Skip to content
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

It amazes me that this implementation of BPE doesn't require LOOP. How to make it correct? #20

Open
lsy641 opened this issue Oct 11, 2022 · 0 comments

Comments

@lsy641
Copy link

lsy641 commented Oct 11, 2022

If a pair is the best in frequency, their combination should reduce the count of both single tokens respectively, which may result in the single tokens' difficulty in keeping high ranks in frequency sorting. But your implementation seems to count all 1-gram, 2-gram .. max-gram in one shot. Is it correct?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant