_target_: deepscreen.models.loss.multitask_loss.MultitaskLoss

loss_fns:
  - _target_: torch.nn.MSELoss
  - _target_: torch.nn.CrossEntropyLoss
    weight: null
reduction: sum