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 7, 17:45 UTC
[Bugfix] Fix AttributeError in SMControlContextManager (#35338)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Lucas Wilkinson committed
Feb 25, 2026 at 21:01 UTC
9511a3f8eec6992d8834ad85af855683bd74ba29
1 file changed
+1
-1
vllm/v1/worker/gpu_ubatch_wrapper.py
+1
-1
@@ -74,7 +74,7 @@ class SMControlContextManager:
74
"SM control is currently only supported on CUDA"
75
)
76
77
- total_sms = num_compute_units(torch.cuda.current_device().index)
77
+ total_sms = num_compute_units(torch.cuda.current_device())
78
79
assert comm_sms < total_sms
80
self.total_sms = total_sms