Add bytelevel normalizer to fix decode when adding tokens to BPE #1555
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
ArthurZucker
commented
Jun 28, 2024
tokenizers/src/tokenizer/mod.rs
Outdated
Comment on lines
850
to
863
| let tokens = ids | ||
| .iter() | ||
| .filter_map(|id| { | ||
| self.added_vocabulary | ||
| .id_to_token(*id, &self.model) | ||
| .filter(|token| { | ||
| !skip_special_tokens || !self.added_vocabulary.is_special_token(token) | ||
| }) | ||
| }) | ||
| .collect::<Vec<_>>(); | ||
|
|
||
| if let Some(decoder) = &self.decoder { | ||
| decoder.decode(tokens) |
There was a problem hiding this comment.
reverted to what we originally had
|
The test passes locally ! |
McPatate
reviewed
Jul 11, 2024
Comment on lines
+33
to
+34
| /// Strip the normalized string inplace | ||
| fn normalize(&self, normalized: &mut NormalizedString) -> Result<()> { |
There was a problem hiding this comment.
Why doesn't this live in NormalizedString like so:
impl NormalizedString {
fn normalize_byte_level(&mut self) -> Result<()> { // ... }
}?
Feels a bit weird to have an empty stateless struct for just a function, but may be due to the structure of tokenizers / python.
add it fix
Co-authored-by: Luc Georges <McPatate@users.noreply.github.com>
ArthurZucker
changed the title
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Outdated suggestions cannot be applied.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.
This revert the previous breaking change.
Also add a new
ByteLevelnormalizer, which replaces the ByteLevel pre_tokenizer.Checked that we can add chines / Cyrillic tokens which are properly encoded and decoder.
Fixes #1392