pub fn backward<T: Tensor<Dtype = f32, Tape = OwnedTape>>(t: T) -> Gradients
Expand description

Runs backprop algorithm with all operations contained in the tape that t has.

This function takes ownership of t and returns Gradients.

Note that t is required to have OwnedTape, which means it currently owns the crate::gradients::GradientTape.