Struct burn_tensor::ops::MaxPool2dBackward
source · pub struct MaxPool2dBackward<B: Backend> {
pub x_grad: B::TensorPrimitive<4>,
}Expand description
Gradient computed during the backward pass for each tensor used by max_pool2d.
Fields§
§x_grad: B::TensorPrimitive<4>Implementations§
source§impl<B: Backend> MaxPool2dBackward<B>
impl<B: Backend> MaxPool2dBackward<B>
sourcepub fn new(x_grad: B::TensorPrimitive<4>) -> Self
pub fn new(x_grad: B::TensorPrimitive<4>) -> Self
Constructs a new MaxPool2dBackward.