AI PlatformsFine-tuning is not a shortcutFine-tuning can help, but it works best after the task, data, evals, and product behavior are already clear.May 19 2026//9 min read#fine-tuning#training#datasetsOpen post→
AI SystemsSmaller models belong in the architectureA practical note on using smaller models for focused jobs instead of sending every request to the largest model.December 30 2025//10 min read#distillation#small models#trainingOpen post→
Deep Learning BasicsBackpropagation is bookkeeping with consequencesA slower explanation of gradients, loss, and weight updates for people who want the mechanism without the fog machine.April 12 2024//6 min read#backpropagation#gradient descent#trainingOpen post→