trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.024600246002460024,
  "eval_steps": 1,
  "global_step": 5,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.2108267843723297,
      "learning_rate": 0.0002,
      "loss": 2.191,
      "step": 1
    },
    {
      "epoch": 0.0,
      "eval_loss": 2.935906171798706,
      "eval_runtime": 132.0925,
      "eval_samples_per_second": 5.277,
      "eval_steps_per_second": 0.333,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.26715168356895447,
      "learning_rate": 0.00015000000000000001,
      "loss": 2.4457,
      "step": 2
    },
    {
      "epoch": 0.01,
      "eval_loss": 2.8475093841552734,
      "eval_runtime": 133.6468,
      "eval_samples_per_second": 5.215,
      "eval_steps_per_second": 0.329,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.264245867729187,
      "learning_rate": 0.0001,
      "loss": 2.6295,
      "step": 3
    },
    {
      "epoch": 0.01,
      "eval_loss": 2.7633216381073,
      "eval_runtime": 133.8751,
      "eval_samples_per_second": 5.206,
      "eval_steps_per_second": 0.329,
      "step": 3
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.2810538709163666,
      "learning_rate": 5e-05,
      "loss": 2.5645,
      "step": 4
    },
    {
      "epoch": 0.02,
      "eval_loss": 2.7106192111968994,
      "eval_runtime": 133.565,
      "eval_samples_per_second": 5.218,
      "eval_steps_per_second": 0.329,
      "step": 4
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3025140166282654,
      "learning_rate": 0.0,
      "loss": 2.6477,
      "step": 5
    },
    {
      "epoch": 0.02,
      "eval_loss": 2.6791343688964844,
      "eval_runtime": 133.83,
      "eval_samples_per_second": 5.208,
      "eval_steps_per_second": 0.329,
      "step": 5
    }
  ],
  "logging_steps": 1,
  "max_steps": 5,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1,
  "total_flos": 182343221821440.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}