stochdepth 0.1.0

A simple hook based implementation of "Deep Networks with Stochastic Depth" for torchvision resnets.

Stochastic Depth with PyTorch Hooks

A simple hook based implementation of Deep Networks with Stochastic Depth for torchvision resnets.

Example

import torch
import torchvision.models as models
resnet = models.resnet152(pretrained=False)

from sdepth import set_hooks
set_hooks(resnet, p=0.2)

x = torch.zeros((8, 3, 224, 224), dtype=torch.float32, device=device)

resnet.train()
y = resnet(x)