{
  "best_metric": 0.24793727695941925,
  "best_model_checkpoint": "./vit-large-binary-isic-patch-32/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": 0.5453,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00019907063197026022,
      "loss": 0.4975,
      "step": 20
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019860594795539035,
      "loss": 0.3572,
      "step": 30
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019814126394052047,
      "loss": 0.4367,
      "step": 40
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019767657992565056,
      "loss": 0.3483,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019721189591078068,
      "loss": 0.4717,
      "step": 60
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019674721189591078,
      "loss": 0.3744,
      "step": 70
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019628252788104092,
      "loss": 0.4076,
      "step": 80
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019581784386617102,
      "loss": 0.3593,
      "step": 90
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0001953531598513011,
      "loss": 0.3606,
      "step": 100
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.8122757560225525,
      "eval_f1": 0.8122757560225525,
      "eval_loss": 0.32245561480522156,
      "eval_precision": 0.8122757560225525,
      "eval_recall": 0.8122757560225525,
      "eval_runtime": 80.6483,
      "eval_samples_per_second": 96.766,
      "eval_steps_per_second": 12.102,
      "step": 100
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019488847583643123,
      "loss": 0.37,
      "step": 110
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00019442379182156135,
      "loss": 0.3961,
      "step": 120
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019395910780669148,
      "loss": 0.2499,
      "step": 130
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019349442379182157,
      "loss": 0.3276,
      "step": 140
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0001930297397769517,
      "loss": 0.2692,
      "step": 150
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019256505576208178,
      "loss": 0.2214,
      "step": 160
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0001921003717472119,
      "loss": 0.3521,
      "step": 170
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019163568773234203,
      "loss": 0.2649,
      "step": 180
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019117100371747212,
      "loss": 0.371,
      "step": 190
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019070631970260224,
      "loss": 0.2867,
      "step": 200
    },
    {
      "epoch": 0.19,
      "eval_accuracy": 0.8564838544336237,
      "eval_f1": 0.8564838544336237,
      "eval_loss": 0.36539900302886963,
      "eval_precision": 0.8564838544336237,
      "eval_recall": 0.8564838544336237,
      "eval_runtime": 81.1123,
      "eval_samples_per_second": 96.212,
      "eval_steps_per_second": 12.033,
      "step": 200
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00019024163568773234,
      "loss": 0.3614,
      "step": 210
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018977695167286246,
      "loss": 0.2786,
      "step": 220
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018931226765799258,
      "loss": 0.226,
      "step": 230
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0001888475836431227,
      "loss": 0.2321,
      "step": 240
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0001883828996282528,
      "loss": 0.2815,
      "step": 250
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018791821561338291,
      "loss": 0.4371,
      "step": 260
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.000187453531598513,
      "loss": 0.3801,
      "step": 270
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00018698884758364313,
      "loss": 0.3431,
      "step": 280
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018652416356877325,
      "loss": 0.3375,
      "step": 290
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018605947955390334,
      "loss": 0.3359,
      "step": 300
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.8502050230650948,
      "eval_f1": 0.8502050230650948,
      "eval_loss": 0.30512556433677673,
      "eval_precision": 0.8502050230650948,
      "eval_recall": 0.8502050230650948,
      "eval_runtime": 81.3358,
      "eval_samples_per_second": 95.948,
      "eval_steps_per_second": 12.0,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018559479553903347,
      "loss": 0.3227,
      "step": 310
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018513011152416359,
      "loss": 0.2997,
      "step": 320
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0001846654275092937,
      "loss": 0.2519,
      "step": 330
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0001842007434944238,
      "loss": 0.2867,
      "step": 340
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0001837360594795539,
      "loss": 0.2631,
      "step": 350
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00018327137546468402,
      "loss": 0.3865,
      "step": 360
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018280669144981414,
      "loss": 0.2142,
      "step": 370
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018234200743494426,
      "loss": 0.2634,
      "step": 380
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018187732342007435,
      "loss": 0.2382,
      "step": 390
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00018141263940520447,
      "loss": 0.2825,
      "step": 400
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.8587903639159405,
      "eval_f1": 0.8587903639159405,
      "eval_loss": 0.35762181878089905,
      "eval_precision": 0.8587903639159405,
      "eval_recall": 0.8587903639159405,
      "eval_runtime": 81.8943,
      "eval_samples_per_second": 95.294,
      "eval_steps_per_second": 11.918,
      "step": 400
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018094795539033457,
      "loss": 0.4065,
      "step": 410
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0001804832713754647,
      "loss": 0.363,
      "step": 420
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0001800185873605948,
      "loss": 0.286,
      "step": 430
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0001795539033457249,
      "loss": 0.1958,
      "step": 440
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017908921933085502,
      "loss": 0.2245,
      "step": 450
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00017862453531598512,
      "loss": 0.2184,
      "step": 460
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017815985130111527,
      "loss": 0.3413,
      "step": 470
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017769516728624536,
      "loss": 0.2517,
      "step": 480
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017723048327137548,
      "loss": 0.2505,
      "step": 490
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017676579925650558,
      "loss": 0.2438,
      "step": 500
    },
    {
      "epoch": 0.46,
      "eval_accuracy": 0.8619938493080471,
      "eval_f1": 0.8619938493080471,
      "eval_loss": 0.30426713824272156,
      "eval_precision": 0.8619938493080471,
      "eval_recall": 0.8619938493080471,
      "eval_runtime": 81.1285,
      "eval_samples_per_second": 96.193,
      "eval_steps_per_second": 12.03,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0001763011152416357,
      "loss": 0.1686,
      "step": 510
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00017583643122676582,
      "loss": 0.2358,
      "step": 520
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0001753717472118959,
      "loss": 0.2654,
      "step": 530
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00017490706319702603,
      "loss": 0.2242,
      "step": 540
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00017444237918215613,
      "loss": 0.2445,
      "step": 550
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00017397769516728625,
      "loss": 0.246,
      "step": 560
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017351301115241637,
      "loss": 0.2333,
      "step": 570
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0001730483271375465,
      "loss": 0.1237,
      "step": 580
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00017258364312267658,
      "loss": 0.2782,
      "step": 590
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0001721189591078067,
      "loss": 0.1465,
      "step": 600
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 0.8798052280881599,
      "eval_f1": 0.8798052280881599,
      "eval_loss": 0.27906984090805054,
      "eval_precision": 0.8798052280881599,
      "eval_recall": 0.8798052280881599,
      "eval_runtime": 81.7512,
      "eval_samples_per_second": 95.46,
      "eval_steps_per_second": 11.939,
      "step": 600
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0001716542750929368,
      "loss": 0.1924,
      "step": 610
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017118959107806692,
      "loss": 0.2634,
      "step": 620
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017072490706319704,
      "loss": 0.2502,
      "step": 630
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017026022304832714,
      "loss": 0.2155,
      "step": 640
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00016979553903345726,
      "loss": 0.249,
      "step": 650
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016933085501858735,
      "loss": 0.2606,
      "step": 660
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0001688661710037175,
      "loss": 0.2149,
      "step": 670
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0001684014869888476,
      "loss": 0.2545,
      "step": 680
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016793680297397771,
      "loss": 0.2159,
      "step": 690
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0001674721189591078,
      "loss": 0.2318,
      "step": 700
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.862890825217837,
      "eval_f1": 0.862890825217837,
      "eval_loss": 0.28155967593193054,
      "eval_precision": 0.862890825217837,
      "eval_recall": 0.862890825217837,
      "eval_runtime": 82.2967,
      "eval_samples_per_second": 94.828,
      "eval_steps_per_second": 11.86,
      "step": 700
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016700743494423793,
      "loss": 0.2853,
      "step": 710
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016654275092936805,
      "loss": 0.2027,
      "step": 720
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00016607806691449814,
      "loss": 0.29,
      "step": 730
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016561338289962827,
      "loss": 0.3087,
      "step": 740
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016514869888475836,
      "loss": 0.1655,
      "step": 750
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016468401486988848,
      "loss": 0.1108,
      "step": 760
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0001642193308550186,
      "loss": 0.2421,
      "step": 770
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016375464684014872,
      "loss": 0.2158,
      "step": 780
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00016328996282527882,
      "loss": 0.2559,
      "step": 790
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0001628252788104089,
      "loss": 0.1846,
      "step": 800
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 0.8913377754997437,
      "eval_f1": 0.8913377754997437,
      "eval_loss": 0.24793727695941925,
      "eval_precision": 0.8913377754997437,
      "eval_recall": 0.8913377754997437,
      "eval_runtime": 82.5571,
      "eval_samples_per_second": 94.528,
      "eval_steps_per_second": 11.822,
      "step": 800
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00016236059479553903,
      "loss": 0.1809,
      "step": 810
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016189591078066915,
      "loss": 0.2702,
      "step": 820
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016143122676579927,
      "loss": 0.2589,
      "step": 830
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016096654275092937,
      "loss": 0.1895,
      "step": 840
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0001605018587360595,
      "loss": 0.2404,
      "step": 850
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0001600371747211896,
      "loss": 0.2182,
      "step": 860
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0001595724907063197,
      "loss": 0.3631,
      "step": 870
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00015910780669144982,
      "loss": 0.2109,
      "step": 880
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015864312267657992,
      "loss": 0.2076,
      "step": 890
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015817843866171004,
      "loss": 0.1451,
      "step": 900
    },
    {
      "epoch": 0.84,
      "eval_accuracy": 0.8828805740645823,
      "eval_f1": 0.8828805740645824,
      "eval_loss": 0.32319504022598267,
      "eval_precision": 0.8828805740645823,
      "eval_recall": 0.8828805740645823,
      "eval_runtime": 81.5657,
      "eval_samples_per_second": 95.677,
      "eval_steps_per_second": 11.966,
      "step": 900
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00015771375464684016,
      "loss": 0.1979,
      "step": 910
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015724907063197028,
      "loss": 0.2097,
      "step": 920
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015678438661710038,
      "loss": 0.2007,
      "step": 930
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0001563197026022305,
      "loss": 0.3018,
      "step": 940
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0001558550185873606,
      "loss": 0.2009,
      "step": 950
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0001553903345724907,
      "loss": 0.2443,
      "step": 960
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00015492565055762083,
      "loss": 0.1993,
      "step": 970
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015446096654275093,
      "loss": 0.1611,
      "step": 980
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015399628252788105,
      "loss": 0.256,
      "step": 990
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00015353159851301114,
      "loss": 0.1157,
      "step": 1000
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.8767298821117375,
      "eval_f1": 0.8767298821117375,
      "eval_loss": 0.2799040973186493,
      "eval_precision": 0.8767298821117375,
      "eval_recall": 0.8767298821117375,
      "eval_runtime": 81.1552,
      "eval_samples_per_second": 96.161,
      "eval_steps_per_second": 12.026,
      "step": 1000
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0001530669144981413,
      "loss": 0.1427,
      "step": 1010
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015260223048327138,
      "loss": 0.1118,
      "step": 1020
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0001521375464684015,
      "loss": 0.1709,
      "step": 1030
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0001516728624535316,
      "loss": 0.1446,
      "step": 1040
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00015120817843866172,
      "loss": 0.3159,
      "step": 1050
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015074349442379184,
      "loss": 0.1712,
      "step": 1060
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015027881040892194,
      "loss": 0.1344,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00014981412639405206,
      "loss": 0.1549,
      "step": 1080
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014934944237918215,
      "loss": 0.1055,
      "step": 1090
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014888475836431227,
      "loss": 0.1175,
      "step": 1100
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.869682214249103,
      "eval_f1": 0.869682214249103,
      "eval_loss": 0.3236228823661804,
      "eval_precision": 0.869682214249103,
      "eval_recall": 0.869682214249103,
      "eval_runtime": 81.7176,
      "eval_samples_per_second": 95.5,
      "eval_steps_per_second": 11.944,
      "step": 1100
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0001484200743494424,
      "loss": 0.1202,
      "step": 1110
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014795539033457251,
      "loss": 0.1381,
      "step": 1120
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0001474907063197026,
      "loss": 0.15,
      "step": 1130
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0001470260223048327,
      "loss": 0.1897,
      "step": 1140
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00014656133828996282,
      "loss": 0.1298,
      "step": 1150
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014609665427509294,
      "loss": 0.1613,
      "step": 1160
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014563197026022307,
      "loss": 0.0893,
      "step": 1170
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00014516728624535316,
      "loss": 0.2234,
      "step": 1180
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014470260223048328,
      "loss": 0.1569,
      "step": 1190
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014423791821561337,
      "loss": 0.2015,
      "step": 1200
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.8705791901588928,
      "eval_f1": 0.870579190158893,
      "eval_loss": 0.3056430220603943,
      "eval_precision": 0.8705791901588928,
      "eval_recall": 0.8705791901588928,
      "eval_runtime": 81.3892,
      "eval_samples_per_second": 95.885,
      "eval_steps_per_second": 11.992,
      "step": 1200
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014377323420074352,
      "loss": 0.1816,
      "step": 1210
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00014330855018587362,
      "loss": 0.1157,
      "step": 1220
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0001428438661710037,
      "loss": 0.1125,
      "step": 1230
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00014237918215613383,
      "loss": 0.1507,
      "step": 1240
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014191449814126395,
      "loss": 0.0971,
      "step": 1250
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00014144981412639407,
      "loss": 0.1003,
      "step": 1260
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014098513011152417,
      "loss": 0.1637,
      "step": 1270
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0001405204460966543,
      "loss": 0.1592,
      "step": 1280
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00014005576208178438,
      "loss": 0.1446,
      "step": 1290
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0001395910780669145,
      "loss": 0.0627,
      "step": 1300
    },
    {
      "epoch": 1.21,
      "eval_accuracy": 0.8757047667862634,
      "eval_f1": 0.8757047667862634,
      "eval_loss": 0.31595587730407715,
      "eval_precision": 0.8757047667862634,
      "eval_recall": 0.8757047667862634,
      "eval_runtime": 82.5326,
      "eval_samples_per_second": 94.557,
      "eval_steps_per_second": 11.826,
      "step": 1300
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00013912639405204463,
      "loss": 0.138,
      "step": 1310
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013866171003717472,
      "loss": 0.0946,
      "step": 1320
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013819702602230484,
      "loss": 0.0869,
      "step": 1330
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013773234200743493,
      "loss": 0.0712,
      "step": 1340
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013726765799256506,
      "loss": 0.1059,
      "step": 1350
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013680297397769518,
      "loss": 0.1168,
      "step": 1360
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0001363382899628253,
      "loss": 0.1482,
      "step": 1370
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0001358736059479554,
      "loss": 0.1544,
      "step": 1380
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013540892193308549,
      "loss": 0.1423,
      "step": 1390
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00013494423791821563,
      "loss": 0.0842,
      "step": 1400
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.8908252178370066,
      "eval_f1": 0.8908252178370066,
      "eval_loss": 0.3299122750759125,
      "eval_precision": 0.8908252178370066,
      "eval_recall": 0.8908252178370066,
      "eval_runtime": 81.3488,
      "eval_samples_per_second": 95.933,
      "eval_steps_per_second": 11.998,
      "step": 1400
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013447955390334573,
      "loss": 0.2055,
      "step": 1410
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00013401486988847585,
      "loss": 0.0821,
      "step": 1420
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013355018587360594,
      "loss": 0.1164,
      "step": 1430
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013308550185873606,
      "loss": 0.095,
      "step": 1440
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00013262081784386618,
      "loss": 0.0803,
      "step": 1450
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0001321561338289963,
      "loss": 0.1729,
      "step": 1460
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001316914498141264,
      "loss": 0.135,
      "step": 1470
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0001312267657992565,
      "loss": 0.107,
      "step": 1480
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013076208178438661,
      "loss": 0.0676,
      "step": 1490
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013029739776951674,
      "loss": 0.0617,
      "step": 1500
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.884674525884162,
      "eval_f1": 0.884674525884162,
      "eval_loss": 0.33529117703437805,
      "eval_precision": 0.884674525884162,
      "eval_recall": 0.884674525884162,
      "eval_runtime": 81.7735,
      "eval_samples_per_second": 95.434,
      "eval_steps_per_second": 11.935,
      "step": 1500
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00012983271375464686,
      "loss": 0.1767,
      "step": 1510
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012936802973977695,
      "loss": 0.1008,
      "step": 1520
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012890334572490707,
      "loss": 0.1224,
      "step": 1530
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012843866171003717,
      "loss": 0.1069,
      "step": 1540
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012797397769516731,
      "loss": 0.1251,
      "step": 1550
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0001275092936802974,
      "loss": 0.0852,
      "step": 1560
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0001270446096654275,
      "loss": 0.1602,
      "step": 1570
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012657992565055762,
      "loss": 0.11,
      "step": 1580
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00012611524163568772,
      "loss": 0.1638,
      "step": 1590
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012565055762081787,
      "loss": 0.1919,
      "step": 1600
    },
    {
      "epoch": 1.49,
      "eval_accuracy": 0.8630189646335212,
      "eval_f1": 0.8630189646335212,
      "eval_loss": 0.342080682516098,
      "eval_precision": 0.8630189646335212,
      "eval_recall": 0.8630189646335212,
      "eval_runtime": 80.8014,
      "eval_samples_per_second": 96.582,
      "eval_steps_per_second": 12.079,
      "step": 1600
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00012518587360594796,
      "loss": 0.1023,
      "step": 1610
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012472118959107808,
      "loss": 0.1326,
      "step": 1620
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012425650557620817,
      "loss": 0.1534,
      "step": 1630
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0001237918215613383,
      "loss": 0.0779,
      "step": 1640
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012332713754646842,
      "loss": 0.0704,
      "step": 1650
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001228624535315985,
      "loss": 0.1526,
      "step": 1660
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.00012239776951672863,
      "loss": 0.0959,
      "step": 1670
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012193308550185874,
      "loss": 0.0807,
      "step": 1680
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00012146840148698886,
      "loss": 0.0652,
      "step": 1690
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012100371747211895,
      "loss": 0.1212,
      "step": 1700
    },
    {
      "epoch": 1.58,
      "eval_accuracy": 0.8692977960020503,
      "eval_f1": 0.8692977960020504,
      "eval_loss": 0.4301460087299347,
      "eval_precision": 0.8692977960020503,
      "eval_recall": 0.8692977960020503,
      "eval_runtime": 81.8213,
      "eval_samples_per_second": 95.379,
      "eval_steps_per_second": 11.928,
      "step": 1700
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012053903345724909,
      "loss": 0.0918,
      "step": 1710
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00012007434944237918,
      "loss": 0.1311,
      "step": 1720
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0001196096654275093,
      "loss": 0.0871,
      "step": 1730
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011914498141263941,
      "loss": 0.0736,
      "step": 1740
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011868029739776952,
      "loss": 0.1844,
      "step": 1750
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011821561338289964,
      "loss": 0.1179,
      "step": 1760
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011775092936802973,
      "loss": 0.1003,
      "step": 1770
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011728624535315987,
      "loss": 0.063,
      "step": 1780
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011682156133828996,
      "loss": 0.1095,
      "step": 1790
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011635687732342008,
      "loss": 0.1012,
      "step": 1800
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.8799333675038442,
      "eval_f1": 0.8799333675038442,
      "eval_loss": 0.3351816236972809,
      "eval_precision": 0.8799333675038442,
      "eval_recall": 0.8799333675038442,
      "eval_runtime": 81.3399,
      "eval_samples_per_second": 95.943,
      "eval_steps_per_second": 11.999,
      "step": 1800
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011589219330855019,
      "loss": 0.0599,
      "step": 1810
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011542750929368031,
      "loss": 0.1574,
      "step": 1820
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011496282527881042,
      "loss": 0.0812,
      "step": 1830
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011449814126394051,
      "loss": 0.1645,
      "step": 1840
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011403345724907064,
      "loss": 0.0667,
      "step": 1850
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011356877323420074,
      "loss": 0.1485,
      "step": 1860
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011310408921933086,
      "loss": 0.083,
      "step": 1870
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00011263940520446097,
      "loss": 0.1055,
      "step": 1880
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011217472118959109,
      "loss": 0.0513,
      "step": 1890
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0001117100371747212,
      "loss": 0.0622,
      "step": 1900
    },
    {
      "epoch": 1.77,
      "eval_accuracy": 0.8881342901076371,
      "eval_f1": 0.8881342901076371,
      "eval_loss": 0.327203631401062,
      "eval_precision": 0.8881342901076371,
      "eval_recall": 0.8881342901076371,
      "eval_runtime": 83.0524,
      "eval_samples_per_second": 93.965,
      "eval_steps_per_second": 11.752,
      "step": 1900
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0001112453531598513,
      "loss": 0.0792,
      "step": 1910
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011078066914498142,
      "loss": 0.0725,
      "step": 1920
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011031598513011152,
      "loss": 0.0755,
      "step": 1930
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00010985130111524164,
      "loss": 0.0726,
      "step": 1940
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010938661710037175,
      "loss": 0.0688,
      "step": 1950
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.00010892193308550187,
      "loss": 0.0433,
      "step": 1960
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010845724907063197,
      "loss": 0.1268,
      "step": 1970
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0001079925650557621,
      "loss": 0.1058,
      "step": 1980
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0001075278810408922,
      "loss": 0.1628,
      "step": 1990
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0001070631970260223,
      "loss": 0.0733,
      "step": 2000
    },
    {
      "epoch": 1.86,
      "eval_accuracy": 0.8867247565351102,
      "eval_f1": 0.8867247565351102,
      "eval_loss": 0.2903330624103546,
      "eval_precision": 0.8867247565351102,
      "eval_recall": 0.8867247565351102,
      "eval_runtime": 81.9645,
      "eval_samples_per_second": 95.212,
      "eval_steps_per_second": 11.908,
      "step": 2000
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.00010659851301115242,
      "loss": 0.0741,
      "step": 2010
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010613382899628253,
      "loss": 0.1168,
      "step": 2020
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010566914498141265,
      "loss": 0.1115,
      "step": 2030
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010520446096654275,
      "loss": 0.0555,
      "step": 2040
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010473977695167288,
      "loss": 0.0806,
      "step": 2050
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010427509293680297,
      "loss": 0.0504,
      "step": 2060
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0001038104089219331,
      "loss": 0.1187,
      "step": 2070
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0001033457249070632,
      "loss": 0.0853,
      "step": 2080
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0001028810408921933,
      "loss": 0.1163,
      "step": 2090
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010241635687732343,
      "loss": 0.0293,
      "step": 2100
    },
    {
      "epoch": 1.95,
      "eval_accuracy": 0.9031266017426961,
      "eval_f1": 0.903126601742696,
      "eval_loss": 0.277230441570282,
      "eval_precision": 0.9031266017426961,
      "eval_recall": 0.9031266017426961,
      "eval_runtime": 81.3072,
      "eval_samples_per_second": 95.982,
      "eval_steps_per_second": 12.004,
      "step": 2100
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010195167286245353,
      "loss": 0.0313,
      "step": 2110
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010148698884758365,
      "loss": 0.1132,
      "step": 2120
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010102230483271375,
      "loss": 0.0729,
      "step": 2130
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010055762081784388,
      "loss": 0.0785,
      "step": 2140
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010009293680297398,
      "loss": 0.0806,
      "step": 2150
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.962825278810409e-05,
      "loss": 0.0931,
      "step": 2160
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.916356877323421e-05,
      "loss": 0.0509,
      "step": 2170
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.869888475836432e-05,
      "loss": 0.0536,
      "step": 2180
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.823420074349443e-05,
      "loss": 0.0104,
      "step": 2190
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.776951672862455e-05,
      "loss": 0.0305,
      "step": 2200
    },
    {
      "epoch": 2.04,
      "eval_accuracy": 0.8928754484879549,
      "eval_f1": 0.8928754484879549,
      "eval_loss": 0.3218231797218323,
      "eval_precision": 0.8928754484879549,
      "eval_recall": 0.8928754484879549,
      "eval_runtime": 81.9077,
      "eval_samples_per_second": 95.278,
      "eval_steps_per_second": 11.916,
      "step": 2200
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.730483271375466e-05,
      "loss": 0.0102,
      "step": 2210
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.684014869888476e-05,
      "loss": 0.0067,
      "step": 2220
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.637546468401487e-05,
      "loss": 0.0437,
      "step": 2230
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.591078066914498e-05,
      "loss": 0.029,
      "step": 2240
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.54460966542751e-05,
      "loss": 0.0504,
      "step": 2250
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.498141263940521e-05,
      "loss": 0.0244,
      "step": 2260
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.451672862453533e-05,
      "loss": 0.0249,
      "step": 2270
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.405204460966544e-05,
      "loss": 0.0104,
      "step": 2280
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.358736059479554e-05,
      "loss": 0.025,
      "step": 2290
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.316914498141265e-05,
      "loss": 0.034,
      "step": 2300
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.891465914915428,
      "eval_f1": 0.8914659149154279,
      "eval_loss": 0.4207420349121094,
      "eval_precision": 0.891465914915428,
      "eval_recall": 0.891465914915428,
      "eval_runtime": 81.489,
      "eval_samples_per_second": 95.767,
      "eval_steps_per_second": 11.977,
      "step": 2300
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.270446096654275e-05,
      "loss": 0.0129,
      "step": 2310
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.223977695167286e-05,
      "loss": 0.0123,
      "step": 2320
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.177509293680297e-05,
      "loss": 0.0169,
      "step": 2330
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.131040892193309e-05,
      "loss": 0.014,
      "step": 2340
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.08457249070632e-05,
      "loss": 0.0259,
      "step": 2350
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.038104089219332e-05,
      "loss": 0.0781,
      "step": 2360
    },
    {
      "epoch": 2.2,
      "learning_rate": 8.991635687732343e-05,
      "loss": 0.0505,
      "step": 2370
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.945167286245353e-05,
      "loss": 0.0429,
      "step": 2380
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.898698884758366e-05,
      "loss": 0.0067,
      "step": 2390
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.852230483271375e-05,
      "loss": 0.0113,
      "step": 2400
    },
    {
      "epoch": 2.23,
      "eval_accuracy": 0.8947975397232188,
      "eval_f1": 0.8947975397232188,
      "eval_loss": 0.41928279399871826,
      "eval_precision": 0.8947975397232188,
      "eval_recall": 0.8947975397232188,
      "eval_runtime": 81.8306,
      "eval_samples_per_second": 95.368,
      "eval_steps_per_second": 11.927,
      "step": 2400
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.805762081784387e-05,
      "loss": 0.0176,
      "step": 2410
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.759293680297398e-05,
      "loss": 0.0046,
      "step": 2420
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.71282527881041e-05,
      "loss": 0.0119,
      "step": 2430
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.66635687732342e-05,
      "loss": 0.0175,
      "step": 2440
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.619888475836431e-05,
      "loss": 0.0819,
      "step": 2450
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.573420074349443e-05,
      "loss": 0.0243,
      "step": 2460
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.526951672862454e-05,
      "loss": 0.0178,
      "step": 2470
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.480483271375465e-05,
      "loss": 0.0425,
      "step": 2480
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.434014869888476e-05,
      "loss": 0.0642,
      "step": 2490
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.387546468401486e-05,
      "loss": 0.0071,
      "step": 2500
    },
    {
      "epoch": 2.32,
      "eval_accuracy": 0.8908252178370066,
      "eval_f1": 0.8908252178370066,
      "eval_loss": 0.4371524751186371,
      "eval_precision": 0.8908252178370066,
      "eval_recall": 0.8908252178370066,
      "eval_runtime": 82.4806,
      "eval_samples_per_second": 94.616,
      "eval_steps_per_second": 11.833,
      "step": 2500
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.341078066914499e-05,
      "loss": 0.0519,
      "step": 2510
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.29460966542751e-05,
      "loss": 0.0304,
      "step": 2520
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.248141263940521e-05,
      "loss": 0.1128,
      "step": 2530
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.201672862453532e-05,
      "loss": 0.0336,
      "step": 2540
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.155204460966543e-05,
      "loss": 0.0224,
      "step": 2550
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.108736059479555e-05,
      "loss": 0.0109,
      "step": 2560
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.062267657992564e-05,
      "loss": 0.0246,
      "step": 2570
    },
    {
      "epoch": 2.4,
      "learning_rate": 8.015799256505577e-05,
      "loss": 0.073,
      "step": 2580
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.969330855018587e-05,
      "loss": 0.007,
      "step": 2590
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.922862453531598e-05,
      "loss": 0.0074,
      "step": 2600
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.9009482316760635,
      "eval_f1": 0.9009482316760635,
      "eval_loss": 0.4252789318561554,
      "eval_precision": 0.9009482316760635,
      "eval_recall": 0.9009482316760635,
      "eval_runtime": 82.2003,
      "eval_samples_per_second": 94.939,
      "eval_steps_per_second": 11.873,
      "step": 2600
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.87639405204461e-05,
      "loss": 0.0127,
      "step": 2610
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.829925650557621e-05,
      "loss": 0.0508,
      "step": 2620
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.783457249070633e-05,
      "loss": 0.0183,
      "step": 2630
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.736988847583644e-05,
      "loss": 0.0127,
      "step": 2640
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.690520446096655e-05,
      "loss": 0.0141,
      "step": 2650
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.644052044609665e-05,
      "loss": 0.0328,
      "step": 2660
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.597583643122676e-05,
      "loss": 0.057,
      "step": 2670
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.551115241635688e-05,
      "loss": 0.025,
      "step": 2680
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.504646840148699e-05,
      "loss": 0.0464,
      "step": 2690
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.458178438661711e-05,
      "loss": 0.079,
      "step": 2700
    },
    {
      "epoch": 2.51,
      "eval_accuracy": 0.9012045105074321,
      "eval_f1": 0.9012045105074321,
      "eval_loss": 0.3814283013343811,
      "eval_precision": 0.9012045105074321,
      "eval_recall": 0.9012045105074321,
      "eval_runtime": 81.7211,
      "eval_samples_per_second": 95.496,
      "eval_steps_per_second": 11.943,
      "step": 2700
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.411710037174722e-05,
      "loss": 0.0409,
      "step": 2710
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.365241635687733e-05,
      "loss": 0.024,
      "step": 2720
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.318773234200745e-05,
      "loss": 0.0124,
      "step": 2730
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.272304832713755e-05,
      "loss": 0.0755,
      "step": 2740
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.225836431226766e-05,
      "loss": 0.0113,
      "step": 2750
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.179368029739777e-05,
      "loss": 0.0431,
      "step": 2760
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.132899628252788e-05,
      "loss": 0.0121,
      "step": 2770
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.0864312267658e-05,
      "loss": 0.008,
      "step": 2780
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.03996282527881e-05,
      "loss": 0.0195,
      "step": 2790
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.993494423791823e-05,
      "loss": 0.0407,
      "step": 2800
    },
    {
      "epoch": 2.6,
      "eval_accuracy": 0.9070989236289082,
      "eval_f1": 0.9070989236289082,
      "eval_loss": 0.3968299627304077,
      "eval_precision": 0.9070989236289082,
      "eval_recall": 0.9070989236289082,
      "eval_runtime": 82.0475,
      "eval_samples_per_second": 95.116,
      "eval_steps_per_second": 11.896,
      "step": 2800
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.947026022304833e-05,
      "loss": 0.0083,
      "step": 2810
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.900557620817844e-05,
      "loss": 0.02,
      "step": 2820
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.854089219330855e-05,
      "loss": 0.0322,
      "step": 2830
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.807620817843866e-05,
      "loss": 0.0441,
      "step": 2840
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.761152416356878e-05,
      "loss": 0.018,
      "step": 2850
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.714684014869889e-05,
      "loss": 0.014,
      "step": 2860
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.668215613382899e-05,
      "loss": 0.0025,
      "step": 2870
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.621747211895911e-05,
      "loss": 0.003,
      "step": 2880
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.575278810408922e-05,
      "loss": 0.0302,
      "step": 2890
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.528810408921934e-05,
      "loss": 0.0096,
      "step": 2900
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 0.9046642747309073,
      "eval_f1": 0.9046642747309073,
      "eval_loss": 0.43183794617652893,
      "eval_precision": 0.9046642747309073,
      "eval_recall": 0.9046642747309073,
      "eval_runtime": 81.009,
      "eval_samples_per_second": 96.335,
      "eval_steps_per_second": 12.048,
      "step": 2900
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.482342007434945e-05,
      "loss": 0.002,
      "step": 2910
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.435873605947956e-05,
      "loss": 0.0603,
      "step": 2920
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.389405204460967e-05,
      "loss": 0.0405,
      "step": 2930
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.342936802973977e-05,
      "loss": 0.0356,
      "step": 2940
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.29646840148699e-05,
      "loss": 0.0019,
      "step": 2950
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.25e-05,
      "loss": 0.0698,
      "step": 2960
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.203531598513012e-05,
      "loss": 0.0305,
      "step": 2970
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.157063197026023e-05,
      "loss": 0.0052,
      "step": 2980
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.110594795539034e-05,
      "loss": 0.1099,
      "step": 2990
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.064126394052045e-05,
      "loss": 0.052,
      "step": 3000
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 0.8986417221937468,
      "eval_f1": 0.8986417221937467,
      "eval_loss": 0.4111975431442261,
      "eval_precision": 0.8986417221937468,
      "eval_recall": 0.8986417221937468,
      "eval_runtime": 81.103,
      "eval_samples_per_second": 96.223,
      "eval_steps_per_second": 12.034,
      "step": 3000
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.017657992565055e-05,
      "loss": 0.0631,
      "step": 3010
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.971189591078067e-05,
      "loss": 0.0278,
      "step": 3020
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.924721189591078e-05,
      "loss": 0.0268,
      "step": 3030
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.8782527881040895e-05,
      "loss": 0.0075,
      "step": 3040
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.831784386617101e-05,
      "loss": 0.0099,
      "step": 3050
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.785315985130112e-05,
      "loss": 0.006,
      "step": 3060
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.738847583643123e-05,
      "loss": 0.0424,
      "step": 3070
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.6923791821561346e-05,
      "loss": 0.003,
      "step": 3080
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.645910780669146e-05,
      "loss": 0.0343,
      "step": 3090
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.599442379182156e-05,
      "loss": 0.0075,
      "step": 3100
    },
    {
      "epoch": 2.88,
      "eval_accuracy": 0.902101486417222,
      "eval_f1": 0.902101486417222,
      "eval_loss": 0.4230574071407318,
      "eval_precision": 0.902101486417222,
      "eval_recall": 0.902101486417222,
      "eval_runtime": 81.6651,
      "eval_samples_per_second": 95.561,
      "eval_steps_per_second": 11.951,
      "step": 3100
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.5529739776951675e-05,
      "loss": 0.0509,
      "step": 3110
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.506505576208178e-05,
      "loss": 0.0022,
      "step": 3120
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.46003717472119e-05,
      "loss": 0.0023,
      "step": 3130
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.413568773234201e-05,
      "loss": 0.0193,
      "step": 3140
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.3671003717472126e-05,
      "loss": 0.0183,
      "step": 3150
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.320631970260223e-05,
      "loss": 0.015,
      "step": 3160
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.274163568773235e-05,
      "loss": 0.0223,
      "step": 3170
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.227695167286245e-05,
      "loss": 0.0057,
      "step": 3180
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.181226765799256e-05,
      "loss": 0.0336,
      "step": 3190
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.134758364312268e-05,
      "loss": 0.0183,
      "step": 3200
    },
    {
      "epoch": 2.97,
      "eval_accuracy": 0.89633521271143,
      "eval_f1": 0.89633521271143,
      "eval_loss": 0.41062062978744507,
      "eval_precision": 0.89633521271143,
      "eval_recall": 0.89633521271143,
      "eval_runtime": 81.3975,
      "eval_samples_per_second": 95.875,
      "eval_steps_per_second": 11.991,
      "step": 3200
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.088289962825279e-05,
      "loss": 0.0211,
      "step": 3210
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.04182156133829e-05,
      "loss": 0.0057,
      "step": 3220
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.995353159851301e-05,
      "loss": 0.0015,
      "step": 3230
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.948884758364313e-05,
      "loss": 0.0021,
      "step": 3240
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.9024163568773235e-05,
      "loss": 0.0027,
      "step": 3250
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.855947955390335e-05,
      "loss": 0.0279,
      "step": 3260
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.8094795539033464e-05,
      "loss": 0.0402,
      "step": 3270
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.763011152416357e-05,
      "loss": 0.0024,
      "step": 3280
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.716542750929368e-05,
      "loss": 0.0025,
      "step": 3290
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.670074349442379e-05,
      "loss": 0.0134,
      "step": 3300
    },
    {
      "epoch": 3.07,
      "eval_accuracy": 0.9031266017426961,
      "eval_f1": 0.903126601742696,
      "eval_loss": 0.42098259925842285,
      "eval_precision": 0.9031266017426961,
      "eval_recall": 0.9031266017426961,
      "eval_runtime": 81.218,
      "eval_samples_per_second": 96.087,
      "eval_steps_per_second": 12.017,
      "step": 3300
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.623605947955391e-05,
      "loss": 0.0009,
      "step": 3310
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.577137546468402e-05,
      "loss": 0.001,
      "step": 3320
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.530669144981413e-05,
      "loss": 0.0018,
      "step": 3330
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.484200743494424e-05,
      "loss": 0.0077,
      "step": 3340
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.437732342007435e-05,
      "loss": 0.0008,
      "step": 3350
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.3912639405204465e-05,
      "loss": 0.004,
      "step": 3360
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.344795539033458e-05,
      "loss": 0.0016,
      "step": 3370
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.298327137546469e-05,
      "loss": 0.0009,
      "step": 3380
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.2518587360594795e-05,
      "loss": 0.0225,
      "step": 3390
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.205390334572491e-05,
      "loss": 0.0387,
      "step": 3400
    },
    {
      "epoch": 3.16,
      "eval_accuracy": 0.9041517170681702,
      "eval_f1": 0.9041517170681702,
      "eval_loss": 0.4335762858390808,
      "eval_precision": 0.9041517170681702,
      "eval_recall": 0.9041517170681702,
      "eval_runtime": 81.7574,
      "eval_samples_per_second": 95.453,
      "eval_steps_per_second": 11.938,
      "step": 3400
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.158921933085502e-05,
      "loss": 0.0013,
      "step": 3410
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.112453531598513e-05,
      "loss": 0.0237,
      "step": 3420
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.0659851301115245e-05,
      "loss": 0.0031,
      "step": 3430
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.019516728624535e-05,
      "loss": 0.0088,
      "step": 3440
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.973048327137547e-05,
      "loss": 0.0021,
      "step": 3450
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9265799256505575e-05,
      "loss": 0.001,
      "step": 3460
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.880111524163569e-05,
      "loss": 0.0014,
      "step": 3470
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.83364312267658e-05,
      "loss": 0.0039,
      "step": 3480
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.787174721189591e-05,
      "loss": 0.0047,
      "step": 3490
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.7407063197026025e-05,
      "loss": 0.001,
      "step": 3500
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.8987698616094311,
      "eval_f1": 0.8987698616094311,
      "eval_loss": 0.46794891357421875,
      "eval_precision": 0.8987698616094311,
      "eval_recall": 0.8987698616094311,
      "eval_runtime": 81.1297,
      "eval_samples_per_second": 96.192,
      "eval_steps_per_second": 12.03,
      "step": 3500
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.694237918215613e-05,
      "loss": 0.0014,
      "step": 3510
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.647769516728625e-05,
      "loss": 0.0008,
      "step": 3520
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.601301115241636e-05,
      "loss": 0.0011,
      "step": 3530
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.5548327137546475e-05,
      "loss": 0.0007,
      "step": 3540
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.5083643122676576e-05,
      "loss": 0.0007,
      "step": 3550
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.461895910780669e-05,
      "loss": 0.0211,
      "step": 3560
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4154275092936805e-05,
      "loss": 0.0406,
      "step": 3570
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.368959107806692e-05,
      "loss": 0.0053,
      "step": 3580
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.322490706319703e-05,
      "loss": 0.0038,
      "step": 3590
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.276022304832714e-05,
      "loss": 0.0292,
      "step": 3600
    },
    {
      "epoch": 3.35,
      "eval_accuracy": 0.8976166068682727,
      "eval_f1": 0.8976166068682727,
      "eval_loss": 0.4691353440284729,
      "eval_precision": 0.8976166068682727,
      "eval_recall": 0.8976166068682727,
      "eval_runtime": 81.6274,
      "eval_samples_per_second": 95.605,
      "eval_steps_per_second": 11.957,
      "step": 3600
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.229553903345725e-05,
      "loss": 0.0015,
      "step": 3610
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.183085501858736e-05,
      "loss": 0.0019,
      "step": 3620
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.136617100371748e-05,
      "loss": 0.007,
      "step": 3630
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.0901486988847585e-05,
      "loss": 0.0037,
      "step": 3640
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.0436802973977696e-05,
      "loss": 0.0025,
      "step": 3650
    },
    {
      "epoch": 3.4,
      "learning_rate": 2.997211895910781e-05,
      "loss": 0.0012,
      "step": 3660
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.950743494423792e-05,
      "loss": 0.0012,
      "step": 3670
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.904275092936803e-05,
      "loss": 0.0011,
      "step": 3680
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.8578066914498143e-05,
      "loss": 0.0006,
      "step": 3690
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8113382899628254e-05,
      "loss": 0.0109,
      "step": 3700
    },
    {
      "epoch": 3.44,
      "eval_accuracy": 0.9060738083034341,
      "eval_f1": 0.9060738083034341,
      "eval_loss": 0.47132211923599243,
      "eval_precision": 0.9060738083034341,
      "eval_recall": 0.9060738083034341,
      "eval_runtime": 81.3336,
      "eval_samples_per_second": 95.951,
      "eval_steps_per_second": 12.0,
      "step": 3700
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.7648698884758368e-05,
      "loss": 0.0007,
      "step": 3710
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.7184014869888475e-05,
      "loss": 0.0006,
      "step": 3720
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6719330855018586e-05,
      "loss": 0.0006,
      "step": 3730
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.62546468401487e-05,
      "loss": 0.0014,
      "step": 3740
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.578996282527881e-05,
      "loss": 0.0006,
      "step": 3750
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5325278810408926e-05,
      "loss": 0.0018,
      "step": 3760
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.4860594795539037e-05,
      "loss": 0.0011,
      "step": 3770
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4395910780669144e-05,
      "loss": 0.0006,
      "step": 3780
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.393122676579926e-05,
      "loss": 0.0156,
      "step": 3790
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.346654275092937e-05,
      "loss": 0.0007,
      "step": 3800
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.9062019477191184,
      "eval_f1": 0.9062019477191184,
      "eval_loss": 0.48416757583618164,
      "eval_precision": 0.9062019477191184,
      "eval_recall": 0.9062019477191184,
      "eval_runtime": 81.9896,
      "eval_samples_per_second": 95.183,
      "eval_steps_per_second": 11.904,
      "step": 3800
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.300185873605948e-05,
      "loss": 0.0007,
      "step": 3810
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.253717472118959e-05,
      "loss": 0.0007,
      "step": 3820
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2072490706319702e-05,
      "loss": 0.0027,
      "step": 3830
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.1607806691449813e-05,
      "loss": 0.0008,
      "step": 3840
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.1143122676579928e-05,
      "loss": 0.0007,
      "step": 3850
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.067843866171004e-05,
      "loss": 0.0005,
      "step": 3860
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.021375464684015e-05,
      "loss": 0.0153,
      "step": 3870
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9749070631970264e-05,
      "loss": 0.0006,
      "step": 3880
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.928438661710037e-05,
      "loss": 0.0006,
      "step": 3890
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.8819702602230486e-05,
      "loss": 0.0023,
      "step": 3900
    },
    {
      "epoch": 3.62,
      "eval_accuracy": 0.9041517170681702,
      "eval_f1": 0.9041517170681702,
      "eval_loss": 0.4973074495792389,
      "eval_precision": 0.9041517170681702,
      "eval_recall": 0.9041517170681702,
      "eval_runtime": 81.5112,
      "eval_samples_per_second": 95.741,
      "eval_steps_per_second": 11.974,
      "step": 3900
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8355018587360597e-05,
      "loss": 0.0006,
      "step": 3910
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.7890334572490707e-05,
      "loss": 0.0006,
      "step": 3920
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.742565055762082e-05,
      "loss": 0.0008,
      "step": 3930
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.696096654275093e-05,
      "loss": 0.0005,
      "step": 3940
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.649628252788104e-05,
      "loss": 0.0006,
      "step": 3950
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.6031598513011155e-05,
      "loss": 0.0005,
      "step": 3960
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5566914498141262e-05,
      "loss": 0.0011,
      "step": 3970
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5102230483271376e-05,
      "loss": 0.0005,
      "step": 3980
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4637546468401489e-05,
      "loss": 0.0006,
      "step": 3990
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.4172862453531598e-05,
      "loss": 0.0039,
      "step": 4000
    },
    {
      "epoch": 3.72,
      "eval_accuracy": 0.9042798564838544,
      "eval_f1": 0.9042798564838544,
      "eval_loss": 0.4993748068809509,
      "eval_precision": 0.9042798564838544,
      "eval_recall": 0.9042798564838544,
      "eval_runtime": 81.7239,
      "eval_samples_per_second": 95.492,
      "eval_steps_per_second": 11.943,
      "step": 4000
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3708178438661713e-05,
      "loss": 0.0009,
      "step": 4010
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.3243494423791822e-05,
      "loss": 0.0012,
      "step": 4020
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2778810408921934e-05,
      "loss": 0.0183,
      "step": 4030
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2314126394052045e-05,
      "loss": 0.0066,
      "step": 4040
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.1849442379182158e-05,
      "loss": 0.0008,
      "step": 4050
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1384758364312269e-05,
      "loss": 0.0005,
      "step": 4060
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.092007434944238e-05,
      "loss": 0.0006,
      "step": 4070
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0455390334572492e-05,
      "loss": 0.001,
      "step": 4080
    },
    {
      "epoch": 3.8,
      "learning_rate": 9.990706319702603e-06,
      "loss": 0.0005,
      "step": 4090
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.526022304832714e-06,
      "loss": 0.0005,
      "step": 4100
    },
    {
      "epoch": 3.81,
      "eval_accuracy": 0.9059456688877499,
      "eval_f1": 0.9059456688877499,
      "eval_loss": 0.4907146096229553,
      "eval_precision": 0.9059456688877499,
      "eval_recall": 0.9059456688877499,
      "eval_runtime": 81.779,
      "eval_samples_per_second": 95.428,
      "eval_steps_per_second": 11.935,
      "step": 4100
    },
    {
      "epoch": 3.82,
      "learning_rate": 9.061338289962825e-06,
      "loss": 0.0019,
      "step": 4110
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.596654275092938e-06,
      "loss": 0.0099,
      "step": 4120
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.131970260223049e-06,
      "loss": 0.0005,
      "step": 4130
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.66728624535316e-06,
      "loss": 0.0005,
      "step": 4140
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.202602230483271e-06,
      "loss": 0.0005,
      "step": 4150
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.737918215613382e-06,
      "loss": 0.0005,
      "step": 4160
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.273234200743494e-06,
      "loss": 0.0014,
      "step": 4170
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.808550185873606e-06,
      "loss": 0.0005,
      "step": 4180
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.343866171003718e-06,
      "loss": 0.0005,
      "step": 4190
    },
    {
      "epoch": 3.9,
      "learning_rate": 4.879182156133829e-06,
      "loss": 0.0005,
      "step": 4200
    },
    {
      "epoch": 3.9,
      "eval_accuracy": 0.9058175294720656,
      "eval_f1": 0.9058175294720656,
      "eval_loss": 0.4918811023235321,
      "eval_precision": 0.9058175294720656,
      "eval_recall": 0.9058175294720656,
      "eval_runtime": 82.9585,
      "eval_samples_per_second": 94.071,
      "eval_steps_per_second": 11.765,
      "step": 4200
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.414498141263941e-06,
      "loss": 0.0007,
      "step": 4210
    },
    {
      "epoch": 3.92,
      "learning_rate": 3.949814126394052e-06,
      "loss": 0.0005,
      "step": 4220
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.485130111524164e-06,
      "loss": 0.0006,
      "step": 4230
    },
    {
      "epoch": 3.94,
      "learning_rate": 3.020446096654275e-06,
      "loss": 0.0007,
      "step": 4240
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.5557620817843865e-06,
      "loss": 0.0005,
      "step": 4250
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.0910780669144983e-06,
      "loss": 0.0006,
      "step": 4260
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.6263940520446099e-06,
      "loss": 0.0028,
      "step": 4270
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.1617100371747212e-06,
      "loss": 0.0004,
      "step": 4280
    },
    {
      "epoch": 3.99,
      "learning_rate": 6.970260223048328e-07,
      "loss": 0.0004,
      "step": 4290
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.3234200743494425e-07,
      "loss": 0.0432,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9059456688877499,
      "eval_f1": 0.9059456688877499,
      "eval_loss": 0.49230173230171204,
      "eval_precision": 0.9059456688877499,
      "eval_recall": 0.9059456688877499,
      "eval_runtime": 81.5662,
      "eval_samples_per_second": 95.677,
      "eval_steps_per_second": 11.966,
      "step": 4300
    },
    {
      "epoch": 4.0,
      "step": 4304,
      "total_flos": 1.899937193661481e+19,
      "train_loss": 0.10234801637709443,
      "train_runtime": 5842.5525,
      "train_samples_per_second": 11.785,
      "train_steps_per_second": 0.737
    }
  ],
  "max_steps": 4304,
  "num_train_epochs": 4,
  "total_flos": 1.899937193661481e+19,
  "trial_name": null,
  "trial_params": null
}