huggingface / huggingface/datasets
Reload dataset that was already downloaded with `load_from_disk` from cloud storage
Open
enhancement
- Dominant language
- Python
- Stars
- 22k
- Forks
- 3.4k
- Avg merge
- 5d 7h
- Merged PRs (30d)
- 17
Description
`load_from_disk` downloads the dataset to a temporary directory without checking if the dataset has already been downloaded once.
It would be nice to have some sort of caching for datasets downloaded this way. This could leverage the fingerprint of the dataset that was saved in the `state.json` file.
Contributor guide
Assessment
This issue has not been assessed yet.