ANTHROPIC Archives - icodebroker

[PYTHON/COMMON] langchain-google-vertexai 패키지 설치하기

■ langchain-google-vertexai 패키지를 설치하는 방법을 보여준다. 1. 명령 프롬프트를 실행한다. 2. 명령 프롬프트에서 아래 명령을 실행한다. ▶ 실행 명령


pip install langchain-google-vertexai

pip install langchain-google-vertexai

[PYTHON/LANGCHAIN] init_chat_model 함수 : 추론 기능을 사용해 모델 만들기

■ init_chat_model 함수에서 추론 기능을 사용해 모델을 만드는 방법을 보여준다. ※ API KEY 환경 변수 값은 .env 파일에 정의한다. ▶ main.py


from dotenv                import load_dotenv
from langchain.chat_models import init_chat_model

load_dotenv()

chatOpenAI    = init_chat_model("gpt-4o"                , temperature = 0)
chatAnthropic = init_chat_model("claude-3-opus-20240229", temperature = 0)
chatVertexAI  = init_chat_model("gemini-1.5-pro"        , temperature = 0)

from dotenv import load_dotenv

from langchain.chat_models import init_chat_model

load_dotenv()

chatOpenAI = init_chat_model("gpt-4o" , temperature = 0)

chatAnthropic = init_chat_model("claude-3-opus-20240229", temperature = 0)

chatVertexAI = init_chat_model("gemini-1.5-pro" , temperature = 0)

[PYTHON/LANGCHAIN] init_chat_model 함수 : 생성자를 사용해 모델 만들기

■ init_chat_model 함수의 생성자를 사용해 모델을 만드는 방법을 보여준다. ※ API KEY 환경 변수 값은 .env 파일에 정의한다. ▶ main.py


from dotenv                import load_dotenv
from langchain.chat_models import init_chat_model

load_dotenv()

chatOpenAI    = init_chat_model("gpt-4o"                , model_provider = "openai"         , temperature = 0)
chatAnthropic = init_chat_model("claude-3-opus-20240229", model_provider = "anthropic"      , temperature = 0)
chatVertexAI  = init_chat_model("gemini-1.5-pro"        , model_provider = "google_vertexai", temperature = 0)

print("GPT-4o      : " + chatOpenAI.invoke("what's your name").content   )
print("Claude Opus : " + chatAnthropic.invoke("what's your name").content)
print("Gemini 1.5  : " + chatVertexAI .invoke("what's your name").content)

"""
GPT-4o      : I'm an AI created by OpenAI, and I don't have a personal name. You can call me Assistant! How can I help you today?
Claude Opus : My name is Claude. It's nice to meet you!
Gemini 1.5  : I am a large language model, trained by Google. I do not have a name.
"""

from dotenv import load_dotenv

from langchain.chat_models import init_chat_model

load_dotenv()

chatOpenAI = init_chat_model("gpt-4o" , model_provider = "openai" , temperature = 0)

chatAnthropic = init_chat_model("claude-3-opus-20240229", model_provider = "anthropic" , temperature = 0)

chatVertexAI = init_chat_model("gemini-1.5-pro" , model_provider = "google_vertexai", temperature = 0)

print("GPT-4o : " + chatOpenAI.invoke("what's your name").content )

print("Claude Opus : " + chatAnthropic.invoke("what's your name").content)

print("Gemini 1.5 : " + chatVertexAI .invoke("what's your name").content)

"""

GPT-4o : I'm an AI created by OpenAI, and I don't have a personal name. You can call me Assistant! How can I help you today?

Claude Opus : My name is Claude. It's nice to meet you!

Gemini 1.5 : I am a large language model, trained by Google. I do not have a name.

"""

▶

[PYTHON/LANGCHAIN] AIMessage 클래스 : response_metadata 속성 사용하기 (ChatBedrock 객체 사용시)

■ AIMessage 클래스의 response_metadata 속성을 사용하는 방법을 보여준다. (ChatBedrock 객체 사용시) ▶ 예제 코드 (PY)


from langchain_aws import ChatBedrock

chatBedrock = ChatBedrock(model_id = "anthropic.claude-v2")

aiMessage = chatBedrock.invoke([("human", "What's the oldest known example of cuneiform")])

print(aiMessage.response_metadata)

"""
{
    'model_id' : 'anthropic.claude-v2',
    'usage' : {'prompt_tokens' : 19, 'completion_tokens' : 371, 'total_tokens' : 390}
}
"""

from langchain_aws import ChatBedrock

chatBedrock = ChatBedrock(model_id = "anthropic.claude-v2")

aiMessage = chatBedrock.invoke([("human", "What's the oldest known example of cuneiform")])

print(aiMessage.response_metadata)

"""

{

'model_id' : 'anthropic.claude-v2',

'usage' : {'prompt_tokens' : 19, 'completion_tokens' : 371, 'total_tokens' : 390}

}

"""

[PYTHON/LANGCHAIN] AIMessage 클래스 : response_metadata 속성 사용하기 (ChatAnthropic 객체 사용시)

■ AIMessage 클래스의 response_metadata 속성을 사용하는 방법을 보여준다. (ChatAnthropic 객체 사용시) ▶ 예제 코드 (PY)


from langchain_anthropic import ChatAnthropic

chatAnthropic = ChatAnthropic(model = "claude-3-sonnet-20240229")

aiMessage = chatAnthropic.invoke([("human", "What's the oldest known example of cuneiform")])

print(aiMessage.response_metadata)

"""
{
    'id'            : 'msg_01CzQyD7BX8nkhDNfT1QqvEp',
    'model'         : 'claude-3-sonnet-20240229',
    'stop_reason'   : 'end_turn',
    'stop_sequence' : None,
    'usage'         : {'input_tokens' : 17, 'output_tokens' : 296}
}
"""

from langchain_anthropic import ChatAnthropic

chatAnthropic = ChatAnthropic(model = "claude-3-sonnet-20240229")

aiMessage = chatAnthropic.invoke([("human", "What's the oldest known example of cuneiform")])

print(aiMessage.response_metadata)

"""

{

'id' : 'msg_01CzQyD7BX8nkhDNfT1QqvEp',

'model' : 'claude-3-sonnet-20240229',

'stop_reason' : 'end_turn',

'stop_sequence' : None,

'usage' : {'input_tokens' : 17, 'output_tokens' : 296}

}

"""

[PYTHON/LANGCHAIN] RunnableSequence 클래스 : with_config 메소드를 사용해 모델이나 프롬프트를 대체한 RunnableBinding 객체 만들기

■ RunnableSequence 클래스의 with_config 메소드를 사용해 모델이나 프롬프트를 대체한 RunnableBinding 객체를 만드는 방법을 보여준다. ※ ANTHROPIC_API_KEY 및 OPENAI_API_KEY 환경 변수 값은 .env

[PYTHON/LANGCHAIN] 대체 가능한 프롬프트 및 LLM 설정하기

■ 대체 가능한 프롬프트 및 LLM을 설정하는 방법을 보여준다. ※ ANTHROPIC_API_KEY 및 OPENAI_API_KEY 환경 변수 값은 .env 파일에 정의한다. ▶ main.py


from dotenv                   import load_dotenv
from langchain_core.prompts   import PromptTemplate
from langchain_core.runnables import ConfigurableField
from langchain_anthropic      import ChatAnthropic
from langchain_openai         import ChatOpenAI

load_dotenv()

promptTemplate = PromptTemplate.from_template("Tell me a joke about {topic}")

promptRunnableConfigurableAlternatives = promptTemplate.configurable_alternatives(
    ConfigurableField(id = "prompt"),
    default_key = "joke",
    poem        = PromptTemplate.from_template("Write a short poem about {topic}")
)

chatAnthropic = ChatAnthropic(model = "claude-3-haiku-20240307", temperature = 0)

llmRunnableConfigurableAlternatives = chatAnthropic.configurable_alternatives(
    ConfigurableField(id = "llm"),
    default_key = "anthropic",
    openai      = ChatOpenAI(),
    gpt4        = ChatOpenAI(model = "gpt-4")
)

runnableSequence = promptRunnableConfigurableAlternatives | llmRunnableConfigurableAlternatives

runnableBinding = runnableSequence.with_config(configurable = {"prompt" : "poem", "llm" : "openai"})

responseAIMessage = runnableBinding.invoke({"topic" : "bears"})

print(responseAIMessage)

"""
content           = 'In the forest deep and wide,\nWhere the shadows softly glide,\nThere roams a creature strong and grand,\nThe mighty bear, ruler of the land.\n\nWith fur as dark as the night,\nAnd eyes that gleam with ancient might,\nHe wanders through the trees so tall,\nA silent guardian over all.\n\nHis paws leave prints in the snow,\nAs he hunts for food below,\nBut do not fear his fearsome roar,\nFor the bear is gentle at his core.\n\nSo let us admire this noble beast,\nAnd hope that in the wild, he finds his feast,\nFor in the heart of every bear,\nThere lies a beauty beyond compare.'
response_metadata = {
    'token_usage'        : {'completion_tokens' : 132, 'prompt_tokens' : 13, 'total_tokens' : 145},
    'model_name'         : 'gpt-3.5-turbo',
    'system_fingerprint' : None,
    'finish_reason'      : 'stop',
    'logprobs'           : None
}
id                = 'run-c71c2737-3b2d-4460-be10-735f74ac7e7f-0'
usage_metadata    = {'input_tokens' : 13, 'output_tokens' : 132, 'total_tokens' : 145}
"""

from dotenv import load_dotenv

from langchain_core.prompts import PromptTemplate

from langchain_core.runnables import ConfigurableField

from langchain_anthropic import ChatAnthropic

from langchain_openai import ChatOpenAI

load_dotenv()

promptTemplate = PromptTemplate.from_template("Tell me a joke about {topic}")

promptRunnableConfigurableAlternatives = promptTemplate.configurable_alternatives(

ConfigurableField(id = "prompt"),

default_key = "joke",

poem = PromptTemplate.from_template("Write a short poem about {topic}")

)

chatAnthropic = ChatAnthropic(model = "claude-3-haiku-20240307", temperature = 0)

llmRunnableConfigurableAlternatives = chatAnthropic.configurable_alternatives(

ConfigurableField(id = "llm"),

default_key = "anthropic",

openai = ChatOpenAI(),

gpt4 = ChatOpenAI(model = "gpt-4")

)

runnableSequence = promptRunnableConfigurableAlternatives | llmRunnableConfigurableAlternatives

runnableBinding = runnableSequence.with_config(configurable = {"prompt" : "poem", "llm" : "openai"})

responseAIMessage = runnableBinding.invoke({"topic" : "bears"})

print(responseAIMessage)

"""

content = 'In the forest deep and wide,\nWhere the shadows softly glide,\nThere roams a creature strong and grand,\nThe mighty bear, ruler of the land.\n\nWith fur as dark as the night,\nAnd eyes that gleam with ancient might,\nHe wanders through the trees so tall,\nA silent guardian over all.\n\nHis paws leave prints in the snow,\nAs he hunts for food below,\nBut do not fear his fearsome roar,\nFor the bear is gentle at his core.\n\nSo let us admire this noble beast,\nAnd hope that in the wild, he finds his feast,\nFor in the heart of every bear,\nThere lies a beauty beyond compare.'

response_metadata = {

'token_usage' : {'completion_tokens' : 132, 'prompt_tokens' : 13, 'total_tokens' : 145},

'model_name' : 'gpt-3.5-turbo',

'system_fingerprint' : None,

'finish_reason' : 'stop',

'logprobs' : None

}

id = 'run-c71c2737-3b2d-4460-be10-735f74ac7e7f-0'

usage_metadata = {'input_tokens' : 13, 'output_tokens' : 132, 'total_tokens' : 145}

"""

[PYTHON/COMMON] langchain-anthropic 패키지 설치하기

■ langchain-anthropic 패키지를 설치하는 방법을 보여준다. 1. 명령 프롬프트를 실행한다. 2. 명령 프롬프트에서 아래 명령을 실행한다. ▶ 실행 명령


pip install langchain-anthropic

pip install langchain-anthropic

[PYTHON/LANGCHAIN] ChatAnthropic 클래스 : configurable_alternatives 메소드를 사용해 대체 가능한 LLM을 설정하는 RunnableConfigurableAlternatives 객체 만들기

■ ChatAnthropic 클래스의 configurable_alternatives 메소드를 사용해 대체 가능한 LLM을 설정하는 RunnableConfigurableAlternatives 객체를 만드는 방법을 보여준다. ※ ANTHROPIC_API_KEY 및 OPENAI_API_KEY 환경 변수 값은