app.py

import os
import gradio as gr
import openai
from openai import OpenAI
import speech_recognition as sr
import threading
import time

# Initialize OpenAI client with API key from environment variable
client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))

# Create an assistant
assistant = client.beta.assistants.create(
    name="Pronunciation Assistant",
    instructions="You are a helpful pronunciation assistant. You compare the generated text with the user's transcription and then provide feedback on how the user can improve their pronunciation accordingly. You also single out specific words they pronounced incorrectly and give tips on how to improve like for example 'schedule' can be pronounced as 'sked-jool'.",
    model="gpt-4-1106-preview"
)

def generate_text():
    response = client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[
            {"role": "system", "content": "Generate a short paragraph (2-3 sentences) for an English learner to read aloud."},
            {"role": "user", "content": "Create a practice text."}
        ]
    )
    return response.choices[0].message.content

def get_pronunciation_feedback(original_text, transcription):
    thread = client.beta.threads.create()
    
    message = client.beta.threads.messages.create(
        thread_id=thread.id,
        role="user",
        content=f"Original text: '{original_text}'\nTranscription: '{transcription}'\nProvide pronunciation feedback."
    )
    
    run = client.beta.threads.runs.create(
        thread_id=thread.id,
        assistant_id=assistant.id
    )
    
    while run.status != "completed":
        time.sleep(1)
        run = client.beta.threads.runs.retrieve(thread_id=thread.id, run_id=run.id)
    
    messages = client.beta.threads.messages.list(thread_id=thread.id)
    return messages.data[0].content[0].text.value

def transcribe_audio_realtime(audio):
    recognizer = sr.Recognizer()
    with sr.AudioFile(audio) as source:
        audio_data = recognizer.record(source)
    try:
        return recognizer.recognize_google(audio_data)
    except sr.UnknownValueError:
        return "Could not understand audio"
    except sr.RequestError:
        return "Could not request results from the speech recognition service"

def practice_pronunciation(audio):
    original_text = generate_text()
    transcription = transcribe_audio_realtime(audio)
    feedback = get_pronunciation_feedback(original_text, transcription)
    return original_text, transcription, feedback

# Gradio interface
demo = gr.Interface(
    fn=practice_pronunciation,
    inputs=[
        gr.Audio(type="filepath")  # Removed 'source="microphone"'
    ],
    outputs=[
        gr.Textbox(label="Text to Read"),
        gr.Textbox(label="Your Transcription"),
        gr.Textbox(label="Pronunciation Feedback")
    ],
    title="Pronunciation Practice Tool",
    description="Read the generated text aloud. The system will transcribe your speech and provide pronunciation feedback.",
    live=True
)

# Launch the app
if __name__ == "__main__":
    demo.launch()