{
  "best_metric": 0.037242207676172256,
  "best_model_checkpoint": "doc-topic-model_eval-01_train-00/checkpoint-14000",
  "epoch": 9.368836291913215,
  "eval_steps": 1000,
  "global_step": 19000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2465483234714004,
      "grad_norm": 0.32839319109916687,
      "learning_rate": 1.9950690335305722e-05,
      "loss": 0.1668,
      "step": 500
    },
    {
      "epoch": 0.4930966469428008,
      "grad_norm": 0.4088590443134308,
      "learning_rate": 1.9901380670611442e-05,
      "loss": 0.0929,
      "step": 1000
    },
    {
      "epoch": 0.4930966469428008,
      "eval_accuracy": 0.981383845685049,
      "eval_f1": 0.0,
      "eval_loss": 0.0910971537232399,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 12.1172,
      "eval_samples_per_second": 669.294,
      "eval_steps_per_second": 2.641,
      "step": 1000
    },
    {
      "epoch": 0.7396449704142012,
      "grad_norm": 0.33391374349594116,
      "learning_rate": 1.9852071005917162e-05,
      "loss": 0.0874,
      "step": 1500
    },
    {
      "epoch": 0.9861932938856016,
      "grad_norm": 0.3162483870983124,
      "learning_rate": 1.980276134122288e-05,
      "loss": 0.0785,
      "step": 2000
    },
    {
      "epoch": 0.9861932938856016,
      "eval_accuracy": 0.981383845685049,
      "eval_f1": 0.0,
      "eval_loss": 0.07068673521280289,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 12.2907,
      "eval_samples_per_second": 659.85,
      "eval_steps_per_second": 2.604,
      "step": 2000
    },
    {
      "epoch": 1.232741617357002,
      "grad_norm": 0.31390729546546936,
      "learning_rate": 1.9753451676528602e-05,
      "loss": 0.0684,
      "step": 2500
    },
    {
      "epoch": 1.4792899408284024,
      "grad_norm": 0.40165144205093384,
      "learning_rate": 1.9704142011834322e-05,
      "loss": 0.0622,
      "step": 3000
    },
    {
      "epoch": 1.4792899408284024,
      "eval_accuracy": 0.9822611505591224,
      "eval_f1": 0.10574449842812232,
      "eval_loss": 0.05743462219834328,
      "eval_precision": 0.859465737514518,
      "eval_recall": 0.056338028169014086,
      "eval_runtime": 14.6119,
      "eval_samples_per_second": 555.025,
      "eval_steps_per_second": 2.19,
      "step": 3000
    },
    {
      "epoch": 1.725838264299803,
      "grad_norm": 0.4135937988758087,
      "learning_rate": 1.965483234714004e-05,
      "loss": 0.0584,
      "step": 3500
    },
    {
      "epoch": 1.972386587771203,
      "grad_norm": 0.324859082698822,
      "learning_rate": 1.9605522682445763e-05,
      "loss": 0.0542,
      "step": 4000
    },
    {
      "epoch": 1.972386587771203,
      "eval_accuracy": 0.9842850461329138,
      "eval_f1": 0.33960690887433,
      "eval_loss": 0.04979131370782852,
      "eval_precision": 0.7800273597811217,
      "eval_recall": 0.21705367339170156,
      "eval_runtime": 14.5784,
      "eval_samples_per_second": 556.302,
      "eval_steps_per_second": 2.195,
      "step": 4000
    },
    {
      "epoch": 2.2189349112426036,
      "grad_norm": 0.3279891312122345,
      "learning_rate": 1.9556213017751483e-05,
      "loss": 0.0488,
      "step": 4500
    },
    {
      "epoch": 2.465483234714004,
      "grad_norm": 0.4424910843372345,
      "learning_rate": 1.95069033530572e-05,
      "loss": 0.048,
      "step": 5000
    },
    {
      "epoch": 2.465483234714004,
      "eval_accuracy": 0.985223294641212,
      "eval_f1": 0.4251213056903397,
      "eval_loss": 0.04605511203408241,
      "eval_precision": 0.7708458308338332,
      "eval_recall": 0.29349067377236393,
      "eval_runtime": 14.4651,
      "eval_samples_per_second": 560.659,
      "eval_steps_per_second": 2.212,
      "step": 5000
    },
    {
      "epoch": 2.712031558185404,
      "grad_norm": 0.3125520646572113,
      "learning_rate": 1.9457593688362923e-05,
      "loss": 0.0453,
      "step": 5500
    },
    {
      "epoch": 2.9585798816568047,
      "grad_norm": 0.4686921536922455,
      "learning_rate": 1.940828402366864e-05,
      "loss": 0.0436,
      "step": 6000
    },
    {
      "epoch": 2.9585798816568047,
      "eval_accuracy": 0.9860113100046771,
      "eval_f1": 0.5031212243254128,
      "eval_loss": 0.04325457662343979,
      "eval_precision": 0.7426066280279388,
      "eval_recall": 0.3804339550818424,
      "eval_runtime": 14.4821,
      "eval_samples_per_second": 560.003,
      "eval_steps_per_second": 2.21,
      "step": 6000
    },
    {
      "epoch": 3.2051282051282053,
      "grad_norm": 0.3562796711921692,
      "learning_rate": 1.935897435897436e-05,
      "loss": 0.0412,
      "step": 6500
    },
    {
      "epoch": 3.4516765285996054,
      "grad_norm": 0.24857360124588013,
      "learning_rate": 1.930966469428008e-05,
      "loss": 0.0384,
      "step": 7000
    },
    {
      "epoch": 3.4516765285996054,
      "eval_accuracy": 0.9864563402639001,
      "eval_f1": 0.5389366013702596,
      "eval_loss": 0.04127802699804306,
      "eval_precision": 0.7357396917402187,
      "eval_recall": 0.425199847735059,
      "eval_runtime": 14.4965,
      "eval_samples_per_second": 559.446,
      "eval_steps_per_second": 2.207,
      "step": 7000
    },
    {
      "epoch": 3.698224852071006,
      "grad_norm": 0.47940972447395325,
      "learning_rate": 1.92603550295858e-05,
      "loss": 0.0394,
      "step": 7500
    },
    {
      "epoch": 3.9447731755424065,
      "grad_norm": 0.39106041193008423,
      "learning_rate": 1.921104536489152e-05,
      "loss": 0.0385,
      "step": 8000
    },
    {
      "epoch": 3.9447731755424065,
      "eval_accuracy": 0.9867043666822569,
      "eval_f1": 0.536168108776267,
      "eval_loss": 0.0398826077580452,
      "eval_precision": 0.7647390691114245,
      "eval_recall": 0.41279025504377614,
      "eval_runtime": 14.6187,
      "eval_samples_per_second": 554.767,
      "eval_steps_per_second": 2.189,
      "step": 8000
    },
    {
      "epoch": 4.191321499013807,
      "grad_norm": 0.4470931887626648,
      "learning_rate": 1.916173570019724e-05,
      "loss": 0.0342,
      "step": 8500
    },
    {
      "epoch": 4.437869822485207,
      "grad_norm": 0.33709824085235596,
      "learning_rate": 1.911242603550296e-05,
      "loss": 0.0343,
      "step": 9000
    },
    {
      "epoch": 4.437869822485207,
      "eval_accuracy": 0.986877276528197,
      "eval_f1": 0.5599125433718333,
      "eval_loss": 0.03963331878185272,
      "eval_precision": 0.7451923076923077,
      "eval_recall": 0.4484202512371526,
      "eval_runtime": 14.6638,
      "eval_samples_per_second": 553.061,
      "eval_steps_per_second": 2.182,
      "step": 9000
    },
    {
      "epoch": 4.684418145956608,
      "grad_norm": 0.3310143053531647,
      "learning_rate": 1.906311637080868e-05,
      "loss": 0.0343,
      "step": 9500
    },
    {
      "epoch": 4.930966469428008,
      "grad_norm": 0.5156289935112,
      "learning_rate": 1.90138067061144e-05,
      "loss": 0.0343,
      "step": 10000
    },
    {
      "epoch": 4.930966469428008,
      "eval_accuracy": 0.9870459344926796,
      "eval_f1": 0.5692336695258743,
      "eval_loss": 0.038666386157274246,
      "eval_precision": 0.7471235927254732,
      "eval_recall": 0.4597639893414541,
      "eval_runtime": 14.4377,
      "eval_samples_per_second": 561.722,
      "eval_steps_per_second": 2.216,
      "step": 10000
    },
    {
      "epoch": 5.177514792899408,
      "grad_norm": 0.3030126988887787,
      "learning_rate": 1.896459566074951e-05,
      "loss": 0.0314,
      "step": 10500
    },
    {
      "epoch": 5.424063116370808,
      "grad_norm": 0.49180835485458374,
      "learning_rate": 1.891528599605523e-05,
      "loss": 0.0304,
      "step": 11000
    },
    {
      "epoch": 5.424063116370808,
      "eval_accuracy": 0.9872783706790255,
      "eval_f1": 0.586054233536248,
      "eval_loss": 0.03851797431707382,
      "eval_precision": 0.7432448239560182,
      "eval_recall": 0.48374571754853446,
      "eval_runtime": 14.5078,
      "eval_samples_per_second": 559.011,
      "eval_steps_per_second": 2.206,
      "step": 11000
    },
    {
      "epoch": 5.670611439842209,
      "grad_norm": 0.4685238003730774,
      "learning_rate": 1.886597633136095e-05,
      "loss": 0.0297,
      "step": 11500
    },
    {
      "epoch": 5.9171597633136095,
      "grad_norm": 0.46912193298339844,
      "learning_rate": 1.881666666666667e-05,
      "loss": 0.0299,
      "step": 12000
    },
    {
      "epoch": 5.9171597633136095,
      "eval_accuracy": 0.9875023031024561,
      "eval_f1": 0.6054939155332856,
      "eval_loss": 0.037345413118600845,
      "eval_precision": 0.7341868286861235,
      "eval_recall": 0.5151884278644842,
      "eval_runtime": 14.7174,
      "eval_samples_per_second": 551.047,
      "eval_steps_per_second": 2.174,
      "step": 12000
    },
    {
      "epoch": 6.16370808678501,
      "grad_norm": 0.3353117108345032,
      "learning_rate": 1.876735700197239e-05,
      "loss": 0.0276,
      "step": 12500
    },
    {
      "epoch": 6.410256410256411,
      "grad_norm": 0.2552751898765564,
      "learning_rate": 1.8718145956607497e-05,
      "loss": 0.0265,
      "step": 13000
    },
    {
      "epoch": 6.410256410256411,
      "eval_accuracy": 0.987298212792494,
      "eval_f1": 0.6069298245614035,
      "eval_loss": 0.03756754845380783,
      "eval_precision": 0.715882048629074,
      "eval_recall": 0.5267605633802817,
      "eval_runtime": 14.9359,
      "eval_samples_per_second": 542.988,
      "eval_steps_per_second": 2.142,
      "step": 13000
    },
    {
      "epoch": 6.65680473372781,
      "grad_norm": 0.2560911774635315,
      "learning_rate": 1.8668836291913217e-05,
      "loss": 0.0273,
      "step": 13500
    },
    {
      "epoch": 6.903353057199211,
      "grad_norm": 0.4542127549648285,
      "learning_rate": 1.8619526627218937e-05,
      "loss": 0.0261,
      "step": 14000
    },
    {
      "epoch": 6.903353057199211,
      "eval_accuracy": 0.987696472355684,
      "eval_f1": 0.6137829781554478,
      "eval_loss": 0.037242207676172256,
      "eval_precision": 0.7383857846285592,
      "eval_recall": 0.5251617814998096,
      "eval_runtime": 14.4235,
      "eval_samples_per_second": 562.276,
      "eval_steps_per_second": 2.219,
      "step": 14000
    },
    {
      "epoch": 7.149901380670611,
      "grad_norm": 0.2716215252876282,
      "learning_rate": 1.8570216962524657e-05,
      "loss": 0.025,
      "step": 14500
    },
    {
      "epoch": 7.396449704142012,
      "grad_norm": 0.2959195077419281,
      "learning_rate": 1.8520907297830377e-05,
      "loss": 0.0236,
      "step": 15000
    },
    {
      "epoch": 7.396449704142012,
      "eval_accuracy": 0.9875859234377878,
      "eval_f1": 0.6186599329531107,
      "eval_loss": 0.037820491939783096,
      "eval_precision": 0.7224933902786251,
      "eval_recall": 0.5409212028930339,
      "eval_runtime": 14.5081,
      "eval_samples_per_second": 558.998,
      "eval_steps_per_second": 2.206,
      "step": 15000
    },
    {
      "epoch": 7.642998027613412,
      "grad_norm": 0.44261014461517334,
      "learning_rate": 1.8471696252465485e-05,
      "loss": 0.0225,
      "step": 15500
    },
    {
      "epoch": 7.889546351084813,
      "grad_norm": 0.721012532711029,
      "learning_rate": 1.8422386587771205e-05,
      "loss": 0.0236,
      "step": 16000
    },
    {
      "epoch": 7.889546351084813,
      "eval_accuracy": 0.9878041866859418,
      "eval_f1": 0.6205406358865811,
      "eval_loss": 0.03788597509264946,
      "eval_precision": 0.7373716202054077,
      "eval_recall": 0.5356680624286259,
      "eval_runtime": 14.537,
      "eval_samples_per_second": 557.887,
      "eval_steps_per_second": 2.201,
      "step": 16000
    },
    {
      "epoch": 8.136094674556213,
      "grad_norm": 0.2590983510017395,
      "learning_rate": 1.8373076923076926e-05,
      "loss": 0.0219,
      "step": 16500
    },
    {
      "epoch": 8.382642998027613,
      "grad_norm": 0.6246200203895569,
      "learning_rate": 1.8323767258382646e-05,
      "loss": 0.0215,
      "step": 17000
    },
    {
      "epoch": 8.382642998027613,
      "eval_accuracy": 0.9875504910923083,
      "eval_f1": 0.6240691603184114,
      "eval_loss": 0.03826015442609787,
      "eval_precision": 0.7126380607956212,
      "eval_recall": 0.5550818424057861,
      "eval_runtime": 14.5306,
      "eval_samples_per_second": 558.131,
      "eval_steps_per_second": 2.202,
      "step": 17000
    },
    {
      "epoch": 8.629191321499015,
      "grad_norm": 0.46473428606987,
      "learning_rate": 1.8274457593688362e-05,
      "loss": 0.0204,
      "step": 17500
    },
    {
      "epoch": 8.875739644970414,
      "grad_norm": 0.5553109049797058,
      "learning_rate": 1.822524654832347e-05,
      "loss": 0.0216,
      "step": 18000
    },
    {
      "epoch": 8.875739644970414,
      "eval_accuracy": 0.9876723783607579,
      "eval_f1": 0.629683242506812,
      "eval_loss": 0.03862897306680679,
      "eval_precision": 0.7142857142857143,
      "eval_recall": 0.5629996193376475,
      "eval_runtime": 14.4315,
      "eval_samples_per_second": 561.967,
      "eval_steps_per_second": 2.217,
      "step": 18000
    },
    {
      "epoch": 9.122287968441814,
      "grad_norm": 0.3819935917854309,
      "learning_rate": 1.8175936883629194e-05,
      "loss": 0.0194,
      "step": 18500
    },
    {
      "epoch": 9.368836291913215,
      "grad_norm": 0.5073143243789673,
      "learning_rate": 1.8126627218934914e-05,
      "loss": 0.0177,
      "step": 19000
    },
    {
      "epoch": 9.368836291913215,
      "eval_accuracy": 0.9877914310415692,
      "eval_f1": 0.6353708093464273,
      "eval_loss": 0.03860396519303322,
      "eval_precision": 0.7155114882257603,
      "eval_recall": 0.571374191092501,
      "eval_runtime": 14.5627,
      "eval_samples_per_second": 556.901,
      "eval_steps_per_second": 2.197,
      "step": 19000
    }
  ],
  "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": 430075849523328.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}