trainer_state.json

{
  "best_metric": 0.6085333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0469/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": 3.0312,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3048,
      "eval_loss": 2.541198492050171,
      "eval_runtime": 6.8868,
      "eval_samples_per_second": 544.517,
      "eval_steps_per_second": 8.567,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.2872,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.42693333333333333,
      "eval_loss": 2.0783562660217285,
      "eval_runtime": 6.9085,
      "eval_samples_per_second": 542.81,
      "eval_steps_per_second": 8.54,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.375285148620605,
      "learning_rate": 0.0005,
      "loss": 1.8438,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4744,
      "eval_loss": 1.847942590713501,
      "eval_runtime": 7.0777,
      "eval_samples_per_second": 529.836,
      "eval_steps_per_second": 8.336,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 14.721891403198242,
      "learning_rate": 0.0005,
      "loss": 1.538,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5365333333333333,
      "eval_loss": 1.6561228036880493,
      "eval_runtime": 8.134,
      "eval_samples_per_second": 461.029,
      "eval_steps_per_second": 7.254,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.696285247802734,
      "learning_rate": 0.0005,
      "loss": 1.255,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5608,
      "eval_loss": 1.5435680150985718,
      "eval_runtime": 6.8858,
      "eval_samples_per_second": 544.602,
      "eval_steps_per_second": 8.568,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.0287,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5906666666666667,
      "eval_loss": 1.480968713760376,
      "eval_runtime": 6.9248,
      "eval_samples_per_second": 541.53,
      "eval_steps_per_second": 8.52,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.1058671474456787,
      "learning_rate": 0.0005,
      "loss": 0.8384,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6085333333333334,
      "eval_loss": 1.4309122562408447,
      "eval_runtime": 9.4201,
      "eval_samples_per_second": 398.086,
      "eval_steps_per_second": 6.263,
      "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
}