[incubator-mxnet] branch master updated (4a13edd -> e1435a3)

[ha...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

haoj pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.


    from 4a13edd  [BUGFIX] fix model zoo parallel download (#17372)
     add e1435a3  [NumPy] Add NumPy support for norm (#17014)

No new revisions were added by this update.

Summary of changes:
 3rdparty/mshadow/mshadow/base.h                    |  10 +
 python/mxnet/ndarray/numpy/linalg.py               | 161 +++-
 python/mxnet/symbol/numpy/linalg.py                | 164 +++-
 .../linalg/broadcast_reduce_customized-inl.cuh     | 416 ++++++++++
 .../numpy/linalg/broadcast_reduce_customized-inl.h | 181 +++++
 .../numpy/linalg/broadcast_reduce_op_customized.h  | 168 +++++
 src/operator/numpy/linalg/np_gesvd-inl.h           |  13 +-
 src/operator/numpy/linalg/np_norm-inl.h            | 836 +++++++++++++++++++++
 src/operator/numpy/linalg/np_norm.cc               | 204 +++++
 .../np_norm_backward.cc}                           |  18 +-
 .../crop.cu => numpy/linalg/np_norm_backward.cu}   |  17 +-
 .../linalg/np_norm_forward.cc}                     |  40 +-
 .../crop.cu => numpy/linalg/np_norm_forward.cu}    |  17 +-
 .../python/unittest/test_numpy_interoperability.py |  16 +-
 tests/python/unittest/test_numpy_op.py             | 120 ++-
 15 files changed, 2268 insertions(+), 113 deletions(-)
 create mode 100644 src/operator/numpy/linalg/broadcast_reduce_customized-inl.cuh
 create mode 100644 src/operator/numpy/linalg/broadcast_reduce_customized-inl.h
 create mode 100644 src/operator/numpy/linalg/broadcast_reduce_op_customized.h
 create mode 100644 src/operator/numpy/linalg/np_norm-inl.h
 create mode 100644 src/operator/numpy/linalg/np_norm.cc
 copy src/operator/numpy/{random/np_bernoulli_op.cu => linalg/np_norm_backward.cc} (64%)
 copy src/operator/{image/crop.cu => numpy/linalg/np_norm_backward.cu} (75%)
 copy src/operator/{contrib/modulated_deformable_convolution.cu => numpy/linalg/np_norm_forward.cc} (53%)
 copy src/operator/{image/crop.cu => numpy/linalg/np_norm_forward.cu} (76%)