langchain/templates/rag-lantern/rag_lantern/chain.py

from langchain_community.chat_models import ChatOpenAI
from langchain_community.embeddings import OpenAIEmbeddings
from langchain_community.vectorstores import Lantern
from langchain_core.output_parsers import StrOutputParser
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.pydantic_v1 import BaseModel
from langchain_core.runnables import RunnableParallel, RunnablePassthrough

CONNECTION_STRING = "postgresql://postgres:postgres@localhost:5432"
COLLECTION_NAME = "documents"
DB_NAME = "postgres"

embeddings = OpenAIEmbeddings()

vectorstore = Lantern(
    collection_name=COLLECTION_NAME,
    connection_string=CONNECTION_STRING,
    embedding_function=embeddings,
)

retriever = vectorstore.as_retriever()


template = """Answer the question based only on the following context:
{context}

Question: {question}
"""

prompt = ChatPromptTemplate.from_template(template)

model = ChatOpenAI()

chain = (
    RunnableParallel({"context": retriever, "question": RunnablePassthrough()})
    | prompt
    | model
    | StrOutputParser()
)


# Add typing for input
class Question(BaseModel):
    __root__: str


chain = chain.with_types(input_type=Question)