Two corrupted words

#2
by gbenson - opened

Hiya, I've been using this dataset (thanks!) but I noticed two of the words in "sorted_by_frequency" are broken:

>>> from datasets import load_dataset
>>> ds = load_dataset("Maximax67/English-Valid-Words", "sorted_by_frequency", split="train")
>>> [row["Rank"] for row in ds if not row["Word"]]
[8932, 16351]

They're null and nan:

$ grep '^\(8932\|16351\),' valid_words_sorted_by_frequency.csv 
8932,null,11628795,nul,0.0193
16351,nan,4460845,,

Probably something in the Hugging Face import code translated them.

The other 172,780 words seem fine :)

Sign up or log in to comment