checkpoint-100/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2518891687657431,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 7.8137,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 2.5574,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 2.3516,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 2.0521,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.5531,
      "step": 50
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 2.0324,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.704,
      "step": 70
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.8253,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 2.1076,
      "step": 90
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.8701,
      "step": 100
    }
  ],
  "logging_steps": 10,
  "max_steps": 100,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 1507829183594496.0,
  "trial_name": null,
  "trial_params": null
}