In this article, we dive into the internal workings of the Accelerate library from HuggingFace, to answer "could Accelerate really be this easy?"

As part of this blog post, we look at the MLP Mixer architecture in detail and also understand why it is not considered conv free.

As part of this post, we look at the Convolution and Linear layers in MS Excel and compare results from Excel with PyTorch implementations.

As part of this blog post, we will look into the ConViT transformer architecture in detail and learn all about it and also the gated positional self-attention (GPSA) layer! We also see how the ConViT architecture gets the best of both worlds and obtains the benefits of both Transformers and CNNs.

As part of this blog post, I introduce conda, pip, docker & conda channels. I showcase an example of how recently I got stuck in a "messy" environment and some possible solutions. Finally, I introduce "fastchan" and show how it can make all of our lives easier.

BatchNorm is a key component in almost all SOTA image classifiers, but it also introduces practical challenges. Is there a way to get similar performance for image classifiers without BatchNorm? Looks like the answer is "yes"!

Do training methods matter more than model architectural changes? In this post we take a dive into the ResNet-RS paper and try to answer this question.

As part of this blog, we are going to be looking into the EfficientNetV2 architecture in detail with code implementation in PyTorch.