{
  "best_metric": 0.04898680001497269,
  "best_model_checkpoint": "./vit-base-beans-demo-v5/checkpoint-2200",
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 2384,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 0.00019916107382550338,
      "loss": 0.5752,
      "step": 10
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019832214765100673,
      "loss": 0.3639,
      "step": 20
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019748322147651007,
      "loss": 0.414,
      "step": 30
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019664429530201342,
      "loss": 0.4994,
      "step": 40
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0001958053691275168,
      "loss": 0.3953,
      "step": 50
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019496644295302013,
      "loss": 0.1514,
      "step": 60
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0001941275167785235,
      "loss": 0.4222,
      "step": 70
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019328859060402688,
      "loss": 0.4522,
      "step": 80
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019244966442953022,
      "loss": 0.39,
      "step": 90
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019161073825503357,
      "loss": 0.1764,
      "step": 100
    },
    {
      "epoch": 0.17,
      "eval_accuracy": 0.9345088161209067,
      "eval_loss": 0.24242226779460907,
      "eval_runtime": 37.0601,
      "eval_samples_per_second": 64.274,
      "eval_steps_per_second": 8.041,
      "step": 100
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0001907718120805369,
      "loss": 0.199,
      "step": 110
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018993288590604028,
      "loss": 0.2534,
      "step": 120
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018909395973154363,
      "loss": 0.2201,
      "step": 130
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.000188255033557047,
      "loss": 0.1656,
      "step": 140
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00018741610738255034,
      "loss": 0.3582,
      "step": 150
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0001865771812080537,
      "loss": 0.3158,
      "step": 160
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018573825503355706,
      "loss": 0.2936,
      "step": 170
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0001848993288590604,
      "loss": 0.3009,
      "step": 180
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018406040268456378,
      "loss": 0.3815,
      "step": 190
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018322147651006712,
      "loss": 0.2401,
      "step": 200
    },
    {
      "epoch": 0.34,
      "eval_accuracy": 0.9160369437447523,
      "eval_loss": 0.29085931181907654,
      "eval_runtime": 37.2208,
      "eval_samples_per_second": 63.996,
      "eval_steps_per_second": 8.006,
      "step": 200
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018238255033557047,
      "loss": 0.3746,
      "step": 210
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00018154362416107384,
      "loss": 0.3423,
      "step": 220
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018070469798657718,
      "loss": 0.1224,
      "step": 230
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00017986577181208056,
      "loss": 0.2699,
      "step": 240
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0001790268456375839,
      "loss": 0.2199,
      "step": 250
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017818791946308727,
      "loss": 0.4065,
      "step": 260
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017734899328859062,
      "loss": 0.4393,
      "step": 270
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00017651006711409396,
      "loss": 0.3829,
      "step": 280
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00017567114093959733,
      "loss": 0.2467,
      "step": 290
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00017483221476510068,
      "loss": 0.1986,
      "step": 300
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.9357682619647355,
      "eval_loss": 0.22584371268749237,
      "eval_runtime": 37.6779,
      "eval_samples_per_second": 63.22,
      "eval_steps_per_second": 7.909,
      "step": 300
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00017399328859060405,
      "loss": 0.3105,
      "step": 310
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0001731543624161074,
      "loss": 0.1708,
      "step": 320
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00017231543624161074,
      "loss": 0.2205,
      "step": 330
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00017147651006711408,
      "loss": 0.1301,
      "step": 340
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017063758389261746,
      "loss": 0.2069,
      "step": 350
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00016979865771812083,
      "loss": 0.098,
      "step": 360
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00016895973154362417,
      "loss": 0.2552,
      "step": 370
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016812080536912754,
      "loss": 0.2953,
      "step": 380
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00016728187919463086,
      "loss": 0.2195,
      "step": 390
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016644295302013423,
      "loss": 0.1579,
      "step": 400
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.9240134340890008,
      "eval_loss": 0.2699319124221802,
      "eval_runtime": 37.7055,
      "eval_samples_per_second": 63.174,
      "eval_steps_per_second": 7.903,
      "step": 400
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016560402684563758,
      "loss": 0.3875,
      "step": 410
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016476510067114095,
      "loss": 0.254,
      "step": 420
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016401006711409396,
      "loss": 0.2794,
      "step": 430
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016317114093959733,
      "loss": 0.1986,
      "step": 440
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016233221476510067,
      "loss": 0.1392,
      "step": 450
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016149328859060404,
      "loss": 0.1405,
      "step": 460
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0001606543624161074,
      "loss": 0.0757,
      "step": 470
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015981543624161076,
      "loss": 0.1505,
      "step": 480
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0001589765100671141,
      "loss": 0.3554,
      "step": 490
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015813758389261745,
      "loss": 0.21,
      "step": 500
    },
    {
      "epoch": 0.84,
      "eval_accuracy": 0.9634760705289672,
      "eval_loss": 0.1416286677122116,
      "eval_runtime": 37.8399,
      "eval_samples_per_second": 62.949,
      "eval_steps_per_second": 7.875,
      "step": 500
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015729865771812082,
      "loss": 0.24,
      "step": 510
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00015645973154362417,
      "loss": 0.161,
      "step": 520
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00015562080536912754,
      "loss": 0.1271,
      "step": 530
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015478187919463088,
      "loss": 0.2553,
      "step": 540
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015394295302013423,
      "loss": 0.223,
      "step": 550
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015310402684563757,
      "loss": 0.1555,
      "step": 560
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00015226510067114095,
      "loss": 0.2612,
      "step": 570
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00015142617449664432,
      "loss": 0.1819,
      "step": 580
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015058724832214766,
      "loss": 0.2398,
      "step": 590
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014974832214765103,
      "loss": 0.0873,
      "step": 600
    },
    {
      "epoch": 1.01,
      "eval_accuracy": 0.9466834592779177,
      "eval_loss": 0.1723610907793045,
      "eval_runtime": 36.2673,
      "eval_samples_per_second": 65.679,
      "eval_steps_per_second": 8.217,
      "step": 600
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014890939597315435,
      "loss": 0.1876,
      "step": 610
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014807046979865772,
      "loss": 0.0848,
      "step": 620
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00014723154362416107,
      "loss": 0.0784,
      "step": 630
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00014639261744966444,
      "loss": 0.0697,
      "step": 640
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0001455536912751678,
      "loss": 0.1411,
      "step": 650
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014471476510067116,
      "loss": 0.0589,
      "step": 660
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0001438758389261745,
      "loss": 0.0867,
      "step": 670
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014303691275167785,
      "loss": 0.0671,
      "step": 680
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014219798657718122,
      "loss": 0.0821,
      "step": 690
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00014135906040268456,
      "loss": 0.1594,
      "step": 700
    },
    {
      "epoch": 1.17,
      "eval_accuracy": 0.9319899244332494,
      "eval_loss": 0.24392056465148926,
      "eval_runtime": 37.3417,
      "eval_samples_per_second": 63.789,
      "eval_steps_per_second": 7.98,
      "step": 700
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014052013422818793,
      "loss": 0.1412,
      "step": 710
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013968120805369128,
      "loss": 0.1345,
      "step": 720
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00013884228187919462,
      "loss": 0.0804,
      "step": 730
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.000138003355704698,
      "loss": 0.0858,
      "step": 740
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013716442953020134,
      "loss": 0.1313,
      "step": 750
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0001363255033557047,
      "loss": 0.0775,
      "step": 760
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013548657718120806,
      "loss": 0.119,
      "step": 770
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013464765100671143,
      "loss": 0.0896,
      "step": 780
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013380872483221477,
      "loss": 0.0962,
      "step": 790
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013296979865771812,
      "loss": 0.0824,
      "step": 800
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.9701931150293871,
      "eval_loss": 0.10386788845062256,
      "eval_runtime": 37.9403,
      "eval_samples_per_second": 62.783,
      "eval_steps_per_second": 7.854,
      "step": 800
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0001321308724832215,
      "loss": 0.0185,
      "step": 810
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013129194630872484,
      "loss": 0.0723,
      "step": 820
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0001304530201342282,
      "loss": 0.1575,
      "step": 830
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012961409395973155,
      "loss": 0.1435,
      "step": 840
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0001287751677852349,
      "loss": 0.0303,
      "step": 850
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012793624161073827,
      "loss": 0.1279,
      "step": 860
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012709731543624161,
      "loss": 0.1075,
      "step": 870
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00012625838926174499,
      "loss": 0.0728,
      "step": 880
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012541946308724833,
      "loss": 0.1591,
      "step": 890
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0001245805369127517,
      "loss": 0.0888,
      "step": 900
    },
    {
      "epoch": 1.51,
      "eval_accuracy": 0.9546599496221663,
      "eval_loss": 0.14956098794937134,
      "eval_runtime": 37.2049,
      "eval_samples_per_second": 64.024,
      "eval_steps_per_second": 8.01,
      "step": 900
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012374161073825502,
      "loss": 0.0903,
      "step": 910
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001229026845637584,
      "loss": 0.0751,
      "step": 920
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012206375838926174,
      "loss": 0.1672,
      "step": 930
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012122483221476511,
      "loss": 0.0507,
      "step": 940
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012038590604026847,
      "loss": 0.1459,
      "step": 950
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00011954697986577181,
      "loss": 0.0659,
      "step": 960
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011870805369127518,
      "loss": 0.0726,
      "step": 970
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011786912751677853,
      "loss": 0.0898,
      "step": 980
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011703020134228189,
      "loss": 0.1465,
      "step": 990
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011619127516778523,
      "loss": 0.1058,
      "step": 1000
    },
    {
      "epoch": 1.68,
      "eval_accuracy": 0.9748110831234257,
      "eval_loss": 0.1099860668182373,
      "eval_runtime": 37.5608,
      "eval_samples_per_second": 63.417,
      "eval_steps_per_second": 7.934,
      "step": 1000
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011535234899328859,
      "loss": 0.1159,
      "step": 1010
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011451342281879196,
      "loss": 0.0583,
      "step": 1020
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0001136744966442953,
      "loss": 0.1075,
      "step": 1030
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011283557046979866,
      "loss": 0.1809,
      "step": 1040
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011199664429530201,
      "loss": 0.1112,
      "step": 1050
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011115771812080538,
      "loss": 0.0408,
      "step": 1060
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00011031879194630873,
      "loss": 0.1428,
      "step": 1070
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010947986577181208,
      "loss": 0.0248,
      "step": 1080
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010864093959731546,
      "loss": 0.0873,
      "step": 1090
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00010780201342281879,
      "loss": 0.077,
      "step": 1100
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.9756507136859782,
      "eval_loss": 0.07842576503753662,
      "eval_runtime": 36.9225,
      "eval_samples_per_second": 64.513,
      "eval_steps_per_second": 8.071,
      "step": 1100
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010696308724832216,
      "loss": 0.1448,
      "step": 1110
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0001061241610738255,
      "loss": 0.1401,
      "step": 1120
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010528523489932886,
      "loss": 0.0186,
      "step": 1130
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010444630872483221,
      "loss": 0.0693,
      "step": 1140
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010360738255033558,
      "loss": 0.0521,
      "step": 1150
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010276845637583894,
      "loss": 0.0286,
      "step": 1160
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010192953020134228,
      "loss": 0.0321,
      "step": 1170
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010109060402684565,
      "loss": 0.1065,
      "step": 1180
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010025167785234899,
      "loss": 0.071,
      "step": 1190
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.941275167785236e-05,
      "loss": 0.0486,
      "step": 1200
    },
    {
      "epoch": 2.01,
      "eval_accuracy": 0.9710327455919395,
      "eval_loss": 0.09972918033599854,
      "eval_runtime": 38.6678,
      "eval_samples_per_second": 61.602,
      "eval_steps_per_second": 7.707,
      "step": 1200
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.857382550335572e-05,
      "loss": 0.04,
      "step": 1210
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.773489932885906e-05,
      "loss": 0.0241,
      "step": 1220
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.689597315436242e-05,
      "loss": 0.0574,
      "step": 1230
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.605704697986578e-05,
      "loss": 0.0196,
      "step": 1240
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.521812080536914e-05,
      "loss": 0.0344,
      "step": 1250
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.43791946308725e-05,
      "loss": 0.0342,
      "step": 1260
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.354026845637585e-05,
      "loss": 0.0121,
      "step": 1270
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.27013422818792e-05,
      "loss": 0.0251,
      "step": 1280
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.186241610738255e-05,
      "loss": 0.0503,
      "step": 1290
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.102348993288591e-05,
      "loss": 0.0087,
      "step": 1300
    },
    {
      "epoch": 2.18,
      "eval_accuracy": 0.9811083123425692,
      "eval_loss": 0.07042522728443146,
      "eval_runtime": 37.2902,
      "eval_samples_per_second": 63.877,
      "eval_steps_per_second": 7.991,
      "step": 1300
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.018456375838926e-05,
      "loss": 0.0676,
      "step": 1310
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.934563758389262e-05,
      "loss": 0.0445,
      "step": 1320
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.850671140939599e-05,
      "loss": 0.0157,
      "step": 1330
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.766778523489933e-05,
      "loss": 0.045,
      "step": 1340
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.682885906040269e-05,
      "loss": 0.0071,
      "step": 1350
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.598993288590605e-05,
      "loss": 0.0219,
      "step": 1360
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.51510067114094e-05,
      "loss": 0.0201,
      "step": 1370
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.431208053691275e-05,
      "loss": 0.0097,
      "step": 1380
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.347315436241611e-05,
      "loss": 0.0552,
      "step": 1390
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.263422818791947e-05,
      "loss": 0.0382,
      "step": 1400
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.9760705289672544,
      "eval_loss": 0.09097278863191605,
      "eval_runtime": 38.1785,
      "eval_samples_per_second": 62.391,
      "eval_steps_per_second": 7.805,
      "step": 1400
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.179530201342283e-05,
      "loss": 0.0136,
      "step": 1410
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.095637583892619e-05,
      "loss": 0.0735,
      "step": 1420
    },
    {
      "epoch": 2.4,
      "learning_rate": 8.011744966442953e-05,
      "loss": 0.0244,
      "step": 1430
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.927852348993289e-05,
      "loss": 0.0208,
      "step": 1440
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.843959731543625e-05,
      "loss": 0.0289,
      "step": 1450
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.760067114093959e-05,
      "loss": 0.023,
      "step": 1460
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.676174496644296e-05,
      "loss": 0.006,
      "step": 1470
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.592281879194631e-05,
      "loss": 0.0415,
      "step": 1480
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.508389261744967e-05,
      "loss": 0.0253,
      "step": 1490
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.424496644295303e-05,
      "loss": 0.046,
      "step": 1500
    },
    {
      "epoch": 2.52,
      "eval_accuracy": 0.984047019311503,
      "eval_loss": 0.05878419801592827,
      "eval_runtime": 37.7397,
      "eval_samples_per_second": 63.116,
      "eval_steps_per_second": 7.896,
      "step": 1500
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.340604026845638e-05,
      "loss": 0.0409,
      "step": 1510
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.256711409395973e-05,
      "loss": 0.0335,
      "step": 1520
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.172818791946309e-05,
      "loss": 0.0049,
      "step": 1530
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.088926174496645e-05,
      "loss": 0.0119,
      "step": 1540
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.00503355704698e-05,
      "loss": 0.0188,
      "step": 1550
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.921140939597316e-05,
      "loss": 0.0047,
      "step": 1560
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.83724832214765e-05,
      "loss": 0.0053,
      "step": 1570
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.753355704697986e-05,
      "loss": 0.0034,
      "step": 1580
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.669463087248322e-05,
      "loss": 0.0026,
      "step": 1590
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.585570469798658e-05,
      "loss": 0.0163,
      "step": 1600
    },
    {
      "epoch": 2.68,
      "eval_accuracy": 0.9836272040302267,
      "eval_loss": 0.06088751554489136,
      "eval_runtime": 38.2646,
      "eval_samples_per_second": 62.251,
      "eval_steps_per_second": 7.788,
      "step": 1600
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.501677852348994e-05,
      "loss": 0.0029,
      "step": 1610
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.41778523489933e-05,
      "loss": 0.0027,
      "step": 1620
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.333892617449664e-05,
      "loss": 0.0394,
      "step": 1630
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.25e-05,
      "loss": 0.0235,
      "step": 1640
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.166107382550336e-05,
      "loss": 0.0199,
      "step": 1650
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.082214765100671e-05,
      "loss": 0.0037,
      "step": 1660
    },
    {
      "epoch": 2.8,
      "learning_rate": 5.998322147651006e-05,
      "loss": 0.0151,
      "step": 1670
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.914429530201343e-05,
      "loss": 0.0472,
      "step": 1680
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.8305369127516786e-05,
      "loss": 0.0406,
      "step": 1690
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.746644295302014e-05,
      "loss": 0.0876,
      "step": 1700
    },
    {
      "epoch": 2.85,
      "eval_accuracy": 0.9811083123425692,
      "eval_loss": 0.06890641152858734,
      "eval_runtime": 38.0085,
      "eval_samples_per_second": 62.67,
      "eval_steps_per_second": 7.84,
      "step": 1700
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.6627516778523496e-05,
      "loss": 0.0033,
      "step": 1710
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.578859060402685e-05,
      "loss": 0.003,
      "step": 1720
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.49496644295302e-05,
      "loss": 0.0024,
      "step": 1730
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.411073825503356e-05,
      "loss": 0.0122,
      "step": 1740
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.327181208053692e-05,
      "loss": 0.0072,
      "step": 1750
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.2432885906040274e-05,
      "loss": 0.0023,
      "step": 1760
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.1593959731543626e-05,
      "loss": 0.0098,
      "step": 1770
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.0755033557046984e-05,
      "loss": 0.035,
      "step": 1780
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.9916107382550336e-05,
      "loss": 0.0349,
      "step": 1790
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.9077181208053694e-05,
      "loss": 0.0072,
      "step": 1800
    },
    {
      "epoch": 3.02,
      "eval_accuracy": 0.9861460957178841,
      "eval_loss": 0.06315400451421738,
      "eval_runtime": 38.6086,
      "eval_samples_per_second": 61.696,
      "eval_steps_per_second": 7.718,
      "step": 1800
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.823825503355705e-05,
      "loss": 0.0022,
      "step": 1810
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.7399328859060404e-05,
      "loss": 0.0131,
      "step": 1820
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.6560402684563755e-05,
      "loss": 0.0031,
      "step": 1830
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.572147651006712e-05,
      "loss": 0.0024,
      "step": 1840
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.488255033557047e-05,
      "loss": 0.0022,
      "step": 1850
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.4043624161073823e-05,
      "loss": 0.0029,
      "step": 1860
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.320469798657718e-05,
      "loss": 0.0018,
      "step": 1870
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.236577181208054e-05,
      "loss": 0.0019,
      "step": 1880
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.152684563758389e-05,
      "loss": 0.0046,
      "step": 1890
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.068791946308725e-05,
      "loss": 0.0022,
      "step": 1900
    },
    {
      "epoch": 3.19,
      "eval_accuracy": 0.9844668345927792,
      "eval_loss": 0.0669856071472168,
      "eval_runtime": 38.9734,
      "eval_samples_per_second": 61.119,
      "eval_steps_per_second": 7.646,
      "step": 1900
    },
    {
      "epoch": 3.2,
      "learning_rate": 3.984899328859061e-05,
      "loss": 0.0031,
      "step": 1910
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.901006711409396e-05,
      "loss": 0.0033,
      "step": 1920
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.817114093959732e-05,
      "loss": 0.0171,
      "step": 1930
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.733221476510067e-05,
      "loss": 0.0022,
      "step": 1940
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.649328859060403e-05,
      "loss": 0.0023,
      "step": 1950
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.5654362416107386e-05,
      "loss": 0.0036,
      "step": 1960
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.481543624161074e-05,
      "loss": 0.0038,
      "step": 1970
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.3976510067114096e-05,
      "loss": 0.0016,
      "step": 1980
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.3137583892617455e-05,
      "loss": 0.0022,
      "step": 1990
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.2298657718120806e-05,
      "loss": 0.0018,
      "step": 2000
    },
    {
      "epoch": 3.36,
      "eval_accuracy": 0.9844668345927792,
      "eval_loss": 0.0704568475484848,
      "eval_runtime": 36.8214,
      "eval_samples_per_second": 64.691,
      "eval_steps_per_second": 8.093,
      "step": 2000
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.145973154362416e-05,
      "loss": 0.0036,
      "step": 2010
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.062080536912752e-05,
      "loss": 0.0018,
      "step": 2020
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9781879194630874e-05,
      "loss": 0.0015,
      "step": 2030
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.894295302013423e-05,
      "loss": 0.0018,
      "step": 2040
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8104026845637588e-05,
      "loss": 0.0021,
      "step": 2050
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.7265100671140943e-05,
      "loss": 0.0016,
      "step": 2060
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6426174496644297e-05,
      "loss": 0.0019,
      "step": 2070
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.558724832214765e-05,
      "loss": 0.0025,
      "step": 2080
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4748322147651007e-05,
      "loss": 0.0071,
      "step": 2090
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.3909395973154362e-05,
      "loss": 0.0023,
      "step": 2100
    },
    {
      "epoch": 3.52,
      "eval_accuracy": 0.9861460957178841,
      "eval_loss": 0.0588483065366745,
      "eval_runtime": 36.7062,
      "eval_samples_per_second": 64.894,
      "eval_steps_per_second": 8.119,
      "step": 2100
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.3070469798657717e-05,
      "loss": 0.0111,
      "step": 2110
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2231543624161076e-05,
      "loss": 0.0019,
      "step": 2120
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.139261744966443e-05,
      "loss": 0.007,
      "step": 2130
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0553691275167785e-05,
      "loss": 0.0022,
      "step": 2140
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9714765100671144e-05,
      "loss": 0.0019,
      "step": 2150
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.8875838926174495e-05,
      "loss": 0.0017,
      "step": 2160
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.8036912751677854e-05,
      "loss": 0.0017,
      "step": 2170
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.719798657718121e-05,
      "loss": 0.0134,
      "step": 2180
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6359060402684563e-05,
      "loss": 0.0017,
      "step": 2190
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5520134228187922e-05,
      "loss": 0.0017,
      "step": 2200
    },
    {
      "epoch": 3.69,
      "eval_accuracy": 0.9882451721242653,
      "eval_loss": 0.04898680001497269,
      "eval_runtime": 36.2766,
      "eval_samples_per_second": 65.662,
      "eval_steps_per_second": 8.215,
      "step": 2200
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4681208053691275e-05,
      "loss": 0.002,
      "step": 2210
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.3842281879194632e-05,
      "loss": 0.0019,
      "step": 2220
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.3003355704697987e-05,
      "loss": 0.0017,
      "step": 2230
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.2164429530201343e-05,
      "loss": 0.0021,
      "step": 2240
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1325503355704698e-05,
      "loss": 0.0027,
      "step": 2250
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0486577181208055e-05,
      "loss": 0.0027,
      "step": 2260
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.64765100671141e-06,
      "loss": 0.0021,
      "step": 2270
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.808724832214765e-06,
      "loss": 0.0017,
      "step": 2280
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.969798657718121e-06,
      "loss": 0.0018,
      "step": 2290
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.130872483221476e-06,
      "loss": 0.0248,
      "step": 2300
    },
    {
      "epoch": 3.86,
      "eval_accuracy": 0.9882451721242653,
      "eval_loss": 0.04982466623187065,
      "eval_runtime": 37.2392,
      "eval_samples_per_second": 63.965,
      "eval_steps_per_second": 8.002,
      "step": 2300
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.291946308724833e-06,
      "loss": 0.0024,
      "step": 2310
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.453020134228188e-06,
      "loss": 0.0017,
      "step": 2320
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.6140939597315445e-06,
      "loss": 0.0315,
      "step": 2330
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.7751677852349e-06,
      "loss": 0.0018,
      "step": 2340
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.936241610738255e-06,
      "loss": 0.0018,
      "step": 2350
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.097315436241611e-06,
      "loss": 0.0016,
      "step": 2360
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.2583892617449663e-06,
      "loss": 0.0027,
      "step": 2370
    },
    {
      "epoch": 3.99,
      "learning_rate": 4.1946308724832216e-07,
      "loss": 0.0133,
      "step": 2380
    },
    {
      "epoch": 4.0,
      "step": 2384,
      "total_flos": 2.953718820314874e+18,
      "train_loss": 0.09724263496534437,
      "train_runtime": 2056.7864,
      "train_samples_per_second": 18.53,
      "train_steps_per_second": 1.159
    }
  ],
  "logging_steps": 10,
  "max_steps": 2384,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 2.953718820314874e+18,
  "trial_name": null,
  "trial_params": null
}