This is a follow-up to my previous post of Depthwise Separable Convolutions in PyTorch. This article is based on the nice CVPR paper titled “Rethinking Depthwise Separable Convolutions: How Intra-Kernel Correlations Lead to Improved MobileNets” by Haase and Amthor. Previously I took a look at depthwise separable convolutions which are a drop-in replacement for standard convolutions, but focused on computational and parameter-based efficiency. Basically, you can gain similar results with a lot less parameters and FLOPs, so they are used in MobileNet style architectures.

Creating pleasant plots with seaborn Seaborn is an awesome Python library to create great-looking data plots. It’s a bit higher level than the often used matplotlib and this blog entry serves as a self-reminder about the most frequently used plots for myself. It’s way to specify in a declarative way what you want to plot rather than plot details like markers, colors etc is refreshing and frees some cognitive space which you can use for other tasks.

Today’s paper: Emerging properties in self-supervised vision transformers by Mathilde Caron et al. Let’s get the dinosaur out of the room: the name DINO refers to self-distillation with no labels. The self-distillation part refers to self-supervised learning in a student-teacher setup as is often seen for distillation. However, the catch is that in contrast to normal distillation setups where a previously trained teacher network is training a student network, here they work without labels and without pre-training the teacher.

Today’s paper: Rethinking ‘Batch’ in BatchNorm by Wu & Johnson BatchNorm is a critical building block in modern convolutional neural networks. Its unique property of operating on “batches” instead of individual samples introduces significantly different behaviors from most other operations in deep learning. As a result, it leads to many hidden caveats that can negatively impact model’s performance in subtle ways. This is a citation from the paper’s abstract and the emphasis is mine which caught my attention.

To compare the previously described approach of hyperparameter tuning using fastai and wandb, today we’ll see how to tackle the same approach, but using PyTorch Lightning instead of fastai. The goal is to have an automated hyperparameter tuning pipeline running on the Numerai data set. What is Numerai? Numerai is a hedge fund which trades stocks in a market neutral fashion. That means that they try to make money without having a lot of risk for their customers.