Professor Uses Codex To Create Lecture Notes On Off-Policy Policy Gradients

Why all this effort? I’m a “deep learner”, of course 🤓

@CSProfKGD Out of curiosity wouldn't it make sense to use HF TRL library https://huggingface.co/docs/trl/en/ppo_trainer. (i)its got an awesome connection with W&B (ii)teaches how you could use it with actual human feedback (iii)you can perhaps use HF spaces with gradio for super easy demos

@CSProfKGD Nothing against codex but I feel the loop to bring in what you want to highlight vs a library well known across community is better. But again you are an awesome teacher so I think you know your stuff well