Language for Test Specification of Machine Learning Models

Domain specific language (DSL) come in handy to fluently express complex business and working assumptions, in a language that reads like English. In industry and academia, usually developers and testers spent a good amount of time in the development and testing of ML models. And sometimes they are required to repeat the same set of operations even with a slight change in the requirement from client end. To tackle this This problem we developed a language for test generation of ML models which would lead to reduced turnaround time related to the development and testing of ML models.

DLTL workflow