[BugFix] Fix: ImportError when building on hopper systems (#20513)

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

csrc/quantization/cutlass_w8a8/moe/blockwise_scaled_group_mm_sm100.cu

@@ -1,3 +1,5 @@
+#include "core/registration.h"
+
 #include <torch/all.h>
 #include <cutlass/arch/arch.h>
 
@@ -364,4 +366,9 @@ void cutlass_blockwise_scaled_grouped_mm(
     TORCH_CHECK(false, "Unsupported output tensor type");
   }
 #endif
-}
+}
+
+TORCH_LIBRARY_IMPL_EXPAND(TORCH_EXTENSION_NAME, CUDA, m) {
+  m.impl("cutlass_blockwise_scaled_grouped_mm",
+         &cutlass_blockwise_scaled_grouped_mm);
+}