update fused_moe.py

vllm/model_executor/layers/fused_moe/fused_moe.py

@@ -915,9 +915,8 @@ def invoke_fused_moe_kernel(A: torch.Tensor,
         assert A_scale is None
         assert B_scale is None
 
-    if use_int4_w4a16:
-        EM = sorted_token_ids.shape[0]
-    elif A.shape[0] < config["BLOCK_SIZE_M"]:
+    EM = sorted_token_ids.shape[0]
+    if A.shape[0] < config["BLOCK_SIZE_M"]:
         # optimize for small batch_size.
         # We assume that top_ids of each token is unique, so
         # so num_valid_experts <= batch_size <= BLOCK_SIZE_M,