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 13, 13:15 UTC
Only register rocm_aiter_ops if aiter is found (#28428)
Signed-off-by: mgoin <mgoin64@gmail.com>
Michael Goin committed
Nov 10, 2025 at 19:53 UTC
f2d9ad0620d9aa71481527dcfafdb8357da00470
1 file changed
+2
-1
vllm/_aiter_ops.py
+2
-1
@@ -938,4 +938,5 @@ class rocm_aiter_ops:
938
return tuple(shuffle_weight(tensor, layout=layout) for tensor in tensors)
939
940
941
-rocm_aiter_ops.register_ops_once()
941
+if IS_AITER_FOUND:
942
+ rocm_aiter_ops.register_ops_once()