trainer_state.json

{
  "best_metric": 0.07586462050676346,
  "best_model_checkpoint": "/data/jcanete/all_results/pos/distillbeto/epochs_4_bs_16_lr_5e-5/checkpoint-1700",
  "epoch": 4.0,
  "global_step": 3580,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "eval_accuracy": 0.9443306584293053,
      "eval_f1": 0.9341365141420992,
      "eval_loss": 0.212116539478302,
      "eval_precision": 0.9329083665338646,
      "eval_recall": 0.9353678996564672,
      "eval_runtime": 1.788,
      "eval_samples_per_second": 925.041,
      "eval_steps_per_second": 58.165,
      "step": 100
    },
    {
      "epoch": 0.22,
      "eval_accuracy": 0.9563677265554176,
      "eval_f1": 0.9504974205973039,
      "eval_loss": 0.15402881801128387,
      "eval_precision": 0.9497298051805619,
      "eval_recall": 0.9512662778621075,
      "eval_runtime": 1.7688,
      "eval_samples_per_second": 935.077,
      "eval_steps_per_second": 58.796,
      "step": 200
    },
    {
      "epoch": 0.34,
      "eval_accuracy": 0.9625961118759023,
      "eval_f1": 0.9571392947982644,
      "eval_loss": 0.13182665407657623,
      "eval_precision": 0.9560045430091458,
      "eval_recall": 0.958276743628665,
      "eval_runtime": 1.7773,
      "eval_samples_per_second": 930.646,
      "eval_steps_per_second": 58.517,
      "step": 300
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.9649632340596985,
      "eval_f1": 0.96070983500324,
      "eval_loss": 0.12322123348712921,
      "eval_precision": 0.9590938949379939,
      "eval_recall": 0.9623312295278421,
      "eval_runtime": 1.7902,
      "eval_samples_per_second": 923.933,
      "eval_steps_per_second": 58.095,
      "step": 400
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.301675977653631e-05,
      "loss": 0.2881,
      "step": 500
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 0.9674814491488433,
      "eval_f1": 0.9628313283082913,
      "eval_loss": 0.11282572150230408,
      "eval_precision": 0.9625526478631455,
      "eval_recall": 0.9631101701685707,
      "eval_runtime": 1.7431,
      "eval_samples_per_second": 948.862,
      "eval_steps_per_second": 59.662,
      "step": 500
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.9697646308296679,
      "eval_f1": 0.9659112436064888,
      "eval_loss": 0.10294512659311295,
      "eval_precision": 0.9643632164685739,
      "eval_recall": 0.9674642486218743,
      "eval_runtime": 1.7656,
      "eval_samples_per_second": 936.792,
      "eval_steps_per_second": 58.903,
      "step": 600
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 0.9717624148003895,
      "eval_f1": 0.9682482698099284,
      "eval_loss": 0.09821932762861252,
      "eval_precision": 0.9668791077474607,
      "eval_recall": 0.9696213150115842,
      "eval_runtime": 1.786,
      "eval_samples_per_second": 926.088,
      "eval_steps_per_second": 58.23,
      "step": 700
    },
    {
      "epoch": 0.89,
      "eval_accuracy": 0.9731894033509049,
      "eval_f1": 0.9696576892779867,
      "eval_loss": 0.09264940768480301,
      "eval_precision": 0.9681806415642858,
      "eval_recall": 0.971139250619158,
      "eval_runtime": 1.7594,
      "eval_samples_per_second": 940.07,
      "eval_steps_per_second": 59.11,
      "step": 800
    },
    {
      "epoch": 1.01,
      "eval_accuracy": 0.9730215223449619,
      "eval_f1": 0.9703566974307807,
      "eval_loss": 0.09371750056743622,
      "eval_precision": 0.9694957832406245,
      "eval_recall": 0.9712191419669249,
      "eval_runtime": 1.7445,
      "eval_samples_per_second": 948.129,
      "eval_steps_per_second": 59.616,
      "step": 900
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.603351955307263e-05,
      "loss": 0.0962,
      "step": 1000
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.9754054326293523,
      "eval_f1": 0.9727950199024332,
      "eval_loss": 0.08893433958292007,
      "eval_precision": 0.9717964561200694,
      "eval_recall": 0.973795637932412,
      "eval_runtime": 1.7661,
      "eval_samples_per_second": 936.527,
      "eval_steps_per_second": 58.887,
      "step": 1000
    },
    {
      "epoch": 1.23,
      "eval_accuracy": 0.9751368230198435,
      "eval_f1": 0.972067262112669,
      "eval_loss": 0.09049103409051895,
      "eval_precision": 0.9713983684703912,
      "eval_recall": 0.9727370775744987,
      "eval_runtime": 1.7822,
      "eval_samples_per_second": 928.079,
      "eval_steps_per_second": 58.356,
      "step": 1100
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.9766141758721418,
      "eval_f1": 0.9741814481533949,
      "eval_loss": 0.08416619151830673,
      "eval_precision": 0.973191150089695,
      "eval_recall": 0.9751737636813933,
      "eval_runtime": 1.8044,
      "eval_samples_per_second": 916.636,
      "eval_steps_per_second": 57.636,
      "step": 1200
    },
    {
      "epoch": 1.45,
      "eval_accuracy": 0.975892287546587,
      "eval_f1": 0.9735277689115864,
      "eval_loss": 0.08117581903934479,
      "eval_precision": 0.972363911692038,
      "eval_recall": 0.9746944155947911,
      "eval_runtime": 1.7582,
      "eval_samples_per_second": 940.735,
      "eval_steps_per_second": 59.151,
      "step": 1300
    },
    {
      "epoch": 1.56,
      "eval_accuracy": 0.9770674545881879,
      "eval_f1": 0.9749930148086058,
      "eval_loss": 0.08163653314113617,
      "eval_precision": 0.9742541480536057,
      "eval_recall": 0.9757330031157626,
      "eval_runtime": 1.7901,
      "eval_samples_per_second": 923.964,
      "eval_steps_per_second": 58.097,
      "step": 1400
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.9050279329608944e-05,
      "loss": 0.0625,
      "step": 1500
    },
    {
      "epoch": 1.68,
      "eval_accuracy": 0.976513447268576,
      "eval_f1": 0.9742249531181423,
      "eval_loss": 0.08262678235769272,
      "eval_precision": 0.9730989956958394,
      "eval_recall": 0.9753535192138691,
      "eval_runtime": 1.755,
      "eval_samples_per_second": 942.465,
      "eval_steps_per_second": 59.26,
      "step": 1500
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.9771513950911593,
      "eval_f1": 0.9745410305248356,
      "eval_loss": 0.08152274787425995,
      "eval_precision": 0.9731532931031049,
      "eval_recall": 0.9759327314851801,
      "eval_runtime": 1.7691,
      "eval_samples_per_second": 934.957,
      "eval_steps_per_second": 58.788,
      "step": 1600
    },
    {
      "epoch": 1.9,
      "eval_accuracy": 0.9779572239196858,
      "eval_f1": 0.9755450236966825,
      "eval_loss": 0.07586462050676346,
      "eval_precision": 0.9746795063500608,
      "eval_recall": 0.9764120795717823,
      "eval_runtime": 1.754,
      "eval_samples_per_second": 942.987,
      "eval_steps_per_second": 59.293,
      "step": 1700
    },
    {
      "epoch": 2.01,
      "eval_accuracy": 0.9781251049256288,
      "eval_f1": 0.9761717152574831,
      "eval_loss": 0.07830089330673218,
      "eval_precision": 0.9749755932338467,
      "eval_recall": 0.9773707757449868,
      "eval_runtime": 1.7982,
      "eval_samples_per_second": 919.825,
      "eval_steps_per_second": 57.837,
      "step": 1800
    },
    {
      "epoch": 2.12,
      "eval_accuracy": 0.9776550381089884,
      "eval_f1": 0.9756199790429619,
      "eval_loss": 0.08235861361026764,
      "eval_precision": 0.9749486407244151,
      "eval_recall": 0.9762922425501318,
      "eval_runtime": 1.7592,
      "eval_samples_per_second": 940.194,
      "eval_steps_per_second": 59.117,
      "step": 1900
    },
    {
      "epoch": 2.23,
      "learning_rate": 2.208100558659218e-05,
      "loss": 0.0492,
      "step": 2000
    },
    {
      "epoch": 2.23,
      "eval_accuracy": 0.9773528522982909,
      "eval_f1": 0.975572275330792,
      "eval_loss": 0.08445031195878983,
      "eval_precision": 0.9748135444502054,
      "eval_recall": 0.9763321882240154,
      "eval_runtime": 1.7661,
      "eval_samples_per_second": 936.546,
      "eval_steps_per_second": 58.888,
      "step": 2000
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.9780579525232516,
      "eval_f1": 0.9760987969594797,
      "eval_loss": 0.08256729692220688,
      "eval_precision": 0.9750288971262306,
      "eval_recall": 0.9771710473755693,
      "eval_runtime": 1.7494,
      "eval_samples_per_second": 945.46,
      "eval_steps_per_second": 59.449,
      "step": 2100
    },
    {
      "epoch": 2.46,
      "eval_accuracy": 0.9782761978309774,
      "eval_f1": 0.9764048308214393,
      "eval_loss": 0.08261139690876007,
      "eval_precision": 0.9758788555923547,
      "eval_recall": 0.9769313733322681,
      "eval_runtime": 1.7667,
      "eval_samples_per_second": 936.186,
      "eval_steps_per_second": 58.865,
      "step": 2200
    },
    {
      "epoch": 2.57,
      "eval_accuracy": 0.9783265621327603,
      "eval_f1": 0.9761928518688512,
      "eval_loss": 0.07977774739265442,
      "eval_precision": 0.9753559038162459,
      "eval_recall": 0.977031237516977,
      "eval_runtime": 1.7935,
      "eval_samples_per_second": 922.203,
      "eval_steps_per_second": 57.986,
      "step": 2300
    },
    {
      "epoch": 2.68,
      "eval_accuracy": 0.9788973575529665,
      "eval_f1": 0.9770031227863634,
      "eval_loss": 0.08085376024246216,
      "eval_precision": 0.9760779857664015,
      "eval_recall": 0.9779300151793561,
      "eval_runtime": 1.8062,
      "eval_samples_per_second": 915.718,
      "eval_steps_per_second": 57.578,
      "step": 2400
    },
    {
      "epoch": 2.79,
      "learning_rate": 1.509776536312849e-05,
      "loss": 0.0356,
      "step": 2500
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 0.9785280193398919,
      "eval_f1": 0.9764671847410792,
      "eval_loss": 0.08119603991508484,
      "eval_precision": 0.9754260089686099,
      "eval_recall": 0.9775105856035792,
      "eval_runtime": 1.7562,
      "eval_samples_per_second": 941.793,
      "eval_steps_per_second": 59.218,
      "step": 2500
    },
    {
      "epoch": 2.91,
      "eval_accuracy": 0.9783937145351375,
      "eval_f1": 0.9762508232382701,
      "eval_loss": 0.079347625374794,
      "eval_precision": 0.9754915646312766,
      "eval_recall": 0.9770112646800352,
      "eval_runtime": 1.8127,
      "eval_samples_per_second": 912.464,
      "eval_steps_per_second": 57.374,
      "step": 2600
    },
    {
      "epoch": 3.02,
      "eval_accuracy": 0.9787966289494007,
      "eval_f1": 0.9768903789738371,
      "eval_loss": 0.07749345153570175,
      "eval_precision": 0.9760917248255234,
      "eval_recall": 0.977690341136055,
      "eval_runtime": 1.7581,
      "eval_samples_per_second": 940.794,
      "eval_steps_per_second": 59.155,
      "step": 2700
    },
    {
      "epoch": 3.13,
      "eval_accuracy": 0.9790148742571265,
      "eval_f1": 0.9772217067257326,
      "eval_loss": 0.08272269368171692,
      "eval_precision": 0.9763352538926214,
      "eval_recall": 0.9781097707118319,
      "eval_runtime": 1.7744,
      "eval_samples_per_second": 932.135,
      "eval_steps_per_second": 58.611,
      "step": 2800
    },
    {
      "epoch": 3.24,
      "eval_accuracy": 0.9795353053755498,
      "eval_f1": 0.9775534228536084,
      "eval_loss": 0.08201353996992111,
      "eval_precision": 0.9765596970300977,
      "eval_recall": 0.9785491731245506,
      "eval_runtime": 1.8046,
      "eval_samples_per_second": 916.528,
      "eval_steps_per_second": 57.629,
      "step": 2900
    },
    {
      "epoch": 3.35,
      "learning_rate": 8.114525139664805e-06,
      "loss": 0.0273,
      "step": 3000
    },
    {
      "epoch": 3.35,
      "eval_accuracy": 0.9786959003458349,
      "eval_f1": 0.9767107848515375,
      "eval_loss": 0.08390253037214279,
      "eval_precision": 0.9754945908792063,
      "eval_recall": 0.9779300151793561,
      "eval_runtime": 1.7708,
      "eval_samples_per_second": 934.046,
      "eval_steps_per_second": 58.731,
      "step": 3000
    },
    {
      "epoch": 3.46,
      "eval_accuracy": 0.9789309337541551,
      "eval_f1": 0.9770059880239521,
      "eval_loss": 0.0809033140540123,
      "eval_precision": 0.9763823505944307,
      "eval_recall": 0.9776304226252297,
      "eval_runtime": 1.7599,
      "eval_samples_per_second": 939.83,
      "eval_steps_per_second": 59.094,
      "step": 3100
    },
    {
      "epoch": 3.58,
      "eval_accuracy": 0.9789645099553437,
      "eval_f1": 0.9768119412515964,
      "eval_loss": 0.08214528858661652,
      "eval_precision": 0.97595502033655,
      "eval_recall": 0.9776703682991132,
      "eval_runtime": 1.7652,
      "eval_samples_per_second": 936.992,
      "eval_steps_per_second": 58.916,
      "step": 3200
    },
    {
      "epoch": 3.69,
      "eval_accuracy": 0.9794849410737669,
      "eval_f1": 0.9775280898876404,
      "eval_loss": 0.08165577799081802,
      "eval_precision": 0.9767678379132931,
      "eval_recall": 0.9782895262443078,
      "eval_runtime": 1.7562,
      "eval_samples_per_second": 941.784,
      "eval_steps_per_second": 59.217,
      "step": 3300
    },
    {
      "epoch": 3.8,
      "eval_accuracy": 0.9792499076654467,
      "eval_f1": 0.9771967466693278,
      "eval_loss": 0.0833316445350647,
      "eval_precision": 0.97652432335401,
      "eval_recall": 0.9778700966685308,
      "eval_runtime": 1.7941,
      "eval_samples_per_second": 921.897,
      "eval_steps_per_second": 57.967,
      "step": 3400
    },
    {
      "epoch": 3.91,
      "learning_rate": 1.1312849162011174e-06,
      "loss": 0.0232,
      "step": 3500
    },
    {
      "epoch": 3.91,
      "eval_accuracy": 0.9791827552630695,
      "eval_f1": 0.977121020125121,
      "eval_loss": 0.08293969184160233,
      "eval_precision": 0.9762735519888346,
      "eval_recall": 0.9779699608532396,
      "eval_runtime": 1.7577,
      "eval_samples_per_second": 941.011,
      "eval_steps_per_second": 59.169,
      "step": 3500
    },
    {
      "epoch": 4.0,
      "step": 3580,
      "total_flos": 1136902583504160.0,
      "train_loss": 0.08176401986756138,
      "train_runtime": 7371.7605,
      "train_samples_per_second": 7.762,
      "train_steps_per_second": 0.486
    }
  ],
  "max_steps": 3580,
  "num_train_epochs": 4,
  "total_flos": 1136902583504160.0,
  "trial_name": null,
  "trial_params": null
}