last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.01972386587771203,
  "eval_steps": 3,
  "global_step": 5,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0039447731755424065,
      "grad_norm": 12.800737380981445,
      "learning_rate": 2e-05,
      "loss": 4.4657,
      "step": 1
    },
    {
      "epoch": 0.0039447731755424065,
      "eval_loss": 4.758890151977539,
      "eval_runtime": 22.2538,
      "eval_samples_per_second": 9.616,
      "eval_steps_per_second": 4.808,
      "step": 1
    },
    {
      "epoch": 0.007889546351084813,
      "grad_norm": 12.97591781616211,
      "learning_rate": 4e-05,
      "loss": 4.9917,
      "step": 2
    },
    {
      "epoch": 0.011834319526627219,
      "grad_norm": 12.59117603302002,
      "learning_rate": 6e-05,
      "loss": 4.6719,
      "step": 3
    },
    {
      "epoch": 0.011834319526627219,
      "eval_loss": 4.388483047485352,
      "eval_runtime": 22.2476,
      "eval_samples_per_second": 9.619,
      "eval_steps_per_second": 4.81,
      "step": 3
    },
    {
      "epoch": 0.015779092702169626,
      "grad_norm": 13.937124252319336,
      "learning_rate": 8e-05,
      "loss": 4.4163,
      "step": 4
    },
    {
      "epoch": 0.01972386587771203,
      "grad_norm": 13.813149452209473,
      "learning_rate": 0.0001,
      "loss": 3.3764,
      "step": 5
    }
  ],
  "logging_steps": 1,
  "max_steps": 25,
  "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": 6534586045562880.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}