[量化]Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constraine

Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1

已复现：https://github.com/onion-rain/MCTS:

train

python trainer_exp.py --gpu 1 --arch resnet14_binary --dataset cifar10 --valuate --binary --visdom

BNN

两种二值化方法：

deterministic，确定法，作则采用，大于0取1，否则-1

stochastic