trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.92,
  "eval_steps": 500,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 5e-05,
      "loss": 2.272,
      "step": 1
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0001995184726672197,
      "loss": 1.954,
      "step": 5
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00018314696123025454,
      "loss": 0.6371,
      "step": 10
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0001471396736825998,
      "loss": 0.2788,
      "step": 15
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.23904697597026825,
      "eval_runtime": 11.1391,
      "eval_samples_per_second": 17.955,
      "eval_steps_per_second": 2.244,
      "step": 18
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001,
      "loss": 0.2252,
      "step": 20
    },
    {
      "epoch": 1.33,
      "learning_rate": 5.286032631740023e-05,
      "loss": 0.2128,
      "step": 25
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.6853038769745467e-05,
      "loss": 0.2122,
      "step": 30
    },
    {
      "epoch": 1.87,
      "learning_rate": 4.815273327803182e-07,
      "loss": 0.1993,
      "step": 35
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.2245262861251831,
      "eval_runtime": 9.6856,
      "eval_samples_per_second": 20.649,
      "eval_steps_per_second": 2.581,
      "step": 36
    },
    {
      "epoch": 1.92,
      "step": 36,
      "total_flos": 40159503319040.0,
      "train_loss": 0.5313632095025645,
      "train_runtime": 554.6126,
      "train_samples_per_second": 2.164,
      "train_steps_per_second": 0.065
    }
  ],
  "logging_steps": 5,
  "max_steps": 36,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 40159503319040.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}