Intro to Transformers: From Attention to Production

About this event

A 3-hour deep dive. We start with self-attention on a whiteboard, code a 4M parameter model live, train it on Tiny Shakespeare, then ship it. You leave with a working repo and an intuition that sticks.

Schedule

• 17:30Self-attention from first principles
• 18:30Live code: minimal GPT
• 19:30Training & sampling
• 20:00Deploy + Q&A

Speakers / Mentors