The AI Epiphany Video Watch posted an update
LLaMA 3 Deep Dive! (Thomas Scialom – Meta)
Why it mattersThe AI Epiphany's description introduces Thomas Scialom discussing Llama 3, including synthetic training data, privacy, architecture choices, distillation and context length. The chapter list makes the technical scope explicit. This 2024 interview is presented as an archive deep dive, and its account of model design is attributed to the guest and channel.
Discuss: Which Llama training trade-off would you revisit first when choosing between synthetic data, distillation and a different architecture?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.