CoreNet is a deep neural network toolkit allowing researchers to train various models for tasks like object classification and semantic segmentation. Noteworthy updates in version 0.1.0 include OpenELM and CatLIP examples. Apple’s research using CoreNet led to publications like “FastVit.” Installation tips for Git LFS and PyTorch are provided. The directory structure organizes tutorials, training recipes, MLX examples, model implementations, and datasets. Maintained by Sachin, Maxwell Horton, Mohammad Sekhavat, and Yanzi Jin, CoreNet welcomes contributions. Originating from CVNets, CoreNet now supports a wider range of applications, such as foundation models like LLMs. Don’t forget to cite their work if you find it helpful!
https://github.com/apple/corenet