Cool demo, but even the smallest model does 2.5 tps on my 7900xtx which is not great.
I get 9-10 tps on my MacBook Pro. Though the small model seem really limited
18.8 tps on an 11th gen mobile i7 here.
Hmm I get 44,7 tokens/s on the smallest 1.7b one in the browser on my m1 max.
I get 9-10 tps on my MacBook Pro. Though the small model seem really limited