trainer_state.json

{
  "best_metric": 0.8954666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0771/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": 4.532847881317139,
      "learning_rate": 8.571428571428571e-05,
      "loss": 1.5342,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7581333333333333,
      "eval_loss": 0.808803141117096,
      "eval_runtime": 7.4801,
      "eval_samples_per_second": 501.328,
      "eval_steps_per_second": 7.888,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7433410286903381,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.5328,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8248,
      "eval_loss": 0.5820872783660889,
      "eval_runtime": 7.3761,
      "eval_samples_per_second": 508.399,
      "eval_steps_per_second": 7.999,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.104405879974365,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.3224,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8376,
      "eval_loss": 0.49828091263771057,
      "eval_runtime": 7.3885,
      "eval_samples_per_second": 507.542,
      "eval_steps_per_second": 7.985,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 4.2900042900042906e-05,
      "loss": 0.2036,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.47393542528152466,
      "eval_runtime": 7.5962,
      "eval_samples_per_second": 493.67,
      "eval_steps_per_second": 7.767,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.36529821157455444,
      "learning_rate": 2.8614328614328617e-05,
      "loss": 0.1222,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.40411490201950073,
      "eval_runtime": 7.3625,
      "eval_samples_per_second": 509.338,
      "eval_steps_per_second": 8.014,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.743577003479004,
      "learning_rate": 1.4328614328614328e-05,
      "loss": 0.0669,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.39391791820526123,
      "eval_runtime": 7.7256,
      "eval_samples_per_second": 485.397,
      "eval_steps_per_second": 7.637,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1872980147600174,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.0368,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8954666666666666,
      "eval_loss": 0.37085509300231934,
      "eval_runtime": 7.7632,
      "eval_samples_per_second": 483.05,
      "eval_steps_per_second": 7.6,
      "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
}