In hydra experiment when I want to sweep over two parameters, why does

I’m using Hydra to create some configuration of layers in my neural network, these layers will then be used in other configs to create neural network architectures. However, I find myself creating multiple repeated configs, and the only difference between them for example is the width.

I’m using Hydra to create some configuration of layers in my neural network, these layers will then be used in other configs to create neural network architectures. However, I find myself creating multiple repeated configs, and the only difference between them for example is the width.

I want to instantiate my own custom class. My directory structure looks like this:

I have the following use case. I am working on a large project with multiple subtasks. For example, we are performing classification on some datasets, segmentation on others, and translation on yet another set of datasets (see below).

My config file has a controller field that I manually set using CLI options: +controller=mpc. The controller is defined inside the controller group. The filename is mpc.yaml. The content is this: