[Flashinfer Autotune] Fix FlashInfer FP4 MoE autotuning crash by removing incorrect flatten on hidden_states_scale (#18500)

2026-02-11 21:31:27 -08:00
parent 20554a0a4f
commit 454676811e
1 changed files with 1 additions and 1 deletions
--- a/python/sglang/srt/layers/moe/fused_moe_triton/layer.py
+++ b/python/sglang/srt/layers/moe/fused_moe_triton/layer.py
@@ -1329,7 +1329,7 @@ class FlashInferFP4MoE(FusedMoE):
            routing_logits=router_logits,
            routing_bias=correction_bias,
            hidden_states=hs_fp4,
-            hidden_states_scale=hs_scale_linear.view(torch.float8_e4m3fn).flatten(),
+            hidden_states_scale=hs_scale_linear.view(torch.float8_e4m3fn),
            gemm1_weights=self.gemm1_weights_fp4_shuffled.data,
            gemm1_weights_scale=self.gemm1_scales_fp4_shuffled.data.view(
                torch.float8_e4m3fn