checkpoint-1000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8,
  "eval_steps": 200,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 0.0,
      "loss": 292.3211,
      "step": 200
    },
    {
      "epoch": 0.16,
      "eval_loss": NaN,
      "eval_runtime": 67.797,
      "eval_samples_per_second": 3.687,
      "eval_steps_per_second": 0.472,
      "step": 200
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0,
      "loss": 76.016,
      "step": 400
    },
    {
      "epoch": 0.32,
      "eval_loss": NaN,
      "eval_runtime": 67.8532,
      "eval_samples_per_second": 3.684,
      "eval_steps_per_second": 0.472,
      "step": 400
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0,
      "loss": 1.954,
      "step": 600
    },
    {
      "epoch": 0.48,
      "eval_loss": NaN,
      "eval_runtime": 67.8787,
      "eval_samples_per_second": 3.683,
      "eval_steps_per_second": 0.471,
      "step": 600
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0,
      "loss": 66.7608,
      "step": 800
    },
    {
      "epoch": 0.64,
      "eval_loss": NaN,
      "eval_runtime": 67.9031,
      "eval_samples_per_second": 3.682,
      "eval_steps_per_second": 0.471,
      "step": 800
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0,
      "loss": 15210.5125,
      "step": 1000
    },
    {
      "epoch": 0.8,
      "eval_loss": NaN,
      "eval_runtime": 67.7641,
      "eval_samples_per_second": 3.689,
      "eval_steps_per_second": 0.472,
      "step": 1000
    }
  ],
  "logging_steps": 200,
  "max_steps": 1250,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 8.1216808157184e+16,
  "trial_name": null,
  "trial_params": null
}