,
+      "learning_rate": 4.65184565374265e-06,
+      "loss": 0.627,
+      "step": 2988
+    },
+    {
+      "epoch": 1.0495084269662922,
+      "grad_norm": 0.8847421407699585,
+      "learning_rate": 4.651610271296298e-06,
+      "loss": 0.617,
+      "step": 2989
+    },
+    {
+      "epoch": 1.0498595505617978,
+      "grad_norm": 0.8955681324005127,
+      "learning_rate": 4.651374815266715e-06,
+      "loss": 0.6153,
+      "step": 2990
+    },
+    {
+      "epoch": 1.0502106741573034,
+      "grad_norm": 0.8633450865745544,
+      "learning_rate": 4.651139285661955e-06,
+      "loss": 0.6217,
+      "step": 2991
+    },
+    {
+      "epoch": 1.050561797752809,
+      "grad_norm": 0.886147677898407,
+      "learning_rate": 4.650903682490072e-06,
+      "loss": 0.5844,
+      "step": 2992
+    },
+    {
+      "epoch": 1.0509129213483146,
+      "grad_norm": 0.9161800146102905,
+      "learning_rate": 4.650668005759124e-06,
+      "loss": 0.6577,
+      "step": 2993
+    },
+    {
+      "epoch": 1.0512640449438202,
+      "grad_norm": 0.9195558428764343,
+      "learning_rate": 4.6504322554771694e-06,
+      "loss": 0.5948,
+      "step": 2994
+    },
+    {
+      "epoch": 1.0516151685393258,
+      "grad_norm": 0.8970798850059509,
+      "learning_rate": 4.650196431652272e-06,
+      "loss": 0.6012,
+      "step": 2995
+    },
+    {
+      "epoch": 1.0519662921348314,
+      "grad_norm": 0.9370929002761841,
+      "learning_rate": 4.649960534292497e-06,
+      "loss": 0.6284,
+      "step": 2996
+    },
+    {
+      "epoch": 1.052317415730337,
+      "grad_norm": 0.8850135207176208,
+      "learning_rate": 4.649724563405912e-06,
+      "loss": 0.6382,
+      "step": 2997
+    },
+    {
+      "epoch": 1.0526685393258426,
+      "grad_norm": 0.8644798398017883,
+      "learning_rate": 4.649488519000586e-06,
+      "loss": 0.6199,
+      "step": 2998
+    },
+    {
+      "epoch": 1.0530196629213484,
+      "grad_norm": 0.8610007166862488,
+      "learning_rate": 4.649252401084592e-06,
+      "loss": 0.6561,
+      "step": 2999
+    },
+    {
+      "epoch": 1.053370786516854,
+      "grad_norm": 0.9027438163757324,
+      "learning_rate": 4.649016209666006e-06,
+      "loss": 0.641,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0537219101123596,
+      "grad_norm": 0.8504754304885864,
+      "learning_rate": 4.6487799447529025e-06,
+      "loss": 0.6128,
+      "step": 3001
+    },
+    {
+      "epoch": 1.0540730337078652,
+      "grad_norm": 0.8692281246185303,
+      "learning_rate": 4.648543606353365e-06,
+      "loss": 0.6375,
+      "step": 3002
+    },
+    {
+      "epoch": 1.0544241573033708,
+      "grad_norm": 0.9047029614448547,
+      "learning_rate": 4.648307194475474e-06,
+      "loss": 0.6364,
+      "step": 3003
+    },
+    {
+      "epoch": 1.0547752808988764,
+      "grad_norm": 0.9719060659408569,
+      "learning_rate": 4.648070709127316e-06,
+      "loss": 0.6765,
+      "step": 3004
+    },
+    {
+      "epoch": 1.055126404494382,
+      "grad_norm": 0.889021635055542,
+      "learning_rate": 4.647834150316978e-06,
+      "loss": 0.6384,
+      "step": 3005
+    },
+    {
+      "epoch": 1.0554775280898876,
+      "grad_norm": 0.8828878998756409,
+      "learning_rate": 4.647597518052549e-06,
+      "loss": 0.628,
+      "step": 3006
+    },
+    {
+      "epoch": 1.0558286516853932,
+      "grad_norm": 0.8976204991340637,
+      "learning_rate": 4.647360812342123e-06,
+      "loss": 0.6497,
+      "step": 3007
+    },
+    {
+      "epoch": 1.0561797752808988,
+      "grad_norm": 0.9564626216888428,
+      "learning_rate": 4.647124033193795e-06,
+      "loss": 0.6396,
+      "step": 3008
+    },
+    {
+      "epoch": 1.0565308988764044,
+      "grad_norm": 0.8927937150001526,
+      "learning_rate": 4.6468871806156625e-06,
+      "loss": 0.6332,
+      "step": 3009
+    },
+    {
+      "epoch": 1.0568820224719102,
+      "grad_norm": 0.9105674624443054,
+      "learning_rate": 4.646650254615825e-06,
+      "loss": 0.6592,
+      "step": 3010
+    },
+    {
+      "epoch": 1.0572331460674158,
+      "grad_norm": 0.869835376739502,
+      "learning_rate": 4.646413255202386e-06,
+      "loss": 0.6322,
+      "step": 3011
+    },
+    {
+      "epoch": 1.0575842696629214,
+      "grad_norm": 0.9017779231071472,
+      "learning_rate": 4.646176182383451e-06,
+      "loss": 0.585,
+      "step": 3012
+    },
+    {
+      "epoch": 1.057935393258427,
+      "grad_norm": 0.8828255534172058,
+      "learning_rate": 4.645939036167127e-06,
+      "loss": 0.6305,
+      "step": 3013
+    },
+    {
+      "epoch": 1.0582865168539326,
+      "grad_norm": 0.8809278011322021,
+      "learning_rate": 4.645701816561523e-06,
+      "loss": 0.6758,
+      "step": 3014
+    },
+    {
+      "epoch": 1.0586376404494382,
+      "grad_norm": 0.8752210736274719,
+      "learning_rate": 4.645464523574753e-06,
+      "loss": 0.6448,
+      "step": 3015
+    },
+    {
+      "epoch": 1.0589887640449438,
+      "grad_norm": 0.8653308749198914,
+      "learning_rate": 4.645227157214933e-06,
+      "loss": 0.6405,
+      "step": 3016
+    },
+    {
+      "epoch": 1.0593398876404494,
+      "grad_norm": 1.4249169826507568,
+      "learning_rate": 4.64498971749018e-06,
+      "loss": 0.6015,
+      "step": 3017
+    },
+    {
+      "epoch": 1.059691011235955,
+      "grad_norm": 0.8942536115646362,
+      "learning_rate": 4.6447522044086144e-06,
+      "loss": 0.6818,
+      "step": 3018
+    },
+    {
+      "epoch": 1.0600421348314606,
+      "grad_norm": 0.8870991468429565,
+      "learning_rate": 4.644514617978359e-06,
+      "loss": 0.6296,
+      "step": 3019
+    },
+    {
+      "epoch": 1.0603932584269662,
+      "grad_norm": 0.9000036716461182,
+      "learning_rate": 4.644276958207538e-06,
+      "loss": 0.6122,
+      "step": 3020
+    },
+    {
+      "epoch": 1.060744382022472,
+      "grad_norm": 0.9074209332466125,
+      "learning_rate": 4.644039225104279e-06,
+      "loss": 0.658,
+      "step": 3021
+    },
+    {
+      "epoch": 1.0610955056179776,
+      "grad_norm": 0.9330301880836487,
+      "learning_rate": 4.643801418676714e-06,
+      "loss": 0.6222,
+      "step": 3022
+    },
+    {
+      "epoch": 1.0614466292134832,
+      "grad_norm": 0.892718493938446,
+      "learning_rate": 4.643563538932976e-06,
+      "loss": 0.6382,
+      "step": 3023
+    },
+    {
+      "epoch": 1.0617977528089888,
+      "grad_norm": 0.8774312734603882,
+      "learning_rate": 4.643325585881198e-06,
+      "loss": 0.61,
+      "step": 3024
+    },
+    {
+      "epoch": 1.0621488764044944,
+      "grad_norm": 0.9106227159500122,
+      "learning_rate": 4.643087559529519e-06,
+      "loss": 0.6362,
+      "step": 3025
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 0.8989148139953613,
+      "learning_rate": 4.642849459886079e-06,
+      "loss": 0.6264,
+      "step": 3026
+    },
+    {
+      "epoch": 1.0628511235955056,
+      "grad_norm": 0.9128755927085876,
+      "learning_rate": 4.6426112869590224e-06,
+      "loss": 0.6537,
+      "step": 3027
+    },
+    {
+      "epoch": 1.0632022471910112,
+      "grad_norm": 0.9222116470336914,
+      "learning_rate": 4.642373040756493e-06,
+      "loss": 0.6387,
+      "step": 3028
+    },
+    {
+      "epoch": 1.0635533707865168,
+      "grad_norm": 0.8915761113166809,
+      "learning_rate": 4.6421347212866385e-06,
+      "loss": 0.6207,
+      "step": 3029
+    },
+    {
+      "epoch": 1.0639044943820224,
+      "grad_norm": 0.9350106120109558,
+      "learning_rate": 4.64189632855761e-06,
+      "loss": 0.6882,
+      "step": 3030
+    },
+    {
+      "epoch": 1.064255617977528,
+      "grad_norm": 0.9001282453536987,
+      "learning_rate": 4.641657862577559e-06,
+      "loss": 0.6126,
+      "step": 3031
+    },
+    {
+      "epoch": 1.0646067415730338,
+      "grad_norm": 0.8905743360519409,
+      "learning_rate": 4.6414193233546426e-06,
+      "loss": 0.6418,
+      "step": 3032
+    },
+    {
+      "epoch": 1.0649578651685394,
+      "grad_norm": 0.9091342687606812,
+      "learning_rate": 4.6411807108970175e-06,
+      "loss": 0.632,
+      "step": 3033
+    },
+    {
+      "epoch": 1.065308988764045,
+      "grad_norm": 0.879478931427002,
+      "learning_rate": 4.640942025212844e-06,
+      "loss": 0.6223,
+      "step": 3034
+    },
+    {
+      "epoch": 1.0656601123595506,
+      "grad_norm": 0.8716598153114319,
+      "learning_rate": 4.640703266310286e-06,
+      "loss": 0.6576,
+      "step": 3035
+    },
+    {
+      "epoch": 1.0660112359550562,
+      "grad_norm": 0.867374837398529,
+      "learning_rate": 4.640464434197507e-06,
+      "loss": 0.618,
+      "step": 3036
+    },
+    {
+      "epoch": 1.0663623595505618,
+      "grad_norm": 0.899344801902771,
+      "learning_rate": 4.640225528882677e-06,
+      "loss": 0.6314,
+      "step": 3037
+    },
+    {
+      "epoch": 1.0667134831460674,
+      "grad_norm": 0.8644077181816101,
+      "learning_rate": 4.639986550373965e-06,
+      "loss": 0.6707,
+      "step": 3038
+    },
+    {
+      "epoch": 1.067064606741573,
+      "grad_norm": 0.8869829177856445,
+      "learning_rate": 4.639747498679544e-06,
+      "loss": 0.6249,
+      "step": 3039
+    },
+    {
+      "epoch": 1.0674157303370786,
+      "grad_norm": 0.9011994004249573,
+      "learning_rate": 4.63950837380759e-06,
+      "loss": 0.6363,
+      "step": 3040
+    },
+    {
+      "epoch": 1.0677668539325842,
+      "grad_norm": 0.870620846748352,
+      "learning_rate": 4.63926917576628e-06,
+      "loss": 0.6256,
+      "step": 3041
+    },
+    {
+      "epoch": 1.0681179775280898,
+      "grad_norm": 0.8807886838912964,
+      "learning_rate": 4.639029904563795e-06,
+      "loss": 0.6254,
+      "step": 3042
+    },
+    {
+      "epoch": 1.0684691011235956,
+      "grad_norm": 0.8803632259368896,
+      "learning_rate": 4.638790560208318e-06,
+      "loss": 0.6605,
+      "step": 3043
+    },
+    {
+      "epoch": 1.0688202247191012,
+      "grad_norm": 0.8775250315666199,
+      "learning_rate": 4.638551142708033e-06,
+      "loss": 0.6239,
+      "step": 3044
+    },
+    {
+      "epoch": 1.0691713483146068,
+      "grad_norm": 0.8915308117866516,
+      "learning_rate": 4.63831165207113e-06,
+      "loss": 0.6403,
+      "step": 3045
+    },
+    {
+      "epoch": 1.0695224719101124,
+      "grad_norm": 0.8814626336097717,
+      "learning_rate": 4.638072088305798e-06,
+      "loss": 0.6237,
+      "step": 3046
+    },
+    {
+      "epoch": 1.069873595505618,
+      "grad_norm": 0.9150183200836182,
+      "learning_rate": 4.63783245142023e-06,
+      "loss": 0.652,
+      "step": 3047
+    },
+    {
+      "epoch": 1.0702247191011236,
+      "grad_norm": 0.876789927482605,
+      "learning_rate": 4.637592741422621e-06,
+      "loss": 0.6419,
+      "step": 3048
+    },
+    {
+      "epoch": 1.0705758426966292,
+      "grad_norm": 0.8699705600738525,
+      "learning_rate": 4.63735295832117e-06,
+      "loss": 0.6399,
+      "step": 3049
+    },
+    {
+      "epoch": 1.0709269662921348,
+      "grad_norm": 0.8871046900749207,
+      "learning_rate": 4.637113102124077e-06,
+      "loss": 0.6427,
+      "step": 3050
+    },
+    {
+      "epoch": 1.0712780898876404,
+      "grad_norm": 0.877758800983429,
+      "learning_rate": 4.6368731728395444e-06,
+      "loss": 0.6534,
+      "step": 3051
+    },
+    {
+      "epoch": 1.071629213483146,
+      "grad_norm": 0.8570012450218201,
+      "learning_rate": 4.636633170475777e-06,
+      "loss": 0.6073,
+      "step": 3052
+    },
+    {
+      "epoch": 1.0719803370786516,
+      "grad_norm": 0.8933810591697693,
+      "learning_rate": 4.636393095040985e-06,
+      "loss": 0.608,
+      "step": 3053
+    },
+    {
+      "epoch": 1.0723314606741572,
+      "grad_norm": 0.8845889568328857,
+      "learning_rate": 4.636152946543377e-06,
+      "loss": 0.6466,
+      "step": 3054
+    },
+    {
+      "epoch": 1.072682584269663,
+      "grad_norm": 0.9182982444763184,
+      "learning_rate": 4.635912724991165e-06,
+      "loss": 0.6424,
+      "step": 3055
+    },
+    {
+      "epoch": 1.0730337078651686,
+      "grad_norm": 0.906626284122467,
+      "learning_rate": 4.635672430392566e-06,
+      "loss": 0.6481,
+      "step": 3056
+    },
+    {
+      "epoch": 1.0733848314606742,
+      "grad_norm": 0.9121250510215759,
+      "learning_rate": 4.635432062755798e-06,
+      "loss": 0.6403,
+      "step": 3057
+    },
+    {
+      "epoch": 1.0737359550561798,
+      "grad_norm": 0.8549202680587769,
+      "learning_rate": 4.63519162208908e-06,
+      "loss": 0.5976,
+      "step": 3058
+    },
+    {
+      "epoch": 1.0740870786516854,
+      "grad_norm": 0.9251717329025269,
+      "learning_rate": 4.634951108400636e-06,
+      "loss": 0.6198,
+      "step": 3059
+    },
+    {
+      "epoch": 1.074438202247191,
+      "grad_norm": 0.8714932799339294,
+      "learning_rate": 4.634710521698692e-06,
+      "loss": 0.6565,
+      "step": 3060
+    },
+    {
+      "epoch": 1.0747893258426966,
+      "grad_norm": 0.8903185725212097,
+      "learning_rate": 4.634469861991473e-06,
+      "loss": 0.6119,
+      "step": 3061
+    },
+    {
+      "epoch": 1.0751404494382022,
+      "grad_norm": 0.8958722352981567,
+      "learning_rate": 4.6342291292872125e-06,
+      "loss": 0.6417,
+      "step": 3062
+    },
+    {
+      "epoch": 1.0754915730337078,
+      "grad_norm": 0.8739234209060669,
+      "learning_rate": 4.633988323594142e-06,
+      "loss": 0.6374,
+      "step": 3063
+    },
+    {
+      "epoch": 1.0758426966292134,
+      "grad_norm": 0.8859426975250244,
+      "learning_rate": 4.633747444920497e-06,
+      "loss": 0.6547,
+      "step": 3064
+    },
+    {
+      "epoch": 1.0761938202247192,
+      "grad_norm": 0.9002133011817932,
+      "learning_rate": 4.633506493274514e-06,
+      "loss": 0.6693,
+      "step": 3065
+    },
+    {
+      "epoch": 1.0765449438202248,
+      "grad_norm": 0.8716120719909668,
+      "learning_rate": 4.633265468664436e-06,
+      "loss": 0.5971,
+      "step": 3066
+    },
+    {
+      "epoch": 1.0768960674157304,
+      "grad_norm": 0.9050167798995972,
+      "learning_rate": 4.633024371098503e-06,
+      "loss": 0.6686,
+      "step": 3067
+    },
+    {
+      "epoch": 1.077247191011236,
+      "grad_norm": 0.8593562841415405,
+      "learning_rate": 4.6327832005849635e-06,
+      "loss": 0.6134,
+      "step": 3068
+    },
+    {
+      "epoch": 1.0775983146067416,
+      "grad_norm": 0.8681020736694336,
+      "learning_rate": 4.632541957132063e-06,
+      "loss": 0.6232,
+      "step": 3069
+    },
+    {
+      "epoch": 1.0779494382022472,
+      "grad_norm": 0.8903972506523132,
+      "learning_rate": 4.632300640748052e-06,
+      "loss": 0.6333,
+      "step": 3070
+    },
+    {
+      "epoch": 1.0783005617977528,
+      "grad_norm": 0.8613920211791992,
+      "learning_rate": 4.632059251441184e-06,
+      "loss": 0.6129,
+      "step": 3071
+    },
+    {
+      "epoch": 1.0786516853932584,
+      "grad_norm": 0.9118038415908813,
+      "learning_rate": 4.631817789219714e-06,
+      "loss": 0.6453,
+      "step": 3072
+    },
+    {
+      "epoch": 1.079002808988764,
+      "grad_norm": 0.9000958800315857,
+      "learning_rate": 4.6315762540919e-06,
+      "loss": 0.6578,
+      "step": 3073
+    },
+    {
+      "epoch": 1.0793539325842696,
+      "grad_norm": 0.8761112093925476,
+      "learning_rate": 4.631334646066001e-06,
+      "loss": 0.6345,
+      "step": 3074
+    },
+    {
+      "epoch": 1.0797050561797752,
+      "grad_norm": 0.9070879817008972,
+      "learning_rate": 4.6310929651502825e-06,
+      "loss": 0.6475,
+      "step": 3075
+    },
+    {
+      "epoch": 1.0800561797752808,
+      "grad_norm": 0.869162380695343,
+      "learning_rate": 4.630851211353007e-06,
+      "loss": 0.6341,
+      "step": 3076
+    },
+    {
+      "epoch": 1.0804073033707866,
+      "grad_norm": 0.9210894703865051,
+      "learning_rate": 4.630609384682444e-06,
+      "loss": 0.6839,
+      "step": 3077
+    },
+    {
+      "epoch": 1.0807584269662922,
+      "grad_norm": 0.9321752786636353,
+      "learning_rate": 4.630367485146863e-06,
+      "loss": 0.6581,
+      "step": 3078
+    },
+    {
+      "epoch": 1.0811095505617978,
+      "grad_norm": 0.8789924383163452,
+      "learning_rate": 4.630125512754537e-06,
+      "loss": 0.6129,
+      "step": 3079
+    },
+    {
+      "epoch": 1.0814606741573034,
+      "grad_norm": 0.8936173915863037,
+      "learning_rate": 4.629883467513741e-06,
+      "loss": 0.6021,
+      "step": 3080
+    },
+    {
+      "epoch": 1.081811797752809,
+      "grad_norm": 0.8786038160324097,
+      "learning_rate": 4.629641349432754e-06,
+      "loss": 0.633,
+      "step": 3081
+    },
+    {
+      "epoch": 1.0821629213483146,
+      "grad_norm": 0.8804035782814026,
+      "learning_rate": 4.629399158519855e-06,
+      "loss": 0.6404,
+      "step": 3082
+    },
+    {
+      "epoch": 1.0825140449438202,
+      "grad_norm": 0.9330071210861206,
+      "learning_rate": 4.629156894783326e-06,
+      "loss": 0.6629,
+      "step": 3083
+    },
+    {
+      "epoch": 1.0828651685393258,
+      "grad_norm": 0.8635822534561157,
+      "learning_rate": 4.628914558231455e-06,
+      "loss": 0.6409,
+      "step": 3084
+    },
+    {
+      "epoch": 1.0832162921348314,
+      "grad_norm": 0.9098319411277771,
+      "learning_rate": 4.628672148872526e-06,
+      "loss": 0.6533,
+      "step": 3085
+    },
+    {
+      "epoch": 1.083567415730337,
+      "grad_norm": 0.8650401830673218,
+      "learning_rate": 4.628429666714831e-06,
+      "loss": 0.6031,
+      "step": 3086
+    },
+    {
+      "epoch": 1.0839185393258428,
+      "grad_norm": 0.8966626524925232,
+      "learning_rate": 4.6281871117666634e-06,
+      "loss": 0.6209,
+      "step": 3087
+    },
+    {
+      "epoch": 1.0842696629213484,
+      "grad_norm": 0.9028195142745972,
+      "learning_rate": 4.627944484036318e-06,
+      "loss": 0.6343,
+      "step": 3088
+    },
+    {
+      "epoch": 1.084620786516854,
+      "grad_norm": 0.9410665035247803,
+      "learning_rate": 4.627701783532092e-06,
+      "loss": 0.6796,
+      "step": 3089
+    },
+    {
+      "epoch": 1.0849719101123596,
+      "grad_norm": 0.9015296697616577,
+      "learning_rate": 4.627459010262285e-06,
+      "loss": 0.6649,
+      "step": 3090
+    },
+    {
+      "epoch": 1.0853230337078652,
+      "grad_norm": 0.8640577793121338,
+      "learning_rate": 4.627216164235201e-06,
+      "loss": 0.6224,
+      "step": 3091
+    },
+    {
+      "epoch": 1.0856741573033708,
+      "grad_norm": 0.8947779536247253,
+      "learning_rate": 4.626973245459144e-06,
+      "loss": 0.6254,
+      "step": 3092
+    },
+    {
+      "epoch": 1.0860252808988764,
+      "grad_norm": 0.8579085469245911,
+      "learning_rate": 4.626730253942422e-06,
+      "loss": 0.6158,
+      "step": 3093
+    },
+    {
+      "epoch": 1.086376404494382,
+      "grad_norm": 0.8828961849212646,
+      "learning_rate": 4.626487189693345e-06,
+      "loss": 0.6429,
+      "step": 3094
+    },
+    {
+      "epoch": 1.0867275280898876,
+      "grad_norm": 0.8998451232910156,
+      "learning_rate": 4.626244052720227e-06,
+      "loss": 0.6176,
+      "step": 3095
+    },
+    {
+      "epoch": 1.0870786516853932,
+      "grad_norm": 0.8975029587745667,
+      "learning_rate": 4.626000843031381e-06,
+      "loss": 0.6106,
+      "step": 3096
+    },
+    {
+      "epoch": 1.0874297752808988,
+      "grad_norm": 0.8873660564422607,
+      "learning_rate": 4.625757560635126e-06,
+      "loss": 0.6279,
+      "step": 3097
+    },
+    {
+      "epoch": 1.0877808988764044,
+      "grad_norm": 0.8927737474441528,
+      "learning_rate": 4.625514205539781e-06,
+      "loss": 0.6208,
+      "step": 3098
+    },
+    {
+      "epoch": 1.0881320224719102,
+      "grad_norm": 0.9284412264823914,
+      "learning_rate": 4.6252707777536685e-06,
+      "loss": 0.6701,
+      "step": 3099
+    },
+    {
+      "epoch": 1.0884831460674158,
+      "grad_norm": 0.8798356652259827,
+      "learning_rate": 4.625027277285114e-06,
+      "loss": 0.5872,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0888342696629214,
+      "grad_norm": 0.9197069406509399,
+      "learning_rate": 4.624783704142446e-06,
+      "loss": 0.6478,
+      "step": 3101
+    },
+    {
+      "epoch": 1.089185393258427,
+      "grad_norm": 0.8777607083320618,
+      "learning_rate": 4.624540058333993e-06,
+      "loss": 0.6248,
+      "step": 3102
+    },
+    {
+      "epoch": 1.0895365168539326,
+      "grad_norm": 0.8981837630271912,
+      "learning_rate": 4.624296339868089e-06,
+      "loss": 0.6398,
+      "step": 3103
+    },
+    {
+      "epoch": 1.0898876404494382,
+      "grad_norm": 1.1026835441589355,
+      "learning_rate": 4.624052548753066e-06,
+      "loss": 0.701,
+      "step": 3104
+    },
+    {
+      "epoch": 1.0902387640449438,
+      "grad_norm": 0.8540518879890442,
+      "learning_rate": 4.6238086849972655e-06,
+      "loss": 0.6133,
+      "step": 3105
+    },
+    {
+      "epoch": 1.0905898876404494,
+      "grad_norm": 0.8801087141036987,
+      "learning_rate": 4.623564748609024e-06,
+      "loss": 0.6557,
+      "step": 3106
+    },
+    {
+      "epoch": 1.090941011235955,
+      "grad_norm": 0.8963208794593811,
+      "learning_rate": 4.623320739596685e-06,
+      "loss": 0.6368,
+      "step": 3107
+    },
+    {
+      "epoch": 1.0912921348314606,
+      "grad_norm": 0.8750952482223511,
+      "learning_rate": 4.623076657968594e-06,
+      "loss": 0.6296,
+      "step": 3108
+    },
+    {
+      "epoch": 1.0916432584269662,
+      "grad_norm": 0.9236363768577576,
+      "learning_rate": 4.622832503733099e-06,
+      "loss": 0.6424,
+      "step": 3109
+    },
+    {
+      "epoch": 1.091994382022472,
+      "grad_norm": 0.8751869797706604,
+      "learning_rate": 4.6225882768985485e-06,
+      "loss": 0.6033,
+      "step": 3110
+    },
+    {
+      "epoch": 1.0923455056179776,
+      "grad_norm": 0.9356908202171326,
+      "learning_rate": 4.6223439774732944e-06,
+      "loss": 0.6372,
+      "step": 3111
+    },
+    {
+      "epoch": 1.0926966292134832,
+      "grad_norm": 0.881905734539032,
+      "learning_rate": 4.622099605465693e-06,
+      "loss": 0.6139,
+      "step": 3112
+    },
+    {
+      "epoch": 1.0930477528089888,
+      "grad_norm": 0.8708788156509399,
+      "learning_rate": 4.6218551608841005e-06,
+      "loss": 0.6441,
+      "step": 3113
+    },
+    {
+      "epoch": 1.0933988764044944,
+      "grad_norm": 0.8892695903778076,
+      "learning_rate": 4.621610643736878e-06,
+      "loss": 0.6329,
+      "step": 3114
+    },
+    {
+      "epoch": 1.09375,
+      "grad_norm": 0.8857613801956177,
+      "learning_rate": 4.6213660540323865e-06,
+      "loss": 0.6215,
+      "step": 3115
+    },
+    {
+      "epoch": 1.0941011235955056,
+      "grad_norm": 0.8977625370025635,
+      "learning_rate": 4.621121391778991e-06,
+      "loss": 0.5709,
+      "step": 3116
+    },
+    {
+      "epoch": 1.0944522471910112,
+      "grad_norm": 0.9355443120002747,
+      "learning_rate": 4.62087665698506e-06,
+      "loss": 0.6367,
+      "step": 3117
+    },
+    {
+      "epoch": 1.0948033707865168,
+      "grad_norm": 0.8514768481254578,
+      "learning_rate": 4.620631849658961e-06,
+      "loss": 0.6059,
+      "step": 3118
+    },
+    {
+      "epoch": 1.0951544943820224,
+      "grad_norm": 0.897706925868988,
+      "learning_rate": 4.6203869698090676e-06,
+      "loss": 0.6327,
+      "step": 3119
+    },
+    {
+      "epoch": 1.095505617977528,
+      "grad_norm": 0.8812986612319946,
+      "learning_rate": 4.620142017443755e-06,
+      "loss": 0.625,
+      "step": 3120
+    },
+    {
+      "epoch": 1.0958567415730338,
+      "grad_norm": 0.878730833530426,
+      "learning_rate": 4.619896992571399e-06,
+      "loss": 0.6629,
+      "step": 3121
+    },
+    {
+      "epoch": 1.0962078651685394,
+      "grad_norm": 0.8637169003486633,
+      "learning_rate": 4.61965189520038e-06,
+      "loss": 0.6148,
+      "step": 3122
+    },
+    {
+      "epoch": 1.096558988764045,
+      "grad_norm": 0.8571529388427734,
+      "learning_rate": 4.61940672533908e-06,
+      "loss": 0.5992,
+      "step": 3123
+    },
+    {
+      "epoch": 1.0969101123595506,
+      "grad_norm": 0.8880246877670288,
+      "learning_rate": 4.619161482995883e-06,
+      "loss": 0.6257,
+      "step": 3124
+    },
+    {
+      "epoch": 1.0972612359550562,
+      "grad_norm": 0.8742407560348511,
+      "learning_rate": 4.618916168179178e-06,
+      "loss": 0.6463,
+      "step": 3125
+    },
+    {
+      "epoch": 1.0976123595505618,
+      "grad_norm": 0.9125159978866577,
+      "learning_rate": 4.6186707808973524e-06,
+      "loss": 0.6534,
+      "step": 3126
+    },
+    {
+      "epoch": 1.0979634831460674,
+      "grad_norm": 0.9120495319366455,
+      "learning_rate": 4.6184253211588e-06,
+      "loss": 0.6858,
+      "step": 3127
+    },
+    {
+      "epoch": 1.098314606741573,
+      "grad_norm": 0.8634641170501709,
+      "learning_rate": 4.618179788971913e-06,
+      "loss": 0.6343,
+      "step": 3128
+    },
+    {
+      "epoch": 1.0986657303370786,
+      "grad_norm": 0.9329657554626465,
+      "learning_rate": 4.6179341843450895e-06,
+      "loss": 0.6626,
+      "step": 3129
+    },
+    {
+      "epoch": 1.0990168539325842,
+      "grad_norm": 0.9169970750808716,
+      "learning_rate": 4.617688507286731e-06,
+      "loss": 0.6737,
+      "step": 3130
+    },
+    {
+      "epoch": 1.0993679775280898,
+      "grad_norm": 0.9113175272941589,
+      "learning_rate": 4.617442757805236e-06,
+      "loss": 0.6285,
+      "step": 3131
+    },
+    {
+      "epoch": 1.0997191011235956,
+      "grad_norm": 0.8834072351455688,
+      "learning_rate": 4.617196935909012e-06,
+      "loss": 0.6344,
+      "step": 3132
+    },
+    {
+      "epoch": 1.1000702247191012,
+      "grad_norm": 0.8792498707771301,
+      "learning_rate": 4.6169510416064635e-06,
+      "loss": 0.635,
+      "step": 3133
+    },
+    {
+      "epoch": 1.1004213483146068,
+      "grad_norm": 0.864789605140686,
+      "learning_rate": 4.6167050749060014e-06,
+      "loss": 0.6498,
+      "step": 3134
+    },
+    {
+      "epoch": 1.1007724719101124,
+      "grad_norm": 0.8957346677780151,
+      "learning_rate": 4.616459035816037e-06,
+      "loss": 0.6484,
+      "step": 3135
+    },
+    {
+      "epoch": 1.101123595505618,
+      "grad_norm": 0.9335662126541138,
+      "learning_rate": 4.6162129243449845e-06,
+      "loss": 0.651,
+      "step": 3136
+    },
+    {
+      "epoch": 1.1014747191011236,
+      "grad_norm": 0.8907780051231384,
+      "learning_rate": 4.615966740501261e-06,
+      "loss": 0.6279,
+      "step": 3137
+    },
+    {
+      "epoch": 1.1018258426966292,
+      "grad_norm": 0.8416582942008972,
+      "learning_rate": 4.615720484293286e-06,
+      "loss": 0.6183,
+      "step": 3138
+    },
+    {
+      "epoch": 1.1021769662921348,
+      "grad_norm": 0.8414410352706909,
+      "learning_rate": 4.6154741557294805e-06,
+      "loss": 0.6091,
+      "step": 3139
+    },
+    {
+      "epoch": 1.1025280898876404,
+      "grad_norm": 0.8924736380577087,
+      "learning_rate": 4.6152277548182685e-06,
+      "loss": 0.6466,
+      "step": 3140
+    },
+    {
+      "epoch": 1.102879213483146,
+      "grad_norm": 0.952515721321106,
+      "learning_rate": 4.614981281568078e-06,
+      "loss": 0.5875,
+      "step": 3141
+    },
+    {
+      "epoch": 1.1032303370786516,
+      "grad_norm": 1.0150952339172363,
+      "learning_rate": 4.6147347359873376e-06,
+      "loss": 0.6295,
+      "step": 3142
+    },
+    {
+      "epoch": 1.1035814606741572,
+      "grad_norm": 0.8959284424781799,
+      "learning_rate": 4.614488118084478e-06,
+      "loss": 0.6483,
+      "step": 3143
+    },
+    {
+      "epoch": 1.103932584269663,
+      "grad_norm": 0.8429542183876038,
+      "learning_rate": 4.614241427867935e-06,
+      "loss": 0.6017,
+      "step": 3144
+    },
+    {
+      "epoch": 1.1042837078651686,
+      "grad_norm": 0.8937910795211792,
+      "learning_rate": 4.613994665346144e-06,
+      "loss": 0.6685,
+      "step": 3145
+    },
+    {
+      "epoch": 1.1046348314606742,
+      "grad_norm": 0.9087120890617371,
+      "learning_rate": 4.6137478305275444e-06,
+      "loss": 0.656,
+      "step": 3146
+    },
+    {
+      "epoch": 1.1049859550561798,
+      "grad_norm": 0.9055848717689514,
+      "learning_rate": 4.613500923420577e-06,
+      "loss": 0.641,
+      "step": 3147
+    },
+    {
+      "epoch": 1.1053370786516854,
+      "grad_norm": 0.8999933004379272,
+      "learning_rate": 4.613253944033688e-06,
+      "loss": 0.6282,
+      "step": 3148
+    },
+    {
+      "epoch": 1.105688202247191,
+      "grad_norm": 0.901351809501648,
+      "learning_rate": 4.6130068923753226e-06,
+      "loss": 0.6409,
+      "step": 3149
+    },
+    {
+      "epoch": 1.1060393258426966,
+      "grad_norm": 0.8660926818847656,
+      "learning_rate": 4.612759768453928e-06,
+      "loss": 0.6238,
+      "step": 3150
+    },
+    {
+      "epoch": 1.1063904494382022,
+      "grad_norm": 0.8839932084083557,
+      "learning_rate": 4.612512572277959e-06,
+      "loss": 0.6394,
+      "step": 3151
+    },
+    {
+      "epoch": 1.1067415730337078,
+      "grad_norm": 0.882322371006012,
+      "learning_rate": 4.6122653038558665e-06,
+      "loss": 0.6587,
+      "step": 3152
+    },
+    {
+      "epoch": 1.1070926966292134,
+      "grad_norm": 0.8952814936637878,
+      "learning_rate": 4.6120179631961086e-06,
+      "loss": 0.647,
+      "step": 3153
+    },
+    {
+      "epoch": 1.1074438202247192,
+      "grad_norm": 0.8898069262504578,
+      "learning_rate": 4.611770550307143e-06,
+      "loss": 0.6238,
+      "step": 3154
+    },
+    {
+      "epoch": 1.1077949438202248,
+      "grad_norm": 0.8928719758987427,
+      "learning_rate": 4.6115230651974315e-06,
+      "loss": 0.6393,
+      "step": 3155
+    },
+    {
+      "epoch": 1.1081460674157304,
+      "grad_norm": 0.908116340637207,
+      "learning_rate": 4.611275507875439e-06,
+      "loss": 0.6431,
+      "step": 3156
+    },
+    {
+      "epoch": 1.108497191011236,
+      "grad_norm": 0.9306859970092773,
+      "learning_rate": 4.611027878349631e-06,
+      "loss": 0.6692,
+      "step": 3157
+    },
+    {
+      "epoch": 1.1088483146067416,
+      "grad_norm": 0.8866667151451111,
+      "learning_rate": 4.610780176628475e-06,
+      "loss": 0.621,
+      "step": 3158
+    },
+    {
+      "epoch": 1.1091994382022472,
+      "grad_norm": 0.8539839386940002,
+      "learning_rate": 4.610532402720444e-06,
+      "loss": 0.601,
+      "step": 3159
+    },
+    {
+      "epoch": 1.1095505617977528,
+      "grad_norm": 0.8730782270431519,
+      "learning_rate": 4.61028455663401e-06,
+      "loss": 0.6,
+      "step": 3160
+    },
+    {
+      "epoch": 1.1099016853932584,
+      "grad_norm": 0.8877293467521667,
+      "learning_rate": 4.6100366383776505e-06,
+      "loss": 0.629,
+      "step": 3161
+    },
+    {
+      "epoch": 1.110252808988764,
+      "grad_norm": 0.8557886481285095,
+      "learning_rate": 4.609788647959843e-06,
+      "loss": 0.6303,
+      "step": 3162
+    },
+    {
+      "epoch": 1.1106039325842696,
+      "grad_norm": 0.8975784778594971,
+      "learning_rate": 4.609540585389069e-06,
+      "loss": 0.645,
+      "step": 3163
+    },
+    {
+      "epoch": 1.1109550561797752,
+      "grad_norm": 0.8701896667480469,
+      "learning_rate": 4.6092924506738134e-06,
+      "loss": 0.6208,
+      "step": 3164
+    },
+    {
+      "epoch": 1.1113061797752808,
+      "grad_norm": 0.8492256999015808,
+      "learning_rate": 4.60904424382256e-06,
+      "loss": 0.6441,
+      "step": 3165
+    },
+    {
+      "epoch": 1.1116573033707866,
+      "grad_norm": 0.8746402263641357,
+      "learning_rate": 4.608795964843798e-06,
+      "loss": 0.6269,
+      "step": 3166
+    },
+    {
+      "epoch": 1.1120084269662922,
+      "grad_norm": 0.9015893936157227,
+      "learning_rate": 4.60854761374602e-06,
+      "loss": 0.6396,
+      "step": 3167
+    },
+    {
+      "epoch": 1.1123595505617978,
+      "grad_norm": 0.9147549867630005,
+      "learning_rate": 4.608299190537717e-06,
+      "loss": 0.6055,
+      "step": 3168
+    },
+    {
+      "epoch": 1.1127106741573034,
+      "grad_norm": 0.8999795317649841,
+      "learning_rate": 4.6080506952273855e-06,
+      "loss": 0.6343,
+      "step": 3169
+    },
+    {
+      "epoch": 1.113061797752809,
+      "grad_norm": 0.8840816617012024,
+      "learning_rate": 4.607802127823524e-06,
+      "loss": 0.6173,
+      "step": 3170
+    },
+    {
+      "epoch": 1.1134129213483146,
+      "grad_norm": 0.9025754332542419,
+      "learning_rate": 4.6075534883346335e-06,
+      "loss": 0.6426,
+      "step": 3171
+    },
+    {
+      "epoch": 1.1137640449438202,
+      "grad_norm": 0.9157222509384155,
+      "learning_rate": 4.607304776769217e-06,
+      "loss": 0.6639,
+      "step": 3172
+    },
+    {
+      "epoch": 1.1141151685393258,
+      "grad_norm": 0.8551729321479797,
+      "learning_rate": 4.607055993135781e-06,
+      "loss": 0.62,
+      "step": 3173
+    },
+    {
+      "epoch": 1.1144662921348314,
+      "grad_norm": 0.9303814172744751,
+      "learning_rate": 4.606807137442833e-06,
+      "loss": 0.6504,
+      "step": 3174
+    },
+    {
+      "epoch": 1.114817415730337,
+      "grad_norm": 0.8886232972145081,
+      "learning_rate": 4.606558209698884e-06,
+      "loss": 0.6301,
+      "step": 3175
+    },
+    {
+      "epoch": 1.1151685393258428,
+      "grad_norm": 0.8753377795219421,
+      "learning_rate": 4.606309209912446e-06,
+      "loss": 0.6206,
+      "step": 3176
+    },
+    {
+      "epoch": 1.1155196629213484,
+      "grad_norm": 0.9052675366401672,
+      "learning_rate": 4.606060138092036e-06,
+      "loss": 0.6563,
+      "step": 3177
+    },
+    {
+      "epoch": 1.115870786516854,
+      "grad_norm": 0.8677799105644226,
+      "learning_rate": 4.6058109942461705e-06,
+      "loss": 0.6203,
+      "step": 3178
+    },
+    {
+      "epoch": 1.1162219101123596,
+      "grad_norm": 0.8847718834877014,
+      "learning_rate": 4.605561778383372e-06,
+      "loss": 0.6473,
+      "step": 3179
+    },
+    {
+      "epoch": 1.1165730337078652,
+      "grad_norm": 0.8971977829933167,
+      "learning_rate": 4.605312490512162e-06,
+      "loss": 0.6363,
+      "step": 3180
+    },
+    {
+      "epoch": 1.1169241573033708,
+      "grad_norm": 0.8800894021987915,
+      "learning_rate": 4.605063130641067e-06,
+      "loss": 0.6071,
+      "step": 3181
+    },
+    {
+      "epoch": 1.1172752808988764,
+      "grad_norm": 0.884075403213501,
+      "learning_rate": 4.604813698778613e-06,
+      "loss": 0.6281,
+      "step": 3182
+    },
+    {
+      "epoch": 1.117626404494382,
+      "grad_norm": 0.8670194745063782,
+      "learning_rate": 4.604564194933332e-06,
+      "loss": 0.644,
+      "step": 3183
+    },
+    {
+      "epoch": 1.1179775280898876,
+      "grad_norm": 0.8559613227844238,
+      "learning_rate": 4.604314619113756e-06,
+      "loss": 0.6348,
+      "step": 3184
+    },
+    {
+      "epoch": 1.1183286516853932,
+      "grad_norm": 0.8893458247184753,
+      "learning_rate": 4.604064971328422e-06,
+      "loss": 0.646,
+      "step": 3185
+    },
+    {
+      "epoch": 1.1186797752808988,
+      "grad_norm": 0.9015454053878784,
+      "learning_rate": 4.603815251585867e-06,
+      "loss": 0.6798,
+      "step": 3186
+    },
+    {
+      "epoch": 1.1190308988764044,
+      "grad_norm": 0.881079912185669,
+      "learning_rate": 4.603565459894629e-06,
+      "loss": 0.6282,
+      "step": 3187
+    },
+    {
+      "epoch": 1.1193820224719102,
+      "grad_norm": 0.8383495211601257,
+      "learning_rate": 4.6033155962632535e-06,
+      "loss": 0.613,
+      "step": 3188
+    },
+    {
+      "epoch": 1.1197331460674158,
+      "grad_norm": 0.845001757144928,
+      "learning_rate": 4.603065660700283e-06,
+      "loss": 0.6104,
+      "step": 3189
+    },
+    {
+      "epoch": 1.1200842696629214,
+      "grad_norm": 0.8651373982429504,
+      "learning_rate": 4.602815653214267e-06,
+      "loss": 0.6286,
+      "step": 3190
+    },
+    {
+      "epoch": 1.120435393258427,
+      "grad_norm": 0.8765587210655212,
+      "learning_rate": 4.6025655738137566e-06,
+      "loss": 0.6206,
+      "step": 3191
+    },
+    {
+      "epoch": 1.1207865168539326,
+      "grad_norm": 0.8903856873512268,
+      "learning_rate": 4.602315422507301e-06,
+      "loss": 0.6162,
+      "step": 3192
+    },
+    {
+      "epoch": 1.1211376404494382,
+      "grad_norm": 0.8526650667190552,
+      "learning_rate": 4.602065199303458e-06,
+      "loss": 0.6418,
+      "step": 3193
+    },
+    {
+      "epoch": 1.1214887640449438,
+      "grad_norm": 0.8819793462753296,
+      "learning_rate": 4.601814904210783e-06,
+      "loss": 0.6513,
+      "step": 3194
+    },
+    {
+      "epoch": 1.1218398876404494,
+      "grad_norm": 0.9131810069084167,
+      "learning_rate": 4.601564537237837e-06,
+      "loss": 0.6234,
+      "step": 3195
+    },
+    {
+      "epoch": 1.122191011235955,
+      "grad_norm": 0.8469668030738831,
+      "learning_rate": 4.601314098393183e-06,
+      "loss": 0.5831,
+      "step": 3196
+    },
+    {
+      "epoch": 1.1225421348314606,
+      "grad_norm": 0.8926812410354614,
+      "learning_rate": 4.601063587685384e-06,
+      "loss": 0.6649,
+      "step": 3197
+    },
+    {
+      "epoch": 1.1228932584269662,
+      "grad_norm": 0.8658381700515747,
+      "learning_rate": 4.600813005123009e-06,
+      "loss": 0.6095,
+      "step": 3198
+    },
+    {
+      "epoch": 1.123244382022472,
+      "grad_norm": 0.8796584606170654,
+      "learning_rate": 4.600562350714627e-06,
+      "loss": 0.6189,
+      "step": 3199
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.8917720913887024,
+      "learning_rate": 4.6003116244688095e-06,
+      "loss": 0.6294,
+      "step": 3200
+    },
+    {
+      "epoch": 1.1239466292134832,
+      "grad_norm": 0.8931936621665955,
+      "learning_rate": 4.600060826394131e-06,
+      "loss": 0.6428,
+      "step": 3201
+    },
+    {
+      "epoch": 1.1242977528089888,
+      "grad_norm": 0.8980125784873962,
+      "learning_rate": 4.599809956499171e-06,
+      "loss": 0.6325,
+      "step": 3202
+    },
+    {
+      "epoch": 1.1246488764044944,
+      "grad_norm": 0.9097473621368408,
+      "learning_rate": 4.599559014792507e-06,
+      "loss": 0.6634,
+      "step": 3203
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.920569121837616,
+      "learning_rate": 4.59930800128272e-06,
+      "loss": 0.6751,
+      "step": 3204
+    },
+    {
+      "epoch": 1.1253511235955056,
+      "grad_norm": 0.8496858477592468,
+      "learning_rate": 4.599056915978396e-06,
+      "loss": 0.6123,
+      "step": 3205
+    },
+    {
+      "epoch": 1.1257022471910112,
+      "grad_norm": 0.9153531193733215,
+      "learning_rate": 4.598805758888122e-06,
+      "loss": 0.6713,
+      "step": 3206
+    },
+    {
+      "epoch": 1.1260533707865168,
+      "grad_norm": 0.8894977569580078,
+      "learning_rate": 4.5985545300204875e-06,
+      "loss": 0.6154,
+      "step": 3207
+    },
+    {
+      "epoch": 1.1264044943820224,
+      "grad_norm": 0.9272782206535339,
+      "learning_rate": 4.598303229384083e-06,
+      "loss": 0.6311,
+      "step": 3208
+    },
+    {
+      "epoch": 1.126755617977528,
+      "grad_norm": 0.8554537296295166,
+      "learning_rate": 4.598051856987504e-06,
+      "loss": 0.6121,
+      "step": 3209
+    },
+    {
+      "epoch": 1.1271067415730336,
+      "grad_norm": 0.8680627346038818,
+      "learning_rate": 4.5978004128393475e-06,
+      "loss": 0.6256,
+      "step": 3210
+    },
+    {
+      "epoch": 1.1274578651685394,
+      "grad_norm": 0.8919551372528076,
+      "learning_rate": 4.597548896948212e-06,
+      "loss": 0.6632,
+      "step": 3211
+    },
+    {
+      "epoch": 1.127808988764045,
+      "grad_norm": 0.90400230884552,
+      "learning_rate": 4.597297309322698e-06,
+      "loss": 0.6558,
+      "step": 3212
+    },
+    {
+      "epoch": 1.1281601123595506,
+      "grad_norm": 0.8687441945075989,
+      "learning_rate": 4.597045649971411e-06,
+      "loss": 0.6301,
+      "step": 3213
+    },
+    {
+      "epoch": 1.1285112359550562,
+      "grad_norm": 0.8631216883659363,
+      "learning_rate": 4.596793918902957e-06,
+      "loss": 0.6264,
+      "step": 3214
+    },
+    {
+      "epoch": 1.1288623595505618,
+      "grad_norm": 0.8835992813110352,
+      "learning_rate": 4.596542116125947e-06,
+      "loss": 0.6416,
+      "step": 3215
+    },
+    {
+      "epoch": 1.1292134831460674,
+      "grad_norm": 0.8880035281181335,
+      "learning_rate": 4.596290241648989e-06,
+      "loss": 0.6072,
+      "step": 3216
+    },
+    {
+      "epoch": 1.129564606741573,
+      "grad_norm": 0.9440239667892456,
+      "learning_rate": 4.596038295480699e-06,
+      "loss": 0.6522,
+      "step": 3217
+    },
+    {
+      "epoch": 1.1299157303370786,
+      "grad_norm": 0.8469729423522949,
+      "learning_rate": 4.595786277629693e-06,
+      "loss": 0.6141,
+      "step": 3218
+    },
+    {
+      "epoch": 1.1302668539325842,
+      "grad_norm": 0.8801910281181335,
+      "learning_rate": 4.595534188104589e-06,
+      "loss": 0.648,
+      "step": 3219
+    },
+    {
+      "epoch": 1.13061797752809,
+      "grad_norm": 0.926419198513031,
+      "learning_rate": 4.59528202691401e-06,
+      "loss": 0.646,
+      "step": 3220
+    },
+    {
+      "epoch": 1.1309691011235956,
+      "grad_norm": 0.8675405383110046,
+      "learning_rate": 4.595029794066578e-06,
+      "loss": 0.626,
+      "step": 3221
+    },
+    {
+      "epoch": 1.1313202247191012,
+      "grad_norm": 0.8670094013214111,
+      "learning_rate": 4.594777489570919e-06,
+      "loss": 0.618,
+      "step": 3222
+    },
+    {
+      "epoch": 1.1316713483146068,
+      "grad_norm": 0.8746402263641357,
+      "learning_rate": 4.5945251134356635e-06,
+      "loss": 0.6172,
+      "step": 3223
+    },
+    {
+      "epoch": 1.1320224719101124,
+      "grad_norm": 0.9028223156929016,
+      "learning_rate": 4.594272665669442e-06,
+      "loss": 0.6329,
+      "step": 3224
+    },
+    {
+      "epoch": 1.132373595505618,
+      "grad_norm": 0.9109153151512146,
+      "learning_rate": 4.594020146280887e-06,
+      "loss": 0.646,
+      "step": 3225
+    },
+    {
+      "epoch": 1.1327247191011236,
+      "grad_norm": 0.8849334716796875,
+      "learning_rate": 4.593767555278634e-06,
+      "loss": 0.6491,
+      "step": 3226
+    },
+    {
+      "epoch": 1.1330758426966292,
+      "grad_norm": 0.9008032083511353,
+      "learning_rate": 4.593514892671323e-06,
+      "loss": 0.642,
+      "step": 3227
+    },
+    {
+      "epoch": 1.1334269662921348,
+      "grad_norm": 0.8774647116661072,
+      "learning_rate": 4.593262158467594e-06,
+      "loss": 0.6362,
+      "step": 3228
+    },
+    {
+      "epoch": 1.1337780898876404,
+      "grad_norm": 0.8808940649032593,
+      "learning_rate": 4.59300935267609e-06,
+      "loss": 0.6469,
+      "step": 3229
+    },
+    {
+      "epoch": 1.134129213483146,
+      "grad_norm": 0.9236598610877991,
+      "learning_rate": 4.592756475305458e-06,
+      "loss": 0.6227,
+      "step": 3230
+    },
+    {
+      "epoch": 1.1344803370786516,
+      "grad_norm": 0.8788805603981018,
+      "learning_rate": 4.592503526364344e-06,
+      "loss": 0.6168,
+      "step": 3231
+    },
+    {
+      "epoch": 1.1348314606741572,
+      "grad_norm": 1.109196424484253,
+      "learning_rate": 4.5922505058614016e-06,
+      "loss": 0.6154,
+      "step": 3232
+    },
+    {
+      "epoch": 1.135182584269663,
+      "grad_norm": 0.8943460583686829,
+      "learning_rate": 4.591997413805281e-06,
+      "loss": 0.6442,
+      "step": 3233
+    },
+    {
+      "epoch": 1.1355337078651686,
+      "grad_norm": 0.8823835849761963,
+      "learning_rate": 4.59174425020464e-06,
+      "loss": 0.6075,
+      "step": 3234
+    },
+    {
+      "epoch": 1.1358848314606742,
+      "grad_norm": 0.9072199463844299,
+      "learning_rate": 4.591491015068135e-06,
+      "loss": 0.6295,
+      "step": 3235
+    },
+    {
+      "epoch": 1.1362359550561798,
+      "grad_norm": 0.8751364946365356,
+      "learning_rate": 4.591237708404427e-06,
+      "loss": 0.6398,
+      "step": 3236
+    },
+    {
+      "epoch": 1.1365870786516854,
+      "grad_norm": 0.8724543452262878,
+      "learning_rate": 4.590984330222179e-06,
+      "loss": 0.5811,
+      "step": 3237
+    },
+    {
+      "epoch": 1.136938202247191,
+      "grad_norm": 0.9179860949516296,
+      "learning_rate": 4.590730880530056e-06,
+      "loss": 0.6846,
+      "step": 3238
+    },
+    {
+      "epoch": 1.1372893258426966,
+      "grad_norm": 0.8932046890258789,
+      "learning_rate": 4.590477359336726e-06,
+      "loss": 0.6447,
+      "step": 3239
+    },
+    {
+      "epoch": 1.1376404494382022,
+      "grad_norm": 0.9021968245506287,
+      "learning_rate": 4.590223766650859e-06,
+      "loss": 0.6503,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1379915730337078,
+      "grad_norm": 0.8907482028007507,
+      "learning_rate": 4.5899701024811285e-06,
+      "loss": 0.6339,
+      "step": 3241
+    },
+    {
+      "epoch": 1.1383426966292134,
+      "grad_norm": 0.9072391390800476,
+      "learning_rate": 4.589716366836208e-06,
+      "loss": 0.65,
+      "step": 3242
+    },
+    {
+      "epoch": 1.1386938202247192,
+      "grad_norm": 0.8935176134109497,
+      "learning_rate": 4.5894625597247765e-06,
+      "loss": 0.6105,
+      "step": 3243
+    },
+    {
+      "epoch": 1.1390449438202248,
+      "grad_norm": 0.9080550670623779,
+      "learning_rate": 4.589208681155513e-06,
+      "loss": 0.6641,
+      "step": 3244
+    },
+    {
+      "epoch": 1.1393960674157304,
+      "grad_norm": 0.8920962810516357,
+      "learning_rate": 4.5889547311371006e-06,
+      "loss": 0.6482,
+      "step": 3245
+    },
+    {
+      "epoch": 1.139747191011236,
+      "grad_norm": 0.868358850479126,
+      "learning_rate": 4.588700709678224e-06,
+      "loss": 0.615,
+      "step": 3246
+    },
+    {
+      "epoch": 1.1400983146067416,
+      "grad_norm": 0.9104098677635193,
+      "learning_rate": 4.588446616787571e-06,
+      "loss": 0.6472,
+      "step": 3247
+    },
+    {
+      "epoch": 1.1404494382022472,
+      "grad_norm": 0.9489714503288269,
+      "learning_rate": 4.58819245247383e-06,
+      "loss": 0.6441,
+      "step": 3248
+    },
+    {
+      "epoch": 1.1408005617977528,
+      "grad_norm": 0.8975799679756165,
+      "learning_rate": 4.587938216745694e-06,
+      "loss": 0.6469,
+      "step": 3249
+    },
+    {
+      "epoch": 1.1411516853932584,
+      "grad_norm": 0.9056989550590515,
+      "learning_rate": 4.587683909611858e-06,
+      "loss": 0.6643,
+      "step": 3250
+    },
+    {
+      "epoch": 1.141502808988764,
+      "grad_norm": 0.890923023223877,
+      "learning_rate": 4.587429531081019e-06,
+      "loss": 0.6421,
+      "step": 3251
+    },
+    {
+      "epoch": 1.1418539325842696,
+      "grad_norm": 0.8975485563278198,
+      "learning_rate": 4.587175081161875e-06,
+      "loss": 0.6422,
+      "step": 3252
+    },
+    {
+      "epoch": 1.1422050561797752,
+      "grad_norm": 0.9034876823425293,
+      "learning_rate": 4.586920559863131e-06,
+      "loss": 0.6406,
+      "step": 3253
+    },
+    {
+      "epoch": 1.1425561797752808,
+      "grad_norm": 0.8907468318939209,
+      "learning_rate": 4.586665967193489e-06,
+      "loss": 0.6191,
+      "step": 3254
+    },
+    {
+      "epoch": 1.1429073033707866,
+      "grad_norm": 0.8739985227584839,
+      "learning_rate": 4.586411303161656e-06,
+      "loss": 0.6278,
+      "step": 3255
+    },
+    {
+      "epoch": 1.1432584269662922,
+      "grad_norm": 0.866232693195343,
+      "learning_rate": 4.586156567776343e-06,
+      "loss": 0.6497,
+      "step": 3256
+    },
+    {
+      "epoch": 1.1436095505617978,
+      "grad_norm": 0.8944258093833923,
+      "learning_rate": 4.585901761046259e-06,
+      "loss": 0.637,
+      "step": 3257
+    },
+    {
+      "epoch": 1.1439606741573034,
+      "grad_norm": 0.8632298707962036,
+      "learning_rate": 4.585646882980121e-06,
+      "loss": 0.6197,
+      "step": 3258
+    },
+    {
+      "epoch": 1.144311797752809,
+      "grad_norm": 0.8586609363555908,
+      "learning_rate": 4.585391933586644e-06,
+      "loss": 0.6163,
+      "step": 3259
+    },
+    {
+      "epoch": 1.1446629213483146,
+      "grad_norm": 0.8893579244613647,
+      "learning_rate": 4.585136912874547e-06,
+      "loss": 0.6173,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1450140449438202,
+      "grad_norm": 0.8902631998062134,
+      "learning_rate": 4.584881820852551e-06,
+      "loss": 0.644,
+      "step": 3261
+    },
+    {
+      "epoch": 1.1453651685393258,
+      "grad_norm": 0.8647865653038025,
+      "learning_rate": 4.584626657529382e-06,
+      "loss": 0.6156,
+      "step": 3262
+    },
+    {
+      "epoch": 1.1457162921348314,
+      "grad_norm": 0.8853922486305237,
+      "learning_rate": 4.584371422913765e-06,
+      "loss": 0.6659,
+      "step": 3263
+    },
+    {
+      "epoch": 1.146067415730337,
+      "grad_norm": 0.8677497506141663,
+      "learning_rate": 4.584116117014429e-06,
+      "loss": 0.6403,
+      "step": 3264
+    },
+    {
+      "epoch": 1.1464185393258428,
+      "grad_norm": 0.8832502961158752,
+      "learning_rate": 4.583860739840105e-06,
+      "loss": 0.6243,
+      "step": 3265
+    },
+    {
+      "epoch": 1.1467696629213484,
+      "grad_norm": 0.952691912651062,
+      "learning_rate": 4.583605291399527e-06,
+      "loss": 0.6237,
+      "step": 3266
+    },
+    {
+      "epoch": 1.147120786516854,
+      "grad_norm": 0.8712973594665527,
+      "learning_rate": 4.5833497717014306e-06,
+      "loss": 0.6303,
+      "step": 3267
+    },
+    {
+      "epoch": 1.1474719101123596,
+      "grad_norm": 0.9440521597862244,
+      "learning_rate": 4.583094180754556e-06,
+      "loss": 0.6276,
+      "step": 3268
+    },
+    {
+      "epoch": 1.1478230337078652,
+      "grad_norm": 0.862451434135437,
+      "learning_rate": 4.582838518567641e-06,
+      "loss": 0.6363,
+      "step": 3269
+    },
+    {
+      "epoch": 1.1481741573033708,
+      "grad_norm": 0.8932682275772095,
+      "learning_rate": 4.582582785149432e-06,
+      "loss": 0.6508,
+      "step": 3270
+    },
+    {
+      "epoch": 1.1485252808988764,
+      "grad_norm": 0.9098778963088989,
+      "learning_rate": 4.582326980508675e-06,
+      "loss": 0.6377,
+      "step": 3271
+    },
+    {
+      "epoch": 1.148876404494382,
+      "grad_norm": 0.9052352905273438,
+      "learning_rate": 4.582071104654115e-06,
+      "loss": 0.6199,
+      "step": 3272
+    },
+    {
+      "epoch": 1.1492275280898876,
+      "grad_norm": 0.8726509213447571,
+      "learning_rate": 4.581815157594507e-06,
+      "loss": 0.6352,
+      "step": 3273
+    },
+    {
+      "epoch": 1.1495786516853932,
+      "grad_norm": 0.8598045110702515,
+      "learning_rate": 4.5815591393386e-06,
+      "loss": 0.6432,
+      "step": 3274
+    },
+    {
+      "epoch": 1.1499297752808988,
+      "grad_norm": 0.9145620465278625,
+      "learning_rate": 4.581303049895153e-06,
+      "loss": 0.6256,
+      "step": 3275
+    },
+    {
+      "epoch": 1.1502808988764044,
+      "grad_norm": 0.9343267679214478,
+      "learning_rate": 4.5810468892729225e-06,
+      "loss": 0.6535,
+      "step": 3276
+    },
+    {
+      "epoch": 1.15063202247191,
+      "grad_norm": 0.891581654548645,
+      "learning_rate": 4.580790657480668e-06,
+      "loss": 0.7085,
+      "step": 3277
+    },
+    {
+      "epoch": 1.1509831460674158,
+      "grad_norm": 0.8569180369377136,
+      "learning_rate": 4.5805343545271555e-06,
+      "loss": 0.6165,
+      "step": 3278
+    },
+    {
+      "epoch": 1.1513342696629214,
+      "grad_norm": 0.856910765171051,
+      "learning_rate": 4.5802779804211475e-06,
+      "loss": 0.6347,
+      "step": 3279
+    },
+    {
+      "epoch": 1.151685393258427,
+      "grad_norm": 0.8790578842163086,
+      "learning_rate": 4.5800215351714124e-06,
+      "loss": 0.6518,
+      "step": 3280
+    },
+    {
+      "epoch": 1.1520365168539326,
+      "grad_norm": 0.8809064626693726,
+      "learning_rate": 4.5797650187867205e-06,
+      "loss": 0.654,
+      "step": 3281
+    },
+    {
+      "epoch": 1.1523876404494382,
+      "grad_norm": 0.8870155215263367,
+      "learning_rate": 4.579508431275845e-06,
+      "loss": 0.651,
+      "step": 3282
+    },
+    {
+      "epoch": 1.1527387640449438,
+      "grad_norm": 0.8868922591209412,
+      "learning_rate": 4.57925177264756e-06,
+      "loss": 0.6077,
+      "step": 3283
+    },
+    {
+      "epoch": 1.1530898876404494,
+      "grad_norm": 0.8815183639526367,
+      "learning_rate": 4.578995042910645e-06,
+      "loss": 0.6332,
+      "step": 3284
+    },
+    {
+      "epoch": 1.153441011235955,
+      "grad_norm": 0.8861362934112549,
+      "learning_rate": 4.578738242073878e-06,
+      "loss": 0.64,
+      "step": 3285
+    },
+    {
+      "epoch": 1.1537921348314606,
+      "grad_norm": 0.8777415156364441,
+      "learning_rate": 4.5784813701460415e-06,
+      "loss": 0.6357,
+      "step": 3286
+    },
+    {
+      "epoch": 1.1541432584269664,
+      "grad_norm": 0.924778163433075,
+      "learning_rate": 4.578224427135921e-06,
+      "loss": 0.6345,
+      "step": 3287
+    },
+    {
+      "epoch": 1.154494382022472,
+      "grad_norm": 0.8765180110931396,
+      "learning_rate": 4.577967413052303e-06,
+      "loss": 0.6345,
+      "step": 3288
+    },
+    {
+      "epoch": 1.1548455056179776,
+      "grad_norm": 0.8863521814346313,
+      "learning_rate": 4.577710327903978e-06,
+      "loss": 0.6391,
+      "step": 3289
+    },
+    {
+      "epoch": 1.1551966292134832,
+      "grad_norm": 0.8893427848815918,
+      "learning_rate": 4.5774531716997384e-06,
+      "loss": 0.6325,
+      "step": 3290
+    },
+    {
+      "epoch": 1.1555477528089888,
+      "grad_norm": 0.8832981586456299,
+      "learning_rate": 4.577195944448377e-06,
+      "loss": 0.625,
+      "step": 3291
+    },
+    {
+      "epoch": 1.1558988764044944,
+      "grad_norm": 0.8558886647224426,
+      "learning_rate": 4.576938646158693e-06,
+      "loss": 0.5583,
+      "step": 3292
+    },
+    {
+      "epoch": 1.15625,
+      "grad_norm": 0.9013493657112122,
+      "learning_rate": 4.576681276839483e-06,
+      "loss": 0.641,
+      "step": 3293
+    },
+    {
+      "epoch": 1.1566011235955056,
+      "grad_norm": 0.8543421626091003,
+      "learning_rate": 4.576423836499552e-06,
+      "loss": 0.6452,
+      "step": 3294
+    },
+    {
+      "epoch": 1.1569522471910112,
+      "grad_norm": 0.8888267874717712,
+      "learning_rate": 4.5761663251477006e-06,
+      "loss": 0.6525,
+      "step": 3295
+    },
+    {
+      "epoch": 1.1573033707865168,
+      "grad_norm": 0.8754894733428955,
+      "learning_rate": 4.575908742792739e-06,
+      "loss": 0.631,
+      "step": 3296
+    },
+    {
+      "epoch": 1.1576544943820224,
+      "grad_norm": 0.8829214572906494,
+      "learning_rate": 4.575651089443474e-06,
+      "loss": 0.6478,
+      "step": 3297
+    },
+    {
+      "epoch": 1.158005617977528,
+      "grad_norm": 0.8613840937614441,
+      "learning_rate": 4.575393365108719e-06,
+      "loss": 0.5818,
+      "step": 3298
+    },
+    {
+      "epoch": 1.1583567415730336,
+      "grad_norm": 0.850951611995697,
+      "learning_rate": 4.575135569797285e-06,
+      "loss": 0.6349,
+      "step": 3299
+    },
+    {
+      "epoch": 1.1587078651685394,
+      "grad_norm": 0.9041875004768372,
+      "learning_rate": 4.574877703517992e-06,
+      "loss": 0.6544,
+      "step": 3300
+    },
+    {
+      "epoch": 1.159058988764045,
+      "grad_norm": 0.8524355888366699,
+      "learning_rate": 4.5746197662796565e-06,
+      "loss": 0.6145,
+      "step": 3301
+    },
+    {
+      "epoch": 1.1594101123595506,
+      "grad_norm": 0.8648415803909302,
+      "learning_rate": 4.5743617580911e-06,
+      "loss": 0.6221,
+      "step": 3302
+    },
+    {
+      "epoch": 1.1597612359550562,
+      "grad_norm": 0.8579859733581543,
+      "learning_rate": 4.574103678961147e-06,
+      "loss": 0.6011,
+      "step": 3303
+    },
+    {
+      "epoch": 1.1601123595505618,
+      "grad_norm": 0.8796312808990479,
+      "learning_rate": 4.5738455288986225e-06,
+      "loss": 0.6453,
+      "step": 3304
+    },
+    {
+      "epoch": 1.1604634831460674,
+      "grad_norm": 0.871971845626831,
+      "learning_rate": 4.573587307912356e-06,
+      "loss": 0.6096,
+      "step": 3305
+    },
+    {
+      "epoch": 1.160814606741573,
+      "grad_norm": 0.8905230760574341,
+      "learning_rate": 4.573329016011177e-06,
+      "loss": 0.6272,
+      "step": 3306
+    },
+    {
+      "epoch": 1.1611657303370786,
+      "grad_norm": 0.8950706720352173,
+      "learning_rate": 4.57307065320392e-06,
+      "loss": 0.6285,
+      "step": 3307
+    },
+    {
+      "epoch": 1.1615168539325842,
+      "grad_norm": 0.8785974979400635,
+      "learning_rate": 4.572812219499422e-06,
+      "loss": 0.6175,
+      "step": 3308
+    },
+    {
+      "epoch": 1.16186797752809,
+      "grad_norm": 0.8762933015823364,
+      "learning_rate": 4.572553714906519e-06,
+      "loss": 0.6128,
+      "step": 3309
+    },
+    {
+      "epoch": 1.1622191011235956,
+      "grad_norm": 0.9125222563743591,
+      "learning_rate": 4.572295139434052e-06,
+      "loss": 0.634,
+      "step": 3310
+    },
+    {
+      "epoch": 1.1625702247191012,
+      "grad_norm": 0.8794988989830017,
+      "learning_rate": 4.572036493090864e-06,
+      "loss": 0.6494,
+      "step": 3311
+    },
+    {
+      "epoch": 1.1629213483146068,
+      "grad_norm": 0.901485800743103,
+      "learning_rate": 4.571777775885803e-06,
+      "loss": 0.6836,
+      "step": 3312
+    },
+    {
+      "epoch": 1.1632724719101124,
+      "grad_norm": 0.9252587556838989,
+      "learning_rate": 4.5715189878277135e-06,
+      "loss": 0.6082,
+      "step": 3313
+    },
+    {
+      "epoch": 1.163623595505618,
+      "grad_norm": 0.8742091655731201,
+      "learning_rate": 4.571260128925448e-06,
+      "loss": 0.6542,
+      "step": 3314
+    },
+    {
+      "epoch": 1.1639747191011236,
+      "grad_norm": 0.8586685657501221,
+      "learning_rate": 4.571001199187857e-06,
+      "loss": 0.6165,
+      "step": 3315
+    },
+    {
+      "epoch": 1.1643258426966292,
+      "grad_norm": 0.8647174835205078,
+      "learning_rate": 4.570742198623798e-06,
+      "loss": 0.6262,
+      "step": 3316
+    },
+    {
+      "epoch": 1.1646769662921348,
+      "grad_norm": 0.8538329601287842,
+      "learning_rate": 4.570483127242129e-06,
+      "loss": 0.6523,
+      "step": 3317
+    },
+    {
+      "epoch": 1.1650280898876404,
+      "grad_norm": 0.8604421019554138,
+      "learning_rate": 4.570223985051707e-06,
+      "loss": 0.6367,
+      "step": 3318
+    },
+    {
+      "epoch": 1.165379213483146,
+      "grad_norm": 0.9220116138458252,
+      "learning_rate": 4.569964772061397e-06,
+      "loss": 0.6565,
+      "step": 3319
+    },
+    {
+      "epoch": 1.1657303370786516,
+      "grad_norm": 0.8763860464096069,
+      "learning_rate": 4.569705488280063e-06,
+      "loss": 0.6075,
+      "step": 3320
+    },
+    {
+      "epoch": 1.1660814606741572,
+      "grad_norm": 0.8693764805793762,
+      "learning_rate": 4.569446133716572e-06,
+      "loss": 0.596,
+      "step": 3321
+    },
+    {
+      "epoch": 1.166432584269663,
+      "grad_norm": 0.8876716494560242,
+      "learning_rate": 4.569186708379795e-06,
+      "loss": 0.6617,
+      "step": 3322
+    },
+    {
+      "epoch": 1.1667837078651686,
+      "grad_norm": 0.9221491813659668,
+      "learning_rate": 4.568927212278602e-06,
+      "loss": 0.6395,
+      "step": 3323
+    },
+    {
+      "epoch": 1.1671348314606742,
+      "grad_norm": 0.9208928346633911,
+      "learning_rate": 4.56866764542187e-06,
+      "loss": 0.6224,
+      "step": 3324
+    },
+    {
+      "epoch": 1.1674859550561798,
+      "grad_norm": 0.8737736344337463,
+      "learning_rate": 4.5684080078184745e-06,
+      "loss": 0.6247,
+      "step": 3325
+    },
+    {
+      "epoch": 1.1678370786516854,
+      "grad_norm": 0.8767399787902832,
+      "learning_rate": 4.568148299477294e-06,
+      "loss": 0.6394,
+      "step": 3326
+    },
+    {
+      "epoch": 1.168188202247191,
+      "grad_norm": 0.8892948031425476,
+      "learning_rate": 4.567888520407212e-06,
+      "loss": 0.6176,
+      "step": 3327
+    },
+    {
+      "epoch": 1.1685393258426966,
+      "grad_norm": 0.858261227607727,
+      "learning_rate": 4.567628670617113e-06,
+      "loss": 0.611,
+      "step": 3328
+    },
+    {
+      "epoch": 1.1688904494382022,
+      "grad_norm": 0.8876168131828308,
+      "learning_rate": 4.567368750115882e-06,
+      "loss": 0.6314,
+      "step": 3329
+    },
+    {
+      "epoch": 1.1692415730337078,
+      "grad_norm": 0.9200357794761658,
+      "learning_rate": 4.567108758912408e-06,
+      "loss": 0.6621,
+      "step": 3330
+    },
+    {
+      "epoch": 1.1695926966292134,
+      "grad_norm": 0.8937010765075684,
+      "learning_rate": 4.5668486970155845e-06,
+      "loss": 0.6381,
+      "step": 3331
+    },
+    {
+      "epoch": 1.1699438202247192,
+      "grad_norm": 0.9102723002433777,
+      "learning_rate": 4.566588564434304e-06,
+      "loss": 0.6355,
+      "step": 3332
+    },
+    {
+      "epoch": 1.1702949438202248,
+      "grad_norm": 0.8944071531295776,
+      "learning_rate": 4.566328361177463e-06,
+      "loss": 0.618,
+      "step": 3333
+    },
+    {
+      "epoch": 1.1706460674157304,
+      "grad_norm": 0.9432087540626526,
+      "learning_rate": 4.56606808725396e-06,
+      "loss": 0.6654,
+      "step": 3334
+    },
+    {
+      "epoch": 1.170997191011236,
+      "grad_norm": 0.8926756978034973,
+      "learning_rate": 4.5658077426726965e-06,
+      "loss": 0.6635,
+      "step": 3335
+    },
+    {
+      "epoch": 1.1713483146067416,
+      "grad_norm": 0.9085322022438049,
+      "learning_rate": 4.5655473274425756e-06,
+      "loss": 0.625,
+      "step": 3336
+    },
+    {
+      "epoch": 1.1716994382022472,
+      "grad_norm": 0.8761731386184692,
+      "learning_rate": 4.565286841572504e-06,
+      "loss": 0.6263,
+      "step": 3337
+    },
+    {
+      "epoch": 1.1720505617977528,
+      "grad_norm": 0.902802586555481,
+      "learning_rate": 4.5650262850713895e-06,
+      "loss": 0.5946,
+      "step": 3338
+    },
+    {
+      "epoch": 1.1724016853932584,
+      "grad_norm": 0.9183904528617859,
+      "learning_rate": 4.564765657948143e-06,
+      "loss": 0.6103,
+      "step": 3339
+    },
+    {
+      "epoch": 1.172752808988764,
+      "grad_norm": 0.8977819085121155,
+      "learning_rate": 4.564504960211678e-06,
+      "loss": 0.6408,
+      "step": 3340
+    },
+    {
+      "epoch": 1.1731039325842696,
+      "grad_norm": 0.9143522381782532,
+      "learning_rate": 4.56424419187091e-06,
+      "loss": 0.6379,
+      "step": 3341
+    },
+    {
+      "epoch": 1.1734550561797752,
+      "grad_norm": 0.9088187217712402,
+      "learning_rate": 4.563983352934758e-06,
+      "loss": 0.6473,
+      "step": 3342
+    },
+    {
+      "epoch": 1.1738061797752808,
+      "grad_norm": 0.8903788328170776,
+      "learning_rate": 4.563722443412139e-06,
+      "loss": 0.5842,
+      "step": 3343
+    },
+    {
+      "epoch": 1.1741573033707866,
+      "grad_norm": 0.8306713700294495,
+      "learning_rate": 4.563461463311981e-06,
+      "loss": 0.5916,
+      "step": 3344
+    },
+    {
+      "epoch": 1.1745084269662922,
+      "grad_norm": 0.8774393200874329,
+      "learning_rate": 4.5632004126432055e-06,
+      "loss": 0.628,
+      "step": 3345
+    },
+    {
+      "epoch": 1.1748595505617978,
+      "grad_norm": 0.8590057492256165,
+      "learning_rate": 4.5629392914147415e-06,
+      "loss": 0.6111,
+      "step": 3346
+    },
+    {
+      "epoch": 1.1752106741573034,
+      "grad_norm": 0.8803392648696899,
+      "learning_rate": 4.562678099635519e-06,
+      "loss": 0.6313,
+      "step": 3347
+    },
+    {
+      "epoch": 1.175561797752809,
+      "grad_norm": 0.8844352960586548,
+      "learning_rate": 4.562416837314471e-06,
+      "loss": 0.6392,
+      "step": 3348
+    },
+    {
+      "epoch": 1.1759129213483146,
+      "grad_norm": 0.9030863642692566,
+      "learning_rate": 4.5621555044605315e-06,
+      "loss": 0.6407,
+      "step": 3349
+    },
+    {
+      "epoch": 1.1762640449438202,
+      "grad_norm": 0.8938463926315308,
+      "learning_rate": 4.561894101082638e-06,
+      "loss": 0.6224,
+      "step": 3350
+    },
+    {
+      "epoch": 1.1766151685393258,
+      "grad_norm": 0.869136393070221,
+      "learning_rate": 4.5616326271897315e-06,
+      "loss": 0.6239,
+      "step": 3351
+    },
+    {
+      "epoch": 1.1769662921348314,
+      "grad_norm": 0.8866480588912964,
+      "learning_rate": 4.5613710827907535e-06,
+      "loss": 0.6319,
+      "step": 3352
+    },
+    {
+      "epoch": 1.177317415730337,
+      "grad_norm": 0.911381185054779,
+      "learning_rate": 4.561109467894647e-06,
+      "loss": 0.6779,
+      "step": 3353
+    },
+    {
+      "epoch": 1.1776685393258428,
+      "grad_norm": 0.8881227374076843,
+      "learning_rate": 4.560847782510361e-06,
+      "loss": 0.6174,
+      "step": 3354
+    },
+    {
+      "epoch": 1.1780196629213484,
+      "grad_norm": 0.8753611445426941,
+      "learning_rate": 4.560586026646845e-06,
+      "loss": 0.6337,
+      "step": 3355
+    },
+    {
+      "epoch": 1.178370786516854,
+      "grad_norm": 0.898257315158844,
+      "learning_rate": 4.56032420031305e-06,
+      "loss": 0.6547,
+      "step": 3356
+    },
+    {
+      "epoch": 1.1787219101123596,
+      "grad_norm": 0.871589720249176,
+      "learning_rate": 4.56006230351793e-06,
+      "loss": 0.6093,
+      "step": 3357
+    },
+    {
+      "epoch": 1.1790730337078652,
+      "grad_norm": 0.8686171174049377,
+      "learning_rate": 4.559800336270443e-06,
+      "loss": 0.6396,
+      "step": 3358
+    },
+    {
+      "epoch": 1.1794241573033708,
+      "grad_norm": 0.8978163599967957,
+      "learning_rate": 4.559538298579546e-06,
+      "loss": 0.6505,
+      "step": 3359
+    },
+    {
+      "epoch": 1.1797752808988764,
+      "grad_norm": 0.9143759608268738,
+      "learning_rate": 4.559276190454201e-06,
+      "loss": 0.6452,
+      "step": 3360
+    },
+    {
+      "epoch": 1.180126404494382,
+      "grad_norm": 0.9073001146316528,
+      "learning_rate": 4.559014011903373e-06,
+      "loss": 0.6252,
+      "step": 3361
+    },
+    {
+      "epoch": 1.1804775280898876,
+      "grad_norm": 0.9082154035568237,
+      "learning_rate": 4.558751762936029e-06,
+      "loss": 0.6401,
+      "step": 3362
+    },
+    {
+      "epoch": 1.1808286516853932,
+      "grad_norm": 0.949808657169342,
+      "learning_rate": 4.558489443561135e-06,
+      "loss": 0.6716,
+      "step": 3363
+    },
+    {
+      "epoch": 1.1811797752808988,
+      "grad_norm": 0.8602761626243591,
+      "learning_rate": 4.558227053787664e-06,
+      "loss": 0.6112,
+      "step": 3364
+    },
+    {
+      "epoch": 1.1815308988764044,
+      "grad_norm": 0.9209701418876648,
+      "learning_rate": 4.5579645936245886e-06,
+      "loss": 0.6205,
+      "step": 3365
+    },
+    {
+      "epoch": 1.18188202247191,
+      "grad_norm": 0.8840071558952332,
+      "learning_rate": 4.557702063080886e-06,
+      "loss": 0.6242,
+      "step": 3366
+    },
+    {
+      "epoch": 1.1822331460674158,
+      "grad_norm": 0.906648576259613,
+      "learning_rate": 4.557439462165532e-06,
+      "loss": 0.6455,
+      "step": 3367
+    },
+    {
+      "epoch": 1.1825842696629214,
+      "grad_norm": 0.8441259264945984,
+      "learning_rate": 4.55717679088751e-06,
+      "loss": 0.5916,
+      "step": 3368
+    },
+    {
+      "epoch": 1.182935393258427,
+      "grad_norm": 0.9262234568595886,
+      "learning_rate": 4.556914049255803e-06,
+      "loss": 0.659,
+      "step": 3369
+    },
+    {
+      "epoch": 1.1832865168539326,
+      "grad_norm": 0.8849537372589111,
+      "learning_rate": 4.556651237279394e-06,
+      "loss": 0.6503,
+      "step": 3370
+    },
+    {
+      "epoch": 1.1836376404494382,
+      "grad_norm": 0.8733585476875305,
+      "learning_rate": 4.5563883549672735e-06,
+      "loss": 0.6309,
+      "step": 3371
+    },
+    {
+      "epoch": 1.1839887640449438,
+      "grad_norm": 0.91927570104599,
+      "learning_rate": 4.556125402328431e-06,
+      "loss": 0.6148,
+      "step": 3372
+    },
+    {
+      "epoch": 1.1843398876404494,
+      "grad_norm": 0.9334448575973511,
+      "learning_rate": 4.555862379371858e-06,
+      "loss": 0.6483,
+      "step": 3373
+    },
+    {
+      "epoch": 1.184691011235955,
+      "grad_norm": 0.8952125906944275,
+      "learning_rate": 4.555599286106553e-06,
+      "loss": 0.6218,
+      "step": 3374
+    },
+    {
+      "epoch": 1.1850421348314606,
+      "grad_norm": 0.8810850381851196,
+      "learning_rate": 4.55533612254151e-06,
+      "loss": 0.6391,
+      "step": 3375
+    },
+    {
+      "epoch": 1.1853932584269664,
+      "grad_norm": 0.9044777154922485,
+      "learning_rate": 4.5550728886857305e-06,
+      "loss": 0.6462,
+      "step": 3376
+    },
+    {
+      "epoch": 1.185744382022472,
+      "grad_norm": 0.9001294374465942,
+      "learning_rate": 4.554809584548217e-06,
+      "loss": 0.609,
+      "step": 3377
+    },
+    {
+      "epoch": 1.1860955056179776,
+      "grad_norm": 0.8865534663200378,
+      "learning_rate": 4.554546210137973e-06,
+      "loss": 0.6016,
+      "step": 3378
+    },
+    {
+      "epoch": 1.1864466292134832,
+      "grad_norm": 0.8913449645042419,
+      "learning_rate": 4.554282765464009e-06,
+      "loss": 0.6301,
+      "step": 3379
+    },
+    {
+      "epoch": 1.1867977528089888,
+      "grad_norm": 0.8913347125053406,
+      "learning_rate": 4.55401925053533e-06,
+      "loss": 0.6369,
+      "step": 3380
+    },
+    {
+      "epoch": 1.1871488764044944,
+      "grad_norm": 0.9004356265068054,
+      "learning_rate": 4.553755665360952e-06,
+      "loss": 0.646,
+      "step": 3381
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 0.85580974817276,
+      "learning_rate": 4.553492009949886e-06,
+      "loss": 0.6096,
+      "step": 3382
+    },
+    {
+      "epoch": 1.1878511235955056,
+      "grad_norm": 0.860244631767273,
+      "learning_rate": 4.553228284311152e-06,
+      "loss": 0.6043,
+      "step": 3383
+    },
+    {
+      "epoch": 1.1882022471910112,
+      "grad_norm": 0.8783630132675171,
+      "learning_rate": 4.552964488453767e-06,
+      "loss": 0.6359,
+      "step": 3384
+    },
+    {
+      "epoch": 1.1885533707865168,
+      "grad_norm": 0.8642095923423767,
+      "learning_rate": 4.552700622386753e-06,
+      "loss": 0.6038,
+      "step": 3385
+    },
+    {
+      "epoch": 1.1889044943820224,
+      "grad_norm": 0.867074728012085,
+      "learning_rate": 4.5524366861191345e-06,
+      "loss": 0.6415,
+      "step": 3386
+    },
+    {
+      "epoch": 1.189255617977528,
+      "grad_norm": 0.8984413743019104,
+      "learning_rate": 4.552172679659938e-06,
+      "loss": 0.6519,
+      "step": 3387
+    },
+    {
+      "epoch": 1.1896067415730336,
+      "grad_norm": 0.869999885559082,
+      "learning_rate": 4.551908603018191e-06,
+      "loss": 0.6269,
+      "step": 3388
+    },
+    {
+      "epoch": 1.1899578651685394,
+      "grad_norm": 0.8916454911231995,
+      "learning_rate": 4.551644456202926e-06,
+      "loss": 0.6486,
+      "step": 3389
+    },
+    {
+      "epoch": 1.190308988764045,
+      "grad_norm": 0.888932466506958,
+      "learning_rate": 4.551380239223176e-06,
+      "loss": 0.6084,
+      "step": 3390
+    },
+    {
+      "epoch": 1.1906601123595506,
+      "grad_norm": 0.9049292802810669,
+      "learning_rate": 4.5511159520879785e-06,
+      "loss": 0.6134,
+      "step": 3391
+    },
+    {
+      "epoch": 1.1910112359550562,
+      "grad_norm": 0.8797271847724915,
+      "learning_rate": 4.550851594806369e-06,
+      "loss": 0.6367,
+      "step": 3392
+    },
+    {
+      "epoch": 1.1913623595505618,
+      "grad_norm": 0.9155986309051514,
+      "learning_rate": 4.550587167387391e-06,
+      "loss": 0.6415,
+      "step": 3393
+    },
+    {
+      "epoch": 1.1917134831460674,
+      "grad_norm": 0.8914098143577576,
+      "learning_rate": 4.550322669840086e-06,
+      "loss": 0.6352,
+      "step": 3394
+    },
+    {
+      "epoch": 1.192064606741573,
+      "grad_norm": 0.8818766474723816,
+      "learning_rate": 4.550058102173501e-06,
+      "loss": 0.6219,
+      "step": 3395
+    },
+    {
+      "epoch": 1.1924157303370786,
+      "grad_norm": 0.8712998628616333,
+      "learning_rate": 4.5497934643966825e-06,
+      "loss": 0.5773,
+      "step": 3396
+    },
+    {
+      "epoch": 1.1927668539325842,
+      "grad_norm": 0.8595727682113647,
+      "learning_rate": 4.5495287565186806e-06,
+      "loss": 0.6171,
+      "step": 3397
+    },
+    {
+      "epoch": 1.19311797752809,
+      "grad_norm": 0.8803826570510864,
+      "learning_rate": 4.549263978548551e-06,
+      "loss": 0.5942,
+      "step": 3398
+    },
+    {
+      "epoch": 1.1934691011235956,
+      "grad_norm": 0.8828043937683105,
+      "learning_rate": 4.548999130495345e-06,
+      "loss": 0.651,
+      "step": 3399
+    },
+    {
+      "epoch": 1.1938202247191012,
+      "grad_norm": 0.8621920347213745,
+      "learning_rate": 4.548734212368123e-06,
+      "loss": 0.5715,
+      "step": 3400
+    },
+    {
+      "epoch": 1.1941713483146068,
+      "grad_norm": 0.9156777858734131,
+      "learning_rate": 4.548469224175944e-06,
+      "loss": 0.6669,
+      "step": 3401
+    },
+    {
+      "epoch": 1.1945224719101124,
+      "grad_norm": 0.8646230101585388,
+      "learning_rate": 4.548204165927871e-06,
+      "loss": 0.5863,
+      "step": 3402
+    },
+    {
+      "epoch": 1.194873595505618,
+      "grad_norm": 0.9154274463653564,
+      "learning_rate": 4.5479390376329676e-06,
+      "loss": 0.6236,
+      "step": 3403
+    },
+    {
+      "epoch": 1.1952247191011236,
+      "grad_norm": 0.8905937671661377,
+      "learning_rate": 4.547673839300301e-06,
+      "loss": 0.5949,
+      "step": 3404
+    },
+    {
+      "epoch": 1.1955758426966292,
+      "grad_norm": 0.9460303783416748,
+      "learning_rate": 4.547408570938942e-06,
+      "loss": 0.6367,
+      "step": 3405
+    },
+    {
+      "epoch": 1.1959269662921348,
+      "grad_norm": 0.8888306617736816,
+      "learning_rate": 4.547143232557962e-06,
+      "loss": 0.6533,
+      "step": 3406
+    },
+    {
+      "epoch": 1.1962780898876404,
+      "grad_norm": 0.8990828394889832,
+      "learning_rate": 4.546877824166436e-06,
+      "loss": 0.6418,
+      "step": 3407
+    },
+    {
+      "epoch": 1.196629213483146,
+      "grad_norm": 0.9120289087295532,
+      "learning_rate": 4.5466123457734385e-06,
+      "loss": 0.6308,
+      "step": 3408
+    },
+    {
+      "epoch": 1.1969803370786516,
+      "grad_norm": 0.8693567514419556,
+      "learning_rate": 4.546346797388051e-06,
+      "loss": 0.6784,
+      "step": 3409
+    },
+    {
+      "epoch": 1.1973314606741572,
+      "grad_norm": 0.8695741295814514,
+      "learning_rate": 4.546081179019353e-06,
+      "loss": 0.6344,
+      "step": 3410
+    },
+    {
+      "epoch": 1.197682584269663,
+      "grad_norm": 0.8742298483848572,
+      "learning_rate": 4.545815490676431e-06,
+      "loss": 0.584,
+      "step": 3411
+    },
+    {
+      "epoch": 1.1980337078651686,
+      "grad_norm": 0.8857693076133728,
+      "learning_rate": 4.545549732368368e-06,
+      "loss": 0.6591,
+      "step": 3412
+    },
+    {
+      "epoch": 1.1983848314606742,
+      "grad_norm": 0.887653648853302,
+      "learning_rate": 4.545283904104257e-06,
+      "loss": 0.6695,
+      "step": 3413
+    },
+    {
+      "epoch": 1.1987359550561798,
+      "grad_norm": 0.9010708332061768,
+      "learning_rate": 4.545018005893185e-06,
+      "loss": 0.6184,
+      "step": 3414
+    },
+    {
+      "epoch": 1.1990870786516854,
+      "grad_norm": 0.9171410799026489,
+      "learning_rate": 4.544752037744247e-06,
+      "loss": 0.6684,
+      "step": 3415
+    },
+    {
+      "epoch": 1.199438202247191,
+      "grad_norm": 0.8993377685546875,
+      "learning_rate": 4.5444859996665395e-06,
+      "loss": 0.6397,
+      "step": 3416
+    },
+    {
+      "epoch": 1.1997893258426966,
+      "grad_norm": 0.8996485471725464,
+      "learning_rate": 4.54421989166916e-06,
+      "loss": 0.6625,
+      "step": 3417
+    },
+    {
+      "epoch": 1.2001404494382022,
+      "grad_norm": 0.8873034715652466,
+      "learning_rate": 4.54395371376121e-06,
+      "loss": 0.5966,
+      "step": 3418
+    },
+    {
+      "epoch": 1.2004915730337078,
+      "grad_norm": 0.9241093397140503,
+      "learning_rate": 4.543687465951791e-06,
+      "loss": 0.6443,
+      "step": 3419
+    },
+    {
+      "epoch": 1.2008426966292134,
+      "grad_norm": 0.9346911907196045,
+      "learning_rate": 4.5434211482500116e-06,
+      "loss": 0.6544,
+      "step": 3420
+    },
+    {
+      "epoch": 1.2011938202247192,
+      "grad_norm": 0.8910781741142273,
+      "learning_rate": 4.543154760664975e-06,
+      "loss": 0.6357,
+      "step": 3421
+    },
+    {
+      "epoch": 1.2015449438202248,
+      "grad_norm": 0.8794376850128174,
+      "learning_rate": 4.542888303205796e-06,
+      "loss": 0.642,
+      "step": 3422
+    },
+    {
+      "epoch": 1.2018960674157304,
+      "grad_norm": 0.8754070401191711,
+      "learning_rate": 4.542621775881584e-06,
+      "loss": 0.6145,
+      "step": 3423
+    },
+    {
+      "epoch": 1.202247191011236,
+      "grad_norm": 0.881120502948761,
+      "learning_rate": 4.542355178701456e-06,
+      "loss": 0.6168,
+      "step": 3424
+    },
+    {
+      "epoch": 1.2025983146067416,
+      "grad_norm": 0.9401150941848755,
+      "learning_rate": 4.5420885116745275e-06,
+      "loss": 0.6636,
+      "step": 3425
+    },
+    {
+      "epoch": 1.2029494382022472,
+      "grad_norm": 0.9032625555992126,
+      "learning_rate": 4.541821774809919e-06,
+      "loss": 0.6496,
+      "step": 3426
+    },
+    {
+      "epoch": 1.2033005617977528,
+      "grad_norm": 0.9384265542030334,
+      "learning_rate": 4.541554968116754e-06,
+      "loss": 0.668,
+      "step": 3427
+    },
+    {
+      "epoch": 1.2036516853932584,
+      "grad_norm": 0.8976498246192932,
+      "learning_rate": 4.541288091604156e-06,
+      "loss": 0.6287,
+      "step": 3428
+    },
+    {
+      "epoch": 1.204002808988764,
+      "grad_norm": 0.9575441479682922,
+      "learning_rate": 4.5410211452812515e-06,
+      "loss": 0.657,
+      "step": 3429
+    },
+    {
+      "epoch": 1.2043539325842696,
+      "grad_norm": 0.8660023212432861,
+      "learning_rate": 4.540754129157171e-06,
+      "loss": 0.6206,
+      "step": 3430
+    },
+    {
+      "epoch": 1.2047050561797752,
+      "grad_norm": 0.8459751605987549,
+      "learning_rate": 4.540487043241045e-06,
+      "loss": 0.6318,
+      "step": 3431
+    },
+    {
+      "epoch": 1.2050561797752808,
+      "grad_norm": 0.8574265241622925,
+      "learning_rate": 4.540219887542009e-06,
+      "loss": 0.6434,
+      "step": 3432
+    },
+    {
+      "epoch": 1.2054073033707866,
+      "grad_norm": 0.8773812055587769,
+      "learning_rate": 4.539952662069198e-06,
+      "loss": 0.6217,
+      "step": 3433
+    },
+    {
+      "epoch": 1.2057584269662922,
+      "grad_norm": 0.8617153763771057,
+      "learning_rate": 4.539685366831752e-06,
+      "loss": 0.6403,
+      "step": 3434
+    },
+    {
+      "epoch": 1.2061095505617978,
+      "grad_norm": 0.8632030487060547,
+      "learning_rate": 4.539418001838812e-06,
+      "loss": 0.6054,
+      "step": 3435
+    },
+    {
+      "epoch": 1.2064606741573034,
+      "grad_norm": 0.8510541319847107,
+      "learning_rate": 4.539150567099521e-06,
+      "loss": 0.6207,
+      "step": 3436
+    },
+    {
+      "epoch": 1.206811797752809,
+      "grad_norm": 0.8945329785346985,
+      "learning_rate": 4.538883062623025e-06,
+      "loss": 0.6413,
+      "step": 3437
+    },
+    {
+      "epoch": 1.2071629213483146,
+      "grad_norm": 0.9274084568023682,
+      "learning_rate": 4.538615488418474e-06,
+      "loss": 0.647,
+      "step": 3438
+    },
+    {
+      "epoch": 1.2075140449438202,
+      "grad_norm": 0.8531306385993958,
+      "learning_rate": 4.538347844495018e-06,
+      "loss": 0.6365,
+      "step": 3439
+    },
+    {
+      "epoch": 1.2078651685393258,
+      "grad_norm": 0.9376130700111389,
+      "learning_rate": 4.538080130861809e-06,
+      "loss": 0.6194,
+      "step": 3440
+    },
+    {
+      "epoch": 1.2082162921348314,
+      "grad_norm": 0.8854097723960876,
+      "learning_rate": 4.537812347528004e-06,
+      "loss": 0.6196,
+      "step": 3441
+    },
+    {
+      "epoch": 1.208567415730337,
+      "grad_norm": 0.8933984637260437,
+      "learning_rate": 4.537544494502761e-06,
+      "loss": 0.6661,
+      "step": 3442
+    },
+    {
+      "epoch": 1.2089185393258428,
+      "grad_norm": 0.8640488386154175,
+      "learning_rate": 4.537276571795238e-06,
+      "loss": 0.5997,
+      "step": 3443
+    },
+    {
+      "epoch": 1.2092696629213484,
+      "grad_norm": 0.8864790201187134,
+      "learning_rate": 4.537008579414601e-06,
+      "loss": 0.6444,
+      "step": 3444
+    },
+    {
+      "epoch": 1.209620786516854,
+      "grad_norm": 0.8434916138648987,
+      "learning_rate": 4.536740517370013e-06,
+      "loss": 0.6138,
+      "step": 3445
+    },
+    {
+      "epoch": 1.2099719101123596,
+      "grad_norm": 0.8927525281906128,
+      "learning_rate": 4.536472385670643e-06,
+      "loss": 0.6309,
+      "step": 3446
+    },
+    {
+      "epoch": 1.2103230337078652,
+      "grad_norm": 0.8696897625923157,
+      "learning_rate": 4.53620418432566e-06,
+      "loss": 0.655,
+      "step": 3447
+    },
+    {
+      "epoch": 1.2106741573033708,
+      "grad_norm": 0.8829634189605713,
+      "learning_rate": 4.535935913344236e-06,
+      "loss": 0.623,
+      "step": 3448
+    },
+    {
+      "epoch": 1.2110252808988764,
+      "grad_norm": 0.8809219598770142,
+      "learning_rate": 4.535667572735546e-06,
+      "loss": 0.6806,
+      "step": 3449
+    },
+    {
+      "epoch": 1.211376404494382,
+      "grad_norm": 0.8975908756256104,
+      "learning_rate": 4.535399162508767e-06,
+      "loss": 0.6567,
+      "step": 3450
+    },
+    {
+      "epoch": 1.2117275280898876,
+      "grad_norm": 0.8763452768325806,
+      "learning_rate": 4.535130682673078e-06,
+      "loss": 0.6758,
+      "step": 3451
+    },
+    {
+      "epoch": 1.2120786516853932,
+      "grad_norm": 0.8537138104438782,
+      "learning_rate": 4.534862133237661e-06,
+      "loss": 0.623,
+      "step": 3452
+    },
+    {
+      "epoch": 1.2124297752808988,
+      "grad_norm": 0.8764957189559937,
+      "learning_rate": 4.534593514211701e-06,
+      "loss": 0.661,
+      "step": 3453
+    },
+    {
+      "epoch": 1.2127808988764044,
+      "grad_norm": 0.8892989158630371,
+      "learning_rate": 4.534324825604383e-06,
+      "loss": 0.6597,
+      "step": 3454
+    },
+    {
+      "epoch": 1.21313202247191,
+      "grad_norm": 0.902550220489502,
+      "learning_rate": 4.534056067424898e-06,
+      "loss": 0.6569,
+      "step": 3455
+    },
+    {
+      "epoch": 1.2134831460674158,
+      "grad_norm": 0.872238039970398,
+      "learning_rate": 4.5337872396824355e-06,
+      "loss": 0.6272,
+      "step": 3456
+    },
+    {
+      "epoch": 1.2138342696629214,
+      "grad_norm": 0.8700047135353088,
+      "learning_rate": 4.533518342386189e-06,
+      "loss": 0.6396,
+      "step": 3457
+    },
+    {
+      "epoch": 1.214185393258427,
+      "grad_norm": 0.8908973336219788,
+      "learning_rate": 4.5332493755453555e-06,
+      "loss": 0.6108,
+      "step": 3458
+    },
+    {
+      "epoch": 1.2145365168539326,
+      "grad_norm": 0.9054476618766785,
+      "learning_rate": 4.532980339169134e-06,
+      "loss": 0.6372,
+      "step": 3459
+    },
+    {
+      "epoch": 1.2148876404494382,
+      "grad_norm": 0.8539584279060364,
+      "learning_rate": 4.532711233266724e-06,
+      "loss": 0.6418,
+      "step": 3460
+    },
+    {
+      "epoch": 1.2152387640449438,
+      "grad_norm": 0.8940576314926147,
+      "learning_rate": 4.532442057847329e-06,
+      "loss": 0.6454,
+      "step": 3461
+    },
+    {
+      "epoch": 1.2155898876404494,
+      "grad_norm": 0.9023147225379944,
+      "learning_rate": 4.532172812920156e-06,
+      "loss": 0.6301,
+      "step": 3462
+    },
+    {
+      "epoch": 1.215941011235955,
+      "grad_norm": 0.9048089385032654,
+      "learning_rate": 4.53190349849441e-06,
+      "loss": 0.6469,
+      "step": 3463
+    },
+    {
+      "epoch": 1.2162921348314606,
+      "grad_norm": 0.8976174592971802,
+      "learning_rate": 4.531634114579304e-06,
+      "loss": 0.6518,
+      "step": 3464
+    },
+    {
+      "epoch": 1.2166432584269664,
+      "grad_norm": 0.8753100633621216,
+      "learning_rate": 4.53136466118405e-06,
+      "loss": 0.6173,
+      "step": 3465
+    },
+    {
+      "epoch": 1.216994382022472,
+      "grad_norm": 0.9097129106521606,
+      "learning_rate": 4.531095138317862e-06,
+      "loss": 0.6753,
+      "step": 3466
+    },
+    {
+      "epoch": 1.2173455056179776,
+      "grad_norm": 0.8648528456687927,
+      "learning_rate": 4.53082554598996e-06,
+      "loss": 0.5863,
+      "step": 3467
+    },
+    {
+      "epoch": 1.2176966292134832,
+      "grad_norm": 0.9543090462684631,
+      "learning_rate": 4.530555884209561e-06,
+      "loss": 0.6562,
+      "step": 3468
+    },
+    {
+      "epoch": 1.2180477528089888,
+      "grad_norm": 0.8979066610336304,
+      "learning_rate": 4.530286152985888e-06,
+      "loss": 0.6224,
+      "step": 3469
+    },
+    {
+      "epoch": 1.2183988764044944,
+      "grad_norm": 0.9094343185424805,
+      "learning_rate": 4.530016352328167e-06,
+      "loss": 0.6453,
+      "step": 3470
+    },
+    {
+      "epoch": 1.21875,
+      "grad_norm": 0.861716628074646,
+      "learning_rate": 4.5297464822456225e-06,
+      "loss": 0.5886,
+      "step": 3471
+    },
+    {
+      "epoch": 1.2191011235955056,
+      "grad_norm": 0.928870439529419,
+      "learning_rate": 4.5294765427474865e-06,
+      "loss": 0.6492,
+      "step": 3472
+    },
+    {
+      "epoch": 1.2194522471910112,
+      "grad_norm": 0.8802708387374878,
+      "learning_rate": 4.529206533842988e-06,
+      "loss": 0.6405,
+      "step": 3473
+    },
+    {
+      "epoch": 1.2198033707865168,
+      "grad_norm": 1.2119126319885254,
+      "learning_rate": 4.528936455541364e-06,
+      "loss": 0.6119,
+      "step": 3474
+    },
+    {
+      "epoch": 1.2201544943820224,
+      "grad_norm": 0.8638689517974854,
+      "learning_rate": 4.528666307851848e-06,
+      "loss": 0.6467,
+      "step": 3475
+    },
+    {
+      "epoch": 1.220505617977528,
+      "grad_norm": 0.8760889172554016,
+      "learning_rate": 4.528396090783681e-06,
+      "loss": 0.6113,
+      "step": 3476
+    },
+    {
+      "epoch": 1.2208567415730336,
+      "grad_norm": 0.8935677409172058,
+      "learning_rate": 4.5281258043461035e-06,
+      "loss": 0.6029,
+      "step": 3477
+    },
+    {
+      "epoch": 1.2212078651685394,
+      "grad_norm": 0.881455659866333,
+      "learning_rate": 4.527855448548359e-06,
+      "loss": 0.6389,
+      "step": 3478
+    },
+    {
+      "epoch": 1.221558988764045,
+      "grad_norm": 0.8549000024795532,
+      "learning_rate": 4.527585023399693e-06,
+      "loss": 0.6052,
+      "step": 3479
+    },
+    {
+      "epoch": 1.2219101123595506,
+      "grad_norm": 0.8719263672828674,
+      "learning_rate": 4.527314528909354e-06,
+      "loss": 0.635,
+      "step": 3480
+    },
+    {
+      "epoch": 1.2222612359550562,
+      "grad_norm": 0.8731090426445007,
+      "learning_rate": 4.527043965086594e-06,
+      "loss": 0.643,
+      "step": 3481
+    },
+    {
+      "epoch": 1.2226123595505618,
+      "grad_norm": 0.8634929060935974,
+      "learning_rate": 4.5267733319406635e-06,
+      "loss": 0.6184,
+      "step": 3482
+    },
+    {
+      "epoch": 1.2229634831460674,
+      "grad_norm": 0.8768059611320496,
+      "learning_rate": 4.52650262948082e-06,
+      "loss": 0.636,
+      "step": 3483
+    },
+    {
+      "epoch": 1.223314606741573,
+      "grad_norm": 0.8799298405647278,
+      "learning_rate": 4.526231857716321e-06,
+      "loss": 0.6273,
+      "step": 3484
+    },
+    {
+      "epoch": 1.2236657303370786,
+      "grad_norm": 0.8915926814079285,
+      "learning_rate": 4.525961016656426e-06,
+      "loss": 0.608,
+      "step": 3485
+    },
+    {
+      "epoch": 1.2240168539325842,
+      "grad_norm": 0.8969928026199341,
+      "learning_rate": 4.525690106310399e-06,
+      "loss": 0.6404,
+      "step": 3486
+    },
+    {
+      "epoch": 1.22436797752809,
+      "grad_norm": 0.9175208210945129,
+      "learning_rate": 4.525419126687503e-06,
+      "loss": 0.6429,
+      "step": 3487
+    },
+    {
+      "epoch": 1.2247191011235956,
+      "grad_norm": 0.8877578377723694,
+      "learning_rate": 4.5251480777970045e-06,
+      "loss": 0.6268,
+      "step": 3488
+    },
+    {
+      "epoch": 1.2250702247191012,
+      "grad_norm": 0.8742526173591614,
+      "learning_rate": 4.524876959648176e-06,
+      "loss": 0.6216,
+      "step": 3489
+    },
+    {
+      "epoch": 1.2254213483146068,
+      "grad_norm": 0.8701285123825073,
+      "learning_rate": 4.524605772250289e-06,
+      "loss": 0.6139,
+      "step": 3490
+    },
+    {
+      "epoch": 1.2257724719101124,
+      "grad_norm": 0.8775697946548462,
+      "learning_rate": 4.5243345156126166e-06,
+      "loss": 0.6187,
+      "step": 3491
+    },
+    {
+      "epoch": 1.226123595505618,
+      "grad_norm": 0.8713624477386475,
+      "learning_rate": 4.524063189744435e-06,
+      "loss": 0.6025,
+      "step": 3492
+    },
+    {
+      "epoch": 1.2264747191011236,
+      "grad_norm": 0.8762874007225037,
+      "learning_rate": 4.5237917946550244e-06,
+      "loss": 0.6542,
+      "step": 3493
+    },
+    {
+      "epoch": 1.2268258426966292,
+      "grad_norm": 0.9075011610984802,
+      "learning_rate": 4.523520330353668e-06,
+      "loss": 0.6346,
+      "step": 3494
+    },
+    {
+      "epoch": 1.2271769662921348,
+      "grad_norm": 0.9128796458244324,
+      "learning_rate": 4.523248796849647e-06,
+      "loss": 0.6474,
+      "step": 3495
+    },
+    {
+      "epoch": 1.2275280898876404,
+      "grad_norm": 0.888933002948761,
+      "learning_rate": 4.522977194152247e-06,
+      "loss": 0.5965,
+      "step": 3496
+    },
+    {
+      "epoch": 1.227879213483146,
+      "grad_norm": 0.8765919804573059,
+      "learning_rate": 4.522705522270759e-06,
+      "loss": 0.644,
+      "step": 3497
+    },
+    {
+      "epoch": 1.2282303370786516,
+      "grad_norm": 0.9075853228569031,
+      "learning_rate": 4.5224337812144735e-06,
+      "loss": 0.628,
+      "step": 3498
+    },
+    {
+      "epoch": 1.2285814606741572,
+      "grad_norm": 0.9263275265693665,
+      "learning_rate": 4.522161970992683e-06,
+      "loss": 0.6301,
+      "step": 3499
+    },
+    {
+      "epoch": 1.228932584269663,
+      "grad_norm": 0.8859231472015381,
+      "learning_rate": 4.521890091614684e-06,
+      "loss": 0.6347,
+      "step": 3500
+    },
+    {
+      "epoch": 1.2292837078651686,
+      "grad_norm": 0.879749596118927,
+      "learning_rate": 4.521618143089773e-06,
+      "loss": 0.649,
+      "step": 3501
+    },
+    {
+      "epoch": 1.2296348314606742,
+      "grad_norm": 0.8523240685462952,
+      "learning_rate": 4.521346125427252e-06,
+      "loss": 0.6284,
+      "step": 3502
+    },
+    {
+      "epoch": 1.2299859550561798,
+      "grad_norm": 0.8656728863716125,
+      "learning_rate": 4.521074038636422e-06,
+      "loss": 0.6338,
+      "step": 3503
+    },
+    {
+      "epoch": 1.2303370786516854,
+      "grad_norm": 0.8819969892501831,
+      "learning_rate": 4.5208018827265915e-06,
+      "loss": 0.6169,
+      "step": 3504
+    },
+    {
+      "epoch": 1.230688202247191,
+      "grad_norm": 0.868812084197998,
+      "learning_rate": 4.520529657707065e-06,
+      "loss": 0.6601,
+      "step": 3505
+    },
+    {
+      "epoch": 1.2310393258426966,
+      "grad_norm": 0.8616583347320557,
+      "learning_rate": 4.520257363587152e-06,
+      "loss": 0.5738,
+      "step": 3506
+    },
+    {
+      "epoch": 1.2313904494382022,
+      "grad_norm": 0.8898042440414429,
+      "learning_rate": 4.519985000376168e-06,
+      "loss": 0.671,
+      "step": 3507
+    },
+    {
+      "epoch": 1.2317415730337078,
+      "grad_norm": 0.8364049196243286,
+      "learning_rate": 4.519712568083424e-06,
+      "loss": 0.63,
+      "step": 3508
+    },
+    {
+      "epoch": 1.2320926966292134,
+      "grad_norm": 0.8617914915084839,
+      "learning_rate": 4.519440066718239e-06,
+      "loss": 0.5971,
+      "step": 3509
+    },
+    {
+      "epoch": 1.2324438202247192,
+      "grad_norm": 0.8895637392997742,
+      "learning_rate": 4.5191674962899314e-06,
+      "loss": 0.5993,
+      "step": 3510
+    },
+    {
+      "epoch": 1.2327949438202248,
+      "grad_norm": 0.910457968711853,
+      "learning_rate": 4.518894856807824e-06,
+      "loss": 0.6661,
+      "step": 3511
+    },
+    {
+      "epoch": 1.2331460674157304,
+      "grad_norm": 0.8428799510002136,
+      "learning_rate": 4.518622148281239e-06,
+      "loss": 0.5736,
+      "step": 3512
+    },
+    {
+      "epoch": 1.233497191011236,
+      "grad_norm": 0.8717475533485413,
+      "learning_rate": 4.518349370719505e-06,
+      "loss": 0.6315,
+      "step": 3513
+    },
+    {
+      "epoch": 1.2338483146067416,
+      "grad_norm": 0.9214996695518494,
+      "learning_rate": 4.518076524131949e-06,
+      "loss": 0.617,
+      "step": 3514
+    },
+    {
+      "epoch": 1.2341994382022472,
+      "grad_norm": 0.88681960105896,
+      "learning_rate": 4.517803608527903e-06,
+      "loss": 0.6427,
+      "step": 3515
+    },
+    {
+      "epoch": 1.2345505617977528,
+      "grad_norm": 0.8749536275863647,
+      "learning_rate": 4.517530623916701e-06,
+      "loss": 0.6326,
+      "step": 3516
+    },
+    {
+      "epoch": 1.2349016853932584,
+      "grad_norm": 0.8672427535057068,
+      "learning_rate": 4.517257570307676e-06,
+      "loss": 0.6264,
+      "step": 3517
+    },
+    {
+      "epoch": 1.235252808988764,
+      "grad_norm": 0.8522306680679321,
+      "learning_rate": 4.5169844477101695e-06,
+      "loss": 0.591,
+      "step": 3518
+    },
+    {
+      "epoch": 1.2356039325842696,
+      "grad_norm": 0.9023152589797974,
+      "learning_rate": 4.516711256133522e-06,
+      "loss": 0.641,
+      "step": 3519
+    },
+    {
+      "epoch": 1.2359550561797752,
+      "grad_norm": 0.8903506398200989,
+      "learning_rate": 4.516437995587074e-06,
+      "loss": 0.6356,
+      "step": 3520
+    },
+    {
+      "epoch": 1.2363061797752808,
+      "grad_norm": 0.8734146356582642,
+      "learning_rate": 4.516164666080172e-06,
+      "loss": 0.6365,
+      "step": 3521
+    },
+    {
+      "epoch": 1.2366573033707866,
+      "grad_norm": 0.8641245365142822,
+      "learning_rate": 4.5158912676221635e-06,
+      "loss": 0.6536,
+      "step": 3522
+    },
+    {
+      "epoch": 1.2370084269662922,
+      "grad_norm": 0.8999871015548706,
+      "learning_rate": 4.515617800222399e-06,
+      "loss": 0.6096,
+      "step": 3523
+    },
+    {
+      "epoch": 1.2373595505617978,
+      "grad_norm": 0.9102576971054077,
+      "learning_rate": 4.515344263890231e-06,
+      "loss": 0.6285,
+      "step": 3524
+    },
+    {
+      "epoch": 1.2377106741573034,
+      "grad_norm": 0.8963742256164551,
+      "learning_rate": 4.515070658635013e-06,
+      "loss": 0.6625,
+      "step": 3525
+    },
+    {
+      "epoch": 1.238061797752809,
+      "grad_norm": 0.9126663208007812,
+      "learning_rate": 4.514796984466103e-06,
+      "loss": 0.6398,
+      "step": 3526
+    },
+    {
+      "epoch": 1.2384129213483146,
+      "grad_norm": 0.8794665336608887,
+      "learning_rate": 4.514523241392861e-06,
+      "loss": 0.6545,
+      "step": 3527
+    },
+    {
+      "epoch": 1.2387640449438202,
+      "grad_norm": 0.9098983407020569,
+      "learning_rate": 4.514249429424647e-06,
+      "loss": 0.6394,
+      "step": 3528
+    },
+    {
+      "epoch": 1.2391151685393258,
+      "grad_norm": 0.8858410120010376,
+      "learning_rate": 4.513975548570826e-06,
+      "loss": 0.6219,
+      "step": 3529
+    },
+    {
+      "epoch": 1.2394662921348314,
+      "grad_norm": 0.9184651970863342,
+      "learning_rate": 4.5137015988407655e-06,
+      "loss": 0.647,
+      "step": 3530
+    },
+    {
+      "epoch": 1.239817415730337,
+      "grad_norm": 0.9131940603256226,
+      "learning_rate": 4.513427580243833e-06,
+      "loss": 0.615,
+      "step": 3531
+    },
+    {
+      "epoch": 1.2401685393258428,
+      "grad_norm": 0.8828357458114624,
+      "learning_rate": 4.5131534927894e-06,
+      "loss": 0.6288,
+      "step": 3532
+    },
+    {
+      "epoch": 1.2405196629213484,
+      "grad_norm": 0.8712863922119141,
+      "learning_rate": 4.512879336486841e-06,
+      "loss": 0.6083,
+      "step": 3533
+    },
+    {
+      "epoch": 1.240870786516854,
+      "grad_norm": 0.8657607436180115,
+      "learning_rate": 4.5126051113455295e-06,
+      "loss": 0.6189,
+      "step": 3534
+    },
+    {
+      "epoch": 1.2412219101123596,
+      "grad_norm": 0.896507203578949,
+      "learning_rate": 4.512330817374847e-06,
+      "loss": 0.6512,
+      "step": 3535
+    },
+    {
+      "epoch": 1.2415730337078652,
+      "grad_norm": 0.8922907710075378,
+      "learning_rate": 4.512056454584172e-06,
+      "loss": 0.6269,
+      "step": 3536
+    },
+    {
+      "epoch": 1.2419241573033708,
+      "grad_norm": 0.9080265760421753,
+      "learning_rate": 4.511782022982888e-06,
+      "loss": 0.6319,
+      "step": 3537
+    },
+    {
+      "epoch": 1.2422752808988764,
+      "grad_norm": 0.8720276355743408,
+      "learning_rate": 4.511507522580381e-06,
+      "loss": 0.6038,
+      "step": 3538
+    },
+    {
+      "epoch": 1.242626404494382,
+      "grad_norm": 0.8793643712997437,
+      "learning_rate": 4.511232953386037e-06,
+      "loss": 0.6672,
+      "step": 3539
+    },
+    {
+      "epoch": 1.2429775280898876,
+      "grad_norm": 0.8374327421188354,
+      "learning_rate": 4.510958315409247e-06,
+      "loss": 0.5826,
+      "step": 3540
+    },
+    {
+      "epoch": 1.2433286516853932,
+      "grad_norm": 0.8608018159866333,
+      "learning_rate": 4.510683608659404e-06,
+      "loss": 0.61,
+      "step": 3541
+    },
+    {
+      "epoch": 1.2436797752808988,
+      "grad_norm": 0.9034033417701721,
+      "learning_rate": 4.510408833145902e-06,
+      "loss": 0.6382,
+      "step": 3542
+    },
+    {
+      "epoch": 1.2440308988764044,
+      "grad_norm": 0.908312201499939,
+      "learning_rate": 4.510133988878138e-06,
+      "loss": 0.6437,
+      "step": 3543
+    },
+    {
+      "epoch": 1.24438202247191,
+      "grad_norm": 0.9202480316162109,
+      "learning_rate": 4.5098590758655115e-06,
+      "loss": 0.6462,
+      "step": 3544
+    },
+    {
+      "epoch": 1.2447331460674158,
+      "grad_norm": 0.8709115386009216,
+      "learning_rate": 4.509584094117424e-06,
+      "loss": 0.6146,
+      "step": 3545
+    },
+    {
+      "epoch": 1.2450842696629214,
+      "grad_norm": 0.8764071464538574,
+      "learning_rate": 4.509309043643282e-06,
+      "loss": 0.6002,
+      "step": 3546
+    },
+    {
+      "epoch": 1.245435393258427,
+      "grad_norm": 0.8512903451919556,
+      "learning_rate": 4.5090339244524886e-06,
+      "loss": 0.5995,
+      "step": 3547
+    },
+    {
+      "epoch": 1.2457865168539326,
+      "grad_norm": 0.846396803855896,
+      "learning_rate": 4.508758736554454e-06,
+      "loss": 0.5954,
+      "step": 3548
+    },
+    {
+      "epoch": 1.2461376404494382,
+      "grad_norm": 0.8854002952575684,
+      "learning_rate": 4.508483479958591e-06,
+      "loss": 0.619,
+      "step": 3549
+    },
+    {
+      "epoch": 1.2464887640449438,
+      "grad_norm": 0.8938324451446533,
+      "learning_rate": 4.50820815467431e-06,
+      "loss": 0.6382,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2468398876404494,
+      "grad_norm": 0.8843767046928406,
+      "learning_rate": 4.50793276071103e-06,
+      "loss": 0.6169,
+      "step": 3551
+    },
+    {
+      "epoch": 1.247191011235955,
+      "grad_norm": 0.8539121747016907,
+      "learning_rate": 4.507657298078167e-06,
+      "loss": 0.6312,
+      "step": 3552
+    },
+    {
+      "epoch": 1.2475421348314606,
+      "grad_norm": 0.8683538436889648,
+      "learning_rate": 4.5073817667851415e-06,
+      "loss": 0.6121,
+      "step": 3553
+    },
+    {
+      "epoch": 1.2478932584269664,
+      "grad_norm": 0.8984917998313904,
+      "learning_rate": 4.507106166841378e-06,
+      "loss": 0.6588,
+      "step": 3554
+    },
+    {
+      "epoch": 1.248244382022472,
+      "grad_norm": 0.9375321865081787,
+      "learning_rate": 4.506830498256302e-06,
+      "loss": 0.6943,
+      "step": 3555
+    },
+    {
+      "epoch": 1.2485955056179776,
+      "grad_norm": 0.8783092498779297,
+      "learning_rate": 4.5065547610393395e-06,
+      "loss": 0.6103,
+      "step": 3556
+    },
+    {
+      "epoch": 1.2489466292134832,
+      "grad_norm": 0.8783857822418213,
+      "learning_rate": 4.506278955199922e-06,
+      "loss": 0.6034,
+      "step": 3557
+    },
+    {
+      "epoch": 1.2492977528089888,
+      "grad_norm": 0.8634523749351501,
+      "learning_rate": 4.506003080747479e-06,
+      "loss": 0.5898,
+      "step": 3558
+    },
+    {
+      "epoch": 1.2496488764044944,
+      "grad_norm": 0.8823550343513489,
+      "learning_rate": 4.5057271376914495e-06,
+      "loss": 0.5877,
+      "step": 3559
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8883979320526123,
+      "learning_rate": 4.505451126041267e-06,
+      "loss": 0.6384,
+      "step": 3560
+    },
+    {
+      "epoch": 1.2503511235955056,
+      "grad_norm": 0.8742533922195435,
+      "learning_rate": 4.5051750458063725e-06,
+      "loss": 0.6099,
+      "step": 3561
+    },
+    {
+      "epoch": 1.2507022471910112,
+      "grad_norm": 0.9132082462310791,
+      "learning_rate": 4.504898896996207e-06,
+      "loss": 0.6418,
+      "step": 3562
+    },
+    {
+      "epoch": 1.2510533707865168,
+      "grad_norm": 0.8753030300140381,
+      "learning_rate": 4.504622679620215e-06,
+      "loss": 0.6074,
+      "step": 3563
+    },
+    {
+      "epoch": 1.2514044943820224,
+      "grad_norm": 0.8916899561882019,
+      "learning_rate": 4.504346393687843e-06,
+      "loss": 0.6614,
+      "step": 3564
+    },
+    {
+      "epoch": 1.251755617977528,
+      "grad_norm": 0.8677763342857361,
+      "learning_rate": 4.504070039208539e-06,
+      "loss": 0.6256,
+      "step": 3565
+    },
+    {
+      "epoch": 1.2521067415730336,
+      "grad_norm": 0.8957414627075195,
+      "learning_rate": 4.503793616191755e-06,
+      "loss": 0.6742,
+      "step": 3566
+    },
+    {
+      "epoch": 1.2524578651685394,
+      "grad_norm": 0.8888939619064331,
+      "learning_rate": 4.5035171246469435e-06,
+      "loss": 0.6346,
+      "step": 3567
+    },
+    {
+      "epoch": 1.252808988764045,
+      "grad_norm": 0.8643205165863037,
+      "learning_rate": 4.5032405645835605e-06,
+      "loss": 0.6214,
+      "step": 3568
+    },
+    {
+      "epoch": 1.2531601123595506,
+      "grad_norm": 0.8865513801574707,
+      "learning_rate": 4.502963936011065e-06,
+      "loss": 0.6469,
+      "step": 3569
+    },
+    {
+      "epoch": 1.2535112359550562,
+      "grad_norm": 0.8715183734893799,
+      "learning_rate": 4.502687238938917e-06,
+      "loss": 0.6104,
+      "step": 3570
+    },
+    {
+      "epoch": 1.2538623595505618,
+      "grad_norm": 0.8890028595924377,
+      "learning_rate": 4.5024104733765795e-06,
+      "loss": 0.6414,
+      "step": 3571
+    },
+    {
+      "epoch": 1.2542134831460674,
+      "grad_norm": 0.8931072354316711,
+      "learning_rate": 4.502133639333516e-06,
+      "loss": 0.6321,
+      "step": 3572
+    },
+    {
+      "epoch": 1.254564606741573,
+      "grad_norm": 0.8449422121047974,
+      "learning_rate": 4.501856736819196e-06,
+      "loss": 0.6246,
+      "step": 3573
+    },
+    {
+      "epoch": 1.2549157303370786,
+      "grad_norm": 0.8641847968101501,
+      "learning_rate": 4.501579765843089e-06,
+      "loss": 0.5858,
+      "step": 3574
+    },
+    {
+      "epoch": 1.2552668539325842,
+      "grad_norm": 0.890620768070221,
+      "learning_rate": 4.501302726414666e-06,
+      "loss": 0.6045,
+      "step": 3575
+    },
+    {
+      "epoch": 1.25561797752809,
+      "grad_norm": 0.8621348738670349,
+      "learning_rate": 4.501025618543403e-06,
+      "loss": 0.622,
+      "step": 3576
+    },
+    {
+      "epoch": 1.2559691011235956,
+      "grad_norm": 0.8901969194412231,
+      "learning_rate": 4.500748442238776e-06,
+      "loss": 0.6475,
+      "step": 3577
+    },
+    {
+      "epoch": 1.2563202247191012,
+      "grad_norm": 0.8945297598838806,
+      "learning_rate": 4.500471197510264e-06,
+      "loss": 0.6374,
+      "step": 3578
+    },
+    {
+      "epoch": 1.2566713483146068,
+      "grad_norm": 0.8774127960205078,
+      "learning_rate": 4.500193884367349e-06,
+      "loss": 0.6243,
+      "step": 3579
+    },
+    {
+      "epoch": 1.2570224719101124,
+      "grad_norm": 0.8826004862785339,
+      "learning_rate": 4.499916502819514e-06,
+      "loss": 0.6227,
+      "step": 3580
+    },
+    {
+      "epoch": 1.257373595505618,
+      "grad_norm": 0.9239538311958313,
+      "learning_rate": 4.499639052876247e-06,
+      "loss": 0.6432,
+      "step": 3581
+    },
+    {
+      "epoch": 1.2577247191011236,
+      "grad_norm": 0.8692966103553772,
+      "learning_rate": 4.499361534547035e-06,
+      "loss": 0.6343,
+      "step": 3582
+    },
+    {
+      "epoch": 1.2580758426966292,
+      "grad_norm": 0.837487518787384,
+      "learning_rate": 4.499083947841371e-06,
+      "loss": 0.599,
+      "step": 3583
+    },
+    {
+      "epoch": 1.2584269662921348,
+      "grad_norm": 0.9392856955528259,
+      "learning_rate": 4.498806292768745e-06,
+      "loss": 0.5877,
+      "step": 3584
+    },
+    {
+      "epoch": 1.2587780898876404,
+      "grad_norm": 0.9112626910209656,
+      "learning_rate": 4.498528569338654e-06,
+      "loss": 0.6188,
+      "step": 3585
+    },
+    {
+      "epoch": 1.259129213483146,
+      "grad_norm": 0.863733172416687,
+      "learning_rate": 4.498250777560596e-06,
+      "loss": 0.6268,
+      "step": 3586
+    },
+    {
+      "epoch": 1.2594803370786516,
+      "grad_norm": 0.8512037992477417,
+      "learning_rate": 4.497972917444072e-06,
+      "loss": 0.6109,
+      "step": 3587
+    },
+    {
+      "epoch": 1.2598314606741572,
+      "grad_norm": 0.8677399158477783,
+      "learning_rate": 4.497694988998584e-06,
+      "loss": 0.6133,
+      "step": 3588
+    },
+    {
+      "epoch": 1.2601825842696628,
+      "grad_norm": 0.8969357013702393,
+      "learning_rate": 4.497416992233637e-06,
+      "loss": 0.6317,
+      "step": 3589
+    },
+    {
+      "epoch": 1.2605337078651686,
+      "grad_norm": 0.8791055679321289,
+      "learning_rate": 4.497138927158737e-06,
+      "loss": 0.6711,
+      "step": 3590
+    },
+    {
+      "epoch": 1.2608848314606742,
+      "grad_norm": 0.8862146735191345,
+      "learning_rate": 4.496860793783397e-06,
+      "loss": 0.6027,
+      "step": 3591
+    },
+    {
+      "epoch": 1.2612359550561798,
+      "grad_norm": 0.8788702487945557,
+      "learning_rate": 4.496582592117123e-06,
+      "loss": 0.6441,
+      "step": 3592
+    },
+    {
+      "epoch": 1.2615870786516854,
+      "grad_norm": 0.879822850227356,
+      "learning_rate": 4.496304322169436e-06,
+      "loss": 0.605,
+      "step": 3593
+    },
+    {
+      "epoch": 1.261938202247191,
+      "grad_norm": 0.870986819267273,
+      "learning_rate": 4.496025983949848e-06,
+      "loss": 0.6143,
+      "step": 3594
+    },
+    {
+      "epoch": 1.2622893258426966,
+      "grad_norm": 0.8678484559059143,
+      "learning_rate": 4.495747577467879e-06,
+      "loss": 0.6081,
+      "step": 3595
+    },
+    {
+      "epoch": 1.2626404494382022,
+      "grad_norm": 0.8320144414901733,
+      "learning_rate": 4.495469102733052e-06,
+      "loss": 0.6052,
+      "step": 3596
+    },
+    {
+      "epoch": 1.2629915730337078,
+      "grad_norm": 0.8532273769378662,
+      "learning_rate": 4.495190559754888e-06,
+      "loss": 0.6202,
+      "step": 3597
+    },
+    {
+      "epoch": 1.2633426966292136,
+      "grad_norm": 0.873294472694397,
+      "learning_rate": 4.494911948542915e-06,
+      "loss": 0.6292,
+      "step": 3598
+    },
+    {
+      "epoch": 1.2636938202247192,
+      "grad_norm": 0.8932523131370544,
+      "learning_rate": 4.4946332691066605e-06,
+      "loss": 0.6189,
+      "step": 3599
+    },
+    {
+      "epoch": 1.2640449438202248,
+      "grad_norm": 0.9090549945831299,
+      "learning_rate": 4.494354521455655e-06,
+      "loss": 0.66,
+      "step": 3600
+    },
+    {
+      "epoch": 1.2643960674157304,
+      "grad_norm": 0.8804965615272522,
+      "learning_rate": 4.49407570559943e-06,
+      "loss": 0.6105,
+      "step": 3601
+    },
+    {
+      "epoch": 1.264747191011236,
+      "grad_norm": 0.8950492143630981,
+      "learning_rate": 4.493796821547523e-06,
+      "loss": 0.6281,
+      "step": 3602
+    },
+    {
+      "epoch": 1.2650983146067416,
+      "grad_norm": 0.9223471283912659,
+      "learning_rate": 4.493517869309472e-06,
+      "loss": 0.6455,
+      "step": 3603
+    },
+    {
+      "epoch": 1.2654494382022472,
+      "grad_norm": 0.9145454168319702,
+      "learning_rate": 4.493238848894815e-06,
+      "loss": 0.637,
+      "step": 3604
+    },
+    {
+      "epoch": 1.2658005617977528,
+      "grad_norm": 0.9027042388916016,
+      "learning_rate": 4.4929597603130944e-06,
+      "loss": 0.6292,
+      "step": 3605
+    },
+    {
+      "epoch": 1.2661516853932584,
+      "grad_norm": 0.9141017198562622,
+      "learning_rate": 4.492680603573857e-06,
+      "loss": 0.621,
+      "step": 3606
+    },
+    {
+      "epoch": 1.266502808988764,
+      "grad_norm": 0.903798520565033,
+      "learning_rate": 4.492401378686646e-06,
+      "loss": 0.6279,
+      "step": 3607
+    },
+    {
+      "epoch": 1.2668539325842696,
+      "grad_norm": 0.9175305962562561,
+      "learning_rate": 4.492122085661014e-06,
+      "loss": 0.6319,
+      "step": 3608
+    },
+    {
+      "epoch": 1.2672050561797752,
+      "grad_norm": 0.9265148043632507,
+      "learning_rate": 4.491842724506511e-06,
+      "loss": 0.6561,
+      "step": 3609
+    },
+    {
+      "epoch": 1.2675561797752808,
+      "grad_norm": 0.8984575867652893,
+      "learning_rate": 4.491563295232692e-06,
+      "loss": 0.6515,
+      "step": 3610
+    },
+    {
+      "epoch": 1.2679073033707864,
+      "grad_norm": 0.9043526649475098,
+      "learning_rate": 4.491283797849112e-06,
+      "loss": 0.6505,
+      "step": 3611
+    },
+    {
+      "epoch": 1.2682584269662922,
+      "grad_norm": 0.8922470211982727,
+      "learning_rate": 4.4910042323653295e-06,
+      "loss": 0.6638,
+      "step": 3612
+    },
+    {
+      "epoch": 1.2686095505617978,
+      "grad_norm": 0.8709413409233093,
+      "learning_rate": 4.4907245987909066e-06,
+      "loss": 0.6109,
+      "step": 3613
+    },
+    {
+      "epoch": 1.2689606741573034,
+      "grad_norm": 0.931884229183197,
+      "learning_rate": 4.490444897135405e-06,
+      "loss": 0.6791,
+      "step": 3614
+    },
+    {
+      "epoch": 1.269311797752809,
+      "grad_norm": 0.892665445804596,
+      "learning_rate": 4.490165127408393e-06,
+      "loss": 0.6312,
+      "step": 3615
+    },
+    {
+      "epoch": 1.2696629213483146,
+      "grad_norm": 0.8642548322677612,
+      "learning_rate": 4.489885289619435e-06,
+      "loss": 0.6095,
+      "step": 3616
+    },
+    {
+      "epoch": 1.2700140449438202,
+      "grad_norm": 0.8878061771392822,
+      "learning_rate": 4.489605383778103e-06,
+      "loss": 0.6334,
+      "step": 3617
+    },
+    {
+      "epoch": 1.2703651685393258,
+      "grad_norm": 0.8912963271141052,
+      "learning_rate": 4.48932540989397e-06,
+      "loss": 0.6384,
+      "step": 3618
+    },
+    {
+      "epoch": 1.2707162921348314,
+      "grad_norm": 0.8930538892745972,
+      "learning_rate": 4.489045367976611e-06,
+      "loss": 0.6209,
+      "step": 3619
+    },
+    {
+      "epoch": 1.2710674157303372,
+      "grad_norm": 0.9626955986022949,
+      "learning_rate": 4.488765258035601e-06,
+      "loss": 0.6584,
+      "step": 3620
+    },
+    {
+      "epoch": 1.2714185393258428,
+      "grad_norm": 0.9230532050132751,
+      "learning_rate": 4.488485080080522e-06,
+      "loss": 0.6289,
+      "step": 3621
+    },
+    {
+      "epoch": 1.2717696629213484,
+      "grad_norm": 0.9188664555549622,
+      "learning_rate": 4.488204834120954e-06,
+      "loss": 0.6185,
+      "step": 3622
+    },
+    {
+      "epoch": 1.272120786516854,
+      "grad_norm": 0.9008336663246155,
+      "learning_rate": 4.4879245201664825e-06,
+      "loss": 0.6234,
+      "step": 3623
+    },
+    {
+      "epoch": 1.2724719101123596,
+      "grad_norm": 0.8780505657196045,
+      "learning_rate": 4.487644138226693e-06,
+      "loss": 0.6431,
+      "step": 3624
+    },
+    {
+      "epoch": 1.2728230337078652,
+      "grad_norm": 0.9006023406982422,
+      "learning_rate": 4.487363688311176e-06,
+      "loss": 0.6304,
+      "step": 3625
+    },
+    {
+      "epoch": 1.2731741573033708,
+      "grad_norm": 0.894845724105835,
+      "learning_rate": 4.487083170429521e-06,
+      "loss": 0.6455,
+      "step": 3626
+    },
+    {
+      "epoch": 1.2735252808988764,
+      "grad_norm": 0.9087368249893188,
+      "learning_rate": 4.486802584591321e-06,
+      "loss": 0.6616,
+      "step": 3627
+    },
+    {
+      "epoch": 1.273876404494382,
+      "grad_norm": 0.8746863007545471,
+      "learning_rate": 4.486521930806174e-06,
+      "loss": 0.6158,
+      "step": 3628
+    },
+    {
+      "epoch": 1.2742275280898876,
+      "grad_norm": 0.8722705245018005,
+      "learning_rate": 4.486241209083676e-06,
+      "loss": 0.59,
+      "step": 3629
+    },
+    {
+      "epoch": 1.2745786516853932,
+      "grad_norm": 0.8320393562316895,
+      "learning_rate": 4.485960419433428e-06,
+      "loss": 0.6003,
+      "step": 3630
+    },
+    {
+      "epoch": 1.2749297752808988,
+      "grad_norm": 0.9099650382995605,
+      "learning_rate": 4.485679561865034e-06,
+      "loss": 0.6279,
+      "step": 3631
+    },
+    {
+      "epoch": 1.2752808988764044,
+      "grad_norm": 0.905540406703949,
+      "learning_rate": 4.485398636388098e-06,
+      "loss": 0.6963,
+      "step": 3632
+    },
+    {
+      "epoch": 1.27563202247191,
+      "grad_norm": 0.9020867347717285,
+      "learning_rate": 4.485117643012227e-06,
+      "loss": 0.6448,
+      "step": 3633
+    },
+    {
+      "epoch": 1.2759831460674158,
+      "grad_norm": 0.8654907941818237,
+      "learning_rate": 4.484836581747032e-06,
+      "loss": 0.6488,
+      "step": 3634
+    },
+    {
+      "epoch": 1.2763342696629214,
+      "grad_norm": 0.878855288028717,
+      "learning_rate": 4.4845554526021235e-06,
+      "loss": 0.6241,
+      "step": 3635
+    },
+    {
+      "epoch": 1.276685393258427,
+      "grad_norm": 0.8361636400222778,
+      "learning_rate": 4.484274255587117e-06,
+      "loss": 0.6078,
+      "step": 3636
+    },
+    {
+      "epoch": 1.2770365168539326,
+      "grad_norm": 0.8940629959106445,
+      "learning_rate": 4.48399299071163e-06,
+      "loss": 0.6096,
+      "step": 3637
+    },
+    {
+      "epoch": 1.2773876404494382,
+      "grad_norm": 0.8670988082885742,
+      "learning_rate": 4.483711657985279e-06,
+      "loss": 0.5922,
+      "step": 3638
+    },
+    {
+      "epoch": 1.2777387640449438,
+      "grad_norm": 0.9459436535835266,
+      "learning_rate": 4.483430257417687e-06,
+      "loss": 0.648,
+      "step": 3639
+    },
+    {
+      "epoch": 1.2780898876404494,
+      "grad_norm": 0.889330267906189,
+      "learning_rate": 4.483148789018478e-06,
+      "loss": 0.6526,
+      "step": 3640
+    },
+    {
+      "epoch": 1.278441011235955,
+      "grad_norm": 0.8704626560211182,
+      "learning_rate": 4.4828672527972775e-06,
+      "loss": 0.6545,
+      "step": 3641
+    },
+    {
+      "epoch": 1.2787921348314606,
+      "grad_norm": 0.8923251628875732,
+      "learning_rate": 4.482585648763713e-06,
+      "loss": 0.6321,
+      "step": 3642
+    },
+    {
+      "epoch": 1.2791432584269664,
+      "grad_norm": 0.8882480263710022,
+      "learning_rate": 4.482303976927417e-06,
+      "loss": 0.6371,
+      "step": 3643
+    },
+    {
+      "epoch": 1.279494382022472,
+      "grad_norm": 0.8774885535240173,
+      "learning_rate": 4.482022237298021e-06,
+      "loss": 0.5873,
+      "step": 3644
+    },
+    {
+      "epoch": 1.2798455056179776,
+      "grad_norm": 0.8971747756004333,
+      "learning_rate": 4.481740429885159e-06,
+      "loss": 0.6418,
+      "step": 3645
+    },
+    {
+      "epoch": 1.2801966292134832,
+      "grad_norm": 0.860084593296051,
+      "learning_rate": 4.481458554698471e-06,
+      "loss": 0.6259,
+      "step": 3646
+    },
+    {
+      "epoch": 1.2805477528089888,
+      "grad_norm": 0.885615885257721,
+      "learning_rate": 4.481176611747597e-06,
+      "loss": 0.5883,
+      "step": 3647
+    },
+    {
+      "epoch": 1.2808988764044944,
+      "grad_norm": 0.9061590433120728,
+      "learning_rate": 4.480894601042177e-06,
+      "loss": 0.6295,
+      "step": 3648
+    },
+    {
+      "epoch": 1.28125,
+      "grad_norm": 0.8891565203666687,
+      "learning_rate": 4.480612522591857e-06,
+      "loss": 0.6505,
+      "step": 3649
+    },
+    {
+      "epoch": 1.2816011235955056,
+      "grad_norm": 0.8980679512023926,
+      "learning_rate": 4.480330376406284e-06,
+      "loss": 0.6035,
+      "step": 3650
+    },
+    {
+      "epoch": 1.2819522471910112,
+      "grad_norm": 0.8772475719451904,
+      "learning_rate": 4.480048162495106e-06,
+      "loss": 0.6141,
+      "step": 3651
+    },
+    {
+      "epoch": 1.2823033707865168,
+      "grad_norm": 0.883439302444458,
+      "learning_rate": 4.479765880867976e-06,
+      "loss": 0.6349,
+      "step": 3652
+    },
+    {
+      "epoch": 1.2826544943820224,
+      "grad_norm": 0.872630774974823,
+      "learning_rate": 4.479483531534546e-06,
+      "loss": 0.6381,
+      "step": 3653
+    },
+    {
+      "epoch": 1.283005617977528,
+      "grad_norm": 0.8689930438995361,
+      "learning_rate": 4.479201114504474e-06,
+      "loss": 0.6256,
+      "step": 3654
+    },
+    {
+      "epoch": 1.2833567415730336,
+      "grad_norm": 0.9189596176147461,
+      "learning_rate": 4.478918629787417e-06,
+      "loss": 0.6296,
+      "step": 3655
+    },
+    {
+      "epoch": 1.2837078651685394,
+      "grad_norm": 0.8905652165412903,
+      "learning_rate": 4.478636077393037e-06,
+      "loss": 0.6283,
+      "step": 3656
+    },
+    {
+      "epoch": 1.284058988764045,
+      "grad_norm": 0.889506995677948,
+      "learning_rate": 4.478353457330995e-06,
+      "loss": 0.6433,
+      "step": 3657
+    },
+    {
+      "epoch": 1.2844101123595506,
+      "grad_norm": 0.881960391998291,
+      "learning_rate": 4.478070769610959e-06,
+      "loss": 0.6477,
+      "step": 3658
+    },
+    {
+      "epoch": 1.2847612359550562,
+      "grad_norm": 0.9423824548721313,
+      "learning_rate": 4.477788014242594e-06,
+      "loss": 0.6636,
+      "step": 3659
+    },
+    {
+      "epoch": 1.2851123595505618,
+      "grad_norm": 0.9165781140327454,
+      "learning_rate": 4.477505191235572e-06,
+      "loss": 0.6498,
+      "step": 3660
+    },
+    {
+      "epoch": 1.2854634831460674,
+      "grad_norm": 0.8826702237129211,
+      "learning_rate": 4.477222300599566e-06,
+      "loss": 0.6255,
+      "step": 3661
+    },
+    {
+      "epoch": 1.285814606741573,
+      "grad_norm": 0.8841883540153503,
+      "learning_rate": 4.476939342344246e-06,
+      "loss": 0.6496,
+      "step": 3662
+    },
+    {
+      "epoch": 1.2861657303370786,
+      "grad_norm": 0.8767368197441101,
+      "learning_rate": 4.476656316479295e-06,
+      "loss": 0.6227,
+      "step": 3663
+    },
+    {
+      "epoch": 1.2865168539325842,
+      "grad_norm": 0.8614138960838318,
+      "learning_rate": 4.4763732230143885e-06,
+      "loss": 0.6168,
+      "step": 3664
+    },
+    {
+      "epoch": 1.28686797752809,
+      "grad_norm": 0.8860598802566528,
+      "learning_rate": 4.476090061959209e-06,
+      "loss": 0.6414,
+      "step": 3665
+    },
+    {
+      "epoch": 1.2872191011235956,
+      "grad_norm": 0.8853031396865845,
+      "learning_rate": 4.4758068333234405e-06,
+      "loss": 0.6207,
+      "step": 3666
+    },
+    {
+      "epoch": 1.2875702247191012,
+      "grad_norm": 0.889927089214325,
+      "learning_rate": 4.4755235371167685e-06,
+      "loss": 0.6297,
+      "step": 3667
+    },
+    {
+      "epoch": 1.2879213483146068,
+      "grad_norm": 0.8700869083404541,
+      "learning_rate": 4.475240173348883e-06,
+      "loss": 0.6291,
+      "step": 3668
+    },
+    {
+      "epoch": 1.2882724719101124,
+      "grad_norm": 0.8681374788284302,
+      "learning_rate": 4.474956742029473e-06,
+      "loss": 0.6304,
+      "step": 3669
+    },
+    {
+      "epoch": 1.288623595505618,
+      "grad_norm": 0.8633899688720703,
+      "learning_rate": 4.474673243168233e-06,
+      "loss": 0.5838,
+      "step": 3670
+    },
+    {
+      "epoch": 1.2889747191011236,
+      "grad_norm": 0.8746448755264282,
+      "learning_rate": 4.4743896767748566e-06,
+      "loss": 0.6498,
+      "step": 3671
+    },
+    {
+      "epoch": 1.2893258426966292,
+      "grad_norm": 0.8993739485740662,
+      "learning_rate": 4.474106042859044e-06,
+      "loss": 0.6524,
+      "step": 3672
+    },
+    {
+      "epoch": 1.2896769662921348,
+      "grad_norm": 0.9252314567565918,
+      "learning_rate": 4.473822341430493e-06,
+      "loss": 0.668,
+      "step": 3673
+    },
+    {
+      "epoch": 1.2900280898876404,
+      "grad_norm": 0.8796794414520264,
+      "learning_rate": 4.473538572498908e-06,
+      "loss": 0.6452,
+      "step": 3674
+    },
+    {
+      "epoch": 1.290379213483146,
+      "grad_norm": 0.8701682686805725,
+      "learning_rate": 4.473254736073992e-06,
+      "loss": 0.6003,
+      "step": 3675
+    },
+    {
+      "epoch": 1.2907303370786516,
+      "grad_norm": 0.9012110233306885,
+      "learning_rate": 4.472970832165453e-06,
+      "loss": 0.6412,
+      "step": 3676
+    },
+    {
+      "epoch": 1.2910814606741572,
+      "grad_norm": 0.8365998864173889,
+      "learning_rate": 4.4726868607829995e-06,
+      "loss": 0.559,
+      "step": 3677
+    },
+    {
+      "epoch": 1.2914325842696628,
+      "grad_norm": 0.8749606609344482,
+      "learning_rate": 4.472402821936344e-06,
+      "loss": 0.6241,
+      "step": 3678
+    },
+    {
+      "epoch": 1.2917837078651686,
+      "grad_norm": 0.8832113146781921,
+      "learning_rate": 4.472118715635199e-06,
+      "loss": 0.6543,
+      "step": 3679
+    },
+    {
+      "epoch": 1.2921348314606742,
+      "grad_norm": 0.8898352384567261,
+      "learning_rate": 4.471834541889282e-06,
+      "loss": 0.6716,
+      "step": 3680
+    },
+    {
+      "epoch": 1.2924859550561798,
+      "grad_norm": 0.8850893974304199,
+      "learning_rate": 4.47155030070831e-06,
+      "loss": 0.6269,
+      "step": 3681
+    },
+    {
+      "epoch": 1.2928370786516854,
+      "grad_norm": 0.8936760425567627,
+      "learning_rate": 4.471265992102005e-06,
+      "loss": 0.658,
+      "step": 3682
+    },
+    {
+      "epoch": 1.293188202247191,
+      "grad_norm": 0.8844362497329712,
+      "learning_rate": 4.4709816160800915e-06,
+      "loss": 0.6128,
+      "step": 3683
+    },
+    {
+      "epoch": 1.2935393258426966,
+      "grad_norm": 0.9699074625968933,
+      "learning_rate": 4.470697172652292e-06,
+      "loss": 0.653,
+      "step": 3684
+    },
+    {
+      "epoch": 1.2938904494382022,
+      "grad_norm": 0.9202075600624084,
+      "learning_rate": 4.470412661828335e-06,
+      "loss": 0.6514,
+      "step": 3685
+    },
+    {
+      "epoch": 1.2942415730337078,
+      "grad_norm": 0.8599914312362671,
+      "learning_rate": 4.470128083617953e-06,
+      "loss": 0.6347,
+      "step": 3686
+    },
+    {
+      "epoch": 1.2945926966292136,
+      "grad_norm": 0.8901776075363159,
+      "learning_rate": 4.469843438030875e-06,
+      "loss": 0.6073,
+      "step": 3687
+    },
+    {
+      "epoch": 1.2949438202247192,
+      "grad_norm": 0.8744489550590515,
+      "learning_rate": 4.4695587250768365e-06,
+      "loss": 0.5886,
+      "step": 3688
+    },
+    {
+      "epoch": 1.2952949438202248,
+      "grad_norm": 0.8581476211547852,
+      "learning_rate": 4.469273944765576e-06,
+      "loss": 0.6239,
+      "step": 3689
+    },
+    {
+      "epoch": 1.2956460674157304,
+      "grad_norm": 0.8890603184700012,
+      "learning_rate": 4.468989097106832e-06,
+      "loss": 0.6113,
+      "step": 3690
+    },
+    {
+      "epoch": 1.295997191011236,
+      "grad_norm": 0.9091203808784485,
+      "learning_rate": 4.468704182110346e-06,
+      "loss": 0.6044,
+      "step": 3691
+    },
+    {
+      "epoch": 1.2963483146067416,
+      "grad_norm": 0.8641930818557739,
+      "learning_rate": 4.46841919978586e-06,
+      "loss": 0.6264,
+      "step": 3692
+    },
+    {
+      "epoch": 1.2966994382022472,
+      "grad_norm": 0.8812434673309326,
+      "learning_rate": 4.468134150143123e-06,
+      "loss": 0.6027,
+      "step": 3693
+    },
+    {
+      "epoch": 1.2970505617977528,
+      "grad_norm": 0.9063153266906738,
+      "learning_rate": 4.467849033191882e-06,
+      "loss": 0.6429,
+      "step": 3694
+    },
+    {
+      "epoch": 1.2974016853932584,
+      "grad_norm": 0.8584569096565247,
+      "learning_rate": 4.467563848941888e-06,
+      "loss": 0.6003,
+      "step": 3695
+    },
+    {
+      "epoch": 1.297752808988764,
+      "grad_norm": 0.8659058809280396,
+      "learning_rate": 4.4672785974028935e-06,
+      "loss": 0.5802,
+      "step": 3696
+    },
+    {
+      "epoch": 1.2981039325842696,
+      "grad_norm": 0.8678346872329712,
+      "learning_rate": 4.4669932785846545e-06,
+      "loss": 0.6162,
+      "step": 3697
+    },
+    {
+      "epoch": 1.2984550561797752,
+      "grad_norm": 0.8793441653251648,
+      "learning_rate": 4.466707892496929e-06,
+      "loss": 0.6187,
+      "step": 3698
+    },
+    {
+      "epoch": 1.2988061797752808,
+      "grad_norm": 0.8773770332336426,
+      "learning_rate": 4.466422439149477e-06,
+      "loss": 0.6057,
+      "step": 3699
+    },
+    {
+      "epoch": 1.2991573033707864,
+      "grad_norm": 0.8449274897575378,
+      "learning_rate": 4.4661369185520594e-06,
+      "loss": 0.6015,
+      "step": 3700
+    },
+    {
+      "epoch": 1.2995084269662922,
+      "grad_norm": 0.9012969136238098,
+      "learning_rate": 4.465851330714441e-06,
+      "loss": 0.6815,
+      "step": 3701
+    },
+    {
+      "epoch": 1.2998595505617978,
+      "grad_norm": 0.8823035359382629,
+      "learning_rate": 4.465565675646391e-06,
+      "loss": 0.6406,
+      "step": 3702
+    },
+    {
+      "epoch": 1.3002106741573034,
+      "grad_norm": 0.8793365955352783,
+      "learning_rate": 4.4652799533576765e-06,
+      "loss": 0.6172,
+      "step": 3703
+    },
+    {
+      "epoch": 1.300561797752809,
+      "grad_norm": 0.9065350890159607,
+      "learning_rate": 4.464994163858069e-06,
+      "loss": 0.661,
+      "step": 3704
+    },
+    {
+      "epoch": 1.3009129213483146,
+      "grad_norm": 0.960344672203064,
+      "learning_rate": 4.464708307157342e-06,
+      "loss": 0.6326,
+      "step": 3705
+    },
+    {
+      "epoch": 1.3012640449438202,
+      "grad_norm": 0.8882971405982971,
+      "learning_rate": 4.4644223832652725e-06,
+      "loss": 0.6869,
+      "step": 3706
+    },
+    {
+      "epoch": 1.3016151685393258,
+      "grad_norm": 0.9107559323310852,
+      "learning_rate": 4.464136392191638e-06,
+      "loss": 0.669,
+      "step": 3707
+    },
+    {
+      "epoch": 1.3019662921348314,
+      "grad_norm": 0.93108069896698,
+      "learning_rate": 4.46385033394622e-06,
+      "loss": 0.6978,
+      "step": 3708
+    },
+    {
+      "epoch": 1.3023174157303372,
+      "grad_norm": 0.8690692782402039,
+      "learning_rate": 4.4635642085388016e-06,
+      "loss": 0.6164,
+      "step": 3709
+    },
+    {
+      "epoch": 1.3026685393258428,
+      "grad_norm": 0.882405698299408,
+      "learning_rate": 4.463278015979166e-06,
+      "loss": 0.6205,
+      "step": 3710
+    },
+    {
+      "epoch": 1.3030196629213484,
+      "grad_norm": 0.8657339215278625,
+      "learning_rate": 4.462991756277104e-06,
+      "loss": 0.6307,
+      "step": 3711
+    },
+    {
+      "epoch": 1.303370786516854,
+      "grad_norm": 0.8758717775344849,
+      "learning_rate": 4.462705429442403e-06,
+      "loss": 0.6318,
+      "step": 3712
+    },
+    {
+      "epoch": 1.3037219101123596,
+      "grad_norm": 0.909834623336792,
+      "learning_rate": 4.4624190354848564e-06,
+      "loss": 0.6526,
+      "step": 3713
+    },
+    {
+      "epoch": 1.3040730337078652,
+      "grad_norm": 0.8672031760215759,
+      "learning_rate": 4.462132574414257e-06,
+      "loss": 0.6516,
+      "step": 3714
+    },
+    {
+      "epoch": 1.3044241573033708,
+      "grad_norm": 0.89857017993927,
+      "learning_rate": 4.461846046240404e-06,
+      "loss": 0.6645,
+      "step": 3715
+    },
+    {
+      "epoch": 1.3047752808988764,
+      "grad_norm": 0.8839460015296936,
+      "learning_rate": 4.461559450973093e-06,
+      "loss": 0.5743,
+      "step": 3716
+    },
+    {
+      "epoch": 1.305126404494382,
+      "grad_norm": 0.8894768953323364,
+      "learning_rate": 4.461272788622129e-06,
+      "loss": 0.6357,
+      "step": 3717
+    },
+    {
+      "epoch": 1.3054775280898876,
+      "grad_norm": 0.9299485683441162,
+      "learning_rate": 4.4609860591973134e-06,
+      "loss": 0.6118,
+      "step": 3718
+    },
+    {
+      "epoch": 1.3058286516853932,
+      "grad_norm": 0.8966097831726074,
+      "learning_rate": 4.460699262708452e-06,
+      "loss": 0.6128,
+      "step": 3719
+    },
+    {
+      "epoch": 1.3061797752808988,
+      "grad_norm": 0.9211109280586243,
+      "learning_rate": 4.460412399165355e-06,
+      "loss": 0.6307,
+      "step": 3720
+    },
+    {
+      "epoch": 1.3065308988764044,
+      "grad_norm": 0.8726346492767334,
+      "learning_rate": 4.46012546857783e-06,
+      "loss": 0.588,
+      "step": 3721
+    },
+    {
+      "epoch": 1.30688202247191,
+      "grad_norm": 0.937791645526886,
+      "learning_rate": 4.459838470955692e-06,
+      "loss": 0.6367,
+      "step": 3722
+    },
+    {
+      "epoch": 1.3072331460674158,
+      "grad_norm": 0.9048507213592529,
+      "learning_rate": 4.459551406308755e-06,
+      "loss": 0.6337,
+      "step": 3723
+    },
+    {
+      "epoch": 1.3075842696629214,
+      "grad_norm": 0.8875606060028076,
+      "learning_rate": 4.459264274646837e-06,
+      "loss": 0.6341,
+      "step": 3724
+    },
+    {
+      "epoch": 1.307935393258427,
+      "grad_norm": 0.8814994096755981,
+      "learning_rate": 4.458977075979758e-06,
+      "loss": 0.6406,
+      "step": 3725
+    },
+    {
+      "epoch": 1.3082865168539326,
+      "grad_norm": 0.8935707211494446,
+      "learning_rate": 4.458689810317338e-06,
+      "loss": 0.683,
+      "step": 3726
+    },
+    {
+      "epoch": 1.3086376404494382,
+      "grad_norm": 0.9421975612640381,
+      "learning_rate": 4.458402477669403e-06,
+      "loss": 0.6653,
+      "step": 3727
+    },
+    {
+      "epoch": 1.3089887640449438,
+      "grad_norm": 0.871554434299469,
+      "learning_rate": 4.45811507804578e-06,
+      "loss": 0.6291,
+      "step": 3728
+    },
+    {
+      "epoch": 1.3093398876404494,
+      "grad_norm": 0.9223468899726868,
+      "learning_rate": 4.457827611456296e-06,
+      "loss": 0.6447,
+      "step": 3729
+    },
+    {
+      "epoch": 1.309691011235955,
+      "grad_norm": 0.9151185154914856,
+      "learning_rate": 4.457540077910784e-06,
+      "loss": 0.6106,
+      "step": 3730
+    },
+    {
+      "epoch": 1.3100421348314606,
+      "grad_norm": 0.8625356554985046,
+      "learning_rate": 4.4572524774190754e-06,
+      "loss": 0.6024,
+      "step": 3731
+    },
+    {
+      "epoch": 1.3103932584269664,
+      "grad_norm": 0.8800901174545288,
+      "learning_rate": 4.456964809991007e-06,
+      "loss": 0.6199,
+      "step": 3732
+    },
+    {
+      "epoch": 1.310744382022472,
+      "grad_norm": 0.8656051754951477,
+      "learning_rate": 4.4566770756364165e-06,
+      "loss": 0.6249,
+      "step": 3733
+    },
+    {
+      "epoch": 1.3110955056179776,
+      "grad_norm": 0.8878893256187439,
+      "learning_rate": 4.456389274365145e-06,
+      "loss": 0.6181,
+      "step": 3734
+    },
+    {
+      "epoch": 1.3114466292134832,
+      "grad_norm": 0.8846556544303894,
+      "learning_rate": 4.456101406187035e-06,
+      "loss": 0.6395,
+      "step": 3735
+    },
+    {
+      "epoch": 1.3117977528089888,
+      "grad_norm": 0.9168640375137329,
+      "learning_rate": 4.4558134711119295e-06,
+      "loss": 0.6336,
+      "step": 3736
+    },
+    {
+      "epoch": 1.3121488764044944,
+      "grad_norm": 0.9098040461540222,
+      "learning_rate": 4.455525469149677e-06,
+      "loss": 0.6639,
+      "step": 3737
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 0.8935442566871643,
+      "learning_rate": 4.455237400310127e-06,
+      "loss": 0.6221,
+      "step": 3738
+    },
+    {
+      "epoch": 1.3128511235955056,
+      "grad_norm": 0.8903330564498901,
+      "learning_rate": 4.454949264603132e-06,
+      "loss": 0.6484,
+      "step": 3739
+    },
+    {
+      "epoch": 1.3132022471910112,
+      "grad_norm": 0.8853354454040527,
+      "learning_rate": 4.454661062038544e-06,
+      "loss": 0.5822,
+      "step": 3740
+    },
+    {
+      "epoch": 1.3135533707865168,
+      "grad_norm": 0.9109371304512024,
+      "learning_rate": 4.45437279262622e-06,
+      "loss": 0.6844,
+      "step": 3741
+    },
+    {
+      "epoch": 1.3139044943820224,
+      "grad_norm": 0.8940867781639099,
+      "learning_rate": 4.45408445637602e-06,
+      "loss": 0.6693,
+      "step": 3742
+    },
+    {
+      "epoch": 1.314255617977528,
+      "grad_norm": 0.9142806529998779,
+      "learning_rate": 4.453796053297803e-06,
+      "loss": 0.6416,
+      "step": 3743
+    },
+    {
+      "epoch": 1.3146067415730336,
+      "grad_norm": 0.9041429162025452,
+      "learning_rate": 4.453507583401433e-06,
+      "loss": 0.6545,
+      "step": 3744
+    },
+    {
+      "epoch": 1.3149578651685394,
+      "grad_norm": 0.9050527811050415,
+      "learning_rate": 4.453219046696775e-06,
+      "loss": 0.6182,
+      "step": 3745
+    },
+    {
+      "epoch": 1.315308988764045,
+      "grad_norm": 0.8687748312950134,
+      "learning_rate": 4.4529304431936976e-06,
+      "loss": 0.5813,
+      "step": 3746
+    },
+    {
+      "epoch": 1.3156601123595506,
+      "grad_norm": 0.8562371730804443,
+      "learning_rate": 4.45264177290207e-06,
+      "loss": 0.6052,
+      "step": 3747
+    },
+    {
+      "epoch": 1.3160112359550562,
+      "grad_norm": 0.8941411375999451,
+      "learning_rate": 4.452353035831764e-06,
+      "loss": 0.6132,
+      "step": 3748
+    },
+    {
+      "epoch": 1.3163623595505618,
+      "grad_norm": 0.8909905552864075,
+      "learning_rate": 4.452064231992655e-06,
+      "loss": 0.6323,
+      "step": 3749
+    },
+    {
+      "epoch": 1.3167134831460674,
+      "grad_norm": 0.8690940737724304,
+      "learning_rate": 4.451775361394621e-06,
+      "loss": 0.6171,
+      "step": 3750
+    },
+    {
+      "epoch": 1.317064606741573,
+      "grad_norm": 0.8960789442062378,
+      "learning_rate": 4.451486424047538e-06,
+      "loss": 0.6539,
+      "step": 3751
+    },
+    {
+      "epoch": 1.3174157303370786,
+      "grad_norm": 0.8996719717979431,
+      "learning_rate": 4.451197419961291e-06,
+      "loss": 0.6298,
+      "step": 3752
+    },
+    {
+      "epoch": 1.3177668539325842,
+      "grad_norm": 0.9086701273918152,
+      "learning_rate": 4.4509083491457605e-06,
+      "loss": 0.6719,
+      "step": 3753
+    },
+    {
+      "epoch": 1.31811797752809,
+      "grad_norm": 0.8379474878311157,
+      "learning_rate": 4.450619211610835e-06,
+      "loss": 0.6051,
+      "step": 3754
+    },
+    {
+      "epoch": 1.3184691011235956,
+      "grad_norm": 0.8519045114517212,
+      "learning_rate": 4.4503300073664e-06,
+      "loss": 0.6239,
+      "step": 3755
+    },
+    {
+      "epoch": 1.3188202247191012,
+      "grad_norm": 0.8749707341194153,
+      "learning_rate": 4.450040736422349e-06,
+      "loss": 0.614,
+      "step": 3756
+    },
+    {
+      "epoch": 1.3191713483146068,
+      "grad_norm": 0.9106235504150391,
+      "learning_rate": 4.4497513987885735e-06,
+      "loss": 0.6247,
+      "step": 3757
+    },
+    {
+      "epoch": 1.3195224719101124,
+      "grad_norm": 0.8531911969184875,
+      "learning_rate": 4.449461994474968e-06,
+      "loss": 0.5988,
+      "step": 3758
+    },
+    {
+      "epoch": 1.319873595505618,
+      "grad_norm": 0.8912167549133301,
+      "learning_rate": 4.449172523491431e-06,
+      "loss": 0.6383,
+      "step": 3759
+    },
+    {
+      "epoch": 1.3202247191011236,
+      "grad_norm": 0.8568180203437805,
+      "learning_rate": 4.448882985847861e-06,
+      "loss": 0.6366,
+      "step": 3760
+    },
+    {
+      "epoch": 1.3205758426966292,
+      "grad_norm": 0.9863873720169067,
+      "learning_rate": 4.448593381554161e-06,
+      "loss": 0.6223,
+      "step": 3761
+    },
+    {
+      "epoch": 1.3209269662921348,
+      "grad_norm": 0.8997707366943359,
+      "learning_rate": 4.448303710620234e-06,
+      "loss": 0.6224,
+      "step": 3762
+    },
+    {
+      "epoch": 1.3212780898876404,
+      "grad_norm": 0.8545982837677002,
+      "learning_rate": 4.448013973055988e-06,
+      "loss": 0.597,
+      "step": 3763
+    },
+    {
+      "epoch": 1.321629213483146,
+      "grad_norm": 0.9015237092971802,
+      "learning_rate": 4.447724168871331e-06,
+      "loss": 0.6257,
+      "step": 3764
+    },
+    {
+      "epoch": 1.3219803370786516,
+      "grad_norm": 0.8829646110534668,
+      "learning_rate": 4.447434298076174e-06,
+      "loss": 0.6348,
+      "step": 3765
+    },
+    {
+      "epoch": 1.3223314606741572,
+      "grad_norm": 0.8937842845916748,
+      "learning_rate": 4.447144360680431e-06,
+      "loss": 0.6363,
+      "step": 3766
+    },
+    {
+      "epoch": 1.3226825842696628,
+      "grad_norm": 0.8497073650360107,
+      "learning_rate": 4.446854356694017e-06,
+      "loss": 0.5725,
+      "step": 3767
+    },
+    {
+      "epoch": 1.3230337078651686,
+      "grad_norm": 0.8760591149330139,
+      "learning_rate": 4.446564286126849e-06,
+      "loss": 0.6221,
+      "step": 3768
+    },
+    {
+      "epoch": 1.3233848314606742,
+      "grad_norm": 0.8975864052772522,
+      "learning_rate": 4.446274148988849e-06,
+      "loss": 0.6172,
+      "step": 3769
+    },
+    {
+      "epoch": 1.3237359550561798,
+      "grad_norm": 0.9180589914321899,
+      "learning_rate": 4.4459839452899385e-06,
+      "loss": 0.6418,
+      "step": 3770
+    },
+    {
+      "epoch": 1.3240870786516854,
+      "grad_norm": 0.8982850313186646,
+      "learning_rate": 4.4456936750400415e-06,
+      "loss": 0.6271,
+      "step": 3771
+    },
+    {
+      "epoch": 1.324438202247191,
+      "grad_norm": 0.8551401495933533,
+      "learning_rate": 4.445403338249087e-06,
+      "loss": 0.6145,
+      "step": 3772
+    },
+    {
+      "epoch": 1.3247893258426966,
+      "grad_norm": 0.9100740551948547,
+      "learning_rate": 4.445112934927002e-06,
+      "loss": 0.6249,
+      "step": 3773
+    },
+    {
+      "epoch": 1.3251404494382022,
+      "grad_norm": 0.8740219473838806,
+      "learning_rate": 4.444822465083719e-06,
+      "loss": 0.6523,
+      "step": 3774
+    },
+    {
+      "epoch": 1.3254915730337078,
+      "grad_norm": 0.9058376550674438,
+      "learning_rate": 4.444531928729173e-06,
+      "loss": 0.6331,
+      "step": 3775
+    },
+    {
+      "epoch": 1.3258426966292136,
+      "grad_norm": 0.9305749535560608,
+      "learning_rate": 4.444241325873298e-06,
+      "loss": 0.6519,
+      "step": 3776
+    },
+    {
+      "epoch": 1.3261938202247192,
+      "grad_norm": 0.8894662261009216,
+      "learning_rate": 4.443950656526034e-06,
+      "loss": 0.6155,
+      "step": 3777
+    },
+    {
+      "epoch": 1.3265449438202248,
+      "grad_norm": 0.9070370197296143,
+      "learning_rate": 4.4436599206973205e-06,
+      "loss": 0.6019,
+      "step": 3778
+    },
+    {
+      "epoch": 1.3268960674157304,
+      "grad_norm": 0.8799253106117249,
+      "learning_rate": 4.4433691183971006e-06,
+      "loss": 0.6568,
+      "step": 3779
+    },
+    {
+      "epoch": 1.327247191011236,
+      "grad_norm": 0.9257767200469971,
+      "learning_rate": 4.4430782496353205e-06,
+      "loss": 0.6571,
+      "step": 3780
+    },
+    {
+      "epoch": 1.3275983146067416,
+      "grad_norm": 0.8728551864624023,
+      "learning_rate": 4.442787314421927e-06,
+      "loss": 0.6364,
+      "step": 3781
+    },
+    {
+      "epoch": 1.3279494382022472,
+      "grad_norm": 0.8830699324607849,
+      "learning_rate": 4.442496312766869e-06,
+      "loss": 0.629,
+      "step": 3782
+    },
+    {
+      "epoch": 1.3283005617977528,
+      "grad_norm": 0.9129642248153687,
+      "learning_rate": 4.4422052446801e-06,
+      "loss": 0.6377,
+      "step": 3783
+    },
+    {
+      "epoch": 1.3286516853932584,
+      "grad_norm": 0.8649726510047913,
+      "learning_rate": 4.441914110171574e-06,
+      "loss": 0.6565,
+      "step": 3784
+    },
+    {
+      "epoch": 1.329002808988764,
+      "grad_norm": 0.8841148018836975,
+      "learning_rate": 4.441622909251246e-06,
+      "loss": 0.6232,
+      "step": 3785
+    },
+    {
+      "epoch": 1.3293539325842696,
+      "grad_norm": 0.9116290211677551,
+      "learning_rate": 4.4413316419290775e-06,
+      "loss": 0.6522,
+      "step": 3786
+    },
+    {
+      "epoch": 1.3297050561797752,
+      "grad_norm": 0.8940332531929016,
+      "learning_rate": 4.441040308215027e-06,
+      "loss": 0.6147,
+      "step": 3787
+    },
+    {
+      "epoch": 1.3300561797752808,
+      "grad_norm": 0.8921476602554321,
+      "learning_rate": 4.4407489081190594e-06,
+      "loss": 0.6297,
+      "step": 3788
+    },
+    {
+      "epoch": 1.3304073033707864,
+      "grad_norm": 0.8484114408493042,
+      "learning_rate": 4.440457441651139e-06,
+      "loss": 0.5943,
+      "step": 3789
+    },
+    {
+      "epoch": 1.3307584269662922,
+      "grad_norm": 0.8836881518363953,
+      "learning_rate": 4.4401659088212356e-06,
+      "loss": 0.613,
+      "step": 3790
+    },
+    {
+      "epoch": 1.3311095505617978,
+      "grad_norm": 0.8993842005729675,
+      "learning_rate": 4.439874309639318e-06,
+      "loss": 0.653,
+      "step": 3791
+    },
+    {
+      "epoch": 1.3314606741573034,
+      "grad_norm": 0.896313488483429,
+      "learning_rate": 4.439582644115359e-06,
+      "loss": 0.6625,
+      "step": 3792
+    },
+    {
+      "epoch": 1.331811797752809,
+      "grad_norm": 0.8841575980186462,
+      "learning_rate": 4.439290912259333e-06,
+      "loss": 0.6328,
+      "step": 3793
+    },
+    {
+      "epoch": 1.3321629213483146,
+      "grad_norm": 0.9332913160324097,
+      "learning_rate": 4.438999114081218e-06,
+      "loss": 0.6589,
+      "step": 3794
+    },
+    {
+      "epoch": 1.3325140449438202,
+      "grad_norm": 0.872435450553894,
+      "learning_rate": 4.438707249590992e-06,
+      "loss": 0.6348,
+      "step": 3795
+    },
+    {
+      "epoch": 1.3328651685393258,
+      "grad_norm": 0.8649969100952148,
+      "learning_rate": 4.4384153187986365e-06,
+      "loss": 0.6116,
+      "step": 3796
+    },
+    {
+      "epoch": 1.3332162921348314,
+      "grad_norm": 0.8806115984916687,
+      "learning_rate": 4.438123321714137e-06,
+      "loss": 0.618,
+      "step": 3797
+    },
+    {
+      "epoch": 1.3335674157303372,
+      "grad_norm": 0.905692994594574,
+      "learning_rate": 4.437831258347477e-06,
+      "loss": 0.6558,
+      "step": 3798
+    },
+    {
+      "epoch": 1.3339185393258428,
+      "grad_norm": 0.906639039516449,
+      "learning_rate": 4.437539128708647e-06,
+      "loss": 0.6338,
+      "step": 3799
+    },
+    {
+      "epoch": 1.3342696629213484,
+      "grad_norm": 0.925403356552124,
+      "learning_rate": 4.4372469328076375e-06,
+      "loss": 0.6615,
+      "step": 3800
+    },
+    {
+      "epoch": 1.334620786516854,
+      "grad_norm": 0.8646488189697266,
+      "learning_rate": 4.436954670654439e-06,
+      "loss": 0.6128,
+      "step": 3801
+    },
+    {
+      "epoch": 1.3349719101123596,
+      "grad_norm": 0.8448339700698853,
+      "learning_rate": 4.436662342259049e-06,
+      "loss": 0.6352,
+      "step": 3802
+    },
+    {
+      "epoch": 1.3353230337078652,
+      "grad_norm": 0.8907808661460876,
+      "learning_rate": 4.436369947631465e-06,
+      "loss": 0.6108,
+      "step": 3803
+    },
+    {
+      "epoch": 1.3356741573033708,
+      "grad_norm": 0.8870221972465515,
+      "learning_rate": 4.436077486781685e-06,
+      "loss": 0.6391,
+      "step": 3804
+    },
+    {
+      "epoch": 1.3360252808988764,
+      "grad_norm": 0.8831346035003662,
+      "learning_rate": 4.435784959719711e-06,
+      "loss": 0.6947,
+      "step": 3805
+    },
+    {
+      "epoch": 1.336376404494382,
+      "grad_norm": 0.9102858304977417,
+      "learning_rate": 4.4354923664555485e-06,
+      "loss": 0.6248,
+      "step": 3806
+    },
+    {
+      "epoch": 1.3367275280898876,
+      "grad_norm": 0.8967962265014648,
+      "learning_rate": 4.435199706999204e-06,
+      "loss": 0.6066,
+      "step": 3807
+    },
+    {
+      "epoch": 1.3370786516853932,
+      "grad_norm": 0.8433191180229187,
+      "learning_rate": 4.434906981360684e-06,
+      "loss": 0.62,
+      "step": 3808
+    },
+    {
+      "epoch": 1.3374297752808988,
+      "grad_norm": 0.8778981566429138,
+      "learning_rate": 4.434614189550003e-06,
+      "loss": 0.6213,
+      "step": 3809
+    },
+    {
+      "epoch": 1.3377808988764044,
+      "grad_norm": 0.9490671753883362,
+      "learning_rate": 4.4343213315771695e-06,
+      "loss": 0.6183,
+      "step": 3810
+    },
+    {
+      "epoch": 1.33813202247191,
+      "grad_norm": 0.8874966502189636,
+      "learning_rate": 4.434028407452203e-06,
+      "loss": 0.6473,
+      "step": 3811
+    },
+    {
+      "epoch": 1.3384831460674158,
+      "grad_norm": 0.8727471828460693,
+      "learning_rate": 4.433735417185121e-06,
+      "loss": 0.6234,
+      "step": 3812
+    },
+    {
+      "epoch": 1.3388342696629214,
+      "grad_norm": 0.9094645977020264,
+      "learning_rate": 4.433442360785941e-06,
+      "loss": 0.6511,
+      "step": 3813
+    },
+    {
+      "epoch": 1.339185393258427,
+      "grad_norm": 0.934386670589447,
+      "learning_rate": 4.433149238264688e-06,
+      "loss": 0.6563,
+      "step": 3814
+    },
+    {
+      "epoch": 1.3395365168539326,
+      "grad_norm": 0.8715890049934387,
+      "learning_rate": 4.432856049631385e-06,
+      "loss": 0.6598,
+      "step": 3815
+    },
+    {
+      "epoch": 1.3398876404494382,
+      "grad_norm": 0.8467699289321899,
+      "learning_rate": 4.432562794896058e-06,
+      "loss": 0.6224,
+      "step": 3816
+    },
+    {
+      "epoch": 1.3402387640449438,
+      "grad_norm": 0.8690305948257446,
+      "learning_rate": 4.4322694740687375e-06,
+      "loss": 0.6075,
+      "step": 3817
+    },
+    {
+      "epoch": 1.3405898876404494,
+      "grad_norm": 0.8769239783287048,
+      "learning_rate": 4.431976087159455e-06,
+      "loss": 0.6145,
+      "step": 3818
+    },
+    {
+      "epoch": 1.340941011235955,
+      "grad_norm": 0.8641560673713684,
+      "learning_rate": 4.431682634178243e-06,
+      "loss": 0.629,
+      "step": 3819
+    },
+    {
+      "epoch": 1.3412921348314606,
+      "grad_norm": 0.8964089155197144,
+      "learning_rate": 4.431389115135138e-06,
+      "loss": 0.6456,
+      "step": 3820
+    },
+    {
+      "epoch": 1.3416432584269664,
+      "grad_norm": 0.8796816468238831,
+      "learning_rate": 4.4310955300401785e-06,
+      "loss": 0.6511,
+      "step": 3821
+    },
+    {
+      "epoch": 1.341994382022472,
+      "grad_norm": 0.8870885372161865,
+      "learning_rate": 4.430801878903403e-06,
+      "loss": 0.6458,
+      "step": 3822
+    },
+    {
+      "epoch": 1.3423455056179776,
+      "grad_norm": 0.8509821891784668,
+      "learning_rate": 4.430508161734857e-06,
+      "loss": 0.6291,
+      "step": 3823
+    },
+    {
+      "epoch": 1.3426966292134832,
+      "grad_norm": 0.8842064738273621,
+      "learning_rate": 4.430214378544583e-06,
+      "loss": 0.6223,
+      "step": 3824
+    },
+    {
+      "epoch": 1.3430477528089888,
+      "grad_norm": 0.882311224937439,
+      "learning_rate": 4.429920529342629e-06,
+      "loss": 0.5903,
+      "step": 3825
+    },
+    {
+      "epoch": 1.3433988764044944,
+      "grad_norm": 0.9211931228637695,
+      "learning_rate": 4.429626614139044e-06,
+      "loss": 0.6096,
+      "step": 3826
+    },
+    {
+      "epoch": 1.34375,
+      "grad_norm": 0.9198620319366455,
+      "learning_rate": 4.429332632943881e-06,
+      "loss": 0.6021,
+      "step": 3827
+    },
+    {
+      "epoch": 1.3441011235955056,
+      "grad_norm": 0.8927902579307556,
+      "learning_rate": 4.4290385857671924e-06,
+      "loss": 0.6681,
+      "step": 3828
+    },
+    {
+      "epoch": 1.3444522471910112,
+      "grad_norm": 0.8720464706420898,
+      "learning_rate": 4.4287444726190346e-06,
+      "loss": 0.643,
+      "step": 3829
+    },
+    {
+      "epoch": 1.3448033707865168,
+      "grad_norm": 0.9204714298248291,
+      "learning_rate": 4.428450293509467e-06,
+      "loss": 0.6289,
+      "step": 3830
+    },
+    {
+      "epoch": 1.3451544943820224,
+      "grad_norm": 0.8725011348724365,
+      "learning_rate": 4.428156048448549e-06,
+      "loss": 0.6461,
+      "step": 3831
+    },
+    {
+      "epoch": 1.345505617977528,
+      "grad_norm": 0.9051564335823059,
+      "learning_rate": 4.427861737446344e-06,
+      "loss": 0.6235,
+      "step": 3832
+    },
+    {
+      "epoch": 1.3458567415730336,
+      "grad_norm": 0.9251704216003418,
+      "learning_rate": 4.427567360512917e-06,
+      "loss": 0.648,
+      "step": 3833
+    },
+    {
+      "epoch": 1.3462078651685394,
+      "grad_norm": 0.8915055990219116,
+      "learning_rate": 4.427272917658337e-06,
+      "loss": 0.6474,
+      "step": 3834
+    },
+    {
+      "epoch": 1.346558988764045,
+      "grad_norm": 0.8907241225242615,
+      "learning_rate": 4.426978408892672e-06,
+      "loss": 0.6431,
+      "step": 3835
+    },
+    {
+      "epoch": 1.3469101123595506,
+      "grad_norm": 0.8840727806091309,
+      "learning_rate": 4.426683834225993e-06,
+      "loss": 0.6302,
+      "step": 3836
+    },
+    {
+      "epoch": 1.3472612359550562,
+      "grad_norm": 0.8734930157661438,
+      "learning_rate": 4.4263891936683765e-06,
+      "loss": 0.5856,
+      "step": 3837
+    },
+    {
+      "epoch": 1.3476123595505618,
+      "grad_norm": 0.9012126326560974,
+      "learning_rate": 4.426094487229898e-06,
+      "loss": 0.6484,
+      "step": 3838
+    },
+    {
+      "epoch": 1.3479634831460674,
+      "grad_norm": 0.8863605856895447,
+      "learning_rate": 4.425799714920636e-06,
+      "loss": 0.6754,
+      "step": 3839
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8811327815055847,
+      "learning_rate": 4.4255048767506725e-06,
+      "loss": 0.6565,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3486657303370786,
+      "grad_norm": 0.9418013691902161,
+      "learning_rate": 4.425209972730089e-06,
+      "loss": 0.6495,
+      "step": 3841
+    },
+    {
+      "epoch": 1.3490168539325842,
+      "grad_norm": 0.9380903244018555,
+      "learning_rate": 4.4249150028689726e-06,
+      "loss": 0.6224,
+      "step": 3842
+    },
+    {
+      "epoch": 1.34936797752809,
+      "grad_norm": 0.8815275430679321,
+      "learning_rate": 4.4246199671774094e-06,
+      "loss": 0.6617,
+      "step": 3843
+    },
+    {
+      "epoch": 1.3497191011235956,
+      "grad_norm": 0.8603273630142212,
+      "learning_rate": 4.424324865665491e-06,
+      "loss": 0.6247,
+      "step": 3844
+    },
+    {
+      "epoch": 1.3500702247191012,
+      "grad_norm": 0.904721736907959,
+      "learning_rate": 4.424029698343309e-06,
+      "loss": 0.6487,
+      "step": 3845
+    },
+    {
+      "epoch": 1.3504213483146068,
+      "grad_norm": 0.873080313205719,
+      "learning_rate": 4.423734465220957e-06,
+      "loss": 0.6169,
+      "step": 3846
+    },
+    {
+      "epoch": 1.3507724719101124,
+      "grad_norm": 0.8521433472633362,
+      "learning_rate": 4.4234391663085324e-06,
+      "loss": 0.6537,
+      "step": 3847
+    },
+    {
+      "epoch": 1.351123595505618,
+      "grad_norm": 0.9350584745407104,
+      "learning_rate": 4.423143801616134e-06,
+      "loss": 0.6279,
+      "step": 3848
+    },
+    {
+      "epoch": 1.3514747191011236,
+      "grad_norm": 0.8658765554428101,
+      "learning_rate": 4.422848371153865e-06,
+      "loss": 0.6363,
+      "step": 3849
+    },
+    {
+      "epoch": 1.3518258426966292,
+      "grad_norm": 0.9064232110977173,
+      "learning_rate": 4.422552874931825e-06,
+      "loss": 0.6337,
+      "step": 3850
+    },
+    {
+      "epoch": 1.3521769662921348,
+      "grad_norm": 0.8607880473136902,
+      "learning_rate": 4.422257312960123e-06,
+      "loss": 0.6058,
+      "step": 3851
+    },
+    {
+      "epoch": 1.3525280898876404,
+      "grad_norm": 0.8825393319129944,
+      "learning_rate": 4.421961685248866e-06,
+      "loss": 0.6546,
+      "step": 3852
+    },
+    {
+      "epoch": 1.352879213483146,
+      "grad_norm": 0.8812253475189209,
+      "learning_rate": 4.421665991808163e-06,
+      "loss": 0.607,
+      "step": 3853
+    },
+    {
+      "epoch": 1.3532303370786516,
+      "grad_norm": 0.8943750262260437,
+      "learning_rate": 4.421370232648128e-06,
+      "loss": 0.6298,
+      "step": 3854
+    },
+    {
+      "epoch": 1.3535814606741572,
+      "grad_norm": 0.8581367135047913,
+      "learning_rate": 4.4210744077788755e-06,
+      "loss": 0.6057,
+      "step": 3855
+    },
+    {
+      "epoch": 1.3539325842696628,
+      "grad_norm": 0.8943094611167908,
+      "learning_rate": 4.4207785172105216e-06,
+      "loss": 0.6333,
+      "step": 3856
+    },
+    {
+      "epoch": 1.3542837078651686,
+      "grad_norm": 0.8830317854881287,
+      "learning_rate": 4.420482560953186e-06,
+      "loss": 0.651,
+      "step": 3857
+    },
+    {
+      "epoch": 1.3546348314606742,
+      "grad_norm": 0.8857917189598083,
+      "learning_rate": 4.420186539016991e-06,
+      "loss": 0.6463,
+      "step": 3858
+    },
+    {
+      "epoch": 1.3549859550561798,
+      "grad_norm": 0.8663979768753052,
+      "learning_rate": 4.419890451412059e-06,
+      "loss": 0.5975,
+      "step": 3859
+    },
+    {
+      "epoch": 1.3553370786516854,
+      "grad_norm": 0.8694441318511963,
+      "learning_rate": 4.4195942981485165e-06,
+      "loss": 0.6093,
+      "step": 3860
+    },
+    {
+      "epoch": 1.355688202247191,
+      "grad_norm": 0.8855713605880737,
+      "learning_rate": 4.419298079236491e-06,
+      "loss": 0.625,
+      "step": 3861
+    },
+    {
+      "epoch": 1.3560393258426966,
+      "grad_norm": 0.9174140095710754,
+      "learning_rate": 4.419001794686114e-06,
+      "loss": 0.6448,
+      "step": 3862
+    },
+    {
+      "epoch": 1.3563904494382022,
+      "grad_norm": 0.8802694082260132,
+      "learning_rate": 4.418705444507518e-06,
+      "loss": 0.681,
+      "step": 3863
+    },
+    {
+      "epoch": 1.3567415730337078,
+      "grad_norm": 0.8532006740570068,
+      "learning_rate": 4.418409028710838e-06,
+      "loss": 0.6378,
+      "step": 3864
+    },
+    {
+      "epoch": 1.3570926966292136,
+      "grad_norm": 0.8874769806861877,
+      "learning_rate": 4.41811254730621e-06,
+      "loss": 0.6397,
+      "step": 3865
+    },
+    {
+      "epoch": 1.3574438202247192,
+      "grad_norm": 0.9477583765983582,
+      "learning_rate": 4.417816000303775e-06,
+      "loss": 0.62,
+      "step": 3866
+    },
+    {
+      "epoch": 1.3577949438202248,
+      "grad_norm": 0.8806896805763245,
+      "learning_rate": 4.417519387713673e-06,
+      "loss": 0.6324,
+      "step": 3867
+    },
+    {
+      "epoch": 1.3581460674157304,
+      "grad_norm": 0.8547712564468384,
+      "learning_rate": 4.41722270954605e-06,
+      "loss": 0.6066,
+      "step": 3868
+    },
+    {
+      "epoch": 1.358497191011236,
+      "grad_norm": 0.895625114440918,
+      "learning_rate": 4.416925965811049e-06,
+      "loss": 0.6373,
+      "step": 3869
+    },
+    {
+      "epoch": 1.3588483146067416,
+      "grad_norm": 0.895209550857544,
+      "learning_rate": 4.416629156518822e-06,
+      "loss": 0.6408,
+      "step": 3870
+    },
+    {
+      "epoch": 1.3591994382022472,
+      "grad_norm": 0.8948414325714111,
+      "learning_rate": 4.416332281679517e-06,
+      "loss": 0.6238,
+      "step": 3871
+    },
+    {
+      "epoch": 1.3595505617977528,
+      "grad_norm": 0.8785004019737244,
+      "learning_rate": 4.4160353413032885e-06,
+      "loss": 0.6141,
+      "step": 3872
+    },
+    {
+      "epoch": 1.3599016853932584,
+      "grad_norm": 0.9018064737319946,
+      "learning_rate": 4.415738335400291e-06,
+      "loss": 0.6668,
+      "step": 3873
+    },
+    {
+      "epoch": 1.360252808988764,
+      "grad_norm": 0.8808990120887756,
+      "learning_rate": 4.415441263980681e-06,
+      "loss": 0.634,
+      "step": 3874
+    },
+    {
+      "epoch": 1.3606039325842696,
+      "grad_norm": 0.8734375834465027,
+      "learning_rate": 4.415144127054619e-06,
+      "loss": 0.65,
+      "step": 3875
+    },
+    {
+      "epoch": 1.3609550561797752,
+      "grad_norm": 0.8722198605537415,
+      "learning_rate": 4.414846924632267e-06,
+      "loss": 0.6227,
+      "step": 3876
+    },
+    {
+      "epoch": 1.3613061797752808,
+      "grad_norm": 0.8828457593917847,
+      "learning_rate": 4.414549656723789e-06,
+      "loss": 0.5958,
+      "step": 3877
+    },
+    {
+      "epoch": 1.3616573033707864,
+      "grad_norm": 0.8679347634315491,
+      "learning_rate": 4.41425232333935e-06,
+      "loss": 0.6228,
+      "step": 3878
+    },
+    {
+      "epoch": 1.3620084269662922,
+      "grad_norm": 0.8740646839141846,
+      "learning_rate": 4.413954924489121e-06,
+      "loss": 0.5995,
+      "step": 3879
+    },
+    {
+      "epoch": 1.3623595505617978,
+      "grad_norm": 0.867852509021759,
+      "learning_rate": 4.41365746018327e-06,
+      "loss": 0.6606,
+      "step": 3880
+    },
+    {
+      "epoch": 1.3627106741573034,
+      "grad_norm": 0.917854905128479,
+      "learning_rate": 4.413359930431972e-06,
+      "loss": 0.6485,
+      "step": 3881
+    },
+    {
+      "epoch": 1.363061797752809,
+      "grad_norm": 0.8932327628135681,
+      "learning_rate": 4.413062335245402e-06,
+      "loss": 0.6356,
+      "step": 3882
+    },
+    {
+      "epoch": 1.3634129213483146,
+      "grad_norm": 0.8385898470878601,
+      "learning_rate": 4.412764674633736e-06,
+      "loss": 0.6072,
+      "step": 3883
+    },
+    {
+      "epoch": 1.3637640449438202,
+      "grad_norm": 0.9037755727767944,
+      "learning_rate": 4.412466948607156e-06,
+      "loss": 0.6339,
+      "step": 3884
+    },
+    {
+      "epoch": 1.3641151685393258,
+      "grad_norm": 0.882864236831665,
+      "learning_rate": 4.412169157175843e-06,
+      "loss": 0.6571,
+      "step": 3885
+    },
+    {
+      "epoch": 1.3644662921348314,
+      "grad_norm": 0.8602117896080017,
+      "learning_rate": 4.41187130034998e-06,
+      "loss": 0.6173,
+      "step": 3886
+    },
+    {
+      "epoch": 1.3648174157303372,
+      "grad_norm": 0.8562754392623901,
+      "learning_rate": 4.411573378139755e-06,
+      "loss": 0.5818,
+      "step": 3887
+    },
+    {
+      "epoch": 1.3651685393258428,
+      "grad_norm": 0.8554957509040833,
+      "learning_rate": 4.4112753905553575e-06,
+      "loss": 0.6,
+      "step": 3888
+    },
+    {
+      "epoch": 1.3655196629213484,
+      "grad_norm": 0.9020095467567444,
+      "learning_rate": 4.410977337606975e-06,
+      "loss": 0.6538,
+      "step": 3889
+    },
+    {
+      "epoch": 1.365870786516854,
+      "grad_norm": 0.9186870455741882,
+      "learning_rate": 4.410679219304804e-06,
+      "loss": 0.6393,
+      "step": 3890
+    },
+    {
+      "epoch": 1.3662219101123596,
+      "grad_norm": 0.8861616849899292,
+      "learning_rate": 4.410381035659039e-06,
+      "loss": 0.6201,
+      "step": 3891
+    },
+    {
+      "epoch": 1.3665730337078652,
+      "grad_norm": 0.8816412687301636,
+      "learning_rate": 4.4100827866798765e-06,
+      "loss": 0.6228,
+      "step": 3892
+    },
+    {
+      "epoch": 1.3669241573033708,
+      "grad_norm": 0.9039531350135803,
+      "learning_rate": 4.4097844723775185e-06,
+      "loss": 0.6368,
+      "step": 3893
+    },
+    {
+      "epoch": 1.3672752808988764,
+      "grad_norm": 0.882317066192627,
+      "learning_rate": 4.409486092762164e-06,
+      "loss": 0.6586,
+      "step": 3894
+    },
+    {
+      "epoch": 1.367626404494382,
+      "grad_norm": 0.8748815655708313,
+      "learning_rate": 4.409187647844021e-06,
+      "loss": 0.6285,
+      "step": 3895
+    },
+    {
+      "epoch": 1.3679775280898876,
+      "grad_norm": 0.8901538848876953,
+      "learning_rate": 4.408889137633292e-06,
+      "loss": 0.608,
+      "step": 3896
+    },
+    {
+      "epoch": 1.3683286516853932,
+      "grad_norm": 0.8996601104736328,
+      "learning_rate": 4.40859056214019e-06,
+      "loss": 0.6687,
+      "step": 3897
+    },
+    {
+      "epoch": 1.3686797752808988,
+      "grad_norm": 0.8397285342216492,
+      "learning_rate": 4.408291921374923e-06,
+      "loss": 0.5793,
+      "step": 3898
+    },
+    {
+      "epoch": 1.3690308988764044,
+      "grad_norm": 0.8991907835006714,
+      "learning_rate": 4.407993215347705e-06,
+      "loss": 0.6129,
+      "step": 3899
+    },
+    {
+      "epoch": 1.36938202247191,
+      "grad_norm": 0.8629971742630005,
+      "learning_rate": 4.407694444068752e-06,
+      "loss": 0.589,
+      "step": 3900
+    },
+    {
+      "epoch": 1.3697331460674158,
+      "grad_norm": 0.9028024077415466,
+      "learning_rate": 4.407395607548281e-06,
+      "loss": 0.6262,
+      "step": 3901
+    },
+    {
+      "epoch": 1.3700842696629214,
+      "grad_norm": 0.9251904487609863,
+      "learning_rate": 4.407096705796513e-06,
+      "loss": 0.645,
+      "step": 3902
+    },
+    {
+      "epoch": 1.370435393258427,
+      "grad_norm": 0.8927795886993408,
+      "learning_rate": 4.40679773882367e-06,
+      "loss": 0.6317,
+      "step": 3903
+    },
+    {
+      "epoch": 1.3707865168539326,
+      "grad_norm": 0.8791918158531189,
+      "learning_rate": 4.406498706639976e-06,
+      "loss": 0.6172,
+      "step": 3904
+    },
+    {
+      "epoch": 1.3711376404494382,
+      "grad_norm": 0.8712174296379089,
+      "learning_rate": 4.406199609255657e-06,
+      "loss": 0.6229,
+      "step": 3905
+    },
+    {
+      "epoch": 1.3714887640449438,
+      "grad_norm": 0.8883311748504639,
+      "learning_rate": 4.405900446680942e-06,
+      "loss": 0.632,
+      "step": 3906
+    },
+    {
+      "epoch": 1.3718398876404494,
+      "grad_norm": 0.8838800191879272,
+      "learning_rate": 4.4056012189260634e-06,
+      "loss": 0.6314,
+      "step": 3907
+    },
+    {
+      "epoch": 1.372191011235955,
+      "grad_norm": 0.8562436699867249,
+      "learning_rate": 4.405301926001253e-06,
+      "loss": 0.6476,
+      "step": 3908
+    },
+    {
+      "epoch": 1.3725421348314606,
+      "grad_norm": 0.8729368448257446,
+      "learning_rate": 4.405002567916749e-06,
+      "loss": 0.631,
+      "step": 3909
+    },
+    {
+      "epoch": 1.3728932584269664,
+      "grad_norm": 0.8762539625167847,
+      "learning_rate": 4.4047031446827855e-06,
+      "loss": 0.6258,
+      "step": 3910
+    },
+    {
+      "epoch": 1.373244382022472,
+      "grad_norm": 0.8716838955879211,
+      "learning_rate": 4.404403656309605e-06,
+      "loss": 0.6441,
+      "step": 3911
+    },
+    {
+      "epoch": 1.3735955056179776,
+      "grad_norm": 0.8772343993186951,
+      "learning_rate": 4.404104102807448e-06,
+      "loss": 0.622,
+      "step": 3912
+    },
+    {
+      "epoch": 1.3739466292134832,
+      "grad_norm": 0.8888761401176453,
+      "learning_rate": 4.403804484186561e-06,
+      "loss": 0.6393,
+      "step": 3913
+    },
+    {
+      "epoch": 1.3742977528089888,
+      "grad_norm": 0.891753613948822,
+      "learning_rate": 4.40350480045719e-06,
+      "loss": 0.6305,
+      "step": 3914
+    },
+    {
+      "epoch": 1.3746488764044944,
+      "grad_norm": 0.8871645927429199,
+      "learning_rate": 4.403205051629584e-06,
+      "loss": 0.6426,
+      "step": 3915
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.8718512654304504,
+      "learning_rate": 4.402905237713993e-06,
+      "loss": 0.6553,
+      "step": 3916
+    },
+    {
+      "epoch": 1.3753511235955056,
+      "grad_norm": 0.8967269062995911,
+      "learning_rate": 4.402605358720672e-06,
+      "loss": 0.654,
+      "step": 3917
+    },
+    {
+      "epoch": 1.3757022471910112,
+      "grad_norm": 0.8855593204498291,
+      "learning_rate": 4.402305414659875e-06,
+      "loss": 0.6592,
+      "step": 3918
+    },
+    {
+      "epoch": 1.3760533707865168,
+      "grad_norm": 0.859167754650116,
+      "learning_rate": 4.40200540554186e-06,
+      "loss": 0.6189,
+      "step": 3919
+    },
+    {
+      "epoch": 1.3764044943820224,
+      "grad_norm": 0.9102888703346252,
+      "learning_rate": 4.4017053313768895e-06,
+      "loss": 0.655,
+      "step": 3920
+    },
+    {
+      "epoch": 1.376755617977528,
+      "grad_norm": 0.8694773316383362,
+      "learning_rate": 4.401405192175223e-06,
+      "loss": 0.6315,
+      "step": 3921
+    },
+    {
+      "epoch": 1.3771067415730336,
+      "grad_norm": 0.8884313106536865,
+      "learning_rate": 4.401104987947126e-06,
+      "loss": 0.6305,
+      "step": 3922
+    },
+    {
+      "epoch": 1.3774578651685394,
+      "grad_norm": 0.8785154819488525,
+      "learning_rate": 4.400804718702865e-06,
+      "loss": 0.6349,
+      "step": 3923
+    },
+    {
+      "epoch": 1.377808988764045,
+      "grad_norm": 0.8957915306091309,
+      "learning_rate": 4.40050438445271e-06,
+      "loss": 0.6323,
+      "step": 3924
+    },
+    {
+      "epoch": 1.3781601123595506,
+      "grad_norm": 0.9052011966705322,
+      "learning_rate": 4.40020398520693e-06,
+      "loss": 0.6171,
+      "step": 3925
+    },
+    {
+      "epoch": 1.3785112359550562,
+      "grad_norm": 0.8835782408714294,
+      "learning_rate": 4.3999035209758015e-06,
+      "loss": 0.6299,
+      "step": 3926
+    },
+    {
+      "epoch": 1.3788623595505618,
+      "grad_norm": 0.911495566368103,
+      "learning_rate": 4.399602991769597e-06,
+      "loss": 0.6131,
+      "step": 3927
+    },
+    {
+      "epoch": 1.3792134831460674,
+      "grad_norm": 0.8838707208633423,
+      "learning_rate": 4.399302397598596e-06,
+      "loss": 0.6246,
+      "step": 3928
+    },
+    {
+      "epoch": 1.379564606741573,
+      "grad_norm": 0.9067579507827759,
+      "learning_rate": 4.399001738473079e-06,
+      "loss": 0.6554,
+      "step": 3929
+    },
+    {
+      "epoch": 1.3799157303370786,
+      "grad_norm": 0.9124329686164856,
+      "learning_rate": 4.398701014403326e-06,
+      "loss": 0.6115,
+      "step": 3930
+    },
+    {
+      "epoch": 1.3802668539325842,
+      "grad_norm": 0.9139940142631531,
+      "learning_rate": 4.398400225399624e-06,
+      "loss": 0.6373,
+      "step": 3931
+    },
+    {
+      "epoch": 1.38061797752809,
+      "grad_norm": 0.8919884562492371,
+      "learning_rate": 4.398099371472259e-06,
+      "loss": 0.6771,
+      "step": 3932
+    },
+    {
+      "epoch": 1.3809691011235956,
+      "grad_norm": 3.221696615219116,
+      "learning_rate": 4.397798452631519e-06,
+      "loss": 0.6222,
+      "step": 3933
+    },
+    {
+      "epoch": 1.3813202247191012,
+      "grad_norm": 0.8897351026535034,
+      "learning_rate": 4.397497468887697e-06,
+      "loss": 0.6688,
+      "step": 3934
+    },
+    {
+      "epoch": 1.3816713483146068,
+      "grad_norm": 0.8792717456817627,
+      "learning_rate": 4.3971964202510844e-06,
+      "loss": 0.6576,
+      "step": 3935
+    },
+    {
+      "epoch": 1.3820224719101124,
+      "grad_norm": 0.8989633321762085,
+      "learning_rate": 4.396895306731978e-06,
+      "loss": 0.6293,
+      "step": 3936
+    },
+    {
+      "epoch": 1.382373595505618,
+      "grad_norm": 0.8742619156837463,
+      "learning_rate": 4.396594128340675e-06,
+      "loss": 0.6013,
+      "step": 3937
+    },
+    {
+      "epoch": 1.3827247191011236,
+      "grad_norm": 0.8698790669441223,
+      "learning_rate": 4.396292885087476e-06,
+      "loss": 0.6172,
+      "step": 3938
+    },
+    {
+      "epoch": 1.3830758426966292,
+      "grad_norm": 0.8680218458175659,
+      "learning_rate": 4.395991576982683e-06,
+      "loss": 0.6318,
+      "step": 3939
+    },
+    {
+      "epoch": 1.3834269662921348,
+      "grad_norm": 0.9022037386894226,
+      "learning_rate": 4.3956902040366e-06,
+      "loss": 0.6423,
+      "step": 3940
+    },
+    {
+      "epoch": 1.3837780898876404,
+      "grad_norm": 0.8297759890556335,
+      "learning_rate": 4.3953887662595356e-06,
+      "loss": 0.5716,
+      "step": 3941
+    },
+    {
+      "epoch": 1.384129213483146,
+      "grad_norm": 0.9098703265190125,
+      "learning_rate": 4.395087263661796e-06,
+      "loss": 0.6599,
+      "step": 3942
+    },
+    {
+      "epoch": 1.3844803370786516,
+      "grad_norm": 0.8878948092460632,
+      "learning_rate": 4.394785696253694e-06,
+      "loss": 0.613,
+      "step": 3943
+    },
+    {
+      "epoch": 1.3848314606741572,
+      "grad_norm": 0.8692125678062439,
+      "learning_rate": 4.394484064045542e-06,
+      "loss": 0.6057,
+      "step": 3944
+    },
+    {
+      "epoch": 1.3851825842696628,
+      "grad_norm": 0.9190526008605957,
+      "learning_rate": 4.394182367047657e-06,
+      "loss": 0.6382,
+      "step": 3945
+    },
+    {
+      "epoch": 1.3855337078651686,
+      "grad_norm": 0.8912534713745117,
+      "learning_rate": 4.393880605270355e-06,
+      "loss": 0.6535,
+      "step": 3946
+    },
+    {
+      "epoch": 1.3858848314606742,
+      "grad_norm": 0.8843017220497131,
+      "learning_rate": 4.393578778723958e-06,
+      "loss": 0.6332,
+      "step": 3947
+    },
+    {
+      "epoch": 1.3862359550561798,
+      "grad_norm": 0.8994221687316895,
+      "learning_rate": 4.393276887418786e-06,
+      "loss": 0.6337,
+      "step": 3948
+    },
+    {
+      "epoch": 1.3865870786516854,
+      "grad_norm": 0.8987597823143005,
+      "learning_rate": 4.392974931365166e-06,
+      "loss": 0.6465,
+      "step": 3949
+    },
+    {
+      "epoch": 1.386938202247191,
+      "grad_norm": 0.8565405011177063,
+      "learning_rate": 4.392672910573422e-06,
+      "loss": 0.63,
+      "step": 3950
+    },
+    {
+      "epoch": 1.3872893258426966,
+      "grad_norm": 0.8897543549537659,
+      "learning_rate": 4.392370825053884e-06,
+      "loss": 0.6248,
+      "step": 3951
+    },
+    {
+      "epoch": 1.3876404494382022,
+      "grad_norm": 0.8679741621017456,
+      "learning_rate": 4.392068674816884e-06,
+      "loss": 0.6087,
+      "step": 3952
+    },
+    {
+      "epoch": 1.3879915730337078,
+      "grad_norm": 0.8487875461578369,
+      "learning_rate": 4.3917664598727535e-06,
+      "loss": 0.6154,
+      "step": 3953
+    },
+    {
+      "epoch": 1.3883426966292136,
+      "grad_norm": 0.9262088537216187,
+      "learning_rate": 4.39146418023183e-06,
+      "loss": 0.6278,
+      "step": 3954
+    },
+    {
+      "epoch": 1.3886938202247192,
+      "grad_norm": 0.9284934997558594,
+      "learning_rate": 4.391161835904449e-06,
+      "loss": 0.6405,
+      "step": 3955
+    },
+    {
+      "epoch": 1.3890449438202248,
+      "grad_norm": 0.8615259528160095,
+      "learning_rate": 4.390859426900952e-06,
+      "loss": 0.6117,
+      "step": 3956
+    },
+    {
+      "epoch": 1.3893960674157304,
+      "grad_norm": 0.9133617877960205,
+      "learning_rate": 4.39055695323168e-06,
+      "loss": 0.6036,
+      "step": 3957
+    },
+    {
+      "epoch": 1.389747191011236,
+      "grad_norm": 0.9007285237312317,
+      "learning_rate": 4.390254414906979e-06,
+      "loss": 0.6211,
+      "step": 3958
+    },
+    {
+      "epoch": 1.3900983146067416,
+      "grad_norm": 0.9187461733818054,
+      "learning_rate": 4.389951811937195e-06,
+      "loss": 0.6685,
+      "step": 3959
+    },
+    {
+      "epoch": 1.3904494382022472,
+      "grad_norm": 0.8911426663398743,
+      "learning_rate": 4.3896491443326746e-06,
+      "loss": 0.6602,
+      "step": 3960
+    },
+    {
+      "epoch": 1.3908005617977528,
+      "grad_norm": 0.9059688448905945,
+      "learning_rate": 4.389346412103771e-06,
+      "loss": 0.5913,
+      "step": 3961
+    },
+    {
+      "epoch": 1.3911516853932584,
+      "grad_norm": 0.871945858001709,
+      "learning_rate": 4.389043615260837e-06,
+      "loss": 0.6377,
+      "step": 3962
+    },
+    {
+      "epoch": 1.391502808988764,
+      "grad_norm": 0.8781225085258484,
+      "learning_rate": 4.388740753814228e-06,
+      "loss": 0.6198,
+      "step": 3963
+    },
+    {
+      "epoch": 1.3918539325842696,
+      "grad_norm": 0.8738672733306885,
+      "learning_rate": 4.388437827774301e-06,
+      "loss": 0.6079,
+      "step": 3964
+    },
+    {
+      "epoch": 1.3922050561797752,
+      "grad_norm": 0.88068026304245,
+      "learning_rate": 4.388134837151417e-06,
+      "loss": 0.628,
+      "step": 3965
+    },
+    {
+      "epoch": 1.3925561797752808,
+      "grad_norm": 0.884495735168457,
+      "learning_rate": 4.3878317819559365e-06,
+      "loss": 0.6055,
+      "step": 3966
+    },
+    {
+      "epoch": 1.3929073033707864,
+      "grad_norm": 0.8998886346817017,
+      "learning_rate": 4.387528662198225e-06,
+      "loss": 0.6411,
+      "step": 3967
+    },
+    {
+      "epoch": 1.3932584269662922,
+      "grad_norm": 0.8848565816879272,
+      "learning_rate": 4.387225477888648e-06,
+      "loss": 0.6399,
+      "step": 3968
+    },
+    {
+      "epoch": 1.3936095505617978,
+      "grad_norm": 0.9099523425102234,
+      "learning_rate": 4.386922229037575e-06,
+      "loss": 0.6303,
+      "step": 3969
+    },
+    {
+      "epoch": 1.3939606741573034,
+      "grad_norm": 0.8989171385765076,
+      "learning_rate": 4.386618915655375e-06,
+      "loss": 0.6432,
+      "step": 3970
+    },
+    {
+      "epoch": 1.394311797752809,
+      "grad_norm": 0.8913509845733643,
+      "learning_rate": 4.386315537752425e-06,
+      "loss": 0.6323,
+      "step": 3971
+    },
+    {
+      "epoch": 1.3946629213483146,
+      "grad_norm": 0.8825880885124207,
+      "learning_rate": 4.3860120953390965e-06,
+      "loss": 0.6687,
+      "step": 3972
+    },
+    {
+      "epoch": 1.3950140449438202,
+      "grad_norm": 0.8669257760047913,
+      "learning_rate": 4.3857085884257675e-06,
+      "loss": 0.6326,
+      "step": 3973
+    },
+    {
+      "epoch": 1.3953651685393258,
+      "grad_norm": 0.8898574113845825,
+      "learning_rate": 4.385405017022819e-06,
+      "loss": 0.6391,
+      "step": 3974
+    },
+    {
+      "epoch": 1.3957162921348314,
+      "grad_norm": 0.8731878399848938,
+      "learning_rate": 4.385101381140633e-06,
+      "loss": 0.613,
+      "step": 3975
+    },
+    {
+      "epoch": 1.3960674157303372,
+      "grad_norm": 0.8665676712989807,
+      "learning_rate": 4.384797680789592e-06,
+      "loss": 0.6171,
+      "step": 3976
+    },
+    {
+      "epoch": 1.3964185393258428,
+      "grad_norm": 0.9031374454498291,
+      "learning_rate": 4.384493915980083e-06,
+      "loss": 0.6174,
+      "step": 3977
+    },
+    {
+      "epoch": 1.3967696629213484,
+      "grad_norm": 0.8580880761146545,
+      "learning_rate": 4.3841900867224955e-06,
+      "loss": 0.6338,
+      "step": 3978
+    },
+    {
+      "epoch": 1.397120786516854,
+      "grad_norm": 0.9038748741149902,
+      "learning_rate": 4.383886193027219e-06,
+      "loss": 0.6258,
+      "step": 3979
+    },
+    {
+      "epoch": 1.3974719101123596,
+      "grad_norm": 0.9072511196136475,
+      "learning_rate": 4.383582234904646e-06,
+      "loss": 0.6096,
+      "step": 3980
+    },
+    {
+      "epoch": 1.3978230337078652,
+      "grad_norm": 0.8500005602836609,
+      "learning_rate": 4.383278212365174e-06,
+      "loss": 0.6366,
+      "step": 3981
+    },
+    {
+      "epoch": 1.3981741573033708,
+      "grad_norm": 0.8831138610839844,
+      "learning_rate": 4.382974125419197e-06,
+      "loss": 0.6157,
+      "step": 3982
+    },
+    {
+      "epoch": 1.3985252808988764,
+      "grad_norm": 0.9004981517791748,
+      "learning_rate": 4.382669974077117e-06,
+      "loss": 0.6431,
+      "step": 3983
+    },
+    {
+      "epoch": 1.398876404494382,
+      "grad_norm": 0.8857392072677612,
+      "learning_rate": 4.382365758349335e-06,
+      "loss": 0.6413,
+      "step": 3984
+    },
+    {
+      "epoch": 1.3992275280898876,
+      "grad_norm": 0.9009760022163391,
+      "learning_rate": 4.382061478246255e-06,
+      "loss": 0.6753,
+      "step": 3985
+    },
+    {
+      "epoch": 1.3995786516853932,
+      "grad_norm": 0.8752977252006531,
+      "learning_rate": 4.381757133778283e-06,
+      "loss": 0.6297,
+      "step": 3986
+    },
+    {
+      "epoch": 1.3999297752808988,
+      "grad_norm": 0.8759447336196899,
+      "learning_rate": 4.381452724955826e-06,
+      "loss": 0.6598,
+      "step": 3987
+    },
+    {
+      "epoch": 1.4002808988764044,
+      "grad_norm": 0.8644441962242126,
+      "learning_rate": 4.381148251789297e-06,
+      "loss": 0.6237,
+      "step": 3988
+    },
+    {
+      "epoch": 1.40063202247191,
+      "grad_norm": 0.8749090433120728,
+      "learning_rate": 4.380843714289107e-06,
+      "loss": 0.6475,
+      "step": 3989
+    },
+    {
+      "epoch": 1.4009831460674158,
+      "grad_norm": 0.8733727335929871,
+      "learning_rate": 4.380539112465672e-06,
+      "loss": 0.6421,
+      "step": 3990
+    },
+    {
+      "epoch": 1.4013342696629214,
+      "grad_norm": 0.8818216919898987,
+      "learning_rate": 4.3802344463294075e-06,
+      "loss": 0.6208,
+      "step": 3991
+    },
+    {
+      "epoch": 1.401685393258427,
+      "grad_norm": 0.8928992748260498,
+      "learning_rate": 4.379929715890735e-06,
+      "loss": 0.6075,
+      "step": 3992
+    },
+    {
+      "epoch": 1.4020365168539326,
+      "grad_norm": 0.9166174530982971,
+      "learning_rate": 4.3796249211600745e-06,
+      "loss": 0.6294,
+      "step": 3993
+    },
+    {
+      "epoch": 1.4023876404494382,
+      "grad_norm": 0.9182286858558655,
+      "learning_rate": 4.37932006214785e-06,
+      "loss": 0.6487,
+      "step": 3994
+    },
+    {
+      "epoch": 1.4027387640449438,
+      "grad_norm": 0.917757511138916,
+      "learning_rate": 4.379015138864487e-06,
+      "loss": 0.6424,
+      "step": 3995
+    },
+    {
+      "epoch": 1.4030898876404494,
+      "grad_norm": 0.912039041519165,
+      "learning_rate": 4.378710151320416e-06,
+      "loss": 0.635,
+      "step": 3996
+    },
+    {
+      "epoch": 1.403441011235955,
+      "grad_norm": 0.8986327648162842,
+      "learning_rate": 4.378405099526064e-06,
+      "loss": 0.6258,
+      "step": 3997
+    },
+    {
+      "epoch": 1.4037921348314606,
+      "grad_norm": 0.8343982100486755,
+      "learning_rate": 4.3780999834918645e-06,
+      "loss": 0.5669,
+      "step": 3998
+    },
+    {
+      "epoch": 1.4041432584269664,
+      "grad_norm": 0.8634805679321289,
+      "learning_rate": 4.377794803228254e-06,
+      "loss": 0.6714,
+      "step": 3999
+    },
+    {
+      "epoch": 1.404494382022472,
+      "grad_norm": 0.8799702525138855,
+      "learning_rate": 4.3774895587456664e-06,
+      "loss": 0.6306,
+      "step": 4000
+    },
+    {
+      "epoch": 1.4048455056179776,
+      "grad_norm": 0.8644582629203796,
+      "learning_rate": 4.377184250054544e-06,
+      "loss": 0.5853,
+      "step": 4001
+    },
+    {
+      "epoch": 1.4051966292134832,
+      "grad_norm": 0.8828877210617065,
+      "learning_rate": 4.3768788771653255e-06,
+      "loss": 0.6351,
+      "step": 4002
+    },
+    {
+      "epoch": 1.4055477528089888,
+      "grad_norm": 0.8081719875335693,
+      "learning_rate": 4.3765734400884565e-06,
+      "loss": 0.5686,
+      "step": 4003
+    },
+    {
+      "epoch": 1.4058988764044944,
+      "grad_norm": 0.873838484287262,
+      "learning_rate": 4.376267938834381e-06,
+      "loss": 0.5937,
+      "step": 4004
+    },
+    {
+      "epoch": 1.40625,
+      "grad_norm": 0.8522796630859375,
+      "learning_rate": 4.3759623734135485e-06,
+      "loss": 0.5996,
+      "step": 4005
+    },
+    {
+      "epoch": 1.4066011235955056,
+      "grad_norm": 0.8799022436141968,
+      "learning_rate": 4.3756567438364075e-06,
+      "loss": 0.61,
+      "step": 4006
+    },
+    {
+      "epoch": 1.4069522471910112,
+      "grad_norm": 0.9355555176734924,
+      "learning_rate": 4.37535105011341e-06,
+      "loss": 0.6227,
+      "step": 4007
+    },
+    {
+      "epoch": 1.4073033707865168,
+      "grad_norm": 0.91303950548172,
+      "learning_rate": 4.375045292255012e-06,
+      "loss": 0.6437,
+      "step": 4008
+    },
+    {
+      "epoch": 1.4076544943820224,
+      "grad_norm": 0.8792590498924255,
+      "learning_rate": 4.37473947027167e-06,
+      "loss": 0.6135,
+      "step": 4009
+    },
+    {
+      "epoch": 1.408005617977528,
+      "grad_norm": 0.9129072427749634,
+      "learning_rate": 4.3744335841738415e-06,
+      "loss": 0.646,
+      "step": 4010
+    },
+    {
+      "epoch": 1.4083567415730336,
+      "grad_norm": 0.8504462838172913,
+      "learning_rate": 4.37412763397199e-06,
+      "loss": 0.5935,
+      "step": 4011
+    },
+    {
+      "epoch": 1.4087078651685394,
+      "grad_norm": 0.8766793608665466,
+      "learning_rate": 4.373821619676575e-06,
+      "loss": 0.652,
+      "step": 4012
+    },
+    {
+      "epoch": 1.409058988764045,
+      "grad_norm": 0.8757210373878479,
+      "learning_rate": 4.373515541298066e-06,
+      "loss": 0.6559,
+      "step": 4013
+    },
+    {
+      "epoch": 1.4094101123595506,
+      "grad_norm": 0.8983198404312134,
+      "learning_rate": 4.373209398846927e-06,
+      "loss": 0.6412,
+      "step": 4014
+    },
+    {
+      "epoch": 1.4097612359550562,
+      "grad_norm": 0.8636668920516968,
+      "learning_rate": 4.37290319233363e-06,
+      "loss": 0.6076,
+      "step": 4015
+    },
+    {
+      "epoch": 1.4101123595505618,
+      "grad_norm": 0.9178102016448975,
+      "learning_rate": 4.372596921768647e-06,
+      "loss": 0.6572,
+      "step": 4016
+    },
+    {
+      "epoch": 1.4104634831460674,
+      "grad_norm": 0.8619410991668701,
+      "learning_rate": 4.372290587162451e-06,
+      "loss": 0.6625,
+      "step": 4017
+    },
+    {
+      "epoch": 1.410814606741573,
+      "grad_norm": 0.8461251854896545,
+      "learning_rate": 4.37198418852552e-06,
+      "loss": 0.623,
+      "step": 4018
+    },
+    {
+      "epoch": 1.4111657303370786,
+      "grad_norm": 0.867741584777832,
+      "learning_rate": 4.37167772586833e-06,
+      "loss": 0.6362,
+      "step": 4019
+    },
+    {
+      "epoch": 1.4115168539325842,
+      "grad_norm": 0.8633602261543274,
+      "learning_rate": 4.3713711992013656e-06,
+      "loss": 0.592,
+      "step": 4020
+    },
+    {
+      "epoch": 1.41186797752809,
+      "grad_norm": 0.8842202425003052,
+      "learning_rate": 4.371064608535106e-06,
+      "loss": 0.632,
+      "step": 4021
+    },
+    {
+      "epoch": 1.4122191011235956,
+      "grad_norm": 0.8827239274978638,
+      "learning_rate": 4.370757953880038e-06,
+      "loss": 0.6164,
+      "step": 4022
+    },
+    {
+      "epoch": 1.4125702247191012,
+      "grad_norm": 0.8834839463233948,
+      "learning_rate": 4.370451235246649e-06,
+      "loss": 0.5874,
+      "step": 4023
+    },
+    {
+      "epoch": 1.4129213483146068,
+      "grad_norm": 0.9291568398475647,
+      "learning_rate": 4.370144452645429e-06,
+      "loss": 0.6186,
+      "step": 4024
+    },
+    {
+      "epoch": 1.4132724719101124,
+      "grad_norm": 0.8816742897033691,
+      "learning_rate": 4.369837606086869e-06,
+      "loss": 0.6692,
+      "step": 4025
+    },
+    {
+      "epoch": 1.413623595505618,
+      "grad_norm": 0.8751959204673767,
+      "learning_rate": 4.369530695581461e-06,
+      "loss": 0.6598,
+      "step": 4026
+    },
+    {
+      "epoch": 1.4139747191011236,
+      "grad_norm": 0.896353006362915,
+      "learning_rate": 4.369223721139705e-06,
+      "loss": 0.6605,
+      "step": 4027
+    },
+    {
+      "epoch": 1.4143258426966292,
+      "grad_norm": 0.8932797908782959,
+      "learning_rate": 4.368916682772097e-06,
+      "loss": 0.6395,
+      "step": 4028
+    },
+    {
+      "epoch": 1.4146769662921348,
+      "grad_norm": 0.8765178918838501,
+      "learning_rate": 4.368609580489137e-06,
+      "loss": 0.6298,
+      "step": 4029
+    },
+    {
+      "epoch": 1.4150280898876404,
+      "grad_norm": 0.872016429901123,
+      "learning_rate": 4.368302414301328e-06,
+      "loss": 0.6291,
+      "step": 4030
+    },
+    {
+      "epoch": 1.415379213483146,
+      "grad_norm": 0.878186821937561,
+      "learning_rate": 4.367995184219176e-06,
+      "loss": 0.6139,
+      "step": 4031
+    },
+    {
+      "epoch": 1.4157303370786516,
+      "grad_norm": 0.8804723024368286,
+      "learning_rate": 4.367687890253187e-06,
+      "loss": 0.6372,
+      "step": 4032
+    },
+    {
+      "epoch": 1.4160814606741572,
+      "grad_norm": 0.8700343370437622,
+      "learning_rate": 4.3673805324138705e-06,
+      "loss": 0.5968,
+      "step": 4033
+    },
+    {
+      "epoch": 1.4164325842696628,
+      "grad_norm": 0.8827090859413147,
+      "learning_rate": 4.367073110711738e-06,
+      "loss": 0.5944,
+      "step": 4034
+    },
+    {
+      "epoch": 1.4167837078651686,
+      "grad_norm": 0.8662370443344116,
+      "learning_rate": 4.366765625157302e-06,
+      "loss": 0.6339,
+      "step": 4035
+    },
+    {
+      "epoch": 1.4171348314606742,
+      "grad_norm": 0.8922637701034546,
+      "learning_rate": 4.3664580757610795e-06,
+      "loss": 0.6869,
+      "step": 4036
+    },
+    {
+      "epoch": 1.4174859550561798,
+      "grad_norm": 0.9283745288848877,
+      "learning_rate": 4.366150462533588e-06,
+      "loss": 0.6271,
+      "step": 4037
+    },
+    {
+      "epoch": 1.4178370786516854,
+      "grad_norm": 0.8803569674491882,
+      "learning_rate": 4.365842785485348e-06,
+      "loss": 0.6337,
+      "step": 4038
+    },
+    {
+      "epoch": 1.418188202247191,
+      "grad_norm": 0.8726440072059631,
+      "learning_rate": 4.36553504462688e-06,
+      "loss": 0.64,
+      "step": 4039
+    },
+    {
+      "epoch": 1.4185393258426966,
+      "grad_norm": 0.882938027381897,
+      "learning_rate": 4.3652272399687105e-06,
+      "loss": 0.6372,
+      "step": 4040
+    },
+    {
+      "epoch": 1.4188904494382022,
+      "grad_norm": 0.8737468123435974,
+      "learning_rate": 4.364919371521365e-06,
+      "loss": 0.6284,
+      "step": 4041
+    },
+    {
+      "epoch": 1.4192415730337078,
+      "grad_norm": 0.8580338954925537,
+      "learning_rate": 4.364611439295374e-06,
+      "loss": 0.5805,
+      "step": 4042
+    },
+    {
+      "epoch": 1.4195926966292136,
+      "grad_norm": 0.8792635202407837,
+      "learning_rate": 4.364303443301267e-06,
+      "loss": 0.5764,
+      "step": 4043
+    },
+    {
+      "epoch": 1.4199438202247192,
+      "grad_norm": 0.9049398303031921,
+      "learning_rate": 4.363995383549578e-06,
+      "loss": 0.6422,
+      "step": 4044
+    },
+    {
+      "epoch": 1.4202949438202248,
+      "grad_norm": 0.9274637699127197,
+      "learning_rate": 4.363687260050841e-06,
+      "loss": 0.6415,
+      "step": 4045
+    },
+    {
+      "epoch": 1.4206460674157304,
+      "grad_norm": 0.9082589149475098,
+      "learning_rate": 4.363379072815594e-06,
+      "loss": 0.6201,
+      "step": 4046
+    },
+    {
+      "epoch": 1.420997191011236,
+      "grad_norm": 0.866824209690094,
+      "learning_rate": 4.363070821854378e-06,
+      "loss": 0.5856,
+      "step": 4047
+    },
+    {
+      "epoch": 1.4213483146067416,
+      "grad_norm": 0.8817905187606812,
+      "learning_rate": 4.362762507177735e-06,
+      "loss": 0.5998,
+      "step": 4048
+    },
+    {
+      "epoch": 1.4216994382022472,
+      "grad_norm": 0.8592423796653748,
+      "learning_rate": 4.3624541287962065e-06,
+      "loss": 0.6326,
+      "step": 4049
+    },
+    {
+      "epoch": 1.4220505617977528,
+      "grad_norm": 0.8892386555671692,
+      "learning_rate": 4.3621456867203414e-06,
+      "loss": 0.5931,
+      "step": 4050
+    },
+    {
+      "epoch": 1.4224016853932584,
+      "grad_norm": 0.8761540055274963,
+      "learning_rate": 4.361837180960687e-06,
+      "loss": 0.6055,
+      "step": 4051
+    },
+    {
+      "epoch": 1.422752808988764,
+      "grad_norm": 0.8538140654563904,
+      "learning_rate": 4.361528611527795e-06,
+      "loss": 0.6098,
+      "step": 4052
+    },
+    {
+      "epoch": 1.4231039325842696,
+      "grad_norm": 0.9002669453620911,
+      "learning_rate": 4.3612199784322165e-06,
+      "loss": 0.6199,
+      "step": 4053
+    },
+    {
+      "epoch": 1.4234550561797752,
+      "grad_norm": 0.8686503767967224,
+      "learning_rate": 4.3609112816845074e-06,
+      "loss": 0.6301,
+      "step": 4054
+    },
+    {
+      "epoch": 1.4238061797752808,
+      "grad_norm": 0.8846096992492676,
+      "learning_rate": 4.360602521295226e-06,
+      "loss": 0.6258,
+      "step": 4055
+    },
+    {
+      "epoch": 1.4241573033707864,
+      "grad_norm": 0.9569324851036072,
+      "learning_rate": 4.36029369727493e-06,
+      "loss": 0.6516,
+      "step": 4056
+    },
+    {
+      "epoch": 1.4245084269662922,
+      "grad_norm": 0.8662309050559998,
+      "learning_rate": 4.35998480963418e-06,
+      "loss": 0.6247,
+      "step": 4057
+    },
+    {
+      "epoch": 1.4248595505617978,
+      "grad_norm": 0.8644338250160217,
+      "learning_rate": 4.359675858383543e-06,
+      "loss": 0.6197,
+      "step": 4058
+    },
+    {
+      "epoch": 1.4252106741573034,
+      "grad_norm": 0.8778071999549866,
+      "learning_rate": 4.359366843533583e-06,
+      "loss": 0.576,
+      "step": 4059
+    },
+    {
+      "epoch": 1.425561797752809,
+      "grad_norm": 0.8852523565292358,
+      "learning_rate": 4.359057765094866e-06,
+      "loss": 0.6389,
+      "step": 4060
+    },
+    {
+      "epoch": 1.4259129213483146,
+      "grad_norm": 0.9235852956771851,
+      "learning_rate": 4.358748623077966e-06,
+      "loss": 0.614,
+      "step": 4061
+    },
+    {
+      "epoch": 1.4262640449438202,
+      "grad_norm": 0.8874046206474304,
+      "learning_rate": 4.358439417493453e-06,
+      "loss": 0.6236,
+      "step": 4062
+    },
+    {
+      "epoch": 1.4266151685393258,
+      "grad_norm": 0.8729009628295898,
+      "learning_rate": 4.3581301483519036e-06,
+      "loss": 0.611,
+      "step": 4063
+    },
+    {
+      "epoch": 1.4269662921348314,
+      "grad_norm": 0.9033231735229492,
+      "learning_rate": 4.3578208156638915e-06,
+      "loss": 0.6216,
+      "step": 4064
+    },
+    {
+      "epoch": 1.4273174157303372,
+      "grad_norm": 0.8786650896072388,
+      "learning_rate": 4.3575114194399974e-06,
+      "loss": 0.6198,
+      "step": 4065
+    },
+    {
+      "epoch": 1.4276685393258428,
+      "grad_norm": 0.887387752532959,
+      "learning_rate": 4.357201959690802e-06,
+      "loss": 0.6346,
+      "step": 4066
+    },
+    {
+      "epoch": 1.4280196629213484,
+      "grad_norm": 0.8734090924263,
+      "learning_rate": 4.356892436426891e-06,
+      "loss": 0.6107,
+      "step": 4067
+    },
+    {
+      "epoch": 1.428370786516854,
+      "grad_norm": 0.8874668478965759,
+      "learning_rate": 4.356582849658845e-06,
+      "loss": 0.6454,
+      "step": 4068
+    },
+    {
+      "epoch": 1.4287219101123596,
+      "grad_norm": 0.8698024749755859,
+      "learning_rate": 4.356273199397256e-06,
+      "loss": 0.6529,
+      "step": 4069
+    },
+    {
+      "epoch": 1.4290730337078652,
+      "grad_norm": 0.8861244916915894,
+      "learning_rate": 4.355963485652711e-06,
+      "loss": 0.594,
+      "step": 4070
+    },
+    {
+      "epoch": 1.4294241573033708,
+      "grad_norm": 0.8716466426849365,
+      "learning_rate": 4.355653708435803e-06,
+      "loss": 0.6367,
+      "step": 4071
+    },
+    {
+      "epoch": 1.4297752808988764,
+      "grad_norm": 0.8842322826385498,
+      "learning_rate": 4.355343867757127e-06,
+      "loss": 0.6313,
+      "step": 4072
+    },
+    {
+      "epoch": 1.430126404494382,
+      "grad_norm": 0.8881300687789917,
+      "learning_rate": 4.355033963627277e-06,
+      "loss": 0.624,
+      "step": 4073
+    },
+    {
+      "epoch": 1.4304775280898876,
+      "grad_norm": 0.8615336418151855,
+      "learning_rate": 4.354723996056855e-06,
+      "loss": 0.6307,
+      "step": 4074
+    },
+    {
+      "epoch": 1.4308286516853932,
+      "grad_norm": 0.8932889699935913,
+      "learning_rate": 4.354413965056457e-06,
+      "loss": 0.6211,
+      "step": 4075
+    },
+    {
+      "epoch": 1.4311797752808988,
+      "grad_norm": 0.8843944668769836,
+      "learning_rate": 4.354103870636689e-06,
+      "loss": 0.6482,
+      "step": 4076
+    },
+    {
+      "epoch": 1.4315308988764044,
+      "grad_norm": 0.8917078375816345,
+      "learning_rate": 4.353793712808156e-06,
+      "loss": 0.6044,
+      "step": 4077
+    },
+    {
+      "epoch": 1.43188202247191,
+      "grad_norm": 0.8852561712265015,
+      "learning_rate": 4.353483491581463e-06,
+      "loss": 0.6213,
+      "step": 4078
+    },
+    {
+      "epoch": 1.4322331460674158,
+      "grad_norm": 0.8828893899917603,
+      "learning_rate": 4.3531732069672215e-06,
+      "loss": 0.6676,
+      "step": 4079
+    },
+    {
+      "epoch": 1.4325842696629214,
+      "grad_norm": 0.8896101713180542,
+      "learning_rate": 4.352862858976043e-06,
+      "loss": 0.6277,
+      "step": 4080
+    },
+    {
+      "epoch": 1.432935393258427,
+      "grad_norm": 0.8827319145202637,
+      "learning_rate": 4.352552447618539e-06,
+      "loss": 0.6316,
+      "step": 4081
+    },
+    {
+      "epoch": 1.4332865168539326,
+      "grad_norm": 0.9116552472114563,
+      "learning_rate": 4.352241972905326e-06,
+      "loss": 0.6307,
+      "step": 4082
+    },
+    {
+      "epoch": 1.4336376404494382,
+      "grad_norm": 0.880622386932373,
+      "learning_rate": 4.351931434847023e-06,
+      "loss": 0.6102,
+      "step": 4083
+    },
+    {
+      "epoch": 1.4339887640449438,
+      "grad_norm": 0.8774731159210205,
+      "learning_rate": 4.351620833454251e-06,
+      "loss": 0.6243,
+      "step": 4084
+    },
+    {
+      "epoch": 1.4343398876404494,
+      "grad_norm": 0.8572145104408264,
+      "learning_rate": 4.351310168737629e-06,
+      "loss": 0.6149,
+      "step": 4085
+    },
+    {
+      "epoch": 1.434691011235955,
+      "grad_norm": 0.8389350175857544,
+      "learning_rate": 4.350999440707785e-06,
+      "loss": 0.597,
+      "step": 4086
+    },
+    {
+      "epoch": 1.4350421348314606,
+      "grad_norm": 0.8829629421234131,
+      "learning_rate": 4.350688649375343e-06,
+      "loss": 0.6257,
+      "step": 4087
+    },
+    {
+      "epoch": 1.4353932584269664,
+      "grad_norm": 0.901107132434845,
+      "learning_rate": 4.350377794750933e-06,
+      "loss": 0.6468,
+      "step": 4088
+    },
+    {
+      "epoch": 1.435744382022472,
+      "grad_norm": 0.8687376379966736,
+      "learning_rate": 4.350066876845186e-06,
+      "loss": 0.6274,
+      "step": 4089
+    },
+    {
+      "epoch": 1.4360955056179776,
+      "grad_norm": 0.8897804617881775,
+      "learning_rate": 4.349755895668735e-06,
+      "loss": 0.6661,
+      "step": 4090
+    },
+    {
+      "epoch": 1.4364466292134832,
+      "grad_norm": 0.8794514536857605,
+      "learning_rate": 4.349444851232215e-06,
+      "loss": 0.6543,
+      "step": 4091
+    },
+    {
+      "epoch": 1.4367977528089888,
+      "grad_norm": 0.8795956969261169,
+      "learning_rate": 4.349133743546265e-06,
+      "loss": 0.6164,
+      "step": 4092
+    },
+    {
+      "epoch": 1.4371488764044944,
+      "grad_norm": 0.8774768114089966,
+      "learning_rate": 4.348822572621521e-06,
+      "loss": 0.6115,
+      "step": 4093
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 0.8642878532409668,
+      "learning_rate": 4.348511338468629e-06,
+      "loss": 0.6548,
+      "step": 4094
+    },
+    {
+      "epoch": 1.4378511235955056,
+      "grad_norm": 0.9389157295227051,
+      "learning_rate": 4.34820004109823e-06,
+      "loss": 0.6053,
+      "step": 4095
+    },
+    {
+      "epoch": 1.4382022471910112,
+      "grad_norm": 0.8678531050682068,
+      "learning_rate": 4.347888680520971e-06,
+      "loss": 0.6254,
+      "step": 4096
+    },
+    {
+      "epoch": 1.4385533707865168,
+      "grad_norm": 0.8993881940841675,
+      "learning_rate": 4.347577256747501e-06,
+      "loss": 0.6723,
+      "step": 4097
+    },
+    {
+      "epoch": 1.4389044943820224,
+      "grad_norm": 0.9129367470741272,
+      "learning_rate": 4.3472657697884705e-06,
+      "loss": 0.6197,
+      "step": 4098
+    },
+    {
+      "epoch": 1.439255617977528,
+      "grad_norm": 0.8977501392364502,
+      "learning_rate": 4.34695421965453e-06,
+      "loss": 0.6027,
+      "step": 4099
+    },
+    {
+      "epoch": 1.4396067415730336,
+      "grad_norm": 0.887247622013092,
+      "learning_rate": 4.346642606356336e-06,
+      "loss": 0.5911,
+      "step": 4100
+    },
+    {
+      "epoch": 1.4399578651685394,
+      "grad_norm": 0.8997041583061218,
+      "learning_rate": 4.346330929904546e-06,
+      "loss": 0.6262,
+      "step": 4101
+    },
+    {
+      "epoch": 1.440308988764045,
+      "grad_norm": 0.885622501373291,
+      "learning_rate": 4.346019190309817e-06,
+      "loss": 0.6288,
+      "step": 4102
+    },
+    {
+      "epoch": 1.4406601123595506,
+      "grad_norm": 0.8826760649681091,
+      "learning_rate": 4.345707387582811e-06,
+      "loss": 0.6669,
+      "step": 4103
+    },
+    {
+      "epoch": 1.4410112359550562,
+      "grad_norm": 0.883783221244812,
+      "learning_rate": 4.345395521734192e-06,
+      "loss": 0.6235,
+      "step": 4104
+    },
+    {
+      "epoch": 1.4413623595505618,
+      "grad_norm": 0.8693239092826843,
+      "learning_rate": 4.345083592774626e-06,
+      "loss": 0.6066,
+      "step": 4105
+    },
+    {
+      "epoch": 1.4417134831460674,
+      "grad_norm": 0.8620202541351318,
+      "learning_rate": 4.34477160071478e-06,
+      "loss": 0.6027,
+      "step": 4106
+    },
+    {
+      "epoch": 1.442064606741573,
+      "grad_norm": 0.8705936074256897,
+      "learning_rate": 4.344459545565324e-06,
+      "loss": 0.6173,
+      "step": 4107
+    },
+    {
+      "epoch": 1.4424157303370786,
+      "grad_norm": 0.8645413517951965,
+      "learning_rate": 4.344147427336929e-06,
+      "loss": 0.6598,
+      "step": 4108
+    },
+    {
+      "epoch": 1.4427668539325842,
+      "grad_norm": 0.8727925419807434,
+      "learning_rate": 4.34383524604027e-06,
+      "loss": 0.6537,
+      "step": 4109
+    },
+    {
+      "epoch": 1.44311797752809,
+      "grad_norm": 0.8704939484596252,
+      "learning_rate": 4.343523001686024e-06,
+      "loss": 0.6272,
+      "step": 4110
+    },
+    {
+      "epoch": 1.4434691011235956,
+      "grad_norm": 0.8718854188919067,
+      "learning_rate": 4.343210694284869e-06,
+      "loss": 0.6491,
+      "step": 4111
+    },
+    {
+      "epoch": 1.4438202247191012,
+      "grad_norm": 0.8282216787338257,
+      "learning_rate": 4.342898323847486e-06,
+      "loss": 0.559,
+      "step": 4112
+    },
+    {
+      "epoch": 1.4441713483146068,
+      "grad_norm": 0.8932397961616516,
+      "learning_rate": 4.3425858903845555e-06,
+      "loss": 0.6539,
+      "step": 4113
+    },
+    {
+      "epoch": 1.4445224719101124,
+      "grad_norm": 0.8598067164421082,
+      "learning_rate": 4.342273393906765e-06,
+      "loss": 0.585,
+      "step": 4114
+    },
+    {
+      "epoch": 1.444873595505618,
+      "grad_norm": 0.9065674543380737,
+      "learning_rate": 4.3419608344248015e-06,
+      "loss": 0.6237,
+      "step": 4115
+    },
+    {
+      "epoch": 1.4452247191011236,
+      "grad_norm": 0.8612712621688843,
+      "learning_rate": 4.341648211949353e-06,
+      "loss": 0.5795,
+      "step": 4116
+    },
+    {
+      "epoch": 1.4455758426966292,
+      "grad_norm": 0.9073783755302429,
+      "learning_rate": 4.341335526491112e-06,
+      "loss": 0.6542,
+      "step": 4117
+    },
+    {
+      "epoch": 1.4459269662921348,
+      "grad_norm": 0.8671945333480835,
+      "learning_rate": 4.341022778060772e-06,
+      "loss": 0.6384,
+      "step": 4118
+    },
+    {
+      "epoch": 1.4462780898876404,
+      "grad_norm": 0.8586475849151611,
+      "learning_rate": 4.340709966669028e-06,
+      "loss": 0.6043,
+      "step": 4119
+    },
+    {
+      "epoch": 1.446629213483146,
+      "grad_norm": 0.8946781754493713,
+      "learning_rate": 4.340397092326578e-06,
+      "loss": 0.635,
+      "step": 4120
+    },
+    {
+      "epoch": 1.4469803370786516,
+      "grad_norm": 0.8610424995422363,
+      "learning_rate": 4.340084155044122e-06,
+      "loss": 0.6365,
+      "step": 4121
+    },
+    {
+      "epoch": 1.4473314606741572,
+      "grad_norm": 0.8891358971595764,
+      "learning_rate": 4.339771154832364e-06,
+      "loss": 0.6054,
+      "step": 4122
+    },
+    {
+      "epoch": 1.4476825842696628,
+      "grad_norm": 0.8908564448356628,
+      "learning_rate": 4.339458091702005e-06,
+      "loss": 0.6168,
+      "step": 4123
+    },
+    {
+      "epoch": 1.4480337078651686,
+      "grad_norm": 0.8783697485923767,
+      "learning_rate": 4.339144965663754e-06,
+      "loss": 0.6002,
+      "step": 4124
+    },
+    {
+      "epoch": 1.4483848314606742,
+      "grad_norm": 0.9083595871925354,
+      "learning_rate": 4.3388317767283195e-06,
+      "loss": 0.6357,
+      "step": 4125
+    },
+    {
+      "epoch": 1.4487359550561798,
+      "grad_norm": 0.8990292549133301,
+      "learning_rate": 4.338518524906411e-06,
+      "loss": 0.6422,
+      "step": 4126
+    },
+    {
+      "epoch": 1.4490870786516854,
+      "grad_norm": 0.9985833168029785,
+      "learning_rate": 4.338205210208742e-06,
+      "loss": 0.5973,
+      "step": 4127
+    },
+    {
+      "epoch": 1.449438202247191,
+      "grad_norm": 0.8731505870819092,
+      "learning_rate": 4.337891832646028e-06,
+      "loss": 0.6414,
+      "step": 4128
+    },
+    {
+      "epoch": 1.4497893258426966,
+      "grad_norm": 0.9245936274528503,
+      "learning_rate": 4.337578392228987e-06,
+      "loss": 0.6543,
+      "step": 4129
+    },
+    {
+      "epoch": 1.4501404494382022,
+      "grad_norm": 0.9523738622665405,
+      "learning_rate": 4.337264888968337e-06,
+      "loss": 0.6389,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4504915730337078,
+      "grad_norm": 0.9052737355232239,
+      "learning_rate": 4.336951322874799e-06,
+      "loss": 0.6343,
+      "step": 4131
+    },
+    {
+      "epoch": 1.4508426966292136,
+      "grad_norm": 0.9124898910522461,
+      "learning_rate": 4.336637693959099e-06,
+      "loss": 0.6346,
+      "step": 4132
+    },
+    {
+      "epoch": 1.4511938202247192,
+      "grad_norm": 0.8696838617324829,
+      "learning_rate": 4.336324002231961e-06,
+      "loss": 0.6181,
+      "step": 4133
+    },
+    {
+      "epoch": 1.4515449438202248,
+      "grad_norm": 0.8571642637252808,
+      "learning_rate": 4.336010247704113e-06,
+      "loss": 0.6331,
+      "step": 4134
+    },
+    {
+      "epoch": 1.4518960674157304,
+      "grad_norm": 0.88227379322052,
+      "learning_rate": 4.335696430386286e-06,
+      "loss": 0.585,
+      "step": 4135
+    },
+    {
+      "epoch": 1.452247191011236,
+      "grad_norm": 0.8858445286750793,
+      "learning_rate": 4.335382550289212e-06,
+      "loss": 0.6268,
+      "step": 4136
+    },
+    {
+      "epoch": 1.4525983146067416,
+      "grad_norm": 0.877275288105011,
+      "learning_rate": 4.335068607423624e-06,
+      "loss": 0.6317,
+      "step": 4137
+    },
+    {
+      "epoch": 1.4529494382022472,
+      "grad_norm": 0.8988421559333801,
+      "learning_rate": 4.334754601800263e-06,
+      "loss": 0.6305,
+      "step": 4138
+    },
+    {
+      "epoch": 1.4533005617977528,
+      "grad_norm": 0.9341934323310852,
+      "learning_rate": 4.334440533429862e-06,
+      "loss": 0.614,
+      "step": 4139
+    },
+    {
+      "epoch": 1.4536516853932584,
+      "grad_norm": 0.8875921964645386,
+      "learning_rate": 4.3341264023231655e-06,
+      "loss": 0.6094,
+      "step": 4140
+    },
+    {
+      "epoch": 1.454002808988764,
+      "grad_norm": 0.8801622986793518,
+      "learning_rate": 4.333812208490915e-06,
+      "loss": 0.5936,
+      "step": 4141
+    },
+    {
+      "epoch": 1.4543539325842696,
+      "grad_norm": 0.9054430723190308,
+      "learning_rate": 4.333497951943856e-06,
+      "loss": 0.6518,
+      "step": 4142
+    },
+    {
+      "epoch": 1.4547050561797752,
+      "grad_norm": 0.8847231268882751,
+      "learning_rate": 4.333183632692737e-06,
+      "loss": 0.644,
+      "step": 4143
+    },
+    {
+      "epoch": 1.4550561797752808,
+      "grad_norm": 0.8920429348945618,
+      "learning_rate": 4.332869250748306e-06,
+      "loss": 0.6457,
+      "step": 4144
+    },
+    {
+      "epoch": 1.4554073033707864,
+      "grad_norm": 0.9018110632896423,
+      "learning_rate": 4.332554806121315e-06,
+      "loss": 0.6295,
+      "step": 4145
+    },
+    {
+      "epoch": 1.4557584269662922,
+      "grad_norm": 0.8690390586853027,
+      "learning_rate": 4.332240298822517e-06,
+      "loss": 0.6273,
+      "step": 4146
+    },
+    {
+      "epoch": 1.4561095505617978,
+      "grad_norm": 0.861956000328064,
+      "learning_rate": 4.33192572886267e-06,
+      "loss": 0.6252,
+      "step": 4147
+    },
+    {
+      "epoch": 1.4564606741573034,
+      "grad_norm": 0.8745825886726379,
+      "learning_rate": 4.33161109625253e-06,
+      "loss": 0.6328,
+      "step": 4148
+    },
+    {
+      "epoch": 1.456811797752809,
+      "grad_norm": 0.8459515571594238,
+      "learning_rate": 4.331296401002858e-06,
+      "loss": 0.5466,
+      "step": 4149
+    },
+    {
+      "epoch": 1.4571629213483146,
+      "grad_norm": 0.8832082152366638,
+      "learning_rate": 4.330981643124416e-06,
+      "loss": 0.6516,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4575140449438202,
+      "grad_norm": 0.8869217038154602,
+      "learning_rate": 4.330666822627968e-06,
+      "loss": 0.5984,
+      "step": 4151
+    },
+    {
+      "epoch": 1.4578651685393258,
+      "grad_norm": 0.8645246624946594,
+      "learning_rate": 4.330351939524282e-06,
+      "loss": 0.5963,
+      "step": 4152
+    },
+    {
+      "epoch": 1.4582162921348314,
+      "grad_norm": 0.8830666542053223,
+      "learning_rate": 4.330036993824126e-06,
+      "loss": 0.6349,
+      "step": 4153
+    },
+    {
+      "epoch": 1.4585674157303372,
+      "grad_norm": 0.9177601933479309,
+      "learning_rate": 4.329721985538271e-06,
+      "loss": 0.6322,
+      "step": 4154
+    },
+    {
+      "epoch": 1.4589185393258428,
+      "grad_norm": 0.9256566762924194,
+      "learning_rate": 4.329406914677489e-06,
+      "loss": 0.6912,
+      "step": 4155
+    },
+    {
+      "epoch": 1.4592696629213484,
+      "grad_norm": 0.879478931427002,
+      "learning_rate": 4.329091781252556e-06,
+      "loss": 0.6133,
+      "step": 4156
+    },
+    {
+      "epoch": 1.459620786516854,
+      "grad_norm": 0.8938345909118652,
+      "learning_rate": 4.32877658527425e-06,
+      "loss": 0.5806,
+      "step": 4157
+    },
+    {
+      "epoch": 1.4599719101123596,
+      "grad_norm": 0.8846132159233093,
+      "learning_rate": 4.328461326753349e-06,
+      "loss": 0.6136,
+      "step": 4158
+    },
+    {
+      "epoch": 1.4603230337078652,
+      "grad_norm": 0.8530312180519104,
+      "learning_rate": 4.328146005700636e-06,
+      "loss": 0.6212,
+      "step": 4159
+    },
+    {
+      "epoch": 1.4606741573033708,
+      "grad_norm": 0.8560202121734619,
+      "learning_rate": 4.327830622126893e-06,
+      "loss": 0.6011,
+      "step": 4160
+    },
+    {
+      "epoch": 1.4610252808988764,
+      "grad_norm": 0.8946338295936584,
+      "learning_rate": 4.327515176042908e-06,
+      "loss": 0.5978,
+      "step": 4161
+    },
+    {
+      "epoch": 1.461376404494382,
+      "grad_norm": 0.89649498462677,
+      "learning_rate": 4.327199667459467e-06,
+      "loss": 0.6511,
+      "step": 4162
+    },
+    {
+      "epoch": 1.4617275280898876,
+      "grad_norm": 0.8728743195533752,
+      "learning_rate": 4.326884096387361e-06,
+      "loss": 0.614,
+      "step": 4163
+    },
+    {
+      "epoch": 1.4620786516853932,
+      "grad_norm": 0.8919811844825745,
+      "learning_rate": 4.326568462837381e-06,
+      "loss": 0.6341,
+      "step": 4164
+    },
+    {
+      "epoch": 1.4624297752808988,
+      "grad_norm": 0.8965336680412292,
+      "learning_rate": 4.326252766820325e-06,
+      "loss": 0.5954,
+      "step": 4165
+    },
+    {
+      "epoch": 1.4627808988764044,
+      "grad_norm": 0.919897198677063,
+      "learning_rate": 4.325937008346985e-06,
+      "loss": 0.6758,
+      "step": 4166
+    },
+    {
+      "epoch": 1.46313202247191,
+      "grad_norm": 0.8864952921867371,
+      "learning_rate": 4.325621187428164e-06,
+      "loss": 0.6116,
+      "step": 4167
+    },
+    {
+      "epoch": 1.4634831460674158,
+      "grad_norm": 0.895219624042511,
+      "learning_rate": 4.325305304074659e-06,
+      "loss": 0.6196,
+      "step": 4168
+    },
+    {
+      "epoch": 1.4638342696629214,
+      "grad_norm": 0.8681285381317139,
+      "learning_rate": 4.3249893582972745e-06,
+      "loss": 0.6074,
+      "step": 4169
+    },
+    {
+      "epoch": 1.464185393258427,
+      "grad_norm": 0.8704788088798523,
+      "learning_rate": 4.324673350106816e-06,
+      "loss": 0.6145,
+      "step": 4170
+    },
+    {
+      "epoch": 1.4645365168539326,
+      "grad_norm": 0.8976466655731201,
+      "learning_rate": 4.324357279514091e-06,
+      "loss": 0.6499,
+      "step": 4171
+    },
+    {
+      "epoch": 1.4648876404494382,
+      "grad_norm": 0.8723233938217163,
+      "learning_rate": 4.324041146529907e-06,
+      "loss": 0.6227,
+      "step": 4172
+    },
+    {
+      "epoch": 1.4652387640449438,
+      "grad_norm": 0.9027929902076721,
+      "learning_rate": 4.323724951165077e-06,
+      "loss": 0.6279,
+      "step": 4173
+    },
+    {
+      "epoch": 1.4655898876404494,
+      "grad_norm": 0.859115481376648,
+      "learning_rate": 4.323408693430414e-06,
+      "loss": 0.6295,
+      "step": 4174
+    },
+    {
+      "epoch": 1.465941011235955,
+      "grad_norm": 1.1383740901947021,
+      "learning_rate": 4.323092373336735e-06,
+      "loss": 0.6494,
+      "step": 4175
+    },
+    {
+      "epoch": 1.4662921348314606,
+      "grad_norm": 0.8614624738693237,
+      "learning_rate": 4.322775990894856e-06,
+      "loss": 0.5949,
+      "step": 4176
+    },
+    {
+      "epoch": 1.4666432584269664,
+      "grad_norm": 0.8514102697372437,
+      "learning_rate": 4.322459546115598e-06,
+      "loss": 0.5966,
+      "step": 4177
+    },
+    {
+      "epoch": 1.466994382022472,
+      "grad_norm": 0.899704098701477,
+      "learning_rate": 4.322143039009783e-06,
+      "loss": 0.6548,
+      "step": 4178
+    },
+    {
+      "epoch": 1.4673455056179776,
+      "grad_norm": 0.8559503555297852,
+      "learning_rate": 4.321826469588236e-06,
+      "loss": 0.6039,
+      "step": 4179
+    },
+    {
+      "epoch": 1.4676966292134832,
+      "grad_norm": 0.8766810297966003,
+      "learning_rate": 4.3215098378617815e-06,
+      "loss": 0.588,
+      "step": 4180
+    },
+    {
+      "epoch": 1.4680477528089888,
+      "grad_norm": 0.8811195492744446,
+      "learning_rate": 4.32119314384125e-06,
+      "loss": 0.6208,
+      "step": 4181
+    },
+    {
+      "epoch": 1.4683988764044944,
+      "grad_norm": 0.9778435826301575,
+      "learning_rate": 4.320876387537472e-06,
+      "loss": 0.6445,
+      "step": 4182
+    },
+    {
+      "epoch": 1.46875,
+      "grad_norm": 0.905834436416626,
+      "learning_rate": 4.320559568961279e-06,
+      "loss": 0.6574,
+      "step": 4183
+    },
+    {
+      "epoch": 1.4691011235955056,
+      "grad_norm": 0.8865072131156921,
+      "learning_rate": 4.320242688123506e-06,
+      "loss": 0.5757,
+      "step": 4184
+    },
+    {
+      "epoch": 1.4694522471910112,
+      "grad_norm": 0.8838146328926086,
+      "learning_rate": 4.319925745034991e-06,
+      "loss": 0.634,
+      "step": 4185
+    },
+    {
+      "epoch": 1.4698033707865168,
+      "grad_norm": 0.8725399971008301,
+      "learning_rate": 4.3196087397065734e-06,
+      "loss": 0.6075,
+      "step": 4186
+    },
+    {
+      "epoch": 1.4701544943820224,
+      "grad_norm": 0.8557268977165222,
+      "learning_rate": 4.3192916721490945e-06,
+      "loss": 0.6138,
+      "step": 4187
+    },
+    {
+      "epoch": 1.470505617977528,
+      "grad_norm": 0.8887699246406555,
+      "learning_rate": 4.318974542373396e-06,
+      "loss": 0.6389,
+      "step": 4188
+    },
+    {
+      "epoch": 1.4708567415730336,
+      "grad_norm": 0.8605521321296692,
+      "learning_rate": 4.318657350390324e-06,
+      "loss": 0.6261,
+      "step": 4189
+    },
+    {
+      "epoch": 1.4712078651685394,
+      "grad_norm": 0.8946020007133484,
+      "learning_rate": 4.318340096210729e-06,
+      "loss": 0.6294,
+      "step": 4190
+    },
+    {
+      "epoch": 1.471558988764045,
+      "grad_norm": 0.8895167708396912,
+      "learning_rate": 4.318022779845457e-06,
+      "loss": 0.629,
+      "step": 4191
+    },
+    {
+      "epoch": 1.4719101123595506,
+      "grad_norm": 0.8478034138679504,
+      "learning_rate": 4.317705401305362e-06,
+      "loss": 0.6042,
+      "step": 4192
+    },
+    {
+      "epoch": 1.4722612359550562,
+      "grad_norm": 0.8884312510490417,
+      "learning_rate": 4.317387960601298e-06,
+      "loss": 0.6217,
+      "step": 4193
+    },
+    {
+      "epoch": 1.4726123595505618,
+      "grad_norm": 0.8619526624679565,
+      "learning_rate": 4.317070457744121e-06,
+      "loss": 0.5889,
+      "step": 4194
+    },
+    {
+      "epoch": 1.4729634831460674,
+      "grad_norm": 0.8887055516242981,
+      "learning_rate": 4.316752892744689e-06,
+      "loss": 0.6346,
+      "step": 4195
+    },
+    {
+      "epoch": 1.473314606741573,
+      "grad_norm": 0.8569642901420593,
+      "learning_rate": 4.316435265613863e-06,
+      "loss": 0.622,
+      "step": 4196
+    },
+    {
+      "epoch": 1.4736657303370786,
+      "grad_norm": 0.8919498920440674,
+      "learning_rate": 4.316117576362505e-06,
+      "loss": 0.6371,
+      "step": 4197
+    },
+    {
+      "epoch": 1.4740168539325842,
+      "grad_norm": 0.8512890338897705,
+      "learning_rate": 4.31579982500148e-06,
+      "loss": 0.6495,
+      "step": 4198
+    },
+    {
+      "epoch": 1.47436797752809,
+      "grad_norm": 0.8871182203292847,
+      "learning_rate": 4.315482011541655e-06,
+      "loss": 0.6166,
+      "step": 4199
+    },
+    {
+      "epoch": 1.4747191011235956,
+      "grad_norm": 0.8572426438331604,
+      "learning_rate": 4.315164135993898e-06,
+      "loss": 0.6101,
+      "step": 4200
+    },
+    {
+      "epoch": 1.4750702247191012,
+      "grad_norm": 0.8520934581756592,
+      "learning_rate": 4.314846198369082e-06,
+      "loss": 0.6039,
+      "step": 4201
+    },
+    {
+      "epoch": 1.4754213483146068,
+      "grad_norm": 0.9315975308418274,
+      "learning_rate": 4.314528198678079e-06,
+      "loss": 0.6408,
+      "step": 4202
+    },
+    {
+      "epoch": 1.4757724719101124,
+      "grad_norm": 0.8353304862976074,
+      "learning_rate": 4.314210136931762e-06,
+      "loss": 0.5972,
+      "step": 4203
+    },
+    {
+      "epoch": 1.476123595505618,
+      "grad_norm": 0.8691760897636414,
+      "learning_rate": 4.313892013141013e-06,
+      "loss": 0.5916,
+      "step": 4204
+    },
+    {
+      "epoch": 1.4764747191011236,
+      "grad_norm": 0.8935512900352478,
+      "learning_rate": 4.313573827316709e-06,
+      "loss": 0.6402,
+      "step": 4205
+    },
+    {
+      "epoch": 1.4768258426966292,
+      "grad_norm": 0.8638583421707153,
+      "learning_rate": 4.313255579469732e-06,
+      "loss": 0.621,
+      "step": 4206
+    },
+    {
+      "epoch": 1.4771769662921348,
+      "grad_norm": 0.9053982496261597,
+      "learning_rate": 4.312937269610965e-06,
+      "loss": 0.652,
+      "step": 4207
+    },
+    {
+      "epoch": 1.4775280898876404,
+      "grad_norm": 0.8318531513214111,
+      "learning_rate": 4.312618897751295e-06,
+      "loss": 0.5976,
+      "step": 4208
+    },
+    {
+      "epoch": 1.477879213483146,
+      "grad_norm": 0.8552941083908081,
+      "learning_rate": 4.3123004639016105e-06,
+      "loss": 0.5748,
+      "step": 4209
+    },
+    {
+      "epoch": 1.4782303370786516,
+      "grad_norm": 0.8920894265174866,
+      "learning_rate": 4.3119819680728e-06,
+      "loss": 0.6487,
+      "step": 4210
+    },
+    {
+      "epoch": 1.4785814606741572,
+      "grad_norm": 0.9560785293579102,
+      "learning_rate": 4.311663410275757e-06,
+      "loss": 0.6285,
+      "step": 4211
+    },
+    {
+      "epoch": 1.4789325842696628,
+      "grad_norm": 1.0505443811416626,
+      "learning_rate": 4.3113447905213765e-06,
+      "loss": 0.5522,
+      "step": 4212
+    },
+    {
+      "epoch": 1.4792837078651686,
+      "grad_norm": 0.8848596811294556,
+      "learning_rate": 4.311026108820553e-06,
+      "loss": 0.6029,
+      "step": 4213
+    },
+    {
+      "epoch": 1.4796348314606742,
+      "grad_norm": 0.8782453536987305,
+      "learning_rate": 4.310707365184187e-06,
+      "loss": 0.6234,
+      "step": 4214
+    },
+    {
+      "epoch": 1.4799859550561798,
+      "grad_norm": 0.8838267922401428,
+      "learning_rate": 4.310388559623179e-06,
+      "loss": 0.6206,
+      "step": 4215
+    },
+    {
+      "epoch": 1.4803370786516854,
+      "grad_norm": 0.9158840179443359,
+      "learning_rate": 4.31006969214843e-06,
+      "loss": 0.6414,
+      "step": 4216
+    },
+    {
+      "epoch": 1.480688202247191,
+      "grad_norm": 0.861907422542572,
+      "learning_rate": 4.309750762770847e-06,
+      "loss": 0.6319,
+      "step": 4217
+    },
+    {
+      "epoch": 1.4810393258426966,
+      "grad_norm": 0.8838783502578735,
+      "learning_rate": 4.3094317715013376e-06,
+      "loss": 0.5957,
+      "step": 4218
+    },
+    {
+      "epoch": 1.4813904494382022,
+      "grad_norm": 0.8620510101318359,
+      "learning_rate": 4.309112718350809e-06,
+      "loss": 0.6096,
+      "step": 4219
+    },
+    {
+      "epoch": 1.4817415730337078,
+      "grad_norm": 0.8699179887771606,
+      "learning_rate": 4.308793603330173e-06,
+      "loss": 0.6448,
+      "step": 4220
+    },
+    {
+      "epoch": 1.4820926966292136,
+      "grad_norm": 0.8801244497299194,
+      "learning_rate": 4.308474426450344e-06,
+      "loss": 0.6395,
+      "step": 4221
+    },
+    {
+      "epoch": 1.4824438202247192,
+      "grad_norm": 0.8840081691741943,
+      "learning_rate": 4.308155187722238e-06,
+      "loss": 0.6654,
+      "step": 4222
+    },
+    {
+      "epoch": 1.4827949438202248,
+      "grad_norm": 0.8977017402648926,
+      "learning_rate": 4.307835887156771e-06,
+      "loss": 0.6508,
+      "step": 4223
+    },
+    {
+      "epoch": 1.4831460674157304,
+      "grad_norm": 0.878200113773346,
+      "learning_rate": 4.307516524764864e-06,
+      "loss": 0.5663,
+      "step": 4224
+    },
+    {
+      "epoch": 1.483497191011236,
+      "grad_norm": 0.8168246150016785,
+      "learning_rate": 4.307197100557438e-06,
+      "loss": 0.6026,
+      "step": 4225
+    },
+    {
+      "epoch": 1.4838483146067416,
+      "grad_norm": 0.8974123597145081,
+      "learning_rate": 4.306877614545419e-06,
+      "loss": 0.6149,
+      "step": 4226
+    },
+    {
+      "epoch": 1.4841994382022472,
+      "grad_norm": 0.8998949527740479,
+      "learning_rate": 4.30655806673973e-06,
+      "loss": 0.6391,
+      "step": 4227
+    },
+    {
+      "epoch": 1.4845505617977528,
+      "grad_norm": 0.9134969115257263,
+      "learning_rate": 4.306238457151302e-06,
+      "loss": 0.6334,
+      "step": 4228
+    },
+    {
+      "epoch": 1.4849016853932584,
+      "grad_norm": 0.8966532349586487,
+      "learning_rate": 4.305918785791064e-06,
+      "loss": 0.6537,
+      "step": 4229
+    },
+    {
+      "epoch": 1.485252808988764,
+      "grad_norm": 0.900520384311676,
+      "learning_rate": 4.305599052669948e-06,
+      "loss": 0.6194,
+      "step": 4230
+    },
+    {
+      "epoch": 1.4856039325842696,
+      "grad_norm": 0.8717507719993591,
+      "learning_rate": 4.3052792577988915e-06,
+      "loss": 0.6231,
+      "step": 4231
+    },
+    {
+      "epoch": 1.4859550561797752,
+      "grad_norm": 0.9033085107803345,
+      "learning_rate": 4.304959401188828e-06,
+      "loss": 0.6555,
+      "step": 4232
+    },
+    {
+      "epoch": 1.4863061797752808,
+      "grad_norm": 0.853325605392456,
+      "learning_rate": 4.304639482850697e-06,
+      "loss": 0.6104,
+      "step": 4233
+    },
+    {
+      "epoch": 1.4866573033707864,
+      "grad_norm": 0.8766210675239563,
+      "learning_rate": 4.304319502795441e-06,
+      "loss": 0.6367,
+      "step": 4234
+    },
+    {
+      "epoch": 1.4870084269662922,
+      "grad_norm": 0.8499912619590759,
+      "learning_rate": 4.303999461034002e-06,
+      "loss": 0.6074,
+      "step": 4235
+    },
+    {
+      "epoch": 1.4873595505617978,
+      "grad_norm": 0.8618026375770569,
+      "learning_rate": 4.303679357577325e-06,
+      "loss": 0.6203,
+      "step": 4236
+    },
+    {
+      "epoch": 1.4877106741573034,
+      "grad_norm": 0.9145092964172363,
+      "learning_rate": 4.303359192436357e-06,
+      "loss": 0.6205,
+      "step": 4237
+    },
+    {
+      "epoch": 1.488061797752809,
+      "grad_norm": 0.8490617275238037,
+      "learning_rate": 4.303038965622048e-06,
+      "loss": 0.5845,
+      "step": 4238
+    },
+    {
+      "epoch": 1.4884129213483146,
+      "grad_norm": 0.8706423044204712,
+      "learning_rate": 4.3027186771453485e-06,
+      "loss": 0.6171,
+      "step": 4239
+    },
+    {
+      "epoch": 1.4887640449438202,
+      "grad_norm": 1.030942440032959,
+      "learning_rate": 4.302398327017214e-06,
+      "loss": 0.6191,
+      "step": 4240
+    },
+    {
+      "epoch": 1.4891151685393258,
+      "grad_norm": 0.8936895728111267,
+      "learning_rate": 4.3020779152485986e-06,
+      "loss": 0.6483,
+      "step": 4241
+    },
+    {
+      "epoch": 1.4894662921348314,
+      "grad_norm": 0.8798382878303528,
+      "learning_rate": 4.30175744185046e-06,
+      "loss": 0.619,
+      "step": 4242
+    },
+    {
+      "epoch": 1.4898174157303372,
+      "grad_norm": 0.8868904709815979,
+      "learning_rate": 4.30143690683376e-06,
+      "loss": 0.6033,
+      "step": 4243
+    },
+    {
+      "epoch": 1.4901685393258428,
+      "grad_norm": 0.8829790353775024,
+      "learning_rate": 4.301116310209459e-06,
+      "loss": 0.6283,
+      "step": 4244
+    },
+    {
+      "epoch": 1.4905196629213484,
+      "grad_norm": 0.8758097887039185,
+      "learning_rate": 4.3007956519885205e-06,
+      "loss": 0.5976,
+      "step": 4245
+    },
+    {
+      "epoch": 1.490870786516854,
+      "grad_norm": 0.865586519241333,
+      "learning_rate": 4.300474932181912e-06,
+      "loss": 0.6278,
+      "step": 4246
+    },
+    {
+      "epoch": 1.4912219101123596,
+      "grad_norm": 0.8885650038719177,
+      "learning_rate": 4.300154150800602e-06,
+      "loss": 0.5966,
+      "step": 4247
+    },
+    {
+      "epoch": 1.4915730337078652,
+      "grad_norm": 0.9123331308364868,
+      "learning_rate": 4.29983330785556e-06,
+      "loss": 0.6209,
+      "step": 4248
+    },
+    {
+      "epoch": 1.4919241573033708,
+      "grad_norm": 0.9102221131324768,
+      "learning_rate": 4.299512403357758e-06,
+      "loss": 0.6052,
+      "step": 4249
+    },
+    {
+      "epoch": 1.4922752808988764,
+      "grad_norm": 0.8728795051574707,
+      "learning_rate": 4.299191437318173e-06,
+      "loss": 0.5967,
+      "step": 4250
+    },
+    {
+      "epoch": 1.492626404494382,
+      "grad_norm": 0.8740513920783997,
+      "learning_rate": 4.29887040974778e-06,
+      "loss": 0.6309,
+      "step": 4251
+    },
+    {
+      "epoch": 1.4929775280898876,
+      "grad_norm": 0.8648720979690552,
+      "learning_rate": 4.298549320657559e-06,
+      "loss": 0.625,
+      "step": 4252
+    },
+    {
+      "epoch": 1.4933286516853932,
+      "grad_norm": 0.9204023480415344,
+      "learning_rate": 4.29822817005849e-06,
+      "loss": 0.6341,
+      "step": 4253
+    },
+    {
+      "epoch": 1.4936797752808988,
+      "grad_norm": 0.877215564250946,
+      "learning_rate": 4.2979069579615566e-06,
+      "loss": 0.6547,
+      "step": 4254
+    },
+    {
+      "epoch": 1.4940308988764044,
+      "grad_norm": 0.8751193881034851,
+      "learning_rate": 4.297585684377742e-06,
+      "loss": 0.6074,
+      "step": 4255
+    },
+    {
+      "epoch": 1.49438202247191,
+      "grad_norm": 0.8720304369926453,
+      "learning_rate": 4.297264349318037e-06,
+      "loss": 0.6214,
+      "step": 4256
+    },
+    {
+      "epoch": 1.4947331460674158,
+      "grad_norm": 0.8909294605255127,
+      "learning_rate": 4.296942952793429e-06,
+      "loss": 0.6194,
+      "step": 4257
+    },
+    {
+      "epoch": 1.4950842696629214,
+      "grad_norm": 0.8968172669410706,
+      "learning_rate": 4.296621494814909e-06,
+      "loss": 0.6196,
+      "step": 4258
+    },
+    {
+      "epoch": 1.495435393258427,
+      "grad_norm": 0.8939587473869324,
+      "learning_rate": 4.296299975393472e-06,
+      "loss": 0.67,
+      "step": 4259
+    },
+    {
+      "epoch": 1.4957865168539326,
+      "grad_norm": 0.9261441826820374,
+      "learning_rate": 4.2959783945401125e-06,
+      "loss": 0.6346,
+      "step": 4260
+    },
+    {
+      "epoch": 1.4961376404494382,
+      "grad_norm": 0.8645102381706238,
+      "learning_rate": 4.295656752265829e-06,
+      "loss": 0.6358,
+      "step": 4261
+    },
+    {
+      "epoch": 1.4964887640449438,
+      "grad_norm": 0.8751295804977417,
+      "learning_rate": 4.2953350485816205e-06,
+      "loss": 0.6288,
+      "step": 4262
+    },
+    {
+      "epoch": 1.4968398876404494,
+      "grad_norm": 0.9099578857421875,
+      "learning_rate": 4.29501328349849e-06,
+      "loss": 0.6367,
+      "step": 4263
+    },
+    {
+      "epoch": 1.497191011235955,
+      "grad_norm": 0.8653101325035095,
+      "learning_rate": 4.294691457027441e-06,
+      "loss": 0.6113,
+      "step": 4264
+    },
+    {
+      "epoch": 1.4975421348314606,
+      "grad_norm": 0.8679119348526001,
+      "learning_rate": 4.29436956917948e-06,
+      "loss": 0.6003,
+      "step": 4265
+    },
+    {
+      "epoch": 1.4978932584269664,
+      "grad_norm": 0.8774658441543579,
+      "learning_rate": 4.294047619965616e-06,
+      "loss": 0.62,
+      "step": 4266
+    },
+    {
+      "epoch": 1.498244382022472,
+      "grad_norm": 0.8665967583656311,
+      "learning_rate": 4.293725609396858e-06,
+      "loss": 0.6274,
+      "step": 4267
+    },
+    {
+      "epoch": 1.4985955056179776,
+      "grad_norm": 0.8959212899208069,
+      "learning_rate": 4.293403537484219e-06,
+      "loss": 0.5988,
+      "step": 4268
+    },
+    {
+      "epoch": 1.4989466292134832,
+      "grad_norm": 0.8877567648887634,
+      "learning_rate": 4.293081404238714e-06,
+      "loss": 0.6771,
+      "step": 4269
+    },
+    {
+      "epoch": 1.4992977528089888,
+      "grad_norm": 0.8664157390594482,
+      "learning_rate": 4.292759209671359e-06,
+      "loss": 0.6581,
+      "step": 4270
+    },
+    {
+      "epoch": 1.4996488764044944,
+      "grad_norm": 0.9112155437469482,
+      "learning_rate": 4.292436953793173e-06,
+      "loss": 0.6558,
+      "step": 4271
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8777109980583191,
+      "learning_rate": 4.292114636615176e-06,
+      "loss": 0.6097,
+      "step": 4272
+    },
+    {
+      "epoch": 1.5003511235955056,
+      "grad_norm": 0.8593363761901855,
+      "learning_rate": 4.291792258148394e-06,
+      "loss": 0.6,
+      "step": 4273
+    },
+    {
+      "epoch": 1.5007022471910112,
+      "grad_norm": 0.8953386545181274,
+      "learning_rate": 4.291469818403849e-06,
+      "loss": 0.6188,
+      "step": 4274
+    },
+    {
+      "epoch": 1.5010533707865168,
+      "grad_norm": 0.8898375630378723,
+      "learning_rate": 4.291147317392569e-06,
+      "loss": 0.6432,
+      "step": 4275
+    },
+    {
+      "epoch": 1.5014044943820224,
+      "grad_norm": 0.8718478679656982,
+      "learning_rate": 4.290824755125584e-06,
+      "loss": 0.5929,
+      "step": 4276
+    },
+    {
+      "epoch": 1.501755617977528,
+      "grad_norm": 0.8857758641242981,
+      "learning_rate": 4.290502131613925e-06,
+      "loss": 0.6434,
+      "step": 4277
+    },
+    {
+      "epoch": 1.5021067415730336,
+      "grad_norm": 0.8710552453994751,
+      "learning_rate": 4.290179446868624e-06,
+      "loss": 0.5877,
+      "step": 4278
+    },
+    {
+      "epoch": 1.5024578651685392,
+      "grad_norm": 0.8936992883682251,
+      "learning_rate": 4.289856700900719e-06,
+      "loss": 0.5977,
+      "step": 4279
+    },
+    {
+      "epoch": 1.5028089887640448,
+      "grad_norm": 0.8890673518180847,
+      "learning_rate": 4.289533893721246e-06,
+      "loss": 0.662,
+      "step": 4280
+    },
+    {
+      "epoch": 1.5031601123595506,
+      "grad_norm": 0.8563557267189026,
+      "learning_rate": 4.289211025341246e-06,
+      "loss": 0.6105,
+      "step": 4281
+    },
+    {
+      "epoch": 1.5035112359550562,
+      "grad_norm": 0.8855664730072021,
+      "learning_rate": 4.288888095771759e-06,
+      "loss": 0.6502,
+      "step": 4282
+    },
+    {
+      "epoch": 1.5038623595505618,
+      "grad_norm": 0.8829025626182556,
+      "learning_rate": 4.2885651050238294e-06,
+      "loss": 0.6375,
+      "step": 4283
+    },
+    {
+      "epoch": 1.5042134831460674,
+      "grad_norm": 0.8763623237609863,
+      "learning_rate": 4.288242053108505e-06,
+      "loss": 0.6173,
+      "step": 4284
+    },
+    {
+      "epoch": 1.504564606741573,
+      "grad_norm": 0.8628937005996704,
+      "learning_rate": 4.287918940036831e-06,
+      "loss": 0.6088,
+      "step": 4285
+    },
+    {
+      "epoch": 1.5049157303370788,
+      "grad_norm": 0.8726745247840881,
+      "learning_rate": 4.287595765819861e-06,
+      "loss": 0.6046,
+      "step": 4286
+    },
+    {
+      "epoch": 1.5052668539325844,
+      "grad_norm": 0.886001706123352,
+      "learning_rate": 4.287272530468644e-06,
+      "loss": 0.6755,
+      "step": 4287
+    },
+    {
+      "epoch": 1.50561797752809,
+      "grad_norm": 0.9122753143310547,
+      "learning_rate": 4.286949233994238e-06,
+      "loss": 0.6319,
+      "step": 4288
+    },
+    {
+      "epoch": 1.5059691011235956,
+      "grad_norm": 0.8860533833503723,
+      "learning_rate": 4.286625876407695e-06,
+      "loss": 0.6237,
+      "step": 4289
+    },
+    {
+      "epoch": 1.5063202247191012,
+      "grad_norm": 0.8959923386573792,
+      "learning_rate": 4.286302457720077e-06,
+      "loss": 0.6397,
+      "step": 4290
+    },
+    {
+      "epoch": 1.5066713483146068,
+      "grad_norm": 0.872547447681427,
+      "learning_rate": 4.285978977942443e-06,
+      "loss": 0.6098,
+      "step": 4291
+    },
+    {
+      "epoch": 1.5070224719101124,
+      "grad_norm": 0.885536253452301,
+      "learning_rate": 4.2856554370858565e-06,
+      "loss": 0.6148,
+      "step": 4292
+    },
+    {
+      "epoch": 1.507373595505618,
+      "grad_norm": 0.8694351315498352,
+      "learning_rate": 4.2853318351613825e-06,
+      "loss": 0.6649,
+      "step": 4293
+    },
+    {
+      "epoch": 1.5077247191011236,
+      "grad_norm": 0.9179470539093018,
+      "learning_rate": 4.285008172180087e-06,
+      "loss": 0.6518,
+      "step": 4294
+    },
+    {
+      "epoch": 1.5080758426966292,
+      "grad_norm": 0.9301983714103699,
+      "learning_rate": 4.2846844481530395e-06,
+      "loss": 0.6733,
+      "step": 4295
+    },
+    {
+      "epoch": 1.5084269662921348,
+      "grad_norm": 0.874769926071167,
+      "learning_rate": 4.28436066309131e-06,
+      "loss": 0.6206,
+      "step": 4296
+    },
+    {
+      "epoch": 1.5087780898876404,
+      "grad_norm": 0.8874285817146301,
+      "learning_rate": 4.284036817005974e-06,
+      "loss": 0.6427,
+      "step": 4297
+    },
+    {
+      "epoch": 1.509129213483146,
+      "grad_norm": 0.8830395340919495,
+      "learning_rate": 4.283712909908105e-06,
+      "loss": 0.6715,
+      "step": 4298
+    },
+    {
+      "epoch": 1.5094803370786516,
+      "grad_norm": 0.8672149777412415,
+      "learning_rate": 4.283388941808782e-06,
+      "loss": 0.6019,
+      "step": 4299
+    },
+    {
+      "epoch": 1.5098314606741572,
+      "grad_norm": 0.9076434969902039,
+      "learning_rate": 4.283064912719081e-06,
+      "loss": 0.6157,
+      "step": 4300
+    },
+    {
+      "epoch": 1.5101825842696628,
+      "grad_norm": 0.8910434246063232,
+      "learning_rate": 4.282740822650087e-06,
+      "loss": 0.614,
+      "step": 4301
+    },
+    {
+      "epoch": 1.5105337078651684,
+      "grad_norm": 0.9061810374259949,
+      "learning_rate": 4.282416671612883e-06,
+      "loss": 0.6202,
+      "step": 4302
+    },
+    {
+      "epoch": 1.5108848314606742,
+      "grad_norm": 0.877909243106842,
+      "learning_rate": 4.282092459618553e-06,
+      "loss": 0.618,
+      "step": 4303
+    },
+    {
+      "epoch": 1.5112359550561798,
+      "grad_norm": 0.8897304534912109,
+      "learning_rate": 4.281768186678187e-06,
+      "loss": 0.646,
+      "step": 4304
+    },
+    {
+      "epoch": 1.5115870786516854,
+      "grad_norm": 0.9377396106719971,
+      "learning_rate": 4.281443852802873e-06,
+      "loss": 0.6323,
+      "step": 4305
+    },
+    {
+      "epoch": 1.511938202247191,
+      "grad_norm": 0.8810194730758667,
+      "learning_rate": 4.281119458003704e-06,
+      "loss": 0.6372,
+      "step": 4306
+    },
+    {
+      "epoch": 1.5122893258426966,
+      "grad_norm": 0.8763102293014526,
+      "learning_rate": 4.280795002291773e-06,
+      "loss": 0.6189,
+      "step": 4307
+    },
+    {
+      "epoch": 1.5126404494382022,
+      "grad_norm": 0.8697223663330078,
+      "learning_rate": 4.280470485678177e-06,
+      "loss": 0.6188,
+      "step": 4308
+    },
+    {
+      "epoch": 1.512991573033708,
+      "grad_norm": 0.8664148449897766,
+      "learning_rate": 4.2801459081740144e-06,
+      "loss": 0.6135,
+      "step": 4309
+    },
+    {
+      "epoch": 1.5133426966292136,
+      "grad_norm": 0.877975583076477,
+      "learning_rate": 4.279821269790384e-06,
+      "loss": 0.607,
+      "step": 4310
+    },
+    {
+      "epoch": 1.5136938202247192,
+      "grad_norm": 0.8779658079147339,
+      "learning_rate": 4.2794965705383906e-06,
+      "loss": 0.5998,
+      "step": 4311
+    },
+    {
+      "epoch": 1.5140449438202248,
+      "grad_norm": 0.8846172094345093,
+      "learning_rate": 4.279171810429136e-06,
+      "loss": 0.6212,
+      "step": 4312
+    },
+    {
+      "epoch": 1.5143960674157304,
+      "grad_norm": 0.8623836040496826,
+      "learning_rate": 4.278846989473729e-06,
+      "loss": 0.6333,
+      "step": 4313
+    },
+    {
+      "epoch": 1.514747191011236,
+      "grad_norm": 0.9182273149490356,
+      "learning_rate": 4.2785221076832765e-06,
+      "loss": 0.6423,
+      "step": 4314
+    },
+    {
+      "epoch": 1.5150983146067416,
+      "grad_norm": 0.879574716091156,
+      "learning_rate": 4.278197165068891e-06,
+      "loss": 0.621,
+      "step": 4315
+    },
+    {
+      "epoch": 1.5154494382022472,
+      "grad_norm": 0.8660203814506531,
+      "learning_rate": 4.277872161641682e-06,
+      "loss": 0.582,
+      "step": 4316
+    },
+    {
+      "epoch": 1.5158005617977528,
+      "grad_norm": 0.8960110545158386,
+      "learning_rate": 4.277547097412768e-06,
+      "loss": 0.6073,
+      "step": 4317
+    },
+    {
+      "epoch": 1.5161516853932584,
+      "grad_norm": 0.8762567639350891,
+      "learning_rate": 4.277221972393263e-06,
+      "loss": 0.5839,
+      "step": 4318
+    },
+    {
+      "epoch": 1.516502808988764,
+      "grad_norm": 0.8783929944038391,
+      "learning_rate": 4.276896786594288e-06,
+      "loss": 0.6342,
+      "step": 4319
+    },
+    {
+      "epoch": 1.5168539325842696,
+      "grad_norm": 0.8779000043869019,
+      "learning_rate": 4.276571540026964e-06,
+      "loss": 0.6274,
+      "step": 4320
+    },
+    {
+      "epoch": 1.5172050561797752,
+      "grad_norm": 0.8748044967651367,
+      "learning_rate": 4.276246232702413e-06,
+      "loss": 0.614,
+      "step": 4321
+    },
+    {
+      "epoch": 1.5175561797752808,
+      "grad_norm": 0.9069294333457947,
+      "learning_rate": 4.275920864631761e-06,
+      "loss": 0.6438,
+      "step": 4322
+    },
+    {
+      "epoch": 1.5179073033707864,
+      "grad_norm": 0.8882784247398376,
+      "learning_rate": 4.275595435826134e-06,
+      "loss": 0.6104,
+      "step": 4323
+    },
+    {
+      "epoch": 1.518258426966292,
+      "grad_norm": 0.8945493698120117,
+      "learning_rate": 4.275269946296664e-06,
+      "loss": 0.5955,
+      "step": 4324
+    },
+    {
+      "epoch": 1.5186095505617978,
+      "grad_norm": 0.9070472121238708,
+      "learning_rate": 4.27494439605448e-06,
+      "loss": 0.6506,
+      "step": 4325
+    },
+    {
+      "epoch": 1.5189606741573034,
+      "grad_norm": 0.8988576531410217,
+      "learning_rate": 4.274618785110718e-06,
+      "loss": 0.619,
+      "step": 4326
+    },
+    {
+      "epoch": 1.519311797752809,
+      "grad_norm": 0.8801273107528687,
+      "learning_rate": 4.274293113476511e-06,
+      "loss": 0.598,
+      "step": 4327
+    },
+    {
+      "epoch": 1.5196629213483146,
+      "grad_norm": 0.8812402486801147,
+      "learning_rate": 4.273967381162998e-06,
+      "loss": 0.6274,
+      "step": 4328
+    },
+    {
+      "epoch": 1.5200140449438202,
+      "grad_norm": 0.8644037246704102,
+      "learning_rate": 4.273641588181317e-06,
+      "loss": 0.5863,
+      "step": 4329
+    },
+    {
+      "epoch": 1.5203651685393258,
+      "grad_norm": 0.8854338526725769,
+      "learning_rate": 4.2733157345426135e-06,
+      "loss": 0.6315,
+      "step": 4330
+    },
+    {
+      "epoch": 1.5207162921348316,
+      "grad_norm": 0.9194030165672302,
+      "learning_rate": 4.272989820258029e-06,
+      "loss": 0.6241,
+      "step": 4331
+    },
+    {
+      "epoch": 1.5210674157303372,
+      "grad_norm": 0.8962981700897217,
+      "learning_rate": 4.27266384533871e-06,
+      "loss": 0.5799,
+      "step": 4332
+    },
+    {
+      "epoch": 1.5214185393258428,
+      "grad_norm": 0.8707279562950134,
+      "learning_rate": 4.2723378097958035e-06,
+      "loss": 0.6517,
+      "step": 4333
+    },
+    {
+      "epoch": 1.5217696629213484,
+      "grad_norm": 0.8625728487968445,
+      "learning_rate": 4.272011713640461e-06,
+      "loss": 0.5913,
+      "step": 4334
+    },
+    {
+      "epoch": 1.522120786516854,
+      "grad_norm": 0.8925690054893494,
+      "learning_rate": 4.271685556883833e-06,
+      "loss": 0.6334,
+      "step": 4335
+    },
+    {
+      "epoch": 1.5224719101123596,
+      "grad_norm": 0.8702839612960815,
+      "learning_rate": 4.271359339537077e-06,
+      "loss": 0.5967,
+      "step": 4336
+    },
+    {
+      "epoch": 1.5228230337078652,
+      "grad_norm": 0.9461783170700073,
+      "learning_rate": 4.271033061611346e-06,
+      "loss": 0.634,
+      "step": 4337
+    },
+    {
+      "epoch": 1.5231741573033708,
+      "grad_norm": 1.0706638097763062,
+      "learning_rate": 4.2707067231178e-06,
+      "loss": 0.6449,
+      "step": 4338
+    },
+    {
+      "epoch": 1.5235252808988764,
+      "grad_norm": 0.9275455474853516,
+      "learning_rate": 4.2703803240676e-06,
+      "loss": 0.6322,
+      "step": 4339
+    },
+    {
+      "epoch": 1.523876404494382,
+      "grad_norm": 0.8851011991500854,
+      "learning_rate": 4.270053864471907e-06,
+      "loss": 0.6354,
+      "step": 4340
+    },
+    {
+      "epoch": 1.5242275280898876,
+      "grad_norm": 0.8991807699203491,
+      "learning_rate": 4.269727344341887e-06,
+      "loss": 0.6617,
+      "step": 4341
+    },
+    {
+      "epoch": 1.5245786516853932,
+      "grad_norm": 0.8674828410148621,
+      "learning_rate": 4.269400763688706e-06,
+      "loss": 0.5909,
+      "step": 4342
+    },
+    {
+      "epoch": 1.5249297752808988,
+      "grad_norm": 0.8902739882469177,
+      "learning_rate": 4.2690741225235336e-06,
+      "loss": 0.6346,
+      "step": 4343
+    },
+    {
+      "epoch": 1.5252808988764044,
+      "grad_norm": 0.9026355743408203,
+      "learning_rate": 4.2687474208575395e-06,
+      "loss": 0.6505,
+      "step": 4344
+    },
+    {
+      "epoch": 1.52563202247191,
+      "grad_norm": 0.8712281584739685,
+      "learning_rate": 4.268420658701897e-06,
+      "loss": 0.6541,
+      "step": 4345
+    },
+    {
+      "epoch": 1.5259831460674156,
+      "grad_norm": 0.874847412109375,
+      "learning_rate": 4.268093836067781e-06,
+      "loss": 0.5961,
+      "step": 4346
+    },
+    {
+      "epoch": 1.5263342696629212,
+      "grad_norm": 0.8955193758010864,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.6221,
+      "step": 4347
+    },
+    {
+      "epoch": 1.526685393258427,
+      "grad_norm": 0.8489906191825867,
+      "learning_rate": 4.2674400094088405e-06,
+      "loss": 0.6062,
+      "step": 4348
+    },
+    {
+      "epoch": 1.5270365168539326,
+      "grad_norm": 0.9131476283073425,
+      "learning_rate": 4.267113005406375e-06,
+      "loss": 0.6277,
+      "step": 4349
+    },
+    {
+      "epoch": 1.5273876404494382,
+      "grad_norm": 0.9100372195243835,
+      "learning_rate": 4.266785940970158e-06,
+      "loss": 0.6168,
+      "step": 4350
+    },
+    {
+      "epoch": 1.5277387640449438,
+      "grad_norm": 0.8778021335601807,
+      "learning_rate": 4.266458816111373e-06,
+      "loss": 0.6156,
+      "step": 4351
+    },
+    {
+      "epoch": 1.5280898876404494,
+      "grad_norm": 0.8872804045677185,
+      "learning_rate": 4.266131630841208e-06,
+      "loss": 0.6325,
+      "step": 4352
+    },
+    {
+      "epoch": 1.5284410112359552,
+      "grad_norm": 0.8724574446678162,
+      "learning_rate": 4.265804385170853e-06,
+      "loss": 0.6053,
+      "step": 4353
+    },
+    {
+      "epoch": 1.5287921348314608,
+      "grad_norm": 0.875560998916626,
+      "learning_rate": 4.265477079111499e-06,
+      "loss": 0.6343,
+      "step": 4354
+    },
+    {
+      "epoch": 1.5291432584269664,
+      "grad_norm": 0.9058078527450562,
+      "learning_rate": 4.26514971267434e-06,
+      "loss": 0.6229,
+      "step": 4355
+    },
+    {
+      "epoch": 1.529494382022472,
+      "grad_norm": 0.8583801984786987,
+      "learning_rate": 4.26482228587057e-06,
+      "loss": 0.6371,
+      "step": 4356
+    },
+    {
+      "epoch": 1.5298455056179776,
+      "grad_norm": 0.8806894421577454,
+      "learning_rate": 4.264494798711389e-06,
+      "loss": 0.6322,
+      "step": 4357
+    },
+    {
+      "epoch": 1.5301966292134832,
+      "grad_norm": 0.8982591032981873,
+      "learning_rate": 4.264167251207995e-06,
+      "loss": 0.6166,
+      "step": 4358
+    },
+    {
+      "epoch": 1.5305477528089888,
+      "grad_norm": 0.8655708432197571,
+      "learning_rate": 4.263839643371591e-06,
+      "loss": 0.6254,
+      "step": 4359
+    },
+    {
+      "epoch": 1.5308988764044944,
+      "grad_norm": 0.8754801154136658,
+      "learning_rate": 4.263511975213381e-06,
+      "loss": 0.6228,
+      "step": 4360
+    },
+    {
+      "epoch": 1.53125,
+      "grad_norm": 0.8953879475593567,
+      "learning_rate": 4.263184246744571e-06,
+      "loss": 0.596,
+      "step": 4361
+    },
+    {
+      "epoch": 1.5316011235955056,
+      "grad_norm": 0.8852868676185608,
+      "learning_rate": 4.2628564579763675e-06,
+      "loss": 0.6318,
+      "step": 4362
+    },
+    {
+      "epoch": 1.5319522471910112,
+      "grad_norm": 0.8706391453742981,
+      "learning_rate": 4.262528608919981e-06,
+      "loss": 0.6429,
+      "step": 4363
+    },
+    {
+      "epoch": 1.5323033707865168,
+      "grad_norm": 0.8404558897018433,
+      "learning_rate": 4.262200699586626e-06,
+      "loss": 0.5882,
+      "step": 4364
+    },
+    {
+      "epoch": 1.5326544943820224,
+      "grad_norm": 0.8769184350967407,
+      "learning_rate": 4.261872729987514e-06,
+      "loss": 0.6284,
+      "step": 4365
+    },
+    {
+      "epoch": 1.533005617977528,
+      "grad_norm": 0.8632640838623047,
+      "learning_rate": 4.261544700133863e-06,
+      "loss": 0.5794,
+      "step": 4366
+    },
+    {
+      "epoch": 1.5333567415730336,
+      "grad_norm": 0.8795828223228455,
+      "learning_rate": 4.26121661003689e-06,
+      "loss": 0.6281,
+      "step": 4367
+    },
+    {
+      "epoch": 1.5337078651685392,
+      "grad_norm": 0.9072569012641907,
+      "learning_rate": 4.260888459707816e-06,
+      "loss": 0.6154,
+      "step": 4368
+    },
+    {
+      "epoch": 1.5340589887640448,
+      "grad_norm": 0.8855779767036438,
+      "learning_rate": 4.260560249157864e-06,
+      "loss": 0.6199,
+      "step": 4369
+    },
+    {
+      "epoch": 1.5344101123595506,
+      "grad_norm": 0.8932167887687683,
+      "learning_rate": 4.260231978398257e-06,
+      "loss": 0.6155,
+      "step": 4370
+    },
+    {
+      "epoch": 1.5347612359550562,
+      "grad_norm": 0.8883824348449707,
+      "learning_rate": 4.259903647440223e-06,
+      "loss": 0.6386,
+      "step": 4371
+    },
+    {
+      "epoch": 1.5351123595505618,
+      "grad_norm": 0.8885374665260315,
+      "learning_rate": 4.259575256294991e-06,
+      "loss": 0.6175,
+      "step": 4372
+    },
+    {
+      "epoch": 1.5354634831460674,
+      "grad_norm": 0.9104458689689636,
+      "learning_rate": 4.25924680497379e-06,
+      "loss": 0.6223,
+      "step": 4373
+    },
+    {
+      "epoch": 1.535814606741573,
+      "grad_norm": 0.8783615231513977,
+      "learning_rate": 4.258918293487854e-06,
+      "loss": 0.645,
+      "step": 4374
+    },
+    {
+      "epoch": 1.5361657303370788,
+      "grad_norm": 0.8793334364891052,
+      "learning_rate": 4.2585897218484175e-06,
+      "loss": 0.6762,
+      "step": 4375
+    },
+    {
+      "epoch": 1.5365168539325844,
+      "grad_norm": 0.8661759495735168,
+      "learning_rate": 4.258261090066718e-06,
+      "loss": 0.5936,
+      "step": 4376
+    },
+    {
+      "epoch": 1.53686797752809,
+      "grad_norm": 0.9374626874923706,
+      "learning_rate": 4.257932398153993e-06,
+      "loss": 0.6332,
+      "step": 4377
+    },
+    {
+      "epoch": 1.5372191011235956,
+      "grad_norm": 0.8962463736534119,
+      "learning_rate": 4.2576036461214844e-06,
+      "loss": 0.6104,
+      "step": 4378
+    },
+    {
+      "epoch": 1.5375702247191012,
+      "grad_norm": 0.8866143822669983,
+      "learning_rate": 4.257274833980435e-06,
+      "loss": 0.6105,
+      "step": 4379
+    },
+    {
+      "epoch": 1.5379213483146068,
+      "grad_norm": 0.9056102633476257,
+      "learning_rate": 4.256945961742091e-06,
+      "loss": 0.6392,
+      "step": 4380
+    },
+    {
+      "epoch": 1.5382724719101124,
+      "grad_norm": 0.8876556158065796,
+      "learning_rate": 4.256617029417696e-06,
+      "loss": 0.5981,
+      "step": 4381
+    },
+    {
+      "epoch": 1.538623595505618,
+      "grad_norm": 0.8786119222640991,
+      "learning_rate": 4.256288037018504e-06,
+      "loss": 0.6489,
+      "step": 4382
+    },
+    {
+      "epoch": 1.5389747191011236,
+      "grad_norm": 0.9107440114021301,
+      "learning_rate": 4.255958984555764e-06,
+      "loss": 0.6233,
+      "step": 4383
+    },
+    {
+      "epoch": 1.5393258426966292,
+      "grad_norm": 0.8715046048164368,
+      "learning_rate": 4.255629872040728e-06,
+      "loss": 0.6196,
+      "step": 4384
+    },
+    {
+      "epoch": 1.5396769662921348,
+      "grad_norm": 0.8908570408821106,
+      "learning_rate": 4.255300699484654e-06,
+      "loss": 0.6353,
+      "step": 4385
+    },
+    {
+      "epoch": 1.5400280898876404,
+      "grad_norm": 0.9205907583236694,
+      "learning_rate": 4.254971466898798e-06,
+      "loss": 0.6375,
+      "step": 4386
+    },
+    {
+      "epoch": 1.540379213483146,
+      "grad_norm": 0.8485553860664368,
+      "learning_rate": 4.254642174294419e-06,
+      "loss": 0.5977,
+      "step": 4387
+    },
+    {
+      "epoch": 1.5407303370786516,
+      "grad_norm": 0.9237390160560608,
+      "learning_rate": 4.254312821682779e-06,
+      "loss": 0.6641,
+      "step": 4388
+    },
+    {
+      "epoch": 1.5410814606741572,
+      "grad_norm": 0.8933415412902832,
+      "learning_rate": 4.253983409075142e-06,
+      "loss": 0.6581,
+      "step": 4389
+    },
+    {
+      "epoch": 1.5414325842696628,
+      "grad_norm": 0.9153465032577515,
+      "learning_rate": 4.2536539364827734e-06,
+      "loss": 0.6331,
+      "step": 4390
+    },
+    {
+      "epoch": 1.5417837078651684,
+      "grad_norm": 0.8514824509620667,
+      "learning_rate": 4.25332440391694e-06,
+      "loss": 0.6189,
+      "step": 4391
+    },
+    {
+      "epoch": 1.5421348314606742,
+      "grad_norm": 0.8722473382949829,
+      "learning_rate": 4.252994811388914e-06,
+      "loss": 0.6107,
+      "step": 4392
+    },
+    {
+      "epoch": 1.5424859550561798,
+      "grad_norm": 0.8800398707389832,
+      "learning_rate": 4.252665158909964e-06,
+      "loss": 0.6291,
+      "step": 4393
+    },
+    {
+      "epoch": 1.5428370786516854,
+      "grad_norm": 0.8885344862937927,
+      "learning_rate": 4.252335446491366e-06,
+      "loss": 0.5903,
+      "step": 4394
+    },
+    {
+      "epoch": 1.543188202247191,
+      "grad_norm": 0.9298628568649292,
+      "learning_rate": 4.2520056741443956e-06,
+      "loss": 0.6482,
+      "step": 4395
+    },
+    {
+      "epoch": 1.5435393258426966,
+      "grad_norm": 0.9046818614006042,
+      "learning_rate": 4.251675841880329e-06,
+      "loss": 0.6256,
+      "step": 4396
+    },
+    {
+      "epoch": 1.5438904494382022,
+      "grad_norm": 0.902118444442749,
+      "learning_rate": 4.251345949710449e-06,
+      "loss": 0.6262,
+      "step": 4397
+    },
+    {
+      "epoch": 1.544241573033708,
+      "grad_norm": 0.8869757652282715,
+      "learning_rate": 4.251015997646035e-06,
+      "loss": 0.6324,
+      "step": 4398
+    },
+    {
+      "epoch": 1.5445926966292136,
+      "grad_norm": 0.8704875707626343,
+      "learning_rate": 4.250685985698374e-06,
+      "loss": 0.6217,
+      "step": 4399
+    },
+    {
+      "epoch": 1.5449438202247192,
+      "grad_norm": 0.8620396852493286,
+      "learning_rate": 4.250355913878749e-06,
+      "loss": 0.653,
+      "step": 4400
+    },
+    {
+      "epoch": 1.5452949438202248,
+      "grad_norm": 0.9138562679290771,
+      "learning_rate": 4.250025782198449e-06,
+      "loss": 0.638,
+      "step": 4401
+    },
+    {
+      "epoch": 1.5456460674157304,
+      "grad_norm": 0.874911904335022,
+      "learning_rate": 4.249695590668765e-06,
+      "loss": 0.6063,
+      "step": 4402
+    },
+    {
+      "epoch": 1.545997191011236,
+      "grad_norm": 0.867830753326416,
+      "learning_rate": 4.249365339300991e-06,
+      "loss": 0.6183,
+      "step": 4403
+    },
+    {
+      "epoch": 1.5463483146067416,
+      "grad_norm": 0.8954204320907593,
+      "learning_rate": 4.249035028106418e-06,
+      "loss": 0.6259,
+      "step": 4404
+    },
+    {
+      "epoch": 1.5466994382022472,
+      "grad_norm": 0.8731714487075806,
+      "learning_rate": 4.248704657096343e-06,
+      "loss": 0.601,
+      "step": 4405
+    },
+    {
+      "epoch": 1.5470505617977528,
+      "grad_norm": 0.8896400332450867,
+      "learning_rate": 4.248374226282066e-06,
+      "loss": 0.6124,
+      "step": 4406
+    },
+    {
+      "epoch": 1.5474016853932584,
+      "grad_norm": 0.8725085258483887,
+      "learning_rate": 4.248043735674886e-06,
+      "loss": 0.6295,
+      "step": 4407
+    },
+    {
+      "epoch": 1.547752808988764,
+      "grad_norm": 0.9611513018608093,
+      "learning_rate": 4.2477131852861064e-06,
+      "loss": 0.6177,
+      "step": 4408
+    },
+    {
+      "epoch": 1.5481039325842696,
+      "grad_norm": 0.8641888499259949,
+      "learning_rate": 4.247382575127031e-06,
+      "loss": 0.6317,
+      "step": 4409
+    },
+    {
+      "epoch": 1.5484550561797752,
+      "grad_norm": 0.9119146466255188,
+      "learning_rate": 4.247051905208967e-06,
+      "loss": 0.6246,
+      "step": 4410
+    },
+    {
+      "epoch": 1.5488061797752808,
+      "grad_norm": 0.8706852197647095,
+      "learning_rate": 4.246721175543224e-06,
+      "loss": 0.6083,
+      "step": 4411
+    },
+    {
+      "epoch": 1.5491573033707864,
+      "grad_norm": 0.8891316056251526,
+      "learning_rate": 4.246390386141109e-06,
+      "loss": 0.6409,
+      "step": 4412
+    },
+    {
+      "epoch": 1.549508426966292,
+      "grad_norm": 0.8789838552474976,
+      "learning_rate": 4.246059537013939e-06,
+      "loss": 0.6488,
+      "step": 4413
+    },
+    {
+      "epoch": 1.5498595505617978,
+      "grad_norm": 0.8697891235351562,
+      "learning_rate": 4.245728628173027e-06,
+      "loss": 0.6155,
+      "step": 4414
+    },
+    {
+      "epoch": 1.5502106741573034,
+      "grad_norm": 0.8879761695861816,
+      "learning_rate": 4.2453976596296895e-06,
+      "loss": 0.6194,
+      "step": 4415
+    },
+    {
+      "epoch": 1.550561797752809,
+      "grad_norm": 0.8760266304016113,
+      "learning_rate": 4.245066631395245e-06,
+      "loss": 0.6037,
+      "step": 4416
+    },
+    {
+      "epoch": 1.5509129213483146,
+      "grad_norm": 0.8705148696899414,
+      "learning_rate": 4.244735543481016e-06,
+      "loss": 0.6262,
+      "step": 4417
+    },
+    {
+      "epoch": 1.5512640449438202,
+      "grad_norm": 0.8751351833343506,
+      "learning_rate": 4.244404395898324e-06,
+      "loss": 0.6264,
+      "step": 4418
+    },
+    {
+      "epoch": 1.5516151685393258,
+      "grad_norm": 0.8804269433021545,
+      "learning_rate": 4.244073188658494e-06,
+      "loss": 0.6613,
+      "step": 4419
+    },
+    {
+      "epoch": 1.5519662921348316,
+      "grad_norm": 0.8883501291275024,
+      "learning_rate": 4.243741921772853e-06,
+      "loss": 0.6459,
+      "step": 4420
+    },
+    {
+      "epoch": 1.5523174157303372,
+      "grad_norm": 0.8871954679489136,
+      "learning_rate": 4.243410595252731e-06,
+      "loss": 0.6259,
+      "step": 4421
+    },
+    {
+      "epoch": 1.5526685393258428,
+      "grad_norm": 0.8802275657653809,
+      "learning_rate": 4.243079209109458e-06,
+      "loss": 0.5989,
+      "step": 4422
+    },
+    {
+      "epoch": 1.5530196629213484,
+      "grad_norm": 0.8572296500205994,
+      "learning_rate": 4.242747763354369e-06,
+      "loss": 0.6087,
+      "step": 4423
+    },
+    {
+      "epoch": 1.553370786516854,
+      "grad_norm": 0.9028926491737366,
+      "learning_rate": 4.242416257998797e-06,
+      "loss": 0.5896,
+      "step": 4424
+    },
+    {
+      "epoch": 1.5537219101123596,
+      "grad_norm": 0.9015838503837585,
+      "learning_rate": 4.2420846930540795e-06,
+      "loss": 0.7101,
+      "step": 4425
+    },
+    {
+      "epoch": 1.5540730337078652,
+      "grad_norm": 0.8723250031471252,
+      "learning_rate": 4.241753068531557e-06,
+      "loss": 0.646,
+      "step": 4426
+    },
+    {
+      "epoch": 1.5544241573033708,
+      "grad_norm": 0.8892289400100708,
+      "learning_rate": 4.241421384442569e-06,
+      "loss": 0.6446,
+      "step": 4427
+    },
+    {
+      "epoch": 1.5547752808988764,
+      "grad_norm": 0.8631646633148193,
+      "learning_rate": 4.24108964079846e-06,
+      "loss": 0.5921,
+      "step": 4428
+    },
+    {
+      "epoch": 1.555126404494382,
+      "grad_norm": 0.8563498258590698,
+      "learning_rate": 4.2407578376105765e-06,
+      "loss": 0.5942,
+      "step": 4429
+    },
+    {
+      "epoch": 1.5554775280898876,
+      "grad_norm": 0.883753776550293,
+      "learning_rate": 4.2404259748902635e-06,
+      "loss": 0.6359,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5558286516853932,
+      "grad_norm": 0.8877705335617065,
+      "learning_rate": 4.240094052648872e-06,
+      "loss": 0.6229,
+      "step": 4431
+    },
+    {
+      "epoch": 1.5561797752808988,
+      "grad_norm": 0.8870841860771179,
+      "learning_rate": 4.239762070897752e-06,
+      "loss": 0.6342,
+      "step": 4432
+    },
+    {
+      "epoch": 1.5565308988764044,
+      "grad_norm": 0.8865689635276794,
+      "learning_rate": 4.239430029648258e-06,
+      "loss": 0.64,
+      "step": 4433
+    },
+    {
+      "epoch": 1.55688202247191,
+      "grad_norm": 0.8670912981033325,
+      "learning_rate": 4.239097928911746e-06,
+      "loss": 0.6176,
+      "step": 4434
+    },
+    {
+      "epoch": 1.5572331460674156,
+      "grad_norm": 0.8641391396522522,
+      "learning_rate": 4.238765768699573e-06,
+      "loss": 0.6329,
+      "step": 4435
+    },
+    {
+      "epoch": 1.5575842696629212,
+      "grad_norm": 0.8882496356964111,
+      "learning_rate": 4.238433549023098e-06,
+      "loss": 0.6506,
+      "step": 4436
+    },
+    {
+      "epoch": 1.557935393258427,
+      "grad_norm": 0.8790398836135864,
+      "learning_rate": 4.238101269893684e-06,
+      "loss": 0.6053,
+      "step": 4437
+    },
+    {
+      "epoch": 1.5582865168539326,
+      "grad_norm": 0.8824111223220825,
+      "learning_rate": 4.237768931322694e-06,
+      "loss": 0.6207,
+      "step": 4438
+    },
+    {
+      "epoch": 1.5586376404494382,
+      "grad_norm": 0.8859694004058838,
+      "learning_rate": 4.237436533321492e-06,
+      "loss": 0.6206,
+      "step": 4439
+    },
+    {
+      "epoch": 1.5589887640449438,
+      "grad_norm": 0.8975111842155457,
+      "learning_rate": 4.237104075901449e-06,
+      "loss": 0.6106,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5593398876404494,
+      "grad_norm": 0.9005839228630066,
+      "learning_rate": 4.236771559073932e-06,
+      "loss": 0.6292,
+      "step": 4441
+    },
+    {
+      "epoch": 1.5596910112359552,
+      "grad_norm": 0.8936960101127625,
+      "learning_rate": 4.236438982850315e-06,
+      "loss": 0.6472,
+      "step": 4442
+    },
+    {
+      "epoch": 1.5600421348314608,
+      "grad_norm": 0.8928436636924744,
+      "learning_rate": 4.23610634724197e-06,
+      "loss": 0.6097,
+      "step": 4443
+    },
+    {
+      "epoch": 1.5603932584269664,
+      "grad_norm": 0.8569137454032898,
+      "learning_rate": 4.235773652260274e-06,
+      "loss": 0.5809,
+      "step": 4444
+    },
+    {
+      "epoch": 1.560744382022472,
+      "grad_norm": 0.8743844032287598,
+      "learning_rate": 4.235440897916604e-06,
+      "loss": 0.6311,
+      "step": 4445
+    },
+    {
+      "epoch": 1.5610955056179776,
+      "grad_norm": 0.9037197828292847,
+      "learning_rate": 4.23510808422234e-06,
+      "loss": 0.6386,
+      "step": 4446
+    },
+    {
+      "epoch": 1.5614466292134832,
+      "grad_norm": 0.8934659361839294,
+      "learning_rate": 4.234775211188864e-06,
+      "loss": 0.5906,
+      "step": 4447
+    },
+    {
+      "epoch": 1.5617977528089888,
+      "grad_norm": 0.8630779981613159,
+      "learning_rate": 4.23444227882756e-06,
+      "loss": 0.6175,
+      "step": 4448
+    },
+    {
+      "epoch": 1.5621488764044944,
+      "grad_norm": 0.9021822214126587,
+      "learning_rate": 4.234109287149816e-06,
+      "loss": 0.6217,
+      "step": 4449
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.9597007036209106,
+      "learning_rate": 4.233776236167017e-06,
+      "loss": 0.6448,
+      "step": 4450
+    },
+    {
+      "epoch": 1.5628511235955056,
+      "grad_norm": 0.8997489809989929,
+      "learning_rate": 4.233443125890555e-06,
+      "loss": 0.6307,
+      "step": 4451
+    },
+    {
+      "epoch": 1.5632022471910112,
+      "grad_norm": 0.8864565491676331,
+      "learning_rate": 4.233109956331821e-06,
+      "loss": 0.6215,
+      "step": 4452
+    },
+    {
+      "epoch": 1.5635533707865168,
+      "grad_norm": 0.8683627843856812,
+      "learning_rate": 4.2327767275022096e-06,
+      "loss": 0.6235,
+      "step": 4453
+    },
+    {
+      "epoch": 1.5639044943820224,
+      "grad_norm": 0.8793233036994934,
+      "learning_rate": 4.232443439413117e-06,
+      "loss": 0.6551,
+      "step": 4454
+    },
+    {
+      "epoch": 1.564255617977528,
+      "grad_norm": 0.8933106064796448,
+      "learning_rate": 4.232110092075941e-06,
+      "loss": 0.6277,
+      "step": 4455
+    },
+    {
+      "epoch": 1.5646067415730336,
+      "grad_norm": 0.8762953281402588,
+      "learning_rate": 4.231776685502083e-06,
+      "loss": 0.6052,
+      "step": 4456
+    },
+    {
+      "epoch": 1.5649578651685392,
+      "grad_norm": 0.8579549193382263,
+      "learning_rate": 4.231443219702943e-06,
+      "loss": 0.5982,
+      "step": 4457
+    },
+    {
+      "epoch": 1.5653089887640448,
+      "grad_norm": 0.8534015417098999,
+      "learning_rate": 4.231109694689927e-06,
+      "loss": 0.6071,
+      "step": 4458
+    },
+    {
+      "epoch": 1.5656601123595506,
+      "grad_norm": 0.8755966424942017,
+      "learning_rate": 4.23077611047444e-06,
+      "loss": 0.5972,
+      "step": 4459
+    },
+    {
+      "epoch": 1.5660112359550562,
+      "grad_norm": 0.8907667398452759,
+      "learning_rate": 4.230442467067892e-06,
+      "loss": 0.6195,
+      "step": 4460
+    },
+    {
+      "epoch": 1.5663623595505618,
+      "grad_norm": 0.8510807156562805,
+      "learning_rate": 4.230108764481693e-06,
+      "loss": 0.6359,
+      "step": 4461
+    },
+    {
+      "epoch": 1.5667134831460674,
+      "grad_norm": 0.8813369274139404,
+      "learning_rate": 4.2297750027272535e-06,
+      "loss": 0.6517,
+      "step": 4462
+    },
+    {
+      "epoch": 1.567064606741573,
+      "grad_norm": 0.8754818439483643,
+      "learning_rate": 4.22944118181599e-06,
+      "loss": 0.633,
+      "step": 4463
+    },
+    {
+      "epoch": 1.5674157303370788,
+      "grad_norm": 0.8787099719047546,
+      "learning_rate": 4.229107301759317e-06,
+      "loss": 0.6439,
+      "step": 4464
+    },
+    {
+      "epoch": 1.5677668539325844,
+      "grad_norm": 0.8702832460403442,
+      "learning_rate": 4.2287733625686545e-06,
+      "loss": 0.6221,
+      "step": 4465
+    },
+    {
+      "epoch": 1.56811797752809,
+      "grad_norm": 0.8664124011993408,
+      "learning_rate": 4.228439364255422e-06,
+      "loss": 0.6139,
+      "step": 4466
+    },
+    {
+      "epoch": 1.5684691011235956,
+      "grad_norm": 0.8977587819099426,
+      "learning_rate": 4.228105306831043e-06,
+      "loss": 0.6185,
+      "step": 4467
+    },
+    {
+      "epoch": 1.5688202247191012,
+      "grad_norm": 0.8700708746910095,
+      "learning_rate": 4.2277711903069405e-06,
+      "loss": 0.6421,
+      "step": 4468
+    },
+    {
+      "epoch": 1.5691713483146068,
+      "grad_norm": 0.8524662256240845,
+      "learning_rate": 4.227437014694542e-06,
+      "loss": 0.6432,
+      "step": 4469
+    },
+    {
+      "epoch": 1.5695224719101124,
+      "grad_norm": 0.8722302317619324,
+      "learning_rate": 4.227102780005275e-06,
+      "loss": 0.6245,
+      "step": 4470
+    },
+    {
+      "epoch": 1.569873595505618,
+      "grad_norm": 0.879640519618988,
+      "learning_rate": 4.226768486250573e-06,
+      "loss": 0.6451,
+      "step": 4471
+    },
+    {
+      "epoch": 1.5702247191011236,
+      "grad_norm": 0.855945348739624,
+      "learning_rate": 4.226434133441865e-06,
+      "loss": 0.6301,
+      "step": 4472
+    },
+    {
+      "epoch": 1.5705758426966292,
+      "grad_norm": 0.8772982954978943,
+      "learning_rate": 4.226099721590586e-06,
+      "loss": 0.6495,
+      "step": 4473
+    },
+    {
+      "epoch": 1.5709269662921348,
+      "grad_norm": 0.9376111626625061,
+      "learning_rate": 4.225765250708175e-06,
+      "loss": 0.6047,
+      "step": 4474
+    },
+    {
+      "epoch": 1.5712780898876404,
+      "grad_norm": 0.8725435733795166,
+      "learning_rate": 4.225430720806069e-06,
+      "loss": 0.634,
+      "step": 4475
+    },
+    {
+      "epoch": 1.571629213483146,
+      "grad_norm": 0.8824449181556702,
+      "learning_rate": 4.225096131895709e-06,
+      "loss": 0.6396,
+      "step": 4476
+    },
+    {
+      "epoch": 1.5719803370786516,
+      "grad_norm": 0.8654618859291077,
+      "learning_rate": 4.2247614839885365e-06,
+      "loss": 0.6144,
+      "step": 4477
+    },
+    {
+      "epoch": 1.5723314606741572,
+      "grad_norm": 0.8552867770195007,
+      "learning_rate": 4.224426777095998e-06,
+      "loss": 0.5836,
+      "step": 4478
+    },
+    {
+      "epoch": 1.5726825842696628,
+      "grad_norm": 0.8514997363090515,
+      "learning_rate": 4.224092011229539e-06,
+      "loss": 0.5915,
+      "step": 4479
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.8661342859268188,
+      "learning_rate": 4.223757186400609e-06,
+      "loss": 0.6288,
+      "step": 4480
+    },
+    {
+      "epoch": 1.5733848314606742,
+      "grad_norm": 0.8661715984344482,
+      "learning_rate": 4.223422302620657e-06,
+      "loss": 0.6201,
+      "step": 4481
+    },
+    {
+      "epoch": 1.5737359550561798,
+      "grad_norm": 0.8975440859794617,
+      "learning_rate": 4.223087359901138e-06,
+      "loss": 0.7046,
+      "step": 4482
+    },
+    {
+      "epoch": 1.5740870786516854,
+      "grad_norm": 0.8667924404144287,
+      "learning_rate": 4.222752358253506e-06,
+      "loss": 0.6245,
+      "step": 4483
+    },
+    {
+      "epoch": 1.574438202247191,
+      "grad_norm": 0.8844180107116699,
+      "learning_rate": 4.222417297689217e-06,
+      "loss": 0.6172,
+      "step": 4484
+    },
+    {
+      "epoch": 1.5747893258426966,
+      "grad_norm": 0.8499074578285217,
+      "learning_rate": 4.22208217821973e-06,
+      "loss": 0.5952,
+      "step": 4485
+    },
+    {
+      "epoch": 1.5751404494382022,
+      "grad_norm": 0.9003639221191406,
+      "learning_rate": 4.221746999856507e-06,
+      "loss": 0.6488,
+      "step": 4486
+    },
+    {
+      "epoch": 1.575491573033708,
+      "grad_norm": 0.8823853731155396,
+      "learning_rate": 4.22141176261101e-06,
+      "loss": 0.6171,
+      "step": 4487
+    },
+    {
+      "epoch": 1.5758426966292136,
+      "grad_norm": 0.8788625597953796,
+      "learning_rate": 4.221076466494702e-06,
+      "loss": 0.6026,
+      "step": 4488
+    },
+    {
+      "epoch": 1.5761938202247192,
+      "grad_norm": 0.8390807509422302,
+      "learning_rate": 4.220741111519053e-06,
+      "loss": 0.5835,
+      "step": 4489
+    },
+    {
+      "epoch": 1.5765449438202248,
+      "grad_norm": 0.8773233294487,
+      "learning_rate": 4.220405697695531e-06,
+      "loss": 0.644,
+      "step": 4490
+    },
+    {
+      "epoch": 1.5768960674157304,
+      "grad_norm": 0.8999008536338806,
+      "learning_rate": 4.220070225035605e-06,
+      "loss": 0.6453,
+      "step": 4491
+    },
+    {
+      "epoch": 1.577247191011236,
+      "grad_norm": 0.8900070190429688,
+      "learning_rate": 4.21973469355075e-06,
+      "loss": 0.6586,
+      "step": 4492
+    },
+    {
+      "epoch": 1.5775983146067416,
+      "grad_norm": 0.8822365403175354,
+      "learning_rate": 4.21939910325244e-06,
+      "loss": 0.639,
+      "step": 4493
+    },
+    {
+      "epoch": 1.5779494382022472,
+      "grad_norm": 0.8647001385688782,
+      "learning_rate": 4.219063454152152e-06,
+      "loss": 0.6038,
+      "step": 4494
+    },
+    {
+      "epoch": 1.5783005617977528,
+      "grad_norm": 0.8594561219215393,
+      "learning_rate": 4.218727746261364e-06,
+      "loss": 0.6143,
+      "step": 4495
+    },
+    {
+      "epoch": 1.5786516853932584,
+      "grad_norm": 0.8780025243759155,
+      "learning_rate": 4.218391979591559e-06,
+      "loss": 0.6539,
+      "step": 4496
+    },
+    {
+      "epoch": 1.579002808988764,
+      "grad_norm": 0.8385159373283386,
+      "learning_rate": 4.218056154154219e-06,
+      "loss": 0.6088,
+      "step": 4497
+    },
+    {
+      "epoch": 1.5793539325842696,
+      "grad_norm": 0.8734780550003052,
+      "learning_rate": 4.217720269960828e-06,
+      "loss": 0.6463,
+      "step": 4498
+    },
+    {
+      "epoch": 1.5797050561797752,
+      "grad_norm": 0.8894116282463074,
+      "learning_rate": 4.217384327022873e-06,
+      "loss": 0.623,
+      "step": 4499
+    },
+    {
+      "epoch": 1.5800561797752808,
+      "grad_norm": 0.9236807227134705,
+      "learning_rate": 4.217048325351844e-06,
+      "loss": 0.6192,
+      "step": 4500
+    },
+    {
+      "epoch": 1.5804073033707864,
+      "grad_norm": 0.8841096758842468,
+      "learning_rate": 4.216712264959232e-06,
+      "loss": 0.6471,
+      "step": 4501
+    },
+    {
+      "epoch": 1.580758426966292,
+      "grad_norm": 0.8719488978385925,
+      "learning_rate": 4.216376145856529e-06,
+      "loss": 0.6733,
+      "step": 4502
+    },
+    {
+      "epoch": 1.5811095505617978,
+      "grad_norm": 0.8742629289627075,
+      "learning_rate": 4.216039968055231e-06,
+      "loss": 0.6004,
+      "step": 4503
+    },
+    {
+      "epoch": 1.5814606741573034,
+      "grad_norm": 0.8602683544158936,
+      "learning_rate": 4.215703731566835e-06,
+      "loss": 0.6221,
+      "step": 4504
+    },
+    {
+      "epoch": 1.581811797752809,
+      "grad_norm": 0.8487911224365234,
+      "learning_rate": 4.215367436402838e-06,
+      "loss": 0.6398,
+      "step": 4505
+    },
+    {
+      "epoch": 1.5821629213483146,
+      "grad_norm": 0.8509633541107178,
+      "learning_rate": 4.215031082574744e-06,
+      "loss": 0.593,
+      "step": 4506
+    },
+    {
+      "epoch": 1.5825140449438202,
+      "grad_norm": 0.9125469923019409,
+      "learning_rate": 4.2146946700940545e-06,
+      "loss": 0.6602,
+      "step": 4507
+    },
+    {
+      "epoch": 1.5828651685393258,
+      "grad_norm": 0.908766508102417,
+      "learning_rate": 4.2143581989722736e-06,
+      "loss": 0.6807,
+      "step": 4508
+    },
+    {
+      "epoch": 1.5832162921348316,
+      "grad_norm": 0.893795371055603,
+      "learning_rate": 4.21402166922091e-06,
+      "loss": 0.6677,
+      "step": 4509
+    },
+    {
+      "epoch": 1.5835674157303372,
+      "grad_norm": 0.8966583609580994,
+      "learning_rate": 4.2136850808514715e-06,
+      "loss": 0.6459,
+      "step": 4510
+    },
+    {
+      "epoch": 1.5839185393258428,
+      "grad_norm": 0.8950080871582031,
+      "learning_rate": 4.21334843387547e-06,
+      "loss": 0.6565,
+      "step": 4511
+    },
+    {
+      "epoch": 1.5842696629213484,
+      "grad_norm": 0.85506671667099,
+      "learning_rate": 4.213011728304418e-06,
+      "loss": 0.6113,
+      "step": 4512
+    },
+    {
+      "epoch": 1.584620786516854,
+      "grad_norm": 0.8737542629241943,
+      "learning_rate": 4.212674964149831e-06,
+      "loss": 0.6275,
+      "step": 4513
+    },
+    {
+      "epoch": 1.5849719101123596,
+      "grad_norm": 0.882484495639801,
+      "learning_rate": 4.212338141423226e-06,
+      "loss": 0.6476,
+      "step": 4514
+    },
+    {
+      "epoch": 1.5853230337078652,
+      "grad_norm": 0.8486616015434265,
+      "learning_rate": 4.212001260136121e-06,
+      "loss": 0.5877,
+      "step": 4515
+    },
+    {
+      "epoch": 1.5856741573033708,
+      "grad_norm": 0.8440214991569519,
+      "learning_rate": 4.2116643203000386e-06,
+      "loss": 0.5961,
+      "step": 4516
+    },
+    {
+      "epoch": 1.5860252808988764,
+      "grad_norm": 0.8951239585876465,
+      "learning_rate": 4.211327321926501e-06,
+      "loss": 0.6463,
+      "step": 4517
+    },
+    {
+      "epoch": 1.586376404494382,
+      "grad_norm": 0.8931921124458313,
+      "learning_rate": 4.210990265027034e-06,
+      "loss": 0.6554,
+      "step": 4518
+    },
+    {
+      "epoch": 1.5867275280898876,
+      "grad_norm": 0.8429591655731201,
+      "learning_rate": 4.210653149613163e-06,
+      "loss": 0.5997,
+      "step": 4519
+    },
+    {
+      "epoch": 1.5870786516853932,
+      "grad_norm": 0.8799921274185181,
+      "learning_rate": 4.210315975696419e-06,
+      "loss": 0.6236,
+      "step": 4520
+    },
+    {
+      "epoch": 1.5874297752808988,
+      "grad_norm": 0.8685789108276367,
+      "learning_rate": 4.209978743288331e-06,
+      "loss": 0.6138,
+      "step": 4521
+    },
+    {
+      "epoch": 1.5877808988764044,
+      "grad_norm": 0.8596302270889282,
+      "learning_rate": 4.2096414524004335e-06,
+      "loss": 0.5799,
+      "step": 4522
+    },
+    {
+      "epoch": 1.58813202247191,
+      "grad_norm": 0.8389763832092285,
+      "learning_rate": 4.209304103044262e-06,
+      "loss": 0.6066,
+      "step": 4523
+    },
+    {
+      "epoch": 1.5884831460674156,
+      "grad_norm": 0.8625580072402954,
+      "learning_rate": 4.2089666952313525e-06,
+      "loss": 0.6263,
+      "step": 4524
+    },
+    {
+      "epoch": 1.5888342696629212,
+      "grad_norm": 0.8957053422927856,
+      "learning_rate": 4.208629228973245e-06,
+      "loss": 0.6237,
+      "step": 4525
+    },
+    {
+      "epoch": 1.589185393258427,
+      "grad_norm": 0.8905929327011108,
+      "learning_rate": 4.208291704281479e-06,
+      "loss": 0.6389,
+      "step": 4526
+    },
+    {
+      "epoch": 1.5895365168539326,
+      "grad_norm": 0.9022085070610046,
+      "learning_rate": 4.207954121167599e-06,
+      "loss": 0.62,
+      "step": 4527
+    },
+    {
+      "epoch": 1.5898876404494382,
+      "grad_norm": 0.8556004762649536,
+      "learning_rate": 4.20761647964315e-06,
+      "loss": 0.5949,
+      "step": 4528
+    },
+    {
+      "epoch": 1.5902387640449438,
+      "grad_norm": 0.8509926795959473,
+      "learning_rate": 4.207278779719677e-06,
+      "loss": 0.603,
+      "step": 4529
+    },
+    {
+      "epoch": 1.5905898876404494,
+      "grad_norm": 0.8614581227302551,
+      "learning_rate": 4.206941021408732e-06,
+      "loss": 0.6292,
+      "step": 4530
+    },
+    {
+      "epoch": 1.5909410112359552,
+      "grad_norm": 0.8607187867164612,
+      "learning_rate": 4.206603204721865e-06,
+      "loss": 0.5729,
+      "step": 4531
+    },
+    {
+      "epoch": 1.5912921348314608,
+      "grad_norm": 0.9069685339927673,
+      "learning_rate": 4.206265329670628e-06,
+      "loss": 0.6275,
+      "step": 4532
+    },
+    {
+      "epoch": 1.5916432584269664,
+      "grad_norm": 0.9023653268814087,
+      "learning_rate": 4.205927396266577e-06,
+      "loss": 0.6463,
+      "step": 4533
+    },
+    {
+      "epoch": 1.591994382022472,
+      "grad_norm": 0.9172576665878296,
+      "learning_rate": 4.205589404521269e-06,
+      "loss": 0.6653,
+      "step": 4534
+    },
+    {
+      "epoch": 1.5923455056179776,
+      "grad_norm": 0.8926611542701721,
+      "learning_rate": 4.205251354446262e-06,
+      "loss": 0.6522,
+      "step": 4535
+    },
+    {
+      "epoch": 1.5926966292134832,
+      "grad_norm": 0.8904423713684082,
+      "learning_rate": 4.204913246053119e-06,
+      "loss": 0.6635,
+      "step": 4536
+    },
+    {
+      "epoch": 1.5930477528089888,
+      "grad_norm": 0.876338005065918,
+      "learning_rate": 4.204575079353403e-06,
+      "loss": 0.628,
+      "step": 4537
+    },
+    {
+      "epoch": 1.5933988764044944,
+      "grad_norm": 0.8750518560409546,
+      "learning_rate": 4.204236854358676e-06,
+      "loss": 0.6365,
+      "step": 4538
+    },
+    {
+      "epoch": 1.59375,
+      "grad_norm": 0.8659214973449707,
+      "learning_rate": 4.203898571080507e-06,
+      "loss": 0.6157,
+      "step": 4539
+    },
+    {
+      "epoch": 1.5941011235955056,
+      "grad_norm": 0.869689404964447,
+      "learning_rate": 4.203560229530465e-06,
+      "loss": 0.605,
+      "step": 4540
+    },
+    {
+      "epoch": 1.5944522471910112,
+      "grad_norm": 0.8688238859176636,
+      "learning_rate": 4.203221829720121e-06,
+      "loss": 0.6453,
+      "step": 4541
+    },
+    {
+      "epoch": 1.5948033707865168,
+      "grad_norm": 0.8444859385490417,
+      "learning_rate": 4.202883371661048e-06,
+      "loss": 0.6125,
+      "step": 4542
+    },
+    {
+      "epoch": 1.5951544943820224,
+      "grad_norm": 0.881079912185669,
+      "learning_rate": 4.202544855364822e-06,
+      "loss": 0.6082,
+      "step": 4543
+    },
+    {
+      "epoch": 1.595505617977528,
+      "grad_norm": 0.8665393590927124,
+      "learning_rate": 4.202206280843017e-06,
+      "loss": 0.6304,
+      "step": 4544
+    },
+    {
+      "epoch": 1.5958567415730336,
+      "grad_norm": 0.8872676491737366,
+      "learning_rate": 4.201867648107215e-06,
+      "loss": 0.6206,
+      "step": 4545
+    },
+    {
+      "epoch": 1.5962078651685392,
+      "grad_norm": 0.8978577852249146,
+      "learning_rate": 4.201528957168995e-06,
+      "loss": 0.6471,
+      "step": 4546
+    },
+    {
+      "epoch": 1.5965589887640448,
+      "grad_norm": 0.9299441576004028,
+      "learning_rate": 4.201190208039942e-06,
+      "loss": 0.6454,
+      "step": 4547
+    },
+    {
+      "epoch": 1.5969101123595506,
+      "grad_norm": 0.8918309807777405,
+      "learning_rate": 4.200851400731638e-06,
+      "loss": 0.6372,
+      "step": 4548
+    },
+    {
+      "epoch": 1.5972612359550562,
+      "grad_norm": 0.8987869024276733,
+      "learning_rate": 4.200512535255672e-06,
+      "loss": 0.6298,
+      "step": 4549
+    },
+    {
+      "epoch": 1.5976123595505618,
+      "grad_norm": 0.8805652260780334,
+      "learning_rate": 4.200173611623632e-06,
+      "loss": 0.6621,
+      "step": 4550
+    },
+    {
+      "epoch": 1.5979634831460674,
+      "grad_norm": 0.8915992975234985,
+      "learning_rate": 4.19983462984711e-06,
+      "loss": 0.65,
+      "step": 4551
+    },
+    {
+      "epoch": 1.598314606741573,
+      "grad_norm": 0.8713775873184204,
+      "learning_rate": 4.199495589937699e-06,
+      "loss": 0.6122,
+      "step": 4552
+    },
+    {
+      "epoch": 1.5986657303370788,
+      "grad_norm": 0.8860073089599609,
+      "learning_rate": 4.199156491906991e-06,
+      "loss": 0.653,
+      "step": 4553
+    },
+    {
+      "epoch": 1.5990168539325844,
+      "grad_norm": 0.8860781788825989,
+      "learning_rate": 4.198817335766587e-06,
+      "loss": 0.6147,
+      "step": 4554
+    },
+    {
+      "epoch": 1.59936797752809,
+      "grad_norm": 0.8467510938644409,
+      "learning_rate": 4.198478121528084e-06,
+      "loss": 0.6153,
+      "step": 4555
+    },
+    {
+      "epoch": 1.5997191011235956,
+      "grad_norm": 0.8688376545906067,
+      "learning_rate": 4.1981388492030815e-06,
+      "loss": 0.6415,
+      "step": 4556
+    },
+    {
+      "epoch": 1.6000702247191012,
+      "grad_norm": 0.8537015318870544,
+      "learning_rate": 4.197799518803184e-06,
+      "loss": 0.6312,
+      "step": 4557
+    },
+    {
+      "epoch": 1.6004213483146068,
+      "grad_norm": 0.883955717086792,
+      "learning_rate": 4.197460130339995e-06,
+      "loss": 0.6212,
+      "step": 4558
+    },
+    {
+      "epoch": 1.6007724719101124,
+      "grad_norm": 0.8804914355278015,
+      "learning_rate": 4.1971206838251225e-06,
+      "loss": 0.6224,
+      "step": 4559
+    },
+    {
+      "epoch": 1.601123595505618,
+      "grad_norm": 0.8551623225212097,
+      "learning_rate": 4.1967811792701745e-06,
+      "loss": 0.6037,
+      "step": 4560
+    },
+    {
+      "epoch": 1.6014747191011236,
+      "grad_norm": 0.9019951820373535,
+      "learning_rate": 4.196441616686763e-06,
+      "loss": 0.6206,
+      "step": 4561
+    },
+    {
+      "epoch": 1.6018258426966292,
+      "grad_norm": 0.8744482398033142,
+      "learning_rate": 4.196101996086499e-06,
+      "loss": 0.6206,
+      "step": 4562
+    },
+    {
+      "epoch": 1.6021769662921348,
+      "grad_norm": 0.8476253747940063,
+      "learning_rate": 4.195762317480998e-06,
+      "loss": 0.5789,
+      "step": 4563
+    },
+    {
+      "epoch": 1.6025280898876404,
+      "grad_norm": 0.8534361720085144,
+      "learning_rate": 4.195422580881878e-06,
+      "loss": 0.6131,
+      "step": 4564
+    },
+    {
+      "epoch": 1.602879213483146,
+      "grad_norm": 0.8521122336387634,
+      "learning_rate": 4.195082786300756e-06,
+      "loss": 0.6133,
+      "step": 4565
+    },
+    {
+      "epoch": 1.6032303370786516,
+      "grad_norm": 0.8880335092544556,
+      "learning_rate": 4.194742933749253e-06,
+      "loss": 0.6433,
+      "step": 4566
+    },
+    {
+      "epoch": 1.6035814606741572,
+      "grad_norm": 0.8391385674476624,
+      "learning_rate": 4.194403023238992e-06,
+      "loss": 0.5859,
+      "step": 4567
+    },
+    {
+      "epoch": 1.6039325842696628,
+      "grad_norm": 0.8838050961494446,
+      "learning_rate": 4.194063054781597e-06,
+      "loss": 0.5979,
+      "step": 4568
+    },
+    {
+      "epoch": 1.6042837078651684,
+      "grad_norm": 0.8805633783340454,
+      "learning_rate": 4.193723028388696e-06,
+      "loss": 0.6112,
+      "step": 4569
+    },
+    {
+      "epoch": 1.6046348314606742,
+      "grad_norm": 0.8977530598640442,
+      "learning_rate": 4.193382944071917e-06,
+      "loss": 0.6199,
+      "step": 4570
+    },
+    {
+      "epoch": 1.6049859550561798,
+      "grad_norm": 0.8989500403404236,
+      "learning_rate": 4.193042801842889e-06,
+      "loss": 0.6176,
+      "step": 4571
+    },
+    {
+      "epoch": 1.6053370786516854,
+      "grad_norm": 0.9075813889503479,
+      "learning_rate": 4.192702601713247e-06,
+      "loss": 0.6429,
+      "step": 4572
+    },
+    {
+      "epoch": 1.605688202247191,
+      "grad_norm": 0.9065417051315308,
+      "learning_rate": 4.192362343694624e-06,
+      "loss": 0.6387,
+      "step": 4573
+    },
+    {
+      "epoch": 1.6060393258426966,
+      "grad_norm": 0.8901644349098206,
+      "learning_rate": 4.1920220277986574e-06,
+      "loss": 0.6332,
+      "step": 4574
+    },
+    {
+      "epoch": 1.6063904494382022,
+      "grad_norm": 0.8711532354354858,
+      "learning_rate": 4.191681654036985e-06,
+      "loss": 0.6324,
+      "step": 4575
+    },
+    {
+      "epoch": 1.606741573033708,
+      "grad_norm": 0.8672968745231628,
+      "learning_rate": 4.191341222421248e-06,
+      "loss": 0.6165,
+      "step": 4576
+    },
+    {
+      "epoch": 1.6070926966292136,
+      "grad_norm": 0.8829415440559387,
+      "learning_rate": 4.191000732963088e-06,
+      "loss": 0.6228,
+      "step": 4577
+    },
+    {
+      "epoch": 1.6074438202247192,
+      "grad_norm": 0.8676837086677551,
+      "learning_rate": 4.19066018567415e-06,
+      "loss": 0.6452,
+      "step": 4578
+    },
+    {
+      "epoch": 1.6077949438202248,
+      "grad_norm": 0.8773974180221558,
+      "learning_rate": 4.19031958056608e-06,
+      "loss": 0.674,
+      "step": 4579
+    },
+    {
+      "epoch": 1.6081460674157304,
+      "grad_norm": 0.8626821041107178,
+      "learning_rate": 4.189978917650528e-06,
+      "loss": 0.6104,
+      "step": 4580
+    },
+    {
+      "epoch": 1.608497191011236,
+      "grad_norm": 0.8596952557563782,
+      "learning_rate": 4.189638196939143e-06,
+      "loss": 0.63,
+      "step": 4581
+    },
+    {
+      "epoch": 1.6088483146067416,
+      "grad_norm": 0.880653977394104,
+      "learning_rate": 4.189297418443576e-06,
+      "loss": 0.6064,
+      "step": 4582
+    },
+    {
+      "epoch": 1.6091994382022472,
+      "grad_norm": 0.8691625595092773,
+      "learning_rate": 4.188956582175484e-06,
+      "loss": 0.6144,
+      "step": 4583
+    },
+    {
+      "epoch": 1.6095505617977528,
+      "grad_norm": 0.8744456171989441,
+      "learning_rate": 4.1886156881465225e-06,
+      "loss": 0.5914,
+      "step": 4584
+    },
+    {
+      "epoch": 1.6099016853932584,
+      "grad_norm": 0.9076862931251526,
+      "learning_rate": 4.188274736368349e-06,
+      "loss": 0.5998,
+      "step": 4585
+    },
+    {
+      "epoch": 1.610252808988764,
+      "grad_norm": 0.8850449323654175,
+      "learning_rate": 4.187933726852624e-06,
+      "loss": 0.613,
+      "step": 4586
+    },
+    {
+      "epoch": 1.6106039325842696,
+      "grad_norm": 0.875081479549408,
+      "learning_rate": 4.187592659611011e-06,
+      "loss": 0.5947,
+      "step": 4587
+    },
+    {
+      "epoch": 1.6109550561797752,
+      "grad_norm": 0.9096956849098206,
+      "learning_rate": 4.187251534655171e-06,
+      "loss": 0.6624,
+      "step": 4588
+    },
+    {
+      "epoch": 1.6113061797752808,
+      "grad_norm": 0.8888775110244751,
+      "learning_rate": 4.1869103519967745e-06,
+      "loss": 0.634,
+      "step": 4589
+    },
+    {
+      "epoch": 1.6116573033707864,
+      "grad_norm": 0.8596838712692261,
+      "learning_rate": 4.186569111647486e-06,
+      "loss": 0.6214,
+      "step": 4590
+    },
+    {
+      "epoch": 1.612008426966292,
+      "grad_norm": 0.8774394392967224,
+      "learning_rate": 4.186227813618977e-06,
+      "loss": 0.6341,
+      "step": 4591
+    },
+    {
+      "epoch": 1.6123595505617978,
+      "grad_norm": 0.902068018913269,
+      "learning_rate": 4.185886457922921e-06,
+      "loss": 0.6276,
+      "step": 4592
+    },
+    {
+      "epoch": 1.6127106741573034,
+      "grad_norm": 0.8684345483779907,
+      "learning_rate": 4.18554504457099e-06,
+      "loss": 0.5965,
+      "step": 4593
+    },
+    {
+      "epoch": 1.613061797752809,
+      "grad_norm": 0.8846623301506042,
+      "learning_rate": 4.185203573574861e-06,
+      "loss": 0.6202,
+      "step": 4594
+    },
+    {
+      "epoch": 1.6134129213483146,
+      "grad_norm": 0.8909761309623718,
+      "learning_rate": 4.1848620449462115e-06,
+      "loss": 0.6117,
+      "step": 4595
+    },
+    {
+      "epoch": 1.6137640449438202,
+      "grad_norm": 0.9091597199440002,
+      "learning_rate": 4.1845204586967225e-06,
+      "loss": 0.6311,
+      "step": 4596
+    },
+    {
+      "epoch": 1.6141151685393258,
+      "grad_norm": 0.8552021384239197,
+      "learning_rate": 4.184178814838075e-06,
+      "loss": 0.5772,
+      "step": 4597
+    },
+    {
+      "epoch": 1.6144662921348316,
+      "grad_norm": 0.8852929472923279,
+      "learning_rate": 4.1838371133819525e-06,
+      "loss": 0.6538,
+      "step": 4598
+    },
+    {
+      "epoch": 1.6148174157303372,
+      "grad_norm": 1.1475235223770142,
+      "learning_rate": 4.183495354340042e-06,
+      "loss": 0.6104,
+      "step": 4599
+    },
+    {
+      "epoch": 1.6151685393258428,
+      "grad_norm": 0.8915920257568359,
+      "learning_rate": 4.1831535377240306e-06,
+      "loss": 0.5814,
+      "step": 4600
+    },
+    {
+      "epoch": 1.6155196629213484,
+      "grad_norm": 0.8609941601753235,
+      "learning_rate": 4.182811663545609e-06,
+      "loss": 0.591,
+      "step": 4601
+    },
+    {
+      "epoch": 1.615870786516854,
+      "grad_norm": 0.8900454640388489,
+      "learning_rate": 4.182469731816468e-06,
+      "loss": 0.611,
+      "step": 4602
+    },
+    {
+      "epoch": 1.6162219101123596,
+      "grad_norm": 0.8634515404701233,
+      "learning_rate": 4.182127742548303e-06,
+      "loss": 0.5957,
+      "step": 4603
+    },
+    {
+      "epoch": 1.6165730337078652,
+      "grad_norm": 0.9392664432525635,
+      "learning_rate": 4.181785695752808e-06,
+      "loss": 0.6798,
+      "step": 4604
+    },
+    {
+      "epoch": 1.6169241573033708,
+      "grad_norm": 0.8867204189300537,
+      "learning_rate": 4.181443591441682e-06,
+      "loss": 0.6288,
+      "step": 4605
+    },
+    {
+      "epoch": 1.6172752808988764,
+      "grad_norm": 0.869739294052124,
+      "learning_rate": 4.181101429626623e-06,
+      "loss": 0.641,
+      "step": 4606
+    },
+    {
+      "epoch": 1.617626404494382,
+      "grad_norm": 0.8330792784690857,
+      "learning_rate": 4.180759210319334e-06,
+      "loss": 0.5901,
+      "step": 4607
+    },
+    {
+      "epoch": 1.6179775280898876,
+      "grad_norm": 0.8733412623405457,
+      "learning_rate": 4.180416933531518e-06,
+      "loss": 0.6115,
+      "step": 4608
+    },
+    {
+      "epoch": 1.6183286516853932,
+      "grad_norm": 0.8878810405731201,
+      "learning_rate": 4.1800745992748815e-06,
+      "loss": 0.6408,
+      "step": 4609
+    },
+    {
+      "epoch": 1.6186797752808988,
+      "grad_norm": 0.8582045435905457,
+      "learning_rate": 4.179732207561131e-06,
+      "loss": 0.6173,
+      "step": 4610
+    },
+    {
+      "epoch": 1.6190308988764044,
+      "grad_norm": 0.8719291687011719,
+      "learning_rate": 4.179389758401976e-06,
+      "loss": 0.6223,
+      "step": 4611
+    },
+    {
+      "epoch": 1.61938202247191,
+      "grad_norm": 0.8811583518981934,
+      "learning_rate": 4.179047251809129e-06,
+      "loss": 0.6012,
+      "step": 4612
+    },
+    {
+      "epoch": 1.6197331460674156,
+      "grad_norm": 0.9122189283370972,
+      "learning_rate": 4.178704687794302e-06,
+      "loss": 0.6253,
+      "step": 4613
+    },
+    {
+      "epoch": 1.6200842696629212,
+      "grad_norm": 0.8523516654968262,
+      "learning_rate": 4.178362066369211e-06,
+      "loss": 0.6133,
+      "step": 4614
+    },
+    {
+      "epoch": 1.620435393258427,
+      "grad_norm": 0.9010724425315857,
+      "learning_rate": 4.1780193875455745e-06,
+      "loss": 0.6225,
+      "step": 4615
+    },
+    {
+      "epoch": 1.6207865168539326,
+      "grad_norm": 0.8976810574531555,
+      "learning_rate": 4.177676651335111e-06,
+      "loss": 0.6051,
+      "step": 4616
+    },
+    {
+      "epoch": 1.6211376404494382,
+      "grad_norm": 0.8751901388168335,
+      "learning_rate": 4.17733385774954e-06,
+      "loss": 0.6189,
+      "step": 4617
+    },
+    {
+      "epoch": 1.6214887640449438,
+      "grad_norm": 0.8850089311599731,
+      "learning_rate": 4.176991006800588e-06,
+      "loss": 0.6241,
+      "step": 4618
+    },
+    {
+      "epoch": 1.6218398876404494,
+      "grad_norm": 0.8844864964485168,
+      "learning_rate": 4.176648098499978e-06,
+      "loss": 0.605,
+      "step": 4619
+    },
+    {
+      "epoch": 1.6221910112359552,
+      "grad_norm": 0.884218156337738,
+      "learning_rate": 4.176305132859438e-06,
+      "loss": 0.6216,
+      "step": 4620
+    },
+    {
+      "epoch": 1.6225421348314608,
+      "grad_norm": 0.8722840547561646,
+      "learning_rate": 4.175962109890697e-06,
+      "loss": 0.6316,
+      "step": 4621
+    },
+    {
+      "epoch": 1.6228932584269664,
+      "grad_norm": 0.8568814992904663,
+      "learning_rate": 4.175619029605485e-06,
+      "loss": 0.6037,
+      "step": 4622
+    },
+    {
+      "epoch": 1.623244382022472,
+      "grad_norm": 0.8812152743339539,
+      "learning_rate": 4.175275892015537e-06,
+      "loss": 0.5933,
+      "step": 4623
+    },
+    {
+      "epoch": 1.6235955056179776,
+      "grad_norm": 0.8655858039855957,
+      "learning_rate": 4.174932697132587e-06,
+      "loss": 0.6138,
+      "step": 4624
+    },
+    {
+      "epoch": 1.6239466292134832,
+      "grad_norm": 0.8626354932785034,
+      "learning_rate": 4.174589444968371e-06,
+      "loss": 0.6076,
+      "step": 4625
+    },
+    {
+      "epoch": 1.6242977528089888,
+      "grad_norm": 0.8693781495094299,
+      "learning_rate": 4.17424613553463e-06,
+      "loss": 0.6139,
+      "step": 4626
+    },
+    {
+      "epoch": 1.6246488764044944,
+      "grad_norm": 0.9022204875946045,
+      "learning_rate": 4.173902768843104e-06,
+      "loss": 0.6254,
+      "step": 4627
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.8757764101028442,
+      "learning_rate": 4.173559344905535e-06,
+      "loss": 0.5988,
+      "step": 4628
+    },
+    {
+      "epoch": 1.6253511235955056,
+      "grad_norm": 0.8923527002334595,
+      "learning_rate": 4.173215863733669e-06,
+      "loss": 0.6333,
+      "step": 4629
+    },
+    {
+      "epoch": 1.6257022471910112,
+      "grad_norm": 0.8588404059410095,
+      "learning_rate": 4.172872325339252e-06,
+      "loss": 0.6117,
+      "step": 4630
+    },
+    {
+      "epoch": 1.6260533707865168,
+      "grad_norm": 0.8929343819618225,
+      "learning_rate": 4.1725287297340335e-06,
+      "loss": 0.6244,
+      "step": 4631
+    },
+    {
+      "epoch": 1.6264044943820224,
+      "grad_norm": 0.8902310132980347,
+      "learning_rate": 4.172185076929763e-06,
+      "loss": 0.6019,
+      "step": 4632
+    },
+    {
+      "epoch": 1.626755617977528,
+      "grad_norm": 0.8931180238723755,
+      "learning_rate": 4.171841366938193e-06,
+      "loss": 0.6306,
+      "step": 4633
+    },
+    {
+      "epoch": 1.6271067415730336,
+      "grad_norm": 0.8756839632987976,
+      "learning_rate": 4.1714975997710804e-06,
+      "loss": 0.6456,
+      "step": 4634
+    },
+    {
+      "epoch": 1.6274578651685392,
+      "grad_norm": 0.8530634045600891,
+      "learning_rate": 4.171153775440179e-06,
+      "loss": 0.5906,
+      "step": 4635
+    },
+    {
+      "epoch": 1.6278089887640448,
+      "grad_norm": 0.8859505653381348,
+      "learning_rate": 4.170809893957249e-06,
+      "loss": 0.6395,
+      "step": 4636
+    },
+    {
+      "epoch": 1.6281601123595506,
+      "grad_norm": 0.8910026550292969,
+      "learning_rate": 4.170465955334051e-06,
+      "loss": 0.6086,
+      "step": 4637
+    },
+    {
+      "epoch": 1.6285112359550562,
+      "grad_norm": 0.8203482031822205,
+      "learning_rate": 4.170121959582346e-06,
+      "loss": 0.5843,
+      "step": 4638
+    },
+    {
+      "epoch": 1.6288623595505618,
+      "grad_norm": 0.8972336649894714,
+      "learning_rate": 4.169777906713899e-06,
+      "loss": 0.6345,
+      "step": 4639
+    },
+    {
+      "epoch": 1.6292134831460674,
+      "grad_norm": 0.9142724275588989,
+      "learning_rate": 4.169433796740477e-06,
+      "loss": 0.6292,
+      "step": 4640
+    },
+    {
+      "epoch": 1.629564606741573,
+      "grad_norm": 0.889117419719696,
+      "learning_rate": 4.169089629673848e-06,
+      "loss": 0.6188,
+      "step": 4641
+    },
+    {
+      "epoch": 1.6299157303370788,
+      "grad_norm": 0.9001839756965637,
+      "learning_rate": 4.168745405525781e-06,
+      "loss": 0.6506,
+      "step": 4642
+    },
+    {
+      "epoch": 1.6302668539325844,
+      "grad_norm": 0.8708594441413879,
+      "learning_rate": 4.168401124308049e-06,
+      "loss": 0.5841,
+      "step": 4643
+    },
+    {
+      "epoch": 1.63061797752809,
+      "grad_norm": 0.8744320273399353,
+      "learning_rate": 4.1680567860324275e-06,
+      "loss": 0.6199,
+      "step": 4644
+    },
+    {
+      "epoch": 1.6309691011235956,
+      "grad_norm": 0.8642039895057678,
+      "learning_rate": 4.1677123907106905e-06,
+      "loss": 0.6318,
+      "step": 4645
+    },
+    {
+      "epoch": 1.6313202247191012,
+      "grad_norm": 0.875656247138977,
+      "learning_rate": 4.167367938354616e-06,
+      "loss": 0.6521,
+      "step": 4646
+    },
+    {
+      "epoch": 1.6316713483146068,
+      "grad_norm": 0.8626296520233154,
+      "learning_rate": 4.167023428975986e-06,
+      "loss": 0.6273,
+      "step": 4647
+    },
+    {
+      "epoch": 1.6320224719101124,
+      "grad_norm": 0.8608964085578918,
+      "learning_rate": 4.1666788625865805e-06,
+      "loss": 0.6305,
+      "step": 4648
+    },
+    {
+      "epoch": 1.632373595505618,
+      "grad_norm": 0.871910572052002,
+      "learning_rate": 4.166334239198185e-06,
+      "loss": 0.6386,
+      "step": 4649
+    },
+    {
+      "epoch": 1.6327247191011236,
+      "grad_norm": 0.8875142931938171,
+      "learning_rate": 4.165989558822584e-06,
+      "loss": 0.6126,
+      "step": 4650
+    },
+    {
+      "epoch": 1.6330758426966292,
+      "grad_norm": 0.8592172861099243,
+      "learning_rate": 4.165644821471565e-06,
+      "loss": 0.6037,
+      "step": 4651
+    },
+    {
+      "epoch": 1.6334269662921348,
+      "grad_norm": 0.871364951133728,
+      "learning_rate": 4.165300027156919e-06,
+      "loss": 0.6198,
+      "step": 4652
+    },
+    {
+      "epoch": 1.6337780898876404,
+      "grad_norm": 0.8956006169319153,
+      "learning_rate": 4.164955175890437e-06,
+      "loss": 0.6009,
+      "step": 4653
+    },
+    {
+      "epoch": 1.634129213483146,
+      "grad_norm": 0.8735149502754211,
+      "learning_rate": 4.164610267683913e-06,
+      "loss": 0.6032,
+      "step": 4654
+    },
+    {
+      "epoch": 1.6344803370786516,
+      "grad_norm": 0.8705981373786926,
+      "learning_rate": 4.164265302549142e-06,
+      "loss": 0.6201,
+      "step": 4655
+    },
+    {
+      "epoch": 1.6348314606741572,
+      "grad_norm": 0.8664510846138,
+      "learning_rate": 4.1639202804979215e-06,
+      "loss": 0.5911,
+      "step": 4656
+    },
+    {
+      "epoch": 1.6351825842696628,
+      "grad_norm": 0.8341373205184937,
+      "learning_rate": 4.163575201542052e-06,
+      "loss": 0.579,
+      "step": 4657
+    },
+    {
+      "epoch": 1.6355337078651684,
+      "grad_norm": 0.8607082962989807,
+      "learning_rate": 4.163230065693334e-06,
+      "loss": 0.5828,
+      "step": 4658
+    },
+    {
+      "epoch": 1.6358848314606742,
+      "grad_norm": 0.8956010341644287,
+      "learning_rate": 4.162884872963571e-06,
+      "loss": 0.6105,
+      "step": 4659
+    },
+    {
+      "epoch": 1.6362359550561798,
+      "grad_norm": 0.8953211903572083,
+      "learning_rate": 4.162539623364568e-06,
+      "loss": 0.6139,
+      "step": 4660
+    },
+    {
+      "epoch": 1.6365870786516854,
+      "grad_norm": 0.8708750605583191,
+      "learning_rate": 4.162194316908132e-06,
+      "loss": 0.6285,
+      "step": 4661
+    },
+    {
+      "epoch": 1.636938202247191,
+      "grad_norm": 0.8955643773078918,
+      "learning_rate": 4.161848953606074e-06,
+      "loss": 0.5956,
+      "step": 4662
+    },
+    {
+      "epoch": 1.6372893258426966,
+      "grad_norm": 0.8670720458030701,
+      "learning_rate": 4.161503533470202e-06,
+      "loss": 0.6217,
+      "step": 4663
+    },
+    {
+      "epoch": 1.6376404494382022,
+      "grad_norm": 0.8754477500915527,
+      "learning_rate": 4.1611580565123325e-06,
+      "loss": 0.6245,
+      "step": 4664
+    },
+    {
+      "epoch": 1.637991573033708,
+      "grad_norm": 0.8910819888114929,
+      "learning_rate": 4.160812522744279e-06,
+      "loss": 0.605,
+      "step": 4665
+    },
+    {
+      "epoch": 1.6383426966292136,
+      "grad_norm": 0.8849613666534424,
+      "learning_rate": 4.1604669321778576e-06,
+      "loss": 0.6272,
+      "step": 4666
+    },
+    {
+      "epoch": 1.6386938202247192,
+      "grad_norm": 0.8790765404701233,
+      "learning_rate": 4.160121284824888e-06,
+      "loss": 0.6461,
+      "step": 4667
+    },
+    {
+      "epoch": 1.6390449438202248,
+      "grad_norm": 0.8697189092636108,
+      "learning_rate": 4.1597755806971914e-06,
+      "loss": 0.6243,
+      "step": 4668
+    },
+    {
+      "epoch": 1.6393960674157304,
+      "grad_norm": 0.8819729089736938,
+      "learning_rate": 4.15942981980659e-06,
+      "loss": 0.6333,
+      "step": 4669
+    },
+    {
+      "epoch": 1.639747191011236,
+      "grad_norm": 0.8244314193725586,
+      "learning_rate": 4.15908400216491e-06,
+      "loss": 0.5749,
+      "step": 4670
+    },
+    {
+      "epoch": 1.6400983146067416,
+      "grad_norm": 0.9082034826278687,
+      "learning_rate": 4.158738127783975e-06,
+      "loss": 0.6297,
+      "step": 4671
+    },
+    {
+      "epoch": 1.6404494382022472,
+      "grad_norm": 0.8924310803413391,
+      "learning_rate": 4.1583921966756154e-06,
+      "loss": 0.6438,
+      "step": 4672
+    },
+    {
+      "epoch": 1.6408005617977528,
+      "grad_norm": 0.9138229489326477,
+      "learning_rate": 4.158046208851663e-06,
+      "loss": 0.6064,
+      "step": 4673
+    },
+    {
+      "epoch": 1.6411516853932584,
+      "grad_norm": 0.8768520355224609,
+      "learning_rate": 4.157700164323949e-06,
+      "loss": 0.5972,
+      "step": 4674
+    },
+    {
+      "epoch": 1.641502808988764,
+      "grad_norm": 0.8785306215286255,
+      "learning_rate": 4.157354063104307e-06,
+      "loss": 0.6111,
+      "step": 4675
+    },
+    {
+      "epoch": 1.6418539325842696,
+      "grad_norm": 0.8370828032493591,
+      "learning_rate": 4.157007905204575e-06,
+      "loss": 0.5849,
+      "step": 4676
+    },
+    {
+      "epoch": 1.6422050561797752,
+      "grad_norm": 0.887261152267456,
+      "learning_rate": 4.15666169063659e-06,
+      "loss": 0.6375,
+      "step": 4677
+    },
+    {
+      "epoch": 1.6425561797752808,
+      "grad_norm": 0.8963328003883362,
+      "learning_rate": 4.156315419412194e-06,
+      "loss": 0.6058,
+      "step": 4678
+    },
+    {
+      "epoch": 1.6429073033707864,
+      "grad_norm": 0.8939655423164368,
+      "learning_rate": 4.155969091543226e-06,
+      "loss": 0.621,
+      "step": 4679
+    },
+    {
+      "epoch": 1.643258426966292,
+      "grad_norm": 0.9163799285888672,
+      "learning_rate": 4.155622707041534e-06,
+      "loss": 0.6343,
+      "step": 4680
+    },
+    {
+      "epoch": 1.6436095505617978,
+      "grad_norm": 0.8742914199829102,
+      "learning_rate": 4.155276265918962e-06,
+      "loss": 0.6141,
+      "step": 4681
+    },
+    {
+      "epoch": 1.6439606741573034,
+      "grad_norm": 0.8842009902000427,
+      "learning_rate": 4.154929768187357e-06,
+      "loss": 0.5868,
+      "step": 4682
+    },
+    {
+      "epoch": 1.644311797752809,
+      "grad_norm": 0.8529406189918518,
+      "learning_rate": 4.15458321385857e-06,
+      "loss": 0.6573,
+      "step": 4683
+    },
+    {
+      "epoch": 1.6446629213483146,
+      "grad_norm": 0.8780986666679382,
+      "learning_rate": 4.154236602944454e-06,
+      "loss": 0.614,
+      "step": 4684
+    },
+    {
+      "epoch": 1.6450140449438202,
+      "grad_norm": 0.8548287749290466,
+      "learning_rate": 4.153889935456862e-06,
+      "loss": 0.622,
+      "step": 4685
+    },
+    {
+      "epoch": 1.6453651685393258,
+      "grad_norm": 0.8868494629859924,
+      "learning_rate": 4.153543211407648e-06,
+      "loss": 0.6204,
+      "step": 4686
+    },
+    {
+      "epoch": 1.6457162921348316,
+      "grad_norm": 0.8862740993499756,
+      "learning_rate": 4.153196430808672e-06,
+      "loss": 0.639,
+      "step": 4687
+    },
+    {
+      "epoch": 1.6460674157303372,
+      "grad_norm": 0.8767764568328857,
+      "learning_rate": 4.152849593671793e-06,
+      "loss": 0.6059,
+      "step": 4688
+    },
+    {
+      "epoch": 1.6464185393258428,
+      "grad_norm": 0.9215349555015564,
+      "learning_rate": 4.1525027000088715e-06,
+      "loss": 0.6434,
+      "step": 4689
+    },
+    {
+      "epoch": 1.6467696629213484,
+      "grad_norm": 0.868288516998291,
+      "learning_rate": 4.152155749831772e-06,
+      "loss": 0.6229,
+      "step": 4690
+    },
+    {
+      "epoch": 1.647120786516854,
+      "grad_norm": 0.851254940032959,
+      "learning_rate": 4.15180874315236e-06,
+      "loss": 0.6259,
+      "step": 4691
+    },
+    {
+      "epoch": 1.6474719101123596,
+      "grad_norm": 0.8357805609703064,
+      "learning_rate": 4.151461679982501e-06,
+      "loss": 0.5411,
+      "step": 4692
+    },
+    {
+      "epoch": 1.6478230337078652,
+      "grad_norm": 0.8943964838981628,
+      "learning_rate": 4.1511145603340675e-06,
+      "loss": 0.6564,
+      "step": 4693
+    },
+    {
+      "epoch": 1.6481741573033708,
+      "grad_norm": 0.8508518934249878,
+      "learning_rate": 4.150767384218929e-06,
+      "loss": 0.6238,
+      "step": 4694
+    },
+    {
+      "epoch": 1.6485252808988764,
+      "grad_norm": 0.9138011932373047,
+      "learning_rate": 4.150420151648957e-06,
+      "loss": 0.6538,
+      "step": 4695
+    },
+    {
+      "epoch": 1.648876404494382,
+      "grad_norm": 0.8589404821395874,
+      "learning_rate": 4.150072862636029e-06,
+      "loss": 0.6307,
+      "step": 4696
+    },
+    {
+      "epoch": 1.6492275280898876,
+      "grad_norm": 0.8829411864280701,
+      "learning_rate": 4.149725517192021e-06,
+      "loss": 0.6113,
+      "step": 4697
+    },
+    {
+      "epoch": 1.6495786516853932,
+      "grad_norm": 0.9030507206916809,
+      "learning_rate": 4.149378115328812e-06,
+      "loss": 0.6494,
+      "step": 4698
+    },
+    {
+      "epoch": 1.6499297752808988,
+      "grad_norm": 0.8570032715797424,
+      "learning_rate": 4.149030657058283e-06,
+      "loss": 0.5724,
+      "step": 4699
+    },
+    {
+      "epoch": 1.6502808988764044,
+      "grad_norm": 0.886920690536499,
+      "learning_rate": 4.148683142392317e-06,
+      "loss": 0.6313,
+      "step": 4700
+    },
+    {
+      "epoch": 1.65063202247191,
+      "grad_norm": 0.8540245890617371,
+      "learning_rate": 4.148335571342798e-06,
+      "loss": 0.624,
+      "step": 4701
+    },
+    {
+      "epoch": 1.6509831460674156,
+      "grad_norm": 0.8701122999191284,
+      "learning_rate": 4.1479879439216116e-06,
+      "loss": 0.6177,
+      "step": 4702
+    },
+    {
+      "epoch": 1.6513342696629212,
+      "grad_norm": 0.8617565035820007,
+      "learning_rate": 4.14764026014065e-06,
+      "loss": 0.5755,
+      "step": 4703
+    },
+    {
+      "epoch": 1.651685393258427,
+      "grad_norm": 0.8952060341835022,
+      "learning_rate": 4.1472925200118e-06,
+      "loss": 0.6087,
+      "step": 4704
+    },
+    {
+      "epoch": 1.6520365168539326,
+      "grad_norm": 0.9085246920585632,
+      "learning_rate": 4.146944723546956e-06,
+      "loss": 0.6407,
+      "step": 4705
+    },
+    {
+      "epoch": 1.6523876404494382,
+      "grad_norm": 0.890703022480011,
+      "learning_rate": 4.146596870758012e-06,
+      "loss": 0.6719,
+      "step": 4706
+    },
+    {
+      "epoch": 1.6527387640449438,
+      "grad_norm": 0.8996905088424683,
+      "learning_rate": 4.146248961656863e-06,
+      "loss": 0.6283,
+      "step": 4707
+    },
+    {
+      "epoch": 1.6530898876404494,
+      "grad_norm": 0.8666749000549316,
+      "learning_rate": 4.145900996255409e-06,
+      "loss": 0.6244,
+      "step": 4708
+    },
+    {
+      "epoch": 1.6534410112359552,
+      "grad_norm": 0.8974918127059937,
+      "learning_rate": 4.145552974565549e-06,
+      "loss": 0.6304,
+      "step": 4709
+    },
+    {
+      "epoch": 1.6537921348314608,
+      "grad_norm": 0.8758713603019714,
+      "learning_rate": 4.145204896599186e-06,
+      "loss": 0.6481,
+      "step": 4710
+    },
+    {
+      "epoch": 1.6541432584269664,
+      "grad_norm": 0.8611330389976501,
+      "learning_rate": 4.1448567623682226e-06,
+      "loss": 0.6096,
+      "step": 4711
+    },
+    {
+      "epoch": 1.654494382022472,
+      "grad_norm": 0.8383393883705139,
+      "learning_rate": 4.144508571884565e-06,
+      "loss": 0.5943,
+      "step": 4712
+    },
+    {
+      "epoch": 1.6548455056179776,
+      "grad_norm": 0.8979886174201965,
+      "learning_rate": 4.144160325160122e-06,
+      "loss": 0.6182,
+      "step": 4713
+    },
+    {
+      "epoch": 1.6551966292134832,
+      "grad_norm": 0.8730772733688354,
+      "learning_rate": 4.143812022206804e-06,
+      "loss": 0.6265,
+      "step": 4714
+    },
+    {
+      "epoch": 1.6555477528089888,
+      "grad_norm": 0.8903005123138428,
+      "learning_rate": 4.14346366303652e-06,
+      "loss": 0.6213,
+      "step": 4715
+    },
+    {
+      "epoch": 1.6558988764044944,
+      "grad_norm": 0.8978590369224548,
+      "learning_rate": 4.1431152476611856e-06,
+      "loss": 0.6048,
+      "step": 4716
+    },
+    {
+      "epoch": 1.65625,
+      "grad_norm": 0.8998496532440186,
+      "learning_rate": 4.1427667760927155e-06,
+      "loss": 0.6435,
+      "step": 4717
+    },
+    {
+      "epoch": 1.6566011235955056,
+      "grad_norm": 0.882354736328125,
+      "learning_rate": 4.142418248343028e-06,
+      "loss": 0.6063,
+      "step": 4718
+    },
+    {
+      "epoch": 1.6569522471910112,
+      "grad_norm": 0.8468122482299805,
+      "learning_rate": 4.142069664424041e-06,
+      "loss": 0.607,
+      "step": 4719
+    },
+    {
+      "epoch": 1.6573033707865168,
+      "grad_norm": 0.8717797994613647,
+      "learning_rate": 4.141721024347677e-06,
+      "loss": 0.6255,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6576544943820224,
+      "grad_norm": 0.8650950193405151,
+      "learning_rate": 4.141372328125859e-06,
+      "loss": 0.5888,
+      "step": 4721
+    },
+    {
+      "epoch": 1.658005617977528,
+      "grad_norm": 0.8369475603103638,
+      "learning_rate": 4.141023575770512e-06,
+      "loss": 0.5846,
+      "step": 4722
+    },
+    {
+      "epoch": 1.6583567415730336,
+      "grad_norm": 0.8811677694320679,
+      "learning_rate": 4.140674767293562e-06,
+      "loss": 0.6263,
+      "step": 4723
+    },
+    {
+      "epoch": 1.6587078651685392,
+      "grad_norm": 0.8905251026153564,
+      "learning_rate": 4.14032590270694e-06,
+      "loss": 0.6204,
+      "step": 4724
+    },
+    {
+      "epoch": 1.6590589887640448,
+      "grad_norm": 0.8818519711494446,
+      "learning_rate": 4.139976982022576e-06,
+      "loss": 0.6142,
+      "step": 4725
+    },
+    {
+      "epoch": 1.6594101123595506,
+      "grad_norm": 0.8485746383666992,
+      "learning_rate": 4.139628005252401e-06,
+      "loss": 0.5798,
+      "step": 4726
+    },
+    {
+      "epoch": 1.6597612359550562,
+      "grad_norm": 0.8504104614257812,
+      "learning_rate": 4.139278972408353e-06,
+      "loss": 0.6418,
+      "step": 4727
+    },
+    {
+      "epoch": 1.6601123595505618,
+      "grad_norm": 0.8766199946403503,
+      "learning_rate": 4.138929883502365e-06,
+      "loss": 0.6014,
+      "step": 4728
+    },
+    {
+      "epoch": 1.6604634831460674,
+      "grad_norm": 0.8618320226669312,
+      "learning_rate": 4.1385807385463785e-06,
+      "loss": 0.5874,
+      "step": 4729
+    },
+    {
+      "epoch": 1.660814606741573,
+      "grad_norm": 0.8511069416999817,
+      "learning_rate": 4.138231537552333e-06,
+      "loss": 0.6173,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6611657303370788,
+      "grad_norm": 0.859161913394928,
+      "learning_rate": 4.137882280532171e-06,
+      "loss": 0.6264,
+      "step": 4731
+    },
+    {
+      "epoch": 1.6615168539325844,
+      "grad_norm": 0.8392783999443054,
+      "learning_rate": 4.137532967497837e-06,
+      "loss": 0.6064,
+      "step": 4732
+    },
+    {
+      "epoch": 1.66186797752809,
+      "grad_norm": 0.8905898332595825,
+      "learning_rate": 4.1371835984612755e-06,
+      "loss": 0.6294,
+      "step": 4733
+    },
+    {
+      "epoch": 1.6622191011235956,
+      "grad_norm": 0.8814263343811035,
+      "learning_rate": 4.1368341734344365e-06,
+      "loss": 0.6053,
+      "step": 4734
+    },
+    {
+      "epoch": 1.6625702247191012,
+      "grad_norm": 0.8924182653427124,
+      "learning_rate": 4.13648469242927e-06,
+      "loss": 0.6454,
+      "step": 4735
+    },
+    {
+      "epoch": 1.6629213483146068,
+      "grad_norm": 0.8582429885864258,
+      "learning_rate": 4.136135155457726e-06,
+      "loss": 0.5989,
+      "step": 4736
+    },
+    {
+      "epoch": 1.6632724719101124,
+      "grad_norm": 0.9076274037361145,
+      "learning_rate": 4.135785562531761e-06,
+      "loss": 0.6454,
+      "step": 4737
+    },
+    {
+      "epoch": 1.663623595505618,
+      "grad_norm": 0.9228194355964661,
+      "learning_rate": 4.135435913663328e-06,
+      "loss": 0.6089,
+      "step": 4738
+    },
+    {
+      "epoch": 1.6639747191011236,
+      "grad_norm": 0.8946560025215149,
+      "learning_rate": 4.135086208864388e-06,
+      "loss": 0.6196,
+      "step": 4739
+    },
+    {
+      "epoch": 1.6643258426966292,
+      "grad_norm": 0.8732385635375977,
+      "learning_rate": 4.134736448146898e-06,
+      "loss": 0.6055,
+      "step": 4740
+    },
+    {
+      "epoch": 1.6646769662921348,
+      "grad_norm": 0.8972743153572083,
+      "learning_rate": 4.13438663152282e-06,
+      "loss": 0.6511,
+      "step": 4741
+    },
+    {
+      "epoch": 1.6650280898876404,
+      "grad_norm": 0.9056386351585388,
+      "learning_rate": 4.134036759004119e-06,
+      "loss": 0.635,
+      "step": 4742
+    },
+    {
+      "epoch": 1.665379213483146,
+      "grad_norm": 0.8671088814735413,
+      "learning_rate": 4.133686830602758e-06,
+      "loss": 0.6208,
+      "step": 4743
+    },
+    {
+      "epoch": 1.6657303370786516,
+      "grad_norm": 0.8547230958938599,
+      "learning_rate": 4.133336846330705e-06,
+      "loss": 0.6195,
+      "step": 4744
+    },
+    {
+      "epoch": 1.6660814606741572,
+      "grad_norm": 0.8662692308425903,
+      "learning_rate": 4.132986806199931e-06,
+      "loss": 0.6028,
+      "step": 4745
+    },
+    {
+      "epoch": 1.6664325842696628,
+      "grad_norm": 0.9029760956764221,
+      "learning_rate": 4.132636710222404e-06,
+      "loss": 0.6218,
+      "step": 4746
+    },
+    {
+      "epoch": 1.6667837078651684,
+      "grad_norm": 0.8908417224884033,
+      "learning_rate": 4.1322865584100994e-06,
+      "loss": 0.6275,
+      "step": 4747
+    },
+    {
+      "epoch": 1.6671348314606742,
+      "grad_norm": 0.8707289099693298,
+      "learning_rate": 4.131936350774991e-06,
+      "loss": 0.6066,
+      "step": 4748
+    },
+    {
+      "epoch": 1.6674859550561798,
+      "grad_norm": 0.8993298411369324,
+      "learning_rate": 4.131586087329057e-06,
+      "loss": 0.5929,
+      "step": 4749
+    },
+    {
+      "epoch": 1.6678370786516854,
+      "grad_norm": 0.8987124562263489,
+      "learning_rate": 4.1312357680842735e-06,
+      "loss": 0.6188,
+      "step": 4750
+    },
+    {
+      "epoch": 1.668188202247191,
+      "grad_norm": 0.8531199097633362,
+      "learning_rate": 4.130885393052623e-06,
+      "loss": 0.6091,
+      "step": 4751
+    },
+    {
+      "epoch": 1.6685393258426966,
+      "grad_norm": 0.8786628246307373,
+      "learning_rate": 4.130534962246088e-06,
+      "loss": 0.6173,
+      "step": 4752
+    },
+    {
+      "epoch": 1.6688904494382022,
+      "grad_norm": 0.8418239951133728,
+      "learning_rate": 4.130184475676652e-06,
+      "loss": 0.6025,
+      "step": 4753
+    },
+    {
+      "epoch": 1.669241573033708,
+      "grad_norm": 0.8778673410415649,
+      "learning_rate": 4.129833933356302e-06,
+      "loss": 0.6347,
+      "step": 4754
+    },
+    {
+      "epoch": 1.6695926966292136,
+      "grad_norm": 0.8813691735267639,
+      "learning_rate": 4.129483335297027e-06,
+      "loss": 0.6317,
+      "step": 4755
+    },
+    {
+      "epoch": 1.6699438202247192,
+      "grad_norm": 0.8676353096961975,
+      "learning_rate": 4.129132681510815e-06,
+      "loss": 0.6313,
+      "step": 4756
+    },
+    {
+      "epoch": 1.6702949438202248,
+      "grad_norm": 0.848374605178833,
+      "learning_rate": 4.128781972009661e-06,
+      "loss": 0.6367,
+      "step": 4757
+    },
+    {
+      "epoch": 1.6706460674157304,
+      "grad_norm": 0.875219464302063,
+      "learning_rate": 4.128431206805556e-06,
+      "loss": 0.6053,
+      "step": 4758
+    },
+    {
+      "epoch": 1.670997191011236,
+      "grad_norm": 0.8850371241569519,
+      "learning_rate": 4.1280803859104975e-06,
+      "loss": 0.6308,
+      "step": 4759
+    },
+    {
+      "epoch": 1.6713483146067416,
+      "grad_norm": 0.8848199248313904,
+      "learning_rate": 4.1277295093364835e-06,
+      "loss": 0.5875,
+      "step": 4760
+    },
+    {
+      "epoch": 1.6716994382022472,
+      "grad_norm": 0.8953977227210999,
+      "learning_rate": 4.1273785770955125e-06,
+      "loss": 0.6082,
+      "step": 4761
+    },
+    {
+      "epoch": 1.6720505617977528,
+      "grad_norm": 0.8523750305175781,
+      "learning_rate": 4.127027589199587e-06,
+      "loss": 0.6132,
+      "step": 4762
+    },
+    {
+      "epoch": 1.6724016853932584,
+      "grad_norm": 0.8619308471679688,
+      "learning_rate": 4.126676545660709e-06,
+      "loss": 0.5715,
+      "step": 4763
+    },
+    {
+      "epoch": 1.672752808988764,
+      "grad_norm": 0.9020019173622131,
+      "learning_rate": 4.126325446490887e-06,
+      "loss": 0.6054,
+      "step": 4764
+    },
+    {
+      "epoch": 1.6731039325842696,
+      "grad_norm": 0.8890277147293091,
+      "learning_rate": 4.125974291702126e-06,
+      "loss": 0.6116,
+      "step": 4765
+    },
+    {
+      "epoch": 1.6734550561797752,
+      "grad_norm": 0.9404541254043579,
+      "learning_rate": 4.125623081306435e-06,
+      "loss": 0.6256,
+      "step": 4766
+    },
+    {
+      "epoch": 1.6738061797752808,
+      "grad_norm": 0.8866811990737915,
+      "learning_rate": 4.125271815315826e-06,
+      "loss": 0.6266,
+      "step": 4767
+    },
+    {
+      "epoch": 1.6741573033707864,
+      "grad_norm": 0.8981043696403503,
+      "learning_rate": 4.1249204937423115e-06,
+      "loss": 0.6314,
+      "step": 4768
+    },
+    {
+      "epoch": 1.674508426966292,
+      "grad_norm": 0.889731228351593,
+      "learning_rate": 4.124569116597907e-06,
+      "loss": 0.5948,
+      "step": 4769
+    },
+    {
+      "epoch": 1.6748595505617978,
+      "grad_norm": 0.9299228191375732,
+      "learning_rate": 4.124217683894628e-06,
+      "loss": 0.6508,
+      "step": 4770
+    },
+    {
+      "epoch": 1.6752106741573034,
+      "grad_norm": 0.909980058670044,
+      "learning_rate": 4.1238661956444946e-06,
+      "loss": 0.6499,
+      "step": 4771
+    },
+    {
+      "epoch": 1.675561797752809,
+      "grad_norm": 0.8873563408851624,
+      "learning_rate": 4.123514651859527e-06,
+      "loss": 0.6122,
+      "step": 4772
+    },
+    {
+      "epoch": 1.6759129213483146,
+      "grad_norm": 0.8655134439468384,
+      "learning_rate": 4.123163052551747e-06,
+      "loss": 0.6395,
+      "step": 4773
+    },
+    {
+      "epoch": 1.6762640449438202,
+      "grad_norm": 0.8635685443878174,
+      "learning_rate": 4.122811397733181e-06,
+      "loss": 0.6194,
+      "step": 4774
+    },
+    {
+      "epoch": 1.6766151685393258,
+      "grad_norm": 0.8841813802719116,
+      "learning_rate": 4.122459687415852e-06,
+      "loss": 0.6524,
+      "step": 4775
+    },
+    {
+      "epoch": 1.6769662921348316,
+      "grad_norm": 0.8351168632507324,
+      "learning_rate": 4.122107921611791e-06,
+      "loss": 0.6083,
+      "step": 4776
+    },
+    {
+      "epoch": 1.6773174157303372,
+      "grad_norm": 0.8590231537818909,
+      "learning_rate": 4.121756100333028e-06,
+      "loss": 0.6243,
+      "step": 4777
+    },
+    {
+      "epoch": 1.6776685393258428,
+      "grad_norm": 0.8575931191444397,
+      "learning_rate": 4.121404223591592e-06,
+      "loss": 0.5959,
+      "step": 4778
+    },
+    {
+      "epoch": 1.6780196629213484,
+      "grad_norm": 0.8839737772941589,
+      "learning_rate": 4.12105229139952e-06,
+      "loss": 0.6432,
+      "step": 4779
+    },
+    {
+      "epoch": 1.678370786516854,
+      "grad_norm": 0.871910572052002,
+      "learning_rate": 4.120700303768848e-06,
+      "loss": 0.6134,
+      "step": 4780
+    },
+    {
+      "epoch": 1.6787219101123596,
+      "grad_norm": 0.8897039294242859,
+      "learning_rate": 4.1203482607116105e-06,
+      "loss": 0.6329,
+      "step": 4781
+    },
+    {
+      "epoch": 1.6790730337078652,
+      "grad_norm": 0.8837077617645264,
+      "learning_rate": 4.119996162239851e-06,
+      "loss": 0.6059,
+      "step": 4782
+    },
+    {
+      "epoch": 1.6794241573033708,
+      "grad_norm": 0.8870334625244141,
+      "learning_rate": 4.119644008365607e-06,
+      "loss": 0.5785,
+      "step": 4783
+    },
+    {
+      "epoch": 1.6797752808988764,
+      "grad_norm": 0.8958740234375,
+      "learning_rate": 4.119291799100925e-06,
+      "loss": 0.6076,
+      "step": 4784
+    },
+    {
+      "epoch": 1.680126404494382,
+      "grad_norm": 0.8951886892318726,
+      "learning_rate": 4.118939534457849e-06,
+      "loss": 0.6134,
+      "step": 4785
+    },
+    {
+      "epoch": 1.6804775280898876,
+      "grad_norm": 0.8920659422874451,
+      "learning_rate": 4.118587214448426e-06,
+      "loss": 0.6176,
+      "step": 4786
+    },
+    {
+      "epoch": 1.6808286516853932,
+      "grad_norm": 0.9017564058303833,
+      "learning_rate": 4.118234839084706e-06,
+      "loss": 0.623,
+      "step": 4787
+    },
+    {
+      "epoch": 1.6811797752808988,
+      "grad_norm": 0.8780067563056946,
+      "learning_rate": 4.117882408378739e-06,
+      "loss": 0.6034,
+      "step": 4788
+    },
+    {
+      "epoch": 1.6815308988764044,
+      "grad_norm": 0.8677077293395996,
+      "learning_rate": 4.117529922342578e-06,
+      "loss": 0.5848,
+      "step": 4789
+    },
+    {
+      "epoch": 1.68188202247191,
+      "grad_norm": 0.9106276035308838,
+      "learning_rate": 4.117177380988277e-06,
+      "loss": 0.634,
+      "step": 4790
+    },
+    {
+      "epoch": 1.6822331460674156,
+      "grad_norm": 0.9075629711151123,
+      "learning_rate": 4.116824784327894e-06,
+      "loss": 0.622,
+      "step": 4791
+    },
+    {
+      "epoch": 1.6825842696629212,
+      "grad_norm": 0.9017618298530579,
+      "learning_rate": 4.116472132373488e-06,
+      "loss": 0.6306,
+      "step": 4792
+    },
+    {
+      "epoch": 1.682935393258427,
+      "grad_norm": 0.846936821937561,
+      "learning_rate": 4.116119425137117e-06,
+      "loss": 0.5728,
+      "step": 4793
+    },
+    {
+      "epoch": 1.6832865168539326,
+      "grad_norm": 0.8908042311668396,
+      "learning_rate": 4.115766662630845e-06,
+      "loss": 0.6305,
+      "step": 4794
+    },
+    {
+      "epoch": 1.6836376404494382,
+      "grad_norm": 0.8723462224006653,
+      "learning_rate": 4.115413844866736e-06,
+      "loss": 0.6326,
+      "step": 4795
+    },
+    {
+      "epoch": 1.6839887640449438,
+      "grad_norm": 0.9132239818572998,
+      "learning_rate": 4.115060971856855e-06,
+      "loss": 0.6706,
+      "step": 4796
+    },
+    {
+      "epoch": 1.6843398876404494,
+      "grad_norm": 0.8534126281738281,
+      "learning_rate": 4.1147080436132716e-06,
+      "loss": 0.597,
+      "step": 4797
+    },
+    {
+      "epoch": 1.6846910112359552,
+      "grad_norm": 0.8956817388534546,
+      "learning_rate": 4.1143550601480546e-06,
+      "loss": 0.6305,
+      "step": 4798
+    },
+    {
+      "epoch": 1.6850421348314608,
+      "grad_norm": 0.8739310503005981,
+      "learning_rate": 4.114002021473276e-06,
+      "loss": 0.6274,
+      "step": 4799
+    },
+    {
+      "epoch": 1.6853932584269664,
+      "grad_norm": 0.8571768403053284,
+      "learning_rate": 4.113648927601008e-06,
+      "loss": 0.6144,
+      "step": 4800
+    },
+    {
+      "epoch": 1.685744382022472,
+      "grad_norm": 0.9097314476966858,
+      "learning_rate": 4.113295778543329e-06,
+      "loss": 0.5915,
+      "step": 4801
+    },
+    {
+      "epoch": 1.6860955056179776,
+      "grad_norm": 0.8725996613502502,
+      "learning_rate": 4.112942574312315e-06,
+      "loss": 0.6315,
+      "step": 4802
+    },
+    {
+      "epoch": 1.6864466292134832,
+      "grad_norm": 0.8608437776565552,
+      "learning_rate": 4.112589314920045e-06,
+      "loss": 0.6396,
+      "step": 4803
+    },
+    {
+      "epoch": 1.6867977528089888,
+      "grad_norm": 0.8394370675086975,
+      "learning_rate": 4.1122360003785995e-06,
+      "loss": 0.5763,
+      "step": 4804
+    },
+    {
+      "epoch": 1.6871488764044944,
+      "grad_norm": 0.8686445951461792,
+      "learning_rate": 4.111882630700064e-06,
+      "loss": 0.6144,
+      "step": 4805
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 0.8637919425964355,
+      "learning_rate": 4.11152920589652e-06,
+      "loss": 0.5964,
+      "step": 4806
+    },
+    {
+      "epoch": 1.6878511235955056,
+      "grad_norm": 0.8992030620574951,
+      "learning_rate": 4.111175725980057e-06,
+      "loss": 0.6218,
+      "step": 4807
+    },
+    {
+      "epoch": 1.6882022471910112,
+      "grad_norm": 0.8575127124786377,
+      "learning_rate": 4.110822190962764e-06,
+      "loss": 0.6043,
+      "step": 4808
+    },
+    {
+      "epoch": 1.6885533707865168,
+      "grad_norm": 0.8570694327354431,
+      "learning_rate": 4.110468600856729e-06,
+      "loss": 0.6018,
+      "step": 4809
+    },
+    {
+      "epoch": 1.6889044943820224,
+      "grad_norm": 0.8761375546455383,
+      "learning_rate": 4.110114955674046e-06,
+      "loss": 0.6141,
+      "step": 4810
+    },
+    {
+      "epoch": 1.689255617977528,
+      "grad_norm": 0.8680332899093628,
+      "learning_rate": 4.10976125542681e-06,
+      "loss": 0.5964,
+      "step": 4811
+    },
+    {
+      "epoch": 1.6896067415730336,
+      "grad_norm": 0.8575436472892761,
+      "learning_rate": 4.1094075001271164e-06,
+      "loss": 0.6062,
+      "step": 4812
+    },
+    {
+      "epoch": 1.6899578651685392,
+      "grad_norm": 0.9025743007659912,
+      "learning_rate": 4.109053689787063e-06,
+      "loss": 0.6434,
+      "step": 4813
+    },
+    {
+      "epoch": 1.6903089887640448,
+      "grad_norm": 0.9137904047966003,
+      "learning_rate": 4.108699824418751e-06,
+      "loss": 0.6291,
+      "step": 4814
+    },
+    {
+      "epoch": 1.6906601123595506,
+      "grad_norm": 0.8753927946090698,
+      "learning_rate": 4.108345904034281e-06,
+      "loss": 0.6151,
+      "step": 4815
+    },
+    {
+      "epoch": 1.6910112359550562,
+      "grad_norm": 0.8979054689407349,
+      "learning_rate": 4.107991928645757e-06,
+      "loss": 0.6381,
+      "step": 4816
+    },
+    {
+      "epoch": 1.6913623595505618,
+      "grad_norm": 0.8909083008766174,
+      "learning_rate": 4.107637898265286e-06,
+      "loss": 0.6121,
+      "step": 4817
+    },
+    {
+      "epoch": 1.6917134831460674,
+      "grad_norm": 0.8720690011978149,
+      "learning_rate": 4.107283812904974e-06,
+      "loss": 0.6342,
+      "step": 4818
+    },
+    {
+      "epoch": 1.692064606741573,
+      "grad_norm": 0.9081621766090393,
+      "learning_rate": 4.106929672576932e-06,
+      "loss": 0.6058,
+      "step": 4819
+    },
+    {
+      "epoch": 1.6924157303370788,
+      "grad_norm": 0.8846354484558105,
+      "learning_rate": 4.106575477293269e-06,
+      "loss": 0.6247,
+      "step": 4820
+    },
+    {
+      "epoch": 1.6927668539325844,
+      "grad_norm": 0.8892889022827148,
+      "learning_rate": 4.1062212270661e-06,
+      "loss": 0.6229,
+      "step": 4821
+    },
+    {
+      "epoch": 1.69311797752809,
+      "grad_norm": 0.8698278069496155,
+      "learning_rate": 4.105866921907539e-06,
+      "loss": 0.6147,
+      "step": 4822
+    },
+    {
+      "epoch": 1.6934691011235956,
+      "grad_norm": 0.8612379431724548,
+      "learning_rate": 4.1055125618297036e-06,
+      "loss": 0.6106,
+      "step": 4823
+    },
+    {
+      "epoch": 1.6938202247191012,
+      "grad_norm": 0.8882355690002441,
+      "learning_rate": 4.105158146844713e-06,
+      "loss": 0.588,
+      "step": 4824
+    },
+    {
+      "epoch": 1.6941713483146068,
+      "grad_norm": 0.9116877317428589,
+      "learning_rate": 4.104803676964686e-06,
+      "loss": 0.663,
+      "step": 4825
+    },
+    {
+      "epoch": 1.6945224719101124,
+      "grad_norm": 0.8949796557426453,
+      "learning_rate": 4.1044491522017485e-06,
+      "loss": 0.6081,
+      "step": 4826
+    },
+    {
+      "epoch": 1.694873595505618,
+      "grad_norm": 0.8727705478668213,
+      "learning_rate": 4.104094572568022e-06,
+      "loss": 0.5666,
+      "step": 4827
+    },
+    {
+      "epoch": 1.6952247191011236,
+      "grad_norm": 0.8831336498260498,
+      "learning_rate": 4.1037399380756324e-06,
+      "loss": 0.627,
+      "step": 4828
+    },
+    {
+      "epoch": 1.6955758426966292,
+      "grad_norm": 0.8737815618515015,
+      "learning_rate": 4.103385248736711e-06,
+      "loss": 0.6177,
+      "step": 4829
+    },
+    {
+      "epoch": 1.6959269662921348,
+      "grad_norm": 0.8915756344795227,
+      "learning_rate": 4.103030504563385e-06,
+      "loss": 0.6209,
+      "step": 4830
+    },
+    {
+      "epoch": 1.6962780898876404,
+      "grad_norm": 0.9170124530792236,
+      "learning_rate": 4.102675705567788e-06,
+      "loss": 0.6134,
+      "step": 4831
+    },
+    {
+      "epoch": 1.696629213483146,
+      "grad_norm": 0.9044004678726196,
+      "learning_rate": 4.1023208517620535e-06,
+      "loss": 0.6432,
+      "step": 4832
+    },
+    {
+      "epoch": 1.6969803370786516,
+      "grad_norm": 0.9003125429153442,
+      "learning_rate": 4.101965943158318e-06,
+      "loss": 0.6029,
+      "step": 4833
+    },
+    {
+      "epoch": 1.6973314606741572,
+      "grad_norm": 0.9001431465148926,
+      "learning_rate": 4.101610979768716e-06,
+      "loss": 0.6043,
+      "step": 4834
+    },
+    {
+      "epoch": 1.6976825842696628,
+      "grad_norm": 0.8822726607322693,
+      "learning_rate": 4.10125596160539e-06,
+      "loss": 0.6257,
+      "step": 4835
+    },
+    {
+      "epoch": 1.6980337078651684,
+      "grad_norm": 0.8968911170959473,
+      "learning_rate": 4.10090088868048e-06,
+      "loss": 0.6129,
+      "step": 4836
+    },
+    {
+      "epoch": 1.6983848314606742,
+      "grad_norm": 0.9139944314956665,
+      "learning_rate": 4.10054576100613e-06,
+      "loss": 0.6493,
+      "step": 4837
+    },
+    {
+      "epoch": 1.6987359550561798,
+      "grad_norm": 0.8980047106742859,
+      "learning_rate": 4.1001905785944855e-06,
+      "loss": 0.615,
+      "step": 4838
+    },
+    {
+      "epoch": 1.6990870786516854,
+      "grad_norm": 0.9069281816482544,
+      "learning_rate": 4.099835341457691e-06,
+      "loss": 0.5864,
+      "step": 4839
+    },
+    {
+      "epoch": 1.699438202247191,
+      "grad_norm": 1.545245885848999,
+      "learning_rate": 4.099480049607897e-06,
+      "loss": 0.6083,
+      "step": 4840
+    },
+    {
+      "epoch": 1.6997893258426966,
+      "grad_norm": 0.8865572214126587,
+      "learning_rate": 4.099124703057254e-06,
+      "loss": 0.6515,
+      "step": 4841
+    },
+    {
+      "epoch": 1.7001404494382022,
+      "grad_norm": 0.84342360496521,
+      "learning_rate": 4.0987693018179145e-06,
+      "loss": 0.6363,
+      "step": 4842
+    },
+    {
+      "epoch": 1.700491573033708,
+      "grad_norm": 0.8635156750679016,
+      "learning_rate": 4.098413845902033e-06,
+      "loss": 0.5762,
+      "step": 4843
+    },
+    {
+      "epoch": 1.7008426966292136,
+      "grad_norm": 0.8980802297592163,
+      "learning_rate": 4.098058335321767e-06,
+      "loss": 0.6078,
+      "step": 4844
+    },
+    {
+      "epoch": 1.7011938202247192,
+      "grad_norm": 0.888487696647644,
+      "learning_rate": 4.0977027700892725e-06,
+      "loss": 0.6108,
+      "step": 4845
+    },
+    {
+      "epoch": 1.7015449438202248,
+      "grad_norm": 0.8712450265884399,
+      "learning_rate": 4.09734715021671e-06,
+      "loss": 0.6579,
+      "step": 4846
+    },
+    {
+      "epoch": 1.7018960674157304,
+      "grad_norm": 0.8814717531204224,
+      "learning_rate": 4.096991475716242e-06,
+      "loss": 0.64,
+      "step": 4847
+    },
+    {
+      "epoch": 1.702247191011236,
+      "grad_norm": 0.8596243262290955,
+      "learning_rate": 4.096635746600032e-06,
+      "loss": 0.594,
+      "step": 4848
+    },
+    {
+      "epoch": 1.7025983146067416,
+      "grad_norm": 0.8710845708847046,
+      "learning_rate": 4.096279962880246e-06,
+      "loss": 0.6383,
+      "step": 4849
+    },
+    {
+      "epoch": 1.7029494382022472,
+      "grad_norm": 0.8820031881332397,
+      "learning_rate": 4.095924124569052e-06,
+      "loss": 0.6421,
+      "step": 4850
+    },
+    {
+      "epoch": 1.7033005617977528,
+      "grad_norm": 0.9059613347053528,
+      "learning_rate": 4.095568231678617e-06,
+      "loss": 0.6502,
+      "step": 4851
+    },
+    {
+      "epoch": 1.7036516853932584,
+      "grad_norm": 0.9789934158325195,
+      "learning_rate": 4.095212284221115e-06,
+      "loss": 0.6336,
+      "step": 4852
+    },
+    {
+      "epoch": 1.704002808988764,
+      "grad_norm": 0.8713051080703735,
+      "learning_rate": 4.094856282208717e-06,
+      "loss": 0.6108,
+      "step": 4853
+    },
+    {
+      "epoch": 1.7043539325842696,
+      "grad_norm": 0.9158304929733276,
+      "learning_rate": 4.0945002256536e-06,
+      "loss": 0.6391,
+      "step": 4854
+    },
+    {
+      "epoch": 1.7047050561797752,
+      "grad_norm": 0.8795137405395508,
+      "learning_rate": 4.094144114567939e-06,
+      "loss": 0.631,
+      "step": 4855
+    },
+    {
+      "epoch": 1.7050561797752808,
+      "grad_norm": 0.9169765710830688,
+      "learning_rate": 4.093787948963913e-06,
+      "loss": 0.6475,
+      "step": 4856
+    },
+    {
+      "epoch": 1.7054073033707864,
+      "grad_norm": 0.8848928213119507,
+      "learning_rate": 4.093431728853704e-06,
+      "loss": 0.6164,
+      "step": 4857
+    },
+    {
+      "epoch": 1.705758426966292,
+      "grad_norm": 0.8515163660049438,
+      "learning_rate": 4.0930754542494935e-06,
+      "loss": 0.6388,
+      "step": 4858
+    },
+    {
+      "epoch": 1.7061095505617978,
+      "grad_norm": 0.8862313628196716,
+      "learning_rate": 4.092719125163464e-06,
+      "loss": 0.6385,
+      "step": 4859
+    },
+    {
+      "epoch": 1.7064606741573034,
+      "grad_norm": 0.8719597458839417,
+      "learning_rate": 4.092362741607805e-06,
+      "loss": 0.5853,
+      "step": 4860
+    },
+    {
+      "epoch": 1.706811797752809,
+      "grad_norm": 0.8930037617683411,
+      "learning_rate": 4.092006303594702e-06,
+      "loss": 0.6277,
+      "step": 4861
+    },
+    {
+      "epoch": 1.7071629213483146,
+      "grad_norm": 0.8765163421630859,
+      "learning_rate": 4.091649811136346e-06,
+      "loss": 0.6454,
+      "step": 4862
+    },
+    {
+      "epoch": 1.7075140449438202,
+      "grad_norm": 0.8377728462219238,
+      "learning_rate": 4.091293264244929e-06,
+      "loss": 0.5957,
+      "step": 4863
+    },
+    {
+      "epoch": 1.7078651685393258,
+      "grad_norm": 0.8431267738342285,
+      "learning_rate": 4.090936662932643e-06,
+      "loss": 0.586,
+      "step": 4864
+    },
+    {
+      "epoch": 1.7082162921348316,
+      "grad_norm": 0.923705518245697,
+      "learning_rate": 4.090580007211686e-06,
+      "loss": 0.6231,
+      "step": 4865
+    },
+    {
+      "epoch": 1.7085674157303372,
+      "grad_norm": 0.8978968262672424,
+      "learning_rate": 4.090223297094251e-06,
+      "loss": 0.6113,
+      "step": 4866
+    },
+    {
+      "epoch": 1.7089185393258428,
+      "grad_norm": 0.8877233266830444,
+      "learning_rate": 4.0898665325925425e-06,
+      "loss": 0.6058,
+      "step": 4867
+    },
+    {
+      "epoch": 1.7092696629213484,
+      "grad_norm": 0.8544127941131592,
+      "learning_rate": 4.089509713718759e-06,
+      "loss": 0.596,
+      "step": 4868
+    },
+    {
+      "epoch": 1.709620786516854,
+      "grad_norm": 0.885586678981781,
+      "learning_rate": 4.089152840485102e-06,
+      "loss": 0.603,
+      "step": 4869
+    },
+    {
+      "epoch": 1.7099719101123596,
+      "grad_norm": 0.8819291591644287,
+      "learning_rate": 4.088795912903779e-06,
+      "loss": 0.6087,
+      "step": 4870
+    },
+    {
+      "epoch": 1.7103230337078652,
+      "grad_norm": 0.8448066115379333,
+      "learning_rate": 4.088438930986995e-06,
+      "loss": 0.6105,
+      "step": 4871
+    },
+    {
+      "epoch": 1.7106741573033708,
+      "grad_norm": 0.8704783320426941,
+      "learning_rate": 4.088081894746959e-06,
+      "loss": 0.664,
+      "step": 4872
+    },
+    {
+      "epoch": 1.7110252808988764,
+      "grad_norm": 0.9045015573501587,
+      "learning_rate": 4.08772480419588e-06,
+      "loss": 0.6106,
+      "step": 4873
+    },
+    {
+      "epoch": 1.711376404494382,
+      "grad_norm": 0.8924446105957031,
+      "learning_rate": 4.087367659345973e-06,
+      "loss": 0.6542,
+      "step": 4874
+    },
+    {
+      "epoch": 1.7117275280898876,
+      "grad_norm": 0.8980352878570557,
+      "learning_rate": 4.0870104602094495e-06,
+      "loss": 0.6459,
+      "step": 4875
+    },
+    {
+      "epoch": 1.7120786516853932,
+      "grad_norm": 0.8418675661087036,
+      "learning_rate": 4.086653206798526e-06,
+      "loss": 0.6147,
+      "step": 4876
+    },
+    {
+      "epoch": 1.7124297752808988,
+      "grad_norm": 0.8605186343193054,
+      "learning_rate": 4.086295899125421e-06,
+      "loss": 0.6553,
+      "step": 4877
+    },
+    {
+      "epoch": 1.7127808988764044,
+      "grad_norm": 0.8851338624954224,
+      "learning_rate": 4.085938537202353e-06,
+      "loss": 0.6092,
+      "step": 4878
+    },
+    {
+      "epoch": 1.71313202247191,
+      "grad_norm": 0.8552588820457458,
+      "learning_rate": 4.085581121041545e-06,
+      "loss": 0.6042,
+      "step": 4879
+    },
+    {
+      "epoch": 1.7134831460674156,
+      "grad_norm": 0.8800450563430786,
+      "learning_rate": 4.085223650655219e-06,
+      "loss": 0.6298,
+      "step": 4880
+    },
+    {
+      "epoch": 1.7138342696629212,
+      "grad_norm": 0.8970910310745239,
+      "learning_rate": 4.084866126055601e-06,
+      "loss": 0.6378,
+      "step": 4881
+    },
+    {
+      "epoch": 1.714185393258427,
+      "grad_norm": 0.8541218638420105,
+      "learning_rate": 4.084508547254917e-06,
+      "loss": 0.5781,
+      "step": 4882
+    },
+    {
+      "epoch": 1.7145365168539326,
+      "grad_norm": 0.87952721118927,
+      "learning_rate": 4.084150914265398e-06,
+      "loss": 0.6068,
+      "step": 4883
+    },
+    {
+      "epoch": 1.7148876404494382,
+      "grad_norm": 0.9047470092773438,
+      "learning_rate": 4.083793227099272e-06,
+      "loss": 0.6484,
+      "step": 4884
+    },
+    {
+      "epoch": 1.7152387640449438,
+      "grad_norm": 0.8786801099777222,
+      "learning_rate": 4.0834354857687735e-06,
+      "loss": 0.6395,
+      "step": 4885
+    },
+    {
+      "epoch": 1.7155898876404494,
+      "grad_norm": 0.821982741355896,
+      "learning_rate": 4.0830776902861354e-06,
+      "loss": 0.5653,
+      "step": 4886
+    },
+    {
+      "epoch": 1.7159410112359552,
+      "grad_norm": 0.8869472742080688,
+      "learning_rate": 4.082719840663596e-06,
+      "loss": 0.6045,
+      "step": 4887
+    },
+    {
+      "epoch": 1.7162921348314608,
+      "grad_norm": 0.8803991079330444,
+      "learning_rate": 4.082361936913393e-06,
+      "loss": 0.5996,
+      "step": 4888
+    },
+    {
+      "epoch": 1.7166432584269664,
+      "grad_norm": 0.8931458592414856,
+      "learning_rate": 4.0820039790477635e-06,
+      "loss": 0.6376,
+      "step": 4889
+    },
+    {
+      "epoch": 1.716994382022472,
+      "grad_norm": 0.8802606463432312,
+      "learning_rate": 4.081645967078953e-06,
+      "loss": 0.6036,
+      "step": 4890
+    },
+    {
+      "epoch": 1.7173455056179776,
+      "grad_norm": 0.8540336489677429,
+      "learning_rate": 4.081287901019204e-06,
+      "loss": 0.6039,
+      "step": 4891
+    },
+    {
+      "epoch": 1.7176966292134832,
+      "grad_norm": 0.8872647285461426,
+      "learning_rate": 4.080929780880762e-06,
+      "loss": 0.6198,
+      "step": 4892
+    },
+    {
+      "epoch": 1.7180477528089888,
+      "grad_norm": 0.858366072177887,
+      "learning_rate": 4.080571606675875e-06,
+      "loss": 0.5973,
+      "step": 4893
+    },
+    {
+      "epoch": 1.7183988764044944,
+      "grad_norm": 0.8896951675415039,
+      "learning_rate": 4.08021337841679e-06,
+      "loss": 0.6307,
+      "step": 4894
+    },
+    {
+      "epoch": 1.71875,
+      "grad_norm": 0.8642908930778503,
+      "learning_rate": 4.07985509611576e-06,
+      "loss": 0.6158,
+      "step": 4895
+    },
+    {
+      "epoch": 1.7191011235955056,
+      "grad_norm": 0.9126440286636353,
+      "learning_rate": 4.079496759785039e-06,
+      "loss": 0.6595,
+      "step": 4896
+    },
+    {
+      "epoch": 1.7194522471910112,
+      "grad_norm": 0.882386326789856,
+      "learning_rate": 4.0791383694368795e-06,
+      "loss": 0.5986,
+      "step": 4897
+    },
+    {
+      "epoch": 1.7198033707865168,
+      "grad_norm": 0.846931517124176,
+      "learning_rate": 4.078779925083538e-06,
+      "loss": 0.6443,
+      "step": 4898
+    },
+    {
+      "epoch": 1.7201544943820224,
+      "grad_norm": 0.8972097635269165,
+      "learning_rate": 4.0784214267372755e-06,
+      "loss": 0.5873,
+      "step": 4899
+    },
+    {
+      "epoch": 1.720505617977528,
+      "grad_norm": 0.8785168528556824,
+      "learning_rate": 4.07806287441035e-06,
+      "loss": 0.6073,
+      "step": 4900
+    },
+    {
+      "epoch": 1.7208567415730336,
+      "grad_norm": 0.8814970254898071,
+      "learning_rate": 4.077704268115025e-06,
+      "loss": 0.6517,
+      "step": 4901
+    },
+    {
+      "epoch": 1.7212078651685392,
+      "grad_norm": 0.8843238353729248,
+      "learning_rate": 4.0773456078635635e-06,
+      "loss": 0.6437,
+      "step": 4902
+    },
+    {
+      "epoch": 1.7215589887640448,
+      "grad_norm": 0.8787217736244202,
+      "learning_rate": 4.076986893668233e-06,
+      "loss": 0.667,
+      "step": 4903
+    },
+    {
+      "epoch": 1.7219101123595506,
+      "grad_norm": 0.860560953617096,
+      "learning_rate": 4.076628125541299e-06,
+      "loss": 0.5985,
+      "step": 4904
+    },
+    {
+      "epoch": 1.7222612359550562,
+      "grad_norm": 0.8581588268280029,
+      "learning_rate": 4.076269303495033e-06,
+      "loss": 0.5936,
+      "step": 4905
+    },
+    {
+      "epoch": 1.7226123595505618,
+      "grad_norm": 0.8703935742378235,
+      "learning_rate": 4.0759104275417056e-06,
+      "loss": 0.6099,
+      "step": 4906
+    },
+    {
+      "epoch": 1.7229634831460674,
+      "grad_norm": 0.8954237103462219,
+      "learning_rate": 4.07555149769359e-06,
+      "loss": 0.5983,
+      "step": 4907
+    },
+    {
+      "epoch": 1.723314606741573,
+      "grad_norm": 0.8742859363555908,
+      "learning_rate": 4.075192513962962e-06,
+      "loss": 0.6193,
+      "step": 4908
+    },
+    {
+      "epoch": 1.7236657303370788,
+      "grad_norm": 0.865284264087677,
+      "learning_rate": 4.0748334763620975e-06,
+      "loss": 0.6157,
+      "step": 4909
+    },
+    {
+      "epoch": 1.7240168539325844,
+      "grad_norm": 0.8605691194534302,
+      "learning_rate": 4.074474384903275e-06,
+      "loss": 0.622,
+      "step": 4910
+    },
+    {
+      "epoch": 1.72436797752809,
+      "grad_norm": 0.8634774088859558,
+      "learning_rate": 4.0741152395987765e-06,
+      "loss": 0.595,
+      "step": 4911
+    },
+    {
+      "epoch": 1.7247191011235956,
+      "grad_norm": 0.8784124851226807,
+      "learning_rate": 4.073756040460884e-06,
+      "loss": 0.6459,
+      "step": 4912
+    },
+    {
+      "epoch": 1.7250702247191012,
+      "grad_norm": 0.8804491758346558,
+      "learning_rate": 4.073396787501881e-06,
+      "loss": 0.6502,
+      "step": 4913
+    },
+    {
+      "epoch": 1.7254213483146068,
+      "grad_norm": 0.8847582340240479,
+      "learning_rate": 4.073037480734054e-06,
+      "loss": 0.6206,
+      "step": 4914
+    },
+    {
+      "epoch": 1.7257724719101124,
+      "grad_norm": 0.8761957287788391,
+      "learning_rate": 4.072678120169692e-06,
+      "loss": 0.6005,
+      "step": 4915
+    },
+    {
+      "epoch": 1.726123595505618,
+      "grad_norm": 0.8767578601837158,
+      "learning_rate": 4.0723187058210845e-06,
+      "loss": 0.6233,
+      "step": 4916
+    },
+    {
+      "epoch": 1.7264747191011236,
+      "grad_norm": 0.8684590458869934,
+      "learning_rate": 4.071959237700522e-06,
+      "loss": 0.6048,
+      "step": 4917
+    },
+    {
+      "epoch": 1.7268258426966292,
+      "grad_norm": 0.8713444471359253,
+      "learning_rate": 4.071599715820299e-06,
+      "loss": 0.6331,
+      "step": 4918
+    },
+    {
+      "epoch": 1.7271769662921348,
+      "grad_norm": 0.8798062205314636,
+      "learning_rate": 4.07124014019271e-06,
+      "loss": 0.5935,
+      "step": 4919
+    },
+    {
+      "epoch": 1.7275280898876404,
+      "grad_norm": 0.892004668712616,
+      "learning_rate": 4.070880510830053e-06,
+      "loss": 0.6378,
+      "step": 4920
+    },
+    {
+      "epoch": 1.727879213483146,
+      "grad_norm": 0.8701374530792236,
+      "learning_rate": 4.070520827744627e-06,
+      "loss": 0.6012,
+      "step": 4921
+    },
+    {
+      "epoch": 1.7282303370786516,
+      "grad_norm": 0.8508331179618835,
+      "learning_rate": 4.0701610909487325e-06,
+      "loss": 0.6149,
+      "step": 4922
+    },
+    {
+      "epoch": 1.7285814606741572,
+      "grad_norm": 0.8751832842826843,
+      "learning_rate": 4.0698013004546725e-06,
+      "loss": 0.6548,
+      "step": 4923
+    },
+    {
+      "epoch": 1.7289325842696628,
+      "grad_norm": 0.8840449452400208,
+      "learning_rate": 4.06944145627475e-06,
+      "loss": 0.6347,
+      "step": 4924
+    },
+    {
+      "epoch": 1.7292837078651684,
+      "grad_norm": 0.8908770084381104,
+      "learning_rate": 4.069081558421274e-06,
+      "loss": 0.6504,
+      "step": 4925
+    },
+    {
+      "epoch": 1.7296348314606742,
+      "grad_norm": 0.9205425977706909,
+      "learning_rate": 4.068721606906552e-06,
+      "loss": 0.673,
+      "step": 4926
+    },
+    {
+      "epoch": 1.7299859550561798,
+      "grad_norm": 0.8873515725135803,
+      "learning_rate": 4.068361601742892e-06,
+      "loss": 0.6288,
+      "step": 4927
+    },
+    {
+      "epoch": 1.7303370786516854,
+      "grad_norm": 0.8750460743904114,
+      "learning_rate": 4.068001542942608e-06,
+      "loss": 0.6161,
+      "step": 4928
+    },
+    {
+      "epoch": 1.730688202247191,
+      "grad_norm": 0.8630272746086121,
+      "learning_rate": 4.067641430518013e-06,
+      "loss": 0.603,
+      "step": 4929
+    },
+    {
+      "epoch": 1.7310393258426966,
+      "grad_norm": 0.8792911171913147,
+      "learning_rate": 4.067281264481422e-06,
+      "loss": 0.6615,
+      "step": 4930
+    },
+    {
+      "epoch": 1.7313904494382022,
+      "grad_norm": 0.8867348432540894,
+      "learning_rate": 4.066921044845153e-06,
+      "loss": 0.623,
+      "step": 4931
+    },
+    {
+      "epoch": 1.731741573033708,
+      "grad_norm": 0.8691920042037964,
+      "learning_rate": 4.066560771621525e-06,
+      "loss": 0.6306,
+      "step": 4932
+    },
+    {
+      "epoch": 1.7320926966292136,
+      "grad_norm": 0.8616251349449158,
+      "learning_rate": 4.06620044482286e-06,
+      "loss": 0.594,
+      "step": 4933
+    },
+    {
+      "epoch": 1.7324438202247192,
+      "grad_norm": 0.8571128845214844,
+      "learning_rate": 4.065840064461479e-06,
+      "loss": 0.5968,
+      "step": 4934
+    },
+    {
+      "epoch": 1.7327949438202248,
+      "grad_norm": 0.8640409111976624,
+      "learning_rate": 4.065479630549709e-06,
+      "loss": 0.6371,
+      "step": 4935
+    },
+    {
+      "epoch": 1.7331460674157304,
+      "grad_norm": 0.8730684518814087,
+      "learning_rate": 4.065119143099874e-06,
+      "loss": 0.5974,
+      "step": 4936
+    },
+    {
+      "epoch": 1.733497191011236,
+      "grad_norm": 0.9529215097427368,
+      "learning_rate": 4.064758602124304e-06,
+      "loss": 0.6094,
+      "step": 4937
+    },
+    {
+      "epoch": 1.7338483146067416,
+      "grad_norm": 0.8958554863929749,
+      "learning_rate": 4.0643980076353286e-06,
+      "loss": 0.6162,
+      "step": 4938
+    },
+    {
+      "epoch": 1.7341994382022472,
+      "grad_norm": 0.9163009524345398,
+      "learning_rate": 4.06403735964528e-06,
+      "loss": 0.6386,
+      "step": 4939
+    },
+    {
+      "epoch": 1.7345505617977528,
+      "grad_norm": 0.9188485145568848,
+      "learning_rate": 4.063676658166492e-06,
+      "loss": 0.6402,
+      "step": 4940
+    },
+    {
+      "epoch": 1.7349016853932584,
+      "grad_norm": 0.8810327053070068,
+      "learning_rate": 4.063315903211301e-06,
+      "loss": 0.5978,
+      "step": 4941
+    },
+    {
+      "epoch": 1.735252808988764,
+      "grad_norm": 0.8675046563148499,
+      "learning_rate": 4.062955094792043e-06,
+      "loss": 0.602,
+      "step": 4942
+    },
+    {
+      "epoch": 1.7356039325842696,
+      "grad_norm": 0.8730883002281189,
+      "learning_rate": 4.062594232921059e-06,
+      "loss": 0.6152,
+      "step": 4943
+    },
+    {
+      "epoch": 1.7359550561797752,
+      "grad_norm": 0.8923463225364685,
+      "learning_rate": 4.062233317610689e-06,
+      "loss": 0.5889,
+      "step": 4944
+    },
+    {
+      "epoch": 1.7363061797752808,
+      "grad_norm": 0.8387205004692078,
+      "learning_rate": 4.061872348873277e-06,
+      "loss": 0.5848,
+      "step": 4945
+    },
+    {
+      "epoch": 1.7366573033707864,
+      "grad_norm": 0.8802045583724976,
+      "learning_rate": 4.061511326721166e-06,
+      "loss": 0.5933,
+      "step": 4946
+    },
+    {
+      "epoch": 1.737008426966292,
+      "grad_norm": 0.8678444623947144,
+      "learning_rate": 4.061150251166705e-06,
+      "loss": 0.6301,
+      "step": 4947
+    },
+    {
+      "epoch": 1.7373595505617978,
+      "grad_norm": 0.8504912853240967,
+      "learning_rate": 4.0607891222222406e-06,
+      "loss": 0.6174,
+      "step": 4948
+    },
+    {
+      "epoch": 1.7377106741573034,
+      "grad_norm": 0.8655530214309692,
+      "learning_rate": 4.060427939900123e-06,
+      "loss": 0.5946,
+      "step": 4949
+    },
+    {
+      "epoch": 1.738061797752809,
+      "grad_norm": 0.8756310939788818,
+      "learning_rate": 4.060066704212706e-06,
+      "loss": 0.6013,
+      "step": 4950
+    },
+    {
+      "epoch": 1.7384129213483146,
+      "grad_norm": 0.8152898550033569,
+      "learning_rate": 4.059705415172343e-06,
+      "loss": 0.5471,
+      "step": 4951
+    },
+    {
+      "epoch": 1.7387640449438202,
+      "grad_norm": 0.9165331721305847,
+      "learning_rate": 4.059344072791389e-06,
+      "loss": 0.6438,
+      "step": 4952
+    },
+    {
+      "epoch": 1.7391151685393258,
+      "grad_norm": 0.8528552055358887,
+      "learning_rate": 4.058982677082203e-06,
+      "loss": 0.5829,
+      "step": 4953
+    },
+    {
+      "epoch": 1.7394662921348316,
+      "grad_norm": 0.8451977968215942,
+      "learning_rate": 4.0586212280571424e-06,
+      "loss": 0.5632,
+      "step": 4954
+    },
+    {
+      "epoch": 1.7398174157303372,
+      "grad_norm": 0.8505921363830566,
+      "learning_rate": 4.0582597257285695e-06,
+      "loss": 0.6139,
+      "step": 4955
+    },
+    {
+      "epoch": 1.7401685393258428,
+      "grad_norm": 0.9377829432487488,
+      "learning_rate": 4.0578981701088485e-06,
+      "loss": 0.6527,
+      "step": 4956
+    },
+    {
+      "epoch": 1.7405196629213484,
+      "grad_norm": 0.914893388748169,
+      "learning_rate": 4.0575365612103425e-06,
+      "loss": 0.6147,
+      "step": 4957
+    },
+    {
+      "epoch": 1.740870786516854,
+      "grad_norm": 0.8996618986129761,
+      "learning_rate": 4.057174899045418e-06,
+      "loss": 0.6074,
+      "step": 4958
+    },
+    {
+      "epoch": 1.7412219101123596,
+      "grad_norm": 0.8936229348182678,
+      "learning_rate": 4.056813183626446e-06,
+      "loss": 0.5993,
+      "step": 4959
+    },
+    {
+      "epoch": 1.7415730337078652,
+      "grad_norm": 0.899512767791748,
+      "learning_rate": 4.056451414965795e-06,
+      "loss": 0.6188,
+      "step": 4960
+    },
+    {
+      "epoch": 1.7419241573033708,
+      "grad_norm": 0.8578722476959229,
+      "learning_rate": 4.056089593075837e-06,
+      "loss": 0.5732,
+      "step": 4961
+    },
+    {
+      "epoch": 1.7422752808988764,
+      "grad_norm": 0.8505608439445496,
+      "learning_rate": 4.055727717968946e-06,
+      "loss": 0.5933,
+      "step": 4962
+    },
+    {
+      "epoch": 1.742626404494382,
+      "grad_norm": 0.8775436878204346,
+      "learning_rate": 4.0553657896575e-06,
+      "loss": 0.604,
+      "step": 4963
+    },
+    {
+      "epoch": 1.7429775280898876,
+      "grad_norm": 0.8631317019462585,
+      "learning_rate": 4.055003808153873e-06,
+      "loss": 0.6071,
+      "step": 4964
+    },
+    {
+      "epoch": 1.7433286516853932,
+      "grad_norm": 0.8578945398330688,
+      "learning_rate": 4.054641773470447e-06,
+      "loss": 0.6116,
+      "step": 4965
+    },
+    {
+      "epoch": 1.7436797752808988,
+      "grad_norm": 0.8524402976036072,
+      "learning_rate": 4.054279685619603e-06,
+      "loss": 0.6225,
+      "step": 4966
+    },
+    {
+      "epoch": 1.7440308988764044,
+      "grad_norm": 0.8535289168357849,
+      "learning_rate": 4.053917544613723e-06,
+      "loss": 0.6195,
+      "step": 4967
+    },
+    {
+      "epoch": 1.74438202247191,
+      "grad_norm": 0.8799499869346619,
+      "learning_rate": 4.0535553504651935e-06,
+      "loss": 0.6221,
+      "step": 4968
+    },
+    {
+      "epoch": 1.7447331460674156,
+      "grad_norm": 0.8846419453620911,
+      "learning_rate": 4.053193103186401e-06,
+      "loss": 0.5955,
+      "step": 4969
+    },
+    {
+      "epoch": 1.7450842696629212,
+      "grad_norm": 0.8822450637817383,
+      "learning_rate": 4.052830802789732e-06,
+      "loss": 0.6114,
+      "step": 4970
+    },
+    {
+      "epoch": 1.745435393258427,
+      "grad_norm": 0.8850097060203552,
+      "learning_rate": 4.0524684492875795e-06,
+      "loss": 0.6098,
+      "step": 4971
+    },
+    {
+      "epoch": 1.7457865168539326,
+      "grad_norm": 0.9146633744239807,
+      "learning_rate": 4.0521060426923345e-06,
+      "loss": 0.6262,
+      "step": 4972
+    },
+    {
+      "epoch": 1.7461376404494382,
+      "grad_norm": 0.8852578997612,
+      "learning_rate": 4.05174358301639e-06,
+      "loss": 0.6145,
+      "step": 4973
+    },
+    {
+      "epoch": 1.7464887640449438,
+      "grad_norm": 0.9027591943740845,
+      "learning_rate": 4.051381070272143e-06,
+      "loss": 0.6252,
+      "step": 4974
+    },
+    {
+      "epoch": 1.7468398876404494,
+      "grad_norm": 0.8800673484802246,
+      "learning_rate": 4.051018504471992e-06,
+      "loss": 0.616,
+      "step": 4975
+    },
+    {
+      "epoch": 1.7471910112359552,
+      "grad_norm": 0.8417078256607056,
+      "learning_rate": 4.050655885628335e-06,
+      "loss": 0.5869,
+      "step": 4976
+    },
+    {
+      "epoch": 1.7475421348314608,
+      "grad_norm": 0.8745079040527344,
+      "learning_rate": 4.050293213753573e-06,
+      "loss": 0.6215,
+      "step": 4977
+    },
+    {
+      "epoch": 1.7478932584269664,
+      "grad_norm": 0.8856931924819946,
+      "learning_rate": 4.04993048886011e-06,
+      "loss": 0.6311,
+      "step": 4978
+    },
+    {
+      "epoch": 1.748244382022472,
+      "grad_norm": 0.868653416633606,
+      "learning_rate": 4.04956771096035e-06,
+      "loss": 0.6088,
+      "step": 4979
+    },
+    {
+      "epoch": 1.7485955056179776,
+      "grad_norm": 0.9026139974594116,
+      "learning_rate": 4.049204880066701e-06,
+      "loss": 0.6032,
+      "step": 4980
+    },
+    {
+      "epoch": 1.7489466292134832,
+      "grad_norm": 1.1023459434509277,
+      "learning_rate": 4.0488419961915704e-06,
+      "loss": 0.6069,
+      "step": 4981
+    },
+    {
+      "epoch": 1.7492977528089888,
+      "grad_norm": 0.9017621278762817,
+      "learning_rate": 4.048479059347369e-06,
+      "loss": 0.652,
+      "step": 4982
+    },
+    {
+      "epoch": 1.7496488764044944,
+      "grad_norm": 0.8769750595092773,
+      "learning_rate": 4.048116069546509e-06,
+      "loss": 0.633,
+      "step": 4983
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8825761675834656,
+      "learning_rate": 4.047753026801404e-06,
+      "loss": 0.63,
+      "step": 4984
+    },
+    {
+      "epoch": 1.7503511235955056,
+      "grad_norm": 0.8574188947677612,
+      "learning_rate": 4.04738993112447e-06,
+      "loss": 0.6121,
+      "step": 4985
+    },
+    {
+      "epoch": 1.7507022471910112,
+      "grad_norm": 0.871944785118103,
+      "learning_rate": 4.047026782528125e-06,
+      "loss": 0.6116,
+      "step": 4986
+    },
+    {
+      "epoch": 1.7510533707865168,
+      "grad_norm": 0.8469028472900391,
+      "learning_rate": 4.046663581024787e-06,
+      "loss": 0.5971,
+      "step": 4987
+    },
+    {
+      "epoch": 1.7514044943820224,
+      "grad_norm": 0.8719568252563477,
+      "learning_rate": 4.046300326626878e-06,
+      "loss": 0.5941,
+      "step": 4988
+    },
+    {
+      "epoch": 1.751755617977528,
+      "grad_norm": 0.879953920841217,
+      "learning_rate": 4.045937019346822e-06,
+      "loss": 0.6169,
+      "step": 4989
+    },
+    {
+      "epoch": 1.7521067415730336,
+      "grad_norm": 0.8611469268798828,
+      "learning_rate": 4.045573659197042e-06,
+      "loss": 0.6287,
+      "step": 4990
+    },
+    {
+      "epoch": 1.7524578651685392,
+      "grad_norm": 0.8662636876106262,
+      "learning_rate": 4.045210246189966e-06,
+      "loss": 0.6159,
+      "step": 4991
+    },
+    {
+      "epoch": 1.7528089887640448,
+      "grad_norm": 0.874043881893158,
+      "learning_rate": 4.044846780338021e-06,
+      "loss": 0.6005,
+      "step": 4992
+    },
+    {
+      "epoch": 1.7531601123595506,
+      "grad_norm": 0.8485775589942932,
+      "learning_rate": 4.0444832616536385e-06,
+      "loss": 0.6224,
+      "step": 4993
+    },
+    {
+      "epoch": 1.7535112359550562,
+      "grad_norm": 0.8543581962585449,
+      "learning_rate": 4.04411969014925e-06,
+      "loss": 0.5885,
+      "step": 4994
+    },
+    {
+      "epoch": 1.7538623595505618,
+      "grad_norm": 0.8683145642280579,
+      "learning_rate": 4.043756065837289e-06,
+      "loss": 0.6394,
+      "step": 4995
+    },
+    {
+      "epoch": 1.7542134831460674,
+      "grad_norm": 0.8860946893692017,
+      "learning_rate": 4.043392388730192e-06,
+      "loss": 0.6372,
+      "step": 4996
+    },
+    {
+      "epoch": 1.754564606741573,
+      "grad_norm": 0.8612002730369568,
+      "learning_rate": 4.043028658840397e-06,
+      "loss": 0.6136,
+      "step": 4997
+    },
+    {
+      "epoch": 1.7549157303370788,
+      "grad_norm": 0.8517924547195435,
+      "learning_rate": 4.042664876180341e-06,
+      "loss": 0.5788,
+      "step": 4998
+    },
+    {
+      "epoch": 1.7552668539325844,
+      "grad_norm": 0.8405445218086243,
+      "learning_rate": 4.042301040762467e-06,
+      "loss": 0.5945,
+      "step": 4999
+    },
+    {
+      "epoch": 1.75561797752809,
+      "grad_norm": 0.8536641597747803,
+      "learning_rate": 4.041937152599216e-06,
+      "loss": 0.5888,
+      "step": 5000
+    },
+    {
+      "epoch": 1.7559691011235956,
+      "grad_norm": 0.9032542705535889,
+      "learning_rate": 4.041573211703035e-06,
+      "loss": 0.6133,
+      "step": 5001
+    },
+    {
+      "epoch": 1.7563202247191012,
+      "grad_norm": 0.8637306094169617,
+      "learning_rate": 4.041209218086369e-06,
+      "loss": 0.6245,
+      "step": 5002
+    },
+    {
+      "epoch": 1.7566713483146068,
+      "grad_norm": 0.8811326026916504,
+      "learning_rate": 4.0408451717616656e-06,
+      "loss": 0.6569,
+      "step": 5003
+    },
+    {
+      "epoch": 1.7570224719101124,
+      "grad_norm": 0.8908126950263977,
+      "learning_rate": 4.040481072741378e-06,
+      "loss": 0.6469,
+      "step": 5004
+    },
+    {
+      "epoch": 1.757373595505618,
+      "grad_norm": 0.886318564414978,
+      "learning_rate": 4.040116921037954e-06,
+      "loss": 0.5808,
+      "step": 5005
+    },
+    {
+      "epoch": 1.7577247191011236,
+      "grad_norm": 0.8739534616470337,
+      "learning_rate": 4.03975271666385e-06,
+      "loss": 0.5976,
+      "step": 5006
+    },
+    {
+      "epoch": 1.7580758426966292,
+      "grad_norm": 0.9307764172554016,
+      "learning_rate": 4.03938845963152e-06,
+      "loss": 0.6418,
+      "step": 5007
+    },
+    {
+      "epoch": 1.7584269662921348,
+      "grad_norm": 0.861684262752533,
+      "learning_rate": 4.039024149953424e-06,
+      "loss": 0.6109,
+      "step": 5008
+    },
+    {
+      "epoch": 1.7587780898876404,
+      "grad_norm": 0.8454620838165283,
+      "learning_rate": 4.038659787642016e-06,
+      "loss": 0.5839,
+      "step": 5009
+    },
+    {
+      "epoch": 1.759129213483146,
+      "grad_norm": 0.8540975451469421,
+      "learning_rate": 4.038295372709762e-06,
+      "loss": 0.6302,
+      "step": 5010
+    },
+    {
+      "epoch": 1.7594803370786516,
+      "grad_norm": 0.8659608364105225,
+      "learning_rate": 4.0379309051691225e-06,
+      "loss": 0.6305,
+      "step": 5011
+    },
+    {
+      "epoch": 1.7598314606741572,
+      "grad_norm": 0.8791552782058716,
+      "learning_rate": 4.037566385032561e-06,
+      "loss": 0.6323,
+      "step": 5012
+    },
+    {
+      "epoch": 1.7601825842696628,
+      "grad_norm": 0.8560447692871094,
+      "learning_rate": 4.037201812312545e-06,
+      "loss": 0.5957,
+      "step": 5013
+    },
+    {
+      "epoch": 1.7605337078651684,
+      "grad_norm": 0.8539266586303711,
+      "learning_rate": 4.036837187021544e-06,
+      "loss": 0.6113,
+      "step": 5014
+    },
+    {
+      "epoch": 1.7608848314606742,
+      "grad_norm": 0.9080570340156555,
+      "learning_rate": 4.036472509172025e-06,
+      "loss": 0.6168,
+      "step": 5015
+    },
+    {
+      "epoch": 1.7612359550561798,
+      "grad_norm": 0.8955601453781128,
+      "learning_rate": 4.036107778776461e-06,
+      "loss": 0.6174,
+      "step": 5016
+    },
+    {
+      "epoch": 1.7615870786516854,
+      "grad_norm": 0.8747341632843018,
+      "learning_rate": 4.0357429958473256e-06,
+      "loss": 0.6357,
+      "step": 5017
+    },
+    {
+      "epoch": 1.761938202247191,
+      "grad_norm": 0.8803605437278748,
+      "learning_rate": 4.035378160397094e-06,
+      "loss": 0.5907,
+      "step": 5018
+    },
+    {
+      "epoch": 1.7622893258426966,
+      "grad_norm": 0.8748698234558105,
+      "learning_rate": 4.035013272438243e-06,
+      "loss": 0.6121,
+      "step": 5019
+    },
+    {
+      "epoch": 1.7626404494382022,
+      "grad_norm": 0.9040994048118591,
+      "learning_rate": 4.034648331983251e-06,
+      "loss": 0.6346,
+      "step": 5020
+    },
+    {
+      "epoch": 1.762991573033708,
+      "grad_norm": 0.8952014446258545,
+      "learning_rate": 4.0342833390446e-06,
+      "loss": 0.6016,
+      "step": 5021
+    },
+    {
+      "epoch": 1.7633426966292136,
+      "grad_norm": 0.8983685374259949,
+      "learning_rate": 4.033918293634772e-06,
+      "loss": 0.6257,
+      "step": 5022
+    },
+    {
+      "epoch": 1.7636938202247192,
+      "grad_norm": 0.8786833882331848,
+      "learning_rate": 4.03355319576625e-06,
+      "loss": 0.6118,
+      "step": 5023
+    },
+    {
+      "epoch": 1.7640449438202248,
+      "grad_norm": 0.8733046054840088,
+      "learning_rate": 4.033188045451522e-06,
+      "loss": 0.6078,
+      "step": 5024
+    },
+    {
+      "epoch": 1.7643960674157304,
+      "grad_norm": 0.8754127025604248,
+      "learning_rate": 4.032822842703075e-06,
+      "loss": 0.6507,
+      "step": 5025
+    },
+    {
+      "epoch": 1.764747191011236,
+      "grad_norm": 0.8841948509216309,
+      "learning_rate": 4.032457587533397e-06,
+      "loss": 0.6371,
+      "step": 5026
+    },
+    {
+      "epoch": 1.7650983146067416,
+      "grad_norm": 0.8662333488464355,
+      "learning_rate": 4.032092279954982e-06,
+      "loss": 0.5939,
+      "step": 5027
+    },
+    {
+      "epoch": 1.7654494382022472,
+      "grad_norm": 0.8241957426071167,
+      "learning_rate": 4.031726919980321e-06,
+      "loss": 0.5896,
+      "step": 5028
+    },
+    {
+      "epoch": 1.7658005617977528,
+      "grad_norm": 0.8761467933654785,
+      "learning_rate": 4.0313615076219105e-06,
+      "loss": 0.6228,
+      "step": 5029
+    },
+    {
+      "epoch": 1.7661516853932584,
+      "grad_norm": 0.8869350552558899,
+      "learning_rate": 4.030996042892248e-06,
+      "loss": 0.6509,
+      "step": 5030
+    },
+    {
+      "epoch": 1.766502808988764,
+      "grad_norm": 0.8962212204933167,
+      "learning_rate": 4.030630525803829e-06,
+      "loss": 0.6377,
+      "step": 5031
+    },
+    {
+      "epoch": 1.7668539325842696,
+      "grad_norm": 0.8528103828430176,
+      "learning_rate": 4.030264956369158e-06,
+      "loss": 0.5963,
+      "step": 5032
+    },
+    {
+      "epoch": 1.7672050561797752,
+      "grad_norm": 0.8822567462921143,
+      "learning_rate": 4.029899334600733e-06,
+      "loss": 0.6135,
+      "step": 5033
+    },
+    {
+      "epoch": 1.7675561797752808,
+      "grad_norm": 0.846155047416687,
+      "learning_rate": 4.029533660511061e-06,
+      "loss": 0.6057,
+      "step": 5034
+    },
+    {
+      "epoch": 1.7679073033707864,
+      "grad_norm": 0.9543619751930237,
+      "learning_rate": 4.029167934112646e-06,
+      "loss": 0.6142,
+      "step": 5035
+    },
+    {
+      "epoch": 1.768258426966292,
+      "grad_norm": 0.9285796880722046,
+      "learning_rate": 4.028802155417997e-06,
+      "loss": 0.6105,
+      "step": 5036
+    },
+    {
+      "epoch": 1.7686095505617978,
+      "grad_norm": 0.857733964920044,
+      "learning_rate": 4.028436324439622e-06,
+      "loss": 0.5895,
+      "step": 5037
+    },
+    {
+      "epoch": 1.7689606741573034,
+      "grad_norm": 0.8516778349876404,
+      "learning_rate": 4.028070441190032e-06,
+      "loss": 0.592,
+      "step": 5038
+    },
+    {
+      "epoch": 1.769311797752809,
+      "grad_norm": 0.8570651412010193,
+      "learning_rate": 4.027704505681741e-06,
+      "loss": 0.6106,
+      "step": 5039
+    },
+    {
+      "epoch": 1.7696629213483146,
+      "grad_norm": 0.8978039622306824,
+      "learning_rate": 4.027338517927264e-06,
+      "loss": 0.6333,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7700140449438202,
+      "grad_norm": 0.8815526366233826,
+      "learning_rate": 4.026972477939115e-06,
+      "loss": 0.6127,
+      "step": 5041
+    },
+    {
+      "epoch": 1.7703651685393258,
+      "grad_norm": 0.8965631723403931,
+      "learning_rate": 4.026606385729816e-06,
+      "loss": 0.6252,
+      "step": 5042
+    },
+    {
+      "epoch": 1.7707162921348316,
+      "grad_norm": 0.8918066620826721,
+      "learning_rate": 4.026240241311883e-06,
+      "loss": 0.6325,
+      "step": 5043
+    },
+    {
+      "epoch": 1.7710674157303372,
+      "grad_norm": 0.8715677857398987,
+      "learning_rate": 4.025874044697841e-06,
+      "loss": 0.591,
+      "step": 5044
+    },
+    {
+      "epoch": 1.7714185393258428,
+      "grad_norm": 0.8741485476493835,
+      "learning_rate": 4.025507795900212e-06,
+      "loss": 0.6137,
+      "step": 5045
+    },
+    {
+      "epoch": 1.7717696629213484,
+      "grad_norm": 0.8711748123168945,
+      "learning_rate": 4.025141494931522e-06,
+      "loss": 0.6098,
+      "step": 5046
+    },
+    {
+      "epoch": 1.772120786516854,
+      "grad_norm": 0.9053213000297546,
+      "learning_rate": 4.024775141804297e-06,
+      "loss": 0.6198,
+      "step": 5047
+    },
+    {
+      "epoch": 1.7724719101123596,
+      "grad_norm": 0.8338163495063782,
+      "learning_rate": 4.024408736531068e-06,
+      "loss": 0.5808,
+      "step": 5048
+    },
+    {
+      "epoch": 1.7728230337078652,
+      "grad_norm": 0.8671820759773254,
+      "learning_rate": 4.024042279124364e-06,
+      "loss": 0.5916,
+      "step": 5049
+    },
+    {
+      "epoch": 1.7731741573033708,
+      "grad_norm": 0.9618306159973145,
+      "learning_rate": 4.023675769596718e-06,
+      "loss": 0.6416,
+      "step": 5050
+    },
+    {
+      "epoch": 1.7735252808988764,
+      "grad_norm": 0.8482176065444946,
+      "learning_rate": 4.023309207960666e-06,
+      "loss": 0.5838,
+      "step": 5051
+    },
+    {
+      "epoch": 1.773876404494382,
+      "grad_norm": 0.891528844833374,
+      "learning_rate": 4.0229425942287414e-06,
+      "loss": 0.6196,
+      "step": 5052
+    },
+    {
+      "epoch": 1.7742275280898876,
+      "grad_norm": 0.8804656267166138,
+      "learning_rate": 4.0225759284134835e-06,
+      "loss": 0.5967,
+      "step": 5053
+    },
+    {
+      "epoch": 1.7745786516853932,
+      "grad_norm": 0.863013744354248,
+      "learning_rate": 4.022209210527432e-06,
+      "loss": 0.6211,
+      "step": 5054
+    },
+    {
+      "epoch": 1.7749297752808988,
+      "grad_norm": 0.851879358291626,
+      "learning_rate": 4.021842440583127e-06,
+      "loss": 0.6305,
+      "step": 5055
+    },
+    {
+      "epoch": 1.7752808988764044,
+      "grad_norm": 0.8675127029418945,
+      "learning_rate": 4.0214756185931135e-06,
+      "loss": 0.6076,
+      "step": 5056
+    },
+    {
+      "epoch": 1.77563202247191,
+      "grad_norm": 0.8953244090080261,
+      "learning_rate": 4.021108744569936e-06,
+      "loss": 0.6286,
+      "step": 5057
+    },
+    {
+      "epoch": 1.7759831460674156,
+      "grad_norm": 0.8707031011581421,
+      "learning_rate": 4.020741818526141e-06,
+      "loss": 0.638,
+      "step": 5058
+    },
+    {
+      "epoch": 1.7763342696629212,
+      "grad_norm": 0.9085872173309326,
+      "learning_rate": 4.020374840474277e-06,
+      "loss": 0.6537,
+      "step": 5059
+    },
+    {
+      "epoch": 1.776685393258427,
+      "grad_norm": 0.8786725401878357,
+      "learning_rate": 4.020007810426895e-06,
+      "loss": 0.6362,
+      "step": 5060
+    },
+    {
+      "epoch": 1.7770365168539326,
+      "grad_norm": 0.889363706111908,
+      "learning_rate": 4.019640728396545e-06,
+      "loss": 0.6149,
+      "step": 5061
+    },
+    {
+      "epoch": 1.7773876404494382,
+      "grad_norm": 0.9281964302062988,
+      "learning_rate": 4.019273594395784e-06,
+      "loss": 0.6798,
+      "step": 5062
+    },
+    {
+      "epoch": 1.7777387640449438,
+      "grad_norm": 0.9001464247703552,
+      "learning_rate": 4.018906408437166e-06,
+      "loss": 0.6325,
+      "step": 5063
+    },
+    {
+      "epoch": 1.7780898876404494,
+      "grad_norm": 0.8923134207725525,
+      "learning_rate": 4.018539170533248e-06,
+      "loss": 0.6288,
+      "step": 5064
+    },
+    {
+      "epoch": 1.7784410112359552,
+      "grad_norm": 0.8892466425895691,
+      "learning_rate": 4.018171880696589e-06,
+      "loss": 0.628,
+      "step": 5065
+    },
+    {
+      "epoch": 1.7787921348314608,
+      "grad_norm": 0.8730875849723816,
+      "learning_rate": 4.017804538939753e-06,
+      "loss": 0.6235,
+      "step": 5066
+    },
+    {
+      "epoch": 1.7791432584269664,
+      "grad_norm": 0.902269721031189,
+      "learning_rate": 4.0174371452752995e-06,
+      "loss": 0.6527,
+      "step": 5067
+    },
+    {
+      "epoch": 1.779494382022472,
+      "grad_norm": 0.8821526169776917,
+      "learning_rate": 4.017069699715794e-06,
+      "loss": 0.5963,
+      "step": 5068
+    },
+    {
+      "epoch": 1.7798455056179776,
+      "grad_norm": 0.8617396354675293,
+      "learning_rate": 4.016702202273802e-06,
+      "loss": 0.6127,
+      "step": 5069
+    },
+    {
+      "epoch": 1.7801966292134832,
+      "grad_norm": 0.9029756784439087,
+      "learning_rate": 4.016334652961894e-06,
+      "loss": 0.6271,
+      "step": 5070
+    },
+    {
+      "epoch": 1.7805477528089888,
+      "grad_norm": 0.8544668555259705,
+      "learning_rate": 4.015967051792638e-06,
+      "loss": 0.6131,
+      "step": 5071
+    },
+    {
+      "epoch": 1.7808988764044944,
+      "grad_norm": 0.8941280841827393,
+      "learning_rate": 4.015599398778606e-06,
+      "loss": 0.6029,
+      "step": 5072
+    },
+    {
+      "epoch": 1.78125,
+      "grad_norm": 0.9132516384124756,
+      "learning_rate": 4.015231693932372e-06,
+      "loss": 0.6349,
+      "step": 5073
+    },
+    {
+      "epoch": 1.7816011235955056,
+      "grad_norm": 0.8514044284820557,
+      "learning_rate": 4.01486393726651e-06,
+      "loss": 0.5818,
+      "step": 5074
+    },
+    {
+      "epoch": 1.7819522471910112,
+      "grad_norm": 0.8844448328018188,
+      "learning_rate": 4.014496128793598e-06,
+      "loss": 0.615,
+      "step": 5075
+    },
+    {
+      "epoch": 1.7823033707865168,
+      "grad_norm": 0.861137330532074,
+      "learning_rate": 4.014128268526214e-06,
+      "loss": 0.5902,
+      "step": 5076
+    },
+    {
+      "epoch": 1.7826544943820224,
+      "grad_norm": 0.86243736743927,
+      "learning_rate": 4.013760356476939e-06,
+      "loss": 0.6108,
+      "step": 5077
+    },
+    {
+      "epoch": 1.783005617977528,
+      "grad_norm": 0.8797655701637268,
+      "learning_rate": 4.013392392658355e-06,
+      "loss": 0.6207,
+      "step": 5078
+    },
+    {
+      "epoch": 1.7833567415730336,
+      "grad_norm": 0.9152445793151855,
+      "learning_rate": 4.013024377083046e-06,
+      "loss": 0.6307,
+      "step": 5079
+    },
+    {
+      "epoch": 1.7837078651685392,
+      "grad_norm": 0.9087762236595154,
+      "learning_rate": 4.012656309763598e-06,
+      "loss": 0.6143,
+      "step": 5080
+    },
+    {
+      "epoch": 1.7840589887640448,
+      "grad_norm": 0.8641596436500549,
+      "learning_rate": 4.0122881907126e-06,
+      "loss": 0.6173,
+      "step": 5081
+    },
+    {
+      "epoch": 1.7844101123595506,
+      "grad_norm": 0.8863744139671326,
+      "learning_rate": 4.011920019942638e-06,
+      "loss": 0.6361,
+      "step": 5082
+    },
+    {
+      "epoch": 1.7847612359550562,
+      "grad_norm": 0.8684104084968567,
+      "learning_rate": 4.011551797466306e-06,
+      "loss": 0.6034,
+      "step": 5083
+    },
+    {
+      "epoch": 1.7851123595505618,
+      "grad_norm": 0.875816822052002,
+      "learning_rate": 4.011183523296197e-06,
+      "loss": 0.6085,
+      "step": 5084
+    },
+    {
+      "epoch": 1.7854634831460674,
+      "grad_norm": 0.8723933100700378,
+      "learning_rate": 4.010815197444903e-06,
+      "loss": 0.6009,
+      "step": 5085
+    },
+    {
+      "epoch": 1.785814606741573,
+      "grad_norm": 0.8689554333686829,
+      "learning_rate": 4.010446819925023e-06,
+      "loss": 0.6131,
+      "step": 5086
+    },
+    {
+      "epoch": 1.7861657303370788,
+      "grad_norm": 0.8560565710067749,
+      "learning_rate": 4.010078390749154e-06,
+      "loss": 0.5902,
+      "step": 5087
+    },
+    {
+      "epoch": 1.7865168539325844,
+      "grad_norm": 0.8722461462020874,
+      "learning_rate": 4.009709909929897e-06,
+      "loss": 0.6335,
+      "step": 5088
+    },
+    {
+      "epoch": 1.78686797752809,
+      "grad_norm": 0.8994833827018738,
+      "learning_rate": 4.009341377479851e-06,
+      "loss": 0.6265,
+      "step": 5089
+    },
+    {
+      "epoch": 1.7872191011235956,
+      "grad_norm": 0.8879294395446777,
+      "learning_rate": 4.008972793411623e-06,
+      "loss": 0.6255,
+      "step": 5090
+    },
+    {
+      "epoch": 1.7875702247191012,
+      "grad_norm": 0.8504826426506042,
+      "learning_rate": 4.008604157737816e-06,
+      "loss": 0.6158,
+      "step": 5091
+    },
+    {
+      "epoch": 1.7879213483146068,
+      "grad_norm": 0.8719592690467834,
+      "learning_rate": 4.008235470471038e-06,
+      "loss": 0.6005,
+      "step": 5092
+    },
+    {
+      "epoch": 1.7882724719101124,
+      "grad_norm": 0.8644405007362366,
+      "learning_rate": 4.007866731623898e-06,
+      "loss": 0.5995,
+      "step": 5093
+    },
+    {
+      "epoch": 1.788623595505618,
+      "grad_norm": 0.8683051466941833,
+      "learning_rate": 4.007497941209006e-06,
+      "loss": 0.6372,
+      "step": 5094
+    },
+    {
+      "epoch": 1.7889747191011236,
+      "grad_norm": 0.8745830655097961,
+      "learning_rate": 4.007129099238973e-06,
+      "loss": 0.6063,
+      "step": 5095
+    },
+    {
+      "epoch": 1.7893258426966292,
+      "grad_norm": 0.8765552639961243,
+      "learning_rate": 4.006760205726416e-06,
+      "loss": 0.6434,
+      "step": 5096
+    },
+    {
+      "epoch": 1.7896769662921348,
+      "grad_norm": 0.8652710318565369,
+      "learning_rate": 4.00639126068395e-06,
+      "loss": 0.6183,
+      "step": 5097
+    },
+    {
+      "epoch": 1.7900280898876404,
+      "grad_norm": 0.863023042678833,
+      "learning_rate": 4.00602226412419e-06,
+      "loss": 0.6265,
+      "step": 5098
+    },
+    {
+      "epoch": 1.790379213483146,
+      "grad_norm": 0.8672717213630676,
+      "learning_rate": 4.005653216059758e-06,
+      "loss": 0.6281,
+      "step": 5099
+    },
+    {
+      "epoch": 1.7907303370786516,
+      "grad_norm": 0.8654349446296692,
+      "learning_rate": 4.005284116503275e-06,
+      "loss": 0.6355,
+      "step": 5100
+    },
+    {
+      "epoch": 1.7910814606741572,
+      "grad_norm": 0.8771569728851318,
+      "learning_rate": 4.004914965467363e-06,
+      "loss": 0.6518,
+      "step": 5101
+    },
+    {
+      "epoch": 1.7914325842696628,
+      "grad_norm": 0.9185158014297485,
+      "learning_rate": 4.004545762964646e-06,
+      "loss": 0.6309,
+      "step": 5102
+    },
+    {
+      "epoch": 1.7917837078651684,
+      "grad_norm": 0.8859367966651917,
+      "learning_rate": 4.004176509007752e-06,
+      "loss": 0.6431,
+      "step": 5103
+    },
+    {
+      "epoch": 1.7921348314606742,
+      "grad_norm": 0.8867841958999634,
+      "learning_rate": 4.0038072036093086e-06,
+      "loss": 0.618,
+      "step": 5104
+    },
+    {
+      "epoch": 1.7924859550561798,
+      "grad_norm": 0.8852524757385254,
+      "learning_rate": 4.003437846781946e-06,
+      "loss": 0.6452,
+      "step": 5105
+    },
+    {
+      "epoch": 1.7928370786516854,
+      "grad_norm": 0.8460947275161743,
+      "learning_rate": 4.003068438538295e-06,
+      "loss": 0.6402,
+      "step": 5106
+    },
+    {
+      "epoch": 1.793188202247191,
+      "grad_norm": 0.8796756267547607,
+      "learning_rate": 4.00269897889099e-06,
+      "loss": 0.6215,
+      "step": 5107
+    },
+    {
+      "epoch": 1.7935393258426966,
+      "grad_norm": 0.8622691035270691,
+      "learning_rate": 4.002329467852666e-06,
+      "loss": 0.6182,
+      "step": 5108
+    },
+    {
+      "epoch": 1.7938904494382022,
+      "grad_norm": 0.8955039978027344,
+      "learning_rate": 4.001959905435959e-06,
+      "loss": 0.6284,
+      "step": 5109
+    },
+    {
+      "epoch": 1.794241573033708,
+      "grad_norm": 0.900213360786438,
+      "learning_rate": 4.001590291653509e-06,
+      "loss": 0.6138,
+      "step": 5110
+    },
+    {
+      "epoch": 1.7945926966292136,
+      "grad_norm": 0.8492575287818909,
+      "learning_rate": 4.0012206265179555e-06,
+      "loss": 0.5906,
+      "step": 5111
+    },
+    {
+      "epoch": 1.7949438202247192,
+      "grad_norm": 0.8939003944396973,
+      "learning_rate": 4.000850910041943e-06,
+      "loss": 0.6461,
+      "step": 5112
+    },
+    {
+      "epoch": 1.7952949438202248,
+      "grad_norm": 0.8977720737457275,
+      "learning_rate": 4.000481142238112e-06,
+      "loss": 0.6255,
+      "step": 5113
+    },
+    {
+      "epoch": 1.7956460674157304,
+      "grad_norm": 0.8857546448707581,
+      "learning_rate": 4.0001113231191095e-06,
+      "loss": 0.6419,
+      "step": 5114
+    },
+    {
+      "epoch": 1.795997191011236,
+      "grad_norm": 0.8970683217048645,
+      "learning_rate": 3.999741452697584e-06,
+      "loss": 0.6544,
+      "step": 5115
+    },
+    {
+      "epoch": 1.7963483146067416,
+      "grad_norm": 0.8892328143119812,
+      "learning_rate": 3.999371530986184e-06,
+      "loss": 0.5811,
+      "step": 5116
+    },
+    {
+      "epoch": 1.7966994382022472,
+      "grad_norm": 0.8691011667251587,
+      "learning_rate": 3.999001557997561e-06,
+      "loss": 0.6118,
+      "step": 5117
+    },
+    {
+      "epoch": 1.7970505617977528,
+      "grad_norm": 0.8691001534461975,
+      "learning_rate": 3.998631533744367e-06,
+      "loss": 0.6209,
+      "step": 5118
+    },
+    {
+      "epoch": 1.7974016853932584,
+      "grad_norm": 0.8801686763763428,
+      "learning_rate": 3.998261458239257e-06,
+      "loss": 0.5993,
+      "step": 5119
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.8820109367370605,
+      "learning_rate": 3.997891331494888e-06,
+      "loss": 0.6182,
+      "step": 5120
+    },
+    {
+      "epoch": 1.7981039325842696,
+      "grad_norm": 0.891557514667511,
+      "learning_rate": 3.997521153523916e-06,
+      "loss": 0.6214,
+      "step": 5121
+    },
+    {
+      "epoch": 1.7984550561797752,
+      "grad_norm": 0.8902502059936523,
+      "learning_rate": 3.997150924339003e-06,
+      "loss": 0.6082,
+      "step": 5122
+    },
+    {
+      "epoch": 1.7988061797752808,
+      "grad_norm": 0.8698774576187134,
+      "learning_rate": 3.99678064395281e-06,
+      "loss": 0.6319,
+      "step": 5123
+    },
+    {
+      "epoch": 1.7991573033707864,
+      "grad_norm": 0.8708075284957886,
+      "learning_rate": 3.996410312377998e-06,
+      "loss": 0.5903,
+      "step": 5124
+    },
+    {
+      "epoch": 1.799508426966292,
+      "grad_norm": 0.860001266002655,
+      "learning_rate": 3.996039929627235e-06,
+      "loss": 0.6108,
+      "step": 5125
+    },
+    {
+      "epoch": 1.7998595505617978,
+      "grad_norm": 0.9757430553436279,
+      "learning_rate": 3.995669495713187e-06,
+      "loss": 0.6222,
+      "step": 5126
+    },
+    {
+      "epoch": 1.8002106741573034,
+      "grad_norm": 0.8471954464912415,
+      "learning_rate": 3.9952990106485216e-06,
+      "loss": 0.6127,
+      "step": 5127
+    },
+    {
+      "epoch": 1.800561797752809,
+      "grad_norm": 0.8945921063423157,
+      "learning_rate": 3.99492847444591e-06,
+      "loss": 0.6047,
+      "step": 5128
+    },
+    {
+      "epoch": 1.8009129213483146,
+      "grad_norm": 0.8653395175933838,
+      "learning_rate": 3.994557887118022e-06,
+      "loss": 0.6214,
+      "step": 5129
+    },
+    {
+      "epoch": 1.8012640449438202,
+      "grad_norm": 0.917481005191803,
+      "learning_rate": 3.994187248677535e-06,
+      "loss": 0.6226,
+      "step": 5130
+    },
+    {
+      "epoch": 1.8016151685393258,
+      "grad_norm": 0.862048864364624,
+      "learning_rate": 3.993816559137122e-06,
+      "loss": 0.6309,
+      "step": 5131
+    },
+    {
+      "epoch": 1.8019662921348316,
+      "grad_norm": 0.8727179765701294,
+      "learning_rate": 3.993445818509461e-06,
+      "loss": 0.602,
+      "step": 5132
+    },
+    {
+      "epoch": 1.8023174157303372,
+      "grad_norm": 0.8686352968215942,
+      "learning_rate": 3.9930750268072315e-06,
+      "loss": 0.5883,
+      "step": 5133
+    },
+    {
+      "epoch": 1.8026685393258428,
+      "grad_norm": 0.8954786658287048,
+      "learning_rate": 3.992704184043113e-06,
+      "loss": 0.6448,
+      "step": 5134
+    },
+    {
+      "epoch": 1.8030196629213484,
+      "grad_norm": 0.8747565150260925,
+      "learning_rate": 3.992333290229789e-06,
+      "loss": 0.5916,
+      "step": 5135
+    },
+    {
+      "epoch": 1.803370786516854,
+      "grad_norm": 0.905752956867218,
+      "learning_rate": 3.9919623453799436e-06,
+      "loss": 0.6148,
+      "step": 5136
+    },
+    {
+      "epoch": 1.8037219101123596,
+      "grad_norm": 0.8951407670974731,
+      "learning_rate": 3.991591349506263e-06,
+      "loss": 0.6234,
+      "step": 5137
+    },
+    {
+      "epoch": 1.8040730337078652,
+      "grad_norm": 0.87555992603302,
+      "learning_rate": 3.991220302621434e-06,
+      "loss": 0.6072,
+      "step": 5138
+    },
+    {
+      "epoch": 1.8044241573033708,
+      "grad_norm": 0.8721452355384827,
+      "learning_rate": 3.990849204738147e-06,
+      "loss": 0.6289,
+      "step": 5139
+    },
+    {
+      "epoch": 1.8047752808988764,
+      "grad_norm": 0.8791200518608093,
+      "learning_rate": 3.990478055869093e-06,
+      "loss": 0.6394,
+      "step": 5140
+    },
+    {
+      "epoch": 1.805126404494382,
+      "grad_norm": 0.8407002091407776,
+      "learning_rate": 3.990106856026965e-06,
+      "loss": 0.6079,
+      "step": 5141
+    },
+    {
+      "epoch": 1.8054775280898876,
+      "grad_norm": 0.9156894683837891,
+      "learning_rate": 3.9897356052244565e-06,
+      "loss": 0.6131,
+      "step": 5142
+    },
+    {
+      "epoch": 1.8058286516853932,
+      "grad_norm": 0.8723463416099548,
+      "learning_rate": 3.9893643034742665e-06,
+      "loss": 0.6361,
+      "step": 5143
+    },
+    {
+      "epoch": 1.8061797752808988,
+      "grad_norm": 0.8676952719688416,
+      "learning_rate": 3.988992950789091e-06,
+      "loss": 0.6159,
+      "step": 5144
+    },
+    {
+      "epoch": 1.8065308988764044,
+      "grad_norm": 0.8790785670280457,
+      "learning_rate": 3.988621547181631e-06,
+      "loss": 0.6162,
+      "step": 5145
+    },
+    {
+      "epoch": 1.80688202247191,
+      "grad_norm": 0.8970304727554321,
+      "learning_rate": 3.988250092664588e-06,
+      "loss": 0.5972,
+      "step": 5146
+    },
+    {
+      "epoch": 1.8072331460674156,
+      "grad_norm": 0.8706192970275879,
+      "learning_rate": 3.987878587250666e-06,
+      "loss": 0.6323,
+      "step": 5147
+    },
+    {
+      "epoch": 1.8075842696629212,
+      "grad_norm": 0.913070559501648,
+      "learning_rate": 3.987507030952568e-06,
+      "loss": 0.6182,
+      "step": 5148
+    },
+    {
+      "epoch": 1.807935393258427,
+      "grad_norm": 0.9056144952774048,
+      "learning_rate": 3.987135423783004e-06,
+      "loss": 0.6193,
+      "step": 5149
+    },
+    {
+      "epoch": 1.8082865168539326,
+      "grad_norm": 0.8835223317146301,
+      "learning_rate": 3.9867637657546795e-06,
+      "loss": 0.604,
+      "step": 5150
+    },
+    {
+      "epoch": 1.8086376404494382,
+      "grad_norm": 0.9045801758766174,
+      "learning_rate": 3.986392056880307e-06,
+      "loss": 0.6386,
+      "step": 5151
+    },
+    {
+      "epoch": 1.8089887640449438,
+      "grad_norm": 0.8738263845443726,
+      "learning_rate": 3.986020297172599e-06,
+      "loss": 0.6158,
+      "step": 5152
+    },
+    {
+      "epoch": 1.8093398876404494,
+      "grad_norm": 0.8748819828033447,
+      "learning_rate": 3.985648486644267e-06,
+      "loss": 0.6018,
+      "step": 5153
+    },
+    {
+      "epoch": 1.8096910112359552,
+      "grad_norm": 0.8693463802337646,
+      "learning_rate": 3.985276625308029e-06,
+      "loss": 0.6557,
+      "step": 5154
+    },
+    {
+      "epoch": 1.8100421348314608,
+      "grad_norm": 0.8723728060722351,
+      "learning_rate": 3.9849047131766015e-06,
+      "loss": 0.5882,
+      "step": 5155
+    },
+    {
+      "epoch": 1.8103932584269664,
+      "grad_norm": 0.8742467164993286,
+      "learning_rate": 3.984532750262703e-06,
+      "loss": 0.6267,
+      "step": 5156
+    },
+    {
+      "epoch": 1.810744382022472,
+      "grad_norm": 0.8632952570915222,
+      "learning_rate": 3.984160736579054e-06,
+      "loss": 0.6136,
+      "step": 5157
+    },
+    {
+      "epoch": 1.8110955056179776,
+      "grad_norm": 0.8474553823471069,
+      "learning_rate": 3.983788672138379e-06,
+      "loss": 0.6133,
+      "step": 5158
+    },
+    {
+      "epoch": 1.8114466292134832,
+      "grad_norm": 0.872573733329773,
+      "learning_rate": 3.983416556953401e-06,
+      "loss": 0.5971,
+      "step": 5159
+    },
+    {
+      "epoch": 1.8117977528089888,
+      "grad_norm": 0.877545952796936,
+      "learning_rate": 3.983044391036846e-06,
+      "loss": 0.6148,
+      "step": 5160
+    },
+    {
+      "epoch": 1.8121488764044944,
+      "grad_norm": 0.908349871635437,
+      "learning_rate": 3.982672174401442e-06,
+      "loss": 0.657,
+      "step": 5161
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 0.8775234818458557,
+      "learning_rate": 3.982299907059918e-06,
+      "loss": 0.6167,
+      "step": 5162
+    },
+    {
+      "epoch": 1.8128511235955056,
+      "grad_norm": 0.9253164529800415,
+      "learning_rate": 3.981927589025006e-06,
+      "loss": 0.6277,
+      "step": 5163
+    },
+    {
+      "epoch": 1.8132022471910112,
+      "grad_norm": 0.8476220965385437,
+      "learning_rate": 3.981555220309439e-06,
+      "loss": 0.6029,
+      "step": 5164
+    },
+    {
+      "epoch": 1.8135533707865168,
+      "grad_norm": 0.8830453157424927,
+      "learning_rate": 3.98118280092595e-06,
+      "loss": 0.6432,
+      "step": 5165
+    },
+    {
+      "epoch": 1.8139044943820224,
+      "grad_norm": 0.8966081738471985,
+      "learning_rate": 3.980810330887277e-06,
+      "loss": 0.6397,
+      "step": 5166
+    },
+    {
+      "epoch": 1.814255617977528,
+      "grad_norm": 0.8749344944953918,
+      "learning_rate": 3.980437810206158e-06,
+      "loss": 0.6352,
+      "step": 5167
+    },
+    {
+      "epoch": 1.8146067415730336,
+      "grad_norm": 0.8861076831817627,
+      "learning_rate": 3.980065238895333e-06,
+      "loss": 0.6306,
+      "step": 5168
+    },
+    {
+      "epoch": 1.8149578651685392,
+      "grad_norm": 0.9134885668754578,
+      "learning_rate": 3.979692616967543e-06,
+      "loss": 0.6036,
+      "step": 5169
+    },
+    {
+      "epoch": 1.8153089887640448,
+      "grad_norm": 0.8736245632171631,
+      "learning_rate": 3.979319944435532e-06,
+      "loss": 0.6032,
+      "step": 5170
+    },
+    {
+      "epoch": 1.8156601123595506,
+      "grad_norm": 0.881210207939148,
+      "learning_rate": 3.978947221312045e-06,
+      "loss": 0.6334,
+      "step": 5171
+    },
+    {
+      "epoch": 1.8160112359550562,
+      "grad_norm": 0.8849275708198547,
+      "learning_rate": 3.978574447609828e-06,
+      "loss": 0.5735,
+      "step": 5172
+    },
+    {
+      "epoch": 1.8163623595505618,
+      "grad_norm": 0.8601321578025818,
+      "learning_rate": 3.978201623341631e-06,
+      "loss": 0.5696,
+      "step": 5173
+    },
+    {
+      "epoch": 1.8167134831460674,
+      "grad_norm": 0.875879168510437,
+      "learning_rate": 3.977828748520203e-06,
+      "loss": 0.6013,
+      "step": 5174
+    },
+    {
+      "epoch": 1.817064606741573,
+      "grad_norm": 0.9210758209228516,
+      "learning_rate": 3.977455823158296e-06,
+      "loss": 0.6343,
+      "step": 5175
+    },
+    {
+      "epoch": 1.8174157303370788,
+      "grad_norm": 0.8983147144317627,
+      "learning_rate": 3.977082847268664e-06,
+      "loss": 0.598,
+      "step": 5176
+    },
+    {
+      "epoch": 1.8177668539325844,
+      "grad_norm": 0.8796141147613525,
+      "learning_rate": 3.976709820864063e-06,
+      "loss": 0.5995,
+      "step": 5177
+    },
+    {
+      "epoch": 1.81811797752809,
+      "grad_norm": 0.9013523459434509,
+      "learning_rate": 3.97633674395725e-06,
+      "loss": 0.6097,
+      "step": 5178
+    },
+    {
+      "epoch": 1.8184691011235956,
+      "grad_norm": 0.8761310577392578,
+      "learning_rate": 3.975963616560983e-06,
+      "loss": 0.6272,
+      "step": 5179
+    },
+    {
+      "epoch": 1.8188202247191012,
+      "grad_norm": 0.8740566968917847,
+      "learning_rate": 3.975590438688024e-06,
+      "loss": 0.6276,
+      "step": 5180
+    },
+    {
+      "epoch": 1.8191713483146068,
+      "grad_norm": 0.8199687600135803,
+      "learning_rate": 3.975217210351133e-06,
+      "loss": 0.5566,
+      "step": 5181
+    },
+    {
+      "epoch": 1.8195224719101124,
+      "grad_norm": 0.8775818347930908,
+      "learning_rate": 3.974843931563078e-06,
+      "loss": 0.6017,
+      "step": 5182
+    },
+    {
+      "epoch": 1.819873595505618,
+      "grad_norm": 0.8951563239097595,
+      "learning_rate": 3.974470602336622e-06,
+      "loss": 0.6004,
+      "step": 5183
+    },
+    {
+      "epoch": 1.8202247191011236,
+      "grad_norm": 0.8890185952186584,
+      "learning_rate": 3.974097222684532e-06,
+      "loss": 0.6362,
+      "step": 5184
+    },
+    {
+      "epoch": 1.8205758426966292,
+      "grad_norm": 0.8939943909645081,
+      "learning_rate": 3.97372379261958e-06,
+      "loss": 0.6268,
+      "step": 5185
+    },
+    {
+      "epoch": 1.8209269662921348,
+      "grad_norm": 0.8659777641296387,
+      "learning_rate": 3.9733503121545345e-06,
+      "loss": 0.624,
+      "step": 5186
+    },
+    {
+      "epoch": 1.8212780898876404,
+      "grad_norm": 0.8676909804344177,
+      "learning_rate": 3.972976781302169e-06,
+      "loss": 0.6138,
+      "step": 5187
+    },
+    {
+      "epoch": 1.821629213483146,
+      "grad_norm": 0.8564205169677734,
+      "learning_rate": 3.972603200075258e-06,
+      "loss": 0.5945,
+      "step": 5188
+    },
+    {
+      "epoch": 1.8219803370786516,
+      "grad_norm": 0.8555643558502197,
+      "learning_rate": 3.972229568486578e-06,
+      "loss": 0.5785,
+      "step": 5189
+    },
+    {
+      "epoch": 1.8223314606741572,
+      "grad_norm": 0.9054913520812988,
+      "learning_rate": 3.971855886548906e-06,
+      "loss": 0.6445,
+      "step": 5190
+    },
+    {
+      "epoch": 1.8226825842696628,
+      "grad_norm": 0.9118171334266663,
+      "learning_rate": 3.9714821542750226e-06,
+      "loss": 0.6134,
+      "step": 5191
+    },
+    {
+      "epoch": 1.8230337078651684,
+      "grad_norm": 0.8640024065971375,
+      "learning_rate": 3.9711083716777085e-06,
+      "loss": 0.6404,
+      "step": 5192
+    },
+    {
+      "epoch": 1.8233848314606742,
+      "grad_norm": 0.8471640348434448,
+      "learning_rate": 3.970734538769747e-06,
+      "loss": 0.5966,
+      "step": 5193
+    },
+    {
+      "epoch": 1.8237359550561798,
+      "grad_norm": 0.8837315440177917,
+      "learning_rate": 3.970360655563923e-06,
+      "loss": 0.6001,
+      "step": 5194
+    },
+    {
+      "epoch": 1.8240870786516854,
+      "grad_norm": 0.8849839568138123,
+      "learning_rate": 3.969986722073023e-06,
+      "loss": 0.6301,
+      "step": 5195
+    },
+    {
+      "epoch": 1.824438202247191,
+      "grad_norm": 0.8799655437469482,
+      "learning_rate": 3.969612738309834e-06,
+      "loss": 0.5701,
+      "step": 5196
+    },
+    {
+      "epoch": 1.8247893258426966,
+      "grad_norm": 0.8617206811904907,
+      "learning_rate": 3.969238704287147e-06,
+      "loss": 0.6126,
+      "step": 5197
+    },
+    {
+      "epoch": 1.8251404494382022,
+      "grad_norm": 0.9338657259941101,
+      "learning_rate": 3.968864620017755e-06,
+      "loss": 0.6022,
+      "step": 5198
+    },
+    {
+      "epoch": 1.825491573033708,
+      "grad_norm": 0.8785296678543091,
+      "learning_rate": 3.968490485514449e-06,
+      "loss": 0.6073,
+      "step": 5199
+    },
+    {
+      "epoch": 1.8258426966292136,
+      "grad_norm": 0.8909986019134521,
+      "learning_rate": 3.9681163007900255e-06,
+      "loss": 0.6152,
+      "step": 5200
+    },
+    {
+      "epoch": 1.8261938202247192,
+      "grad_norm": 0.9001063704490662,
+      "learning_rate": 3.96774206585728e-06,
+      "loss": 0.6351,
+      "step": 5201
+    },
+    {
+      "epoch": 1.8265449438202248,
+      "grad_norm": 0.8840953707695007,
+      "learning_rate": 3.967367780729012e-06,
+      "loss": 0.6452,
+      "step": 5202
+    },
+    {
+      "epoch": 1.8268960674157304,
+      "grad_norm": 0.8713771104812622,
+      "learning_rate": 3.966993445418022e-06,
+      "loss": 0.6099,
+      "step": 5203
+    },
+    {
+      "epoch": 1.827247191011236,
+      "grad_norm": 0.8601224422454834,
+      "learning_rate": 3.9666190599371116e-06,
+      "loss": 0.5861,
+      "step": 5204
+    },
+    {
+      "epoch": 1.8275983146067416,
+      "grad_norm": 0.8551958799362183,
+      "learning_rate": 3.9662446242990834e-06,
+      "loss": 0.5977,
+      "step": 5205
+    },
+    {
+      "epoch": 1.8279494382022472,
+      "grad_norm": 0.8728621602058411,
+      "learning_rate": 3.965870138516744e-06,
+      "loss": 0.6523,
+      "step": 5206
+    },
+    {
+      "epoch": 1.8283005617977528,
+      "grad_norm": 0.8756167888641357,
+      "learning_rate": 3.965495602602901e-06,
+      "loss": 0.6479,
+      "step": 5207
+    },
+    {
+      "epoch": 1.8286516853932584,
+      "grad_norm": 0.8521246910095215,
+      "learning_rate": 3.965121016570361e-06,
+      "loss": 0.6009,
+      "step": 5208
+    },
+    {
+      "epoch": 1.829002808988764,
+      "grad_norm": 0.8759570121765137,
+      "learning_rate": 3.964746380431937e-06,
+      "loss": 0.6221,
+      "step": 5209
+    },
+    {
+      "epoch": 1.8293539325842696,
+      "grad_norm": 0.8832354545593262,
+      "learning_rate": 3.964371694200441e-06,
+      "loss": 0.6124,
+      "step": 5210
+    },
+    {
+      "epoch": 1.8297050561797752,
+      "grad_norm": 0.8482329845428467,
+      "learning_rate": 3.963996957888685e-06,
+      "loss": 0.5887,
+      "step": 5211
+    },
+    {
+      "epoch": 1.8300561797752808,
+      "grad_norm": 0.8687066435813904,
+      "learning_rate": 3.963622171509486e-06,
+      "loss": 0.6097,
+      "step": 5212
+    },
+    {
+      "epoch": 1.8304073033707864,
+      "grad_norm": 0.8731704950332642,
+      "learning_rate": 3.963247335075661e-06,
+      "loss": 0.5865,
+      "step": 5213
+    },
+    {
+      "epoch": 1.830758426966292,
+      "grad_norm": 0.8566392660140991,
+      "learning_rate": 3.9628724486000295e-06,
+      "loss": 0.6283,
+      "step": 5214
+    },
+    {
+      "epoch": 1.8311095505617978,
+      "grad_norm": 0.867788553237915,
+      "learning_rate": 3.962497512095412e-06,
+      "loss": 0.6278,
+      "step": 5215
+    },
+    {
+      "epoch": 1.8314606741573034,
+      "grad_norm": 0.913202166557312,
+      "learning_rate": 3.9621225255746306e-06,
+      "loss": 0.6382,
+      "step": 5216
+    },
+    {
+      "epoch": 1.831811797752809,
+      "grad_norm": 0.8661263585090637,
+      "learning_rate": 3.9617474890505114e-06,
+      "loss": 0.619,
+      "step": 5217
+    },
+    {
+      "epoch": 1.8321629213483146,
+      "grad_norm": 0.867225170135498,
+      "learning_rate": 3.961372402535878e-06,
+      "loss": 0.6006,
+      "step": 5218
+    },
+    {
+      "epoch": 1.8325140449438202,
+      "grad_norm": 0.9224227070808411,
+      "learning_rate": 3.9609972660435586e-06,
+      "loss": 0.6413,
+      "step": 5219
+    },
+    {
+      "epoch": 1.8328651685393258,
+      "grad_norm": 0.9165964722633362,
+      "learning_rate": 3.960622079586384e-06,
+      "loss": 0.5787,
+      "step": 5220
+    },
+    {
+      "epoch": 1.8332162921348316,
+      "grad_norm": 0.9111368060112,
+      "learning_rate": 3.960246843177182e-06,
+      "loss": 0.637,
+      "step": 5221
+    },
+    {
+      "epoch": 1.8335674157303372,
+      "grad_norm": 0.892612874507904,
+      "learning_rate": 3.959871556828789e-06,
+      "loss": 0.6237,
+      "step": 5222
+    },
+    {
+      "epoch": 1.8339185393258428,
+      "grad_norm": 0.8713697791099548,
+      "learning_rate": 3.9594962205540385e-06,
+      "loss": 0.6418,
+      "step": 5223
+    },
+    {
+      "epoch": 1.8342696629213484,
+      "grad_norm": 0.8447199463844299,
+      "learning_rate": 3.959120834365766e-06,
+      "loss": 0.606,
+      "step": 5224
+    },
+    {
+      "epoch": 1.834620786516854,
+      "grad_norm": 0.8594667911529541,
+      "learning_rate": 3.958745398276808e-06,
+      "loss": 0.6233,
+      "step": 5225
+    },
+    {
+      "epoch": 1.8349719101123596,
+      "grad_norm": 0.8984555602073669,
+      "learning_rate": 3.958369912300007e-06,
+      "loss": 0.6336,
+      "step": 5226
+    },
+    {
+      "epoch": 1.8353230337078652,
+      "grad_norm": 0.8514502048492432,
+      "learning_rate": 3.957994376448203e-06,
+      "loss": 0.5858,
+      "step": 5227
+    },
+    {
+      "epoch": 1.8356741573033708,
+      "grad_norm": 0.8587605357170105,
+      "learning_rate": 3.957618790734239e-06,
+      "loss": 0.6066,
+      "step": 5228
+    },
+    {
+      "epoch": 1.8360252808988764,
+      "grad_norm": 0.8868040442466736,
+      "learning_rate": 3.957243155170959e-06,
+      "loss": 0.6111,
+      "step": 5229
+    },
+    {
+      "epoch": 1.836376404494382,
+      "grad_norm": 0.8550049662590027,
+      "learning_rate": 3.956867469771211e-06,
+      "loss": 0.5917,
+      "step": 5230
+    },
+    {
+      "epoch": 1.8367275280898876,
+      "grad_norm": 0.885949969291687,
+      "learning_rate": 3.956491734547841e-06,
+      "loss": 0.6348,
+      "step": 5231
+    },
+    {
+      "epoch": 1.8370786516853932,
+      "grad_norm": 0.8566646575927734,
+      "learning_rate": 3.9561159495137e-06,
+      "loss": 0.597,
+      "step": 5232
+    },
+    {
+      "epoch": 1.8374297752808988,
+      "grad_norm": 0.9448535442352295,
+      "learning_rate": 3.95574011468164e-06,
+      "loss": 0.6376,
+      "step": 5233
+    },
+    {
+      "epoch": 1.8377808988764044,
+      "grad_norm": 0.8502497673034668,
+      "learning_rate": 3.955364230064515e-06,
+      "loss": 0.6009,
+      "step": 5234
+    },
+    {
+      "epoch": 1.83813202247191,
+      "grad_norm": 0.8837283849716187,
+      "learning_rate": 3.9549882956751766e-06,
+      "loss": 0.6434,
+      "step": 5235
+    },
+    {
+      "epoch": 1.8384831460674156,
+      "grad_norm": 0.9013311862945557,
+      "learning_rate": 3.954612311526485e-06,
+      "loss": 0.5851,
+      "step": 5236
+    },
+    {
+      "epoch": 1.8388342696629212,
+      "grad_norm": 0.9204567074775696,
+      "learning_rate": 3.9542362776312974e-06,
+      "loss": 0.655,
+      "step": 5237
+    },
+    {
+      "epoch": 1.839185393258427,
+      "grad_norm": 0.8798168301582336,
+      "learning_rate": 3.953860194002472e-06,
+      "loss": 0.6303,
+      "step": 5238
+    },
+    {
+      "epoch": 1.8395365168539326,
+      "grad_norm": 0.9072458744049072,
+      "learning_rate": 3.953484060652872e-06,
+      "loss": 0.6276,
+      "step": 5239
+    },
+    {
+      "epoch": 1.8398876404494382,
+      "grad_norm": 0.8829042911529541,
+      "learning_rate": 3.9531078775953625e-06,
+      "loss": 0.631,
+      "step": 5240
+    },
+    {
+      "epoch": 1.8402387640449438,
+      "grad_norm": 0.8745750784873962,
+      "learning_rate": 3.952731644842806e-06,
+      "loss": 0.6143,
+      "step": 5241
+    },
+    {
+      "epoch": 1.8405898876404494,
+      "grad_norm": 0.9073604345321655,
+      "learning_rate": 3.95235536240807e-06,
+      "loss": 0.627,
+      "step": 5242
+    },
+    {
+      "epoch": 1.8409410112359552,
+      "grad_norm": 0.8719262480735779,
+      "learning_rate": 3.951979030304024e-06,
+      "loss": 0.6162,
+      "step": 5243
+    },
+    {
+      "epoch": 1.8412921348314608,
+      "grad_norm": 0.8346160650253296,
+      "learning_rate": 3.9516026485435375e-06,
+      "loss": 0.6049,
+      "step": 5244
+    },
+    {
+      "epoch": 1.8416432584269664,
+      "grad_norm": 0.8686418533325195,
+      "learning_rate": 3.9512262171394835e-06,
+      "loss": 0.6157,
+      "step": 5245
+    },
+    {
+      "epoch": 1.841994382022472,
+      "grad_norm": 0.8777024745941162,
+      "learning_rate": 3.9508497361047334e-06,
+      "loss": 0.6434,
+      "step": 5246
+    },
+    {
+      "epoch": 1.8423455056179776,
+      "grad_norm": 0.8348974585533142,
+      "learning_rate": 3.950473205452165e-06,
+      "loss": 0.5557,
+      "step": 5247
+    },
+    {
+      "epoch": 1.8426966292134832,
+      "grad_norm": 0.856793224811554,
+      "learning_rate": 3.950096625194653e-06,
+      "loss": 0.5629,
+      "step": 5248
+    },
+    {
+      "epoch": 1.8430477528089888,
+      "grad_norm": 0.8882337212562561,
+      "learning_rate": 3.949719995345078e-06,
+      "loss": 0.6041,
+      "step": 5249
+    },
+    {
+      "epoch": 1.8433988764044944,
+      "grad_norm": 0.9027885794639587,
+      "learning_rate": 3.9493433159163195e-06,
+      "loss": 0.6082,
+      "step": 5250
+    },
+    {
+      "epoch": 1.84375,
+      "grad_norm": 0.8926260471343994,
+      "learning_rate": 3.94896658692126e-06,
+      "loss": 0.5864,
+      "step": 5251
+    },
+    {
+      "epoch": 1.8441011235955056,
+      "grad_norm": 0.8764528632164001,
+      "learning_rate": 3.948589808372783e-06,
+      "loss": 0.6275,
+      "step": 5252
+    },
+    {
+      "epoch": 1.8444522471910112,
+      "grad_norm": 0.8873112797737122,
+      "learning_rate": 3.948212980283773e-06,
+      "loss": 0.6121,
+      "step": 5253
+    },
+    {
+      "epoch": 1.8448033707865168,
+      "grad_norm": 0.8742668628692627,
+      "learning_rate": 3.947836102667119e-06,
+      "loss": 0.6105,
+      "step": 5254
+    },
+    {
+      "epoch": 1.8451544943820224,
+      "grad_norm": 0.8881484270095825,
+      "learning_rate": 3.947459175535711e-06,
+      "loss": 0.6315,
+      "step": 5255
+    },
+    {
+      "epoch": 1.845505617977528,
+      "grad_norm": 0.8762322068214417,
+      "learning_rate": 3.9470821989024356e-06,
+      "loss": 0.61,
+      "step": 5256
+    },
+    {
+      "epoch": 1.8458567415730336,
+      "grad_norm": 0.8735952973365784,
+      "learning_rate": 3.946705172780189e-06,
+      "loss": 0.6066,
+      "step": 5257
+    },
+    {
+      "epoch": 1.8462078651685392,
+      "grad_norm": 0.8771315813064575,
+      "learning_rate": 3.946328097181862e-06,
+      "loss": 0.6196,
+      "step": 5258
+    },
+    {
+      "epoch": 1.8465589887640448,
+      "grad_norm": 0.8824196457862854,
+      "learning_rate": 3.945950972120351e-06,
+      "loss": 0.6753,
+      "step": 5259
+    },
+    {
+      "epoch": 1.8469101123595506,
+      "grad_norm": 0.8754235506057739,
+      "learning_rate": 3.9455737976085564e-06,
+      "loss": 0.6097,
+      "step": 5260
+    },
+    {
+      "epoch": 1.8472612359550562,
+      "grad_norm": 0.8685277700424194,
+      "learning_rate": 3.945196573659374e-06,
+      "loss": 0.6107,
+      "step": 5261
+    },
+    {
+      "epoch": 1.8476123595505618,
+      "grad_norm": 0.8714946508407593,
+      "learning_rate": 3.944819300285705e-06,
+      "loss": 0.6071,
+      "step": 5262
+    },
+    {
+      "epoch": 1.8479634831460674,
+      "grad_norm": 0.8681133985519409,
+      "learning_rate": 3.9444419775004525e-06,
+      "loss": 0.6028,
+      "step": 5263
+    },
+    {
+      "epoch": 1.848314606741573,
+      "grad_norm": 0.8595680594444275,
+      "learning_rate": 3.9440646053165204e-06,
+      "loss": 0.5931,
+      "step": 5264
+    },
+    {
+      "epoch": 1.8486657303370788,
+      "grad_norm": 0.9302463531494141,
+      "learning_rate": 3.943687183746815e-06,
+      "loss": 0.6239,
+      "step": 5265
+    },
+    {
+      "epoch": 1.8490168539325844,
+      "grad_norm": 0.8686720132827759,
+      "learning_rate": 3.943309712804243e-06,
+      "loss": 0.6177,
+      "step": 5266
+    },
+    {
+      "epoch": 1.84936797752809,
+      "grad_norm": 0.8854331970214844,
+      "learning_rate": 3.9429321925017136e-06,
+      "loss": 0.6291,
+      "step": 5267
+    },
+    {
+      "epoch": 1.8497191011235956,
+      "grad_norm": 0.8775175213813782,
+      "learning_rate": 3.942554622852138e-06,
+      "loss": 0.6006,
+      "step": 5268
+    },
+    {
+      "epoch": 1.8500702247191012,
+      "grad_norm": 0.8732326030731201,
+      "learning_rate": 3.942177003868428e-06,
+      "loss": 0.6282,
+      "step": 5269
+    },
+    {
+      "epoch": 1.8504213483146068,
+      "grad_norm": 0.8572641015052795,
+      "learning_rate": 3.9417993355635e-06,
+      "loss": 0.613,
+      "step": 5270
+    },
+    {
+      "epoch": 1.8507724719101124,
+      "grad_norm": 0.8818247318267822,
+      "learning_rate": 3.941421617950268e-06,
+      "loss": 0.5826,
+      "step": 5271
+    },
+    {
+      "epoch": 1.851123595505618,
+      "grad_norm": 0.8704795837402344,
+      "learning_rate": 3.9410438510416494e-06,
+      "loss": 0.6254,
+      "step": 5272
+    },
+    {
+      "epoch": 1.8514747191011236,
+      "grad_norm": 0.8848576545715332,
+      "learning_rate": 3.940666034850565e-06,
+      "loss": 0.6199,
+      "step": 5273
+    },
+    {
+      "epoch": 1.8518258426966292,
+      "grad_norm": 0.8868470191955566,
+      "learning_rate": 3.940288169389935e-06,
+      "loss": 0.617,
+      "step": 5274
+    },
+    {
+      "epoch": 1.8521769662921348,
+      "grad_norm": 0.875606894493103,
+      "learning_rate": 3.939910254672682e-06,
+      "loss": 0.5775,
+      "step": 5275
+    },
+    {
+      "epoch": 1.8525280898876404,
+      "grad_norm": 0.8833727240562439,
+      "learning_rate": 3.93953229071173e-06,
+      "loss": 0.607,
+      "step": 5276
+    },
+    {
+      "epoch": 1.852879213483146,
+      "grad_norm": 0.9035220742225647,
+      "learning_rate": 3.939154277520006e-06,
+      "loss": 0.6213,
+      "step": 5277
+    },
+    {
+      "epoch": 1.8532303370786516,
+      "grad_norm": 0.8979851007461548,
+      "learning_rate": 3.938776215110437e-06,
+      "loss": 0.6331,
+      "step": 5278
+    },
+    {
+      "epoch": 1.8535814606741572,
+      "grad_norm": 0.8894830346107483,
+      "learning_rate": 3.938398103495952e-06,
+      "loss": 0.585,
+      "step": 5279
+    },
+    {
+      "epoch": 1.8539325842696628,
+      "grad_norm": 0.8893154859542847,
+      "learning_rate": 3.938019942689484e-06,
+      "loss": 0.5978,
+      "step": 5280
+    },
+    {
+      "epoch": 1.8542837078651684,
+      "grad_norm": 0.8567217588424683,
+      "learning_rate": 3.9376417327039636e-06,
+      "loss": 0.6029,
+      "step": 5281
+    },
+    {
+      "epoch": 1.8546348314606742,
+      "grad_norm": 0.8965117335319519,
+      "learning_rate": 3.937263473552326e-06,
+      "loss": 0.6029,
+      "step": 5282
+    },
+    {
+      "epoch": 1.8549859550561798,
+      "grad_norm": 0.8815075755119324,
+      "learning_rate": 3.936885165247508e-06,
+      "loss": 0.6741,
+      "step": 5283
+    },
+    {
+      "epoch": 1.8553370786516854,
+      "grad_norm": 0.8753764629364014,
+      "learning_rate": 3.936506807802446e-06,
+      "loss": 0.6105,
+      "step": 5284
+    },
+    {
+      "epoch": 1.855688202247191,
+      "grad_norm": 0.884854257106781,
+      "learning_rate": 3.9361284012300816e-06,
+      "loss": 0.6191,
+      "step": 5285
+    },
+    {
+      "epoch": 1.8560393258426966,
+      "grad_norm": 0.8499030470848083,
+      "learning_rate": 3.935749945543354e-06,
+      "loss": 0.6179,
+      "step": 5286
+    },
+    {
+      "epoch": 1.8563904494382022,
+      "grad_norm": 0.8841039538383484,
+      "learning_rate": 3.935371440755208e-06,
+      "loss": 0.6265,
+      "step": 5287
+    },
+    {
+      "epoch": 1.856741573033708,
+      "grad_norm": 0.855475902557373,
+      "learning_rate": 3.934992886878586e-06,
+      "loss": 0.5774,
+      "step": 5288
+    },
+    {
+      "epoch": 1.8570926966292136,
+      "grad_norm": 0.8800686597824097,
+      "learning_rate": 3.934614283926435e-06,
+      "loss": 0.6131,
+      "step": 5289
+    },
+    {
+      "epoch": 1.8574438202247192,
+      "grad_norm": 0.8433512449264526,
+      "learning_rate": 3.934235631911704e-06,
+      "loss": 0.5718,
+      "step": 5290
+    },
+    {
+      "epoch": 1.8577949438202248,
+      "grad_norm": 0.9216337203979492,
+      "learning_rate": 3.933856930847341e-06,
+      "loss": 0.6243,
+      "step": 5291
+    },
+    {
+      "epoch": 1.8581460674157304,
+      "grad_norm": 0.8634188175201416,
+      "learning_rate": 3.9334781807462984e-06,
+      "loss": 0.6016,
+      "step": 5292
+    },
+    {
+      "epoch": 1.858497191011236,
+      "grad_norm": 0.8891706466674805,
+      "learning_rate": 3.933099381621528e-06,
+      "loss": 0.6323,
+      "step": 5293
+    },
+    {
+      "epoch": 1.8588483146067416,
+      "grad_norm": 0.8881804943084717,
+      "learning_rate": 3.9327205334859855e-06,
+      "loss": 0.6182,
+      "step": 5294
+    },
+    {
+      "epoch": 1.8591994382022472,
+      "grad_norm": 0.8558664917945862,
+      "learning_rate": 3.932341636352627e-06,
+      "loss": 0.6333,
+      "step": 5295
+    },
+    {
+      "epoch": 1.8595505617977528,
+      "grad_norm": 0.8729931116104126,
+      "learning_rate": 3.93196269023441e-06,
+      "loss": 0.6282,
+      "step": 5296
+    },
+    {
+      "epoch": 1.8599016853932584,
+      "grad_norm": 0.9086493849754333,
+      "learning_rate": 3.931583695144294e-06,
+      "loss": 0.6378,
+      "step": 5297
+    },
+    {
+      "epoch": 1.860252808988764,
+      "grad_norm": 0.8665076494216919,
+      "learning_rate": 3.931204651095241e-06,
+      "loss": 0.6327,
+      "step": 5298
+    },
+    {
+      "epoch": 1.8606039325842696,
+      "grad_norm": 0.9046199917793274,
+      "learning_rate": 3.9308255581002145e-06,
+      "loss": 0.6651,
+      "step": 5299
+    },
+    {
+      "epoch": 1.8609550561797752,
+      "grad_norm": 0.89439857006073,
+      "learning_rate": 3.930446416172177e-06,
+      "loss": 0.6414,
+      "step": 5300
+    },
+    {
+      "epoch": 1.8613061797752808,
+      "grad_norm": 0.8834724426269531,
+      "learning_rate": 3.930067225324096e-06,
+      "loss": 0.5866,
+      "step": 5301
+    },
+    {
+      "epoch": 1.8616573033707864,
+      "grad_norm": 0.882892906665802,
+      "learning_rate": 3.92968798556894e-06,
+      "loss": 0.6551,
+      "step": 5302
+    },
+    {
+      "epoch": 1.862008426966292,
+      "grad_norm": 0.8716971278190613,
+      "learning_rate": 3.929308696919678e-06,
+      "loss": 0.6237,
+      "step": 5303
+    },
+    {
+      "epoch": 1.8623595505617978,
+      "grad_norm": 0.8676100373268127,
+      "learning_rate": 3.928929359389282e-06,
+      "loss": 0.5929,
+      "step": 5304
+    },
+    {
+      "epoch": 1.8627106741573034,
+      "grad_norm": 0.922857403755188,
+      "learning_rate": 3.928549972990725e-06,
+      "loss": 0.679,
+      "step": 5305
+    },
+    {
+      "epoch": 1.863061797752809,
+      "grad_norm": 0.8612826466560364,
+      "learning_rate": 3.9281705377369814e-06,
+      "loss": 0.6116,
+      "step": 5306
+    },
+    {
+      "epoch": 1.8634129213483146,
+      "grad_norm": 0.8769382238388062,
+      "learning_rate": 3.927791053641027e-06,
+      "loss": 0.6057,
+      "step": 5307
+    },
+    {
+      "epoch": 1.8637640449438202,
+      "grad_norm": 0.8767523169517517,
+      "learning_rate": 3.92741152071584e-06,
+      "loss": 0.6181,
+      "step": 5308
+    },
+    {
+      "epoch": 1.8641151685393258,
+      "grad_norm": 0.897783637046814,
+      "learning_rate": 3.927031938974401e-06,
+      "loss": 0.5993,
+      "step": 5309
+    },
+    {
+      "epoch": 1.8644662921348316,
+      "grad_norm": 0.8882983922958374,
+      "learning_rate": 3.926652308429691e-06,
+      "loss": 0.6075,
+      "step": 5310
+    },
+    {
+      "epoch": 1.8648174157303372,
+      "grad_norm": 0.9197914600372314,
+      "learning_rate": 3.926272629094692e-06,
+      "loss": 0.6058,
+      "step": 5311
+    },
+    {
+      "epoch": 1.8651685393258428,
+      "grad_norm": 0.8851439356803894,
+      "learning_rate": 3.9258929009823906e-06,
+      "loss": 0.6073,
+      "step": 5312
+    },
+    {
+      "epoch": 1.8655196629213484,
+      "grad_norm": 0.8552735447883606,
+      "learning_rate": 3.925513124105772e-06,
+      "loss": 0.6214,
+      "step": 5313
+    },
+    {
+      "epoch": 1.865870786516854,
+      "grad_norm": 0.8668675422668457,
+      "learning_rate": 3.925133298477824e-06,
+      "loss": 0.6129,
+      "step": 5314
+    },
+    {
+      "epoch": 1.8662219101123596,
+      "grad_norm": 0.8721663951873779,
+      "learning_rate": 3.924753424111537e-06,
+      "loss": 0.6205,
+      "step": 5315
+    },
+    {
+      "epoch": 1.8665730337078652,
+      "grad_norm": 0.8686200976371765,
+      "learning_rate": 3.9243735010199014e-06,
+      "loss": 0.6144,
+      "step": 5316
+    },
+    {
+      "epoch": 1.8669241573033708,
+      "grad_norm": 0.8643429279327393,
+      "learning_rate": 3.923993529215912e-06,
+      "loss": 0.618,
+      "step": 5317
+    },
+    {
+      "epoch": 1.8672752808988764,
+      "grad_norm": 0.8749399185180664,
+      "learning_rate": 3.923613508712562e-06,
+      "loss": 0.6436,
+      "step": 5318
+    },
+    {
+      "epoch": 1.867626404494382,
+      "grad_norm": 0.8470568060874939,
+      "learning_rate": 3.923233439522847e-06,
+      "loss": 0.6082,
+      "step": 5319
+    },
+    {
+      "epoch": 1.8679775280898876,
+      "grad_norm": 0.8877463340759277,
+      "learning_rate": 3.922853321659767e-06,
+      "loss": 0.6028,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8683286516853932,
+      "grad_norm": 0.8590446710586548,
+      "learning_rate": 3.922473155136322e-06,
+      "loss": 0.601,
+      "step": 5321
+    },
+    {
+      "epoch": 1.8686797752808988,
+      "grad_norm": 0.9017147421836853,
+      "learning_rate": 3.922092939965511e-06,
+      "loss": 0.6602,
+      "step": 5322
+    },
+    {
+      "epoch": 1.8690308988764044,
+      "grad_norm": 0.8631618022918701,
+      "learning_rate": 3.921712676160339e-06,
+      "loss": 0.6088,
+      "step": 5323
+    },
+    {
+      "epoch": 1.86938202247191,
+      "grad_norm": 0.8838371634483337,
+      "learning_rate": 3.921332363733811e-06,
+      "loss": 0.6288,
+      "step": 5324
+    },
+    {
+      "epoch": 1.8697331460674156,
+      "grad_norm": 0.8923317193984985,
+      "learning_rate": 3.9209520026989305e-06,
+      "loss": 0.6259,
+      "step": 5325
+    },
+    {
+      "epoch": 1.8700842696629212,
+      "grad_norm": 0.8856344223022461,
+      "learning_rate": 3.920571593068708e-06,
+      "loss": 0.6069,
+      "step": 5326
+    },
+    {
+      "epoch": 1.870435393258427,
+      "grad_norm": 0.8869331479072571,
+      "learning_rate": 3.920191134856153e-06,
+      "loss": 0.5994,
+      "step": 5327
+    },
+    {
+      "epoch": 1.8707865168539326,
+      "grad_norm": 0.8856245875358582,
+      "learning_rate": 3.9198106280742765e-06,
+      "loss": 0.6061,
+      "step": 5328
+    },
+    {
+      "epoch": 1.8711376404494382,
+      "grad_norm": 0.8708471059799194,
+      "learning_rate": 3.919430072736091e-06,
+      "loss": 0.597,
+      "step": 5329
+    },
+    {
+      "epoch": 1.8714887640449438,
+      "grad_norm": 0.9098658561706543,
+      "learning_rate": 3.919049468854612e-06,
+      "loss": 0.6303,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8718398876404494,
+      "grad_norm": 0.8640565276145935,
+      "learning_rate": 3.918668816442856e-06,
+      "loss": 0.6217,
+      "step": 5331
+    },
+    {
+      "epoch": 1.8721910112359552,
+      "grad_norm": 0.8827778697013855,
+      "learning_rate": 3.918288115513839e-06,
+      "loss": 0.6286,
+      "step": 5332
+    },
+    {
+      "epoch": 1.8725421348314608,
+      "grad_norm": 0.867546021938324,
+      "learning_rate": 3.917907366080583e-06,
+      "loss": 0.6291,
+      "step": 5333
+    },
+    {
+      "epoch": 1.8728932584269664,
+      "grad_norm": 0.8924286961555481,
+      "learning_rate": 3.917526568156109e-06,
+      "loss": 0.6058,
+      "step": 5334
+    },
+    {
+      "epoch": 1.873244382022472,
+      "grad_norm": 0.8416984677314758,
+      "learning_rate": 3.917145721753438e-06,
+      "loss": 0.5685,
+      "step": 5335
+    },
+    {
+      "epoch": 1.8735955056179776,
+      "grad_norm": 0.836463212966919,
+      "learning_rate": 3.916764826885597e-06,
+      "loss": 0.5749,
+      "step": 5336
+    },
+    {
+      "epoch": 1.8739466292134832,
+      "grad_norm": 0.888582706451416,
+      "learning_rate": 3.916383883565612e-06,
+      "loss": 0.5987,
+      "step": 5337
+    },
+    {
+      "epoch": 1.8742977528089888,
+      "grad_norm": 0.8800485134124756,
+      "learning_rate": 3.916002891806508e-06,
+      "loss": 0.5966,
+      "step": 5338
+    },
+    {
+      "epoch": 1.8746488764044944,
+      "grad_norm": 0.8612186908721924,
+      "learning_rate": 3.915621851621318e-06,
+      "loss": 0.5848,
+      "step": 5339
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.8935150504112244,
+      "learning_rate": 3.915240763023072e-06,
+      "loss": 0.6036,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8753511235955056,
+      "grad_norm": 0.8847930431365967,
+      "learning_rate": 3.914859626024804e-06,
+      "loss": 0.6171,
+      "step": 5341
+    },
+    {
+      "epoch": 1.8757022471910112,
+      "grad_norm": 0.8781599998474121,
+      "learning_rate": 3.914478440639546e-06,
+      "loss": 0.615,
+      "step": 5342
+    },
+    {
+      "epoch": 1.8760533707865168,
+      "grad_norm": 0.902090847492218,
+      "learning_rate": 3.914097206880336e-06,
+      "loss": 0.6234,
+      "step": 5343
+    },
+    {
+      "epoch": 1.8764044943820224,
+      "grad_norm": 0.8890019655227661,
+      "learning_rate": 3.913715924760212e-06,
+      "loss": 0.6005,
+      "step": 5344
+    },
+    {
+      "epoch": 1.876755617977528,
+      "grad_norm": 0.853225827217102,
+      "learning_rate": 3.913334594292213e-06,
+      "loss": 0.6023,
+      "step": 5345
+    },
+    {
+      "epoch": 1.8771067415730336,
+      "grad_norm": 0.8685701489448547,
+      "learning_rate": 3.91295321548938e-06,
+      "loss": 0.6291,
+      "step": 5346
+    },
+    {
+      "epoch": 1.8774578651685392,
+      "grad_norm": 0.8687042593955994,
+      "learning_rate": 3.912571788364756e-06,
+      "loss": 0.5823,
+      "step": 5347
+    },
+    {
+      "epoch": 1.8778089887640448,
+      "grad_norm": 0.8955281972885132,
+      "learning_rate": 3.9121903129313864e-06,
+      "loss": 0.6188,
+      "step": 5348
+    },
+    {
+      "epoch": 1.8781601123595506,
+      "grad_norm": 0.911578893661499,
+      "learning_rate": 3.911808789202316e-06,
+      "loss": 0.5788,
+      "step": 5349
+    },
+    {
+      "epoch": 1.8785112359550562,
+      "grad_norm": 0.8692489862442017,
+      "learning_rate": 3.911427217190593e-06,
+      "loss": 0.6135,
+      "step": 5350
+    },
+    {
+      "epoch": 1.8788623595505618,
+      "grad_norm": 0.8962387442588806,
+      "learning_rate": 3.911045596909267e-06,
+      "loss": 0.6012,
+      "step": 5351
+    },
+    {
+      "epoch": 1.8792134831460674,
+      "grad_norm": 0.8675644993782043,
+      "learning_rate": 3.910663928371389e-06,
+      "loss": 0.6014,
+      "step": 5352
+    },
+    {
+      "epoch": 1.879564606741573,
+      "grad_norm": 0.8898077011108398,
+      "learning_rate": 3.910282211590012e-06,
+      "loss": 0.6502,
+      "step": 5353
+    },
+    {
+      "epoch": 1.8799157303370788,
+      "grad_norm": 0.876844584941864,
+      "learning_rate": 3.9099004465781895e-06,
+      "loss": 0.6123,
+      "step": 5354
+    },
+    {
+      "epoch": 1.8802668539325844,
+      "grad_norm": 0.8768188953399658,
+      "learning_rate": 3.909518633348979e-06,
+      "loss": 0.611,
+      "step": 5355
+    },
+    {
+      "epoch": 1.88061797752809,
+      "grad_norm": 0.9023895859718323,
+      "learning_rate": 3.909136771915437e-06,
+      "loss": 0.6135,
+      "step": 5356
+    },
+    {
+      "epoch": 1.8809691011235956,
+      "grad_norm": 0.9012351632118225,
+      "learning_rate": 3.9087548622906235e-06,
+      "loss": 0.6493,
+      "step": 5357
+    },
+    {
+      "epoch": 1.8813202247191012,
+      "grad_norm": 0.8937700986862183,
+      "learning_rate": 3.908372904487599e-06,
+      "loss": 0.6015,
+      "step": 5358
+    },
+    {
+      "epoch": 1.8816713483146068,
+      "grad_norm": 0.8769406676292419,
+      "learning_rate": 3.9079908985194255e-06,
+      "loss": 0.6112,
+      "step": 5359
+    },
+    {
+      "epoch": 1.8820224719101124,
+      "grad_norm": 0.887407124042511,
+      "learning_rate": 3.90760884439917e-06,
+      "loss": 0.5949,
+      "step": 5360
+    },
+    {
+      "epoch": 1.882373595505618,
+      "grad_norm": 0.8920381665229797,
+      "learning_rate": 3.907226742139895e-06,
+      "loss": 0.6141,
+      "step": 5361
+    },
+    {
+      "epoch": 1.8827247191011236,
+      "grad_norm": 0.9120955467224121,
+      "learning_rate": 3.906844591754671e-06,
+      "loss": 0.6315,
+      "step": 5362
+    },
+    {
+      "epoch": 1.8830758426966292,
+      "grad_norm": 0.8675406575202942,
+      "learning_rate": 3.906462393256565e-06,
+      "loss": 0.6008,
+      "step": 5363
+    },
+    {
+      "epoch": 1.8834269662921348,
+      "grad_norm": 0.8848645091056824,
+      "learning_rate": 3.906080146658648e-06,
+      "loss": 0.6364,
+      "step": 5364
+    },
+    {
+      "epoch": 1.8837780898876404,
+      "grad_norm": 0.873410165309906,
+      "learning_rate": 3.905697851973994e-06,
+      "loss": 0.6179,
+      "step": 5365
+    },
+    {
+      "epoch": 1.884129213483146,
+      "grad_norm": 0.8644540905952454,
+      "learning_rate": 3.905315509215676e-06,
+      "loss": 0.5992,
+      "step": 5366
+    },
+    {
+      "epoch": 1.8844803370786516,
+      "grad_norm": 0.8811582326889038,
+      "learning_rate": 3.9049331183967694e-06,
+      "loss": 0.6385,
+      "step": 5367
+    },
+    {
+      "epoch": 1.8848314606741572,
+      "grad_norm": 0.8806861042976379,
+      "learning_rate": 3.904550679530354e-06,
+      "loss": 0.5863,
+      "step": 5368
+    },
+    {
+      "epoch": 1.8851825842696628,
+      "grad_norm": 0.8641405701637268,
+      "learning_rate": 3.904168192629506e-06,
+      "loss": 0.6297,
+      "step": 5369
+    },
+    {
+      "epoch": 1.8855337078651684,
+      "grad_norm": 0.8805460929870605,
+      "learning_rate": 3.903785657707308e-06,
+      "loss": 0.6499,
+      "step": 5370
+    },
+    {
+      "epoch": 1.8858848314606742,
+      "grad_norm": 0.8589033484458923,
+      "learning_rate": 3.9034030747768416e-06,
+      "loss": 0.5999,
+      "step": 5371
+    },
+    {
+      "epoch": 1.8862359550561798,
+      "grad_norm": 0.8469181060791016,
+      "learning_rate": 3.903020443851191e-06,
+      "loss": 0.6029,
+      "step": 5372
+    },
+    {
+      "epoch": 1.8865870786516854,
+      "grad_norm": 0.8949747681617737,
+      "learning_rate": 3.902637764943441e-06,
+      "loss": 0.633,
+      "step": 5373
+    },
+    {
+      "epoch": 1.886938202247191,
+      "grad_norm": 0.8772246241569519,
+      "learning_rate": 3.902255038066681e-06,
+      "loss": 0.6144,
+      "step": 5374
+    },
+    {
+      "epoch": 1.8872893258426966,
+      "grad_norm": 0.8535354733467102,
+      "learning_rate": 3.901872263233998e-06,
+      "loss": 0.6059,
+      "step": 5375
+    },
+    {
+      "epoch": 1.8876404494382022,
+      "grad_norm": 0.8883733153343201,
+      "learning_rate": 3.901489440458483e-06,
+      "loss": 0.5846,
+      "step": 5376
+    },
+    {
+      "epoch": 1.887991573033708,
+      "grad_norm": 0.8794171810150146,
+      "learning_rate": 3.901106569753229e-06,
+      "loss": 0.5991,
+      "step": 5377
+    },
+    {
+      "epoch": 1.8883426966292136,
+      "grad_norm": 0.8670855164527893,
+      "learning_rate": 3.900723651131328e-06,
+      "loss": 0.6293,
+      "step": 5378
+    },
+    {
+      "epoch": 1.8886938202247192,
+      "grad_norm": 0.8933359384536743,
+      "learning_rate": 3.900340684605878e-06,
+      "loss": 0.6334,
+      "step": 5379
+    },
+    {
+      "epoch": 1.8890449438202248,
+      "grad_norm": 0.8721469640731812,
+      "learning_rate": 3.899957670189974e-06,
+      "loss": 0.6133,
+      "step": 5380
+    },
+    {
+      "epoch": 1.8893960674157304,
+      "grad_norm": 0.8683612942695618,
+      "learning_rate": 3.899574607896716e-06,
+      "loss": 0.6315,
+      "step": 5381
+    },
+    {
+      "epoch": 1.889747191011236,
+      "grad_norm": 0.8970484137535095,
+      "learning_rate": 3.8991914977392055e-06,
+      "loss": 0.6686,
+      "step": 5382
+    },
+    {
+      "epoch": 1.8900983146067416,
+      "grad_norm": 0.8801787495613098,
+      "learning_rate": 3.898808339730541e-06,
+      "loss": 0.6471,
+      "step": 5383
+    },
+    {
+      "epoch": 1.8904494382022472,
+      "grad_norm": 0.907516360282898,
+      "learning_rate": 3.89842513388383e-06,
+      "loss": 0.654,
+      "step": 5384
+    },
+    {
+      "epoch": 1.8908005617977528,
+      "grad_norm": 0.8796942830085754,
+      "learning_rate": 3.898041880212174e-06,
+      "loss": 0.5926,
+      "step": 5385
+    },
+    {
+      "epoch": 1.8911516853932584,
+      "grad_norm": 0.873957633972168,
+      "learning_rate": 3.897658578728683e-06,
+      "loss": 0.6269,
+      "step": 5386
+    },
+    {
+      "epoch": 1.891502808988764,
+      "grad_norm": 0.8868711590766907,
+      "learning_rate": 3.897275229446466e-06,
+      "loss": 0.6081,
+      "step": 5387
+    },
+    {
+      "epoch": 1.8918539325842696,
+      "grad_norm": 0.8819994926452637,
+      "learning_rate": 3.89689183237863e-06,
+      "loss": 0.6204,
+      "step": 5388
+    },
+    {
+      "epoch": 1.8922050561797752,
+      "grad_norm": 0.8778470158576965,
+      "learning_rate": 3.896508387538289e-06,
+      "loss": 0.6256,
+      "step": 5389
+    },
+    {
+      "epoch": 1.8925561797752808,
+      "grad_norm": 0.8505924344062805,
+      "learning_rate": 3.896124894938556e-06,
+      "loss": 0.6361,
+      "step": 5390
+    },
+    {
+      "epoch": 1.8929073033707864,
+      "grad_norm": 0.8763235211372375,
+      "learning_rate": 3.8957413545925475e-06,
+      "loss": 0.6421,
+      "step": 5391
+    },
+    {
+      "epoch": 1.893258426966292,
+      "grad_norm": 0.8803117871284485,
+      "learning_rate": 3.895357766513378e-06,
+      "loss": 0.6148,
+      "step": 5392
+    },
+    {
+      "epoch": 1.8936095505617978,
+      "grad_norm": 0.8600309491157532,
+      "learning_rate": 3.894974130714167e-06,
+      "loss": 0.6275,
+      "step": 5393
+    },
+    {
+      "epoch": 1.8939606741573034,
+      "grad_norm": 0.8821873664855957,
+      "learning_rate": 3.894590447208034e-06,
+      "loss": 0.6356,
+      "step": 5394
+    },
+    {
+      "epoch": 1.894311797752809,
+      "grad_norm": 0.8747897148132324,
+      "learning_rate": 3.894206716008102e-06,
+      "loss": 0.6109,
+      "step": 5395
+    },
+    {
+      "epoch": 1.8946629213483146,
+      "grad_norm": 0.8427042365074158,
+      "learning_rate": 3.893822937127492e-06,
+      "loss": 0.5866,
+      "step": 5396
+    },
+    {
+      "epoch": 1.8950140449438202,
+      "grad_norm": 0.8987433910369873,
+      "learning_rate": 3.893439110579332e-06,
+      "loss": 0.6339,
+      "step": 5397
+    },
+    {
+      "epoch": 1.8953651685393258,
+      "grad_norm": 0.8769526481628418,
+      "learning_rate": 3.893055236376746e-06,
+      "loss": 0.6208,
+      "step": 5398
+    },
+    {
+      "epoch": 1.8957162921348316,
+      "grad_norm": 0.8786547183990479,
+      "learning_rate": 3.892671314532863e-06,
+      "loss": 0.655,
+      "step": 5399
+    },
+    {
+      "epoch": 1.8960674157303372,
+      "grad_norm": 0.8721256256103516,
+      "learning_rate": 3.892287345060814e-06,
+      "loss": 0.6026,
+      "step": 5400
+    },
+    {
+      "epoch": 1.8964185393258428,
+      "grad_norm": 0.9021056294441223,
+      "learning_rate": 3.891903327973727e-06,
+      "loss": 0.6335,
+      "step": 5401
+    },
+    {
+      "epoch": 1.8967696629213484,
+      "grad_norm": 0.8656460642814636,
+      "learning_rate": 3.891519263284738e-06,
+      "loss": 0.5922,
+      "step": 5402
+    },
+    {
+      "epoch": 1.897120786516854,
+      "grad_norm": 0.8590431213378906,
+      "learning_rate": 3.891135151006981e-06,
+      "loss": 0.6004,
+      "step": 5403
+    },
+    {
+      "epoch": 1.8974719101123596,
+      "grad_norm": 0.8588739037513733,
+      "learning_rate": 3.890750991153592e-06,
+      "loss": 0.5901,
+      "step": 5404
+    },
+    {
+      "epoch": 1.8978230337078652,
+      "grad_norm": 0.8661170601844788,
+      "learning_rate": 3.89036678373771e-06,
+      "loss": 0.6204,
+      "step": 5405
+    },
+    {
+      "epoch": 1.8981741573033708,
+      "grad_norm": 0.8917558193206787,
+      "learning_rate": 3.889982528772474e-06,
+      "loss": 0.6184,
+      "step": 5406
+    },
+    {
+      "epoch": 1.8985252808988764,
+      "grad_norm": 0.8691694736480713,
+      "learning_rate": 3.8895982262710236e-06,
+      "loss": 0.6294,
+      "step": 5407
+    },
+    {
+      "epoch": 1.898876404494382,
+      "grad_norm": 0.9101895093917847,
+      "learning_rate": 3.8892138762465034e-06,
+      "loss": 0.6445,
+      "step": 5408
+    },
+    {
+      "epoch": 1.8992275280898876,
+      "grad_norm": 0.8621460795402527,
+      "learning_rate": 3.888829478712058e-06,
+      "loss": 0.5856,
+      "step": 5409
+    },
+    {
+      "epoch": 1.8995786516853932,
+      "grad_norm": 0.8767841458320618,
+      "learning_rate": 3.8884450336808325e-06,
+      "loss": 0.6094,
+      "step": 5410
+    },
+    {
+      "epoch": 1.8999297752808988,
+      "grad_norm": 0.8723804354667664,
+      "learning_rate": 3.888060541165974e-06,
+      "loss": 0.6424,
+      "step": 5411
+    },
+    {
+      "epoch": 1.9002808988764044,
+      "grad_norm": 0.914440393447876,
+      "learning_rate": 3.887676001180635e-06,
+      "loss": 0.6079,
+      "step": 5412
+    },
+    {
+      "epoch": 1.90063202247191,
+      "grad_norm": 0.8787926435470581,
+      "learning_rate": 3.887291413737962e-06,
+      "loss": 0.5855,
+      "step": 5413
+    },
+    {
+      "epoch": 1.9009831460674156,
+      "grad_norm": 0.8624381422996521,
+      "learning_rate": 3.886906778851111e-06,
+      "loss": 0.5986,
+      "step": 5414
+    },
+    {
+      "epoch": 1.9013342696629212,
+      "grad_norm": 0.8594074845314026,
+      "learning_rate": 3.886522096533235e-06,
+      "loss": 0.5894,
+      "step": 5415
+    },
+    {
+      "epoch": 1.901685393258427,
+      "grad_norm": 0.8813564777374268,
+      "learning_rate": 3.8861373667974885e-06,
+      "loss": 0.6339,
+      "step": 5416
+    },
+    {
+      "epoch": 1.9020365168539326,
+      "grad_norm": 0.900949239730835,
+      "learning_rate": 3.885752589657032e-06,
+      "loss": 0.6144,
+      "step": 5417
+    },
+    {
+      "epoch": 1.9023876404494382,
+      "grad_norm": 0.9057722091674805,
+      "learning_rate": 3.885367765125021e-06,
+      "loss": 0.5978,
+      "step": 5418
+    },
+    {
+      "epoch": 1.9027387640449438,
+      "grad_norm": 0.8934159874916077,
+      "learning_rate": 3.884982893214619e-06,
+      "loss": 0.642,
+      "step": 5419
+    },
+    {
+      "epoch": 1.9030898876404494,
+      "grad_norm": 0.888180673122406,
+      "learning_rate": 3.884597973938987e-06,
+      "loss": 0.6387,
+      "step": 5420
+    },
+    {
+      "epoch": 1.9034410112359552,
+      "grad_norm": 0.8636809587478638,
+      "learning_rate": 3.884213007311289e-06,
+      "loss": 0.6016,
+      "step": 5421
+    },
+    {
+      "epoch": 1.9037921348314608,
+      "grad_norm": 0.8603489995002747,
+      "learning_rate": 3.883827993344691e-06,
+      "loss": 0.5827,
+      "step": 5422
+    },
+    {
+      "epoch": 1.9041432584269664,
+      "grad_norm": 0.8876716494560242,
+      "learning_rate": 3.88344293205236e-06,
+      "loss": 0.5968,
+      "step": 5423
+    },
+    {
+      "epoch": 1.904494382022472,
+      "grad_norm": 0.9033610820770264,
+      "learning_rate": 3.883057823447464e-06,
+      "loss": 0.6301,
+      "step": 5424
+    },
+    {
+      "epoch": 1.9048455056179776,
+      "grad_norm": 0.9055176377296448,
+      "learning_rate": 3.882672667543174e-06,
+      "loss": 0.6085,
+      "step": 5425
+    },
+    {
+      "epoch": 1.9051966292134832,
+      "grad_norm": 0.8801239132881165,
+      "learning_rate": 3.882287464352662e-06,
+      "loss": 0.5865,
+      "step": 5426
+    },
+    {
+      "epoch": 1.9055477528089888,
+      "grad_norm": 0.8756425380706787,
+      "learning_rate": 3.8819022138891e-06,
+      "loss": 0.6097,
+      "step": 5427
+    },
+    {
+      "epoch": 1.9058988764044944,
+      "grad_norm": 0.8778197169303894,
+      "learning_rate": 3.881516916165668e-06,
+      "loss": 0.6026,
+      "step": 5428
+    },
+    {
+      "epoch": 1.90625,
+      "grad_norm": 0.8809471726417542,
+      "learning_rate": 3.881131571195537e-06,
+      "loss": 0.5967,
+      "step": 5429
+    },
+    {
+      "epoch": 1.9066011235955056,
+      "grad_norm": 0.854507565498352,
+      "learning_rate": 3.880746178991889e-06,
+      "loss": 0.5973,
+      "step": 5430
+    },
+    {
+      "epoch": 1.9069522471910112,
+      "grad_norm": 0.8528857827186584,
+      "learning_rate": 3.8803607395679035e-06,
+      "loss": 0.5809,
+      "step": 5431
+    },
+    {
+      "epoch": 1.9073033707865168,
+      "grad_norm": 0.8797081112861633,
+      "learning_rate": 3.879975252936761e-06,
+      "loss": 0.6207,
+      "step": 5432
+    },
+    {
+      "epoch": 1.9076544943820224,
+      "grad_norm": 0.862160325050354,
+      "learning_rate": 3.879589719111646e-06,
+      "loss": 0.5845,
+      "step": 5433
+    },
+    {
+      "epoch": 1.908005617977528,
+      "grad_norm": 0.8742191791534424,
+      "learning_rate": 3.879204138105744e-06,
+      "loss": 0.6332,
+      "step": 5434
+    },
+    {
+      "epoch": 1.9083567415730336,
+      "grad_norm": 0.9287744760513306,
+      "learning_rate": 3.87881850993224e-06,
+      "loss": 0.6571,
+      "step": 5435
+    },
+    {
+      "epoch": 1.9087078651685392,
+      "grad_norm": 0.8513810634613037,
+      "learning_rate": 3.878432834604324e-06,
+      "loss": 0.5712,
+      "step": 5436
+    },
+    {
+      "epoch": 1.9090589887640448,
+      "grad_norm": 0.9044253826141357,
+      "learning_rate": 3.878047112135183e-06,
+      "loss": 0.6297,
+      "step": 5437
+    },
+    {
+      "epoch": 1.9094101123595506,
+      "grad_norm": 0.8363608717918396,
+      "learning_rate": 3.877661342538011e-06,
+      "loss": 0.6032,
+      "step": 5438
+    },
+    {
+      "epoch": 1.9097612359550562,
+      "grad_norm": 0.9041305184364319,
+      "learning_rate": 3.877275525826e-06,
+      "loss": 0.6192,
+      "step": 5439
+    },
+    {
+      "epoch": 1.9101123595505618,
+      "grad_norm": 0.8639614582061768,
+      "learning_rate": 3.876889662012343e-06,
+      "loss": 0.6213,
+      "step": 5440
+    },
+    {
+      "epoch": 1.9104634831460674,
+      "grad_norm": 0.8700418472290039,
+      "learning_rate": 3.876503751110239e-06,
+      "loss": 0.6209,
+      "step": 5441
+    },
+    {
+      "epoch": 1.910814606741573,
+      "grad_norm": 0.899722158908844,
+      "learning_rate": 3.876117793132884e-06,
+      "loss": 0.6461,
+      "step": 5442
+    },
+    {
+      "epoch": 1.9111657303370788,
+      "grad_norm": 0.8731403350830078,
+      "learning_rate": 3.875731788093478e-06,
+      "loss": 0.6135,
+      "step": 5443
+    },
+    {
+      "epoch": 1.9115168539325844,
+      "grad_norm": 0.8795063495635986,
+      "learning_rate": 3.8753457360052235e-06,
+      "loss": 0.6442,
+      "step": 5444
+    },
+    {
+      "epoch": 1.91186797752809,
+      "grad_norm": 0.8663721680641174,
+      "learning_rate": 3.874959636881321e-06,
+      "loss": 0.6225,
+      "step": 5445
+    },
+    {
+      "epoch": 1.9122191011235956,
+      "grad_norm": 0.863272488117218,
+      "learning_rate": 3.874573490734975e-06,
+      "loss": 0.6031,
+      "step": 5446
+    },
+    {
+      "epoch": 1.9125702247191012,
+      "grad_norm": 0.8670037388801575,
+      "learning_rate": 3.874187297579392e-06,
+      "loss": 0.6105,
+      "step": 5447
+    },
+    {
+      "epoch": 1.9129213483146068,
+      "grad_norm": 0.898658812046051,
+      "learning_rate": 3.87380105742778e-06,
+      "loss": 0.6176,
+      "step": 5448
+    },
+    {
+      "epoch": 1.9132724719101124,
+      "grad_norm": 0.8771390914916992,
+      "learning_rate": 3.873414770293347e-06,
+      "loss": 0.5724,
+      "step": 5449
+    },
+    {
+      "epoch": 1.913623595505618,
+      "grad_norm": 0.8916599750518799,
+      "learning_rate": 3.873028436189304e-06,
+      "loss": 0.6308,
+      "step": 5450
+    },
+    {
+      "epoch": 1.9139747191011236,
+      "grad_norm": 0.8692615628242493,
+      "learning_rate": 3.872642055128863e-06,
+      "loss": 0.5857,
+      "step": 5451
+    },
+    {
+      "epoch": 1.9143258426966292,
+      "grad_norm": 0.9091446399688721,
+      "learning_rate": 3.872255627125238e-06,
+      "loss": 0.6443,
+      "step": 5452
+    },
+    {
+      "epoch": 1.9146769662921348,
+      "grad_norm": 0.8689329624176025,
+      "learning_rate": 3.871869152191646e-06,
+      "loss": 0.6319,
+      "step": 5453
+    },
+    {
+      "epoch": 1.9150280898876404,
+      "grad_norm": 0.8973026275634766,
+      "learning_rate": 3.871482630341303e-06,
+      "loss": 0.6078,
+      "step": 5454
+    },
+    {
+      "epoch": 1.915379213483146,
+      "grad_norm": 0.8688374161720276,
+      "learning_rate": 3.871096061587427e-06,
+      "loss": 0.5782,
+      "step": 5455
+    },
+    {
+      "epoch": 1.9157303370786516,
+      "grad_norm": 0.9019613862037659,
+      "learning_rate": 3.8707094459432384e-06,
+      "loss": 0.6484,
+      "step": 5456
+    },
+    {
+      "epoch": 1.9160814606741572,
+      "grad_norm": 0.8789473176002502,
+      "learning_rate": 3.870322783421961e-06,
+      "loss": 0.5992,
+      "step": 5457
+    },
+    {
+      "epoch": 1.9164325842696628,
+      "grad_norm": 0.8779850602149963,
+      "learning_rate": 3.869936074036816e-06,
+      "loss": 0.5993,
+      "step": 5458
+    },
+    {
+      "epoch": 1.9167837078651684,
+      "grad_norm": 0.8998771905899048,
+      "learning_rate": 3.86954931780103e-06,
+      "loss": 0.6341,
+      "step": 5459
+    },
+    {
+      "epoch": 1.9171348314606742,
+      "grad_norm": 0.8945295214653015,
+      "learning_rate": 3.869162514727829e-06,
+      "loss": 0.6197,
+      "step": 5460
+    },
+    {
+      "epoch": 1.9174859550561798,
+      "grad_norm": 0.8839888572692871,
+      "learning_rate": 3.868775664830442e-06,
+      "loss": 0.6411,
+      "step": 5461
+    },
+    {
+      "epoch": 1.9178370786516854,
+      "grad_norm": 0.8995748162269592,
+      "learning_rate": 3.868388768122098e-06,
+      "loss": 0.6471,
+      "step": 5462
+    },
+    {
+      "epoch": 1.918188202247191,
+      "grad_norm": 0.8744586706161499,
+      "learning_rate": 3.868001824616029e-06,
+      "loss": 0.6023,
+      "step": 5463
+    },
+    {
+      "epoch": 1.9185393258426966,
+      "grad_norm": 0.8731626272201538,
+      "learning_rate": 3.867614834325469e-06,
+      "loss": 0.6145,
+      "step": 5464
+    },
+    {
+      "epoch": 1.9188904494382022,
+      "grad_norm": 0.8779453635215759,
+      "learning_rate": 3.867227797263652e-06,
+      "loss": 0.5688,
+      "step": 5465
+    },
+    {
+      "epoch": 1.919241573033708,
+      "grad_norm": 0.894486129283905,
+      "learning_rate": 3.866840713443814e-06,
+      "loss": 0.5936,
+      "step": 5466
+    },
+    {
+      "epoch": 1.9195926966292136,
+      "grad_norm": 0.9034679532051086,
+      "learning_rate": 3.866453582879193e-06,
+      "loss": 0.6005,
+      "step": 5467
+    },
+    {
+      "epoch": 1.9199438202247192,
+      "grad_norm": 0.8829547166824341,
+      "learning_rate": 3.866066405583028e-06,
+      "loss": 0.6141,
+      "step": 5468
+    },
+    {
+      "epoch": 1.9202949438202248,
+      "grad_norm": 0.8914617896080017,
+      "learning_rate": 3.865679181568561e-06,
+      "loss": 0.6344,
+      "step": 5469
+    },
+    {
+      "epoch": 1.9206460674157304,
+      "grad_norm": 0.8478212952613831,
+      "learning_rate": 3.865291910849036e-06,
+      "loss": 0.5906,
+      "step": 5470
+    },
+    {
+      "epoch": 1.920997191011236,
+      "grad_norm": 0.9483685493469238,
+      "learning_rate": 3.864904593437696e-06,
+      "loss": 0.6285,
+      "step": 5471
+    },
+    {
+      "epoch": 1.9213483146067416,
+      "grad_norm": 0.8849635124206543,
+      "learning_rate": 3.8645172293477855e-06,
+      "loss": 0.6054,
+      "step": 5472
+    },
+    {
+      "epoch": 1.9216994382022472,
+      "grad_norm": 0.8696728348731995,
+      "learning_rate": 3.864129818592554e-06,
+      "loss": 0.63,
+      "step": 5473
+    },
+    {
+      "epoch": 1.9220505617977528,
+      "grad_norm": 0.8640512824058533,
+      "learning_rate": 3.86374236118525e-06,
+      "loss": 0.5815,
+      "step": 5474
+    },
+    {
+      "epoch": 1.9224016853932584,
+      "grad_norm": 0.9038639068603516,
+      "learning_rate": 3.8633548571391245e-06,
+      "loss": 0.6459,
+      "step": 5475
+    },
+    {
+      "epoch": 1.922752808988764,
+      "grad_norm": 0.8907269835472107,
+      "learning_rate": 3.862967306467429e-06,
+      "loss": 0.6009,
+      "step": 5476
+    },
+    {
+      "epoch": 1.9231039325842696,
+      "grad_norm": 0.9073399305343628,
+      "learning_rate": 3.862579709183418e-06,
+      "loss": 0.6359,
+      "step": 5477
+    },
+    {
+      "epoch": 1.9234550561797752,
+      "grad_norm": 0.8706175684928894,
+      "learning_rate": 3.862192065300347e-06,
+      "loss": 0.6203,
+      "step": 5478
+    },
+    {
+      "epoch": 1.9238061797752808,
+      "grad_norm": 0.874502956867218,
+      "learning_rate": 3.861804374831474e-06,
+      "loss": 0.5929,
+      "step": 5479
+    },
+    {
+      "epoch": 1.9241573033707864,
+      "grad_norm": 0.846248984336853,
+      "learning_rate": 3.861416637790056e-06,
+      "loss": 0.5648,
+      "step": 5480
+    },
+    {
+      "epoch": 1.924508426966292,
+      "grad_norm": 0.8775991797447205,
+      "learning_rate": 3.861028854189353e-06,
+      "loss": 0.6009,
+      "step": 5481
+    },
+    {
+      "epoch": 1.9248595505617978,
+      "grad_norm": 0.9060062766075134,
+      "learning_rate": 3.860641024042628e-06,
+      "loss": 0.6432,
+      "step": 5482
+    },
+    {
+      "epoch": 1.9252106741573034,
+      "grad_norm": 0.8684884309768677,
+      "learning_rate": 3.8602531473631456e-06,
+      "loss": 0.5969,
+      "step": 5483
+    },
+    {
+      "epoch": 1.925561797752809,
+      "grad_norm": 0.9077668190002441,
+      "learning_rate": 3.859865224164169e-06,
+      "loss": 0.6311,
+      "step": 5484
+    },
+    {
+      "epoch": 1.9259129213483146,
+      "grad_norm": 0.8688258528709412,
+      "learning_rate": 3.8594772544589656e-06,
+      "loss": 0.6016,
+      "step": 5485
+    },
+    {
+      "epoch": 1.9262640449438202,
+      "grad_norm": 0.8717395663261414,
+      "learning_rate": 3.8590892382608025e-06,
+      "loss": 0.6134,
+      "step": 5486
+    },
+    {
+      "epoch": 1.9266151685393258,
+      "grad_norm": 0.8912414312362671,
+      "learning_rate": 3.858701175582951e-06,
+      "loss": 0.6481,
+      "step": 5487
+    },
+    {
+      "epoch": 1.9269662921348316,
+      "grad_norm": 0.858781099319458,
+      "learning_rate": 3.858313066438682e-06,
+      "loss": 0.6464,
+      "step": 5488
+    },
+    {
+      "epoch": 1.9273174157303372,
+      "grad_norm": 0.858666718006134,
+      "learning_rate": 3.857924910841269e-06,
+      "loss": 0.6094,
+      "step": 5489
+    },
+    {
+      "epoch": 1.9276685393258428,
+      "grad_norm": 0.8739560842514038,
+      "learning_rate": 3.857536708803985e-06,
+      "loss": 0.6047,
+      "step": 5490
+    },
+    {
+      "epoch": 1.9280196629213484,
+      "grad_norm": 0.8531132936477661,
+      "learning_rate": 3.857148460340107e-06,
+      "loss": 0.586,
+      "step": 5491
+    },
+    {
+      "epoch": 1.928370786516854,
+      "grad_norm": 0.8856545686721802,
+      "learning_rate": 3.856760165462914e-06,
+      "loss": 0.5874,
+      "step": 5492
+    },
+    {
+      "epoch": 1.9287219101123596,
+      "grad_norm": 0.9217509627342224,
+      "learning_rate": 3.856371824185684e-06,
+      "loss": 0.5857,
+      "step": 5493
+    },
+    {
+      "epoch": 1.9290730337078652,
+      "grad_norm": 0.845321774482727,
+      "learning_rate": 3.855983436521698e-06,
+      "loss": 0.6045,
+      "step": 5494
+    },
+    {
+      "epoch": 1.9294241573033708,
+      "grad_norm": 0.84823077917099,
+      "learning_rate": 3.855595002484239e-06,
+      "loss": 0.621,
+      "step": 5495
+    },
+    {
+      "epoch": 1.9297752808988764,
+      "grad_norm": 1.008187174797058,
+      "learning_rate": 3.855206522086591e-06,
+      "loss": 0.6289,
+      "step": 5496
+    },
+    {
+      "epoch": 1.930126404494382,
+      "grad_norm": 0.8827589154243469,
+      "learning_rate": 3.8548179953420405e-06,
+      "loss": 0.5957,
+      "step": 5497
+    },
+    {
+      "epoch": 1.9304775280898876,
+      "grad_norm": 0.8767725825309753,
+      "learning_rate": 3.854429422263872e-06,
+      "loss": 0.5905,
+      "step": 5498
+    },
+    {
+      "epoch": 1.9308286516853932,
+      "grad_norm": 0.8842640519142151,
+      "learning_rate": 3.854040802865377e-06,
+      "loss": 0.6496,
+      "step": 5499
+    },
+    {
+      "epoch": 1.9311797752808988,
+      "grad_norm": 0.8716596364974976,
+      "learning_rate": 3.853652137159846e-06,
+      "loss": 0.6259,
+      "step": 5500
+    },
+    {
+      "epoch": 1.9315308988764044,
+      "grad_norm": 0.8906645774841309,
+      "learning_rate": 3.85326342516057e-06,
+      "loss": 0.5954,
+      "step": 5501
+    },
+    {
+      "epoch": 1.93188202247191,
+      "grad_norm": 0.8781253099441528,
+      "learning_rate": 3.8528746668808425e-06,
+      "loss": 0.6137,
+      "step": 5502
+    },
+    {
+      "epoch": 1.9322331460674156,
+      "grad_norm": 0.9453568458557129,
+      "learning_rate": 3.8524858623339585e-06,
+      "loss": 0.6333,
+      "step": 5503
+    },
+    {
+      "epoch": 1.9325842696629212,
+      "grad_norm": 0.8589761257171631,
+      "learning_rate": 3.852097011533217e-06,
+      "loss": 0.6573,
+      "step": 5504
+    },
+    {
+      "epoch": 1.932935393258427,
+      "grad_norm": 1.0276621580123901,
+      "learning_rate": 3.851708114491913e-06,
+      "loss": 0.594,
+      "step": 5505
+    },
+    {
+      "epoch": 1.9332865168539326,
+      "grad_norm": 0.8756514191627502,
+      "learning_rate": 3.851319171223349e-06,
+      "loss": 0.6206,
+      "step": 5506
+    },
+    {
+      "epoch": 1.9336376404494382,
+      "grad_norm": 0.8564157485961914,
+      "learning_rate": 3.850930181740826e-06,
+      "loss": 0.6062,
+      "step": 5507
+    },
+    {
+      "epoch": 1.9339887640449438,
+      "grad_norm": 0.8514534831047058,
+      "learning_rate": 3.8505411460576455e-06,
+      "loss": 0.5625,
+      "step": 5508
+    },
+    {
+      "epoch": 1.9343398876404494,
+      "grad_norm": 0.9018797874450684,
+      "learning_rate": 3.850152064187115e-06,
+      "loss": 0.6357,
+      "step": 5509
+    },
+    {
+      "epoch": 1.9346910112359552,
+      "grad_norm": 0.8431733250617981,
+      "learning_rate": 3.849762936142538e-06,
+      "loss": 0.6041,
+      "step": 5510
+    },
+    {
+      "epoch": 1.9350421348314608,
+      "grad_norm": 0.8911080360412598,
+      "learning_rate": 3.849373761937225e-06,
+      "loss": 0.6077,
+      "step": 5511
+    },
+    {
+      "epoch": 1.9353932584269664,
+      "grad_norm": 0.8887476921081543,
+      "learning_rate": 3.848984541584483e-06,
+      "loss": 0.6041,
+      "step": 5512
+    },
+    {
+      "epoch": 1.935744382022472,
+      "grad_norm": 0.8878927826881409,
+      "learning_rate": 3.848595275097624e-06,
+      "loss": 0.6169,
+      "step": 5513
+    },
+    {
+      "epoch": 1.9360955056179776,
+      "grad_norm": 0.9070707559585571,
+      "learning_rate": 3.848205962489961e-06,
+      "loss": 0.5982,
+      "step": 5514
+    },
+    {
+      "epoch": 1.9364466292134832,
+      "grad_norm": 1.2515376806259155,
+      "learning_rate": 3.847816603774807e-06,
+      "loss": 0.6475,
+      "step": 5515
+    },
+    {
+      "epoch": 1.9367977528089888,
+      "grad_norm": 0.8596694469451904,
+      "learning_rate": 3.84742719896548e-06,
+      "loss": 0.6293,
+      "step": 5516
+    },
+    {
+      "epoch": 1.9371488764044944,
+      "grad_norm": 0.9003183841705322,
+      "learning_rate": 3.847037748075294e-06,
+      "loss": 0.6235,
+      "step": 5517
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 0.8805832266807556,
+      "learning_rate": 3.846648251117571e-06,
+      "loss": 0.5892,
+      "step": 5518
+    },
+    {
+      "epoch": 1.9378511235955056,
+      "grad_norm": 0.8691025972366333,
+      "learning_rate": 3.846258708105629e-06,
+      "loss": 0.5987,
+      "step": 5519
+    },
+    {
+      "epoch": 1.9382022471910112,
+      "grad_norm": 0.8848376870155334,
+      "learning_rate": 3.8458691190527915e-06,
+      "loss": 0.5966,
+      "step": 5520
+    },
+    {
+      "epoch": 1.9385533707865168,
+      "grad_norm": 0.8585199117660522,
+      "learning_rate": 3.845479483972383e-06,
+      "loss": 0.5966,
+      "step": 5521
+    },
+    {
+      "epoch": 1.9389044943820224,
+      "grad_norm": 0.8973793983459473,
+      "learning_rate": 3.845089802877727e-06,
+      "loss": 0.6159,
+      "step": 5522
+    },
+    {
+      "epoch": 1.939255617977528,
+      "grad_norm": 0.8836932182312012,
+      "learning_rate": 3.84470007578215e-06,
+      "loss": 0.6267,
+      "step": 5523
+    },
+    {
+      "epoch": 1.9396067415730336,
+      "grad_norm": 0.861569344997406,
+      "learning_rate": 3.844310302698982e-06,
+      "loss": 0.6287,
+      "step": 5524
+    },
+    {
+      "epoch": 1.9399578651685392,
+      "grad_norm": 0.9185940027236938,
+      "learning_rate": 3.843920483641551e-06,
+      "loss": 0.6334,
+      "step": 5525
+    },
+    {
+      "epoch": 1.9403089887640448,
+      "grad_norm": 0.8938425779342651,
+      "learning_rate": 3.8435306186231905e-06,
+      "loss": 0.609,
+      "step": 5526
+    },
+    {
+      "epoch": 1.9406601123595506,
+      "grad_norm": 0.8964993357658386,
+      "learning_rate": 3.8431407076572316e-06,
+      "loss": 0.5996,
+      "step": 5527
+    },
+    {
+      "epoch": 1.9410112359550562,
+      "grad_norm": 0.8518646359443665,
+      "learning_rate": 3.8427507507570104e-06,
+      "loss": 0.5862,
+      "step": 5528
+    },
+    {
+      "epoch": 1.9413623595505618,
+      "grad_norm": 0.8727825880050659,
+      "learning_rate": 3.842360747935863e-06,
+      "loss": 0.627,
+      "step": 5529
+    },
+    {
+      "epoch": 1.9417134831460674,
+      "grad_norm": 0.8543184399604797,
+      "learning_rate": 3.841970699207125e-06,
+      "loss": 0.5737,
+      "step": 5530
+    },
+    {
+      "epoch": 1.942064606741573,
+      "grad_norm": 0.8790775537490845,
+      "learning_rate": 3.841580604584139e-06,
+      "loss": 0.5938,
+      "step": 5531
+    },
+    {
+      "epoch": 1.9424157303370788,
+      "grad_norm": 0.8977394104003906,
+      "learning_rate": 3.8411904640802435e-06,
+      "loss": 0.6028,
+      "step": 5532
+    },
+    {
+      "epoch": 1.9427668539325844,
+      "grad_norm": 0.8888459205627441,
+      "learning_rate": 3.8408002777087825e-06,
+      "loss": 0.5876,
+      "step": 5533
+    },
+    {
+      "epoch": 1.94311797752809,
+      "grad_norm": 0.8542887568473816,
+      "learning_rate": 3.840410045483098e-06,
+      "loss": 0.5762,
+      "step": 5534
+    },
+    {
+      "epoch": 1.9434691011235956,
+      "grad_norm": 0.9694353342056274,
+      "learning_rate": 3.8400197674165385e-06,
+      "loss": 0.6231,
+      "step": 5535
+    },
+    {
+      "epoch": 1.9438202247191012,
+      "grad_norm": 0.9098255634307861,
+      "learning_rate": 3.839629443522449e-06,
+      "loss": 0.6231,
+      "step": 5536
+    },
+    {
+      "epoch": 1.9441713483146068,
+      "grad_norm": 0.9094727635383606,
+      "learning_rate": 3.8392390738141784e-06,
+      "loss": 0.6039,
+      "step": 5537
+    },
+    {
+      "epoch": 1.9445224719101124,
+      "grad_norm": 0.8750001788139343,
+      "learning_rate": 3.838848658305079e-06,
+      "loss": 0.6298,
+      "step": 5538
+    },
+    {
+      "epoch": 1.944873595505618,
+      "grad_norm": 0.892129123210907,
+      "learning_rate": 3.8384581970085e-06,
+      "loss": 0.6056,
+      "step": 5539
+    },
+    {
+      "epoch": 1.9452247191011236,
+      "grad_norm": 0.8831581473350525,
+      "learning_rate": 3.8380676899377965e-06,
+      "loss": 0.6372,
+      "step": 5540
+    },
+    {
+      "epoch": 1.9455758426966292,
+      "grad_norm": 0.8667299151420593,
+      "learning_rate": 3.837677137106322e-06,
+      "loss": 0.621,
+      "step": 5541
+    },
+    {
+      "epoch": 1.9459269662921348,
+      "grad_norm": 0.8704750537872314,
+      "learning_rate": 3.837286538527435e-06,
+      "loss": 0.6052,
+      "step": 5542
+    },
+    {
+      "epoch": 1.9462780898876404,
+      "grad_norm": 0.9060012102127075,
+      "learning_rate": 3.836895894214492e-06,
+      "loss": 0.5917,
+      "step": 5543
+    },
+    {
+      "epoch": 1.946629213483146,
+      "grad_norm": 0.8708364963531494,
+      "learning_rate": 3.836505204180854e-06,
+      "loss": 0.6443,
+      "step": 5544
+    },
+    {
+      "epoch": 1.9469803370786516,
+      "grad_norm": 0.881945788860321,
+      "learning_rate": 3.836114468439882e-06,
+      "loss": 0.6284,
+      "step": 5545
+    },
+    {
+      "epoch": 1.9473314606741572,
+      "grad_norm": 0.8595880270004272,
+      "learning_rate": 3.835723687004938e-06,
+      "loss": 0.6034,
+      "step": 5546
+    },
+    {
+      "epoch": 1.9476825842696628,
+      "grad_norm": 0.8725127577781677,
+      "learning_rate": 3.8353328598893865e-06,
+      "loss": 0.6082,
+      "step": 5547
+    },
+    {
+      "epoch": 1.9480337078651684,
+      "grad_norm": 0.8611936569213867,
+      "learning_rate": 3.834941987106594e-06,
+      "loss": 0.5814,
+      "step": 5548
+    },
+    {
+      "epoch": 1.9483848314606742,
+      "grad_norm": 0.8612635731697083,
+      "learning_rate": 3.834551068669928e-06,
+      "loss": 0.5771,
+      "step": 5549
+    },
+    {
+      "epoch": 1.9487359550561798,
+      "grad_norm": 0.8790583610534668,
+      "learning_rate": 3.834160104592758e-06,
+      "loss": 0.5968,
+      "step": 5550
+    },
+    {
+      "epoch": 1.9490870786516854,
+      "grad_norm": 0.9288917183876038,
+      "learning_rate": 3.833769094888454e-06,
+      "loss": 0.6283,
+      "step": 5551
+    },
+    {
+      "epoch": 1.949438202247191,
+      "grad_norm": 0.8635450005531311,
+      "learning_rate": 3.833378039570388e-06,
+      "loss": 0.6024,
+      "step": 5552
+    },
+    {
+      "epoch": 1.9497893258426966,
+      "grad_norm": 0.9010553359985352,
+      "learning_rate": 3.8329869386519335e-06,
+      "loss": 0.5933,
+      "step": 5553
+    },
+    {
+      "epoch": 1.9501404494382022,
+      "grad_norm": 0.8576677441596985,
+      "learning_rate": 3.832595792146468e-06,
+      "loss": 0.6018,
+      "step": 5554
+    },
+    {
+      "epoch": 1.950491573033708,
+      "grad_norm": 0.9220290780067444,
+      "learning_rate": 3.832204600067365e-06,
+      "loss": 0.6292,
+      "step": 5555
+    },
+    {
+      "epoch": 1.9508426966292136,
+      "grad_norm": 0.8435717821121216,
+      "learning_rate": 3.831813362428005e-06,
+      "loss": 0.581,
+      "step": 5556
+    },
+    {
+      "epoch": 1.9511938202247192,
+      "grad_norm": 0.9167920351028442,
+      "learning_rate": 3.831422079241768e-06,
+      "loss": 0.621,
+      "step": 5557
+    },
+    {
+      "epoch": 1.9515449438202248,
+      "grad_norm": 0.8408527374267578,
+      "learning_rate": 3.831030750522034e-06,
+      "loss": 0.5921,
+      "step": 5558
+    },
+    {
+      "epoch": 1.9518960674157304,
+      "grad_norm": 0.8575167655944824,
+      "learning_rate": 3.830639376282189e-06,
+      "loss": 0.5961,
+      "step": 5559
+    },
+    {
+      "epoch": 1.952247191011236,
+      "grad_norm": 0.8758642673492432,
+      "learning_rate": 3.8302479565356154e-06,
+      "loss": 0.657,
+      "step": 5560
+    },
+    {
+      "epoch": 1.9525983146067416,
+      "grad_norm": 0.8745916485786438,
+      "learning_rate": 3.8298564912957e-06,
+      "loss": 0.6317,
+      "step": 5561
+    },
+    {
+      "epoch": 1.9529494382022472,
+      "grad_norm": 0.867461621761322,
+      "learning_rate": 3.829464980575831e-06,
+      "loss": 0.6211,
+      "step": 5562
+    },
+    {
+      "epoch": 1.9533005617977528,
+      "grad_norm": 0.8979873657226562,
+      "learning_rate": 3.829073424389396e-06,
+      "loss": 0.6253,
+      "step": 5563
+    },
+    {
+      "epoch": 1.9536516853932584,
+      "grad_norm": 0.8740954399108887,
+      "learning_rate": 3.828681822749789e-06,
+      "loss": 0.6189,
+      "step": 5564
+    },
+    {
+      "epoch": 1.954002808988764,
+      "grad_norm": 0.8818367123603821,
+      "learning_rate": 3.828290175670399e-06,
+      "loss": 0.6066,
+      "step": 5565
+    },
+    {
+      "epoch": 1.9543539325842696,
+      "grad_norm": 0.9065709114074707,
+      "learning_rate": 3.8278984831646225e-06,
+      "loss": 0.638,
+      "step": 5566
+    },
+    {
+      "epoch": 1.9547050561797752,
+      "grad_norm": 0.8510881066322327,
+      "learning_rate": 3.827506745245854e-06,
+      "loss": 0.587,
+      "step": 5567
+    },
+    {
+      "epoch": 1.9550561797752808,
+      "grad_norm": 0.8661137819290161,
+      "learning_rate": 3.827114961927491e-06,
+      "loss": 0.6402,
+      "step": 5568
+    },
+    {
+      "epoch": 1.9554073033707864,
+      "grad_norm": 0.8897769451141357,
+      "learning_rate": 3.8267231332229315e-06,
+      "loss": 0.6144,
+      "step": 5569
+    },
+    {
+      "epoch": 1.955758426966292,
+      "grad_norm": 0.9311615824699402,
+      "learning_rate": 3.826331259145576e-06,
+      "loss": 0.5961,
+      "step": 5570
+    },
+    {
+      "epoch": 1.9561095505617978,
+      "grad_norm": 0.8808618187904358,
+      "learning_rate": 3.825939339708827e-06,
+      "loss": 0.6238,
+      "step": 5571
+    },
+    {
+      "epoch": 1.9564606741573034,
+      "grad_norm": 0.8663437962532043,
+      "learning_rate": 3.825547374926087e-06,
+      "loss": 0.5774,
+      "step": 5572
+    },
+    {
+      "epoch": 1.956811797752809,
+      "grad_norm": 0.8695099353790283,
+      "learning_rate": 3.825155364810761e-06,
+      "loss": 0.6166,
+      "step": 5573
+    },
+    {
+      "epoch": 1.9571629213483146,
+      "grad_norm": 0.890457272529602,
+      "learning_rate": 3.824763309376256e-06,
+      "loss": 0.6004,
+      "step": 5574
+    },
+    {
+      "epoch": 1.9575140449438202,
+      "grad_norm": 0.8851537108421326,
+      "learning_rate": 3.8243712086359785e-06,
+      "loss": 0.5441,
+      "step": 5575
+    },
+    {
+      "epoch": 1.9578651685393258,
+      "grad_norm": 0.8680998086929321,
+      "learning_rate": 3.823979062603339e-06,
+      "loss": 0.634,
+      "step": 5576
+    },
+    {
+      "epoch": 1.9582162921348316,
+      "grad_norm": 0.8861505389213562,
+      "learning_rate": 3.823586871291749e-06,
+      "loss": 0.6083,
+      "step": 5577
+    },
+    {
+      "epoch": 1.9585674157303372,
+      "grad_norm": 0.853298544883728,
+      "learning_rate": 3.823194634714621e-06,
+      "loss": 0.6065,
+      "step": 5578
+    },
+    {
+      "epoch": 1.9589185393258428,
+      "grad_norm": 0.896482527256012,
+      "learning_rate": 3.822802352885367e-06,
+      "loss": 0.6085,
+      "step": 5579
+    },
+    {
+      "epoch": 1.9592696629213484,
+      "grad_norm": 0.8784202337265015,
+      "learning_rate": 3.8224100258174066e-06,
+      "loss": 0.6141,
+      "step": 5580
+    },
+    {
+      "epoch": 1.959620786516854,
+      "grad_norm": 0.8632230162620544,
+      "learning_rate": 3.8220176535241546e-06,
+      "loss": 0.6062,
+      "step": 5581
+    },
+    {
+      "epoch": 1.9599719101123596,
+      "grad_norm": 0.8777154684066772,
+      "learning_rate": 3.821625236019029e-06,
+      "loss": 0.5994,
+      "step": 5582
+    },
+    {
+      "epoch": 1.9603230337078652,
+      "grad_norm": 0.8801113963127136,
+      "learning_rate": 3.821232773315452e-06,
+      "loss": 0.6037,
+      "step": 5583
+    },
+    {
+      "epoch": 1.9606741573033708,
+      "grad_norm": 0.8997433185577393,
+      "learning_rate": 3.820840265426844e-06,
+      "loss": 0.641,
+      "step": 5584
+    },
+    {
+      "epoch": 1.9610252808988764,
+      "grad_norm": 0.9067327976226807,
+      "learning_rate": 3.82044771236663e-06,
+      "loss": 0.6045,
+      "step": 5585
+    },
+    {
+      "epoch": 1.961376404494382,
+      "grad_norm": 0.8806515336036682,
+      "learning_rate": 3.820055114148234e-06,
+      "loss": 0.6139,
+      "step": 5586
+    },
+    {
+      "epoch": 1.9617275280898876,
+      "grad_norm": 0.8809478282928467,
+      "learning_rate": 3.819662470785082e-06,
+      "loss": 0.6166,
+      "step": 5587
+    },
+    {
+      "epoch": 1.9620786516853932,
+      "grad_norm": 0.8271637558937073,
+      "learning_rate": 3.819269782290603e-06,
+      "loss": 0.5549,
+      "step": 5588
+    },
+    {
+      "epoch": 1.9624297752808988,
+      "grad_norm": 0.866726815700531,
+      "learning_rate": 3.8188770486782275e-06,
+      "loss": 0.5978,
+      "step": 5589
+    },
+    {
+      "epoch": 1.9627808988764044,
+      "grad_norm": 0.8871544003486633,
+      "learning_rate": 3.818484269961385e-06,
+      "loss": 0.6149,
+      "step": 5590
+    },
+    {
+      "epoch": 1.96313202247191,
+      "grad_norm": 0.9260693788528442,
+      "learning_rate": 3.818091446153508e-06,
+      "loss": 0.6066,
+      "step": 5591
+    },
+    {
+      "epoch": 1.9634831460674156,
+      "grad_norm": 0.867003858089447,
+      "learning_rate": 3.817698577268032e-06,
+      "loss": 0.6245,
+      "step": 5592
+    },
+    {
+      "epoch": 1.9638342696629212,
+      "grad_norm": 0.8828530311584473,
+      "learning_rate": 3.817305663318393e-06,
+      "loss": 0.6173,
+      "step": 5593
+    },
+    {
+      "epoch": 1.964185393258427,
+      "grad_norm": 0.8841493725776672,
+      "learning_rate": 3.8169127043180255e-06,
+      "loss": 0.5633,
+      "step": 5594
+    },
+    {
+      "epoch": 1.9645365168539326,
+      "grad_norm": 0.881924569606781,
+      "learning_rate": 3.816519700280372e-06,
+      "loss": 0.5844,
+      "step": 5595
+    },
+    {
+      "epoch": 1.9648876404494382,
+      "grad_norm": 0.8986069560050964,
+      "learning_rate": 3.816126651218871e-06,
+      "loss": 0.6792,
+      "step": 5596
+    },
+    {
+      "epoch": 1.9652387640449438,
+      "grad_norm": 0.8712230920791626,
+      "learning_rate": 3.815733557146964e-06,
+      "loss": 0.6117,
+      "step": 5597
+    },
+    {
+      "epoch": 1.9655898876404494,
+      "grad_norm": 0.9183915853500366,
+      "learning_rate": 3.815340418078096e-06,
+      "loss": 0.6554,
+      "step": 5598
+    },
+    {
+      "epoch": 1.9659410112359552,
+      "grad_norm": 0.8475895524024963,
+      "learning_rate": 3.814947234025711e-06,
+      "loss": 0.5581,
+      "step": 5599
+    },
+    {
+      "epoch": 1.9662921348314608,
+      "grad_norm": 0.881952702999115,
+      "learning_rate": 3.8145540050032558e-06,
+      "loss": 0.6173,
+      "step": 5600
+    },
+    {
+      "epoch": 1.9666432584269664,
+      "grad_norm": 0.8803291916847229,
+      "learning_rate": 3.814160731024179e-06,
+      "loss": 0.6322,
+      "step": 5601
+    },
+    {
+      "epoch": 1.966994382022472,
+      "grad_norm": 0.8611246347427368,
+      "learning_rate": 3.8137674121019296e-06,
+      "loss": 0.6011,
+      "step": 5602
+    },
+    {
+      "epoch": 1.9673455056179776,
+      "grad_norm": 0.8588524460792542,
+      "learning_rate": 3.8133740482499582e-06,
+      "loss": 0.614,
+      "step": 5603
+    },
+    {
+      "epoch": 1.9676966292134832,
+      "grad_norm": 0.872290313243866,
+      "learning_rate": 3.8129806394817186e-06,
+      "loss": 0.6002,
+      "step": 5604
+    },
+    {
+      "epoch": 1.9680477528089888,
+      "grad_norm": 0.892127513885498,
+      "learning_rate": 3.812587185810665e-06,
+      "loss": 0.616,
+      "step": 5605
+    },
+    {
+      "epoch": 1.9683988764044944,
+      "grad_norm": 0.8674480319023132,
+      "learning_rate": 3.8121936872502523e-06,
+      "loss": 0.6111,
+      "step": 5606
+    },
+    {
+      "epoch": 1.96875,
+      "grad_norm": 0.8589068651199341,
+      "learning_rate": 3.8118001438139384e-06,
+      "loss": 0.5768,
+      "step": 5607
+    },
+    {
+      "epoch": 1.9691011235955056,
+      "grad_norm": 0.8877671957015991,
+      "learning_rate": 3.8114065555151826e-06,
+      "loss": 0.6497,
+      "step": 5608
+    },
+    {
+      "epoch": 1.9694522471910112,
+      "grad_norm": 0.8697142004966736,
+      "learning_rate": 3.8110129223674443e-06,
+      "loss": 0.6118,
+      "step": 5609
+    },
+    {
+      "epoch": 1.9698033707865168,
+      "grad_norm": 0.8770889043807983,
+      "learning_rate": 3.810619244384186e-06,
+      "loss": 0.6308,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9701544943820224,
+      "grad_norm": 0.8840969204902649,
+      "learning_rate": 3.8102255215788713e-06,
+      "loss": 0.5886,
+      "step": 5611
+    },
+    {
+      "epoch": 1.970505617977528,
+      "grad_norm": 0.8892825841903687,
+      "learning_rate": 3.8098317539649656e-06,
+      "loss": 0.6546,
+      "step": 5612
+    },
+    {
+      "epoch": 1.9708567415730336,
+      "grad_norm": 0.8668626546859741,
+      "learning_rate": 3.8094379415559333e-06,
+      "loss": 0.6218,
+      "step": 5613
+    },
+    {
+      "epoch": 1.9712078651685392,
+      "grad_norm": 0.8517277836799622,
+      "learning_rate": 3.8090440843652454e-06,
+      "loss": 0.6075,
+      "step": 5614
+    },
+    {
+      "epoch": 1.9715589887640448,
+      "grad_norm": 0.9507593512535095,
+      "learning_rate": 3.8086501824063687e-06,
+      "loss": 0.6418,
+      "step": 5615
+    },
+    {
+      "epoch": 1.9719101123595506,
+      "grad_norm": 0.8373234868049622,
+      "learning_rate": 3.808256235692776e-06,
+      "loss": 0.6001,
+      "step": 5616
+    },
+    {
+      "epoch": 1.9722612359550562,
+      "grad_norm": 0.8520222306251526,
+      "learning_rate": 3.80786224423794e-06,
+      "loss": 0.5982,
+      "step": 5617
+    },
+    {
+      "epoch": 1.9726123595505618,
+      "grad_norm": 0.873327374458313,
+      "learning_rate": 3.807468208055334e-06,
+      "loss": 0.6114,
+      "step": 5618
+    },
+    {
+      "epoch": 1.9729634831460674,
+      "grad_norm": 0.8597673773765564,
+      "learning_rate": 3.807074127158434e-06,
+      "loss": 0.5953,
+      "step": 5619
+    },
+    {
+      "epoch": 1.973314606741573,
+      "grad_norm": 0.8622745275497437,
+      "learning_rate": 3.8066800015607173e-06,
+      "loss": 0.6354,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9736657303370788,
+      "grad_norm": 0.868476390838623,
+      "learning_rate": 3.8062858312756625e-06,
+      "loss": 0.6406,
+      "step": 5621
+    },
+    {
+      "epoch": 1.9740168539325844,
+      "grad_norm": 0.8725282549858093,
+      "learning_rate": 3.8058916163167504e-06,
+      "loss": 0.6252,
+      "step": 5622
+    },
+    {
+      "epoch": 1.97436797752809,
+      "grad_norm": 0.8920749425888062,
+      "learning_rate": 3.805497356697462e-06,
+      "loss": 0.6377,
+      "step": 5623
+    },
+    {
+      "epoch": 1.9747191011235956,
+      "grad_norm": 0.8895111083984375,
+      "learning_rate": 3.805103052431282e-06,
+      "loss": 0.5911,
+      "step": 5624
+    },
+    {
+      "epoch": 1.9750702247191012,
+      "grad_norm": 0.868254542350769,
+      "learning_rate": 3.804708703531694e-06,
+      "loss": 0.6213,
+      "step": 5625
+    },
+    {
+      "epoch": 1.9754213483146068,
+      "grad_norm": 0.8568325638771057,
+      "learning_rate": 3.8043143100121847e-06,
+      "loss": 0.6014,
+      "step": 5626
+    },
+    {
+      "epoch": 1.9757724719101124,
+      "grad_norm": 0.8779681921005249,
+      "learning_rate": 3.8039198718862424e-06,
+      "loss": 0.6314,
+      "step": 5627
+    },
+    {
+      "epoch": 1.976123595505618,
+      "grad_norm": 0.8798127174377441,
+      "learning_rate": 3.803525389167356e-06,
+      "loss": 0.6109,
+      "step": 5628
+    },
+    {
+      "epoch": 1.9764747191011236,
+      "grad_norm": 0.8850237131118774,
+      "learning_rate": 3.803130861869017e-06,
+      "loss": 0.6351,
+      "step": 5629
+    },
+    {
+      "epoch": 1.9768258426966292,
+      "grad_norm": 0.8631384372711182,
+      "learning_rate": 3.8027362900047177e-06,
+      "loss": 0.63,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9771769662921348,
+      "grad_norm": 0.8860955834388733,
+      "learning_rate": 3.8023416735879513e-06,
+      "loss": 0.6042,
+      "step": 5631
+    },
+    {
+      "epoch": 1.9775280898876404,
+      "grad_norm": 0.8704652190208435,
+      "learning_rate": 3.8019470126322146e-06,
+      "loss": 0.6089,
+      "step": 5632
+    },
+    {
+      "epoch": 1.977879213483146,
+      "grad_norm": 0.8756489157676697,
+      "learning_rate": 3.801552307151004e-06,
+      "loss": 0.6121,
+      "step": 5633
+    },
+    {
+      "epoch": 1.9782303370786516,
+      "grad_norm": 0.8789172768592834,
+      "learning_rate": 3.8011575571578186e-06,
+      "loss": 0.5878,
+      "step": 5634
+    },
+    {
+      "epoch": 1.9785814606741572,
+      "grad_norm": 0.8992910385131836,
+      "learning_rate": 3.800762762666158e-06,
+      "loss": 0.6156,
+      "step": 5635
+    },
+    {
+      "epoch": 1.9789325842696628,
+      "grad_norm": 0.8658064603805542,
+      "learning_rate": 3.800367923689524e-06,
+      "loss": 0.6162,
+      "step": 5636
+    },
+    {
+      "epoch": 1.9792837078651684,
+      "grad_norm": 0.861645519733429,
+      "learning_rate": 3.7999730402414193e-06,
+      "loss": 0.572,
+      "step": 5637
+    },
+    {
+      "epoch": 1.9796348314606742,
+      "grad_norm": 0.8935385346412659,
+      "learning_rate": 3.79957811233535e-06,
+      "loss": 0.6163,
+      "step": 5638
+    },
+    {
+      "epoch": 1.9799859550561798,
+      "grad_norm": 0.8381475210189819,
+      "learning_rate": 3.799183139984821e-06,
+      "loss": 0.5775,
+      "step": 5639
+    },
+    {
+      "epoch": 1.9803370786516854,
+      "grad_norm": 0.8763298988342285,
+      "learning_rate": 3.7987881232033397e-06,
+      "loss": 0.6381,
+      "step": 5640
+    },
+    {
+      "epoch": 1.980688202247191,
+      "grad_norm": 0.8951460123062134,
+      "learning_rate": 3.7983930620044164e-06,
+      "loss": 0.6457,
+      "step": 5641
+    },
+    {
+      "epoch": 1.9810393258426966,
+      "grad_norm": 0.8614299893379211,
+      "learning_rate": 3.797997956401561e-06,
+      "loss": 0.5989,
+      "step": 5642
+    },
+    {
+      "epoch": 1.9813904494382022,
+      "grad_norm": 0.8521272540092468,
+      "learning_rate": 3.797602806408287e-06,
+      "loss": 0.6016,
+      "step": 5643
+    },
+    {
+      "epoch": 1.981741573033708,
+      "grad_norm": 0.8658344745635986,
+      "learning_rate": 3.7972076120381075e-06,
+      "loss": 0.6222,
+      "step": 5644
+    },
+    {
+      "epoch": 1.9820926966292136,
+      "grad_norm": 0.8566646575927734,
+      "learning_rate": 3.796812373304537e-06,
+      "loss": 0.6105,
+      "step": 5645
+    },
+    {
+      "epoch": 1.9824438202247192,
+      "grad_norm": 0.8909837007522583,
+      "learning_rate": 3.7964170902210927e-06,
+      "loss": 0.6047,
+      "step": 5646
+    },
+    {
+      "epoch": 1.9827949438202248,
+      "grad_norm": 0.8529945611953735,
+      "learning_rate": 3.796021762801294e-06,
+      "loss": 0.6196,
+      "step": 5647
+    },
+    {
+      "epoch": 1.9831460674157304,
+      "grad_norm": 0.878425657749176,
+      "learning_rate": 3.79562639105866e-06,
+      "loss": 0.5868,
+      "step": 5648
+    },
+    {
+      "epoch": 1.983497191011236,
+      "grad_norm": 0.8768588900566101,
+      "learning_rate": 3.795230975006712e-06,
+      "loss": 0.5811,
+      "step": 5649
+    },
+    {
+      "epoch": 1.9838483146067416,
+      "grad_norm": 0.8540440201759338,
+      "learning_rate": 3.794835514658972e-06,
+      "loss": 0.6095,
+      "step": 5650
+    },
+    {
+      "epoch": 1.9841994382022472,
+      "grad_norm": 0.8583154082298279,
+      "learning_rate": 3.794440010028967e-06,
+      "loss": 0.5874,
+      "step": 5651
+    },
+    {
+      "epoch": 1.9845505617977528,
+      "grad_norm": 0.8400346040725708,
+      "learning_rate": 3.7940444611302206e-06,
+      "loss": 0.5943,
+      "step": 5652
+    },
+    {
+      "epoch": 1.9849016853932584,
+      "grad_norm": 0.8862133026123047,
+      "learning_rate": 3.7936488679762617e-06,
+      "loss": 0.5951,
+      "step": 5653
+    },
+    {
+      "epoch": 1.985252808988764,
+      "grad_norm": 0.9019565582275391,
+      "learning_rate": 3.7932532305806175e-06,
+      "loss": 0.6084,
+      "step": 5654
+    },
+    {
+      "epoch": 1.9856039325842696,
+      "grad_norm": 0.8929518461227417,
+      "learning_rate": 3.7928575489568197e-06,
+      "loss": 0.589,
+      "step": 5655
+    },
+    {
+      "epoch": 1.9859550561797752,
+      "grad_norm": 0.8899116516113281,
+      "learning_rate": 3.7924618231184e-06,
+      "loss": 0.6176,
+      "step": 5656
+    },
+    {
+      "epoch": 1.9863061797752808,
+      "grad_norm": 0.8458961844444275,
+      "learning_rate": 3.7920660530788923e-06,
+      "loss": 0.5669,
+      "step": 5657
+    },
+    {
+      "epoch": 1.9866573033707864,
+      "grad_norm": 0.855290412902832,
+      "learning_rate": 3.7916702388518313e-06,
+      "loss": 0.5848,
+      "step": 5658
+    },
+    {
+      "epoch": 1.987008426966292,
+      "grad_norm": 0.8671900629997253,
+      "learning_rate": 3.7912743804507522e-06,
+      "loss": 0.5838,
+      "step": 5659
+    },
+    {
+      "epoch": 1.9873595505617978,
+      "grad_norm": 0.8991999626159668,
+      "learning_rate": 3.7908784778891954e-06,
+      "loss": 0.6014,
+      "step": 5660
+    },
+    {
+      "epoch": 1.9877106741573034,
+      "grad_norm": 0.8997818827629089,
+      "learning_rate": 3.7904825311807e-06,
+      "loss": 0.6297,
+      "step": 5661
+    },
+    {
+      "epoch": 1.988061797752809,
+      "grad_norm": 0.8679925203323364,
+      "learning_rate": 3.7900865403388054e-06,
+      "loss": 0.6217,
+      "step": 5662
+    },
+    {
+      "epoch": 1.9884129213483146,
+      "grad_norm": 0.8854179382324219,
+      "learning_rate": 3.789690505377055e-06,
+      "loss": 0.592,
+      "step": 5663
+    },
+    {
+      "epoch": 1.9887640449438202,
+      "grad_norm": 0.8907507658004761,
+      "learning_rate": 3.7892944263089927e-06,
+      "loss": 0.6142,
+      "step": 5664
+    },
+    {
+      "epoch": 1.9891151685393258,
+      "grad_norm": 1.1212937831878662,
+      "learning_rate": 3.788898303148164e-06,
+      "loss": 0.6249,
+      "step": 5665
+    },
+    {
+      "epoch": 1.9894662921348316,
+      "grad_norm": 0.8652292490005493,
+      "learning_rate": 3.788502135908117e-06,
+      "loss": 0.6164,
+      "step": 5666
+    },
+    {
+      "epoch": 1.9898174157303372,
+      "grad_norm": 0.8923054337501526,
+      "learning_rate": 3.7881059246023993e-06,
+      "loss": 0.6409,
+      "step": 5667
+    },
+    {
+      "epoch": 1.9901685393258428,
+      "grad_norm": 0.8936302065849304,
+      "learning_rate": 3.7877096692445613e-06,
+      "loss": 0.6208,
+      "step": 5668
+    },
+    {
+      "epoch": 1.9905196629213484,
+      "grad_norm": 0.8705040216445923,
+      "learning_rate": 3.7873133698481547e-06,
+      "loss": 0.6326,
+      "step": 5669
+    },
+    {
+      "epoch": 1.990870786516854,
+      "grad_norm": 0.8782395124435425,
+      "learning_rate": 3.7869170264267325e-06,
+      "loss": 0.6189,
+      "step": 5670
+    },
+    {
+      "epoch": 1.9912219101123596,
+      "grad_norm": 0.8925029635429382,
+      "learning_rate": 3.786520638993849e-06,
+      "loss": 0.6151,
+      "step": 5671
+    },
+    {
+      "epoch": 1.9915730337078652,
+      "grad_norm": 0.8680460453033447,
+      "learning_rate": 3.78612420756306e-06,
+      "loss": 0.6144,
+      "step": 5672
+    },
+    {
+      "epoch": 1.9919241573033708,
+      "grad_norm": 0.859771728515625,
+      "learning_rate": 3.785727732147925e-06,
+      "loss": 0.623,
+      "step": 5673
+    },
+    {
+      "epoch": 1.9922752808988764,
+      "grad_norm": 0.8385575413703918,
+      "learning_rate": 3.7853312127620003e-06,
+      "loss": 0.5991,
+      "step": 5674
+    },
+    {
+      "epoch": 1.992626404494382,
+      "grad_norm": 0.8689984083175659,
+      "learning_rate": 3.7849346494188487e-06,
+      "loss": 0.5658,
+      "step": 5675
+    },
+    {
+      "epoch": 1.9929775280898876,
+      "grad_norm": 0.8608708381652832,
+      "learning_rate": 3.784538042132031e-06,
+      "loss": 0.6215,
+      "step": 5676
+    },
+    {
+      "epoch": 1.9933286516853932,
+      "grad_norm": 0.884060800075531,
+      "learning_rate": 3.7841413909151127e-06,
+      "loss": 0.6248,
+      "step": 5677
+    },
+    {
+      "epoch": 1.9936797752808988,
+      "grad_norm": 0.86598801612854,
+      "learning_rate": 3.783744695781657e-06,
+      "loss": 0.5958,
+      "step": 5678
+    },
+    {
+      "epoch": 1.9940308988764044,
+      "grad_norm": 0.8934966325759888,
+      "learning_rate": 3.7833479567452308e-06,
+      "loss": 0.6226,
+      "step": 5679
+    },
+    {
+      "epoch": 1.99438202247191,
+      "grad_norm": 0.8695813417434692,
+      "learning_rate": 3.782951173819403e-06,
+      "loss": 0.59,
+      "step": 5680
+    },
+    {
+      "epoch": 1.9947331460674156,
+      "grad_norm": 0.9348827004432678,
+      "learning_rate": 3.782554347017743e-06,
+      "loss": 0.6402,
+      "step": 5681
+    },
+    {
+      "epoch": 1.9950842696629212,
+      "grad_norm": 0.8460621237754822,
+      "learning_rate": 3.7821574763538216e-06,
+      "loss": 0.5891,
+      "step": 5682
+    },
+    {
+      "epoch": 1.995435393258427,
+      "grad_norm": 0.8422776460647583,
+      "learning_rate": 3.7817605618412125e-06,
+      "loss": 0.6031,
+      "step": 5683
+    },
+    {
+      "epoch": 1.9957865168539326,
+      "grad_norm": 0.87843257188797,
+      "learning_rate": 3.7813636034934875e-06,
+      "loss": 0.6175,
+      "step": 5684
+    },
+    {
+      "epoch": 1.9961376404494382,
+      "grad_norm": 0.9025244116783142,
+      "learning_rate": 3.7809666013242245e-06,
+      "loss": 0.6066,
+      "step": 5685
+    },
+    {
+      "epoch": 1.9964887640449438,
+      "grad_norm": 0.8787137866020203,
+      "learning_rate": 3.7805695553470002e-06,
+      "loss": 0.62,
+      "step": 5686
+    },
+    {
+      "epoch": 1.9968398876404494,
+      "grad_norm": 0.8768424391746521,
+      "learning_rate": 3.7801724655753922e-06,
+      "loss": 0.6046,
+      "step": 5687
+    },
+    {
+      "epoch": 1.9971910112359552,
+      "grad_norm": 0.8371465802192688,
+      "learning_rate": 3.7797753320229822e-06,
+      "loss": 0.582,
+      "step": 5688
+    },
+    {
+      "epoch": 1.9975421348314608,
+      "grad_norm": 0.8751241564750671,
+      "learning_rate": 3.7793781547033503e-06,
+      "loss": 0.597,
+      "step": 5689
+    },
+    {
+      "epoch": 1.9978932584269664,
+      "grad_norm": 0.8778484463691711,
+      "learning_rate": 3.778980933630081e-06,
+      "loss": 0.6299,
+      "step": 5690
+    },
+    {
+      "epoch": 1.998244382022472,
+      "grad_norm": 0.8325814008712769,
+      "learning_rate": 3.7785836688167576e-06,
+      "loss": 0.5842,
+      "step": 5691
+    },
+    {
+      "epoch": 1.9985955056179776,
+      "grad_norm": 0.8934423923492432,
+      "learning_rate": 3.778186360276967e-06,
+      "loss": 0.6192,
+      "step": 5692
+    },
+    {
+      "epoch": 1.9989466292134832,
+      "grad_norm": 0.8526570796966553,
+      "learning_rate": 3.7777890080242963e-06,
+      "loss": 0.5843,
+      "step": 5693
+    },
+    {
+      "epoch": 1.9992977528089888,
+      "grad_norm": 0.896981418132782,
+      "learning_rate": 3.7773916120723354e-06,
+      "loss": 0.651,
+      "step": 5694
+    },
+    {
+      "epoch": 1.9996488764044944,
+      "grad_norm": 0.8761165738105774,
+      "learning_rate": 3.776994172434675e-06,
+      "loss": 0.649,
+      "step": 5695
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8966251611709595,
+      "learning_rate": 3.776596689124905e-06,
+      "loss": 0.5836,
+      "step": 5696
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 17088,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 2848,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4388305024746783e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}