-
Notifications
You must be signed in to change notification settings - Fork 1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Integrate stateful dataset in tinyllama.py #797
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Did you consider backward-compatibility with existing tinyllama checkpoints?
We would have to leave the for loop we have here and run it conditionally based on whether the checkpoint has a dataloader state saved or not. I don't think this is worth it. |
Replaces the fast-forward resumption of streaming dataset in tinyllama.py with the state-dict approach for faster reloading of the dataset state.
Requires fixing of