xlnstorch.csrc.conv3d_backward

xlnstorch.csrc.conv3d_backward#

xlnstorch.csrc.conv3d_backward(grad, x, weight, base, bias_defined, stride_d, stride_h, stride_w, padding_d, padding_h, padding_w, dilation_d, dilation_h, dilation_w, groups)#

Perform the backward pass for LNS 3D convolution.

Parameters:

  • grad (torch.Tensor) – The gradient tensor in LNS format (torch.int64).

  • x (torch.Tensor) – The input tensor in LNS format (torch.int64).

  • weight (torch.Tensor) – The convolution kernel in LNS format (torch.int64).

  • base (torch.Tensor) – The base for the LNS representation.

  • bias_defined (bool) – Whether the bias tensor is defined.

  • stride_d (int) – The depth stride of the convolution.

  • stride_h (int) – The height stride of the convolution.

  • stride_w (int) – The width stride of the convolution.

  • padding_d (int) – The depth padding applied to the input tensor.

  • padding_h (int) – The height padding applied to the input tensor.

  • padding_w (int) – The width padding applied to the input tensor.

  • dilation_d (int) – The depth dilation factor for the convolution.

  • dilation_h (int) – The height dilation factor for the convolution.

  • dilation_w (int) – The width dilation factor for the convolution.

  • groups (int) – The number of groups for the convolution.

Return type:

Tensor

Contents