[jokoon]

Now do the same for image classifiers. I tried a few of them, they're just horribly slow.

This is pretty outrageous considering the first robust image image classifiers appeared around 2007.

[ipsum2]

Doesn't work on image classifiers, because there's no KV cache. Also, standard image classifiers can do 100-1000 images/sec without any optimizations.

[jokoon]

It's not really fast if makes intense usage of a GPU

[ipsum2]

...what?