README.md

metadata

title: Respite API
emoji: 🤖
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 6.3.0
app_file: app.py
pinned: false

Respite API

A CPU-based Hugging Face Space for generating music using Stability AI's Stable Audio 3 Small model.

Features

• Model: small-music (433M parameters)
• Duration: 1–120 seconds
• Sample rate: 44.1 kHz stereo
• Runs on CPU (only a 1s ZeroGPU touch at startup; no GPU quota used per request)
• Gradio 6.3.0 interface

This Space is configured for CPU inference. ZeroGPU Spaces fail to boot unless at least one spaces.GPU function is called, so app.py calls one minimal @spaces.GPU(duration=1) function a single time at startup. Keep every other function (especially generation) undecorated so no GPU allocation happens per request.

API discovery endpoints

Respite API exposes machine-readable API metadata alongside the Gradio UI:

• GET /api/resources lists callable capabilities, including each resource's method, endpoint, inputs, and output.
• GET /api/specs lists server-wide specifications such as version, supported content types, limits, discovery URLs, CPU cores, total RAM, and storage total/used/free space.

Both endpoints return JSON and currently require no authentication. The metadata is designed to grow as additional AI capabilities are added.

Usage

1. Enter a text prompt describing the music you want
2. Adjust duration, steps, CFG scale, and seed
3. Click Generate and wait for the audio

Setup

pip install -r requirements.txt
python app.py

License

This project uses the Stable Audio 3 model. Please refer to Stability AI's license for usage terms.