trainer_state.json

{
  "best_metric": 0.9029333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0013/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": 1.0357416868209839,
      "learning_rate": 7.714285714285714e-05,
      "loss": 1.5571,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7837333333333333,
      "eval_loss": 0.768093466758728,
      "eval_runtime": 11.9247,
      "eval_samples_per_second": 314.473,
      "eval_steps_per_second": 4.948,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.321319580078125,
      "learning_rate": 6.428571428571429e-05,
      "loss": 0.4987,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8381333333333333,
      "eval_loss": 0.5606697797775269,
      "eval_runtime": 12.2254,
      "eval_samples_per_second": 306.738,
      "eval_steps_per_second": 4.826,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.014869689941406,
      "learning_rate": 5.142857142857143e-05,
      "loss": 0.2744,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.4812227785587311,
      "eval_runtime": 11.9786,
      "eval_samples_per_second": 313.058,
      "eval_steps_per_second": 4.925,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5090433359146118,
      "learning_rate": 3.857142857142857e-05,
      "loss": 0.1518,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.41656291484832764,
      "eval_runtime": 12.2442,
      "eval_samples_per_second": 306.268,
      "eval_steps_per_second": 4.819,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2740810811519623,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 0.0776,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.4299846887588501,
      "eval_runtime": 11.9588,
      "eval_samples_per_second": 313.576,
      "eval_steps_per_second": 4.934,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.07098124921321869,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 0.0351,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8973333333333333,
      "eval_loss": 0.3948049545288086,
      "eval_runtime": 12.3212,
      "eval_samples_per_second": 304.354,
      "eval_steps_per_second": 4.788,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.861003861003861e-08,
      "loss": 0.0146,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9029333333333334,
      "eval_loss": 0.38785508275032043,
      "eval_runtime": 12.2048,
      "eval_samples_per_second": 307.256,
      "eval_steps_per_second": 4.834,
      "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
}