app.py

import gradio as gr
import torch
from transformers import pipeline
from transformers import AutoProcessor, MusicgenForConditionalGeneration
import scipy
import wave
import io

def generate_music(text):
    processor = AutoProcessor.from_pretrained("facebook/musicgen-small")
    model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-small")
    
    inputs = processor(
        text=text,
        padding=True,
        return_tensors="pt",
    )
    audio_values = model.generate(**inputs, do_sample=True, guidance_scale=3, max_new_tokens=256)
    sampling_rate = model.config.audio_encoder.sampling_rate
    scipy.io.wavfile.write("musicgen_out.wav", rate=sampling_rate, data=audio_values[0, 0].numpy())
    file_path= "musicgen_out.wav"
    try:
        return file_path
    except Exception as e:
        return str(e)

demo = gr.Interface(
  fn=generate_music,
  inputs='text',
  outputs='file',
)

demo.launch(share = True)