For managing configs of ML experiments (where each experiment can override a base config, and "variant" configs can further override the experiment config, etc)

For managing configs of ML experiments (where each experiment can override a base config, and "variant" configs can further override the experiment config, etc), Hydra + Yaml + OmegaConf is really nice. https://hydra.cc/ https://hydra.cc/ I admit I don't fully understand all the advanced options in Hydra, but the basic usage is already very useful. A nice guide is here: https://florianwilhelm.info/2022/01/configuration_via_yaml_and_cli_with_hydra/ https://florianwilhelm.info/2022/01/configuration_via_yaml_a...