small refactor

#1
by davanstrien HF staff - opened
BigLAM: BigScience Libraries, Archives and Museums org
No description provided.
BigLAM: BigScience Libraries, Archives and Museums org

Would this cause any issues in generating the splits? I had left the name kwarg in for that reason

BigLAM: BigScience Libraries, Archives and Museums org

Not in this case, the split is set on this line: https://huggingface.co/datasets/biglam/contentious_contexts/blob/main/contentious_contexts.py#L99.

If you had multiple splits you would usually have more things inside the return statement, i.e. name=datasets.Split.VALID, the main reason that split is sometimes passed as a kwarg is that the different splits might be processed differently or contains different downloads so if you, for example, had one file downloaded for the training split and another you for the the validation split it's easier to check for that in _generate_examples. For example, in the GLUE script, the split info is used to determine how some processing steps are done: https://github.com/huggingface/datasets/blob/0c1d099f87a883e52c42d3fd1f1052ad3967e647/datasets/glue/glue.py#L552

BigLAM: BigScience Libraries, Archives and Museums org

Ah, I see. Thanks for the detailed explanation. Merging this

shamikbose89 changed pull request status to merged

Sign up or log in to comment