pytorch Causal Conv2d

CausalConv2d.py

from torch.nn.modules.utils import _pair

class CausalConv2d(nn.Conv2d):

    def __init__(self, in_channels, out_channels, kernel_size, stride=1, padding=None, dilation=1, groups=1, bias=True):
        kernel_size = _pair(kernel_size)
        stride = _pair(stride)
        dilation = _pair(dilation)
        if padding is None:
            padding = [int((kernel_size[i] -1) * dilation[i]) for i in range(len(kernel_size))]
        else:
           padding = padding * 2
        self.left_padding = _pair(padding)
        super().__init__(in_channels, out_channels, kernel_size,
                                           stride=stride, padding=0, dilation=dilation,
                                           groups=groups, bias=bias)
    def forward(self, inputs):
        inputs = F.pad(inputs, (self.left_padding[1], 0, self.left_padding[0], 0))
        output = super().forward(inputs)
        return output

Note from a random person: this expects padding to be different from what is traditionally passed to nn.Conv2d. I believe line 11 should include:

else:
   padding = padding * 2

That's the best kind of note! Thanks, and it's updated