[singa] branch dev updated (d95b0a6 -> c2e98d5)

[wa...@apache.org]

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

wangwei pushed a change to branch dev
in repository https://gitbox.apache.org/repos/asf/singa.git.


    from d95b0a6  Merge branch 'master' into dev
     new eecd2b3  rm mkldnn
     new 48b006c  added dnnl device
     new 710b4ee  added and tested softmax
     new c3a2eac  added bn fwd testing training
     new ed64bdc  added batchnorm bwd
     new 621021e  todo: gradient check
     new 0def5c3  added dnnl pooling
     new 41bb4a9  added conv
     new 5a5d543  added batchnorm bwd check
     new 9446881  update dockerfile
     new b94d12f  renamed dnnl obj
     new bb66097  fix pooling bug
     new 4fefb69  update dnnl dockerfile
     new 05c35d1  added pooling tests
     new f85e2f0  remove dup code
     new 8cdf2a4  dockerfile singa dir name
     new fc69867  fix conflicts
     new c2e98d5  Merge pull request #548 from dcslin/dnnl

The 1609 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                             |   2 +-
 cmake/Dependencies.cmake                   |  12 +-
 cmake/Templates/singa_config.h.in          |   2 +-
 include/singa/core/common.h                |  15 +-
 include/singa/core/device.h                |   3 -
 include/singa/utils/dnnl_utils.h           |  55 ++++++
 include/singa/utils/mkldnn_utils.h         |  56 ------
 src/api/config.i.in                        |   2 +-
 src/api/model_operation.i                  |  15 +-
 src/core/device/cpp_cpu.cc                 |  11 +-
 src/core/tensor/tensor_math_cpp.h          |  41 +++++
 src/model/operation/batchnorm.cc           | 232 ++++++++++++-------------
 src/model/operation/batchnorm.h            |  37 ++--
 src/model/operation/convolution.cc         | 228 ++++++++++++------------
 src/model/operation/convolution.h          |  44 ++---
 src/model/operation/pooling.cc             | 146 ++++++----------
 src/model/operation/pooling.h              |  37 ++--
 test/python/test_api.py                    | 267 +++++++++++++++++++++++++++++
 test/singa/test_operation_batchnorm.cc     | 237 ++++---------------------
 test/singa/test_operation_convolution.cc   |  56 +++---
 test/singa/test_operation_pooling.cc       |  89 +---------
 test/singa/test_tensor_math.cc             |  22 ++-
 tool/docker/devel/ubuntu/cuda10/Dockerfile |  12 +-
 tool/docker/devel/ubuntu/cuda9/Dockerfile  |  12 +-
 24 files changed, 825 insertions(+), 808 deletions(-)
 create mode 100644 include/singa/utils/dnnl_utils.h
 delete mode 100644 include/singa/utils/mkldnn_utils.h
 mode change 100644 => 100755 src/api/config.i.in
 mode change 100644 => 100755 test/singa/test_operation_batchnorm.cc