fix: use greedy longest-match for SentencePiece (matches llama.cpp)#6
Merged
fix: use greedy longest-match for SentencePiece (matches llama.cpp)#6
Conversation
Replace Viterbi DP with greedy leftmost-longest match in sentencePieceEncode. This matches llama.cpp's llm_tokenizer_spm behavior and fixes token splitting where sub-token score sums beat whole-token scores (e.g., "▁What" splitting into "Wh"+"at").
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Replace Viterbi DP with greedy longest match. Viterbi picked sub-word splits over whole words because sum(sub-scores) > single-token-score. Greedy prefers longest match, matching llama.cpp behavior.