Update app.py

app.py

@@ -1,7 +1,7 @@
-import gradio as gr
+import streamlit as st
 import torch
 import fitz  # PyMuPDF
-from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+from transformers import AutoTokenizer, pipeline, AutoModelForSeq2SeqLM # Import for T5 model
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain_community.vectorstores import FAISS
 from langchain_community.embeddings import HuggingFaceEmbeddings
@@ -9,161 +9,202 @@ from langchain_community.llms.huggingface_pipeline import HuggingFacePipeline
 from langchain.chains import ConversationalRetrievalChain
 from langchain.memory import ConversationBufferMemory
 from langchain.prompts import PromptTemplate
+
+# For Fairness Audit
 import pandas as pd
 from aif360.datasets import StandardDataset
 from aif360.metrics import BinaryLabelDatasetMetric
-import time
 
-# --- Caching (simple global variables for Gradio) ---
-_llm = None
-_qa_chain = None
+# --- Page Configuration ---
+st.set_page_config(
+    page_title="Sahay AI 🇮🇳",
+    page_icon="🤖",
+    layout="wide",
+    initial_sidebar_state="expanded"
+)
 
-# --- Core AI and Data Processing Functions (Unchanged) ---
+# --- Caching for Performance ---
+@st.cache_resource
 def load_llm():
-    """Loads the IBM Granite LLM, forcing it to use the GPU."""
-    global _llm
-    if _llm is None:
-        print("Loading LLM for the first time...")
-        llm_model_name = "ibm-granite/granite-3.3-8b-instruct"
-        
-        if not torch.cuda.is_available():
-            raise RuntimeError("ZeroGPU requires a GPU. Please ensure hardware is set correctly.")
+    """
+    Loads a smaller, CPU-friendly model (FLAN-T5-Base) for better performance 
+    on the free Hugging Face Spaces hardware.
+    """
+    # Using a smaller, CPU-compatible model to ensure the app is fast and responsive.
+    llm_model_name = "google/flan-t5-base" 
+    
+    tokenizer = AutoTokenizer.from_pretrained(llm_model_name)
+    # Use AutoModelForSeq2SeqLM for T5 models
+    model = AutoModelForSeq2SeqLM.from_pretrained(llm_model_name)
+    
+    pipe = pipeline(
+        "text2text-generation", # T5 models use this pipeline type
+        model=model, 
+        tokenizer=tokenizer, 
+        max_length=512
+    )
+    return HuggingFacePipeline(pipeline=pipe)
+
+@st.cache_resource
+def load_and_process_pdf(pdf_path):
+    """Loads, chunks, and embeds the PDF into a FAISS vector store using IBM's model."""
+    try:
+        doc = fitz.open(pdf_path)
+        text = "".join(page.get_text() for page in doc)
+        if not text:
+            st.error("Could not extract text from the PDF.")
+            return None
+    except Exception as e:
+        st.error(f"Error reading PDF file: {e}. Make sure 'PMKisanSamanNidhi.PDF' is uploaded to the Space.")
+        return None
 
-        model = AutoModelForCausalLM.from_pretrained(
-            llm_model_name, torch_dtype=torch.bfloat16, load_in_4bit=True
-        )
-        tokenizer = AutoTokenizer.from_pretrained(llm_model_name)
-        
-        pipe = pipeline(
-            "text-generation", model=model, tokenizer=tokenizer, 
-            max_new_tokens=512, temperature=0.1, device=0
-        )
-        _llm = HuggingFacePipeline(pipeline=pipe)
-    return _llm
-
-def load_and_process_pdf(pdf_path="PMKisanSamanNidhi.PDF"):
-    """Loads and processes the PDF into a FAISS vector store."""
-    print("Loading and processing PDF...")
-    doc = fitz.open(pdf_path)
-    text = "".join(page.get_text() for page in doc)
     text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=150)
     docs = text_splitter.create_documents([text])
-    embedding_model = HuggingFaceEmbeddings(model_name="ibm-granite/granite-embedding-278m-multilingual")
+    
+    # Still using the powerful IBM embedding model for multilingual understanding
+    model_name = "ibm-granite/granite-embedding-278m-multilingual"
+    embedding_model = HuggingFaceEmbeddings(model_name=model_name)
+    
     vector_db = FAISS.from_documents(docs, embedding_model)
     return vector_db
 
-def create_conversational_chain(llm, vector_db):
+# --- Conversational Chain ---
+def create_conversational_chain(_llm, _vector_db):
     """Creates the LangChain conversational retrieval chain."""
-    prompt_template = """You are a polite and professional AI assistant for the PM-KISAN scheme... (rest of prompt)"""
+    prompt_template = """You are a polite and professional AI assistant for the PM-KISAN scheme.
+    Use the following context to answer the user's question precisely.
+    If the question is not related to the provided context, you must state: "I can only answer questions related to the PM-KISAN scheme."
+    Do not make up information.
+    
+    Context: {context}
+    Question: {question}
+    
+    Helpful Answer:"""
+    
     QA_PROMPT = PromptTemplate(template=prompt_template, input_variables=["context", "question"])
     memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True, output_key='answer')
+    
     chain = ConversationalRetrievalChain.from_llm(
-        llm=llm, retriever=vector_db.as_retriever(), memory=memory,
-        return_source_documents=True, combine_docs_chain_kwargs={"prompt": QA_PROMPT}
+        llm=_llm,
+        retriever=_vector_db.as_retriever(search_kwargs={'k': 3}),
+        memory=memory,
+        return_source_documents=True,
+        combine_docs_chain_kwargs={"prompt": QA_PROMPT}
     )
     return chain
 
-def get_qa_chain():
-    """Initializes and returns the QA chain."""
-    global _qa_chain
-    if _qa_chain is None:
-        llm = load_llm()
-        vector_db = load_and_process_pdf()
-        _qa_chain = create_conversational_chain(llm, vector_db)
-    return _qa_chain
-
+# --- IBM AIF360 Fairness Audit ---
 def run_fairness_audit():
-    """Performs the fairness audit and returns a formatted string."""
-    df_display = pd.DataFrame({
+    """Performs and displays a simulated fairness audit."""
+    st.subheader("🤖 IBM AIF360 - Fairness Audit")
+    st.info("""
+    This is a simulation to demonstrate how we can check for bias in our information retriever. 
+    A fair system should provide equally good information to all demographic groups.
+    """)
+    test_data = {
         'query': ["loan for my farm", "help for my crops", "scheme for women", "grant for female farmer"],
         'gender_text': ['male', 'male', 'female', 'female'],
         'expected_doc': ['doc1', 'doc1', 'doc2', 'doc2']
-    })
+    }
+    df_display = pd.DataFrame(test_data)
+
     def simulate_retriever(query):
         return "doc2" if "women" in query or "female" in query else "doc1"
     df_display['retrieved_doc'] = df_display['query'].apply(simulate_retriever)
     df_display['favorable_outcome'] = (df_display['retrieved_doc'] == df_display['expected_doc']).astype(int)
-    
+
     df_for_aif = pd.DataFrame()
     df_for_aif['gender'] = df_display['gender_text'].map({'male': 1, 'female': 0})
     df_for_aif['favorable_outcome'] = df_display['favorable_outcome']
     
-    aif_dataset = StandardDataset(df_for_aif, label_name='favorable_outcome', favorable_classes=[1],
-                                  protected_attribute_names=['gender'], privileged_classes=[[1]])
+    aif_dataset = StandardDataset(df_for_aif, 
+                                  label_name='favorable_outcome', 
+                                  favorable_classes=[1],
+                                  protected_attribute_names=['gender'], 
+                                  privileged_classes=[[1]])
+    
     metric = BinaryLabelDatasetMetric(aif_dataset, unprivileged_groups=[{'gender': 0}], privileged_groups=[{'gender': 1}])
     spd = metric.statistical_parity_difference()
+
+    st.markdown("---")
+    col1, col2 = st.columns(2)
+    with col1:
+        st.metric(label="**Metric: Statistical Parity Difference (SPD)**", value=f"{spd:.4f}")
+    with col2:
+        st.success("An SPD of **0.0** indicates perfect fairness in this simulation.")
     
-    report = f"""
-    ### 🤖 IBM AIF360 - Fairness Audit Results
-    **Metric: Statistical Parity Difference (SPD):** {spd:.4f}
-    **Interpretation:** An SPD of 0.0 indicates perfect fairness in this simulation.
-    ---
-    **Raw Audit Data:**
-    ```
-    {df_display.to_string()}
-    ```
-    """
-    return report
-
-# --- Gradio UI ---
-def chat_response(message, history):
-    """Handles the user's message and returns the bot's response."""
-    qa_chain = get_qa_chain()
-    result = qa_chain.invoke({"question": message})
-    response = result["answer"]
-    
-    # Add sources to the response
-    source_docs = result.get("source_documents", [])
-    if source_docs:
-        response += "\n\n--- \n*Sources used to generate this answer:*"
-        for i, doc in enumerate(source_docs):
-            cleaned_content = ' '.join(doc.page_content.split())
-            response += f"\n\n> **Source {i+1}:** \"{cleaned_content[:150]}...\""
-            
-    # Yield response for streaming effect
-    for i in range(len(response)):
-        time.sleep(0.005)
-        yield response[:i+1]
-
-# Initialize the AI model on startup
-print("Initializing AI Chain...")
-get_qa_chain()
-print("AI Chain Ready.")
-
-with gr.Blocks(theme=gr.themes.Soft(), title="Sahay AI") as demo:
-    gr.Markdown("# 🇮🇳 Chat with Sahay AI 💬")
-    gr.Markdown("Your trusted guide to the PM-KISAN scheme, powered by IBM Granite.")
+    with st.expander("Show Raw Audit Data"):
+        st.dataframe(df_display)
+
+# --- Main Application UI ---
+if __name__ == "__main__":
     
-    with gr.Row():
-        with gr.Column(scale=3):
-            chatbot = gr.Chatbot(
-                value=[[None, "Welcome! Ask me a question about the PM-KISAN scheme."]],
-                label="Conversation",
-                bubble_full_width=False,
-                avatar_images=(None, "https://upload.wikimedia.org/wikipedia/commons/5/51/IBM_logo.svg")
-            )
-            msg = gr.Textbox(label="Your Question", placeholder="e.g., Who is eligible for this scheme?")
-            submit_btn = gr.Button("Send", variant="primary")
-            
-        with gr.Column(scale=1):
-            gr.Markdown("### Actions & Connect")
-            audit_button = gr.Button("Run Fairness Audit")
-            audit_report = gr.Markdown(visible=False)
-            
-            whatsapp_link = "https://wa.me/15551234567?text=Hello%20Sahay%20AI!"
-            gr.Markdown(f"📱 [Chat on WhatsApp]({whatsapp_link})")
-            gr.Markdown("⭐ [View Project on GitHub](https://github.com)")
-
-    # Event handlers
-    msg.submit(chat_response, [msg, chatbot], chatbot)
-    submit_btn.click(chat_response, [msg, chatbot], chatbot)
-    msg.submit(lambda: "", None, msg) # Clear textbox on submit
-    submit_btn.click(lambda: "", None, msg) # Clear textbox on submit
+    with st.sidebar:
+        st.image("https://upload.wikimedia.org/wikipedia/commons/5/51/IBM_logo.svg", width=100)
+        st.title("🇮🇳 Sahay AI")
+        st.markdown("### About")
+        st.markdown("An AI assistant for the **PM-KISAN** scheme, built with IBM's multilingual embedding model.")
+        st.markdown("---")
+        
+        st.markdown("### Actions")
+        if st.button("Run Fairness Audit", use_container_width=True):
+            st.session_state.run_audit = True
+        st.markdown("---")
+
+        st.markdown("### Connect")
+        st.markdown("📱 [Try the WhatsApp Bot](https://wa.me/15551234567?text=Hello%20Sahay%20AI!)") # Replace with your number
+        st.markdown("⭐ [View Project on GitHub](https://github.com)")
+        st.markdown("---")
+
+    st.header("Chat with Sahay AI 💬")
+    st.markdown("Your trusted guide to the PM-KISAN scheme.")
+
+    if st.session_state.get('run_audit', False):
+        run_fairness_audit()
+        st.session_state.run_audit = False
     
-    def show_audit():
-        report = run_fairness_audit()
-        return gr.update(value=report, visible=True)
-    audit_button.click(show_audit, outputs=audit_report)
+    if "messages" not in st.session_state:
+        st.session_state.messages = []
+        st.session_state.messages.append({
+            "role": "assistant", 
+            "content": "Welcome! How can I help you understand the PM-KISAN scheme today? You can ask me questions like:\n- What is this scheme about?\n- Who is eligible?\n- *इस योजना के लिए कौन पात्र है?*"
+        })
 
-if __name__ == "__main__":
-    demo.launch()
+    if "qa_chain" not in st.session_state:
+        with st.spinner("🚀 Initializing Sahay AI... This may take a moment."):
+            llm = load_llm()
+            vector_db = load_and_process_pdf("PMKisanSamanNidhi.PDF")
+            if vector_db:
+                st.session_state.qa_chain = create_conversational_chain(llm, vector_db)
+            else:
+                st.error("Application could not start. Please check the PDF file is uploaded correctly.")
+                st.stop()
+
+    for message in st.session_state.messages:
+        with st.chat_message(message["role"]):
+            st.markdown(message["content"])
+
+    if prompt := st.chat_input("Ask a question about the PM-KISAN scheme..."):
+        st.session_state.messages.append({"role": "user", "content": prompt})
+        with st.chat_message("user"):
+            st.markdown(prompt)
+
+        with st.chat_message("assistant"):
+            with st.spinner("🧠 Thinking..."):
+                if "qa_chain" in st.session_state:
+                    result = st.session_state.qa_chain.invoke({"question": prompt})
+                    response = result["answer"]
+                    source_docs = result.get("source_documents", [])
+                    
+                    if source_docs:
+                        response += "\n\n--- \n*Sources used to generate this answer:*"
+                        for i, doc in enumerate(source_docs):
+                            cleaned_content = ' '.join(doc.page_content.split())
+                            response += f"\n\n> **Source {i+1}:** \"{cleaned_content[:150]}...\""
+                    
+                    st.markdown(response)
+                else:
+                    response = "Sorry, the application is not properly initialized."
+                    st.error(response)
+        
+        st.session_state.messages.append({"role": "assistant", "content": response})