fix: eos token ids for Qwen3.8/Kimi-Linear + auto trust_remote_code - #2265
Open
adamteale wants to merge 1 commit into
Open
fix: eos token ids for Qwen3.8/Kimi-Linear + auto trust_remote_code#2265adamteale wants to merge 1 commit into
adamteale wants to merge 1 commit into
Conversation
Three model-loading fixes in utils_mlx.py: 1. Add Qwen3.8 to the eos token branch ([248046, 248044]) — without this, <|im_end|> leaked into generated content on all Qwen3.8 quants. 2. Add kimi-linear to the kimi-k2 eos branch ([163586]) — the auto-resolved eos (163585) was off-by-one, causing stop-token failures. 3. Auto-enable trust_remote_code when a model directory ships modeling_*.py custom code (e.g. kimi_linear). Without this, such models fail to load with "custom code which must be executed" errors.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Three model-loading fixes in utils_mlx.py:
Add Qwen3.8 to the eos token branch ([248046, 248044]) — without this, <|im_end|> leaked into generated content on all Qwen3.8 quants.
Add kimi-linear to the kimi-k2 eos branch ([163586]) — the auto-resolved eos (163585) was off-by-one, causing stop-token failures.
Auto-enable trust_remote_code when a model directory ships modeling_*.py custom code (e.g. kimi_linear). Without this, such models fail to load with "custom code which must be executed" errors.
Motivation
Changes
Why It Works
Test Plan
Manual Testing
Automated Testing