trainer_state.json

{
  "best_metric": 0.8992,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0197/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": 6.655434848731736e-05,
      "loss": 1.6028,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7957333333333333,
      "eval_loss": 0.7549870014190674,
      "eval_runtime": 7.5137,
      "eval_samples_per_second": 499.088,
      "eval_steps_per_second": 7.852,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.32313346862793,
      "learning_rate": 5.685900305529986e-05,
      "loss": 0.5424,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5393970012664795,
      "eval_runtime": 7.3474,
      "eval_samples_per_second": 510.385,
      "eval_steps_per_second": 8.03,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.3035201132297516,
      "learning_rate": 4.28342139851007e-05,
      "loss": 0.3223,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.46326297521591187,
      "eval_runtime": 7.3692,
      "eval_samples_per_second": 508.876,
      "eval_steps_per_second": 8.006,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.14131274819374084,
      "learning_rate": 2.7257762754823102e-05,
      "loss": 0.1914,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.42293351888656616,
      "eval_runtime": 7.3942,
      "eval_samples_per_second": 507.151,
      "eval_steps_per_second": 7.979,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.3139183521270752,
      "learning_rate": 1.3214756563261342e-05,
      "loss": 0.1061,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.39409881830215454,
      "eval_runtime": 7.5894,
      "eval_samples_per_second": 494.109,
      "eval_steps_per_second": 7.774,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.0620877742767334,
      "learning_rate": 3.4865850128228307e-06,
      "loss": 0.057,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.36947229504585266,
      "eval_runtime": 7.2918,
      "eval_samples_per_second": 514.279,
      "eval_steps_per_second": 8.091,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1292727142572403,
      "learning_rate": 3.178726637742546e-11,
      "loss": 0.0392,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8992,
      "eval_loss": 0.3686140477657318,
      "eval_runtime": 8.95,
      "eval_samples_per_second": 418.997,
      "eval_steps_per_second": 6.592,
      "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
}