sj
1824f2d14b
# Conflicts: # scripts/legacy_startup/restart_with_prompts.sh # scripts/legacy_startup/start-webui.sh # scripts/legacy_startup/start_celery.sh # scripts/legacy_startup/start_celery_7880.sh # scripts/legacy_startup/start_celery_port9000.sh
111 lines
4.0 KiB
Python
Executable File
111 lines
4.0 KiB
Python
Executable File
"""DeepSeek Provider - 使用 DeepSeek OpenAI-compatible API"""
|
|
import asyncio
|
|
from typing import AsyncIterator, Iterator
|
|
|
|
from openai import OpenAI
|
|
|
|
from ..base import BaseLLMProvider, LLMMessage, LLMResponse, LLMToolCall
|
|
|
|
|
|
class DeepSeekProvider(BaseLLMProvider):
|
|
"""DeepSeek模型Provider - 通过 DeepSeek OpenAI-compatible API 调用
|
|
|
|
支持的模型:
|
|
- deepseek-chat: 通用对话模型
|
|
- deepseek-reasoner: 推理模型（适合复杂分析）
|
|
"""
|
|
provider_name = "deepseek"
|
|
DEFAULT_BASE_URL = "https://api.deepseek.com"
|
|
|
|
# 百炼平台模型名称映射
|
|
MODEL_MAPPING = {
|
|
"deepseek-chat": "deepseek-chat",
|
|
"deepseek-reasoner": "deepseek-reasoner",
|
|
"deepseek-v3": "deepseek-chat",
|
|
"deepseek-r1": "deepseek-reasoner",
|
|
"deepseek-v4-flash": "deepseek-v4-flash",
|
|
"deepseek-v4-pro": "deepseek-v4-pro",
|
|
}
|
|
|
|
def __init__(
|
|
self,
|
|
api_key: str,
|
|
model: str = "deepseek-chat",
|
|
temperature: float = 0.7,
|
|
max_tokens: int = 4096,
|
|
api_base: str = None,
|
|
):
|
|
# 映射模型名称
|
|
deepseek_model = self.MODEL_MAPPING.get(model, model)
|
|
base_url = api_base or self.DEFAULT_BASE_URL
|
|
|
|
super().__init__(
|
|
api_key=api_key,
|
|
api_base=base_url,
|
|
model=deepseek_model,
|
|
temperature=temperature,
|
|
max_tokens=max_tokens,
|
|
)
|
|
self.client = OpenAI(api_key=api_key, base_url=base_url)
|
|
|
|
def call(self, messages: list[LLMMessage], **kwargs) -> LLMResponse:
|
|
request_kwargs = {
|
|
"model": self.model,
|
|
"messages": self._build_messages(messages),
|
|
"temperature": kwargs.get("temperature", self.temperature),
|
|
"max_tokens": kwargs.get("max_tokens", self.max_tokens),
|
|
}
|
|
tools = kwargs.get("tools")
|
|
if tools is not None:
|
|
request_kwargs["tools"] = tools
|
|
if tools is not None or kwargs.get("disable_thinking", False):
|
|
request_kwargs["extra_body"] = {"thinking": {"type": "disabled"}}
|
|
tool_choice = kwargs.get("tool_choice")
|
|
if tool_choice is not None:
|
|
request_kwargs["tool_choice"] = tool_choice
|
|
response_format = kwargs.get("response_format")
|
|
if response_format is not None:
|
|
request_kwargs["response_format"] = response_format
|
|
|
|
response = self.client.chat.completions.create(**request_kwargs)
|
|
message = response.choices[0].message
|
|
tool_calls = [
|
|
LLMToolCall(
|
|
id=tool_call.id,
|
|
name=tool_call.function.name,
|
|
arguments=tool_call.function.arguments,
|
|
)
|
|
for tool_call in (message.tool_calls or [])
|
|
]
|
|
return LLMResponse(
|
|
content=message.content or "",
|
|
model=self.model,
|
|
provider=self.provider_name,
|
|
usage={
|
|
"prompt_tokens": response.usage.prompt_tokens,
|
|
"completion_tokens": response.usage.completion_tokens,
|
|
"total_tokens": response.usage.total_tokens,
|
|
},
|
|
finish_reason=response.choices[0].finish_reason,
|
|
tool_calls=tool_calls,
|
|
)
|
|
|
|
async def acall(self, messages: list[LLMMessage], **kwargs) -> LLMResponse:
|
|
return await asyncio.to_thread(self.call, messages, **kwargs)
|
|
|
|
def stream(self, messages: list[LLMMessage], **kwargs) -> Iterator[str]:
|
|
response = self.client.chat.completions.create(
|
|
model=self.model,
|
|
messages=self._build_messages(messages),
|
|
temperature=kwargs.get("temperature", self.temperature),
|
|
max_tokens=kwargs.get("max_tokens", self.max_tokens),
|
|
stream=True,
|
|
)
|
|
for chunk in response:
|
|
if chunk.choices[0].delta.content:
|
|
yield chunk.choices[0].delta.content
|
|
|
|
async def astream(self, messages: list[LLMMessage], **kwargs) -> AsyncIterator[str]:
|
|
for content in self.stream(messages, **kwargs):
|
|
yield content
|