This is an automated email from the ASF dual-hosted git repository.
anirudh2290 pushed a change to branch v1.2.0
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 546a233 [MXNET-491] Use depthwise convolution by cuDNNv7 if
available, updated version (#11076) (#11233)
new bc9a2d2 Add Windows MKLDNN Building Instruction (#10613)
new db24cc0 [MXNET-33] SSD example not working with mkl-dnn (#10021)
new 162cc78 add unittest for gluon mkldnn ndarray slice computation
new 4dcdb17 [MXNET-362] ensure same mkldnn engine is used for consistency
(#10616)
new 1647b70 Add more gluon computation on MKLDNN with memory
operation(slice, reshape etc.) (#10764)
new 18d239c [MXNET-365] handle inplace in mkldnn FallBackCompute (#10591)
new 0111a36 Fix a bug in getting MKLDNN memory (#10731)
new be7f358 fix a bug in cudnn softmax activation. (#10918)
new 22966a0 handle NDArray slice properly for mkldnn layout
new 9e97a96 invalidate MKLDNN memory for reused NDArrays. (#10706)
new dfc847f handle fallback correctly for write inplace when the array is
MKLDNN. (#10651)
new 62a47a7 Fix bugs in MKLDNN. (#10979)
The 12 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "add" were already present in the repository and have only
been added to this reference.
Summary of changes:
CMakeLists.txt | 8 +-
Jenkinsfile | 21 +-
MKL_README.md | 96 ++-
ci/docker/runtime_functions.sh | 15 +
docs/install/windows_setup.md | 4 +-
src/common/exec_utils.h | 15 +-
src/executor/attach_op_execs_pass.cc | 16 +-
src/imperative/imperative_utils.h | 23 +-
src/ndarray/ndarray.cc | 171 +++--
.../nn/cudnn/cudnn_softmax_activation-inl.h | 13 +-
src/operator/nn/mkldnn/mkldnn_base-inl.h | 39 +-
src/operator/nn/mkldnn/mkldnn_base.cc | 30 +-
src/operator/nn/mkldnn/mkldnn_copy.cc | 8 +-
src/operator/nn/mkldnn/mkldnn_pooling-inl.h | 6 +
src/operator/nn/mkldnn/mkldnn_sum.cc | 22 +-
src/operator/tensor/elemwise_binary_op_basic.cc | 12 +-
tests/cpp/include/test_core_op.h | 10 +-
tests/cpp/operator/mkldnn.cc | 565 +++++++++++++++
tests/python/gpu/test_gluon_model_zoo_gpu.py | 19 +-
tests/python/gpu/test_operator_gpu.py | 21 +
.../data/test_mkldnn_test_mkldnn_model_model1.json | 770 +++++++++++++++++++++
tests/python/mkl/test_mkldnn.py | 217 ++++++
tests/python/mkl/test_mkldnn_install.py | 56 ++
23 files changed, 2005 insertions(+), 152 deletions(-)
create mode 100644
tests/python/mkl/data/test_mkldnn_test_mkldnn_model_model1.json
create mode 100644 tests/python/mkl/test_mkldnn.py
create mode 100644 tests/python/mkl/test_mkldnn_install.py
--
To stop receiving notification emails like this one, please contact
anirudh2...@apache.org.
