From 1689f7ef86c3e87842fbbb649aa580606bf0fa4b Mon Sep 17 00:00:00 2001
From: kxz2002 <115912648+kxz2002@users.noreply.github.com>
Date: Wed, 5 Nov 2025 10:59:47 +0800
Subject: [PATCH] [Cherry-Pick] Fix ernie4_5_vl_processor.py and
 qwen_vl_processor.py can not disable thinking  (#4762) (#4798)

* [Feature] add a new reasoning parser (#4571)

* add new reasoning_parser initial commit

* add parser file content

* add register

* ernie_test_reasoning_parser

* support <tool_call> token and add tool_parser

* add and fix unit tests

* modify reasoning_parser

* modify reasoning parser and tool parser

* modify unit tests

* modify reasoning_parser and tool_parser

* modify unit tests

* fix tool_parser

* modify the logic of reasoning_parser and tool_parser

* add and modify unit tests

* standardize code style

* simplify reasoning_parser and tool_parser

* modify unit test

* [BugFix] Fix finish reason in _create_chat_completion_choice  (#4582)

* fix n_param _create_chat_completion_choicel

* fix unit test

* fix final_res

* modify unit tests

* [BugFix] fix offline llm chat "enable_thinking" is always "False" (#4686)

* fix enable_thinking

* recover ernie4_5_vl_processor

* [BugFix] Fix ernie4_5_vl_processor.py and qwen_vl_processor.py can not disable thinking (#4762)

* fix ernie4_5_vl_processor.py and qwen_vl_processor.py

* add unit test
---
 .../ernie4_5_vl_processor.py                  |  2 +-
 .../qwen_vl_processor/qwen_vl_processor.py    |  2 +-
 tests/input/test_qwen_vl_processor.py         | 20 +++++++++++++++++++
 3 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/fastdeploy/input/ernie4_5_vl_processor/ernie4_5_vl_processor.py b/fastdeploy/input/ernie4_5_vl_processor/ernie4_5_vl_processor.py
index 317980038..862709757 100644
--- a/fastdeploy/input/ernie4_5_vl_processor/ernie4_5_vl_processor.py
+++ b/fastdeploy/input/ernie4_5_vl_processor/ernie4_5_vl_processor.py
@@ -233,7 +233,7 @@ class Ernie4_5_VLProcessor(Ernie4_5Processor):
             if chat_template_kwargs:
                 if isinstance(chat_template_kwargs, dict):
                     for k, v in chat_template_kwargs.items():
-                        if k not in request:
+                        if k not in request or request[k] is None:
                             request[k] = v
                 else:
                     raise ValueError("Invalid input: chat_template_kwargs must be a dict")
diff --git a/fastdeploy/input/qwen_vl_processor/qwen_vl_processor.py b/fastdeploy/input/qwen_vl_processor/qwen_vl_processor.py
index cc649c9ba..06f43f335 100644
--- a/fastdeploy/input/qwen_vl_processor/qwen_vl_processor.py
+++ b/fastdeploy/input/qwen_vl_processor/qwen_vl_processor.py
@@ -237,7 +237,7 @@ class QwenVLProcessor(TextProcessor):
             if chat_template_kwargs:
                 if isinstance(chat_template_kwargs, dict):
                     for k, v in chat_template_kwargs.items():
-                        if k not in request:
+                        if k not in request or request[k] is None:
                             request[k] = v
                 else:
                     raise ValueError("Invalid input: chat_template_kwargs must be a dict")
diff --git a/tests/input/test_qwen_vl_processor.py b/tests/input/test_qwen_vl_processor.py
index a6db80cde..083f7017c 100644
--- a/tests/input/test_qwen_vl_processor.py
+++ b/tests/input/test_qwen_vl_processor.py
@@ -204,6 +204,26 @@ class TestQwenVLProcessor(unittest.TestCase):
             result["multimodal_inputs"]["image_type_ids"].shape[0], result["multimodal_inputs"]["grid_thw"][:, 0].sum()
         )
 
+    def test_process_request_dict_enable_thinking(self):
+        num_completion_token_ids = 10
+        request = {
+            "request_id": "12345",
+            "completion_token_ids": [1] * num_completion_token_ids,
+            "stop": ["stop", "eof"],
+            "messages": [
+                {
+                    "role": "user",
+                    "content": [
+                        {"type": "text", "text": "Hello"},
+                    ],
+                }
+            ],
+            "chat_template_kwargs": {"enable_thinking": True},
+        }
+
+        result = self.processor.process_request_dict(request, 100)
+        self.assertEqual(result.get("enable_thinking"), False)
+
     def test_prompt(self):
         """
         Test processing of prompt with image and video placeholders