#192192nd of 242 in ML frameworks and libraries
gin-config
Python config library using decorators and .gin files for dependency injection
- Stars
- 2.2k
- License
- Apache-2.0
- Last commit
- Oct 2026
- Language
- Python
2,145 on 25 Apr 20262,160 on 14 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Gin lets you decorate Python functions or classes with `@gin.configurable` and override their default parameters from `.gin` files or the command line. It supports references to other configurables, scopes, macros, file includes and logging of the operative config. It targets ML experiments with many nested parameters, with optional `gin.tf` and `gin.torch` modules.
Who it is for: Python ML engineers managing many experiment parameters via config files
Strengths
- Needs only a decorator and a parse call; no config objects or protos to maintain
- Scopes let one function be configured differently, e.g. generator and discriminator networks
- Logs the operative config, with TensorBoard integration
- Installs with pip; core import works without TensorFlow
Weaknesses
- README says it is in alpha; corner-case behavior may change incompatibly
- Evaluated references are not cached, which can surprise users, especially with macros
- Nested scopes have ongoing debate over ordering and behavior
- Heavily hierarchical configs reduce readability; no tagged releases
What it needs
- no GPU
- Models: TensorFlow (optional gin.tf), PyTorch (optional gin.torch)
Also in ML frameworks and libraries
See all 242| Rank | Project | Score |
|---|---|---|
| 1 | transformersModel-definition library for pretrained text, vision, audio and multimodal models | 94 out of 100 |
| 2 | ultralyticsPython library and CLI for training and running YOLO vision models | 94 out of 100 |
| 3 | pytorch-image-modelsPyTorch library of image models, pretrained weights and training scripts | 94 out of 100 |
| #4 | pytorchPython tensor library with GPU support and tape-based autograd | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 92 out of 100 |