torch.nn.functional.max_pool2d¶
- torch.nn.functional.max_pool2d(input, kernel_size, stride=None, padding=0, dilation=1, ceil_mode=False, return_indices=False)[source]¶
Applies a 2D max pooling over an input signal composed of several input planes.
Note
The order of
ceil_mode
andreturn_indices
is different from what seen inMaxPool2d
, and will change in a future release.See
MaxPool2d
for details.- Parameters
input – input tensor , minibatch dim optional.
kernel_size – size of the pooling region. Can be a single number or a tuple (kH, kW)
stride – stride of the pooling operation. Can be a single number or a tuple (sH, sW). Default:
kernel_size
padding – Implicit negative infinity padding to be added on both sides, must be >= 0 and <= kernel_size / 2.
dilation – The stride between elements within a sliding window, must be > 0.
ceil_mode – If
True
, will use ceil instead of floor to compute the output shape. This ensures that every element in the input tensor is covered by a sliding window.return_indices – If
True
, will return the argmax along with the max values. Useful fortorch.nn.functional.max_unpool2d
later