Update app.py

app.py

@@ -1,11 +1,10 @@
-from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 import gradio as gr
+from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 import torch
 
-# Agent 1: Intent Classifier
+# ---------------- Agent 1: Intent Classifier ----------------
 intent_classifier = pipeline("text-classification", model="MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli")
 
-
 def detect_intent(text):
     labels = {
         "weather": "The user wants to know the weather.",
@@ -21,41 +20,40 @@ def detect_intent(text):
             best_intent = label
     return best_intent
 
-# Agent 2: Domain Logic
+# ---------------- Agent 2: Domain Logic ----------------
 def handle_logic(intent):
     if intent == "weather":
-        return "It’s currently sunny and 26°C."
+        return "It's sunny and 26°C today."
     elif intent == "faq":
-        return "Please click on 'Forgot Password' to reset it."
+        return "To reset your password, use the 'Forgot Password' option."
     else:
-        return "Haha, that’s funny! Tell me more."
+        return "That's great! Anything else you'd like to talk about?"
 
-# Agent 3: NLG with DialoGPT
+# ---------------- Agent 3: Natural Language Generation ----------------
 tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-small")
 model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-small")
 
-def generate_reply(text):
-    input_ids = tokenizer.encode(text + tokenizer.eos_token, return_tensors='pt')
-    output = model.generate(input_ids, max_length=100, pad_token_id=tokenizer.eos_token_id)
-    reply = tokenizer.decode(output[:, input_ids.shape[-1]:][0], skip_special_tokens=True)
-    return reply
+def generate_reply(prompt):
+    input_ids = tokenizer.encode(prompt + tokenizer.eos_token, return_tensors='pt')
+    output_ids = model.generate(input_ids, max_length=100, pad_token_id=tokenizer.eos_token_id)
+    response = tokenizer.decode(output_ids[:, input_ids.shape[-1]:][0], skip_special_tokens=True)
+    return response
 
-# Combined chatbot
+# ---------------- Chatbot Pipeline ----------------
 def chatbot(user_input):
     intent = detect_intent(user_input)
     logic = handle_logic(intent)
-    reply = generate_reply(logic)
-    return reply
-
-# Gradio interface
-iface = gr.Interface(fn=chatbot,
-                     inputs=gr.Textbox(lines=2, placeholder="Type your message..."),
-                     outputs="text",
-                     title="Three-Agent Hugging Face Chatbot",
-                     description="Intent detection + domain logic + natural generation")
-
-iface.launch()
-
+    response = generate_reply(logic)
+    return response
+
+# ---------------- Gradio UI ----------------
+gr.Interface(
+    fn=chatbot,
+    inputs=gr.Textbox(label="User Input"),
+    outputs=gr.Textbox(label="Chatbot Response"),
+    title="3-Agent Chatbot",
+    description="Intent Detection → Domain Logic → Natural Language Generation"
+).launch()