huggingface / huggingface/datasets

Easy way to create a Kaggle dataset from a Huggingface dataset?

Open
#5,815 4 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
22k
Forks
3.4k
Avg merge
5d 7h
Merged PRs (30d)
17

Description

I'm not sure whether this is more appropriately addressed with HuggingFace or Kaggle. I would like to somehow directly create a Kaggle dataset from a HuggingFace Dataset.

While Kaggle does provide the option to create a dataset from a URI, that URI must point to a single file. For example:

![image](https://user-images.githubusercontent.com/5355286/235792394-7c559d07-4aff-45b7-ad2b-9c5280c88415.png)

Is there some mechanism from huggingface to represent a dataset (such as that from `load_dataset('wmt14', 'de-en', split='train')` as a single file? Or, some other way to get that into a Kaggle dataset so that I can use the huggingface `datasets` module to process and consume it inside of a Kaggle notebook?

Thanks in advance!

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.