{
  "best_metric": 0.49814796447753906,
  "best_model_checkpoint": "./vit-kimchi/checkpoint-1300",
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 1928,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 0.00019896265560165978,
      "loss": 2.2446,
      "step": 10
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019792531120331952,
      "loss": 1.8517,
      "step": 20
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019688796680497928,
      "loss": 1.5648,
      "step": 30
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.000195850622406639,
      "loss": 1.2418,
      "step": 40
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019481327800829876,
      "loss": 1.1769,
      "step": 50
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019377593360995853,
      "loss": 1.273,
      "step": 60
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019273858921161827,
      "loss": 0.972,
      "step": 70
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019170124481327803,
      "loss": 1.0242,
      "step": 80
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019066390041493777,
      "loss": 0.9587,
      "step": 90
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0001896265560165975,
      "loss": 0.8988,
      "step": 100
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.7145454545454546,
      "eval_loss": 0.9406929016113281,
      "eval_runtime": 49.3896,
      "eval_samples_per_second": 22.272,
      "eval_steps_per_second": 2.794,
      "step": 100
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018858921161825728,
      "loss": 0.925,
      "step": 110
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00018755186721991702,
      "loss": 0.8407,
      "step": 120
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018651452282157679,
      "loss": 0.7866,
      "step": 130
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018547717842323653,
      "loss": 0.7472,
      "step": 140
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0001844398340248963,
      "loss": 0.8416,
      "step": 150
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.000183402489626556,
      "loss": 0.8749,
      "step": 160
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018236514522821577,
      "loss": 0.8141,
      "step": 170
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00018132780082987554,
      "loss": 0.7578,
      "step": 180
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018029045643153528,
      "loss": 0.8922,
      "step": 190
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017925311203319504,
      "loss": 0.6914,
      "step": 200
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.8081818181818182,
      "eval_loss": 0.6929934024810791,
      "eval_runtime": 53.9344,
      "eval_samples_per_second": 20.395,
      "eval_steps_per_second": 2.559,
      "step": 200
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017821576763485478,
      "loss": 0.6568,
      "step": 210
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017717842323651452,
      "loss": 0.7086,
      "step": 220
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0001761410788381743,
      "loss": 0.8155,
      "step": 230
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00017510373443983403,
      "loss": 0.7773,
      "step": 240
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0001740663900414938,
      "loss": 0.6456,
      "step": 250
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017302904564315354,
      "loss": 0.59,
      "step": 260
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0001719917012448133,
      "loss": 0.6845,
      "step": 270
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017095435684647304,
      "loss": 0.8958,
      "step": 280
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00016991701244813278,
      "loss": 0.61,
      "step": 290
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00016887966804979255,
      "loss": 0.5689,
      "step": 300
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 0.7827272727272727,
      "eval_loss": 0.691547691822052,
      "eval_runtime": 52.4462,
      "eval_samples_per_second": 20.974,
      "eval_steps_per_second": 2.631,
      "step": 300
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0001678423236514523,
      "loss": 0.5978,
      "step": 310
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016680497925311205,
      "loss": 0.6593,
      "step": 320
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0001657676348547718,
      "loss": 0.6557,
      "step": 330
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016473029045643153,
      "loss": 0.6392,
      "step": 340
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0001636929460580913,
      "loss": 0.7019,
      "step": 350
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00016265560165975104,
      "loss": 0.713,
      "step": 360
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0001616182572614108,
      "loss": 0.7038,
      "step": 370
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016058091286307055,
      "loss": 0.8199,
      "step": 380
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0001595435684647303,
      "loss": 0.7091,
      "step": 390
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015850622406639005,
      "loss": 0.6755,
      "step": 400
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.7627272727272727,
      "eval_loss": 0.7460294961929321,
      "eval_runtime": 48.0299,
      "eval_samples_per_second": 22.902,
      "eval_steps_per_second": 2.873,
      "step": 400
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0001574688796680498,
      "loss": 0.7875,
      "step": 410
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00015643153526970956,
      "loss": 0.6854,
      "step": 420
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0001553941908713693,
      "loss": 0.5474,
      "step": 430
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015435684647302906,
      "loss": 0.7665,
      "step": 440
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0001533195020746888,
      "loss": 0.6185,
      "step": 450
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015228215767634854,
      "loss": 0.8083,
      "step": 460
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0001512448132780083,
      "loss": 0.5685,
      "step": 470
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00015020746887966805,
      "loss": 0.5072,
      "step": 480
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014917012448132782,
      "loss": 0.3516,
      "step": 490
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014813278008298756,
      "loss": 0.3364,
      "step": 500
    },
    {
      "epoch": 1.04,
      "eval_accuracy": 0.8227272727272728,
      "eval_loss": 0.5715581774711609,
      "eval_runtime": 52.5205,
      "eval_samples_per_second": 20.944,
      "eval_steps_per_second": 2.628,
      "step": 500
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00014709543568464732,
      "loss": 0.3399,
      "step": 510
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014605809128630706,
      "loss": 0.3942,
      "step": 520
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0001450207468879668,
      "loss": 0.3344,
      "step": 530
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014398340248962657,
      "loss": 0.3416,
      "step": 540
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0001429460580912863,
      "loss": 0.3703,
      "step": 550
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014190871369294607,
      "loss": 0.3983,
      "step": 560
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014087136929460581,
      "loss": 0.3765,
      "step": 570
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00013983402489626555,
      "loss": 0.4639,
      "step": 580
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00013879668049792532,
      "loss": 0.3725,
      "step": 590
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013775933609958506,
      "loss": 0.3872,
      "step": 600
    },
    {
      "epoch": 1.24,
      "eval_accuracy": 0.8009090909090909,
      "eval_loss": 0.6399267315864563,
      "eval_runtime": 48.4264,
      "eval_samples_per_second": 22.715,
      "eval_steps_per_second": 2.85,
      "step": 600
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00013672199170124483,
      "loss": 0.3308,
      "step": 610
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013568464730290457,
      "loss": 0.4045,
      "step": 620
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013464730290456433,
      "loss": 0.4431,
      "step": 630
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013360995850622407,
      "loss": 0.4729,
      "step": 640
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0001325726141078838,
      "loss": 0.3641,
      "step": 650
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013153526970954358,
      "loss": 0.2814,
      "step": 660
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013049792531120332,
      "loss": 0.4076,
      "step": 670
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012946058091286308,
      "loss": 0.5521,
      "step": 680
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012842323651452285,
      "loss": 0.4752,
      "step": 690
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00012738589211618256,
      "loss": 0.4035,
      "step": 700
    },
    {
      "epoch": 1.45,
      "eval_accuracy": 0.7927272727272727,
      "eval_loss": 0.6669933795928955,
      "eval_runtime": 232.0646,
      "eval_samples_per_second": 4.74,
      "eval_steps_per_second": 0.595,
      "step": 700
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012634854771784233,
      "loss": 0.356,
      "step": 710
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012531120331950207,
      "loss": 0.3443,
      "step": 720
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012427385892116184,
      "loss": 0.3427,
      "step": 730
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00012323651452282158,
      "loss": 0.4563,
      "step": 740
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012219917012448134,
      "loss": 0.4225,
      "step": 750
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012116182572614108,
      "loss": 0.2857,
      "step": 760
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00012012448132780084,
      "loss": 0.3558,
      "step": 770
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011908713692946059,
      "loss": 0.3012,
      "step": 780
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011804979253112034,
      "loss": 0.2585,
      "step": 790
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0001170124481327801,
      "loss": 0.2907,
      "step": 800
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 0.8018181818181818,
      "eval_loss": 0.619886040687561,
      "eval_runtime": 182.6836,
      "eval_samples_per_second": 6.021,
      "eval_steps_per_second": 0.755,
      "step": 800
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011597510373443985,
      "loss": 0.2441,
      "step": 810
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011493775933609959,
      "loss": 0.3021,
      "step": 820
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011390041493775934,
      "loss": 0.4212,
      "step": 830
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0001128630705394191,
      "loss": 0.3419,
      "step": 840
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011182572614107885,
      "loss": 0.4709,
      "step": 850
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0001107883817427386,
      "loss": 0.3137,
      "step": 860
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00010975103734439835,
      "loss": 0.2889,
      "step": 870
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010871369294605809,
      "loss": 0.3452,
      "step": 880
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00010767634854771785,
      "loss": 0.2411,
      "step": 890
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0001066390041493776,
      "loss": 0.2934,
      "step": 900
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.8109090909090909,
      "eval_loss": 0.5889019966125488,
      "eval_runtime": 74.2537,
      "eval_samples_per_second": 14.814,
      "eval_steps_per_second": 1.858,
      "step": 900
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010560165975103735,
      "loss": 0.3304,
      "step": 910
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0001045643153526971,
      "loss": 0.2853,
      "step": 920
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010352697095435686,
      "loss": 0.3694,
      "step": 930
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0001024896265560166,
      "loss": 0.2806,
      "step": 940
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010145228215767635,
      "loss": 0.3782,
      "step": 950
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0001004149377593361,
      "loss": 0.2552,
      "step": 960
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.937759336099586e-05,
      "loss": 0.3536,
      "step": 970
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.83402489626556e-05,
      "loss": 0.0899,
      "step": 980
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.730290456431535e-05,
      "loss": 0.1806,
      "step": 990
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.626556016597512e-05,
      "loss": 0.1166,
      "step": 1000
    },
    {
      "epoch": 2.07,
      "eval_accuracy": 0.8390909090909091,
      "eval_loss": 0.5368185043334961,
      "eval_runtime": 70.8059,
      "eval_samples_per_second": 15.535,
      "eval_steps_per_second": 1.949,
      "step": 1000
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.522821576763486e-05,
      "loss": 0.1473,
      "step": 1010
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.419087136929461e-05,
      "loss": 0.1777,
      "step": 1020
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.315352697095436e-05,
      "loss": 0.2064,
      "step": 1030
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.21161825726141e-05,
      "loss": 0.1514,
      "step": 1040
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.107883817427387e-05,
      "loss": 0.2532,
      "step": 1050
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.004149377593362e-05,
      "loss": 0.1578,
      "step": 1060
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.900414937759336e-05,
      "loss": 0.1755,
      "step": 1070
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.796680497925311e-05,
      "loss": 0.1563,
      "step": 1080
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.692946058091287e-05,
      "loss": 0.133,
      "step": 1090
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.58921161825726e-05,
      "loss": 0.2452,
      "step": 1100
    },
    {
      "epoch": 2.28,
      "eval_accuracy": 0.83,
      "eval_loss": 0.5816710591316223,
      "eval_runtime": 82.9166,
      "eval_samples_per_second": 13.266,
      "eval_steps_per_second": 1.664,
      "step": 1100
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.485477178423237e-05,
      "loss": 0.155,
      "step": 1110
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.381742738589213e-05,
      "loss": 0.1068,
      "step": 1120
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.278008298755187e-05,
      "loss": 0.0867,
      "step": 1130
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.174273858921162e-05,
      "loss": 0.193,
      "step": 1140
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.070539419087137e-05,
      "loss": 0.1487,
      "step": 1150
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.966804979253112e-05,
      "loss": 0.1718,
      "step": 1160
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.863070539419088e-05,
      "loss": 0.2072,
      "step": 1170
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.759336099585063e-05,
      "loss": 0.1637,
      "step": 1180
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.655601659751037e-05,
      "loss": 0.0945,
      "step": 1190
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.551867219917012e-05,
      "loss": 0.1838,
      "step": 1200
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 0.8336363636363636,
      "eval_loss": 0.5901106595993042,
      "eval_runtime": 65.6982,
      "eval_samples_per_second": 16.743,
      "eval_steps_per_second": 2.101,
      "step": 1200
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.448132780082988e-05,
      "loss": 0.2265,
      "step": 1210
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.344398340248963e-05,
      "loss": 0.1791,
      "step": 1220
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.240663900414938e-05,
      "loss": 0.1296,
      "step": 1230
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.136929460580914e-05,
      "loss": 0.1491,
      "step": 1240
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.033195020746889e-05,
      "loss": 0.2888,
      "step": 1250
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.929460580912863e-05,
      "loss": 0.1386,
      "step": 1260
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.825726141078838e-05,
      "loss": 0.0571,
      "step": 1270
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.721991701244813e-05,
      "loss": 0.1371,
      "step": 1280
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.618257261410789e-05,
      "loss": 0.1443,
      "step": 1290
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.514522821576764e-05,
      "loss": 0.168,
      "step": 1300
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 0.86,
      "eval_loss": 0.49814796447753906,
      "eval_runtime": 74.5165,
      "eval_samples_per_second": 14.762,
      "eval_steps_per_second": 1.852,
      "step": 1300
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.41078838174274e-05,
      "loss": 0.1147,
      "step": 1310
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.307053941908713e-05,
      "loss": 0.1334,
      "step": 1320
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.203319502074689e-05,
      "loss": 0.0889,
      "step": 1330
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.099585062240665e-05,
      "loss": 0.1234,
      "step": 1340
    },
    {
      "epoch": 2.8,
      "learning_rate": 5.995850622406639e-05,
      "loss": 0.0907,
      "step": 1350
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.8921161825726146e-05,
      "loss": 0.1191,
      "step": 1360
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.78838174273859e-05,
      "loss": 0.0852,
      "step": 1370
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.6846473029045646e-05,
      "loss": 0.1637,
      "step": 1380
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.58091286307054e-05,
      "loss": 0.0901,
      "step": 1390
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.477178423236515e-05,
      "loss": 0.1578,
      "step": 1400
    },
    {
      "epoch": 2.9,
      "eval_accuracy": 0.8581818181818182,
      "eval_loss": 0.5170807242393494,
      "eval_runtime": 44.8814,
      "eval_samples_per_second": 24.509,
      "eval_steps_per_second": 3.075,
      "step": 1400
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.37344398340249e-05,
      "loss": 0.2121,
      "step": 1410
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.269709543568465e-05,
      "loss": 0.1638,
      "step": 1420
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.1659751037344404e-05,
      "loss": 0.1259,
      "step": 1430
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.062240663900415e-05,
      "loss": 0.2338,
      "step": 1440
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.9585062240663904e-05,
      "loss": 0.0752,
      "step": 1450
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.854771784232366e-05,
      "loss": 0.032,
      "step": 1460
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.75103734439834e-05,
      "loss": 0.0438,
      "step": 1470
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.6473029045643156e-05,
      "loss": 0.02,
      "step": 1480
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.543568464730291e-05,
      "loss": 0.0427,
      "step": 1490
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.4398340248962656e-05,
      "loss": 0.0172,
      "step": 1500
    },
    {
      "epoch": 3.11,
      "eval_accuracy": 0.8536363636363636,
      "eval_loss": 0.5450488924980164,
      "eval_runtime": 48.9818,
      "eval_samples_per_second": 22.457,
      "eval_steps_per_second": 2.817,
      "step": 1500
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.336099585062241e-05,
      "loss": 0.0196,
      "step": 1510
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.232365145228216e-05,
      "loss": 0.0644,
      "step": 1520
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.128630705394191e-05,
      "loss": 0.0208,
      "step": 1530
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.024896265560166e-05,
      "loss": 0.0379,
      "step": 1540
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9211618257261414e-05,
      "loss": 0.0425,
      "step": 1550
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.817427385892116e-05,
      "loss": 0.0271,
      "step": 1560
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.7136929460580914e-05,
      "loss": 0.0711,
      "step": 1570
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.609958506224067e-05,
      "loss": 0.0198,
      "step": 1580
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.506224066390041e-05,
      "loss": 0.0543,
      "step": 1590
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4024896265560166e-05,
      "loss": 0.052,
      "step": 1600
    },
    {
      "epoch": 3.32,
      "eval_accuracy": 0.8463636363636363,
      "eval_loss": 0.560087263584137,
      "eval_runtime": 52.4163,
      "eval_samples_per_second": 20.986,
      "eval_steps_per_second": 2.633,
      "step": 1600
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.298755186721992e-05,
      "loss": 0.021,
      "step": 1610
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.1950207468879666e-05,
      "loss": 0.044,
      "step": 1620
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.091286307053942e-05,
      "loss": 0.0595,
      "step": 1630
    },
    {
      "epoch": 3.4,
      "learning_rate": 2.9875518672199172e-05,
      "loss": 0.0108,
      "step": 1640
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.883817427385892e-05,
      "loss": 0.0932,
      "step": 1650
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.7800829875518675e-05,
      "loss": 0.0853,
      "step": 1660
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.6763485477178424e-05,
      "loss": 0.021,
      "step": 1670
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5726141078838174e-05,
      "loss": 0.038,
      "step": 1680
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4688796680497927e-05,
      "loss": 0.039,
      "step": 1690
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.3651452282157677e-05,
      "loss": 0.0263,
      "step": 1700
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.8690909090909091,
      "eval_loss": 0.5324931740760803,
      "eval_runtime": 56.072,
      "eval_samples_per_second": 19.618,
      "eval_steps_per_second": 2.461,
      "step": 1700
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.261410788381743e-05,
      "loss": 0.024,
      "step": 1710
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.157676348547718e-05,
      "loss": 0.0129,
      "step": 1720
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.053941908713693e-05,
      "loss": 0.0889,
      "step": 1730
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9502074688796682e-05,
      "loss": 0.1226,
      "step": 1740
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8464730290456432e-05,
      "loss": 0.0127,
      "step": 1750
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.7427385892116182e-05,
      "loss": 0.0624,
      "step": 1760
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6390041493775935e-05,
      "loss": 0.0191,
      "step": 1770
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5352697095435685e-05,
      "loss": 0.0122,
      "step": 1780
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4315352697095436e-05,
      "loss": 0.0154,
      "step": 1790
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3278008298755187e-05,
      "loss": 0.0113,
      "step": 1800
    },
    {
      "epoch": 3.73,
      "eval_accuracy": 0.8663636363636363,
      "eval_loss": 0.5351160764694214,
      "eval_runtime": 55.1313,
      "eval_samples_per_second": 19.952,
      "eval_steps_per_second": 2.503,
      "step": 1800
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.2240663900414937e-05,
      "loss": 0.089,
      "step": 1810
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.120331950207469e-05,
      "loss": 0.0482,
      "step": 1820
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.016597510373444e-05,
      "loss": 0.089,
      "step": 1830
    },
    {
      "epoch": 3.82,
      "learning_rate": 9.128630705394191e-06,
      "loss": 0.0537,
      "step": 1840
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.091286307053943e-06,
      "loss": 0.023,
      "step": 1850
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.053941908713693e-06,
      "loss": 0.0528,
      "step": 1860
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.016597510373445e-06,
      "loss": 0.1387,
      "step": 1870
    },
    {
      "epoch": 3.9,
      "learning_rate": 4.979253112033195e-06,
      "loss": 0.0683,
      "step": 1880
    },
    {
      "epoch": 3.92,
      "learning_rate": 3.941908713692946e-06,
      "loss": 0.0383,
      "step": 1890
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.9045643153526973e-06,
      "loss": 0.0254,
      "step": 1900
    },
    {
      "epoch": 3.94,
      "eval_accuracy": 0.8690909090909091,
      "eval_loss": 0.5347076654434204,
      "eval_runtime": 52.6092,
      "eval_samples_per_second": 20.909,
      "eval_steps_per_second": 2.623,
      "step": 1900
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.8672199170124482e-06,
      "loss": 0.037,
      "step": 1910
    },
    {
      "epoch": 3.98,
      "learning_rate": 8.298755186721992e-07,
      "loss": 0.0492,
      "step": 1920
    },
    {
      "epoch": 4.0,
      "step": 1928,
      "total_flos": 2.386945805561856e+18,
      "train_loss": 0.3497509906586275,
      "train_runtime": 9933.302,
      "train_samples_per_second": 3.101,
      "train_steps_per_second": 0.194
    }
  ],
  "logging_steps": 10,
  "max_steps": 1928,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 2.386945805561856e+18,
  "trial_name": null,
  "trial_params": null
}