{
  "best_metric": 0.2583008110523224,
  "best_model_checkpoint": "./vit-base-binary-isic-sharpened-patch-16/checkpoint-1400",
  "epoch": 4.0,
  "global_step": 4328,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.00019953789279112755,
      "loss": 0.5548,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0001990757855822551,
      "loss": 0.4224,
      "step": 20
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019861367837338263,
      "loss": 0.3661,
      "step": 30
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019815157116451017,
      "loss": 0.3027,
      "step": 40
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0001976894639556377,
      "loss": 0.7411,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019722735674676528,
      "loss": 0.4699,
      "step": 60
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0001967652495378928,
      "loss": 0.4816,
      "step": 70
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019630314232902034,
      "loss": 0.3819,
      "step": 80
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019584103512014788,
      "loss": 0.3487,
      "step": 90
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019537892791127544,
      "loss": 0.3281,
      "step": 100
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.818298308559713,
      "eval_f1": 0.8182983085597131,
      "eval_loss": 0.4381178617477417,
      "eval_precision": 0.818298308559713,
      "eval_recall": 0.818298308559713,
      "eval_runtime": 68.2283,
      "eval_samples_per_second": 114.381,
      "eval_steps_per_second": 14.305,
      "step": 100
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019491682070240296,
      "loss": 0.4754,
      "step": 110
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0001944547134935305,
      "loss": 0.344,
      "step": 120
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019399260628465804,
      "loss": 0.4745,
      "step": 130
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0001935304990757856,
      "loss": 0.3565,
      "step": 140
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019306839186691312,
      "loss": 0.3888,
      "step": 150
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019260628465804066,
      "loss": 0.2515,
      "step": 160
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019214417744916823,
      "loss": 0.3237,
      "step": 170
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019168207024029577,
      "loss": 0.2789,
      "step": 180
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0001912199630314233,
      "loss": 0.3318,
      "step": 190
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019075785582255082,
      "loss": 0.3212,
      "step": 200
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.8503331624807791,
      "eval_f1": 0.8503331624807791,
      "eval_loss": 0.31786420941352844,
      "eval_precision": 0.8503331624807791,
      "eval_recall": 0.8503331624807791,
      "eval_runtime": 68.6097,
      "eval_samples_per_second": 113.745,
      "eval_steps_per_second": 14.225,
      "step": 200
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0001902957486136784,
      "loss": 0.29,
      "step": 210
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018983364140480593,
      "loss": 0.4371,
      "step": 220
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018937153419593347,
      "loss": 0.2557,
      "step": 230
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.000188909426987061,
      "loss": 0.3501,
      "step": 240
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018844731977818855,
      "loss": 0.3539,
      "step": 250
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0001879852125693161,
      "loss": 0.2254,
      "step": 260
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00018752310536044364,
      "loss": 0.3563,
      "step": 270
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00018706099815157118,
      "loss": 0.3356,
      "step": 280
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018659889094269872,
      "loss": 0.2957,
      "step": 290
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018613678373382626,
      "loss": 0.2864,
      "step": 300
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.8654536135315223,
      "eval_f1": 0.8654536135315223,
      "eval_loss": 0.31259891390800476,
      "eval_precision": 0.8654536135315223,
      "eval_recall": 0.8654536135315223,
      "eval_runtime": 68.4623,
      "eval_samples_per_second": 113.99,
      "eval_steps_per_second": 14.256,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0001856746765249538,
      "loss": 0.4686,
      "step": 310
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018521256931608134,
      "loss": 0.2769,
      "step": 320
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018475046210720888,
      "loss": 0.4208,
      "step": 330
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018428835489833642,
      "loss": 0.2864,
      "step": 340
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018382624768946396,
      "loss": 0.2303,
      "step": 350
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0001833641404805915,
      "loss": 0.4072,
      "step": 360
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018290203327171904,
      "loss": 0.2665,
      "step": 370
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018243992606284658,
      "loss": 0.4042,
      "step": 380
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018197781885397413,
      "loss": 0.3958,
      "step": 390
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0001815157116451017,
      "loss": 0.2692,
      "step": 400
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.8599436186570989,
      "eval_f1": 0.8599436186570989,
      "eval_loss": 0.32165130972862244,
      "eval_precision": 0.8599436186570989,
      "eval_recall": 0.8599436186570989,
      "eval_runtime": 68.4788,
      "eval_samples_per_second": 113.962,
      "eval_steps_per_second": 14.253,
      "step": 400
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0001810536044362292,
      "loss": 0.2838,
      "step": 410
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018059149722735675,
      "loss": 0.2625,
      "step": 420
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0001801293900184843,
      "loss": 0.2881,
      "step": 430
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017966728280961186,
      "loss": 0.3115,
      "step": 440
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017920517560073937,
      "loss": 0.2082,
      "step": 450
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0001787430683918669,
      "loss": 0.289,
      "step": 460
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00017828096118299448,
      "loss": 0.2381,
      "step": 470
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017781885397412202,
      "loss": 0.2562,
      "step": 480
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017735674676524953,
      "loss": 0.2725,
      "step": 490
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017689463955637707,
      "loss": 0.3195,
      "step": 500
    },
    {
      "epoch": 0.46,
      "eval_accuracy": 0.8694259354177345,
      "eval_f1": 0.8694259354177345,
      "eval_loss": 0.3061072826385498,
      "eval_precision": 0.8694259354177345,
      "eval_recall": 0.8694259354177345,
      "eval_runtime": 68.1818,
      "eval_samples_per_second": 114.459,
      "eval_steps_per_second": 14.315,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00017643253234750464,
      "loss": 0.2288,
      "step": 510
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00017597042513863218,
      "loss": 0.398,
      "step": 520
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0001755083179297597,
      "loss": 0.2576,
      "step": 530
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00017504621072088724,
      "loss": 0.2785,
      "step": 540
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0001745841035120148,
      "loss": 0.2616,
      "step": 550
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00017412199630314234,
      "loss": 0.3359,
      "step": 560
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017365988909426989,
      "loss": 0.1817,
      "step": 570
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017319778188539743,
      "loss": 0.281,
      "step": 580
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00017273567467652497,
      "loss": 0.2725,
      "step": 590
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0001722735674676525,
      "loss": 0.2095,
      "step": 600
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.8668631471040492,
      "eval_f1": 0.8668631471040492,
      "eval_loss": 0.29103726148605347,
      "eval_precision": 0.8668631471040492,
      "eval_recall": 0.8668631471040492,
      "eval_runtime": 68.1241,
      "eval_samples_per_second": 114.556,
      "eval_steps_per_second": 14.327,
      "step": 600
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017181146025878005,
      "loss": 0.3749,
      "step": 610
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0001713493530499076,
      "loss": 0.224,
      "step": 620
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017088724584103513,
      "loss": 0.2121,
      "step": 630
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017042513863216267,
      "loss": 0.3307,
      "step": 640
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0001699630314232902,
      "loss": 0.2623,
      "step": 650
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016950092421441775,
      "loss": 0.2967,
      "step": 660
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0001690388170055453,
      "loss": 0.1768,
      "step": 670
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00016857670979667283,
      "loss": 0.3169,
      "step": 680
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016811460258780037,
      "loss": 0.2648,
      "step": 690
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00016765249537892791,
      "loss": 0.2168,
      "step": 700
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.8730138390568939,
      "eval_f1": 0.8730138390568939,
      "eval_loss": 0.32482409477233887,
      "eval_precision": 0.8730138390568939,
      "eval_recall": 0.8730138390568939,
      "eval_runtime": 68.2733,
      "eval_samples_per_second": 114.305,
      "eval_steps_per_second": 14.295,
      "step": 700
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016719038817005546,
      "loss": 0.2543,
      "step": 710
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.000166728280961183,
      "loss": 0.2349,
      "step": 720
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016626617375231054,
      "loss": 0.2346,
      "step": 730
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00016580406654343808,
      "loss": 0.2432,
      "step": 740
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016534195933456562,
      "loss": 0.3282,
      "step": 750
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016487985212569316,
      "loss": 0.2576,
      "step": 760
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016441774491682073,
      "loss": 0.2225,
      "step": 770
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016395563770794827,
      "loss": 0.1587,
      "step": 780
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00016349353049907578,
      "loss": 0.284,
      "step": 790
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016303142329020332,
      "loss": 0.2288,
      "step": 800
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 0.8553305996924654,
      "eval_f1": 0.8553305996924655,
      "eval_loss": 0.306738018989563,
      "eval_precision": 0.8553305996924654,
      "eval_recall": 0.8553305996924654,
      "eval_runtime": 68.2633,
      "eval_samples_per_second": 114.322,
      "eval_steps_per_second": 14.298,
      "step": 800
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0001625693160813309,
      "loss": 0.2511,
      "step": 810
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016210720887245843,
      "loss": 0.2133,
      "step": 820
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016164510166358594,
      "loss": 0.1996,
      "step": 830
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016118299445471348,
      "loss": 0.2501,
      "step": 840
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016072088724584105,
      "loss": 0.1805,
      "step": 850
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0001602587800369686,
      "loss": 0.1993,
      "step": 860
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0001597966728280961,
      "loss": 0.3014,
      "step": 870
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015933456561922367,
      "loss": 0.3067,
      "step": 880
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00015887245841035122,
      "loss": 0.1731,
      "step": 890
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015841035120147876,
      "loss": 0.2521,
      "step": 900
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.8689133777549974,
      "eval_f1": 0.8689133777549973,
      "eval_loss": 0.27230527997016907,
      "eval_precision": 0.8689133777549974,
      "eval_recall": 0.8689133777549974,
      "eval_runtime": 68.0157,
      "eval_samples_per_second": 114.738,
      "eval_steps_per_second": 14.35,
      "step": 900
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015794824399260627,
      "loss": 0.1691,
      "step": 910
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00015748613678373384,
      "loss": 0.249,
      "step": 920
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015702402957486138,
      "loss": 0.1896,
      "step": 930
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00015656192236598892,
      "loss": 0.2891,
      "step": 940
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00015609981515711646,
      "loss": 0.3825,
      "step": 950
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.000155637707948244,
      "loss": 0.2239,
      "step": 960
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00015517560073937154,
      "loss": 0.2502,
      "step": 970
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015471349353049908,
      "loss": 0.268,
      "step": 980
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015425138632162662,
      "loss": 0.2304,
      "step": 990
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015378927911275416,
      "loss": 0.1953,
      "step": 1000
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 0.8723731419784726,
      "eval_f1": 0.8723731419784726,
      "eval_loss": 0.27285629510879517,
      "eval_precision": 0.8723731419784726,
      "eval_recall": 0.8723731419784726,
      "eval_runtime": 67.9159,
      "eval_samples_per_second": 114.907,
      "eval_steps_per_second": 14.371,
      "step": 1000
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0001533271719038817,
      "loss": 0.1812,
      "step": 1010
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015286506469500925,
      "loss": 0.2239,
      "step": 1020
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015240295748613679,
      "loss": 0.2245,
      "step": 1030
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00015194085027726433,
      "loss": 0.1914,
      "step": 1040
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00015147874306839187,
      "loss": 0.2178,
      "step": 1050
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0001510166358595194,
      "loss": 0.2064,
      "step": 1060
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015055452865064698,
      "loss": 0.1492,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0001500924214417745,
      "loss": 0.2063,
      "step": 1080
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014963031423290203,
      "loss": 0.1508,
      "step": 1090
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014916820702402957,
      "loss": 0.2845,
      "step": 1100
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.8666068682726806,
      "eval_f1": 0.8666068682726806,
      "eval_loss": 0.43923676013946533,
      "eval_precision": 0.8666068682726806,
      "eval_recall": 0.8666068682726806,
      "eval_runtime": 68.2732,
      "eval_samples_per_second": 114.305,
      "eval_steps_per_second": 14.296,
      "step": 1100
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00014870609981515714,
      "loss": 0.2509,
      "step": 1110
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014824399260628468,
      "loss": 0.2282,
      "step": 1120
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0001477818853974122,
      "loss": 0.1907,
      "step": 1130
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00014731977818853976,
      "loss": 0.2258,
      "step": 1140
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0001468576709796673,
      "loss": 0.223,
      "step": 1150
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00014639556377079484,
      "loss": 0.2137,
      "step": 1160
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014593345656192236,
      "loss": 0.2253,
      "step": 1170
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014547134935304992,
      "loss": 0.137,
      "step": 1180
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00014500924214417746,
      "loss": 0.1647,
      "step": 1190
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.000144547134935305,
      "loss": 0.1484,
      "step": 1200
    },
    {
      "epoch": 1.11,
      "eval_accuracy": 0.8883905689390056,
      "eval_f1": 0.8883905689390056,
      "eval_loss": 0.30305391550064087,
      "eval_precision": 0.8883905689390056,
      "eval_recall": 0.8883905689390056,
      "eval_runtime": 68.1495,
      "eval_samples_per_second": 114.513,
      "eval_steps_per_second": 14.321,
      "step": 1200
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014408502772643252,
      "loss": 0.1431,
      "step": 1210
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0001436229205175601,
      "loss": 0.2111,
      "step": 1220
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014316081330868763,
      "loss": 0.1222,
      "step": 1230
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00014269870609981517,
      "loss": 0.1183,
      "step": 1240
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0001422365988909427,
      "loss": 0.1697,
      "step": 1250
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014177449168207025,
      "loss": 0.1608,
      "step": 1260
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0001413123844731978,
      "loss": 0.1648,
      "step": 1270
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014085027726432533,
      "loss": 0.0847,
      "step": 1280
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014038817005545287,
      "loss": 0.1868,
      "step": 1290
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0001399260628465804,
      "loss": 0.153,
      "step": 1300
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.8991542798564839,
      "eval_f1": 0.8991542798564839,
      "eval_loss": 0.2849101126194,
      "eval_precision": 0.8991542798564839,
      "eval_recall": 0.8991542798564839,
      "eval_runtime": 68.0056,
      "eval_samples_per_second": 114.755,
      "eval_steps_per_second": 14.352,
      "step": 1300
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013946395563770795,
      "loss": 0.2958,
      "step": 1310
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001390018484288355,
      "loss": 0.1571,
      "step": 1320
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013853974121996306,
      "loss": 0.125,
      "step": 1330
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013807763401109058,
      "loss": 0.1282,
      "step": 1340
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013761552680221812,
      "loss": 0.1299,
      "step": 1350
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013715341959334566,
      "loss": 0.2503,
      "step": 1360
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00013669131238447322,
      "loss": 0.1461,
      "step": 1370
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013622920517560074,
      "loss": 0.1362,
      "step": 1380
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013576709796672828,
      "loss": 0.2611,
      "step": 1390
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013530499075785582,
      "loss": 0.1648,
      "step": 1400
    },
    {
      "epoch": 1.29,
      "eval_accuracy": 0.8912096360840595,
      "eval_f1": 0.8912096360840595,
      "eval_loss": 0.2583008110523224,
      "eval_precision": 0.8912096360840595,
      "eval_recall": 0.8912096360840595,
      "eval_runtime": 68.2491,
      "eval_samples_per_second": 114.346,
      "eval_steps_per_second": 14.301,
      "step": 1400
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0001348428835489834,
      "loss": 0.1683,
      "step": 1410
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0001343807763401109,
      "loss": 0.1355,
      "step": 1420
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00013391866913123844,
      "loss": 0.1657,
      "step": 1430
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.000133456561922366,
      "loss": 0.0833,
      "step": 1440
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013299445471349355,
      "loss": 0.1859,
      "step": 1450
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00013253234750462106,
      "loss": 0.1593,
      "step": 1460
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0001320702402957486,
      "loss": 0.1254,
      "step": 1470
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013160813308687617,
      "loss": 0.2359,
      "step": 1480
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013114602587800371,
      "loss": 0.1055,
      "step": 1490
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013068391866913125,
      "loss": 0.1627,
      "step": 1500
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.8932598667350077,
      "eval_f1": 0.8932598667350077,
      "eval_loss": 0.27056819200515747,
      "eval_precision": 0.8932598667350077,
      "eval_recall": 0.8932598667350077,
      "eval_runtime": 68.3811,
      "eval_samples_per_second": 114.125,
      "eval_steps_per_second": 14.273,
      "step": 1500
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00013022181146025877,
      "loss": 0.159,
      "step": 1510
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00012975970425138634,
      "loss": 0.124,
      "step": 1520
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012929759704251388,
      "loss": 0.1114,
      "step": 1530
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012883548983364142,
      "loss": 0.1002,
      "step": 1540
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012837338262476896,
      "loss": 0.1672,
      "step": 1550
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0001279112754158965,
      "loss": 0.1301,
      "step": 1560
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00012744916820702404,
      "loss": 0.1178,
      "step": 1570
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012698706099815158,
      "loss": 0.104,
      "step": 1580
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012652495378927912,
      "loss": 0.0748,
      "step": 1590
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00012606284658040666,
      "loss": 0.0943,
      "step": 1600
    },
    {
      "epoch": 1.48,
      "eval_accuracy": 0.9033828805740646,
      "eval_f1": 0.9033828805740646,
      "eval_loss": 0.27828386425971985,
      "eval_precision": 0.9033828805740646,
      "eval_recall": 0.9033828805740646,
      "eval_runtime": 67.7545,
      "eval_samples_per_second": 115.181,
      "eval_steps_per_second": 14.405,
      "step": 1600
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0001256007393715342,
      "loss": 0.0878,
      "step": 1610
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00012513863216266174,
      "loss": 0.1153,
      "step": 1620
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012467652495378928,
      "loss": 0.2319,
      "step": 1630
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00012421441774491682,
      "loss": 0.1918,
      "step": 1640
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00012375231053604437,
      "loss": 0.1432,
      "step": 1650
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0001232902033271719,
      "loss": 0.1487,
      "step": 1660
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00012282809611829945,
      "loss": 0.155,
      "step": 1670
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.000122365988909427,
      "loss": 0.1157,
      "step": 1680
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012190388170055453,
      "loss": 0.164,
      "step": 1690
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00012144177449168208,
      "loss": 0.0624,
      "step": 1700
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 0.8926191696565864,
      "eval_f1": 0.8926191696565864,
      "eval_loss": 0.2920725643634796,
      "eval_precision": 0.8926191696565864,
      "eval_recall": 0.8926191696565864,
      "eval_runtime": 68.1055,
      "eval_samples_per_second": 114.587,
      "eval_steps_per_second": 14.331,
      "step": 1700
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012097966728280962,
      "loss": 0.1088,
      "step": 1710
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012051756007393715,
      "loss": 0.167,
      "step": 1720
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0001200554528650647,
      "loss": 0.1346,
      "step": 1730
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00011959334565619225,
      "loss": 0.1275,
      "step": 1740
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011913123844731979,
      "loss": 0.0868,
      "step": 1750
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011866913123844731,
      "loss": 0.2667,
      "step": 1760
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011820702402957487,
      "loss": 0.1119,
      "step": 1770
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011774491682070241,
      "loss": 0.0471,
      "step": 1780
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011728280961182996,
      "loss": 0.1091,
      "step": 1790
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011682070240295748,
      "loss": 0.12,
      "step": 1800
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 0.9005638134290108,
      "eval_f1": 0.9005638134290108,
      "eval_loss": 0.29145559668540955,
      "eval_precision": 0.9005638134290108,
      "eval_recall": 0.9005638134290108,
      "eval_runtime": 68.3975,
      "eval_samples_per_second": 114.098,
      "eval_steps_per_second": 14.27,
      "step": 1800
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011635859519408503,
      "loss": 0.1147,
      "step": 1810
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011589648798521257,
      "loss": 0.1017,
      "step": 1820
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011543438077634013,
      "loss": 0.181,
      "step": 1830
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011497227356746765,
      "loss": 0.1062,
      "step": 1840
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0001145101663585952,
      "loss": 0.1605,
      "step": 1850
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011404805914972275,
      "loss": 0.0934,
      "step": 1860
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011358595194085029,
      "loss": 0.0757,
      "step": 1870
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011312384473197783,
      "loss": 0.0682,
      "step": 1880
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00011266173752310536,
      "loss": 0.1369,
      "step": 1890
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011219963031423291,
      "loss": 0.0735,
      "step": 1900
    },
    {
      "epoch": 1.76,
      "eval_accuracy": 0.8896719630958483,
      "eval_f1": 0.8896719630958483,
      "eval_loss": 0.3102613687515259,
      "eval_precision": 0.8896719630958483,
      "eval_recall": 0.8896719630958483,
      "eval_runtime": 68.2707,
      "eval_samples_per_second": 114.31,
      "eval_steps_per_second": 14.296,
      "step": 1900
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011173752310536045,
      "loss": 0.1849,
      "step": 1910
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.000111275415896488,
      "loss": 0.1584,
      "step": 1920
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011081330868761552,
      "loss": 0.1299,
      "step": 1930
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011035120147874307,
      "loss": 0.1857,
      "step": 1940
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00010988909426987061,
      "loss": 0.1487,
      "step": 1950
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010942698706099817,
      "loss": 0.1121,
      "step": 1960
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0001089648798521257,
      "loss": 0.0669,
      "step": 1970
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010850277264325324,
      "loss": 0.1051,
      "step": 1980
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010804066543438078,
      "loss": 0.1454,
      "step": 1990
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00010757855822550833,
      "loss": 0.0609,
      "step": 2000
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.8971040492055357,
      "eval_f1": 0.8971040492055357,
      "eval_loss": 0.3381568193435669,
      "eval_precision": 0.8971040492055357,
      "eval_recall": 0.8971040492055357,
      "eval_runtime": 68.0421,
      "eval_samples_per_second": 114.694,
      "eval_steps_per_second": 14.344,
      "step": 2000
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010711645101663586,
      "loss": 0.0916,
      "step": 2010
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0001066543438077634,
      "loss": 0.074,
      "step": 2020
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010619223659889095,
      "loss": 0.0809,
      "step": 2030
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0001057301293900185,
      "loss": 0.1194,
      "step": 2040
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010526802218114602,
      "loss": 0.0866,
      "step": 2050
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010480591497227356,
      "loss": 0.1301,
      "step": 2060
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010434380776340112,
      "loss": 0.1557,
      "step": 2070
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010388170055452866,
      "loss": 0.1552,
      "step": 2080
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010341959334565621,
      "loss": 0.1308,
      "step": 2090
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010295748613678373,
      "loss": 0.1645,
      "step": 2100
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 0.8900563813429011,
      "eval_f1": 0.8900563813429011,
      "eval_loss": 0.26747891306877136,
      "eval_precision": 0.8900563813429011,
      "eval_recall": 0.8900563813429011,
      "eval_runtime": 68.0572,
      "eval_samples_per_second": 114.668,
      "eval_steps_per_second": 14.341,
      "step": 2100
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010254158964879852,
      "loss": 0.1046,
      "step": 2110
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010207948243992607,
      "loss": 0.165,
      "step": 2120
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010161737523105361,
      "loss": 0.0636,
      "step": 2130
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010115526802218117,
      "loss": 0.0757,
      "step": 2140
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010069316081330868,
      "loss": 0.1345,
      "step": 2150
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010023105360443623,
      "loss": 0.0419,
      "step": 2160
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.976894639556378e-05,
      "loss": 0.0908,
      "step": 2170
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.930683918669132e-05,
      "loss": 0.0179,
      "step": 2180
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.884473197781886e-05,
      "loss": 0.1491,
      "step": 2190
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.83826247689464e-05,
      "loss": 0.0839,
      "step": 2200
    },
    {
      "epoch": 2.03,
      "eval_accuracy": 0.8962070732957458,
      "eval_f1": 0.8962070732957458,
      "eval_loss": 0.3940969705581665,
      "eval_precision": 0.8962070732957458,
      "eval_recall": 0.8962070732957458,
      "eval_runtime": 68.4541,
      "eval_samples_per_second": 114.003,
      "eval_steps_per_second": 14.258,
      "step": 2200
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.792051756007394e-05,
      "loss": 0.0517,
      "step": 2210
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.745841035120148e-05,
      "loss": 0.0144,
      "step": 2220
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.699630314232902e-05,
      "loss": 0.0761,
      "step": 2230
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.653419593345656e-05,
      "loss": 0.0355,
      "step": 2240
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.607208872458411e-05,
      "loss": 0.0046,
      "step": 2250
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.560998151571166e-05,
      "loss": 0.0475,
      "step": 2260
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.51478743068392e-05,
      "loss": 0.0351,
      "step": 2270
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.468576709796674e-05,
      "loss": 0.0597,
      "step": 2280
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.422365988909428e-05,
      "loss": 0.0157,
      "step": 2290
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.376155268022182e-05,
      "loss": 0.0571,
      "step": 2300
    },
    {
      "epoch": 2.13,
      "eval_accuracy": 0.9046642747309073,
      "eval_f1": 0.9046642747309073,
      "eval_loss": 0.3887563943862915,
      "eval_precision": 0.9046642747309073,
      "eval_recall": 0.9046642747309073,
      "eval_runtime": 68.596,
      "eval_samples_per_second": 113.768,
      "eval_steps_per_second": 14.228,
      "step": 2300
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.329944547134936e-05,
      "loss": 0.0338,
      "step": 2310
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.28373382624769e-05,
      "loss": 0.0569,
      "step": 2320
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.237523105360444e-05,
      "loss": 0.0047,
      "step": 2330
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.191312384473198e-05,
      "loss": 0.0603,
      "step": 2340
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.145101663585952e-05,
      "loss": 0.0247,
      "step": 2350
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.098890942698706e-05,
      "loss": 0.0375,
      "step": 2360
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.05268022181146e-05,
      "loss": 0.0259,
      "step": 2370
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.006469500924214e-05,
      "loss": 0.1091,
      "step": 2380
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.960258780036968e-05,
      "loss": 0.0454,
      "step": 2390
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.914048059149724e-05,
      "loss": 0.0929,
      "step": 2400
    },
    {
      "epoch": 2.22,
      "eval_accuracy": 0.9009482316760635,
      "eval_f1": 0.9009482316760635,
      "eval_loss": 0.37732452154159546,
      "eval_precision": 0.9009482316760635,
      "eval_recall": 0.9009482316760635,
      "eval_runtime": 68.175,
      "eval_samples_per_second": 114.47,
      "eval_steps_per_second": 14.316,
      "step": 2400
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.867837338262477e-05,
      "loss": 0.0662,
      "step": 2410
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.821626617375232e-05,
      "loss": 0.0249,
      "step": 2420
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.775415896487985e-05,
      "loss": 0.0774,
      "step": 2430
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.72920517560074e-05,
      "loss": 0.022,
      "step": 2440
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.682994454713494e-05,
      "loss": 0.0128,
      "step": 2450
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.636783733826248e-05,
      "loss": 0.0056,
      "step": 2460
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.590573012939002e-05,
      "loss": 0.052,
      "step": 2470
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.544362292051756e-05,
      "loss": 0.0087,
      "step": 2480
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.49815157116451e-05,
      "loss": 0.0552,
      "step": 2490
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.451940850277265e-05,
      "loss": 0.0378,
      "step": 2500
    },
    {
      "epoch": 2.31,
      "eval_accuracy": 0.9028703229113275,
      "eval_f1": 0.9028703229113276,
      "eval_loss": 0.4577423334121704,
      "eval_precision": 0.9028703229113275,
      "eval_recall": 0.9028703229113275,
      "eval_runtime": 68.0487,
      "eval_samples_per_second": 114.683,
      "eval_steps_per_second": 14.343,
      "step": 2500
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.405730129390019e-05,
      "loss": 0.0389,
      "step": 2510
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.359519408502773e-05,
      "loss": 0.08,
      "step": 2520
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.313308687615527e-05,
      "loss": 0.1181,
      "step": 2530
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.267097966728281e-05,
      "loss": 0.0461,
      "step": 2540
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.220887245841036e-05,
      "loss": 0.049,
      "step": 2550
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.179297597042514e-05,
      "loss": 0.0492,
      "step": 2560
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.133086876155268e-05,
      "loss": 0.0382,
      "step": 2570
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.086876155268022e-05,
      "loss": 0.064,
      "step": 2580
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.040665434380776e-05,
      "loss": 0.0385,
      "step": 2590
    },
    {
      "epoch": 2.4,
      "learning_rate": 7.99445471349353e-05,
      "loss": 0.0085,
      "step": 2600
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.9202972834443875,
      "eval_f1": 0.9202972834443875,
      "eval_loss": 0.3182913661003113,
      "eval_precision": 0.9202972834443875,
      "eval_recall": 0.9202972834443875,
      "eval_runtime": 68.3876,
      "eval_samples_per_second": 114.114,
      "eval_steps_per_second": 14.272,
      "step": 2600
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.948243992606285e-05,
      "loss": 0.0233,
      "step": 2610
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.902033271719039e-05,
      "loss": 0.0726,
      "step": 2620
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.855822550831793e-05,
      "loss": 0.0309,
      "step": 2630
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.809611829944548e-05,
      "loss": 0.0107,
      "step": 2640
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.763401109057301e-05,
      "loss": 0.0093,
      "step": 2650
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.717190388170056e-05,
      "loss": 0.038,
      "step": 2660
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.67097966728281e-05,
      "loss": 0.0343,
      "step": 2670
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.624768946395564e-05,
      "loss": 0.0888,
      "step": 2680
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.578558225508319e-05,
      "loss": 0.0912,
      "step": 2690
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.532347504621073e-05,
      "loss": 0.06,
      "step": 2700
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.9126089185033316,
      "eval_f1": 0.9126089185033316,
      "eval_loss": 0.3547973334789276,
      "eval_precision": 0.9126089185033316,
      "eval_recall": 0.9126089185033316,
      "eval_runtime": 68.5081,
      "eval_samples_per_second": 113.914,
      "eval_steps_per_second": 14.246,
      "step": 2700
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.486136783733827e-05,
      "loss": 0.0463,
      "step": 2710
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.439926062846581e-05,
      "loss": 0.0523,
      "step": 2720
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.393715341959335e-05,
      "loss": 0.0254,
      "step": 2730
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.347504621072089e-05,
      "loss": 0.0618,
      "step": 2740
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.301293900184843e-05,
      "loss": 0.042,
      "step": 2750
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.255083179297597e-05,
      "loss": 0.0127,
      "step": 2760
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.208872458410351e-05,
      "loss": 0.0312,
      "step": 2770
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.162661737523105e-05,
      "loss": 0.0226,
      "step": 2780
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.11645101663586e-05,
      "loss": 0.0441,
      "step": 2790
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.070240295748613e-05,
      "loss": 0.0139,
      "step": 2800
    },
    {
      "epoch": 2.59,
      "eval_accuracy": 0.9197847257816504,
      "eval_f1": 0.9197847257816503,
      "eval_loss": 0.3212650418281555,
      "eval_precision": 0.9197847257816504,
      "eval_recall": 0.9197847257816504,
      "eval_runtime": 68.2461,
      "eval_samples_per_second": 114.351,
      "eval_steps_per_second": 14.301,
      "step": 2800
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.024029574861369e-05,
      "loss": 0.0127,
      "step": 2810
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.977818853974121e-05,
      "loss": 0.0737,
      "step": 2820
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.931608133086877e-05,
      "loss": 0.0033,
      "step": 2830
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.885397412199631e-05,
      "loss": 0.0991,
      "step": 2840
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.839186691312385e-05,
      "loss": 0.0167,
      "step": 2850
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.792975970425139e-05,
      "loss": 0.0242,
      "step": 2860
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.746765249537893e-05,
      "loss": 0.0384,
      "step": 2870
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.700554528650647e-05,
      "loss": 0.006,
      "step": 2880
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.654343807763401e-05,
      "loss": 0.0607,
      "step": 2890
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.608133086876155e-05,
      "loss": 0.056,
      "step": 2900
    },
    {
      "epoch": 2.68,
      "eval_accuracy": 0.9131214761660686,
      "eval_f1": 0.9131214761660685,
      "eval_loss": 0.35575374960899353,
      "eval_precision": 0.9131214761660686,
      "eval_recall": 0.9131214761660686,
      "eval_runtime": 67.8806,
      "eval_samples_per_second": 114.967,
      "eval_steps_per_second": 14.378,
      "step": 2900
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.56192236598891e-05,
      "loss": 0.0563,
      "step": 2910
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.515711645101664e-05,
      "loss": 0.0499,
      "step": 2920
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.469500924214418e-05,
      "loss": 0.0304,
      "step": 2930
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.423290203327173e-05,
      "loss": 0.0178,
      "step": 2940
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.377079482439926e-05,
      "loss": 0.0304,
      "step": 2950
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.330868761552681e-05,
      "loss": 0.028,
      "step": 2960
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.284658040665434e-05,
      "loss": 0.0567,
      "step": 2970
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.23844731977819e-05,
      "loss": 0.0198,
      "step": 2980
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.192236598890942e-05,
      "loss": 0.015,
      "step": 2990
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.146025878003697e-05,
      "loss": 0.0433,
      "step": 3000
    },
    {
      "epoch": 2.77,
      "eval_accuracy": 0.9214505381855459,
      "eval_f1": 0.9214505381855459,
      "eval_loss": 0.3101392388343811,
      "eval_precision": 0.9214505381855459,
      "eval_recall": 0.9214505381855459,
      "eval_runtime": 68.1627,
      "eval_samples_per_second": 114.491,
      "eval_steps_per_second": 14.319,
      "step": 3000
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.099815157116451e-05,
      "loss": 0.0055,
      "step": 3010
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.0536044362292056e-05,
      "loss": 0.0518,
      "step": 3020
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.0073937153419604e-05,
      "loss": 0.0027,
      "step": 3030
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.961182994454714e-05,
      "loss": 0.0249,
      "step": 3040
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.9149722735674685e-05,
      "loss": 0.0414,
      "step": 3050
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.868761552680222e-05,
      "loss": 0.0233,
      "step": 3060
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.8225508317929767e-05,
      "loss": 0.0039,
      "step": 3070
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.77634011090573e-05,
      "loss": 0.0433,
      "step": 3080
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.730129390018485e-05,
      "loss": 0.0479,
      "step": 3090
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.683918669131238e-05,
      "loss": 0.0074,
      "step": 3100
    },
    {
      "epoch": 2.87,
      "eval_accuracy": 0.9176063557150179,
      "eval_f1": 0.9176063557150179,
      "eval_loss": 0.3139953315258026,
      "eval_precision": 0.9176063557150179,
      "eval_recall": 0.9176063557150179,
      "eval_runtime": 68.4275,
      "eval_samples_per_second": 114.048,
      "eval_steps_per_second": 14.263,
      "step": 3100
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.637707948243993e-05,
      "loss": 0.0745,
      "step": 3110
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.591497227356747e-05,
      "loss": 0.0395,
      "step": 3120
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.545286506469501e-05,
      "loss": 0.0151,
      "step": 3130
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.499075785582255e-05,
      "loss": 0.0227,
      "step": 3140
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.45286506469501e-05,
      "loss": 0.0399,
      "step": 3150
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.406654343807763e-05,
      "loss": 0.0165,
      "step": 3160
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.360443622920518e-05,
      "loss": 0.0383,
      "step": 3170
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.3142329020332715e-05,
      "loss": 0.0159,
      "step": 3180
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.268022181146026e-05,
      "loss": 0.0039,
      "step": 3190
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.2218114602587796e-05,
      "loss": 0.0216,
      "step": 3200
    },
    {
      "epoch": 2.96,
      "eval_accuracy": 0.918631471040492,
      "eval_f1": 0.918631471040492,
      "eval_loss": 0.3656544089317322,
      "eval_precision": 0.918631471040492,
      "eval_recall": 0.918631471040492,
      "eval_runtime": 68.2059,
      "eval_samples_per_second": 114.418,
      "eval_steps_per_second": 14.31,
      "step": 3200
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.1756007393715344e-05,
      "loss": 0.0133,
      "step": 3210
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.129390018484289e-05,
      "loss": 0.0278,
      "step": 3220
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.0831792975970425e-05,
      "loss": 0.0476,
      "step": 3230
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.036968576709797e-05,
      "loss": 0.0467,
      "step": 3240
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.9907578558225506e-05,
      "loss": 0.0748,
      "step": 3250
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.9445471349353054e-05,
      "loss": 0.0039,
      "step": 3260
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.8983364140480595e-05,
      "loss": 0.0016,
      "step": 3270
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.8521256931608135e-05,
      "loss": 0.0069,
      "step": 3280
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.8059149722735676e-05,
      "loss": 0.002,
      "step": 3290
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.759704251386322e-05,
      "loss": 0.0118,
      "step": 3300
    },
    {
      "epoch": 3.05,
      "eval_accuracy": 0.9195284469502819,
      "eval_f1": 0.9195284469502819,
      "eval_loss": 0.3722282350063324,
      "eval_precision": 0.9195284469502819,
      "eval_recall": 0.9195284469502819,
      "eval_runtime": 68.0438,
      "eval_samples_per_second": 114.691,
      "eval_steps_per_second": 14.344,
      "step": 3300
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.7134935304990764e-05,
      "loss": 0.0033,
      "step": 3310
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.6672828096118305e-05,
      "loss": 0.0047,
      "step": 3320
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.6210720887245846e-05,
      "loss": 0.0075,
      "step": 3330
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.5748613678373387e-05,
      "loss": 0.0261,
      "step": 3340
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.528650646950093e-05,
      "loss": 0.0015,
      "step": 3350
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.482439926062847e-05,
      "loss": 0.0031,
      "step": 3360
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.436229205175601e-05,
      "loss": 0.0425,
      "step": 3370
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.390018484288355e-05,
      "loss": 0.0013,
      "step": 3380
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.343807763401109e-05,
      "loss": 0.0011,
      "step": 3390
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.297597042513863e-05,
      "loss": 0.0014,
      "step": 3400
    },
    {
      "epoch": 3.14,
      "eval_accuracy": 0.9141465914915428,
      "eval_f1": 0.9141465914915428,
      "eval_loss": 0.40892842411994934,
      "eval_precision": 0.9141465914915428,
      "eval_recall": 0.9141465914915428,
      "eval_runtime": 68.314,
      "eval_samples_per_second": 114.237,
      "eval_steps_per_second": 14.287,
      "step": 3400
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.251386321626618e-05,
      "loss": 0.0015,
      "step": 3410
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.205175600739372e-05,
      "loss": 0.001,
      "step": 3420
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.158964879852126e-05,
      "loss": 0.0024,
      "step": 3430
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.11275415896488e-05,
      "loss": 0.0266,
      "step": 3440
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.066543438077634e-05,
      "loss": 0.0009,
      "step": 3450
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.020332717190388e-05,
      "loss": 0.001,
      "step": 3460
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.974121996303142e-05,
      "loss": 0.0009,
      "step": 3470
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9279112754158964e-05,
      "loss": 0.0031,
      "step": 3480
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.8817005545286504e-05,
      "loss": 0.0016,
      "step": 3490
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.835489833641405e-05,
      "loss": 0.001,
      "step": 3500
    },
    {
      "epoch": 3.23,
      "eval_accuracy": 0.9188877498718606,
      "eval_f1": 0.9188877498718606,
      "eval_loss": 0.4045088589191437,
      "eval_precision": 0.9188877498718606,
      "eval_recall": 0.9188877498718606,
      "eval_runtime": 69.1504,
      "eval_samples_per_second": 112.855,
      "eval_steps_per_second": 14.114,
      "step": 3500
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.789279112754159e-05,
      "loss": 0.0086,
      "step": 3510
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.743068391866913e-05,
      "loss": 0.0009,
      "step": 3520
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.6968576709796674e-05,
      "loss": 0.0093,
      "step": 3530
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.6506469500924215e-05,
      "loss": 0.0078,
      "step": 3540
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.6044362292051755e-05,
      "loss": 0.0184,
      "step": 3550
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.55822550831793e-05,
      "loss": 0.0011,
      "step": 3560
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.5120147874306844e-05,
      "loss": 0.0008,
      "step": 3570
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.4658040665434384e-05,
      "loss": 0.0017,
      "step": 3580
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.4195933456561925e-05,
      "loss": 0.0008,
      "step": 3590
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.3733826247689466e-05,
      "loss": 0.0009,
      "step": 3600
    },
    {
      "epoch": 3.33,
      "eval_accuracy": 0.9229882111737571,
      "eval_f1": 0.9229882111737571,
      "eval_loss": 0.3931851387023926,
      "eval_precision": 0.9229882111737571,
      "eval_recall": 0.9229882111737571,
      "eval_runtime": 68.0238,
      "eval_samples_per_second": 114.725,
      "eval_steps_per_second": 14.348,
      "step": 3600
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.3271719038817007e-05,
      "loss": 0.002,
      "step": 3610
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.280961182994455e-05,
      "loss": 0.0009,
      "step": 3620
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.234750462107209e-05,
      "loss": 0.0008,
      "step": 3630
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.188539741219963e-05,
      "loss": 0.0008,
      "step": 3640
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.142329020332717e-05,
      "loss": 0.0007,
      "step": 3650
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.096118299445471e-05,
      "loss": 0.0103,
      "step": 3660
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.0499075785582254e-05,
      "loss": 0.0008,
      "step": 3670
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.0036968576709802e-05,
      "loss": 0.0008,
      "step": 3680
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9574861367837343e-05,
      "loss": 0.0421,
      "step": 3690
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.9112754158964883e-05,
      "loss": 0.0009,
      "step": 3700
    },
    {
      "epoch": 3.42,
      "eval_accuracy": 0.9173500768836494,
      "eval_f1": 0.9173500768836496,
      "eval_loss": 0.4256540536880493,
      "eval_precision": 0.9173500768836494,
      "eval_recall": 0.9173500768836494,
      "eval_runtime": 68.8035,
      "eval_samples_per_second": 113.425,
      "eval_steps_per_second": 14.185,
      "step": 3700
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.8650646950092424e-05,
      "loss": 0.0189,
      "step": 3710
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8188539741219965e-05,
      "loss": 0.001,
      "step": 3720
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.7726432532347505e-05,
      "loss": 0.0012,
      "step": 3730
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.726432532347505e-05,
      "loss": 0.0011,
      "step": 3740
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.680221811460259e-05,
      "loss": 0.0007,
      "step": 3750
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.634011090573013e-05,
      "loss": 0.0007,
      "step": 3760
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.5878003696857672e-05,
      "loss": 0.0476,
      "step": 3770
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5415896487985213e-05,
      "loss": 0.0007,
      "step": 3780
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.4953789279112753e-05,
      "loss": 0.0007,
      "step": 3790
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4491682070240297e-05,
      "loss": 0.03,
      "step": 3800
    },
    {
      "epoch": 3.51,
      "eval_accuracy": 0.9222193746796514,
      "eval_f1": 0.9222193746796514,
      "eval_loss": 0.3981027603149414,
      "eval_precision": 0.9222193746796514,
      "eval_recall": 0.9222193746796514,
      "eval_runtime": 68.2323,
      "eval_samples_per_second": 114.374,
      "eval_steps_per_second": 14.304,
      "step": 3800
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.4029574861367838e-05,
      "loss": 0.0007,
      "step": 3810
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.3567467652495382e-05,
      "loss": 0.0032,
      "step": 3820
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.3105360443622923e-05,
      "loss": 0.04,
      "step": 3830
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.2643253234750464e-05,
      "loss": 0.0008,
      "step": 3840
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2181146025878004e-05,
      "loss": 0.0008,
      "step": 3850
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.1719038817005545e-05,
      "loss": 0.0007,
      "step": 3860
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.125693160813309e-05,
      "loss": 0.0007,
      "step": 3870
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.079482439926063e-05,
      "loss": 0.0337,
      "step": 3880
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.033271719038817e-05,
      "loss": 0.0007,
      "step": 3890
    },
    {
      "epoch": 3.6,
      "learning_rate": 1.987060998151571e-05,
      "loss": 0.0007,
      "step": 3900
    },
    {
      "epoch": 3.6,
      "eval_accuracy": 0.9188877498718606,
      "eval_f1": 0.9188877498718606,
      "eval_loss": 0.42112305760383606,
      "eval_precision": 0.9188877498718606,
      "eval_recall": 0.9188877498718606,
      "eval_runtime": 67.8886,
      "eval_samples_per_second": 114.953,
      "eval_steps_per_second": 14.376,
      "step": 3900
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9408502772643252e-05,
      "loss": 0.0719,
      "step": 3910
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.8946395563770796e-05,
      "loss": 0.0036,
      "step": 3920
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8484288354898337e-05,
      "loss": 0.0009,
      "step": 3930
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.8022181146025878e-05,
      "loss": 0.0009,
      "step": 3940
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.7560073937153422e-05,
      "loss": 0.0008,
      "step": 3950
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.7097966728280963e-05,
      "loss": 0.0012,
      "step": 3960
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6635859519408503e-05,
      "loss": 0.0009,
      "step": 3970
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.6173752310536044e-05,
      "loss": 0.001,
      "step": 3980
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5711645101663585e-05,
      "loss": 0.0009,
      "step": 3990
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5249537892791127e-05,
      "loss": 0.0494,
      "step": 4000
    },
    {
      "epoch": 3.7,
      "eval_accuracy": 0.9206817016914403,
      "eval_f1": 0.9206817016914403,
      "eval_loss": 0.40287914872169495,
      "eval_precision": 0.9206817016914403,
      "eval_recall": 0.9206817016914403,
      "eval_runtime": 68.3875,
      "eval_samples_per_second": 114.114,
      "eval_steps_per_second": 14.272,
      "step": 4000
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4787430683918671e-05,
      "loss": 0.0011,
      "step": 4010
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.4325323475046212e-05,
      "loss": 0.001,
      "step": 4020
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.3863216266173753e-05,
      "loss": 0.001,
      "step": 4030
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3401109057301295e-05,
      "loss": 0.0318,
      "step": 4040
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.2939001848428836e-05,
      "loss": 0.0009,
      "step": 4050
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2476894639556377e-05,
      "loss": 0.001,
      "step": 4060
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.2060998151571165e-05,
      "loss": 0.0082,
      "step": 4070
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1598890942698706e-05,
      "loss": 0.0011,
      "step": 4080
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1136783733826248e-05,
      "loss": 0.0067,
      "step": 4090
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0674676524953791e-05,
      "loss": 0.0009,
      "step": 4100
    },
    {
      "epoch": 3.79,
      "eval_accuracy": 0.9226037929267042,
      "eval_f1": 0.9226037929267042,
      "eval_loss": 0.3950587511062622,
      "eval_precision": 0.9226037929267042,
      "eval_recall": 0.9226037929267042,
      "eval_runtime": 68.4949,
      "eval_samples_per_second": 113.936,
      "eval_steps_per_second": 14.249,
      "step": 4100
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.0212569316081332e-05,
      "loss": 0.0249,
      "step": 4110
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.750462107208872e-06,
      "loss": 0.0008,
      "step": 4120
    },
    {
      "epoch": 3.82,
      "learning_rate": 9.288354898336415e-06,
      "loss": 0.0007,
      "step": 4130
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.826247689463957e-06,
      "loss": 0.0007,
      "step": 4140
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.364140480591498e-06,
      "loss": 0.0008,
      "step": 4150
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.902033271719039e-06,
      "loss": 0.0188,
      "step": 4160
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.43992606284658e-06,
      "loss": 0.001,
      "step": 4170
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.977818853974122e-06,
      "loss": 0.0007,
      "step": 4180
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.515711645101664e-06,
      "loss": 0.0944,
      "step": 4190
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.053604436229206e-06,
      "loss": 0.0319,
      "step": 4200
    },
    {
      "epoch": 3.88,
      "eval_accuracy": 0.9220912352639672,
      "eval_f1": 0.9220912352639672,
      "eval_loss": 0.3943887948989868,
      "eval_precision": 0.9220912352639672,
      "eval_recall": 0.9220912352639672,
      "eval_runtime": 68.3629,
      "eval_samples_per_second": 114.155,
      "eval_steps_per_second": 14.277,
      "step": 4200
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.591497227356747e-06,
      "loss": 0.0008,
      "step": 4210
    },
    {
      "epoch": 3.9,
      "learning_rate": 5.129390018484288e-06,
      "loss": 0.0246,
      "step": 4220
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.6672828096118305e-06,
      "loss": 0.001,
      "step": 4230
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.205175600739371e-06,
      "loss": 0.0087,
      "step": 4240
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.7430683918669137e-06,
      "loss": 0.0309,
      "step": 4250
    },
    {
      "epoch": 3.94,
      "learning_rate": 3.280961182994455e-06,
      "loss": 0.0007,
      "step": 4260
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.8188539741219964e-06,
      "loss": 0.0008,
      "step": 4270
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.356746765249538e-06,
      "loss": 0.0009,
      "step": 4280
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.8946395563770795e-06,
      "loss": 0.0096,
      "step": 4290
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.4325323475046211e-06,
      "loss": 0.0013,
      "step": 4300
    },
    {
      "epoch": 3.97,
      "eval_accuracy": 0.92247565351102,
      "eval_f1": 0.92247565351102,
      "eval_loss": 0.389446884393692,
      "eval_precision": 0.92247565351102,
      "eval_recall": 0.92247565351102,
      "eval_runtime": 68.2335,
      "eval_samples_per_second": 114.372,
      "eval_steps_per_second": 14.304,
      "step": 4300
    },
    {
      "epoch": 3.98,
      "learning_rate": 9.704251386321627e-07,
      "loss": 0.0009,
      "step": 4310
    },
    {
      "epoch": 3.99,
      "learning_rate": 5.083179297597043e-07,
      "loss": 0.0008,
      "step": 4320
    },
    {
      "epoch": 4.0,
      "step": 4328,
      "total_flos": 5.3636855531636e+18,
      "train_loss": 0.12048932520910244,
      "train_runtime": 4412.0416,
      "train_samples_per_second": 15.688,
      "train_steps_per_second": 0.981
    }
  ],
  "max_steps": 4328,
  "num_train_epochs": 4,
  "total_flos": 5.3636855531636e+18,
  "trial_name": null,
  "trial_params": null
}