Datasets:

Modalities:
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
Dask
License:

What is the size of each subset?

#7
by agdhruv - opened

How many rows and tokens exist in each subset of the v2 dataset? The dataset viewer says ~806M rows in the High-Quality subset, but reading the parquet footers gives a higher number. It would be great to get these numbers.

Sign up or log in to comment