Hey
Is there a PyTorch equivalent to theano’s hard sigmoid?
http://deeplearning.net/software/theano/library/tensor/nnet/nnet.html
Thanks
Hey
Is there a PyTorch equivalent to theano’s hard sigmoid?
http://deeplearning.net/software/theano/library/tensor/nnet/nnet.html
Thanks
nn.Hardtanh
has optional parameters for min and max linear values. depending on the details of your problem this might be sufficient.