app.py

"""Gradio demo for denoisers."""
import gradio as gr
import numpy as np
import torch
import torchaudio
from denoisers import UNet1DModel, WaveUNetModel
from tqdm import tqdm

MODELS = [
    "wrice/unet1d-vctk-48khz",
    "wrice/waveunet-vctk-48khz",
    "wrice/waveunet-vctk-24khz",
]


def denoise(model_name, inputs):
    """Denoise audio."""
    if "unet1d" in model_name:
        model = UNet1DModel.from_pretrained(model_name)
    else:
        model = WaveUNetModel.from_pretrained(model_name)
    sr, audio = inputs
    audio = torch.from_numpy(audio)[None]
    audio = audio / 32768.0
    audio = audio.permute(0, 2, 1)

    print(f"Audio shape: {audio.shape}")
    print(f"Sample rate: {sr}")

    if audio.shape[1] > 1:
        audio = audio.mean(1, keepdim=True)

    print(f"Audio shape: {audio.shape}")

    if sr != model.config.sample_rate:
        audio = torchaudio.functional.resample(audio, sr, model.config.sample_rate)

    chunk_size = model.config.max_length

    padding = abs(audio.size(-1) % chunk_size - chunk_size)
    padded = torch.nn.functional.pad(audio, (0, padding))

    clean = []
    for i in tqdm(range(0, padded.shape[-1], chunk_size)):
        audio_chunk = padded[:, :, i : i + chunk_size]
        with torch.no_grad():
            clean_chunk = model(audio_chunk).logits
        clean.append(clean_chunk.squeeze(0))

    denoised = torch.concat(clean, 1)[:, : audio.shape[-1]].clamp(-1.0, 1.0)
    denoised = (denoised * 32767.0).numpy().astype(np.int16)

    print(f"Denoised shape: {denoised.shape}")

    return model.config.sample_rate, denoised.transpose()


iface = gr.Interface(
    fn=denoise,
    inputs=[gr.Dropdown(choices=MODELS, value=MODELS[0]), "audio"],
    outputs="audio",
)
iface.launch()