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

[BugFix][QWEN-VL]fix wrong apply_rotary_emb_torch selection introduced by #24642 (#26123)

Browse Source 
Signed-off-by: Chendi Xue <Chendi.Xue@intel.com>
Signed-off-by: Chendi.Xue <chendi.xue@intel.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
This commit is contained in:
Chendi.Xue
2025-10-03 10:52:26 -05:00
committed by GitHub
parent 4f8f47e87e
commit dd96465fd7
2 changed files with 10 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
vllm/model_executor/models/qwen2_vl.py
View File

@@ -276,7 +276,8 @@ def apply_rotary_emb_torch(x: torch.Tensor,
def apply_rotary_pos_emb_vision(t: torch.Tensor,
freqs: torch.Tensor) -> torch.Tensor:
rotary_emb_function = dispatch_rotary_emb_function()
rotary_emb_function = dispatch_rotary_emb_function(
default=apply_rotary_emb_torch)
t_ = t.float()
cos = freqs.cos()
sin = freqs.sin()