Skip to content

[BUGFIX] Fix crash in Eagle Speculative Decoding models when exceedin… #24662

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
merged 3 commits into from
Sep 25, 2025
+16 −2
Merged

[BUGFIX] Fix crash in Eagle Speculative Decoding models when exceedin… #24662

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
88617ef
[BUGFIX] Fix crash in Eagle Speculative Decoding models when exceedin…
AlonKejzman Sep 11, 2025
515a675
Merge branch 'main' into main
AlonKejzman Sep 25, 2025
93ca8fd
Merge branch 'main' into main
AlonKejzman Sep 25, 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
18 changes: 16 additions & 2 deletions vllm/v1/worker/gpu_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2310,7 +2310,20 @@ def propose_draft_token_ids(sampled_token_ids):
use_padded_batch_for_eagle = self.speculative_config and \
self.speculative_config.use_eagle() and \
not self.speculative_config.disable_padded_drafter_batch
if use_padded_batch_for_eagle:
effective_drafter_max_model_len = self.max_model_len
if effective_drafter_max_model_len is None:
effective_drafter_max_model_len = self.model_config.max_model_len
if (self.speculative_config
and self.speculative_config.draft_model_config is not None
and self.speculative_config.draft_model_config.max_model_len
is not None):
effective_drafter_max_model_len = (
self.speculative_config.draft_model_config.max_model_len)
input_fits_in_drafter = spec_decode_common_attn_metadata and (
spec_decode_common_attn_metadata.seq_lens.max() +
self.speculative_config.num_speculative_tokens
<= effective_drafter_max_model_len)
if use_padded_batch_for_eagle and input_fits_in_drafter:
# EAGLE speculative decoding can use the GPU sampled tokens
# as inputs, and does not need to wait for bookkeeping to finish.
propose_draft_token_ids(sampler_output.sampled_token_ids)
Expand All @@ -2328,7 +2341,8 @@ def propose_draft_token_ids(sampled_token_ids):
logits, hidden_states,
num_scheduled_tokens)

if self.speculative_config and not use_padded_batch_for_eagle:
if (self.speculative_config and not use_padded_batch_for_eagle
and input_fits_in_drafter):
# ngram and other speculative decoding methods use the sampled
# tokens on the CPU, so they are run after bookkeeping.
propose_draft_token_ids(valid_sampled_token_ids)
Expand Down