Update app.py

app.py

@@ -138,46 +138,77 @@
 
 
 
-import streamlit as st from langchain_community.document_loaders import PyPDFLoader from langchain.text_splitter import RecursiveCharacterTextSplitter from langchain_community.vectorstores import FAISS from langchain.embeddings import HuggingFaceEmbeddings from langchain.chains import RetrievalQA from langchain.prompts import PromptTemplate from langchain.llms import HuggingFaceHub import os
-
-Set Hugging Face API Token
-
+import streamlit as st
+from langchain_community.document_loaders import PyPDFLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import FAISS
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.chains import RetrievalQA
+from langchain.prompts import PromptTemplate
+from langchain.llms import HuggingFaceHub
+import os
+
+# Set Hugging Face API Token
 os.environ["HUGGINGFACEHUB_API_TOKEN"] = "your_huggingfacehub_api_token_here"
 
-Custom Prompt
-
-custom_prompt = PromptTemplate( input_variables=["context", "question"], template=""" You are a helpful assistant. Use the context below to answer the question. If the answer is not in the context, say "I don't know."
-
-Context: {context}
-
-Question: {question}
-
-Answer: """ )
-
-Load PDF and split into chunks
-
-def load_and_split_pdf(uploaded_file): loader = PyPDFLoader(uploaded_file.name) documents = loader.load() text_splitter = RecursiveCharacterTextSplitter(chunk_size=800, chunk_overlap=100) chunks = text_splitter.split_documents(documents) return chunks
-
-Build vectorstore from document chunks
-
-def build_vectorstore(chunks): embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2") db = FAISS.from_documents(chunks, embedding=embeddings) return db
-
-Build QA chain
-
-def build_qa_chain(vectorstore): llm = HuggingFaceHub(repo_id="mistralai/Mistral-7B-Instruct-v0.1", model_kwargs={"temperature": 0.2, "max_length": 512}) qa_chain = RetrievalQA.from_chain_type( llm=llm, retriever=vectorstore.as_retriever(), chain_type="stuff", chain_type_kwargs={"prompt": custom_prompt} ) return qa_chain
-
-Streamlit App
-
-st.set_page_config(page_title="Accurate PDF Chatbot", layout="centered") st.title("PDF QA Chatbot - RAG Powered")
+# Custom Prompt
+custom_prompt = PromptTemplate(
+    input_variables=["context", "question"],
+    template="""
+You are a helpful assistant. Use the context below to answer the question.
+If the answer is not in the context, say "I don't know."
+
+Context:
+{context}
+
+Question:
+{question}
+
+Answer:
+"""
+)
+
+# Load PDF and split into chunks
+def load_and_split_pdf(uploaded_file):
+    loader = PyPDFLoader(uploaded_file.name)
+    documents = loader.load()
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=800, chunk_overlap=100)
+    chunks = text_splitter.split_documents(documents)
+    return chunks
+
+# Build vectorstore from document chunks
+def build_vectorstore(chunks):
+    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
+    db = FAISS.from_documents(chunks, embedding=embeddings)
+    return db
+
+# Build QA chain
+def build_qa_chain(vectorstore):
+    llm = HuggingFaceHub(repo_id="mistralai/Mistral-7B-Instruct-v0.1", model_kwargs={"temperature": 0.2, "max_length": 512})
+    qa_chain = RetrievalQA.from_chain_type(
+        llm=llm,
+        retriever=vectorstore.as_retriever(),
+        chain_type="stuff",
+        chain_type_kwargs={"prompt": custom_prompt}
+    )
+    return qa_chain
+
+# Streamlit App
+st.set_page_config(page_title="Accurate PDF Chatbot", layout="centered")
+st.title("PDF QA Chatbot - RAG Powered")
 
 uploaded_file = st.file_uploader("Upload a PDF", type="pdf")
 
-if uploaded_file: with st.spinner("Reading and processing PDF..."): chunks = load_and_split_pdf(uploaded_file) vectorstore = build_vectorstore(chunks) qa_chain = build_qa_chain(vectorstore) st.success("PDF processed. Ask your question below.")
-
-question = st.text_input("Ask a question from the PDF:")
+if uploaded_file:
+    with st.spinner("Reading and processing PDF..."):
+        chunks = load_and_split_pdf(uploaded_file)
+        vectorstore = build_vectorstore(chunks)
+        qa_chain = build_qa_chain(vectorstore)
+        st.success("PDF processed. Ask your question below.")
 
-if question:
-    with st.spinner("Searching answer..."):
-        answer = qa_chain.run(question)
-        st.markdown(f"**Answer:** {answer}")
+    question = st.text_input("Ask a question from the PDF:")
 
+    if question:
+        with st.spinner("Searching answer..."):
+            answer = qa_chain.run(question)
+            st.markdown(f"**Answer:** {answer}")