[Bugfix][MRV2] Fix speculative warmup deadlock with pipeline parallelism (#62129)
Signed-off-by: chenjr97 <chenjr97@mail2.sysu.edu.cn> Co-authored-by: Nick Hill <nickhill123@gmail.com> Co-authored-by: vllm-agent <claw@inferact.ai>
This commit is contained in:
co-authored by
Nick Hill
vllm-agent
parent
da9134bc89
commit
7c9042c05a
@@ -88,6 +88,7 @@ def _make_runner(
|
||||
"""Stub model runner exposing only what the warmup entry points read."""
|
||||
return SimpleNamespace(
|
||||
num_speculative_steps=num_spec_steps,
|
||||
speculative_config=SimpleNamespace() if num_spec_steps else None,
|
||||
adaptive_verification=None,
|
||||
rejection_sampler=None,
|
||||
decode_query_len=num_spec_steps + 1,
|
||||
|
||||
Reference in New Issue
Block a user