Closed helfardy closed 2 years ago
Sorry, I'm still refactoring the code. Will share the ckpts using the new code soon.
Sorry, I'm still refactoring the code. Will share the ckpts using the new code soon.
So,we can not run this project ?
You can, but you have to process raw text data to tensor files. New pipeline allows you to load from the raw data directly and tokenize it on the fly.
data.zip doesn't have the models/checkpoints (as specified in the readme)