trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.125,
  "eval_steps": 2500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "grad_norm": 2.9973244667053223,
      "learning_rate": 1.971072463768116e-05,
      "loss": 1.6969,
      "step": 1000
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.261388778686523,
      "learning_rate": 1.9131014492753626e-05,
      "loss": 1.5602,
      "step": 2000
    },
    {
      "epoch": 0.06,
      "eval_loss": 1.513683795928955,
      "eval_runtime": 202.2713,
      "eval_samples_per_second": 4.944,
      "eval_steps_per_second": 4.944,
      "step": 2500
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.314629316329956,
      "learning_rate": 1.8551304347826088e-05,
      "loss": 1.5222,
      "step": 3000
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.6707098484039307,
      "learning_rate": 1.7971594202898553e-05,
      "loss": 1.4933,
      "step": 4000
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.160583257675171,
      "learning_rate": 1.7392463768115943e-05,
      "loss": 1.5056,
      "step": 5000
    },
    {
      "epoch": 0.12,
      "eval_loss": 1.4657692909240723,
      "eval_runtime": 201.9515,
      "eval_samples_per_second": 4.952,
      "eval_steps_per_second": 4.952,
      "step": 5000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 35000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2500,
  "total_flos": 8.051062996992e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}