Bug: apply final_hidden_states*=self.routed_scaling_factor at MoE lay… (#8511)

Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com>
2025-08-01 00:07:41 -07:00
parent 7e831efee8
commit c8d3a402c1
1 changed files with 1 additions and 1 deletions
--- a/python/sglang/srt/layers/moe/ep_moe/layer.py
+++ b/python/sglang/srt/layers/moe/ep_moe/layer.py
@@ -304,7 +304,7 @@ class EPMoE(FusedMoE):
            m_max * self.start_expert_id,
            BLOCK_SIZE=512,
        )
-        return output
+        return output * self.routed_scaling_factor


 class DeepEPMoE(EPMoE):