trainer_state.json

{
  "best_metric": 0.4962666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0991/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 0.0005,
      "loss": 3.2417,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2072,
      "eval_loss": 2.9549760818481445,
      "eval_runtime": 12.3716,
      "eval_samples_per_second": 303.113,
      "eval_steps_per_second": 4.769,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.7117,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2866666666666667,
      "eval_loss": 2.5452799797058105,
      "eval_runtime": 12.2239,
      "eval_samples_per_second": 306.777,
      "eval_steps_per_second": 4.827,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.4035,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3416,
      "eval_loss": 2.33097767829895,
      "eval_runtime": 12.1691,
      "eval_samples_per_second": 308.157,
      "eval_steps_per_second": 4.848,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 11.317265510559082,
      "learning_rate": 0.0005,
      "loss": 2.1543,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3874666666666667,
      "eval_loss": 2.142082452774048,
      "eval_runtime": 12.2282,
      "eval_samples_per_second": 306.669,
      "eval_steps_per_second": 4.825,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.27786636352539,
      "learning_rate": 0.0005,
      "loss": 1.9648,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4112,
      "eval_loss": 2.0533969402313232,
      "eval_runtime": 12.2872,
      "eval_samples_per_second": 305.197,
      "eval_steps_per_second": 4.802,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 10.003772735595703,
      "learning_rate": 0.0005,
      "loss": 1.7501,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.45893333333333336,
      "eval_loss": 1.87117338180542,
      "eval_runtime": 12.1357,
      "eval_samples_per_second": 309.006,
      "eval_steps_per_second": 4.862,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 6.45909309387207,
      "learning_rate": 0.0005,
      "loss": 1.5655,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.444,
      "eval_loss": 1.9491734504699707,
      "eval_runtime": 12.4066,
      "eval_samples_per_second": 302.258,
      "eval_steps_per_second": 4.756,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.3715,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4962666666666667,
      "eval_loss": 1.7829171419143677,
      "eval_runtime": 12.1853,
      "eval_samples_per_second": 307.747,
      "eval_steps_per_second": 4.842,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}