trainer_state.json

{
  "best_metric": 0.13029101490974426,
  "best_model_checkpoint": "/home/jcanete/ft-data/all_results/pos/roberta_large_bne/epochs_2_bs_16_lr_5e-5/checkpoint-1600",
  "epoch": 2.0,
  "global_step": 1790,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.22,
      "eval_accuracy": 0.8726584582954056,
      "eval_f1": 0.8425117528542645,
      "eval_loss": 0.38713014125823975,
      "eval_precision": 0.8330501357395363,
      "eval_recall": 0.8521907654192723,
      "eval_runtime": 3.9474,
      "eval_samples_per_second": 419.014,
      "eval_steps_per_second": 26.347,
      "step": 200
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.9134448010987081,
      "eval_f1": 0.8921425085407516,
      "eval_loss": 0.26501840353012085,
      "eval_precision": 0.889493545183714,
      "eval_recall": 0.8948072965575113,
      "eval_runtime": 3.9579,
      "eval_samples_per_second": 417.895,
      "eval_steps_per_second": 26.276,
      "step": 400
    },
    {
      "epoch": 0.56,
      "learning_rate": 3.608938547486034e-05,
      "loss": 0.4031,
      "step": 500
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.9339657292515051,
      "eval_f1": 0.915745815140092,
      "eval_loss": 0.20921829342842102,
      "eval_precision": 0.9117647058823529,
      "eval_recall": 0.9197618429202213,
      "eval_runtime": 3.8819,
      "eval_samples_per_second": 426.082,
      "eval_steps_per_second": 26.791,
      "step": 600
    },
    {
      "epoch": 0.89,
      "eval_accuracy": 0.9399543269614654,
      "eval_f1": 0.9255756947696724,
      "eval_loss": 0.18737097084522247,
      "eval_precision": 0.9232298334194728,
      "eval_recall": 0.9279335078220216,
      "eval_runtime": 3.8865,
      "eval_samples_per_second": 425.58,
      "eval_steps_per_second": 26.76,
      "step": 800
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.212290502793296e-05,
      "loss": 0.1622,
      "step": 1000
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.9501908366470241,
      "eval_f1": 0.9376133984647592,
      "eval_loss": 0.16575469076633453,
      "eval_precision": 0.935673212757914,
      "eval_recall": 0.9395616471199376,
      "eval_runtime": 3.8706,
      "eval_samples_per_second": 427.321,
      "eval_steps_per_second": 26.869,
      "step": 1000
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.9536402689279612,
      "eval_f1": 0.9416929400965798,
      "eval_loss": 0.1559223085641861,
      "eval_precision": 0.940510592502541,
      "eval_recall": 0.942878264170546,
      "eval_runtime": 3.799,
      "eval_samples_per_second": 435.378,
      "eval_steps_per_second": 27.376,
      "step": 1200
    },
    {
      "epoch": 1.56,
      "eval_accuracy": 0.9577604241524138,
      "eval_f1": 0.9459559227001088,
      "eval_loss": 0.13980740308761597,
      "eval_precision": 0.9447212147782074,
      "eval_recall": 0.9471938622604943,
      "eval_runtime": 3.8091,
      "eval_samples_per_second": 434.224,
      "eval_steps_per_second": 27.303,
      "step": 1400
    },
    {
      "epoch": 1.68,
      "learning_rate": 8.156424581005587e-06,
      "loss": 0.0793,
      "step": 1500
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.960938373337166,
      "eval_f1": 0.951597772499551,
      "eval_loss": 0.13029101490974426,
      "eval_precision": 0.9506490398995034,
      "eval_recall": 0.952548400631356,
      "eval_runtime": 3.835,
      "eval_samples_per_second": 431.293,
      "eval_steps_per_second": 27.119,
      "step": 1600
    },
    {
      "epoch": 2.0,
      "step": 1790,
      "total_flos": 4231215425384640.0,
      "train_loss": 0.19050939762392524,
      "train_runtime": 272.28,
      "train_samples_per_second": 105.076,
      "train_steps_per_second": 6.574
    }
  ],
  "max_steps": 1790,
  "num_train_epochs": 2,
  "total_flos": 4231215425384640.0,
  "trial_name": null,
  "trial_params": null
}