Depth study notes deep learning basic parameters --Epoch, Iteration, Batchsize

Deep learning basic --Epoch, Iteration, Batchsize

The power of the GPU batch 2 may play a better performance, and therefore is often provided to the whole than 10 ... 64, 128, the better the performance of the whole multiple of 100
https: // zhuanlan .zhihu.com / p / 27763696

Guess you like

Origin www.cnblogs.com/cekong/p/11233432.html