The reference implementation
Join Bluegum's new Applied AI team. Before anyone trains a big model, the team builds a small reference implementation from scratch: tensors, autograd, a network with a checked backward pass, and a training loop that learns. Includes 4 hands-on Python coding tasks (NumPy), checked by automated tests.
Season 2: The Applied AI team
Bluegum has set up a deep-learning team, and you've been invited to join it. Over four milestones you will build a reference implementation from scratch, rescue a training run that won't converge, build the core of a form reader and a ward early-warning model, and ship a compressed transformer safely. Every coding task is real NumPy, checked by tests.
You can play this season without finishing season 1. Your skill scores carry over if you did.
Bluegum Health, Riverside Clinics and every person here are fictional. The decisions, tools and practices are modelled on how real ML teams work.