{
  "best_metric": 0.022856945171952248,
  "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles1_seed1_classic_image_classification_local\\checkpoint-1407",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2345,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010660980810234541,
      "grad_norm": 1.3030591011047363,
      "learning_rate": 4.989339019189766e-05,
      "loss": 0.6439,
      "step": 5
    },
    {
      "epoch": 0.021321961620469083,
      "grad_norm": 1.8761012554168701,
      "learning_rate": 4.978678038379531e-05,
      "loss": 0.488,
      "step": 10
    },
    {
      "epoch": 0.031982942430703626,
      "grad_norm": 1.1793400049209595,
      "learning_rate": 4.9680170575692967e-05,
      "loss": 0.337,
      "step": 15
    },
    {
      "epoch": 0.042643923240938165,
      "grad_norm": 1.3804587125778198,
      "learning_rate": 4.957356076759062e-05,
      "loss": 0.2404,
      "step": 20
    },
    {
      "epoch": 0.053304904051172705,
      "grad_norm": 0.987789511680603,
      "learning_rate": 4.9466950959488276e-05,
      "loss": 0.1539,
      "step": 25
    },
    {
      "epoch": 0.06396588486140725,
      "grad_norm": 2.2496049404144287,
      "learning_rate": 4.936034115138593e-05,
      "loss": 0.1444,
      "step": 30
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.39667999744415283,
      "learning_rate": 4.9253731343283586e-05,
      "loss": 0.079,
      "step": 35
    },
    {
      "epoch": 0.08528784648187633,
      "grad_norm": 1.3956135511398315,
      "learning_rate": 4.914712153518124e-05,
      "loss": 0.0746,
      "step": 40
    },
    {
      "epoch": 0.09594882729211088,
      "grad_norm": 4.631561756134033,
      "learning_rate": 4.904051172707889e-05,
      "loss": 0.0829,
      "step": 45
    },
    {
      "epoch": 0.10660980810234541,
      "grad_norm": 2.049582004547119,
      "learning_rate": 4.893390191897655e-05,
      "loss": 0.0496,
      "step": 50
    },
    {
      "epoch": 0.11727078891257996,
      "grad_norm": 3.339857816696167,
      "learning_rate": 4.88272921108742e-05,
      "loss": 0.0542,
      "step": 55
    },
    {
      "epoch": 0.1279317697228145,
      "grad_norm": 3.415274143218994,
      "learning_rate": 4.872068230277186e-05,
      "loss": 0.1034,
      "step": 60
    },
    {
      "epoch": 0.13859275053304904,
      "grad_norm": 0.7254932522773743,
      "learning_rate": 4.861407249466951e-05,
      "loss": 0.1021,
      "step": 65
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.7911928296089172,
      "learning_rate": 4.850746268656717e-05,
      "loss": 0.0581,
      "step": 70
    },
    {
      "epoch": 0.15991471215351813,
      "grad_norm": 2.9554224014282227,
      "learning_rate": 4.840085287846482e-05,
      "loss": 0.0626,
      "step": 75
    },
    {
      "epoch": 0.17057569296375266,
      "grad_norm": 0.17032524943351746,
      "learning_rate": 4.829424307036248e-05,
      "loss": 0.0366,
      "step": 80
    },
    {
      "epoch": 0.1812366737739872,
      "grad_norm": 4.795769214630127,
      "learning_rate": 4.8187633262260126e-05,
      "loss": 0.0359,
      "step": 85
    },
    {
      "epoch": 0.19189765458422176,
      "grad_norm": 0.13845813274383545,
      "learning_rate": 4.808102345415779e-05,
      "loss": 0.0231,
      "step": 90
    },
    {
      "epoch": 0.2025586353944563,
      "grad_norm": 0.14712730050086975,
      "learning_rate": 4.7974413646055436e-05,
      "loss": 0.046,
      "step": 95
    },
    {
      "epoch": 0.21321961620469082,
      "grad_norm": 0.1327529400587082,
      "learning_rate": 4.78678038379531e-05,
      "loss": 0.0522,
      "step": 100
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.12169656902551651,
      "learning_rate": 4.7761194029850745e-05,
      "loss": 0.039,
      "step": 105
    },
    {
      "epoch": 0.2345415778251599,
      "grad_norm": 4.029494762420654,
      "learning_rate": 4.765458422174841e-05,
      "loss": 0.0986,
      "step": 110
    },
    {
      "epoch": 0.24520255863539445,
      "grad_norm": 0.569814145565033,
      "learning_rate": 4.7547974413646055e-05,
      "loss": 0.0501,
      "step": 115
    },
    {
      "epoch": 0.255863539445629,
      "grad_norm": 2.2544350624084473,
      "learning_rate": 4.7441364605543716e-05,
      "loss": 0.0229,
      "step": 120
    },
    {
      "epoch": 0.26652452025586354,
      "grad_norm": 1.0173734426498413,
      "learning_rate": 4.7334754797441364e-05,
      "loss": 0.0453,
      "step": 125
    },
    {
      "epoch": 0.2771855010660981,
      "grad_norm": 0.32047396898269653,
      "learning_rate": 4.7228144989339026e-05,
      "loss": 0.0371,
      "step": 130
    },
    {
      "epoch": 0.2878464818763326,
      "grad_norm": 2.2969608306884766,
      "learning_rate": 4.7121535181236674e-05,
      "loss": 0.0379,
      "step": 135
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.11387304216623306,
      "learning_rate": 4.7014925373134335e-05,
      "loss": 0.0304,
      "step": 140
    },
    {
      "epoch": 0.3091684434968017,
      "grad_norm": 0.2686951756477356,
      "learning_rate": 4.690831556503198e-05,
      "loss": 0.0677,
      "step": 145
    },
    {
      "epoch": 0.31982942430703626,
      "grad_norm": 0.10119659453630447,
      "learning_rate": 4.6801705756929645e-05,
      "loss": 0.0423,
      "step": 150
    },
    {
      "epoch": 0.3304904051172708,
      "grad_norm": 3.7759270668029785,
      "learning_rate": 4.669509594882729e-05,
      "loss": 0.0497,
      "step": 155
    },
    {
      "epoch": 0.3411513859275053,
      "grad_norm": 0.09366951137781143,
      "learning_rate": 4.658848614072495e-05,
      "loss": 0.0196,
      "step": 160
    },
    {
      "epoch": 0.35181236673773986,
      "grad_norm": 0.3446415364742279,
      "learning_rate": 4.64818763326226e-05,
      "loss": 0.0411,
      "step": 165
    },
    {
      "epoch": 0.3624733475479744,
      "grad_norm": 0.0982455462217331,
      "learning_rate": 4.637526652452026e-05,
      "loss": 0.0206,
      "step": 170
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 2.6638343334198,
      "learning_rate": 4.626865671641791e-05,
      "loss": 0.0489,
      "step": 175
    },
    {
      "epoch": 0.3837953091684435,
      "grad_norm": 0.2727707326412201,
      "learning_rate": 4.6162046908315566e-05,
      "loss": 0.0153,
      "step": 180
    },
    {
      "epoch": 0.39445628997867804,
      "grad_norm": 0.17455758154392242,
      "learning_rate": 4.605543710021322e-05,
      "loss": 0.0446,
      "step": 185
    },
    {
      "epoch": 0.4051172707889126,
      "grad_norm": 0.10797569900751114,
      "learning_rate": 4.5948827292110876e-05,
      "loss": 0.0139,
      "step": 190
    },
    {
      "epoch": 0.4157782515991471,
      "grad_norm": 0.802625298500061,
      "learning_rate": 4.584221748400853e-05,
      "loss": 0.0366,
      "step": 195
    },
    {
      "epoch": 0.42643923240938164,
      "grad_norm": 0.33289510011672974,
      "learning_rate": 4.5735607675906185e-05,
      "loss": 0.0832,
      "step": 200
    },
    {
      "epoch": 0.43710021321961623,
      "grad_norm": 0.07750815153121948,
      "learning_rate": 4.562899786780384e-05,
      "loss": 0.017,
      "step": 205
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.07023430615663528,
      "learning_rate": 4.5522388059701495e-05,
      "loss": 0.0474,
      "step": 210
    },
    {
      "epoch": 0.4584221748400853,
      "grad_norm": 7.5608625411987305,
      "learning_rate": 4.541577825159915e-05,
      "loss": 0.0238,
      "step": 215
    },
    {
      "epoch": 0.4690831556503198,
      "grad_norm": 0.07039786130189896,
      "learning_rate": 4.5309168443496804e-05,
      "loss": 0.0177,
      "step": 220
    },
    {
      "epoch": 0.47974413646055436,
      "grad_norm": 1.5955876111984253,
      "learning_rate": 4.520255863539446e-05,
      "loss": 0.0529,
      "step": 225
    },
    {
      "epoch": 0.4904051172707889,
      "grad_norm": 0.29084673523902893,
      "learning_rate": 4.5095948827292114e-05,
      "loss": 0.0389,
      "step": 230
    },
    {
      "epoch": 0.5010660980810234,
      "grad_norm": 4.458634853363037,
      "learning_rate": 4.498933901918977e-05,
      "loss": 0.028,
      "step": 235
    },
    {
      "epoch": 0.511727078891258,
      "grad_norm": 0.24322013556957245,
      "learning_rate": 4.488272921108742e-05,
      "loss": 0.014,
      "step": 240
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 1.5862743854522705,
      "learning_rate": 4.477611940298508e-05,
      "loss": 0.0273,
      "step": 245
    },
    {
      "epoch": 0.5330490405117271,
      "grad_norm": 2.472909927368164,
      "learning_rate": 4.466950959488273e-05,
      "loss": 0.0572,
      "step": 250
    },
    {
      "epoch": 0.5437100213219617,
      "grad_norm": 0.1901959925889969,
      "learning_rate": 4.456289978678039e-05,
      "loss": 0.0095,
      "step": 255
    },
    {
      "epoch": 0.5543710021321961,
      "grad_norm": 1.9824448823928833,
      "learning_rate": 4.445628997867804e-05,
      "loss": 0.0374,
      "step": 260
    },
    {
      "epoch": 0.5650319829424307,
      "grad_norm": 0.0593295581638813,
      "learning_rate": 4.43496801705757e-05,
      "loss": 0.0357,
      "step": 265
    },
    {
      "epoch": 0.5756929637526652,
      "grad_norm": 0.060530565679073334,
      "learning_rate": 4.424307036247335e-05,
      "loss": 0.0222,
      "step": 270
    },
    {
      "epoch": 0.5863539445628998,
      "grad_norm": 1.5108375549316406,
      "learning_rate": 4.4136460554371006e-05,
      "loss": 0.0104,
      "step": 275
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 1.9284145832061768,
      "learning_rate": 4.402985074626866e-05,
      "loss": 0.0312,
      "step": 280
    },
    {
      "epoch": 0.6076759061833689,
      "grad_norm": 0.061266589909791946,
      "learning_rate": 4.3923240938166316e-05,
      "loss": 0.0625,
      "step": 285
    },
    {
      "epoch": 0.6183368869936035,
      "grad_norm": 0.05914849042892456,
      "learning_rate": 4.381663113006397e-05,
      "loss": 0.0446,
      "step": 290
    },
    {
      "epoch": 0.6289978678038379,
      "grad_norm": 0.4006780982017517,
      "learning_rate": 4.3710021321961625e-05,
      "loss": 0.0497,
      "step": 295
    },
    {
      "epoch": 0.6396588486140725,
      "grad_norm": 0.0631353035569191,
      "learning_rate": 4.360341151385928e-05,
      "loss": 0.0098,
      "step": 300
    },
    {
      "epoch": 0.650319829424307,
      "grad_norm": 0.2760462760925293,
      "learning_rate": 4.3496801705756935e-05,
      "loss": 0.0444,
      "step": 305
    },
    {
      "epoch": 0.6609808102345416,
      "grad_norm": 0.05796327069401741,
      "learning_rate": 4.339019189765459e-05,
      "loss": 0.0107,
      "step": 310
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 5.437504291534424,
      "learning_rate": 4.328358208955224e-05,
      "loss": 0.027,
      "step": 315
    },
    {
      "epoch": 0.6823027718550106,
      "grad_norm": 0.05112091079354286,
      "learning_rate": 4.31769722814499e-05,
      "loss": 0.0296,
      "step": 320
    },
    {
      "epoch": 0.6929637526652452,
      "grad_norm": 0.0484708808362484,
      "learning_rate": 4.307036247334755e-05,
      "loss": 0.0419,
      "step": 325
    },
    {
      "epoch": 0.7036247334754797,
      "grad_norm": 0.05781492963433266,
      "learning_rate": 4.29637526652452e-05,
      "loss": 0.0191,
      "step": 330
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.07626837491989136,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.032,
      "step": 335
    },
    {
      "epoch": 0.7249466950959488,
      "grad_norm": 0.046046238392591476,
      "learning_rate": 4.275053304904051e-05,
      "loss": 0.0315,
      "step": 340
    },
    {
      "epoch": 0.7356076759061834,
      "grad_norm": 0.07111696898937225,
      "learning_rate": 4.2643923240938166e-05,
      "loss": 0.0617,
      "step": 345
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.2557612359523773,
      "learning_rate": 4.253731343283582e-05,
      "loss": 0.0183,
      "step": 350
    },
    {
      "epoch": 0.7569296375266524,
      "grad_norm": 0.20927634835243225,
      "learning_rate": 4.2430703624733475e-05,
      "loss": 0.0275,
      "step": 355
    },
    {
      "epoch": 0.767590618336887,
      "grad_norm": 1.5735142230987549,
      "learning_rate": 4.232409381663113e-05,
      "loss": 0.0319,
      "step": 360
    },
    {
      "epoch": 0.7782515991471215,
      "grad_norm": 3.6991448402404785,
      "learning_rate": 4.2217484008528785e-05,
      "loss": 0.0247,
      "step": 365
    },
    {
      "epoch": 0.7889125799573561,
      "grad_norm": 0.05717187374830246,
      "learning_rate": 4.211087420042644e-05,
      "loss": 0.0127,
      "step": 370
    },
    {
      "epoch": 0.7995735607675906,
      "grad_norm": 0.07145973294973373,
      "learning_rate": 4.2004264392324094e-05,
      "loss": 0.0091,
      "step": 375
    },
    {
      "epoch": 0.8102345415778252,
      "grad_norm": 4.079801082611084,
      "learning_rate": 4.189765458422175e-05,
      "loss": 0.0135,
      "step": 380
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.5012163519859314,
      "learning_rate": 4.1791044776119404e-05,
      "loss": 0.0447,
      "step": 385
    },
    {
      "epoch": 0.8315565031982942,
      "grad_norm": 0.07103706151247025,
      "learning_rate": 4.168443496801706e-05,
      "loss": 0.0522,
      "step": 390
    },
    {
      "epoch": 0.8422174840085288,
      "grad_norm": 0.045256324112415314,
      "learning_rate": 4.157782515991471e-05,
      "loss": 0.0238,
      "step": 395
    },
    {
      "epoch": 0.8528784648187633,
      "grad_norm": 1.065915584564209,
      "learning_rate": 4.147121535181237e-05,
      "loss": 0.0762,
      "step": 400
    },
    {
      "epoch": 0.8635394456289979,
      "grad_norm": 7.5331597328186035,
      "learning_rate": 4.136460554371002e-05,
      "loss": 0.0476,
      "step": 405
    },
    {
      "epoch": 0.8742004264392325,
      "grad_norm": 3.337735176086426,
      "learning_rate": 4.125799573560768e-05,
      "loss": 0.0595,
      "step": 410
    },
    {
      "epoch": 0.8848614072494669,
      "grad_norm": 0.09289413690567017,
      "learning_rate": 4.115138592750533e-05,
      "loss": 0.0096,
      "step": 415
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.04636334627866745,
      "learning_rate": 4.104477611940299e-05,
      "loss": 0.0262,
      "step": 420
    },
    {
      "epoch": 0.906183368869936,
      "grad_norm": 5.231013298034668,
      "learning_rate": 4.093816631130064e-05,
      "loss": 0.0344,
      "step": 425
    },
    {
      "epoch": 0.9168443496801706,
      "grad_norm": 0.558976948261261,
      "learning_rate": 4.0831556503198296e-05,
      "loss": 0.0635,
      "step": 430
    },
    {
      "epoch": 0.9275053304904051,
      "grad_norm": 0.05399622395634651,
      "learning_rate": 4.072494669509595e-05,
      "loss": 0.0208,
      "step": 435
    },
    {
      "epoch": 0.9381663113006397,
      "grad_norm": 0.04404484108090401,
      "learning_rate": 4.0618336886993606e-05,
      "loss": 0.0312,
      "step": 440
    },
    {
      "epoch": 0.9488272921108742,
      "grad_norm": 0.3174332082271576,
      "learning_rate": 4.051172707889126e-05,
      "loss": 0.0447,
      "step": 445
    },
    {
      "epoch": 0.9594882729211087,
      "grad_norm": 0.06034679710865021,
      "learning_rate": 4.0405117270788915e-05,
      "loss": 0.006,
      "step": 450
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 2.384519577026367,
      "learning_rate": 4.029850746268657e-05,
      "loss": 0.0217,
      "step": 455
    },
    {
      "epoch": 0.9808102345415778,
      "grad_norm": 2.445675849914551,
      "learning_rate": 4.0191897654584225e-05,
      "loss": 0.0107,
      "step": 460
    },
    {
      "epoch": 0.9914712153518124,
      "grad_norm": 1.2337150573730469,
      "learning_rate": 4.008528784648188e-05,
      "loss": 0.0735,
      "step": 465
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9762666666666666,
      "eval_loss": 0.07783844321966171,
      "eval_runtime": 16.5274,
      "eval_samples_per_second": 226.896,
      "eval_steps_per_second": 7.14,
      "step": 469
    },
    {
      "epoch": 1.0021321961620469,
      "grad_norm": 0.03896041214466095,
      "learning_rate": 3.997867803837953e-05,
      "loss": 0.0095,
      "step": 470
    },
    {
      "epoch": 1.0127931769722816,
      "grad_norm": 0.30038806796073914,
      "learning_rate": 3.987206823027719e-05,
      "loss": 0.0497,
      "step": 475
    },
    {
      "epoch": 1.023454157782516,
      "grad_norm": 0.04247400909662247,
      "learning_rate": 3.976545842217484e-05,
      "loss": 0.0644,
      "step": 480
    },
    {
      "epoch": 1.0341151385927505,
      "grad_norm": 0.1584388017654419,
      "learning_rate": 3.96588486140725e-05,
      "loss": 0.0136,
      "step": 485
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 1.492022156715393,
      "learning_rate": 3.9552238805970146e-05,
      "loss": 0.0354,
      "step": 490
    },
    {
      "epoch": 1.0554371002132197,
      "grad_norm": 0.07293384522199631,
      "learning_rate": 3.944562899786781e-05,
      "loss": 0.0095,
      "step": 495
    },
    {
      "epoch": 1.0660980810234542,
      "grad_norm": 0.12515315413475037,
      "learning_rate": 3.9339019189765456e-05,
      "loss": 0.0061,
      "step": 500
    },
    {
      "epoch": 1.0767590618336886,
      "grad_norm": 0.04049481078982353,
      "learning_rate": 3.923240938166312e-05,
      "loss": 0.0048,
      "step": 505
    },
    {
      "epoch": 1.0874200426439233,
      "grad_norm": 0.3493141531944275,
      "learning_rate": 3.9125799573560765e-05,
      "loss": 0.0072,
      "step": 510
    },
    {
      "epoch": 1.0980810234541578,
      "grad_norm": 0.040050432085990906,
      "learning_rate": 3.901918976545843e-05,
      "loss": 0.0062,
      "step": 515
    },
    {
      "epoch": 1.1087420042643923,
      "grad_norm": 0.03947841376066208,
      "learning_rate": 3.8912579957356075e-05,
      "loss": 0.0043,
      "step": 520
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 0.03871216997504234,
      "learning_rate": 3.8805970149253736e-05,
      "loss": 0.0341,
      "step": 525
    },
    {
      "epoch": 1.1300639658848615,
      "grad_norm": 0.03848418593406677,
      "learning_rate": 3.8699360341151384e-05,
      "loss": 0.0061,
      "step": 530
    },
    {
      "epoch": 1.140724946695096,
      "grad_norm": 0.03081074170768261,
      "learning_rate": 3.8592750533049046e-05,
      "loss": 0.0044,
      "step": 535
    },
    {
      "epoch": 1.1513859275053304,
      "grad_norm": 0.036050815135240555,
      "learning_rate": 3.8486140724946694e-05,
      "loss": 0.0044,
      "step": 540
    },
    {
      "epoch": 1.1620469083155651,
      "grad_norm": 0.27078014612197876,
      "learning_rate": 3.8379530916844355e-05,
      "loss": 0.0193,
      "step": 545
    },
    {
      "epoch": 1.1727078891257996,
      "grad_norm": 0.03798219561576843,
      "learning_rate": 3.8272921108742e-05,
      "loss": 0.0042,
      "step": 550
    },
    {
      "epoch": 1.183368869936034,
      "grad_norm": 0.03544671833515167,
      "learning_rate": 3.8166311300639665e-05,
      "loss": 0.0074,
      "step": 555
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 0.028461942449212074,
      "learning_rate": 3.805970149253731e-05,
      "loss": 0.007,
      "step": 560
    },
    {
      "epoch": 1.2046908315565032,
      "grad_norm": 0.03148175776004791,
      "learning_rate": 3.7953091684434974e-05,
      "loss": 0.0044,
      "step": 565
    },
    {
      "epoch": 1.2153518123667377,
      "grad_norm": 0.027858976274728775,
      "learning_rate": 3.784648187633262e-05,
      "loss": 0.0038,
      "step": 570
    },
    {
      "epoch": 1.2260127931769722,
      "grad_norm": 3.376032590866089,
      "learning_rate": 3.7739872068230284e-05,
      "loss": 0.0057,
      "step": 575
    },
    {
      "epoch": 1.236673773987207,
      "grad_norm": 0.026746030896902084,
      "learning_rate": 3.763326226012793e-05,
      "loss": 0.0035,
      "step": 580
    },
    {
      "epoch": 1.2473347547974414,
      "grad_norm": 0.03644810989499092,
      "learning_rate": 3.752665245202559e-05,
      "loss": 0.0047,
      "step": 585
    },
    {
      "epoch": 1.2579957356076759,
      "grad_norm": 0.7633171081542969,
      "learning_rate": 3.742004264392324e-05,
      "loss": 0.0039,
      "step": 590
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 0.027774915099143982,
      "learning_rate": 3.73134328358209e-05,
      "loss": 0.0036,
      "step": 595
    },
    {
      "epoch": 1.279317697228145,
      "grad_norm": 1.0888952016830444,
      "learning_rate": 3.720682302771855e-05,
      "loss": 0.0112,
      "step": 600
    },
    {
      "epoch": 1.2899786780383795,
      "grad_norm": 0.030503449961543083,
      "learning_rate": 3.710021321961621e-05,
      "loss": 0.0082,
      "step": 605
    },
    {
      "epoch": 1.3006396588486142,
      "grad_norm": 0.025810692459344864,
      "learning_rate": 3.699360341151386e-05,
      "loss": 0.0036,
      "step": 610
    },
    {
      "epoch": 1.3113006396588487,
      "grad_norm": 0.024351516738533974,
      "learning_rate": 3.6886993603411515e-05,
      "loss": 0.0336,
      "step": 615
    },
    {
      "epoch": 1.3219616204690832,
      "grad_norm": 0.0555342361330986,
      "learning_rate": 3.678038379530917e-05,
      "loss": 0.0036,
      "step": 620
    },
    {
      "epoch": 1.3326226012793176,
      "grad_norm": 0.06257825344800949,
      "learning_rate": 3.6673773987206824e-05,
      "loss": 0.0084,
      "step": 625
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 0.02445884980261326,
      "learning_rate": 3.656716417910448e-05,
      "loss": 0.0045,
      "step": 630
    },
    {
      "epoch": 1.3539445628997868,
      "grad_norm": 0.025071511045098305,
      "learning_rate": 3.6460554371002134e-05,
      "loss": 0.003,
      "step": 635
    },
    {
      "epoch": 1.3646055437100213,
      "grad_norm": 0.15359072387218475,
      "learning_rate": 3.635394456289979e-05,
      "loss": 0.0034,
      "step": 640
    },
    {
      "epoch": 1.375266524520256,
      "grad_norm": 0.631420373916626,
      "learning_rate": 3.624733475479744e-05,
      "loss": 0.0035,
      "step": 645
    },
    {
      "epoch": 1.3859275053304905,
      "grad_norm": 0.024551689624786377,
      "learning_rate": 3.61407249466951e-05,
      "loss": 0.0029,
      "step": 650
    },
    {
      "epoch": 1.396588486140725,
      "grad_norm": 0.02399507351219654,
      "learning_rate": 3.603411513859275e-05,
      "loss": 0.0057,
      "step": 655
    },
    {
      "epoch": 1.4072494669509594,
      "grad_norm": 6.411247730255127,
      "learning_rate": 3.592750533049041e-05,
      "loss": 0.0067,
      "step": 660
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 0.022543814033269882,
      "learning_rate": 3.582089552238806e-05,
      "loss": 0.027,
      "step": 665
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.021749133244156837,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.0116,
      "step": 670
    },
    {
      "epoch": 1.439232409381663,
      "grad_norm": 0.022446081042289734,
      "learning_rate": 3.560767590618337e-05,
      "loss": 0.0032,
      "step": 675
    },
    {
      "epoch": 1.4498933901918978,
      "grad_norm": 0.14957591891288757,
      "learning_rate": 3.5501066098081026e-05,
      "loss": 0.0227,
      "step": 680
    },
    {
      "epoch": 1.4605543710021323,
      "grad_norm": 0.2059992253780365,
      "learning_rate": 3.539445628997868e-05,
      "loss": 0.0042,
      "step": 685
    },
    {
      "epoch": 1.4712153518123667,
      "grad_norm": 0.07327228784561157,
      "learning_rate": 3.5287846481876336e-05,
      "loss": 0.0033,
      "step": 690
    },
    {
      "epoch": 1.4818763326226012,
      "grad_norm": 0.022074097767472267,
      "learning_rate": 3.518123667377399e-05,
      "loss": 0.0031,
      "step": 695
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 0.035878945142030716,
      "learning_rate": 3.5074626865671645e-05,
      "loss": 0.0348,
      "step": 700
    },
    {
      "epoch": 1.5031982942430704,
      "grad_norm": 0.020859839394688606,
      "learning_rate": 3.496801705756929e-05,
      "loss": 0.0816,
      "step": 705
    },
    {
      "epoch": 1.5138592750533049,
      "grad_norm": 0.024958930909633636,
      "learning_rate": 3.4861407249466955e-05,
      "loss": 0.0028,
      "step": 710
    },
    {
      "epoch": 1.5245202558635396,
      "grad_norm": 0.02834257483482361,
      "learning_rate": 3.47547974413646e-05,
      "loss": 0.0027,
      "step": 715
    },
    {
      "epoch": 1.535181236673774,
      "grad_norm": 0.022769009694457054,
      "learning_rate": 3.4648187633262264e-05,
      "loss": 0.0174,
      "step": 720
    },
    {
      "epoch": 1.5458422174840085,
      "grad_norm": 0.02601616457104683,
      "learning_rate": 3.454157782515991e-05,
      "loss": 0.0027,
      "step": 725
    },
    {
      "epoch": 1.556503198294243,
      "grad_norm": 0.24697370827198029,
      "learning_rate": 3.4434968017057574e-05,
      "loss": 0.0056,
      "step": 730
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 0.01992082968354225,
      "learning_rate": 3.432835820895522e-05,
      "loss": 0.0056,
      "step": 735
    },
    {
      "epoch": 1.5778251599147122,
      "grad_norm": 0.020425518974661827,
      "learning_rate": 3.422174840085288e-05,
      "loss": 0.0026,
      "step": 740
    },
    {
      "epoch": 1.5884861407249466,
      "grad_norm": 0.021747827529907227,
      "learning_rate": 3.411513859275053e-05,
      "loss": 0.0034,
      "step": 745
    },
    {
      "epoch": 1.5991471215351813,
      "grad_norm": 0.02213311195373535,
      "learning_rate": 3.400852878464819e-05,
      "loss": 0.0389,
      "step": 750
    },
    {
      "epoch": 1.6098081023454158,
      "grad_norm": 0.01999504491686821,
      "learning_rate": 3.390191897654584e-05,
      "loss": 0.0043,
      "step": 755
    },
    {
      "epoch": 1.6204690831556503,
      "grad_norm": 0.019559355452656746,
      "learning_rate": 3.37953091684435e-05,
      "loss": 0.0026,
      "step": 760
    },
    {
      "epoch": 1.6311300639658848,
      "grad_norm": 0.019753890112042427,
      "learning_rate": 3.368869936034115e-05,
      "loss": 0.0024,
      "step": 765
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 0.024971093982458115,
      "learning_rate": 3.358208955223881e-05,
      "loss": 0.0026,
      "step": 770
    },
    {
      "epoch": 1.652452025586354,
      "grad_norm": 0.109837107360363,
      "learning_rate": 3.347547974413646e-05,
      "loss": 0.0296,
      "step": 775
    },
    {
      "epoch": 1.6631130063965884,
      "grad_norm": 0.01920832134783268,
      "learning_rate": 3.336886993603412e-05,
      "loss": 0.0024,
      "step": 780
    },
    {
      "epoch": 1.6737739872068231,
      "grad_norm": 0.019149158149957657,
      "learning_rate": 3.326226012793177e-05,
      "loss": 0.0026,
      "step": 785
    },
    {
      "epoch": 1.6844349680170576,
      "grad_norm": 0.01869184523820877,
      "learning_rate": 3.3155650319829424e-05,
      "loss": 0.0034,
      "step": 790
    },
    {
      "epoch": 1.695095948827292,
      "grad_norm": 0.22865666449069977,
      "learning_rate": 3.304904051172708e-05,
      "loss": 0.0027,
      "step": 795
    },
    {
      "epoch": 1.7057569296375266,
      "grad_norm": 0.020090360194444656,
      "learning_rate": 3.294243070362473e-05,
      "loss": 0.0024,
      "step": 800
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 0.018748512491583824,
      "learning_rate": 3.283582089552239e-05,
      "loss": 0.0022,
      "step": 805
    },
    {
      "epoch": 1.7270788912579957,
      "grad_norm": 0.020859690383076668,
      "learning_rate": 3.272921108742004e-05,
      "loss": 0.0265,
      "step": 810
    },
    {
      "epoch": 1.7377398720682304,
      "grad_norm": 0.020721951499581337,
      "learning_rate": 3.26226012793177e-05,
      "loss": 0.0025,
      "step": 815
    },
    {
      "epoch": 1.748400852878465,
      "grad_norm": 0.019519254565238953,
      "learning_rate": 3.251599147121535e-05,
      "loss": 0.0022,
      "step": 820
    },
    {
      "epoch": 1.7590618336886994,
      "grad_norm": 0.019015362486243248,
      "learning_rate": 3.240938166311301e-05,
      "loss": 0.0276,
      "step": 825
    },
    {
      "epoch": 1.7697228144989339,
      "grad_norm": 7.688128471374512,
      "learning_rate": 3.230277185501066e-05,
      "loss": 0.0149,
      "step": 830
    },
    {
      "epoch": 1.7803837953091683,
      "grad_norm": 0.017749857157468796,
      "learning_rate": 3.2196162046908317e-05,
      "loss": 0.0024,
      "step": 835
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 0.017808672040700912,
      "learning_rate": 3.208955223880597e-05,
      "loss": 0.0023,
      "step": 840
    },
    {
      "epoch": 1.8017057569296375,
      "grad_norm": 0.0173553004860878,
      "learning_rate": 3.1982942430703626e-05,
      "loss": 0.0021,
      "step": 845
    },
    {
      "epoch": 1.8123667377398722,
      "grad_norm": 0.018168503418564796,
      "learning_rate": 3.187633262260128e-05,
      "loss": 0.0026,
      "step": 850
    },
    {
      "epoch": 1.8230277185501067,
      "grad_norm": 0.016489172354340553,
      "learning_rate": 3.1769722814498935e-05,
      "loss": 0.0045,
      "step": 855
    },
    {
      "epoch": 1.8336886993603412,
      "grad_norm": 0.018007785081863403,
      "learning_rate": 3.166311300639659e-05,
      "loss": 0.0021,
      "step": 860
    },
    {
      "epoch": 1.8443496801705757,
      "grad_norm": 0.017267387360334396,
      "learning_rate": 3.1556503198294245e-05,
      "loss": 0.0023,
      "step": 865
    },
    {
      "epoch": 1.8550106609808101,
      "grad_norm": 0.016420844942331314,
      "learning_rate": 3.14498933901919e-05,
      "loss": 0.0104,
      "step": 870
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 0.016852952539920807,
      "learning_rate": 3.1343283582089554e-05,
      "loss": 0.002,
      "step": 875
    },
    {
      "epoch": 1.8763326226012793,
      "grad_norm": 0.017920179292559624,
      "learning_rate": 3.123667377398721e-05,
      "loss": 0.002,
      "step": 880
    },
    {
      "epoch": 1.886993603411514,
      "grad_norm": 0.016802307218313217,
      "learning_rate": 3.1130063965884864e-05,
      "loss": 0.0019,
      "step": 885
    },
    {
      "epoch": 1.8976545842217485,
      "grad_norm": 0.016610046848654747,
      "learning_rate": 3.102345415778252e-05,
      "loss": 0.002,
      "step": 890
    },
    {
      "epoch": 1.908315565031983,
      "grad_norm": 0.019902439787983894,
      "learning_rate": 3.0916844349680173e-05,
      "loss": 0.0021,
      "step": 895
    },
    {
      "epoch": 1.9189765458422174,
      "grad_norm": 0.01531369797885418,
      "learning_rate": 3.081023454157783e-05,
      "loss": 0.0025,
      "step": 900
    },
    {
      "epoch": 1.929637526652452,
      "grad_norm": 0.015508510172367096,
      "learning_rate": 3.070362473347548e-05,
      "loss": 0.0019,
      "step": 905
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 0.5575276613235474,
      "learning_rate": 3.059701492537314e-05,
      "loss": 0.0122,
      "step": 910
    },
    {
      "epoch": 1.950959488272921,
      "grad_norm": 0.015613945201039314,
      "learning_rate": 3.0490405117270792e-05,
      "loss": 0.0019,
      "step": 915
    },
    {
      "epoch": 1.9616204690831558,
      "grad_norm": 0.015503398142755032,
      "learning_rate": 3.0383795309168444e-05,
      "loss": 0.0055,
      "step": 920
    },
    {
      "epoch": 1.9722814498933903,
      "grad_norm": 3.0477633476257324,
      "learning_rate": 3.0277185501066102e-05,
      "loss": 0.0294,
      "step": 925
    },
    {
      "epoch": 1.9829424307036247,
      "grad_norm": 0.015289404429495335,
      "learning_rate": 3.0170575692963753e-05,
      "loss": 0.0018,
      "step": 930
    },
    {
      "epoch": 1.9936034115138592,
      "grad_norm": 6.750314235687256,
      "learning_rate": 3.006396588486141e-05,
      "loss": 0.0205,
      "step": 935
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9896,
      "eval_loss": 0.04148600623011589,
      "eval_runtime": 17.7849,
      "eval_samples_per_second": 210.852,
      "eval_steps_per_second": 6.635,
      "step": 938
    },
    {
      "epoch": 2.0042643923240937,
      "grad_norm": 0.08804287761449814,
      "learning_rate": 2.9957356076759063e-05,
      "loss": 0.0022,
      "step": 940
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 0.015377379953861237,
      "learning_rate": 2.9850746268656714e-05,
      "loss": 0.0326,
      "step": 945
    },
    {
      "epoch": 2.025586353944563,
      "grad_norm": 0.024583661928772926,
      "learning_rate": 2.9744136460554372e-05,
      "loss": 0.0019,
      "step": 950
    },
    {
      "epoch": 2.0362473347547976,
      "grad_norm": 0.015179556794464588,
      "learning_rate": 2.9637526652452023e-05,
      "loss": 0.0019,
      "step": 955
    },
    {
      "epoch": 2.046908315565032,
      "grad_norm": 0.014737558551132679,
      "learning_rate": 2.953091684434968e-05,
      "loss": 0.0018,
      "step": 960
    },
    {
      "epoch": 2.0575692963752665,
      "grad_norm": 0.059039849787950516,
      "learning_rate": 2.9424307036247333e-05,
      "loss": 0.0022,
      "step": 965
    },
    {
      "epoch": 2.068230277185501,
      "grad_norm": 0.014863478019833565,
      "learning_rate": 2.931769722814499e-05,
      "loss": 0.0018,
      "step": 970
    },
    {
      "epoch": 2.0788912579957355,
      "grad_norm": 0.015474963933229446,
      "learning_rate": 2.9211087420042642e-05,
      "loss": 0.0018,
      "step": 975
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 0.015029369853436947,
      "learning_rate": 2.91044776119403e-05,
      "loss": 0.0017,
      "step": 980
    },
    {
      "epoch": 2.100213219616205,
      "grad_norm": 0.014501337893307209,
      "learning_rate": 2.8997867803837952e-05,
      "loss": 0.0017,
      "step": 985
    },
    {
      "epoch": 2.1108742004264394,
      "grad_norm": 0.014287133701145649,
      "learning_rate": 2.889125799573561e-05,
      "loss": 0.0017,
      "step": 990
    },
    {
      "epoch": 2.121535181236674,
      "grad_norm": 0.013947279192507267,
      "learning_rate": 2.878464818763326e-05,
      "loss": 0.0032,
      "step": 995
    },
    {
      "epoch": 2.1321961620469083,
      "grad_norm": 0.0139941880479455,
      "learning_rate": 2.867803837953092e-05,
      "loss": 0.002,
      "step": 1000
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.014081170782446861,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.0017,
      "step": 1005
    },
    {
      "epoch": 2.1535181236673773,
      "grad_norm": 0.015363846905529499,
      "learning_rate": 2.846481876332623e-05,
      "loss": 0.007,
      "step": 1010
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 0.01426735706627369,
      "learning_rate": 2.835820895522388e-05,
      "loss": 0.0017,
      "step": 1015
    },
    {
      "epoch": 2.1748400852878467,
      "grad_norm": 0.04286054149270058,
      "learning_rate": 2.825159914712154e-05,
      "loss": 0.0018,
      "step": 1020
    },
    {
      "epoch": 2.185501066098081,
      "grad_norm": 0.013738875277340412,
      "learning_rate": 2.814498933901919e-05,
      "loss": 0.0026,
      "step": 1025
    },
    {
      "epoch": 2.1961620469083156,
      "grad_norm": 0.4904980957508087,
      "learning_rate": 2.8038379530916848e-05,
      "loss": 0.0051,
      "step": 1030
    },
    {
      "epoch": 2.20682302771855,
      "grad_norm": 0.013008911162614822,
      "learning_rate": 2.79317697228145e-05,
      "loss": 0.0016,
      "step": 1035
    },
    {
      "epoch": 2.2174840085287846,
      "grad_norm": 0.013201174326241016,
      "learning_rate": 2.7825159914712157e-05,
      "loss": 0.0016,
      "step": 1040
    },
    {
      "epoch": 2.228144989339019,
      "grad_norm": 0.03662689030170441,
      "learning_rate": 2.771855010660981e-05,
      "loss": 0.0019,
      "step": 1045
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 0.012772216461598873,
      "learning_rate": 2.7611940298507467e-05,
      "loss": 0.0015,
      "step": 1050
    },
    {
      "epoch": 2.2494669509594885,
      "grad_norm": 0.012919310480356216,
      "learning_rate": 2.7505330490405118e-05,
      "loss": 0.0019,
      "step": 1055
    },
    {
      "epoch": 2.260127931769723,
      "grad_norm": 0.01390413660556078,
      "learning_rate": 2.7398720682302776e-05,
      "loss": 0.0017,
      "step": 1060
    },
    {
      "epoch": 2.2707889125799574,
      "grad_norm": 0.013280141167342663,
      "learning_rate": 2.7292110874200428e-05,
      "loss": 0.0015,
      "step": 1065
    },
    {
      "epoch": 2.281449893390192,
      "grad_norm": 0.013000053353607655,
      "learning_rate": 2.7185501066098086e-05,
      "loss": 0.0015,
      "step": 1070
    },
    {
      "epoch": 2.2921108742004264,
      "grad_norm": 0.012426944449543953,
      "learning_rate": 2.7078891257995737e-05,
      "loss": 0.0016,
      "step": 1075
    },
    {
      "epoch": 2.302771855010661,
      "grad_norm": 0.012249843217432499,
      "learning_rate": 2.6972281449893395e-05,
      "loss": 0.0015,
      "step": 1080
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 0.012140439823269844,
      "learning_rate": 2.6865671641791047e-05,
      "loss": 0.0014,
      "step": 1085
    },
    {
      "epoch": 2.3240938166311302,
      "grad_norm": 0.01209353469312191,
      "learning_rate": 2.6759061833688705e-05,
      "loss": 0.0015,
      "step": 1090
    },
    {
      "epoch": 2.3347547974413647,
      "grad_norm": 0.01500478945672512,
      "learning_rate": 2.6652452025586356e-05,
      "loss": 0.0024,
      "step": 1095
    },
    {
      "epoch": 2.345415778251599,
      "grad_norm": 0.012158229947090149,
      "learning_rate": 2.6545842217484007e-05,
      "loss": 0.0029,
      "step": 1100
    },
    {
      "epoch": 2.3560767590618337,
      "grad_norm": 0.012071085162460804,
      "learning_rate": 2.6439232409381666e-05,
      "loss": 0.0014,
      "step": 1105
    },
    {
      "epoch": 2.366737739872068,
      "grad_norm": 0.012218043208122253,
      "learning_rate": 2.6332622601279317e-05,
      "loss": 0.0049,
      "step": 1110
    },
    {
      "epoch": 2.3773987206823026,
      "grad_norm": 0.01214613951742649,
      "learning_rate": 2.6226012793176975e-05,
      "loss": 0.0016,
      "step": 1115
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 0.012283292599022388,
      "learning_rate": 2.6119402985074626e-05,
      "loss": 0.0015,
      "step": 1120
    },
    {
      "epoch": 2.398720682302772,
      "grad_norm": 0.011755337938666344,
      "learning_rate": 2.6012793176972285e-05,
      "loss": 0.0014,
      "step": 1125
    },
    {
      "epoch": 2.4093816631130065,
      "grad_norm": 0.01584736444056034,
      "learning_rate": 2.5906183368869936e-05,
      "loss": 0.0014,
      "step": 1130
    },
    {
      "epoch": 2.420042643923241,
      "grad_norm": 0.01156493742018938,
      "learning_rate": 2.5799573560767594e-05,
      "loss": 0.0014,
      "step": 1135
    },
    {
      "epoch": 2.4307036247334755,
      "grad_norm": 0.02171207033097744,
      "learning_rate": 2.5692963752665245e-05,
      "loss": 0.0015,
      "step": 1140
    },
    {
      "epoch": 2.44136460554371,
      "grad_norm": 0.019722815603017807,
      "learning_rate": 2.5586353944562904e-05,
      "loss": 0.0014,
      "step": 1145
    },
    {
      "epoch": 2.4520255863539444,
      "grad_norm": 0.013081517070531845,
      "learning_rate": 2.5479744136460555e-05,
      "loss": 0.0014,
      "step": 1150
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 0.011529254727065563,
      "learning_rate": 2.537313432835821e-05,
      "loss": 0.0014,
      "step": 1155
    },
    {
      "epoch": 2.473347547974414,
      "grad_norm": 0.011184630915522575,
      "learning_rate": 2.5266524520255864e-05,
      "loss": 0.0058,
      "step": 1160
    },
    {
      "epoch": 2.4840085287846483,
      "grad_norm": 0.014834458939731121,
      "learning_rate": 2.515991471215352e-05,
      "loss": 0.0013,
      "step": 1165
    },
    {
      "epoch": 2.4946695095948828,
      "grad_norm": 0.011027764528989792,
      "learning_rate": 2.5053304904051174e-05,
      "loss": 0.0014,
      "step": 1170
    },
    {
      "epoch": 2.5053304904051172,
      "grad_norm": 0.030367983505129814,
      "learning_rate": 2.494669509594883e-05,
      "loss": 0.0015,
      "step": 1175
    },
    {
      "epoch": 2.5159914712153517,
      "grad_norm": 0.011093689128756523,
      "learning_rate": 2.4840085287846483e-05,
      "loss": 0.0013,
      "step": 1180
    },
    {
      "epoch": 2.526652452025586,
      "grad_norm": 0.010750957764685154,
      "learning_rate": 2.4733475479744138e-05,
      "loss": 0.0029,
      "step": 1185
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 0.010756141506135464,
      "learning_rate": 2.4626865671641793e-05,
      "loss": 0.0013,
      "step": 1190
    },
    {
      "epoch": 2.5479744136460556,
      "grad_norm": 0.011194289661943913,
      "learning_rate": 2.4520255863539444e-05,
      "loss": 0.0013,
      "step": 1195
    },
    {
      "epoch": 2.55863539445629,
      "grad_norm": 0.010614649392664433,
      "learning_rate": 2.44136460554371e-05,
      "loss": 0.0013,
      "step": 1200
    },
    {
      "epoch": 2.5692963752665245,
      "grad_norm": 0.01083033811300993,
      "learning_rate": 2.4307036247334754e-05,
      "loss": 0.0013,
      "step": 1205
    },
    {
      "epoch": 2.579957356076759,
      "grad_norm": 0.010807648301124573,
      "learning_rate": 2.420042643923241e-05,
      "loss": 0.0013,
      "step": 1210
    },
    {
      "epoch": 2.5906183368869935,
      "grad_norm": 0.022627221420407295,
      "learning_rate": 2.4093816631130063e-05,
      "loss": 0.0013,
      "step": 1215
    },
    {
      "epoch": 2.6012793176972284,
      "grad_norm": 0.010628443211317062,
      "learning_rate": 2.3987206823027718e-05,
      "loss": 0.0048,
      "step": 1220
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 0.010282629169523716,
      "learning_rate": 2.3880597014925373e-05,
      "loss": 0.0012,
      "step": 1225
    },
    {
      "epoch": 2.6226012793176974,
      "grad_norm": 0.01436874084174633,
      "learning_rate": 2.3773987206823027e-05,
      "loss": 0.0017,
      "step": 1230
    },
    {
      "epoch": 2.633262260127932,
      "grad_norm": 0.010025793686509132,
      "learning_rate": 2.3667377398720682e-05,
      "loss": 0.0012,
      "step": 1235
    },
    {
      "epoch": 2.6439232409381663,
      "grad_norm": 0.2057594358921051,
      "learning_rate": 2.3560767590618337e-05,
      "loss": 0.0067,
      "step": 1240
    },
    {
      "epoch": 2.654584221748401,
      "grad_norm": 0.010275648906826973,
      "learning_rate": 2.345415778251599e-05,
      "loss": 0.0012,
      "step": 1245
    },
    {
      "epoch": 2.6652452025586353,
      "grad_norm": 0.029614688828587532,
      "learning_rate": 2.3347547974413646e-05,
      "loss": 0.0015,
      "step": 1250
    },
    {
      "epoch": 2.6759061833688698,
      "grad_norm": 0.01017677504569292,
      "learning_rate": 2.32409381663113e-05,
      "loss": 0.0012,
      "step": 1255
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 0.010082006454467773,
      "learning_rate": 2.3134328358208956e-05,
      "loss": 0.0012,
      "step": 1260
    },
    {
      "epoch": 2.697228144989339,
      "grad_norm": 0.01118564885109663,
      "learning_rate": 2.302771855010661e-05,
      "loss": 0.0012,
      "step": 1265
    },
    {
      "epoch": 2.7078891257995736,
      "grad_norm": 0.009883342310786247,
      "learning_rate": 2.2921108742004265e-05,
      "loss": 0.0015,
      "step": 1270
    },
    {
      "epoch": 2.718550106609808,
      "grad_norm": 0.010351916775107384,
      "learning_rate": 2.281449893390192e-05,
      "loss": 0.0012,
      "step": 1275
    },
    {
      "epoch": 2.7292110874200426,
      "grad_norm": 0.011892825365066528,
      "learning_rate": 2.2707889125799575e-05,
      "loss": 0.0027,
      "step": 1280
    },
    {
      "epoch": 2.739872068230277,
      "grad_norm": 0.010221291333436966,
      "learning_rate": 2.260127931769723e-05,
      "loss": 0.0012,
      "step": 1285
    },
    {
      "epoch": 2.750533049040512,
      "grad_norm": 0.010303476825356483,
      "learning_rate": 2.2494669509594884e-05,
      "loss": 0.0012,
      "step": 1290
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 0.019318614155054092,
      "learning_rate": 2.238805970149254e-05,
      "loss": 0.0013,
      "step": 1295
    },
    {
      "epoch": 2.771855010660981,
      "grad_norm": 0.009472822770476341,
      "learning_rate": 2.2281449893390194e-05,
      "loss": 0.0011,
      "step": 1300
    },
    {
      "epoch": 2.7825159914712154,
      "grad_norm": 0.009815825149416924,
      "learning_rate": 2.217484008528785e-05,
      "loss": 0.0012,
      "step": 1305
    },
    {
      "epoch": 2.79317697228145,
      "grad_norm": 0.009688694961369038,
      "learning_rate": 2.2068230277185503e-05,
      "loss": 0.0011,
      "step": 1310
    },
    {
      "epoch": 2.8038379530916844,
      "grad_norm": 0.28238606452941895,
      "learning_rate": 2.1961620469083158e-05,
      "loss": 0.0435,
      "step": 1315
    },
    {
      "epoch": 2.814498933901919,
      "grad_norm": 0.01650206558406353,
      "learning_rate": 2.1855010660980813e-05,
      "loss": 0.0011,
      "step": 1320
    },
    {
      "epoch": 2.8251599147121533,
      "grad_norm": 0.009697536937892437,
      "learning_rate": 2.1748400852878467e-05,
      "loss": 0.0012,
      "step": 1325
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 0.010664798319339752,
      "learning_rate": 2.164179104477612e-05,
      "loss": 0.0011,
      "step": 1330
    },
    {
      "epoch": 2.8464818763326227,
      "grad_norm": 0.009492074139416218,
      "learning_rate": 2.1535181236673773e-05,
      "loss": 0.0012,
      "step": 1335
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.009845663793385029,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.0011,
      "step": 1340
    },
    {
      "epoch": 2.8678038379530917,
      "grad_norm": 0.009600410237908363,
      "learning_rate": 2.1321961620469083e-05,
      "loss": 0.0011,
      "step": 1345
    },
    {
      "epoch": 2.878464818763326,
      "grad_norm": 0.009601897560060024,
      "learning_rate": 2.1215351812366738e-05,
      "loss": 0.0011,
      "step": 1350
    },
    {
      "epoch": 2.8891257995735606,
      "grad_norm": 0.010103050619363785,
      "learning_rate": 2.1108742004264392e-05,
      "loss": 0.0029,
      "step": 1355
    },
    {
      "epoch": 2.8997867803837956,
      "grad_norm": 0.009285909123718739,
      "learning_rate": 2.1002132196162047e-05,
      "loss": 0.0131,
      "step": 1360
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 0.01248752698302269,
      "learning_rate": 2.0895522388059702e-05,
      "loss": 0.0011,
      "step": 1365
    },
    {
      "epoch": 2.9211087420042645,
      "grad_norm": 0.010123838670551777,
      "learning_rate": 2.0788912579957357e-05,
      "loss": 0.0011,
      "step": 1370
    },
    {
      "epoch": 2.931769722814499,
      "grad_norm": 0.009308276697993279,
      "learning_rate": 2.068230277185501e-05,
      "loss": 0.0011,
      "step": 1375
    },
    {
      "epoch": 2.9424307036247335,
      "grad_norm": 0.009351525455713272,
      "learning_rate": 2.0575692963752666e-05,
      "loss": 0.0013,
      "step": 1380
    },
    {
      "epoch": 2.953091684434968,
      "grad_norm": 0.016460668295621872,
      "learning_rate": 2.046908315565032e-05,
      "loss": 0.0012,
      "step": 1385
    },
    {
      "epoch": 2.9637526652452024,
      "grad_norm": 0.00919937714934349,
      "learning_rate": 2.0362473347547976e-05,
      "loss": 0.0011,
      "step": 1390
    },
    {
      "epoch": 2.974413646055437,
      "grad_norm": 0.009609651751816273,
      "learning_rate": 2.025586353944563e-05,
      "loss": 0.0011,
      "step": 1395
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 0.013855514116585255,
      "learning_rate": 2.0149253731343285e-05,
      "loss": 0.0011,
      "step": 1400
    },
    {
      "epoch": 2.9957356076759063,
      "grad_norm": 0.009133475832641125,
      "learning_rate": 2.004264392324094e-05,
      "loss": 0.0011,
      "step": 1405
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9946666666666667,
      "eval_loss": 0.022856945171952248,
      "eval_runtime": 17.0183,
      "eval_samples_per_second": 220.351,
      "eval_steps_per_second": 6.934,
      "step": 1407
    },
    {
      "epoch": 3.0063965884861408,
      "grad_norm": 0.009368853643536568,
      "learning_rate": 1.9936034115138594e-05,
      "loss": 0.0011,
      "step": 1410
    },
    {
      "epoch": 3.0170575692963753,
      "grad_norm": 0.00928573589771986,
      "learning_rate": 1.982942430703625e-05,
      "loss": 0.0011,
      "step": 1415
    },
    {
      "epoch": 3.0277185501066097,
      "grad_norm": 0.009126279503107071,
      "learning_rate": 1.9722814498933904e-05,
      "loss": 0.0012,
      "step": 1420
    },
    {
      "epoch": 3.038379530916844,
      "grad_norm": 0.008992459625005722,
      "learning_rate": 1.961620469083156e-05,
      "loss": 0.001,
      "step": 1425
    },
    {
      "epoch": 3.0490405117270787,
      "grad_norm": 0.009131360799074173,
      "learning_rate": 1.9509594882729213e-05,
      "loss": 0.0011,
      "step": 1430
    },
    {
      "epoch": 3.0597014925373136,
      "grad_norm": 0.00903151836246252,
      "learning_rate": 1.9402985074626868e-05,
      "loss": 0.0071,
      "step": 1435
    },
    {
      "epoch": 3.070362473347548,
      "grad_norm": 0.008744587190449238,
      "learning_rate": 1.9296375266524523e-05,
      "loss": 0.001,
      "step": 1440
    },
    {
      "epoch": 3.0810234541577826,
      "grad_norm": 0.00879298523068428,
      "learning_rate": 1.9189765458422178e-05,
      "loss": 0.001,
      "step": 1445
    },
    {
      "epoch": 3.091684434968017,
      "grad_norm": 0.008923621848225594,
      "learning_rate": 1.9083155650319832e-05,
      "loss": 0.001,
      "step": 1450
    },
    {
      "epoch": 3.1023454157782515,
      "grad_norm": 0.008955532684922218,
      "learning_rate": 1.8976545842217487e-05,
      "loss": 0.0016,
      "step": 1455
    },
    {
      "epoch": 3.113006396588486,
      "grad_norm": 0.008777831681072712,
      "learning_rate": 1.8869936034115142e-05,
      "loss": 0.001,
      "step": 1460
    },
    {
      "epoch": 3.1236673773987205,
      "grad_norm": 0.009090494364500046,
      "learning_rate": 1.8763326226012797e-05,
      "loss": 0.001,
      "step": 1465
    },
    {
      "epoch": 3.1343283582089554,
      "grad_norm": 0.00862858910113573,
      "learning_rate": 1.865671641791045e-05,
      "loss": 0.001,
      "step": 1470
    },
    {
      "epoch": 3.14498933901919,
      "grad_norm": 0.009000065736472607,
      "learning_rate": 1.8550106609808106e-05,
      "loss": 0.001,
      "step": 1475
    },
    {
      "epoch": 3.1556503198294243,
      "grad_norm": 0.008759795688092709,
      "learning_rate": 1.8443496801705757e-05,
      "loss": 0.0035,
      "step": 1480
    },
    {
      "epoch": 3.166311300639659,
      "grad_norm": 0.008655063807964325,
      "learning_rate": 1.8336886993603412e-05,
      "loss": 0.001,
      "step": 1485
    },
    {
      "epoch": 3.1769722814498933,
      "grad_norm": 0.008448773995041847,
      "learning_rate": 1.8230277185501067e-05,
      "loss": 0.001,
      "step": 1490
    },
    {
      "epoch": 3.1876332622601278,
      "grad_norm": 0.00870588980615139,
      "learning_rate": 1.812366737739872e-05,
      "loss": 0.0011,
      "step": 1495
    },
    {
      "epoch": 3.1982942430703627,
      "grad_norm": 0.008393441326916218,
      "learning_rate": 1.8017057569296376e-05,
      "loss": 0.001,
      "step": 1500
    },
    {
      "epoch": 3.208955223880597,
      "grad_norm": 0.009119479916989803,
      "learning_rate": 1.791044776119403e-05,
      "loss": 0.001,
      "step": 1505
    },
    {
      "epoch": 3.2196162046908317,
      "grad_norm": 0.008332905359566212,
      "learning_rate": 1.7803837953091686e-05,
      "loss": 0.001,
      "step": 1510
    },
    {
      "epoch": 3.230277185501066,
      "grad_norm": 0.011142686009407043,
      "learning_rate": 1.769722814498934e-05,
      "loss": 0.001,
      "step": 1515
    },
    {
      "epoch": 3.2409381663113006,
      "grad_norm": 0.008251124061644077,
      "learning_rate": 1.7590618336886995e-05,
      "loss": 0.001,
      "step": 1520
    },
    {
      "epoch": 3.251599147121535,
      "grad_norm": 0.008427221328020096,
      "learning_rate": 1.7484008528784647e-05,
      "loss": 0.001,
      "step": 1525
    },
    {
      "epoch": 3.2622601279317696,
      "grad_norm": 0.00855677668005228,
      "learning_rate": 1.73773987206823e-05,
      "loss": 0.001,
      "step": 1530
    },
    {
      "epoch": 3.272921108742004,
      "grad_norm": 0.008078392595052719,
      "learning_rate": 1.7270788912579956e-05,
      "loss": 0.001,
      "step": 1535
    },
    {
      "epoch": 3.283582089552239,
      "grad_norm": 0.008482052944600582,
      "learning_rate": 1.716417910447761e-05,
      "loss": 0.001,
      "step": 1540
    },
    {
      "epoch": 3.2942430703624734,
      "grad_norm": 0.009341388940811157,
      "learning_rate": 1.7057569296375266e-05,
      "loss": 0.001,
      "step": 1545
    },
    {
      "epoch": 3.304904051172708,
      "grad_norm": 0.008152157068252563,
      "learning_rate": 1.695095948827292e-05,
      "loss": 0.001,
      "step": 1550
    },
    {
      "epoch": 3.3155650319829424,
      "grad_norm": 0.008092079311609268,
      "learning_rate": 1.6844349680170575e-05,
      "loss": 0.001,
      "step": 1555
    },
    {
      "epoch": 3.326226012793177,
      "grad_norm": 0.008420913480222225,
      "learning_rate": 1.673773987206823e-05,
      "loss": 0.001,
      "step": 1560
    },
    {
      "epoch": 3.3368869936034113,
      "grad_norm": 0.008343126624822617,
      "learning_rate": 1.6631130063965885e-05,
      "loss": 0.001,
      "step": 1565
    },
    {
      "epoch": 3.3475479744136463,
      "grad_norm": 0.008144230581820011,
      "learning_rate": 1.652452025586354e-05,
      "loss": 0.001,
      "step": 1570
    },
    {
      "epoch": 3.3582089552238807,
      "grad_norm": 0.007995973341166973,
      "learning_rate": 1.6417910447761194e-05,
      "loss": 0.0009,
      "step": 1575
    },
    {
      "epoch": 3.368869936034115,
      "grad_norm": 0.007915866561233997,
      "learning_rate": 1.631130063965885e-05,
      "loss": 0.0009,
      "step": 1580
    },
    {
      "epoch": 3.3795309168443497,
      "grad_norm": 0.007987083867192268,
      "learning_rate": 1.6204690831556504e-05,
      "loss": 0.0009,
      "step": 1585
    },
    {
      "epoch": 3.390191897654584,
      "grad_norm": 0.008029541000723839,
      "learning_rate": 1.6098081023454158e-05,
      "loss": 0.0009,
      "step": 1590
    },
    {
      "epoch": 3.4008528784648187,
      "grad_norm": 0.008263756521046162,
      "learning_rate": 1.5991471215351813e-05,
      "loss": 0.0009,
      "step": 1595
    },
    {
      "epoch": 3.411513859275053,
      "grad_norm": 0.00779961422085762,
      "learning_rate": 1.5884861407249468e-05,
      "loss": 0.0009,
      "step": 1600
    },
    {
      "epoch": 3.4221748400852876,
      "grad_norm": 0.00786847248673439,
      "learning_rate": 1.5778251599147122e-05,
      "loss": 0.0009,
      "step": 1605
    },
    {
      "epoch": 3.4328358208955225,
      "grad_norm": 0.007777598220854998,
      "learning_rate": 1.5671641791044777e-05,
      "loss": 0.0009,
      "step": 1610
    },
    {
      "epoch": 3.443496801705757,
      "grad_norm": 0.007693154737353325,
      "learning_rate": 1.5565031982942432e-05,
      "loss": 0.001,
      "step": 1615
    },
    {
      "epoch": 3.4541577825159915,
      "grad_norm": 0.00797303393483162,
      "learning_rate": 1.5458422174840087e-05,
      "loss": 0.0009,
      "step": 1620
    },
    {
      "epoch": 3.464818763326226,
      "grad_norm": 0.007763400208204985,
      "learning_rate": 1.535181236673774e-05,
      "loss": 0.0009,
      "step": 1625
    },
    {
      "epoch": 3.4754797441364604,
      "grad_norm": 0.007800037506967783,
      "learning_rate": 1.5245202558635396e-05,
      "loss": 0.0009,
      "step": 1630
    },
    {
      "epoch": 3.486140724946695,
      "grad_norm": 0.007904596626758575,
      "learning_rate": 1.5138592750533051e-05,
      "loss": 0.0009,
      "step": 1635
    },
    {
      "epoch": 3.49680170575693,
      "grad_norm": 0.008192162029445171,
      "learning_rate": 1.5031982942430706e-05,
      "loss": 0.0013,
      "step": 1640
    },
    {
      "epoch": 3.5074626865671643,
      "grad_norm": 0.007729103323072195,
      "learning_rate": 1.4925373134328357e-05,
      "loss": 0.0009,
      "step": 1645
    },
    {
      "epoch": 3.518123667377399,
      "grad_norm": 0.2967156171798706,
      "learning_rate": 1.4818763326226012e-05,
      "loss": 0.044,
      "step": 1650
    },
    {
      "epoch": 3.5287846481876333,
      "grad_norm": 0.008105948567390442,
      "learning_rate": 1.4712153518123666e-05,
      "loss": 0.0009,
      "step": 1655
    },
    {
      "epoch": 3.5394456289978677,
      "grad_norm": 0.00775302154943347,
      "learning_rate": 1.4605543710021321e-05,
      "loss": 0.0009,
      "step": 1660
    },
    {
      "epoch": 3.550106609808102,
      "grad_norm": 0.0077361236326396465,
      "learning_rate": 1.4498933901918976e-05,
      "loss": 0.0009,
      "step": 1665
    },
    {
      "epoch": 3.5607675906183367,
      "grad_norm": 0.0088364128023386,
      "learning_rate": 1.439232409381663e-05,
      "loss": 0.0009,
      "step": 1670
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.009987774305045605,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.001,
      "step": 1675
    },
    {
      "epoch": 3.582089552238806,
      "grad_norm": 0.0083088343963027,
      "learning_rate": 1.417910447761194e-05,
      "loss": 0.0009,
      "step": 1680
    },
    {
      "epoch": 3.5927505330490406,
      "grad_norm": 0.008930937387049198,
      "learning_rate": 1.4072494669509595e-05,
      "loss": 0.0052,
      "step": 1685
    },
    {
      "epoch": 3.603411513859275,
      "grad_norm": 0.007975965738296509,
      "learning_rate": 1.396588486140725e-05,
      "loss": 0.0009,
      "step": 1690
    },
    {
      "epoch": 3.6140724946695095,
      "grad_norm": 0.007811375893652439,
      "learning_rate": 1.3859275053304904e-05,
      "loss": 0.001,
      "step": 1695
    },
    {
      "epoch": 3.624733475479744,
      "grad_norm": 0.007806049659848213,
      "learning_rate": 1.3752665245202559e-05,
      "loss": 0.0009,
      "step": 1700
    },
    {
      "epoch": 3.635394456289979,
      "grad_norm": 0.007902558892965317,
      "learning_rate": 1.3646055437100214e-05,
      "loss": 0.0009,
      "step": 1705
    },
    {
      "epoch": 3.6460554371002134,
      "grad_norm": 0.007543427404016256,
      "learning_rate": 1.3539445628997869e-05,
      "loss": 0.0009,
      "step": 1710
    },
    {
      "epoch": 3.656716417910448,
      "grad_norm": 0.013299783691763878,
      "learning_rate": 1.3432835820895523e-05,
      "loss": 0.001,
      "step": 1715
    },
    {
      "epoch": 3.6673773987206824,
      "grad_norm": 0.007508652750402689,
      "learning_rate": 1.3326226012793178e-05,
      "loss": 0.0009,
      "step": 1720
    },
    {
      "epoch": 3.678038379530917,
      "grad_norm": 0.00801705289632082,
      "learning_rate": 1.3219616204690833e-05,
      "loss": 0.0009,
      "step": 1725
    },
    {
      "epoch": 3.6886993603411513,
      "grad_norm": 0.007567389402538538,
      "learning_rate": 1.3113006396588488e-05,
      "loss": 0.0009,
      "step": 1730
    },
    {
      "epoch": 3.699360341151386,
      "grad_norm": 0.007756579201668501,
      "learning_rate": 1.3006396588486142e-05,
      "loss": 0.0082,
      "step": 1735
    },
    {
      "epoch": 3.7100213219616203,
      "grad_norm": 0.007432934828102589,
      "learning_rate": 1.2899786780383797e-05,
      "loss": 0.0009,
      "step": 1740
    },
    {
      "epoch": 3.7206823027718547,
      "grad_norm": 0.007582199759781361,
      "learning_rate": 1.2793176972281452e-05,
      "loss": 0.0009,
      "step": 1745
    },
    {
      "epoch": 3.7313432835820897,
      "grad_norm": 0.007463539484888315,
      "learning_rate": 1.2686567164179105e-05,
      "loss": 0.0008,
      "step": 1750
    },
    {
      "epoch": 3.742004264392324,
      "grad_norm": 0.00755223399028182,
      "learning_rate": 1.257995735607676e-05,
      "loss": 0.0009,
      "step": 1755
    },
    {
      "epoch": 3.7526652452025586,
      "grad_norm": 0.007484814617782831,
      "learning_rate": 1.2473347547974414e-05,
      "loss": 0.001,
      "step": 1760
    },
    {
      "epoch": 3.763326226012793,
      "grad_norm": 0.007422641385346651,
      "learning_rate": 1.2366737739872069e-05,
      "loss": 0.0009,
      "step": 1765
    },
    {
      "epoch": 3.7739872068230276,
      "grad_norm": 0.007435275241732597,
      "learning_rate": 1.2260127931769722e-05,
      "loss": 0.0009,
      "step": 1770
    },
    {
      "epoch": 3.7846481876332625,
      "grad_norm": 0.00717807887122035,
      "learning_rate": 1.2153518123667377e-05,
      "loss": 0.0009,
      "step": 1775
    },
    {
      "epoch": 3.795309168443497,
      "grad_norm": 0.007124255411326885,
      "learning_rate": 1.2046908315565032e-05,
      "loss": 0.0008,
      "step": 1780
    },
    {
      "epoch": 3.8059701492537314,
      "grad_norm": 0.007450129836797714,
      "learning_rate": 1.1940298507462686e-05,
      "loss": 0.0008,
      "step": 1785
    },
    {
      "epoch": 3.816631130063966,
      "grad_norm": 0.0072946385480463505,
      "learning_rate": 1.1833688699360341e-05,
      "loss": 0.0008,
      "step": 1790
    },
    {
      "epoch": 3.8272921108742004,
      "grad_norm": 0.00721264211460948,
      "learning_rate": 1.1727078891257996e-05,
      "loss": 0.0008,
      "step": 1795
    },
    {
      "epoch": 3.837953091684435,
      "grad_norm": 0.007415530737489462,
      "learning_rate": 1.162046908315565e-05,
      "loss": 0.0008,
      "step": 1800
    },
    {
      "epoch": 3.8486140724946694,
      "grad_norm": 0.007288595661520958,
      "learning_rate": 1.1513859275053305e-05,
      "loss": 0.0037,
      "step": 1805
    },
    {
      "epoch": 3.859275053304904,
      "grad_norm": 0.00771984551101923,
      "learning_rate": 1.140724946695096e-05,
      "loss": 0.0008,
      "step": 1810
    },
    {
      "epoch": 3.8699360341151388,
      "grad_norm": 0.018424108624458313,
      "learning_rate": 1.1300639658848615e-05,
      "loss": 0.0013,
      "step": 1815
    },
    {
      "epoch": 3.8805970149253732,
      "grad_norm": 0.03557995706796646,
      "learning_rate": 1.119402985074627e-05,
      "loss": 0.0012,
      "step": 1820
    },
    {
      "epoch": 3.8912579957356077,
      "grad_norm": 0.007265831809490919,
      "learning_rate": 1.1087420042643924e-05,
      "loss": 0.0008,
      "step": 1825
    },
    {
      "epoch": 3.901918976545842,
      "grad_norm": 0.00740558747202158,
      "learning_rate": 1.0980810234541579e-05,
      "loss": 0.0008,
      "step": 1830
    },
    {
      "epoch": 3.9125799573560767,
      "grad_norm": 0.007415784057229757,
      "learning_rate": 1.0874200426439234e-05,
      "loss": 0.0008,
      "step": 1835
    },
    {
      "epoch": 3.923240938166311,
      "grad_norm": 0.00691164331510663,
      "learning_rate": 1.0767590618336887e-05,
      "loss": 0.0013,
      "step": 1840
    },
    {
      "epoch": 3.933901918976546,
      "grad_norm": 0.007646833546459675,
      "learning_rate": 1.0660980810234541e-05,
      "loss": 0.0008,
      "step": 1845
    },
    {
      "epoch": 3.9445628997867805,
      "grad_norm": 0.006887191440910101,
      "learning_rate": 1.0554371002132196e-05,
      "loss": 0.0008,
      "step": 1850
    },
    {
      "epoch": 3.955223880597015,
      "grad_norm": 0.007158663123846054,
      "learning_rate": 1.0447761194029851e-05,
      "loss": 0.0008,
      "step": 1855
    },
    {
      "epoch": 3.9658848614072495,
      "grad_norm": 0.007033709902316332,
      "learning_rate": 1.0341151385927506e-05,
      "loss": 0.0008,
      "step": 1860
    },
    {
      "epoch": 3.976545842217484,
      "grad_norm": 0.00696564232930541,
      "learning_rate": 1.023454157782516e-05,
      "loss": 0.0008,
      "step": 1865
    },
    {
      "epoch": 3.9872068230277184,
      "grad_norm": 0.006933414377272129,
      "learning_rate": 1.0127931769722815e-05,
      "loss": 0.001,
      "step": 1870
    },
    {
      "epoch": 3.997867803837953,
      "grad_norm": 0.0068906634114682674,
      "learning_rate": 1.002132196162047e-05,
      "loss": 0.0008,
      "step": 1875
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9949333333333333,
      "eval_loss": 0.02470521070063114,
      "eval_runtime": 20.5328,
      "eval_samples_per_second": 182.634,
      "eval_steps_per_second": 5.747,
      "step": 1876
    },
    {
      "epoch": 4.008528784648187,
      "grad_norm": 0.007092602085322142,
      "learning_rate": 9.914712153518125e-06,
      "loss": 0.0008,
      "step": 1880
    },
    {
      "epoch": 4.019189765458422,
      "grad_norm": 0.007114503066986799,
      "learning_rate": 9.80810234541578e-06,
      "loss": 0.0009,
      "step": 1885
    },
    {
      "epoch": 4.029850746268656,
      "grad_norm": 0.007054983638226986,
      "learning_rate": 9.701492537313434e-06,
      "loss": 0.0008,
      "step": 1890
    },
    {
      "epoch": 4.040511727078891,
      "grad_norm": 0.028849003836512566,
      "learning_rate": 9.594882729211089e-06,
      "loss": 0.001,
      "step": 1895
    },
    {
      "epoch": 4.051172707889126,
      "grad_norm": 0.007059966679662466,
      "learning_rate": 9.488272921108744e-06,
      "loss": 0.0012,
      "step": 1900
    },
    {
      "epoch": 4.061833688699361,
      "grad_norm": 0.006836135871708393,
      "learning_rate": 9.381663113006398e-06,
      "loss": 0.0009,
      "step": 1905
    },
    {
      "epoch": 4.072494669509595,
      "grad_norm": 0.0068479483015835285,
      "learning_rate": 9.275053304904053e-06,
      "loss": 0.0008,
      "step": 1910
    },
    {
      "epoch": 4.08315565031983,
      "grad_norm": 0.006866186857223511,
      "learning_rate": 9.168443496801706e-06,
      "loss": 0.0008,
      "step": 1915
    },
    {
      "epoch": 4.093816631130064,
      "grad_norm": 0.006824200972914696,
      "learning_rate": 9.06183368869936e-06,
      "loss": 0.001,
      "step": 1920
    },
    {
      "epoch": 4.104477611940299,
      "grad_norm": 0.006811595056205988,
      "learning_rate": 8.955223880597016e-06,
      "loss": 0.0008,
      "step": 1925
    },
    {
      "epoch": 4.115138592750533,
      "grad_norm": 0.00677255354821682,
      "learning_rate": 8.84861407249467e-06,
      "loss": 0.0008,
      "step": 1930
    },
    {
      "epoch": 4.1257995735607675,
      "grad_norm": 0.007070044055581093,
      "learning_rate": 8.742004264392323e-06,
      "loss": 0.0008,
      "step": 1935
    },
    {
      "epoch": 4.136460554371002,
      "grad_norm": 0.0068147857673466206,
      "learning_rate": 8.635394456289978e-06,
      "loss": 0.0008,
      "step": 1940
    },
    {
      "epoch": 4.1471215351812365,
      "grad_norm": 0.0068505811505019665,
      "learning_rate": 8.528784648187633e-06,
      "loss": 0.0008,
      "step": 1945
    },
    {
      "epoch": 4.157782515991471,
      "grad_norm": 0.006723499391227961,
      "learning_rate": 8.422174840085288e-06,
      "loss": 0.0008,
      "step": 1950
    },
    {
      "epoch": 4.1684434968017055,
      "grad_norm": 0.008438466116786003,
      "learning_rate": 8.315565031982942e-06,
      "loss": 0.0041,
      "step": 1955
    },
    {
      "epoch": 4.17910447761194,
      "grad_norm": 0.006672563962638378,
      "learning_rate": 8.208955223880597e-06,
      "loss": 0.0008,
      "step": 1960
    },
    {
      "epoch": 4.189765458422174,
      "grad_norm": 0.006688879802823067,
      "learning_rate": 8.102345415778252e-06,
      "loss": 0.0008,
      "step": 1965
    },
    {
      "epoch": 4.20042643923241,
      "grad_norm": 0.006508278660476208,
      "learning_rate": 7.995735607675907e-06,
      "loss": 0.0008,
      "step": 1970
    },
    {
      "epoch": 4.211087420042644,
      "grad_norm": 0.006888619624078274,
      "learning_rate": 7.889125799573561e-06,
      "loss": 0.0008,
      "step": 1975
    },
    {
      "epoch": 4.221748400852879,
      "grad_norm": 0.006600632332265377,
      "learning_rate": 7.782515991471216e-06,
      "loss": 0.0008,
      "step": 1980
    },
    {
      "epoch": 4.232409381663113,
      "grad_norm": 0.16250911355018616,
      "learning_rate": 7.67590618336887e-06,
      "loss": 0.007,
      "step": 1985
    },
    {
      "epoch": 4.243070362473348,
      "grad_norm": 0.006614594720304012,
      "learning_rate": 7.5692963752665255e-06,
      "loss": 0.0008,
      "step": 1990
    },
    {
      "epoch": 4.253731343283582,
      "grad_norm": 0.006743111647665501,
      "learning_rate": 7.4626865671641785e-06,
      "loss": 0.0008,
      "step": 1995
    },
    {
      "epoch": 4.264392324093817,
      "grad_norm": 0.006710912100970745,
      "learning_rate": 7.356076759061833e-06,
      "loss": 0.0008,
      "step": 2000
    },
    {
      "epoch": 4.275053304904051,
      "grad_norm": 0.0064783464185893536,
      "learning_rate": 7.249466950959488e-06,
      "loss": 0.001,
      "step": 2005
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.006849293597042561,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.0008,
      "step": 2010
    },
    {
      "epoch": 4.29637526652452,
      "grad_norm": 0.006612967699766159,
      "learning_rate": 7.0362473347547975e-06,
      "loss": 0.0008,
      "step": 2015
    },
    {
      "epoch": 4.3070362473347545,
      "grad_norm": 0.006666961126029491,
      "learning_rate": 6.929637526652452e-06,
      "loss": 0.0008,
      "step": 2020
    },
    {
      "epoch": 4.317697228144989,
      "grad_norm": 0.09179400652647018,
      "learning_rate": 6.823027718550107e-06,
      "loss": 0.003,
      "step": 2025
    },
    {
      "epoch": 4.3283582089552235,
      "grad_norm": 0.006482471711933613,
      "learning_rate": 6.716417910447762e-06,
      "loss": 0.0008,
      "step": 2030
    },
    {
      "epoch": 4.339019189765459,
      "grad_norm": 0.006697284523397684,
      "learning_rate": 6.609808102345416e-06,
      "loss": 0.0008,
      "step": 2035
    },
    {
      "epoch": 4.349680170575693,
      "grad_norm": 0.007023471407592297,
      "learning_rate": 6.503198294243071e-06,
      "loss": 0.0008,
      "step": 2040
    },
    {
      "epoch": 4.360341151385928,
      "grad_norm": 0.006777188275009394,
      "learning_rate": 6.396588486140726e-06,
      "loss": 0.0008,
      "step": 2045
    },
    {
      "epoch": 4.371002132196162,
      "grad_norm": 0.006447089836001396,
      "learning_rate": 6.28997867803838e-06,
      "loss": 0.0008,
      "step": 2050
    },
    {
      "epoch": 4.381663113006397,
      "grad_norm": 0.006505937315523624,
      "learning_rate": 6.1833688699360345e-06,
      "loss": 0.0008,
      "step": 2055
    },
    {
      "epoch": 4.392324093816631,
      "grad_norm": 0.006575414910912514,
      "learning_rate": 6.076759061833688e-06,
      "loss": 0.0007,
      "step": 2060
    },
    {
      "epoch": 4.402985074626866,
      "grad_norm": 0.006808110978454351,
      "learning_rate": 5.970149253731343e-06,
      "loss": 0.0008,
      "step": 2065
    },
    {
      "epoch": 4.4136460554371,
      "grad_norm": 0.00644000293686986,
      "learning_rate": 5.863539445628998e-06,
      "loss": 0.0007,
      "step": 2070
    },
    {
      "epoch": 4.424307036247335,
      "grad_norm": 0.015466631390154362,
      "learning_rate": 5.756929637526653e-06,
      "loss": 0.0009,
      "step": 2075
    },
    {
      "epoch": 4.434968017057569,
      "grad_norm": 0.006661558523774147,
      "learning_rate": 5.650319829424307e-06,
      "loss": 0.0008,
      "step": 2080
    },
    {
      "epoch": 4.445628997867804,
      "grad_norm": 0.006613501347601414,
      "learning_rate": 5.543710021321962e-06,
      "loss": 0.0008,
      "step": 2085
    },
    {
      "epoch": 4.456289978678038,
      "grad_norm": 0.007009352091699839,
      "learning_rate": 5.437100213219617e-06,
      "loss": 0.0008,
      "step": 2090
    },
    {
      "epoch": 4.466950959488273,
      "grad_norm": 0.006467157509177923,
      "learning_rate": 5.330490405117271e-06,
      "loss": 0.0007,
      "step": 2095
    },
    {
      "epoch": 4.477611940298507,
      "grad_norm": 0.006350195035338402,
      "learning_rate": 5.2238805970149255e-06,
      "loss": 0.0007,
      "step": 2100
    },
    {
      "epoch": 4.4882729211087415,
      "grad_norm": 0.006541652604937553,
      "learning_rate": 5.11727078891258e-06,
      "loss": 0.0007,
      "step": 2105
    },
    {
      "epoch": 4.498933901918977,
      "grad_norm": 0.006451810710132122,
      "learning_rate": 5.010660980810235e-06,
      "loss": 0.0009,
      "step": 2110
    },
    {
      "epoch": 4.509594882729211,
      "grad_norm": 0.006328529678285122,
      "learning_rate": 4.90405117270789e-06,
      "loss": 0.0008,
      "step": 2115
    },
    {
      "epoch": 4.520255863539446,
      "grad_norm": 0.006482933647930622,
      "learning_rate": 4.797441364605544e-06,
      "loss": 0.0008,
      "step": 2120
    },
    {
      "epoch": 4.53091684434968,
      "grad_norm": 0.006646956317126751,
      "learning_rate": 4.690831556503199e-06,
      "loss": 0.0007,
      "step": 2125
    },
    {
      "epoch": 4.541577825159915,
      "grad_norm": 0.006530864164233208,
      "learning_rate": 4.584221748400853e-06,
      "loss": 0.0007,
      "step": 2130
    },
    {
      "epoch": 4.552238805970149,
      "grad_norm": 0.006505405064672232,
      "learning_rate": 4.477611940298508e-06,
      "loss": 0.0007,
      "step": 2135
    },
    {
      "epoch": 4.562899786780384,
      "grad_norm": 0.006481026764959097,
      "learning_rate": 4.371002132196162e-06,
      "loss": 0.0007,
      "step": 2140
    },
    {
      "epoch": 4.573560767590618,
      "grad_norm": 0.00649348646402359,
      "learning_rate": 4.264392324093816e-06,
      "loss": 0.0007,
      "step": 2145
    },
    {
      "epoch": 4.584221748400853,
      "grad_norm": 0.006477253511548042,
      "learning_rate": 4.157782515991471e-06,
      "loss": 0.0007,
      "step": 2150
    },
    {
      "epoch": 4.594882729211087,
      "grad_norm": 0.006447369698435068,
      "learning_rate": 4.051172707889126e-06,
      "loss": 0.0007,
      "step": 2155
    },
    {
      "epoch": 4.605543710021322,
      "grad_norm": 0.006516118999570608,
      "learning_rate": 3.944562899786781e-06,
      "loss": 0.0007,
      "step": 2160
    },
    {
      "epoch": 4.616204690831556,
      "grad_norm": 0.006422132253646851,
      "learning_rate": 3.837953091684435e-06,
      "loss": 0.0007,
      "step": 2165
    },
    {
      "epoch": 4.6268656716417915,
      "grad_norm": 0.006467184517532587,
      "learning_rate": 3.7313432835820893e-06,
      "loss": 0.0007,
      "step": 2170
    },
    {
      "epoch": 4.637526652452026,
      "grad_norm": 0.0063330139964818954,
      "learning_rate": 3.624733475479744e-06,
      "loss": 0.0008,
      "step": 2175
    },
    {
      "epoch": 4.6481876332622605,
      "grad_norm": 0.006526220589876175,
      "learning_rate": 3.5181236673773987e-06,
      "loss": 0.0007,
      "step": 2180
    },
    {
      "epoch": 4.658848614072495,
      "grad_norm": 0.006420309655368328,
      "learning_rate": 3.4115138592750535e-06,
      "loss": 0.0007,
      "step": 2185
    },
    {
      "epoch": 4.669509594882729,
      "grad_norm": 0.006409101188182831,
      "learning_rate": 3.304904051172708e-06,
      "loss": 0.0007,
      "step": 2190
    },
    {
      "epoch": 4.680170575692964,
      "grad_norm": 0.006357685197144747,
      "learning_rate": 3.198294243070363e-06,
      "loss": 0.0007,
      "step": 2195
    },
    {
      "epoch": 4.690831556503198,
      "grad_norm": 0.006533981300890446,
      "learning_rate": 3.0916844349680173e-06,
      "loss": 0.0007,
      "step": 2200
    },
    {
      "epoch": 4.701492537313433,
      "grad_norm": 0.006736787501722574,
      "learning_rate": 2.9850746268656716e-06,
      "loss": 0.0007,
      "step": 2205
    },
    {
      "epoch": 4.712153518123667,
      "grad_norm": 0.006252239923924208,
      "learning_rate": 2.8784648187633263e-06,
      "loss": 0.0007,
      "step": 2210
    },
    {
      "epoch": 4.722814498933902,
      "grad_norm": 0.19979675114154816,
      "learning_rate": 2.771855010660981e-06,
      "loss": 0.0075,
      "step": 2215
    },
    {
      "epoch": 4.733475479744136,
      "grad_norm": 0.006329044234007597,
      "learning_rate": 2.6652452025586354e-06,
      "loss": 0.0007,
      "step": 2220
    },
    {
      "epoch": 4.744136460554371,
      "grad_norm": 0.006287151016294956,
      "learning_rate": 2.55863539445629e-06,
      "loss": 0.0007,
      "step": 2225
    },
    {
      "epoch": 4.754797441364605,
      "grad_norm": 0.006514046806842089,
      "learning_rate": 2.452025586353945e-06,
      "loss": 0.0007,
      "step": 2230
    },
    {
      "epoch": 4.76545842217484,
      "grad_norm": 0.006423839833587408,
      "learning_rate": 2.3454157782515996e-06,
      "loss": 0.0009,
      "step": 2235
    },
    {
      "epoch": 4.776119402985074,
      "grad_norm": 0.006428393069654703,
      "learning_rate": 2.238805970149254e-06,
      "loss": 0.0007,
      "step": 2240
    },
    {
      "epoch": 4.786780383795309,
      "grad_norm": 0.00650773337110877,
      "learning_rate": 2.132196162046908e-06,
      "loss": 0.0007,
      "step": 2245
    },
    {
      "epoch": 4.797441364605544,
      "grad_norm": 0.0062748147174716,
      "learning_rate": 2.025586353944563e-06,
      "loss": 0.0007,
      "step": 2250
    },
    {
      "epoch": 4.8081023454157785,
      "grad_norm": 0.006249801255762577,
      "learning_rate": 1.9189765458422177e-06,
      "loss": 0.0007,
      "step": 2255
    },
    {
      "epoch": 4.818763326226013,
      "grad_norm": 0.006277838721871376,
      "learning_rate": 1.812366737739872e-06,
      "loss": 0.0007,
      "step": 2260
    },
    {
      "epoch": 4.8294243070362475,
      "grad_norm": 0.006403633393347263,
      "learning_rate": 1.7057569296375267e-06,
      "loss": 0.0007,
      "step": 2265
    },
    {
      "epoch": 4.840085287846482,
      "grad_norm": 0.0062882257625460625,
      "learning_rate": 1.5991471215351815e-06,
      "loss": 0.0007,
      "step": 2270
    },
    {
      "epoch": 4.850746268656716,
      "grad_norm": 0.006579338572919369,
      "learning_rate": 1.4925373134328358e-06,
      "loss": 0.0007,
      "step": 2275
    },
    {
      "epoch": 4.861407249466951,
      "grad_norm": 0.006250455509871244,
      "learning_rate": 1.3859275053304905e-06,
      "loss": 0.0442,
      "step": 2280
    },
    {
      "epoch": 4.872068230277185,
      "grad_norm": 0.006211218424141407,
      "learning_rate": 1.279317697228145e-06,
      "loss": 0.0007,
      "step": 2285
    },
    {
      "epoch": 4.88272921108742,
      "grad_norm": 0.006650151219218969,
      "learning_rate": 1.1727078891257998e-06,
      "loss": 0.0007,
      "step": 2290
    },
    {
      "epoch": 4.893390191897654,
      "grad_norm": 0.006827252451330423,
      "learning_rate": 1.066098081023454e-06,
      "loss": 0.0029,
      "step": 2295
    },
    {
      "epoch": 4.904051172707889,
      "grad_norm": 0.006380705162882805,
      "learning_rate": 9.594882729211088e-07,
      "loss": 0.0007,
      "step": 2300
    },
    {
      "epoch": 4.914712153518123,
      "grad_norm": 0.007317108102142811,
      "learning_rate": 8.528784648187634e-07,
      "loss": 0.0007,
      "step": 2305
    },
    {
      "epoch": 4.925373134328359,
      "grad_norm": 0.006272484548389912,
      "learning_rate": 7.462686567164179e-07,
      "loss": 0.0007,
      "step": 2310
    },
    {
      "epoch": 4.936034115138593,
      "grad_norm": 0.008044550195336342,
      "learning_rate": 6.396588486140725e-07,
      "loss": 0.0008,
      "step": 2315
    },
    {
      "epoch": 4.946695095948828,
      "grad_norm": 0.006260822061449289,
      "learning_rate": 5.33049040511727e-07,
      "loss": 0.0007,
      "step": 2320
    },
    {
      "epoch": 4.957356076759062,
      "grad_norm": 0.0062827724032104015,
      "learning_rate": 4.264392324093817e-07,
      "loss": 0.0007,
      "step": 2325
    },
    {
      "epoch": 4.968017057569297,
      "grad_norm": 0.006444897036999464,
      "learning_rate": 3.1982942430703626e-07,
      "loss": 0.0007,
      "step": 2330
    },
    {
      "epoch": 4.978678038379531,
      "grad_norm": 0.015175371430814266,
      "learning_rate": 2.1321961620469084e-07,
      "loss": 0.0008,
      "step": 2335
    },
    {
      "epoch": 4.9893390191897655,
      "grad_norm": 0.006180144846439362,
      "learning_rate": 1.0660980810234542e-07,
      "loss": 0.0007,
      "step": 2340
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.006668759509921074,
      "learning_rate": 0.0,
      "loss": 0.0007,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9949333333333333,
      "eval_loss": 0.02496175467967987,
      "eval_runtime": 16.5113,
      "eval_samples_per_second": 227.117,
      "eval_steps_per_second": 7.147,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "step": 2345,
      "total_flos": 5.8118992210944e+18,
      "train_loss": 0.014759776034136253,
      "train_runtime": 832.1114,
      "train_samples_per_second": 90.132,
      "train_steps_per_second": 2.818
    }
  ],
  "logging_steps": 5,
  "max_steps": 2345,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.8118992210944e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}