---
title: AudioSR
emoji: 📶
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 6.17.3
python_version: "3.10"
preload_from_hub:
  - haoheliu/audiosr_basic pytorch_model.bin 74a47f49061a1e788e968cc43ad45c0b6243f37d
app_file: app.py
pinned: false
license: mit
short_description: Upsamples any audio to 48 kHz
tags:
  - category:enhancement
  - subcategory:bandwidth-extension
  - domain:music
  - domain:speech
  - domain:general-audio
  - input:audio
  - input:text
  - output:audio
  - sample-rate:48000
  - channels:mono
---

Upsamples any audio to 48 kHz with audio super-resolution.

Gradio is held at 6.17.3, the last release that allows the transformers 4.30.2 that audiosr pins.

A HARP deployment. See https://github.com/TEAMuP-dev/HARP