Implementing Custom Modules in Pytorch
Tutorial on custom designing a new module for your network in pytorch
Image credit: Li et. al.Overview
There are many posts and codes on how to build your own deep neural network using built-in functions and modules such as torch.nn.Conv2d or F.relu etc.but what if you want to design a completely new module for your network. In this blog post, we will learn about things that you need to keep in mind while implementing your own module and some common errors that might drive you up the wall.
Vidit Goel
Machine Learning Engineer
My research interests include generative models and segmentation.