Spaces:
Sleeping
Sleeping
import torch | |
import torch.nn as nn | |
import torch.functional as F | |
from bn import batch_norm | |
from residual import residual | |
class decoder(nn.Module): | |
def __init__(self, inp, out): | |
super().__init__() | |
self.upsample = nn.Upsample(scale_factor=2, mode = 'bilinear', align_corners = True) | |
self.block = residual(inp+out, out) | |
def forward(self, x, skip): | |
x = self.upsample(x) | |
x = torch.cat([x, skip], axis = 1) | |
x = self.block(x) | |
return x |