Top level commands

The main top-level command train and eval/test are described on the Train, Eval, Test page. Further top-level commands are detailed below.

Adapt Command

This command adapts an already trained model using KL divergence regularization. It is advised that all other adaptations should be carried out using [MEL](Model Editing Language (MEL)). The adapt command is very similar to the train command except that it carries two more parameters:

• originalModelFileName: the file name of the model that will be adapted.

• refNodeName: the name of the node in the computational network which will be used for KL divergence regularization.