trainer_state.json

{
  "best_metric": 0.8762666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0951/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": 2.9100759384327822e-05,
      "loss": 2.2165,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7466666666666667,
      "eval_loss": 1.057502269744873,
      "eval_runtime": 5.7595,
      "eval_samples_per_second": 651.093,
      "eval_steps_per_second": 10.244,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 2.6510855361961218e-05,
      "loss": 0.7107,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8306666666666667,
      "eval_loss": 0.6313174962997437,
      "eval_runtime": 6.6018,
      "eval_samples_per_second": 568.025,
      "eval_steps_per_second": 8.937,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.6357440948486328,
      "learning_rate": 2.2527217722455156e-05,
      "loss": 0.3679,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5111252069473267,
      "eval_runtime": 5.8204,
      "eval_samples_per_second": 644.287,
      "eval_steps_per_second": 10.137,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 1.7651164135367127e-05,
      "loss": 0.2047,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8629333333333333,
      "eval_loss": 0.4554196894168854,
      "eval_runtime": 5.7978,
      "eval_samples_per_second": 646.802,
      "eval_steps_per_second": 10.176,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3753751516342163,
      "learning_rate": 1.2441744564412656e-05,
      "loss": 0.1042,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.4423435628414154,
      "eval_runtime": 5.8305,
      "eval_samples_per_second": 643.167,
      "eval_steps_per_second": 10.119,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.953532874584198,
      "learning_rate": 7.5408883548191384e-06,
      "loss": 0.0472,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8693333333333333,
      "eval_loss": 0.4683886170387268,
      "eval_runtime": 5.9516,
      "eval_samples_per_second": 630.079,
      "eval_steps_per_second": 9.913,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.2808518409729004,
      "learning_rate": 3.5397110943979598e-06,
      "loss": 0.0232,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8762666666666666,
      "eval_loss": 0.4666622579097748,
      "eval_runtime": 5.7878,
      "eval_samples_per_second": 647.914,
      "eval_steps_per_second": 10.194,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}