{
  "best_metric": 0.08173014223575592,
  "best_model_checkpoint": "./vit-lage-beans-demo-v5/checkpoint-1200",
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 1440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 0.00019861111111111113,
      "loss": 1.4551,
      "step": 10
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019722222222222225,
      "loss": 1.148,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019583333333333334,
      "loss": 0.9667,
      "step": 30
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00019444444444444446,
      "loss": 0.7212,
      "step": 40
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019305555555555558,
      "loss": 1.0176,
      "step": 50
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019166666666666667,
      "loss": 0.6942,
      "step": 60
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019027777777777776,
      "loss": 0.8307,
      "step": 70
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018888888888888888,
      "loss": 0.5401,
      "step": 80
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0001875,
      "loss": 0.7918,
      "step": 90
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018611111111111112,
      "loss": 0.7861,
      "step": 100
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.725,
      "eval_loss": 0.7609096765518188,
      "eval_runtime": 45.8495,
      "eval_samples_per_second": 15.704,
      "eval_steps_per_second": 1.963,
      "step": 100
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018472222222222224,
      "loss": 0.6622,
      "step": 110
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00018333333333333334,
      "loss": 0.8922,
      "step": 120
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018194444444444445,
      "loss": 0.7098,
      "step": 130
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018055555555555557,
      "loss": 0.6424,
      "step": 140
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0001791666666666667,
      "loss": 0.6945,
      "step": 150
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017777777777777779,
      "loss": 0.634,
      "step": 160
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0001763888888888889,
      "loss": 0.5476,
      "step": 170
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.000175,
      "loss": 0.6335,
      "step": 180
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017361111111111112,
      "loss": 0.4134,
      "step": 190
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017222222222222224,
      "loss": 0.6276,
      "step": 200
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 0.7763888888888889,
      "eval_loss": 0.6835257411003113,
      "eval_runtime": 46.4022,
      "eval_samples_per_second": 15.517,
      "eval_steps_per_second": 1.94,
      "step": 200
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017083333333333333,
      "loss": 0.6095,
      "step": 210
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016944444444444445,
      "loss": 0.5328,
      "step": 220
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016805555555555557,
      "loss": 0.4397,
      "step": 230
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.5338,
      "step": 240
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016527777777777778,
      "loss": 0.5808,
      "step": 250
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0001638888888888889,
      "loss": 0.601,
      "step": 260
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00016250000000000002,
      "loss": 0.445,
      "step": 270
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0001611111111111111,
      "loss": 0.3964,
      "step": 280
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015972222222222223,
      "loss": 0.5958,
      "step": 290
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015833333333333332,
      "loss": 0.6089,
      "step": 300
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.7930555555555555,
      "eval_loss": 0.5567876100540161,
      "eval_runtime": 46.1606,
      "eval_samples_per_second": 15.598,
      "eval_steps_per_second": 1.95,
      "step": 300
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015694444444444444,
      "loss": 0.518,
      "step": 310
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00015555555555555556,
      "loss": 0.4721,
      "step": 320
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015416666666666668,
      "loss": 0.4247,
      "step": 330
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015277777777777777,
      "loss": 0.4503,
      "step": 340
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0001513888888888889,
      "loss": 0.344,
      "step": 350
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.3077,
      "step": 360
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00014861111111111113,
      "loss": 0.2311,
      "step": 370
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00014722222222222223,
      "loss": 0.2271,
      "step": 380
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014583333333333335,
      "loss": 0.2595,
      "step": 390
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014444444444444444,
      "loss": 0.302,
      "step": 400
    },
    {
      "epoch": 1.11,
      "eval_accuracy": 0.8194444444444444,
      "eval_loss": 0.5886520147323608,
      "eval_runtime": 46.0306,
      "eval_samples_per_second": 15.642,
      "eval_steps_per_second": 1.955,
      "step": 400
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014305555555555556,
      "loss": 0.2532,
      "step": 410
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00014166666666666668,
      "loss": 0.3509,
      "step": 420
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014027777777777777,
      "loss": 0.2317,
      "step": 430
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001388888888888889,
      "loss": 0.2371,
      "step": 440
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0001375,
      "loss": 0.2115,
      "step": 450
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013611111111111113,
      "loss": 0.1403,
      "step": 460
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013472222222222225,
      "loss": 0.163,
      "step": 470
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.4951,
      "step": 480
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00013194444444444446,
      "loss": 0.274,
      "step": 490
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013055555555555555,
      "loss": 0.1723,
      "step": 500
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.8902777777777777,
      "eval_loss": 0.298811137676239,
      "eval_runtime": 45.7665,
      "eval_samples_per_second": 15.732,
      "eval_steps_per_second": 1.967,
      "step": 500
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012916666666666667,
      "loss": 0.291,
      "step": 510
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012777777777777776,
      "loss": 0.2423,
      "step": 520
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012638888888888888,
      "loss": 0.1181,
      "step": 530
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.000125,
      "loss": 0.2404,
      "step": 540
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012361111111111112,
      "loss": 0.1568,
      "step": 550
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012222222222222224,
      "loss": 0.1433,
      "step": 560
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012083333333333333,
      "loss": 0.1673,
      "step": 570
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00011944444444444445,
      "loss": 0.1575,
      "step": 580
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011805555555555556,
      "loss": 0.1429,
      "step": 590
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011666666666666668,
      "loss": 0.2522,
      "step": 600
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.9166666666666666,
      "eval_loss": 0.25636595487594604,
      "eval_runtime": 45.8795,
      "eval_samples_per_second": 15.693,
      "eval_steps_per_second": 1.962,
      "step": 600
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011527777777777777,
      "loss": 0.11,
      "step": 610
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011388888888888889,
      "loss": 0.1725,
      "step": 620
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00011250000000000001,
      "loss": 0.159,
      "step": 630
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.0854,
      "step": 640
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010972222222222224,
      "loss": 0.0999,
      "step": 650
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010833333333333333,
      "loss": 0.2217,
      "step": 660
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010694444444444445,
      "loss": 0.1712,
      "step": 670
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010555555555555557,
      "loss": 0.1299,
      "step": 680
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010416666666666667,
      "loss": 0.1376,
      "step": 690
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010277777777777778,
      "loss": 0.1115,
      "step": 700
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 0.9472222222222222,
      "eval_loss": 0.16800908744335175,
      "eval_runtime": 46.034,
      "eval_samples_per_second": 15.641,
      "eval_steps_per_second": 1.955,
      "step": 700
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010138888888888889,
      "loss": 0.1504,
      "step": 710
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001,
      "loss": 0.0608,
      "step": 720
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.861111111111112e-05,
      "loss": 0.0344,
      "step": 730
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.722222222222223e-05,
      "loss": 0.0326,
      "step": 740
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.583333333333334e-05,
      "loss": 0.2028,
      "step": 750
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.444444444444444e-05,
      "loss": 0.0335,
      "step": 760
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.305555555555556e-05,
      "loss": 0.0554,
      "step": 770
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.0525,
      "step": 780
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.027777777777779e-05,
      "loss": 0.0358,
      "step": 790
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.1445,
      "step": 800
    },
    {
      "epoch": 2.22,
      "eval_accuracy": 0.9402777777777778,
      "eval_loss": 0.2064695954322815,
      "eval_runtime": 45.6748,
      "eval_samples_per_second": 15.764,
      "eval_steps_per_second": 1.97,
      "step": 800
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.75e-05,
      "loss": 0.0905,
      "step": 810
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.611111111111112e-05,
      "loss": 0.006,
      "step": 820
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.472222222222222e-05,
      "loss": 0.0438,
      "step": 830
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.0172,
      "step": 840
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.194444444444445e-05,
      "loss": 0.021,
      "step": 850
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.055555555555556e-05,
      "loss": 0.0362,
      "step": 860
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.916666666666666e-05,
      "loss": 0.0007,
      "step": 870
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.0898,
      "step": 880
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.638888888888889e-05,
      "loss": 0.0117,
      "step": 890
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0648,
      "step": 900
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.9569444444444445,
      "eval_loss": 0.1673145890235901,
      "eval_runtime": 44.888,
      "eval_samples_per_second": 16.04,
      "eval_steps_per_second": 2.005,
      "step": 900
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.361111111111111e-05,
      "loss": 0.0248,
      "step": 910
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.222222222222222e-05,
      "loss": 0.0785,
      "step": 920
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.083333333333334e-05,
      "loss": 0.0009,
      "step": 930
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.944444444444444e-05,
      "loss": 0.0249,
      "step": 940
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.805555555555556e-05,
      "loss": 0.018,
      "step": 950
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0016,
      "step": 960
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.527777777777778e-05,
      "loss": 0.0616,
      "step": 970
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.388888888888888e-05,
      "loss": 0.0318,
      "step": 980
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.25e-05,
      "loss": 0.0253,
      "step": 990
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.111111111111112e-05,
      "loss": 0.0209,
      "step": 1000
    },
    {
      "epoch": 2.78,
      "eval_accuracy": 0.9569444444444445,
      "eval_loss": 0.16363362967967987,
      "eval_runtime": 45.529,
      "eval_samples_per_second": 15.814,
      "eval_steps_per_second": 1.977,
      "step": 1000
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.972222222222223e-05,
      "loss": 0.0209,
      "step": 1010
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.833333333333334e-05,
      "loss": 0.0082,
      "step": 1020
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.6944444444444445e-05,
      "loss": 0.051,
      "step": 1030
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.0023,
      "step": 1040
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.4166666666666664e-05,
      "loss": 0.0072,
      "step": 1050
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.2777777777777784e-05,
      "loss": 0.0777,
      "step": 1060
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.138888888888889e-05,
      "loss": 0.0186,
      "step": 1070
    },
    {
      "epoch": 3.0,
      "learning_rate": 5e-05,
      "loss": 0.0652,
      "step": 1080
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.0041,
      "step": 1090
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.0003,
      "step": 1100
    },
    {
      "epoch": 3.06,
      "eval_accuracy": 0.9694444444444444,
      "eval_loss": 0.12931323051452637,
      "eval_runtime": 45.4456,
      "eval_samples_per_second": 15.843,
      "eval_steps_per_second": 1.98,
      "step": 1100
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.0004,
      "step": 1110
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.0121,
      "step": 1120
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.002,
      "step": 1130
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.0035,
      "step": 1140
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.0057,
      "step": 1150
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.0002,
      "step": 1160
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0005,
      "step": 1170
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.0008,
      "step": 1180
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.017,
      "step": 1190
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0034,
      "step": 1200
    },
    {
      "epoch": 3.33,
      "eval_accuracy": 0.9791666666666666,
      "eval_loss": 0.08173014223575592,
      "eval_runtime": 44.7405,
      "eval_samples_per_second": 16.093,
      "eval_steps_per_second": 2.012,
      "step": 1200
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.194444444444444e-05,
      "loss": 0.0008,
      "step": 1210
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.0003,
      "step": 1220
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.0201,
      "step": 1230
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.0003,
      "step": 1240
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.0001,
      "step": 1250
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.5e-05,
      "loss": 0.0002,
      "step": 1260
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.361111111111111e-05,
      "loss": 0.0001,
      "step": 1270
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.0022,
      "step": 1280
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.0099,
      "step": 1290
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.0006,
      "step": 1300
    },
    {
      "epoch": 3.61,
      "eval_accuracy": 0.9833333333333333,
      "eval_loss": 0.08741980046033859,
      "eval_runtime": 44.9726,
      "eval_samples_per_second": 16.01,
      "eval_steps_per_second": 2.001,
      "step": 1300
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 0.0003,
      "step": 1310
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0003,
      "step": 1320
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.527777777777778e-05,
      "loss": 0.0016,
      "step": 1330
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.001,
      "step": 1340
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.25e-05,
      "loss": 0.0001,
      "step": 1350
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.0001,
      "step": 1360
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.0005,
      "step": 1370
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0001,
      "step": 1380
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.0002,
      "step": 1390
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.0023,
      "step": 1400
    },
    {
      "epoch": 3.89,
      "eval_accuracy": 0.9777777777777777,
      "eval_loss": 0.10755519568920135,
      "eval_runtime": 51.7521,
      "eval_samples_per_second": 13.912,
      "eval_steps_per_second": 1.739,
      "step": 1400
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.0266,
      "step": 1410
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.0128,
      "step": 1420
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.388888888888889e-06,
      "loss": 0.0001,
      "step": 1430
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.0,
      "loss": 0.0002,
      "step": 1440
    },
    {
      "epoch": 4.0,
      "step": 1440,
      "total_flos": 6.311528820941783e+18,
      "train_loss": 0.2221095090903772,
      "train_runtime": 4823.2045,
      "train_samples_per_second": 4.777,
      "train_steps_per_second": 0.299
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9791666666666666,
      "eval_loss": 0.10265839099884033,
      "eval_runtime": 48.2725,
      "eval_samples_per_second": 14.915,
      "eval_steps_per_second": 1.864,
      "step": 1440
    }
  ],
  "logging_steps": 10,
  "max_steps": 1440,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 6.311528820941783e+18,
  "trial_name": null,
  "trial_params": null
}