last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.00043835616438356166,
  "eval_steps": 5,
  "global_step": 5,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.767123287671233e-05,
      "grad_norm": 17.826961517333984,
      "learning_rate": 1e-05,
      "loss": 4.9034,
      "step": 1
    },
    {
      "epoch": 8.767123287671233e-05,
      "eval_loss": 5.864974498748779,
      "eval_runtime": 4180.0734,
      "eval_samples_per_second": 9.192,
      "eval_steps_per_second": 4.596,
      "step": 1
    },
    {
      "epoch": 0.00017534246575342467,
      "grad_norm": 15.42636489868164,
      "learning_rate": 2e-05,
      "loss": 4.5406,
      "step": 2
    },
    {
      "epoch": 0.00026301369863013696,
      "grad_norm": 15.912243843078613,
      "learning_rate": 3e-05,
      "loss": 4.8244,
      "step": 3
    },
    {
      "epoch": 0.00035068493150684934,
      "grad_norm": 15.158151626586914,
      "learning_rate": 4e-05,
      "loss": 4.8005,
      "step": 4
    },
    {
      "epoch": 0.00043835616438356166,
      "grad_norm": 14.64368724822998,
      "learning_rate": 5e-05,
      "loss": 4.4149,
      "step": 5
    },
    {
      "epoch": 0.00043835616438356166,
      "eval_loss": 4.748353958129883,
      "eval_runtime": 4179.9748,
      "eval_samples_per_second": 9.192,
      "eval_steps_per_second": 4.596,
      "step": 5
    }
  ],
  "logging_steps": 1,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2883817203302400.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}