Comment

Comment by whatshisface

RL is barely even a training method, its more of a dataset generation method.
parent whatshisface 11 months ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton