trainer_state.json

{
  "best_metric": 0.8877333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0056/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": 3e-05,
      "loss": 2.2106,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7653333333333333,
      "eval_loss": 0.9871073365211487,
      "eval_runtime": 6.5586,
      "eval_samples_per_second": 571.769,
      "eval_steps_per_second": 8.996,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.6368,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5573188066482544,
      "eval_runtime": 5.688,
      "eval_samples_per_second": 659.28,
      "eval_steps_per_second": 10.373,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 23.385278701782227,
      "learning_rate": 3e-05,
      "loss": 0.3084,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.44014641642570496,
      "eval_runtime": 5.7096,
      "eval_samples_per_second": 656.788,
      "eval_steps_per_second": 10.333,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2468045949935913,
      "learning_rate": 3e-05,
      "loss": 0.1653,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.4452268183231354,
      "eval_runtime": 5.9057,
      "eval_samples_per_second": 634.982,
      "eval_steps_per_second": 9.99,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6750131249427795,
      "learning_rate": 3e-05,
      "loss": 0.0888,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8685333333333334,
      "eval_loss": 0.4883524179458618,
      "eval_runtime": 5.7027,
      "eval_samples_per_second": 657.588,
      "eval_steps_per_second": 10.346,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6460413932800293,
      "learning_rate": 3e-05,
      "loss": 0.0585,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8877333333333334,
      "eval_loss": 0.4296153783798218,
      "eval_runtime": 6.0842,
      "eval_samples_per_second": 616.352,
      "eval_steps_per_second": 9.697,
      "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
}