Load CelebA-HQ in diffusers_training_example.ipynb

Dear Hugging Face friends:

I was trying to train diffusers on the CelebA-HQ dataset in training_example.ipynb. I used

from datasets import load_dataset
config.dataset_name = "huggan/CelebA-HQ"
dataset = load_dataset(config.dataset_name, split="train")

as suggested in the notebook:

But upon running I got error ArrowInvalid: Parquet magic bytes not found in footer.

Do you know what is going on and how to train on the CelebA-HQ dataset in this notebook? Thank you very much!

Not very sure but this may be what I was looking for:

config.dataset_name = "huggan/CelebA-faces"
dataset = load_dataset(config.dataset_name, split="train")