Only 1.2B tokens in the dataset

#2
by ShuoZheLi - opened

The dataset contains around ONLY ~ 1.2b tokens by pythia tokenizer.

Sign up or log in to comment