Skip to content

Conversation

@zotanika
Copy link
Contributor

  • Handling group > 1 cases, assuming group == output channels
  • Simply decomposed into Relay split, conv2d_transposed, and multi-leveled concatenate ops
  • Added some test cases

Signed-off-by: zotanika [email protected]

Thanks for contributing to TVM! Please refer to guideline https://tvm.apache.org/docs/contribute/ for useful information and tips. After the pull request is submitted, please request code reviews from Reviewers by @ them in the pull request thread.

- Handling group > 1 cases, assuming group == output channels
- Simply decomposed into Relay split, conv2d_transposed, and multi-leveled concatenate ops
- Added some test cases

Signed-off-by: zotanika <[email protected]>
@mshr-h
Copy link
Contributor

mshr-h commented Jul 19, 2021

I want to try your PR.

Is there any pretrained models which contains Deconv with group > 1?

@masahi
Copy link
Member

masahi commented Jan 9, 2022

@zotanika please resolve the conflict.

@masahi masahi merged commit be0677d into apache:main Jan 17, 2022
yuanfz98 pushed a commit to yuanfz98/tvm that referenced this pull request Jan 24, 2022
* [Caffe Frontend] supporting group > 1 cases for Deconv op

- Handling group > 1 cases, assuming group == output channels
- Simply decomposed into Relay split, conv2d_transposed, and multi-leveled concatenate ops
- Added some test cases

Signed-off-by: zotanika <[email protected]>

* [Caffe Frontend] amending a test case for Deconv op

Signed-off-by: zotanika <[email protected]>

* explicit importing tvm.testing

* changing split axis to 0, according to PR apache#9336
crazydemo pushed a commit to crazydemo/tvm that referenced this pull request Jan 27, 2022
* [Caffe Frontend] supporting group > 1 cases for Deconv op

- Handling group > 1 cases, assuming group == output channels
- Simply decomposed into Relay split, conv2d_transposed, and multi-leveled concatenate ops
- Added some test cases

Signed-off-by: zotanika <[email protected]>

* [Caffe Frontend] amending a test case for Deconv op

Signed-off-by: zotanika <[email protected]>

* explicit importing tvm.testing

* changing split axis to 0, according to PR apache#9336
ylc pushed a commit to ylc/tvm that referenced this pull request Feb 16, 2022
* [Caffe Frontend] supporting group > 1 cases for Deconv op

- Handling group > 1 cases, assuming group == output channels
- Simply decomposed into Relay split, conv2d_transposed, and multi-leveled concatenate ops
- Added some test cases

Signed-off-by: zotanika <[email protected]>

* [Caffe Frontend] amending a test case for Deconv op

Signed-off-by: zotanika <[email protected]>

* explicit importing tvm.testing

* changing split axis to 0, according to PR apache#9336
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants