{
  "best_metric": 0.03719973564147949,
  "best_model_checkpoint": "doc-topic-model_eval-04_train-01/checkpoint-16000",
  "epoch": 10.355029585798816,
  "eval_steps": 1000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2465483234714004,
      "grad_norm": 0.32817578315734863,
      "learning_rate": 1.9950690335305722e-05,
      "loss": 0.1659,
      "step": 500
    },
    {
      "epoch": 0.4930966469428008,
      "grad_norm": 0.34106990694999695,
      "learning_rate": 1.9901380670611442e-05,
      "loss": 0.0941,
      "step": 1000
    },
    {
      "epoch": 0.4930966469428008,
      "eval_accuracy": 0.981496987239155,
      "eval_f1": 0.0,
      "eval_loss": 0.09015170484781265,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 12.0789,
      "eval_samples_per_second": 671.667,
      "eval_steps_per_second": 2.649,
      "step": 1000
    },
    {
      "epoch": 0.7396449704142012,
      "grad_norm": 0.4004117548465729,
      "learning_rate": 1.9852071005917162e-05,
      "loss": 0.088,
      "step": 1500
    },
    {
      "epoch": 0.9861932938856016,
      "grad_norm": 0.37588435411453247,
      "learning_rate": 1.980276134122288e-05,
      "loss": 0.0787,
      "step": 2000
    },
    {
      "epoch": 0.9861932938856016,
      "eval_accuracy": 0.981496987239155,
      "eval_f1": 0.0,
      "eval_loss": 0.07028835266828537,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 12.1247,
      "eval_samples_per_second": 669.129,
      "eval_steps_per_second": 2.639,
      "step": 2000
    },
    {
      "epoch": 1.232741617357002,
      "grad_norm": 0.43946701288223267,
      "learning_rate": 1.9753451676528602e-05,
      "loss": 0.0696,
      "step": 2500
    },
    {
      "epoch": 1.4792899408284024,
      "grad_norm": 0.4252438545227051,
      "learning_rate": 1.9704142011834322e-05,
      "loss": 0.0628,
      "step": 3000
    },
    {
      "epoch": 1.4792899408284024,
      "eval_accuracy": 0.9823399652324707,
      "eval_f1": 0.12347936150762956,
      "eval_loss": 0.05721545219421387,
      "eval_precision": 0.756244616709733,
      "eval_recall": 0.06722817764165391,
      "eval_runtime": 14.3987,
      "eval_samples_per_second": 563.455,
      "eval_steps_per_second": 2.222,
      "step": 3000
    },
    {
      "epoch": 1.725838264299803,
      "grad_norm": 0.3535268306732178,
      "learning_rate": 1.965483234714004e-05,
      "loss": 0.0584,
      "step": 3500
    },
    {
      "epoch": 1.972386587771203,
      "grad_norm": 0.3976692855358124,
      "learning_rate": 1.9605522682445763e-05,
      "loss": 0.0537,
      "step": 4000
    },
    {
      "epoch": 1.972386587771203,
      "eval_accuracy": 0.984258271455915,
      "eval_f1": 0.322045274269327,
      "eval_loss": 0.04996015876531601,
      "eval_precision": 0.792730549714629,
      "eval_recall": 0.20206738131699847,
      "eval_runtime": 14.3904,
      "eval_samples_per_second": 563.78,
      "eval_steps_per_second": 2.224,
      "step": 4000
    },
    {
      "epoch": 2.2189349112426036,
      "grad_norm": 0.24940212070941925,
      "learning_rate": 1.9556213017751483e-05,
      "loss": 0.0499,
      "step": 4500
    },
    {
      "epoch": 2.465483234714004,
      "grad_norm": 0.40094295144081116,
      "learning_rate": 1.95069033530572e-05,
      "loss": 0.0478,
      "step": 5000
    },
    {
      "epoch": 2.465483234714004,
      "eval_accuracy": 0.9853151816794672,
      "eval_f1": 0.4339468079296598,
      "eval_loss": 0.04658225178718567,
      "eval_precision": 0.7566177870881737,
      "eval_recall": 0.30421133231240427,
      "eval_runtime": 14.4801,
      "eval_samples_per_second": 560.288,
      "eval_steps_per_second": 2.21,
      "step": 5000
    },
    {
      "epoch": 2.712031558185404,
      "grad_norm": 0.42049625515937805,
      "learning_rate": 1.9457593688362923e-05,
      "loss": 0.0463,
      "step": 5500
    },
    {
      "epoch": 2.9585798816568047,
      "grad_norm": 0.41673073172569275,
      "learning_rate": 1.940828402366864e-05,
      "loss": 0.0453,
      "step": 6000
    },
    {
      "epoch": 2.9585798816568047,
      "eval_accuracy": 0.9859003075807098,
      "eval_f1": 0.5020016012810248,
      "eval_loss": 0.044097188860177994,
      "eval_precision": 0.7244367417677643,
      "eval_recall": 0.38407350689127107,
      "eval_runtime": 14.4621,
      "eval_samples_per_second": 560.982,
      "eval_steps_per_second": 2.213,
      "step": 6000
    },
    {
      "epoch": 3.2051282051282053,
      "grad_norm": 0.3240315914154053,
      "learning_rate": 1.935897435897436e-05,
      "loss": 0.0424,
      "step": 6500
    },
    {
      "epoch": 3.4516765285996054,
      "grad_norm": 0.39667147397994995,
      "learning_rate": 1.930966469428008e-05,
      "loss": 0.0389,
      "step": 7000
    },
    {
      "epoch": 3.4516765285996054,
      "eval_accuracy": 0.9864840167122159,
      "eval_f1": 0.5425338064639877,
      "eval_loss": 0.04140521585941315,
      "eval_precision": 0.7258147292789325,
      "eval_recall": 0.43315467075038283,
      "eval_runtime": 14.4115,
      "eval_samples_per_second": 562.952,
      "eval_steps_per_second": 2.22,
      "step": 7000
    },
    {
      "epoch": 3.698224852071006,
      "grad_norm": 0.39966270327568054,
      "learning_rate": 1.92603550295858e-05,
      "loss": 0.0403,
      "step": 7500
    },
    {
      "epoch": 3.9447731755424065,
      "grad_norm": 0.42601555585861206,
      "learning_rate": 1.921104536489152e-05,
      "loss": 0.0393,
      "step": 8000
    },
    {
      "epoch": 3.9447731755424065,
      "eval_accuracy": 0.9863295888109194,
      "eval_f1": 0.5470165719919252,
      "eval_loss": 0.04064860939979553,
      "eval_precision": 0.7069530396796505,
      "eval_recall": 0.4460949464012251,
      "eval_runtime": 14.3504,
      "eval_samples_per_second": 565.352,
      "eval_steps_per_second": 2.23,
      "step": 8000
    },
    {
      "epoch": 4.191321499013807,
      "grad_norm": 0.41485270857810974,
      "learning_rate": 1.916173570019724e-05,
      "loss": 0.0361,
      "step": 8500
    },
    {
      "epoch": 4.437869822485207,
      "grad_norm": 0.4282112717628479,
      "learning_rate": 1.911242603550296e-05,
      "loss": 0.0349,
      "step": 9000
    },
    {
      "epoch": 4.437869822485207,
      "eval_accuracy": 0.9869572178042619,
      "eval_f1": 0.5758776375195799,
      "eval_loss": 0.0391574390232563,
      "eval_precision": 0.7228776312745778,
      "eval_recall": 0.4785604900459418,
      "eval_runtime": 14.3708,
      "eval_samples_per_second": 564.548,
      "eval_steps_per_second": 2.227,
      "step": 9000
    },
    {
      "epoch": 4.684418145956608,
      "grad_norm": 0.5737874507904053,
      "learning_rate": 1.906311637080868e-05,
      "loss": 0.0358,
      "step": 9500
    },
    {
      "epoch": 4.930966469428008,
      "grad_norm": 0.42196542024612427,
      "learning_rate": 1.90138067061144e-05,
      "loss": 0.0344,
      "step": 10000
    },
    {
      "epoch": 4.930966469428008,
      "eval_accuracy": 0.9871839009621284,
      "eval_f1": 0.5806989895244276,
      "eval_loss": 0.03859451785683632,
      "eval_precision": 0.7357293868921776,
      "eval_recall": 0.4796324655436447,
      "eval_runtime": 14.346,
      "eval_samples_per_second": 565.525,
      "eval_steps_per_second": 2.231,
      "step": 10000
    },
    {
      "epoch": 5.177514792899408,
      "grad_norm": 0.2073846012353897,
      "learning_rate": 1.896459566074951e-05,
      "loss": 0.0315,
      "step": 10500
    },
    {
      "epoch": 5.424063116370808,
      "grad_norm": 0.3816153109073639,
      "learning_rate": 1.891528599605523e-05,
      "loss": 0.0302,
      "step": 11000
    },
    {
      "epoch": 5.424063116370808,
      "eval_accuracy": 0.9873298282450048,
      "eval_f1": 0.5949911688782211,
      "eval_loss": 0.03807613626122475,
      "eval_precision": 0.7281897794036137,
      "eval_recall": 0.5029862174578866,
      "eval_runtime": 14.3913,
      "eval_samples_per_second": 563.744,
      "eval_steps_per_second": 2.224,
      "step": 11000
    },
    {
      "epoch": 5.670611439842209,
      "grad_norm": 0.4880998730659485,
      "learning_rate": 1.886597633136095e-05,
      "loss": 0.0321,
      "step": 11500
    },
    {
      "epoch": 5.9171597633136095,
      "grad_norm": 0.31077542901039124,
      "learning_rate": 1.881666666666667e-05,
      "loss": 0.0305,
      "step": 12000
    },
    {
      "epoch": 5.9171597633136095,
      "eval_accuracy": 0.987210819587125,
      "eval_f1": 0.5974581939799332,
      "eval_loss": 0.038143206387758255,
      "eval_precision": 0.7153230112119594,
      "eval_recall": 0.5129402756508422,
      "eval_runtime": 14.3895,
      "eval_samples_per_second": 563.816,
      "eval_steps_per_second": 2.224,
      "step": 12000
    },
    {
      "epoch": 6.16370808678501,
      "grad_norm": 0.4404219686985016,
      "learning_rate": 1.876735700197239e-05,
      "loss": 0.0281,
      "step": 12500
    },
    {
      "epoch": 6.410256410256411,
      "grad_norm": 0.39860355854034424,
      "learning_rate": 1.8718145956607497e-05,
      "loss": 0.027,
      "step": 13000
    },
    {
      "epoch": 6.410256410256411,
      "eval_accuracy": 0.987472921988408,
      "eval_f1": 0.6029636281993713,
      "eval_loss": 0.03775615990161896,
      "eval_precision": 0.7289902280130293,
      "eval_recall": 0.5140888208269525,
      "eval_runtime": 14.5744,
      "eval_samples_per_second": 556.661,
      "eval_steps_per_second": 2.196,
      "step": 13000
    },
    {
      "epoch": 6.65680473372781,
      "grad_norm": 0.3162294924259186,
      "learning_rate": 1.8668836291913217e-05,
      "loss": 0.0267,
      "step": 13500
    },
    {
      "epoch": 6.903353057199211,
      "grad_norm": 0.3698256313800812,
      "learning_rate": 1.8619526627218937e-05,
      "loss": 0.0282,
      "step": 14000
    },
    {
      "epoch": 6.903353057199211,
      "eval_accuracy": 0.9875990144949712,
      "eval_f1": 0.6094328677881398,
      "eval_loss": 0.03744753450155258,
      "eval_precision": 0.7302962250026735,
      "eval_recall": 0.5228943338437979,
      "eval_runtime": 14.4711,
      "eval_samples_per_second": 560.633,
      "eval_steps_per_second": 2.211,
      "step": 14000
    },
    {
      "epoch": 7.149901380670611,
      "grad_norm": 0.42440325021743774,
      "learning_rate": 1.8570216962524657e-05,
      "loss": 0.0252,
      "step": 14500
    },
    {
      "epoch": 7.396449704142012,
      "grad_norm": 0.5403586626052856,
      "learning_rate": 1.8521005917159765e-05,
      "loss": 0.0235,
      "step": 15000
    },
    {
      "epoch": 7.396449704142012,
      "eval_accuracy": 0.9875805964883946,
      "eval_f1": 0.6213390928725702,
      "eval_loss": 0.037772245705127716,
      "eval_precision": 0.712784935579782,
      "eval_recall": 0.5506891271056662,
      "eval_runtime": 14.5254,
      "eval_samples_per_second": 558.54,
      "eval_steps_per_second": 2.203,
      "step": 15000
    },
    {
      "epoch": 7.642998027613412,
      "grad_norm": 0.3601667582988739,
      "learning_rate": 1.8471696252465485e-05,
      "loss": 0.0243,
      "step": 15500
    },
    {
      "epoch": 7.889546351084813,
      "grad_norm": 0.4192611575126648,
      "learning_rate": 1.8422386587771205e-05,
      "loss": 0.0255,
      "step": 16000
    },
    {
      "epoch": 7.889546351084813,
      "eval_accuracy": 0.9878200305738909,
      "eval_f1": 0.6303478522595347,
      "eval_loss": 0.03719973564147949,
      "eval_precision": 0.71883887417868,
      "eval_recall": 0.5612557427258805,
      "eval_runtime": 14.5726,
      "eval_samples_per_second": 556.728,
      "eval_steps_per_second": 2.196,
      "step": 16000
    },
    {
      "epoch": 8.136094674556213,
      "grad_norm": 0.13463503122329712,
      "learning_rate": 1.8373076923076926e-05,
      "loss": 0.0224,
      "step": 16500
    },
    {
      "epoch": 8.382642998027613,
      "grad_norm": 0.25073572993278503,
      "learning_rate": 1.8323865877712033e-05,
      "loss": 0.0214,
      "step": 17000
    },
    {
      "epoch": 8.382642998027613,
      "eval_accuracy": 0.9878299479620476,
      "eval_f1": 0.6356464200882254,
      "eval_loss": 0.03777812048792839,
      "eval_precision": 0.7125332826169646,
      "eval_recall": 0.5737366003062787,
      "eval_runtime": 14.3295,
      "eval_samples_per_second": 566.174,
      "eval_steps_per_second": 2.233,
      "step": 17000
    },
    {
      "epoch": 8.629191321499015,
      "grad_norm": 0.39663228392601013,
      "learning_rate": 1.8274556213017754e-05,
      "loss": 0.0215,
      "step": 17500
    },
    {
      "epoch": 8.875739644970414,
      "grad_norm": 0.7587671279907227,
      "learning_rate": 1.822524654832347e-05,
      "loss": 0.0222,
      "step": 18000
    },
    {
      "epoch": 8.875739644970414,
      "eval_accuracy": 0.9877746939423176,
      "eval_f1": 0.6313495962746187,
      "eval_loss": 0.03808865323662758,
      "eval_precision": 0.7141200347926935,
      "eval_recall": 0.5657733537519143,
      "eval_runtime": 14.4716,
      "eval_samples_per_second": 560.614,
      "eval_steps_per_second": 2.211,
      "step": 18000
    },
    {
      "epoch": 9.122287968441814,
      "grad_norm": 0.14514634013175964,
      "learning_rate": 1.8175936883629194e-05,
      "loss": 0.0198,
      "step": 18500
    },
    {
      "epoch": 9.368836291913215,
      "grad_norm": 0.44190242886543274,
      "learning_rate": 1.8126725838264302e-05,
      "loss": 0.0192,
      "step": 19000
    },
    {
      "epoch": 9.368836291913215,
      "eval_accuracy": 0.9874545039818313,
      "eval_f1": 0.628518689432395,
      "eval_loss": 0.038962751626968384,
      "eval_precision": 0.6950913983483344,
      "eval_recall": 0.573583460949464,
      "eval_runtime": 14.4229,
      "eval_samples_per_second": 562.509,
      "eval_steps_per_second": 2.219,
      "step": 19000
    },
    {
      "epoch": 9.615384615384615,
      "grad_norm": 0.3838333785533905,
      "learning_rate": 1.8077416173570022e-05,
      "loss": 0.0193,
      "step": 19500
    },
    {
      "epoch": 9.861932938856016,
      "grad_norm": 0.36484745144844055,
      "learning_rate": 1.802810650887574e-05,
      "loss": 0.0189,
      "step": 20000
    },
    {
      "epoch": 9.861932938856016,
      "eval_accuracy": 0.9877888616396843,
      "eval_f1": 0.6364978280123149,
      "eval_loss": 0.039096660912036896,
      "eval_precision": 0.708478077175852,
      "eval_recall": 0.5777947932618683,
      "eval_runtime": 14.4016,
      "eval_samples_per_second": 563.339,
      "eval_steps_per_second": 2.222,
      "step": 20000
    },
    {
      "epoch": 10.108481262327416,
      "grad_norm": 0.18375088274478912,
      "learning_rate": 1.7978796844181462e-05,
      "loss": 0.0175,
      "step": 20500
    },
    {
      "epoch": 10.355029585798816,
      "grad_norm": 0.13978295028209686,
      "learning_rate": 1.7929487179487182e-05,
      "loss": 0.0159,
      "step": 21000
    },
    {
      "epoch": 10.355029585798816,
      "eval_accuracy": 0.9879291218436141,
      "eval_f1": 0.641504670537743,
      "eval_loss": 0.03958018124103546,
      "eval_precision": 0.7120306370259667,
      "eval_recall": 0.5836906584992343,
      "eval_runtime": 14.3748,
      "eval_samples_per_second": 564.392,
      "eval_steps_per_second": 2.226,
      "step": 21000
    }
  ],
  "logging_steps": 500,
  "max_steps": 202800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 1000,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 474932252325492.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}