trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 6315,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9357798165137615,
      "eval_loss": 0.2056138515472412,
      "eval_runtime": 0.7658,
      "eval_samples_per_second": 1138.649,
      "eval_steps_per_second": 36.562,
      "step": 2105
    },
    {
      "epoch": 1.19,
      "learning_rate": 2.6523754345307068e-05,
      "loss": 0.2549,
      "step": 2500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9438073394495413,
      "eval_loss": 0.18503105640411377,
      "eval_runtime": 0.7645,
      "eval_samples_per_second": 1140.671,
      "eval_steps_per_second": 36.627,
      "step": 4210
    },
    {
      "epoch": 2.38,
      "learning_rate": 9.142526071842411e-06,
      "loss": 0.1162,
      "step": 5000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9506880733944955,
      "eval_loss": 0.1753561794757843,
      "eval_runtime": 0.7638,
      "eval_samples_per_second": 1141.657,
      "eval_steps_per_second": 36.659,
      "step": 6315
    },
    {
      "epoch": 3.0,
      "step": 6315,
      "total_flos": 1449426901211460.0,
      "train_loss": 0.16375755859761604,
      "train_runtime": 455.2846,
      "train_samples_per_second": 443.782,
      "train_steps_per_second": 13.87
    }
  ],
  "max_steps": 6315,
  "num_train_epochs": 3,
  "total_flos": 1449426901211460.0,
  "trial_name": null,
  "trial_params": null
}