togethercomputer

/

RedPajama-INCITE-7B-Base

like 93

Hi! Could earlier checkpoints (i.e. before 240B tokens) please be added? Thank you!!

I'd like this data to better study scaling properties of models during pretraining.