Update app.py

app.py

@@ -1,64 +1,101 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
+from llama_index.core import VectorStoreIndex, SimpleDirectoryReader
+from llama_index.llms.huggingface import HuggingFaceInferenceAPI
+from llama_index.embeddings.huggingface import HuggingFaceEmbedding
+import os
+from pathlib import Path
+import shutil
+
+# Set up Hugging Face API token (replace with your token or set in HF Spaces secrets)
+os.environ["HF_TOKEN"] = os.getenv("HF_TOKEN", "your-huggingface-api-token")
+
+# Initialize Mistral model for text generation
+llm = HuggingFaceInferenceAPI(
+    model_name="mistralai/Mistral-7B-Instruct-v0.3",
+    api_key=os.environ["HF_TOKEN"]
 )
 
+# Initialize embedding model for document indexing
+embed_model = HuggingFaceEmbedding(model_name="sentence-transformers/all-MiniLM-L6-v2")
+
+# Directory to store uploaded documents
+DOCS_DIR = "policy_docs"
+if not os.path.exists(DOCS_DIR):
+    os.makedirs(DOCS_DIR)
+
+# Global variable to store the index
+index = None
+
+def process_document(file):
+    """Process uploaded policy document and create an index."""
+    global index
+    try:
+        # Clear previous documents
+        if os.path.exists(DOCS_DIR):
+            shutil.rmtree(DOCS_DIR)
+        os.makedirs(DOCS_DIR)
+
+        # Save uploaded file
+        file_path = os.path.join(DOCS_DIR, file.name)
+        shutil.copy(file.name, file_path)
+
+        # Load documents
+        documents = SimpleDirectoryReader(DOCS_DIR).load_data()
+
+        # Create index
+        index = VectorStoreIndex.from_documents(documents, embed_model=embed_model)
+
+        return "Document processed successfully! You can now ask questions about the policy."
+    except Exception as e:
+        return f"Error processing document: {str(e)}"
+
+def policy_chat(message, history):
+    """Handle user queries about the policy with context-aware responses."""
+    global index
+    if index is None:
+        return "Please upload a policy document first."
+
+    try:
+        # Create query engine
+        query_engine = index.as_query_engine(llm=llm, similarity_top_k=3)
+
+        # Craft a prompt that considers financial and other perspectives
+        prompt = (
+            f"You are a policy expert. A user has asked: '{message}'. "
+            "Based on the uploaded policy document, provide a concise response addressing the user's query. "
+            "Consider financial, practical, and other relevant perspectives. "
+            "If the query is about whether the policy works for the user, evaluate eligibility and benefits clearly. "
+            "Keep the response brief and clear."
+        )
+
+        # Query the index
+        response = query_engine.query(prompt)
+
+        # Append to history
+        history.append({"role": "user", "content": message})
+        history.append({"role": "assistant", "content": str(response)})
+        return history
+    except Exception as e:
+        return f"Error processing query: {str(e)}"
+
+# Gradio interface
+with gr.Blocks() as demo:
+    gr.Markdown("# Policy Bot")
+    gr.Markdown("Upload a policy document (PDF, text, etc.) and ask questions about it. The bot will analyze the policy and respond from financial, practical, and other perspectives.")
+
+    # File upload for policy documents
+    file_input = gr.File(label="Upload Policy Document")
+    upload_button = gr.Button("Process Document")
+    upload_output = gr.Textbox(label="Upload Status")
+
+    # Chat interface
+    chatbot = gr.Chatbot(type="messages")
+    msg = gr.Textbox(placeholder="Ask about the policy (e.g., 'Will this policy cover my medical expenses?')")
+    clear = gr.ClearButton([msg, chatbot])
+
+    # Event handlers
+    upload_button.click(process_document, inputs=file_input, outputs=upload_output)
+    msg.submit(policy_chat, inputs=[msg, chatbot], outputs=chatbot)
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()