app.py

import gradio as gr
import numpy as np
from elevenlabs import clone, generate, set_api_key
from pydub import AudioSegment

# Set up ElevenLabs API Key
set_api_key("73bb17b223e2b0f90a403eaeaa3b4d35")

# Function to convert MP3 to NumPy array
def mp3_to_numpy(file_path):
    audio = AudioSegment.from_mp3(file_path)
    samples = np.array(audio.get_array_of_samples())
    return audio.frame_rate, samples

# Load and clone voice
voice = clone(
    name="Voice Name",
    description="An old American male voice with a slight hoarseness in his throat. Perfect for news.",
    files=["./sample1.mp3", "./sample2.mp3"],
)

def generate_voice_output(text):
    try:
        # Generate audio for the provided text
        audio = generate(text=text, voice=voice)

        # Convert audio for Gradio
        audio_numpy = (44100, np.frombuffer(audio, dtype=np.int16))
        return audio_numpy
    except Exception as e:
        return str(e)

# Set up Gradio components and interface
input_text = gr.Textbox(label="Input Text", lines=2)
output_audio = gr.Audio(label="Generated Voice", type="numpy")

iface = gr.Interface(
    fn=generate_voice_output,
    inputs=input_text,
    outputs=output_audio,
    theme="Monochrome",
)

# Launch Gradio app
iface.launch()