{
  "best_metric": 0.26652154326438904,
  "best_model_checkpoint": "./vit-base-binary-isic-patch-16/checkpoint-800",
  "epoch": 4.0,
  "global_step": 4304,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.0001995353159851301,
      "loss": 0.5547,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00019907063197026022,
      "loss": 0.4747,
      "step": 20
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019860594795539035,
      "loss": 0.3674,
      "step": 30
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019814126394052047,
      "loss": 0.3681,
      "step": 40
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019767657992565056,
      "loss": 0.3661,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019721189591078068,
      "loss": 0.4179,
      "step": 60
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019674721189591078,
      "loss": 0.3013,
      "step": 70
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019628252788104092,
      "loss": 0.4635,
      "step": 80
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019581784386617102,
      "loss": 0.4503,
      "step": 90
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0001953531598513011,
      "loss": 0.3887,
      "step": 100
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.8539210661199385,
      "eval_f1": 0.8539210661199385,
      "eval_loss": 0.32007861137390137,
      "eval_precision": 0.8539210661199385,
      "eval_recall": 0.8539210661199385,
      "eval_runtime": 65.3823,
      "eval_samples_per_second": 119.36,
      "eval_steps_per_second": 14.928,
      "step": 100
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019488847583643123,
      "loss": 0.3623,
      "step": 110
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00019442379182156135,
      "loss": 0.3477,
      "step": 120
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019395910780669148,
      "loss": 0.2489,
      "step": 130
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019349442379182157,
      "loss": 0.3897,
      "step": 140
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0001930297397769517,
      "loss": 0.2382,
      "step": 150
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019256505576208178,
      "loss": 0.2525,
      "step": 160
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0001921003717472119,
      "loss": 0.3432,
      "step": 170
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019163568773234203,
      "loss": 0.2406,
      "step": 180
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019117100371747212,
      "loss": 0.2909,
      "step": 190
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019070631970260224,
      "loss": 0.2725,
      "step": 200
    },
    {
      "epoch": 0.19,
      "eval_accuracy": 0.8526396719630959,
      "eval_f1": 0.8526396719630959,
      "eval_loss": 0.32080867886543274,
      "eval_precision": 0.8526396719630959,
      "eval_recall": 0.8526396719630959,
      "eval_runtime": 65.5227,
      "eval_samples_per_second": 119.104,
      "eval_steps_per_second": 14.896,
      "step": 200
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00019024163568773234,
      "loss": 0.265,
      "step": 210
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018977695167286246,
      "loss": 0.3581,
      "step": 220
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018931226765799258,
      "loss": 0.249,
      "step": 230
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0001888475836431227,
      "loss": 0.2204,
      "step": 240
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0001883828996282528,
      "loss": 0.2033,
      "step": 250
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018791821561338291,
      "loss": 0.3672,
      "step": 260
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.000187453531598513,
      "loss": 0.3686,
      "step": 270
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00018698884758364313,
      "loss": 0.3134,
      "step": 280
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018652416356877325,
      "loss": 0.4176,
      "step": 290
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018605947955390334,
      "loss": 0.2653,
      "step": 300
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.8531522296258329,
      "eval_f1": 0.8531522296258329,
      "eval_loss": 0.31128329038619995,
      "eval_precision": 0.8531522296258329,
      "eval_recall": 0.8531522296258329,
      "eval_runtime": 64.8428,
      "eval_samples_per_second": 120.353,
      "eval_steps_per_second": 15.052,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018559479553903347,
      "loss": 0.2225,
      "step": 310
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018513011152416359,
      "loss": 0.2297,
      "step": 320
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018471189591078068,
      "loss": 0.2706,
      "step": 330
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0001842472118959108,
      "loss": 0.3019,
      "step": 340
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00018378252788104092,
      "loss": 0.3417,
      "step": 350
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.000183317843866171,
      "loss": 0.2706,
      "step": 360
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0001828531598513011,
      "loss": 0.2599,
      "step": 370
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018238847583643123,
      "loss": 0.2019,
      "step": 380
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018192379182156135,
      "loss": 0.2046,
      "step": 390
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00018145910780669147,
      "loss": 0.2439,
      "step": 400
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.8749359302921579,
      "eval_f1": 0.8749359302921579,
      "eval_loss": 0.3156268298625946,
      "eval_precision": 0.8749359302921579,
      "eval_recall": 0.8749359302921579,
      "eval_runtime": 65.9023,
      "eval_samples_per_second": 118.418,
      "eval_steps_per_second": 14.81,
      "step": 400
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018099442379182156,
      "loss": 0.232,
      "step": 410
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018052973977695168,
      "loss": 0.2763,
      "step": 420
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00018006505576208178,
      "loss": 0.2423,
      "step": 430
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017960037174721193,
      "loss": 0.1792,
      "step": 440
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017913568773234202,
      "loss": 0.3386,
      "step": 450
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00017867100371747211,
      "loss": 0.2346,
      "step": 460
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017820631970260224,
      "loss": 0.2047,
      "step": 470
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017774163568773236,
      "loss": 0.2911,
      "step": 480
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017727695167286248,
      "loss": 0.2839,
      "step": 490
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017681226765799257,
      "loss": 0.2183,
      "step": 500
    },
    {
      "epoch": 0.46,
      "eval_accuracy": 0.8390568939005638,
      "eval_f1": 0.8390568939005638,
      "eval_loss": 0.3145322799682617,
      "eval_precision": 0.8390568939005638,
      "eval_recall": 0.8390568939005638,
      "eval_runtime": 65.806,
      "eval_samples_per_second": 118.591,
      "eval_steps_per_second": 14.831,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0001763475836431227,
      "loss": 0.1713,
      "step": 510
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0001758828996282528,
      "loss": 0.2892,
      "step": 520
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0001754182156133829,
      "loss": 0.2889,
      "step": 530
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00017495353159851303,
      "loss": 0.1804,
      "step": 540
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00017448884758364312,
      "loss": 0.2312,
      "step": 550
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00017402416356877324,
      "loss": 0.1858,
      "step": 560
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017355947955390334,
      "loss": 0.1769,
      "step": 570
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017309479553903346,
      "loss": 0.1604,
      "step": 580
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00017263011152416358,
      "loss": 0.2147,
      "step": 590
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0001721654275092937,
      "loss": 0.2233,
      "step": 600
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 0.8543054843669913,
      "eval_f1": 0.8543054843669913,
      "eval_loss": 0.3082994520664215,
      "eval_precision": 0.8543054843669913,
      "eval_recall": 0.8543054843669913,
      "eval_runtime": 66.1562,
      "eval_samples_per_second": 117.963,
      "eval_steps_per_second": 14.753,
      "step": 600
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0001717007434944238,
      "loss": 0.1553,
      "step": 610
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0001712360594795539,
      "loss": 0.205,
      "step": 620
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017077137546468404,
      "loss": 0.202,
      "step": 630
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017030669144981413,
      "loss": 0.2538,
      "step": 640
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00016984200743494425,
      "loss": 0.2256,
      "step": 650
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016937732342007435,
      "loss": 0.2723,
      "step": 660
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00016891263940520447,
      "loss": 0.2022,
      "step": 670
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0001684479553903346,
      "loss": 0.2223,
      "step": 680
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0001679832713754647,
      "loss": 0.1468,
      "step": 690
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0001675185873605948,
      "loss": 0.2234,
      "step": 700
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.8598154792414147,
      "eval_f1": 0.8598154792414147,
      "eval_loss": 0.37567147612571716,
      "eval_precision": 0.8598154792414147,
      "eval_recall": 0.8598154792414147,
      "eval_runtime": 66.048,
      "eval_samples_per_second": 118.156,
      "eval_steps_per_second": 14.777,
      "step": 700
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0001670539033457249,
      "loss": 0.3206,
      "step": 710
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016658921933085502,
      "loss": 0.2027,
      "step": 720
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00016612453531598514,
      "loss": 0.2562,
      "step": 730
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016565985130111526,
      "loss": 0.2237,
      "step": 740
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016519516728624536,
      "loss": 0.1546,
      "step": 750
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016473048327137548,
      "loss": 0.1116,
      "step": 760
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016426579925650557,
      "loss": 0.2155,
      "step": 770
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016380111524163572,
      "loss": 0.3059,
      "step": 780
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0001633364312267658,
      "loss": 0.225,
      "step": 790
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0001628717472118959,
      "loss": 0.1825,
      "step": 800
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 0.8799333675038442,
      "eval_f1": 0.8799333675038442,
      "eval_loss": 0.26652154326438904,
      "eval_precision": 0.8799333675038442,
      "eval_recall": 0.8799333675038442,
      "eval_runtime": 65.7811,
      "eval_samples_per_second": 118.636,
      "eval_steps_per_second": 14.837,
      "step": 800
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00016240706319702603,
      "loss": 0.1389,
      "step": 810
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016194237918215612,
      "loss": 0.2437,
      "step": 820
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016147769516728627,
      "loss": 0.2605,
      "step": 830
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016101301115241636,
      "loss": 0.1945,
      "step": 840
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016054832713754648,
      "loss": 0.2126,
      "step": 850
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00016008364312267658,
      "loss": 0.1279,
      "step": 860
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0001596189591078067,
      "loss": 0.2797,
      "step": 870
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00015915427509293682,
      "loss": 0.1624,
      "step": 880
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015868959107806691,
      "loss": 0.2071,
      "step": 890
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015822490706319704,
      "loss": 0.1244,
      "step": 900
    },
    {
      "epoch": 0.84,
      "eval_accuracy": 0.8886468477703742,
      "eval_f1": 0.8886468477703742,
      "eval_loss": 0.30883264541625977,
      "eval_precision": 0.8886468477703742,
      "eval_recall": 0.8886468477703742,
      "eval_runtime": 66.22,
      "eval_samples_per_second": 117.85,
      "eval_steps_per_second": 14.739,
      "step": 900
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00015776022304832713,
      "loss": 0.1724,
      "step": 910
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015729553903345725,
      "loss": 0.2183,
      "step": 920
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015683085501858737,
      "loss": 0.1506,
      "step": 930
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0001563661710037175,
      "loss": 0.3948,
      "step": 940
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0001559014869888476,
      "loss": 0.1716,
      "step": 950
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0001554368029739777,
      "loss": 0.2014,
      "step": 960
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0001549721189591078,
      "loss": 0.1122,
      "step": 970
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015450743494423792,
      "loss": 0.1644,
      "step": 980
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015404275092936804,
      "loss": 0.1991,
      "step": 990
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00015357806691449814,
      "loss": 0.1454,
      "step": 1000
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.8544336237826755,
      "eval_f1": 0.8544336237826755,
      "eval_loss": 0.4594305157661438,
      "eval_precision": 0.8544336237826755,
      "eval_recall": 0.8544336237826755,
      "eval_runtime": 66.4924,
      "eval_samples_per_second": 117.367,
      "eval_steps_per_second": 14.678,
      "step": 1000
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015311338289962826,
      "loss": 0.1785,
      "step": 1010
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015264869888475838,
      "loss": 0.1385,
      "step": 1020
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0001521840148698885,
      "loss": 0.1649,
      "step": 1030
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0001517193308550186,
      "loss": 0.1453,
      "step": 1040
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00015125464684014872,
      "loss": 0.1751,
      "step": 1050
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0001507899628252788,
      "loss": 0.1477,
      "step": 1060
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015032527881040893,
      "loss": 0.1723,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00014986059479553905,
      "loss": 0.1179,
      "step": 1080
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014939591078066915,
      "loss": 0.101,
      "step": 1090
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014893122676579927,
      "loss": 0.043,
      "step": 1100
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.8819835981547924,
      "eval_f1": 0.8819835981547925,
      "eval_loss": 0.34242600202560425,
      "eval_precision": 0.8819835981547924,
      "eval_recall": 0.8819835981547924,
      "eval_runtime": 65.9295,
      "eval_samples_per_second": 118.369,
      "eval_steps_per_second": 14.804,
      "step": 1100
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00014846654275092936,
      "loss": 0.0525,
      "step": 1110
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014800185873605948,
      "loss": 0.0857,
      "step": 1120
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0001475371747211896,
      "loss": 0.0544,
      "step": 1130
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0001470724907063197,
      "loss": 0.125,
      "step": 1140
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00014660780669144982,
      "loss": 0.1786,
      "step": 1150
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0001461431226765799,
      "loss": 0.1693,
      "step": 1160
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014567843866171006,
      "loss": 0.0443,
      "step": 1170
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00014521375464684016,
      "loss": 0.102,
      "step": 1180
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014474907063197028,
      "loss": 0.1092,
      "step": 1190
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014428438661710037,
      "loss": 0.1911,
      "step": 1200
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.8630189646335212,
      "eval_f1": 0.8630189646335212,
      "eval_loss": 0.4267546236515045,
      "eval_precision": 0.8630189646335212,
      "eval_recall": 0.8630189646335212,
      "eval_runtime": 66.425,
      "eval_samples_per_second": 117.486,
      "eval_steps_per_second": 14.693,
      "step": 1200
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0001438197026022305,
      "loss": 0.165,
      "step": 1210
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0001433550185873606,
      "loss": 0.1083,
      "step": 1220
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0001428903345724907,
      "loss": 0.1184,
      "step": 1230
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00014242565055762083,
      "loss": 0.1208,
      "step": 1240
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014196096654275092,
      "loss": 0.1142,
      "step": 1250
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00014149628252788104,
      "loss": 0.0976,
      "step": 1260
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014103159851301116,
      "loss": 0.087,
      "step": 1270
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014056691449814129,
      "loss": 0.1507,
      "step": 1280
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00014010223048327138,
      "loss": 0.1603,
      "step": 1290
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0001396375464684015,
      "loss": 0.1091,
      "step": 1300
    },
    {
      "epoch": 1.21,
      "eval_accuracy": 0.877883136852896,
      "eval_f1": 0.877883136852896,
      "eval_loss": 0.29402631521224976,
      "eval_precision": 0.877883136852896,
      "eval_recall": 0.877883136852896,
      "eval_runtime": 65.9601,
      "eval_samples_per_second": 118.314,
      "eval_steps_per_second": 14.797,
      "step": 1300
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001391728624535316,
      "loss": 0.0846,
      "step": 1310
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013870817843866172,
      "loss": 0.0417,
      "step": 1320
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013824349442379184,
      "loss": 0.1398,
      "step": 1330
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013777881040892193,
      "loss": 0.1127,
      "step": 1340
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013731412639405205,
      "loss": 0.1656,
      "step": 1350
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013684944237918215,
      "loss": 0.0533,
      "step": 1360
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0001363847583643123,
      "loss": 0.0843,
      "step": 1370
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0001359200743494424,
      "loss": 0.1083,
      "step": 1380
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0001354553903345725,
      "loss": 0.0481,
      "step": 1390
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0001349907063197026,
      "loss": 0.0426,
      "step": 1400
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.891465914915428,
      "eval_f1": 0.8914659149154279,
      "eval_loss": 0.3423434793949127,
      "eval_precision": 0.891465914915428,
      "eval_recall": 0.891465914915428,
      "eval_runtime": 65.7174,
      "eval_samples_per_second": 118.751,
      "eval_steps_per_second": 14.851,
      "step": 1400
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013452602230483272,
      "loss": 0.0537,
      "step": 1410
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00013406133828996284,
      "loss": 0.0605,
      "step": 1420
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013359665427509294,
      "loss": 0.0624,
      "step": 1430
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013313197026022306,
      "loss": 0.0825,
      "step": 1440
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00013266728624535315,
      "loss": 0.133,
      "step": 1450
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00013220260223048327,
      "loss": 0.1118,
      "step": 1460
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001317379182156134,
      "loss": 0.1296,
      "step": 1470
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013127323420074352,
      "loss": 0.1391,
      "step": 1480
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0001308085501858736,
      "loss": 0.109,
      "step": 1490
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0001303438661710037,
      "loss": 0.0831,
      "step": 1500
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.8625064069707842,
      "eval_f1": 0.8625064069707842,
      "eval_loss": 0.40863361954689026,
      "eval_precision": 0.8625064069707842,
      "eval_recall": 0.8625064069707842,
      "eval_runtime": 65.0899,
      "eval_samples_per_second": 119.896,
      "eval_steps_per_second": 14.995,
      "step": 1500
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00012987918215613383,
      "loss": 0.1479,
      "step": 1510
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012941449814126395,
      "loss": 0.0766,
      "step": 1520
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012894981412639407,
      "loss": 0.0498,
      "step": 1530
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012848513011152416,
      "loss": 0.1506,
      "step": 1540
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012802044609665428,
      "loss": 0.1968,
      "step": 1550
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0001275557620817844,
      "loss": 0.1047,
      "step": 1560
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012709107806691453,
      "loss": 0.0901,
      "step": 1570
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012662639405204462,
      "loss": 0.0387,
      "step": 1580
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0001261617100371747,
      "loss": 0.0545,
      "step": 1590
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012569702602230483,
      "loss": 0.106,
      "step": 1600
    },
    {
      "epoch": 1.49,
      "eval_accuracy": 0.8972321886212199,
      "eval_f1": 0.8972321886212199,
      "eval_loss": 0.302043616771698,
      "eval_precision": 0.8972321886212199,
      "eval_recall": 0.8972321886212199,
      "eval_runtime": 65.3274,
      "eval_samples_per_second": 119.46,
      "eval_steps_per_second": 14.94,
      "step": 1600
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00012523234200743496,
      "loss": 0.1566,
      "step": 1610
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012476765799256508,
      "loss": 0.0706,
      "step": 1620
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012430297397769517,
      "loss": 0.1137,
      "step": 1630
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0001238382899628253,
      "loss": 0.0776,
      "step": 1640
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012337360594795539,
      "loss": 0.1014,
      "step": 1650
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001229089219330855,
      "loss": 0.1057,
      "step": 1660
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.00012244423791821563,
      "loss": 0.0913,
      "step": 1670
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012197955390334572,
      "loss": 0.0881,
      "step": 1680
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00012151486988847584,
      "loss": 0.0524,
      "step": 1690
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012105018587360595,
      "loss": 0.0717,
      "step": 1700
    },
    {
      "epoch": 1.58,
      "eval_accuracy": 0.8874935930292158,
      "eval_f1": 0.8874935930292158,
      "eval_loss": 0.3970746397972107,
      "eval_precision": 0.8874935930292158,
      "eval_recall": 0.8874935930292158,
      "eval_runtime": 65.4763,
      "eval_samples_per_second": 119.188,
      "eval_steps_per_second": 14.906,
      "step": 1700
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012058550185873607,
      "loss": 0.0432,
      "step": 1710
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00012012081784386618,
      "loss": 0.1672,
      "step": 1720
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0001196561338289963,
      "loss": 0.1003,
      "step": 1730
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0001191914498141264,
      "loss": 0.0985,
      "step": 1740
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0001187267657992565,
      "loss": 0.0614,
      "step": 1750
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011826208178438662,
      "loss": 0.107,
      "step": 1760
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011779739776951673,
      "loss": 0.0364,
      "step": 1770
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011733271375464685,
      "loss": 0.0813,
      "step": 1780
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011686802973977695,
      "loss": 0.064,
      "step": 1790
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011640334572490708,
      "loss": 0.1134,
      "step": 1800
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.8923628908252178,
      "eval_f1": 0.8923628908252178,
      "eval_loss": 0.35297414660453796,
      "eval_precision": 0.8923628908252178,
      "eval_recall": 0.8923628908252178,
      "eval_runtime": 64.6499,
      "eval_samples_per_second": 120.712,
      "eval_steps_per_second": 15.097,
      "step": 1800
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011593866171003717,
      "loss": 0.0179,
      "step": 1810
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0001154739776951673,
      "loss": 0.0665,
      "step": 1820
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0001150092936802974,
      "loss": 0.0372,
      "step": 1830
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011454460966542751,
      "loss": 0.1263,
      "step": 1840
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011407992565055763,
      "loss": 0.0293,
      "step": 1850
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011361524163568773,
      "loss": 0.1185,
      "step": 1860
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011315055762081786,
      "loss": 0.0788,
      "step": 1870
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00011268587360594795,
      "loss": 0.0587,
      "step": 1880
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011222118959107807,
      "loss": 0.0585,
      "step": 1890
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011175650557620818,
      "loss": 0.0702,
      "step": 1900
    },
    {
      "epoch": 1.77,
      "eval_accuracy": 0.8887749871860584,
      "eval_f1": 0.8887749871860584,
      "eval_loss": 0.3878268897533417,
      "eval_precision": 0.8887749871860584,
      "eval_recall": 0.8887749871860584,
      "eval_runtime": 65.5021,
      "eval_samples_per_second": 119.141,
      "eval_steps_per_second": 14.9,
      "step": 1900
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0001112918215613383,
      "loss": 0.084,
      "step": 1910
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011082713754646841,
      "loss": 0.039,
      "step": 1920
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0001103624535315985,
      "loss": 0.0466,
      "step": 1930
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00010989776951672863,
      "loss": 0.0496,
      "step": 1940
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010943308550185873,
      "loss": 0.1033,
      "step": 1950
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.00010896840148698885,
      "loss": 0.0784,
      "step": 1960
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010850371747211896,
      "loss": 0.0542,
      "step": 1970
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010803903345724908,
      "loss": 0.0652,
      "step": 1980
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00010757434944237919,
      "loss": 0.1441,
      "step": 1990
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010710966542750931,
      "loss": 0.0828,
      "step": 2000
    },
    {
      "epoch": 1.86,
      "eval_accuracy": 0.885187083546899,
      "eval_f1": 0.885187083546899,
      "eval_loss": 0.3822377622127533,
      "eval_precision": 0.885187083546899,
      "eval_recall": 0.885187083546899,
      "eval_runtime": 65.3969,
      "eval_samples_per_second": 119.333,
      "eval_steps_per_second": 14.924,
      "step": 2000
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0001066449814126394,
      "loss": 0.0508,
      "step": 2010
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010618029739776951,
      "loss": 0.129,
      "step": 2020
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010571561338289963,
      "loss": 0.1331,
      "step": 2030
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010525092936802974,
      "loss": 0.0763,
      "step": 2040
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010478624535315986,
      "loss": 0.0495,
      "step": 2050
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010432156133828996,
      "loss": 0.0353,
      "step": 2060
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010385687732342009,
      "loss": 0.0923,
      "step": 2070
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010339219330855019,
      "loss": 0.0461,
      "step": 2080
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010292750929368031,
      "loss": 0.0763,
      "step": 2090
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010246282527881041,
      "loss": 0.0497,
      "step": 2100
    },
    {
      "epoch": 1.95,
      "eval_accuracy": 0.8991542798564839,
      "eval_f1": 0.8991542798564839,
      "eval_loss": 0.3363974690437317,
      "eval_precision": 0.8991542798564839,
      "eval_recall": 0.8991542798564839,
      "eval_runtime": 65.489,
      "eval_samples_per_second": 119.165,
      "eval_steps_per_second": 14.903,
      "step": 2100
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010199814126394052,
      "loss": 0.0843,
      "step": 2110
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010153345724907064,
      "loss": 0.0746,
      "step": 2120
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010106877323420074,
      "loss": 0.0642,
      "step": 2130
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010060408921933087,
      "loss": 0.0207,
      "step": 2140
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010013940520446097,
      "loss": 0.0126,
      "step": 2150
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.967472118959109e-05,
      "loss": 0.0574,
      "step": 2160
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.92100371747212e-05,
      "loss": 0.0037,
      "step": 2170
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.87453531598513e-05,
      "loss": 0.0794,
      "step": 2180
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.828066914498142e-05,
      "loss": 0.0326,
      "step": 2190
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.781598513011153e-05,
      "loss": 0.0191,
      "step": 2200
    },
    {
      "epoch": 2.04,
      "eval_accuracy": 0.8924910302409022,
      "eval_f1": 0.8924910302409022,
      "eval_loss": 0.4041886031627655,
      "eval_precision": 0.8924910302409022,
      "eval_recall": 0.8924910302409022,
      "eval_runtime": 65.2142,
      "eval_samples_per_second": 119.667,
      "eval_steps_per_second": 14.966,
      "step": 2200
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.735130111524164e-05,
      "loss": 0.0039,
      "step": 2210
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.688661710037176e-05,
      "loss": 0.0294,
      "step": 2220
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.642193308550185e-05,
      "loss": 0.0465,
      "step": 2230
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.595724907063197e-05,
      "loss": 0.0069,
      "step": 2240
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.549256505576208e-05,
      "loss": 0.0108,
      "step": 2250
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.50278810408922e-05,
      "loss": 0.0205,
      "step": 2260
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.456319702602231e-05,
      "loss": 0.0044,
      "step": 2270
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.409851301115242e-05,
      "loss": 0.02,
      "step": 2280
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.363382899628254e-05,
      "loss": 0.0356,
      "step": 2290
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.316914498141265e-05,
      "loss": 0.0388,
      "step": 2300
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.8969759097898513,
      "eval_f1": 0.8969759097898513,
      "eval_loss": 0.4484836459159851,
      "eval_precision": 0.8969759097898513,
      "eval_recall": 0.8969759097898513,
      "eval_runtime": 65.1292,
      "eval_samples_per_second": 119.823,
      "eval_steps_per_second": 14.986,
      "step": 2300
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.270446096654275e-05,
      "loss": 0.0118,
      "step": 2310
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.223977695167286e-05,
      "loss": 0.007,
      "step": 2320
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.177509293680297e-05,
      "loss": 0.0716,
      "step": 2330
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.131040892193309e-05,
      "loss": 0.004,
      "step": 2340
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.08457249070632e-05,
      "loss": 0.0501,
      "step": 2350
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.038104089219332e-05,
      "loss": 0.0478,
      "step": 2360
    },
    {
      "epoch": 2.2,
      "learning_rate": 8.991635687732343e-05,
      "loss": 0.0994,
      "step": 2370
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.945167286245353e-05,
      "loss": 0.0133,
      "step": 2380
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.898698884758366e-05,
      "loss": 0.0215,
      "step": 2390
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.852230483271375e-05,
      "loss": 0.0273,
      "step": 2400
    },
    {
      "epoch": 2.23,
      "eval_accuracy": 0.8953100973859559,
      "eval_f1": 0.8953100973859559,
      "eval_loss": 0.3862035572528839,
      "eval_precision": 0.8953100973859559,
      "eval_recall": 0.8953100973859559,
      "eval_runtime": 65.6051,
      "eval_samples_per_second": 118.954,
      "eval_steps_per_second": 14.877,
      "step": 2400
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.805762081784387e-05,
      "loss": 0.0041,
      "step": 2410
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.759293680297398e-05,
      "loss": 0.0112,
      "step": 2420
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.71282527881041e-05,
      "loss": 0.0148,
      "step": 2430
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.66635687732342e-05,
      "loss": 0.0259,
      "step": 2440
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.619888475836431e-05,
      "loss": 0.036,
      "step": 2450
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.573420074349443e-05,
      "loss": 0.0237,
      "step": 2460
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.526951672862454e-05,
      "loss": 0.0177,
      "step": 2470
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.480483271375465e-05,
      "loss": 0.0027,
      "step": 2480
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.434014869888476e-05,
      "loss": 0.0198,
      "step": 2490
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.387546468401486e-05,
      "loss": 0.0032,
      "step": 2500
    },
    {
      "epoch": 2.32,
      "eval_accuracy": 0.8949256791389031,
      "eval_f1": 0.8949256791389031,
      "eval_loss": 0.4709579646587372,
      "eval_precision": 0.8949256791389031,
      "eval_recall": 0.8949256791389031,
      "eval_runtime": 65.1155,
      "eval_samples_per_second": 119.849,
      "eval_steps_per_second": 14.989,
      "step": 2500
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.341078066914499e-05,
      "loss": 0.0337,
      "step": 2510
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.29460966542751e-05,
      "loss": 0.0346,
      "step": 2520
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.248141263940521e-05,
      "loss": 0.0032,
      "step": 2530
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.201672862453532e-05,
      "loss": 0.007,
      "step": 2540
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.155204460966543e-05,
      "loss": 0.0023,
      "step": 2550
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.108736059479555e-05,
      "loss": 0.0014,
      "step": 2560
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.062267657992564e-05,
      "loss": 0.0146,
      "step": 2570
    },
    {
      "epoch": 2.4,
      "learning_rate": 8.015799256505577e-05,
      "loss": 0.0474,
      "step": 2580
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.969330855018587e-05,
      "loss": 0.0259,
      "step": 2590
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.922862453531598e-05,
      "loss": 0.0242,
      "step": 2600
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.8937724243977447,
      "eval_f1": 0.8937724243977447,
      "eval_loss": 0.48591741919517517,
      "eval_precision": 0.8937724243977447,
      "eval_recall": 0.8937724243977447,
      "eval_runtime": 64.6141,
      "eval_samples_per_second": 120.779,
      "eval_steps_per_second": 15.105,
      "step": 2600
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.87639405204461e-05,
      "loss": 0.0226,
      "step": 2610
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.829925650557621e-05,
      "loss": 0.0012,
      "step": 2620
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.783457249070633e-05,
      "loss": 0.0034,
      "step": 2630
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.736988847583644e-05,
      "loss": 0.0013,
      "step": 2640
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.690520446096655e-05,
      "loss": 0.0448,
      "step": 2650
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.644052044609665e-05,
      "loss": 0.0348,
      "step": 2660
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.597583643122676e-05,
      "loss": 0.0381,
      "step": 2670
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.551115241635688e-05,
      "loss": 0.0064,
      "step": 2680
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.504646840148699e-05,
      "loss": 0.0393,
      "step": 2690
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.458178438661711e-05,
      "loss": 0.065,
      "step": 2700
    },
    {
      "epoch": 2.51,
      "eval_accuracy": 0.898257303946694,
      "eval_f1": 0.898257303946694,
      "eval_loss": 0.444608598947525,
      "eval_precision": 0.898257303946694,
      "eval_recall": 0.898257303946694,
      "eval_runtime": 64.9542,
      "eval_samples_per_second": 120.146,
      "eval_steps_per_second": 15.026,
      "step": 2700
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.411710037174722e-05,
      "loss": 0.0063,
      "step": 2710
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.365241635687733e-05,
      "loss": 0.0398,
      "step": 2720
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.318773234200745e-05,
      "loss": 0.0088,
      "step": 2730
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.272304832713755e-05,
      "loss": 0.034,
      "step": 2740
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.225836431226766e-05,
      "loss": 0.0162,
      "step": 2750
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.179368029739777e-05,
      "loss": 0.0013,
      "step": 2760
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.132899628252788e-05,
      "loss": 0.0125,
      "step": 2770
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.0864312267658e-05,
      "loss": 0.0024,
      "step": 2780
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.03996282527881e-05,
      "loss": 0.0407,
      "step": 2790
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.993494423791823e-05,
      "loss": 0.0258,
      "step": 2800
    },
    {
      "epoch": 2.6,
      "eval_accuracy": 0.8924910302409022,
      "eval_f1": 0.8924910302409022,
      "eval_loss": 0.5354290008544922,
      "eval_precision": 0.8924910302409022,
      "eval_recall": 0.8924910302409022,
      "eval_runtime": 65.3052,
      "eval_samples_per_second": 119.501,
      "eval_steps_per_second": 14.945,
      "step": 2800
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.947026022304833e-05,
      "loss": 0.0016,
      "step": 2810
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.900557620817844e-05,
      "loss": 0.0144,
      "step": 2820
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.854089219330855e-05,
      "loss": 0.0009,
      "step": 2830
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.807620817843866e-05,
      "loss": 0.0498,
      "step": 2840
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.761152416356878e-05,
      "loss": 0.0079,
      "step": 2850
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.714684014869889e-05,
      "loss": 0.0318,
      "step": 2860
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.668215613382899e-05,
      "loss": 0.0471,
      "step": 2870
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.621747211895911e-05,
      "loss": 0.0011,
      "step": 2880
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.575278810408922e-05,
      "loss": 0.003,
      "step": 2890
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.528810408921934e-05,
      "loss": 0.001,
      "step": 2900
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 0.895822655048693,
      "eval_f1": 0.895822655048693,
      "eval_loss": 0.5200697183609009,
      "eval_precision": 0.895822655048693,
      "eval_recall": 0.895822655048693,
      "eval_runtime": 65.3033,
      "eval_samples_per_second": 119.504,
      "eval_steps_per_second": 14.946,
      "step": 2900
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.482342007434945e-05,
      "loss": 0.0027,
      "step": 2910
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.435873605947956e-05,
      "loss": 0.0009,
      "step": 2920
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.389405204460967e-05,
      "loss": 0.0334,
      "step": 2930
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.342936802973977e-05,
      "loss": 0.0418,
      "step": 2940
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.29646840148699e-05,
      "loss": 0.0012,
      "step": 2950
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.25e-05,
      "loss": 0.0017,
      "step": 2960
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.203531598513012e-05,
      "loss": 0.039,
      "step": 2970
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.157063197026023e-05,
      "loss": 0.0293,
      "step": 2980
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.110594795539034e-05,
      "loss": 0.001,
      "step": 2990
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.064126394052045e-05,
      "loss": 0.0336,
      "step": 3000
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 0.8923628908252178,
      "eval_f1": 0.8923628908252178,
      "eval_loss": 0.5137433409690857,
      "eval_precision": 0.8923628908252178,
      "eval_recall": 0.8923628908252178,
      "eval_runtime": 64.8327,
      "eval_samples_per_second": 120.371,
      "eval_steps_per_second": 15.054,
      "step": 3000
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.017657992565055e-05,
      "loss": 0.003,
      "step": 3010
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.971189591078067e-05,
      "loss": 0.0054,
      "step": 3020
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.924721189591078e-05,
      "loss": 0.0125,
      "step": 3030
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.8782527881040895e-05,
      "loss": 0.001,
      "step": 3040
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.831784386617101e-05,
      "loss": 0.0025,
      "step": 3050
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.785315985130112e-05,
      "loss": 0.001,
      "step": 3060
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.738847583643123e-05,
      "loss": 0.0317,
      "step": 3070
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.6923791821561346e-05,
      "loss": 0.0008,
      "step": 3080
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.645910780669146e-05,
      "loss": 0.0159,
      "step": 3090
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.599442379182156e-05,
      "loss": 0.0205,
      "step": 3100
    },
    {
      "epoch": 2.88,
      "eval_accuracy": 0.8972321886212199,
      "eval_f1": 0.8972321886212199,
      "eval_loss": 0.5250884294509888,
      "eval_precision": 0.8972321886212199,
      "eval_recall": 0.8972321886212199,
      "eval_runtime": 65.4971,
      "eval_samples_per_second": 119.15,
      "eval_steps_per_second": 14.901,
      "step": 3100
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.5529739776951675e-05,
      "loss": 0.0284,
      "step": 3110
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.506505576208178e-05,
      "loss": 0.0009,
      "step": 3120
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.46003717472119e-05,
      "loss": 0.0166,
      "step": 3130
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.413568773234201e-05,
      "loss": 0.0347,
      "step": 3140
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.3671003717472126e-05,
      "loss": 0.0387,
      "step": 3150
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.320631970260223e-05,
      "loss": 0.003,
      "step": 3160
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.274163568773235e-05,
      "loss": 0.0248,
      "step": 3170
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.227695167286245e-05,
      "loss": 0.0079,
      "step": 3180
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.181226765799256e-05,
      "loss": 0.0012,
      "step": 3190
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.134758364312268e-05,
      "loss": 0.0033,
      "step": 3200
    },
    {
      "epoch": 2.97,
      "eval_accuracy": 0.8906970784213224,
      "eval_f1": 0.8906970784213224,
      "eval_loss": 0.5083357691764832,
      "eval_precision": 0.8906970784213224,
      "eval_recall": 0.8906970784213224,
      "eval_runtime": 66.2102,
      "eval_samples_per_second": 117.867,
      "eval_steps_per_second": 14.741,
      "step": 3200
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.088289962825279e-05,
      "loss": 0.0275,
      "step": 3210
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.04182156133829e-05,
      "loss": 0.028,
      "step": 3220
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.995353159851301e-05,
      "loss": 0.0013,
      "step": 3230
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.948884758364313e-05,
      "loss": 0.0017,
      "step": 3240
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.9024163568773235e-05,
      "loss": 0.002,
      "step": 3250
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.855947955390335e-05,
      "loss": 0.0009,
      "step": 3260
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.8094795539033464e-05,
      "loss": 0.0008,
      "step": 3270
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.763011152416357e-05,
      "loss": 0.0009,
      "step": 3280
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.716542750929368e-05,
      "loss": 0.0007,
      "step": 3290
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.670074349442379e-05,
      "loss": 0.0009,
      "step": 3300
    },
    {
      "epoch": 3.07,
      "eval_accuracy": 0.900691952844695,
      "eval_f1": 0.900691952844695,
      "eval_loss": 0.4909333288669586,
      "eval_precision": 0.900691952844695,
      "eval_recall": 0.900691952844695,
      "eval_runtime": 65.7075,
      "eval_samples_per_second": 118.769,
      "eval_steps_per_second": 14.854,
      "step": 3300
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.623605947955391e-05,
      "loss": 0.0008,
      "step": 3310
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.577137546468402e-05,
      "loss": 0.0007,
      "step": 3320
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.530669144981413e-05,
      "loss": 0.0007,
      "step": 3330
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.484200743494424e-05,
      "loss": 0.0007,
      "step": 3340
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.437732342007435e-05,
      "loss": 0.0008,
      "step": 3350
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.3912639405204465e-05,
      "loss": 0.0005,
      "step": 3360
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.344795539033458e-05,
      "loss": 0.0007,
      "step": 3370
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.298327137546469e-05,
      "loss": 0.0007,
      "step": 3380
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.2518587360594795e-05,
      "loss": 0.0006,
      "step": 3390
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.205390334572491e-05,
      "loss": 0.0465,
      "step": 3400
    },
    {
      "epoch": 3.16,
      "eval_accuracy": 0.8983854433623782,
      "eval_f1": 0.8983854433623782,
      "eval_loss": 0.5175954103469849,
      "eval_precision": 0.8983854433623782,
      "eval_recall": 0.8983854433623782,
      "eval_runtime": 65.9687,
      "eval_samples_per_second": 118.298,
      "eval_steps_per_second": 14.795,
      "step": 3400
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.158921933085502e-05,
      "loss": 0.0008,
      "step": 3410
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.112453531598513e-05,
      "loss": 0.0038,
      "step": 3420
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.0659851301115245e-05,
      "loss": 0.0009,
      "step": 3430
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.019516728624535e-05,
      "loss": 0.0008,
      "step": 3440
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.973048327137547e-05,
      "loss": 0.0007,
      "step": 3450
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9265799256505575e-05,
      "loss": 0.001,
      "step": 3460
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.880111524163569e-05,
      "loss": 0.0006,
      "step": 3470
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.83364312267658e-05,
      "loss": 0.0008,
      "step": 3480
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.787174721189591e-05,
      "loss": 0.0009,
      "step": 3490
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.7407063197026025e-05,
      "loss": 0.0007,
      "step": 3500
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.897744746283957,
      "eval_f1": 0.897744746283957,
      "eval_loss": 0.5411400198936462,
      "eval_precision": 0.897744746283957,
      "eval_recall": 0.897744746283957,
      "eval_runtime": 66.0953,
      "eval_samples_per_second": 118.072,
      "eval_steps_per_second": 14.767,
      "step": 3500
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.694237918215613e-05,
      "loss": 0.0008,
      "step": 3510
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.647769516728625e-05,
      "loss": 0.0007,
      "step": 3520
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.601301115241636e-05,
      "loss": 0.0005,
      "step": 3530
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.5548327137546475e-05,
      "loss": 0.0007,
      "step": 3540
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.5083643122676576e-05,
      "loss": 0.0219,
      "step": 3550
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.461895910780669e-05,
      "loss": 0.0061,
      "step": 3560
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4154275092936805e-05,
      "loss": 0.0462,
      "step": 3570
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.368959107806692e-05,
      "loss": 0.0309,
      "step": 3580
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.322490706319703e-05,
      "loss": 0.0006,
      "step": 3590
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.276022304832714e-05,
      "loss": 0.0406,
      "step": 3600
    },
    {
      "epoch": 3.35,
      "eval_accuracy": 0.9008200922603793,
      "eval_f1": 0.9008200922603794,
      "eval_loss": 0.49289974570274353,
      "eval_precision": 0.9008200922603793,
      "eval_recall": 0.9008200922603793,
      "eval_runtime": 66.1603,
      "eval_samples_per_second": 117.956,
      "eval_steps_per_second": 14.752,
      "step": 3600
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.229553903345725e-05,
      "loss": 0.0008,
      "step": 3610
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.183085501858736e-05,
      "loss": 0.0008,
      "step": 3620
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.136617100371748e-05,
      "loss": 0.0282,
      "step": 3630
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.0901486988847585e-05,
      "loss": 0.0009,
      "step": 3640
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.0436802973977696e-05,
      "loss": 0.0287,
      "step": 3650
    },
    {
      "epoch": 3.4,
      "learning_rate": 2.997211895910781e-05,
      "loss": 0.0011,
      "step": 3660
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.950743494423792e-05,
      "loss": 0.0327,
      "step": 3670
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.904275092936803e-05,
      "loss": 0.0008,
      "step": 3680
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.8578066914498143e-05,
      "loss": 0.0022,
      "step": 3690
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8113382899628254e-05,
      "loss": 0.0016,
      "step": 3700
    },
    {
      "epoch": 3.44,
      "eval_accuracy": 0.8992824192721681,
      "eval_f1": 0.8992824192721681,
      "eval_loss": 0.5065130591392517,
      "eval_precision": 0.8992824192721681,
      "eval_recall": 0.8992824192721681,
      "eval_runtime": 65.825,
      "eval_samples_per_second": 118.557,
      "eval_steps_per_second": 14.827,
      "step": 3700
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.7648698884758368e-05,
      "loss": 0.0007,
      "step": 3710
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.7184014869888475e-05,
      "loss": 0.0009,
      "step": 3720
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6719330855018586e-05,
      "loss": 0.0008,
      "step": 3730
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.62546468401487e-05,
      "loss": 0.0008,
      "step": 3740
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.578996282527881e-05,
      "loss": 0.0007,
      "step": 3750
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5325278810408926e-05,
      "loss": 0.0007,
      "step": 3760
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.4860594795539037e-05,
      "loss": 0.0006,
      "step": 3770
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4442379182156137e-05,
      "loss": 0.0092,
      "step": 3780
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.3977695167286245e-05,
      "loss": 0.0007,
      "step": 3790
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.351301115241636e-05,
      "loss": 0.0006,
      "step": 3800
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.8985135827780625,
      "eval_f1": 0.8985135827780625,
      "eval_loss": 0.5403147339820862,
      "eval_precision": 0.8985135827780625,
      "eval_recall": 0.8985135827780625,
      "eval_runtime": 65.9895,
      "eval_samples_per_second": 118.261,
      "eval_steps_per_second": 14.79,
      "step": 3800
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.304832713754647e-05,
      "loss": 0.0033,
      "step": 3810
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.258364312267658e-05,
      "loss": 0.0005,
      "step": 3820
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.211895910780669e-05,
      "loss": 0.0006,
      "step": 3830
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.1654275092936806e-05,
      "loss": 0.0006,
      "step": 3840
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.1189591078066913e-05,
      "loss": 0.0006,
      "step": 3850
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0724907063197028e-05,
      "loss": 0.0006,
      "step": 3860
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.026022304832714e-05,
      "loss": 0.0005,
      "step": 3870
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.979553903345725e-05,
      "loss": 0.0007,
      "step": 3880
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.9330855018587364e-05,
      "loss": 0.0006,
      "step": 3890
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.886617100371747e-05,
      "loss": 0.0272,
      "step": 3900
    },
    {
      "epoch": 3.62,
      "eval_accuracy": 0.8991542798564839,
      "eval_f1": 0.8991542798564839,
      "eval_loss": 0.5398901104927063,
      "eval_precision": 0.8991542798564839,
      "eval_recall": 0.8991542798564839,
      "eval_runtime": 65.8876,
      "eval_samples_per_second": 118.444,
      "eval_steps_per_second": 14.813,
      "step": 3900
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8401486988847586e-05,
      "loss": 0.0006,
      "step": 3910
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.7936802973977697e-05,
      "loss": 0.0005,
      "step": 3920
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.7472118959107808e-05,
      "loss": 0.0011,
      "step": 3930
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.700743494423792e-05,
      "loss": 0.0176,
      "step": 3940
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.654275092936803e-05,
      "loss": 0.0061,
      "step": 3950
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.607806691449814e-05,
      "loss": 0.0005,
      "step": 3960
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5613382899628255e-05,
      "loss": 0.0017,
      "step": 3970
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5148698884758364e-05,
      "loss": 0.0005,
      "step": 3980
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4684014869888477e-05,
      "loss": 0.0005,
      "step": 3990
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.421933085501859e-05,
      "loss": 0.0223,
      "step": 4000
    },
    {
      "epoch": 3.72,
      "eval_accuracy": 0.9044079958995387,
      "eval_f1": 0.9044079958995387,
      "eval_loss": 0.5075308084487915,
      "eval_precision": 0.9044079958995387,
      "eval_recall": 0.9044079958995387,
      "eval_runtime": 66.1084,
      "eval_samples_per_second": 118.049,
      "eval_steps_per_second": 14.764,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.37546468401487e-05,
      "loss": 0.0006,
      "step": 4010
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.3289962825278813e-05,
      "loss": 0.0007,
      "step": 4020
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2825278810408922e-05,
      "loss": 0.0008,
      "step": 4030
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2360594795539033e-05,
      "loss": 0.0133,
      "step": 4040
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.1895910780669145e-05,
      "loss": 0.0074,
      "step": 4050
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1431226765799256e-05,
      "loss": 0.0016,
      "step": 4060
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.0966542750929369e-05,
      "loss": 0.0007,
      "step": 4070
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.050185873605948e-05,
      "loss": 0.0005,
      "step": 4080
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.0037174721189593e-05,
      "loss": 0.0005,
      "step": 4090
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.572490706319703e-06,
      "loss": 0.0006,
      "step": 4100
    },
    {
      "epoch": 3.81,
      "eval_accuracy": 0.8992824192721681,
      "eval_f1": 0.8992824192721681,
      "eval_loss": 0.5432131290435791,
      "eval_precision": 0.8992824192721681,
      "eval_recall": 0.8992824192721681,
      "eval_runtime": 66.0363,
      "eval_samples_per_second": 118.177,
      "eval_steps_per_second": 14.78,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "learning_rate": 9.107806691449814e-06,
      "loss": 0.0125,
      "step": 4110
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.643122676579925e-06,
      "loss": 0.0005,
      "step": 4120
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.178438661710038e-06,
      "loss": 0.0005,
      "step": 4130
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.713754646840149e-06,
      "loss": 0.0006,
      "step": 4140
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.249070631970261e-06,
      "loss": 0.0008,
      "step": 4150
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.7843866171003716e-06,
      "loss": 0.0005,
      "step": 4160
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.319702602230483e-06,
      "loss": 0.0006,
      "step": 4170
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.855018587360595e-06,
      "loss": 0.0006,
      "step": 4180
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.390334572490707e-06,
      "loss": 0.0005,
      "step": 4190
    },
    {
      "epoch": 3.9,
      "learning_rate": 4.925650557620818e-06,
      "loss": 0.0153,
      "step": 4200
    },
    {
      "epoch": 3.9,
      "eval_accuracy": 0.902101486417222,
      "eval_f1": 0.902101486417222,
      "eval_loss": 0.5262740850448608,
      "eval_precision": 0.902101486417222,
      "eval_recall": 0.902101486417222,
      "eval_runtime": 66.1026,
      "eval_samples_per_second": 118.059,
      "eval_steps_per_second": 14.765,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.4609665427509296e-06,
      "loss": 0.0005,
      "step": 4210
    },
    {
      "epoch": 3.92,
      "learning_rate": 3.9962825278810405e-06,
      "loss": 0.0006,
      "step": 4220
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.5315985130111527e-06,
      "loss": 0.0005,
      "step": 4230
    },
    {
      "epoch": 3.94,
      "learning_rate": 3.0669144981412644e-06,
      "loss": 0.0006,
      "step": 4240
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.6022304832713758e-06,
      "loss": 0.0272,
      "step": 4250
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.137546468401487e-06,
      "loss": 0.0004,
      "step": 4260
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.6728624535315985e-06,
      "loss": 0.0005,
      "step": 4270
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.2081784386617102e-06,
      "loss": 0.0006,
      "step": 4280
    },
    {
      "epoch": 3.99,
      "learning_rate": 7.434944237918216e-07,
      "loss": 0.0012,
      "step": 4290
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.788104089219331e-07,
      "loss": 0.0273,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9028703229113275,
      "eval_f1": 0.9028703229113276,
      "eval_loss": 0.5241953134536743,
      "eval_precision": 0.9028703229113275,
      "eval_recall": 0.9028703229113275,
      "eval_runtime": 66.1578,
      "eval_samples_per_second": 117.96,
      "eval_steps_per_second": 14.753,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "step": 4304,
      "total_flos": 5.335788436902347e+18,
      "train_loss": 0.09078962347299496,
      "train_runtime": 4275.1726,
      "train_samples_per_second": 16.106,
      "train_steps_per_second": 1.007
    }
  ],
  "max_steps": 4304,
  "num_train_epochs": 4,
  "total_flos": 5.335788436902347e+18,
  "trial_name": null,
  "trial_params": null
}