Add support for Hugging Face Datasets

[NielsRogge]

Hi there,

Would be cool to support interoperability with 🤗 datasets since the Dataset you're working with (TEKGEN) is available here: https://huggingface.co/datasets/google-research-datasets/kelm.

Hence it can be easily loaded as follows:

from datasets import load_dataset

dataset = load_dataset("google-research-datasets/kelm", split="test")

This could reduce the burden of having to separately download the dataset.

Just an FYI :)

Kind regards,

Niels