trainer_state.json

{
  "best_metric": 0.7168,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0556/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": 75.2810287475586,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 3.6539,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5704,
      "eval_loss": 3.159108877182007,
      "eval_runtime": 12.5507,
      "eval_samples_per_second": 298.788,
      "eval_steps_per_second": 4.701,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 457.2284240722656,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 2.2987,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6954666666666667,
      "eval_loss": 1.7309819459915161,
      "eval_runtime": 12.8257,
      "eval_samples_per_second": 292.381,
      "eval_steps_per_second": 4.6,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 117.46577453613281,
      "learning_rate": 0.0,
      "loss": 1.5606,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7168,
      "eval_loss": 1.5515094995498657,
      "eval_runtime": 12.5402,
      "eval_samples_per_second": 299.038,
      "eval_steps_per_second": 4.705,
      "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
}