diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 30b637e23..5e68958a9 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -110,6 +110,12 @@ jobs:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
   
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v1
+
+      - name: set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
       - name: Build backend and push
         id: docker_build_backend
         run: |
@@ -158,6 +164,12 @@ jobs:
       - name: Set Environment Variable
         run: echo "RELEASE_VERSION=${{ steps.get_version.outputs.VERSION }}" >> $GITHUB_ENV
 
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v1
+
+      - name: set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
       # 登录 docker hub
       - name: Login to DockerHub
         uses: docker/login-action@v1
diff --git a/src/backend/bisheng/interface/llms/custom.py b/src/backend/bisheng/interface/llms/custom.py
index 60f99ead9..0fd20ed16 100644
--- a/src/backend/bisheng/interface/llms/custom.py
+++ b/src/backend/bisheng/interface/llms/custom.py
@@ -207,52 +207,6 @@ class BishengLLM(BaseChatModel):
         params = params_handler(default_params, server_config, model_config)
         return params
 
-        params = {}
-        if server_info.config:
-            params.update(server_info.config)
-        enable_web_search = False
-        if model_info.config:
-            enable_web_search = model_info.config.get('enable_web_search', False)
-            if model_info.config.get('max_tokens'):
-                params['max_tokens'] = model_info.config.get('max_tokens')
-
-        params.update({
-            'model_name': model_info.model_name,
-            'streaming': self.streaming,
-            'temperature': self.temperature,
-            'top_p': self.top_p,
-            'cache': self.cache
-        })
-        if server_info.type == LLMServerType.OLLAMA.value:
-            params['model'] = params.pop('model_name')
-            if params.get('max_tokens'):
-                params['num_ctx'] = params.pop('max_tokens')
-        elif server_info.type == LLMServerType.AZURE_OPENAI.value:
-            params['azure_deployment'] = params.pop('model_name')
-        elif server_info.type == LLMServerType.QIAN_FAN.value:
-            params['model'] = params.pop('model_name')
-            params['qianfan_ak'] = params.pop('wenxin_api_key')
-            params['qianfan_sk'] = params.pop('wenxin_secret_key')
-            if params.get('max_tokens'):
-                params['model_kwargs'] = {"max_output_tokens": params.pop('max_tokens')}
-        elif server_info.type == LLMServerType.SPARK.value:
-            params['openai_api_key'] = f'{params.pop("api_key")}:{params.pop("api_secret")}'
-        elif server_info.type in [LLMServerType.XINFERENCE.value, LLMServerType.LLAMACPP.value,
-                                  LLMServerType.VLLM.value]:
-            params['openai_api_key'] = params.pop('openai_api_key', None) or "EMPTY"
-        elif server_info.type == LLMServerType.QWEN.value:
-            params['dashscope_api_key'] = params.pop('openai_api_key')
-            params.pop('openai_api_base', None)
-            params['model_kwargs'] = {'enable_search': enable_web_search}
-            if params.get('max_tokens'):
-                params['model_kwargs']['max_tokens'] = params.pop('max_tokens')
-        elif server_info.type == LLMServerType.TENCENT.value:
-            params['extra_body'] = {'enable_enhancement': enable_web_search}
-        elif server_info.type == LLMServerType.MINIMAX.value:
-            params['api_key'] = params.pop('openai_api_key', None)
-            params.pop('openai_api_base', None)
-        return params
-
     def _get_default_params(self, server_config: dict, model_config: dict) -> dict:
         default_params = {
             'model': self.model_info.model_name,
@@ -446,7 +400,6 @@ class BishengLLM(BaseChatModel):
 
     def _update_model_status(self, status: int, remark: str = ''):
         """更新模型状态"""
-        # todo 接入到异步任务模块 累计5分钟更新一次
         if self.model_info.status != status:
             self.model_info.status = status
             LLMDao.update_model_status(self.model_id, status, remark[:500])