trainer_state.json

{
  "best_metric": 0.8594666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0271/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": 0.0001,
      "loss": 1.8033,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7584,
      "eval_loss": 0.8167320489883423,
      "eval_runtime": 12.4422,
      "eval_samples_per_second": 301.393,
      "eval_steps_per_second": 4.742,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.5591,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8317333333333333,
      "eval_loss": 0.5810436010360718,
      "eval_runtime": 13.0229,
      "eval_samples_per_second": 287.953,
      "eval_steps_per_second": 4.53,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 65.91923522949219,
      "learning_rate": 0.0001,
      "loss": 0.3364,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8349333333333333,
      "eval_loss": 0.551714301109314,
      "eval_runtime": 12.174,
      "eval_samples_per_second": 308.034,
      "eval_steps_per_second": 4.846,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5156259536743164,
      "learning_rate": 0.0001,
      "loss": 0.2185,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8410666666666666,
      "eval_loss": 0.5220444798469543,
      "eval_runtime": 11.9668,
      "eval_samples_per_second": 313.368,
      "eval_steps_per_second": 4.93,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.13729357719421387,
      "learning_rate": 0.0001,
      "loss": 0.1577,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5389065742492676,
      "eval_runtime": 12.4422,
      "eval_samples_per_second": 301.395,
      "eval_steps_per_second": 4.742,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 11.49265193939209,
      "learning_rate": 0.0001,
      "loss": 0.1221,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.5339897871017456,
      "eval_runtime": 12.5233,
      "eval_samples_per_second": 299.441,
      "eval_steps_per_second": 4.711,
      "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
}