trainer_state.json

{
  "best_metric": 0.9021333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0952/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": Infinity,
      "learning_rate": 4.2878592878592884e-05,
      "loss": 1.8321,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8381333333333333,
      "eval_loss": 0.7437000274658203,
      "eval_runtime": 7.0335,
      "eval_samples_per_second": 533.163,
      "eval_steps_per_second": 8.388,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 18.446109771728516,
      "learning_rate": 3.573573573573574e-05,
      "loss": 0.4872,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.511381208896637,
      "eval_runtime": 6.8307,
      "eval_samples_per_second": 548.992,
      "eval_steps_per_second": 8.637,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.660834312438965,
      "learning_rate": 2.8592878592878596e-05,
      "loss": 0.2248,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.876,
      "eval_loss": 0.43698668479919434,
      "eval_runtime": 6.3809,
      "eval_samples_per_second": 587.691,
      "eval_steps_per_second": 9.246,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5049573183059692,
      "learning_rate": 2.1450021450021453e-05,
      "loss": 0.1,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.888,
      "eval_loss": 0.3872281312942505,
      "eval_runtime": 6.3727,
      "eval_samples_per_second": 588.443,
      "eval_steps_per_second": 9.258,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8028401136398315,
      "learning_rate": 1.4307164307164309e-05,
      "loss": 0.0433,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8917333333333334,
      "eval_loss": 0.4058274030685425,
      "eval_runtime": 7.1826,
      "eval_samples_per_second": 522.096,
      "eval_steps_per_second": 8.214,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5357438921928406,
      "learning_rate": 7.164307164307164e-06,
      "loss": 0.017,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8973333333333333,
      "eval_loss": 0.3828280568122864,
      "eval_runtime": 6.3085,
      "eval_samples_per_second": 594.434,
      "eval_steps_per_second": 9.352,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3225955665111542,
      "learning_rate": 2.1450021450021452e-08,
      "loss": 0.0079,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9021333333333333,
      "eval_loss": 0.3703397810459137,
      "eval_runtime": 6.2936,
      "eval_samples_per_second": 595.84,
      "eval_steps_per_second": 9.375,
      "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
}