app.py

import torch
import torchaudio
from einops import rearrange
from stable_audio_tools import get_pretrained_model
from stable_audio_tools.inference.generation import generate_diffusion_cond
import gradio as gr

device = "cuda" if torch.cuda.is_available() else "cpu"

# تحميل النموذج
model, model_config = get_pretrained_model("stabilityai/stable-audio-open-small")
sample_rate = model_config["sample_rate"]
sample_size = model_config["sample_size"]
model = model.to(device)

def generate_audio(prompt, duration):
    conditioning = [{
        "prompt": prompt,
        "seconds_total": duration
    }]
    output = generate_diffusion_cond(
        model,
        steps=8,
        conditioning=conditioning,
        sample_size=sample_size,
        sampler_type="pingpong",
        device=device
    )
    output = rearrange(output, "b d n -> d (b n)")
    output = output.to(torch.float32).div(torch.max(torch.abs(output))).clamp(-1, 1).mul(32767).to(torch.int16).cpu()
    torchaudio.save("output.wav", output, sample_rate)
    return "output.wav"

gr.Interface(
    fn=generate_audio,
    inputs=[
        gr.Textbox(label="Prompt", placeholder="مثال: 128 BPM tech house drum loop"),
        gr.Slider(1, 11, step=1, label="مدة الصوت بالثواني")
    ],
    outputs=gr.Audio(label="الصوت الناتج"),
    title="تجربة Stable Audio Open Small",
    description="أدخل وصفًا صوتيًا، وجرب توليد مقطع صوتي من النموذج."
).launch()