Skip to content

Latest commit

 

History

History
269 lines (214 loc) · 11.8 KB

File metadata and controls

269 lines (214 loc) · 11.8 KB

for rýchly prehľad.

Nasledujúci úryvok v Pythone demonštruje jednoduchého plánovacieho agenta, ktorý rozkladá cieľ na podúlohy a generuje štruktúrovaný plán:

from pydantic import BaseModel
from enum import Enum
from typing import List, Optional, Union
import json
import os
from typing import Optional
from pprint import pprint
from autogen_core.models import UserMessage, SystemMessage, AssistantMessage
from autogen_ext.models.azure import AzureAIChatCompletionClient
from azure.core.credentials import AzureKeyCredential

class AgentEnum(str, Enum):
    FlightBooking = "flight_booking"
    HotelBooking = "hotel_booking"
    CarRental = "car_rental"
    ActivitiesBooking = "activities_booking"
    DestinationInfo = "destination_info"
    DefaultAgent = "default_agent"
    GroupChatManager = "group_chat_manager"

# Travel SubTask Model
class TravelSubTask(BaseModel):
    task_details: str
    assigned_agent: AgentEnum  # we want to assign the task to the agent

class TravelPlan(BaseModel):
    main_task: str
    subtasks: List[TravelSubTask]
    is_greeting: bool

client = AzureAIChatCompletionClient(
    model="gpt-4o-mini",
    endpoint="https://models.inference.ai.azure.com",
    # To authenticate with the model you will need to generate a personal access token (PAT) in your GitHub settings.
    # Create your PAT token by following instructions here: https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/managing-your-personal-access-tokens
    credential=AzureKeyCredential(os.environ["GITHUB_TOKEN"]),
    model_info={
        "json_output": False,
        "function_calling": True,
        "vision": True,
        "family": "unknown",
    },
)

# Define the user message
messages = [
    SystemMessage(content="""You are an planner agent.
    Your job is to decide which agents to run based on the user's request.
                      Provide your response in JSON format with the following structure:
{'main_task': 'Plan a family trip from Singapore to Melbourne.',
 'subtasks': [{'assigned_agent': 'flight_booking',
               'task_details': 'Book round-trip flights from Singapore to '
                               'Melbourne.'}
    Below are the available agents specialised in different tasks:
    - FlightBooking: For booking flights and providing flight information
    - HotelBooking: For booking hotels and providing hotel information
    - CarRental: For booking cars and providing car rental information
    - ActivitiesBooking: For booking activities and providing activity information
    - DestinationInfo: For providing information about destinations
    - DefaultAgent: For handling general requests""", source="system"),
    UserMessage(
        content="Create a travel plan for a family of 2 kids from Singapore to Melboune", source="user"),
]

response = await client.create(messages=messages, extra_create_args={"response_format": 'json_object'})

response_content: Optional[str] = response.content if isinstance(
    response.content, str) else None
if response_content is None:
    raise ValueError("Response content is not a valid JSON string" )

pprint(json.loads(response_content))

# # Ensure the response content is a valid JSON string before loading it
# response_content: Optional[str] = response.content if isinstance(
#     response.content, str) else None
# if response_content is None:
#     raise ValueError("Response content is not a valid JSON string")

# # Print the response content after loading it as JSON
# pprint(json.loads(response_content))

# Validate the response content with the MathReasoning model
# TravelPlan.model_validate(json.loads(response_content))

Plánovací agent s multi-agentnou orchestráciou

V tomto príklade Semantic Router Agent prijíma požiadavku používateľa (napr. „Potrebujem plán hotela na moju cestu.“).

Plánovač potom:

  • Prijme plán hotela: Plánovač vezme správu používateľa a na základe systémového promptu (vrátane dostupných detailov agentov) vygeneruje štruktúrovaný cestovný plán.
  • Vypíše agentov a ich nástroje: Registr agentov obsahuje zoznam agentov (napr. pre lety, hotely, požičovne áut a aktivity) spolu s funkciami alebo nástrojmi, ktoré ponúkajú.
  • Smeruje plán príslušným agentom: V závislosti od počtu podúloh plánovač buď pošle správu priamo vyhradenému agentovi (v prípade jednoulohových scenárov), alebo koordinuje cez správcu skupinového chatu pre spoluprácu viacerých agentov.
  • Zhrnie výsledok: Nakoniec plánovač zhrnie vygenerovaný plán pre prehľadnosť. Nasledujúci príklad kódu v Pythone ilustruje tieto kroky:
from pydantic import BaseModel

from enum import Enum
from typing import List, Optional, Union

class AgentEnum(str, Enum):
    FlightBooking = "flight_booking"
    HotelBooking = "hotel_booking"
    CarRental = "car_rental"
    ActivitiesBooking = "activities_booking"
    DestinationInfo = "destination_info"
    DefaultAgent = "default_agent"
    GroupChatManager = "group_chat_manager"

# Travel SubTask Model

class TravelSubTask(BaseModel):
    task_details: str
    assigned_agent: AgentEnum # we want to assign the task to the agent

class TravelPlan(BaseModel):
    main_task: str
    subtasks: List[TravelSubTask]
    is_greeting: bool
import json
import os
from typing import Optional

from autogen_core.models import UserMessage, SystemMessage, AssistantMessage
from autogen_ext.models.openai import AzureOpenAIChatCompletionClient

# Create the client with type-checked environment variables

client = AzureOpenAIChatCompletionClient(
    azure_deployment=os.getenv("AZURE_OPENAI_DEPLOYMENT_NAME"),
    model=os.getenv("AZURE_OPENAI_DEPLOYMENT_NAME"),
    api_version=os.getenv("AZURE_OPENAI_API_VERSION"),
    azure_endpoint=os.getenv("AZURE_OPENAI_ENDPOINT"),
    api_key=os.getenv("AZURE_OPENAI_API_KEY"),
)

from pprint import pprint

# Define the user message

messages = [
    SystemMessage(content="""You are an planner agent.
    Your job is to decide which agents to run based on the user's request.
    Below are the available agents specialized in different tasks:
    - FlightBooking: For booking flights and providing flight information
    - HotelBooking: For booking hotels and providing hotel information
    - CarRental: For booking cars and providing car rental information
    - ActivitiesBooking: For booking activities and providing activity information
    - DestinationInfo: For providing information about destinations
    - DefaultAgent: For handling general requests""", source="system"),
    UserMessage(content="Create a travel plan for a family of 2 kids from Singapore to Melbourne", source="user"),
]

response = await client.create(messages=messages, extra_create_args={"response_format": TravelPlan})

# Ensure the response content is a valid JSON string before loading it

response_content: Optional[str] = response.content if isinstance(response.content, str) else None
if response_content is None:
    raise ValueError("Response content is not a valid JSON string")

# Print the response content after loading it as JSON

pprint(json.loads(response_content))

Nižšie je výstup z predchádzajúceho kódu, ktorý môžete použiť na smerovanie k assigned_agent a zhrnutie cestovného plánu pre koncového používateľa.

{
    "is_greeting": "False",
    "main_task": "Plan a family trip from Singapore to Melbourne.",
    "subtasks": [
        {
            "assigned_agent": "flight_booking",
            "task_details": "Book round-trip flights from Singapore to Melbourne."
        },
        {
            "assigned_agent": "hotel_booking",
            "task_details": "Find family-friendly hotels in Melbourne."
        },
        {
            "assigned_agent": "car_rental",
            "task_details": "Arrange a car rental suitable for a family of four in Melbourne."
        },
        {
            "assigned_agent": "activities_booking",
            "task_details": "List family-friendly activities in Melbourne."
        },
        {
            "assigned_agent": "destination_info",
            "task_details": "Provide information about Melbourne as a travel destination."
        }
    ]
}

Príklad notebooku s predchádzajúcim kódom je dostupný tu.

Iteratívne plánovanie

Niektoré úlohy vyžadujú spätnú väzbu alebo preplánovanie, kde výsledok jednej podúlohy ovplyvňuje ďalšiu. Napríklad, ak agent objaví neočakávaný formát dát pri rezervácii letov, môže potrebovať upraviť svoju stratégiu pred pokračovaním k rezervácii hotela.

Okrem toho spätná väzba používateľa (napr. človek rozhodne, že uprednostňuje skorší let) môže spustiť čiastočné preplánovanie. Tento dynamický, iteratívny prístup zabezpečuje, že konečné riešenie zodpovedá reálnym obmedzeniam a meniacim sa preferenciám používateľa.

napr. ukážkový kód

from autogen_core.models import UserMessage, SystemMessage, AssistantMessage
#.. same as previous code and pass on the user history, current plan
messages = [
    SystemMessage(content="""You are a planner agent to optimize the
    Your job is to decide which agents to run based on the user's request.
    Below are the available agents specialized in different tasks:
    - FlightBooking: For booking flights and providing flight information
    - HotelBooking: For booking hotels and providing hotel information
    - CarRental: For booking cars and providing car rental information
    - ActivitiesBooking: For booking activities and providing activity information
    - DestinationInfo: For providing information about destinations
    - DefaultAgent: For handling general requests""", source="system"),
    UserMessage(content="Create a travel plan for a family of 2 kids from Singapore to Melbourne", source="user"),
    AssistantMessage(content=f"Previous travel plan - {TravelPlan}", source="assistant")
]
# .. re-plan and send the tasks to respective agents

Pre komplexnejšie plánovanie odporúčame Magnetic One

pre riešenie zložitých úloh.

Zhrnutie

V tomto článku sme si ukázali príklad, ako môžeme vytvoriť plánovač, ktorý dynamicky vyberá dostupných definovaných agentov. Výstup plánovača rozkladá úlohy a priraďuje agentov, aby mohli byť vykonané. Predpokladá sa, že agenti majú prístup k funkciám/nástrojom potrebným na vykonanie úlohy. Okrem agentov môžete zahrnúť aj ďalšie vzory ako reflexia, sumarizátor a round robin chat pre ďalšie prispôsobenie.

Dodatočné zdroje

  • AutoGen Magnetic One – Generalistický multi-agentný systém na riešenie zložitých úloh, ktorý dosiahol pôsobivé výsledky na viacerých náročných agentných benchmarkoch. Referencia:

. V tejto implementácii orchestrátor vytvára plán špecifický pre úlohu a deleguje tieto úlohy dostupným agentom. Okrem plánovania orchestrátor tiež využíva mechanizmus sledovania na monitorovanie priebehu úlohy a podľa potreby preplánováva.

Predchádzajúca lekcia

Building Trustworthy AI Agents

Nasledujúca lekcia

Multi-Agent Design Pattern

Vyhlásenie o zodpovednosti:
Tento dokument bol preložený pomocou AI prekladateľskej služby Co-op Translator. Hoci sa snažíme o presnosť, prosím, majte na pamäti, že automatizované preklady môžu obsahovať chyby alebo nepresnosti. Originálny dokument v jeho pôvodnom jazyku by mal byť považovaný za autoritatívny zdroj. Pre kritické informácie sa odporúča profesionálny ľudský preklad. Nie sme zodpovední za akékoľvek nedorozumenia alebo nesprávne interpretácie vyplývajúce z použitia tohto prekladu.