{
  "best_metric": 0.9410023545240498,
  "best_model_checkpoint": "/mnt/data4_HDD_14TB/yang/voxceleb-checkpoints/xvector/voxceleb1/finetune/ce-len3-bs256-lr1e-3/checkpoint-5230",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 5230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03824091778202677,
      "grad_norm": 4.7170538902282715,
      "learning_rate": 3.824091778202677e-05,
      "loss": 7.1456,
      "step": 20
    },
    {
      "epoch": 0.07648183556405354,
      "grad_norm": 4.5134358406066895,
      "learning_rate": 7.648183556405354e-05,
      "loss": 7.1158,
      "step": 40
    },
    {
      "epoch": 0.1147227533460803,
      "grad_norm": 3.9098806381225586,
      "learning_rate": 0.0001147227533460803,
      "loss": 7.0575,
      "step": 60
    },
    {
      "epoch": 0.15296367112810708,
      "grad_norm": 3.2200050354003906,
      "learning_rate": 0.00015296367112810707,
      "loss": 6.9743,
      "step": 80
    },
    {
      "epoch": 0.19120458891013384,
      "grad_norm": 2.5479934215545654,
      "learning_rate": 0.00019120458891013384,
      "loss": 6.874,
      "step": 100
    },
    {
      "epoch": 0.2294455066921606,
      "grad_norm": 2.168301820755005,
      "learning_rate": 0.0002294455066921606,
      "loss": 6.7278,
      "step": 120
    },
    {
      "epoch": 0.2676864244741874,
      "grad_norm": 2.0083394050598145,
      "learning_rate": 0.0002676864244741874,
      "loss": 6.5778,
      "step": 140
    },
    {
      "epoch": 0.30592734225621415,
      "grad_norm": 1.8298897743225098,
      "learning_rate": 0.00030592734225621415,
      "loss": 6.4192,
      "step": 160
    },
    {
      "epoch": 0.3441682600382409,
      "grad_norm": 1.782423734664917,
      "learning_rate": 0.00034416826003824094,
      "loss": 6.2416,
      "step": 180
    },
    {
      "epoch": 0.3824091778202677,
      "grad_norm": 1.8139146566390991,
      "learning_rate": 0.0003824091778202677,
      "loss": 6.0624,
      "step": 200
    },
    {
      "epoch": 0.42065009560229444,
      "grad_norm": 1.7659958600997925,
      "learning_rate": 0.0004206500956022944,
      "loss": 5.9686,
      "step": 220
    },
    {
      "epoch": 0.4588910133843212,
      "grad_norm": 1.8660094738006592,
      "learning_rate": 0.0004588910133843212,
      "loss": 5.8142,
      "step": 240
    },
    {
      "epoch": 0.497131931166348,
      "grad_norm": 1.831566333770752,
      "learning_rate": 0.0004971319311663481,
      "loss": 5.6944,
      "step": 260
    },
    {
      "epoch": 0.5353728489483748,
      "grad_norm": 1.7546241283416748,
      "learning_rate": 0.0005353728489483748,
      "loss": 5.6101,
      "step": 280
    },
    {
      "epoch": 0.5736137667304015,
      "grad_norm": 1.8890600204467773,
      "learning_rate": 0.0005736137667304016,
      "loss": 5.5192,
      "step": 300
    },
    {
      "epoch": 0.6118546845124283,
      "grad_norm": 1.7542874813079834,
      "learning_rate": 0.0006118546845124283,
      "loss": 5.3822,
      "step": 320
    },
    {
      "epoch": 0.6500956022944551,
      "grad_norm": 1.8762731552124023,
      "learning_rate": 0.000650095602294455,
      "loss": 5.3236,
      "step": 340
    },
    {
      "epoch": 0.6883365200764818,
      "grad_norm": 1.886903166770935,
      "learning_rate": 0.0006883365200764819,
      "loss": 5.2483,
      "step": 360
    },
    {
      "epoch": 0.7265774378585086,
      "grad_norm": 1.9873583316802979,
      "learning_rate": 0.0007265774378585086,
      "loss": 5.1245,
      "step": 380
    },
    {
      "epoch": 0.7648183556405354,
      "grad_norm": 1.953506350517273,
      "learning_rate": 0.0007648183556405354,
      "loss": 5.0771,
      "step": 400
    },
    {
      "epoch": 0.8030592734225621,
      "grad_norm": 1.851192831993103,
      "learning_rate": 0.0008030592734225621,
      "loss": 5.0354,
      "step": 420
    },
    {
      "epoch": 0.8413001912045889,
      "grad_norm": 1.861971139907837,
      "learning_rate": 0.0008413001912045888,
      "loss": 4.9532,
      "step": 440
    },
    {
      "epoch": 0.8795411089866156,
      "grad_norm": 1.9388970136642456,
      "learning_rate": 0.0008795411089866157,
      "loss": 4.8698,
      "step": 460
    },
    {
      "epoch": 0.9177820267686424,
      "grad_norm": 1.919184684753418,
      "learning_rate": 0.0009177820267686424,
      "loss": 4.847,
      "step": 480
    },
    {
      "epoch": 0.9560229445506692,
      "grad_norm": 1.8886795043945312,
      "learning_rate": 0.0009560229445506692,
      "loss": 4.7754,
      "step": 500
    },
    {
      "epoch": 0.994263862332696,
      "grad_norm": 1.9832813739776611,
      "learning_rate": 0.0009942638623326961,
      "loss": 4.6728,
      "step": 520
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.15042045072317525,
      "eval_loss": 4.345639228820801,
      "eval_runtime": 211.2136,
      "eval_samples_per_second": 70.379,
      "eval_steps_per_second": 70.379,
      "step": 523
    },
    {
      "epoch": 1.0325047801147227,
      "grad_norm": 1.831260323524475,
      "learning_rate": 0.0009963883577650309,
      "loss": 4.5657,
      "step": 540
    },
    {
      "epoch": 1.0707456978967496,
      "grad_norm": 1.934414029121399,
      "learning_rate": 0.0009921393669003612,
      "loss": 4.5047,
      "step": 560
    },
    {
      "epoch": 1.1089866156787762,
      "grad_norm": 1.7718721628189087,
      "learning_rate": 0.0009878903760356915,
      "loss": 4.4165,
      "step": 580
    },
    {
      "epoch": 1.147227533460803,
      "grad_norm": 1.741455078125,
      "learning_rate": 0.0009836413851710218,
      "loss": 4.3933,
      "step": 600
    },
    {
      "epoch": 1.1854684512428297,
      "grad_norm": 1.8857481479644775,
      "learning_rate": 0.0009793923943063523,
      "loss": 4.3249,
      "step": 620
    },
    {
      "epoch": 1.2237093690248566,
      "grad_norm": 1.8325748443603516,
      "learning_rate": 0.0009751434034416827,
      "loss": 4.2147,
      "step": 640
    },
    {
      "epoch": 1.2619502868068833,
      "grad_norm": 1.8758591413497925,
      "learning_rate": 0.000970894412577013,
      "loss": 4.1569,
      "step": 660
    },
    {
      "epoch": 1.3001912045889101,
      "grad_norm": 1.899542212486267,
      "learning_rate": 0.0009666454217123433,
      "loss": 4.131,
      "step": 680
    },
    {
      "epoch": 1.338432122370937,
      "grad_norm": 1.8188538551330566,
      "learning_rate": 0.0009623964308476737,
      "loss": 4.0467,
      "step": 700
    },
    {
      "epoch": 1.3766730401529637,
      "grad_norm": 1.7679705619812012,
      "learning_rate": 0.000958147439983004,
      "loss": 3.9904,
      "step": 720
    },
    {
      "epoch": 1.4149139579349903,
      "grad_norm": 1.849482774734497,
      "learning_rate": 0.0009538984491183344,
      "loss": 3.9464,
      "step": 740
    },
    {
      "epoch": 1.4531548757170172,
      "grad_norm": 1.8237632513046265,
      "learning_rate": 0.0009496494582536647,
      "loss": 3.9104,
      "step": 760
    },
    {
      "epoch": 1.491395793499044,
      "grad_norm": 1.8175936937332153,
      "learning_rate": 0.0009454004673889951,
      "loss": 3.8441,
      "step": 780
    },
    {
      "epoch": 1.5296367112810707,
      "grad_norm": 1.7967997789382935,
      "learning_rate": 0.0009411514765243255,
      "loss": 3.7898,
      "step": 800
    },
    {
      "epoch": 1.5678776290630974,
      "grad_norm": 1.7681634426116943,
      "learning_rate": 0.0009369024856596558,
      "loss": 3.6894,
      "step": 820
    },
    {
      "epoch": 1.6061185468451242,
      "grad_norm": 1.8655925989151,
      "learning_rate": 0.0009326534947949862,
      "loss": 3.6798,
      "step": 840
    },
    {
      "epoch": 1.644359464627151,
      "grad_norm": 1.853769302368164,
      "learning_rate": 0.0009284045039303166,
      "loss": 3.6297,
      "step": 860
    },
    {
      "epoch": 1.682600382409178,
      "grad_norm": 1.8198288679122925,
      "learning_rate": 0.0009241555130656469,
      "loss": 3.5592,
      "step": 880
    },
    {
      "epoch": 1.7208413001912046,
      "grad_norm": 1.7744460105895996,
      "learning_rate": 0.0009199065222009773,
      "loss": 3.5056,
      "step": 900
    },
    {
      "epoch": 1.7590822179732313,
      "grad_norm": 1.797914981842041,
      "learning_rate": 0.0009156575313363077,
      "loss": 3.4635,
      "step": 920
    },
    {
      "epoch": 1.7973231357552581,
      "grad_norm": 1.8479169607162476,
      "learning_rate": 0.000911408540471638,
      "loss": 3.4434,
      "step": 940
    },
    {
      "epoch": 1.835564053537285,
      "grad_norm": 1.818405032157898,
      "learning_rate": 0.0009071595496069684,
      "loss": 3.441,
      "step": 960
    },
    {
      "epoch": 1.8738049713193117,
      "grad_norm": 1.7609572410583496,
      "learning_rate": 0.0009029105587422988,
      "loss": 3.3934,
      "step": 980
    },
    {
      "epoch": 1.9120458891013383,
      "grad_norm": 1.7228211164474487,
      "learning_rate": 0.0008986615678776291,
      "loss": 3.2961,
      "step": 1000
    },
    {
      "epoch": 1.9502868068833652,
      "grad_norm": 1.8148291110992432,
      "learning_rate": 0.0008944125770129595,
      "loss": 3.2611,
      "step": 1020
    },
    {
      "epoch": 1.988527724665392,
      "grad_norm": 1.933300495147705,
      "learning_rate": 0.0008901635861482899,
      "loss": 3.224,
      "step": 1040
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5140935082408342,
      "eval_loss": 2.258894205093384,
      "eval_runtime": 203.9361,
      "eval_samples_per_second": 72.89,
      "eval_steps_per_second": 72.89,
      "step": 1046
    },
    {
      "epoch": 2.026768642447419,
      "grad_norm": 1.819346308708191,
      "learning_rate": 0.0008859145952836202,
      "loss": 3.1667,
      "step": 1060
    },
    {
      "epoch": 2.0650095602294454,
      "grad_norm": 1.7024896144866943,
      "learning_rate": 0.0008816656044189504,
      "loss": 3.0232,
      "step": 1080
    },
    {
      "epoch": 2.1032504780114722,
      "grad_norm": 1.7023948431015015,
      "learning_rate": 0.000877416613554281,
      "loss": 3.0489,
      "step": 1100
    },
    {
      "epoch": 2.141491395793499,
      "grad_norm": 1.804140329360962,
      "learning_rate": 0.0008731676226896112,
      "loss": 2.9732,
      "step": 1120
    },
    {
      "epoch": 2.179732313575526,
      "grad_norm": 1.7260992527008057,
      "learning_rate": 0.0008689186318249415,
      "loss": 2.9562,
      "step": 1140
    },
    {
      "epoch": 2.2179732313575524,
      "grad_norm": 1.7970356941223145,
      "learning_rate": 0.000864669640960272,
      "loss": 2.8875,
      "step": 1160
    },
    {
      "epoch": 2.2562141491395793,
      "grad_norm": 1.8579261302947998,
      "learning_rate": 0.0008604206500956023,
      "loss": 2.916,
      "step": 1180
    },
    {
      "epoch": 2.294455066921606,
      "grad_norm": 1.852342128753662,
      "learning_rate": 0.0008561716592309326,
      "loss": 2.8963,
      "step": 1200
    },
    {
      "epoch": 2.332695984703633,
      "grad_norm": 1.8845752477645874,
      "learning_rate": 0.000851922668366263,
      "loss": 2.8029,
      "step": 1220
    },
    {
      "epoch": 2.3709369024856595,
      "grad_norm": 1.883952260017395,
      "learning_rate": 0.0008476736775015934,
      "loss": 2.8237,
      "step": 1240
    },
    {
      "epoch": 2.4091778202676863,
      "grad_norm": 1.8383756875991821,
      "learning_rate": 0.0008434246866369237,
      "loss": 2.8473,
      "step": 1260
    },
    {
      "epoch": 2.447418738049713,
      "grad_norm": 1.8900470733642578,
      "learning_rate": 0.0008391756957722541,
      "loss": 2.7722,
      "step": 1280
    },
    {
      "epoch": 2.48565965583174,
      "grad_norm": 1.8097845315933228,
      "learning_rate": 0.0008349267049075845,
      "loss": 2.7584,
      "step": 1300
    },
    {
      "epoch": 2.5239005736137665,
      "grad_norm": 1.7215895652770996,
      "learning_rate": 0.0008306777140429148,
      "loss": 2.7134,
      "step": 1320
    },
    {
      "epoch": 2.5621414913957934,
      "grad_norm": 1.8249051570892334,
      "learning_rate": 0.0008264287231782451,
      "loss": 2.6531,
      "step": 1340
    },
    {
      "epoch": 2.6003824091778203,
      "grad_norm": 1.8082237243652344,
      "learning_rate": 0.0008221797323135756,
      "loss": 2.6675,
      "step": 1360
    },
    {
      "epoch": 2.638623326959847,
      "grad_norm": 1.7981261014938354,
      "learning_rate": 0.0008179307414489059,
      "loss": 2.5702,
      "step": 1380
    },
    {
      "epoch": 2.676864244741874,
      "grad_norm": 1.6964036226272583,
      "learning_rate": 0.0008136817505842362,
      "loss": 2.6339,
      "step": 1400
    },
    {
      "epoch": 2.7151051625239004,
      "grad_norm": 1.755050778388977,
      "learning_rate": 0.0008094327597195667,
      "loss": 2.5489,
      "step": 1420
    },
    {
      "epoch": 2.7533460803059273,
      "grad_norm": 1.7242581844329834,
      "learning_rate": 0.000805183768854897,
      "loss": 2.5908,
      "step": 1440
    },
    {
      "epoch": 2.791586998087954,
      "grad_norm": 1.819612741470337,
      "learning_rate": 0.0008009347779902273,
      "loss": 2.5143,
      "step": 1460
    },
    {
      "epoch": 2.8298279158699806,
      "grad_norm": 1.7033363580703735,
      "learning_rate": 0.0007966857871255578,
      "loss": 2.4662,
      "step": 1480
    },
    {
      "epoch": 2.8680688336520075,
      "grad_norm": 1.7662159204483032,
      "learning_rate": 0.000792436796260888,
      "loss": 2.4044,
      "step": 1500
    },
    {
      "epoch": 2.9063097514340344,
      "grad_norm": 1.7460269927978516,
      "learning_rate": 0.0007881878053962183,
      "loss": 2.4636,
      "step": 1520
    },
    {
      "epoch": 2.9445506692160612,
      "grad_norm": 1.8268380165100098,
      "learning_rate": 0.0007839388145315488,
      "loss": 2.3955,
      "step": 1540
    },
    {
      "epoch": 2.982791586998088,
      "grad_norm": 1.796981930732727,
      "learning_rate": 0.0007796898236668791,
      "loss": 2.3964,
      "step": 1560
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6835519677093844,
      "eval_loss": 1.4662528038024902,
      "eval_runtime": 419.2044,
      "eval_samples_per_second": 35.46,
      "eval_steps_per_second": 35.46,
      "step": 1569
    },
    {
      "epoch": 3.0210325047801145,
      "grad_norm": 1.7852272987365723,
      "learning_rate": 0.0007754408328022094,
      "loss": 2.3174,
      "step": 1580
    },
    {
      "epoch": 3.0592734225621414,
      "grad_norm": 1.8464534282684326,
      "learning_rate": 0.0007711918419375399,
      "loss": 2.2913,
      "step": 1600
    },
    {
      "epoch": 3.0975143403441683,
      "grad_norm": 1.7783145904541016,
      "learning_rate": 0.0007669428510728702,
      "loss": 2.2856,
      "step": 1620
    },
    {
      "epoch": 3.135755258126195,
      "grad_norm": 1.744454264640808,
      "learning_rate": 0.0007626938602082005,
      "loss": 2.2099,
      "step": 1640
    },
    {
      "epoch": 3.173996175908222,
      "grad_norm": 1.8276797533035278,
      "learning_rate": 0.0007584448693435309,
      "loss": 2.23,
      "step": 1660
    },
    {
      "epoch": 3.2122370936902485,
      "grad_norm": 1.8144315481185913,
      "learning_rate": 0.0007541958784788613,
      "loss": 2.1912,
      "step": 1680
    },
    {
      "epoch": 3.2504780114722753,
      "grad_norm": 1.8499830961227417,
      "learning_rate": 0.0007499468876141916,
      "loss": 2.1818,
      "step": 1700
    },
    {
      "epoch": 3.288718929254302,
      "grad_norm": 1.7623099088668823,
      "learning_rate": 0.000745697896749522,
      "loss": 2.1349,
      "step": 1720
    },
    {
      "epoch": 3.3269598470363286,
      "grad_norm": 1.8180640935897827,
      "learning_rate": 0.0007414489058848524,
      "loss": 2.1055,
      "step": 1740
    },
    {
      "epoch": 3.3652007648183555,
      "grad_norm": 1.8159964084625244,
      "learning_rate": 0.0007371999150201827,
      "loss": 2.1077,
      "step": 1760
    },
    {
      "epoch": 3.4034416826003824,
      "grad_norm": 1.7902129888534546,
      "learning_rate": 0.0007329509241555131,
      "loss": 2.0999,
      "step": 1780
    },
    {
      "epoch": 3.4416826003824093,
      "grad_norm": 1.7685898542404175,
      "learning_rate": 0.0007287019332908435,
      "loss": 2.1188,
      "step": 1800
    },
    {
      "epoch": 3.479923518164436,
      "grad_norm": 1.758325219154358,
      "learning_rate": 0.0007244529424261738,
      "loss": 2.0956,
      "step": 1820
    },
    {
      "epoch": 3.5181644359464626,
      "grad_norm": 1.7802537679672241,
      "learning_rate": 0.0007202039515615042,
      "loss": 2.0488,
      "step": 1840
    },
    {
      "epoch": 3.5564053537284894,
      "grad_norm": 1.8220280408859253,
      "learning_rate": 0.0007159549606968346,
      "loss": 2.0776,
      "step": 1860
    },
    {
      "epoch": 3.5946462715105163,
      "grad_norm": 1.8494378328323364,
      "learning_rate": 0.0007117059698321649,
      "loss": 2.0257,
      "step": 1880
    },
    {
      "epoch": 3.632887189292543,
      "grad_norm": 1.719109296798706,
      "learning_rate": 0.0007074569789674953,
      "loss": 2.0332,
      "step": 1900
    },
    {
      "epoch": 3.67112810707457,
      "grad_norm": 1.9517509937286377,
      "learning_rate": 0.0007032079881028257,
      "loss": 2.0052,
      "step": 1920
    },
    {
      "epoch": 3.7093690248565965,
      "grad_norm": 1.7318940162658691,
      "learning_rate": 0.0006989589972381559,
      "loss": 1.9901,
      "step": 1940
    },
    {
      "epoch": 3.7476099426386233,
      "grad_norm": 1.767015814781189,
      "learning_rate": 0.0006947100063734863,
      "loss": 1.9411,
      "step": 1960
    },
    {
      "epoch": 3.78585086042065,
      "grad_norm": 1.761806607246399,
      "learning_rate": 0.0006904610155088166,
      "loss": 2.0048,
      "step": 1980
    },
    {
      "epoch": 3.8240917782026767,
      "grad_norm": 1.7126002311706543,
      "learning_rate": 0.000686212024644147,
      "loss": 1.9312,
      "step": 2000
    },
    {
      "epoch": 3.8623326959847035,
      "grad_norm": 1.7167437076568604,
      "learning_rate": 0.0006819630337794774,
      "loss": 1.9443,
      "step": 2020
    },
    {
      "epoch": 3.9005736137667304,
      "grad_norm": 1.749881386756897,
      "learning_rate": 0.0006777140429148077,
      "loss": 1.893,
      "step": 2040
    },
    {
      "epoch": 3.9388145315487573,
      "grad_norm": 1.6846592426300049,
      "learning_rate": 0.0006734650520501381,
      "loss": 1.8876,
      "step": 2060
    },
    {
      "epoch": 3.977055449330784,
      "grad_norm": 1.8149057626724243,
      "learning_rate": 0.0006692160611854685,
      "loss": 1.8474,
      "step": 2080
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7926673393878237,
      "eval_loss": 0.9547563195228577,
      "eval_runtime": 183.3001,
      "eval_samples_per_second": 81.097,
      "eval_steps_per_second": 81.097,
      "step": 2092
    },
    {
      "epoch": 4.015296367112811,
      "grad_norm": 1.6455098390579224,
      "learning_rate": 0.0006649670703207988,
      "loss": 1.813,
      "step": 2100
    },
    {
      "epoch": 4.053537284894838,
      "grad_norm": 1.6958200931549072,
      "learning_rate": 0.0006607180794561292,
      "loss": 1.7354,
      "step": 2120
    },
    {
      "epoch": 4.091778202676864,
      "grad_norm": 1.7456037998199463,
      "learning_rate": 0.0006564690885914596,
      "loss": 1.7479,
      "step": 2140
    },
    {
      "epoch": 4.130019120458891,
      "grad_norm": 1.7887734174728394,
      "learning_rate": 0.0006522200977267899,
      "loss": 1.7138,
      "step": 2160
    },
    {
      "epoch": 4.168260038240918,
      "grad_norm": 1.7080284357070923,
      "learning_rate": 0.0006479711068621203,
      "loss": 1.7023,
      "step": 2180
    },
    {
      "epoch": 4.2065009560229445,
      "grad_norm": 1.8061983585357666,
      "learning_rate": 0.0006437221159974506,
      "loss": 1.7526,
      "step": 2200
    },
    {
      "epoch": 4.244741873804971,
      "grad_norm": 1.7831811904907227,
      "learning_rate": 0.000639473125132781,
      "loss": 1.7474,
      "step": 2220
    },
    {
      "epoch": 4.282982791586998,
      "grad_norm": 1.752357840538025,
      "learning_rate": 0.0006352241342681113,
      "loss": 1.6688,
      "step": 2240
    },
    {
      "epoch": 4.321223709369025,
      "grad_norm": 1.7843034267425537,
      "learning_rate": 0.0006309751434034417,
      "loss": 1.7009,
      "step": 2260
    },
    {
      "epoch": 4.359464627151052,
      "grad_norm": 1.7608367204666138,
      "learning_rate": 0.0006267261525387721,
      "loss": 1.6727,
      "step": 2280
    },
    {
      "epoch": 4.397705544933078,
      "grad_norm": 1.6877254247665405,
      "learning_rate": 0.0006224771616741024,
      "loss": 1.6801,
      "step": 2300
    },
    {
      "epoch": 4.435946462715105,
      "grad_norm": 1.7891350984573364,
      "learning_rate": 0.0006182281708094328,
      "loss": 1.7108,
      "step": 2320
    },
    {
      "epoch": 4.474187380497132,
      "grad_norm": 1.7104123830795288,
      "learning_rate": 0.0006139791799447631,
      "loss": 1.6442,
      "step": 2340
    },
    {
      "epoch": 4.512428298279159,
      "grad_norm": 1.7026969194412231,
      "learning_rate": 0.0006097301890800934,
      "loss": 1.6531,
      "step": 2360
    },
    {
      "epoch": 4.550669216061186,
      "grad_norm": 1.7890552282333374,
      "learning_rate": 0.0006054811982154238,
      "loss": 1.6539,
      "step": 2380
    },
    {
      "epoch": 4.588910133843212,
      "grad_norm": 1.8423861265182495,
      "learning_rate": 0.0006012322073507542,
      "loss": 1.6681,
      "step": 2400
    },
    {
      "epoch": 4.627151051625239,
      "grad_norm": 1.6434499025344849,
      "learning_rate": 0.0005969832164860845,
      "loss": 1.5935,
      "step": 2420
    },
    {
      "epoch": 4.665391969407266,
      "grad_norm": 1.7261130809783936,
      "learning_rate": 0.0005927342256214149,
      "loss": 1.6273,
      "step": 2440
    },
    {
      "epoch": 4.7036328871892925,
      "grad_norm": 1.7288273572921753,
      "learning_rate": 0.0005884852347567453,
      "loss": 1.6181,
      "step": 2460
    },
    {
      "epoch": 4.741873804971319,
      "grad_norm": 1.773258924484253,
      "learning_rate": 0.0005842362438920756,
      "loss": 1.5719,
      "step": 2480
    },
    {
      "epoch": 4.780114722753346,
      "grad_norm": 1.7676658630371094,
      "learning_rate": 0.000579987253027406,
      "loss": 1.578,
      "step": 2500
    },
    {
      "epoch": 4.818355640535373,
      "grad_norm": 1.8115794658660889,
      "learning_rate": 0.0005757382621627364,
      "loss": 1.535,
      "step": 2520
    },
    {
      "epoch": 4.8565965583174,
      "grad_norm": 1.7989414930343628,
      "learning_rate": 0.0005714892712980667,
      "loss": 1.5493,
      "step": 2540
    },
    {
      "epoch": 4.894837476099426,
      "grad_norm": 1.6607849597930908,
      "learning_rate": 0.000567240280433397,
      "loss": 1.5489,
      "step": 2560
    },
    {
      "epoch": 4.933078393881453,
      "grad_norm": 1.630257487297058,
      "learning_rate": 0.0005629912895687275,
      "loss": 1.5091,
      "step": 2580
    },
    {
      "epoch": 4.97131931166348,
      "grad_norm": 1.7995944023132324,
      "learning_rate": 0.0005587422987040578,
      "loss": 1.5275,
      "step": 2600
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8571140262361251,
      "eval_loss": 0.6697778105735779,
      "eval_runtime": 181.3784,
      "eval_samples_per_second": 81.956,
      "eval_steps_per_second": 81.956,
      "step": 2615
    },
    {
      "epoch": 5.009560229445507,
      "grad_norm": 1.7868553400039673,
      "learning_rate": 0.0005544933078393881,
      "loss": 1.4774,
      "step": 2620
    },
    {
      "epoch": 5.047801147227533,
      "grad_norm": 1.6380654573440552,
      "learning_rate": 0.0005502443169747186,
      "loss": 1.3955,
      "step": 2640
    },
    {
      "epoch": 5.08604206500956,
      "grad_norm": 1.7844533920288086,
      "learning_rate": 0.0005459953261100489,
      "loss": 1.4414,
      "step": 2660
    },
    {
      "epoch": 5.124282982791587,
      "grad_norm": 1.779080867767334,
      "learning_rate": 0.0005417463352453792,
      "loss": 1.3782,
      "step": 2680
    },
    {
      "epoch": 5.162523900573614,
      "grad_norm": 1.741326928138733,
      "learning_rate": 0.0005374973443807097,
      "loss": 1.4152,
      "step": 2700
    },
    {
      "epoch": 5.2007648183556405,
      "grad_norm": 1.7447401285171509,
      "learning_rate": 0.00053324835351604,
      "loss": 1.3996,
      "step": 2720
    },
    {
      "epoch": 5.239005736137667,
      "grad_norm": 1.8067736625671387,
      "learning_rate": 0.0005289993626513702,
      "loss": 1.4137,
      "step": 2740
    },
    {
      "epoch": 5.277246653919694,
      "grad_norm": 1.7393046617507935,
      "learning_rate": 0.0005247503717867008,
      "loss": 1.3937,
      "step": 2760
    },
    {
      "epoch": 5.315487571701721,
      "grad_norm": 1.756184458732605,
      "learning_rate": 0.000520501380922031,
      "loss": 1.3912,
      "step": 2780
    },
    {
      "epoch": 5.353728489483748,
      "grad_norm": 1.7133733034133911,
      "learning_rate": 0.0005162523900573613,
      "loss": 1.387,
      "step": 2800
    },
    {
      "epoch": 5.3919694072657744,
      "grad_norm": 1.6597713232040405,
      "learning_rate": 0.0005120033991926918,
      "loss": 1.3551,
      "step": 2820
    },
    {
      "epoch": 5.430210325047801,
      "grad_norm": 1.8462845087051392,
      "learning_rate": 0.0005077544083280221,
      "loss": 1.3557,
      "step": 2840
    },
    {
      "epoch": 5.468451242829828,
      "grad_norm": 1.6737143993377686,
      "learning_rate": 0.0005035054174633524,
      "loss": 1.3495,
      "step": 2860
    },
    {
      "epoch": 5.506692160611855,
      "grad_norm": 1.7071157693862915,
      "learning_rate": 0.0004992564265986828,
      "loss": 1.394,
      "step": 2880
    },
    {
      "epoch": 5.544933078393882,
      "grad_norm": 1.663072943687439,
      "learning_rate": 0.0004950074357340132,
      "loss": 1.3263,
      "step": 2900
    },
    {
      "epoch": 5.583173996175908,
      "grad_norm": 1.640093207359314,
      "learning_rate": 0.0004907584448693436,
      "loss": 1.3474,
      "step": 2920
    },
    {
      "epoch": 5.621414913957935,
      "grad_norm": 1.762568712234497,
      "learning_rate": 0.0004865094540046739,
      "loss": 1.3375,
      "step": 2940
    },
    {
      "epoch": 5.659655831739962,
      "grad_norm": 1.6714434623718262,
      "learning_rate": 0.00048226046314000425,
      "loss": 1.3218,
      "step": 2960
    },
    {
      "epoch": 5.6978967495219885,
      "grad_norm": 1.7594107389450073,
      "learning_rate": 0.0004780114722753346,
      "loss": 1.3008,
      "step": 2980
    },
    {
      "epoch": 5.736137667304015,
      "grad_norm": 1.6483973264694214,
      "learning_rate": 0.000473762481410665,
      "loss": 1.3331,
      "step": 3000
    },
    {
      "epoch": 5.774378585086042,
      "grad_norm": 1.7252651453018188,
      "learning_rate": 0.00046951349054599533,
      "loss": 1.2775,
      "step": 3020
    },
    {
      "epoch": 5.812619502868069,
      "grad_norm": 1.7860745191574097,
      "learning_rate": 0.0004652644996813257,
      "loss": 1.2747,
      "step": 3040
    },
    {
      "epoch": 5.850860420650095,
      "grad_norm": 1.749874234199524,
      "learning_rate": 0.0004610155088166561,
      "loss": 1.2946,
      "step": 3060
    },
    {
      "epoch": 5.8891013384321225,
      "grad_norm": 1.7197644710540771,
      "learning_rate": 0.0004567665179519864,
      "loss": 1.2849,
      "step": 3080
    },
    {
      "epoch": 5.927342256214149,
      "grad_norm": 1.6396132707595825,
      "learning_rate": 0.00045251752708731676,
      "loss": 1.2544,
      "step": 3100
    },
    {
      "epoch": 5.965583173996176,
      "grad_norm": 1.720376968383789,
      "learning_rate": 0.0004482685362226471,
      "loss": 1.248,
      "step": 3120
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8899428187016482,
      "eval_loss": 0.5270123481750488,
      "eval_runtime": 183.7621,
      "eval_samples_per_second": 80.893,
      "eval_steps_per_second": 80.893,
      "step": 3138
    },
    {
      "epoch": 6.003824091778203,
      "grad_norm": 1.5206599235534668,
      "learning_rate": 0.0004440195453579775,
      "loss": 1.2398,
      "step": 3140
    },
    {
      "epoch": 6.042065009560229,
      "grad_norm": 1.7172082662582397,
      "learning_rate": 0.00043977055449330785,
      "loss": 1.2097,
      "step": 3160
    },
    {
      "epoch": 6.080305927342256,
      "grad_norm": 1.5570909976959229,
      "learning_rate": 0.0004355215636286382,
      "loss": 1.1669,
      "step": 3180
    },
    {
      "epoch": 6.118546845124283,
      "grad_norm": 1.7044614553451538,
      "learning_rate": 0.0004312725727639686,
      "loss": 1.1647,
      "step": 3200
    },
    {
      "epoch": 6.15678776290631,
      "grad_norm": 1.5819571018218994,
      "learning_rate": 0.0004270235818992989,
      "loss": 1.1627,
      "step": 3220
    },
    {
      "epoch": 6.195028680688337,
      "grad_norm": 1.7076871395111084,
      "learning_rate": 0.0004227745910346293,
      "loss": 1.1728,
      "step": 3240
    },
    {
      "epoch": 6.233269598470363,
      "grad_norm": 1.7301490306854248,
      "learning_rate": 0.0004185256001699597,
      "loss": 1.1459,
      "step": 3260
    },
    {
      "epoch": 6.27151051625239,
      "grad_norm": 1.7135626077651978,
      "learning_rate": 0.00041427660930528997,
      "loss": 1.1676,
      "step": 3280
    },
    {
      "epoch": 6.309751434034417,
      "grad_norm": 1.602142572402954,
      "learning_rate": 0.00041002761844062037,
      "loss": 1.1488,
      "step": 3300
    },
    {
      "epoch": 6.347992351816444,
      "grad_norm": 1.755293846130371,
      "learning_rate": 0.00040577862757595076,
      "loss": 1.1395,
      "step": 3320
    },
    {
      "epoch": 6.3862332695984705,
      "grad_norm": 1.663662314414978,
      "learning_rate": 0.00040152963671128105,
      "loss": 1.1321,
      "step": 3340
    },
    {
      "epoch": 6.424474187380497,
      "grad_norm": 1.7366993427276611,
      "learning_rate": 0.00039728064584661145,
      "loss": 1.1317,
      "step": 3360
    },
    {
      "epoch": 6.462715105162524,
      "grad_norm": 1.7560149431228638,
      "learning_rate": 0.0003930316549819418,
      "loss": 1.1449,
      "step": 3380
    },
    {
      "epoch": 6.500956022944551,
      "grad_norm": 1.7576582431793213,
      "learning_rate": 0.00038878266411727214,
      "loss": 1.13,
      "step": 3400
    },
    {
      "epoch": 6.539196940726577,
      "grad_norm": 1.7916873693466187,
      "learning_rate": 0.00038453367325260254,
      "loss": 1.1419,
      "step": 3420
    },
    {
      "epoch": 6.577437858508604,
      "grad_norm": 1.5987508296966553,
      "learning_rate": 0.0003802846823879329,
      "loss": 1.1107,
      "step": 3440
    },
    {
      "epoch": 6.615678776290631,
      "grad_norm": 1.8192518949508667,
      "learning_rate": 0.0003760356915232632,
      "loss": 1.1162,
      "step": 3460
    },
    {
      "epoch": 6.653919694072657,
      "grad_norm": 1.7236486673355103,
      "learning_rate": 0.0003717867006585936,
      "loss": 1.1255,
      "step": 3480
    },
    {
      "epoch": 6.692160611854685,
      "grad_norm": 1.8209389448165894,
      "learning_rate": 0.0003675377097939239,
      "loss": 1.0629,
      "step": 3500
    },
    {
      "epoch": 6.730401529636711,
      "grad_norm": 1.652782678604126,
      "learning_rate": 0.0003632887189292543,
      "loss": 1.0809,
      "step": 3520
    },
    {
      "epoch": 6.768642447418738,
      "grad_norm": 1.6148645877838135,
      "learning_rate": 0.00035903972806458466,
      "loss": 1.1286,
      "step": 3540
    },
    {
      "epoch": 6.806883365200765,
      "grad_norm": 1.6869423389434814,
      "learning_rate": 0.000354790737199915,
      "loss": 1.1069,
      "step": 3560
    },
    {
      "epoch": 6.845124282982791,
      "grad_norm": 1.6373172998428345,
      "learning_rate": 0.0003505417463352454,
      "loss": 1.0911,
      "step": 3580
    },
    {
      "epoch": 6.8833652007648185,
      "grad_norm": 1.6761549711227417,
      "learning_rate": 0.00034629275547057574,
      "loss": 1.0808,
      "step": 3600
    },
    {
      "epoch": 6.921606118546845,
      "grad_norm": 1.6510460376739502,
      "learning_rate": 0.0003420437646059061,
      "loss": 1.0809,
      "step": 3620
    },
    {
      "epoch": 6.959847036328872,
      "grad_norm": 1.7351855039596558,
      "learning_rate": 0.0003377947737412365,
      "loss": 1.0912,
      "step": 3640
    },
    {
      "epoch": 6.998087954110899,
      "grad_norm": 1.7165274620056152,
      "learning_rate": 0.00033354578287656683,
      "loss": 1.0991,
      "step": 3660
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9037336024217961,
      "eval_loss": 0.44995447993278503,
      "eval_runtime": 189.5634,
      "eval_samples_per_second": 78.417,
      "eval_steps_per_second": 78.417,
      "step": 3661
    },
    {
      "epoch": 7.036328871892925,
      "grad_norm": 1.6468501091003418,
      "learning_rate": 0.0003292967920118972,
      "loss": 1.0154,
      "step": 3680
    },
    {
      "epoch": 7.074569789674952,
      "grad_norm": 1.79421067237854,
      "learning_rate": 0.0003250478011472275,
      "loss": 1.0378,
      "step": 3700
    },
    {
      "epoch": 7.112810707456979,
      "grad_norm": 1.7234885692596436,
      "learning_rate": 0.0003207988102825579,
      "loss": 1.0145,
      "step": 3720
    },
    {
      "epoch": 7.151051625239006,
      "grad_norm": 1.6947157382965088,
      "learning_rate": 0.00031654981941788826,
      "loss": 1.0012,
      "step": 3740
    },
    {
      "epoch": 7.189292543021033,
      "grad_norm": 1.6818758249282837,
      "learning_rate": 0.0003123008285532186,
      "loss": 1.0191,
      "step": 3760
    },
    {
      "epoch": 7.227533460803059,
      "grad_norm": 1.557080864906311,
      "learning_rate": 0.000308051837688549,
      "loss": 1.0437,
      "step": 3780
    },
    {
      "epoch": 7.265774378585086,
      "grad_norm": 1.6532793045043945,
      "learning_rate": 0.00030380284682387935,
      "loss": 1.0079,
      "step": 3800
    },
    {
      "epoch": 7.304015296367113,
      "grad_norm": 1.646686315536499,
      "learning_rate": 0.0002995538559592097,
      "loss": 0.9904,
      "step": 3820
    },
    {
      "epoch": 7.342256214149139,
      "grad_norm": 1.6772829294204712,
      "learning_rate": 0.0002953048650945401,
      "loss": 1.0002,
      "step": 3840
    },
    {
      "epoch": 7.3804971319311665,
      "grad_norm": 1.6452054977416992,
      "learning_rate": 0.0002910558742298704,
      "loss": 0.9674,
      "step": 3860
    },
    {
      "epoch": 7.418738049713193,
      "grad_norm": 1.592207908630371,
      "learning_rate": 0.0002868068833652008,
      "loss": 0.9642,
      "step": 3880
    },
    {
      "epoch": 7.45697896749522,
      "grad_norm": 1.7015941143035889,
      "learning_rate": 0.0002825578925005312,
      "loss": 0.973,
      "step": 3900
    },
    {
      "epoch": 7.495219885277247,
      "grad_norm": 1.6589232683181763,
      "learning_rate": 0.00027830890163586146,
      "loss": 0.9803,
      "step": 3920
    },
    {
      "epoch": 7.533460803059273,
      "grad_norm": 1.660190463066101,
      "learning_rate": 0.00027405991077119186,
      "loss": 0.9702,
      "step": 3940
    },
    {
      "epoch": 7.5717017208413,
      "grad_norm": 1.7052509784698486,
      "learning_rate": 0.00026981091990652226,
      "loss": 0.9919,
      "step": 3960
    },
    {
      "epoch": 7.609942638623327,
      "grad_norm": 1.6874445676803589,
      "learning_rate": 0.00026556192904185255,
      "loss": 0.938,
      "step": 3980
    },
    {
      "epoch": 7.648183556405353,
      "grad_norm": 1.811640739440918,
      "learning_rate": 0.00026131293817718295,
      "loss": 0.9564,
      "step": 4000
    },
    {
      "epoch": 7.686424474187381,
      "grad_norm": 1.741968035697937,
      "learning_rate": 0.00025706394731251324,
      "loss": 0.9432,
      "step": 4020
    },
    {
      "epoch": 7.724665391969407,
      "grad_norm": 1.6731518507003784,
      "learning_rate": 0.00025281495644784364,
      "loss": 0.9082,
      "step": 4040
    },
    {
      "epoch": 7.762906309751434,
      "grad_norm": 1.7399870157241821,
      "learning_rate": 0.00024856596558317403,
      "loss": 0.92,
      "step": 4060
    },
    {
      "epoch": 7.801147227533461,
      "grad_norm": 1.580674171447754,
      "learning_rate": 0.0002443169747185044,
      "loss": 0.9433,
      "step": 4080
    },
    {
      "epoch": 7.839388145315487,
      "grad_norm": 1.5683550834655762,
      "learning_rate": 0.00024006798385383472,
      "loss": 0.9584,
      "step": 4100
    },
    {
      "epoch": 7.8776290630975145,
      "grad_norm": 1.7664682865142822,
      "learning_rate": 0.00023581899298916507,
      "loss": 0.9318,
      "step": 4120
    },
    {
      "epoch": 7.915869980879541,
      "grad_norm": 1.5522887706756592,
      "learning_rate": 0.00023157000212449544,
      "loss": 0.8852,
      "step": 4140
    },
    {
      "epoch": 7.954110898661568,
      "grad_norm": 1.626836895942688,
      "learning_rate": 0.00022732101125982578,
      "loss": 0.9121,
      "step": 4160
    },
    {
      "epoch": 7.992351816443595,
      "grad_norm": 1.8042898178100586,
      "learning_rate": 0.00022307202039515615,
      "loss": 0.9221,
      "step": 4180
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9266733938782374,
      "eval_loss": 0.3572401702404022,
      "eval_runtime": 168.0336,
      "eval_samples_per_second": 88.464,
      "eval_steps_per_second": 88.464,
      "step": 4184
    },
    {
      "epoch": 8.030592734225621,
      "grad_norm": 1.545024037361145,
      "learning_rate": 0.00021882302953048652,
      "loss": 0.8979,
      "step": 4200
    },
    {
      "epoch": 8.068833652007648,
      "grad_norm": 1.592607855796814,
      "learning_rate": 0.00021457403866581687,
      "loss": 0.8464,
      "step": 4220
    },
    {
      "epoch": 8.107074569789676,
      "grad_norm": 1.5347646474838257,
      "learning_rate": 0.0002103250478011472,
      "loss": 0.8634,
      "step": 4240
    },
    {
      "epoch": 8.145315487571702,
      "grad_norm": 1.621201515197754,
      "learning_rate": 0.0002060760569364776,
      "loss": 0.8425,
      "step": 4260
    },
    {
      "epoch": 8.183556405353729,
      "grad_norm": 1.7381062507629395,
      "learning_rate": 0.00020182706607180795,
      "loss": 0.8776,
      "step": 4280
    },
    {
      "epoch": 8.221797323135755,
      "grad_norm": 1.5798373222351074,
      "learning_rate": 0.0001975780752071383,
      "loss": 0.854,
      "step": 4300
    },
    {
      "epoch": 8.260038240917781,
      "grad_norm": 1.5751338005065918,
      "learning_rate": 0.00019332908434246867,
      "loss": 0.8646,
      "step": 4320
    },
    {
      "epoch": 8.29827915869981,
      "grad_norm": 1.570742130279541,
      "learning_rate": 0.00018908009347779904,
      "loss": 0.8521,
      "step": 4340
    },
    {
      "epoch": 8.336520076481836,
      "grad_norm": 1.7959846258163452,
      "learning_rate": 0.00018483110261312938,
      "loss": 0.8858,
      "step": 4360
    },
    {
      "epoch": 8.374760994263863,
      "grad_norm": 1.7537908554077148,
      "learning_rate": 0.00018058211174845973,
      "loss": 0.8707,
      "step": 4380
    },
    {
      "epoch": 8.413001912045889,
      "grad_norm": 1.635578989982605,
      "learning_rate": 0.0001763331208837901,
      "loss": 0.8375,
      "step": 4400
    },
    {
      "epoch": 8.451242829827915,
      "grad_norm": 1.5729222297668457,
      "learning_rate": 0.00017208413001912047,
      "loss": 0.8631,
      "step": 4420
    },
    {
      "epoch": 8.489483747609942,
      "grad_norm": 1.6586476564407349,
      "learning_rate": 0.00016783513915445082,
      "loss": 0.8551,
      "step": 4440
    },
    {
      "epoch": 8.52772466539197,
      "grad_norm": 1.6118619441986084,
      "learning_rate": 0.00016358614828978119,
      "loss": 0.8637,
      "step": 4460
    },
    {
      "epoch": 8.565965583173996,
      "grad_norm": 1.5538595914840698,
      "learning_rate": 0.00015933715742511153,
      "loss": 0.8484,
      "step": 4480
    },
    {
      "epoch": 8.604206500956023,
      "grad_norm": 1.5646642446517944,
      "learning_rate": 0.0001550881665604419,
      "loss": 0.8433,
      "step": 4500
    },
    {
      "epoch": 8.64244741873805,
      "grad_norm": 1.7190415859222412,
      "learning_rate": 0.00015083917569577227,
      "loss": 0.8592,
      "step": 4520
    },
    {
      "epoch": 8.680688336520076,
      "grad_norm": 1.4950307607650757,
      "learning_rate": 0.00014659018483110262,
      "loss": 0.8236,
      "step": 4540
    },
    {
      "epoch": 8.718929254302104,
      "grad_norm": 1.5117732286453247,
      "learning_rate": 0.00014234119396643296,
      "loss": 0.8421,
      "step": 4560
    },
    {
      "epoch": 8.75717017208413,
      "grad_norm": 1.5558750629425049,
      "learning_rate": 0.00013809220310176336,
      "loss": 0.8287,
      "step": 4580
    },
    {
      "epoch": 8.795411089866157,
      "grad_norm": 1.7955564260482788,
      "learning_rate": 0.0001338432122370937,
      "loss": 0.8492,
      "step": 4600
    },
    {
      "epoch": 8.833652007648183,
      "grad_norm": 1.6532599925994873,
      "learning_rate": 0.00012959422137242405,
      "loss": 0.8419,
      "step": 4620
    },
    {
      "epoch": 8.87189292543021,
      "grad_norm": 1.7040739059448242,
      "learning_rate": 0.0001253452305077544,
      "loss": 0.8125,
      "step": 4640
    },
    {
      "epoch": 8.910133843212238,
      "grad_norm": 1.7040703296661377,
      "learning_rate": 0.00012109623964308478,
      "loss": 0.8187,
      "step": 4660
    },
    {
      "epoch": 8.948374760994264,
      "grad_norm": 1.7090845108032227,
      "learning_rate": 0.00011684724877841513,
      "loss": 0.8155,
      "step": 4680
    },
    {
      "epoch": 8.98661567877629,
      "grad_norm": 1.6070616245269775,
      "learning_rate": 0.00011259825791374549,
      "loss": 0.7997,
      "step": 4700
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9352842246888665,
      "eval_loss": 0.3138497769832611,
      "eval_runtime": 177.561,
      "eval_samples_per_second": 83.718,
      "eval_steps_per_second": 83.718,
      "step": 4707
    },
    {
      "epoch": 9.024856596558317,
      "grad_norm": 1.5590825080871582,
      "learning_rate": 0.00010834926704907585,
      "loss": 0.7906,
      "step": 4720
    },
    {
      "epoch": 9.063097514340344,
      "grad_norm": 1.4745252132415771,
      "learning_rate": 0.0001041002761844062,
      "loss": 0.7494,
      "step": 4740
    },
    {
      "epoch": 9.101338432122372,
      "grad_norm": 1.61099112033844,
      "learning_rate": 9.985128531973658e-05,
      "loss": 0.7854,
      "step": 4760
    },
    {
      "epoch": 9.139579349904398,
      "grad_norm": 1.5839650630950928,
      "learning_rate": 9.560229445506692e-05,
      "loss": 0.7636,
      "step": 4780
    },
    {
      "epoch": 9.177820267686425,
      "grad_norm": 1.7259138822555542,
      "learning_rate": 9.135330359039729e-05,
      "loss": 0.776,
      "step": 4800
    },
    {
      "epoch": 9.216061185468451,
      "grad_norm": 1.5495970249176025,
      "learning_rate": 8.710431272572764e-05,
      "loss": 0.7444,
      "step": 4820
    },
    {
      "epoch": 9.254302103250478,
      "grad_norm": 1.5250838994979858,
      "learning_rate": 8.2855321861058e-05,
      "loss": 0.7603,
      "step": 4840
    },
    {
      "epoch": 9.292543021032504,
      "grad_norm": 1.6244220733642578,
      "learning_rate": 7.860633099638836e-05,
      "loss": 0.7561,
      "step": 4860
    },
    {
      "epoch": 9.330783938814532,
      "grad_norm": 1.6825993061065674,
      "learning_rate": 7.435734013171871e-05,
      "loss": 0.7908,
      "step": 4880
    },
    {
      "epoch": 9.369024856596559,
      "grad_norm": 1.563707947731018,
      "learning_rate": 7.010834926704908e-05,
      "loss": 0.7517,
      "step": 4900
    },
    {
      "epoch": 9.407265774378585,
      "grad_norm": 1.7463629245758057,
      "learning_rate": 6.585935840237942e-05,
      "loss": 0.7679,
      "step": 4920
    },
    {
      "epoch": 9.445506692160611,
      "grad_norm": 1.5689053535461426,
      "learning_rate": 6.16103675377098e-05,
      "loss": 0.7426,
      "step": 4940
    },
    {
      "epoch": 9.483747609942638,
      "grad_norm": 1.6512914896011353,
      "learning_rate": 5.736137667304015e-05,
      "loss": 0.7695,
      "step": 4960
    },
    {
      "epoch": 9.521988527724666,
      "grad_norm": 1.6542084217071533,
      "learning_rate": 5.311238580837052e-05,
      "loss": 0.7603,
      "step": 4980
    },
    {
      "epoch": 9.560229445506693,
      "grad_norm": 1.6929945945739746,
      "learning_rate": 4.8863394943700874e-05,
      "loss": 0.754,
      "step": 5000
    },
    {
      "epoch": 9.598470363288719,
      "grad_norm": 1.4880517721176147,
      "learning_rate": 4.461440407903123e-05,
      "loss": 0.7597,
      "step": 5020
    },
    {
      "epoch": 9.636711281070745,
      "grad_norm": 1.578971266746521,
      "learning_rate": 4.036541321436159e-05,
      "loss": 0.7624,
      "step": 5040
    },
    {
      "epoch": 9.674952198852772,
      "grad_norm": 1.616727352142334,
      "learning_rate": 3.6116422349691954e-05,
      "loss": 0.7653,
      "step": 5060
    },
    {
      "epoch": 9.7131931166348,
      "grad_norm": 1.6762784719467163,
      "learning_rate": 3.186743148502231e-05,
      "loss": 0.7352,
      "step": 5080
    },
    {
      "epoch": 9.751434034416826,
      "grad_norm": 1.5666388273239136,
      "learning_rate": 2.7618440620352666e-05,
      "loss": 0.732,
      "step": 5100
    },
    {
      "epoch": 9.789674952198853,
      "grad_norm": 1.641012191772461,
      "learning_rate": 2.3369449755683023e-05,
      "loss": 0.7631,
      "step": 5120
    },
    {
      "epoch": 9.82791586998088,
      "grad_norm": 1.7024327516555786,
      "learning_rate": 1.9120458891013384e-05,
      "loss": 0.7153,
      "step": 5140
    },
    {
      "epoch": 9.866156787762906,
      "grad_norm": 1.4840829372406006,
      "learning_rate": 1.4871468026343743e-05,
      "loss": 0.7247,
      "step": 5160
    },
    {
      "epoch": 9.904397705544934,
      "grad_norm": 1.627562165260315,
      "learning_rate": 1.0622477161674103e-05,
      "loss": 0.7289,
      "step": 5180
    },
    {
      "epoch": 9.94263862332696,
      "grad_norm": 1.6473079919815063,
      "learning_rate": 6.373486297004461e-06,
      "loss": 0.7563,
      "step": 5200
    },
    {
      "epoch": 9.980879541108987,
      "grad_norm": 1.577776312828064,
      "learning_rate": 2.1244954323348204e-06,
      "loss": 0.7603,
      "step": 5220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9410023545240498,
      "eval_loss": 0.29460111260414124,
      "eval_runtime": 187.2975,
      "eval_samples_per_second": 79.366,
      "eval_steps_per_second": 79.366,
      "step": 5230
    },
    {
      "epoch": 10.0,
      "step": 5230,
      "total_flos": 1.96318398191328e+18,
      "train_loss": 2.1099621225725396,
      "train_runtime": 19361.582,
      "train_samples_per_second": 69.094,
      "train_steps_per_second": 0.27
    }
  ],
  "logging_steps": 20,
  "max_steps": 5230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 1.96318398191328e+18,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}