Skip to content

[Perf][Qwen4Exp] Route the PLE shard copy through copy_weight_ #65692

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
gau-nernst merged 2 commits into from
Nov 5, 2026
+2 −1
Merged

[Perf][Qwen4Exp] Route the PLE shard copy through copy_weight_ #65692

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6605fa9
[Perf][Qwen4Exp] Route the PLE shard copy through copy_weight_
JimmyToluene Nov 3, 2026
197bdf0
Merge branch 'main' into qwen4exp-ple-copy-weight
vllm-agent Nov 5, 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
3 changes: 2 additions & 1 deletion vllm/models/qwen4_exp/common/ple.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@
from vllm.model_executor.layers.vocab_parallel_embedding import (
VocabParallelEmbedding,
)
from vllm.model_executor.utils import copy_weight_


@dataclass(frozen=True)
Expand Down Expand Up @@ -75,7 +76,7 @@ def _copy_embedding_shard_(
source = loaded_weight.narrow(0, overlap.source_start, overlap.row_count)
target = destination.narrow(0, overlap.destination_start, overlap.row_count)
with torch.no_grad():
target.copy_(source.to(device=target.device, dtype=target.dtype))
copy_weight_(target, source)
return overlap.row_count

def weight_loader(
Expand Down
Loading