trainer_state.json

{
  "best_metric": 0.8341333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0074/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 216.7678985595703,
      "learning_rate": 5e-05,
      "loss": 3.7096,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.47786666666666666,
      "eval_loss": 3.165989637374878,
      "eval_runtime": 8.1134,
      "eval_samples_per_second": 462.2,
      "eval_steps_per_second": 7.272,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 230.4749755859375,
      "learning_rate": 5e-05,
      "loss": 1.8842,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7637333333333334,
      "eval_loss": 0.9839171767234802,
      "eval_runtime": 7.9559,
      "eval_samples_per_second": 471.348,
      "eval_steps_per_second": 7.416,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 120.39430236816406,
      "learning_rate": 5e-05,
      "loss": 0.7759,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8341333333333333,
      "eval_loss": 0.5666536688804626,
      "eval_runtime": 7.6335,
      "eval_samples_per_second": 491.257,
      "eval_steps_per_second": 7.729,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}