{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7524454477050414,
  "eval_steps": 50,
  "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": 5e-07,
      "loss": 1.1001,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 1e-06,
      "loss": 1.0661,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.5e-06,
      "loss": 1.0782,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 2e-06,
      "loss": 1.0578,
      "step": 40
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.5e-06,
      "loss": 1.0395,
      "step": 50
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.4754781102152329,
      "eval_loss": 1.0322505235671997,
      "eval_runtime": 49.5848,
      "eval_samples_per_second": 605.307,
      "eval_steps_per_second": 1.19,
      "step": 50
    },
    {
      "epoch": 0.05,
      "learning_rate": 3e-06,
      "loss": 1.0308,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 3.5e-06,
      "loss": 1.0044,
      "step": 70
    },
    {
      "epoch": 0.06,
      "learning_rate": 4e-06,
      "loss": 1.0398,
      "step": 80
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.5e-06,
      "loss": 1.0116,
      "step": 90
    },
    {
      "epoch": 0.08,
      "learning_rate": 5e-06,
      "loss": 1.0067,
      "step": 100
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.5313187179316319,
      "eval_loss": 0.9863238334655762,
      "eval_runtime": 49.2356,
      "eval_samples_per_second": 609.599,
      "eval_steps_per_second": 1.198,
      "step": 100
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.5e-06,
      "loss": 1.0131,
      "step": 110
    },
    {
      "epoch": 0.09,
      "learning_rate": 6e-06,
      "loss": 1.0235,
      "step": 120
    },
    {
      "epoch": 0.1,
      "learning_rate": 6.5e-06,
      "loss": 1.0073,
      "step": 130
    },
    {
      "epoch": 0.11,
      "learning_rate": 7e-06,
      "loss": 0.9856,
      "step": 140
    },
    {
      "epoch": 0.11,
      "learning_rate": 7.5e-06,
      "loss": 1.0062,
      "step": 150
    },
    {
      "epoch": 0.11,
      "eval_accuracy": 0.5358499366962084,
      "eval_loss": 0.9735665917396545,
      "eval_runtime": 49.4568,
      "eval_samples_per_second": 606.874,
      "eval_steps_per_second": 1.193,
      "step": 150
    },
    {
      "epoch": 0.12,
      "learning_rate": 8e-06,
      "loss": 0.9622,
      "step": 160
    },
    {
      "epoch": 0.13,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.9583,
      "step": 170
    },
    {
      "epoch": 0.14,
      "learning_rate": 9e-06,
      "loss": 0.987,
      "step": 180
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.5e-06,
      "loss": 0.9995,
      "step": 190
    },
    {
      "epoch": 0.15,
      "learning_rate": 1e-05,
      "loss": 0.9953,
      "step": 200
    },
    {
      "epoch": 0.15,
      "eval_accuracy": 0.5406476977410541,
      "eval_loss": 0.9847397804260254,
      "eval_runtime": 49.5469,
      "eval_samples_per_second": 605.769,
      "eval_steps_per_second": 1.191,
      "step": 200
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.0500000000000001e-05,
      "loss": 0.9894,
      "step": 210
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.1e-05,
      "loss": 0.9857,
      "step": 220
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.15e-05,
      "loss": 0.9806,
      "step": 230
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.2e-05,
      "loss": 0.9802,
      "step": 240
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.25e-05,
      "loss": 0.9559,
      "step": 250
    },
    {
      "epoch": 0.19,
      "eval_accuracy": 0.5136269740787632,
      "eval_loss": 1.0206754207611084,
      "eval_runtime": 49.5108,
      "eval_samples_per_second": 606.212,
      "eval_steps_per_second": 1.192,
      "step": 250
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3e-05,
      "loss": 0.9877,
      "step": 260
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.35e-05,
      "loss": 0.9878,
      "step": 270
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.4e-05,
      "loss": 0.9646,
      "step": 280
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 0.9598,
      "step": 290
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.5e-05,
      "loss": 0.9567,
      "step": 300
    },
    {
      "epoch": 0.23,
      "eval_accuracy": 0.5693676284400613,
      "eval_loss": 0.9298574924468994,
      "eval_runtime": 49.5798,
      "eval_samples_per_second": 605.367,
      "eval_steps_per_second": 1.19,
      "step": 300
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.55e-05,
      "loss": 0.9659,
      "step": 310
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.6e-05,
      "loss": 0.9803,
      "step": 320
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.65e-05,
      "loss": 0.959,
      "step": 330
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.9298,
      "step": 340
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.8923,
      "step": 350
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.603285133604318,
      "eval_loss": 0.8731092810630798,
      "eval_runtime": 49.5126,
      "eval_samples_per_second": 606.189,
      "eval_steps_per_second": 1.192,
      "step": 350
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.8e-05,
      "loss": 0.9255,
      "step": 360
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.85e-05,
      "loss": 0.9567,
      "step": 370
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9e-05,
      "loss": 0.9135,
      "step": 380
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.95e-05,
      "loss": 0.8978,
      "step": 390
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 0.9617,
      "step": 400
    },
    {
      "epoch": 0.3,
      "eval_accuracy": 0.5401479309655494,
      "eval_loss": 0.9500759840011597,
      "eval_runtime": 49.5248,
      "eval_samples_per_second": 606.04,
      "eval_steps_per_second": 1.191,
      "step": 400
    },
    {
      "epoch": 0.31,
      "learning_rate": 2.05e-05,
      "loss": 0.9387,
      "step": 410
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.1000000000000002e-05,
      "loss": 0.8832,
      "step": 420
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.1499999999999997e-05,
      "loss": 0.9087,
      "step": 430
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2e-05,
      "loss": 0.9192,
      "step": 440
    },
    {
      "epoch": 0.34,
      "learning_rate": 2.2499999999999998e-05,
      "loss": 0.8456,
      "step": 450
    },
    {
      "epoch": 0.34,
      "eval_accuracy": 0.6245085626707536,
      "eval_loss": 0.8298193216323853,
      "eval_runtime": 49.5226,
      "eval_samples_per_second": 606.066,
      "eval_steps_per_second": 1.191,
      "step": 450
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.3e-05,
      "loss": 0.8511,
      "step": 460
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.3500000000000002e-05,
      "loss": 0.9211,
      "step": 470
    },
    {
      "epoch": 0.36,
      "learning_rate": 2.4e-05,
      "loss": 0.8416,
      "step": 480
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 0.849,
      "step": 490
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.5e-05,
      "loss": 0.8098,
      "step": 500
    },
    {
      "epoch": 0.38,
      "eval_accuracy": 0.6523622309588859,
      "eval_loss": 0.771039605140686,
      "eval_runtime": 49.4859,
      "eval_samples_per_second": 606.517,
      "eval_steps_per_second": 1.192,
      "step": 500
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.55e-05,
      "loss": 0.8406,
      "step": 510
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.6e-05,
      "loss": 0.8868,
      "step": 520
    },
    {
      "epoch": 0.4,
      "learning_rate": 2.65e-05,
      "loss": 0.8206,
      "step": 530
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.7e-05,
      "loss": 0.7607,
      "step": 540
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.75e-05,
      "loss": 0.6892,
      "step": 550
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.7312254281335376,
      "eval_loss": 0.670229971408844,
      "eval_runtime": 49.5662,
      "eval_samples_per_second": 605.533,
      "eval_steps_per_second": 1.19,
      "step": 550
    },
    {
      "epoch": 0.42,
      "learning_rate": 2.8e-05,
      "loss": 0.7065,
      "step": 560
    },
    {
      "epoch": 0.43,
      "learning_rate": 2.85e-05,
      "loss": 0.7075,
      "step": 570
    },
    {
      "epoch": 0.44,
      "learning_rate": 2.9000000000000004e-05,
      "loss": 0.6793,
      "step": 580
    },
    {
      "epoch": 0.44,
      "learning_rate": 2.95e-05,
      "loss": 0.6057,
      "step": 590
    },
    {
      "epoch": 0.45,
      "learning_rate": 3e-05,
      "loss": 0.5794,
      "step": 600
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.7694409275671353,
      "eval_loss": 0.6066040396690369,
      "eval_runtime": 49.5056,
      "eval_samples_per_second": 606.275,
      "eval_steps_per_second": 1.192,
      "step": 600
    },
    {
      "epoch": 0.46,
      "learning_rate": 3.05e-05,
      "loss": 0.609,
      "step": 610
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.1e-05,
      "loss": 0.6058,
      "step": 620
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.15e-05,
      "loss": 0.5518,
      "step": 630
    },
    {
      "epoch": 0.48,
      "learning_rate": 3.2e-05,
      "loss": 0.6132,
      "step": 640
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.5262,
      "step": 650
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.8218498034250683,
      "eval_loss": 0.4961170256137848,
      "eval_runtime": 49.5394,
      "eval_samples_per_second": 605.861,
      "eval_steps_per_second": 1.191,
      "step": 650
    },
    {
      "epoch": 0.5,
      "learning_rate": 3.3e-05,
      "loss": 0.5355,
      "step": 660
    },
    {
      "epoch": 0.5,
      "learning_rate": 3.35e-05,
      "loss": 0.4905,
      "step": 670
    },
    {
      "epoch": 0.51,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.4992,
      "step": 680
    },
    {
      "epoch": 0.52,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 0.4959,
      "step": 690
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.5000000000000004e-05,
      "loss": 0.4693,
      "step": 700
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.851469314319984,
      "eval_loss": 0.43606042861938477,
      "eval_runtime": 49.5557,
      "eval_samples_per_second": 605.662,
      "eval_steps_per_second": 1.191,
      "step": 700
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.5499999999999996e-05,
      "loss": 0.4725,
      "step": 710
    },
    {
      "epoch": 0.54,
      "learning_rate": 3.6e-05,
      "loss": 0.5033,
      "step": 720
    },
    {
      "epoch": 0.55,
      "learning_rate": 3.65e-05,
      "loss": 0.4679,
      "step": 730
    },
    {
      "epoch": 0.56,
      "learning_rate": 3.7e-05,
      "loss": 0.4639,
      "step": 740
    },
    {
      "epoch": 0.56,
      "learning_rate": 3.75e-05,
      "loss": 0.434,
      "step": 750
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 0.8657293263143866,
      "eval_loss": 0.41026708483695984,
      "eval_runtime": 49.5112,
      "eval_samples_per_second": 606.206,
      "eval_steps_per_second": 1.192,
      "step": 750
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.8e-05,
      "loss": 0.3815,
      "step": 760
    },
    {
      "epoch": 0.58,
      "learning_rate": 3.85e-05,
      "loss": 0.4073,
      "step": 770
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.9e-05,
      "loss": 0.424,
      "step": 780
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.95e-05,
      "loss": 0.3867,
      "step": 790
    },
    {
      "epoch": 0.6,
      "learning_rate": 4e-05,
      "loss": 0.463,
      "step": 800
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.8671286732858,
      "eval_loss": 0.3895261585712433,
      "eval_runtime": 49.5727,
      "eval_samples_per_second": 605.455,
      "eval_steps_per_second": 1.19,
      "step": 800
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.05e-05,
      "loss": 0.407,
      "step": 810
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.1e-05,
      "loss": 0.4077,
      "step": 820
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.1500000000000006e-05,
      "loss": 0.4015,
      "step": 830
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 0.3652,
      "step": 840
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.25e-05,
      "loss": 0.411,
      "step": 850
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.8794895715332844,
      "eval_loss": 0.36858388781547546,
      "eval_runtime": 49.5593,
      "eval_samples_per_second": 605.618,
      "eval_steps_per_second": 1.19,
      "step": 850
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.2999999999999995e-05,
      "loss": 0.4162,
      "step": 860
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.35e-05,
      "loss": 0.3985,
      "step": 870
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.4e-05,
      "loss": 0.3925,
      "step": 880
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.45e-05,
      "loss": 0.3621,
      "step": 890
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 0.3715,
      "step": 900
    },
    {
      "epoch": 0.68,
      "eval_accuracy": 0.8932164989671486,
      "eval_loss": 0.33608904480934143,
      "eval_runtime": 49.5424,
      "eval_samples_per_second": 605.825,
      "eval_steps_per_second": 1.191,
      "step": 900
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.55e-05,
      "loss": 0.3337,
      "step": 910
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.6e-05,
      "loss": 0.3592,
      "step": 920
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.65e-05,
      "loss": 0.3845,
      "step": 930
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 0.4175,
      "step": 940
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.75e-05,
      "loss": 0.4039,
      "step": 950
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 0.8966815486106484,
      "eval_loss": 0.3278920352458954,
      "eval_runtime": 49.6606,
      "eval_samples_per_second": 604.383,
      "eval_steps_per_second": 1.188,
      "step": 950
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.8e-05,
      "loss": 0.3474,
      "step": 960
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.85e-05,
      "loss": 0.3804,
      "step": 970
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.9000000000000005e-05,
      "loss": 0.3432,
      "step": 980
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 0.3184,
      "step": 990
    },
    {
      "epoch": 0.75,
      "learning_rate": 5e-05,
      "loss": 0.369,
      "step": 1000
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.9011461318051576,
      "eval_loss": 0.33548465371131897,
      "eval_runtime": 49.6573,
      "eval_samples_per_second": 604.422,
      "eval_steps_per_second": 1.188,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 3987,
  "num_train_epochs": 3,
  "save_steps": 250,
  "total_flos": 3.3678516289536e+16,
  "trial_name": null,
  "trial_params": null
}