# About

Machine learning code can be hard to manage. Research often starts with hacking a model prototype in an improvised setup that grows in complexity over time. Configuration options and hyperparameters diffuse through the code base and continued changes can feel increasingly brittle. Often, technical debt takes its toll when modifications take an unreasonably long time due to poor structure and lack of modularity.

machinable provides an alternative approach. Using straight-forward conventions and a powerful configuration engine, it can help structure your projects in a principled way so you can move quickly while enabling reuse and collaboration.

# Features

Explore key features at a glance →

Powerful configuration

  • YAML-based project-wide configuration files with expressive syntax
  • Efficient configuration manipulation
  • Modular code organisation to allow for encapsulation and re-use
  • Import system to use 3rd party configuration and code without overhead
  • 'Mixins' for horizontal inheritance structure

Efficient execution

  • Works with existing code
  • Support for seamless cloud execution
  • Automatic code backups
  • Managed randomness and reproducibility
  • Advanced hyperparameter tuning using Ray Tune

Effective result collection and analysis

  • Logging, tabular record writer and storage API
  • File system abstraction (in-memory, AWS S3, and more)
  • Flat-file result database with SQL-like query interface
  • Convenient configuration and result retrieval