torch.nn.functional.hardshrink

torch.nn.functional.hardshrink(input, lambd=0.5) → Tensor

Applies the hard shrinkage function element-wise

See Hardshrink for more details.