Runs and Converters

The Run Object

DeepCAVE utilizes Run objects as a fundamental unit for data interpretation. A run represents a hyperparameter optimization process, encompassing a collection of trials, each corresponding to a specific hyperparameter configuration with its associated objective value, budget, and seed.

Converters

Converters are used to access the optimizer data stored on the file system and transform it into run objects.

Currently, DeepCAVE offers the following converters:

  • SMAC (v1.4)

  • SMAC (v2.0.0)

  • AMLTK

  • Optuna

  • BOHB

  • DeepCAVE (native)

  • Pandas DataFrame

In the logs directory, you can find example runs for each of the converters.

Note

DeepCAVE observes optimizer data on the file system and therefore allows for monitoring of both finished processes and running processes that regularly write new results to disk.

Adding a Converter

If you would like to add your own converter, please have a look at our tutorial on how to add a converter and use any of our provided converters as a starting point.