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

Merge remote-tracking branch 'upstream/debug' into feature

Browse Source
This commit is contained in:
tcmofashi
2025-03-04 18:10:08 +08:00
parent 0294ef421b f4c383caf3
commit f1a2ccfccf
3 changed files with 43 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
.github/workflows/docker-image.yml vendored
View File

@@ -34,3 +34,5 @@ jobs:
${{ secrets.DOCKERHUB_USERNAME }}/maimbot:${{ github.ref_name }} ${{ secrets.DOCKERHUB_USERNAME }}/maimbot:${{ github.ref_name }}
${{ secrets.DOCKERHUB_USERNAME }}/maimbot:latest ${{ secrets.DOCKERHUB_USERNAME }}/maimbot:latest
push: true push: true
cache-from: type=registry,ref=${{ secrets.DOCKERHUB_USERNAME }}/maimbot:latest
cache-to: type=inline

15
src/plugins/memory_system/llm_module_memory_make.py
View File

@@ -5,13 +5,13 @@ import time
from nonebot import get_driver from nonebot import get_driver
import aiohttp import aiohttp
import asyncio import asyncio
from loguru import logger
from src.plugins.chat.config import BotConfig, global_config from src.plugins.chat.config import BotConfig, global_config
driver = get_driver() driver = get_driver()
config = driver.config config = driver.config
class LLMModel: class LLMModel:
# def __init__(self, model_name="deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", **kwargs):
def __init__(self, model_name=global_config.SILICONFLOW_MODEL_V3, **kwargs): def __init__(self, model_name=global_config.SILICONFLOW_MODEL_V3, **kwargs):
self.model_name = model_name self.model_name = model_name
self.params = kwargs self.params = kwargs
@@ -21,7 +21,7 @@ class LLMModel:
if not self.api_key or not self.base_url: if not self.api_key or not self.base_url:
raise ValueError("环境变量未正确加载：SILICONFLOW_KEY 或 SILICONFLOW_BASE_URL 未设置") raise ValueError("环境变量未正确加载：SILICONFLOW_KEY 或 SILICONFLOW_BASE_URL 未设置")
print(f"API URL: {self.base_url}") # 打印 base_url 用于调试 logger.info(f"API URL: {self.base_url}") # 使用 logger 记录 base_url
async def generate_response(self, prompt: str) -> Tuple[str, str]: async def generate_response(self, prompt: str) -> Tuple[str, str]:
"""根据输入的提示生成模型的响应""" """根据输入的提示生成模型的响应"""
@@ -38,8 +38,9 @@ class LLMModel:
**self.params **self.params
} }
# 发送请求到完整的chat/completions端点 # 发送请求到完整的 chat/completions 端点
api_url = f"{self.base_url.rstrip('/')}/chat/completions" 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 max_retries = 3
base_wait_time = 15 # 基础等待时间（秒） base_wait_time = 15 # 基础等待时间（秒）
@@ -50,7 +51,7 @@ class LLMModel:
async with session.post(api_url, headers=headers, json=data) as response: async with session.post(api_url, headers=headers, json=data) as response:
if response.status == 429: if response.status == 429:
wait_time = base_wait_time * (2 ** retry) # 指数退避 wait_time = base_wait_time * (2 ** retry) # 指数退避
print(f"遇到请求限制(429)，等待{wait_time}秒后重试...") logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
continue continue
@@ -66,9 +67,11 @@ class LLMModel:
except Exception as e: except Exception as e:
if retry < max_retries - 1: # 如果还有重试机会 if retry < max_retries - 1: # 如果还有重试机会
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"[回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}") logger.error(f"[回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}")
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
else: else:
logger.error(f"请求失败: {str(e)}")
return f"请求失败: {str(e)}", "" return f"请求失败: {str(e)}", ""
return "达到最大重试次数，请求仍然失败", "" 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", ""

50
src/plugins/models/utils_model.py
View File

@@ -4,18 +4,21 @@ import requests
import time import time
from typing import Tuple, Union from typing import Tuple, Union
from nonebot import get_driver from nonebot import get_driver
from loguru import logger
from ..chat.config import global_config from ..chat.config import global_config
driver = get_driver() driver = get_driver()
config = driver.config config = driver.config
class LLM_request: class LLM_request:
def __init__(self, model ,**kwargs): 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 # 将大写的配置键转换为小写并从config中获取实际值
try: try:
self.api_key = getattr(config, model["key"]) self.api_key = getattr(config, model["key"])
self.base_url = getattr(config, model["base_url"]) self.base_url = getattr(config, model["base_url"])
except AttributeError as e: except AttributeError as e:
raise ValueError(f"配置错误：找不到对应的配置项 - {str(e)}") logger.error(f"配置错误：找不到对应的配置项 - {str(e)}")
raise ValueError(f"配置错误：找不到对应的配置项 - {str(e)}") from e
self.model_name = model["name"] self.model_name = model["name"]
self.params = kwargs self.params = kwargs
@@ -35,6 +38,7 @@ class LLM_request:
# 发送请求到完整的chat/completions端点 # 发送请求到完整的chat/completions端点
api_url = f"{self.base_url.rstrip('/')}/chat/completions" 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 max_retries = 3
base_wait_time = 15 base_wait_time = 15
@@ -45,7 +49,7 @@ class LLM_request:
async with session.post(api_url, headers=headers, json=data) as response: async with session.post(api_url, headers=headers, json=data) as response:
if response.status == 429: if response.status == 429:
wait_time = base_wait_time * (2 ** retry) # 指数退避 wait_time = base_wait_time * (2 ** retry) # 指数退避
print(f"遇到请求限制(429)，等待{wait_time}秒后重试...") logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
continue continue
@@ -61,11 +65,13 @@ class LLM_request:
except Exception as e: except Exception as e:
if retry < max_retries - 1: # 如果还有重试机会 if retry < max_retries - 1: # 如果还有重试机会
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"[回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}") logger.error(f"[回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}", exc_info=True)
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
else: else:
logger.critical(f"请求失败: {str(e)}", exc_info=True)
return 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", "" return "达到最大重试次数，请求仍然失败", ""
async def generate_response_for_image(self, prompt: str, image_base64: str) -> Tuple[str, str]: async def generate_response_for_image(self, prompt: str, image_base64: str) -> Tuple[str, str]:
@@ -100,6 +106,7 @@ class LLM_request:
# 发送请求到完整的chat/completions端点 # 发送请求到完整的chat/completions端点
api_url = f"{self.base_url.rstrip('/')}/chat/completions" 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 max_retries = 3
base_wait_time = 15 base_wait_time = 15
@@ -110,7 +117,7 @@ class LLM_request:
async with session.post(api_url, headers=headers, json=data) as response: async with session.post(api_url, headers=headers, json=data) as response:
if response.status == 429: if response.status == 429:
wait_time = base_wait_time * (2 ** retry) # 指数退避 wait_time = base_wait_time * (2 ** retry) # 指数退避
print(f"遇到请求限制(429)，等待{wait_time}秒后重试...") logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
continue continue
@@ -126,11 +133,13 @@ class LLM_request:
except Exception as e: except Exception as e:
if retry < max_retries - 1: # 如果还有重试机会 if retry < max_retries - 1: # 如果还有重试机会
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"[image回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}") logger.error(f"[image回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}", exc_info=True)
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
else: else:
logger.critical(f"请求失败: {str(e)}", exc_info=True)
return 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", "" return "达到最大重试次数，请求仍然失败", ""
def generate_response_for_image_sync(self, prompt: str, image_base64: str) -> Tuple[str, str]: def generate_response_for_image_sync(self, prompt: str, image_base64: str) -> Tuple[str, str]:
@@ -165,6 +174,7 @@ class LLM_request:
# 发送请求到完整的chat/completions端点 # 发送请求到完整的chat/completions端点
api_url = f"{self.base_url.rstrip('/')}/chat/completions" 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 max_retries = 2
base_wait_time = 6 base_wait_time = 6
@@ -174,8 +184,8 @@ class LLM_request:
response = requests.post(api_url, headers=headers, json=data, timeout=30) response = requests.post(api_url, headers=headers, json=data, timeout=30)
if response.status_code == 429: if response.status_code == 429:
wait_time = base_wait_time * (2 ** retry) # 指数退避 wait_time = base_wait_time * (2 ** retry)
print(f"遇到请求限制(429)，等待{wait_time}秒后重试...") logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
time.sleep(wait_time) time.sleep(wait_time)
continue continue
@@ -191,11 +201,13 @@ class LLM_request:
except Exception as e: except Exception as e:
if retry < max_retries - 1: # 如果还有重试机会 if retry < max_retries - 1: # 如果还有重试机会
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"[image_sync回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}") logger.error(f"[image_sync回复]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}", exc_info=True)
time.sleep(wait_time) time.sleep(wait_time)
else: else:
logger.critical(f"请求失败: {str(e)}", exc_info=True)
return 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", "" return "达到最大重试次数，请求仍然失败", ""
def get_embedding_sync(self, text: str, model: str = "BAAI/bge-m3") -> Union[list, None]: def get_embedding_sync(self, text: str, model: str = "BAAI/bge-m3") -> Union[list, None]:
@@ -220,6 +232,7 @@ class LLM_request:
} }
api_url = f"{self.base_url.rstrip('/')}/embeddings" 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 max_retries = 2
base_wait_time = 6 base_wait_time = 6
@@ -230,7 +243,7 @@ class LLM_request:
if response.status_code == 429: if response.status_code == 429:
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"遇到请求限制(429)，等待{wait_time}秒后重试...") logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
time.sleep(wait_time) time.sleep(wait_time)
continue continue
@@ -244,13 +257,13 @@ class LLM_request:
except Exception as e: except Exception as e:
if retry < max_retries - 1: if retry < max_retries - 1:
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"[embedding_sync]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}") logger.error(f"[embedding_sync]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}", exc_info=True)
time.sleep(wait_time) time.sleep(wait_time)
else: else:
print(f"embedding请求失败: {str(e)}") logger.critical(f"embedding请求失败: {str(e)}", exc_info=True)
return None return None
print("达到最大重试次数，embedding请求仍然失败") logger.error("达到最大重试次数，embedding请求仍然失败")
return None return None
async def get_embedding(self, text: str, model: str = "BAAI/bge-m3") -> Union[list, None]: async def get_embedding(self, text: str, model: str = "BAAI/bge-m3") -> Union[list, None]:
@@ -275,6 +288,7 @@ class LLM_request:
} }
api_url = f"{self.base_url.rstrip('/')}/embeddings" 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 max_retries = 3
base_wait_time = 15 base_wait_time = 15
@@ -285,7 +299,7 @@ class LLM_request:
async with session.post(api_url, headers=headers, json=data) as response: async with session.post(api_url, headers=headers, json=data) as response:
if response.status == 429: if response.status == 429:
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"遇到请求限制(429)，等待{wait_time}秒后重试...") logger.warning(f"遇到请求限制(429)，等待{wait_time}秒后重试...")
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
continue continue
@@ -299,11 +313,11 @@ class LLM_request:
except Exception as e: except Exception as e:
if retry < max_retries - 1: if retry < max_retries - 1:
wait_time = base_wait_time * (2 ** retry) wait_time = base_wait_time * (2 ** retry)
print(f"[embedding]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}") logger.error(f"[embedding]请求失败，等待{wait_time}秒后重试... 错误: {str(e)}", exc_info=True)
await asyncio.sleep(wait_time) await asyncio.sleep(wait_time)
else: else:
print(f"embedding请求失败: {str(e)}") logger.critical(f"embedding请求失败: {str(e)}", exc_info=True)
return None return None
print("达到最大重试次数，embedding请求仍然失败") logger.error("达到最大重试次数，embedding请求仍然失败")
return None return None