query_data.py

import argparse
from langchain.vectorstores.chroma import Chroma
from langchain.prompts import ChatPromptTemplate
from langchain_community.llms.ollama import Ollama
from utils import *
from pdb import set_trace as bp

CHROMA_PATH = "chroma"
PROMPT_TEMPLATE = """
Answer the question based only on the following context:

{context}

---
Now, Answer the question based on the above context: {question}
"""

def query_rag(query_text: str):
    # Prepare the DB
    print("Preparing the DB")
    embedding_function = get_embedding_function()
    db = Chroma(
        persist_directory=CHROMA_PATH, embedding_function=embedding_function
    )

    # Search the DB
    print("Searching the DB")
    results = db.similarity_search_with_score(query_text, k = 3)
    context_text = "\n\n---\n\n".join([doc.page_content for doc, _score in results])
    prompt_template = ChatPromptTemplate.from_template(PROMPT_TEMPLATE)
    prompt = prompt_template.format(context = context_text, question = query_text)

    model = Ollama(model="llama3")
    response_text = model.invoke(prompt)

    sources = [doc.metadata.get("id", None) for doc, _score in results]
    formatted_response = f"Response: {response_text} \n\nSources: {sources}"
    print(formatted_response)
    return response_text

def main():
    parser = argparse.ArgumentParser()
    parser.add_argument("query_text", type=str, help="The query/question to the model")
    args = parser.parse_args()
    query_text = args.query_text
    query_rag(query_text)


if __name__ == "__main__":
    main()