{
  "best_metric": 0.21288499236106873,
  "best_model_checkpoint": "./vit-huge-binary-isic-sharpened-patch-14/checkpoint-2000",
  "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": 0.5732,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0001990757855822551,
      "loss": 0.4357,
      "step": 20
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019861367837338263,
      "loss": 0.3054,
      "step": 30
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00019815157116451017,
      "loss": 0.292,
      "step": 40
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0001976894639556377,
      "loss": 0.3734,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019722735674676528,
      "loss": 0.3744,
      "step": 60
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0001967652495378928,
      "loss": 0.3538,
      "step": 70
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019630314232902034,
      "loss": 0.3201,
      "step": 80
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019584103512014788,
      "loss": 0.3039,
      "step": 90
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019537892791127544,
      "loss": 0.3116,
      "step": 100
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.8553305996924654,
      "eval_f1": 0.8553305996924655,
      "eval_loss": 0.3191303610801697,
      "eval_precision": 0.8553305996924654,
      "eval_recall": 0.8553305996924654,
      "eval_runtime": 93.9458,
      "eval_samples_per_second": 83.069,
      "eval_steps_per_second": 10.389,
      "step": 100
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019491682070240296,
      "loss": 0.354,
      "step": 110
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0001944547134935305,
      "loss": 0.2593,
      "step": 120
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019399260628465804,
      "loss": 0.374,
      "step": 130
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0001935304990757856,
      "loss": 0.2888,
      "step": 140
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019306839186691312,
      "loss": 0.2536,
      "step": 150
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019260628465804066,
      "loss": 0.2267,
      "step": 160
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019214417744916823,
      "loss": 0.2944,
      "step": 170
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019168207024029577,
      "loss": 0.2981,
      "step": 180
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0001912199630314233,
      "loss": 0.302,
      "step": 190
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019075785582255082,
      "loss": 0.3328,
      "step": 200
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.8152229625832906,
      "eval_f1": 0.8152229625832906,
      "eval_loss": 0.3595452904701233,
      "eval_precision": 0.8152229625832906,
      "eval_recall": 0.8152229625832906,
      "eval_runtime": 94.0698,
      "eval_samples_per_second": 82.96,
      "eval_steps_per_second": 10.375,
      "step": 200
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0001902957486136784,
      "loss": 0.2301,
      "step": 210
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018983364140480593,
      "loss": 0.4238,
      "step": 220
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018937153419593347,
      "loss": 0.237,
      "step": 230
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.000188909426987061,
      "loss": 0.2305,
      "step": 240
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018844731977818855,
      "loss": 0.2962,
      "step": 250
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0001879852125693161,
      "loss": 0.1984,
      "step": 260
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00018752310536044364,
      "loss": 0.2392,
      "step": 270
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00018706099815157118,
      "loss": 0.2821,
      "step": 280
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018659889094269872,
      "loss": 0.2291,
      "step": 290
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018613678373382626,
      "loss": 0.2597,
      "step": 300
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.8732701178882625,
      "eval_f1": 0.8732701178882625,
      "eval_loss": 0.29005151987075806,
      "eval_precision": 0.8732701178882625,
      "eval_recall": 0.8732701178882625,
      "eval_runtime": 93.7287,
      "eval_samples_per_second": 83.262,
      "eval_steps_per_second": 10.413,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0001856746765249538,
      "loss": 0.2859,
      "step": 310
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018521256931608134,
      "loss": 0.2641,
      "step": 320
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018475046210720888,
      "loss": 0.2536,
      "step": 330
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018428835489833642,
      "loss": 0.1996,
      "step": 340
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018382624768946396,
      "loss": 0.2005,
      "step": 350
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0001833641404805915,
      "loss": 0.29,
      "step": 360
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018290203327171904,
      "loss": 0.209,
      "step": 370
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018243992606284658,
      "loss": 0.2627,
      "step": 380
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018197781885397413,
      "loss": 0.2659,
      "step": 390
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0001815157116451017,
      "loss": 0.2037,
      "step": 400
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.8912096360840595,
      "eval_f1": 0.8912096360840595,
      "eval_loss": 0.24775776267051697,
      "eval_precision": 0.8912096360840595,
      "eval_recall": 0.8912096360840595,
      "eval_runtime": 94.6511,
      "eval_samples_per_second": 82.45,
      "eval_steps_per_second": 10.312,
      "step": 400
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0001810536044362292,
      "loss": 0.1986,
      "step": 410
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00018059149722735675,
      "loss": 0.1758,
      "step": 420
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0001801293900184843,
      "loss": 0.2557,
      "step": 430
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00017966728280961186,
      "loss": 0.2713,
      "step": 440
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017920517560073937,
      "loss": 0.1341,
      "step": 450
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0001787430683918669,
      "loss": 0.2298,
      "step": 460
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00017828096118299448,
      "loss": 0.1818,
      "step": 470
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017781885397412202,
      "loss": 0.1684,
      "step": 480
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017735674676524953,
      "loss": 0.2039,
      "step": 490
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017689463955637707,
      "loss": 0.2415,
      "step": 500
    },
    {
      "epoch": 0.46,
      "eval_accuracy": 0.8965914915427986,
      "eval_f1": 0.8965914915427986,
      "eval_loss": 0.24882297217845917,
      "eval_precision": 0.8965914915427986,
      "eval_recall": 0.8965914915427986,
      "eval_runtime": 94.8507,
      "eval_samples_per_second": 82.277,
      "eval_steps_per_second": 10.29,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00017643253234750464,
      "loss": 0.1601,
      "step": 510
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00017597042513863218,
      "loss": 0.2957,
      "step": 520
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0001755083179297597,
      "loss": 0.165,
      "step": 530
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00017504621072088724,
      "loss": 0.1761,
      "step": 540
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0001745841035120148,
      "loss": 0.1448,
      "step": 550
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00017412199630314234,
      "loss": 0.2043,
      "step": 560
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017365988909426989,
      "loss": 0.1434,
      "step": 570
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00017319778188539743,
      "loss": 0.1463,
      "step": 580
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00017273567467652497,
      "loss": 0.162,
      "step": 590
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0001722735674676525,
      "loss": 0.1483,
      "step": 600
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.8844182470527935,
      "eval_f1": 0.8844182470527935,
      "eval_loss": 0.2943035662174225,
      "eval_precision": 0.8844182470527935,
      "eval_recall": 0.8844182470527935,
      "eval_runtime": 94.8571,
      "eval_samples_per_second": 82.271,
      "eval_steps_per_second": 10.289,
      "step": 600
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017181146025878005,
      "loss": 0.2827,
      "step": 610
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0001713493530499076,
      "loss": 0.1908,
      "step": 620
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017088724584103513,
      "loss": 0.134,
      "step": 630
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00017042513863216267,
      "loss": 0.1652,
      "step": 640
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0001699630314232902,
      "loss": 0.1357,
      "step": 650
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016950092421441775,
      "loss": 0.2055,
      "step": 660
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0001690388170055453,
      "loss": 0.1381,
      "step": 670
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00016857670979667283,
      "loss": 0.1973,
      "step": 680
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016811460258780037,
      "loss": 0.17,
      "step": 690
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00016765249537892791,
      "loss": 0.1126,
      "step": 700
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.8751922091235264,
      "eval_f1": 0.8751922091235264,
      "eval_loss": 0.3269084393978119,
      "eval_precision": 0.8751922091235264,
      "eval_recall": 0.8751922091235264,
      "eval_runtime": 94.1477,
      "eval_samples_per_second": 82.891,
      "eval_steps_per_second": 10.367,
      "step": 700
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016719038817005546,
      "loss": 0.1857,
      "step": 710
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.000166728280961183,
      "loss": 0.1593,
      "step": 720
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016626617375231054,
      "loss": 0.118,
      "step": 730
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00016580406654343808,
      "loss": 0.1207,
      "step": 740
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016534195933456562,
      "loss": 0.1067,
      "step": 750
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016487985212569316,
      "loss": 0.1537,
      "step": 760
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016441774491682073,
      "loss": 0.1113,
      "step": 770
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016395563770794827,
      "loss": 0.1337,
      "step": 780
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00016349353049907578,
      "loss": 0.1811,
      "step": 790
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016303142329020332,
      "loss": 0.207,
      "step": 800
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 0.8868528959507944,
      "eval_f1": 0.8868528959507944,
      "eval_loss": 0.29836413264274597,
      "eval_precision": 0.8868528959507944,
      "eval_recall": 0.8868528959507944,
      "eval_runtime": 95.3225,
      "eval_samples_per_second": 81.869,
      "eval_steps_per_second": 10.239,
      "step": 800
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0001625693160813309,
      "loss": 0.183,
      "step": 810
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00016210720887245843,
      "loss": 0.168,
      "step": 820
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016164510166358594,
      "loss": 0.147,
      "step": 830
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016118299445471348,
      "loss": 0.1924,
      "step": 840
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016072088724584105,
      "loss": 0.1122,
      "step": 850
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0001602587800369686,
      "loss": 0.1421,
      "step": 860
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0001597966728280961,
      "loss": 0.1391,
      "step": 870
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015933456561922367,
      "loss": 0.2099,
      "step": 880
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00015887245841035122,
      "loss": 0.1028,
      "step": 890
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015841035120147876,
      "loss": 0.234,
      "step": 900
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.8856996412096361,
      "eval_f1": 0.8856996412096362,
      "eval_loss": 0.23936809599399567,
      "eval_precision": 0.8856996412096361,
      "eval_recall": 0.8856996412096361,
      "eval_runtime": 94.789,
      "eval_samples_per_second": 82.33,
      "eval_steps_per_second": 10.297,
      "step": 900
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015794824399260627,
      "loss": 0.1027,
      "step": 910
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00015748613678373384,
      "loss": 0.1248,
      "step": 920
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015702402957486138,
      "loss": 0.1334,
      "step": 930
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00015656192236598892,
      "loss": 0.1109,
      "step": 940
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00015609981515711646,
      "loss": 0.0983,
      "step": 950
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.000155637707948244,
      "loss": 0.0804,
      "step": 960
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00015517560073937154,
      "loss": 0.0837,
      "step": 970
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015471349353049908,
      "loss": 0.1788,
      "step": 980
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00015425138632162662,
      "loss": 0.1913,
      "step": 990
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015378927911275416,
      "loss": 0.1625,
      "step": 1000
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 0.8930035879036392,
      "eval_f1": 0.8930035879036392,
      "eval_loss": 0.2706281542778015,
      "eval_precision": 0.8930035879036392,
      "eval_recall": 0.8930035879036392,
      "eval_runtime": 94.6616,
      "eval_samples_per_second": 82.441,
      "eval_steps_per_second": 10.31,
      "step": 1000
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0001533271719038817,
      "loss": 0.1515,
      "step": 1010
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015286506469500925,
      "loss": 0.0941,
      "step": 1020
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015240295748613679,
      "loss": 0.1245,
      "step": 1030
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00015194085027726433,
      "loss": 0.0967,
      "step": 1040
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00015147874306839187,
      "loss": 0.1997,
      "step": 1050
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0001510166358595194,
      "loss": 0.1668,
      "step": 1060
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015055452865064698,
      "loss": 0.1185,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0001500924214417745,
      "loss": 0.1416,
      "step": 1080
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014963031423290203,
      "loss": 0.0646,
      "step": 1090
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014916820702402957,
      "loss": 0.1131,
      "step": 1100
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.9208098411071246,
      "eval_f1": 0.9208098411071246,
      "eval_loss": 0.2155652791261673,
      "eval_precision": 0.9208098411071246,
      "eval_recall": 0.9208098411071246,
      "eval_runtime": 96.0421,
      "eval_samples_per_second": 81.256,
      "eval_steps_per_second": 10.162,
      "step": 1100
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00014870609981515714,
      "loss": 0.1247,
      "step": 1110
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014824399260628468,
      "loss": 0.1129,
      "step": 1120
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0001477818853974122,
      "loss": 0.0801,
      "step": 1130
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00014731977818853976,
      "loss": 0.0829,
      "step": 1140
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0001468576709796673,
      "loss": 0.0948,
      "step": 1150
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00014639556377079484,
      "loss": 0.0756,
      "step": 1160
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014593345656192236,
      "loss": 0.1084,
      "step": 1170
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014547134935304992,
      "loss": 0.0584,
      "step": 1180
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00014500924214417746,
      "loss": 0.0641,
      "step": 1190
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.000144547134935305,
      "loss": 0.048,
      "step": 1200
    },
    {
      "epoch": 1.11,
      "eval_accuracy": 0.902614044079959,
      "eval_f1": 0.902614044079959,
      "eval_loss": 0.2676137685775757,
      "eval_precision": 0.902614044079959,
      "eval_recall": 0.902614044079959,
      "eval_runtime": 95.4306,
      "eval_samples_per_second": 81.777,
      "eval_steps_per_second": 10.227,
      "step": 1200
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014408502772643252,
      "loss": 0.1314,
      "step": 1210
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0001436229205175601,
      "loss": 0.0481,
      "step": 1220
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014316081330868763,
      "loss": 0.0473,
      "step": 1230
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00014269870609981517,
      "loss": 0.0504,
      "step": 1240
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0001422365988909427,
      "loss": 0.0526,
      "step": 1250
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014177449168207025,
      "loss": 0.1015,
      "step": 1260
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0001413123844731978,
      "loss": 0.0173,
      "step": 1270
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014085027726432533,
      "loss": 0.0714,
      "step": 1280
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014038817005545287,
      "loss": 0.0518,
      "step": 1290
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0001399260628465804,
      "loss": 0.1235,
      "step": 1300
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.9024859046642747,
      "eval_f1": 0.9024859046642747,
      "eval_loss": 0.3325667083263397,
      "eval_precision": 0.9024859046642747,
      "eval_recall": 0.9024859046642747,
      "eval_runtime": 94.6894,
      "eval_samples_per_second": 82.417,
      "eval_steps_per_second": 10.307,
      "step": 1300
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013946395563770795,
      "loss": 0.1105,
      "step": 1310
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001390018484288355,
      "loss": 0.0317,
      "step": 1320
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013853974121996306,
      "loss": 0.0119,
      "step": 1330
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00013807763401109058,
      "loss": 0.0451,
      "step": 1340
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013761552680221812,
      "loss": 0.0338,
      "step": 1350
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013715341959334566,
      "loss": 0.0925,
      "step": 1360
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00013669131238447322,
      "loss": 0.05,
      "step": 1370
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013622920517560074,
      "loss": 0.0053,
      "step": 1380
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00013576709796672828,
      "loss": 0.0155,
      "step": 1390
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013530499075785582,
      "loss": 0.0093,
      "step": 1400
    },
    {
      "epoch": 1.29,
      "eval_accuracy": 0.9106868272680677,
      "eval_f1": 0.9106868272680677,
      "eval_loss": 0.3238871097564697,
      "eval_precision": 0.9106868272680677,
      "eval_recall": 0.9106868272680677,
      "eval_runtime": 94.2987,
      "eval_samples_per_second": 82.758,
      "eval_steps_per_second": 10.35,
      "step": 1400
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0001348428835489834,
      "loss": 0.0329,
      "step": 1410
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0001343807763401109,
      "loss": 0.1286,
      "step": 1420
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00013391866913123844,
      "loss": 0.0362,
      "step": 1430
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.000133456561922366,
      "loss": 0.056,
      "step": 1440
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013299445471349355,
      "loss": 0.0902,
      "step": 1450
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00013253234750462106,
      "loss": 0.0593,
      "step": 1460
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0001320702402957486,
      "loss": 0.0531,
      "step": 1470
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013160813308687617,
      "loss": 0.0572,
      "step": 1480
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013114602587800371,
      "loss": 0.1474,
      "step": 1490
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013068391866913125,
      "loss": 0.0851,
      "step": 1500
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.8971040492055357,
      "eval_f1": 0.8971040492055357,
      "eval_loss": 0.30670446157455444,
      "eval_precision": 0.8971040492055357,
      "eval_recall": 0.8971040492055357,
      "eval_runtime": 94.9658,
      "eval_samples_per_second": 82.177,
      "eval_steps_per_second": 10.277,
      "step": 1500
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00013022181146025877,
      "loss": 0.0714,
      "step": 1510
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00012975970425138634,
      "loss": 0.0505,
      "step": 1520
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012929759704251388,
      "loss": 0.0827,
      "step": 1530
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012883548983364142,
      "loss": 0.0374,
      "step": 1540
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012837338262476896,
      "loss": 0.0715,
      "step": 1550
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0001279112754158965,
      "loss": 0.0216,
      "step": 1560
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00012744916820702404,
      "loss": 0.0223,
      "step": 1570
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012698706099815158,
      "loss": 0.0945,
      "step": 1580
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012652495378927912,
      "loss": 0.0817,
      "step": 1590
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00012606284658040666,
      "loss": 0.0962,
      "step": 1600
    },
    {
      "epoch": 1.48,
      "eval_accuracy": 0.8997949769349052,
      "eval_f1": 0.8997949769349052,
      "eval_loss": 0.31642210483551025,
      "eval_precision": 0.8997949769349052,
      "eval_recall": 0.8997949769349052,
      "eval_runtime": 94.5252,
      "eval_samples_per_second": 82.56,
      "eval_steps_per_second": 10.325,
      "step": 1600
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0001256007393715342,
      "loss": 0.0105,
      "step": 1610
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00012513863216266174,
      "loss": 0.0262,
      "step": 1620
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012467652495378928,
      "loss": 0.1526,
      "step": 1630
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00012421441774491682,
      "loss": 0.0759,
      "step": 1640
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00012375231053604437,
      "loss": 0.016,
      "step": 1650
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0001232902033271719,
      "loss": 0.0586,
      "step": 1660
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00012282809611829945,
      "loss": 0.1242,
      "step": 1670
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.000122365988909427,
      "loss": 0.0294,
      "step": 1680
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00012190388170055453,
      "loss": 0.0231,
      "step": 1690
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00012144177449168208,
      "loss": 0.0143,
      "step": 1700
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 0.9265761148129165,
      "eval_f1": 0.9265761148129165,
      "eval_loss": 0.22694824635982513,
      "eval_precision": 0.9265761148129165,
      "eval_recall": 0.9265761148129165,
      "eval_runtime": 94.6116,
      "eval_samples_per_second": 82.485,
      "eval_steps_per_second": 10.316,
      "step": 1700
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012097966728280962,
      "loss": 0.0102,
      "step": 1710
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012051756007393715,
      "loss": 0.0731,
      "step": 1720
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0001200554528650647,
      "loss": 0.0087,
      "step": 1730
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00011959334565619225,
      "loss": 0.053,
      "step": 1740
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011913123844731979,
      "loss": 0.0179,
      "step": 1750
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011866913123844731,
      "loss": 0.1918,
      "step": 1760
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011820702402957487,
      "loss": 0.0386,
      "step": 1770
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011774491682070241,
      "loss": 0.0208,
      "step": 1780
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011728280961182996,
      "loss": 0.0044,
      "step": 1790
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011682070240295748,
      "loss": 0.071,
      "step": 1800
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 0.9118400820092261,
      "eval_f1": 0.9118400820092261,
      "eval_loss": 0.3436349630355835,
      "eval_precision": 0.9118400820092261,
      "eval_recall": 0.9118400820092261,
      "eval_runtime": 94.8946,
      "eval_samples_per_second": 82.239,
      "eval_steps_per_second": 10.285,
      "step": 1800
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00011635859519408503,
      "loss": 0.0363,
      "step": 1810
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011589648798521257,
      "loss": 0.0218,
      "step": 1820
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011543438077634013,
      "loss": 0.0769,
      "step": 1830
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011497227356746765,
      "loss": 0.0351,
      "step": 1840
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0001145101663585952,
      "loss": 0.077,
      "step": 1850
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011404805914972275,
      "loss": 0.0142,
      "step": 1860
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011358595194085029,
      "loss": 0.0371,
      "step": 1870
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011312384473197783,
      "loss": 0.0498,
      "step": 1880
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00011266173752310536,
      "loss": 0.0085,
      "step": 1890
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00011219963031423291,
      "loss": 0.0594,
      "step": 1900
    },
    {
      "epoch": 1.76,
      "eval_accuracy": 0.8996668375192209,
      "eval_f1": 0.8996668375192209,
      "eval_loss": 0.3813391327857971,
      "eval_precision": 0.8996668375192209,
      "eval_recall": 0.8996668375192209,
      "eval_runtime": 94.4198,
      "eval_samples_per_second": 82.652,
      "eval_steps_per_second": 10.337,
      "step": 1900
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011173752310536045,
      "loss": 0.1587,
      "step": 1910
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.000111275415896488,
      "loss": 0.0589,
      "step": 1920
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00011081330868761552,
      "loss": 0.0239,
      "step": 1930
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011035120147874307,
      "loss": 0.0649,
      "step": 1940
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00010988909426987061,
      "loss": 0.0337,
      "step": 1950
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010942698706099817,
      "loss": 0.0777,
      "step": 1960
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0001089648798521257,
      "loss": 0.0788,
      "step": 1970
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00010850277264325324,
      "loss": 0.0358,
      "step": 1980
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010804066543438078,
      "loss": 0.059,
      "step": 1990
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00010757855822550833,
      "loss": 0.0406,
      "step": 2000
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.933623782675551,
      "eval_f1": 0.933623782675551,
      "eval_loss": 0.21288499236106873,
      "eval_precision": 0.933623782675551,
      "eval_recall": 0.933623782675551,
      "eval_runtime": 95.9077,
      "eval_samples_per_second": 81.37,
      "eval_steps_per_second": 10.176,
      "step": 2000
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010711645101663586,
      "loss": 0.0426,
      "step": 2010
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0001066543438077634,
      "loss": 0.0358,
      "step": 2020
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010619223659889095,
      "loss": 0.0231,
      "step": 2030
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0001057301293900185,
      "loss": 0.0349,
      "step": 2040
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00010526802218114602,
      "loss": 0.0079,
      "step": 2050
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010480591497227356,
      "loss": 0.0262,
      "step": 2060
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010434380776340112,
      "loss": 0.0441,
      "step": 2070
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010388170055452866,
      "loss": 0.1052,
      "step": 2080
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010341959334565621,
      "loss": 0.0417,
      "step": 2090
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010295748613678373,
      "loss": 0.048,
      "step": 2100
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 0.9160686827268067,
      "eval_f1": 0.9160686827268067,
      "eval_loss": 0.31165143847465515,
      "eval_precision": 0.9160686827268067,
      "eval_recall": 0.9160686827268067,
      "eval_runtime": 94.3406,
      "eval_samples_per_second": 82.722,
      "eval_steps_per_second": 10.345,
      "step": 2100
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010249537892791128,
      "loss": 0.0299,
      "step": 2110
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010203327171903882,
      "loss": 0.0579,
      "step": 2120
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010157116451016637,
      "loss": 0.015,
      "step": 2130
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0001011090573012939,
      "loss": 0.0403,
      "step": 2140
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010064695009242144,
      "loss": 0.0375,
      "step": 2150
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.000100184842883549,
      "loss": 0.0328,
      "step": 2160
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.972273567467652e-05,
      "loss": 0.0407,
      "step": 2170
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.926062846580408e-05,
      "loss": 0.0067,
      "step": 2180
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.87985212569316e-05,
      "loss": 0.0374,
      "step": 2190
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.833641404805916e-05,
      "loss": 0.0071,
      "step": 2200
    },
    {
      "epoch": 2.03,
      "eval_accuracy": 0.9246540235776525,
      "eval_f1": 0.9246540235776525,
      "eval_loss": 0.2732369899749756,
      "eval_precision": 0.9246540235776525,
      "eval_recall": 0.9246540235776525,
      "eval_runtime": 95.2591,
      "eval_samples_per_second": 81.924,
      "eval_steps_per_second": 10.246,
      "step": 2200
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.787430683918669e-05,
      "loss": 0.0076,
      "step": 2210
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.741219963031424e-05,
      "loss": 0.0266,
      "step": 2220
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.695009242144177e-05,
      "loss": 0.0467,
      "step": 2230
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.648798521256932e-05,
      "loss": 0.009,
      "step": 2240
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.602587800369686e-05,
      "loss": 0.0048,
      "step": 2250
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.55637707948244e-05,
      "loss": 0.0019,
      "step": 2260
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.510166358595194e-05,
      "loss": 0.0131,
      "step": 2270
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.463955637707949e-05,
      "loss": 0.005,
      "step": 2280
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.417744916820703e-05,
      "loss": 0.002,
      "step": 2290
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.371534195933457e-05,
      "loss": 0.0021,
      "step": 2300
    },
    {
      "epoch": 2.13,
      "eval_accuracy": 0.9274730907227063,
      "eval_f1": 0.9274730907227063,
      "eval_loss": 0.3055046498775482,
      "eval_precision": 0.9274730907227063,
      "eval_recall": 0.9274730907227063,
      "eval_runtime": 94.2262,
      "eval_samples_per_second": 82.822,
      "eval_steps_per_second": 10.358,
      "step": 2300
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.325323475046212e-05,
      "loss": 0.0015,
      "step": 2310
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.279112754158965e-05,
      "loss": 0.0252,
      "step": 2320
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.23290203327172e-05,
      "loss": 0.0019,
      "step": 2330
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.186691312384473e-05,
      "loss": 0.0507,
      "step": 2340
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.140480591497228e-05,
      "loss": 0.0024,
      "step": 2350
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.094269870609981e-05,
      "loss": 0.008,
      "step": 2360
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.048059149722737e-05,
      "loss": 0.008,
      "step": 2370
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.001848428835489e-05,
      "loss": 0.0038,
      "step": 2380
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.955637707948245e-05,
      "loss": 0.0024,
      "step": 2390
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.909426987060999e-05,
      "loss": 0.0722,
      "step": 2400
    },
    {
      "epoch": 2.22,
      "eval_accuracy": 0.9250384418247053,
      "eval_f1": 0.9250384418247052,
      "eval_loss": 0.3441855013370514,
      "eval_precision": 0.9250384418247053,
      "eval_recall": 0.9250384418247053,
      "eval_runtime": 94.2676,
      "eval_samples_per_second": 82.786,
      "eval_steps_per_second": 10.354,
      "step": 2400
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.863216266173753e-05,
      "loss": 0.0172,
      "step": 2410
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.817005545286507e-05,
      "loss": 0.0038,
      "step": 2420
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.770794824399261e-05,
      "loss": 0.0211,
      "step": 2430
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.724584103512015e-05,
      "loss": 0.0568,
      "step": 2440
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.678373382624769e-05,
      "loss": 0.007,
      "step": 2450
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.632162661737525e-05,
      "loss": 0.0037,
      "step": 2460
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.585951940850277e-05,
      "loss": 0.0247,
      "step": 2470
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.539741219963033e-05,
      "loss": 0.0011,
      "step": 2480
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.493530499075785e-05,
      "loss": 0.0025,
      "step": 2490
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.447319778188541e-05,
      "loss": 0.0146,
      "step": 2500
    },
    {
      "epoch": 2.31,
      "eval_accuracy": 0.9314454126089186,
      "eval_f1": 0.9314454126089186,
      "eval_loss": 0.29841846227645874,
      "eval_precision": 0.9314454126089186,
      "eval_recall": 0.9314454126089186,
      "eval_runtime": 93.978,
      "eval_samples_per_second": 83.041,
      "eval_steps_per_second": 10.385,
      "step": 2500
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.401109057301294e-05,
      "loss": 0.0621,
      "step": 2510
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.354898336414049e-05,
      "loss": 0.0361,
      "step": 2520
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.308687615526802e-05,
      "loss": 0.0517,
      "step": 2530
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.262476894639557e-05,
      "loss": 0.0026,
      "step": 2540
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.216266173752311e-05,
      "loss": 0.0047,
      "step": 2550
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.170055452865065e-05,
      "loss": 0.0038,
      "step": 2560
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.12384473197782e-05,
      "loss": 0.0109,
      "step": 2570
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.077634011090573e-05,
      "loss": 0.051,
      "step": 2580
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.031423290203328e-05,
      "loss": 0.0026,
      "step": 2590
    },
    {
      "epoch": 2.4,
      "learning_rate": 7.985212569316082e-05,
      "loss": 0.0031,
      "step": 2600
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.929779600205023,
      "eval_f1": 0.929779600205023,
      "eval_loss": 0.3211658000946045,
      "eval_precision": 0.929779600205023,
      "eval_recall": 0.929779600205023,
      "eval_runtime": 94.0484,
      "eval_samples_per_second": 82.979,
      "eval_steps_per_second": 10.378,
      "step": 2600
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.939001848428836e-05,
      "loss": 0.0153,
      "step": 2610
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.89279112754159e-05,
      "loss": 0.0019,
      "step": 2620
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.846580406654345e-05,
      "loss": 0.0014,
      "step": 2630
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.800369685767098e-05,
      "loss": 0.0012,
      "step": 2640
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.754158964879853e-05,
      "loss": 0.0011,
      "step": 2650
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.707948243992606e-05,
      "loss": 0.001,
      "step": 2660
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.661737523105361e-05,
      "loss": 0.0024,
      "step": 2670
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.615526802218114e-05,
      "loss": 0.001,
      "step": 2680
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.56931608133087e-05,
      "loss": 0.0611,
      "step": 2690
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.523105360443624e-05,
      "loss": 0.0167,
      "step": 2700
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.9319579702716556,
      "eval_f1": 0.9319579702716556,
      "eval_loss": 0.3110896348953247,
      "eval_precision": 0.9319579702716556,
      "eval_recall": 0.9319579702716556,
      "eval_runtime": 94.4758,
      "eval_samples_per_second": 82.603,
      "eval_steps_per_second": 10.331,
      "step": 2700
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.476894639556378e-05,
      "loss": 0.0286,
      "step": 2710
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.430683918669132e-05,
      "loss": 0.0365,
      "step": 2720
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.384473197781886e-05,
      "loss": 0.0041,
      "step": 2730
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.33826247689464e-05,
      "loss": 0.0449,
      "step": 2740
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.292051756007394e-05,
      "loss": 0.0126,
      "step": 2750
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.245841035120148e-05,
      "loss": 0.0011,
      "step": 2760
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.199630314232902e-05,
      "loss": 0.0283,
      "step": 2770
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.153419593345656e-05,
      "loss": 0.0013,
      "step": 2780
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.10720887245841e-05,
      "loss": 0.0012,
      "step": 2790
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.060998151571166e-05,
      "loss": 0.0017,
      "step": 2800
    },
    {
      "epoch": 2.59,
      "eval_accuracy": 0.9306765761148129,
      "eval_f1": 0.9306765761148129,
      "eval_loss": 0.2883150577545166,
      "eval_precision": 0.9306765761148129,
      "eval_recall": 0.9306765761148129,
      "eval_runtime": 94.2594,
      "eval_samples_per_second": 82.793,
      "eval_steps_per_second": 10.354,
      "step": 2800
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.014787430683918e-05,
      "loss": 0.0019,
      "step": 2810
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.968576709796674e-05,
      "loss": 0.0032,
      "step": 2820
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.922365988909427e-05,
      "loss": 0.0009,
      "step": 2830
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.876155268022182e-05,
      "loss": 0.0015,
      "step": 2840
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.829944547134936e-05,
      "loss": 0.0008,
      "step": 2850
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.78373382624769e-05,
      "loss": 0.0008,
      "step": 2860
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.737523105360444e-05,
      "loss": 0.0021,
      "step": 2870
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.691312384473198e-05,
      "loss": 0.0009,
      "step": 2880
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.645101663585952e-05,
      "loss": 0.0016,
      "step": 2890
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.598890942698706e-05,
      "loss": 0.0007,
      "step": 2900
    },
    {
      "epoch": 2.68,
      "eval_accuracy": 0.9320861096873398,
      "eval_f1": 0.9320861096873398,
      "eval_loss": 0.31885990500450134,
      "eval_precision": 0.9320861096873398,
      "eval_recall": 0.9320861096873398,
      "eval_runtime": 93.7458,
      "eval_samples_per_second": 83.246,
      "eval_steps_per_second": 10.411,
      "step": 2900
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.55268022181146e-05,
      "loss": 0.0008,
      "step": 2910
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.506469500924215e-05,
      "loss": 0.0016,
      "step": 2920
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.460258780036969e-05,
      "loss": 0.0006,
      "step": 2930
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.414048059149723e-05,
      "loss": 0.0037,
      "step": 2940
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.372458410351202e-05,
      "loss": 0.0266,
      "step": 2950
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.326247689463956e-05,
      "loss": 0.0007,
      "step": 2960
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.28003696857671e-05,
      "loss": 0.0007,
      "step": 2970
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.233826247689464e-05,
      "loss": 0.001,
      "step": 2980
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.187615526802218e-05,
      "loss": 0.0143,
      "step": 2990
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.141404805914972e-05,
      "loss": 0.0012,
      "step": 3000
    },
    {
      "epoch": 2.77,
      "eval_accuracy": 0.9383649410558688,
      "eval_f1": 0.9383649410558688,
      "eval_loss": 0.29957300424575806,
      "eval_precision": 0.9383649410558688,
      "eval_recall": 0.9383649410558688,
      "eval_runtime": 94.4006,
      "eval_samples_per_second": 82.669,
      "eval_steps_per_second": 10.339,
      "step": 3000
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.0951940850277264e-05,
      "loss": 0.0035,
      "step": 3010
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.048983364140481e-05,
      "loss": 0.0008,
      "step": 3020
    },
    {
      "epoch": 2.8,
      "learning_rate": 6.002772643253235e-05,
      "loss": 0.0006,
      "step": 3030
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.956561922365989e-05,
      "loss": 0.01,
      "step": 3040
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.9103512014787434e-05,
      "loss": 0.0011,
      "step": 3050
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.864140480591498e-05,
      "loss": 0.0155,
      "step": 3060
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.8179297597042515e-05,
      "loss": 0.0009,
      "step": 3070
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.771719038817006e-05,
      "loss": 0.0005,
      "step": 3080
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.72550831792976e-05,
      "loss": 0.0012,
      "step": 3090
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.6792975970425144e-05,
      "loss": 0.0006,
      "step": 3100
    },
    {
      "epoch": 2.87,
      "eval_accuracy": 0.9281137878011276,
      "eval_f1": 0.9281137878011276,
      "eval_loss": 0.3553410470485687,
      "eval_precision": 0.9281137878011276,
      "eval_recall": 0.9281137878011276,
      "eval_runtime": 93.6408,
      "eval_samples_per_second": 83.34,
      "eval_steps_per_second": 10.423,
      "step": 3100
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.633086876155268e-05,
      "loss": 0.0481,
      "step": 3110
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.5868761552680226e-05,
      "loss": 0.0064,
      "step": 3120
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.540665434380776e-05,
      "loss": 0.0008,
      "step": 3130
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.494454713493531e-05,
      "loss": 0.0008,
      "step": 3140
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.448243992606285e-05,
      "loss": 0.0393,
      "step": 3150
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.402033271719039e-05,
      "loss": 0.0102,
      "step": 3160
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.355822550831793e-05,
      "loss": 0.0015,
      "step": 3170
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.309611829944548e-05,
      "loss": 0.0009,
      "step": 3180
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.263401109057301e-05,
      "loss": 0.0007,
      "step": 3190
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.217190388170056e-05,
      "loss": 0.0007,
      "step": 3200
    },
    {
      "epoch": 2.96,
      "eval_accuracy": 0.9284982060481805,
      "eval_f1": 0.9284982060481805,
      "eval_loss": 0.3595002293586731,
      "eval_precision": 0.9284982060481805,
      "eval_recall": 0.9284982060481805,
      "eval_runtime": 94.7075,
      "eval_samples_per_second": 82.401,
      "eval_steps_per_second": 10.305,
      "step": 3200
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.1709796672828106e-05,
      "loss": 0.0008,
      "step": 3210
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.124768946395564e-05,
      "loss": 0.0122,
      "step": 3220
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.078558225508319e-05,
      "loss": 0.0015,
      "step": 3230
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.032347504621072e-05,
      "loss": 0.0462,
      "step": 3240
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.986136783733826e-05,
      "loss": 0.0007,
      "step": 3250
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.93992606284658e-05,
      "loss": 0.0008,
      "step": 3260
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.893715341959334e-05,
      "loss": 0.0007,
      "step": 3270
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.8475046210720884e-05,
      "loss": 0.0007,
      "step": 3280
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.801293900184843e-05,
      "loss": 0.0007,
      "step": 3290
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.755083179297597e-05,
      "loss": 0.0238,
      "step": 3300
    },
    {
      "epoch": 3.05,
      "eval_accuracy": 0.9349051768323936,
      "eval_f1": 0.9349051768323936,
      "eval_loss": 0.3350887596607208,
      "eval_precision": 0.9349051768323936,
      "eval_recall": 0.9349051768323936,
      "eval_runtime": 93.8092,
      "eval_samples_per_second": 83.19,
      "eval_steps_per_second": 10.404,
      "step": 3300
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.708872458410351e-05,
      "loss": 0.0008,
      "step": 3310
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.662661737523106e-05,
      "loss": 0.0008,
      "step": 3320
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.61645101663586e-05,
      "loss": 0.0007,
      "step": 3330
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.570240295748614e-05,
      "loss": 0.0007,
      "step": 3340
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.524029574861368e-05,
      "loss": 0.0032,
      "step": 3350
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.4778188539741223e-05,
      "loss": 0.0085,
      "step": 3360
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.4316081330868764e-05,
      "loss": 0.0006,
      "step": 3370
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.3853974121996305e-05,
      "loss": 0.0218,
      "step": 3380
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.3391866913123846e-05,
      "loss": 0.0006,
      "step": 3390
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.2929759704251386e-05,
      "loss": 0.0065,
      "step": 3400
    },
    {
      "epoch": 3.14,
      "eval_accuracy": 0.9375961045617632,
      "eval_f1": 0.9375961045617632,
      "eval_loss": 0.3165816366672516,
      "eval_precision": 0.9375961045617632,
      "eval_recall": 0.9375961045617632,
      "eval_runtime": 93.6238,
      "eval_samples_per_second": 83.355,
      "eval_steps_per_second": 10.425,
      "step": 3400
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.246765249537893e-05,
      "loss": 0.0006,
      "step": 3410
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.200554528650647e-05,
      "loss": 0.0005,
      "step": 3420
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.154343807763401e-05,
      "loss": 0.0005,
      "step": 3430
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.1081330868761556e-05,
      "loss": 0.0005,
      "step": 3440
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.06192236598891e-05,
      "loss": 0.0006,
      "step": 3450
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.015711645101664e-05,
      "loss": 0.0005,
      "step": 3460
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.969500924214418e-05,
      "loss": 0.0445,
      "step": 3470
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9232902033271726e-05,
      "loss": 0.0007,
      "step": 3480
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.8770794824399266e-05,
      "loss": 0.0009,
      "step": 3490
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.830868761552681e-05,
      "loss": 0.0009,
      "step": 3500
    },
    {
      "epoch": 3.23,
      "eval_accuracy": 0.933111225012814,
      "eval_f1": 0.933111225012814,
      "eval_loss": 0.33037713170051575,
      "eval_precision": 0.933111225012814,
      "eval_recall": 0.933111225012814,
      "eval_runtime": 94.2067,
      "eval_samples_per_second": 82.839,
      "eval_steps_per_second": 10.36,
      "step": 3500
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.784658040665435e-05,
      "loss": 0.0008,
      "step": 3510
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.738447319778189e-05,
      "loss": 0.0055,
      "step": 3520
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.692236598890943e-05,
      "loss": 0.0008,
      "step": 3530
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.646025878003697e-05,
      "loss": 0.0006,
      "step": 3540
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.599815157116451e-05,
      "loss": 0.0209,
      "step": 3550
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.553604436229205e-05,
      "loss": 0.0007,
      "step": 3560
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.507393715341959e-05,
      "loss": 0.0006,
      "step": 3570
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.461182994454713e-05,
      "loss": 0.0009,
      "step": 3580
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.414972273567468e-05,
      "loss": 0.0013,
      "step": 3590
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.368761552680222e-05,
      "loss": 0.0006,
      "step": 3600
    },
    {
      "epoch": 3.33,
      "eval_accuracy": 0.9347770374167094,
      "eval_f1": 0.9347770374167094,
      "eval_loss": 0.3214694857597351,
      "eval_precision": 0.9347770374167094,
      "eval_recall": 0.9347770374167094,
      "eval_runtime": 93.9759,
      "eval_samples_per_second": 83.043,
      "eval_steps_per_second": 10.386,
      "step": 3600
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.322550831792976e-05,
      "loss": 0.0014,
      "step": 3610
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.27634011090573e-05,
      "loss": 0.0013,
      "step": 3620
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.2301293900184843e-05,
      "loss": 0.0005,
      "step": 3630
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.1839186691312384e-05,
      "loss": 0.0005,
      "step": 3640
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.1377079482439925e-05,
      "loss": 0.0006,
      "step": 3650
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.091497227356747e-05,
      "loss": 0.0005,
      "step": 3660
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.0452865064695013e-05,
      "loss": 0.0005,
      "step": 3670
    },
    {
      "epoch": 3.4,
      "learning_rate": 2.9990757855822554e-05,
      "loss": 0.0009,
      "step": 3680
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9528650646950095e-05,
      "loss": 0.001,
      "step": 3690
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.9066543438077635e-05,
      "loss": 0.0005,
      "step": 3700
    },
    {
      "epoch": 3.42,
      "eval_accuracy": 0.9352895950794464,
      "eval_f1": 0.9352895950794464,
      "eval_loss": 0.34141379594802856,
      "eval_precision": 0.9352895950794464,
      "eval_recall": 0.9352895950794464,
      "eval_runtime": 94.4131,
      "eval_samples_per_second": 82.658,
      "eval_steps_per_second": 10.338,
      "step": 3700
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.860443622920518e-05,
      "loss": 0.0006,
      "step": 3710
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.814232902033272e-05,
      "loss": 0.0098,
      "step": 3720
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.768022181146026e-05,
      "loss": 0.0004,
      "step": 3730
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.72181146025878e-05,
      "loss": 0.0005,
      "step": 3740
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6756007393715342e-05,
      "loss": 0.019,
      "step": 3750
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.6293900184842883e-05,
      "loss": 0.0006,
      "step": 3760
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.5831792975970427e-05,
      "loss": 0.0448,
      "step": 3770
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5369685767097968e-05,
      "loss": 0.0006,
      "step": 3780
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.490757855822551e-05,
      "loss": 0.0006,
      "step": 3790
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.4445471349353053e-05,
      "loss": 0.0006,
      "step": 3800
    },
    {
      "epoch": 3.51,
      "eval_accuracy": 0.935033316248078,
      "eval_f1": 0.935033316248078,
      "eval_loss": 0.3321364223957062,
      "eval_precision": 0.935033316248078,
      "eval_recall": 0.935033316248078,
      "eval_runtime": 94.2057,
      "eval_samples_per_second": 82.84,
      "eval_steps_per_second": 10.36,
      "step": 3800
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.3983364140480594e-05,
      "loss": 0.0006,
      "step": 3810
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.3521256931608134e-05,
      "loss": 0.0006,
      "step": 3820
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.3059149722735675e-05,
      "loss": 0.0006,
      "step": 3830
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.259704251386322e-05,
      "loss": 0.0016,
      "step": 3840
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.213493530499076e-05,
      "loss": 0.0017,
      "step": 3850
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.16728280961183e-05,
      "loss": 0.0005,
      "step": 3860
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.121072088724584e-05,
      "loss": 0.0384,
      "step": 3870
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0748613678373382e-05,
      "loss": 0.0009,
      "step": 3880
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.0286506469500923e-05,
      "loss": 0.0005,
      "step": 3890
    },
    {
      "epoch": 3.6,
      "learning_rate": 1.9824399260628467e-05,
      "loss": 0.0006,
      "step": 3900
    },
    {
      "epoch": 3.6,
      "eval_accuracy": 0.9378523833931317,
      "eval_f1": 0.9378523833931317,
      "eval_loss": 0.33057039976119995,
      "eval_precision": 0.9378523833931317,
      "eval_recall": 0.9378523833931317,
      "eval_runtime": 93.9934,
      "eval_samples_per_second": 83.027,
      "eval_steps_per_second": 10.384,
      "step": 3900
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9362292051756008e-05,
      "loss": 0.0056,
      "step": 3910
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.890018484288355e-05,
      "loss": 0.0005,
      "step": 3920
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8438077634011092e-05,
      "loss": 0.0005,
      "step": 3930
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.7975970425138633e-05,
      "loss": 0.0005,
      "step": 3940
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.7513863216266174e-05,
      "loss": 0.0005,
      "step": 3950
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.7051756007393715e-05,
      "loss": 0.0005,
      "step": 3960
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6589648798521255e-05,
      "loss": 0.0005,
      "step": 3970
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.61275415896488e-05,
      "loss": 0.0005,
      "step": 3980
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.5665434380776344e-05,
      "loss": 0.0005,
      "step": 3990
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5203327171903883e-05,
      "loss": 0.0411,
      "step": 4000
    },
    {
      "epoch": 3.7,
      "eval_accuracy": 0.9374679651460789,
      "eval_f1": 0.9374679651460789,
      "eval_loss": 0.3225729167461395,
      "eval_precision": 0.9374679651460789,
      "eval_recall": 0.9374679651460789,
      "eval_runtime": 93.7031,
      "eval_samples_per_second": 83.284,
      "eval_steps_per_second": 10.416,
      "step": 4000
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4741219963031425e-05,
      "loss": 0.0005,
      "step": 4010
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.4279112754158966e-05,
      "loss": 0.0006,
      "step": 4020
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.3817005545286506e-05,
      "loss": 0.0006,
      "step": 4030
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3354898336414049e-05,
      "loss": 0.0005,
      "step": 4040
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.289279112754159e-05,
      "loss": 0.0005,
      "step": 4050
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2430683918669132e-05,
      "loss": 0.0005,
      "step": 4060
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.1968576709796673e-05,
      "loss": 0.0006,
      "step": 4070
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1506469500924215e-05,
      "loss": 0.0006,
      "step": 4080
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1044362292051756e-05,
      "loss": 0.0005,
      "step": 4090
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0582255083179297e-05,
      "loss": 0.0006,
      "step": 4100
    },
    {
      "epoch": 3.79,
      "eval_accuracy": 0.9377242439774475,
      "eval_f1": 0.9377242439774475,
      "eval_loss": 0.3234710097312927,
      "eval_precision": 0.9377242439774475,
      "eval_recall": 0.9377242439774475,
      "eval_runtime": 93.455,
      "eval_samples_per_second": 83.505,
      "eval_steps_per_second": 10.444,
      "step": 4100
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.012014787430684e-05,
      "loss": 0.0005,
      "step": 4110
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.658040665434381e-06,
      "loss": 0.0005,
      "step": 4120
    },
    {
      "epoch": 3.82,
      "learning_rate": 9.195933456561922e-06,
      "loss": 0.0005,
      "step": 4130
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.733826247689465e-06,
      "loss": 0.0005,
      "step": 4140
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.271719038817007e-06,
      "loss": 0.0005,
      "step": 4150
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.809611829944548e-06,
      "loss": 0.0005,
      "step": 4160
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.347504621072089e-06,
      "loss": 0.0006,
      "step": 4170
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.88539741219963e-06,
      "loss": 0.0005,
      "step": 4180
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.423290203327172e-06,
      "loss": 0.0729,
      "step": 4190
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.961182994454713e-06,
      "loss": 0.0354,
      "step": 4200
    },
    {
      "epoch": 3.88,
      "eval_accuracy": 0.9356740133264992,
      "eval_f1": 0.9356740133264994,
      "eval_loss": 0.326831191778183,
      "eval_precision": 0.9356740133264992,
      "eval_recall": 0.9356740133264992,
      "eval_runtime": 94.0495,
      "eval_samples_per_second": 82.978,
      "eval_steps_per_second": 10.378,
      "step": 4200
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.499075785582256e-06,
      "loss": 0.0007,
      "step": 4210
    },
    {
      "epoch": 3.9,
      "learning_rate": 5.036968576709796e-06,
      "loss": 0.0009,
      "step": 4220
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.574861367837339e-06,
      "loss": 0.0007,
      "step": 4230
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.11275415896488e-06,
      "loss": 0.0007,
      "step": 4240
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.650646950092421e-06,
      "loss": 0.0008,
      "step": 4250
    },
    {
      "epoch": 3.94,
      "learning_rate": 3.188539741219963e-06,
      "loss": 0.0007,
      "step": 4260
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.7264325323475047e-06,
      "loss": 0.0007,
      "step": 4270
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.2643253234750463e-06,
      "loss": 0.0037,
      "step": 4280
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.8022181146025879e-06,
      "loss": 0.0007,
      "step": 4290
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.3401109057301294e-06,
      "loss": 0.0328,
      "step": 4300
    },
    {
      "epoch": 3.97,
      "eval_accuracy": 0.9358021527421835,
      "eval_f1": 0.9358021527421835,
      "eval_loss": 0.3264217674732208,
      "eval_precision": 0.9358021527421835,
      "eval_recall": 0.9358021527421835,
      "eval_runtime": 94.2652,
      "eval_samples_per_second": 82.788,
      "eval_steps_per_second": 10.354,
      "step": 4300
    },
    {
      "epoch": 3.98,
      "learning_rate": 8.78003696857671e-07,
      "loss": 0.0008,
      "step": 4310
    },
    {
      "epoch": 3.99,
      "learning_rate": 4.158964879852126e-07,
      "loss": 0.0006,
      "step": 4320
    },
    {
      "epoch": 4.0,
      "step": 4328,
      "total_flos": 3.943158668110371e+19,
      "train_loss": 0.07091429132740952,
      "train_runtime": 8488.5621,
      "train_samples_per_second": 8.154,
      "train_steps_per_second": 0.51
    }
  ],
  "max_steps": 4328,
  "num_train_epochs": 4,
  "total_flos": 3.943158668110371e+19,
  "trial_name": null,
  "trial_params": null
}