Skip to content

[v0.13.0][Eagle3]Extend PR #5786 to eagle3 #6443

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
wangxiyuan merged 1 commit into from
Jan 30, 2026
Merged

[v0.13.0][Eagle3]Extend PR #5786 to eagle3 #6443

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
2 changes: 1 addition & 1 deletion vllm_ascend/patch/worker/patch_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -275,7 +275,7 @@ def _update_states(self, scheduler_output: "SchedulerOutput") -> None:
# request state. NOTE: The spec tokens are placeholders and not
# added to token_ids_cpu.
if self.is_kv_consumer and self.speculative_config and \
self.speculative_config.method == "mtp" and self.use_async_scheduling:
self.use_async_scheduling:
req_state = self.requests[request.req_id]
spec_token_ids = scheduler_output.scheduled_spec_decode_tokens.get(
request.req_id, [])
Expand Down
3 changes: 1 addition & 2 deletions vllm_ascend/platform.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -336,8 +336,7 @@ def check_and_update_config(cls, vllm_config: VllmConfig) -> None:
# When open PD separation + MTP + Full Graph + asynchronous scheduling,
# patch gpu_model_runner._update_states for solving
# decode nodes may be in eager mode.
if vllm_config.speculative_config and vllm_config.kv_transfer_config and \
vllm_config.speculative_config.method == "mtp":
if vllm_config.speculative_config and vllm_config.kv_transfer_config:
import vllm_ascend.patch.worker.patch_model_runner # noqa

# Extend original scheduler_config to use SchedulerDynamicBatch.
Expand Down
Loading