deepseek-r1-w4a8 mlp/moe调用silu-mul-quant融合

vllm/envs.py

@@ -205,6 +205,7 @@ if TYPE_CHECKING:
     VLLM_USE_TRITON_CAT: bool = False
     VLLM_USE_MERGE_ATTN_STATES_OPT: bool = False
     USE_FUSED_RMS_QUANT: bool = False
+    USE_FUSED_SILU_MUL_QUANT: bool = False
 
 
 def get_default_cache_root():
@@ -1401,6 +1402,10 @@ environment_variables: dict[str, Callable[[], Any]] = {
     "USE_FUSED_RMS_QUANT": 
     lambda: (os.getenv('USE_FUSED_RMS_QUANT', '0').lower() in
              ("true", "1")),
+    # vllm will use silu_mul_quant fused op 
+    "USE_FUSED_SILU_MUL_QUANT": 
+    lambda: (os.getenv('USE_FUSED_SILU_MUL_QUANT', '0').lower() in
+             ("true", "1")),
 }
 
 # --8<-- [end:env-vars-definition]

vllm/model_executor/layers/linear.py

@@ -39,6 +39,13 @@ if envs.USE_FUSED_RMS_QUANT:
     except Exception as e:
         print(f"Error: Import fused rmsquant error: {e}") 
         
+if envs.USE_FUSED_SILU_MUL_QUANT:        
+    try:
+        # from lightop import fuse_silu_mul_quant
+        from lmslim.quantize.quant_ops import lm_fuse_silu_mul_quant
+    except Exception as e:
+        print(f"Error: Import fused silu_mul_qunat error: {e}")
+
 logger = init_logger(__name__)
 
 WEIGHT_LOADER_V2_SUPPORTED = [
@@ -1500,7 +1507,8 @@ class RowParallelLinear(LinearBase):
         param.load_row_parallel_weight(loaded_weight=loaded_weight)
 
     def forward(
-        self, input_
+        self, input_,
+        use_fused_silu_mul_quant: Optional[bool] = False
     ) -> Union[torch.Tensor, tuple[torch.Tensor, Optional[Parameter]]]:
         if self.input_is_parallel:
             input_parallel = input_
@@ -1514,6 +1522,15 @@ class RowParallelLinear(LinearBase):
         # Only fuse bias add into GEMM for rank 0 (this ensures that
         # bias will not get added more than once in TP>1 case)
         bias_ = None if (self.tp_rank > 0 or self.skip_bias_add) else self.bias
+        if use_fused_silu_mul_quant:
+            xq, xs = lm_fuse_silu_mul_quant(input_parallel)
+            
+            silu_quant_args = [xq, xs]
+            output_parallel = self.quant_method.apply(self,
+                                                      input_parallel,
+                                                      bias=bias_,
+                                                      silu_quant_args=silu_quant_args)
+        else:
             output_parallel = self.quant_method.apply(self,
                                                       input_parallel,
                                                       bias=bias_)

vllm/model_executor/layers/quantization/slimquant_w4a8.py

@@ -153,11 +153,14 @@ class SlimQuantW4A8Int8LinearMethod(LinearMethodBase):
         layer: torch.nn.Module,
         x: torch.Tensor,
         bias: Optional[torch.Tensor] = None,
-        input_quant_args: Optional[list[torch.Tensor]] = None 
+        input_quant_args: Optional[list[torch.Tensor]] = None,
+        silu_quant_args: Optional[list[torch.Tensor]] = None 
     ):
         if envs.USE_FUSED_RMS_QUANT and input_quant_args is not None:
             assert len(input_quant_args) == 2
             x_q, x_scale = input_quant_args
+        elif envs.USE_FUSED_SILU_MUL_QUANT and silu_quant_args is not None:
+            x_q, x_scale = silu_quant_args
         else:
             x_q, x_scale = per_token_quant_int8(x)
 

vllm/model_executor/models/deepseek_v2.py

@@ -116,6 +116,9 @@ class DeepseekV2MLP(nn.Module):
                 ):
         if envs.USE_FUSED_RMS_QUANT:
             gate_up, new_resi, _  = self.gate_up_proj(x, rms_weight, residual, update_hd=update_hd)
+            if envs.USE_FUSED_SILU_MUL_QUANT:
+                x, _ = self.down_proj(gate_up, use_fused_silu_mul_quant=True)
+            else:
                 x = self.act_fn(gate_up)
                 x, _ = self.down_proj(x)
             return x, new_resi