2 ms·
They can always finetune using RL later. Superversied training was the first step at making AlphaGo work.
by dontreact 9y ago
They can always finetune using RL later. Superversied training was the first step at making AlphaGo work.
loading thread…