app.py

import gradio as gr
import pyttsx3
import speech_recognition as sr
from transformers import pipeline

# Initialize Text-to-Speech engine
engine = pyttsx3.init()
engine.setProperty('rate', 150)

# Load a conversational AI model
chatbot = pipeline("conversational", model="microsoft/DialoGPT-medium")

def tts_response(user_query):
    """Generate AI response, convert to speech, and return text."""
    try:
        response = chatbot(user_query)
        bot_reply = response[0]['generated_text']
        engine.say(bot_reply)
        engine.runAndWait()
        return bot_reply
    except Exception as e:
        return f"Error: {str(e)}"

def stt_to_tts(audio):
    """Convert speech to text, process with AI, then convert response to speech."""
    recognizer = sr.Recognizer()
    try:
        with sr.AudioFile(audio) as source:
            audio_data = recognizer.record(source)
            user_query = recognizer.recognize_google(audio_data)
            return tts_response(user_query)
    except sr.UnknownValueError:
        return "Sorry, I could not understand the audio."
    except sr.RequestError as e:
        return f"Request error from Speech Recognition service; {e}"

# Gradio interface
def gradio_ui():
    with gr.Blocks() as app:
        gr.Markdown("## AI Voice Agent")
        with gr.Tab("Text-to-Speech"):
            user_query = gr.Textbox(label="Your Query")
            response = gr.Textbox(label="AI Response", interactive=False)
            submit_button = gr.Button("Submit")
            submit_button.click(tts_response, inputs=user_query, outputs=response)
        with gr.Tab("Speech-to-Speech"):
            audio_input = gr.Audio(source="microphone", type="filepath", label="Speak Now")
            audio_response = gr.Textbox(label="AI Response")
            audio_button = gr.Button("Submit")
            audio_button.click(stt_to_tts, inputs=audio_input, outputs=audio_response)
    return app

if __name__ == "__main__":
    app = gradio_ui()
    app.launch()