{
  "best_metric": 0.9945213646234112,
  "best_model_checkpoint": "output_dev/v39-albert_xxlarge_v2-f1-warmup-greater-patience-seqlen-240-6epoch-2e-5/checkpoint-12600",
  "epoch": 4.973037747153985,
  "eval_steps": 200,
  "global_step": 16600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.9960079840319364e-07,
      "loss": 1.4634,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 3.9920159680638727e-07,
      "loss": 1.4384,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 5.988023952095809e-07,
      "loss": 1.3844,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 7.984031936127745e-07,
      "loss": 1.3105,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 9.980039920159682e-07,
      "loss": 1.2166,
      "step": 50
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.1976047904191619e-06,
      "loss": 1.096,
      "step": 60
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.3972055888223554e-06,
      "loss": 0.9712,
      "step": 70
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.596806387225549e-06,
      "loss": 0.8719,
      "step": 80
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.7964071856287426e-06,
      "loss": 0.7869,
      "step": 90
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.9960079840319363e-06,
      "loss": 0.6573,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.19560878243513e-06,
      "loss": 0.7487,
      "step": 110
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.3952095808383237e-06,
      "loss": 0.7163,
      "step": 120
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.594810379241517e-06,
      "loss": 0.6521,
      "step": 130
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.7944111776447107e-06,
      "loss": 0.5949,
      "step": 140
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.9940119760479042e-06,
      "loss": 0.565,
      "step": 150
    },
    {
      "epoch": 0.05,
      "learning_rate": 3.193612774451098e-06,
      "loss": 0.5002,
      "step": 160
    },
    {
      "epoch": 0.05,
      "learning_rate": 3.3932135728542912e-06,
      "loss": 0.357,
      "step": 170
    },
    {
      "epoch": 0.05,
      "learning_rate": 3.592814371257485e-06,
      "loss": 0.3758,
      "step": 180
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.7924151696606787e-06,
      "loss": 0.3239,
      "step": 190
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.992015968063873e-06,
      "loss": 0.3293,
      "step": 200
    },
    {
      "epoch": 0.06,
      "eval_f1": 0.7083361387123823,
      "eval_f1_0": 0.9510483607396013,
      "eval_f1_1": 0.9524008736274194,
      "eval_f1_2": 0.0,
      "eval_f1_3": 0.9298953204825086,
      "eval_loss": 0.2948536276817322,
      "eval_runtime": 246.9405,
      "eval_samples_per_second": 48.06,
      "eval_steps_per_second": 0.19,
      "step": 200
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.191616766467066e-06,
      "loss": 0.2598,
      "step": 210
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.39121756487026e-06,
      "loss": 0.2467,
      "step": 220
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.5908183632734535e-06,
      "loss": 0.2446,
      "step": 230
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.7904191616766475e-06,
      "loss": 0.255,
      "step": 240
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.9900199600798405e-06,
      "loss": 0.2325,
      "step": 250
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.189620758483034e-06,
      "loss": 0.2231,
      "step": 260
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.389221556886228e-06,
      "loss": 0.2053,
      "step": 270
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.5888223552894215e-06,
      "loss": 0.1621,
      "step": 280
    },
    {
      "epoch": 0.09,
      "learning_rate": 5.788423153692615e-06,
      "loss": 0.1894,
      "step": 290
    },
    {
      "epoch": 0.09,
      "learning_rate": 5.9880239520958085e-06,
      "loss": 0.1616,
      "step": 300
    },
    {
      "epoch": 0.09,
      "learning_rate": 6.187624750499002e-06,
      "loss": 0.1562,
      "step": 310
    },
    {
      "epoch": 0.1,
      "learning_rate": 6.387225548902196e-06,
      "loss": 0.1327,
      "step": 320
    },
    {
      "epoch": 0.1,
      "learning_rate": 6.586826347305389e-06,
      "loss": 0.1985,
      "step": 330
    },
    {
      "epoch": 0.1,
      "learning_rate": 6.7864271457085825e-06,
      "loss": 0.1468,
      "step": 340
    },
    {
      "epoch": 0.1,
      "learning_rate": 6.986027944111777e-06,
      "loss": 0.1134,
      "step": 350
    },
    {
      "epoch": 0.11,
      "learning_rate": 7.18562874251497e-06,
      "loss": 0.1251,
      "step": 360
    },
    {
      "epoch": 0.11,
      "learning_rate": 7.385229540918164e-06,
      "loss": 0.1375,
      "step": 370
    },
    {
      "epoch": 0.11,
      "learning_rate": 7.584830339321357e-06,
      "loss": 0.1153,
      "step": 380
    },
    {
      "epoch": 0.12,
      "learning_rate": 7.784431137724551e-06,
      "loss": 0.1418,
      "step": 390
    },
    {
      "epoch": 0.12,
      "learning_rate": 7.984031936127745e-06,
      "loss": 0.1455,
      "step": 400
    },
    {
      "epoch": 0.12,
      "eval_f1": 0.9228543978673958,
      "eval_f1_0": 0.9778625129746034,
      "eval_f1_1": 0.9741895022164059,
      "eval_f1_2": 0.7576093720407998,
      "eval_f1_3": 0.981756204237774,
      "eval_loss": 0.11195129156112671,
      "eval_runtime": 246.6995,
      "eval_samples_per_second": 48.107,
      "eval_steps_per_second": 0.191,
      "step": 400
    },
    {
      "epoch": 0.12,
      "learning_rate": 8.183632734530939e-06,
      "loss": 0.1266,
      "step": 410
    },
    {
      "epoch": 0.13,
      "learning_rate": 8.383233532934131e-06,
      "loss": 0.1442,
      "step": 420
    },
    {
      "epoch": 0.13,
      "learning_rate": 8.582834331337327e-06,
      "loss": 0.1122,
      "step": 430
    },
    {
      "epoch": 0.13,
      "learning_rate": 8.78243512974052e-06,
      "loss": 0.1085,
      "step": 440
    },
    {
      "epoch": 0.13,
      "learning_rate": 8.982035928143713e-06,
      "loss": 0.1448,
      "step": 450
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.181636726546907e-06,
      "loss": 0.1231,
      "step": 460
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.381237524950101e-06,
      "loss": 0.0911,
      "step": 470
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.580838323353295e-06,
      "loss": 0.1129,
      "step": 480
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.780439121756487e-06,
      "loss": 0.1313,
      "step": 490
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.980039920159681e-06,
      "loss": 0.1467,
      "step": 500
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.0179640718562875e-05,
      "loss": 0.0686,
      "step": 510
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.0379241516966067e-05,
      "loss": 0.097,
      "step": 520
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.0578842315369263e-05,
      "loss": 0.1367,
      "step": 530
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.0778443113772457e-05,
      "loss": 0.1031,
      "step": 540
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.0978043912175649e-05,
      "loss": 0.0803,
      "step": 550
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.1177644710578843e-05,
      "loss": 0.0423,
      "step": 560
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.1377245508982037e-05,
      "loss": 0.1045,
      "step": 570
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.157684630738523e-05,
      "loss": 0.0637,
      "step": 580
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.1776447105788425e-05,
      "loss": 0.0959,
      "step": 590
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.1976047904191617e-05,
      "loss": 0.0474,
      "step": 600
    },
    {
      "epoch": 0.18,
      "eval_f1": 0.9674870062636107,
      "eval_f1_0": 0.9895694665223942,
      "eval_f1_1": 0.9843515711932425,
      "eval_f1_2": 0.9052149189002555,
      "eval_f1_3": 0.9908120684385506,
      "eval_loss": 0.05463177338242531,
      "eval_runtime": 246.6808,
      "eval_samples_per_second": 48.111,
      "eval_steps_per_second": 0.191,
      "step": 600
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.2175648702594811e-05,
      "loss": 0.0389,
      "step": 610
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.2375249500998005e-05,
      "loss": 0.14,
      "step": 620
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.2574850299401197e-05,
      "loss": 0.0626,
      "step": 630
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.2774451097804393e-05,
      "loss": 0.0754,
      "step": 640
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.2974051896207587e-05,
      "loss": 0.0714,
      "step": 650
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3173652694610779e-05,
      "loss": 0.0847,
      "step": 660
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3373253493013973e-05,
      "loss": 0.0346,
      "step": 670
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3572854291417165e-05,
      "loss": 0.0848,
      "step": 680
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.377245508982036e-05,
      "loss": 0.0326,
      "step": 690
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.3972055888223555e-05,
      "loss": 0.0635,
      "step": 700
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.4171656686626747e-05,
      "loss": 0.0522,
      "step": 710
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.437125748502994e-05,
      "loss": 0.0387,
      "step": 720
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.4570858283433136e-05,
      "loss": 0.0442,
      "step": 730
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.4770459081836329e-05,
      "loss": 0.041,
      "step": 740
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.4970059880239522e-05,
      "loss": 0.0528,
      "step": 750
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.5169660678642715e-05,
      "loss": 0.0594,
      "step": 760
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.536926147704591e-05,
      "loss": 0.0732,
      "step": 770
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.5568862275449103e-05,
      "loss": 0.0665,
      "step": 780
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.5768463073852296e-05,
      "loss": 0.078,
      "step": 790
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.596806387225549e-05,
      "loss": 0.1295,
      "step": 800
    },
    {
      "epoch": 0.24,
      "eval_f1": 0.9685681697321592,
      "eval_f1_0": 0.9895594983846618,
      "eval_f1_1": 0.9916923059183913,
      "eval_f1_2": 0.8975888192995841,
      "eval_f1_3": 0.9954320553259993,
      "eval_loss": 0.05254097282886505,
      "eval_runtime": 246.573,
      "eval_samples_per_second": 48.132,
      "eval_steps_per_second": 0.191,
      "step": 800
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.6167664670658684e-05,
      "loss": 0.0733,
      "step": 810
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.6367265469061878e-05,
      "loss": 0.0518,
      "step": 820
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.6566866267465072e-05,
      "loss": 0.0179,
      "step": 830
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.6766467065868263e-05,
      "loss": 0.022,
      "step": 840
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.6966067864271457e-05,
      "loss": 0.0546,
      "step": 850
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.7165668662674654e-05,
      "loss": 0.0468,
      "step": 860
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.7365269461077845e-05,
      "loss": 0.0356,
      "step": 870
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.756487025948104e-05,
      "loss": 0.0334,
      "step": 880
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.7764471057884232e-05,
      "loss": 0.0719,
      "step": 890
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.7964071856287426e-05,
      "loss": 0.019,
      "step": 900
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.816367265469062e-05,
      "loss": 0.0735,
      "step": 910
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.8363273453093814e-05,
      "loss": 0.0296,
      "step": 920
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.8562874251497008e-05,
      "loss": 0.0632,
      "step": 930
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.8762475049900202e-05,
      "loss": 0.097,
      "step": 940
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.8962075848303393e-05,
      "loss": 0.0646,
      "step": 950
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.916167664670659e-05,
      "loss": 0.0557,
      "step": 960
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9361277445109784e-05,
      "loss": 0.0734,
      "step": 970
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9560878243512974e-05,
      "loss": 0.015,
      "step": 980
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9760479041916168e-05,
      "loss": 0.0926,
      "step": 990
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9960079840319362e-05,
      "loss": 0.0183,
      "step": 1000
    },
    {
      "epoch": 0.3,
      "eval_f1": 0.9743006838660146,
      "eval_f1_0": 0.9923671722411238,
      "eval_f1_1": 0.99328476816095,
      "eval_f1_2": 0.9141021246434236,
      "eval_f1_3": 0.997448670418561,
      "eval_loss": 0.04545324668288231,
      "eval_runtime": 246.8525,
      "eval_samples_per_second": 48.077,
      "eval_steps_per_second": 0.19,
      "step": 1000
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9991590455166615e-05,
      "loss": 0.0387,
      "step": 1010
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9981078524124883e-05,
      "loss": 0.0566,
      "step": 1020
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.997056659308315e-05,
      "loss": 0.0629,
      "step": 1030
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.996005466204142e-05,
      "loss": 0.0328,
      "step": 1040
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9949542730999687e-05,
      "loss": 0.0494,
      "step": 1050
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.9939030799957955e-05,
      "loss": 0.0657,
      "step": 1060
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.9928518868916223e-05,
      "loss": 0.0267,
      "step": 1070
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.991800693787449e-05,
      "loss": 0.0953,
      "step": 1080
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.990749500683276e-05,
      "loss": 0.0764,
      "step": 1090
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.9896983075791027e-05,
      "loss": 0.013,
      "step": 1100
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.9886471144749292e-05,
      "loss": 0.0564,
      "step": 1110
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.987595921370756e-05,
      "loss": 0.0465,
      "step": 1120
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.9865447282665828e-05,
      "loss": 0.0439,
      "step": 1130
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.9854935351624092e-05,
      "loss": 0.074,
      "step": 1140
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.984442342058236e-05,
      "loss": 0.0524,
      "step": 1150
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.983391148954063e-05,
      "loss": 0.0616,
      "step": 1160
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.9823399558498897e-05,
      "loss": 0.0869,
      "step": 1170
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.9812887627457165e-05,
      "loss": 0.0689,
      "step": 1180
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.9802375696415433e-05,
      "loss": 0.0209,
      "step": 1190
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.97918637653737e-05,
      "loss": 0.0379,
      "step": 1200
    },
    {
      "epoch": 0.36,
      "eval_f1": 0.9820939596049222,
      "eval_f1_0": 0.9942978550058829,
      "eval_f1_1": 0.9875073502911663,
      "eval_f1_2": 0.9505474052988833,
      "eval_f1_3": 0.9960232278237564,
      "eval_loss": 0.033771250396966934,
      "eval_runtime": 246.8921,
      "eval_samples_per_second": 48.07,
      "eval_steps_per_second": 0.19,
      "step": 1200
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.978135183433197e-05,
      "loss": 0.0284,
      "step": 1210
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.9770839903290237e-05,
      "loss": 0.0198,
      "step": 1220
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.97603279722485e-05,
      "loss": 0.0112,
      "step": 1230
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.974981604120677e-05,
      "loss": 0.0247,
      "step": 1240
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.9739304110165038e-05,
      "loss": 0.057,
      "step": 1250
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.9728792179123306e-05,
      "loss": 0.0607,
      "step": 1260
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.9718280248081574e-05,
      "loss": 0.0521,
      "step": 1270
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.970776831703984e-05,
      "loss": 0.0375,
      "step": 1280
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.969725638599811e-05,
      "loss": 0.0643,
      "step": 1290
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.9686744454956378e-05,
      "loss": 0.0274,
      "step": 1300
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.9676232523914646e-05,
      "loss": 0.0302,
      "step": 1310
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9665720592872914e-05,
      "loss": 0.0508,
      "step": 1320
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.965520866183118e-05,
      "loss": 0.0544,
      "step": 1330
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9644696730789446e-05,
      "loss": 0.0413,
      "step": 1340
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9634184799747715e-05,
      "loss": 0.0194,
      "step": 1350
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.9623672868705983e-05,
      "loss": 0.0297,
      "step": 1360
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.961316093766425e-05,
      "loss": 0.0242,
      "step": 1370
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.960264900662252e-05,
      "loss": 0.02,
      "step": 1380
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.9592137075580787e-05,
      "loss": 0.0381,
      "step": 1390
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.9581625144539055e-05,
      "loss": 0.056,
      "step": 1400
    },
    {
      "epoch": 0.42,
      "eval_f1": 0.9780869363316178,
      "eval_f1_0": 0.9947891529230842,
      "eval_f1_1": 0.9768774467468426,
      "eval_f1_2": 0.9582269358895914,
      "eval_f1_3": 0.9824542097669527,
      "eval_loss": 0.036463286727666855,
      "eval_runtime": 246.9324,
      "eval_samples_per_second": 48.062,
      "eval_steps_per_second": 0.19,
      "step": 1400
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.9571113213497323e-05,
      "loss": 0.044,
      "step": 1410
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.9560601282455587e-05,
      "loss": 0.0407,
      "step": 1420
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.9550089351413855e-05,
      "loss": 0.0146,
      "step": 1430
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.9539577420372124e-05,
      "loss": 0.0488,
      "step": 1440
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.952906548933039e-05,
      "loss": 0.0378,
      "step": 1450
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.951855355828866e-05,
      "loss": 0.0232,
      "step": 1460
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.9508041627246928e-05,
      "loss": 0.0833,
      "step": 1470
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.9497529696205196e-05,
      "loss": 0.0928,
      "step": 1480
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.9487017765163464e-05,
      "loss": 0.0542,
      "step": 1490
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.9476505834121732e-05,
      "loss": 0.066,
      "step": 1500
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.946599390308e-05,
      "loss": 0.0571,
      "step": 1510
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.9455481972038264e-05,
      "loss": 0.0314,
      "step": 1520
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.9444970040996532e-05,
      "loss": 0.0255,
      "step": 1530
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.94344581099548e-05,
      "loss": 0.0473,
      "step": 1540
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.942394617891307e-05,
      "loss": 0.0654,
      "step": 1550
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.9413434247871333e-05,
      "loss": 0.0465,
      "step": 1560
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.94029223168296e-05,
      "loss": 0.0504,
      "step": 1570
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.939241038578787e-05,
      "loss": 0.0873,
      "step": 1580
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.9381898454746137e-05,
      "loss": 0.0262,
      "step": 1590
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.9371386523704405e-05,
      "loss": 0.0652,
      "step": 1600
    },
    {
      "epoch": 0.48,
      "eval_f1": 0.969560375179775,
      "eval_f1_0": 0.9895699528636753,
      "eval_f1_1": 0.9949528484741482,
      "eval_f1_2": 0.8963498614105327,
      "eval_f1_3": 0.9973688379707435,
      "eval_loss": 0.05588960275053978,
      "eval_runtime": 246.8129,
      "eval_samples_per_second": 48.085,
      "eval_steps_per_second": 0.19,
      "step": 1600
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.9360874592662673e-05,
      "loss": 0.0621,
      "step": 1610
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.935036266162094e-05,
      "loss": 0.0405,
      "step": 1620
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.933985073057921e-05,
      "loss": 0.0356,
      "step": 1630
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.9329338799537478e-05,
      "loss": 0.0226,
      "step": 1640
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.9318826868495742e-05,
      "loss": 0.0533,
      "step": 1650
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.930831493745401e-05,
      "loss": 0.0689,
      "step": 1660
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.9297803006412278e-05,
      "loss": 0.0602,
      "step": 1670
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.9287291075370546e-05,
      "loss": 0.0302,
      "step": 1680
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.9276779144328814e-05,
      "loss": 0.0355,
      "step": 1690
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.9266267213287082e-05,
      "loss": 0.009,
      "step": 1700
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.925575528224535e-05,
      "loss": 0.0446,
      "step": 1710
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.924524335120362e-05,
      "loss": 0.0185,
      "step": 1720
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.9234731420161887e-05,
      "loss": 0.0322,
      "step": 1730
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.922421948912015e-05,
      "loss": 0.0363,
      "step": 1740
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.921370755807842e-05,
      "loss": 0.0461,
      "step": 1750
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.9203195627036687e-05,
      "loss": 0.0316,
      "step": 1760
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.9192683695994955e-05,
      "loss": 0.0271,
      "step": 1770
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.9182171764953223e-05,
      "loss": 0.0239,
      "step": 1780
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.917165983391149e-05,
      "loss": 0.0446,
      "step": 1790
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.916114790286976e-05,
      "loss": 0.0186,
      "step": 1800
    },
    {
      "epoch": 0.54,
      "eval_f1": 0.9839843532040691,
      "eval_f1_0": 0.9951651963763937,
      "eval_f1_1": 0.9957723639643443,
      "eval_f1_2": 0.9474666869118015,
      "eval_f1_3": 0.997533165563737,
      "eval_loss": 0.03348803147673607,
      "eval_runtime": 247.0218,
      "eval_samples_per_second": 48.044,
      "eval_steps_per_second": 0.19,
      "step": 1800
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.9150635971828027e-05,
      "loss": 0.0505,
      "step": 1810
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.9140124040786295e-05,
      "loss": 0.0206,
      "step": 1820
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.9129612109744564e-05,
      "loss": 0.0231,
      "step": 1830
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.9119100178702828e-05,
      "loss": 0.0617,
      "step": 1840
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.9108588247661096e-05,
      "loss": 0.0634,
      "step": 1850
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.9098076316619364e-05,
      "loss": 0.0196,
      "step": 1860
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.9087564385577632e-05,
      "loss": 0.046,
      "step": 1870
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.90770524545359e-05,
      "loss": 0.0989,
      "step": 1880
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.906654052349417e-05,
      "loss": 0.0267,
      "step": 1890
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.9056028592452436e-05,
      "loss": 0.0312,
      "step": 1900
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.9045516661410704e-05,
      "loss": 0.0394,
      "step": 1910
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.9035004730368973e-05,
      "loss": 0.0496,
      "step": 1920
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.9024492799327237e-05,
      "loss": 0.0217,
      "step": 1930
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.9013980868285505e-05,
      "loss": 0.0327,
      "step": 1940
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.9003468937243773e-05,
      "loss": 0.0378,
      "step": 1950
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.899295700620204e-05,
      "loss": 0.0121,
      "step": 1960
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.898244507516031e-05,
      "loss": 0.0146,
      "step": 1970
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.8971933144118577e-05,
      "loss": 0.0239,
      "step": 1980
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.8961421213076842e-05,
      "loss": 0.0647,
      "step": 1990
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.895090928203511e-05,
      "loss": 0.0354,
      "step": 2000
    },
    {
      "epoch": 0.6,
      "eval_f1": 0.9850313778222377,
      "eval_f1_0": 0.995182602344066,
      "eval_f1_1": 0.9960927984348177,
      "eval_f1_2": 0.9506635498343746,
      "eval_f1_3": 0.9981865606756924,
      "eval_loss": 0.02872946672141552,
      "eval_runtime": 246.7783,
      "eval_samples_per_second": 48.092,
      "eval_steps_per_second": 0.19,
      "step": 2000
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.8940397350993378e-05,
      "loss": 0.019,
      "step": 2010
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.8929885419951646e-05,
      "loss": 0.0405,
      "step": 2020
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.8919373488909914e-05,
      "loss": 0.0371,
      "step": 2030
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.8908861557868182e-05,
      "loss": 0.0263,
      "step": 2040
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.889834962682645e-05,
      "loss": 0.0307,
      "step": 2050
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8887837695784715e-05,
      "loss": 0.0282,
      "step": 2060
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8877325764742983e-05,
      "loss": 0.0202,
      "step": 2070
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.886681383370125e-05,
      "loss": 0.0917,
      "step": 2080
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.885630190265952e-05,
      "loss": 0.0567,
      "step": 2090
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8845789971617787e-05,
      "loss": 0.013,
      "step": 2100
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8835278040576055e-05,
      "loss": 0.0299,
      "step": 2110
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.8824766109534323e-05,
      "loss": 0.0068,
      "step": 2120
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.881425417849259e-05,
      "loss": 0.0425,
      "step": 2130
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.880374224745086e-05,
      "loss": 0.0519,
      "step": 2140
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.8793230316409127e-05,
      "loss": 0.0297,
      "step": 2150
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.8782718385367392e-05,
      "loss": 0.0117,
      "step": 2160
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.877220645432566e-05,
      "loss": 0.0112,
      "step": 2170
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.8761694523283928e-05,
      "loss": 0.0313,
      "step": 2180
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.8751182592242196e-05,
      "loss": 0.0132,
      "step": 2190
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.8740670661200464e-05,
      "loss": 0.0084,
      "step": 2200
    },
    {
      "epoch": 0.66,
      "eval_f1": 0.9894205624581704,
      "eval_f1_0": 0.9967117531743364,
      "eval_f1_1": 0.9955099152850286,
      "eval_f1_2": 0.9673987946140468,
      "eval_f1_3": 0.9980617867592702,
      "eval_loss": 0.02015283890068531,
      "eval_runtime": 246.9806,
      "eval_samples_per_second": 48.052,
      "eval_steps_per_second": 0.19,
      "step": 2200
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.8730158730158732e-05,
      "loss": 0.0244,
      "step": 2210
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.8719646799117e-05,
      "loss": 0.0159,
      "step": 2220
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.8709134868075268e-05,
      "loss": 0.0268,
      "step": 2230
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.8698622937033536e-05,
      "loss": 0.0358,
      "step": 2240
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.86881110059918e-05,
      "loss": 0.0132,
      "step": 2250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.867759907495007e-05,
      "loss": 0.0236,
      "step": 2260
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.8667087143908337e-05,
      "loss": 0.0106,
      "step": 2270
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.8656575212866605e-05,
      "loss": 0.0348,
      "step": 2280
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.8646063281824873e-05,
      "loss": 0.0174,
      "step": 2290
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.863555135078314e-05,
      "loss": 0.0765,
      "step": 2300
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.862503941974141e-05,
      "loss": 0.0175,
      "step": 2310
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.8614527488699677e-05,
      "loss": 0.0235,
      "step": 2320
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.8604015557657945e-05,
      "loss": 0.0287,
      "step": 2330
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.8593503626616213e-05,
      "loss": 0.0069,
      "step": 2340
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.8582991695574478e-05,
      "loss": 0.031,
      "step": 2350
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.8572479764532746e-05,
      "loss": 0.0065,
      "step": 2360
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.8561967833491014e-05,
      "loss": 0.0333,
      "step": 2370
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.8551455902449282e-05,
      "loss": 0.0189,
      "step": 2380
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.854094397140755e-05,
      "loss": 0.0098,
      "step": 2390
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.8530432040365818e-05,
      "loss": 0.0329,
      "step": 2400
    },
    {
      "epoch": 0.72,
      "eval_f1": 0.9904604467932125,
      "eval_f1_0": 0.9971255232585917,
      "eval_f1_1": 0.9961039309992745,
      "eval_f1_2": 0.97150592142995,
      "eval_f1_3": 0.9971064114850343,
      "eval_loss": 0.02072666585445404,
      "eval_runtime": 247.0645,
      "eval_samples_per_second": 48.036,
      "eval_steps_per_second": 0.19,
      "step": 2400
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.8519920109324083e-05,
      "loss": 0.0384,
      "step": 2410
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.850940817828235e-05,
      "loss": 0.0064,
      "step": 2420
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.849889624724062e-05,
      "loss": 0.0308,
      "step": 2430
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.8488384316198887e-05,
      "loss": 0.0294,
      "step": 2440
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.8477872385157155e-05,
      "loss": 0.0193,
      "step": 2450
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.8467360454115423e-05,
      "loss": 0.0235,
      "step": 2460
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.845684852307369e-05,
      "loss": 0.0211,
      "step": 2470
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.8446336592031956e-05,
      "loss": 0.0155,
      "step": 2480
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.8435824660990224e-05,
      "loss": 0.0064,
      "step": 2490
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.8425312729948492e-05,
      "loss": 0.0576,
      "step": 2500
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.841480079890676e-05,
      "loss": 0.0168,
      "step": 2510
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.8404288867865028e-05,
      "loss": 0.0084,
      "step": 2520
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.8393776936823296e-05,
      "loss": 0.0646,
      "step": 2530
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.8383265005781564e-05,
      "loss": 0.0175,
      "step": 2540
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.8372753074739832e-05,
      "loss": 0.0206,
      "step": 2550
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.83622411436981e-05,
      "loss": 0.0298,
      "step": 2560
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.8351729212656365e-05,
      "loss": 0.0173,
      "step": 2570
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.8341217281614633e-05,
      "loss": 0.0218,
      "step": 2580
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.83307053505729e-05,
      "loss": 0.0119,
      "step": 2590
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.832019341953117e-05,
      "loss": 0.0188,
      "step": 2600
    },
    {
      "epoch": 0.78,
      "eval_f1": 0.9881349379681611,
      "eval_f1_0": 0.9963530192776859,
      "eval_f1_1": 0.9967045164802932,
      "eval_f1_2": 0.9638271848277762,
      "eval_f1_3": 0.9956550312868894,
      "eval_loss": 0.02433362789452076,
      "eval_runtime": 247.0007,
      "eval_samples_per_second": 48.048,
      "eval_steps_per_second": 0.19,
      "step": 2600
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.8309681488489437e-05,
      "loss": 0.0143,
      "step": 2610
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.8299169557447705e-05,
      "loss": 0.0382,
      "step": 2620
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.8288657626405973e-05,
      "loss": 0.0519,
      "step": 2630
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.827814569536424e-05,
      "loss": 0.0063,
      "step": 2640
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.826763376432251e-05,
      "loss": 0.0213,
      "step": 2650
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.8257121833280777e-05,
      "loss": 0.0755,
      "step": 2660
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.824660990223904e-05,
      "loss": 0.0286,
      "step": 2670
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.823609797119731e-05,
      "loss": 0.0381,
      "step": 2680
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.8225586040155578e-05,
      "loss": 0.0078,
      "step": 2690
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.8215074109113846e-05,
      "loss": 0.0404,
      "step": 2700
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.8204562178072114e-05,
      "loss": 0.0494,
      "step": 2710
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.8194050247030382e-05,
      "loss": 0.0429,
      "step": 2720
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.818353831598865e-05,
      "loss": 0.0062,
      "step": 2730
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.8173026384946918e-05,
      "loss": 0.0267,
      "step": 2740
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.8162514453905186e-05,
      "loss": 0.0263,
      "step": 2750
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.815200252286345e-05,
      "loss": 0.0263,
      "step": 2760
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.814149059182172e-05,
      "loss": 0.0222,
      "step": 2770
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.8130978660779987e-05,
      "loss": 0.0516,
      "step": 2780
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.8120466729738255e-05,
      "loss": 0.0384,
      "step": 2790
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.8109954798696523e-05,
      "loss": 0.0226,
      "step": 2800
    },
    {
      "epoch": 0.84,
      "eval_f1": 0.9838635097796292,
      "eval_f1_0": 0.9951853839018016,
      "eval_f1_1": 0.9970718349642465,
      "eval_f1_2": 0.946153195090986,
      "eval_f1_3": 0.9970436251614827,
      "eval_loss": 0.03046056441962719,
      "eval_runtime": 246.9789,
      "eval_samples_per_second": 48.053,
      "eval_steps_per_second": 0.19,
      "step": 2800
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.809944286765479e-05,
      "loss": 0.0489,
      "step": 2810
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.808893093661306e-05,
      "loss": 0.0516,
      "step": 2820
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.8078419005571327e-05,
      "loss": 0.0364,
      "step": 2830
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.806790707452959e-05,
      "loss": 0.0179,
      "step": 2840
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.805739514348786e-05,
      "loss": 0.0151,
      "step": 2850
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.8046883212446128e-05,
      "loss": 0.0181,
      "step": 2860
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.8036371281404396e-05,
      "loss": 0.024,
      "step": 2870
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.8025859350362664e-05,
      "loss": 0.0657,
      "step": 2880
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.801534741932093e-05,
      "loss": 0.0146,
      "step": 2890
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.8004835488279196e-05,
      "loss": 0.0503,
      "step": 2900
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.7994323557237464e-05,
      "loss": 0.0084,
      "step": 2910
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.7983811626195732e-05,
      "loss": 0.0243,
      "step": 2920
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.7973299695154e-05,
      "loss": 0.1062,
      "step": 2930
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.796278776411227e-05,
      "loss": 0.0297,
      "step": 2940
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.7952275833070537e-05,
      "loss": 0.0177,
      "step": 2950
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.7941763902028805e-05,
      "loss": 0.0282,
      "step": 2960
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.7931251970987073e-05,
      "loss": 0.0134,
      "step": 2970
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.792074003994534e-05,
      "loss": 0.0165,
      "step": 2980
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.7910228108903605e-05,
      "loss": 0.0163,
      "step": 2990
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.7899716177861873e-05,
      "loss": 0.0209,
      "step": 3000
    },
    {
      "epoch": 0.9,
      "eval_f1": 0.9863035374364736,
      "eval_f1_0": 0.9958155810269954,
      "eval_f1_1": 0.9973127199838878,
      "eval_f1_2": 0.9540116494953917,
      "eval_f1_3": 0.9980741992396193,
      "eval_loss": 0.029018014669418335,
      "eval_runtime": 246.9779,
      "eval_samples_per_second": 48.053,
      "eval_steps_per_second": 0.19,
      "step": 3000
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.788920424682014e-05,
      "loss": 0.0409,
      "step": 3010
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.787869231577841e-05,
      "loss": 0.0245,
      "step": 3020
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.7868180384736678e-05,
      "loss": 0.0061,
      "step": 3030
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.7857668453694946e-05,
      "loss": 0.0101,
      "step": 3040
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.7847156522653214e-05,
      "loss": 0.0142,
      "step": 3050
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.783664459161148e-05,
      "loss": 0.0084,
      "step": 3060
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.782613266056975e-05,
      "loss": 0.0186,
      "step": 3070
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.7815620729528014e-05,
      "loss": 0.0165,
      "step": 3080
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.7805108798486282e-05,
      "loss": 0.0126,
      "step": 3090
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.779459686744455e-05,
      "loss": 0.0346,
      "step": 3100
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.778408493640282e-05,
      "loss": 0.0071,
      "step": 3110
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.7773573005361087e-05,
      "loss": 0.0165,
      "step": 3120
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.7763061074319355e-05,
      "loss": 0.039,
      "step": 3130
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.7752549143277623e-05,
      "loss": 0.0127,
      "step": 3140
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.774203721223589e-05,
      "loss": 0.0472,
      "step": 3150
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.773152528119416e-05,
      "loss": 0.0259,
      "step": 3160
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.7721013350152427e-05,
      "loss": 0.0283,
      "step": 3170
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.771050141911069e-05,
      "loss": 0.0251,
      "step": 3180
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.769998948806896e-05,
      "loss": 0.0168,
      "step": 3190
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.7689477557027227e-05,
      "loss": 0.0146,
      "step": 3200
    },
    {
      "epoch": 0.96,
      "eval_f1": 0.9887686718149741,
      "eval_f1_0": 0.9965374547230755,
      "eval_f1_1": 0.9966574062163412,
      "eval_f1_2": 0.9635970629668684,
      "eval_f1_3": 0.998282763353611,
      "eval_loss": 0.025605084374547005,
      "eval_runtime": 247.0678,
      "eval_samples_per_second": 48.035,
      "eval_steps_per_second": 0.19,
      "step": 3200
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.7678965625985495e-05,
      "loss": 0.01,
      "step": 3210
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.7668453694943764e-05,
      "loss": 0.0149,
      "step": 3220
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.765794176390203e-05,
      "loss": 0.0347,
      "step": 3230
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.76474298328603e-05,
      "loss": 0.0318,
      "step": 3240
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.7636917901818568e-05,
      "loss": 0.0413,
      "step": 3250
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.7626405970776836e-05,
      "loss": 0.0128,
      "step": 3260
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.76158940397351e-05,
      "loss": 0.0134,
      "step": 3270
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.760538210869337e-05,
      "loss": 0.0295,
      "step": 3280
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.7594870177651636e-05,
      "loss": 0.0246,
      "step": 3290
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.75843582466099e-05,
      "loss": 0.0205,
      "step": 3300
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.757384631556817e-05,
      "loss": 0.0163,
      "step": 3310
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.7563334384526437e-05,
      "loss": 0.0185,
      "step": 3320
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.7552822453484705e-05,
      "loss": 0.0333,
      "step": 3330
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.7542310522442973e-05,
      "loss": 0.0367,
      "step": 3340
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.753179859140124e-05,
      "loss": 0.0083,
      "step": 3350
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.752128666035951e-05,
      "loss": 0.009,
      "step": 3360
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.7510774729317777e-05,
      "loss": 0.0421,
      "step": 3370
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.7500262798276045e-05,
      "loss": 0.0283,
      "step": 3380
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.7489750867234313e-05,
      "loss": 0.0178,
      "step": 3390
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.7479238936192578e-05,
      "loss": 0.0071,
      "step": 3400
    },
    {
      "epoch": 1.02,
      "eval_f1": 0.98872757142487,
      "eval_f1_0": 0.9965074830138135,
      "eval_f1_1": 0.9957934902537697,
      "eval_f1_2": 0.9646168328644767,
      "eval_f1_3": 0.9979924795674198,
      "eval_loss": 0.024837549775838852,
      "eval_runtime": 247.3264,
      "eval_samples_per_second": 47.985,
      "eval_steps_per_second": 0.19,
      "step": 3400
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.7468727005150846e-05,
      "loss": 0.0113,
      "step": 3410
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.7458215074109114e-05,
      "loss": 0.0224,
      "step": 3420
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.7447703143067382e-05,
      "loss": 0.0171,
      "step": 3430
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.743719121202565e-05,
      "loss": 0.0075,
      "step": 3440
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.7426679280983918e-05,
      "loss": 0.0138,
      "step": 3450
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.7416167349942186e-05,
      "loss": 0.0242,
      "step": 3460
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.7405655418900454e-05,
      "loss": 0.0128,
      "step": 3470
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.7395143487858722e-05,
      "loss": 0.0507,
      "step": 3480
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.738463155681699e-05,
      "loss": 0.0447,
      "step": 3490
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.7374119625775255e-05,
      "loss": 0.0409,
      "step": 3500
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.7363607694733523e-05,
      "loss": 0.0431,
      "step": 3510
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.735309576369179e-05,
      "loss": 0.0105,
      "step": 3520
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.734258383265006e-05,
      "loss": 0.0329,
      "step": 3530
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.7332071901608327e-05,
      "loss": 0.0207,
      "step": 3540
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.7321559970566595e-05,
      "loss": 0.0483,
      "step": 3550
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.7311048039524863e-05,
      "loss": 0.0032,
      "step": 3560
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.730053610848313e-05,
      "loss": 0.0152,
      "step": 3570
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.72900241774414e-05,
      "loss": 0.0144,
      "step": 3580
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.7279512246399664e-05,
      "loss": 0.0122,
      "step": 3590
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.7269000315357932e-05,
      "loss": 0.008,
      "step": 3600
    },
    {
      "epoch": 1.08,
      "eval_f1": 0.9883165335752913,
      "eval_f1_0": 0.996354645144335,
      "eval_f1_1": 0.9947466257172877,
      "eval_f1_2": 0.9639287290379523,
      "eval_f1_3": 0.99823613440159,
      "eval_loss": 0.023318510502576828,
      "eval_runtime": 246.8997,
      "eval_samples_per_second": 48.068,
      "eval_steps_per_second": 0.19,
      "step": 3600
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.72584883843162e-05,
      "loss": 0.0138,
      "step": 3610
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.7247976453274468e-05,
      "loss": 0.0062,
      "step": 3620
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.7237464522232736e-05,
      "loss": 0.005,
      "step": 3630
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.7226952591191004e-05,
      "loss": 0.0393,
      "step": 3640
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.7216440660149272e-05,
      "loss": 0.0264,
      "step": 3650
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.720592872910754e-05,
      "loss": 0.0358,
      "step": 3660
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.719541679806581e-05,
      "loss": 0.0326,
      "step": 3670
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.7184904867024076e-05,
      "loss": 0.0109,
      "step": 3680
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.717439293598234e-05,
      "loss": 0.0054,
      "step": 3690
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.716388100494061e-05,
      "loss": 0.019,
      "step": 3700
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.7153369073898877e-05,
      "loss": 0.0085,
      "step": 3710
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.0081,
      "step": 3720
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.713234521181541e-05,
      "loss": 0.0456,
      "step": 3730
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.7121833280773678e-05,
      "loss": 0.0227,
      "step": 3740
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.7111321349731946e-05,
      "loss": 0.025,
      "step": 3750
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.7100809418690214e-05,
      "loss": 0.0157,
      "step": 3760
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.7090297487648482e-05,
      "loss": 0.0378,
      "step": 3770
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.707978555660675e-05,
      "loss": 0.0048,
      "step": 3780
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.7069273625565018e-05,
      "loss": 0.019,
      "step": 3790
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.7058761694523286e-05,
      "loss": 0.0257,
      "step": 3800
    },
    {
      "epoch": 1.14,
      "eval_f1": 0.9919045681689276,
      "eval_f1_0": 0.9974442132736524,
      "eval_f1_1": 0.996116933895651,
      "eval_f1_2": 0.9756226540586659,
      "eval_f1_3": 0.9984344714477411,
      "eval_loss": 0.017140811309218407,
      "eval_runtime": 246.737,
      "eval_samples_per_second": 48.1,
      "eval_steps_per_second": 0.19,
      "step": 3800
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.704824976348155e-05,
      "loss": 0.0088,
      "step": 3810
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.703773783243982e-05,
      "loss": 0.0395,
      "step": 3820
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.7027225901398087e-05,
      "loss": 0.0256,
      "step": 3830
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.7016713970356355e-05,
      "loss": 0.0053,
      "step": 3840
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.7006202039314623e-05,
      "loss": 0.0429,
      "step": 3850
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.699569010827289e-05,
      "loss": 0.0059,
      "step": 3860
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.698517817723116e-05,
      "loss": 0.0203,
      "step": 3870
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.6974666246189427e-05,
      "loss": 0.0229,
      "step": 3880
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.6964154315147695e-05,
      "loss": 0.0038,
      "step": 3890
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.6953642384105963e-05,
      "loss": 0.0093,
      "step": 3900
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.6943130453064228e-05,
      "loss": 0.0029,
      "step": 3910
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.6932618522022496e-05,
      "loss": 0.0041,
      "step": 3920
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.6922106590980764e-05,
      "loss": 0.0067,
      "step": 3930
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.6911594659939032e-05,
      "loss": 0.0066,
      "step": 3940
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.69010827288973e-05,
      "loss": 0.0031,
      "step": 3950
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.6890570797855568e-05,
      "loss": 0.0115,
      "step": 3960
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.6880058866813836e-05,
      "loss": 0.0226,
      "step": 3970
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.6869546935772104e-05,
      "loss": 0.0107,
      "step": 3980
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.6859035004730372e-05,
      "loss": 0.0226,
      "step": 3990
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.6848523073688637e-05,
      "loss": 0.0458,
      "step": 4000
    },
    {
      "epoch": 1.2,
      "eval_f1": 0.9861112549078215,
      "eval_f1_0": 0.9955435867638419,
      "eval_f1_1": 0.9978893206526428,
      "eval_f1_2": 0.9530572796919834,
      "eval_f1_3": 0.997954832522818,
      "eval_loss": 0.02355680800974369,
      "eval_runtime": 247.1958,
      "eval_samples_per_second": 48.011,
      "eval_steps_per_second": 0.19,
      "step": 4000
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.6838011142646905e-05,
      "loss": 0.0142,
      "step": 4010
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.6827499211605173e-05,
      "loss": 0.0115,
      "step": 4020
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.681698728056344e-05,
      "loss": 0.0331,
      "step": 4030
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.680647534952171e-05,
      "loss": 0.006,
      "step": 4040
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.6795963418479977e-05,
      "loss": 0.0187,
      "step": 4050
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.6785451487438245e-05,
      "loss": 0.0057,
      "step": 4060
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.6774939556396513e-05,
      "loss": 0.0183,
      "step": 4070
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.676442762535478e-05,
      "loss": 0.0176,
      "step": 4080
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.675391569431305e-05,
      "loss": 0.0466,
      "step": 4090
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.6743403763271314e-05,
      "loss": 0.0081,
      "step": 4100
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.6732891832229582e-05,
      "loss": 0.0195,
      "step": 4110
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.672237990118785e-05,
      "loss": 0.0301,
      "step": 4120
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.6711867970146115e-05,
      "loss": 0.0083,
      "step": 4130
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.6701356039104383e-05,
      "loss": 0.03,
      "step": 4140
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.669084410806265e-05,
      "loss": 0.0432,
      "step": 4150
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.668033217702092e-05,
      "loss": 0.0246,
      "step": 4160
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.6669820245979187e-05,
      "loss": 0.0258,
      "step": 4170
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.6659308314937455e-05,
      "loss": 0.0253,
      "step": 4180
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.6648796383895723e-05,
      "loss": 0.0059,
      "step": 4190
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.663828445285399e-05,
      "loss": 0.0108,
      "step": 4200
    },
    {
      "epoch": 1.26,
      "eval_f1": 0.9902302676067312,
      "eval_f1_0": 0.9969721720978607,
      "eval_f1_1": 0.996723948111627,
      "eval_f1_2": 0.9691849123227656,
      "eval_f1_3": 0.9980400378946714,
      "eval_loss": 0.017736753448843956,
      "eval_runtime": 246.9667,
      "eval_samples_per_second": 48.055,
      "eval_steps_per_second": 0.19,
      "step": 4200
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.662777252181226e-05,
      "loss": 0.0026,
      "step": 4210
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.6617260590770527e-05,
      "loss": 0.0179,
      "step": 4220
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.660674865972879e-05,
      "loss": 0.0084,
      "step": 4230
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.659623672868706e-05,
      "loss": 0.0105,
      "step": 4240
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.6585724797645328e-05,
      "loss": 0.0219,
      "step": 4250
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.6575212866603596e-05,
      "loss": 0.0074,
      "step": 4260
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.6564700935561864e-05,
      "loss": 0.0409,
      "step": 4270
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.6554189004520132e-05,
      "loss": 0.0067,
      "step": 4280
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.65436770734784e-05,
      "loss": 0.0046,
      "step": 4290
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.6533165142436668e-05,
      "loss": 0.0044,
      "step": 4300
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.6522653211394936e-05,
      "loss": 0.0068,
      "step": 4310
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.65121412803532e-05,
      "loss": 0.0262,
      "step": 4320
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.650162934931147e-05,
      "loss": 0.0218,
      "step": 4330
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.6491117418269737e-05,
      "loss": 0.0198,
      "step": 4340
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.6480605487228005e-05,
      "loss": 0.0382,
      "step": 4350
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.6470093556186273e-05,
      "loss": 0.0086,
      "step": 4360
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.645958162514454e-05,
      "loss": 0.0118,
      "step": 4370
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.644906969410281e-05,
      "loss": 0.0041,
      "step": 4380
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.6438557763061077e-05,
      "loss": 0.0252,
      "step": 4390
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.6428045832019345e-05,
      "loss": 0.01,
      "step": 4400
    },
    {
      "epoch": 1.32,
      "eval_f1": 0.989207624104371,
      "eval_f1_0": 0.9966190646485159,
      "eval_f1_1": 0.9970713657875646,
      "eval_f1_2": 0.9650929302340525,
      "eval_f1_3": 0.9980471357473509,
      "eval_loss": 0.020230012014508247,
      "eval_runtime": 247.0792,
      "eval_samples_per_second": 48.033,
      "eval_steps_per_second": 0.19,
      "step": 4400
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.6417533900977613e-05,
      "loss": 0.0025,
      "step": 4410
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.6407021969935878e-05,
      "loss": 0.0122,
      "step": 4420
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.6396510038894146e-05,
      "loss": 0.0082,
      "step": 4430
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.6385998107852414e-05,
      "loss": 0.0209,
      "step": 4440
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.637548617681068e-05,
      "loss": 0.0054,
      "step": 4450
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.636497424576895e-05,
      "loss": 0.0037,
      "step": 4460
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.6354462314727218e-05,
      "loss": 0.0135,
      "step": 4470
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.6343950383685486e-05,
      "loss": 0.0102,
      "step": 4480
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.6333438452643754e-05,
      "loss": 0.0038,
      "step": 4490
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.6322926521602022e-05,
      "loss": 0.0077,
      "step": 4500
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.6312414590560287e-05,
      "loss": 0.0376,
      "step": 4510
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.6301902659518555e-05,
      "loss": 0.0165,
      "step": 4520
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.6291390728476823e-05,
      "loss": 0.0035,
      "step": 4530
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.628087879743509e-05,
      "loss": 0.0041,
      "step": 4540
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.6270366866393355e-05,
      "loss": 0.0404,
      "step": 4550
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.6259854935351623e-05,
      "loss": 0.0035,
      "step": 4560
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.624934300430989e-05,
      "loss": 0.0076,
      "step": 4570
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.623883107326816e-05,
      "loss": 0.0251,
      "step": 4580
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.6228319142226427e-05,
      "loss": 0.0051,
      "step": 4590
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.6217807211184696e-05,
      "loss": 0.0028,
      "step": 4600
    },
    {
      "epoch": 1.38,
      "eval_f1": 0.9833248306597419,
      "eval_f1_0": 0.9949951970645448,
      "eval_f1_1": 0.9960817942205029,
      "eval_f1_2": 0.9438158108235053,
      "eval_f1_3": 0.9984065205304143,
      "eval_loss": 0.03907902538776398,
      "eval_runtime": 247.1623,
      "eval_samples_per_second": 48.017,
      "eval_steps_per_second": 0.19,
      "step": 4600
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.6207295280142964e-05,
      "loss": 0.0464,
      "step": 4610
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.619678334910123e-05,
      "loss": 0.0058,
      "step": 4620
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.61862714180595e-05,
      "loss": 0.0033,
      "step": 4630
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.6175759487017764e-05,
      "loss": 0.0038,
      "step": 4640
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.6165247555976032e-05,
      "loss": 0.0147,
      "step": 4650
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.61547356249343e-05,
      "loss": 0.0094,
      "step": 4660
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.614422369389257e-05,
      "loss": 0.0061,
      "step": 4670
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.6133711762850836e-05,
      "loss": 0.0233,
      "step": 4680
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.6123199831809104e-05,
      "loss": 0.0091,
      "step": 4690
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.6112687900767373e-05,
      "loss": 0.025,
      "step": 4700
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.610217596972564e-05,
      "loss": 0.0066,
      "step": 4710
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.609166403868391e-05,
      "loss": 0.0066,
      "step": 4720
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.6081152107642177e-05,
      "loss": 0.0338,
      "step": 4730
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.607064017660044e-05,
      "loss": 0.023,
      "step": 4740
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.606012824555871e-05,
      "loss": 0.0072,
      "step": 4750
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.6049616314516977e-05,
      "loss": 0.0268,
      "step": 4760
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.6039104383475245e-05,
      "loss": 0.0052,
      "step": 4770
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.6028592452433513e-05,
      "loss": 0.0161,
      "step": 4780
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.601808052139178e-05,
      "loss": 0.0203,
      "step": 4790
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.600756859035005e-05,
      "loss": 0.0292,
      "step": 4800
    },
    {
      "epoch": 1.44,
      "eval_f1": 0.9910608583904451,
      "eval_f1_0": 0.997222323619616,
      "eval_f1_1": 0.9967902448014252,
      "eval_f1_2": 0.9717467226252857,
      "eval_f1_3": 0.9984841425154537,
      "eval_loss": 0.01618151366710663,
      "eval_runtime": 247.2094,
      "eval_samples_per_second": 48.008,
      "eval_steps_per_second": 0.19,
      "step": 4800
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.5997056659308318e-05,
      "loss": 0.0251,
      "step": 4810
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.5986544728266586e-05,
      "loss": 0.0069,
      "step": 4820
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.597603279722485e-05,
      "loss": 0.0116,
      "step": 4830
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.596552086618312e-05,
      "loss": 0.0171,
      "step": 4840
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.5955008935141386e-05,
      "loss": 0.0148,
      "step": 4850
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.5944497004099654e-05,
      "loss": 0.0069,
      "step": 4860
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.5933985073057922e-05,
      "loss": 0.0242,
      "step": 4870
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.592347314201619e-05,
      "loss": 0.0043,
      "step": 4880
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.591296121097446e-05,
      "loss": 0.0308,
      "step": 4890
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.5902449279932727e-05,
      "loss": 0.0193,
      "step": 4900
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.5891937348890995e-05,
      "loss": 0.0132,
      "step": 4910
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.5881425417849263e-05,
      "loss": 0.0042,
      "step": 4920
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.5870913486807527e-05,
      "loss": 0.0146,
      "step": 4930
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.5860401555765795e-05,
      "loss": 0.0055,
      "step": 4940
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.5849889624724063e-05,
      "loss": 0.006,
      "step": 4950
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.583937769368233e-05,
      "loss": 0.0066,
      "step": 4960
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.5828865762640596e-05,
      "loss": 0.0029,
      "step": 4970
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.5818353831598864e-05,
      "loss": 0.0216,
      "step": 4980
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.5807841900557132e-05,
      "loss": 0.0099,
      "step": 4990
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.57973299695154e-05,
      "loss": 0.009,
      "step": 5000
    },
    {
      "epoch": 1.5,
      "eval_f1": 0.9898035232194219,
      "eval_f1_0": 0.9969628882556604,
      "eval_f1_1": 0.9971197803209279,
      "eval_f1_2": 0.9678906964498517,
      "eval_f1_3": 0.9972407278512478,
      "eval_loss": 0.02300242707133293,
      "eval_runtime": 247.1,
      "eval_samples_per_second": 48.029,
      "eval_steps_per_second": 0.19,
      "step": 5000
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.5786818038473668e-05,
      "loss": 0.0162,
      "step": 5010
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.5776306107431936e-05,
      "loss": 0.0047,
      "step": 5020
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.5765794176390204e-05,
      "loss": 0.0053,
      "step": 5030
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.5755282245348472e-05,
      "loss": 0.0034,
      "step": 5040
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.574477031430674e-05,
      "loss": 0.0084,
      "step": 5050
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.5734258383265005e-05,
      "loss": 0.0083,
      "step": 5060
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.5723746452223273e-05,
      "loss": 0.0052,
      "step": 5070
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.571323452118154e-05,
      "loss": 0.0248,
      "step": 5080
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.570272259013981e-05,
      "loss": 0.0129,
      "step": 5090
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.5692210659098077e-05,
      "loss": 0.0068,
      "step": 5100
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.5681698728056345e-05,
      "loss": 0.0077,
      "step": 5110
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.5671186797014613e-05,
      "loss": 0.0031,
      "step": 5120
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.566067486597288e-05,
      "loss": 0.0042,
      "step": 5130
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.565016293493115e-05,
      "loss": 0.0181,
      "step": 5140
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.5639651003889414e-05,
      "loss": 0.0218,
      "step": 5150
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.5629139072847682e-05,
      "loss": 0.0068,
      "step": 5160
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.561862714180595e-05,
      "loss": 0.006,
      "step": 5170
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.5608115210764218e-05,
      "loss": 0.0048,
      "step": 5180
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.5597603279722486e-05,
      "loss": 0.0133,
      "step": 5190
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.5587091348680754e-05,
      "loss": 0.0224,
      "step": 5200
    },
    {
      "epoch": 1.56,
      "eval_f1": 0.9908468465140768,
      "eval_f1_0": 0.9971399721149006,
      "eval_f1_1": 0.9976318935654946,
      "eval_f1_2": 0.9709939000969645,
      "eval_f1_3": 0.9976216202789473,
      "eval_loss": 0.018998125568032265,
      "eval_runtime": 246.9675,
      "eval_samples_per_second": 48.055,
      "eval_steps_per_second": 0.19,
      "step": 5200
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.5576579417639022e-05,
      "loss": 0.0093,
      "step": 5210
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.556606748659729e-05,
      "loss": 0.0021,
      "step": 5220
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.0058,
      "step": 5230
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.5545043624513826e-05,
      "loss": 0.0207,
      "step": 5240
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.553453169347209e-05,
      "loss": 0.0321,
      "step": 5250
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.552401976243036e-05,
      "loss": 0.006,
      "step": 5260
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.5513507831388627e-05,
      "loss": 0.0172,
      "step": 5270
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.5502995900346895e-05,
      "loss": 0.0083,
      "step": 5280
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.5492483969305163e-05,
      "loss": 0.0269,
      "step": 5290
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.548197203826343e-05,
      "loss": 0.0172,
      "step": 5300
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.54714601072217e-05,
      "loss": 0.0158,
      "step": 5310
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.5460948176179967e-05,
      "loss": 0.0242,
      "step": 5320
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.5450436245138235e-05,
      "loss": 0.0075,
      "step": 5330
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.54399243140965e-05,
      "loss": 0.0048,
      "step": 5340
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.5429412383054768e-05,
      "loss": 0.003,
      "step": 5350
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.5418900452013036e-05,
      "loss": 0.0459,
      "step": 5360
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.5408388520971304e-05,
      "loss": 0.019,
      "step": 5370
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.5397876589929572e-05,
      "loss": 0.0049,
      "step": 5380
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.538736465888784e-05,
      "loss": 0.0332,
      "step": 5390
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.5376852727846105e-05,
      "loss": 0.0027,
      "step": 5400
    },
    {
      "epoch": 1.62,
      "eval_f1": 0.9923619033762792,
      "eval_f1_0": 0.9976219742618546,
      "eval_f1_1": 0.9973801138163009,
      "eval_f1_2": 0.9758560152459584,
      "eval_f1_3": 0.9985895101810025,
      "eval_loss": 0.015381346456706524,
      "eval_runtime": 247.1283,
      "eval_samples_per_second": 48.024,
      "eval_steps_per_second": 0.19,
      "step": 5400
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.5366340796804373e-05,
      "loss": 0.0114,
      "step": 5410
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.535582886576264e-05,
      "loss": 0.0068,
      "step": 5420
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.534531693472091e-05,
      "loss": 0.0184,
      "step": 5430
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.5334805003679177e-05,
      "loss": 0.0148,
      "step": 5440
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.5324293072637445e-05,
      "loss": 0.0055,
      "step": 5450
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.5313781141595713e-05,
      "loss": 0.0073,
      "step": 5460
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.5303269210553978e-05,
      "loss": 0.0065,
      "step": 5470
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.5292757279512246e-05,
      "loss": 0.0206,
      "step": 5480
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.5282245348470514e-05,
      "loss": 0.0048,
      "step": 5490
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.5271733417428782e-05,
      "loss": 0.0354,
      "step": 5500
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.526122148638705e-05,
      "loss": 0.0052,
      "step": 5510
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.5250709555345318e-05,
      "loss": 0.0036,
      "step": 5520
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.5240197624303586e-05,
      "loss": 0.0543,
      "step": 5530
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.5229685693261852e-05,
      "loss": 0.0301,
      "step": 5540
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.521917376222012e-05,
      "loss": 0.0166,
      "step": 5550
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.5208661831178388e-05,
      "loss": 0.0035,
      "step": 5560
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.5198149900136656e-05,
      "loss": 0.0332,
      "step": 5570
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.5187637969094924e-05,
      "loss": 0.0155,
      "step": 5580
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.517712603805319e-05,
      "loss": 0.0358,
      "step": 5590
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.5166614107011459e-05,
      "loss": 0.0094,
      "step": 5600
    },
    {
      "epoch": 1.68,
      "eval_f1": 0.993332322133445,
      "eval_f1_0": 0.9978957647265436,
      "eval_f1_1": 0.9969867891155678,
      "eval_f1_2": 0.9798261907385952,
      "eval_f1_3": 0.9986205439530736,
      "eval_loss": 0.014064018614590168,
      "eval_runtime": 247.0411,
      "eval_samples_per_second": 48.041,
      "eval_steps_per_second": 0.19,
      "step": 5600
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.5156102175969727e-05,
      "loss": 0.0449,
      "step": 5610
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.5145590244927995e-05,
      "loss": 0.005,
      "step": 5620
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.5135078313886263e-05,
      "loss": 0.0345,
      "step": 5630
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.512456638284453e-05,
      "loss": 0.0132,
      "step": 5640
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.5114054451802797e-05,
      "loss": 0.018,
      "step": 5650
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.5103542520761065e-05,
      "loss": 0.0161,
      "step": 5660
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.5093030589719333e-05,
      "loss": 0.0174,
      "step": 5670
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.5082518658677602e-05,
      "loss": 0.0185,
      "step": 5680
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.5072006727635868e-05,
      "loss": 0.031,
      "step": 5690
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.5061494796594136e-05,
      "loss": 0.035,
      "step": 5700
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.5050982865552404e-05,
      "loss": 0.0054,
      "step": 5710
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.5040470934510672e-05,
      "loss": 0.0137,
      "step": 5720
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.5029959003468938e-05,
      "loss": 0.0291,
      "step": 5730
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.5019447072427206e-05,
      "loss": 0.0115,
      "step": 5740
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.5008935141385474e-05,
      "loss": 0.0246,
      "step": 5750
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.4998423210343742e-05,
      "loss": 0.0145,
      "step": 5760
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.498791127930201e-05,
      "loss": 0.017,
      "step": 5770
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.4977399348260277e-05,
      "loss": 0.0149,
      "step": 5780
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.4966887417218545e-05,
      "loss": 0.0063,
      "step": 5790
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.4956375486176813e-05,
      "loss": 0.011,
      "step": 5800
    },
    {
      "epoch": 1.74,
      "eval_f1": 0.9904861636946208,
      "eval_f1_0": 0.9970166636308943,
      "eval_f1_1": 0.9972486731937233,
      "eval_f1_2": 0.969550603359083,
      "eval_f1_3": 0.9981287145947828,
      "eval_loss": 0.018452756106853485,
      "eval_runtime": 246.9711,
      "eval_samples_per_second": 48.054,
      "eval_steps_per_second": 0.19,
      "step": 5800
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.4945863555135081e-05,
      "loss": 0.0092,
      "step": 5810
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.4935351624093349e-05,
      "loss": 0.0041,
      "step": 5820
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.4924839693051614e-05,
      "loss": 0.0162,
      "step": 5830
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.4914327762009882e-05,
      "loss": 0.0065,
      "step": 5840
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.490381583096815e-05,
      "loss": 0.04,
      "step": 5850
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.4893303899926416e-05,
      "loss": 0.0325,
      "step": 5860
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.4882791968884684e-05,
      "loss": 0.0078,
      "step": 5870
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.4872280037842952e-05,
      "loss": 0.0305,
      "step": 5880
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.486176810680122e-05,
      "loss": 0.0084,
      "step": 5890
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.4851256175759488e-05,
      "loss": 0.0124,
      "step": 5900
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.4840744244717755e-05,
      "loss": 0.0068,
      "step": 5910
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.4830232313676023e-05,
      "loss": 0.0163,
      "step": 5920
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.481972038263429e-05,
      "loss": 0.006,
      "step": 5930
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.4809208451592559e-05,
      "loss": 0.0266,
      "step": 5940
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.4798696520550827e-05,
      "loss": 0.0104,
      "step": 5950
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.4788184589509093e-05,
      "loss": 0.0243,
      "step": 5960
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.4777672658467361e-05,
      "loss": 0.034,
      "step": 5970
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.476716072742563e-05,
      "loss": 0.003,
      "step": 5980
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.4756648796383897e-05,
      "loss": 0.0214,
      "step": 5990
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.4746136865342164e-05,
      "loss": 0.0145,
      "step": 6000
    },
    {
      "epoch": 1.8,
      "eval_f1": 0.984540798405043,
      "eval_f1_0": 0.9953676397859114,
      "eval_f1_1": 0.9976339047086255,
      "eval_f1_2": 0.9466528232388529,
      "eval_f1_3": 0.9985088258867826,
      "eval_loss": 0.02828732319176197,
      "eval_runtime": 247.1278,
      "eval_samples_per_second": 48.024,
      "eval_steps_per_second": 0.19,
      "step": 6000
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.4735624934300432e-05,
      "loss": 0.0092,
      "step": 6010
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.47251130032587e-05,
      "loss": 0.0508,
      "step": 6020
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.4714601072216968e-05,
      "loss": 0.0356,
      "step": 6030
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.4704089141175236e-05,
      "loss": 0.0076,
      "step": 6040
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.4693577210133502e-05,
      "loss": 0.0094,
      "step": 6050
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.468306527909177e-05,
      "loss": 0.0234,
      "step": 6060
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.4672553348050038e-05,
      "loss": 0.0063,
      "step": 6070
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.4662041417008306e-05,
      "loss": 0.0338,
      "step": 6080
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.4651529485966574e-05,
      "loss": 0.0038,
      "step": 6090
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.464101755492484e-05,
      "loss": 0.006,
      "step": 6100
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.4630505623883109e-05,
      "loss": 0.007,
      "step": 6110
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.4619993692841377e-05,
      "loss": 0.0055,
      "step": 6120
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.4609481761799645e-05,
      "loss": 0.0027,
      "step": 6130
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.4598969830757913e-05,
      "loss": 0.0024,
      "step": 6140
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.4588457899716179e-05,
      "loss": 0.013,
      "step": 6150
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.4577945968674447e-05,
      "loss": 0.0177,
      "step": 6160
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.4567434037632715e-05,
      "loss": 0.0446,
      "step": 6170
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.4556922106590983e-05,
      "loss": 0.0145,
      "step": 6180
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.454641017554925e-05,
      "loss": 0.039,
      "step": 6190
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.4535898244507518e-05,
      "loss": 0.0074,
      "step": 6200
    },
    {
      "epoch": 1.86,
      "eval_f1": 0.9923867203617436,
      "eval_f1_0": 0.9976212858022695,
      "eval_f1_1": 0.9973340601170403,
      "eval_f1_2": 0.9761297723664167,
      "eval_f1_3": 0.9984617631612476,
      "eval_loss": 0.013705220073461533,
      "eval_runtime": 247.2873,
      "eval_samples_per_second": 47.993,
      "eval_steps_per_second": 0.19,
      "step": 6200
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.4525386313465786e-05,
      "loss": 0.0104,
      "step": 6210
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.4514874382424054e-05,
      "loss": 0.0284,
      "step": 6220
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.4504362451382322e-05,
      "loss": 0.0117,
      "step": 6230
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.4493850520340588e-05,
      "loss": 0.0063,
      "step": 6240
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.4483338589298854e-05,
      "loss": 0.0041,
      "step": 6250
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.4472826658257122e-05,
      "loss": 0.0028,
      "step": 6260
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.446231472721539e-05,
      "loss": 0.0027,
      "step": 6270
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.4451802796173657e-05,
      "loss": 0.0036,
      "step": 6280
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.4441290865131925e-05,
      "loss": 0.0018,
      "step": 6290
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.4430778934090193e-05,
      "loss": 0.0362,
      "step": 6300
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.4420267003048461e-05,
      "loss": 0.0097,
      "step": 6310
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.4409755072006727e-05,
      "loss": 0.0129,
      "step": 6320
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.4399243140964995e-05,
      "loss": 0.0035,
      "step": 6330
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.4388731209923263e-05,
      "loss": 0.0022,
      "step": 6340
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.4378219278881531e-05,
      "loss": 0.0039,
      "step": 6350
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.43677073478398e-05,
      "loss": 0.0021,
      "step": 6360
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.4357195416798066e-05,
      "loss": 0.0037,
      "step": 6370
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.4346683485756334e-05,
      "loss": 0.0028,
      "step": 6380
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.4336171554714602e-05,
      "loss": 0.0221,
      "step": 6390
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.432565962367287e-05,
      "loss": 0.0307,
      "step": 6400
    },
    {
      "epoch": 1.92,
      "eval_f1": 0.9936541817880021,
      "eval_f1_0": 0.9980114982083728,
      "eval_f1_1": 0.9968411863355464,
      "eval_f1_2": 0.9811366963639431,
      "eval_f1_3": 0.9986273462441461,
      "eval_loss": 0.01458107028156519,
      "eval_runtime": 246.7881,
      "eval_samples_per_second": 48.09,
      "eval_steps_per_second": 0.19,
      "step": 6400
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.4315147692631138e-05,
      "loss": 0.0189,
      "step": 6410
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.4304635761589404e-05,
      "loss": 0.0691,
      "step": 6420
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.4294123830547672e-05,
      "loss": 0.0283,
      "step": 6430
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.428361189950594e-05,
      "loss": 0.0355,
      "step": 6440
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.4273099968464208e-05,
      "loss": 0.0081,
      "step": 6450
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.4262588037422476e-05,
      "loss": 0.0103,
      "step": 6460
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.4252076106380743e-05,
      "loss": 0.0136,
      "step": 6470
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.4241564175339011e-05,
      "loss": 0.0168,
      "step": 6480
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.4231052244297279e-05,
      "loss": 0.0043,
      "step": 6490
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.4220540313255547e-05,
      "loss": 0.0066,
      "step": 6500
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.4210028382213813e-05,
      "loss": 0.0235,
      "step": 6510
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.4199516451172081e-05,
      "loss": 0.0359,
      "step": 6520
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.418900452013035e-05,
      "loss": 0.0246,
      "step": 6530
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.4178492589088617e-05,
      "loss": 0.0078,
      "step": 6540
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.4167980658046885e-05,
      "loss": 0.0198,
      "step": 6550
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.4157468727005152e-05,
      "loss": 0.004,
      "step": 6560
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.414695679596342e-05,
      "loss": 0.0062,
      "step": 6570
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.4136444864921688e-05,
      "loss": 0.0053,
      "step": 6580
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.4125932933879956e-05,
      "loss": 0.0049,
      "step": 6590
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.4115421002838224e-05,
      "loss": 0.0245,
      "step": 6600
    },
    {
      "epoch": 1.98,
      "eval_f1": 0.993306313654685,
      "eval_f1_0": 0.9979105907534561,
      "eval_f1_1": 0.9976182166157198,
      "eval_f1_2": 0.9789854232435508,
      "eval_f1_3": 0.9987110240060132,
      "eval_loss": 0.015001187101006508,
      "eval_runtime": 247.0583,
      "eval_samples_per_second": 48.037,
      "eval_steps_per_second": 0.19,
      "step": 6600
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.410490907179649e-05,
      "loss": 0.0106,
      "step": 6610
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.4094397140754758e-05,
      "loss": 0.0181,
      "step": 6620
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.4083885209713026e-05,
      "loss": 0.0053,
      "step": 6630
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.4073373278671294e-05,
      "loss": 0.0032,
      "step": 6640
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.4062861347629562e-05,
      "loss": 0.0021,
      "step": 6650
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.4052349416587829e-05,
      "loss": 0.0156,
      "step": 6660
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.4041837485546097e-05,
      "loss": 0.0345,
      "step": 6670
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.4031325554504363e-05,
      "loss": 0.0016,
      "step": 6680
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.402081362346263e-05,
      "loss": 0.012,
      "step": 6690
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.4010301692420898e-05,
      "loss": 0.0433,
      "step": 6700
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.3999789761379166e-05,
      "loss": 0.0016,
      "step": 6710
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.3989277830337434e-05,
      "loss": 0.0027,
      "step": 6720
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.3978765899295702e-05,
      "loss": 0.0029,
      "step": 6730
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.3968253968253968e-05,
      "loss": 0.002,
      "step": 6740
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.3957742037212236e-05,
      "loss": 0.0018,
      "step": 6750
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.3947230106170504e-05,
      "loss": 0.0113,
      "step": 6760
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.3936718175128772e-05,
      "loss": 0.0039,
      "step": 6770
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.3926206244087039e-05,
      "loss": 0.0219,
      "step": 6780
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.3915694313045307e-05,
      "loss": 0.0021,
      "step": 6790
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.3905182382003575e-05,
      "loss": 0.0077,
      "step": 6800
    },
    {
      "epoch": 2.04,
      "eval_f1": 0.992785186587577,
      "eval_f1_0": 0.9977657438097319,
      "eval_f1_1": 0.9976483947320973,
      "eval_f1_2": 0.9770061029668933,
      "eval_f1_3": 0.9987205048415848,
      "eval_loss": 0.014934339560568333,
      "eval_runtime": 246.9475,
      "eval_samples_per_second": 48.059,
      "eval_steps_per_second": 0.19,
      "step": 6800
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.3894670450961843e-05,
      "loss": 0.0026,
      "step": 6810
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.388415851992011e-05,
      "loss": 0.016,
      "step": 6820
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.3873646588878377e-05,
      "loss": 0.0022,
      "step": 6830
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.3863134657836645e-05,
      "loss": 0.0016,
      "step": 6840
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.3852622726794913e-05,
      "loss": 0.0229,
      "step": 6850
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.3842110795753181e-05,
      "loss": 0.0024,
      "step": 6860
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.383159886471145e-05,
      "loss": 0.0023,
      "step": 6870
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.3821086933669716e-05,
      "loss": 0.0058,
      "step": 6880
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.3810575002627984e-05,
      "loss": 0.0023,
      "step": 6890
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.3800063071586252e-05,
      "loss": 0.0011,
      "step": 6900
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.378955114054452e-05,
      "loss": 0.0023,
      "step": 6910
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.3779039209502788e-05,
      "loss": 0.0052,
      "step": 6920
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.3768527278461054e-05,
      "loss": 0.0078,
      "step": 6930
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.3758015347419322e-05,
      "loss": 0.0027,
      "step": 6940
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.374750341637759e-05,
      "loss": 0.0019,
      "step": 6950
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.3736991485335858e-05,
      "loss": 0.0042,
      "step": 6960
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.3726479554294126e-05,
      "loss": 0.002,
      "step": 6970
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.3715967623252393e-05,
      "loss": 0.0014,
      "step": 6980
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.370545569221066e-05,
      "loss": 0.0017,
      "step": 6990
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.3694943761168929e-05,
      "loss": 0.0154,
      "step": 7000
    },
    {
      "epoch": 2.1,
      "eval_f1": 0.9915422349901406,
      "eval_f1_0": 0.9973726433738388,
      "eval_f1_1": 0.9978057693892374,
      "eval_f1_2": 0.9730353773710962,
      "eval_f1_3": 0.9979551498263894,
      "eval_loss": 0.01724139414727688,
      "eval_runtime": 247.21,
      "eval_samples_per_second": 48.008,
      "eval_steps_per_second": 0.19,
      "step": 7000
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.3684431830127197e-05,
      "loss": 0.0186,
      "step": 7010
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.3673919899085463e-05,
      "loss": 0.0109,
      "step": 7020
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.3663407968043731e-05,
      "loss": 0.0025,
      "step": 7030
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.3652896037001999e-05,
      "loss": 0.0012,
      "step": 7040
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.3642384105960267e-05,
      "loss": 0.0021,
      "step": 7050
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.3631872174918535e-05,
      "loss": 0.0076,
      "step": 7060
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.3621360243876802e-05,
      "loss": 0.0019,
      "step": 7070
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.361084831283507e-05,
      "loss": 0.0115,
      "step": 7080
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.3600336381793338e-05,
      "loss": 0.0016,
      "step": 7090
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.3589824450751606e-05,
      "loss": 0.0021,
      "step": 7100
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.357931251970987e-05,
      "loss": 0.0084,
      "step": 7110
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.3568800588668138e-05,
      "loss": 0.0015,
      "step": 7120
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.3558288657626406e-05,
      "loss": 0.0028,
      "step": 7130
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.3547776726584674e-05,
      "loss": 0.0016,
      "step": 7140
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.353726479554294e-05,
      "loss": 0.0009,
      "step": 7150
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.3526752864501209e-05,
      "loss": 0.0017,
      "step": 7160
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.3516240933459477e-05,
      "loss": 0.0108,
      "step": 7170
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.3505729002417745e-05,
      "loss": 0.0013,
      "step": 7180
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.3495217071376013e-05,
      "loss": 0.0152,
      "step": 7190
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.348470514033428e-05,
      "loss": 0.011,
      "step": 7200
    },
    {
      "epoch": 2.16,
      "eval_f1": 0.9883555230503425,
      "eval_f1_0": 0.9962823960164975,
      "eval_f1_1": 0.9975579344147374,
      "eval_f1_2": 0.9609607551050002,
      "eval_f1_3": 0.9986210066651344,
      "eval_loss": 0.0203411802649498,
      "eval_runtime": 247.0594,
      "eval_samples_per_second": 48.037,
      "eval_steps_per_second": 0.19,
      "step": 7200
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.3474193209292547e-05,
      "loss": 0.0042,
      "step": 7210
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.3463681278250815e-05,
      "loss": 0.0216,
      "step": 7220
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3453169347209083e-05,
      "loss": 0.0211,
      "step": 7230
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3442657416167351e-05,
      "loss": 0.0017,
      "step": 7240
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3432145485125618e-05,
      "loss": 0.0159,
      "step": 7250
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3421633554083886e-05,
      "loss": 0.0197,
      "step": 7260
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.3411121623042154e-05,
      "loss": 0.0032,
      "step": 7270
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.3400609692000422e-05,
      "loss": 0.0037,
      "step": 7280
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.3390097760958688e-05,
      "loss": 0.004,
      "step": 7290
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3379585829916956e-05,
      "loss": 0.0224,
      "step": 7300
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3369073898875224e-05,
      "loss": 0.0053,
      "step": 7310
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3358561967833492e-05,
      "loss": 0.0024,
      "step": 7320
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.334805003679176e-05,
      "loss": 0.0094,
      "step": 7330
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3337538105750027e-05,
      "loss": 0.0168,
      "step": 7340
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3327026174708295e-05,
      "loss": 0.0027,
      "step": 7350
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3316514243666563e-05,
      "loss": 0.0022,
      "step": 7360
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3306002312624831e-05,
      "loss": 0.021,
      "step": 7370
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3295490381583099e-05,
      "loss": 0.0019,
      "step": 7380
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3284978450541365e-05,
      "loss": 0.0111,
      "step": 7390
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3274466519499633e-05,
      "loss": 0.0043,
      "step": 7400
    },
    {
      "epoch": 2.22,
      "eval_f1": 0.9932852814650001,
      "eval_f1_0": 0.997923309472833,
      "eval_f1_1": 0.9977863003616668,
      "eval_f1_2": 0.9796786093916088,
      "eval_f1_3": 0.9977529066338914,
      "eval_loss": 0.013370517641305923,
      "eval_runtime": 247.295,
      "eval_samples_per_second": 47.991,
      "eval_steps_per_second": 0.19,
      "step": 7400
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3263954588457901e-05,
      "loss": 0.0034,
      "step": 7410
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.325344265741617e-05,
      "loss": 0.0019,
      "step": 7420
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.3242930726374437e-05,
      "loss": 0.0017,
      "step": 7430
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.3232418795332704e-05,
      "loss": 0.0054,
      "step": 7440
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.3221906864290972e-05,
      "loss": 0.006,
      "step": 7450
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.321139493324924e-05,
      "loss": 0.0021,
      "step": 7460
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.3200883002207508e-05,
      "loss": 0.0323,
      "step": 7470
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.3190371071165776e-05,
      "loss": 0.0061,
      "step": 7480
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.3179859140124042e-05,
      "loss": 0.0026,
      "step": 7490
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.316934720908231e-05,
      "loss": 0.002,
      "step": 7500
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.3158835278040578e-05,
      "loss": 0.0013,
      "step": 7510
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.3148323346998846e-05,
      "loss": 0.0028,
      "step": 7520
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.3137811415957111e-05,
      "loss": 0.0216,
      "step": 7530
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.3127299484915379e-05,
      "loss": 0.0024,
      "step": 7540
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.3116787553873647e-05,
      "loss": 0.0215,
      "step": 7550
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.3106275622831915e-05,
      "loss": 0.0013,
      "step": 7560
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.3095763691790182e-05,
      "loss": 0.03,
      "step": 7570
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.308525176074845e-05,
      "loss": 0.028,
      "step": 7580
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.3074739829706718e-05,
      "loss": 0.0036,
      "step": 7590
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.3064227898664986e-05,
      "loss": 0.0113,
      "step": 7600
    },
    {
      "epoch": 2.28,
      "eval_f1": 0.9932981324686819,
      "eval_f1_0": 0.9979256203451299,
      "eval_f1_1": 0.9976930113393383,
      "eval_f1_2": 0.9794637997850386,
      "eval_f1_3": 0.998110098405221,
      "eval_loss": 0.014016643166542053,
      "eval_runtime": 247.3903,
      "eval_samples_per_second": 47.973,
      "eval_steps_per_second": 0.19,
      "step": 7600
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.3053715967623252e-05,
      "loss": 0.0019,
      "step": 7610
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.304320403658152e-05,
      "loss": 0.0019,
      "step": 7620
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.3032692105539788e-05,
      "loss": 0.0133,
      "step": 7630
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.3022180174498056e-05,
      "loss": 0.0219,
      "step": 7640
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.3011668243456324e-05,
      "loss": 0.0255,
      "step": 7650
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.300115631241459e-05,
      "loss": 0.0032,
      "step": 7660
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.2990644381372859e-05,
      "loss": 0.0078,
      "step": 7670
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.2980132450331127e-05,
      "loss": 0.0085,
      "step": 7680
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.2969620519289395e-05,
      "loss": 0.0177,
      "step": 7690
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.2959108588247663e-05,
      "loss": 0.0055,
      "step": 7700
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.2948596657205929e-05,
      "loss": 0.0143,
      "step": 7710
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.2938084726164197e-05,
      "loss": 0.0163,
      "step": 7720
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.2927572795122465e-05,
      "loss": 0.0111,
      "step": 7730
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.2917060864080733e-05,
      "loss": 0.0044,
      "step": 7740
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.2906548933039001e-05,
      "loss": 0.0033,
      "step": 7750
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.2896037001997268e-05,
      "loss": 0.0027,
      "step": 7760
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.2885525070955536e-05,
      "loss": 0.0032,
      "step": 7770
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.2875013139913804e-05,
      "loss": 0.0026,
      "step": 7780
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.2864501208872072e-05,
      "loss": 0.002,
      "step": 7790
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.2853989277830338e-05,
      "loss": 0.0075,
      "step": 7800
    },
    {
      "epoch": 2.34,
      "eval_f1": 0.9928924145037707,
      "eval_f1_0": 0.9977736145869261,
      "eval_f1_1": 0.9973392205797658,
      "eval_f1_2": 0.9783996431033783,
      "eval_f1_3": 0.9980571797450126,
      "eval_loss": 0.015785295516252518,
      "eval_runtime": 247.1483,
      "eval_samples_per_second": 48.02,
      "eval_steps_per_second": 0.19,
      "step": 7800
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.2843477346788606e-05,
      "loss": 0.0017,
      "step": 7810
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.2832965415746874e-05,
      "loss": 0.0056,
      "step": 7820
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.2822453484705142e-05,
      "loss": 0.0068,
      "step": 7830
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.281194155366341e-05,
      "loss": 0.0227,
      "step": 7840
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.2801429622621676e-05,
      "loss": 0.0119,
      "step": 7850
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.2790917691579945e-05,
      "loss": 0.0022,
      "step": 7860
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.2780405760538213e-05,
      "loss": 0.0024,
      "step": 7870
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.276989382949648e-05,
      "loss": 0.0016,
      "step": 7880
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.2759381898454749e-05,
      "loss": 0.0248,
      "step": 7890
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.2748869967413015e-05,
      "loss": 0.0009,
      "step": 7900
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.2738358036371283e-05,
      "loss": 0.004,
      "step": 7910
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.2727846105329551e-05,
      "loss": 0.0038,
      "step": 7920
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.2717334174287819e-05,
      "loss": 0.0182,
      "step": 7930
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.2706822243246087e-05,
      "loss": 0.0056,
      "step": 7940
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.2696310312204354e-05,
      "loss": 0.0123,
      "step": 7950
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.268579838116262e-05,
      "loss": 0.0027,
      "step": 7960
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.2675286450120888e-05,
      "loss": 0.0047,
      "step": 7970
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.2664774519079154e-05,
      "loss": 0.0187,
      "step": 7980
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.2654262588037422e-05,
      "loss": 0.0034,
      "step": 7990
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.264375065699569e-05,
      "loss": 0.0109,
      "step": 8000
    },
    {
      "epoch": 2.4,
      "eval_f1": 0.9912616191930683,
      "eval_f1_0": 0.997229491242211,
      "eval_f1_1": 0.997458262756164,
      "eval_f1_2": 0.9719331053643941,
      "eval_f1_3": 0.9984256174095041,
      "eval_loss": 0.02033372037112713,
      "eval_runtime": 247.2179,
      "eval_samples_per_second": 48.006,
      "eval_steps_per_second": 0.19,
      "step": 8000
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.2633238725953958e-05,
      "loss": 0.002,
      "step": 8010
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.2622726794912226e-05,
      "loss": 0.0021,
      "step": 8020
    },
    {
      "epoch": 2.41,
      "learning_rate": 1.2612214863870493e-05,
      "loss": 0.0041,
      "step": 8030
    },
    {
      "epoch": 2.41,
      "learning_rate": 1.260170293282876e-05,
      "loss": 0.0014,
      "step": 8040
    },
    {
      "epoch": 2.41,
      "learning_rate": 1.2591191001787029e-05,
      "loss": 0.001,
      "step": 8050
    },
    {
      "epoch": 2.41,
      "learning_rate": 1.2580679070745297e-05,
      "loss": 0.001,
      "step": 8060
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.2570167139703565e-05,
      "loss": 0.0012,
      "step": 8070
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.2559655208661831e-05,
      "loss": 0.0028,
      "step": 8080
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.25491432776201e-05,
      "loss": 0.0011,
      "step": 8090
    },
    {
      "epoch": 2.43,
      "learning_rate": 1.2538631346578367e-05,
      "loss": 0.0016,
      "step": 8100
    },
    {
      "epoch": 2.43,
      "learning_rate": 1.2528119415536635e-05,
      "loss": 0.0083,
      "step": 8110
    },
    {
      "epoch": 2.43,
      "learning_rate": 1.2517607484494902e-05,
      "loss": 0.0019,
      "step": 8120
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.250709555345317e-05,
      "loss": 0.0217,
      "step": 8130
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.2496583622411438e-05,
      "loss": 0.0032,
      "step": 8140
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.2486071691369706e-05,
      "loss": 0.0163,
      "step": 8150
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.2475559760327974e-05,
      "loss": 0.0106,
      "step": 8160
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.246504782928624e-05,
      "loss": 0.0167,
      "step": 8170
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.2454535898244508e-05,
      "loss": 0.0011,
      "step": 8180
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.2444023967202776e-05,
      "loss": 0.0031,
      "step": 8190
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.2433512036161044e-05,
      "loss": 0.0043,
      "step": 8200
    },
    {
      "epoch": 2.46,
      "eval_f1": 0.9924375590787866,
      "eval_f1_0": 0.9976433033081736,
      "eval_f1_1": 0.9978344308235876,
      "eval_f1_2": 0.9756106831886663,
      "eval_f1_3": 0.9986618189947187,
      "eval_loss": 0.016243239864706993,
      "eval_runtime": 247.4015,
      "eval_samples_per_second": 47.971,
      "eval_steps_per_second": 0.19,
      "step": 8200
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.2423000105119312e-05,
      "loss": 0.002,
      "step": 8210
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.2412488174077579e-05,
      "loss": 0.0346,
      "step": 8220
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.2401976243035847e-05,
      "loss": 0.0023,
      "step": 8230
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.2391464311994115e-05,
      "loss": 0.0055,
      "step": 8240
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.2380952380952383e-05,
      "loss": 0.0022,
      "step": 8250
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.2370440449910651e-05,
      "loss": 0.0015,
      "step": 8260
    },
    {
      "epoch": 2.48,
      "learning_rate": 1.2359928518868917e-05,
      "loss": 0.0105,
      "step": 8270
    },
    {
      "epoch": 2.48,
      "learning_rate": 1.2349416587827185e-05,
      "loss": 0.0018,
      "step": 8280
    },
    {
      "epoch": 2.48,
      "learning_rate": 1.2338904656785453e-05,
      "loss": 0.0026,
      "step": 8290
    },
    {
      "epoch": 2.49,
      "learning_rate": 1.2328392725743721e-05,
      "loss": 0.0013,
      "step": 8300
    },
    {
      "epoch": 2.49,
      "learning_rate": 1.2317880794701988e-05,
      "loss": 0.0029,
      "step": 8310
    },
    {
      "epoch": 2.49,
      "learning_rate": 1.2307368863660256e-05,
      "loss": 0.0026,
      "step": 8320
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.2296856932618524e-05,
      "loss": 0.0019,
      "step": 8330
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.2286345001576792e-05,
      "loss": 0.0014,
      "step": 8340
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.227583307053506e-05,
      "loss": 0.0125,
      "step": 8350
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.2265321139493326e-05,
      "loss": 0.0015,
      "step": 8360
    },
    {
      "epoch": 2.51,
      "learning_rate": 1.2254809208451594e-05,
      "loss": 0.0015,
      "step": 8370
    },
    {
      "epoch": 2.51,
      "learning_rate": 1.2244297277409862e-05,
      "loss": 0.0017,
      "step": 8380
    },
    {
      "epoch": 2.51,
      "learning_rate": 1.2233785346368127e-05,
      "loss": 0.0034,
      "step": 8390
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.2223273415326395e-05,
      "loss": 0.002,
      "step": 8400
    },
    {
      "epoch": 2.52,
      "eval_f1": 0.9921570352782105,
      "eval_f1_0": 0.9975598984211107,
      "eval_f1_1": 0.9979125094167103,
      "eval_f1_2": 0.9747557133917731,
      "eval_f1_3": 0.998400019883248,
      "eval_loss": 0.021348189562559128,
      "eval_runtime": 247.28,
      "eval_samples_per_second": 47.994,
      "eval_steps_per_second": 0.19,
      "step": 8400
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.2212761484284663e-05,
      "loss": 0.0016,
      "step": 8410
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.2202249553242931e-05,
      "loss": 0.0017,
      "step": 8420
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.2191737622201199e-05,
      "loss": 0.0118,
      "step": 8430
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.2181225691159465e-05,
      "loss": 0.0024,
      "step": 8440
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.2170713760117733e-05,
      "loss": 0.0013,
      "step": 8450
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.2160201829076002e-05,
      "loss": 0.0021,
      "step": 8460
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.214968989803427e-05,
      "loss": 0.0093,
      "step": 8470
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.2139177966992538e-05,
      "loss": 0.005,
      "step": 8480
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.2128666035950804e-05,
      "loss": 0.0159,
      "step": 8490
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.2118154104909072e-05,
      "loss": 0.0232,
      "step": 8500
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.210764217386734e-05,
      "loss": 0.0044,
      "step": 8510
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.2097130242825608e-05,
      "loss": 0.0118,
      "step": 8520
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.2086618311783876e-05,
      "loss": 0.0112,
      "step": 8530
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.2076106380742142e-05,
      "loss": 0.0021,
      "step": 8540
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.206559444970041e-05,
      "loss": 0.035,
      "step": 8550
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.2055082518658679e-05,
      "loss": 0.0156,
      "step": 8560
    },
    {
      "epoch": 2.57,
      "learning_rate": 1.2044570587616947e-05,
      "loss": 0.0073,
      "step": 8570
    },
    {
      "epoch": 2.57,
      "learning_rate": 1.2034058656575213e-05,
      "loss": 0.0181,
      "step": 8580
    },
    {
      "epoch": 2.57,
      "learning_rate": 1.2023546725533481e-05,
      "loss": 0.0035,
      "step": 8590
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.2013034794491749e-05,
      "loss": 0.0069,
      "step": 8600
    },
    {
      "epoch": 2.58,
      "eval_f1": 0.9931501671298951,
      "eval_f1_0": 0.9978880401926541,
      "eval_f1_1": 0.9978346019116824,
      "eval_f1_2": 0.9782102790401546,
      "eval_f1_3": 0.9986677473750897,
      "eval_loss": 0.014435278251767159,
      "eval_runtime": 247.2251,
      "eval_samples_per_second": 48.005,
      "eval_steps_per_second": 0.19,
      "step": 8600
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.2002522863450017e-05,
      "loss": 0.0027,
      "step": 8610
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.1992010932408285e-05,
      "loss": 0.0027,
      "step": 8620
    },
    {
      "epoch": 2.59,
      "learning_rate": 1.1981499001366551e-05,
      "loss": 0.0031,
      "step": 8630
    },
    {
      "epoch": 2.59,
      "learning_rate": 1.197098707032482e-05,
      "loss": 0.0155,
      "step": 8640
    },
    {
      "epoch": 2.59,
      "learning_rate": 1.1960475139283088e-05,
      "loss": 0.0064,
      "step": 8650
    },
    {
      "epoch": 2.59,
      "learning_rate": 1.1949963208241356e-05,
      "loss": 0.004,
      "step": 8660
    },
    {
      "epoch": 2.6,
      "learning_rate": 1.1939451277199624e-05,
      "loss": 0.0018,
      "step": 8670
    },
    {
      "epoch": 2.6,
      "learning_rate": 1.192893934615789e-05,
      "loss": 0.0049,
      "step": 8680
    },
    {
      "epoch": 2.6,
      "learning_rate": 1.1918427415116158e-05,
      "loss": 0.0034,
      "step": 8690
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.1907915484074426e-05,
      "loss": 0.002,
      "step": 8700
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.1897403553032694e-05,
      "loss": 0.0398,
      "step": 8710
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.1886891621990962e-05,
      "loss": 0.0193,
      "step": 8720
    },
    {
      "epoch": 2.62,
      "learning_rate": 1.1876379690949228e-05,
      "loss": 0.0029,
      "step": 8730
    },
    {
      "epoch": 2.62,
      "learning_rate": 1.1865867759907497e-05,
      "loss": 0.0259,
      "step": 8740
    },
    {
      "epoch": 2.62,
      "learning_rate": 1.1855355828865765e-05,
      "loss": 0.0169,
      "step": 8750
    },
    {
      "epoch": 2.62,
      "learning_rate": 1.1844843897824033e-05,
      "loss": 0.0015,
      "step": 8760
    },
    {
      "epoch": 2.63,
      "learning_rate": 1.18343319667823e-05,
      "loss": 0.0269,
      "step": 8770
    },
    {
      "epoch": 2.63,
      "learning_rate": 1.1823820035740567e-05,
      "loss": 0.0031,
      "step": 8780
    },
    {
      "epoch": 2.63,
      "learning_rate": 1.1813308104698835e-05,
      "loss": 0.0093,
      "step": 8790
    },
    {
      "epoch": 2.64,
      "learning_rate": 1.1802796173657103e-05,
      "loss": 0.0081,
      "step": 8800
    },
    {
      "epoch": 2.64,
      "eval_f1": 0.9924072469517862,
      "eval_f1_0": 0.9976613989411902,
      "eval_f1_1": 0.997730547066689,
      "eval_f1_2": 0.976424302325263,
      "eval_f1_3": 0.9978127394740026,
      "eval_loss": 0.014092279598116875,
      "eval_runtime": 246.9623,
      "eval_samples_per_second": 48.056,
      "eval_steps_per_second": 0.19,
      "step": 8800
    },
    {
      "epoch": 2.64,
      "learning_rate": 1.1792284242615368e-05,
      "loss": 0.0131,
      "step": 8810
    },
    {
      "epoch": 2.64,
      "learning_rate": 1.1781772311573636e-05,
      "loss": 0.0032,
      "step": 8820
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.1771260380531904e-05,
      "loss": 0.0021,
      "step": 8830
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.1760748449490172e-05,
      "loss": 0.0016,
      "step": 8840
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.175023651844844e-05,
      "loss": 0.0161,
      "step": 8850
    },
    {
      "epoch": 2.65,
      "learning_rate": 1.1739724587406706e-05,
      "loss": 0.0015,
      "step": 8860
    },
    {
      "epoch": 2.66,
      "learning_rate": 1.1729212656364974e-05,
      "loss": 0.0015,
      "step": 8870
    },
    {
      "epoch": 2.66,
      "learning_rate": 1.1718700725323242e-05,
      "loss": 0.0023,
      "step": 8880
    },
    {
      "epoch": 2.66,
      "learning_rate": 1.170818879428151e-05,
      "loss": 0.0011,
      "step": 8890
    },
    {
      "epoch": 2.67,
      "learning_rate": 1.1697676863239777e-05,
      "loss": 0.0019,
      "step": 8900
    },
    {
      "epoch": 2.67,
      "learning_rate": 1.1687164932198045e-05,
      "loss": 0.0198,
      "step": 8910
    },
    {
      "epoch": 2.67,
      "learning_rate": 1.1676653001156313e-05,
      "loss": 0.0196,
      "step": 8920
    },
    {
      "epoch": 2.68,
      "learning_rate": 1.166614107011458e-05,
      "loss": 0.0135,
      "step": 8930
    },
    {
      "epoch": 2.68,
      "learning_rate": 1.1655629139072849e-05,
      "loss": 0.0033,
      "step": 8940
    },
    {
      "epoch": 2.68,
      "learning_rate": 1.1645117208031115e-05,
      "loss": 0.0026,
      "step": 8950
    },
    {
      "epoch": 2.68,
      "learning_rate": 1.1634605276989383e-05,
      "loss": 0.0027,
      "step": 8960
    },
    {
      "epoch": 2.69,
      "learning_rate": 1.1624093345947651e-05,
      "loss": 0.0032,
      "step": 8970
    },
    {
      "epoch": 2.69,
      "learning_rate": 1.161358141490592e-05,
      "loss": 0.0058,
      "step": 8980
    },
    {
      "epoch": 2.69,
      "learning_rate": 1.1603069483864187e-05,
      "loss": 0.0238,
      "step": 8990
    },
    {
      "epoch": 2.7,
      "learning_rate": 1.1592557552822454e-05,
      "loss": 0.004,
      "step": 9000
    },
    {
      "epoch": 2.7,
      "eval_f1": 0.9921203014698665,
      "eval_f1_0": 0.9975569091257982,
      "eval_f1_1": 0.9979021569048568,
      "eval_f1_2": 0.9747215150876396,
      "eval_f1_3": 0.998300624761171,
      "eval_loss": 0.016352033242583275,
      "eval_runtime": 247.1891,
      "eval_samples_per_second": 48.012,
      "eval_steps_per_second": 0.19,
      "step": 9000
    },
    {
      "epoch": 2.7,
      "learning_rate": 1.1582045621780722e-05,
      "loss": 0.0049,
      "step": 9010
    },
    {
      "epoch": 2.7,
      "learning_rate": 1.157153369073899e-05,
      "loss": 0.0058,
      "step": 9020
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.1561021759697258e-05,
      "loss": 0.016,
      "step": 9030
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.1550509828655526e-05,
      "loss": 0.0025,
      "step": 9040
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.1539997897613792e-05,
      "loss": 0.0025,
      "step": 9050
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.152948596657206e-05,
      "loss": 0.0215,
      "step": 9060
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.1518974035530328e-05,
      "loss": 0.0036,
      "step": 9070
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.1508462104488596e-05,
      "loss": 0.0016,
      "step": 9080
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.1497950173446863e-05,
      "loss": 0.0036,
      "step": 9090
    },
    {
      "epoch": 2.73,
      "learning_rate": 1.148743824240513e-05,
      "loss": 0.0021,
      "step": 9100
    },
    {
      "epoch": 2.73,
      "learning_rate": 1.1476926311363399e-05,
      "loss": 0.0016,
      "step": 9110
    },
    {
      "epoch": 2.73,
      "learning_rate": 1.1466414380321667e-05,
      "loss": 0.003,
      "step": 9120
    },
    {
      "epoch": 2.74,
      "learning_rate": 1.1455902449279935e-05,
      "loss": 0.0032,
      "step": 9130
    },
    {
      "epoch": 2.74,
      "learning_rate": 1.1445390518238201e-05,
      "loss": 0.0048,
      "step": 9140
    },
    {
      "epoch": 2.74,
      "learning_rate": 1.143487858719647e-05,
      "loss": 0.0059,
      "step": 9150
    },
    {
      "epoch": 2.74,
      "learning_rate": 1.1424366656154737e-05,
      "loss": 0.0063,
      "step": 9160
    },
    {
      "epoch": 2.75,
      "learning_rate": 1.1413854725113005e-05,
      "loss": 0.0019,
      "step": 9170
    },
    {
      "epoch": 2.75,
      "learning_rate": 1.1403342794071273e-05,
      "loss": 0.0026,
      "step": 9180
    },
    {
      "epoch": 2.75,
      "learning_rate": 1.139283086302954e-05,
      "loss": 0.0015,
      "step": 9190
    },
    {
      "epoch": 2.76,
      "learning_rate": 1.1382318931987808e-05,
      "loss": 0.004,
      "step": 9200
    },
    {
      "epoch": 2.76,
      "eval_f1": 0.9925668540109466,
      "eval_f1_0": 0.9976983207508967,
      "eval_f1_1": 0.997859994554108,
      "eval_f1_2": 0.9764069357663286,
      "eval_f1_3": 0.9983021649724528,
      "eval_loss": 0.018431581556797028,
      "eval_runtime": 247.2626,
      "eval_samples_per_second": 47.998,
      "eval_steps_per_second": 0.19,
      "step": 9200
    },
    {
      "epoch": 2.76,
      "learning_rate": 1.1371807000946076e-05,
      "loss": 0.0012,
      "step": 9210
    },
    {
      "epoch": 2.76,
      "learning_rate": 1.1361295069904344e-05,
      "loss": 0.0175,
      "step": 9220
    },
    {
      "epoch": 2.77,
      "learning_rate": 1.1350783138862612e-05,
      "loss": 0.0046,
      "step": 9230
    },
    {
      "epoch": 2.77,
      "learning_rate": 1.1340271207820876e-05,
      "loss": 0.0041,
      "step": 9240
    },
    {
      "epoch": 2.77,
      "learning_rate": 1.1329759276779145e-05,
      "loss": 0.0082,
      "step": 9250
    },
    {
      "epoch": 2.77,
      "learning_rate": 1.1319247345737413e-05,
      "loss": 0.0017,
      "step": 9260
    },
    {
      "epoch": 2.78,
      "learning_rate": 1.1308735414695679e-05,
      "loss": 0.0135,
      "step": 9270
    },
    {
      "epoch": 2.78,
      "learning_rate": 1.1298223483653947e-05,
      "loss": 0.0037,
      "step": 9280
    },
    {
      "epoch": 2.78,
      "learning_rate": 1.1287711552612215e-05,
      "loss": 0.0032,
      "step": 9290
    },
    {
      "epoch": 2.79,
      "learning_rate": 1.1277199621570483e-05,
      "loss": 0.0204,
      "step": 9300
    },
    {
      "epoch": 2.79,
      "learning_rate": 1.1266687690528751e-05,
      "loss": 0.0105,
      "step": 9310
    },
    {
      "epoch": 2.79,
      "learning_rate": 1.1256175759487017e-05,
      "loss": 0.0031,
      "step": 9320
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.1245663828445285e-05,
      "loss": 0.0051,
      "step": 9330
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.1235151897403554e-05,
      "loss": 0.0105,
      "step": 9340
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.1224639966361822e-05,
      "loss": 0.0022,
      "step": 9350
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.121412803532009e-05,
      "loss": 0.0051,
      "step": 9360
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.1203616104278356e-05,
      "loss": 0.0016,
      "step": 9370
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.1193104173236624e-05,
      "loss": 0.002,
      "step": 9380
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.1182592242194892e-05,
      "loss": 0.0156,
      "step": 9390
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.117208031115316e-05,
      "loss": 0.0206,
      "step": 9400
    },
    {
      "epoch": 2.82,
      "eval_f1": 0.9903516533025725,
      "eval_f1_0": 0.996923897127459,
      "eval_f1_1": 0.9974810986476972,
      "eval_f1_2": 0.9681475339043446,
      "eval_f1_3": 0.9988540835307891,
      "eval_loss": 0.02095732092857361,
      "eval_runtime": 246.81,
      "eval_samples_per_second": 48.086,
      "eval_steps_per_second": 0.19,
      "step": 9400
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.1161568380111426e-05,
      "loss": 0.015,
      "step": 9410
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.1151056449069694e-05,
      "loss": 0.0018,
      "step": 9420
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.1140544518027962e-05,
      "loss": 0.0269,
      "step": 9430
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.113003258698623e-05,
      "loss": 0.005,
      "step": 9440
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.1119520655944499e-05,
      "loss": 0.002,
      "step": 9450
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.1109008724902765e-05,
      "loss": 0.0037,
      "step": 9460
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.1098496793861033e-05,
      "loss": 0.0078,
      "step": 9470
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.1087984862819301e-05,
      "loss": 0.0016,
      "step": 9480
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.1077472931777569e-05,
      "loss": 0.0158,
      "step": 9490
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.1066961000735837e-05,
      "loss": 0.0014,
      "step": 9500
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.1056449069694103e-05,
      "loss": 0.0018,
      "step": 9510
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.1045937138652371e-05,
      "loss": 0.0045,
      "step": 9520
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.103542520761064e-05,
      "loss": 0.0018,
      "step": 9530
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.1024913276568908e-05,
      "loss": 0.0066,
      "step": 9540
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.1014401345527176e-05,
      "loss": 0.0187,
      "step": 9550
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.1003889414485442e-05,
      "loss": 0.0018,
      "step": 9560
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.099337748344371e-05,
      "loss": 0.0033,
      "step": 9570
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.0982865552401978e-05,
      "loss": 0.0291,
      "step": 9580
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.0972353621360246e-05,
      "loss": 0.0019,
      "step": 9590
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.0961841690318512e-05,
      "loss": 0.0214,
      "step": 9600
    },
    {
      "epoch": 2.88,
      "eval_f1": 0.9929649458931602,
      "eval_f1_0": 0.9977650476444936,
      "eval_f1_1": 0.9969581632221962,
      "eval_f1_2": 0.9783259656807685,
      "eval_f1_3": 0.9988106070251822,
      "eval_loss": 0.013837916776537895,
      "eval_runtime": 247.1416,
      "eval_samples_per_second": 48.021,
      "eval_steps_per_second": 0.19,
      "step": 9600
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.095132975927678e-05,
      "loss": 0.0022,
      "step": 9610
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.0940817828235048e-05,
      "loss": 0.0124,
      "step": 9620
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.0930305897193317e-05,
      "loss": 0.0046,
      "step": 9630
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.0919793966151585e-05,
      "loss": 0.0041,
      "step": 9640
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.0909282035109851e-05,
      "loss": 0.0075,
      "step": 9650
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.0898770104068119e-05,
      "loss": 0.0032,
      "step": 9660
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.0888258173026385e-05,
      "loss": 0.0039,
      "step": 9670
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.0877746241984652e-05,
      "loss": 0.0017,
      "step": 9680
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.086723431094292e-05,
      "loss": 0.0027,
      "step": 9690
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.0856722379901188e-05,
      "loss": 0.0013,
      "step": 9700
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.0846210448859456e-05,
      "loss": 0.0292,
      "step": 9710
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.0835698517817724e-05,
      "loss": 0.0019,
      "step": 9720
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.082518658677599e-05,
      "loss": 0.0104,
      "step": 9730
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.0814674655734258e-05,
      "loss": 0.0031,
      "step": 9740
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.0804162724692526e-05,
      "loss": 0.0088,
      "step": 9750
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.0793650793650794e-05,
      "loss": 0.0014,
      "step": 9760
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.0783138862609062e-05,
      "loss": 0.0179,
      "step": 9770
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.0772626931567329e-05,
      "loss": 0.002,
      "step": 9780
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.0762115000525597e-05,
      "loss": 0.0032,
      "step": 9790
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0751603069483865e-05,
      "loss": 0.0014,
      "step": 9800
    },
    {
      "epoch": 2.94,
      "eval_f1": 0.9940370205678817,
      "eval_f1_0": 0.9981610428595508,
      "eval_f1_1": 0.9977305731885351,
      "eval_f1_2": 0.9818766785271222,
      "eval_f1_3": 0.9983797876963189,
      "eval_loss": 0.012908019125461578,
      "eval_runtime": 247.3504,
      "eval_samples_per_second": 47.981,
      "eval_steps_per_second": 0.19,
      "step": 9800
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0741091138442133e-05,
      "loss": 0.0021,
      "step": 9810
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.07305792074004e-05,
      "loss": 0.0023,
      "step": 9820
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0720067276358667e-05,
      "loss": 0.0023,
      "step": 9830
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.0709555345316935e-05,
      "loss": 0.0093,
      "step": 9840
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.0699043414275203e-05,
      "loss": 0.0018,
      "step": 9850
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.0688531483233471e-05,
      "loss": 0.0017,
      "step": 9860
    },
    {
      "epoch": 2.96,
      "learning_rate": 1.0678019552191738e-05,
      "loss": 0.0083,
      "step": 9870
    },
    {
      "epoch": 2.96,
      "learning_rate": 1.0667507621150006e-05,
      "loss": 0.0062,
      "step": 9880
    },
    {
      "epoch": 2.96,
      "learning_rate": 1.0656995690108274e-05,
      "loss": 0.0021,
      "step": 9890
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.0646483759066542e-05,
      "loss": 0.0015,
      "step": 9900
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.063597182802481e-05,
      "loss": 0.0012,
      "step": 9910
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.0625459896983076e-05,
      "loss": 0.0101,
      "step": 9920
    },
    {
      "epoch": 2.97,
      "learning_rate": 1.0614947965941344e-05,
      "loss": 0.0034,
      "step": 9930
    },
    {
      "epoch": 2.98,
      "learning_rate": 1.0604436034899612e-05,
      "loss": 0.0013,
      "step": 9940
    },
    {
      "epoch": 2.98,
      "learning_rate": 1.059392410385788e-05,
      "loss": 0.0008,
      "step": 9950
    },
    {
      "epoch": 2.98,
      "learning_rate": 1.0583412172816148e-05,
      "loss": 0.0033,
      "step": 9960
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.0572900241774415e-05,
      "loss": 0.0029,
      "step": 9970
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.0562388310732683e-05,
      "loss": 0.0008,
      "step": 9980
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.055187637969095e-05,
      "loss": 0.002,
      "step": 9990
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.0541364448649219e-05,
      "loss": 0.002,
      "step": 10000
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.9932719171456413,
      "eval_f1_0": 0.997915452249938,
      "eval_f1_1": 0.9979133423346017,
      "eval_f1_2": 0.9788386724418152,
      "eval_f1_3": 0.9984202015562101,
      "eval_loss": 0.015370514243841171,
      "eval_runtime": 247.0497,
      "eval_samples_per_second": 48.039,
      "eval_steps_per_second": 0.19,
      "step": 10000
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.0530852517607487e-05,
      "loss": 0.0019,
      "step": 10010
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.0520340586565753e-05,
      "loss": 0.0145,
      "step": 10020
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.0509828655524021e-05,
      "loss": 0.0017,
      "step": 10030
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.049931672448229e-05,
      "loss": 0.0103,
      "step": 10040
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.0488804793440557e-05,
      "loss": 0.0116,
      "step": 10050
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.0478292862398825e-05,
      "loss": 0.0008,
      "step": 10060
    },
    {
      "epoch": 3.02,
      "learning_rate": 1.0467780931357092e-05,
      "loss": 0.0011,
      "step": 10070
    },
    {
      "epoch": 3.02,
      "learning_rate": 1.045726900031536e-05,
      "loss": 0.0011,
      "step": 10080
    },
    {
      "epoch": 3.02,
      "learning_rate": 1.0446757069273626e-05,
      "loss": 0.0026,
      "step": 10090
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.0436245138231892e-05,
      "loss": 0.0026,
      "step": 10100
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.042573320719016e-05,
      "loss": 0.0014,
      "step": 10110
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.0415221276148428e-05,
      "loss": 0.0009,
      "step": 10120
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.0404709345106697e-05,
      "loss": 0.0075,
      "step": 10130
    },
    {
      "epoch": 3.04,
      "learning_rate": 1.0394197414064965e-05,
      "loss": 0.0019,
      "step": 10140
    },
    {
      "epoch": 3.04,
      "learning_rate": 1.0383685483023231e-05,
      "loss": 0.0016,
      "step": 10150
    },
    {
      "epoch": 3.04,
      "learning_rate": 1.0373173551981499e-05,
      "loss": 0.004,
      "step": 10160
    },
    {
      "epoch": 3.05,
      "learning_rate": 1.0362661620939767e-05,
      "loss": 0.003,
      "step": 10170
    },
    {
      "epoch": 3.05,
      "learning_rate": 1.0352149689898035e-05,
      "loss": 0.0363,
      "step": 10180
    },
    {
      "epoch": 3.05,
      "learning_rate": 1.0341637758856301e-05,
      "loss": 0.001,
      "step": 10190
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.033112582781457e-05,
      "loss": 0.0025,
      "step": 10200
    },
    {
      "epoch": 3.06,
      "eval_f1": 0.9942635493742759,
      "eval_f1_0": 0.9982225565091258,
      "eval_f1_1": 0.997949447847112,
      "eval_f1_2": 0.9821862655727742,
      "eval_f1_3": 0.9986959275680912,
      "eval_loss": 0.01297374814748764,
      "eval_runtime": 247.1323,
      "eval_samples_per_second": 48.023,
      "eval_steps_per_second": 0.19,
      "step": 10200
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.0320613896772837e-05,
      "loss": 0.0027,
      "step": 10210
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.0310101965731105e-05,
      "loss": 0.0016,
      "step": 10220
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.0299590034689374e-05,
      "loss": 0.0021,
      "step": 10230
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.028907810364764e-05,
      "loss": 0.0017,
      "step": 10240
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.0278566172605908e-05,
      "loss": 0.0018,
      "step": 10250
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.0268054241564176e-05,
      "loss": 0.0015,
      "step": 10260
    },
    {
      "epoch": 3.08,
      "learning_rate": 1.0257542310522444e-05,
      "loss": 0.006,
      "step": 10270
    },
    {
      "epoch": 3.08,
      "learning_rate": 1.0247030379480712e-05,
      "loss": 0.0018,
      "step": 10280
    },
    {
      "epoch": 3.08,
      "learning_rate": 1.0236518448438978e-05,
      "loss": 0.017,
      "step": 10290
    },
    {
      "epoch": 3.09,
      "learning_rate": 1.0226006517397246e-05,
      "loss": 0.0175,
      "step": 10300
    },
    {
      "epoch": 3.09,
      "learning_rate": 1.0215494586355514e-05,
      "loss": 0.0019,
      "step": 10310
    },
    {
      "epoch": 3.09,
      "learning_rate": 1.0204982655313783e-05,
      "loss": 0.0013,
      "step": 10320
    },
    {
      "epoch": 3.09,
      "learning_rate": 1.019447072427205e-05,
      "loss": 0.001,
      "step": 10330
    },
    {
      "epoch": 3.1,
      "learning_rate": 1.0183958793230317e-05,
      "loss": 0.0011,
      "step": 10340
    },
    {
      "epoch": 3.1,
      "learning_rate": 1.0173446862188585e-05,
      "loss": 0.0014,
      "step": 10350
    },
    {
      "epoch": 3.1,
      "learning_rate": 1.0162934931146853e-05,
      "loss": 0.0014,
      "step": 10360
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.0152423000105121e-05,
      "loss": 0.0011,
      "step": 10370
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.0141911069063387e-05,
      "loss": 0.0039,
      "step": 10380
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.0131399138021655e-05,
      "loss": 0.0105,
      "step": 10390
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.0120887206979923e-05,
      "loss": 0.0035,
      "step": 10400
    },
    {
      "epoch": 3.12,
      "eval_f1": 0.9935808549769547,
      "eval_f1_0": 0.9980201750959351,
      "eval_f1_1": 0.9980946100851483,
      "eval_f1_2": 0.9798504975737691,
      "eval_f1_3": 0.9983581371529664,
      "eval_loss": 0.014101088047027588,
      "eval_runtime": 247.0444,
      "eval_samples_per_second": 48.04,
      "eval_steps_per_second": 0.19,
      "step": 10400
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.0110375275938191e-05,
      "loss": 0.002,
      "step": 10410
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.009986334489646e-05,
      "loss": 0.0168,
      "step": 10420
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.0089351413854726e-05,
      "loss": 0.005,
      "step": 10430
    },
    {
      "epoch": 3.13,
      "learning_rate": 1.0078839482812994e-05,
      "loss": 0.0019,
      "step": 10440
    },
    {
      "epoch": 3.13,
      "learning_rate": 1.0068327551771262e-05,
      "loss": 0.0012,
      "step": 10450
    },
    {
      "epoch": 3.13,
      "learning_rate": 1.005781562072953e-05,
      "loss": 0.0065,
      "step": 10460
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.0047303689687798e-05,
      "loss": 0.0019,
      "step": 10470
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.0036791758646064e-05,
      "loss": 0.0026,
      "step": 10480
    },
    {
      "epoch": 3.14,
      "learning_rate": 1.0026279827604332e-05,
      "loss": 0.0015,
      "step": 10490
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.00157678965626e-05,
      "loss": 0.0023,
      "step": 10500
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.0005255965520868e-05,
      "loss": 0.0037,
      "step": 10510
    },
    {
      "epoch": 3.15,
      "learning_rate": 9.994744034479135e-06,
      "loss": 0.0026,
      "step": 10520
    },
    {
      "epoch": 3.15,
      "learning_rate": 9.984232103437403e-06,
      "loss": 0.0022,
      "step": 10530
    },
    {
      "epoch": 3.16,
      "learning_rate": 9.97372017239567e-06,
      "loss": 0.002,
      "step": 10540
    },
    {
      "epoch": 3.16,
      "learning_rate": 9.963208241353937e-06,
      "loss": 0.0012,
      "step": 10550
    },
    {
      "epoch": 3.16,
      "learning_rate": 9.952696310312205e-06,
      "loss": 0.0058,
      "step": 10560
    },
    {
      "epoch": 3.17,
      "learning_rate": 9.942184379270473e-06,
      "loss": 0.0012,
      "step": 10570
    },
    {
      "epoch": 3.17,
      "learning_rate": 9.931672448228741e-06,
      "loss": 0.0022,
      "step": 10580
    },
    {
      "epoch": 3.17,
      "learning_rate": 9.921160517187008e-06,
      "loss": 0.0013,
      "step": 10590
    },
    {
      "epoch": 3.18,
      "learning_rate": 9.910648586145276e-06,
      "loss": 0.0015,
      "step": 10600
    },
    {
      "epoch": 3.18,
      "eval_f1": 0.9931216929057893,
      "eval_f1_0": 0.9978751751544805,
      "eval_f1_1": 0.9981269279509511,
      "eval_f1_2": 0.9785917086383761,
      "eval_f1_3": 0.9978929598793495,
      "eval_loss": 0.01746038906276226,
      "eval_runtime": 247.0078,
      "eval_samples_per_second": 48.047,
      "eval_steps_per_second": 0.19,
      "step": 10600
    },
    {
      "epoch": 3.18,
      "learning_rate": 9.900136655103544e-06,
      "loss": 0.0059,
      "step": 10610
    },
    {
      "epoch": 3.18,
      "learning_rate": 9.889624724061812e-06,
      "loss": 0.0012,
      "step": 10620
    },
    {
      "epoch": 3.18,
      "learning_rate": 9.879112793020078e-06,
      "loss": 0.0018,
      "step": 10630
    },
    {
      "epoch": 3.19,
      "learning_rate": 9.868600861978346e-06,
      "loss": 0.0041,
      "step": 10640
    },
    {
      "epoch": 3.19,
      "learning_rate": 9.858088930936614e-06,
      "loss": 0.0009,
      "step": 10650
    },
    {
      "epoch": 3.19,
      "learning_rate": 9.84757699989488e-06,
      "loss": 0.003,
      "step": 10660
    },
    {
      "epoch": 3.2,
      "learning_rate": 9.837065068853149e-06,
      "loss": 0.0243,
      "step": 10670
    },
    {
      "epoch": 3.2,
      "learning_rate": 9.826553137811417e-06,
      "loss": 0.0012,
      "step": 10680
    },
    {
      "epoch": 3.2,
      "learning_rate": 9.816041206769685e-06,
      "loss": 0.0029,
      "step": 10690
    },
    {
      "epoch": 3.21,
      "learning_rate": 9.805529275727951e-06,
      "loss": 0.0013,
      "step": 10700
    },
    {
      "epoch": 3.21,
      "learning_rate": 9.795017344686219e-06,
      "loss": 0.0048,
      "step": 10710
    },
    {
      "epoch": 3.21,
      "learning_rate": 9.784505413644487e-06,
      "loss": 0.0016,
      "step": 10720
    },
    {
      "epoch": 3.21,
      "learning_rate": 9.773993482602755e-06,
      "loss": 0.002,
      "step": 10730
    },
    {
      "epoch": 3.22,
      "learning_rate": 9.763481551561023e-06,
      "loss": 0.0018,
      "step": 10740
    },
    {
      "epoch": 3.22,
      "learning_rate": 9.75296962051929e-06,
      "loss": 0.0013,
      "step": 10750
    },
    {
      "epoch": 3.22,
      "learning_rate": 9.742457689477558e-06,
      "loss": 0.026,
      "step": 10760
    },
    {
      "epoch": 3.23,
      "learning_rate": 9.731945758435826e-06,
      "loss": 0.001,
      "step": 10770
    },
    {
      "epoch": 3.23,
      "learning_rate": 9.721433827394094e-06,
      "loss": 0.0008,
      "step": 10780
    },
    {
      "epoch": 3.23,
      "learning_rate": 9.710921896352362e-06,
      "loss": 0.0019,
      "step": 10790
    },
    {
      "epoch": 3.24,
      "learning_rate": 9.700409965310628e-06,
      "loss": 0.0012,
      "step": 10800
    },
    {
      "epoch": 3.24,
      "eval_f1": 0.9927619551750094,
      "eval_f1_0": 0.9977062461634947,
      "eval_f1_1": 0.9961181811015598,
      "eval_f1_2": 0.9784811300680376,
      "eval_f1_3": 0.9987422633669455,
      "eval_loss": 0.01910644769668579,
      "eval_runtime": 247.1,
      "eval_samples_per_second": 48.029,
      "eval_steps_per_second": 0.19,
      "step": 10800
    },
    {
      "epoch": 3.24,
      "learning_rate": 9.689898034268896e-06,
      "loss": 0.0026,
      "step": 10810
    },
    {
      "epoch": 3.24,
      "learning_rate": 9.679386103227164e-06,
      "loss": 0.0028,
      "step": 10820
    },
    {
      "epoch": 3.24,
      "learning_rate": 9.668874172185432e-06,
      "loss": 0.001,
      "step": 10830
    },
    {
      "epoch": 3.25,
      "learning_rate": 9.658362241143699e-06,
      "loss": 0.0015,
      "step": 10840
    },
    {
      "epoch": 3.25,
      "learning_rate": 9.647850310101967e-06,
      "loss": 0.0013,
      "step": 10850
    },
    {
      "epoch": 3.25,
      "learning_rate": 9.637338379060233e-06,
      "loss": 0.0011,
      "step": 10860
    },
    {
      "epoch": 3.26,
      "learning_rate": 9.626826448018501e-06,
      "loss": 0.0058,
      "step": 10870
    },
    {
      "epoch": 3.26,
      "learning_rate": 9.616314516976769e-06,
      "loss": 0.002,
      "step": 10880
    },
    {
      "epoch": 3.26,
      "learning_rate": 9.605802585935037e-06,
      "loss": 0.0018,
      "step": 10890
    },
    {
      "epoch": 3.27,
      "learning_rate": 9.595290654893305e-06,
      "loss": 0.001,
      "step": 10900
    },
    {
      "epoch": 3.27,
      "learning_rate": 9.584778723851571e-06,
      "loss": 0.0015,
      "step": 10910
    },
    {
      "epoch": 3.27,
      "learning_rate": 9.57426679280984e-06,
      "loss": 0.0076,
      "step": 10920
    },
    {
      "epoch": 3.27,
      "learning_rate": 9.563754861768108e-06,
      "loss": 0.0013,
      "step": 10930
    },
    {
      "epoch": 3.28,
      "learning_rate": 9.553242930726376e-06,
      "loss": 0.0217,
      "step": 10940
    },
    {
      "epoch": 3.28,
      "learning_rate": 9.542730999684644e-06,
      "loss": 0.0019,
      "step": 10950
    },
    {
      "epoch": 3.28,
      "learning_rate": 9.53221906864291e-06,
      "loss": 0.0017,
      "step": 10960
    },
    {
      "epoch": 3.29,
      "learning_rate": 9.521707137601178e-06,
      "loss": 0.0017,
      "step": 10970
    },
    {
      "epoch": 3.29,
      "learning_rate": 9.511195206559446e-06,
      "loss": 0.0016,
      "step": 10980
    },
    {
      "epoch": 3.29,
      "learning_rate": 9.500683275517714e-06,
      "loss": 0.034,
      "step": 10990
    },
    {
      "epoch": 3.3,
      "learning_rate": 9.490171344475982e-06,
      "loss": 0.0014,
      "step": 11000
    },
    {
      "epoch": 3.3,
      "eval_f1": 0.9911482583328698,
      "eval_f1_0": 0.9971865425406552,
      "eval_f1_1": 0.9981415954661834,
      "eval_f1_2": 0.9704077595660335,
      "eval_f1_3": 0.9988571357586072,
      "eval_loss": 0.019447069615125656,
      "eval_runtime": 247.07,
      "eval_samples_per_second": 48.035,
      "eval_steps_per_second": 0.19,
      "step": 11000
    },
    {
      "epoch": 3.3,
      "learning_rate": 9.479659413434248e-06,
      "loss": 0.002,
      "step": 11010
    },
    {
      "epoch": 3.3,
      "learning_rate": 9.469147482392517e-06,
      "loss": 0.0016,
      "step": 11020
    },
    {
      "epoch": 3.3,
      "learning_rate": 9.458635551350785e-06,
      "loss": 0.0012,
      "step": 11030
    },
    {
      "epoch": 3.31,
      "learning_rate": 9.448123620309053e-06,
      "loss": 0.0009,
      "step": 11040
    },
    {
      "epoch": 3.31,
      "learning_rate": 9.437611689267319e-06,
      "loss": 0.0015,
      "step": 11050
    },
    {
      "epoch": 3.31,
      "learning_rate": 9.427099758225587e-06,
      "loss": 0.0174,
      "step": 11060
    },
    {
      "epoch": 3.32,
      "learning_rate": 9.416587827183853e-06,
      "loss": 0.0014,
      "step": 11070
    },
    {
      "epoch": 3.32,
      "learning_rate": 9.406075896142121e-06,
      "loss": 0.0014,
      "step": 11080
    },
    {
      "epoch": 3.32,
      "learning_rate": 9.39556396510039e-06,
      "loss": 0.0014,
      "step": 11090
    },
    {
      "epoch": 3.33,
      "learning_rate": 9.385052034058657e-06,
      "loss": 0.0051,
      "step": 11100
    },
    {
      "epoch": 3.33,
      "learning_rate": 9.374540103016926e-06,
      "loss": 0.0012,
      "step": 11110
    },
    {
      "epoch": 3.33,
      "learning_rate": 9.364028171975192e-06,
      "loss": 0.0012,
      "step": 11120
    },
    {
      "epoch": 3.33,
      "learning_rate": 9.35351624093346e-06,
      "loss": 0.0014,
      "step": 11130
    },
    {
      "epoch": 3.34,
      "learning_rate": 9.343004309891728e-06,
      "loss": 0.0016,
      "step": 11140
    },
    {
      "epoch": 3.34,
      "learning_rate": 9.332492378849996e-06,
      "loss": 0.0015,
      "step": 11150
    },
    {
      "epoch": 3.34,
      "learning_rate": 9.321980447808264e-06,
      "loss": 0.0015,
      "step": 11160
    },
    {
      "epoch": 3.35,
      "learning_rate": 9.31146851676653e-06,
      "loss": 0.0017,
      "step": 11170
    },
    {
      "epoch": 3.35,
      "learning_rate": 9.300956585724798e-06,
      "loss": 0.0109,
      "step": 11180
    },
    {
      "epoch": 3.35,
      "learning_rate": 9.290444654683066e-06,
      "loss": 0.0021,
      "step": 11190
    },
    {
      "epoch": 3.36,
      "learning_rate": 9.279932723641334e-06,
      "loss": 0.0028,
      "step": 11200
    },
    {
      "epoch": 3.36,
      "eval_f1": 0.9909173410703432,
      "eval_f1_0": 0.9972425041474811,
      "eval_f1_1": 0.9983874215042423,
      "eval_f1_2": 0.9692691835507409,
      "eval_f1_3": 0.9987702550789086,
      "eval_loss": 0.025893166661262512,
      "eval_runtime": 247.0717,
      "eval_samples_per_second": 48.035,
      "eval_steps_per_second": 0.19,
      "step": 11200
    },
    {
      "epoch": 3.36,
      "learning_rate": 9.2694207925996e-06,
      "loss": 0.0204,
      "step": 11210
    },
    {
      "epoch": 3.36,
      "learning_rate": 9.258908861557869e-06,
      "loss": 0.0019,
      "step": 11220
    },
    {
      "epoch": 3.36,
      "learning_rate": 9.248396930516137e-06,
      "loss": 0.0017,
      "step": 11230
    },
    {
      "epoch": 3.37,
      "learning_rate": 9.237884999474405e-06,
      "loss": 0.0074,
      "step": 11240
    },
    {
      "epoch": 3.37,
      "learning_rate": 9.227373068432673e-06,
      "loss": 0.0035,
      "step": 11250
    },
    {
      "epoch": 3.37,
      "learning_rate": 9.21686113739094e-06,
      "loss": 0.0015,
      "step": 11260
    },
    {
      "epoch": 3.38,
      "learning_rate": 9.206349206349207e-06,
      "loss": 0.0016,
      "step": 11270
    },
    {
      "epoch": 3.38,
      "learning_rate": 9.195837275307474e-06,
      "loss": 0.0013,
      "step": 11280
    },
    {
      "epoch": 3.38,
      "learning_rate": 9.185325344265742e-06,
      "loss": 0.0011,
      "step": 11290
    },
    {
      "epoch": 3.39,
      "learning_rate": 9.17481341322401e-06,
      "loss": 0.0047,
      "step": 11300
    },
    {
      "epoch": 3.39,
      "learning_rate": 9.164301482182278e-06,
      "loss": 0.0014,
      "step": 11310
    },
    {
      "epoch": 3.39,
      "learning_rate": 9.153789551140544e-06,
      "loss": 0.0013,
      "step": 11320
    },
    {
      "epoch": 3.39,
      "learning_rate": 9.143277620098812e-06,
      "loss": 0.0211,
      "step": 11330
    },
    {
      "epoch": 3.4,
      "learning_rate": 9.13276568905708e-06,
      "loss": 0.0019,
      "step": 11340
    },
    {
      "epoch": 3.4,
      "learning_rate": 9.122253758015348e-06,
      "loss": 0.0041,
      "step": 11350
    },
    {
      "epoch": 3.4,
      "learning_rate": 9.111741826973616e-06,
      "loss": 0.0012,
      "step": 11360
    },
    {
      "epoch": 3.41,
      "learning_rate": 9.101229895931883e-06,
      "loss": 0.0015,
      "step": 11370
    },
    {
      "epoch": 3.41,
      "learning_rate": 9.09071796489015e-06,
      "loss": 0.0011,
      "step": 11380
    },
    {
      "epoch": 3.41,
      "learning_rate": 9.080206033848419e-06,
      "loss": 0.0027,
      "step": 11390
    },
    {
      "epoch": 3.42,
      "learning_rate": 9.069694102806687e-06,
      "loss": 0.0016,
      "step": 11400
    },
    {
      "epoch": 3.42,
      "eval_f1": 0.9938205849207054,
      "eval_f1_0": 0.9980876044839875,
      "eval_f1_1": 0.9983475193620122,
      "eval_f1_2": 0.9802326876427813,
      "eval_f1_3": 0.9986145281940406,
      "eval_loss": 0.014844870194792747,
      "eval_runtime": 247.3455,
      "eval_samples_per_second": 47.981,
      "eval_steps_per_second": 0.19,
      "step": 11400
    },
    {
      "epoch": 3.42,
      "learning_rate": 9.059182171764955e-06,
      "loss": 0.0025,
      "step": 11410
    },
    {
      "epoch": 3.42,
      "learning_rate": 9.048670240723221e-06,
      "loss": 0.0082,
      "step": 11420
    },
    {
      "epoch": 3.42,
      "learning_rate": 9.03815830968149e-06,
      "loss": 0.0011,
      "step": 11430
    },
    {
      "epoch": 3.43,
      "learning_rate": 9.027646378639757e-06,
      "loss": 0.0188,
      "step": 11440
    },
    {
      "epoch": 3.43,
      "learning_rate": 9.017134447598025e-06,
      "loss": 0.001,
      "step": 11450
    },
    {
      "epoch": 3.43,
      "learning_rate": 9.006622516556293e-06,
      "loss": 0.0053,
      "step": 11460
    },
    {
      "epoch": 3.44,
      "learning_rate": 8.99611058551456e-06,
      "loss": 0.0015,
      "step": 11470
    },
    {
      "epoch": 3.44,
      "learning_rate": 8.985598654472826e-06,
      "loss": 0.0054,
      "step": 11480
    },
    {
      "epoch": 3.44,
      "learning_rate": 8.975086723431094e-06,
      "loss": 0.0009,
      "step": 11490
    },
    {
      "epoch": 3.45,
      "learning_rate": 8.964574792389362e-06,
      "loss": 0.001,
      "step": 11500
    },
    {
      "epoch": 3.45,
      "learning_rate": 8.95406286134763e-06,
      "loss": 0.0117,
      "step": 11510
    },
    {
      "epoch": 3.45,
      "learning_rate": 8.943550930305898e-06,
      "loss": 0.0018,
      "step": 11520
    },
    {
      "epoch": 3.45,
      "learning_rate": 8.933038999264165e-06,
      "loss": 0.0014,
      "step": 11530
    },
    {
      "epoch": 3.46,
      "learning_rate": 8.922527068222433e-06,
      "loss": 0.0251,
      "step": 11540
    },
    {
      "epoch": 3.46,
      "learning_rate": 8.9120151371807e-06,
      "loss": 0.0057,
      "step": 11550
    },
    {
      "epoch": 3.46,
      "learning_rate": 8.901503206138969e-06,
      "loss": 0.002,
      "step": 11560
    },
    {
      "epoch": 3.47,
      "learning_rate": 8.890991275097237e-06,
      "loss": 0.0014,
      "step": 11570
    },
    {
      "epoch": 3.47,
      "learning_rate": 8.880479344055503e-06,
      "loss": 0.0046,
      "step": 11580
    },
    {
      "epoch": 3.47,
      "learning_rate": 8.869967413013771e-06,
      "loss": 0.0015,
      "step": 11590
    },
    {
      "epoch": 3.48,
      "learning_rate": 8.859455481972039e-06,
      "loss": 0.0013,
      "step": 11600
    },
    {
      "epoch": 3.48,
      "eval_f1": 0.9941317271995485,
      "eval_f1_0": 0.998170681126581,
      "eval_f1_1": 0.9981401353429832,
      "eval_f1_2": 0.981455226611655,
      "eval_f1_3": 0.9987608657169743,
      "eval_loss": 0.015224499627947807,
      "eval_runtime": 247.2187,
      "eval_samples_per_second": 48.006,
      "eval_steps_per_second": 0.19,
      "step": 11600
    },
    {
      "epoch": 3.48,
      "learning_rate": 8.848943550930307e-06,
      "loss": 0.0009,
      "step": 11610
    },
    {
      "epoch": 3.48,
      "learning_rate": 8.838431619888575e-06,
      "loss": 0.001,
      "step": 11620
    },
    {
      "epoch": 3.48,
      "learning_rate": 8.827919688846842e-06,
      "loss": 0.0027,
      "step": 11630
    },
    {
      "epoch": 3.49,
      "learning_rate": 8.81740775780511e-06,
      "loss": 0.0226,
      "step": 11640
    },
    {
      "epoch": 3.49,
      "learning_rate": 8.806895826763378e-06,
      "loss": 0.0027,
      "step": 11650
    },
    {
      "epoch": 3.49,
      "learning_rate": 8.796383895721646e-06,
      "loss": 0.0018,
      "step": 11660
    },
    {
      "epoch": 3.5,
      "learning_rate": 8.785871964679912e-06,
      "loss": 0.0064,
      "step": 11670
    },
    {
      "epoch": 3.5,
      "learning_rate": 8.77536003363818e-06,
      "loss": 0.0017,
      "step": 11680
    },
    {
      "epoch": 3.5,
      "learning_rate": 8.764848102596446e-06,
      "loss": 0.0029,
      "step": 11690
    },
    {
      "epoch": 3.51,
      "learning_rate": 8.754336171554714e-06,
      "loss": 0.0055,
      "step": 11700
    },
    {
      "epoch": 3.51,
      "learning_rate": 8.743824240512983e-06,
      "loss": 0.0017,
      "step": 11710
    },
    {
      "epoch": 3.51,
      "learning_rate": 8.73331230947125e-06,
      "loss": 0.0089,
      "step": 11720
    },
    {
      "epoch": 3.51,
      "learning_rate": 8.722800378429519e-06,
      "loss": 0.0018,
      "step": 11730
    },
    {
      "epoch": 3.52,
      "learning_rate": 8.712288447387785e-06,
      "loss": 0.0045,
      "step": 11740
    },
    {
      "epoch": 3.52,
      "learning_rate": 8.701776516346053e-06,
      "loss": 0.0018,
      "step": 11750
    },
    {
      "epoch": 3.52,
      "learning_rate": 8.691264585304321e-06,
      "loss": 0.0055,
      "step": 11760
    },
    {
      "epoch": 3.53,
      "learning_rate": 8.680752654262589e-06,
      "loss": 0.0014,
      "step": 11770
    },
    {
      "epoch": 3.53,
      "learning_rate": 8.670240723220857e-06,
      "loss": 0.0022,
      "step": 11780
    },
    {
      "epoch": 3.53,
      "learning_rate": 8.659728792179123e-06,
      "loss": 0.0044,
      "step": 11790
    },
    {
      "epoch": 3.54,
      "learning_rate": 8.649216861137391e-06,
      "loss": 0.0058,
      "step": 11800
    },
    {
      "epoch": 3.54,
      "eval_f1": 0.9939266874515199,
      "eval_f1_0": 0.9981113766177901,
      "eval_f1_1": 0.9984416181868716,
      "eval_f1_2": 0.9804367422181437,
      "eval_f1_3": 0.9987170127832745,
      "eval_loss": 0.01344754733145237,
      "eval_runtime": 247.2355,
      "eval_samples_per_second": 48.003,
      "eval_steps_per_second": 0.19,
      "step": 11800
    },
    {
      "epoch": 3.54,
      "learning_rate": 8.63870493009566e-06,
      "loss": 0.0015,
      "step": 11810
    },
    {
      "epoch": 3.54,
      "learning_rate": 8.628192999053928e-06,
      "loss": 0.0013,
      "step": 11820
    },
    {
      "epoch": 3.54,
      "learning_rate": 8.617681068012194e-06,
      "loss": 0.0015,
      "step": 11830
    },
    {
      "epoch": 3.55,
      "learning_rate": 8.607169136970462e-06,
      "loss": 0.001,
      "step": 11840
    },
    {
      "epoch": 3.55,
      "learning_rate": 8.59665720592873e-06,
      "loss": 0.0018,
      "step": 11850
    },
    {
      "epoch": 3.55,
      "learning_rate": 8.586145274886998e-06,
      "loss": 0.0016,
      "step": 11860
    },
    {
      "epoch": 3.56,
      "learning_rate": 8.575633343845266e-06,
      "loss": 0.0013,
      "step": 11870
    },
    {
      "epoch": 3.56,
      "learning_rate": 8.565121412803532e-06,
      "loss": 0.0015,
      "step": 11880
    },
    {
      "epoch": 3.56,
      "learning_rate": 8.5546094817618e-06,
      "loss": 0.0017,
      "step": 11890
    },
    {
      "epoch": 3.57,
      "learning_rate": 8.544097550720069e-06,
      "loss": 0.0019,
      "step": 11900
    },
    {
      "epoch": 3.57,
      "learning_rate": 8.533585619678335e-06,
      "loss": 0.0017,
      "step": 11910
    },
    {
      "epoch": 3.57,
      "learning_rate": 8.523073688636603e-06,
      "loss": 0.0019,
      "step": 11920
    },
    {
      "epoch": 3.57,
      "learning_rate": 8.512561757594871e-06,
      "loss": 0.0018,
      "step": 11930
    },
    {
      "epoch": 3.58,
      "learning_rate": 8.502049826553139e-06,
      "loss": 0.0011,
      "step": 11940
    },
    {
      "epoch": 3.58,
      "learning_rate": 8.491537895511405e-06,
      "loss": 0.0223,
      "step": 11950
    },
    {
      "epoch": 3.58,
      "learning_rate": 8.481025964469673e-06,
      "loss": 0.0012,
      "step": 11960
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.470514033427941e-06,
      "loss": 0.0016,
      "step": 11970
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.46000210238621e-06,
      "loss": 0.0016,
      "step": 11980
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.449490171344476e-06,
      "loss": 0.0013,
      "step": 11990
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.438978240302744e-06,
      "loss": 0.0011,
      "step": 12000
    },
    {
      "epoch": 3.59,
      "eval_f1": 0.9937027807438935,
      "eval_f1_0": 0.9981029410616087,
      "eval_f1_1": 0.9984551751632033,
      "eval_f1_2": 0.9807739917023737,
      "eval_f1_3": 0.9974790150483887,
      "eval_loss": 0.016883485019207,
      "eval_runtime": 247.1203,
      "eval_samples_per_second": 48.025,
      "eval_steps_per_second": 0.19,
      "step": 12000
    },
    {
      "epoch": 3.6,
      "learning_rate": 8.428466309261012e-06,
      "loss": 0.0132,
      "step": 12010
    },
    {
      "epoch": 3.6,
      "learning_rate": 8.41795437821928e-06,
      "loss": 0.0019,
      "step": 12020
    },
    {
      "epoch": 3.6,
      "learning_rate": 8.407442447177548e-06,
      "loss": 0.0031,
      "step": 12030
    },
    {
      "epoch": 3.61,
      "learning_rate": 8.396930516135814e-06,
      "loss": 0.0019,
      "step": 12040
    },
    {
      "epoch": 3.61,
      "learning_rate": 8.386418585094082e-06,
      "loss": 0.0368,
      "step": 12050
    },
    {
      "epoch": 3.61,
      "learning_rate": 8.37590665405235e-06,
      "loss": 0.0019,
      "step": 12060
    },
    {
      "epoch": 3.62,
      "learning_rate": 8.365394723010618e-06,
      "loss": 0.0034,
      "step": 12070
    },
    {
      "epoch": 3.62,
      "learning_rate": 8.354882791968886e-06,
      "loss": 0.0009,
      "step": 12080
    },
    {
      "epoch": 3.62,
      "learning_rate": 8.344370860927153e-06,
      "loss": 0.0022,
      "step": 12090
    },
    {
      "epoch": 3.62,
      "learning_rate": 8.33385892988542e-06,
      "loss": 0.0006,
      "step": 12100
    },
    {
      "epoch": 3.63,
      "learning_rate": 8.323346998843689e-06,
      "loss": 0.0031,
      "step": 12110
    },
    {
      "epoch": 3.63,
      "learning_rate": 8.312835067801955e-06,
      "loss": 0.0018,
      "step": 12120
    },
    {
      "epoch": 3.63,
      "learning_rate": 8.302323136760223e-06,
      "loss": 0.0014,
      "step": 12130
    },
    {
      "epoch": 3.64,
      "learning_rate": 8.291811205718491e-06,
      "loss": 0.0026,
      "step": 12140
    },
    {
      "epoch": 3.64,
      "learning_rate": 8.281299274676758e-06,
      "loss": 0.0031,
      "step": 12150
    },
    {
      "epoch": 3.64,
      "learning_rate": 8.270787343635026e-06,
      "loss": 0.0071,
      "step": 12160
    },
    {
      "epoch": 3.65,
      "learning_rate": 8.260275412593294e-06,
      "loss": 0.0011,
      "step": 12170
    },
    {
      "epoch": 3.65,
      "learning_rate": 8.249763481551562e-06,
      "loss": 0.0164,
      "step": 12180
    },
    {
      "epoch": 3.65,
      "learning_rate": 8.23925155050983e-06,
      "loss": 0.0009,
      "step": 12190
    },
    {
      "epoch": 3.65,
      "learning_rate": 8.228739619468096e-06,
      "loss": 0.0236,
      "step": 12200
    },
    {
      "epoch": 3.65,
      "eval_f1": 0.9939173696692444,
      "eval_f1_0": 0.9980961276283965,
      "eval_f1_1": 0.9980212068337392,
      "eval_f1_2": 0.9806423383687922,
      "eval_f1_3": 0.9989098058460497,
      "eval_loss": 0.013924673199653625,
      "eval_runtime": 246.9834,
      "eval_samples_per_second": 48.052,
      "eval_steps_per_second": 0.19,
      "step": 12200
    },
    {
      "epoch": 3.66,
      "learning_rate": 8.218227688426364e-06,
      "loss": 0.007,
      "step": 12210
    },
    {
      "epoch": 3.66,
      "learning_rate": 8.207715757384632e-06,
      "loss": 0.0177,
      "step": 12220
    },
    {
      "epoch": 3.66,
      "learning_rate": 8.1972038263429e-06,
      "loss": 0.0022,
      "step": 12230
    },
    {
      "epoch": 3.67,
      "learning_rate": 8.186691895301168e-06,
      "loss": 0.0223,
      "step": 12240
    },
    {
      "epoch": 3.67,
      "learning_rate": 8.176179964259435e-06,
      "loss": 0.0046,
      "step": 12250
    },
    {
      "epoch": 3.67,
      "learning_rate": 8.165668033217703e-06,
      "loss": 0.0014,
      "step": 12260
    },
    {
      "epoch": 3.68,
      "learning_rate": 8.15515610217597e-06,
      "loss": 0.0016,
      "step": 12270
    },
    {
      "epoch": 3.68,
      "learning_rate": 8.144644171134239e-06,
      "loss": 0.0056,
      "step": 12280
    },
    {
      "epoch": 3.68,
      "learning_rate": 8.134132240092507e-06,
      "loss": 0.0024,
      "step": 12290
    },
    {
      "epoch": 3.68,
      "learning_rate": 8.123620309050773e-06,
      "loss": 0.0011,
      "step": 12300
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.113108378009041e-06,
      "loss": 0.0012,
      "step": 12310
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.10259644696731e-06,
      "loss": 0.0152,
      "step": 12320
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.092084515925576e-06,
      "loss": 0.0012,
      "step": 12330
    },
    {
      "epoch": 3.7,
      "learning_rate": 8.081572584883844e-06,
      "loss": 0.001,
      "step": 12340
    },
    {
      "epoch": 3.7,
      "learning_rate": 8.071060653842112e-06,
      "loss": 0.0011,
      "step": 12350
    },
    {
      "epoch": 3.7,
      "learning_rate": 8.060548722800378e-06,
      "loss": 0.0022,
      "step": 12360
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.050036791758646e-06,
      "loss": 0.0028,
      "step": 12370
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.039524860716914e-06,
      "loss": 0.0016,
      "step": 12380
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.029012929675182e-06,
      "loss": 0.0019,
      "step": 12390
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.01850099863345e-06,
      "loss": 0.0021,
      "step": 12400
    },
    {
      "epoch": 3.71,
      "eval_f1": 0.9942686517328441,
      "eval_f1_0": 0.998221730721652,
      "eval_f1_1": 0.9982992903803493,
      "eval_f1_2": 0.9816252440418489,
      "eval_f1_3": 0.9989283417875259,
      "eval_loss": 0.013739947229623795,
      "eval_runtime": 247.2439,
      "eval_samples_per_second": 48.001,
      "eval_steps_per_second": 0.19,
      "step": 12400
    },
    {
      "epoch": 3.72,
      "learning_rate": 8.007989067591717e-06,
      "loss": 0.0016,
      "step": 12410
    },
    {
      "epoch": 3.72,
      "learning_rate": 7.997477136549985e-06,
      "loss": 0.0024,
      "step": 12420
    },
    {
      "epoch": 3.72,
      "learning_rate": 7.986965205508253e-06,
      "loss": 0.0146,
      "step": 12430
    },
    {
      "epoch": 3.73,
      "learning_rate": 7.97645327446652e-06,
      "loss": 0.001,
      "step": 12440
    },
    {
      "epoch": 3.73,
      "learning_rate": 7.965941343424789e-06,
      "loss": 0.0009,
      "step": 12450
    },
    {
      "epoch": 3.73,
      "learning_rate": 7.955429412383055e-06,
      "loss": 0.0056,
      "step": 12460
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.944917481341323e-06,
      "loss": 0.0011,
      "step": 12470
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.934405550299591e-06,
      "loss": 0.001,
      "step": 12480
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.92389361925786e-06,
      "loss": 0.0014,
      "step": 12490
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.913381688216126e-06,
      "loss": 0.0007,
      "step": 12500
    },
    {
      "epoch": 3.75,
      "learning_rate": 7.902869757174394e-06,
      "loss": 0.001,
      "step": 12510
    },
    {
      "epoch": 3.75,
      "learning_rate": 7.892357826132662e-06,
      "loss": 0.0044,
      "step": 12520
    },
    {
      "epoch": 3.75,
      "learning_rate": 7.88184589509093e-06,
      "loss": 0.0007,
      "step": 12530
    },
    {
      "epoch": 3.76,
      "learning_rate": 7.871333964049198e-06,
      "loss": 0.0006,
      "step": 12540
    },
    {
      "epoch": 3.76,
      "learning_rate": 7.860822033007464e-06,
      "loss": 0.0019,
      "step": 12550
    },
    {
      "epoch": 3.76,
      "learning_rate": 7.850310101965732e-06,
      "loss": 0.0014,
      "step": 12560
    },
    {
      "epoch": 3.77,
      "learning_rate": 7.839798170923998e-06,
      "loss": 0.0024,
      "step": 12570
    },
    {
      "epoch": 3.77,
      "learning_rate": 7.829286239882266e-06,
      "loss": 0.0013,
      "step": 12580
    },
    {
      "epoch": 3.77,
      "learning_rate": 7.818774308840534e-06,
      "loss": 0.007,
      "step": 12590
    },
    {
      "epoch": 3.77,
      "learning_rate": 7.808262377798803e-06,
      "loss": 0.0095,
      "step": 12600
    },
    {
      "epoch": 3.77,
      "eval_f1": 0.9945213646234112,
      "eval_f1_0": 0.9983082790074349,
      "eval_f1_1": 0.9982366898835832,
      "eval_f1_2": 0.9827421477045036,
      "eval_f1_3": 0.998798341898123,
      "eval_loss": 0.015260514803230762,
      "eval_runtime": 247.3084,
      "eval_samples_per_second": 47.989,
      "eval_steps_per_second": 0.19,
      "step": 12600
    },
    {
      "epoch": 3.78,
      "learning_rate": 7.797750446757069e-06,
      "loss": 0.002,
      "step": 12610
    },
    {
      "epoch": 3.78,
      "learning_rate": 7.787238515715337e-06,
      "loss": 0.003,
      "step": 12620
    },
    {
      "epoch": 3.78,
      "learning_rate": 7.776726584673605e-06,
      "loss": 0.002,
      "step": 12630
    },
    {
      "epoch": 3.79,
      "learning_rate": 7.766214653631873e-06,
      "loss": 0.0022,
      "step": 12640
    },
    {
      "epoch": 3.79,
      "learning_rate": 7.755702722590141e-06,
      "loss": 0.0013,
      "step": 12650
    },
    {
      "epoch": 3.79,
      "learning_rate": 7.745190791548407e-06,
      "loss": 0.0012,
      "step": 12660
    },
    {
      "epoch": 3.8,
      "learning_rate": 7.734678860506675e-06,
      "loss": 0.0018,
      "step": 12670
    },
    {
      "epoch": 3.8,
      "learning_rate": 7.724166929464943e-06,
      "loss": 0.002,
      "step": 12680
    },
    {
      "epoch": 3.8,
      "learning_rate": 7.713654998423212e-06,
      "loss": 0.0012,
      "step": 12690
    },
    {
      "epoch": 3.8,
      "learning_rate": 7.70314306738148e-06,
      "loss": 0.0029,
      "step": 12700
    },
    {
      "epoch": 3.81,
      "learning_rate": 7.692631136339746e-06,
      "loss": 0.0015,
      "step": 12710
    },
    {
      "epoch": 3.81,
      "learning_rate": 7.682119205298014e-06,
      "loss": 0.0019,
      "step": 12720
    },
    {
      "epoch": 3.81,
      "learning_rate": 7.671607274256282e-06,
      "loss": 0.0074,
      "step": 12730
    },
    {
      "epoch": 3.82,
      "learning_rate": 7.66109534321455e-06,
      "loss": 0.0008,
      "step": 12740
    },
    {
      "epoch": 3.82,
      "learning_rate": 7.650583412172818e-06,
      "loss": 0.0017,
      "step": 12750
    },
    {
      "epoch": 3.82,
      "learning_rate": 7.640071481131084e-06,
      "loss": 0.0017,
      "step": 12760
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.62955955008935e-06,
      "loss": 0.0009,
      "step": 12770
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.61904761904762e-06,
      "loss": 0.0013,
      "step": 12780
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.608535688005887e-06,
      "loss": 0.0063,
      "step": 12790
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.598023756964155e-06,
      "loss": 0.0023,
      "step": 12800
    },
    {
      "epoch": 3.83,
      "eval_f1": 0.993694587654125,
      "eval_f1_0": 0.998020563621245,
      "eval_f1_1": 0.9978318348049995,
      "eval_f1_2": 0.9799570721355936,
      "eval_f1_3": 0.9989688800546618,
      "eval_loss": 0.014967081137001514,
      "eval_runtime": 247.0896,
      "eval_samples_per_second": 48.031,
      "eval_steps_per_second": 0.19,
      "step": 12800
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.587511825922422e-06,
      "loss": 0.0047,
      "step": 12810
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.57699989488069e-06,
      "loss": 0.001,
      "step": 12820
    },
    {
      "epoch": 3.84,
      "learning_rate": 7.566487963838957e-06,
      "loss": 0.0015,
      "step": 12830
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.555976032797225e-06,
      "loss": 0.0015,
      "step": 12840
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.545464101755493e-06,
      "loss": 0.0012,
      "step": 12850
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.5349521707137606e-06,
      "loss": 0.0009,
      "step": 12860
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.524440239672029e-06,
      "loss": 0.0029,
      "step": 12870
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.513928308630296e-06,
      "loss": 0.0027,
      "step": 12880
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.503416377588564e-06,
      "loss": 0.0065,
      "step": 12890
    },
    {
      "epoch": 3.86,
      "learning_rate": 7.492904446546832e-06,
      "loss": 0.0012,
      "step": 12900
    },
    {
      "epoch": 3.87,
      "learning_rate": 7.482392515505099e-06,
      "loss": 0.0015,
      "step": 12910
    },
    {
      "epoch": 3.87,
      "learning_rate": 7.471880584463367e-06,
      "loss": 0.0018,
      "step": 12920
    },
    {
      "epoch": 3.87,
      "learning_rate": 7.461368653421634e-06,
      "loss": 0.0023,
      "step": 12930
    },
    {
      "epoch": 3.88,
      "learning_rate": 7.450856722379902e-06,
      "loss": 0.0016,
      "step": 12940
    },
    {
      "epoch": 3.88,
      "learning_rate": 7.4403447913381695e-06,
      "loss": 0.0235,
      "step": 12950
    },
    {
      "epoch": 3.88,
      "learning_rate": 7.429832860296438e-06,
      "loss": 0.001,
      "step": 12960
    },
    {
      "epoch": 3.89,
      "learning_rate": 7.419320929254704e-06,
      "loss": 0.0017,
      "step": 12970
    },
    {
      "epoch": 3.89,
      "learning_rate": 7.408808998212972e-06,
      "loss": 0.0034,
      "step": 12980
    },
    {
      "epoch": 3.89,
      "learning_rate": 7.398297067171239e-06,
      "loss": 0.0014,
      "step": 12990
    },
    {
      "epoch": 3.89,
      "learning_rate": 7.387785136129507e-06,
      "loss": 0.0159,
      "step": 13000
    },
    {
      "epoch": 3.89,
      "eval_f1": 0.994513118134118,
      "eval_f1_0": 0.9982883700335918,
      "eval_f1_1": 0.9979322118427867,
      "eval_f1_2": 0.9828598505479184,
      "eval_f1_3": 0.998972040112175,
      "eval_loss": 0.013586748391389847,
      "eval_runtime": 247.2347,
      "eval_samples_per_second": 48.003,
      "eval_steps_per_second": 0.19,
      "step": 13000
    },
    {
      "epoch": 3.9,
      "learning_rate": 7.377273205087775e-06,
      "loss": 0.0019,
      "step": 13010
    },
    {
      "epoch": 3.9,
      "learning_rate": 7.3667612740460424e-06,
      "loss": 0.0016,
      "step": 13020
    },
    {
      "epoch": 3.9,
      "learning_rate": 7.3562493430043105e-06,
      "loss": 0.007,
      "step": 13030
    },
    {
      "epoch": 3.91,
      "learning_rate": 7.345737411962578e-06,
      "loss": 0.0012,
      "step": 13040
    },
    {
      "epoch": 3.91,
      "learning_rate": 7.335225480920846e-06,
      "loss": 0.0027,
      "step": 13050
    },
    {
      "epoch": 3.91,
      "learning_rate": 7.324713549879113e-06,
      "loss": 0.0038,
      "step": 13060
    },
    {
      "epoch": 3.92,
      "learning_rate": 7.314201618837381e-06,
      "loss": 0.0026,
      "step": 13070
    },
    {
      "epoch": 3.92,
      "learning_rate": 7.303689687795649e-06,
      "loss": 0.0013,
      "step": 13080
    },
    {
      "epoch": 3.92,
      "learning_rate": 7.293177756753916e-06,
      "loss": 0.0023,
      "step": 13090
    },
    {
      "epoch": 3.92,
      "learning_rate": 7.282665825712184e-06,
      "loss": 0.0017,
      "step": 13100
    },
    {
      "epoch": 3.93,
      "learning_rate": 7.272153894670451e-06,
      "loss": 0.001,
      "step": 13110
    },
    {
      "epoch": 3.93,
      "learning_rate": 7.2616419636287194e-06,
      "loss": 0.0024,
      "step": 13120
    },
    {
      "epoch": 3.93,
      "learning_rate": 7.2511300325869875e-06,
      "loss": 0.0157,
      "step": 13130
    },
    {
      "epoch": 3.94,
      "learning_rate": 7.240618101545255e-06,
      "loss": 0.0023,
      "step": 13140
    },
    {
      "epoch": 3.94,
      "learning_rate": 7.230106170503523e-06,
      "loss": 0.0015,
      "step": 13150
    },
    {
      "epoch": 3.94,
      "learning_rate": 7.21959423946179e-06,
      "loss": 0.0019,
      "step": 13160
    },
    {
      "epoch": 3.95,
      "learning_rate": 7.209082308420058e-06,
      "loss": 0.0042,
      "step": 13170
    },
    {
      "epoch": 3.95,
      "learning_rate": 7.198570377378325e-06,
      "loss": 0.0065,
      "step": 13180
    },
    {
      "epoch": 3.95,
      "learning_rate": 7.188058446336592e-06,
      "loss": 0.0007,
      "step": 13190
    },
    {
      "epoch": 3.95,
      "learning_rate": 7.1775465152948595e-06,
      "loss": 0.0011,
      "step": 13200
    },
    {
      "epoch": 3.95,
      "eval_f1": 0.9939547121650559,
      "eval_f1_0": 0.9981259071571787,
      "eval_f1_1": 0.9981306691054245,
      "eval_f1_2": 0.9812964734530591,
      "eval_f1_3": 0.9982657989445615,
      "eval_loss": 0.014702709391713142,
      "eval_runtime": 247.143,
      "eval_samples_per_second": 48.021,
      "eval_steps_per_second": 0.19,
      "step": 13200
    },
    {
      "epoch": 3.96,
      "learning_rate": 7.167034584253128e-06,
      "loss": 0.0015,
      "step": 13210
    },
    {
      "epoch": 3.96,
      "learning_rate": 7.156522653211395e-06,
      "loss": 0.0013,
      "step": 13220
    },
    {
      "epoch": 3.96,
      "learning_rate": 7.146010722169663e-06,
      "loss": 0.0016,
      "step": 13230
    },
    {
      "epoch": 3.97,
      "learning_rate": 7.135498791127931e-06,
      "loss": 0.0028,
      "step": 13240
    },
    {
      "epoch": 3.97,
      "learning_rate": 7.124986860086198e-06,
      "loss": 0.0015,
      "step": 13250
    },
    {
      "epoch": 3.97,
      "learning_rate": 7.114474929044466e-06,
      "loss": 0.0011,
      "step": 13260
    },
    {
      "epoch": 3.98,
      "learning_rate": 7.103962998002733e-06,
      "loss": 0.002,
      "step": 13270
    },
    {
      "epoch": 3.98,
      "learning_rate": 7.093451066961001e-06,
      "loss": 0.0096,
      "step": 13280
    },
    {
      "epoch": 3.98,
      "learning_rate": 7.082939135919269e-06,
      "loss": 0.0056,
      "step": 13290
    },
    {
      "epoch": 3.98,
      "learning_rate": 7.0724272048775366e-06,
      "loss": 0.0016,
      "step": 13300
    },
    {
      "epoch": 3.99,
      "learning_rate": 7.061915273835805e-06,
      "loss": 0.0024,
      "step": 13310
    },
    {
      "epoch": 3.99,
      "learning_rate": 7.051403342794072e-06,
      "loss": 0.0168,
      "step": 13320
    },
    {
      "epoch": 3.99,
      "learning_rate": 7.04089141175234e-06,
      "loss": 0.0017,
      "step": 13330
    },
    {
      "epoch": 4.0,
      "learning_rate": 7.030379480710607e-06,
      "loss": 0.0022,
      "step": 13340
    },
    {
      "epoch": 4.0,
      "learning_rate": 7.019867549668875e-06,
      "loss": 0.001,
      "step": 13350
    },
    {
      "epoch": 4.0,
      "learning_rate": 7.009355618627143e-06,
      "loss": 0.0013,
      "step": 13360
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.99884368758541e-06,
      "loss": 0.001,
      "step": 13370
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.988331756543678e-06,
      "loss": 0.0013,
      "step": 13380
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.9778198255019455e-06,
      "loss": 0.0025,
      "step": 13390
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.967307894460213e-06,
      "loss": 0.0008,
      "step": 13400
    },
    {
      "epoch": 4.01,
      "eval_f1": 0.9931913568988674,
      "eval_f1_0": 0.9979095964146147,
      "eval_f1_1": 0.9985399445868733,
      "eval_f1_2": 0.9773470065393196,
      "eval_f1_3": 0.9989688800546618,
      "eval_loss": 0.01953274942934513,
      "eval_runtime": 247.1321,
      "eval_samples_per_second": 48.023,
      "eval_steps_per_second": 0.19,
      "step": 13400
    },
    {
      "epoch": 4.02,
      "learning_rate": 6.95679596341848e-06,
      "loss": 0.0008,
      "step": 13410
    },
    {
      "epoch": 4.02,
      "learning_rate": 6.946284032376748e-06,
      "loss": 0.0017,
      "step": 13420
    },
    {
      "epoch": 4.02,
      "learning_rate": 6.935772101335015e-06,
      "loss": 0.0014,
      "step": 13430
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.925260170293283e-06,
      "loss": 0.0008,
      "step": 13440
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.914748239251551e-06,
      "loss": 0.0006,
      "step": 13450
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.904236308209818e-06,
      "loss": 0.0008,
      "step": 13460
    },
    {
      "epoch": 4.04,
      "learning_rate": 6.8937243771680865e-06,
      "loss": 0.0023,
      "step": 13470
    },
    {
      "epoch": 4.04,
      "learning_rate": 6.883212446126354e-06,
      "loss": 0.0027,
      "step": 13480
    },
    {
      "epoch": 4.04,
      "learning_rate": 6.872700515084622e-06,
      "loss": 0.0007,
      "step": 13490
    },
    {
      "epoch": 4.04,
      "learning_rate": 6.862188584042889e-06,
      "loss": 0.0014,
      "step": 13500
    },
    {
      "epoch": 4.05,
      "learning_rate": 6.851676653001157e-06,
      "loss": 0.0008,
      "step": 13510
    },
    {
      "epoch": 4.05,
      "learning_rate": 6.841164721959425e-06,
      "loss": 0.0015,
      "step": 13520
    },
    {
      "epoch": 4.05,
      "learning_rate": 6.830652790917692e-06,
      "loss": 0.0008,
      "step": 13530
    },
    {
      "epoch": 4.06,
      "learning_rate": 6.82014085987596e-06,
      "loss": 0.0018,
      "step": 13540
    },
    {
      "epoch": 4.06,
      "learning_rate": 6.809628928834227e-06,
      "loss": 0.0015,
      "step": 13550
    },
    {
      "epoch": 4.06,
      "learning_rate": 6.7991169977924954e-06,
      "loss": 0.0013,
      "step": 13560
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.788605066750763e-06,
      "loss": 0.001,
      "step": 13570
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.778093135709031e-06,
      "loss": 0.0011,
      "step": 13580
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.767581204667299e-06,
      "loss": 0.0011,
      "step": 13590
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.757069273625566e-06,
      "loss": 0.0011,
      "step": 13600
    },
    {
      "epoch": 4.07,
      "eval_f1": 0.9941192617742906,
      "eval_f1_0": 0.998165200865003,
      "eval_f1_1": 0.9982913252190436,
      "eval_f1_2": 0.9812346521330658,
      "eval_f1_3": 0.9987858688800494,
      "eval_loss": 0.015765877440571785,
      "eval_runtime": 247.065,
      "eval_samples_per_second": 48.036,
      "eval_steps_per_second": 0.19,
      "step": 13600
    },
    {
      "epoch": 4.08,
      "learning_rate": 6.746557342583832e-06,
      "loss": 0.001,
      "step": 13610
    },
    {
      "epoch": 4.08,
      "learning_rate": 6.7360454115421e-06,
      "loss": 0.0109,
      "step": 13620
    },
    {
      "epoch": 4.08,
      "learning_rate": 6.725533480500368e-06,
      "loss": 0.0018,
      "step": 13630
    },
    {
      "epoch": 4.09,
      "learning_rate": 6.7150215494586355e-06,
      "loss": 0.0168,
      "step": 13640
    },
    {
      "epoch": 4.09,
      "learning_rate": 6.7045096184169036e-06,
      "loss": 0.0078,
      "step": 13650
    },
    {
      "epoch": 4.09,
      "learning_rate": 6.693997687375171e-06,
      "loss": 0.0012,
      "step": 13660
    },
    {
      "epoch": 4.1,
      "learning_rate": 6.683485756333439e-06,
      "loss": 0.0032,
      "step": 13670
    },
    {
      "epoch": 4.1,
      "learning_rate": 6.672973825291707e-06,
      "loss": 0.001,
      "step": 13680
    },
    {
      "epoch": 4.1,
      "learning_rate": 6.662461894249974e-06,
      "loss": 0.0008,
      "step": 13690
    },
    {
      "epoch": 4.1,
      "learning_rate": 6.651949963208242e-06,
      "loss": 0.0009,
      "step": 13700
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.641438032166509e-06,
      "loss": 0.0006,
      "step": 13710
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.630926101124777e-06,
      "loss": 0.0011,
      "step": 13720
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.6204141700830445e-06,
      "loss": 0.0018,
      "step": 13730
    },
    {
      "epoch": 4.12,
      "learning_rate": 6.6099022390413125e-06,
      "loss": 0.001,
      "step": 13740
    },
    {
      "epoch": 4.12,
      "learning_rate": 6.599390307999581e-06,
      "loss": 0.0006,
      "step": 13750
    },
    {
      "epoch": 4.12,
      "learning_rate": 6.588878376957848e-06,
      "loss": 0.0012,
      "step": 13760
    },
    {
      "epoch": 4.13,
      "learning_rate": 6.578366445916116e-06,
      "loss": 0.0012,
      "step": 13770
    },
    {
      "epoch": 4.13,
      "learning_rate": 6.567854514874383e-06,
      "loss": 0.0016,
      "step": 13780
    },
    {
      "epoch": 4.13,
      "learning_rate": 6.557342583832651e-06,
      "loss": 0.001,
      "step": 13790
    },
    {
      "epoch": 4.13,
      "learning_rate": 6.546830652790919e-06,
      "loss": 0.001,
      "step": 13800
    },
    {
      "epoch": 4.13,
      "eval_f1": 0.9927719235604164,
      "eval_f1_0": 0.9977020534509167,
      "eval_f1_1": 0.9964359342367424,
      "eval_f1_2": 0.9780585502815855,
      "eval_f1_3": 0.9988911562724214,
      "eval_loss": 0.01883442886173725,
      "eval_runtime": 247.4624,
      "eval_samples_per_second": 47.959,
      "eval_steps_per_second": 0.19,
      "step": 13800
    },
    {
      "epoch": 4.14,
      "learning_rate": 6.536318721749186e-06,
      "loss": 0.0116,
      "step": 13810
    },
    {
      "epoch": 4.14,
      "learning_rate": 6.525806790707454e-06,
      "loss": 0.001,
      "step": 13820
    },
    {
      "epoch": 4.14,
      "learning_rate": 6.515294859665721e-06,
      "loss": 0.0017,
      "step": 13830
    },
    {
      "epoch": 4.15,
      "learning_rate": 6.504782928623989e-06,
      "loss": 0.0013,
      "step": 13840
    },
    {
      "epoch": 4.15,
      "learning_rate": 6.494270997582256e-06,
      "loss": 0.001,
      "step": 13850
    },
    {
      "epoch": 4.15,
      "learning_rate": 6.483759066540524e-06,
      "loss": 0.0017,
      "step": 13860
    },
    {
      "epoch": 4.16,
      "learning_rate": 6.473247135498791e-06,
      "loss": 0.0011,
      "step": 13870
    },
    {
      "epoch": 4.16,
      "learning_rate": 6.462735204457059e-06,
      "loss": 0.0076,
      "step": 13880
    },
    {
      "epoch": 4.16,
      "learning_rate": 6.452223273415326e-06,
      "loss": 0.0013,
      "step": 13890
    },
    {
      "epoch": 4.16,
      "learning_rate": 6.441711342373594e-06,
      "loss": 0.0084,
      "step": 13900
    },
    {
      "epoch": 4.17,
      "learning_rate": 6.4311994113318625e-06,
      "loss": 0.001,
      "step": 13910
    },
    {
      "epoch": 4.17,
      "learning_rate": 6.42068748029013e-06,
      "loss": 0.001,
      "step": 13920
    },
    {
      "epoch": 4.17,
      "learning_rate": 6.410175549248398e-06,
      "loss": 0.0007,
      "step": 13930
    },
    {
      "epoch": 4.18,
      "learning_rate": 6.399663618206665e-06,
      "loss": 0.0016,
      "step": 13940
    },
    {
      "epoch": 4.18,
      "learning_rate": 6.389151687164933e-06,
      "loss": 0.0011,
      "step": 13950
    },
    {
      "epoch": 4.18,
      "learning_rate": 6.3786397561232e-06,
      "loss": 0.0006,
      "step": 13960
    },
    {
      "epoch": 4.19,
      "learning_rate": 6.368127825081468e-06,
      "loss": 0.001,
      "step": 13970
    },
    {
      "epoch": 4.19,
      "learning_rate": 6.357615894039736e-06,
      "loss": 0.0057,
      "step": 13980
    },
    {
      "epoch": 4.19,
      "learning_rate": 6.347103962998003e-06,
      "loss": 0.0103,
      "step": 13990
    },
    {
      "epoch": 4.19,
      "learning_rate": 6.3365920319562714e-06,
      "loss": 0.001,
      "step": 14000
    },
    {
      "epoch": 4.19,
      "eval_f1": 0.9936074540372482,
      "eval_f1_0": 0.9980276741115003,
      "eval_f1_1": 0.9983116427658396,
      "eval_f1_2": 0.9792301694194372,
      "eval_f1_3": 0.9988603298522155,
      "eval_loss": 0.018208958208560944,
      "eval_runtime": 247.0284,
      "eval_samples_per_second": 48.043,
      "eval_steps_per_second": 0.19,
      "step": 14000
    },
    {
      "epoch": 4.2,
      "learning_rate": 6.326080100914539e-06,
      "loss": 0.0014,
      "step": 14010
    },
    {
      "epoch": 4.2,
      "learning_rate": 6.315568169872807e-06,
      "loss": 0.001,
      "step": 14020
    },
    {
      "epoch": 4.2,
      "learning_rate": 6.305056238831075e-06,
      "loss": 0.0015,
      "step": 14030
    },
    {
      "epoch": 4.21,
      "learning_rate": 6.294544307789341e-06,
      "loss": 0.0019,
      "step": 14040
    },
    {
      "epoch": 4.21,
      "learning_rate": 6.284032376747608e-06,
      "loss": 0.001,
      "step": 14050
    },
    {
      "epoch": 4.21,
      "learning_rate": 6.273520445705876e-06,
      "loss": 0.001,
      "step": 14060
    },
    {
      "epoch": 4.22,
      "learning_rate": 6.263008514664144e-06,
      "loss": 0.0013,
      "step": 14070
    },
    {
      "epoch": 4.22,
      "learning_rate": 6.2524965836224115e-06,
      "loss": 0.0025,
      "step": 14080
    },
    {
      "epoch": 4.22,
      "learning_rate": 6.2419846525806796e-06,
      "loss": 0.0011,
      "step": 14090
    },
    {
      "epoch": 4.22,
      "learning_rate": 6.231472721538947e-06,
      "loss": 0.0013,
      "step": 14100
    },
    {
      "epoch": 4.23,
      "learning_rate": 6.220960790497215e-06,
      "loss": 0.0008,
      "step": 14110
    },
    {
      "epoch": 4.23,
      "learning_rate": 6.210448859455482e-06,
      "loss": 0.0009,
      "step": 14120
    },
    {
      "epoch": 4.23,
      "learning_rate": 6.19993692841375e-06,
      "loss": 0.0011,
      "step": 14130
    },
    {
      "epoch": 4.24,
      "learning_rate": 6.189424997372018e-06,
      "loss": 0.0022,
      "step": 14140
    },
    {
      "epoch": 4.24,
      "learning_rate": 6.178913066330285e-06,
      "loss": 0.0023,
      "step": 14150
    },
    {
      "epoch": 4.24,
      "learning_rate": 6.168401135288553e-06,
      "loss": 0.0015,
      "step": 14160
    },
    {
      "epoch": 4.25,
      "learning_rate": 6.1578892042468205e-06,
      "loss": 0.0009,
      "step": 14170
    },
    {
      "epoch": 4.25,
      "learning_rate": 6.1473772732050885e-06,
      "loss": 0.0146,
      "step": 14180
    },
    {
      "epoch": 4.25,
      "learning_rate": 6.1368653421633566e-06,
      "loss": 0.0013,
      "step": 14190
    },
    {
      "epoch": 4.25,
      "learning_rate": 6.126353411121624e-06,
      "loss": 0.0011,
      "step": 14200
    },
    {
      "epoch": 4.25,
      "eval_f1": 0.9912374633941842,
      "eval_f1_0": 0.9973362751888736,
      "eval_f1_1": 0.9982965987541603,
      "eval_f1_2": 0.9704163653441726,
      "eval_f1_3": 0.9989006142895297,
      "eval_loss": 0.025182632729411125,
      "eval_runtime": 247.1593,
      "eval_samples_per_second": 48.018,
      "eval_steps_per_second": 0.19,
      "step": 14200
    },
    {
      "epoch": 4.26,
      "learning_rate": 6.115841480079892e-06,
      "loss": 0.0018,
      "step": 14210
    },
    {
      "epoch": 4.26,
      "learning_rate": 6.105329549038159e-06,
      "loss": 0.005,
      "step": 14220
    },
    {
      "epoch": 4.26,
      "learning_rate": 6.094817617996427e-06,
      "loss": 0.0013,
      "step": 14230
    },
    {
      "epoch": 4.27,
      "learning_rate": 6.084305686954694e-06,
      "loss": 0.0032,
      "step": 14240
    },
    {
      "epoch": 4.27,
      "learning_rate": 6.073793755912961e-06,
      "loss": 0.0011,
      "step": 14250
    },
    {
      "epoch": 4.27,
      "learning_rate": 6.063281824871229e-06,
      "loss": 0.0013,
      "step": 14260
    },
    {
      "epoch": 4.28,
      "learning_rate": 6.052769893829497e-06,
      "loss": 0.0008,
      "step": 14270
    },
    {
      "epoch": 4.28,
      "learning_rate": 6.042257962787764e-06,
      "loss": 0.0031,
      "step": 14280
    },
    {
      "epoch": 4.28,
      "learning_rate": 6.031746031746032e-06,
      "loss": 0.0015,
      "step": 14290
    },
    {
      "epoch": 4.28,
      "learning_rate": 6.0212341007043e-06,
      "loss": 0.0017,
      "step": 14300
    },
    {
      "epoch": 4.29,
      "learning_rate": 6.010722169662567e-06,
      "loss": 0.0013,
      "step": 14310
    },
    {
      "epoch": 4.29,
      "learning_rate": 6.000210238620835e-06,
      "loss": 0.0018,
      "step": 14320
    },
    {
      "epoch": 4.29,
      "learning_rate": 5.989698307579102e-06,
      "loss": 0.0016,
      "step": 14330
    },
    {
      "epoch": 4.3,
      "learning_rate": 5.97918637653737e-06,
      "loss": 0.0011,
      "step": 14340
    },
    {
      "epoch": 4.3,
      "learning_rate": 5.968674445495638e-06,
      "loss": 0.0015,
      "step": 14350
    },
    {
      "epoch": 4.3,
      "learning_rate": 5.958162514453906e-06,
      "loss": 0.0014,
      "step": 14360
    },
    {
      "epoch": 4.3,
      "learning_rate": 5.947650583412174e-06,
      "loss": 0.0012,
      "step": 14370
    },
    {
      "epoch": 4.31,
      "learning_rate": 5.937138652370441e-06,
      "loss": 0.0012,
      "step": 14380
    },
    {
      "epoch": 4.31,
      "learning_rate": 5.926626721328709e-06,
      "loss": 0.0008,
      "step": 14390
    },
    {
      "epoch": 4.31,
      "learning_rate": 5.916114790286976e-06,
      "loss": 0.0006,
      "step": 14400
    },
    {
      "epoch": 4.31,
      "eval_f1": 0.9933210575730853,
      "eval_f1_0": 0.9979401105448755,
      "eval_f1_1": 0.9984010307025112,
      "eval_f1_2": 0.9779711319432276,
      "eval_f1_3": 0.9989719571017266,
      "eval_loss": 0.01996074803173542,
      "eval_runtime": 247.3247,
      "eval_samples_per_second": 47.986,
      "eval_steps_per_second": 0.19,
      "step": 14400
    },
    {
      "epoch": 4.32,
      "learning_rate": 5.905602859245244e-06,
      "loss": 0.0008,
      "step": 14410
    },
    {
      "epoch": 4.32,
      "learning_rate": 5.895090928203512e-06,
      "loss": 0.0011,
      "step": 14420
    },
    {
      "epoch": 4.32,
      "learning_rate": 5.884578997161779e-06,
      "loss": 0.0007,
      "step": 14430
    },
    {
      "epoch": 4.33,
      "learning_rate": 5.874067066120047e-06,
      "loss": 0.0008,
      "step": 14440
    },
    {
      "epoch": 4.33,
      "learning_rate": 5.863555135078315e-06,
      "loss": 0.0006,
      "step": 14450
    },
    {
      "epoch": 4.33,
      "learning_rate": 5.853043204036583e-06,
      "loss": 0.0018,
      "step": 14460
    },
    {
      "epoch": 4.33,
      "learning_rate": 5.842531272994849e-06,
      "loss": 0.001,
      "step": 14470
    },
    {
      "epoch": 4.34,
      "learning_rate": 5.832019341953117e-06,
      "loss": 0.0016,
      "step": 14480
    },
    {
      "epoch": 4.34,
      "learning_rate": 5.821507410911384e-06,
      "loss": 0.002,
      "step": 14490
    },
    {
      "epoch": 4.34,
      "learning_rate": 5.810995479869652e-06,
      "loss": 0.0014,
      "step": 14500
    },
    {
      "epoch": 4.35,
      "learning_rate": 5.8004835488279195e-06,
      "loss": 0.001,
      "step": 14510
    },
    {
      "epoch": 4.35,
      "learning_rate": 5.7899716177861875e-06,
      "loss": 0.0015,
      "step": 14520
    },
    {
      "epoch": 4.35,
      "learning_rate": 5.7794596867444555e-06,
      "loss": 0.0013,
      "step": 14530
    },
    {
      "epoch": 4.36,
      "learning_rate": 5.768947755702723e-06,
      "loss": 0.0008,
      "step": 14540
    },
    {
      "epoch": 4.36,
      "learning_rate": 5.758435824660991e-06,
      "loss": 0.0006,
      "step": 14550
    },
    {
      "epoch": 4.36,
      "learning_rate": 5.747923893619258e-06,
      "loss": 0.001,
      "step": 14560
    },
    {
      "epoch": 4.36,
      "learning_rate": 5.737411962577526e-06,
      "loss": 0.001,
      "step": 14570
    },
    {
      "epoch": 4.37,
      "learning_rate": 5.726900031535794e-06,
      "loss": 0.0016,
      "step": 14580
    },
    {
      "epoch": 4.37,
      "learning_rate": 5.716388100494061e-06,
      "loss": 0.0008,
      "step": 14590
    },
    {
      "epoch": 4.37,
      "learning_rate": 5.705876169452329e-06,
      "loss": 0.0006,
      "step": 14600
    },
    {
      "epoch": 4.37,
      "eval_f1": 0.9935126869535487,
      "eval_f1_0": 0.9979995496404354,
      "eval_f1_1": 0.9983877059221613,
      "eval_f1_2": 0.9786853741916205,
      "eval_f1_3": 0.9989781180599773,
      "eval_loss": 0.019463276490569115,
      "eval_runtime": 245.9955,
      "eval_samples_per_second": 48.245,
      "eval_steps_per_second": 0.191,
      "step": 14600
    },
    {
      "epoch": 4.38,
      "learning_rate": 5.6953642384105965e-06,
      "loss": 0.0164,
      "step": 14610
    },
    {
      "epoch": 4.38,
      "learning_rate": 5.6848523073688645e-06,
      "loss": 0.0008,
      "step": 14620
    },
    {
      "epoch": 4.38,
      "learning_rate": 5.674340376327132e-06,
      "loss": 0.002,
      "step": 14630
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.6638284452854e-06,
      "loss": 0.0014,
      "step": 14640
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.653316514243668e-06,
      "loss": 0.013,
      "step": 14650
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.642804583201935e-06,
      "loss": 0.0012,
      "step": 14660
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.632292652160203e-06,
      "loss": 0.0009,
      "step": 14670
    },
    {
      "epoch": 4.4,
      "learning_rate": 5.621780721118469e-06,
      "loss": 0.0044,
      "step": 14680
    },
    {
      "epoch": 4.4,
      "learning_rate": 5.611268790076737e-06,
      "loss": 0.0028,
      "step": 14690
    },
    {
      "epoch": 4.4,
      "learning_rate": 5.600756859035005e-06,
      "loss": 0.0008,
      "step": 14700
    },
    {
      "epoch": 4.41,
      "learning_rate": 5.590244927993273e-06,
      "loss": 0.0027,
      "step": 14710
    },
    {
      "epoch": 4.41,
      "learning_rate": 5.57973299695154e-06,
      "loss": 0.0007,
      "step": 14720
    },
    {
      "epoch": 4.41,
      "learning_rate": 5.569221065909808e-06,
      "loss": 0.0011,
      "step": 14730
    },
    {
      "epoch": 4.42,
      "learning_rate": 5.558709134868076e-06,
      "loss": 0.0019,
      "step": 14740
    },
    {
      "epoch": 4.42,
      "learning_rate": 5.548197203826343e-06,
      "loss": 0.026,
      "step": 14750
    },
    {
      "epoch": 4.42,
      "learning_rate": 5.537685272784611e-06,
      "loss": 0.0012,
      "step": 14760
    },
    {
      "epoch": 4.42,
      "learning_rate": 5.527173341742878e-06,
      "loss": 0.0007,
      "step": 14770
    },
    {
      "epoch": 4.43,
      "learning_rate": 5.516661410701146e-06,
      "loss": 0.0005,
      "step": 14780
    },
    {
      "epoch": 4.43,
      "learning_rate": 5.506149479659414e-06,
      "loss": 0.0005,
      "step": 14790
    },
    {
      "epoch": 4.43,
      "learning_rate": 5.495637548617682e-06,
      "loss": 0.0009,
      "step": 14800
    },
    {
      "epoch": 4.43,
      "eval_f1": 0.9934621837870611,
      "eval_f1_0": 0.9979738625280767,
      "eval_f1_1": 0.9983513534238172,
      "eval_f1_2": 0.9785205467497191,
      "eval_f1_3": 0.999002972446631,
      "eval_loss": 0.018196821212768555,
      "eval_runtime": 247.283,
      "eval_samples_per_second": 47.994,
      "eval_steps_per_second": 0.19,
      "step": 14800
    },
    {
      "epoch": 4.44,
      "learning_rate": 5.48512561757595e-06,
      "loss": 0.0012,
      "step": 14810
    },
    {
      "epoch": 4.44,
      "learning_rate": 5.474613686534217e-06,
      "loss": 0.0036,
      "step": 14820
    },
    {
      "epoch": 4.44,
      "learning_rate": 5.464101755492485e-06,
      "loss": 0.0016,
      "step": 14830
    },
    {
      "epoch": 4.45,
      "learning_rate": 5.453589824450752e-06,
      "loss": 0.0112,
      "step": 14840
    },
    {
      "epoch": 4.45,
      "learning_rate": 5.44307789340902e-06,
      "loss": 0.0045,
      "step": 14850
    },
    {
      "epoch": 4.45,
      "learning_rate": 5.432565962367287e-06,
      "loss": 0.0021,
      "step": 14860
    },
    {
      "epoch": 4.45,
      "learning_rate": 5.422054031325555e-06,
      "loss": 0.0017,
      "step": 14870
    },
    {
      "epoch": 4.46,
      "learning_rate": 5.411542100283823e-06,
      "loss": 0.0009,
      "step": 14880
    },
    {
      "epoch": 4.46,
      "learning_rate": 5.40103016924209e-06,
      "loss": 0.0014,
      "step": 14890
    },
    {
      "epoch": 4.46,
      "learning_rate": 5.390518238200357e-06,
      "loss": 0.0013,
      "step": 14900
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.380006307158625e-06,
      "loss": 0.0017,
      "step": 14910
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.369494376116893e-06,
      "loss": 0.0013,
      "step": 14920
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.35898244507516e-06,
      "loss": 0.0009,
      "step": 14930
    },
    {
      "epoch": 4.48,
      "learning_rate": 5.348470514033428e-06,
      "loss": 0.0037,
      "step": 14940
    },
    {
      "epoch": 4.48,
      "learning_rate": 5.3379585829916954e-06,
      "loss": 0.0009,
      "step": 14950
    },
    {
      "epoch": 4.48,
      "learning_rate": 5.3274466519499635e-06,
      "loss": 0.0023,
      "step": 14960
    },
    {
      "epoch": 4.48,
      "learning_rate": 5.3169347209082315e-06,
      "loss": 0.0023,
      "step": 14970
    },
    {
      "epoch": 4.49,
      "learning_rate": 5.306422789866499e-06,
      "loss": 0.0017,
      "step": 14980
    },
    {
      "epoch": 4.49,
      "learning_rate": 5.295910858824767e-06,
      "loss": 0.002,
      "step": 14990
    },
    {
      "epoch": 4.49,
      "learning_rate": 5.285398927783034e-06,
      "loss": 0.0008,
      "step": 15000
    },
    {
      "epoch": 4.49,
      "eval_f1": 0.9935254649586541,
      "eval_f1_0": 0.9979843646981257,
      "eval_f1_1": 0.9980474369253077,
      "eval_f1_2": 0.9791602794546564,
      "eval_f1_3": 0.9989097787565265,
      "eval_loss": 0.018152348697185516,
      "eval_runtime": 247.4004,
      "eval_samples_per_second": 47.971,
      "eval_steps_per_second": 0.19,
      "step": 15000
    },
    {
      "epoch": 4.5,
      "learning_rate": 5.274886996741302e-06,
      "loss": 0.0009,
      "step": 15010
    },
    {
      "epoch": 4.5,
      "learning_rate": 5.264375065699569e-06,
      "loss": 0.0014,
      "step": 15020
    },
    {
      "epoch": 4.5,
      "learning_rate": 5.253863134657837e-06,
      "loss": 0.0012,
      "step": 15030
    },
    {
      "epoch": 4.51,
      "learning_rate": 5.243351203616105e-06,
      "loss": 0.0007,
      "step": 15040
    },
    {
      "epoch": 4.51,
      "learning_rate": 5.2328392725743725e-06,
      "loss": 0.0008,
      "step": 15050
    },
    {
      "epoch": 4.51,
      "learning_rate": 5.2223273415326405e-06,
      "loss": 0.001,
      "step": 15060
    },
    {
      "epoch": 4.51,
      "learning_rate": 5.211815410490908e-06,
      "loss": 0.0015,
      "step": 15070
    },
    {
      "epoch": 4.52,
      "learning_rate": 5.201303479449176e-06,
      "loss": 0.001,
      "step": 15080
    },
    {
      "epoch": 4.52,
      "learning_rate": 5.190791548407444e-06,
      "loss": 0.0012,
      "step": 15090
    },
    {
      "epoch": 4.52,
      "learning_rate": 5.180279617365711e-06,
      "loss": 0.0007,
      "step": 15100
    },
    {
      "epoch": 4.53,
      "learning_rate": 5.169767686323977e-06,
      "loss": 0.0017,
      "step": 15110
    },
    {
      "epoch": 4.53,
      "learning_rate": 5.159255755282245e-06,
      "loss": 0.0032,
      "step": 15120
    },
    {
      "epoch": 4.53,
      "learning_rate": 5.148743824240513e-06,
      "loss": 0.0011,
      "step": 15130
    },
    {
      "epoch": 4.54,
      "learning_rate": 5.138231893198781e-06,
      "loss": 0.0012,
      "step": 15140
    },
    {
      "epoch": 4.54,
      "learning_rate": 5.127719962157049e-06,
      "loss": 0.005,
      "step": 15150
    },
    {
      "epoch": 4.54,
      "learning_rate": 5.117208031115316e-06,
      "loss": 0.0014,
      "step": 15160
    },
    {
      "epoch": 4.54,
      "learning_rate": 5.106696100073584e-06,
      "loss": 0.001,
      "step": 15170
    },
    {
      "epoch": 4.55,
      "learning_rate": 5.096184169031851e-06,
      "loss": 0.018,
      "step": 15180
    },
    {
      "epoch": 4.55,
      "learning_rate": 5.085672237990119e-06,
      "loss": 0.0009,
      "step": 15190
    },
    {
      "epoch": 4.55,
      "learning_rate": 5.075160306948387e-06,
      "loss": 0.001,
      "step": 15200
    },
    {
      "epoch": 4.55,
      "eval_f1": 0.9942465285733024,
      "eval_f1_0": 0.9981978271417208,
      "eval_f1_1": 0.9982744273974914,
      "eval_f1_2": 0.981498377666323,
      "eval_f1_3": 0.9990154820876749,
      "eval_loss": 0.016366170719265938,
      "eval_runtime": 247.0453,
      "eval_samples_per_second": 48.04,
      "eval_steps_per_second": 0.19,
      "step": 15200
    },
    {
      "epoch": 4.56,
      "learning_rate": 5.064648375906654e-06,
      "loss": 0.0162,
      "step": 15210
    },
    {
      "epoch": 4.56,
      "learning_rate": 5.054136444864922e-06,
      "loss": 0.0016,
      "step": 15220
    },
    {
      "epoch": 4.56,
      "learning_rate": 5.0436245138231896e-06,
      "loss": 0.0017,
      "step": 15230
    },
    {
      "epoch": 4.57,
      "learning_rate": 5.033112582781458e-06,
      "loss": 0.0011,
      "step": 15240
    },
    {
      "epoch": 4.57,
      "learning_rate": 5.022600651739725e-06,
      "loss": 0.0009,
      "step": 15250
    },
    {
      "epoch": 4.57,
      "learning_rate": 5.012088720697993e-06,
      "loss": 0.0013,
      "step": 15260
    },
    {
      "epoch": 4.57,
      "learning_rate": 5.001576789656261e-06,
      "loss": 0.0026,
      "step": 15270
    },
    {
      "epoch": 4.58,
      "learning_rate": 4.991064858614528e-06,
      "loss": 0.0008,
      "step": 15280
    },
    {
      "epoch": 4.58,
      "learning_rate": 4.980552927572795e-06,
      "loss": 0.0008,
      "step": 15290
    },
    {
      "epoch": 4.58,
      "learning_rate": 4.970040996531063e-06,
      "loss": 0.0006,
      "step": 15300
    },
    {
      "epoch": 4.59,
      "learning_rate": 4.9595290654893305e-06,
      "loss": 0.0149,
      "step": 15310
    },
    {
      "epoch": 4.59,
      "learning_rate": 4.9490171344475985e-06,
      "loss": 0.0012,
      "step": 15320
    },
    {
      "epoch": 4.59,
      "learning_rate": 4.938505203405866e-06,
      "loss": 0.0103,
      "step": 15330
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.927993272364134e-06,
      "loss": 0.0007,
      "step": 15340
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.917481341322402e-06,
      "loss": 0.0008,
      "step": 15350
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.906969410280669e-06,
      "loss": 0.012,
      "step": 15360
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.896457479238936e-06,
      "loss": 0.0021,
      "step": 15370
    },
    {
      "epoch": 4.61,
      "learning_rate": 4.885945548197204e-06,
      "loss": 0.0019,
      "step": 15380
    },
    {
      "epoch": 4.61,
      "learning_rate": 4.8754336171554714e-06,
      "loss": 0.0011,
      "step": 15390
    },
    {
      "epoch": 4.61,
      "learning_rate": 4.8649216861137395e-06,
      "loss": 0.0028,
      "step": 15400
    },
    {
      "epoch": 4.61,
      "eval_f1": 0.9938303910677069,
      "eval_f1_0": 0.9980855662265384,
      "eval_f1_1": 0.9982303124658477,
      "eval_f1_2": 0.980108372095536,
      "eval_f1_3": 0.9988973134829052,
      "eval_loss": 0.015507446601986885,
      "eval_runtime": 246.8306,
      "eval_samples_per_second": 48.082,
      "eval_steps_per_second": 0.19,
      "step": 15400
    },
    {
      "epoch": 4.62,
      "learning_rate": 4.854409755072007e-06,
      "loss": 0.0022,
      "step": 15410
    },
    {
      "epoch": 4.62,
      "learning_rate": 4.843897824030275e-06,
      "loss": 0.001,
      "step": 15420
    },
    {
      "epoch": 4.62,
      "learning_rate": 4.833385892988543e-06,
      "loss": 0.001,
      "step": 15430
    },
    {
      "epoch": 4.63,
      "learning_rate": 4.82287396194681e-06,
      "loss": 0.0014,
      "step": 15440
    },
    {
      "epoch": 4.63,
      "learning_rate": 4.812362030905078e-06,
      "loss": 0.0011,
      "step": 15450
    },
    {
      "epoch": 4.63,
      "learning_rate": 4.801850099863345e-06,
      "loss": 0.0017,
      "step": 15460
    },
    {
      "epoch": 4.63,
      "learning_rate": 4.791338168821613e-06,
      "loss": 0.0048,
      "step": 15470
    },
    {
      "epoch": 4.64,
      "learning_rate": 4.78082623777988e-06,
      "loss": 0.0011,
      "step": 15480
    },
    {
      "epoch": 4.64,
      "learning_rate": 4.770314306738148e-06,
      "loss": 0.001,
      "step": 15490
    },
    {
      "epoch": 4.64,
      "learning_rate": 4.759802375696416e-06,
      "loss": 0.0007,
      "step": 15500
    },
    {
      "epoch": 4.65,
      "learning_rate": 4.749290444654684e-06,
      "loss": 0.0117,
      "step": 15510
    },
    {
      "epoch": 4.65,
      "learning_rate": 4.738778513612951e-06,
      "loss": 0.0012,
      "step": 15520
    },
    {
      "epoch": 4.65,
      "learning_rate": 4.728266582571219e-06,
      "loss": 0.001,
      "step": 15530
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.717754651529486e-06,
      "loss": 0.013,
      "step": 15540
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.707242720487754e-06,
      "loss": 0.0008,
      "step": 15550
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.696730789446022e-06,
      "loss": 0.0006,
      "step": 15560
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.686218858404289e-06,
      "loss": 0.0042,
      "step": 15570
    },
    {
      "epoch": 4.67,
      "learning_rate": 4.6757069273625574e-06,
      "loss": 0.0025,
      "step": 15580
    },
    {
      "epoch": 4.67,
      "learning_rate": 4.665194996320825e-06,
      "loss": 0.0009,
      "step": 15590
    },
    {
      "epoch": 4.67,
      "learning_rate": 4.654683065279092e-06,
      "loss": 0.002,
      "step": 15600
    },
    {
      "epoch": 4.67,
      "eval_f1": 0.9935568331127168,
      "eval_f1_0": 0.9979808316932188,
      "eval_f1_1": 0.9981607881569333,
      "eval_f1_2": 0.9791199672385551,
      "eval_f1_3": 0.99896574536216,
      "eval_loss": 0.015655139461159706,
      "eval_runtime": 246.978,
      "eval_samples_per_second": 48.053,
      "eval_steps_per_second": 0.19,
      "step": 15600
    },
    {
      "epoch": 4.68,
      "learning_rate": 4.64417113423736e-06,
      "loss": 0.0011,
      "step": 15610
    },
    {
      "epoch": 4.68,
      "learning_rate": 4.633659203195627e-06,
      "loss": 0.0014,
      "step": 15620
    },
    {
      "epoch": 4.68,
      "learning_rate": 4.623147272153895e-06,
      "loss": 0.0014,
      "step": 15630
    },
    {
      "epoch": 4.69,
      "learning_rate": 4.612635341112163e-06,
      "loss": 0.0008,
      "step": 15640
    },
    {
      "epoch": 4.69,
      "learning_rate": 4.60212341007043e-06,
      "loss": 0.0009,
      "step": 15650
    },
    {
      "epoch": 4.69,
      "learning_rate": 4.591611479028698e-06,
      "loss": 0.0013,
      "step": 15660
    },
    {
      "epoch": 4.69,
      "learning_rate": 4.5810995479869656e-06,
      "loss": 0.0006,
      "step": 15670
    },
    {
      "epoch": 4.7,
      "learning_rate": 4.570587616945234e-06,
      "loss": 0.0011,
      "step": 15680
    },
    {
      "epoch": 4.7,
      "learning_rate": 4.560075685903501e-06,
      "loss": 0.0037,
      "step": 15690
    },
    {
      "epoch": 4.7,
      "learning_rate": 4.549563754861768e-06,
      "loss": 0.0015,
      "step": 15700
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.539051823820036e-06,
      "loss": 0.0007,
      "step": 15710
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.528539892778303e-06,
      "loss": 0.0008,
      "step": 15720
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.518027961736571e-06,
      "loss": 0.0014,
      "step": 15730
    },
    {
      "epoch": 4.72,
      "learning_rate": 4.507516030694839e-06,
      "loss": 0.0014,
      "step": 15740
    },
    {
      "epoch": 4.72,
      "learning_rate": 4.4970040996531065e-06,
      "loss": 0.0011,
      "step": 15750
    },
    {
      "epoch": 4.72,
      "learning_rate": 4.4864921686113745e-06,
      "loss": 0.001,
      "step": 15760
    },
    {
      "epoch": 4.72,
      "learning_rate": 4.475980237569642e-06,
      "loss": 0.001,
      "step": 15770
    },
    {
      "epoch": 4.73,
      "learning_rate": 4.46546830652791e-06,
      "loss": 0.0007,
      "step": 15780
    },
    {
      "epoch": 4.73,
      "learning_rate": 4.454956375486178e-06,
      "loss": 0.0028,
      "step": 15790
    },
    {
      "epoch": 4.73,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.001,
      "step": 15800
    },
    {
      "epoch": 4.73,
      "eval_f1": 0.9938767978233104,
      "eval_f1_0": 0.9981261152231388,
      "eval_f1_1": 0.9983157660773445,
      "eval_f1_2": 0.9801369881786712,
      "eval_f1_3": 0.9989283218140871,
      "eval_loss": 0.01843990944325924,
      "eval_runtime": 246.9918,
      "eval_samples_per_second": 48.05,
      "eval_steps_per_second": 0.19,
      "step": 15800
    },
    {
      "epoch": 4.74,
      "learning_rate": 4.433932513402712e-06,
      "loss": 0.0023,
      "step": 15810
    },
    {
      "epoch": 4.74,
      "learning_rate": 4.42342058236098e-06,
      "loss": 0.0009,
      "step": 15820
    },
    {
      "epoch": 4.74,
      "learning_rate": 4.412908651319247e-06,
      "loss": 0.0009,
      "step": 15830
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.4023967202775155e-06,
      "loss": 0.0049,
      "step": 15840
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.391884789235783e-06,
      "loss": 0.0071,
      "step": 15850
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.381372858194051e-06,
      "loss": 0.0018,
      "step": 15860
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.370860927152319e-06,
      "loss": 0.0024,
      "step": 15870
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.360348996110586e-06,
      "loss": 0.0012,
      "step": 15880
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.349837065068854e-06,
      "loss": 0.0013,
      "step": 15890
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.339325134027121e-06,
      "loss": 0.002,
      "step": 15900
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.328813202985388e-06,
      "loss": 0.002,
      "step": 15910
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.318301271943656e-06,
      "loss": 0.0011,
      "step": 15920
    },
    {
      "epoch": 4.77,
      "learning_rate": 4.307789340901924e-06,
      "loss": 0.0019,
      "step": 15930
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.297277409860192e-06,
      "loss": 0.002,
      "step": 15940
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.28676547881846e-06,
      "loss": 0.001,
      "step": 15950
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.276253547776727e-06,
      "loss": 0.0007,
      "step": 15960
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.265741616734995e-06,
      "loss": 0.0007,
      "step": 15970
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.255229685693262e-06,
      "loss": 0.0023,
      "step": 15980
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.24471775465153e-06,
      "loss": 0.0012,
      "step": 15990
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.234205823609797e-06,
      "loss": 0.0009,
      "step": 16000
    },
    {
      "epoch": 4.79,
      "eval_f1": 0.9940040629929541,
      "eval_f1_0": 0.9981262899852337,
      "eval_f1_1": 0.9981356952974256,
      "eval_f1_2": 0.9807885277515572,
      "eval_f1_3": 0.9989657389375995,
      "eval_loss": 0.01693156361579895,
      "eval_runtime": 247.0688,
      "eval_samples_per_second": 48.035,
      "eval_steps_per_second": 0.19,
      "step": 16000
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.2236938925680645e-06,
      "loss": 0.0013,
      "step": 16010
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.2131819615263326e-06,
      "loss": 0.0013,
      "step": 16020
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.202670030484601e-06,
      "loss": 0.0007,
      "step": 16030
    },
    {
      "epoch": 4.81,
      "learning_rate": 4.192158099442868e-06,
      "loss": 0.0007,
      "step": 16040
    },
    {
      "epoch": 4.81,
      "learning_rate": 4.181646168401136e-06,
      "loss": 0.0008,
      "step": 16050
    },
    {
      "epoch": 4.81,
      "learning_rate": 4.171134237359403e-06,
      "loss": 0.0026,
      "step": 16060
    },
    {
      "epoch": 4.81,
      "learning_rate": 4.160622306317671e-06,
      "loss": 0.001,
      "step": 16070
    },
    {
      "epoch": 4.82,
      "learning_rate": 4.150110375275938e-06,
      "loss": 0.001,
      "step": 16080
    },
    {
      "epoch": 4.82,
      "learning_rate": 4.139598444234206e-06,
      "loss": 0.0008,
      "step": 16090
    },
    {
      "epoch": 4.82,
      "learning_rate": 4.129086513192474e-06,
      "loss": 0.0008,
      "step": 16100
    },
    {
      "epoch": 4.83,
      "learning_rate": 4.1185745821507415e-06,
      "loss": 0.0007,
      "step": 16110
    },
    {
      "epoch": 4.83,
      "learning_rate": 4.108062651109009e-06,
      "loss": 0.0019,
      "step": 16120
    },
    {
      "epoch": 4.83,
      "learning_rate": 4.097550720067277e-06,
      "loss": 0.0019,
      "step": 16130
    },
    {
      "epoch": 4.84,
      "learning_rate": 4.087038789025544e-06,
      "loss": 0.0015,
      "step": 16140
    },
    {
      "epoch": 4.84,
      "learning_rate": 4.076526857983812e-06,
      "loss": 0.0021,
      "step": 16150
    },
    {
      "epoch": 4.84,
      "learning_rate": 4.066014926942079e-06,
      "loss": 0.0019,
      "step": 16160
    },
    {
      "epoch": 4.84,
      "learning_rate": 4.055502995900347e-06,
      "loss": 0.0008,
      "step": 16170
    },
    {
      "epoch": 4.85,
      "learning_rate": 4.044991064858615e-06,
      "loss": 0.0019,
      "step": 16180
    },
    {
      "epoch": 4.85,
      "learning_rate": 4.0344791338168825e-06,
      "loss": 0.0026,
      "step": 16190
    },
    {
      "epoch": 4.85,
      "learning_rate": 4.0239672027751505e-06,
      "loss": 0.0007,
      "step": 16200
    },
    {
      "epoch": 4.85,
      "eval_f1": 0.9937509665798826,
      "eval_f1_0": 0.998028197941083,
      "eval_f1_1": 0.9976161868970891,
      "eval_f1_2": 0.9803688828907932,
      "eval_f1_3": 0.998990598590565,
      "eval_loss": 0.019963156431913376,
      "eval_runtime": 244.5505,
      "eval_samples_per_second": 48.53,
      "eval_steps_per_second": 0.192,
      "step": 16200
    },
    {
      "epoch": 4.86,
      "learning_rate": 4.013455271733418e-06,
      "loss": 0.0011,
      "step": 16210
    },
    {
      "epoch": 4.86,
      "learning_rate": 4.002943340691686e-06,
      "loss": 0.0034,
      "step": 16220
    },
    {
      "epoch": 4.86,
      "learning_rate": 3.992431409649953e-06,
      "loss": 0.0008,
      "step": 16230
    },
    {
      "epoch": 4.87,
      "learning_rate": 3.98191947860822e-06,
      "loss": 0.0012,
      "step": 16240
    },
    {
      "epoch": 4.87,
      "learning_rate": 3.971407547566488e-06,
      "loss": 0.0009,
      "step": 16250
    },
    {
      "epoch": 4.87,
      "learning_rate": 3.960895616524756e-06,
      "loss": 0.0096,
      "step": 16260
    },
    {
      "epoch": 4.87,
      "learning_rate": 3.950383685483023e-06,
      "loss": 0.0021,
      "step": 16270
    },
    {
      "epoch": 4.88,
      "learning_rate": 3.9398717544412915e-06,
      "loss": 0.0016,
      "step": 16280
    },
    {
      "epoch": 4.88,
      "learning_rate": 3.929359823399559e-06,
      "loss": 0.001,
      "step": 16290
    },
    {
      "epoch": 4.88,
      "learning_rate": 3.918847892357827e-06,
      "loss": 0.002,
      "step": 16300
    },
    {
      "epoch": 4.89,
      "learning_rate": 3.908335961316094e-06,
      "loss": 0.0011,
      "step": 16310
    },
    {
      "epoch": 4.89,
      "learning_rate": 3.897824030274362e-06,
      "loss": 0.0013,
      "step": 16320
    },
    {
      "epoch": 4.89,
      "learning_rate": 3.887312099232629e-06,
      "loss": 0.0011,
      "step": 16330
    },
    {
      "epoch": 4.9,
      "learning_rate": 3.876800168190897e-06,
      "loss": 0.0007,
      "step": 16340
    },
    {
      "epoch": 4.9,
      "learning_rate": 3.866288237149164e-06,
      "loss": 0.0011,
      "step": 16350
    },
    {
      "epoch": 4.9,
      "learning_rate": 3.855776306107432e-06,
      "loss": 0.0039,
      "step": 16360
    },
    {
      "epoch": 4.9,
      "learning_rate": 3.8452643750657e-06,
      "loss": 0.001,
      "step": 16370
    },
    {
      "epoch": 4.91,
      "learning_rate": 3.834752444023968e-06,
      "loss": 0.0021,
      "step": 16380
    },
    {
      "epoch": 4.91,
      "learning_rate": 3.824240512982235e-06,
      "loss": 0.0006,
      "step": 16390
    },
    {
      "epoch": 4.91,
      "learning_rate": 3.813728581940503e-06,
      "loss": 0.0007,
      "step": 16400
    },
    {
      "epoch": 4.91,
      "eval_f1": 0.9938115465044991,
      "eval_f1_0": 0.9980795383541,
      "eval_f1_1": 0.9981918620369642,
      "eval_f1_2": 0.9800091237902832,
      "eval_f1_3": 0.9989656618366491,
      "eval_loss": 0.01877663843333721,
      "eval_runtime": 246.9983,
      "eval_samples_per_second": 48.049,
      "eval_steps_per_second": 0.19,
      "step": 16400
    },
    {
      "epoch": 4.92,
      "learning_rate": 3.8032166508987705e-06,
      "loss": 0.0016,
      "step": 16410
    },
    {
      "epoch": 4.92,
      "learning_rate": 3.792704719857038e-06,
      "loss": 0.0018,
      "step": 16420
    },
    {
      "epoch": 4.92,
      "learning_rate": 3.782192788815306e-06,
      "loss": 0.0011,
      "step": 16430
    },
    {
      "epoch": 4.93,
      "learning_rate": 3.771680857773573e-06,
      "loss": 0.0027,
      "step": 16440
    },
    {
      "epoch": 4.93,
      "learning_rate": 3.761168926731841e-06,
      "loss": 0.0009,
      "step": 16450
    },
    {
      "epoch": 4.93,
      "learning_rate": 3.7506569956901086e-06,
      "loss": 0.001,
      "step": 16460
    },
    {
      "epoch": 4.93,
      "learning_rate": 3.740145064648376e-06,
      "loss": 0.0015,
      "step": 16470
    },
    {
      "epoch": 4.94,
      "learning_rate": 3.7296331336066438e-06,
      "loss": 0.001,
      "step": 16480
    },
    {
      "epoch": 4.94,
      "learning_rate": 3.7191212025649114e-06,
      "loss": 0.0008,
      "step": 16490
    },
    {
      "epoch": 4.94,
      "learning_rate": 3.708609271523179e-06,
      "loss": 0.0009,
      "step": 16500
    },
    {
      "epoch": 4.95,
      "learning_rate": 3.698097340481447e-06,
      "loss": 0.0008,
      "step": 16510
    },
    {
      "epoch": 4.95,
      "learning_rate": 3.6875854094397147e-06,
      "loss": 0.0016,
      "step": 16520
    },
    {
      "epoch": 4.95,
      "learning_rate": 3.6770734783979823e-06,
      "loss": 0.022,
      "step": 16530
    },
    {
      "epoch": 4.96,
      "learning_rate": 3.66656154735625e-06,
      "loss": 0.0017,
      "step": 16540
    },
    {
      "epoch": 4.96,
      "learning_rate": 3.656049616314517e-06,
      "loss": 0.0016,
      "step": 16550
    },
    {
      "epoch": 4.96,
      "learning_rate": 3.6455376852727847e-06,
      "loss": 0.0009,
      "step": 16560
    },
    {
      "epoch": 4.96,
      "learning_rate": 3.6350257542310523e-06,
      "loss": 0.0037,
      "step": 16570
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.62451382318932e-06,
      "loss": 0.0007,
      "step": 16580
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.614001892147588e-06,
      "loss": 0.001,
      "step": 16590
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.6034899611058556e-06,
      "loss": 0.0093,
      "step": 16600
    },
    {
      "epoch": 4.97,
      "eval_f1": 0.9936563029540371,
      "eval_f1_0": 0.9980271122740356,
      "eval_f1_1": 0.9981113299113571,
      "eval_f1_2": 0.9797076115457067,
      "eval_f1_3": 0.9987791580850489,
      "eval_loss": 0.016260361298918724,
      "eval_runtime": 247.0416,
      "eval_samples_per_second": 48.04,
      "eval_steps_per_second": 0.19,
      "step": 16600
    },
    {
      "epoch": 4.97,
      "step": 16600,
      "total_flos": 2.5850795861737472e+17,
      "train_loss": 0.025379759394816202,
      "train_runtime": 50449.2502,
      "train_samples_per_second": 12.702,
      "train_steps_per_second": 0.397
    }
  ],
  "logging_steps": 10,
  "max_steps": 20028,
  "num_train_epochs": 6,
  "save_steps": 200,
  "total_flos": 2.5850795861737472e+17,
  "trial_name": null,
  "trial_params": null
}