From 67f544ce9ef519b43c52c3a760e2048527f4bc0c Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 28 Nov 2025 14:03:04 +0800
Subject: [PATCH 1/9] add support for ModelScope API
---
app/common/config.py | 6 ++++
app/core/entities.py | 1 +
app/core/llm/check_llm.py | 45 ++++++++++++++++++++++-------
app/core/llm/client.py | 36 +++++++++++++++++++----
app/core/task_factory.py | 4 +++
app/view/setting_interface.py | 8 +++++
resource/subtitle_style/default.txt | 2 +-
7 files changed, 84 insertions(+), 18 deletions(-)
diff --git a/app/common/config.py b/app/common/config.py
index 8b59c032..b64450b2 100644
--- a/app/common/config.py
+++ b/app/common/config.py
@@ -125,6 +125,12 @@ class Config(QConfig):
"LLM", "ChatGLM_API_Base", "https://open.bigmodel.cn/api/paas/v4"
)
+ modelscope_model = ConfigItem("LLM", "ModelScope_Model", "Qwen/Qwen3-8B")
+ modelscope_api_key = ConfigItem("LLM", "ModelScope_API_Key", "")
+ modelscope_api_base = ConfigItem(
+ "LLM", "ModelScope_API_Base", "https://api-inference.modelscope.cn/v1"
+ )
+
# ------------------- 翻译配置 -------------------
translator_service = OptionsConfigItem(
"Translate",
diff --git a/app/core/entities.py b/app/core/entities.py
index a24872e9..acacb455 100644
--- a/app/core/entities.py
+++ b/app/core/entities.py
@@ -105,6 +105,7 @@ class LLMServiceEnum(Enum):
LM_STUDIO = "LM Studio"
GEMINI = "Gemini"
CHATGLM = "ChatGLM"
+ MODELSCOPE = "ModelScope"
class TranscribeModelEnum(Enum):
diff --git a/app/core/llm/check_llm.py b/app/core/llm/check_llm.py
index 572571f6..f972dd6b 100644
--- a/app/core/llm/check_llm.py
+++ b/app/core/llm/check_llm.py
@@ -26,17 +26,40 @@ def check_llm_connection(
# 创建OpenAI客户端并发送请求到API
base_url = normalize_base_url(base_url)
api_key = api_key.strip()
- response = openai.OpenAI(
- base_url=base_url, api_key=api_key, timeout=60
- ).chat.completions.create(
- model=model,
- messages=[
- {"role": "system", "content": "You are a helpful assistant."},
- {"role": "user", "content": 'Just respond with "Hello"!'},
- ],
- timeout=30,
- )
- return True, response.choices[0].message.content
+ # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameters
+ if "modelscope" in base_url:
+ extra_body = {"enable_thinking": True}
+ response_stream = openai.OpenAI(
+ base_url=base_url, api_key=api_key, timeout=60
+ ).chat.completions.create(
+ model=model,
+ messages=[
+ {"role": "system", "content": "You are a helpful assistant."},
+ {"role": "user", 'content': 'Just respond with "Hello"!'},
+ ],
+ stream=True,
+ extra_body=extra_body,
+ timeout=30,
+ )
+
+ full_answer = ""
+ for chunk in response_stream:
+ if chunk.choices and chunk.choices[0].delta.content:
+ full_answer += chunk.choices[0].delta.content
+
+ return True, full_answer.strip()
+ else:
+ response = openai.OpenAI(
+ base_url=base_url, api_key=api_key, timeout=60
+ ).chat.completions.create(
+ model=model,
+ messages=[
+ {"role": "system", "content": "You are a helpful assistant."},
+ {"role": "user", "content": 'Just respond with "Hello"!'},
+ ],
+ timeout=30,
+ )
+ return True, response.choices[0].message.content
except openai.APIConnectionError:
return False, "API Connection Error. Please check your network or VPN."
except openai.RateLimitError as e:
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index 327a8108..00bafb8a 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -17,6 +17,7 @@
from app.core.utils.cache import get_llm_cache, memoize
from app.core.utils.logger import setup_logger
+from types import SimpleNamespace
_global_client: Optional[OpenAI] = None
_client_lock = threading.Lock()
@@ -135,13 +136,36 @@ def call_llm(
ValueError: If response is invalid (empty choices or content)
"""
client = get_llm_client()
+ # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameters
+ if "modelscope" in str(client.base_url):
+ logger.info("Detected ModelScope API, using stream mode with enable_thinking=True.")
+
+ extra_body = {"enable_thinking": True}
+
+ response_stream = client.chat.completions.create(
+ model=model,
+ messages=messages, # pyright: ignore[reportArgumentType]
+ temperature=temperature,
+ stream=True,
+ extra_body=extra_body,
+ **kwargs,
+ )
- response = client.chat.completions.create(
- model=model,
- messages=messages, # pyright: ignore[reportArgumentType]
- temperature=temperature,
- **kwargs,
- )
+ full_content = ""
+ for chunk in response_stream:
+ if chunk.choices and chunk.choices[0].delta.content:
+ full_content += chunk.choices[0].delta.content
+
+ fake_message = SimpleNamespace(content=full_content)
+ fake_choice = SimpleNamespace(message=fake_message)
+ response = SimpleNamespace(choices=[fake_choice])
+ else:
+ response = client.chat.completions.create(
+ model=model,
+ messages=messages, # pyright: ignore[reportArgumentType]
+ temperature=temperature,
+ **kwargs,
+ )
# Validate response (exceptions are not cached by diskcache)
if not (
diff --git a/app/core/task_factory.py b/app/core/task_factory.py
index db493539..019504eb 100644
--- a/app/core/task_factory.py
+++ b/app/core/task_factory.py
@@ -142,6 +142,10 @@ def create_subtitle_task(
base_url = cfg.chatglm_api_base.value
api_key = cfg.chatglm_api_key.value
llm_model = cfg.chatglm_model.value
+ elif current_service == LLMServiceEnum.MODELSCOPE:
+ base_url = cfg.modelscope_api_base.value
+ api_key = cfg.modelscope_api_key.value
+ llm_model = cfg.modelscope_model.value
else:
base_url = ""
api_key = ""
diff --git a/app/view/setting_interface.py b/app/view/setting_interface.py
index df5db3b3..fdc37f9a 100644
--- a/app/view/setting_interface.py
+++ b/app/view/setting_interface.py
@@ -353,6 +353,14 @@ def __createLLMServiceCards(self):
"default_base": "https://open.bigmodel.cn/api/paas/v4",
"default_models": ["glm-4-plus", "glm-4-air-250414", "glm-4-flash"],
},
+ LLMServiceEnum.MODELSCOPE: {
+ "prefix": "modelscope",
+ "api_key_cfg": cfg.modelscope_api_key,
+ "api_base_cfg": cfg.modelscope_api_base,
+ "model_cfg": cfg.modelscope_model,
+ "default_base": "https://api-inference.modelscope.cn/v1",
+ "default_models": ["Qwen/Qwen3-8B", "Qwen/Qwen3-30B-A3B-Instruct-2507", "deepseek-ai/DeepSeek-V3.1"],
+ },
}
# 创建服务配置映射
diff --git a/resource/subtitle_style/default.txt b/resource/subtitle_style/default.txt
index 38deb631..e31ddcd1 100644
--- a/resource/subtitle_style/default.txt
+++ b/resource/subtitle_style/default.txt
@@ -1,4 +1,4 @@
[V4+ Styles]
Format: Name,Fontname,Fontsize,PrimaryColour,SecondaryColour,OutlineColour,BackColour,Bold,Italic,Underline,StrikeOut,ScaleX,ScaleY,Spacing,Angle,BorderStyle,Outline,Shadow,Alignment,MarginL,MarginR,MarginV,Encoding
-Style: Default,Arial,42,&H005aff65,&H000000FF,&H00000000,&H00000000,-1,0,0,0,100,100,3.2,0,1,2.0,0,2,10,10,30,1,\q1
+Style: Default,Arial,30,&H005aff65,&H000000FF,&H00000000,&H00000000,-1,0,0,0,100,100,3.2,0,1,2.0,0,2,10,10,30,1,\q1
Style: Secondary,Arial,30,&H00ffffff,&H000000FF,&H00000000,&H00000000,-1,0,0,0,100,100,0.8,0,1,2.0,0,2,10,10,30,1,\q1
\ No newline at end of file
From 964b9d4c913fbff14e148b77697d0fb38d3a9255 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 28 Nov 2025 14:08:49 +0800
Subject: [PATCH 2/9] restore default subtitle settings
---
resource/subtitle_style/default.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/resource/subtitle_style/default.txt b/resource/subtitle_style/default.txt
index e31ddcd1..38deb631 100644
--- a/resource/subtitle_style/default.txt
+++ b/resource/subtitle_style/default.txt
@@ -1,4 +1,4 @@
[V4+ Styles]
Format: Name,Fontname,Fontsize,PrimaryColour,SecondaryColour,OutlineColour,BackColour,Bold,Italic,Underline,StrikeOut,ScaleX,ScaleY,Spacing,Angle,BorderStyle,Outline,Shadow,Alignment,MarginL,MarginR,MarginV,Encoding
-Style: Default,Arial,30,&H005aff65,&H000000FF,&H00000000,&H00000000,-1,0,0,0,100,100,3.2,0,1,2.0,0,2,10,10,30,1,\q1
+Style: Default,Arial,42,&H005aff65,&H000000FF,&H00000000,&H00000000,-1,0,0,0,100,100,3.2,0,1,2.0,0,2,10,10,30,1,\q1
Style: Secondary,Arial,30,&H00ffffff,&H000000FF,&H00000000,&H00000000,-1,0,0,0,100,100,0.8,0,1,2.0,0,2,10,10,30,1,\q1
\ No newline at end of file
From bd043bddb6eac349175794b9a4e58e61fda282c8 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 28 Nov 2025 14:27:42 +0800
Subject: [PATCH 3/9] update README.md
---
README.md | 1 +
1 file changed, 1 insertion(+)
diff --git a/README.md b/README.md
index 2c4c8256..d8d6792d 100644
--- a/README.md
+++ b/README.md
@@ -138,6 +138,7 @@ LLM 大模型是用来字幕段句、字幕优化、以及字幕翻译（如果
| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| SiliconCloud | [SiliconCloud 官网](https://cloud.siliconflow.cn/i/onCHcaDx)配置方法请参考[配置文档](./docs/llm_config.md)
该并发较低，建议把线程设置为5以下。 |
| DeepSeek | [DeepSeek 官网](https://platform.deepseek.com)，建议使用 `deepseek-v3` 模型，
官方网站最近服务好像并不太稳定。 |
+| ModelScope | [ModelScope 官网](https://modelscope.cn/models?filter=inference_type&page=1&tabKey=task)配置方法请参考[配置文档](https://modelscope.cn/docs/model-service/API-Inference/intro)
该并发较低，建议把线程设置为5以下。 |
| OpenAI兼容接口 | 如果有其他服务商的API，可直接在软件中填写。base_url 和api_key [VideoCaptioner API](https://api.videocaptioner.cn) |
注：如果用的 API 服务商不支持高并发，请在软件设置中将“线程数”调低，避免请求错误。
From 8434abdfba0f8f366126bf4f56c8cbc7947e31a4 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 28 Nov 2025 16:40:06 +0800
Subject: [PATCH 4/9] modify some details
---
README.md | 2 +-
app/core/llm/check_llm.py | 5 +++--
app/core/llm/client.py | 5 +++--
3 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/README.md b/README.md
index d8d6792d..1c58851e 100644
--- a/README.md
+++ b/README.md
@@ -138,7 +138,7 @@ LLM 大模型是用来字幕段句、字幕优化、以及字幕翻译（如果
| -------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| SiliconCloud | [SiliconCloud 官网](https://cloud.siliconflow.cn/i/onCHcaDx)配置方法请参考[配置文档](./docs/llm_config.md)
该并发较低，建议把线程设置为5以下。 |
| DeepSeek | [DeepSeek 官网](https://platform.deepseek.com)，建议使用 `deepseek-v3` 模型，
官方网站最近服务好像并不太稳定。 |
-| ModelScope | [ModelScope 官网](https://modelscope.cn/models?filter=inference_type&page=1&tabKey=task)配置方法请参考[配置文档](https://modelscope.cn/docs/model-service/API-Inference/intro)
该并发较低，建议把线程设置为5以下。 |
+| ModelScope | [ModelScope 官网](https://modelscope.cn/models?filter=inference_type&page=1&tabKey=task)配置方法请参考[配置文档](https://modelscope.cn/docs/model-service/API-Inference/intro)
该并发较低，建议把线程设置为5以下。 |
| OpenAI兼容接口 | 如果有其他服务商的API，可直接在软件中填写。base_url 和api_key [VideoCaptioner API](https://api.videocaptioner.cn) |
注：如果用的 API 服务商不支持高并发，请在软件设置中将“线程数”调低，避免请求错误。
diff --git a/app/core/llm/check_llm.py b/app/core/llm/check_llm.py
index f972dd6b..238d993e 100644
--- a/app/core/llm/check_llm.py
+++ b/app/core/llm/check_llm.py
@@ -26,7 +26,7 @@ def check_llm_connection(
# 创建OpenAI客户端并发送请求到API
base_url = normalize_base_url(base_url)
api_key = api_key.strip()
- # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameters
+ # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameter
if "modelscope" in base_url:
extra_body = {"enable_thinking": True}
response_stream = openai.OpenAI(
@@ -46,7 +46,8 @@ def check_llm_connection(
for chunk in response_stream:
if chunk.choices and chunk.choices[0].delta.content:
full_answer += chunk.choices[0].delta.content
-
+ if not full_answer:
+ raise ValueError("ModelScope streaming response yielded no content")
return True, full_answer.strip()
else:
response = openai.OpenAI(
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index 00bafb8a..7856835f 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -136,7 +136,7 @@ def call_llm(
ValueError: If response is invalid (empty choices or content)
"""
client = get_llm_client()
- # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameters
+ # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameters
if "modelscope" in str(client.base_url):
logger.info("Detected ModelScope API, using stream mode with enable_thinking=True.")
@@ -155,7 +155,8 @@ def call_llm(
for chunk in response_stream:
if chunk.choices and chunk.choices[0].delta.content:
full_content += chunk.choices[0].delta.content
-
+ if not full_content:
+ raise ValueError("ModelScope streaming response yielded no content")
fake_message = SimpleNamespace(content=full_content)
fake_choice = SimpleNamespace(message=fake_message)
response = SimpleNamespace(choices=[fake_choice])
From 32dcb319934d4f7974a4ef729dff738bd0a9dbc2 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 28 Nov 2025 16:44:21 +0800
Subject: [PATCH 5/9] change import statement position
---
app/core/llm/client.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index 7856835f..bd7a2f1c 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -3,6 +3,7 @@
import os
import threading
from typing import Any, List, Optional
+from types import SimpleNamespace
from urllib.parse import urlparse, urlunparse
import openai
@@ -17,7 +18,6 @@
from app.core.utils.cache import get_llm_cache, memoize
from app.core.utils.logger import setup_logger
-from types import SimpleNamespace
_global_client: Optional[OpenAI] = None
_client_lock = threading.Lock()
From 7e650a78b732240580e3e46372a72077a61d13e4 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 19 Dec 2025 14:50:06 +0800
Subject: [PATCH 6/9] modify extra parameter logic
---
app/common/config.py | 7 ++++++
app/core/entities.py | 3 ++-
app/core/llm/check_llm.py | 46 ++++++++-------------------------
app/core/llm/client.py | 53 ++++++++++++++++-----------------------
app/core/task_factory.py | 3 +++
5 files changed, 45 insertions(+), 67 deletions(-)
diff --git a/app/common/config.py b/app/common/config.py
index b64450b2..a811a1aa 100644
--- a/app/common/config.py
+++ b/app/common/config.py
@@ -130,6 +130,13 @@ class Config(QConfig):
modelscope_api_base = ConfigItem(
"LLM", "ModelScope_API_Base", "https://api-inference.modelscope.cn/v1"
)
+ modelscope_extra_body = ConfigItem(
+ "LLM",
+ "ModelScope_Extra_Body",
+ {
+ "enable_thinking": False,
+ },
+ )
# ------------------- 翻译配置 -------------------
translator_service = OptionsConfigItem(
diff --git a/app/core/entities.py b/app/core/entities.py
index acacb455..f1d04a6f 100644
--- a/app/core/entities.py
+++ b/app/core/entities.py
@@ -1,7 +1,7 @@
import datetime
from dataclasses import dataclass, field
from enum import Enum
-from typing import TYPE_CHECKING, Literal, Optional
+from typing import TYPE_CHECKING, Literal, Optional, Dict, Any
if TYPE_CHECKING:
from app.core.translate.types import TargetLanguage
@@ -559,6 +559,7 @@ class SubtitleConfig:
base_url: Optional[str] = None
api_key: Optional[str] = None
llm_model: Optional[str] = None
+ extra_body: Optional[Dict[str, Any]] = None
deeplx_endpoint: Optional[str] = None
# 翻译服务
translator_service: Optional[TranslatorServiceEnum] = None
diff --git a/app/core/llm/check_llm.py b/app/core/llm/check_llm.py
index 238d993e..572571f6 100644
--- a/app/core/llm/check_llm.py
+++ b/app/core/llm/check_llm.py
@@ -26,41 +26,17 @@ def check_llm_connection(
# 创建OpenAI客户端并发送请求到API
base_url = normalize_base_url(base_url)
api_key = api_key.strip()
- # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameter
- if "modelscope" in base_url:
- extra_body = {"enable_thinking": True}
- response_stream = openai.OpenAI(
- base_url=base_url, api_key=api_key, timeout=60
- ).chat.completions.create(
- model=model,
- messages=[
- {"role": "system", "content": "You are a helpful assistant."},
- {"role": "user", 'content': 'Just respond with "Hello"!'},
- ],
- stream=True,
- extra_body=extra_body,
- timeout=30,
- )
-
- full_answer = ""
- for chunk in response_stream:
- if chunk.choices and chunk.choices[0].delta.content:
- full_answer += chunk.choices[0].delta.content
- if not full_answer:
- raise ValueError("ModelScope streaming response yielded no content")
- return True, full_answer.strip()
- else:
- response = openai.OpenAI(
- base_url=base_url, api_key=api_key, timeout=60
- ).chat.completions.create(
- model=model,
- messages=[
- {"role": "system", "content": "You are a helpful assistant."},
- {"role": "user", "content": 'Just respond with "Hello"!'},
- ],
- timeout=30,
- )
- return True, response.choices[0].message.content
+ response = openai.OpenAI(
+ base_url=base_url, api_key=api_key, timeout=60
+ ).chat.completions.create(
+ model=model,
+ messages=[
+ {"role": "system", "content": "You are a helpful assistant."},
+ {"role": "user", "content": 'Just respond with "Hello"!'},
+ ],
+ timeout=30,
+ )
+ return True, response.choices[0].message.content
except openai.APIConnectionError:
return False, "API Connection Error. Please check your network or VPN."
except openai.RateLimitError as e:
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index bd7a2f1c..6e2e8d0c 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -1,5 +1,6 @@
"""Unified LLM client for the application."""
-
+from app.core.utils.logger import setup_logger
+logger = setup_logger(__name__) # 使用当前模块名作为logger名称
import os
import threading
from typing import Any, List, Optional
@@ -136,37 +137,27 @@ def call_llm(
ValueError: If response is invalid (empty choices or content)
"""
client = get_llm_client()
- # Check whether it is the ModelScope platform, as some models require the stream and enable_thinking parameters
- if "modelscope" in str(client.base_url):
- logger.info("Detected ModelScope API, using stream mode with enable_thinking=True.")
-
- extra_body = {"enable_thinking": True}
-
- response_stream = client.chat.completions.create(
- model=model,
- messages=messages, # pyright: ignore[reportArgumentType]
- temperature=temperature,
- stream=True,
- extra_body=extra_body,
- **kwargs,
- )
- full_content = ""
- for chunk in response_stream:
- if chunk.choices and chunk.choices[0].delta.content:
- full_content += chunk.choices[0].delta.content
- if not full_content:
- raise ValueError("ModelScope streaming response yielded no content")
- fake_message = SimpleNamespace(content=full_content)
- fake_choice = SimpleNamespace(message=fake_message)
- response = SimpleNamespace(choices=[fake_choice])
- else:
- response = client.chat.completions.create(
- model=model,
- messages=messages, # pyright: ignore[reportArgumentType]
- temperature=temperature,
- **kwargs,
- )
+ # # 在需要打印的地方
+ # extra_body = kwargs.get("extra_body", {})
+ # logger.info(f"DEBUG: Passing extra_body to API: {extra_body}") # 使用 logger
+ # --- 核心：智能处理平台特定参数 ---
+ # 检查是否为 ModelScope 平台
+ # if "modelscope" in str(client.base_url):
+ # logger.info("Detected ModelScope API, ensuring 'enable_thinking' is set to false.")
+ # # 从 kwargs 中获取现有的 extra_body，如果没有则创建一个空字典
+ # extra_body = kwargs.get("extra_body", {})
+ # # 强制设置 enable_thinking 为 false，以满足其 API 要求
+ # extra_body["enable_thinking"] = False
+ # # 将修改后的 extra_body 放回 kwargs
+ # kwargs["extra_body"] = extra_body
+
+ response = client.chat.completions.create(
+ model=model,
+ messages=messages, # pyright: ignore[reportArgumentType]
+ temperature=temperature,
+ **kwargs,
+ )
# Validate response (exceptions are not cached by diskcache)
if not (
diff --git a/app/core/task_factory.py b/app/core/task_factory.py
index 019504eb..8cd7c584 100644
--- a/app/core/task_factory.py
+++ b/app/core/task_factory.py
@@ -112,6 +112,7 @@ def create_subtitle_task(
Path(file_path).parent / f"【字幕】{output_name}{suffix}.srt"
)
+ extra_body = None # 默认值为None
# 根据当前选择的LLM服务获取对应的配置
current_service = cfg.llm_service.value
if current_service == LLMServiceEnum.OPENAI:
@@ -146,6 +147,7 @@ def create_subtitle_task(
base_url = cfg.modelscope_api_base.value
api_key = cfg.modelscope_api_key.value
llm_model = cfg.modelscope_model.value
+ extra_body = cfg.modelscope_extra_body.value
else:
base_url = ""
api_key = ""
@@ -156,6 +158,7 @@ def create_subtitle_task(
base_url=base_url,
api_key=api_key,
llm_model=llm_model,
+ extra_body=extra_body,
deeplx_endpoint=cfg.deeplx_endpoint.value,
# 翻译服务
translator_service=cfg.translator_service.value,
From 68504b51fcc437d4603fa10815206401f283142e Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 19 Dec 2025 14:53:37 +0800
Subject: [PATCH 7/9] update
---
app/core/llm/client.py | 14 --------------
1 file changed, 14 deletions(-)
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index 6e2e8d0c..d2c714f7 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -138,20 +138,6 @@ def call_llm(
"""
client = get_llm_client()
- # # 在需要打印的地方
- # extra_body = kwargs.get("extra_body", {})
- # logger.info(f"DEBUG: Passing extra_body to API: {extra_body}") # 使用 logger
- # --- 核心：智能处理平台特定参数 ---
- # 检查是否为 ModelScope 平台
- # if "modelscope" in str(client.base_url):
- # logger.info("Detected ModelScope API, ensuring 'enable_thinking' is set to false.")
- # # 从 kwargs 中获取现有的 extra_body，如果没有则创建一个空字典
- # extra_body = kwargs.get("extra_body", {})
- # # 强制设置 enable_thinking 为 false，以满足其 API 要求
- # extra_body["enable_thinking"] = False
- # # 将修改后的 extra_body 放回 kwargs
- # kwargs["extra_body"] = extra_body
-
response = client.chat.completions.create(
model=model,
messages=messages, # pyright: ignore[reportArgumentType]
From 5f91ca59a9f4724afe0d37bf1815a8e87dab89a1 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 19 Dec 2025 14:55:28 +0800
Subject: [PATCH 8/9] delete some unused packages
---
app/core/llm/client.py | 3 ---
1 file changed, 3 deletions(-)
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index d2c714f7..e99bb2f6 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -1,10 +1,7 @@
"""Unified LLM client for the application."""
-from app.core.utils.logger import setup_logger
-logger = setup_logger(__name__) # 使用当前模块名作为logger名称
import os
import threading
from typing import Any, List, Optional
-from types import SimpleNamespace
from urllib.parse import urlparse, urlunparse
import openai
From 2c0ab2342c96703af4452155087aca42af714201 Mon Sep 17 00:00:00 2001
From: yrk <2493404415@qq.com>
Date: Fri, 19 Dec 2025 15:02:01 +0800
Subject: [PATCH 9/9] restore the deleted blank lines
---
app/core/llm/client.py | 1 +
1 file changed, 1 insertion(+)
diff --git a/app/core/llm/client.py b/app/core/llm/client.py
index e99bb2f6..327a8108 100644
--- a/app/core/llm/client.py
+++ b/app/core/llm/client.py
@@ -1,4 +1,5 @@
"""Unified LLM client for the application."""
+
import os
import threading
from typing import Any, List, Optional