Revert "[1/2] sgl-kernel: Fuse routed scaling factor into select_experts" (#8706)

2025-08-02 20:14:30 +08:00
parent ac6962ccd6
commit f9f0138f80
5 changed files with 12 additions and 38 deletions
--- a/sgl-kernel/include/sgl_kernel_ops.h
+++ b/sgl-kernel/include/sgl_kernel_ops.h
@@ -243,8 +243,7 @@ std::vector<at::Tensor> moe_fused_gate(
    int64_t topk_group,
    int64_t topk,
    int64_t num_fused_shared_experts,
-    double routed_scaling_factor,
-    bool apply_routed_scaling_factor_on_output);
+    double routed_scaling_factor);

 void fp8_blockwise_scaled_grouped_mm(
    torch::Tensor& output,