trainer_state.json

{
  "best_metric": 0.5373333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0593/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": 0.0005,
      "loss": 3.1493,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2432,
      "eval_loss": 2.7911136150360107,
      "eval_runtime": 6.2676,
      "eval_samples_per_second": 598.315,
      "eval_steps_per_second": 9.413,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.886582374572754,
      "learning_rate": 0.0005,
      "loss": 2.5333,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3389333333333333,
      "eval_loss": 2.3651070594787598,
      "eval_runtime": 6.2661,
      "eval_samples_per_second": 598.462,
      "eval_steps_per_second": 9.416,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.1345,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4128,
      "eval_loss": 2.0311005115509033,
      "eval_runtime": 7.1059,
      "eval_samples_per_second": 527.728,
      "eval_steps_per_second": 8.303,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.083292961120605,
      "learning_rate": 0.0005,
      "loss": 1.7746,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4624,
      "eval_loss": 1.891615867614746,
      "eval_runtime": 6.7493,
      "eval_samples_per_second": 555.614,
      "eval_steps_per_second": 8.742,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.012608528137207,
      "learning_rate": 0.0005,
      "loss": 1.4699,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.49333333333333335,
      "eval_loss": 1.7610983848571777,
      "eval_runtime": 6.7655,
      "eval_samples_per_second": 554.283,
      "eval_steps_per_second": 8.721,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.1882,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5162666666666667,
      "eval_loss": 1.713046908378601,
      "eval_runtime": 6.1649,
      "eval_samples_per_second": 608.278,
      "eval_steps_per_second": 9.57,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.1461029052734375,
      "learning_rate": 0.0005,
      "loss": 0.9039,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5373333333333333,
      "eval_loss": 1.7244958877563477,
      "eval_runtime": 6.1421,
      "eval_samples_per_second": 610.54,
      "eval_steps_per_second": 9.606,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}