trainer_state.json

{
  "best_metric": 0.8885333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0674/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": 4.7538820348083834e-05,
      "loss": 1.8564,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7978666666666666,
      "eval_loss": 0.8038123250007629,
      "eval_runtime": 5.7769,
      "eval_samples_per_second": 649.132,
      "eval_steps_per_second": 10.213,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 4.063987381465584e-05,
      "loss": 0.5392,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5619820356369019,
      "eval_runtime": 5.896,
      "eval_samples_per_second": 636.02,
      "eval_steps_per_second": 10.007,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.6167054176330566,
      "learning_rate": 3.0628700751103485e-05,
      "loss": 0.2552,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.47739341855049133,
      "eval_runtime": 5.7365,
      "eval_samples_per_second": 653.711,
      "eval_steps_per_second": 10.285,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.856703758239746,
      "learning_rate": 1.950269447231056e-05,
      "loss": 0.1202,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.42670589685440063,
      "eval_runtime": 6.267,
      "eval_samples_per_second": 598.373,
      "eval_steps_per_second": 9.414,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4040428400039673,
      "learning_rate": 9.465496973304394e-06,
      "loss": 0.0539,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8848,
      "eval_loss": 0.42523840069770813,
      "eval_runtime": 5.8744,
      "eval_samples_per_second": 638.366,
      "eval_steps_per_second": 10.044,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3693388104438782,
      "learning_rate": 2.5050983169121382e-06,
      "loss": 0.0223,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8874666666666666,
      "eval_loss": 0.4197372794151306,
      "eval_runtime": 5.7636,
      "eval_samples_per_second": 650.632,
      "eval_steps_per_second": 10.237,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3791907727718353,
      "learning_rate": 9.082071983079754e-11,
      "loss": 0.0146,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8885333333333333,
      "eval_loss": 0.418268084526062,
      "eval_runtime": 5.7775,
      "eval_samples_per_second": 649.067,
      "eval_steps_per_second": 10.212,
      "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
}