trainer_state.json

{
  "best_metric": 0.8984,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0954/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": 58.31099319458008,
      "learning_rate": 0.0003,
      "loss": 1.4301,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8698666666666667,
      "eval_loss": 0.45358923077583313,
      "eval_runtime": 8.9048,
      "eval_samples_per_second": 421.119,
      "eval_steps_per_second": 6.626,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 38.741695404052734,
      "learning_rate": 0.0003,
      "loss": 0.3501,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.38950836658477783,
      "eval_runtime": 8.8373,
      "eval_samples_per_second": 424.338,
      "eval_steps_per_second": 6.676,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 81.23963165283203,
      "learning_rate": 0.0003,
      "loss": 0.1928,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8984,
      "eval_loss": 0.369695782661438,
      "eval_runtime": 8.8079,
      "eval_samples_per_second": 425.756,
      "eval_steps_per_second": 6.699,
      "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
}