Can't read public dataset from HuggingFace

[asmith26]

Describe the bug
I'm trying to read the following HuggingFace dataset into Daft: https://huggingface.co/datasets/HuggingFaceM4/DocumentVQA

To Reproduce
Following https://www.getdaft.io/projects/docs/en/stable/user_guide/integrations/huggingface.html#reading-public-datasets

import daft

df = daft.read_parquet("hf://HuggingFaceM4/DocumentVQA")
df

unfortunately yields:

DaftCoreException: DaftError::External Generic http error: Invalid path: hf://HuggingFaceM4/DocumentVQA

Desktop (please complete the following information):

• OS: Linux Mint 22 (based on Ubuntu 24.04)
• Version: getdaft v0.3.1

Additional context
Many thanks for any help, and this amazing lib! :)

[Vince7778]

Hello,

Seems like our docs might be slightly wrong - you may need to add /datasets/ into the URI, like hf://datasets/HuggingFaceM4/DocumentVQA.

[asmith26]

Can confirm this works! Many thanks :)

import daft

df = daft.read_parquet("hf://datasets/HuggingFaceM4/DocumentVQA")