Update app.py

app.py

@@ -1,20 +1,25 @@
-from transformers import pipeline
 import gradio as gr
 
-# Use a very small, CPU-friendly model
-chatbot_pipeline = pipeline("text-generation", model="google/flan-t5-small")
+# Use Hugging Face hosted inference API (no heavy model in the Space)
+import requests
+
+API_URL = "https://api-inference.huggingface.co/models/google/flan-t5-small"
+headers = {"Authorization": f"Bearer hf_XXXXXXXXXXXXXXXXXXXX"}  # optional if public model
 
-# Chat function
 def chat(message, history=[]):
-    # Generate reply
-    response = chatbot_pipeline(message, max_new_tokens=50)
-    reply = response[0]['generated_text'].strip()
+    payload = {"inputs": message}
+    try:
+        response = requests.post(API_URL, headers=headers, json=payload, timeout=30)
+        result = response.json()
+        reply = result[0]['generated_text']
+    except Exception as e:
+        reply = f"⚠️ Error: {str(e)}"
     history.append((message, reply))
     return history, history
 
 # Gradio UI
 with gr.Blocks() as demo:
-    gr.Markdown("## 🤖 Permanent AI Chatbot (CPU-Friendly)")
+    gr.Markdown("## 🤖 Permanent CPU-Friendly AI Chatbot")
     chat_ui = gr.Chatbot()
     msg = gr.Textbox(label="Type your message here")
     clear = gr.Button("Clear Chat")