252 lines
9.1 KiB
Python
252 lines
9.1 KiB
Python
|
"""
|
||
|
|
MiniMax适配器
|
||
|
|
支持MiniMax大模型API
|
||
|
|
"""
|
||
|
|
import json
|
||
|
|
from datetime import datetime
|
||
|
|
from typing import List, Dict, Any, Optional, AsyncGenerator
|
||
|
|
from loguru import logger
|
||
|
|
|
||
|
|
from .base_adapter import BaseAdapter, ChatMessage, AdapterResponse
|
||
|
|
from utils.proxy_handler import get_http_client
|
||
|
|
|
||
|
|
|
||
|
|
class MiniMaxAdapter(BaseAdapter):
|
||
|
|
"""
|
||
|
|
MiniMax API适配器
|
||
|
|
支持abab系列模型
|
||
|
|
"""
|
||
|
|
|
||
|
|
DEFAULT_BASE_URL = "https://api.minimax.chat/v1"
|
||
|
|
|
||
|
|
def __init__(
|
||
|
|
self,
|
||
|
|
api_key: str,
|
||
|
|
base_url: str = "",
|
||
|
|
model: str = "abab6.5-chat",
|
||
|
|
use_proxy: bool = False,
|
||
|
|
proxy_config: Optional[Dict[str, Any]] = None,
|
||
|
|
timeout: int = 60,
|
||
|
|
**kwargs
|
||
|
|
):
|
||
|
|
super().__init__(
|
||
|
|
api_key=api_key,
|
||
|
|
base_url=base_url or self.DEFAULT_BASE_URL,
|
||
|
|
model=model,
|
||
|
|
use_proxy=use_proxy,
|
||
|
|
proxy_config=proxy_config,
|
||
|
|
timeout=timeout,
|
||
|
|
**kwargs
|
||
|
|
)
|
||
|
|
# MiniMax需要group_id
|
||
|
|
self.group_id = kwargs.get("group_id", "")
|
||
|
|
|
||
|
|
async def chat(
|
||
|
|
self,
|
||
|
|
messages: List[ChatMessage],
|
||
|
|
temperature: float = 0.7,
|
||
|
|
max_tokens: int = 2000,
|
||
|
|
**kwargs
|
||
|
|
) -> AdapterResponse:
|
||
|
|
"""发送聊天请求"""
|
||
|
|
start_time = datetime.utcnow()
|
||
|
|
|
||
|
|
try:
|
||
|
|
async with get_http_client(
|
||
|
|
use_proxy=self.use_proxy,
|
||
|
|
proxy_config=self.proxy_config,
|
||
|
|
timeout=self.timeout
|
||
|
|
) as client:
|
||
|
|
headers = {
|
||
|
|
"Authorization": f"Bearer {self.api_key}",
|
||
|
|
"Content-Type": "application/json"
|
||
|
|
}
|
||
|
|
|
||
|
|
# MiniMax使用特殊的消息格式
|
||
|
|
minimax_messages = []
|
||
|
|
bot_setting = []
|
||
|
|
|
||
|
|
for msg in messages:
|
||
|
|
if msg.role == "system":
|
||
|
|
bot_setting.append({
|
||
|
|
"bot_name": "assistant",
|
||
|
|
"content": msg.content
|
||
|
|
})
|
||
|
|
else:
|
||
|
|
minimax_messages.append({
|
||
|
|
"sender_type": "USER" if msg.role == "user" else "BOT",
|
||
|
|
"sender_name": msg.name or ("用户" if msg.role == "user" else "assistant"),
|
||
|
|
"text": msg.content
|
||
|
|
})
|
||
|
|
|
||
|
|
payload = {
|
||
|
|
"model": self.model,
|
||
|
|
"messages": minimax_messages,
|
||
|
|
"bot_setting": bot_setting if bot_setting else [{"bot_name": "assistant", "content": "你是一个有帮助的助手"}],
|
||
|
|
"temperature": temperature,
|
||
|
|
"tokens_to_generate": max_tokens,
|
||
|
|
"mask_sensitive_info": False,
|
||
|
|
**kwargs
|
||
|
|
}
|
||
|
|
|
||
|
|
url = f"{self.base_url}/text/chatcompletion_v2"
|
||
|
|
if self.group_id:
|
||
|
|
url = f"{url}?GroupId={self.group_id}"
|
||
|
|
|
||
|
|
response = await client.post(
|
||
|
|
url,
|
||
|
|
headers=headers,
|
||
|
|
json=payload
|
||
|
|
)
|
||
|
|
|
||
|
|
if response.status_code != 200:
|
||
|
|
error_text = response.text
|
||
|
|
logger.error(f"MiniMax API错误: {response.status_code} - {error_text}")
|
||
|
|
return AdapterResponse(
|
||
|
|
success=False,
|
||
|
|
error=f"API错误: {response.status_code} - {error_text}",
|
||
|
|
latency_ms=self._calculate_latency(start_time)
|
||
|
|
)
|
||
|
|
|
||
|
|
data = response.json()
|
||
|
|
|
||
|
|
# 检查API返回的错误
|
||
|
|
if data.get("base_resp", {}).get("status_code", 0) != 0:
|
||
|
|
error_msg = data.get("base_resp", {}).get("status_msg", "未知错误")
|
||
|
|
return AdapterResponse(
|
||
|
|
success=False,
|
||
|
|
error=f"API错误: {error_msg}",
|
||
|
|
latency_ms=self._calculate_latency(start_time)
|
||
|
|
)
|
||
|
|
|
||
|
|
reply = data.get("reply", "")
|
||
|
|
usage = data.get("usage", {})
|
||
|
|
|
||
|
|
return AdapterResponse(
|
||
|
|
success=True,
|
||
|
|
content=reply,
|
||
|
|
model=self.model,
|
||
|
|
finish_reason=data.get("output_sensitive", False) and "content_filter" or "stop",
|
||
|
|
prompt_tokens=usage.get("prompt_tokens", 0),
|
||
|
|
completion_tokens=usage.get("completion_tokens", 0),
|
||
|
|
total_tokens=usage.get("total_tokens", 0),
|
||
|
|
latency_ms=self._calculate_latency(start_time)
|
||
|
|
)
|
||
|
|
|
||
|
|
except Exception as e:
|
||
|
|
logger.error(f"MiniMax请求异常: {e}")
|
||
|
|
return AdapterResponse(
|
||
|
|
success=False,
|
||
|
|
error=str(e),
|
||
|
|
latency_ms=self._calculate_latency(start_time)
|
||
|
|
)
|
||
|
|
|
||
|
|
async def chat_stream(
|
||
|
|
self,
|
||
|
|
messages: List[ChatMessage],
|
||
|
|
temperature: float = 0.7,
|
||
|
|
max_tokens: int = 2000,
|
||
|
|
**kwargs
|
||
|
|
) -> AsyncGenerator[str, None]:
|
||
|
|
"""发送流式聊天请求"""
|
||
|
|
try:
|
||
|
|
async with get_http_client(
|
||
|
|
use_proxy=self.use_proxy,
|
||
|
|
proxy_config=self.proxy_config,
|
||
|
|
timeout=self.timeout
|
||
|
|
) as client:
|
||
|
|
headers = {
|
||
|
|
"Authorization": f"Bearer {self.api_key}",
|
||
|
|
"Content-Type": "application/json"
|
||
|
|
}
|
||
|
|
|
||
|
|
minimax_messages = []
|
||
|
|
bot_setting = []
|
||
|
|
|
||
|
|
for msg in messages:
|
||
|
|
if msg.role == "system":
|
||
|
|
bot_setting.append({
|
||
|
|
"bot_name": "assistant",
|
||
|
|
"content": msg.content
|
||
|
|
})
|
||
|
|
else:
|
||
|
|
minimax_messages.append({
|
||
|
|
"sender_type": "USER" if msg.role == "user" else "BOT",
|
||
|
|
"sender_name": msg.name or ("用户" if msg.role == "user" else "assistant"),
|
||
|
|
"text": msg.content
|
||
|
|
})
|
||
|
|
|
||
|
|
payload = {
|
||
|
|
"model": self.model,
|
||
|
|
"messages": minimax_messages,
|
||
|
|
"bot_setting": bot_setting if bot_setting else [{"bot_name": "assistant", "content": "你是一个有帮助的助手"}],
|
||
|
|
"temperature": temperature,
|
||
|
|
"tokens_to_generate": max_tokens,
|
||
|
|
"stream": True,
|
||
|
|
**kwargs
|
||
|
|
}
|
||
|
|
|
||
|
|
url = f"{self.base_url}/text/chatcompletion_v2"
|
||
|
|
if self.group_id:
|
||
|
|
url = f"{url}?GroupId={self.group_id}"
|
||
|
|
|
||
|
|
async with client.stream(
|
||
|
|
"POST",
|
||
|
|
url,
|
||
|
|
headers=headers,
|
||
|
|
json=payload
|
||
|
|
) as response:
|
||
|
|
async for line in response.aiter_lines():
|
||
|
|
if line.startswith("data: "):
|
||
|
|
data_str = line[6:]
|
||
|
|
if data_str.strip() == "[DONE]":
|
||
|
|
break
|
||
|
|
try:
|
||
|
|
data = json.loads(data_str)
|
||
|
|
delta = data.get("choices", [{}])[0].get("delta", {})
|
||
|
|
content = delta.get("content", "")
|
||
|
|
if content:
|
||
|
|
yield content
|
||
|
|
except json.JSONDecodeError:
|
||
|
|
continue
|
||
|
|
|
||
|
|
except Exception as e:
|
||
|
|
logger.error(f"MiniMax流式请求异常: {e}")
|
||
|
|
yield f"[错误: {str(e)}]"
|
||
|
|
|
||
|
|
async def test_connection(self) -> Dict[str, Any]:
|
||
|
|
"""测试API连接"""
|
||
|
|
start_time = datetime.utcnow()
|
||
|
|
|
||
|
|
try:
|
||
|
|
test_messages = [
|
||
|
|
ChatMessage(role="user", content="你好，请回复'OK'")
|
||
|
|
]
|
||
|
|
|
||
|
|
response = await self.chat(
|
||
|
|
messages=test_messages,
|
||
|
|
temperature=0,
|
||
|
|
max_tokens=10
|
||
|
|
)
|
||
|
|
|
||
|
|
if response.success:
|
||
|
|
return {
|
||
|
|
"success": True,
|
||
|
|
"message": "连接成功",
|
||
|
|
"model": response.model,
|
||
|
|
"latency_ms": response.latency_ms
|
||
|
|
}
|
||
|
|
else:
|
||
|
|
return {
|
||
|
|
"success": False,
|
||
|
|
"message": response.error,
|
||
|
|
"latency_ms": response.latency_ms
|
||
|
|
}
|
||
|
|
|
||
|
|
except Exception as e:
|
||
|
|
return {
|
||
|
|
"success": False,
|
||
|
|
"message": str(e),
|
||
|
|
"latency_ms": self._calculate_latency(start_time)
|
||
|
|
}
|