gardel/Mofox-Core
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #63 from SaigyoujiYusora/debug

Browse Source 
fix: <think> was not generated
This commit is contained in:
HYY
2025-03-06 08:30:31 +08:00
committed by GitHub
parent 3ef20ac1c9 bd982ba556
commit 59af990923
1 changed files with 27 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
src/plugins/models/utils_model.py
View File

@@ -55,7 +55,7 @@ class LLM_request:
logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
await asyncio.sleep(wait_time)
continue
if response.status in [500, 503]:
logger.error(f"服务器错误: {response.status}")
raise RuntimeError("服务器负载过高，模型恢复失败QAQ")
@@ -69,10 +69,10 @@ class LLM_request:
think_match = None
reasoning_content = message.get("reasoning_content", "")
if not reasoning_content:
think_match = re.search(r'<think>(.*?)</think>', content, re.DOTALL)
think_match = re.search(r'(?:<think>)?(.*?)</think>', content, re.DOTALL)
if think_match:
reasoning_content = think_match.group(1).strip()
content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL).strip()
content = re.sub(r'(?:<think>)?.*?</think>', '', content, flags=re.DOTALL, count=1).strip()
return content, reasoning_content
return "没有返回结果", ""
@@ -118,7 +118,7 @@ class LLM_request:
],
**self.params
}
# 发送请求到完整的chat/completions端点
api_url = f"{self.base_url.rstrip('/')}/chat/completions"
@@ -126,10 +126,9 @@ class LLM_request:
max_retries = 3
base_wait_time = 15
current_image_base64 = image_base64
current_image_base64 = compress_base64_image_by_scale(current_image_base64)
for retry in range(max_retries):
try:
@@ -146,7 +145,7 @@ class LLM_request:
logger.warning("图片太大(413)，尝试压缩...")
current_image_base64 = compress_base64_image_by_scale(current_image_base64)
continue
response.raise_for_status() # 检查其他响应状态
result = await response.json()
@@ -156,10 +155,10 @@ class LLM_request:
think_match = None
reasoning_content = message.get("reasoning_content", "")
if not reasoning_content:
think_match = re.search(r'<think>(.*?)</think>', content, re.DOTALL)
think_match = re.search(r'(?:<think>)?(.*?)</think>', content, re.DOTALL)
if think_match:
reasoning_content = think_match.group(1).strip()
content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL).strip()
content = re.sub(r'(?:<think>)?.*?</think>', '', content, flags=re.DOTALL, count=1).strip()
return content, reasoning_content
return "没有返回结果", ""
@@ -181,7 +180,7 @@ class LLM_request:
"Authorization": f"Bearer {self.api_key}",
"Content-Type": "application/json"
}
# 构建请求体
data = {
"model": self.model_name,
@@ -189,14 +188,14 @@ class LLM_request:
"temperature": 0.5,
**self.params
}
# 发送请求到完整的 chat/completions 端点
api_url = f"{self.base_url.rstrip('/')}/chat/completions"
logger.info(f"Request URL: {api_url}") # 记录请求的 URL
max_retries = 3
base_wait_time = 15
async with aiohttp.ClientSession() as session:
for retry in range(max_retries):
try:
@@ -206,16 +205,23 @@ class LLM_request:
logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
await asyncio.sleep(wait_time)
continue
response.raise_for_status() # 检查其他响应状态
result = await response.json()
if "choices" in result and len(result["choices"]) > 0:
content = result["choices"][0]["message"]["content"]
reasoning_content = result["choices"][0]["message"].get("reasoning_content", "")
message = result["choices"][0]["message"]
content = message.get("content", "")
think_match = None
reasoning_content = message.get("reasoning_content", "")
if not reasoning_content:
think_match = re.search(r'(?:<think>)?(.*?)</think>', content, re.DOTALL)
if think_match:
reasoning_content = think_match.group(1).strip()
content = re.sub(r'(?:<think>)?.*?</think>', '', content, flags=re.DOTALL, count=1).strip()
return content, reasoning_content
return "没有返回结果", ""
except Exception as e:
if retry < max_retries - 1: # 如果还有重试机会
wait_time = base_wait_time * (2 ** retry)
@@ -224,7 +230,7 @@ class LLM_request:
else:
logger.error(f"请求失败: {str(e)}")
return f"请求失败: {str(e)}", ""
logger.error("达到最大重试次数，请求仍然失败")
return "达到最大重试次数，请求仍然失败", ""
@@ -288,10 +294,10 @@ class LLM_request:
think_match = None
reasoning_content = message.get("reasoning_content", "")
if not reasoning_content:
think_match = re.search(r'<think>(.*?)</think>', content, re.DOTALL)
think_match = re.search(r'(?:<think>)?(.*?)</think>', content, re.DOTALL)
if think_match:
reasoning_content = think_match.group(1).strip()
content = re.sub(r'<think>.*?</think>', '', content, flags=re.DOTALL).strip()
content = re.sub(r'(?:<think>)?.*?</think>', '', content, flags=re.DOTALL, count=1).strip()
return content, reasoning_content
return "没有返回结果", ""