Skip to content

[CI] fix pre-commit #55630

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
vllm-bot merged 1 commit into from
Sep 7, 2026
Merged

[CI] fix pre-commit #55630

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions csrc/libtorch_stable/quantization/w8a8/cutlass/c3x/scaled_mm_blockwise_sm120_fp8.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,8 +11,8 @@ namespace {
// On a GB10 (24 MiB L2) the blockwise kernel loses most of its throughput once
// the weight is re-streamed from DRAM per row of M tiles: a 16384x2560 FP8
// weight runs at 165 TFLOPS at M=4096 but 90 at M=8192 and 52 at M>=16384;
// 8192x8192 is at 54 from M=6144. With the tile scheduler's max_swizzle_size = 8
// the same launches run at 150-174 TFLOPS at every M, bit-identical to the
// 8192x8192 is at 54 from M=6144. With the tile scheduler's max_swizzle_size =
// 8 the same launches run at 150-174 TFLOPS at every M, bit-identical to the
Comment on lines +14 to +15

@chatgpt-codex-connector chatgpt-codex-connector Bot Sep 7, 2026

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Drop the standalone formatting-only change

This commit only reflows an existing comment and makes no behavioral change, so as a standalone contribution it is exactly the one-off mechanical cleanup that the repository requires contributors not to submit; bundle it with substantive work instead of proceeding with this change alone.

AGENTS.md reference: AGENTS.md:L21-L23

Useful? React with 👍 / 👎.

// default order (ten N/K shapes, M 2048-16384, all cells identical). The one
// place the default order is better is a narrow band around M=4096 on the
// 2560-wide weights (167 vs 153 at 16384x2560, 160 vs 154 at 12288x2560);
Expand Down
Loading