trainer_state.json

{
  "best_metric": 0.92,
  "best_model_checkpoint": "roberta-base/checkpoint-24",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.67,
      "eval_f1": 0.5376047904191618,
      "eval_loss": 0.6319047808647156,
      "eval_precision": 0.4489,
      "eval_recall": 0.67,
      "eval_runtime": 0.4159,
      "eval_samples_per_second": 240.427,
      "eval_steps_per_second": 2.404,
      "step": 3
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.67,
      "eval_f1": 0.5376047904191618,
      "eval_loss": 0.6274762153625488,
      "eval_precision": 0.4489,
      "eval_recall": 0.67,
      "eval_runtime": 0.4183,
      "eval_samples_per_second": 239.081,
      "eval_steps_per_second": 2.391,
      "step": 6
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.67,
      "eval_f1": 0.5376047904191618,
      "eval_loss": 0.6107555627822876,
      "eval_precision": 0.4489,
      "eval_recall": 0.67,
      "eval_runtime": 0.4167,
      "eval_samples_per_second": 240.008,
      "eval_steps_per_second": 2.4,
      "step": 9
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.69,
      "eval_f1": 0.5818354978354978,
      "eval_loss": 0.6286537051200867,
      "eval_precision": 0.788061224489796,
      "eval_recall": 0.69,
      "eval_runtime": 0.4169,
      "eval_samples_per_second": 239.86,
      "eval_steps_per_second": 2.399,
      "step": 12
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.72,
      "eval_f1": 0.7282758620689656,
      "eval_loss": 0.5415282249450684,
      "eval_precision": 0.7828091236494598,
      "eval_recall": 0.72,
      "eval_runtime": 0.4173,
      "eval_samples_per_second": 239.629,
      "eval_steps_per_second": 2.396,
      "step": 15
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.85,
      "eval_f1": 0.851543312313309,
      "eval_loss": 0.35827144980430603,
      "eval_precision": 0.8546875,
      "eval_recall": 0.85,
      "eval_runtime": 0.4185,
      "eval_samples_per_second": 238.934,
      "eval_steps_per_second": 2.389,
      "step": 18
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.88,
      "eval_f1": 0.8815824175824176,
      "eval_loss": 0.3416476845741272,
      "eval_precision": 0.8861089661089662,
      "eval_recall": 0.88,
      "eval_runtime": 0.4254,
      "eval_samples_per_second": 235.093,
      "eval_steps_per_second": 2.351,
      "step": 21
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.92,
      "eval_f1": 0.92,
      "eval_loss": 0.2903079688549042,
      "eval_precision": 0.92,
      "eval_recall": 0.92,
      "eval_runtime": 0.4202,
      "eval_samples_per_second": 237.971,
      "eval_steps_per_second": 2.38,
      "step": 24
    }
  ],
  "logging_steps": 500,
  "max_steps": 60,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 631466532864000.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}