biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Models][Qwen3VL] Optimise _validate_and_reshape_mm_tensor (#26426)

Browse Source 
Signed-off-by: Lukas Geiger <lukas.geiger94@gmail.com>
This commit is contained in:
Lukas Geiger
2025-10-09 11:25:48 +01:00
committed by GitHub
parent 4bdf7ac593
commit 0426e3c5e1
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/model_executor/models/qwen3_vl.py
View File

@@ -1304,7 +1304,7 @@ class Qwen3VLForConditionalGeneration(
f"Got ndim: {mm_input.ndim} " f"Got ndim: {mm_input.ndim} "
f"(shape={mm_input.shape})" f"(shape={mm_input.shape})"
) )
return torch.concat(list(mm_input)) return mm_input.reshape(-1, mm_input.shape[-1])
else: else:
return torch.concat(mm_input) return torch.concat(mm_input)