{
  "best_metric": 0.7018599562363238,
  "best_model_checkpoint": "/tmp/test-ner1_/checkpoint-52109",
  "epoch": 120.0,
  "global_step": 58440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7978578755277784,
      "eval_f1": 0.5672056594239515,
      "eval_loss": 0.6516662240028381,
      "eval_precision": 0.5179972311952007,
      "eval_recall": 0.6267448352875489,
      "eval_runtime": 6.9697,
      "eval_samples_per_second": 159.549,
      "eval_steps_per_second": 10.044,
      "step": 487
    },
    {
      "epoch": 1.03,
      "learning_rate": 2.9743326488706367e-05,
      "loss": 1.0091,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8116457145583359,
      "eval_f1": 0.5956049507451376,
      "eval_loss": 0.6197972893714905,
      "eval_precision": 0.5438191881918819,
      "eval_recall": 0.6582914572864321,
      "eval_runtime": 6.8446,
      "eval_samples_per_second": 162.464,
      "eval_steps_per_second": 10.227,
      "step": 974
    },
    {
      "epoch": 2.05,
      "learning_rate": 2.9486652977412733e-05,
      "loss": 0.5042,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8094272560291977,
      "eval_f1": 0.616483796886961,
      "eval_loss": 0.6614459753036499,
      "eval_precision": 0.5676691729323309,
      "eval_recall": 0.6744835287548855,
      "eval_runtime": 6.7608,
      "eval_samples_per_second": 164.478,
      "eval_steps_per_second": 10.354,
      "step": 1461
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.9229979466119096e-05,
      "loss": 0.3543,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8230719687030367,
      "eval_f1": 0.6266820453671665,
      "eval_loss": 0.6793712377548218,
      "eval_precision": 0.5792466240227434,
      "eval_recall": 0.6825795644891123,
      "eval_runtime": 7.0887,
      "eval_samples_per_second": 156.869,
      "eval_steps_per_second": 9.875,
      "step": 1948
    },
    {
      "epoch": 4.11,
      "learning_rate": 2.8973305954825462e-05,
      "loss": 0.2476,
      "step": 2000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8249326113403783,
      "eval_f1": 0.653114074841383,
      "eval_loss": 0.7132434248924255,
      "eval_precision": 0.6090316348708041,
      "eval_recall": 0.7040759352317141,
      "eval_runtime": 6.8077,
      "eval_samples_per_second": 163.344,
      "eval_steps_per_second": 10.282,
      "step": 2435
    },
    {
      "epoch": 5.13,
      "learning_rate": 2.8716632443531828e-05,
      "loss": 0.1849,
      "step": 2500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8218553946709286,
      "eval_f1": 0.6443319049474094,
      "eval_loss": 0.776131272315979,
      "eval_precision": 0.6023306627822287,
      "eval_recall": 0.6926298157453936,
      "eval_runtime": 6.8192,
      "eval_samples_per_second": 163.07,
      "eval_steps_per_second": 10.265,
      "step": 2922
    },
    {
      "epoch": 6.16,
      "learning_rate": 2.8459958932238194e-05,
      "loss": 0.1465,
      "step": 3000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8172992056487202,
      "eval_f1": 0.6444159178433889,
      "eval_loss": 0.8293828368186951,
      "eval_precision": 0.5964828897338403,
      "eval_recall": 0.7007258514796203,
      "eval_runtime": 6.7817,
      "eval_samples_per_second": 163.971,
      "eval_steps_per_second": 10.322,
      "step": 3409
    },
    {
      "epoch": 7.19,
      "learning_rate": 2.8203285420944557e-05,
      "loss": 0.1176,
      "step": 3500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8229765511318909,
      "eval_f1": 0.6518829549927831,
      "eval_loss": 0.8652628064155579,
      "eval_precision": 0.6150037137905422,
      "eval_recall": 0.6934673366834171,
      "eval_runtime": 6.8373,
      "eval_samples_per_second": 162.638,
      "eval_steps_per_second": 10.238,
      "step": 3896
    },
    {
      "epoch": 8.21,
      "learning_rate": 2.7946611909650923e-05,
      "loss": 0.1023,
      "step": 4000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.822642589632881,
      "eval_f1": 0.6499869007073618,
      "eval_loss": 0.8613839745521545,
      "eval_precision": 0.6122902270483712,
      "eval_recall": 0.6926298157453936,
      "eval_runtime": 6.8784,
      "eval_samples_per_second": 161.665,
      "eval_steps_per_second": 10.177,
      "step": 4383
    },
    {
      "epoch": 9.24,
      "learning_rate": 2.7689938398357292e-05,
      "loss": 0.0823,
      "step": 4500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8216168507430643,
      "eval_f1": 0.6437475396929537,
      "eval_loss": 0.9824637174606323,
      "eval_precision": 0.6073285466699678,
      "eval_recall": 0.6848129536571748,
      "eval_runtime": 6.8843,
      "eval_samples_per_second": 161.527,
      "eval_steps_per_second": 10.168,
      "step": 4870
    },
    {
      "epoch": 10.27,
      "learning_rate": 2.743326488706366e-05,
      "loss": 0.069,
      "step": 5000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8247894849836598,
      "eval_f1": 0.6582144743792923,
      "eval_loss": 0.9782905578613281,
      "eval_precision": 0.624561403508772,
      "eval_recall": 0.6957007258514796,
      "eval_runtime": 6.9683,
      "eval_samples_per_second": 159.58,
      "eval_steps_per_second": 10.046,
      "step": 5357
    },
    {
      "epoch": 11.29,
      "learning_rate": 2.717659137577002e-05,
      "loss": 0.0578,
      "step": 5500
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8224040457050166,
      "eval_f1": 0.654025974025974,
      "eval_loss": 1.003711462020874,
      "eval_precision": 0.6114618746964546,
      "eval_recall": 0.7029592406476829,
      "eval_runtime": 7.0443,
      "eval_samples_per_second": 157.859,
      "eval_steps_per_second": 9.937,
      "step": 5844
    },
    {
      "epoch": 12.32,
      "learning_rate": 2.6919917864476387e-05,
      "loss": 0.0522,
      "step": 6000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8160826316166122,
      "eval_f1": 0.6486343145054362,
      "eval_loss": 1.0798953771591187,
      "eval_precision": 0.6176767676767677,
      "eval_recall": 0.68285873813512,
      "eval_runtime": 6.9483,
      "eval_samples_per_second": 160.039,
      "eval_steps_per_second": 10.074,
      "step": 6331
    },
    {
      "epoch": 13.35,
      "learning_rate": 2.6663244353182754e-05,
      "loss": 0.0461,
      "step": 6500
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8203048591398105,
      "eval_f1": 0.651880674448768,
      "eval_loss": 1.0693352222442627,
      "eval_precision": 0.6087693798449613,
      "eval_recall": 0.7015633724176438,
      "eval_runtime": 7.0475,
      "eval_samples_per_second": 157.787,
      "eval_steps_per_second": 9.933,
      "step": 6818
    },
    {
      "epoch": 14.37,
      "learning_rate": 2.640657084188912e-05,
      "loss": 0.0402,
      "step": 7000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8230242599174638,
      "eval_f1": 0.6548117154811716,
      "eval_loss": 1.055952548980713,
      "eval_precision": 0.6158386620757501,
      "eval_recall": 0.6990508096035735,
      "eval_runtime": 6.7343,
      "eval_samples_per_second": 165.125,
      "eval_steps_per_second": 10.395,
      "step": 7305
    },
    {
      "epoch": 15.4,
      "learning_rate": 2.6149897330595482e-05,
      "loss": 0.0369,
      "step": 7500
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.819660790534577,
      "eval_f1": 0.6594724220623501,
      "eval_loss": 1.104617714881897,
      "eval_precision": 0.6307339449541285,
      "eval_recall": 0.6909547738693468,
      "eval_runtime": 6.9981,
      "eval_samples_per_second": 158.901,
      "eval_steps_per_second": 10.003,
      "step": 7792
    },
    {
      "epoch": 16.43,
      "learning_rate": 2.589322381930185e-05,
      "loss": 0.0391,
      "step": 8000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8232628038453281,
      "eval_f1": 0.6534837425348374,
      "eval_loss": 1.1480329036712646,
      "eval_precision": 0.622818112825702,
      "eval_recall": 0.6873255164712451,
      "eval_runtime": 6.8282,
      "eval_samples_per_second": 162.854,
      "eval_steps_per_second": 10.252,
      "step": 8279
    },
    {
      "epoch": 17.45,
      "learning_rate": 2.5636550308008215e-05,
      "loss": 0.0537,
      "step": 8500
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8203764223181699,
      "eval_f1": 0.6538308325587572,
      "eval_loss": 1.2140594720840454,
      "eval_precision": 0.6234489744239048,
      "eval_recall": 0.6873255164712451,
      "eval_runtime": 6.8978,
      "eval_samples_per_second": 161.212,
      "eval_steps_per_second": 10.148,
      "step": 8766
    },
    {
      "epoch": 18.48,
      "learning_rate": 2.537987679671458e-05,
      "loss": 0.0497,
      "step": 9000
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.818945158750984,
      "eval_f1": 0.6579537953795379,
      "eval_loss": 1.222991943359375,
      "eval_precision": 0.6240921612822439,
      "eval_recall": 0.6957007258514796,
      "eval_runtime": 6.7892,
      "eval_samples_per_second": 163.791,
      "eval_steps_per_second": 10.311,
      "step": 9253
    },
    {
      "epoch": 19.51,
      "learning_rate": 2.5123203285420943e-05,
      "loss": 0.0512,
      "step": 9500
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8268171083705065,
      "eval_f1": 0.6670205706702057,
      "eval_loss": 1.207837462425232,
      "eval_precision": 0.6357197065519858,
      "eval_recall": 0.7015633724176438,
      "eval_runtime": 6.7858,
      "eval_samples_per_second": 163.871,
      "eval_steps_per_second": 10.316,
      "step": 9740
    },
    {
      "epoch": 20.53,
      "learning_rate": 2.486652977412731e-05,
      "loss": 0.0508,
      "step": 10000
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.8178240022900217,
      "eval_f1": 0.6514255682564709,
      "eval_loss": 1.1940995454788208,
      "eval_precision": 0.615289153636138,
      "eval_recall": 0.692071468453378,
      "eval_runtime": 6.9352,
      "eval_samples_per_second": 160.342,
      "eval_steps_per_second": 10.093,
      "step": 10227
    },
    {
      "epoch": 21.56,
      "learning_rate": 2.4609856262833676e-05,
      "loss": 0.044,
      "step": 10500
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.8161064860093986,
      "eval_f1": 0.6639004149377594,
      "eval_loss": 1.311377763748169,
      "eval_precision": 0.6376960658266907,
      "eval_recall": 0.6923506420993858,
      "eval_runtime": 6.93,
      "eval_samples_per_second": 160.462,
      "eval_steps_per_second": 10.101,
      "step": 10714
    },
    {
      "epoch": 22.59,
      "learning_rate": 2.4353182751540042e-05,
      "loss": 0.041,
      "step": 11000
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.8165120106867679,
      "eval_f1": 0.6519497686715136,
      "eval_loss": 1.2640310525894165,
      "eval_precision": 0.6191313080592519,
      "eval_recall": 0.6884422110552764,
      "eval_runtime": 6.7498,
      "eval_samples_per_second": 164.745,
      "eval_steps_per_second": 10.371,
      "step": 11201
    },
    {
      "epoch": 23.61,
      "learning_rate": 2.4096509240246405e-05,
      "loss": 0.0216,
      "step": 11500
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.824026144414494,
      "eval_f1": 0.6626618608997463,
      "eval_loss": 1.3126922845840454,
      "eval_precision": 0.6349449987209005,
      "eval_recall": 0.6929089893914014,
      "eval_runtime": 6.9538,
      "eval_samples_per_second": 159.913,
      "eval_steps_per_second": 10.066,
      "step": 11688
    },
    {
      "epoch": 24.64,
      "learning_rate": 2.383983572895277e-05,
      "loss": 0.0187,
      "step": 12000
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.8228572791679588,
      "eval_f1": 0.6716637665640476,
      "eval_loss": 1.3328564167022705,
      "eval_precision": 0.6451529956286963,
      "eval_recall": 0.7004466778336125,
      "eval_runtime": 6.8192,
      "eval_samples_per_second": 163.069,
      "eval_steps_per_second": 10.265,
      "step": 12175
    },
    {
      "epoch": 25.67,
      "learning_rate": 2.3583162217659137e-05,
      "loss": 0.0158,
      "step": 12500
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.8177047303260896,
      "eval_f1": 0.6601762925930797,
      "eval_loss": 1.2958290576934814,
      "eval_precision": 0.6242846479223687,
      "eval_recall": 0.7004466778336125,
      "eval_runtime": 6.9466,
      "eval_samples_per_second": 160.077,
      "eval_steps_per_second": 10.077,
      "step": 12662
    },
    {
      "epoch": 26.69,
      "learning_rate": 2.3326488706365506e-05,
      "loss": 0.0151,
      "step": 13000
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.818062546217886,
      "eval_f1": 0.6571240971766251,
      "eval_loss": 1.3275514841079712,
      "eval_precision": 0.6203818497396479,
      "eval_recall": 0.6984924623115578,
      "eval_runtime": 6.7421,
      "eval_samples_per_second": 164.934,
      "eval_steps_per_second": 10.383,
      "step": 13149
    },
    {
      "epoch": 27.72,
      "learning_rate": 2.306981519507187e-05,
      "loss": 0.016,
      "step": 13500
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.8251473008754562,
      "eval_f1": 0.6730201342281878,
      "eval_loss": 1.2671375274658203,
      "eval_precision": 0.6481385729058945,
      "eval_recall": 0.6998883305415968,
      "eval_runtime": 6.8365,
      "eval_samples_per_second": 162.657,
      "eval_steps_per_second": 10.239,
      "step": 13636
    },
    {
      "epoch": 28.75,
      "learning_rate": 2.2813141683778235e-05,
      "loss": 0.0157,
      "step": 14000
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.8204241311037428,
      "eval_f1": 0.654650703854756,
      "eval_loss": 1.3373730182647705,
      "eval_precision": 0.6190594675292361,
      "eval_recall": 0.6945840312674484,
      "eval_runtime": 7.169,
      "eval_samples_per_second": 155.113,
      "eval_steps_per_second": 9.764,
      "step": 14123
    },
    {
      "epoch": 29.77,
      "learning_rate": 2.25564681724846e-05,
      "loss": 0.0146,
      "step": 14500
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.8191598482860619,
      "eval_f1": 0.6739956568946797,
      "eval_loss": 1.394074559211731,
      "eval_precision": 0.655837295298468,
      "eval_recall": 0.6931881630374093,
      "eval_runtime": 6.8998,
      "eval_samples_per_second": 161.164,
      "eval_steps_per_second": 10.145,
      "step": 14610
    },
    {
      "epoch": 30.8,
      "learning_rate": 2.2299794661190967e-05,
      "loss": 0.0134,
      "step": 15000
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.8163927387228358,
      "eval_f1": 0.6588835212023617,
      "eval_loss": 1.4215149879455566,
      "eval_precision": 0.6343669250645995,
      "eval_recall": 0.6853713009491904,
      "eval_runtime": 6.8752,
      "eval_samples_per_second": 161.741,
      "eval_steps_per_second": 10.182,
      "step": 15097
    },
    {
      "epoch": 31.83,
      "learning_rate": 2.204312114989733e-05,
      "loss": 0.0146,
      "step": 15500
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.8197562081057227,
      "eval_f1": 0.6717123935666982,
      "eval_loss": 1.4602118730545044,
      "eval_precision": 0.6510348441184176,
      "eval_recall": 0.6937465103294249,
      "eval_runtime": 6.8475,
      "eval_samples_per_second": 162.396,
      "eval_steps_per_second": 10.223,
      "step": 15584
    },
    {
      "epoch": 32.85,
      "learning_rate": 2.1786447638603696e-05,
      "loss": 0.0105,
      "step": 16000
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.823978435628921,
      "eval_f1": 0.6736138944555778,
      "eval_loss": 1.4084781408309937,
      "eval_precision": 0.6459133999487574,
      "eval_recall": 0.7037967615857063,
      "eval_runtime": 6.812,
      "eval_samples_per_second": 163.241,
      "eval_steps_per_second": 10.276,
      "step": 16071
    },
    {
      "epoch": 33.88,
      "learning_rate": 2.1529774127310062e-05,
      "loss": 0.0135,
      "step": 16500
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.8166312826507002,
      "eval_f1": 0.6652519893899204,
      "eval_loss": 1.3592619895935059,
      "eval_precision": 0.6336533602829711,
      "eval_recall": 0.7001675041876047,
      "eval_runtime": 6.8441,
      "eval_samples_per_second": 162.475,
      "eval_steps_per_second": 10.228,
      "step": 16558
    },
    {
      "epoch": 34.91,
      "learning_rate": 2.127310061601643e-05,
      "loss": 0.0155,
      "step": 17000
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.8221893561699387,
      "eval_f1": 0.6724347708530485,
      "eval_loss": 1.3412460088729858,
      "eval_precision": 0.6519003931847969,
      "eval_recall": 0.6943048576214406,
      "eval_runtime": 6.8473,
      "eval_samples_per_second": 162.4,
      "eval_steps_per_second": 10.223,
      "step": 17045
    },
    {
      "epoch": 35.93,
      "learning_rate": 2.1016427104722795e-05,
      "loss": 0.0141,
      "step": 17500
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.8219031034565015,
      "eval_f1": 0.6687940433452998,
      "eval_loss": 1.3676053285598755,
      "eval_precision": 0.6384869256156385,
      "eval_recall": 0.7021217197096594,
      "eval_runtime": 6.9597,
      "eval_samples_per_second": 159.777,
      "eval_steps_per_second": 10.058,
      "step": 17532
    },
    {
      "epoch": 36.96,
      "learning_rate": 2.0759753593429157e-05,
      "loss": 0.0145,
      "step": 18000
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.825075737697097,
      "eval_f1": 0.6776680643852293,
      "eval_loss": 1.3878055810928345,
      "eval_precision": 0.6573077932301233,
      "eval_recall": 0.6993299832495813,
      "eval_runtime": 6.8928,
      "eval_samples_per_second": 161.328,
      "eval_steps_per_second": 10.156,
      "step": 18019
    },
    {
      "epoch": 37.99,
      "learning_rate": 2.0503080082135523e-05,
      "loss": 0.0106,
      "step": 18500
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.8239307268433482,
      "eval_f1": 0.6637612255678816,
      "eval_loss": 1.4313843250274658,
      "eval_precision": 0.6298245614035087,
      "eval_recall": 0.7015633724176438,
      "eval_runtime": 6.9946,
      "eval_samples_per_second": 158.98,
      "eval_steps_per_second": 10.008,
      "step": 18506
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.8281529543665466,
      "eval_f1": 0.6862638851259821,
      "eval_loss": 1.3729307651519775,
      "eval_precision": 0.666578947368421,
      "eval_recall": 0.7071468453378001,
      "eval_runtime": 6.8415,
      "eval_samples_per_second": 162.538,
      "eval_steps_per_second": 10.232,
      "step": 18993
    },
    {
      "epoch": 39.01,
      "learning_rate": 2.024640657084189e-05,
      "loss": 0.0112,
      "step": 19000
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.8282960807232652,
      "eval_f1": 0.6758787228333781,
      "eval_loss": 1.3454548120498657,
      "eval_precision": 0.6505681818181818,
      "eval_recall": 0.7032384142936907,
      "eval_runtime": 6.9483,
      "eval_samples_per_second": 160.038,
      "eval_steps_per_second": 10.074,
      "step": 19480
    },
    {
      "epoch": 40.04,
      "learning_rate": 1.9989733059548256e-05,
      "loss": 0.0109,
      "step": 19500
    },
    {
      "epoch": 41.0,
      "eval_accuracy": 0.8278189928675366,
      "eval_f1": 0.6729643427354975,
      "eval_loss": 1.3883967399597168,
      "eval_precision": 0.6428571428571429,
      "eval_recall": 0.7060301507537688,
      "eval_runtime": 6.8639,
      "eval_samples_per_second": 162.007,
      "eval_steps_per_second": 10.198,
      "step": 19967
    },
    {
      "epoch": 41.07,
      "learning_rate": 1.973305954825462e-05,
      "loss": 0.0084,
      "step": 20000
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.8254812623744663,
      "eval_f1": 0.6738408396439485,
      "eval_loss": 1.4240161180496216,
      "eval_precision": 0.6428390367553866,
      "eval_recall": 0.7079843662758235,
      "eval_runtime": 6.8987,
      "eval_samples_per_second": 161.191,
      "eval_steps_per_second": 10.147,
      "step": 20454
    },
    {
      "epoch": 42.09,
      "learning_rate": 1.9476386036960984e-05,
      "loss": 0.0082,
      "step": 20500
    },
    {
      "epoch": 43.0,
      "eval_accuracy": 0.8176331671477303,
      "eval_f1": 0.6438536456962358,
      "eval_loss": 1.4279741048812866,
      "eval_precision": 0.6090637450199203,
      "eval_recall": 0.68285873813512,
      "eval_runtime": 6.884,
      "eval_samples_per_second": 161.533,
      "eval_steps_per_second": 10.168,
      "step": 20941
    },
    {
      "epoch": 43.12,
      "learning_rate": 1.921971252566735e-05,
      "loss": 0.0122,
      "step": 21000
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.8239068724505618,
      "eval_f1": 0.6773326162947029,
      "eval_loss": 1.4722952842712402,
      "eval_precision": 0.6532676348547718,
      "eval_recall": 0.7032384142936907,
      "eval_runtime": 6.8269,
      "eval_samples_per_second": 162.886,
      "eval_steps_per_second": 10.254,
      "step": 21428
    },
    {
      "epoch": 44.15,
      "learning_rate": 1.896303901437372e-05,
      "loss": 0.0082,
      "step": 21500
    },
    {
      "epoch": 45.0,
      "eval_accuracy": 0.8180386918250996,
      "eval_f1": 0.6551905387647833,
      "eval_loss": 1.5150511264801025,
      "eval_precision": 0.6189175769612711,
      "eval_recall": 0.6959798994974874,
      "eval_runtime": 6.807,
      "eval_samples_per_second": 163.361,
      "eval_steps_per_second": 10.284,
      "step": 21915
    },
    {
      "epoch": 45.17,
      "learning_rate": 1.8706365503080083e-05,
      "loss": 0.0068,
      "step": 22000
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.8210681997089764,
      "eval_f1": 0.6669309023649095,
      "eval_loss": 1.4440630674362183,
      "eval_precision": 0.6330574366691748,
      "eval_recall": 0.7046342825237297,
      "eval_runtime": 6.8346,
      "eval_samples_per_second": 162.702,
      "eval_steps_per_second": 10.242,
      "step": 22402
    },
    {
      "epoch": 46.2,
      "learning_rate": 1.844969199178645e-05,
      "loss": 0.0074,
      "step": 22500
    },
    {
      "epoch": 47.0,
      "eval_accuracy": 0.8202571503542377,
      "eval_f1": 0.6726807593914097,
      "eval_loss": 1.4752631187438965,
      "eval_precision": 0.6496749024707412,
      "eval_recall": 0.6973757677275265,
      "eval_runtime": 6.7986,
      "eval_samples_per_second": 163.564,
      "eval_steps_per_second": 10.296,
      "step": 22889
    },
    {
      "epoch": 47.23,
      "learning_rate": 1.8193018480492815e-05,
      "loss": 0.0076,
      "step": 23000
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.8214975787791322,
      "eval_f1": 0.6728770082354529,
      "eval_loss": 1.5147960186004639,
      "eval_precision": 0.6515032679738562,
      "eval_recall": 0.6957007258514796,
      "eval_runtime": 6.8796,
      "eval_samples_per_second": 161.637,
      "eval_steps_per_second": 10.175,
      "step": 23376
    },
    {
      "epoch": 48.25,
      "learning_rate": 1.793634496919918e-05,
      "loss": 0.0098,
      "step": 23500
    },
    {
      "epoch": 49.0,
      "eval_accuracy": 0.8233343670236875,
      "eval_f1": 0.6630391506303915,
      "eval_loss": 1.4481282234191895,
      "eval_precision": 0.6319251201619024,
      "eval_recall": 0.6973757677275265,
      "eval_runtime": 6.9541,
      "eval_samples_per_second": 159.906,
      "eval_steps_per_second": 10.066,
      "step": 23863
    },
    {
      "epoch": 49.28,
      "learning_rate": 1.7679671457905544e-05,
      "loss": 0.0104,
      "step": 24000
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.823477493380406,
      "eval_f1": 0.682099596231494,
      "eval_loss": 1.4813882112503052,
      "eval_precision": 0.6585239085239085,
      "eval_recall": 0.707426018983808,
      "eval_runtime": 7.0699,
      "eval_samples_per_second": 157.287,
      "eval_steps_per_second": 9.901,
      "step": 24350
    },
    {
      "epoch": 50.31,
      "learning_rate": 1.742299794661191e-05,
      "loss": 0.0119,
      "step": 24500
    },
    {
      "epoch": 51.0,
      "eval_accuracy": 0.8263638749075642,
      "eval_f1": 0.6831550802139038,
      "eval_loss": 1.4050037860870361,
      "eval_precision": 0.6554643406875321,
      "eval_recall": 0.7132886655499721,
      "eval_runtime": 6.9082,
      "eval_samples_per_second": 160.968,
      "eval_steps_per_second": 10.133,
      "step": 24837
    },
    {
      "epoch": 51.33,
      "learning_rate": 1.7166324435318276e-05,
      "loss": 0.0078,
      "step": 25000
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.8233582214164739,
      "eval_f1": 0.6825246654953372,
      "eval_loss": 1.4853544235229492,
      "eval_precision": 0.6615142782289757,
      "eval_recall": 0.7049134561697376,
      "eval_runtime": 6.9587,
      "eval_samples_per_second": 159.8,
      "eval_steps_per_second": 10.059,
      "step": 25324
    },
    {
      "epoch": 52.36,
      "learning_rate": 1.6909650924024642e-05,
      "loss": 0.007,
      "step": 25500
    },
    {
      "epoch": 53.0,
      "eval_accuracy": 0.8203764223181699,
      "eval_f1": 0.6733681812089531,
      "eval_loss": 1.4941233396530151,
      "eval_precision": 0.6475895849445733,
      "eval_recall": 0.701284198771636,
      "eval_runtime": 6.8301,
      "eval_samples_per_second": 162.81,
      "eval_steps_per_second": 10.249,
      "step": 25811
    },
    {
      "epoch": 53.39,
      "learning_rate": 1.6652977412731005e-05,
      "loss": 0.0079,
      "step": 26000
    },
    {
      "epoch": 54.0,
      "eval_accuracy": 0.8228095703823859,
      "eval_f1": 0.6796948199705528,
      "eval_loss": 1.4137930870056152,
      "eval_precision": 0.652867060941116,
      "eval_recall": 0.708821887213847,
      "eval_runtime": 7.1554,
      "eval_samples_per_second": 155.406,
      "eval_steps_per_second": 9.783,
      "step": 26298
    },
    {
      "epoch": 54.41,
      "learning_rate": 1.639630390143737e-05,
      "loss": 0.0092,
      "step": 26500
    },
    {
      "epoch": 55.0,
      "eval_accuracy": 0.8217599770997829,
      "eval_f1": 0.6887671232876712,
      "eval_loss": 1.4301496744155884,
      "eval_precision": 0.6761699838622915,
      "eval_recall": 0.7018425460636516,
      "eval_runtime": 7.1182,
      "eval_samples_per_second": 156.219,
      "eval_steps_per_second": 9.834,
      "step": 26785
    },
    {
      "epoch": 55.44,
      "learning_rate": 1.6139630390143737e-05,
      "loss": 0.0097,
      "step": 27000
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.8181579637890317,
      "eval_f1": 0.6752263819435058,
      "eval_loss": 1.5275614261627197,
      "eval_precision": 0.654440660204349,
      "eval_recall": 0.6973757677275265,
      "eval_runtime": 7.0427,
      "eval_samples_per_second": 157.894,
      "eval_steps_per_second": 9.939,
      "step": 27272
    },
    {
      "epoch": 56.47,
      "learning_rate": 1.5882956878850103e-05,
      "loss": 0.0076,
      "step": 27500
    },
    {
      "epoch": 57.0,
      "eval_accuracy": 0.8257913694806899,
      "eval_f1": 0.6765140324963074,
      "eval_loss": 1.430177927017212,
      "eval_precision": 0.6517464424320828,
      "eval_recall": 0.7032384142936907,
      "eval_runtime": 6.9201,
      "eval_samples_per_second": 160.691,
      "eval_steps_per_second": 10.115,
      "step": 27759
    },
    {
      "epoch": 57.49,
      "learning_rate": 1.5626283367556466e-05,
      "loss": 0.0056,
      "step": 28000
    },
    {
      "epoch": 58.0,
      "eval_accuracy": 0.8264831468714964,
      "eval_f1": 0.6855901127257911,
      "eval_loss": 1.4996072053909302,
      "eval_precision": 0.6675482676540597,
      "eval_recall": 0.7046342825237297,
      "eval_runtime": 6.8256,
      "eval_samples_per_second": 162.916,
      "eval_steps_per_second": 10.256,
      "step": 28246
    },
    {
      "epoch": 58.52,
      "learning_rate": 1.5369609856262832e-05,
      "loss": 0.0047,
      "step": 28500
    },
    {
      "epoch": 59.0,
      "eval_accuracy": 0.8240977075928533,
      "eval_f1": 0.6822860238353197,
      "eval_loss": 1.4308714866638184,
      "eval_precision": 0.6625460284061021,
      "eval_recall": 0.7032384142936907,
      "eval_runtime": 6.8106,
      "eval_samples_per_second": 163.274,
      "eval_steps_per_second": 10.278,
      "step": 28733
    },
    {
      "epoch": 59.55,
      "learning_rate": 1.5112936344969198e-05,
      "loss": 0.0126,
      "step": 29000
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.817179933684788,
      "eval_f1": 0.671845700508974,
      "eval_loss": 1.490278720855713,
      "eval_precision": 0.6457260556127703,
      "eval_recall": 0.7001675041876047,
      "eval_runtime": 6.8728,
      "eval_samples_per_second": 161.798,
      "eval_steps_per_second": 10.185,
      "step": 29220
    },
    {
      "epoch": 60.57,
      "learning_rate": 1.4856262833675564e-05,
      "loss": 0.0054,
      "step": 29500
    },
    {
      "epoch": 61.0,
      "eval_accuracy": 0.8218315402781422,
      "eval_f1": 0.6701236537694455,
      "eval_loss": 1.4318019151687622,
      "eval_precision": 0.6397562833206397,
      "eval_recall": 0.7035175879396985,
      "eval_runtime": 6.8056,
      "eval_samples_per_second": 163.394,
      "eval_steps_per_second": 10.286,
      "step": 29707
    },
    {
      "epoch": 61.6,
      "learning_rate": 1.459958932238193e-05,
      "loss": 0.0076,
      "step": 30000
    },
    {
      "epoch": 62.0,
      "eval_accuracy": 0.8196369361417906,
      "eval_f1": 0.6820163487738419,
      "eval_loss": 1.574474811553955,
      "eval_precision": 0.6660457690260777,
      "eval_recall": 0.6987716359575656,
      "eval_runtime": 6.995,
      "eval_samples_per_second": 158.97,
      "eval_steps_per_second": 10.007,
      "step": 30194
    },
    {
      "epoch": 62.63,
      "learning_rate": 1.4342915811088295e-05,
      "loss": 0.0043,
      "step": 30500
    },
    {
      "epoch": 63.0,
      "eval_accuracy": 0.82679325397772,
      "eval_f1": 0.6825053995680346,
      "eval_loss": 1.5101566314697266,
      "eval_precision": 0.6607422895974908,
      "eval_recall": 0.705750977107761,
      "eval_runtime": 6.9716,
      "eval_samples_per_second": 159.503,
      "eval_steps_per_second": 10.041,
      "step": 30681
    },
    {
      "epoch": 63.66,
      "learning_rate": 1.4086242299794661e-05,
      "loss": 0.0046,
      "step": 31000
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.8252427184466019,
      "eval_f1": 0.6798688882818902,
      "eval_loss": 1.5499790906906128,
      "eval_precision": 0.6655080213903743,
      "eval_recall": 0.6948632049134562,
      "eval_runtime": 6.9216,
      "eval_samples_per_second": 160.656,
      "eval_steps_per_second": 10.113,
      "step": 31168
    },
    {
      "epoch": 64.68,
      "learning_rate": 1.3829568788501027e-05,
      "loss": 0.0042,
      "step": 31500
    },
    {
      "epoch": 65.0,
      "eval_accuracy": 0.8273896137973807,
      "eval_f1": 0.6834157423493251,
      "eval_loss": 1.5357410907745361,
      "eval_precision": 0.6554729556523968,
      "eval_recall": 0.7138470128419877,
      "eval_runtime": 6.9877,
      "eval_samples_per_second": 159.137,
      "eval_steps_per_second": 10.018,
      "step": 31655
    },
    {
      "epoch": 65.71,
      "learning_rate": 1.3572895277207393e-05,
      "loss": 0.0039,
      "step": 32000
    },
    {
      "epoch": 66.0,
      "eval_accuracy": 0.8219985210276473,
      "eval_f1": 0.6870022945066812,
      "eval_loss": 1.5469226837158203,
      "eval_precision": 0.665011758557617,
      "eval_recall": 0.7104969290898939,
      "eval_runtime": 6.9394,
      "eval_samples_per_second": 160.243,
      "eval_steps_per_second": 10.087,
      "step": 32142
    },
    {
      "epoch": 66.74,
      "learning_rate": 1.331622176591376e-05,
      "loss": 0.004,
      "step": 32500
    },
    {
      "epoch": 67.0,
      "eval_accuracy": 0.8289162949357124,
      "eval_f1": 0.6831220813875917,
      "eval_loss": 1.481360673904419,
      "eval_precision": 0.6542294914387937,
      "eval_recall": 0.7146845337800112,
      "eval_runtime": 6.9709,
      "eval_samples_per_second": 159.521,
      "eval_steps_per_second": 10.042,
      "step": 32629
    },
    {
      "epoch": 67.76,
      "learning_rate": 1.3059548254620124e-05,
      "loss": 0.0031,
      "step": 33000
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.825028028911524,
      "eval_f1": 0.68095365657648,
      "eval_loss": 1.5209916830062866,
      "eval_precision": 0.6544799176107106,
      "eval_recall": 0.7096594081518705,
      "eval_runtime": 7.0091,
      "eval_samples_per_second": 158.65,
      "eval_steps_per_second": 9.987,
      "step": 33116
    },
    {
      "epoch": 68.79,
      "learning_rate": 1.280287474332649e-05,
      "loss": 0.0047,
      "step": 33500
    },
    {
      "epoch": 69.0,
      "eval_accuracy": 0.8271510698695165,
      "eval_f1": 0.6805257510729614,
      "eval_loss": 1.5326005220413208,
      "eval_precision": 0.6548786783686112,
      "eval_recall": 0.7082635399218313,
      "eval_runtime": 6.9749,
      "eval_samples_per_second": 159.429,
      "eval_steps_per_second": 10.036,
      "step": 33603
    },
    {
      "epoch": 69.82,
      "learning_rate": 1.2546201232032854e-05,
      "loss": 0.0029,
      "step": 34000
    },
    {
      "epoch": 70.0,
      "eval_accuracy": 0.8225710264545216,
      "eval_f1": 0.6829466829466829,
      "eval_loss": 1.6056736707687378,
      "eval_precision": 0.6642913697545526,
      "eval_recall": 0.7026800670016751,
      "eval_runtime": 7.4038,
      "eval_samples_per_second": 150.193,
      "eval_steps_per_second": 9.455,
      "step": 34090
    },
    {
      "epoch": 70.84,
      "learning_rate": 1.228952772073922e-05,
      "loss": 0.0027,
      "step": 34500
    },
    {
      "epoch": 71.0,
      "eval_accuracy": 0.8255289711600391,
      "eval_f1": 0.6856989679667604,
      "eval_loss": 1.5919586420059204,
      "eval_precision": 0.6594483114204692,
      "eval_recall": 0.7141261864879955,
      "eval_runtime": 6.9324,
      "eval_samples_per_second": 160.406,
      "eval_steps_per_second": 10.098,
      "step": 34577
    },
    {
      "epoch": 71.87,
      "learning_rate": 1.2032854209445585e-05,
      "loss": 0.0016,
      "step": 35000
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.8255051167672527,
      "eval_f1": 0.6841604350781781,
      "eval_loss": 1.6220448017120361,
      "eval_precision": 0.6668433607209118,
      "eval_recall": 0.7024008933556672,
      "eval_runtime": 6.9179,
      "eval_samples_per_second": 160.743,
      "eval_steps_per_second": 10.119,
      "step": 35064
    },
    {
      "epoch": 72.9,
      "learning_rate": 1.1776180698151951e-05,
      "loss": 0.0025,
      "step": 35500
    },
    {
      "epoch": 73.0,
      "eval_accuracy": 0.8238591636649889,
      "eval_f1": 0.6912936006591596,
      "eval_loss": 1.626072645187378,
      "eval_precision": 0.6802702702702703,
      "eval_recall": 0.7026800670016751,
      "eval_runtime": 7.3201,
      "eval_samples_per_second": 151.911,
      "eval_steps_per_second": 9.563,
      "step": 35551
    },
    {
      "epoch": 73.92,
      "learning_rate": 1.1519507186858315e-05,
      "loss": 0.0037,
      "step": 36000
    },
    {
      "epoch": 74.0,
      "eval_accuracy": 0.8206865294243935,
      "eval_f1": 0.6906455142231948,
      "eval_loss": 1.6439733505249023,
      "eval_precision": 0.6769436997319035,
      "eval_recall": 0.7049134561697376,
      "eval_runtime": 6.8715,
      "eval_samples_per_second": 161.827,
      "eval_steps_per_second": 10.187,
      "step": 36038
    },
    {
      "epoch": 74.95,
      "learning_rate": 1.1262833675564683e-05,
      "loss": 0.003,
      "step": 36500
    },
    {
      "epoch": 75.0,
      "eval_accuracy": 0.8263400205147778,
      "eval_f1": 0.6881282260255366,
      "eval_loss": 1.6027307510375977,
      "eval_precision": 0.6701058201058201,
      "eval_recall": 0.7071468453378001,
      "eval_runtime": 7.2528,
      "eval_samples_per_second": 153.321,
      "eval_steps_per_second": 9.651,
      "step": 36525
    },
    {
      "epoch": 75.98,
      "learning_rate": 1.1006160164271048e-05,
      "loss": 0.0031,
      "step": 37000
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.8261730397652728,
      "eval_f1": 0.6897667520560874,
      "eval_loss": 1.6012972593307495,
      "eval_precision": 0.6670143415906128,
      "eval_recall": 0.7141261864879955,
      "eval_runtime": 6.9267,
      "eval_samples_per_second": 160.539,
      "eval_steps_per_second": 10.106,
      "step": 37012
    },
    {
      "epoch": 77.0,
      "eval_accuracy": 0.8184919252880418,
      "eval_f1": 0.6771591059383678,
      "eval_loss": 1.6714273691177368,
      "eval_precision": 0.6433777330987686,
      "eval_recall": 0.7146845337800112,
      "eval_runtime": 7.0081,
      "eval_samples_per_second": 158.674,
      "eval_steps_per_second": 9.988,
      "step": 37499
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.0749486652977414e-05,
      "loss": 0.002,
      "step": 37500
    },
    {
      "epoch": 78.0,
      "eval_accuracy": 0.8266501276210014,
      "eval_f1": 0.6862638851259821,
      "eval_loss": 1.6293165683746338,
      "eval_precision": 0.666578947368421,
      "eval_recall": 0.7071468453378001,
      "eval_runtime": 6.9855,
      "eval_samples_per_second": 159.186,
      "eval_steps_per_second": 10.021,
      "step": 37986
    },
    {
      "epoch": 78.03,
      "learning_rate": 1.0492813141683778e-05,
      "loss": 0.0024,
      "step": 38000
    },
    {
      "epoch": 79.0,
      "eval_accuracy": 0.8221893561699387,
      "eval_f1": 0.6826057756883814,
      "eval_loss": 1.679621934890747,
      "eval_precision": 0.6577789282940719,
      "eval_recall": 0.7093802345058626,
      "eval_runtime": 7.0739,
      "eval_samples_per_second": 157.199,
      "eval_steps_per_second": 9.896,
      "step": 38473
    },
    {
      "epoch": 79.06,
      "learning_rate": 1.0236139630390144e-05,
      "loss": 0.003,
      "step": 38500
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.828343789508838,
      "eval_f1": 0.6891781936533767,
      "eval_loss": 1.6462854146957397,
      "eval_precision": 0.6700949367088608,
      "eval_recall": 0.7093802345058626,
      "eval_runtime": 6.9376,
      "eval_samples_per_second": 160.285,
      "eval_steps_per_second": 10.09,
      "step": 38960
    },
    {
      "epoch": 80.08,
      "learning_rate": 9.979466119096509e-06,
      "loss": 0.0015,
      "step": 39000
    },
    {
      "epoch": 81.0,
      "eval_accuracy": 0.826626273228215,
      "eval_f1": 0.6915938864628821,
      "eval_loss": 1.6634407043457031,
      "eval_precision": 0.6764548852108916,
      "eval_recall": 0.707426018983808,
      "eval_runtime": 6.9101,
      "eval_samples_per_second": 160.923,
      "eval_steps_per_second": 10.13,
      "step": 39447
    },
    {
      "epoch": 81.11,
      "learning_rate": 9.722792607802875e-06,
      "loss": 0.003,
      "step": 39500
    },
    {
      "epoch": 82.0,
      "eval_accuracy": 0.8255051167672527,
      "eval_f1": 0.6838971583220568,
      "eval_loss": 1.6947256326675415,
      "eval_precision": 0.6636029411764706,
      "eval_recall": 0.7054718034617532,
      "eval_runtime": 6.9976,
      "eval_samples_per_second": 158.913,
      "eval_steps_per_second": 10.003,
      "step": 39934
    },
    {
      "epoch": 82.14,
      "learning_rate": 9.46611909650924e-06,
      "loss": 0.0036,
      "step": 40000
    },
    {
      "epoch": 83.0,
      "eval_accuracy": 0.8227141528112402,
      "eval_f1": 0.6791335934346832,
      "eval_loss": 1.651548981666565,
      "eval_precision": 0.6554141781355493,
      "eval_recall": 0.7046342825237297,
      "eval_runtime": 7.0674,
      "eval_samples_per_second": 157.343,
      "eval_steps_per_second": 9.905,
      "step": 40421
    },
    {
      "epoch": 83.16,
      "learning_rate": 9.209445585215607e-06,
      "loss": 0.0018,
      "step": 40500
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.8266024188354285,
      "eval_f1": 0.6863617968433833,
      "eval_loss": 1.6855460405349731,
      "eval_precision": 0.6640563821456539,
      "eval_recall": 0.7102177554438861,
      "eval_runtime": 7.0529,
      "eval_samples_per_second": 157.667,
      "eval_steps_per_second": 9.925,
      "step": 40908
    },
    {
      "epoch": 84.19,
      "learning_rate": 8.952772073921972e-06,
      "loss": 0.0012,
      "step": 41000
    },
    {
      "epoch": 85.0,
      "eval_accuracy": 0.8240738532000668,
      "eval_f1": 0.6814775160599572,
      "eval_loss": 1.6966240406036377,
      "eval_precision": 0.6544987146529563,
      "eval_recall": 0.7107761027359017,
      "eval_runtime": 6.8863,
      "eval_samples_per_second": 161.48,
      "eval_steps_per_second": 10.165,
      "step": 41395
    },
    {
      "epoch": 85.22,
      "learning_rate": 8.696098562628338e-06,
      "loss": 0.0019,
      "step": 41500
    },
    {
      "epoch": 86.0,
      "eval_accuracy": 0.8255051167672527,
      "eval_f1": 0.6833355858899851,
      "eval_loss": 1.656359314918518,
      "eval_precision": 0.6623002357872675,
      "eval_recall": 0.705750977107761,
      "eval_runtime": 6.9513,
      "eval_samples_per_second": 159.971,
      "eval_steps_per_second": 10.07,
      "step": 41882
    },
    {
      "epoch": 86.24,
      "learning_rate": 8.439425051334702e-06,
      "loss": 0.0015,
      "step": 42000
    },
    {
      "epoch": 87.0,
      "eval_accuracy": 0.8239307268433482,
      "eval_f1": 0.6778030201790726,
      "eval_loss": 1.6362555027008057,
      "eval_precision": 0.6500897205844656,
      "eval_recall": 0.7079843662758235,
      "eval_runtime": 7.2734,
      "eval_samples_per_second": 152.885,
      "eval_steps_per_second": 9.624,
      "step": 42369
    },
    {
      "epoch": 87.27,
      "learning_rate": 8.182751540041068e-06,
      "loss": 0.0022,
      "step": 42500
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.8260299134085541,
      "eval_f1": 0.6931833767658757,
      "eval_loss": 1.6878582239151,
      "eval_precision": 0.6813157185225128,
      "eval_recall": 0.7054718034617532,
      "eval_runtime": 7.1154,
      "eval_samples_per_second": 156.28,
      "eval_steps_per_second": 9.838,
      "step": 42856
    },
    {
      "epoch": 88.3,
      "learning_rate": 7.926078028747433e-06,
      "loss": 0.0011,
      "step": 43000
    },
    {
      "epoch": 89.0,
      "eval_accuracy": 0.829441091577014,
      "eval_f1": 0.6879049676025918,
      "eval_loss": 1.686987042427063,
      "eval_precision": 0.6659696811291166,
      "eval_recall": 0.7113344500279174,
      "eval_runtime": 6.9712,
      "eval_samples_per_second": 159.514,
      "eval_steps_per_second": 10.041,
      "step": 43343
    },
    {
      "epoch": 89.32,
      "learning_rate": 7.669404517453799e-06,
      "loss": 0.0017,
      "step": 43500
    },
    {
      "epoch": 90.0,
      "eval_accuracy": 0.8275804489396722,
      "eval_f1": 0.6870062653228004,
      "eval_loss": 1.7017887830734253,
      "eval_precision": 0.6707446808510639,
      "eval_recall": 0.7040759352317141,
      "eval_runtime": 6.9862,
      "eval_samples_per_second": 159.17,
      "eval_steps_per_second": 10.02,
      "step": 43830
    },
    {
      "epoch": 90.35,
      "learning_rate": 7.412731006160165e-06,
      "loss": 0.0016,
      "step": 44000
    },
    {
      "epoch": 91.0,
      "eval_accuracy": 0.8280813911881874,
      "eval_f1": 0.6910074374577416,
      "eval_loss": 1.6699299812316895,
      "eval_precision": 0.6700760555992656,
      "eval_recall": 0.7132886655499721,
      "eval_runtime": 6.9237,
      "eval_samples_per_second": 160.607,
      "eval_steps_per_second": 10.11,
      "step": 44317
    },
    {
      "epoch": 91.38,
      "learning_rate": 7.15605749486653e-06,
      "loss": 0.0015,
      "step": 44500
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.8320412203907349,
      "eval_f1": 0.694421768707483,
      "eval_loss": 1.6736540794372559,
      "eval_precision": 0.6772823779193206,
      "eval_recall": 0.7124511446119486,
      "eval_runtime": 7.0271,
      "eval_samples_per_second": 158.244,
      "eval_steps_per_second": 9.961,
      "step": 44804
    },
    {
      "epoch": 92.4,
      "learning_rate": 6.899383983572895e-06,
      "loss": 0.0017,
      "step": 45000
    },
    {
      "epoch": 93.0,
      "eval_accuracy": 0.8280336824026144,
      "eval_f1": 0.6972650961278094,
      "eval_loss": 1.727142572402954,
      "eval_precision": 0.676919032597266,
      "eval_recall": 0.7188721384701284,
      "eval_runtime": 7.0065,
      "eval_samples_per_second": 158.709,
      "eval_steps_per_second": 9.991,
      "step": 45291
    },
    {
      "epoch": 93.43,
      "learning_rate": 6.642710472279261e-06,
      "loss": 0.0005,
      "step": 45500
    },
    {
      "epoch": 94.0,
      "eval_accuracy": 0.8260537678013407,
      "eval_f1": 0.6882329154872624,
      "eval_loss": 1.7245275974273682,
      "eval_precision": 0.6653635652853792,
      "eval_recall": 0.7127303182579564,
      "eval_runtime": 6.9585,
      "eval_samples_per_second": 159.804,
      "eval_steps_per_second": 10.06,
      "step": 45778
    },
    {
      "epoch": 94.46,
      "learning_rate": 6.386036960985627e-06,
      "loss": 0.0013,
      "step": 46000
    },
    {
      "epoch": 95.0,
      "eval_accuracy": 0.8235013477731924,
      "eval_f1": 0.6909190371991247,
      "eval_loss": 1.8143095970153809,
      "eval_precision": 0.6772117962466488,
      "eval_recall": 0.7051926298157454,
      "eval_runtime": 6.8559,
      "eval_samples_per_second": 162.196,
      "eval_steps_per_second": 10.21,
      "step": 46265
    },
    {
      "epoch": 95.48,
      "learning_rate": 6.129363449691992e-06,
      "loss": 0.0012,
      "step": 46500
    },
    {
      "epoch": 96.0,
      "eval_accuracy": 0.8262207485508456,
      "eval_f1": 0.6908744730042159,
      "eval_loss": 1.7299076318740845,
      "eval_precision": 0.673561389551843,
      "eval_recall": 0.7091010608598548,
      "eval_runtime": 7.0371,
      "eval_samples_per_second": 158.02,
      "eval_steps_per_second": 9.947,
      "step": 46752
    },
    {
      "epoch": 96.51,
      "learning_rate": 5.872689938398357e-06,
      "loss": 0.002,
      "step": 47000
    },
    {
      "epoch": 97.0,
      "eval_accuracy": 0.8272703418334486,
      "eval_f1": 0.6936667572709975,
      "eval_loss": 1.7251330614089966,
      "eval_precision": 0.6758474576271186,
      "eval_recall": 0.7124511446119486,
      "eval_runtime": 6.9526,
      "eval_samples_per_second": 159.94,
      "eval_steps_per_second": 10.068,
      "step": 47239
    },
    {
      "epoch": 97.54,
      "learning_rate": 5.6160164271047226e-06,
      "loss": 0.0009,
      "step": 47500
    },
    {
      "epoch": 98.0,
      "eval_accuracy": 0.8262207485508456,
      "eval_f1": 0.6860418610851887,
      "eval_loss": 1.718263864517212,
      "eval_precision": 0.6565450369992345,
      "eval_recall": 0.7183137911781128,
      "eval_runtime": 6.9474,
      "eval_samples_per_second": 160.059,
      "eval_steps_per_second": 10.076,
      "step": 47726
    },
    {
      "epoch": 98.56,
      "learning_rate": 5.359342915811089e-06,
      "loss": 0.0009,
      "step": 48000
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.8278905560458959,
      "eval_f1": 0.6933224534203727,
      "eval_loss": 1.7801190614700317,
      "eval_precision": 0.6759480243967118,
      "eval_recall": 0.7116136236739252,
      "eval_runtime": 6.9784,
      "eval_samples_per_second": 159.35,
      "eval_steps_per_second": 10.031,
      "step": 48213
    },
    {
      "epoch": 99.59,
      "learning_rate": 5.102669404517454e-06,
      "loss": 0.0008,
      "step": 48500
    },
    {
      "epoch": 100.0,
      "eval_accuracy": 0.8263400205147778,
      "eval_f1": 0.6959136258029246,
      "eval_loss": 1.7749357223510742,
      "eval_precision": 0.6816599732262383,
      "eval_recall": 0.7107761027359017,
      "eval_runtime": 6.9549,
      "eval_samples_per_second": 159.888,
      "eval_steps_per_second": 10.065,
      "step": 48700
    },
    {
      "epoch": 100.62,
      "learning_rate": 4.845995893223819e-06,
      "loss": 0.0006,
      "step": 49000
    },
    {
      "epoch": 101.0,
      "eval_accuracy": 0.8272226330478758,
      "eval_f1": 0.6917334057282477,
      "eval_loss": 1.7413458824157715,
      "eval_precision": 0.6731836195508587,
      "eval_recall": 0.7113344500279174,
      "eval_runtime": 6.9416,
      "eval_samples_per_second": 160.194,
      "eval_steps_per_second": 10.084,
      "step": 49187
    },
    {
      "epoch": 101.64,
      "learning_rate": 4.5893223819301845e-06,
      "loss": 0.0005,
      "step": 49500
    },
    {
      "epoch": 102.0,
      "eval_accuracy": 0.826960234727225,
      "eval_f1": 0.6887363746467502,
      "eval_loss": 1.7939246892929077,
      "eval_precision": 0.6648480124707716,
      "eval_recall": 0.7144053601340034,
      "eval_runtime": 7.0029,
      "eval_samples_per_second": 158.792,
      "eval_steps_per_second": 9.996,
      "step": 49674
    },
    {
      "epoch": 102.67,
      "learning_rate": 4.332648870636551e-06,
      "loss": 0.0008,
      "step": 50000
    },
    {
      "epoch": 103.0,
      "eval_accuracy": 0.8237160373082704,
      "eval_f1": 0.6846774193548387,
      "eval_loss": 1.7955106496810913,
      "eval_precision": 0.6601866251944012,
      "eval_recall": 0.7110552763819096,
      "eval_runtime": 6.8701,
      "eval_samples_per_second": 161.86,
      "eval_steps_per_second": 10.189,
      "step": 50161
    },
    {
      "epoch": 103.7,
      "learning_rate": 4.075975359342916e-06,
      "loss": 0.0007,
      "step": 50500
    },
    {
      "epoch": 104.0,
      "eval_accuracy": 0.826626273228215,
      "eval_f1": 0.6900837611456363,
      "eval_loss": 1.78438401222229,
      "eval_precision": 0.668586387434555,
      "eval_recall": 0.7130094919039642,
      "eval_runtime": 7.0026,
      "eval_samples_per_second": 158.798,
      "eval_steps_per_second": 9.996,
      "step": 50648
    },
    {
      "epoch": 104.72,
      "learning_rate": 3.819301848049281e-06,
      "loss": 0.0005,
      "step": 51000
    },
    {
      "epoch": 105.0,
      "eval_accuracy": 0.8279382648314687,
      "eval_f1": 0.6963993453355155,
      "eval_loss": 1.798330307006836,
      "eval_precision": 0.6808,
      "eval_recall": 0.7127303182579564,
      "eval_runtime": 6.8685,
      "eval_samples_per_second": 161.899,
      "eval_steps_per_second": 10.191,
      "step": 51135
    },
    {
      "epoch": 105.75,
      "learning_rate": 3.5626283367556473e-06,
      "loss": 0.0004,
      "step": 51500
    },
    {
      "epoch": 106.0,
      "eval_accuracy": 0.8256243887311848,
      "eval_f1": 0.696007630467366,
      "eval_loss": 1.7945122718811035,
      "eval_precision": 0.6797977109395794,
      "eval_recall": 0.7130094919039642,
      "eval_runtime": 6.9592,
      "eval_samples_per_second": 159.789,
      "eval_steps_per_second": 10.059,
      "step": 51622
    },
    {
      "epoch": 106.78,
      "learning_rate": 3.3059548254620125e-06,
      "loss": 0.0005,
      "step": 52000
    },
    {
      "epoch": 107.0,
      "eval_accuracy": 0.8296796355048782,
      "eval_f1": 0.7018599562363238,
      "eval_loss": 1.8209294080734253,
      "eval_precision": 0.6879356568364611,
      "eval_recall": 0.7163595756560581,
      "eval_runtime": 7.0252,
      "eval_samples_per_second": 158.287,
      "eval_steps_per_second": 9.964,
      "step": 52109
    },
    {
      "epoch": 107.8,
      "learning_rate": 3.049281314168378e-06,
      "loss": 0.0004,
      "step": 52500
    },
    {
      "epoch": 108.0,
      "eval_accuracy": 0.8281052455809738,
      "eval_f1": 0.6960098724804608,
      "eval_loss": 1.8149752616882324,
      "eval_precision": 0.6839126919967664,
      "eval_recall": 0.7085427135678392,
      "eval_runtime": 7.0444,
      "eval_samples_per_second": 157.855,
      "eval_steps_per_second": 9.937,
      "step": 52596
    },
    {
      "epoch": 108.83,
      "learning_rate": 2.7926078028747435e-06,
      "loss": 0.0006,
      "step": 53000
    },
    {
      "epoch": 109.0,
      "eval_accuracy": 0.8287016054006345,
      "eval_f1": 0.6967024019541321,
      "eval_loss": 1.7784167528152466,
      "eval_precision": 0.6778452601003433,
      "eval_recall": 0.7166387493020658,
      "eval_runtime": 6.8299,
      "eval_samples_per_second": 162.815,
      "eval_steps_per_second": 10.249,
      "step": 53083
    },
    {
      "epoch": 109.86,
      "learning_rate": 2.5359342915811088e-06,
      "loss": 0.0009,
      "step": 53500
    },
    {
      "epoch": 110.0,
      "eval_accuracy": 0.8292741108275089,
      "eval_f1": 0.6964527484430002,
      "eval_loss": 1.7941442728042603,
      "eval_precision": 0.676130389064143,
      "eval_recall": 0.718034617532105,
      "eval_runtime": 7.0109,
      "eval_samples_per_second": 158.61,
      "eval_steps_per_second": 9.984,
      "step": 53570
    },
    {
      "epoch": 110.88,
      "learning_rate": 2.2792607802874745e-06,
      "loss": 0.0006,
      "step": 54000
    },
    {
      "epoch": 111.0,
      "eval_accuracy": 0.828009828009828,
      "eval_f1": 0.6974040021633315,
      "eval_loss": 1.807924509048462,
      "eval_precision": 0.6761929732564237,
      "eval_recall": 0.7199888330541597,
      "eval_runtime": 6.9069,
      "eval_samples_per_second": 160.999,
      "eval_steps_per_second": 10.135,
      "step": 54057
    },
    {
      "epoch": 111.91,
      "learning_rate": 2.0225872689938397e-06,
      "loss": 0.0006,
      "step": 54500
    },
    {
      "epoch": 112.0,
      "eval_accuracy": 0.8276520121180315,
      "eval_f1": 0.6952871072589383,
      "eval_loss": 1.796797513961792,
      "eval_precision": 0.6751709626512362,
      "eval_recall": 0.7166387493020658,
      "eval_runtime": 6.9819,
      "eval_samples_per_second": 159.269,
      "eval_steps_per_second": 10.026,
      "step": 54544
    },
    {
      "epoch": 112.94,
      "learning_rate": 1.7659137577002054e-06,
      "loss": 0.0003,
      "step": 55000
    },
    {
      "epoch": 113.0,
      "eval_accuracy": 0.8284630614727702,
      "eval_f1": 0.6953812813219559,
      "eval_loss": 1.7971785068511963,
      "eval_precision": 0.6753485924756643,
      "eval_recall": 0.7166387493020658,
      "eval_runtime": 7.1267,
      "eval_samples_per_second": 156.032,
      "eval_steps_per_second": 9.822,
      "step": 55031
    },
    {
      "epoch": 113.96,
      "learning_rate": 1.509240246406571e-06,
      "loss": 0.0003,
      "step": 55500
    },
    {
      "epoch": 114.0,
      "eval_accuracy": 0.8297989074688104,
      "eval_f1": 0.6962059620596205,
      "eval_loss": 1.7985066175460815,
      "eval_precision": 0.6764086361242759,
      "eval_recall": 0.7171970965940815,
      "eval_runtime": 7.0596,
      "eval_samples_per_second": 157.517,
      "eval_steps_per_second": 9.916,
      "step": 55518
    },
    {
      "epoch": 114.99,
      "learning_rate": 1.2525667351129362e-06,
      "loss": 0.0006,
      "step": 56000
    },
    {
      "epoch": 115.0,
      "eval_accuracy": 0.8286777510078481,
      "eval_f1": 0.695923066504131,
      "eval_loss": 1.8047789335250854,
      "eval_precision": 0.6758747697974218,
      "eval_recall": 0.7171970965940815,
      "eval_runtime": 6.9955,
      "eval_samples_per_second": 158.959,
      "eval_steps_per_second": 10.006,
      "step": 56005
    },
    {
      "epoch": 116.0,
      "eval_accuracy": 0.8298466162543833,
      "eval_f1": 0.6949681269496812,
      "eval_loss": 1.7985312938690186,
      "eval_precision": 0.6758111316275389,
      "eval_recall": 0.7152428810720268,
      "eval_runtime": 7.0552,
      "eval_samples_per_second": 157.615,
      "eval_steps_per_second": 9.922,
      "step": 56492
    },
    {
      "epoch": 116.02,
      "learning_rate": 9.958932238193019e-07,
      "loss": 0.0004,
      "step": 56500
    },
    {
      "epoch": 117.0,
      "eval_accuracy": 0.8314448605710741,
      "eval_f1": 0.6995637949836424,
      "eval_loss": 1.788254976272583,
      "eval_precision": 0.6835375599360682,
      "eval_recall": 0.7163595756560581,
      "eval_runtime": 7.0043,
      "eval_samples_per_second": 158.759,
      "eval_steps_per_second": 9.994,
      "step": 56979
    },
    {
      "epoch": 117.04,
      "learning_rate": 7.392197125256674e-07,
      "loss": 0.0009,
      "step": 57000
    },
    {
      "epoch": 118.0,
      "eval_accuracy": 0.8311347534648506,
      "eval_f1": 0.7000544365813827,
      "eval_loss": 1.785180687904358,
      "eval_precision": 0.6829527349973447,
      "eval_recall": 0.718034617532105,
      "eval_runtime": 6.9229,
      "eval_samples_per_second": 160.627,
      "eval_steps_per_second": 10.111,
      "step": 57466
    },
    {
      "epoch": 118.07,
      "learning_rate": 4.82546201232033e-07,
      "loss": 0.0002,
      "step": 57500
    },
    {
      "epoch": 119.0,
      "eval_accuracy": 0.8308962095369863,
      "eval_f1": 0.7012951601908658,
      "eval_loss": 1.7869131565093994,
      "eval_precision": 0.685318411937117,
      "eval_recall": 0.718034617532105,
      "eval_runtime": 7.1691,
      "eval_samples_per_second": 155.111,
      "eval_steps_per_second": 9.764,
      "step": 57953
    },
    {
      "epoch": 119.1,
      "learning_rate": 2.2587268993839835e-07,
      "loss": 0.0003,
      "step": 58000
    },
    {
      "epoch": 120.0,
      "eval_accuracy": 0.8312063166432099,
      "eval_f1": 0.7009129309170186,
      "eval_loss": 1.7865298986434937,
      "eval_precision": 0.6845887676337503,
      "eval_recall": 0.718034617532105,
      "eval_runtime": 7.0357,
      "eval_samples_per_second": 158.051,
      "eval_steps_per_second": 9.949,
      "step": 58440
    },
    {
      "epoch": 120.0,
      "step": 58440,
      "total_flos": 4.338269813081088e+17,
      "train_loss": 0.033376370980433,
      "train_runtime": 20083.166,
      "train_samples_per_second": 46.517,
      "train_steps_per_second": 2.91
    }
  ],
  "max_steps": 58440,
  "num_train_epochs": 120,
  "total_flos": 4.338269813081088e+17,
  "trial_name": null,
  "trial_params": null
}