trainer_state.json

{
  "best_metric": 0.8626666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0395/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": 0.0001,
      "loss": 1.6399,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7912,
      "eval_loss": 0.7538937330245972,
      "eval_runtime": 6.3576,
      "eval_samples_per_second": 589.843,
      "eval_steps_per_second": 9.28,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.5246,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8453333333333334,
      "eval_loss": 0.5109111070632935,
      "eval_runtime": 6.0341,
      "eval_samples_per_second": 621.469,
      "eval_steps_per_second": 9.778,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.3424631655216217,
      "learning_rate": 0.0001,
      "loss": 0.3246,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8624,
      "eval_loss": 0.48568791151046753,
      "eval_runtime": 6.0834,
      "eval_samples_per_second": 616.437,
      "eval_steps_per_second": 9.699,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 14.565607070922852,
      "learning_rate": 0.0001,
      "loss": 0.2149,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.46648430824279785,
      "eval_runtime": 5.9913,
      "eval_samples_per_second": 625.909,
      "eval_steps_per_second": 9.848,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.09621153771877289,
      "learning_rate": 0.0001,
      "loss": 0.1549,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.5491984486579895,
      "eval_runtime": 6.0157,
      "eval_samples_per_second": 623.368,
      "eval_steps_per_second": 9.808,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.20303036272525787,
      "learning_rate": 0.0001,
      "loss": 0.1175,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8496,
      "eval_loss": 0.6006083488464355,
      "eval_runtime": 6.0314,
      "eval_samples_per_second": 621.742,
      "eval_steps_per_second": 9.782,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.048609450459480286,
      "learning_rate": 0.0001,
      "loss": 0.096,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.5183741450309753,
      "eval_runtime": 6.0168,
      "eval_samples_per_second": 623.255,
      "eval_steps_per_second": 9.806,
      "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
}