trainer_state.json

{
  "best_metric": 0.8802370452039692,
  "best_model_checkpoint": "test/checkpoint-3000",
  "epoch": 12.539184952978056,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.31,
      "eval_accuracy": 0.6750517260154634,
      "eval_f1": 0.38478707258280515,
      "eval_loss": 1.2847243547439575,
      "eval_precision": 0.5487649605296664,
      "eval_recall": 0.29626065438548255,
      "eval_runtime": 48.6236,
      "eval_samples_per_second": 9.728,
      "eval_steps_per_second": 2.447,
      "step": 100
    },
    {
      "epoch": 0.63,
      "eval_accuracy": 0.7829140803658935,
      "eval_f1": 0.589244320815948,
      "eval_loss": 0.8382514119148254,
      "eval_precision": 0.6727240649258998,
      "eval_recall": 0.5241957657409954,
      "eval_runtime": 47.451,
      "eval_samples_per_second": 9.968,
      "eval_steps_per_second": 2.508,
      "step": 200
    },
    {
      "epoch": 0.94,
      "eval_accuracy": 0.8379070020690407,
      "eval_f1": 0.7082447783070723,
      "eval_loss": 0.6504582166671753,
      "eval_precision": 0.7584366661434626,
      "eval_recall": 0.6642837503436898,
      "eval_runtime": 47.1785,
      "eval_samples_per_second": 10.026,
      "eval_steps_per_second": 2.522,
      "step": 300
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 0.8560383316998802,
      "eval_f1": 0.74649361523969,
      "eval_loss": 0.5616394877433777,
      "eval_precision": 0.7579708091256908,
      "eval_recall": 0.7353588122078636,
      "eval_runtime": 47.4319,
      "eval_samples_per_second": 9.972,
      "eval_steps_per_second": 2.509,
      "step": 400
    },
    {
      "epoch": 1.57,
      "learning_rate": 8.750000000000001e-06,
      "loss": 1.0044,
      "step": 500
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 0.8640967004246978,
      "eval_f1": 0.7635898529098273,
      "eval_loss": 0.48347049951553345,
      "eval_precision": 0.7902632740108839,
      "eval_recall": 0.7386582348089085,
      "eval_runtime": 47.4677,
      "eval_samples_per_second": 9.965,
      "eval_steps_per_second": 2.507,
      "step": 500
    },
    {
      "epoch": 1.88,
      "eval_accuracy": 0.8795600566263748,
      "eval_f1": 0.7909914858555341,
      "eval_loss": 0.4531821012496948,
      "eval_precision": 0.7901234567901234,
      "eval_recall": 0.7918614242507561,
      "eval_runtime": 48.4837,
      "eval_samples_per_second": 9.756,
      "eval_steps_per_second": 2.454,
      "step": 600
    },
    {
      "epoch": 2.19,
      "eval_accuracy": 0.8883262550364804,
      "eval_f1": 0.8112951398316589,
      "eval_loss": 0.4272719919681549,
      "eval_precision": 0.8012872083668544,
      "eval_recall": 0.8215562276601595,
      "eval_runtime": 47.6145,
      "eval_samples_per_second": 9.934,
      "eval_steps_per_second": 2.499,
      "step": 700
    },
    {
      "epoch": 2.51,
      "eval_accuracy": 0.8954590003266907,
      "eval_f1": 0.8212751449028298,
      "eval_loss": 0.3849042057991028,
      "eval_precision": 0.8147747260181302,
      "eval_recall": 0.8278801209788287,
      "eval_runtime": 47.9087,
      "eval_samples_per_second": 9.873,
      "eval_steps_per_second": 2.484,
      "step": 800
    },
    {
      "epoch": 2.82,
      "eval_accuracy": 0.8950234128280519,
      "eval_f1": 0.8168306752629793,
      "eval_loss": 0.3808012306690216,
      "eval_precision": 0.8065942903096099,
      "eval_recall": 0.8273302172119879,
      "eval_runtime": 47.5595,
      "eval_samples_per_second": 9.945,
      "eval_steps_per_second": 2.502,
      "step": 900
    },
    {
      "epoch": 3.13,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4068,
      "step": 1000
    },
    {
      "epoch": 3.13,
      "eval_accuracy": 0.9042796471741261,
      "eval_f1": 0.838237306692213,
      "eval_loss": 0.35156166553497314,
      "eval_precision": 0.8344686648501363,
      "eval_recall": 0.8420401429749794,
      "eval_runtime": 48.1704,
      "eval_samples_per_second": 9.819,
      "eval_steps_per_second": 2.47,
      "step": 1000
    },
    {
      "epoch": 3.45,
      "eval_accuracy": 0.9095066971577915,
      "eval_f1": 0.8464466126911634,
      "eval_loss": 0.33900508284568787,
      "eval_precision": 0.8521666434443361,
      "eval_recall": 0.8408028594995876,
      "eval_runtime": 48.0055,
      "eval_samples_per_second": 9.853,
      "eval_steps_per_second": 2.479,
      "step": 1100
    },
    {
      "epoch": 3.76,
      "eval_accuracy": 0.9087444190351737,
      "eval_f1": 0.8463414634146341,
      "eval_loss": 0.32527512311935425,
      "eval_precision": 0.8343574672722415,
      "eval_recall": 0.8586747319219137,
      "eval_runtime": 48.0035,
      "eval_samples_per_second": 9.853,
      "eval_steps_per_second": 2.479,
      "step": 1200
    },
    {
      "epoch": 4.08,
      "eval_accuracy": 0.9081999346618752,
      "eval_f1": 0.8466865833785986,
      "eval_loss": 0.33180123567581177,
      "eval_precision": 0.8364636436812449,
      "eval_recall": 0.8571624965631015,
      "eval_runtime": 47.8611,
      "eval_samples_per_second": 9.883,
      "eval_steps_per_second": 2.486,
      "step": 1300
    },
    {
      "epoch": 4.39,
      "eval_accuracy": 0.9187629315038658,
      "eval_f1": 0.8616123499142367,
      "eval_loss": 0.3064488470554352,
      "eval_precision": 0.8600191754554171,
      "eval_recall": 0.8632114379983503,
      "eval_runtime": 47.6825,
      "eval_samples_per_second": 9.92,
      "eval_steps_per_second": 2.496,
      "step": 1400
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.25e-06,
      "loss": 0.2678,
      "step": 1500
    },
    {
      "epoch": 4.7,
      "eval_accuracy": 0.9184906893172166,
      "eval_f1": 0.8611149536588739,
      "eval_loss": 0.30125683546066284,
      "eval_precision": 0.8665089086859689,
      "eval_recall": 0.8557877371459994,
      "eval_runtime": 47.6577,
      "eval_samples_per_second": 9.925,
      "eval_steps_per_second": 2.497,
      "step": 1500
    },
    {
      "epoch": 5.02,
      "eval_accuracy": 0.9157138190133943,
      "eval_f1": 0.8587268993839836,
      "eval_loss": 0.30989399552345276,
      "eval_precision": 0.8550981461286805,
      "eval_recall": 0.8623865823480891,
      "eval_runtime": 47.7978,
      "eval_samples_per_second": 9.896,
      "eval_steps_per_second": 2.49,
      "step": 1600
    },
    {
      "epoch": 5.33,
      "eval_accuracy": 0.9191440705651748,
      "eval_f1": 0.8643950889777902,
      "eval_loss": 0.3114277124404907,
      "eval_precision": 0.8673864894795127,
      "eval_recall": 0.8614242507561177,
      "eval_runtime": 47.8331,
      "eval_samples_per_second": 9.889,
      "eval_steps_per_second": 2.488,
      "step": 1700
    },
    {
      "epoch": 5.64,
      "eval_accuracy": 0.9207230752477404,
      "eval_f1": 0.8651245795867938,
      "eval_loss": 0.3051866888999939,
      "eval_precision": 0.8638793694311172,
      "eval_recall": 0.8663733846576849,
      "eval_runtime": 47.7044,
      "eval_samples_per_second": 9.915,
      "eval_steps_per_second": 2.495,
      "step": 1800
    },
    {
      "epoch": 5.96,
      "eval_accuracy": 0.9236088424262223,
      "eval_f1": 0.8706572930175415,
      "eval_loss": 0.2927141785621643,
      "eval_precision": 0.8645790971939813,
      "eval_recall": 0.8768215562276601,
      "eval_runtime": 47.4403,
      "eval_samples_per_second": 9.97,
      "eval_steps_per_second": 2.508,
      "step": 1900
    },
    {
      "epoch": 6.27,
      "learning_rate": 5e-06,
      "loss": 0.2068,
      "step": 2000
    },
    {
      "epoch": 6.27,
      "eval_accuracy": 0.9194707611891538,
      "eval_f1": 0.8640943047083819,
      "eval_loss": 0.30293476581573486,
      "eval_precision": 0.8615552822194888,
      "eval_recall": 0.8666483365411053,
      "eval_runtime": 47.7322,
      "eval_samples_per_second": 9.909,
      "eval_steps_per_second": 2.493,
      "step": 2000
    },
    {
      "epoch": 6.58,
      "eval_accuracy": 0.9211042143090493,
      "eval_f1": 0.8684515293957353,
      "eval_loss": 0.3005865216255188,
      "eval_precision": 0.8607697501688049,
      "eval_recall": 0.8762716524608194,
      "eval_runtime": 47.8357,
      "eval_samples_per_second": 9.888,
      "eval_steps_per_second": 2.488,
      "step": 2100
    },
    {
      "epoch": 6.9,
      "eval_accuracy": 0.923391048676903,
      "eval_f1": 0.8718300205620287,
      "eval_loss": 0.2952803373336792,
      "eval_precision": 0.8693275013668671,
      "eval_recall": 0.8743469892768766,
      "eval_runtime": 47.6233,
      "eval_samples_per_second": 9.932,
      "eval_steps_per_second": 2.499,
      "step": 2200
    },
    {
      "epoch": 7.21,
      "eval_accuracy": 0.9212675596210389,
      "eval_f1": 0.8683849739511927,
      "eval_loss": 0.3120380938053131,
      "eval_precision": 0.8660103910308996,
      "eval_recall": 0.8707726147924113,
      "eval_runtime": 48.4779,
      "eval_samples_per_second": 9.757,
      "eval_steps_per_second": 2.455,
      "step": 2300
    },
    {
      "epoch": 7.52,
      "eval_accuracy": 0.9225198736796254,
      "eval_f1": 0.8689645739296951,
      "eval_loss": 0.3099469840526581,
      "eval_precision": 0.8645890038105607,
      "eval_recall": 0.8733846576849051,
      "eval_runtime": 47.5215,
      "eval_samples_per_second": 9.953,
      "eval_steps_per_second": 2.504,
      "step": 2400
    },
    {
      "epoch": 7.84,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.16,
      "step": 2500
    },
    {
      "epoch": 7.84,
      "eval_accuracy": 0.9236088424262223,
      "eval_f1": 0.8730713844887883,
      "eval_loss": 0.299545019865036,
      "eval_precision": 0.8709809823505268,
      "eval_recall": 0.8751718449271377,
      "eval_runtime": 47.9228,
      "eval_samples_per_second": 9.87,
      "eval_steps_per_second": 2.483,
      "step": 2500
    },
    {
      "epoch": 8.15,
      "eval_accuracy": 0.9233366002395731,
      "eval_f1": 0.8713222745579903,
      "eval_loss": 0.3018472492694855,
      "eval_precision": 0.8653559322033898,
      "eval_recall": 0.877371459994501,
      "eval_runtime": 47.7011,
      "eval_samples_per_second": 9.916,
      "eval_steps_per_second": 2.495,
      "step": 2600
    },
    {
      "epoch": 8.46,
      "eval_accuracy": 0.9234999455515627,
      "eval_f1": 0.8743416102332581,
      "eval_loss": 0.305177241563797,
      "eval_precision": 0.8701157249829816,
      "eval_recall": 0.8786087434698928,
      "eval_runtime": 48.329,
      "eval_samples_per_second": 9.787,
      "eval_steps_per_second": 2.462,
      "step": 2700
    },
    {
      "epoch": 8.78,
      "eval_accuracy": 0.9252967439834476,
      "eval_f1": 0.8783895960154953,
      "eval_loss": 0.2957703769207001,
      "eval_precision": 0.8840155945419104,
      "eval_recall": 0.8728347539180643,
      "eval_runtime": 47.6638,
      "eval_samples_per_second": 9.924,
      "eval_steps_per_second": 2.497,
      "step": 2800
    },
    {
      "epoch": 9.09,
      "eval_accuracy": 0.9270935424153327,
      "eval_f1": 0.8792522850663185,
      "eval_loss": 0.2976396679878235,
      "eval_precision": 0.8790710457606157,
      "eval_recall": 0.879433599120154,
      "eval_runtime": 48.2883,
      "eval_samples_per_second": 9.795,
      "eval_steps_per_second": 2.464,
      "step": 2900
    },
    {
      "epoch": 9.4,
      "learning_rate": 2.5e-06,
      "loss": 0.1364,
      "step": 3000
    },
    {
      "epoch": 9.4,
      "eval_accuracy": 0.9288358924098878,
      "eval_f1": 0.8802370452039692,
      "eval_loss": 0.30077481269836426,
      "eval_precision": 0.8824260845537442,
      "eval_recall": 0.878058839703052,
      "eval_runtime": 47.4506,
      "eval_samples_per_second": 9.968,
      "eval_steps_per_second": 2.508,
      "step": 3000
    },
    {
      "epoch": 9.72,
      "eval_accuracy": 0.9276380267886312,
      "eval_f1": 0.8794581144594968,
      "eval_loss": 0.295279324054718,
      "eval_precision": 0.8843480678343064,
      "eval_recall": 0.8746219411602969,
      "eval_runtime": 48.0211,
      "eval_samples_per_second": 9.85,
      "eval_steps_per_second": 2.478,
      "step": 3100
    },
    {
      "epoch": 10.03,
      "eval_accuracy": 0.9261134705433954,
      "eval_f1": 0.8766407807023572,
      "eval_loss": 0.30134841799736023,
      "eval_precision": 0.8764600797031744,
      "eval_recall": 0.8768215562276601,
      "eval_runtime": 47.9205,
      "eval_samples_per_second": 9.871,
      "eval_steps_per_second": 2.483,
      "step": 3200
    },
    {
      "epoch": 10.34,
      "eval_accuracy": 0.9256234346074268,
      "eval_f1": 0.876704818038517,
      "eval_loss": 0.306386798620224,
      "eval_precision": 0.8741287412874129,
      "eval_recall": 0.8792961231784437,
      "eval_runtime": 47.48,
      "eval_samples_per_second": 9.962,
      "eval_steps_per_second": 2.506,
      "step": 3300
    },
    {
      "epoch": 10.66,
      "eval_accuracy": 0.9298159642818251,
      "eval_f1": 0.8808639427706698,
      "eval_loss": 0.2949816882610321,
      "eval_precision": 0.8814702643171806,
      "eval_recall": 0.8802584547704152,
      "eval_runtime": 47.5486,
      "eval_samples_per_second": 9.948,
      "eval_steps_per_second": 2.503,
      "step": 3400
    },
    {
      "epoch": 10.97,
      "learning_rate": 1.25e-06,
      "loss": 0.1192,
      "step": 3500
    },
    {
      "epoch": 10.97,
      "eval_accuracy": 0.9262223674180551,
      "eval_f1": 0.8796156485929993,
      "eval_loss": 0.3046295940876007,
      "eval_precision": 0.8782894736842105,
      "eval_recall": 0.8809458344789661,
      "eval_runtime": 47.8942,
      "eval_samples_per_second": 9.876,
      "eval_steps_per_second": 2.485,
      "step": 3500
    },
    {
      "epoch": 11.29,
      "eval_accuracy": 0.9261134705433954,
      "eval_f1": 0.8789643126241523,
      "eval_loss": 0.30884549021720886,
      "eval_precision": 0.8759044368600682,
      "eval_recall": 0.8820456420126478,
      "eval_runtime": 48.0682,
      "eval_samples_per_second": 9.84,
      "eval_steps_per_second": 2.476,
      "step": 3600
    },
    {
      "epoch": 11.6,
      "eval_accuracy": 0.9268757486660133,
      "eval_f1": 0.8796163069544365,
      "eval_loss": 0.30328136682510376,
      "eval_precision": 0.8767927878705095,
      "eval_recall": 0.8824580698377784,
      "eval_runtime": 47.6224,
      "eval_samples_per_second": 9.932,
      "eval_steps_per_second": 2.499,
      "step": 3700
    },
    {
      "epoch": 11.91,
      "eval_accuracy": 0.9269301971033431,
      "eval_f1": 0.8791058081327573,
      "eval_loss": 0.30441808700561523,
      "eval_precision": 0.8770009577233547,
      "eval_recall": 0.8812207863623865,
      "eval_runtime": 47.6174,
      "eval_samples_per_second": 9.933,
      "eval_steps_per_second": 2.499,
      "step": 3800
    },
    {
      "epoch": 12.23,
      "eval_accuracy": 0.9272568877273222,
      "eval_f1": 0.8795933786661172,
      "eval_loss": 0.3058357238769531,
      "eval_precision": 0.8789293067947838,
      "eval_recall": 0.8802584547704152,
      "eval_runtime": 47.5523,
      "eval_samples_per_second": 9.947,
      "eval_steps_per_second": 2.503,
      "step": 3900
    },
    {
      "epoch": 12.54,
      "learning_rate": 0.0,
      "loss": 0.1044,
      "step": 4000
    },
    {
      "epoch": 12.54,
      "eval_accuracy": 0.9272024392899924,
      "eval_f1": 0.8791826659352715,
      "eval_loss": 0.3065986931324005,
      "eval_precision": 0.8770177838577291,
      "eval_recall": 0.8813582623040968,
      "eval_runtime": 47.8877,
      "eval_samples_per_second": 9.877,
      "eval_steps_per_second": 2.485,
      "step": 4000
    }
  ],
  "max_steps": 4000,
  "num_train_epochs": 13,
  "total_flos": 4247432429568000.0,
  "trial_name": null,
  "trial_params": null
}