Skip to content

[Revert] Fix performance regression for GLM-4.7-GPTQ decode and MTP acceptance rate #33771

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
LucasWilkinson merged 2 commits into from
Feb 8, 2026
Merged

[Revert] Fix performance regression for GLM-4.7-GPTQ decode and MTP acceptance rate #33771

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5397ea8
[Performance] Revert seq_lens.cpu() to seq_lens_cpu for flashinfer.py
aabbccddwasd Feb 5, 2026
c3f54d7
Merge branch 'main' into perf-regression-fix
aabbccddwasd Feb 8, 2026
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: 1 addition & 3 deletions vllm/v1/attention/backends/flashinfer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -919,9 +919,7 @@ def build(
# Guard access to seq_lens_cpu, which may not always be needed
# and can be expensive to retrieve in async mode.
needs_seq_lens_cpu = self.use_dcp or use_cascade or not is_only_trtllm_decode
seq_lens_cpu = (
common_attn_metadata.seq_lens.cpu() if needs_seq_lens_cpu else None
)
seq_lens_cpu = common_attn_metadata.seq_lens_cpu if needs_seq_lens_cpu else None
seq_lens_np = seq_lens_cpu.numpy() if seq_lens_cpu is not None else None
num_blocks_np = (
(seq_lens_np + (page_size - 1)) // page_size
Expand Down