Skip to content

fix bailing_moe with enable_dp_attention #10860

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
zhyncs merged 1 commit into from
Sep 24, 2025
Merged

fix bailing_moe with enable_dp_attention #10860

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
4 changes: 2 additions & 2 deletions python/sglang/srt/models/bailing_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,12 +45,12 @@
get_attention_dp_size,
get_attention_tp_rank,
get_attention_tp_size,
is_dp_attention_enabled,
)
from sglang.srt.layers.layernorm import RMSNorm
from sglang.srt.layers.linear import (
MergedColumnParallelLinear,
QKVParallelLinear,
ReplicatedLinear,
RowParallelLinear,
)
from sglang.srt.layers.logits_processor import LogitsProcessor
Expand Down Expand Up @@ -702,7 +702,7 @@ def __init__(
self.embed_dim,
quant_config=quant_config,
prefix=add_prefix("word_embeddings", prefix),
use_attn_tp_group=global_server_args_dict["enable_dp_lm_head"],
enable_tp=not is_dp_attention_enabled(),
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist bot Sep 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

This change disables tensor parallelism for the word_embeddings layer when DP attention is enabled. While this correctly fixes the hang, be aware that it might have a minor performance impact as the embedding layer will be replicated across all TP ranks instead of being sharded.

)
else:
self.word_embeddings = PPMissingLayer()
Expand Down
Loading