[TensorFlow][Frontend] Adding InversePermutation Op#8277
Conversation
| tvm.testing.assert_allclose(tf_output[i], tvm_output[i], atol=1e-4, rtol=1e-5) | ||
|
|
||
| sess.close() | ||
| # @tvm.testing.parametrize_targets("llvm", "cuda") |
There was a problem hiding this comment.
Is this test commented out intentionally?
There was a problem hiding this comment.
Sorry, I forgot to uncomment that line.
There was a problem hiding this comment.
Sorry, I've just noticed that I forgot to uncomment the whole test_read_variable_op.
cailun01
force-pushed
the
invert_permutation
branch
from
bd796ae to
13917bc
Compare
|
Hi @cailun01, there are some minor linting issues with your change. Have a look on the documentation at https://tvm.apache.org/docs/contribute/pull_request.html#submit-a-pull-request to see instructions on how to run the linter on your own machine. |
OK, Thanks! |
Computes the inverse permutation of a tensor. This Op is used by Mask R-CNN or other object detection models.
cailun01
force-pushed
the
invert_permutation
branch
from
3d52b71 to
15a4564
Compare
* [TensorFlow][Frontend] Adding InversePermutation Op Computes the inverse permutation of a tensor. This Op is used by Mask R-CNN or other object detection models. * uncomment test_read_variable_op * restore several tests * fix lint error * fix python linting error * fix lint error * restore mistakenly deleted codes
* [TensorFlow][Frontend] Adding InversePermutation Op Computes the inverse permutation of a tensor. This Op is used by Mask R-CNN or other object detection models. * uncomment test_read_variable_op * restore several tests * fix lint error * fix python linting error * fix lint error * restore mistakenly deleted codes
3 participants
Computes the inverse permutation of a tensor. This Op is used by Mask R-CNN
or other object detection models.
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.