Update batch size limitation of dsv3_router_gemm kernel to 16 (#8051)

2025-07-31 20:53:31 -07:00
parent dd7ca00601
commit e7e5a3050a
1 changed files with 1 additions and 2 deletions
--- a/python/sglang/srt/models/deepseek_v2.py
+++ b/python/sglang/srt/models/deepseek_v2.py
@@ -252,8 +252,7 @@ class MoEGate(nn.Module):
        # NOTE: For some unknown reason, router_gemm seems degrade accept length.
        if (
            _is_cuda
-            and not self.is_nextn
-            and hidden_states.shape[0] < 4
+            and hidden_states.shape[0] <= 16
            and hidden_states.shape[1] == 7168
            and self.weight.shape[0] == 256
            and _device_sm >= 90