Skip to main content

一眼看清谁在霸占 GPU — a tiny Linux CLI that prints which user is using how much GPU memory.

Project description

whovidia

一眼看清谁在霸占 GPU。 A tiny Linux CLI that prints which user is using how much GPU memory.

┌─────────┬──────────┬─────────┬─────────────────────────────────────────────┐
│   PID   │   用户   │   占用  │                    跑的啥                   │
╞═════════╪══════════╪═════════╪═════════════════════════════════════════════╡
│ 2422445 │ rui.zhao │ 17.8 GB │ reconstruction_pipeline_node(已经跑 6.0 小时)│
│ 1719272 │ root     │  5.0 GB │ carter.py(已经跑 9.5 小时)                  │
└─────────┴──────────┴─────────┴─────────────────────────────────────────────┘

安装 / Install

pip install whovidia

需要 Python 3.9+ 和 nvidia-smi(装了 NVIDIA 驱动的 Linux)。

Requires Python 3.9+ and nvidia-smi on PATH. Linux only. Windows is unsupported because nvidia-smi --query-compute-apps on Windows includes graphics processes (DWM, browsers, games) without usable memory accounting, so the table can't be trusted. macOS has no NVIDIA GPUs.

用法 / Usage

whovidia

也可以这么跑 / or run as a module:

python -m whovidia

输出按显存占用从大到小排序,包括: Output is sorted by GPU memory usage (descending) and shows:

  • PID — 进程号
  • 用户 / user — 占用 GPU 的系统用户
  • 占用 / VRAM — 显存占用 (MB / GB)
  • 跑的啥 / process — 进程命令名(自动解开 python script.py 这种壳),并附上已运行时长

颜色输出在 TTY 下自动启用;可设置 NO_COLOR=1 关闭。 ANSI colors are enabled automatically on a TTY; set NO_COLOR=1 to disable.

工作原理 / How it works

  1. 调用 nvidia-smi --query-compute-apps=pid,used_memory,process_name 拿到所有占用 GPU 的进程;
  2. ps -o user=,etimes=,args= 补上用户名、运行时长、命令行;
  3. 渲染成一张 Unicode 表格,按显存排序。

整个工具零运行时依赖 — 只用 Python 标准库。 Zero runtime dependencies — pure Python stdlib.

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

whovidia-0.1.0.tar.gz (6.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

whovidia-0.1.0-py3-none-any.whl (7.0 kB view details)

Uploaded Python 3

File details

Details for the file whovidia-0.1.0.tar.gz.

File metadata

  • Download URL: whovidia-0.1.0.tar.gz
  • Upload date:
  • Size: 6.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.5

File hashes

Hashes for whovidia-0.1.0.tar.gz
Algorithm Hash digest
SHA256 99ddb743eb551e457d7d91e270c69903d8532ac5cdf21368f71df80243b342b3
MD5 36082f7b6247d5f9efbf039c2024d540
BLAKE2b-256 cbb5f3e1903483b104b9c1798bb66c18564deb573f3bad142ee5ff8da2a63458

See more details on using hashes here.

File details

Details for the file whovidia-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: whovidia-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 7.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.5

File hashes

Hashes for whovidia-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 956f33923df0768257387f31a091c21d503fd3ff71da7eab2f51400dfee7c1b0
MD5 19eee8c08775c495f71513679ec91c95
BLAKE2b-256 d8f8ec9822c3f2b1c600ee633eb2b59c5b1cef543fe970b00d504a35a5f3e9ba

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page