import gradio as gr
from transformers import pipeline
import librosa
import numpy as np
from scipy.io.wavfile import write

def narrate_text(text):
    narrator = pipeline("text-to-speech", model="kakao-enterprise/vits-ljs")
    narrated_text = narrator(text)
    audio_data = narrated_text["audio"][0]
    sampling_rate = narrated_text["sampling_rate"]
    audio_data = librosa.resample(audio_data, orig_sr=sampling_rate, target_sr=22050)
    audio_data = (audio_data * 32767).astype(np.int16)
    with open("output.wav", "wb") as f:
        write(f, 22050, audio_data)
    return "output.wav"

gr.Interface(
    fn=narrate_text, 
    inputs=gr.Textbox(label="Enter Text Here",
                       lines=7), 
    outputs="audio", 
    title="Speak Out Loud - Text to Speech Assistant", 
    description="Upload the text that you want to hear out loud!!!", 
    article = 
                        '''<div>
                            <p style="text-align: center"> All you need to do is to upload your text and hit submit, then wait for compiling. After that click on Play/Pause for listing to the audio. The audio is saved in a wav format.</p>
                        </div>''',
).launch()