from pydantic.dataclasses import dataclass
from ..base import BaseSamplerConfig
[docs]@dataclass
class PixelCNNSamplerConfig(BaseSamplerConfig):
"""This is the PixelCNN sampler configuration instance.
Parameters:
input_dim (tuple): The input data dimension. Default: None.
n_layers (int): The number of convolutional layers in the model. Default: 10.
kernel_size (int): The kernel size in the convolutional layers. It must be odd. Default: 5
"""
n_layers: int = 10
kernel_size: int = 5
def __post_init__(self):
super().__post_init__()
assert (
self.kernel_size % 2 == 1
), f"Wrong kernel size provided. The kernel size must be odd. Got {self.kernel_size}."