Llama 2 lead and Llama 3 post-training lead Thomas Scialom of Meta/FAIR, on the Chinchilla trap, why Synthetic Data and RLHF works, and how Llama4's focus on Agents will lead us to Open Source AGI.
Holy shit. This section on SFT vs RLHF, and scaling RLHF. We are just getting started.
This is incredibly useful thank you so much!
Holy shit. This section on SFT vs RLHF, and scaling RLHF. We are just getting started.
This is incredibly useful thank you so much!