mirror of
https://github.com/csunny/DB-GPT.git
synced 2025-09-15 05:59:59 +00:00
feat(agent): Multi agent sdk (#976)
Co-authored-by: xtyuns <xtyuns@163.com> Co-authored-by: Fangyin Cheng <staneyffer@gmail.com> Co-authored-by: csunny <cfqsunny@163.com> Co-authored-by: qidanrui <qidanrui@gmail.com>
This commit is contained in:
明天
GitHub
338
dbgpt/serve/agent/agents/controller.py
Normal file
338
dbgpt/serve/agent/agents/controller.py
Normal file
@@ -0,0 +1,338 @@
|
||||
import logging
|
||||
import json
|
||||
import asyncio
|
||||
import uuid
|
||||
from collections import defaultdict
|
||||
from typing import Any, Callable, Dict, List, Literal, Optional, Tuple, Type, Union
|
||||
from fastapi import (
|
||||
APIRouter,
|
||||
Body,
|
||||
UploadFile,
|
||||
File,
|
||||
)
|
||||
from fastapi.responses import StreamingResponse
|
||||
|
||||
from abc import ABC
|
||||
from typing import List
|
||||
from dbgpt.core.awel import BaseOperator, SimpleCallDataInputSource, InputOperator, DAG
|
||||
from dbgpt.model.operator.model_operator import ModelOperator, ModelStreamOperator
|
||||
from dbgpt.app.openapi.api_view_model import Result, ConversationVo
|
||||
from dbgpt.util.json_utils import EnhancedJSONEncoder
|
||||
from dbgpt.serve.agent.model import (
|
||||
PluginHubParam,
|
||||
PagenationFilter,
|
||||
PagenationResult,
|
||||
PluginHubFilter,
|
||||
)
|
||||
|
||||
from dbgpt.agent.common.schema import Status
|
||||
from dbgpt.agent.agents.agents_mange import AgentsMange
|
||||
from dbgpt.agent.agents.planner_agent import PlannerAgent
|
||||
from dbgpt.agent.agents.user_proxy_agent import UserProxyAgent
|
||||
from dbgpt.agent.agents.plan_group_chat import PlanChat, PlanChatManager
|
||||
from dbgpt.agent.agents.agent import AgentContext
|
||||
from dbgpt.agent.memory.gpts_memory import GptsMemory
|
||||
|
||||
from .db_gpts_memory import MetaDbGptsPlansMemory, MetaDbGptsMessageMemory
|
||||
|
||||
from ..db.gpts_mange_db import GptsInstanceDao, GptsInstanceEntity
|
||||
from ..db.gpts_conversations_db import GptsConversationsDao, GptsConversationsEntity
|
||||
|
||||
from .dbgpts import DbGptsCompletion, DbGptsTaskStep, DbGptsMessage, DbGptsInstance
|
||||
from dbgpt.component import BaseComponent, ComponentType, SystemApp
|
||||
from dbgpt.agent.agents.agents_mange import agent_mange
|
||||
from dbgpt._private.config import Config
|
||||
from dbgpt.model.cluster.controller.controller import BaseModelController
|
||||
from dbgpt.agent.memory.gpts_memory import GptsMessage
|
||||
from dbgpt.model.cluster import WorkerManager, WorkerManagerFactory
|
||||
|
||||
from dbgpt.model.cluster.client import DefaultLLMClient
|
||||
|
||||
CFG = Config()
|
||||
|
||||
import asyncio
|
||||
|
||||
router = APIRouter()
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class MultiAgents(BaseComponent, ABC):
|
||||
name = ComponentType.MULTI_AGENTS
|
||||
|
||||
def init_app(self, system_app: SystemApp):
|
||||
system_app.app.include_router(router, prefix="/api", tags=["Multi-Agents"])
|
||||
|
||||
def __init__(self):
|
||||
self.gpts_intance = GptsInstanceDao()
|
||||
self.gpts_conversations = GptsConversationsDao()
|
||||
self.memory = GptsMemory(
|
||||
plans_memory=MetaDbGptsPlansMemory(),
|
||||
message_memory=MetaDbGptsMessageMemory(),
|
||||
)
|
||||
|
||||
def gpts_create(self, entity: GptsInstanceEntity):
|
||||
self.gpts_intance.add(entity)
|
||||
|
||||
async def plan_chat(
|
||||
self,
|
||||
name: str,
|
||||
user_query: str,
|
||||
conv_id: str,
|
||||
user_code: str = None,
|
||||
sys_code: str = None,
|
||||
):
|
||||
gpts_instance: GptsInstanceEntity = self.gpts_intance.get_by_name(name)
|
||||
if gpts_instance is None:
|
||||
raise ValueError(f"can't find dbgpts!{name}")
|
||||
agents_names = json.loads(gpts_instance.gpts_agents)
|
||||
llm_models_priority = json.loads(gpts_instance.gpts_models)
|
||||
resource_db = (
|
||||
json.loads(gpts_instance.resource_db) if gpts_instance.resource_db else None
|
||||
)
|
||||
resource_knowledge = (
|
||||
json.loads(gpts_instance.resource_knowledge)
|
||||
if gpts_instance.resource_knowledge
|
||||
else None
|
||||
)
|
||||
resource_internet = (
|
||||
json.loads(gpts_instance.resource_internet)
|
||||
if gpts_instance.resource_internet
|
||||
else None
|
||||
)
|
||||
|
||||
### init chat param
|
||||
worker_manager = CFG.SYSTEM_APP.get_component(
|
||||
ComponentType.WORKER_MANAGER_FACTORY, WorkerManagerFactory
|
||||
).create()
|
||||
llm_task = DefaultLLMClient(worker_manager)
|
||||
context: AgentContext = AgentContext(conv_id=conv_id, llm_provider=llm_task)
|
||||
context.gpts_name = gpts_instance.gpts_name
|
||||
context.resource_db = resource_db
|
||||
context.resource_internet = resource_internet
|
||||
context.resource_knowledge = resource_knowledge
|
||||
context.agents = agents_names
|
||||
|
||||
context.llm_models = await llm_task.models()
|
||||
context.model_priority = llm_models_priority
|
||||
|
||||
agent_map = defaultdict()
|
||||
|
||||
### default plan excute mode
|
||||
agents = []
|
||||
for name in agents_names:
|
||||
cls = agent_mange.get_by_name(name)
|
||||
agent = cls(
|
||||
agent_context=context,
|
||||
memory=self.memory,
|
||||
)
|
||||
agents.append(agent)
|
||||
agent_map[name] = agent
|
||||
|
||||
groupchat = PlanChat(agents=agents, messages=[], max_round=50)
|
||||
planner = PlannerAgent(
|
||||
agent_context=context,
|
||||
memory=self.memory,
|
||||
plan_chat=groupchat,
|
||||
)
|
||||
agent_map[planner.name] = planner
|
||||
|
||||
manager = PlanChatManager(
|
||||
agent_context=context,
|
||||
memory=self.memory,
|
||||
plan_chat=groupchat,
|
||||
planner=planner,
|
||||
)
|
||||
agent_map[manager.name] = manager
|
||||
|
||||
user_proxy = UserProxyAgent(memory=self.memory, agent_context=context)
|
||||
agent_map[user_proxy.name] = user_proxy
|
||||
|
||||
gpts_conversation = self.gpts_conversations.get_by_conv_id(conv_id)
|
||||
if gpts_conversation is None:
|
||||
self.gpts_conversations.add(
|
||||
GptsConversationsEntity(
|
||||
conv_id=conv_id,
|
||||
user_goal=user_query,
|
||||
gpts_name=gpts_instance.gpts_name,
|
||||
state=Status.RUNNING.value,
|
||||
max_auto_reply_round=context.max_chat_round,
|
||||
auto_reply_count=0,
|
||||
user_code=user_code,
|
||||
sys_code=sys_code,
|
||||
)
|
||||
)
|
||||
|
||||
## dbgpts conversation save
|
||||
try:
|
||||
await user_proxy.a_initiate_chat(
|
||||
recipient=manager,
|
||||
message=user_query,
|
||||
memory=self.memory,
|
||||
)
|
||||
except Exception as e:
|
||||
logger.error(f"chat abnormal termination！{str(e)}", e)
|
||||
self.gpts_conversations.update(conv_id, Status.FAILED.value)
|
||||
|
||||
else:
|
||||
# retry chat
|
||||
self.gpts_conversations.update(conv_id, Status.RUNNING.value)
|
||||
try:
|
||||
await user_proxy.a_retry_chat(
|
||||
recipient=manager,
|
||||
agent_map=agent_map,
|
||||
memory=self.memory,
|
||||
)
|
||||
except Exception as e:
|
||||
logger.error(f"chat abnormal termination！{str(e)}", e)
|
||||
self.gpts_conversations.update(conv_id, Status.FAILED.value)
|
||||
|
||||
self.gpts_conversations.update(conv_id, Status.COMPLETE.value)
|
||||
return conv_id
|
||||
|
||||
async def chat_completions(
|
||||
self, conv_id: str, user_code: str = None, system_app: str = None
|
||||
):
|
||||
is_complete = False
|
||||
while True:
|
||||
gpts_conv = self.gpts_conversations.get_by_conv_id(conv_id)
|
||||
if gpts_conv:
|
||||
is_complete = (
|
||||
True
|
||||
if gpts_conv.state
|
||||
in [
|
||||
Status.COMPLETE.value,
|
||||
Status.WAITING.value,
|
||||
Status.FAILED.value,
|
||||
]
|
||||
else False
|
||||
)
|
||||
yield await self.memory.one_plan_chat_competions(conv_id)
|
||||
if is_complete:
|
||||
return
|
||||
else:
|
||||
await asyncio.sleep(5)
|
||||
|
||||
async def stable_message(
|
||||
self, conv_id: str, user_code: str = None, system_app: str = None
|
||||
):
|
||||
gpts_conv = self.gpts_conversations.get_by_conv_id(conv_id)
|
||||
if gpts_conv:
|
||||
is_complete = (
|
||||
True
|
||||
if gpts_conv.state
|
||||
in [Status.COMPLETE.value, Status.WAITING.value, Status.FAILED.value]
|
||||
else False
|
||||
)
|
||||
if is_complete:
|
||||
return await self.self.memory.one_plan_chat_competions(conv_id)
|
||||
else:
|
||||
raise ValueError(
|
||||
"The conversation has not been completed yet, so we cannot directly obtain information."
|
||||
)
|
||||
else:
|
||||
raise ValueError("No conversation record found!")
|
||||
|
||||
def gpts_conv_list(self, user_code: str = None, system_app: str = None):
|
||||
return self.gpts_conversations.get_convs(user_code, system_app)
|
||||
|
||||
|
||||
multi_agents = MultiAgents()
|
||||
|
||||
|
||||
@router.post("/v1/dbbgpts/agents/list", response_model=Result[str])
|
||||
async def agents_list():
|
||||
logger.info("agents_list!")
|
||||
try:
|
||||
agents = agent_mange.all_agents()
|
||||
return Result.succ(agents)
|
||||
except Exception as e:
|
||||
return Result.failed(code="E30001", msg=str(e))
|
||||
|
||||
|
||||
@router.post("/v1/dbbgpts/create", response_model=Result[str])
|
||||
async def create_dbgpts(gpts_instance: DbGptsInstance = Body()):
|
||||
logger.info(f"create_dbgpts:{gpts_instance}")
|
||||
try:
|
||||
multi_agents.gpts_create(
|
||||
GptsInstanceEntity(
|
||||
gpts_name=gpts_instance.gpts_name,
|
||||
gpts_describe=gpts_instance.gpts_describe,
|
||||
resource_db=json.dumps(gpts_instance.resource_db.to_dict()),
|
||||
resource_internet=json.dumps(gpts_instance.resource_internet.to_dict()),
|
||||
resource_knowledge=json.dumps(
|
||||
gpts_instance.resource_knowledge.to_dict()
|
||||
),
|
||||
gpts_agents=json.dumps(gpts_instance.gpts_agents),
|
||||
gpts_models=json.dumps(gpts_instance.gpts_models),
|
||||
language=gpts_instance.language,
|
||||
user_code=gpts_instance.user_code,
|
||||
sys_code=gpts_instance.sys_code,
|
||||
)
|
||||
)
|
||||
return Result.succ(None)
|
||||
except Exception as e:
|
||||
logger.error(f"create_dbgpts failed:{str(e)}")
|
||||
return Result.failed(msg=str(e), code="E300002")
|
||||
|
||||
|
||||
async def stream_generator(conv_id: str):
|
||||
async for chunk in multi_agents.chat_completions(conv_id):
|
||||
if chunk:
|
||||
yield f"data: {chunk}\n\n"
|
||||
|
||||
|
||||
@router.post("/v1/dbbgpts/chat/plan/completions", response_model=Result[str])
|
||||
async def dgpts_completions(
|
||||
gpts_name: str,
|
||||
user_query: str,
|
||||
conv_id: str = None,
|
||||
user_code: str = None,
|
||||
sys_code: str = None,
|
||||
):
|
||||
logger.info(f"dgpts_completions:{gpts_name},{user_query},{conv_id}")
|
||||
if conv_id is None:
|
||||
conv_id = str(uuid.uuid1())
|
||||
asyncio.create_task(
|
||||
multi_agents.plan_chat(gpts_name, user_query, conv_id, user_code, sys_code)
|
||||
)
|
||||
|
||||
headers = {
|
||||
"Content-Type": "text/event-stream",
|
||||
"Cache-Control": "no-cache",
|
||||
"Connection": "keep-alive",
|
||||
"Transfer-Encoding": "chunked",
|
||||
}
|
||||
return StreamingResponse(
|
||||
stream_generator(conv_id),
|
||||
headers=headers,
|
||||
media_type="text/plain",
|
||||
)
|
||||
|
||||
|
||||
@router.post("/v1/dbbgpts/plan/chat/cancel", response_model=Result[str])
|
||||
async def dgpts_plan_chat_cancel(
|
||||
conv_id: str = None, user_code: str = None, sys_code: str = None
|
||||
):
|
||||
pass
|
||||
|
||||
|
||||
@router.get("/v1/dbbgpts/chat/plan/messages", response_model=Result[str])
|
||||
async def plan_chat_messages(conv_id: str, user_code: str = None, sys_code: str = None):
|
||||
logger.info(f"plan_chat_messages:{conv_id},{user_code},{sys_code}")
|
||||
|
||||
headers = {
|
||||
"Content-Type": "text/event-stream",
|
||||
"Cache-Control": "no-cache",
|
||||
"Connection": "keep-alive",
|
||||
"Transfer-Encoding": "chunked",
|
||||
}
|
||||
return StreamingResponse(
|
||||
stream_generator(conv_id),
|
||||
headers=headers,
|
||||
media_type="text/plain",
|
||||
)
|
||||
|
||||
|
||||
@router.post("/v1/dbbgpts/chat/feedback", response_model=Result[str])
|
||||
async def dgpts_chat_feedback(filter: PagenationFilter[PluginHubFilter] = Body()):
|
||||
pass
|
||||
Reference in New Issue
Block a user