{
  "best_metric": 3.4589149951934814,
  "best_model_checkpoint": "./exper2_mesum5/checkpoint-1700",
  "epoch": 4.0,
  "global_step": 1720,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 0.0019883720930232557,
      "loss": 4.9453,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0019767441860465114,
      "loss": 4.9054,
      "step": 20
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0019651162790697676,
      "loss": 4.6079,
      "step": 30
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0019534883720930232,
      "loss": 4.6129,
      "step": 40
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.001941860465116279,
      "loss": 4.613,
      "step": 50
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.001930232558139535,
      "loss": 4.3921,
      "step": 60
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0019186046511627908,
      "loss": 4.459,
      "step": 70
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0019069767441860467,
      "loss": 4.3617,
      "step": 80
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0018953488372093024,
      "loss": 4.3242,
      "step": 90
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0018837209302325583,
      "loss": 4.4265,
      "step": 100
    },
    {
      "epoch": 0.23,
      "eval_accuracy": 0.029585798816568046,
      "eval_loss": 4.367562294006348,
      "eval_runtime": 19.0014,
      "eval_samples_per_second": 88.941,
      "eval_steps_per_second": 11.157,
      "step": 100
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.001872093023255814,
      "loss": 4.3222,
      "step": 110
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0018604651162790699,
      "loss": 4.2917,
      "step": 120
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0018488372093023256,
      "loss": 4.2636,
      "step": 130
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0018372093023255815,
      "loss": 4.2516,
      "step": 140
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0018255813953488372,
      "loss": 4.2823,
      "step": 150
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0018139534883720929,
      "loss": 4.1987,
      "step": 160
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.001802325581395349,
      "loss": 4.3236,
      "step": 170
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0017906976744186047,
      "loss": 4.359,
      "step": 180
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0017790697674418606,
      "loss": 4.2877,
      "step": 190
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0017674418604651163,
      "loss": 4.1144,
      "step": 200
    },
    {
      "epoch": 0.47,
      "eval_accuracy": 0.054437869822485205,
      "eval_loss": 4.160641670227051,
      "eval_runtime": 21.5399,
      "eval_samples_per_second": 78.459,
      "eval_steps_per_second": 9.842,
      "step": 200
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0017558139534883722,
      "loss": 4.1521,
      "step": 210
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0017441860465116279,
      "loss": 4.1644,
      "step": 220
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0017325581395348838,
      "loss": 4.1735,
      "step": 230
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0017209302325581395,
      "loss": 4.2105,
      "step": 240
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0017093023255813954,
      "loss": 4.1582,
      "step": 250
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.001697674418604651,
      "loss": 4.1161,
      "step": 260
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0016860465116279072,
      "loss": 4.0732,
      "step": 270
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.001674418604651163,
      "loss": 4.1367,
      "step": 280
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0016627906976744186,
      "loss": 4.2132,
      "step": 290
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0016511627906976745,
      "loss": 4.0912,
      "step": 300
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.05088757396449704,
      "eval_loss": 4.107085704803467,
      "eval_runtime": 21.8387,
      "eval_samples_per_second": 77.386,
      "eval_steps_per_second": 9.708,
      "step": 300
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0016395348837209302,
      "loss": 4.1099,
      "step": 310
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0016279069767441861,
      "loss": 4.0904,
      "step": 320
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0016162790697674418,
      "loss": 4.1553,
      "step": 330
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0016046511627906977,
      "loss": 4.0868,
      "step": 340
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0015930232558139534,
      "loss": 4.0147,
      "step": 350
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0015813953488372093,
      "loss": 4.0375,
      "step": 360
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.001569767441860465,
      "loss": 4.1162,
      "step": 370
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0015581395348837211,
      "loss": 4.1113,
      "step": 380
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0015465116279069768,
      "loss": 3.9,
      "step": 390
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0015348837209302327,
      "loss": 4.0361,
      "step": 400
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.06686390532544378,
      "eval_loss": 4.062524318695068,
      "eval_runtime": 21.795,
      "eval_samples_per_second": 77.541,
      "eval_steps_per_second": 9.727,
      "step": 400
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0015232558139534884,
      "loss": 4.0367,
      "step": 410
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0015116279069767443,
      "loss": 4.1145,
      "step": 420
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0015,
      "loss": 4.0218,
      "step": 430
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0014883720930232557,
      "loss": 3.9887,
      "step": 440
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0014767441860465116,
      "loss": 3.9717,
      "step": 450
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0014651162790697673,
      "loss": 3.9346,
      "step": 460
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0014534883720930232,
      "loss": 4.0204,
      "step": 470
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.001441860465116279,
      "loss": 3.9421,
      "step": 480
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.001430232558139535,
      "loss": 3.9545,
      "step": 490
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0014186046511627907,
      "loss": 4.0257,
      "step": 500
    },
    {
      "epoch": 1.16,
      "eval_accuracy": 0.08224852071005917,
      "eval_loss": 3.9682300090789795,
      "eval_runtime": 22.2982,
      "eval_samples_per_second": 75.791,
      "eval_steps_per_second": 9.507,
      "step": 500
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0014069767441860467,
      "loss": 3.9265,
      "step": 510
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0013953488372093023,
      "loss": 3.8827,
      "step": 520
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0013837209302325583,
      "loss": 4.0043,
      "step": 530
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.001372093023255814,
      "loss": 3.9064,
      "step": 540
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0013604651162790699,
      "loss": 3.8487,
      "step": 550
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0013488372093023256,
      "loss": 3.872,
      "step": 560
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0013372093023255815,
      "loss": 4.0328,
      "step": 570
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0013255813953488372,
      "loss": 4.0341,
      "step": 580
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0013139534883720928,
      "loss": 3.88,
      "step": 590
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.001302325581395349,
      "loss": 3.8846,
      "step": 600
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.0834319526627219,
      "eval_loss": 3.9310760498046875,
      "eval_runtime": 22.1502,
      "eval_samples_per_second": 76.297,
      "eval_steps_per_second": 9.571,
      "step": 600
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0012906976744186047,
      "loss": 4.0106,
      "step": 610
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0012790697674418606,
      "loss": 3.9745,
      "step": 620
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0012674418604651163,
      "loss": 3.9328,
      "step": 630
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0012558139534883722,
      "loss": 3.9979,
      "step": 640
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0012441860465116279,
      "loss": 3.9394,
      "step": 650
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0012325581395348838,
      "loss": 3.8107,
      "step": 660
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0012209302325581395,
      "loss": 3.8761,
      "step": 670
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0012093023255813954,
      "loss": 4.0095,
      "step": 680
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.001197674418604651,
      "loss": 3.9969,
      "step": 690
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0011860465116279072,
      "loss": 3.9504,
      "step": 700
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.06982248520710059,
      "eval_loss": 3.9254539012908936,
      "eval_runtime": 19.9581,
      "eval_samples_per_second": 84.677,
      "eval_steps_per_second": 10.622,
      "step": 700
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.001174418604651163,
      "loss": 3.9426,
      "step": 710
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0011627906976744186,
      "loss": 3.9817,
      "step": 720
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0011511627906976745,
      "loss": 4.0043,
      "step": 730
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0011395348837209302,
      "loss": 3.9829,
      "step": 740
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.001127906976744186,
      "loss": 3.9204,
      "step": 750
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0011162790697674418,
      "loss": 3.958,
      "step": 760
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0011046511627906977,
      "loss": 3.9127,
      "step": 770
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0010930232558139534,
      "loss": 3.8387,
      "step": 780
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0010813953488372093,
      "loss": 3.8673,
      "step": 790
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.001069767441860465,
      "loss": 3.9884,
      "step": 800
    },
    {
      "epoch": 1.86,
      "eval_accuracy": 0.07218934911242604,
      "eval_loss": 3.940431594848633,
      "eval_runtime": 19.188,
      "eval_samples_per_second": 88.076,
      "eval_steps_per_second": 11.049,
      "step": 800
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0010581395348837211,
      "loss": 3.8334,
      "step": 810
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0010465116279069768,
      "loss": 3.9347,
      "step": 820
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0010348837209302327,
      "loss": 3.9363,
      "step": 830
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0010232558139534884,
      "loss": 3.8543,
      "step": 840
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0010116279069767443,
      "loss": 3.94,
      "step": 850
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.001,
      "loss": 3.7869,
      "step": 860
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0009883720930232557,
      "loss": 3.8011,
      "step": 870
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0009767441860465116,
      "loss": 3.7155,
      "step": 880
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0009651162790697675,
      "loss": 3.909,
      "step": 890
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0009534883720930233,
      "loss": 3.7191,
      "step": 900
    },
    {
      "epoch": 2.09,
      "eval_accuracy": 0.09349112426035502,
      "eval_loss": 3.826247215270996,
      "eval_runtime": 20.1595,
      "eval_samples_per_second": 83.831,
      "eval_steps_per_second": 10.516,
      "step": 900
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0009418604651162791,
      "loss": 3.8592,
      "step": 910
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0009302325581395349,
      "loss": 3.8276,
      "step": 920
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0009186046511627907,
      "loss": 3.7752,
      "step": 930
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0009069767441860464,
      "loss": 3.7793,
      "step": 940
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0008953488372093023,
      "loss": 3.7912,
      "step": 950
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0008837209302325581,
      "loss": 3.6608,
      "step": 960
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0008720930232558139,
      "loss": 3.8321,
      "step": 970
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0008604651162790697,
      "loss": 3.6955,
      "step": 980
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0008488372093023255,
      "loss": 3.7179,
      "step": 990
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0008372093023255815,
      "loss": 3.7952,
      "step": 1000
    },
    {
      "epoch": 2.33,
      "eval_accuracy": 0.07337278106508875,
      "eval_loss": 3.823629379272461,
      "eval_runtime": 19.0027,
      "eval_samples_per_second": 88.935,
      "eval_steps_per_second": 11.156,
      "step": 1000
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.0008255813953488373,
      "loss": 3.6852,
      "step": 1010
    },
    {
      "epoch": 2.37,
      "learning_rate": 0.0008139534883720931,
      "loss": 3.8711,
      "step": 1020
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.0008023255813953489,
      "loss": 3.7056,
      "step": 1030
    },
    {
      "epoch": 2.42,
      "learning_rate": 0.0007906976744186047,
      "loss": 3.8829,
      "step": 1040
    },
    {
      "epoch": 2.44,
      "learning_rate": 0.0007790697674418606,
      "loss": 3.7857,
      "step": 1050
    },
    {
      "epoch": 2.47,
      "learning_rate": 0.0007674418604651164,
      "loss": 3.7401,
      "step": 1060
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.0007558139534883722,
      "loss": 3.7043,
      "step": 1070
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.0007441860465116279,
      "loss": 3.6616,
      "step": 1080
    },
    {
      "epoch": 2.53,
      "learning_rate": 0.0007325581395348837,
      "loss": 3.7174,
      "step": 1090
    },
    {
      "epoch": 2.56,
      "learning_rate": 0.0007209302325581395,
      "loss": 3.8085,
      "step": 1100
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 0.09644970414201183,
      "eval_loss": 3.7693703174591064,
      "eval_runtime": 20.1388,
      "eval_samples_per_second": 83.918,
      "eval_steps_per_second": 10.527,
      "step": 1100
    },
    {
      "epoch": 2.58,
      "learning_rate": 0.0007093023255813954,
      "loss": 3.6466,
      "step": 1110
    },
    {
      "epoch": 2.6,
      "learning_rate": 0.0006976744186046512,
      "loss": 3.8068,
      "step": 1120
    },
    {
      "epoch": 2.63,
      "learning_rate": 0.000686046511627907,
      "loss": 3.6777,
      "step": 1130
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.0006744186046511628,
      "loss": 3.5985,
      "step": 1140
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0006627906976744186,
      "loss": 3.7016,
      "step": 1150
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.0006511627906976745,
      "loss": 3.7256,
      "step": 1160
    },
    {
      "epoch": 2.72,
      "learning_rate": 0.0006395348837209303,
      "loss": 3.6213,
      "step": 1170
    },
    {
      "epoch": 2.74,
      "learning_rate": 0.0006279069767441861,
      "loss": 3.5336,
      "step": 1180
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0006162790697674419,
      "loss": 3.6883,
      "step": 1190
    },
    {
      "epoch": 2.79,
      "learning_rate": 0.0006046511627906977,
      "loss": 3.7535,
      "step": 1200
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 0.1059171597633136,
      "eval_loss": 3.675692319869995,
      "eval_runtime": 19.295,
      "eval_samples_per_second": 87.587,
      "eval_steps_per_second": 10.987,
      "step": 1200
    },
    {
      "epoch": 2.81,
      "learning_rate": 0.0005930232558139536,
      "loss": 3.6236,
      "step": 1210
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0005813953488372093,
      "loss": 3.5595,
      "step": 1220
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.0005697674418604651,
      "loss": 3.6437,
      "step": 1230
    },
    {
      "epoch": 2.88,
      "learning_rate": 0.0005581395348837209,
      "loss": 3.6395,
      "step": 1240
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0005465116279069767,
      "loss": 3.7838,
      "step": 1250
    },
    {
      "epoch": 2.93,
      "learning_rate": 0.0005348837209302325,
      "loss": 3.6164,
      "step": 1260
    },
    {
      "epoch": 2.95,
      "learning_rate": 0.0005232558139534884,
      "loss": 3.6732,
      "step": 1270
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0005116279069767442,
      "loss": 3.5979,
      "step": 1280
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0005,
      "loss": 3.6063,
      "step": 1290
    },
    {
      "epoch": 3.02,
      "learning_rate": 0.0004883720930232558,
      "loss": 3.4218,
      "step": 1300
    },
    {
      "epoch": 3.02,
      "eval_accuracy": 0.10946745562130178,
      "eval_loss": 3.6474382877349854,
      "eval_runtime": 20.4645,
      "eval_samples_per_second": 82.582,
      "eval_steps_per_second": 10.359,
      "step": 1300
    },
    {
      "epoch": 3.05,
      "learning_rate": 0.00047674418604651167,
      "loss": 3.517,
      "step": 1310
    },
    {
      "epoch": 3.07,
      "learning_rate": 0.00046511627906976747,
      "loss": 3.5429,
      "step": 1320
    },
    {
      "epoch": 3.09,
      "learning_rate": 0.0004534883720930232,
      "loss": 3.6631,
      "step": 1330
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.00044186046511627907,
      "loss": 3.6517,
      "step": 1340
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.00043023255813953487,
      "loss": 3.5048,
      "step": 1350
    },
    {
      "epoch": 3.16,
      "learning_rate": 0.0004186046511627907,
      "loss": 3.4971,
      "step": 1360
    },
    {
      "epoch": 3.19,
      "learning_rate": 0.00040697674418604653,
      "loss": 3.4994,
      "step": 1370
    },
    {
      "epoch": 3.21,
      "learning_rate": 0.00039534883720930233,
      "loss": 3.6071,
      "step": 1380
    },
    {
      "epoch": 3.23,
      "learning_rate": 0.0003837209302325582,
      "loss": 3.573,
      "step": 1390
    },
    {
      "epoch": 3.26,
      "learning_rate": 0.00037209302325581393,
      "loss": 3.5172,
      "step": 1400
    },
    {
      "epoch": 3.26,
      "eval_accuracy": 0.1165680473372781,
      "eval_loss": 3.5621113777160645,
      "eval_runtime": 19.4458,
      "eval_samples_per_second": 86.908,
      "eval_steps_per_second": 10.902,
      "step": 1400
    },
    {
      "epoch": 3.28,
      "learning_rate": 0.00036046511627906973,
      "loss": 3.4566,
      "step": 1410
    },
    {
      "epoch": 3.3,
      "learning_rate": 0.0003488372093023256,
      "loss": 3.4021,
      "step": 1420
    },
    {
      "epoch": 3.33,
      "learning_rate": 0.0003372093023255814,
      "loss": 3.4908,
      "step": 1430
    },
    {
      "epoch": 3.35,
      "learning_rate": 0.00032558139534883724,
      "loss": 3.4437,
      "step": 1440
    },
    {
      "epoch": 3.37,
      "learning_rate": 0.00031395348837209304,
      "loss": 3.4836,
      "step": 1450
    },
    {
      "epoch": 3.4,
      "learning_rate": 0.00030232558139534885,
      "loss": 3.5235,
      "step": 1460
    },
    {
      "epoch": 3.42,
      "learning_rate": 0.00029069767441860465,
      "loss": 3.5462,
      "step": 1470
    },
    {
      "epoch": 3.44,
      "learning_rate": 0.00027906976744186045,
      "loss": 3.5786,
      "step": 1480
    },
    {
      "epoch": 3.47,
      "learning_rate": 0.00026744186046511625,
      "loss": 3.4383,
      "step": 1490
    },
    {
      "epoch": 3.49,
      "learning_rate": 0.0002558139534883721,
      "loss": 3.5173,
      "step": 1500
    },
    {
      "epoch": 3.49,
      "eval_accuracy": 0.12071005917159763,
      "eval_loss": 3.557870626449585,
      "eval_runtime": 21.2599,
      "eval_samples_per_second": 79.492,
      "eval_steps_per_second": 9.972,
      "step": 1500
    },
    {
      "epoch": 3.51,
      "learning_rate": 0.0002441860465116279,
      "loss": 3.3559,
      "step": 1510
    },
    {
      "epoch": 3.53,
      "learning_rate": 0.00023255813953488373,
      "loss": 3.5068,
      "step": 1520
    },
    {
      "epoch": 3.56,
      "learning_rate": 0.00022093023255813954,
      "loss": 3.2757,
      "step": 1530
    },
    {
      "epoch": 3.58,
      "learning_rate": 0.00020930232558139536,
      "loss": 3.5227,
      "step": 1540
    },
    {
      "epoch": 3.6,
      "learning_rate": 0.00019767441860465116,
      "loss": 3.361,
      "step": 1550
    },
    {
      "epoch": 3.63,
      "learning_rate": 0.00018604651162790697,
      "loss": 3.327,
      "step": 1560
    },
    {
      "epoch": 3.65,
      "learning_rate": 0.0001744186046511628,
      "loss": 3.5469,
      "step": 1570
    },
    {
      "epoch": 3.67,
      "learning_rate": 0.00016279069767441862,
      "loss": 3.4543,
      "step": 1580
    },
    {
      "epoch": 3.7,
      "learning_rate": 0.00015116279069767442,
      "loss": 3.3873,
      "step": 1590
    },
    {
      "epoch": 3.72,
      "learning_rate": 0.00013953488372093022,
      "loss": 3.4346,
      "step": 1600
    },
    {
      "epoch": 3.72,
      "eval_accuracy": 0.12485207100591716,
      "eval_loss": 3.4816746711730957,
      "eval_runtime": 19.358,
      "eval_samples_per_second": 87.303,
      "eval_steps_per_second": 10.952,
      "step": 1600
    },
    {
      "epoch": 3.74,
      "learning_rate": 0.00012790697674418605,
      "loss": 3.4285,
      "step": 1610
    },
    {
      "epoch": 3.77,
      "learning_rate": 0.00011627906976744187,
      "loss": 3.423,
      "step": 1620
    },
    {
      "epoch": 3.79,
      "learning_rate": 0.00010465116279069768,
      "loss": 3.3739,
      "step": 1630
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.302325581395348e-05,
      "loss": 3.3856,
      "step": 1640
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.139534883720931e-05,
      "loss": 3.497,
      "step": 1650
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.976744186046511e-05,
      "loss": 3.4255,
      "step": 1660
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.8139534883720933e-05,
      "loss": 3.3663,
      "step": 1670
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.651162790697674e-05,
      "loss": 3.3245,
      "step": 1680
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.4883720930232556e-05,
      "loss": 3.4052,
      "step": 1690
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.325581395348837e-05,
      "loss": 3.3995,
      "step": 1700
    },
    {
      "epoch": 3.95,
      "eval_accuracy": 0.13076923076923078,
      "eval_loss": 3.4589149951934814,
      "eval_runtime": 20.5788,
      "eval_samples_per_second": 82.123,
      "eval_steps_per_second": 10.302,
      "step": 1700
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.1627906976744185e-05,
      "loss": 3.3319,
      "step": 1710
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.0,
      "loss": 3.3026,
      "step": 1720
    },
    {
      "epoch": 4.0,
      "step": 1720,
      "total_flos": 2.1353892586861363e+18,
      "train_loss": 3.8407227360925007,
      "train_runtime": 1045.5226,
      "train_samples_per_second": 26.322,
      "train_steps_per_second": 1.645
    }
  ],
  "max_steps": 1720,
  "num_train_epochs": 4,
  "total_flos": 2.1353892586861363e+18,
  "trial_name": null,
  "trial_params": null
}