AI Dynamics

Global AI News Aggregator

About

Code-Davinci-2 vs Text-Davinci-3: Instruction Tuning and PPO Performance

– code-davinci-2 > text-davinci-3 means that their instruction finetuning overall hurts performance on academic benchmarks
– text-davinci-3 > text-davinci-2 means that PPO improves performance

→ View original post on X — @_jasonwei