Fix legacy behaviour #234
Merged
Fix legacy behaviour #234
Conversation
This file contains hidden or 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
There was a case I had not considered 😢 Fixes #96
pcuenca
commented
Sep 11, 2025
There was a problem hiding this comment.
Pull Request Overview
This PR fixes legacy behavior for Llama tokenizers by implementing proper handling of prefix tokens when isLegacy is false. The fix addresses issue #96 related to tokenization differences.
- Adds test case to verify legacy Llama tokenizer behavior
- Modifies property visibility in PreTrainedTokenizer for subclass access
- Implements tokenize override in LlamaPreTrainedTokenizer to handle prefix tokens correctly
Reviewed Changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.
| File | Description |
|---|---|
| Tests/TokenizersTests/TokenizerTests.swift | Adds test case to verify legacy Llama tokenization behavior |
| Sources/Tokenizers/Tokenizer.swift | Changes property visibility and adds tokenize override for proper prefix token handling |
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Sources/Tokenizers/Tokenizer.swift
Outdated
| let tokens = super.tokenize(text: sentencePieceUnderline + text.replacingOccurrences(of: sentencePieceUnderline, with: " ")) | ||
|
|
||
| let second = tokens.dropFirst().first | ||
| if tokens.first == sentencePieceUnderline, second != nil, specialTokens[second!] != nil { |
There was a problem hiding this comment.
Using force unwrap (second!) is risky. Consider using optional binding or guard statement: if tokens.first == sentencePieceUnderline, let secondToken = second, specialTokens[secondToken] != nil {
Suggested change
| if tokens.first == sentencePieceUnderline, second != nil, specialTokens[second!] != nil { | |
| if tokens.first == sentencePieceUnderline, let secondToken = second, specialTokens[secondToken] != nil { |
…nsformers into llama-legacy-fixes
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.
There was a case I had not considered 😢
Fixes #96