diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10486 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998743455497382,
+  "eval_steps": 500,
+  "global_step": 1492,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006701570680628272,
+      "grad_norm": 4.200109958648682,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 8.7721,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013403141361256544,
+      "grad_norm": 4.861227035522461,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 9.0224,
+      "step": 2
+    },
+    {
+      "epoch": 0.002010471204188482,
+      "grad_norm": 6.782981872558594,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 9.2282,
+      "step": 3
+    },
+    {
+      "epoch": 0.0026806282722513087,
+      "grad_norm": 7.665585994720459,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 8.9244,
+      "step": 4
+    },
+    {
+      "epoch": 0.003350785340314136,
+      "grad_norm": 5.701632499694824,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 8.7144,
+      "step": 5
+    },
+    {
+      "epoch": 0.004020942408376964,
+      "grad_norm": 4.979744911193848,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 8.8509,
+      "step": 6
+    },
+    {
+      "epoch": 0.00469109947643979,
+      "grad_norm": 5.203133583068848,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 8.0535,
+      "step": 7
+    },
+    {
+      "epoch": 0.0053612565445026175,
+      "grad_norm": 4.624353408813477,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 8.5312,
+      "step": 8
+    },
+    {
+      "epoch": 0.006031413612565445,
+      "grad_norm": 3.5958287715911865,
+      "learning_rate": 3e-06,
+      "loss": 8.3258,
+      "step": 9
+    },
+    {
+      "epoch": 0.006701570680628272,
+      "grad_norm": 3.4215784072875977,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 8.416,
+      "step": 10
+    },
+    {
+      "epoch": 0.007371727748691099,
+      "grad_norm": 3.484177589416504,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 8.2023,
+      "step": 11
+    },
+    {
+      "epoch": 0.008041884816753928,
+      "grad_norm": 3.484971046447754,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 7.813,
+      "step": 12
+    },
+    {
+      "epoch": 0.008712041884816753,
+      "grad_norm": 2.726947069168091,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 8.134,
+      "step": 13
+    },
+    {
+      "epoch": 0.00938219895287958,
+      "grad_norm": 3.3426618576049805,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 7.7547,
+      "step": 14
+    },
+    {
+      "epoch": 0.010052356020942408,
+      "grad_norm": 3.043964385986328,
+      "learning_rate": 5e-06,
+      "loss": 7.7399,
+      "step": 15
+    },
+    {
+      "epoch": 0.010722513089005235,
+      "grad_norm": 3.101304531097412,
+      "learning_rate": 4.999994344791726e-06,
+      "loss": 7.1639,
+      "step": 16
+    },
+    {
+      "epoch": 0.011392670157068062,
+      "grad_norm": 3.683811902999878,
+      "learning_rate": 4.999977379192491e-06,
+      "loss": 7.595,
+      "step": 17
+    },
+    {
+      "epoch": 0.01206282722513089,
+      "grad_norm": 3.79826021194458,
+      "learning_rate": 4.9999491032790475e-06,
+      "loss": 7.5361,
+      "step": 18
+    },
+    {
+      "epoch": 0.012732984293193717,
+      "grad_norm": 3.463408946990967,
+      "learning_rate": 4.999909517179322e-06,
+      "loss": 7.1776,
+      "step": 19
+    },
+    {
+      "epoch": 0.013403141361256544,
+      "grad_norm": 3.486097574234009,
+      "learning_rate": 4.9998586210724085e-06,
+      "loss": 7.0846,
+      "step": 20
+    },
+    {
+      "epoch": 0.014073298429319371,
+      "grad_norm": 3.8364169597625732,
+      "learning_rate": 4.99979641518857e-06,
+      "loss": 7.3558,
+      "step": 21
+    },
+    {
+      "epoch": 0.014743455497382199,
+      "grad_norm": 2.9487595558166504,
+      "learning_rate": 4.999722899809235e-06,
+      "loss": 6.9944,
+      "step": 22
+    },
+    {
+      "epoch": 0.015413612565445026,
+      "grad_norm": 2.877228260040283,
+      "learning_rate": 4.9996380752670006e-06,
+      "loss": 6.8514,
+      "step": 23
+    },
+    {
+      "epoch": 0.016083769633507855,
+      "grad_norm": 2.7856905460357666,
+      "learning_rate": 4.999541941945626e-06,
+      "loss": 6.9527,
+      "step": 24
+    },
+    {
+      "epoch": 0.016753926701570682,
+      "grad_norm": 2.6983463764190674,
+      "learning_rate": 4.999434500280035e-06,
+      "loss": 6.8831,
+      "step": 25
+    },
+    {
+      "epoch": 0.017424083769633506,
+      "grad_norm": 3.0367815494537354,
+      "learning_rate": 4.999315750756312e-06,
+      "loss": 6.6457,
+      "step": 26
+    },
+    {
+      "epoch": 0.018094240837696333,
+      "grad_norm": 2.566683292388916,
+      "learning_rate": 4.999185693911699e-06,
+      "loss": 6.8638,
+      "step": 27
+    },
+    {
+      "epoch": 0.01876439790575916,
+      "grad_norm": 3.147542953491211,
+      "learning_rate": 4.9990443303345945e-06,
+      "loss": 6.5231,
+      "step": 28
+    },
+    {
+      "epoch": 0.019434554973821988,
+      "grad_norm": 2.622145175933838,
+      "learning_rate": 4.998891660664552e-06,
+      "loss": 6.8036,
+      "step": 29
+    },
+    {
+      "epoch": 0.020104712041884815,
+      "grad_norm": 2.8038251399993896,
+      "learning_rate": 4.998727685592273e-06,
+      "loss": 6.6147,
+      "step": 30
+    },
+    {
+      "epoch": 0.020774869109947643,
+      "grad_norm": 3.022918939590454,
+      "learning_rate": 4.9985524058596105e-06,
+      "loss": 6.6855,
+      "step": 31
+    },
+    {
+      "epoch": 0.02144502617801047,
+      "grad_norm": 2.2576181888580322,
+      "learning_rate": 4.998365822259556e-06,
+      "loss": 6.5477,
+      "step": 32
+    },
+    {
+      "epoch": 0.022115183246073297,
+      "grad_norm": 2.9812848567962646,
+      "learning_rate": 4.998167935636247e-06,
+      "loss": 6.4677,
+      "step": 33
+    },
+    {
+      "epoch": 0.022785340314136125,
+      "grad_norm": 2.937847375869751,
+      "learning_rate": 4.997958746884956e-06,
+      "loss": 6.4969,
+      "step": 34
+    },
+    {
+      "epoch": 0.023455497382198952,
+      "grad_norm": 2.3638932704925537,
+      "learning_rate": 4.997738256952086e-06,
+      "loss": 6.6306,
+      "step": 35
+    },
+    {
+      "epoch": 0.02412565445026178,
+      "grad_norm": 2.4940426349639893,
+      "learning_rate": 4.997506466835171e-06,
+      "loss": 6.6289,
+      "step": 36
+    },
+    {
+      "epoch": 0.024795811518324606,
+      "grad_norm": 2.751082420349121,
+      "learning_rate": 4.997263377582868e-06,
+      "loss": 6.4163,
+      "step": 37
+    },
+    {
+      "epoch": 0.025465968586387434,
+      "grad_norm": 2.156564474105835,
+      "learning_rate": 4.997008990294953e-06,
+      "loss": 6.5205,
+      "step": 38
+    },
+    {
+      "epoch": 0.02613612565445026,
+      "grad_norm": 2.9955501556396484,
+      "learning_rate": 4.996743306122317e-06,
+      "loss": 6.4886,
+      "step": 39
+    },
+    {
+      "epoch": 0.02680628272251309,
+      "grad_norm": 2.5269675254821777,
+      "learning_rate": 4.99646632626696e-06,
+      "loss": 6.5692,
+      "step": 40
+    },
+    {
+      "epoch": 0.027476439790575916,
+      "grad_norm": 2.1620020866394043,
+      "learning_rate": 4.996178051981984e-06,
+      "loss": 6.5636,
+      "step": 41
+    },
+    {
+      "epoch": 0.028146596858638743,
+      "grad_norm": 2.7254207134246826,
+      "learning_rate": 4.99587848457159e-06,
+      "loss": 6.1583,
+      "step": 42
+    },
+    {
+      "epoch": 0.02881675392670157,
+      "grad_norm": 2.702894926071167,
+      "learning_rate": 4.9955676253910716e-06,
+      "loss": 6.4469,
+      "step": 43
+    },
+    {
+      "epoch": 0.029486910994764397,
+      "grad_norm": 2.119009494781494,
+      "learning_rate": 4.995245475846806e-06,
+      "loss": 6.3458,
+      "step": 44
+    },
+    {
+      "epoch": 0.030157068062827225,
+      "grad_norm": 2.6456499099731445,
+      "learning_rate": 4.994912037396254e-06,
+      "loss": 6.8394,
+      "step": 45
+    },
+    {
+      "epoch": 0.030827225130890052,
+      "grad_norm": 2.4168996810913086,
+      "learning_rate": 4.994567311547944e-06,
+      "loss": 6.5947,
+      "step": 46
+    },
+    {
+      "epoch": 0.03149738219895288,
+      "grad_norm": 2.0806760787963867,
+      "learning_rate": 4.994211299861475e-06,
+      "loss": 6.1653,
+      "step": 47
+    },
+    {
+      "epoch": 0.03216753926701571,
+      "grad_norm": 2.130005121231079,
+      "learning_rate": 4.993844003947501e-06,
+      "loss": 6.4416,
+      "step": 48
+    },
+    {
+      "epoch": 0.032837696335078534,
+      "grad_norm": 2.2185635566711426,
+      "learning_rate": 4.993465425467732e-06,
+      "loss": 6.3064,
+      "step": 49
+    },
+    {
+      "epoch": 0.033507853403141365,
+      "grad_norm": 2.5894346237182617,
+      "learning_rate": 4.9930755661349215e-06,
+      "loss": 6.5322,
+      "step": 50
+    },
+    {
+      "epoch": 0.03417801047120419,
+      "grad_norm": 1.9817583560943604,
+      "learning_rate": 4.992674427712855e-06,
+      "loss": 6.0912,
+      "step": 51
+    },
+    {
+      "epoch": 0.03484816753926701,
+      "grad_norm": 2.4749717712402344,
+      "learning_rate": 4.99226201201635e-06,
+      "loss": 6.2575,
+      "step": 52
+    },
+    {
+      "epoch": 0.03551832460732984,
+      "grad_norm": 3.163346767425537,
+      "learning_rate": 4.991838320911245e-06,
+      "loss": 6.3432,
+      "step": 53
+    },
+    {
+      "epoch": 0.03618848167539267,
+      "grad_norm": 2.872176170349121,
+      "learning_rate": 4.991403356314388e-06,
+      "loss": 6.181,
+      "step": 54
+    },
+    {
+      "epoch": 0.0368586387434555,
+      "grad_norm": 2.5608084201812744,
+      "learning_rate": 4.990957120193632e-06,
+      "loss": 6.2428,
+      "step": 55
+    },
+    {
+      "epoch": 0.03752879581151832,
+      "grad_norm": 2.9240713119506836,
+      "learning_rate": 4.990499614567825e-06,
+      "loss": 6.1696,
+      "step": 56
+    },
+    {
+      "epoch": 0.03819895287958115,
+      "grad_norm": 2.4163713455200195,
+      "learning_rate": 4.990030841506797e-06,
+      "loss": 6.1887,
+      "step": 57
+    },
+    {
+      "epoch": 0.038869109947643976,
+      "grad_norm": 2.1572868824005127,
+      "learning_rate": 4.9895508031313545e-06,
+      "loss": 6.4599,
+      "step": 58
+    },
+    {
+      "epoch": 0.03953926701570681,
+      "grad_norm": 2.173875570297241,
+      "learning_rate": 4.989059501613273e-06,
+      "loss": 5.9834,
+      "step": 59
+    },
+    {
+      "epoch": 0.04020942408376963,
+      "grad_norm": 2.419245719909668,
+      "learning_rate": 4.9885569391752805e-06,
+      "loss": 6.1352,
+      "step": 60
+    },
+    {
+      "epoch": 0.04087958115183246,
+      "grad_norm": 2.2928881645202637,
+      "learning_rate": 4.988043118091056e-06,
+      "loss": 5.7896,
+      "step": 61
+    },
+    {
+      "epoch": 0.041549738219895285,
+      "grad_norm": 1.8945914506912231,
+      "learning_rate": 4.9875180406852085e-06,
+      "loss": 6.0007,
+      "step": 62
+    },
+    {
+      "epoch": 0.042219895287958116,
+      "grad_norm": 2.0440707206726074,
+      "learning_rate": 4.986981709333278e-06,
+      "loss": 5.9541,
+      "step": 63
+    },
+    {
+      "epoch": 0.04289005235602094,
+      "grad_norm": 2.1350975036621094,
+      "learning_rate": 4.9864341264617165e-06,
+      "loss": 6.2672,
+      "step": 64
+    },
+    {
+      "epoch": 0.04356020942408377,
+      "grad_norm": 2.444735527038574,
+      "learning_rate": 4.98587529454788e-06,
+      "loss": 6.1391,
+      "step": 65
+    },
+    {
+      "epoch": 0.044230366492146594,
+      "grad_norm": 2.0678515434265137,
+      "learning_rate": 4.985305216120017e-06,
+      "loss": 6.136,
+      "step": 66
+    },
+    {
+      "epoch": 0.044900523560209425,
+      "grad_norm": 2.892256021499634,
+      "learning_rate": 4.984723893757256e-06,
+      "loss": 6.0538,
+      "step": 67
+    },
+    {
+      "epoch": 0.04557068062827225,
+      "grad_norm": 2.0744709968566895,
+      "learning_rate": 4.9841313300895985e-06,
+      "loss": 6.425,
+      "step": 68
+    },
+    {
+      "epoch": 0.04624083769633508,
+      "grad_norm": 2.185494899749756,
+      "learning_rate": 4.9835275277979e-06,
+      "loss": 6.1453,
+      "step": 69
+    },
+    {
+      "epoch": 0.046910994764397904,
+      "grad_norm": 2.4761366844177246,
+      "learning_rate": 4.9829124896138635e-06,
+      "loss": 6.1714,
+      "step": 70
+    },
+    {
+      "epoch": 0.047581151832460734,
+      "grad_norm": 2.2010130882263184,
+      "learning_rate": 4.982286218320023e-06,
+      "loss": 6.3217,
+      "step": 71
+    },
+    {
+      "epoch": 0.04825130890052356,
+      "grad_norm": 2.60612154006958,
+      "learning_rate": 4.981648716749735e-06,
+      "loss": 6.2632,
+      "step": 72
+    },
+    {
+      "epoch": 0.04892146596858639,
+      "grad_norm": 2.6755013465881348,
+      "learning_rate": 4.980999987787163e-06,
+      "loss": 5.8599,
+      "step": 73
+    },
+    {
+      "epoch": 0.04959162303664921,
+      "grad_norm": 2.565596103668213,
+      "learning_rate": 4.980340034367264e-06,
+      "loss": 5.9165,
+      "step": 74
+    },
+    {
+      "epoch": 0.050261780104712044,
+      "grad_norm": 1.9897733926773071,
+      "learning_rate": 4.979668859475779e-06,
+      "loss": 6.2114,
+      "step": 75
+    },
+    {
+      "epoch": 0.05093193717277487,
+      "grad_norm": 2.3420677185058594,
+      "learning_rate": 4.978986466149212e-06,
+      "loss": 6.0591,
+      "step": 76
+    },
+    {
+      "epoch": 0.0516020942408377,
+      "grad_norm": 2.258755683898926,
+      "learning_rate": 4.978292857474828e-06,
+      "loss": 6.1753,
+      "step": 77
+    },
+    {
+      "epoch": 0.05227225130890052,
+      "grad_norm": 2.6298441886901855,
+      "learning_rate": 4.977588036590625e-06,
+      "loss": 6.3313,
+      "step": 78
+    },
+    {
+      "epoch": 0.05294240837696335,
+      "grad_norm": 2.3326869010925293,
+      "learning_rate": 4.976872006685332e-06,
+      "loss": 5.9557,
+      "step": 79
+    },
+    {
+      "epoch": 0.05361256544502618,
+      "grad_norm": 2.4547061920166016,
+      "learning_rate": 4.976144770998386e-06,
+      "loss": 5.9302,
+      "step": 80
+    },
+    {
+      "epoch": 0.05428272251308901,
+      "grad_norm": 2.0991933345794678,
+      "learning_rate": 4.975406332819923e-06,
+      "loss": 5.817,
+      "step": 81
+    },
+    {
+      "epoch": 0.05495287958115183,
+      "grad_norm": 3.2543742656707764,
+      "learning_rate": 4.974656695490762e-06,
+      "loss": 6.128,
+      "step": 82
+    },
+    {
+      "epoch": 0.05562303664921466,
+      "grad_norm": 2.2632484436035156,
+      "learning_rate": 4.973895862402385e-06,
+      "loss": 6.0342,
+      "step": 83
+    },
+    {
+      "epoch": 0.056293193717277486,
+      "grad_norm": 2.243244171142578,
+      "learning_rate": 4.973123836996929e-06,
+      "loss": 6.0889,
+      "step": 84
+    },
+    {
+      "epoch": 0.056963350785340316,
+      "grad_norm": 2.0775105953216553,
+      "learning_rate": 4.9723406227671645e-06,
+      "loss": 5.9834,
+      "step": 85
+    },
+    {
+      "epoch": 0.05763350785340314,
+      "grad_norm": 2.1865806579589844,
+      "learning_rate": 4.971546223256484e-06,
+      "loss": 5.8297,
+      "step": 86
+    },
+    {
+      "epoch": 0.05830366492146597,
+      "grad_norm": 2.533968925476074,
+      "learning_rate": 4.9707406420588836e-06,
+      "loss": 6.1779,
+      "step": 87
+    },
+    {
+      "epoch": 0.058973821989528795,
+      "grad_norm": 2.2566957473754883,
+      "learning_rate": 4.969923882818946e-06,
+      "loss": 5.8309,
+      "step": 88
+    },
+    {
+      "epoch": 0.059643979057591626,
+      "grad_norm": 3.06984806060791,
+      "learning_rate": 4.969095949231826e-06,
+      "loss": 5.7423,
+      "step": 89
+    },
+    {
+      "epoch": 0.06031413612565445,
+      "grad_norm": 2.543743848800659,
+      "learning_rate": 4.968256845043233e-06,
+      "loss": 6.4709,
+      "step": 90
+    },
+    {
+      "epoch": 0.06098429319371728,
+      "grad_norm": 2.6973841190338135,
+      "learning_rate": 4.967406574049416e-06,
+      "loss": 6.2436,
+      "step": 91
+    },
+    {
+      "epoch": 0.061654450261780104,
+      "grad_norm": 2.71950101852417,
+      "learning_rate": 4.96654514009714e-06,
+      "loss": 5.8146,
+      "step": 92
+    },
+    {
+      "epoch": 0.062324607329842935,
+      "grad_norm": 2.4400877952575684,
+      "learning_rate": 4.965672547083678e-06,
+      "loss": 5.9425,
+      "step": 93
+    },
+    {
+      "epoch": 0.06299476439790576,
+      "grad_norm": 2.360313892364502,
+      "learning_rate": 4.964788798956784e-06,
+      "loss": 5.9114,
+      "step": 94
+    },
+    {
+      "epoch": 0.06366492146596858,
+      "grad_norm": 2.3321006298065186,
+      "learning_rate": 4.963893899714683e-06,
+      "loss": 5.9243,
+      "step": 95
+    },
+    {
+      "epoch": 0.06433507853403142,
+      "grad_norm": 2.638052463531494,
+      "learning_rate": 4.9629878534060494e-06,
+      "loss": 6.1771,
+      "step": 96
+    },
+    {
+      "epoch": 0.06500523560209424,
+      "grad_norm": 2.1472249031066895,
+      "learning_rate": 4.962070664129986e-06,
+      "loss": 5.955,
+      "step": 97
+    },
+    {
+      "epoch": 0.06567539267015707,
+      "grad_norm": 2.5043721199035645,
+      "learning_rate": 4.961142336036011e-06,
+      "loss": 6.346,
+      "step": 98
+    },
+    {
+      "epoch": 0.06634554973821989,
+      "grad_norm": 2.469748020172119,
+      "learning_rate": 4.960202873324033e-06,
+      "loss": 5.9722,
+      "step": 99
+    },
+    {
+      "epoch": 0.06701570680628273,
+      "grad_norm": 2.429661512374878,
+      "learning_rate": 4.959252280244342e-06,
+      "loss": 5.89,
+      "step": 100
+    },
+    {
+      "epoch": 0.06768586387434555,
+      "grad_norm": 2.4955296516418457,
+      "learning_rate": 4.958290561097575e-06,
+      "loss": 5.8828,
+      "step": 101
+    },
+    {
+      "epoch": 0.06835602094240838,
+      "grad_norm": 1.904392123222351,
+      "learning_rate": 4.9573177202347135e-06,
+      "loss": 6.0817,
+      "step": 102
+    },
+    {
+      "epoch": 0.0690261780104712,
+      "grad_norm": 2.5589585304260254,
+      "learning_rate": 4.956333762057048e-06,
+      "loss": 5.8343,
+      "step": 103
+    },
+    {
+      "epoch": 0.06969633507853402,
+      "grad_norm": 2.1008498668670654,
+      "learning_rate": 4.955338691016173e-06,
+      "loss": 5.7119,
+      "step": 104
+    },
+    {
+      "epoch": 0.07036649214659686,
+      "grad_norm": 2.3903372287750244,
+      "learning_rate": 4.954332511613953e-06,
+      "loss": 6.1662,
+      "step": 105
+    },
+    {
+      "epoch": 0.07103664921465969,
+      "grad_norm": 2.2429683208465576,
+      "learning_rate": 4.953315228402512e-06,
+      "loss": 5.9493,
+      "step": 106
+    },
+    {
+      "epoch": 0.07170680628272251,
+      "grad_norm": 2.764125347137451,
+      "learning_rate": 4.952286845984209e-06,
+      "loss": 5.8358,
+      "step": 107
+    },
+    {
+      "epoch": 0.07237696335078533,
+      "grad_norm": 2.5214192867279053,
+      "learning_rate": 4.9512473690116165e-06,
+      "loss": 5.7791,
+      "step": 108
+    },
+    {
+      "epoch": 0.07304712041884817,
+      "grad_norm": 2.370891571044922,
+      "learning_rate": 4.950196802187503e-06,
+      "loss": 5.7436,
+      "step": 109
+    },
+    {
+      "epoch": 0.073717277486911,
+      "grad_norm": 2.2167561054229736,
+      "learning_rate": 4.9491351502648075e-06,
+      "loss": 5.9407,
+      "step": 110
+    },
+    {
+      "epoch": 0.07438743455497382,
+      "grad_norm": 2.805875301361084,
+      "learning_rate": 4.94806241804662e-06,
+      "loss": 6.0697,
+      "step": 111
+    },
+    {
+      "epoch": 0.07505759162303664,
+      "grad_norm": 1.881156325340271,
+      "learning_rate": 4.946978610386159e-06,
+      "loss": 5.7838,
+      "step": 112
+    },
+    {
+      "epoch": 0.07572774869109948,
+      "grad_norm": 2.642446517944336,
+      "learning_rate": 4.9458837321867515e-06,
+      "loss": 6.1066,
+      "step": 113
+    },
+    {
+      "epoch": 0.0763979057591623,
+      "grad_norm": 2.5777878761291504,
+      "learning_rate": 4.944777788401809e-06,
+      "loss": 5.6759,
+      "step": 114
+    },
+    {
+      "epoch": 0.07706806282722513,
+      "grad_norm": 2.3588509559631348,
+      "learning_rate": 4.943660784034806e-06,
+      "loss": 6.0307,
+      "step": 115
+    },
+    {
+      "epoch": 0.07773821989528795,
+      "grad_norm": 2.7860770225524902,
+      "learning_rate": 4.942532724139255e-06,
+      "loss": 5.6816,
+      "step": 116
+    },
+    {
+      "epoch": 0.07840837696335079,
+      "grad_norm": 2.4214789867401123,
+      "learning_rate": 4.941393613818688e-06,
+      "loss": 5.9614,
+      "step": 117
+    },
+    {
+      "epoch": 0.07907853403141361,
+      "grad_norm": 1.9265861511230469,
+      "learning_rate": 4.94024345822663e-06,
+      "loss": 5.4188,
+      "step": 118
+    },
+    {
+      "epoch": 0.07974869109947644,
+      "grad_norm": 2.178675413131714,
+      "learning_rate": 4.939082262566575e-06,
+      "loss": 5.6996,
+      "step": 119
+    },
+    {
+      "epoch": 0.08041884816753926,
+      "grad_norm": 2.9240055084228516,
+      "learning_rate": 4.937910032091968e-06,
+      "loss": 6.0002,
+      "step": 120
+    },
+    {
+      "epoch": 0.0810890052356021,
+      "grad_norm": 3.041872262954712,
+      "learning_rate": 4.936726772106174e-06,
+      "loss": 6.1273,
+      "step": 121
+    },
+    {
+      "epoch": 0.08175916230366492,
+      "grad_norm": 2.833592414855957,
+      "learning_rate": 4.935532487962457e-06,
+      "loss": 5.9378,
+      "step": 122
+    },
+    {
+      "epoch": 0.08242931937172775,
+      "grad_norm": 2.215791702270508,
+      "learning_rate": 4.934327185063959e-06,
+      "loss": 5.8925,
+      "step": 123
+    },
+    {
+      "epoch": 0.08309947643979057,
+      "grad_norm": 2.210946559906006,
+      "learning_rate": 4.933110868863671e-06,
+      "loss": 6.0775,
+      "step": 124
+    },
+    {
+      "epoch": 0.08376963350785341,
+      "grad_norm": 1.9565507173538208,
+      "learning_rate": 4.93188354486441e-06,
+      "loss": 6.0097,
+      "step": 125
+    },
+    {
+      "epoch": 0.08443979057591623,
+      "grad_norm": 2.499535083770752,
+      "learning_rate": 4.9306452186187925e-06,
+      "loss": 5.7272,
+      "step": 126
+    },
+    {
+      "epoch": 0.08510994764397906,
+      "grad_norm": 2.260690927505493,
+      "learning_rate": 4.9293958957292155e-06,
+      "loss": 5.8022,
+      "step": 127
+    },
+    {
+      "epoch": 0.08578010471204188,
+      "grad_norm": 2.7281482219696045,
+      "learning_rate": 4.928135581847823e-06,
+      "loss": 5.6282,
+      "step": 128
+    },
+    {
+      "epoch": 0.08645026178010472,
+      "grad_norm": 2.7198517322540283,
+      "learning_rate": 4.926864282676485e-06,
+      "loss": 5.8988,
+      "step": 129
+    },
+    {
+      "epoch": 0.08712041884816754,
+      "grad_norm": 2.2805449962615967,
+      "learning_rate": 4.9255820039667695e-06,
+      "loss": 5.8828,
+      "step": 130
+    },
+    {
+      "epoch": 0.08779057591623037,
+      "grad_norm": 2.526623010635376,
+      "learning_rate": 4.9242887515199215e-06,
+      "loss": 6.1022,
+      "step": 131
+    },
+    {
+      "epoch": 0.08846073298429319,
+      "grad_norm": 2.3147377967834473,
+      "learning_rate": 4.9229845311868275e-06,
+      "loss": 5.8275,
+      "step": 132
+    },
+    {
+      "epoch": 0.08913089005235603,
+      "grad_norm": 2.0316696166992188,
+      "learning_rate": 4.921669348868e-06,
+      "loss": 5.7102,
+      "step": 133
+    },
+    {
+      "epoch": 0.08980104712041885,
+      "grad_norm": 2.1787824630737305,
+      "learning_rate": 4.9203432105135425e-06,
+      "loss": 5.8846,
+      "step": 134
+    },
+    {
+      "epoch": 0.09047120418848167,
+      "grad_norm": 2.2715516090393066,
+      "learning_rate": 4.919006122123125e-06,
+      "loss": 6.0401,
+      "step": 135
+    },
+    {
+      "epoch": 0.0911413612565445,
+      "grad_norm": 2.4892923831939697,
+      "learning_rate": 4.917658089745959e-06,
+      "loss": 6.322,
+      "step": 136
+    },
+    {
+      "epoch": 0.09181151832460734,
+      "grad_norm": 2.3479976654052734,
+      "learning_rate": 4.916299119480767e-06,
+      "loss": 5.578,
+      "step": 137
+    },
+    {
+      "epoch": 0.09248167539267016,
+      "grad_norm": 1.8516892194747925,
+      "learning_rate": 4.914929217475758e-06,
+      "loss": 5.6311,
+      "step": 138
+    },
+    {
+      "epoch": 0.09315183246073298,
+      "grad_norm": 2.260326623916626,
+      "learning_rate": 4.913548389928595e-06,
+      "loss": 5.7489,
+      "step": 139
+    },
+    {
+      "epoch": 0.09382198952879581,
+      "grad_norm": 2.2286529541015625,
+      "learning_rate": 4.912156643086374e-06,
+      "loss": 5.7533,
+      "step": 140
+    },
+    {
+      "epoch": 0.09449214659685864,
+      "grad_norm": 2.654252767562866,
+      "learning_rate": 4.910753983245589e-06,
+      "loss": 5.8891,
+      "step": 141
+    },
+    {
+      "epoch": 0.09516230366492147,
+      "grad_norm": 2.1620259284973145,
+      "learning_rate": 4.909340416752105e-06,
+      "loss": 5.9395,
+      "step": 142
+    },
+    {
+      "epoch": 0.09583246073298429,
+      "grad_norm": 2.2099599838256836,
+      "learning_rate": 4.907915950001135e-06,
+      "loss": 5.8191,
+      "step": 143
+    },
+    {
+      "epoch": 0.09650261780104712,
+      "grad_norm": 2.1290504932403564,
+      "learning_rate": 4.906480589437202e-06,
+      "loss": 5.6433,
+      "step": 144
+    },
+    {
+      "epoch": 0.09717277486910995,
+      "grad_norm": 2.3433656692504883,
+      "learning_rate": 4.905034341554117e-06,
+      "loss": 5.6745,
+      "step": 145
+    },
+    {
+      "epoch": 0.09784293193717278,
+      "grad_norm": 2.5840182304382324,
+      "learning_rate": 4.903577212894947e-06,
+      "loss": 5.9487,
+      "step": 146
+    },
+    {
+      "epoch": 0.0985130890052356,
+      "grad_norm": 2.1299641132354736,
+      "learning_rate": 4.902109210051984e-06,
+      "loss": 5.7433,
+      "step": 147
+    },
+    {
+      "epoch": 0.09918324607329843,
+      "grad_norm": 2.7531559467315674,
+      "learning_rate": 4.900630339666718e-06,
+      "loss": 5.9031,
+      "step": 148
+    },
+    {
+      "epoch": 0.09985340314136126,
+      "grad_norm": 2.4779627323150635,
+      "learning_rate": 4.899140608429804e-06,
+      "loss": 5.7958,
+      "step": 149
+    },
+    {
+      "epoch": 0.10052356020942409,
+      "grad_norm": 2.1861352920532227,
+      "learning_rate": 4.897640023081036e-06,
+      "loss": 5.7467,
+      "step": 150
+    },
+    {
+      "epoch": 0.10119371727748691,
+      "grad_norm": 2.305194854736328,
+      "learning_rate": 4.896128590409311e-06,
+      "loss": 5.8453,
+      "step": 151
+    },
+    {
+      "epoch": 0.10186387434554973,
+      "grad_norm": 2.6004295349121094,
+      "learning_rate": 4.894606317252603e-06,
+      "loss": 5.5061,
+      "step": 152
+    },
+    {
+      "epoch": 0.10253403141361256,
+      "grad_norm": 2.268364191055298,
+      "learning_rate": 4.893073210497928e-06,
+      "loss": 5.8876,
+      "step": 153
+    },
+    {
+      "epoch": 0.1032041884816754,
+      "grad_norm": 2.2736222743988037,
+      "learning_rate": 4.8915292770813184e-06,
+      "loss": 5.9355,
+      "step": 154
+    },
+    {
+      "epoch": 0.10387434554973822,
+      "grad_norm": 2.465298652648926,
+      "learning_rate": 4.889974523987784e-06,
+      "loss": 6.0257,
+      "step": 155
+    },
+    {
+      "epoch": 0.10454450261780104,
+      "grad_norm": 2.676025629043579,
+      "learning_rate": 4.888408958251289e-06,
+      "loss": 5.5879,
+      "step": 156
+    },
+    {
+      "epoch": 0.10521465968586387,
+      "grad_norm": 2.2422494888305664,
+      "learning_rate": 4.886832586954712e-06,
+      "loss": 5.8153,
+      "step": 157
+    },
+    {
+      "epoch": 0.1058848167539267,
+      "grad_norm": 2.0813255310058594,
+      "learning_rate": 4.8852454172298205e-06,
+      "loss": 5.4443,
+      "step": 158
+    },
+    {
+      "epoch": 0.10655497382198953,
+      "grad_norm": 2.5441648960113525,
+      "learning_rate": 4.883647456257234e-06,
+      "loss": 5.6157,
+      "step": 159
+    },
+    {
+      "epoch": 0.10722513089005235,
+      "grad_norm": 2.377225637435913,
+      "learning_rate": 4.882038711266395e-06,
+      "loss": 5.6878,
+      "step": 160
+    },
+    {
+      "epoch": 0.10789528795811518,
+      "grad_norm": 2.6996095180511475,
+      "learning_rate": 4.880419189535532e-06,
+      "loss": 5.6048,
+      "step": 161
+    },
+    {
+      "epoch": 0.10856544502617801,
+      "grad_norm": 2.653292655944824,
+      "learning_rate": 4.878788898391634e-06,
+      "loss": 5.6416,
+      "step": 162
+    },
+    {
+      "epoch": 0.10923560209424084,
+      "grad_norm": 2.5808253288269043,
+      "learning_rate": 4.877147845210407e-06,
+      "loss": 5.6885,
+      "step": 163
+    },
+    {
+      "epoch": 0.10990575916230366,
+      "grad_norm": 2.5603091716766357,
+      "learning_rate": 4.8754960374162516e-06,
+      "loss": 5.6075,
+      "step": 164
+    },
+    {
+      "epoch": 0.11057591623036649,
+      "grad_norm": 2.2412588596343994,
+      "learning_rate": 4.87383348248222e-06,
+      "loss": 5.1883,
+      "step": 165
+    },
+    {
+      "epoch": 0.11124607329842932,
+      "grad_norm": 2.5953209400177,
+      "learning_rate": 4.872160187929987e-06,
+      "loss": 5.4755,
+      "step": 166
+    },
+    {
+      "epoch": 0.11191623036649215,
+      "grad_norm": 2.453779697418213,
+      "learning_rate": 4.870476161329818e-06,
+      "loss": 5.9441,
+      "step": 167
+    },
+    {
+      "epoch": 0.11258638743455497,
+      "grad_norm": 2.0972816944122314,
+      "learning_rate": 4.868781410300529e-06,
+      "loss": 5.7357,
+      "step": 168
+    },
+    {
+      "epoch": 0.1132565445026178,
+      "grad_norm": 2.21097469329834,
+      "learning_rate": 4.867075942509455e-06,
+      "loss": 5.4357,
+      "step": 169
+    },
+    {
+      "epoch": 0.11392670157068063,
+      "grad_norm": 2.468629837036133,
+      "learning_rate": 4.8653597656724185e-06,
+      "loss": 5.8272,
+      "step": 170
+    },
+    {
+      "epoch": 0.11459685863874346,
+      "grad_norm": 2.063910961151123,
+      "learning_rate": 4.8636328875536875e-06,
+      "loss": 5.4154,
+      "step": 171
+    },
+    {
+      "epoch": 0.11526701570680628,
+      "grad_norm": 2.602414846420288,
+      "learning_rate": 4.861895315965946e-06,
+      "loss": 5.9394,
+      "step": 172
+    },
+    {
+      "epoch": 0.1159371727748691,
+      "grad_norm": 5.8961005210876465,
+      "learning_rate": 4.86014705877026e-06,
+      "loss": 5.4304,
+      "step": 173
+    },
+    {
+      "epoch": 0.11660732984293194,
+      "grad_norm": 2.347381353378296,
+      "learning_rate": 4.858388123876035e-06,
+      "loss": 5.4542,
+      "step": 174
+    },
+    {
+      "epoch": 0.11727748691099477,
+      "grad_norm": 2.1685433387756348,
+      "learning_rate": 4.856618519240985e-06,
+      "loss": 5.7699,
+      "step": 175
+    },
+    {
+      "epoch": 0.11794764397905759,
+      "grad_norm": 2.9305124282836914,
+      "learning_rate": 4.854838252871097e-06,
+      "loss": 5.4849,
+      "step": 176
+    },
+    {
+      "epoch": 0.11861780104712041,
+      "grad_norm": 2.922797441482544,
+      "learning_rate": 4.853047332820592e-06,
+      "loss": 5.521,
+      "step": 177
+    },
+    {
+      "epoch": 0.11928795811518325,
+      "grad_norm": 2.1050190925598145,
+      "learning_rate": 4.85124576719189e-06,
+      "loss": 5.6155,
+      "step": 178
+    },
+    {
+      "epoch": 0.11995811518324608,
+      "grad_norm": 2.704467535018921,
+      "learning_rate": 4.849433564135576e-06,
+      "loss": 5.5141,
+      "step": 179
+    },
+    {
+      "epoch": 0.1206282722513089,
+      "grad_norm": 2.1763556003570557,
+      "learning_rate": 4.847610731850358e-06,
+      "loss": 5.7,
+      "step": 180
+    },
+    {
+      "epoch": 0.12129842931937172,
+      "grad_norm": 2.1507253646850586,
+      "learning_rate": 4.845777278583032e-06,
+      "loss": 5.9349,
+      "step": 181
+    },
+    {
+      "epoch": 0.12196858638743456,
+      "grad_norm": 2.007770299911499,
+      "learning_rate": 4.843933212628447e-06,
+      "loss": 5.8838,
+      "step": 182
+    },
+    {
+      "epoch": 0.12263874345549738,
+      "grad_norm": 2.1414952278137207,
+      "learning_rate": 4.8420785423294645e-06,
+      "loss": 5.7521,
+      "step": 183
+    },
+    {
+      "epoch": 0.12330890052356021,
+      "grad_norm": 1.8680200576782227,
+      "learning_rate": 4.840213276076922e-06,
+      "loss": 5.7706,
+      "step": 184
+    },
+    {
+      "epoch": 0.12397905759162303,
+      "grad_norm": 2.4503700733184814,
+      "learning_rate": 4.8383374223095934e-06,
+      "loss": 5.8948,
+      "step": 185
+    },
+    {
+      "epoch": 0.12464921465968587,
+      "grad_norm": 1.9897749423980713,
+      "learning_rate": 4.836450989514155e-06,
+      "loss": 5.7293,
+      "step": 186
+    },
+    {
+      "epoch": 0.12531937172774868,
+      "grad_norm": 2.661517381668091,
+      "learning_rate": 4.834553986225145e-06,
+      "loss": 5.9449,
+      "step": 187
+    },
+    {
+      "epoch": 0.12598952879581152,
+      "grad_norm": 2.1688408851623535,
+      "learning_rate": 4.832646421024918e-06,
+      "loss": 5.9073,
+      "step": 188
+    },
+    {
+      "epoch": 0.12665968586387436,
+      "grad_norm": 3.2971296310424805,
+      "learning_rate": 4.83072830254362e-06,
+      "loss": 5.6963,
+      "step": 189
+    },
+    {
+      "epoch": 0.12732984293193716,
+      "grad_norm": 2.2415287494659424,
+      "learning_rate": 4.828799639459139e-06,
+      "loss": 5.5034,
+      "step": 190
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 2.502039909362793,
+      "learning_rate": 4.826860440497065e-06,
+      "loss": 5.5988,
+      "step": 191
+    },
+    {
+      "epoch": 0.12867015706806284,
+      "grad_norm": 2.392843723297119,
+      "learning_rate": 4.824910714430659e-06,
+      "loss": 5.6986,
+      "step": 192
+    },
+    {
+      "epoch": 0.12934031413612565,
+      "grad_norm": 2.7530651092529297,
+      "learning_rate": 4.822950470080808e-06,
+      "loss": 5.8121,
+      "step": 193
+    },
+    {
+      "epoch": 0.1300104712041885,
+      "grad_norm": 2.3004884719848633,
+      "learning_rate": 4.820979716315983e-06,
+      "loss": 5.6649,
+      "step": 194
+    },
+    {
+      "epoch": 0.1306806282722513,
+      "grad_norm": 2.560392379760742,
+      "learning_rate": 4.8189984620522014e-06,
+      "loss": 5.7792,
+      "step": 195
+    },
+    {
+      "epoch": 0.13135078534031414,
+      "grad_norm": 2.2837352752685547,
+      "learning_rate": 4.8170067162529884e-06,
+      "loss": 5.7563,
+      "step": 196
+    },
+    {
+      "epoch": 0.13202094240837697,
+      "grad_norm": 2.555257558822632,
+      "learning_rate": 4.8150044879293336e-06,
+      "loss": 5.6809,
+      "step": 197
+    },
+    {
+      "epoch": 0.13269109947643978,
+      "grad_norm": 2.5194332599639893,
+      "learning_rate": 4.812991786139652e-06,
+      "loss": 5.8421,
+      "step": 198
+    },
+    {
+      "epoch": 0.13336125654450262,
+      "grad_norm": 2.4896628856658936,
+      "learning_rate": 4.810968619989742e-06,
+      "loss": 5.7051,
+      "step": 199
+    },
+    {
+      "epoch": 0.13403141361256546,
+      "grad_norm": 2.1717307567596436,
+      "learning_rate": 4.808934998632743e-06,
+      "loss": 5.5694,
+      "step": 200
+    },
+    {
+      "epoch": 0.13470157068062827,
+      "grad_norm": 1.9135693311691284,
+      "learning_rate": 4.806890931269098e-06,
+      "loss": 5.5108,
+      "step": 201
+    },
+    {
+      "epoch": 0.1353717277486911,
+      "grad_norm": 2.1709461212158203,
+      "learning_rate": 4.804836427146509e-06,
+      "loss": 5.7672,
+      "step": 202
+    },
+    {
+      "epoch": 0.13604188481675392,
+      "grad_norm": 3.3644444942474365,
+      "learning_rate": 4.8027714955598935e-06,
+      "loss": 5.6492,
+      "step": 203
+    },
+    {
+      "epoch": 0.13671204188481675,
+      "grad_norm": 2.060128927230835,
+      "learning_rate": 4.800696145851346e-06,
+      "loss": 5.8783,
+      "step": 204
+    },
+    {
+      "epoch": 0.1373821989528796,
+      "grad_norm": 2.0333595275878906,
+      "learning_rate": 4.798610387410097e-06,
+      "loss": 5.5442,
+      "step": 205
+    },
+    {
+      "epoch": 0.1380523560209424,
+      "grad_norm": 2.221376419067383,
+      "learning_rate": 4.7965142296724616e-06,
+      "loss": 5.582,
+      "step": 206
+    },
+    {
+      "epoch": 0.13872251308900524,
+      "grad_norm": 2.498304605484009,
+      "learning_rate": 4.794407682121809e-06,
+      "loss": 5.8169,
+      "step": 207
+    },
+    {
+      "epoch": 0.13939267015706805,
+      "grad_norm": 2.1738154888153076,
+      "learning_rate": 4.79229075428851e-06,
+      "loss": 5.65,
+      "step": 208
+    },
+    {
+      "epoch": 0.1400628272251309,
+      "grad_norm": 2.296131134033203,
+      "learning_rate": 4.7901634557499e-06,
+      "loss": 5.4729,
+      "step": 209
+    },
+    {
+      "epoch": 0.14073298429319372,
+      "grad_norm": 2.077704906463623,
+      "learning_rate": 4.788025796130232e-06,
+      "loss": 5.7508,
+      "step": 210
+    },
+    {
+      "epoch": 0.14140314136125653,
+      "grad_norm": 2.2950336933135986,
+      "learning_rate": 4.785877785100633e-06,
+      "loss": 5.6617,
+      "step": 211
+    },
+    {
+      "epoch": 0.14207329842931937,
+      "grad_norm": 2.310399293899536,
+      "learning_rate": 4.783719432379064e-06,
+      "loss": 5.4773,
+      "step": 212
+    },
+    {
+      "epoch": 0.1427434554973822,
+      "grad_norm": 2.098602771759033,
+      "learning_rate": 4.781550747730271e-06,
+      "loss": 5.8573,
+      "step": 213
+    },
+    {
+      "epoch": 0.14341361256544502,
+      "grad_norm": 2.5833826065063477,
+      "learning_rate": 4.779371740965747e-06,
+      "loss": 5.5664,
+      "step": 214
+    },
+    {
+      "epoch": 0.14408376963350786,
+      "grad_norm": 2.150275707244873,
+      "learning_rate": 4.7771824219436805e-06,
+      "loss": 5.4671,
+      "step": 215
+    },
+    {
+      "epoch": 0.14475392670157067,
+      "grad_norm": 2.187807559967041,
+      "learning_rate": 4.774982800568913e-06,
+      "loss": 5.5364,
+      "step": 216
+    },
+    {
+      "epoch": 0.1454240837696335,
+      "grad_norm": 2.631636619567871,
+      "learning_rate": 4.772772886792903e-06,
+      "loss": 5.5136,
+      "step": 217
+    },
+    {
+      "epoch": 0.14609424083769634,
+      "grad_norm": 2.5570361614227295,
+      "learning_rate": 4.770552690613665e-06,
+      "loss": 5.6872,
+      "step": 218
+    },
+    {
+      "epoch": 0.14676439790575915,
+      "grad_norm": 2.5417332649230957,
+      "learning_rate": 4.7683222220757375e-06,
+      "loss": 5.7036,
+      "step": 219
+    },
+    {
+      "epoch": 0.147434554973822,
+      "grad_norm": 2.375767230987549,
+      "learning_rate": 4.766081491270132e-06,
+      "loss": 5.5643,
+      "step": 220
+    },
+    {
+      "epoch": 0.14810471204188483,
+      "grad_norm": 1.9284825325012207,
+      "learning_rate": 4.7638305083342885e-06,
+      "loss": 5.5584,
+      "step": 221
+    },
+    {
+      "epoch": 0.14877486910994764,
+      "grad_norm": 2.257730484008789,
+      "learning_rate": 4.761569283452027e-06,
+      "loss": 5.8337,
+      "step": 222
+    },
+    {
+      "epoch": 0.14944502617801048,
+      "grad_norm": 2.2399933338165283,
+      "learning_rate": 4.7592978268535075e-06,
+      "loss": 5.6251,
+      "step": 223
+    },
+    {
+      "epoch": 0.15011518324607329,
+      "grad_norm": 2.2639455795288086,
+      "learning_rate": 4.757016148815177e-06,
+      "loss": 5.5898,
+      "step": 224
+    },
+    {
+      "epoch": 0.15078534031413612,
+      "grad_norm": 2.5909502506256104,
+      "learning_rate": 4.754724259659727e-06,
+      "loss": 5.4121,
+      "step": 225
+    },
+    {
+      "epoch": 0.15145549738219896,
+      "grad_norm": 1.9159077405929565,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 5.462,
+      "step": 226
+    },
+    {
+      "epoch": 0.15212565445026177,
+      "grad_norm": 1.896048665046692,
+      "learning_rate": 4.750109889519176e-06,
+      "loss": 5.7881,
+      "step": 227
+    },
+    {
+      "epoch": 0.1527958115183246,
+      "grad_norm": 2.2421371936798096,
+      "learning_rate": 4.747787429410253e-06,
+      "loss": 5.4056,
+      "step": 228
+    },
+    {
+      "epoch": 0.15346596858638745,
+      "grad_norm": 2.3694958686828613,
+      "learning_rate": 4.745454799936475e-06,
+      "loss": 5.4631,
+      "step": 229
+    },
+    {
+      "epoch": 0.15413612565445026,
+      "grad_norm": 2.3106985092163086,
+      "learning_rate": 4.743112011651047e-06,
+      "loss": 5.7385,
+      "step": 230
+    },
+    {
+      "epoch": 0.1548062827225131,
+      "grad_norm": 2.0970394611358643,
+      "learning_rate": 4.740759075153134e-06,
+      "loss": 5.5615,
+      "step": 231
+    },
+    {
+      "epoch": 0.1554764397905759,
+      "grad_norm": 2.0424554347991943,
+      "learning_rate": 4.738396001087811e-06,
+      "loss": 5.7584,
+      "step": 232
+    },
+    {
+      "epoch": 0.15614659685863874,
+      "grad_norm": 2.465092420578003,
+      "learning_rate": 4.73602280014602e-06,
+      "loss": 5.5121,
+      "step": 233
+    },
+    {
+      "epoch": 0.15681675392670158,
+      "grad_norm": 2.286259412765503,
+      "learning_rate": 4.733639483064517e-06,
+      "loss": 5.7908,
+      "step": 234
+    },
+    {
+      "epoch": 0.1574869109947644,
+      "grad_norm": 2.41261887550354,
+      "learning_rate": 4.731246060625826e-06,
+      "loss": 5.8801,
+      "step": 235
+    },
+    {
+      "epoch": 0.15815706806282723,
+      "grad_norm": 2.4263551235198975,
+      "learning_rate": 4.72884254365819e-06,
+      "loss": 5.4863,
+      "step": 236
+    },
+    {
+      "epoch": 0.15882722513089007,
+      "grad_norm": 2.1641643047332764,
+      "learning_rate": 4.7264289430355185e-06,
+      "loss": 5.7159,
+      "step": 237
+    },
+    {
+      "epoch": 0.15949738219895288,
+      "grad_norm": 2.152360200881958,
+      "learning_rate": 4.724005269677342e-06,
+      "loss": 5.3524,
+      "step": 238
+    },
+    {
+      "epoch": 0.1601675392670157,
+      "grad_norm": 2.6608517169952393,
+      "learning_rate": 4.721571534548766e-06,
+      "loss": 5.6378,
+      "step": 239
+    },
+    {
+      "epoch": 0.16083769633507852,
+      "grad_norm": 2.169257640838623,
+      "learning_rate": 4.71912774866041e-06,
+      "loss": 5.8246,
+      "step": 240
+    },
+    {
+      "epoch": 0.16150785340314136,
+      "grad_norm": 2.7111823558807373,
+      "learning_rate": 4.716673923068371e-06,
+      "loss": 5.6595,
+      "step": 241
+    },
+    {
+      "epoch": 0.1621780104712042,
+      "grad_norm": 2.112722635269165,
+      "learning_rate": 4.7142100688741645e-06,
+      "loss": 5.5694,
+      "step": 242
+    },
+    {
+      "epoch": 0.162848167539267,
+      "grad_norm": 2.59005069732666,
+      "learning_rate": 4.711736197224677e-06,
+      "loss": 5.6073,
+      "step": 243
+    },
+    {
+      "epoch": 0.16351832460732985,
+      "grad_norm": 2.3710269927978516,
+      "learning_rate": 4.709252319312117e-06,
+      "loss": 5.4579,
+      "step": 244
+    },
+    {
+      "epoch": 0.16418848167539268,
+      "grad_norm": 2.2614245414733887,
+      "learning_rate": 4.706758446373959e-06,
+      "loss": 5.5483,
+      "step": 245
+    },
+    {
+      "epoch": 0.1648586387434555,
+      "grad_norm": 2.088327407836914,
+      "learning_rate": 4.704254589692903e-06,
+      "loss": 5.4686,
+      "step": 246
+    },
+    {
+      "epoch": 0.16552879581151833,
+      "grad_norm": 2.09621524810791,
+      "learning_rate": 4.7017407605968125e-06,
+      "loss": 5.7135,
+      "step": 247
+    },
+    {
+      "epoch": 0.16619895287958114,
+      "grad_norm": 1.8950719833374023,
+      "learning_rate": 4.6992169704586685e-06,
+      "loss": 5.4741,
+      "step": 248
+    },
+    {
+      "epoch": 0.16686910994764398,
+      "grad_norm": 2.1435563564300537,
+      "learning_rate": 4.69668323069652e-06,
+      "loss": 5.4941,
+      "step": 249
+    },
+    {
+      "epoch": 0.16753926701570682,
+      "grad_norm": 2.2771897315979004,
+      "learning_rate": 4.694139552773426e-06,
+      "loss": 5.526,
+      "step": 250
+    },
+    {
+      "epoch": 0.16820942408376963,
+      "grad_norm": 2.7859890460968018,
+      "learning_rate": 4.69158594819741e-06,
+      "loss": 5.5536,
+      "step": 251
+    },
+    {
+      "epoch": 0.16887958115183246,
+      "grad_norm": 2.765713691711426,
+      "learning_rate": 4.689022428521405e-06,
+      "loss": 5.6265,
+      "step": 252
+    },
+    {
+      "epoch": 0.1695497382198953,
+      "grad_norm": 2.27315616607666,
+      "learning_rate": 4.6864490053432e-06,
+      "loss": 5.6416,
+      "step": 253
+    },
+    {
+      "epoch": 0.1702198952879581,
+      "grad_norm": 2.1748857498168945,
+      "learning_rate": 4.683865690305392e-06,
+      "loss": 5.688,
+      "step": 254
+    },
+    {
+      "epoch": 0.17089005235602095,
+      "grad_norm": 2.1356191635131836,
+      "learning_rate": 4.681272495095327e-06,
+      "loss": 5.5414,
+      "step": 255
+    },
+    {
+      "epoch": 0.17156020942408376,
+      "grad_norm": 2.1574511528015137,
+      "learning_rate": 4.6786694314450535e-06,
+      "loss": 5.6417,
+      "step": 256
+    },
+    {
+      "epoch": 0.1722303664921466,
+      "grad_norm": 2.3209385871887207,
+      "learning_rate": 4.6760565111312625e-06,
+      "loss": 5.5285,
+      "step": 257
+    },
+    {
+      "epoch": 0.17290052356020943,
+      "grad_norm": 2.666755437850952,
+      "learning_rate": 4.673433745975245e-06,
+      "loss": 5.6753,
+      "step": 258
+    },
+    {
+      "epoch": 0.17357068062827224,
+      "grad_norm": 2.35461163520813,
+      "learning_rate": 4.670801147842824e-06,
+      "loss": 5.4898,
+      "step": 259
+    },
+    {
+      "epoch": 0.17424083769633508,
+      "grad_norm": 2.502377986907959,
+      "learning_rate": 4.668158728644315e-06,
+      "loss": 5.1188,
+      "step": 260
+    },
+    {
+      "epoch": 0.1749109947643979,
+      "grad_norm": 2.217586040496826,
+      "learning_rate": 4.66550650033446e-06,
+      "loss": 5.5627,
+      "step": 261
+    },
+    {
+      "epoch": 0.17558115183246073,
+      "grad_norm": 2.145159959793091,
+      "learning_rate": 4.662844474912384e-06,
+      "loss": 5.2494,
+      "step": 262
+    },
+    {
+      "epoch": 0.17625130890052357,
+      "grad_norm": 2.508312702178955,
+      "learning_rate": 4.660172664421532e-06,
+      "loss": 5.6636,
+      "step": 263
+    },
+    {
+      "epoch": 0.17692146596858638,
+      "grad_norm": 2.927377939224243,
+      "learning_rate": 4.65749108094962e-06,
+      "loss": 5.6348,
+      "step": 264
+    },
+    {
+      "epoch": 0.17759162303664922,
+      "grad_norm": 2.3884453773498535,
+      "learning_rate": 4.654799736628579e-06,
+      "loss": 5.7108,
+      "step": 265
+    },
+    {
+      "epoch": 0.17826178010471205,
+      "grad_norm": 2.3559861183166504,
+      "learning_rate": 4.6520986436345e-06,
+      "loss": 5.5178,
+      "step": 266
+    },
+    {
+      "epoch": 0.17893193717277486,
+      "grad_norm": 1.8426997661590576,
+      "learning_rate": 4.649387814187575e-06,
+      "loss": 5.4479,
+      "step": 267
+    },
+    {
+      "epoch": 0.1796020942408377,
+      "grad_norm": 2.4238853454589844,
+      "learning_rate": 4.646667260552051e-06,
+      "loss": 5.5709,
+      "step": 268
+    },
+    {
+      "epoch": 0.1802722513089005,
+      "grad_norm": 6.363220691680908,
+      "learning_rate": 4.643936995036164e-06,
+      "loss": 5.6736,
+      "step": 269
+    },
+    {
+      "epoch": 0.18094240837696335,
+      "grad_norm": 2.6399664878845215,
+      "learning_rate": 4.641197029992093e-06,
+      "loss": 5.4715,
+      "step": 270
+    },
+    {
+      "epoch": 0.1816125654450262,
+      "grad_norm": 1.954572081565857,
+      "learning_rate": 4.638447377815892e-06,
+      "loss": 5.3817,
+      "step": 271
+    },
+    {
+      "epoch": 0.182282722513089,
+      "grad_norm": 2.6681385040283203,
+      "learning_rate": 4.635688050947449e-06,
+      "loss": 5.6239,
+      "step": 272
+    },
+    {
+      "epoch": 0.18295287958115183,
+      "grad_norm": 2.382075786590576,
+      "learning_rate": 4.632919061870417e-06,
+      "loss": 5.6783,
+      "step": 273
+    },
+    {
+      "epoch": 0.18362303664921467,
+      "grad_norm": 2.142279624938965,
+      "learning_rate": 4.630140423112164e-06,
+      "loss": 5.3735,
+      "step": 274
+    },
+    {
+      "epoch": 0.18429319371727748,
+      "grad_norm": 2.151935338973999,
+      "learning_rate": 4.627352147243715e-06,
+      "loss": 5.5223,
+      "step": 275
+    },
+    {
+      "epoch": 0.18496335078534032,
+      "grad_norm": 1.9508206844329834,
+      "learning_rate": 4.624554246879695e-06,
+      "loss": 5.2748,
+      "step": 276
+    },
+    {
+      "epoch": 0.18563350785340313,
+      "grad_norm": 2.103137969970703,
+      "learning_rate": 4.6217467346782714e-06,
+      "loss": 5.6259,
+      "step": 277
+    },
+    {
+      "epoch": 0.18630366492146597,
+      "grad_norm": 2.361621618270874,
+      "learning_rate": 4.618929623341097e-06,
+      "loss": 5.6894,
+      "step": 278
+    },
+    {
+      "epoch": 0.1869738219895288,
+      "grad_norm": 2.6873679161071777,
+      "learning_rate": 4.616102925613252e-06,
+      "loss": 5.4963,
+      "step": 279
+    },
+    {
+      "epoch": 0.18764397905759161,
+      "grad_norm": 2.3075380325317383,
+      "learning_rate": 4.613266654283189e-06,
+      "loss": 5.6025,
+      "step": 280
+    },
+    {
+      "epoch": 0.18831413612565445,
+      "grad_norm": 2.0175352096557617,
+      "learning_rate": 4.610420822182671e-06,
+      "loss": 5.5559,
+      "step": 281
+    },
+    {
+      "epoch": 0.1889842931937173,
+      "grad_norm": 2.1352314949035645,
+      "learning_rate": 4.607565442186719e-06,
+      "loss": 5.4375,
+      "step": 282
+    },
+    {
+      "epoch": 0.1896544502617801,
+      "grad_norm": 2.4613332748413086,
+      "learning_rate": 4.604700527213545e-06,
+      "loss": 5.6111,
+      "step": 283
+    },
+    {
+      "epoch": 0.19032460732984294,
+      "grad_norm": 2.250486135482788,
+      "learning_rate": 4.601826090224503e-06,
+      "loss": 5.8779,
+      "step": 284
+    },
+    {
+      "epoch": 0.19099476439790575,
+      "grad_norm": 2.21348237991333,
+      "learning_rate": 4.598942144224025e-06,
+      "loss": 5.3281,
+      "step": 285
+    },
+    {
+      "epoch": 0.19166492146596859,
+      "grad_norm": 2.185913562774658,
+      "learning_rate": 4.596048702259563e-06,
+      "loss": 5.6081,
+      "step": 286
+    },
+    {
+      "epoch": 0.19233507853403142,
+      "grad_norm": 2.3666489124298096,
+      "learning_rate": 4.593145777421529e-06,
+      "loss": 5.4173,
+      "step": 287
+    },
+    {
+      "epoch": 0.19300523560209423,
+      "grad_norm": 1.9570116996765137,
+      "learning_rate": 4.590233382843242e-06,
+      "loss": 5.6447,
+      "step": 288
+    },
+    {
+      "epoch": 0.19367539267015707,
+      "grad_norm": 3.0195152759552,
+      "learning_rate": 4.587311531700858e-06,
+      "loss": 5.495,
+      "step": 289
+    },
+    {
+      "epoch": 0.1943455497382199,
+      "grad_norm": 2.4136688709259033,
+      "learning_rate": 4.584380237213318e-06,
+      "loss": 5.2192,
+      "step": 290
+    },
+    {
+      "epoch": 0.19501570680628272,
+      "grad_norm": 2.426723003387451,
+      "learning_rate": 4.581439512642288e-06,
+      "loss": 5.522,
+      "step": 291
+    },
+    {
+      "epoch": 0.19568586387434556,
+      "grad_norm": 2.700432062149048,
+      "learning_rate": 4.578489371292097e-06,
+      "loss": 5.5662,
+      "step": 292
+    },
+    {
+      "epoch": 0.19635602094240837,
+      "grad_norm": 2.3877530097961426,
+      "learning_rate": 4.5755298265096725e-06,
+      "loss": 5.5128,
+      "step": 293
+    },
+    {
+      "epoch": 0.1970261780104712,
+      "grad_norm": 2.2122280597686768,
+      "learning_rate": 4.572560891684491e-06,
+      "loss": 5.7546,
+      "step": 294
+    },
+    {
+      "epoch": 0.19769633507853404,
+      "grad_norm": 2.132734775543213,
+      "learning_rate": 4.569582580248509e-06,
+      "loss": 5.5579,
+      "step": 295
+    },
+    {
+      "epoch": 0.19836649214659685,
+      "grad_norm": 2.368992328643799,
+      "learning_rate": 4.566594905676099e-06,
+      "loss": 5.6109,
+      "step": 296
+    },
+    {
+      "epoch": 0.1990366492146597,
+      "grad_norm": 2.3953449726104736,
+      "learning_rate": 4.563597881484004e-06,
+      "loss": 5.5182,
+      "step": 297
+    },
+    {
+      "epoch": 0.19970680628272253,
+      "grad_norm": 2.268723487854004,
+      "learning_rate": 4.560591521231259e-06,
+      "loss": 5.5564,
+      "step": 298
+    },
+    {
+      "epoch": 0.20037696335078534,
+      "grad_norm": 2.2275869846343994,
+      "learning_rate": 4.557575838519137e-06,
+      "loss": 5.5888,
+      "step": 299
+    },
+    {
+      "epoch": 0.20104712041884817,
+      "grad_norm": 1.8049523830413818,
+      "learning_rate": 4.554550846991091e-06,
+      "loss": 5.4499,
+      "step": 300
+    },
+    {
+      "epoch": 0.20171727748691098,
+      "grad_norm": 1.9729605913162231,
+      "learning_rate": 4.551516560332686e-06,
+      "loss": 5.5766,
+      "step": 301
+    },
+    {
+      "epoch": 0.20238743455497382,
+      "grad_norm": 1.977363109588623,
+      "learning_rate": 4.548472992271541e-06,
+      "loss": 5.6398,
+      "step": 302
+    },
+    {
+      "epoch": 0.20305759162303666,
+      "grad_norm": 2.736560821533203,
+      "learning_rate": 4.5454201565772635e-06,
+      "loss": 5.4898,
+      "step": 303
+    },
+    {
+      "epoch": 0.20372774869109947,
+      "grad_norm": 2.1042449474334717,
+      "learning_rate": 4.5423580670613925e-06,
+      "loss": 5.4519,
+      "step": 304
+    },
+    {
+      "epoch": 0.2043979057591623,
+      "grad_norm": 2.3877830505371094,
+      "learning_rate": 4.539286737577331e-06,
+      "loss": 5.402,
+      "step": 305
+    },
+    {
+      "epoch": 0.20506806282722512,
+      "grad_norm": 2.189049243927002,
+      "learning_rate": 4.536206182020284e-06,
+      "loss": 5.1588,
+      "step": 306
+    },
+    {
+      "epoch": 0.20573821989528795,
+      "grad_norm": 2.372809648513794,
+      "learning_rate": 4.533116414327199e-06,
+      "loss": 5.4298,
+      "step": 307
+    },
+    {
+      "epoch": 0.2064083769633508,
+      "grad_norm": 2.6864686012268066,
+      "learning_rate": 4.5300174484767e-06,
+      "loss": 5.056,
+      "step": 308
+    },
+    {
+      "epoch": 0.2070785340314136,
+      "grad_norm": 2.790855646133423,
+      "learning_rate": 4.5269092984890254e-06,
+      "loss": 5.8052,
+      "step": 309
+    },
+    {
+      "epoch": 0.20774869109947644,
+      "grad_norm": 2.2243454456329346,
+      "learning_rate": 4.523791978425962e-06,
+      "loss": 5.7378,
+      "step": 310
+    },
+    {
+      "epoch": 0.20841884816753928,
+      "grad_norm": 2.1781084537506104,
+      "learning_rate": 4.520665502390786e-06,
+      "loss": 5.4125,
+      "step": 311
+    },
+    {
+      "epoch": 0.2090890052356021,
+      "grad_norm": 2.8137004375457764,
+      "learning_rate": 4.517529884528196e-06,
+      "loss": 5.8347,
+      "step": 312
+    },
+    {
+      "epoch": 0.20975916230366493,
+      "grad_norm": 2.380362033843994,
+      "learning_rate": 4.51438513902425e-06,
+      "loss": 5.7121,
+      "step": 313
+    },
+    {
+      "epoch": 0.21042931937172774,
+      "grad_norm": 2.651210069656372,
+      "learning_rate": 4.5112312801063e-06,
+      "loss": 5.2444,
+      "step": 314
+    },
+    {
+      "epoch": 0.21109947643979057,
+      "grad_norm": 2.1358680725097656,
+      "learning_rate": 4.50806832204293e-06,
+      "loss": 5.4798,
+      "step": 315
+    },
+    {
+      "epoch": 0.2117696335078534,
+      "grad_norm": 1.9323924779891968,
+      "learning_rate": 4.5048962791438885e-06,
+      "loss": 5.2646,
+      "step": 316
+    },
+    {
+      "epoch": 0.21243979057591622,
+      "grad_norm": 2.2739667892456055,
+      "learning_rate": 4.501715165760027e-06,
+      "loss": 5.6293,
+      "step": 317
+    },
+    {
+      "epoch": 0.21310994764397906,
+      "grad_norm": 2.425351142883301,
+      "learning_rate": 4.4985249962832316e-06,
+      "loss": 5.1849,
+      "step": 318
+    },
+    {
+      "epoch": 0.2137801047120419,
+      "grad_norm": 1.904436469078064,
+      "learning_rate": 4.4953257851463595e-06,
+      "loss": 5.2887,
+      "step": 319
+    },
+    {
+      "epoch": 0.2144502617801047,
+      "grad_norm": 1.8905036449432373,
+      "learning_rate": 4.492117546823178e-06,
+      "loss": 5.2969,
+      "step": 320
+    },
+    {
+      "epoch": 0.21512041884816754,
+      "grad_norm": 2.0948808193206787,
+      "learning_rate": 4.488900295828289e-06,
+      "loss": 5.5535,
+      "step": 321
+    },
+    {
+      "epoch": 0.21579057591623035,
+      "grad_norm": 2.1106090545654297,
+      "learning_rate": 4.485674046717074e-06,
+      "loss": 5.5733,
+      "step": 322
+    },
+    {
+      "epoch": 0.2164607329842932,
+      "grad_norm": 2.204378128051758,
+      "learning_rate": 4.48243881408562e-06,
+      "loss": 5.5237,
+      "step": 323
+    },
+    {
+      "epoch": 0.21713089005235603,
+      "grad_norm": 2.3489532470703125,
+      "learning_rate": 4.479194612570659e-06,
+      "loss": 5.624,
+      "step": 324
+    },
+    {
+      "epoch": 0.21780104712041884,
+      "grad_norm": 2.2582662105560303,
+      "learning_rate": 4.475941456849501e-06,
+      "loss": 5.2971,
+      "step": 325
+    },
+    {
+      "epoch": 0.21847120418848168,
+      "grad_norm": 2.7756502628326416,
+      "learning_rate": 4.472679361639961e-06,
+      "loss": 5.1575,
+      "step": 326
+    },
+    {
+      "epoch": 0.21914136125654451,
+      "grad_norm": 2.4752912521362305,
+      "learning_rate": 4.469408341700304e-06,
+      "loss": 5.4496,
+      "step": 327
+    },
+    {
+      "epoch": 0.21981151832460732,
+      "grad_norm": 2.538774013519287,
+      "learning_rate": 4.466128411829168e-06,
+      "loss": 5.2787,
+      "step": 328
+    },
+    {
+      "epoch": 0.22048167539267016,
+      "grad_norm": 2.3332388401031494,
+      "learning_rate": 4.462839586865503e-06,
+      "loss": 5.6372,
+      "step": 329
+    },
+    {
+      "epoch": 0.22115183246073297,
+      "grad_norm": 2.3022541999816895,
+      "learning_rate": 4.459541881688501e-06,
+      "loss": 5.5492,
+      "step": 330
+    },
+    {
+      "epoch": 0.2218219895287958,
+      "grad_norm": 2.0971004962921143,
+      "learning_rate": 4.456235311217528e-06,
+      "loss": 5.4539,
+      "step": 331
+    },
+    {
+      "epoch": 0.22249214659685865,
+      "grad_norm": 2.689119815826416,
+      "learning_rate": 4.452919890412062e-06,
+      "loss": 5.43,
+      "step": 332
+    },
+    {
+      "epoch": 0.22316230366492146,
+      "grad_norm": 2.7383434772491455,
+      "learning_rate": 4.449595634271619e-06,
+      "loss": 5.4929,
+      "step": 333
+    },
+    {
+      "epoch": 0.2238324607329843,
+      "grad_norm": 2.2287087440490723,
+      "learning_rate": 4.446262557835686e-06,
+      "loss": 5.6617,
+      "step": 334
+    },
+    {
+      "epoch": 0.22450261780104713,
+      "grad_norm": 2.4107394218444824,
+      "learning_rate": 4.442920676183657e-06,
+      "loss": 5.4094,
+      "step": 335
+    },
+    {
+      "epoch": 0.22517277486910994,
+      "grad_norm": 2.970778465270996,
+      "learning_rate": 4.439570004434762e-06,
+      "loss": 5.59,
+      "step": 336
+    },
+    {
+      "epoch": 0.22584293193717278,
+      "grad_norm": 2.0888166427612305,
+      "learning_rate": 4.436210557747998e-06,
+      "loss": 5.2617,
+      "step": 337
+    },
+    {
+      "epoch": 0.2265130890052356,
+      "grad_norm": 1.821423053741455,
+      "learning_rate": 4.432842351322061e-06,
+      "loss": 5.1656,
+      "step": 338
+    },
+    {
+      "epoch": 0.22718324607329843,
+      "grad_norm": 2.560389757156372,
+      "learning_rate": 4.4294654003952785e-06,
+      "loss": 5.6349,
+      "step": 339
+    },
+    {
+      "epoch": 0.22785340314136127,
+      "grad_norm": 1.9959956407546997,
+      "learning_rate": 4.426079720245539e-06,
+      "loss": 5.5315,
+      "step": 340
+    },
+    {
+      "epoch": 0.22852356020942408,
+      "grad_norm": 2.0525870323181152,
+      "learning_rate": 4.422685326190224e-06,
+      "loss": 5.5211,
+      "step": 341
+    },
+    {
+      "epoch": 0.2291937172774869,
+      "grad_norm": 2.3681249618530273,
+      "learning_rate": 4.419282233586137e-06,
+      "loss": 5.5563,
+      "step": 342
+    },
+    {
+      "epoch": 0.22986387434554975,
+      "grad_norm": 2.2519259452819824,
+      "learning_rate": 4.415870457829436e-06,
+      "loss": 5.5249,
+      "step": 343
+    },
+    {
+      "epoch": 0.23053403141361256,
+      "grad_norm": 2.687836170196533,
+      "learning_rate": 4.412450014355564e-06,
+      "loss": 5.4244,
+      "step": 344
+    },
+    {
+      "epoch": 0.2312041884816754,
+      "grad_norm": 2.29353666305542,
+      "learning_rate": 4.409020918639177e-06,
+      "loss": 5.4925,
+      "step": 345
+    },
+    {
+      "epoch": 0.2318743455497382,
+      "grad_norm": 2.130478858947754,
+      "learning_rate": 4.405583186194074e-06,
+      "loss": 5.3689,
+      "step": 346
+    },
+    {
+      "epoch": 0.23254450261780105,
+      "grad_norm": 1.848215103149414,
+      "learning_rate": 4.402136832573132e-06,
+      "loss": 5.3763,
+      "step": 347
+    },
+    {
+      "epoch": 0.23321465968586388,
+      "grad_norm": 1.9483565092086792,
+      "learning_rate": 4.398681873368226e-06,
+      "loss": 5.8511,
+      "step": 348
+    },
+    {
+      "epoch": 0.2338848167539267,
+      "grad_norm": 1.8991060256958008,
+      "learning_rate": 4.395218324210169e-06,
+      "loss": 5.4366,
+      "step": 349
+    },
+    {
+      "epoch": 0.23455497382198953,
+      "grad_norm": 2.2358591556549072,
+      "learning_rate": 4.391746200768634e-06,
+      "loss": 5.3692,
+      "step": 350
+    },
+    {
+      "epoch": 0.23522513089005237,
+      "grad_norm": 2.185328960418701,
+      "learning_rate": 4.388265518752085e-06,
+      "loss": 5.1356,
+      "step": 351
+    },
+    {
+      "epoch": 0.23589528795811518,
+      "grad_norm": 2.137399911880493,
+      "learning_rate": 4.384776293907709e-06,
+      "loss": 5.2214,
+      "step": 352
+    },
+    {
+      "epoch": 0.23656544502617802,
+      "grad_norm": 1.9800633192062378,
+      "learning_rate": 4.38127854202134e-06,
+      "loss": 5.4082,
+      "step": 353
+    },
+    {
+      "epoch": 0.23723560209424083,
+      "grad_norm": 2.312138795852661,
+      "learning_rate": 4.3777722789173895e-06,
+      "loss": 5.6454,
+      "step": 354
+    },
+    {
+      "epoch": 0.23790575916230366,
+      "grad_norm": 2.105069160461426,
+      "learning_rate": 4.3742575204587774e-06,
+      "loss": 5.4411,
+      "step": 355
+    },
+    {
+      "epoch": 0.2385759162303665,
+      "grad_norm": 2.20829701423645,
+      "learning_rate": 4.370734282546856e-06,
+      "loss": 5.4063,
+      "step": 356
+    },
+    {
+      "epoch": 0.2392460732984293,
+      "grad_norm": 2.180429220199585,
+      "learning_rate": 4.36720258112134e-06,
+      "loss": 5.542,
+      "step": 357
+    },
+    {
+      "epoch": 0.23991623036649215,
+      "grad_norm": 2.4989399909973145,
+      "learning_rate": 4.363662432160236e-06,
+      "loss": 5.2325,
+      "step": 358
+    },
+    {
+      "epoch": 0.24058638743455496,
+      "grad_norm": 2.1118667125701904,
+      "learning_rate": 4.360113851679768e-06,
+      "loss": 5.5798,
+      "step": 359
+    },
+    {
+      "epoch": 0.2412565445026178,
+      "grad_norm": 2.0734126567840576,
+      "learning_rate": 4.356556855734304e-06,
+      "loss": 5.4093,
+      "step": 360
+    },
+    {
+      "epoch": 0.24192670157068064,
+      "grad_norm": 2.4159913063049316,
+      "learning_rate": 4.352991460416287e-06,
+      "loss": 5.3141,
+      "step": 361
+    },
+    {
+      "epoch": 0.24259685863874345,
+      "grad_norm": 2.5480008125305176,
+      "learning_rate": 4.349417681856159e-06,
+      "loss": 5.4403,
+      "step": 362
+    },
+    {
+      "epoch": 0.24326701570680628,
+      "grad_norm": 2.2297019958496094,
+      "learning_rate": 4.345835536222291e-06,
+      "loss": 5.4523,
+      "step": 363
+    },
+    {
+      "epoch": 0.24393717277486912,
+      "grad_norm": 2.1222262382507324,
+      "learning_rate": 4.3422450397209065e-06,
+      "loss": 5.3598,
+      "step": 364
+    },
+    {
+      "epoch": 0.24460732984293193,
+      "grad_norm": 2.1692678928375244,
+      "learning_rate": 4.338646208596009e-06,
+      "loss": 5.5585,
+      "step": 365
+    },
+    {
+      "epoch": 0.24527748691099477,
+      "grad_norm": 2.4140853881835938,
+      "learning_rate": 4.33503905912931e-06,
+      "loss": 5.0457,
+      "step": 366
+    },
+    {
+      "epoch": 0.24594764397905758,
+      "grad_norm": 2.552990436553955,
+      "learning_rate": 4.331423607640155e-06,
+      "loss": 5.3984,
+      "step": 367
+    },
+    {
+      "epoch": 0.24661780104712042,
+      "grad_norm": 2.685238838195801,
+      "learning_rate": 4.32779987048545e-06,
+      "loss": 5.3119,
+      "step": 368
+    },
+    {
+      "epoch": 0.24728795811518325,
+      "grad_norm": 2.309687376022339,
+      "learning_rate": 4.324167864059584e-06,
+      "loss": 5.4823,
+      "step": 369
+    },
+    {
+      "epoch": 0.24795811518324606,
+      "grad_norm": 2.3168299198150635,
+      "learning_rate": 4.32052760479436e-06,
+      "loss": 5.4587,
+      "step": 370
+    },
+    {
+      "epoch": 0.2486282722513089,
+      "grad_norm": 2.093782424926758,
+      "learning_rate": 4.316879109158918e-06,
+      "loss": 5.4821,
+      "step": 371
+    },
+    {
+      "epoch": 0.24929842931937174,
+      "grad_norm": 2.271437406539917,
+      "learning_rate": 4.31322239365966e-06,
+      "loss": 5.4722,
+      "step": 372
+    },
+    {
+      "epoch": 0.24996858638743455,
+      "grad_norm": 2.1802608966827393,
+      "learning_rate": 4.309557474840174e-06,
+      "loss": 5.5796,
+      "step": 373
+    },
+    {
+      "epoch": 0.25063874345549736,
+      "grad_norm": 2.491943597793579,
+      "learning_rate": 4.305884369281167e-06,
+      "loss": 5.5483,
+      "step": 374
+    },
+    {
+      "epoch": 0.2513089005235602,
+      "grad_norm": 2.3541171550750732,
+      "learning_rate": 4.302203093600379e-06,
+      "loss": 5.4499,
+      "step": 375
+    },
+    {
+      "epoch": 0.25197905759162303,
+      "grad_norm": 1.8869248628616333,
+      "learning_rate": 4.298513664452513e-06,
+      "loss": 5.2755,
+      "step": 376
+    },
+    {
+      "epoch": 0.2526492146596859,
+      "grad_norm": 2.284665822982788,
+      "learning_rate": 4.294816098529163e-06,
+      "loss": 5.3206,
+      "step": 377
+    },
+    {
+      "epoch": 0.2533193717277487,
+      "grad_norm": 2.177415609359741,
+      "learning_rate": 4.291110412558733e-06,
+      "loss": 5.5821,
+      "step": 378
+    },
+    {
+      "epoch": 0.2539895287958115,
+      "grad_norm": 2.1283631324768066,
+      "learning_rate": 4.287396623306363e-06,
+      "loss": 5.226,
+      "step": 379
+    },
+    {
+      "epoch": 0.25465968586387433,
+      "grad_norm": 2.4899802207946777,
+      "learning_rate": 4.283674747573856e-06,
+      "loss": 5.343,
+      "step": 380
+    },
+    {
+      "epoch": 0.25532984293193717,
+      "grad_norm": 2.8830642700195312,
+      "learning_rate": 4.279944802199597e-06,
+      "loss": 5.2881,
+      "step": 381
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 2.234779119491577,
+      "learning_rate": 4.2762068040584795e-06,
+      "loss": 5.6279,
+      "step": 382
+    },
+    {
+      "epoch": 0.25667015706806284,
+      "grad_norm": 2.346729278564453,
+      "learning_rate": 4.272460770061832e-06,
+      "loss": 5.3688,
+      "step": 383
+    },
+    {
+      "epoch": 0.2573403141361257,
+      "grad_norm": 2.206702470779419,
+      "learning_rate": 4.268706717157334e-06,
+      "loss": 5.6164,
+      "step": 384
+    },
+    {
+      "epoch": 0.25801047120418846,
+      "grad_norm": 2.44124698638916,
+      "learning_rate": 4.264944662328947e-06,
+      "loss": 5.4314,
+      "step": 385
+    },
+    {
+      "epoch": 0.2586806282722513,
+      "grad_norm": 1.9755254983901978,
+      "learning_rate": 4.261174622596835e-06,
+      "loss": 5.2068,
+      "step": 386
+    },
+    {
+      "epoch": 0.25935078534031414,
+      "grad_norm": 1.844270944595337,
+      "learning_rate": 4.257396615017285e-06,
+      "loss": 5.3666,
+      "step": 387
+    },
+    {
+      "epoch": 0.260020942408377,
+      "grad_norm": 2.3097944259643555,
+      "learning_rate": 4.253610656682633e-06,
+      "loss": 5.3687,
+      "step": 388
+    },
+    {
+      "epoch": 0.2606910994764398,
+      "grad_norm": 2.1606905460357666,
+      "learning_rate": 4.2498167647211865e-06,
+      "loss": 5.3283,
+      "step": 389
+    },
+    {
+      "epoch": 0.2613612565445026,
+      "grad_norm": 2.251303195953369,
+      "learning_rate": 4.246014956297142e-06,
+      "loss": 5.4695,
+      "step": 390
+    },
+    {
+      "epoch": 0.26203141361256543,
+      "grad_norm": 2.2528789043426514,
+      "learning_rate": 4.242205248610518e-06,
+      "loss": 5.4339,
+      "step": 391
+    },
+    {
+      "epoch": 0.26270157068062827,
+      "grad_norm": 1.8444974422454834,
+      "learning_rate": 4.238387658897063e-06,
+      "loss": 5.5597,
+      "step": 392
+    },
+    {
+      "epoch": 0.2633717277486911,
+      "grad_norm": 2.023601531982422,
+      "learning_rate": 4.234562204428192e-06,
+      "loss": 5.4275,
+      "step": 393
+    },
+    {
+      "epoch": 0.26404188481675395,
+      "grad_norm": 2.0824105739593506,
+      "learning_rate": 4.230728902510896e-06,
+      "loss": 5.3526,
+      "step": 394
+    },
+    {
+      "epoch": 0.26471204188481673,
+      "grad_norm": 2.2396888732910156,
+      "learning_rate": 4.226887770487675e-06,
+      "loss": 5.5526,
+      "step": 395
+    },
+    {
+      "epoch": 0.26538219895287957,
+      "grad_norm": 2.153627395629883,
+      "learning_rate": 4.223038825736446e-06,
+      "loss": 5.3724,
+      "step": 396
+    },
+    {
+      "epoch": 0.2660523560209424,
+      "grad_norm": 2.5046567916870117,
+      "learning_rate": 4.2191820856704786e-06,
+      "loss": 5.426,
+      "step": 397
+    },
+    {
+      "epoch": 0.26672251308900524,
+      "grad_norm": 2.2566487789154053,
+      "learning_rate": 4.215317567738308e-06,
+      "loss": 5.2797,
+      "step": 398
+    },
+    {
+      "epoch": 0.2673926701570681,
+      "grad_norm": 2.192222833633423,
+      "learning_rate": 4.2114452894236565e-06,
+      "loss": 5.2759,
+      "step": 399
+    },
+    {
+      "epoch": 0.2680628272251309,
+      "grad_norm": 2.016427993774414,
+      "learning_rate": 4.207565268245356e-06,
+      "loss": 5.5729,
+      "step": 400
+    },
+    {
+      "epoch": 0.2687329842931937,
+      "grad_norm": 1.8648808002471924,
+      "learning_rate": 4.203677521757269e-06,
+      "loss": 5.6048,
+      "step": 401
+    },
+    {
+      "epoch": 0.26940314136125654,
+      "grad_norm": 2.325082778930664,
+      "learning_rate": 4.1997820675482085e-06,
+      "loss": 5.3213,
+      "step": 402
+    },
+    {
+      "epoch": 0.2700732984293194,
+      "grad_norm": 2.05472469329834,
+      "learning_rate": 4.19587892324186e-06,
+      "loss": 5.2503,
+      "step": 403
+    },
+    {
+      "epoch": 0.2707434554973822,
+      "grad_norm": 2.525400400161743,
+      "learning_rate": 4.191968106496696e-06,
+      "loss": 5.3963,
+      "step": 404
+    },
+    {
+      "epoch": 0.27141361256544505,
+      "grad_norm": 2.3485240936279297,
+      "learning_rate": 4.188049635005904e-06,
+      "loss": 5.4384,
+      "step": 405
+    },
+    {
+      "epoch": 0.27208376963350783,
+      "grad_norm": 2.287038803100586,
+      "learning_rate": 4.184123526497303e-06,
+      "loss": 5.2257,
+      "step": 406
+    },
+    {
+      "epoch": 0.27275392670157067,
+      "grad_norm": 2.2160794734954834,
+      "learning_rate": 4.180189798733259e-06,
+      "loss": 5.2822,
+      "step": 407
+    },
+    {
+      "epoch": 0.2734240837696335,
+      "grad_norm": 2.236074209213257,
+      "learning_rate": 4.176248469510616e-06,
+      "loss": 5.1092,
+      "step": 408
+    },
+    {
+      "epoch": 0.27409424083769635,
+      "grad_norm": 2.395686149597168,
+      "learning_rate": 4.172299556660601e-06,
+      "loss": 5.2411,
+      "step": 409
+    },
+    {
+      "epoch": 0.2747643979057592,
+      "grad_norm": 2.5743703842163086,
+      "learning_rate": 4.168343078048756e-06,
+      "loss": 5.4214,
+      "step": 410
+    },
+    {
+      "epoch": 0.27543455497382197,
+      "grad_norm": 2.148124933242798,
+      "learning_rate": 4.164379051574848e-06,
+      "loss": 5.258,
+      "step": 411
+    },
+    {
+      "epoch": 0.2761047120418848,
+      "grad_norm": 1.816332459449768,
+      "learning_rate": 4.160407495172794e-06,
+      "loss": 5.2502,
+      "step": 412
+    },
+    {
+      "epoch": 0.27677486910994764,
+      "grad_norm": 2.5482685565948486,
+      "learning_rate": 4.156428426810576e-06,
+      "loss": 5.2039,
+      "step": 413
+    },
+    {
+      "epoch": 0.2774450261780105,
+      "grad_norm": 2.4875996112823486,
+      "learning_rate": 4.152441864490162e-06,
+      "loss": 5.2924,
+      "step": 414
+    },
+    {
+      "epoch": 0.2781151832460733,
+      "grad_norm": 2.5088934898376465,
+      "learning_rate": 4.1484478262474255e-06,
+      "loss": 5.3185,
+      "step": 415
+    },
+    {
+      "epoch": 0.2787853403141361,
+      "grad_norm": 2.1406543254852295,
+      "learning_rate": 4.144446330152061e-06,
+      "loss": 5.3695,
+      "step": 416
+    },
+    {
+      "epoch": 0.27945549738219894,
+      "grad_norm": 2.8924715518951416,
+      "learning_rate": 4.140437394307502e-06,
+      "loss": 5.4743,
+      "step": 417
+    },
+    {
+      "epoch": 0.2801256544502618,
+      "grad_norm": 2.35742449760437,
+      "learning_rate": 4.136421036850844e-06,
+      "loss": 5.259,
+      "step": 418
+    },
+    {
+      "epoch": 0.2807958115183246,
+      "grad_norm": 2.7321271896362305,
+      "learning_rate": 4.132397275952756e-06,
+      "loss": 5.4331,
+      "step": 419
+    },
+    {
+      "epoch": 0.28146596858638745,
+      "grad_norm": 2.255457878112793,
+      "learning_rate": 4.128366129817403e-06,
+      "loss": 5.6379,
+      "step": 420
+    },
+    {
+      "epoch": 0.2821361256544503,
+      "grad_norm": 2.1841540336608887,
+      "learning_rate": 4.124327616682362e-06,
+      "loss": 5.4079,
+      "step": 421
+    },
+    {
+      "epoch": 0.28280628272251307,
+      "grad_norm": 1.8602941036224365,
+      "learning_rate": 4.12028175481854e-06,
+      "loss": 5.156,
+      "step": 422
+    },
+    {
+      "epoch": 0.2834764397905759,
+      "grad_norm": 4.119997024536133,
+      "learning_rate": 4.116228562530089e-06,
+      "loss": 5.3936,
+      "step": 423
+    },
+    {
+      "epoch": 0.28414659685863874,
+      "grad_norm": 2.1611454486846924,
+      "learning_rate": 4.112168058154327e-06,
+      "loss": 5.5529,
+      "step": 424
+    },
+    {
+      "epoch": 0.2848167539267016,
+      "grad_norm": 2.1483073234558105,
+      "learning_rate": 4.108100260061652e-06,
+      "loss": 5.1862,
+      "step": 425
+    },
+    {
+      "epoch": 0.2854869109947644,
+      "grad_norm": 1.9627047777175903,
+      "learning_rate": 4.10402518665546e-06,
+      "loss": 5.278,
+      "step": 426
+    },
+    {
+      "epoch": 0.2861570680628272,
+      "grad_norm": 2.049126386642456,
+      "learning_rate": 4.099942856372064e-06,
+      "loss": 5.1944,
+      "step": 427
+    },
+    {
+      "epoch": 0.28682722513089004,
+      "grad_norm": 2.7416749000549316,
+      "learning_rate": 4.095853287680604e-06,
+      "loss": 5.2854,
+      "step": 428
+    },
+    {
+      "epoch": 0.2874973821989529,
+      "grad_norm": 2.39632511138916,
+      "learning_rate": 4.0917564990829705e-06,
+      "loss": 5.5541,
+      "step": 429
+    },
+    {
+      "epoch": 0.2881675392670157,
+      "grad_norm": 2.582184076309204,
+      "learning_rate": 4.0876525091137195e-06,
+      "loss": 5.3696,
+      "step": 430
+    },
+    {
+      "epoch": 0.28883769633507855,
+      "grad_norm": 2.024606227874756,
+      "learning_rate": 4.083541336339984e-06,
+      "loss": 5.2294,
+      "step": 431
+    },
+    {
+      "epoch": 0.28950785340314134,
+      "grad_norm": 2.250382423400879,
+      "learning_rate": 4.079422999361395e-06,
+      "loss": 5.4614,
+      "step": 432
+    },
+    {
+      "epoch": 0.2901780104712042,
+      "grad_norm": 2.6855008602142334,
+      "learning_rate": 4.075297516809994e-06,
+      "loss": 5.3138,
+      "step": 433
+    },
+    {
+      "epoch": 0.290848167539267,
+      "grad_norm": 2.4782090187072754,
+      "learning_rate": 4.071164907350153e-06,
+      "loss": 5.3277,
+      "step": 434
+    },
+    {
+      "epoch": 0.29151832460732985,
+      "grad_norm": 2.666271448135376,
+      "learning_rate": 4.067025189678485e-06,
+      "loss": 5.3539,
+      "step": 435
+    },
+    {
+      "epoch": 0.2921884816753927,
+      "grad_norm": 2.2352638244628906,
+      "learning_rate": 4.062878382523764e-06,
+      "loss": 5.2922,
+      "step": 436
+    },
+    {
+      "epoch": 0.2928586387434555,
+      "grad_norm": 2.0030295848846436,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 5.2213,
+      "step": 437
+    },
+    {
+      "epoch": 0.2935287958115183,
+      "grad_norm": 2.5756025314331055,
+      "learning_rate": 4.054563574840532e-06,
+      "loss": 5.4836,
+      "step": 438
+    },
+    {
+      "epoch": 0.29419895287958114,
+      "grad_norm": 2.685851573944092,
+      "learning_rate": 4.050395611929599e-06,
+      "loss": 5.1647,
+      "step": 439
+    },
+    {
+      "epoch": 0.294869109947644,
+      "grad_norm": 2.5466887950897217,
+      "learning_rate": 4.046220634770591e-06,
+      "loss": 5.3462,
+      "step": 440
+    },
+    {
+      "epoch": 0.2955392670157068,
+      "grad_norm": 2.186382293701172,
+      "learning_rate": 4.042038662251801e-06,
+      "loss": 5.2909,
+      "step": 441
+    },
+    {
+      "epoch": 0.29620942408376966,
+      "grad_norm": 2.2794177532196045,
+      "learning_rate": 4.0378497132931724e-06,
+      "loss": 5.347,
+      "step": 442
+    },
+    {
+      "epoch": 0.29687958115183244,
+      "grad_norm": 2.45033860206604,
+      "learning_rate": 4.033653806846204e-06,
+      "loss": 5.5143,
+      "step": 443
+    },
+    {
+      "epoch": 0.2975497382198953,
+      "grad_norm": 1.970142126083374,
+      "learning_rate": 4.0294509618938785e-06,
+      "loss": 5.1962,
+      "step": 444
+    },
+    {
+      "epoch": 0.2982198952879581,
+      "grad_norm": 1.9088056087493896,
+      "learning_rate": 4.025241197450566e-06,
+      "loss": 5.2455,
+      "step": 445
+    },
+    {
+      "epoch": 0.29889005235602095,
+      "grad_norm": 1.9862786531448364,
+      "learning_rate": 4.021024532561941e-06,
+      "loss": 5.3017,
+      "step": 446
+    },
+    {
+      "epoch": 0.2995602094240838,
+      "grad_norm": 1.6905484199523926,
+      "learning_rate": 4.0168009863049e-06,
+      "loss": 5.4328,
+      "step": 447
+    },
+    {
+      "epoch": 0.30023036649214657,
+      "grad_norm": 2.0439505577087402,
+      "learning_rate": 4.012570577787469e-06,
+      "loss": 5.1759,
+      "step": 448
+    },
+    {
+      "epoch": 0.3009005235602094,
+      "grad_norm": 2.0029120445251465,
+      "learning_rate": 4.008333326148721e-06,
+      "loss": 5.1908,
+      "step": 449
+    },
+    {
+      "epoch": 0.30157068062827225,
+      "grad_norm": 2.239750385284424,
+      "learning_rate": 4.004089250558688e-06,
+      "loss": 5.2547,
+      "step": 450
+    },
+    {
+      "epoch": 0.3022408376963351,
+      "grad_norm": 2.2577266693115234,
+      "learning_rate": 3.9998383702182764e-06,
+      "loss": 5.3508,
+      "step": 451
+    },
+    {
+      "epoch": 0.3029109947643979,
+      "grad_norm": 2.021261692047119,
+      "learning_rate": 3.9955807043591765e-06,
+      "loss": 5.4991,
+      "step": 452
+    },
+    {
+      "epoch": 0.30358115183246076,
+      "grad_norm": 2.3646843433380127,
+      "learning_rate": 3.991316272243778e-06,
+      "loss": 5.5244,
+      "step": 453
+    },
+    {
+      "epoch": 0.30425130890052354,
+      "grad_norm": 2.3185791969299316,
+      "learning_rate": 3.987045093165083e-06,
+      "loss": 5.5034,
+      "step": 454
+    },
+    {
+      "epoch": 0.3049214659685864,
+      "grad_norm": 2.2033495903015137,
+      "learning_rate": 3.982767186446615e-06,
+      "loss": 5.4536,
+      "step": 455
+    },
+    {
+      "epoch": 0.3055916230366492,
+      "grad_norm": 2.128403663635254,
+      "learning_rate": 3.978482571442339e-06,
+      "loss": 5.3952,
+      "step": 456
+    },
+    {
+      "epoch": 0.30626178010471206,
+      "grad_norm": 2.455321788787842,
+      "learning_rate": 3.974191267536567e-06,
+      "loss": 5.3201,
+      "step": 457
+    },
+    {
+      "epoch": 0.3069319371727749,
+      "grad_norm": 2.422396659851074,
+      "learning_rate": 3.9698932941438714e-06,
+      "loss": 5.3798,
+      "step": 458
+    },
+    {
+      "epoch": 0.3076020942408377,
+      "grad_norm": 2.6799330711364746,
+      "learning_rate": 3.965588670709002e-06,
+      "loss": 5.2903,
+      "step": 459
+    },
+    {
+      "epoch": 0.3082722513089005,
+      "grad_norm": 2.433835506439209,
+      "learning_rate": 3.9612774167067905e-06,
+      "loss": 5.4659,
+      "step": 460
+    },
+    {
+      "epoch": 0.30894240837696335,
+      "grad_norm": 2.2600831985473633,
+      "learning_rate": 3.95695955164207e-06,
+      "loss": 5.4558,
+      "step": 461
+    },
+    {
+      "epoch": 0.3096125654450262,
+      "grad_norm": 2.3049960136413574,
+      "learning_rate": 3.95263509504958e-06,
+      "loss": 5.5923,
+      "step": 462
+    },
+    {
+      "epoch": 0.310282722513089,
+      "grad_norm": 2.1892147064208984,
+      "learning_rate": 3.948304066493885e-06,
+      "loss": 5.5232,
+      "step": 463
+    },
+    {
+      "epoch": 0.3109528795811518,
+      "grad_norm": 2.799586057662964,
+      "learning_rate": 3.9439664855692774e-06,
+      "loss": 5.2964,
+      "step": 464
+    },
+    {
+      "epoch": 0.31162303664921465,
+      "grad_norm": 2.493480682373047,
+      "learning_rate": 3.939622371899697e-06,
+      "loss": 5.489,
+      "step": 465
+    },
+    {
+      "epoch": 0.3122931937172775,
+      "grad_norm": 2.2202532291412354,
+      "learning_rate": 3.935271745138637e-06,
+      "loss": 5.3784,
+      "step": 466
+    },
+    {
+      "epoch": 0.3129633507853403,
+      "grad_norm": 1.9341351985931396,
+      "learning_rate": 3.9309146249690595e-06,
+      "loss": 5.3933,
+      "step": 467
+    },
+    {
+      "epoch": 0.31363350785340316,
+      "grad_norm": 1.8867202997207642,
+      "learning_rate": 3.9265510311033005e-06,
+      "loss": 5.3793,
+      "step": 468
+    },
+    {
+      "epoch": 0.31430366492146594,
+      "grad_norm": 2.1868042945861816,
+      "learning_rate": 3.9221809832829866e-06,
+      "loss": 5.2335,
+      "step": 469
+    },
+    {
+      "epoch": 0.3149738219895288,
+      "grad_norm": 2.428607940673828,
+      "learning_rate": 3.917804501278942e-06,
+      "loss": 5.6313,
+      "step": 470
+    },
+    {
+      "epoch": 0.3156439790575916,
+      "grad_norm": 2.402775287628174,
+      "learning_rate": 3.9134216048911e-06,
+      "loss": 5.4608,
+      "step": 471
+    },
+    {
+      "epoch": 0.31631413612565445,
+      "grad_norm": 1.9883973598480225,
+      "learning_rate": 3.909032313948415e-06,
+      "loss": 5.3552,
+      "step": 472
+    },
+    {
+      "epoch": 0.3169842931937173,
+      "grad_norm": 2.1537022590637207,
+      "learning_rate": 3.90463664830877e-06,
+      "loss": 5.3092,
+      "step": 473
+    },
+    {
+      "epoch": 0.31765445026178013,
+      "grad_norm": 2.195376396179199,
+      "learning_rate": 3.90023462785889e-06,
+      "loss": 5.1523,
+      "step": 474
+    },
+    {
+      "epoch": 0.3183246073298429,
+      "grad_norm": 1.9080901145935059,
+      "learning_rate": 3.895826272514247e-06,
+      "loss": 5.1277,
+      "step": 475
+    },
+    {
+      "epoch": 0.31899476439790575,
+      "grad_norm": 2.5250730514526367,
+      "learning_rate": 3.891411602218976e-06,
+      "loss": 5.3874,
+      "step": 476
+    },
+    {
+      "epoch": 0.3196649214659686,
+      "grad_norm": 2.1713311672210693,
+      "learning_rate": 3.8869906369457815e-06,
+      "loss": 5.1199,
+      "step": 477
+    },
+    {
+      "epoch": 0.3203350785340314,
+      "grad_norm": 2.3029656410217285,
+      "learning_rate": 3.882563396695846e-06,
+      "loss": 5.2659,
+      "step": 478
+    },
+    {
+      "epoch": 0.32100523560209426,
+      "grad_norm": 2.2043564319610596,
+      "learning_rate": 3.878129901498742e-06,
+      "loss": 5.2733,
+      "step": 479
+    },
+    {
+      "epoch": 0.32167539267015705,
+      "grad_norm": 2.431480884552002,
+      "learning_rate": 3.8736901714123415e-06,
+      "loss": 5.2519,
+      "step": 480
+    },
+    {
+      "epoch": 0.3223455497382199,
+      "grad_norm": 2.155893325805664,
+      "learning_rate": 3.869244226522723e-06,
+      "loss": 5.4178,
+      "step": 481
+    },
+    {
+      "epoch": 0.3230157068062827,
+      "grad_norm": 2.1351659297943115,
+      "learning_rate": 3.864792086944081e-06,
+      "loss": 5.5612,
+      "step": 482
+    },
+    {
+      "epoch": 0.32368586387434556,
+      "grad_norm": 2.427992105484009,
+      "learning_rate": 3.860333772818639e-06,
+      "loss": 5.2741,
+      "step": 483
+    },
+    {
+      "epoch": 0.3243560209424084,
+      "grad_norm": 2.6776323318481445,
+      "learning_rate": 3.85586930431655e-06,
+      "loss": 5.4813,
+      "step": 484
+    },
+    {
+      "epoch": 0.3250261780104712,
+      "grad_norm": 2.2945733070373535,
+      "learning_rate": 3.851398701635815e-06,
+      "loss": 5.5534,
+      "step": 485
+    },
+    {
+      "epoch": 0.325696335078534,
+      "grad_norm": 2.452486515045166,
+      "learning_rate": 3.8469219850021845e-06,
+      "loss": 5.3351,
+      "step": 486
+    },
+    {
+      "epoch": 0.32636649214659685,
+      "grad_norm": 2.1009156703948975,
+      "learning_rate": 3.842439174669072e-06,
+      "loss": 4.9801,
+      "step": 487
+    },
+    {
+      "epoch": 0.3270366492146597,
+      "grad_norm": 2.837606906890869,
+      "learning_rate": 3.837950290917457e-06,
+      "loss": 5.4142,
+      "step": 488
+    },
+    {
+      "epoch": 0.32770680628272253,
+      "grad_norm": 1.9545905590057373,
+      "learning_rate": 3.8334553540557986e-06,
+      "loss": 4.9649,
+      "step": 489
+    },
+    {
+      "epoch": 0.32837696335078537,
+      "grad_norm": 1.956103801727295,
+      "learning_rate": 3.828954384419939e-06,
+      "loss": 5.4059,
+      "step": 490
+    },
+    {
+      "epoch": 0.32904712041884815,
+      "grad_norm": 2.232672691345215,
+      "learning_rate": 3.8244474023730155e-06,
+      "loss": 5.074,
+      "step": 491
+    },
+    {
+      "epoch": 0.329717277486911,
+      "grad_norm": 2.234177827835083,
+      "learning_rate": 3.819934428305365e-06,
+      "loss": 5.3179,
+      "step": 492
+    },
+    {
+      "epoch": 0.3303874345549738,
+      "grad_norm": 1.9164519309997559,
+      "learning_rate": 3.815415482634436e-06,
+      "loss": 5.2802,
+      "step": 493
+    },
+    {
+      "epoch": 0.33105759162303666,
+      "grad_norm": 1.753132700920105,
+      "learning_rate": 3.8108905858046896e-06,
+      "loss": 5.1147,
+      "step": 494
+    },
+    {
+      "epoch": 0.3317277486910995,
+      "grad_norm": 2.30802845954895,
+      "learning_rate": 3.8063597582875135e-06,
+      "loss": 5.4144,
+      "step": 495
+    },
+    {
+      "epoch": 0.3323979057591623,
+      "grad_norm": 1.794412612915039,
+      "learning_rate": 3.8018230205811267e-06,
+      "loss": 5.255,
+      "step": 496
+    },
+    {
+      "epoch": 0.3330680628272251,
+      "grad_norm": 2.070859670639038,
+      "learning_rate": 3.797280393210487e-06,
+      "loss": 5.4724,
+      "step": 497
+    },
+    {
+      "epoch": 0.33373821989528796,
+      "grad_norm": 2.355644464492798,
+      "learning_rate": 3.7927318967271966e-06,
+      "loss": 5.443,
+      "step": 498
+    },
+    {
+      "epoch": 0.3344083769633508,
+      "grad_norm": 2.239596128463745,
+      "learning_rate": 3.788177551709411e-06,
+      "loss": 5.1885,
+      "step": 499
+    },
+    {
+      "epoch": 0.33507853403141363,
+      "grad_norm": 2.1752090454101562,
+      "learning_rate": 3.7836173787617475e-06,
+      "loss": 5.3933,
+      "step": 500
+    },
+    {
+      "epoch": 0.3357486910994764,
+      "grad_norm": 2.6792967319488525,
+      "learning_rate": 3.779051398515188e-06,
+      "loss": 5.2386,
+      "step": 501
+    },
+    {
+      "epoch": 0.33641884816753925,
+      "grad_norm": 2.3228695392608643,
+      "learning_rate": 3.774479631626986e-06,
+      "loss": 5.3472,
+      "step": 502
+    },
+    {
+      "epoch": 0.3370890052356021,
+      "grad_norm": 2.5239968299865723,
+      "learning_rate": 3.7699020987805784e-06,
+      "loss": 5.2552,
+      "step": 503
+    },
+    {
+      "epoch": 0.33775916230366493,
+      "grad_norm": 2.4592580795288086,
+      "learning_rate": 3.7653188206854867e-06,
+      "loss": 5.211,
+      "step": 504
+    },
+    {
+      "epoch": 0.33842931937172777,
+      "grad_norm": 2.3289928436279297,
+      "learning_rate": 3.760729818077224e-06,
+      "loss": 5.3682,
+      "step": 505
+    },
+    {
+      "epoch": 0.3390994764397906,
+      "grad_norm": 2.0905239582061768,
+      "learning_rate": 3.756135111717202e-06,
+      "loss": 5.3095,
+      "step": 506
+    },
+    {
+      "epoch": 0.3397696335078534,
+      "grad_norm": 2.084735870361328,
+      "learning_rate": 3.75153472239264e-06,
+      "loss": 5.0262,
+      "step": 507
+    },
+    {
+      "epoch": 0.3404397905759162,
+      "grad_norm": 2.0062642097473145,
+      "learning_rate": 3.746928670916464e-06,
+      "loss": 5.3134,
+      "step": 508
+    },
+    {
+      "epoch": 0.34110994764397906,
+      "grad_norm": 2.108950138092041,
+      "learning_rate": 3.7423169781272177e-06,
+      "loss": 5.439,
+      "step": 509
+    },
+    {
+      "epoch": 0.3417801047120419,
+      "grad_norm": 2.0042457580566406,
+      "learning_rate": 3.7376996648889703e-06,
+      "loss": 5.5338,
+      "step": 510
+    },
+    {
+      "epoch": 0.34245026178010474,
+      "grad_norm": 2.4976985454559326,
+      "learning_rate": 3.733076752091214e-06,
+      "loss": 5.3151,
+      "step": 511
+    },
+    {
+      "epoch": 0.3431204188481675,
+      "grad_norm": 2.1041195392608643,
+      "learning_rate": 3.728448260648777e-06,
+      "loss": 5.2065,
+      "step": 512
+    },
+    {
+      "epoch": 0.34379057591623036,
+      "grad_norm": 2.285289764404297,
+      "learning_rate": 3.7238142115017262e-06,
+      "loss": 5.2395,
+      "step": 513
+    },
+    {
+      "epoch": 0.3444607329842932,
+      "grad_norm": 2.4077186584472656,
+      "learning_rate": 3.719174625615273e-06,
+      "loss": 5.1368,
+      "step": 514
+    },
+    {
+      "epoch": 0.34513089005235603,
+      "grad_norm": 2.0603976249694824,
+      "learning_rate": 3.7145295239796747e-06,
+      "loss": 5.273,
+      "step": 515
+    },
+    {
+      "epoch": 0.34580104712041887,
+      "grad_norm": 1.9497524499893188,
+      "learning_rate": 3.709878927610146e-06,
+      "loss": 5.2622,
+      "step": 516
+    },
+    {
+      "epoch": 0.34647120418848165,
+      "grad_norm": 2.6009366512298584,
+      "learning_rate": 3.705222857546761e-06,
+      "loss": 5.48,
+      "step": 517
+    },
+    {
+      "epoch": 0.3471413612565445,
+      "grad_norm": 2.330073833465576,
+      "learning_rate": 3.700561334854355e-06,
+      "loss": 5.2746,
+      "step": 518
+    },
+    {
+      "epoch": 0.3478115183246073,
+      "grad_norm": 2.2093427181243896,
+      "learning_rate": 3.6958943806224333e-06,
+      "loss": 5.4129,
+      "step": 519
+    },
+    {
+      "epoch": 0.34848167539267016,
+      "grad_norm": 2.3132288455963135,
+      "learning_rate": 3.6912220159650757e-06,
+      "loss": 5.3603,
+      "step": 520
+    },
+    {
+      "epoch": 0.349151832460733,
+      "grad_norm": 2.2647898197174072,
+      "learning_rate": 3.686544262020838e-06,
+      "loss": 5.294,
+      "step": 521
+    },
+    {
+      "epoch": 0.3498219895287958,
+      "grad_norm": 2.105992317199707,
+      "learning_rate": 3.6818611399526576e-06,
+      "loss": 5.2418,
+      "step": 522
+    },
+    {
+      "epoch": 0.3504921465968586,
+      "grad_norm": 2.493412494659424,
+      "learning_rate": 3.6771726709477594e-06,
+      "loss": 5.1907,
+      "step": 523
+    },
+    {
+      "epoch": 0.35116230366492146,
+      "grad_norm": 2.1870672702789307,
+      "learning_rate": 3.6724788762175587e-06,
+      "loss": 5.2098,
+      "step": 524
+    },
+    {
+      "epoch": 0.3518324607329843,
+      "grad_norm": 2.307648181915283,
+      "learning_rate": 3.6677797769975653e-06,
+      "loss": 5.3077,
+      "step": 525
+    },
+    {
+      "epoch": 0.35250261780104714,
+      "grad_norm": 2.2626419067382812,
+      "learning_rate": 3.6630753945472854e-06,
+      "loss": 5.1255,
+      "step": 526
+    },
+    {
+      "epoch": 0.35317277486911,
+      "grad_norm": 2.2895169258117676,
+      "learning_rate": 3.658365750150131e-06,
+      "loss": 5.3941,
+      "step": 527
+    },
+    {
+      "epoch": 0.35384293193717276,
+      "grad_norm": 2.3610305786132812,
+      "learning_rate": 3.653650865113317e-06,
+      "loss": 5.4692,
+      "step": 528
+    },
+    {
+      "epoch": 0.3545130890052356,
+      "grad_norm": 2.173130750656128,
+      "learning_rate": 3.6489307607677694e-06,
+      "loss": 5.1767,
+      "step": 529
+    },
+    {
+      "epoch": 0.35518324607329843,
+      "grad_norm": 2.2770626544952393,
+      "learning_rate": 3.644205458468026e-06,
+      "loss": 5.5926,
+      "step": 530
+    },
+    {
+      "epoch": 0.35585340314136127,
+      "grad_norm": 2.6587777137756348,
+      "learning_rate": 3.6394749795921425e-06,
+      "loss": 5.3162,
+      "step": 531
+    },
+    {
+      "epoch": 0.3565235602094241,
+      "grad_norm": 2.2764198780059814,
+      "learning_rate": 3.6347393455415925e-06,
+      "loss": 5.4997,
+      "step": 532
+    },
+    {
+      "epoch": 0.3571937172774869,
+      "grad_norm": 2.466233491897583,
+      "learning_rate": 3.6299985777411744e-06,
+      "loss": 5.3123,
+      "step": 533
+    },
+    {
+      "epoch": 0.3578638743455497,
+      "grad_norm": 2.2345104217529297,
+      "learning_rate": 3.625252697638912e-06,
+      "loss": 5.2243,
+      "step": 534
+    },
+    {
+      "epoch": 0.35853403141361256,
+      "grad_norm": 1.9270764589309692,
+      "learning_rate": 3.6205017267059567e-06,
+      "loss": 5.11,
+      "step": 535
+    },
+    {
+      "epoch": 0.3592041884816754,
+      "grad_norm": 2.272279977798462,
+      "learning_rate": 3.6157456864364925e-06,
+      "loss": 5.2157,
+      "step": 536
+    },
+    {
+      "epoch": 0.35987434554973824,
+      "grad_norm": 2.258368968963623,
+      "learning_rate": 3.6109845983476393e-06,
+      "loss": 5.117,
+      "step": 537
+    },
+    {
+      "epoch": 0.360544502617801,
+      "grad_norm": 2.28820538520813,
+      "learning_rate": 3.606218483979352e-06,
+      "loss": 5.4147,
+      "step": 538
+    },
+    {
+      "epoch": 0.36121465968586386,
+      "grad_norm": 2.199136734008789,
+      "learning_rate": 3.601447364894326e-06,
+      "loss": 5.4522,
+      "step": 539
+    },
+    {
+      "epoch": 0.3618848167539267,
+      "grad_norm": 2.169766664505005,
+      "learning_rate": 3.596671262677898e-06,
+      "loss": 5.2305,
+      "step": 540
+    },
+    {
+      "epoch": 0.36255497382198953,
+      "grad_norm": 2.2157297134399414,
+      "learning_rate": 3.5918901989379527e-06,
+      "loss": 5.3279,
+      "step": 541
+    },
+    {
+      "epoch": 0.3632251308900524,
+      "grad_norm": 2.254014015197754,
+      "learning_rate": 3.587104195304818e-06,
+      "loss": 5.0062,
+      "step": 542
+    },
+    {
+      "epoch": 0.3638952879581152,
+      "grad_norm": 2.3738510608673096,
+      "learning_rate": 3.5823132734311704e-06,
+      "loss": 5.7765,
+      "step": 543
+    },
+    {
+      "epoch": 0.364565445026178,
+      "grad_norm": 2.2331602573394775,
+      "learning_rate": 3.5775174549919413e-06,
+      "loss": 5.5786,
+      "step": 544
+    },
+    {
+      "epoch": 0.36523560209424083,
+      "grad_norm": 2.5791873931884766,
+      "learning_rate": 3.57271676168421e-06,
+      "loss": 5.3063,
+      "step": 545
+    },
+    {
+      "epoch": 0.36590575916230367,
+      "grad_norm": 2.402435779571533,
+      "learning_rate": 3.5679112152271144e-06,
+      "loss": 5.2438,
+      "step": 546
+    },
+    {
+      "epoch": 0.3665759162303665,
+      "grad_norm": 2.1838717460632324,
+      "learning_rate": 3.5631008373617464e-06,
+      "loss": 5.3002,
+      "step": 547
+    },
+    {
+      "epoch": 0.36724607329842934,
+      "grad_norm": 2.4075636863708496,
+      "learning_rate": 3.5582856498510574e-06,
+      "loss": 5.3944,
+      "step": 548
+    },
+    {
+      "epoch": 0.3679162303664921,
+      "grad_norm": 2.2561862468719482,
+      "learning_rate": 3.5534656744797583e-06,
+      "loss": 5.3127,
+      "step": 549
+    },
+    {
+      "epoch": 0.36858638743455496,
+      "grad_norm": 2.3960776329040527,
+      "learning_rate": 3.54864093305422e-06,
+      "loss": 5.147,
+      "step": 550
+    },
+    {
+      "epoch": 0.3692565445026178,
+      "grad_norm": 2.1438934803009033,
+      "learning_rate": 3.543811447402378e-06,
+      "loss": 5.2816,
+      "step": 551
+    },
+    {
+      "epoch": 0.36992670157068064,
+      "grad_norm": 1.8878827095031738,
+      "learning_rate": 3.538977239373629e-06,
+      "loss": 5.3005,
+      "step": 552
+    },
+    {
+      "epoch": 0.3705968586387435,
+      "grad_norm": 2.35463809967041,
+      "learning_rate": 3.534138330838735e-06,
+      "loss": 5.2632,
+      "step": 553
+    },
+    {
+      "epoch": 0.37126701570680626,
+      "grad_norm": 2.1728053092956543,
+      "learning_rate": 3.5292947436897257e-06,
+      "loss": 5.1822,
+      "step": 554
+    },
+    {
+      "epoch": 0.3719371727748691,
+      "grad_norm": 2.3696134090423584,
+      "learning_rate": 3.5244464998397958e-06,
+      "loss": 5.3895,
+      "step": 555
+    },
+    {
+      "epoch": 0.37260732984293193,
+      "grad_norm": 2.2455461025238037,
+      "learning_rate": 3.5195936212232086e-06,
+      "loss": 5.4655,
+      "step": 556
+    },
+    {
+      "epoch": 0.37327748691099477,
+      "grad_norm": 2.0229246616363525,
+      "learning_rate": 3.5147361297951955e-06,
+      "loss": 5.0173,
+      "step": 557
+    },
+    {
+      "epoch": 0.3739476439790576,
+      "grad_norm": 2.8633840084075928,
+      "learning_rate": 3.509874047531856e-06,
+      "loss": 5.3669,
+      "step": 558
+    },
+    {
+      "epoch": 0.3746178010471204,
+      "grad_norm": 2.104095697402954,
+      "learning_rate": 3.5050073964300624e-06,
+      "loss": 5.1716,
+      "step": 559
+    },
+    {
+      "epoch": 0.37528795811518323,
+      "grad_norm": 2.1818394660949707,
+      "learning_rate": 3.5001361985073532e-06,
+      "loss": 5.327,
+      "step": 560
+    },
+    {
+      "epoch": 0.37595811518324607,
+      "grad_norm": 2.133225679397583,
+      "learning_rate": 3.495260475801841e-06,
+      "loss": 5.4634,
+      "step": 561
+    },
+    {
+      "epoch": 0.3766282722513089,
+      "grad_norm": 2.2934346199035645,
+      "learning_rate": 3.490380250372107e-06,
+      "loss": 5.3697,
+      "step": 562
+    },
+    {
+      "epoch": 0.37729842931937174,
+      "grad_norm": 2.2782511711120605,
+      "learning_rate": 3.485495544297104e-06,
+      "loss": 5.4309,
+      "step": 563
+    },
+    {
+      "epoch": 0.3779685863874346,
+      "grad_norm": 2.150676965713501,
+      "learning_rate": 3.4806063796760557e-06,
+      "loss": 5.2855,
+      "step": 564
+    },
+    {
+      "epoch": 0.37863874345549736,
+      "grad_norm": 2.218550682067871,
+      "learning_rate": 3.4757127786283583e-06,
+      "loss": 5.0748,
+      "step": 565
+    },
+    {
+      "epoch": 0.3793089005235602,
+      "grad_norm": 1.859222173690796,
+      "learning_rate": 3.4708147632934793e-06,
+      "loss": 5.3948,
+      "step": 566
+    },
+    {
+      "epoch": 0.37997905759162304,
+      "grad_norm": 2.387467384338379,
+      "learning_rate": 3.465912355830853e-06,
+      "loss": 5.0676,
+      "step": 567
+    },
+    {
+      "epoch": 0.3806492146596859,
+      "grad_norm": 2.3780243396759033,
+      "learning_rate": 3.4610055784197917e-06,
+      "loss": 5.4155,
+      "step": 568
+    },
+    {
+      "epoch": 0.3813193717277487,
+      "grad_norm": 2.393306255340576,
+      "learning_rate": 3.456094453259371e-06,
+      "loss": 5.2968,
+      "step": 569
+    },
+    {
+      "epoch": 0.3819895287958115,
+      "grad_norm": 2.1577823162078857,
+      "learning_rate": 3.4511790025683396e-06,
+      "loss": 5.257,
+      "step": 570
+    },
+    {
+      "epoch": 0.38265968586387433,
+      "grad_norm": 2.37388277053833,
+      "learning_rate": 3.4462592485850168e-06,
+      "loss": 5.3045,
+      "step": 571
+    },
+    {
+      "epoch": 0.38332984293193717,
+      "grad_norm": 2.1195552349090576,
+      "learning_rate": 3.4413352135671886e-06,
+      "loss": 5.2868,
+      "step": 572
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 2.1157307624816895,
+      "learning_rate": 3.43640691979201e-06,
+      "loss": 5.377,
+      "step": 573
+    },
+    {
+      "epoch": 0.38467015706806285,
+      "grad_norm": 2.0737154483795166,
+      "learning_rate": 3.4314743895559033e-06,
+      "loss": 5.2294,
+      "step": 574
+    },
+    {
+      "epoch": 0.38534031413612563,
+      "grad_norm": 2.3180153369903564,
+      "learning_rate": 3.4265376451744564e-06,
+      "loss": 5.4814,
+      "step": 575
+    },
+    {
+      "epoch": 0.38601047120418847,
+      "grad_norm": 1.7790483236312866,
+      "learning_rate": 3.4215967089823243e-06,
+      "loss": 5.2332,
+      "step": 576
+    },
+    {
+      "epoch": 0.3866806282722513,
+      "grad_norm": 2.8925139904022217,
+      "learning_rate": 3.4166516033331255e-06,
+      "loss": 5.0873,
+      "step": 577
+    },
+    {
+      "epoch": 0.38735078534031414,
+      "grad_norm": 2.228334426879883,
+      "learning_rate": 3.411702350599341e-06,
+      "loss": 5.0322,
+      "step": 578
+    },
+    {
+      "epoch": 0.388020942408377,
+      "grad_norm": 1.9167518615722656,
+      "learning_rate": 3.406748973172216e-06,
+      "loss": 5.152,
+      "step": 579
+    },
+    {
+      "epoch": 0.3886910994764398,
+      "grad_norm": 2.6737799644470215,
+      "learning_rate": 3.4017914934616547e-06,
+      "loss": 5.2398,
+      "step": 580
+    },
+    {
+      "epoch": 0.3893612565445026,
+      "grad_norm": 1.8289151191711426,
+      "learning_rate": 3.3968299338961213e-06,
+      "loss": 5.2169,
+      "step": 581
+    },
+    {
+      "epoch": 0.39003141361256544,
+      "grad_norm": 2.4280478954315186,
+      "learning_rate": 3.391864316922538e-06,
+      "loss": 5.1239,
+      "step": 582
+    },
+    {
+      "epoch": 0.3907015706806283,
+      "grad_norm": 2.227719783782959,
+      "learning_rate": 3.386894665006184e-06,
+      "loss": 5.3494,
+      "step": 583
+    },
+    {
+      "epoch": 0.3913717277486911,
+      "grad_norm": 2.0614211559295654,
+      "learning_rate": 3.381921000630591e-06,
+      "loss": 4.9243,
+      "step": 584
+    },
+    {
+      "epoch": 0.39204188481675395,
+      "grad_norm": 2.4239277839660645,
+      "learning_rate": 3.3769433462974476e-06,
+      "loss": 5.4025,
+      "step": 585
+    },
+    {
+      "epoch": 0.39271204188481673,
+      "grad_norm": 2.205012798309326,
+      "learning_rate": 3.371961724526489e-06,
+      "loss": 5.1364,
+      "step": 586
+    },
+    {
+      "epoch": 0.39338219895287957,
+      "grad_norm": 2.2407400608062744,
+      "learning_rate": 3.3669761578554037e-06,
+      "loss": 5.303,
+      "step": 587
+    },
+    {
+      "epoch": 0.3940523560209424,
+      "grad_norm": 2.5046958923339844,
+      "learning_rate": 3.361986668839726e-06,
+      "loss": 5.3919,
+      "step": 588
+    },
+    {
+      "epoch": 0.39472251308900524,
+      "grad_norm": 2.4827637672424316,
+      "learning_rate": 3.356993280052735e-06,
+      "loss": 5.4525,
+      "step": 589
+    },
+    {
+      "epoch": 0.3953926701570681,
+      "grad_norm": 2.063934803009033,
+      "learning_rate": 3.3519960140853542e-06,
+      "loss": 5.326,
+      "step": 590
+    },
+    {
+      "epoch": 0.39606282722513086,
+      "grad_norm": 2.1795217990875244,
+      "learning_rate": 3.346994893546046e-06,
+      "loss": 5.1236,
+      "step": 591
+    },
+    {
+      "epoch": 0.3967329842931937,
+      "grad_norm": 1.890066385269165,
+      "learning_rate": 3.3419899410607142e-06,
+      "loss": 5.1188,
+      "step": 592
+    },
+    {
+      "epoch": 0.39740314136125654,
+      "grad_norm": 2.031525135040283,
+      "learning_rate": 3.3369811792725977e-06,
+      "loss": 5.0546,
+      "step": 593
+    },
+    {
+      "epoch": 0.3980732984293194,
+      "grad_norm": 2.3749258518218994,
+      "learning_rate": 3.3319686308421697e-06,
+      "loss": 5.2319,
+      "step": 594
+    },
+    {
+      "epoch": 0.3987434554973822,
+      "grad_norm": 2.0839717388153076,
+      "learning_rate": 3.326952318447033e-06,
+      "loss": 5.1436,
+      "step": 595
+    },
+    {
+      "epoch": 0.39941361256544505,
+      "grad_norm": 2.0904011726379395,
+      "learning_rate": 3.321932264781822e-06,
+      "loss": 5.2929,
+      "step": 596
+    },
+    {
+      "epoch": 0.40008376963350784,
+      "grad_norm": 2.1882741451263428,
+      "learning_rate": 3.3169084925580963e-06,
+      "loss": 5.0201,
+      "step": 597
+    },
+    {
+      "epoch": 0.4007539267015707,
+      "grad_norm": 1.9673819541931152,
+      "learning_rate": 3.3118810245042376e-06,
+      "loss": 4.9752,
+      "step": 598
+    },
+    {
+      "epoch": 0.4014240837696335,
+      "grad_norm": 1.6698298454284668,
+      "learning_rate": 3.3068498833653482e-06,
+      "loss": 5.1009,
+      "step": 599
+    },
+    {
+      "epoch": 0.40209424083769635,
+      "grad_norm": 2.1457369327545166,
+      "learning_rate": 3.301815091903151e-06,
+      "loss": 5.4315,
+      "step": 600
+    },
+    {
+      "epoch": 0.4027643979057592,
+      "grad_norm": 1.8999826908111572,
+      "learning_rate": 3.2967766728958805e-06,
+      "loss": 5.4815,
+      "step": 601
+    },
+    {
+      "epoch": 0.40343455497382197,
+      "grad_norm": 2.4097654819488525,
+      "learning_rate": 3.2917346491381837e-06,
+      "loss": 5.228,
+      "step": 602
+    },
+    {
+      "epoch": 0.4041047120418848,
+      "grad_norm": 2.1515157222747803,
+      "learning_rate": 3.2866890434410153e-06,
+      "loss": 4.81,
+      "step": 603
+    },
+    {
+      "epoch": 0.40477486910994764,
+      "grad_norm": 2.127140998840332,
+      "learning_rate": 3.2816398786315377e-06,
+      "loss": 5.3845,
+      "step": 604
+    },
+    {
+      "epoch": 0.4054450261780105,
+      "grad_norm": 2.041879653930664,
+      "learning_rate": 3.2765871775530135e-06,
+      "loss": 5.1179,
+      "step": 605
+    },
+    {
+      "epoch": 0.4061151832460733,
+      "grad_norm": 2.8136348724365234,
+      "learning_rate": 3.271530963064703e-06,
+      "loss": 5.3501,
+      "step": 606
+    },
+    {
+      "epoch": 0.4067853403141361,
+      "grad_norm": 2.181309938430786,
+      "learning_rate": 3.2664712580417648e-06,
+      "loss": 5.1919,
+      "step": 607
+    },
+    {
+      "epoch": 0.40745549738219894,
+      "grad_norm": 2.3441267013549805,
+      "learning_rate": 3.261408085375146e-06,
+      "loss": 5.3501,
+      "step": 608
+    },
+    {
+      "epoch": 0.4081256544502618,
+      "grad_norm": 2.380401849746704,
+      "learning_rate": 3.2563414679714844e-06,
+      "loss": 5.2335,
+      "step": 609
+    },
+    {
+      "epoch": 0.4087958115183246,
+      "grad_norm": 2.020721912384033,
+      "learning_rate": 3.2512714287530007e-06,
+      "loss": 5.279,
+      "step": 610
+    },
+    {
+      "epoch": 0.40946596858638745,
+      "grad_norm": 2.2213196754455566,
+      "learning_rate": 3.246197990657397e-06,
+      "loss": 5.4033,
+      "step": 611
+    },
+    {
+      "epoch": 0.41013612565445023,
+      "grad_norm": 2.499452829360962,
+      "learning_rate": 3.241121176637753e-06,
+      "loss": 5.4596,
+      "step": 612
+    },
+    {
+      "epoch": 0.41080628272251307,
+      "grad_norm": 2.035451650619507,
+      "learning_rate": 3.2360410096624204e-06,
+      "loss": 4.818,
+      "step": 613
+    },
+    {
+      "epoch": 0.4114764397905759,
+      "grad_norm": 2.1440494060516357,
+      "learning_rate": 3.230957512714923e-06,
+      "loss": 5.3732,
+      "step": 614
+    },
+    {
+      "epoch": 0.41214659685863875,
+      "grad_norm": 2.139416456222534,
+      "learning_rate": 3.2258707087938464e-06,
+      "loss": 5.1974,
+      "step": 615
+    },
+    {
+      "epoch": 0.4128167539267016,
+      "grad_norm": 2.4063401222229004,
+      "learning_rate": 3.2207806209127383e-06,
+      "loss": 5.228,
+      "step": 616
+    },
+    {
+      "epoch": 0.4134869109947644,
+      "grad_norm": 2.257823944091797,
+      "learning_rate": 3.2156872721000066e-06,
+      "loss": 5.185,
+      "step": 617
+    },
+    {
+      "epoch": 0.4141570680628272,
+      "grad_norm": 2.1968698501586914,
+      "learning_rate": 3.2105906853988083e-06,
+      "loss": 5.4655,
+      "step": 618
+    },
+    {
+      "epoch": 0.41482722513089004,
+      "grad_norm": 2.221487045288086,
+      "learning_rate": 3.205490883866952e-06,
+      "loss": 5.2044,
+      "step": 619
+    },
+    {
+      "epoch": 0.4154973821989529,
+      "grad_norm": 2.286722183227539,
+      "learning_rate": 3.2003878905767877e-06,
+      "loss": 5.3126,
+      "step": 620
+    },
+    {
+      "epoch": 0.4161675392670157,
+      "grad_norm": 2.044743299484253,
+      "learning_rate": 3.1952817286151093e-06,
+      "loss": 5.501,
+      "step": 621
+    },
+    {
+      "epoch": 0.41683769633507856,
+      "grad_norm": 2.2196061611175537,
+      "learning_rate": 3.1901724210830437e-06,
+      "loss": 5.372,
+      "step": 622
+    },
+    {
+      "epoch": 0.41750785340314134,
+      "grad_norm": 2.22896146774292,
+      "learning_rate": 3.185059991095949e-06,
+      "loss": 5.0381,
+      "step": 623
+    },
+    {
+      "epoch": 0.4181780104712042,
+      "grad_norm": 2.2274248600006104,
+      "learning_rate": 3.1799444617833103e-06,
+      "loss": 5.3017,
+      "step": 624
+    },
+    {
+      "epoch": 0.418848167539267,
+      "grad_norm": 2.2849974632263184,
+      "learning_rate": 3.1748258562886356e-06,
+      "loss": 5.3214,
+      "step": 625
+    },
+    {
+      "epoch": 0.41951832460732985,
+      "grad_norm": 2.5994486808776855,
+      "learning_rate": 3.1697041977693475e-06,
+      "loss": 5.1276,
+      "step": 626
+    },
+    {
+      "epoch": 0.4201884816753927,
+      "grad_norm": 2.21824049949646,
+      "learning_rate": 3.1645795093966836e-06,
+      "loss": 5.0308,
+      "step": 627
+    },
+    {
+      "epoch": 0.42085863874345547,
+      "grad_norm": 2.221174716949463,
+      "learning_rate": 3.159451814355588e-06,
+      "loss": 5.2276,
+      "step": 628
+    },
+    {
+      "epoch": 0.4215287958115183,
+      "grad_norm": 2.3974714279174805,
+      "learning_rate": 3.1543211358446068e-06,
+      "loss": 5.1327,
+      "step": 629
+    },
+    {
+      "epoch": 0.42219895287958115,
+      "grad_norm": 2.2819957733154297,
+      "learning_rate": 3.149187497075784e-06,
+      "loss": 5.3803,
+      "step": 630
+    },
+    {
+      "epoch": 0.422869109947644,
+      "grad_norm": 2.1973443031311035,
+      "learning_rate": 3.1440509212745584e-06,
+      "loss": 5.1437,
+      "step": 631
+    },
+    {
+      "epoch": 0.4235392670157068,
+      "grad_norm": 2.5798709392547607,
+      "learning_rate": 3.138911431679654e-06,
+      "loss": 5.0404,
+      "step": 632
+    },
+    {
+      "epoch": 0.42420942408376966,
+      "grad_norm": 2.094707727432251,
+      "learning_rate": 3.133769051542977e-06,
+      "loss": 5.3171,
+      "step": 633
+    },
+    {
+      "epoch": 0.42487958115183244,
+      "grad_norm": 1.9110760688781738,
+      "learning_rate": 3.1286238041295134e-06,
+      "loss": 5.0615,
+      "step": 634
+    },
+    {
+      "epoch": 0.4255497382198953,
+      "grad_norm": 2.4531033039093018,
+      "learning_rate": 3.1234757127172195e-06,
+      "loss": 5.1769,
+      "step": 635
+    },
+    {
+      "epoch": 0.4262198952879581,
+      "grad_norm": 2.4736881256103516,
+      "learning_rate": 3.1183248005969186e-06,
+      "loss": 5.267,
+      "step": 636
+    },
+    {
+      "epoch": 0.42689005235602095,
+      "grad_norm": 2.087996244430542,
+      "learning_rate": 3.1131710910721943e-06,
+      "loss": 4.9783,
+      "step": 637
+    },
+    {
+      "epoch": 0.4275602094240838,
+      "grad_norm": 1.8800019025802612,
+      "learning_rate": 3.1080146074592882e-06,
+      "loss": 4.9921,
+      "step": 638
+    },
+    {
+      "epoch": 0.4282303664921466,
+      "grad_norm": 2.587238311767578,
+      "learning_rate": 3.1028553730869913e-06,
+      "loss": 5.2164,
+      "step": 639
+    },
+    {
+      "epoch": 0.4289005235602094,
+      "grad_norm": 2.528468132019043,
+      "learning_rate": 3.0976934112965394e-06,
+      "loss": 5.2577,
+      "step": 640
+    },
+    {
+      "epoch": 0.42957068062827225,
+      "grad_norm": 1.9641128778457642,
+      "learning_rate": 3.0925287454415083e-06,
+      "loss": 5.2473,
+      "step": 641
+    },
+    {
+      "epoch": 0.4302408376963351,
+      "grad_norm": 1.896745204925537,
+      "learning_rate": 3.087361398887706e-06,
+      "loss": 5.0601,
+      "step": 642
+    },
+    {
+      "epoch": 0.4309109947643979,
+      "grad_norm": 2.3236024379730225,
+      "learning_rate": 3.0821913950130694e-06,
+      "loss": 5.2882,
+      "step": 643
+    },
+    {
+      "epoch": 0.4315811518324607,
+      "grad_norm": 1.563217282295227,
+      "learning_rate": 3.077018757207558e-06,
+      "loss": 5.1928,
+      "step": 644
+    },
+    {
+      "epoch": 0.43225130890052355,
+      "grad_norm": 2.0513296127319336,
+      "learning_rate": 3.071843508873046e-06,
+      "loss": 5.1316,
+      "step": 645
+    },
+    {
+      "epoch": 0.4329214659685864,
+      "grad_norm": 2.402965545654297,
+      "learning_rate": 3.06666567342322e-06,
+      "loss": 5.153,
+      "step": 646
+    },
+    {
+      "epoch": 0.4335916230366492,
+      "grad_norm": 2.0920908451080322,
+      "learning_rate": 3.0614852742834706e-06,
+      "loss": 5.1931,
+      "step": 647
+    },
+    {
+      "epoch": 0.43426178010471206,
+      "grad_norm": 2.192448616027832,
+      "learning_rate": 3.056302334890786e-06,
+      "loss": 4.8578,
+      "step": 648
+    },
+    {
+      "epoch": 0.4349319371727749,
+      "grad_norm": 1.793662667274475,
+      "learning_rate": 3.0511168786936483e-06,
+      "loss": 5.2735,
+      "step": 649
+    },
+    {
+      "epoch": 0.4356020942408377,
+      "grad_norm": 1.9571502208709717,
+      "learning_rate": 3.0459289291519244e-06,
+      "loss": 5.2437,
+      "step": 650
+    },
+    {
+      "epoch": 0.4362722513089005,
+      "grad_norm": 2.033674478530884,
+      "learning_rate": 3.040738509736763e-06,
+      "loss": 5.1984,
+      "step": 651
+    },
+    {
+      "epoch": 0.43694240837696335,
+      "grad_norm": 2.2626571655273438,
+      "learning_rate": 3.0355456439304865e-06,
+      "loss": 5.2065,
+      "step": 652
+    },
+    {
+      "epoch": 0.4376125654450262,
+      "grad_norm": 2.125727415084839,
+      "learning_rate": 3.0303503552264846e-06,
+      "loss": 5.3969,
+      "step": 653
+    },
+    {
+      "epoch": 0.43828272251308903,
+      "grad_norm": 2.778316020965576,
+      "learning_rate": 3.025152667129109e-06,
+      "loss": 5.6389,
+      "step": 654
+    },
+    {
+      "epoch": 0.4389528795811518,
+      "grad_norm": 2.4682669639587402,
+      "learning_rate": 3.0199526031535675e-06,
+      "loss": 5.2054,
+      "step": 655
+    },
+    {
+      "epoch": 0.43962303664921465,
+      "grad_norm": 1.9458839893341064,
+      "learning_rate": 3.014750186825815e-06,
+      "loss": 5.0855,
+      "step": 656
+    },
+    {
+      "epoch": 0.4402931937172775,
+      "grad_norm": 2.1652393341064453,
+      "learning_rate": 3.00954544168245e-06,
+      "loss": 5.1186,
+      "step": 657
+    },
+    {
+      "epoch": 0.4409633507853403,
+      "grad_norm": 2.2295773029327393,
+      "learning_rate": 3.004338391270607e-06,
+      "loss": 5.5311,
+      "step": 658
+    },
+    {
+      "epoch": 0.44163350785340316,
+      "grad_norm": 2.231815814971924,
+      "learning_rate": 2.9991290591478495e-06,
+      "loss": 5.2779,
+      "step": 659
+    },
+    {
+      "epoch": 0.44230366492146594,
+      "grad_norm": 2.0944957733154297,
+      "learning_rate": 2.993917468882064e-06,
+      "loss": 5.4016,
+      "step": 660
+    },
+    {
+      "epoch": 0.4429738219895288,
+      "grad_norm": 2.0941920280456543,
+      "learning_rate": 2.9887036440513534e-06,
+      "loss": 5.1275,
+      "step": 661
+    },
+    {
+      "epoch": 0.4436439790575916,
+      "grad_norm": 2.7081048488616943,
+      "learning_rate": 2.98348760824393e-06,
+      "loss": 5.0913,
+      "step": 662
+    },
+    {
+      "epoch": 0.44431413612565446,
+      "grad_norm": 3.268285036087036,
+      "learning_rate": 2.9782693850580084e-06,
+      "loss": 5.1262,
+      "step": 663
+    },
+    {
+      "epoch": 0.4449842931937173,
+      "grad_norm": 2.2694761753082275,
+      "learning_rate": 2.9730489981017e-06,
+      "loss": 5.0765,
+      "step": 664
+    },
+    {
+      "epoch": 0.4456544502617801,
+      "grad_norm": 2.0705652236938477,
+      "learning_rate": 2.9678264709929054e-06,
+      "loss": 5.2526,
+      "step": 665
+    },
+    {
+      "epoch": 0.4463246073298429,
+      "grad_norm": 2.3029444217681885,
+      "learning_rate": 2.962601827359208e-06,
+      "loss": 4.9703,
+      "step": 666
+    },
+    {
+      "epoch": 0.44699476439790575,
+      "grad_norm": 2.4257922172546387,
+      "learning_rate": 2.957375090837764e-06,
+      "loss": 5.1845,
+      "step": 667
+    },
+    {
+      "epoch": 0.4476649214659686,
+      "grad_norm": 2.0682663917541504,
+      "learning_rate": 2.9521462850752027e-06,
+      "loss": 5.1377,
+      "step": 668
+    },
+    {
+      "epoch": 0.44833507853403143,
+      "grad_norm": 2.2123470306396484,
+      "learning_rate": 2.946915433727511e-06,
+      "loss": 5.1044,
+      "step": 669
+    },
+    {
+      "epoch": 0.44900523560209427,
+      "grad_norm": 2.343519926071167,
+      "learning_rate": 2.941682560459933e-06,
+      "loss": 5.2421,
+      "step": 670
+    },
+    {
+      "epoch": 0.44967539267015705,
+      "grad_norm": 2.291065216064453,
+      "learning_rate": 2.9364476889468585e-06,
+      "loss": 5.5774,
+      "step": 671
+    },
+    {
+      "epoch": 0.4503455497382199,
+      "grad_norm": 2.0356202125549316,
+      "learning_rate": 2.9312108428717177e-06,
+      "loss": 5.174,
+      "step": 672
+    },
+    {
+      "epoch": 0.4510157068062827,
+      "grad_norm": 2.0141336917877197,
+      "learning_rate": 2.9259720459268782e-06,
+      "loss": 5.0388,
+      "step": 673
+    },
+    {
+      "epoch": 0.45168586387434556,
+      "grad_norm": 2.0286974906921387,
+      "learning_rate": 2.920731321813526e-06,
+      "loss": 5.3889,
+      "step": 674
+    },
+    {
+      "epoch": 0.4523560209424084,
+      "grad_norm": 1.9957947731018066,
+      "learning_rate": 2.9154886942415727e-06,
+      "loss": 5.1728,
+      "step": 675
+    },
+    {
+      "epoch": 0.4530261780104712,
+      "grad_norm": 2.285508632659912,
+      "learning_rate": 2.910244186929539e-06,
+      "loss": 5.2326,
+      "step": 676
+    },
+    {
+      "epoch": 0.453696335078534,
+      "grad_norm": 2.101499080657959,
+      "learning_rate": 2.9049978236044486e-06,
+      "loss": 5.1891,
+      "step": 677
+    },
+    {
+      "epoch": 0.45436649214659686,
+      "grad_norm": 2.504451274871826,
+      "learning_rate": 2.8997496280017233e-06,
+      "loss": 4.8588,
+      "step": 678
+    },
+    {
+      "epoch": 0.4550366492146597,
+      "grad_norm": 2.49049973487854,
+      "learning_rate": 2.894499623865075e-06,
+      "loss": 5.4943,
+      "step": 679
+    },
+    {
+      "epoch": 0.45570680628272253,
+      "grad_norm": 2.0159759521484375,
+      "learning_rate": 2.8892478349463987e-06,
+      "loss": 5.2162,
+      "step": 680
+    },
+    {
+      "epoch": 0.4563769633507853,
+      "grad_norm": 2.0550146102905273,
+      "learning_rate": 2.8839942850056596e-06,
+      "loss": 5.5218,
+      "step": 681
+    },
+    {
+      "epoch": 0.45704712041884815,
+      "grad_norm": 2.3331401348114014,
+      "learning_rate": 2.8787389978107947e-06,
+      "loss": 5.2914,
+      "step": 682
+    },
+    {
+      "epoch": 0.457717277486911,
+      "grad_norm": 1.9607702493667603,
+      "learning_rate": 2.873481997137599e-06,
+      "loss": 5.0108,
+      "step": 683
+    },
+    {
+      "epoch": 0.4583874345549738,
+      "grad_norm": 2.2367026805877686,
+      "learning_rate": 2.868223306769619e-06,
+      "loss": 5.0986,
+      "step": 684
+    },
+    {
+      "epoch": 0.45905759162303666,
+      "grad_norm": 2.385270357131958,
+      "learning_rate": 2.862962950498046e-06,
+      "loss": 5.2578,
+      "step": 685
+    },
+    {
+      "epoch": 0.4597277486910995,
+      "grad_norm": 2.34967303276062,
+      "learning_rate": 2.8577009521216083e-06,
+      "loss": 5.0958,
+      "step": 686
+    },
+    {
+      "epoch": 0.4603979057591623,
+      "grad_norm": 2.663029432296753,
+      "learning_rate": 2.8524373354464647e-06,
+      "loss": 5.1852,
+      "step": 687
+    },
+    {
+      "epoch": 0.4610680628272251,
+      "grad_norm": 2.3971621990203857,
+      "learning_rate": 2.847172124286091e-06,
+      "loss": 5.5013,
+      "step": 688
+    },
+    {
+      "epoch": 0.46173821989528796,
+      "grad_norm": 2.6697640419006348,
+      "learning_rate": 2.841905342461182e-06,
+      "loss": 5.3391,
+      "step": 689
+    },
+    {
+      "epoch": 0.4624083769633508,
+      "grad_norm": 2.132344961166382,
+      "learning_rate": 2.836637013799537e-06,
+      "loss": 5.0274,
+      "step": 690
+    },
+    {
+      "epoch": 0.46307853403141364,
+      "grad_norm": 2.4103102684020996,
+      "learning_rate": 2.8313671621359503e-06,
+      "loss": 5.1447,
+      "step": 691
+    },
+    {
+      "epoch": 0.4637486910994764,
+      "grad_norm": 2.3897173404693604,
+      "learning_rate": 2.82609581131211e-06,
+      "loss": 5.2839,
+      "step": 692
+    },
+    {
+      "epoch": 0.46441884816753926,
+      "grad_norm": 1.9705911874771118,
+      "learning_rate": 2.8208229851764857e-06,
+      "loss": 5.1606,
+      "step": 693
+    },
+    {
+      "epoch": 0.4650890052356021,
+      "grad_norm": 2.035224199295044,
+      "learning_rate": 2.815548707584222e-06,
+      "loss": 5.4598,
+      "step": 694
+    },
+    {
+      "epoch": 0.46575916230366493,
+      "grad_norm": 2.2948644161224365,
+      "learning_rate": 2.810273002397028e-06,
+      "loss": 5.0036,
+      "step": 695
+    },
+    {
+      "epoch": 0.46642931937172777,
+      "grad_norm": 2.438422441482544,
+      "learning_rate": 2.8049958934830735e-06,
+      "loss": 4.8402,
+      "step": 696
+    },
+    {
+      "epoch": 0.46709947643979055,
+      "grad_norm": 1.948908805847168,
+      "learning_rate": 2.79971740471688e-06,
+      "loss": 4.953,
+      "step": 697
+    },
+    {
+      "epoch": 0.4677696335078534,
+      "grad_norm": 1.7720941305160522,
+      "learning_rate": 2.794437559979208e-06,
+      "loss": 5.3874,
+      "step": 698
+    },
+    {
+      "epoch": 0.4684397905759162,
+      "grad_norm": 2.545875310897827,
+      "learning_rate": 2.7891563831569562e-06,
+      "loss": 5.4277,
+      "step": 699
+    },
+    {
+      "epoch": 0.46910994764397906,
+      "grad_norm": 2.0979082584381104,
+      "learning_rate": 2.7838738981430475e-06,
+      "loss": 5.128,
+      "step": 700
+    },
+    {
+      "epoch": 0.4697801047120419,
+      "grad_norm": 2.4145262241363525,
+      "learning_rate": 2.7785901288363253e-06,
+      "loss": 5.3202,
+      "step": 701
+    },
+    {
+      "epoch": 0.47045026178010474,
+      "grad_norm": 2.7727010250091553,
+      "learning_rate": 2.7733050991414424e-06,
+      "loss": 5.197,
+      "step": 702
+    },
+    {
+      "epoch": 0.4711204188481675,
+      "grad_norm": 2.412595748901367,
+      "learning_rate": 2.7680188329687517e-06,
+      "loss": 5.3592,
+      "step": 703
+    },
+    {
+      "epoch": 0.47179057591623036,
+      "grad_norm": 2.443044900894165,
+      "learning_rate": 2.7627313542342053e-06,
+      "loss": 5.2874,
+      "step": 704
+    },
+    {
+      "epoch": 0.4724607329842932,
+      "grad_norm": 2.2860937118530273,
+      "learning_rate": 2.7574426868592354e-06,
+      "loss": 5.24,
+      "step": 705
+    },
+    {
+      "epoch": 0.47313089005235603,
+      "grad_norm": 2.3260374069213867,
+      "learning_rate": 2.752152854770655e-06,
+      "loss": 5.0235,
+      "step": 706
+    },
+    {
+      "epoch": 0.4738010471204189,
+      "grad_norm": 2.500518560409546,
+      "learning_rate": 2.7468618819005464e-06,
+      "loss": 5.1536,
+      "step": 707
+    },
+    {
+      "epoch": 0.47447120418848165,
+      "grad_norm": 2.2053189277648926,
+      "learning_rate": 2.7415697921861527e-06,
+      "loss": 5.1939,
+      "step": 708
+    },
+    {
+      "epoch": 0.4751413612565445,
+      "grad_norm": 1.915648102760315,
+      "learning_rate": 2.736276609569769e-06,
+      "loss": 5.0484,
+      "step": 709
+    },
+    {
+      "epoch": 0.47581151832460733,
+      "grad_norm": 2.492061138153076,
+      "learning_rate": 2.7309823579986345e-06,
+      "loss": 5.2883,
+      "step": 710
+    },
+    {
+      "epoch": 0.47648167539267017,
+      "grad_norm": 2.494213819503784,
+      "learning_rate": 2.7256870614248277e-06,
+      "loss": 5.3296,
+      "step": 711
+    },
+    {
+      "epoch": 0.477151832460733,
+      "grad_norm": 2.026763916015625,
+      "learning_rate": 2.7203907438051504e-06,
+      "loss": 5.3168,
+      "step": 712
+    },
+    {
+      "epoch": 0.4778219895287958,
+      "grad_norm": 2.2470083236694336,
+      "learning_rate": 2.7150934291010267e-06,
+      "loss": 5.3063,
+      "step": 713
+    },
+    {
+      "epoch": 0.4784921465968586,
+      "grad_norm": 2.2697417736053467,
+      "learning_rate": 2.7097951412783913e-06,
+      "loss": 5.3035,
+      "step": 714
+    },
+    {
+      "epoch": 0.47916230366492146,
+      "grad_norm": 2.0392396450042725,
+      "learning_rate": 2.7044959043075815e-06,
+      "loss": 5.1482,
+      "step": 715
+    },
+    {
+      "epoch": 0.4798324607329843,
+      "grad_norm": 2.2121872901916504,
+      "learning_rate": 2.6991957421632274e-06,
+      "loss": 5.2075,
+      "step": 716
+    },
+    {
+      "epoch": 0.48050261780104714,
+      "grad_norm": 2.4546327590942383,
+      "learning_rate": 2.6938946788241454e-06,
+      "loss": 5.4877,
+      "step": 717
+    },
+    {
+      "epoch": 0.4811727748691099,
+      "grad_norm": 2.6870744228363037,
+      "learning_rate": 2.68859273827323e-06,
+      "loss": 5.4175,
+      "step": 718
+    },
+    {
+      "epoch": 0.48184293193717276,
+      "grad_norm": 2.482137441635132,
+      "learning_rate": 2.6832899444973444e-06,
+      "loss": 5.1865,
+      "step": 719
+    },
+    {
+      "epoch": 0.4825130890052356,
+      "grad_norm": 2.210994243621826,
+      "learning_rate": 2.6779863214872094e-06,
+      "loss": 5.1259,
+      "step": 720
+    },
+    {
+      "epoch": 0.48318324607329843,
+      "grad_norm": 2.396907329559326,
+      "learning_rate": 2.672681893237301e-06,
+      "loss": 5.4887,
+      "step": 721
+    },
+    {
+      "epoch": 0.48385340314136127,
+      "grad_norm": 2.131406545639038,
+      "learning_rate": 2.667376683745735e-06,
+      "loss": 5.1091,
+      "step": 722
+    },
+    {
+      "epoch": 0.4845235602094241,
+      "grad_norm": 2.202169895172119,
+      "learning_rate": 2.662070717014163e-06,
+      "loss": 4.9898,
+      "step": 723
+    },
+    {
+      "epoch": 0.4851937172774869,
+      "grad_norm": 2.0075578689575195,
+      "learning_rate": 2.6567640170476637e-06,
+      "loss": 5.0948,
+      "step": 724
+    },
+    {
+      "epoch": 0.48586387434554973,
+      "grad_norm": 2.250730514526367,
+      "learning_rate": 2.651456607854632e-06,
+      "loss": 5.1126,
+      "step": 725
+    },
+    {
+      "epoch": 0.48653403141361257,
+      "grad_norm": 2.187277317047119,
+      "learning_rate": 2.64614851344667e-06,
+      "loss": 5.2793,
+      "step": 726
+    },
+    {
+      "epoch": 0.4872041884816754,
+      "grad_norm": 2.349581480026245,
+      "learning_rate": 2.640839757838482e-06,
+      "loss": 5.1725,
+      "step": 727
+    },
+    {
+      "epoch": 0.48787434554973824,
+      "grad_norm": 2.1224112510681152,
+      "learning_rate": 2.635530365047764e-06,
+      "loss": 5.043,
+      "step": 728
+    },
+    {
+      "epoch": 0.488544502617801,
+      "grad_norm": 2.0967764854431152,
+      "learning_rate": 2.6302203590950914e-06,
+      "loss": 5.0547,
+      "step": 729
+    },
+    {
+      "epoch": 0.48921465968586386,
+      "grad_norm": 2.5871574878692627,
+      "learning_rate": 2.6249097640038175e-06,
+      "loss": 5.4152,
+      "step": 730
+    },
+    {
+      "epoch": 0.4898848167539267,
+      "grad_norm": 2.2016632556915283,
+      "learning_rate": 2.6195986037999587e-06,
+      "loss": 4.9081,
+      "step": 731
+    },
+    {
+      "epoch": 0.49055497382198954,
+      "grad_norm": 2.4175848960876465,
+      "learning_rate": 2.6142869025120893e-06,
+      "loss": 5.2261,
+      "step": 732
+    },
+    {
+      "epoch": 0.4912251308900524,
+      "grad_norm": 2.046898126602173,
+      "learning_rate": 2.608974684171231e-06,
+      "loss": 5.3583,
+      "step": 733
+    },
+    {
+      "epoch": 0.49189528795811516,
+      "grad_norm": 2.4127092361450195,
+      "learning_rate": 2.6036619728107434e-06,
+      "loss": 5.4236,
+      "step": 734
+    },
+    {
+      "epoch": 0.492565445026178,
+      "grad_norm": 2.0978167057037354,
+      "learning_rate": 2.598348792466219e-06,
+      "loss": 5.2757,
+      "step": 735
+    },
+    {
+      "epoch": 0.49323560209424083,
+      "grad_norm": 2.4408397674560547,
+      "learning_rate": 2.5930351671753707e-06,
+      "loss": 5.2243,
+      "step": 736
+    },
+    {
+      "epoch": 0.49390575916230367,
+      "grad_norm": 2.442858934402466,
+      "learning_rate": 2.5877211209779247e-06,
+      "loss": 5.2263,
+      "step": 737
+    },
+    {
+      "epoch": 0.4945759162303665,
+      "grad_norm": 2.2256414890289307,
+      "learning_rate": 2.5824066779155118e-06,
+      "loss": 5.0348,
+      "step": 738
+    },
+    {
+      "epoch": 0.49524607329842935,
+      "grad_norm": 2.098453998565674,
+      "learning_rate": 2.5770918620315573e-06,
+      "loss": 5.2263,
+      "step": 739
+    },
+    {
+      "epoch": 0.49591623036649213,
+      "grad_norm": 2.2181296348571777,
+      "learning_rate": 2.5717766973711734e-06,
+      "loss": 5.3056,
+      "step": 740
+    },
+    {
+      "epoch": 0.49658638743455497,
+      "grad_norm": 2.378608465194702,
+      "learning_rate": 2.5664612079810512e-06,
+      "loss": 5.1194,
+      "step": 741
+    },
+    {
+      "epoch": 0.4972565445026178,
+      "grad_norm": 2.3113150596618652,
+      "learning_rate": 2.5611454179093515e-06,
+      "loss": 5.1407,
+      "step": 742
+    },
+    {
+      "epoch": 0.49792670157068064,
+      "grad_norm": 2.7640299797058105,
+      "learning_rate": 2.5558293512055927e-06,
+      "loss": 5.2319,
+      "step": 743
+    },
+    {
+      "epoch": 0.4985968586387435,
+      "grad_norm": 1.936281442642212,
+      "learning_rate": 2.5505130319205464e-06,
+      "loss": 5.2259,
+      "step": 744
+    },
+    {
+      "epoch": 0.49926701570680626,
+      "grad_norm": 1.9551467895507812,
+      "learning_rate": 2.545196484106127e-06,
+      "loss": 5.0921,
+      "step": 745
+    },
+    {
+      "epoch": 0.4999371727748691,
+      "grad_norm": 3.0447819232940674,
+      "learning_rate": 2.539879731815284e-06,
+      "loss": 5.1542,
+      "step": 746
+    },
+    {
+      "epoch": 0.500607329842932,
+      "grad_norm": 2.248399496078491,
+      "learning_rate": 2.5345627991018884e-06,
+      "loss": 4.9197,
+      "step": 747
+    },
+    {
+      "epoch": 0.5012774869109947,
+      "grad_norm": 2.5293211936950684,
+      "learning_rate": 2.529245710020632e-06,
+      "loss": 5.1887,
+      "step": 748
+    },
+    {
+      "epoch": 0.5019476439790576,
+      "grad_norm": 1.9050483703613281,
+      "learning_rate": 2.52392848862691e-06,
+      "loss": 5.198,
+      "step": 749
+    },
+    {
+      "epoch": 0.5026178010471204,
+      "grad_norm": 1.7937599420547485,
+      "learning_rate": 2.5186111589767187e-06,
+      "loss": 5.1386,
+      "step": 750
+    },
+    {
+      "epoch": 0.5032879581151832,
+      "grad_norm": 2.233506679534912,
+      "learning_rate": 2.5132937451265427e-06,
+      "loss": 5.3234,
+      "step": 751
+    },
+    {
+      "epoch": 0.5039581151832461,
+      "grad_norm": 2.8709378242492676,
+      "learning_rate": 2.507976271133249e-06,
+      "loss": 4.9134,
+      "step": 752
+    },
+    {
+      "epoch": 0.5046282722513089,
+      "grad_norm": 2.2755050659179688,
+      "learning_rate": 2.502658761053976e-06,
+      "loss": 4.88,
+      "step": 753
+    },
+    {
+      "epoch": 0.5052984293193717,
+      "grad_norm": 2.3214619159698486,
+      "learning_rate": 2.4973412389460245e-06,
+      "loss": 4.9027,
+      "step": 754
+    },
+    {
+      "epoch": 0.5059685863874346,
+      "grad_norm": 2.645906925201416,
+      "learning_rate": 2.4920237288667516e-06,
+      "loss": 5.1749,
+      "step": 755
+    },
+    {
+      "epoch": 0.5066387434554974,
+      "grad_norm": 2.431973457336426,
+      "learning_rate": 2.4867062548734585e-06,
+      "loss": 5.3679,
+      "step": 756
+    },
+    {
+      "epoch": 0.5073089005235603,
+      "grad_norm": 2.163480758666992,
+      "learning_rate": 2.4813888410232825e-06,
+      "loss": 5.0453,
+      "step": 757
+    },
+    {
+      "epoch": 0.507979057591623,
+      "grad_norm": 2.323777437210083,
+      "learning_rate": 2.476071511373091e-06,
+      "loss": 4.8367,
+      "step": 758
+    },
+    {
+      "epoch": 0.5086492146596858,
+      "grad_norm": 2.0933356285095215,
+      "learning_rate": 2.470754289979369e-06,
+      "loss": 5.1988,
+      "step": 759
+    },
+    {
+      "epoch": 0.5093193717277487,
+      "grad_norm": 2.36794376373291,
+      "learning_rate": 2.4654372008981116e-06,
+      "loss": 5.2736,
+      "step": 760
+    },
+    {
+      "epoch": 0.5099895287958115,
+      "grad_norm": 2.062737464904785,
+      "learning_rate": 2.4601202681847173e-06,
+      "loss": 5.1134,
+      "step": 761
+    },
+    {
+      "epoch": 0.5106596858638743,
+      "grad_norm": 2.358090400695801,
+      "learning_rate": 2.4548035158938734e-06,
+      "loss": 5.2011,
+      "step": 762
+    },
+    {
+      "epoch": 0.5113298429319372,
+      "grad_norm": 2.187567949295044,
+      "learning_rate": 2.4494869680794544e-06,
+      "loss": 5.0013,
+      "step": 763
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 2.2072947025299072,
+      "learning_rate": 2.444170648794408e-06,
+      "loss": 5.4118,
+      "step": 764
+    },
+    {
+      "epoch": 0.5126701570680628,
+      "grad_norm": 2.464012622833252,
+      "learning_rate": 2.438854582090649e-06,
+      "loss": 5.101,
+      "step": 765
+    },
+    {
+      "epoch": 0.5133403141361257,
+      "grad_norm": 1.9377411603927612,
+      "learning_rate": 2.433538792018949e-06,
+      "loss": 5.2372,
+      "step": 766
+    },
+    {
+      "epoch": 0.5140104712041885,
+      "grad_norm": 2.2407162189483643,
+      "learning_rate": 2.428223302628828e-06,
+      "loss": 5.1168,
+      "step": 767
+    },
+    {
+      "epoch": 0.5146806282722514,
+      "grad_norm": 2.3990302085876465,
+      "learning_rate": 2.422908137968444e-06,
+      "loss": 5.0664,
+      "step": 768
+    },
+    {
+      "epoch": 0.5153507853403141,
+      "grad_norm": 2.019167900085449,
+      "learning_rate": 2.4175933220844895e-06,
+      "loss": 4.9689,
+      "step": 769
+    },
+    {
+      "epoch": 0.5160209424083769,
+      "grad_norm": 2.4319581985473633,
+      "learning_rate": 2.4122788790220757e-06,
+      "loss": 5.0069,
+      "step": 770
+    },
+    {
+      "epoch": 0.5166910994764398,
+      "grad_norm": 2.629899024963379,
+      "learning_rate": 2.4069648328246305e-06,
+      "loss": 5.1161,
+      "step": 771
+    },
+    {
+      "epoch": 0.5173612565445026,
+      "grad_norm": 2.352264404296875,
+      "learning_rate": 2.4016512075337815e-06,
+      "loss": 5.1608,
+      "step": 772
+    },
+    {
+      "epoch": 0.5180314136125654,
+      "grad_norm": 2.2626523971557617,
+      "learning_rate": 2.3963380271892575e-06,
+      "loss": 5.1764,
+      "step": 773
+    },
+    {
+      "epoch": 0.5187015706806283,
+      "grad_norm": 2.327106475830078,
+      "learning_rate": 2.39102531582877e-06,
+      "loss": 4.9147,
+      "step": 774
+    },
+    {
+      "epoch": 0.5193717277486911,
+      "grad_norm": 2.278066635131836,
+      "learning_rate": 2.3857130974879107e-06,
+      "loss": 5.1306,
+      "step": 775
+    },
+    {
+      "epoch": 0.520041884816754,
+      "grad_norm": 2.14404296875,
+      "learning_rate": 2.3804013962000417e-06,
+      "loss": 5.1931,
+      "step": 776
+    },
+    {
+      "epoch": 0.5207120418848168,
+      "grad_norm": 2.2046420574188232,
+      "learning_rate": 2.3750902359961834e-06,
+      "loss": 5.2374,
+      "step": 777
+    },
+    {
+      "epoch": 0.5213821989528796,
+      "grad_norm": 2.2356772422790527,
+      "learning_rate": 2.3697796409049094e-06,
+      "loss": 5.1609,
+      "step": 778
+    },
+    {
+      "epoch": 0.5220523560209424,
+      "grad_norm": 2.101590633392334,
+      "learning_rate": 2.3644696349522367e-06,
+      "loss": 5.1114,
+      "step": 779
+    },
+    {
+      "epoch": 0.5227225130890052,
+      "grad_norm": 2.1280152797698975,
+      "learning_rate": 2.359160242161518e-06,
+      "loss": 5.1482,
+      "step": 780
+    },
+    {
+      "epoch": 0.523392670157068,
+      "grad_norm": 2.1777684688568115,
+      "learning_rate": 2.3538514865533315e-06,
+      "loss": 5.2686,
+      "step": 781
+    },
+    {
+      "epoch": 0.5240628272251309,
+      "grad_norm": 2.2483649253845215,
+      "learning_rate": 2.348543392145369e-06,
+      "loss": 5.3181,
+      "step": 782
+    },
+    {
+      "epoch": 0.5247329842931937,
+      "grad_norm": 2.2942185401916504,
+      "learning_rate": 2.343235982952337e-06,
+      "loss": 5.0594,
+      "step": 783
+    },
+    {
+      "epoch": 0.5254031413612565,
+      "grad_norm": 1.8971693515777588,
+      "learning_rate": 2.3379292829858373e-06,
+      "loss": 4.9603,
+      "step": 784
+    },
+    {
+      "epoch": 0.5260732984293194,
+      "grad_norm": 2.3738677501678467,
+      "learning_rate": 2.3326233162542655e-06,
+      "loss": 5.2229,
+      "step": 785
+    },
+    {
+      "epoch": 0.5267434554973822,
+      "grad_norm": 2.0925230979919434,
+      "learning_rate": 2.3273181067627e-06,
+      "loss": 5.1843,
+      "step": 786
+    },
+    {
+      "epoch": 0.5274136125654451,
+      "grad_norm": 2.359753131866455,
+      "learning_rate": 2.3220136785127914e-06,
+      "loss": 5.294,
+      "step": 787
+    },
+    {
+      "epoch": 0.5280837696335079,
+      "grad_norm": 3.4682345390319824,
+      "learning_rate": 2.316710055502656e-06,
+      "loss": 5.2098,
+      "step": 788
+    },
+    {
+      "epoch": 0.5287539267015707,
+      "grad_norm": 2.415839433670044,
+      "learning_rate": 2.3114072617267695e-06,
+      "loss": 5.2111,
+      "step": 789
+    },
+    {
+      "epoch": 0.5294240837696335,
+      "grad_norm": 2.775794267654419,
+      "learning_rate": 2.3061053211758554e-06,
+      "loss": 5.2005,
+      "step": 790
+    },
+    {
+      "epoch": 0.5300942408376963,
+      "grad_norm": 1.9597004652023315,
+      "learning_rate": 2.300804257836774e-06,
+      "loss": 4.9533,
+      "step": 791
+    },
+    {
+      "epoch": 0.5307643979057591,
+      "grad_norm": 2.0995726585388184,
+      "learning_rate": 2.2955040956924194e-06,
+      "loss": 5.0624,
+      "step": 792
+    },
+    {
+      "epoch": 0.531434554973822,
+      "grad_norm": 2.133856773376465,
+      "learning_rate": 2.290204858721609e-06,
+      "loss": 5.1067,
+      "step": 793
+    },
+    {
+      "epoch": 0.5321047120418848,
+      "grad_norm": 2.7437381744384766,
+      "learning_rate": 2.2849065708989737e-06,
+      "loss": 5.1492,
+      "step": 794
+    },
+    {
+      "epoch": 0.5327748691099476,
+      "grad_norm": 2.054792881011963,
+      "learning_rate": 2.279609256194851e-06,
+      "loss": 5.16,
+      "step": 795
+    },
+    {
+      "epoch": 0.5334450261780105,
+      "grad_norm": 2.819891929626465,
+      "learning_rate": 2.2743129385751736e-06,
+      "loss": 5.0842,
+      "step": 796
+    },
+    {
+      "epoch": 0.5341151832460733,
+      "grad_norm": 1.7283567190170288,
+      "learning_rate": 2.2690176420013663e-06,
+      "loss": 5.0836,
+      "step": 797
+    },
+    {
+      "epoch": 0.5347853403141362,
+      "grad_norm": 2.002793550491333,
+      "learning_rate": 2.263723390430232e-06,
+      "loss": 5.2371,
+      "step": 798
+    },
+    {
+      "epoch": 0.535455497382199,
+      "grad_norm": 1.953526496887207,
+      "learning_rate": 2.2584302078138477e-06,
+      "loss": 4.928,
+      "step": 799
+    },
+    {
+      "epoch": 0.5361256544502618,
+      "grad_norm": 2.3337416648864746,
+      "learning_rate": 2.2531381180994544e-06,
+      "loss": 5.1453,
+      "step": 800
+    },
+    {
+      "epoch": 0.5367958115183246,
+      "grad_norm": 2.313509464263916,
+      "learning_rate": 2.247847145229346e-06,
+      "loss": 5.4026,
+      "step": 801
+    },
+    {
+      "epoch": 0.5374659685863874,
+      "grad_norm": 2.409526824951172,
+      "learning_rate": 2.242557313140766e-06,
+      "loss": 5.394,
+      "step": 802
+    },
+    {
+      "epoch": 0.5381361256544502,
+      "grad_norm": 2.319575548171997,
+      "learning_rate": 2.2372686457657955e-06,
+      "loss": 5.0763,
+      "step": 803
+    },
+    {
+      "epoch": 0.5388062827225131,
+      "grad_norm": 2.0233211517333984,
+      "learning_rate": 2.2319811670312483e-06,
+      "loss": 4.9404,
+      "step": 804
+    },
+    {
+      "epoch": 0.5394764397905759,
+      "grad_norm": 2.419832468032837,
+      "learning_rate": 2.2266949008585593e-06,
+      "loss": 5.0889,
+      "step": 805
+    },
+    {
+      "epoch": 0.5401465968586388,
+      "grad_norm": 2.5951621532440186,
+      "learning_rate": 2.221409871163675e-06,
+      "loss": 4.7896,
+      "step": 806
+    },
+    {
+      "epoch": 0.5408167539267016,
+      "grad_norm": 2.0952606201171875,
+      "learning_rate": 2.2161261018569534e-06,
+      "loss": 5.1611,
+      "step": 807
+    },
+    {
+      "epoch": 0.5414869109947644,
+      "grad_norm": 2.287743330001831,
+      "learning_rate": 2.2108436168430446e-06,
+      "loss": 5.0981,
+      "step": 808
+    },
+    {
+      "epoch": 0.5421570680628273,
+      "grad_norm": 2.6961002349853516,
+      "learning_rate": 2.205562440020792e-06,
+      "loss": 5.0804,
+      "step": 809
+    },
+    {
+      "epoch": 0.5428272251308901,
+      "grad_norm": 2.004816770553589,
+      "learning_rate": 2.200282595283121e-06,
+      "loss": 5.1105,
+      "step": 810
+    },
+    {
+      "epoch": 0.5434973821989528,
+      "grad_norm": 2.5916943550109863,
+      "learning_rate": 2.195004106516927e-06,
+      "loss": 5.4152,
+      "step": 811
+    },
+    {
+      "epoch": 0.5441675392670157,
+      "grad_norm": 2.091641664505005,
+      "learning_rate": 2.1897269976029727e-06,
+      "loss": 5.3991,
+      "step": 812
+    },
+    {
+      "epoch": 0.5448376963350785,
+      "grad_norm": 1.8039250373840332,
+      "learning_rate": 2.1844512924157783e-06,
+      "loss": 5.1846,
+      "step": 813
+    },
+    {
+      "epoch": 0.5455078534031413,
+      "grad_norm": 2.0316944122314453,
+      "learning_rate": 2.1791770148235143e-06,
+      "loss": 5.3211,
+      "step": 814
+    },
+    {
+      "epoch": 0.5461780104712042,
+      "grad_norm": 2.534327745437622,
+      "learning_rate": 2.173904188687891e-06,
+      "loss": 5.0661,
+      "step": 815
+    },
+    {
+      "epoch": 0.546848167539267,
+      "grad_norm": 2.272688627243042,
+      "learning_rate": 2.1686328378640506e-06,
+      "loss": 5.0007,
+      "step": 816
+    },
+    {
+      "epoch": 0.5475183246073299,
+      "grad_norm": 2.319639205932617,
+      "learning_rate": 2.163362986200464e-06,
+      "loss": 5.2632,
+      "step": 817
+    },
+    {
+      "epoch": 0.5481884816753927,
+      "grad_norm": 2.070636510848999,
+      "learning_rate": 2.158094657538818e-06,
+      "loss": 4.8467,
+      "step": 818
+    },
+    {
+      "epoch": 0.5488586387434555,
+      "grad_norm": 2.220353364944458,
+      "learning_rate": 2.152827875713909e-06,
+      "loss": 5.0162,
+      "step": 819
+    },
+    {
+      "epoch": 0.5495287958115184,
+      "grad_norm": 2.3918845653533936,
+      "learning_rate": 2.147562664553537e-06,
+      "loss": 5.0224,
+      "step": 820
+    },
+    {
+      "epoch": 0.5501989528795812,
+      "grad_norm": 1.9870012998580933,
+      "learning_rate": 2.142299047878392e-06,
+      "loss": 5.0995,
+      "step": 821
+    },
+    {
+      "epoch": 0.5508691099476439,
+      "grad_norm": 2.232142448425293,
+      "learning_rate": 2.1370370495019543e-06,
+      "loss": 5.2807,
+      "step": 822
+    },
+    {
+      "epoch": 0.5515392670157068,
+      "grad_norm": 2.1024062633514404,
+      "learning_rate": 2.1317766932303812e-06,
+      "loss": 5.2492,
+      "step": 823
+    },
+    {
+      "epoch": 0.5522094240837696,
+      "grad_norm": 2.2490932941436768,
+      "learning_rate": 2.1265180028624017e-06,
+      "loss": 5.3037,
+      "step": 824
+    },
+    {
+      "epoch": 0.5528795811518324,
+      "grad_norm": 2.358008623123169,
+      "learning_rate": 2.121261002189206e-06,
+      "loss": 5.0627,
+      "step": 825
+    },
+    {
+      "epoch": 0.5535497382198953,
+      "grad_norm": 2.569955825805664,
+      "learning_rate": 2.1160057149943408e-06,
+      "loss": 5.2377,
+      "step": 826
+    },
+    {
+      "epoch": 0.5542198952879581,
+      "grad_norm": 2.1548383235931396,
+      "learning_rate": 2.110752165053602e-06,
+      "loss": 5.0036,
+      "step": 827
+    },
+    {
+      "epoch": 0.554890052356021,
+      "grad_norm": 2.5272021293640137,
+      "learning_rate": 2.105500376134925e-06,
+      "loss": 4.9154,
+      "step": 828
+    },
+    {
+      "epoch": 0.5555602094240838,
+      "grad_norm": 2.289073944091797,
+      "learning_rate": 2.100250371998278e-06,
+      "loss": 4.9875,
+      "step": 829
+    },
+    {
+      "epoch": 0.5562303664921466,
+      "grad_norm": 2.1486709117889404,
+      "learning_rate": 2.0950021763955526e-06,
+      "loss": 4.9589,
+      "step": 830
+    },
+    {
+      "epoch": 0.5569005235602095,
+      "grad_norm": 2.581888437271118,
+      "learning_rate": 2.089755813070462e-06,
+      "loss": 4.994,
+      "step": 831
+    },
+    {
+      "epoch": 0.5575706806282722,
+      "grad_norm": 2.5435903072357178,
+      "learning_rate": 2.0845113057584277e-06,
+      "loss": 5.399,
+      "step": 832
+    },
+    {
+      "epoch": 0.558240837696335,
+      "grad_norm": 2.547701835632324,
+      "learning_rate": 2.079268678186474e-06,
+      "loss": 4.9896,
+      "step": 833
+    },
+    {
+      "epoch": 0.5589109947643979,
+      "grad_norm": 2.241283655166626,
+      "learning_rate": 2.0740279540731234e-06,
+      "loss": 4.9348,
+      "step": 834
+    },
+    {
+      "epoch": 0.5595811518324607,
+      "grad_norm": 2.11767315864563,
+      "learning_rate": 2.0687891571282827e-06,
+      "loss": 5.0824,
+      "step": 835
+    },
+    {
+      "epoch": 0.5602513089005235,
+      "grad_norm": 2.2595574855804443,
+      "learning_rate": 2.0635523110531424e-06,
+      "loss": 5.0977,
+      "step": 836
+    },
+    {
+      "epoch": 0.5609214659685864,
+      "grad_norm": 2.1792337894439697,
+      "learning_rate": 2.0583174395400675e-06,
+      "loss": 5.1129,
+      "step": 837
+    },
+    {
+      "epoch": 0.5615916230366492,
+      "grad_norm": 1.6265062093734741,
+      "learning_rate": 2.0530845662724897e-06,
+      "loss": 5.2009,
+      "step": 838
+    },
+    {
+      "epoch": 0.5622617801047121,
+      "grad_norm": 2.008075475692749,
+      "learning_rate": 2.0478537149247985e-06,
+      "loss": 5.0015,
+      "step": 839
+    },
+    {
+      "epoch": 0.5629319371727749,
+      "grad_norm": 2.2118005752563477,
+      "learning_rate": 2.042624909162237e-06,
+      "loss": 5.0713,
+      "step": 840
+    },
+    {
+      "epoch": 0.5636020942408377,
+      "grad_norm": 2.1996700763702393,
+      "learning_rate": 2.037398172640793e-06,
+      "loss": 5.0876,
+      "step": 841
+    },
+    {
+      "epoch": 0.5642722513089006,
+      "grad_norm": 2.2208292484283447,
+      "learning_rate": 2.032173529007095e-06,
+      "loss": 5.0323,
+      "step": 842
+    },
+    {
+      "epoch": 0.5649424083769633,
+      "grad_norm": 2.339991569519043,
+      "learning_rate": 2.0269510018983003e-06,
+      "loss": 5.2139,
+      "step": 843
+    },
+    {
+      "epoch": 0.5656125654450261,
+      "grad_norm": 2.166781187057495,
+      "learning_rate": 2.0217306149419925e-06,
+      "loss": 5.0463,
+      "step": 844
+    },
+    {
+      "epoch": 0.566282722513089,
+      "grad_norm": 2.1559500694274902,
+      "learning_rate": 2.0165123917560708e-06,
+      "loss": 5.1012,
+      "step": 845
+    },
+    {
+      "epoch": 0.5669528795811518,
+      "grad_norm": 2.5520191192626953,
+      "learning_rate": 2.011296355948647e-06,
+      "loss": 5.0468,
+      "step": 846
+    },
+    {
+      "epoch": 0.5676230366492147,
+      "grad_norm": 2.157261848449707,
+      "learning_rate": 2.006082531117936e-06,
+      "loss": 5.1208,
+      "step": 847
+    },
+    {
+      "epoch": 0.5682931937172775,
+      "grad_norm": 2.015791416168213,
+      "learning_rate": 2.000870940852151e-06,
+      "loss": 5.0915,
+      "step": 848
+    },
+    {
+      "epoch": 0.5689633507853403,
+      "grad_norm": 2.281003713607788,
+      "learning_rate": 1.9956616087293943e-06,
+      "loss": 4.9733,
+      "step": 849
+    },
+    {
+      "epoch": 0.5696335078534032,
+      "grad_norm": 2.614745616912842,
+      "learning_rate": 1.990454558317551e-06,
+      "loss": 5.0554,
+      "step": 850
+    },
+    {
+      "epoch": 0.570303664921466,
+      "grad_norm": 1.9381040334701538,
+      "learning_rate": 1.985249813174186e-06,
+      "loss": 5.4856,
+      "step": 851
+    },
+    {
+      "epoch": 0.5709738219895288,
+      "grad_norm": 1.8871889114379883,
+      "learning_rate": 1.9800473968464333e-06,
+      "loss": 4.8442,
+      "step": 852
+    },
+    {
+      "epoch": 0.5716439790575917,
+      "grad_norm": 2.239786148071289,
+      "learning_rate": 1.974847332870891e-06,
+      "loss": 4.9804,
+      "step": 853
+    },
+    {
+      "epoch": 0.5723141361256544,
+      "grad_norm": 2.7184863090515137,
+      "learning_rate": 1.9696496447735162e-06,
+      "loss": 5.2111,
+      "step": 854
+    },
+    {
+      "epoch": 0.5729842931937172,
+      "grad_norm": 1.9862799644470215,
+      "learning_rate": 1.964454356069514e-06,
+      "loss": 5.0734,
+      "step": 855
+    },
+    {
+      "epoch": 0.5736544502617801,
+      "grad_norm": 2.061177968978882,
+      "learning_rate": 1.9592614902632374e-06,
+      "loss": 4.8461,
+      "step": 856
+    },
+    {
+      "epoch": 0.5743246073298429,
+      "grad_norm": 2.305881977081299,
+      "learning_rate": 1.954071070848076e-06,
+      "loss": 4.9771,
+      "step": 857
+    },
+    {
+      "epoch": 0.5749947643979058,
+      "grad_norm": 2.083867311477661,
+      "learning_rate": 1.948883121306353e-06,
+      "loss": 4.9486,
+      "step": 858
+    },
+    {
+      "epoch": 0.5756649214659686,
+      "grad_norm": 2.0957860946655273,
+      "learning_rate": 1.9436976651092143e-06,
+      "loss": 5.0704,
+      "step": 859
+    },
+    {
+      "epoch": 0.5763350785340314,
+      "grad_norm": 2.01287579536438,
+      "learning_rate": 1.93851472571653e-06,
+      "loss": 4.7694,
+      "step": 860
+    },
+    {
+      "epoch": 0.5770052356020943,
+      "grad_norm": 1.8653457164764404,
+      "learning_rate": 1.9333343265767803e-06,
+      "loss": 5.2827,
+      "step": 861
+    },
+    {
+      "epoch": 0.5776753926701571,
+      "grad_norm": 6.047257900238037,
+      "learning_rate": 1.9281564911269544e-06,
+      "loss": 5.1922,
+      "step": 862
+    },
+    {
+      "epoch": 0.5783455497382199,
+      "grad_norm": 1.9761441946029663,
+      "learning_rate": 1.9229812427924438e-06,
+      "loss": 5.304,
+      "step": 863
+    },
+    {
+      "epoch": 0.5790157068062827,
+      "grad_norm": 2.323146343231201,
+      "learning_rate": 1.9178086049869314e-06,
+      "loss": 5.2857,
+      "step": 864
+    },
+    {
+      "epoch": 0.5796858638743455,
+      "grad_norm": 2.059492349624634,
+      "learning_rate": 1.912638601112295e-06,
+      "loss": 4.705,
+      "step": 865
+    },
+    {
+      "epoch": 0.5803560209424083,
+      "grad_norm": 2.288398504257202,
+      "learning_rate": 1.9074712545584925e-06,
+      "loss": 5.0554,
+      "step": 866
+    },
+    {
+      "epoch": 0.5810261780104712,
+      "grad_norm": 2.4416613578796387,
+      "learning_rate": 1.9023065887034606e-06,
+      "loss": 5.2227,
+      "step": 867
+    },
+    {
+      "epoch": 0.581696335078534,
+      "grad_norm": 2.665241241455078,
+      "learning_rate": 1.8971446269130095e-06,
+      "loss": 5.1925,
+      "step": 868
+    },
+    {
+      "epoch": 0.5823664921465969,
+      "grad_norm": 2.535707950592041,
+      "learning_rate": 1.8919853925407122e-06,
+      "loss": 5.0185,
+      "step": 869
+    },
+    {
+      "epoch": 0.5830366492146597,
+      "grad_norm": 1.6442601680755615,
+      "learning_rate": 1.8868289089278066e-06,
+      "loss": 5.1138,
+      "step": 870
+    },
+    {
+      "epoch": 0.5837068062827225,
+      "grad_norm": 2.151127815246582,
+      "learning_rate": 1.8816751994030825e-06,
+      "loss": 4.977,
+      "step": 871
+    },
+    {
+      "epoch": 0.5843769633507854,
+      "grad_norm": 2.169908285140991,
+      "learning_rate": 1.876524287282781e-06,
+      "loss": 5.198,
+      "step": 872
+    },
+    {
+      "epoch": 0.5850471204188482,
+      "grad_norm": 2.1878390312194824,
+      "learning_rate": 1.8713761958704874e-06,
+      "loss": 4.8579,
+      "step": 873
+    },
+    {
+      "epoch": 0.585717277486911,
+      "grad_norm": 2.040858745574951,
+      "learning_rate": 1.8662309484570237e-06,
+      "loss": 5.185,
+      "step": 874
+    },
+    {
+      "epoch": 0.5863874345549738,
+      "grad_norm": 1.8941375017166138,
+      "learning_rate": 1.8610885683203467e-06,
+      "loss": 4.9263,
+      "step": 875
+    },
+    {
+      "epoch": 0.5870575916230366,
+      "grad_norm": 2.8969922065734863,
+      "learning_rate": 1.8559490787254423e-06,
+      "loss": 5.1801,
+      "step": 876
+    },
+    {
+      "epoch": 0.5877277486910994,
+      "grad_norm": 2.199195623397827,
+      "learning_rate": 1.8508125029242158e-06,
+      "loss": 5.1154,
+      "step": 877
+    },
+    {
+      "epoch": 0.5883979057591623,
+      "grad_norm": 2.813744306564331,
+      "learning_rate": 1.8456788641553947e-06,
+      "loss": 5.0422,
+      "step": 878
+    },
+    {
+      "epoch": 0.5890680628272251,
+      "grad_norm": 1.960178256034851,
+      "learning_rate": 1.840548185644413e-06,
+      "loss": 5.0718,
+      "step": 879
+    },
+    {
+      "epoch": 0.589738219895288,
+      "grad_norm": 2.070737361907959,
+      "learning_rate": 1.835420490603317e-06,
+      "loss": 5.2001,
+      "step": 880
+    },
+    {
+      "epoch": 0.5904083769633508,
+      "grad_norm": 2.2630679607391357,
+      "learning_rate": 1.8302958022306527e-06,
+      "loss": 5.285,
+      "step": 881
+    },
+    {
+      "epoch": 0.5910785340314136,
+      "grad_norm": 1.8324917554855347,
+      "learning_rate": 1.8251741437113646e-06,
+      "loss": 4.9438,
+      "step": 882
+    },
+    {
+      "epoch": 0.5917486910994765,
+      "grad_norm": 2.199841260910034,
+      "learning_rate": 1.8200555382166901e-06,
+      "loss": 4.932,
+      "step": 883
+    },
+    {
+      "epoch": 0.5924188481675393,
+      "grad_norm": 2.254441499710083,
+      "learning_rate": 1.8149400089040519e-06,
+      "loss": 4.9877,
+      "step": 884
+    },
+    {
+      "epoch": 0.593089005235602,
+      "grad_norm": 2.3117265701293945,
+      "learning_rate": 1.809827578916957e-06,
+      "loss": 5.1466,
+      "step": 885
+    },
+    {
+      "epoch": 0.5937591623036649,
+      "grad_norm": 2.0028316974639893,
+      "learning_rate": 1.8047182713848913e-06,
+      "loss": 5.1821,
+      "step": 886
+    },
+    {
+      "epoch": 0.5944293193717277,
+      "grad_norm": 2.3601784706115723,
+      "learning_rate": 1.7996121094232125e-06,
+      "loss": 5.0561,
+      "step": 887
+    },
+    {
+      "epoch": 0.5950994764397906,
+      "grad_norm": 1.9924535751342773,
+      "learning_rate": 1.7945091161330498e-06,
+      "loss": 5.1431,
+      "step": 888
+    },
+    {
+      "epoch": 0.5957696335078534,
+      "grad_norm": 2.1438329219818115,
+      "learning_rate": 1.7894093146011923e-06,
+      "loss": 5.374,
+      "step": 889
+    },
+    {
+      "epoch": 0.5964397905759162,
+      "grad_norm": 2.116299629211426,
+      "learning_rate": 1.7843127278999944e-06,
+      "loss": 5.1171,
+      "step": 890
+    },
+    {
+      "epoch": 0.5971099476439791,
+      "grad_norm": 2.721327781677246,
+      "learning_rate": 1.779219379087262e-06,
+      "loss": 4.9894,
+      "step": 891
+    },
+    {
+      "epoch": 0.5977801047120419,
+      "grad_norm": 2.1536357402801514,
+      "learning_rate": 1.774129291206155e-06,
+      "loss": 5.2512,
+      "step": 892
+    },
+    {
+      "epoch": 0.5984502617801047,
+      "grad_norm": 2.7440242767333984,
+      "learning_rate": 1.7690424872850776e-06,
+      "loss": 5.3032,
+      "step": 893
+    },
+    {
+      "epoch": 0.5991204188481676,
+      "grad_norm": 2.2082743644714355,
+      "learning_rate": 1.7639589903375798e-06,
+      "loss": 5.1788,
+      "step": 894
+    },
+    {
+      "epoch": 0.5997905759162304,
+      "grad_norm": 2.348146915435791,
+      "learning_rate": 1.7588788233622471e-06,
+      "loss": 4.9295,
+      "step": 895
+    },
+    {
+      "epoch": 0.6004607329842931,
+      "grad_norm": 2.086449146270752,
+      "learning_rate": 1.7538020093426028e-06,
+      "loss": 5.0697,
+      "step": 896
+    },
+    {
+      "epoch": 0.601130890052356,
+      "grad_norm": 2.2433080673217773,
+      "learning_rate": 1.7487285712470004e-06,
+      "loss": 5.1808,
+      "step": 897
+    },
+    {
+      "epoch": 0.6018010471204188,
+      "grad_norm": 2.728177547454834,
+      "learning_rate": 1.7436585320285165e-06,
+      "loss": 5.2392,
+      "step": 898
+    },
+    {
+      "epoch": 0.6024712041884817,
+      "grad_norm": 2.220369815826416,
+      "learning_rate": 1.7385919146248542e-06,
+      "loss": 4.7992,
+      "step": 899
+    },
+    {
+      "epoch": 0.6031413612565445,
+      "grad_norm": 1.9938991069793701,
+      "learning_rate": 1.7335287419582359e-06,
+      "loss": 4.9839,
+      "step": 900
+    },
+    {
+      "epoch": 0.6038115183246073,
+      "grad_norm": 2.1020009517669678,
+      "learning_rate": 1.728469036935297e-06,
+      "loss": 5.3504,
+      "step": 901
+    },
+    {
+      "epoch": 0.6044816753926702,
+      "grad_norm": 2.0839734077453613,
+      "learning_rate": 1.7234128224469876e-06,
+      "loss": 5.1468,
+      "step": 902
+    },
+    {
+      "epoch": 0.605151832460733,
+      "grad_norm": 1.8644421100616455,
+      "learning_rate": 1.7183601213684625e-06,
+      "loss": 4.9848,
+      "step": 903
+    },
+    {
+      "epoch": 0.6058219895287958,
+      "grad_norm": 2.2959697246551514,
+      "learning_rate": 1.7133109565589851e-06,
+      "loss": 5.0542,
+      "step": 904
+    },
+    {
+      "epoch": 0.6064921465968587,
+      "grad_norm": 2.755241632461548,
+      "learning_rate": 1.708265350861817e-06,
+      "loss": 5.2195,
+      "step": 905
+    },
+    {
+      "epoch": 0.6071623036649215,
+      "grad_norm": 2.443690776824951,
+      "learning_rate": 1.7032233271041195e-06,
+      "loss": 5.3558,
+      "step": 906
+    },
+    {
+      "epoch": 0.6078324607329842,
+      "grad_norm": 2.377474784851074,
+      "learning_rate": 1.6981849080968495e-06,
+      "loss": 5.1478,
+      "step": 907
+    },
+    {
+      "epoch": 0.6085026178010471,
+      "grad_norm": 2.394932985305786,
+      "learning_rate": 1.6931501166346522e-06,
+      "loss": 5.1642,
+      "step": 908
+    },
+    {
+      "epoch": 0.6091727748691099,
+      "grad_norm": 2.0526111125946045,
+      "learning_rate": 1.6881189754957634e-06,
+      "loss": 4.9269,
+      "step": 909
+    },
+    {
+      "epoch": 0.6098429319371728,
+      "grad_norm": 2.377007484436035,
+      "learning_rate": 1.6830915074419041e-06,
+      "loss": 5.1605,
+      "step": 910
+    },
+    {
+      "epoch": 0.6105130890052356,
+      "grad_norm": 2.189647912979126,
+      "learning_rate": 1.6780677352181781e-06,
+      "loss": 5.4189,
+      "step": 911
+    },
+    {
+      "epoch": 0.6111832460732984,
+      "grad_norm": 2.9951906204223633,
+      "learning_rate": 1.673047681552968e-06,
+      "loss": 5.0452,
+      "step": 912
+    },
+    {
+      "epoch": 0.6118534031413613,
+      "grad_norm": 2.1930408477783203,
+      "learning_rate": 1.6680313691578314e-06,
+      "loss": 4.9555,
+      "step": 913
+    },
+    {
+      "epoch": 0.6125235602094241,
+      "grad_norm": 2.541778802871704,
+      "learning_rate": 1.6630188207274034e-06,
+      "loss": 5.347,
+      "step": 914
+    },
+    {
+      "epoch": 0.613193717277487,
+      "grad_norm": 2.424107074737549,
+      "learning_rate": 1.6580100589392862e-06,
+      "loss": 4.8499,
+      "step": 915
+    },
+    {
+      "epoch": 0.6138638743455498,
+      "grad_norm": 2.1423468589782715,
+      "learning_rate": 1.6530051064539544e-06,
+      "loss": 4.891,
+      "step": 916
+    },
+    {
+      "epoch": 0.6145340314136125,
+      "grad_norm": 2.1596438884735107,
+      "learning_rate": 1.648003985914647e-06,
+      "loss": 5.2224,
+      "step": 917
+    },
+    {
+      "epoch": 0.6152041884816754,
+      "grad_norm": 2.459508180618286,
+      "learning_rate": 1.643006719947266e-06,
+      "loss": 4.9728,
+      "step": 918
+    },
+    {
+      "epoch": 0.6158743455497382,
+      "grad_norm": 2.7368266582489014,
+      "learning_rate": 1.6380133311602745e-06,
+      "loss": 5.0703,
+      "step": 919
+    },
+    {
+      "epoch": 0.616544502617801,
+      "grad_norm": 2.2180020809173584,
+      "learning_rate": 1.6330238421445959e-06,
+      "loss": 4.8884,
+      "step": 920
+    },
+    {
+      "epoch": 0.6172146596858639,
+      "grad_norm": 2.030007839202881,
+      "learning_rate": 1.628038275473511e-06,
+      "loss": 5.0728,
+      "step": 921
+    },
+    {
+      "epoch": 0.6178848167539267,
+      "grad_norm": 2.1178250312805176,
+      "learning_rate": 1.6230566537025539e-06,
+      "loss": 5.336,
+      "step": 922
+    },
+    {
+      "epoch": 0.6185549738219895,
+      "grad_norm": 2.596489191055298,
+      "learning_rate": 1.6180789993694092e-06,
+      "loss": 5.3047,
+      "step": 923
+    },
+    {
+      "epoch": 0.6192251308900524,
+      "grad_norm": 2.662217855453491,
+      "learning_rate": 1.6131053349938167e-06,
+      "loss": 5.0644,
+      "step": 924
+    },
+    {
+      "epoch": 0.6198952879581152,
+      "grad_norm": 2.6869699954986572,
+      "learning_rate": 1.6081356830774625e-06,
+      "loss": 4.8568,
+      "step": 925
+    },
+    {
+      "epoch": 0.620565445026178,
+      "grad_norm": 2.0684654712677,
+      "learning_rate": 1.60317006610388e-06,
+      "loss": 4.9038,
+      "step": 926
+    },
+    {
+      "epoch": 0.6212356020942409,
+      "grad_norm": 2.492082357406616,
+      "learning_rate": 1.5982085065383463e-06,
+      "loss": 4.9686,
+      "step": 927
+    },
+    {
+      "epoch": 0.6219057591623036,
+      "grad_norm": 1.890755534172058,
+      "learning_rate": 1.593251026827785e-06,
+      "loss": 5.0921,
+      "step": 928
+    },
+    {
+      "epoch": 0.6225759162303665,
+      "grad_norm": 2.640437126159668,
+      "learning_rate": 1.5882976494006597e-06,
+      "loss": 5.0312,
+      "step": 929
+    },
+    {
+      "epoch": 0.6232460732984293,
+      "grad_norm": 2.528960704803467,
+      "learning_rate": 1.583348396666875e-06,
+      "loss": 5.3454,
+      "step": 930
+    },
+    {
+      "epoch": 0.6239162303664921,
+      "grad_norm": 1.8781567811965942,
+      "learning_rate": 1.5784032910176767e-06,
+      "loss": 4.9509,
+      "step": 931
+    },
+    {
+      "epoch": 0.624586387434555,
+      "grad_norm": 2.1682565212249756,
+      "learning_rate": 1.5734623548255446e-06,
+      "loss": 5.0253,
+      "step": 932
+    },
+    {
+      "epoch": 0.6252565445026178,
+      "grad_norm": 2.2101004123687744,
+      "learning_rate": 1.5685256104440977e-06,
+      "loss": 5.0726,
+      "step": 933
+    },
+    {
+      "epoch": 0.6259267015706806,
+      "grad_norm": 2.197580337524414,
+      "learning_rate": 1.5635930802079905e-06,
+      "loss": 5.0462,
+      "step": 934
+    },
+    {
+      "epoch": 0.6265968586387435,
+      "grad_norm": 2.051706075668335,
+      "learning_rate": 1.558664786432812e-06,
+      "loss": 5.0463,
+      "step": 935
+    },
+    {
+      "epoch": 0.6272670157068063,
+      "grad_norm": 2.4994492530822754,
+      "learning_rate": 1.5537407514149843e-06,
+      "loss": 5.2271,
+      "step": 936
+    },
+    {
+      "epoch": 0.6279371727748692,
+      "grad_norm": 2.1373913288116455,
+      "learning_rate": 1.5488209974316608e-06,
+      "loss": 5.1098,
+      "step": 937
+    },
+    {
+      "epoch": 0.6286073298429319,
+      "grad_norm": 3.135277509689331,
+      "learning_rate": 1.5439055467406299e-06,
+      "loss": 5.1169,
+      "step": 938
+    },
+    {
+      "epoch": 0.6292774869109947,
+      "grad_norm": 2.1849703788757324,
+      "learning_rate": 1.5389944215802093e-06,
+      "loss": 5.1895,
+      "step": 939
+    },
+    {
+      "epoch": 0.6299476439790576,
+      "grad_norm": 2.2958791255950928,
+      "learning_rate": 1.5340876441691465e-06,
+      "loss": 5.2696,
+      "step": 940
+    },
+    {
+      "epoch": 0.6306178010471204,
+      "grad_norm": 2.3132193088531494,
+      "learning_rate": 1.5291852367065224e-06,
+      "loss": 5.2749,
+      "step": 941
+    },
+    {
+      "epoch": 0.6312879581151832,
+      "grad_norm": 3.056617021560669,
+      "learning_rate": 1.5242872213716426e-06,
+      "loss": 4.9438,
+      "step": 942
+    },
+    {
+      "epoch": 0.6319581151832461,
+      "grad_norm": 1.7447956800460815,
+      "learning_rate": 1.5193936203239452e-06,
+      "loss": 5.0243,
+      "step": 943
+    },
+    {
+      "epoch": 0.6326282722513089,
+      "grad_norm": 2.151508092880249,
+      "learning_rate": 1.514504455702897e-06,
+      "loss": 5.2879,
+      "step": 944
+    },
+    {
+      "epoch": 0.6332984293193717,
+      "grad_norm": 2.2350332736968994,
+      "learning_rate": 1.5096197496278942e-06,
+      "loss": 5.2136,
+      "step": 945
+    },
+    {
+      "epoch": 0.6339685863874346,
+      "grad_norm": 2.3778679370880127,
+      "learning_rate": 1.5047395241981606e-06,
+      "loss": 5.215,
+      "step": 946
+    },
+    {
+      "epoch": 0.6346387434554974,
+      "grad_norm": 2.2321999073028564,
+      "learning_rate": 1.4998638014926476e-06,
+      "loss": 5.1403,
+      "step": 947
+    },
+    {
+      "epoch": 0.6353089005235603,
+      "grad_norm": 2.958186388015747,
+      "learning_rate": 1.4949926035699382e-06,
+      "loss": 4.9567,
+      "step": 948
+    },
+    {
+      "epoch": 0.635979057591623,
+      "grad_norm": 1.9767028093338013,
+      "learning_rate": 1.4901259524681444e-06,
+      "loss": 5.1582,
+      "step": 949
+    },
+    {
+      "epoch": 0.6366492146596858,
+      "grad_norm": 5.674414157867432,
+      "learning_rate": 1.485263870204805e-06,
+      "loss": 4.9456,
+      "step": 950
+    },
+    {
+      "epoch": 0.6373193717277487,
+      "grad_norm": 2.754943370819092,
+      "learning_rate": 1.4804063787767924e-06,
+      "loss": 5.4178,
+      "step": 951
+    },
+    {
+      "epoch": 0.6379895287958115,
+      "grad_norm": 2.0564680099487305,
+      "learning_rate": 1.4755535001602044e-06,
+      "loss": 5.1593,
+      "step": 952
+    },
+    {
+      "epoch": 0.6386596858638743,
+      "grad_norm": 2.3718740940093994,
+      "learning_rate": 1.470705256310275e-06,
+      "loss": 5.0125,
+      "step": 953
+    },
+    {
+      "epoch": 0.6393298429319372,
+      "grad_norm": 2.3798463344573975,
+      "learning_rate": 1.4658616691612653e-06,
+      "loss": 5.1758,
+      "step": 954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.253955602645874,
+      "learning_rate": 1.4610227606263718e-06,
+      "loss": 5.0712,
+      "step": 955
+    },
+    {
+      "epoch": 0.6406701570680629,
+      "grad_norm": 2.273934841156006,
+      "learning_rate": 1.4561885525976227e-06,
+      "loss": 5.1689,
+      "step": 956
+    },
+    {
+      "epoch": 0.6413403141361257,
+      "grad_norm": 2.539167881011963,
+      "learning_rate": 1.4513590669457798e-06,
+      "loss": 5.1739,
+      "step": 957
+    },
+    {
+      "epoch": 0.6420104712041885,
+      "grad_norm": 2.5597174167633057,
+      "learning_rate": 1.4465343255202425e-06,
+      "loss": 5.2156,
+      "step": 958
+    },
+    {
+      "epoch": 0.6426806282722514,
+      "grad_norm": 2.087078094482422,
+      "learning_rate": 1.4417143501489428e-06,
+      "loss": 4.8461,
+      "step": 959
+    },
+    {
+      "epoch": 0.6433507853403141,
+      "grad_norm": 2.6639723777770996,
+      "learning_rate": 1.436899162638255e-06,
+      "loss": 5.0139,
+      "step": 960
+    },
+    {
+      "epoch": 0.6440209424083769,
+      "grad_norm": 2.5338354110717773,
+      "learning_rate": 1.4320887847728864e-06,
+      "loss": 5.381,
+      "step": 961
+    },
+    {
+      "epoch": 0.6446910994764398,
+      "grad_norm": 2.2862353324890137,
+      "learning_rate": 1.4272832383157902e-06,
+      "loss": 5.1397,
+      "step": 962
+    },
+    {
+      "epoch": 0.6453612565445026,
+      "grad_norm": 2.38358998298645,
+      "learning_rate": 1.422482545008059e-06,
+      "loss": 5.1036,
+      "step": 963
+    },
+    {
+      "epoch": 0.6460314136125654,
+      "grad_norm": 2.191732168197632,
+      "learning_rate": 1.41768672656883e-06,
+      "loss": 5.3168,
+      "step": 964
+    },
+    {
+      "epoch": 0.6467015706806283,
+      "grad_norm": 2.504222869873047,
+      "learning_rate": 1.4128958046951835e-06,
+      "loss": 5.0436,
+      "step": 965
+    },
+    {
+      "epoch": 0.6473717277486911,
+      "grad_norm": 2.3530819416046143,
+      "learning_rate": 1.4081098010620481e-06,
+      "loss": 5.3618,
+      "step": 966
+    },
+    {
+      "epoch": 0.648041884816754,
+      "grad_norm": 2.2972545623779297,
+      "learning_rate": 1.4033287373221022e-06,
+      "loss": 5.1409,
+      "step": 967
+    },
+    {
+      "epoch": 0.6487120418848168,
+      "grad_norm": 2.2007195949554443,
+      "learning_rate": 1.398552635105675e-06,
+      "loss": 5.3514,
+      "step": 968
+    },
+    {
+      "epoch": 0.6493821989528796,
+      "grad_norm": 2.092994213104248,
+      "learning_rate": 1.3937815160206485e-06,
+      "loss": 5.1033,
+      "step": 969
+    },
+    {
+      "epoch": 0.6500523560209424,
+      "grad_norm": 1.8113712072372437,
+      "learning_rate": 1.3890154016523615e-06,
+      "loss": 4.8722,
+      "step": 970
+    },
+    {
+      "epoch": 0.6507225130890052,
+      "grad_norm": 2.0513923168182373,
+      "learning_rate": 1.3842543135635075e-06,
+      "loss": 5.2054,
+      "step": 971
+    },
+    {
+      "epoch": 0.651392670157068,
+      "grad_norm": 2.532233238220215,
+      "learning_rate": 1.3794982732940444e-06,
+      "loss": 5.0319,
+      "step": 972
+    },
+    {
+      "epoch": 0.6520628272251309,
+      "grad_norm": 2.2876851558685303,
+      "learning_rate": 1.3747473023610885e-06,
+      "loss": 5.084,
+      "step": 973
+    },
+    {
+      "epoch": 0.6527329842931937,
+      "grad_norm": 1.9787242412567139,
+      "learning_rate": 1.3700014222588256e-06,
+      "loss": 4.9917,
+      "step": 974
+    },
+    {
+      "epoch": 0.6534031413612565,
+      "grad_norm": 2.3899760246276855,
+      "learning_rate": 1.3652606544584086e-06,
+      "loss": 5.1363,
+      "step": 975
+    },
+    {
+      "epoch": 0.6540732984293194,
+      "grad_norm": 2.2005436420440674,
+      "learning_rate": 1.3605250204078585e-06,
+      "loss": 5.0114,
+      "step": 976
+    },
+    {
+      "epoch": 0.6547434554973822,
+      "grad_norm": 2.094499349594116,
+      "learning_rate": 1.3557945415319746e-06,
+      "loss": 5.2157,
+      "step": 977
+    },
+    {
+      "epoch": 0.6554136125654451,
+      "grad_norm": 2.4904541969299316,
+      "learning_rate": 1.3510692392322316e-06,
+      "loss": 5.2516,
+      "step": 978
+    },
+    {
+      "epoch": 0.6560837696335079,
+      "grad_norm": 1.9249300956726074,
+      "learning_rate": 1.3463491348866837e-06,
+      "loss": 5.1889,
+      "step": 979
+    },
+    {
+      "epoch": 0.6567539267015707,
+      "grad_norm": 2.2775089740753174,
+      "learning_rate": 1.3416342498498703e-06,
+      "loss": 5.2225,
+      "step": 980
+    },
+    {
+      "epoch": 0.6574240837696335,
+      "grad_norm": 2.057075262069702,
+      "learning_rate": 1.3369246054527152e-06,
+      "loss": 5.279,
+      "step": 981
+    },
+    {
+      "epoch": 0.6580942408376963,
+      "grad_norm": 3.090810775756836,
+      "learning_rate": 1.3322202230024356e-06,
+      "loss": 5.0438,
+      "step": 982
+    },
+    {
+      "epoch": 0.6587643979057591,
+      "grad_norm": 2.0964043140411377,
+      "learning_rate": 1.3275211237824415e-06,
+      "loss": 5.1803,
+      "step": 983
+    },
+    {
+      "epoch": 0.659434554973822,
+      "grad_norm": 2.631629228591919,
+      "learning_rate": 1.3228273290522403e-06,
+      "loss": 4.9872,
+      "step": 984
+    },
+    {
+      "epoch": 0.6601047120418848,
+      "grad_norm": 2.3109307289123535,
+      "learning_rate": 1.3181388600473433e-06,
+      "loss": 5.0663,
+      "step": 985
+    },
+    {
+      "epoch": 0.6607748691099476,
+      "grad_norm": 2.0783398151397705,
+      "learning_rate": 1.3134557379791636e-06,
+      "loss": 5.0157,
+      "step": 986
+    },
+    {
+      "epoch": 0.6614450261780105,
+      "grad_norm": 1.9334152936935425,
+      "learning_rate": 1.308777984034925e-06,
+      "loss": 4.9186,
+      "step": 987
+    },
+    {
+      "epoch": 0.6621151832460733,
+      "grad_norm": 2.0614535808563232,
+      "learning_rate": 1.3041056193775667e-06,
+      "loss": 5.1108,
+      "step": 988
+    },
+    {
+      "epoch": 0.6627853403141362,
+      "grad_norm": 2.4771032333374023,
+      "learning_rate": 1.2994386651456454e-06,
+      "loss": 5.0904,
+      "step": 989
+    },
+    {
+      "epoch": 0.663455497382199,
+      "grad_norm": 2.3995018005371094,
+      "learning_rate": 1.2947771424532399e-06,
+      "loss": 5.1486,
+      "step": 990
+    },
+    {
+      "epoch": 0.6641256544502617,
+      "grad_norm": 2.0989136695861816,
+      "learning_rate": 1.290121072389854e-06,
+      "loss": 5.1089,
+      "step": 991
+    },
+    {
+      "epoch": 0.6647958115183246,
+      "grad_norm": 2.0078938007354736,
+      "learning_rate": 1.2854704760203265e-06,
+      "loss": 5.327,
+      "step": 992
+    },
+    {
+      "epoch": 0.6654659685863874,
+      "grad_norm": 2.3669705390930176,
+      "learning_rate": 1.280825374384728e-06,
+      "loss": 5.0939,
+      "step": 993
+    },
+    {
+      "epoch": 0.6661361256544502,
+      "grad_norm": 2.0225884914398193,
+      "learning_rate": 1.2761857884982748e-06,
+      "loss": 5.1236,
+      "step": 994
+    },
+    {
+      "epoch": 0.6668062827225131,
+      "grad_norm": 2.120720863342285,
+      "learning_rate": 1.2715517393512239e-06,
+      "loss": 5.0135,
+      "step": 995
+    },
+    {
+      "epoch": 0.6674764397905759,
+      "grad_norm": 2.4720404148101807,
+      "learning_rate": 1.2669232479087868e-06,
+      "loss": 5.202,
+      "step": 996
+    },
+    {
+      "epoch": 0.6681465968586388,
+      "grad_norm": 2.271608352661133,
+      "learning_rate": 1.2623003351110302e-06,
+      "loss": 4.9941,
+      "step": 997
+    },
+    {
+      "epoch": 0.6688167539267016,
+      "grad_norm": 2.137437343597412,
+      "learning_rate": 1.2576830218727817e-06,
+      "loss": 4.9493,
+      "step": 998
+    },
+    {
+      "epoch": 0.6694869109947644,
+      "grad_norm": 2.634817600250244,
+      "learning_rate": 1.253071329083537e-06,
+      "loss": 5.2495,
+      "step": 999
+    },
+    {
+      "epoch": 0.6701570680628273,
+      "grad_norm": 2.1700284481048584,
+      "learning_rate": 1.2484652776073613e-06,
+      "loss": 5.0001,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6708272251308901,
+      "grad_norm": 2.2349369525909424,
+      "learning_rate": 1.2438648882827984e-06,
+      "loss": 5.0308,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6714973821989528,
+      "grad_norm": 2.265810966491699,
+      "learning_rate": 1.2392701819227768e-06,
+      "loss": 5.1102,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6721675392670157,
+      "grad_norm": 2.1368231773376465,
+      "learning_rate": 1.2346811793145135e-06,
+      "loss": 4.8558,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6728376963350785,
+      "grad_norm": 2.780846118927002,
+      "learning_rate": 1.2300979012194222e-06,
+      "loss": 5.0876,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6735078534031413,
+      "grad_norm": 2.227250814437866,
+      "learning_rate": 1.2255203683730144e-06,
+      "loss": 5.2692,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6741780104712042,
+      "grad_norm": 1.9963433742523193,
+      "learning_rate": 1.2209486014848138e-06,
+      "loss": 5.0172,
+      "step": 1006
+    },
+    {
+      "epoch": 0.674848167539267,
+      "grad_norm": 2.1841299533843994,
+      "learning_rate": 1.216382621238253e-06,
+      "loss": 5.1824,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6755183246073299,
+      "grad_norm": 2.150474786758423,
+      "learning_rate": 1.2118224482905892e-06,
+      "loss": 5.2183,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6761884816753927,
+      "grad_norm": 2.589063882827759,
+      "learning_rate": 1.2072681032728048e-06,
+      "loss": 5.0682,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6768586387434555,
+      "grad_norm": 2.179386615753174,
+      "learning_rate": 1.2027196067895138e-06,
+      "loss": 4.9421,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6775287958115184,
+      "grad_norm": 1.9265069961547852,
+      "learning_rate": 1.1981769794188735e-06,
+      "loss": 5.1369,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6781989528795812,
+      "grad_norm": 2.187723159790039,
+      "learning_rate": 1.1936402417124865e-06,
+      "loss": 5.0849,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6788691099476439,
+      "grad_norm": 2.1362497806549072,
+      "learning_rate": 1.1891094141953114e-06,
+      "loss": 5.1315,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6795392670157068,
+      "grad_norm": 2.778053045272827,
+      "learning_rate": 1.1845845173655658e-06,
+      "loss": 5.1386,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6802094240837696,
+      "grad_norm": 2.422011137008667,
+      "learning_rate": 1.1800655716946355e-06,
+      "loss": 5.2709,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6808795811518324,
+      "grad_norm": 3.0462863445281982,
+      "learning_rate": 1.1755525976269851e-06,
+      "loss": 5.0757,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6815497382198953,
+      "grad_norm": 2.319657564163208,
+      "learning_rate": 1.1710456155800612e-06,
+      "loss": 5.2053,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6822198952879581,
+      "grad_norm": 1.9604647159576416,
+      "learning_rate": 1.1665446459442014e-06,
+      "loss": 4.9627,
+      "step": 1018
+    },
+    {
+      "epoch": 0.682890052356021,
+      "grad_norm": 2.1266629695892334,
+      "learning_rate": 1.1620497090825434e-06,
+      "loss": 5.0072,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6835602094240838,
+      "grad_norm": 2.379615068435669,
+      "learning_rate": 1.157560825330929e-06,
+      "loss": 4.9827,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6842303664921466,
+      "grad_norm": 1.7763530015945435,
+      "learning_rate": 1.153078014997816e-06,
+      "loss": 4.9343,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6849005235602095,
+      "grad_norm": 1.9431225061416626,
+      "learning_rate": 1.1486012983641858e-06,
+      "loss": 5.3294,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6855706806282722,
+      "grad_norm": 2.3577218055725098,
+      "learning_rate": 1.1441306956834506e-06,
+      "loss": 4.8451,
+      "step": 1023
+    },
+    {
+      "epoch": 0.686240837696335,
+      "grad_norm": 2.1077463626861572,
+      "learning_rate": 1.1396662271813625e-06,
+      "loss": 5.1231,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6869109947643979,
+      "grad_norm": 2.119215250015259,
+      "learning_rate": 1.135207913055919e-06,
+      "loss": 5.1054,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6875811518324607,
+      "grad_norm": 1.9343498945236206,
+      "learning_rate": 1.130755773477278e-06,
+      "loss": 4.9137,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6882513089005236,
+      "grad_norm": 1.9753541946411133,
+      "learning_rate": 1.1263098285876592e-06,
+      "loss": 5.0259,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6889214659685864,
+      "grad_norm": 1.7094663381576538,
+      "learning_rate": 1.1218700985012584e-06,
+      "loss": 5.3622,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6895916230366492,
+      "grad_norm": 2.331881046295166,
+      "learning_rate": 1.1174366033041555e-06,
+      "loss": 5.1145,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6902617801047121,
+      "grad_norm": 2.7443439960479736,
+      "learning_rate": 1.11300936305422e-06,
+      "loss": 5.2896,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6909319371727749,
+      "grad_norm": 2.5008018016815186,
+      "learning_rate": 1.1085883977810247e-06,
+      "loss": 5.1273,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6916020942408377,
+      "grad_norm": 2.0488998889923096,
+      "learning_rate": 1.1041737274857533e-06,
+      "loss": 5.3134,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6922722513089006,
+      "grad_norm": 1.9997189044952393,
+      "learning_rate": 1.099765372141111e-06,
+      "loss": 4.9215,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6929424083769633,
+      "grad_norm": 2.278038740158081,
+      "learning_rate": 1.095363351691231e-06,
+      "loss": 5.3046,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6936125654450261,
+      "grad_norm": 2.224224805831909,
+      "learning_rate": 1.090967686051586e-06,
+      "loss": 4.8629,
+      "step": 1035
+    },
+    {
+      "epoch": 0.694282722513089,
+      "grad_norm": 1.8891581296920776,
+      "learning_rate": 1.0865783951089005e-06,
+      "loss": 4.8682,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6949528795811518,
+      "grad_norm": 2.5251624584198,
+      "learning_rate": 1.0821954987210584e-06,
+      "loss": 5.1834,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6956230366492147,
+      "grad_norm": 1.9501714706420898,
+      "learning_rate": 1.0778190167170145e-06,
+      "loss": 4.988,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6962931937172775,
+      "grad_norm": 2.3251214027404785,
+      "learning_rate": 1.0734489688967001e-06,
+      "loss": 4.9515,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6969633507853403,
+      "grad_norm": 2.245398998260498,
+      "learning_rate": 1.0690853750309417e-06,
+      "loss": 5.034,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6976335078534032,
+      "grad_norm": 2.169013261795044,
+      "learning_rate": 1.0647282548613636e-06,
+      "loss": 5.1129,
+      "step": 1041
+    },
+    {
+      "epoch": 0.698303664921466,
+      "grad_norm": 2.124728202819824,
+      "learning_rate": 1.0603776281003037e-06,
+      "loss": 4.9932,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6989738219895288,
+      "grad_norm": 2.4038686752319336,
+      "learning_rate": 1.0560335144307238e-06,
+      "loss": 5.0806,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6996439790575916,
+      "grad_norm": 2.3839521408081055,
+      "learning_rate": 1.0516959335061157e-06,
+      "loss": 5.2674,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7003141361256544,
+      "grad_norm": 2.4349517822265625,
+      "learning_rate": 1.0473649049504195e-06,
+      "loss": 5.0355,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7009842931937172,
+      "grad_norm": 2.201889991760254,
+      "learning_rate": 1.0430404483579299e-06,
+      "loss": 4.9743,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7016544502617801,
+      "grad_norm": 1.9042011499404907,
+      "learning_rate": 1.03872258329321e-06,
+      "loss": 4.9538,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7023246073298429,
+      "grad_norm": 1.9022135734558105,
+      "learning_rate": 1.0344113292909993e-06,
+      "loss": 5.0116,
+      "step": 1048
+    },
+    {
+      "epoch": 0.7029947643979058,
+      "grad_norm": 2.583439588546753,
+      "learning_rate": 1.0301067058561296e-06,
+      "loss": 5.1247,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7036649214659686,
+      "grad_norm": 2.090341329574585,
+      "learning_rate": 1.025808732463434e-06,
+      "loss": 5.0901,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7043350785340314,
+      "grad_norm": 2.315129280090332,
+      "learning_rate": 1.0215174285576615e-06,
+      "loss": 5.0314,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7050052356020943,
+      "grad_norm": 2.396319627761841,
+      "learning_rate": 1.0172328135533852e-06,
+      "loss": 5.2357,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7056753926701571,
+      "grad_norm": 2.5406458377838135,
+      "learning_rate": 1.0129549068349185e-06,
+      "loss": 5.1052,
+      "step": 1053
+    },
+    {
+      "epoch": 0.70634554973822,
+      "grad_norm": 1.9954371452331543,
+      "learning_rate": 1.008683727756223e-06,
+      "loss": 5.1479,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7070157068062827,
+      "grad_norm": 2.01151967048645,
+      "learning_rate": 1.0044192956408241e-06,
+      "loss": 5.1476,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7076858638743455,
+      "grad_norm": 1.9611108303070068,
+      "learning_rate": 1.0001616297817238e-06,
+      "loss": 5.0162,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7083560209424083,
+      "grad_norm": 2.1401920318603516,
+      "learning_rate": 9.959107494413118e-07,
+      "loss": 5.1986,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7090261780104712,
+      "grad_norm": 2.2252449989318848,
+      "learning_rate": 9.9166667385128e-07,
+      "loss": 5.2287,
+      "step": 1058
+    },
+    {
+      "epoch": 0.709696335078534,
+      "grad_norm": 2.3955202102661133,
+      "learning_rate": 9.874294222125316e-07,
+      "loss": 5.1653,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7103664921465969,
+      "grad_norm": 2.327396869659424,
+      "learning_rate": 9.831990136951003e-07,
+      "loss": 4.9633,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7110366492146597,
+      "grad_norm": 1.979533314704895,
+      "learning_rate": 9.789754674380597e-07,
+      "loss": 5.1874,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7117068062827225,
+      "grad_norm": 2.0694804191589355,
+      "learning_rate": 9.747588025494351e-07,
+      "loss": 5.1106,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7123769633507854,
+      "grad_norm": 2.1586427688598633,
+      "learning_rate": 9.705490381061227e-07,
+      "loss": 5.7843,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7130471204188482,
+      "grad_norm": 2.1827874183654785,
+      "learning_rate": 9.663461931537965e-07,
+      "loss": 5.0521,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7137172774869109,
+      "grad_norm": 1.917332649230957,
+      "learning_rate": 9.621502867068286e-07,
+      "loss": 4.7877,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7143874345549738,
+      "grad_norm": 2.061023473739624,
+      "learning_rate": 9.57961337748198e-07,
+      "loss": 5.1603,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7150575916230366,
+      "grad_norm": 2.5405564308166504,
+      "learning_rate": 9.537793652294096e-07,
+      "loss": 4.9064,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7157277486910995,
+      "grad_norm": 2.10152268409729,
+      "learning_rate": 9.496043880704025e-07,
+      "loss": 4.9259,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7163979057591623,
+      "grad_norm": 2.14532732963562,
+      "learning_rate": 9.454364251594683e-07,
+      "loss": 4.8689,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7170680628272251,
+      "grad_norm": 2.5526065826416016,
+      "learning_rate": 9.412754953531664e-07,
+      "loss": 5.0766,
+      "step": 1070
+    },
+    {
+      "epoch": 0.717738219895288,
+      "grad_norm": 2.3036630153656006,
+      "learning_rate": 9.371216174762365e-07,
+      "loss": 5.0425,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7184083769633508,
+      "grad_norm": 2.6235127449035645,
+      "learning_rate": 9.329748103215153e-07,
+      "loss": 5.1771,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7190785340314136,
+      "grad_norm": 2.0794284343719482,
+      "learning_rate": 9.288350926498474e-07,
+      "loss": 5.0205,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7197486910994765,
+      "grad_norm": 2.227818489074707,
+      "learning_rate": 9.247024831900065e-07,
+      "loss": 5.1375,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7204188481675393,
+      "grad_norm": 2.365229606628418,
+      "learning_rate": 9.205770006386066e-07,
+      "loss": 4.9339,
+      "step": 1075
+    },
+    {
+      "epoch": 0.721089005235602,
+      "grad_norm": 2.539462089538574,
+      "learning_rate": 9.164586636600168e-07,
+      "loss": 5.1168,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7217591623036649,
+      "grad_norm": 1.930780053138733,
+      "learning_rate": 9.123474908862817e-07,
+      "loss": 4.8905,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7224293193717277,
+      "grad_norm": 2.19105863571167,
+      "learning_rate": 9.082435009170298e-07,
+      "loss": 5.0776,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7230994764397906,
+      "grad_norm": 2.0612986087799072,
+      "learning_rate": 9.041467123193967e-07,
+      "loss": 4.9747,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7237696335078534,
+      "grad_norm": 2.2077128887176514,
+      "learning_rate": 9.000571436279363e-07,
+      "loss": 5.2412,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7244397905759162,
+      "grad_norm": 2.2451539039611816,
+      "learning_rate": 8.959748133445401e-07,
+      "loss": 5.2,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7251099476439791,
+      "grad_norm": 2.2490241527557373,
+      "learning_rate": 8.918997399383491e-07,
+      "loss": 4.871,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7257801047120419,
+      "grad_norm": 2.010239839553833,
+      "learning_rate": 8.878319418456738e-07,
+      "loss": 5.3053,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7264502617801047,
+      "grad_norm": 2.4550492763519287,
+      "learning_rate": 8.837714374699116e-07,
+      "loss": 5.1626,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7271204188481676,
+      "grad_norm": 2.0976195335388184,
+      "learning_rate": 8.797182451814603e-07,
+      "loss": 5.2316,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7277905759162304,
+      "grad_norm": 1.84067964553833,
+      "learning_rate": 8.756723833176376e-07,
+      "loss": 4.9795,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7284607329842931,
+      "grad_norm": 2.3378286361694336,
+      "learning_rate": 8.716338701825974e-07,
+      "loss": 4.9382,
+      "step": 1087
+    },
+    {
+      "epoch": 0.729130890052356,
+      "grad_norm": 2.4951651096343994,
+      "learning_rate": 8.676027240472445e-07,
+      "loss": 4.9807,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7298010471204188,
+      "grad_norm": 2.410188913345337,
+      "learning_rate": 8.63578963149157e-07,
+      "loss": 4.8992,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7304712041884817,
+      "grad_norm": 2.0807738304138184,
+      "learning_rate": 8.595626056924983e-07,
+      "loss": 5.0869,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7311413612565445,
+      "grad_norm": 2.450817108154297,
+      "learning_rate": 8.555536698479395e-07,
+      "loss": 5.0584,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7318115183246073,
+      "grad_norm": 2.6124391555786133,
+      "learning_rate": 8.515521737525748e-07,
+      "loss": 5.0537,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7324816753926702,
+      "grad_norm": 2.224575996398926,
+      "learning_rate": 8.47558135509838e-07,
+      "loss": 5.165,
+      "step": 1093
+    },
+    {
+      "epoch": 0.733151832460733,
+      "grad_norm": 2.324970006942749,
+      "learning_rate": 8.435715731894245e-07,
+      "loss": 5.2625,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7338219895287958,
+      "grad_norm": 3.057370185852051,
+      "learning_rate": 8.395925048272069e-07,
+      "loss": 4.9905,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7344921465968587,
+      "grad_norm": 2.1883432865142822,
+      "learning_rate": 8.356209484251521e-07,
+      "loss": 4.9127,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7351623036649214,
+      "grad_norm": 2.510040283203125,
+      "learning_rate": 8.316569219512446e-07,
+      "loss": 5.1409,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7358324607329843,
+      "grad_norm": 2.030768394470215,
+      "learning_rate": 8.27700443339399e-07,
+      "loss": 4.9366,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7365026178010471,
+      "grad_norm": 2.413208246231079,
+      "learning_rate": 8.237515304893845e-07,
+      "loss": 4.7532,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7371727748691099,
+      "grad_norm": 2.9401113986968994,
+      "learning_rate": 8.198102012667409e-07,
+      "loss": 5.2135,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7378429319371728,
+      "grad_norm": 2.438283681869507,
+      "learning_rate": 8.158764735026989e-07,
+      "loss": 4.8833,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7385130890052356,
+      "grad_norm": 1.9578114748001099,
+      "learning_rate": 8.119503649940969e-07,
+      "loss": 4.7994,
+      "step": 1102
+    },
+    {
+      "epoch": 0.7391832460732984,
+      "grad_norm": 2.337479591369629,
+      "learning_rate": 8.080318935033057e-07,
+      "loss": 5.0201,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7398534031413613,
+      "grad_norm": 2.590935707092285,
+      "learning_rate": 8.041210767581415e-07,
+      "loss": 5.0649,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7405235602094241,
+      "grad_norm": 2.108517646789551,
+      "learning_rate": 8.002179324517914e-07,
+      "loss": 5.0238,
+      "step": 1105
+    },
+    {
+      "epoch": 0.741193717277487,
+      "grad_norm": 2.358654022216797,
+      "learning_rate": 7.96322478242732e-07,
+      "loss": 4.8828,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7418638743455498,
+      "grad_norm": 2.0274851322174072,
+      "learning_rate": 7.92434731754645e-07,
+      "loss": 5.159,
+      "step": 1107
+    },
+    {
+      "epoch": 0.7425340314136125,
+      "grad_norm": 1.9625911712646484,
+      "learning_rate": 7.885547105763444e-07,
+      "loss": 5.2291,
+      "step": 1108
+    },
+    {
+      "epoch": 0.7432041884816754,
+      "grad_norm": 2.2201850414276123,
+      "learning_rate": 7.84682432261693e-07,
+      "loss": 4.7749,
+      "step": 1109
+    },
+    {
+      "epoch": 0.7438743455497382,
+      "grad_norm": 2.2542552947998047,
+      "learning_rate": 7.808179143295219e-07,
+      "loss": 4.7366,
+      "step": 1110
+    },
+    {
+      "epoch": 0.744544502617801,
+      "grad_norm": 2.288684368133545,
+      "learning_rate": 7.769611742635555e-07,
+      "loss": 5.3394,
+      "step": 1111
+    },
+    {
+      "epoch": 0.7452146596858639,
+      "grad_norm": 1.952508807182312,
+      "learning_rate": 7.731122295123267e-07,
+      "loss": 5.0954,
+      "step": 1112
+    },
+    {
+      "epoch": 0.7458848167539267,
+      "grad_norm": 2.026873826980591,
+      "learning_rate": 7.692710974891041e-07,
+      "loss": 5.2031,
+      "step": 1113
+    },
+    {
+      "epoch": 0.7465549738219895,
+      "grad_norm": 2.2017242908477783,
+      "learning_rate": 7.654377955718084e-07,
+      "loss": 5.0217,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7472251308900524,
+      "grad_norm": 2.8049333095550537,
+      "learning_rate": 7.616123411029367e-07,
+      "loss": 5.5266,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7478952879581152,
+      "grad_norm": 2.2660751342773438,
+      "learning_rate": 7.577947513894829e-07,
+      "loss": 5.1222,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7485654450261781,
+      "grad_norm": 2.282757520675659,
+      "learning_rate": 7.539850437028584e-07,
+      "loss": 4.9671,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7492356020942408,
+      "grad_norm": 2.5191409587860107,
+      "learning_rate": 7.501832352788147e-07,
+      "loss": 5.0362,
+      "step": 1118
+    },
+    {
+      "epoch": 0.7499057591623036,
+      "grad_norm": 2.01081919670105,
+      "learning_rate": 7.463893433173669e-07,
+      "loss": 5.1149,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7505759162303665,
+      "grad_norm": 2.890888214111328,
+      "learning_rate": 7.426033849827149e-07,
+      "loss": 4.9724,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7512460732984293,
+      "grad_norm": 2.2375943660736084,
+      "learning_rate": 7.388253774031659e-07,
+      "loss": 5.0092,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7519162303664921,
+      "grad_norm": 2.171166181564331,
+      "learning_rate": 7.350553376710534e-07,
+      "loss": 5.2243,
+      "step": 1122
+    },
+    {
+      "epoch": 0.752586387434555,
+      "grad_norm": 2.401020050048828,
+      "learning_rate": 7.312932828426677e-07,
+      "loss": 5.0838,
+      "step": 1123
+    },
+    {
+      "epoch": 0.7532565445026178,
+      "grad_norm": 2.0519063472747803,
+      "learning_rate": 7.275392299381695e-07,
+      "loss": 4.9773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.7539267015706806,
+      "grad_norm": 2.1711297035217285,
+      "learning_rate": 7.237931959415207e-07,
+      "loss": 4.922,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7545968586387435,
+      "grad_norm": 2.1511411666870117,
+      "learning_rate": 7.200551978004039e-07,
+      "loss": 5.2377,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7552670157068063,
+      "grad_norm": 2.283114433288574,
+      "learning_rate": 7.163252524261443e-07,
+      "loss": 5.052,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7559371727748692,
+      "grad_norm": 2.2724831104278564,
+      "learning_rate": 7.126033766936366e-07,
+      "loss": 4.9807,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7566073298429319,
+      "grad_norm": 2.3890180587768555,
+      "learning_rate": 7.088895874412668e-07,
+      "loss": 5.1586,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7572774869109947,
+      "grad_norm": 2.5954933166503906,
+      "learning_rate": 7.051839014708372e-07,
+      "loss": 5.064,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7579476439790576,
+      "grad_norm": 2.3010849952697754,
+      "learning_rate": 7.01486335547488e-07,
+      "loss": 4.944,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7586178010471204,
+      "grad_norm": 1.916985034942627,
+      "learning_rate": 6.977969063996223e-07,
+      "loss": 4.9986,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7592879581151832,
+      "grad_norm": 2.1439950466156006,
+      "learning_rate": 6.941156307188332e-07,
+      "loss": 5.0494,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7599581151832461,
+      "grad_norm": 2.3296618461608887,
+      "learning_rate": 6.904425251598254e-07,
+      "loss": 5.1706,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7606282722513089,
+      "grad_norm": 2.1923646926879883,
+      "learning_rate": 6.867776063403411e-07,
+      "loss": 4.9433,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7612984293193717,
+      "grad_norm": 2.6342825889587402,
+      "learning_rate": 6.831208908410825e-07,
+      "loss": 5.0178,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7619685863874346,
+      "grad_norm": 2.1129515171051025,
+      "learning_rate": 6.794723952056406e-07,
+      "loss": 5.5518,
+      "step": 1137
+    },
+    {
+      "epoch": 0.7626387434554974,
+      "grad_norm": 2.1442720890045166,
+      "learning_rate": 6.758321359404166e-07,
+      "loss": 4.8999,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7633089005235603,
+      "grad_norm": 2.1809842586517334,
+      "learning_rate": 6.722001295145509e-07,
+      "loss": 4.9485,
+      "step": 1139
+    },
+    {
+      "epoch": 0.763979057591623,
+      "grad_norm": 2.5500996112823486,
+      "learning_rate": 6.685763923598457e-07,
+      "loss": 5.2035,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7646492146596858,
+      "grad_norm": 2.273024320602417,
+      "learning_rate": 6.649609408706909e-07,
+      "loss": 5.169,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7653193717277487,
+      "grad_norm": 2.008328676223755,
+      "learning_rate": 6.613537914039916e-07,
+      "loss": 4.6944,
+      "step": 1142
+    },
+    {
+      "epoch": 0.7659895287958115,
+      "grad_norm": 1.9586178064346313,
+      "learning_rate": 6.577549602790936e-07,
+      "loss": 4.9895,
+      "step": 1143
+    },
+    {
+      "epoch": 0.7666596858638743,
+      "grad_norm": 2.1195785999298096,
+      "learning_rate": 6.541644637777089e-07,
+      "loss": 5.0358,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7673298429319372,
+      "grad_norm": 1.9128762483596802,
+      "learning_rate": 6.505823181438414e-07,
+      "loss": 4.7646,
+      "step": 1145
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 2.1416807174682617,
+      "learning_rate": 6.470085395837139e-07,
+      "loss": 4.8051,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7686701570680629,
+      "grad_norm": 2.2263646125793457,
+      "learning_rate": 6.434431442656966e-07,
+      "loss": 4.9021,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7693403141361257,
+      "grad_norm": 2.5426077842712402,
+      "learning_rate": 6.398861483202326e-07,
+      "loss": 4.9737,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7700104712041885,
+      "grad_norm": 1.969666600227356,
+      "learning_rate": 6.363375678397638e-07,
+      "loss": 4.8842,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7706806282722513,
+      "grad_norm": 1.8267338275909424,
+      "learning_rate": 6.3279741887866e-07,
+      "loss": 5.1273,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7713507853403141,
+      "grad_norm": 2.2450010776519775,
+      "learning_rate": 6.292657174531449e-07,
+      "loss": 4.918,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7720209424083769,
+      "grad_norm": 2.3020267486572266,
+      "learning_rate": 6.257424795412229e-07,
+      "loss": 5.1702,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7726910994764398,
+      "grad_norm": 2.17891788482666,
+      "learning_rate": 6.222277210826105e-07,
+      "loss": 4.9306,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7733612565445026,
+      "grad_norm": 2.244767904281616,
+      "learning_rate": 6.187214579786602e-07,
+      "loss": 5.2499,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7740314136125654,
+      "grad_norm": 2.6995861530303955,
+      "learning_rate": 6.152237060922916e-07,
+      "loss": 4.4595,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7747015706806283,
+      "grad_norm": 2.22466778755188,
+      "learning_rate": 6.117344812479154e-07,
+      "loss": 4.7913,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7753717277486911,
+      "grad_norm": 2.3127236366271973,
+      "learning_rate": 6.082537992313667e-07,
+      "loss": 5.1861,
+      "step": 1157
+    },
+    {
+      "epoch": 0.776041884816754,
+      "grad_norm": 2.3279953002929688,
+      "learning_rate": 6.04781675789832e-07,
+      "loss": 4.8803,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7767120418848168,
+      "grad_norm": 1.9932987689971924,
+      "learning_rate": 6.013181266317747e-07,
+      "loss": 4.8846,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7773821989528796,
+      "grad_norm": 2.2041752338409424,
+      "learning_rate": 5.978631674268693e-07,
+      "loss": 5.0406,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7780523560209424,
+      "grad_norm": 2.7398922443389893,
+      "learning_rate": 5.944168138059261e-07,
+      "loss": 4.8435,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7787225130890052,
+      "grad_norm": 2.2988779544830322,
+      "learning_rate": 5.909790813608235e-07,
+      "loss": 4.9794,
+      "step": 1162
+    },
+    {
+      "epoch": 0.779392670157068,
+      "grad_norm": 2.4860384464263916,
+      "learning_rate": 5.875499856444358e-07,
+      "loss": 4.9771,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7800628272251309,
+      "grad_norm": 2.2307345867156982,
+      "learning_rate": 5.841295421705642e-07,
+      "loss": 5.1321,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7807329842931937,
+      "grad_norm": 1.9979619979858398,
+      "learning_rate": 5.807177664138644e-07,
+      "loss": 5.0066,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7814031413612565,
+      "grad_norm": 2.7662293910980225,
+      "learning_rate": 5.773146738097773e-07,
+      "loss": 5.1328,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7820732984293194,
+      "grad_norm": 1.8403937816619873,
+      "learning_rate": 5.739202797544619e-07,
+      "loss": 5.1026,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7827434554973822,
+      "grad_norm": 2.0250093936920166,
+      "learning_rate": 5.70534599604722e-07,
+      "loss": 5.0364,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7834136125654451,
+      "grad_norm": 2.4427311420440674,
+      "learning_rate": 5.671576486779398e-07,
+      "loss": 5.0155,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7840837696335079,
+      "grad_norm": 2.440207004547119,
+      "learning_rate": 5.637894422520027e-07,
+      "loss": 5.0432,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7847539267015706,
+      "grad_norm": 2.796550750732422,
+      "learning_rate": 5.604299955652381e-07,
+      "loss": 5.1718,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7854240837696335,
+      "grad_norm": 1.7754937410354614,
+      "learning_rate": 5.570793238163436e-07,
+      "loss": 5.0406,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7860942408376963,
+      "grad_norm": 2.1862494945526123,
+      "learning_rate": 5.537374421643146e-07,
+      "loss": 4.9814,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7867643979057591,
+      "grad_norm": 2.3494925498962402,
+      "learning_rate": 5.504043657283823e-07,
+      "loss": 5.0955,
+      "step": 1174
+    },
+    {
+      "epoch": 0.787434554973822,
+      "grad_norm": 2.3137881755828857,
+      "learning_rate": 5.470801095879382e-07,
+      "loss": 5.4067,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7881047120418848,
+      "grad_norm": 2.568342685699463,
+      "learning_rate": 5.437646887824721e-07,
+      "loss": 5.2062,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7887748691099477,
+      "grad_norm": 2.170259952545166,
+      "learning_rate": 5.404581183114998e-07,
+      "loss": 5.1373,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7894450261780105,
+      "grad_norm": 2.1512749195098877,
+      "learning_rate": 5.371604131344979e-07,
+      "loss": 5.1238,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7901151832460733,
+      "grad_norm": 2.1920998096466064,
+      "learning_rate": 5.338715881708331e-07,
+      "loss": 4.895,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7907853403141362,
+      "grad_norm": 2.220663070678711,
+      "learning_rate": 5.30591658299697e-07,
+      "loss": 5.0831,
+      "step": 1180
+    },
+    {
+      "epoch": 0.791455497382199,
+      "grad_norm": 1.8620455265045166,
+      "learning_rate": 5.273206383600396e-07,
+      "loss": 4.9863,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7921256544502617,
+      "grad_norm": 1.9880365133285522,
+      "learning_rate": 5.240585431505002e-07,
+      "loss": 4.5929,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7927958115183246,
+      "grad_norm": 2.086104154586792,
+      "learning_rate": 5.208053874293406e-07,
+      "loss": 4.9963,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7934659685863874,
+      "grad_norm": 2.240558624267578,
+      "learning_rate": 5.175611859143806e-07,
+      "loss": 4.7543,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7941361256544502,
+      "grad_norm": 2.365817070007324,
+      "learning_rate": 5.143259532829267e-07,
+      "loss": 5.3058,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7948062827225131,
+      "grad_norm": 2.296628475189209,
+      "learning_rate": 5.110997041717116e-07,
+      "loss": 5.0126,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7954764397905759,
+      "grad_norm": 2.3187358379364014,
+      "learning_rate": 5.078824531768228e-07,
+      "loss": 4.8611,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7961465968586388,
+      "grad_norm": 2.1453304290771484,
+      "learning_rate": 5.046742148536404e-07,
+      "loss": 4.8574,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7968167539267016,
+      "grad_norm": 2.0868799686431885,
+      "learning_rate": 5.014750037167698e-07,
+      "loss": 5.1505,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7974869109947644,
+      "grad_norm": 1.784549593925476,
+      "learning_rate": 4.982848342399741e-07,
+      "loss": 4.8991,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7981570680628273,
+      "grad_norm": 1.9321773052215576,
+      "learning_rate": 4.951037208561116e-07,
+      "loss": 4.928,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7988272251308901,
+      "grad_norm": 2.221848726272583,
+      "learning_rate": 4.919316779570707e-07,
+      "loss": 5.1757,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7994973821989528,
+      "grad_norm": 2.8202672004699707,
+      "learning_rate": 4.887687198937002e-07,
+      "loss": 5.0547,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8001675392670157,
+      "grad_norm": 2.555051803588867,
+      "learning_rate": 4.856148609757508e-07,
+      "loss": 5.2369,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8008376963350785,
+      "grad_norm": 1.9884748458862305,
+      "learning_rate": 4.824701154718045e-07,
+      "loss": 5.1295,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8015078534031413,
+      "grad_norm": 2.368027925491333,
+      "learning_rate": 4.793344976092146e-07,
+      "loss": 4.9066,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8021780104712042,
+      "grad_norm": 2.2029364109039307,
+      "learning_rate": 4.762080215740386e-07,
+      "loss": 5.1582,
+      "step": 1197
+    },
+    {
+      "epoch": 0.802848167539267,
+      "grad_norm": 2.4221737384796143,
+      "learning_rate": 4.730907015109759e-07,
+      "loss": 5.1814,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8035183246073299,
+      "grad_norm": 3.0611066818237305,
+      "learning_rate": 4.699825515233003e-07,
+      "loss": 5.1989,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8041884816753927,
+      "grad_norm": 1.8770856857299805,
+      "learning_rate": 4.6688358567280186e-07,
+      "loss": 4.8958,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8048586387434555,
+      "grad_norm": 2.4339778423309326,
+      "learning_rate": 4.637938179797166e-07,
+      "loss": 5.1851,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8055287958115184,
+      "grad_norm": 2.19170880317688,
+      "learning_rate": 4.6071326242266984e-07,
+      "loss": 5.1224,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8061989528795811,
+      "grad_norm": 2.668105125427246,
+      "learning_rate": 4.576419329386081e-07,
+      "loss": 5.1409,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8068691099476439,
+      "grad_norm": 2.373500108718872,
+      "learning_rate": 4.545798434227366e-07,
+      "loss": 5.1758,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8075392670157068,
+      "grad_norm": 2.703608989715576,
+      "learning_rate": 4.5152700772845947e-07,
+      "loss": 5.1138,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8082094240837696,
+      "grad_norm": 2.1617355346679688,
+      "learning_rate": 4.484834396673146e-07,
+      "loss": 5.1007,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8088795811518324,
+      "grad_norm": 2.0932817459106445,
+      "learning_rate": 4.4544915300890915e-07,
+      "loss": 5.0408,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8095497382198953,
+      "grad_norm": 2.0439183712005615,
+      "learning_rate": 4.424241614808636e-07,
+      "loss": 5.0576,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8102198952879581,
+      "grad_norm": 2.4854025840759277,
+      "learning_rate": 4.3940847876874193e-07,
+      "loss": 4.9091,
+      "step": 1209
+    },
+    {
+      "epoch": 0.810890052356021,
+      "grad_norm": 1.8081549406051636,
+      "learning_rate": 4.3640211851599614e-07,
+      "loss": 5.0549,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8115602094240838,
+      "grad_norm": 2.7424466609954834,
+      "learning_rate": 4.3340509432390043e-07,
+      "loss": 5.025,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8122303664921466,
+      "grad_norm": 1.8654872179031372,
+      "learning_rate": 4.3041741975149294e-07,
+      "loss": 5.0045,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8129005235602095,
+      "grad_norm": 2.3444607257843018,
+      "learning_rate": 4.274391083155091e-07,
+      "loss": 4.8565,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8135706806282722,
+      "grad_norm": 2.078972339630127,
+      "learning_rate": 4.244701734903281e-07,
+      "loss": 5.0893,
+      "step": 1214
+    },
+    {
+      "epoch": 0.814240837696335,
+      "grad_norm": 1.8032852411270142,
+      "learning_rate": 4.215106287079043e-07,
+      "loss": 4.9006,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8149109947643979,
+      "grad_norm": 2.370866060256958,
+      "learning_rate": 4.185604873577118e-07,
+      "loss": 4.6488,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8155811518324607,
+      "grad_norm": 5.689456939697266,
+      "learning_rate": 4.1561976278668174e-07,
+      "loss": 4.8656,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8162513089005236,
+      "grad_norm": 2.522265672683716,
+      "learning_rate": 4.1268846829914305e-07,
+      "loss": 5.106,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8169214659685864,
+      "grad_norm": 2.0813515186309814,
+      "learning_rate": 4.0976661715675854e-07,
+      "loss": 5.1492,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8175916230366492,
+      "grad_norm": 2.3402888774871826,
+      "learning_rate": 4.0685422257847134e-07,
+      "loss": 5.0009,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8182617801047121,
+      "grad_norm": 2.1948485374450684,
+      "learning_rate": 4.0395129774043836e-07,
+      "loss": 4.7593,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8189319371727749,
+      "grad_norm": 2.4026002883911133,
+      "learning_rate": 4.010578557759759e-07,
+      "loss": 4.9038,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8196020942408377,
+      "grad_norm": 2.156548500061035,
+      "learning_rate": 3.98173909775498e-07,
+      "loss": 4.9309,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8202722513089005,
+      "grad_norm": 2.5898830890655518,
+      "learning_rate": 3.9529947278645563e-07,
+      "loss": 4.859,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8209424083769633,
+      "grad_norm": 2.2762200832366943,
+      "learning_rate": 3.9243455781328153e-07,
+      "loss": 4.9816,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8216125654450261,
+      "grad_norm": 2.2160186767578125,
+      "learning_rate": 3.8957917781732883e-07,
+      "loss": 5.0222,
+      "step": 1226
+    },
+    {
+      "epoch": 0.822282722513089,
+      "grad_norm": 1.7954459190368652,
+      "learning_rate": 3.867333457168115e-07,
+      "loss": 4.9616,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8229528795811518,
+      "grad_norm": 2.0955536365509033,
+      "learning_rate": 3.8389707438674875e-07,
+      "loss": 5.2389,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8236230366492147,
+      "grad_norm": 2.1967079639434814,
+      "learning_rate": 3.8107037665890386e-07,
+      "loss": 5.0087,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8242931937172775,
+      "grad_norm": 2.6518898010253906,
+      "learning_rate": 3.7825326532172887e-07,
+      "loss": 4.9795,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8249633507853403,
+      "grad_norm": 2.064209461212158,
+      "learning_rate": 3.754457531203048e-07,
+      "loss": 5.0624,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8256335078534032,
+      "grad_norm": 1.7200958728790283,
+      "learning_rate": 3.726478527562846e-07,
+      "loss": 4.7556,
+      "step": 1232
+    },
+    {
+      "epoch": 0.826303664921466,
+      "grad_norm": 2.319408655166626,
+      "learning_rate": 3.698595768878363e-07,
+      "loss": 4.7299,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8269738219895288,
+      "grad_norm": 2.6238155364990234,
+      "learning_rate": 3.6708093812958396e-07,
+      "loss": 5.0846,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8276439790575916,
+      "grad_norm": 2.5130534172058105,
+      "learning_rate": 3.643119490525518e-07,
+      "loss": 5.0706,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8283141361256544,
+      "grad_norm": 2.467212438583374,
+      "learning_rate": 3.615526221841084e-07,
+      "loss": 4.9499,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8289842931937172,
+      "grad_norm": 2.0987708568573,
+      "learning_rate": 3.5880297000790847e-07,
+      "loss": 5.1668,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8296544502617801,
+      "grad_norm": 2.1706671714782715,
+      "learning_rate": 3.5606300496383584e-07,
+      "loss": 4.9499,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8303246073298429,
+      "grad_norm": 2.0631673336029053,
+      "learning_rate": 3.533327394479491e-07,
+      "loss": 4.9016,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8309947643979058,
+      "grad_norm": 2.171215772628784,
+      "learning_rate": 3.506121858124253e-07,
+      "loss": 5.0346,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8316649214659686,
+      "grad_norm": 2.081542730331421,
+      "learning_rate": 3.479013563655012e-07,
+      "loss": 4.8801,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8323350785340314,
+      "grad_norm": 2.2110610008239746,
+      "learning_rate": 3.452002633714219e-07,
+      "loss": 5.0428,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8330052356020943,
+      "grad_norm": 1.9400110244750977,
+      "learning_rate": 3.425089190503808e-07,
+      "loss": 4.9467,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8336753926701571,
+      "grad_norm": 2.2724132537841797,
+      "learning_rate": 3.39827335578469e-07,
+      "loss": 4.9223,
+      "step": 1244
+    },
+    {
+      "epoch": 0.83434554973822,
+      "grad_norm": 1.9748417139053345,
+      "learning_rate": 3.371555250876168e-07,
+      "loss": 5.0014,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8350157068062827,
+      "grad_norm": 2.634237766265869,
+      "learning_rate": 3.3449349966553995e-07,
+      "loss": 4.7432,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8356858638743455,
+      "grad_norm": 2.3327419757843018,
+      "learning_rate": 3.3184127135568596e-07,
+      "loss": 5.1241,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8363560209424084,
+      "grad_norm": 2.305849075317383,
+      "learning_rate": 3.2919885215717626e-07,
+      "loss": 4.9703,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8370261780104712,
+      "grad_norm": 2.3055591583251953,
+      "learning_rate": 3.2656625402475576e-07,
+      "loss": 5.0756,
+      "step": 1249
+    },
+    {
+      "epoch": 0.837696335078534,
+      "grad_norm": 2.3185768127441406,
+      "learning_rate": 3.2394348886873735e-07,
+      "loss": 5.052,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8383664921465969,
+      "grad_norm": 2.2942299842834473,
+      "learning_rate": 3.2133056855494753e-07,
+      "loss": 5.14,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8390366492146597,
+      "grad_norm": 2.267603874206543,
+      "learning_rate": 3.187275049046737e-07,
+      "loss": 5.2036,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8397068062827225,
+      "grad_norm": 2.073228120803833,
+      "learning_rate": 3.1613430969460844e-07,
+      "loss": 4.7733,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8403769633507854,
+      "grad_norm": 2.136348009109497,
+      "learning_rate": 3.135509946568002e-07,
+      "loss": 5.0415,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8410471204188482,
+      "grad_norm": 2.260221481323242,
+      "learning_rate": 3.1097757147859566e-07,
+      "loss": 5.1981,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8417172774869109,
+      "grad_norm": 2.2212343215942383,
+      "learning_rate": 3.084140518025902e-07,
+      "loss": 4.9098,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8423874345549738,
+      "grad_norm": 2.0503504276275635,
+      "learning_rate": 3.058604472265747e-07,
+      "loss": 5.1323,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8430575916230366,
+      "grad_norm": 2.308495044708252,
+      "learning_rate": 3.033167693034808e-07,
+      "loss": 4.7553,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8437277486910995,
+      "grad_norm": 2.055201768875122,
+      "learning_rate": 3.0078302954133166e-07,
+      "loss": 5.0294,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8443979057591623,
+      "grad_norm": 2.3220279216766357,
+      "learning_rate": 2.98259239403188e-07,
+      "loss": 4.8975,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8450680628272251,
+      "grad_norm": 1.8509376049041748,
+      "learning_rate": 2.957454103070978e-07,
+      "loss": 5.0668,
+      "step": 1261
+    },
+    {
+      "epoch": 0.845738219895288,
+      "grad_norm": 2.187986373901367,
+      "learning_rate": 2.932415536260419e-07,
+      "loss": 4.8605,
+      "step": 1262
+    },
+    {
+      "epoch": 0.8464083769633508,
+      "grad_norm": 2.2834601402282715,
+      "learning_rate": 2.907476806878845e-07,
+      "loss": 5.3201,
+      "step": 1263
+    },
+    {
+      "epoch": 0.8470785340314136,
+      "grad_norm": 2.043067216873169,
+      "learning_rate": 2.882638027753232e-07,
+      "loss": 4.841,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8477486910994765,
+      "grad_norm": 2.645731210708618,
+      "learning_rate": 2.8578993112583525e-07,
+      "loss": 5.2072,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8484188481675393,
+      "grad_norm": 2.133944511413574,
+      "learning_rate": 2.8332607693162845e-07,
+      "loss": 4.9593,
+      "step": 1266
+    },
+    {
+      "epoch": 0.849089005235602,
+      "grad_norm": 2.55283784866333,
+      "learning_rate": 2.808722513395901e-07,
+      "loss": 4.985,
+      "step": 1267
+    },
+    {
+      "epoch": 0.8497591623036649,
+      "grad_norm": 2.150108575820923,
+      "learning_rate": 2.784284654512351e-07,
+      "loss": 5.5307,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8504293193717277,
+      "grad_norm": 2.09555721282959,
+      "learning_rate": 2.7599473032265817e-07,
+      "loss": 5.1041,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8510994764397906,
+      "grad_norm": 2.0397253036499023,
+      "learning_rate": 2.7357105696448254e-07,
+      "loss": 5.3425,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8517696335078534,
+      "grad_norm": 2.5002498626708984,
+      "learning_rate": 2.7115745634181097e-07,
+      "loss": 5.1043,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8524397905759162,
+      "grad_norm": 2.1711158752441406,
+      "learning_rate": 2.687539393741739e-07,
+      "loss": 4.9387,
+      "step": 1272
+    },
+    {
+      "epoch": 0.8531099476439791,
+      "grad_norm": 1.9462813138961792,
+      "learning_rate": 2.663605169354833e-07,
+      "loss": 4.9784,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8537801047120419,
+      "grad_norm": 2.0433311462402344,
+      "learning_rate": 2.639771998539806e-07,
+      "loss": 4.9194,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8544502617801047,
+      "grad_norm": 2.780001640319824,
+      "learning_rate": 2.616039989121899e-07,
+      "loss": 5.0706,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8551204188481676,
+      "grad_norm": 2.471343517303467,
+      "learning_rate": 2.5924092484686746e-07,
+      "loss": 5.118,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8557905759162303,
+      "grad_norm": 2.0888404846191406,
+      "learning_rate": 2.5688798834895335e-07,
+      "loss": 5.0195,
+      "step": 1277
+    },
+    {
+      "epoch": 0.8564607329842931,
+      "grad_norm": 1.878853440284729,
+      "learning_rate": 2.545452000635251e-07,
+      "loss": 5.0919,
+      "step": 1278
+    },
+    {
+      "epoch": 0.857130890052356,
+      "grad_norm": 1.875441312789917,
+      "learning_rate": 2.522125705897471e-07,
+      "loss": 4.9351,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8578010471204188,
+      "grad_norm": 2.194225788116455,
+      "learning_rate": 2.498901104808241e-07,
+      "loss": 5.1229,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8584712041884817,
+      "grad_norm": 2.382415294647217,
+      "learning_rate": 2.4757783024395244e-07,
+      "loss": 5.056,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8591413612565445,
+      "grad_norm": 2.61612868309021,
+      "learning_rate": 2.452757403402728e-07,
+      "loss": 4.9424,
+      "step": 1282
+    },
+    {
+      "epoch": 0.8598115183246073,
+      "grad_norm": 2.4941701889038086,
+      "learning_rate": 2.4298385118482345e-07,
+      "loss": 4.8454,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8604816753926702,
+      "grad_norm": 2.6090404987335205,
+      "learning_rate": 2.4070217314649316e-07,
+      "loss": 5.3559,
+      "step": 1284
+    },
+    {
+      "epoch": 0.861151832460733,
+      "grad_norm": 2.0773582458496094,
+      "learning_rate": 2.3843071654797325e-07,
+      "loss": 5.2277,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8618219895287959,
+      "grad_norm": 2.038578510284424,
+      "learning_rate": 2.3616949166571256e-07,
+      "loss": 5.1219,
+      "step": 1286
+    },
+    {
+      "epoch": 0.8624921465968587,
+      "grad_norm": 2.4837217330932617,
+      "learning_rate": 2.3391850872986815e-07,
+      "loss": 5.0083,
+      "step": 1287
+    },
+    {
+      "epoch": 0.8631623036649214,
+      "grad_norm": 2.193636655807495,
+      "learning_rate": 2.3167777792426237e-07,
+      "loss": 5.0886,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8638324607329843,
+      "grad_norm": 2.2123429775238037,
+      "learning_rate": 2.294473093863353e-07,
+      "loss": 5.1775,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8645026178010471,
+      "grad_norm": 2.225257158279419,
+      "learning_rate": 2.2722711320709745e-07,
+      "loss": 4.8214,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8651727748691099,
+      "grad_norm": 2.601940870285034,
+      "learning_rate": 2.250171994310868e-07,
+      "loss": 5.1491,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8658429319371728,
+      "grad_norm": 2.152582883834839,
+      "learning_rate": 2.2281757805632076e-07,
+      "loss": 5.1693,
+      "step": 1292
+    },
+    {
+      "epoch": 0.8665130890052356,
+      "grad_norm": 2.2037830352783203,
+      "learning_rate": 2.2062825903425316e-07,
+      "loss": 5.0747,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8671832460732984,
+      "grad_norm": 2.5105886459350586,
+      "learning_rate": 2.1844925226972847e-07,
+      "loss": 5.0326,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8678534031413613,
+      "grad_norm": 2.2535853385925293,
+      "learning_rate": 2.1628056762093674e-07,
+      "loss": 4.9301,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8685235602094241,
+      "grad_norm": 2.282524824142456,
+      "learning_rate": 2.1412221489936796e-07,
+      "loss": 4.9776,
+      "step": 1296
+    },
+    {
+      "epoch": 0.869193717277487,
+      "grad_norm": 2.4529056549072266,
+      "learning_rate": 2.119742038697689e-07,
+      "loss": 5.0795,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8698638743455498,
+      "grad_norm": 2.4816536903381348,
+      "learning_rate": 2.098365442501002e-07,
+      "loss": 5.0823,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8705340314136125,
+      "grad_norm": 2.298016309738159,
+      "learning_rate": 2.0770924571149014e-07,
+      "loss": 5.1371,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8712041884816754,
+      "grad_norm": 1.9586299657821655,
+      "learning_rate": 2.0559231787819128e-07,
+      "loss": 4.8267,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8718743455497382,
+      "grad_norm": 2.62386155128479,
+      "learning_rate": 2.0348577032753892e-07,
+      "loss": 4.8385,
+      "step": 1301
+    },
+    {
+      "epoch": 0.872544502617801,
+      "grad_norm": 1.8399372100830078,
+      "learning_rate": 2.0138961258990397e-07,
+      "loss": 5.1666,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8732146596858639,
+      "grad_norm": 2.1892104148864746,
+      "learning_rate": 1.9930385414865388e-07,
+      "loss": 5.0949,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8738848167539267,
+      "grad_norm": 2.136589527130127,
+      "learning_rate": 1.9722850444010704e-07,
+      "loss": 4.9039,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8745549738219895,
+      "grad_norm": 2.5759475231170654,
+      "learning_rate": 1.9516357285349202e-07,
+      "loss": 5.0665,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8752251308900524,
+      "grad_norm": 2.103248119354248,
+      "learning_rate": 1.931090687309023e-07,
+      "loss": 4.8426,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8758952879581152,
+      "grad_norm": 2.2694711685180664,
+      "learning_rate": 1.9106500136725742e-07,
+      "loss": 5.2192,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8765654450261781,
+      "grad_norm": 1.8070567846298218,
+      "learning_rate": 1.890313800102589e-07,
+      "loss": 5.0335,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8772356020942408,
+      "grad_norm": 1.9571316242218018,
+      "learning_rate": 1.870082138603485e-07,
+      "loss": 5.0606,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8779057591623036,
+      "grad_norm": 2.1865131855010986,
+      "learning_rate": 1.849955120706673e-07,
+      "loss": 4.8543,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8785759162303665,
+      "grad_norm": 2.1236016750335693,
+      "learning_rate": 1.8299328374701246e-07,
+      "loss": 4.9756,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8792460732984293,
+      "grad_norm": 2.2098305225372314,
+      "learning_rate": 1.8100153794779946e-07,
+      "loss": 5.0787,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8799162303664921,
+      "grad_norm": 2.4154696464538574,
+      "learning_rate": 1.790202836840177e-07,
+      "loss": 4.9587,
+      "step": 1313
+    },
+    {
+      "epoch": 0.880586387434555,
+      "grad_norm": 2.911160707473755,
+      "learning_rate": 1.7704952991919184e-07,
+      "loss": 4.8429,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8812565445026178,
+      "grad_norm": 2.913119077682495,
+      "learning_rate": 1.7508928556934062e-07,
+      "loss": 5.0933,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8819267015706806,
+      "grad_norm": 2.209881544113159,
+      "learning_rate": 1.731395595029353e-07,
+      "loss": 5.0659,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8825968586387435,
+      "grad_norm": 1.6870235204696655,
+      "learning_rate": 1.712003605408624e-07,
+      "loss": 4.7823,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8832670157068063,
+      "grad_norm": 2.3089218139648438,
+      "learning_rate": 1.6927169745638016e-07,
+      "loss": 4.9966,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8839371727748692,
+      "grad_norm": 2.4036853313446045,
+      "learning_rate": 1.6735357897508216e-07,
+      "loss": 5.3279,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8846073298429319,
+      "grad_norm": 2.5057334899902344,
+      "learning_rate": 1.6544601377485648e-07,
+      "loss": 5.2201,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8852774869109947,
+      "grad_norm": 2.2471020221710205,
+      "learning_rate": 1.6354901048584477e-07,
+      "loss": 5.0172,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8859476439790576,
+      "grad_norm": 2.294185161590576,
+      "learning_rate": 1.6166257769040712e-07,
+      "loss": 5.1012,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8866178010471204,
+      "grad_norm": 2.232290267944336,
+      "learning_rate": 1.5978672392307935e-07,
+      "loss": 4.9362,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8872879581151832,
+      "grad_norm": 3.2191405296325684,
+      "learning_rate": 1.5792145767053623e-07,
+      "loss": 4.772,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8879581151832461,
+      "grad_norm": 2.6714560985565186,
+      "learning_rate": 1.560667873715538e-07,
+      "loss": 4.962,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8886282722513089,
+      "grad_norm": 2.2959508895874023,
+      "learning_rate": 1.5422272141696866e-07,
+      "loss": 5.0707,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8892984293193718,
+      "grad_norm": 2.202547073364258,
+      "learning_rate": 1.5238926814964276e-07,
+      "loss": 5.0551,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8899685863874346,
+      "grad_norm": 2.037724256515503,
+      "learning_rate": 1.5056643586442403e-07,
+      "loss": 4.9772,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8906387434554974,
+      "grad_norm": 2.147706985473633,
+      "learning_rate": 1.4875423280810986e-07,
+      "loss": 4.984,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8913089005235602,
+      "grad_norm": 2.203908681869507,
+      "learning_rate": 1.4695266717940887e-07,
+      "loss": 4.8907,
+      "step": 1330
+    },
+    {
+      "epoch": 0.891979057591623,
+      "grad_norm": 2.007119655609131,
+      "learning_rate": 1.4516174712890406e-07,
+      "loss": 5.0029,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8926492146596858,
+      "grad_norm": 2.464463949203491,
+      "learning_rate": 1.4338148075901558e-07,
+      "loss": 5.1036,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8933193717277487,
+      "grad_norm": 2.6474461555480957,
+      "learning_rate": 1.4161187612396543e-07,
+      "loss": 4.9669,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8939895287958115,
+      "grad_norm": 2.2933366298675537,
+      "learning_rate": 1.3985294122973974e-07,
+      "loss": 5.3269,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8946596858638743,
+      "grad_norm": 2.7613108158111572,
+      "learning_rate": 1.3810468403405358e-07,
+      "loss": 4.5326,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8953298429319372,
+      "grad_norm": 2.812737464904785,
+      "learning_rate": 1.3636711244631312e-07,
+      "loss": 4.9299,
+      "step": 1336
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 2.1628639698028564,
+      "learning_rate": 1.346402343275824e-07,
+      "loss": 5.2307,
+      "step": 1337
+    },
+    {
+      "epoch": 0.8966701570680629,
+      "grad_norm": 2.1737706661224365,
+      "learning_rate": 1.329240574905452e-07,
+      "loss": 5.2518,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8973403141361257,
+      "grad_norm": 2.3390467166900635,
+      "learning_rate": 1.3121858969947194e-07,
+      "loss": 5.104,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8980104712041885,
+      "grad_norm": 2.431335210800171,
+      "learning_rate": 1.295238386701822e-07,
+      "loss": 5.0695,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8986806282722513,
+      "grad_norm": 2.235074043273926,
+      "learning_rate": 1.2783981207001277e-07,
+      "loss": 4.969,
+      "step": 1341
+    },
+    {
+      "epoch": 0.8993507853403141,
+      "grad_norm": 2.1779608726501465,
+      "learning_rate": 1.2616651751778035e-07,
+      "loss": 5.036,
+      "step": 1342
+    },
+    {
+      "epoch": 0.9000209424083769,
+      "grad_norm": 2.16412353515625,
+      "learning_rate": 1.2450396258374854e-07,
+      "loss": 5.0072,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9006910994764398,
+      "grad_norm": 2.2823896408081055,
+      "learning_rate": 1.2285215478959278e-07,
+      "loss": 4.9887,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9013612565445026,
+      "grad_norm": 2.1547162532806396,
+      "learning_rate": 1.2121110160836697e-07,
+      "loss": 4.7598,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9020314136125654,
+      "grad_norm": 2.2301156520843506,
+      "learning_rate": 1.1958081046446828e-07,
+      "loss": 4.9489,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9027015706806283,
+      "grad_norm": 2.637312412261963,
+      "learning_rate": 1.1796128873360624e-07,
+      "loss": 5.1715,
+      "step": 1347
+    },
+    {
+      "epoch": 0.9033717277486911,
+      "grad_norm": 1.954455018043518,
+      "learning_rate": 1.1635254374276649e-07,
+      "loss": 4.8564,
+      "step": 1348
+    },
+    {
+      "epoch": 0.904041884816754,
+      "grad_norm": 2.5579347610473633,
+      "learning_rate": 1.1475458277018042e-07,
+      "loss": 4.9343,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9047120418848168,
+      "grad_norm": 2.069652557373047,
+      "learning_rate": 1.1316741304528839e-07,
+      "loss": 5.1306,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9053821989528796,
+      "grad_norm": 2.0289247035980225,
+      "learning_rate": 1.1159104174871177e-07,
+      "loss": 5.1948,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9060523560209424,
+      "grad_norm": 2.011836528778076,
+      "learning_rate": 1.100254760122163e-07,
+      "loss": 5.0701,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9067225130890052,
+      "grad_norm": 2.4269700050354004,
+      "learning_rate": 1.084707229186821e-07,
+      "loss": 5.0405,
+      "step": 1353
+    },
+    {
+      "epoch": 0.907392670157068,
+      "grad_norm": 1.9985160827636719,
+      "learning_rate": 1.069267895020723e-07,
+      "loss": 4.858,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9080628272251309,
+      "grad_norm": 2.3926138877868652,
+      "learning_rate": 1.0539368274739753e-07,
+      "loss": 5.2444,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9087329842931937,
+      "grad_norm": 2.214724540710449,
+      "learning_rate": 1.0387140959068898e-07,
+      "loss": 4.9224,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9094031413612566,
+      "grad_norm": 2.054424285888672,
+      "learning_rate": 1.0235997691896398e-07,
+      "loss": 4.9698,
+      "step": 1357
+    },
+    {
+      "epoch": 0.9100732984293194,
+      "grad_norm": 2.2345471382141113,
+      "learning_rate": 1.0085939157019609e-07,
+      "loss": 5.2301,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9107434554973822,
+      "grad_norm": 2.892404794692993,
+      "learning_rate": 9.936966033328277e-08,
+      "loss": 4.9301,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9114136125654451,
+      "grad_norm": 2.082559823989868,
+      "learning_rate": 9.789078994801665e-08,
+      "loss": 5.0013,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9120837696335079,
+      "grad_norm": 2.0129618644714355,
+      "learning_rate": 9.642278710505354e-08,
+      "loss": 4.9005,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9127539267015706,
+      "grad_norm": 2.6280977725982666,
+      "learning_rate": 9.496565844588329e-08,
+      "loss": 5.0724,
+      "step": 1362
+    },
+    {
+      "epoch": 0.9134240837696335,
+      "grad_norm": 1.7958965301513672,
+      "learning_rate": 9.351941056279845e-08,
+      "loss": 5.0119,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9140942408376963,
+      "grad_norm": 2.725520372390747,
+      "learning_rate": 9.208404999886594e-08,
+      "loss": 5.0868,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9147643979057591,
+      "grad_norm": 2.4962949752807617,
+      "learning_rate": 9.065958324789542e-08,
+      "loss": 4.9649,
+      "step": 1365
+    },
+    {
+      "epoch": 0.915434554973822,
+      "grad_norm": 1.9726396799087524,
+      "learning_rate": 8.924601675441207e-08,
+      "loss": 4.7758,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9161047120418848,
+      "grad_norm": 1.9447152614593506,
+      "learning_rate": 8.784335691362638e-08,
+      "loss": 5.0546,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9167748691099477,
+      "grad_norm": 1.860182523727417,
+      "learning_rate": 8.645161007140468e-08,
+      "loss": 5.1615,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9174450261780105,
+      "grad_norm": 1.9219021797180176,
+      "learning_rate": 8.50707825242425e-08,
+      "loss": 5.3586,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9181151832460733,
+      "grad_norm": 2.7881152629852295,
+      "learning_rate": 8.370088051923297e-08,
+      "loss": 5.122,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9187853403141362,
+      "grad_norm": 2.0921435356140137,
+      "learning_rate": 8.234191025404126e-08,
+      "loss": 4.9688,
+      "step": 1371
+    },
+    {
+      "epoch": 0.919455497382199,
+      "grad_norm": 1.88462495803833,
+      "learning_rate": 8.099387787687541e-08,
+      "loss": 4.6999,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9201256544502617,
+      "grad_norm": 2.3097009658813477,
+      "learning_rate": 7.965678948645833e-08,
+      "loss": 5.2093,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9207958115183246,
+      "grad_norm": 1.9321610927581787,
+      "learning_rate": 7.833065113200034e-08,
+      "loss": 5.155,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9214659685863874,
+      "grad_norm": 2.0562613010406494,
+      "learning_rate": 7.701546881317273e-08,
+      "loss": 4.8406,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9221361256544502,
+      "grad_norm": 2.438230037689209,
+      "learning_rate": 7.571124848007954e-08,
+      "loss": 5.0072,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9228062827225131,
+      "grad_norm": 2.011482000350952,
+      "learning_rate": 7.441799603323058e-08,
+      "loss": 4.9704,
+      "step": 1377
+    },
+    {
+      "epoch": 0.9234764397905759,
+      "grad_norm": 2.19347882270813,
+      "learning_rate": 7.313571732351588e-08,
+      "loss": 4.7632,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9241465968586388,
+      "grad_norm": 1.9491517543792725,
+      "learning_rate": 7.186441815217771e-08,
+      "loss": 5.0813,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9248167539267016,
+      "grad_norm": 2.243320941925049,
+      "learning_rate": 7.060410427078473e-08,
+      "loss": 4.8204,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9254869109947644,
+      "grad_norm": 2.298475503921509,
+      "learning_rate": 6.93547813812076e-08,
+      "loss": 4.7645,
+      "step": 1381
+    },
+    {
+      "epoch": 0.9261570680628273,
+      "grad_norm": 1.9371122121810913,
+      "learning_rate": 6.811645513559118e-08,
+      "loss": 5.1322,
+      "step": 1382
+    },
+    {
+      "epoch": 0.92682722513089,
+      "grad_norm": 2.0925161838531494,
+      "learning_rate": 6.688913113632983e-08,
+      "loss": 5.0049,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9274973821989528,
+      "grad_norm": 2.1542677879333496,
+      "learning_rate": 6.567281493604139e-08,
+      "loss": 5.146,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9281675392670157,
+      "grad_norm": 2.136967897415161,
+      "learning_rate": 6.446751203754347e-08,
+      "loss": 4.8245,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9288376963350785,
+      "grad_norm": 2.5279605388641357,
+      "learning_rate": 6.327322789382695e-08,
+      "loss": 4.8866,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9295078534031413,
+      "grad_norm": 2.2963929176330566,
+      "learning_rate": 6.208996790803223e-08,
+      "loss": 4.6499,
+      "step": 1387
+    },
+    {
+      "epoch": 0.9301780104712042,
+      "grad_norm": 2.017291784286499,
+      "learning_rate": 6.091773743342493e-08,
+      "loss": 5.1354,
+      "step": 1388
+    },
+    {
+      "epoch": 0.930848167539267,
+      "grad_norm": 2.0148367881774902,
+      "learning_rate": 5.97565417733706e-08,
+      "loss": 4.6275,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9315183246073299,
+      "grad_norm": 2.604938268661499,
+      "learning_rate": 5.8606386181312147e-08,
+      "loss": 5.2862,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9321884816753927,
+      "grad_norm": 1.9097956418991089,
+      "learning_rate": 5.74672758607453e-08,
+      "loss": 4.7748,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9328586387434555,
+      "grad_norm": 2.368349552154541,
+      "learning_rate": 5.633921596519487e-08,
+      "loss": 5.1818,
+      "step": 1392
+    },
+    {
+      "epoch": 0.9335287958115184,
+      "grad_norm": 2.3302204608917236,
+      "learning_rate": 5.522221159819152e-08,
+      "loss": 5.0694,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9341989528795811,
+      "grad_norm": 2.0187971591949463,
+      "learning_rate": 5.411626781324897e-08,
+      "loss": 5.2536,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9348691099476439,
+      "grad_norm": 2.059431314468384,
+      "learning_rate": 5.302138961384179e-08,
+      "loss": 4.9142,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9355392670157068,
+      "grad_norm": 2.138587713241577,
+      "learning_rate": 5.193758195338072e-08,
+      "loss": 5.185,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9362094240837696,
+      "grad_norm": 1.9335336685180664,
+      "learning_rate": 5.0864849735192676e-08,
+      "loss": 4.9316,
+      "step": 1397
+    },
+    {
+      "epoch": 0.9368795811518325,
+      "grad_norm": 1.8651350736618042,
+      "learning_rate": 4.980319781249687e-08,
+      "loss": 4.8454,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9375497382198953,
+      "grad_norm": 2.0417473316192627,
+      "learning_rate": 4.875263098838345e-08,
+      "loss": 4.9571,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9382198952879581,
+      "grad_norm": 1.8700650930404663,
+      "learning_rate": 4.771315401579185e-08,
+      "loss": 4.6554,
+      "step": 1400
+    },
+    {
+      "epoch": 0.938890052356021,
+      "grad_norm": 2.453132390975952,
+      "learning_rate": 4.668477159748858e-08,
+      "loss": 4.8134,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9395602094240838,
+      "grad_norm": 1.904445767402649,
+      "learning_rate": 4.5667488386047784e-08,
+      "loss": 5.1193,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9402303664921466,
+      "grad_norm": 1.8896658420562744,
+      "learning_rate": 4.466130898382798e-08,
+      "loss": 4.8977,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9409005235602095,
+      "grad_norm": 2.006617546081543,
+      "learning_rate": 4.366623794295199e-08,
+      "loss": 5.0047,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9415706806282722,
+      "grad_norm": 1.9613410234451294,
+      "learning_rate": 4.2682279765287324e-08,
+      "loss": 5.0321,
+      "step": 1405
+    },
+    {
+      "epoch": 0.942240837696335,
+      "grad_norm": 2.4916062355041504,
+      "learning_rate": 4.170943890242501e-08,
+      "loss": 4.7501,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9429109947643979,
+      "grad_norm": 2.0068774223327637,
+      "learning_rate": 4.074771975565911e-08,
+      "loss": 4.8988,
+      "step": 1407
+    },
+    {
+      "epoch": 0.9435811518324607,
+      "grad_norm": 2.543067216873169,
+      "learning_rate": 3.979712667596669e-08,
+      "loss": 5.1493,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9442513089005236,
+      "grad_norm": 2.2659168243408203,
+      "learning_rate": 3.8857663963989825e-08,
+      "loss": 4.9611,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9449214659685864,
+      "grad_norm": 2.5053818225860596,
+      "learning_rate": 3.792933587001418e-08,
+      "loss": 4.8455,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9455916230366492,
+      "grad_norm": 2.22271990776062,
+      "learning_rate": 3.701214659395047e-08,
+      "loss": 4.995,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9462617801047121,
+      "grad_norm": 1.9299851655960083,
+      "learning_rate": 3.6106100285316634e-08,
+      "loss": 4.7568,
+      "step": 1412
+    },
+    {
+      "epoch": 0.9469319371727749,
+      "grad_norm": 2.163027763366699,
+      "learning_rate": 3.5211201043216234e-08,
+      "loss": 4.8645,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9476020942408377,
+      "grad_norm": 2.464020252227783,
+      "learning_rate": 3.432745291632289e-08,
+      "loss": 4.8882,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9482722513089005,
+      "grad_norm": 2.6676039695739746,
+      "learning_rate": 3.3454859902860295e-08,
+      "loss": 4.9683,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9489424083769633,
+      "grad_norm": 2.4233171939849854,
+      "learning_rate": 3.2593425950584736e-08,
+      "loss": 5.2675,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9496125654450261,
+      "grad_norm": 1.5752209424972534,
+      "learning_rate": 3.174315495676705e-08,
+      "loss": 4.576,
+      "step": 1417
+    },
+    {
+      "epoch": 0.950282722513089,
+      "grad_norm": 2.3115358352661133,
+      "learning_rate": 3.0904050768174596e-08,
+      "loss": 4.9546,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9509528795811518,
+      "grad_norm": 2.4787373542785645,
+      "learning_rate": 3.007611718105485e-08,
+      "loss": 5.0709,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9516230366492147,
+      "grad_norm": 2.1984169483184814,
+      "learning_rate": 2.9259357941117117e-08,
+      "loss": 4.6754,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9522931937172775,
+      "grad_norm": 2.377700090408325,
+      "learning_rate": 2.8453776743516127e-08,
+      "loss": 5.0916,
+      "step": 1421
+    },
+    {
+      "epoch": 0.9529633507853403,
+      "grad_norm": 1.8702716827392578,
+      "learning_rate": 2.7659377232835683e-08,
+      "loss": 4.9365,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9536335078534032,
+      "grad_norm": 1.924991250038147,
+      "learning_rate": 2.687616300307172e-08,
+      "loss": 4.8889,
+      "step": 1423
+    },
+    {
+      "epoch": 0.954303664921466,
+      "grad_norm": 2.0711193084716797,
+      "learning_rate": 2.6104137597615375e-08,
+      "loss": 4.8053,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9549738219895288,
+      "grad_norm": 1.9230101108551025,
+      "learning_rate": 2.5343304509238552e-08,
+      "loss": 5.077,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9556439790575916,
+      "grad_norm": 1.9874987602233887,
+      "learning_rate": 2.4593667180076996e-08,
+      "loss": 4.8518,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9563141361256544,
+      "grad_norm": 2.0824201107025146,
+      "learning_rate": 2.3855229001614744e-08,
+      "loss": 5.1043,
+      "step": 1427
+    },
+    {
+      "epoch": 0.9569842931937173,
+      "grad_norm": 2.187917947769165,
+      "learning_rate": 2.3127993314669138e-08,
+      "loss": 4.8822,
+      "step": 1428
+    },
+    {
+      "epoch": 0.9576544502617801,
+      "grad_norm": 1.7651984691619873,
+      "learning_rate": 2.241196340937557e-08,
+      "loss": 5.0213,
+      "step": 1429
+    },
+    {
+      "epoch": 0.9583246073298429,
+      "grad_norm": 2.359253168106079,
+      "learning_rate": 2.170714252517275e-08,
+      "loss": 5.1379,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9589947643979058,
+      "grad_norm": 1.8880469799041748,
+      "learning_rate": 2.1013533850787747e-08,
+      "loss": 4.7338,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9596649214659686,
+      "grad_norm": 2.318695068359375,
+      "learning_rate": 2.033114052422208e-08,
+      "loss": 4.7872,
+      "step": 1432
+    },
+    {
+      "epoch": 0.9603350785340314,
+      "grad_norm": 2.3342957496643066,
+      "learning_rate": 1.96599656327362e-08,
+      "loss": 4.7796,
+      "step": 1433
+    },
+    {
+      "epoch": 0.9610052356020943,
+      "grad_norm": 1.8491644859313965,
+      "learning_rate": 1.9000012212837537e-08,
+      "loss": 4.8163,
+      "step": 1434
+    },
+    {
+      "epoch": 0.9616753926701571,
+      "grad_norm": 2.1779308319091797,
+      "learning_rate": 1.835128325026525e-08,
+      "loss": 5.0043,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9623455497382198,
+      "grad_norm": 2.096520185470581,
+      "learning_rate": 1.771378167997745e-08,
+      "loss": 4.9094,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9630157068062827,
+      "grad_norm": 2.06124210357666,
+      "learning_rate": 1.7087510386137042e-08,
+      "loss": 5.0354,
+      "step": 1437
+    },
+    {
+      "epoch": 0.9636858638743455,
+      "grad_norm": 2.6854307651519775,
+      "learning_rate": 1.6472472202100074e-08,
+      "loss": 5.0915,
+      "step": 1438
+    },
+    {
+      "epoch": 0.9643560209424084,
+      "grad_norm": 2.3049659729003906,
+      "learning_rate": 1.586866991040187e-08,
+      "loss": 5.0065,
+      "step": 1439
+    },
+    {
+      "epoch": 0.9650261780104712,
+      "grad_norm": 2.1497740745544434,
+      "learning_rate": 1.5276106242744225e-08,
+      "loss": 4.9907,
+      "step": 1440
+    },
+    {
+      "epoch": 0.965696335078534,
+      "grad_norm": 2.1462044715881348,
+      "learning_rate": 1.4694783879984076e-08,
+      "loss": 4.8249,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9663664921465969,
+      "grad_norm": 2.2082231044769287,
+      "learning_rate": 1.4124705452120691e-08,
+      "loss": 5.104,
+      "step": 1442
+    },
+    {
+      "epoch": 0.9670366492146597,
+      "grad_norm": 2.118767023086548,
+      "learning_rate": 1.3565873538283758e-08,
+      "loss": 4.9168,
+      "step": 1443
+    },
+    {
+      "epoch": 0.9677068062827225,
+      "grad_norm": 1.9635920524597168,
+      "learning_rate": 1.3018290666721723e-08,
+      "loss": 5.0158,
+      "step": 1444
+    },
+    {
+      "epoch": 0.9683769633507854,
+      "grad_norm": 2.3437256813049316,
+      "learning_rate": 1.2481959314791514e-08,
+      "loss": 5.1937,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9690471204188482,
+      "grad_norm": 1.8330011367797852,
+      "learning_rate": 1.195688190894495e-08,
+      "loss": 4.8805,
+      "step": 1446
+    },
+    {
+      "epoch": 0.969717277486911,
+      "grad_norm": 2.258427858352661,
+      "learning_rate": 1.1443060824719576e-08,
+      "loss": 4.9438,
+      "step": 1447
+    },
+    {
+      "epoch": 0.9703874345549738,
+      "grad_norm": 2.3859939575195312,
+      "learning_rate": 1.0940498386728116e-08,
+      "loss": 5.1079,
+      "step": 1448
+    },
+    {
+      "epoch": 0.9710575916230366,
+      "grad_norm": 1.903747797012329,
+      "learning_rate": 1.0449196868646261e-08,
+      "loss": 5.1166,
+      "step": 1449
+    },
+    {
+      "epoch": 0.9717277486910995,
+      "grad_norm": 2.1217775344848633,
+      "learning_rate": 9.969158493204067e-09,
+      "loss": 4.7963,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9723979057591623,
+      "grad_norm": 2.065089464187622,
+      "learning_rate": 9.500385432175684e-09,
+      "loss": 5.1951,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9730680628272251,
+      "grad_norm": 2.248767852783203,
+      "learning_rate": 9.0428798063677e-09,
+      "loss": 4.9754,
+      "step": 1452
+    },
+    {
+      "epoch": 0.973738219895288,
+      "grad_norm": 2.0079345703125,
+      "learning_rate": 8.596643685612194e-09,
+      "loss": 4.9236,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9744083769633508,
+      "grad_norm": 2.086193799972534,
+      "learning_rate": 8.161679088755925e-09,
+      "loss": 5.1244,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9750785340314136,
+      "grad_norm": 2.2422029972076416,
+      "learning_rate": 7.737987983650607e-09,
+      "loss": 5.0332,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9757486910994765,
+      "grad_norm": 2.0613884925842285,
+      "learning_rate": 7.325572287146254e-09,
+      "loss": 4.8539,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9764188481675393,
+      "grad_norm": 2.1857516765594482,
+      "learning_rate": 6.9244338650792385e-09,
+      "loss": 4.835,
+      "step": 1457
+    },
+    {
+      "epoch": 0.977089005235602,
+      "grad_norm": 2.3052470684051514,
+      "learning_rate": 6.534574532267302e-09,
+      "loss": 4.8526,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9777591623036649,
+      "grad_norm": 2.3822693824768066,
+      "learning_rate": 6.155996052499002e-09,
+      "loss": 5.2308,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9784293193717277,
+      "grad_norm": 2.422942876815796,
+      "learning_rate": 5.788700138525949e-09,
+      "loss": 4.9636,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9790994764397906,
+      "grad_norm": 1.6775134801864624,
+      "learning_rate": 5.432688452056689e-09,
+      "loss": 5.1564,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9797696335078534,
+      "grad_norm": 2.2838222980499268,
+      "learning_rate": 5.087962603746999e-09,
+      "loss": 5.0561,
+      "step": 1462
+    },
+    {
+      "epoch": 0.9804397905759162,
+      "grad_norm": 2.447178363800049,
+      "learning_rate": 4.754524153194051e-09,
+      "loss": 5.1673,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9811099476439791,
+      "grad_norm": 2.1302945613861084,
+      "learning_rate": 4.432374608929202e-09,
+      "loss": 4.8763,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9817801047120419,
+      "grad_norm": 1.8921369314193726,
+      "learning_rate": 4.121515428410217e-09,
+      "loss": 5.0894,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9824502617801047,
+      "grad_norm": 1.892519235610962,
+      "learning_rate": 3.821948018016552e-09,
+      "loss": 4.8003,
+      "step": 1466
+    },
+    {
+      "epoch": 0.9831204188481676,
+      "grad_norm": 1.7763696908950806,
+      "learning_rate": 3.533673733040477e-09,
+      "loss": 4.9118,
+      "step": 1467
+    },
+    {
+      "epoch": 0.9837905759162303,
+      "grad_norm": 2.504063844680786,
+      "learning_rate": 3.2566938776831814e-09,
+      "loss": 5.0741,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9844607329842932,
+      "grad_norm": 2.338676929473877,
+      "learning_rate": 2.9910097050472877e-09,
+      "loss": 4.751,
+      "step": 1469
+    },
+    {
+      "epoch": 0.985130890052356,
+      "grad_norm": 2.104109048843384,
+      "learning_rate": 2.736622417132684e-09,
+      "loss": 5.079,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9858010471204188,
+      "grad_norm": 1.9825674295425415,
+      "learning_rate": 2.4935331648298644e-09,
+      "loss": 4.7784,
+      "step": 1471
+    },
+    {
+      "epoch": 0.9864712041884817,
+      "grad_norm": 2.10986328125,
+      "learning_rate": 2.2617430479149327e-09,
+      "loss": 4.8066,
+      "step": 1472
+    },
+    {
+      "epoch": 0.9871413612565445,
+      "grad_norm": 2.6451609134674072,
+      "learning_rate": 2.0412531150448833e-09,
+      "loss": 5.0877,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9878115183246073,
+      "grad_norm": 2.6569290161132812,
+      "learning_rate": 1.832064363753161e-09,
+      "loss": 4.9329,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9884816753926702,
+      "grad_norm": 2.2146356105804443,
+      "learning_rate": 1.634177740444387e-09,
+      "loss": 5.0452,
+      "step": 1475
+    },
+    {
+      "epoch": 0.989151832460733,
+      "grad_norm": 2.198579788208008,
+      "learning_rate": 1.4475941403904737e-09,
+      "loss": 4.8974,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9898219895287959,
+      "grad_norm": 2.7071197032928467,
+      "learning_rate": 1.272314407727293e-09,
+      "loss": 4.9834,
+      "step": 1477
+    },
+    {
+      "epoch": 0.9904921465968587,
+      "grad_norm": 1.946290135383606,
+      "learning_rate": 1.1083393354488492e-09,
+      "loss": 4.9904,
+      "step": 1478
+    },
+    {
+      "epoch": 0.9911623036649214,
+      "grad_norm": 2.483842611312866,
+      "learning_rate": 9.556696654058894e-10,
+      "loss": 4.8588,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9918324607329843,
+      "grad_norm": 2.2516233921051025,
+      "learning_rate": 8.143060883017418e-10,
+      "loss": 4.8331,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9925026178010471,
+      "grad_norm": 2.159393787384033,
+      "learning_rate": 6.842492436887061e-10,
+      "loss": 5.2052,
+      "step": 1481
+    },
+    {
+      "epoch": 0.9931727748691099,
+      "grad_norm": 2.3281965255737305,
+      "learning_rate": 5.654997199652789e-10,
+      "loss": 5.2328,
+      "step": 1482
+    },
+    {
+      "epoch": 0.9938429319371728,
+      "grad_norm": 2.1843552589416504,
+      "learning_rate": 4.5805805437448834e-10,
+      "loss": 4.9036,
+      "step": 1483
+    },
+    {
+      "epoch": 0.9945130890052356,
+      "grad_norm": 2.202324867248535,
+      "learning_rate": 3.6192473300000754e-10,
+      "loss": 4.9878,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9951832460732984,
+      "grad_norm": 2.2752175331115723,
+      "learning_rate": 2.7710019076532257e-10,
+      "loss": 4.9163,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9958534031413613,
+      "grad_norm": 2.1237142086029053,
+      "learning_rate": 2.0358481143067932e-10,
+      "loss": 4.9672,
+      "step": 1486
+    },
+    {
+      "epoch": 0.9965235602094241,
+      "grad_norm": 1.9874749183654785,
+      "learning_rate": 1.4137892759169546e-10,
+      "loss": 4.8258,
+      "step": 1487
+    },
+    {
+      "epoch": 0.997193717277487,
+      "grad_norm": 2.537884473800659,
+      "learning_rate": 9.048282067825043e-11,
+      "loss": 5.03,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9978638743455497,
+      "grad_norm": 1.966549277305603,
+      "learning_rate": 5.0896720952819945e-11,
+      "loss": 5.0325,
+      "step": 1489
+    },
+    {
+      "epoch": 0.9985340314136125,
+      "grad_norm": 2.476804733276367,
+      "learning_rate": 2.262080750964346e-11,
+      "loss": 5.1559,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9992041884816754,
+      "grad_norm": 1.8347021341323853,
+      "learning_rate": 5.655208273613877e-12,
+      "loss": 5.0006,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9998743455497382,
+      "grad_norm": 2.3418211936950684,
+      "learning_rate": 0.0,
+      "loss": 5.0768,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9998743455497382,
+      "step": 1492,
+      "total_flos": 2.186841498544033e+19,
+      "train_loss": 5.3191524389282305,
+      "train_runtime": 50561.1303,
+      "train_samples_per_second": 7.555,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1492,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.186841498544033e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}