Datasets:

ArXiv:
License:

Getting around ~2T tokens

#13
by adityakallappaolakrutrim - opened

Hello,

I have downloaded the dataset and after tokenization, I am getting around 2T tokens. Is that the complete dataset? Because the README says 4T tokens. Please help.

Thanks

which tokenizer is that with?

This was with a different tokenizer than the one mentioned in the paper. There were multiple files that were not downloaded which resulted in this. The issue is resolved.
Thanks.

adityakallappaolakrutrim changed discussion status to closed

Sign up or log in to comment