{
  "best_metric": 0.44273167848587036,
  "best_model_checkpoint": "rm_ckptreward-p100-w0.3-s0/checkpoint-1000",
  "epoch": 0.5583712739455373,
  "eval_steps": 200,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 0.6608,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 6.505149978319905e-06,
      "loss": 0.5639,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 7.385606273598311e-06,
      "loss": 0.5156,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 8.010299956639811e-06,
      "loss": 0.4808,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 8.494850021680093e-06,
      "loss": 0.4853,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 8.890756251918216e-06,
      "loss": 0.4662,
      "step": 60
    },
    {
      "epoch": 0.04,
      "learning_rate": 9.225490200071284e-06,
      "loss": 0.4775,
      "step": 70
    },
    {
      "epoch": 0.04,
      "learning_rate": 9.515449934959717e-06,
      "loss": 0.4649,
      "step": 80
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.771212547196623e-06,
      "loss": 0.4546,
      "step": 90
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.4609,
      "step": 100
    },
    {
      "epoch": 0.06,
      "learning_rate": 1e-05,
      "loss": 0.4577,
      "step": 110
    },
    {
      "epoch": 0.07,
      "learning_rate": 1e-05,
      "loss": 0.4729,
      "step": 120
    },
    {
      "epoch": 0.07,
      "learning_rate": 1e-05,
      "loss": 0.4653,
      "step": 130
    },
    {
      "epoch": 0.08,
      "learning_rate": 1e-05,
      "loss": 0.4504,
      "step": 140
    },
    {
      "epoch": 0.08,
      "learning_rate": 1e-05,
      "loss": 0.4574,
      "step": 150
    },
    {
      "epoch": 0.09,
      "learning_rate": 1e-05,
      "loss": 0.4482,
      "step": 160
    },
    {
      "epoch": 0.09,
      "learning_rate": 1e-05,
      "loss": 0.4592,
      "step": 170
    },
    {
      "epoch": 0.1,
      "learning_rate": 1e-05,
      "loss": 0.4446,
      "step": 180
    },
    {
      "epoch": 0.11,
      "learning_rate": 1e-05,
      "loss": 0.4612,
      "step": 190
    },
    {
      "epoch": 0.11,
      "learning_rate": 1e-05,
      "loss": 0.4493,
      "step": 200
    },
    {
      "epoch": 0.11,
      "eval_val_accuracy": 893.1666666666667,
      "eval_val_loss": 0.4625067412853241,
      "eval_val_runtime": 2785.6581,
      "eval_val_samples_per_second": 3.59,
      "eval_val_steps_per_second": 0.513,
      "step": 200
    },
    {
      "epoch": 0.11,
      "eval_test_accuracy": 897.0952380952382,
      "eval_test_loss": 0.4605662226676941,
      "eval_test_runtime": 2786.3481,
      "eval_test_samples_per_second": 3.589,
      "eval_test_steps_per_second": 0.513,
      "step": 200
    },
    {
      "epoch": 0.12,
      "learning_rate": 1e-05,
      "loss": 0.449,
      "step": 210
    },
    {
      "epoch": 0.12,
      "learning_rate": 1e-05,
      "loss": 0.4511,
      "step": 220
    },
    {
      "epoch": 0.13,
      "learning_rate": 1e-05,
      "loss": 0.4534,
      "step": 230
    },
    {
      "epoch": 0.13,
      "learning_rate": 1e-05,
      "loss": 0.4676,
      "step": 240
    },
    {
      "epoch": 0.14,
      "learning_rate": 1e-05,
      "loss": 0.4498,
      "step": 250
    },
    {
      "epoch": 0.15,
      "learning_rate": 1e-05,
      "loss": 0.4505,
      "step": 260
    },
    {
      "epoch": 0.15,
      "learning_rate": 1e-05,
      "loss": 0.4553,
      "step": 270
    },
    {
      "epoch": 0.16,
      "learning_rate": 1e-05,
      "loss": 0.4449,
      "step": 280
    },
    {
      "epoch": 0.16,
      "learning_rate": 1e-05,
      "loss": 0.4517,
      "step": 290
    },
    {
      "epoch": 0.17,
      "learning_rate": 1e-05,
      "loss": 0.4546,
      "step": 300
    },
    {
      "epoch": 0.17,
      "learning_rate": 1e-05,
      "loss": 0.4616,
      "step": 310
    },
    {
      "epoch": 0.18,
      "learning_rate": 1e-05,
      "loss": 0.4518,
      "step": 320
    },
    {
      "epoch": 0.18,
      "learning_rate": 1e-05,
      "loss": 0.4531,
      "step": 330
    },
    {
      "epoch": 0.19,
      "learning_rate": 1e-05,
      "loss": 0.4486,
      "step": 340
    },
    {
      "epoch": 0.2,
      "learning_rate": 1e-05,
      "loss": 0.4516,
      "step": 350
    },
    {
      "epoch": 0.2,
      "learning_rate": 1e-05,
      "loss": 0.4486,
      "step": 360
    },
    {
      "epoch": 0.21,
      "learning_rate": 1e-05,
      "loss": 0.4468,
      "step": 370
    },
    {
      "epoch": 0.21,
      "learning_rate": 1e-05,
      "loss": 0.4397,
      "step": 380
    },
    {
      "epoch": 0.22,
      "learning_rate": 1e-05,
      "loss": 0.4348,
      "step": 390
    },
    {
      "epoch": 0.22,
      "learning_rate": 1e-05,
      "loss": 0.444,
      "step": 400
    },
    {
      "epoch": 0.22,
      "eval_val_accuracy": 0.63225,
      "eval_val_loss": 0.45088696479797363,
      "eval_val_runtime": 2816.5532,
      "eval_val_samples_per_second": 3.55,
      "eval_val_steps_per_second": 0.507,
      "step": 400
    },
    {
      "epoch": 0.22,
      "eval_test_accuracy": 0.63275,
      "eval_test_loss": 0.4495917558670044,
      "eval_test_runtime": 2819.1352,
      "eval_test_samples_per_second": 3.547,
      "eval_test_steps_per_second": 0.507,
      "step": 400
    },
    {
      "epoch": 0.23,
      "learning_rate": 1e-05,
      "loss": 0.4549,
      "step": 410
    },
    {
      "epoch": 0.23,
      "learning_rate": 1e-05,
      "loss": 0.4479,
      "step": 420
    },
    {
      "epoch": 0.24,
      "learning_rate": 1e-05,
      "loss": 0.4396,
      "step": 430
    },
    {
      "epoch": 0.25,
      "learning_rate": 1e-05,
      "loss": 0.4464,
      "step": 440
    },
    {
      "epoch": 0.25,
      "learning_rate": 1e-05,
      "loss": 0.4412,
      "step": 450
    },
    {
      "epoch": 0.26,
      "learning_rate": 1e-05,
      "loss": 0.453,
      "step": 460
    },
    {
      "epoch": 0.26,
      "learning_rate": 1e-05,
      "loss": 0.449,
      "step": 470
    },
    {
      "epoch": 0.27,
      "learning_rate": 1e-05,
      "loss": 0.4446,
      "step": 480
    },
    {
      "epoch": 0.27,
      "learning_rate": 1e-05,
      "loss": 0.4352,
      "step": 490
    },
    {
      "epoch": 0.28,
      "learning_rate": 1e-05,
      "loss": 0.4518,
      "step": 500
    },
    {
      "epoch": 0.28,
      "learning_rate": 1e-05,
      "loss": 0.4378,
      "step": 510
    },
    {
      "epoch": 0.29,
      "learning_rate": 1e-05,
      "loss": 0.4428,
      "step": 520
    },
    {
      "epoch": 0.3,
      "learning_rate": 1e-05,
      "loss": 0.4543,
      "step": 530
    },
    {
      "epoch": 0.3,
      "learning_rate": 1e-05,
      "loss": 0.4368,
      "step": 540
    },
    {
      "epoch": 0.31,
      "learning_rate": 1e-05,
      "loss": 0.4442,
      "step": 550
    },
    {
      "epoch": 0.31,
      "learning_rate": 1e-05,
      "loss": 0.4473,
      "step": 560
    },
    {
      "epoch": 0.32,
      "learning_rate": 1e-05,
      "loss": 0.4472,
      "step": 570
    },
    {
      "epoch": 0.32,
      "learning_rate": 1e-05,
      "loss": 0.4339,
      "step": 580
    },
    {
      "epoch": 0.33,
      "learning_rate": 1e-05,
      "loss": 0.4397,
      "step": 590
    },
    {
      "epoch": 0.34,
      "learning_rate": 1e-05,
      "loss": 0.4468,
      "step": 600
    },
    {
      "epoch": 0.34,
      "eval_val_accuracy": 0.6332333333333333,
      "eval_val_loss": 0.4452223777770996,
      "eval_val_runtime": 2816.3837,
      "eval_val_samples_per_second": 3.551,
      "eval_val_steps_per_second": 0.507,
      "step": 600
    },
    {
      "epoch": 0.34,
      "eval_test_accuracy": 0.63525,
      "eval_test_loss": 0.4449952244758606,
      "eval_test_runtime": 2819.9617,
      "eval_test_samples_per_second": 3.546,
      "eval_test_steps_per_second": 0.507,
      "step": 600
    },
    {
      "epoch": 0.34,
      "learning_rate": 1e-05,
      "loss": 0.4372,
      "step": 610
    },
    {
      "epoch": 0.35,
      "learning_rate": 1e-05,
      "loss": 0.4433,
      "step": 620
    },
    {
      "epoch": 0.35,
      "learning_rate": 1e-05,
      "loss": 0.4442,
      "step": 630
    },
    {
      "epoch": 0.36,
      "learning_rate": 1e-05,
      "loss": 0.4427,
      "step": 640
    },
    {
      "epoch": 0.36,
      "learning_rate": 1e-05,
      "loss": 0.4367,
      "step": 650
    },
    {
      "epoch": 0.37,
      "learning_rate": 1e-05,
      "loss": 0.4535,
      "step": 660
    },
    {
      "epoch": 0.37,
      "learning_rate": 1e-05,
      "loss": 0.4428,
      "step": 670
    },
    {
      "epoch": 0.38,
      "learning_rate": 1e-05,
      "loss": 0.4434,
      "step": 680
    },
    {
      "epoch": 0.39,
      "learning_rate": 1e-05,
      "loss": 0.4416,
      "step": 690
    },
    {
      "epoch": 0.39,
      "learning_rate": 1e-05,
      "loss": 0.4432,
      "step": 700
    },
    {
      "epoch": 0.4,
      "learning_rate": 1e-05,
      "loss": 0.4345,
      "step": 710
    },
    {
      "epoch": 0.4,
      "learning_rate": 1e-05,
      "loss": 0.4403,
      "step": 720
    },
    {
      "epoch": 0.41,
      "learning_rate": 1e-05,
      "loss": 0.4395,
      "step": 730
    },
    {
      "epoch": 0.41,
      "learning_rate": 1e-05,
      "loss": 0.444,
      "step": 740
    },
    {
      "epoch": 0.42,
      "learning_rate": 1e-05,
      "loss": 0.4503,
      "step": 750
    },
    {
      "epoch": 0.42,
      "learning_rate": 1e-05,
      "loss": 0.443,
      "step": 760
    },
    {
      "epoch": 0.43,
      "learning_rate": 1e-05,
      "loss": 0.4507,
      "step": 770
    },
    {
      "epoch": 0.44,
      "learning_rate": 1e-05,
      "loss": 0.4536,
      "step": 780
    },
    {
      "epoch": 0.44,
      "learning_rate": 1e-05,
      "loss": 0.4578,
      "step": 790
    },
    {
      "epoch": 0.45,
      "learning_rate": 1e-05,
      "loss": 0.4276,
      "step": 800
    },
    {
      "epoch": 0.45,
      "eval_val_accuracy": 0.6333166666666667,
      "eval_val_loss": 0.4469798803329468,
      "eval_val_runtime": 2816.5985,
      "eval_val_samples_per_second": 3.55,
      "eval_val_steps_per_second": 0.507,
      "step": 800
    },
    {
      "epoch": 0.45,
      "eval_test_accuracy": 0.634,
      "eval_test_loss": 0.4446450173854828,
      "eval_test_runtime": 2817.8588,
      "eval_test_samples_per_second": 3.549,
      "eval_test_steps_per_second": 0.507,
      "step": 800
    },
    {
      "epoch": 0.45,
      "learning_rate": 1e-05,
      "loss": 0.449,
      "step": 810
    },
    {
      "epoch": 0.46,
      "learning_rate": 1e-05,
      "loss": 0.4335,
      "step": 820
    },
    {
      "epoch": 0.46,
      "learning_rate": 1e-05,
      "loss": 0.4587,
      "step": 830
    },
    {
      "epoch": 0.47,
      "learning_rate": 1e-05,
      "loss": 0.4519,
      "step": 840
    },
    {
      "epoch": 0.47,
      "learning_rate": 1e-05,
      "loss": 0.439,
      "step": 850
    },
    {
      "epoch": 0.48,
      "learning_rate": 1e-05,
      "loss": 0.4443,
      "step": 860
    },
    {
      "epoch": 0.49,
      "learning_rate": 1e-05,
      "loss": 0.4533,
      "step": 870
    },
    {
      "epoch": 0.49,
      "learning_rate": 1e-05,
      "loss": 0.435,
      "step": 880
    },
    {
      "epoch": 0.5,
      "learning_rate": 1e-05,
      "loss": 0.4397,
      "step": 890
    },
    {
      "epoch": 0.5,
      "learning_rate": 1e-05,
      "loss": 0.4461,
      "step": 900
    },
    {
      "epoch": 0.51,
      "learning_rate": 1e-05,
      "loss": 0.4575,
      "step": 910
    },
    {
      "epoch": 0.51,
      "learning_rate": 1e-05,
      "loss": 0.4541,
      "step": 920
    },
    {
      "epoch": 0.52,
      "learning_rate": 1e-05,
      "loss": 0.4394,
      "step": 930
    },
    {
      "epoch": 0.52,
      "learning_rate": 1e-05,
      "loss": 0.4503,
      "step": 940
    },
    {
      "epoch": 0.53,
      "learning_rate": 1e-05,
      "loss": 0.446,
      "step": 950
    },
    {
      "epoch": 0.54,
      "learning_rate": 1e-05,
      "loss": 0.4425,
      "step": 960
    },
    {
      "epoch": 0.54,
      "learning_rate": 1e-05,
      "loss": 0.433,
      "step": 970
    },
    {
      "epoch": 0.55,
      "learning_rate": 1e-05,
      "loss": 0.444,
      "step": 980
    },
    {
      "epoch": 0.55,
      "learning_rate": 1e-05,
      "loss": 0.4322,
      "step": 990
    },
    {
      "epoch": 0.56,
      "learning_rate": 1e-05,
      "loss": 0.4385,
      "step": 1000
    },
    {
      "epoch": 0.56,
      "eval_val_accuracy": 0.63495,
      "eval_val_loss": 0.44273167848587036,
      "eval_val_runtime": 2783.3749,
      "eval_val_samples_per_second": 3.593,
      "eval_val_steps_per_second": 0.513,
      "step": 1000
    },
    {
      "epoch": 0.56,
      "eval_test_accuracy": 0.6344,
      "eval_test_loss": 0.4418855905532837,
      "eval_test_runtime": 2784.0684,
      "eval_test_samples_per_second": 3.592,
      "eval_test_steps_per_second": 0.513,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 8950,
  "num_train_epochs": 5,
  "save_steps": 200,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}