[Kernel] moe wna16 marlin kernel (#14447)

Signed-off-by: Jinzhen Lin <linjinzhen@hotmail.com>
Co-authored-by: Michael Goin <michael@neuralmagic.com>
Co-authored-by: mgoin <mgoin64@gmail.com>

vllm/model_executor/layers/fused_moe/layer.py

@@ -472,6 +472,7 @@ class FusedMoE(torch.nn.Module):
         self.global_num_experts = num_experts
 
         assert intermediate_size % self.tp_size == 0
+        self.hidden_size = hidden_size
         self.intermediate_size_per_partition = intermediate_size // self.tp_size
         self.reduce_results = reduce_results
         self.renormalize = renormalize