{
  "best_metric": 0.6760911345481873,
  "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles1_seed1_vit_lr\\checkpoint-2345",
  "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": 0.01,
      "grad_norm": 0.5828699469566345,
      "learning_rate": 4.989339019189766e-05,
      "loss": 0.8221,
      "step": 5
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.39033976197242737,
      "learning_rate": 4.978678038379531e-05,
      "loss": 0.8309,
      "step": 10
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5242815613746643,
      "learning_rate": 4.9680170575692967e-05,
      "loss": 0.7594,
      "step": 15
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8361539244651794,
      "learning_rate": 4.957356076759062e-05,
      "loss": 0.8066,
      "step": 20
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1711714118719101,
      "learning_rate": 4.9466950959488276e-05,
      "loss": 0.7071,
      "step": 25
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.19998708367347717,
      "learning_rate": 4.936034115138593e-05,
      "loss": 0.6991,
      "step": 30
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.30334341526031494,
      "learning_rate": 4.9253731343283586e-05,
      "loss": 0.7404,
      "step": 35
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.3605813682079315,
      "learning_rate": 4.914712153518124e-05,
      "loss": 0.7351,
      "step": 40
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.06580419093370438,
      "learning_rate": 4.904051172707889e-05,
      "loss": 0.6959,
      "step": 45
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.05234284698963165,
      "learning_rate": 4.893390191897655e-05,
      "loss": 0.6917,
      "step": 50
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10359219461679459,
      "learning_rate": 4.88272921108742e-05,
      "loss": 0.6982,
      "step": 55
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.10600591450929642,
      "learning_rate": 4.872068230277186e-05,
      "loss": 0.7052,
      "step": 60
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.14884132146835327,
      "learning_rate": 4.861407249466951e-05,
      "loss": 0.6971,
      "step": 65
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.16608068346977234,
      "learning_rate": 4.850746268656717e-05,
      "loss": 0.696,
      "step": 70
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1025952398777008,
      "learning_rate": 4.840085287846482e-05,
      "loss": 0.6952,
      "step": 75
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.032446928322315216,
      "learning_rate": 4.829424307036248e-05,
      "loss": 0.7,
      "step": 80
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.20473094284534454,
      "learning_rate": 4.8187633262260126e-05,
      "loss": 0.6982,
      "step": 85
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.24259920418262482,
      "learning_rate": 4.808102345415779e-05,
      "loss": 0.6941,
      "step": 90
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13239941000938416,
      "learning_rate": 4.7974413646055436e-05,
      "loss": 0.6893,
      "step": 95
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11972621828317642,
      "learning_rate": 4.78678038379531e-05,
      "loss": 0.689,
      "step": 100
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.014825105667114258,
      "learning_rate": 4.7761194029850745e-05,
      "loss": 0.6935,
      "step": 105
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2749171555042267,
      "learning_rate": 4.765458422174841e-05,
      "loss": 0.7003,
      "step": 110
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.05277346074581146,
      "learning_rate": 4.7547974413646055e-05,
      "loss": 0.692,
      "step": 115
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.06357314437627792,
      "learning_rate": 4.7441364605543716e-05,
      "loss": 0.696,
      "step": 120
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.18096879124641418,
      "learning_rate": 4.7334754797441364e-05,
      "loss": 0.7011,
      "step": 125
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1589221954345703,
      "learning_rate": 4.7228144989339026e-05,
      "loss": 0.6944,
      "step": 130
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1216045394539833,
      "learning_rate": 4.7121535181236674e-05,
      "loss": 0.6947,
      "step": 135
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.022791074588894844,
      "learning_rate": 4.7014925373134335e-05,
      "loss": 0.6942,
      "step": 140
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.07503971457481384,
      "learning_rate": 4.690831556503198e-05,
      "loss": 0.6937,
      "step": 145
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.22517453134059906,
      "learning_rate": 4.6801705756929645e-05,
      "loss": 0.6958,
      "step": 150
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.006116226315498352,
      "learning_rate": 4.669509594882729e-05,
      "loss": 0.693,
      "step": 155
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.04895751550793648,
      "learning_rate": 4.658848614072495e-05,
      "loss": 0.6929,
      "step": 160
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3163030743598938,
      "learning_rate": 4.64818763326226e-05,
      "loss": 0.694,
      "step": 165
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.04843531921505928,
      "learning_rate": 4.637526652452026e-05,
      "loss": 0.6926,
      "step": 170
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.05857136473059654,
      "learning_rate": 4.626865671641791e-05,
      "loss": 0.6943,
      "step": 175
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.05212704837322235,
      "learning_rate": 4.6162046908315566e-05,
      "loss": 0.694,
      "step": 180
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.057463593780994415,
      "learning_rate": 4.605543710021322e-05,
      "loss": 0.6937,
      "step": 185
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.11181072145700455,
      "learning_rate": 4.5948827292110876e-05,
      "loss": 0.6927,
      "step": 190
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.16910992562770844,
      "learning_rate": 4.584221748400853e-05,
      "loss": 0.694,
      "step": 195
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.16462725400924683,
      "learning_rate": 4.5735607675906185e-05,
      "loss": 0.6938,
      "step": 200
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.04628448933362961,
      "learning_rate": 4.562899786780384e-05,
      "loss": 0.6932,
      "step": 205
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2542133629322052,
      "learning_rate": 4.5522388059701495e-05,
      "loss": 0.6933,
      "step": 210
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.022086847573518753,
      "learning_rate": 4.541577825159915e-05,
      "loss": 0.6929,
      "step": 215
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1755961775779724,
      "learning_rate": 4.5309168443496804e-05,
      "loss": 0.695,
      "step": 220
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1452350616455078,
      "learning_rate": 4.520255863539446e-05,
      "loss": 0.6953,
      "step": 225
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2042374312877655,
      "learning_rate": 4.5095948827292114e-05,
      "loss": 0.6929,
      "step": 230
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1066068783402443,
      "learning_rate": 4.498933901918977e-05,
      "loss": 0.692,
      "step": 235
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15991215407848358,
      "learning_rate": 4.488272921108742e-05,
      "loss": 0.6944,
      "step": 240
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.20871728658676147,
      "learning_rate": 4.477611940298508e-05,
      "loss": 0.693,
      "step": 245
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2562817335128784,
      "learning_rate": 4.466950959488273e-05,
      "loss": 0.6933,
      "step": 250
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.04552245885133743,
      "learning_rate": 4.456289978678039e-05,
      "loss": 0.6929,
      "step": 255
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.012926601804792881,
      "learning_rate": 4.445628997867804e-05,
      "loss": 0.6921,
      "step": 260
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.06915324926376343,
      "learning_rate": 4.43496801705757e-05,
      "loss": 0.6936,
      "step": 265
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11693533509969711,
      "learning_rate": 4.424307036247335e-05,
      "loss": 0.6975,
      "step": 270
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.25377050042152405,
      "learning_rate": 4.4136460554371006e-05,
      "loss": 0.6928,
      "step": 275
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.22589242458343506,
      "learning_rate": 4.402985074626866e-05,
      "loss": 0.6936,
      "step": 280
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.17468225955963135,
      "learning_rate": 4.3923240938166316e-05,
      "loss": 0.6934,
      "step": 285
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.041436854749917984,
      "learning_rate": 4.381663113006397e-05,
      "loss": 0.6949,
      "step": 290
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16202151775360107,
      "learning_rate": 4.3710021321961625e-05,
      "loss": 0.6925,
      "step": 295
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.10329734534025192,
      "learning_rate": 4.360341151385928e-05,
      "loss": 0.6947,
      "step": 300
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.06689364463090897,
      "learning_rate": 4.3496801705756935e-05,
      "loss": 0.6923,
      "step": 305
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.03244592249393463,
      "learning_rate": 4.339019189765459e-05,
      "loss": 0.6943,
      "step": 310
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.08199141174554825,
      "learning_rate": 4.328358208955224e-05,
      "loss": 0.6926,
      "step": 315
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.07944226264953613,
      "learning_rate": 4.31769722814499e-05,
      "loss": 0.6915,
      "step": 320
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.08416552096605301,
      "learning_rate": 4.307036247334755e-05,
      "loss": 0.6913,
      "step": 325
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1907459795475006,
      "learning_rate": 4.29637526652452e-05,
      "loss": 0.6964,
      "step": 330
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.08609720319509506,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.6878,
      "step": 335
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.036432862281799316,
      "learning_rate": 4.275053304904051e-05,
      "loss": 0.6942,
      "step": 340
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.031249074265360832,
      "learning_rate": 4.2643923240938166e-05,
      "loss": 0.6932,
      "step": 345
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1355859935283661,
      "learning_rate": 4.253731343283582e-05,
      "loss": 0.6911,
      "step": 350
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1345302313566208,
      "learning_rate": 4.2430703624733475e-05,
      "loss": 0.6947,
      "step": 355
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.07437537610530853,
      "learning_rate": 4.232409381663113e-05,
      "loss": 0.695,
      "step": 360
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.01497673336416483,
      "learning_rate": 4.2217484008528785e-05,
      "loss": 0.694,
      "step": 365
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.14993418753147125,
      "learning_rate": 4.211087420042644e-05,
      "loss": 0.6929,
      "step": 370
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.017090369015932083,
      "learning_rate": 4.2004264392324094e-05,
      "loss": 0.6923,
      "step": 375
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.07324916124343872,
      "learning_rate": 4.189765458422175e-05,
      "loss": 0.6925,
      "step": 380
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.035330481827259064,
      "learning_rate": 4.1791044776119404e-05,
      "loss": 0.6965,
      "step": 385
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2187144011259079,
      "learning_rate": 4.168443496801706e-05,
      "loss": 0.6933,
      "step": 390
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.05486568436026573,
      "learning_rate": 4.157782515991471e-05,
      "loss": 0.6934,
      "step": 395
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.20385077595710754,
      "learning_rate": 4.147121535181237e-05,
      "loss": 0.6928,
      "step": 400
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.011761588044464588,
      "learning_rate": 4.136460554371002e-05,
      "loss": 0.6928,
      "step": 405
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.34647056460380554,
      "learning_rate": 4.125799573560768e-05,
      "loss": 0.6924,
      "step": 410
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.24146543443202972,
      "learning_rate": 4.115138592750533e-05,
      "loss": 0.6909,
      "step": 415
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.07559261471033096,
      "learning_rate": 4.104477611940299e-05,
      "loss": 0.6914,
      "step": 420
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.031942449510097504,
      "learning_rate": 4.093816631130064e-05,
      "loss": 0.6911,
      "step": 425
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.033947765827178955,
      "learning_rate": 4.0831556503198296e-05,
      "loss": 0.6923,
      "step": 430
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.08389558643102646,
      "learning_rate": 4.072494669509595e-05,
      "loss": 0.6969,
      "step": 435
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12675201892852783,
      "learning_rate": 4.0618336886993606e-05,
      "loss": 0.6897,
      "step": 440
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.02003253623843193,
      "learning_rate": 4.051172707889126e-05,
      "loss": 0.691,
      "step": 445
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18815450370311737,
      "learning_rate": 4.0405117270788915e-05,
      "loss": 0.693,
      "step": 450
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.02639641985297203,
      "learning_rate": 4.029850746268657e-05,
      "loss": 0.6902,
      "step": 455
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.021094592288136482,
      "learning_rate": 4.0191897654584225e-05,
      "loss": 0.6958,
      "step": 460
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12382938712835312,
      "learning_rate": 4.008528784648188e-05,
      "loss": 0.6952,
      "step": 465
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.49706666666666666,
      "eval_loss": 0.6928426027297974,
      "eval_runtime": 17.2988,
      "eval_samples_per_second": 216.778,
      "eval_steps_per_second": 6.821,
      "step": 469
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1660253256559372,
      "learning_rate": 3.997867803837953e-05,
      "loss": 0.6912,
      "step": 470
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2098858803510666,
      "learning_rate": 3.987206823027719e-05,
      "loss": 0.6937,
      "step": 475
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.2579794228076935,
      "learning_rate": 3.976545842217484e-05,
      "loss": 0.6929,
      "step": 480
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.15190201997756958,
      "learning_rate": 3.96588486140725e-05,
      "loss": 0.6913,
      "step": 485
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.11310646682977676,
      "learning_rate": 3.9552238805970146e-05,
      "loss": 0.6922,
      "step": 490
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.13565397262573242,
      "learning_rate": 3.944562899786781e-05,
      "loss": 0.6916,
      "step": 495
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.028749778866767883,
      "learning_rate": 3.9339019189765456e-05,
      "loss": 0.6919,
      "step": 500
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.03068220429122448,
      "learning_rate": 3.923240938166312e-05,
      "loss": 0.6904,
      "step": 505
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.38801059126853943,
      "learning_rate": 3.9125799573560765e-05,
      "loss": 0.699,
      "step": 510
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.17091982066631317,
      "learning_rate": 3.901918976545843e-05,
      "loss": 0.692,
      "step": 515
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.19640736281871796,
      "learning_rate": 3.8912579957356075e-05,
      "loss": 0.6906,
      "step": 520
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.11164979636669159,
      "learning_rate": 3.8805970149253736e-05,
      "loss": 0.6919,
      "step": 525
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.09658452868461609,
      "learning_rate": 3.8699360341151384e-05,
      "loss": 0.6907,
      "step": 530
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.2699227035045624,
      "learning_rate": 3.8592750533049046e-05,
      "loss": 0.6921,
      "step": 535
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.16162264347076416,
      "learning_rate": 3.8486140724946694e-05,
      "loss": 0.694,
      "step": 540
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.10174128413200378,
      "learning_rate": 3.8379530916844355e-05,
      "loss": 0.6909,
      "step": 545
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.04992598667740822,
      "learning_rate": 3.8272921108742e-05,
      "loss": 0.692,
      "step": 550
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.054032549262046814,
      "learning_rate": 3.8166311300639665e-05,
      "loss": 0.6913,
      "step": 555
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.11708575487136841,
      "learning_rate": 3.805970149253731e-05,
      "loss": 0.692,
      "step": 560
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.012999237515032291,
      "learning_rate": 3.7953091684434974e-05,
      "loss": 0.6931,
      "step": 565
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.3079542815685272,
      "learning_rate": 3.784648187633262e-05,
      "loss": 0.6913,
      "step": 570
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.006371928378939629,
      "learning_rate": 3.7739872068230284e-05,
      "loss": 0.6927,
      "step": 575
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.10761117190122604,
      "learning_rate": 3.763326226012793e-05,
      "loss": 0.6913,
      "step": 580
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.05754609405994415,
      "learning_rate": 3.752665245202559e-05,
      "loss": 0.6923,
      "step": 585
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.11597250401973724,
      "learning_rate": 3.742004264392324e-05,
      "loss": 0.6932,
      "step": 590
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.06282239407300949,
      "learning_rate": 3.73134328358209e-05,
      "loss": 0.6926,
      "step": 595
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.21621769666671753,
      "learning_rate": 3.720682302771855e-05,
      "loss": 0.6913,
      "step": 600
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.042546581476926804,
      "learning_rate": 3.710021321961621e-05,
      "loss": 0.6919,
      "step": 605
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.06986606121063232,
      "learning_rate": 3.699360341151386e-05,
      "loss": 0.6913,
      "step": 610
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.0846039429306984,
      "learning_rate": 3.6886993603411515e-05,
      "loss": 0.6899,
      "step": 615
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.08857711404561996,
      "learning_rate": 3.678038379530917e-05,
      "loss": 0.6932,
      "step": 620
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.17316599190235138,
      "learning_rate": 3.6673773987206824e-05,
      "loss": 0.6908,
      "step": 625
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.05025525391101837,
      "learning_rate": 3.656716417910448e-05,
      "loss": 0.6927,
      "step": 630
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.2874080240726471,
      "learning_rate": 3.6460554371002134e-05,
      "loss": 0.6909,
      "step": 635
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.19556842744350433,
      "learning_rate": 3.635394456289979e-05,
      "loss": 0.694,
      "step": 640
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.0742514356970787,
      "learning_rate": 3.624733475479744e-05,
      "loss": 0.6935,
      "step": 645
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.09203707426786423,
      "learning_rate": 3.61407249466951e-05,
      "loss": 0.6863,
      "step": 650
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.022830668836832047,
      "learning_rate": 3.603411513859275e-05,
      "loss": 0.6968,
      "step": 655
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.10433661937713623,
      "learning_rate": 3.592750533049041e-05,
      "loss": 0.6915,
      "step": 660
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.266999751329422,
      "learning_rate": 3.582089552238806e-05,
      "loss": 0.6929,
      "step": 665
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.21351861953735352,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.6935,
      "step": 670
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.21409828960895538,
      "learning_rate": 3.560767590618337e-05,
      "loss": 0.6916,
      "step": 675
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.10528093576431274,
      "learning_rate": 3.5501066098081026e-05,
      "loss": 0.6892,
      "step": 680
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.05427803099155426,
      "learning_rate": 3.539445628997868e-05,
      "loss": 0.6904,
      "step": 685
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.10534835606813431,
      "learning_rate": 3.5287846481876336e-05,
      "loss": 0.6916,
      "step": 690
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.21933074295520782,
      "learning_rate": 3.518123667377399e-05,
      "loss": 0.6911,
      "step": 695
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.09845433384180069,
      "learning_rate": 3.5074626865671645e-05,
      "loss": 0.6905,
      "step": 700
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.06464946269989014,
      "learning_rate": 3.496801705756929e-05,
      "loss": 0.6925,
      "step": 705
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.1835612803697586,
      "learning_rate": 3.4861407249466955e-05,
      "loss": 0.6882,
      "step": 710
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.043870821595191956,
      "learning_rate": 3.47547974413646e-05,
      "loss": 0.6909,
      "step": 715
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.19715048372745514,
      "learning_rate": 3.4648187633262264e-05,
      "loss": 0.6881,
      "step": 720
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.08701707422733307,
      "learning_rate": 3.454157782515991e-05,
      "loss": 0.689,
      "step": 725
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.026389161124825478,
      "learning_rate": 3.4434968017057574e-05,
      "loss": 0.6923,
      "step": 730
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.15285784006118774,
      "learning_rate": 3.432835820895522e-05,
      "loss": 0.69,
      "step": 735
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.050972145050764084,
      "learning_rate": 3.422174840085288e-05,
      "loss": 0.6866,
      "step": 740
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.10428329557180405,
      "learning_rate": 3.411513859275053e-05,
      "loss": 0.692,
      "step": 745
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.19904804229736328,
      "learning_rate": 3.400852878464819e-05,
      "loss": 0.6953,
      "step": 750
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.2752636671066284,
      "learning_rate": 3.390191897654584e-05,
      "loss": 0.6895,
      "step": 755
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.07439960539340973,
      "learning_rate": 3.37953091684435e-05,
      "loss": 0.6841,
      "step": 760
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.16306805610656738,
      "learning_rate": 3.368869936034115e-05,
      "loss": 0.687,
      "step": 765
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.10966607183218002,
      "learning_rate": 3.358208955223881e-05,
      "loss": 0.6883,
      "step": 770
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.05831507220864296,
      "learning_rate": 3.347547974413646e-05,
      "loss": 0.6852,
      "step": 775
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.31338346004486084,
      "learning_rate": 3.336886993603412e-05,
      "loss": 0.6906,
      "step": 780
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.10868186503648758,
      "learning_rate": 3.326226012793177e-05,
      "loss": 0.6948,
      "step": 785
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.12757889926433563,
      "learning_rate": 3.3155650319829424e-05,
      "loss": 0.6776,
      "step": 790
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.07820712774991989,
      "learning_rate": 3.304904051172708e-05,
      "loss": 0.6944,
      "step": 795
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.25035345554351807,
      "learning_rate": 3.294243070362473e-05,
      "loss": 0.6879,
      "step": 800
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.15776149928569794,
      "learning_rate": 3.283582089552239e-05,
      "loss": 0.6788,
      "step": 805
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.10203266143798828,
      "learning_rate": 3.272921108742004e-05,
      "loss": 0.6847,
      "step": 810
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.19664335250854492,
      "learning_rate": 3.26226012793177e-05,
      "loss": 0.6888,
      "step": 815
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.17209216952323914,
      "learning_rate": 3.251599147121535e-05,
      "loss": 0.6995,
      "step": 820
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.06759248673915863,
      "learning_rate": 3.240938166311301e-05,
      "loss": 0.6901,
      "step": 825
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.18248839676380157,
      "learning_rate": 3.230277185501066e-05,
      "loss": 0.6906,
      "step": 830
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.049247272312641144,
      "learning_rate": 3.2196162046908317e-05,
      "loss": 0.6822,
      "step": 835
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.0517529733479023,
      "learning_rate": 3.208955223880597e-05,
      "loss": 0.6778,
      "step": 840
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.43367257714271545,
      "learning_rate": 3.1982942430703626e-05,
      "loss": 0.7034,
      "step": 845
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.37519243359565735,
      "learning_rate": 3.187633262260128e-05,
      "loss": 0.6829,
      "step": 850
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.07908283174037933,
      "learning_rate": 3.1769722814498935e-05,
      "loss": 0.6873,
      "step": 855
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.22837957739830017,
      "learning_rate": 3.166311300639659e-05,
      "loss": 0.6907,
      "step": 860
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.2421358823776245,
      "learning_rate": 3.1556503198294245e-05,
      "loss": 0.6893,
      "step": 865
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.26317983865737915,
      "learning_rate": 3.14498933901919e-05,
      "loss": 0.687,
      "step": 870
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.3738718032836914,
      "learning_rate": 3.1343283582089554e-05,
      "loss": 0.6857,
      "step": 875
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.14564202725887299,
      "learning_rate": 3.123667377398721e-05,
      "loss": 0.6821,
      "step": 880
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.13704021275043488,
      "learning_rate": 3.1130063965884864e-05,
      "loss": 0.6951,
      "step": 885
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.10348877310752869,
      "learning_rate": 3.102345415778252e-05,
      "loss": 0.6793,
      "step": 890
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.10818944126367569,
      "learning_rate": 3.0916844349680173e-05,
      "loss": 0.6884,
      "step": 895
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.17710749804973602,
      "learning_rate": 3.081023454157783e-05,
      "loss": 0.6861,
      "step": 900
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.11066386103630066,
      "learning_rate": 3.070362473347548e-05,
      "loss": 0.6761,
      "step": 905
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.14435610175132751,
      "learning_rate": 3.059701492537314e-05,
      "loss": 0.6747,
      "step": 910
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.15541020035743713,
      "learning_rate": 3.0490405117270792e-05,
      "loss": 0.6806,
      "step": 915
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.08867447078227997,
      "learning_rate": 3.0383795309168444e-05,
      "loss": 0.69,
      "step": 920
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.32354313135147095,
      "learning_rate": 3.0277185501066102e-05,
      "loss": 0.6812,
      "step": 925
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.32110169529914856,
      "learning_rate": 3.0170575692963753e-05,
      "loss": 0.6733,
      "step": 930
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.0616983063519001,
      "learning_rate": 3.006396588486141e-05,
      "loss": 0.6736,
      "step": 935
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5552,
      "eval_loss": 0.6843511462211609,
      "eval_runtime": 18.2604,
      "eval_samples_per_second": 205.363,
      "eval_steps_per_second": 6.462,
      "step": 938
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.13492348790168762,
      "learning_rate": 2.9957356076759063e-05,
      "loss": 0.6925,
      "step": 940
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.2928030490875244,
      "learning_rate": 2.9850746268656714e-05,
      "loss": 0.6677,
      "step": 945
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.11364587396383286,
      "learning_rate": 2.9744136460554372e-05,
      "loss": 0.6838,
      "step": 950
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.16877992451190948,
      "learning_rate": 2.9637526652452023e-05,
      "loss": 0.6952,
      "step": 955
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.38469552993774414,
      "learning_rate": 2.953091684434968e-05,
      "loss": 0.6981,
      "step": 960
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.2534121572971344,
      "learning_rate": 2.9424307036247333e-05,
      "loss": 0.6846,
      "step": 965
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.23564128577709198,
      "learning_rate": 2.931769722814499e-05,
      "loss": 0.6697,
      "step": 970
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.1798560619354248,
      "learning_rate": 2.9211087420042642e-05,
      "loss": 0.681,
      "step": 975
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.2816617786884308,
      "learning_rate": 2.91044776119403e-05,
      "loss": 0.6721,
      "step": 980
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.15949612855911255,
      "learning_rate": 2.8997867803837952e-05,
      "loss": 0.6684,
      "step": 985
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.3336601257324219,
      "learning_rate": 2.889125799573561e-05,
      "loss": 0.6789,
      "step": 990
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.15128879249095917,
      "learning_rate": 2.878464818763326e-05,
      "loss": 0.6679,
      "step": 995
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.2831324636936188,
      "learning_rate": 2.867803837953092e-05,
      "loss": 0.683,
      "step": 1000
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.09263738244771957,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.6769,
      "step": 1005
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.10221715271472931,
      "learning_rate": 2.846481876332623e-05,
      "loss": 0.6633,
      "step": 1010
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.3450409471988678,
      "learning_rate": 2.835820895522388e-05,
      "loss": 0.6859,
      "step": 1015
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.15574760735034943,
      "learning_rate": 2.825159914712154e-05,
      "loss": 0.6703,
      "step": 1020
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.18141822516918182,
      "learning_rate": 2.814498933901919e-05,
      "loss": 0.6829,
      "step": 1025
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.2733962833881378,
      "learning_rate": 2.8038379530916848e-05,
      "loss": 0.6668,
      "step": 1030
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.406516969203949,
      "learning_rate": 2.79317697228145e-05,
      "loss": 0.6932,
      "step": 1035
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.26090744137763977,
      "learning_rate": 2.7825159914712157e-05,
      "loss": 0.6624,
      "step": 1040
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.30359649658203125,
      "learning_rate": 2.771855010660981e-05,
      "loss": 0.678,
      "step": 1045
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.6121388077735901,
      "learning_rate": 2.7611940298507467e-05,
      "loss": 0.6777,
      "step": 1050
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.25791430473327637,
      "learning_rate": 2.7505330490405118e-05,
      "loss": 0.6562,
      "step": 1055
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.28375235199928284,
      "learning_rate": 2.7398720682302776e-05,
      "loss": 0.6836,
      "step": 1060
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.34015530347824097,
      "learning_rate": 2.7292110874200428e-05,
      "loss": 0.6845,
      "step": 1065
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.4710615277290344,
      "learning_rate": 2.7185501066098086e-05,
      "loss": 0.6593,
      "step": 1070
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.2846826910972595,
      "learning_rate": 2.7078891257995737e-05,
      "loss": 0.664,
      "step": 1075
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.599051296710968,
      "learning_rate": 2.6972281449893395e-05,
      "loss": 0.6747,
      "step": 1080
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.26196974515914917,
      "learning_rate": 2.6865671641791047e-05,
      "loss": 0.6693,
      "step": 1085
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.6016497611999512,
      "learning_rate": 2.6759061833688705e-05,
      "loss": 0.6736,
      "step": 1090
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.21997526288032532,
      "learning_rate": 2.6652452025586356e-05,
      "loss": 0.7005,
      "step": 1095
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.10501424223184586,
      "learning_rate": 2.6545842217484007e-05,
      "loss": 0.6859,
      "step": 1100
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.4104955196380615,
      "learning_rate": 2.6439232409381666e-05,
      "loss": 0.6939,
      "step": 1105
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.16587625443935394,
      "learning_rate": 2.6332622601279317e-05,
      "loss": 0.6703,
      "step": 1110
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.15244054794311523,
      "learning_rate": 2.6226012793176975e-05,
      "loss": 0.6841,
      "step": 1115
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.14376255869865417,
      "learning_rate": 2.6119402985074626e-05,
      "loss": 0.6817,
      "step": 1120
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.28233468532562256,
      "learning_rate": 2.6012793176972285e-05,
      "loss": 0.6702,
      "step": 1125
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.21459686756134033,
      "learning_rate": 2.5906183368869936e-05,
      "loss": 0.6552,
      "step": 1130
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.29934316873550415,
      "learning_rate": 2.5799573560767594e-05,
      "loss": 0.6684,
      "step": 1135
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.1567791849374771,
      "learning_rate": 2.5692963752665245e-05,
      "loss": 0.6877,
      "step": 1140
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.7527381777763367,
      "learning_rate": 2.5586353944562904e-05,
      "loss": 0.6815,
      "step": 1145
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.2388136237859726,
      "learning_rate": 2.5479744136460555e-05,
      "loss": 0.7029,
      "step": 1150
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.7899766564369202,
      "learning_rate": 2.537313432835821e-05,
      "loss": 0.6938,
      "step": 1155
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.5360927581787109,
      "learning_rate": 2.5266524520255864e-05,
      "loss": 0.6813,
      "step": 1160
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.2651243805885315,
      "learning_rate": 2.515991471215352e-05,
      "loss": 0.6877,
      "step": 1165
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.40308693051338196,
      "learning_rate": 2.5053304904051174e-05,
      "loss": 0.6936,
      "step": 1170
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.25783565640449524,
      "learning_rate": 2.494669509594883e-05,
      "loss": 0.6626,
      "step": 1175
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.2718556523323059,
      "learning_rate": 2.4840085287846483e-05,
      "loss": 0.6858,
      "step": 1180
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.139323428273201,
      "learning_rate": 2.4733475479744138e-05,
      "loss": 0.6612,
      "step": 1185
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.256630539894104,
      "learning_rate": 2.4626865671641793e-05,
      "loss": 0.6875,
      "step": 1190
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.2941284477710724,
      "learning_rate": 2.4520255863539444e-05,
      "loss": 0.6905,
      "step": 1195
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.2886035144329071,
      "learning_rate": 2.44136460554371e-05,
      "loss": 0.6565,
      "step": 1200
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.18019695580005646,
      "learning_rate": 2.4307036247334754e-05,
      "loss": 0.6781,
      "step": 1205
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.4069792628288269,
      "learning_rate": 2.420042643923241e-05,
      "loss": 0.6899,
      "step": 1210
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.4986726641654968,
      "learning_rate": 2.4093816631130063e-05,
      "loss": 0.6843,
      "step": 1215
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.34947189688682556,
      "learning_rate": 2.3987206823027718e-05,
      "loss": 0.6828,
      "step": 1220
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.12795394659042358,
      "learning_rate": 2.3880597014925373e-05,
      "loss": 0.6805,
      "step": 1225
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.3556659519672394,
      "learning_rate": 2.3773987206823027e-05,
      "loss": 0.6659,
      "step": 1230
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.2913447618484497,
      "learning_rate": 2.3667377398720682e-05,
      "loss": 0.661,
      "step": 1235
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.33494627475738525,
      "learning_rate": 2.3560767590618337e-05,
      "loss": 0.6621,
      "step": 1240
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.19486567378044128,
      "learning_rate": 2.345415778251599e-05,
      "loss": 0.6631,
      "step": 1245
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.16531342267990112,
      "learning_rate": 2.3347547974413646e-05,
      "loss": 0.689,
      "step": 1250
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.4831390976905823,
      "learning_rate": 2.32409381663113e-05,
      "loss": 0.6789,
      "step": 1255
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.15912608802318573,
      "learning_rate": 2.3134328358208956e-05,
      "loss": 0.6605,
      "step": 1260
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.21471765637397766,
      "learning_rate": 2.302771855010661e-05,
      "loss": 0.6865,
      "step": 1265
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.333564817905426,
      "learning_rate": 2.2921108742004265e-05,
      "loss": 0.6757,
      "step": 1270
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.23908694088459015,
      "learning_rate": 2.281449893390192e-05,
      "loss": 0.7025,
      "step": 1275
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.22186516225337982,
      "learning_rate": 2.2707889125799575e-05,
      "loss": 0.6907,
      "step": 1280
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.7879774570465088,
      "learning_rate": 2.260127931769723e-05,
      "loss": 0.685,
      "step": 1285
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.562380850315094,
      "learning_rate": 2.2494669509594884e-05,
      "loss": 0.6728,
      "step": 1290
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.22762958705425262,
      "learning_rate": 2.238805970149254e-05,
      "loss": 0.6782,
      "step": 1295
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.26743292808532715,
      "learning_rate": 2.2281449893390194e-05,
      "loss": 0.663,
      "step": 1300
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.21805542707443237,
      "learning_rate": 2.217484008528785e-05,
      "loss": 0.6609,
      "step": 1305
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.2946370542049408,
      "learning_rate": 2.2068230277185503e-05,
      "loss": 0.6802,
      "step": 1310
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.23653313517570496,
      "learning_rate": 2.1961620469083158e-05,
      "loss": 0.6943,
      "step": 1315
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.27479249238967896,
      "learning_rate": 2.1855010660980813e-05,
      "loss": 0.6599,
      "step": 1320
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.3639926314353943,
      "learning_rate": 2.1748400852878467e-05,
      "loss": 0.6773,
      "step": 1325
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.17922058701515198,
      "learning_rate": 2.164179104477612e-05,
      "loss": 0.6758,
      "step": 1330
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.39532744884490967,
      "learning_rate": 2.1535181236673773e-05,
      "loss": 0.6746,
      "step": 1335
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.4862898290157318,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.675,
      "step": 1340
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.46271321177482605,
      "learning_rate": 2.1321961620469083e-05,
      "loss": 0.6846,
      "step": 1345
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.15888762474060059,
      "learning_rate": 2.1215351812366738e-05,
      "loss": 0.6843,
      "step": 1350
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.45253995060920715,
      "learning_rate": 2.1108742004264392e-05,
      "loss": 0.658,
      "step": 1355
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.1543508768081665,
      "learning_rate": 2.1002132196162047e-05,
      "loss": 0.6722,
      "step": 1360
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.2126993089914322,
      "learning_rate": 2.0895522388059702e-05,
      "loss": 0.6526,
      "step": 1365
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.25720474123954773,
      "learning_rate": 2.0788912579957357e-05,
      "loss": 0.6596,
      "step": 1370
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.3520115613937378,
      "learning_rate": 2.068230277185501e-05,
      "loss": 0.6623,
      "step": 1375
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.3493128716945648,
      "learning_rate": 2.0575692963752666e-05,
      "loss": 0.666,
      "step": 1380
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.44602665305137634,
      "learning_rate": 2.046908315565032e-05,
      "loss": 0.6703,
      "step": 1385
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.5327634215354919,
      "learning_rate": 2.0362473347547976e-05,
      "loss": 0.6953,
      "step": 1390
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.7421027421951294,
      "learning_rate": 2.025586353944563e-05,
      "loss": 0.6827,
      "step": 1395
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.516767144203186,
      "learning_rate": 2.0149253731343285e-05,
      "loss": 0.6975,
      "step": 1400
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.417506605386734,
      "learning_rate": 2.004264392324094e-05,
      "loss": 0.66,
      "step": 1405
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5522666666666667,
      "eval_loss": 0.6899899840354919,
      "eval_runtime": 17.8871,
      "eval_samples_per_second": 209.648,
      "eval_steps_per_second": 6.597,
      "step": 1407
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.26411086320877075,
      "learning_rate": 1.9936034115138594e-05,
      "loss": 0.6743,
      "step": 1410
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.2531265318393707,
      "learning_rate": 1.982942430703625e-05,
      "loss": 0.6764,
      "step": 1415
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.3299374580383301,
      "learning_rate": 1.9722814498933904e-05,
      "loss": 0.6946,
      "step": 1420
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.1010500192642212,
      "learning_rate": 1.961620469083156e-05,
      "loss": 0.6612,
      "step": 1425
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.5194435715675354,
      "learning_rate": 1.9509594882729213e-05,
      "loss": 0.6768,
      "step": 1430
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.32416439056396484,
      "learning_rate": 1.9402985074626868e-05,
      "loss": 0.691,
      "step": 1435
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.1599058210849762,
      "learning_rate": 1.9296375266524523e-05,
      "loss": 0.6603,
      "step": 1440
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.2524496018886566,
      "learning_rate": 1.9189765458422178e-05,
      "loss": 0.688,
      "step": 1445
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.12636564671993256,
      "learning_rate": 1.9083155650319832e-05,
      "loss": 0.679,
      "step": 1450
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.26544949412345886,
      "learning_rate": 1.8976545842217487e-05,
      "loss": 0.6691,
      "step": 1455
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.22614602744579315,
      "learning_rate": 1.8869936034115142e-05,
      "loss": 0.6858,
      "step": 1460
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.39157307147979736,
      "learning_rate": 1.8763326226012797e-05,
      "loss": 0.6774,
      "step": 1465
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.7258807420730591,
      "learning_rate": 1.865671641791045e-05,
      "loss": 0.6577,
      "step": 1470
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.3102707862854004,
      "learning_rate": 1.8550106609808106e-05,
      "loss": 0.6719,
      "step": 1475
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.6749600768089294,
      "learning_rate": 1.8443496801705757e-05,
      "loss": 0.6728,
      "step": 1480
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.3821292519569397,
      "learning_rate": 1.8336886993603412e-05,
      "loss": 0.6486,
      "step": 1485
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.3780907094478607,
      "learning_rate": 1.8230277185501067e-05,
      "loss": 0.6672,
      "step": 1490
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.23514017462730408,
      "learning_rate": 1.812366737739872e-05,
      "loss": 0.6668,
      "step": 1495
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.18084169924259186,
      "learning_rate": 1.8017057569296376e-05,
      "loss": 0.6618,
      "step": 1500
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.5182689428329468,
      "learning_rate": 1.791044776119403e-05,
      "loss": 0.6417,
      "step": 1505
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.801882266998291,
      "learning_rate": 1.7803837953091686e-05,
      "loss": 0.6645,
      "step": 1510
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.18955937027931213,
      "learning_rate": 1.769722814498934e-05,
      "loss": 0.6887,
      "step": 1515
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.33656734228134155,
      "learning_rate": 1.7590618336886995e-05,
      "loss": 0.6597,
      "step": 1520
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.19224336743354797,
      "learning_rate": 1.7484008528784647e-05,
      "loss": 0.6939,
      "step": 1525
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.30128347873687744,
      "learning_rate": 1.73773987206823e-05,
      "loss": 0.6853,
      "step": 1530
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.4165492355823517,
      "learning_rate": 1.7270788912579956e-05,
      "loss": 0.6606,
      "step": 1535
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.38350632786750793,
      "learning_rate": 1.716417910447761e-05,
      "loss": 0.6526,
      "step": 1540
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.5553402900695801,
      "learning_rate": 1.7057569296375266e-05,
      "loss": 0.6917,
      "step": 1545
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.38175731897354126,
      "learning_rate": 1.695095948827292e-05,
      "loss": 0.637,
      "step": 1550
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.6655339598655701,
      "learning_rate": 1.6844349680170575e-05,
      "loss": 0.6645,
      "step": 1555
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.31345900893211365,
      "learning_rate": 1.673773987206823e-05,
      "loss": 0.6746,
      "step": 1560
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.5527360439300537,
      "learning_rate": 1.6631130063965885e-05,
      "loss": 0.6596,
      "step": 1565
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.2297668755054474,
      "learning_rate": 1.652452025586354e-05,
      "loss": 0.6798,
      "step": 1570
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.4209613502025604,
      "learning_rate": 1.6417910447761194e-05,
      "loss": 0.6866,
      "step": 1575
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.1880345642566681,
      "learning_rate": 1.631130063965885e-05,
      "loss": 0.6687,
      "step": 1580
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.2493944764137268,
      "learning_rate": 1.6204690831556504e-05,
      "loss": 0.673,
      "step": 1585
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.1546977460384369,
      "learning_rate": 1.6098081023454158e-05,
      "loss": 0.6654,
      "step": 1590
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.4542850852012634,
      "learning_rate": 1.5991471215351813e-05,
      "loss": 0.6897,
      "step": 1595
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.30015575885772705,
      "learning_rate": 1.5884861407249468e-05,
      "loss": 0.6784,
      "step": 1600
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.19910399615764618,
      "learning_rate": 1.5778251599147122e-05,
      "loss": 0.6695,
      "step": 1605
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.0947527885437012,
      "learning_rate": 1.5671641791044777e-05,
      "loss": 0.6715,
      "step": 1610
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.29615241289138794,
      "learning_rate": 1.5565031982942432e-05,
      "loss": 0.6926,
      "step": 1615
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.7357563376426697,
      "learning_rate": 1.5458422174840087e-05,
      "loss": 0.676,
      "step": 1620
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.3963560461997986,
      "learning_rate": 1.535181236673774e-05,
      "loss": 0.6943,
      "step": 1625
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.7890651226043701,
      "learning_rate": 1.5245202558635396e-05,
      "loss": 0.6674,
      "step": 1630
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.34833666682243347,
      "learning_rate": 1.5138592750533051e-05,
      "loss": 0.6803,
      "step": 1635
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.3659195899963379,
      "learning_rate": 1.5031982942430706e-05,
      "loss": 0.6559,
      "step": 1640
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.4654899835586548,
      "learning_rate": 1.4925373134328357e-05,
      "loss": 0.6859,
      "step": 1645
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.29908713698387146,
      "learning_rate": 1.4818763326226012e-05,
      "loss": 0.6877,
      "step": 1650
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.20764610171318054,
      "learning_rate": 1.4712153518123666e-05,
      "loss": 0.6714,
      "step": 1655
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.3057771623134613,
      "learning_rate": 1.4605543710021321e-05,
      "loss": 0.68,
      "step": 1660
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.4783516824245453,
      "learning_rate": 1.4498933901918976e-05,
      "loss": 0.6582,
      "step": 1665
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.44634905457496643,
      "learning_rate": 1.439232409381663e-05,
      "loss": 0.6602,
      "step": 1670
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.23565933108329773,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.6701,
      "step": 1675
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.3964284658432007,
      "learning_rate": 1.417910447761194e-05,
      "loss": 0.6803,
      "step": 1680
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.1792701780796051,
      "learning_rate": 1.4072494669509595e-05,
      "loss": 0.69,
      "step": 1685
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.2967628538608551,
      "learning_rate": 1.396588486140725e-05,
      "loss": 0.6374,
      "step": 1690
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.2924692928791046,
      "learning_rate": 1.3859275053304904e-05,
      "loss": 0.6933,
      "step": 1695
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.44005072116851807,
      "learning_rate": 1.3752665245202559e-05,
      "loss": 0.6913,
      "step": 1700
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.27338188886642456,
      "learning_rate": 1.3646055437100214e-05,
      "loss": 0.6607,
      "step": 1705
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.54692143201828,
      "learning_rate": 1.3539445628997869e-05,
      "loss": 0.6474,
      "step": 1710
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.2899147570133209,
      "learning_rate": 1.3432835820895523e-05,
      "loss": 0.6617,
      "step": 1715
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.853637158870697,
      "learning_rate": 1.3326226012793178e-05,
      "loss": 0.6652,
      "step": 1720
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.16075211763381958,
      "learning_rate": 1.3219616204690833e-05,
      "loss": 0.6644,
      "step": 1725
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.5768300890922546,
      "learning_rate": 1.3113006396588488e-05,
      "loss": 0.6407,
      "step": 1730
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.36618074774742126,
      "learning_rate": 1.3006396588486142e-05,
      "loss": 0.631,
      "step": 1735
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.393250048160553,
      "learning_rate": 1.2899786780383797e-05,
      "loss": 0.6894,
      "step": 1740
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.16580410301685333,
      "learning_rate": 1.2793176972281452e-05,
      "loss": 0.6921,
      "step": 1745
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.3606794476509094,
      "learning_rate": 1.2686567164179105e-05,
      "loss": 0.6484,
      "step": 1750
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.28691309690475464,
      "learning_rate": 1.257995735607676e-05,
      "loss": 0.6781,
      "step": 1755
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.6777899861335754,
      "learning_rate": 1.2473347547974414e-05,
      "loss": 0.6987,
      "step": 1760
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.2805548310279846,
      "learning_rate": 1.2366737739872069e-05,
      "loss": 0.661,
      "step": 1765
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.44818803668022156,
      "learning_rate": 1.2260127931769722e-05,
      "loss": 0.6558,
      "step": 1770
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.564558207988739,
      "learning_rate": 1.2153518123667377e-05,
      "loss": 0.6883,
      "step": 1775
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.49266985058784485,
      "learning_rate": 1.2046908315565032e-05,
      "loss": 0.6723,
      "step": 1780
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.5589272379875183,
      "learning_rate": 1.1940298507462686e-05,
      "loss": 0.6727,
      "step": 1785
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.4831165373325348,
      "learning_rate": 1.1833688699360341e-05,
      "loss": 0.6545,
      "step": 1790
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.2775633931159973,
      "learning_rate": 1.1727078891257996e-05,
      "loss": 0.6586,
      "step": 1795
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.43390023708343506,
      "learning_rate": 1.162046908315565e-05,
      "loss": 0.671,
      "step": 1800
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.1472308337688446,
      "learning_rate": 1.1513859275053305e-05,
      "loss": 0.6629,
      "step": 1805
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.21258753538131714,
      "learning_rate": 1.140724946695096e-05,
      "loss": 0.7005,
      "step": 1810
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.13851375877857208,
      "learning_rate": 1.1300639658848615e-05,
      "loss": 0.6705,
      "step": 1815
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.2994244694709778,
      "learning_rate": 1.119402985074627e-05,
      "loss": 0.6877,
      "step": 1820
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.3223628103733063,
      "learning_rate": 1.1087420042643924e-05,
      "loss": 0.6503,
      "step": 1825
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.3159940242767334,
      "learning_rate": 1.0980810234541579e-05,
      "loss": 0.6736,
      "step": 1830
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.47484922409057617,
      "learning_rate": 1.0874200426439234e-05,
      "loss": 0.6631,
      "step": 1835
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.052727222442627,
      "learning_rate": 1.0767590618336887e-05,
      "loss": 0.6596,
      "step": 1840
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.315295934677124,
      "learning_rate": 1.0660980810234541e-05,
      "loss": 0.686,
      "step": 1845
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.2225022166967392,
      "learning_rate": 1.0554371002132196e-05,
      "loss": 0.6945,
      "step": 1850
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.28455567359924316,
      "learning_rate": 1.0447761194029851e-05,
      "loss": 0.6731,
      "step": 1855
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.34978070855140686,
      "learning_rate": 1.0341151385927506e-05,
      "loss": 0.6743,
      "step": 1860
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.44299939274787903,
      "learning_rate": 1.023454157782516e-05,
      "loss": 0.7111,
      "step": 1865
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.7539863586425781,
      "learning_rate": 1.0127931769722815e-05,
      "loss": 0.6464,
      "step": 1870
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5005848407745361,
      "learning_rate": 1.002132196162047e-05,
      "loss": 0.6937,
      "step": 1875
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5789333333333333,
      "eval_loss": 0.6769688129425049,
      "eval_runtime": 17.8906,
      "eval_samples_per_second": 209.608,
      "eval_steps_per_second": 6.596,
      "step": 1876
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.39526981115341187,
      "learning_rate": 9.914712153518125e-06,
      "loss": 0.6529,
      "step": 1880
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.5626300573348999,
      "learning_rate": 9.80810234541578e-06,
      "loss": 0.6536,
      "step": 1885
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.011904239654541,
      "learning_rate": 9.701492537313434e-06,
      "loss": 0.6832,
      "step": 1890
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.20950503647327423,
      "learning_rate": 9.594882729211089e-06,
      "loss": 0.6743,
      "step": 1895
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.4118889570236206,
      "learning_rate": 9.488272921108744e-06,
      "loss": 0.6909,
      "step": 1900
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.18626834452152252,
      "learning_rate": 9.381663113006398e-06,
      "loss": 0.705,
      "step": 1905
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.2255084216594696,
      "learning_rate": 9.275053304904053e-06,
      "loss": 0.6689,
      "step": 1910
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.32827380299568176,
      "learning_rate": 9.168443496801706e-06,
      "loss": 0.6667,
      "step": 1915
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.452370285987854,
      "learning_rate": 9.06183368869936e-06,
      "loss": 0.6925,
      "step": 1920
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.42745348811149597,
      "learning_rate": 8.955223880597016e-06,
      "loss": 0.6615,
      "step": 1925
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.1782078742980957,
      "learning_rate": 8.84861407249467e-06,
      "loss": 0.6755,
      "step": 1930
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.2486790418624878,
      "learning_rate": 8.742004264392323e-06,
      "loss": 0.6598,
      "step": 1935
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.19132310152053833,
      "learning_rate": 8.635394456289978e-06,
      "loss": 0.6543,
      "step": 1940
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.796686589717865,
      "learning_rate": 8.528784648187633e-06,
      "loss": 0.6576,
      "step": 1945
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.17676232755184174,
      "learning_rate": 8.422174840085288e-06,
      "loss": 0.6948,
      "step": 1950
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.22592103481292725,
      "learning_rate": 8.315565031982942e-06,
      "loss": 0.678,
      "step": 1955
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.3258933126926422,
      "learning_rate": 8.208955223880597e-06,
      "loss": 0.6767,
      "step": 1960
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.2841905653476715,
      "learning_rate": 8.102345415778252e-06,
      "loss": 0.6907,
      "step": 1965
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.19143372774124146,
      "learning_rate": 7.995735607675907e-06,
      "loss": 0.6424,
      "step": 1970
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.5144029259681702,
      "learning_rate": 7.889125799573561e-06,
      "loss": 0.6666,
      "step": 1975
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.20798300206661224,
      "learning_rate": 7.782515991471216e-06,
      "loss": 0.6561,
      "step": 1980
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.5507712960243225,
      "learning_rate": 7.67590618336887e-06,
      "loss": 0.6582,
      "step": 1985
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.2972014546394348,
      "learning_rate": 7.5692963752665255e-06,
      "loss": 0.6903,
      "step": 1990
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.5243217349052429,
      "learning_rate": 7.4626865671641785e-06,
      "loss": 0.6516,
      "step": 1995
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.2954091727733612,
      "learning_rate": 7.356076759061833e-06,
      "loss": 0.6873,
      "step": 2000
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.42244890332221985,
      "learning_rate": 7.249466950959488e-06,
      "loss": 0.6707,
      "step": 2005
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.7245458364486694,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.6705,
      "step": 2010
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.5897809863090515,
      "learning_rate": 7.0362473347547975e-06,
      "loss": 0.6536,
      "step": 2015
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.7268871665000916,
      "learning_rate": 6.929637526652452e-06,
      "loss": 0.6987,
      "step": 2020
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7317787408828735,
      "learning_rate": 6.823027718550107e-06,
      "loss": 0.67,
      "step": 2025
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.21886880695819855,
      "learning_rate": 6.716417910447762e-06,
      "loss": 0.667,
      "step": 2030
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.3600420355796814,
      "learning_rate": 6.609808102345416e-06,
      "loss": 0.6832,
      "step": 2035
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.708007276058197,
      "learning_rate": 6.503198294243071e-06,
      "loss": 0.6739,
      "step": 2040
    },
    {
      "epoch": 4.36,
      "grad_norm": 1.6473708152770996,
      "learning_rate": 6.396588486140726e-06,
      "loss": 0.6487,
      "step": 2045
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.6396905779838562,
      "learning_rate": 6.28997867803838e-06,
      "loss": 0.6816,
      "step": 2050
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.16320891678333282,
      "learning_rate": 6.1833688699360345e-06,
      "loss": 0.6707,
      "step": 2055
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.5632389187812805,
      "learning_rate": 6.076759061833688e-06,
      "loss": 0.6798,
      "step": 2060
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.322733610868454,
      "learning_rate": 5.970149253731343e-06,
      "loss": 0.6486,
      "step": 2065
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.9542133212089539,
      "learning_rate": 5.863539445628998e-06,
      "loss": 0.6671,
      "step": 2070
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.21736344695091248,
      "learning_rate": 5.756929637526653e-06,
      "loss": 0.6549,
      "step": 2075
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.28586217761039734,
      "learning_rate": 5.650319829424307e-06,
      "loss": 0.6457,
      "step": 2080
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7068216800689697,
      "learning_rate": 5.543710021321962e-06,
      "loss": 0.6562,
      "step": 2085
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.2828645408153534,
      "learning_rate": 5.437100213219617e-06,
      "loss": 0.6714,
      "step": 2090
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.4406694173812866,
      "learning_rate": 5.330490405117271e-06,
      "loss": 0.6552,
      "step": 2095
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.3138512670993805,
      "learning_rate": 5.2238805970149255e-06,
      "loss": 0.646,
      "step": 2100
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.3833194971084595,
      "learning_rate": 5.11727078891258e-06,
      "loss": 0.6898,
      "step": 2105
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.5406767129898071,
      "learning_rate": 5.010660980810235e-06,
      "loss": 0.6603,
      "step": 2110
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.3065316379070282,
      "learning_rate": 4.90405117270789e-06,
      "loss": 0.688,
      "step": 2115
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.17264671623706818,
      "learning_rate": 4.797441364605544e-06,
      "loss": 0.6723,
      "step": 2120
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.2249528020620346,
      "learning_rate": 4.690831556503199e-06,
      "loss": 0.6521,
      "step": 2125
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.21267692744731903,
      "learning_rate": 4.584221748400853e-06,
      "loss": 0.6599,
      "step": 2130
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.24561788141727448,
      "learning_rate": 4.477611940298508e-06,
      "loss": 0.6713,
      "step": 2135
    },
    {
      "epoch": 4.56,
      "grad_norm": 1.0180352926254272,
      "learning_rate": 4.371002132196162e-06,
      "loss": 0.6565,
      "step": 2140
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.5230167508125305,
      "learning_rate": 4.264392324093816e-06,
      "loss": 0.6575,
      "step": 2145
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6080590486526489,
      "learning_rate": 4.157782515991471e-06,
      "loss": 0.6781,
      "step": 2150
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.394972026348114,
      "learning_rate": 4.051172707889126e-06,
      "loss": 0.6932,
      "step": 2155
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.9278691411018372,
      "learning_rate": 3.944562899786781e-06,
      "loss": 0.6283,
      "step": 2160
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.3256765305995941,
      "learning_rate": 3.837953091684435e-06,
      "loss": 0.6811,
      "step": 2165
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.43106502294540405,
      "learning_rate": 3.7313432835820893e-06,
      "loss": 0.6642,
      "step": 2170
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.26512110233306885,
      "learning_rate": 3.624733475479744e-06,
      "loss": 0.6548,
      "step": 2175
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.7567458152770996,
      "learning_rate": 3.5181236673773987e-06,
      "loss": 0.693,
      "step": 2180
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.6813324689865112,
      "learning_rate": 3.4115138592750535e-06,
      "loss": 0.6453,
      "step": 2185
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5289140939712524,
      "learning_rate": 3.304904051172708e-06,
      "loss": 0.6952,
      "step": 2190
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.32925984263420105,
      "learning_rate": 3.198294243070363e-06,
      "loss": 0.6705,
      "step": 2195
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.15043401718139648,
      "learning_rate": 3.0916844349680173e-06,
      "loss": 0.6799,
      "step": 2200
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.21513478457927704,
      "learning_rate": 2.9850746268656716e-06,
      "loss": 0.6346,
      "step": 2205
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.5901110172271729,
      "learning_rate": 2.8784648187633263e-06,
      "loss": 0.6564,
      "step": 2210
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.18733665347099304,
      "learning_rate": 2.771855010660981e-06,
      "loss": 0.6512,
      "step": 2215
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.7170325517654419,
      "learning_rate": 2.6652452025586354e-06,
      "loss": 0.6693,
      "step": 2220
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.5361260175704956,
      "learning_rate": 2.55863539445629e-06,
      "loss": 0.6719,
      "step": 2225
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.5559740662574768,
      "learning_rate": 2.452025586353945e-06,
      "loss": 0.6548,
      "step": 2230
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.4540775716304779,
      "learning_rate": 2.3454157782515996e-06,
      "loss": 0.6915,
      "step": 2235
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.16333653032779694,
      "learning_rate": 2.238805970149254e-06,
      "loss": 0.661,
      "step": 2240
    },
    {
      "epoch": 4.79,
      "grad_norm": 1.2140674591064453,
      "learning_rate": 2.132196162046908e-06,
      "loss": 0.6485,
      "step": 2245
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.48497453331947327,
      "learning_rate": 2.025586353944563e-06,
      "loss": 0.682,
      "step": 2250
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.5769209265708923,
      "learning_rate": 1.9189765458422177e-06,
      "loss": 0.6674,
      "step": 2255
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.3388761579990387,
      "learning_rate": 1.812366737739872e-06,
      "loss": 0.6814,
      "step": 2260
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.380491703748703,
      "learning_rate": 1.7057569296375267e-06,
      "loss": 0.6464,
      "step": 2265
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.4902530610561371,
      "learning_rate": 1.5991471215351815e-06,
      "loss": 0.6572,
      "step": 2270
    },
    {
      "epoch": 4.85,
      "grad_norm": 1.4383553266525269,
      "learning_rate": 1.4925373134328358e-06,
      "loss": 0.6834,
      "step": 2275
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.4202342927455902,
      "learning_rate": 1.3859275053304905e-06,
      "loss": 0.6597,
      "step": 2280
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.6906626224517822,
      "learning_rate": 1.279317697228145e-06,
      "loss": 0.6891,
      "step": 2285
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.36534583568573,
      "learning_rate": 1.1727078891257998e-06,
      "loss": 0.6654,
      "step": 2290
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.22818098962306976,
      "learning_rate": 1.066098081023454e-06,
      "loss": 0.641,
      "step": 2295
    },
    {
      "epoch": 4.9,
      "grad_norm": 1.2233145236968994,
      "learning_rate": 9.594882729211088e-07,
      "loss": 0.6369,
      "step": 2300
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.2983638346195221,
      "learning_rate": 8.528784648187634e-07,
      "loss": 0.6589,
      "step": 2305
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.5600255131721497,
      "learning_rate": 7.462686567164179e-07,
      "loss": 0.642,
      "step": 2310
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.271295964717865,
      "learning_rate": 6.396588486140725e-07,
      "loss": 0.6882,
      "step": 2315
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.3047320246696472,
      "learning_rate": 5.33049040511727e-07,
      "loss": 0.6622,
      "step": 2320
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.2932121753692627,
      "learning_rate": 4.264392324093817e-07,
      "loss": 0.6658,
      "step": 2325
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.9874823689460754,
      "learning_rate": 3.1982942430703626e-07,
      "loss": 0.6619,
      "step": 2330
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.6264281272888184,
      "learning_rate": 2.1321961620469084e-07,
      "loss": 0.678,
      "step": 2335
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.49118179082870483,
      "learning_rate": 1.0660980810234542e-07,
      "loss": 0.702,
      "step": 2340
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.45167651772499084,
      "learning_rate": 0.0,
      "loss": 0.6723,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5736,
      "eval_loss": 0.6760911345481873,
      "eval_runtime": 17.4911,
      "eval_samples_per_second": 214.395,
      "eval_steps_per_second": 6.746,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "step": 2345,
      "total_flos": 5.8118992210944e+18,
      "train_loss": 0.6811762883973275,
      "train_runtime": 808.2731,
      "train_samples_per_second": 92.79,
      "train_steps_per_second": 2.901
    }
  ],
  "logging_steps": 5,
  "max_steps": 2345,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 5.8118992210944e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}