Skip to content

[perf][spec decoding] Skip full-vocab softmax in EAGLE draft when topk == 1 #26235

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
Qiaolin-Yu merged 1 commit into from
May 25, 2026
Merged

[perf][spec decoding] Skip full-vocab softmax in EAGLE draft when topk == 1 #26235

Show file tree
Hide file tree
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
10 changes: 8 additions & 2 deletions python/sglang/srt/speculative/eagle_draft_extend_cuda_graph_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -401,8 +401,14 @@ def run_once():
forward_batch.positions,
forward_batch,
)
probs = torch.softmax(ret.next_token_logits, dim=-1)
ret.topk_p, ret.topk_index = fast_topk(probs, self.topk, dim=-1)
if self.topk == 1:
ret.topk_index = torch.argmax(
ret.next_token_logits, dim=-1, keepdim=True
)
ret.topk_p = torch.ones_like(ret.topk_index, dtype=torch.float32)
else:
probs = torch.softmax(ret.next_token_logits, dim=-1)
ret.topk_p, ret.topk_index = fast_topk(probs, self.topk, dim=-1)

forward_batch.out_cache_loc = output_cache_loc_backup
forward_batch.spec_info.hidden_states = hidden_states_backup
Expand Down
22 changes: 18 additions & 4 deletions python/sglang/srt/speculative/eagle_worker_v2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -479,8 +479,16 @@ def draft_forward(self, forward_batch: ForwardBatch):
forward_batch, skip_attn_backend_init=True
).logits_output
maybe_detect_nan(logits_output.next_token_logits, f"draft_forward step {i}")
probs = torch.softmax(logits_output.next_token_logits, dim=-1)
topk_p, topk_index = fast_topk(probs, self.topk, dim=-1)
if self.topk == 1:
# topk=1 → degenerate single-path tree; `topk_p` is unused
# downstream, so skip softmax and just argmax over logits.
topk_index = torch.argmax(
logits_output.next_token_logits, dim=-1, keepdim=True
)
topk_p = torch.ones_like(topk_index, dtype=torch.float32)
else:
probs = torch.softmax(logits_output.next_token_logits, dim=-1)
topk_p, topk_index = fast_topk(probs, self.topk, dim=-1)
maybe_detect_oob(
topk_index,
0,
Expand Down Expand Up @@ -647,8 +655,14 @@ def _draft_extend_for_decode(
draft_logits_output.hidden_states = draft_logits_output.hidden_states[
select_index
]
probs = torch.softmax(draft_logits_output.next_token_logits, dim=-1)
ret_topk_p, ret_topk_index = fast_topk(probs, self.topk, dim=-1)
if self.topk == 1:
ret_topk_index = torch.argmax(
draft_logits_output.next_token_logits, dim=-1, keepdim=True
)
ret_topk_p = torch.ones_like(ret_topk_index, dtype=torch.float32)
else:
probs = torch.softmax(draft_logits_output.next_token_logits, dim=-1)
ret_topk_p, ret_topk_index = fast_topk(probs, self.topk, dim=-1)
ret_hidden_states = draft_logits_output.hidden_states

# Construct the return values
Expand Down
Loading