diff --git "a/outputs_pretrained/checkpoint-846/trainer_state.json" "b/outputs_pretrained/checkpoint-846/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/outputs_pretrained/checkpoint-846/trainer_state.json"
@@ -0,0 +1,5956 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 846,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002364066193853428,
+      "grad_norm": 9.786882400512695,
+      "learning_rate": 0.0,
+      "loss": 3.2811,
+      "step": 1
+    },
+    {
+      "epoch": 0.004728132387706856,
+      "grad_norm": 10.056135177612305,
+      "learning_rate": 5e-06,
+      "loss": 3.2727,
+      "step": 2
+    },
+    {
+      "epoch": 0.0070921985815602835,
+      "grad_norm": 10.060759544372559,
+      "learning_rate": 1e-05,
+      "loss": 3.4002,
+      "step": 3
+    },
+    {
+      "epoch": 0.009456264775413711,
+      "grad_norm": 8.956620216369629,
+      "learning_rate": 1.5e-05,
+      "loss": 3.3498,
+      "step": 4
+    },
+    {
+      "epoch": 0.01182033096926714,
+      "grad_norm": 7.450173854827881,
+      "learning_rate": 2e-05,
+      "loss": 3.3485,
+      "step": 5
+    },
+    {
+      "epoch": 0.014184397163120567,
+      "grad_norm": 7.106024265289307,
+      "learning_rate": 2.5e-05,
+      "loss": 3.1664,
+      "step": 6
+    },
+    {
+      "epoch": 0.016548463356973995,
+      "grad_norm": 7.5257415771484375,
+      "learning_rate": 3e-05,
+      "loss": 3.087,
+      "step": 7
+    },
+    {
+      "epoch": 0.018912529550827423,
+      "grad_norm": 7.290092468261719,
+      "learning_rate": 3.5e-05,
+      "loss": 2.9256,
+      "step": 8
+    },
+    {
+      "epoch": 0.02127659574468085,
+      "grad_norm": 6.230826377868652,
+      "learning_rate": 4e-05,
+      "loss": 2.83,
+      "step": 9
+    },
+    {
+      "epoch": 0.02364066193853428,
+      "grad_norm": 7.3913679122924805,
+      "learning_rate": 4.5e-05,
+      "loss": 2.8809,
+      "step": 10
+    },
+    {
+      "epoch": 0.026004728132387706,
+      "grad_norm": 7.500810146331787,
+      "learning_rate": 5e-05,
+      "loss": 2.9329,
+      "step": 11
+    },
+    {
+      "epoch": 0.028368794326241134,
+      "grad_norm": 6.774663925170898,
+      "learning_rate": 4.994019138755981e-05,
+      "loss": 2.8633,
+      "step": 12
+    },
+    {
+      "epoch": 0.030732860520094562,
+      "grad_norm": 8.217077255249023,
+      "learning_rate": 4.988038277511962e-05,
+      "loss": 2.779,
+      "step": 13
+    },
+    {
+      "epoch": 0.03309692671394799,
+      "grad_norm": 6.935091495513916,
+      "learning_rate": 4.982057416267943e-05,
+      "loss": 2.7578,
+      "step": 14
+    },
+    {
+      "epoch": 0.03546099290780142,
+      "grad_norm": 7.576630592346191,
+      "learning_rate": 4.9760765550239234e-05,
+      "loss": 2.7023,
+      "step": 15
+    },
+    {
+      "epoch": 0.037825059101654845,
+      "grad_norm": 5.840672492980957,
+      "learning_rate": 4.970095693779905e-05,
+      "loss": 2.6723,
+      "step": 16
+    },
+    {
+      "epoch": 0.04018912529550828,
+      "grad_norm": 5.728421211242676,
+      "learning_rate": 4.964114832535885e-05,
+      "loss": 2.6571,
+      "step": 17
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 5.6147990226745605,
+      "learning_rate": 4.958133971291866e-05,
+      "loss": 2.712,
+      "step": 18
+    },
+    {
+      "epoch": 0.04491725768321513,
+      "grad_norm": 5.601494789123535,
+      "learning_rate": 4.952153110047847e-05,
+      "loss": 2.6546,
+      "step": 19
+    },
+    {
+      "epoch": 0.04728132387706856,
+      "grad_norm": 5.493714332580566,
+      "learning_rate": 4.946172248803828e-05,
+      "loss": 2.599,
+      "step": 20
+    },
+    {
+      "epoch": 0.04964539007092199,
+      "grad_norm": 5.742792129516602,
+      "learning_rate": 4.940191387559809e-05,
+      "loss": 2.6123,
+      "step": 21
+    },
+    {
+      "epoch": 0.05200945626477541,
+      "grad_norm": 4.954222202301025,
+      "learning_rate": 4.9342105263157894e-05,
+      "loss": 2.4297,
+      "step": 22
+    },
+    {
+      "epoch": 0.054373522458628844,
+      "grad_norm": 5.282480239868164,
+      "learning_rate": 4.928229665071771e-05,
+      "loss": 2.5279,
+      "step": 23
+    },
+    {
+      "epoch": 0.05673758865248227,
+      "grad_norm": 5.26470947265625,
+      "learning_rate": 4.922248803827751e-05,
+      "loss": 2.4522,
+      "step": 24
+    },
+    {
+      "epoch": 0.0591016548463357,
+      "grad_norm": 5.375247001647949,
+      "learning_rate": 4.916267942583732e-05,
+      "loss": 2.5739,
+      "step": 25
+    },
+    {
+      "epoch": 0.061465721040189124,
+      "grad_norm": 5.001519203186035,
+      "learning_rate": 4.910287081339713e-05,
+      "loss": 2.5496,
+      "step": 26
+    },
+    {
+      "epoch": 0.06382978723404255,
+      "grad_norm": 5.700991630554199,
+      "learning_rate": 4.904306220095694e-05,
+      "loss": 2.5462,
+      "step": 27
+    },
+    {
+      "epoch": 0.06619385342789598,
+      "grad_norm": 5.051942348480225,
+      "learning_rate": 4.898325358851675e-05,
+      "loss": 2.3694,
+      "step": 28
+    },
+    {
+      "epoch": 0.06855791962174941,
+      "grad_norm": 5.411677837371826,
+      "learning_rate": 4.892344497607656e-05,
+      "loss": 2.5297,
+      "step": 29
+    },
+    {
+      "epoch": 0.07092198581560284,
+      "grad_norm": 4.419405937194824,
+      "learning_rate": 4.886363636363637e-05,
+      "loss": 2.5009,
+      "step": 30
+    },
+    {
+      "epoch": 0.07328605200945626,
+      "grad_norm": 5.185110569000244,
+      "learning_rate": 4.880382775119617e-05,
+      "loss": 2.4033,
+      "step": 31
+    },
+    {
+      "epoch": 0.07565011820330969,
+      "grad_norm": 5.229886054992676,
+      "learning_rate": 4.874401913875598e-05,
+      "loss": 2.4766,
+      "step": 32
+    },
+    {
+      "epoch": 0.07801418439716312,
+      "grad_norm": 4.489797115325928,
+      "learning_rate": 4.868421052631579e-05,
+      "loss": 2.4468,
+      "step": 33
+    },
+    {
+      "epoch": 0.08037825059101655,
+      "grad_norm": 4.7152419090271,
+      "learning_rate": 4.86244019138756e-05,
+      "loss": 2.4374,
+      "step": 34
+    },
+    {
+      "epoch": 0.08274231678486997,
+      "grad_norm": 5.1662278175354,
+      "learning_rate": 4.8564593301435404e-05,
+      "loss": 2.3898,
+      "step": 35
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 4.726010322570801,
+      "learning_rate": 4.850478468899522e-05,
+      "loss": 2.5208,
+      "step": 36
+    },
+    {
+      "epoch": 0.08747044917257683,
+      "grad_norm": 4.929504871368408,
+      "learning_rate": 4.844497607655503e-05,
+      "loss": 2.4105,
+      "step": 37
+    },
+    {
+      "epoch": 0.08983451536643026,
+      "grad_norm": 4.676382064819336,
+      "learning_rate": 4.838516746411483e-05,
+      "loss": 2.3955,
+      "step": 38
+    },
+    {
+      "epoch": 0.09219858156028368,
+      "grad_norm": 5.068330764770508,
+      "learning_rate": 4.832535885167465e-05,
+      "loss": 2.3171,
+      "step": 39
+    },
+    {
+      "epoch": 0.09456264775413711,
+      "grad_norm": 4.925624847412109,
+      "learning_rate": 4.826555023923445e-05,
+      "loss": 2.3936,
+      "step": 40
+    },
+    {
+      "epoch": 0.09692671394799054,
+      "grad_norm": 4.296761989593506,
+      "learning_rate": 4.820574162679426e-05,
+      "loss": 2.3958,
+      "step": 41
+    },
+    {
+      "epoch": 0.09929078014184398,
+      "grad_norm": 4.61798095703125,
+      "learning_rate": 4.8145933014354064e-05,
+      "loss": 2.4068,
+      "step": 42
+    },
+    {
+      "epoch": 0.1016548463356974,
+      "grad_norm": 5.4643025398254395,
+      "learning_rate": 4.808612440191388e-05,
+      "loss": 2.3982,
+      "step": 43
+    },
+    {
+      "epoch": 0.10401891252955082,
+      "grad_norm": 4.580990314483643,
+      "learning_rate": 4.802631578947368e-05,
+      "loss": 2.346,
+      "step": 44
+    },
+    {
+      "epoch": 0.10638297872340426,
+      "grad_norm": 4.549551963806152,
+      "learning_rate": 4.796650717703349e-05,
+      "loss": 2.2992,
+      "step": 45
+    },
+    {
+      "epoch": 0.10874704491725769,
+      "grad_norm": 4.603874206542969,
+      "learning_rate": 4.790669856459331e-05,
+      "loss": 2.3436,
+      "step": 46
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 4.180109977722168,
+      "learning_rate": 4.784688995215311e-05,
+      "loss": 2.3465,
+      "step": 47
+    },
+    {
+      "epoch": 0.11347517730496454,
+      "grad_norm": 4.274153232574463,
+      "learning_rate": 4.778708133971292e-05,
+      "loss": 2.3173,
+      "step": 48
+    },
+    {
+      "epoch": 0.11583924349881797,
+      "grad_norm": 4.787023544311523,
+      "learning_rate": 4.772727272727273e-05,
+      "loss": 2.3225,
+      "step": 49
+    },
+    {
+      "epoch": 0.1182033096926714,
+      "grad_norm": 4.5308146476745605,
+      "learning_rate": 4.766746411483254e-05,
+      "loss": 2.2716,
+      "step": 50
+    },
+    {
+      "epoch": 0.12056737588652482,
+      "grad_norm": 4.548710823059082,
+      "learning_rate": 4.760765550239234e-05,
+      "loss": 2.277,
+      "step": 51
+    },
+    {
+      "epoch": 0.12293144208037825,
+      "grad_norm": 4.569216251373291,
+      "learning_rate": 4.754784688995216e-05,
+      "loss": 2.3076,
+      "step": 52
+    },
+    {
+      "epoch": 0.12529550827423167,
+      "grad_norm": 4.4847493171691895,
+      "learning_rate": 4.748803827751196e-05,
+      "loss": 2.3058,
+      "step": 53
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 4.511356353759766,
+      "learning_rate": 4.742822966507177e-05,
+      "loss": 2.1524,
+      "step": 54
+    },
+    {
+      "epoch": 0.13002364066193853,
+      "grad_norm": 4.245441436767578,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 2.2795,
+      "step": 55
+    },
+    {
+      "epoch": 0.13238770685579196,
+      "grad_norm": 4.903697490692139,
+      "learning_rate": 4.730861244019139e-05,
+      "loss": 2.2934,
+      "step": 56
+    },
+    {
+      "epoch": 0.1347517730496454,
+      "grad_norm": 4.915858745574951,
+      "learning_rate": 4.72488038277512e-05,
+      "loss": 2.271,
+      "step": 57
+    },
+    {
+      "epoch": 0.13711583924349882,
+      "grad_norm": 4.528074264526367,
+      "learning_rate": 4.7188995215311e-05,
+      "loss": 2.2712,
+      "step": 58
+    },
+    {
+      "epoch": 0.13947990543735225,
+      "grad_norm": 4.402848720550537,
+      "learning_rate": 4.712918660287082e-05,
+      "loss": 2.2441,
+      "step": 59
+    },
+    {
+      "epoch": 0.14184397163120568,
+      "grad_norm": 4.317126274108887,
+      "learning_rate": 4.706937799043062e-05,
+      "loss": 2.3081,
+      "step": 60
+    },
+    {
+      "epoch": 0.14420803782505912,
+      "grad_norm": 4.551846027374268,
+      "learning_rate": 4.700956937799043e-05,
+      "loss": 2.33,
+      "step": 61
+    },
+    {
+      "epoch": 0.14657210401891252,
+      "grad_norm": 4.294714450836182,
+      "learning_rate": 4.694976076555024e-05,
+      "loss": 2.2291,
+      "step": 62
+    },
+    {
+      "epoch": 0.14893617021276595,
+      "grad_norm": 4.264118194580078,
+      "learning_rate": 4.688995215311005e-05,
+      "loss": 2.187,
+      "step": 63
+    },
+    {
+      "epoch": 0.15130023640661938,
+      "grad_norm": 4.644858360290527,
+      "learning_rate": 4.683014354066986e-05,
+      "loss": 2.2767,
+      "step": 64
+    },
+    {
+      "epoch": 0.1536643026004728,
+      "grad_norm": 4.4240522384643555,
+      "learning_rate": 4.677033492822967e-05,
+      "loss": 2.2365,
+      "step": 65
+    },
+    {
+      "epoch": 0.15602836879432624,
+      "grad_norm": 4.994328022003174,
+      "learning_rate": 4.671052631578948e-05,
+      "loss": 2.1967,
+      "step": 66
+    },
+    {
+      "epoch": 0.15839243498817968,
+      "grad_norm": 4.091360092163086,
+      "learning_rate": 4.665071770334928e-05,
+      "loss": 2.1983,
+      "step": 67
+    },
+    {
+      "epoch": 0.1607565011820331,
+      "grad_norm": 4.4661760330200195,
+      "learning_rate": 4.659090909090909e-05,
+      "loss": 2.2006,
+      "step": 68
+    },
+    {
+      "epoch": 0.16312056737588654,
+      "grad_norm": 4.753389358520508,
+      "learning_rate": 4.65311004784689e-05,
+      "loss": 2.2816,
+      "step": 69
+    },
+    {
+      "epoch": 0.16548463356973994,
+      "grad_norm": 4.429995059967041,
+      "learning_rate": 4.647129186602871e-05,
+      "loss": 2.2732,
+      "step": 70
+    },
+    {
+      "epoch": 0.16784869976359337,
+      "grad_norm": 4.254610061645508,
+      "learning_rate": 4.641148325358852e-05,
+      "loss": 2.1097,
+      "step": 71
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 4.007920265197754,
+      "learning_rate": 4.635167464114833e-05,
+      "loss": 2.2057,
+      "step": 72
+    },
+    {
+      "epoch": 0.17257683215130024,
+      "grad_norm": 4.3318772315979,
+      "learning_rate": 4.629186602870814e-05,
+      "loss": 2.2074,
+      "step": 73
+    },
+    {
+      "epoch": 0.17494089834515367,
+      "grad_norm": 5.013289928436279,
+      "learning_rate": 4.623205741626794e-05,
+      "loss": 2.3091,
+      "step": 74
+    },
+    {
+      "epoch": 0.1773049645390071,
+      "grad_norm": 4.14231538772583,
+      "learning_rate": 4.617224880382776e-05,
+      "loss": 2.2149,
+      "step": 75
+    },
+    {
+      "epoch": 0.17966903073286053,
+      "grad_norm": 4.8862762451171875,
+      "learning_rate": 4.611244019138756e-05,
+      "loss": 2.2153,
+      "step": 76
+    },
+    {
+      "epoch": 0.18203309692671396,
+      "grad_norm": 4.4072489738464355,
+      "learning_rate": 4.605263157894737e-05,
+      "loss": 2.256,
+      "step": 77
+    },
+    {
+      "epoch": 0.18439716312056736,
+      "grad_norm": 4.19016170501709,
+      "learning_rate": 4.599282296650718e-05,
+      "loss": 2.1268,
+      "step": 78
+    },
+    {
+      "epoch": 0.1867612293144208,
+      "grad_norm": 4.425796031951904,
+      "learning_rate": 4.593301435406699e-05,
+      "loss": 2.2441,
+      "step": 79
+    },
+    {
+      "epoch": 0.18912529550827423,
+      "grad_norm": 4.253169536590576,
+      "learning_rate": 4.58732057416268e-05,
+      "loss": 2.2179,
+      "step": 80
+    },
+    {
+      "epoch": 0.19148936170212766,
+      "grad_norm": 4.226646900177002,
+      "learning_rate": 4.58133971291866e-05,
+      "loss": 2.2049,
+      "step": 81
+    },
+    {
+      "epoch": 0.1938534278959811,
+      "grad_norm": 4.134627342224121,
+      "learning_rate": 4.575358851674642e-05,
+      "loss": 2.1709,
+      "step": 82
+    },
+    {
+      "epoch": 0.19621749408983452,
+      "grad_norm": 3.9358890056610107,
+      "learning_rate": 4.569377990430622e-05,
+      "loss": 2.1614,
+      "step": 83
+    },
+    {
+      "epoch": 0.19858156028368795,
+      "grad_norm": 4.453876972198486,
+      "learning_rate": 4.563397129186603e-05,
+      "loss": 2.2435,
+      "step": 84
+    },
+    {
+      "epoch": 0.20094562647754138,
+      "grad_norm": 4.1704535484313965,
+      "learning_rate": 4.557416267942584e-05,
+      "loss": 2.1753,
+      "step": 85
+    },
+    {
+      "epoch": 0.2033096926713948,
+      "grad_norm": 4.117349624633789,
+      "learning_rate": 4.551435406698565e-05,
+      "loss": 2.196,
+      "step": 86
+    },
+    {
+      "epoch": 0.20567375886524822,
+      "grad_norm": 3.9573557376861572,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 2.1901,
+      "step": 87
+    },
+    {
+      "epoch": 0.20803782505910165,
+      "grad_norm": 4.639995098114014,
+      "learning_rate": 4.539473684210527e-05,
+      "loss": 2.2147,
+      "step": 88
+    },
+    {
+      "epoch": 0.21040189125295508,
+      "grad_norm": 4.055258750915527,
+      "learning_rate": 4.533492822966508e-05,
+      "loss": 2.1608,
+      "step": 89
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 4.474514484405518,
+      "learning_rate": 4.527511961722488e-05,
+      "loss": 2.1507,
+      "step": 90
+    },
+    {
+      "epoch": 0.21513002364066194,
+      "grad_norm": 4.639707088470459,
+      "learning_rate": 4.521531100478469e-05,
+      "loss": 2.2826,
+      "step": 91
+    },
+    {
+      "epoch": 0.21749408983451538,
+      "grad_norm": 4.863740921020508,
+      "learning_rate": 4.51555023923445e-05,
+      "loss": 2.1766,
+      "step": 92
+    },
+    {
+      "epoch": 0.2198581560283688,
+      "grad_norm": 4.412019729614258,
+      "learning_rate": 4.509569377990431e-05,
+      "loss": 2.1956,
+      "step": 93
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 4.765653610229492,
+      "learning_rate": 4.503588516746411e-05,
+      "loss": 2.2213,
+      "step": 94
+    },
+    {
+      "epoch": 0.22458628841607564,
+      "grad_norm": 4.181426525115967,
+      "learning_rate": 4.497607655502393e-05,
+      "loss": 2.1097,
+      "step": 95
+    },
+    {
+      "epoch": 0.22695035460992907,
+      "grad_norm": 4.018885612487793,
+      "learning_rate": 4.491626794258373e-05,
+      "loss": 2.1055,
+      "step": 96
+    },
+    {
+      "epoch": 0.2293144208037825,
+      "grad_norm": 4.294673919677734,
+      "learning_rate": 4.485645933014354e-05,
+      "loss": 2.1984,
+      "step": 97
+    },
+    {
+      "epoch": 0.23167848699763594,
+      "grad_norm": 4.040910720825195,
+      "learning_rate": 4.4796650717703357e-05,
+      "loss": 2.1209,
+      "step": 98
+    },
+    {
+      "epoch": 0.23404255319148937,
+      "grad_norm": 4.413674354553223,
+      "learning_rate": 4.473684210526316e-05,
+      "loss": 2.177,
+      "step": 99
+    },
+    {
+      "epoch": 0.2364066193853428,
+      "grad_norm": 4.586551189422607,
+      "learning_rate": 4.467703349282297e-05,
+      "loss": 2.2419,
+      "step": 100
+    },
+    {
+      "epoch": 0.23877068557919623,
+      "grad_norm": 3.753817319869995,
+      "learning_rate": 4.461722488038278e-05,
+      "loss": 2.1044,
+      "step": 101
+    },
+    {
+      "epoch": 0.24113475177304963,
+      "grad_norm": 4.756337642669678,
+      "learning_rate": 4.455741626794259e-05,
+      "loss": 2.2236,
+      "step": 102
+    },
+    {
+      "epoch": 0.24349881796690306,
+      "grad_norm": 4.174933433532715,
+      "learning_rate": 4.449760765550239e-05,
+      "loss": 2.1759,
+      "step": 103
+    },
+    {
+      "epoch": 0.2458628841607565,
+      "grad_norm": 4.282530784606934,
+      "learning_rate": 4.44377990430622e-05,
+      "loss": 2.16,
+      "step": 104
+    },
+    {
+      "epoch": 0.24822695035460993,
+      "grad_norm": 4.547479629516602,
+      "learning_rate": 4.437799043062201e-05,
+      "loss": 2.0593,
+      "step": 105
+    },
+    {
+      "epoch": 0.25059101654846333,
+      "grad_norm": 4.146774768829346,
+      "learning_rate": 4.431818181818182e-05,
+      "loss": 2.1806,
+      "step": 106
+    },
+    {
+      "epoch": 0.25295508274231676,
+      "grad_norm": 4.418383598327637,
+      "learning_rate": 4.425837320574163e-05,
+      "loss": 2.1936,
+      "step": 107
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 4.2527947425842285,
+      "learning_rate": 4.419856459330144e-05,
+      "loss": 2.0821,
+      "step": 108
+    },
+    {
+      "epoch": 0.2576832151300236,
+      "grad_norm": 4.108152389526367,
+      "learning_rate": 4.413875598086125e-05,
+      "loss": 2.1622,
+      "step": 109
+    },
+    {
+      "epoch": 0.26004728132387706,
+      "grad_norm": 4.083064556121826,
+      "learning_rate": 4.407894736842105e-05,
+      "loss": 2.1178,
+      "step": 110
+    },
+    {
+      "epoch": 0.2624113475177305,
+      "grad_norm": 4.213859558105469,
+      "learning_rate": 4.401913875598087e-05,
+      "loss": 2.1575,
+      "step": 111
+    },
+    {
+      "epoch": 0.2647754137115839,
+      "grad_norm": 5.222052574157715,
+      "learning_rate": 4.395933014354067e-05,
+      "loss": 2.0866,
+      "step": 112
+    },
+    {
+      "epoch": 0.26713947990543735,
+      "grad_norm": 4.5867743492126465,
+      "learning_rate": 4.389952153110048e-05,
+      "loss": 2.156,
+      "step": 113
+    },
+    {
+      "epoch": 0.2695035460992908,
+      "grad_norm": 4.399920463562012,
+      "learning_rate": 4.383971291866029e-05,
+      "loss": 2.1911,
+      "step": 114
+    },
+    {
+      "epoch": 0.2718676122931442,
+      "grad_norm": 3.858452081680298,
+      "learning_rate": 4.37799043062201e-05,
+      "loss": 2.152,
+      "step": 115
+    },
+    {
+      "epoch": 0.27423167848699764,
+      "grad_norm": 4.103243350982666,
+      "learning_rate": 4.372009569377991e-05,
+      "loss": 2.1825,
+      "step": 116
+    },
+    {
+      "epoch": 0.2765957446808511,
+      "grad_norm": 4.9622344970703125,
+      "learning_rate": 4.366028708133971e-05,
+      "loss": 2.1182,
+      "step": 117
+    },
+    {
+      "epoch": 0.2789598108747045,
+      "grad_norm": 4.199765205383301,
+      "learning_rate": 4.360047846889953e-05,
+      "loss": 2.0638,
+      "step": 118
+    },
+    {
+      "epoch": 0.28132387706855794,
+      "grad_norm": 4.943136215209961,
+      "learning_rate": 4.354066985645933e-05,
+      "loss": 2.1906,
+      "step": 119
+    },
+    {
+      "epoch": 0.28368794326241137,
+      "grad_norm": 4.076112270355225,
+      "learning_rate": 4.348086124401914e-05,
+      "loss": 2.1208,
+      "step": 120
+    },
+    {
+      "epoch": 0.2860520094562648,
+      "grad_norm": 4.323102951049805,
+      "learning_rate": 4.342105263157895e-05,
+      "loss": 2.1642,
+      "step": 121
+    },
+    {
+      "epoch": 0.28841607565011823,
+      "grad_norm": 3.8718385696411133,
+      "learning_rate": 4.336124401913876e-05,
+      "loss": 2.0946,
+      "step": 122
+    },
+    {
+      "epoch": 0.2907801418439716,
+      "grad_norm": 4.6664862632751465,
+      "learning_rate": 4.330143540669857e-05,
+      "loss": 2.1411,
+      "step": 123
+    },
+    {
+      "epoch": 0.29314420803782504,
+      "grad_norm": 4.570885181427002,
+      "learning_rate": 4.324162679425838e-05,
+      "loss": 2.1268,
+      "step": 124
+    },
+    {
+      "epoch": 0.29550827423167847,
+      "grad_norm": 3.7459053993225098,
+      "learning_rate": 4.318181818181819e-05,
+      "loss": 2.042,
+      "step": 125
+    },
+    {
+      "epoch": 0.2978723404255319,
+      "grad_norm": 4.332980155944824,
+      "learning_rate": 4.312200956937799e-05,
+      "loss": 2.1462,
+      "step": 126
+    },
+    {
+      "epoch": 0.30023640661938533,
+      "grad_norm": 4.2524919509887695,
+      "learning_rate": 4.3062200956937806e-05,
+      "loss": 2.136,
+      "step": 127
+    },
+    {
+      "epoch": 0.30260047281323876,
+      "grad_norm": 4.230786323547363,
+      "learning_rate": 4.300239234449761e-05,
+      "loss": 2.1397,
+      "step": 128
+    },
+    {
+      "epoch": 0.3049645390070922,
+      "grad_norm": 4.281345844268799,
+      "learning_rate": 4.294258373205742e-05,
+      "loss": 2.1,
+      "step": 129
+    },
+    {
+      "epoch": 0.3073286052009456,
+      "grad_norm": 4.491425514221191,
+      "learning_rate": 4.288277511961723e-05,
+      "loss": 2.1304,
+      "step": 130
+    },
+    {
+      "epoch": 0.30969267139479906,
+      "grad_norm": 4.068315029144287,
+      "learning_rate": 4.282296650717704e-05,
+      "loss": 2.1201,
+      "step": 131
+    },
+    {
+      "epoch": 0.3120567375886525,
+      "grad_norm": 4.424421787261963,
+      "learning_rate": 4.2763157894736847e-05,
+      "loss": 2.0387,
+      "step": 132
+    },
+    {
+      "epoch": 0.3144208037825059,
+      "grad_norm": 4.6968865394592285,
+      "learning_rate": 4.270334928229665e-05,
+      "loss": 2.0306,
+      "step": 133
+    },
+    {
+      "epoch": 0.31678486997635935,
+      "grad_norm": 3.468080520629883,
+      "learning_rate": 4.2643540669856466e-05,
+      "loss": 2.0499,
+      "step": 134
+    },
+    {
+      "epoch": 0.3191489361702128,
+      "grad_norm": 4.1545186042785645,
+      "learning_rate": 4.258373205741627e-05,
+      "loss": 2.0953,
+      "step": 135
+    },
+    {
+      "epoch": 0.3215130023640662,
+      "grad_norm": 4.665497779846191,
+      "learning_rate": 4.252392344497608e-05,
+      "loss": 2.0821,
+      "step": 136
+    },
+    {
+      "epoch": 0.32387706855791965,
+      "grad_norm": 4.307028770446777,
+      "learning_rate": 4.246411483253589e-05,
+      "loss": 2.0566,
+      "step": 137
+    },
+    {
+      "epoch": 0.3262411347517731,
+      "grad_norm": 4.265110969543457,
+      "learning_rate": 4.24043062200957e-05,
+      "loss": 2.1148,
+      "step": 138
+    },
+    {
+      "epoch": 0.32860520094562645,
+      "grad_norm": 4.940046787261963,
+      "learning_rate": 4.2344497607655506e-05,
+      "loss": 2.0617,
+      "step": 139
+    },
+    {
+      "epoch": 0.3309692671394799,
+      "grad_norm": 4.576653480529785,
+      "learning_rate": 4.2284688995215316e-05,
+      "loss": 2.1433,
+      "step": 140
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 4.502660274505615,
+      "learning_rate": 4.2224880382775126e-05,
+      "loss": 2.0935,
+      "step": 141
+    },
+    {
+      "epoch": 0.33569739952718675,
+      "grad_norm": 5.039083480834961,
+      "learning_rate": 4.216507177033493e-05,
+      "loss": 2.1362,
+      "step": 142
+    },
+    {
+      "epoch": 0.3380614657210402,
+      "grad_norm": 4.402792930603027,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 2.0993,
+      "step": 143
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 4.509486675262451,
+      "learning_rate": 4.204545454545455e-05,
+      "loss": 2.0299,
+      "step": 144
+    },
+    {
+      "epoch": 0.34278959810874704,
+      "grad_norm": 4.011101722717285,
+      "learning_rate": 4.198564593301436e-05,
+      "loss": 2.0929,
+      "step": 145
+    },
+    {
+      "epoch": 0.34515366430260047,
+      "grad_norm": 5.007168292999268,
+      "learning_rate": 4.192583732057416e-05,
+      "loss": 2.0842,
+      "step": 146
+    },
+    {
+      "epoch": 0.3475177304964539,
+      "grad_norm": 4.631533145904541,
+      "learning_rate": 4.1866028708133976e-05,
+      "loss": 2.0474,
+      "step": 147
+    },
+    {
+      "epoch": 0.34988179669030733,
+      "grad_norm": 4.278231620788574,
+      "learning_rate": 4.1806220095693785e-05,
+      "loss": 2.0819,
+      "step": 148
+    },
+    {
+      "epoch": 0.35224586288416077,
+      "grad_norm": 4.484694480895996,
+      "learning_rate": 4.174641148325359e-05,
+      "loss": 2.0502,
+      "step": 149
+    },
+    {
+      "epoch": 0.3546099290780142,
+      "grad_norm": 4.756961345672607,
+      "learning_rate": 4.1686602870813404e-05,
+      "loss": 2.0529,
+      "step": 150
+    },
+    {
+      "epoch": 0.35697399527186763,
+      "grad_norm": 4.347701549530029,
+      "learning_rate": 4.162679425837321e-05,
+      "loss": 2.0666,
+      "step": 151
+    },
+    {
+      "epoch": 0.35933806146572106,
+      "grad_norm": 3.76229190826416,
+      "learning_rate": 4.156698564593302e-05,
+      "loss": 2.0763,
+      "step": 152
+    },
+    {
+      "epoch": 0.3617021276595745,
+      "grad_norm": 3.8186378479003906,
+      "learning_rate": 4.150717703349282e-05,
+      "loss": 2.0392,
+      "step": 153
+    },
+    {
+      "epoch": 0.3640661938534279,
+      "grad_norm": 4.353139877319336,
+      "learning_rate": 4.1447368421052636e-05,
+      "loss": 1.9854,
+      "step": 154
+    },
+    {
+      "epoch": 0.3664302600472813,
+      "grad_norm": 5.3801422119140625,
+      "learning_rate": 4.138755980861244e-05,
+      "loss": 1.9959,
+      "step": 155
+    },
+    {
+      "epoch": 0.36879432624113473,
+      "grad_norm": 4.44751501083374,
+      "learning_rate": 4.132775119617225e-05,
+      "loss": 2.0911,
+      "step": 156
+    },
+    {
+      "epoch": 0.37115839243498816,
+      "grad_norm": 4.025294303894043,
+      "learning_rate": 4.1267942583732064e-05,
+      "loss": 2.0192,
+      "step": 157
+    },
+    {
+      "epoch": 0.3735224586288416,
+      "grad_norm": 4.605421543121338,
+      "learning_rate": 4.120813397129187e-05,
+      "loss": 2.1464,
+      "step": 158
+    },
+    {
+      "epoch": 0.375886524822695,
+      "grad_norm": 5.045236110687256,
+      "learning_rate": 4.114832535885168e-05,
+      "loss": 2.0756,
+      "step": 159
+    },
+    {
+      "epoch": 0.37825059101654845,
+      "grad_norm": 4.459848403930664,
+      "learning_rate": 4.1088516746411486e-05,
+      "loss": 2.1544,
+      "step": 160
+    },
+    {
+      "epoch": 0.3806146572104019,
+      "grad_norm": 4.420864105224609,
+      "learning_rate": 4.1028708133971296e-05,
+      "loss": 2.0296,
+      "step": 161
+    },
+    {
+      "epoch": 0.3829787234042553,
+      "grad_norm": 3.998237133026123,
+      "learning_rate": 4.09688995215311e-05,
+      "loss": 2.1039,
+      "step": 162
+    },
+    {
+      "epoch": 0.38534278959810875,
+      "grad_norm": 4.448668003082275,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 2.0762,
+      "step": 163
+    },
+    {
+      "epoch": 0.3877068557919622,
+      "grad_norm": 4.266834259033203,
+      "learning_rate": 4.084928229665072e-05,
+      "loss": 2.0557,
+      "step": 164
+    },
+    {
+      "epoch": 0.3900709219858156,
+      "grad_norm": 4.796644687652588,
+      "learning_rate": 4.078947368421053e-05,
+      "loss": 2.1027,
+      "step": 165
+    },
+    {
+      "epoch": 0.39243498817966904,
+      "grad_norm": 4.560353755950928,
+      "learning_rate": 4.0729665071770337e-05,
+      "loss": 2.0245,
+      "step": 166
+    },
+    {
+      "epoch": 0.3947990543735225,
+      "grad_norm": 3.673262119293213,
+      "learning_rate": 4.0669856459330146e-05,
+      "loss": 2.033,
+      "step": 167
+    },
+    {
+      "epoch": 0.3971631205673759,
+      "grad_norm": 4.134026050567627,
+      "learning_rate": 4.0610047846889956e-05,
+      "loss": 2.0079,
+      "step": 168
+    },
+    {
+      "epoch": 0.39952718676122934,
+      "grad_norm": 4.677055835723877,
+      "learning_rate": 4.055023923444976e-05,
+      "loss": 2.0874,
+      "step": 169
+    },
+    {
+      "epoch": 0.40189125295508277,
+      "grad_norm": 4.731474876403809,
+      "learning_rate": 4.0490430622009575e-05,
+      "loss": 1.9878,
+      "step": 170
+    },
+    {
+      "epoch": 0.40425531914893614,
+      "grad_norm": 4.689650058746338,
+      "learning_rate": 4.043062200956938e-05,
+      "loss": 2.0933,
+      "step": 171
+    },
+    {
+      "epoch": 0.4066193853427896,
+      "grad_norm": 4.23575496673584,
+      "learning_rate": 4.037081339712919e-05,
+      "loss": 2.0436,
+      "step": 172
+    },
+    {
+      "epoch": 0.408983451536643,
+      "grad_norm": 3.8167171478271484,
+      "learning_rate": 4.0311004784688996e-05,
+      "loss": 1.9843,
+      "step": 173
+    },
+    {
+      "epoch": 0.41134751773049644,
+      "grad_norm": 4.35087251663208,
+      "learning_rate": 4.0251196172248806e-05,
+      "loss": 2.0386,
+      "step": 174
+    },
+    {
+      "epoch": 0.41371158392434987,
+      "grad_norm": 4.764196395874023,
+      "learning_rate": 4.0191387559808616e-05,
+      "loss": 2.0783,
+      "step": 175
+    },
+    {
+      "epoch": 0.4160756501182033,
+      "grad_norm": 4.374221324920654,
+      "learning_rate": 4.0131578947368425e-05,
+      "loss": 1.9711,
+      "step": 176
+    },
+    {
+      "epoch": 0.41843971631205673,
+      "grad_norm": 4.542954444885254,
+      "learning_rate": 4.0071770334928235e-05,
+      "loss": 2.0036,
+      "step": 177
+    },
+    {
+      "epoch": 0.42080378250591016,
+      "grad_norm": 3.849266767501831,
+      "learning_rate": 4.001196172248804e-05,
+      "loss": 2.0571,
+      "step": 178
+    },
+    {
+      "epoch": 0.4231678486997636,
+      "grad_norm": 4.014004230499268,
+      "learning_rate": 3.995215311004785e-05,
+      "loss": 2.0642,
+      "step": 179
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 4.369685649871826,
+      "learning_rate": 3.9892344497607656e-05,
+      "loss": 2.0502,
+      "step": 180
+    },
+    {
+      "epoch": 0.42789598108747046,
+      "grad_norm": 4.280391216278076,
+      "learning_rate": 3.9832535885167466e-05,
+      "loss": 2.0025,
+      "step": 181
+    },
+    {
+      "epoch": 0.4302600472813239,
+      "grad_norm": 4.444298267364502,
+      "learning_rate": 3.9772727272727275e-05,
+      "loss": 1.9828,
+      "step": 182
+    },
+    {
+      "epoch": 0.4326241134751773,
+      "grad_norm": 3.794656991958618,
+      "learning_rate": 3.9712918660287085e-05,
+      "loss": 2.048,
+      "step": 183
+    },
+    {
+      "epoch": 0.43498817966903075,
+      "grad_norm": 4.498697280883789,
+      "learning_rate": 3.9653110047846894e-05,
+      "loss": 2.1017,
+      "step": 184
+    },
+    {
+      "epoch": 0.4373522458628842,
+      "grad_norm": 4.182841777801514,
+      "learning_rate": 3.95933014354067e-05,
+      "loss": 2.0588,
+      "step": 185
+    },
+    {
+      "epoch": 0.4397163120567376,
+      "grad_norm": 4.118316173553467,
+      "learning_rate": 3.9533492822966514e-05,
+      "loss": 2.0654,
+      "step": 186
+    },
+    {
+      "epoch": 0.44208037825059104,
+      "grad_norm": 5.026358127593994,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 2.0808,
+      "step": 187
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 4.171792984008789,
+      "learning_rate": 3.9413875598086126e-05,
+      "loss": 1.9832,
+      "step": 188
+    },
+    {
+      "epoch": 0.44680851063829785,
+      "grad_norm": 4.625768661499023,
+      "learning_rate": 3.9354066985645935e-05,
+      "loss": 2.0244,
+      "step": 189
+    },
+    {
+      "epoch": 0.4491725768321513,
+      "grad_norm": 4.119114398956299,
+      "learning_rate": 3.9294258373205745e-05,
+      "loss": 2.011,
+      "step": 190
+    },
+    {
+      "epoch": 0.4515366430260047,
+      "grad_norm": 4.071448802947998,
+      "learning_rate": 3.9234449760765554e-05,
+      "loss": 2.0469,
+      "step": 191
+    },
+    {
+      "epoch": 0.45390070921985815,
+      "grad_norm": 4.00705099105835,
+      "learning_rate": 3.917464114832536e-05,
+      "loss": 1.9889,
+      "step": 192
+    },
+    {
+      "epoch": 0.4562647754137116,
+      "grad_norm": 4.969683647155762,
+      "learning_rate": 3.9114832535885173e-05,
+      "loss": 2.0097,
+      "step": 193
+    },
+    {
+      "epoch": 0.458628841607565,
+      "grad_norm": 4.4712605476379395,
+      "learning_rate": 3.9055023923444976e-05,
+      "loss": 2.0413,
+      "step": 194
+    },
+    {
+      "epoch": 0.46099290780141844,
+      "grad_norm": 4.194468021392822,
+      "learning_rate": 3.8995215311004786e-05,
+      "loss": 2.0623,
+      "step": 195
+    },
+    {
+      "epoch": 0.46335697399527187,
+      "grad_norm": 3.9785311222076416,
+      "learning_rate": 3.8935406698564595e-05,
+      "loss": 2.0606,
+      "step": 196
+    },
+    {
+      "epoch": 0.4657210401891253,
+      "grad_norm": 4.314589977264404,
+      "learning_rate": 3.8875598086124405e-05,
+      "loss": 2.0411,
+      "step": 197
+    },
+    {
+      "epoch": 0.46808510638297873,
+      "grad_norm": 3.732752561569214,
+      "learning_rate": 3.8815789473684214e-05,
+      "loss": 1.8543,
+      "step": 198
+    },
+    {
+      "epoch": 0.47044917257683216,
+      "grad_norm": 4.228579998016357,
+      "learning_rate": 3.8755980861244024e-05,
+      "loss": 2.0262,
+      "step": 199
+    },
+    {
+      "epoch": 0.4728132387706856,
+      "grad_norm": 3.9367430210113525,
+      "learning_rate": 3.869617224880383e-05,
+      "loss": 2.0398,
+      "step": 200
+    },
+    {
+      "epoch": 0.475177304964539,
+      "grad_norm": 3.8661906719207764,
+      "learning_rate": 3.8636363636363636e-05,
+      "loss": 1.9708,
+      "step": 201
+    },
+    {
+      "epoch": 0.47754137115839246,
+      "grad_norm": 3.8464865684509277,
+      "learning_rate": 3.8576555023923446e-05,
+      "loss": 2.0388,
+      "step": 202
+    },
+    {
+      "epoch": 0.4799054373522459,
+      "grad_norm": 3.867969274520874,
+      "learning_rate": 3.8516746411483255e-05,
+      "loss": 2.0592,
+      "step": 203
+    },
+    {
+      "epoch": 0.48226950354609927,
+      "grad_norm": 4.027868270874023,
+      "learning_rate": 3.8456937799043065e-05,
+      "loss": 2.0641,
+      "step": 204
+    },
+    {
+      "epoch": 0.4846335697399527,
+      "grad_norm": 4.061384677886963,
+      "learning_rate": 3.839712918660287e-05,
+      "loss": 2.0542,
+      "step": 205
+    },
+    {
+      "epoch": 0.48699763593380613,
+      "grad_norm": 3.954545021057129,
+      "learning_rate": 3.8337320574162684e-05,
+      "loss": 1.9951,
+      "step": 206
+    },
+    {
+      "epoch": 0.48936170212765956,
+      "grad_norm": 3.8164191246032715,
+      "learning_rate": 3.8277511961722486e-05,
+      "loss": 2.0094,
+      "step": 207
+    },
+    {
+      "epoch": 0.491725768321513,
+      "grad_norm": 4.0749053955078125,
+      "learning_rate": 3.8217703349282296e-05,
+      "loss": 1.9861,
+      "step": 208
+    },
+    {
+      "epoch": 0.4940898345153664,
+      "grad_norm": 3.8733890056610107,
+      "learning_rate": 3.815789473684211e-05,
+      "loss": 1.9821,
+      "step": 209
+    },
+    {
+      "epoch": 0.49645390070921985,
+      "grad_norm": 3.9854531288146973,
+      "learning_rate": 3.8098086124401915e-05,
+      "loss": 1.9549,
+      "step": 210
+    },
+    {
+      "epoch": 0.4988179669030733,
+      "grad_norm": 4.046158313751221,
+      "learning_rate": 3.8038277511961725e-05,
+      "loss": 2.02,
+      "step": 211
+    },
+    {
+      "epoch": 0.5011820330969267,
+      "grad_norm": 4.107903003692627,
+      "learning_rate": 3.7978468899521534e-05,
+      "loss": 1.9814,
+      "step": 212
+    },
+    {
+      "epoch": 0.5035460992907801,
+      "grad_norm": 4.039123058319092,
+      "learning_rate": 3.7918660287081344e-05,
+      "loss": 1.9969,
+      "step": 213
+    },
+    {
+      "epoch": 0.5059101654846335,
+      "grad_norm": 4.200449466705322,
+      "learning_rate": 3.7858851674641146e-05,
+      "loss": 1.9756,
+      "step": 214
+    },
+    {
+      "epoch": 0.508274231678487,
+      "grad_norm": 4.346997261047363,
+      "learning_rate": 3.7799043062200956e-05,
+      "loss": 2.0102,
+      "step": 215
+    },
+    {
+      "epoch": 0.5106382978723404,
+      "grad_norm": 4.00998067855835,
+      "learning_rate": 3.7739234449760765e-05,
+      "loss": 2.0895,
+      "step": 216
+    },
+    {
+      "epoch": 0.5130023640661938,
+      "grad_norm": 4.083645820617676,
+      "learning_rate": 3.7679425837320575e-05,
+      "loss": 1.9927,
+      "step": 217
+    },
+    {
+      "epoch": 0.5153664302600472,
+      "grad_norm": 4.116786956787109,
+      "learning_rate": 3.7619617224880384e-05,
+      "loss": 1.9798,
+      "step": 218
+    },
+    {
+      "epoch": 0.5177304964539007,
+      "grad_norm": 3.8103268146514893,
+      "learning_rate": 3.7559808612440194e-05,
+      "loss": 1.9616,
+      "step": 219
+    },
+    {
+      "epoch": 0.5200945626477541,
+      "grad_norm": 3.789548635482788,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.995,
+      "step": 220
+    },
+    {
+      "epoch": 0.5224586288416075,
+      "grad_norm": 4.23507022857666,
+      "learning_rate": 3.7440191387559806e-05,
+      "loss": 2.0384,
+      "step": 221
+    },
+    {
+      "epoch": 0.524822695035461,
+      "grad_norm": 4.319752216339111,
+      "learning_rate": 3.738038277511962e-05,
+      "loss": 2.0136,
+      "step": 222
+    },
+    {
+      "epoch": 0.5271867612293144,
+      "grad_norm": 4.502188682556152,
+      "learning_rate": 3.7320574162679425e-05,
+      "loss": 1.993,
+      "step": 223
+    },
+    {
+      "epoch": 0.5295508274231678,
+      "grad_norm": 3.8698880672454834,
+      "learning_rate": 3.7260765550239235e-05,
+      "loss": 1.9742,
+      "step": 224
+    },
+    {
+      "epoch": 0.5319148936170213,
+      "grad_norm": 3.9415745735168457,
+      "learning_rate": 3.7200956937799044e-05,
+      "loss": 1.9259,
+      "step": 225
+    },
+    {
+      "epoch": 0.5342789598108747,
+      "grad_norm": 4.509599208831787,
+      "learning_rate": 3.7141148325358854e-05,
+      "loss": 1.9809,
+      "step": 226
+    },
+    {
+      "epoch": 0.5366430260047281,
+      "grad_norm": 4.538198471069336,
+      "learning_rate": 3.7081339712918663e-05,
+      "loss": 2.0455,
+      "step": 227
+    },
+    {
+      "epoch": 0.5390070921985816,
+      "grad_norm": 4.268890857696533,
+      "learning_rate": 3.7021531100478466e-05,
+      "loss": 2.0385,
+      "step": 228
+    },
+    {
+      "epoch": 0.541371158392435,
+      "grad_norm": 4.729723930358887,
+      "learning_rate": 3.696172248803828e-05,
+      "loss": 2.0132,
+      "step": 229
+    },
+    {
+      "epoch": 0.5437352245862884,
+      "grad_norm": 3.6644513607025146,
+      "learning_rate": 3.6901913875598085e-05,
+      "loss": 1.9425,
+      "step": 230
+    },
+    {
+      "epoch": 0.5460992907801419,
+      "grad_norm": 3.711594343185425,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 2.0302,
+      "step": 231
+    },
+    {
+      "epoch": 0.5484633569739953,
+      "grad_norm": 4.120182991027832,
+      "learning_rate": 3.6782296650717704e-05,
+      "loss": 1.9863,
+      "step": 232
+    },
+    {
+      "epoch": 0.5508274231678487,
+      "grad_norm": 4.521392822265625,
+      "learning_rate": 3.6722488038277514e-05,
+      "loss": 1.9855,
+      "step": 233
+    },
+    {
+      "epoch": 0.5531914893617021,
+      "grad_norm": 4.531548023223877,
+      "learning_rate": 3.666267942583732e-05,
+      "loss": 1.9397,
+      "step": 234
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 3.672027111053467,
+      "learning_rate": 3.660287081339713e-05,
+      "loss": 2.0197,
+      "step": 235
+    },
+    {
+      "epoch": 0.557919621749409,
+      "grad_norm": 3.798299789428711,
+      "learning_rate": 3.654306220095694e-05,
+      "loss": 1.9749,
+      "step": 236
+    },
+    {
+      "epoch": 0.5602836879432624,
+      "grad_norm": 4.396960258483887,
+      "learning_rate": 3.6483253588516745e-05,
+      "loss": 1.9643,
+      "step": 237
+    },
+    {
+      "epoch": 0.5626477541371159,
+      "grad_norm": 3.955106019973755,
+      "learning_rate": 3.642344497607656e-05,
+      "loss": 1.9963,
+      "step": 238
+    },
+    {
+      "epoch": 0.5650118203309693,
+      "grad_norm": 4.263886451721191,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 1.9553,
+      "step": 239
+    },
+    {
+      "epoch": 0.5673758865248227,
+      "grad_norm": 4.148135662078857,
+      "learning_rate": 3.6303827751196174e-05,
+      "loss": 1.9899,
+      "step": 240
+    },
+    {
+      "epoch": 0.5697399527186762,
+      "grad_norm": 4.090165138244629,
+      "learning_rate": 3.624401913875598e-05,
+      "loss": 1.9922,
+      "step": 241
+    },
+    {
+      "epoch": 0.5721040189125296,
+      "grad_norm": 3.7597265243530273,
+      "learning_rate": 3.618421052631579e-05,
+      "loss": 1.9846,
+      "step": 242
+    },
+    {
+      "epoch": 0.574468085106383,
+      "grad_norm": 4.532017707824707,
+      "learning_rate": 3.61244019138756e-05,
+      "loss": 1.9955,
+      "step": 243
+    },
+    {
+      "epoch": 0.5768321513002365,
+      "grad_norm": 4.268376350402832,
+      "learning_rate": 3.6064593301435405e-05,
+      "loss": 1.928,
+      "step": 244
+    },
+    {
+      "epoch": 0.5791962174940898,
+      "grad_norm": 4.414393901824951,
+      "learning_rate": 3.600478468899522e-05,
+      "loss": 1.9792,
+      "step": 245
+    },
+    {
+      "epoch": 0.5815602836879432,
+      "grad_norm": 4.180037498474121,
+      "learning_rate": 3.5944976076555024e-05,
+      "loss": 1.9661,
+      "step": 246
+    },
+    {
+      "epoch": 0.5839243498817966,
+      "grad_norm": 4.057360649108887,
+      "learning_rate": 3.5885167464114834e-05,
+      "loss": 1.9004,
+      "step": 247
+    },
+    {
+      "epoch": 0.5862884160756501,
+      "grad_norm": 3.8866353034973145,
+      "learning_rate": 3.582535885167464e-05,
+      "loss": 1.9582,
+      "step": 248
+    },
+    {
+      "epoch": 0.5886524822695035,
+      "grad_norm": 4.187384605407715,
+      "learning_rate": 3.576555023923445e-05,
+      "loss": 1.992,
+      "step": 249
+    },
+    {
+      "epoch": 0.5910165484633569,
+      "grad_norm": 3.99149751663208,
+      "learning_rate": 3.570574162679426e-05,
+      "loss": 1.9069,
+      "step": 250
+    },
+    {
+      "epoch": 0.5933806146572104,
+      "grad_norm": 3.89758563041687,
+      "learning_rate": 3.5645933014354065e-05,
+      "loss": 2.0696,
+      "step": 251
+    },
+    {
+      "epoch": 0.5957446808510638,
+      "grad_norm": 4.572390556335449,
+      "learning_rate": 3.558612440191388e-05,
+      "loss": 2.0519,
+      "step": 252
+    },
+    {
+      "epoch": 0.5981087470449172,
+      "grad_norm": 3.7999329566955566,
+      "learning_rate": 3.5526315789473684e-05,
+      "loss": 1.9143,
+      "step": 253
+    },
+    {
+      "epoch": 0.6004728132387707,
+      "grad_norm": 3.8050220012664795,
+      "learning_rate": 3.5466507177033493e-05,
+      "loss": 1.919,
+      "step": 254
+    },
+    {
+      "epoch": 0.6028368794326241,
+      "grad_norm": 4.685467720031738,
+      "learning_rate": 3.54066985645933e-05,
+      "loss": 2.0639,
+      "step": 255
+    },
+    {
+      "epoch": 0.6052009456264775,
+      "grad_norm": 4.132735252380371,
+      "learning_rate": 3.534688995215311e-05,
+      "loss": 2.0094,
+      "step": 256
+    },
+    {
+      "epoch": 0.607565011820331,
+      "grad_norm": 3.779338836669922,
+      "learning_rate": 3.5287081339712915e-05,
+      "loss": 1.9425,
+      "step": 257
+    },
+    {
+      "epoch": 0.6099290780141844,
+      "grad_norm": 3.988375186920166,
+      "learning_rate": 3.522727272727273e-05,
+      "loss": 1.9529,
+      "step": 258
+    },
+    {
+      "epoch": 0.6122931442080378,
+      "grad_norm": 3.8638248443603516,
+      "learning_rate": 3.516746411483254e-05,
+      "loss": 1.9326,
+      "step": 259
+    },
+    {
+      "epoch": 0.6146572104018913,
+      "grad_norm": 3.718116283416748,
+      "learning_rate": 3.5107655502392344e-05,
+      "loss": 1.9376,
+      "step": 260
+    },
+    {
+      "epoch": 0.6170212765957447,
+      "grad_norm": 3.8833279609680176,
+      "learning_rate": 3.504784688995216e-05,
+      "loss": 2.0297,
+      "step": 261
+    },
+    {
+      "epoch": 0.6193853427895981,
+      "grad_norm": 3.518829345703125,
+      "learning_rate": 3.498803827751196e-05,
+      "loss": 1.991,
+      "step": 262
+    },
+    {
+      "epoch": 0.6217494089834515,
+      "grad_norm": 3.8007776737213135,
+      "learning_rate": 3.492822966507177e-05,
+      "loss": 1.9722,
+      "step": 263
+    },
+    {
+      "epoch": 0.624113475177305,
+      "grad_norm": 3.51373028755188,
+      "learning_rate": 3.4868421052631575e-05,
+      "loss": 1.8961,
+      "step": 264
+    },
+    {
+      "epoch": 0.6264775413711584,
+      "grad_norm": 3.6159632205963135,
+      "learning_rate": 3.480861244019139e-05,
+      "loss": 1.9719,
+      "step": 265
+    },
+    {
+      "epoch": 0.6288416075650118,
+      "grad_norm": 3.923194408416748,
+      "learning_rate": 3.4748803827751194e-05,
+      "loss": 1.9133,
+      "step": 266
+    },
+    {
+      "epoch": 0.6312056737588653,
+      "grad_norm": 3.849912166595459,
+      "learning_rate": 3.4688995215311004e-05,
+      "loss": 1.9508,
+      "step": 267
+    },
+    {
+      "epoch": 0.6335697399527187,
+      "grad_norm": 3.8065907955169678,
+      "learning_rate": 3.462918660287082e-05,
+      "loss": 1.9426,
+      "step": 268
+    },
+    {
+      "epoch": 0.6359338061465721,
+      "grad_norm": 4.141129493713379,
+      "learning_rate": 3.456937799043062e-05,
+      "loss": 2.0031,
+      "step": 269
+    },
+    {
+      "epoch": 0.6382978723404256,
+      "grad_norm": 4.0821123123168945,
+      "learning_rate": 3.450956937799043e-05,
+      "loss": 1.9503,
+      "step": 270
+    },
+    {
+      "epoch": 0.640661938534279,
+      "grad_norm": 3.863445997238159,
+      "learning_rate": 3.444976076555024e-05,
+      "loss": 1.918,
+      "step": 271
+    },
+    {
+      "epoch": 0.6430260047281324,
+      "grad_norm": 3.9590418338775635,
+      "learning_rate": 3.438995215311005e-05,
+      "loss": 1.9768,
+      "step": 272
+    },
+    {
+      "epoch": 0.6453900709219859,
+      "grad_norm": 3.7065553665161133,
+      "learning_rate": 3.4330143540669854e-05,
+      "loss": 1.95,
+      "step": 273
+    },
+    {
+      "epoch": 0.6477541371158393,
+      "grad_norm": 3.657320737838745,
+      "learning_rate": 3.427033492822967e-05,
+      "loss": 1.9733,
+      "step": 274
+    },
+    {
+      "epoch": 0.6501182033096927,
+      "grad_norm": 4.384260654449463,
+      "learning_rate": 3.421052631578947e-05,
+      "loss": 2.0237,
+      "step": 275
+    },
+    {
+      "epoch": 0.6524822695035462,
+      "grad_norm": 4.627539157867432,
+      "learning_rate": 3.415071770334928e-05,
+      "loss": 1.9492,
+      "step": 276
+    },
+    {
+      "epoch": 0.6548463356973995,
+      "grad_norm": 3.8552463054656982,
+      "learning_rate": 3.409090909090909e-05,
+      "loss": 1.9489,
+      "step": 277
+    },
+    {
+      "epoch": 0.6572104018912529,
+      "grad_norm": 3.8745601177215576,
+      "learning_rate": 3.40311004784689e-05,
+      "loss": 1.9182,
+      "step": 278
+    },
+    {
+      "epoch": 0.6595744680851063,
+      "grad_norm": 4.04236364364624,
+      "learning_rate": 3.397129186602871e-05,
+      "loss": 1.9581,
+      "step": 279
+    },
+    {
+      "epoch": 0.6619385342789598,
+      "grad_norm": 3.5719153881073,
+      "learning_rate": 3.3911483253588514e-05,
+      "loss": 1.9321,
+      "step": 280
+    },
+    {
+      "epoch": 0.6643026004728132,
+      "grad_norm": 4.358823776245117,
+      "learning_rate": 3.385167464114833e-05,
+      "loss": 1.9567,
+      "step": 281
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 4.5859293937683105,
+      "learning_rate": 3.379186602870813e-05,
+      "loss": 1.966,
+      "step": 282
+    },
+    {
+      "epoch": 0.6690307328605201,
+      "grad_norm": 4.17390251159668,
+      "learning_rate": 3.373205741626794e-05,
+      "loss": 2.0129,
+      "step": 283
+    },
+    {
+      "epoch": 0.6713947990543735,
+      "grad_norm": 4.25508975982666,
+      "learning_rate": 3.367224880382775e-05,
+      "loss": 1.9553,
+      "step": 284
+    },
+    {
+      "epoch": 0.6737588652482269,
+      "grad_norm": 3.8857264518737793,
+      "learning_rate": 3.361244019138756e-05,
+      "loss": 1.9755,
+      "step": 285
+    },
+    {
+      "epoch": 0.6761229314420804,
+      "grad_norm": 4.436855792999268,
+      "learning_rate": 3.355263157894737e-05,
+      "loss": 1.9313,
+      "step": 286
+    },
+    {
+      "epoch": 0.6784869976359338,
+      "grad_norm": 3.8671867847442627,
+      "learning_rate": 3.349282296650718e-05,
+      "loss": 1.9402,
+      "step": 287
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 3.6638526916503906,
+      "learning_rate": 3.343301435406699e-05,
+      "loss": 1.871,
+      "step": 288
+    },
+    {
+      "epoch": 0.6832151300236406,
+      "grad_norm": 4.047444820404053,
+      "learning_rate": 3.337320574162679e-05,
+      "loss": 1.9283,
+      "step": 289
+    },
+    {
+      "epoch": 0.6855791962174941,
+      "grad_norm": 3.7109580039978027,
+      "learning_rate": 3.33133971291866e-05,
+      "loss": 1.9663,
+      "step": 290
+    },
+    {
+      "epoch": 0.6879432624113475,
+      "grad_norm": 3.8624162673950195,
+      "learning_rate": 3.325358851674641e-05,
+      "loss": 1.9802,
+      "step": 291
+    },
+    {
+      "epoch": 0.6903073286052009,
+      "grad_norm": 3.6883432865142822,
+      "learning_rate": 3.319377990430622e-05,
+      "loss": 1.9165,
+      "step": 292
+    },
+    {
+      "epoch": 0.6926713947990544,
+      "grad_norm": 3.7664642333984375,
+      "learning_rate": 3.313397129186603e-05,
+      "loss": 1.9822,
+      "step": 293
+    },
+    {
+      "epoch": 0.6950354609929078,
+      "grad_norm": 3.6505606174468994,
+      "learning_rate": 3.307416267942584e-05,
+      "loss": 1.8814,
+      "step": 294
+    },
+    {
+      "epoch": 0.6973995271867612,
+      "grad_norm": 3.5809507369995117,
+      "learning_rate": 3.301435406698565e-05,
+      "loss": 1.9044,
+      "step": 295
+    },
+    {
+      "epoch": 0.6997635933806147,
+      "grad_norm": 3.8675575256347656,
+      "learning_rate": 3.295454545454545e-05,
+      "loss": 1.8942,
+      "step": 296
+    },
+    {
+      "epoch": 0.7021276595744681,
+      "grad_norm": 3.916057586669922,
+      "learning_rate": 3.289473684210527e-05,
+      "loss": 2.0424,
+      "step": 297
+    },
+    {
+      "epoch": 0.7044917257683215,
+      "grad_norm": 3.671261787414551,
+      "learning_rate": 3.283492822966507e-05,
+      "loss": 1.9595,
+      "step": 298
+    },
+    {
+      "epoch": 0.706855791962175,
+      "grad_norm": 4.06015682220459,
+      "learning_rate": 3.277511961722488e-05,
+      "loss": 1.8396,
+      "step": 299
+    },
+    {
+      "epoch": 0.7092198581560284,
+      "grad_norm": 3.9253146648406982,
+      "learning_rate": 3.271531100478469e-05,
+      "loss": 1.9823,
+      "step": 300
+    },
+    {
+      "epoch": 0.7115839243498818,
+      "grad_norm": 4.20823335647583,
+      "learning_rate": 3.26555023923445e-05,
+      "loss": 1.9768,
+      "step": 301
+    },
+    {
+      "epoch": 0.7139479905437353,
+      "grad_norm": 4.035942554473877,
+      "learning_rate": 3.259569377990431e-05,
+      "loss": 2.0076,
+      "step": 302
+    },
+    {
+      "epoch": 0.7163120567375887,
+      "grad_norm": 3.9924399852752686,
+      "learning_rate": 3.253588516746411e-05,
+      "loss": 1.9217,
+      "step": 303
+    },
+    {
+      "epoch": 0.7186761229314421,
+      "grad_norm": 3.8201253414154053,
+      "learning_rate": 3.247607655502393e-05,
+      "loss": 1.9539,
+      "step": 304
+    },
+    {
+      "epoch": 0.7210401891252955,
+      "grad_norm": 4.5720696449279785,
+      "learning_rate": 3.241626794258373e-05,
+      "loss": 1.9805,
+      "step": 305
+    },
+    {
+      "epoch": 0.723404255319149,
+      "grad_norm": 4.060619354248047,
+      "learning_rate": 3.235645933014354e-05,
+      "loss": 1.9667,
+      "step": 306
+    },
+    {
+      "epoch": 0.7257683215130024,
+      "grad_norm": 4.097210884094238,
+      "learning_rate": 3.229665071770335e-05,
+      "loss": 1.9609,
+      "step": 307
+    },
+    {
+      "epoch": 0.7281323877068558,
+      "grad_norm": 3.757749080657959,
+      "learning_rate": 3.223684210526316e-05,
+      "loss": 2.0129,
+      "step": 308
+    },
+    {
+      "epoch": 0.7304964539007093,
+      "grad_norm": 4.25848388671875,
+      "learning_rate": 3.217703349282297e-05,
+      "loss": 2.0603,
+      "step": 309
+    },
+    {
+      "epoch": 0.7328605200945626,
+      "grad_norm": 3.9824750423431396,
+      "learning_rate": 3.211722488038278e-05,
+      "loss": 1.8766,
+      "step": 310
+    },
+    {
+      "epoch": 0.735224586288416,
+      "grad_norm": 4.179236888885498,
+      "learning_rate": 3.205741626794259e-05,
+      "loss": 1.9354,
+      "step": 311
+    },
+    {
+      "epoch": 0.7375886524822695,
+      "grad_norm": 3.9761154651641846,
+      "learning_rate": 3.199760765550239e-05,
+      "loss": 1.9749,
+      "step": 312
+    },
+    {
+      "epoch": 0.7399527186761229,
+      "grad_norm": 3.731112241744995,
+      "learning_rate": 3.19377990430622e-05,
+      "loss": 1.9597,
+      "step": 313
+    },
+    {
+      "epoch": 0.7423167848699763,
+      "grad_norm": 3.762098789215088,
+      "learning_rate": 3.187799043062201e-05,
+      "loss": 1.8983,
+      "step": 314
+    },
+    {
+      "epoch": 0.7446808510638298,
+      "grad_norm": 4.024747371673584,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 1.9149,
+      "step": 315
+    },
+    {
+      "epoch": 0.7470449172576832,
+      "grad_norm": 4.643849849700928,
+      "learning_rate": 3.175837320574162e-05,
+      "loss": 1.9244,
+      "step": 316
+    },
+    {
+      "epoch": 0.7494089834515366,
+      "grad_norm": 4.206188201904297,
+      "learning_rate": 3.169856459330144e-05,
+      "loss": 1.9201,
+      "step": 317
+    },
+    {
+      "epoch": 0.75177304964539,
+      "grad_norm": 3.647243022918701,
+      "learning_rate": 3.163875598086124e-05,
+      "loss": 1.8717,
+      "step": 318
+    },
+    {
+      "epoch": 0.7541371158392435,
+      "grad_norm": 3.7720701694488525,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 1.8891,
+      "step": 319
+    },
+    {
+      "epoch": 0.7565011820330969,
+      "grad_norm": 3.8284690380096436,
+      "learning_rate": 3.151913875598087e-05,
+      "loss": 1.9221,
+      "step": 320
+    },
+    {
+      "epoch": 0.7588652482269503,
+      "grad_norm": 4.109698295593262,
+      "learning_rate": 3.145933014354067e-05,
+      "loss": 1.8875,
+      "step": 321
+    },
+    {
+      "epoch": 0.7612293144208038,
+      "grad_norm": 3.684807062149048,
+      "learning_rate": 3.139952153110048e-05,
+      "loss": 1.9002,
+      "step": 322
+    },
+    {
+      "epoch": 0.7635933806146572,
+      "grad_norm": 3.700782299041748,
+      "learning_rate": 3.133971291866029e-05,
+      "loss": 1.9455,
+      "step": 323
+    },
+    {
+      "epoch": 0.7659574468085106,
+      "grad_norm": 3.9737021923065186,
+      "learning_rate": 3.12799043062201e-05,
+      "loss": 1.9065,
+      "step": 324
+    },
+    {
+      "epoch": 0.7683215130023641,
+      "grad_norm": 3.3986127376556396,
+      "learning_rate": 3.12200956937799e-05,
+      "loss": 1.8713,
+      "step": 325
+    },
+    {
+      "epoch": 0.7706855791962175,
+      "grad_norm": 3.8407750129699707,
+      "learning_rate": 3.116028708133971e-05,
+      "loss": 1.8779,
+      "step": 326
+    },
+    {
+      "epoch": 0.7730496453900709,
+      "grad_norm": 4.058707237243652,
+      "learning_rate": 3.110047846889952e-05,
+      "loss": 1.8802,
+      "step": 327
+    },
+    {
+      "epoch": 0.7754137115839244,
+      "grad_norm": 3.9604055881500244,
+      "learning_rate": 3.104066985645933e-05,
+      "loss": 1.8403,
+      "step": 328
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 4.052743434906006,
+      "learning_rate": 3.098086124401914e-05,
+      "loss": 1.9294,
+      "step": 329
+    },
+    {
+      "epoch": 0.7801418439716312,
+      "grad_norm": 3.3787264823913574,
+      "learning_rate": 3.092105263157895e-05,
+      "loss": 1.8994,
+      "step": 330
+    },
+    {
+      "epoch": 0.7825059101654847,
+      "grad_norm": 3.8170766830444336,
+      "learning_rate": 3.086124401913876e-05,
+      "loss": 1.9975,
+      "step": 331
+    },
+    {
+      "epoch": 0.7848699763593381,
+      "grad_norm": 3.8066794872283936,
+      "learning_rate": 3.080143540669856e-05,
+      "loss": 1.968,
+      "step": 332
+    },
+    {
+      "epoch": 0.7872340425531915,
+      "grad_norm": 4.192262172698975,
+      "learning_rate": 3.074162679425838e-05,
+      "loss": 1.9784,
+      "step": 333
+    },
+    {
+      "epoch": 0.789598108747045,
+      "grad_norm": 3.9428586959838867,
+      "learning_rate": 3.068181818181818e-05,
+      "loss": 1.9362,
+      "step": 334
+    },
+    {
+      "epoch": 0.7919621749408984,
+      "grad_norm": 4.186812400817871,
+      "learning_rate": 3.062200956937799e-05,
+      "loss": 1.9212,
+      "step": 335
+    },
+    {
+      "epoch": 0.7943262411347518,
+      "grad_norm": 4.045288562774658,
+      "learning_rate": 3.05622009569378e-05,
+      "loss": 1.9017,
+      "step": 336
+    },
+    {
+      "epoch": 0.7966903073286052,
+      "grad_norm": 4.203200340270996,
+      "learning_rate": 3.050239234449761e-05,
+      "loss": 1.9844,
+      "step": 337
+    },
+    {
+      "epoch": 0.7990543735224587,
+      "grad_norm": 4.394277095794678,
+      "learning_rate": 3.0442583732057416e-05,
+      "loss": 1.9332,
+      "step": 338
+    },
+    {
+      "epoch": 0.8014184397163121,
+      "grad_norm": 3.7559564113616943,
+      "learning_rate": 3.0382775119617225e-05,
+      "loss": 1.9142,
+      "step": 339
+    },
+    {
+      "epoch": 0.8037825059101655,
+      "grad_norm": 3.5842766761779785,
+      "learning_rate": 3.0322966507177035e-05,
+      "loss": 1.8843,
+      "step": 340
+    },
+    {
+      "epoch": 0.806146572104019,
+      "grad_norm": 3.8942437171936035,
+      "learning_rate": 3.0263157894736844e-05,
+      "loss": 1.825,
+      "step": 341
+    },
+    {
+      "epoch": 0.8085106382978723,
+      "grad_norm": 4.06583833694458,
+      "learning_rate": 3.020334928229665e-05,
+      "loss": 1.8859,
+      "step": 342
+    },
+    {
+      "epoch": 0.8108747044917257,
+      "grad_norm": 3.9897472858428955,
+      "learning_rate": 3.0143540669856463e-05,
+      "loss": 2.0088,
+      "step": 343
+    },
+    {
+      "epoch": 0.8132387706855791,
+      "grad_norm": 4.120972633361816,
+      "learning_rate": 3.008373205741627e-05,
+      "loss": 1.9644,
+      "step": 344
+    },
+    {
+      "epoch": 0.8156028368794326,
+      "grad_norm": 3.761667013168335,
+      "learning_rate": 3.0023923444976076e-05,
+      "loss": 1.8468,
+      "step": 345
+    },
+    {
+      "epoch": 0.817966903073286,
+      "grad_norm": 4.167830944061279,
+      "learning_rate": 2.996411483253589e-05,
+      "loss": 1.8995,
+      "step": 346
+    },
+    {
+      "epoch": 0.8203309692671394,
+      "grad_norm": 3.732332944869995,
+      "learning_rate": 2.9904306220095695e-05,
+      "loss": 1.8529,
+      "step": 347
+    },
+    {
+      "epoch": 0.8226950354609929,
+      "grad_norm": 4.0308003425598145,
+      "learning_rate": 2.9844497607655504e-05,
+      "loss": 1.9401,
+      "step": 348
+    },
+    {
+      "epoch": 0.8250591016548463,
+      "grad_norm": 3.981724500656128,
+      "learning_rate": 2.9784688995215314e-05,
+      "loss": 1.8821,
+      "step": 349
+    },
+    {
+      "epoch": 0.8274231678486997,
+      "grad_norm": 3.8602871894836426,
+      "learning_rate": 2.9724880382775123e-05,
+      "loss": 1.8909,
+      "step": 350
+    },
+    {
+      "epoch": 0.8297872340425532,
+      "grad_norm": 3.8556690216064453,
+      "learning_rate": 2.966507177033493e-05,
+      "loss": 1.9972,
+      "step": 351
+    },
+    {
+      "epoch": 0.8321513002364066,
+      "grad_norm": 3.716454029083252,
+      "learning_rate": 2.9605263157894735e-05,
+      "loss": 1.9235,
+      "step": 352
+    },
+    {
+      "epoch": 0.83451536643026,
+      "grad_norm": 4.057294845581055,
+      "learning_rate": 2.954545454545455e-05,
+      "loss": 1.8717,
+      "step": 353
+    },
+    {
+      "epoch": 0.8368794326241135,
+      "grad_norm": 3.5962278842926025,
+      "learning_rate": 2.9485645933014355e-05,
+      "loss": 1.9414,
+      "step": 354
+    },
+    {
+      "epoch": 0.8392434988179669,
+      "grad_norm": 4.190985679626465,
+      "learning_rate": 2.942583732057416e-05,
+      "loss": 1.9724,
+      "step": 355
+    },
+    {
+      "epoch": 0.8416075650118203,
+      "grad_norm": 3.9760379791259766,
+      "learning_rate": 2.9366028708133974e-05,
+      "loss": 1.839,
+      "step": 356
+    },
+    {
+      "epoch": 0.8439716312056738,
+      "grad_norm": 3.629091501235962,
+      "learning_rate": 2.9306220095693783e-05,
+      "loss": 1.8866,
+      "step": 357
+    },
+    {
+      "epoch": 0.8463356973995272,
+      "grad_norm": 3.752070188522339,
+      "learning_rate": 2.924641148325359e-05,
+      "loss": 1.8868,
+      "step": 358
+    },
+    {
+      "epoch": 0.8486997635933806,
+      "grad_norm": 3.5992238521575928,
+      "learning_rate": 2.9186602870813402e-05,
+      "loss": 1.8586,
+      "step": 359
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 3.47458553314209,
+      "learning_rate": 2.912679425837321e-05,
+      "loss": 1.9659,
+      "step": 360
+    },
+    {
+      "epoch": 0.8534278959810875,
+      "grad_norm": 3.6117656230926514,
+      "learning_rate": 2.9066985645933014e-05,
+      "loss": 1.8892,
+      "step": 361
+    },
+    {
+      "epoch": 0.8557919621749409,
+      "grad_norm": 4.080473899841309,
+      "learning_rate": 2.900717703349282e-05,
+      "loss": 1.8809,
+      "step": 362
+    },
+    {
+      "epoch": 0.8581560283687943,
+      "grad_norm": 4.260461330413818,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 1.9461,
+      "step": 363
+    },
+    {
+      "epoch": 0.8605200945626478,
+      "grad_norm": 4.231245994567871,
+      "learning_rate": 2.888755980861244e-05,
+      "loss": 1.9389,
+      "step": 364
+    },
+    {
+      "epoch": 0.8628841607565012,
+      "grad_norm": 3.64261794090271,
+      "learning_rate": 2.882775119617225e-05,
+      "loss": 1.922,
+      "step": 365
+    },
+    {
+      "epoch": 0.8652482269503546,
+      "grad_norm": 3.591475009918213,
+      "learning_rate": 2.8767942583732062e-05,
+      "loss": 1.9616,
+      "step": 366
+    },
+    {
+      "epoch": 0.8676122931442081,
+      "grad_norm": 3.9587414264678955,
+      "learning_rate": 2.8708133971291868e-05,
+      "loss": 1.9045,
+      "step": 367
+    },
+    {
+      "epoch": 0.8699763593380615,
+      "grad_norm": 3.6751394271850586,
+      "learning_rate": 2.8648325358851674e-05,
+      "loss": 1.9077,
+      "step": 368
+    },
+    {
+      "epoch": 0.8723404255319149,
+      "grad_norm": 4.2092790603637695,
+      "learning_rate": 2.8588516746411487e-05,
+      "loss": 1.9413,
+      "step": 369
+    },
+    {
+      "epoch": 0.8747044917257684,
+      "grad_norm": 3.814706325531006,
+      "learning_rate": 2.8528708133971293e-05,
+      "loss": 1.9203,
+      "step": 370
+    },
+    {
+      "epoch": 0.8770685579196218,
+      "grad_norm": 3.674201250076294,
+      "learning_rate": 2.84688995215311e-05,
+      "loss": 1.8488,
+      "step": 371
+    },
+    {
+      "epoch": 0.8794326241134752,
+      "grad_norm": 3.5166468620300293,
+      "learning_rate": 2.8409090909090912e-05,
+      "loss": 1.9175,
+      "step": 372
+    },
+    {
+      "epoch": 0.8817966903073287,
+      "grad_norm": 3.619014024734497,
+      "learning_rate": 2.834928229665072e-05,
+      "loss": 1.8396,
+      "step": 373
+    },
+    {
+      "epoch": 0.8841607565011821,
+      "grad_norm": 3.923396110534668,
+      "learning_rate": 2.8289473684210528e-05,
+      "loss": 1.8317,
+      "step": 374
+    },
+    {
+      "epoch": 0.8865248226950354,
+      "grad_norm": 3.934695243835449,
+      "learning_rate": 2.8229665071770334e-05,
+      "loss": 1.9562,
+      "step": 375
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 3.761104106903076,
+      "learning_rate": 2.8169856459330147e-05,
+      "loss": 1.7913,
+      "step": 376
+    },
+    {
+      "epoch": 0.8912529550827423,
+      "grad_norm": 3.7853753566741943,
+      "learning_rate": 2.8110047846889953e-05,
+      "loss": 1.8132,
+      "step": 377
+    },
+    {
+      "epoch": 0.8936170212765957,
+      "grad_norm": 3.526927947998047,
+      "learning_rate": 2.805023923444976e-05,
+      "loss": 1.6913,
+      "step": 378
+    },
+    {
+      "epoch": 0.8959810874704491,
+      "grad_norm": 3.75763201713562,
+      "learning_rate": 2.7990430622009572e-05,
+      "loss": 1.8787,
+      "step": 379
+    },
+    {
+      "epoch": 0.8983451536643026,
+      "grad_norm": 3.601562023162842,
+      "learning_rate": 2.793062200956938e-05,
+      "loss": 1.9166,
+      "step": 380
+    },
+    {
+      "epoch": 0.900709219858156,
+      "grad_norm": 3.5951952934265137,
+      "learning_rate": 2.7870813397129185e-05,
+      "loss": 1.885,
+      "step": 381
+    },
+    {
+      "epoch": 0.9030732860520094,
+      "grad_norm": 3.5643539428710938,
+      "learning_rate": 2.7811004784688998e-05,
+      "loss": 1.9044,
+      "step": 382
+    },
+    {
+      "epoch": 0.9054373522458629,
+      "grad_norm": 3.7953860759735107,
+      "learning_rate": 2.7751196172248807e-05,
+      "loss": 1.9872,
+      "step": 383
+    },
+    {
+      "epoch": 0.9078014184397163,
+      "grad_norm": 4.0880913734436035,
+      "learning_rate": 2.7691387559808613e-05,
+      "loss": 1.9078,
+      "step": 384
+    },
+    {
+      "epoch": 0.9101654846335697,
+      "grad_norm": 3.8961236476898193,
+      "learning_rate": 2.7631578947368426e-05,
+      "loss": 1.8843,
+      "step": 385
+    },
+    {
+      "epoch": 0.9125295508274232,
+      "grad_norm": 3.7427453994750977,
+      "learning_rate": 2.7571770334928232e-05,
+      "loss": 1.8713,
+      "step": 386
+    },
+    {
+      "epoch": 0.9148936170212766,
+      "grad_norm": 3.7328555583953857,
+      "learning_rate": 2.751196172248804e-05,
+      "loss": 1.8305,
+      "step": 387
+    },
+    {
+      "epoch": 0.91725768321513,
+      "grad_norm": 3.890418291091919,
+      "learning_rate": 2.7452153110047845e-05,
+      "loss": 1.8197,
+      "step": 388
+    },
+    {
+      "epoch": 0.9196217494089834,
+      "grad_norm": 3.7216286659240723,
+      "learning_rate": 2.7392344497607657e-05,
+      "loss": 1.9201,
+      "step": 389
+    },
+    {
+      "epoch": 0.9219858156028369,
+      "grad_norm": 3.705873489379883,
+      "learning_rate": 2.7332535885167464e-05,
+      "loss": 1.8885,
+      "step": 390
+    },
+    {
+      "epoch": 0.9243498817966903,
+      "grad_norm": 3.5170631408691406,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 1.895,
+      "step": 391
+    },
+    {
+      "epoch": 0.9267139479905437,
+      "grad_norm": 3.632924795150757,
+      "learning_rate": 2.7212918660287086e-05,
+      "loss": 1.9286,
+      "step": 392
+    },
+    {
+      "epoch": 0.9290780141843972,
+      "grad_norm": 4.132338523864746,
+      "learning_rate": 2.7153110047846892e-05,
+      "loss": 1.9079,
+      "step": 393
+    },
+    {
+      "epoch": 0.9314420803782506,
+      "grad_norm": 3.8694465160369873,
+      "learning_rate": 2.7093301435406698e-05,
+      "loss": 1.856,
+      "step": 394
+    },
+    {
+      "epoch": 0.933806146572104,
+      "grad_norm": 4.146971702575684,
+      "learning_rate": 2.703349282296651e-05,
+      "loss": 1.9795,
+      "step": 395
+    },
+    {
+      "epoch": 0.9361702127659575,
+      "grad_norm": 3.581249952316284,
+      "learning_rate": 2.6973684210526317e-05,
+      "loss": 1.868,
+      "step": 396
+    },
+    {
+      "epoch": 0.9385342789598109,
+      "grad_norm": 3.779081106185913,
+      "learning_rate": 2.6913875598086123e-05,
+      "loss": 1.8462,
+      "step": 397
+    },
+    {
+      "epoch": 0.9408983451536643,
+      "grad_norm": 3.373218536376953,
+      "learning_rate": 2.6854066985645936e-05,
+      "loss": 1.8336,
+      "step": 398
+    },
+    {
+      "epoch": 0.9432624113475178,
+      "grad_norm": 3.7768990993499756,
+      "learning_rate": 2.6794258373205743e-05,
+      "loss": 1.9062,
+      "step": 399
+    },
+    {
+      "epoch": 0.9456264775413712,
+      "grad_norm": 3.4512805938720703,
+      "learning_rate": 2.6734449760765552e-05,
+      "loss": 1.873,
+      "step": 400
+    },
+    {
+      "epoch": 0.9479905437352246,
+      "grad_norm": 3.38236927986145,
+      "learning_rate": 2.6674641148325358e-05,
+      "loss": 1.8911,
+      "step": 401
+    },
+    {
+      "epoch": 0.950354609929078,
+      "grad_norm": 3.191875696182251,
+      "learning_rate": 2.661483253588517e-05,
+      "loss": 1.8732,
+      "step": 402
+    },
+    {
+      "epoch": 0.9527186761229315,
+      "grad_norm": 3.671778440475464,
+      "learning_rate": 2.6555023923444977e-05,
+      "loss": 1.8879,
+      "step": 403
+    },
+    {
+      "epoch": 0.9550827423167849,
+      "grad_norm": 3.831817150115967,
+      "learning_rate": 2.6495215311004783e-05,
+      "loss": 1.8256,
+      "step": 404
+    },
+    {
+      "epoch": 0.9574468085106383,
+      "grad_norm": 3.432061195373535,
+      "learning_rate": 2.6435406698564596e-05,
+      "loss": 1.8245,
+      "step": 405
+    },
+    {
+      "epoch": 0.9598108747044918,
+      "grad_norm": 3.591796398162842,
+      "learning_rate": 2.6375598086124402e-05,
+      "loss": 1.8954,
+      "step": 406
+    },
+    {
+      "epoch": 0.9621749408983451,
+      "grad_norm": 3.4541237354278564,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 1.8886,
+      "step": 407
+    },
+    {
+      "epoch": 0.9645390070921985,
+      "grad_norm": 3.4565842151641846,
+      "learning_rate": 2.625598086124402e-05,
+      "loss": 1.8766,
+      "step": 408
+    },
+    {
+      "epoch": 0.966903073286052,
+      "grad_norm": 3.812185049057007,
+      "learning_rate": 2.619617224880383e-05,
+      "loss": 1.8797,
+      "step": 409
+    },
+    {
+      "epoch": 0.9692671394799054,
+      "grad_norm": 4.211532115936279,
+      "learning_rate": 2.6136363636363637e-05,
+      "loss": 1.8563,
+      "step": 410
+    },
+    {
+      "epoch": 0.9716312056737588,
+      "grad_norm": 3.5806126594543457,
+      "learning_rate": 2.6076555023923443e-05,
+      "loss": 1.9192,
+      "step": 411
+    },
+    {
+      "epoch": 0.9739952718676123,
+      "grad_norm": 3.7554843425750732,
+      "learning_rate": 2.6016746411483256e-05,
+      "loss": 1.8267,
+      "step": 412
+    },
+    {
+      "epoch": 0.9763593380614657,
+      "grad_norm": 3.9262287616729736,
+      "learning_rate": 2.5956937799043062e-05,
+      "loss": 1.9415,
+      "step": 413
+    },
+    {
+      "epoch": 0.9787234042553191,
+      "grad_norm": 3.754761219024658,
+      "learning_rate": 2.589712918660287e-05,
+      "loss": 1.8816,
+      "step": 414
+    },
+    {
+      "epoch": 0.9810874704491725,
+      "grad_norm": 3.463529586791992,
+      "learning_rate": 2.583732057416268e-05,
+      "loss": 1.8487,
+      "step": 415
+    },
+    {
+      "epoch": 0.983451536643026,
+      "grad_norm": 3.6155738830566406,
+      "learning_rate": 2.5777511961722488e-05,
+      "loss": 1.8339,
+      "step": 416
+    },
+    {
+      "epoch": 0.9858156028368794,
+      "grad_norm": 3.745180130004883,
+      "learning_rate": 2.5717703349282297e-05,
+      "loss": 1.8555,
+      "step": 417
+    },
+    {
+      "epoch": 0.9881796690307328,
+      "grad_norm": 3.9255855083465576,
+      "learning_rate": 2.565789473684211e-05,
+      "loss": 1.8148,
+      "step": 418
+    },
+    {
+      "epoch": 0.9905437352245863,
+      "grad_norm": 4.076484203338623,
+      "learning_rate": 2.5598086124401916e-05,
+      "loss": 1.8527,
+      "step": 419
+    },
+    {
+      "epoch": 0.9929078014184397,
+      "grad_norm": 3.6310737133026123,
+      "learning_rate": 2.5538277511961722e-05,
+      "loss": 1.8875,
+      "step": 420
+    },
+    {
+      "epoch": 0.9952718676122931,
+      "grad_norm": 3.757092237472534,
+      "learning_rate": 2.5478468899521535e-05,
+      "loss": 1.9363,
+      "step": 421
+    },
+    {
+      "epoch": 0.9976359338061466,
+      "grad_norm": 3.754251003265381,
+      "learning_rate": 2.541866028708134e-05,
+      "loss": 1.8591,
+      "step": 422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.948606014251709,
+      "learning_rate": 2.5358851674641147e-05,
+      "loss": 1.8516,
+      "step": 423
+    },
+    {
+      "epoch": 1.0023640661938533,
+      "grad_norm": 3.393385171890259,
+      "learning_rate": 2.5299043062200957e-05,
+      "loss": 1.7089,
+      "step": 424
+    },
+    {
+      "epoch": 1.0047281323877069,
+      "grad_norm": 3.5038678646087646,
+      "learning_rate": 2.5239234449760766e-05,
+      "loss": 1.6513,
+      "step": 425
+    },
+    {
+      "epoch": 1.0070921985815602,
+      "grad_norm": 3.546327590942383,
+      "learning_rate": 2.5179425837320576e-05,
+      "loss": 1.6913,
+      "step": 426
+    },
+    {
+      "epoch": 1.0094562647754137,
+      "grad_norm": 3.012467384338379,
+      "learning_rate": 2.5119617224880382e-05,
+      "loss": 1.671,
+      "step": 427
+    },
+    {
+      "epoch": 1.011820330969267,
+      "grad_norm": 3.5642528533935547,
+      "learning_rate": 2.5059808612440195e-05,
+      "loss": 1.7187,
+      "step": 428
+    },
+    {
+      "epoch": 1.0141843971631206,
+      "grad_norm": 3.3508832454681396,
+      "learning_rate": 2.5e-05,
+      "loss": 1.6118,
+      "step": 429
+    },
+    {
+      "epoch": 1.016548463356974,
+      "grad_norm": 3.450350761413574,
+      "learning_rate": 2.494019138755981e-05,
+      "loss": 1.6704,
+      "step": 430
+    },
+    {
+      "epoch": 1.0189125295508275,
+      "grad_norm": 3.859874725341797,
+      "learning_rate": 2.4880382775119617e-05,
+      "loss": 1.6719,
+      "step": 431
+    },
+    {
+      "epoch": 1.0212765957446808,
+      "grad_norm": 3.572866439819336,
+      "learning_rate": 2.4820574162679426e-05,
+      "loss": 1.6766,
+      "step": 432
+    },
+    {
+      "epoch": 1.0236406619385343,
+      "grad_norm": 3.1181817054748535,
+      "learning_rate": 2.4760765550239236e-05,
+      "loss": 1.623,
+      "step": 433
+    },
+    {
+      "epoch": 1.0260047281323876,
+      "grad_norm": 3.3449785709381104,
+      "learning_rate": 2.4700956937799045e-05,
+      "loss": 1.6767,
+      "step": 434
+    },
+    {
+      "epoch": 1.0283687943262412,
+      "grad_norm": 3.494570732116699,
+      "learning_rate": 2.4641148325358855e-05,
+      "loss": 1.6433,
+      "step": 435
+    },
+    {
+      "epoch": 1.0307328605200945,
+      "grad_norm": 3.3296971321105957,
+      "learning_rate": 2.458133971291866e-05,
+      "loss": 1.6977,
+      "step": 436
+    },
+    {
+      "epoch": 1.033096926713948,
+      "grad_norm": 3.5671586990356445,
+      "learning_rate": 2.452153110047847e-05,
+      "loss": 1.6876,
+      "step": 437
+    },
+    {
+      "epoch": 1.0354609929078014,
+      "grad_norm": 3.4455606937408447,
+      "learning_rate": 2.446172248803828e-05,
+      "loss": 1.5463,
+      "step": 438
+    },
+    {
+      "epoch": 1.037825059101655,
+      "grad_norm": 3.561481237411499,
+      "learning_rate": 2.4401913875598086e-05,
+      "loss": 1.6733,
+      "step": 439
+    },
+    {
+      "epoch": 1.0401891252955082,
+      "grad_norm": 3.6918563842773438,
+      "learning_rate": 2.4342105263157896e-05,
+      "loss": 1.6511,
+      "step": 440
+    },
+    {
+      "epoch": 1.0425531914893618,
+      "grad_norm": 3.360909938812256,
+      "learning_rate": 2.4282296650717702e-05,
+      "loss": 1.6931,
+      "step": 441
+    },
+    {
+      "epoch": 1.044917257683215,
+      "grad_norm": 3.5804953575134277,
+      "learning_rate": 2.4222488038277515e-05,
+      "loss": 1.6333,
+      "step": 442
+    },
+    {
+      "epoch": 1.0472813238770686,
+      "grad_norm": 3.5099520683288574,
+      "learning_rate": 2.4162679425837324e-05,
+      "loss": 1.5951,
+      "step": 443
+    },
+    {
+      "epoch": 1.049645390070922,
+      "grad_norm": 3.4295504093170166,
+      "learning_rate": 2.410287081339713e-05,
+      "loss": 1.6764,
+      "step": 444
+    },
+    {
+      "epoch": 1.0520094562647755,
+      "grad_norm": 3.5318591594696045,
+      "learning_rate": 2.404306220095694e-05,
+      "loss": 1.6894,
+      "step": 445
+    },
+    {
+      "epoch": 1.0543735224586288,
+      "grad_norm": 3.4848272800445557,
+      "learning_rate": 2.3983253588516746e-05,
+      "loss": 1.6769,
+      "step": 446
+    },
+    {
+      "epoch": 1.0567375886524824,
+      "grad_norm": 3.7782180309295654,
+      "learning_rate": 2.3923444976076556e-05,
+      "loss": 1.6239,
+      "step": 447
+    },
+    {
+      "epoch": 1.0591016548463357,
+      "grad_norm": 3.2487025260925293,
+      "learning_rate": 2.3863636363636365e-05,
+      "loss": 1.5877,
+      "step": 448
+    },
+    {
+      "epoch": 1.0614657210401892,
+      "grad_norm": 3.6180076599121094,
+      "learning_rate": 2.380382775119617e-05,
+      "loss": 1.685,
+      "step": 449
+    },
+    {
+      "epoch": 1.0638297872340425,
+      "grad_norm": 3.6394782066345215,
+      "learning_rate": 2.374401913875598e-05,
+      "loss": 1.6699,
+      "step": 450
+    },
+    {
+      "epoch": 1.066193853427896,
+      "grad_norm": 3.7263615131378174,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 1.6681,
+      "step": 451
+    },
+    {
+      "epoch": 1.0685579196217494,
+      "grad_norm": 3.455543279647827,
+      "learning_rate": 2.36244019138756e-05,
+      "loss": 1.5929,
+      "step": 452
+    },
+    {
+      "epoch": 1.070921985815603,
+      "grad_norm": 3.379056930541992,
+      "learning_rate": 2.356459330143541e-05,
+      "loss": 1.7262,
+      "step": 453
+    },
+    {
+      "epoch": 1.0732860520094563,
+      "grad_norm": 3.415682792663574,
+      "learning_rate": 2.3504784688995216e-05,
+      "loss": 1.6502,
+      "step": 454
+    },
+    {
+      "epoch": 1.0756501182033098,
+      "grad_norm": 3.3975017070770264,
+      "learning_rate": 2.3444976076555025e-05,
+      "loss": 1.6846,
+      "step": 455
+    },
+    {
+      "epoch": 1.0780141843971631,
+      "grad_norm": 3.844403028488159,
+      "learning_rate": 2.3385167464114835e-05,
+      "loss": 1.6581,
+      "step": 456
+    },
+    {
+      "epoch": 1.0803782505910164,
+      "grad_norm": 3.237973690032959,
+      "learning_rate": 2.332535885167464e-05,
+      "loss": 1.6452,
+      "step": 457
+    },
+    {
+      "epoch": 1.08274231678487,
+      "grad_norm": 3.138275384902954,
+      "learning_rate": 2.326555023923445e-05,
+      "loss": 1.671,
+      "step": 458
+    },
+    {
+      "epoch": 1.0851063829787233,
+      "grad_norm": 3.2867116928100586,
+      "learning_rate": 2.320574162679426e-05,
+      "loss": 1.7039,
+      "step": 459
+    },
+    {
+      "epoch": 1.0874704491725768,
+      "grad_norm": 3.331429958343506,
+      "learning_rate": 2.314593301435407e-05,
+      "loss": 1.6251,
+      "step": 460
+    },
+    {
+      "epoch": 1.0898345153664302,
+      "grad_norm": 3.517249822616577,
+      "learning_rate": 2.308612440191388e-05,
+      "loss": 1.6588,
+      "step": 461
+    },
+    {
+      "epoch": 1.0921985815602837,
+      "grad_norm": 3.447352170944214,
+      "learning_rate": 2.3026315789473685e-05,
+      "loss": 1.6494,
+      "step": 462
+    },
+    {
+      "epoch": 1.094562647754137,
+      "grad_norm": 3.820619583129883,
+      "learning_rate": 2.2966507177033495e-05,
+      "loss": 1.6767,
+      "step": 463
+    },
+    {
+      "epoch": 1.0969267139479906,
+      "grad_norm": 3.612136125564575,
+      "learning_rate": 2.29066985645933e-05,
+      "loss": 1.6123,
+      "step": 464
+    },
+    {
+      "epoch": 1.099290780141844,
+      "grad_norm": 3.2653629779815674,
+      "learning_rate": 2.284688995215311e-05,
+      "loss": 1.639,
+      "step": 465
+    },
+    {
+      "epoch": 1.1016548463356974,
+      "grad_norm": 3.241689443588257,
+      "learning_rate": 2.278708133971292e-05,
+      "loss": 1.6561,
+      "step": 466
+    },
+    {
+      "epoch": 1.1040189125295508,
+      "grad_norm": 3.3771729469299316,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 1.6457,
+      "step": 467
+    },
+    {
+      "epoch": 1.1063829787234043,
+      "grad_norm": 3.3176181316375732,
+      "learning_rate": 2.266746411483254e-05,
+      "loss": 1.6047,
+      "step": 468
+    },
+    {
+      "epoch": 1.1087470449172576,
+      "grad_norm": 3.281697988510132,
+      "learning_rate": 2.2607655502392345e-05,
+      "loss": 1.6211,
+      "step": 469
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 3.4810190200805664,
+      "learning_rate": 2.2547846889952154e-05,
+      "loss": 1.7409,
+      "step": 470
+    },
+    {
+      "epoch": 1.1134751773049645,
+      "grad_norm": 3.873317003250122,
+      "learning_rate": 2.2488038277511964e-05,
+      "loss": 1.6596,
+      "step": 471
+    },
+    {
+      "epoch": 1.115839243498818,
+      "grad_norm": 3.9520647525787354,
+      "learning_rate": 2.242822966507177e-05,
+      "loss": 1.7332,
+      "step": 472
+    },
+    {
+      "epoch": 1.1182033096926713,
+      "grad_norm": 3.922635555267334,
+      "learning_rate": 2.236842105263158e-05,
+      "loss": 1.6934,
+      "step": 473
+    },
+    {
+      "epoch": 1.1205673758865249,
+      "grad_norm": 3.404571056365967,
+      "learning_rate": 2.230861244019139e-05,
+      "loss": 1.6456,
+      "step": 474
+    },
+    {
+      "epoch": 1.1229314420803782,
+      "grad_norm": 3.497051239013672,
+      "learning_rate": 2.2248803827751195e-05,
+      "loss": 1.712,
+      "step": 475
+    },
+    {
+      "epoch": 1.1252955082742317,
+      "grad_norm": 3.632838249206543,
+      "learning_rate": 2.2188995215311005e-05,
+      "loss": 1.6521,
+      "step": 476
+    },
+    {
+      "epoch": 1.127659574468085,
+      "grad_norm": 3.8431527614593506,
+      "learning_rate": 2.2129186602870814e-05,
+      "loss": 1.6738,
+      "step": 477
+    },
+    {
+      "epoch": 1.1300236406619386,
+      "grad_norm": 3.709177255630493,
+      "learning_rate": 2.2069377990430624e-05,
+      "loss": 1.622,
+      "step": 478
+    },
+    {
+      "epoch": 1.132387706855792,
+      "grad_norm": 3.3974366188049316,
+      "learning_rate": 2.2009569377990433e-05,
+      "loss": 1.7082,
+      "step": 479
+    },
+    {
+      "epoch": 1.1347517730496455,
+      "grad_norm": 3.6680588722229004,
+      "learning_rate": 2.194976076555024e-05,
+      "loss": 1.6759,
+      "step": 480
+    },
+    {
+      "epoch": 1.1371158392434988,
+      "grad_norm": 3.3660480976104736,
+      "learning_rate": 2.188995215311005e-05,
+      "loss": 1.6893,
+      "step": 481
+    },
+    {
+      "epoch": 1.1394799054373523,
+      "grad_norm": 3.4249792098999023,
+      "learning_rate": 2.1830143540669855e-05,
+      "loss": 1.6269,
+      "step": 482
+    },
+    {
+      "epoch": 1.1418439716312057,
+      "grad_norm": 3.5676686763763428,
+      "learning_rate": 2.1770334928229665e-05,
+      "loss": 1.5851,
+      "step": 483
+    },
+    {
+      "epoch": 1.1442080378250592,
+      "grad_norm": 3.6361424922943115,
+      "learning_rate": 2.1710526315789474e-05,
+      "loss": 1.683,
+      "step": 484
+    },
+    {
+      "epoch": 1.1465721040189125,
+      "grad_norm": 3.530165910720825,
+      "learning_rate": 2.1650717703349284e-05,
+      "loss": 1.6617,
+      "step": 485
+    },
+    {
+      "epoch": 1.148936170212766,
+      "grad_norm": 3.3330204486846924,
+      "learning_rate": 2.1590909090909093e-05,
+      "loss": 1.6294,
+      "step": 486
+    },
+    {
+      "epoch": 1.1513002364066194,
+      "grad_norm": 3.3433423042297363,
+      "learning_rate": 2.1531100478468903e-05,
+      "loss": 1.5875,
+      "step": 487
+    },
+    {
+      "epoch": 1.1536643026004727,
+      "grad_norm": 3.511631488800049,
+      "learning_rate": 2.147129186602871e-05,
+      "loss": 1.6799,
+      "step": 488
+    },
+    {
+      "epoch": 1.1560283687943262,
+      "grad_norm": 3.3675262928009033,
+      "learning_rate": 2.141148325358852e-05,
+      "loss": 1.7247,
+      "step": 489
+    },
+    {
+      "epoch": 1.1583924349881798,
+      "grad_norm": 3.7090232372283936,
+      "learning_rate": 2.1351674641148325e-05,
+      "loss": 1.6302,
+      "step": 490
+    },
+    {
+      "epoch": 1.160756501182033,
+      "grad_norm": 3.7816193103790283,
+      "learning_rate": 2.1291866028708134e-05,
+      "loss": 1.7691,
+      "step": 491
+    },
+    {
+      "epoch": 1.1631205673758864,
+      "grad_norm": 3.2552871704101562,
+      "learning_rate": 2.1232057416267944e-05,
+      "loss": 1.5899,
+      "step": 492
+    },
+    {
+      "epoch": 1.16548463356974,
+      "grad_norm": 3.853459119796753,
+      "learning_rate": 2.1172248803827753e-05,
+      "loss": 1.7019,
+      "step": 493
+    },
+    {
+      "epoch": 1.1678486997635933,
+      "grad_norm": 3.5649783611297607,
+      "learning_rate": 2.1112440191387563e-05,
+      "loss": 1.6619,
+      "step": 494
+    },
+    {
+      "epoch": 1.1702127659574468,
+      "grad_norm": 3.476576805114746,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 1.611,
+      "step": 495
+    },
+    {
+      "epoch": 1.1725768321513002,
+      "grad_norm": 3.5537772178649902,
+      "learning_rate": 2.099282296650718e-05,
+      "loss": 1.637,
+      "step": 496
+    },
+    {
+      "epoch": 1.1749408983451537,
+      "grad_norm": 3.6302125453948975,
+      "learning_rate": 2.0933014354066988e-05,
+      "loss": 1.6317,
+      "step": 497
+    },
+    {
+      "epoch": 1.177304964539007,
+      "grad_norm": 3.622593879699707,
+      "learning_rate": 2.0873205741626794e-05,
+      "loss": 1.6086,
+      "step": 498
+    },
+    {
+      "epoch": 1.1796690307328606,
+      "grad_norm": 3.44309663772583,
+      "learning_rate": 2.0813397129186604e-05,
+      "loss": 1.6525,
+      "step": 499
+    },
+    {
+      "epoch": 1.1820330969267139,
+      "grad_norm": 3.0509703159332275,
+      "learning_rate": 2.075358851674641e-05,
+      "loss": 1.6,
+      "step": 500
+    },
+    {
+      "epoch": 1.1843971631205674,
+      "grad_norm": 3.322601556777954,
+      "learning_rate": 2.069377990430622e-05,
+      "loss": 1.5762,
+      "step": 501
+    },
+    {
+      "epoch": 1.1867612293144207,
+      "grad_norm": 3.471799612045288,
+      "learning_rate": 2.0633971291866032e-05,
+      "loss": 1.6196,
+      "step": 502
+    },
+    {
+      "epoch": 1.1891252955082743,
+      "grad_norm": 3.549744129180908,
+      "learning_rate": 2.057416267942584e-05,
+      "loss": 1.6898,
+      "step": 503
+    },
+    {
+      "epoch": 1.1914893617021276,
+      "grad_norm": 3.763566493988037,
+      "learning_rate": 2.0514354066985648e-05,
+      "loss": 1.6695,
+      "step": 504
+    },
+    {
+      "epoch": 1.1938534278959811,
+      "grad_norm": 3.3992598056793213,
+      "learning_rate": 2.0454545454545457e-05,
+      "loss": 1.6777,
+      "step": 505
+    },
+    {
+      "epoch": 1.1962174940898345,
+      "grad_norm": 3.303372621536255,
+      "learning_rate": 2.0394736842105264e-05,
+      "loss": 1.6111,
+      "step": 506
+    },
+    {
+      "epoch": 1.198581560283688,
+      "grad_norm": 3.443369150161743,
+      "learning_rate": 2.0334928229665073e-05,
+      "loss": 1.5834,
+      "step": 507
+    },
+    {
+      "epoch": 1.2009456264775413,
+      "grad_norm": 3.5040507316589355,
+      "learning_rate": 2.027511961722488e-05,
+      "loss": 1.6172,
+      "step": 508
+    },
+    {
+      "epoch": 1.2033096926713949,
+      "grad_norm": 3.334400177001953,
+      "learning_rate": 2.021531100478469e-05,
+      "loss": 1.6028,
+      "step": 509
+    },
+    {
+      "epoch": 1.2056737588652482,
+      "grad_norm": 3.5346176624298096,
+      "learning_rate": 2.0155502392344498e-05,
+      "loss": 1.696,
+      "step": 510
+    },
+    {
+      "epoch": 1.2080378250591017,
+      "grad_norm": 3.8326451778411865,
+      "learning_rate": 2.0095693779904308e-05,
+      "loss": 1.6668,
+      "step": 511
+    },
+    {
+      "epoch": 1.210401891252955,
+      "grad_norm": 3.3983545303344727,
+      "learning_rate": 2.0035885167464117e-05,
+      "loss": 1.6307,
+      "step": 512
+    },
+    {
+      "epoch": 1.2127659574468086,
+      "grad_norm": 3.451599597930908,
+      "learning_rate": 1.9976076555023923e-05,
+      "loss": 1.6307,
+      "step": 513
+    },
+    {
+      "epoch": 1.215130023640662,
+      "grad_norm": 3.855468988418579,
+      "learning_rate": 1.9916267942583733e-05,
+      "loss": 1.6825,
+      "step": 514
+    },
+    {
+      "epoch": 1.2174940898345155,
+      "grad_norm": 3.5130748748779297,
+      "learning_rate": 1.9856459330143542e-05,
+      "loss": 1.6431,
+      "step": 515
+    },
+    {
+      "epoch": 1.2198581560283688,
+      "grad_norm": 3.419294595718384,
+      "learning_rate": 1.979665071770335e-05,
+      "loss": 1.6686,
+      "step": 516
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 3.5099575519561768,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 1.6568,
+      "step": 517
+    },
+    {
+      "epoch": 1.2245862884160756,
+      "grad_norm": 3.4354982376098633,
+      "learning_rate": 1.9677033492822968e-05,
+      "loss": 1.6038,
+      "step": 518
+    },
+    {
+      "epoch": 1.226950354609929,
+      "grad_norm": 3.4908761978149414,
+      "learning_rate": 1.9617224880382777e-05,
+      "loss": 1.6258,
+      "step": 519
+    },
+    {
+      "epoch": 1.2293144208037825,
+      "grad_norm": 3.4986984729766846,
+      "learning_rate": 1.9557416267942587e-05,
+      "loss": 1.6308,
+      "step": 520
+    },
+    {
+      "epoch": 1.231678486997636,
+      "grad_norm": 3.4746501445770264,
+      "learning_rate": 1.9497607655502393e-05,
+      "loss": 1.5862,
+      "step": 521
+    },
+    {
+      "epoch": 1.2340425531914894,
+      "grad_norm": 3.19508957862854,
+      "learning_rate": 1.9437799043062202e-05,
+      "loss": 1.6315,
+      "step": 522
+    },
+    {
+      "epoch": 1.2364066193853427,
+      "grad_norm": 3.50010347366333,
+      "learning_rate": 1.9377990430622012e-05,
+      "loss": 1.6374,
+      "step": 523
+    },
+    {
+      "epoch": 1.2387706855791962,
+      "grad_norm": 3.509312391281128,
+      "learning_rate": 1.9318181818181818e-05,
+      "loss": 1.6422,
+      "step": 524
+    },
+    {
+      "epoch": 1.2411347517730495,
+      "grad_norm": 3.383615732192993,
+      "learning_rate": 1.9258373205741628e-05,
+      "loss": 1.6133,
+      "step": 525
+    },
+    {
+      "epoch": 1.243498817966903,
+      "grad_norm": 3.5364954471588135,
+      "learning_rate": 1.9198564593301434e-05,
+      "loss": 1.6137,
+      "step": 526
+    },
+    {
+      "epoch": 1.2458628841607564,
+      "grad_norm": 3.60488224029541,
+      "learning_rate": 1.9138755980861243e-05,
+      "loss": 1.6931,
+      "step": 527
+    },
+    {
+      "epoch": 1.24822695035461,
+      "grad_norm": 3.1032845973968506,
+      "learning_rate": 1.9078947368421056e-05,
+      "loss": 1.6491,
+      "step": 528
+    },
+    {
+      "epoch": 1.2505910165484633,
+      "grad_norm": 3.539658784866333,
+      "learning_rate": 1.9019138755980862e-05,
+      "loss": 1.5981,
+      "step": 529
+    },
+    {
+      "epoch": 1.2529550827423168,
+      "grad_norm": 3.2954676151275635,
+      "learning_rate": 1.8959330143540672e-05,
+      "loss": 1.702,
+      "step": 530
+    },
+    {
+      "epoch": 1.2553191489361701,
+      "grad_norm": 3.566760540008545,
+      "learning_rate": 1.8899521531100478e-05,
+      "loss": 1.7553,
+      "step": 531
+    },
+    {
+      "epoch": 1.2576832151300237,
+      "grad_norm": 3.6372859477996826,
+      "learning_rate": 1.8839712918660287e-05,
+      "loss": 1.6513,
+      "step": 532
+    },
+    {
+      "epoch": 1.260047281323877,
+      "grad_norm": 3.449582576751709,
+      "learning_rate": 1.8779904306220097e-05,
+      "loss": 1.6247,
+      "step": 533
+    },
+    {
+      "epoch": 1.2624113475177305,
+      "grad_norm": 3.4251697063446045,
+      "learning_rate": 1.8720095693779903e-05,
+      "loss": 1.6476,
+      "step": 534
+    },
+    {
+      "epoch": 1.2647754137115839,
+      "grad_norm": 3.401045799255371,
+      "learning_rate": 1.8660287081339713e-05,
+      "loss": 1.6616,
+      "step": 535
+    },
+    {
+      "epoch": 1.2671394799054374,
+      "grad_norm": 3.2664437294006348,
+      "learning_rate": 1.8600478468899522e-05,
+      "loss": 1.5472,
+      "step": 536
+    },
+    {
+      "epoch": 1.2695035460992907,
+      "grad_norm": 3.605675458908081,
+      "learning_rate": 1.8540669856459332e-05,
+      "loss": 1.673,
+      "step": 537
+    },
+    {
+      "epoch": 1.2718676122931443,
+      "grad_norm": 3.743053674697876,
+      "learning_rate": 1.848086124401914e-05,
+      "loss": 1.7022,
+      "step": 538
+    },
+    {
+      "epoch": 1.2742316784869976,
+      "grad_norm": 3.6139564514160156,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 1.7027,
+      "step": 539
+    },
+    {
+      "epoch": 1.2765957446808511,
+      "grad_norm": 3.537170648574829,
+      "learning_rate": 1.8361244019138757e-05,
+      "loss": 1.6479,
+      "step": 540
+    },
+    {
+      "epoch": 1.2789598108747045,
+      "grad_norm": 3.652331590652466,
+      "learning_rate": 1.8301435406698566e-05,
+      "loss": 1.6372,
+      "step": 541
+    },
+    {
+      "epoch": 1.281323877068558,
+      "grad_norm": 3.5538816452026367,
+      "learning_rate": 1.8241626794258373e-05,
+      "loss": 1.6375,
+      "step": 542
+    },
+    {
+      "epoch": 1.2836879432624113,
+      "grad_norm": 3.6458072662353516,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.6305,
+      "step": 543
+    },
+    {
+      "epoch": 1.2860520094562649,
+      "grad_norm": 3.471625328063965,
+      "learning_rate": 1.812200956937799e-05,
+      "loss": 1.5957,
+      "step": 544
+    },
+    {
+      "epoch": 1.2884160756501182,
+      "grad_norm": 3.323765516281128,
+      "learning_rate": 1.80622009569378e-05,
+      "loss": 1.5751,
+      "step": 545
+    },
+    {
+      "epoch": 1.2907801418439715,
+      "grad_norm": 3.2532002925872803,
+      "learning_rate": 1.800239234449761e-05,
+      "loss": 1.5689,
+      "step": 546
+    },
+    {
+      "epoch": 1.293144208037825,
+      "grad_norm": 3.2593157291412354,
+      "learning_rate": 1.7942583732057417e-05,
+      "loss": 1.5601,
+      "step": 547
+    },
+    {
+      "epoch": 1.2955082742316786,
+      "grad_norm": 3.3987958431243896,
+      "learning_rate": 1.7882775119617226e-05,
+      "loss": 1.6927,
+      "step": 548
+    },
+    {
+      "epoch": 1.297872340425532,
+      "grad_norm": 3.6468160152435303,
+      "learning_rate": 1.7822966507177032e-05,
+      "loss": 1.6122,
+      "step": 549
+    },
+    {
+      "epoch": 1.3002364066193852,
+      "grad_norm": 3.36958646774292,
+      "learning_rate": 1.7763157894736842e-05,
+      "loss": 1.702,
+      "step": 550
+    },
+    {
+      "epoch": 1.3026004728132388,
+      "grad_norm": 3.3941123485565186,
+      "learning_rate": 1.770334928229665e-05,
+      "loss": 1.6855,
+      "step": 551
+    },
+    {
+      "epoch": 1.3049645390070923,
+      "grad_norm": 3.241328001022339,
+      "learning_rate": 1.7643540669856458e-05,
+      "loss": 1.6427,
+      "step": 552
+    },
+    {
+      "epoch": 1.3073286052009456,
+      "grad_norm": 3.470787763595581,
+      "learning_rate": 1.758373205741627e-05,
+      "loss": 1.6664,
+      "step": 553
+    },
+    {
+      "epoch": 1.309692671394799,
+      "grad_norm": 3.4991133213043213,
+      "learning_rate": 1.752392344497608e-05,
+      "loss": 1.6903,
+      "step": 554
+    },
+    {
+      "epoch": 1.3120567375886525,
+      "grad_norm": 3.3299062252044678,
+      "learning_rate": 1.7464114832535886e-05,
+      "loss": 1.6424,
+      "step": 555
+    },
+    {
+      "epoch": 1.314420803782506,
+      "grad_norm": 3.72178316116333,
+      "learning_rate": 1.7404306220095696e-05,
+      "loss": 1.6782,
+      "step": 556
+    },
+    {
+      "epoch": 1.3167848699763594,
+      "grad_norm": 3.3736157417297363,
+      "learning_rate": 1.7344497607655502e-05,
+      "loss": 1.6784,
+      "step": 557
+    },
+    {
+      "epoch": 1.3191489361702127,
+      "grad_norm": 3.2792179584503174,
+      "learning_rate": 1.728468899521531e-05,
+      "loss": 1.6388,
+      "step": 558
+    },
+    {
+      "epoch": 1.3215130023640662,
+      "grad_norm": 3.2111778259277344,
+      "learning_rate": 1.722488038277512e-05,
+      "loss": 1.7016,
+      "step": 559
+    },
+    {
+      "epoch": 1.3238770685579198,
+      "grad_norm": 3.1302785873413086,
+      "learning_rate": 1.7165071770334927e-05,
+      "loss": 1.6088,
+      "step": 560
+    },
+    {
+      "epoch": 1.326241134751773,
+      "grad_norm": 3.296743869781494,
+      "learning_rate": 1.7105263157894737e-05,
+      "loss": 1.6173,
+      "step": 561
+    },
+    {
+      "epoch": 1.3286052009456264,
+      "grad_norm": 3.36970853805542,
+      "learning_rate": 1.7045454545454546e-05,
+      "loss": 1.6424,
+      "step": 562
+    },
+    {
+      "epoch": 1.33096926713948,
+      "grad_norm": 3.7065136432647705,
+      "learning_rate": 1.6985645933014356e-05,
+      "loss": 1.6204,
+      "step": 563
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 3.59222412109375,
+      "learning_rate": 1.6925837320574165e-05,
+      "loss": 1.6838,
+      "step": 564
+    },
+    {
+      "epoch": 1.3356973995271868,
+      "grad_norm": 3.18104887008667,
+      "learning_rate": 1.686602870813397e-05,
+      "loss": 1.6503,
+      "step": 565
+    },
+    {
+      "epoch": 1.3380614657210401,
+      "grad_norm": 3.9115757942199707,
+      "learning_rate": 1.680622009569378e-05,
+      "loss": 1.6485,
+      "step": 566
+    },
+    {
+      "epoch": 1.3404255319148937,
+      "grad_norm": 3.2171802520751953,
+      "learning_rate": 1.674641148325359e-05,
+      "loss": 1.6396,
+      "step": 567
+    },
+    {
+      "epoch": 1.342789598108747,
+      "grad_norm": 3.288642406463623,
+      "learning_rate": 1.6686602870813396e-05,
+      "loss": 1.6966,
+      "step": 568
+    },
+    {
+      "epoch": 1.3451536643026005,
+      "grad_norm": 3.4783761501312256,
+      "learning_rate": 1.6626794258373206e-05,
+      "loss": 1.6339,
+      "step": 569
+    },
+    {
+      "epoch": 1.3475177304964538,
+      "grad_norm": 3.4193568229675293,
+      "learning_rate": 1.6566985645933016e-05,
+      "loss": 1.666,
+      "step": 570
+    },
+    {
+      "epoch": 1.3498817966903074,
+      "grad_norm": 3.327106237411499,
+      "learning_rate": 1.6507177033492825e-05,
+      "loss": 1.6518,
+      "step": 571
+    },
+    {
+      "epoch": 1.3522458628841607,
+      "grad_norm": 3.579259157180786,
+      "learning_rate": 1.6447368421052635e-05,
+      "loss": 1.6352,
+      "step": 572
+    },
+    {
+      "epoch": 1.3546099290780143,
+      "grad_norm": 3.3811655044555664,
+      "learning_rate": 1.638755980861244e-05,
+      "loss": 1.6311,
+      "step": 573
+    },
+    {
+      "epoch": 1.3569739952718676,
+      "grad_norm": 3.1911275386810303,
+      "learning_rate": 1.632775119617225e-05,
+      "loss": 1.6124,
+      "step": 574
+    },
+    {
+      "epoch": 1.3593380614657211,
+      "grad_norm": 3.3752293586730957,
+      "learning_rate": 1.6267942583732056e-05,
+      "loss": 1.6206,
+      "step": 575
+    },
+    {
+      "epoch": 1.3617021276595744,
+      "grad_norm": 3.6627418994903564,
+      "learning_rate": 1.6208133971291866e-05,
+      "loss": 1.568,
+      "step": 576
+    },
+    {
+      "epoch": 1.364066193853428,
+      "grad_norm": 3.476933717727661,
+      "learning_rate": 1.6148325358851675e-05,
+      "loss": 1.6019,
+      "step": 577
+    },
+    {
+      "epoch": 1.3664302600472813,
+      "grad_norm": 3.566840648651123,
+      "learning_rate": 1.6088516746411485e-05,
+      "loss": 1.6348,
+      "step": 578
+    },
+    {
+      "epoch": 1.3687943262411348,
+      "grad_norm": 3.4907350540161133,
+      "learning_rate": 1.6028708133971294e-05,
+      "loss": 1.6803,
+      "step": 579
+    },
+    {
+      "epoch": 1.3711583924349882,
+      "grad_norm": 3.323463201522827,
+      "learning_rate": 1.59688995215311e-05,
+      "loss": 1.6502,
+      "step": 580
+    },
+    {
+      "epoch": 1.3735224586288415,
+      "grad_norm": 3.230318069458008,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 1.5683,
+      "step": 581
+    },
+    {
+      "epoch": 1.375886524822695,
+      "grad_norm": 3.3330845832824707,
+      "learning_rate": 1.584928229665072e-05,
+      "loss": 1.6215,
+      "step": 582
+    },
+    {
+      "epoch": 1.3782505910165486,
+      "grad_norm": 3.4185128211975098,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 1.6787,
+      "step": 583
+    },
+    {
+      "epoch": 1.3806146572104019,
+      "grad_norm": 3.7025232315063477,
+      "learning_rate": 1.5729665071770335e-05,
+      "loss": 1.6466,
+      "step": 584
+    },
+    {
+      "epoch": 1.3829787234042552,
+      "grad_norm": 3.770089626312256,
+      "learning_rate": 1.5669856459330145e-05,
+      "loss": 1.673,
+      "step": 585
+    },
+    {
+      "epoch": 1.3853427895981087,
+      "grad_norm": 3.472599506378174,
+      "learning_rate": 1.561004784688995e-05,
+      "loss": 1.6181,
+      "step": 586
+    },
+    {
+      "epoch": 1.3877068557919623,
+      "grad_norm": 3.397357702255249,
+      "learning_rate": 1.555023923444976e-05,
+      "loss": 1.6728,
+      "step": 587
+    },
+    {
+      "epoch": 1.3900709219858156,
+      "grad_norm": 3.5820982456207275,
+      "learning_rate": 1.549043062200957e-05,
+      "loss": 1.6025,
+      "step": 588
+    },
+    {
+      "epoch": 1.392434988179669,
+      "grad_norm": 3.512296438217163,
+      "learning_rate": 1.543062200956938e-05,
+      "loss": 1.6413,
+      "step": 589
+    },
+    {
+      "epoch": 1.3947990543735225,
+      "grad_norm": 3.4280846118927,
+      "learning_rate": 1.537081339712919e-05,
+      "loss": 1.6388,
+      "step": 590
+    },
+    {
+      "epoch": 1.397163120567376,
+      "grad_norm": 3.3687398433685303,
+      "learning_rate": 1.5311004784688995e-05,
+      "loss": 1.651,
+      "step": 591
+    },
+    {
+      "epoch": 1.3995271867612293,
+      "grad_norm": 3.170372486114502,
+      "learning_rate": 1.5251196172248805e-05,
+      "loss": 1.6449,
+      "step": 592
+    },
+    {
+      "epoch": 1.4018912529550827,
+      "grad_norm": 3.245079755783081,
+      "learning_rate": 1.5191387559808613e-05,
+      "loss": 1.6725,
+      "step": 593
+    },
+    {
+      "epoch": 1.4042553191489362,
+      "grad_norm": 3.2267603874206543,
+      "learning_rate": 1.5131578947368422e-05,
+      "loss": 1.6259,
+      "step": 594
+    },
+    {
+      "epoch": 1.4066193853427895,
+      "grad_norm": 3.304009199142456,
+      "learning_rate": 1.5071770334928232e-05,
+      "loss": 1.6362,
+      "step": 595
+    },
+    {
+      "epoch": 1.408983451536643,
+      "grad_norm": 3.2211010456085205,
+      "learning_rate": 1.5011961722488038e-05,
+      "loss": 1.6423,
+      "step": 596
+    },
+    {
+      "epoch": 1.4113475177304964,
+      "grad_norm": 3.5229361057281494,
+      "learning_rate": 1.4952153110047847e-05,
+      "loss": 1.604,
+      "step": 597
+    },
+    {
+      "epoch": 1.41371158392435,
+      "grad_norm": 3.3903708457946777,
+      "learning_rate": 1.4892344497607657e-05,
+      "loss": 1.54,
+      "step": 598
+    },
+    {
+      "epoch": 1.4160756501182032,
+      "grad_norm": 3.6783154010772705,
+      "learning_rate": 1.4832535885167465e-05,
+      "loss": 1.6891,
+      "step": 599
+    },
+    {
+      "epoch": 1.4184397163120568,
+      "grad_norm": 3.3614747524261475,
+      "learning_rate": 1.4772727272727274e-05,
+      "loss": 1.5927,
+      "step": 600
+    },
+    {
+      "epoch": 1.42080378250591,
+      "grad_norm": 3.4144439697265625,
+      "learning_rate": 1.471291866028708e-05,
+      "loss": 1.5759,
+      "step": 601
+    },
+    {
+      "epoch": 1.4231678486997636,
+      "grad_norm": 3.5212013721466064,
+      "learning_rate": 1.4653110047846892e-05,
+      "loss": 1.5951,
+      "step": 602
+    },
+    {
+      "epoch": 1.425531914893617,
+      "grad_norm": 3.3338160514831543,
+      "learning_rate": 1.4593301435406701e-05,
+      "loss": 1.5867,
+      "step": 603
+    },
+    {
+      "epoch": 1.4278959810874705,
+      "grad_norm": 3.3701469898223877,
+      "learning_rate": 1.4533492822966507e-05,
+      "loss": 1.6722,
+      "step": 604
+    },
+    {
+      "epoch": 1.4302600472813238,
+      "grad_norm": 3.4370622634887695,
+      "learning_rate": 1.4473684210526317e-05,
+      "loss": 1.6592,
+      "step": 605
+    },
+    {
+      "epoch": 1.4326241134751774,
+      "grad_norm": 3.3094122409820557,
+      "learning_rate": 1.4413875598086125e-05,
+      "loss": 1.6039,
+      "step": 606
+    },
+    {
+      "epoch": 1.4349881796690307,
+      "grad_norm": 3.21444034576416,
+      "learning_rate": 1.4354066985645934e-05,
+      "loss": 1.6889,
+      "step": 607
+    },
+    {
+      "epoch": 1.4373522458628842,
+      "grad_norm": 3.4152231216430664,
+      "learning_rate": 1.4294258373205744e-05,
+      "loss": 1.6215,
+      "step": 608
+    },
+    {
+      "epoch": 1.4397163120567376,
+      "grad_norm": 3.1800973415374756,
+      "learning_rate": 1.423444976076555e-05,
+      "loss": 1.5442,
+      "step": 609
+    },
+    {
+      "epoch": 1.442080378250591,
+      "grad_norm": 3.6048243045806885,
+      "learning_rate": 1.417464114832536e-05,
+      "loss": 1.6962,
+      "step": 610
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 3.259493589401245,
+      "learning_rate": 1.4114832535885167e-05,
+      "loss": 1.5805,
+      "step": 611
+    },
+    {
+      "epoch": 1.4468085106382977,
+      "grad_norm": 3.352950096130371,
+      "learning_rate": 1.4055023923444977e-05,
+      "loss": 1.5834,
+      "step": 612
+    },
+    {
+      "epoch": 1.4491725768321513,
+      "grad_norm": 3.435502529144287,
+      "learning_rate": 1.3995215311004786e-05,
+      "loss": 1.6109,
+      "step": 613
+    },
+    {
+      "epoch": 1.4515366430260048,
+      "grad_norm": 3.175809621810913,
+      "learning_rate": 1.3935406698564592e-05,
+      "loss": 1.5312,
+      "step": 614
+    },
+    {
+      "epoch": 1.4539007092198581,
+      "grad_norm": 3.328200340270996,
+      "learning_rate": 1.3875598086124404e-05,
+      "loss": 1.5559,
+      "step": 615
+    },
+    {
+      "epoch": 1.4562647754137115,
+      "grad_norm": 3.227952480316162,
+      "learning_rate": 1.3815789473684213e-05,
+      "loss": 1.5322,
+      "step": 616
+    },
+    {
+      "epoch": 1.458628841607565,
+      "grad_norm": 3.272395610809326,
+      "learning_rate": 1.375598086124402e-05,
+      "loss": 1.6339,
+      "step": 617
+    },
+    {
+      "epoch": 1.4609929078014185,
+      "grad_norm": 3.417051315307617,
+      "learning_rate": 1.3696172248803829e-05,
+      "loss": 1.6676,
+      "step": 618
+    },
+    {
+      "epoch": 1.4633569739952719,
+      "grad_norm": 3.299630641937256,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.6343,
+      "step": 619
+    },
+    {
+      "epoch": 1.4657210401891252,
+      "grad_norm": 3.3762664794921875,
+      "learning_rate": 1.3576555023923446e-05,
+      "loss": 1.5982,
+      "step": 620
+    },
+    {
+      "epoch": 1.4680851063829787,
+      "grad_norm": 2.958192825317383,
+      "learning_rate": 1.3516746411483256e-05,
+      "loss": 1.598,
+      "step": 621
+    },
+    {
+      "epoch": 1.4704491725768323,
+      "grad_norm": 3.270052671432495,
+      "learning_rate": 1.3456937799043062e-05,
+      "loss": 1.7025,
+      "step": 622
+    },
+    {
+      "epoch": 1.4728132387706856,
+      "grad_norm": 3.4359071254730225,
+      "learning_rate": 1.3397129186602871e-05,
+      "loss": 1.6344,
+      "step": 623
+    },
+    {
+      "epoch": 1.475177304964539,
+      "grad_norm": 3.2563626766204834,
+      "learning_rate": 1.3337320574162679e-05,
+      "loss": 1.6154,
+      "step": 624
+    },
+    {
+      "epoch": 1.4775413711583925,
+      "grad_norm": 3.3198134899139404,
+      "learning_rate": 1.3277511961722489e-05,
+      "loss": 1.5953,
+      "step": 625
+    },
+    {
+      "epoch": 1.479905437352246,
+      "grad_norm": 3.1364660263061523,
+      "learning_rate": 1.3217703349282298e-05,
+      "loss": 1.572,
+      "step": 626
+    },
+    {
+      "epoch": 1.4822695035460993,
+      "grad_norm": 3.346433162689209,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 1.6064,
+      "step": 627
+    },
+    {
+      "epoch": 1.4846335697399526,
+      "grad_norm": 3.1672523021698,
+      "learning_rate": 1.3098086124401916e-05,
+      "loss": 1.6319,
+      "step": 628
+    },
+    {
+      "epoch": 1.4869976359338062,
+      "grad_norm": 3.061640977859497,
+      "learning_rate": 1.3038277511961722e-05,
+      "loss": 1.6271,
+      "step": 629
+    },
+    {
+      "epoch": 1.4893617021276595,
+      "grad_norm": 3.298517942428589,
+      "learning_rate": 1.2978468899521531e-05,
+      "loss": 1.5318,
+      "step": 630
+    },
+    {
+      "epoch": 1.491725768321513,
+      "grad_norm": 3.5346083641052246,
+      "learning_rate": 1.291866028708134e-05,
+      "loss": 1.6172,
+      "step": 631
+    },
+    {
+      "epoch": 1.4940898345153664,
+      "grad_norm": 3.271876335144043,
+      "learning_rate": 1.2858851674641149e-05,
+      "loss": 1.6122,
+      "step": 632
+    },
+    {
+      "epoch": 1.49645390070922,
+      "grad_norm": 3.443845748901367,
+      "learning_rate": 1.2799043062200958e-05,
+      "loss": 1.6147,
+      "step": 633
+    },
+    {
+      "epoch": 1.4988179669030732,
+      "grad_norm": 3.336913585662842,
+      "learning_rate": 1.2739234449760768e-05,
+      "loss": 1.5498,
+      "step": 634
+    },
+    {
+      "epoch": 1.5011820330969265,
+      "grad_norm": 3.355677366256714,
+      "learning_rate": 1.2679425837320574e-05,
+      "loss": 1.619,
+      "step": 635
+    },
+    {
+      "epoch": 1.50354609929078,
+      "grad_norm": 3.250521421432495,
+      "learning_rate": 1.2619617224880383e-05,
+      "loss": 1.5994,
+      "step": 636
+    },
+    {
+      "epoch": 1.5059101654846336,
+      "grad_norm": 3.4614198207855225,
+      "learning_rate": 1.2559808612440191e-05,
+      "loss": 1.6519,
+      "step": 637
+    },
+    {
+      "epoch": 1.508274231678487,
+      "grad_norm": 3.2225258350372314,
+      "learning_rate": 1.25e-05,
+      "loss": 1.678,
+      "step": 638
+    },
+    {
+      "epoch": 1.5106382978723403,
+      "grad_norm": 3.3898487091064453,
+      "learning_rate": 1.2440191387559808e-05,
+      "loss": 1.6185,
+      "step": 639
+    },
+    {
+      "epoch": 1.5130023640661938,
+      "grad_norm": 3.226099967956543,
+      "learning_rate": 1.2380382775119618e-05,
+      "loss": 1.5081,
+      "step": 640
+    },
+    {
+      "epoch": 1.5153664302600474,
+      "grad_norm": 3.1962661743164062,
+      "learning_rate": 1.2320574162679427e-05,
+      "loss": 1.645,
+      "step": 641
+    },
+    {
+      "epoch": 1.5177304964539007,
+      "grad_norm": 3.3228158950805664,
+      "learning_rate": 1.2260765550239235e-05,
+      "loss": 1.6331,
+      "step": 642
+    },
+    {
+      "epoch": 1.520094562647754,
+      "grad_norm": 3.6790809631347656,
+      "learning_rate": 1.2200956937799043e-05,
+      "loss": 1.5544,
+      "step": 643
+    },
+    {
+      "epoch": 1.5224586288416075,
+      "grad_norm": 3.375044822692871,
+      "learning_rate": 1.2141148325358851e-05,
+      "loss": 1.6076,
+      "step": 644
+    },
+    {
+      "epoch": 1.524822695035461,
+      "grad_norm": 3.324845790863037,
+      "learning_rate": 1.2081339712918662e-05,
+      "loss": 1.5041,
+      "step": 645
+    },
+    {
+      "epoch": 1.5271867612293144,
+      "grad_norm": 3.6021666526794434,
+      "learning_rate": 1.202153110047847e-05,
+      "loss": 1.6022,
+      "step": 646
+    },
+    {
+      "epoch": 1.5295508274231677,
+      "grad_norm": 3.412616729736328,
+      "learning_rate": 1.1961722488038278e-05,
+      "loss": 1.5649,
+      "step": 647
+    },
+    {
+      "epoch": 1.5319148936170213,
+      "grad_norm": 3.1025540828704834,
+      "learning_rate": 1.1901913875598086e-05,
+      "loss": 1.5535,
+      "step": 648
+    },
+    {
+      "epoch": 1.5342789598108748,
+      "grad_norm": 3.1715688705444336,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 1.5612,
+      "step": 649
+    },
+    {
+      "epoch": 1.5366430260047281,
+      "grad_norm": 3.3327136039733887,
+      "learning_rate": 1.1782296650717705e-05,
+      "loss": 1.602,
+      "step": 650
+    },
+    {
+      "epoch": 1.5390070921985815,
+      "grad_norm": 3.278064727783203,
+      "learning_rate": 1.1722488038277513e-05,
+      "loss": 1.62,
+      "step": 651
+    },
+    {
+      "epoch": 1.541371158392435,
+      "grad_norm": 3.3567280769348145,
+      "learning_rate": 1.166267942583732e-05,
+      "loss": 1.6035,
+      "step": 652
+    },
+    {
+      "epoch": 1.5437352245862885,
+      "grad_norm": 3.2232353687286377,
+      "learning_rate": 1.160287081339713e-05,
+      "loss": 1.569,
+      "step": 653
+    },
+    {
+      "epoch": 1.5460992907801419,
+      "grad_norm": 3.1278183460235596,
+      "learning_rate": 1.154306220095694e-05,
+      "loss": 1.6563,
+      "step": 654
+    },
+    {
+      "epoch": 1.5484633569739952,
+      "grad_norm": 3.2195639610290527,
+      "learning_rate": 1.1483253588516747e-05,
+      "loss": 1.5537,
+      "step": 655
+    },
+    {
+      "epoch": 1.5508274231678487,
+      "grad_norm": 3.479424238204956,
+      "learning_rate": 1.1423444976076555e-05,
+      "loss": 1.6301,
+      "step": 656
+    },
+    {
+      "epoch": 1.5531914893617023,
+      "grad_norm": 3.478548765182495,
+      "learning_rate": 1.1363636363636365e-05,
+      "loss": 1.6581,
+      "step": 657
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 3.495012044906616,
+      "learning_rate": 1.1303827751196172e-05,
+      "loss": 1.6673,
+      "step": 658
+    },
+    {
+      "epoch": 1.557919621749409,
+      "grad_norm": 3.327958822250366,
+      "learning_rate": 1.1244019138755982e-05,
+      "loss": 1.6435,
+      "step": 659
+    },
+    {
+      "epoch": 1.5602836879432624,
+      "grad_norm": 3.364741325378418,
+      "learning_rate": 1.118421052631579e-05,
+      "loss": 1.6802,
+      "step": 660
+    },
+    {
+      "epoch": 1.562647754137116,
+      "grad_norm": 3.15824031829834,
+      "learning_rate": 1.1124401913875598e-05,
+      "loss": 1.6495,
+      "step": 661
+    },
+    {
+      "epoch": 1.5650118203309693,
+      "grad_norm": 3.103550910949707,
+      "learning_rate": 1.1064593301435407e-05,
+      "loss": 1.6271,
+      "step": 662
+    },
+    {
+      "epoch": 1.5673758865248226,
+      "grad_norm": 3.3356754779815674,
+      "learning_rate": 1.1004784688995217e-05,
+      "loss": 1.5932,
+      "step": 663
+    },
+    {
+      "epoch": 1.5697399527186762,
+      "grad_norm": 3.3036398887634277,
+      "learning_rate": 1.0944976076555025e-05,
+      "loss": 1.6617,
+      "step": 664
+    },
+    {
+      "epoch": 1.5721040189125297,
+      "grad_norm": 3.20703387260437,
+      "learning_rate": 1.0885167464114832e-05,
+      "loss": 1.6287,
+      "step": 665
+    },
+    {
+      "epoch": 1.574468085106383,
+      "grad_norm": 3.3921687602996826,
+      "learning_rate": 1.0825358851674642e-05,
+      "loss": 1.5584,
+      "step": 666
+    },
+    {
+      "epoch": 1.5768321513002364,
+      "grad_norm": 3.2371623516082764,
+      "learning_rate": 1.0765550239234451e-05,
+      "loss": 1.5664,
+      "step": 667
+    },
+    {
+      "epoch": 1.57919621749409,
+      "grad_norm": 3.1653621196746826,
+      "learning_rate": 1.070574162679426e-05,
+      "loss": 1.5695,
+      "step": 668
+    },
+    {
+      "epoch": 1.5815602836879432,
+      "grad_norm": 3.467496871948242,
+      "learning_rate": 1.0645933014354067e-05,
+      "loss": 1.5485,
+      "step": 669
+    },
+    {
+      "epoch": 1.5839243498817965,
+      "grad_norm": 3.112905263900757,
+      "learning_rate": 1.0586124401913877e-05,
+      "loss": 1.5365,
+      "step": 670
+    },
+    {
+      "epoch": 1.58628841607565,
+      "grad_norm": 3.1907830238342285,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 1.5456,
+      "step": 671
+    },
+    {
+      "epoch": 1.5886524822695036,
+      "grad_norm": 3.382817268371582,
+      "learning_rate": 1.0466507177033494e-05,
+      "loss": 1.6402,
+      "step": 672
+    },
+    {
+      "epoch": 1.591016548463357,
+      "grad_norm": 3.159687042236328,
+      "learning_rate": 1.0406698564593302e-05,
+      "loss": 1.5578,
+      "step": 673
+    },
+    {
+      "epoch": 1.5933806146572103,
+      "grad_norm": 3.244880199432373,
+      "learning_rate": 1.034688995215311e-05,
+      "loss": 1.6561,
+      "step": 674
+    },
+    {
+      "epoch": 1.5957446808510638,
+      "grad_norm": 2.9089481830596924,
+      "learning_rate": 1.028708133971292e-05,
+      "loss": 1.5732,
+      "step": 675
+    },
+    {
+      "epoch": 1.5981087470449173,
+      "grad_norm": 3.061993360519409,
+      "learning_rate": 1.0227272727272729e-05,
+      "loss": 1.5751,
+      "step": 676
+    },
+    {
+      "epoch": 1.6004728132387707,
+      "grad_norm": 3.3294806480407715,
+      "learning_rate": 1.0167464114832537e-05,
+      "loss": 1.6372,
+      "step": 677
+    },
+    {
+      "epoch": 1.602836879432624,
+      "grad_norm": 3.1197781562805176,
+      "learning_rate": 1.0107655502392344e-05,
+      "loss": 1.6208,
+      "step": 678
+    },
+    {
+      "epoch": 1.6052009456264775,
+      "grad_norm": 3.3565282821655273,
+      "learning_rate": 1.0047846889952154e-05,
+      "loss": 1.6708,
+      "step": 679
+    },
+    {
+      "epoch": 1.607565011820331,
+      "grad_norm": 3.3783833980560303,
+      "learning_rate": 9.988038277511962e-06,
+      "loss": 1.6521,
+      "step": 680
+    },
+    {
+      "epoch": 1.6099290780141844,
+      "grad_norm": 3.4118947982788086,
+      "learning_rate": 9.928229665071771e-06,
+      "loss": 1.5684,
+      "step": 681
+    },
+    {
+      "epoch": 1.6122931442080377,
+      "grad_norm": 3.194164752960205,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 1.5699,
+      "step": 682
+    },
+    {
+      "epoch": 1.6146572104018913,
+      "grad_norm": 3.446110725402832,
+      "learning_rate": 9.808612440191389e-06,
+      "loss": 1.5395,
+      "step": 683
+    },
+    {
+      "epoch": 1.6170212765957448,
+      "grad_norm": 3.4659085273742676,
+      "learning_rate": 9.748803827751196e-06,
+      "loss": 1.5732,
+      "step": 684
+    },
+    {
+      "epoch": 1.6193853427895981,
+      "grad_norm": 3.388087749481201,
+      "learning_rate": 9.688995215311006e-06,
+      "loss": 1.6181,
+      "step": 685
+    },
+    {
+      "epoch": 1.6217494089834514,
+      "grad_norm": 3.297330379486084,
+      "learning_rate": 9.629186602870814e-06,
+      "loss": 1.6206,
+      "step": 686
+    },
+    {
+      "epoch": 1.624113475177305,
+      "grad_norm": 3.3777146339416504,
+      "learning_rate": 9.569377990430622e-06,
+      "loss": 1.6021,
+      "step": 687
+    },
+    {
+      "epoch": 1.6264775413711585,
+      "grad_norm": 3.2018673419952393,
+      "learning_rate": 9.509569377990431e-06,
+      "loss": 1.5968,
+      "step": 688
+    },
+    {
+      "epoch": 1.6288416075650118,
+      "grad_norm": 3.416879892349243,
+      "learning_rate": 9.449760765550239e-06,
+      "loss": 1.6182,
+      "step": 689
+    },
+    {
+      "epoch": 1.6312056737588652,
+      "grad_norm": 3.445861339569092,
+      "learning_rate": 9.389952153110048e-06,
+      "loss": 1.7168,
+      "step": 690
+    },
+    {
+      "epoch": 1.6335697399527187,
+      "grad_norm": 3.3097856044769287,
+      "learning_rate": 9.330143540669856e-06,
+      "loss": 1.617,
+      "step": 691
+    },
+    {
+      "epoch": 1.6359338061465722,
+      "grad_norm": 3.3081119060516357,
+      "learning_rate": 9.270334928229666e-06,
+      "loss": 1.6162,
+      "step": 692
+    },
+    {
+      "epoch": 1.6382978723404256,
+      "grad_norm": 3.2553486824035645,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 1.5849,
+      "step": 693
+    },
+    {
+      "epoch": 1.6406619385342789,
+      "grad_norm": 3.4111063480377197,
+      "learning_rate": 9.150717703349283e-06,
+      "loss": 1.5933,
+      "step": 694
+    },
+    {
+      "epoch": 1.6430260047281324,
+      "grad_norm": 3.365490436553955,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.694,
+      "step": 695
+    },
+    {
+      "epoch": 1.645390070921986,
+      "grad_norm": 3.191035032272339,
+      "learning_rate": 9.0311004784689e-06,
+      "loss": 1.5617,
+      "step": 696
+    },
+    {
+      "epoch": 1.6477541371158393,
+      "grad_norm": 3.4338693618774414,
+      "learning_rate": 8.971291866028708e-06,
+      "loss": 1.6369,
+      "step": 697
+    },
+    {
+      "epoch": 1.6501182033096926,
+      "grad_norm": 3.3550338745117188,
+      "learning_rate": 8.911483253588516e-06,
+      "loss": 1.6562,
+      "step": 698
+    },
+    {
+      "epoch": 1.6524822695035462,
+      "grad_norm": 3.4408254623413086,
+      "learning_rate": 8.851674641148326e-06,
+      "loss": 1.5803,
+      "step": 699
+    },
+    {
+      "epoch": 1.6548463356973995,
+      "grad_norm": 3.332296371459961,
+      "learning_rate": 8.791866028708135e-06,
+      "loss": 1.6199,
+      "step": 700
+    },
+    {
+      "epoch": 1.6572104018912528,
+      "grad_norm": 3.1053788661956787,
+      "learning_rate": 8.732057416267943e-06,
+      "loss": 1.5932,
+      "step": 701
+    },
+    {
+      "epoch": 1.6595744680851063,
+      "grad_norm": 3.337559700012207,
+      "learning_rate": 8.672248803827751e-06,
+      "loss": 1.6057,
+      "step": 702
+    },
+    {
+      "epoch": 1.6619385342789599,
+      "grad_norm": 3.354663372039795,
+      "learning_rate": 8.61244019138756e-06,
+      "loss": 1.6222,
+      "step": 703
+    },
+    {
+      "epoch": 1.6643026004728132,
+      "grad_norm": 3.2696797847747803,
+      "learning_rate": 8.552631578947368e-06,
+      "loss": 1.6551,
+      "step": 704
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 3.4090421199798584,
+      "learning_rate": 8.492822966507178e-06,
+      "loss": 1.6064,
+      "step": 705
+    },
+    {
+      "epoch": 1.66903073286052,
+      "grad_norm": 3.2877285480499268,
+      "learning_rate": 8.433014354066986e-06,
+      "loss": 1.5979,
+      "step": 706
+    },
+    {
+      "epoch": 1.6713947990543736,
+      "grad_norm": 3.5218770503997803,
+      "learning_rate": 8.373205741626795e-06,
+      "loss": 1.6531,
+      "step": 707
+    },
+    {
+      "epoch": 1.673758865248227,
+      "grad_norm": 3.3162100315093994,
+      "learning_rate": 8.313397129186603e-06,
+      "loss": 1.6029,
+      "step": 708
+    },
+    {
+      "epoch": 1.6761229314420802,
+      "grad_norm": 3.282989978790283,
+      "learning_rate": 8.253588516746413e-06,
+      "loss": 1.6072,
+      "step": 709
+    },
+    {
+      "epoch": 1.6784869976359338,
+      "grad_norm": 3.0652577877044678,
+      "learning_rate": 8.19377990430622e-06,
+      "loss": 1.5495,
+      "step": 710
+    },
+    {
+      "epoch": 1.6808510638297873,
+      "grad_norm": 3.1217284202575684,
+      "learning_rate": 8.133971291866028e-06,
+      "loss": 1.5687,
+      "step": 711
+    },
+    {
+      "epoch": 1.6832151300236406,
+      "grad_norm": 3.3797202110290527,
+      "learning_rate": 8.074162679425838e-06,
+      "loss": 1.6669,
+      "step": 712
+    },
+    {
+      "epoch": 1.685579196217494,
+      "grad_norm": 3.2859535217285156,
+      "learning_rate": 8.014354066985647e-06,
+      "loss": 1.6696,
+      "step": 713
+    },
+    {
+      "epoch": 1.6879432624113475,
+      "grad_norm": 3.3486392498016357,
+      "learning_rate": 7.954545454545455e-06,
+      "loss": 1.5806,
+      "step": 714
+    },
+    {
+      "epoch": 1.690307328605201,
+      "grad_norm": 3.3833365440368652,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 1.5656,
+      "step": 715
+    },
+    {
+      "epoch": 1.6926713947990544,
+      "grad_norm": 3.4777348041534424,
+      "learning_rate": 7.834928229665072e-06,
+      "loss": 1.6621,
+      "step": 716
+    },
+    {
+      "epoch": 1.6950354609929077,
+      "grad_norm": 3.3407208919525146,
+      "learning_rate": 7.77511961722488e-06,
+      "loss": 1.6283,
+      "step": 717
+    },
+    {
+      "epoch": 1.6973995271867612,
+      "grad_norm": 3.4209749698638916,
+      "learning_rate": 7.71531100478469e-06,
+      "loss": 1.6341,
+      "step": 718
+    },
+    {
+      "epoch": 1.6997635933806148,
+      "grad_norm": 3.4124741554260254,
+      "learning_rate": 7.655502392344498e-06,
+      "loss": 1.5899,
+      "step": 719
+    },
+    {
+      "epoch": 1.702127659574468,
+      "grad_norm": 3.362297296524048,
+      "learning_rate": 7.595693779904306e-06,
+      "loss": 1.6455,
+      "step": 720
+    },
+    {
+      "epoch": 1.7044917257683214,
+      "grad_norm": 3.364933967590332,
+      "learning_rate": 7.535885167464116e-06,
+      "loss": 1.6634,
+      "step": 721
+    },
+    {
+      "epoch": 1.706855791962175,
+      "grad_norm": 3.102181911468506,
+      "learning_rate": 7.476076555023924e-06,
+      "loss": 1.5939,
+      "step": 722
+    },
+    {
+      "epoch": 1.7092198581560285,
+      "grad_norm": 3.2613308429718018,
+      "learning_rate": 7.416267942583732e-06,
+      "loss": 1.6737,
+      "step": 723
+    },
+    {
+      "epoch": 1.7115839243498818,
+      "grad_norm": 3.2267072200775146,
+      "learning_rate": 7.35645933014354e-06,
+      "loss": 1.6096,
+      "step": 724
+    },
+    {
+      "epoch": 1.7139479905437351,
+      "grad_norm": 3.3645620346069336,
+      "learning_rate": 7.2966507177033505e-06,
+      "loss": 1.5925,
+      "step": 725
+    },
+    {
+      "epoch": 1.7163120567375887,
+      "grad_norm": 3.501962184906006,
+      "learning_rate": 7.236842105263158e-06,
+      "loss": 1.6251,
+      "step": 726
+    },
+    {
+      "epoch": 1.7186761229314422,
+      "grad_norm": 3.2537877559661865,
+      "learning_rate": 7.177033492822967e-06,
+      "loss": 1.6472,
+      "step": 727
+    },
+    {
+      "epoch": 1.7210401891252955,
+      "grad_norm": 3.2607951164245605,
+      "learning_rate": 7.117224880382775e-06,
+      "loss": 1.5956,
+      "step": 728
+    },
+    {
+      "epoch": 1.7234042553191489,
+      "grad_norm": 3.262928009033203,
+      "learning_rate": 7.0574162679425836e-06,
+      "loss": 1.6301,
+      "step": 729
+    },
+    {
+      "epoch": 1.7257683215130024,
+      "grad_norm": 3.369541883468628,
+      "learning_rate": 6.997607655502393e-06,
+      "loss": 1.6375,
+      "step": 730
+    },
+    {
+      "epoch": 1.728132387706856,
+      "grad_norm": 3.094522714614868,
+      "learning_rate": 6.937799043062202e-06,
+      "loss": 1.6489,
+      "step": 731
+    },
+    {
+      "epoch": 1.7304964539007093,
+      "grad_norm": 3.576575517654419,
+      "learning_rate": 6.87799043062201e-06,
+      "loss": 1.5887,
+      "step": 732
+    },
+    {
+      "epoch": 1.7328605200945626,
+      "grad_norm": 3.258538246154785,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 1.605,
+      "step": 733
+    },
+    {
+      "epoch": 1.7352245862884161,
+      "grad_norm": 3.1668145656585693,
+      "learning_rate": 6.758373205741628e-06,
+      "loss": 1.5805,
+      "step": 734
+    },
+    {
+      "epoch": 1.7375886524822695,
+      "grad_norm": 3.51381778717041,
+      "learning_rate": 6.698564593301436e-06,
+      "loss": 1.5987,
+      "step": 735
+    },
+    {
+      "epoch": 1.7399527186761228,
+      "grad_norm": 3.3210718631744385,
+      "learning_rate": 6.638755980861244e-06,
+      "loss": 1.5751,
+      "step": 736
+    },
+    {
+      "epoch": 1.7423167848699763,
+      "grad_norm": 3.149791717529297,
+      "learning_rate": 6.578947368421053e-06,
+      "loss": 1.5559,
+      "step": 737
+    },
+    {
+      "epoch": 1.7446808510638299,
+      "grad_norm": 3.152360200881958,
+      "learning_rate": 6.519138755980861e-06,
+      "loss": 1.4452,
+      "step": 738
+    },
+    {
+      "epoch": 1.7470449172576832,
+      "grad_norm": 3.0398778915405273,
+      "learning_rate": 6.45933014354067e-06,
+      "loss": 1.5646,
+      "step": 739
+    },
+    {
+      "epoch": 1.7494089834515365,
+      "grad_norm": 3.1679794788360596,
+      "learning_rate": 6.399521531100479e-06,
+      "loss": 1.5783,
+      "step": 740
+    },
+    {
+      "epoch": 1.75177304964539,
+      "grad_norm": 3.173020839691162,
+      "learning_rate": 6.339712918660287e-06,
+      "loss": 1.549,
+      "step": 741
+    },
+    {
+      "epoch": 1.7541371158392436,
+      "grad_norm": 3.451342821121216,
+      "learning_rate": 6.2799043062200955e-06,
+      "loss": 1.6184,
+      "step": 742
+    },
+    {
+      "epoch": 1.756501182033097,
+      "grad_norm": 3.26377010345459,
+      "learning_rate": 6.220095693779904e-06,
+      "loss": 1.6053,
+      "step": 743
+    },
+    {
+      "epoch": 1.7588652482269502,
+      "grad_norm": 3.3404619693756104,
+      "learning_rate": 6.160287081339714e-06,
+      "loss": 1.585,
+      "step": 744
+    },
+    {
+      "epoch": 1.7612293144208038,
+      "grad_norm": 3.3841774463653564,
+      "learning_rate": 6.1004784688995216e-06,
+      "loss": 1.5859,
+      "step": 745
+    },
+    {
+      "epoch": 1.7635933806146573,
+      "grad_norm": 3.4458372592926025,
+      "learning_rate": 6.040669856459331e-06,
+      "loss": 1.5913,
+      "step": 746
+    },
+    {
+      "epoch": 1.7659574468085106,
+      "grad_norm": 3.323127508163452,
+      "learning_rate": 5.980861244019139e-06,
+      "loss": 1.5518,
+      "step": 747
+    },
+    {
+      "epoch": 1.768321513002364,
+      "grad_norm": 3.858598232269287,
+      "learning_rate": 5.921052631578948e-06,
+      "loss": 1.6864,
+      "step": 748
+    },
+    {
+      "epoch": 1.7706855791962175,
+      "grad_norm": 2.968590259552002,
+      "learning_rate": 5.861244019138756e-06,
+      "loss": 1.6128,
+      "step": 749
+    },
+    {
+      "epoch": 1.773049645390071,
+      "grad_norm": 3.0373711585998535,
+      "learning_rate": 5.801435406698565e-06,
+      "loss": 1.543,
+      "step": 750
+    },
+    {
+      "epoch": 1.7754137115839244,
+      "grad_norm": 3.524451494216919,
+      "learning_rate": 5.741626794258374e-06,
+      "loss": 1.5866,
+      "step": 751
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 3.0094456672668457,
+      "learning_rate": 5.681818181818182e-06,
+      "loss": 1.5617,
+      "step": 752
+    },
+    {
+      "epoch": 1.7801418439716312,
+      "grad_norm": 3.0851380825042725,
+      "learning_rate": 5.622009569377991e-06,
+      "loss": 1.4814,
+      "step": 753
+    },
+    {
+      "epoch": 1.7825059101654848,
+      "grad_norm": 3.3271324634552,
+      "learning_rate": 5.562200956937799e-06,
+      "loss": 1.6402,
+      "step": 754
+    },
+    {
+      "epoch": 1.784869976359338,
+      "grad_norm": 3.0717716217041016,
+      "learning_rate": 5.502392344497608e-06,
+      "loss": 1.6037,
+      "step": 755
+    },
+    {
+      "epoch": 1.7872340425531914,
+      "grad_norm": 3.099236249923706,
+      "learning_rate": 5.442583732057416e-06,
+      "loss": 1.5689,
+      "step": 756
+    },
+    {
+      "epoch": 1.789598108747045,
+      "grad_norm": 3.1861002445220947,
+      "learning_rate": 5.382775119617226e-06,
+      "loss": 1.5622,
+      "step": 757
+    },
+    {
+      "epoch": 1.7919621749408985,
+      "grad_norm": 3.010728359222412,
+      "learning_rate": 5.3229665071770335e-06,
+      "loss": 1.6116,
+      "step": 758
+    },
+    {
+      "epoch": 1.7943262411347518,
+      "grad_norm": 3.0583670139312744,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.5781,
+      "step": 759
+    },
+    {
+      "epoch": 1.7966903073286051,
+      "grad_norm": 3.046917676925659,
+      "learning_rate": 5.203349282296651e-06,
+      "loss": 1.4681,
+      "step": 760
+    },
+    {
+      "epoch": 1.7990543735224587,
+      "grad_norm": 3.2571654319763184,
+      "learning_rate": 5.14354066985646e-06,
+      "loss": 1.6099,
+      "step": 761
+    },
+    {
+      "epoch": 1.8014184397163122,
+      "grad_norm": 3.372274160385132,
+      "learning_rate": 5.083732057416268e-06,
+      "loss": 1.5611,
+      "step": 762
+    },
+    {
+      "epoch": 1.8037825059101655,
+      "grad_norm": 3.326112747192383,
+      "learning_rate": 5.023923444976077e-06,
+      "loss": 1.653,
+      "step": 763
+    },
+    {
+      "epoch": 1.8061465721040189,
+      "grad_norm": 3.1222290992736816,
+      "learning_rate": 4.964114832535886e-06,
+      "loss": 1.5228,
+      "step": 764
+    },
+    {
+      "epoch": 1.8085106382978724,
+      "grad_norm": 3.0824484825134277,
+      "learning_rate": 4.904306220095694e-06,
+      "loss": 1.6368,
+      "step": 765
+    },
+    {
+      "epoch": 1.8108747044917257,
+      "grad_norm": 3.1424765586853027,
+      "learning_rate": 4.844497607655503e-06,
+      "loss": 1.5466,
+      "step": 766
+    },
+    {
+      "epoch": 1.813238770685579,
+      "grad_norm": 3.1939899921417236,
+      "learning_rate": 4.784688995215311e-06,
+      "loss": 1.6732,
+      "step": 767
+    },
+    {
+      "epoch": 1.8156028368794326,
+      "grad_norm": 3.301724672317505,
+      "learning_rate": 4.7248803827751195e-06,
+      "loss": 1.6021,
+      "step": 768
+    },
+    {
+      "epoch": 1.8179669030732861,
+      "grad_norm": 3.067897319793701,
+      "learning_rate": 4.665071770334928e-06,
+      "loss": 1.6226,
+      "step": 769
+    },
+    {
+      "epoch": 1.8203309692671394,
+      "grad_norm": 3.0447728633880615,
+      "learning_rate": 4.605263157894737e-06,
+      "loss": 1.6111,
+      "step": 770
+    },
+    {
+      "epoch": 1.8226950354609928,
+      "grad_norm": 3.2400856018066406,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.5478,
+      "step": 771
+    },
+    {
+      "epoch": 1.8250591016548463,
+      "grad_norm": 3.619302749633789,
+      "learning_rate": 4.485645933014354e-06,
+      "loss": 1.5489,
+      "step": 772
+    },
+    {
+      "epoch": 1.8274231678486998,
+      "grad_norm": 3.1459033489227295,
+      "learning_rate": 4.425837320574163e-06,
+      "loss": 1.6253,
+      "step": 773
+    },
+    {
+      "epoch": 1.8297872340425532,
+      "grad_norm": 3.0557096004486084,
+      "learning_rate": 4.3660287081339716e-06,
+      "loss": 1.6382,
+      "step": 774
+    },
+    {
+      "epoch": 1.8321513002364065,
+      "grad_norm": 3.4371113777160645,
+      "learning_rate": 4.30622009569378e-06,
+      "loss": 1.6007,
+      "step": 775
+    },
+    {
+      "epoch": 1.83451536643026,
+      "grad_norm": 3.019963026046753,
+      "learning_rate": 4.246411483253589e-06,
+      "loss": 1.5758,
+      "step": 776
+    },
+    {
+      "epoch": 1.8368794326241136,
+      "grad_norm": 3.315235137939453,
+      "learning_rate": 4.186602870813398e-06,
+      "loss": 1.6083,
+      "step": 777
+    },
+    {
+      "epoch": 1.839243498817967,
+      "grad_norm": 2.9758803844451904,
+      "learning_rate": 4.126794258373206e-06,
+      "loss": 1.5595,
+      "step": 778
+    },
+    {
+      "epoch": 1.8416075650118202,
+      "grad_norm": 3.254204273223877,
+      "learning_rate": 4.066985645933014e-06,
+      "loss": 1.5773,
+      "step": 779
+    },
+    {
+      "epoch": 1.8439716312056738,
+      "grad_norm": 3.199317216873169,
+      "learning_rate": 4.007177033492824e-06,
+      "loss": 1.53,
+      "step": 780
+    },
+    {
+      "epoch": 1.8463356973995273,
+      "grad_norm": 2.9678032398223877,
+      "learning_rate": 3.9473684210526315e-06,
+      "loss": 1.5161,
+      "step": 781
+    },
+    {
+      "epoch": 1.8486997635933806,
+      "grad_norm": 3.191781759262085,
+      "learning_rate": 3.88755980861244e-06,
+      "loss": 1.659,
+      "step": 782
+    },
+    {
+      "epoch": 1.851063829787234,
+      "grad_norm": 3.0926475524902344,
+      "learning_rate": 3.827751196172249e-06,
+      "loss": 1.5918,
+      "step": 783
+    },
+    {
+      "epoch": 1.8534278959810875,
+      "grad_norm": 3.0538089275360107,
+      "learning_rate": 3.767942583732058e-06,
+      "loss": 1.5246,
+      "step": 784
+    },
+    {
+      "epoch": 1.855791962174941,
+      "grad_norm": 3.3812143802642822,
+      "learning_rate": 3.708133971291866e-06,
+      "loss": 1.6095,
+      "step": 785
+    },
+    {
+      "epoch": 1.8581560283687943,
+      "grad_norm": 3.296877861022949,
+      "learning_rate": 3.6483253588516753e-06,
+      "loss": 1.6096,
+      "step": 786
+    },
+    {
+      "epoch": 1.8605200945626477,
+      "grad_norm": 3.0140068531036377,
+      "learning_rate": 3.5885167464114835e-06,
+      "loss": 1.5117,
+      "step": 787
+    },
+    {
+      "epoch": 1.8628841607565012,
+      "grad_norm": 3.225166082382202,
+      "learning_rate": 3.5287081339712918e-06,
+      "loss": 1.5789,
+      "step": 788
+    },
+    {
+      "epoch": 1.8652482269503547,
+      "grad_norm": 3.1967225074768066,
+      "learning_rate": 3.468899521531101e-06,
+      "loss": 1.5403,
+      "step": 789
+    },
+    {
+      "epoch": 1.867612293144208,
+      "grad_norm": 3.2307610511779785,
+      "learning_rate": 3.409090909090909e-06,
+      "loss": 1.5959,
+      "step": 790
+    },
+    {
+      "epoch": 1.8699763593380614,
+      "grad_norm": 3.252063274383545,
+      "learning_rate": 3.349282296650718e-06,
+      "loss": 1.6492,
+      "step": 791
+    },
+    {
+      "epoch": 1.872340425531915,
+      "grad_norm": 3.333066940307617,
+      "learning_rate": 3.2894736842105265e-06,
+      "loss": 1.6399,
+      "step": 792
+    },
+    {
+      "epoch": 1.8747044917257685,
+      "grad_norm": 3.0975208282470703,
+      "learning_rate": 3.229665071770335e-06,
+      "loss": 1.544,
+      "step": 793
+    },
+    {
+      "epoch": 1.8770685579196218,
+      "grad_norm": 3.1042416095733643,
+      "learning_rate": 3.1698564593301434e-06,
+      "loss": 1.5577,
+      "step": 794
+    },
+    {
+      "epoch": 1.8794326241134751,
+      "grad_norm": 3.2555060386657715,
+      "learning_rate": 3.110047846889952e-06,
+      "loss": 1.655,
+      "step": 795
+    },
+    {
+      "epoch": 1.8817966903073287,
+      "grad_norm": 3.113212823867798,
+      "learning_rate": 3.0502392344497608e-06,
+      "loss": 1.6079,
+      "step": 796
+    },
+    {
+      "epoch": 1.8841607565011822,
+      "grad_norm": 3.1746702194213867,
+      "learning_rate": 2.9904306220095695e-06,
+      "loss": 1.5433,
+      "step": 797
+    },
+    {
+      "epoch": 1.8865248226950353,
+      "grad_norm": 3.156456708908081,
+      "learning_rate": 2.930622009569378e-06,
+      "loss": 1.5069,
+      "step": 798
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 3.4995877742767334,
+      "learning_rate": 2.870813397129187e-06,
+      "loss": 1.557,
+      "step": 799
+    },
+    {
+      "epoch": 1.8912529550827424,
+      "grad_norm": 3.097641706466675,
+      "learning_rate": 2.8110047846889955e-06,
+      "loss": 1.5456,
+      "step": 800
+    },
+    {
+      "epoch": 1.8936170212765957,
+      "grad_norm": 3.0806541442871094,
+      "learning_rate": 2.751196172248804e-06,
+      "loss": 1.5219,
+      "step": 801
+    },
+    {
+      "epoch": 1.895981087470449,
+      "grad_norm": 3.0909721851348877,
+      "learning_rate": 2.691387559808613e-06,
+      "loss": 1.6775,
+      "step": 802
+    },
+    {
+      "epoch": 1.8983451536643026,
+      "grad_norm": 3.172172784805298,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 1.58,
+      "step": 803
+    },
+    {
+      "epoch": 1.900709219858156,
+      "grad_norm": 3.4585001468658447,
+      "learning_rate": 2.57177033492823e-06,
+      "loss": 1.6051,
+      "step": 804
+    },
+    {
+      "epoch": 1.9030732860520094,
+      "grad_norm": 3.037222385406494,
+      "learning_rate": 2.5119617224880385e-06,
+      "loss": 1.6135,
+      "step": 805
+    },
+    {
+      "epoch": 1.9054373522458627,
+      "grad_norm": 3.203216791152954,
+      "learning_rate": 2.452153110047847e-06,
+      "loss": 1.5594,
+      "step": 806
+    },
+    {
+      "epoch": 1.9078014184397163,
+      "grad_norm": 3.219123363494873,
+      "learning_rate": 2.3923444976076554e-06,
+      "loss": 1.6568,
+      "step": 807
+    },
+    {
+      "epoch": 1.9101654846335698,
+      "grad_norm": 3.3200933933258057,
+      "learning_rate": 2.332535885167464e-06,
+      "loss": 1.5685,
+      "step": 808
+    },
+    {
+      "epoch": 1.9125295508274232,
+      "grad_norm": 3.247377395629883,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.616,
+      "step": 809
+    },
+    {
+      "epoch": 1.9148936170212765,
+      "grad_norm": 3.124732255935669,
+      "learning_rate": 2.2129186602870814e-06,
+      "loss": 1.599,
+      "step": 810
+    },
+    {
+      "epoch": 1.91725768321513,
+      "grad_norm": 3.5280542373657227,
+      "learning_rate": 2.15311004784689e-06,
+      "loss": 1.6445,
+      "step": 811
+    },
+    {
+      "epoch": 1.9196217494089836,
+      "grad_norm": 3.254999876022339,
+      "learning_rate": 2.093301435406699e-06,
+      "loss": 1.6179,
+      "step": 812
+    },
+    {
+      "epoch": 1.9219858156028369,
+      "grad_norm": 2.8884024620056152,
+      "learning_rate": 2.033492822966507e-06,
+      "loss": 1.5094,
+      "step": 813
+    },
+    {
+      "epoch": 1.9243498817966902,
+      "grad_norm": 3.245943546295166,
+      "learning_rate": 1.9736842105263157e-06,
+      "loss": 1.6167,
+      "step": 814
+    },
+    {
+      "epoch": 1.9267139479905437,
+      "grad_norm": 3.1370460987091064,
+      "learning_rate": 1.9138755980861244e-06,
+      "loss": 1.5627,
+      "step": 815
+    },
+    {
+      "epoch": 1.9290780141843973,
+      "grad_norm": 3.339740753173828,
+      "learning_rate": 1.854066985645933e-06,
+      "loss": 1.6373,
+      "step": 816
+    },
+    {
+      "epoch": 1.9314420803782506,
+      "grad_norm": 3.221043348312378,
+      "learning_rate": 1.7942583732057418e-06,
+      "loss": 1.6013,
+      "step": 817
+    },
+    {
+      "epoch": 1.933806146572104,
+      "grad_norm": 3.43192458152771,
+      "learning_rate": 1.7344497607655504e-06,
+      "loss": 1.638,
+      "step": 818
+    },
+    {
+      "epoch": 1.9361702127659575,
+      "grad_norm": 3.0341379642486572,
+      "learning_rate": 1.674641148325359e-06,
+      "loss": 1.5758,
+      "step": 819
+    },
+    {
+      "epoch": 1.938534278959811,
+      "grad_norm": 3.183203935623169,
+      "learning_rate": 1.6148325358851676e-06,
+      "loss": 1.5632,
+      "step": 820
+    },
+    {
+      "epoch": 1.9408983451536643,
+      "grad_norm": 3.2515642642974854,
+      "learning_rate": 1.555023923444976e-06,
+      "loss": 1.5571,
+      "step": 821
+    },
+    {
+      "epoch": 1.9432624113475176,
+      "grad_norm": 3.4239914417266846,
+      "learning_rate": 1.4952153110047847e-06,
+      "loss": 1.6051,
+      "step": 822
+    },
+    {
+      "epoch": 1.9456264775413712,
+      "grad_norm": 3.138653039932251,
+      "learning_rate": 1.4354066985645934e-06,
+      "loss": 1.5776,
+      "step": 823
+    },
+    {
+      "epoch": 1.9479905437352247,
+      "grad_norm": 3.192990779876709,
+      "learning_rate": 1.375598086124402e-06,
+      "loss": 1.6117,
+      "step": 824
+    },
+    {
+      "epoch": 1.950354609929078,
+      "grad_norm": 3.2178995609283447,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 1.6727,
+      "step": 825
+    },
+    {
+      "epoch": 1.9527186761229314,
+      "grad_norm": 3.2581796646118164,
+      "learning_rate": 1.2559808612440192e-06,
+      "loss": 1.6165,
+      "step": 826
+    },
+    {
+      "epoch": 1.955082742316785,
+      "grad_norm": 3.419954538345337,
+      "learning_rate": 1.1961722488038277e-06,
+      "loss": 1.581,
+      "step": 827
+    },
+    {
+      "epoch": 1.9574468085106385,
+      "grad_norm": 3.3294622898101807,
+      "learning_rate": 1.1363636363636364e-06,
+      "loss": 1.6156,
+      "step": 828
+    },
+    {
+      "epoch": 1.9598108747044918,
+      "grad_norm": 3.2191033363342285,
+      "learning_rate": 1.076555023923445e-06,
+      "loss": 1.5636,
+      "step": 829
+    },
+    {
+      "epoch": 1.962174940898345,
+      "grad_norm": 3.038212537765503,
+      "learning_rate": 1.0167464114832535e-06,
+      "loss": 1.5989,
+      "step": 830
+    },
+    {
+      "epoch": 1.9645390070921986,
+      "grad_norm": 3.1238808631896973,
+      "learning_rate": 9.569377990430622e-07,
+      "loss": 1.6156,
+      "step": 831
+    },
+    {
+      "epoch": 1.966903073286052,
+      "grad_norm": 3.3783342838287354,
+      "learning_rate": 8.971291866028709e-07,
+      "loss": 1.5802,
+      "step": 832
+    },
+    {
+      "epoch": 1.9692671394799053,
+      "grad_norm": 3.407663106918335,
+      "learning_rate": 8.373205741626795e-07,
+      "loss": 1.6369,
+      "step": 833
+    },
+    {
+      "epoch": 1.9716312056737588,
+      "grad_norm": 3.149191379547119,
+      "learning_rate": 7.77511961722488e-07,
+      "loss": 1.5683,
+      "step": 834
+    },
+    {
+      "epoch": 1.9739952718676124,
+      "grad_norm": 3.09270977973938,
+      "learning_rate": 7.177033492822967e-07,
+      "loss": 1.5947,
+      "step": 835
+    },
+    {
+      "epoch": 1.9763593380614657,
+      "grad_norm": 3.1009361743927,
+      "learning_rate": 6.578947368421053e-07,
+      "loss": 1.5997,
+      "step": 836
+    },
+    {
+      "epoch": 1.978723404255319,
+      "grad_norm": 3.1177103519439697,
+      "learning_rate": 5.980861244019139e-07,
+      "loss": 1.645,
+      "step": 837
+    },
+    {
+      "epoch": 1.9810874704491725,
+      "grad_norm": 3.130715847015381,
+      "learning_rate": 5.382775119617225e-07,
+      "loss": 1.564,
+      "step": 838
+    },
+    {
+      "epoch": 1.983451536643026,
+      "grad_norm": 3.048459529876709,
+      "learning_rate": 4.784688995215311e-07,
+      "loss": 1.6382,
+      "step": 839
+    },
+    {
+      "epoch": 1.9858156028368794,
+      "grad_norm": 2.779651403427124,
+      "learning_rate": 4.1866028708133973e-07,
+      "loss": 1.408,
+      "step": 840
+    },
+    {
+      "epoch": 1.9881796690307327,
+      "grad_norm": 3.279956579208374,
+      "learning_rate": 3.5885167464114835e-07,
+      "loss": 1.585,
+      "step": 841
+    },
+    {
+      "epoch": 1.9905437352245863,
+      "grad_norm": 3.1854429244995117,
+      "learning_rate": 2.990430622009569e-07,
+      "loss": 1.6264,
+      "step": 842
+    },
+    {
+      "epoch": 1.9929078014184398,
+      "grad_norm": 3.018812417984009,
+      "learning_rate": 2.3923444976076555e-07,
+      "loss": 1.6454,
+      "step": 843
+    },
+    {
+      "epoch": 1.9952718676122931,
+      "grad_norm": 2.9629745483398438,
+      "learning_rate": 1.7942583732057418e-07,
+      "loss": 1.5372,
+      "step": 844
+    },
+    {
+      "epoch": 1.9976359338061465,
+      "grad_norm": 3.0809407234191895,
+      "learning_rate": 1.1961722488038278e-07,
+      "loss": 1.5961,
+      "step": 845
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.675367593765259,
+      "learning_rate": 5.980861244019139e-08,
+      "loss": 1.585,
+      "step": 846
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 846,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0786650102367027e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}