{
  "best_metric": 0.8618181818181818,
  "best_model_checkpoint": "swiftformer-xs-RD\\checkpoint-1167",
  "epoch": 39.75155279503105,
  "eval_steps": 500,
  "global_step": 1600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "learning_rate": 1.875e-05,
      "loss": 1.6091,
      "step": 10
    },
    {
      "epoch": 0.5,
      "learning_rate": 3.75e-05,
      "loss": 1.6066,
      "step": 20
    },
    {
      "epoch": 0.75,
      "learning_rate": 5.625e-05,
      "loss": 1.5954,
      "step": 30
    },
    {
      "epoch": 0.99,
      "learning_rate": 7.5e-05,
      "loss": 1.5571,
      "step": 40
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.5181818181818182,
      "eval_loss": 1.519059658050537,
      "eval_runtime": 2.8787,
      "eval_samples_per_second": 191.057,
      "eval_steps_per_second": 12.158,
      "step": 40
    },
    {
      "epoch": 1.24,
      "learning_rate": 9.374999999999999e-05,
      "loss": 1.4441,
      "step": 50
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0001125,
      "loss": 1.255,
      "step": 60
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00013125,
      "loss": 1.0283,
      "step": 70
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00015,
      "loss": 0.979,
      "step": 80
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 0.7090909090909091,
      "eval_loss": 0.8952864408493042,
      "eval_runtime": 1.7304,
      "eval_samples_per_second": 317.84,
      "eval_steps_per_second": 20.226,
      "step": 80
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.00016874999999999998,
      "loss": 0.864,
      "step": 90
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.00018749999999999998,
      "loss": 0.859,
      "step": 100
    },
    {
      "epoch": 2.73,
      "learning_rate": 0.00020624999999999997,
      "loss": 0.7978,
      "step": 110
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.000225,
      "loss": 0.8343,
      "step": 120
    },
    {
      "epoch": 2.98,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.7519798874855042,
      "eval_runtime": 1.3968,
      "eval_samples_per_second": 393.746,
      "eval_steps_per_second": 25.057,
      "step": 120
    },
    {
      "epoch": 3.23,
      "learning_rate": 0.00024375,
      "loss": 0.7879,
      "step": 130
    },
    {
      "epoch": 3.48,
      "learning_rate": 0.0002625,
      "loss": 0.7814,
      "step": 140
    },
    {
      "epoch": 3.73,
      "learning_rate": 0.00028125,
      "loss": 0.7806,
      "step": 150
    },
    {
      "epoch": 3.98,
      "learning_rate": 0.0003,
      "loss": 0.7755,
      "step": 160
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7527272727272727,
      "eval_loss": 0.7222614288330078,
      "eval_runtime": 1.3418,
      "eval_samples_per_second": 409.89,
      "eval_steps_per_second": 26.084,
      "step": 161
    },
    {
      "epoch": 4.22,
      "learning_rate": 0.00029791666666666665,
      "loss": 0.7536,
      "step": 170
    },
    {
      "epoch": 4.47,
      "learning_rate": 0.00029583333333333333,
      "loss": 0.6664,
      "step": 180
    },
    {
      "epoch": 4.72,
      "learning_rate": 0.00029374999999999996,
      "loss": 0.6218,
      "step": 190
    },
    {
      "epoch": 4.97,
      "learning_rate": 0.00029166666666666664,
      "loss": 0.7064,
      "step": 200
    },
    {
      "epoch": 4.99,
      "eval_accuracy": 0.74,
      "eval_loss": 0.6760307550430298,
      "eval_runtime": 1.3443,
      "eval_samples_per_second": 409.125,
      "eval_steps_per_second": 26.035,
      "step": 201
    },
    {
      "epoch": 5.22,
      "learning_rate": 0.0002895833333333333,
      "loss": 0.715,
      "step": 210
    },
    {
      "epoch": 5.47,
      "learning_rate": 0.0002875,
      "loss": 0.7128,
      "step": 220
    },
    {
      "epoch": 5.71,
      "learning_rate": 0.0002854166666666666,
      "loss": 0.6436,
      "step": 230
    },
    {
      "epoch": 5.96,
      "learning_rate": 0.0002833333333333333,
      "loss": 0.6296,
      "step": 240
    },
    {
      "epoch": 5.99,
      "eval_accuracy": 0.8018181818181818,
      "eval_loss": 0.5624179840087891,
      "eval_runtime": 1.5624,
      "eval_samples_per_second": 352.03,
      "eval_steps_per_second": 22.402,
      "step": 241
    },
    {
      "epoch": 6.21,
      "learning_rate": 0.00028125,
      "loss": 0.6154,
      "step": 250
    },
    {
      "epoch": 6.46,
      "learning_rate": 0.00027916666666666666,
      "loss": 0.6749,
      "step": 260
    },
    {
      "epoch": 6.71,
      "learning_rate": 0.00027708333333333334,
      "loss": 0.6336,
      "step": 270
    },
    {
      "epoch": 6.96,
      "learning_rate": 0.00027499999999999996,
      "loss": 0.6387,
      "step": 280
    },
    {
      "epoch": 6.98,
      "eval_accuracy": 0.8072727272727273,
      "eval_loss": 0.5362208485603333,
      "eval_runtime": 1.4093,
      "eval_samples_per_second": 390.255,
      "eval_steps_per_second": 24.834,
      "step": 281
    },
    {
      "epoch": 7.2,
      "learning_rate": 0.00027291666666666664,
      "loss": 0.5918,
      "step": 290
    },
    {
      "epoch": 7.45,
      "learning_rate": 0.0002708333333333333,
      "loss": 0.5901,
      "step": 300
    },
    {
      "epoch": 7.7,
      "learning_rate": 0.00026875,
      "loss": 0.6415,
      "step": 310
    },
    {
      "epoch": 7.95,
      "learning_rate": 0.0002666666666666666,
      "loss": 0.5793,
      "step": 320
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7963636363636364,
      "eval_loss": 0.5241875648498535,
      "eval_runtime": 1.3758,
      "eval_samples_per_second": 399.758,
      "eval_steps_per_second": 25.439,
      "step": 322
    },
    {
      "epoch": 8.2,
      "learning_rate": 0.0002645833333333333,
      "loss": 0.5718,
      "step": 330
    },
    {
      "epoch": 8.45,
      "learning_rate": 0.0002625,
      "loss": 0.5651,
      "step": 340
    },
    {
      "epoch": 8.7,
      "learning_rate": 0.00026041666666666666,
      "loss": 0.5828,
      "step": 350
    },
    {
      "epoch": 8.94,
      "learning_rate": 0.00025833333333333334,
      "loss": 0.5723,
      "step": 360
    },
    {
      "epoch": 8.99,
      "eval_accuracy": 0.8309090909090909,
      "eval_loss": 0.454416960477829,
      "eval_runtime": 1.3498,
      "eval_samples_per_second": 407.461,
      "eval_steps_per_second": 25.929,
      "step": 362
    },
    {
      "epoch": 9.19,
      "learning_rate": 0.00025624999999999997,
      "loss": 0.5246,
      "step": 370
    },
    {
      "epoch": 9.44,
      "learning_rate": 0.00025416666666666665,
      "loss": 0.5618,
      "step": 380
    },
    {
      "epoch": 9.69,
      "learning_rate": 0.0002520833333333333,
      "loss": 0.5462,
      "step": 390
    },
    {
      "epoch": 9.94,
      "learning_rate": 0.00025,
      "loss": 0.5897,
      "step": 400
    },
    {
      "epoch": 9.99,
      "eval_accuracy": 0.8272727272727273,
      "eval_loss": 0.4509519636631012,
      "eval_runtime": 1.3383,
      "eval_samples_per_second": 410.963,
      "eval_steps_per_second": 26.152,
      "step": 402
    },
    {
      "epoch": 10.19,
      "learning_rate": 0.00024791666666666663,
      "loss": 0.5194,
      "step": 410
    },
    {
      "epoch": 10.43,
      "learning_rate": 0.0002458333333333333,
      "loss": 0.4946,
      "step": 420
    },
    {
      "epoch": 10.68,
      "learning_rate": 0.00024375,
      "loss": 0.5858,
      "step": 430
    },
    {
      "epoch": 10.93,
      "learning_rate": 0.00024166666666666664,
      "loss": 0.5443,
      "step": 440
    },
    {
      "epoch": 10.98,
      "eval_accuracy": 0.8054545454545454,
      "eval_loss": 0.46854037046432495,
      "eval_runtime": 1.3393,
      "eval_samples_per_second": 410.65,
      "eval_steps_per_second": 26.132,
      "step": 442
    },
    {
      "epoch": 11.18,
      "learning_rate": 0.00023958333333333332,
      "loss": 0.5482,
      "step": 450
    },
    {
      "epoch": 11.43,
      "learning_rate": 0.00023749999999999997,
      "loss": 0.4724,
      "step": 460
    },
    {
      "epoch": 11.68,
      "learning_rate": 0.00023541666666666665,
      "loss": 0.5285,
      "step": 470
    },
    {
      "epoch": 11.93,
      "learning_rate": 0.0002333333333333333,
      "loss": 0.4959,
      "step": 480
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8254545454545454,
      "eval_loss": 0.42480796575546265,
      "eval_runtime": 1.4113,
      "eval_samples_per_second": 389.699,
      "eval_steps_per_second": 24.799,
      "step": 483
    },
    {
      "epoch": 12.17,
      "learning_rate": 0.00023124999999999998,
      "loss": 0.6238,
      "step": 490
    },
    {
      "epoch": 12.42,
      "learning_rate": 0.00022916666666666664,
      "loss": 0.5296,
      "step": 500
    },
    {
      "epoch": 12.67,
      "learning_rate": 0.00022708333333333331,
      "loss": 0.4759,
      "step": 510
    },
    {
      "epoch": 12.92,
      "learning_rate": 0.000225,
      "loss": 0.4743,
      "step": 520
    },
    {
      "epoch": 12.99,
      "eval_accuracy": 0.8309090909090909,
      "eval_loss": 0.4307559132575989,
      "eval_runtime": 1.3628,
      "eval_samples_per_second": 403.57,
      "eval_steps_per_second": 25.682,
      "step": 523
    },
    {
      "epoch": 13.17,
      "learning_rate": 0.00022291666666666665,
      "loss": 0.6277,
      "step": 530
    },
    {
      "epoch": 13.42,
      "learning_rate": 0.00022083333333333333,
      "loss": 0.5356,
      "step": 540
    },
    {
      "epoch": 13.66,
      "learning_rate": 0.00021874999999999998,
      "loss": 0.5112,
      "step": 550
    },
    {
      "epoch": 13.91,
      "learning_rate": 0.00021666666666666666,
      "loss": 0.4679,
      "step": 560
    },
    {
      "epoch": 13.99,
      "eval_accuracy": 0.84,
      "eval_loss": 0.39991825819015503,
      "eval_runtime": 1.3308,
      "eval_samples_per_second": 413.283,
      "eval_steps_per_second": 26.3,
      "step": 563
    },
    {
      "epoch": 14.16,
      "learning_rate": 0.0002145833333333333,
      "loss": 0.4224,
      "step": 570
    },
    {
      "epoch": 14.41,
      "learning_rate": 0.0002125,
      "loss": 0.4699,
      "step": 580
    },
    {
      "epoch": 14.66,
      "learning_rate": 0.00021041666666666664,
      "loss": 0.4727,
      "step": 590
    },
    {
      "epoch": 14.91,
      "learning_rate": 0.00020833333333333332,
      "loss": 0.4997,
      "step": 600
    },
    {
      "epoch": 14.98,
      "eval_accuracy": 0.8381818181818181,
      "eval_loss": 0.41513749957084656,
      "eval_runtime": 1.3563,
      "eval_samples_per_second": 405.507,
      "eval_steps_per_second": 25.805,
      "step": 603
    },
    {
      "epoch": 15.16,
      "learning_rate": 0.00020624999999999997,
      "loss": 0.4647,
      "step": 610
    },
    {
      "epoch": 15.4,
      "learning_rate": 0.00020416666666666665,
      "loss": 0.507,
      "step": 620
    },
    {
      "epoch": 15.65,
      "learning_rate": 0.00020208333333333333,
      "loss": 0.4575,
      "step": 630
    },
    {
      "epoch": 15.9,
      "learning_rate": 0.00019999999999999998,
      "loss": 0.4602,
      "step": 640
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8218181818181818,
      "eval_loss": 0.41115185618400574,
      "eval_runtime": 1.3393,
      "eval_samples_per_second": 410.657,
      "eval_steps_per_second": 26.133,
      "step": 644
    },
    {
      "epoch": 16.15,
      "learning_rate": 0.00019791666666666663,
      "loss": 0.4685,
      "step": 650
    },
    {
      "epoch": 16.4,
      "learning_rate": 0.00019583333333333331,
      "loss": 0.4071,
      "step": 660
    },
    {
      "epoch": 16.65,
      "learning_rate": 0.00019375,
      "loss": 0.4358,
      "step": 670
    },
    {
      "epoch": 16.89,
      "learning_rate": 0.00019166666666666665,
      "loss": 0.4459,
      "step": 680
    },
    {
      "epoch": 16.99,
      "eval_accuracy": 0.8345454545454546,
      "eval_loss": 0.4195989668369293,
      "eval_runtime": 1.4108,
      "eval_samples_per_second": 389.839,
      "eval_steps_per_second": 24.808,
      "step": 684
    },
    {
      "epoch": 17.14,
      "learning_rate": 0.0001895833333333333,
      "loss": 0.4683,
      "step": 690
    },
    {
      "epoch": 17.39,
      "learning_rate": 0.00018749999999999998,
      "loss": 0.4322,
      "step": 700
    },
    {
      "epoch": 17.64,
      "learning_rate": 0.00018541666666666666,
      "loss": 0.4099,
      "step": 710
    },
    {
      "epoch": 17.89,
      "learning_rate": 0.00018333333333333334,
      "loss": 0.4668,
      "step": 720
    },
    {
      "epoch": 17.99,
      "eval_accuracy": 0.8290909090909091,
      "eval_loss": 0.4041709899902344,
      "eval_runtime": 1.5199,
      "eval_samples_per_second": 361.876,
      "eval_steps_per_second": 23.028,
      "step": 724
    },
    {
      "epoch": 18.14,
      "learning_rate": 0.00018124999999999996,
      "loss": 0.5344,
      "step": 730
    },
    {
      "epoch": 18.39,
      "learning_rate": 0.00017916666666666664,
      "loss": 0.4369,
      "step": 740
    },
    {
      "epoch": 18.63,
      "learning_rate": 0.00017708333333333332,
      "loss": 0.4014,
      "step": 750
    },
    {
      "epoch": 18.88,
      "learning_rate": 0.000175,
      "loss": 0.41,
      "step": 760
    },
    {
      "epoch": 18.98,
      "eval_accuracy": 0.8436363636363636,
      "eval_loss": 0.4112071990966797,
      "eval_runtime": 1.3473,
      "eval_samples_per_second": 408.216,
      "eval_steps_per_second": 25.977,
      "step": 764
    },
    {
      "epoch": 19.13,
      "learning_rate": 0.00017291666666666662,
      "loss": 0.3628,
      "step": 770
    },
    {
      "epoch": 19.38,
      "learning_rate": 0.0001708333333333333,
      "loss": 0.3968,
      "step": 780
    },
    {
      "epoch": 19.63,
      "learning_rate": 0.00016874999999999998,
      "loss": 0.4301,
      "step": 790
    },
    {
      "epoch": 19.88,
      "learning_rate": 0.00016666666666666666,
      "loss": 0.4349,
      "step": 800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8418181818181818,
      "eval_loss": 0.4119725525379181,
      "eval_runtime": 1.3453,
      "eval_samples_per_second": 408.821,
      "eval_steps_per_second": 26.016,
      "step": 805
    },
    {
      "epoch": 20.12,
      "learning_rate": 0.00016458333333333334,
      "loss": 0.3921,
      "step": 810
    },
    {
      "epoch": 20.37,
      "learning_rate": 0.00016249999999999997,
      "loss": 0.4351,
      "step": 820
    },
    {
      "epoch": 20.62,
      "learning_rate": 0.00016041666666666664,
      "loss": 0.4144,
      "step": 830
    },
    {
      "epoch": 20.87,
      "learning_rate": 0.00015833333333333332,
      "loss": 0.4558,
      "step": 840
    },
    {
      "epoch": 20.99,
      "eval_accuracy": 0.8218181818181818,
      "eval_loss": 0.4554423391819,
      "eval_runtime": 1.3368,
      "eval_samples_per_second": 411.426,
      "eval_steps_per_second": 26.182,
      "step": 845
    },
    {
      "epoch": 21.12,
      "learning_rate": 0.00015625,
      "loss": 0.4053,
      "step": 850
    },
    {
      "epoch": 21.37,
      "learning_rate": 0.00015416666666666663,
      "loss": 0.3483,
      "step": 860
    },
    {
      "epoch": 21.61,
      "learning_rate": 0.0001520833333333333,
      "loss": 0.4139,
      "step": 870
    },
    {
      "epoch": 21.86,
      "learning_rate": 0.00015,
      "loss": 0.3952,
      "step": 880
    },
    {
      "epoch": 21.99,
      "eval_accuracy": 0.850909090909091,
      "eval_loss": 0.3774945139884949,
      "eval_runtime": 1.3708,
      "eval_samples_per_second": 401.212,
      "eval_steps_per_second": 25.532,
      "step": 885
    },
    {
      "epoch": 22.11,
      "learning_rate": 0.00014791666666666667,
      "loss": 0.3661,
      "step": 890
    },
    {
      "epoch": 22.36,
      "learning_rate": 0.00014583333333333332,
      "loss": 0.3706,
      "step": 900
    },
    {
      "epoch": 22.61,
      "learning_rate": 0.00014375,
      "loss": 0.4127,
      "step": 910
    },
    {
      "epoch": 22.86,
      "learning_rate": 0.00014166666666666665,
      "loss": 0.3395,
      "step": 920
    },
    {
      "epoch": 22.98,
      "eval_accuracy": 0.84,
      "eval_loss": 0.4027109742164612,
      "eval_runtime": 1.4218,
      "eval_samples_per_second": 386.825,
      "eval_steps_per_second": 24.616,
      "step": 925
    },
    {
      "epoch": 23.11,
      "learning_rate": 0.00013958333333333333,
      "loss": 0.3821,
      "step": 930
    },
    {
      "epoch": 23.35,
      "learning_rate": 0.00013749999999999998,
      "loss": 0.3408,
      "step": 940
    },
    {
      "epoch": 23.6,
      "learning_rate": 0.00013541666666666666,
      "loss": 0.3344,
      "step": 950
    },
    {
      "epoch": 23.85,
      "learning_rate": 0.0001333333333333333,
      "loss": 0.3525,
      "step": 960
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.850909090909091,
      "eval_loss": 0.3956778645515442,
      "eval_runtime": 1.3433,
      "eval_samples_per_second": 409.433,
      "eval_steps_per_second": 26.055,
      "step": 966
    },
    {
      "epoch": 24.1,
      "learning_rate": 0.00013125,
      "loss": 0.3803,
      "step": 970
    },
    {
      "epoch": 24.35,
      "learning_rate": 0.00012916666666666667,
      "loss": 0.3213,
      "step": 980
    },
    {
      "epoch": 24.6,
      "learning_rate": 0.00012708333333333332,
      "loss": 0.3845,
      "step": 990
    },
    {
      "epoch": 24.84,
      "learning_rate": 0.000125,
      "loss": 0.3497,
      "step": 1000
    },
    {
      "epoch": 24.99,
      "eval_accuracy": 0.8527272727272728,
      "eval_loss": 0.4020462930202484,
      "eval_runtime": 1.3428,
      "eval_samples_per_second": 409.587,
      "eval_steps_per_second": 26.065,
      "step": 1006
    },
    {
      "epoch": 25.09,
      "learning_rate": 0.00012291666666666665,
      "loss": 0.375,
      "step": 1010
    },
    {
      "epoch": 25.34,
      "learning_rate": 0.00012083333333333332,
      "loss": 0.3141,
      "step": 1020
    },
    {
      "epoch": 25.59,
      "learning_rate": 0.00011874999999999999,
      "loss": 0.3899,
      "step": 1030
    },
    {
      "epoch": 25.84,
      "learning_rate": 0.00011666666666666665,
      "loss": 0.3625,
      "step": 1040
    },
    {
      "epoch": 25.99,
      "eval_accuracy": 0.8563636363636363,
      "eval_loss": 0.39101672172546387,
      "eval_runtime": 1.3918,
      "eval_samples_per_second": 395.16,
      "eval_steps_per_second": 25.147,
      "step": 1046
    },
    {
      "epoch": 26.09,
      "learning_rate": 0.00011458333333333332,
      "loss": 0.3777,
      "step": 1050
    },
    {
      "epoch": 26.34,
      "learning_rate": 0.0001125,
      "loss": 0.3572,
      "step": 1060
    },
    {
      "epoch": 26.58,
      "learning_rate": 0.00011041666666666666,
      "loss": 0.3486,
      "step": 1070
    },
    {
      "epoch": 26.83,
      "learning_rate": 0.00010833333333333333,
      "loss": 0.3008,
      "step": 1080
    },
    {
      "epoch": 26.98,
      "eval_accuracy": 0.8581818181818182,
      "eval_loss": 0.3789851665496826,
      "eval_runtime": 1.4113,
      "eval_samples_per_second": 389.701,
      "eval_steps_per_second": 24.799,
      "step": 1086
    },
    {
      "epoch": 27.08,
      "learning_rate": 0.00010625,
      "loss": 0.32,
      "step": 1090
    },
    {
      "epoch": 27.33,
      "learning_rate": 0.00010416666666666666,
      "loss": 0.356,
      "step": 1100
    },
    {
      "epoch": 27.58,
      "learning_rate": 0.00010208333333333333,
      "loss": 0.3053,
      "step": 1110
    },
    {
      "epoch": 27.83,
      "learning_rate": 9.999999999999999e-05,
      "loss": 0.2907,
      "step": 1120
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.8563636363636363,
      "eval_loss": 0.38305726647377014,
      "eval_runtime": 1.3603,
      "eval_samples_per_second": 404.315,
      "eval_steps_per_second": 25.729,
      "step": 1127
    },
    {
      "epoch": 28.07,
      "learning_rate": 9.791666666666666e-05,
      "loss": 0.3384,
      "step": 1130
    },
    {
      "epoch": 28.32,
      "learning_rate": 9.583333333333332e-05,
      "loss": 0.2956,
      "step": 1140
    },
    {
      "epoch": 28.57,
      "learning_rate": 9.374999999999999e-05,
      "loss": 0.2923,
      "step": 1150
    },
    {
      "epoch": 28.82,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.3165,
      "step": 1160
    },
    {
      "epoch": 28.99,
      "eval_accuracy": 0.8618181818181818,
      "eval_loss": 0.4005330204963684,
      "eval_runtime": 1.3778,
      "eval_samples_per_second": 399.18,
      "eval_steps_per_second": 25.402,
      "step": 1167
    },
    {
      "epoch": 29.07,
      "learning_rate": 8.958333333333332e-05,
      "loss": 0.302,
      "step": 1170
    },
    {
      "epoch": 29.32,
      "learning_rate": 8.75e-05,
      "loss": 0.2723,
      "step": 1180
    },
    {
      "epoch": 29.57,
      "learning_rate": 8.541666666666665e-05,
      "loss": 0.2953,
      "step": 1190
    },
    {
      "epoch": 29.81,
      "learning_rate": 8.333333333333333e-05,
      "loss": 0.3231,
      "step": 1200
    },
    {
      "epoch": 29.99,
      "eval_accuracy": 0.8327272727272728,
      "eval_loss": 0.4512370228767395,
      "eval_runtime": 1.3443,
      "eval_samples_per_second": 409.127,
      "eval_steps_per_second": 26.035,
      "step": 1207
    },
    {
      "epoch": 30.06,
      "learning_rate": 8.124999999999998e-05,
      "loss": 0.3583,
      "step": 1210
    },
    {
      "epoch": 30.31,
      "learning_rate": 7.916666666666666e-05,
      "loss": 0.3356,
      "step": 1220
    },
    {
      "epoch": 30.56,
      "learning_rate": 7.708333333333331e-05,
      "loss": 0.2687,
      "step": 1230
    },
    {
      "epoch": 30.81,
      "learning_rate": 7.5e-05,
      "loss": 0.2819,
      "step": 1240
    },
    {
      "epoch": 30.98,
      "eval_accuracy": 0.8454545454545455,
      "eval_loss": 0.4227891266345978,
      "eval_runtime": 1.4114,
      "eval_samples_per_second": 389.695,
      "eval_steps_per_second": 24.799,
      "step": 1247
    },
    {
      "epoch": 31.06,
      "learning_rate": 7.291666666666666e-05,
      "loss": 0.2949,
      "step": 1250
    },
    {
      "epoch": 31.3,
      "learning_rate": 7.083333333333332e-05,
      "loss": 0.2654,
      "step": 1260
    },
    {
      "epoch": 31.55,
      "learning_rate": 6.874999999999999e-05,
      "loss": 0.2745,
      "step": 1270
    },
    {
      "epoch": 31.8,
      "learning_rate": 6.666666666666666e-05,
      "loss": 0.2704,
      "step": 1280
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.4074237644672394,
      "eval_runtime": 1.3463,
      "eval_samples_per_second": 408.52,
      "eval_steps_per_second": 25.997,
      "step": 1288
    },
    {
      "epoch": 32.05,
      "learning_rate": 6.458333333333334e-05,
      "loss": 0.3105,
      "step": 1290
    },
    {
      "epoch": 32.3,
      "learning_rate": 6.25e-05,
      "loss": 0.2614,
      "step": 1300
    },
    {
      "epoch": 32.55,
      "learning_rate": 6.041666666666666e-05,
      "loss": 0.2842,
      "step": 1310
    },
    {
      "epoch": 32.8,
      "learning_rate": 5.8333333333333326e-05,
      "loss": 0.2429,
      "step": 1320
    },
    {
      "epoch": 32.99,
      "eval_accuracy": 0.8545454545454545,
      "eval_loss": 0.4405260682106018,
      "eval_runtime": 1.6104,
      "eval_samples_per_second": 341.534,
      "eval_steps_per_second": 21.734,
      "step": 1328
    },
    {
      "epoch": 33.04,
      "learning_rate": 5.625e-05,
      "loss": 0.2956,
      "step": 1330
    },
    {
      "epoch": 33.29,
      "learning_rate": 5.4166666666666664e-05,
      "loss": 0.2767,
      "step": 1340
    },
    {
      "epoch": 33.54,
      "learning_rate": 5.208333333333333e-05,
      "loss": 0.2676,
      "step": 1350
    },
    {
      "epoch": 33.79,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 0.2421,
      "step": 1360
    },
    {
      "epoch": 33.99,
      "eval_accuracy": 0.8527272727272728,
      "eval_loss": 0.43370744585990906,
      "eval_runtime": 1.4724,
      "eval_samples_per_second": 373.551,
      "eval_steps_per_second": 23.771,
      "step": 1368
    },
    {
      "epoch": 34.04,
      "learning_rate": 4.791666666666666e-05,
      "loss": 0.3275,
      "step": 1370
    },
    {
      "epoch": 34.29,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.2475,
      "step": 1380
    },
    {
      "epoch": 34.53,
      "learning_rate": 4.375e-05,
      "loss": 0.288,
      "step": 1390
    },
    {
      "epoch": 34.78,
      "learning_rate": 4.1666666666666665e-05,
      "loss": 0.3039,
      "step": 1400
    },
    {
      "epoch": 34.98,
      "eval_accuracy": 0.8472727272727273,
      "eval_loss": 0.46279987692832947,
      "eval_runtime": 1.3718,
      "eval_samples_per_second": 400.919,
      "eval_steps_per_second": 25.513,
      "step": 1408
    },
    {
      "epoch": 35.03,
      "learning_rate": 3.958333333333333e-05,
      "loss": 0.2737,
      "step": 1410
    },
    {
      "epoch": 35.28,
      "learning_rate": 3.75e-05,
      "loss": 0.2746,
      "step": 1420
    },
    {
      "epoch": 35.53,
      "learning_rate": 3.541666666666666e-05,
      "loss": 0.2906,
      "step": 1430
    },
    {
      "epoch": 35.78,
      "learning_rate": 3.333333333333333e-05,
      "loss": 0.2677,
      "step": 1440
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.8545454545454545,
      "eval_loss": 0.4411250948905945,
      "eval_runtime": 1.3953,
      "eval_samples_per_second": 394.172,
      "eval_steps_per_second": 25.084,
      "step": 1449
    },
    {
      "epoch": 36.02,
      "learning_rate": 3.125e-05,
      "loss": 0.2467,
      "step": 1450
    },
    {
      "epoch": 36.27,
      "learning_rate": 2.9166666666666663e-05,
      "loss": 0.2513,
      "step": 1460
    },
    {
      "epoch": 36.52,
      "learning_rate": 2.7083333333333332e-05,
      "loss": 0.271,
      "step": 1470
    },
    {
      "epoch": 36.77,
      "learning_rate": 2.4999999999999998e-05,
      "loss": 0.2171,
      "step": 1480
    },
    {
      "epoch": 36.99,
      "eval_accuracy": 0.8618181818181818,
      "eval_loss": 0.4755066931247711,
      "eval_runtime": 1.4153,
      "eval_samples_per_second": 388.599,
      "eval_steps_per_second": 24.729,
      "step": 1489
    },
    {
      "epoch": 37.02,
      "learning_rate": 2.2916666666666667e-05,
      "loss": 0.2207,
      "step": 1490
    },
    {
      "epoch": 37.27,
      "learning_rate": 2.0833333333333333e-05,
      "loss": 0.2346,
      "step": 1500
    },
    {
      "epoch": 37.52,
      "learning_rate": 1.875e-05,
      "loss": 0.2055,
      "step": 1510
    },
    {
      "epoch": 37.76,
      "learning_rate": 1.6666666666666664e-05,
      "loss": 0.2268,
      "step": 1520
    },
    {
      "epoch": 37.99,
      "eval_accuracy": 0.86,
      "eval_loss": 0.45060980319976807,
      "eval_runtime": 1.4394,
      "eval_samples_per_second": 382.115,
      "eval_steps_per_second": 24.316,
      "step": 1529
    },
    {
      "epoch": 38.01,
      "learning_rate": 1.4583333333333331e-05,
      "loss": 0.2844,
      "step": 1530
    },
    {
      "epoch": 38.26,
      "learning_rate": 1.2499999999999999e-05,
      "loss": 0.2419,
      "step": 1540
    },
    {
      "epoch": 38.51,
      "learning_rate": 1.0416666666666666e-05,
      "loss": 0.256,
      "step": 1550
    },
    {
      "epoch": 38.76,
      "learning_rate": 8.333333333333332e-06,
      "loss": 0.2378,
      "step": 1560
    },
    {
      "epoch": 38.98,
      "eval_accuracy": 0.8527272727272728,
      "eval_loss": 0.4633024334907532,
      "eval_runtime": 1.3843,
      "eval_samples_per_second": 397.304,
      "eval_steps_per_second": 25.283,
      "step": 1569
    },
    {
      "epoch": 39.01,
      "learning_rate": 6.2499999999999995e-06,
      "loss": 0.3098,
      "step": 1570
    },
    {
      "epoch": 39.25,
      "learning_rate": 4.166666666666666e-06,
      "loss": 0.2416,
      "step": 1580
    },
    {
      "epoch": 39.5,
      "learning_rate": 2.083333333333333e-06,
      "loss": 0.2014,
      "step": 1590
    },
    {
      "epoch": 39.75,
      "learning_rate": 0.0,
      "loss": 0.2021,
      "step": 1600
    },
    {
      "epoch": 39.75,
      "eval_accuracy": 0.8563636363636363,
      "eval_loss": 0.4492277503013611,
      "eval_runtime": 1.3993,
      "eval_samples_per_second": 393.042,
      "eval_steps_per_second": 25.012,
      "step": 1600
    },
    {
      "epoch": 39.75,
      "step": 1600,
      "total_flos": 2.7935287357114368e+17,
      "train_loss": 0.4774747739732266,
      "train_runtime": 572.0649,
      "train_samples_per_second": 179.21,
      "train_steps_per_second": 2.797
    }
  ],
  "logging_steps": 10,
  "max_steps": 1600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 2.7935287357114368e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}