3 ms·
This is ridiculous. 32B and beating deepseek and o1. And yet I'm trying it out and, yeah, it seems pretty intelligent... Remember when models this size could j
by Alex-Programs 2y ago
This is ridiculous. 32B and beating deepseek and o1. And yet I'm trying it out and, yeah, it seems pretty intelligent...
Remember when models this size could just about maintain a conversation?
- moffkalast 2y agoI still remember Vicuna-33B, that one stayed on the leaderboards for quite a while. Today it looks like a Model T, with 1B models being more coherent.
- dcreater 2y agoHave you tried it as yet? Don't fall for benchmark scores.