Skip to content

[Kimi-K3] Allow DSPARK verify on cutedsl_mla (fold_sq) #33650

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
hnyls2002 merged 2 commits into from
Aug 6, 2026
Merged

[Kimi-K3] Allow DSPARK verify on cutedsl_mla (fold_sq) #33650

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
17d5c6a
[Kimi-K3] Allow DSPARK q_len<=8 verify on cutedsl_mla
yhyang201 Aug 5, 2026
d22384e
[Kimi-K3] cutedsl_mla serves any DSPARK verify width (drop q_len cap)
yhyang201 Aug 5, 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
6 changes: 4 additions & 2 deletions python/sglang/srt/arg_groups/overrides.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -327,8 +327,10 @@ def _dspark_verify_on_decode_backend(
if backend == "tokenspeed_mla":
return kv_cache_dtype == "fp8_e4m3" and q_len <= 8
if backend == "cutedsl_mla":
# The cute-dsl kernel rejects q_len >= 5 with no fallback.
return q_len <= 4
# cute-dsl monolithic MLA decode folds the verify tokens into the head
# dim (fold_sq), so it serves any DSPARK verify width. Needs flashinfer
# >= 0.6.15 (older builds reject q_len >= 5).
return True
return False


Expand Down
Loading