Hacker News new | ask | show | jobs
by anon373839 889 days ago
With the exception of task specialization. Fine-tuning a small model such as Mistral 7B on a specific set of tasks can outperform using GPT-4 on those tasks, and with cheaper and faster inference.
1 comments

Not on the leaderboards mentioned here. That’s my point, you can overfit for specific tasks, you can’t beat them on multi-task leaderboards without training on the test data.