add fused_experts_impl of int8

vllm/model_executor/layers/fused_moe/fused_moe.py

@@ -1481,6 +1481,31 @@ def fused_experts_impl(
     use_nn_moe: Optional[bool] = False,
 ) -> torch.Tensor:
     # Check constraints.
+    if use_int8_w8a8 is True:
+        return fused_experts_impl_int8(hidden_states=hidden_states,
+                                       w1=w1,
+                                       w2=w2,
+                                       topk_weights=topk_weights,
+                                       topk_ids=topk_ids,
+                                       inplace=inplace,
+                                       activation=activation,
+                                       apply_router_weight_on_input=apply_router_weight_on_input,
+                                       use_fp8_w8a8=False,
+                                       use_int8_w8a8=True,
+                                       use_int8_w8a16=False,
+                                       use_int4_w4a16=False,
+                                       per_channel_quant=per_channel_quant,
+                                       global_num_experts=global_num_experts,
+                                       expert_map=expert_map,
+                                       w1_scale=w1_scale,
+                                       w2_scale=w2_scale,
+                                       w1_zp=w1_zp,
+                                       w2_zp=w2_zp,
+                                       a1_scale=a1_scale,
+                                       a2_scale=a2_scale,
+                                       block_shape=block_shape,
+                                       use_nn_moe=False
+                                       )
     if use_int4_w4a16:
         assert hidden_states.size(1) // 2 == w1.size(2), (
             "Hidden size mismatch")