lagent
v0.5.0rc1
únete a nosotros en ? (Twitter), Discord y WeChat
Instalar desde la fuente:
git clone https://github.com/InternLM/lagent.git
cd lagent
pip install -e . Lagent se inspira en la filosofía de diseño de PyTorch. Esperamos que la analogía de las capas de redes neuronales haga que el flujo de trabajo sea más claro e intuitivo, de modo que los usuarios sólo tengan que centrarse en crear capas y definir el mensaje que pasa entre ellas de forma pitónica. Este es un tutorial sencillo que le permitirá comenzar rápidamente a crear aplicaciones multiagente.
Los agentes utilizan AgentMessage para comunicarse.
from typing import Dict , List
from lagent . agents import Agent
from lagent . schema import AgentMessage
from lagent . llms import VllmModel , INTERNLM2_META
llm = VllmModel (
path = 'Qwen/Qwen2-7B-Instruct' ,
meta_template = INTERNLM2_META ,
tp = 1 ,
top_k = 1 ,
temperature = 1.0 ,
stop_words = [ '<|im_end|>' ],
max_new_tokens = 1024 ,
)
system_prompt = '你的回答只能从“典”、“孝”、“急”三个字中选一个。'
agent = Agent ( llm , system_prompt )
user_msg = AgentMessage ( sender = 'user' , content = '今天天气情况' )
bot_msg = agent ( user_msg )
print ( bot_msg ) content='急' sender='Agent' formatted=None extra_info=None type=None receiver=None stream_state=<AgentStatusCode.END: 0>
Tanto los mensajes de entrada como los de salida se agregarán a la memoria del Agent en cada paso directo. Esto se realiza en __call__ en lugar de forward . Vea el siguiente pseudocódigo
def __call__ ( self , * message ):
message = pre_hooks ( message )
add_memory ( message )
message = self . forward ( * message )
add_memory ( message )
message = post_hooks ( message )
return messageInspeccionar la memoria de dos maneras.
memory : List [ AgentMessage ] = agent . memory . get_memory ()
print ( memory )
print ( '-' * 120 )
dumped_memory : Dict [ str , List [ dict ]] = agent . state_dict ()
print ( dumped_memory [ 'memory' ]) [AgentMessage(content='今天天气情况', sender='user', formatted=None, extra_info=None, type=None, receiver=None, stream_state=<AgentStatusCode.END: 0>), AgentMessage(content='急', sender='Agent', formatted=None, extra_info=None, type=None, receiver=None, stream_state=<AgentStatusCode.END: 0>)]
------------------------------------------------------------------------------------------------------------------------
[{'content': '今天天气情况', 'sender': 'user', 'formatted': None, 'extra_info': None, 'type': None, 'receiver': None, 'stream_state': <AgentStatusCode.END: 0>}, {'content': '急', 'sender': 'Agent', 'formatted': None, 'extra_info': None, 'type': None, 'receiver': None, 'stream_state': <AgentStatusCode.END: 0>}]
Borre la memoria de esta sesión ( session_id=0 por defecto):
agent . reset () Se llama DefaultAggregator de forma interna para ensamblar y convertir AgentMessage al formato de mensaje OpenAI.
def forward ( self , * message : AgentMessage , session_id = 0 , ** kwargs ) -> Union [ AgentMessage , str ]:
formatted_messages = self . aggregator . aggregate (
self . memory . get ( session_id ),
self . name ,
self . output_format ,
self . template ,
)
llm_response = self . llm . chat ( formatted_messages , ** kwargs )
...Implementar un agregador simple que pueda recibir algunos disparos.
from typing import List , Union
from lagent . memory import Memory
from lagent . prompts import StrParser
from lagent . agents . aggregator import DefaultAggregator
class FewshotAggregator ( DefaultAggregator ):
def __init__ ( self , few_shot : List [ dict ] = None ):
self . few_shot = few_shot or []
def aggregate ( self ,
messages : Memory ,
name : str ,
parser : StrParser = None ,
system_instruction : Union [ str , dict , List [ dict ]] = None ) -> List [ dict ]:
_message = []
if system_instruction :
_message . extend (
self . aggregate_system_intruction ( system_instruction ))
_message . extend ( self . few_shot )
messages = messages . get_memory ()
for message in messages :
if message . sender == name :
_message . append (
dict ( role = 'assistant' , content = str ( message . content )))
else :
user_message = message . content
if len ( _message ) > 0 and _message [ - 1 ][ 'role' ] == 'user' :
_message [ - 1 ][ 'content' ] += user_message
else :
_message . append ( dict ( role = 'user' , content = user_message ))
return _message
agent = Agent (
llm ,
aggregator = FewshotAggregator (
[
{ "role" : "user" , "content" : "今天天气" },
{ "role" : "assistant" , "content" : "【晴】" },
]
)
)
user_msg = AgentMessage ( sender = 'user' , content = '昨天天气' )
bot_msg = agent ( user_msg )
print ( bot_msg ) content='【多云转晴，夜间有轻微降温】' sender='Agent' formatted=None extra_info=None type=None receiver=None stream_state=<AgentStatusCode.END: 0>
En AgentMessage , formatted está reservado para almacenar información analizada por output_format de la salida del modelo.
def forward ( self , * message : AgentMessage , session_id = 0 , ** kwargs ) -> Union [ AgentMessage , str ]:
...
llm_response = self . llm . chat ( formatted_messages , ** kwargs )
if self . output_format :
formatted_messages = self . output_format . parse_response ( llm_response )
return AgentMessage (
sender = self . name ,
content = llm_response ,
formatted = formatted_messages ,
)
...Utilice un analizador de herramientas de la siguiente manera
from lagent . prompts . parsers import ToolParser
system_prompt = "逐步分析并编写Python代码解决以下问题。"
parser = ToolParser ( tool_type = 'code interpreter' , begin = '```python n ' , end = ' n ``` n ' )
llm . gen_params [ 'stop_words' ]. append ( ' n ``` n ' )
agent = Agent ( llm , system_prompt , output_format = parser )
user_msg = AgentMessage (
sender = 'user' ,
content = 'Marie is thinking of a multiple of 63, while Jay is thinking of a '
'factor of 63. They happen to be thinking of the same number. There are '
'two possibilities for the number that each of them is thinking of, one '
'positive and one negative. Find the product of these two numbers.' )
bot_msg = agent ( user_msg )
print ( bot_msg . model_dump_json ( indent = 4 )) {
"content": "首先，我们需要找出63的所有正因数和负因数。63的正因数可以通过分解63的质因数来找出，即\(63 = 3^2 \times 7\)。因此，63的正因数包括1, 3, 7, 9, 21, 和 63。对于负因数，我们只需将上述正因数乘以-1。nn接下来，我们需要找出与63的正因数相乘的结果为63的数，以及与63的负因数相乘的结果为63的数。这可以通过将63除以每个正因数和负因数来实现。nn最后，我们将找到的两个数相乘得到最终答案。nn下面是Python代码实现：nn```pythonndef find_numbers():n # 正因数n positive_factors = [1, 3, 7, 9, 21, 63]n # 负因数n negative_factors = [-1, -3, -7, -9, -21, -63]n n # 找到与正因数相乘的结果为63的数n positive_numbers = [63 / factor for factor in positive_factors]n # 找到与负因数相乘的结果为63的数n negative_numbers = [-63 / factor for factor in negative_factors]n n # 计算两个数的乘积n product = positive_numbers[0] * negative_numbers[0]n n return productnnresult = find_numbers()nprint(result)",
"sender": "Agent",
"formatted": {
"tool_type": "code interpreter",
"thought": "首先，我们需要找出63的所有正因数和负因数。63的正因数可以通过分解63的质因数来找出，即\(63 = 3^2 \times 7\)。因此，63的正因数包括1, 3, 7, 9, 21, 和 63。对于负因数，我们只需将上述正因数乘以-1。nn接下来，我们需要找出与63的正因数相乘的结果为63的数，以及与63的负因数相乘的结果为63的数。这可以通过将63除以每个正因数和负因数来实现。nn最后，我们将找到的两个数相乘得到最终答案。nn下面是Python代码实现：nn",
"action": "def find_numbers():n # 正因数n positive_factors = [1, 3, 7, 9, 21, 63]n # 负因数n negative_factors = [-1, -3, -7, -9, -21, -63]n n # 找到与正因数相乘的结果为63的数n positive_numbers = [63 / factor for factor in positive_factors]n # 找到与负因数相乘的结果为63的数n negative_numbers = [-63 / factor for factor in negative_factors]n n # 计算两个数的乘积n product = positive_numbers[0] * negative_numbers[0]n n return productnnresult = find_numbers()nprint(result)",
"status": 1
},
"extra_info": null,
"type": null,
"receiver": null,
"stream_state": 0
}
ActionExecutor utiliza la misma estructura de datos de comunicación que Agent , pero requiere que el contenido de entrada AgentMessage sea un dict que contenga:
name : nombre de la herramienta, por ejemplo, 'IPythonInterpreter' , 'WebBrowser.search' .parameters : argumentos de palabras clave de la API de la herramienta, por ejemplo, {'command': 'import math;math.sqrt(2)'} , {'query': ['recent progress in AI']} .Puede registrar ganchos personalizados para la conversión de mensajes.
from lagent . hooks import Hook
from lagent . schema import ActionReturn , ActionStatusCode , AgentMessage
from lagent . actions import ActionExecutor , IPythonInteractive
class CodeProcessor ( Hook ):
def before_action ( self , executor , message , session_id ):
message = message . copy ( deep = True )
message . content = dict (
name = 'IPythonInteractive' , parameters = { 'command' : message . formatted [ 'action' ]}
)
return message
def after_action ( self , executor , message , session_id ):
action_return = message . content
if isinstance ( action_return , ActionReturn ):
if action_return . state == ActionStatusCode . SUCCESS :
response = action_return . format_result ()
else :
response = action_return . errmsg
else :
response = action_return
message . content = response
return message
executor = ActionExecutor ( actions = [ IPythonInteractive ()], hooks = [ CodeProcessor ()])
bot_msg = AgentMessage (
sender = 'Agent' ,
content = '首先，我们需要...' ,
formatted = {
'tool_type' : 'code interpreter' ,
'thought' : '首先，我们需要...' ,
'action' : 'def find_numbers(): n # 正因数n positive_factors = [1, 3, 7, 9, 21, 63] n # 负因数n negative_factors = [-1, -3, -7, -9, -21, -63] n n # 找到与正因数相乘的结果为63的数n positive_numbers = [63 / factor for factor in positive_factors] n # 找到与负因数相乘的结果为63的数n negative_numbers = [-63 / factor for factor in negative_factors] n n # 计算两个数的乘积n product = positive_numbers[0] * negative_numbers[0] n n return product n n result = find_numbers() n print(result)' ,
'status' : 1
})
executor_msg = executor ( bot_msg )
print ( executor_msg ) content='3969.0' sender='ActionExecutor' formatted=None extra_info=None type=None receiver=None stream_state=<AgentStatusCode.END: 0>
Para mayor comodidad, Lagent proporciona InternLMActionProcessor que se adapta a mensajes formateados por ToolParser como se mencionó anteriormente.
Lagent adopta un diseño de interfaz dual, donde casi todos los componentes (LLM, acciones, ejecutores de acciones...) tienen la variante asincrónica correspondiente anteponiendo a su identificador el prefijo 'Async'. Se recomienda utilizar agentes sincrónicos para la depuración y asincrónicos para la inferencia a gran escala para aprovechar al máximo los recursos inactivos de CPU y GPU.
Sin embargo, asegúrese de que la coherencia interna de los agentes, es decir, los agentes asincrónicos deben estar equipados con LLM asincrónicos y ejecutores de acciones asincrónicos que impulsen herramientas asincrónicas.
from lagent . llms import VllmModel , AsyncVllmModel , LMDeployPipeline , AsyncLMDeployPipeline
from lagent . actions import ActionExecutor , AsyncActionExecutor , WebBrowser , AsyncWebBrowser
from lagent . agents import Agent , AsyncAgent , AgentForInternLM , AsyncAgentForInternLM forward en lugar de __call__ de subclases a menos que sea necesario.session_id , que está diseñado para aislar la memoria, las solicitudes de LLM y la invocación de herramientas (por ejemplo, mantener múltiples entornos IPython independientes) de forma simultánea.Agentes matemáticos que resuelven problemas mediante programación.
from lagent . agents . aggregator import InternLMToolAggregator
class Coder ( Agent ):
def __init__ ( self , model_path , system_prompt , max_turn = 3 ):
super (). __init__ ()
llm = VllmModel (
path = model_path ,
meta_template = INTERNLM2_META ,
tp = 1 ,
top_k = 1 ,
temperature = 1.0 ,
stop_words = [ ' n ``` n ' , '<|im_end|>' ],
max_new_tokens = 1024 ,
)
self . agent = Agent (
llm ,
system_prompt ,
output_format = ToolParser (
tool_type = 'code interpreter' , begin = '```python n ' , end = ' n ``` n '
),
# `InternLMToolAggregator` is adapted to `ToolParser` for aggregating
# messages with tool invocations and execution results
aggregator = InternLMToolAggregator (),
)
self . executor = ActionExecutor ([ IPythonInteractive ()], hooks = [ CodeProcessor ()])
self . max_turn = max_turn
def forward ( self , message : AgentMessage , session_id = 0 ) -> AgentMessage :
for _ in range ( self . max_turn ):
message = self . agent ( message , session_id = session_id )
if message . formatted [ 'tool_type' ] is None :
return message
message = self . executor ( message , session_id = session_id )
return message
coder = Coder ( 'Qwen/Qwen2-7B-Instruct' , 'Solve the problem step by step with assistance of Python code' )
query = AgentMessage (
sender = 'user' ,
content = 'Find the projection of $ \ mathbf{a}$ onto $ \ mathbf{b} = '
' \ begin{pmatrix} 1 \ \ -3 \ end{pmatrix}$ if $ \ mathbf{a} \ cdot \ mathbf{b} = 2.$'
)
answer = coder ( query )
print ( answer . content )
print ( '-' * 120 )
for msg in coder . state_dict ()[ 'agent.memory' ]:
print ( '*' * 80 )
print ( f' { msg [ "sender" ] } : n n { msg [ "content" ] } ' )Agentes de blogs asincrónicos que mejoran la calidad de la escritura mediante el autorefinamiento (ejemplo original de AutoGen)
import asyncio
import os
from lagent . llms import AsyncGPTAPI
from lagent . agents import AsyncAgent
os . environ [ 'OPENAI_API_KEY' ] = 'YOUR_API_KEY'
class PrefixedMessageHook ( Hook ):
def __init__ ( self , prefix : str , senders : list = None ):
self . prefix = prefix
self . senders = senders or []
def before_agent ( self , agent , messages , session_id ):
for message in messages :
if message . sender in self . senders :
message . content = self . prefix + message . content
class AsyncBlogger ( AsyncAgent ):
def __init__ ( self , model_path , writer_prompt , critic_prompt , critic_prefix = '' , max_turn = 3 ):
super (). __init__ ()
llm = AsyncGPTAPI ( model_type = model_path , retry = 5 , max_new_tokens = 2048 )
self . writer = AsyncAgent ( llm , writer_prompt , name = 'writer' )
self . critic = AsyncAgent (
llm , critic_prompt , name = 'critic' , hooks = [ PrefixedMessageHook ( critic_prefix , [ 'writer' ])]
)
self . max_turn = max_turn
async def forward ( self , message : AgentMessage , session_id = 0 ) -> AgentMessage :
for _ in range ( self . max_turn ):
message = await self . writer ( message , session_id = session_id )
message = await self . critic ( message , session_id = session_id )
return await self . writer ( message , session_id = session_id )
blogger = AsyncBlogger (
'gpt-4o-2024-05-13' ,
writer_prompt = "You are an writing assistant tasked to write engaging blogpost. You try to generate the best blogpost possible for the user's request. "
"If the user provides critique, then respond with a revised version of your previous attempts" ,
critic_prompt = "Generate critique and recommendations on the writing. Provide detailed recommendations, including requests for length, depth, style, etc.." ,
critic_prefix = 'Reflect and provide critique on the following writing. n n ' ,
)
user_prompt = (
"Write an engaging blogpost on the recent updates in {topic}. "
"The blogpost should be engaging and understandable for general audience. "
"Should have more than 3 paragraphes but no longer than 1000 words." )
bot_msgs = asyncio . get_event_loop (). run_until_complete (
asyncio . gather (
* [
blogger ( AgentMessage ( sender = 'user' , content = user_prompt . format ( topic = topic )), session_id = i )
for i , topic in enumerate ([ 'AI' , 'Biotechnology' , 'New Energy' , 'Video Games' , 'Pop Music' ])
]
)
)
print ( bot_msgs [ 0 ]. content )
print ( '-' * 120 )
for msg in blogger . state_dict ( session_id = 0 )[ 'writer.memory' ]:
print ( '*' * 80 )
print ( f' { msg [ "sender" ] } : n n { msg [ "content" ] } ' )
print ( '-' * 120 )
for msg in blogger . state_dict ( session_id = 0 )[ 'critic.memory' ]:
print ( '*' * 80 )
print ( f' { msg [ "sender" ] } : n n { msg [ "content" ] } ' )Un flujo de trabajo de múltiples agentes que realiza recuperación de información, recopilación de datos y trazado de gráficos (ejemplo original de LangGraph)
