Skip to content

[fix] Handle transposed w13_weight #357

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
RuixuanZhang06 merged 1 commit into from
Feb 26, 2026
Merged

[fix] Handle transposed w13_weight #357

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/sgl_kernel_npu/sgl_kernel_npu/activation/swiglu_oai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,7 +83,7 @@ def swiglu_oai_triton(


def swiglu_oai_native(layer, hidden_states):
E, N, _ = layer.w13_weight.size()
E, _, N = layer.w13_weight.size()
gate_up = hidden_states.view(-1, N)
alpha = layer.moe_runner_config.gemm1_alpha
limit = layer.moe_runner_config.gemm1_clamp_limit
Expand All @@ -98,7 +98,7 @@ def swiglu_oai_native(layer, hidden_states):
def swiglu_oai(layer, hidden_states):
return swiglu_oai_triton(
hidden_states,
layer.w13_weight.shape[1],
layer.w13_weight.shape[2],
layer.moe_runner_config.gemm1_alpha,
layer.moe_runner_config.gemm1_clamp_limit,
)