google/gin-config

Health v2 · maintenance only

Adoption not part of the score

2155 stars · 120 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Gin-config is a lightweight Python configuration framework based on dependency injection. Functions or classes decorated with @gin.configurable can have default parameter values supplied from .gin config files or the command line, which is especially useful for machine learning experiments.

Use cases

• configure hyperparameters for ML experiments from a file
• remove boilerplate parameter plumbing in Python projects
• swap model architectures via config bindings
• manage nested configuration for TensorFlow or PyTorch training
• override function defaults from the command line

When to choose

• your ML codebase has many deeply nested parameters
• you want config-as-code without defining proto or config objects
• you use TensorFlow or PyTorch and want framework-specific configurability

When to avoid

• you need runtime-reloadable or hierarchical YAML/JSON config for services
• your project is not Python
• you prefer explicit dependency injection containers

Facets

library · maturity stable

configuration-management machine-learning developer-tools deep-learning python cross-platform dependency-injection config-files tensorflow pytorch experiment-configuration

Member repositories

For agents

markdown · JSON · MCP: product_card(name="google/gin-config")