Update app.py

app.py

@@ -1,68 +1,78 @@
-import gradio as gr
-from langchain_community.document_loaders import PyPDFLoader
-from langchain_community.embeddings import HuggingFaceEmbeddings
-from langchain_community.vectorstores import Chroma
-from langchain_community.chat_models import ChatGroq
-from langchain.chains import ConversationalRetrievalChain
-from langchain.memory import ConversationBufferMemory
 import os
-import shutil
-
-# PDF upload folder
-UPLOAD_DIR = "pdf_uploads"
-if not os.path.exists(UPLOAD_DIR):
-    os.makedirs(UPLOAD_DIR)
-
-# Initialize LLaMA 3 (no API key directly in code)
-llm = ChatGroq(model_name="llama3-8b-8192")
+import gradio as gr
+from transformers import pipeline
+from PyPDF2 import PdfReader
 
-embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
-memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
+# Load a summarization pipeline using a free HF model (e.g., "facebook/bart-large-cnn")
+summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
 
-qa_chain = None
+# Load a QA pipeline
+qa_pipeline = pipeline("question-answering", model="distilbert-base-cased-distilled-squad")
 
-def process_pdf(file_path):
-    loader = PyPDFLoader(file_path)
-    pages = loader.load_and_split()
+# Function to extract text from PDF
+def extract_text_from_pdf(pdf_file):
+    reader = PdfReader(pdf_file)
+    text = ""
+    for page in reader.pages:
+        text += page.extract_text() + "\n"
+    return text
 
-    vectordb = Chroma.from_documents(pages, embedding=embeddings)
-    qa_chain = ConversationalRetrievalChain.from_llm(
-        llm=llm,
-        retriever=vectordb.as_retriever(),
-        memory=memory
-    )
-    return qa_chain
+# Process the uploaded PDF and store the text
+pdf_text_store = ""
 
 def upload_pdf(file):
-    global qa_chain
-    file_path = os.path.join(UPLOAD_DIR, os.path.basename(file))
-    shutil.copy(file, file_path)
-    qa_chain = process_pdf(file_path)
-    return "✅ PDF uploaded and processed! Ask me anything about it."
+    global pdf_text_store
+    pdf_text_store = extract_text_from_pdf(file)
+    # Summarize for quick overview
+    summary = summarizer(pdf_text_store[:1000], max_length=200, min_length=50, do_sample=False)[0]['summary_text']
+    return f"✅ PDF processed! Summary:\n\n{summary}"
 
 def chatbot(user_message, history):
-    if qa_chain is None:
+    if not pdf_text_store:
         return "❌ Please upload a PDF first.", history
-
-    response = qa_chain({"question": user_message, "chat_history": history})
-    answer = response["answer"]
+    
+    result = qa_pipeline({
+        'context': pdf_text_store,
+        'question': user_message
+    })
+    answer = result["answer"]
     history.append((user_message, answer))
     return "", history
 
-with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("<h1 style='text-align:center;'>📄 LLaMA 3 PDF Chatbot</h1>")
+with gr.Blocks(theme=gr.themes.Soft(primary_hue="cyan", secondary_hue="blue")) as demo:
+    gr.HTML(
+        """
+        <div style="text-align:center;">
+            <h1 style="font-size:2.2rem; margin-bottom:0.3em;">📚 Free PDF Chatbot</h1>
+            <p style="font-size:1rem; color:#4b5563; margin:0;">Upload your PDF and ask questions – powered by Hugging Face models only.</p>
+        </div>
+        """
+    )
 
     with gr.Row():
-        pdf_upload = gr.File(label="Upload PDF", file_types=[".pdf"])
-        upload_btn = gr.Button("Process PDF")
+        with gr.Column(scale=2):
+            pdf_upload = gr.File(label="Upload PDF", file_types=[".pdf"])
+        with gr.Column(scale=1, min_width=100):
+            upload_btn = gr.Button("🚀 Process PDF", variant="primary")
 
-    chatbot_ui = gr.Chatbot(height=400)
-    user_input = gr.Textbox(label="Ask something about the PDF...", placeholder="Type your question here and hit Enter")
+    chatbot_ui = gr.Chatbot(label="💬 Chat with your PDF", height=450, show_copy_button=True)
+
+    user_input = gr.Textbox(
+        label="Ask a question about your PDF...",
+        placeholder="Type your question and press Enter",
+        show_label=False
+    )
 
     upload_btn.click(upload_pdf, inputs=pdf_upload, outputs=chatbot_ui)
     user_input.submit(chatbot, [user_input, chatbot_ui], [user_input, chatbot_ui])
 
-    gr.Markdown("<footer style='text-align:center; font-size:0.85rem; color:#64748b;'>Built with LLaMA 3 + LangChain on Hugging Face Spaces 🚀</footer>")
+    gr.HTML(
+        """
+        <div style="text-align:center; font-size:0.85rem; color:#9ca3af; margin-top:2em;">
+            Free, fast and simple – no API keys needed!
+        </div>
+        """
+    )
 
 if __name__ == "__main__":
     demo.launch()