trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.9689440993788825,
  "eval_steps": 200,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.83,
      "learning_rate": 1e-05,
      "loss": 6.109,
      "step": 100
    },
    {
      "epoch": 1.66,
      "learning_rate": 2e-05,
      "loss": 1.7606,
      "step": 200
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.9147409796714783,
      "eval_runtime": 3.2133,
      "eval_samples_per_second": 120.438,
      "eval_steps_per_second": 15.249,
      "step": 200
    },
    {
      "epoch": 2.48,
      "learning_rate": 3e-05,
      "loss": 0.8584,
      "step": 300
    },
    {
      "epoch": 3.31,
      "learning_rate": 4e-05,
      "loss": 0.579,
      "step": 400
    },
    {
      "epoch": 3.31,
      "eval_loss": 0.3369111120700836,
      "eval_runtime": 3.2981,
      "eval_samples_per_second": 117.339,
      "eval_steps_per_second": 14.857,
      "step": 400
    },
    {
      "epoch": 4.14,
      "learning_rate": 5e-05,
      "loss": 0.4074,
      "step": 500
    },
    {
      "epoch": 4.97,
      "learning_rate": 0.0,
      "loss": 0.2482,
      "step": 600
    },
    {
      "epoch": 4.97,
      "eval_loss": 0.11852475255727768,
      "eval_runtime": 3.2809,
      "eval_samples_per_second": 117.954,
      "eval_steps_per_second": 14.935,
      "step": 600
    },
    {
      "epoch": 4.97,
      "step": 600,
      "total_flos": 510083481600000.0,
      "train_loss": 1.6604227765401205,
      "train_runtime": 352.86,
      "train_samples_per_second": 27.376,
      "train_steps_per_second": 1.7
    }
  ],
  "logging_steps": 100,
  "max_steps": 600,
  "num_train_epochs": 5,
  "save_steps": 1000,
  "total_flos": 510083481600000.0,
  "trial_name": null,
  "trial_params": null
}