The gap between the promising theory and the janky practice is wiiiiide.

My hope is that the next time you read an article about the latest natural language processing conquest, you'll be able to nod contentedly, having pretty darn good mental model of what's going on under the hood. https://t.co/QO6n3FsUVG

Timeline cleanser. The abstraction for reinforcement learning is beautiful to me. from mind import Mind from world import World mind = Mind() world = World() actions = zeros(n_actions) while True: sensors, reward = world.step(actions) actions = mind.step(sensors, reward)

Now with a table of contents. (Thanks for the suggestion @sri_ramanathan!) https://t.co/VaFt8md9hJ

Not pictured here is the four weeks I spent reading the original paper and a handful of exceptional blog posts and trying to get all the pieces to fit together in my head.

Thanks to all of you who followed along during the birthing process. https://t.co/K3r2ssNR2K

If you've always wondered how Transformers work but know nothing about machine learning, I wrote this peek behind the curtain for you. (Actually I wrote it for me, but you might find it useful.) Here's the beta release. https://t.co/VaFt8md9hJ

An approachable guide to building ML models for research. Covers the five stages of of the machine learning process which, I was surprised to discover, are not denial, anger, bargaining, depression, acceptance. https://t.co/XImgM90f5Z h/t @hardmaru, @vboykis