Implementation of Bi-Real-Net Model in Pytorch https://arxiv.org/abs/1808.00278
Below you can find pretrained weights on imagenet for full-precision bireal18 and bireal34 network.
Both networks are trained by setting initial lr=1.0e-1
, weight_decay=1e-4
and momentum=0.9
. The learning rate is decayed by 0.1 using a step-lr schedule every 30 epochs for a total of 120 epochs.
BNN+: Improved Binary Network Training: https://arxiv.org/abs/1812.11800