3 ms·
It's not too surprising, honestly! I've poked around with similar in the past and am of a perspective that ternary is a very good thing for a lot of neural netw
by tbalsam 3y ago
It's not too surprising, honestly! I've poked around with similar in the past and am of a perspective that ternary is a very good thing for a lot of neural networks.
Training CIFAR-10 speedily w/ ternary weights on an fp16 interface (using fp16 buffers, and norm params unchanged): https://gist.github.com/tysam-code/a43c0fab332e50163b74141bccdf4588 https://gist.github.com/tysam-code/a43c0fab332e50163b74141bc...