@arylwen.bsky.social
Reposted
I just tried it on deepseek-r1:32b and the full 607b (or whatever), and both stabilize at a consistent 85% confidence
going bigger/smarter doesn't seem to make it any more or less confident after some point. The sweet spot seems to be 7b-8b
going bigger/smarter doesn't seem to make it any more or less confident after some point. The sweet spot seems to be 7b-8b
February 9, 2025 at 6:28 PM
I just tried it on deepseek-r1:32b and the full 607b (or whatever), and both stabilize at a consistent 85% confidence
going bigger/smarter doesn't seem to make it any more or less confident after some point. The sweet spot seems to be 7b-8b
going bigger/smarter doesn't seem to make it any more or less confident after some point. The sweet spot seems to be 7b-8b