Optimizing is a great way to map a new system's territory! Current trying to make a non-interactive machine learning systems interactive: a treasure hunt in a strange land!
2.5s -> 250ms: reimplement on GPU
250ms -> 50ms: fewer stages
50ms -> 25ms: fewer textures
25ms -> 2.5ms: ?
Conversation
Doing this has also made me take the model apart and put it back together in different ways a couple times, which has helped me understand how it works much better than if I'd implemented it very literally from the paper. Active learning is powerful!
2
25
I'm generally interested in the way that interactions transform when they become continuous and immediate, rather than discrete and delayed. There's a lot of unexplored opportunity there in machine learning—this pen's ink comes from both past and future; it draws with you.
1
8
Show replies
Making them behave and respond through this sort of mechanism would be a very interesting project…
My mind is just exploding with potentials and ideas watching this 10 second without context!