app.py

import gradio as gr
import soundfile as sf
from tango import Tango
import IPython.display as ipd

# Initialize the Tango model
tango = Tango("declare-lab/tango-full-ft-audiocaps")

# Function to generate audio from text prompt
def generate_audio(prompt):
    # Generate audio from the prompt
    audio = tango.generate(prompt)
    # Save the audio to a file
    file_path = f"{prompt}.wav"
    sf.write(file_path, audio, samplerate=16000)
    return file_path

# Gradio interface
interface = gr.Interface(
    fn=generate_audio,
    inputs="text",
    outputs="audio",
    title="Audio Generation with Tango",
    description="Enter a prompt to generate an audio description using the Tango model."
)

# Launch the app
interface.launch()