PDF-Insight-PRO / preprocessing.py
Jatin Mehra
Refactor agentic_rag function to include memory parameter and enhance prompt clarity with detailed instructions for context usage and search behavior.
9492bcd
raw
history blame
7.34 kB
import os
from langchain_community.document_loaders import PyMuPDFLoader
import faiss
from langchain_groq import ChatGroq
from langchain.agents import AgentExecutor, create_tool_calling_agent
from langchain_community.tools.tavily_search import TavilySearchResults
from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain.memory import ConversationBufferMemory
from sentence_transformers import SentenceTransformer
import dotenv
dotenv.load_dotenv()
# Initialize LLM and tools globally
def model_selection(model_name):
llm = ChatGroq(model=model_name, api_key=os.getenv("GROQ_API_KEY"))
return llm
tools = [TavilySearchResults(max_results=5)]
# Initialize memory for conversation history
memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
def estimate_tokens(text):
"""Estimate the number of tokens in a text (rough approximation)."""
return len(text) // 4
def process_pdf_file(file_path):
"""Load a PDF file and extract its text with metadata."""
if not os.path.exists(file_path):
raise FileNotFoundError(f"The file {file_path} does not exist.")
loader = PyMuPDFLoader(file_path)
documents = loader.load()
return documents # Return list of Document objects with metadata
def chunk_text(documents, max_length=1000):
"""Split documents into chunks with metadata."""
chunks = []
for doc in documents:
text = doc.page_content
metadata = doc.metadata
paragraphs = text.split("\n\n")
current_chunk = ""
current_metadata = metadata.copy()
for paragraph in paragraphs:
if estimate_tokens(current_chunk + paragraph) <= max_length // 4:
current_chunk += paragraph + "\n\n"
else:
chunks.append({"text": current_chunk.strip(), "metadata": current_metadata})
current_chunk = paragraph + "\n\n"
if current_chunk:
chunks.append({"text": current_chunk.strip(), "metadata": current_metadata})
return chunks
def create_embeddings(chunks, model):
"""Create embeddings for a list of chunk texts."""
texts = [chunk["text"] for chunk in chunks]
embeddings = model.encode(texts, show_progress_bar=True, convert_to_tensor=True)
return embeddings.cpu().numpy(), chunks
def build_faiss_index(embeddings):
"""Build a FAISS HNSW index from embeddings for similarity search."""
dim = embeddings.shape[1]
index = faiss.IndexHNSWFlat(dim, 32) # 32 = number of neighbors in HNSW graph
index.hnsw.efConstruction = 200 # Higher = better quality, slower build
index.hnsw.efSearch = 50 # Higher = better accuracy, slower search
index.add(embeddings)
return index
def retrieve_similar_chunks(query, index, chunks, model, k=10, max_chunk_length=1000):
"""Retrieve top k similar chunks to the query from the FAISS index."""
query_embedding = model.encode([query], convert_to_tensor=True).cpu().numpy()
distances, indices = index.search(query_embedding, k)
return [(chunks[i]["text"][:max_chunk_length], distances[0][j], chunks[i]["metadata"]) for j, i in enumerate(indices[0])]
def agentic_rag(llm, tools, query, context_chunks, memory, Use_Tavily=False):
# Sort chunks by relevance (lower distance = more relevant)
context_chunks = sorted(context_chunks, key=lambda x: x[1]) # Sort by distance
context = ""
total_tokens = 0
max_tokens = 7000 # Leave room for prompt and response
# Aggregate relevant chunks until token limit is reached
for chunk, _, _ in context_chunks: # Unpack three elements
chunk_tokens = estimate_tokens(chunk)
if total_tokens + chunk_tokens <= max_tokens:
context += chunk + "\n\n"
total_tokens += chunk_tokens
else:
break
# Set up the search behavior
search_behavior = (
"If the context is insufficient, *then* use the 'search' tool to find the answer."
if Use_Tavily
else "If the context is insufficient, you *must* state that you don't know."
)
# Define prompt template
prompt = ChatPromptTemplate.from_messages([
("system", """
You are an expert Q&A system. Your primary function is to answer questions using a given set of documents (Context).
**Your Process:**
1. **Analyze the Question:** Understand exactly what the user is asking.
2. **Scan the Context:** Thoroughly review the 'Context' provided to find relevant information.
3. **Formulate the Answer:**
* If the context contains a clear answer, synthesize it into a concise response.
* **Always** start your answer with "Based on the Document, ...".
* {search_behavior}
* If, after all steps, you cannot find an answer, respond with: "Based on the Document, I don't know the answer."
4. **Clarity:** Ensure your final answer is clear, direct, and avoids jargon if possible.
**Important Rules:**
* **Stick to the Context:** Unless you use the search tool, do *not* use any information outside of the provided 'Context'.
* **No Speculation:** Do not make assumptions or infer information not explicitly present.
* **Cite Sources (If Searching):** If you use the search tool, you MUST include the source links in your response.
"""),
("human", "Context: {context}\n\nQuestion: {input}"),
MessagesPlaceholder(variable_name="chat_history"),
MessagesPlaceholder(variable_name="agent_scratchpad"),
])
agent_tools = tools if Use_Tavily else []
try:
agent = create_tool_calling_agent(llm, agent_tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=agent_tools, memory=memory, verbose=True)
return agent_executor.invoke({
"input": query,
"context": context,
"search_behavior": search_behavior
})
except Exception as e:
print(f"Error during agent execution: {str(e)}")
fallback_prompt = ChatPromptTemplate.from_messages([
("system", "You are a helpful assistant. Use the provided context to answer the user's question."),
("human", "Context: {context}\n\nQuestion: {input}")
])
response = llm.invoke(fallback_prompt.format(context=context, input=query))
return {"output": response.content}
"""if __name__ == "__main__":
# Process PDF and prepare index
dotenv.load_dotenv()
pdf_file = "JatinCV.pdf"
llm = model_selection("meta-llama/llama-4-scout-17b-16e-instruct")
texts = process_pdf_file(pdf_file)
chunks = chunk_text(texts, max_length=1500)
model = SentenceTransformer('all-MiniLM-L6-v2')
embeddings = create_embeddings(chunks, model)
index = build_faiss_index(embeddings)
# Chat loop
print("Chat with the assistant (type 'exit' or 'quit' to stop):")
while True:
query = input("User: ")
if query.lower() in ["exit", "quit"]:
break
# Retrieve similar chunks
similar_chunks = retrieve_similar_chunks(query, index, chunks, model, k=3)
# context = "\n".join([chunk for chunk, _ in similar_chunks])
# Generate response
response = agentic_rag(llm, tools, query=query, context=similar_chunks, Use_Tavily=True, memory=memory)
print("Assistant:", response["output"])"""