trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.046207497820401,
  "eval_steps": 500,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08718395815170009,
      "grad_norm": 492.52783203125,
      "learning_rate": 5.833333333333333e-06,
      "loss": 21.8384,
      "step": 100
    },
    {
      "epoch": 0.17436791630340018,
      "grad_norm": 143.71047973632812,
      "learning_rate": 6.481481481481481e-06,
      "loss": 19.3308,
      "step": 200
    },
    {
      "epoch": 0.26155187445510025,
      "grad_norm": 805.2228393554688,
      "learning_rate": 5.833333333333333e-06,
      "loss": 21.3734,
      "step": 300
    },
    {
      "epoch": 0.34873583260680036,
      "grad_norm": 405.0468444824219,
      "learning_rate": 5.185185185185185e-06,
      "loss": 19.1667,
      "step": 400
    },
    {
      "epoch": 0.43591979075850046,
      "grad_norm": 632.6600952148438,
      "learning_rate": 4.537037037037037e-06,
      "loss": 17.4984,
      "step": 500
    },
    {
      "epoch": 0.5231037489102005,
      "grad_norm": 890.8648071289062,
      "learning_rate": 3.888888888888889e-06,
      "loss": 16.9422,
      "step": 600
    },
    {
      "epoch": 0.6102877070619006,
      "grad_norm": 243.8746337890625,
      "learning_rate": 3.2407407407407406e-06,
      "loss": 19.9537,
      "step": 700
    },
    {
      "epoch": 0.6974716652136007,
      "grad_norm": 408.9120178222656,
      "learning_rate": 2.5925925925925925e-06,
      "loss": 18.6301,
      "step": 800
    },
    {
      "epoch": 0.7846556233653008,
      "grad_norm": 171.10406494140625,
      "learning_rate": 1.9444444444444444e-06,
      "loss": 15.2552,
      "step": 900
    },
    {
      "epoch": 0.8718395815170009,
      "grad_norm": 236.40130615234375,
      "learning_rate": 1.2962962962962962e-06,
      "loss": 14.5902,
      "step": 1000
    },
    {
      "epoch": 0.9590235396687009,
      "grad_norm": 386.51702880859375,
      "learning_rate": 6.481481481481481e-07,
      "loss": 12.5311,
      "step": 1100
    },
    {
      "epoch": 1.0,
      "eval_loss": 508.4683532714844,
      "eval_runtime": 15.2151,
      "eval_samples_per_second": 67.038,
      "eval_steps_per_second": 8.413,
      "step": 1147
    },
    {
      "epoch": 1.046207497820401,
      "grad_norm": 72.9695053100586,
      "learning_rate": 0.0,
      "loss": 13.603,
      "step": 1200
    }
  ],
  "logging_steps": 100,
  "max_steps": 1200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 300,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}