Skip to content

[ModelOpt] Fix broken Qwen3-235B-A22B-Instruct-2507-NVFP4 launch #18189

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
b8zhong merged 1 commit into from
Feb 8, 2026
+8 −0
Merged

[ModelOpt] Fix broken Qwen3-235B-A22B-Instruct-2507-NVFP4 launch #18189

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions python/sglang/srt/models/qwen3_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -890,6 +890,14 @@ def __init__(
class Qwen3MoeForCausalLM(nn.Module):
fall_back_to_pt_during_load = False

# Mapping from fused module names to their component weight names.
# Required for quantization configs (e.g., ModelOpt FP4) to correctly identify
# which layers should be skipped based on the exclude_modules/ignore list.
packed_modules_mapping = {
Copy link
Copy Markdown
Contributor

@ssshinigami ssshinigami Feb 4, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same question as for this PR #18224

Copy link
Copy Markdown
Contributor Author

@vincentzed vincentzed Feb 4, 2026 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's because qkv and o proj are NVFP4 in these recipes, which is not the case for the nv chkpt

Nv does not show recipe for NVFP4 model.

"qkv_proj": ["q_proj", "k_proj", "v_proj"],
"gate_up_proj": ["gate_proj", "up_proj"],
}
Comment thread
vincentzed marked this conversation as resolved.

def __init__(
self,
config: Qwen3MoeConfig,
Expand Down
Loading