The easiest way would be to plot them before normalizing.
However, if that’s not possible, you could also undo the normalization:
x = torch.empty(3, 224, 224).uniform_(0, 1)
mean = (0.5, 0.5, 0.5)
std = (0.5, 0.5, 0.5)
norm = transforms.Normalize(mean, std)
x_norm = norm(x)
x_restore = x_norm * torch.tensor(std).view(3, 1, 1) + torch.tensor(mean).view(3, 1, 1)
print((x_restore - x).abs().max())
> tensor(0.)