xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-03-21 04:16:49 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Chore]: qwen3-moe-type-hints-mistake (#19860)

Browse Source 
Co-authored-by: xinnan.hou <hxn02029096@alibaba-inc.com>
This commit is contained in:
Xerxes 2025-06-20 12:43:07 +08:00 committed by GitHub
parent 5aa4a015ce
commit e41bf15cd0
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/model_executor/models/qwen3_moe.py
View File

@ -294,7 +294,7 @@ class Qwen3MoeDecoderLayer(nn.Module):
positions: torch.Tensor,
hidden_states: torch.Tensor,
residual: Optional[torch.Tensor],
) -> torch.Tensor:
) -> tuple[torch.Tensor, torch.Tensor]:
# Self Attention
if residual is None:
residual = hidden_states