Update app.py

app.py

@@ -1,150 +1,33 @@
-import os
-import re
-import gradio as gr
-from gtts import gTTS
-from datetime import datetime
-from openpyxl import Workbook, load_workbook
-from langchain.memory import ConversationBufferMemory
-
-from transformers import AutoTokenizer, AutoModelForCausalLM
+from flask import Flask, request, jsonify
+from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 import torch
 
-# --- GPT-2 Setup ---
-device = "cuda" if torch.cuda.is_available() else "cpu"
-tokenizer = AutoTokenizer.from_pretrained("distilgpt2")
-model = AutoModelForCausalLM.from_pretrained("distilgpt2").to(device)
-model.eval()
-
-# --- Restaurant menu ---
-MENU = {
-    "Cheeseburger": 5.99,
-    "Fries": 2.99,
-    "Coke": 1.99,
-    "Pizza": 12.99,
-    "Chicken Wings": 7.99,
-    "Salad": 6.99
-}
-
-# --- Memory with LangChain ---
-memory = ConversationBufferMemory(return_messages=True)
-order = []
-customer_name = ""
-
-# --- Excel Setup ---
-EXCEL_FILE = "orders.xlsx"
-def setup_excel():
-    if not os.path.exists(EXCEL_FILE):
-        wb = Workbook()
-        ws = wb.active
-        ws.title = "Orders"
-        ws.append(["Order ID", "Date", "Customer", "Items", "Total", "Time"])
-        wb.save(EXCEL_FILE)
-setup_excel()
-
-def save_to_excel(name, items):
-    wb = load_workbook(EXCEL_FILE)
-    ws = wb.active
-    order_id = f"ORD{ws.max_row:04d}"
-    now = datetime.now()
-    total = sum(qty * MENU[item] for item, qty in items)
-    items_str = ", ".join(f"{qty} x {item}" for item, qty in items)
-    ws.append([order_id, now.strftime("%Y-%m-%d"), name, items_str, f"${total:.2f}", now.strftime("%H:%M:%S")])
-    wb.save(EXCEL_FILE)
-    return order_id
-
-# --- TTS ---
-def clean_text(text):
-    text = re.sub(r"\*\*(.*?)\*\*", r"\1", text)
-    text = re.sub(r"Bot\s*:\s*", "", text, flags=re.IGNORECASE)
-    return text.strip()
-
-def speak(text, filename="response.mp3"):
-    tts = gTTS(text=clean_text(text))
-    tts.save(filename)
-    return filename
-
-# --- Generate GPT-2 response ---
-def generate_response(user_input):
-    global order, customer_name
-
-    menu_str = "\n".join([f"{item}: ${price}" for item, price in MENU.items()])
-    order_summary = ", ".join([f"{qty} x {item}" for item, qty in order]) if order else "No items yet"
-
-    prompt = f"""
-You are a friendly restaurant assistant at 'Systaurant'.
-Here is the menu:
-{menu_str}
-
-Customer name: {customer_name or "Unknown"}
-Current order: {order_summary}
-
-Instructions:
-- Ask for the customer's name if not known
-- Show the menu if asked
-- Detect food items and quantity
-- Provide order summary and ask to confirm
-
-Conversation history:
-"""
-
-    for m in memory.chat_memory.messages:
-        role = "Customer" if m.type == "human" else "Bot"
-        prompt += f"{role}: {m.content}\n"
-
-    prompt += f"Customer: {user_input}\nBot:"
-
-    inputs = tokenizer(prompt, return_tensors="pt", truncation=True).to(device)
-
-    outputs = model.generate(
-        **inputs,
-        max_new_tokens=80,
-        temperature=0.7,
-        top_p=0.9,
-        do_sample=True,
-        pad_token_id=tokenizer.eos_token_id
-    )
-
-    output_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    bot_reply = output_text.split("Bot:")[-1].strip()
-    return bot_reply
-
-# --- Chat handler ---
-def handle_chat(user_input):
-    global customer_name, order
+app = Flask(__name__)
 
-    memory.chat_memory.add_user_message(user_input)
-    bot_reply = generate_response(user_input)
-    memory.chat_memory.add_ai_message(bot_reply)
+# Load OpenChat model and tokenizer
+MODEL_NAME = "openchat/openchat-3.5-0106"
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.float16, device_map="auto")
 
-    if "my name is" in user_input.lower():
-        customer_name = user_input.split("my name is")[-1].strip().split()[0].title()
+chatbot = pipeline("text-generation", model=model, tokenizer=tokenizer)
 
-    for item in MENU:
-        if item.lower() in user_input.lower():
-            qty = 1
-            for word in user_input.lower().split():
-                if word.isdigit():
-                    qty = int(word)
-                    break
-            order.append((item, qty))
+@app.route("/chat", methods=["POST"])
+def chat():
+    data = request.get_json()
+    prompt = data.get("message", "")
+    if not prompt:
+        return jsonify({"error": "Empty message"}), 400
 
-    if "confirm" in user_input.lower() or "yes" in user_input.lower():
-        if customer_name and order:
-            order_id = save_to_excel(customer_name, order)
-            bot_reply += f"\n✅ Your order ID is {order_id}. Thank you for ordering from Systaurant!"
+    system_prompt = "<|system|>\nYou are a helpful assistant for food ordering.\n<|user|>\n" + prompt + "\n<|assistant|>\n"
+    output = chatbot(system_prompt, max_new_tokens=200, do_sample=True, temperature=0.7)[0]["generated_text"]
+    
+    # Extract response after <|assistant|>
+    if "<|assistant|>" in output:
+        reply = output.split("<|assistant|>")[-1].strip()
+    else:
+        reply = output
 
-    audio_file = speak(bot_reply)
-    return bot_reply, audio_file
+    return jsonify({"response": reply})
 
-# --- Gradio UI ---
-gr.Interface(
-    fn=handle_chat,
-    inputs=gr.Textbox(label="👤 You", placeholder="Type your order here..."),
-    outputs=[
-        gr.Textbox(label="🤖 Bot Response"),
-        gr.Audio(label="🔊 Speaking", autoplay=True)
-    ],
-    title="🍔 Systaurant Voice Bot",
-    description="A smart voice-enabled assistant to take food orders.",
-    theme="soft"
-).launch(share=True)
+if __name__ == "__main__":
+    app.run(debug=True)