diff options
| author | Chandler Zuo <chandlerzuo@fb.com> | 2019-01-16 14:01:39 -0800 |
|---|---|---|
| committer | Facebook Github Bot <facebook-github-bot@users.noreply.github.com> | 2019-01-16 14:16:30 -0800 |
| commit | 237c0c3c7a551c6b27c5108ce30769bd41a542be (patch) | |
| tree | bf6abe0adf2f11d4f567aef9263e2cfd6b40e8a8 /tools/autograd/derivatives.yaml | |
| parent | aff0964ee78215cf11bbd46883159c1f8694a3de (diff) | |
| download | pytorch-237c0c3c7a551c6b27c5108ce30769bd41a542be.tar.gz pytorch-237c0c3c7a551c6b27c5108ce30769bd41a542be.tar.bz2 pytorch-237c0c3c7a551c6b27c5108ce30769bd41a542be.zip | |
Port the backend of FractionalMaxPool3d from TH to ATen (#15575)
Summary:
1. Port the FractionalMaxPool3d implementation from THNN/THCUNN to ATen.
2. Expose this function to Python module nn.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/15575
Differential Revision: D13612848
Pulled By: chandlerzuo
fbshipit-source-id: 5f474b39005efa7788e984e8a805456dcdc43f6c
Diffstat (limited to 'tools/autograd/derivatives.yaml')
| -rw-r--r-- | tools/autograd/derivatives.yaml | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/tools/autograd/derivatives.yaml b/tools/autograd/derivatives.yaml index 9ec8f9efea..5eef168ee2 100644 --- a/tools/autograd/derivatives.yaml +++ b/tools/autograd/derivatives.yaml @@ -1069,6 +1069,9 @@ - name: fractional_max_pool2d(Tensor self, IntList kernel_size, IntList output_size, Tensor random_samples) self: fractional_max_pool2d_backward(grad, self, kernel_size, output_size, indices) +- name: fractional_max_pool3d(Tensor self, IntList kernel_size, IntList output_size, Tensor random_samples) + self: fractional_max_pool3d_backward(grad, self, kernel_size, output_size, indices) + - name: max_pool2d_with_indices(Tensor self, IntList kernel_size, IntList stride, IntList padding, IntList dilation, bool ceil_mode) self: max_pool2d_with_indices_backward(grad, self, kernel_size, stride, padding, dilation, ceil_mode, indices) @@ -1164,6 +1167,10 @@ grad_output: max_pool_double_backward(grad, indices, 2) self: zeros_like(self) +- name: fractional_max_pool3d_backward(Tensor grad_output, Tensor self, IntList kernel_size, IntList output_size, Tensor indices) + grad_output: max_pool_double_backward(grad, indices, 3) + self: zeros_like(self) + - name: glu_backward(Tensor grad_output, Tensor self, int64_t dim) grad_output: glu_double_backward_grad_output(grad, self, dim) self: glu_double_backward(grad, grad_output, self, dim) |