load_dataset function returns an error #3

by idar - opened

load_dataset('sil-ai/bloom-speech', 'es')
returns the following error:
FileNotFoundError: Couldn't find a dataset script at /home/ubuntu/*******/*********/sil-ai/bloom-speech/bloom-speech.py or any data file in the same directory. Couldn't find 'sil-ai/bloom-speech' on the Hugging Face Hub either: FileNotFoundError: Dataset 'sil-ai/bloom-speech' doesn't exist on the Hub. If the repo is private, make sure you are authenticated with use_auth_token=True after logging in with huggingface-cli login.

Because there is a requirement to agree to the terms to use the dataset, you have to also pass use_auth_token=True. I think I need to update the example on the page to reflect that. Thanks for raising the issue!

Also ... if you're in an environment like colab, you'll need to log in to huggingface with something like import huggingface_hub and then huggingface_hub.notebook_login() before attempting to connect. If you're in your own environment, you can just login with huggingface cli before downloading.

jnemecek changed discussion status to closed

Sign up or log in to comment