A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[Bugfix] Fix Qwen25VL packed_modules_mapping (#23604)
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com> Co-authored-by: Michael Goin <mgoin64@gmail.com>
Jee Jee Li committed
Aug 26, 2025 at 16:09 UTC
9b5f64238fbd0f98928587b3426cbf69eea96ae7
1 file changed
+1
vllm/model_executor/models/qwen2_5_vl.py
+1
@@ -853,6 +853,7 @@ class Qwen2_5_VLForConditionalGeneration(nn.Module, SupportsMultiModal,
853
SupportsQuant):
854
855
packed_modules_mapping = {
856
+ "qkv_proj": ["q_proj", "k_proj", "v_proj"],
857
"gate_up_proj": ["gate_proj", "up_proj"],
858
}
859