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

Merge pull request #52 from SaigyoujiYusora/debug

Browse Source 
增加<think>标签处理
This commit is contained in:
SengokuCola
2025-03-04 23:36:35 +08:00
committed by GitHub
parent 57cde4de24 ac9bd7af63
commit e74e128874
1 changed files with 64 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files

108
src/plugins/models/utils_model.py
View File

@@ -2,6 +2,7 @@ import aiohttp
import asyncio
import requests
import time
import re
from typing import Tuple, Union
from nonebot import get_driver
from loguru import logger
@@ -11,6 +12,7 @@ from ..chat.utils_image import compress_base64_image_by_scale
driver = get_driver()
config = driver.config
class LLM_request:
def __init__(self, model, **kwargs):
# 将大写的配置键转换为小写并从config中获取实际值
@@ -29,21 +31,21 @@ class LLM_request:
"Authorization": f"Bearer {self.api_key}",
"Content-Type": "application/json"
}
# 构建请求体
data = {
"model": self.model_name,
"messages": [{"role": "user", "content": prompt}],
**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"发送请求到URL: {api_url}") # 记录请求的URL
max_retries = 3
base_wait_time = 15
for retry in range(max_retries):
try:
async with aiohttp.ClientSession() as session:
@@ -53,16 +55,16 @@ 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", "")
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)
@@ -71,7 +73,7 @@ class LLM_request:
else:
logger.critical(f"请求失败: {str(e)}", exc_info=True)
raise RuntimeError(f"API请求失败: {str(e)}")
logger.error("达到最大重试次数，请求仍然失败")
raise RuntimeError("达到最大重试次数，API请求仍然失败")
@@ -81,7 +83,7 @@ class LLM_request:
"Authorization": f"Bearer {self.api_key}",
"Content-Type": "application/json"
}
# 构建请求体
def build_request_data(img_base64: str):
return {
@@ -106,14 +108,17 @@ 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"
logger.info(f"发送请求到URL: {api_url}") # 记录请求的URL
max_retries = 3
base_wait_time = 15
current_image_base64 = image_base64
for retry in range(max_retries):
try:
data = build_request_data(current_image_base64)
@@ -124,20 +129,28 @@ 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
elif response.status == 413:
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()
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).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)
@@ -146,7 +159,7 @@ class LLM_request:
else:
logger.critical(f"请求失败: {str(e)}", exc_info=True)
raise RuntimeError(f"API请求失败: {str(e)}")
logger.error("达到最大重试次数，请求仍然失败")
raise RuntimeError("达到最大重试次数，API请求仍然失败")
@@ -156,7 +169,7 @@ class LLM_request:
"Authorization": f"Bearer {self.api_key}",
"Content-Type": "application/json"
}
# 构建请求体
data = {
"model": self.model_name,
@@ -179,33 +192,40 @@ 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"
logger.info(f"发送请求到URL: {api_url}") # 记录请求的URL
max_retries = 2
base_wait_time = 6
for retry in range(max_retries):
try:
response = requests.post(api_url, headers=headers, json=data, timeout=30)
if response.status_code == 429:
wait_time = base_wait_time * (2 ** retry)
logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
time.sleep(wait_time)
continue
response.raise_for_status() # 检查其他响应状态
result = 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).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)
@@ -214,7 +234,7 @@ class LLM_request:
else:
logger.critical(f"请求失败: {str(e)}", exc_info=True)
raise RuntimeError(f"API请求失败: {str(e)}")
logger.error("达到最大重试次数，请求仍然失败")
raise RuntimeError("达到最大重试次数，API请求仍然失败")
@@ -232,36 +252,36 @@ class LLM_request:
"Authorization": f"Bearer {self.api_key}",
"Content-Type": "application/json"
}
data = {
"model": model,
"input": text,
"encoding_format": "float"
}
api_url = f"{self.base_url.rstrip('/')}/embeddings"
logger.info(f"发送请求到URL: {api_url}") # 记录请求的URL
max_retries = 2
base_wait_time = 6
for retry in range(max_retries):
try:
response = requests.post(api_url, headers=headers, json=data, timeout=30)
if response.status_code == 429:
wait_time = base_wait_time * (2 ** retry)
logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
time.sleep(wait_time)
continue
response.raise_for_status()
result = response.json()
if 'data' in result and len(result['data']) > 0:
return result['data'][0]['embedding']
return None
except Exception as e:
if retry < max_retries - 1:
wait_time = base_wait_time * (2 ** retry)
@@ -270,7 +290,7 @@ class LLM_request:
else:
logger.critical(f"embedding请求失败: {str(e)}", exc_info=True)
return None
logger.error("达到最大重试次数，embedding请求仍然失败")
return None
@@ -288,19 +308,19 @@ class LLM_request:
"Authorization": f"Bearer {self.api_key}",
"Content-Type": "application/json"
}
data = {
"model": model,
"input": text,
"encoding_format": "float"
}
api_url = f"{self.base_url.rstrip('/')}/embeddings"
logger.info(f"发送请求到URL: {api_url}") # 记录请求的URL
max_retries = 3
base_wait_time = 15
for retry in range(max_retries):
try:
async with aiohttp.ClientSession() as session:
@@ -310,14 +330,14 @@ 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()
result = await response.json()
if 'data' in result and len(result['data']) > 0:
return result['data'][0]['embedding']
return None
except Exception as e:
if retry < max_retries - 1:
wait_time = base_wait_time * (2 ** retry)
@@ -326,6 +346,6 @@ class LLM_request:
else:
logger.critical(f"embedding请求失败: {str(e)}", exc_info=True)
return None
logger.error("达到最大重试次数，embedding请求仍然失败")
return None
return None