B blog.research.google Outperforming larger language models with less training data and smaller models 320 points atg_abhishek 3 years ago 123 comments
