4 ms·And this is even worse on locally ran models. I get 2k tokens of context, max.by highduc 3y agoAnd this is even worse on locally ran models. I get 2k tokens of context, max.