diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4978 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1410,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00425531914893617,
+      "grad_norm": 1.125,
+      "learning_rate": 8.450704225352114e-07,
+      "loss": 1.4721390008926392,
+      "step": 2
+    },
+    {
+      "epoch": 0.00851063829787234,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.535211267605634e-06,
+      "loss": 1.9258784055709839,
+      "step": 4
+    },
+    {
+      "epoch": 0.01276595744680851,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.225352112676057e-06,
+      "loss": 1.9619797468185425,
+      "step": 6
+    },
+    {
+      "epoch": 0.01702127659574468,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.915492957746478e-06,
+      "loss": 1.6346324682235718,
+      "step": 8
+    },
+    {
+      "epoch": 0.02127659574468085,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.6056338028169015e-06,
+      "loss": 1.9556920528411865,
+      "step": 10
+    },
+    {
+      "epoch": 0.02553191489361702,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.295774647887323e-06,
+      "loss": 1.470306634902954,
+      "step": 12
+    },
+    {
+      "epoch": 0.029787234042553193,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.0985915492957748e-05,
+      "loss": 2.068174362182617,
+      "step": 14
+    },
+    {
+      "epoch": 0.03404255319148936,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.267605633802817e-05,
+      "loss": 1.903701663017273,
+      "step": 16
+    },
+    {
+      "epoch": 0.03829787234042553,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.436619718309859e-05,
+      "loss": 1.765465259552002,
+      "step": 18
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.6056338028169014e-05,
+      "loss": 1.9205501079559326,
+      "step": 20
+    },
+    {
+      "epoch": 0.04680851063829787,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.7746478873239435e-05,
+      "loss": 1.6455962657928467,
+      "step": 22
+    },
+    {
+      "epoch": 0.05106382978723404,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.943661971830986e-05,
+      "loss": 1.6872483491897583,
+      "step": 24
+    },
+    {
+      "epoch": 0.05531914893617021,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.112676056338028e-05,
+      "loss": 1.4278151988983154,
+      "step": 26
+    },
+    {
+      "epoch": 0.059574468085106386,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.2816901408450703e-05,
+      "loss": 1.7643554210662842,
+      "step": 28
+    },
+    {
+      "epoch": 0.06382978723404255,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.4507042253521128e-05,
+      "loss": 1.686973214149475,
+      "step": 30
+    },
+    {
+      "epoch": 0.06808510638297872,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.619718309859155e-05,
+      "loss": 1.4633193016052246,
+      "step": 32
+    },
+    {
+      "epoch": 0.07234042553191489,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.788732394366197e-05,
+      "loss": 1.6693249940872192,
+      "step": 34
+    },
+    {
+      "epoch": 0.07659574468085106,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9577464788732395e-05,
+      "loss": 1.657246708869934,
+      "step": 36
+    },
+    {
+      "epoch": 0.08085106382978724,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.126760563380282e-05,
+      "loss": 1.882036805152893,
+      "step": 38
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.2957746478873245e-05,
+      "loss": 1.618768572807312,
+      "step": 40
+    },
+    {
+      "epoch": 0.08936170212765958,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.4647887323943666e-05,
+      "loss": 1.29584801197052,
+      "step": 42
+    },
+    {
+      "epoch": 0.09361702127659574,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.633802816901408e-05,
+      "loss": 1.5222253799438477,
+      "step": 44
+    },
+    {
+      "epoch": 0.09787234042553192,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.802816901408451e-05,
+      "loss": 1.3532958030700684,
+      "step": 46
+    },
+    {
+      "epoch": 0.10212765957446808,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.971830985915493e-05,
+      "loss": 1.5933119058609009,
+      "step": 48
+    },
+    {
+      "epoch": 0.10638297872340426,
+      "grad_norm": 1.671875,
+      "learning_rate": 4.140845070422535e-05,
+      "loss": 1.6478880643844604,
+      "step": 50
+    },
+    {
+      "epoch": 0.11063829787234042,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.309859154929578e-05,
+      "loss": 1.3426867723464966,
+      "step": 52
+    },
+    {
+      "epoch": 0.1148936170212766,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.47887323943662e-05,
+      "loss": 1.4301996231079102,
+      "step": 54
+    },
+    {
+      "epoch": 0.11914893617021277,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.6478873239436617e-05,
+      "loss": 1.2384978532791138,
+      "step": 56
+    },
+    {
+      "epoch": 0.12340425531914893,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.8169014084507045e-05,
+      "loss": 1.4298828840255737,
+      "step": 58
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.9859154929577466e-05,
+      "loss": 1.3327827453613281,
+      "step": 60
+    },
+    {
+      "epoch": 0.13191489361702127,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.154929577464789e-05,
+      "loss": 1.281138300895691,
+      "step": 62
+    },
+    {
+      "epoch": 0.13617021276595745,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.3239436619718316e-05,
+      "loss": 1.064106822013855,
+      "step": 64
+    },
+    {
+      "epoch": 0.14042553191489363,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.492957746478874e-05,
+      "loss": 1.4334735870361328,
+      "step": 66
+    },
+    {
+      "epoch": 0.14468085106382977,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.661971830985915e-05,
+      "loss": 1.3736847639083862,
+      "step": 68
+    },
+    {
+      "epoch": 0.14893617021276595,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.830985915492958e-05,
+      "loss": 1.3596861362457275,
+      "step": 70
+    },
+    {
+      "epoch": 0.15319148936170213,
+      "grad_norm": 0.671875,
+      "learning_rate": 6e-05,
+      "loss": 1.4395864009857178,
+      "step": 72
+    },
+    {
+      "epoch": 0.1574468085106383,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.9999735771881775e-05,
+      "loss": 1.5454157590866089,
+      "step": 74
+    },
+    {
+      "epoch": 0.16170212765957448,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.99989430933451e-05,
+      "loss": 1.3853117227554321,
+      "step": 76
+    },
+    {
+      "epoch": 0.16595744680851063,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.9997621981843994e-05,
+      "loss": 1.2754864692687988,
+      "step": 78
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.999577246646804e-05,
+      "loss": 1.5860013961791992,
+      "step": 80
+    },
+    {
+      "epoch": 0.17446808510638298,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.99933945879417e-05,
+      "loss": 1.237565517425537,
+      "step": 82
+    },
+    {
+      "epoch": 0.17872340425531916,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.999048839862352e-05,
+      "loss": 1.2814902067184448,
+      "step": 84
+    },
+    {
+      "epoch": 0.1829787234042553,
+      "grad_norm": 1.4296875,
+      "learning_rate": 5.998705396250493e-05,
+      "loss": 1.3300844430923462,
+      "step": 86
+    },
+    {
+      "epoch": 0.18723404255319148,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.998309135520878e-05,
+      "loss": 1.3447200059890747,
+      "step": 88
+    },
+    {
+      "epoch": 0.19148936170212766,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.997860066398778e-05,
+      "loss": 1.3371561765670776,
+      "step": 90
+    },
+    {
+      "epoch": 0.19574468085106383,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.997358198772249e-05,
+      "loss": 1.3265268802642822,
+      "step": 92
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.9968035436919206e-05,
+      "loss": 1.1992994546890259,
+      "step": 94
+    },
+    {
+      "epoch": 0.20425531914893616,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.996196113370748e-05,
+      "loss": 1.2414522171020508,
+      "step": 96
+    },
+    {
+      "epoch": 0.20851063829787234,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.9955359211837465e-05,
+      "loss": 1.314779281616211,
+      "step": 98
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.994822981667691e-05,
+      "loss": 1.2792837619781494,
+      "step": 100
+    },
+    {
+      "epoch": 0.2170212765957447,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.994057310520807e-05,
+      "loss": 1.234731912612915,
+      "step": 102
+    },
+    {
+      "epoch": 0.22127659574468084,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.993238924602414e-05,
+      "loss": 1.1855965852737427,
+      "step": 104
+    },
+    {
+      "epoch": 0.225531914893617,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.9923678419325616e-05,
+      "loss": 1.3185398578643799,
+      "step": 106
+    },
+    {
+      "epoch": 0.2297872340425532,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.9914440816916236e-05,
+      "loss": 1.3415206670761108,
+      "step": 108
+    },
+    {
+      "epoch": 0.23404255319148937,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.990467664219887e-05,
+      "loss": 1.7355936765670776,
+      "step": 110
+    },
+    {
+      "epoch": 0.23829787234042554,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.989438611017101e-05,
+      "loss": 1.299019455909729,
+      "step": 112
+    },
+    {
+      "epoch": 0.2425531914893617,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.9883569447419946e-05,
+      "loss": 1.2734484672546387,
+      "step": 114
+    },
+    {
+      "epoch": 0.24680851063829787,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.98722268921179e-05,
+      "loss": 1.3722704648971558,
+      "step": 116
+    },
+    {
+      "epoch": 0.251063829787234,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.9860358694016695e-05,
+      "loss": 1.241133451461792,
+      "step": 118
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.984796511444231e-05,
+      "loss": 1.2068897485733032,
+      "step": 120
+    },
+    {
+      "epoch": 0.25957446808510637,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.983504642628911e-05,
+      "loss": 1.0534322261810303,
+      "step": 122
+    },
+    {
+      "epoch": 0.26382978723404255,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.9821602914013794e-05,
+      "loss": 1.306460976600647,
+      "step": 124
+    },
+    {
+      "epoch": 0.2680851063829787,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.9807634873629205e-05,
+      "loss": 1.3372706174850464,
+      "step": 126
+    },
+    {
+      "epoch": 0.2723404255319149,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.979314261269775e-05,
+      "loss": 1.3372623920440674,
+      "step": 128
+    },
+    {
+      "epoch": 0.2765957446808511,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.9778126450324674e-05,
+      "loss": 1.303137183189392,
+      "step": 130
+    },
+    {
+      "epoch": 0.28085106382978725,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.9762586717151e-05,
+      "loss": 1.2485641241073608,
+      "step": 132
+    },
+    {
+      "epoch": 0.2851063829787234,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.9746523755346284e-05,
+      "loss": 1.2784473896026611,
+      "step": 134
+    },
+    {
+      "epoch": 0.28936170212765955,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.972993791860101e-05,
+      "loss": 1.3001835346221924,
+      "step": 136
+    },
+    {
+      "epoch": 0.2936170212765957,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.97128295721189e-05,
+      "loss": 1.2696309089660645,
+      "step": 138
+    },
+    {
+      "epoch": 0.2978723404255319,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.969519909260879e-05,
+      "loss": 1.4176753759384155,
+      "step": 140
+    },
+    {
+      "epoch": 0.3021276595744681,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.967704686827641e-05,
+      "loss": 1.3172787427902222,
+      "step": 142
+    },
+    {
+      "epoch": 0.30638297872340425,
+      "grad_norm": 14.0,
+      "learning_rate": 5.965837329881574e-05,
+      "loss": 1.4879260063171387,
+      "step": 144
+    },
+    {
+      "epoch": 0.31063829787234043,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.9639178795400295e-05,
+      "loss": 1.1774471998214722,
+      "step": 146
+    },
+    {
+      "epoch": 0.3148936170212766,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.9619463780674034e-05,
+      "loss": 1.2439987659454346,
+      "step": 148
+    },
+    {
+      "epoch": 0.3191489361702128,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.959922868874206e-05,
+      "loss": 1.205959439277649,
+      "step": 150
+    },
+    {
+      "epoch": 0.32340425531914896,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.9578473965161075e-05,
+      "loss": 1.330673098564148,
+      "step": 152
+    },
+    {
+      "epoch": 0.3276595744680851,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.9557200066929534e-05,
+      "loss": 1.2126266956329346,
+      "step": 154
+    },
+    {
+      "epoch": 0.33191489361702126,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.95354074624776e-05,
+      "loss": 1.2708547115325928,
+      "step": 156
+    },
+    {
+      "epoch": 0.33617021276595743,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.951309663165686e-05,
+      "loss": 1.2793720960617065,
+      "step": 158
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 0.310546875,
+      "learning_rate": 5.9490268065729724e-05,
+      "loss": 1.2078324556350708,
+      "step": 160
+    },
+    {
+      "epoch": 0.3446808510638298,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.946692226735858e-05,
+      "loss": 1.2282007932662964,
+      "step": 162
+    },
+    {
+      "epoch": 0.34893617021276596,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.944305975059482e-05,
+      "loss": 1.2256958484649658,
+      "step": 164
+    },
+    {
+      "epoch": 0.35319148936170214,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.9418681040867445e-05,
+      "loss": 1.2895530462265015,
+      "step": 166
+    },
+    {
+      "epoch": 0.3574468085106383,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.939378667497147e-05,
+      "loss": 1.1319845914840698,
+      "step": 168
+    },
+    {
+      "epoch": 0.3617021276595745,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.9368377201056195e-05,
+      "loss": 1.3089065551757812,
+      "step": 170
+    },
+    {
+      "epoch": 0.3659574468085106,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.934245317861308e-05,
+      "loss": 1.2677334547042847,
+      "step": 172
+    },
+    {
+      "epoch": 0.3702127659574468,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.9316015178463446e-05,
+      "loss": 1.2161953449249268,
+      "step": 174
+    },
+    {
+      "epoch": 0.37446808510638296,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.928906378274585e-05,
+      "loss": 1.4091310501098633,
+      "step": 176
+    },
+    {
+      "epoch": 0.37872340425531914,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.9261599584903375e-05,
+      "loss": 1.3989202976226807,
+      "step": 178
+    },
+    {
+      "epoch": 0.3829787234042553,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.923362318967043e-05,
+      "loss": 1.1342294216156006,
+      "step": 180
+    },
+    {
+      "epoch": 0.3872340425531915,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.920513521305955e-05,
+      "loss": 1.2139039039611816,
+      "step": 182
+    },
+    {
+      "epoch": 0.39148936170212767,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.9176136282347777e-05,
+      "loss": 1.12371027469635,
+      "step": 184
+    },
+    {
+      "epoch": 0.39574468085106385,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.914662703606285e-05,
+      "loss": 1.3441473245620728,
+      "step": 186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.911660812396916e-05,
+      "loss": 1.2006664276123047,
+      "step": 188
+    },
+    {
+      "epoch": 0.40425531914893614,
+      "grad_norm": 0.5,
+      "learning_rate": 5.90860802070534e-05,
+      "loss": 1.2136142253875732,
+      "step": 190
+    },
+    {
+      "epoch": 0.4085106382978723,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.905504395751011e-05,
+      "loss": 1.2749568223953247,
+      "step": 192
+    },
+    {
+      "epoch": 0.4127659574468085,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.9023500058726754e-05,
+      "loss": 1.369842529296875,
+      "step": 194
+    },
+    {
+      "epoch": 0.41702127659574467,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.899144920526876e-05,
+      "loss": 1.2179538011550903,
+      "step": 196
+    },
+    {
+      "epoch": 0.42127659574468085,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.895889210286416e-05,
+      "loss": 1.3043186664581299,
+      "step": 198
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.8925829468388156e-05,
+      "loss": 1.309710144996643,
+      "step": 200
+    },
+    {
+      "epoch": 0.4297872340425532,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.8892262029847186e-05,
+      "loss": 1.233137607574463,
+      "step": 202
+    },
+    {
+      "epoch": 0.4340425531914894,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.885819052636303e-05,
+      "loss": 1.1934456825256348,
+      "step": 204
+    },
+    {
+      "epoch": 0.43829787234042555,
+      "grad_norm": 0.625,
+      "learning_rate": 5.882361570815645e-05,
+      "loss": 1.2160348892211914,
+      "step": 206
+    },
+    {
+      "epoch": 0.4425531914893617,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.8788538336530715e-05,
+      "loss": 1.2657216787338257,
+      "step": 208
+    },
+    {
+      "epoch": 0.44680851063829785,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.8752959183854816e-05,
+      "loss": 1.2517626285552979,
+      "step": 210
+    },
+    {
+      "epoch": 0.451063829787234,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.871687903354646e-05,
+      "loss": 1.1521042585372925,
+      "step": 212
+    },
+    {
+      "epoch": 0.4553191489361702,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.8680298680054824e-05,
+      "loss": 1.2568317651748657,
+      "step": 214
+    },
+    {
+      "epoch": 0.4595744680851064,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.864321892884309e-05,
+      "loss": 1.2429876327514648,
+      "step": 216
+    },
+    {
+      "epoch": 0.46382978723404256,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.860564059637066e-05,
+      "loss": 1.327577829360962,
+      "step": 218
+    },
+    {
+      "epoch": 0.46808510638297873,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.85675645100752e-05,
+      "loss": 1.416573166847229,
+      "step": 220
+    },
+    {
+      "epoch": 0.4723404255319149,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.852899150835445e-05,
+      "loss": 1.241747498512268,
+      "step": 222
+    },
+    {
+      "epoch": 0.4765957446808511,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.848992244054768e-05,
+      "loss": 1.2418628931045532,
+      "step": 224
+    },
+    {
+      "epoch": 0.4808510638297872,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.845035816691711e-05,
+      "loss": 1.1999211311340332,
+      "step": 226
+    },
+    {
+      "epoch": 0.4851063829787234,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.841029955862885e-05,
+      "loss": 1.4046475887298584,
+      "step": 228
+    },
+    {
+      "epoch": 0.48936170212765956,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.836974749773382e-05,
+      "loss": 1.2456966638565063,
+      "step": 230
+    },
+    {
+      "epoch": 0.49361702127659574,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.83287028771482e-05,
+      "loss": 1.2392256259918213,
+      "step": 232
+    },
+    {
+      "epoch": 0.4978723404255319,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.828716660063395e-05,
+      "loss": 1.3959895372390747,
+      "step": 234
+    },
+    {
+      "epoch": 0.502127659574468,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.824513958277871e-05,
+      "loss": 1.2444478273391724,
+      "step": 236
+    },
+    {
+      "epoch": 0.5063829787234042,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.8202622748975817e-05,
+      "loss": 1.294707179069519,
+      "step": 238
+    },
+    {
+      "epoch": 0.5106382978723404,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.815961703540386e-05,
+      "loss": 1.3415395021438599,
+      "step": 240
+    },
+    {
+      "epoch": 0.5148936170212766,
+      "grad_norm": 1.328125,
+      "learning_rate": 5.811612338900606e-05,
+      "loss": 1.1040756702423096,
+      "step": 242
+    },
+    {
+      "epoch": 0.5191489361702127,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.807214276746943e-05,
+      "loss": 1.1671684980392456,
+      "step": 244
+    },
+    {
+      "epoch": 0.5234042553191489,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.802767613920375e-05,
+      "loss": 1.1621744632720947,
+      "step": 246
+    },
+    {
+      "epoch": 0.5276595744680851,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.7982724483320105e-05,
+      "loss": 1.2460886240005493,
+      "step": 248
+    },
+    {
+      "epoch": 0.5319148936170213,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.793728878960947e-05,
+      "loss": 1.3446158170700073,
+      "step": 250
+    },
+    {
+      "epoch": 0.5361702127659574,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.789137005852084e-05,
+      "loss": 1.4579079151153564,
+      "step": 252
+    },
+    {
+      "epoch": 0.5404255319148936,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.7844969301139194e-05,
+      "loss": 1.4334136247634888,
+      "step": 254
+    },
+    {
+      "epoch": 0.5446808510638298,
+      "grad_norm": 1.421875,
+      "learning_rate": 5.779808753916325e-05,
+      "loss": 1.6397161483764648,
+      "step": 256
+    },
+    {
+      "epoch": 0.548936170212766,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.7750725804883034e-05,
+      "loss": 1.301457166671753,
+      "step": 258
+    },
+    {
+      "epoch": 0.5531914893617021,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.770288514115698e-05,
+      "loss": 1.2889872789382935,
+      "step": 260
+    },
+    {
+      "epoch": 0.5574468085106383,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.765456660138919e-05,
+      "loss": 1.2275917530059814,
+      "step": 262
+    },
+    {
+      "epoch": 0.5617021276595745,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.760577124950603e-05,
+      "loss": 1.2264875173568726,
+      "step": 264
+    },
+    {
+      "epoch": 0.5659574468085107,
+      "grad_norm": 0.75,
+      "learning_rate": 5.7556500159932835e-05,
+      "loss": 1.3288713693618774,
+      "step": 266
+    },
+    {
+      "epoch": 0.5702127659574469,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.750675441757024e-05,
+      "loss": 1.1868747472763062,
+      "step": 268
+    },
+    {
+      "epoch": 0.574468085106383,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.7456535117770204e-05,
+      "loss": 1.3627235889434814,
+      "step": 270
+    },
+    {
+      "epoch": 0.5787234042553191,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.7405843366312e-05,
+      "loss": 1.2165353298187256,
+      "step": 272
+    },
+    {
+      "epoch": 0.5829787234042553,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.735468027937782e-05,
+      "loss": 1.3962359428405762,
+      "step": 274
+    },
+    {
+      "epoch": 0.5872340425531914,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.730304698352815e-05,
+      "loss": 1.2368425130844116,
+      "step": 276
+    },
+    {
+      "epoch": 0.5914893617021276,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.7250944615677036e-05,
+      "loss": 1.2761774063110352,
+      "step": 278
+    },
+    {
+      "epoch": 0.5957446808510638,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.719837432306707e-05,
+      "loss": 1.146309494972229,
+      "step": 280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.7145337263244e-05,
+      "loss": 1.3055988550186157,
+      "step": 282
+    },
+    {
+      "epoch": 0.6042553191489362,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.7091834604031386e-05,
+      "loss": 1.2853789329528809,
+      "step": 284
+    },
+    {
+      "epoch": 0.6085106382978723,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.703786752350482e-05,
+      "loss": 1.2626879215240479,
+      "step": 286
+    },
+    {
+      "epoch": 0.6127659574468085,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.698343720996596e-05,
+      "loss": 1.0880780220031738,
+      "step": 288
+    },
+    {
+      "epoch": 0.6170212765957447,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.6928544861916416e-05,
+      "loss": 1.4804942607879639,
+      "step": 290
+    },
+    {
+      "epoch": 0.6212765957446809,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.687319168803136e-05,
+      "loss": 1.2268519401550293,
+      "step": 292
+    },
+    {
+      "epoch": 0.625531914893617,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.681737890713286e-05,
+      "loss": 1.2773425579071045,
+      "step": 294
+    },
+    {
+      "epoch": 0.6297872340425532,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.676110774816309e-05,
+      "loss": 1.3065530061721802,
+      "step": 296
+    },
+    {
+      "epoch": 0.6340425531914894,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.670437945015725e-05,
+      "loss": 1.2505621910095215,
+      "step": 298
+    },
+    {
+      "epoch": 0.6382978723404256,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.6647195262216296e-05,
+      "loss": 1.2247607707977295,
+      "step": 300
+    },
+    {
+      "epoch": 0.6425531914893617,
+      "grad_norm": 0.296875,
+      "learning_rate": 5.658955644347944e-05,
+      "loss": 1.3000612258911133,
+      "step": 302
+    },
+    {
+      "epoch": 0.6468085106382979,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.653146426309637e-05,
+      "loss": 1.2301831245422363,
+      "step": 304
+    },
+    {
+      "epoch": 0.6510638297872341,
+      "grad_norm": 2.875,
+      "learning_rate": 5.64729200001994e-05,
+      "loss": 1.23581862449646,
+      "step": 306
+    },
+    {
+      "epoch": 0.6553191489361702,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.6413924943875225e-05,
+      "loss": 1.2415351867675781,
+      "step": 308
+    },
+    {
+      "epoch": 0.6595744680851063,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.635448039313658e-05,
+      "loss": 1.054430365562439,
+      "step": 310
+    },
+    {
+      "epoch": 0.6638297872340425,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.62945876568936e-05,
+      "loss": 1.1587709188461304,
+      "step": 312
+    },
+    {
+      "epoch": 0.6680851063829787,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.623424805392504e-05,
+      "loss": 1.2375378608703613,
+      "step": 314
+    },
+    {
+      "epoch": 0.6723404255319149,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.617346291284922e-05,
+      "loss": 1.2096360921859741,
+      "step": 316
+    },
+    {
+      "epoch": 0.676595744680851,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.611223357209474e-05,
+      "loss": 1.4315375089645386,
+      "step": 318
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.6050561379871077e-05,
+      "loss": 1.2253354787826538,
+      "step": 320
+    },
+    {
+      "epoch": 0.6851063829787234,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.5988447694138834e-05,
+      "loss": 1.1945621967315674,
+      "step": 322
+    },
+    {
+      "epoch": 0.6893617021276596,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.5925893882579864e-05,
+      "loss": 1.2644530534744263,
+      "step": 324
+    },
+    {
+      "epoch": 0.6936170212765957,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.5862901322567175e-05,
+      "loss": 1.2183250188827515,
+      "step": 326
+    },
+    {
+      "epoch": 0.6978723404255319,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.5799471401134543e-05,
+      "loss": 1.4243056774139404,
+      "step": 328
+    },
+    {
+      "epoch": 0.7021276595744681,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.5735605514946046e-05,
+      "loss": 1.2356775999069214,
+      "step": 330
+    },
+    {
+      "epoch": 0.7063829787234043,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.567130507026527e-05,
+      "loss": 1.245442509651184,
+      "step": 332
+    },
+    {
+      "epoch": 0.7106382978723405,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.560657148292432e-05,
+      "loss": 1.0984729528427124,
+      "step": 334
+    },
+    {
+      "epoch": 0.7148936170212766,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.554140617829271e-05,
+      "loss": 1.2282965183258057,
+      "step": 336
+    },
+    {
+      "epoch": 0.7191489361702128,
+      "grad_norm": 0.875,
+      "learning_rate": 5.547581059124591e-05,
+      "loss": 1.3061587810516357,
+      "step": 338
+    },
+    {
+      "epoch": 0.723404255319149,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.540978616613381e-05,
+      "loss": 1.2117385864257812,
+      "step": 340
+    },
+    {
+      "epoch": 0.7276595744680852,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.534333435674889e-05,
+      "loss": 1.105977177619934,
+      "step": 342
+    },
+    {
+      "epoch": 0.7319148936170212,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.527645662629417e-05,
+      "loss": 1.1953120231628418,
+      "step": 344
+    },
+    {
+      "epoch": 0.7361702127659574,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.520915444735106e-05,
+      "loss": 1.2646780014038086,
+      "step": 346
+    },
+    {
+      "epoch": 0.7404255319148936,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.514142930184689e-05,
+      "loss": 1.2484480142593384,
+      "step": 348
+    },
+    {
+      "epoch": 0.7446808510638298,
+      "grad_norm": 1.8203125,
+      "learning_rate": 5.5073282681022354e-05,
+      "loss": 1.274402379989624,
+      "step": 350
+    },
+    {
+      "epoch": 0.7489361702127659,
+      "grad_norm": 0.37890625,
+      "learning_rate": 5.5004716085398515e-05,
+      "loss": 1.182061791419983,
+      "step": 352
+    },
+    {
+      "epoch": 0.7531914893617021,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.493573102474395e-05,
+      "loss": 1.1438305377960205,
+      "step": 354
+    },
+    {
+      "epoch": 0.7574468085106383,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.486632901804137e-05,
+      "loss": 1.2968688011169434,
+      "step": 356
+    },
+    {
+      "epoch": 0.7617021276595745,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.4796511593454254e-05,
+      "loss": 1.1422514915466309,
+      "step": 358
+    },
+    {
+      "epoch": 0.7659574468085106,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.4726280288293156e-05,
+      "loss": 1.20034658908844,
+      "step": 360
+    },
+    {
+      "epoch": 0.7702127659574468,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.4655636648981876e-05,
+      "loss": 1.224547266960144,
+      "step": 362
+    },
+    {
+      "epoch": 0.774468085106383,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.458458223102342e-05,
+      "loss": 1.2625551223754883,
+      "step": 364
+    },
+    {
+      "epoch": 0.7787234042553192,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.45131185989657e-05,
+      "loss": 1.170459508895874,
+      "step": 366
+    },
+    {
+      "epoch": 0.7829787234042553,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.444124732636714e-05,
+      "loss": 1.1499803066253662,
+      "step": 368
+    },
+    {
+      "epoch": 0.7872340425531915,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.4368969995762e-05,
+      "loss": 1.2377893924713135,
+      "step": 370
+    },
+    {
+      "epoch": 0.7914893617021277,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.429628819862553e-05,
+      "loss": 1.1148861646652222,
+      "step": 372
+    },
+    {
+      "epoch": 0.7957446808510639,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.4223203535338945e-05,
+      "loss": 1.0856741666793823,
+      "step": 374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.414971761515417e-05,
+      "loss": 1.150919795036316,
+      "step": 376
+    },
+    {
+      "epoch": 0.8042553191489362,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.4075832056158395e-05,
+      "loss": 1.21478271484375,
+      "step": 378
+    },
+    {
+      "epoch": 0.8085106382978723,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.400154848523847e-05,
+      "loss": 1.2340768575668335,
+      "step": 380
+    },
+    {
+      "epoch": 0.8127659574468085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.392686853804508e-05,
+      "loss": 1.1131625175476074,
+      "step": 382
+    },
+    {
+      "epoch": 0.8170212765957446,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.385179385895671e-05,
+      "loss": 1.203751564025879,
+      "step": 384
+    },
+    {
+      "epoch": 0.8212765957446808,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.377632610104347e-05,
+      "loss": 1.090701937675476,
+      "step": 386
+    },
+    {
+      "epoch": 0.825531914893617,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.370046692603067e-05,
+      "loss": 1.1695449352264404,
+      "step": 388
+    },
+    {
+      "epoch": 0.8297872340425532,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.362421800426221e-05,
+      "loss": 1.2987892627716064,
+      "step": 390
+    },
+    {
+      "epoch": 0.8340425531914893,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.354758101466387e-05,
+      "loss": 1.145818829536438,
+      "step": 392
+    },
+    {
+      "epoch": 0.8382978723404255,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.347055764470628e-05,
+      "loss": 1.0881308317184448,
+      "step": 394
+    },
+    {
+      "epoch": 0.8425531914893617,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.3393149590367764e-05,
+      "loss": 1.312870740890503,
+      "step": 396
+    },
+    {
+      "epoch": 0.8468085106382979,
+      "grad_norm": 1.3203125,
+      "learning_rate": 5.331535855609704e-05,
+      "loss": 1.239678144454956,
+      "step": 398
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.323718625477566e-05,
+      "loss": 1.1566566228866577,
+      "step": 400
+    },
+    {
+      "epoch": 0.8553191489361702,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.315863440768031e-05,
+      "loss": 1.2142958641052246,
+      "step": 402
+    },
+    {
+      "epoch": 0.8595744680851064,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.307970474444487e-05,
+      "loss": 1.1810637712478638,
+      "step": 404
+    },
+    {
+      "epoch": 0.8638297872340426,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.300039900302237e-05,
+      "loss": 1.216611623764038,
+      "step": 406
+    },
+    {
+      "epoch": 0.8680851063829788,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.292071892964672e-05,
+      "loss": 1.2652095556259155,
+      "step": 408
+    },
+    {
+      "epoch": 0.8723404255319149,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.2840666278794244e-05,
+      "loss": 1.1579747200012207,
+      "step": 410
+    },
+    {
+      "epoch": 0.8765957446808511,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.276024281314504e-05,
+      "loss": 1.2436968088150024,
+      "step": 412
+    },
+    {
+      "epoch": 0.8808510638297873,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.267945030354419e-05,
+      "loss": 1.1159509420394897,
+      "step": 414
+    },
+    {
+      "epoch": 0.8851063829787233,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.259829052896277e-05,
+      "loss": 1.1389038562774658,
+      "step": 416
+    },
+    {
+      "epoch": 0.8893617021276595,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.251676527645864e-05,
+      "loss": 1.179782509803772,
+      "step": 418
+    },
+    {
+      "epoch": 0.8936170212765957,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.243487634113716e-05,
+      "loss": 1.2923263311386108,
+      "step": 420
+    },
+    {
+      "epoch": 0.8978723404255319,
+      "grad_norm": 2.09375,
+      "learning_rate": 5.235262552611159e-05,
+      "loss": 1.0829358100891113,
+      "step": 422
+    },
+    {
+      "epoch": 0.902127659574468,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.227001464246343e-05,
+      "loss": 1.155044674873352,
+      "step": 424
+    },
+    {
+      "epoch": 0.9063829787234042,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.218704550920257e-05,
+      "loss": 1.3217086791992188,
+      "step": 426
+    },
+    {
+      "epoch": 0.9106382978723404,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.210371995322713e-05,
+      "loss": 1.2187201976776123,
+      "step": 428
+    },
+    {
+      "epoch": 0.9148936170212766,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.202003980928338e-05,
+      "loss": 1.1969107389450073,
+      "step": 430
+    },
+    {
+      "epoch": 0.9191489361702128,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.1936006919925216e-05,
+      "loss": 1.2421386241912842,
+      "step": 432
+    },
+    {
+      "epoch": 0.9234042553191489,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.185162313547368e-05,
+      "loss": 1.5777398347854614,
+      "step": 434
+    },
+    {
+      "epoch": 0.9276595744680851,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.176689031397612e-05,
+      "loss": 1.2189395427703857,
+      "step": 436
+    },
+    {
+      "epoch": 0.9319148936170213,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.16818103211654e-05,
+      "loss": 1.191472053527832,
+      "step": 438
+    },
+    {
+      "epoch": 0.9361702127659575,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.1596385030418686e-05,
+      "loss": 1.0312530994415283,
+      "step": 440
+    },
+    {
+      "epoch": 0.9404255319148936,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.151061632271633e-05,
+      "loss": 1.2013260126113892,
+      "step": 442
+    },
+    {
+      "epoch": 0.9446808510638298,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.142450608660035e-05,
+      "loss": 1.1814100742340088,
+      "step": 444
+    },
+    {
+      "epoch": 0.948936170212766,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.133805621813285e-05,
+      "loss": 1.3198750019073486,
+      "step": 446
+    },
+    {
+      "epoch": 0.9531914893617022,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.1251268620854375e-05,
+      "loss": 1.1754573583602905,
+      "step": 448
+    },
+    {
+      "epoch": 0.9574468085106383,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.116414520574186e-05,
+      "loss": 1.1766588687896729,
+      "step": 450
+    },
+    {
+      "epoch": 0.9617021276595744,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.107668789116664e-05,
+      "loss": 1.1669294834136963,
+      "step": 452
+    },
+    {
+      "epoch": 0.9659574468085106,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.0988898602852204e-05,
+      "loss": 1.0226119756698608,
+      "step": 454
+    },
+    {
+      "epoch": 0.9702127659574468,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.090077927383173e-05,
+      "loss": 1.184398889541626,
+      "step": 456
+    },
+    {
+      "epoch": 0.9744680851063829,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.081233184440562e-05,
+      "loss": 1.2115238904953003,
+      "step": 458
+    },
+    {
+      "epoch": 0.9787234042553191,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.072355826209869e-05,
+      "loss": 1.169049620628357,
+      "step": 460
+    },
+    {
+      "epoch": 0.9829787234042553,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.063446048161731e-05,
+      "loss": 1.2979567050933838,
+      "step": 462
+    },
+    {
+      "epoch": 0.9872340425531915,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.054504046480638e-05,
+      "loss": 1.046877145767212,
+      "step": 464
+    },
+    {
+      "epoch": 0.9914893617021276,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.0455300180606165e-05,
+      "loss": 1.2601969242095947,
+      "step": 466
+    },
+    {
+      "epoch": 0.9957446808510638,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.036524160500883e-05,
+      "loss": 1.1848105192184448,
+      "step": 468
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.027486672101506e-05,
+      "loss": 1.2240030765533447,
+      "step": 470
+    },
+    {
+      "epoch": 1.004255319148936,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.0184177518590294e-05,
+      "loss": 0.9678438901901245,
+      "step": 472
+    },
+    {
+      "epoch": 1.0085106382978724,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.0093175994621006e-05,
+      "loss": 0.8098158836364746,
+      "step": 474
+    },
+    {
+      "epoch": 1.0127659574468084,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.000186415287064e-05,
+      "loss": 1.0377544164657593,
+      "step": 476
+    },
+    {
+      "epoch": 1.0170212765957447,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.9910244003935514e-05,
+      "loss": 0.9136829376220703,
+      "step": 478
+    },
+    {
+      "epoch": 1.0212765957446808,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.981831756520061e-05,
+      "loss": 0.8520618081092834,
+      "step": 480
+    },
+    {
+      "epoch": 1.025531914893617,
+      "grad_norm": 0.875,
+      "learning_rate": 4.972608686079509e-05,
+      "loss": 0.7956000566482544,
+      "step": 482
+    },
+    {
+      "epoch": 1.0297872340425531,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.963355392154774e-05,
+      "loss": 0.9468518495559692,
+      "step": 484
+    },
+    {
+      "epoch": 1.0340425531914894,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.954072078494226e-05,
+      "loss": 1.131507158279419,
+      "step": 486
+    },
+    {
+      "epoch": 1.0382978723404255,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.944758949507241e-05,
+      "loss": 1.0291956663131714,
+      "step": 488
+    },
+    {
+      "epoch": 1.0425531914893618,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.9354162102596946e-05,
+      "loss": 1.1023054122924805,
+      "step": 490
+    },
+    {
+      "epoch": 1.0468085106382978,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.926044066469459e-05,
+      "loss": 0.9480533599853516,
+      "step": 492
+    },
+    {
+      "epoch": 1.0510638297872341,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.916642724501856e-05,
+      "loss": 0.898105263710022,
+      "step": 494
+    },
+    {
+      "epoch": 1.0553191489361702,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.9072123913651306e-05,
+      "loss": 0.9935942888259888,
+      "step": 496
+    },
+    {
+      "epoch": 1.0595744680851065,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.89775327470588e-05,
+      "loss": 0.7786449193954468,
+      "step": 498
+    },
+    {
+      "epoch": 1.0638297872340425,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.888265582804487e-05,
+      "loss": 0.8598610758781433,
+      "step": 500
+    },
+    {
+      "epoch": 1.0680851063829788,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.878749524570533e-05,
+      "loss": 0.9269306063652039,
+      "step": 502
+    },
+    {
+      "epoch": 1.0723404255319149,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.869205309538197e-05,
+      "loss": 0.6500522494316101,
+      "step": 504
+    },
+    {
+      "epoch": 1.076595744680851,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.8596331478616454e-05,
+      "loss": 0.9429140090942383,
+      "step": 506
+    },
+    {
+      "epoch": 1.0808510638297872,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.8500332503103996e-05,
+      "loss": 0.9708827137947083,
+      "step": 508
+    },
+    {
+      "epoch": 1.0851063829787233,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.8404058282646985e-05,
+      "loss": 0.9373714327812195,
+      "step": 510
+    },
+    {
+      "epoch": 1.0893617021276596,
+      "grad_norm": 1.671875,
+      "learning_rate": 4.830751093710844e-05,
+      "loss": 1.1464526653289795,
+      "step": 512
+    },
+    {
+      "epoch": 1.0936170212765957,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.8210692592365296e-05,
+      "loss": 0.8451336026191711,
+      "step": 514
+    },
+    {
+      "epoch": 1.097872340425532,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.811360538026165e-05,
+      "loss": 1.058915376663208,
+      "step": 516
+    },
+    {
+      "epoch": 1.102127659574468,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.801625143856179e-05,
+      "loss": 1.225139856338501,
+      "step": 518
+    },
+    {
+      "epoch": 1.1063829787234043,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.79186329109031e-05,
+      "loss": 0.6482216119766235,
+      "step": 520
+    },
+    {
+      "epoch": 1.1106382978723404,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.782075194674892e-05,
+      "loss": 0.802341103553772,
+      "step": 522
+    },
+    {
+      "epoch": 1.1148936170212767,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.772261070134113e-05,
+      "loss": 0.8717508316040039,
+      "step": 524
+    },
+    {
+      "epoch": 1.1191489361702127,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.762421133565284e-05,
+      "loss": 0.8982354402542114,
+      "step": 526
+    },
+    {
+      "epoch": 1.123404255319149,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.752555601634059e-05,
+      "loss": 1.0001122951507568,
+      "step": 528
+    },
+    {
+      "epoch": 1.127659574468085,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.742664691569685e-05,
+      "loss": 0.8902131915092468,
+      "step": 530
+    },
+    {
+      "epoch": 1.1319148936170214,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.732748621160212e-05,
+      "loss": 1.2693915367126465,
+      "step": 532
+    },
+    {
+      "epoch": 1.1361702127659574,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.7228076087476864e-05,
+      "loss": 0.8774423599243164,
+      "step": 534
+    },
+    {
+      "epoch": 1.1404255319148937,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.71284187322336e-05,
+      "loss": 0.9756125211715698,
+      "step": 536
+    },
+    {
+      "epoch": 1.1446808510638298,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.702851634022864e-05,
+      "loss": 0.8841724991798401,
+      "step": 538
+    },
+    {
+      "epoch": 1.148936170212766,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.692837111121371e-05,
+      "loss": 0.794608473777771,
+      "step": 540
+    },
+    {
+      "epoch": 1.1531914893617021,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.6827985250287616e-05,
+      "loss": 0.9677450060844421,
+      "step": 542
+    },
+    {
+      "epoch": 1.1574468085106382,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.672736096784759e-05,
+      "loss": 0.9043182134628296,
+      "step": 544
+    },
+    {
+      "epoch": 1.1617021276595745,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.662650047954073e-05,
+      "loss": 0.9864247441291809,
+      "step": 546
+    },
+    {
+      "epoch": 1.1659574468085105,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.652540600621512e-05,
+      "loss": 0.7724499106407166,
+      "step": 548
+    },
+    {
+      "epoch": 1.1702127659574468,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.642407977387093e-05,
+      "loss": 0.8427982330322266,
+      "step": 550
+    },
+    {
+      "epoch": 1.174468085106383,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.632252401361149e-05,
+      "loss": 0.9444398283958435,
+      "step": 552
+    },
+    {
+      "epoch": 1.1787234042553192,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.622074096159409e-05,
+      "loss": 0.9146491289138794,
+      "step": 554
+    },
+    {
+      "epoch": 1.1829787234042553,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.6118732858980764e-05,
+      "loss": 0.9469373226165771,
+      "step": 556
+    },
+    {
+      "epoch": 1.1872340425531915,
+      "grad_norm": 0.3046875,
+      "learning_rate": 4.6016501951888916e-05,
+      "loss": 0.7791444659233093,
+      "step": 558
+    },
+    {
+      "epoch": 1.1914893617021276,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.591405049134189e-05,
+      "loss": 0.8431161046028137,
+      "step": 560
+    },
+    {
+      "epoch": 1.195744680851064,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.5811380733219405e-05,
+      "loss": 0.9778469204902649,
+      "step": 562
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.570849493820789e-05,
+      "loss": 1.0207704305648804,
+      "step": 564
+    },
+    {
+      "epoch": 1.2042553191489362,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.560539537175068e-05,
+      "loss": 0.9259340167045593,
+      "step": 566
+    },
+    {
+      "epoch": 1.2085106382978723,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.550208430399813e-05,
+      "loss": 0.908126711845398,
+      "step": 568
+    },
+    {
+      "epoch": 1.2127659574468086,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.539856400975767e-05,
+      "loss": 0.9130632877349854,
+      "step": 570
+    },
+    {
+      "epoch": 1.2170212765957447,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.529483676844366e-05,
+      "loss": 0.9781907796859741,
+      "step": 572
+    },
+    {
+      "epoch": 1.2212765957446807,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.519090486402727e-05,
+      "loss": 1.0180959701538086,
+      "step": 574
+    },
+    {
+      "epoch": 1.225531914893617,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.5086770584986135e-05,
+      "loss": 0.8333351016044617,
+      "step": 576
+    },
+    {
+      "epoch": 1.2297872340425533,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.498243622425395e-05,
+      "loss": 0.7459168434143066,
+      "step": 578
+    },
+    {
+      "epoch": 1.2340425531914894,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.4877904079170046e-05,
+      "loss": 0.7476868033409119,
+      "step": 580
+    },
+    {
+      "epoch": 1.2382978723404254,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.477317645142874e-05,
+      "loss": 0.7878999710083008,
+      "step": 582
+    },
+    {
+      "epoch": 1.2425531914893617,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.4668255647028706e-05,
+      "loss": 0.8205560445785522,
+      "step": 584
+    },
+    {
+      "epoch": 1.2468085106382978,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.456314397622217e-05,
+      "loss": 0.9549564719200134,
+      "step": 586
+    },
+    {
+      "epoch": 1.251063829787234,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.445784375346406e-05,
+      "loss": 0.7480695247650146,
+      "step": 588
+    },
+    {
+      "epoch": 1.2553191489361701,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.435235729736101e-05,
+      "loss": 1.0009480714797974,
+      "step": 590
+    },
+    {
+      "epoch": 1.2595744680851064,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.4246686930620326e-05,
+      "loss": 0.7730486989021301,
+      "step": 592
+    },
+    {
+      "epoch": 1.2638297872340425,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.4140834979998853e-05,
+      "loss": 1.0018788576126099,
+      "step": 594
+    },
+    {
+      "epoch": 1.2680851063829788,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.403480377625176e-05,
+      "loss": 0.9256978631019592,
+      "step": 596
+    },
+    {
+      "epoch": 1.2723404255319148,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.392859565408114e-05,
+      "loss": 0.9084795713424683,
+      "step": 598
+    },
+    {
+      "epoch": 1.2765957446808511,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.38222129520847e-05,
+      "loss": 1.0896780490875244,
+      "step": 600
+    },
+    {
+      "epoch": 1.2808510638297872,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.3715658012704184e-05,
+      "loss": 0.992649495601654,
+      "step": 602
+    },
+    {
+      "epoch": 1.2851063829787235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.360893318217386e-05,
+      "loss": 0.8589400053024292,
+      "step": 604
+    },
+    {
+      "epoch": 1.2893617021276595,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.350204081046885e-05,
+      "loss": 1.02162766456604,
+      "step": 606
+    },
+    {
+      "epoch": 1.2936170212765958,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.3394983251253303e-05,
+      "loss": 0.83441162109375,
+      "step": 608
+    },
+    {
+      "epoch": 1.297872340425532,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.328776286182869e-05,
+      "loss": 1.0881268978118896,
+      "step": 610
+    },
+    {
+      "epoch": 1.302127659574468,
+      "grad_norm": 0.375,
+      "learning_rate": 4.3180382003081806e-05,
+      "loss": 0.9433637857437134,
+      "step": 612
+    },
+    {
+      "epoch": 1.3063829787234043,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.307284303943286e-05,
+      "loss": 1.0009987354278564,
+      "step": 614
+    },
+    {
+      "epoch": 1.3106382978723405,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.296514833878333e-05,
+      "loss": 0.9870180487632751,
+      "step": 616
+    },
+    {
+      "epoch": 1.3148936170212766,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.2857300272463896e-05,
+      "loss": 0.7985630631446838,
+      "step": 618
+    },
+    {
+      "epoch": 1.3191489361702127,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.274930121518221e-05,
+      "loss": 0.9055668115615845,
+      "step": 620
+    },
+    {
+      "epoch": 1.323404255319149,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.264115354497057e-05,
+      "loss": 0.9184504747390747,
+      "step": 622
+    },
+    {
+      "epoch": 1.327659574468085,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.253285964313358e-05,
+      "loss": 0.8388556241989136,
+      "step": 624
+    },
+    {
+      "epoch": 1.3319148936170213,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.2424421894195746e-05,
+      "loss": 0.9362179040908813,
+      "step": 626
+    },
+    {
+      "epoch": 1.3361702127659574,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.2315842685848914e-05,
+      "loss": 0.9783342480659485,
+      "step": 628
+    },
+    {
+      "epoch": 1.3404255319148937,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.220712440889975e-05,
+      "loss": 1.012178659439087,
+      "step": 630
+    },
+    {
+      "epoch": 1.3446808510638297,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.2098269457217074e-05,
+      "loss": 0.9540913701057434,
+      "step": 632
+    },
+    {
+      "epoch": 1.348936170212766,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.1989280227679136e-05,
+      "loss": 0.9662917852401733,
+      "step": 634
+    },
+    {
+      "epoch": 1.353191489361702,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.188015912012085e-05,
+      "loss": 0.9110101461410522,
+      "step": 636
+    },
+    {
+      "epoch": 1.3574468085106384,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.177090853728096e-05,
+      "loss": 0.9384497404098511,
+      "step": 638
+    },
+    {
+      "epoch": 1.3617021276595744,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.1661530884749125e-05,
+      "loss": 0.7670997977256775,
+      "step": 640
+    },
+    {
+      "epoch": 1.3659574468085105,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.155202857091296e-05,
+      "loss": 0.8632846474647522,
+      "step": 642
+    },
+    {
+      "epoch": 1.3702127659574468,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.144240400690499e-05,
+      "loss": 0.6109669804573059,
+      "step": 644
+    },
+    {
+      "epoch": 1.374468085106383,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.133265960654956e-05,
+      "loss": 0.9433888792991638,
+      "step": 646
+    },
+    {
+      "epoch": 1.3787234042553191,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.122279778630972e-05,
+      "loss": 0.9552747011184692,
+      "step": 648
+    },
+    {
+      "epoch": 1.3829787234042552,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.1112820965233954e-05,
+      "loss": 1.0235886573791504,
+      "step": 650
+    },
+    {
+      "epoch": 1.3872340425531915,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.100273156490298e-05,
+      "loss": 0.9218517541885376,
+      "step": 652
+    },
+    {
+      "epoch": 1.3914893617021278,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.089253200937639e-05,
+      "loss": 0.9519026875495911,
+      "step": 654
+    },
+    {
+      "epoch": 1.3957446808510638,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.078222472513928e-05,
+      "loss": 1.0031726360321045,
+      "step": 656
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.067181214104883e-05,
+      "loss": 0.9014978408813477,
+      "step": 658
+    },
+    {
+      "epoch": 1.4042553191489362,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.05612966882808e-05,
+      "loss": 0.8434354066848755,
+      "step": 660
+    },
+    {
+      "epoch": 1.4085106382978723,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.0450680800276065e-05,
+      "loss": 0.9269939064979553,
+      "step": 662
+    },
+    {
+      "epoch": 1.4127659574468086,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.033996691268693e-05,
+      "loss": 1.1910121440887451,
+      "step": 664
+    },
+    {
+      "epoch": 1.4170212765957446,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.022915746332358e-05,
+      "loss": 0.9383742213249207,
+      "step": 666
+    },
+    {
+      "epoch": 1.421276595744681,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.011825489210038e-05,
+      "loss": 0.9260106086730957,
+      "step": 668
+    },
+    {
+      "epoch": 1.425531914893617,
+      "grad_norm": 0.5,
+      "learning_rate": 4.000726164098213e-05,
+      "loss": 0.8756862282752991,
+      "step": 670
+    },
+    {
+      "epoch": 1.4297872340425533,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.9896180153930326e-05,
+      "loss": 0.9409113526344299,
+      "step": 672
+    },
+    {
+      "epoch": 1.4340425531914893,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.978501287684933e-05,
+      "loss": 0.9107382893562317,
+      "step": 674
+    },
+    {
+      "epoch": 1.4382978723404256,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.9673762257532496e-05,
+      "loss": 0.9422937035560608,
+      "step": 676
+    },
+    {
+      "epoch": 1.4425531914893617,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.9562430745608315e-05,
+      "loss": 0.9622249007225037,
+      "step": 678
+    },
+    {
+      "epoch": 1.4468085106382977,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.945102079248645e-05,
+      "loss": 0.9257479906082153,
+      "step": 680
+    },
+    {
+      "epoch": 1.451063829787234,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.9339534851303746e-05,
+      "loss": 0.8421862125396729,
+      "step": 682
+    },
+    {
+      "epoch": 1.4553191489361703,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.9227975376870264e-05,
+      "loss": 0.8206483125686646,
+      "step": 684
+    },
+    {
+      "epoch": 1.4595744680851064,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.911634482561514e-05,
+      "loss": 0.8526564836502075,
+      "step": 686
+    },
+    {
+      "epoch": 1.4638297872340424,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.900464565553259e-05,
+      "loss": 1.036257266998291,
+      "step": 688
+    },
+    {
+      "epoch": 1.4680851063829787,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.889288032612775e-05,
+      "loss": 0.6551811695098877,
+      "step": 690
+    },
+    {
+      "epoch": 1.472340425531915,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.878105129836252e-05,
+      "loss": 0.9458101391792297,
+      "step": 692
+    },
+    {
+      "epoch": 1.476595744680851,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.8669161034601336e-05,
+      "loss": 0.8864946961402893,
+      "step": 694
+    },
+    {
+      "epoch": 1.4808510638297872,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.8557211998557025e-05,
+      "loss": 0.9007453322410583,
+      "step": 696
+    },
+    {
+      "epoch": 1.4851063829787234,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.844520665523654e-05,
+      "loss": 0.7618290185928345,
+      "step": 698
+    },
+    {
+      "epoch": 1.4893617021276595,
+      "grad_norm": 0.625,
+      "learning_rate": 3.8333147470886604e-05,
+      "loss": 0.7166705131530762,
+      "step": 700
+    },
+    {
+      "epoch": 1.4936170212765958,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.822103691293953e-05,
+      "loss": 0.9165350794792175,
+      "step": 702
+    },
+    {
+      "epoch": 1.4978723404255319,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.810887744995878e-05,
+      "loss": 0.6734737157821655,
+      "step": 704
+    },
+    {
+      "epoch": 1.5021276595744681,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.7996671551584686e-05,
+      "loss": 1.038259506225586,
+      "step": 706
+    },
+    {
+      "epoch": 1.5063829787234042,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.788442168848002e-05,
+      "loss": 1.081926941871643,
+      "step": 708
+    },
+    {
+      "epoch": 1.5106382978723403,
+      "grad_norm": 0.375,
+      "learning_rate": 3.777213033227562e-05,
+      "loss": 0.9264968633651733,
+      "step": 710
+    },
+    {
+      "epoch": 1.5148936170212766,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.765979995551599e-05,
+      "loss": 0.9294235110282898,
+      "step": 712
+    },
+    {
+      "epoch": 1.5191489361702128,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.7547433031604774e-05,
+      "loss": 0.9344536662101746,
+      "step": 714
+    },
+    {
+      "epoch": 1.523404255319149,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.7435032034750385e-05,
+      "loss": 0.9146227240562439,
+      "step": 716
+    },
+    {
+      "epoch": 1.527659574468085,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.73225994399115e-05,
+      "loss": 0.9178793430328369,
+      "step": 718
+    },
+    {
+      "epoch": 1.5319148936170213,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.721013772274251e-05,
+      "loss": 0.8365082144737244,
+      "step": 720
+    },
+    {
+      "epoch": 1.5361702127659576,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.7097649359539075e-05,
+      "loss": 0.8762301802635193,
+      "step": 722
+    },
+    {
+      "epoch": 1.5404255319148936,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.6985136827183575e-05,
+      "loss": 0.8902574181556702,
+      "step": 724
+    },
+    {
+      "epoch": 1.5446808510638297,
+      "grad_norm": 0.5,
+      "learning_rate": 3.6872602603090564e-05,
+      "loss": 0.9932896494865417,
+      "step": 726
+    },
+    {
+      "epoch": 1.548936170212766,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.6760049165152256e-05,
+      "loss": 0.896535336971283,
+      "step": 728
+    },
+    {
+      "epoch": 1.5531914893617023,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.6647478991683885e-05,
+      "loss": 1.0637593269348145,
+      "step": 730
+    },
+    {
+      "epoch": 1.5574468085106383,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.6534894561369214e-05,
+      "loss": 0.7936927676200867,
+      "step": 732
+    },
+    {
+      "epoch": 1.5617021276595744,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.642229835320593e-05,
+      "loss": 0.8339595794677734,
+      "step": 734
+    },
+    {
+      "epoch": 1.5659574468085107,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.630969284645103e-05,
+      "loss": 0.9330006837844849,
+      "step": 736
+    },
+    {
+      "epoch": 1.570212765957447,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.6197080520566315e-05,
+      "loss": 0.8958064913749695,
+      "step": 738
+    },
+    {
+      "epoch": 1.574468085106383,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.6084463855163666e-05,
+      "loss": 0.8008121848106384,
+      "step": 740
+    },
+    {
+      "epoch": 1.578723404255319,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.597184532995055e-05,
+      "loss": 0.950303316116333,
+      "step": 742
+    },
+    {
+      "epoch": 1.5829787234042554,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.58592274246754e-05,
+      "loss": 0.9128738641738892,
+      "step": 744
+    },
+    {
+      "epoch": 1.5872340425531914,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.574661261907297e-05,
+      "loss": 0.8636385798454285,
+      "step": 746
+    },
+    {
+      "epoch": 1.5914893617021275,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.563400339280979e-05,
+      "loss": 0.8232119083404541,
+      "step": 748
+    },
+    {
+      "epoch": 1.5957446808510638,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.5521402225429485e-05,
+      "loss": 0.8971787095069885,
+      "step": 750
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.540881159629831e-05,
+      "loss": 0.9066179394721985,
+      "step": 752
+    },
+    {
+      "epoch": 1.6042553191489362,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.529623398455042e-05,
+      "loss": 0.8429763913154602,
+      "step": 754
+    },
+    {
+      "epoch": 1.6085106382978722,
+      "grad_norm": 1.25,
+      "learning_rate": 3.5183671869033355e-05,
+      "loss": 0.8769048452377319,
+      "step": 756
+    },
+    {
+      "epoch": 1.6127659574468085,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.507112772825345e-05,
+      "loss": 1.064902424812317,
+      "step": 758
+    },
+    {
+      "epoch": 1.6170212765957448,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.4958604040321254e-05,
+      "loss": 1.0025050640106201,
+      "step": 760
+    },
+    {
+      "epoch": 1.6212765957446809,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.4846103282896956e-05,
+      "loss": 0.9082741737365723,
+      "step": 762
+    },
+    {
+      "epoch": 1.625531914893617,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.473362793313583e-05,
+      "loss": 1.00468909740448,
+      "step": 764
+    },
+    {
+      "epoch": 1.6297872340425532,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.4621180467633736e-05,
+      "loss": 1.1201874017715454,
+      "step": 766
+    },
+    {
+      "epoch": 1.6340425531914895,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.4508763362372495e-05,
+      "loss": 0.8420827388763428,
+      "step": 768
+    },
+    {
+      "epoch": 1.6382978723404256,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.4396379092665476e-05,
+      "loss": 0.9371001720428467,
+      "step": 770
+    },
+    {
+      "epoch": 1.6425531914893616,
+      "grad_norm": 0.375,
+      "learning_rate": 3.428403013310303e-05,
+      "loss": 1.0956753492355347,
+      "step": 772
+    },
+    {
+      "epoch": 1.646808510638298,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.4171718957497976e-05,
+      "loss": 1.0000417232513428,
+      "step": 774
+    },
+    {
+      "epoch": 1.6510638297872342,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.405944803883121e-05,
+      "loss": 0.7758415937423706,
+      "step": 776
+    },
+    {
+      "epoch": 1.65531914893617,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.394721984919721e-05,
+      "loss": 0.9292746186256409,
+      "step": 778
+    },
+    {
+      "epoch": 1.6595744680851063,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.383503685974956e-05,
+      "loss": 0.8925681114196777,
+      "step": 780
+    },
+    {
+      "epoch": 1.6638297872340426,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.3722901540646634e-05,
+      "loss": 0.9213247299194336,
+      "step": 782
+    },
+    {
+      "epoch": 1.6680851063829787,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.361081636099712e-05,
+      "loss": 1.0045268535614014,
+      "step": 784
+    },
+    {
+      "epoch": 1.6723404255319148,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.34987837888057e-05,
+      "loss": 1.0464414358139038,
+      "step": 786
+    },
+    {
+      "epoch": 1.676595744680851,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.338680629091867e-05,
+      "loss": 0.7823745608329773,
+      "step": 788
+    },
+    {
+      "epoch": 1.6808510638297873,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.327488633296967e-05,
+      "loss": 1.0422827005386353,
+      "step": 790
+    },
+    {
+      "epoch": 1.6851063829787234,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.316302637932537e-05,
+      "loss": 0.8115458488464355,
+      "step": 792
+    },
+    {
+      "epoch": 1.6893617021276595,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.305122889303116e-05,
+      "loss": 1.0222609043121338,
+      "step": 794
+    },
+    {
+      "epoch": 1.6936170212765957,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.2939496335757004e-05,
+      "loss": 0.9007856249809265,
+      "step": 796
+    },
+    {
+      "epoch": 1.697872340425532,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.28278311677432e-05,
+      "loss": 0.9559044241905212,
+      "step": 798
+    },
+    {
+      "epoch": 1.702127659574468,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.271623584774616e-05,
+      "loss": 0.9334125518798828,
+      "step": 800
+    },
+    {
+      "epoch": 1.7063829787234042,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.260471283298434e-05,
+      "loss": 0.9849218130111694,
+      "step": 802
+    },
+    {
+      "epoch": 1.7106382978723405,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.249326457908411e-05,
+      "loss": 0.8784253001213074,
+      "step": 804
+    },
+    {
+      "epoch": 1.7148936170212767,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.2381893540025676e-05,
+      "loss": 0.9033365845680237,
+      "step": 806
+    },
+    {
+      "epoch": 1.7191489361702128,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.227060216808902e-05,
+      "loss": 1.0640370845794678,
+      "step": 808
+    },
+    {
+      "epoch": 1.7234042553191489,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.2159392913799974e-05,
+      "loss": 0.8058955669403076,
+      "step": 810
+    },
+    {
+      "epoch": 1.7276595744680852,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.2048268225876204e-05,
+      "loss": 0.9892784357070923,
+      "step": 812
+    },
+    {
+      "epoch": 1.7319148936170212,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.1937230551173295e-05,
+      "loss": 0.7961447834968567,
+      "step": 814
+    },
+    {
+      "epoch": 1.7361702127659573,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.182628233463091e-05,
+      "loss": 0.9170913100242615,
+      "step": 816
+    },
+    {
+      "epoch": 1.7404255319148936,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1715426019218936e-05,
+      "loss": 0.9619283080101013,
+      "step": 818
+    },
+    {
+      "epoch": 1.7446808510638299,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.1604664045883653e-05,
+      "loss": 0.9417982697486877,
+      "step": 820
+    },
+    {
+      "epoch": 1.748936170212766,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.1493998853494055e-05,
+      "loss": 0.8605011105537415,
+      "step": 822
+    },
+    {
+      "epoch": 1.753191489361702,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.1383432878788086e-05,
+      "loss": 0.9855290651321411,
+      "step": 824
+    },
+    {
+      "epoch": 1.7574468085106383,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.127296855631906e-05,
+      "loss": 1.0079026222229004,
+      "step": 826
+    },
+    {
+      "epoch": 1.7617021276595746,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.116260831840196e-05,
+      "loss": 0.9938555955886841,
+      "step": 828
+    },
+    {
+      "epoch": 1.7659574468085106,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.105235459505996e-05,
+      "loss": 0.930496871471405,
+      "step": 830
+    },
+    {
+      "epoch": 1.7702127659574467,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.0942209813970894e-05,
+      "loss": 0.9760335683822632,
+      "step": 832
+    },
+    {
+      "epoch": 1.774468085106383,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.0832176400413745e-05,
+      "loss": 0.9947340488433838,
+      "step": 834
+    },
+    {
+      "epoch": 1.7787234042553193,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.072225677721537e-05,
+      "loss": 0.994687020778656,
+      "step": 836
+    },
+    {
+      "epoch": 1.7829787234042553,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.0612453364697025e-05,
+      "loss": 0.6977970004081726,
+      "step": 838
+    },
+    {
+      "epoch": 1.7872340425531914,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.050276858062112e-05,
+      "loss": 0.9086024761199951,
+      "step": 840
+    },
+    {
+      "epoch": 1.7914893617021277,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.039320484013799e-05,
+      "loss": 0.7375553250312805,
+      "step": 842
+    },
+    {
+      "epoch": 1.795744680851064,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.028376455573274e-05,
+      "loss": 0.9830509424209595,
+      "step": 844
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.0174450137172063e-05,
+      "loss": 0.8254852890968323,
+      "step": 846
+    },
+    {
+      "epoch": 1.804255319148936,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.0065263991451205e-05,
+      "loss": 0.865581750869751,
+      "step": 848
+    },
+    {
+      "epoch": 1.8085106382978724,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.9956208522740998e-05,
+      "loss": 0.9538697600364685,
+      "step": 850
+    },
+    {
+      "epoch": 1.8127659574468085,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.9847286132334884e-05,
+      "loss": 0.9136937856674194,
+      "step": 852
+    },
+    {
+      "epoch": 1.8170212765957445,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.9738499218596033e-05,
+      "loss": 0.9238496422767639,
+      "step": 854
+    },
+    {
+      "epoch": 1.8212765957446808,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.962985017690459e-05,
+      "loss": 0.870966374874115,
+      "step": 856
+    },
+    {
+      "epoch": 1.825531914893617,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.9521341399604866e-05,
+      "loss": 0.9596646428108215,
+      "step": 858
+    },
+    {
+      "epoch": 1.8297872340425532,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.9412975275952698e-05,
+      "loss": 0.8268457651138306,
+      "step": 860
+    },
+    {
+      "epoch": 1.8340425531914892,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9304754192062825e-05,
+      "loss": 0.9566388130187988,
+      "step": 862
+    },
+    {
+      "epoch": 1.8382978723404255,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.919668053085637e-05,
+      "loss": 0.9026366472244263,
+      "step": 864
+    },
+    {
+      "epoch": 1.8425531914893618,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.908875667200835e-05,
+      "loss": 1.1470530033111572,
+      "step": 866
+    },
+    {
+      "epoch": 1.8468085106382979,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.898098499189525e-05,
+      "loss": 0.9321303367614746,
+      "step": 868
+    },
+    {
+      "epoch": 1.851063829787234,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.887336786354277e-05,
+      "loss": 0.8843462467193604,
+      "step": 870
+    },
+    {
+      "epoch": 1.8553191489361702,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.8765907656573538e-05,
+      "loss": 0.9543187022209167,
+      "step": 872
+    },
+    {
+      "epoch": 1.8595744680851065,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.8658606737154885e-05,
+      "loss": 0.9508803486824036,
+      "step": 874
+    },
+    {
+      "epoch": 1.8638297872340426,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.8551467467946817e-05,
+      "loss": 0.7834142446517944,
+      "step": 876
+    },
+    {
+      "epoch": 1.8680851063829786,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.844449220804997e-05,
+      "loss": 0.9817957282066345,
+      "step": 878
+    },
+    {
+      "epoch": 1.872340425531915,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8337683312953634e-05,
+      "loss": 0.9959681630134583,
+      "step": 880
+    },
+    {
+      "epoch": 1.8765957446808512,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.823104313448392e-05,
+      "loss": 0.8245176672935486,
+      "step": 882
+    },
+    {
+      "epoch": 1.8808510638297873,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.8124574020751983e-05,
+      "loss": 0.9558523297309875,
+      "step": 884
+    },
+    {
+      "epoch": 1.8851063829787233,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.8018278316102283e-05,
+      "loss": 0.8767422437667847,
+      "step": 886
+    },
+    {
+      "epoch": 1.8893617021276596,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.7912158361060976e-05,
+      "loss": 0.966520369052887,
+      "step": 888
+    },
+    {
+      "epoch": 1.8936170212765957,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.7806216492284407e-05,
+      "loss": 0.6318687200546265,
+      "step": 890
+    },
+    {
+      "epoch": 1.8978723404255318,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.7700455042507637e-05,
+      "loss": 0.9377596378326416,
+      "step": 892
+    },
+    {
+      "epoch": 1.902127659574468,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.759487634049306e-05,
+      "loss": 1.0316444635391235,
+      "step": 894
+    },
+    {
+      "epoch": 1.9063829787234043,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.7489482710979147e-05,
+      "loss": 0.7911099195480347,
+      "step": 896
+    },
+    {
+      "epoch": 1.9106382978723404,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.7384276474629283e-05,
+      "loss": 0.8916456699371338,
+      "step": 898
+    },
+    {
+      "epoch": 1.9148936170212765,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.7279259947980615e-05,
+      "loss": 1.0238900184631348,
+      "step": 900
+    },
+    {
+      "epoch": 1.9191489361702128,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.717443544339307e-05,
+      "loss": 0.9721688628196716,
+      "step": 902
+    },
+    {
+      "epoch": 1.923404255319149,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.7069805268998467e-05,
+      "loss": 0.8731375336647034,
+      "step": 904
+    },
+    {
+      "epoch": 1.9276595744680851,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.6965371728649632e-05,
+      "loss": 0.9867290258407593,
+      "step": 906
+    },
+    {
+      "epoch": 1.9319148936170212,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.686113712186971e-05,
+      "loss": 0.8383113741874695,
+      "step": 908
+    },
+    {
+      "epoch": 1.9361702127659575,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.6757103743801555e-05,
+      "loss": 0.9159626960754395,
+      "step": 910
+    },
+    {
+      "epoch": 1.9404255319148938,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.665327388515714e-05,
+      "loss": 0.9763681292533875,
+      "step": 912
+    },
+    {
+      "epoch": 1.9446808510638298,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.6549649832167138e-05,
+      "loss": 1.0418541431427002,
+      "step": 914
+    },
+    {
+      "epoch": 1.9489361702127659,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.644623386653059e-05,
+      "loss": 1.002961277961731,
+      "step": 916
+    },
+    {
+      "epoch": 1.9531914893617022,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.6343028265364677e-05,
+      "loss": 0.8668839335441589,
+      "step": 918
+    },
+    {
+      "epoch": 1.9574468085106385,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.6240035301154533e-05,
+      "loss": 1.0359348058700562,
+      "step": 920
+    },
+    {
+      "epoch": 1.9617021276595743,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.6137257241703254e-05,
+      "loss": 0.9270017147064209,
+      "step": 922
+    },
+    {
+      "epoch": 1.9659574468085106,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.6034696350081965e-05,
+      "loss": 0.9452486634254456,
+      "step": 924
+    },
+    {
+      "epoch": 1.9702127659574469,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.5932354884579916e-05,
+      "loss": 0.9649438261985779,
+      "step": 926
+    },
+    {
+      "epoch": 1.974468085106383,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.5830235098654857e-05,
+      "loss": 0.7984356880187988,
+      "step": 928
+    },
+    {
+      "epoch": 1.978723404255319,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.5728339240883376e-05,
+      "loss": 1.008255958557129,
+      "step": 930
+    },
+    {
+      "epoch": 1.9829787234042553,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.5626669554911353e-05,
+      "loss": 1.021424651145935,
+      "step": 932
+    },
+    {
+      "epoch": 1.9872340425531916,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.55252282794046e-05,
+      "loss": 0.8926823735237122,
+      "step": 934
+    },
+    {
+      "epoch": 1.9914893617021276,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.5424017647999574e-05,
+      "loss": 0.7046434283256531,
+      "step": 936
+    },
+    {
+      "epoch": 1.9957446808510637,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.532303988925417e-05,
+      "loss": 1.097957730293274,
+      "step": 938
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.5222297226598633e-05,
+      "loss": 0.8573867082595825,
+      "step": 940
+    },
+    {
+      "epoch": 2.0042553191489363,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.512179187828667e-05,
+      "loss": 0.6483351588249207,
+      "step": 942
+    },
+    {
+      "epoch": 2.008510638297872,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.5021526057346518e-05,
+      "loss": 0.6100251078605652,
+      "step": 944
+    },
+    {
+      "epoch": 2.0127659574468084,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.4921501971532267e-05,
+      "loss": 0.6785602569580078,
+      "step": 946
+    },
+    {
+      "epoch": 2.0170212765957447,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.4821721823275266e-05,
+      "loss": 0.6127380728721619,
+      "step": 948
+    },
+    {
+      "epoch": 2.021276595744681,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.472218780963559e-05,
+      "loss": 0.845111072063446,
+      "step": 950
+    },
+    {
+      "epoch": 2.025531914893617,
+      "grad_norm": 0.375,
+      "learning_rate": 2.462290212225365e-05,
+      "loss": 0.6631003022193909,
+      "step": 952
+    },
+    {
+      "epoch": 2.029787234042553,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.452386694730197e-05,
+      "loss": 0.5201372504234314,
+      "step": 954
+    },
+    {
+      "epoch": 2.0340425531914894,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.4425084465437053e-05,
+      "loss": 0.7710279226303101,
+      "step": 956
+    },
+    {
+      "epoch": 2.0382978723404257,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.4326556851751346e-05,
+      "loss": 0.6157738566398621,
+      "step": 958
+    },
+    {
+      "epoch": 2.0425531914893615,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.422828627572534e-05,
+      "loss": 0.7093011736869812,
+      "step": 960
+    },
+    {
+      "epoch": 2.046808510638298,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4130274901179803e-05,
+      "loss": 0.6860568523406982,
+      "step": 962
+    },
+    {
+      "epoch": 2.051063829787234,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.4032524886228184e-05,
+      "loss": 0.719150960445404,
+      "step": 964
+    },
+    {
+      "epoch": 2.0553191489361704,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.3935038383229e-05,
+      "loss": 0.6379442811012268,
+      "step": 966
+    },
+    {
+      "epoch": 2.0595744680851062,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.3837817538738525e-05,
+      "loss": 0.7576701641082764,
+      "step": 968
+    },
+    {
+      "epoch": 2.0638297872340425,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.374086449346352e-05,
+      "loss": 0.4662551283836365,
+      "step": 970
+    },
+    {
+      "epoch": 2.068085106382979,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3644181382214013e-05,
+      "loss": 0.5384807586669922,
+      "step": 972
+    },
+    {
+      "epoch": 2.072340425531915,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.3547770333856386e-05,
+      "loss": 0.6011054515838623,
+      "step": 974
+    },
+    {
+      "epoch": 2.076595744680851,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.345163347126647e-05,
+      "loss": 0.5131646990776062,
+      "step": 976
+    },
+    {
+      "epoch": 2.0808510638297872,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3355772911282796e-05,
+      "loss": 0.8288999199867249,
+      "step": 978
+    },
+    {
+      "epoch": 2.0851063829787235,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.3260190764659976e-05,
+      "loss": 0.47640979290008545,
+      "step": 980
+    },
+    {
+      "epoch": 2.0893617021276594,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.316488913602222e-05,
+      "loss": 0.5666584372520447,
+      "step": 982
+    },
+    {
+      "epoch": 2.0936170212765957,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.3069870123817056e-05,
+      "loss": 0.5758649110794067,
+      "step": 984
+    },
+    {
+      "epoch": 2.097872340425532,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.2975135820269026e-05,
+      "loss": 0.5656971335411072,
+      "step": 986
+    },
+    {
+      "epoch": 2.1021276595744682,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2880688311333702e-05,
+      "loss": 0.3263399302959442,
+      "step": 988
+    },
+    {
+      "epoch": 2.106382978723404,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.278652967665173e-05,
+      "loss": 0.690028965473175,
+      "step": 990
+    },
+    {
+      "epoch": 2.1106382978723404,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.2692661989502995e-05,
+      "loss": 0.6520633101463318,
+      "step": 992
+    },
+    {
+      "epoch": 2.1148936170212767,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.2599087316761018e-05,
+      "loss": 0.7235679626464844,
+      "step": 994
+    },
+    {
+      "epoch": 2.119148936170213,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.2505807718847456e-05,
+      "loss": 0.6303759217262268,
+      "step": 996
+    },
+    {
+      "epoch": 2.123404255319149,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.24128252496867e-05,
+      "loss": 0.6719092130661011,
+      "step": 998
+    },
+    {
+      "epoch": 2.127659574468085,
+      "grad_norm": 0.5,
+      "learning_rate": 2.2320141956660646e-05,
+      "loss": 0.28417664766311646,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1319148936170214,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2227759880563626e-05,
+      "loss": 0.7599141597747803,
+      "step": 1002
+    },
+    {
+      "epoch": 2.1361702127659576,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2135681055557504e-05,
+      "loss": 0.6200002431869507,
+      "step": 1004
+    },
+    {
+      "epoch": 2.1404255319148935,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.2043907509126812e-05,
+      "loss": 0.7208251953125,
+      "step": 1006
+    },
+    {
+      "epoch": 2.1446808510638298,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.1952441262034188e-05,
+      "loss": 0.5411021709442139,
+      "step": 1008
+    },
+    {
+      "epoch": 2.148936170212766,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.1861284328275845e-05,
+      "loss": 0.5026164650917053,
+      "step": 1010
+    },
+    {
+      "epoch": 2.153191489361702,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.1770438715037165e-05,
+      "loss": 0.6443688273429871,
+      "step": 1012
+    },
+    {
+      "epoch": 2.157446808510638,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1679906422648626e-05,
+      "loss": 0.6200368404388428,
+      "step": 1014
+    },
+    {
+      "epoch": 2.1617021276595745,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1589689444541662e-05,
+      "loss": 0.6573978066444397,
+      "step": 1016
+    },
+    {
+      "epoch": 2.1659574468085108,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.1499789767204812e-05,
+      "loss": 0.6872032880783081,
+      "step": 1018
+    },
+    {
+      "epoch": 2.1702127659574466,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.1410209370139945e-05,
+      "loss": 0.5005927085876465,
+      "step": 1020
+    },
+    {
+      "epoch": 2.174468085106383,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1320950225818707e-05,
+      "loss": 0.7184480428695679,
+      "step": 1022
+    },
+    {
+      "epoch": 2.178723404255319,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1232014299639085e-05,
+      "loss": 0.6513587832450867,
+      "step": 1024
+    },
+    {
+      "epoch": 2.1829787234042555,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.11434035498821e-05,
+      "loss": 0.5126093626022339,
+      "step": 1026
+    },
+    {
+      "epoch": 2.1872340425531913,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.105511992766874e-05,
+      "loss": 0.6408154964447021,
+      "step": 1028
+    },
+    {
+      "epoch": 2.1914893617021276,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.0967165376916945e-05,
+      "loss": 0.7048395872116089,
+      "step": 1030
+    },
+    {
+      "epoch": 2.195744680851064,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.0879541834298828e-05,
+      "loss": 0.662340521812439,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.0792251229198035e-05,
+      "loss": 0.7315186858177185,
+      "step": 1034
+    },
+    {
+      "epoch": 2.204255319148936,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.0705295483667274e-05,
+      "loss": 0.47347530722618103,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2085106382978723,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.0618676512385965e-05,
+      "loss": 0.5597242712974548,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2127659574468086,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.0532396222618083e-05,
+      "loss": 0.6009190082550049,
+      "step": 1040
+    },
+    {
+      "epoch": 2.217021276595745,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.0446456514170166e-05,
+      "loss": 0.7078320384025574,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2212765957446807,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.0360859279349523e-05,
+      "loss": 0.6109620928764343,
+      "step": 1044
+    },
+    {
+      "epoch": 2.225531914893617,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.0275606402922496e-05,
+      "loss": 0.6881995797157288,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2297872340425533,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.0190699762073015e-05,
+      "loss": 0.8076979517936707,
+      "step": 1048
+    },
+    {
+      "epoch": 2.2340425531914896,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.010614122636125e-05,
+      "loss": 0.5962254405021667,
+      "step": 1050
+    },
+    {
+      "epoch": 2.2382978723404254,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.002193265768241e-05,
+      "loss": 0.5154112577438354,
+      "step": 1052
+    },
+    {
+      "epoch": 2.2425531914893617,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.9938075910225816e-05,
+      "loss": 0.5857576131820679,
+      "step": 1054
+    },
+    {
+      "epoch": 2.246808510638298,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.985457283043401e-05,
+      "loss": 0.7445710301399231,
+      "step": 1056
+    },
+    {
+      "epoch": 2.251063829787234,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.977142525696214e-05,
+      "loss": 0.4012032449245453,
+      "step": 1058
+    },
+    {
+      "epoch": 2.25531914893617,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9688635020637438e-05,
+      "loss": 0.7297635078430176,
+      "step": 1060
+    },
+    {
+      "epoch": 2.2595744680851064,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9606203944418924e-05,
+      "loss": 0.5968733429908752,
+      "step": 1062
+    },
+    {
+      "epoch": 2.2638297872340427,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.9524133843357294e-05,
+      "loss": 0.7591136693954468,
+      "step": 1064
+    },
+    {
+      "epoch": 2.2680851063829786,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9442426524554893e-05,
+      "loss": 0.6319488286972046,
+      "step": 1066
+    },
+    {
+      "epoch": 2.272340425531915,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9361083787126e-05,
+      "loss": 0.6187411546707153,
+      "step": 1068
+    },
+    {
+      "epoch": 2.276595744680851,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.9280107422157143e-05,
+      "loss": 0.6891050934791565,
+      "step": 1070
+    },
+    {
+      "epoch": 2.2808510638297874,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9199499212667688e-05,
+      "loss": 0.507449209690094,
+      "step": 1072
+    },
+    {
+      "epoch": 2.2851063829787233,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9119260933570603e-05,
+      "loss": 0.563790500164032,
+      "step": 1074
+    },
+    {
+      "epoch": 2.2893617021276595,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.903939435163335e-05,
+      "loss": 0.5918843150138855,
+      "step": 1076
+    },
+    {
+      "epoch": 2.293617021276596,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8959901225439e-05,
+      "loss": 0.6510270833969116,
+      "step": 1078
+    },
+    {
+      "epoch": 2.297872340425532,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.888078330534744e-05,
+      "loss": 0.4389875531196594,
+      "step": 1080
+    },
+    {
+      "epoch": 2.302127659574468,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.880204233345696e-05,
+      "loss": 0.3430854082107544,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3063829787234043,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8723680043565798e-05,
+      "loss": 0.7696226835250854,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3106382978723405,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.8645698161133972e-05,
+      "loss": 0.5286341905593872,
+      "step": 1086
+    },
+    {
+      "epoch": 2.3148936170212764,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8568098403245336e-05,
+      "loss": 0.567283570766449,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3191489361702127,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.8490882478569716e-05,
+      "loss": 0.5737777352333069,
+      "step": 1090
+    },
+    {
+      "epoch": 2.323404255319149,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8414052087325308e-05,
+      "loss": 0.728583037853241,
+      "step": 1092
+    },
+    {
+      "epoch": 2.3276595744680852,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8337608921241267e-05,
+      "loss": 0.600398600101471,
+      "step": 1094
+    },
+    {
+      "epoch": 2.331914893617021,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8261554663520416e-05,
+      "loss": 0.6284059286117554,
+      "step": 1096
+    },
+    {
+      "epoch": 2.3361702127659574,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8185890988802214e-05,
+      "loss": 0.6262116432189941,
+      "step": 1098
+    },
+    {
+      "epoch": 2.3404255319148937,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.8110619563125844e-05,
+      "loss": 0.5292909741401672,
+      "step": 1100
+    },
+    {
+      "epoch": 2.34468085106383,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8035742043893575e-05,
+      "loss": 0.625741720199585,
+      "step": 1102
+    },
+    {
+      "epoch": 2.348936170212766,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.796126007983425e-05,
+      "loss": 0.7988000512123108,
+      "step": 1104
+    },
+    {
+      "epoch": 2.353191489361702,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7887175310966956e-05,
+      "loss": 0.6517726182937622,
+      "step": 1106
+    },
+    {
+      "epoch": 2.3574468085106384,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.7813489368564965e-05,
+      "loss": 0.6426142454147339,
+      "step": 1108
+    },
+    {
+      "epoch": 2.3617021276595747,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.7740203875119755e-05,
+      "loss": 0.5290454626083374,
+      "step": 1110
+    },
+    {
+      "epoch": 2.3659574468085105,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7667320444305326e-05,
+      "loss": 0.5435438752174377,
+      "step": 1112
+    },
+    {
+      "epoch": 2.370212765957447,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7594840680942667e-05,
+      "loss": 0.4323336184024811,
+      "step": 1114
+    },
+    {
+      "epoch": 2.374468085106383,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.752276618096441e-05,
+      "loss": 0.5053521394729614,
+      "step": 1116
+    },
+    {
+      "epoch": 2.378723404255319,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.7451098531379666e-05,
+      "loss": 0.63356614112854,
+      "step": 1118
+    },
+    {
+      "epoch": 2.382978723404255,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7379839310239118e-05,
+      "loss": 0.7377380728721619,
+      "step": 1120
+    },
+    {
+      "epoch": 2.3872340425531915,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.7308990086600258e-05,
+      "loss": 0.7407448887825012,
+      "step": 1122
+    },
+    {
+      "epoch": 2.391489361702128,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7238552420492854e-05,
+      "loss": 0.5100683569908142,
+      "step": 1124
+    },
+    {
+      "epoch": 2.395744680851064,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.716852786288455e-05,
+      "loss": 0.6132201552391052,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.709891795564679e-05,
+      "loss": 0.5412701368331909,
+      "step": 1128
+    },
+    {
+      "epoch": 2.404255319148936,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.7029724231520792e-05,
+      "loss": 0.8436723351478577,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4085106382978725,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.696094821408385e-05,
+      "loss": 0.6744006872177124,
+      "step": 1132
+    },
+    {
+      "epoch": 2.4127659574468083,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.6892591417715775e-05,
+      "loss": 0.7410330176353455,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4170212765957446,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.682465534756555e-05,
+      "loss": 0.6016606092453003,
+      "step": 1136
+    },
+    {
+      "epoch": 2.421276595744681,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.6757141499518153e-05,
+      "loss": 0.4425470530986786,
+      "step": 1138
+    },
+    {
+      "epoch": 2.425531914893617,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6690051360161673e-05,
+      "loss": 0.7017032504081726,
+      "step": 1140
+    },
+    {
+      "epoch": 2.429787234042553,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6623386406754555e-05,
+      "loss": 0.5560771822929382,
+      "step": 1142
+    },
+    {
+      "epoch": 2.4340425531914893,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.655714810719307e-05,
+      "loss": 0.6498077511787415,
+      "step": 1144
+    },
+    {
+      "epoch": 2.4382978723404256,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.6491337919978978e-05,
+      "loss": 0.6854323148727417,
+      "step": 1146
+    },
+    {
+      "epoch": 2.4425531914893615,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.642595729418745e-05,
+      "loss": 0.4551085829734802,
+      "step": 1148
+    },
+    {
+      "epoch": 2.4468085106382977,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6361007669435126e-05,
+      "loss": 0.5184612274169922,
+      "step": 1150
+    },
+    {
+      "epoch": 2.451063829787234,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.6296490475848424e-05,
+      "loss": 0.5938658118247986,
+      "step": 1152
+    },
+    {
+      "epoch": 2.4553191489361703,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.623240713403207e-05,
+      "loss": 0.5654492974281311,
+      "step": 1154
+    },
+    {
+      "epoch": 2.4595744680851066,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.6168759055037817e-05,
+      "loss": 0.6496747732162476,
+      "step": 1156
+    },
+    {
+      "epoch": 2.4638297872340424,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.610554764033332e-05,
+      "loss": 0.7038140296936035,
+      "step": 1158
+    },
+    {
+      "epoch": 2.4680851063829787,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6042774281771345e-05,
+      "loss": 0.641715407371521,
+      "step": 1160
+    },
+    {
+      "epoch": 2.472340425531915,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.59804403615591e-05,
+      "loss": 0.5276774168014526,
+      "step": 1162
+    },
+    {
+      "epoch": 2.476595744680851,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.5918547252227793e-05,
+      "loss": 0.6699836850166321,
+      "step": 1164
+    },
+    {
+      "epoch": 2.480851063829787,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.58570963166024e-05,
+      "loss": 0.6293801069259644,
+      "step": 1166
+    },
+    {
+      "epoch": 2.4851063829787234,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5796088907771674e-05,
+      "loss": 0.6323214769363403,
+      "step": 1168
+    },
+    {
+      "epoch": 2.4893617021276597,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5735526369058364e-05,
+      "loss": 0.6430104970932007,
+      "step": 1170
+    },
+    {
+      "epoch": 2.4936170212765956,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5675410033989592e-05,
+      "loss": 0.5185415744781494,
+      "step": 1172
+    },
+    {
+      "epoch": 2.497872340425532,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.561574122626754e-05,
+      "loss": 0.47881028056144714,
+      "step": 1174
+    },
+    {
+      "epoch": 2.502127659574468,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.555652125974028e-05,
+      "loss": 0.7108798027038574,
+      "step": 1176
+    },
+    {
+      "epoch": 2.506382978723404,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.5497751438372827e-05,
+      "loss": 0.5829865336418152,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5106382978723403,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.543943305621846e-05,
+      "loss": 0.5929511785507202,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5148936170212766,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.538156739739021e-05,
+      "loss": 0.5175199508666992,
+      "step": 1182
+    },
+    {
+      "epoch": 2.519148936170213,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.5324155736032595e-05,
+      "loss": 0.7886143326759338,
+      "step": 1184
+    },
+    {
+      "epoch": 2.523404255319149,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.526719933629355e-05,
+      "loss": 0.6642839908599854,
+      "step": 1186
+    },
+    {
+      "epoch": 2.527659574468085,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5210699452296592e-05,
+      "loss": 0.8588666915893555,
+      "step": 1188
+    },
+    {
+      "epoch": 2.5319148936170213,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5154657328113233e-05,
+      "loss": 0.7579631209373474,
+      "step": 1190
+    },
+    {
+      "epoch": 2.5361702127659576,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5099074197735552e-05,
+      "loss": 0.6104500889778137,
+      "step": 1192
+    },
+    {
+      "epoch": 2.5404255319148934,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.504395128504905e-05,
+      "loss": 0.34646666049957275,
+      "step": 1194
+    },
+    {
+      "epoch": 2.5446808510638297,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4989289803805685e-05,
+      "loss": 0.5463817119598389,
+      "step": 1196
+    },
+    {
+      "epoch": 2.548936170212766,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4935090957597149e-05,
+      "loss": 0.61968594789505,
+      "step": 1198
+    },
+    {
+      "epoch": 2.5531914893617023,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4881355939828364e-05,
+      "loss": 0.6242573857307434,
+      "step": 1200
+    },
+    {
+      "epoch": 2.5574468085106385,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4828085933691223e-05,
+      "loss": 0.6220880150794983,
+      "step": 1202
+    },
+    {
+      "epoch": 2.5617021276595744,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.477528211213852e-05,
+      "loss": 0.6225618124008179,
+      "step": 1204
+    },
+    {
+      "epoch": 2.5659574468085107,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4722945637858116e-05,
+      "loss": 0.4512316584587097,
+      "step": 1206
+    },
+    {
+      "epoch": 2.570212765957447,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.4671077663247351e-05,
+      "loss": 0.5912795662879944,
+      "step": 1208
+    },
+    {
+      "epoch": 2.574468085106383,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.4619679330387679e-05,
+      "loss": 0.6551855802536011,
+      "step": 1210
+    },
+    {
+      "epoch": 2.578723404255319,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4568751771019482e-05,
+      "loss": 0.5984119772911072,
+      "step": 1212
+    },
+    {
+      "epoch": 2.5829787234042554,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4518296106517206e-05,
+      "loss": 0.6674041152000427,
+      "step": 1214
+    },
+    {
+      "epoch": 2.5872340425531917,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4468313447864624e-05,
+      "loss": 0.48521387577056885,
+      "step": 1216
+    },
+    {
+      "epoch": 2.5914893617021275,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.441880489563038e-05,
+      "loss": 0.6522884964942932,
+      "step": 1218
+    },
+    {
+      "epoch": 2.595744680851064,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.4369771539943776e-05,
+      "loss": 0.7783772945404053,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4321214460470755e-05,
+      "loss": 0.5732651948928833,
+      "step": 1222
+    },
+    {
+      "epoch": 2.604255319148936,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4273134726390138e-05,
+      "loss": 0.6476449966430664,
+      "step": 1224
+    },
+    {
+      "epoch": 2.608510638297872,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.4225533396370053e-05,
+      "loss": 0.7929537296295166,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6127659574468085,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.4178411518544654e-05,
+      "loss": 0.3998229205608368,
+      "step": 1228
+    },
+    {
+      "epoch": 2.617021276595745,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.413177013049104e-05,
+      "loss": 0.4514097273349762,
+      "step": 1230
+    },
+    {
+      "epoch": 2.621276595744681,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4085610259206387e-05,
+      "loss": 0.6864388585090637,
+      "step": 1232
+    },
+    {
+      "epoch": 2.625531914893617,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4039932921085362e-05,
+      "loss": 0.5575997233390808,
+      "step": 1234
+    },
+    {
+      "epoch": 2.629787234042553,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.3994739121897718e-05,
+      "loss": 0.653709888458252,
+      "step": 1236
+    },
+    {
+      "epoch": 2.6340425531914895,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3950029856766165e-05,
+      "loss": 0.6980820298194885,
+      "step": 1238
+    },
+    {
+      "epoch": 2.6382978723404253,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3905806110144452e-05,
+      "loss": 0.6501242518424988,
+      "step": 1240
+    },
+    {
+      "epoch": 2.6425531914893616,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3862068855795701e-05,
+      "loss": 0.7544458508491516,
+      "step": 1242
+    },
+    {
+      "epoch": 2.646808510638298,
+      "grad_norm": 1.0,
+      "learning_rate": 1.3818819056770932e-05,
+      "loss": 0.5266544222831726,
+      "step": 1244
+    },
+    {
+      "epoch": 2.651063829787234,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3776057665387907e-05,
+      "loss": 0.6442818641662598,
+      "step": 1246
+    },
+    {
+      "epoch": 2.65531914893617,
+      "grad_norm": 0.375,
+      "learning_rate": 1.373378562321012e-05,
+      "loss": 0.6095808148384094,
+      "step": 1248
+    },
+    {
+      "epoch": 2.6595744680851063,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3692003861026083e-05,
+      "loss": 0.7218018770217896,
+      "step": 1250
+    },
+    {
+      "epoch": 2.6638297872340426,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.365071329882883e-05,
+      "loss": 0.6180848479270935,
+      "step": 1252
+    },
+    {
+      "epoch": 2.6680851063829785,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.360991484579566e-05,
+      "loss": 0.4784125089645386,
+      "step": 1254
+    },
+    {
+      "epoch": 2.6723404255319148,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3569609400268112e-05,
+      "loss": 0.7386208772659302,
+      "step": 1256
+    },
+    {
+      "epoch": 2.676595744680851,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3529797849732183e-05,
+      "loss": 0.5939998626708984,
+      "step": 1258
+    },
+    {
+      "epoch": 2.6808510638297873,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.3490481070798797e-05,
+      "loss": 0.6285054683685303,
+      "step": 1260
+    },
+    {
+      "epoch": 2.6851063829787236,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.34516599291845e-05,
+      "loss": 0.7830681800842285,
+      "step": 1262
+    },
+    {
+      "epoch": 2.6893617021276595,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.3413335279692392e-05,
+      "loss": 0.7019430994987488,
+      "step": 1264
+    },
+    {
+      "epoch": 2.6936170212765957,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.3375507966193309e-05,
+      "loss": 0.5392411947250366,
+      "step": 1266
+    },
+    {
+      "epoch": 2.697872340425532,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3338178821607234e-05,
+      "loss": 0.8191847205162048,
+      "step": 1268
+    },
+    {
+      "epoch": 2.702127659574468,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3301348667884975e-05,
+      "loss": 0.30988848209381104,
+      "step": 1270
+    },
+    {
+      "epoch": 2.706382978723404,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.3265018315990046e-05,
+      "loss": 0.5801700353622437,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7106382978723405,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3229188565880827e-05,
+      "loss": 0.6832275390625,
+      "step": 1274
+    },
+    {
+      "epoch": 2.7148936170212767,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3193860206492936e-05,
+      "loss": 0.6654208898544312,
+      "step": 1276
+    },
+    {
+      "epoch": 2.719148936170213,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.3159034015721865e-05,
+      "loss": 0.7127547860145569,
+      "step": 1278
+    },
+    {
+      "epoch": 2.723404255319149,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3124710760405853e-05,
+      "loss": 0.537979781627655,
+      "step": 1280
+    },
+    {
+      "epoch": 2.727659574468085,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3090891196309e-05,
+      "loss": 0.6351765990257263,
+      "step": 1282
+    },
+    {
+      "epoch": 2.731914893617021,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3057576068104621e-05,
+      "loss": 0.601453959941864,
+      "step": 1284
+    },
+    {
+      "epoch": 2.7361702127659573,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3024766109358845e-05,
+      "loss": 0.7181084752082825,
+      "step": 1286
+    },
+    {
+      "epoch": 2.7404255319148936,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2992462042514483e-05,
+      "loss": 0.5218726396560669,
+      "step": 1288
+    },
+    {
+      "epoch": 2.74468085106383,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2960664578875104e-05,
+      "loss": 0.7979943156242371,
+      "step": 1290
+    },
+    {
+      "epoch": 2.748936170212766,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2929374418589363e-05,
+      "loss": 0.600561797618866,
+      "step": 1292
+    },
+    {
+      "epoch": 2.753191489361702,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.289859225063562e-05,
+      "loss": 0.5805540084838867,
+      "step": 1294
+    },
+    {
+      "epoch": 2.7574468085106383,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2868318752806726e-05,
+      "loss": 0.5937775373458862,
+      "step": 1296
+    },
+    {
+      "epoch": 2.7617021276595746,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2838554591695126e-05,
+      "loss": 0.4572460353374481,
+      "step": 1298
+    },
+    {
+      "epoch": 2.7659574468085104,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2809300422678187e-05,
+      "loss": 0.7637752294540405,
+      "step": 1300
+    },
+    {
+      "epoch": 2.7702127659574467,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2780556889903737e-05,
+      "loss": 0.48286503553390503,
+      "step": 1302
+    },
+    {
+      "epoch": 2.774468085106383,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.275232462627591e-05,
+      "loss": 0.5692922472953796,
+      "step": 1304
+    },
+    {
+      "epoch": 2.7787234042553193,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2724604253441195e-05,
+      "loss": 0.49659186601638794,
+      "step": 1306
+    },
+    {
+      "epoch": 2.7829787234042556,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2697396381774753e-05,
+      "loss": 0.7237148880958557,
+      "step": 1308
+    },
+    {
+      "epoch": 2.7872340425531914,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2670701610366985e-05,
+      "loss": 0.5669375658035278,
+      "step": 1310
+    },
+    {
+      "epoch": 2.7914893617021277,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2644520527010319e-05,
+      "loss": 0.49942082166671753,
+      "step": 1312
+    },
+    {
+      "epoch": 2.795744680851064,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2618853708186294e-05,
+      "loss": 0.5854433178901672,
+      "step": 1314
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.2593701719052839e-05,
+      "loss": 0.4207378625869751,
+      "step": 1316
+    },
+    {
+      "epoch": 2.804255319148936,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2569065113431854e-05,
+      "loss": 0.6822559237480164,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8085106382978724,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2544944433797002e-05,
+      "loss": 0.5026736855506897,
+      "step": 1320
+    },
+    {
+      "epoch": 2.8127659574468087,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.252134021126177e-05,
+      "loss": 0.7221361398696899,
+      "step": 1322
+    },
+    {
+      "epoch": 2.8170212765957445,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2498252965567755e-05,
+      "loss": 0.6420372724533081,
+      "step": 1324
+    },
+    {
+      "epoch": 2.821276595744681,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2475683205073255e-05,
+      "loss": 0.5234490633010864,
+      "step": 1326
+    },
+    {
+      "epoch": 2.825531914893617,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.2453631426742047e-05,
+      "loss": 0.51298987865448,
+      "step": 1328
+    },
+    {
+      "epoch": 2.829787234042553,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2432098116132458e-05,
+      "loss": 0.5665377378463745,
+      "step": 1330
+    },
+    {
+      "epoch": 2.8340425531914892,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2411083747386662e-05,
+      "loss": 0.7207722067832947,
+      "step": 1332
+    },
+    {
+      "epoch": 2.8382978723404255,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2390588783220257e-05,
+      "loss": 0.595878005027771,
+      "step": 1334
+    },
+    {
+      "epoch": 2.842553191489362,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.2370613674912055e-05,
+      "loss": 0.7001873850822449,
+      "step": 1336
+    },
+    {
+      "epoch": 2.846808510638298,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2351158862294165e-05,
+      "loss": 0.5404136776924133,
+      "step": 1338
+    },
+    {
+      "epoch": 2.851063829787234,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2332224773742298e-05,
+      "loss": 0.5846556425094604,
+      "step": 1340
+    },
+    {
+      "epoch": 2.8553191489361702,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.231381182616634e-05,
+      "loss": 0.4478878080844879,
+      "step": 1342
+    },
+    {
+      "epoch": 2.8595744680851065,
+      "grad_norm": 0.625,
+      "learning_rate": 1.229592042500116e-05,
+      "loss": 0.49717509746551514,
+      "step": 1344
+    },
+    {
+      "epoch": 2.8638297872340424,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2278550964197693e-05,
+      "loss": 0.650644063949585,
+      "step": 1346
+    },
+    {
+      "epoch": 2.8680851063829786,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2261703826214275e-05,
+      "loss": 0.6666759252548218,
+      "step": 1348
+    },
+    {
+      "epoch": 2.872340425531915,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2245379382008206e-05,
+      "loss": 0.7478127479553223,
+      "step": 1350
+    },
+    {
+      "epoch": 2.876595744680851,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.2229577991027574e-05,
+      "loss": 0.7170307636260986,
+      "step": 1352
+    },
+    {
+      "epoch": 2.8808510638297875,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2214300001203369e-05,
+      "loss": 0.8161506056785583,
+      "step": 1354
+    },
+    {
+      "epoch": 2.8851063829787233,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2199545748941797e-05,
+      "loss": 0.49168941378593445,
+      "step": 1356
+    },
+    {
+      "epoch": 2.8893617021276596,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2185315559116883e-05,
+      "loss": 0.46535876393318176,
+      "step": 1358
+    },
+    {
+      "epoch": 2.8936170212765955,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.217160974506332e-05,
+      "loss": 0.770456850528717,
+      "step": 1360
+    },
+    {
+      "epoch": 2.8978723404255318,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2158428608569563e-05,
+      "loss": 0.6479332447052002,
+      "step": 1362
+    },
+    {
+      "epoch": 2.902127659574468,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2145772439871186e-05,
+      "loss": 0.5973125100135803,
+      "step": 1364
+    },
+    {
+      "epoch": 2.9063829787234043,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2133641517644488e-05,
+      "loss": 0.5936951637268066,
+      "step": 1366
+    },
+    {
+      "epoch": 2.9106382978723406,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2122036109000369e-05,
+      "loss": 0.5945168733596802,
+      "step": 1368
+    },
+    {
+      "epoch": 2.9148936170212765,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2110956469478434e-05,
+      "loss": 0.799973726272583,
+      "step": 1370
+    },
+    {
+      "epoch": 2.9191489361702128,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2100402843041378e-05,
+      "loss": 0.5132302641868591,
+      "step": 1372
+    },
+    {
+      "epoch": 2.923404255319149,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2090375462069602e-05,
+      "loss": 0.6212731599807739,
+      "step": 1374
+    },
+    {
+      "epoch": 2.927659574468085,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.208087454735611e-05,
+      "loss": 0.6145610809326172,
+      "step": 1376
+    },
+    {
+      "epoch": 2.931914893617021,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2071900308101635e-05,
+      "loss": 0.7530128359794617,
+      "step": 1378
+    },
+    {
+      "epoch": 2.9361702127659575,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2063452941910037e-05,
+      "loss": 0.5762054324150085,
+      "step": 1380
+    },
+    {
+      "epoch": 2.9404255319148938,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.205553263478396e-05,
+      "loss": 0.7723484039306641,
+      "step": 1382
+    },
+    {
+      "epoch": 2.94468085106383,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2048139561120719e-05,
+      "loss": 0.6207926869392395,
+      "step": 1384
+    },
+    {
+      "epoch": 2.948936170212766,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2041273883708483e-05,
+      "loss": 0.6677027940750122,
+      "step": 1386
+    },
+    {
+      "epoch": 2.953191489361702,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2034935753722666e-05,
+      "loss": 0.7704895734786987,
+      "step": 1388
+    },
+    {
+      "epoch": 2.9574468085106385,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2029125310722613e-05,
+      "loss": 0.7197574377059937,
+      "step": 1390
+    },
+    {
+      "epoch": 2.9617021276595743,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.202384268264853e-05,
+      "loss": 0.7283585667610168,
+      "step": 1392
+    },
+    {
+      "epoch": 2.9659574468085106,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2019087985818655e-05,
+      "loss": 0.5656808018684387,
+      "step": 1394
+    },
+    {
+      "epoch": 2.970212765957447,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2014861324926705e-05,
+      "loss": 0.49777036905288696,
+      "step": 1396
+    },
+    {
+      "epoch": 2.974468085106383,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2011162793039564e-05,
+      "loss": 0.6693958640098572,
+      "step": 1398
+    },
+    {
+      "epoch": 2.978723404255319,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2007992471595247e-05,
+      "loss": 0.5529562830924988,
+      "step": 1400
+    },
+    {
+      "epoch": 2.9829787234042553,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.2005350430401092e-05,
+      "loss": 0.5693247318267822,
+      "step": 1402
+    },
+    {
+      "epoch": 2.9872340425531916,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2003236727632224e-05,
+      "loss": 0.5829908847808838,
+      "step": 1404
+    },
+    {
+      "epoch": 2.9914893617021274,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.2001651409830295e-05,
+      "loss": 0.6101734638214111,
+      "step": 1406
+    },
+    {
+      "epoch": 2.9957446808510637,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2000594511902426e-05,
+      "loss": 0.6352876424789429,
+      "step": 1408
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2000066057120467e-05,
+      "loss": 0.3924168646335602,
+      "step": 1410
+    },
+    {
+      "epoch": 3.0,
+      "step": 1410,
+      "total_flos": 4.1743170019314893e+18,
+      "train_loss": 0.9459603985573383,
+      "train_runtime": 10552.9203,
+      "train_samples_per_second": 4.276,
+      "train_steps_per_second": 0.134
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1410,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.1743170019314893e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}