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 12, 14:45 UTC
[CI/Test Fix] Fix CP tests on Blackwell (#29338)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Lucas Wilkinson committed
Nov 24, 2025 at 23:55 UTC
2d9ee28cab204b90aa304f60fd7083ea45204bd7
1 file changed
-1
vllm/attention/ops/common.py
-1
@@ -194,7 +194,6 @@ def _cp_lse_common(
194
cp_attn_lse = cp_attn_lse.contiguous()
195
lses = cp_group.all_gather(cp_attn_lse, dim=0).view_as(lses)
196
out, lse = correct_attn_out(cp_attn_out, lses, cp_group.rank_in_group, ctx)
197
- assert out.is_contiguous()
197
return out, lse
198
199