Fix repeated characters handling in BPE tokenization (e.g., 'RR' in 'Strawberry') #336
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…Strawberry') - Resolved an issue in `_byte_pair_merge` where repeated characters like 'RR' in 'Strawberry' were not correctly handled. - Updated the logic in `_byte_pair_merge` to ensure missing pairs are logged and handled properly. - Added a test case `test_repeated_r_in_strawberry` to verify correct tokenization.
There was a problem hiding this comment.
nstead of making the ranks.get call twice (once for unwrap_or and again for unwrap_or_else), it would be better to store the result in a variable and then use it.
let rank = ranks.get(&piece[i..i + 2]).unwrap_or_else(|| {
eprintln!("Missing pair: {:?}", &piece[i..i + 2]);
&Rank::MAX
});
let rank = *rank;
|
Thanks, but there isn't a bug here |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fix repeated characters handling in BPE tokenization (e.g., 'RR' in 'Strawberry')
Resolved an issue in
_byte_pair_mergewhere repeated characters like 'RR' in 'Strawberry' were incorrectly handled.Updated the logic in
_byte_pair_mergeto ensure missing pairs are logged and handled properly.Added a test case
test_repeated_r_in_strawberryto verify correct tokenization.