app.py

import streamlit as st
from transformers import pipeline
import scipy.io.wavfile
import numpy as np
import os

# Load the text-to-speech pipeline
synthesizer = pipeline("text-to-speech", model="suno/bark")

# Streamlit app
st.title("Text-to-Speech with Suno/Bark Model")

# Text input from the user
text = st.text_area("Enter the text you want to convert to speech:", "")

if st.button("Generate Speech"):
    if text:
        with st.spinner("Generating speech..."):
            # Generate speech from text
            speech = synthesizer(text, forward_params={"do_sample": True})
            
            # Save the speech as a WAV file
            output_path = "output.wav"
            scipy.io.wavfile.write(output_path, rate=speech["sampling_rate"], data=np.array(speech["audio"]))
            
            st.success("Speech generated successfully!")
            
            # Provide options to play and download the audio file
            audio_file = open(output_path, "rb").read()
            st.audio(audio_file, format="audio/wav")
            st.download_button(label="Download WAV file", data=audio_file, file_name="output.wav", mime="audio/wav")
    else:
        st.error("Please enter some text to generate speech.")