trainer_state.json

{
  "best_metric": 0.8722666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0164/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "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": 0.0001,
      "loss": 1.4396,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.82,
      "eval_loss": 0.6539126038551331,
      "eval_runtime": 5.991,
      "eval_samples_per_second": 625.934,
      "eval_steps_per_second": 9.848,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.277584552764893,
      "learning_rate": 0.0001,
      "loss": 0.4661,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8352,
      "eval_loss": 0.5388305187225342,
      "eval_runtime": 6.2017,
      "eval_samples_per_second": 604.672,
      "eval_steps_per_second": 9.514,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.2928,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.4703882932662964,
      "eval_runtime": 5.3582,
      "eval_samples_per_second": 699.858,
      "eval_steps_per_second": 11.011,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.771178245544434,
      "learning_rate": 0.0001,
      "loss": 0.2058,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.5026739239692688,
      "eval_runtime": 5.3223,
      "eval_samples_per_second": 704.582,
      "eval_steps_per_second": 11.085,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.378175735473633,
      "learning_rate": 0.0001,
      "loss": 0.1479,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8722666666666666,
      "eval_loss": 0.4683147668838501,
      "eval_runtime": 5.3861,
      "eval_samples_per_second": 696.233,
      "eval_steps_per_second": 10.954,
      "step": 1665
    }
  ],
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}