Skip to content

Add aiter attention support in prefill-attention-backend of gpt-oss #18282

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
HaiShaw merged 3 commits into from
Mar 2, 2026
Merged

Add aiter attention support in prefill-attention-backend of gpt-oss #18282

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
5efdefd
aiter MHA prefill support sliding window attention with sink
Jan 23, 2026
00afda2
Merge branch 'main' into gpt-oss-with-aiter-prefill-swa
kkHuang-amd Feb 5, 2026
d2fa223
Merge branch 'main' into gpt-oss-with-aiter-prefill-swa
kkHuang-amd Mar 2, 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
7 changes: 7 additions & 0 deletions python/sglang/srt/layers/attention/aiter_backend.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1430,6 +1430,7 @@ def forward_extend(
layer: RadixAttention,
forward_batch: ForwardBatch,
save_kv_cache=True,
sinks=None,
):
cache_loc = (
forward_batch.out_cache_loc
Expand Down Expand Up @@ -1798,6 +1799,10 @@ def forward_extend(
k_cache = k_cache.to(dtype)
v_cache = v_cache.to(dtype)

window_size = (-1, -1)
if layer.sliding_window_size is not None and layer.sliding_window_size > -1:
window_size = (layer.sliding_window_size, -1)

o = mha_batch_prefill_func(
q.contiguous().view(-1, layer.tp_q_head_num, layer.head_dim),
k_cache,
Expand All @@ -1812,6 +1817,8 @@ def forward_extend(
alibi_slopes=None,
return_lse=False,
return_attn_probs=False,
window_size=window_size,
sink_ptr=sinks,
)

return o.view(-1, layer.tp_q_head_num * layer.head_dim)
Expand Down
9 changes: 8 additions & 1 deletion python/sglang/srt/server_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1368,7 +1368,14 @@ def _handle_model_specific_adjustments(self):
else:
self.attention_backend = "triton"

supported_backends = ["triton", "trtllm_mha", "fa3", "fa4", "ascend"]
supported_backends = [
"triton",
"trtllm_mha",
"fa3",
"fa4",
"ascend",
"aiter",
]
prefill_attn_backend, decode_attn_backend = self.get_attention_backends()
assert (
prefill_attn_backend in supported_backends
Expand Down
Loading