trainer_state.json

{
  "best_metric": 0.8898666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0319/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 6.224557891224557e-05,
      "loss": 2.0439,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7704,
      "eval_loss": 0.8544617295265198,
      "eval_runtime": 6.9891,
      "eval_samples_per_second": 536.553,
      "eval_steps_per_second": 8.442,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1471295356750488,
      "learning_rate": 5.44678011344678e-05,
      "loss": 0.5943,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8349333333333333,
      "eval_loss": 0.5699132084846497,
      "eval_runtime": 6.526,
      "eval_samples_per_second": 574.624,
      "eval_steps_per_second": 9.041,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.557382583618164,
      "learning_rate": 4.669002335669002e-05,
      "loss": 0.3317,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8448,
      "eval_loss": 0.5187342166900635,
      "eval_runtime": 6.2146,
      "eval_samples_per_second": 603.415,
      "eval_steps_per_second": 9.494,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9496538639068604,
      "learning_rate": 3.891224557891224e-05,
      "loss": 0.1962,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8653333333333333,
      "eval_loss": 0.4565514922142029,
      "eval_runtime": 6.1685,
      "eval_samples_per_second": 607.925,
      "eval_steps_per_second": 9.565,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9770507216453552,
      "learning_rate": 3.113446780113447e-05,
      "loss": 0.1182,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.4696522355079651,
      "eval_runtime": 6.1896,
      "eval_samples_per_second": 605.857,
      "eval_steps_per_second": 9.532,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.6338449716567993,
      "learning_rate": 2.3356690023356685e-05,
      "loss": 0.0626,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.4494333863258362,
      "eval_runtime": 6.7413,
      "eval_samples_per_second": 556.276,
      "eval_steps_per_second": 8.752,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.44788604974746704,
      "learning_rate": 1.557891224557891e-05,
      "loss": 0.0324,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.4466918110847473,
      "eval_runtime": 6.9041,
      "eval_samples_per_second": 543.156,
      "eval_steps_per_second": 8.546,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.41444581747055054,
      "learning_rate": 7.801134467801133e-06,
      "loss": 0.0143,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8882666666666666,
      "eval_loss": 0.42573481798171997,
      "eval_runtime": 6.3137,
      "eval_samples_per_second": 593.944,
      "eval_steps_per_second": 9.345,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.0547945611178875,
      "learning_rate": 2.3356690023356688e-08,
      "loss": 0.0084,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8898666666666667,
      "eval_loss": 0.42515188455581665,
      "eval_runtime": 6.4164,
      "eval_samples_per_second": 584.44,
      "eval_steps_per_second": 9.195,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}