How to pass arguments to a custom aggregator using hydra configuration

Physics Simulation Modulus Physics-ML Model Framework Technical Support (Modulus Only)
1

I am using a custom aggregator, and its constructor has arguments other than “params”, “num_losses”, and “weights”. I want to set the values of these arguments in the hydra configuration, but it does not accept any arguments other than the standard three.
How can I set the values of arbitrary arguments by the hydra configuration?

### config.yaml
defaults:
  - loss : custom-aggregator

### loss/custom-aggregator.yaml
_target_ : custom_aggregtor.CustomAggregator
hoge : 10 #<- this setting raises error

### custom_aggregator.py
class CustomAggregator(Aggregator):
    def __init__(self, params, num_losses, weights=None, hoge=1):
        super().__init__(params, num_losses, weights)
        self.hoge = hoge

    def forward(self, losses: Dict[str, torch.Tensor], step: int) -> torch.Tensor:
        loss = ...
        return loss