diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9471 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9941650458460685,
+  "eval_steps": 500,
+  "global_step": 1347,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022228396776882466,
+      "grad_norm": 0.31254997849464417,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 1.8161,
+      "step": 1
+    },
+    {
+      "epoch": 0.004445679355376493,
+      "grad_norm": 0.3643212616443634,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.8137,
+      "step": 2
+    },
+    {
+      "epoch": 0.0066685190330647405,
+      "grad_norm": 0.3575803339481354,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.7748,
+      "step": 3
+    },
+    {
+      "epoch": 0.008891358710752986,
+      "grad_norm": 0.31876519322395325,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 1.7532,
+      "step": 4
+    },
+    {
+      "epoch": 0.011114198388441235,
+      "grad_norm": 0.3756287693977356,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 1.7228,
+      "step": 5
+    },
+    {
+      "epoch": 0.013337038066129481,
+      "grad_norm": 0.39089882373809814,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 2.0789,
+      "step": 6
+    },
+    {
+      "epoch": 0.015559877743817728,
+      "grad_norm": 0.3725750148296356,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 1.71,
+      "step": 7
+    },
+    {
+      "epoch": 0.017782717421505972,
+      "grad_norm": 0.3968382179737091,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 1.8689,
+      "step": 8
+    },
+    {
+      "epoch": 0.02000555709919422,
+      "grad_norm": 0.3096616566181183,
+      "learning_rate": 0.0002,
+      "loss": 1.9166,
+      "step": 9
+    },
+    {
+      "epoch": 0.02222839677688247,
+      "grad_norm": 0.40228065848350525,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 1.7125,
+      "step": 10
+    },
+    {
+      "epoch": 0.024451236454570716,
+      "grad_norm": 0.3223334848880768,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 1.929,
+      "step": 11
+    },
+    {
+      "epoch": 0.026674076132258962,
+      "grad_norm": 0.28250545263290405,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 1.6014,
+      "step": 12
+    },
+    {
+      "epoch": 0.02889691580994721,
+      "grad_norm": 0.3532133400440216,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 1.8351,
+      "step": 13
+    },
+    {
+      "epoch": 0.031119755487635455,
+      "grad_norm": 0.28702786564826965,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 1.7652,
+      "step": 14
+    },
+    {
+      "epoch": 0.0333425951653237,
+      "grad_norm": 0.30105912685394287,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 1.6464,
+      "step": 15
+    },
+    {
+      "epoch": 0.035565434843011945,
+      "grad_norm": 0.2549647092819214,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 1.7022,
+      "step": 16
+    },
+    {
+      "epoch": 0.037788274520700195,
+      "grad_norm": 0.280660480260849,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 1.7242,
+      "step": 17
+    },
+    {
+      "epoch": 0.04001111419838844,
+      "grad_norm": 0.3075920641422272,
+      "learning_rate": 0.0004,
+      "loss": 1.5918,
+      "step": 18
+    },
+    {
+      "epoch": 0.04223395387607669,
+      "grad_norm": 0.2627553939819336,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 1.6612,
+      "step": 19
+    },
+    {
+      "epoch": 0.04445679355376494,
+      "grad_norm": 0.24408374726772308,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 1.7682,
+      "step": 20
+    },
+    {
+      "epoch": 0.04667963323145318,
+      "grad_norm": 0.27137765288352966,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 1.6334,
+      "step": 21
+    },
+    {
+      "epoch": 0.04890247290914143,
+      "grad_norm": 0.25347405672073364,
+      "learning_rate": 0.000488888888888889,
+      "loss": 1.8713,
+      "step": 22
+    },
+    {
+      "epoch": 0.051125312586829674,
+      "grad_norm": 0.21500591933727264,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 1.637,
+      "step": 23
+    },
+    {
+      "epoch": 0.053348152264517924,
+      "grad_norm": 0.23220418393611908,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 1.6928,
+      "step": 24
+    },
+    {
+      "epoch": 0.05557099194220617,
+      "grad_norm": 0.21109041571617126,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 1.5009,
+      "step": 25
+    },
+    {
+      "epoch": 0.05779383161989442,
+      "grad_norm": 0.19518163800239563,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 1.5758,
+      "step": 26
+    },
+    {
+      "epoch": 0.06001667129758266,
+      "grad_norm": 0.19399304687976837,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 1.6354,
+      "step": 27
+    },
+    {
+      "epoch": 0.06223951097527091,
+      "grad_norm": 0.2238541692495346,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 1.4809,
+      "step": 28
+    },
+    {
+      "epoch": 0.06446235065295916,
+      "grad_norm": 0.17391888797283173,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 1.3821,
+      "step": 29
+    },
+    {
+      "epoch": 0.0666851903306474,
+      "grad_norm": 0.17666837573051453,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 1.3057,
+      "step": 30
+    },
+    {
+      "epoch": 0.06890803000833565,
+      "grad_norm": 0.19777463376522064,
+      "learning_rate": 0.000688888888888889,
+      "loss": 1.617,
+      "step": 31
+    },
+    {
+      "epoch": 0.07113086968602389,
+      "grad_norm": 0.16105899214744568,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 1.5269,
+      "step": 32
+    },
+    {
+      "epoch": 0.07335370936371215,
+      "grad_norm": 0.17360083758831024,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 1.4468,
+      "step": 33
+    },
+    {
+      "epoch": 0.07557654904140039,
+      "grad_norm": 0.1594723016023636,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 1.4326,
+      "step": 34
+    },
+    {
+      "epoch": 0.07779938871908863,
+      "grad_norm": 0.1737031638622284,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 1.405,
+      "step": 35
+    },
+    {
+      "epoch": 0.08002222839677688,
+      "grad_norm": 0.1713709533214569,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 36
+    },
+    {
+      "epoch": 0.08224506807446513,
+      "grad_norm": 0.1633702665567398,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 1.3118,
+      "step": 37
+    },
+    {
+      "epoch": 0.08446790775215338,
+      "grad_norm": 0.17267894744873047,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 1.3091,
+      "step": 38
+    },
+    {
+      "epoch": 0.08669074742984162,
+      "grad_norm": 0.14479492604732513,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 1.378,
+      "step": 39
+    },
+    {
+      "epoch": 0.08891358710752988,
+      "grad_norm": 0.13730104267597198,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 1.244,
+      "step": 40
+    },
+    {
+      "epoch": 0.09113642678521812,
+      "grad_norm": 0.17174997925758362,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 1.3365,
+      "step": 41
+    },
+    {
+      "epoch": 0.09335926646290636,
+      "grad_norm": 0.13429133594036102,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 1.0797,
+      "step": 42
+    },
+    {
+      "epoch": 0.0955821061405946,
+      "grad_norm": 0.1332227885723114,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 1.3447,
+      "step": 43
+    },
+    {
+      "epoch": 0.09780494581828286,
+      "grad_norm": 0.10597763955593109,
+      "learning_rate": 0.000977777777777778,
+      "loss": 1.2595,
+      "step": 44
+    },
+    {
+      "epoch": 0.1000277854959711,
+      "grad_norm": 0.14317654073238373,
+      "learning_rate": 0.001,
+      "loss": 1.2789,
+      "step": 45
+    },
+    {
+      "epoch": 0.10225062517365935,
+      "grad_norm": 0.13641610741615295,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 1.3444,
+      "step": 46
+    },
+    {
+      "epoch": 0.10447346485134759,
+      "grad_norm": 0.125591441988945,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 1.1095,
+      "step": 47
+    },
+    {
+      "epoch": 0.10669630452903585,
+      "grad_norm": 0.14460553228855133,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 1.2323,
+      "step": 48
+    },
+    {
+      "epoch": 0.10891914420672409,
+      "grad_norm": 0.12514247000217438,
+      "learning_rate": 0.001088888888888889,
+      "loss": 1.1664,
+      "step": 49
+    },
+    {
+      "epoch": 0.11114198388441233,
+      "grad_norm": 0.1322765350341797,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 1.3675,
+      "step": 50
+    },
+    {
+      "epoch": 0.11336482356210058,
+      "grad_norm": 0.14224638044834137,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 1.24,
+      "step": 51
+    },
+    {
+      "epoch": 0.11558766323978883,
+      "grad_norm": 0.11452258378267288,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 1.3013,
+      "step": 52
+    },
+    {
+      "epoch": 0.11781050291747708,
+      "grad_norm": 0.1195889487862587,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 1.1503,
+      "step": 53
+    },
+    {
+      "epoch": 0.12003334259516532,
+      "grad_norm": 0.1277323216199875,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 1.1958,
+      "step": 54
+    },
+    {
+      "epoch": 0.12225618227285356,
+      "grad_norm": 0.09402500092983246,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 0.9762,
+      "step": 55
+    },
+    {
+      "epoch": 0.12447902195054182,
+      "grad_norm": 0.10695713013410568,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 1.1013,
+      "step": 56
+    },
+    {
+      "epoch": 0.12670186162823005,
+      "grad_norm": 0.10557325184345245,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 1.0793,
+      "step": 57
+    },
+    {
+      "epoch": 0.12892470130591832,
+      "grad_norm": 0.12055651843547821,
+      "learning_rate": 0.001288888888888889,
+      "loss": 1.2221,
+      "step": 58
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 0.08459929376840591,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 1.0976,
+      "step": 59
+    },
+    {
+      "epoch": 0.1333703806612948,
+      "grad_norm": 0.11813811212778091,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 1.1168,
+      "step": 60
+    },
+    {
+      "epoch": 0.13559322033898305,
+      "grad_norm": 0.1075403243303299,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 0.9292,
+      "step": 61
+    },
+    {
+      "epoch": 0.1378160600166713,
+      "grad_norm": 0.07610759884119034,
+      "learning_rate": 0.001377777777777778,
+      "loss": 1.0947,
+      "step": 62
+    },
+    {
+      "epoch": 0.14003889969435954,
+      "grad_norm": 0.09791035205125809,
+      "learning_rate": 0.0014,
+      "loss": 1.1571,
+      "step": 63
+    },
+    {
+      "epoch": 0.14226173937204778,
+      "grad_norm": 0.09495139867067337,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 0.8295,
+      "step": 64
+    },
+    {
+      "epoch": 0.14448457904973605,
+      "grad_norm": 0.07287945598363876,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 0.9453,
+      "step": 65
+    },
+    {
+      "epoch": 0.1467074187274243,
+      "grad_norm": 0.08833761513233185,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 1.1003,
+      "step": 66
+    },
+    {
+      "epoch": 0.14893025840511254,
+      "grad_norm": 0.09454136341810226,
+      "learning_rate": 0.001488888888888889,
+      "loss": 1.1327,
+      "step": 67
+    },
+    {
+      "epoch": 0.15115309808280078,
+      "grad_norm": 0.09609611332416534,
+      "learning_rate": 0.001511111111111111,
+      "loss": 0.9176,
+      "step": 68
+    },
+    {
+      "epoch": 0.15337593776048902,
+      "grad_norm": 0.10027166455984116,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 1.0967,
+      "step": 69
+    },
+    {
+      "epoch": 0.15559877743817727,
+      "grad_norm": 0.08003437519073486,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 0.9908,
+      "step": 70
+    },
+    {
+      "epoch": 0.1578216171158655,
+      "grad_norm": 0.06798948347568512,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 0.8607,
+      "step": 71
+    },
+    {
+      "epoch": 0.16004445679355375,
+      "grad_norm": 0.07019151747226715,
+      "learning_rate": 0.0016,
+      "loss": 0.9327,
+      "step": 72
+    },
+    {
+      "epoch": 0.16226729647124202,
+      "grad_norm": 0.09492333978414536,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 0.8007,
+      "step": 73
+    },
+    {
+      "epoch": 0.16449013614893027,
+      "grad_norm": 0.06816477328538895,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 0.8318,
+      "step": 74
+    },
+    {
+      "epoch": 0.1667129758266185,
+      "grad_norm": 0.05683727562427521,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 0.816,
+      "step": 75
+    },
+    {
+      "epoch": 0.16893581550430675,
+      "grad_norm": 0.06219564750790596,
+      "learning_rate": 0.001688888888888889,
+      "loss": 0.8008,
+      "step": 76
+    },
+    {
+      "epoch": 0.171158655181995,
+      "grad_norm": 0.06426230072975159,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 1.0033,
+      "step": 77
+    },
+    {
+      "epoch": 0.17338149485968324,
+      "grad_norm": 0.057185135781764984,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 0.9249,
+      "step": 78
+    },
+    {
+      "epoch": 0.17560433453737148,
+      "grad_norm": 0.0551241934299469,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 0.891,
+      "step": 79
+    },
+    {
+      "epoch": 0.17782717421505975,
+      "grad_norm": 0.05191711336374283,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 0.9672,
+      "step": 80
+    },
+    {
+      "epoch": 0.180050013892748,
+      "grad_norm": 0.06342898309230804,
+      "learning_rate": 0.0018,
+      "loss": 0.9372,
+      "step": 81
+    },
+    {
+      "epoch": 0.18227285357043624,
+      "grad_norm": 0.05517818406224251,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 0.8477,
+      "step": 82
+    },
+    {
+      "epoch": 0.18449569324812448,
+      "grad_norm": 0.058783963322639465,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 1.0301,
+      "step": 83
+    },
+    {
+      "epoch": 0.18671853292581272,
+      "grad_norm": 0.05436277762055397,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 0.8878,
+      "step": 84
+    },
+    {
+      "epoch": 0.18894137260350097,
+      "grad_norm": 0.041651614010334015,
+      "learning_rate": 0.001888888888888889,
+      "loss": 0.9357,
+      "step": 85
+    },
+    {
+      "epoch": 0.1911642122811892,
+      "grad_norm": 0.050092216581106186,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 0.9938,
+      "step": 86
+    },
+    {
+      "epoch": 0.19338705195887745,
+      "grad_norm": 0.053640950471162796,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 1.0159,
+      "step": 87
+    },
+    {
+      "epoch": 0.19560989163656572,
+      "grad_norm": 0.05056702345609665,
+      "learning_rate": 0.001955555555555556,
+      "loss": 0.9827,
+      "step": 88
+    },
+    {
+      "epoch": 0.19783273131425397,
+      "grad_norm": 0.04525972530245781,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 0.7511,
+      "step": 89
+    },
+    {
+      "epoch": 0.2000555709919422,
+      "grad_norm": 0.04984632506966591,
+      "learning_rate": 0.002,
+      "loss": 0.8568,
+      "step": 90
+    },
+    {
+      "epoch": 0.20227841066963045,
+      "grad_norm": 0.04132724925875664,
+      "learning_rate": 0.002022222222222222,
+      "loss": 0.9806,
+      "step": 91
+    },
+    {
+      "epoch": 0.2045012503473187,
+      "grad_norm": 0.039617739617824554,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 0.8346,
+      "step": 92
+    },
+    {
+      "epoch": 0.20672409002500694,
+      "grad_norm": 0.03755780681967735,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 0.813,
+      "step": 93
+    },
+    {
+      "epoch": 0.20894692970269518,
+      "grad_norm": 0.038126129657030106,
+      "learning_rate": 0.002088888888888889,
+      "loss": 0.9437,
+      "step": 94
+    },
+    {
+      "epoch": 0.21116976938038343,
+      "grad_norm": 0.03865043818950653,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 0.7994,
+      "step": 95
+    },
+    {
+      "epoch": 0.2133926090580717,
+      "grad_norm": 0.038602616637945175,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 0.9596,
+      "step": 96
+    },
+    {
+      "epoch": 0.21561544873575994,
+      "grad_norm": 0.03432125225663185,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 0.7337,
+      "step": 97
+    },
+    {
+      "epoch": 0.21783828841344818,
+      "grad_norm": 0.04551972821354866,
+      "learning_rate": 0.002177777777777778,
+      "loss": 0.9933,
+      "step": 98
+    },
+    {
+      "epoch": 0.22006112809113643,
+      "grad_norm": 0.03250821307301521,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 0.8251,
+      "step": 99
+    },
+    {
+      "epoch": 0.22228396776882467,
+      "grad_norm": 0.03453311696648598,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 0.8074,
+      "step": 100
+    },
+    {
+      "epoch": 0.2245068074465129,
+      "grad_norm": 0.041663624346256256,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 0.7755,
+      "step": 101
+    },
+    {
+      "epoch": 0.22672964712420116,
+      "grad_norm": 0.035251155495643616,
+      "learning_rate": 0.002266666666666667,
+      "loss": 0.9015,
+      "step": 102
+    },
+    {
+      "epoch": 0.22895248680188943,
+      "grad_norm": 0.026816092431545258,
+      "learning_rate": 0.002288888888888889,
+      "loss": 0.6089,
+      "step": 103
+    },
+    {
+      "epoch": 0.23117532647957767,
+      "grad_norm": 0.025371363386511803,
+      "learning_rate": 0.002311111111111111,
+      "loss": 0.6579,
+      "step": 104
+    },
+    {
+      "epoch": 0.2333981661572659,
+      "grad_norm": 0.030030515044927597,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 0.7587,
+      "step": 105
+    },
+    {
+      "epoch": 0.23562100583495416,
+      "grad_norm": 0.03042268380522728,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 0.673,
+      "step": 106
+    },
+    {
+      "epoch": 0.2378438455126424,
+      "grad_norm": 0.03374524414539337,
+      "learning_rate": 0.002377777777777778,
+      "loss": 0.7638,
+      "step": 107
+    },
+    {
+      "epoch": 0.24006668519033064,
+      "grad_norm": 0.05823088809847832,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 1.0445,
+      "step": 108
+    },
+    {
+      "epoch": 0.24228952486801889,
+      "grad_norm": 0.02552204020321369,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 0.758,
+      "step": 109
+    },
+    {
+      "epoch": 0.24451236454570713,
+      "grad_norm": 0.030858848243951797,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 0.8163,
+      "step": 110
+    },
+    {
+      "epoch": 0.2467352042233954,
+      "grad_norm": 0.030134430155158043,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 0.8197,
+      "step": 111
+    },
+    {
+      "epoch": 0.24895804390108364,
+      "grad_norm": 0.03654756769537926,
+      "learning_rate": 0.002488888888888889,
+      "loss": 0.8028,
+      "step": 112
+    },
+    {
+      "epoch": 0.2511808835787719,
+      "grad_norm": 0.04132033884525299,
+      "learning_rate": 0.002511111111111111,
+      "loss": 0.8087,
+      "step": 113
+    },
+    {
+      "epoch": 0.2534037232564601,
+      "grad_norm": 0.02729404903948307,
+      "learning_rate": 0.002533333333333333,
+      "loss": 0.7327,
+      "step": 114
+    },
+    {
+      "epoch": 0.25562656293414837,
+      "grad_norm": 0.037324242293834686,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 0.8203,
+      "step": 115
+    },
+    {
+      "epoch": 0.25784940261183664,
+      "grad_norm": 0.045340005308389664,
+      "learning_rate": 0.002577777777777778,
+      "loss": 1.1208,
+      "step": 116
+    },
+    {
+      "epoch": 0.26007224228952486,
+      "grad_norm": 0.03202158212661743,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 0.7704,
+      "step": 117
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 0.031156064942479134,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 0.6212,
+      "step": 118
+    },
+    {
+      "epoch": 0.26451792164490134,
+      "grad_norm": 0.0396842360496521,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 0.8821,
+      "step": 119
+    },
+    {
+      "epoch": 0.2667407613225896,
+      "grad_norm": 0.024765778332948685,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 0.7624,
+      "step": 120
+    },
+    {
+      "epoch": 0.26896360100027783,
+      "grad_norm": 0.0588197335600853,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 0.7491,
+      "step": 121
+    },
+    {
+      "epoch": 0.2711864406779661,
+      "grad_norm": 0.04317701607942581,
+      "learning_rate": 0.002711111111111111,
+      "loss": 0.9822,
+      "step": 122
+    },
+    {
+      "epoch": 0.27340928035565437,
+      "grad_norm": 0.023474140092730522,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 0.6543,
+      "step": 123
+    },
+    {
+      "epoch": 0.2756321200333426,
+      "grad_norm": 0.03395792841911316,
+      "learning_rate": 0.002755555555555556,
+      "loss": 0.9753,
+      "step": 124
+    },
+    {
+      "epoch": 0.27785495971103086,
+      "grad_norm": 0.03774833306670189,
+      "learning_rate": 0.002777777777777778,
+      "loss": 0.7568,
+      "step": 125
+    },
+    {
+      "epoch": 0.2800777993887191,
+      "grad_norm": 0.019662071019411087,
+      "learning_rate": 0.0028,
+      "loss": 0.683,
+      "step": 126
+    },
+    {
+      "epoch": 0.28230063906640734,
+      "grad_norm": 0.02699323743581772,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 0.7626,
+      "step": 127
+    },
+    {
+      "epoch": 0.28452347874409556,
+      "grad_norm": 0.02954452484846115,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 0.8323,
+      "step": 128
+    },
+    {
+      "epoch": 0.28674631842178383,
+      "grad_norm": 0.021271225064992905,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 0.7864,
+      "step": 129
+    },
+    {
+      "epoch": 0.2889691580994721,
+      "grad_norm": 0.02944800816476345,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 0.9257,
+      "step": 130
+    },
+    {
+      "epoch": 0.2911919977771603,
+      "grad_norm": 0.029844852164387703,
+      "learning_rate": 0.002911111111111111,
+      "loss": 0.7552,
+      "step": 131
+    },
+    {
+      "epoch": 0.2934148374548486,
+      "grad_norm": 0.023612311109900475,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 0.6756,
+      "step": 132
+    },
+    {
+      "epoch": 0.2956376771325368,
+      "grad_norm": 0.017975427210330963,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 0.7141,
+      "step": 133
+    },
+    {
+      "epoch": 0.2978605168102251,
+      "grad_norm": 0.038071002811193466,
+      "learning_rate": 0.002977777777777778,
+      "loss": 0.9247,
+      "step": 134
+    },
+    {
+      "epoch": 0.3000833564879133,
+      "grad_norm": 0.027641968801617622,
+      "learning_rate": 0.003,
+      "loss": 0.6383,
+      "step": 135
+    },
+    {
+      "epoch": 0.30230619616560156,
+      "grad_norm": 0.02428288944065571,
+      "learning_rate": 0.0029975247524752476,
+      "loss": 0.803,
+      "step": 136
+    },
+    {
+      "epoch": 0.30452903584328983,
+      "grad_norm": 0.055082373321056366,
+      "learning_rate": 0.002995049504950495,
+      "loss": 0.6262,
+      "step": 137
+    },
+    {
+      "epoch": 0.30675187552097805,
+      "grad_norm": 0.019244015216827393,
+      "learning_rate": 0.0029925742574257426,
+      "loss": 0.7297,
+      "step": 138
+    },
+    {
+      "epoch": 0.3089747151986663,
+      "grad_norm": 0.0185896847397089,
+      "learning_rate": 0.00299009900990099,
+      "loss": 0.775,
+      "step": 139
+    },
+    {
+      "epoch": 0.31119755487635453,
+      "grad_norm": 0.04311481863260269,
+      "learning_rate": 0.0029876237623762377,
+      "loss": 0.9192,
+      "step": 140
+    },
+    {
+      "epoch": 0.3134203945540428,
+      "grad_norm": 0.02153279259800911,
+      "learning_rate": 0.002985148514851485,
+      "loss": 0.7181,
+      "step": 141
+    },
+    {
+      "epoch": 0.315643234231731,
+      "grad_norm": 0.030037125572562218,
+      "learning_rate": 0.0029826732673267327,
+      "loss": 0.7947,
+      "step": 142
+    },
+    {
+      "epoch": 0.3178660739094193,
+      "grad_norm": 0.03192693367600441,
+      "learning_rate": 0.0029801980198019802,
+      "loss": 1.0311,
+      "step": 143
+    },
+    {
+      "epoch": 0.3200889135871075,
+      "grad_norm": 0.021701497957110405,
+      "learning_rate": 0.0029777227722772278,
+      "loss": 0.6511,
+      "step": 144
+    },
+    {
+      "epoch": 0.3223117532647958,
+      "grad_norm": 0.025493718683719635,
+      "learning_rate": 0.0029752475247524753,
+      "loss": 0.6231,
+      "step": 145
+    },
+    {
+      "epoch": 0.32453459294248405,
+      "grad_norm": 0.02308356948196888,
+      "learning_rate": 0.002972772277227723,
+      "loss": 0.7476,
+      "step": 146
+    },
+    {
+      "epoch": 0.32675743262017226,
+      "grad_norm": 0.01986018754541874,
+      "learning_rate": 0.0029702970297029703,
+      "loss": 0.7325,
+      "step": 147
+    },
+    {
+      "epoch": 0.32898027229786053,
+      "grad_norm": 0.017985232174396515,
+      "learning_rate": 0.002967821782178218,
+      "loss": 0.7233,
+      "step": 148
+    },
+    {
+      "epoch": 0.33120311197554875,
+      "grad_norm": 0.01952328160405159,
+      "learning_rate": 0.0029653465346534654,
+      "loss": 0.6821,
+      "step": 149
+    },
+    {
+      "epoch": 0.333425951653237,
+      "grad_norm": 0.020746653899550438,
+      "learning_rate": 0.002962871287128713,
+      "loss": 0.8254,
+      "step": 150
+    },
+    {
+      "epoch": 0.33564879133092523,
+      "grad_norm": 0.018172353506088257,
+      "learning_rate": 0.0029603960396039604,
+      "loss": 0.6146,
+      "step": 151
+    },
+    {
+      "epoch": 0.3378716310086135,
+      "grad_norm": 0.01994176022708416,
+      "learning_rate": 0.002957920792079208,
+      "loss": 0.8184,
+      "step": 152
+    },
+    {
+      "epoch": 0.3400944706863018,
+      "grad_norm": 0.026261357590556145,
+      "learning_rate": 0.0029554455445544555,
+      "loss": 0.7213,
+      "step": 153
+    },
+    {
+      "epoch": 0.34231731036399,
+      "grad_norm": 0.019761715084314346,
+      "learning_rate": 0.002952970297029703,
+      "loss": 0.7215,
+      "step": 154
+    },
+    {
+      "epoch": 0.34454015004167826,
+      "grad_norm": 0.022664040327072144,
+      "learning_rate": 0.0029504950495049505,
+      "loss": 0.8718,
+      "step": 155
+    },
+    {
+      "epoch": 0.3467629897193665,
+      "grad_norm": 0.027874581515789032,
+      "learning_rate": 0.002948019801980198,
+      "loss": 0.8164,
+      "step": 156
+    },
+    {
+      "epoch": 0.34898582939705475,
+      "grad_norm": 0.02273603528738022,
+      "learning_rate": 0.0029455445544554456,
+      "loss": 0.7851,
+      "step": 157
+    },
+    {
+      "epoch": 0.35120866907474296,
+      "grad_norm": 0.0187845416367054,
+      "learning_rate": 0.002943069306930693,
+      "loss": 0.7604,
+      "step": 158
+    },
+    {
+      "epoch": 0.35343150875243123,
+      "grad_norm": 0.019273824989795685,
+      "learning_rate": 0.0029405940594059406,
+      "loss": 0.5622,
+      "step": 159
+    },
+    {
+      "epoch": 0.3556543484301195,
+      "grad_norm": 0.02631755918264389,
+      "learning_rate": 0.002938118811881188,
+      "loss": 0.8661,
+      "step": 160
+    },
+    {
+      "epoch": 0.3578771881078077,
+      "grad_norm": 0.022346017882227898,
+      "learning_rate": 0.0029356435643564356,
+      "loss": 0.6781,
+      "step": 161
+    },
+    {
+      "epoch": 0.360100027785496,
+      "grad_norm": 0.021816574037075043,
+      "learning_rate": 0.002933168316831683,
+      "loss": 0.9022,
+      "step": 162
+    },
+    {
+      "epoch": 0.3623228674631842,
+      "grad_norm": 0.021531615406274796,
+      "learning_rate": 0.002930693069306931,
+      "loss": 0.6871,
+      "step": 163
+    },
+    {
+      "epoch": 0.3645457071408725,
+      "grad_norm": 0.022265218198299408,
+      "learning_rate": 0.002928217821782178,
+      "loss": 0.9247,
+      "step": 164
+    },
+    {
+      "epoch": 0.3667685468185607,
+      "grad_norm": 0.01669679768383503,
+      "learning_rate": 0.0029257425742574257,
+      "loss": 0.6005,
+      "step": 165
+    },
+    {
+      "epoch": 0.36899138649624896,
+      "grad_norm": 0.020891698077321053,
+      "learning_rate": 0.0029232673267326737,
+      "loss": 0.7063,
+      "step": 166
+    },
+    {
+      "epoch": 0.3712142261739372,
+      "grad_norm": 0.01642497256398201,
+      "learning_rate": 0.002920792079207921,
+      "loss": 0.7598,
+      "step": 167
+    },
+    {
+      "epoch": 0.37343706585162545,
+      "grad_norm": 0.022815469652414322,
+      "learning_rate": 0.0029183168316831683,
+      "loss": 0.7631,
+      "step": 168
+    },
+    {
+      "epoch": 0.3756599055293137,
+      "grad_norm": 0.02138301357626915,
+      "learning_rate": 0.0029158415841584163,
+      "loss": 0.7098,
+      "step": 169
+    },
+    {
+      "epoch": 0.37788274520700194,
+      "grad_norm": 0.022588403895497322,
+      "learning_rate": 0.0029133663366336634,
+      "loss": 0.8482,
+      "step": 170
+    },
+    {
+      "epoch": 0.3801055848846902,
+      "grad_norm": 0.017089666798710823,
+      "learning_rate": 0.002910891089108911,
+      "loss": 0.6223,
+      "step": 171
+    },
+    {
+      "epoch": 0.3823284245623784,
+      "grad_norm": 0.014961940236389637,
+      "learning_rate": 0.002908415841584159,
+      "loss": 0.6889,
+      "step": 172
+    },
+    {
+      "epoch": 0.3845512642400667,
+      "grad_norm": 0.022101087495684624,
+      "learning_rate": 0.002905940594059406,
+      "loss": 0.7824,
+      "step": 173
+    },
+    {
+      "epoch": 0.3867741039177549,
+      "grad_norm": 0.020935313776135445,
+      "learning_rate": 0.0029034653465346534,
+      "loss": 0.7214,
+      "step": 174
+    },
+    {
+      "epoch": 0.3889969435954432,
+      "grad_norm": 0.027509761974215508,
+      "learning_rate": 0.0029009900990099014,
+      "loss": 0.7776,
+      "step": 175
+    },
+    {
+      "epoch": 0.39121978327313145,
+      "grad_norm": 0.03217627480626106,
+      "learning_rate": 0.0028985148514851485,
+      "loss": 0.7934,
+      "step": 176
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 0.01634255051612854,
+      "learning_rate": 0.002896039603960396,
+      "loss": 0.6374,
+      "step": 177
+    },
+    {
+      "epoch": 0.39566546262850794,
+      "grad_norm": 0.016692742705345154,
+      "learning_rate": 0.002893564356435644,
+      "loss": 0.6864,
+      "step": 178
+    },
+    {
+      "epoch": 0.39788830230619615,
+      "grad_norm": 0.019375508651137352,
+      "learning_rate": 0.002891089108910891,
+      "loss": 0.7127,
+      "step": 179
+    },
+    {
+      "epoch": 0.4001111419838844,
+      "grad_norm": 0.02516164816915989,
+      "learning_rate": 0.0028886138613861386,
+      "loss": 0.6287,
+      "step": 180
+    },
+    {
+      "epoch": 0.40233398166157264,
+      "grad_norm": 0.024371830746531487,
+      "learning_rate": 0.0028861386138613865,
+      "loss": 0.6914,
+      "step": 181
+    },
+    {
+      "epoch": 0.4045568213392609,
+      "grad_norm": 0.023201432079076767,
+      "learning_rate": 0.0028836633663366336,
+      "loss": 0.8256,
+      "step": 182
+    },
+    {
+      "epoch": 0.4067796610169492,
+      "grad_norm": 0.03344796970486641,
+      "learning_rate": 0.002881188118811881,
+      "loss": 0.7973,
+      "step": 183
+    },
+    {
+      "epoch": 0.4090025006946374,
+      "grad_norm": 0.019620005041360855,
+      "learning_rate": 0.002878712871287129,
+      "loss": 0.7596,
+      "step": 184
+    },
+    {
+      "epoch": 0.41122534037232567,
+      "grad_norm": 0.017026478424668312,
+      "learning_rate": 0.002876237623762376,
+      "loss": 0.7195,
+      "step": 185
+    },
+    {
+      "epoch": 0.4134481800500139,
+      "grad_norm": 0.01848086714744568,
+      "learning_rate": 0.0028737623762376237,
+      "loss": 0.887,
+      "step": 186
+    },
+    {
+      "epoch": 0.41567101972770215,
+      "grad_norm": 0.02092432789504528,
+      "learning_rate": 0.0028712871287128717,
+      "loss": 0.7142,
+      "step": 187
+    },
+    {
+      "epoch": 0.41789385940539037,
+      "grad_norm": 0.060593266040086746,
+      "learning_rate": 0.0028688118811881188,
+      "loss": 0.6788,
+      "step": 188
+    },
+    {
+      "epoch": 0.42011669908307864,
+      "grad_norm": 0.018896559253335,
+      "learning_rate": 0.0028663366336633663,
+      "loss": 0.5818,
+      "step": 189
+    },
+    {
+      "epoch": 0.42233953876076685,
+      "grad_norm": 0.015938229858875275,
+      "learning_rate": 0.0028638613861386142,
+      "loss": 0.6667,
+      "step": 190
+    },
+    {
+      "epoch": 0.4245623784384551,
+      "grad_norm": 0.04060890153050423,
+      "learning_rate": 0.0028613861386138613,
+      "loss": 0.7224,
+      "step": 191
+    },
+    {
+      "epoch": 0.4267852181161434,
+      "grad_norm": 0.019955582916736603,
+      "learning_rate": 0.002858910891089109,
+      "loss": 0.7129,
+      "step": 192
+    },
+    {
+      "epoch": 0.4290080577938316,
+      "grad_norm": 0.02846820093691349,
+      "learning_rate": 0.002856435643564357,
+      "loss": 0.8338,
+      "step": 193
+    },
+    {
+      "epoch": 0.4312308974715199,
+      "grad_norm": 0.024327551946043968,
+      "learning_rate": 0.002853960396039604,
+      "loss": 0.7844,
+      "step": 194
+    },
+    {
+      "epoch": 0.4334537371492081,
+      "grad_norm": 0.022156810387969017,
+      "learning_rate": 0.0028514851485148514,
+      "loss": 0.7783,
+      "step": 195
+    },
+    {
+      "epoch": 0.43567657682689637,
+      "grad_norm": 0.020973121747374535,
+      "learning_rate": 0.0028490099009900994,
+      "loss": 0.7924,
+      "step": 196
+    },
+    {
+      "epoch": 0.4378994165045846,
+      "grad_norm": 0.02271064557135105,
+      "learning_rate": 0.0028465346534653465,
+      "loss": 0.8475,
+      "step": 197
+    },
+    {
+      "epoch": 0.44012225618227285,
+      "grad_norm": 0.017409365624189377,
+      "learning_rate": 0.002844059405940594,
+      "loss": 0.6592,
+      "step": 198
+    },
+    {
+      "epoch": 0.4423450958599611,
+      "grad_norm": 0.016952740028500557,
+      "learning_rate": 0.002841584158415842,
+      "loss": 0.7419,
+      "step": 199
+    },
+    {
+      "epoch": 0.44456793553764934,
+      "grad_norm": 0.022074569016695023,
+      "learning_rate": 0.002839108910891089,
+      "loss": 0.9592,
+      "step": 200
+    },
+    {
+      "epoch": 0.4467907752153376,
+      "grad_norm": 0.019633406773209572,
+      "learning_rate": 0.0028366336633663366,
+      "loss": 0.846,
+      "step": 201
+    },
+    {
+      "epoch": 0.4490136148930258,
+      "grad_norm": 0.02506411075592041,
+      "learning_rate": 0.0028341584158415845,
+      "loss": 0.8544,
+      "step": 202
+    },
+    {
+      "epoch": 0.4512364545707141,
+      "grad_norm": 0.011820493265986443,
+      "learning_rate": 0.0028316831683168316,
+      "loss": 0.6832,
+      "step": 203
+    },
+    {
+      "epoch": 0.4534592942484023,
+      "grad_norm": 0.013546622358262539,
+      "learning_rate": 0.002829207920792079,
+      "loss": 0.7245,
+      "step": 204
+    },
+    {
+      "epoch": 0.4556821339260906,
+      "grad_norm": 0.016001511365175247,
+      "learning_rate": 0.002826732673267327,
+      "loss": 0.6386,
+      "step": 205
+    },
+    {
+      "epoch": 0.45790497360377885,
+      "grad_norm": 0.016753453761339188,
+      "learning_rate": 0.002824257425742574,
+      "loss": 0.681,
+      "step": 206
+    },
+    {
+      "epoch": 0.46012781328146707,
+      "grad_norm": 0.021365856751799583,
+      "learning_rate": 0.0028217821782178217,
+      "loss": 0.7229,
+      "step": 207
+    },
+    {
+      "epoch": 0.46235065295915534,
+      "grad_norm": 0.016552921384572983,
+      "learning_rate": 0.0028193069306930696,
+      "loss": 0.741,
+      "step": 208
+    },
+    {
+      "epoch": 0.46457349263684355,
+      "grad_norm": 0.0160622950643301,
+      "learning_rate": 0.0028168316831683167,
+      "loss": 0.8179,
+      "step": 209
+    },
+    {
+      "epoch": 0.4667963323145318,
+      "grad_norm": 0.01873830333352089,
+      "learning_rate": 0.0028143564356435643,
+      "loss": 0.7208,
+      "step": 210
+    },
+    {
+      "epoch": 0.46901917199222004,
+      "grad_norm": 0.01617845520377159,
+      "learning_rate": 0.002811881188118812,
+      "loss": 0.7291,
+      "step": 211
+    },
+    {
+      "epoch": 0.4712420116699083,
+      "grad_norm": 0.016890326514840126,
+      "learning_rate": 0.0028094059405940593,
+      "loss": 0.7388,
+      "step": 212
+    },
+    {
+      "epoch": 0.4734648513475965,
+      "grad_norm": 0.01602407731115818,
+      "learning_rate": 0.0028069306930693073,
+      "loss": 0.6657,
+      "step": 213
+    },
+    {
+      "epoch": 0.4756876910252848,
+      "grad_norm": 0.02520548366010189,
+      "learning_rate": 0.0028044554455445548,
+      "loss": 0.7801,
+      "step": 214
+    },
+    {
+      "epoch": 0.47791053070297307,
+      "grad_norm": 0.017677821218967438,
+      "learning_rate": 0.002801980198019802,
+      "loss": 0.7508,
+      "step": 215
+    },
+    {
+      "epoch": 0.4801333703806613,
+      "grad_norm": 0.01971537061035633,
+      "learning_rate": 0.00279950495049505,
+      "loss": 0.8671,
+      "step": 216
+    },
+    {
+      "epoch": 0.48235621005834955,
+      "grad_norm": 0.020906252786517143,
+      "learning_rate": 0.0027970297029702973,
+      "loss": 0.8528,
+      "step": 217
+    },
+    {
+      "epoch": 0.48457904973603777,
+      "grad_norm": 0.020072396844625473,
+      "learning_rate": 0.0027945544554455444,
+      "loss": 0.6818,
+      "step": 218
+    },
+    {
+      "epoch": 0.48680188941372604,
+      "grad_norm": 0.015342594124376774,
+      "learning_rate": 0.0027920792079207924,
+      "loss": 0.7102,
+      "step": 219
+    },
+    {
+      "epoch": 0.48902472909141426,
+      "grad_norm": 0.018210800364613533,
+      "learning_rate": 0.00278960396039604,
+      "loss": 0.6902,
+      "step": 220
+    },
+    {
+      "epoch": 0.4912475687691025,
+      "grad_norm": 0.02456044964492321,
+      "learning_rate": 0.002787128712871287,
+      "loss": 0.7501,
+      "step": 221
+    },
+    {
+      "epoch": 0.4934704084467908,
+      "grad_norm": 0.010794063098728657,
+      "learning_rate": 0.002784653465346535,
+      "loss": 0.6494,
+      "step": 222
+    },
+    {
+      "epoch": 0.495693248124479,
+      "grad_norm": 0.014852452091872692,
+      "learning_rate": 0.0027821782178217825,
+      "loss": 0.6509,
+      "step": 223
+    },
+    {
+      "epoch": 0.4979160878021673,
+      "grad_norm": 0.023020023480057716,
+      "learning_rate": 0.0027797029702970296,
+      "loss": 0.7606,
+      "step": 224
+    },
+    {
+      "epoch": 0.5001389274798556,
+      "grad_norm": 0.018114762380719185,
+      "learning_rate": 0.0027772277227722775,
+      "loss": 0.7541,
+      "step": 225
+    },
+    {
+      "epoch": 0.5023617671575438,
+      "grad_norm": 0.018622858449816704,
+      "learning_rate": 0.002774752475247525,
+      "loss": 0.8621,
+      "step": 226
+    },
+    {
+      "epoch": 0.504584606835232,
+      "grad_norm": 0.018405485898256302,
+      "learning_rate": 0.002772277227722772,
+      "loss": 0.6815,
+      "step": 227
+    },
+    {
+      "epoch": 0.5068074465129202,
+      "grad_norm": 0.019367920234799385,
+      "learning_rate": 0.00276980198019802,
+      "loss": 0.8323,
+      "step": 228
+    },
+    {
+      "epoch": 0.5090302861906085,
+      "grad_norm": 0.016442907974123955,
+      "learning_rate": 0.0027673267326732676,
+      "loss": 0.6732,
+      "step": 229
+    },
+    {
+      "epoch": 0.5112531258682967,
+      "grad_norm": 0.02419120818376541,
+      "learning_rate": 0.0027648514851485147,
+      "loss": 0.7696,
+      "step": 230
+    },
+    {
+      "epoch": 0.513475965545985,
+      "grad_norm": 0.023511352017521858,
+      "learning_rate": 0.0027623762376237627,
+      "loss": 0.6797,
+      "step": 231
+    },
+    {
+      "epoch": 0.5156988052236733,
+      "grad_norm": 0.019304659217596054,
+      "learning_rate": 0.00275990099009901,
+      "loss": 0.7255,
+      "step": 232
+    },
+    {
+      "epoch": 0.5179216449013615,
+      "grad_norm": 0.01737944595515728,
+      "learning_rate": 0.0027574257425742573,
+      "loss": 0.6911,
+      "step": 233
+    },
+    {
+      "epoch": 0.5201444845790497,
+      "grad_norm": 0.017303744331002235,
+      "learning_rate": 0.0027549504950495052,
+      "loss": 0.7673,
+      "step": 234
+    },
+    {
+      "epoch": 0.5223673242567379,
+      "grad_norm": 0.018068619072437286,
+      "learning_rate": 0.0027524752475247528,
+      "loss": 0.7392,
+      "step": 235
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 0.01970060169696808,
+      "learning_rate": 0.00275,
+      "loss": 0.6399,
+      "step": 236
+    },
+    {
+      "epoch": 0.5268130036121145,
+      "grad_norm": 0.016325384378433228,
+      "learning_rate": 0.002747524752475248,
+      "loss": 0.8503,
+      "step": 237
+    },
+    {
+      "epoch": 0.5290358432898027,
+      "grad_norm": 0.020598750561475754,
+      "learning_rate": 0.002745049504950495,
+      "loss": 0.7524,
+      "step": 238
+    },
+    {
+      "epoch": 0.531258682967491,
+      "grad_norm": 0.016212867572903633,
+      "learning_rate": 0.0027425742574257424,
+      "loss": 0.6051,
+      "step": 239
+    },
+    {
+      "epoch": 0.5334815226451792,
+      "grad_norm": 0.015269650146365166,
+      "learning_rate": 0.0027400990099009904,
+      "loss": 0.642,
+      "step": 240
+    },
+    {
+      "epoch": 0.5357043623228674,
+      "grad_norm": 0.023223020136356354,
+      "learning_rate": 0.0027376237623762375,
+      "loss": 0.5973,
+      "step": 241
+    },
+    {
+      "epoch": 0.5379272020005557,
+      "grad_norm": 0.015060219913721085,
+      "learning_rate": 0.002735148514851485,
+      "loss": 0.6679,
+      "step": 242
+    },
+    {
+      "epoch": 0.540150041678244,
+      "grad_norm": 0.016781222075223923,
+      "learning_rate": 0.002732673267326733,
+      "loss": 0.6876,
+      "step": 243
+    },
+    {
+      "epoch": 0.5423728813559322,
+      "grad_norm": 0.01453195046633482,
+      "learning_rate": 0.00273019801980198,
+      "loss": 0.6025,
+      "step": 244
+    },
+    {
+      "epoch": 0.5445957210336204,
+      "grad_norm": 0.017990389838814735,
+      "learning_rate": 0.0027277227722772275,
+      "loss": 0.7093,
+      "step": 245
+    },
+    {
+      "epoch": 0.5468185607113087,
+      "grad_norm": 0.013350987806916237,
+      "learning_rate": 0.0027252475247524755,
+      "loss": 0.659,
+      "step": 246
+    },
+    {
+      "epoch": 0.549041400388997,
+      "grad_norm": 0.016740279272198677,
+      "learning_rate": 0.0027227722772277226,
+      "loss": 0.6952,
+      "step": 247
+    },
+    {
+      "epoch": 0.5512642400666852,
+      "grad_norm": 0.016187356784939766,
+      "learning_rate": 0.00272029702970297,
+      "loss": 0.6865,
+      "step": 248
+    },
+    {
+      "epoch": 0.5534870797443734,
+      "grad_norm": 0.015389680862426758,
+      "learning_rate": 0.002717821782178218,
+      "loss": 0.7032,
+      "step": 249
+    },
+    {
+      "epoch": 0.5557099194220617,
+      "grad_norm": 0.01613827608525753,
+      "learning_rate": 0.002715346534653465,
+      "loss": 0.8342,
+      "step": 250
+    },
+    {
+      "epoch": 0.5579327590997499,
+      "grad_norm": 0.02487931028008461,
+      "learning_rate": 0.0027128712871287127,
+      "loss": 0.7652,
+      "step": 251
+    },
+    {
+      "epoch": 0.5601555987774381,
+      "grad_norm": 0.01547365915030241,
+      "learning_rate": 0.0027103960396039606,
+      "loss": 0.6371,
+      "step": 252
+    },
+    {
+      "epoch": 0.5623784384551265,
+      "grad_norm": 0.08315558731555939,
+      "learning_rate": 0.0027079207920792077,
+      "loss": 0.7897,
+      "step": 253
+    },
+    {
+      "epoch": 0.5646012781328147,
+      "grad_norm": 0.014980711974203587,
+      "learning_rate": 0.0027054455445544552,
+      "loss": 0.8446,
+      "step": 254
+    },
+    {
+      "epoch": 0.5668241178105029,
+      "grad_norm": 0.01912595145404339,
+      "learning_rate": 0.002702970297029703,
+      "loss": 0.6859,
+      "step": 255
+    },
+    {
+      "epoch": 0.5690469574881911,
+      "grad_norm": 0.015234099701046944,
+      "learning_rate": 0.0027004950495049503,
+      "loss": 0.6902,
+      "step": 256
+    },
+    {
+      "epoch": 0.5712697971658794,
+      "grad_norm": 0.016250934451818466,
+      "learning_rate": 0.002698019801980198,
+      "loss": 0.6429,
+      "step": 257
+    },
+    {
+      "epoch": 0.5734926368435677,
+      "grad_norm": 0.013874251395463943,
+      "learning_rate": 0.0026955445544554458,
+      "loss": 0.6679,
+      "step": 258
+    },
+    {
+      "epoch": 0.5757154765212559,
+      "grad_norm": 0.020642723888158798,
+      "learning_rate": 0.002693069306930693,
+      "loss": 0.7782,
+      "step": 259
+    },
+    {
+      "epoch": 0.5779383161989442,
+      "grad_norm": 0.018470142036676407,
+      "learning_rate": 0.0026905940594059404,
+      "loss": 0.726,
+      "step": 260
+    },
+    {
+      "epoch": 0.5801611558766324,
+      "grad_norm": 0.01383270975202322,
+      "learning_rate": 0.0026881188118811883,
+      "loss": 0.6343,
+      "step": 261
+    },
+    {
+      "epoch": 0.5823839955543206,
+      "grad_norm": 0.02608071267604828,
+      "learning_rate": 0.0026856435643564354,
+      "loss": 0.7594,
+      "step": 262
+    },
+    {
+      "epoch": 0.5846068352320088,
+      "grad_norm": 0.01618632860481739,
+      "learning_rate": 0.0026831683168316834,
+      "loss": 0.6656,
+      "step": 263
+    },
+    {
+      "epoch": 0.5868296749096972,
+      "grad_norm": 0.022311942651867867,
+      "learning_rate": 0.002680693069306931,
+      "loss": 0.9195,
+      "step": 264
+    },
+    {
+      "epoch": 0.5890525145873854,
+      "grad_norm": 0.01792202889919281,
+      "learning_rate": 0.002678217821782178,
+      "loss": 0.7247,
+      "step": 265
+    },
+    {
+      "epoch": 0.5912753542650736,
+      "grad_norm": 0.017917366698384285,
+      "learning_rate": 0.002675742574257426,
+      "loss": 0.6302,
+      "step": 266
+    },
+    {
+      "epoch": 0.5934981939427619,
+      "grad_norm": 0.022937409579753876,
+      "learning_rate": 0.0026732673267326735,
+      "loss": 0.9526,
+      "step": 267
+    },
+    {
+      "epoch": 0.5957210336204501,
+      "grad_norm": 0.015674270689487457,
+      "learning_rate": 0.0026707920792079206,
+      "loss": 0.7029,
+      "step": 268
+    },
+    {
+      "epoch": 0.5979438732981384,
+      "grad_norm": 0.01607857085764408,
+      "learning_rate": 0.0026683168316831685,
+      "loss": 0.8267,
+      "step": 269
+    },
+    {
+      "epoch": 0.6001667129758266,
+      "grad_norm": 0.018459970131516457,
+      "learning_rate": 0.002665841584158416,
+      "loss": 0.7411,
+      "step": 270
+    },
+    {
+      "epoch": 0.6023895526535149,
+      "grad_norm": 0.01604890450835228,
+      "learning_rate": 0.002663366336633663,
+      "loss": 0.7138,
+      "step": 271
+    },
+    {
+      "epoch": 0.6046123923312031,
+      "grad_norm": 0.021214580163359642,
+      "learning_rate": 0.002660891089108911,
+      "loss": 0.6114,
+      "step": 272
+    },
+    {
+      "epoch": 0.6068352320088913,
+      "grad_norm": 0.019329868257045746,
+      "learning_rate": 0.0026584158415841586,
+      "loss": 0.723,
+      "step": 273
+    },
+    {
+      "epoch": 0.6090580716865797,
+      "grad_norm": 0.012522357515990734,
+      "learning_rate": 0.0026559405940594057,
+      "loss": 0.7007,
+      "step": 274
+    },
+    {
+      "epoch": 0.6112809113642679,
+      "grad_norm": 0.01763787679374218,
+      "learning_rate": 0.0026534653465346537,
+      "loss": 0.7885,
+      "step": 275
+    },
+    {
+      "epoch": 0.6135037510419561,
+      "grad_norm": 0.018133271485567093,
+      "learning_rate": 0.002650990099009901,
+      "loss": 0.7477,
+      "step": 276
+    },
+    {
+      "epoch": 0.6157265907196443,
+      "grad_norm": 0.013666578568518162,
+      "learning_rate": 0.0026485148514851483,
+      "loss": 0.6034,
+      "step": 277
+    },
+    {
+      "epoch": 0.6179494303973326,
+      "grad_norm": 0.014519906602799892,
+      "learning_rate": 0.0026460396039603962,
+      "loss": 0.6982,
+      "step": 278
+    },
+    {
+      "epoch": 0.6201722700750208,
+      "grad_norm": 0.016167080029845238,
+      "learning_rate": 0.0026435643564356437,
+      "loss": 0.7584,
+      "step": 279
+    },
+    {
+      "epoch": 0.6223951097527091,
+      "grad_norm": 0.0192260704934597,
+      "learning_rate": 0.002641089108910891,
+      "loss": 0.8423,
+      "step": 280
+    },
+    {
+      "epoch": 0.6246179494303973,
+      "grad_norm": 0.02769404463469982,
+      "learning_rate": 0.002638613861386139,
+      "loss": 0.7093,
+      "step": 281
+    },
+    {
+      "epoch": 0.6268407891080856,
+      "grad_norm": 0.023452669382095337,
+      "learning_rate": 0.0026361386138613863,
+      "loss": 0.7986,
+      "step": 282
+    },
+    {
+      "epoch": 0.6290636287857738,
+      "grad_norm": 0.02152738720178604,
+      "learning_rate": 0.0026336633663366334,
+      "loss": 0.7953,
+      "step": 283
+    },
+    {
+      "epoch": 0.631286468463462,
+      "grad_norm": 0.022495241835713387,
+      "learning_rate": 0.0026311881188118814,
+      "loss": 0.8022,
+      "step": 284
+    },
+    {
+      "epoch": 0.6335093081411504,
+      "grad_norm": 0.01805182732641697,
+      "learning_rate": 0.002628712871287129,
+      "loss": 0.6848,
+      "step": 285
+    },
+    {
+      "epoch": 0.6357321478188386,
+      "grad_norm": 0.020260373130440712,
+      "learning_rate": 0.002626237623762376,
+      "loss": 0.6499,
+      "step": 286
+    },
+    {
+      "epoch": 0.6379549874965268,
+      "grad_norm": 0.01385380420833826,
+      "learning_rate": 0.002623762376237624,
+      "loss": 0.6989,
+      "step": 287
+    },
+    {
+      "epoch": 0.640177827174215,
+      "grad_norm": 0.015470602549612522,
+      "learning_rate": 0.0026212871287128714,
+      "loss": 0.745,
+      "step": 288
+    },
+    {
+      "epoch": 0.6424006668519033,
+      "grad_norm": 0.011990759521722794,
+      "learning_rate": 0.0026188118811881185,
+      "loss": 0.6064,
+      "step": 289
+    },
+    {
+      "epoch": 0.6446235065295916,
+      "grad_norm": 0.01651688665151596,
+      "learning_rate": 0.0026163366336633665,
+      "loss": 0.6985,
+      "step": 290
+    },
+    {
+      "epoch": 0.6468463462072798,
+      "grad_norm": 0.013933480717241764,
+      "learning_rate": 0.002613861386138614,
+      "loss": 0.6943,
+      "step": 291
+    },
+    {
+      "epoch": 0.6490691858849681,
+      "grad_norm": 0.014627222903072834,
+      "learning_rate": 0.002611386138613861,
+      "loss": 0.5978,
+      "step": 292
+    },
+    {
+      "epoch": 0.6512920255626563,
+      "grad_norm": 0.016101978719234467,
+      "learning_rate": 0.002608910891089109,
+      "loss": 0.7254,
+      "step": 293
+    },
+    {
+      "epoch": 0.6535148652403445,
+      "grad_norm": 0.012672492302954197,
+      "learning_rate": 0.0026064356435643566,
+      "loss": 0.686,
+      "step": 294
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 0.019365599378943443,
+      "learning_rate": 0.0026039603960396037,
+      "loss": 0.6987,
+      "step": 295
+    },
+    {
+      "epoch": 0.6579605445957211,
+      "grad_norm": 0.013569362461566925,
+      "learning_rate": 0.0026014851485148516,
+      "loss": 0.6362,
+      "step": 296
+    },
+    {
+      "epoch": 0.6601833842734093,
+      "grad_norm": 0.016133926808834076,
+      "learning_rate": 0.002599009900990099,
+      "loss": 0.6661,
+      "step": 297
+    },
+    {
+      "epoch": 0.6624062239510975,
+      "grad_norm": 0.016300931572914124,
+      "learning_rate": 0.0025965346534653462,
+      "loss": 0.64,
+      "step": 298
+    },
+    {
+      "epoch": 0.6646290636287858,
+      "grad_norm": 0.012371404096484184,
+      "learning_rate": 0.002594059405940594,
+      "loss": 0.6825,
+      "step": 299
+    },
+    {
+      "epoch": 0.666851903306474,
+      "grad_norm": 0.018244942650198936,
+      "learning_rate": 0.0025915841584158417,
+      "loss": 0.735,
+      "step": 300
+    },
+    {
+      "epoch": 0.6690747429841623,
+      "grad_norm": 0.05636992305517197,
+      "learning_rate": 0.002589108910891089,
+      "loss": 0.7862,
+      "step": 301
+    },
+    {
+      "epoch": 0.6712975826618505,
+      "grad_norm": 0.01677551493048668,
+      "learning_rate": 0.0025866336633663368,
+      "loss": 0.7861,
+      "step": 302
+    },
+    {
+      "epoch": 0.6735204223395388,
+      "grad_norm": 0.01706579327583313,
+      "learning_rate": 0.0025841584158415843,
+      "loss": 0.8371,
+      "step": 303
+    },
+    {
+      "epoch": 0.675743262017227,
+      "grad_norm": 0.014940536580979824,
+      "learning_rate": 0.0025816831683168314,
+      "loss": 0.6754,
+      "step": 304
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 0.013865659013390541,
+      "learning_rate": 0.0025792079207920793,
+      "loss": 0.9877,
+      "step": 305
+    },
+    {
+      "epoch": 0.6801889413726036,
+      "grad_norm": 0.010642016306519508,
+      "learning_rate": 0.002576732673267327,
+      "loss": 0.6201,
+      "step": 306
+    },
+    {
+      "epoch": 0.6824117810502918,
+      "grad_norm": 0.01661493070423603,
+      "learning_rate": 0.002574257425742574,
+      "loss": 0.6654,
+      "step": 307
+    },
+    {
+      "epoch": 0.68463462072798,
+      "grad_norm": 0.018468208611011505,
+      "learning_rate": 0.002571782178217822,
+      "loss": 0.7863,
+      "step": 308
+    },
+    {
+      "epoch": 0.6868574604056682,
+      "grad_norm": 0.022226519882678986,
+      "learning_rate": 0.0025693069306930694,
+      "loss": 0.7191,
+      "step": 309
+    },
+    {
+      "epoch": 0.6890803000833565,
+      "grad_norm": 0.019259922206401825,
+      "learning_rate": 0.0025668316831683165,
+      "loss": 0.8028,
+      "step": 310
+    },
+    {
+      "epoch": 0.6913031397610447,
+      "grad_norm": 0.017490798607468605,
+      "learning_rate": 0.0025643564356435645,
+      "loss": 0.7814,
+      "step": 311
+    },
+    {
+      "epoch": 0.693525979438733,
+      "grad_norm": 0.02297484688460827,
+      "learning_rate": 0.002561881188118812,
+      "loss": 0.6705,
+      "step": 312
+    },
+    {
+      "epoch": 0.6957488191164213,
+      "grad_norm": 0.01425850111991167,
+      "learning_rate": 0.0025594059405940595,
+      "loss": 0.7235,
+      "step": 313
+    },
+    {
+      "epoch": 0.6979716587941095,
+      "grad_norm": 0.012471362948417664,
+      "learning_rate": 0.002556930693069307,
+      "loss": 0.7424,
+      "step": 314
+    },
+    {
+      "epoch": 0.7001944984717977,
+      "grad_norm": 0.013804659247398376,
+      "learning_rate": 0.0025544554455445546,
+      "loss": 0.6745,
+      "step": 315
+    },
+    {
+      "epoch": 0.7024173381494859,
+      "grad_norm": 0.01753597892820835,
+      "learning_rate": 0.002551980198019802,
+      "loss": 0.7136,
+      "step": 316
+    },
+    {
+      "epoch": 0.7046401778271743,
+      "grad_norm": 0.01735616661608219,
+      "learning_rate": 0.0025495049504950496,
+      "loss": 0.6987,
+      "step": 317
+    },
+    {
+      "epoch": 0.7068630175048625,
+      "grad_norm": 0.013026660308241844,
+      "learning_rate": 0.002547029702970297,
+      "loss": 0.5521,
+      "step": 318
+    },
+    {
+      "epoch": 0.7090858571825507,
+      "grad_norm": 0.011274688877165318,
+      "learning_rate": 0.0025445544554455446,
+      "loss": 0.5929,
+      "step": 319
+    },
+    {
+      "epoch": 0.711308696860239,
+      "grad_norm": 0.013325052335858345,
+      "learning_rate": 0.002542079207920792,
+      "loss": 0.6723,
+      "step": 320
+    },
+    {
+      "epoch": 0.7135315365379272,
+      "grad_norm": 0.012779835611581802,
+      "learning_rate": 0.0025396039603960397,
+      "loss": 0.6871,
+      "step": 321
+    },
+    {
+      "epoch": 0.7157543762156154,
+      "grad_norm": 0.019513219594955444,
+      "learning_rate": 0.002537128712871287,
+      "loss": 0.7549,
+      "step": 322
+    },
+    {
+      "epoch": 0.7179772158933037,
+      "grad_norm": 0.014735296368598938,
+      "learning_rate": 0.0025346534653465347,
+      "loss": 0.7868,
+      "step": 323
+    },
+    {
+      "epoch": 0.720200055570992,
+      "grad_norm": 0.015642767772078514,
+      "learning_rate": 0.0025321782178217823,
+      "loss": 0.6612,
+      "step": 324
+    },
+    {
+      "epoch": 0.7224228952486802,
+      "grad_norm": 0.013669556006789207,
+      "learning_rate": 0.0025297029702970298,
+      "loss": 0.6112,
+      "step": 325
+    },
+    {
+      "epoch": 0.7246457349263684,
+      "grad_norm": 0.012594704516232014,
+      "learning_rate": 0.0025272277227722773,
+      "loss": 0.7393,
+      "step": 326
+    },
+    {
+      "epoch": 0.7268685746040566,
+      "grad_norm": 0.015079697594046593,
+      "learning_rate": 0.002524752475247525,
+      "loss": 0.793,
+      "step": 327
+    },
+    {
+      "epoch": 0.729091414281745,
+      "grad_norm": 0.013951257802546024,
+      "learning_rate": 0.0025222772277227723,
+      "loss": 0.6706,
+      "step": 328
+    },
+    {
+      "epoch": 0.7313142539594332,
+      "grad_norm": 0.019177842885255814,
+      "learning_rate": 0.00251980198019802,
+      "loss": 0.7779,
+      "step": 329
+    },
+    {
+      "epoch": 0.7335370936371214,
+      "grad_norm": 0.013305417262017727,
+      "learning_rate": 0.0025173267326732674,
+      "loss": 0.7178,
+      "step": 330
+    },
+    {
+      "epoch": 0.7357599333148097,
+      "grad_norm": 0.02146240696310997,
+      "learning_rate": 0.002514851485148515,
+      "loss": 0.7293,
+      "step": 331
+    },
+    {
+      "epoch": 0.7379827729924979,
+      "grad_norm": 0.013251793570816517,
+      "learning_rate": 0.0025123762376237624,
+      "loss": 0.7747,
+      "step": 332
+    },
+    {
+      "epoch": 0.7402056126701861,
+      "grad_norm": 0.01089315302670002,
+      "learning_rate": 0.00250990099009901,
+      "loss": 0.5445,
+      "step": 333
+    },
+    {
+      "epoch": 0.7424284523478744,
+      "grad_norm": 0.01445831824094057,
+      "learning_rate": 0.0025074257425742575,
+      "loss": 0.6669,
+      "step": 334
+    },
+    {
+      "epoch": 0.7446512920255627,
+      "grad_norm": 0.014526651240885258,
+      "learning_rate": 0.002504950495049505,
+      "loss": 0.7194,
+      "step": 335
+    },
+    {
+      "epoch": 0.7468741317032509,
+      "grad_norm": 0.018426531925797462,
+      "learning_rate": 0.0025024752475247525,
+      "loss": 0.8315,
+      "step": 336
+    },
+    {
+      "epoch": 0.7490969713809391,
+      "grad_norm": 0.01586451753973961,
+      "learning_rate": 0.0025,
+      "loss": 0.572,
+      "step": 337
+    },
+    {
+      "epoch": 0.7513198110586274,
+      "grad_norm": 0.014794967137277126,
+      "learning_rate": 0.0024975247524752476,
+      "loss": 0.6875,
+      "step": 338
+    },
+    {
+      "epoch": 0.7535426507363157,
+      "grad_norm": 0.013549881055951118,
+      "learning_rate": 0.002495049504950495,
+      "loss": 0.6709,
+      "step": 339
+    },
+    {
+      "epoch": 0.7557654904140039,
+      "grad_norm": 0.0105968052521348,
+      "learning_rate": 0.0024925742574257426,
+      "loss": 0.5964,
+      "step": 340
+    },
+    {
+      "epoch": 0.7579883300916921,
+      "grad_norm": 0.015588056296110153,
+      "learning_rate": 0.00249009900990099,
+      "loss": 0.6671,
+      "step": 341
+    },
+    {
+      "epoch": 0.7602111697693804,
+      "grad_norm": 0.01798832416534424,
+      "learning_rate": 0.0024876237623762377,
+      "loss": 0.8818,
+      "step": 342
+    },
+    {
+      "epoch": 0.7624340094470686,
+      "grad_norm": 0.013795461505651474,
+      "learning_rate": 0.002485148514851485,
+      "loss": 0.5748,
+      "step": 343
+    },
+    {
+      "epoch": 0.7646568491247568,
+      "grad_norm": 0.016307495534420013,
+      "learning_rate": 0.0024826732673267327,
+      "loss": 0.7323,
+      "step": 344
+    },
+    {
+      "epoch": 0.7668796888024452,
+      "grad_norm": 0.012711350806057453,
+      "learning_rate": 0.0024801980198019802,
+      "loss": 0.5185,
+      "step": 345
+    },
+    {
+      "epoch": 0.7691025284801334,
+      "grad_norm": 0.01902935467660427,
+      "learning_rate": 0.0024777227722772278,
+      "loss": 0.7524,
+      "step": 346
+    },
+    {
+      "epoch": 0.7713253681578216,
+      "grad_norm": 0.013338466174900532,
+      "learning_rate": 0.0024752475247524753,
+      "loss": 0.6606,
+      "step": 347
+    },
+    {
+      "epoch": 0.7735482078355098,
+      "grad_norm": 0.017752062529325485,
+      "learning_rate": 0.002472772277227723,
+      "loss": 0.85,
+      "step": 348
+    },
+    {
+      "epoch": 0.7757710475131981,
+      "grad_norm": 0.013239924795925617,
+      "learning_rate": 0.0024702970297029703,
+      "loss": 0.623,
+      "step": 349
+    },
+    {
+      "epoch": 0.7779938871908864,
+      "grad_norm": 0.017011214047670364,
+      "learning_rate": 0.002467821782178218,
+      "loss": 0.6873,
+      "step": 350
+    },
+    {
+      "epoch": 0.7802167268685746,
+      "grad_norm": 0.015225573442876339,
+      "learning_rate": 0.0024653465346534654,
+      "loss": 0.6334,
+      "step": 351
+    },
+    {
+      "epoch": 0.7824395665462629,
+      "grad_norm": 0.020225588232278824,
+      "learning_rate": 0.002462871287128713,
+      "loss": 0.7056,
+      "step": 352
+    },
+    {
+      "epoch": 0.7846624062239511,
+      "grad_norm": 0.01229569036513567,
+      "learning_rate": 0.0024603960396039604,
+      "loss": 0.7022,
+      "step": 353
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 0.013165128417313099,
+      "learning_rate": 0.002457920792079208,
+      "loss": 0.6532,
+      "step": 354
+    },
+    {
+      "epoch": 0.7891080855793275,
+      "grad_norm": 0.026388289406895638,
+      "learning_rate": 0.0024554455445544555,
+      "loss": 0.6812,
+      "step": 355
+    },
+    {
+      "epoch": 0.7913309252570159,
+      "grad_norm": 0.014291019178926945,
+      "learning_rate": 0.002452970297029703,
+      "loss": 0.7062,
+      "step": 356
+    },
+    {
+      "epoch": 0.7935537649347041,
+      "grad_norm": 0.017437303438782692,
+      "learning_rate": 0.0024504950495049505,
+      "loss": 0.7651,
+      "step": 357
+    },
+    {
+      "epoch": 0.7957766046123923,
+      "grad_norm": 0.016056453809142113,
+      "learning_rate": 0.002448019801980198,
+      "loss": 0.6909,
+      "step": 358
+    },
+    {
+      "epoch": 0.7979994442900806,
+      "grad_norm": 0.017870964482426643,
+      "learning_rate": 0.0024455445544554455,
+      "loss": 0.541,
+      "step": 359
+    },
+    {
+      "epoch": 0.8002222839677688,
+      "grad_norm": 0.014736250974237919,
+      "learning_rate": 0.002443069306930693,
+      "loss": 0.7619,
+      "step": 360
+    },
+    {
+      "epoch": 0.8024451236454571,
+      "grad_norm": 0.013172541745007038,
+      "learning_rate": 0.0024405940594059406,
+      "loss": 0.5607,
+      "step": 361
+    },
+    {
+      "epoch": 0.8046679633231453,
+      "grad_norm": 0.017459526658058167,
+      "learning_rate": 0.002438118811881188,
+      "loss": 0.7483,
+      "step": 362
+    },
+    {
+      "epoch": 0.8068908030008336,
+      "grad_norm": 0.016839707270264626,
+      "learning_rate": 0.002435643564356436,
+      "loss": 0.7191,
+      "step": 363
+    },
+    {
+      "epoch": 0.8091136426785218,
+      "grad_norm": 0.017826620489358902,
+      "learning_rate": 0.002433168316831683,
+      "loss": 0.8137,
+      "step": 364
+    },
+    {
+      "epoch": 0.81133648235621,
+      "grad_norm": 0.019586768001317978,
+      "learning_rate": 0.0024306930693069307,
+      "loss": 0.8679,
+      "step": 365
+    },
+    {
+      "epoch": 0.8135593220338984,
+      "grad_norm": 0.03678233548998833,
+      "learning_rate": 0.0024282178217821786,
+      "loss": 0.6402,
+      "step": 366
+    },
+    {
+      "epoch": 0.8157821617115866,
+      "grad_norm": 0.02029457874596119,
+      "learning_rate": 0.0024257425742574257,
+      "loss": 0.794,
+      "step": 367
+    },
+    {
+      "epoch": 0.8180050013892748,
+      "grad_norm": 0.014588624238967896,
+      "learning_rate": 0.0024232673267326733,
+      "loss": 0.5804,
+      "step": 368
+    },
+    {
+      "epoch": 0.820227841066963,
+      "grad_norm": 0.014623894356191158,
+      "learning_rate": 0.002420792079207921,
+      "loss": 0.6849,
+      "step": 369
+    },
+    {
+      "epoch": 0.8224506807446513,
+      "grad_norm": 0.014702382497489452,
+      "learning_rate": 0.0024183168316831683,
+      "loss": 0.6318,
+      "step": 370
+    },
+    {
+      "epoch": 0.8246735204223395,
+      "grad_norm": 0.026238251477479935,
+      "learning_rate": 0.002415841584158416,
+      "loss": 0.6903,
+      "step": 371
+    },
+    {
+      "epoch": 0.8268963601000278,
+      "grad_norm": 0.017133679240942,
+      "learning_rate": 0.0024133663366336638,
+      "loss": 0.6215,
+      "step": 372
+    },
+    {
+      "epoch": 0.8291191997777161,
+      "grad_norm": 0.011881059035658836,
+      "learning_rate": 0.002410891089108911,
+      "loss": 0.6326,
+      "step": 373
+    },
+    {
+      "epoch": 0.8313420394554043,
+      "grad_norm": 0.015931686386466026,
+      "learning_rate": 0.0024084158415841584,
+      "loss": 0.5843,
+      "step": 374
+    },
+    {
+      "epoch": 0.8335648791330925,
+      "grad_norm": 0.014707183465361595,
+      "learning_rate": 0.0024059405940594063,
+      "loss": 0.5787,
+      "step": 375
+    },
+    {
+      "epoch": 0.8357877188107807,
+      "grad_norm": 0.015076966024935246,
+      "learning_rate": 0.0024034653465346534,
+      "loss": 0.6184,
+      "step": 376
+    },
+    {
+      "epoch": 0.8380105584884691,
+      "grad_norm": 0.015617873519659042,
+      "learning_rate": 0.002400990099009901,
+      "loss": 0.6095,
+      "step": 377
+    },
+    {
+      "epoch": 0.8402333981661573,
+      "grad_norm": 0.016791800037026405,
+      "learning_rate": 0.002398514851485149,
+      "loss": 0.7002,
+      "step": 378
+    },
+    {
+      "epoch": 0.8424562378438455,
+      "grad_norm": 0.01942884549498558,
+      "learning_rate": 0.002396039603960396,
+      "loss": 0.6643,
+      "step": 379
+    },
+    {
+      "epoch": 0.8446790775215337,
+      "grad_norm": 0.02692730724811554,
+      "learning_rate": 0.0023935643564356435,
+      "loss": 0.6722,
+      "step": 380
+    },
+    {
+      "epoch": 0.846901917199222,
+      "grad_norm": 0.016538864001631737,
+      "learning_rate": 0.0023910891089108915,
+      "loss": 0.592,
+      "step": 381
+    },
+    {
+      "epoch": 0.8491247568769102,
+      "grad_norm": 0.0320366770029068,
+      "learning_rate": 0.0023886138613861386,
+      "loss": 0.6845,
+      "step": 382
+    },
+    {
+      "epoch": 0.8513475965545985,
+      "grad_norm": 0.011704586446285248,
+      "learning_rate": 0.002386138613861386,
+      "loss": 0.6338,
+      "step": 383
+    },
+    {
+      "epoch": 0.8535704362322868,
+      "grad_norm": 0.012988455593585968,
+      "learning_rate": 0.002383663366336634,
+      "loss": 0.6345,
+      "step": 384
+    },
+    {
+      "epoch": 0.855793275909975,
+      "grad_norm": 0.036786969751119614,
+      "learning_rate": 0.002381188118811881,
+      "loss": 0.6513,
+      "step": 385
+    },
+    {
+      "epoch": 0.8580161155876632,
+      "grad_norm": 0.01267944648861885,
+      "learning_rate": 0.0023787128712871287,
+      "loss": 0.6335,
+      "step": 386
+    },
+    {
+      "epoch": 0.8602389552653514,
+      "grad_norm": 0.021348008885979652,
+      "learning_rate": 0.0023762376237623766,
+      "loss": 0.6572,
+      "step": 387
+    },
+    {
+      "epoch": 0.8624617949430398,
+      "grad_norm": 0.016306119039654732,
+      "learning_rate": 0.0023737623762376237,
+      "loss": 0.6352,
+      "step": 388
+    },
+    {
+      "epoch": 0.864684634620728,
+      "grad_norm": 0.016549356281757355,
+      "learning_rate": 0.0023712871287128712,
+      "loss": 0.6428,
+      "step": 389
+    },
+    {
+      "epoch": 0.8669074742984162,
+      "grad_norm": 0.01346430741250515,
+      "learning_rate": 0.002368811881188119,
+      "loss": 0.6152,
+      "step": 390
+    },
+    {
+      "epoch": 0.8691303139761045,
+      "grad_norm": 0.01475781761109829,
+      "learning_rate": 0.0023663366336633663,
+      "loss": 0.6869,
+      "step": 391
+    },
+    {
+      "epoch": 0.8713531536537927,
+      "grad_norm": 0.021658863872289658,
+      "learning_rate": 0.002363861386138614,
+      "loss": 0.8617,
+      "step": 392
+    },
+    {
+      "epoch": 0.873575993331481,
+      "grad_norm": 0.014893801882863045,
+      "learning_rate": 0.0023613861386138617,
+      "loss": 0.6334,
+      "step": 393
+    },
+    {
+      "epoch": 0.8757988330091692,
+      "grad_norm": 0.015763310715556145,
+      "learning_rate": 0.002358910891089109,
+      "loss": 0.6145,
+      "step": 394
+    },
+    {
+      "epoch": 0.8780216726868575,
+      "grad_norm": 0.01235837209969759,
+      "learning_rate": 0.0023564356435643564,
+      "loss": 0.6377,
+      "step": 395
+    },
+    {
+      "epoch": 0.8802445123645457,
+      "grad_norm": 0.08386275917291641,
+      "learning_rate": 0.0023539603960396043,
+      "loss": 0.6466,
+      "step": 396
+    },
+    {
+      "epoch": 0.8824673520422339,
+      "grad_norm": 0.024875586852431297,
+      "learning_rate": 0.0023514851485148514,
+      "loss": 0.772,
+      "step": 397
+    },
+    {
+      "epoch": 0.8846901917199222,
+      "grad_norm": 0.022778382524847984,
+      "learning_rate": 0.002349009900990099,
+      "loss": 0.77,
+      "step": 398
+    },
+    {
+      "epoch": 0.8869130313976105,
+      "grad_norm": 0.01597260683774948,
+      "learning_rate": 0.002346534653465347,
+      "loss": 0.8057,
+      "step": 399
+    },
+    {
+      "epoch": 0.8891358710752987,
+      "grad_norm": 0.023420272395014763,
+      "learning_rate": 0.002344059405940594,
+      "loss": 0.6152,
+      "step": 400
+    },
+    {
+      "epoch": 0.8913587107529869,
+      "grad_norm": 0.014347629621624947,
+      "learning_rate": 0.0023415841584158415,
+      "loss": 0.6851,
+      "step": 401
+    },
+    {
+      "epoch": 0.8935815504306752,
+      "grad_norm": 0.015029153786599636,
+      "learning_rate": 0.0023391089108910895,
+      "loss": 0.667,
+      "step": 402
+    },
+    {
+      "epoch": 0.8958043901083634,
+      "grad_norm": 0.014809265732765198,
+      "learning_rate": 0.0023366336633663365,
+      "loss": 0.5953,
+      "step": 403
+    },
+    {
+      "epoch": 0.8980272297860517,
+      "grad_norm": 0.015516281127929688,
+      "learning_rate": 0.002334158415841584,
+      "loss": 0.7201,
+      "step": 404
+    },
+    {
+      "epoch": 0.90025006946374,
+      "grad_norm": 0.016222365200519562,
+      "learning_rate": 0.002331683168316832,
+      "loss": 0.6713,
+      "step": 405
+    },
+    {
+      "epoch": 0.9024729091414282,
+      "grad_norm": 0.017578039318323135,
+      "learning_rate": 0.002329207920792079,
+      "loss": 0.5812,
+      "step": 406
+    },
+    {
+      "epoch": 0.9046957488191164,
+      "grad_norm": 0.018017487600445747,
+      "learning_rate": 0.0023267326732673266,
+      "loss": 0.6542,
+      "step": 407
+    },
+    {
+      "epoch": 0.9069185884968046,
+      "grad_norm": 0.020421525463461876,
+      "learning_rate": 0.0023242574257425746,
+      "loss": 0.8204,
+      "step": 408
+    },
+    {
+      "epoch": 0.909141428174493,
+      "grad_norm": 0.016109053045511246,
+      "learning_rate": 0.0023217821782178217,
+      "loss": 0.6965,
+      "step": 409
+    },
+    {
+      "epoch": 0.9113642678521812,
+      "grad_norm": 0.014515829272568226,
+      "learning_rate": 0.002319306930693069,
+      "loss": 0.6253,
+      "step": 410
+    },
+    {
+      "epoch": 0.9135871075298694,
+      "grad_norm": 0.021663261577486992,
+      "learning_rate": 0.002316831683168317,
+      "loss": 0.8357,
+      "step": 411
+    },
+    {
+      "epoch": 0.9158099472075577,
+      "grad_norm": 0.014531668275594711,
+      "learning_rate": 0.0023143564356435642,
+      "loss": 0.6608,
+      "step": 412
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 0.018239939585328102,
+      "learning_rate": 0.002311881188118812,
+      "loss": 0.7747,
+      "step": 413
+    },
+    {
+      "epoch": 0.9202556265629341,
+      "grad_norm": 0.012773513793945312,
+      "learning_rate": 0.0023094059405940597,
+      "loss": 0.5356,
+      "step": 414
+    },
+    {
+      "epoch": 0.9224784662406224,
+      "grad_norm": 0.012568398378789425,
+      "learning_rate": 0.002306930693069307,
+      "loss": 0.6577,
+      "step": 415
+    },
+    {
+      "epoch": 0.9247013059183107,
+      "grad_norm": 0.017879627645015717,
+      "learning_rate": 0.0023044554455445548,
+      "loss": 0.5712,
+      "step": 416
+    },
+    {
+      "epoch": 0.9269241455959989,
+      "grad_norm": 0.0374915786087513,
+      "learning_rate": 0.0023019801980198023,
+      "loss": 0.6147,
+      "step": 417
+    },
+    {
+      "epoch": 0.9291469852736871,
+      "grad_norm": 0.01358667854219675,
+      "learning_rate": 0.0022995049504950494,
+      "loss": 0.6152,
+      "step": 418
+    },
+    {
+      "epoch": 0.9313698249513754,
+      "grad_norm": 0.013912348076701164,
+      "learning_rate": 0.0022970297029702973,
+      "loss": 0.6511,
+      "step": 419
+    },
+    {
+      "epoch": 0.9335926646290637,
+      "grad_norm": 0.014229892753064632,
+      "learning_rate": 0.002294554455445545,
+      "loss": 0.6068,
+      "step": 420
+    },
+    {
+      "epoch": 0.9358155043067519,
+      "grad_norm": 0.017046229913830757,
+      "learning_rate": 0.002292079207920792,
+      "loss": 0.7221,
+      "step": 421
+    },
+    {
+      "epoch": 0.9380383439844401,
+      "grad_norm": 0.013679401949048042,
+      "learning_rate": 0.00228960396039604,
+      "loss": 0.6077,
+      "step": 422
+    },
+    {
+      "epoch": 0.9402611836621284,
+      "grad_norm": 0.018443197011947632,
+      "learning_rate": 0.0022871287128712874,
+      "loss": 0.7146,
+      "step": 423
+    },
+    {
+      "epoch": 0.9424840233398166,
+      "grad_norm": 0.021021803840994835,
+      "learning_rate": 0.0022846534653465345,
+      "loss": 0.829,
+      "step": 424
+    },
+    {
+      "epoch": 0.9447068630175048,
+      "grad_norm": 0.013253494165837765,
+      "learning_rate": 0.0022821782178217825,
+      "loss": 0.6044,
+      "step": 425
+    },
+    {
+      "epoch": 0.946929702695193,
+      "grad_norm": 0.015879614278674126,
+      "learning_rate": 0.00227970297029703,
+      "loss": 0.6039,
+      "step": 426
+    },
+    {
+      "epoch": 0.9491525423728814,
+      "grad_norm": 0.02068069390952587,
+      "learning_rate": 0.002277227722772277,
+      "loss": 0.7617,
+      "step": 427
+    },
+    {
+      "epoch": 0.9513753820505696,
+      "grad_norm": 0.013184352777898312,
+      "learning_rate": 0.002274752475247525,
+      "loss": 0.6523,
+      "step": 428
+    },
+    {
+      "epoch": 0.9535982217282578,
+      "grad_norm": 0.0113218380138278,
+      "learning_rate": 0.0022722772277227726,
+      "loss": 0.6023,
+      "step": 429
+    },
+    {
+      "epoch": 0.9558210614059461,
+      "grad_norm": 0.018204597756266594,
+      "learning_rate": 0.0022698019801980196,
+      "loss": 0.6851,
+      "step": 430
+    },
+    {
+      "epoch": 0.9580439010836344,
+      "grad_norm": 0.024960225448012352,
+      "learning_rate": 0.0022673267326732676,
+      "loss": 0.7379,
+      "step": 431
+    },
+    {
+      "epoch": 0.9602667407613226,
+      "grad_norm": 0.01937321573495865,
+      "learning_rate": 0.002264851485148515,
+      "loss": 0.6909,
+      "step": 432
+    },
+    {
+      "epoch": 0.9624895804390108,
+      "grad_norm": 0.017059147357940674,
+      "learning_rate": 0.0022623762376237622,
+      "loss": 0.7239,
+      "step": 433
+    },
+    {
+      "epoch": 0.9647124201166991,
+      "grad_norm": 0.011201994493603706,
+      "learning_rate": 0.00225990099009901,
+      "loss": 0.5845,
+      "step": 434
+    },
+    {
+      "epoch": 0.9669352597943873,
+      "grad_norm": 0.01506833452731371,
+      "learning_rate": 0.0022574257425742577,
+      "loss": 0.7139,
+      "step": 435
+    },
+    {
+      "epoch": 0.9691580994720755,
+      "grad_norm": 0.017956897616386414,
+      "learning_rate": 0.002254950495049505,
+      "loss": 0.6997,
+      "step": 436
+    },
+    {
+      "epoch": 0.9713809391497639,
+      "grad_norm": 0.018598267808556557,
+      "learning_rate": 0.0022524752475247527,
+      "loss": 0.5882,
+      "step": 437
+    },
+    {
+      "epoch": 0.9736037788274521,
+      "grad_norm": 0.016982879489660263,
+      "learning_rate": 0.0022500000000000003,
+      "loss": 0.5976,
+      "step": 438
+    },
+    {
+      "epoch": 0.9758266185051403,
+      "grad_norm": 0.01876719854772091,
+      "learning_rate": 0.0022475247524752474,
+      "loss": 0.7,
+      "step": 439
+    },
+    {
+      "epoch": 0.9780494581828285,
+      "grad_norm": 0.020731210708618164,
+      "learning_rate": 0.0022450495049504953,
+      "loss": 0.705,
+      "step": 440
+    },
+    {
+      "epoch": 0.9802722978605168,
+      "grad_norm": 0.014992648735642433,
+      "learning_rate": 0.0022425742574257424,
+      "loss": 0.6393,
+      "step": 441
+    },
+    {
+      "epoch": 0.982495137538205,
+      "grad_norm": 0.012561212293803692,
+      "learning_rate": 0.00224009900990099,
+      "loss": 0.6634,
+      "step": 442
+    },
+    {
+      "epoch": 0.9847179772158933,
+      "grad_norm": 0.019787229597568512,
+      "learning_rate": 0.002237623762376238,
+      "loss": 0.6422,
+      "step": 443
+    },
+    {
+      "epoch": 0.9869408168935816,
+      "grad_norm": 0.015230356715619564,
+      "learning_rate": 0.002235148514851485,
+      "loss": 0.6319,
+      "step": 444
+    },
+    {
+      "epoch": 0.9891636565712698,
+      "grad_norm": 0.013144402764737606,
+      "learning_rate": 0.0022326732673267325,
+      "loss": 0.5625,
+      "step": 445
+    },
+    {
+      "epoch": 0.991386496248958,
+      "grad_norm": 0.010007484816014767,
+      "learning_rate": 0.0022301980198019804,
+      "loss": 0.5449,
+      "step": 446
+    },
+    {
+      "epoch": 0.9936093359266462,
+      "grad_norm": 0.014984616078436375,
+      "learning_rate": 0.0022277227722772275,
+      "loss": 0.7016,
+      "step": 447
+    },
+    {
+      "epoch": 0.9958321756043346,
+      "grad_norm": 0.014676968567073345,
+      "learning_rate": 0.002225247524752475,
+      "loss": 0.6125,
+      "step": 448
+    },
+    {
+      "epoch": 0.9980550152820228,
+      "grad_norm": 0.01267558429390192,
+      "learning_rate": 0.002222772277227723,
+      "loss": 0.6128,
+      "step": 449
+    },
+    {
+      "epoch": 1.000277854959711,
+      "grad_norm": 0.014883238822221756,
+      "learning_rate": 0.00222029702970297,
+      "loss": 0.6702,
+      "step": 450
+    },
+    {
+      "epoch": 1.0025006946373993,
+      "grad_norm": 0.020183809101581573,
+      "learning_rate": 0.0022178217821782176,
+      "loss": 0.7836,
+      "step": 451
+    },
+    {
+      "epoch": 1.0047235343150875,
+      "grad_norm": 0.018987402319908142,
+      "learning_rate": 0.0022153465346534656,
+      "loss": 0.6474,
+      "step": 452
+    },
+    {
+      "epoch": 1.0069463739927758,
+      "grad_norm": 0.013754444196820259,
+      "learning_rate": 0.0022128712871287127,
+      "loss": 0.5332,
+      "step": 453
+    },
+    {
+      "epoch": 1.009169213670464,
+      "grad_norm": 0.024167753756046295,
+      "learning_rate": 0.00221039603960396,
+      "loss": 0.6738,
+      "step": 454
+    },
+    {
+      "epoch": 1.0113920533481522,
+      "grad_norm": 0.013680294156074524,
+      "learning_rate": 0.002207920792079208,
+      "loss": 0.6034,
+      "step": 455
+    },
+    {
+      "epoch": 1.0136148930258404,
+      "grad_norm": 0.01513979583978653,
+      "learning_rate": 0.0022054455445544552,
+      "loss": 0.6736,
+      "step": 456
+    },
+    {
+      "epoch": 1.0158377327035288,
+      "grad_norm": 0.01789553090929985,
+      "learning_rate": 0.0022029702970297028,
+      "loss": 0.5954,
+      "step": 457
+    },
+    {
+      "epoch": 1.018060572381217,
+      "grad_norm": 0.01587206870317459,
+      "learning_rate": 0.0022004950495049507,
+      "loss": 0.6374,
+      "step": 458
+    },
+    {
+      "epoch": 1.0202834120589053,
+      "grad_norm": 0.013578852638602257,
+      "learning_rate": 0.002198019801980198,
+      "loss": 0.6105,
+      "step": 459
+    },
+    {
+      "epoch": 1.0225062517365935,
+      "grad_norm": 0.01525961235165596,
+      "learning_rate": 0.0021955445544554453,
+      "loss": 0.5187,
+      "step": 460
+    },
+    {
+      "epoch": 1.0247290914142817,
+      "grad_norm": 0.01209926512092352,
+      "learning_rate": 0.0021930693069306933,
+      "loss": 0.5858,
+      "step": 461
+    },
+    {
+      "epoch": 1.02695193109197,
+      "grad_norm": 0.013947585597634315,
+      "learning_rate": 0.0021905940594059404,
+      "loss": 0.7486,
+      "step": 462
+    },
+    {
+      "epoch": 1.0291747707696581,
+      "grad_norm": 0.013879354111850262,
+      "learning_rate": 0.002188118811881188,
+      "loss": 0.5767,
+      "step": 463
+    },
+    {
+      "epoch": 1.0313976104473466,
+      "grad_norm": 0.015864117071032524,
+      "learning_rate": 0.002185643564356436,
+      "loss": 0.5945,
+      "step": 464
+    },
+    {
+      "epoch": 1.0336204501250348,
+      "grad_norm": 0.012343275360763073,
+      "learning_rate": 0.002183168316831683,
+      "loss": 0.6243,
+      "step": 465
+    },
+    {
+      "epoch": 1.035843289802723,
+      "grad_norm": 0.011329158209264278,
+      "learning_rate": 0.002180693069306931,
+      "loss": 0.7206,
+      "step": 466
+    },
+    {
+      "epoch": 1.0380661294804112,
+      "grad_norm": 0.014862005598843098,
+      "learning_rate": 0.0021782178217821784,
+      "loss": 0.7483,
+      "step": 467
+    },
+    {
+      "epoch": 1.0402889691580994,
+      "grad_norm": 0.013391937129199505,
+      "learning_rate": 0.0021757425742574255,
+      "loss": 0.7626,
+      "step": 468
+    },
+    {
+      "epoch": 1.0425118088357876,
+      "grad_norm": 0.018975989893078804,
+      "learning_rate": 0.0021732673267326735,
+      "loss": 0.7409,
+      "step": 469
+    },
+    {
+      "epoch": 1.0447346485134759,
+      "grad_norm": 0.01241536159068346,
+      "learning_rate": 0.002170792079207921,
+      "loss": 0.753,
+      "step": 470
+    },
+    {
+      "epoch": 1.0469574881911643,
+      "grad_norm": 0.015430880710482597,
+      "learning_rate": 0.002168316831683168,
+      "loss": 0.776,
+      "step": 471
+    },
+    {
+      "epoch": 1.0491803278688525,
+      "grad_norm": 0.014778917655348778,
+      "learning_rate": 0.002165841584158416,
+      "loss": 0.7363,
+      "step": 472
+    },
+    {
+      "epoch": 1.0514031675465407,
+      "grad_norm": 0.016265999525785446,
+      "learning_rate": 0.0021633663366336636,
+      "loss": 0.6905,
+      "step": 473
+    },
+    {
+      "epoch": 1.053626007224229,
+      "grad_norm": 0.01429293118417263,
+      "learning_rate": 0.0021608910891089106,
+      "loss": 0.6937,
+      "step": 474
+    },
+    {
+      "epoch": 1.0558488469019172,
+      "grad_norm": 0.015557540580630302,
+      "learning_rate": 0.0021584158415841586,
+      "loss": 0.5898,
+      "step": 475
+    },
+    {
+      "epoch": 1.0580716865796054,
+      "grad_norm": 0.010969873517751694,
+      "learning_rate": 0.002155940594059406,
+      "loss": 0.6872,
+      "step": 476
+    },
+    {
+      "epoch": 1.0602945262572936,
+      "grad_norm": 0.02743585780262947,
+      "learning_rate": 0.002153465346534653,
+      "loss": 0.6932,
+      "step": 477
+    },
+    {
+      "epoch": 1.062517365934982,
+      "grad_norm": 0.018156539648771286,
+      "learning_rate": 0.002150990099009901,
+      "loss": 0.6349,
+      "step": 478
+    },
+    {
+      "epoch": 1.0647402056126702,
+      "grad_norm": 0.029770391061902046,
+      "learning_rate": 0.0021485148514851487,
+      "loss": 0.8011,
+      "step": 479
+    },
+    {
+      "epoch": 1.0669630452903585,
+      "grad_norm": 0.028706863522529602,
+      "learning_rate": 0.0021460396039603958,
+      "loss": 0.5518,
+      "step": 480
+    },
+    {
+      "epoch": 1.0691858849680467,
+      "grad_norm": 0.01390407606959343,
+      "learning_rate": 0.0021435643564356437,
+      "loss": 0.6653,
+      "step": 481
+    },
+    {
+      "epoch": 1.071408724645735,
+      "grad_norm": 0.02037845179438591,
+      "learning_rate": 0.0021410891089108913,
+      "loss": 0.7506,
+      "step": 482
+    },
+    {
+      "epoch": 1.073631564323423,
+      "grad_norm": 0.012380843982100487,
+      "learning_rate": 0.0021386138613861383,
+      "loss": 0.5933,
+      "step": 483
+    },
+    {
+      "epoch": 1.0758544040011113,
+      "grad_norm": 0.01470144558697939,
+      "learning_rate": 0.0021361386138613863,
+      "loss": 0.8263,
+      "step": 484
+    },
+    {
+      "epoch": 1.0780772436787998,
+      "grad_norm": 0.015167983248829842,
+      "learning_rate": 0.002133663366336634,
+      "loss": 0.7961,
+      "step": 485
+    },
+    {
+      "epoch": 1.080300083356488,
+      "grad_norm": 0.016553625464439392,
+      "learning_rate": 0.002131188118811881,
+      "loss": 0.6522,
+      "step": 486
+    },
+    {
+      "epoch": 1.0825229230341762,
+      "grad_norm": 0.029548676684498787,
+      "learning_rate": 0.002128712871287129,
+      "loss": 0.7258,
+      "step": 487
+    },
+    {
+      "epoch": 1.0847457627118644,
+      "grad_norm": 0.012250442057847977,
+      "learning_rate": 0.0021262376237623764,
+      "loss": 0.5983,
+      "step": 488
+    },
+    {
+      "epoch": 1.0869686023895526,
+      "grad_norm": 0.013489637523889542,
+      "learning_rate": 0.0021237623762376235,
+      "loss": 0.6837,
+      "step": 489
+    },
+    {
+      "epoch": 1.0891914420672408,
+      "grad_norm": 0.01595526933670044,
+      "learning_rate": 0.0021212871287128714,
+      "loss": 0.6818,
+      "step": 490
+    },
+    {
+      "epoch": 1.091414281744929,
+      "grad_norm": 0.020508253946900368,
+      "learning_rate": 0.002118811881188119,
+      "loss": 0.6479,
+      "step": 491
+    },
+    {
+      "epoch": 1.0936371214226175,
+      "grad_norm": 0.01163517590612173,
+      "learning_rate": 0.002116336633663366,
+      "loss": 0.5976,
+      "step": 492
+    },
+    {
+      "epoch": 1.0958599611003057,
+      "grad_norm": 0.015765825286507607,
+      "learning_rate": 0.002113861386138614,
+      "loss": 0.6612,
+      "step": 493
+    },
+    {
+      "epoch": 1.098082800777994,
+      "grad_norm": 0.018122674897313118,
+      "learning_rate": 0.0021113861386138615,
+      "loss": 0.6547,
+      "step": 494
+    },
+    {
+      "epoch": 1.1003056404556821,
+      "grad_norm": 0.011971593834459782,
+      "learning_rate": 0.0021089108910891086,
+      "loss": 0.5676,
+      "step": 495
+    },
+    {
+      "epoch": 1.1025284801333703,
+      "grad_norm": 0.01705964468419552,
+      "learning_rate": 0.0021064356435643566,
+      "loss": 0.6342,
+      "step": 496
+    },
+    {
+      "epoch": 1.1047513198110586,
+      "grad_norm": 0.018494384363293648,
+      "learning_rate": 0.002103960396039604,
+      "loss": 0.7893,
+      "step": 497
+    },
+    {
+      "epoch": 1.1069741594887468,
+      "grad_norm": 0.015471387654542923,
+      "learning_rate": 0.002101485148514851,
+      "loss": 0.5907,
+      "step": 498
+    },
+    {
+      "epoch": 1.1091969991664352,
+      "grad_norm": 0.013633133843541145,
+      "learning_rate": 0.002099009900990099,
+      "loss": 0.7585,
+      "step": 499
+    },
+    {
+      "epoch": 1.1114198388441234,
+      "grad_norm": 0.013887847773730755,
+      "learning_rate": 0.0020965346534653467,
+      "loss": 0.6958,
+      "step": 500
+    },
+    {
+      "epoch": 1.1136426785218116,
+      "grad_norm": 0.015942484140396118,
+      "learning_rate": 0.0020940594059405938,
+      "loss": 0.7464,
+      "step": 501
+    },
+    {
+      "epoch": 1.1158655181994999,
+      "grad_norm": 0.01781686581671238,
+      "learning_rate": 0.0020915841584158417,
+      "loss": 0.5278,
+      "step": 502
+    },
+    {
+      "epoch": 1.118088357877188,
+      "grad_norm": 0.017188075929880142,
+      "learning_rate": 0.0020891089108910892,
+      "loss": 0.6229,
+      "step": 503
+    },
+    {
+      "epoch": 1.1203111975548763,
+      "grad_norm": 0.01655665598809719,
+      "learning_rate": 0.0020866336633663363,
+      "loss": 0.6761,
+      "step": 504
+    },
+    {
+      "epoch": 1.1225340372325645,
+      "grad_norm": 0.01681992970407009,
+      "learning_rate": 0.0020841584158415843,
+      "loss": 0.6118,
+      "step": 505
+    },
+    {
+      "epoch": 1.124756876910253,
+      "grad_norm": 0.021237896755337715,
+      "learning_rate": 0.002081683168316832,
+      "loss": 0.8598,
+      "step": 506
+    },
+    {
+      "epoch": 1.1269797165879412,
+      "grad_norm": 0.012300066649913788,
+      "learning_rate": 0.002079207920792079,
+      "loss": 0.5905,
+      "step": 507
+    },
+    {
+      "epoch": 1.1292025562656294,
+      "grad_norm": 0.018155410885810852,
+      "learning_rate": 0.002076732673267327,
+      "loss": 0.6204,
+      "step": 508
+    },
+    {
+      "epoch": 1.1314253959433176,
+      "grad_norm": 0.011526006273925304,
+      "learning_rate": 0.0020742574257425744,
+      "loss": 0.666,
+      "step": 509
+    },
+    {
+      "epoch": 1.1336482356210058,
+      "grad_norm": 0.014204266481101513,
+      "learning_rate": 0.0020717821782178215,
+      "loss": 0.6174,
+      "step": 510
+    },
+    {
+      "epoch": 1.135871075298694,
+      "grad_norm": 0.015811122953891754,
+      "learning_rate": 0.0020693069306930694,
+      "loss": 0.6967,
+      "step": 511
+    },
+    {
+      "epoch": 1.1380939149763822,
+      "grad_norm": 0.014839411713182926,
+      "learning_rate": 0.002066831683168317,
+      "loss": 0.5885,
+      "step": 512
+    },
+    {
+      "epoch": 1.1403167546540707,
+      "grad_norm": 0.015573407523334026,
+      "learning_rate": 0.002064356435643564,
+      "loss": 0.6261,
+      "step": 513
+    },
+    {
+      "epoch": 1.142539594331759,
+      "grad_norm": 0.01776973530650139,
+      "learning_rate": 0.002061881188118812,
+      "loss": 0.8348,
+      "step": 514
+    },
+    {
+      "epoch": 1.144762434009447,
+      "grad_norm": 0.02191152609884739,
+      "learning_rate": 0.0020594059405940595,
+      "loss": 0.6325,
+      "step": 515
+    },
+    {
+      "epoch": 1.1469852736871353,
+      "grad_norm": 0.03398964926600456,
+      "learning_rate": 0.002056930693069307,
+      "loss": 0.784,
+      "step": 516
+    },
+    {
+      "epoch": 1.1492081133648235,
+      "grad_norm": 0.01424696296453476,
+      "learning_rate": 0.0020544554455445545,
+      "loss": 0.7358,
+      "step": 517
+    },
+    {
+      "epoch": 1.1514309530425118,
+      "grad_norm": 0.013430099934339523,
+      "learning_rate": 0.002051980198019802,
+      "loss": 0.6242,
+      "step": 518
+    },
+    {
+      "epoch": 1.1536537927202,
+      "grad_norm": 0.010374044068157673,
+      "learning_rate": 0.0020495049504950496,
+      "loss": 0.5702,
+      "step": 519
+    },
+    {
+      "epoch": 1.1558766323978884,
+      "grad_norm": 0.0130368173122406,
+      "learning_rate": 0.002047029702970297,
+      "loss": 0.6321,
+      "step": 520
+    },
+    {
+      "epoch": 1.1580994720755766,
+      "grad_norm": 0.016580238938331604,
+      "learning_rate": 0.0020445544554455446,
+      "loss": 0.7411,
+      "step": 521
+    },
+    {
+      "epoch": 1.1603223117532648,
+      "grad_norm": 0.015030154958367348,
+      "learning_rate": 0.002042079207920792,
+      "loss": 0.572,
+      "step": 522
+    },
+    {
+      "epoch": 1.162545151430953,
+      "grad_norm": 0.014109233394265175,
+      "learning_rate": 0.0020396039603960397,
+      "loss": 0.7282,
+      "step": 523
+    },
+    {
+      "epoch": 1.1647679911086413,
+      "grad_norm": 0.019948862493038177,
+      "learning_rate": 0.002037128712871287,
+      "loss": 0.8526,
+      "step": 524
+    },
+    {
+      "epoch": 1.1669908307863295,
+      "grad_norm": 0.03067428059875965,
+      "learning_rate": 0.0020346534653465347,
+      "loss": 0.77,
+      "step": 525
+    },
+    {
+      "epoch": 1.1692136704640177,
+      "grad_norm": 0.014951253309845924,
+      "learning_rate": 0.0020321782178217822,
+      "loss": 0.5926,
+      "step": 526
+    },
+    {
+      "epoch": 1.1714365101417061,
+      "grad_norm": 0.020071959123015404,
+      "learning_rate": 0.0020297029702970298,
+      "loss": 0.6401,
+      "step": 527
+    },
+    {
+      "epoch": 1.1736593498193943,
+      "grad_norm": 0.013084011152386665,
+      "learning_rate": 0.0020272277227722773,
+      "loss": 0.5778,
+      "step": 528
+    },
+    {
+      "epoch": 1.1758821894970826,
+      "grad_norm": 0.012277776375412941,
+      "learning_rate": 0.002024752475247525,
+      "loss": 0.6096,
+      "step": 529
+    },
+    {
+      "epoch": 1.1781050291747708,
+      "grad_norm": 0.012979078106582165,
+      "learning_rate": 0.0020222772277227723,
+      "loss": 0.5634,
+      "step": 530
+    },
+    {
+      "epoch": 1.180327868852459,
+      "grad_norm": 0.015513453632593155,
+      "learning_rate": 0.00201980198019802,
+      "loss": 0.5527,
+      "step": 531
+    },
+    {
+      "epoch": 1.1825507085301472,
+      "grad_norm": 0.020396575331687927,
+      "learning_rate": 0.0020173267326732674,
+      "loss": 0.9307,
+      "step": 532
+    },
+    {
+      "epoch": 1.1847735482078354,
+      "grad_norm": 0.018580187112092972,
+      "learning_rate": 0.002014851485148515,
+      "loss": 0.6786,
+      "step": 533
+    },
+    {
+      "epoch": 1.1869963878855239,
+      "grad_norm": 0.015678012743592262,
+      "learning_rate": 0.0020123762376237624,
+      "loss": 0.5932,
+      "step": 534
+    },
+    {
+      "epoch": 1.189219227563212,
+      "grad_norm": 0.01222003810107708,
+      "learning_rate": 0.00200990099009901,
+      "loss": 0.5452,
+      "step": 535
+    },
+    {
+      "epoch": 1.1914420672409003,
+      "grad_norm": 0.0165384691208601,
+      "learning_rate": 0.0020074257425742575,
+      "loss": 0.655,
+      "step": 536
+    },
+    {
+      "epoch": 1.1936649069185885,
+      "grad_norm": 0.01701522432267666,
+      "learning_rate": 0.002004950495049505,
+      "loss": 0.512,
+      "step": 537
+    },
+    {
+      "epoch": 1.1958877465962767,
+      "grad_norm": 0.014413165859878063,
+      "learning_rate": 0.0020024752475247525,
+      "loss": 0.6258,
+      "step": 538
+    },
+    {
+      "epoch": 1.198110586273965,
+      "grad_norm": 0.011610453017055988,
+      "learning_rate": 0.002,
+      "loss": 0.6445,
+      "step": 539
+    },
+    {
+      "epoch": 1.2003334259516532,
+      "grad_norm": 0.016592789441347122,
+      "learning_rate": 0.0019975247524752476,
+      "loss": 0.7951,
+      "step": 540
+    },
+    {
+      "epoch": 1.2025562656293416,
+      "grad_norm": 0.018560299649834633,
+      "learning_rate": 0.001995049504950495,
+      "loss": 0.6788,
+      "step": 541
+    },
+    {
+      "epoch": 1.2047791053070298,
+      "grad_norm": 0.0228038989007473,
+      "learning_rate": 0.0019925742574257426,
+      "loss": 0.7306,
+      "step": 542
+    },
+    {
+      "epoch": 1.207001944984718,
+      "grad_norm": 0.013510987162590027,
+      "learning_rate": 0.00199009900990099,
+      "loss": 0.5083,
+      "step": 543
+    },
+    {
+      "epoch": 1.2092247846624062,
+      "grad_norm": 0.018051352351903915,
+      "learning_rate": 0.0019876237623762377,
+      "loss": 0.6776,
+      "step": 544
+    },
+    {
+      "epoch": 1.2114476243400945,
+      "grad_norm": 0.01775428093969822,
+      "learning_rate": 0.001985148514851485,
+      "loss": 0.702,
+      "step": 545
+    },
+    {
+      "epoch": 1.2136704640177827,
+      "grad_norm": 0.013823853805661201,
+      "learning_rate": 0.0019826732673267327,
+      "loss": 0.6253,
+      "step": 546
+    },
+    {
+      "epoch": 1.2158933036954709,
+      "grad_norm": 0.019051026552915573,
+      "learning_rate": 0.0019801980198019802,
+      "loss": 0.5959,
+      "step": 547
+    },
+    {
+      "epoch": 1.2181161433731593,
+      "grad_norm": 0.014647615142166615,
+      "learning_rate": 0.0019777227722772277,
+      "loss": 0.4997,
+      "step": 548
+    },
+    {
+      "epoch": 1.2203389830508475,
+      "grad_norm": 0.019500313326716423,
+      "learning_rate": 0.0019752475247524753,
+      "loss": 0.6763,
+      "step": 549
+    },
+    {
+      "epoch": 1.2225618227285358,
+      "grad_norm": 0.018273385241627693,
+      "learning_rate": 0.001972772277227723,
+      "loss": 0.5713,
+      "step": 550
+    },
+    {
+      "epoch": 1.224784662406224,
+      "grad_norm": 0.01844954676926136,
+      "learning_rate": 0.0019702970297029703,
+      "loss": 0.6973,
+      "step": 551
+    },
+    {
+      "epoch": 1.2270075020839122,
+      "grad_norm": 0.013612812384963036,
+      "learning_rate": 0.001967821782178218,
+      "loss": 0.53,
+      "step": 552
+    },
+    {
+      "epoch": 1.2292303417616004,
+      "grad_norm": 0.018429776653647423,
+      "learning_rate": 0.0019653465346534654,
+      "loss": 0.6369,
+      "step": 553
+    },
+    {
+      "epoch": 1.2314531814392886,
+      "grad_norm": 0.018885424360632896,
+      "learning_rate": 0.001962871287128713,
+      "loss": 0.7193,
+      "step": 554
+    },
+    {
+      "epoch": 1.233676021116977,
+      "grad_norm": 0.019281677901744843,
+      "learning_rate": 0.0019603960396039604,
+      "loss": 0.6858,
+      "step": 555
+    },
+    {
+      "epoch": 1.2358988607946653,
+      "grad_norm": 0.01671980880200863,
+      "learning_rate": 0.001957920792079208,
+      "loss": 0.708,
+      "step": 556
+    },
+    {
+      "epoch": 1.2381217004723535,
+      "grad_norm": 0.015330369584262371,
+      "learning_rate": 0.0019554455445544554,
+      "loss": 0.6055,
+      "step": 557
+    },
+    {
+      "epoch": 1.2403445401500417,
+      "grad_norm": 0.015489472076296806,
+      "learning_rate": 0.0019529702970297032,
+      "loss": 0.6312,
+      "step": 558
+    },
+    {
+      "epoch": 1.24256737982773,
+      "grad_norm": 0.01861627586185932,
+      "learning_rate": 0.0019504950495049505,
+      "loss": 0.6892,
+      "step": 559
+    },
+    {
+      "epoch": 1.2447902195054181,
+      "grad_norm": 0.01912919245660305,
+      "learning_rate": 0.0019480198019801978,
+      "loss": 0.5856,
+      "step": 560
+    },
+    {
+      "epoch": 1.2470130591831063,
+      "grad_norm": 0.01634219102561474,
+      "learning_rate": 0.0019455445544554458,
+      "loss": 0.7168,
+      "step": 561
+    },
+    {
+      "epoch": 1.2492358988607948,
+      "grad_norm": 0.01822001487016678,
+      "learning_rate": 0.001943069306930693,
+      "loss": 0.5872,
+      "step": 562
+    },
+    {
+      "epoch": 1.2514587385384828,
+      "grad_norm": 0.01985522173345089,
+      "learning_rate": 0.0019405940594059404,
+      "loss": 0.6333,
+      "step": 563
+    },
+    {
+      "epoch": 1.2536815782161712,
+      "grad_norm": 0.016396082937717438,
+      "learning_rate": 0.0019381188118811883,
+      "loss": 0.7005,
+      "step": 564
+    },
+    {
+      "epoch": 1.2559044178938594,
+      "grad_norm": 0.012350345961749554,
+      "learning_rate": 0.0019356435643564356,
+      "loss": 0.6913,
+      "step": 565
+    },
+    {
+      "epoch": 1.2581272575715476,
+      "grad_norm": 0.01816396601498127,
+      "learning_rate": 0.0019331683168316834,
+      "loss": 0.5174,
+      "step": 566
+    },
+    {
+      "epoch": 1.2603500972492359,
+      "grad_norm": 0.01564900204539299,
+      "learning_rate": 0.0019306930693069309,
+      "loss": 0.6686,
+      "step": 567
+    },
+    {
+      "epoch": 1.262572936926924,
+      "grad_norm": 0.014737764373421669,
+      "learning_rate": 0.0019282178217821782,
+      "loss": 0.5977,
+      "step": 568
+    },
+    {
+      "epoch": 1.2647957766046125,
+      "grad_norm": 0.02006666548550129,
+      "learning_rate": 0.001925742574257426,
+      "loss": 0.6276,
+      "step": 569
+    },
+    {
+      "epoch": 1.2670186162823005,
+      "grad_norm": 0.013658919371664524,
+      "learning_rate": 0.0019232673267326735,
+      "loss": 0.5861,
+      "step": 570
+    },
+    {
+      "epoch": 1.269241455959989,
+      "grad_norm": 0.011005178093910217,
+      "learning_rate": 0.0019207920792079208,
+      "loss": 0.8343,
+      "step": 571
+    },
+    {
+      "epoch": 1.2714642956376772,
+      "grad_norm": 0.016614580526947975,
+      "learning_rate": 0.0019183168316831685,
+      "loss": 0.7405,
+      "step": 572
+    },
+    {
+      "epoch": 1.2736871353153654,
+      "grad_norm": 0.013580153696238995,
+      "learning_rate": 0.001915841584158416,
+      "loss": 0.5955,
+      "step": 573
+    },
+    {
+      "epoch": 1.2759099749930536,
+      "grad_norm": 0.01571527309715748,
+      "learning_rate": 0.0019133663366336633,
+      "loss": 0.5994,
+      "step": 574
+    },
+    {
+      "epoch": 1.2781328146707418,
+      "grad_norm": 0.01320920791476965,
+      "learning_rate": 0.001910891089108911,
+      "loss": 0.5398,
+      "step": 575
+    },
+    {
+      "epoch": 1.2803556543484302,
+      "grad_norm": 0.02257618121802807,
+      "learning_rate": 0.0019084158415841586,
+      "loss": 0.6662,
+      "step": 576
+    },
+    {
+      "epoch": 1.2825784940261182,
+      "grad_norm": 0.020821325480937958,
+      "learning_rate": 0.001905940594059406,
+      "loss": 0.6462,
+      "step": 577
+    },
+    {
+      "epoch": 1.2848013337038067,
+      "grad_norm": 0.01534060388803482,
+      "learning_rate": 0.0019034653465346536,
+      "loss": 0.6875,
+      "step": 578
+    },
+    {
+      "epoch": 1.2870241733814949,
+      "grad_norm": 0.01106045302003622,
+      "learning_rate": 0.0019009900990099012,
+      "loss": 0.6197,
+      "step": 579
+    },
+    {
+      "epoch": 1.289247013059183,
+      "grad_norm": 0.023773077875375748,
+      "learning_rate": 0.0018985148514851485,
+      "loss": 0.6881,
+      "step": 580
+    },
+    {
+      "epoch": 1.2914698527368713,
+      "grad_norm": 0.014630904421210289,
+      "learning_rate": 0.0018960396039603962,
+      "loss": 0.8041,
+      "step": 581
+    },
+    {
+      "epoch": 1.2936926924145595,
+      "grad_norm": 0.021165568381547928,
+      "learning_rate": 0.0018935643564356437,
+      "loss": 0.7045,
+      "step": 582
+    },
+    {
+      "epoch": 1.295915532092248,
+      "grad_norm": 0.015476713888347149,
+      "learning_rate": 0.001891089108910891,
+      "loss": 0.5461,
+      "step": 583
+    },
+    {
+      "epoch": 1.298138371769936,
+      "grad_norm": 0.013936633244156837,
+      "learning_rate": 0.0018886138613861388,
+      "loss": 0.6052,
+      "step": 584
+    },
+    {
+      "epoch": 1.3003612114476244,
+      "grad_norm": 0.014164215885102749,
+      "learning_rate": 0.0018861386138613863,
+      "loss": 0.8044,
+      "step": 585
+    },
+    {
+      "epoch": 1.3025840511253126,
+      "grad_norm": 0.014623990282416344,
+      "learning_rate": 0.0018836633663366336,
+      "loss": 0.6213,
+      "step": 586
+    },
+    {
+      "epoch": 1.3048068908030008,
+      "grad_norm": 0.015673892572522163,
+      "learning_rate": 0.0018811881188118813,
+      "loss": 0.7006,
+      "step": 587
+    },
+    {
+      "epoch": 1.307029730480689,
+      "grad_norm": 0.020650116726756096,
+      "learning_rate": 0.0018787128712871289,
+      "loss": 0.6096,
+      "step": 588
+    },
+    {
+      "epoch": 1.3092525701583773,
+      "grad_norm": 0.019570233300328255,
+      "learning_rate": 0.0018762376237623762,
+      "loss": 0.7876,
+      "step": 589
+    },
+    {
+      "epoch": 1.3114754098360657,
+      "grad_norm": 0.024105733260512352,
+      "learning_rate": 0.001873762376237624,
+      "loss": 0.6646,
+      "step": 590
+    },
+    {
+      "epoch": 1.3136982495137537,
+      "grad_norm": 0.014969500713050365,
+      "learning_rate": 0.0018712871287128712,
+      "loss": 0.5828,
+      "step": 591
+    },
+    {
+      "epoch": 1.3159210891914421,
+      "grad_norm": 0.012277282774448395,
+      "learning_rate": 0.0018688118811881187,
+      "loss": 0.5273,
+      "step": 592
+    },
+    {
+      "epoch": 1.3181439288691303,
+      "grad_norm": 0.013523784466087818,
+      "learning_rate": 0.0018663366336633665,
+      "loss": 0.6394,
+      "step": 593
+    },
+    {
+      "epoch": 1.3203667685468186,
+      "grad_norm": 0.022080164402723312,
+      "learning_rate": 0.0018638613861386138,
+      "loss": 0.7142,
+      "step": 594
+    },
+    {
+      "epoch": 1.3225896082245068,
+      "grad_norm": 0.014192742295563221,
+      "learning_rate": 0.0018613861386138613,
+      "loss": 0.56,
+      "step": 595
+    },
+    {
+      "epoch": 1.324812447902195,
+      "grad_norm": 0.018689125776290894,
+      "learning_rate": 0.001858910891089109,
+      "loss": 0.6791,
+      "step": 596
+    },
+    {
+      "epoch": 1.3270352875798834,
+      "grad_norm": 0.01835326850414276,
+      "learning_rate": 0.0018564356435643563,
+      "loss": 0.6733,
+      "step": 597
+    },
+    {
+      "epoch": 1.3292581272575714,
+      "grad_norm": 0.022089499980211258,
+      "learning_rate": 0.0018539603960396039,
+      "loss": 0.6607,
+      "step": 598
+    },
+    {
+      "epoch": 1.3314809669352599,
+      "grad_norm": 0.011036070995032787,
+      "learning_rate": 0.0018514851485148516,
+      "loss": 0.5217,
+      "step": 599
+    },
+    {
+      "epoch": 1.333703806612948,
+      "grad_norm": 0.018533801659941673,
+      "learning_rate": 0.001849009900990099,
+      "loss": 0.7096,
+      "step": 600
+    },
+    {
+      "epoch": 1.3359266462906363,
+      "grad_norm": 0.017293326556682587,
+      "learning_rate": 0.0018465346534653464,
+      "loss": 0.6189,
+      "step": 601
+    },
+    {
+      "epoch": 1.3381494859683245,
+      "grad_norm": 0.01351864356547594,
+      "learning_rate": 0.0018440594059405942,
+      "loss": 0.7259,
+      "step": 602
+    },
+    {
+      "epoch": 1.3403723256460127,
+      "grad_norm": 0.018879210576415062,
+      "learning_rate": 0.0018415841584158415,
+      "loss": 0.5974,
+      "step": 603
+    },
+    {
+      "epoch": 1.3425951653237012,
+      "grad_norm": 0.023037323728203773,
+      "learning_rate": 0.001839108910891089,
+      "loss": 0.6754,
+      "step": 604
+    },
+    {
+      "epoch": 1.3448180050013891,
+      "grad_norm": 0.017983853816986084,
+      "learning_rate": 0.0018366336633663367,
+      "loss": 0.7237,
+      "step": 605
+    },
+    {
+      "epoch": 1.3470408446790776,
+      "grad_norm": 0.01909794472157955,
+      "learning_rate": 0.001834158415841584,
+      "loss": 0.7222,
+      "step": 606
+    },
+    {
+      "epoch": 1.3492636843567658,
+      "grad_norm": 0.014953205361962318,
+      "learning_rate": 0.0018316831683168316,
+      "loss": 0.6433,
+      "step": 607
+    },
+    {
+      "epoch": 1.351486524034454,
+      "grad_norm": 0.015649525448679924,
+      "learning_rate": 0.0018292079207920793,
+      "loss": 0.7764,
+      "step": 608
+    },
+    {
+      "epoch": 1.3537093637121422,
+      "grad_norm": 0.017241261899471283,
+      "learning_rate": 0.0018267326732673266,
+      "loss": 0.6004,
+      "step": 609
+    },
+    {
+      "epoch": 1.3559322033898304,
+      "grad_norm": 0.016156010329723358,
+      "learning_rate": 0.0018242574257425741,
+      "loss": 0.6187,
+      "step": 610
+    },
+    {
+      "epoch": 1.3581550430675189,
+      "grad_norm": 0.017765775322914124,
+      "learning_rate": 0.0018217821782178219,
+      "loss": 0.7092,
+      "step": 611
+    },
+    {
+      "epoch": 1.3603778827452069,
+      "grad_norm": 0.01482133287936449,
+      "learning_rate": 0.0018193069306930692,
+      "loss": 0.6921,
+      "step": 612
+    },
+    {
+      "epoch": 1.3626007224228953,
+      "grad_norm": 0.0338587686419487,
+      "learning_rate": 0.0018168316831683167,
+      "loss": 0.8791,
+      "step": 613
+    },
+    {
+      "epoch": 1.3648235621005835,
+      "grad_norm": 0.01598595455288887,
+      "learning_rate": 0.0018143564356435645,
+      "loss": 0.6677,
+      "step": 614
+    },
+    {
+      "epoch": 1.3670464017782717,
+      "grad_norm": 0.020157670602202415,
+      "learning_rate": 0.0018118811881188118,
+      "loss": 0.6577,
+      "step": 615
+    },
+    {
+      "epoch": 1.36926924145596,
+      "grad_norm": 0.017931604757905006,
+      "learning_rate": 0.0018094059405940597,
+      "loss": 0.5995,
+      "step": 616
+    },
+    {
+      "epoch": 1.3714920811336482,
+      "grad_norm": 0.015047606080770493,
+      "learning_rate": 0.001806930693069307,
+      "loss": 0.5683,
+      "step": 617
+    },
+    {
+      "epoch": 1.3737149208113366,
+      "grad_norm": 0.01254578959196806,
+      "learning_rate": 0.0018044554455445543,
+      "loss": 0.5574,
+      "step": 618
+    },
+    {
+      "epoch": 1.3759377604890246,
+      "grad_norm": 0.0250204186886549,
+      "learning_rate": 0.0018019801980198023,
+      "loss": 0.7746,
+      "step": 619
+    },
+    {
+      "epoch": 1.378160600166713,
+      "grad_norm": 0.0146646061912179,
+      "learning_rate": 0.0017995049504950496,
+      "loss": 0.5632,
+      "step": 620
+    },
+    {
+      "epoch": 1.3803834398444013,
+      "grad_norm": 0.019512226805090904,
+      "learning_rate": 0.0017970297029702969,
+      "loss": 0.6738,
+      "step": 621
+    },
+    {
+      "epoch": 1.3826062795220895,
+      "grad_norm": 0.013420785777270794,
+      "learning_rate": 0.0017945544554455448,
+      "loss": 0.6316,
+      "step": 622
+    },
+    {
+      "epoch": 1.3848291191997777,
+      "grad_norm": 0.023486951366066933,
+      "learning_rate": 0.0017920792079207922,
+      "loss": 0.5786,
+      "step": 623
+    },
+    {
+      "epoch": 1.387051958877466,
+      "grad_norm": 0.013036476448178291,
+      "learning_rate": 0.0017896039603960395,
+      "loss": 0.5637,
+      "step": 624
+    },
+    {
+      "epoch": 1.3892747985551543,
+      "grad_norm": 0.016444619745016098,
+      "learning_rate": 0.0017871287128712874,
+      "loss": 0.5816,
+      "step": 625
+    },
+    {
+      "epoch": 1.3914976382328423,
+      "grad_norm": 0.012313454411923885,
+      "learning_rate": 0.0017846534653465347,
+      "loss": 0.6912,
+      "step": 626
+    },
+    {
+      "epoch": 1.3937204779105308,
+      "grad_norm": 0.029828950762748718,
+      "learning_rate": 0.001782178217821782,
+      "loss": 0.5757,
+      "step": 627
+    },
+    {
+      "epoch": 1.395943317588219,
+      "grad_norm": 0.019931312650442123,
+      "learning_rate": 0.00177970297029703,
+      "loss": 0.833,
+      "step": 628
+    },
+    {
+      "epoch": 1.3981661572659072,
+      "grad_norm": 0.023598700761795044,
+      "learning_rate": 0.0017772277227722773,
+      "loss": 0.7982,
+      "step": 629
+    },
+    {
+      "epoch": 1.4003889969435954,
+      "grad_norm": 0.013721930794417858,
+      "learning_rate": 0.0017747524752475246,
+      "loss": 0.5859,
+      "step": 630
+    },
+    {
+      "epoch": 1.4026118366212836,
+      "grad_norm": 0.013073272071778774,
+      "learning_rate": 0.0017722772277227726,
+      "loss": 0.6802,
+      "step": 631
+    },
+    {
+      "epoch": 1.404834676298972,
+      "grad_norm": 0.015466667711734772,
+      "learning_rate": 0.0017698019801980199,
+      "loss": 0.6876,
+      "step": 632
+    },
+    {
+      "epoch": 1.40705751597666,
+      "grad_norm": 0.013860577717423439,
+      "learning_rate": 0.0017673267326732672,
+      "loss": 0.6569,
+      "step": 633
+    },
+    {
+      "epoch": 1.4092803556543485,
+      "grad_norm": 0.01486227661371231,
+      "learning_rate": 0.0017648514851485151,
+      "loss": 0.6061,
+      "step": 634
+    },
+    {
+      "epoch": 1.4115031953320367,
+      "grad_norm": 0.014290432445704937,
+      "learning_rate": 0.0017623762376237624,
+      "loss": 0.7512,
+      "step": 635
+    },
+    {
+      "epoch": 1.413726035009725,
+      "grad_norm": 0.014747899025678635,
+      "learning_rate": 0.0017599009900990097,
+      "loss": 0.6048,
+      "step": 636
+    },
+    {
+      "epoch": 1.4159488746874132,
+      "grad_norm": 0.02269178442656994,
+      "learning_rate": 0.0017574257425742577,
+      "loss": 0.5836,
+      "step": 637
+    },
+    {
+      "epoch": 1.4181717143651014,
+      "grad_norm": 0.019799571484327316,
+      "learning_rate": 0.001754950495049505,
+      "loss": 0.7016,
+      "step": 638
+    },
+    {
+      "epoch": 1.4203945540427896,
+      "grad_norm": 0.014343987219035625,
+      "learning_rate": 0.0017524752475247523,
+      "loss": 0.62,
+      "step": 639
+    },
+    {
+      "epoch": 1.4226173937204778,
+      "grad_norm": 0.01983143761754036,
+      "learning_rate": 0.0017500000000000003,
+      "loss": 0.5534,
+      "step": 640
+    },
+    {
+      "epoch": 1.4248402333981662,
+      "grad_norm": 0.012549810111522675,
+      "learning_rate": 0.0017475247524752476,
+      "loss": 0.6738,
+      "step": 641
+    },
+    {
+      "epoch": 1.4270630730758544,
+      "grad_norm": 0.019573699682950974,
+      "learning_rate": 0.0017450495049504949,
+      "loss": 0.6609,
+      "step": 642
+    },
+    {
+      "epoch": 1.4292859127535427,
+      "grad_norm": 0.017040664330124855,
+      "learning_rate": 0.0017425742574257428,
+      "loss": 0.5657,
+      "step": 643
+    },
+    {
+      "epoch": 1.4315087524312309,
+      "grad_norm": 0.021280160173773766,
+      "learning_rate": 0.0017400990099009901,
+      "loss": 0.7476,
+      "step": 644
+    },
+    {
+      "epoch": 1.433731592108919,
+      "grad_norm": 0.01822512224316597,
+      "learning_rate": 0.0017376237623762374,
+      "loss": 0.566,
+      "step": 645
+    },
+    {
+      "epoch": 1.4359544317866073,
+      "grad_norm": 0.015408231876790524,
+      "learning_rate": 0.0017351485148514854,
+      "loss": 0.6393,
+      "step": 646
+    },
+    {
+      "epoch": 1.4381772714642955,
+      "grad_norm": 0.012197799980640411,
+      "learning_rate": 0.0017326732673267327,
+      "loss": 0.6905,
+      "step": 647
+    },
+    {
+      "epoch": 1.440400111141984,
+      "grad_norm": 0.014727549627423286,
+      "learning_rate": 0.00173019801980198,
+      "loss": 0.6197,
+      "step": 648
+    },
+    {
+      "epoch": 1.4426229508196722,
+      "grad_norm": 0.021275896579027176,
+      "learning_rate": 0.001727722772277228,
+      "loss": 0.7485,
+      "step": 649
+    },
+    {
+      "epoch": 1.4448457904973604,
+      "grad_norm": 0.012766880914568901,
+      "learning_rate": 0.0017252475247524753,
+      "loss": 0.6318,
+      "step": 650
+    },
+    {
+      "epoch": 1.4470686301750486,
+      "grad_norm": 0.020205190405249596,
+      "learning_rate": 0.0017227722772277226,
+      "loss": 0.5067,
+      "step": 651
+    },
+    {
+      "epoch": 1.4492914698527368,
+      "grad_norm": 0.014094137586653233,
+      "learning_rate": 0.0017202970297029705,
+      "loss": 0.6953,
+      "step": 652
+    },
+    {
+      "epoch": 1.451514309530425,
+      "grad_norm": 0.015574413351714611,
+      "learning_rate": 0.0017178217821782178,
+      "loss": 0.5785,
+      "step": 653
+    },
+    {
+      "epoch": 1.4537371492081133,
+      "grad_norm": 0.01962168887257576,
+      "learning_rate": 0.0017153465346534651,
+      "loss": 0.5311,
+      "step": 654
+    },
+    {
+      "epoch": 1.4559599888858017,
+      "grad_norm": 0.014445985667407513,
+      "learning_rate": 0.001712871287128713,
+      "loss": 0.5768,
+      "step": 655
+    },
+    {
+      "epoch": 1.45818282856349,
+      "grad_norm": 0.013283908367156982,
+      "learning_rate": 0.0017103960396039604,
+      "loss": 0.6906,
+      "step": 656
+    },
+    {
+      "epoch": 1.4604056682411781,
+      "grad_norm": 0.014568808488547802,
+      "learning_rate": 0.0017079207920792077,
+      "loss": 0.5427,
+      "step": 657
+    },
+    {
+      "epoch": 1.4626285079188663,
+      "grad_norm": 0.01634564995765686,
+      "learning_rate": 0.0017054455445544557,
+      "loss": 0.7341,
+      "step": 658
+    },
+    {
+      "epoch": 1.4648513475965546,
+      "grad_norm": 0.015426786616444588,
+      "learning_rate": 0.001702970297029703,
+      "loss": 0.6963,
+      "step": 659
+    },
+    {
+      "epoch": 1.4670741872742428,
+      "grad_norm": 0.017544064670801163,
+      "learning_rate": 0.0017004950495049503,
+      "loss": 0.7626,
+      "step": 660
+    },
+    {
+      "epoch": 1.469297026951931,
+      "grad_norm": 0.017237959429621696,
+      "learning_rate": 0.0016980198019801982,
+      "loss": 0.5935,
+      "step": 661
+    },
+    {
+      "epoch": 1.4715198666296194,
+      "grad_norm": 0.019028373062610626,
+      "learning_rate": 0.0016955445544554455,
+      "loss": 0.6749,
+      "step": 662
+    },
+    {
+      "epoch": 1.4737427063073076,
+      "grad_norm": 0.012185648083686829,
+      "learning_rate": 0.0016930693069306928,
+      "loss": 0.6386,
+      "step": 663
+    },
+    {
+      "epoch": 1.4759655459849959,
+      "grad_norm": 0.01708870567381382,
+      "learning_rate": 0.0016905940594059408,
+      "loss": 0.691,
+      "step": 664
+    },
+    {
+      "epoch": 1.478188385662684,
+      "grad_norm": 0.013403747230768204,
+      "learning_rate": 0.001688118811881188,
+      "loss": 0.6614,
+      "step": 665
+    },
+    {
+      "epoch": 1.4804112253403723,
+      "grad_norm": 0.021224593743681908,
+      "learning_rate": 0.0016856435643564358,
+      "loss": 0.7882,
+      "step": 666
+    },
+    {
+      "epoch": 1.4826340650180605,
+      "grad_norm": 0.07430857419967651,
+      "learning_rate": 0.0016831683168316834,
+      "loss": 0.8449,
+      "step": 667
+    },
+    {
+      "epoch": 1.4848569046957487,
+      "grad_norm": 0.01444665901362896,
+      "learning_rate": 0.0016806930693069307,
+      "loss": 0.5309,
+      "step": 668
+    },
+    {
+      "epoch": 1.4870797443734372,
+      "grad_norm": 0.014784655533730984,
+      "learning_rate": 0.0016782178217821784,
+      "loss": 0.6546,
+      "step": 669
+    },
+    {
+      "epoch": 1.4893025840511254,
+      "grad_norm": 0.02172185480594635,
+      "learning_rate": 0.001675742574257426,
+      "loss": 0.673,
+      "step": 670
+    },
+    {
+      "epoch": 1.4915254237288136,
+      "grad_norm": 0.010736542753875256,
+      "learning_rate": 0.0016732673267326732,
+      "loss": 0.5521,
+      "step": 671
+    },
+    {
+      "epoch": 1.4937482634065018,
+      "grad_norm": 0.013517359271645546,
+      "learning_rate": 0.001670792079207921,
+      "loss": 0.6336,
+      "step": 672
+    },
+    {
+      "epoch": 1.49597110308419,
+      "grad_norm": 0.021300826221704483,
+      "learning_rate": 0.0016683168316831685,
+      "loss": 0.6419,
+      "step": 673
+    },
+    {
+      "epoch": 1.4981939427618782,
+      "grad_norm": 0.019188571721315384,
+      "learning_rate": 0.0016658415841584158,
+      "loss": 0.6618,
+      "step": 674
+    },
+    {
+      "epoch": 1.5004167824395664,
+      "grad_norm": 0.0127353984862566,
+      "learning_rate": 0.0016633663366336635,
+      "loss": 0.6328,
+      "step": 675
+    },
+    {
+      "epoch": 1.5026396221172549,
+      "grad_norm": 0.012695038691163063,
+      "learning_rate": 0.001660891089108911,
+      "loss": 0.5579,
+      "step": 676
+    },
+    {
+      "epoch": 1.5048624617949429,
+      "grad_norm": 0.02976105734705925,
+      "learning_rate": 0.0016584158415841584,
+      "loss": 0.8424,
+      "step": 677
+    },
+    {
+      "epoch": 1.5070853014726313,
+      "grad_norm": 0.016801824793219566,
+      "learning_rate": 0.0016559405940594061,
+      "loss": 0.5539,
+      "step": 678
+    },
+    {
+      "epoch": 1.5093081411503195,
+      "grad_norm": 0.021268382668495178,
+      "learning_rate": 0.0016534653465346536,
+      "loss": 0.7329,
+      "step": 679
+    },
+    {
+      "epoch": 1.5115309808280077,
+      "grad_norm": 0.014745811931788921,
+      "learning_rate": 0.001650990099009901,
+      "loss": 0.6433,
+      "step": 680
+    },
+    {
+      "epoch": 1.5137538205056962,
+      "grad_norm": 0.018007369711995125,
+      "learning_rate": 0.0016485148514851487,
+      "loss": 0.5846,
+      "step": 681
+    },
+    {
+      "epoch": 1.5159766601833842,
+      "grad_norm": 0.014517346397042274,
+      "learning_rate": 0.0016460396039603962,
+      "loss": 0.6494,
+      "step": 682
+    },
+    {
+      "epoch": 1.5181994998610726,
+      "grad_norm": 0.011836903169751167,
+      "learning_rate": 0.0016435643564356435,
+      "loss": 0.6144,
+      "step": 683
+    },
+    {
+      "epoch": 1.5204223395387606,
+      "grad_norm": 0.015668461099267006,
+      "learning_rate": 0.0016410891089108912,
+      "loss": 0.697,
+      "step": 684
+    },
+    {
+      "epoch": 1.522645179216449,
+      "grad_norm": 0.0159301720559597,
+      "learning_rate": 0.0016386138613861388,
+      "loss": 0.6344,
+      "step": 685
+    },
+    {
+      "epoch": 1.5248680188941373,
+      "grad_norm": 0.021761801093816757,
+      "learning_rate": 0.001636138613861386,
+      "loss": 0.7244,
+      "step": 686
+    },
+    {
+      "epoch": 1.5270908585718255,
+      "grad_norm": 0.017973683774471283,
+      "learning_rate": 0.0016336633663366338,
+      "loss": 0.6616,
+      "step": 687
+    },
+    {
+      "epoch": 1.529313698249514,
+      "grad_norm": 0.03193248063325882,
+      "learning_rate": 0.0016311881188118813,
+      "loss": 0.5955,
+      "step": 688
+    },
+    {
+      "epoch": 1.531536537927202,
+      "grad_norm": 0.022393716499209404,
+      "learning_rate": 0.0016287128712871286,
+      "loss": 0.6299,
+      "step": 689
+    },
+    {
+      "epoch": 1.5337593776048903,
+      "grad_norm": 0.022320229560136795,
+      "learning_rate": 0.0016262376237623764,
+      "loss": 0.7729,
+      "step": 690
+    },
+    {
+      "epoch": 1.5359822172825783,
+      "grad_norm": 0.014977824874222279,
+      "learning_rate": 0.0016237623762376237,
+      "loss": 0.6697,
+      "step": 691
+    },
+    {
+      "epoch": 1.5382050569602668,
+      "grad_norm": 0.01872638612985611,
+      "learning_rate": 0.0016212871287128712,
+      "loss": 0.5893,
+      "step": 692
+    },
+    {
+      "epoch": 1.540427896637955,
+      "grad_norm": 0.015523020178079605,
+      "learning_rate": 0.001618811881188119,
+      "loss": 0.6822,
+      "step": 693
+    },
+    {
+      "epoch": 1.5426507363156432,
+      "grad_norm": 0.014033573679625988,
+      "learning_rate": 0.0016163366336633663,
+      "loss": 0.6213,
+      "step": 694
+    },
+    {
+      "epoch": 1.5448735759933316,
+      "grad_norm": 0.014586709439754486,
+      "learning_rate": 0.0016138613861386138,
+      "loss": 0.7703,
+      "step": 695
+    },
+    {
+      "epoch": 1.5470964156710196,
+      "grad_norm": 0.014374952763319016,
+      "learning_rate": 0.0016113861386138615,
+      "loss": 0.6198,
+      "step": 696
+    },
+    {
+      "epoch": 1.549319255348708,
+      "grad_norm": 0.017982954159379005,
+      "learning_rate": 0.0016089108910891088,
+      "loss": 0.6394,
+      "step": 697
+    },
+    {
+      "epoch": 1.551542095026396,
+      "grad_norm": 0.014446941204369068,
+      "learning_rate": 0.0016064356435643563,
+      "loss": 0.6657,
+      "step": 698
+    },
+    {
+      "epoch": 1.5537649347040845,
+      "grad_norm": 0.018195228651165962,
+      "learning_rate": 0.001603960396039604,
+      "loss": 0.5129,
+      "step": 699
+    },
+    {
+      "epoch": 1.5559877743817727,
+      "grad_norm": 0.01925376057624817,
+      "learning_rate": 0.0016014851485148514,
+      "loss": 0.643,
+      "step": 700
+    },
+    {
+      "epoch": 1.558210614059461,
+      "grad_norm": 0.036713700741529465,
+      "learning_rate": 0.001599009900990099,
+      "loss": 0.6965,
+      "step": 701
+    },
+    {
+      "epoch": 1.5604334537371494,
+      "grad_norm": 0.01623792201280594,
+      "learning_rate": 0.0015965346534653467,
+      "loss": 0.5376,
+      "step": 702
+    },
+    {
+      "epoch": 1.5626562934148374,
+      "grad_norm": 0.0159526988863945,
+      "learning_rate": 0.001594059405940594,
+      "loss": 0.6076,
+      "step": 703
+    },
+    {
+      "epoch": 1.5648791330925258,
+      "grad_norm": 0.017126891762018204,
+      "learning_rate": 0.0015915841584158415,
+      "loss": 0.6329,
+      "step": 704
+    },
+    {
+      "epoch": 1.5671019727702138,
+      "grad_norm": 0.019651105627417564,
+      "learning_rate": 0.0015891089108910892,
+      "loss": 0.4878,
+      "step": 705
+    },
+    {
+      "epoch": 1.5693248124479022,
+      "grad_norm": 0.01890754885971546,
+      "learning_rate": 0.0015866336633663365,
+      "loss": 0.677,
+      "step": 706
+    },
+    {
+      "epoch": 1.5715476521255904,
+      "grad_norm": 0.014043249189853668,
+      "learning_rate": 0.001584158415841584,
+      "loss": 0.5893,
+      "step": 707
+    },
+    {
+      "epoch": 1.5737704918032787,
+      "grad_norm": 0.014365495182573795,
+      "learning_rate": 0.0015816831683168318,
+      "loss": 0.6114,
+      "step": 708
+    },
+    {
+      "epoch": 1.575993331480967,
+      "grad_norm": 0.021364012733101845,
+      "learning_rate": 0.001579207920792079,
+      "loss": 0.616,
+      "step": 709
+    },
+    {
+      "epoch": 1.578216171158655,
+      "grad_norm": 0.021359330043196678,
+      "learning_rate": 0.0015767326732673266,
+      "loss": 0.7248,
+      "step": 710
+    },
+    {
+      "epoch": 1.5804390108363435,
+      "grad_norm": 0.01866665668785572,
+      "learning_rate": 0.0015742574257425744,
+      "loss": 0.7569,
+      "step": 711
+    },
+    {
+      "epoch": 1.5826618505140315,
+      "grad_norm": 0.013409837149083614,
+      "learning_rate": 0.0015717821782178217,
+      "loss": 0.5397,
+      "step": 712
+    },
+    {
+      "epoch": 1.58488469019172,
+      "grad_norm": 0.014709453098475933,
+      "learning_rate": 0.0015693069306930692,
+      "loss": 0.5613,
+      "step": 713
+    },
+    {
+      "epoch": 1.5871075298694082,
+      "grad_norm": 0.014858447946608067,
+      "learning_rate": 0.001566831683168317,
+      "loss": 0.6546,
+      "step": 714
+    },
+    {
+      "epoch": 1.5893303695470964,
+      "grad_norm": 0.01660234108567238,
+      "learning_rate": 0.0015643564356435642,
+      "loss": 0.8238,
+      "step": 715
+    },
+    {
+      "epoch": 1.5915532092247848,
+      "grad_norm": 0.01475167740136385,
+      "learning_rate": 0.0015618811881188122,
+      "loss": 0.6471,
+      "step": 716
+    },
+    {
+      "epoch": 1.5937760489024728,
+      "grad_norm": 0.024248579517006874,
+      "learning_rate": 0.0015594059405940595,
+      "loss": 0.7741,
+      "step": 717
+    },
+    {
+      "epoch": 1.5959988885801613,
+      "grad_norm": 0.02008378878235817,
+      "learning_rate": 0.0015569306930693068,
+      "loss": 0.6338,
+      "step": 718
+    },
+    {
+      "epoch": 1.5982217282578493,
+      "grad_norm": 0.013149694539606571,
+      "learning_rate": 0.0015544554455445548,
+      "loss": 0.5793,
+      "step": 719
+    },
+    {
+      "epoch": 1.6004445679355377,
+      "grad_norm": 0.016647828742861748,
+      "learning_rate": 0.001551980198019802,
+      "loss": 0.571,
+      "step": 720
+    },
+    {
+      "epoch": 1.602667407613226,
+      "grad_norm": 0.020569197833538055,
+      "learning_rate": 0.0015495049504950494,
+      "loss": 0.747,
+      "step": 721
+    },
+    {
+      "epoch": 1.6048902472909141,
+      "grad_norm": 0.015371223911643028,
+      "learning_rate": 0.0015470297029702973,
+      "loss": 0.7479,
+      "step": 722
+    },
+    {
+      "epoch": 1.6071130869686026,
+      "grad_norm": 0.018335159868001938,
+      "learning_rate": 0.0015445544554455446,
+      "loss": 0.7619,
+      "step": 723
+    },
+    {
+      "epoch": 1.6093359266462905,
+      "grad_norm": 0.012718189507722855,
+      "learning_rate": 0.001542079207920792,
+      "loss": 0.6172,
+      "step": 724
+    },
+    {
+      "epoch": 1.611558766323979,
+      "grad_norm": 0.0307784266769886,
+      "learning_rate": 0.0015396039603960399,
+      "loss": 0.6105,
+      "step": 725
+    },
+    {
+      "epoch": 1.613781606001667,
+      "grad_norm": 0.017151063308119774,
+      "learning_rate": 0.0015371287128712872,
+      "loss": 0.6892,
+      "step": 726
+    },
+    {
+      "epoch": 1.6160044456793554,
+      "grad_norm": 0.020293843001127243,
+      "learning_rate": 0.0015346534653465345,
+      "loss": 0.6951,
+      "step": 727
+    },
+    {
+      "epoch": 1.6182272853570436,
+      "grad_norm": 0.018611643463373184,
+      "learning_rate": 0.0015321782178217825,
+      "loss": 0.7882,
+      "step": 728
+    },
+    {
+      "epoch": 1.6204501250347318,
+      "grad_norm": 0.0173718873411417,
+      "learning_rate": 0.0015297029702970298,
+      "loss": 0.6516,
+      "step": 729
+    },
+    {
+      "epoch": 1.62267296471242,
+      "grad_norm": 0.014312409795820713,
+      "learning_rate": 0.001527227722772277,
+      "loss": 0.5164,
+      "step": 730
+    },
+    {
+      "epoch": 1.6248958043901083,
+      "grad_norm": 0.01838870160281658,
+      "learning_rate": 0.001524752475247525,
+      "loss": 0.6303,
+      "step": 731
+    },
+    {
+      "epoch": 1.6271186440677967,
+      "grad_norm": 0.013072308152914047,
+      "learning_rate": 0.0015222772277227723,
+      "loss": 0.6555,
+      "step": 732
+    },
+    {
+      "epoch": 1.6293414837454847,
+      "grad_norm": 0.013176838867366314,
+      "learning_rate": 0.0015198019801980196,
+      "loss": 0.6303,
+      "step": 733
+    },
+    {
+      "epoch": 1.6315643234231731,
+      "grad_norm": 0.016583677381277084,
+      "learning_rate": 0.0015173267326732676,
+      "loss": 0.6323,
+      "step": 734
+    },
+    {
+      "epoch": 1.6337871631008614,
+      "grad_norm": 0.015155021101236343,
+      "learning_rate": 0.001514851485148515,
+      "loss": 0.6323,
+      "step": 735
+    },
+    {
+      "epoch": 1.6360100027785496,
+      "grad_norm": 0.013991404324769974,
+      "learning_rate": 0.0015123762376237622,
+      "loss": 0.6144,
+      "step": 736
+    },
+    {
+      "epoch": 1.6382328424562378,
+      "grad_norm": 0.021456275135278702,
+      "learning_rate": 0.0015099009900990102,
+      "loss": 0.6232,
+      "step": 737
+    },
+    {
+      "epoch": 1.640455682133926,
+      "grad_norm": 0.016199328005313873,
+      "learning_rate": 0.0015074257425742575,
+      "loss": 0.6021,
+      "step": 738
+    },
+    {
+      "epoch": 1.6426785218116144,
+      "grad_norm": 0.02143920212984085,
+      "learning_rate": 0.0015049504950495048,
+      "loss": 0.6952,
+      "step": 739
+    },
+    {
+      "epoch": 1.6449013614893024,
+      "grad_norm": 0.016033926978707314,
+      "learning_rate": 0.0015024752475247527,
+      "loss": 0.7172,
+      "step": 740
+    },
+    {
+      "epoch": 1.6471242011669909,
+      "grad_norm": 0.012828083708882332,
+      "learning_rate": 0.0015,
+      "loss": 0.6543,
+      "step": 741
+    },
+    {
+      "epoch": 1.649347040844679,
+      "grad_norm": 0.0132541349157691,
+      "learning_rate": 0.0014975247524752476,
+      "loss": 0.6105,
+      "step": 742
+    },
+    {
+      "epoch": 1.6515698805223673,
+      "grad_norm": 0.01240797434002161,
+      "learning_rate": 0.001495049504950495,
+      "loss": 0.6074,
+      "step": 743
+    },
+    {
+      "epoch": 1.6537927202000555,
+      "grad_norm": 0.0144476518034935,
+      "learning_rate": 0.0014925742574257426,
+      "loss": 0.6292,
+      "step": 744
+    },
+    {
+      "epoch": 1.6560155598777437,
+      "grad_norm": 0.013765274547040462,
+      "learning_rate": 0.0014900990099009901,
+      "loss": 0.6266,
+      "step": 745
+    },
+    {
+      "epoch": 1.6582383995554322,
+      "grad_norm": 0.015940168872475624,
+      "learning_rate": 0.0014876237623762376,
+      "loss": 0.6191,
+      "step": 746
+    },
+    {
+      "epoch": 1.6604612392331202,
+      "grad_norm": 0.013493590988218784,
+      "learning_rate": 0.0014851485148514852,
+      "loss": 0.6048,
+      "step": 747
+    },
+    {
+      "epoch": 1.6626840789108086,
+      "grad_norm": 0.01261868141591549,
+      "learning_rate": 0.0014826732673267327,
+      "loss": 0.6823,
+      "step": 748
+    },
+    {
+      "epoch": 1.6649069185884968,
+      "grad_norm": 0.024276770651340485,
+      "learning_rate": 0.0014801980198019802,
+      "loss": 0.7041,
+      "step": 749
+    },
+    {
+      "epoch": 1.667129758266185,
+      "grad_norm": 0.019383903592824936,
+      "learning_rate": 0.0014777227722772277,
+      "loss": 0.7646,
+      "step": 750
+    },
+    {
+      "epoch": 1.6693525979438733,
+      "grad_norm": 0.015850719064474106,
+      "learning_rate": 0.0014752475247524753,
+      "loss": 0.5642,
+      "step": 751
+    },
+    {
+      "epoch": 1.6715754376215615,
+      "grad_norm": 0.014828231185674667,
+      "learning_rate": 0.0014727722772277228,
+      "loss": 0.728,
+      "step": 752
+    },
+    {
+      "epoch": 1.67379827729925,
+      "grad_norm": 0.014795539900660515,
+      "learning_rate": 0.0014702970297029703,
+      "loss": 0.514,
+      "step": 753
+    },
+    {
+      "epoch": 1.676021116976938,
+      "grad_norm": 0.03070404753088951,
+      "learning_rate": 0.0014678217821782178,
+      "loss": 0.659,
+      "step": 754
+    },
+    {
+      "epoch": 1.6782439566546263,
+      "grad_norm": 0.01784861460328102,
+      "learning_rate": 0.0014653465346534656,
+      "loss": 0.6316,
+      "step": 755
+    },
+    {
+      "epoch": 1.6804667963323145,
+      "grad_norm": 0.020610976964235306,
+      "learning_rate": 0.0014628712871287129,
+      "loss": 0.5602,
+      "step": 756
+    },
+    {
+      "epoch": 1.6826896360100028,
+      "grad_norm": 0.01749056577682495,
+      "learning_rate": 0.0014603960396039604,
+      "loss": 0.7298,
+      "step": 757
+    },
+    {
+      "epoch": 1.684912475687691,
+      "grad_norm": 0.019512880593538284,
+      "learning_rate": 0.0014579207920792081,
+      "loss": 0.6296,
+      "step": 758
+    },
+    {
+      "epoch": 1.6871353153653792,
+      "grad_norm": 0.017783217132091522,
+      "learning_rate": 0.0014554455445544554,
+      "loss": 0.6959,
+      "step": 759
+    },
+    {
+      "epoch": 1.6893581550430676,
+      "grad_norm": 0.015032476745545864,
+      "learning_rate": 0.001452970297029703,
+      "loss": 0.54,
+      "step": 760
+    },
+    {
+      "epoch": 1.6915809947207556,
+      "grad_norm": 0.010041123256087303,
+      "learning_rate": 0.0014504950495049507,
+      "loss": 0.5299,
+      "step": 761
+    },
+    {
+      "epoch": 1.693803834398444,
+      "grad_norm": 0.013333450071513653,
+      "learning_rate": 0.001448019801980198,
+      "loss": 0.5786,
+      "step": 762
+    },
+    {
+      "epoch": 1.6960266740761323,
+      "grad_norm": 0.017033929005265236,
+      "learning_rate": 0.0014455445544554455,
+      "loss": 0.6885,
+      "step": 763
+    },
+    {
+      "epoch": 1.6982495137538205,
+      "grad_norm": 0.01851385273039341,
+      "learning_rate": 0.0014430693069306933,
+      "loss": 0.8939,
+      "step": 764
+    },
+    {
+      "epoch": 1.7004723534315087,
+      "grad_norm": 0.021369075402617455,
+      "learning_rate": 0.0014405940594059406,
+      "loss": 0.6529,
+      "step": 765
+    },
+    {
+      "epoch": 1.702695193109197,
+      "grad_norm": 0.012632249854505062,
+      "learning_rate": 0.001438118811881188,
+      "loss": 0.5608,
+      "step": 766
+    },
+    {
+      "epoch": 1.7049180327868854,
+      "grad_norm": 0.01696857064962387,
+      "learning_rate": 0.0014356435643564358,
+      "loss": 0.5658,
+      "step": 767
+    },
+    {
+      "epoch": 1.7071408724645734,
+      "grad_norm": 0.02316252514719963,
+      "learning_rate": 0.0014331683168316831,
+      "loss": 0.6628,
+      "step": 768
+    },
+    {
+      "epoch": 1.7093637121422618,
+      "grad_norm": 0.014780178666114807,
+      "learning_rate": 0.0014306930693069307,
+      "loss": 0.5469,
+      "step": 769
+    },
+    {
+      "epoch": 1.71158655181995,
+      "grad_norm": 0.02393134869635105,
+      "learning_rate": 0.0014282178217821784,
+      "loss": 0.6237,
+      "step": 770
+    },
+    {
+      "epoch": 1.7138093914976382,
+      "grad_norm": 0.017437728121876717,
+      "learning_rate": 0.0014257425742574257,
+      "loss": 0.598,
+      "step": 771
+    },
+    {
+      "epoch": 1.7160322311753264,
+      "grad_norm": 0.014657423831522465,
+      "learning_rate": 0.0014232673267326732,
+      "loss": 0.6574,
+      "step": 772
+    },
+    {
+      "epoch": 1.7182550708530147,
+      "grad_norm": 0.014511947520077229,
+      "learning_rate": 0.001420792079207921,
+      "loss": 0.4105,
+      "step": 773
+    },
+    {
+      "epoch": 1.720477910530703,
+      "grad_norm": 0.0161074697971344,
+      "learning_rate": 0.0014183168316831683,
+      "loss": 0.5853,
+      "step": 774
+    },
+    {
+      "epoch": 1.722700750208391,
+      "grad_norm": 0.015577802434563637,
+      "learning_rate": 0.0014158415841584158,
+      "loss": 0.7821,
+      "step": 775
+    },
+    {
+      "epoch": 1.7249235898860795,
+      "grad_norm": 0.013394194655120373,
+      "learning_rate": 0.0014133663366336635,
+      "loss": 0.6393,
+      "step": 776
+    },
+    {
+      "epoch": 1.7271464295637677,
+      "grad_norm": 0.011438438668847084,
+      "learning_rate": 0.0014108910891089108,
+      "loss": 0.5918,
+      "step": 777
+    },
+    {
+      "epoch": 1.729369269241456,
+      "grad_norm": 0.01416991837322712,
+      "learning_rate": 0.0014084158415841584,
+      "loss": 0.5726,
+      "step": 778
+    },
+    {
+      "epoch": 1.7315921089191442,
+      "grad_norm": 0.011613192036747932,
+      "learning_rate": 0.001405940594059406,
+      "loss": 0.6207,
+      "step": 779
+    },
+    {
+      "epoch": 1.7338149485968324,
+      "grad_norm": 0.014506768435239792,
+      "learning_rate": 0.0014034653465346536,
+      "loss": 0.7833,
+      "step": 780
+    },
+    {
+      "epoch": 1.7360377882745208,
+      "grad_norm": 0.016172049567103386,
+      "learning_rate": 0.001400990099009901,
+      "loss": 0.6071,
+      "step": 781
+    },
+    {
+      "epoch": 1.7382606279522088,
+      "grad_norm": 0.014774010516703129,
+      "learning_rate": 0.0013985148514851487,
+      "loss": 0.7103,
+      "step": 782
+    },
+    {
+      "epoch": 1.7404834676298973,
+      "grad_norm": 0.019349824637174606,
+      "learning_rate": 0.0013960396039603962,
+      "loss": 0.648,
+      "step": 783
+    },
+    {
+      "epoch": 1.7427063073075855,
+      "grad_norm": 0.012977352365851402,
+      "learning_rate": 0.0013935643564356435,
+      "loss": 0.5717,
+      "step": 784
+    },
+    {
+      "epoch": 1.7449291469852737,
+      "grad_norm": 0.01410915981978178,
+      "learning_rate": 0.0013910891089108912,
+      "loss": 0.5183,
+      "step": 785
+    },
+    {
+      "epoch": 1.747151986662962,
+      "grad_norm": 0.018457330763339996,
+      "learning_rate": 0.0013886138613861388,
+      "loss": 0.5707,
+      "step": 786
+    },
+    {
+      "epoch": 1.7493748263406501,
+      "grad_norm": 0.020965803414583206,
+      "learning_rate": 0.001386138613861386,
+      "loss": 0.6848,
+      "step": 787
+    },
+    {
+      "epoch": 1.7515976660183386,
+      "grad_norm": 0.01630076766014099,
+      "learning_rate": 0.0013836633663366338,
+      "loss": 0.8393,
+      "step": 788
+    },
+    {
+      "epoch": 1.7538205056960265,
+      "grad_norm": 0.013406305573880672,
+      "learning_rate": 0.0013811881188118813,
+      "loss": 0.4215,
+      "step": 789
+    },
+    {
+      "epoch": 1.756043345373715,
+      "grad_norm": 0.013798919506371021,
+      "learning_rate": 0.0013787128712871286,
+      "loss": 0.6094,
+      "step": 790
+    },
+    {
+      "epoch": 1.7582661850514032,
+      "grad_norm": 0.028619924560189247,
+      "learning_rate": 0.0013762376237623764,
+      "loss": 0.6902,
+      "step": 791
+    },
+    {
+      "epoch": 1.7604890247290914,
+      "grad_norm": 0.035553187131881714,
+      "learning_rate": 0.001373762376237624,
+      "loss": 0.5251,
+      "step": 792
+    },
+    {
+      "epoch": 1.7627118644067796,
+      "grad_norm": 0.013903150334954262,
+      "learning_rate": 0.0013712871287128712,
+      "loss": 0.7792,
+      "step": 793
+    },
+    {
+      "epoch": 1.7649347040844678,
+      "grad_norm": 0.012611381709575653,
+      "learning_rate": 0.0013688118811881187,
+      "loss": 0.592,
+      "step": 794
+    },
+    {
+      "epoch": 1.7671575437621563,
+      "grad_norm": 0.014770290814340115,
+      "learning_rate": 0.0013663366336633665,
+      "loss": 0.5483,
+      "step": 795
+    },
+    {
+      "epoch": 1.7693803834398443,
+      "grad_norm": 0.02011730521917343,
+      "learning_rate": 0.0013638613861386138,
+      "loss": 0.7013,
+      "step": 796
+    },
+    {
+      "epoch": 1.7716032231175327,
+      "grad_norm": 0.024151911959052086,
+      "learning_rate": 0.0013613861386138613,
+      "loss": 0.6834,
+      "step": 797
+    },
+    {
+      "epoch": 1.773826062795221,
+      "grad_norm": 0.020802835002541542,
+      "learning_rate": 0.001358910891089109,
+      "loss": 0.6703,
+      "step": 798
+    },
+    {
+      "epoch": 1.7760489024729091,
+      "grad_norm": 0.027099374681711197,
+      "learning_rate": 0.0013564356435643563,
+      "loss": 0.6492,
+      "step": 799
+    },
+    {
+      "epoch": 1.7782717421505974,
+      "grad_norm": 0.015476135537028313,
+      "learning_rate": 0.0013539603960396039,
+      "loss": 0.7148,
+      "step": 800
+    },
+    {
+      "epoch": 1.7804945818282856,
+      "grad_norm": 0.016380343586206436,
+      "learning_rate": 0.0013514851485148516,
+      "loss": 0.65,
+      "step": 801
+    },
+    {
+      "epoch": 1.782717421505974,
+      "grad_norm": 0.01689559780061245,
+      "learning_rate": 0.001349009900990099,
+      "loss": 0.5779,
+      "step": 802
+    },
+    {
+      "epoch": 1.784940261183662,
+      "grad_norm": 0.021533837541937828,
+      "learning_rate": 0.0013465346534653464,
+      "loss": 0.5734,
+      "step": 803
+    },
+    {
+      "epoch": 1.7871631008613504,
+      "grad_norm": 0.025379789993166924,
+      "learning_rate": 0.0013440594059405942,
+      "loss": 0.554,
+      "step": 804
+    },
+    {
+      "epoch": 1.7893859405390387,
+      "grad_norm": 0.022600479423999786,
+      "learning_rate": 0.0013415841584158417,
+      "loss": 0.7596,
+      "step": 805
+    },
+    {
+      "epoch": 1.7916087802167269,
+      "grad_norm": 0.029088463634252548,
+      "learning_rate": 0.001339108910891089,
+      "loss": 0.6969,
+      "step": 806
+    },
+    {
+      "epoch": 1.793831619894415,
+      "grad_norm": 0.02145700715482235,
+      "learning_rate": 0.0013366336633663367,
+      "loss": 0.7427,
+      "step": 807
+    },
+    {
+      "epoch": 1.7960544595721033,
+      "grad_norm": 0.019825611263513565,
+      "learning_rate": 0.0013341584158415843,
+      "loss": 0.5842,
+      "step": 808
+    },
+    {
+      "epoch": 1.7982772992497917,
+      "grad_norm": 0.018616992980241776,
+      "learning_rate": 0.0013316831683168316,
+      "loss": 0.5571,
+      "step": 809
+    },
+    {
+      "epoch": 1.8005001389274797,
+      "grad_norm": 0.014252702705562115,
+      "learning_rate": 0.0013292079207920793,
+      "loss": 0.6742,
+      "step": 810
+    },
+    {
+      "epoch": 1.8027229786051682,
+      "grad_norm": 0.010596544481813908,
+      "learning_rate": 0.0013267326732673268,
+      "loss": 0.4869,
+      "step": 811
+    },
+    {
+      "epoch": 1.8049458182828564,
+      "grad_norm": 0.02070692926645279,
+      "learning_rate": 0.0013242574257425741,
+      "loss": 0.7206,
+      "step": 812
+    },
+    {
+      "epoch": 1.8071686579605446,
+      "grad_norm": 0.017880035564303398,
+      "learning_rate": 0.0013217821782178219,
+      "loss": 0.6991,
+      "step": 813
+    },
+    {
+      "epoch": 1.8093914976382328,
+      "grad_norm": 0.02070513367652893,
+      "learning_rate": 0.0013193069306930694,
+      "loss": 0.8145,
+      "step": 814
+    },
+    {
+      "epoch": 1.811614337315921,
+      "grad_norm": 0.016880514100193977,
+      "learning_rate": 0.0013168316831683167,
+      "loss": 0.5353,
+      "step": 815
+    },
+    {
+      "epoch": 1.8138371769936095,
+      "grad_norm": 0.017728716135025024,
+      "learning_rate": 0.0013143564356435644,
+      "loss": 0.6704,
+      "step": 816
+    },
+    {
+      "epoch": 1.8160600166712975,
+      "grad_norm": 0.026395466178655624,
+      "learning_rate": 0.001311881188118812,
+      "loss": 0.672,
+      "step": 817
+    },
+    {
+      "epoch": 1.818282856348986,
+      "grad_norm": 0.019495896995067596,
+      "learning_rate": 0.0013094059405940593,
+      "loss": 0.5702,
+      "step": 818
+    },
+    {
+      "epoch": 1.8205056960266741,
+      "grad_norm": 0.027116477489471436,
+      "learning_rate": 0.001306930693069307,
+      "loss": 0.5935,
+      "step": 819
+    },
+    {
+      "epoch": 1.8227285357043623,
+      "grad_norm": 0.014737212099134922,
+      "learning_rate": 0.0013044554455445545,
+      "loss": 0.5925,
+      "step": 820
+    },
+    {
+      "epoch": 1.8249513753820505,
+      "grad_norm": 0.015115857124328613,
+      "learning_rate": 0.0013019801980198018,
+      "loss": 0.614,
+      "step": 821
+    },
+    {
+      "epoch": 1.8271742150597388,
+      "grad_norm": 0.011744903400540352,
+      "learning_rate": 0.0012995049504950496,
+      "loss": 0.5326,
+      "step": 822
+    },
+    {
+      "epoch": 1.8293970547374272,
+      "grad_norm": 0.013531588949263096,
+      "learning_rate": 0.001297029702970297,
+      "loss": 0.6255,
+      "step": 823
+    },
+    {
+      "epoch": 1.8316198944151152,
+      "grad_norm": 0.014545679092407227,
+      "learning_rate": 0.0012945544554455444,
+      "loss": 0.6711,
+      "step": 824
+    },
+    {
+      "epoch": 1.8338427340928036,
+      "grad_norm": 0.02696896158158779,
+      "learning_rate": 0.0012920792079207921,
+      "loss": 0.7379,
+      "step": 825
+    },
+    {
+      "epoch": 1.8360655737704918,
+      "grad_norm": 0.017231903970241547,
+      "learning_rate": 0.0012896039603960397,
+      "loss": 0.6318,
+      "step": 826
+    },
+    {
+      "epoch": 1.83828841344818,
+      "grad_norm": 0.020397018641233444,
+      "learning_rate": 0.001287128712871287,
+      "loss": 0.7356,
+      "step": 827
+    },
+    {
+      "epoch": 1.8405112531258683,
+      "grad_norm": 0.014277629554271698,
+      "learning_rate": 0.0012846534653465347,
+      "loss": 0.6417,
+      "step": 828
+    },
+    {
+      "epoch": 1.8427340928035565,
+      "grad_norm": 0.014533923007547855,
+      "learning_rate": 0.0012821782178217822,
+      "loss": 0.558,
+      "step": 829
+    },
+    {
+      "epoch": 1.844956932481245,
+      "grad_norm": 0.014783985912799835,
+      "learning_rate": 0.0012797029702970298,
+      "loss": 0.7035,
+      "step": 830
+    },
+    {
+      "epoch": 1.847179772158933,
+      "grad_norm": 0.017703086137771606,
+      "learning_rate": 0.0012772277227722773,
+      "loss": 0.7231,
+      "step": 831
+    },
+    {
+      "epoch": 1.8494026118366214,
+      "grad_norm": 0.03160193935036659,
+      "learning_rate": 0.0012747524752475248,
+      "loss": 0.6218,
+      "step": 832
+    },
+    {
+      "epoch": 1.8516254515143096,
+      "grad_norm": 0.01616073027253151,
+      "learning_rate": 0.0012722772277227723,
+      "loss": 0.5358,
+      "step": 833
+    },
+    {
+      "epoch": 1.8538482911919978,
+      "grad_norm": 0.015707818791270256,
+      "learning_rate": 0.0012698019801980198,
+      "loss": 0.6487,
+      "step": 834
+    },
+    {
+      "epoch": 1.856071130869686,
+      "grad_norm": 0.027285683900117874,
+      "learning_rate": 0.0012673267326732674,
+      "loss": 0.7571,
+      "step": 835
+    },
+    {
+      "epoch": 1.8582939705473742,
+      "grad_norm": 0.013436183333396912,
+      "learning_rate": 0.0012648514851485149,
+      "loss": 0.5062,
+      "step": 836
+    },
+    {
+      "epoch": 1.8605168102250627,
+      "grad_norm": 0.012100230902433395,
+      "learning_rate": 0.0012623762376237624,
+      "loss": 0.6001,
+      "step": 837
+    },
+    {
+      "epoch": 1.8627396499027506,
+      "grad_norm": 0.016179142519831657,
+      "learning_rate": 0.00125990099009901,
+      "loss": 0.5925,
+      "step": 838
+    },
+    {
+      "epoch": 1.864962489580439,
+      "grad_norm": 0.01836666651070118,
+      "learning_rate": 0.0012574257425742575,
+      "loss": 0.7492,
+      "step": 839
+    },
+    {
+      "epoch": 1.8671853292581273,
+      "grad_norm": 0.013563703745603561,
+      "learning_rate": 0.001254950495049505,
+      "loss": 0.5668,
+      "step": 840
+    },
+    {
+      "epoch": 1.8694081689358155,
+      "grad_norm": 0.02543150633573532,
+      "learning_rate": 0.0012524752475247525,
+      "loss": 0.5982,
+      "step": 841
+    },
+    {
+      "epoch": 1.8716310086135037,
+      "grad_norm": 0.017902787774801254,
+      "learning_rate": 0.00125,
+      "loss": 0.6887,
+      "step": 842
+    },
+    {
+      "epoch": 1.873853848291192,
+      "grad_norm": 0.018437763676047325,
+      "learning_rate": 0.0012475247524752475,
+      "loss": 0.6153,
+      "step": 843
+    },
+    {
+      "epoch": 1.8760766879688804,
+      "grad_norm": 0.01389164850115776,
+      "learning_rate": 0.001245049504950495,
+      "loss": 0.6548,
+      "step": 844
+    },
+    {
+      "epoch": 1.8782995276465684,
+      "grad_norm": 0.03561052680015564,
+      "learning_rate": 0.0012425742574257426,
+      "loss": 0.5777,
+      "step": 845
+    },
+    {
+      "epoch": 1.8805223673242568,
+      "grad_norm": 0.014207260683178902,
+      "learning_rate": 0.0012400990099009901,
+      "loss": 0.5582,
+      "step": 846
+    },
+    {
+      "epoch": 1.882745207001945,
+      "grad_norm": 0.046381477266550064,
+      "learning_rate": 0.0012376237623762376,
+      "loss": 0.605,
+      "step": 847
+    },
+    {
+      "epoch": 1.8849680466796332,
+      "grad_norm": 0.015709543600678444,
+      "learning_rate": 0.0012351485148514852,
+      "loss": 0.5713,
+      "step": 848
+    },
+    {
+      "epoch": 1.8871908863573215,
+      "grad_norm": 0.01592281460762024,
+      "learning_rate": 0.0012326732673267327,
+      "loss": 0.618,
+      "step": 849
+    },
+    {
+      "epoch": 1.8894137260350097,
+      "grad_norm": 0.01385457906872034,
+      "learning_rate": 0.0012301980198019802,
+      "loss": 0.5394,
+      "step": 850
+    },
+    {
+      "epoch": 1.8916365657126981,
+      "grad_norm": 0.02153988927602768,
+      "learning_rate": 0.0012277227722772277,
+      "loss": 0.5842,
+      "step": 851
+    },
+    {
+      "epoch": 1.893859405390386,
+      "grad_norm": 0.02466188557446003,
+      "learning_rate": 0.0012252475247524753,
+      "loss": 0.5083,
+      "step": 852
+    },
+    {
+      "epoch": 1.8960822450680745,
+      "grad_norm": 0.015681538730859756,
+      "learning_rate": 0.0012227722772277228,
+      "loss": 0.7941,
+      "step": 853
+    },
+    {
+      "epoch": 1.8983050847457628,
+      "grad_norm": 0.012338684871792793,
+      "learning_rate": 0.0012202970297029703,
+      "loss": 0.5342,
+      "step": 854
+    },
+    {
+      "epoch": 1.900527924423451,
+      "grad_norm": 0.01564796082675457,
+      "learning_rate": 0.001217821782178218,
+      "loss": 0.7584,
+      "step": 855
+    },
+    {
+      "epoch": 1.9027507641011392,
+      "grad_norm": 0.02049081400036812,
+      "learning_rate": 0.0012153465346534653,
+      "loss": 0.644,
+      "step": 856
+    },
+    {
+      "epoch": 1.9049736037788274,
+      "grad_norm": 0.029596436768770218,
+      "learning_rate": 0.0012128712871287129,
+      "loss": 0.7099,
+      "step": 857
+    },
+    {
+      "epoch": 1.9071964434565158,
+      "grad_norm": 0.013466271571815014,
+      "learning_rate": 0.0012103960396039606,
+      "loss": 0.649,
+      "step": 858
+    },
+    {
+      "epoch": 1.9094192831342038,
+      "grad_norm": 0.02464134432375431,
+      "learning_rate": 0.001207920792079208,
+      "loss": 0.5849,
+      "step": 859
+    },
+    {
+      "epoch": 1.9116421228118923,
+      "grad_norm": 0.0143961850553751,
+      "learning_rate": 0.0012054455445544554,
+      "loss": 0.6556,
+      "step": 860
+    },
+    {
+      "epoch": 1.9138649624895805,
+      "grad_norm": 0.016098588705062866,
+      "learning_rate": 0.0012029702970297032,
+      "loss": 0.6991,
+      "step": 861
+    },
+    {
+      "epoch": 1.9160878021672687,
+      "grad_norm": 0.013091151602566242,
+      "learning_rate": 0.0012004950495049505,
+      "loss": 0.6811,
+      "step": 862
+    },
+    {
+      "epoch": 1.918310641844957,
+      "grad_norm": 0.029850434511899948,
+      "learning_rate": 0.001198019801980198,
+      "loss": 0.703,
+      "step": 863
+    },
+    {
+      "epoch": 1.9205334815226451,
+      "grad_norm": 0.01120046991854906,
+      "learning_rate": 0.0011955445544554457,
+      "loss": 0.4938,
+      "step": 864
+    },
+    {
+      "epoch": 1.9227563212003336,
+      "grad_norm": 0.014592386782169342,
+      "learning_rate": 0.001193069306930693,
+      "loss": 0.6209,
+      "step": 865
+    },
+    {
+      "epoch": 1.9249791608780216,
+      "grad_norm": 0.015447413548827171,
+      "learning_rate": 0.0011905940594059406,
+      "loss": 0.5492,
+      "step": 866
+    },
+    {
+      "epoch": 1.92720200055571,
+      "grad_norm": 0.015509175136685371,
+      "learning_rate": 0.0011881188118811883,
+      "loss": 0.54,
+      "step": 867
+    },
+    {
+      "epoch": 1.9294248402333982,
+      "grad_norm": 0.01778079755604267,
+      "learning_rate": 0.0011856435643564356,
+      "loss": 0.6965,
+      "step": 868
+    },
+    {
+      "epoch": 1.9316476799110864,
+      "grad_norm": 0.02252313494682312,
+      "learning_rate": 0.0011831683168316831,
+      "loss": 0.5413,
+      "step": 869
+    },
+    {
+      "epoch": 1.9338705195887747,
+      "grad_norm": 0.012748087756335735,
+      "learning_rate": 0.0011806930693069309,
+      "loss": 0.5079,
+      "step": 870
+    },
+    {
+      "epoch": 1.9360933592664629,
+      "grad_norm": 0.016808731481432915,
+      "learning_rate": 0.0011782178217821782,
+      "loss": 0.6487,
+      "step": 871
+    },
+    {
+      "epoch": 1.9383161989441513,
+      "grad_norm": 0.017723703756928444,
+      "learning_rate": 0.0011757425742574257,
+      "loss": 0.7172,
+      "step": 872
+    },
+    {
+      "epoch": 1.9405390386218393,
+      "grad_norm": 0.014089664444327354,
+      "learning_rate": 0.0011732673267326734,
+      "loss": 0.4578,
+      "step": 873
+    },
+    {
+      "epoch": 1.9427618782995277,
+      "grad_norm": 0.019459979608654976,
+      "learning_rate": 0.0011707920792079207,
+      "loss": 0.5246,
+      "step": 874
+    },
+    {
+      "epoch": 1.9449847179772157,
+      "grad_norm": 0.017233047634363174,
+      "learning_rate": 0.0011683168316831683,
+      "loss": 0.6245,
+      "step": 875
+    },
+    {
+      "epoch": 1.9472075576549042,
+      "grad_norm": 0.019499676302075386,
+      "learning_rate": 0.001165841584158416,
+      "loss": 0.6322,
+      "step": 876
+    },
+    {
+      "epoch": 1.9494303973325924,
+      "grad_norm": 0.022460227832198143,
+      "learning_rate": 0.0011633663366336633,
+      "loss": 0.6601,
+      "step": 877
+    },
+    {
+      "epoch": 1.9516532370102806,
+      "grad_norm": 0.020194463431835175,
+      "learning_rate": 0.0011608910891089108,
+      "loss": 0.7348,
+      "step": 878
+    },
+    {
+      "epoch": 1.953876076687969,
+      "grad_norm": 0.018907731398940086,
+      "learning_rate": 0.0011584158415841586,
+      "loss": 0.6646,
+      "step": 879
+    },
+    {
+      "epoch": 1.956098916365657,
+      "grad_norm": 0.015820452943444252,
+      "learning_rate": 0.001155940594059406,
+      "loss": 0.5312,
+      "step": 880
+    },
+    {
+      "epoch": 1.9583217560433455,
+      "grad_norm": 0.018061013892292976,
+      "learning_rate": 0.0011534653465346534,
+      "loss": 0.6744,
+      "step": 881
+    },
+    {
+      "epoch": 1.9605445957210335,
+      "grad_norm": 0.02651524357497692,
+      "learning_rate": 0.0011509900990099011,
+      "loss": 0.6111,
+      "step": 882
+    },
+    {
+      "epoch": 1.962767435398722,
+      "grad_norm": 0.014137241058051586,
+      "learning_rate": 0.0011485148514851487,
+      "loss": 0.5668,
+      "step": 883
+    },
+    {
+      "epoch": 1.96499027507641,
+      "grad_norm": 0.02786339819431305,
+      "learning_rate": 0.001146039603960396,
+      "loss": 0.6426,
+      "step": 884
+    },
+    {
+      "epoch": 1.9672131147540983,
+      "grad_norm": 0.012125310488045216,
+      "learning_rate": 0.0011435643564356437,
+      "loss": 0.6304,
+      "step": 885
+    },
+    {
+      "epoch": 1.9694359544317868,
+      "grad_norm": 0.019771946594119072,
+      "learning_rate": 0.0011410891089108912,
+      "loss": 0.6049,
+      "step": 886
+    },
+    {
+      "epoch": 1.9716587941094748,
+      "grad_norm": 0.028338497504591942,
+      "learning_rate": 0.0011386138613861385,
+      "loss": 0.7123,
+      "step": 887
+    },
+    {
+      "epoch": 1.9738816337871632,
+      "grad_norm": 0.01594017632305622,
+      "learning_rate": 0.0011361386138613863,
+      "loss": 0.6499,
+      "step": 888
+    },
+    {
+      "epoch": 1.9761044734648512,
+      "grad_norm": 0.015689987689256668,
+      "learning_rate": 0.0011336633663366338,
+      "loss": 0.6692,
+      "step": 889
+    },
+    {
+      "epoch": 1.9783273131425396,
+      "grad_norm": 0.014462132938206196,
+      "learning_rate": 0.0011311881188118811,
+      "loss": 0.7722,
+      "step": 890
+    },
+    {
+      "epoch": 1.9805501528202278,
+      "grad_norm": 0.01385975256562233,
+      "learning_rate": 0.0011287128712871288,
+      "loss": 0.4916,
+      "step": 891
+    },
+    {
+      "epoch": 1.982772992497916,
+      "grad_norm": 0.01998106576502323,
+      "learning_rate": 0.0011262376237623764,
+      "loss": 0.4741,
+      "step": 892
+    },
+    {
+      "epoch": 1.9849958321756045,
+      "grad_norm": 0.0200062096118927,
+      "learning_rate": 0.0011237623762376237,
+      "loss": 0.6419,
+      "step": 893
+    },
+    {
+      "epoch": 1.9872186718532925,
+      "grad_norm": 0.01526566967368126,
+      "learning_rate": 0.0011212871287128712,
+      "loss": 0.7451,
+      "step": 894
+    },
+    {
+      "epoch": 1.989441511530981,
+      "grad_norm": 0.022502193227410316,
+      "learning_rate": 0.001118811881188119,
+      "loss": 0.7621,
+      "step": 895
+    },
+    {
+      "epoch": 1.991664351208669,
+      "grad_norm": 0.01716960035264492,
+      "learning_rate": 0.0011163366336633662,
+      "loss": 0.522,
+      "step": 896
+    },
+    {
+      "epoch": 1.9938871908863574,
+      "grad_norm": 0.013558654114603996,
+      "learning_rate": 0.0011138613861386138,
+      "loss": 0.5354,
+      "step": 897
+    },
+    {
+      "epoch": 1.9961100305640456,
+      "grad_norm": 0.012157892808318138,
+      "learning_rate": 0.0011113861386138615,
+      "loss": 0.5341,
+      "step": 898
+    },
+    {
+      "epoch": 1.9983328702417338,
+      "grad_norm": 0.020689917728304863,
+      "learning_rate": 0.0011089108910891088,
+      "loss": 0.7788,
+      "step": 899
+    },
+    {
+      "epoch": 2.000555709919422,
+      "grad_norm": 0.013700471259653568,
+      "learning_rate": 0.0011064356435643563,
+      "loss": 0.4876,
+      "step": 900
+    },
+    {
+      "epoch": 2.00277854959711,
+      "grad_norm": 0.022665875032544136,
+      "learning_rate": 0.001103960396039604,
+      "loss": 0.549,
+      "step": 901
+    },
+    {
+      "epoch": 2.0050013892747987,
+      "grad_norm": 0.012792624533176422,
+      "learning_rate": 0.0011014851485148514,
+      "loss": 0.5934,
+      "step": 902
+    },
+    {
+      "epoch": 2.0072242289524866,
+      "grad_norm": 0.014675344340503216,
+      "learning_rate": 0.001099009900990099,
+      "loss": 0.6914,
+      "step": 903
+    },
+    {
+      "epoch": 2.009447068630175,
+      "grad_norm": 0.02120724506676197,
+      "learning_rate": 0.0010965346534653466,
+      "loss": 0.6156,
+      "step": 904
+    },
+    {
+      "epoch": 2.011669908307863,
+      "grad_norm": 0.014748603105545044,
+      "learning_rate": 0.001094059405940594,
+      "loss": 0.622,
+      "step": 905
+    },
+    {
+      "epoch": 2.0138927479855515,
+      "grad_norm": 0.014371293596923351,
+      "learning_rate": 0.0010915841584158415,
+      "loss": 0.5474,
+      "step": 906
+    },
+    {
+      "epoch": 2.01611558766324,
+      "grad_norm": 0.017763959243893623,
+      "learning_rate": 0.0010891089108910892,
+      "loss": 0.5761,
+      "step": 907
+    },
+    {
+      "epoch": 2.018338427340928,
+      "grad_norm": 0.01812128722667694,
+      "learning_rate": 0.0010866336633663367,
+      "loss": 0.7893,
+      "step": 908
+    },
+    {
+      "epoch": 2.0205612670186164,
+      "grad_norm": 0.012955384328961372,
+      "learning_rate": 0.001084158415841584,
+      "loss": 0.5467,
+      "step": 909
+    },
+    {
+      "epoch": 2.0227841066963044,
+      "grad_norm": 0.01600860245525837,
+      "learning_rate": 0.0010816831683168318,
+      "loss": 0.6869,
+      "step": 910
+    },
+    {
+      "epoch": 2.025006946373993,
+      "grad_norm": 0.018638163805007935,
+      "learning_rate": 0.0010792079207920793,
+      "loss": 0.5909,
+      "step": 911
+    },
+    {
+      "epoch": 2.027229786051681,
+      "grad_norm": 0.018940769135951996,
+      "learning_rate": 0.0010767326732673266,
+      "loss": 0.7257,
+      "step": 912
+    },
+    {
+      "epoch": 2.0294526257293692,
+      "grad_norm": 0.016438838094472885,
+      "learning_rate": 0.0010742574257425743,
+      "loss": 0.6541,
+      "step": 913
+    },
+    {
+      "epoch": 2.0316754654070577,
+      "grad_norm": 0.015446559526026249,
+      "learning_rate": 0.0010717821782178219,
+      "loss": 0.4858,
+      "step": 914
+    },
+    {
+      "epoch": 2.0338983050847457,
+      "grad_norm": 0.01849900372326374,
+      "learning_rate": 0.0010693069306930692,
+      "loss": 0.642,
+      "step": 915
+    },
+    {
+      "epoch": 2.036121144762434,
+      "grad_norm": 0.016022328287363052,
+      "learning_rate": 0.001066831683168317,
+      "loss": 0.5797,
+      "step": 916
+    },
+    {
+      "epoch": 2.038343984440122,
+      "grad_norm": 0.011567950248718262,
+      "learning_rate": 0.0010643564356435644,
+      "loss": 0.5816,
+      "step": 917
+    },
+    {
+      "epoch": 2.0405668241178105,
+      "grad_norm": 0.013285503722727299,
+      "learning_rate": 0.0010618811881188117,
+      "loss": 0.651,
+      "step": 918
+    },
+    {
+      "epoch": 2.0427896637954985,
+      "grad_norm": 0.020052414387464523,
+      "learning_rate": 0.0010594059405940595,
+      "loss": 0.6495,
+      "step": 919
+    },
+    {
+      "epoch": 2.045012503473187,
+      "grad_norm": 0.02657526172697544,
+      "learning_rate": 0.001056930693069307,
+      "loss": 0.6453,
+      "step": 920
+    },
+    {
+      "epoch": 2.0472353431508754,
+      "grad_norm": 0.019164010882377625,
+      "learning_rate": 0.0010544554455445543,
+      "loss": 0.5374,
+      "step": 921
+    },
+    {
+      "epoch": 2.0494581828285634,
+      "grad_norm": 0.016299502924084663,
+      "learning_rate": 0.001051980198019802,
+      "loss": 0.6959,
+      "step": 922
+    },
+    {
+      "epoch": 2.051681022506252,
+      "grad_norm": 0.01600020006299019,
+      "learning_rate": 0.0010495049504950496,
+      "loss": 0.645,
+      "step": 923
+    },
+    {
+      "epoch": 2.05390386218394,
+      "grad_norm": 0.018981320783495903,
+      "learning_rate": 0.0010470297029702969,
+      "loss": 0.7247,
+      "step": 924
+    },
+    {
+      "epoch": 2.0561267018616283,
+      "grad_norm": 0.018881479278206825,
+      "learning_rate": 0.0010445544554455446,
+      "loss": 0.668,
+      "step": 925
+    },
+    {
+      "epoch": 2.0583495415393163,
+      "grad_norm": 0.030967555940151215,
+      "learning_rate": 0.0010420792079207921,
+      "loss": 0.5847,
+      "step": 926
+    },
+    {
+      "epoch": 2.0605723812170047,
+      "grad_norm": 0.017560211941599846,
+      "learning_rate": 0.0010396039603960394,
+      "loss": 0.6954,
+      "step": 927
+    },
+    {
+      "epoch": 2.062795220894693,
+      "grad_norm": 0.01360076479613781,
+      "learning_rate": 0.0010371287128712872,
+      "loss": 0.6609,
+      "step": 928
+    },
+    {
+      "epoch": 2.065018060572381,
+      "grad_norm": 0.016593527048826218,
+      "learning_rate": 0.0010346534653465347,
+      "loss": 0.6066,
+      "step": 929
+    },
+    {
+      "epoch": 2.0672409002500696,
+      "grad_norm": 0.014666872099041939,
+      "learning_rate": 0.001032178217821782,
+      "loss": 0.5952,
+      "step": 930
+    },
+    {
+      "epoch": 2.0694637399277576,
+      "grad_norm": 0.014250427484512329,
+      "learning_rate": 0.0010297029702970298,
+      "loss": 0.5504,
+      "step": 931
+    },
+    {
+      "epoch": 2.071686579605446,
+      "grad_norm": 0.02060864120721817,
+      "learning_rate": 0.0010272277227722773,
+      "loss": 0.5399,
+      "step": 932
+    },
+    {
+      "epoch": 2.073909419283134,
+      "grad_norm": 0.020173972472548485,
+      "learning_rate": 0.0010247524752475248,
+      "loss": 0.6484,
+      "step": 933
+    },
+    {
+      "epoch": 2.0761322589608224,
+      "grad_norm": 0.017474539577960968,
+      "learning_rate": 0.0010222772277227723,
+      "loss": 0.5387,
+      "step": 934
+    },
+    {
+      "epoch": 2.078355098638511,
+      "grad_norm": 0.014941662549972534,
+      "learning_rate": 0.0010198019801980198,
+      "loss": 0.5956,
+      "step": 935
+    },
+    {
+      "epoch": 2.080577938316199,
+      "grad_norm": 0.027800098061561584,
+      "learning_rate": 0.0010173267326732674,
+      "loss": 0.601,
+      "step": 936
+    },
+    {
+      "epoch": 2.0828007779938873,
+      "grad_norm": 0.01730935461819172,
+      "learning_rate": 0.0010148514851485149,
+      "loss": 0.7134,
+      "step": 937
+    },
+    {
+      "epoch": 2.0850236176715753,
+      "grad_norm": 0.01698504388332367,
+      "learning_rate": 0.0010123762376237624,
+      "loss": 0.5066,
+      "step": 938
+    },
+    {
+      "epoch": 2.0872464573492637,
+      "grad_norm": 0.015243062749505043,
+      "learning_rate": 0.00100990099009901,
+      "loss": 0.657,
+      "step": 939
+    },
+    {
+      "epoch": 2.0894692970269517,
+      "grad_norm": 0.025019872933626175,
+      "learning_rate": 0.0010074257425742575,
+      "loss": 0.6808,
+      "step": 940
+    },
+    {
+      "epoch": 2.09169213670464,
+      "grad_norm": 0.012755641713738441,
+      "learning_rate": 0.001004950495049505,
+      "loss": 0.5509,
+      "step": 941
+    },
+    {
+      "epoch": 2.0939149763823286,
+      "grad_norm": 0.01922733336687088,
+      "learning_rate": 0.0010024752475247525,
+      "loss": 0.8053,
+      "step": 942
+    },
+    {
+      "epoch": 2.0961378160600166,
+      "grad_norm": 0.03526071086525917,
+      "learning_rate": 0.001,
+      "loss": 0.6377,
+      "step": 943
+    },
+    {
+      "epoch": 2.098360655737705,
+      "grad_norm": 0.021753890439867973,
+      "learning_rate": 0.0009975247524752475,
+      "loss": 0.727,
+      "step": 944
+    },
+    {
+      "epoch": 2.100583495415393,
+      "grad_norm": 0.016603440046310425,
+      "learning_rate": 0.000995049504950495,
+      "loss": 0.6584,
+      "step": 945
+    },
+    {
+      "epoch": 2.1028063350930815,
+      "grad_norm": 0.021078625693917274,
+      "learning_rate": 0.0009925742574257426,
+      "loss": 0.6133,
+      "step": 946
+    },
+    {
+      "epoch": 2.1050291747707695,
+      "grad_norm": 0.037239789962768555,
+      "learning_rate": 0.0009900990099009901,
+      "loss": 0.6421,
+      "step": 947
+    },
+    {
+      "epoch": 2.107252014448458,
+      "grad_norm": 0.012814635410904884,
+      "learning_rate": 0.0009876237623762376,
+      "loss": 0.6296,
+      "step": 948
+    },
+    {
+      "epoch": 2.1094748541261463,
+      "grad_norm": 0.018249889835715294,
+      "learning_rate": 0.0009851485148514852,
+      "loss": 0.6202,
+      "step": 949
+    },
+    {
+      "epoch": 2.1116976938038343,
+      "grad_norm": 0.0173304695636034,
+      "learning_rate": 0.0009826732673267327,
+      "loss": 0.5801,
+      "step": 950
+    },
+    {
+      "epoch": 2.1139205334815228,
+      "grad_norm": 0.020902493968605995,
+      "learning_rate": 0.0009801980198019802,
+      "loss": 0.7202,
+      "step": 951
+    },
+    {
+      "epoch": 2.1161433731592107,
+      "grad_norm": 0.020156148821115494,
+      "learning_rate": 0.0009777227722772277,
+      "loss": 0.6196,
+      "step": 952
+    },
+    {
+      "epoch": 2.118366212836899,
+      "grad_norm": 0.020738055929541588,
+      "learning_rate": 0.0009752475247524752,
+      "loss": 0.573,
+      "step": 953
+    },
+    {
+      "epoch": 2.120589052514587,
+      "grad_norm": 0.015748674049973488,
+      "learning_rate": 0.0009727722772277229,
+      "loss": 0.5665,
+      "step": 954
+    },
+    {
+      "epoch": 2.1228118921922756,
+      "grad_norm": 0.01676134020090103,
+      "learning_rate": 0.0009702970297029702,
+      "loss": 0.6478,
+      "step": 955
+    },
+    {
+      "epoch": 2.125034731869964,
+      "grad_norm": 0.014315289445221424,
+      "learning_rate": 0.0009678217821782178,
+      "loss": 0.6188,
+      "step": 956
+    },
+    {
+      "epoch": 2.127257571547652,
+      "grad_norm": 0.013696223497390747,
+      "learning_rate": 0.0009653465346534654,
+      "loss": 0.6517,
+      "step": 957
+    },
+    {
+      "epoch": 2.1294804112253405,
+      "grad_norm": 0.01995706558227539,
+      "learning_rate": 0.000962871287128713,
+      "loss": 0.6499,
+      "step": 958
+    },
+    {
+      "epoch": 2.1317032509030285,
+      "grad_norm": 0.01407367642968893,
+      "learning_rate": 0.0009603960396039604,
+      "loss": 0.5505,
+      "step": 959
+    },
+    {
+      "epoch": 2.133926090580717,
+      "grad_norm": 0.013528737239539623,
+      "learning_rate": 0.000957920792079208,
+      "loss": 0.4975,
+      "step": 960
+    },
+    {
+      "epoch": 2.136148930258405,
+      "grad_norm": 0.020059697329998016,
+      "learning_rate": 0.0009554455445544555,
+      "loss": 0.7414,
+      "step": 961
+    },
+    {
+      "epoch": 2.1383717699360933,
+      "grad_norm": 0.01345049124211073,
+      "learning_rate": 0.000952970297029703,
+      "loss": 0.5884,
+      "step": 962
+    },
+    {
+      "epoch": 2.140594609613782,
+      "grad_norm": 0.013034477829933167,
+      "learning_rate": 0.0009504950495049506,
+      "loss": 0.5133,
+      "step": 963
+    },
+    {
+      "epoch": 2.14281744929147,
+      "grad_norm": 0.01752249337732792,
+      "learning_rate": 0.0009480198019801981,
+      "loss": 0.6086,
+      "step": 964
+    },
+    {
+      "epoch": 2.145040288969158,
+      "grad_norm": 0.02016667276620865,
+      "learning_rate": 0.0009455445544554455,
+      "loss": 0.6729,
+      "step": 965
+    },
+    {
+      "epoch": 2.147263128646846,
+      "grad_norm": 0.01611597090959549,
+      "learning_rate": 0.0009430693069306931,
+      "loss": 0.5042,
+      "step": 966
+    },
+    {
+      "epoch": 2.1494859683245346,
+      "grad_norm": 0.015491276048123837,
+      "learning_rate": 0.0009405940594059407,
+      "loss": 0.4197,
+      "step": 967
+    },
+    {
+      "epoch": 2.1517088080022226,
+      "grad_norm": 0.0135455671697855,
+      "learning_rate": 0.0009381188118811881,
+      "loss": 0.6672,
+      "step": 968
+    },
+    {
+      "epoch": 2.153931647679911,
+      "grad_norm": 0.018375838175415993,
+      "learning_rate": 0.0009356435643564356,
+      "loss": 0.6353,
+      "step": 969
+    },
+    {
+      "epoch": 2.1561544873575995,
+      "grad_norm": 0.019288305193185806,
+      "learning_rate": 0.0009331683168316832,
+      "loss": 0.5963,
+      "step": 970
+    },
+    {
+      "epoch": 2.1583773270352875,
+      "grad_norm": 0.018923213705420494,
+      "learning_rate": 0.0009306930693069307,
+      "loss": 0.7183,
+      "step": 971
+    },
+    {
+      "epoch": 2.160600166712976,
+      "grad_norm": 0.02921975590288639,
+      "learning_rate": 0.0009282178217821782,
+      "loss": 0.5991,
+      "step": 972
+    },
+    {
+      "epoch": 2.162823006390664,
+      "grad_norm": 0.020133277401328087,
+      "learning_rate": 0.0009257425742574258,
+      "loss": 0.5406,
+      "step": 973
+    },
+    {
+      "epoch": 2.1650458460683524,
+      "grad_norm": 0.014706325717270374,
+      "learning_rate": 0.0009232673267326732,
+      "loss": 0.5322,
+      "step": 974
+    },
+    {
+      "epoch": 2.1672686857460404,
+      "grad_norm": 0.030446428805589676,
+      "learning_rate": 0.0009207920792079207,
+      "loss": 0.6796,
+      "step": 975
+    },
+    {
+      "epoch": 2.169491525423729,
+      "grad_norm": 0.015138603746891022,
+      "learning_rate": 0.0009183168316831684,
+      "loss": 0.5747,
+      "step": 976
+    },
+    {
+      "epoch": 2.1717143651014172,
+      "grad_norm": 0.01757650263607502,
+      "learning_rate": 0.0009158415841584158,
+      "loss": 0.7159,
+      "step": 977
+    },
+    {
+      "epoch": 2.1739372047791052,
+      "grad_norm": 0.017245231196284294,
+      "learning_rate": 0.0009133663366336633,
+      "loss": 0.5669,
+      "step": 978
+    },
+    {
+      "epoch": 2.1761600444567937,
+      "grad_norm": 0.018706029281020164,
+      "learning_rate": 0.0009108910891089109,
+      "loss": 0.5436,
+      "step": 979
+    },
+    {
+      "epoch": 2.1783828841344817,
+      "grad_norm": 0.010909450240433216,
+      "learning_rate": 0.0009084158415841584,
+      "loss": 0.5101,
+      "step": 980
+    },
+    {
+      "epoch": 2.18060572381217,
+      "grad_norm": 0.01844753511250019,
+      "learning_rate": 0.0009059405940594059,
+      "loss": 0.5766,
+      "step": 981
+    },
+    {
+      "epoch": 2.182828563489858,
+      "grad_norm": 0.01436743326485157,
+      "learning_rate": 0.0009034653465346535,
+      "loss": 0.4609,
+      "step": 982
+    },
+    {
+      "epoch": 2.1850514031675465,
+      "grad_norm": 0.015225561335682869,
+      "learning_rate": 0.0009009900990099011,
+      "loss": 0.5411,
+      "step": 983
+    },
+    {
+      "epoch": 2.187274242845235,
+      "grad_norm": 0.020290294662117958,
+      "learning_rate": 0.0008985148514851484,
+      "loss": 0.6755,
+      "step": 984
+    },
+    {
+      "epoch": 2.189497082522923,
+      "grad_norm": 0.013398604467511177,
+      "learning_rate": 0.0008960396039603961,
+      "loss": 0.5362,
+      "step": 985
+    },
+    {
+      "epoch": 2.1917199222006114,
+      "grad_norm": 0.01631542108952999,
+      "learning_rate": 0.0008935643564356437,
+      "loss": 0.5372,
+      "step": 986
+    },
+    {
+      "epoch": 2.1939427618782994,
+      "grad_norm": 0.014349683187901974,
+      "learning_rate": 0.000891089108910891,
+      "loss": 0.4823,
+      "step": 987
+    },
+    {
+      "epoch": 2.196165601555988,
+      "grad_norm": 0.015834227204322815,
+      "learning_rate": 0.0008886138613861386,
+      "loss": 0.4641,
+      "step": 988
+    },
+    {
+      "epoch": 2.198388441233676,
+      "grad_norm": 0.022029682993888855,
+      "learning_rate": 0.0008861386138613863,
+      "loss": 0.5842,
+      "step": 989
+    },
+    {
+      "epoch": 2.2006112809113643,
+      "grad_norm": 0.014018687419593334,
+      "learning_rate": 0.0008836633663366336,
+      "loss": 0.5924,
+      "step": 990
+    },
+    {
+      "epoch": 2.2028341205890527,
+      "grad_norm": 0.0164606012403965,
+      "learning_rate": 0.0008811881188118812,
+      "loss": 0.4709,
+      "step": 991
+    },
+    {
+      "epoch": 2.2050569602667407,
+      "grad_norm": 0.014225888065993786,
+      "learning_rate": 0.0008787128712871288,
+      "loss": 0.6046,
+      "step": 992
+    },
+    {
+      "epoch": 2.207279799944429,
+      "grad_norm": 0.018137982115149498,
+      "learning_rate": 0.0008762376237623761,
+      "loss": 0.6395,
+      "step": 993
+    },
+    {
+      "epoch": 2.209502639622117,
+      "grad_norm": 0.015276088379323483,
+      "learning_rate": 0.0008737623762376238,
+      "loss": 0.6237,
+      "step": 994
+    },
+    {
+      "epoch": 2.2117254792998056,
+      "grad_norm": 0.019907942041754723,
+      "learning_rate": 0.0008712871287128714,
+      "loss": 0.6448,
+      "step": 995
+    },
+    {
+      "epoch": 2.2139483189774936,
+      "grad_norm": 0.024012433364987373,
+      "learning_rate": 0.0008688118811881187,
+      "loss": 0.7278,
+      "step": 996
+    },
+    {
+      "epoch": 2.216171158655182,
+      "grad_norm": 0.021498383954167366,
+      "learning_rate": 0.0008663366336633663,
+      "loss": 0.4841,
+      "step": 997
+    },
+    {
+      "epoch": 2.2183939983328704,
+      "grad_norm": 0.017866021022200584,
+      "learning_rate": 0.000863861386138614,
+      "loss": 0.6018,
+      "step": 998
+    },
+    {
+      "epoch": 2.2206168380105584,
+      "grad_norm": 0.01377987489104271,
+      "learning_rate": 0.0008613861386138613,
+      "loss": 0.5827,
+      "step": 999
+    },
+    {
+      "epoch": 2.222839677688247,
+      "grad_norm": 0.016971712931990623,
+      "learning_rate": 0.0008589108910891089,
+      "loss": 0.5961,
+      "step": 1000
+    },
+    {
+      "epoch": 2.225062517365935,
+      "grad_norm": 0.02038002572953701,
+      "learning_rate": 0.0008564356435643565,
+      "loss": 0.7805,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2272853570436233,
+      "grad_norm": 0.016022466123104095,
+      "learning_rate": 0.0008539603960396039,
+      "loss": 0.5474,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2295081967213113,
+      "grad_norm": 0.020060956478118896,
+      "learning_rate": 0.0008514851485148515,
+      "loss": 0.5361,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2317310363989997,
+      "grad_norm": 0.03471532464027405,
+      "learning_rate": 0.0008490099009900991,
+      "loss": 0.6665,
+      "step": 1004
+    },
+    {
+      "epoch": 2.233953876076688,
+      "grad_norm": 0.013896548189222813,
+      "learning_rate": 0.0008465346534653464,
+      "loss": 0.7585,
+      "step": 1005
+    },
+    {
+      "epoch": 2.236176715754376,
+      "grad_norm": 0.01731082797050476,
+      "learning_rate": 0.000844059405940594,
+      "loss": 0.6095,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2383995554320646,
+      "grad_norm": 0.022298304364085197,
+      "learning_rate": 0.0008415841584158417,
+      "loss": 0.6763,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2406223951097526,
+      "grad_norm": 0.02241886593401432,
+      "learning_rate": 0.0008391089108910892,
+      "loss": 0.6923,
+      "step": 1008
+    },
+    {
+      "epoch": 2.242845234787441,
+      "grad_norm": 0.01641622744500637,
+      "learning_rate": 0.0008366336633663366,
+      "loss": 0.4781,
+      "step": 1009
+    },
+    {
+      "epoch": 2.245068074465129,
+      "grad_norm": 0.01847672462463379,
+      "learning_rate": 0.0008341584158415842,
+      "loss": 0.7357,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2472909141428175,
+      "grad_norm": 0.0181216262280941,
+      "learning_rate": 0.0008316831683168318,
+      "loss": 0.6434,
+      "step": 1011
+    },
+    {
+      "epoch": 2.249513753820506,
+      "grad_norm": 0.016178471967577934,
+      "learning_rate": 0.0008292079207920792,
+      "loss": 0.4554,
+      "step": 1012
+    },
+    {
+      "epoch": 2.251736593498194,
+      "grad_norm": 0.017215240746736526,
+      "learning_rate": 0.0008267326732673268,
+      "loss": 0.4935,
+      "step": 1013
+    },
+    {
+      "epoch": 2.2539594331758823,
+      "grad_norm": 0.014085826463997364,
+      "learning_rate": 0.0008242574257425743,
+      "loss": 0.6251,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2561822728535703,
+      "grad_norm": 0.02178933471441269,
+      "learning_rate": 0.0008217821782178218,
+      "loss": 0.4698,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2584051125312588,
+      "grad_norm": 0.019534152001142502,
+      "learning_rate": 0.0008193069306930694,
+      "loss": 0.5568,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2606279522089467,
+      "grad_norm": 0.012782914564013481,
+      "learning_rate": 0.0008168316831683169,
+      "loss": 0.6018,
+      "step": 1017
+    },
+    {
+      "epoch": 2.262850791886635,
+      "grad_norm": 0.014194606803357601,
+      "learning_rate": 0.0008143564356435643,
+      "loss": 0.5657,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2650736315643236,
+      "grad_norm": 0.011684983968734741,
+      "learning_rate": 0.0008118811881188118,
+      "loss": 0.572,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2672964712420116,
+      "grad_norm": 0.017416108399629593,
+      "learning_rate": 0.0008094059405940595,
+      "loss": 0.6647,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2695193109197,
+      "grad_norm": 0.021715115755796432,
+      "learning_rate": 0.0008069306930693069,
+      "loss": 0.5813,
+      "step": 1021
+    },
+    {
+      "epoch": 2.271742150597388,
+      "grad_norm": 0.013240072876214981,
+      "learning_rate": 0.0008044554455445544,
+      "loss": 0.5668,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2739649902750765,
+      "grad_norm": 0.020175835117697716,
+      "learning_rate": 0.000801980198019802,
+      "loss": 0.656,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2761878299527645,
+      "grad_norm": 0.017000149935483932,
+      "learning_rate": 0.0007995049504950495,
+      "loss": 0.676,
+      "step": 1024
+    },
+    {
+      "epoch": 2.278410669630453,
+      "grad_norm": 0.016977574676275253,
+      "learning_rate": 0.000797029702970297,
+      "loss": 0.4404,
+      "step": 1025
+    },
+    {
+      "epoch": 2.2806335093081413,
+      "grad_norm": 0.01674072816967964,
+      "learning_rate": 0.0007945544554455446,
+      "loss": 0.7254,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2828563489858293,
+      "grad_norm": 0.01809377409517765,
+      "learning_rate": 0.000792079207920792,
+      "loss": 0.6138,
+      "step": 1027
+    },
+    {
+      "epoch": 2.285079188663518,
+      "grad_norm": 0.01745619997382164,
+      "learning_rate": 0.0007896039603960395,
+      "loss": 0.5348,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2873020283412058,
+      "grad_norm": 0.013998817652463913,
+      "learning_rate": 0.0007871287128712872,
+      "loss": 0.6114,
+      "step": 1029
+    },
+    {
+      "epoch": 2.289524868018894,
+      "grad_norm": 0.017970522865653038,
+      "learning_rate": 0.0007846534653465346,
+      "loss": 0.5384,
+      "step": 1030
+    },
+    {
+      "epoch": 2.291747707696582,
+      "grad_norm": 0.01691269688308239,
+      "learning_rate": 0.0007821782178217821,
+      "loss": 0.6129,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2939705473742706,
+      "grad_norm": 0.015712127089500427,
+      "learning_rate": 0.0007797029702970297,
+      "loss": 0.5979,
+      "step": 1032
+    },
+    {
+      "epoch": 2.296193387051959,
+      "grad_norm": 0.01665470376610756,
+      "learning_rate": 0.0007772277227722774,
+      "loss": 0.6595,
+      "step": 1033
+    },
+    {
+      "epoch": 2.298416226729647,
+      "grad_norm": 0.018899960443377495,
+      "learning_rate": 0.0007747524752475247,
+      "loss": 0.6313,
+      "step": 1034
+    },
+    {
+      "epoch": 2.3006390664073355,
+      "grad_norm": 0.01907293312251568,
+      "learning_rate": 0.0007722772277227723,
+      "loss": 0.6861,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3028619060850235,
+      "grad_norm": 0.01808004081249237,
+      "learning_rate": 0.0007698019801980199,
+      "loss": 0.5455,
+      "step": 1036
+    },
+    {
+      "epoch": 2.305084745762712,
+      "grad_norm": 0.02773713693022728,
+      "learning_rate": 0.0007673267326732672,
+      "loss": 0.6716,
+      "step": 1037
+    },
+    {
+      "epoch": 2.3073075854404,
+      "grad_norm": 0.0136567959561944,
+      "learning_rate": 0.0007648514851485149,
+      "loss": 0.5891,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3095304251180884,
+      "grad_norm": 0.01570400409400463,
+      "learning_rate": 0.0007623762376237625,
+      "loss": 0.648,
+      "step": 1039
+    },
+    {
+      "epoch": 2.311753264795777,
+      "grad_norm": 0.020001349970698357,
+      "learning_rate": 0.0007599009900990098,
+      "loss": 0.8376,
+      "step": 1040
+    },
+    {
+      "epoch": 2.313976104473465,
+      "grad_norm": 0.025027025490999222,
+      "learning_rate": 0.0007574257425742574,
+      "loss": 0.7757,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3161989441511532,
+      "grad_norm": 0.023692984133958817,
+      "learning_rate": 0.0007549504950495051,
+      "loss": 0.6278,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3184217838288412,
+      "grad_norm": 0.015494839288294315,
+      "learning_rate": 0.0007524752475247524,
+      "loss": 0.5303,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3206446235065297,
+      "grad_norm": 0.01560200285166502,
+      "learning_rate": 0.00075,
+      "loss": 0.5107,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3228674631842177,
+      "grad_norm": 0.03477834537625313,
+      "learning_rate": 0.0007475247524752475,
+      "loss": 0.6573,
+      "step": 1045
+    },
+    {
+      "epoch": 2.325090302861906,
+      "grad_norm": 0.030586643144488335,
+      "learning_rate": 0.0007450495049504951,
+      "loss": 0.5893,
+      "step": 1046
+    },
+    {
+      "epoch": 2.3273131425395945,
+      "grad_norm": 0.0189833864569664,
+      "learning_rate": 0.0007425742574257426,
+      "loss": 0.4884,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3295359822172825,
+      "grad_norm": 0.01656891033053398,
+      "learning_rate": 0.0007400990099009901,
+      "loss": 0.545,
+      "step": 1048
+    },
+    {
+      "epoch": 2.331758821894971,
+      "grad_norm": 0.017621248960494995,
+      "learning_rate": 0.0007376237623762376,
+      "loss": 0.5904,
+      "step": 1049
+    },
+    {
+      "epoch": 2.333981661572659,
+      "grad_norm": 0.029514534398913383,
+      "learning_rate": 0.0007351485148514852,
+      "loss": 0.7345,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3362045012503474,
+      "grad_norm": 0.014921414665877819,
+      "learning_rate": 0.0007326732673267328,
+      "loss": 0.5596,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3384273409280354,
+      "grad_norm": 0.020877940580248833,
+      "learning_rate": 0.0007301980198019802,
+      "loss": 0.598,
+      "step": 1052
+    },
+    {
+      "epoch": 2.340650180605724,
+      "grad_norm": 0.01902448944747448,
+      "learning_rate": 0.0007277227722772277,
+      "loss": 0.5555,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3428730202834123,
+      "grad_norm": 0.01561102457344532,
+      "learning_rate": 0.0007252475247524753,
+      "loss": 0.547,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3450958599611003,
+      "grad_norm": 0.0154028395190835,
+      "learning_rate": 0.0007227722772277228,
+      "loss": 0.5978,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3473186996387887,
+      "grad_norm": 0.014076720923185349,
+      "learning_rate": 0.0007202970297029703,
+      "loss": 0.6162,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3495415393164767,
+      "grad_norm": 0.01202862337231636,
+      "learning_rate": 0.0007178217821782179,
+      "loss": 0.5687,
+      "step": 1057
+    },
+    {
+      "epoch": 2.351764378994165,
+      "grad_norm": 0.015466923825442791,
+      "learning_rate": 0.0007153465346534653,
+      "loss": 0.499,
+      "step": 1058
+    },
+    {
+      "epoch": 2.353987218671853,
+      "grad_norm": 0.014686975628137589,
+      "learning_rate": 0.0007128712871287129,
+      "loss": 0.509,
+      "step": 1059
+    },
+    {
+      "epoch": 2.3562100583495416,
+      "grad_norm": 0.019549714401364326,
+      "learning_rate": 0.0007103960396039605,
+      "loss": 0.5659,
+      "step": 1060
+    },
+    {
+      "epoch": 2.35843289802723,
+      "grad_norm": 0.010959336534142494,
+      "learning_rate": 0.0007079207920792079,
+      "loss": 0.4723,
+      "step": 1061
+    },
+    {
+      "epoch": 2.360655737704918,
+      "grad_norm": 0.017420019954442978,
+      "learning_rate": 0.0007054455445544554,
+      "loss": 0.6579,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3628785773826064,
+      "grad_norm": 0.016592038795351982,
+      "learning_rate": 0.000702970297029703,
+      "loss": 0.624,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3651014170602944,
+      "grad_norm": 0.015422000549733639,
+      "learning_rate": 0.0007004950495049505,
+      "loss": 0.6795,
+      "step": 1064
+    },
+    {
+      "epoch": 2.367324256737983,
+      "grad_norm": 0.02857932634651661,
+      "learning_rate": 0.0006980198019801981,
+      "loss": 0.6188,
+      "step": 1065
+    },
+    {
+      "epoch": 2.369547096415671,
+      "grad_norm": 0.020330024883151054,
+      "learning_rate": 0.0006955445544554456,
+      "loss": 0.5202,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3717699360933593,
+      "grad_norm": 0.04005992412567139,
+      "learning_rate": 0.000693069306930693,
+      "loss": 0.5341,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3739927757710477,
+      "grad_norm": 0.014854200184345245,
+      "learning_rate": 0.0006905940594059407,
+      "loss": 0.55,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3762156154487357,
+      "grad_norm": 0.012588760815560818,
+      "learning_rate": 0.0006881188118811882,
+      "loss": 0.5753,
+      "step": 1069
+    },
+    {
+      "epoch": 2.378438455126424,
+      "grad_norm": 0.034118834882974625,
+      "learning_rate": 0.0006856435643564356,
+      "loss": 0.602,
+      "step": 1070
+    },
+    {
+      "epoch": 2.380661294804112,
+      "grad_norm": 0.021048780530691147,
+      "learning_rate": 0.0006831683168316832,
+      "loss": 0.7728,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3828841344818006,
+      "grad_norm": 0.020414873957633972,
+      "learning_rate": 0.0006806930693069306,
+      "loss": 0.6985,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3851069741594886,
+      "grad_norm": 0.012938344851136208,
+      "learning_rate": 0.0006782178217821782,
+      "loss": 0.5337,
+      "step": 1073
+    },
+    {
+      "epoch": 2.387329813837177,
+      "grad_norm": 0.01313922181725502,
+      "learning_rate": 0.0006757425742574258,
+      "loss": 0.4339,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3895526535148655,
+      "grad_norm": 0.01818598248064518,
+      "learning_rate": 0.0006732673267326732,
+      "loss": 0.6091,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3917754931925534,
+      "grad_norm": 0.017337147146463394,
+      "learning_rate": 0.0006707920792079208,
+      "loss": 0.4966,
+      "step": 1076
+    },
+    {
+      "epoch": 2.393998332870242,
+      "grad_norm": 0.017781397327780724,
+      "learning_rate": 0.0006683168316831684,
+      "loss": 0.6172,
+      "step": 1077
+    },
+    {
+      "epoch": 2.39622117254793,
+      "grad_norm": 0.028531668707728386,
+      "learning_rate": 0.0006658415841584158,
+      "loss": 0.6679,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3984440122256183,
+      "grad_norm": 0.023334413766860962,
+      "learning_rate": 0.0006633663366336634,
+      "loss": 0.558,
+      "step": 1079
+    },
+    {
+      "epoch": 2.4006668519033063,
+      "grad_norm": 0.018032846972346306,
+      "learning_rate": 0.0006608910891089109,
+      "loss": 0.5269,
+      "step": 1080
+    },
+    {
+      "epoch": 2.4028896915809947,
+      "grad_norm": 0.015402448363602161,
+      "learning_rate": 0.0006584158415841584,
+      "loss": 0.7979,
+      "step": 1081
+    },
+    {
+      "epoch": 2.405112531258683,
+      "grad_norm": 0.019905133172869682,
+      "learning_rate": 0.000655940594059406,
+      "loss": 0.6255,
+      "step": 1082
+    },
+    {
+      "epoch": 2.407335370936371,
+      "grad_norm": 0.022906949743628502,
+      "learning_rate": 0.0006534653465346535,
+      "loss": 0.5736,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4095582106140596,
+      "grad_norm": 0.014906423166394234,
+      "learning_rate": 0.0006509900990099009,
+      "loss": 0.5573,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4117810502917476,
+      "grad_norm": 0.014077015221118927,
+      "learning_rate": 0.0006485148514851485,
+      "loss": 0.4867,
+      "step": 1085
+    },
+    {
+      "epoch": 2.414003889969436,
+      "grad_norm": 0.02250657044351101,
+      "learning_rate": 0.0006460396039603961,
+      "loss": 0.6064,
+      "step": 1086
+    },
+    {
+      "epoch": 2.416226729647124,
+      "grad_norm": 0.019261633977293968,
+      "learning_rate": 0.0006435643564356435,
+      "loss": 0.6381,
+      "step": 1087
+    },
+    {
+      "epoch": 2.4184495693248125,
+      "grad_norm": 0.019322820007801056,
+      "learning_rate": 0.0006410891089108911,
+      "loss": 0.5413,
+      "step": 1088
+    },
+    {
+      "epoch": 2.420672409002501,
+      "grad_norm": 0.021883083507418633,
+      "learning_rate": 0.0006386138613861386,
+      "loss": 0.6144,
+      "step": 1089
+    },
+    {
+      "epoch": 2.422895248680189,
+      "grad_norm": 0.02103458158671856,
+      "learning_rate": 0.0006361386138613862,
+      "loss": 0.5646,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4251180883578773,
+      "grad_norm": 0.01773696579039097,
+      "learning_rate": 0.0006336633663366337,
+      "loss": 0.5891,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4273409280355653,
+      "grad_norm": 0.024175936356186867,
+      "learning_rate": 0.0006311881188118812,
+      "loss": 0.6319,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4295637677132538,
+      "grad_norm": 0.01638614386320114,
+      "learning_rate": 0.0006287128712871287,
+      "loss": 0.6784,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4317866073909418,
+      "grad_norm": 0.015058539807796478,
+      "learning_rate": 0.0006262376237623763,
+      "loss": 0.5012,
+      "step": 1094
+    },
+    {
+      "epoch": 2.43400944706863,
+      "grad_norm": 0.012811798602342606,
+      "learning_rate": 0.0006237623762376238,
+      "loss": 0.6107,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4362322867463186,
+      "grad_norm": 0.0197319146245718,
+      "learning_rate": 0.0006212871287128713,
+      "loss": 0.5236,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4384551264240066,
+      "grad_norm": 0.018934501335024834,
+      "learning_rate": 0.0006188118811881188,
+      "loss": 0.521,
+      "step": 1097
+    },
+    {
+      "epoch": 2.440677966101695,
+      "grad_norm": 0.014251308515667915,
+      "learning_rate": 0.0006163366336633663,
+      "loss": 0.7587,
+      "step": 1098
+    },
+    {
+      "epoch": 2.442900805779383,
+      "grad_norm": 0.018055707216262817,
+      "learning_rate": 0.0006138613861386139,
+      "loss": 0.5067,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4451236454570715,
+      "grad_norm": 0.018532631918787956,
+      "learning_rate": 0.0006113861386138614,
+      "loss": 0.5683,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4473464851347595,
+      "grad_norm": 0.01650720275938511,
+      "learning_rate": 0.000608910891089109,
+      "loss": 0.7914,
+      "step": 1101
+    },
+    {
+      "epoch": 2.449569324812448,
+      "grad_norm": 0.017233526334166527,
+      "learning_rate": 0.0006064356435643564,
+      "loss": 0.5865,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4517921644901364,
+      "grad_norm": 0.012895757332444191,
+      "learning_rate": 0.000603960396039604,
+      "loss": 0.689,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4540150041678244,
+      "grad_norm": 0.01163373701274395,
+      "learning_rate": 0.0006014851485148516,
+      "loss": 0.5351,
+      "step": 1104
+    },
+    {
+      "epoch": 2.456237843845513,
+      "grad_norm": 0.02952023595571518,
+      "learning_rate": 0.000599009900990099,
+      "loss": 0.6504,
+      "step": 1105
+    },
+    {
+      "epoch": 2.458460683523201,
+      "grad_norm": 0.017913196235895157,
+      "learning_rate": 0.0005965346534653465,
+      "loss": 0.7684,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4606835232008892,
+      "grad_norm": 0.015341251157224178,
+      "learning_rate": 0.0005940594059405942,
+      "loss": 0.6273,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4629063628785772,
+      "grad_norm": 0.02385122701525688,
+      "learning_rate": 0.0005915841584158416,
+      "loss": 0.641,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4651292025562657,
+      "grad_norm": 0.020036252215504646,
+      "learning_rate": 0.0005891089108910891,
+      "loss": 0.5889,
+      "step": 1109
+    },
+    {
+      "epoch": 2.467352042233954,
+      "grad_norm": 0.019238587468862534,
+      "learning_rate": 0.0005866336633663367,
+      "loss": 0.6194,
+      "step": 1110
+    },
+    {
+      "epoch": 2.469574881911642,
+      "grad_norm": 0.01825280860066414,
+      "learning_rate": 0.0005841584158415841,
+      "loss": 0.6515,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4717977215893305,
+      "grad_norm": 0.019387997686862946,
+      "learning_rate": 0.0005816831683168317,
+      "loss": 0.5575,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4740205612670185,
+      "grad_norm": 0.019788188859820366,
+      "learning_rate": 0.0005792079207920793,
+      "loss": 0.5544,
+      "step": 1113
+    },
+    {
+      "epoch": 2.476243400944707,
+      "grad_norm": 0.011799083091318607,
+      "learning_rate": 0.0005767326732673267,
+      "loss": 0.5047,
+      "step": 1114
+    },
+    {
+      "epoch": 2.478466240622395,
+      "grad_norm": 0.01735115796327591,
+      "learning_rate": 0.0005742574257425743,
+      "loss": 0.6194,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4806890803000834,
+      "grad_norm": 0.017334984615445137,
+      "learning_rate": 0.0005717821782178219,
+      "loss": 0.6659,
+      "step": 1116
+    },
+    {
+      "epoch": 2.482911919977772,
+      "grad_norm": 0.013627518899738789,
+      "learning_rate": 0.0005693069306930693,
+      "loss": 0.6804,
+      "step": 1117
+    },
+    {
+      "epoch": 2.48513475965546,
+      "grad_norm": 0.015926288440823555,
+      "learning_rate": 0.0005668316831683169,
+      "loss": 0.552,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4873575993331483,
+      "grad_norm": 0.031809452921152115,
+      "learning_rate": 0.0005643564356435644,
+      "loss": 0.5977,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4895804390108363,
+      "grad_norm": 0.013694511726498604,
+      "learning_rate": 0.0005618811881188118,
+      "loss": 0.4524,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4918032786885247,
+      "grad_norm": 0.02304377593100071,
+      "learning_rate": 0.0005594059405940595,
+      "loss": 0.5375,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4940261183662127,
+      "grad_norm": 0.014395623467862606,
+      "learning_rate": 0.0005569306930693069,
+      "loss": 0.6207,
+      "step": 1122
+    },
+    {
+      "epoch": 2.496248958043901,
+      "grad_norm": 0.019711505621671677,
+      "learning_rate": 0.0005544554455445544,
+      "loss": 0.7473,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4984717977215896,
+      "grad_norm": 0.019002612680196762,
+      "learning_rate": 0.000551980198019802,
+      "loss": 0.5221,
+      "step": 1124
+    },
+    {
+      "epoch": 2.5006946373992776,
+      "grad_norm": 0.02028425969183445,
+      "learning_rate": 0.0005495049504950495,
+      "loss": 0.6702,
+      "step": 1125
+    },
+    {
+      "epoch": 2.5029174770769655,
+      "grad_norm": 0.018800638616085052,
+      "learning_rate": 0.000547029702970297,
+      "loss": 0.566,
+      "step": 1126
+    },
+    {
+      "epoch": 2.505140316754654,
+      "grad_norm": 0.01781066693365574,
+      "learning_rate": 0.0005445544554455446,
+      "loss": 0.5553,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5073631564323424,
+      "grad_norm": 0.02512913942337036,
+      "learning_rate": 0.000542079207920792,
+      "loss": 0.5782,
+      "step": 1128
+    },
+    {
+      "epoch": 2.5095859961100304,
+      "grad_norm": 0.016941504552960396,
+      "learning_rate": 0.0005396039603960396,
+      "loss": 0.4379,
+      "step": 1129
+    },
+    {
+      "epoch": 2.511808835787719,
+      "grad_norm": 0.018688999116420746,
+      "learning_rate": 0.0005371287128712872,
+      "loss": 0.5909,
+      "step": 1130
+    },
+    {
+      "epoch": 2.5140316754654073,
+      "grad_norm": 0.014856253750622272,
+      "learning_rate": 0.0005346534653465346,
+      "loss": 0.6463,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5162545151430953,
+      "grad_norm": 0.01583769917488098,
+      "learning_rate": 0.0005321782178217822,
+      "loss": 0.5773,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5184773548207833,
+      "grad_norm": 0.017723005264997482,
+      "learning_rate": 0.0005297029702970297,
+      "loss": 0.6589,
+      "step": 1133
+    },
+    {
+      "epoch": 2.5207001944984717,
+      "grad_norm": 0.01602606289088726,
+      "learning_rate": 0.0005272277227722772,
+      "loss": 0.5239,
+      "step": 1134
+    },
+    {
+      "epoch": 2.52292303417616,
+      "grad_norm": 0.02923739328980446,
+      "learning_rate": 0.0005247524752475248,
+      "loss": 0.6907,
+      "step": 1135
+    },
+    {
+      "epoch": 2.525145873853848,
+      "grad_norm": 0.018406685441732407,
+      "learning_rate": 0.0005222772277227723,
+      "loss": 0.5979,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5273687135315366,
+      "grad_norm": 0.01883380301296711,
+      "learning_rate": 0.0005198019801980197,
+      "loss": 0.5477,
+      "step": 1137
+    },
+    {
+      "epoch": 2.529591553209225,
+      "grad_norm": 0.021243581548333168,
+      "learning_rate": 0.0005173267326732674,
+      "loss": 0.6081,
+      "step": 1138
+    },
+    {
+      "epoch": 2.531814392886913,
+      "grad_norm": 0.0210957620292902,
+      "learning_rate": 0.0005148514851485149,
+      "loss": 0.6677,
+      "step": 1139
+    },
+    {
+      "epoch": 2.534037232564601,
+      "grad_norm": 0.014948558993637562,
+      "learning_rate": 0.0005123762376237624,
+      "loss": 0.6664,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5362600722422894,
+      "grad_norm": 0.0169658400118351,
+      "learning_rate": 0.0005099009900990099,
+      "loss": 0.5975,
+      "step": 1141
+    },
+    {
+      "epoch": 2.538482911919978,
+      "grad_norm": 0.01910773292183876,
+      "learning_rate": 0.0005074257425742574,
+      "loss": 0.7887,
+      "step": 1142
+    },
+    {
+      "epoch": 2.540705751597666,
+      "grad_norm": 0.022958895191550255,
+      "learning_rate": 0.000504950495049505,
+      "loss": 0.6042,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5429285912753543,
+      "grad_norm": 0.015756452456116676,
+      "learning_rate": 0.0005024752475247525,
+      "loss": 0.5066,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5451514309530427,
+      "grad_norm": 0.023179860785603523,
+      "learning_rate": 0.0005,
+      "loss": 0.7382,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5473742706307307,
+      "grad_norm": 0.018427975475788116,
+      "learning_rate": 0.0004975247524752475,
+      "loss": 0.7563,
+      "step": 1146
+    },
+    {
+      "epoch": 2.5495971103084187,
+      "grad_norm": 0.016231978312134743,
+      "learning_rate": 0.0004950495049504951,
+      "loss": 0.5783,
+      "step": 1147
+    },
+    {
+      "epoch": 2.551819949986107,
+      "grad_norm": 0.014167056418955326,
+      "learning_rate": 0.0004925742574257426,
+      "loss": 0.5216,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5540427896637956,
+      "grad_norm": 0.022291019558906555,
+      "learning_rate": 0.0004900990099009901,
+      "loss": 0.6444,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5562656293414836,
+      "grad_norm": 0.0183990690857172,
+      "learning_rate": 0.0004876237623762376,
+      "loss": 0.5068,
+      "step": 1150
+    },
+    {
+      "epoch": 2.558488469019172,
+      "grad_norm": 0.0216660313308239,
+      "learning_rate": 0.0004851485148514851,
+      "loss": 0.6325,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5607113086968605,
+      "grad_norm": 0.018358584493398666,
+      "learning_rate": 0.0004826732673267327,
+      "loss": 0.6248,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5629341483745485,
+      "grad_norm": 0.018779192119836807,
+      "learning_rate": 0.0004801980198019802,
+      "loss": 0.6257,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5651569880522365,
+      "grad_norm": 0.02526138722896576,
+      "learning_rate": 0.00047772277227722777,
+      "loss": 0.6264,
+      "step": 1154
+    },
+    {
+      "epoch": 2.567379827729925,
+      "grad_norm": 0.016142774373292923,
+      "learning_rate": 0.0004752475247524753,
+      "loss": 0.4821,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5696026674076133,
+      "grad_norm": 0.010809332132339478,
+      "learning_rate": 0.00047277227722772276,
+      "loss": 0.6248,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5718255070853013,
+      "grad_norm": 0.01412492711097002,
+      "learning_rate": 0.00047029702970297034,
+      "loss": 0.5484,
+      "step": 1157
+    },
+    {
+      "epoch": 2.5740483467629898,
+      "grad_norm": 0.01630065217614174,
+      "learning_rate": 0.0004678217821782178,
+      "loss": 0.5974,
+      "step": 1158
+    },
+    {
+      "epoch": 2.576271186440678,
+      "grad_norm": 0.023423103615641594,
+      "learning_rate": 0.0004653465346534653,
+      "loss": 0.5882,
+      "step": 1159
+    },
+    {
+      "epoch": 2.578494026118366,
+      "grad_norm": 0.019574182108044624,
+      "learning_rate": 0.0004628712871287129,
+      "loss": 0.691,
+      "step": 1160
+    },
+    {
+      "epoch": 2.580716865796054,
+      "grad_norm": 0.02245352789759636,
+      "learning_rate": 0.00046039603960396037,
+      "loss": 0.5872,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5829397054737426,
+      "grad_norm": 0.021161986514925957,
+      "learning_rate": 0.0004579207920792079,
+      "loss": 0.6218,
+      "step": 1162
+    },
+    {
+      "epoch": 2.585162545151431,
+      "grad_norm": 0.015376755967736244,
+      "learning_rate": 0.00045544554455445547,
+      "loss": 0.5453,
+      "step": 1163
+    },
+    {
+      "epoch": 2.587385384829119,
+      "grad_norm": 0.020387450233101845,
+      "learning_rate": 0.00045297029702970294,
+      "loss": 0.6943,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5896082245068075,
+      "grad_norm": 0.017321715131402016,
+      "learning_rate": 0.00045049504950495057,
+      "loss": 0.5572,
+      "step": 1165
+    },
+    {
+      "epoch": 2.591831064184496,
+      "grad_norm": 0.013957494869828224,
+      "learning_rate": 0.00044801980198019804,
+      "loss": 0.4917,
+      "step": 1166
+    },
+    {
+      "epoch": 2.594053903862184,
+      "grad_norm": 0.024132177233695984,
+      "learning_rate": 0.0004455445544554455,
+      "loss": 0.5059,
+      "step": 1167
+    },
+    {
+      "epoch": 2.596276743539872,
+      "grad_norm": 0.014031797647476196,
+      "learning_rate": 0.00044306930693069314,
+      "loss": 0.5578,
+      "step": 1168
+    },
+    {
+      "epoch": 2.5984995832175604,
+      "grad_norm": 0.015123395249247551,
+      "learning_rate": 0.0004405940594059406,
+      "loss": 0.6005,
+      "step": 1169
+    },
+    {
+      "epoch": 2.600722422895249,
+      "grad_norm": 0.016194505617022514,
+      "learning_rate": 0.0004381188118811881,
+      "loss": 0.6959,
+      "step": 1170
+    },
+    {
+      "epoch": 2.602945262572937,
+      "grad_norm": 0.015521598979830742,
+      "learning_rate": 0.0004356435643564357,
+      "loss": 0.6823,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6051681022506252,
+      "grad_norm": 0.015657033771276474,
+      "learning_rate": 0.0004331683168316832,
+      "loss": 0.6982,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6073909419283137,
+      "grad_norm": 0.017940111458301544,
+      "learning_rate": 0.00043069306930693064,
+      "loss": 0.6088,
+      "step": 1173
+    },
+    {
+      "epoch": 2.6096137816060017,
+      "grad_norm": 0.023241877555847168,
+      "learning_rate": 0.0004282178217821783,
+      "loss": 0.551,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6118366212836897,
+      "grad_norm": 0.015215943567454815,
+      "learning_rate": 0.00042574257425742574,
+      "loss": 0.6157,
+      "step": 1175
+    },
+    {
+      "epoch": 2.614059460961378,
+      "grad_norm": 0.016419822350144386,
+      "learning_rate": 0.0004232673267326732,
+      "loss": 0.6533,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6162823006390665,
+      "grad_norm": 0.01776900142431259,
+      "learning_rate": 0.00042079207920792084,
+      "loss": 0.7526,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6185051403167545,
+      "grad_norm": 0.04952996224164963,
+      "learning_rate": 0.0004183168316831683,
+      "loss": 0.6754,
+      "step": 1178
+    },
+    {
+      "epoch": 2.620727979994443,
+      "grad_norm": 0.013856332749128342,
+      "learning_rate": 0.0004158415841584159,
+      "loss": 0.5468,
+      "step": 1179
+    },
+    {
+      "epoch": 2.6229508196721314,
+      "grad_norm": 0.018015973269939423,
+      "learning_rate": 0.0004133663366336634,
+      "loss": 0.6287,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6251736593498194,
+      "grad_norm": 0.017965400591492653,
+      "learning_rate": 0.0004108910891089109,
+      "loss": 0.5121,
+      "step": 1181
+    },
+    {
+      "epoch": 2.6273964990275074,
+      "grad_norm": 0.033831071108579636,
+      "learning_rate": 0.00040841584158415845,
+      "loss": 0.5647,
+      "step": 1182
+    },
+    {
+      "epoch": 2.629619338705196,
+      "grad_norm": 0.01705920323729515,
+      "learning_rate": 0.0004059405940594059,
+      "loss": 0.6454,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6318421783828843,
+      "grad_norm": 0.014178541488945484,
+      "learning_rate": 0.00040346534653465344,
+      "loss": 0.5411,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6340650180605722,
+      "grad_norm": 0.018709300085902214,
+      "learning_rate": 0.000400990099009901,
+      "loss": 0.7137,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6362878577382607,
+      "grad_norm": 0.019753027707338333,
+      "learning_rate": 0.0003985148514851485,
+      "loss": 0.6662,
+      "step": 1186
+    },
+    {
+      "epoch": 2.638510697415949,
+      "grad_norm": 0.021427631378173828,
+      "learning_rate": 0.000396039603960396,
+      "loss": 0.557,
+      "step": 1187
+    },
+    {
+      "epoch": 2.640733537093637,
+      "grad_norm": 0.014400881715118885,
+      "learning_rate": 0.0003935643564356436,
+      "loss": 0.5467,
+      "step": 1188
+    },
+    {
+      "epoch": 2.642956376771325,
+      "grad_norm": 0.0270222220569849,
+      "learning_rate": 0.00039108910891089106,
+      "loss": 0.5585,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6451792164490135,
+      "grad_norm": 0.022014157846570015,
+      "learning_rate": 0.0003886138613861387,
+      "loss": 0.433,
+      "step": 1190
+    },
+    {
+      "epoch": 2.647402056126702,
+      "grad_norm": 0.015185764990746975,
+      "learning_rate": 0.00038613861386138616,
+      "loss": 0.6045,
+      "step": 1191
+    },
+    {
+      "epoch": 2.64962489580439,
+      "grad_norm": 0.018142223358154297,
+      "learning_rate": 0.0003836633663366336,
+      "loss": 0.5797,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6518477354820784,
+      "grad_norm": 0.015990694984793663,
+      "learning_rate": 0.00038118811881188126,
+      "loss": 0.6177,
+      "step": 1193
+    },
+    {
+      "epoch": 2.654070575159767,
+      "grad_norm": 0.01589980535209179,
+      "learning_rate": 0.0003787128712871287,
+      "loss": 0.5231,
+      "step": 1194
+    },
+    {
+      "epoch": 2.656293414837455,
+      "grad_norm": 0.020976124331355095,
+      "learning_rate": 0.0003762376237623762,
+      "loss": 0.6988,
+      "step": 1195
+    },
+    {
+      "epoch": 2.658516254515143,
+      "grad_norm": 0.01627793163061142,
+      "learning_rate": 0.00037376237623762377,
+      "loss": 0.6378,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6607390941928313,
+      "grad_norm": 0.02804560214281082,
+      "learning_rate": 0.0003712871287128713,
+      "loss": 0.62,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6629619338705197,
+      "grad_norm": 0.033843785524368286,
+      "learning_rate": 0.0003688118811881188,
+      "loss": 0.6269,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6651847735482077,
+      "grad_norm": 0.0171663835644722,
+      "learning_rate": 0.0003663366336633664,
+      "loss": 0.6803,
+      "step": 1199
+    },
+    {
+      "epoch": 2.667407613225896,
+      "grad_norm": 0.01909332536160946,
+      "learning_rate": 0.00036386138613861386,
+      "loss": 0.5866,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6696304529035846,
+      "grad_norm": 0.014213655143976212,
+      "learning_rate": 0.0003613861386138614,
+      "loss": 0.5731,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6718532925812726,
+      "grad_norm": 0.024629587307572365,
+      "learning_rate": 0.00035891089108910896,
+      "loss": 0.6168,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6740761322589606,
+      "grad_norm": 0.018423261120915413,
+      "learning_rate": 0.0003564356435643564,
+      "loss": 0.566,
+      "step": 1203
+    },
+    {
+      "epoch": 2.676298971936649,
+      "grad_norm": 0.01580965891480446,
+      "learning_rate": 0.00035396039603960395,
+      "loss": 0.5403,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6785218116143374,
+      "grad_norm": 0.01555953361093998,
+      "learning_rate": 0.0003514851485148515,
+      "loss": 0.5308,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6807446512920254,
+      "grad_norm": 0.023963620886206627,
+      "learning_rate": 0.00034900990099009905,
+      "loss": 0.5788,
+      "step": 1206
+    },
+    {
+      "epoch": 2.682967490969714,
+      "grad_norm": 0.015511863864958286,
+      "learning_rate": 0.0003465346534653465,
+      "loss": 0.707,
+      "step": 1207
+    },
+    {
+      "epoch": 2.6851903306474023,
+      "grad_norm": 0.01718342863023281,
+      "learning_rate": 0.0003440594059405941,
+      "loss": 0.5708,
+      "step": 1208
+    },
+    {
+      "epoch": 2.6874131703250903,
+      "grad_norm": 0.027808725833892822,
+      "learning_rate": 0.0003415841584158416,
+      "loss": 0.7655,
+      "step": 1209
+    },
+    {
+      "epoch": 2.6896360100027783,
+      "grad_norm": 0.016733597964048386,
+      "learning_rate": 0.0003391089108910891,
+      "loss": 0.5219,
+      "step": 1210
+    },
+    {
+      "epoch": 2.6918588496804667,
+      "grad_norm": 0.015038656070828438,
+      "learning_rate": 0.0003366336633663366,
+      "loss": 0.6082,
+      "step": 1211
+    },
+    {
+      "epoch": 2.694081689358155,
+      "grad_norm": 0.021239133551716805,
+      "learning_rate": 0.0003341584158415842,
+      "loss": 0.587,
+      "step": 1212
+    },
+    {
+      "epoch": 2.696304529035843,
+      "grad_norm": 0.018834024667739868,
+      "learning_rate": 0.0003316831683168317,
+      "loss": 0.5195,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6985273687135316,
+      "grad_norm": 0.04993009567260742,
+      "learning_rate": 0.0003292079207920792,
+      "loss": 0.6812,
+      "step": 1214
+    },
+    {
+      "epoch": 2.70075020839122,
+      "grad_norm": 0.021933620795607567,
+      "learning_rate": 0.00032673267326732675,
+      "loss": 0.4748,
+      "step": 1215
+    },
+    {
+      "epoch": 2.702973048068908,
+      "grad_norm": 0.014722834341228008,
+      "learning_rate": 0.0003242574257425743,
+      "loss": 0.438,
+      "step": 1216
+    },
+    {
+      "epoch": 2.705195887746596,
+      "grad_norm": 0.013122837990522385,
+      "learning_rate": 0.00032178217821782174,
+      "loss": 0.6231,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7074187274242845,
+      "grad_norm": 0.016432546079158783,
+      "learning_rate": 0.0003193069306930693,
+      "loss": 0.5908,
+      "step": 1218
+    },
+    {
+      "epoch": 2.709641567101973,
+      "grad_norm": 0.016996275633573532,
+      "learning_rate": 0.00031683168316831684,
+      "loss": 0.7369,
+      "step": 1219
+    },
+    {
+      "epoch": 2.711864406779661,
+      "grad_norm": 0.02186731994152069,
+      "learning_rate": 0.00031435643564356436,
+      "loss": 0.6629,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7140872464573493,
+      "grad_norm": 0.016827069222927094,
+      "learning_rate": 0.0003118811881188119,
+      "loss": 0.5866,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7163100861350378,
+      "grad_norm": 0.015328455716371536,
+      "learning_rate": 0.0003094059405940594,
+      "loss": 0.6301,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7185329258127258,
+      "grad_norm": 0.02612834982573986,
+      "learning_rate": 0.00030693069306930693,
+      "loss": 0.4968,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7207557654904138,
+      "grad_norm": 0.018258119001984596,
+      "learning_rate": 0.0003044554455445545,
+      "loss": 0.5323,
+      "step": 1224
+    },
+    {
+      "epoch": 2.722978605168102,
+      "grad_norm": 0.02044299617409706,
+      "learning_rate": 0.000301980198019802,
+      "loss": 0.7331,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7252014448457906,
+      "grad_norm": 0.015368648804724216,
+      "learning_rate": 0.0002995049504950495,
+      "loss": 0.6208,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7274242845234786,
+      "grad_norm": 0.018496476113796234,
+      "learning_rate": 0.0002970297029702971,
+      "loss": 0.5622,
+      "step": 1227
+    },
+    {
+      "epoch": 2.729647124201167,
+      "grad_norm": 0.021723395213484764,
+      "learning_rate": 0.00029455445544554455,
+      "loss": 0.6789,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7318699638788555,
+      "grad_norm": 0.01531304232776165,
+      "learning_rate": 0.00029207920792079207,
+      "loss": 0.6396,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7340928035565435,
+      "grad_norm": 0.02245800755918026,
+      "learning_rate": 0.00028960396039603964,
+      "loss": 0.6289,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7363156432342315,
+      "grad_norm": 0.0517827607691288,
+      "learning_rate": 0.00028712871287128717,
+      "loss": 0.7297,
+      "step": 1231
+    },
+    {
+      "epoch": 2.73853848291192,
+      "grad_norm": 0.014105240814387798,
+      "learning_rate": 0.00028465346534653464,
+      "loss": 0.5948,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7407613225896084,
+      "grad_norm": 0.018385961651802063,
+      "learning_rate": 0.0002821782178217822,
+      "loss": 0.6336,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7429841622672964,
+      "grad_norm": 0.014009015634655952,
+      "learning_rate": 0.00027970297029702973,
+      "loss": 0.5458,
+      "step": 1234
+    },
+    {
+      "epoch": 2.745207001944985,
+      "grad_norm": 0.014900448732078075,
+      "learning_rate": 0.0002772277227722772,
+      "loss": 0.6029,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7474298416226732,
+      "grad_norm": 0.01685742661356926,
+      "learning_rate": 0.0002747524752475247,
+      "loss": 0.5911,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7496526813003612,
+      "grad_norm": 0.01696729101240635,
+      "learning_rate": 0.0002722772277227723,
+      "loss": 0.489,
+      "step": 1237
+    },
+    {
+      "epoch": 2.751875520978049,
+      "grad_norm": 0.01900607720017433,
+      "learning_rate": 0.0002698019801980198,
+      "loss": 0.6149,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7540983606557377,
+      "grad_norm": 0.012541387230157852,
+      "learning_rate": 0.0002673267326732673,
+      "loss": 0.6309,
+      "step": 1239
+    },
+    {
+      "epoch": 2.756321200333426,
+      "grad_norm": 0.018375830724835396,
+      "learning_rate": 0.00026485148514851487,
+      "loss": 0.5464,
+      "step": 1240
+    },
+    {
+      "epoch": 2.758544040011114,
+      "grad_norm": 0.015132361091673374,
+      "learning_rate": 0.0002623762376237624,
+      "loss": 0.6358,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7607668796888025,
+      "grad_norm": 0.0193571038544178,
+      "learning_rate": 0.00025990099009900986,
+      "loss": 0.5601,
+      "step": 1242
+    },
+    {
+      "epoch": 2.762989719366491,
+      "grad_norm": 0.024456117302179337,
+      "learning_rate": 0.00025742574257425744,
+      "loss": 0.7201,
+      "step": 1243
+    },
+    {
+      "epoch": 2.765212559044179,
+      "grad_norm": 0.014225888065993786,
+      "learning_rate": 0.00025495049504950496,
+      "loss": 0.6177,
+      "step": 1244
+    },
+    {
+      "epoch": 2.767435398721867,
+      "grad_norm": 0.0177957434207201,
+      "learning_rate": 0.0002524752475247525,
+      "loss": 0.6382,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7696582383995554,
+      "grad_norm": 0.022806817665696144,
+      "learning_rate": 0.00025,
+      "loss": 0.565,
+      "step": 1246
+    },
+    {
+      "epoch": 2.771881078077244,
+      "grad_norm": 0.016606708988547325,
+      "learning_rate": 0.00024752475247524753,
+      "loss": 0.524,
+      "step": 1247
+    },
+    {
+      "epoch": 2.774103917754932,
+      "grad_norm": 0.013806411065161228,
+      "learning_rate": 0.00024504950495049505,
+      "loss": 0.4843,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7763267574326203,
+      "grad_norm": 0.01461612805724144,
+      "learning_rate": 0.00024257425742574255,
+      "loss": 0.5291,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7785495971103087,
+      "grad_norm": 0.02402338944375515,
+      "learning_rate": 0.0002400990099009901,
+      "loss": 0.5735,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7807724367879967,
+      "grad_norm": 0.01479054894298315,
+      "learning_rate": 0.00023762376237623765,
+      "loss": 0.574,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7829952764656847,
+      "grad_norm": 0.02644219435751438,
+      "learning_rate": 0.00023514851485148517,
+      "loss": 0.5969,
+      "step": 1252
+    },
+    {
+      "epoch": 2.785218116143373,
+      "grad_norm": 0.05949250981211662,
+      "learning_rate": 0.00023267326732673266,
+      "loss": 0.6232,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7874409558210616,
+      "grad_norm": 0.015891224145889282,
+      "learning_rate": 0.00023019801980198019,
+      "loss": 0.5959,
+      "step": 1254
+    },
+    {
+      "epoch": 2.7896637954987495,
+      "grad_norm": 0.012178865261375904,
+      "learning_rate": 0.00022772277227722774,
+      "loss": 0.5764,
+      "step": 1255
+    },
+    {
+      "epoch": 2.791886635176438,
+      "grad_norm": 0.014039144851267338,
+      "learning_rate": 0.00022524752475247528,
+      "loss": 0.6795,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7941094748541264,
+      "grad_norm": 0.012444275431334972,
+      "learning_rate": 0.00022277227722772275,
+      "loss": 0.538,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7963323145318144,
+      "grad_norm": 0.016555974259972572,
+      "learning_rate": 0.0002202970297029703,
+      "loss": 0.5226,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7985551542095024,
+      "grad_norm": 0.02042076364159584,
+      "learning_rate": 0.00021782178217821785,
+      "loss": 0.6182,
+      "step": 1259
+    },
+    {
+      "epoch": 2.800777993887191,
+      "grad_norm": 0.018156493082642555,
+      "learning_rate": 0.00021534653465346532,
+      "loss": 0.6077,
+      "step": 1260
+    },
+    {
+      "epoch": 2.8030008335648793,
+      "grad_norm": 0.015342988073825836,
+      "learning_rate": 0.00021287128712871287,
+      "loss": 0.6097,
+      "step": 1261
+    },
+    {
+      "epoch": 2.8052236732425673,
+      "grad_norm": 0.017595073208212852,
+      "learning_rate": 0.00021039603960396042,
+      "loss": 0.6639,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8074465129202557,
+      "grad_norm": 0.015470032580196857,
+      "learning_rate": 0.00020792079207920794,
+      "loss": 0.6588,
+      "step": 1263
+    },
+    {
+      "epoch": 2.809669352597944,
+      "grad_norm": 0.016241345554590225,
+      "learning_rate": 0.00020544554455445544,
+      "loss": 0.5132,
+      "step": 1264
+    },
+    {
+      "epoch": 2.811892192275632,
+      "grad_norm": 0.024083595722913742,
+      "learning_rate": 0.00020297029702970296,
+      "loss": 0.6011,
+      "step": 1265
+    },
+    {
+      "epoch": 2.81411503195332,
+      "grad_norm": 0.013913623988628387,
+      "learning_rate": 0.0002004950495049505,
+      "loss": 0.5177,
+      "step": 1266
+    },
+    {
+      "epoch": 2.8163378716310086,
+      "grad_norm": 0.020045112818479538,
+      "learning_rate": 0.000198019801980198,
+      "loss": 0.6554,
+      "step": 1267
+    },
+    {
+      "epoch": 2.818560711308697,
+      "grad_norm": 0.02670402266085148,
+      "learning_rate": 0.00019554455445544553,
+      "loss": 0.5696,
+      "step": 1268
+    },
+    {
+      "epoch": 2.820783550986385,
+      "grad_norm": 0.015266619622707367,
+      "learning_rate": 0.00019306930693069308,
+      "loss": 0.5624,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8230063906640734,
+      "grad_norm": 0.019668592140078545,
+      "learning_rate": 0.00019059405940594063,
+      "loss": 0.568,
+      "step": 1270
+    },
+    {
+      "epoch": 2.825229230341762,
+      "grad_norm": 0.04979453608393669,
+      "learning_rate": 0.0001881188118811881,
+      "loss": 0.5098,
+      "step": 1271
+    },
+    {
+      "epoch": 2.82745207001945,
+      "grad_norm": 0.028548695147037506,
+      "learning_rate": 0.00018564356435643565,
+      "loss": 0.6911,
+      "step": 1272
+    },
+    {
+      "epoch": 2.829674909697138,
+      "grad_norm": 0.02356465719640255,
+      "learning_rate": 0.0001831683168316832,
+      "loss": 0.5654,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8318977493748263,
+      "grad_norm": 0.021715005859732628,
+      "learning_rate": 0.0001806930693069307,
+      "loss": 0.4995,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8341205890525147,
+      "grad_norm": 0.014024381525814533,
+      "learning_rate": 0.0001782178217821782,
+      "loss": 0.6592,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8363434287302027,
+      "grad_norm": 0.020154127851128578,
+      "learning_rate": 0.00017574257425742576,
+      "loss": 0.6721,
+      "step": 1276
+    },
+    {
+      "epoch": 2.838566268407891,
+      "grad_norm": 0.0298357754945755,
+      "learning_rate": 0.00017326732673267326,
+      "loss": 0.6803,
+      "step": 1277
+    },
+    {
+      "epoch": 2.840789108085579,
+      "grad_norm": 0.026704518124461174,
+      "learning_rate": 0.0001707920792079208,
+      "loss": 0.7333,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8430119477632676,
+      "grad_norm": 0.017025306820869446,
+      "learning_rate": 0.0001683168316831683,
+      "loss": 0.6408,
+      "step": 1279
+    },
+    {
+      "epoch": 2.8452347874409556,
+      "grad_norm": 0.014133095741271973,
+      "learning_rate": 0.00016584158415841585,
+      "loss": 0.6176,
+      "step": 1280
+    },
+    {
+      "epoch": 2.847457627118644,
+      "grad_norm": 0.019554702565073967,
+      "learning_rate": 0.00016336633663366338,
+      "loss": 0.4898,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8496804667963325,
+      "grad_norm": 0.027809258550405502,
+      "learning_rate": 0.00016089108910891087,
+      "loss": 0.5767,
+      "step": 1282
+    },
+    {
+      "epoch": 2.8519033064740205,
+      "grad_norm": 0.016597462818026543,
+      "learning_rate": 0.00015841584158415842,
+      "loss": 0.4662,
+      "step": 1283
+    },
+    {
+      "epoch": 2.854126146151709,
+      "grad_norm": 0.04323901981115341,
+      "learning_rate": 0.00015594059405940594,
+      "loss": 0.6596,
+      "step": 1284
+    },
+    {
+      "epoch": 2.856348985829397,
+      "grad_norm": 0.020414169877767563,
+      "learning_rate": 0.00015346534653465347,
+      "loss": 0.601,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8585718255070853,
+      "grad_norm": 0.020976142957806587,
+      "learning_rate": 0.000150990099009901,
+      "loss": 0.6277,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8607946651847733,
+      "grad_norm": 0.016226671636104584,
+      "learning_rate": 0.00014851485148514854,
+      "loss": 0.6328,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8630175048624618,
+      "grad_norm": 0.022331437095999718,
+      "learning_rate": 0.00014603960396039603,
+      "loss": 0.5317,
+      "step": 1288
+    },
+    {
+      "epoch": 2.86524034454015,
+      "grad_norm": 0.028064247220754623,
+      "learning_rate": 0.00014356435643564358,
+      "loss": 0.5249,
+      "step": 1289
+    },
+    {
+      "epoch": 2.867463184217838,
+      "grad_norm": 0.011397906579077244,
+      "learning_rate": 0.0001410891089108911,
+      "loss": 0.5118,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8696860238955266,
+      "grad_norm": 0.015670351684093475,
+      "learning_rate": 0.0001386138613861386,
+      "loss": 0.5159,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8719088635732146,
+      "grad_norm": 0.015906330198049545,
+      "learning_rate": 0.00013613861386138615,
+      "loss": 0.5301,
+      "step": 1292
+    },
+    {
+      "epoch": 2.874131703250903,
+      "grad_norm": 0.014323754236102104,
+      "learning_rate": 0.00013366336633663365,
+      "loss": 0.6323,
+      "step": 1293
+    },
+    {
+      "epoch": 2.876354542928591,
+      "grad_norm": 0.017434075474739075,
+      "learning_rate": 0.0001311881188118812,
+      "loss": 0.7477,
+      "step": 1294
+    },
+    {
+      "epoch": 2.8785773826062795,
+      "grad_norm": 0.013494400307536125,
+      "learning_rate": 0.00012871287128712872,
+      "loss": 0.5316,
+      "step": 1295
+    },
+    {
+      "epoch": 2.880800222283968,
+      "grad_norm": 0.018436387181282043,
+      "learning_rate": 0.00012623762376237624,
+      "loss": 0.6895,
+      "step": 1296
+    },
+    {
+      "epoch": 2.883023061961656,
+      "grad_norm": 0.014047752134501934,
+      "learning_rate": 0.00012376237623762376,
+      "loss": 0.5798,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8852459016393444,
+      "grad_norm": 0.014521469362080097,
+      "learning_rate": 0.00012128712871287127,
+      "loss": 0.7352,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8874687413170324,
+      "grad_norm": 0.018711743876338005,
+      "learning_rate": 0.00011881188118811882,
+      "loss": 0.4314,
+      "step": 1299
+    },
+    {
+      "epoch": 2.889691580994721,
+      "grad_norm": 0.017838047817349434,
+      "learning_rate": 0.00011633663366336633,
+      "loss": 0.5504,
+      "step": 1300
+    },
+    {
+      "epoch": 2.891914420672409,
+      "grad_norm": 0.023302067071199417,
+      "learning_rate": 0.00011386138613861387,
+      "loss": 0.5713,
+      "step": 1301
+    },
+    {
+      "epoch": 2.894137260350097,
+      "grad_norm": 0.015559905208647251,
+      "learning_rate": 0.00011138613861386138,
+      "loss": 0.6113,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8963601000277857,
+      "grad_norm": 0.023848090320825577,
+      "learning_rate": 0.00010891089108910893,
+      "loss": 0.7445,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8985829397054736,
+      "grad_norm": 0.014597102999687195,
+      "learning_rate": 0.00010643564356435644,
+      "loss": 0.5805,
+      "step": 1304
+    },
+    {
+      "epoch": 2.900805779383162,
+      "grad_norm": 0.01882697455585003,
+      "learning_rate": 0.00010396039603960397,
+      "loss": 0.6368,
+      "step": 1305
+    },
+    {
+      "epoch": 2.90302861906085,
+      "grad_norm": 0.013598893769085407,
+      "learning_rate": 0.00010148514851485148,
+      "loss": 0.5894,
+      "step": 1306
+    },
+    {
+      "epoch": 2.9052514587385385,
+      "grad_norm": 0.019652416929602623,
+      "learning_rate": 9.9009900990099e-05,
+      "loss": 0.6455,
+      "step": 1307
+    },
+    {
+      "epoch": 2.9074742984162265,
+      "grad_norm": 0.014602046459913254,
+      "learning_rate": 9.653465346534654e-05,
+      "loss": 0.5992,
+      "step": 1308
+    },
+    {
+      "epoch": 2.909697138093915,
+      "grad_norm": 0.012488738633692265,
+      "learning_rate": 9.405940594059405e-05,
+      "loss": 0.5696,
+      "step": 1309
+    },
+    {
+      "epoch": 2.9119199777716034,
+      "grad_norm": 0.018612246960401535,
+      "learning_rate": 9.15841584158416e-05,
+      "loss": 0.5856,
+      "step": 1310
+    },
+    {
+      "epoch": 2.9141428174492914,
+      "grad_norm": 0.018312592059373856,
+      "learning_rate": 8.91089108910891e-05,
+      "loss": 0.6731,
+      "step": 1311
+    },
+    {
+      "epoch": 2.91636565712698,
+      "grad_norm": 0.011301212944090366,
+      "learning_rate": 8.663366336633663e-05,
+      "loss": 0.6148,
+      "step": 1312
+    },
+    {
+      "epoch": 2.918588496804668,
+      "grad_norm": 0.014970047399401665,
+      "learning_rate": 8.415841584158415e-05,
+      "loss": 0.5364,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9208113364823562,
+      "grad_norm": 0.017927002161741257,
+      "learning_rate": 8.168316831683169e-05,
+      "loss": 0.7138,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9230341761600442,
+      "grad_norm": 0.012390673160552979,
+      "learning_rate": 7.920792079207921e-05,
+      "loss": 0.5939,
+      "step": 1315
+    },
+    {
+      "epoch": 2.9252570158377327,
+      "grad_norm": 0.021033355966210365,
+      "learning_rate": 7.673267326732673e-05,
+      "loss": 0.5774,
+      "step": 1316
+    },
+    {
+      "epoch": 2.927479855515421,
+      "grad_norm": 0.019287073984742165,
+      "learning_rate": 7.425742574257427e-05,
+      "loss": 0.6687,
+      "step": 1317
+    },
+    {
+      "epoch": 2.929702695193109,
+      "grad_norm": 0.01410654280334711,
+      "learning_rate": 7.178217821782179e-05,
+      "loss": 0.5655,
+      "step": 1318
+    },
+    {
+      "epoch": 2.9319255348707975,
+      "grad_norm": 0.015142921358346939,
+      "learning_rate": 6.93069306930693e-05,
+      "loss": 0.6458,
+      "step": 1319
+    },
+    {
+      "epoch": 2.9341483745484855,
+      "grad_norm": 0.014644642360508442,
+      "learning_rate": 6.683168316831682e-05,
+      "loss": 0.5915,
+      "step": 1320
+    },
+    {
+      "epoch": 2.936371214226174,
+      "grad_norm": 0.018795736134052277,
+      "learning_rate": 6.435643564356436e-05,
+      "loss": 0.5382,
+      "step": 1321
+    },
+    {
+      "epoch": 2.938594053903862,
+      "grad_norm": 0.01939059980213642,
+      "learning_rate": 6.188118811881188e-05,
+      "loss": 0.5383,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9408168935815504,
+      "grad_norm": 0.016617722809314728,
+      "learning_rate": 5.940594059405941e-05,
+      "loss": 0.5883,
+      "step": 1323
+    },
+    {
+      "epoch": 2.943039733259239,
+      "grad_norm": 0.01931897923350334,
+      "learning_rate": 5.6930693069306934e-05,
+      "loss": 0.5775,
+      "step": 1324
+    },
+    {
+      "epoch": 2.945262572936927,
+      "grad_norm": 0.018571315333247185,
+      "learning_rate": 5.445544554455446e-05,
+      "loss": 0.6164,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9474854126146153,
+      "grad_norm": 0.018728012219071388,
+      "learning_rate": 5.1980198019801986e-05,
+      "loss": 0.6216,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9497082522923033,
+      "grad_norm": 0.01738075353205204,
+      "learning_rate": 4.95049504950495e-05,
+      "loss": 0.6433,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9519310919699917,
+      "grad_norm": 0.012660011649131775,
+      "learning_rate": 4.7029702970297024e-05,
+      "loss": 0.6901,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9541539316476797,
+      "grad_norm": 0.01377065945416689,
+      "learning_rate": 4.455445544554455e-05,
+      "loss": 0.5237,
+      "step": 1329
+    },
+    {
+      "epoch": 2.956376771325368,
+      "grad_norm": 0.030198214575648308,
+      "learning_rate": 4.2079207920792076e-05,
+      "loss": 0.6441,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9585996110030566,
+      "grad_norm": 0.01597333326935768,
+      "learning_rate": 3.9603960396039605e-05,
+      "loss": 0.6057,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9608224506807446,
+      "grad_norm": 0.021931879222393036,
+      "learning_rate": 3.7128712871287135e-05,
+      "loss": 0.5539,
+      "step": 1332
+    },
+    {
+      "epoch": 2.963045290358433,
+      "grad_norm": 0.01602141186594963,
+      "learning_rate": 3.465346534653465e-05,
+      "loss": 0.6342,
+      "step": 1333
+    },
+    {
+      "epoch": 2.965268130036121,
+      "grad_norm": 0.026012420654296875,
+      "learning_rate": 3.217821782178218e-05,
+      "loss": 0.7019,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9674909697138094,
+      "grad_norm": 0.016496378928422928,
+      "learning_rate": 2.9702970297029706e-05,
+      "loss": 0.586,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9697138093914974,
+      "grad_norm": 0.014965620823204517,
+      "learning_rate": 2.722772277227723e-05,
+      "loss": 0.5953,
+      "step": 1336
+    },
+    {
+      "epoch": 2.971936649069186,
+      "grad_norm": 0.016745861619710922,
+      "learning_rate": 2.475247524752475e-05,
+      "loss": 0.5842,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9741594887468743,
+      "grad_norm": 0.021782133728265762,
+      "learning_rate": 2.2277227722772277e-05,
+      "loss": 0.668,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9763823284245623,
+      "grad_norm": 0.017215097323060036,
+      "learning_rate": 1.9801980198019803e-05,
+      "loss": 0.6326,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9786051681022507,
+      "grad_norm": 0.012050891295075417,
+      "learning_rate": 1.7326732673267325e-05,
+      "loss": 0.6311,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9808280077799387,
+      "grad_norm": 0.01476576179265976,
+      "learning_rate": 1.4851485148514853e-05,
+      "loss": 0.6109,
+      "step": 1341
+    },
+    {
+      "epoch": 2.983050847457627,
+      "grad_norm": 0.015637947246432304,
+      "learning_rate": 1.2376237623762375e-05,
+      "loss": 0.6767,
+      "step": 1342
+    },
+    {
+      "epoch": 2.985273687135315,
+      "grad_norm": 0.01803119294345379,
+      "learning_rate": 9.900990099009901e-06,
+      "loss": 0.587,
+      "step": 1343
+    },
+    {
+      "epoch": 2.9874965268130036,
+      "grad_norm": 0.021125512197613716,
+      "learning_rate": 7.425742574257426e-06,
+      "loss": 0.8802,
+      "step": 1344
+    },
+    {
+      "epoch": 2.989719366490692,
+      "grad_norm": 0.024220464751124382,
+      "learning_rate": 4.950495049504951e-06,
+      "loss": 0.7215,
+      "step": 1345
+    },
+    {
+      "epoch": 2.99194220616838,
+      "grad_norm": 0.022379010915756226,
+      "learning_rate": 2.4752475247524753e-06,
+      "loss": 0.6175,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9941650458460685,
+      "grad_norm": 0.0142740523442626,
+      "learning_rate": 0.0,
+      "loss": 0.5943,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9941650458460685,
+      "step": 1347,
+      "total_flos": 5.0749830055591936e+17,
+      "train_loss": 0.6960906616253771,
+      "train_runtime": 3387.1352,
+      "train_samples_per_second": 6.374,
+      "train_steps_per_second": 0.398
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1347,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.0749830055591936e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}