Function autograd::ops::stop_gradients
[−]
[src]
pub fn stop_gradients(x: &Tensor) -> Tensor
Stops gradients.
Guarantees that the gradient is not propagated to the tensors behind this during gradient computation.