fix(inference): preserve AddedToken metadata for WordPiece pre-splitting#108
Closed
ohdearquant wants to merge 1 commit into
Closed
fix(inference): preserve AddedToken metadata for WordPiece pre-splitting#108ohdearquant wants to merge 1 commit into
ohdearquant wants to merge 1 commit into
Conversation
Parse added_tokens from tokenizer.json and match them in raw text before BERT normalization. Special tokens like [CLS] and [SEP] appearing literally in input text are now recognized as whole tokens instead of being split character-by-character through the normalizer. BGE (bge-small-en-v1.5) parity moves from 8/9 to 9/9. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Owner
Author
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Layer
L1 — tokenizer fix (PR4 of 11)
What
AddedTokenwithcontent,lstrip,rstrip,normalized,single_wordfieldsWhy
Audit gap: special tokens (
[CLS],[SEP]) appearing literally in input text were being split by BERT normalization instead of matched as whole tokens. Pre-fix lattice produced 8/9 PASS on BGE WordPiece audit cases.Result
Stack
Base: #107 (PR3 Qwen BPE)
Umbrella: #104
🤖 Generated with Claude Code