Aung Ko·6d agoI Compared a Small Trained Vision Model with a Vision LLM. Here’s What I Learned.Classifying and grouping property photos using DINOv2, lightweight classifiers, and geometric matching — and comparing the results with a…A response icon1A response icon1
Aung Ko·May 24From the Netflix Prize to a TikTok-Style Job FeedBuilding a recommender with Funk’s SGD, an LLM cold-start layer, and a swipe UI
