Key terms

Dotscience takes data science from experiments and research to full enterprise production. As such, there may be terminology that is not familiar to all users. This section will define some of those terms to help you get more value from using Dotscience.


A real or virtual machine used to run models that utilize Dotscience. Runners can have GPUs in them, in which case Dotscience will enable usage of the GPUs from inside JupyterLab or ds run scripts.


Or hyperparameter. A configurable input to a model, recorded in Dotscience using, for example, the ds() functions of the Dotscience Python library.

Summary statistic

A quantitative value recording the success of a model, also recordable via ds().

(More to follow.)