{
  "best_metric": 0.08173693716526031,
  "best_model_checkpoint": "./vit-lage-beans-demo-v5/checkpoint-1200",
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 1260,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 0.00019841269841269844,
      "loss": 1.4196,
      "step": 10
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019682539682539682,
      "loss": 1.1295,
      "step": 20
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019523809523809525,
      "loss": 0.9118,
      "step": 30
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019365079365079365,
      "loss": 0.7541,
      "step": 40
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019206349206349208,
      "loss": 0.7979,
      "step": 50
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019047619047619048,
      "loss": 0.7402,
      "step": 60
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018888888888888888,
      "loss": 0.7126,
      "step": 70
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00018730158730158731,
      "loss": 0.6237,
      "step": 80
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018571428571428572,
      "loss": 0.6575,
      "step": 90
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018412698412698412,
      "loss": 0.6573,
      "step": 100
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.7861111111111111,
      "eval_loss": 0.5740312933921814,
      "eval_runtime": 45.9753,
      "eval_samples_per_second": 31.321,
      "eval_steps_per_second": 3.915,
      "step": 100
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018253968253968255,
      "loss": 0.487,
      "step": 110
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018095238095238095,
      "loss": 0.5934,
      "step": 120
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017936507936507938,
      "loss": 0.4899,
      "step": 130
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017777777777777779,
      "loss": 0.6438,
      "step": 140
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0001761904761904762,
      "loss": 0.4279,
      "step": 150
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00017460317460317462,
      "loss": 0.4603,
      "step": 160
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017301587301587302,
      "loss": 0.453,
      "step": 170
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00017142857142857143,
      "loss": 0.2507,
      "step": 180
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00016984126984126986,
      "loss": 0.5763,
      "step": 190
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00016825396825396826,
      "loss": 0.327,
      "step": 200
    },
    {
      "epoch": 0.63,
      "eval_accuracy": 0.8513888888888889,
      "eval_loss": 0.4065500795841217,
      "eval_runtime": 46.0712,
      "eval_samples_per_second": 31.256,
      "eval_steps_per_second": 3.907,
      "step": 200
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.5312,
      "step": 210
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0001650793650793651,
      "loss": 0.3813,
      "step": 220
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0001634920634920635,
      "loss": 0.2829,
      "step": 230
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016190476190476192,
      "loss": 0.3556,
      "step": 240
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016031746031746033,
      "loss": 0.5408,
      "step": 250
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015873015873015873,
      "loss": 0.3737,
      "step": 260
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015714285714285716,
      "loss": 0.2048,
      "step": 270
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00015555555555555556,
      "loss": 0.4684,
      "step": 280
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.000153968253968254,
      "loss": 0.2974,
      "step": 290
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015238095238095237,
      "loss": 0.3854,
      "step": 300
    },
    {
      "epoch": 0.95,
      "eval_accuracy": 0.8270833333333333,
      "eval_loss": 0.49766069650650024,
      "eval_runtime": 45.942,
      "eval_samples_per_second": 31.344,
      "eval_steps_per_second": 3.918,
      "step": 300
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0001507936507936508,
      "loss": 0.3011,
      "step": 310
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014920634920634923,
      "loss": 0.2746,
      "step": 320
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00014761904761904763,
      "loss": 0.2346,
      "step": 330
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014603174603174603,
      "loss": 0.1104,
      "step": 340
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014444444444444444,
      "loss": 0.1131,
      "step": 350
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014285714285714287,
      "loss": 0.149,
      "step": 360
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0001412698412698413,
      "loss": 0.1193,
      "step": 370
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013968253968253967,
      "loss": 0.1402,
      "step": 380
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0001380952380952381,
      "loss": 0.1694,
      "step": 390
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0001365079365079365,
      "loss": 0.266,
      "step": 400
    },
    {
      "epoch": 1.27,
      "eval_accuracy": 0.9034722222222222,
      "eval_loss": 0.3753337562084198,
      "eval_runtime": 47.0012,
      "eval_samples_per_second": 30.637,
      "eval_steps_per_second": 3.83,
      "step": 400
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00013492063492063494,
      "loss": 0.1028,
      "step": 410
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.1415,
      "step": 420
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013174603174603174,
      "loss": 0.0935,
      "step": 430
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00013015873015873017,
      "loss": 0.0799,
      "step": 440
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012857142857142858,
      "loss": 0.1661,
      "step": 450
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012698412698412698,
      "loss": 0.1452,
      "step": 460
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0001253968253968254,
      "loss": 0.148,
      "step": 470
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0001238095238095238,
      "loss": 0.1659,
      "step": 480
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012222222222222224,
      "loss": 0.1124,
      "step": 490
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012063492063492063,
      "loss": 0.1229,
      "step": 500
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.93125,
      "eval_loss": 0.20605972409248352,
      "eval_runtime": 46.864,
      "eval_samples_per_second": 30.727,
      "eval_steps_per_second": 3.841,
      "step": 500
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011904761904761905,
      "loss": 0.1222,
      "step": 510
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011746031746031746,
      "loss": 0.1434,
      "step": 520
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0001158730158730159,
      "loss": 0.0954,
      "step": 530
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011428571428571428,
      "loss": 0.1391,
      "step": 540
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0001126984126984127,
      "loss": 0.1533,
      "step": 550
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.0617,
      "step": 560
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010952380952380953,
      "loss": 0.1743,
      "step": 570
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010793650793650794,
      "loss": 0.071,
      "step": 580
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.00010634920634920635,
      "loss": 0.0937,
      "step": 590
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010476190476190477,
      "loss": 0.097,
      "step": 600
    },
    {
      "epoch": 1.9,
      "eval_accuracy": 0.9451388888888889,
      "eval_loss": 0.17500616610050201,
      "eval_runtime": 46.666,
      "eval_samples_per_second": 30.858,
      "eval_steps_per_second": 3.857,
      "step": 600
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010317460317460319,
      "loss": 0.166,
      "step": 610
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010158730158730159,
      "loss": 0.0641,
      "step": 620
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001,
      "loss": 0.0904,
      "step": 630
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.841269841269841e-05,
      "loss": 0.0235,
      "step": 640
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.682539682539682e-05,
      "loss": 0.0531,
      "step": 650
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.523809523809524e-05,
      "loss": 0.0062,
      "step": 660
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.365079365079366e-05,
      "loss": 0.0178,
      "step": 670
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.206349206349206e-05,
      "loss": 0.0191,
      "step": 680
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.047619047619048e-05,
      "loss": 0.0367,
      "step": 690
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.0655,
      "step": 700
    },
    {
      "epoch": 2.22,
      "eval_accuracy": 0.9652777777777778,
      "eval_loss": 0.13490158319473267,
      "eval_runtime": 46.9498,
      "eval_samples_per_second": 30.671,
      "eval_steps_per_second": 3.834,
      "step": 700
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.730158730158731e-05,
      "loss": 0.0379,
      "step": 710
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.0018,
      "step": 720
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.412698412698413e-05,
      "loss": 0.037,
      "step": 730
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.253968253968255e-05,
      "loss": 0.0137,
      "step": 740
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.095238095238096e-05,
      "loss": 0.0088,
      "step": 750
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.936507936507937e-05,
      "loss": 0.0031,
      "step": 760
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.0149,
      "step": 770
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.619047619047618e-05,
      "loss": 0.042,
      "step": 780
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.460317460317461e-05,
      "loss": 0.0029,
      "step": 790
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.301587301587302e-05,
      "loss": 0.017,
      "step": 800
    },
    {
      "epoch": 2.54,
      "eval_accuracy": 0.9673611111111111,
      "eval_loss": 0.14617331326007843,
      "eval_runtime": 46.5667,
      "eval_samples_per_second": 30.923,
      "eval_steps_per_second": 3.865,
      "step": 800
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.0438,
      "step": 810
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.984126984126984e-05,
      "loss": 0.0109,
      "step": 820
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.825396825396825e-05,
      "loss": 0.0011,
      "step": 830
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0127,
      "step": 840
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.507936507936509e-05,
      "loss": 0.0143,
      "step": 850
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.349206349206349e-05,
      "loss": 0.0011,
      "step": 860
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.19047619047619e-05,
      "loss": 0.0299,
      "step": 870
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.0317460317460316e-05,
      "loss": 0.0157,
      "step": 880
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.873015873015873e-05,
      "loss": 0.0339,
      "step": 890
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.0041,
      "step": 900
    },
    {
      "epoch": 2.86,
      "eval_accuracy": 0.9798611111111111,
      "eval_loss": 0.08337713032960892,
      "eval_runtime": 45.8768,
      "eval_samples_per_second": 31.388,
      "eval_steps_per_second": 3.924,
      "step": 900
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.0202,
      "step": 910
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.396825396825397e-05,
      "loss": 0.0046,
      "step": 920
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.2380952380952384e-05,
      "loss": 0.0016,
      "step": 930
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.0793650793650794e-05,
      "loss": 0.0005,
      "step": 940
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.9206349206349204e-05,
      "loss": 0.0006,
      "step": 950
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.761904761904762e-05,
      "loss": 0.0003,
      "step": 960
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.603174603174603e-05,
      "loss": 0.0158,
      "step": 970
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.0002,
      "step": 980
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.0007,
      "step": 990
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.126984126984127e-05,
      "loss": 0.0003,
      "step": 1000
    },
    {
      "epoch": 3.17,
      "eval_accuracy": 0.98125,
      "eval_loss": 0.08399645984172821,
      "eval_runtime": 46.3453,
      "eval_samples_per_second": 31.071,
      "eval_steps_per_second": 3.884,
      "step": 1000
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.968253968253968e-05,
      "loss": 0.0004,
      "step": 1010
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.809523809523809e-05,
      "loss": 0.0002,
      "step": 1020
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.650793650793651e-05,
      "loss": 0.0004,
      "step": 1030
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.492063492063492e-05,
      "loss": 0.0002,
      "step": 1040
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0004,
      "step": 1050
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.1746031746031745e-05,
      "loss": 0.0002,
      "step": 1060
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.0158730158730158e-05,
      "loss": 0.0002,
      "step": 1070
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.0002,
      "step": 1080
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.6984126984126984e-05,
      "loss": 0.0002,
      "step": 1090
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5396825396825397e-05,
      "loss": 0.0002,
      "step": 1100
    },
    {
      "epoch": 3.49,
      "eval_accuracy": 0.9805555555555555,
      "eval_loss": 0.08516526967287064,
      "eval_runtime": 46.626,
      "eval_samples_per_second": 30.884,
      "eval_steps_per_second": 3.861,
      "step": 1100
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.380952380952381e-05,
      "loss": 0.0004,
      "step": 1110
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.0003,
      "step": 1120
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0634920634920636e-05,
      "loss": 0.0003,
      "step": 1130
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 0.0004,
      "step": 1140
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.746031746031746e-05,
      "loss": 0.0002,
      "step": 1150
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.5873015873015872e-05,
      "loss": 0.0002,
      "step": 1160
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.0004,
      "step": 1170
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2698412698412699e-05,
      "loss": 0.0053,
      "step": 1180
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.0002,
      "step": 1190
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.523809523809523e-06,
      "loss": 0.0023,
      "step": 1200
    },
    {
      "epoch": 3.81,
      "eval_accuracy": 0.98125,
      "eval_loss": 0.08173693716526031,
      "eval_runtime": 46.7754,
      "eval_samples_per_second": 30.785,
      "eval_steps_per_second": 3.848,
      "step": 1200
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.0003,
      "step": 1210
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.349206349206349e-06,
      "loss": 0.0004,
      "step": 1220
    },
    {
      "epoch": 3.9,
      "learning_rate": 4.7619047619047615e-06,
      "loss": 0.0008,
      "step": 1230
    },
    {
      "epoch": 3.94,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 0.0002,
      "step": 1240
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 0.0004,
      "step": 1250
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.0,
      "loss": 0.0005,
      "step": 1260
    },
    {
      "epoch": 4.0,
      "step": 1260,
      "total_flos": 1.5619845389576602e+18,
      "train_loss": 0.17611938929111356,
      "train_runtime": 2152.8949,
      "train_samples_per_second": 9.362,
      "train_steps_per_second": 0.585
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9833564493758669,
      "eval_loss": 0.07597729563713074,
      "eval_runtime": 24.5791,
      "eval_samples_per_second": 29.334,
      "eval_steps_per_second": 3.702,
      "step": 1260
    }
  ],
  "logging_steps": 10,
  "max_steps": 1260,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 1.5619845389576602e+18,
  "trial_name": null,
  "trial_params": null
}