MConvNormReLU

class torchelie.nn.MConvNormReLU(in_ch: int, out_ch: int, ks: int, norm, center: bool = True)

Experimental: A packed block with Masked Conv-Norm-ReLU

Parameters

• in_ch (int) – input channels

• out_ch (int) – output channels

• ks (int) – kernel size

• norm (ctor) – A normalization layer constructor in the form (num_layers) -> norm layer or None

• center (bool) – whether the masked conv has access to the central pixel or not

Returns

A packed block with MaskedConv-Norm-ReLU as a CondSeq

Warning

MConvNormReLU() is experimental, and may change or be deleted soon if not already broken