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.

An example of a layer config is the following:

diffusion128|gcn128|concurent:
  name: ProteinEncoder
  instanciate:
    _target_: atomsurf.networks.ProteinEncoderBlock
  kwargs:
    surface_encoder:
      name: DiffusionNetBlockBatch
      instanciate:
        _target_: atomsurf.network_utils.DiffusionNetBlockBatch  # diffusion_net.DiffusionNet
      kwargs:
        C_width: 128
        mlp_hidden_dims: [128, 128]
        dropout: 0.0
        use_bn: true
        init_time: 2.0  # either null (for constant init) or a float
        init_std: 2.0
    graph_encoder:
      name: GCNx2Block
      instanciate:
        _target_: atomsurf.network_utils.GCNx2Block
      kwargs:
        dim_in: 128
        hidden_dims: 128
        dim_out: 128
        dropout: 0.0
        use_bn: true
        use_weighted_edge_distance: false
    communication_block:
      name: ConcurrentCommunication
[....]

Is there a way to make the layer names customizable in the configuration? For example, changing diffusion128|gcn128|concurrent to diffusion{$width1}|gcn{$width2}|concurrent with width1 and width2 as parameters that can be defined when creating an architecture.

Thank you for your help!