In this session, I will talk about using Lance file format to manage deep learning artefacts. More specifically, saving, loading and versioning model weights. I will also be demonstrating a Proof-of-Concept on this topic.