trainer_state.json

{
  "best_metric": 0.8968,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0871/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 6.003003003003003e-05,
      "loss": 1.6681,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7824,
      "eval_loss": 0.7962325811386108,
      "eval_runtime": 7.0344,
      "eval_samples_per_second": 533.095,
      "eval_steps_per_second": 8.387,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.086453437805176,
      "learning_rate": 5.0030030030030024e-05,
      "loss": 0.5448,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8405333333333334,
      "eval_loss": 0.551358699798584,
      "eval_runtime": 7.0345,
      "eval_samples_per_second": 533.084,
      "eval_steps_per_second": 8.387,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5800143480300903,
      "learning_rate": 4.003003003003003e-05,
      "loss": 0.3217,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.4897666871547699,
      "eval_runtime": 7.034,
      "eval_samples_per_second": 533.128,
      "eval_steps_per_second": 8.388,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6008933782577515,
      "learning_rate": 3.003003003003003e-05,
      "loss": 0.2052,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.4203386902809143,
      "eval_runtime": 7.0128,
      "eval_samples_per_second": 534.74,
      "eval_steps_per_second": 8.413,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 2.006006006006006e-05,
      "loss": 0.1245,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8826666666666667,
      "eval_loss": 0.40465182065963745,
      "eval_runtime": 6.8995,
      "eval_samples_per_second": 543.519,
      "eval_steps_per_second": 8.551,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.10143309086561203,
      "learning_rate": 1.006006006006006e-05,
      "loss": 0.0715,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8957333333333334,
      "eval_loss": 0.39617347717285156,
      "eval_runtime": 7.34,
      "eval_samples_per_second": 510.898,
      "eval_steps_per_second": 8.038,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.11448998749256134,
      "learning_rate": 6.006006006006005e-08,
      "loss": 0.0394,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.3844410181045532,
      "eval_runtime": 8.2886,
      "eval_samples_per_second": 452.428,
      "eval_steps_per_second": 7.118,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}