trainer_state.json

{
  "best_metric": 0.8821333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0025/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": 9e-05,
      "loss": 1.5454,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8184,
      "eval_loss": 0.657539963722229,
      "eval_runtime": 5.514,
      "eval_samples_per_second": 680.087,
      "eval_steps_per_second": 10.7,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.538280725479126,
      "learning_rate": 9e-05,
      "loss": 0.5122,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8448,
      "eval_loss": 0.515932023525238,
      "eval_runtime": 4.4194,
      "eval_samples_per_second": 848.524,
      "eval_steps_per_second": 13.35,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.3405,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.5010052919387817,
      "eval_runtime": 4.4207,
      "eval_samples_per_second": 848.279,
      "eval_steps_per_second": 13.346,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9339044690132141,
      "learning_rate": 9e-05,
      "loss": 0.2463,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8685333333333334,
      "eval_loss": 0.4415110647678375,
      "eval_runtime": 4.424,
      "eval_samples_per_second": 847.657,
      "eval_steps_per_second": 13.336,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.35819149017334,
      "learning_rate": 9e-05,
      "loss": 0.1787,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8688,
      "eval_loss": 0.4555664658546448,
      "eval_runtime": 4.4668,
      "eval_samples_per_second": 839.529,
      "eval_steps_per_second": 13.209,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4336201548576355,
      "learning_rate": 9e-05,
      "loss": 0.1442,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8773333333333333,
      "eval_loss": 0.4365544617176056,
      "eval_runtime": 4.4703,
      "eval_samples_per_second": 838.877,
      "eval_steps_per_second": 13.198,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.04936341196298599,
      "learning_rate": 9e-05,
      "loss": 0.1181,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8821333333333333,
      "eval_loss": 0.44292840361595154,
      "eval_runtime": 4.7278,
      "eval_samples_per_second": 793.186,
      "eval_steps_per_second": 12.479,
      "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
}