diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4957 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999288610656613,
+  "eval_steps": 500,
+  "global_step": 7028,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001422778686775272,
+      "grad_norm": 227.20426891789228,
+      "learning_rate": 2.132701421800948e-07,
+      "loss": 14.582,
+      "step": 10
+    },
+    {
+      "epoch": 0.002845557373550544,
+      "grad_norm": 288.4752938926377,
+      "learning_rate": 4.502369668246446e-07,
+      "loss": 14.1726,
+      "step": 20
+    },
+    {
+      "epoch": 0.004268336060325817,
+      "grad_norm": 182.76059043427875,
+      "learning_rate": 6.872037914691944e-07,
+      "loss": 13.4078,
+      "step": 30
+    },
+    {
+      "epoch": 0.005691114747101088,
+      "grad_norm": 127.33519082588683,
+      "learning_rate": 9.241706161137441e-07,
+      "loss": 11.4628,
+      "step": 40
+    },
+    {
+      "epoch": 0.007113893433876361,
+      "grad_norm": 155.10626329925768,
+      "learning_rate": 1.161137440758294e-06,
+      "loss": 9.1475,
+      "step": 50
+    },
+    {
+      "epoch": 0.008536672120651633,
+      "grad_norm": 138.90229390796665,
+      "learning_rate": 1.3981042654028437e-06,
+      "loss": 7.7319,
+      "step": 60
+    },
+    {
+      "epoch": 0.009959450807426905,
+      "grad_norm": 109.7463772232744,
+      "learning_rate": 1.6350710900473934e-06,
+      "loss": 6.4941,
+      "step": 70
+    },
+    {
+      "epoch": 0.011382229494202176,
+      "grad_norm": 83.21567835995891,
+      "learning_rate": 1.8720379146919433e-06,
+      "loss": 6.1299,
+      "step": 80
+    },
+    {
+      "epoch": 0.01280500818097745,
+      "grad_norm": 276.7319598681411,
+      "learning_rate": 2.109004739336493e-06,
+      "loss": 5.8824,
+      "step": 90
+    },
+    {
+      "epoch": 0.014227786867752721,
+      "grad_norm": 150.10206828081996,
+      "learning_rate": 2.345971563981043e-06,
+      "loss": 5.4126,
+      "step": 100
+    },
+    {
+      "epoch": 0.015650565554527995,
+      "grad_norm": 85.51637256924744,
+      "learning_rate": 2.5829383886255925e-06,
+      "loss": 5.4679,
+      "step": 110
+    },
+    {
+      "epoch": 0.017073344241303266,
+      "grad_norm": 54.01631684738819,
+      "learning_rate": 2.8199052132701426e-06,
+      "loss": 5.2629,
+      "step": 120
+    },
+    {
+      "epoch": 0.018496122928078538,
+      "grad_norm": 48.046263311608506,
+      "learning_rate": 3.0568720379146923e-06,
+      "loss": 5.1994,
+      "step": 130
+    },
+    {
+      "epoch": 0.01991890161485381,
+      "grad_norm": 48.29009312747582,
+      "learning_rate": 3.293838862559242e-06,
+      "loss": 4.9431,
+      "step": 140
+    },
+    {
+      "epoch": 0.02134168030162908,
+      "grad_norm": 68.96565083843548,
+      "learning_rate": 3.5308056872037916e-06,
+      "loss": 4.9651,
+      "step": 150
+    },
+    {
+      "epoch": 0.022764458988404353,
+      "grad_norm": 57.886348280789576,
+      "learning_rate": 3.7677725118483417e-06,
+      "loss": 4.8191,
+      "step": 160
+    },
+    {
+      "epoch": 0.024187237675179624,
+      "grad_norm": 33.13905030540343,
+      "learning_rate": 4.004739336492891e-06,
+      "loss": 4.601,
+      "step": 170
+    },
+    {
+      "epoch": 0.0256100163619549,
+      "grad_norm": 36.383587300681455,
+      "learning_rate": 4.2417061611374415e-06,
+      "loss": 4.6011,
+      "step": 180
+    },
+    {
+      "epoch": 0.02703279504873017,
+      "grad_norm": 93.93751717462193,
+      "learning_rate": 4.478672985781991e-06,
+      "loss": 4.5838,
+      "step": 190
+    },
+    {
+      "epoch": 0.028455573735505443,
+      "grad_norm": 161.18260012854552,
+      "learning_rate": 4.715639810426541e-06,
+      "loss": 4.473,
+      "step": 200
+    },
+    {
+      "epoch": 0.029878352422280714,
+      "grad_norm": 48.03951820879001,
+      "learning_rate": 4.952606635071091e-06,
+      "loss": 4.328,
+      "step": 210
+    },
+    {
+      "epoch": 0.03130113110905599,
+      "grad_norm": 48.351614233644206,
+      "learning_rate": 4.999983009637205e-06,
+      "loss": 3.8117,
+      "step": 220
+    },
+    {
+      "epoch": 0.03272390979583126,
+      "grad_norm": 40.879446824882315,
+      "learning_rate": 4.999913986684146e-06,
+      "loss": 3.666,
+      "step": 230
+    },
+    {
+      "epoch": 0.03414668848260653,
+      "grad_norm": 71.68709145083035,
+      "learning_rate": 4.999791870707925e-06,
+      "loss": 3.3076,
+      "step": 240
+    },
+    {
+      "epoch": 0.0355694671693818,
+      "grad_norm": 41.1760727865846,
+      "learning_rate": 4.999616664302036e-06,
+      "loss": 3.0586,
+      "step": 250
+    },
+    {
+      "epoch": 0.036992245856157076,
+      "grad_norm": 35.99541853661167,
+      "learning_rate": 4.9993883711875e-06,
+      "loss": 2.6416,
+      "step": 260
+    },
+    {
+      "epoch": 0.038415024542932344,
+      "grad_norm": 30.668465660210362,
+      "learning_rate": 4.999106996212795e-06,
+      "loss": 2.4984,
+      "step": 270
+    },
+    {
+      "epoch": 0.03983780322970762,
+      "grad_norm": 36.932680181660984,
+      "learning_rate": 4.998772545353743e-06,
+      "loss": 2.3382,
+      "step": 280
+    },
+    {
+      "epoch": 0.041260581916482894,
+      "grad_norm": 61.90188051017363,
+      "learning_rate": 4.998385025713392e-06,
+      "loss": 2.0779,
+      "step": 290
+    },
+    {
+      "epoch": 0.04268336060325816,
+      "grad_norm": 69.48759279430097,
+      "learning_rate": 4.9979444455218615e-06,
+      "loss": 2.028,
+      "step": 300
+    },
+    {
+      "epoch": 0.04410613929003344,
+      "grad_norm": 24.129222083009303,
+      "learning_rate": 4.997450814136168e-06,
+      "loss": 2.0022,
+      "step": 310
+    },
+    {
+      "epoch": 0.045528917976808705,
+      "grad_norm": 57.107184624594545,
+      "learning_rate": 4.9969041420400235e-06,
+      "loss": 1.7308,
+      "step": 320
+    },
+    {
+      "epoch": 0.04695169666358398,
+      "grad_norm": 25.286429840125543,
+      "learning_rate": 4.996304440843618e-06,
+      "loss": 1.8255,
+      "step": 330
+    },
+    {
+      "epoch": 0.04837447535035925,
+      "grad_norm": 41.33604492294289,
+      "learning_rate": 4.9956517232833696e-06,
+      "loss": 1.6635,
+      "step": 340
+    },
+    {
+      "epoch": 0.049797254037134524,
+      "grad_norm": 25.057988737297727,
+      "learning_rate": 4.994946003221654e-06,
+      "loss": 1.6383,
+      "step": 350
+    },
+    {
+      "epoch": 0.0512200327239098,
+      "grad_norm": 29.141194168593408,
+      "learning_rate": 4.994187295646512e-06,
+      "loss": 1.6471,
+      "step": 360
+    },
+    {
+      "epoch": 0.05264281141068507,
+      "grad_norm": 32.28184288482766,
+      "learning_rate": 4.9933756166713295e-06,
+      "loss": 1.5427,
+      "step": 370
+    },
+    {
+      "epoch": 0.05406559009746034,
+      "grad_norm": 21.403858874487227,
+      "learning_rate": 4.992510983534496e-06,
+      "loss": 1.4988,
+      "step": 380
+    },
+    {
+      "epoch": 0.05548836878423561,
+      "grad_norm": 33.94744437090256,
+      "learning_rate": 4.9915934145990385e-06,
+      "loss": 1.4508,
+      "step": 390
+    },
+    {
+      "epoch": 0.056911147471010885,
+      "grad_norm": 19.822673933241806,
+      "learning_rate": 4.99062292935223e-06,
+      "loss": 1.4209,
+      "step": 400
+    },
+    {
+      "epoch": 0.05833392615778615,
+      "grad_norm": 41.542065289192145,
+      "learning_rate": 4.989599548405178e-06,
+      "loss": 1.5373,
+      "step": 410
+    },
+    {
+      "epoch": 0.05975670484456143,
+      "grad_norm": 46.3095442295298,
+      "learning_rate": 4.9885232934923865e-06,
+      "loss": 1.3934,
+      "step": 420
+    },
+    {
+      "epoch": 0.061179483531336704,
+      "grad_norm": 25.997158063369152,
+      "learning_rate": 4.987394187471292e-06,
+      "loss": 1.4154,
+      "step": 430
+    },
+    {
+      "epoch": 0.06260226221811198,
+      "grad_norm": 46.76059720724752,
+      "learning_rate": 4.98621225432178e-06,
+      "loss": 1.2577,
+      "step": 440
+    },
+    {
+      "epoch": 0.06402504090488724,
+      "grad_norm": 39.72838305337924,
+      "learning_rate": 4.984977519145677e-06,
+      "loss": 1.3735,
+      "step": 450
+    },
+    {
+      "epoch": 0.06544781959166251,
+      "grad_norm": 31.8000552139225,
+      "learning_rate": 4.983690008166213e-06,
+      "loss": 1.378,
+      "step": 460
+    },
+    {
+      "epoch": 0.06687059827843779,
+      "grad_norm": 54.99100856311778,
+      "learning_rate": 4.982349748727471e-06,
+      "loss": 1.2952,
+      "step": 470
+    },
+    {
+      "epoch": 0.06829337696521307,
+      "grad_norm": 34.7813184333919,
+      "learning_rate": 4.980956769293799e-06,
+      "loss": 1.3304,
+      "step": 480
+    },
+    {
+      "epoch": 0.06971615565198833,
+      "grad_norm": 56.07368019486356,
+      "learning_rate": 4.9795110994492104e-06,
+      "loss": 1.2965,
+      "step": 490
+    },
+    {
+      "epoch": 0.0711389343387636,
+      "grad_norm": 24.812325854128844,
+      "learning_rate": 4.978012769896754e-06,
+      "loss": 1.2318,
+      "step": 500
+    },
+    {
+      "epoch": 0.07256171302553888,
+      "grad_norm": 16.4693703560664,
+      "learning_rate": 4.976461812457866e-06,
+      "loss": 1.3849,
+      "step": 510
+    },
+    {
+      "epoch": 0.07398449171231415,
+      "grad_norm": 17.511580201601642,
+      "learning_rate": 4.974858260071684e-06,
+      "loss": 1.2936,
+      "step": 520
+    },
+    {
+      "epoch": 0.07540727039908943,
+      "grad_norm": 18.829140041990236,
+      "learning_rate": 4.973202146794358e-06,
+      "loss": 1.2266,
+      "step": 530
+    },
+    {
+      "epoch": 0.07683004908586469,
+      "grad_norm": 30.404720510571355,
+      "learning_rate": 4.971493507798324e-06,
+      "loss": 1.2773,
+      "step": 540
+    },
+    {
+      "epoch": 0.07825282777263996,
+      "grad_norm": 14.60687195849512,
+      "learning_rate": 4.9697323793715526e-06,
+      "loss": 1.25,
+      "step": 550
+    },
+    {
+      "epoch": 0.07967560645941524,
+      "grad_norm": 71.6497384118756,
+      "learning_rate": 4.967918798916785e-06,
+      "loss": 1.3495,
+      "step": 560
+    },
+    {
+      "epoch": 0.08109838514619051,
+      "grad_norm": 16.64365379567024,
+      "learning_rate": 4.966052804950733e-06,
+      "loss": 1.1047,
+      "step": 570
+    },
+    {
+      "epoch": 0.08252116383296579,
+      "grad_norm": 33.87717811951715,
+      "learning_rate": 4.964134437103265e-06,
+      "loss": 1.2592,
+      "step": 580
+    },
+    {
+      "epoch": 0.08394394251974105,
+      "grad_norm": 70.63821742601606,
+      "learning_rate": 4.962163736116562e-06,
+      "loss": 1.1634,
+      "step": 590
+    },
+    {
+      "epoch": 0.08536672120651632,
+      "grad_norm": 32.75072239557704,
+      "learning_rate": 4.960140743844252e-06,
+      "loss": 1.2711,
+      "step": 600
+    },
+    {
+      "epoch": 0.0867894998932916,
+      "grad_norm": 35.5844324891052,
+      "learning_rate": 4.958065503250526e-06,
+      "loss": 1.2501,
+      "step": 610
+    },
+    {
+      "epoch": 0.08821227858006687,
+      "grad_norm": 14.266674138271055,
+      "learning_rate": 4.955938058409217e-06,
+      "loss": 1.1745,
+      "step": 620
+    },
+    {
+      "epoch": 0.08963505726684214,
+      "grad_norm": 18.873108069596782,
+      "learning_rate": 4.953758454502872e-06,
+      "loss": 1.1528,
+      "step": 630
+    },
+    {
+      "epoch": 0.09105783595361741,
+      "grad_norm": 31.56236779309872,
+      "learning_rate": 4.951526737821787e-06,
+      "loss": 1.2088,
+      "step": 640
+    },
+    {
+      "epoch": 0.09248061464039269,
+      "grad_norm": 18.70989762385317,
+      "learning_rate": 4.949242955763028e-06,
+      "loss": 1.211,
+      "step": 650
+    },
+    {
+      "epoch": 0.09390339332716796,
+      "grad_norm": 39.25798153988374,
+      "learning_rate": 4.94690715682942e-06,
+      "loss": 1.1713,
+      "step": 660
+    },
+    {
+      "epoch": 0.09532617201394324,
+      "grad_norm": 22.778346344296285,
+      "learning_rate": 4.9445193906285206e-06,
+      "loss": 1.2458,
+      "step": 670
+    },
+    {
+      "epoch": 0.0967489507007185,
+      "grad_norm": 23.69275420448688,
+      "learning_rate": 4.942079707871566e-06,
+      "loss": 1.1972,
+      "step": 680
+    },
+    {
+      "epoch": 0.09817172938749377,
+      "grad_norm": 39.306258513979415,
+      "learning_rate": 4.939588160372392e-06,
+      "loss": 1.1222,
+      "step": 690
+    },
+    {
+      "epoch": 0.09959450807426905,
+      "grad_norm": 22.896068053456,
+      "learning_rate": 4.937044801046334e-06,
+      "loss": 1.1845,
+      "step": 700
+    },
+    {
+      "epoch": 0.10101728676104432,
+      "grad_norm": 60.82968098497559,
+      "learning_rate": 4.934449683909106e-06,
+      "loss": 1.2123,
+      "step": 710
+    },
+    {
+      "epoch": 0.1024400654478196,
+      "grad_norm": 56.56768516919535,
+      "learning_rate": 4.931802864075648e-06,
+      "loss": 0.9821,
+      "step": 720
+    },
+    {
+      "epoch": 0.10386284413459486,
+      "grad_norm": 24.624330732318725,
+      "learning_rate": 4.929104397758962e-06,
+      "loss": 1.2275,
+      "step": 730
+    },
+    {
+      "epoch": 0.10528562282137013,
+      "grad_norm": 38.57583464398156,
+      "learning_rate": 4.926354342268913e-06,
+      "loss": 1.0655,
+      "step": 740
+    },
+    {
+      "epoch": 0.10670840150814541,
+      "grad_norm": 22.07451047416335,
+      "learning_rate": 4.9235527560110156e-06,
+      "loss": 1.0444,
+      "step": 750
+    },
+    {
+      "epoch": 0.10813118019492068,
+      "grad_norm": 60.36637431473924,
+      "learning_rate": 4.9206996984851895e-06,
+      "loss": 1.1408,
+      "step": 760
+    },
+    {
+      "epoch": 0.10955395888169595,
+      "grad_norm": 20.94795499081589,
+      "learning_rate": 4.917795230284499e-06,
+      "loss": 1.0697,
+      "step": 770
+    },
+    {
+      "epoch": 0.11097673756847122,
+      "grad_norm": 15.489949414714857,
+      "learning_rate": 4.914839413093867e-06,
+      "loss": 1.2892,
+      "step": 780
+    },
+    {
+      "epoch": 0.1123995162552465,
+      "grad_norm": 40.226087592606554,
+      "learning_rate": 4.911832309688758e-06,
+      "loss": 0.9953,
+      "step": 790
+    },
+    {
+      "epoch": 0.11382229494202177,
+      "grad_norm": 14.411243294953653,
+      "learning_rate": 4.9087739839338586e-06,
+      "loss": 1.0995,
+      "step": 800
+    },
+    {
+      "epoch": 0.11524507362879705,
+      "grad_norm": 25.746283972263587,
+      "learning_rate": 4.905664500781704e-06,
+      "loss": 1.1964,
+      "step": 810
+    },
+    {
+      "epoch": 0.1166678523155723,
+      "grad_norm": 24.066909183738016,
+      "learning_rate": 4.902503926271311e-06,
+      "loss": 1.0434,
+      "step": 820
+    },
+    {
+      "epoch": 0.11809063100234758,
+      "grad_norm": 10.58109917774702,
+      "learning_rate": 4.899292327526771e-06,
+      "loss": 1.1068,
+      "step": 830
+    },
+    {
+      "epoch": 0.11951340968912286,
+      "grad_norm": 31.468690458560978,
+      "learning_rate": 4.896029772755827e-06,
+      "loss": 1.1184,
+      "step": 840
+    },
+    {
+      "epoch": 0.12093618837589813,
+      "grad_norm": 34.07373775220134,
+      "learning_rate": 4.892716331248421e-06,
+      "loss": 1.131,
+      "step": 850
+    },
+    {
+      "epoch": 0.12235896706267341,
+      "grad_norm": 32.13791878005405,
+      "learning_rate": 4.889352073375224e-06,
+      "loss": 1.0096,
+      "step": 860
+    },
+    {
+      "epoch": 0.12378174574944867,
+      "grad_norm": 25.032238348130715,
+      "learning_rate": 4.885937070586143e-06,
+      "loss": 1.0091,
+      "step": 870
+    },
+    {
+      "epoch": 0.12520452443622396,
+      "grad_norm": 43.796780744116205,
+      "learning_rate": 4.882471395408802e-06,
+      "loss": 1.1524,
+      "step": 880
+    },
+    {
+      "epoch": 0.12662730312299922,
+      "grad_norm": 39.73622388351608,
+      "learning_rate": 4.878955121447003e-06,
+      "loss": 1.0613,
+      "step": 890
+    },
+    {
+      "epoch": 0.12805008180977448,
+      "grad_norm": 23.415663074214173,
+      "learning_rate": 4.875388323379161e-06,
+      "loss": 1.0664,
+      "step": 900
+    },
+    {
+      "epoch": 0.12947286049654977,
+      "grad_norm": 26.895406711294633,
+      "learning_rate": 4.871771076956718e-06,
+      "loss": 1.1549,
+      "step": 910
+    },
+    {
+      "epoch": 0.13089563918332503,
+      "grad_norm": 21.255238865940857,
+      "learning_rate": 4.8681034590025385e-06,
+      "loss": 0.979,
+      "step": 920
+    },
+    {
+      "epoch": 0.13231841787010032,
+      "grad_norm": 18.490585034014213,
+      "learning_rate": 4.864385547409272e-06,
+      "loss": 1.2308,
+      "step": 930
+    },
+    {
+      "epoch": 0.13374119655687558,
+      "grad_norm": 31.35551833525057,
+      "learning_rate": 4.8606174211377e-06,
+      "loss": 1.125,
+      "step": 940
+    },
+    {
+      "epoch": 0.13516397524365084,
+      "grad_norm": 18.48243362822039,
+      "learning_rate": 4.8567991602150635e-06,
+      "loss": 1.0853,
+      "step": 950
+    },
+    {
+      "epoch": 0.13658675393042613,
+      "grad_norm": 20.45592551281928,
+      "learning_rate": 4.852930845733358e-06,
+      "loss": 1.1877,
+      "step": 960
+    },
+    {
+      "epoch": 0.1380095326172014,
+      "grad_norm": 13.073784179027522,
+      "learning_rate": 4.849012559847615e-06,
+      "loss": 1.0812,
+      "step": 970
+    },
+    {
+      "epoch": 0.13943231130397665,
+      "grad_norm": 58.464808300471525,
+      "learning_rate": 4.845044385774154e-06,
+      "loss": 1.0702,
+      "step": 980
+    },
+    {
+      "epoch": 0.14085508999075194,
+      "grad_norm": 32.284426339203456,
+      "learning_rate": 4.841026407788818e-06,
+      "loss": 1.0646,
+      "step": 990
+    },
+    {
+      "epoch": 0.1422778686775272,
+      "grad_norm": 23.861973812252366,
+      "learning_rate": 4.836958711225183e-06,
+      "loss": 1.1582,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1437006473643025,
+      "grad_norm": 35.10042096090748,
+      "learning_rate": 4.832841382472744e-06,
+      "loss": 1.0257,
+      "step": 1010
+    },
+    {
+      "epoch": 0.14512342605107775,
+      "grad_norm": 14.61198523399814,
+      "learning_rate": 4.828674508975081e-06,
+      "loss": 0.9973,
+      "step": 1020
+    },
+    {
+      "epoch": 0.14654620473785301,
+      "grad_norm": 25.535678355280265,
+      "learning_rate": 4.8244581792280045e-06,
+      "loss": 1.1356,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1479689834246283,
+      "grad_norm": 47.776324802406265,
+      "learning_rate": 4.820192482777672e-06,
+      "loss": 0.9028,
+      "step": 1040
+    },
+    {
+      "epoch": 0.14939176211140356,
+      "grad_norm": 14.799263331252178,
+      "learning_rate": 4.815877510218687e-06,
+      "loss": 1.1014,
+      "step": 1050
+    },
+    {
+      "epoch": 0.15081454079817885,
+      "grad_norm": 24.813065365774627,
+      "learning_rate": 4.811513353192181e-06,
+      "loss": 1.1839,
+      "step": 1060
+    },
+    {
+      "epoch": 0.15223731948495411,
+      "grad_norm": 25.60795480877662,
+      "learning_rate": 4.807100104383856e-06,
+      "loss": 0.9862,
+      "step": 1070
+    },
+    {
+      "epoch": 0.15366009817172938,
+      "grad_norm": 81.74250558980262,
+      "learning_rate": 4.802637857522026e-06,
+      "loss": 1.0582,
+      "step": 1080
+    },
+    {
+      "epoch": 0.15508287685850466,
+      "grad_norm": 32.87867343488564,
+      "learning_rate": 4.798126707375623e-06,
+      "loss": 1.1183,
+      "step": 1090
+    },
+    {
+      "epoch": 0.15650565554527993,
+      "grad_norm": 11.38171030977416,
+      "learning_rate": 4.793566749752183e-06,
+      "loss": 1.1165,
+      "step": 1100
+    },
+    {
+      "epoch": 0.15792843423205521,
+      "grad_norm": 21.175570579434233,
+      "learning_rate": 4.78895808149581e-06,
+      "loss": 1.1336,
+      "step": 1110
+    },
+    {
+      "epoch": 0.15935121291883048,
+      "grad_norm": 17.668418598817492,
+      "learning_rate": 4.784300800485126e-06,
+      "loss": 1.0142,
+      "step": 1120
+    },
+    {
+      "epoch": 0.16077399160560574,
+      "grad_norm": 29.88399695204879,
+      "learning_rate": 4.779595005631185e-06,
+      "loss": 1.1087,
+      "step": 1130
+    },
+    {
+      "epoch": 0.16219677029238103,
+      "grad_norm": 28.1710561261029,
+      "learning_rate": 4.7748407968753744e-06,
+      "loss": 1.0295,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1636195489791563,
+      "grad_norm": 25.185712797257235,
+      "learning_rate": 4.770038275187295e-06,
+      "loss": 1.0201,
+      "step": 1150
+    },
+    {
+      "epoch": 0.16504232766593158,
+      "grad_norm": 59.25807393338673,
+      "learning_rate": 4.765187542562615e-06,
+      "loss": 1.0377,
+      "step": 1160
+    },
+    {
+      "epoch": 0.16646510635270684,
+      "grad_norm": 29.97235785066762,
+      "learning_rate": 4.760288702020901e-06,
+      "loss": 1.0311,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1678878850394821,
+      "grad_norm": 37.35879781998382,
+      "learning_rate": 4.755341857603436e-06,
+      "loss": 1.0862,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1693106637262574,
+      "grad_norm": 23.672541588017463,
+      "learning_rate": 4.750347114371004e-06,
+      "loss": 0.9988,
+      "step": 1190
+    },
+    {
+      "epoch": 0.17073344241303265,
+      "grad_norm": 19.550633147562124,
+      "learning_rate": 4.745304578401662e-06,
+      "loss": 0.9844,
+      "step": 1200
+    },
+    {
+      "epoch": 0.17215622109980794,
+      "grad_norm": 32.695401824329494,
+      "learning_rate": 4.740214356788486e-06,
+      "loss": 1.0509,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1735789997865832,
+      "grad_norm": 9.646363052517763,
+      "learning_rate": 4.735076557637297e-06,
+      "loss": 1.0341,
+      "step": 1220
+    },
+    {
+      "epoch": 0.17500177847335846,
+      "grad_norm": 24.741345371927583,
+      "learning_rate": 4.729891290064365e-06,
+      "loss": 1.0208,
+      "step": 1230
+    },
+    {
+      "epoch": 0.17642455716013375,
+      "grad_norm": 45.651678686892964,
+      "learning_rate": 4.724658664194091e-06,
+      "loss": 1.0502,
+      "step": 1240
+    },
+    {
+      "epoch": 0.177847335846909,
+      "grad_norm": 19.24141408740206,
+      "learning_rate": 4.71937879115667e-06,
+      "loss": 0.9975,
+      "step": 1250
+    },
+    {
+      "epoch": 0.17927011453368427,
+      "grad_norm": 31.556248725504627,
+      "learning_rate": 4.714051783085728e-06,
+      "loss": 1.1013,
+      "step": 1260
+    },
+    {
+      "epoch": 0.18069289322045956,
+      "grad_norm": 20.17633982870767,
+      "learning_rate": 4.708677753115942e-06,
+      "loss": 0.9531,
+      "step": 1270
+    },
+    {
+      "epoch": 0.18211567190723482,
+      "grad_norm": 19.4051188478354,
+      "learning_rate": 4.703256815380639e-06,
+      "loss": 1.0627,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1835384505940101,
+      "grad_norm": 31.784954367517898,
+      "learning_rate": 4.6977890850093684e-06,
+      "loss": 0.9364,
+      "step": 1290
+    },
+    {
+      "epoch": 0.18496122928078537,
+      "grad_norm": 59.59839123261177,
+      "learning_rate": 4.692274678125461e-06,
+      "loss": 0.9959,
+      "step": 1300
+    },
+    {
+      "epoch": 0.18638400796756063,
+      "grad_norm": 86.09375297334182,
+      "learning_rate": 4.6867137118435565e-06,
+      "loss": 1.0006,
+      "step": 1310
+    },
+    {
+      "epoch": 0.18780678665433592,
+      "grad_norm": 22.56759533641289,
+      "learning_rate": 4.6811063042671245e-06,
+      "loss": 0.917,
+      "step": 1320
+    },
+    {
+      "epoch": 0.18922956534111118,
+      "grad_norm": 42.93003422068975,
+      "learning_rate": 4.675452574485952e-06,
+      "loss": 0.9586,
+      "step": 1330
+    },
+    {
+      "epoch": 0.19065234402788647,
+      "grad_norm": 25.098646626316107,
+      "learning_rate": 4.669752642573609e-06,
+      "loss": 0.958,
+      "step": 1340
+    },
+    {
+      "epoch": 0.19207512271466173,
+      "grad_norm": 26.200356526413692,
+      "learning_rate": 4.664006629584909e-06,
+      "loss": 0.9578,
+      "step": 1350
+    },
+    {
+      "epoch": 0.193497901401437,
+      "grad_norm": 40.45559950946015,
+      "learning_rate": 4.658214657553333e-06,
+      "loss": 1.1494,
+      "step": 1360
+    },
+    {
+      "epoch": 0.19492068008821228,
+      "grad_norm": 11.30097276948413,
+      "learning_rate": 4.652376849488432e-06,
+      "loss": 1.0556,
+      "step": 1370
+    },
+    {
+      "epoch": 0.19634345877498754,
+      "grad_norm": 20.822794274521176,
+      "learning_rate": 4.646493329373226e-06,
+      "loss": 1.0884,
+      "step": 1380
+    },
+    {
+      "epoch": 0.19776623746176283,
+      "grad_norm": 29.222350209439792,
+      "learning_rate": 4.640564222161559e-06,
+      "loss": 0.9724,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1991890161485381,
+      "grad_norm": 17.347939039481435,
+      "learning_rate": 4.634589653775456e-06,
+      "loss": 0.9809,
+      "step": 1400
+    },
+    {
+      "epoch": 0.20061179483531336,
+      "grad_norm": 35.203954530331465,
+      "learning_rate": 4.628569751102442e-06,
+      "loss": 0.9926,
+      "step": 1410
+    },
+    {
+      "epoch": 0.20203457352208865,
+      "grad_norm": 21.145870608823014,
+      "learning_rate": 4.622504641992847e-06,
+      "loss": 0.9443,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2034573522088639,
+      "grad_norm": 109.94274559504879,
+      "learning_rate": 4.616394455257095e-06,
+      "loss": 1.0751,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2048801308956392,
+      "grad_norm": 72.18769613220586,
+      "learning_rate": 4.610239320662965e-06,
+      "loss": 0.9313,
+      "step": 1440
+    },
+    {
+      "epoch": 0.20630290958241446,
+      "grad_norm": 13.0262713604652,
+      "learning_rate": 4.6040393689328365e-06,
+      "loss": 1.0373,
+      "step": 1450
+    },
+    {
+      "epoch": 0.20772568826918972,
+      "grad_norm": 24.568104013079804,
+      "learning_rate": 4.5977947317409115e-06,
+      "loss": 0.9896,
+      "step": 1460
+    },
+    {
+      "epoch": 0.209148466955965,
+      "grad_norm": 19.83377026210839,
+      "learning_rate": 4.591505541710419e-06,
+      "loss": 0.9114,
+      "step": 1470
+    },
+    {
+      "epoch": 0.21057124564274027,
+      "grad_norm": 15.803261030668528,
+      "learning_rate": 4.585171932410802e-06,
+      "loss": 1.0671,
+      "step": 1480
+    },
+    {
+      "epoch": 0.21199402432951556,
+      "grad_norm": 73.8034968144292,
+      "learning_rate": 4.57879403835487e-06,
+      "loss": 0.9116,
+      "step": 1490
+    },
+    {
+      "epoch": 0.21341680301629082,
+      "grad_norm": 50.65746396577896,
+      "learning_rate": 4.572371994995959e-06,
+      "loss": 0.9233,
+      "step": 1500
+    },
+    {
+      "epoch": 0.21483958170306608,
+      "grad_norm": 19.756346599520104,
+      "learning_rate": 4.565905938725036e-06,
+      "loss": 0.9569,
+      "step": 1510
+    },
+    {
+      "epoch": 0.21626236038984137,
+      "grad_norm": 9.29993489756742,
+      "learning_rate": 4.559396006867819e-06,
+      "loss": 0.8349,
+      "step": 1520
+    },
+    {
+      "epoch": 0.21768513907661663,
+      "grad_norm": 25.972739952389897,
+      "learning_rate": 4.552842337681848e-06,
+      "loss": 1.004,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2191079177633919,
+      "grad_norm": 88.34667254185341,
+      "learning_rate": 4.546245070353559e-06,
+      "loss": 0.9903,
+      "step": 1540
+    },
+    {
+      "epoch": 0.22053069645016718,
+      "grad_norm": 31.07756438176318,
+      "learning_rate": 4.539604344995315e-06,
+      "loss": 1.0547,
+      "step": 1550
+    },
+    {
+      "epoch": 0.22195347513694244,
+      "grad_norm": 37.65219781115293,
+      "learning_rate": 4.532920302642444e-06,
+      "loss": 0.9818,
+      "step": 1560
+    },
+    {
+      "epoch": 0.22337625382371773,
+      "grad_norm": 11.736978172112924,
+      "learning_rate": 4.5261930852502374e-06,
+      "loss": 0.8711,
+      "step": 1570
+    },
+    {
+      "epoch": 0.224799032510493,
+      "grad_norm": 53.43167755603555,
+      "learning_rate": 4.519422835690934e-06,
+      "loss": 1.0607,
+      "step": 1580
+    },
+    {
+      "epoch": 0.22622181119726825,
+      "grad_norm": 33.172428433273375,
+      "learning_rate": 4.512609697750687e-06,
+      "loss": 0.8635,
+      "step": 1590
+    },
+    {
+      "epoch": 0.22764458988404354,
+      "grad_norm": 30.294310375393177,
+      "learning_rate": 4.505753816126511e-06,
+      "loss": 1.0003,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2290673685708188,
+      "grad_norm": 16.381735397778595,
+      "learning_rate": 4.4988553364232065e-06,
+      "loss": 0.9944,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2304901472575941,
+      "grad_norm": 25.441517619108396,
+      "learning_rate": 4.491914405150275e-06,
+      "loss": 0.8836,
+      "step": 1620
+    },
+    {
+      "epoch": 0.23191292594436935,
+      "grad_norm": 22.84586846575256,
+      "learning_rate": 4.4849311697187955e-06,
+      "loss": 0.9763,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2333357046311446,
+      "grad_norm": 22.741071369838796,
+      "learning_rate": 4.477905778438303e-06,
+      "loss": 0.9,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2347584833179199,
+      "grad_norm": 45.274883279817054,
+      "learning_rate": 4.470838380513637e-06,
+      "loss": 0.9619,
+      "step": 1650
+    },
+    {
+      "epoch": 0.23618126200469516,
+      "grad_norm": 28.411705472066803,
+      "learning_rate": 4.463729126041769e-06,
+      "loss": 0.9951,
+      "step": 1660
+    },
+    {
+      "epoch": 0.23760404069147045,
+      "grad_norm": 23.03515293998754,
+      "learning_rate": 4.45657816600862e-06,
+      "loss": 0.9833,
+      "step": 1670
+    },
+    {
+      "epoch": 0.23902681937824571,
+      "grad_norm": 36.26988848442576,
+      "learning_rate": 4.449385652285849e-06,
+      "loss": 0.9385,
+      "step": 1680
+    },
+    {
+      "epoch": 0.24044959806502098,
+      "grad_norm": 29.98622414763686,
+      "learning_rate": 4.44215173762763e-06,
+      "loss": 1.0314,
+      "step": 1690
+    },
+    {
+      "epoch": 0.24187237675179626,
+      "grad_norm": 25.99130726723674,
+      "learning_rate": 4.434876575667409e-06,
+      "loss": 1.0867,
+      "step": 1700
+    },
+    {
+      "epoch": 0.24329515543857153,
+      "grad_norm": 19.5639741652505,
+      "learning_rate": 4.42756032091464e-06,
+      "loss": 1.0046,
+      "step": 1710
+    },
+    {
+      "epoch": 0.24471793412534681,
+      "grad_norm": 34.129588654435274,
+      "learning_rate": 4.4202031287515e-06,
+      "loss": 0.9393,
+      "step": 1720
+    },
+    {
+      "epoch": 0.24614071281212208,
+      "grad_norm": 18.561263502065916,
+      "learning_rate": 4.412805155429596e-06,
+      "loss": 0.9873,
+      "step": 1730
+    },
+    {
+      "epoch": 0.24756349149889734,
+      "grad_norm": 37.94544849948392,
+      "learning_rate": 4.4053665580666406e-06,
+      "loss": 0.9624,
+      "step": 1740
+    },
+    {
+      "epoch": 0.24898627018567263,
+      "grad_norm": 84.98992365379094,
+      "learning_rate": 4.397887494643119e-06,
+      "loss": 0.9292,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2504090488724479,
+      "grad_norm": 41.9440027405956,
+      "learning_rate": 4.390368123998928e-06,
+      "loss": 1.1042,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2518318275592232,
+      "grad_norm": 18.271645141393737,
+      "learning_rate": 4.382808605830014e-06,
+      "loss": 0.8905,
+      "step": 1770
+    },
+    {
+      "epoch": 0.25325460624599844,
+      "grad_norm": 25.615500179028004,
+      "learning_rate": 4.375209100684968e-06,
+      "loss": 0.9344,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2546773849327737,
+      "grad_norm": 34.000225650472906,
+      "learning_rate": 4.367569769961624e-06,
+      "loss": 1.0618,
+      "step": 1790
+    },
+    {
+      "epoch": 0.25610016361954896,
+      "grad_norm": 17.946869566933312,
+      "learning_rate": 4.35989077590363e-06,
+      "loss": 0.9339,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2575229423063243,
+      "grad_norm": 39.112615826338605,
+      "learning_rate": 4.352172281596999e-06,
+      "loss": 0.887,
+      "step": 1810
+    },
+    {
+      "epoch": 0.25894572099309954,
+      "grad_norm": 10.173767469816399,
+      "learning_rate": 4.344414450966652e-06,
+      "loss": 0.9607,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2603684996798748,
+      "grad_norm": 20.08966991211306,
+      "learning_rate": 4.336617448772929e-06,
+      "loss": 1.0094,
+      "step": 1830
+    },
+    {
+      "epoch": 0.26179127836665006,
+      "grad_norm": 23.382294735645083,
+      "learning_rate": 4.3287814406080955e-06,
+      "loss": 0.9473,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2632140570534253,
+      "grad_norm": 13.416514028844402,
+      "learning_rate": 4.320906592892821e-06,
+      "loss": 0.8759,
+      "step": 1850
+    },
+    {
+      "epoch": 0.26463683574020064,
+      "grad_norm": 18.21997944762085,
+      "learning_rate": 4.3129930728726475e-06,
+      "loss": 0.9567,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2660596144269759,
+      "grad_norm": 26.833449827216317,
+      "learning_rate": 4.305041048614441e-06,
+      "loss": 0.8975,
+      "step": 1870
+    },
+    {
+      "epoch": 0.26748239311375116,
+      "grad_norm": 17.479289371987786,
+      "learning_rate": 4.297050689002814e-06,
+      "loss": 0.8648,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2689051718005264,
+      "grad_norm": 43.85707088350821,
+      "learning_rate": 4.289022163736544e-06,
+      "loss": 0.9487,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2703279504873017,
+      "grad_norm": 59.69136365583277,
+      "learning_rate": 4.280955643324969e-06,
+      "loss": 0.8577,
+      "step": 1900
+    },
+    {
+      "epoch": 0.271750729174077,
+      "grad_norm": 24.95179608305563,
+      "learning_rate": 4.272851299084365e-06,
+      "loss": 0.9763,
+      "step": 1910
+    },
+    {
+      "epoch": 0.27317350786085226,
+      "grad_norm": 25.905615353023105,
+      "learning_rate": 4.264709303134308e-06,
+      "loss": 0.8112,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2745962865476275,
+      "grad_norm": 16.44416792587216,
+      "learning_rate": 4.256529828394019e-06,
+      "loss": 0.9131,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2760190652344028,
+      "grad_norm": 21.581395452993693,
+      "learning_rate": 4.2483130485786924e-06,
+      "loss": 0.8873,
+      "step": 1940
+    },
+    {
+      "epoch": 0.27744184392117804,
+      "grad_norm": 19.647619107063836,
+      "learning_rate": 4.2400591381958044e-06,
+      "loss": 0.8884,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2788646226079533,
+      "grad_norm": 23.162495970616806,
+      "learning_rate": 4.231768272541408e-06,
+      "loss": 0.8656,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2802874012947286,
+      "grad_norm": 24.785879598129082,
+      "learning_rate": 4.22344062769641e-06,
+      "loss": 0.8738,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2817101799815039,
+      "grad_norm": 41.50134900928378,
+      "learning_rate": 4.215076380522832e-06,
+      "loss": 0.9892,
+      "step": 1980
+    },
+    {
+      "epoch": 0.28313295866827914,
+      "grad_norm": 26.038557737763075,
+      "learning_rate": 4.206675708660053e-06,
+      "loss": 0.9443,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2845557373550544,
+      "grad_norm": 22.726648702579652,
+      "learning_rate": 4.198238790521037e-06,
+      "loss": 0.9251,
+      "step": 2000
+    },
+    {
+      "epoch": 0.28597851604182967,
+      "grad_norm": 53.35902150926828,
+      "learning_rate": 4.189765805288547e-06,
+      "loss": 0.8751,
+      "step": 2010
+    },
+    {
+      "epoch": 0.287401294728605,
+      "grad_norm": 17.06572011162811,
+      "learning_rate": 4.181256932911333e-06,
+      "loss": 0.9443,
+      "step": 2020
+    },
+    {
+      "epoch": 0.28882407341538024,
+      "grad_norm": 11.916586599185443,
+      "learning_rate": 4.172712354100315e-06,
+      "loss": 0.9999,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2902468521021555,
+      "grad_norm": 23.991331588369764,
+      "learning_rate": 4.164132250324747e-06,
+      "loss": 0.9772,
+      "step": 2040
+    },
+    {
+      "epoch": 0.29166963078893077,
+      "grad_norm": 12.585371928264612,
+      "learning_rate": 4.155516803808357e-06,
+      "loss": 0.91,
+      "step": 2050
+    },
+    {
+      "epoch": 0.29309240947570603,
+      "grad_norm": 28.322986319509504,
+      "learning_rate": 4.14686619752548e-06,
+      "loss": 1.0151,
+      "step": 2060
+    },
+    {
+      "epoch": 0.29451518816248135,
+      "grad_norm": 13.797656671652906,
+      "learning_rate": 4.138180615197173e-06,
+      "loss": 0.8534,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2959379668492566,
+      "grad_norm": 12.278884499684143,
+      "learning_rate": 4.129460241287313e-06,
+      "loss": 0.92,
+      "step": 2080
+    },
+    {
+      "epoch": 0.29736074553603187,
+      "grad_norm": 21.325682869764425,
+      "learning_rate": 4.120705260998677e-06,
+      "loss": 0.9308,
+      "step": 2090
+    },
+    {
+      "epoch": 0.29878352422280713,
+      "grad_norm": 15.28691034387365,
+      "learning_rate": 4.111915860269011e-06,
+      "loss": 0.9071,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3002063029095824,
+      "grad_norm": 17.27001412320014,
+      "learning_rate": 4.103092225767079e-06,
+      "loss": 1.0093,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3016290815963577,
+      "grad_norm": 9.580279459737769,
+      "learning_rate": 4.094234544888702e-06,
+      "loss": 0.8731,
+      "step": 2120
+    },
+    {
+      "epoch": 0.30305186028313297,
+      "grad_norm": 39.44419290450929,
+      "learning_rate": 4.085343005752774e-06,
+      "loss": 1.0284,
+      "step": 2130
+    },
+    {
+      "epoch": 0.30447463896990823,
+      "grad_norm": 21.611840041612204,
+      "learning_rate": 4.07641779719727e-06,
+      "loss": 0.9289,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3058974176566835,
+      "grad_norm": 18.649439161622322,
+      "learning_rate": 4.067459108775233e-06,
+      "loss": 0.8894,
+      "step": 2150
+    },
+    {
+      "epoch": 0.30732019634345875,
+      "grad_norm": 13.95416605872919,
+      "learning_rate": 4.058467130750751e-06,
+      "loss": 0.9318,
+      "step": 2160
+    },
+    {
+      "epoch": 0.30874297503023407,
+      "grad_norm": 20.547631527897213,
+      "learning_rate": 4.049442054094911e-06,
+      "loss": 0.9263,
+      "step": 2170
+    },
+    {
+      "epoch": 0.31016575371700933,
+      "grad_norm": 35.25360992586632,
+      "learning_rate": 4.040384070481751e-06,
+      "loss": 0.9331,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3115885324037846,
+      "grad_norm": 19.707802730532986,
+      "learning_rate": 4.031293372284182e-06,
+      "loss": 0.9877,
+      "step": 2190
+    },
+    {
+      "epoch": 0.31301131109055985,
+      "grad_norm": 51.74288791532746,
+      "learning_rate": 4.022170152569908e-06,
+      "loss": 0.9195,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3144340897773351,
+      "grad_norm": 12.795646156712829,
+      "learning_rate": 4.013014605097321e-06,
+      "loss": 0.9595,
+      "step": 2210
+    },
+    {
+      "epoch": 0.31585686846411043,
+      "grad_norm": 26.854021865817128,
+      "learning_rate": 4.0038269243113875e-06,
+      "loss": 0.8867,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3172796471508857,
+      "grad_norm": 17.896171542975722,
+      "learning_rate": 3.994607305339522e-06,
+      "loss": 0.9779,
+      "step": 2230
+    },
+    {
+      "epoch": 0.31870242583766095,
+      "grad_norm": 55.11927308390862,
+      "learning_rate": 3.985355943987437e-06,
+      "loss": 0.8624,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3201252045244362,
+      "grad_norm": 17.36025025212368,
+      "learning_rate": 3.976073036734992e-06,
+      "loss": 0.9457,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3215479832112115,
+      "grad_norm": 23.497001094477394,
+      "learning_rate": 3.966758780732012e-06,
+      "loss": 0.8444,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3229707618979868,
+      "grad_norm": 24.99517493980936,
+      "learning_rate": 3.957413373794106e-06,
+      "loss": 0.8614,
+      "step": 2270
+    },
+    {
+      "epoch": 0.32439354058476205,
+      "grad_norm": 13.13877674656499,
+      "learning_rate": 3.948037014398467e-06,
+      "loss": 0.8804,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3258163192715373,
+      "grad_norm": 57.07057351332041,
+      "learning_rate": 3.938629901679652e-06,
+      "loss": 0.8595,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3272390979583126,
+      "grad_norm": 23.765236030390856,
+      "learning_rate": 3.929192235425358e-06,
+      "loss": 0.9584,
+      "step": 2300
+    },
+    {
+      "epoch": 0.32866187664508784,
+      "grad_norm": 28.279346440606176,
+      "learning_rate": 3.919724216072173e-06,
+      "loss": 0.9113,
+      "step": 2310
+    },
+    {
+      "epoch": 0.33008465533186315,
+      "grad_norm": 11.409844719072586,
+      "learning_rate": 3.910226044701323e-06,
+      "loss": 0.8503,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3315074340186384,
+      "grad_norm": 22.617931436775116,
+      "learning_rate": 3.900697923034401e-06,
+      "loss": 0.9235,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3329302127054137,
+      "grad_norm": 15.776516228855538,
+      "learning_rate": 3.891140053429085e-06,
+      "loss": 0.8564,
+      "step": 2340
+    },
+    {
+      "epoch": 0.33435299139218894,
+      "grad_norm": 15.955596620116578,
+      "learning_rate": 3.881552638874835e-06,
+      "loss": 0.8281,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3357757700789642,
+      "grad_norm": 17.290067596352323,
+      "learning_rate": 3.8719358829885845e-06,
+      "loss": 0.8976,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3371985487657395,
+      "grad_norm": 11.323070820416119,
+      "learning_rate": 3.862289990010418e-06,
+      "loss": 0.7097,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3386213274525148,
+      "grad_norm": 13.479043421646962,
+      "learning_rate": 3.852615164799233e-06,
+      "loss": 0.9754,
+      "step": 2380
+    },
+    {
+      "epoch": 0.34004410613929004,
+      "grad_norm": 30.713987560244302,
+      "learning_rate": 3.842911612828385e-06,
+      "loss": 0.7999,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3414668848260653,
+      "grad_norm": 13.737577446750928,
+      "learning_rate": 3.8331795401813285e-06,
+      "loss": 0.7721,
+      "step": 2400
+    },
+    {
+      "epoch": 0.34288966351284056,
+      "grad_norm": 22.24678775851343,
+      "learning_rate": 3.823419153547238e-06,
+      "loss": 0.8526,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3443124421996159,
+      "grad_norm": 17.04844934249826,
+      "learning_rate": 3.8136306602166194e-06,
+      "loss": 0.8595,
+      "step": 2420
+    },
+    {
+      "epoch": 0.34573522088639114,
+      "grad_norm": 25.308993688836463,
+      "learning_rate": 3.8038142680769063e-06,
+      "loss": 1.0052,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3471579995731664,
+      "grad_norm": 14.393502318885858,
+      "learning_rate": 3.7939701856080457e-06,
+      "loss": 0.8819,
+      "step": 2440
+    },
+    {
+      "epoch": 0.34858077825994166,
+      "grad_norm": 10.4265150253465,
+      "learning_rate": 3.7840986218780685e-06,
+      "loss": 0.8679,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3500035569467169,
+      "grad_norm": 17.551190684773992,
+      "learning_rate": 3.774199786538654e-06,
+      "loss": 0.8846,
+      "step": 2460
+    },
+    {
+      "epoch": 0.35142633563349224,
+      "grad_norm": 18.0417618263592,
+      "learning_rate": 3.764273889820673e-06,
+      "loss": 0.8088,
+      "step": 2470
+    },
+    {
+      "epoch": 0.3528491143202675,
+      "grad_norm": 46.10956168861658,
+      "learning_rate": 3.754321142529724e-06,
+      "loss": 0.8241,
+      "step": 2480
+    },
+    {
+      "epoch": 0.35427189300704276,
+      "grad_norm": 25.26864515730438,
+      "learning_rate": 3.744341756041655e-06,
+      "loss": 0.8299,
+      "step": 2490
+    },
+    {
+      "epoch": 0.355694671693818,
+      "grad_norm": 21.076814874729575,
+      "learning_rate": 3.7343359422980795e-06,
+      "loss": 0.8453,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3571174503805933,
+      "grad_norm": 14.730127479771735,
+      "learning_rate": 3.724303913801867e-06,
+      "loss": 0.8901,
+      "step": 2510
+    },
+    {
+      "epoch": 0.35854022906736854,
+      "grad_norm": 19.02164497852309,
+      "learning_rate": 3.714245883612637e-06,
+      "loss": 0.7881,
+      "step": 2520
+    },
+    {
+      "epoch": 0.35996300775414386,
+      "grad_norm": 25.87504804339562,
+      "learning_rate": 3.7041620653422318e-06,
+      "loss": 1.0057,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3613857864409191,
+      "grad_norm": 18.900367608478,
+      "learning_rate": 3.6940526731501765e-06,
+      "loss": 0.8415,
+      "step": 2540
+    },
+    {
+      "epoch": 0.3628085651276944,
+      "grad_norm": 13.818205544374647,
+      "learning_rate": 3.6839179217391376e-06,
+      "loss": 0.8001,
+      "step": 2550
+    },
+    {
+      "epoch": 0.36423134381446964,
+      "grad_norm": 15.922467404725849,
+      "learning_rate": 3.673758026350355e-06,
+      "loss": 0.9267,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3656541225012449,
+      "grad_norm": 14.467916102674197,
+      "learning_rate": 3.6635732027590787e-06,
+      "loss": 0.8614,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3670769011880202,
+      "grad_norm": 11.21170529268701,
+      "learning_rate": 3.6533636672699795e-06,
+      "loss": 0.9035,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3684996798747955,
+      "grad_norm": 27.606309272195567,
+      "learning_rate": 3.6431296367125612e-06,
+      "loss": 0.7962,
+      "step": 2590
+    },
+    {
+      "epoch": 0.36992245856157074,
+      "grad_norm": 19.853262159190642,
+      "learning_rate": 3.632871328436548e-06,
+      "loss": 0.8038,
+      "step": 2600
+    },
+    {
+      "epoch": 0.371345237248346,
+      "grad_norm": 19.731745048336382,
+      "learning_rate": 3.622588960307275e-06,
+      "loss": 0.9304,
+      "step": 2610
+    },
+    {
+      "epoch": 0.37276801593512127,
+      "grad_norm": 21.55654616341265,
+      "learning_rate": 3.612282750701062e-06,
+      "loss": 0.8366,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3741907946218966,
+      "grad_norm": 26.397576545609592,
+      "learning_rate": 3.601952918500567e-06,
+      "loss": 0.9263,
+      "step": 2630
+    },
+    {
+      "epoch": 0.37561357330867184,
+      "grad_norm": 25.926948460471703,
+      "learning_rate": 3.591599683090148e-06,
+      "loss": 0.9104,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3770363519954471,
+      "grad_norm": 15.998065211902164,
+      "learning_rate": 3.5812232643511958e-06,
+      "loss": 0.7771,
+      "step": 2650
+    },
+    {
+      "epoch": 0.37845913068222237,
+      "grad_norm": 26.96727158385917,
+      "learning_rate": 3.5708238826574693e-06,
+      "loss": 0.9005,
+      "step": 2660
+    },
+    {
+      "epoch": 0.37988190936899763,
+      "grad_norm": 20.10325840992243,
+      "learning_rate": 3.560401758870412e-06,
+      "loss": 0.7469,
+      "step": 2670
+    },
+    {
+      "epoch": 0.38130468805577294,
+      "grad_norm": 50.12252765157174,
+      "learning_rate": 3.549957114334464e-06,
+      "loss": 0.844,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3827274667425482,
+      "grad_norm": 35.46308598534144,
+      "learning_rate": 3.5394901708723583e-06,
+      "loss": 0.713,
+      "step": 2690
+    },
+    {
+      "epoch": 0.38415024542932347,
+      "grad_norm": 13.37358811877764,
+      "learning_rate": 3.529001150780412e-06,
+      "loss": 0.8835,
+      "step": 2700
+    },
+    {
+      "epoch": 0.38557302411609873,
+      "grad_norm": 67.31815559112255,
+      "learning_rate": 3.5184902768238036e-06,
+      "loss": 0.8744,
+      "step": 2710
+    },
+    {
+      "epoch": 0.386995802802874,
+      "grad_norm": 20.83885896533027,
+      "learning_rate": 3.5079577722318438e-06,
+      "loss": 0.8298,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3884185814896493,
+      "grad_norm": 35.929755336977074,
+      "learning_rate": 3.497403860693232e-06,
+      "loss": 0.824,
+      "step": 2730
+    },
+    {
+      "epoch": 0.38984136017642457,
+      "grad_norm": 22.769671014939643,
+      "learning_rate": 3.4868287663513088e-06,
+      "loss": 0.8447,
+      "step": 2740
+    },
+    {
+      "epoch": 0.39126413886319983,
+      "grad_norm": 33.4314261483403,
+      "learning_rate": 3.4762327137992915e-06,
+      "loss": 0.7872,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3926869175499751,
+      "grad_norm": 30.356120631372097,
+      "learning_rate": 3.46561592807551e-06,
+      "loss": 0.8708,
+      "step": 2760
+    },
+    {
+      "epoch": 0.39410969623675035,
+      "grad_norm": 13.978203051247352,
+      "learning_rate": 3.4549786346586212e-06,
+      "loss": 0.7915,
+      "step": 2770
+    },
+    {
+      "epoch": 0.39553247492352567,
+      "grad_norm": 29.198172693027075,
+      "learning_rate": 3.444321059462823e-06,
+      "loss": 0.9637,
+      "step": 2780
+    },
+    {
+      "epoch": 0.39695525361030093,
+      "grad_norm": 38.94328556809456,
+      "learning_rate": 3.4336434288330595e-06,
+      "loss": 0.7995,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3983780322970762,
+      "grad_norm": 12.890373512474305,
+      "learning_rate": 3.4229459695402068e-06,
+      "loss": 0.9066,
+      "step": 2800
+    },
+    {
+      "epoch": 0.39980081098385145,
+      "grad_norm": 29.611565622555542,
+      "learning_rate": 3.4122289087762646e-06,
+      "loss": 0.8502,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4012235896706267,
+      "grad_norm": 19.90868187031473,
+      "learning_rate": 3.4014924741495263e-06,
+      "loss": 0.8319,
+      "step": 2820
+    },
+    {
+      "epoch": 0.40264636835740203,
+      "grad_norm": 19.810385554932147,
+      "learning_rate": 3.3907368936797443e-06,
+      "loss": 0.8489,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4040691470441773,
+      "grad_norm": 33.35123940706816,
+      "learning_rate": 3.3799623957932924e-06,
+      "loss": 0.8622,
+      "step": 2840
+    },
+    {
+      "epoch": 0.40549192573095255,
+      "grad_norm": 24.606830290642474,
+      "learning_rate": 3.3691692093183093e-06,
+      "loss": 0.8794,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4069147044177278,
+      "grad_norm": 24.341434672257584,
+      "learning_rate": 3.358357563479841e-06,
+      "loss": 0.7623,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4083374831045031,
+      "grad_norm": 20.93359541770851,
+      "learning_rate": 3.3475276878949724e-06,
+      "loss": 0.7811,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4097602617912784,
+      "grad_norm": 11.64588748669064,
+      "learning_rate": 3.336679812567953e-06,
+      "loss": 0.8174,
+      "step": 2880
+    },
+    {
+      "epoch": 0.41118304047805365,
+      "grad_norm": 29.809570461704386,
+      "learning_rate": 3.3258141678853064e-06,
+      "loss": 0.7583,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4126058191648289,
+      "grad_norm": 19.944685397539168,
+      "learning_rate": 3.3149309846109403e-06,
+      "loss": 0.8549,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4140285978516042,
+      "grad_norm": 13.417267929086359,
+      "learning_rate": 3.3040304938812506e-06,
+      "loss": 0.7833,
+      "step": 2910
+    },
+    {
+      "epoch": 0.41545137653837944,
+      "grad_norm": 11.077322871687823,
+      "learning_rate": 3.2931129272002033e-06,
+      "loss": 0.7773,
+      "step": 2920
+    },
+    {
+      "epoch": 0.41687415522515475,
+      "grad_norm": 12.517933094558126,
+      "learning_rate": 3.2821785164344242e-06,
+      "loss": 0.8651,
+      "step": 2930
+    },
+    {
+      "epoch": 0.41829693391193,
+      "grad_norm": 28.71996579019108,
+      "learning_rate": 3.2712274938082746e-06,
+      "loss": 0.7801,
+      "step": 2940
+    },
+    {
+      "epoch": 0.4197197125987053,
+      "grad_norm": 16.51415564607136,
+      "learning_rate": 3.2602600918989126e-06,
+      "loss": 0.7822,
+      "step": 2950
+    },
+    {
+      "epoch": 0.42114249128548054,
+      "grad_norm": 24.564509764438558,
+      "learning_rate": 3.249276543631363e-06,
+      "loss": 0.8514,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4225652699722558,
+      "grad_norm": 15.847385700324564,
+      "learning_rate": 3.238277082273564e-06,
+      "loss": 0.8609,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4239880486590311,
+      "grad_norm": 38.683130093540626,
+      "learning_rate": 3.2272619414314167e-06,
+      "loss": 0.777,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4254108273458064,
+      "grad_norm": 22.996043924193113,
+      "learning_rate": 3.2162313550438197e-06,
+      "loss": 0.7298,
+      "step": 2990
+    },
+    {
+      "epoch": 0.42683360603258164,
+      "grad_norm": 33.18898104375557,
+      "learning_rate": 3.2051855573777045e-06,
+      "loss": 0.846,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4282563847193569,
+      "grad_norm": 30.006547705606142,
+      "learning_rate": 3.19412478302306e-06,
+      "loss": 0.8202,
+      "step": 3010
+    },
+    {
+      "epoch": 0.42967916340613216,
+      "grad_norm": 21.366134925657423,
+      "learning_rate": 3.1830492668879464e-06,
+      "loss": 0.805,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4311019420929075,
+      "grad_norm": 15.471273661530745,
+      "learning_rate": 3.171959244193512e-06,
+      "loss": 0.906,
+      "step": 3030
+    },
+    {
+      "epoch": 0.43252472077968274,
+      "grad_norm": 62.7358588675245,
+      "learning_rate": 3.160854950468991e-06,
+      "loss": 0.9361,
+      "step": 3040
+    },
+    {
+      "epoch": 0.433947499466458,
+      "grad_norm": 87.71714813403881,
+      "learning_rate": 3.149736621546709e-06,
+      "loss": 0.8383,
+      "step": 3050
+    },
+    {
+      "epoch": 0.43537027815323326,
+      "grad_norm": 27.520759090414302,
+      "learning_rate": 3.1386044935570665e-06,
+      "loss": 0.876,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4367930568400085,
+      "grad_norm": 13.9736259962008,
+      "learning_rate": 3.1274588029235297e-06,
+      "loss": 0.6913,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4382158355267838,
+      "grad_norm": 15.111893897955833,
+      "learning_rate": 3.1162997863576076e-06,
+      "loss": 0.8171,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4396386142135591,
+      "grad_norm": 25.631492477295634,
+      "learning_rate": 3.1051276808538238e-06,
+      "loss": 0.8492,
+      "step": 3090
+    },
+    {
+      "epoch": 0.44106139290033436,
+      "grad_norm": 43.326724052088984,
+      "learning_rate": 3.093942723684684e-06,
+      "loss": 0.754,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4424841715871096,
+      "grad_norm": 31.65852965127963,
+      "learning_rate": 3.0827451523956385e-06,
+      "loss": 0.781,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4439069502738849,
+      "grad_norm": 10.557234288549818,
+      "learning_rate": 3.071535204800032e-06,
+      "loss": 0.8802,
+      "step": 3120
+    },
+    {
+      "epoch": 0.44532972896066014,
+      "grad_norm": 7.741616062639269,
+      "learning_rate": 3.060313118974059e-06,
+      "loss": 0.9144,
+      "step": 3130
+    },
+    {
+      "epoch": 0.44675250764743546,
+      "grad_norm": 17.118385663085036,
+      "learning_rate": 3.049079133251706e-06,
+      "loss": 0.8421,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4481752863342107,
+      "grad_norm": 15.048257306198911,
+      "learning_rate": 3.0378334862196857e-06,
+      "loss": 0.8138,
+      "step": 3150
+    },
+    {
+      "epoch": 0.449598065020986,
+      "grad_norm": 31.256968405353184,
+      "learning_rate": 3.026576416712376e-06,
+      "loss": 0.9385,
+      "step": 3160
+    },
+    {
+      "epoch": 0.45102084370776124,
+      "grad_norm": 11.182338189217198,
+      "learning_rate": 3.015308163806741e-06,
+      "loss": 0.791,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4524436223945365,
+      "grad_norm": 17.60192918913859,
+      "learning_rate": 3.0040289668172616e-06,
+      "loss": 0.7683,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4538664010813118,
+      "grad_norm": 30.05089385535281,
+      "learning_rate": 2.992739065290845e-06,
+      "loss": 0.803,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4552891797680871,
+      "grad_norm": 13.204065583525782,
+      "learning_rate": 2.9814386990017434e-06,
+      "loss": 0.9288,
+      "step": 3200
+    },
+    {
+      "epoch": 0.45671195845486234,
+      "grad_norm": 13.859245888833582,
+      "learning_rate": 2.970128107946456e-06,
+      "loss": 0.8033,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4581347371416376,
+      "grad_norm": 21.38437791515857,
+      "learning_rate": 2.958807532338637e-06,
+      "loss": 0.7454,
+      "step": 3220
+    },
+    {
+      "epoch": 0.45955751582841287,
+      "grad_norm": 17.25273898261663,
+      "learning_rate": 2.947477212603993e-06,
+      "loss": 0.8496,
+      "step": 3230
+    },
+    {
+      "epoch": 0.4609802945151882,
+      "grad_norm": 17.365610931311284,
+      "learning_rate": 2.9361373893751714e-06,
+      "loss": 0.7646,
+      "step": 3240
+    },
+    {
+      "epoch": 0.46240307320196344,
+      "grad_norm": 22.98057561120779,
+      "learning_rate": 2.92478830348666e-06,
+      "loss": 0.7607,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4638258518887387,
+      "grad_norm": 65.67946753021502,
+      "learning_rate": 2.913430195969662e-06,
+      "loss": 0.7932,
+      "step": 3260
+    },
+    {
+      "epoch": 0.46524863057551397,
+      "grad_norm": 10.472509256267601,
+      "learning_rate": 2.9020633080469832e-06,
+      "loss": 0.6801,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4666714092622892,
+      "grad_norm": 8.728102914060528,
+      "learning_rate": 2.890687881127907e-06,
+      "loss": 0.7957,
+      "step": 3280
+    },
+    {
+      "epoch": 0.46809418794906454,
+      "grad_norm": 36.73652541253815,
+      "learning_rate": 2.8793041568030676e-06,
+      "loss": 0.7548,
+      "step": 3290
+    },
+    {
+      "epoch": 0.4695169666358398,
+      "grad_norm": 15.554148968457284,
+      "learning_rate": 2.867912376839318e-06,
+      "loss": 0.74,
+      "step": 3300
+    },
+    {
+      "epoch": 0.47093974532261507,
+      "grad_norm": 22.587356388546713,
+      "learning_rate": 2.856512783174598e-06,
+      "loss": 0.8052,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4723625240093903,
+      "grad_norm": 12.635529890330547,
+      "learning_rate": 2.845105617912793e-06,
+      "loss": 0.733,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4737853026961656,
+      "grad_norm": 26.954609723737445,
+      "learning_rate": 2.8336911233185934e-06,
+      "loss": 0.8855,
+      "step": 3330
+    },
+    {
+      "epoch": 0.4752080813829409,
+      "grad_norm": 86.32637501223262,
+      "learning_rate": 2.8222695418123508e-06,
+      "loss": 0.8312,
+      "step": 3340
+    },
+    {
+      "epoch": 0.47663086006971617,
+      "grad_norm": 14.733465114880971,
+      "learning_rate": 2.810841115964928e-06,
+      "loss": 0.7234,
+      "step": 3350
+    },
+    {
+      "epoch": 0.47805363875649143,
+      "grad_norm": 19.34741466004809,
+      "learning_rate": 2.7994060884925457e-06,
+      "loss": 0.8293,
+      "step": 3360
+    },
+    {
+      "epoch": 0.4794764174432667,
+      "grad_norm": 7.637471653925083,
+      "learning_rate": 2.787964702251632e-06,
+      "loss": 0.803,
+      "step": 3370
+    },
+    {
+      "epoch": 0.48089919613004195,
+      "grad_norm": 29.511379406914546,
+      "learning_rate": 2.7765172002336605e-06,
+      "loss": 0.7792,
+      "step": 3380
+    },
+    {
+      "epoch": 0.48232197481681727,
+      "grad_norm": 34.4635844809523,
+      "learning_rate": 2.765063825559991e-06,
+      "loss": 0.8358,
+      "step": 3390
+    },
+    {
+      "epoch": 0.48374475350359253,
+      "grad_norm": 13.354157874294469,
+      "learning_rate": 2.7536048214767095e-06,
+      "loss": 0.8608,
+      "step": 3400
+    },
+    {
+      "epoch": 0.4851675321903678,
+      "grad_norm": 27.02117147205228,
+      "learning_rate": 2.742140431349454e-06,
+      "loss": 0.9341,
+      "step": 3410
+    },
+    {
+      "epoch": 0.48659031087714305,
+      "grad_norm": 25.28820023979218,
+      "learning_rate": 2.730670898658255e-06,
+      "loss": 0.7388,
+      "step": 3420
+    },
+    {
+      "epoch": 0.4880130895639183,
+      "grad_norm": 18.21003423824356,
+      "learning_rate": 2.71919646699236e-06,
+      "loss": 0.8631,
+      "step": 3430
+    },
+    {
+      "epoch": 0.48943586825069363,
+      "grad_norm": 21.784602372887232,
+      "learning_rate": 2.7077173800450586e-06,
+      "loss": 0.7476,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4908586469374689,
+      "grad_norm": 10.68299097369915,
+      "learning_rate": 2.6962338816085103e-06,
+      "loss": 0.8056,
+      "step": 3450
+    },
+    {
+      "epoch": 0.49228142562424415,
+      "grad_norm": 28.685154064471742,
+      "learning_rate": 2.684746215568566e-06,
+      "loss": 0.782,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4937042043110194,
+      "grad_norm": 8.053005322889824,
+      "learning_rate": 2.6732546258995875e-06,
+      "loss": 0.7254,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4951269829977947,
+      "grad_norm": 13.318128655622132,
+      "learning_rate": 2.661759356659265e-06,
+      "loss": 0.8543,
+      "step": 3480
+    },
+    {
+      "epoch": 0.49654976168457,
+      "grad_norm": 8.05788198458803,
+      "learning_rate": 2.6502606519834382e-06,
+      "loss": 0.7091,
+      "step": 3490
+    },
+    {
+      "epoch": 0.49797254037134525,
+      "grad_norm": 12.644292212569455,
+      "learning_rate": 2.638758756080906e-06,
+      "loss": 0.8854,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4993953190581205,
+      "grad_norm": 20.711604720983626,
+      "learning_rate": 2.6272539132282413e-06,
+      "loss": 0.7534,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5008180977448958,
+      "grad_norm": 11.915903750704732,
+      "learning_rate": 2.6157463677646094e-06,
+      "loss": 0.799,
+      "step": 3520
+    },
+    {
+      "epoch": 0.502240876431671,
+      "grad_norm": 12.013523847402071,
+      "learning_rate": 2.604236364086567e-06,
+      "loss": 0.7497,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5036636551184464,
+      "grad_norm": 29.372864262754756,
+      "learning_rate": 2.5927241466428822e-06,
+      "loss": 0.7534,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5050864338052216,
+      "grad_norm": 11.132378268449663,
+      "learning_rate": 2.581209959929339e-06,
+      "loss": 0.8214,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5065092124919969,
+      "grad_norm": 60.09517726214323,
+      "learning_rate": 2.5696940484835408e-06,
+      "loss": 0.663,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5079319911787722,
+      "grad_norm": 27.6512864848014,
+      "learning_rate": 2.5581766568797254e-06,
+      "loss": 0.6626,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5093547698655474,
+      "grad_norm": 63.10018193684602,
+      "learning_rate": 2.546658029723565e-06,
+      "loss": 0.8257,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5107775485523227,
+      "grad_norm": 27.616273418999626,
+      "learning_rate": 2.5351384116469703e-06,
+      "loss": 0.7819,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5122003272390979,
+      "grad_norm": 13.470993574295683,
+      "learning_rate": 2.5236180473028983e-06,
+      "loss": 0.7722,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5136231059258732,
+      "grad_norm": 14.99732405273878,
+      "learning_rate": 2.5120971813601553e-06,
+      "loss": 0.7715,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5150458846126486,
+      "grad_norm": 9.656051019345648,
+      "learning_rate": 2.5005760584982015e-06,
+      "loss": 0.6556,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5164686632994238,
+      "grad_norm": 13.74072835052684,
+      "learning_rate": 2.489054923401952e-06,
+      "loss": 0.8078,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5178914419861991,
+      "grad_norm": 25.080939371100442,
+      "learning_rate": 2.477534020756581e-06,
+      "loss": 0.7669,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5193142206729743,
+      "grad_norm": 12.0826969622057,
+      "learning_rate": 2.46601359524233e-06,
+      "loss": 0.7708,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5207369993597496,
+      "grad_norm": 19.16440576327706,
+      "learning_rate": 2.454493891529302e-06,
+      "loss": 0.7616,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5221597780465249,
+      "grad_norm": 7.519923215474658,
+      "learning_rate": 2.442975154272273e-06,
+      "loss": 0.8015,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5235825567333001,
+      "grad_norm": 12.175173720340789,
+      "learning_rate": 2.4314576281054954e-06,
+      "loss": 0.7262,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5250053354200754,
+      "grad_norm": 22.52963523965808,
+      "learning_rate": 2.419941557637496e-06,
+      "loss": 0.7193,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5264281141068506,
+      "grad_norm": 13.643765438313824,
+      "learning_rate": 2.4084271874458885e-06,
+      "loss": 0.7809,
+      "step": 3700
+    },
+    {
+      "epoch": 0.527850892793626,
+      "grad_norm": 20.518020446748704,
+      "learning_rate": 2.396914762072177e-06,
+      "loss": 0.7998,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5292736714804013,
+      "grad_norm": 15.048773713943275,
+      "learning_rate": 2.3854045260165605e-06,
+      "loss": 0.776,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5306964501671765,
+      "grad_norm": 16.141296496771535,
+      "learning_rate": 2.373896723732739e-06,
+      "loss": 0.7137,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5321192288539518,
+      "grad_norm": 14.4660866817992,
+      "learning_rate": 2.3623915996227278e-06,
+      "loss": 0.7506,
+      "step": 3740
+    },
+    {
+      "epoch": 0.533542007540727,
+      "grad_norm": 17.017665454072137,
+      "learning_rate": 2.350889398031664e-06,
+      "loss": 0.7292,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5349647862275023,
+      "grad_norm": 15.28793249418399,
+      "learning_rate": 2.339390363242611e-06,
+      "loss": 0.6674,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5363875649142776,
+      "grad_norm": 12.014033119525148,
+      "learning_rate": 2.327894739471381e-06,
+      "loss": 0.6896,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5378103436010528,
+      "grad_norm": 13.815575430114484,
+      "learning_rate": 2.3164027708613424e-06,
+      "loss": 0.7688,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5392331222878282,
+      "grad_norm": 18.681441273504802,
+      "learning_rate": 2.3049147014782323e-06,
+      "loss": 0.6952,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5406559009746034,
+      "grad_norm": 9.652036109574787,
+      "learning_rate": 2.2934307753049796e-06,
+      "loss": 0.8163,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5420786796613787,
+      "grad_norm": 21.08779764453933,
+      "learning_rate": 2.28195123623652e-06,
+      "loss": 0.8639,
+      "step": 3810
+    },
+    {
+      "epoch": 0.543501458348154,
+      "grad_norm": 14.423681782810991,
+      "learning_rate": 2.270476328074615e-06,
+      "loss": 0.8514,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5449242370349292,
+      "grad_norm": 13.108333913672809,
+      "learning_rate": 2.2590062945226728e-06,
+      "loss": 0.8109,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5463470157217045,
+      "grad_norm": 24.62140487582271,
+      "learning_rate": 2.2475413791805793e-06,
+      "loss": 0.7165,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5477697944084797,
+      "grad_norm": 14.631735750619352,
+      "learning_rate": 2.236081825539518e-06,
+      "loss": 0.841,
+      "step": 3850
+    },
+    {
+      "epoch": 0.549192573095255,
+      "grad_norm": 28.771238476821182,
+      "learning_rate": 2.2246278769768007e-06,
+      "loss": 0.8079,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5506153517820304,
+      "grad_norm": 11.04153065779966,
+      "learning_rate": 2.2131797767507e-06,
+      "loss": 0.6889,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5520381304688056,
+      "grad_norm": 15.667877055627352,
+      "learning_rate": 2.2017377679952797e-06,
+      "loss": 0.8782,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5534609091555809,
+      "grad_norm": 17.915285579619628,
+      "learning_rate": 2.190302093715237e-06,
+      "loss": 0.6891,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5548836878423561,
+      "grad_norm": 8.177091476378212,
+      "learning_rate": 2.1788729967807343e-06,
+      "loss": 0.6736,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5563064665291314,
+      "grad_norm": 14.742743308283796,
+      "learning_rate": 2.1674507199222457e-06,
+      "loss": 0.7508,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5577292452159066,
+      "grad_norm": 11.775787837102852,
+      "learning_rate": 2.156035505725403e-06,
+      "loss": 0.7598,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5591520239026819,
+      "grad_norm": 13.551003247447502,
+      "learning_rate": 2.1446275966258387e-06,
+      "loss": 0.7771,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5605748025894572,
+      "grad_norm": 18.121404299267205,
+      "learning_rate": 2.133227234904041e-06,
+      "loss": 0.7978,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5619975812762324,
+      "grad_norm": 15.58681745499423,
+      "learning_rate": 2.1218346626802093e-06,
+      "loss": 0.685,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5634203599630078,
+      "grad_norm": 22.175071520706346,
+      "learning_rate": 2.110450121909107e-06,
+      "loss": 0.752,
+      "step": 3960
+    },
+    {
+      "epoch": 0.564843138649783,
+      "grad_norm": 15.293504173125388,
+      "learning_rate": 2.0990738543749275e-06,
+      "loss": 0.7029,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5662659173365583,
+      "grad_norm": 8.048785808139954,
+      "learning_rate": 2.087706101686159e-06,
+      "loss": 0.816,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5676886960233336,
+      "grad_norm": 23.32897951386727,
+      "learning_rate": 2.07634710527045e-06,
+      "loss": 0.7524,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5691114747101088,
+      "grad_norm": 11.411010901933299,
+      "learning_rate": 2.0649971063694834e-06,
+      "loss": 0.7571,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5705342533968841,
+      "grad_norm": 23.62461099105402,
+      "learning_rate": 2.053656346033856e-06,
+      "loss": 0.7853,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5719570320836593,
+      "grad_norm": 14.060482033029814,
+      "learning_rate": 2.0423250651179545e-06,
+      "loss": 0.7517,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5733798107704347,
+      "grad_norm": 20.428898918513838,
+      "learning_rate": 2.0310035042748416e-06,
+      "loss": 0.7691,
+      "step": 4030
+    },
+    {
+      "epoch": 0.57480258945721,
+      "grad_norm": 15.764443220749541,
+      "learning_rate": 2.019691903951148e-06,
+      "loss": 0.7349,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5762253681439852,
+      "grad_norm": 17.47815964928467,
+      "learning_rate": 2.0083905043819623e-06,
+      "loss": 0.7858,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5776481468307605,
+      "grad_norm": 66.34722872295633,
+      "learning_rate": 1.997099545585729e-06,
+      "loss": 0.8182,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5790709255175357,
+      "grad_norm": 24.5569265136684,
+      "learning_rate": 1.9858192673591536e-06,
+      "loss": 0.7509,
+      "step": 4070
+    },
+    {
+      "epoch": 0.580493704204311,
+      "grad_norm": 20.773532557907654,
+      "learning_rate": 1.974549909272108e-06,
+      "loss": 0.71,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5819164828910863,
+      "grad_norm": 22.26196471143872,
+      "learning_rate": 1.9632917106625434e-06,
+      "loss": 0.646,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5833392615778615,
+      "grad_norm": 24.53608217109384,
+      "learning_rate": 1.952044910631405e-06,
+      "loss": 0.799,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5847620402646369,
+      "grad_norm": 39.53398159892062,
+      "learning_rate": 1.9408097480375557e-06,
+      "loss": 0.7533,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5861848189514121,
+      "grad_norm": 18.337674444590398,
+      "learning_rate": 1.929586461492707e-06,
+      "loss": 0.6658,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5876075976381874,
+      "grad_norm": 39.29566731972662,
+      "learning_rate": 1.918375289356342e-06,
+      "loss": 0.8095,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5890303763249627,
+      "grad_norm": 21.6126934252539,
+      "learning_rate": 1.9071764697306619e-06,
+      "loss": 0.6963,
+      "step": 4140
+    },
+    {
+      "epoch": 0.5904531550117379,
+      "grad_norm": 17.062659972993746,
+      "learning_rate": 1.895990240455527e-06,
+      "loss": 0.7375,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5918759336985132,
+      "grad_norm": 25.470882444505943,
+      "learning_rate": 1.8848168391034021e-06,
+      "loss": 0.7067,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5932987123852884,
+      "grad_norm": 14.034661319660938,
+      "learning_rate": 1.8736565029743138e-06,
+      "loss": 0.6564,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5947214910720637,
+      "grad_norm": 21.67807376393738,
+      "learning_rate": 1.8625094690908118e-06,
+      "loss": 0.8221,
+      "step": 4180
+    },
+    {
+      "epoch": 0.596144269758839,
+      "grad_norm": 27.935245493154525,
+      "learning_rate": 1.8513759741929321e-06,
+      "loss": 0.7042,
+      "step": 4190
+    },
+    {
+      "epoch": 0.5975670484456143,
+      "grad_norm": 17.1910089435646,
+      "learning_rate": 1.8402562547331704e-06,
+      "loss": 0.7366,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5989898271323896,
+      "grad_norm": 84.46253693971576,
+      "learning_rate": 1.8291505468714615e-06,
+      "loss": 0.7689,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6004126058191648,
+      "grad_norm": 15.71473456214827,
+      "learning_rate": 1.8180590864701628e-06,
+      "loss": 0.8016,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6018353845059401,
+      "grad_norm": 10.447588979119375,
+      "learning_rate": 1.8069821090890427e-06,
+      "loss": 0.7882,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6032581631927154,
+      "grad_norm": 56.25796085812837,
+      "learning_rate": 1.7959198499802832e-06,
+      "loss": 0.6945,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6046809418794906,
+      "grad_norm": 6.285689054380614,
+      "learning_rate": 1.7848725440834783e-06,
+      "loss": 0.8445,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6061037205662659,
+      "grad_norm": 58.716542467523524,
+      "learning_rate": 1.7738404260206487e-06,
+      "loss": 0.7795,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6075264992530411,
+      "grad_norm": 16.08617915947653,
+      "learning_rate": 1.7628237300912542e-06,
+      "loss": 0.6113,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6089492779398165,
+      "grad_norm": 14.376354723346575,
+      "learning_rate": 1.7518226902672214e-06,
+      "loss": 0.827,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6103720566265918,
+      "grad_norm": 17.953652516839444,
+      "learning_rate": 1.7408375401879748e-06,
+      "loss": 0.7564,
+      "step": 4290
+    },
+    {
+      "epoch": 0.611794835313367,
+      "grad_norm": 21.00900096740313,
+      "learning_rate": 1.7298685131554716e-06,
+      "loss": 0.7692,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6132176140001423,
+      "grad_norm": 18.309612056889513,
+      "learning_rate": 1.7189158421292484e-06,
+      "loss": 0.6907,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6146403926869175,
+      "grad_norm": 16.619415766247535,
+      "learning_rate": 1.707979759721476e-06,
+      "loss": 0.6678,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6160631713736928,
+      "grad_norm": 19.779361226835274,
+      "learning_rate": 1.6970604981920174e-06,
+      "loss": 0.7107,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6174859500604681,
+      "grad_norm": 19.91886633881263,
+      "learning_rate": 1.6861582894434902e-06,
+      "loss": 0.713,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6189087287472433,
+      "grad_norm": 130.5283349642206,
+      "learning_rate": 1.675273365016351e-06,
+      "loss": 0.6687,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6203315074340187,
+      "grad_norm": 12.030263554177433,
+      "learning_rate": 1.6644059560839726e-06,
+      "loss": 0.717,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6217542861207939,
+      "grad_norm": 13.146734464615218,
+      "learning_rate": 1.6535562934477307e-06,
+      "loss": 0.721,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6231770648075692,
+      "grad_norm": 13.276782675467445,
+      "learning_rate": 1.6427246075321099e-06,
+      "loss": 0.8076,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6245998434943445,
+      "grad_norm": 40.769061214515816,
+      "learning_rate": 1.6319111283798078e-06,
+      "loss": 0.6972,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6260226221811197,
+      "grad_norm": 12.294066061763774,
+      "learning_rate": 1.6211160856468416e-06,
+      "loss": 0.6645,
+      "step": 4400
+    },
+    {
+      "epoch": 0.627445400867895,
+      "grad_norm": 14.365616243961242,
+      "learning_rate": 1.6103397085976824e-06,
+      "loss": 0.7766,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6288681795546702,
+      "grad_norm": 25.766483277820093,
+      "learning_rate": 1.59958222610038e-06,
+      "loss": 0.7166,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6302909582414455,
+      "grad_norm": 11.836958911544688,
+      "learning_rate": 1.5888438666217012e-06,
+      "loss": 0.7651,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6317137369282209,
+      "grad_norm": 13.9380916036791,
+      "learning_rate": 1.5781248582222787e-06,
+      "loss": 0.68,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6331365156149961,
+      "grad_norm": 18.812370433142473,
+      "learning_rate": 1.5674254285517712e-06,
+      "loss": 0.7157,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6345592943017714,
+      "grad_norm": 29.237875217787078,
+      "learning_rate": 1.5567458048440232e-06,
+      "loss": 0.6739,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6359820729885466,
+      "grad_norm": 10.488421538243566,
+      "learning_rate": 1.54608621391224e-06,
+      "loss": 0.7193,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6374048516753219,
+      "grad_norm": 11.137577579014978,
+      "learning_rate": 1.5354468821441762e-06,
+      "loss": 0.7398,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6388276303620972,
+      "grad_norm": 14.916947670390048,
+      "learning_rate": 1.5248280354973195e-06,
+      "loss": 0.8009,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6402504090488724,
+      "grad_norm": 9.492480685750538,
+      "learning_rate": 1.514229899494099e-06,
+      "loss": 0.6761,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6416731877356477,
+      "grad_norm": 17.176525722076306,
+      "learning_rate": 1.50365269921709e-06,
+      "loss": 0.6576,
+      "step": 4510
+    },
+    {
+      "epoch": 0.643095966422423,
+      "grad_norm": 9.346985316225243,
+      "learning_rate": 1.4930966593042362e-06,
+      "loss": 0.8006,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6445187451091983,
+      "grad_norm": 14.91630735957912,
+      "learning_rate": 1.4825620039440812e-06,
+      "loss": 0.8409,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6459415237959736,
+      "grad_norm": 11.473298576346068,
+      "learning_rate": 1.4720489568710016e-06,
+      "loss": 0.7263,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6473643024827488,
+      "grad_norm": 15.939839366948728,
+      "learning_rate": 1.4615577413604604e-06,
+      "loss": 0.6488,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6487870811695241,
+      "grad_norm": 16.191145799664508,
+      "learning_rate": 1.4510885802242637e-06,
+      "loss": 0.7685,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6502098598562993,
+      "grad_norm": 18.698577322891694,
+      "learning_rate": 1.440641695805825e-06,
+      "loss": 0.668,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6516326385430746,
+      "grad_norm": 13.016791806898487,
+      "learning_rate": 1.4302173099754506e-06,
+      "loss": 0.7125,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6530554172298499,
+      "grad_norm": 40.70752668906216,
+      "learning_rate": 1.4198156441256194e-06,
+      "loss": 0.7818,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6544781959166251,
+      "grad_norm": 7.281754729573183,
+      "learning_rate": 1.4094369191662876e-06,
+      "loss": 0.7283,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6559009746034005,
+      "grad_norm": 21.2940004124432,
+      "learning_rate": 1.3990813555201916e-06,
+      "loss": 0.7082,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6573237532901757,
+      "grad_norm": 27.176553938994964,
+      "learning_rate": 1.38874917311817e-06,
+      "loss": 0.6501,
+      "step": 4620
+    },
+    {
+      "epoch": 0.658746531976951,
+      "grad_norm": 13.520934938840828,
+      "learning_rate": 1.3784405913944948e-06,
+      "loss": 0.7522,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6601693106637263,
+      "grad_norm": 27.795368348387175,
+      "learning_rate": 1.368155829282204e-06,
+      "loss": 0.6796,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6615920893505015,
+      "grad_norm": 13.139909160482343,
+      "learning_rate": 1.3578951052084566e-06,
+      "loss": 0.6848,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6630148680372768,
+      "grad_norm": 20.60306083940586,
+      "learning_rate": 1.347658637089896e-06,
+      "loss": 0.7656,
+      "step": 4660
+    },
+    {
+      "epoch": 0.664437646724052,
+      "grad_norm": 14.25958816780152,
+      "learning_rate": 1.3374466423280185e-06,
+      "loss": 0.6432,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6658604254108274,
+      "grad_norm": 19.038734308507603,
+      "learning_rate": 1.3272593378045534e-06,
+      "loss": 0.8042,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6672832040976027,
+      "grad_norm": 29.21367724067394,
+      "learning_rate": 1.3170969398768649e-06,
+      "loss": 0.6578,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6687059827843779,
+      "grad_norm": 533.8160870471642,
+      "learning_rate": 1.306959664373349e-06,
+      "loss": 0.7387,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6701287614711532,
+      "grad_norm": 14.771200737749313,
+      "learning_rate": 1.296847726588853e-06,
+      "loss": 0.6526,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6715515401579284,
+      "grad_norm": 16.43963547785031,
+      "learning_rate": 1.2867613412801055e-06,
+      "loss": 0.7479,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6729743188447037,
+      "grad_norm": 12.308142199707067,
+      "learning_rate": 1.2767007226611533e-06,
+      "loss": 0.7722,
+      "step": 4730
+    },
+    {
+      "epoch": 0.674397097531479,
+      "grad_norm": 21.747552443599744,
+      "learning_rate": 1.2666660843988088e-06,
+      "loss": 0.7633,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6758198762182542,
+      "grad_norm": 9.250080675840156,
+      "learning_rate": 1.2566576396081188e-06,
+      "loss": 0.6876,
+      "step": 4750
+    },
+    {
+      "epoch": 0.6772426549050296,
+      "grad_norm": 17.884494630302683,
+      "learning_rate": 1.2466756008478303e-06,
+      "loss": 0.6994,
+      "step": 4760
+    },
+    {
+      "epoch": 0.6786654335918048,
+      "grad_norm": 9.122887241170309,
+      "learning_rate": 1.236720180115886e-06,
+      "loss": 0.6575,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6800882122785801,
+      "grad_norm": 7.956747557348674,
+      "learning_rate": 1.2267915888449114e-06,
+      "loss": 0.7269,
+      "step": 4780
+    },
+    {
+      "epoch": 0.6815109909653554,
+      "grad_norm": 20.631525051921074,
+      "learning_rate": 1.216890037897733e-06,
+      "loss": 0.75,
+      "step": 4790
+    },
+    {
+      "epoch": 0.6829337696521306,
+      "grad_norm": 35.036205257844856,
+      "learning_rate": 1.207015737562896e-06,
+      "loss": 0.7037,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6843565483389059,
+      "grad_norm": 19.008308176713022,
+      "learning_rate": 1.1971688975501971e-06,
+      "loss": 0.7218,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6857793270256811,
+      "grad_norm": 7.128767920750745,
+      "learning_rate": 1.1873497269862327e-06,
+      "loss": 0.6508,
+      "step": 4820
+    },
+    {
+      "epoch": 0.6872021057124564,
+      "grad_norm": 9.501177883754682,
+      "learning_rate": 1.1775584344099586e-06,
+      "loss": 0.7482,
+      "step": 4830
+    },
+    {
+      "epoch": 0.6886248843992318,
+      "grad_norm": 14.480692474454747,
+      "learning_rate": 1.1677952277682597e-06,
+      "loss": 0.6636,
+      "step": 4840
+    },
+    {
+      "epoch": 0.690047663086007,
+      "grad_norm": 8.260427118275636,
+      "learning_rate": 1.1580603144115301e-06,
+      "loss": 0.7237,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6914704417727823,
+      "grad_norm": 17.5013994162,
+      "learning_rate": 1.1483539010892777e-06,
+      "loss": 0.6309,
+      "step": 4860
+    },
+    {
+      "epoch": 0.6928932204595575,
+      "grad_norm": 13.22124517933893,
+      "learning_rate": 1.1386761939457247e-06,
+      "loss": 0.6407,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6943159991463328,
+      "grad_norm": 25.784036778389154,
+      "learning_rate": 1.1290273985154333e-06,
+      "loss": 0.7658,
+      "step": 4880
+    },
+    {
+      "epoch": 0.6957387778331081,
+      "grad_norm": 31.750375019656428,
+      "learning_rate": 1.1194077197189427e-06,
+      "loss": 0.6312,
+      "step": 4890
+    },
+    {
+      "epoch": 0.6971615565198833,
+      "grad_norm": 13.27769247274,
+      "learning_rate": 1.1098173618584147e-06,
+      "loss": 0.656,
+      "step": 4900
+    },
+    {
+      "epoch": 0.6985843352066586,
+      "grad_norm": 24.745992727476718,
+      "learning_rate": 1.1002565286132914e-06,
+      "loss": 0.6578,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7000071138934338,
+      "grad_norm": 6.572338606115534,
+      "learning_rate": 1.0907254230359777e-06,
+      "loss": 0.6173,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7014298925802092,
+      "grad_norm": 21.375498414332217,
+      "learning_rate": 1.0812242475475192e-06,
+      "loss": 0.6443,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7028526712669845,
+      "grad_norm": 18.135153621580898,
+      "learning_rate": 1.071753203933313e-06,
+      "loss": 0.733,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7042754499537597,
+      "grad_norm": 14.042627759924816,
+      "learning_rate": 1.0623124933388126e-06,
+      "loss": 0.6542,
+      "step": 4950
+    },
+    {
+      "epoch": 0.705698228640535,
+      "grad_norm": 30.16915992497821,
+      "learning_rate": 1.0529023162652635e-06,
+      "loss": 0.7769,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7071210073273102,
+      "grad_norm": 14.388444772922258,
+      "learning_rate": 1.0435228725654426e-06,
+      "loss": 0.6462,
+      "step": 4970
+    },
+    {
+      "epoch": 0.7085437860140855,
+      "grad_norm": 18.03414525268103,
+      "learning_rate": 1.0341743614394105e-06,
+      "loss": 0.729,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7099665647008608,
+      "grad_norm": 13.85227892598916,
+      "learning_rate": 1.0248569814302864e-06,
+      "loss": 0.6415,
+      "step": 4990
+    },
+    {
+      "epoch": 0.711389343387636,
+      "grad_norm": 15.215365219685497,
+      "learning_rate": 1.015570930420027e-06,
+      "loss": 0.6695,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7128121220744114,
+      "grad_norm": 97.40345906478952,
+      "learning_rate": 1.0063164056252254e-06,
+      "loss": 0.6863,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7142349007611866,
+      "grad_norm": 16.79920759950203,
+      "learning_rate": 9.97093603592924e-07,
+      "loss": 0.6956,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7156576794479619,
+      "grad_norm": 14.27314471502695,
+      "learning_rate": 9.879027201964392e-07,
+      "loss": 0.7643,
+      "step": 5030
+    },
+    {
+      "epoch": 0.7170804581347371,
+      "grad_norm": 34.06257052430994,
+      "learning_rate": 9.787439506312016e-07,
+      "loss": 0.6817,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7185032368215124,
+      "grad_norm": 44.832955767998286,
+      "learning_rate": 9.696174894106081e-07,
+      "loss": 0.7809,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7199260155082877,
+      "grad_norm": 16.245080415300535,
+      "learning_rate": 9.605235303618935e-07,
+      "loss": 0.727,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7213487941950629,
+      "grad_norm": 14.111285705311632,
+      "learning_rate": 9.514622666220158e-07,
+      "loss": 0.6188,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7227715728818382,
+      "grad_norm": 15.162572043299535,
+      "learning_rate": 9.424338906335486e-07,
+      "loss": 0.7122,
+      "step": 5080
+    },
+    {
+      "epoch": 0.7241943515686134,
+      "grad_norm": 27.668155036946754,
+      "learning_rate": 9.334385941406002e-07,
+      "loss": 0.7043,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7256171302553888,
+      "grad_norm": 11.64701587775779,
+      "learning_rate": 9.244765681847379e-07,
+      "loss": 0.6523,
+      "step": 5100
+    },
+    {
+      "epoch": 0.7270399089421641,
+      "grad_norm": 17.376356821899527,
+      "learning_rate": 9.155480031009304e-07,
+      "loss": 0.6972,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7284626876289393,
+      "grad_norm": 8.690171383763024,
+      "learning_rate": 9.066530885135055e-07,
+      "loss": 0.7344,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7298854663157146,
+      "grad_norm": 22.08471670413809,
+      "learning_rate": 8.977920133321266e-07,
+      "loss": 0.7304,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7313082450024898,
+      "grad_norm": 51.96286186100988,
+      "learning_rate": 8.889649657477773e-07,
+      "loss": 0.6957,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7327310236892651,
+      "grad_norm": 14.379879755090343,
+      "learning_rate": 8.801721332287619e-07,
+      "loss": 0.7775,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7341538023760404,
+      "grad_norm": 26.855334316137547,
+      "learning_rate": 8.714137025167319e-07,
+      "loss": 0.714,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7355765810628156,
+      "grad_norm": 16.00017490250412,
+      "learning_rate": 8.626898596227118e-07,
+      "loss": 0.652,
+      "step": 5170
+    },
+    {
+      "epoch": 0.736999359749591,
+      "grad_norm": 8.432034143264096,
+      "learning_rate": 8.540007898231525e-07,
+      "loss": 0.8362,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7384221384363662,
+      "grad_norm": 22.25669162158987,
+      "learning_rate": 8.453466776559974e-07,
+      "loss": 0.6658,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7398449171231415,
+      "grad_norm": 8.791913538046265,
+      "learning_rate": 8.367277069167612e-07,
+      "loss": 0.7508,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7412676958099168,
+      "grad_norm": 19.57210800521164,
+      "learning_rate": 8.281440606546278e-07,
+      "loss": 0.7802,
+      "step": 5210
+    },
+    {
+      "epoch": 0.742690474496692,
+      "grad_norm": 11.354804669570045,
+      "learning_rate": 8.195959211685606e-07,
+      "loss": 0.6344,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7441132531834673,
+      "grad_norm": 15.17145858767825,
+      "learning_rate": 8.110834700034315e-07,
+      "loss": 0.7363,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7455360318702425,
+      "grad_norm": 29.579246308700657,
+      "learning_rate": 8.026068879461693e-07,
+      "loss": 0.6516,
+      "step": 5240
+    },
+    {
+      "epoch": 0.7469588105570178,
+      "grad_norm": 18.39668533555811,
+      "learning_rate": 7.941663550219134e-07,
+      "loss": 0.6841,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7483815892437932,
+      "grad_norm": 22.82729737786706,
+      "learning_rate": 7.857620504901953e-07,
+      "loss": 0.68,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7498043679305684,
+      "grad_norm": 6.6003652423985155,
+      "learning_rate": 7.773941528411316e-07,
+      "loss": 0.655,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7512271466173437,
+      "grad_norm": 22.211334230492756,
+      "learning_rate": 7.690628397916291e-07,
+      "loss": 0.7384,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7526499253041189,
+      "grad_norm": 12.074461866500236,
+      "learning_rate": 7.607682882816136e-07,
+      "loss": 0.6816,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7540727039908942,
+      "grad_norm": 16.231173521043722,
+      "learning_rate": 7.525106744702725e-07,
+      "loss": 0.6975,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7554954826776695,
+      "grad_norm": 31.721613985668142,
+      "learning_rate": 7.442901737323133e-07,
+      "loss": 0.681,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7569182613644447,
+      "grad_norm": 17.681404615301002,
+      "learning_rate": 7.361069606542354e-07,
+      "loss": 0.6685,
+      "step": 5320
+    },
+    {
+      "epoch": 0.75834104005122,
+      "grad_norm": 9.124774910267016,
+      "learning_rate": 7.279612090306285e-07,
+      "loss": 0.5928,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7597638187379953,
+      "grad_norm": 30.734994130943175,
+      "learning_rate": 7.198530918604754e-07,
+      "loss": 0.709,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7611865974247706,
+      "grad_norm": 11.112472348569415,
+      "learning_rate": 7.117827813434813e-07,
+      "loss": 0.6674,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7626093761115459,
+      "grad_norm": 16.480693812459887,
+      "learning_rate": 7.037504488764174e-07,
+      "loss": 0.6491,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7640321547983211,
+      "grad_norm": 12.34032417959078,
+      "learning_rate": 6.957562650494781e-07,
+      "loss": 0.616,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7654549334850964,
+      "grad_norm": 12.791383068851031,
+      "learning_rate": 6.878003996426605e-07,
+      "loss": 0.7438,
+      "step": 5380
+    },
+    {
+      "epoch": 0.7668777121718716,
+      "grad_norm": 37.40607230940538,
+      "learning_rate": 6.798830216221552e-07,
+      "loss": 0.6666,
+      "step": 5390
+    },
+    {
+      "epoch": 0.7683004908586469,
+      "grad_norm": 23.140755286118655,
+      "learning_rate": 6.720042991367614e-07,
+      "loss": 0.7311,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7697232695454223,
+      "grad_norm": 20.51227690551173,
+      "learning_rate": 6.641643995143152e-07,
+      "loss": 0.7196,
+      "step": 5410
+    },
+    {
+      "epoch": 0.7711460482321975,
+      "grad_norm": 11.565002812640744,
+      "learning_rate": 6.563634892581324e-07,
+      "loss": 0.6755,
+      "step": 5420
+    },
+    {
+      "epoch": 0.7725688269189728,
+      "grad_norm": 20.44574240987166,
+      "learning_rate": 6.486017340434777e-07,
+      "loss": 0.6708,
+      "step": 5430
+    },
+    {
+      "epoch": 0.773991605605748,
+      "grad_norm": 23.82537630461409,
+      "learning_rate": 6.408792987140433e-07,
+      "loss": 0.6907,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7754143842925233,
+      "grad_norm": 24.844791674029324,
+      "learning_rate": 6.331963472784458e-07,
+      "loss": 0.6266,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7768371629792986,
+      "grad_norm": 34.51982306880169,
+      "learning_rate": 6.255530429067463e-07,
+      "loss": 0.68,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7782599416660738,
+      "grad_norm": 22.6817440592934,
+      "learning_rate": 6.179495479269848e-07,
+      "loss": 0.5493,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7796827203528491,
+      "grad_norm": 19.193324207599662,
+      "learning_rate": 6.103860238217316e-07,
+      "loss": 0.7456,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7811054990396243,
+      "grad_norm": 31.92374383346149,
+      "learning_rate": 6.028626312246563e-07,
+      "loss": 0.6894,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7825282777263997,
+      "grad_norm": 14.402098783610006,
+      "learning_rate": 5.953795299171208e-07,
+      "loss": 0.6572,
+      "step": 5500
+    },
+    {
+      "epoch": 0.783951056413175,
+      "grad_norm": 16.58225414656111,
+      "learning_rate": 5.879368788247805e-07,
+      "loss": 0.6419,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7853738350999502,
+      "grad_norm": 15.51842210754784,
+      "learning_rate": 5.805348360142119e-07,
+      "loss": 0.6338,
+      "step": 5520
+    },
+    {
+      "epoch": 0.7867966137867255,
+      "grad_norm": 18.454262994429182,
+      "learning_rate": 5.731735586895562e-07,
+      "loss": 0.6657,
+      "step": 5530
+    },
+    {
+      "epoch": 0.7882193924735007,
+      "grad_norm": 19.16725689823867,
+      "learning_rate": 5.658532031891798e-07,
+      "loss": 0.5431,
+      "step": 5540
+    },
+    {
+      "epoch": 0.789642171160276,
+      "grad_norm": 14.314115828548886,
+      "learning_rate": 5.585739249823532e-07,
+      "loss": 0.7796,
+      "step": 5550
+    },
+    {
+      "epoch": 0.7910649498470513,
+      "grad_norm": 18.83332788207389,
+      "learning_rate": 5.5133587866595e-07,
+      "loss": 0.7011,
+      "step": 5560
+    },
+    {
+      "epoch": 0.7924877285338265,
+      "grad_norm": 6.571592737822653,
+      "learning_rate": 5.441392179611623e-07,
+      "loss": 0.6035,
+      "step": 5570
+    },
+    {
+      "epoch": 0.7939105072206019,
+      "grad_norm": 23.07747991496131,
+      "learning_rate": 5.369840957102399e-07,
+      "loss": 0.6608,
+      "step": 5580
+    },
+    {
+      "epoch": 0.7953332859073771,
+      "grad_norm": 17.394620247817357,
+      "learning_rate": 5.298706638732373e-07,
+      "loss": 0.5769,
+      "step": 5590
+    },
+    {
+      "epoch": 0.7967560645941524,
+      "grad_norm": 37.99269008869083,
+      "learning_rate": 5.227990735247942e-07,
+      "loss": 0.7183,
+      "step": 5600
+    },
+    {
+      "epoch": 0.7981788432809277,
+      "grad_norm": 14.27327658796223,
+      "learning_rate": 5.157694748509218e-07,
+      "loss": 0.6888,
+      "step": 5610
+    },
+    {
+      "epoch": 0.7996016219677029,
+      "grad_norm": 25.80673651224746,
+      "learning_rate": 5.087820171458149e-07,
+      "loss": 0.6342,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8010244006544782,
+      "grad_norm": 314.11857800515276,
+      "learning_rate": 5.018368488086808e-07,
+      "loss": 0.7554,
+      "step": 5630
+    },
+    {
+      "epoch": 0.8024471793412534,
+      "grad_norm": 14.451508569333676,
+      "learning_rate": 4.949341173405863e-07,
+      "loss": 0.6255,
+      "step": 5640
+    },
+    {
+      "epoch": 0.8038699580280287,
+      "grad_norm": 27.119138215469455,
+      "learning_rate": 4.880739693413294e-07,
+      "loss": 0.7462,
+      "step": 5650
+    },
+    {
+      "epoch": 0.8052927367148041,
+      "grad_norm": 19.870097640286513,
+      "learning_rate": 4.812565505063207e-07,
+      "loss": 0.6811,
+      "step": 5660
+    },
+    {
+      "epoch": 0.8067155154015793,
+      "grad_norm": 10.389938860504465,
+      "learning_rate": 4.7448200562349197e-07,
+      "loss": 0.6477,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8081382940883546,
+      "grad_norm": 13.92659812488069,
+      "learning_rate": 4.677504785702219e-07,
+      "loss": 0.6891,
+      "step": 5680
+    },
+    {
+      "epoch": 0.8095610727751298,
+      "grad_norm": 32.85051628694543,
+      "learning_rate": 4.610621123102768e-07,
+      "loss": 0.6472,
+      "step": 5690
+    },
+    {
+      "epoch": 0.8109838514619051,
+      "grad_norm": 7.863872935498879,
+      "learning_rate": 4.544170488907776e-07,
+      "loss": 0.6665,
+      "step": 5700
+    },
+    {
+      "epoch": 0.8124066301486804,
+      "grad_norm": 22.860753676803004,
+      "learning_rate": 4.478154294391826e-07,
+      "loss": 0.6891,
+      "step": 5710
+    },
+    {
+      "epoch": 0.8138294088354556,
+      "grad_norm": 11.732849180082756,
+      "learning_rate": 4.412573941602902e-07,
+      "loss": 0.6427,
+      "step": 5720
+    },
+    {
+      "epoch": 0.8152521875222309,
+      "grad_norm": 17.869770041617677,
+      "learning_rate": 4.3474308233325823e-07,
+      "loss": 0.7251,
+      "step": 5730
+    },
+    {
+      "epoch": 0.8166749662090061,
+      "grad_norm": 19.42876313677897,
+      "learning_rate": 4.2827263230865195e-07,
+      "loss": 0.6646,
+      "step": 5740
+    },
+    {
+      "epoch": 0.8180977448957815,
+      "grad_norm": 8.590975505290642,
+      "learning_rate": 4.218461815054994e-07,
+      "loss": 0.7297,
+      "step": 5750
+    },
+    {
+      "epoch": 0.8195205235825568,
+      "grad_norm": 19.349716561967885,
+      "learning_rate": 4.154638664083768e-07,
+      "loss": 0.6522,
+      "step": 5760
+    },
+    {
+      "epoch": 0.820943302269332,
+      "grad_norm": 9.483732611169318,
+      "learning_rate": 4.091258225645095e-07,
+      "loss": 0.5925,
+      "step": 5770
+    },
+    {
+      "epoch": 0.8223660809561073,
+      "grad_norm": 17.221316669496034,
+      "learning_rate": 4.0283218458089164e-07,
+      "loss": 0.6834,
+      "step": 5780
+    },
+    {
+      "epoch": 0.8237888596428825,
+      "grad_norm": 16.143634622370726,
+      "learning_rate": 3.965830861214301e-07,
+      "loss": 0.7427,
+      "step": 5790
+    },
+    {
+      "epoch": 0.8252116383296578,
+      "grad_norm": 15.675226060931111,
+      "learning_rate": 3.903786599041018e-07,
+      "loss": 0.7599,
+      "step": 5800
+    },
+    {
+      "epoch": 0.8266344170164331,
+      "grad_norm": 25.996033801532082,
+      "learning_rate": 3.8421903769813777e-07,
+      "loss": 0.7081,
+      "step": 5810
+    },
+    {
+      "epoch": 0.8280571957032083,
+      "grad_norm": 11.960706714057824,
+      "learning_rate": 3.781043503212259e-07,
+      "loss": 0.6384,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8294799743899837,
+      "grad_norm": 26.473024863638063,
+      "learning_rate": 3.720347276367281e-07,
+      "loss": 0.6261,
+      "step": 5830
+    },
+    {
+      "epoch": 0.8309027530767589,
+      "grad_norm": 33.97170669671055,
+      "learning_rate": 3.6601029855092754e-07,
+      "loss": 0.6457,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8323255317635342,
+      "grad_norm": 13.60176717003026,
+      "learning_rate": 3.600311910102877e-07,
+      "loss": 0.6363,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8337483104503095,
+      "grad_norm": 15.646223876804923,
+      "learning_rate": 3.5409753199873524e-07,
+      "loss": 0.6772,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8351710891370847,
+      "grad_norm": 14.887721160003261,
+      "learning_rate": 3.4820944753496316e-07,
+      "loss": 0.6259,
+      "step": 5870
+    },
+    {
+      "epoch": 0.83659386782386,
+      "grad_norm": 8.180457549268645,
+      "learning_rate": 3.4236706266975673e-07,
+      "loss": 0.7065,
+      "step": 5880
+    },
+    {
+      "epoch": 0.8380166465106352,
+      "grad_norm": 20.131585730116342,
+      "learning_rate": 3.3657050148333507e-07,
+      "loss": 0.7261,
+      "step": 5890
+    },
+    {
+      "epoch": 0.8394394251974105,
+      "grad_norm": 15.01194640883317,
+      "learning_rate": 3.308198870827159e-07,
+      "loss": 0.7131,
+      "step": 5900
+    },
+    {
+      "epoch": 0.8408622038841859,
+      "grad_norm": 25.50169407499349,
+      "learning_rate": 3.2511534159910347e-07,
+      "loss": 0.7371,
+      "step": 5910
+    },
+    {
+      "epoch": 0.8422849825709611,
+      "grad_norm": 8.932934403329037,
+      "learning_rate": 3.1945698618529195e-07,
+      "loss": 0.604,
+      "step": 5920
+    },
+    {
+      "epoch": 0.8437077612577364,
+      "grad_norm": 21.804238920752727,
+      "learning_rate": 3.138449410130931e-07,
+      "loss": 0.661,
+      "step": 5930
+    },
+    {
+      "epoch": 0.8451305399445116,
+      "grad_norm": 31.003829882682634,
+      "learning_rate": 3.082793252707861e-07,
+      "loss": 0.6599,
+      "step": 5940
+    },
+    {
+      "epoch": 0.8465533186312869,
+      "grad_norm": 13.798244734908524,
+      "learning_rate": 3.027602571605845e-07,
+      "loss": 0.7096,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8479760973180622,
+      "grad_norm": 21.048536525617802,
+      "learning_rate": 2.972878538961241e-07,
+      "loss": 0.6568,
+      "step": 5960
+    },
+    {
+      "epoch": 0.8493988760048374,
+      "grad_norm": 35.14605252696549,
+      "learning_rate": 2.9186223169997835e-07,
+      "loss": 0.6926,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8508216546916127,
+      "grad_norm": 6.287182276066726,
+      "learning_rate": 2.8648350580118445e-07,
+      "loss": 0.677,
+      "step": 5980
+    },
+    {
+      "epoch": 0.852244433378388,
+      "grad_norm": 21.352684241357913,
+      "learning_rate": 2.811517904328012e-07,
+      "loss": 0.6748,
+      "step": 5990
+    },
+    {
+      "epoch": 0.8536672120651633,
+      "grad_norm": 8.739882575649565,
+      "learning_rate": 2.758671988294784e-07,
+      "loss": 0.6458,
+      "step": 6000
+    },
+    {
+      "epoch": 0.8550899907519386,
+      "grad_norm": 49.48961762778536,
+      "learning_rate": 2.7062984322505564e-07,
+      "loss": 0.5956,
+      "step": 6010
+    },
+    {
+      "epoch": 0.8565127694387138,
+      "grad_norm": 9.765359265138134,
+      "learning_rate": 2.654398348501777e-07,
+      "loss": 0.6783,
+      "step": 6020
+    },
+    {
+      "epoch": 0.8579355481254891,
+      "grad_norm": 10.353183038698942,
+      "learning_rate": 2.6029728392993e-07,
+      "loss": 0.7489,
+      "step": 6030
+    },
+    {
+      "epoch": 0.8593583268122643,
+      "grad_norm": 9.492744746937786,
+      "learning_rate": 2.552022996814998e-07,
+      "loss": 0.6283,
+      "step": 6040
+    },
+    {
+      "epoch": 0.8607811054990396,
+      "grad_norm": 15.573884263724349,
+      "learning_rate": 2.5015499031185746e-07,
+      "loss": 0.7418,
+      "step": 6050
+    },
+    {
+      "epoch": 0.862203884185815,
+      "grad_norm": 55.47118233050117,
+      "learning_rate": 2.451554630154565e-07,
+      "loss": 0.6575,
+      "step": 6060
+    },
+    {
+      "epoch": 0.8636266628725902,
+      "grad_norm": 11.171500931865225,
+      "learning_rate": 2.4020382397195675e-07,
+      "loss": 0.6171,
+      "step": 6070
+    },
+    {
+      "epoch": 0.8650494415593655,
+      "grad_norm": 10.465530747854013,
+      "learning_rate": 2.353001783439715e-07,
+      "loss": 0.7969,
+      "step": 6080
+    },
+    {
+      "epoch": 0.8664722202461407,
+      "grad_norm": 17.91155449474328,
+      "learning_rate": 2.30444630274832e-07,
+      "loss": 0.6153,
+      "step": 6090
+    },
+    {
+      "epoch": 0.867894998932916,
+      "grad_norm": 14.885236298278624,
+      "learning_rate": 2.2563728288637648e-07,
+      "loss": 0.7035,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8693177776196913,
+      "grad_norm": 18.276684017399496,
+      "learning_rate": 2.208782382767599e-07,
+      "loss": 0.6747,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8707405563064665,
+      "grad_norm": 10.293023049649177,
+      "learning_rate": 2.1616759751828697e-07,
+      "loss": 0.7219,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8721633349932418,
+      "grad_norm": 14.965682612306166,
+      "learning_rate": 2.1150546065526245e-07,
+      "loss": 0.7804,
+      "step": 6130
+    },
+    {
+      "epoch": 0.873586113680017,
+      "grad_norm": 29.038048950021082,
+      "learning_rate": 2.0689192670186957e-07,
+      "loss": 0.6408,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8750088923667924,
+      "grad_norm": 15.060073917261208,
+      "learning_rate": 2.0232709364006482e-07,
+      "loss": 0.5891,
+      "step": 6150
+    },
+    {
+      "epoch": 0.8764316710535676,
+      "grad_norm": 21.41557776637373,
+      "learning_rate": 1.978110584174994e-07,
+      "loss": 0.6116,
+      "step": 6160
+    },
+    {
+      "epoch": 0.8778544497403429,
+      "grad_norm": 7.518014180571353,
+      "learning_rate": 1.933439169454579e-07,
+      "loss": 0.5941,
+      "step": 6170
+    },
+    {
+      "epoch": 0.8792772284271182,
+      "grad_norm": 11.884159235216817,
+      "learning_rate": 1.8892576409682256e-07,
+      "loss": 0.7353,
+      "step": 6180
+    },
+    {
+      "epoch": 0.8807000071138934,
+      "grad_norm": 21.83174021803315,
+      "learning_rate": 1.8455669370405882e-07,
+      "loss": 0.6475,
+      "step": 6190
+    },
+    {
+      "epoch": 0.8821227858006687,
+      "grad_norm": 14.057248171008613,
+      "learning_rate": 1.8023679855722093e-07,
+      "loss": 0.7441,
+      "step": 6200
+    },
+    {
+      "epoch": 0.8835455644874439,
+      "grad_norm": 21.79237041631171,
+      "learning_rate": 1.7596617040198327e-07,
+      "loss": 0.7151,
+      "step": 6210
+    },
+    {
+      "epoch": 0.8849683431742192,
+      "grad_norm": 10.609700883252497,
+      "learning_rate": 1.7174489993768917e-07,
+      "loss": 0.6349,
+      "step": 6220
+    },
+    {
+      "epoch": 0.8863911218609946,
+      "grad_norm": 13.311107416979928,
+      "learning_rate": 1.6757307681542821e-07,
+      "loss": 0.6425,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8878139005477698,
+      "grad_norm": 34.676737788107566,
+      "learning_rate": 1.6345078963612897e-07,
+      "loss": 0.6191,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8892366792345451,
+      "grad_norm": 10.877524165504067,
+      "learning_rate": 1.59378125948679e-07,
+      "loss": 0.6359,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8906594579213203,
+      "grad_norm": 169.81539374163754,
+      "learning_rate": 1.553551722480659e-07,
+      "loss": 0.6926,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8920822366080956,
+      "grad_norm": 33.283503418457634,
+      "learning_rate": 1.5138201397353834e-07,
+      "loss": 0.7244,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8935050152948709,
+      "grad_norm": 16.66512043307651,
+      "learning_rate": 1.4745873550679274e-07,
+      "loss": 0.7099,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8949277939816461,
+      "grad_norm": 39.05716141566033,
+      "learning_rate": 1.4358542017018222e-07,
+      "loss": 0.6016,
+      "step": 6290
+    },
+    {
+      "epoch": 0.8963505726684214,
+      "grad_norm": 11.399154977290856,
+      "learning_rate": 1.397621502249452e-07,
+      "loss": 0.7192,
+      "step": 6300
+    },
+    {
+      "epoch": 0.8977733513551966,
+      "grad_norm": 15.074253686543178,
+      "learning_rate": 1.3598900686945916e-07,
+      "loss": 0.6877,
+      "step": 6310
+    },
+    {
+      "epoch": 0.899196130041972,
+      "grad_norm": 42.56082012662866,
+      "learning_rate": 1.3226607023751642e-07,
+      "loss": 0.5992,
+      "step": 6320
+    },
+    {
+      "epoch": 0.9006189087287473,
+      "grad_norm": 26.049955980737074,
+      "learning_rate": 1.2859341939662178e-07,
+      "loss": 0.7737,
+      "step": 6330
+    },
+    {
+      "epoch": 0.9020416874155225,
+      "grad_norm": 30.559789796724967,
+      "learning_rate": 1.2497113234631238e-07,
+      "loss": 0.6726,
+      "step": 6340
+    },
+    {
+      "epoch": 0.9034644661022978,
+      "grad_norm": 11.508415485790165,
+      "learning_rate": 1.2139928601650393e-07,
+      "loss": 0.5407,
+      "step": 6350
+    },
+    {
+      "epoch": 0.904887244789073,
+      "grad_norm": 22.16281781432751,
+      "learning_rate": 1.1787795626585441e-07,
+      "loss": 0.6747,
+      "step": 6360
+    },
+    {
+      "epoch": 0.9063100234758483,
+      "grad_norm": 8.512128854917334,
+      "learning_rate": 1.1440721788015308e-07,
+      "loss": 0.5865,
+      "step": 6370
+    },
+    {
+      "epoch": 0.9077328021626236,
+      "grad_norm": 9.064303945494332,
+      "learning_rate": 1.1098714457073417e-07,
+      "loss": 0.7327,
+      "step": 6380
+    },
+    {
+      "epoch": 0.9091555808493988,
+      "grad_norm": 18.62534412436195,
+      "learning_rate": 1.0761780897290852e-07,
+      "loss": 0.683,
+      "step": 6390
+    },
+    {
+      "epoch": 0.9105783595361742,
+      "grad_norm": 15.04597150265621,
+      "learning_rate": 1.0429928264442362e-07,
+      "loss": 0.6818,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9120011382229494,
+      "grad_norm": 17.9224272604521,
+      "learning_rate": 1.0103163606394239e-07,
+      "loss": 0.6399,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9134239169097247,
+      "grad_norm": 15.200205393349849,
+      "learning_rate": 9.781493862954655e-08,
+      "loss": 0.6882,
+      "step": 6420
+    },
+    {
+      "epoch": 0.9148466955965,
+      "grad_norm": 61.61576418334018,
+      "learning_rate": 9.464925865726366e-08,
+      "loss": 0.7736,
+      "step": 6430
+    },
+    {
+      "epoch": 0.9162694742832752,
+      "grad_norm": 23.93288153019585,
+      "learning_rate": 9.153466337961436e-08,
+      "loss": 0.7299,
+      "step": 6440
+    },
+    {
+      "epoch": 0.9176922529700505,
+      "grad_norm": 16.01184842044598,
+      "learning_rate": 8.847121894418609e-08,
+      "loss": 0.6777,
+      "step": 6450
+    },
+    {
+      "epoch": 0.9191150316568257,
+      "grad_norm": 11.579637767092105,
+      "learning_rate": 8.545899041222827e-08,
+      "loss": 0.6941,
+      "step": 6460
+    },
+    {
+      "epoch": 0.920537810343601,
+      "grad_norm": 17.965375169975463,
+      "learning_rate": 8.249804175726934e-08,
+      "loss": 0.6288,
+      "step": 6470
+    },
+    {
+      "epoch": 0.9219605890303764,
+      "grad_norm": 15.134319011043464,
+      "learning_rate": 7.958843586375886e-08,
+      "loss": 0.6137,
+      "step": 6480
+    },
+    {
+      "epoch": 0.9233833677171516,
+      "grad_norm": 17.336267124248604,
+      "learning_rate": 7.673023452573286e-08,
+      "loss": 0.6739,
+      "step": 6490
+    },
+    {
+      "epoch": 0.9248061464039269,
+      "grad_norm": 17.494109579354692,
+      "learning_rate": 7.392349844549923e-08,
+      "loss": 0.7235,
+      "step": 6500
+    },
+    {
+      "epoch": 0.9262289250907021,
+      "grad_norm": 25.849714031596356,
+      "learning_rate": 7.116828723234991e-08,
+      "loss": 0.7266,
+      "step": 6510
+    },
+    {
+      "epoch": 0.9276517037774774,
+      "grad_norm": 11.328018184834953,
+      "learning_rate": 6.846465940129609e-08,
+      "loss": 0.5282,
+      "step": 6520
+    },
+    {
+      "epoch": 0.9290744824642527,
+      "grad_norm": 8.210571934863204,
+      "learning_rate": 6.581267237182304e-08,
+      "loss": 0.7225,
+      "step": 6530
+    },
+    {
+      "epoch": 0.9304972611510279,
+      "grad_norm": 26.615240326619148,
+      "learning_rate": 6.321238246667166e-08,
+      "loss": 0.6524,
+      "step": 6540
+    },
+    {
+      "epoch": 0.9319200398378032,
+      "grad_norm": 10.506543994481495,
+      "learning_rate": 6.066384491064337e-08,
+      "loss": 0.6497,
+      "step": 6550
+    },
+    {
+      "epoch": 0.9333428185245785,
+      "grad_norm": 32.578044982132084,
+      "learning_rate": 5.8167113829425416e-08,
+      "loss": 0.7905,
+      "step": 6560
+    },
+    {
+      "epoch": 0.9347655972113538,
+      "grad_norm": 26.139082460876924,
+      "learning_rate": 5.572224224844297e-08,
+      "loss": 0.6009,
+      "step": 6570
+    },
+    {
+      "epoch": 0.9361883758981291,
+      "grad_norm": 25.70715498977883,
+      "learning_rate": 5.332928209173138e-08,
+      "loss": 0.7175,
+      "step": 6580
+    },
+    {
+      "epoch": 0.9376111545849043,
+      "grad_norm": 21.18820776002297,
+      "learning_rate": 5.0988284180835404e-08,
+      "loss": 0.6472,
+      "step": 6590
+    },
+    {
+      "epoch": 0.9390339332716796,
+      "grad_norm": 30.441715942004116,
+      "learning_rate": 4.8699298233727546e-08,
+      "loss": 0.7279,
+      "step": 6600
+    },
+    {
+      "epoch": 0.9404567119584548,
+      "grad_norm": 21.34748999867495,
+      "learning_rate": 4.646237286375449e-08,
+      "loss": 0.701,
+      "step": 6610
+    },
+    {
+      "epoch": 0.9418794906452301,
+      "grad_norm": 13.676207981782268,
+      "learning_rate": 4.427755557860236e-08,
+      "loss": 0.5594,
+      "step": 6620
+    },
+    {
+      "epoch": 0.9433022693320054,
+      "grad_norm": 13.877508873616051,
+      "learning_rate": 4.2144892779290015e-08,
+      "loss": 0.7165,
+      "step": 6630
+    },
+    {
+      "epoch": 0.9447250480187807,
+      "grad_norm": 34.9111327739568,
+      "learning_rate": 4.006442975918123e-08,
+      "loss": 0.7097,
+      "step": 6640
+    },
+    {
+      "epoch": 0.946147826705556,
+      "grad_norm": 8.860238134882508,
+      "learning_rate": 3.8036210703025486e-08,
+      "loss": 0.6653,
+      "step": 6650
+    },
+    {
+      "epoch": 0.9475706053923312,
+      "grad_norm": 12.97838301460571,
+      "learning_rate": 3.606027868601702e-08,
+      "loss": 0.6635,
+      "step": 6660
+    },
+    {
+      "epoch": 0.9489933840791065,
+      "grad_norm": 13.245086828871894,
+      "learning_rate": 3.413667567288143e-08,
+      "loss": 0.5832,
+      "step": 6670
+    },
+    {
+      "epoch": 0.9504161627658818,
+      "grad_norm": 14.890537012064636,
+      "learning_rate": 3.2265442516983826e-08,
+      "loss": 0.6808,
+      "step": 6680
+    },
+    {
+      "epoch": 0.951838941452657,
+      "grad_norm": 30.709204073798546,
+      "learning_rate": 3.0446618959461826e-08,
+      "loss": 0.6955,
+      "step": 6690
+    },
+    {
+      "epoch": 0.9532617201394323,
+      "grad_norm": 9.62781816402037,
+      "learning_rate": 2.8680243628381177e-08,
+      "loss": 0.6631,
+      "step": 6700
+    },
+    {
+      "epoch": 0.9546844988262075,
+      "grad_norm": 10.056987346093363,
+      "learning_rate": 2.6966354037914477e-08,
+      "loss": 0.6459,
+      "step": 6710
+    },
+    {
+      "epoch": 0.9561072775129829,
+      "grad_norm": 9.393608339470743,
+      "learning_rate": 2.5304986587546544e-08,
+      "loss": 0.5365,
+      "step": 6720
+    },
+    {
+      "epoch": 0.9575300561997582,
+      "grad_norm": 9.350612385267924,
+      "learning_rate": 2.369617656129919e-08,
+      "loss": 0.7408,
+      "step": 6730
+    },
+    {
+      "epoch": 0.9589528348865334,
+      "grad_norm": 17.57921704990656,
+      "learning_rate": 2.2139958126983217e-08,
+      "loss": 0.6266,
+      "step": 6740
+    },
+    {
+      "epoch": 0.9603756135733087,
+      "grad_norm": 9.372225143512935,
+      "learning_rate": 2.0636364335472614e-08,
+      "loss": 0.7491,
+      "step": 6750
+    },
+    {
+      "epoch": 0.9617983922600839,
+      "grad_norm": 21.57591391004256,
+      "learning_rate": 1.9185427120001487e-08,
+      "loss": 0.6907,
+      "step": 6760
+    },
+    {
+      "epoch": 0.9632211709468592,
+      "grad_norm": 17.18911624015644,
+      "learning_rate": 1.7787177295487956e-08,
+      "loss": 0.5657,
+      "step": 6770
+    },
+    {
+      "epoch": 0.9646439496336345,
+      "grad_norm": 36.75906035966361,
+      "learning_rate": 1.6441644557877444e-08,
+      "loss": 0.7114,
+      "step": 6780
+    },
+    {
+      "epoch": 0.9660667283204097,
+      "grad_norm": 21.24728321301748,
+      "learning_rate": 1.514885748351319e-08,
+      "loss": 0.6113,
+      "step": 6790
+    },
+    {
+      "epoch": 0.9674895070071851,
+      "grad_norm": 7.101447822320762,
+      "learning_rate": 1.3908843528530337e-08,
+      "loss": 0.6532,
+      "step": 6800
+    },
+    {
+      "epoch": 0.9689122856939603,
+      "grad_norm": 18.090559454068472,
+      "learning_rate": 1.2721629028269733e-08,
+      "loss": 0.6345,
+      "step": 6810
+    },
+    {
+      "epoch": 0.9703350643807356,
+      "grad_norm": 16.802326235518294,
+      "learning_rate": 1.1587239196722555e-08,
+      "loss": 0.6792,
+      "step": 6820
+    },
+    {
+      "epoch": 0.9717578430675109,
+      "grad_norm": 10.580873804156397,
+      "learning_rate": 1.050569812599156e-08,
+      "loss": 0.728,
+      "step": 6830
+    },
+    {
+      "epoch": 0.9731806217542861,
+      "grad_norm": 23.751550135831035,
+      "learning_rate": 9.47702878578094e-09,
+      "loss": 0.6499,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9746034004410614,
+      "grad_norm": 15.286902696444267,
+      "learning_rate": 8.501253022908662e-09,
+      "loss": 0.7415,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9760261791278366,
+      "grad_norm": 23.750942999798756,
+      "learning_rate": 7.578391560841836e-09,
+      "loss": 0.6233,
+      "step": 6860
+    },
+    {
+      "epoch": 0.9774489578146119,
+      "grad_norm": 13.098266803006062,
+      "learning_rate": 6.70846399925651e-09,
+      "loss": 0.7737,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9788717365013873,
+      "grad_norm": 11.660221339653082,
+      "learning_rate": 5.891488813621893e-09,
+      "loss": 0.777,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9802945151881625,
+      "grad_norm": 44.79723668339218,
+      "learning_rate": 5.127483354808449e-09,
+      "loss": 0.6923,
+      "step": 6890
+    },
+    {
+      "epoch": 0.9817172938749378,
+      "grad_norm": 13.137774107102898,
+      "learning_rate": 4.416463848717911e-09,
+      "loss": 0.6114,
+      "step": 6900
+    },
+    {
+      "epoch": 0.983140072561713,
+      "grad_norm": 13.86061937649414,
+      "learning_rate": 3.758445395939669e-09,
+      "loss": 0.6833,
+      "step": 6910
+    },
+    {
+      "epoch": 0.9845628512484883,
+      "grad_norm": 13.909081899756915,
+      "learning_rate": 3.1534419714307484e-09,
+      "loss": 0.6655,
+      "step": 6920
+    },
+    {
+      "epoch": 0.9859856299352636,
+      "grad_norm": 9.192505102390527,
+      "learning_rate": 2.6014664242168806e-09,
+      "loss": 0.6961,
+      "step": 6930
+    },
+    {
+      "epoch": 0.9874084086220388,
+      "grad_norm": 14.296169677609232,
+      "learning_rate": 2.102530477121889e-09,
+      "loss": 0.6308,
+      "step": 6940
+    },
+    {
+      "epoch": 0.9888311873088141,
+      "grad_norm": 20.607220237529024,
+      "learning_rate": 1.6566447265176089e-09,
+      "loss": 0.8115,
+      "step": 6950
+    },
+    {
+      "epoch": 0.9902539659955893,
+      "grad_norm": 18.63472394489305,
+      "learning_rate": 1.2638186420985132e-09,
+      "loss": 0.6796,
+      "step": 6960
+    },
+    {
+      "epoch": 0.9916767446823647,
+      "grad_norm": 21.10612820302392,
+      "learning_rate": 9.240605666818725e-10,
+      "loss": 0.6622,
+      "step": 6970
+    },
+    {
+      "epoch": 0.99309952336914,
+      "grad_norm": 31.484150350563915,
+      "learning_rate": 6.373777160287309e-10,
+      "loss": 0.6983,
+      "step": 6980
+    },
+    {
+      "epoch": 0.9945223020559152,
+      "grad_norm": 20.57209304211363,
+      "learning_rate": 4.0377617869263863e-10,
+      "loss": 0.6418,
+      "step": 6990
+    },
+    {
+      "epoch": 0.9959450807426905,
+      "grad_norm": 129.38234765145998,
+      "learning_rate": 2.232609158889232e-10,
+      "loss": 0.7882,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9973678594294657,
+      "grad_norm": 11.228945900702065,
+      "learning_rate": 9.583576138977313e-11,
+      "loss": 0.6529,
+      "step": 7010
+    },
+    {
+      "epoch": 0.998790638116241,
+      "grad_norm": 31.57256099590383,
+      "learning_rate": 2.150342144319195e-11,
+      "loss": 0.5964,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9999288610656613,
+      "step": 7028,
+      "total_flos": 1.4353626173538304e+16,
+      "train_loss": 1.0448219544112716,
+      "train_runtime": 120927.7644,
+      "train_samples_per_second": 3.72,
+      "train_steps_per_second": 0.058
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 7028,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4353626173538304e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}