Fix: auto-keyword and auto-question fail with qwq model (#8190)

### What problem does this PR solve?

Fix auto-keyword and auto-question fail with qwq model. #8189 

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

rag/llm/chat_model.py

@@ -636,21 +636,49 @@ class QWenChat(Base):
                 return "".join(result_list[:-1]), result_list[-1]
 
     def _chat(self, history, gen_conf):
-        if self.is_reasoning_model(self.model_name) or self.model_name in ["qwen-vl-plus", "qwen-vl-plus-latest", "qwen-vl-max", "qwen-vl-max-latest"]:
-            return super()._chat(history, gen_conf)
-        response = Generation.call(self.model_name, messages=history, result_format="message", **gen_conf)
-        ans = ""
         tk_count = 0
-        if response.status_code == HTTPStatus.OK:
+        if self.is_reasoning_model(self.model_name) or self.model_name in ["qwen-vl-plus", "qwen-vl-plus-latest", "qwen-vl-max", "qwen-vl-max-latest"]:
-            ans += response.output.choices[0]["message"]["content"]
+            try:
-            tk_count += self.total_token_count(response)
+                response = super()._chat(history, gen_conf)
-            if response.output.choices[0].get("finish_reason", "") == "length":
+                return response
-                if is_chinese([ans]):
+            except Exception as e:
-                    ans += LENGTH_NOTIFICATION_CN
+                error_msg = str(e).lower()
+                if "invalid_parameter_error" in error_msg and "only support stream mode" in error_msg:
+                    return self._simulate_one_shot_from_stream(history, gen_conf)
                 else:
-                    ans += LENGTH_NOTIFICATION_EN
+                    return "**ERROR**: " + str(e), tk_count
-            return ans, tk_count
+
-        return "**ERROR**: " + response.message, tk_count
+        try:
+            ans = ""
+            response = Generation.call(self.model_name, messages=history, result_format="message", **gen_conf)
+            if response.status_code == HTTPStatus.OK:
+                ans += response.output.choices[0]["message"]["content"]
+                tk_count += self.total_token_count(response)
+                if response.output.choices[0].get("finish_reason", "") == "length":
+                    if is_chinese([ans]):
+                        ans += LENGTH_NOTIFICATION_CN
+                    else:
+                        ans += LENGTH_NOTIFICATION_EN
+                return ans, tk_count
+            return "**ERROR**: " + response.message, tk_count
+        except Exception as e:
+            error_msg = str(e).lower()
+            if "invalid_parameter_error" in error_msg and "only support stream mode" in error_msg:
+                return self._simulate_one_shot_from_stream(history, gen_conf)
+            else:
+                return "**ERROR**: " + str(e), tk_count
+
+    def _simulate_one_shot_from_stream(self, history, gen_conf):
+        """
+        Handles models that require streaming output but need one-shot response.
+        """
+        g = self._chat_streamly("", history, gen_conf, incremental_output=True)
+        result_list = list(g)
+        error_msg_list = [item for item in result_list if str(item).find("**ERROR**") >= 0]
+        if len(error_msg_list) > 0:
+            return "**ERROR**: " + "".join(error_msg_list), 0
+        else:
+            return "".join(result_list[:-1]), result_list[-1]
 
     def _wrap_toolcall_message(self, old_message, message):
         if not old_message:
@@ -943,10 +971,10 @@ class LocalAIChat(Base):
 
 
 class LocalLLM(Base):
-
     def __init__(self, key, model_name, base_url=None, **kwargs):
         super().__init__(key, model_name, base_url=base_url, **kwargs)
         from jina import Client
+
         self.client = Client(port=12345, protocol="grpc", asyncio=True)
 
     def _prepare_prompt(self, system, history, gen_conf):
@@ -1003,13 +1031,7 @@ class VolcEngineChat(Base):
 
 
 class MiniMaxChat(Base):
-    def __init__(
+    def __init__(self, key, model_name, base_url="https://api.minimax.chat/v1/text/chatcompletion_v2", **kwargs):
-        self,
-        key,
-        model_name,
-        base_url="https://api.minimax.chat/v1/text/chatcompletion_v2",
-        **kwargs
-    ):
         super().__init__(key, model_name, base_url=base_url, **kwargs)
 
         if not base_url:
@@ -1241,6 +1263,7 @@ class GeminiChat(Base):
 
     def _chat(self, history, gen_conf):
         from google.generativeai.types import content_types
+
         system = history[0]["content"] if history and history[0]["role"] == "system" else ""
         hist = []
         for item in history: