[MXNET-670] shape_array and size_array operator is non-differentiable

[ijkguo]

Description

mxnet.base.MXNetError: [00:21:26] src/pass/gradient.cc:192: Operator shape_array is non-differentiable because it didn't register FGradient attribute.

shape_array and size_array are generating new array, similar to ones_like. To use them in training, zero gradient should be attached.

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant JIRA issue created (except PRs with tiny changes)
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage:
• Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
• Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
• Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
• Code is well-documented:
• For user-facing API changes, API doc string has been updated.
• For new C++ functions in header files, their functionalities and arguments are documented.
• For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
• Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html
• To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

• Feature1, tests, (and when applicable, API doc)
• Feature2, tests, (and when applicable, API doc)

Comments

• If this change is a backward incompatible change, why must this change be made.
• Interesting edge cases to note here

[ijkguo]

#10789 This issue discussed generating array using a runtime shape information is impossible right now.

#10889 This pr added shape_array and size_array to generate output of shape information.

Use case in training

img = mx.sym.var('img')
box = mx.sym.var('box')
im_shape = mx.sym.shape_array(img)
box = mx.sym.maximum(box, im_shape)