last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0335195530726257,
  "eval_steps": 9,
  "global_step": 9,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0037243947858473,
      "eval_loss": 2.4661662578582764,
      "eval_runtime": 20.016,
      "eval_samples_per_second": 22.632,
      "eval_steps_per_second": 2.848,
      "step": 1
    },
    {
      "epoch": 0.0111731843575419,
      "grad_norm": 2.409907341003418,
      "learning_rate": 1.5e-05,
      "loss": 2.226,
      "step": 3
    },
    {
      "epoch": 0.0223463687150838,
      "grad_norm": 3.097743034362793,
      "learning_rate": 3e-05,
      "loss": 2.3836,
      "step": 6
    },
    {
      "epoch": 0.0335195530726257,
      "grad_norm": 2.5673141479492188,
      "learning_rate": 4.5e-05,
      "loss": 2.4005,
      "step": 9
    },
    {
      "epoch": 0.0335195530726257,
      "eval_loss": 2.459803342819214,
      "eval_runtime": 20.2549,
      "eval_samples_per_second": 22.365,
      "eval_steps_per_second": 2.814,
      "step": 9
    }
  ],
  "logging_steps": 3,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 9,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4728391571865600.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}