Get the latest tech news
Intermediate Activations – the forward hook (2020)
Table of contents
As a researcher actively developing deep learning models, I have come to prefer PyTorch for its ease of usage, stemming primarily from its similarity to Python, especially Numpy. A basic method discussed in PyTorch forums is to reconstruct a new classifier from the original one with the architecture you desire. To give a concrete example for ResNet, let’s say we want the concatenated outputs of model.maxpool, our old friend model.layer3[0].downsample[1], the model.avgpool and the popular layer that feeds into the softmax (i.e., model.fc) for every forward pass of an image.
Or read this on Hacker News