checkpoint-744/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 744,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8351485729217529,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.4848,
      "step": 186
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.31648269295692444,
      "eval_runtime": 3.2931,
      "eval_samples_per_second": 4.251,
      "eval_steps_per_second": 4.251,
      "step": 186
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2281255722045898,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.2556,
      "step": 372
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2928766906261444,
      "eval_runtime": 3.0078,
      "eval_samples_per_second": 4.655,
      "eval_steps_per_second": 4.655,
      "step": 372
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9520564675331116,
      "learning_rate": 8e-05,
      "loss": 0.1603,
      "step": 558
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.3044079840183258,
      "eval_runtime": 3.4408,
      "eval_samples_per_second": 4.069,
      "eval_steps_per_second": 4.069,
      "step": 558
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4696643352508545,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.0987,
      "step": 744
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.30785349011421204,
      "eval_runtime": 3.5553,
      "eval_samples_per_second": 3.938,
      "eval_steps_per_second": 3.938,
      "step": 744
    }
  ],
  "logging_steps": 500,
  "max_steps": 2790,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3120092141322240.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}