Skip to content

[Perf] Support Flashinfer RoPE+Quant+KV update kernel for trtllm_mha backend for GPT-OSS #15729

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

Open
elvischenv wants to merge 3 commits into
base: main
Choose a base branch
from
+298 −104
Open

[Perf] Support Flashinfer RoPE+Quant+KV update kernel for trtllm_mha backend for GPT-OSS #15729

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
8e942ab
clean up trtllm mha backend and use NHD kv layout
elvischenv Dec 18, 2025
91e1a1b
support Flashinfer rope_quantize_fp8 + append_paged_kv_cache for GPT-OSS
elvischenv Dec 19, 2025
e59267f
support Flashinfer rope_quantize_fp8_append_paged_kv_cache
elvischenv Dec 22, 2025
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: 4 additions & 0 deletions python/sglang/srt/layers/attention/base_attn_backend.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -167,3 +167,7 @@ def get_indexer_metadata(
) -> Optional[BaseIndexerMetadata]:
"""Get the indexer metadata. None means don't support indexer."""
return None

def support_rope_fusion(self) -> bool:
"""Check if the current backend supports RoPE fusion."""
Copy link
Collaborator

@Fridge003 Fridge003 Mar 20, 2026

Choose a reason for hiding this comment

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

Instead of adding this method in base class, can we control this fusion with an environ flag?
Now it is set to False by default. After this feature stabilizes it can be turned on by default

return False
Loading
Loading