{
  "best_metric": 0.7031954526901245,
  "best_model_checkpoint": "./vit-stanford-cars/checkpoint-2500",
  "epoch": 9.803921568627452,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 5.3806,
      "step": 10
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 5.4127,
      "step": 20
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.6400000000000005e-06,
      "loss": 5.3923,
      "step": 30
    },
    {
      "epoch": 0.16,
      "learning_rate": 6.24e-06,
      "loss": 5.4098,
      "step": 40
    },
    {
      "epoch": 0.2,
      "learning_rate": 7.840000000000001e-06,
      "loss": 5.3607,
      "step": 50
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.440000000000001e-06,
      "loss": 5.3706,
      "step": 60
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.1040000000000001e-05,
      "loss": 5.3293,
      "step": 70
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.2640000000000001e-05,
      "loss": 5.3037,
      "step": 80
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.4240000000000001e-05,
      "loss": 5.2453,
      "step": 90
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.584e-05,
      "loss": 5.2552,
      "step": 100
    },
    {
      "epoch": 0.39,
      "eval_accuracy": 0.0125,
      "eval_loss": 5.2306718826293945,
      "eval_runtime": 39.7246,
      "eval_samples_per_second": 50.347,
      "eval_steps_per_second": 1.586,
      "step": 100
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.7440000000000002e-05,
      "loss": 5.1976,
      "step": 110
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.904e-05,
      "loss": 5.1826,
      "step": 120
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.0640000000000002e-05,
      "loss": 5.1389,
      "step": 130
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.2240000000000004e-05,
      "loss": 5.0934,
      "step": 140
    },
    {
      "epoch": 0.59,
      "learning_rate": 2.3840000000000002e-05,
      "loss": 5.0457,
      "step": 150
    },
    {
      "epoch": 0.63,
      "learning_rate": 2.5440000000000004e-05,
      "loss": 5.0993,
      "step": 160
    },
    {
      "epoch": 0.67,
      "learning_rate": 2.7040000000000005e-05,
      "loss": 4.9771,
      "step": 170
    },
    {
      "epoch": 0.71,
      "learning_rate": 2.864e-05,
      "loss": 4.9128,
      "step": 180
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.0240000000000002e-05,
      "loss": 4.8988,
      "step": 190
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.184000000000001e-05,
      "loss": 4.7653,
      "step": 200
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 0.079,
      "eval_loss": 4.759223461151123,
      "eval_runtime": 40.216,
      "eval_samples_per_second": 49.732,
      "eval_steps_per_second": 1.567,
      "step": 200
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.344e-05,
      "loss": 4.7304,
      "step": 210
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.504e-05,
      "loss": 4.7613,
      "step": 220
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.664e-05,
      "loss": 4.6521,
      "step": 230
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.824e-05,
      "loss": 4.6012,
      "step": 240
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.9840000000000005e-05,
      "loss": 4.4954,
      "step": 250
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.144e-05,
      "loss": 4.3733,
      "step": 260
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.304000000000001e-05,
      "loss": 4.1471,
      "step": 270
    },
    {
      "epoch": 1.1,
      "learning_rate": 4.4640000000000006e-05,
      "loss": 4.1306,
      "step": 280
    },
    {
      "epoch": 1.14,
      "learning_rate": 4.624e-05,
      "loss": 4.0396,
      "step": 290
    },
    {
      "epoch": 1.18,
      "learning_rate": 4.784e-05,
      "loss": 3.9779,
      "step": 300
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.193,
      "eval_loss": 4.070125579833984,
      "eval_runtime": 40.3837,
      "eval_samples_per_second": 49.525,
      "eval_steps_per_second": 1.56,
      "step": 300
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.944e-05,
      "loss": 3.8729,
      "step": 310
    },
    {
      "epoch": 1.25,
      "learning_rate": 5.1040000000000006e-05,
      "loss": 3.7538,
      "step": 320
    },
    {
      "epoch": 1.29,
      "learning_rate": 5.2640000000000004e-05,
      "loss": 3.6952,
      "step": 330
    },
    {
      "epoch": 1.33,
      "learning_rate": 5.424000000000001e-05,
      "loss": 3.6388,
      "step": 340
    },
    {
      "epoch": 1.37,
      "learning_rate": 5.584e-05,
      "loss": 3.6777,
      "step": 350
    },
    {
      "epoch": 1.41,
      "learning_rate": 5.7440000000000006e-05,
      "loss": 3.5145,
      "step": 360
    },
    {
      "epoch": 1.45,
      "learning_rate": 5.9040000000000004e-05,
      "loss": 3.5315,
      "step": 370
    },
    {
      "epoch": 1.49,
      "learning_rate": 6.064e-05,
      "loss": 3.4288,
      "step": 380
    },
    {
      "epoch": 1.53,
      "learning_rate": 6.224e-05,
      "loss": 3.3541,
      "step": 390
    },
    {
      "epoch": 1.57,
      "learning_rate": 6.384000000000001e-05,
      "loss": 3.2754,
      "step": 400
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 0.3305,
      "eval_loss": 3.3328330516815186,
      "eval_runtime": 40.7474,
      "eval_samples_per_second": 49.083,
      "eval_steps_per_second": 1.546,
      "step": 400
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.544e-05,
      "loss": 3.1604,
      "step": 410
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.704000000000001e-05,
      "loss": 3.1043,
      "step": 420
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.864000000000001e-05,
      "loss": 2.9978,
      "step": 430
    },
    {
      "epoch": 1.73,
      "learning_rate": 7.024e-05,
      "loss": 2.9723,
      "step": 440
    },
    {
      "epoch": 1.76,
      "learning_rate": 7.184e-05,
      "loss": 2.9747,
      "step": 450
    },
    {
      "epoch": 1.8,
      "learning_rate": 7.344000000000002e-05,
      "loss": 2.882,
      "step": 460
    },
    {
      "epoch": 1.84,
      "learning_rate": 7.504e-05,
      "loss": 2.7142,
      "step": 470
    },
    {
      "epoch": 1.88,
      "learning_rate": 7.664e-05,
      "loss": 2.7157,
      "step": 480
    },
    {
      "epoch": 1.92,
      "learning_rate": 7.824000000000001e-05,
      "loss": 2.6695,
      "step": 490
    },
    {
      "epoch": 1.96,
      "learning_rate": 7.984000000000001e-05,
      "loss": 2.5607,
      "step": 500
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.491,
      "eval_loss": 2.6118950843811035,
      "eval_runtime": 40.1814,
      "eval_samples_per_second": 49.774,
      "eval_steps_per_second": 1.568,
      "step": 500
    },
    {
      "epoch": 2.0,
      "learning_rate": 7.964878048780489e-05,
      "loss": 2.6025,
      "step": 510
    },
    {
      "epoch": 2.04,
      "learning_rate": 7.925853658536585e-05,
      "loss": 1.9918,
      "step": 520
    },
    {
      "epoch": 2.08,
      "learning_rate": 7.886829268292684e-05,
      "loss": 1.8799,
      "step": 530
    },
    {
      "epoch": 2.12,
      "learning_rate": 7.84780487804878e-05,
      "loss": 1.7169,
      "step": 540
    },
    {
      "epoch": 2.16,
      "learning_rate": 7.808780487804879e-05,
      "loss": 1.7148,
      "step": 550
    },
    {
      "epoch": 2.2,
      "learning_rate": 7.769756097560977e-05,
      "loss": 1.7091,
      "step": 560
    },
    {
      "epoch": 2.24,
      "learning_rate": 7.730731707317074e-05,
      "loss": 1.569,
      "step": 570
    },
    {
      "epoch": 2.27,
      "learning_rate": 7.691707317073172e-05,
      "loss": 1.5601,
      "step": 580
    },
    {
      "epoch": 2.31,
      "learning_rate": 7.652682926829269e-05,
      "loss": 1.5463,
      "step": 590
    },
    {
      "epoch": 2.35,
      "learning_rate": 7.613658536585367e-05,
      "loss": 1.5249,
      "step": 600
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.61,
      "eval_loss": 2.0372462272644043,
      "eval_runtime": 40.501,
      "eval_samples_per_second": 49.381,
      "eval_steps_per_second": 1.556,
      "step": 600
    },
    {
      "epoch": 2.39,
      "learning_rate": 7.574634146341463e-05,
      "loss": 1.4028,
      "step": 610
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.535609756097562e-05,
      "loss": 1.3612,
      "step": 620
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.49658536585366e-05,
      "loss": 1.3848,
      "step": 630
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.457560975609757e-05,
      "loss": 1.3545,
      "step": 640
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.418536585365855e-05,
      "loss": 1.2483,
      "step": 650
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.379512195121952e-05,
      "loss": 1.2979,
      "step": 660
    },
    {
      "epoch": 2.63,
      "learning_rate": 7.34048780487805e-05,
      "loss": 1.2087,
      "step": 670
    },
    {
      "epoch": 2.67,
      "learning_rate": 7.301463414634147e-05,
      "loss": 1.2661,
      "step": 680
    },
    {
      "epoch": 2.71,
      "learning_rate": 7.262439024390245e-05,
      "loss": 1.2181,
      "step": 690
    },
    {
      "epoch": 2.75,
      "learning_rate": 7.223414634146343e-05,
      "loss": 1.1915,
      "step": 700
    },
    {
      "epoch": 2.75,
      "eval_accuracy": 0.663,
      "eval_loss": 1.6082996129989624,
      "eval_runtime": 39.9953,
      "eval_samples_per_second": 50.006,
      "eval_steps_per_second": 1.575,
      "step": 700
    },
    {
      "epoch": 2.78,
      "learning_rate": 7.18439024390244e-05,
      "loss": 1.1685,
      "step": 710
    },
    {
      "epoch": 2.82,
      "learning_rate": 7.145365853658538e-05,
      "loss": 1.0873,
      "step": 720
    },
    {
      "epoch": 2.86,
      "learning_rate": 7.106341463414635e-05,
      "loss": 1.1013,
      "step": 730
    },
    {
      "epoch": 2.9,
      "learning_rate": 7.067317073170733e-05,
      "loss": 1.1133,
      "step": 740
    },
    {
      "epoch": 2.94,
      "learning_rate": 7.02829268292683e-05,
      "loss": 1.0955,
      "step": 750
    },
    {
      "epoch": 2.98,
      "learning_rate": 6.989268292682928e-05,
      "loss": 0.9469,
      "step": 760
    },
    {
      "epoch": 3.02,
      "learning_rate": 6.950243902439025e-05,
      "loss": 0.7376,
      "step": 770
    },
    {
      "epoch": 3.06,
      "learning_rate": 6.911219512195123e-05,
      "loss": 0.4818,
      "step": 780
    },
    {
      "epoch": 3.1,
      "learning_rate": 6.87219512195122e-05,
      "loss": 0.4825,
      "step": 790
    },
    {
      "epoch": 3.14,
      "learning_rate": 6.833170731707318e-05,
      "loss": 0.4414,
      "step": 800
    },
    {
      "epoch": 3.14,
      "eval_accuracy": 0.725,
      "eval_loss": 1.320522665977478,
      "eval_runtime": 40.1363,
      "eval_samples_per_second": 49.83,
      "eval_steps_per_second": 1.57,
      "step": 800
    },
    {
      "epoch": 3.18,
      "learning_rate": 6.794146341463415e-05,
      "loss": 0.4289,
      "step": 810
    },
    {
      "epoch": 3.22,
      "learning_rate": 6.755121951219513e-05,
      "loss": 0.3919,
      "step": 820
    },
    {
      "epoch": 3.25,
      "learning_rate": 6.71609756097561e-05,
      "loss": 0.3715,
      "step": 830
    },
    {
      "epoch": 3.29,
      "learning_rate": 6.677073170731708e-05,
      "loss": 0.3895,
      "step": 840
    },
    {
      "epoch": 3.33,
      "learning_rate": 6.638048780487805e-05,
      "loss": 0.3804,
      "step": 850
    },
    {
      "epoch": 3.37,
      "learning_rate": 6.599024390243903e-05,
      "loss": 0.3973,
      "step": 860
    },
    {
      "epoch": 3.41,
      "learning_rate": 6.56e-05,
      "loss": 0.3987,
      "step": 870
    },
    {
      "epoch": 3.45,
      "learning_rate": 6.520975609756098e-05,
      "loss": 0.3903,
      "step": 880
    },
    {
      "epoch": 3.49,
      "learning_rate": 6.481951219512196e-05,
      "loss": 0.3469,
      "step": 890
    },
    {
      "epoch": 3.53,
      "learning_rate": 6.442926829268294e-05,
      "loss": 0.323,
      "step": 900
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.7615,
      "eval_loss": 1.1008645296096802,
      "eval_runtime": 41.4264,
      "eval_samples_per_second": 48.278,
      "eval_steps_per_second": 1.521,
      "step": 900
    },
    {
      "epoch": 3.57,
      "learning_rate": 6.403902439024391e-05,
      "loss": 0.3168,
      "step": 910
    },
    {
      "epoch": 3.61,
      "learning_rate": 6.364878048780489e-05,
      "loss": 0.3404,
      "step": 920
    },
    {
      "epoch": 3.65,
      "learning_rate": 6.325853658536586e-05,
      "loss": 0.3553,
      "step": 930
    },
    {
      "epoch": 3.69,
      "learning_rate": 6.286829268292684e-05,
      "loss": 0.3665,
      "step": 940
    },
    {
      "epoch": 3.73,
      "learning_rate": 6.247804878048781e-05,
      "loss": 0.2983,
      "step": 950
    },
    {
      "epoch": 3.76,
      "learning_rate": 6.208780487804879e-05,
      "loss": 0.2886,
      "step": 960
    },
    {
      "epoch": 3.8,
      "learning_rate": 6.169756097560976e-05,
      "loss": 0.2721,
      "step": 970
    },
    {
      "epoch": 3.84,
      "learning_rate": 6.130731707317074e-05,
      "loss": 0.2898,
      "step": 980
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.0917073170731714e-05,
      "loss": 0.3222,
      "step": 990
    },
    {
      "epoch": 3.92,
      "learning_rate": 6.052682926829269e-05,
      "loss": 0.2993,
      "step": 1000
    },
    {
      "epoch": 3.92,
      "eval_accuracy": 0.766,
      "eval_loss": 1.0115569829940796,
      "eval_runtime": 40.1299,
      "eval_samples_per_second": 49.838,
      "eval_steps_per_second": 1.57,
      "step": 1000
    },
    {
      "epoch": 3.96,
      "learning_rate": 6.0136585365853664e-05,
      "loss": 0.2903,
      "step": 1010
    },
    {
      "epoch": 4.0,
      "learning_rate": 5.974634146341464e-05,
      "loss": 0.2707,
      "step": 1020
    },
    {
      "epoch": 4.04,
      "learning_rate": 5.9356097560975614e-05,
      "loss": 0.0953,
      "step": 1030
    },
    {
      "epoch": 4.08,
      "learning_rate": 5.896585365853659e-05,
      "loss": 0.1066,
      "step": 1040
    },
    {
      "epoch": 4.12,
      "learning_rate": 5.8575609756097564e-05,
      "loss": 0.0944,
      "step": 1050
    },
    {
      "epoch": 4.16,
      "learning_rate": 5.818536585365854e-05,
      "loss": 0.0943,
      "step": 1060
    },
    {
      "epoch": 4.2,
      "learning_rate": 5.779512195121951e-05,
      "loss": 0.0985,
      "step": 1070
    },
    {
      "epoch": 4.24,
      "learning_rate": 5.740487804878049e-05,
      "loss": 0.1034,
      "step": 1080
    },
    {
      "epoch": 4.27,
      "learning_rate": 5.701463414634146e-05,
      "loss": 0.1015,
      "step": 1090
    },
    {
      "epoch": 4.31,
      "learning_rate": 5.662439024390244e-05,
      "loss": 0.1024,
      "step": 1100
    },
    {
      "epoch": 4.31,
      "eval_accuracy": 0.7895,
      "eval_loss": 0.9027227759361267,
      "eval_runtime": 40.6004,
      "eval_samples_per_second": 49.261,
      "eval_steps_per_second": 1.552,
      "step": 1100
    },
    {
      "epoch": 4.35,
      "learning_rate": 5.623414634146341e-05,
      "loss": 0.0779,
      "step": 1110
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.584390243902439e-05,
      "loss": 0.0754,
      "step": 1120
    },
    {
      "epoch": 4.43,
      "learning_rate": 5.545365853658537e-05,
      "loss": 0.0754,
      "step": 1130
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.5063414634146344e-05,
      "loss": 0.0724,
      "step": 1140
    },
    {
      "epoch": 4.51,
      "learning_rate": 5.4673170731707326e-05,
      "loss": 0.0765,
      "step": 1150
    },
    {
      "epoch": 4.55,
      "learning_rate": 5.42829268292683e-05,
      "loss": 0.0854,
      "step": 1160
    },
    {
      "epoch": 4.59,
      "learning_rate": 5.3892682926829276e-05,
      "loss": 0.0619,
      "step": 1170
    },
    {
      "epoch": 4.63,
      "learning_rate": 5.350243902439025e-05,
      "loss": 0.0679,
      "step": 1180
    },
    {
      "epoch": 4.67,
      "learning_rate": 5.3112195121951225e-05,
      "loss": 0.0836,
      "step": 1190
    },
    {
      "epoch": 4.71,
      "learning_rate": 5.27219512195122e-05,
      "loss": 0.097,
      "step": 1200
    },
    {
      "epoch": 4.71,
      "eval_accuracy": 0.7825,
      "eval_loss": 0.8696740865707397,
      "eval_runtime": 39.9428,
      "eval_samples_per_second": 50.072,
      "eval_steps_per_second": 1.577,
      "step": 1200
    },
    {
      "epoch": 4.75,
      "learning_rate": 5.2331707317073175e-05,
      "loss": 0.0763,
      "step": 1210
    },
    {
      "epoch": 4.78,
      "learning_rate": 5.194146341463415e-05,
      "loss": 0.0863,
      "step": 1220
    },
    {
      "epoch": 4.82,
      "learning_rate": 5.1551219512195125e-05,
      "loss": 0.0827,
      "step": 1230
    },
    {
      "epoch": 4.86,
      "learning_rate": 5.11609756097561e-05,
      "loss": 0.0929,
      "step": 1240
    },
    {
      "epoch": 4.9,
      "learning_rate": 5.0770731707317075e-05,
      "loss": 0.0619,
      "step": 1250
    },
    {
      "epoch": 4.94,
      "learning_rate": 5.0380487804878056e-05,
      "loss": 0.0827,
      "step": 1260
    },
    {
      "epoch": 4.98,
      "learning_rate": 4.999024390243903e-05,
      "loss": 0.0694,
      "step": 1270
    },
    {
      "epoch": 5.02,
      "learning_rate": 4.9600000000000006e-05,
      "loss": 0.0603,
      "step": 1280
    },
    {
      "epoch": 5.06,
      "learning_rate": 4.920975609756098e-05,
      "loss": 0.0233,
      "step": 1290
    },
    {
      "epoch": 5.1,
      "learning_rate": 4.8819512195121956e-05,
      "loss": 0.0273,
      "step": 1300
    },
    {
      "epoch": 5.1,
      "eval_accuracy": 0.805,
      "eval_loss": 0.7953792810440063,
      "eval_runtime": 39.9571,
      "eval_samples_per_second": 50.054,
      "eval_steps_per_second": 1.577,
      "step": 1300
    },
    {
      "epoch": 5.14,
      "learning_rate": 4.842926829268293e-05,
      "loss": 0.0197,
      "step": 1310
    },
    {
      "epoch": 5.18,
      "learning_rate": 4.8039024390243906e-05,
      "loss": 0.0194,
      "step": 1320
    },
    {
      "epoch": 5.22,
      "learning_rate": 4.764878048780488e-05,
      "loss": 0.0326,
      "step": 1330
    },
    {
      "epoch": 5.25,
      "learning_rate": 4.7258536585365856e-05,
      "loss": 0.0367,
      "step": 1340
    },
    {
      "epoch": 5.29,
      "learning_rate": 4.686829268292683e-05,
      "loss": 0.0172,
      "step": 1350
    },
    {
      "epoch": 5.33,
      "learning_rate": 4.6478048780487805e-05,
      "loss": 0.0259,
      "step": 1360
    },
    {
      "epoch": 5.37,
      "learning_rate": 4.608780487804878e-05,
      "loss": 0.0176,
      "step": 1370
    },
    {
      "epoch": 5.41,
      "learning_rate": 4.5697560975609755e-05,
      "loss": 0.0233,
      "step": 1380
    },
    {
      "epoch": 5.45,
      "learning_rate": 4.530731707317073e-05,
      "loss": 0.0393,
      "step": 1390
    },
    {
      "epoch": 5.49,
      "learning_rate": 4.491707317073172e-05,
      "loss": 0.0386,
      "step": 1400
    },
    {
      "epoch": 5.49,
      "eval_accuracy": 0.8095,
      "eval_loss": 0.782217264175415,
      "eval_runtime": 40.3652,
      "eval_samples_per_second": 49.548,
      "eval_steps_per_second": 1.561,
      "step": 1400
    },
    {
      "epoch": 5.53,
      "learning_rate": 4.452682926829269e-05,
      "loss": 0.0222,
      "step": 1410
    },
    {
      "epoch": 5.57,
      "learning_rate": 4.413658536585367e-05,
      "loss": 0.041,
      "step": 1420
    },
    {
      "epoch": 5.61,
      "learning_rate": 4.374634146341464e-05,
      "loss": 0.0188,
      "step": 1430
    },
    {
      "epoch": 5.65,
      "learning_rate": 4.335609756097562e-05,
      "loss": 0.032,
      "step": 1440
    },
    {
      "epoch": 5.69,
      "learning_rate": 4.296585365853659e-05,
      "loss": 0.0159,
      "step": 1450
    },
    {
      "epoch": 5.73,
      "learning_rate": 4.257560975609757e-05,
      "loss": 0.0433,
      "step": 1460
    },
    {
      "epoch": 5.76,
      "learning_rate": 4.218536585365854e-05,
      "loss": 0.0174,
      "step": 1470
    },
    {
      "epoch": 5.8,
      "learning_rate": 4.179512195121952e-05,
      "loss": 0.0307,
      "step": 1480
    },
    {
      "epoch": 5.84,
      "learning_rate": 4.140487804878049e-05,
      "loss": 0.0221,
      "step": 1490
    },
    {
      "epoch": 5.88,
      "learning_rate": 4.101463414634147e-05,
      "loss": 0.0304,
      "step": 1500
    },
    {
      "epoch": 5.88,
      "eval_accuracy": 0.8025,
      "eval_loss": 0.7874125242233276,
      "eval_runtime": 40.5243,
      "eval_samples_per_second": 49.353,
      "eval_steps_per_second": 1.555,
      "step": 1500
    },
    {
      "epoch": 5.92,
      "learning_rate": 4.062439024390244e-05,
      "loss": 0.0429,
      "step": 1510
    },
    {
      "epoch": 5.96,
      "learning_rate": 4.023414634146342e-05,
      "loss": 0.0244,
      "step": 1520
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.984390243902439e-05,
      "loss": 0.0368,
      "step": 1530
    },
    {
      "epoch": 6.04,
      "learning_rate": 3.945365853658537e-05,
      "loss": 0.0104,
      "step": 1540
    },
    {
      "epoch": 6.08,
      "learning_rate": 3.906341463414634e-05,
      "loss": 0.0105,
      "step": 1550
    },
    {
      "epoch": 6.12,
      "learning_rate": 3.8673170731707317e-05,
      "loss": 0.0164,
      "step": 1560
    },
    {
      "epoch": 6.16,
      "learning_rate": 3.828292682926829e-05,
      "loss": 0.009,
      "step": 1570
    },
    {
      "epoch": 6.2,
      "learning_rate": 3.7892682926829266e-05,
      "loss": 0.0101,
      "step": 1580
    },
    {
      "epoch": 6.24,
      "learning_rate": 3.750243902439025e-05,
      "loss": 0.0184,
      "step": 1590
    },
    {
      "epoch": 6.27,
      "learning_rate": 3.711219512195122e-05,
      "loss": 0.0163,
      "step": 1600
    },
    {
      "epoch": 6.27,
      "eval_accuracy": 0.816,
      "eval_loss": 0.7488501071929932,
      "eval_runtime": 40.3879,
      "eval_samples_per_second": 49.52,
      "eval_steps_per_second": 1.56,
      "step": 1600
    },
    {
      "epoch": 6.31,
      "learning_rate": 3.67219512195122e-05,
      "loss": 0.0118,
      "step": 1610
    },
    {
      "epoch": 6.35,
      "learning_rate": 3.633170731707317e-05,
      "loss": 0.0082,
      "step": 1620
    },
    {
      "epoch": 6.39,
      "learning_rate": 3.594146341463415e-05,
      "loss": 0.011,
      "step": 1630
    },
    {
      "epoch": 6.43,
      "learning_rate": 3.555121951219512e-05,
      "loss": 0.0088,
      "step": 1640
    },
    {
      "epoch": 6.47,
      "learning_rate": 3.51609756097561e-05,
      "loss": 0.0403,
      "step": 1650
    },
    {
      "epoch": 6.51,
      "learning_rate": 3.477073170731708e-05,
      "loss": 0.0089,
      "step": 1660
    },
    {
      "epoch": 6.55,
      "learning_rate": 3.4380487804878054e-05,
      "loss": 0.0203,
      "step": 1670
    },
    {
      "epoch": 6.59,
      "learning_rate": 3.399024390243903e-05,
      "loss": 0.0157,
      "step": 1680
    },
    {
      "epoch": 6.63,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 0.0145,
      "step": 1690
    },
    {
      "epoch": 6.67,
      "learning_rate": 3.320975609756098e-05,
      "loss": 0.0171,
      "step": 1700
    },
    {
      "epoch": 6.67,
      "eval_accuracy": 0.816,
      "eval_loss": 0.7466872930526733,
      "eval_runtime": 40.0569,
      "eval_samples_per_second": 49.929,
      "eval_steps_per_second": 1.573,
      "step": 1700
    },
    {
      "epoch": 6.71,
      "learning_rate": 3.281951219512195e-05,
      "loss": 0.0115,
      "step": 1710
    },
    {
      "epoch": 6.75,
      "learning_rate": 3.2429268292682935e-05,
      "loss": 0.0207,
      "step": 1720
    },
    {
      "epoch": 6.78,
      "learning_rate": 3.203902439024391e-05,
      "loss": 0.038,
      "step": 1730
    },
    {
      "epoch": 6.82,
      "learning_rate": 3.1648780487804885e-05,
      "loss": 0.0289,
      "step": 1740
    },
    {
      "epoch": 6.86,
      "learning_rate": 3.125853658536586e-05,
      "loss": 0.0261,
      "step": 1750
    },
    {
      "epoch": 6.9,
      "learning_rate": 3.0868292682926835e-05,
      "loss": 0.0191,
      "step": 1760
    },
    {
      "epoch": 6.94,
      "learning_rate": 3.047804878048781e-05,
      "loss": 0.008,
      "step": 1770
    },
    {
      "epoch": 6.98,
      "learning_rate": 3.0087804878048784e-05,
      "loss": 0.0076,
      "step": 1780
    },
    {
      "epoch": 7.02,
      "learning_rate": 2.969756097560976e-05,
      "loss": 0.0085,
      "step": 1790
    },
    {
      "epoch": 7.06,
      "learning_rate": 2.9307317073170734e-05,
      "loss": 0.0068,
      "step": 1800
    },
    {
      "epoch": 7.06,
      "eval_accuracy": 0.8155,
      "eval_loss": 0.7402506470680237,
      "eval_runtime": 40.0633,
      "eval_samples_per_second": 49.921,
      "eval_steps_per_second": 1.573,
      "step": 1800
    },
    {
      "epoch": 7.1,
      "learning_rate": 2.891707317073171e-05,
      "loss": 0.0132,
      "step": 1810
    },
    {
      "epoch": 7.14,
      "learning_rate": 2.8526829268292684e-05,
      "loss": 0.0136,
      "step": 1820
    },
    {
      "epoch": 7.18,
      "learning_rate": 2.813658536585366e-05,
      "loss": 0.0107,
      "step": 1830
    },
    {
      "epoch": 7.22,
      "learning_rate": 2.7746341463414634e-05,
      "loss": 0.0069,
      "step": 1840
    },
    {
      "epoch": 7.25,
      "learning_rate": 2.7356097560975615e-05,
      "loss": 0.0152,
      "step": 1850
    },
    {
      "epoch": 7.29,
      "learning_rate": 2.696585365853659e-05,
      "loss": 0.0135,
      "step": 1860
    },
    {
      "epoch": 7.33,
      "learning_rate": 2.6575609756097565e-05,
      "loss": 0.0134,
      "step": 1870
    },
    {
      "epoch": 7.37,
      "learning_rate": 2.618536585365854e-05,
      "loss": 0.0174,
      "step": 1880
    },
    {
      "epoch": 7.41,
      "learning_rate": 2.5795121951219515e-05,
      "loss": 0.0062,
      "step": 1890
    },
    {
      "epoch": 7.45,
      "learning_rate": 2.540487804878049e-05,
      "loss": 0.0064,
      "step": 1900
    },
    {
      "epoch": 7.45,
      "eval_accuracy": 0.8225,
      "eval_loss": 0.7162501215934753,
      "eval_runtime": 40.5713,
      "eval_samples_per_second": 49.296,
      "eval_steps_per_second": 1.553,
      "step": 1900
    },
    {
      "epoch": 7.49,
      "learning_rate": 2.5014634146341465e-05,
      "loss": 0.0058,
      "step": 1910
    },
    {
      "epoch": 7.53,
      "learning_rate": 2.462439024390244e-05,
      "loss": 0.0056,
      "step": 1920
    },
    {
      "epoch": 7.57,
      "learning_rate": 2.4234146341463414e-05,
      "loss": 0.006,
      "step": 1930
    },
    {
      "epoch": 7.61,
      "learning_rate": 2.3843902439024393e-05,
      "loss": 0.0175,
      "step": 1940
    },
    {
      "epoch": 7.65,
      "learning_rate": 2.3453658536585367e-05,
      "loss": 0.0062,
      "step": 1950
    },
    {
      "epoch": 7.69,
      "learning_rate": 2.3063414634146342e-05,
      "loss": 0.0062,
      "step": 1960
    },
    {
      "epoch": 7.73,
      "learning_rate": 2.2673170731707317e-05,
      "loss": 0.0166,
      "step": 1970
    },
    {
      "epoch": 7.76,
      "learning_rate": 2.2282926829268296e-05,
      "loss": 0.0121,
      "step": 1980
    },
    {
      "epoch": 7.8,
      "learning_rate": 2.189268292682927e-05,
      "loss": 0.0119,
      "step": 1990
    },
    {
      "epoch": 7.84,
      "learning_rate": 2.150243902439025e-05,
      "loss": 0.0236,
      "step": 2000
    },
    {
      "epoch": 7.84,
      "eval_accuracy": 0.824,
      "eval_loss": 0.7153184413909912,
      "eval_runtime": 40.1832,
      "eval_samples_per_second": 49.772,
      "eval_steps_per_second": 1.568,
      "step": 2000
    },
    {
      "epoch": 7.88,
      "learning_rate": 2.1112195121951224e-05,
      "loss": 0.0141,
      "step": 2010
    },
    {
      "epoch": 7.92,
      "learning_rate": 2.07219512195122e-05,
      "loss": 0.016,
      "step": 2020
    },
    {
      "epoch": 7.96,
      "learning_rate": 2.0331707317073173e-05,
      "loss": 0.0173,
      "step": 2030
    },
    {
      "epoch": 8.0,
      "learning_rate": 1.9941463414634148e-05,
      "loss": 0.0058,
      "step": 2040
    },
    {
      "epoch": 8.04,
      "learning_rate": 1.9551219512195123e-05,
      "loss": 0.007,
      "step": 2050
    },
    {
      "epoch": 8.08,
      "learning_rate": 1.9160975609756098e-05,
      "loss": 0.007,
      "step": 2060
    },
    {
      "epoch": 8.12,
      "learning_rate": 1.8770731707317073e-05,
      "loss": 0.0052,
      "step": 2070
    },
    {
      "epoch": 8.16,
      "learning_rate": 1.838048780487805e-05,
      "loss": 0.0059,
      "step": 2080
    },
    {
      "epoch": 8.2,
      "learning_rate": 1.7990243902439026e-05,
      "loss": 0.0051,
      "step": 2090
    },
    {
      "epoch": 8.24,
      "learning_rate": 1.76e-05,
      "loss": 0.0075,
      "step": 2100
    },
    {
      "epoch": 8.24,
      "eval_accuracy": 0.8255,
      "eval_loss": 0.7127741575241089,
      "eval_runtime": 40.1888,
      "eval_samples_per_second": 49.765,
      "eval_steps_per_second": 1.568,
      "step": 2100
    },
    {
      "epoch": 8.27,
      "learning_rate": 1.7209756097560976e-05,
      "loss": 0.0049,
      "step": 2110
    },
    {
      "epoch": 8.31,
      "learning_rate": 1.681951219512195e-05,
      "loss": 0.0087,
      "step": 2120
    },
    {
      "epoch": 8.35,
      "learning_rate": 1.6429268292682926e-05,
      "loss": 0.0065,
      "step": 2130
    },
    {
      "epoch": 8.39,
      "learning_rate": 1.6039024390243904e-05,
      "loss": 0.0053,
      "step": 2140
    },
    {
      "epoch": 8.43,
      "learning_rate": 1.564878048780488e-05,
      "loss": 0.0133,
      "step": 2150
    },
    {
      "epoch": 8.47,
      "learning_rate": 1.5258536585365855e-05,
      "loss": 0.0186,
      "step": 2160
    },
    {
      "epoch": 8.51,
      "learning_rate": 1.486829268292683e-05,
      "loss": 0.0178,
      "step": 2170
    },
    {
      "epoch": 8.55,
      "learning_rate": 1.4478048780487805e-05,
      "loss": 0.0051,
      "step": 2180
    },
    {
      "epoch": 8.59,
      "learning_rate": 1.4087804878048782e-05,
      "loss": 0.0051,
      "step": 2190
    },
    {
      "epoch": 8.63,
      "learning_rate": 1.3697560975609758e-05,
      "loss": 0.0154,
      "step": 2200
    },
    {
      "epoch": 8.63,
      "eval_accuracy": 0.8235,
      "eval_loss": 0.7104279398918152,
      "eval_runtime": 39.6742,
      "eval_samples_per_second": 50.411,
      "eval_steps_per_second": 1.588,
      "step": 2200
    },
    {
      "epoch": 8.67,
      "learning_rate": 1.3307317073170733e-05,
      "loss": 0.0051,
      "step": 2210
    },
    {
      "epoch": 8.71,
      "learning_rate": 1.2917073170731708e-05,
      "loss": 0.0097,
      "step": 2220
    },
    {
      "epoch": 8.75,
      "learning_rate": 1.2526829268292685e-05,
      "loss": 0.0213,
      "step": 2230
    },
    {
      "epoch": 8.78,
      "learning_rate": 1.213658536585366e-05,
      "loss": 0.0091,
      "step": 2240
    },
    {
      "epoch": 8.82,
      "learning_rate": 1.1746341463414634e-05,
      "loss": 0.0087,
      "step": 2250
    },
    {
      "epoch": 8.86,
      "learning_rate": 1.135609756097561e-05,
      "loss": 0.005,
      "step": 2260
    },
    {
      "epoch": 8.9,
      "learning_rate": 1.0965853658536587e-05,
      "loss": 0.0119,
      "step": 2270
    },
    {
      "epoch": 8.94,
      "learning_rate": 1.0575609756097562e-05,
      "loss": 0.0049,
      "step": 2280
    },
    {
      "epoch": 8.98,
      "learning_rate": 1.0185365853658537e-05,
      "loss": 0.0152,
      "step": 2290
    },
    {
      "epoch": 9.02,
      "learning_rate": 9.795121951219512e-06,
      "loss": 0.0082,
      "step": 2300
    },
    {
      "epoch": 9.02,
      "eval_accuracy": 0.824,
      "eval_loss": 0.7048487067222595,
      "eval_runtime": 39.8204,
      "eval_samples_per_second": 50.226,
      "eval_steps_per_second": 1.582,
      "step": 2300
    },
    {
      "epoch": 9.06,
      "learning_rate": 9.404878048780489e-06,
      "loss": 0.0046,
      "step": 2310
    },
    {
      "epoch": 9.1,
      "learning_rate": 9.014634146341464e-06,
      "loss": 0.0076,
      "step": 2320
    },
    {
      "epoch": 9.14,
      "learning_rate": 8.62439024390244e-06,
      "loss": 0.0107,
      "step": 2330
    },
    {
      "epoch": 9.18,
      "learning_rate": 8.234146341463415e-06,
      "loss": 0.0182,
      "step": 2340
    },
    {
      "epoch": 9.22,
      "learning_rate": 7.843902439024392e-06,
      "loss": 0.0045,
      "step": 2350
    },
    {
      "epoch": 9.25,
      "learning_rate": 7.4536585365853665e-06,
      "loss": 0.0047,
      "step": 2360
    },
    {
      "epoch": 9.29,
      "learning_rate": 7.063414634146341e-06,
      "loss": 0.0047,
      "step": 2370
    },
    {
      "epoch": 9.33,
      "learning_rate": 6.673170731707318e-06,
      "loss": 0.0045,
      "step": 2380
    },
    {
      "epoch": 9.37,
      "learning_rate": 6.282926829268293e-06,
      "loss": 0.0162,
      "step": 2390
    },
    {
      "epoch": 9.41,
      "learning_rate": 5.892682926829269e-06,
      "loss": 0.007,
      "step": 2400
    },
    {
      "epoch": 9.41,
      "eval_accuracy": 0.823,
      "eval_loss": 0.7026473879814148,
      "eval_runtime": 40.761,
      "eval_samples_per_second": 49.067,
      "eval_steps_per_second": 1.546,
      "step": 2400
    },
    {
      "epoch": 9.45,
      "learning_rate": 5.502439024390245e-06,
      "loss": 0.0074,
      "step": 2410
    },
    {
      "epoch": 9.49,
      "learning_rate": 5.11219512195122e-06,
      "loss": 0.0045,
      "step": 2420
    },
    {
      "epoch": 9.53,
      "learning_rate": 4.721951219512195e-06,
      "loss": 0.0044,
      "step": 2430
    },
    {
      "epoch": 9.57,
      "learning_rate": 4.331707317073171e-06,
      "loss": 0.0046,
      "step": 2440
    },
    {
      "epoch": 9.61,
      "learning_rate": 3.941463414634147e-06,
      "loss": 0.0066,
      "step": 2450
    },
    {
      "epoch": 9.65,
      "learning_rate": 3.551219512195122e-06,
      "loss": 0.009,
      "step": 2460
    },
    {
      "epoch": 9.69,
      "learning_rate": 3.160975609756098e-06,
      "loss": 0.0045,
      "step": 2470
    },
    {
      "epoch": 9.73,
      "learning_rate": 2.7707317073170736e-06,
      "loss": 0.0046,
      "step": 2480
    },
    {
      "epoch": 9.76,
      "learning_rate": 2.380487804878049e-06,
      "loss": 0.0147,
      "step": 2490
    },
    {
      "epoch": 9.8,
      "learning_rate": 1.9902439024390246e-06,
      "loss": 0.0045,
      "step": 2500
    },
    {
      "epoch": 9.8,
      "eval_accuracy": 0.8235,
      "eval_loss": 0.7031954526901245,
      "eval_runtime": 39.643,
      "eval_samples_per_second": 50.45,
      "eval_steps_per_second": 1.589,
      "step": 2500
    }
  ],
  "max_steps": 2550,
  "num_train_epochs": 10,
  "total_flos": 6.198960121799639e+18,
  "trial_name": null,
  "trial_params": null
}