trainer_state.json

{
  "best_metric": 0.8824,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0041/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 5e-05,
      "loss": 1.8966,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7864,
      "eval_loss": 0.8616273403167725,
      "eval_runtime": 6.5002,
      "eval_samples_per_second": 576.906,
      "eval_steps_per_second": 9.077,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.6347,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5556043386459351,
      "eval_runtime": 6.4051,
      "eval_samples_per_second": 585.474,
      "eval_steps_per_second": 9.211,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 27.954235076904297,
      "learning_rate": 5e-05,
      "loss": 0.3921,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8509333333333333,
      "eval_loss": 0.5099478363990784,
      "eval_runtime": 6.368,
      "eval_samples_per_second": 588.885,
      "eval_steps_per_second": 9.265,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0573689937591553,
      "learning_rate": 5e-05,
      "loss": 0.2729,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.4906845986843109,
      "eval_runtime": 6.3188,
      "eval_samples_per_second": 593.471,
      "eval_steps_per_second": 9.337,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7037006616592407,
      "learning_rate": 5e-05,
      "loss": 0.2046,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.4918658137321472,
      "eval_runtime": 6.3328,
      "eval_samples_per_second": 592.159,
      "eval_steps_per_second": 9.317,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 29.72886848449707,
      "learning_rate": 5e-05,
      "loss": 0.1542,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.44236132502555847,
      "eval_runtime": 6.567,
      "eval_samples_per_second": 571.041,
      "eval_steps_per_second": 8.984,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}