{
  "best_metric": 0.6401488780975342,
  "best_model_checkpoint": "./exper1_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": 0.0001988372093023256,
      "loss": 4.9469,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019767441860465116,
      "loss": 4.8606,
      "step": 20
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019651162790697676,
      "loss": 4.6917,
      "step": 30
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019534883720930232,
      "loss": 4.5534,
      "step": 40
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0001941860465116279,
      "loss": 4.4467,
      "step": 50
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0001930232558139535,
      "loss": 4.2795,
      "step": 60
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0001918604651162791,
      "loss": 4.2333,
      "step": 70
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019069767441860466,
      "loss": 4.099,
      "step": 80
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018953488372093025,
      "loss": 3.9539,
      "step": 90
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018837209302325584,
      "loss": 3.9352,
      "step": 100
    },
    {
      "epoch": 0.23,
      "eval_accuracy": 0.19585798816568048,
      "eval_loss": 3.8550209999084473,
      "eval_runtime": 19.1544,
      "eval_samples_per_second": 88.23,
      "eval_steps_per_second": 11.068,
      "step": 100
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0001872093023255814,
      "loss": 3.8303,
      "step": 110
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.000186046511627907,
      "loss": 3.7242,
      "step": 120
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018488372093023256,
      "loss": 3.5167,
      "step": 130
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00018372093023255815,
      "loss": 3.57,
      "step": 140
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0001825581395348837,
      "loss": 3.5231,
      "step": 150
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0001813953488372093,
      "loss": 3.4206,
      "step": 160
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0001802325581395349,
      "loss": 3.4456,
      "step": 170
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017906976744186048,
      "loss": 3.3114,
      "step": 180
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017790697674418605,
      "loss": 3.3379,
      "step": 190
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00017674418604651164,
      "loss": 3.1536,
      "step": 200
    },
    {
      "epoch": 0.47,
      "eval_accuracy": 0.28875739644970416,
      "eval_loss": 3.175450086593628,
      "eval_runtime": 21.4845,
      "eval_samples_per_second": 78.662,
      "eval_steps_per_second": 9.868,
      "step": 200
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00017558139534883723,
      "loss": 3.0307,
      "step": 210
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0001744186046511628,
      "loss": 3.0889,
      "step": 220
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017325581395348838,
      "loss": 3.0574,
      "step": 230
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017209302325581395,
      "loss": 3.1058,
      "step": 240
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017093023255813954,
      "loss": 2.9981,
      "step": 250
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0001697674418604651,
      "loss": 2.8662,
      "step": 260
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00016860465116279072,
      "loss": 2.7418,
      "step": 270
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00016744186046511629,
      "loss": 2.9494,
      "step": 280
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016639534883720932,
      "loss": 2.7704,
      "step": 290
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016523255813953488,
      "loss": 2.6937,
      "step": 300
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.4272189349112426,
      "eval_loss": 2.633167266845703,
      "eval_runtime": 21.2397,
      "eval_samples_per_second": 79.568,
      "eval_steps_per_second": 9.981,
      "step": 300
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016406976744186047,
      "loss": 2.6292,
      "step": 310
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016290697674418604,
      "loss": 2.641,
      "step": 320
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016174418604651163,
      "loss": 2.6534,
      "step": 330
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016058139534883722,
      "loss": 2.5655,
      "step": 340
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0001594186046511628,
      "loss": 2.4129,
      "step": 350
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015825581395348837,
      "loss": 2.3514,
      "step": 360
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015709302325581396,
      "loss": 2.4527,
      "step": 370
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00015593023255813955,
      "loss": 2.4411,
      "step": 380
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015476744186046512,
      "loss": 2.2639,
      "step": 390
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0001536046511627907,
      "loss": 2.3748,
      "step": 400
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.4970414201183432,
      "eval_loss": 2.28334641456604,
      "eval_runtime": 21.0937,
      "eval_samples_per_second": 80.119,
      "eval_steps_per_second": 10.05,
      "step": 400
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015244186046511627,
      "loss": 2.3598,
      "step": 410
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00015127906976744186,
      "loss": 2.3091,
      "step": 420
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00015011627906976743,
      "loss": 2.1695,
      "step": 430
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014895348837209302,
      "loss": 1.951,
      "step": 440
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0001477906976744186,
      "loss": 1.8001,
      "step": 450
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001466279069767442,
      "loss": 1.7164,
      "step": 460
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0001454651162790698,
      "loss": 1.8812,
      "step": 470
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014430232558139536,
      "loss": 1.7576,
      "step": 480
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014313953488372095,
      "loss": 1.7459,
      "step": 490
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0001419767441860465,
      "loss": 1.5575,
      "step": 500
    },
    {
      "epoch": 1.16,
      "eval_accuracy": 0.5887573964497042,
      "eval_loss": 1.8712348937988281,
      "eval_runtime": 21.4456,
      "eval_samples_per_second": 78.804,
      "eval_steps_per_second": 9.885,
      "step": 500
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0001408139534883721,
      "loss": 1.6987,
      "step": 510
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013965116279069767,
      "loss": 1.6239,
      "step": 520
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013848837209302326,
      "loss": 1.656,
      "step": 530
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013732558139534885,
      "loss": 1.43,
      "step": 540
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013616279069767444,
      "loss": 1.5609,
      "step": 550
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00013500000000000003,
      "loss": 1.5339,
      "step": 560
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0001338372093023256,
      "loss": 1.4529,
      "step": 570
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00013267441860465118,
      "loss": 1.453,
      "step": 580
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013151162790697675,
      "loss": 1.4002,
      "step": 590
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00013034883720930234,
      "loss": 1.4063,
      "step": 600
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.6313609467455621,
      "eval_loss": 1.6047611236572266,
      "eval_runtime": 21.4699,
      "eval_samples_per_second": 78.715,
      "eval_steps_per_second": 9.874,
      "step": 600
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0001291860465116279,
      "loss": 1.4063,
      "step": 610
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0001280232558139535,
      "loss": 1.3656,
      "step": 620
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012686046511627906,
      "loss": 1.2191,
      "step": 630
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012569767441860465,
      "loss": 1.3991,
      "step": 640
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012453488372093024,
      "loss": 1.2237,
      "step": 650
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012337209302325583,
      "loss": 1.1939,
      "step": 660
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012220930232558142,
      "loss": 1.2683,
      "step": 670
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012104651162790698,
      "loss": 1.2969,
      "step": 680
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00011988372093023256,
      "loss": 1.1475,
      "step": 690
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011872093023255815,
      "loss": 1.1841,
      "step": 700
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.6621301775147929,
      "eval_loss": 1.4109498262405396,
      "eval_runtime": 19.5984,
      "eval_samples_per_second": 86.231,
      "eval_steps_per_second": 10.817,
      "step": 700
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011755813953488373,
      "loss": 1.2236,
      "step": 710
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011639534883720931,
      "loss": 1.3392,
      "step": 720
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011523255813953489,
      "loss": 1.2305,
      "step": 730
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011406976744186046,
      "loss": 1.1766,
      "step": 740
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011290697674418604,
      "loss": 1.164,
      "step": 750
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011174418604651162,
      "loss": 1.2152,
      "step": 760
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011058139534883722,
      "loss": 1.0498,
      "step": 770
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0001094186046511628,
      "loss": 1.1407,
      "step": 780
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010825581395348838,
      "loss": 1.1325,
      "step": 790
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010709302325581397,
      "loss": 1.0857,
      "step": 800
    },
    {
      "epoch": 1.86,
      "eval_accuracy": 0.7112426035502959,
      "eval_loss": 1.1831614971160889,
      "eval_runtime": 19.5421,
      "eval_samples_per_second": 86.48,
      "eval_steps_per_second": 10.848,
      "step": 800
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010593023255813955,
      "loss": 0.9263,
      "step": 810
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010476744186046512,
      "loss": 0.9742,
      "step": 820
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0001036046511627907,
      "loss": 1.0176,
      "step": 830
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010244186046511628,
      "loss": 1.0105,
      "step": 840
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010127906976744185,
      "loss": 1.0131,
      "step": 850
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010011627906976743,
      "loss": 0.9204,
      "step": 860
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.895348837209302e-05,
      "loss": 0.6221,
      "step": 870
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.779069767441861e-05,
      "loss": 0.6229,
      "step": 880
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.662790697674419e-05,
      "loss": 0.7227,
      "step": 890
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.546511627906978e-05,
      "loss": 0.582,
      "step": 900
    },
    {
      "epoch": 2.09,
      "eval_accuracy": 0.7479289940828402,
      "eval_loss": 1.0371233224868774,
      "eval_runtime": 19.7303,
      "eval_samples_per_second": 85.655,
      "eval_steps_per_second": 10.745,
      "step": 900
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.430232558139536e-05,
      "loss": 0.6454,
      "step": 910
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.313953488372094e-05,
      "loss": 0.6227,
      "step": 920
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.197674418604651e-05,
      "loss": 0.4867,
      "step": 930
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.081395348837209e-05,
      "loss": 0.4923,
      "step": 940
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.965116279069767e-05,
      "loss": 0.5217,
      "step": 950
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.848837209302326e-05,
      "loss": 0.4372,
      "step": 960
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.732558139534884e-05,
      "loss": 0.567,
      "step": 970
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.616279069767443e-05,
      "loss": 0.5837,
      "step": 980
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.5e-05,
      "loss": 0.5339,
      "step": 990
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.383720930232558e-05,
      "loss": 0.5971,
      "step": 1000
    },
    {
      "epoch": 2.33,
      "eval_accuracy": 0.7461538461538462,
      "eval_loss": 0.9839160442352295,
      "eval_runtime": 19.6324,
      "eval_samples_per_second": 86.082,
      "eval_steps_per_second": 10.798,
      "step": 1000
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.267441860465117e-05,
      "loss": 0.5588,
      "step": 1010
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.151162790697675e-05,
      "loss": 0.5793,
      "step": 1020
    },
    {
      "epoch": 2.4,
      "learning_rate": 8.034883720930233e-05,
      "loss": 0.475,
      "step": 1030
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.91860465116279e-05,
      "loss": 0.5391,
      "step": 1040
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.802325581395348e-05,
      "loss": 0.487,
      "step": 1050
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.686046511627908e-05,
      "loss": 0.4474,
      "step": 1060
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.569767441860465e-05,
      "loss": 0.4753,
      "step": 1070
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.453488372093024e-05,
      "loss": 0.3932,
      "step": 1080
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.337209302325582e-05,
      "loss": 0.4907,
      "step": 1090
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.22093023255814e-05,
      "loss": 0.4617,
      "step": 1100
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 0.7656804733727811,
      "eval_loss": 0.9233230948448181,
      "eval_runtime": 19.3219,
      "eval_samples_per_second": 87.465,
      "eval_steps_per_second": 10.972,
      "step": 1100
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.104651162790698e-05,
      "loss": 0.4143,
      "step": 1110
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.988372093023257e-05,
      "loss": 0.5355,
      "step": 1120
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.872093023255814e-05,
      "loss": 0.4717,
      "step": 1130
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.755813953488372e-05,
      "loss": 0.3827,
      "step": 1140
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.63953488372093e-05,
      "loss": 0.443,
      "step": 1150
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.523255813953488e-05,
      "loss": 0.4257,
      "step": 1160
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.406976744186047e-05,
      "loss": 0.3875,
      "step": 1170
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.290697674418606e-05,
      "loss": 0.3956,
      "step": 1180
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.174418604651164e-05,
      "loss": 0.3629,
      "step": 1190
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.058139534883721e-05,
      "loss": 0.4621,
      "step": 1200
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 0.7828402366863906,
      "eval_loss": 0.8416844606399536,
      "eval_runtime": 19.2964,
      "eval_samples_per_second": 87.581,
      "eval_steps_per_second": 10.986,
      "step": 1200
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.941860465116279e-05,
      "loss": 0.4833,
      "step": 1210
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.825581395348837e-05,
      "loss": 0.3749,
      "step": 1220
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.709302325581396e-05,
      "loss": 0.4963,
      "step": 1230
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.5930232558139536e-05,
      "loss": 0.2743,
      "step": 1240
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.476744186046512e-05,
      "loss": 0.3608,
      "step": 1250
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.36046511627907e-05,
      "loss": 0.3941,
      "step": 1260
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.2441860465116275e-05,
      "loss": 0.3464,
      "step": 1270
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.1279069767441866e-05,
      "loss": 0.3461,
      "step": 1280
    },
    {
      "epoch": 3.0,
      "learning_rate": 5.011627906976745e-05,
      "loss": 0.368,
      "step": 1290
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.895348837209303e-05,
      "loss": 0.2128,
      "step": 1300
    },
    {
      "epoch": 3.02,
      "eval_accuracy": 0.7970414201183432,
      "eval_loss": 0.76435786485672,
      "eval_runtime": 19.4425,
      "eval_samples_per_second": 86.923,
      "eval_steps_per_second": 10.904,
      "step": 1300
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.7790697674418605e-05,
      "loss": 0.2484,
      "step": 1310
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.662790697674419e-05,
      "loss": 0.2213,
      "step": 1320
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.5465116279069766e-05,
      "loss": 0.2305,
      "step": 1330
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.430232558139535e-05,
      "loss": 0.208,
      "step": 1340
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.3139534883720935e-05,
      "loss": 0.2191,
      "step": 1350
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.197674418604651e-05,
      "loss": 0.2032,
      "step": 1360
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.0813953488372096e-05,
      "loss": 0.216,
      "step": 1370
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.9651162790697674e-05,
      "loss": 0.1977,
      "step": 1380
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.848837209302326e-05,
      "loss": 0.1767,
      "step": 1390
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.732558139534884e-05,
      "loss": 0.1883,
      "step": 1400
    },
    {
      "epoch": 3.26,
      "eval_accuracy": 0.8183431952662722,
      "eval_loss": 0.7001345157623291,
      "eval_runtime": 19.4125,
      "eval_samples_per_second": 87.057,
      "eval_steps_per_second": 10.921,
      "step": 1400
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.616279069767442e-05,
      "loss": 0.1768,
      "step": 1410
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.5e-05,
      "loss": 0.1436,
      "step": 1420
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.383720930232558e-05,
      "loss": 0.1781,
      "step": 1430
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.2674418604651165e-05,
      "loss": 0.1816,
      "step": 1440
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.151162790697675e-05,
      "loss": 0.1516,
      "step": 1450
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.0348837209302327e-05,
      "loss": 0.182,
      "step": 1460
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.9186046511627908e-05,
      "loss": 0.1435,
      "step": 1470
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8023255813953492e-05,
      "loss": 0.1575,
      "step": 1480
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6860465116279073e-05,
      "loss": 0.1361,
      "step": 1490
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.569767441860465e-05,
      "loss": 0.1501,
      "step": 1500
    },
    {
      "epoch": 3.49,
      "eval_accuracy": 0.8201183431952662,
      "eval_loss": 0.6825571060180664,
      "eval_runtime": 19.3331,
      "eval_samples_per_second": 87.415,
      "eval_steps_per_second": 10.966,
      "step": 1500
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.453488372093023e-05,
      "loss": 0.1719,
      "step": 1510
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.3372093023255815e-05,
      "loss": 0.2121,
      "step": 1520
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2209302325581396e-05,
      "loss": 0.1895,
      "step": 1530
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.104651162790698e-05,
      "loss": 0.1466,
      "step": 1540
    },
    {
      "epoch": 3.6,
      "learning_rate": 1.9883720930232557e-05,
      "loss": 0.2179,
      "step": 1550
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8720930232558138e-05,
      "loss": 0.1219,
      "step": 1560
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.7558139534883722e-05,
      "loss": 0.1275,
      "step": 1570
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6395348837209303e-05,
      "loss": 0.177,
      "step": 1580
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5232558139534884e-05,
      "loss": 0.1242,
      "step": 1590
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.4069767441860465e-05,
      "loss": 0.1626,
      "step": 1600
    },
    {
      "epoch": 3.72,
      "eval_accuracy": 0.8254437869822485,
      "eval_loss": 0.6568493843078613,
      "eval_runtime": 19.3691,
      "eval_samples_per_second": 87.253,
      "eval_steps_per_second": 10.945,
      "step": 1600
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.2906976744186047e-05,
      "loss": 0.1078,
      "step": 1610
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1744186046511628e-05,
      "loss": 0.1073,
      "step": 1620
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.058139534883721e-05,
      "loss": 0.106,
      "step": 1630
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.418604651162791e-06,
      "loss": 0.1707,
      "step": 1640
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.255813953488372e-06,
      "loss": 0.1292,
      "step": 1650
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.093023255813954e-06,
      "loss": 0.1311,
      "step": 1660
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.930232558139535e-06,
      "loss": 0.1477,
      "step": 1670
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.767441860465117e-06,
      "loss": 0.1223,
      "step": 1680
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.604651162790698e-06,
      "loss": 0.1227,
      "step": 1690
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.4418604651162793e-06,
      "loss": 0.1053,
      "step": 1700
    },
    {
      "epoch": 3.95,
      "eval_accuracy": 0.827810650887574,
      "eval_loss": 0.6401488780975342,
      "eval_runtime": 19.3377,
      "eval_samples_per_second": 87.394,
      "eval_steps_per_second": 10.963,
      "step": 1700
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.2790697674418605e-06,
      "loss": 0.1335,
      "step": 1710
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.1627906976744187e-07,
      "loss": 0.0918,
      "step": 1720
    },
    {
      "epoch": 4.0,
      "step": 1720,
      "total_flos": 2.1353892586861363e+18,
      "train_loss": 1.3106378566387087,
      "train_runtime": 1021.3737,
      "train_samples_per_second": 26.944,
      "train_steps_per_second": 1.684
    }
  ],
  "max_steps": 1720,
  "num_train_epochs": 4,
  "total_flos": 2.1353892586861363e+18,
  "trial_name": null,
  "trial_params": null
}