trainer_state.json

{
  "best_metric": 0.8824,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0314/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": 44.44316482543945,
      "learning_rate": 0.000375,
      "loss": 1.4442,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8144,
      "eval_loss": 0.5967246294021606,
      "eval_runtime": 9.1361,
      "eval_samples_per_second": 410.458,
      "eval_steps_per_second": 6.458,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 53.689815521240234,
      "learning_rate": 0.00012500000000000006,
      "loss": 0.4506,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8701333333333333,
      "eval_loss": 0.40578535199165344,
      "eval_runtime": 9.6067,
      "eval_samples_per_second": 390.352,
      "eval_steps_per_second": 6.142,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 22.607912063598633,
      "learning_rate": 0.0,
      "loss": 0.2247,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.3803592324256897,
      "eval_runtime": 9.3139,
      "eval_samples_per_second": 402.622,
      "eval_steps_per_second": 6.335,
      "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
}