trainer_state.json

{
  "best_metric": 0.44053333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0420/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": 0.0005,
      "loss": 3.242,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.21466666666666667,
      "eval_loss": 2.847616195678711,
      "eval_runtime": 5.6394,
      "eval_samples_per_second": 664.968,
      "eval_steps_per_second": 10.462,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.702,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2792,
      "eval_loss": 2.542882204055786,
      "eval_runtime": 5.5278,
      "eval_samples_per_second": 678.395,
      "eval_steps_per_second": 10.673,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.3455,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3602666666666667,
      "eval_loss": 2.2607693672180176,
      "eval_runtime": 5.5041,
      "eval_samples_per_second": 681.305,
      "eval_steps_per_second": 10.719,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 14.644773483276367,
      "learning_rate": 0.0005,
      "loss": 2.0549,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3864,
      "eval_loss": 2.165658712387085,
      "eval_runtime": 5.566,
      "eval_samples_per_second": 673.735,
      "eval_steps_per_second": 10.6,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 12.385472297668457,
      "learning_rate": 0.0005,
      "loss": 1.8115,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.44053333333333333,
      "eval_loss": 1.9493929147720337,
      "eval_runtime": 5.2302,
      "eval_samples_per_second": 716.987,
      "eval_steps_per_second": 11.281,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}