fix Qwen3.5 BF16 packed_modules_mapping#1
Open
ningding01 wants to merge 9 commits into
Open
Conversation
…ight shape check failed
…ATOMAttnBackendForSgl
wanzhenchn
force-pushed
the
feat/qwen3.5-sgl-plugin-final
branch
8 times, most recently
from
ddf304b to
2f5f006
Compare
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Fix a BF16-specific weight loading issue in the Qwen3.5 SGLang plugin.
The plugin did not adapt
packed_modules_mappingfor the BF16 fused GDN layoutand instead reused the default split mapping used by non-BF16 / FP8 paths. As a
result, BF16 checkpoint shards could be routed to the wrong parameters, leaving
some GDN input projection weights unloaded and causing garbled generation output.