megengine.functional.nn.sliding_window_transpose

sliding_window_transpose(inp, output_size, kernel_size, padding=0, stride=1, dilation=1)

Sum over the sliding windows on the corresponding input location.

Refer to SlidingWindowTranspose for more information.

Parameters:

• inp (Tensor) – input tensor.

• output_size (Union[int, Tuple[int, int]]) – shape of output tensor.

• kernel_size (Union[int, Tuple[int, int]]) – size of the window.

• padding (Union[int, Tuple[int, int]]) – implicit zero padding added on both sides of input. Default: 0

• stride (Union[int, Tuple[int, int]]) – stride of the window. Default: 1

• dilation (Union[int, Tuple[int, int]]) – dilation of the window. Default: 1

Return type:

Tensor