peal.architectures.interfaces.ResnetConfig

class peal.architectures.interfaces.ResnetConfig

Bases: BaseModel

The config template for a ResNet layer.

Methods

Attributes

config_name	The type of config.
num_blocks	Number of blocks per layer.
num_neurons	The number of neurons in the layer.
tensor_dim	The dimension of the tensor.
use_batchnorm	Whether to use batchnorm or not.

config_name: str = 'ResnetConfig'

The type of config. This is necessary to find config class from yaml config

num_neurons: pydantic.PositiveInt = 512

The number of neurons in the layer.

num_blocks: pydantic.PositiveInt = 2

Number of blocks per layer.

use_batchnorm: bool = True

Whether to use batchnorm or not.

tensor_dim: pydantic.PositiveInt = 2

The dimension of the tensor. Options: [1, 2, 3]