trainer_state.json

{
  "best_metric": 0.8573333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0001/checkpoints/checkpoint-1332",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1332,
  "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": 2.7151869176294504e-05,
      "loss": 2.4701,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6773333333333333,
      "eval_loss": 1.3558939695358276,
      "eval_runtime": 12.2324,
      "eval_samples_per_second": 306.562,
      "eval_steps_per_second": 4.823,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 45.40098190307617,
      "learning_rate": 1.9662164055386567e-05,
      "loss": 0.9045,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8074666666666667,
      "eval_loss": 0.7500902414321899,
      "eval_runtime": 12.5613,
      "eval_samples_per_second": 298.536,
      "eval_steps_per_second": 4.697,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2521793842315674,
      "learning_rate": 1.0391670726449015e-05,
      "loss": 0.5125,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8368,
      "eval_loss": 0.5874084830284119,
      "eval_runtime": 12.4405,
      "eval_samples_per_second": 301.435,
      "eval_steps_per_second": 4.743,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.738951206207275,
      "learning_rate": 2.881402548656824e-06,
      "loss": 0.3493,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.5122963190078735,
      "eval_runtime": 12.2861,
      "eval_samples_per_second": 305.224,
      "eval_steps_per_second": 4.802,
      "step": 1332
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.58965282490368e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}