This is a library for simplifying parameter search for machine learning. The library is designed to be as flexible as possible. It is designed to be a set of tools rather than a framework.
There are 3 key goals for this module
- Be Flexible: make as few assumptions as possible about the program to be optimized.
- Be Modular: this is not a framework, but it can be used as one. The individual components (deployment, logging, analysis) should work on their own.
- Show Data: we don't hide the analysis from the user. We show the effect that the parameters have across all runs and perform significance tests.
This information will be added soon. For now, if you find this project and are interested in trying it just post an issue on the repo and I will upload some instructions.