clinicadl.networks.config.ConvEncoderOptions¶

clinicadl.networks.config.ConvEncoderOptions[source]¶

Config class for conv_args of CNNConfig, AutoEncoderConfig, and VAEConfig.

parameter channels: Sequence[PositiveInt] [Required]¶

parameter kernel_size: Union[int, tuple[int, ...], list[Union[int, tuple[int, ...]]]] = 3¶

parameter stride: Union[int, tuple[int, ...], list[Union[int, tuple[int, ...]]]] = 1¶

parameter padding: Union[int, tuple[int, ...], list[Union[int, tuple[int, ...]]]] = 0¶

parameter dilation: Union[int, tuple[int, ...], list[Union[int, tuple[int, ...]]]] = 1¶

parameter pooling: Union[tuple[PoolingLayer, dict[str, Any]], list[tuple[PoolingLayer, dict[str, Any]]], None] = (PoolingLayer.MAX, {'kernel_size': 2})¶

parameter pooling_indices: Sequence[int] = None¶

parameter act: Union[ActFunction, tuple[ActFunction, dict[str, Any]], None] = ActFunction.PRELU¶

parameter output_act: Union[ActFunction, tuple[ActFunction, dict[str, Any]], None] = None¶

parameter norm: Union[ConvNormLayer, tuple[ConvNormLayer, dict[str, Any]], None] = ConvNormLayer.INSTANCE¶

parameter dropout: Optional[NonNegativeFloat] = None¶

parameter bias: bool = True¶

parameter adn_ordering: str = 'NDA'¶