[BugFix] Enable moe_gate_fp32 using FD_ENABLE_RL (#7130)

* rl gate fp32

* clean

fastdeploy/engine/args_utils.py

@@ -624,6 +624,8 @@ class EngineArgs:
             raise NotImplementedError(
                 f"not support model_impl: '{self.model_impl}'. " f"Must be one of: {', '.join(valid_model_impls)}"
             )
+        if envs.FD_ENABLE_RL == 1:
+            self.moe_gate_fp32 = True
 
         self.post_init_all_ports()
 

fastdeploy/envs.py

@@ -266,6 +266,8 @@ environment_variables: dict[str, Callable[[], Any]] = {
     "FD_SAVE_OUTPUT_CACHE_FOR_PREEMPTED_REQUEST": lambda: bool(
         int(os.getenv("FD_SAVE_OUTPUT_CACHE_FOR_PREEMPTED_REQUEST", "1"))
     ),
+    # Whether to align RoPE and moe gate precision with training
+    "FD_ENABLE_RL": lambda: int(os.getenv("FD_ENABLE_RL", "0")),
 }
 
 

fastdeploy/model_executor/models/glm4_moe.py

@@ -151,9 +151,7 @@ class Glm4Moe(nn.Layer):
             output_size=fd_config.model_config.n_routed_experts,
             with_bias=False,
             skip_quant=True,
-            weight_dtype=(
-                "float32" if fd_config.load_config.dynamic_load_weight or fd_config.model_config.moe_gate_fp32 else ""
-            ),
+            weight_dtype=("float32" if fd_config.model_config.moe_gate_fp32 else ""),
         )
         self.gate.e_score_correction_bias = self.create_parameter(
             shape=[1, fd_config.model_config.n_routed_experts],

fastdeploy/model_executor/models/qwen3moe.py

@@ -77,9 +77,7 @@ class Qwen3MoeBlock(nn.Layer):
             output_size=fd_config.model_config.num_experts,
             with_bias=False,
             skip_quant=True,
-            weight_dtype=(
-                "float32" if fd_config.load_config.dynamic_load_weight or fd_config.model_config.moe_gate_fp32 else ""
-            ),
+            weight_dtype=("float32" if fd_config.model_config.moe_gate_fp32 else ""),
         )
 
     def forward(self, x, forward_meta):