[PYTHON/LANGCHAIN] RunnableLambda 클래스 : 커스텀 함수 만들기

■ RunnableLambda 클래스를 사용해 커스텀 함수를 만드는 방법을 보여준다.

※ OPENAI_API_KEY 환경 변수 값은 .env 파일에 정의한다.

▶ main.py


from operator import itemgetter

from dotenv                   import load_dotenv
from langchain_core.prompts   import ChatPromptTemplate
from langchain_core.runnables import RunnableLambda
from langchain_openai         import ChatOpenAI

load_dotenv()

def getStringLength(text):
    return len(text)

def multiplyStringLengths(text1, text2):
    return len(text1) * len(text2)

def multiplyStringLengthsFromDictionary(dictionary):
    return multiplyStringLengths(dictionary["text1"], dictionary["text2"])

chatOpenAI = ChatOpenAI()

chatPromptTemplate = ChatPromptTemplate.from_template("what is {a} + {b}")

runnableSequence = (
    {
        "a" : itemgetter("foo") | RunnableLambda(getStringLength),
        "b" : {"text1" : itemgetter("foo"), "text2" : itemgetter("bar")} | RunnableLambda(multiplyStringLengthsFromDictionary),
    }
    | chatPromptTemplate
    | chatOpenAI
)

responseAIMessage = runnableSequence.invoke({"foo" : "bar", "bar" : "gah"})

print(responseAIMessage)

"""
content           = '3 + 9 = 12'
response_metadata = {
    'token_usage'        : {completion_tokens' : 7, 'prompt_tokens' : 14, 'total_tokens' : 21},
    'model_name'         : 'gpt-3.5-turbo',
    'system_fingerprint' : None,
    'finish_reason'      : 'stop',
    'logprobs'           : None
}
id                = 'run-c93ab559-db3d-4ff7-828d-15d27444b606-0'
usage_metadata    = {'input_tokens' : 14, 'output_tokens' : 7, 'total_tokens' : 21}
"""

from operator import itemgetter

from dotenv import load_dotenv

from langchain_core.prompts import ChatPromptTemplate

from langchain_core.runnables import RunnableLambda

from langchain_openai import ChatOpenAI

load_dotenv()

def getStringLength(text):

return len(text)

def multiplyStringLengths(text1, text2):

return len(text1) * len(text2)

def multiplyStringLengthsFromDictionary(dictionary):

return multiplyStringLengths(dictionary["text1"], dictionary["text2"])

chatOpenAI = ChatOpenAI()

chatPromptTemplate = ChatPromptTemplate.from_template("what is {a} + {b}")

runnableSequence = (

{

"a" : itemgetter("foo") | RunnableLambda(getStringLength),

"b" : {"text1" : itemgetter("foo"), "text2" : itemgetter("bar")} | RunnableLambda(multiplyStringLengthsFromDictionary),

}

| chatPromptTemplate

| chatOpenAI

)

responseAIMessage = runnableSequence.invoke({"foo" : "bar", "bar" : "gah"})

print(responseAIMessage)

"""

content = '3 + 9 = 12'

response_metadata = {

'token_usage' : {completion_tokens' : 7, 'prompt_tokens' : 14, 'total_tokens' : 21},

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

'system_fingerprint' : None,

'finish_reason' : 'stop',

'logprobs' : None

}

id = 'run-c93ab559-db3d-4ff7-828d-15d27444b606-0'

usage_metadata = {'input_tokens' : 14, 'output_tokens' : 7, 'total_tokens' : 21}

"""

▶ requirements.txt


aiohttp==3.9.5
aiosignal==1.3.1
annotated-types==0.7.0
anyio==4.4.0
async-timeout==4.0.3
attrs==23.2.0
certifi==2024.6.2
charset-normalizer==3.3.2
distro==1.9.0
exceptiongroup==1.2.1
frozenlist==1.4.1
greenlet==3.0.3
h11==0.14.0
httpcore==1.0.5
httpx==0.27.0
idna==3.7
jsonpatch==1.33
jsonpointer==3.0.0
langchain==0.2.5
langchain-core==0.2.9
langchain-openai==0.1.8
langchain-text-splitters==0.2.1
langsmith==0.1.81
multidict==6.0.5
numpy==1.26.4
openai==1.35.3
orjson==3.10.5
packaging==24.1
pydantic==2.7.4
pydantic_core==2.18.4
python-dotenv==1.0.1
PyYAML==6.0.1
regex==2024.5.15
requests==2.32.3
sniffio==1.3.1
SQLAlchemy==2.0.31
tenacity==8.4.1
tiktoken==0.7.0
tqdm==4.66.4
typing_extensions==4.12.2
urllib3==2.2.2
yarl==1.9.4

aiohttp==3.9.5

aiosignal==1.3.1

annotated-types==0.7.0

anyio==4.4.0

async-timeout==4.0.3

attrs==23.2.0

certifi==2024.6.2

charset-normalizer==3.3.2

distro==1.9.0

exceptiongroup==1.2.1

frozenlist==1.4.1

greenlet==3.0.3

h11==0.14.0

httpcore==1.0.5

httpx==0.27.0

idna==3.7

jsonpatch==1.33

jsonpointer==3.0.0

langchain==0.2.5

langchain-core==0.2.9

langchain-openai==0.1.8

langchain-text-splitters==0.2.1

langsmith==0.1.81

multidict==6.0.5

numpy==1.26.4

openai==1.35.3

orjson==3.10.5

packaging==24.1

pydantic==2.7.4

pydantic_core==2.18.4

python-dotenv==1.0.1

PyYAML==6.0.1

regex==2024.5.15

requests==2.32.3

sniffio==1.3.1

SQLAlchemy==2.0.31

tenacity==8.4.1

tiktoken==0.7.0

tqdm==4.66.4

typing_extensions==4.12.2

urllib3==2.2.2

yarl==1.9.4

※ pip install python-dotenv langchain langchain_openai 명령을 실행했다.