basic_rag.py

# import os

# files = os.listdir("../Aihacks/txt_files")

# print(files)


# %%
"""
# Setup
"""


# %%
# OctoAI
# ! pip install langchain langchain-community faiss-cpu sentence-transformers octoai-sdk langchain-text-splitters lxml tiktoken python-dotenv 'arize-phoenix[evals]'

# %%
from dotenv import load_dotenv
import os
import PyPDF2

load_dotenv()
OCTOAI_API_TOKEN = os.environ["OCTOAI_API_TOKEN"]

# %%
"""
# change pdf to textfile
"""

def pdf_to_text(pdf_path, txt_path):
    # Open the PDF file
    with open(pdf_path, 'rb') as pdf_file:
        # Create a PDF reader object
        pdf_reader = PyPDF2.PdfFileReader(pdf_file)
        
        # Open the text file in write mode
        with open(txt_path, 'w') as text_file:
            # Iterate through each page in the PDF
            for page_num in range(pdf_reader.numPages):
                # Get the page
                page = pdf_reader.getPage(page_num)
                # Extract the text from the page
                text = page.extract_text()
                # Write the text to the text file
                text_file.write(text)

# %%
"""
# Ingest Data
"""

# %%
from langchain.text_splitter import CharacterTextSplitter
from langchain.schema import Document

# %%
files = os.listdir("../Aihacks/txt_files")
file_texts = []
for file in files:
    with open(f"../Aihacks/txt_files/{file}") as f:
        file_text = f.read()
    text_splitter = CharacterTextSplitter.from_tiktoken_encoder(
        chunk_size=512, chunk_overlap=64, 
    )
    texts = text_splitter.split_text(file_text)
    for i, chunked_text in enumerate(texts):
        file_texts.append(Document(page_content=chunked_text, 
                metadata={"doc_title": file.split(".")[0], "chunk_num": i}))

# %%
from langchain_huggingface import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS

# %%
embeddings = HuggingFaceEmbeddings()

# %%
vector_store = FAISS.from_documents(
    file_texts,
    embedding=embeddings
)

# %%
"""
# Search the Data
"""

# %%
from langchain_community.llms.octoai_endpoint import OctoAIEndpoint
llm = OctoAIEndpoint(
        model="meta-llama-3-8b-instruct",
        max_tokens=1024,
        presence_penalty=0,
        temperature=0.1,
        top_p=0.9,
    )

# %%
retriever = vector_store.as_retriever()

# %%
from langchain.prompts import ChatPromptTemplate
template="""You are a investment banker. If you don't know the answer, just say that you don't know. Use three sentences maximum and keep the answer concise
Question: {question} 
Context: {context} 
Answer:"""
prompt = ChatPromptTemplate.from_template(template)

# %%
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser
chain = (
    {"context": retriever, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

# %%
print(chain.invoke("What is microsofts outlook in 2025?"))

# %%