Skip to content

[NVIDIA]Fix local_num_experts for EP #8779

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
Aug 5, 2025
Merged

[NVIDIA]Fix local_num_experts for EP #8779

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
3 changes: 2 additions & 1 deletion python/sglang/srt/layers/moe/fused_moe_triton/layer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -200,7 +200,8 @@ def __init__(
self.quant_config = quant_config
self.quant_method.create_weights(
layer=self,
num_experts=self.num_local_experts,
num_experts=self.num_experts,
num_local_experts=self.num_local_experts,
hidden_size=hidden_size,
# FIXME: figure out which intermediate_size to use
intermediate_size=self.intermediate_size_per_partition,
Expand Down
3 changes: 2 additions & 1 deletion python/sglang/srt/layers/quantization/modelopt_quant.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -752,6 +752,7 @@ def create_weights(
self,
layer: torch.nn.Module,
num_experts: int,
Copy link
Copy Markdown
Collaborator

@kaixih kaixih Aug 5, 2025

Choose a reason for hiding this comment

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

Why not simply change the num_experts to num_local_experts? I think we just need the actual num of experts when allocating weights, right?

num_local_experts: int,
hidden_size: int,
intermediate_size_per_partition: int,
params_dtype: torch.dtype,
Expand All @@ -765,7 +766,7 @@ def create_weights(

# TODO(ch-wan): check if this is needed
layer.num_experts = num_experts
layer.num_local_experts = num_experts
layer.num_local_experts = num_local_experts
layer.intermediate_size_per_partition = intermediate_size_per_partition
layer.params_dtype = params_dtype
layer.quant_config = self.quant_config
Expand Down
Loading