You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by ta...@apache.org on 2019/08/26 13:11:05 UTC
[incubator-mxnet] branch mkldnn-v1.0 updated (ced2bdb -> 03b734b)
This is an automated email from the ASF dual-hosted git repository.
taolv pushed a change to branch mkldnn-v1.0
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from ced2bdb Merge remote-tracking branch 'origin' into mkldnn-v1.0
add 843c3ab Add Large Tensor Support for Sequence, NN Ops (#15807)
add 3d8035b Numpy-compatible concatenate upstream (#15894)
add 78733f0 Numpy-compatible concatenate upstream (#15894)
add b98f297 Numpy-compatible concatenate upstream (#15894)
add 9dbfc2d Numpy-compatible concatenate upstream (#15894)
add 40593c6 Fix ConcatType backward type inference (#15829)
add f586c07 [OpPerf] Profiler flag for Python, Cpp (#15881)
add acc7865 added check for empty params file and unknown param (not arg/aux) (#15917)
add c8f0b21 Remove Scala package test on build (#15915)
add 09cf75b Add power, exponent, log ops large tensor support (#15794)
add 5a4c01b [Opperf] Filter out deprecated ops (#15541)
add a8b9728 Add disable attr to subgraph property (#15926)
add 1a6fe60 Refactor for windows CI 'out of heap space' errors (#15922)
add 8eb0f61 Numpy-compatible stack upstream (#15842)
add bcbdc1c Re-enable flaky test_prelu (#15777)
add 64a0502 declare explicitly the tblob default assign operator and copy constructor (#15937)
add c50bfe2 [OpPerf] Handle positional arguments (#15761)
add 92d2319 [OpPerf] Take care of 4d param (#15736)
add 3b7f888 removing unnecessary int64 C apis that were added to support Large Tensors and Vectors (#15944)
add 3dfb19a Discard needless test cases in `test_convolution_independent_gradients` (#15939)
add 5656a18 creating ndarray directly using mxnet ndarray primitives to reduce memory footprint of tests for topk, sort and argsort (#15900)
add cd397a3 Benchmark doc fix (#15769)
add 308e4ac Adding tests to verify support for Large Tensors in additional Ops along with new C_Apis supporting 64bit indexing (#15895)
add d225074 fix naive engine for multi-threaded inference (#15574)
add 0b5526b [Quantization]support exclude operators while quantization (#15910)
add 521e925 [Numpy] Numpy behavior random.uniform() (#15858)
add bdeb7bc Softmax optimization for GPU (#15545)
add 681cfc4 [Dependency Update] [Doc] move the general prerequisite software to the top (#15896)
add 434f185 [MKLDNN]Support fullyconnected and element-wise ops fusion (#15950)
add 9023256 Tvm broadcast backward (#15938)
add fa16a9e Disable test coverage for Clang MKLDNN (#15977)
add fade159 Fix get_rows_per_block (#15979)
add 73a692e Fix a memory misalignment in topk operator (#15948)
add d8c2d85 update support MKLDNN BN conditions (#15870)
add 717d883 Remove requests_failed_to_import handling
add 8d0b895 np elemwise unary ops upstream (#15831)
add 1eb1925 remove Julia cat image for license issue (#15964)
add cba7c4e Add fp16 support for topk (#15560)
add 77a633f Fix Nightly Maven GPU (#15989)
add 2e4242e Add symbol api for randn and fix shape issue for randn ndarray and symbol api (#15772)
add 3b7e484 Decouple dtype from shape for Random multinomial (#15980)
add c7a8a78 Fix dtype inference in arange_like operator (#15930)
add 5042d2a [MKLDNN] Fix out of bound access of req vector (#16000)
add 7d8c2ec Disable laop_6 (#15976)
add d8b6e47 Update CODEOWNERS. (#15972)
add 9410cc4 add uint8 bn mkldnn implementation (#16003)
new 03b734b Merge remote-tracking branch 'origin' into mkldnn-v1.0
The 1 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:
3rdparty/mshadow/mshadow/tensor.h | 6 +-
CODEOWNERS | 3 +-
benchmark/opperf/README.md | 9 +-
benchmark/opperf/nd_operations/array_rearrange.py | 4 +-
benchmark/opperf/nd_operations/binary_operators.py | 8 +-
benchmark/opperf/nd_operations/gemm_operators.py | 6 +-
.../nd_operations/nn_activation_operators.py | 5 +-
.../opperf/nd_operations/nn_basic_operators.py | 5 +-
.../opperf/nd_operations/nn_conv_operators.py | 12 +-
.../opperf/nd_operations/nn_optimizer_operators.py | 4 +-
.../nd_operations/random_sampling_operators.py | 4 +-
.../opperf/nd_operations/reduction_operators.py | 4 +-
.../nd_operations/sorting_searching_operators.py | 4 +-
benchmark/opperf/nd_operations/unary_operators.py | 4 +-
benchmark/opperf/opperf.py | 40 +-
benchmark/opperf/rules/default_params.py | 15 +-
benchmark/opperf/utils/benchmark_utils.py | 67 +-
benchmark/opperf/utils/common_utils.py | 38 +-
benchmark/opperf/utils/ndarray_utils.py | 14 +-
benchmark/opperf/utils/op_registry_utils.py | 14 +-
benchmark/opperf/utils/profiler_utils.py | 57 +-
ci/jenkins/Jenkinsfile_clang | 4 +-
ci/jenkins/Jenkinsfile_unix_cpu | 2 +-
ci/publish/scala/build.sh | 2 +-
contrib/tvmop/__init__.py | 1 +
contrib/tvmop/basic/ufunc.py | 54 +-
contrib/tvmop/utils.py | 29 +
docs/api/python/symbol/symbol.md | 1 +
docs/tutorials/c++/subgraphAPI.md | 6 +-
docs/tutorials/mkldnn/mkldnn_quantization.md | 8 +-
example/quantization/imagenet_gen_qsym.py | 13 +-
example/quantization/imagenet_gen_qsym_mkldnn.py | 3 +-
include/mxnet/c_api.h | 45 +-
include/mxnet/tensor_blob.h | 19 +
.../Prediction with Pre-trained Model.ipynb | 4 +-
.../imagenet/ijulia-pretrained-predict/cat.png | Bin 123126 -> 0 bytes
make/maven/maven_linux_cu90mkl.mk | 1 +
make/maven/maven_linux_cu92mkl.mk | 1 +
make/pip/pip_linux_cu100.mk | 2 +
make/pip/pip_linux_cu100mkl.mk | 2 +
make/pip/pip_linux_cu101.mk | 2 +
make/pip/pip_linux_cu101mkl.mk | 2 +
make/pip/pip_linux_cu91.mk | 2 +
make/pip/pip_linux_cu91mkl.mk | 2 +
make/pip/pip_linux_cu92.mk | 2 +
make/pip/pip_linux_cu92mkl.mk | 2 +
python/mxnet/contrib/quantization.py | 130 +-
python/mxnet/gluon/utils.py | 7 +-
python/mxnet/model.py | 19 +-
python/mxnet/ndarray/ndarray.py | 32 +-
python/mxnet/ndarray/numpy/_op.py | 1227 ++++++++++++++++++-
python/mxnet/ndarray/numpy/random.py | 62 +-
python/mxnet/ndarray/random.py | 4 +-
python/mxnet/numpy/multiarray.py | 1237 +++++++++++++++++++-
python/mxnet/numpy/random.py | 37 +-
python/mxnet/symbol/numpy/_symbol.py | 1022 +++++++++++++++-
python/mxnet/symbol/numpy/random.py | 62 +-
python/mxnet/symbol/random.py | 32 +-
src/c_api/c_api.cc | 33 +-
src/c_api/c_api_symbolic.cc | 20 +-
src/common/cuda_utils.cc | 90 ++
src/common/cuda_utils.h | 50 +-
src/engine/naive_engine.cc | 12 +-
src/executor/graph_executor.cc | 9 +-
src/operator/contrib/tvmop/ufunc.cc | 99 +-
src/operator/mxnet_op.h | 30 +
src/operator/nn/batch_norm.cc | 10 +-
src/operator/nn/concat.cc | 32 +-
src/operator/nn/log_softmax.cc | 78 ++
.../{tensor/cast_storage.cu => nn/log_softmax.cu} | 16 +-
src/operator/nn/mkldnn/mkldnn_base-inl.h | 7 +
src/operator/nn/mkldnn/mkldnn_batch_norm-inl.h | 32 +-
src/operator/nn/mkldnn/mkldnn_convolution-inl.h | 11 +-
.../nn/mkldnn/mkldnn_fully_connected-inl.h | 8 +-
src/operator/nn/mkldnn/mkldnn_fully_connected.cc | 10 +-
src/operator/nn/softmax-inl.h | 723 +++++++-----
src/operator/nn/softmax.cc | 99 --
src/operator/nn/softmax.cu | 18 +-
src/operator/nn/softmin.cc | 89 ++
.../{tensor/cast_storage.cu => nn/softmin.cu} | 16 +-
src/operator/numpy/np_elemwise_unary_op_basic.cc | 366 ++++++
src/operator/numpy/np_elemwise_unary_op_basic.cu | 110 ++
src/operator/numpy/np_matrix_op-inl.h | 1 +
src/operator/numpy/np_matrix_op.cc | 99 +-
src/operator/numpy/np_matrix_op.cu | 11 +-
src/operator/numpy/random/dist_common.h | 177 +++
src/operator/numpy/random/np_uniform_op.cc | 63 +
.../random/np_uniform_op.cu} | 13 +-
src/operator/numpy/random/np_uniform_op.h | 207 ++++
.../mkldnn/mkldnn_quantized_batch_norm.cc | 23 +-
src/operator/quantization/quantize_graph_pass.cc | 68 +-
src/operator/quantization/quantized_batch_norm.cc | 6 +
src/operator/random/sample_multinomial_op.h | 7 +-
src/operator/subgraph/mkldnn/mkldnn_conv.cc | 5 +-
.../subgraph/mkldnn/mkldnn_conv_property.h | 7 +-
src/operator/subgraph/mkldnn/mkldnn_fc-inl.h | 83 ++
src/operator/subgraph/mkldnn/mkldnn_fc.cc | 34 +-
src/operator/subgraph/mkldnn/mkldnn_fc_property.h | 102 +-
src/operator/subgraph/subgraph_property.h | 11 +-
.../tensor/broadcast_reduce_minmax_value.cc | 59 +
...p_index.cu => broadcast_reduce_minmax_value.cu} | 24 +-
src/operator/tensor/broadcast_reduce_norm_value.cc | 118 ++
..._op_value.cu => broadcast_reduce_norm_value.cu} | 58 +-
src/operator/tensor/broadcast_reduce_op.h | 15 +
src/operator/tensor/broadcast_reduce_op_value.cc | 239 ----
src/operator/tensor/broadcast_reduce_op_value.cu | 71 --
src/operator/tensor/broadcast_reduce_prod_value.cc | 59 +
..._op_index.cu => broadcast_reduce_prod_value.cu} | 24 +-
src/operator/tensor/broadcast_reduce_sum_value.cc | 113 ++
...e_op_index.cu => broadcast_reduce_sum_value.cu} | 28 +-
src/operator/tensor/dot-inl.h | 2 +-
src/operator/tensor/elemwise_unary_op_basic.cc | 377 ------
src/operator/tensor/elemwise_unary_op_basic.cu | 101 --
src/operator/tensor/elemwise_unary_op_logexp.cc | 230 ++++
src/operator/tensor/elemwise_unary_op_logexp.cu | 76 ++
src/operator/tensor/elemwise_unary_op_pow.cc | 212 ++++
src/operator/tensor/elemwise_unary_op_pow.cu | 82 ++
src/operator/tensor/init_op.cc | 2 +-
src/operator/tensor/init_op.h | 4 -
src/operator/tensor/ordering_op-inl.h | 46 +-
src/operator/tensor/sort_op-inl.cuh | 68 +-
tests/nightly/test_large_array.py | 692 ++++++++++-
tests/nightly/test_large_vector.py | 1 +
tests/python/gpu/test_contrib_amp.py | 19 +-
tests/python/gpu/test_operator_gpu.py | 16 +
tests/python/mkl/test_subgraph.py | 48 +-
tests/python/quantization/test_quantization.py | 69 +-
tests/python/unittest/test_numpy_ndarray.py | 36 +
tests/python/unittest/test_numpy_op.py | 251 ++++
tests/python/unittest/test_operator.py | 72 +-
tests/python/unittest/test_random.py | 35 +-
tests/python/unittest/test_tvm_op.py | 49 +-
tools/dependencies/README.md | 36 +-
tools/dependencies/openblas.sh | 2 +-
tools/setup_gpu_build_tools.sh | 9 +-
135 files changed, 8697 insertions(+), 1934 deletions(-)
delete mode 100644 julia/examples/imagenet/ijulia-pretrained-predict/cat.png
create mode 100644 src/common/cuda_utils.cc
create mode 100644 src/operator/nn/log_softmax.cc
copy src/operator/{tensor/cast_storage.cu => nn/log_softmax.cu} (67%)
create mode 100644 src/operator/nn/softmin.cc
copy src/operator/{tensor/cast_storage.cu => nn/softmin.cu} (67%)
create mode 100644 src/operator/numpy/np_elemwise_unary_op_basic.cc
create mode 100644 src/operator/numpy/np_elemwise_unary_op_basic.cu
create mode 100644 src/operator/numpy/random/dist_common.h
create mode 100644 src/operator/numpy/random/np_uniform_op.cc
copy src/operator/{quantization/quantize_v2.cu => numpy/random/np_uniform_op.cu} (77%)
create mode 100644 src/operator/numpy/random/np_uniform_op.h
create mode 100644 src/operator/subgraph/mkldnn/mkldnn_fc-inl.h
create mode 100644 src/operator/tensor/broadcast_reduce_minmax_value.cc
copy src/operator/tensor/{broadcast_reduce_op_index.cu => broadcast_reduce_minmax_value.cu} (62%)
create mode 100644 src/operator/tensor/broadcast_reduce_norm_value.cc
copy src/operator/tensor/{broadcast_reduce_op_value.cu => broadcast_reduce_norm_value.cu} (50%)
create mode 100644 src/operator/tensor/broadcast_reduce_prod_value.cc
copy src/operator/tensor/{broadcast_reduce_op_index.cu => broadcast_reduce_prod_value.cu} (60%)
create mode 100644 src/operator/tensor/broadcast_reduce_sum_value.cc
copy src/operator/tensor/{broadcast_reduce_op_index.cu => broadcast_reduce_sum_value.cu} (54%)
create mode 100644 src/operator/tensor/elemwise_unary_op_logexp.cc
create mode 100644 src/operator/tensor/elemwise_unary_op_logexp.cu
create mode 100644 src/operator/tensor/elemwise_unary_op_pow.cc
create mode 100644 src/operator/tensor/elemwise_unary_op_pow.cu
[incubator-mxnet] 01/01: Merge remote-tracking branch 'origin' into
mkldnn-v1.0
Posted by ta...@apache.org.
This is an automated email from the ASF dual-hosted git repository.
taolv pushed a commit to branch mkldnn-v1.0
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
commit 03b734b7f40109c28c0a7af94a2598acdcfdee2b
Merge: ced2bdb 9410cc4
Author: Tao Lv <ta...@intel.com>
AuthorDate: Mon Aug 26 21:04:40 2019 +0800
Merge remote-tracking branch 'origin' into mkldnn-v1.0
Conflicts:
tests/python/mkl/test_subgraph.py
3rdparty/mshadow/mshadow/tensor.h | 6 +-
CODEOWNERS | 3 +-
benchmark/opperf/README.md | 9 +-
benchmark/opperf/nd_operations/array_rearrange.py | 4 +-
benchmark/opperf/nd_operations/binary_operators.py | 8 +-
benchmark/opperf/nd_operations/gemm_operators.py | 6 +-
.../nd_operations/nn_activation_operators.py | 5 +-
.../opperf/nd_operations/nn_basic_operators.py | 5 +-
.../opperf/nd_operations/nn_conv_operators.py | 12 +-
.../opperf/nd_operations/nn_optimizer_operators.py | 4 +-
.../nd_operations/random_sampling_operators.py | 4 +-
.../opperf/nd_operations/reduction_operators.py | 4 +-
.../nd_operations/sorting_searching_operators.py | 4 +-
benchmark/opperf/nd_operations/unary_operators.py | 4 +-
benchmark/opperf/opperf.py | 40 +-
benchmark/opperf/rules/default_params.py | 15 +-
benchmark/opperf/utils/benchmark_utils.py | 67 +-
benchmark/opperf/utils/common_utils.py | 38 +-
benchmark/opperf/utils/ndarray_utils.py | 14 +-
benchmark/opperf/utils/op_registry_utils.py | 14 +-
benchmark/opperf/utils/profiler_utils.py | 57 +-
ci/jenkins/Jenkinsfile_clang | 4 +-
ci/jenkins/Jenkinsfile_unix_cpu | 2 +-
ci/publish/scala/build.sh | 2 +-
contrib/tvmop/__init__.py | 1 +
contrib/tvmop/basic/ufunc.py | 54 +-
contrib/tvmop/utils.py | 29 +
docs/api/python/symbol/symbol.md | 1 +
docs/tutorials/c++/subgraphAPI.md | 6 +-
docs/tutorials/mkldnn/mkldnn_quantization.md | 8 +-
example/quantization/imagenet_gen_qsym.py | 13 +-
example/quantization/imagenet_gen_qsym_mkldnn.py | 3 +-
include/mxnet/c_api.h | 45 +-
include/mxnet/tensor_blob.h | 19 +
.../Prediction with Pre-trained Model.ipynb | 4 +-
.../imagenet/ijulia-pretrained-predict/cat.png | Bin 123126 -> 0 bytes
make/maven/maven_linux_cu90mkl.mk | 1 +
make/maven/maven_linux_cu92mkl.mk | 1 +
make/pip/pip_linux_cu100.mk | 2 +
make/pip/pip_linux_cu100mkl.mk | 2 +
make/pip/pip_linux_cu101.mk | 2 +
make/pip/pip_linux_cu101mkl.mk | 2 +
make/pip/pip_linux_cu91.mk | 2 +
make/pip/pip_linux_cu91mkl.mk | 2 +
make/pip/pip_linux_cu92.mk | 2 +
make/pip/pip_linux_cu92mkl.mk | 2 +
python/mxnet/contrib/quantization.py | 130 +-
python/mxnet/gluon/utils.py | 7 +-
python/mxnet/model.py | 19 +-
python/mxnet/ndarray/ndarray.py | 32 +-
python/mxnet/ndarray/numpy/_op.py | 1227 ++++++++++++++++++-
python/mxnet/ndarray/numpy/random.py | 62 +-
python/mxnet/ndarray/random.py | 4 +-
python/mxnet/numpy/multiarray.py | 1237 +++++++++++++++++++-
python/mxnet/numpy/random.py | 37 +-
python/mxnet/symbol/numpy/_symbol.py | 1022 +++++++++++++++-
python/mxnet/symbol/numpy/random.py | 62 +-
python/mxnet/symbol/random.py | 32 +-
src/c_api/c_api.cc | 33 +-
src/c_api/c_api_symbolic.cc | 20 +-
src/common/cuda_utils.cc | 90 ++
src/common/cuda_utils.h | 50 +-
src/engine/naive_engine.cc | 12 +-
src/executor/graph_executor.cc | 9 +-
src/operator/contrib/tvmop/ufunc.cc | 99 +-
src/operator/mxnet_op.h | 30 +
src/operator/nn/batch_norm.cc | 10 +-
src/operator/nn/concat.cc | 32 +-
src/operator/nn/log_softmax.cc | 78 ++
src/operator/nn/{softmax.cu => log_softmax.cu} | 18 +-
src/operator/nn/mkldnn/mkldnn_base-inl.h | 7 +
src/operator/nn/mkldnn/mkldnn_batch_norm-inl.h | 32 +-
src/operator/nn/mkldnn/mkldnn_convolution-inl.h | 11 +-
.../nn/mkldnn/mkldnn_fully_connected-inl.h | 8 +-
src/operator/nn/mkldnn/mkldnn_fully_connected.cc | 10 +-
src/operator/nn/softmax-inl.h | 723 +++++++-----
src/operator/nn/softmax.cc | 99 --
src/operator/nn/softmax.cu | 18 +-
src/operator/nn/softmin.cc | 89 ++
src/operator/nn/{softmax.cu => softmin.cu} | 18 +-
src/operator/numpy/np_elemwise_unary_op_basic.cc | 366 ++++++
src/operator/numpy/np_elemwise_unary_op_basic.cu | 110 ++
src/operator/numpy/np_matrix_op-inl.h | 1 +
src/operator/numpy/np_matrix_op.cc | 99 +-
src/operator/numpy/np_matrix_op.cu | 11 +-
src/operator/numpy/random/dist_common.h | 177 +++
src/operator/numpy/random/np_uniform_op.cc | 63 +
.../{np_matrix_op.cu => random/np_uniform_op.cu} | 20 +-
src/operator/numpy/random/np_uniform_op.h | 207 ++++
.../mkldnn/mkldnn_quantized_batch_norm.cc | 23 +-
src/operator/quantization/quantize_graph_pass.cc | 68 +-
src/operator/quantization/quantized_batch_norm.cc | 6 +
src/operator/random/sample_multinomial_op.h | 7 +-
src/operator/subgraph/mkldnn/mkldnn_conv.cc | 5 +-
.../subgraph/mkldnn/mkldnn_conv_property.h | 7 +-
src/operator/subgraph/mkldnn/mkldnn_fc-inl.h | 83 ++
src/operator/subgraph/mkldnn/mkldnn_fc.cc | 34 +-
src/operator/subgraph/mkldnn/mkldnn_fc_property.h | 102 +-
src/operator/subgraph/subgraph_property.h | 11 +-
.../tensor/broadcast_reduce_minmax_value.cc | 59 +
.../broadcast_reduce_minmax_value.cu} | 24 +-
src/operator/tensor/broadcast_reduce_norm_value.cc | 118 ++
..._op_value.cu => broadcast_reduce_norm_value.cu} | 58 +-
src/operator/tensor/broadcast_reduce_op.h | 15 +
src/operator/tensor/broadcast_reduce_op_value.cc | 239 ----
src/operator/tensor/broadcast_reduce_op_value.cu | 71 --
src/operator/tensor/broadcast_reduce_prod_value.cc | 59 +
.../broadcast_reduce_prod_value.cu} | 24 +-
src/operator/tensor/broadcast_reduce_sum_value.cc | 113 ++
.../broadcast_reduce_sum_value.cu} | 30 +-
src/operator/tensor/dot-inl.h | 2 +-
src/operator/tensor/elemwise_unary_op_basic.cc | 377 ------
src/operator/tensor/elemwise_unary_op_basic.cu | 101 --
src/operator/tensor/elemwise_unary_op_logexp.cc | 230 ++++
src/operator/tensor/elemwise_unary_op_logexp.cu | 76 ++
src/operator/tensor/elemwise_unary_op_pow.cc | 212 ++++
src/operator/tensor/elemwise_unary_op_pow.cu | 82 ++
src/operator/tensor/init_op.cc | 2 +-
src/operator/tensor/init_op.h | 4 -
src/operator/tensor/ordering_op-inl.h | 46 +-
src/operator/tensor/sort_op-inl.cuh | 68 +-
tests/nightly/test_large_array.py | 692 ++++++++++-
tests/nightly/test_large_vector.py | 1 +
tests/python/gpu/test_contrib_amp.py | 19 +-
tests/python/gpu/test_operator_gpu.py | 16 +
tests/python/mkl/test_subgraph.py | 48 +-
tests/python/quantization/test_quantization.py | 69 +-
tests/python/unittest/test_numpy_ndarray.py | 36 +
tests/python/unittest/test_numpy_op.py | 251 ++++
tests/python/unittest/test_operator.py | 72 +-
tests/python/unittest/test_random.py | 35 +-
tests/python/unittest/test_tvm_op.py | 49 +-
tools/dependencies/README.md | 36 +-
tools/dependencies/openblas.sh | 2 +-
tools/setup_gpu_build_tools.sh | 9 +-
135 files changed, 8691 insertions(+), 1953 deletions(-)
diff --cc tests/python/mkl/test_subgraph.py
index 6bd18c5,eba77f2..af8f54b
--- a/tests/python/mkl/test_subgraph.py
+++ b/tests/python/mkl/test_subgraph.py
@@@ -821,10 -828,12 +844,13 @@@ def test_single_fc()
@with_seed()
+@unittest.skip('skip for MKL-DNN 1.0 integration: https://github.com/apache/incubator-mxnet/projects/16')
- def test_fc_relu():
- for dshape, no_bias, flatten in itertools.product(DATA_SHAPE, [True, False], [True, False]):
- syms, attrs = fc_relu(no_bias, dshape, flatten)
+ def test_fc_eltwise():
+ for dshape, no_bias, flatten, alg in itertools.product(DATA_SHAPE,
+ [True, False],
+ [True, False],
+ fc_post_ops_list):
+ syms, attrs = fc_eltwise(no_bias, dshape, flatten, alg)
if flatten is True:
check_fusion(syms, dshape, attrs, check_quantization=True)
else: