{
  "best_metric": 0.9324748546966902,
  "best_model_checkpoint": "/home/nikola/projects/neuroticla/result/ner/xlmrb-sl_hr_sr_bs_mk_sq_cs_bg_pl_ru_sk_uk/checkpoint-433760",
  "epoch": 40.0,
  "eval_steps": 500,
  "global_step": 433760,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 1.95e-05,
      "loss": 0.1128,
      "step": 10844
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9769370487250145,
      "eval_f1": 0.9010081585905575,
      "eval_loss": 0.09421534836292267,
      "eval_precision": 0.9111733920553665,
      "eval_recall": 0.8910672334713056,
      "eval_runtime": 89.8183,
      "eval_samples_per_second": 301.776,
      "eval_steps_per_second": 15.097,
      "step": 10844
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.9e-05,
      "loss": 0.0676,
      "step": 21688
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9799811843427059,
      "eval_f1": 0.9143064683983713,
      "eval_loss": 0.08378221094608307,
      "eval_precision": 0.9133559397719819,
      "eval_recall": 0.9152589775142633,
      "eval_runtime": 89.3428,
      "eval_samples_per_second": 303.382,
      "eval_steps_per_second": 15.177,
      "step": 21688
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.8500000000000002e-05,
      "loss": 0.0507,
      "step": 32532
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9804305567434127,
      "eval_f1": 0.9163790332723132,
      "eval_loss": 0.08407577127218246,
      "eval_precision": 0.9157519899455383,
      "eval_recall": 0.9170069358988702,
      "eval_runtime": 90.9562,
      "eval_samples_per_second": 298.001,
      "eval_steps_per_second": 14.908,
      "step": 32532
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.8e-05,
      "loss": 0.0401,
      "step": 43376
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9814757099721534,
      "eval_f1": 0.9221617781244994,
      "eval_loss": 0.08777288347482681,
      "eval_precision": 0.9186871140101311,
      "eval_recall": 0.9256628258194429,
      "eval_runtime": 89.4976,
      "eval_samples_per_second": 302.857,
      "eval_steps_per_second": 15.151,
      "step": 43376
    },
    {
      "epoch": 5.0,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.032,
      "step": 54220
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9822106512855675,
      "eval_f1": 0.9250753982350196,
      "eval_loss": 0.09507809579372406,
      "eval_precision": 0.921564273720146,
      "eval_recall": 0.9286133795726591,
      "eval_runtime": 90.924,
      "eval_samples_per_second": 298.106,
      "eval_steps_per_second": 14.914,
      "step": 54220
    },
    {
      "epoch": 6.0,
      "learning_rate": 1.7e-05,
      "loss": 0.0264,
      "step": 65064
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9809553077403671,
      "eval_f1": 0.9212521094420008,
      "eval_loss": 0.1004280373454094,
      "eval_precision": 0.9113936176035908,
      "eval_recall": 0.9313262109855689,
      "eval_runtime": 90.6899,
      "eval_samples_per_second": 298.875,
      "eval_steps_per_second": 14.952,
      "step": 65064
    },
    {
      "epoch": 7.0,
      "learning_rate": 1.65e-05,
      "loss": 0.0215,
      "step": 75908
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9824875226679385,
      "eval_f1": 0.92642423737521,
      "eval_loss": 0.1009296178817749,
      "eval_precision": 0.9199227905694195,
      "eval_recall": 0.9330182347018682,
      "eval_runtime": 90.3482,
      "eval_samples_per_second": 300.006,
      "eval_steps_per_second": 15.009,
      "step": 75908
    },
    {
      "epoch": 8.0,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0177,
      "step": 86752
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9821120792750898,
      "eval_f1": 0.9245811025699784,
      "eval_loss": 0.11745402961969376,
      "eval_precision": 0.9183000675890037,
      "eval_recall": 0.9309486519744938,
      "eval_runtime": 89.2897,
      "eval_samples_per_second": 303.563,
      "eval_steps_per_second": 15.187,
      "step": 86752
    },
    {
      "epoch": 9.0,
      "learning_rate": 1.55e-05,
      "loss": 0.015,
      "step": 97596
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9825324599080091,
      "eval_f1": 0.926065441605738,
      "eval_loss": 0.11713194102048874,
      "eval_precision": 0.9205632012380479,
      "eval_recall": 0.9316338516612597,
      "eval_runtime": 89.7586,
      "eval_samples_per_second": 301.977,
      "eval_steps_per_second": 15.107,
      "step": 97596
    },
    {
      "epoch": 10.0,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.0133,
      "step": 108440
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9820569949162935,
      "eval_f1": 0.9250865136228494,
      "eval_loss": 0.13051354885101318,
      "eval_precision": 0.9212697796330453,
      "eval_recall": 0.9289350039154268,
      "eval_runtime": 89.1452,
      "eval_samples_per_second": 304.054,
      "eval_steps_per_second": 15.211,
      "step": 108440
    },
    {
      "epoch": 11.0,
      "learning_rate": 1.45e-05,
      "loss": 0.0112,
      "step": 119284
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9825614516757967,
      "eval_f1": 0.9272675473035418,
      "eval_loss": 0.1286834180355072,
      "eval_precision": 0.9215873147419164,
      "eval_recall": 0.9330182347018682,
      "eval_runtime": 89.577,
      "eval_samples_per_second": 302.589,
      "eval_steps_per_second": 15.138,
      "step": 119284
    },
    {
      "epoch": 12.0,
      "learning_rate": 1.4e-05,
      "loss": 0.01,
      "step": 130128
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.982439686251089,
      "eval_f1": 0.9262543854764159,
      "eval_loss": 0.13614533841609955,
      "eval_precision": 0.9221972721224219,
      "eval_recall": 0.9303473542901891,
      "eval_runtime": 88.3792,
      "eval_samples_per_second": 306.69,
      "eval_steps_per_second": 15.343,
      "step": 130128
    },
    {
      "epoch": 13.0,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.0086,
      "step": 140972
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9824309887207527,
      "eval_f1": 0.9259202794101342,
      "eval_loss": 0.13529813289642334,
      "eval_precision": 0.9177794263105836,
      "eval_recall": 0.9342068464034008,
      "eval_runtime": 88.5532,
      "eval_samples_per_second": 306.087,
      "eval_steps_per_second": 15.313,
      "step": 140972
    },
    {
      "epoch": 14.0,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.0079,
      "step": 151816
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9827180072218493,
      "eval_f1": 0.9273193405173312,
      "eval_loss": 0.14126552641391754,
      "eval_precision": 0.9215123103053067,
      "eval_recall": 0.9332000223738673,
      "eval_runtime": 89.1586,
      "eval_samples_per_second": 304.009,
      "eval_steps_per_second": 15.209,
      "step": 151816
    },
    {
      "epoch": 15.0,
      "learning_rate": 1.25e-05,
      "loss": 0.007,
      "step": 162660
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.982528111142841,
      "eval_f1": 0.9268309701909121,
      "eval_loss": 0.1539030820131302,
      "eval_precision": 0.9249721448467967,
      "eval_recall": 0.9286972815751202,
      "eval_runtime": 88.721,
      "eval_samples_per_second": 305.508,
      "eval_steps_per_second": 15.284,
      "step": 162660
    },
    {
      "epoch": 16.0,
      "learning_rate": 1.2e-05,
      "loss": 0.0064,
      "step": 173504
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.982581745913248,
      "eval_f1": 0.9264010897364617,
      "eval_loss": 0.15460434556007385,
      "eval_precision": 0.9208715596330275,
      "eval_recall": 0.9319974270052579,
      "eval_runtime": 88.6075,
      "eval_samples_per_second": 305.9,
      "eval_steps_per_second": 15.303,
      "step": 173504
    },
    {
      "epoch": 17.0,
      "learning_rate": 1.15e-05,
      "loss": 0.0057,
      "step": 184348
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.982916600831194,
      "eval_f1": 0.9276149465203501,
      "eval_loss": 0.1507822722196579,
      "eval_precision": 0.9214963846111387,
      "eval_recall": 0.9338153037252489,
      "eval_runtime": 89.1007,
      "eval_samples_per_second": 304.206,
      "eval_steps_per_second": 15.219,
      "step": 184348
    },
    {
      "epoch": 18.0,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.0053,
      "step": 195192
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.9827933858180969,
      "eval_f1": 0.927893329629501,
      "eval_loss": 0.15079163014888763,
      "eval_precision": 0.9216781866092739,
      "eval_recall": 0.934192862736324,
      "eval_runtime": 90.146,
      "eval_samples_per_second": 300.679,
      "eval_steps_per_second": 15.042,
      "step": 195192
    },
    {
      "epoch": 19.0,
      "learning_rate": 1.0500000000000001e-05,
      "loss": 0.0048,
      "step": 206036
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.9825150648473366,
      "eval_f1": 0.9270733168889446,
      "eval_loss": 0.16124233603477478,
      "eval_precision": 0.9239187800338879,
      "eval_recall": 0.930249468620651,
      "eval_runtime": 89.3088,
      "eval_samples_per_second": 303.497,
      "eval_steps_per_second": 15.183,
      "step": 206036
    },
    {
      "epoch": 20.0,
      "learning_rate": 1e-05,
      "loss": 0.0043,
      "step": 216880
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.9827440998128582,
      "eval_f1": 0.9272743748997957,
      "eval_loss": 0.1557987779378891,
      "eval_precision": 0.9244839808186809,
      "eval_recall": 0.9300816646157288,
      "eval_runtime": 88.3664,
      "eval_samples_per_second": 306.734,
      "eval_steps_per_second": 15.345,
      "step": 216880
    },
    {
      "epoch": 21.0,
      "learning_rate": 9.5e-06,
      "loss": 0.0041,
      "step": 227724
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.9828412222349464,
      "eval_f1": 0.92726855170398,
      "eval_loss": 0.1576606184244156,
      "eval_precision": 0.9229714190715008,
      "eval_recall": 0.9316058843271059,
      "eval_runtime": 88.9473,
      "eval_samples_per_second": 304.731,
      "eval_steps_per_second": 15.245,
      "step": 227724
    },
    {
      "epoch": 22.0,
      "learning_rate": 9e-06,
      "loss": 0.0036,
      "step": 238568
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.9831050473218129,
      "eval_f1": 0.9286320918900104,
      "eval_loss": 0.16860993206501007,
      "eval_precision": 0.9246028445479497,
      "eval_recall": 0.9326966103591006,
      "eval_runtime": 88.4212,
      "eval_samples_per_second": 306.544,
      "eval_steps_per_second": 15.336,
      "step": 238568
    },
    {
      "epoch": 23.0,
      "learning_rate": 8.5e-06,
      "loss": 0.0033,
      "step": 249412
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.9828832602982384,
      "eval_f1": 0.9276945586288581,
      "eval_loss": 0.16733527183532715,
      "eval_precision": 0.9244355577772224,
      "eval_recall": 0.9309766193086475,
      "eval_runtime": 88.9341,
      "eval_samples_per_second": 304.776,
      "eval_steps_per_second": 15.247,
      "step": 249412
    },
    {
      "epoch": 24.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0031,
      "step": 260256
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.9835123816592278,
      "eval_f1": 0.9298053519496189,
      "eval_loss": 0.16969779133796692,
      "eval_precision": 0.9274573913043478,
      "eval_recall": 0.9321652310101801,
      "eval_runtime": 87.785,
      "eval_samples_per_second": 308.766,
      "eval_steps_per_second": 15.447,
      "step": 260256
    },
    {
      "epoch": 25.0,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.0029,
      "step": 271100
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.983328283933777,
      "eval_f1": 0.9295600197618832,
      "eval_loss": 0.16894972324371338,
      "eval_precision": 0.9251374672779401,
      "eval_recall": 0.9340250587314017,
      "eval_runtime": 88.675,
      "eval_samples_per_second": 305.667,
      "eval_steps_per_second": 15.292,
      "step": 271100
    },
    {
      "epoch": 26.0,
      "learning_rate": 7e-06,
      "loss": 0.0026,
      "step": 281944
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.9831079464985917,
      "eval_f1": 0.9281980972530569,
      "eval_loss": 0.1714058518409729,
      "eval_precision": 0.9239348804987877,
      "eval_recall": 0.9325008390200246,
      "eval_runtime": 89.606,
      "eval_samples_per_second": 302.491,
      "eval_steps_per_second": 15.133,
      "step": 281944
    },
    {
      "epoch": 27.0,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.0022,
      "step": 292788
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.9831311399128217,
      "eval_f1": 0.9291277150061997,
      "eval_loss": 0.16884349286556244,
      "eval_precision": 0.9257100024985425,
      "eval_recall": 0.9325707573554088,
      "eval_runtime": 88.5198,
      "eval_samples_per_second": 306.203,
      "eval_steps_per_second": 15.319,
      "step": 292788
    },
    {
      "epoch": 28.0,
      "learning_rate": 6e-06,
      "loss": 0.002,
      "step": 303632
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.9833819187041839,
      "eval_f1": 0.9299846945874496,
      "eval_loss": 0.17876744270324707,
      "eval_precision": 0.9253752007531705,
      "eval_recall": 0.9346403400827833,
      "eval_runtime": 88.8069,
      "eval_samples_per_second": 305.213,
      "eval_steps_per_second": 15.269,
      "step": 303632
    },
    {
      "epoch": 29.0,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.0019,
      "step": 314476
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.9836268991419886,
      "eval_f1": 0.9311766262342112,
      "eval_loss": 0.1778053343296051,
      "eval_precision": 0.9273667859421375,
      "eval_recall": 0.9350178990938584,
      "eval_runtime": 89.2258,
      "eval_samples_per_second": 303.78,
      "eval_steps_per_second": 15.197,
      "step": 314476
    },
    {
      "epoch": 30.0,
      "learning_rate": 5e-06,
      "loss": 0.0018,
      "step": 325320
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.9834732427727146,
      "eval_f1": 0.9307018886832106,
      "eval_loss": 0.18135882914066315,
      "eval_precision": 0.9263569113124429,
      "eval_recall": 0.9350878174292426,
      "eval_runtime": 88.7746,
      "eval_samples_per_second": 305.324,
      "eval_steps_per_second": 15.275,
      "step": 325320
    },
    {
      "epoch": 31.0,
      "learning_rate": 4.5e-06,
      "loss": 0.0016,
      "step": 336164
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.983370321997069,
      "eval_f1": 0.9299260330348701,
      "eval_loss": 0.18739104270935059,
      "eval_precision": 0.9246295067462951,
      "eval_recall": 0.9352835887683186,
      "eval_runtime": 89.214,
      "eval_samples_per_second": 303.82,
      "eval_steps_per_second": 15.199,
      "step": 336164
    },
    {
      "epoch": 32.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0014,
      "step": 347008
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.9836066049045373,
      "eval_f1": 0.9308149819620296,
      "eval_loss": 0.18249443173408508,
      "eval_precision": 0.927185947775897,
      "eval_recall": 0.9344725360778611,
      "eval_runtime": 89.9015,
      "eval_samples_per_second": 301.497,
      "eval_steps_per_second": 15.083,
      "step": 347008
    },
    {
      "epoch": 33.0,
      "learning_rate": 3.5e-06,
      "loss": 0.0015,
      "step": 357852
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.9833732211738477,
      "eval_f1": 0.9303621946290954,
      "eval_loss": 0.18371780216693878,
      "eval_precision": 0.927196466764812,
      "eval_recall": 0.9335496140507887,
      "eval_runtime": 89.8085,
      "eval_samples_per_second": 301.809,
      "eval_steps_per_second": 15.099,
      "step": 357852
    },
    {
      "epoch": 34.0,
      "learning_rate": 3e-06,
      "loss": 0.0013,
      "step": 368696
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.9834457005933165,
      "eval_f1": 0.9307165143748607,
      "eval_loss": 0.1884261518716812,
      "eval_precision": 0.9271006771006771,
      "eval_recall": 0.9343606667412462,
      "eval_runtime": 90.3295,
      "eval_samples_per_second": 300.068,
      "eval_steps_per_second": 15.012,
      "step": 368696
    },
    {
      "epoch": 35.0,
      "learning_rate": 2.5e-06,
      "loss": 0.0012,
      "step": 379540
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.9836544413213868,
      "eval_f1": 0.9312036572448013,
      "eval_loss": 0.18962261080741882,
      "eval_precision": 0.9281507001555901,
      "eval_recall": 0.9342767647387851,
      "eval_runtime": 89.7245,
      "eval_samples_per_second": 302.092,
      "eval_steps_per_second": 15.113,
      "step": 379540
    },
    {
      "epoch": 36.0,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0011,
      "step": 390384
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.9836124032580948,
      "eval_f1": 0.9311286323238515,
      "eval_loss": 0.18686576187610626,
      "eval_precision": 0.9275881210102692,
      "eval_recall": 0.9346962747510907,
      "eval_runtime": 90.2423,
      "eval_samples_per_second": 300.358,
      "eval_steps_per_second": 15.026,
      "step": 390384
    },
    {
      "epoch": 37.0,
      "learning_rate": 1.5e-06,
      "loss": 0.001,
      "step": 401228
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.9836442942026612,
      "eval_f1": 0.9311170842443909,
      "eval_loss": 0.18996329605579376,
      "eval_precision": 0.9273587261075817,
      "eval_recall": 0.9349060297572436,
      "eval_runtime": 88.7307,
      "eval_samples_per_second": 305.475,
      "eval_steps_per_second": 15.282,
      "step": 401228
    },
    {
      "epoch": 38.0,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.001,
      "step": 412072
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.9837356182711919,
      "eval_f1": 0.9316876434183994,
      "eval_loss": 0.19159720838069916,
      "eval_precision": 0.9266231431021605,
      "eval_recall": 0.9368078084796957,
      "eval_runtime": 88.7508,
      "eval_samples_per_second": 305.406,
      "eval_steps_per_second": 15.279,
      "step": 412072
    },
    {
      "epoch": 39.0,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.001,
      "step": 422916
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.9837994001603245,
      "eval_f1": 0.9321045231167601,
      "eval_loss": 0.1949097365140915,
      "eval_precision": 0.927983367983368,
      "eval_recall": 0.9362624454636984,
      "eval_runtime": 88.5519,
      "eval_samples_per_second": 306.092,
      "eval_steps_per_second": 15.313,
      "step": 422916
    },
    {
      "epoch": 40.0,
      "learning_rate": 0.0,
      "loss": 0.0009,
      "step": 433760
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.9839153672314747,
      "eval_f1": 0.9324748546966902,
      "eval_loss": 0.19464968144893646,
      "eval_precision": 0.9283328482530179,
      "eval_recall": 0.9366539881418503,
      "eval_runtime": 88.4748,
      "eval_samples_per_second": 306.358,
      "eval_steps_per_second": 15.326,
      "step": 433760
    }
  ],
  "logging_steps": 500,
  "max_steps": 433760,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 1.1334359572090675e+18,
  "trial_name": null,
  "trial_params": null
}