Metadata-Version: 2.4
Name: keep_gpu
Version: 1.0.0
Summary: KeepGPU is a polite GPU keeper with CLI, service, and MCP interfaces
Author-email: Siyuan Wang <sywang0227@gmail.com>, Yaorui Shi <yaoruishi@gmail.com>, Yiduck Liu <geodesicseiran@gmail.com>
Maintainer-email: Siyuan Wang <sywang0227@gmail.com>, Yaorui Shi <yaoruishi@gmail.com>, Yiduck Liu <geodesicseiran@gmail.com>
License-Expression: MIT
Project-URL: bugs, https://github.com/Wangmerlyn/KeepGPU/issues
Project-URL: changelog, https://github.com/Wangmerlyn/KeepGPU/releases
Project-URL: homepage, https://github.com/Wangmerlyn/KeepGPU
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Requires-Python: <3.14,>=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: nvidia-ml-py
Requires-Dist: typer
Requires-Dist: rich>=13.8.0
Requires-Dist: torch
Provides-Extra: dev
Requires-Dist: coverage; extra == "dev"
Requires-Dist: mypy; extra == "dev"
Requires-Dist: pytest; extra == "dev"
Requires-Dist: ruff; extra == "dev"
Requires-Dist: black; extra == "dev"
Requires-Dist: pre-commit; extra == "dev"
Requires-Dist: bump-my-version; extra == "dev"
Requires-Dist: setuptools<82,>=77; extra == "dev"
Provides-Extra: rocm
Provides-Extra: macm
Requires-Dist: psutil; extra == "macm"
Dynamic: license-file

# KeepGPU

[![PyPI Version](https://img.shields.io/pypi/v/keep-gpu.svg)](https://pypi.org/project/keep-gpu/)
[![Docs Status](https://readthedocs.org/projects/keepgpu/badge/?version=latest)](https://keepgpu.readthedocs.io/en/latest/?version=latest)
[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.17129114.svg)](https://doi.org/10.5281/zenodo.17129114)

KeepGPU is a small, polite GPU keeper for shared machines. It allocates only
when utilization backoff permits, keeps the requested VRAM signal lightweight,
and releases cleanly when you are done.

## Quick Start

```console
pip install keep-gpu
keep-gpu --gpu-ids 0 --vram 1GiB --interval 60
```

Press `Ctrl+C` to release. CUDA, ROCm/HIP, and Apple Silicon MPS are supported
when PyTorch can see the target device.

## Documentation

- [Documentation](https://keepgpu.readthedocs.io/en/latest/) is the full guide hub.
- [Getting Started](https://keepgpu.readthedocs.io/en/latest/getting-started/) covers install options and the first hardware check.
- [Citation](https://keepgpu.readthedocs.io/en/latest/citation/) has the BibTeX entry.
