Articles

• ---

  Week 26 Digest - Poet Engineer Edition

• ---

  Week 25 Digest - Pixel Art Edition

• ---

  Zero shot learning with CLIP

  A fun property of machine learning is that this reasoning works in reverse too: If meaningful generalities can help you represent your data with fewer numbers, finding a way to represent your data in fewer numbers can often help you find meaningful generalities. Compression is akin to understanding and all that.

• ---

  Transfer Learning with ConvNext, HuggingFace, and Ignite

  The outcome of this exploration is a family of pure ConvNet models dubbed ConvNeXt. Constructed entirely from standard ConvNet modules, ConvNeXts compete favorably with Transformers in terms of accuracy and scalability, achieving 87.8% ImageNet top-1 accuracy and outperforming Swin Transformers on COCO detection and ADE20K segmentation, while maintaining the simplicity and efficiency of standard ConvNets.

  TLDR: I put together a quick-n-dirty example (Github gist here) if you don’t feel like reading.

• ---

  A 2022 Guide to Installing Emacs on Windows

• ---

  Getting Started