models/ebc/utils.py

from torch import nn
from typing import  Union, List, List


vgg_urls = {
    "vgg11": "https://download.pytorch.org/models/vgg11-8a719046.pth",
    "vgg11_bn": "https://download.pytorch.org/models/vgg11_bn-6002323d.pth",
    "vgg13": "https://download.pytorch.org/models/vgg13-19584684.pth",
    "vgg13_bn": "https://download.pytorch.org/models/vgg13_bn-abd245e5.pth",
    "vgg16": "https://download.pytorch.org/models/vgg16-397923af.pth",
    "vgg16_bn": "https://download.pytorch.org/models/vgg16_bn-6c64b313.pth",
    "vgg19": "https://download.pytorch.org/models/vgg19-dcbb9e9d.pth",
    "vgg19_bn": "https://download.pytorch.org/models/vgg19_bn-c79401a0.pth",
}


vgg_cfgs = {
    "A": [64, "M", 128, "M", 256, 256, "M", 512, 512, "M", 512, 512],
    "B": [64, 64, "M", 128, 128, "M", 256, 256, "M", 512, 512, "M", 512, 512],
    "D": [64, 64, "M", 128, 128, "M", 256, 256, 256, "M", 512, 512, 512, "M", 512, 512, 512],
    "E": [64, 64, "M", 128, 128, "M", 256, 256, 256, 256, "M", 512, 512, 512, 512, "M", 512, 512, 512, 512]
}


def make_vgg_layers(cfg: List[Union[str, int]], in_channels: int = 3, batch_norm: bool = False, dilation: int = 1) -> nn.Sequential:
    layers = []
    for v in cfg:
        if v == "M":
            layers += [nn.MaxPool2d(kernel_size=2, stride=2)]
        else:
            conv2d = nn.Conv2d(in_channels, v, kernel_size=3, padding=dilation, dilation=dilation)
            if batch_norm:
                layers += [conv2d, nn.BatchNorm2d(v), nn.ReLU(inplace=True)]
            else:
                layers += [conv2d, nn.ReLU(inplace=True)]
            in_channels = v
    return nn.Sequential(*layers)