You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by re...@apache.org on 2019/03/13 18:53:10 UTC
[incubator-mxnet] branch numpy updated (d8e9235 -> b4497e7)
This is an automated email from the ASF dual-hosted git repository.
reminisce pushed a change to branch numpy
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
discard d8e9235 [numpy] Shape support scalar tensor (#14315)
new f2497aa Updated news.md with the latest mkldnn submodule version (#14298)
new 49d7fc6 Enhance gpu quantization (#14094)
new d6eafca Bypass ThreadedEngine in test_operator_gpu.py:test_convolution_multiple_streams. (#14338)
new 111b881 Limit workspace for cudnnGet results (#14326)
new 19d737f [MXNET-1331] Removal of non-MXNET classes from JAR (#14303)
new 184c2a5 fix render issue in NDArray linalg docs (#14258)
new fccce20 Add more support for mxnet_to_coreml (#14222)
new a0f3f92 Add default parameters for Scala NDArray.arange (#13816)
new b486594 Register fake grad to subgraph and quantized operators (#14275)
new 39412b3 corrected a spellign (#14247)
new 83d2c2d [MXNET-1324] Add NaiveRunGraph to imperative utils (#14192)
new 8ab7998 Updates build_lib.sh to copy the cub library license (#14347)
new efb8823 Add MKLDNN headers to pip package (#14339)
new 7b8e3a9 compatibility with opencv4 (#14313)
new f4ab2d7 [MXNET-1291] solve pylint errors in examples with issue no.12205 (#13848)
new 49932fa #14199: catch subprocess.CalledProcessError in get_gpus() (#14212)
new 6caaa38 print error message for mxnet::cpp::Operator::Invoke when failed (#14318)
new 8beea18 Bulked op segments to allow Variable nodes (#14200)
new e703694 Fixes #14181, validate model output shape for ObjectDetector. (#14215)
new 12c41e6 Optimizer MXKVStoreUpdater bug fix in serializeState method (#14337)
new 2b7d57d Installs qemu pip requirements from qemu requirements file (#14355)
new 838e256 Optimize NMS part 2 (#14352)
new 30b1cbc add exception (#14362)
new 8668db7 MKLDNN based Quantized FullyConnected Operator and its fusion (#14128)
new ce9e3cf add pos_weight for SigmoidBinaryCrossEntropyLoss (#13612)
new ed83071 Julia: split symbolic-node.jl into several snippets (#14024)
new c645591 Fix NaN value comparisons in relu, max and min ops (#14262)
new 8be97d7 [clojure-package][wip] add `->nd-vec` function in `ndarray.clj` (#14308)
new 35098b8 support leading dimension of -1 in ravel/unravel (#14356)
new 4f5cba5 fix engine crash in shutdown phase (#14382)
new 47d4d66 Flaky test https://github.com/apache/incubator-mxnet/issues/14189 (#14190)
new af41af5 Julia: rename `mx.clip` to `clamp` for `NDArray` (#14027)
new 2df5756 add backgroud class in box_nms (#14058)
new 0e8c270 CI Changes for Codified Windows AMIs (#14336)
new a4b9802 [Clojure] Helper function for n-dim vector to ndarray (#14305)
new 89bebd1 [DOC] fix sym.arange doc (#14237)
new 73b29fa Julia: add binding for runtime feature detection (#13992)
new ab0ca86 [MXNET-1093] Add python3 Docker images for each MXNet release (#12791)
new 66c74cc Enable bulking test on windows (#14392)
new 6aa8c27 [MXNET-1327] Allow RNN Layers to be initialized to fp16 (#14219)
new c4cae6e Disables flaky test_operator.test_sgld test (#14410)
new 82504ad Fix relative difference scala (#14417)
new b4497e7 [numpy] Shape support scalar tensor (#14315)
This update added new revisions after undoing existing revisions.
That is to say, some revisions that were in the old version of the
branch are not in the new version. This situation occurs
when a user --force pushes a change and generates a repository
containing something like this:
* -- * -- B -- O -- O -- O (d8e9235)
\
N -- N -- N refs/heads/numpy (b4497e7)
You should already have received notification emails for all of the O
revisions, and so the following emails describe only the N revisions
from the common base, B.
Any revisions marked "omit" are not gone; other references still
refer to them. Any revisions marked "discard" are gone forever.
The 9389 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:
CONTRIBUTORS.md | 5 +-
Makefile | 30 +-
NEWS.md | 2 +-
benchmark/python/control_flow/rnn.py | 7 +-
ci/build_windows.py | 6 +-
ci/docker/qemu/runtime_functions.py | 4 +-
ci/docker/qemu/vmcontrol.py | 1 +
ci/jenkins/Jenkins_steps.groovy | 32 +-
ci/jenkins/Jenkinsfile_windows_cpu | 10 +-
ci/jenkins/Jenkinsfile_windows_gpu | 10 +-
ci/qemu/README.md | 4 +
ci/qemu/test_requirements.txt | 3 +
ci/windows/test_py2_cpu.ps1 | 6 +-
ci/windows/test_py2_gpu.ps1 | 10 +-
ci/windows/test_py3_cpu.ps1 | 6 +-
ci/windows/test_py3_gpu.ps1 | 10 +-
.../src/org/apache/clojure_mxnet/ndarray.clj | 79 +-
.../src/org/apache/clojure_mxnet/util.clj | 20 +
.../test/org/apache/clojure_mxnet/ndarray_test.clj | 24 +
.../test/org/apache/clojure_mxnet/util_test.clj | 27 +
cpp-package/example/alexnet.cpp | 75 +-
cpp-package/example/inception_bn.cpp | 35 +-
cpp-package/example/lenet_with_mxdataiter.cpp | 32 +-
cpp-package/example/resnet.cpp | 55 +-
cpp-package/include/mxnet-cpp/operator.hpp | 8 +-
docker/docker-python/Dockerfile.mxnet.python.cpu | 11 +-
.../docker-python/Dockerfile.mxnet.python.cpu.mkl | 11 +-
.../docker-python/Dockerfile.mxnet.python.gpu.cu80 | 11 +-
.../Dockerfile.mxnet.python.gpu.cu80.mkl | 11 +-
.../docker-python/Dockerfile.mxnet.python.gpu.cu90 | 11 +-
.../Dockerfile.mxnet.python.gpu.cu90.mkl | 11 +-
.../docker-python/Dockerfile.mxnet.python.gpu.cu92 | 11 +-
.../Dockerfile.mxnet.python.gpu.cu92.mkl | 11 +-
...net.python.cpu => Dockerfile.mxnet.python3.cpu} | 11 +-
...on.cpu.mkl => Dockerfile.mxnet.python3.cpu.mkl} | 11 +-
....gpu.cu80 => Dockerfile.mxnet.python3.gpu.cu80} | 11 +-
...0.mkl => Dockerfile.mxnet.python3.gpu.cu80.mkl} | 11 +-
....gpu.cu90 => Dockerfile.mxnet.python3.gpu.cu90} | 11 +-
...0.mkl => Dockerfile.mxnet.python3.gpu.cu90.mkl} | 11 +-
....gpu.cu92 => Dockerfile.mxnet.python3.gpu.cu92} | 11 +-
...2.mkl => Dockerfile.mxnet.python3.gpu.cu92.mkl} | 11 +-
docker/docker-python/README.md | 58 +-
docker/docker-python/build_python_dockerfile.sh | 178 +++-
docs/faq/env_var.md | 8 +-
example/bayesian-methods/bdk_demo.py | 1 -
example/caffe/caffe_net.py | 4 +-
example/caffe/train_model.py | 4 +-
.../data_helpers.py | 3 +-
.../cnn_chinese_text_classification/text_cnn.py | 9 +-
example/cnn_text_classification/data_helpers.py | 8 +-
example/cnn_text_classification/text_cnn.py | 162 +++-
example/ctc/captcha_generator.py | 45 +-
example/ctc/ctc_metrics.py | 20 +-
example/ctc/hyperparams.py | 6 +-
example/ctc/lstm_ocr_infer.py | 7 +-
example/ctc/lstm_ocr_train.py | 4 +-
example/ctc/multiproc_data.py | 36 +-
example/ctc/ocr_iter.py | 17 +-
example/ctc/ocr_predict.py | 22 +-
example/deep-embedded-clustering/autoencoder.py | 29 +-
example/deep-embedded-clustering/data.py | 8 +-
example/deep-embedded-clustering/dec.py | 31 +-
example/deep-embedded-clustering/model.py | 3 +-
example/distributed_training/cifar10_dist.py | 78 +-
example/image-classification/README.md | 2 +-
example/image-classification/common/util.py | 8 +-
example/quantization/imagenet_gen_qsym_mkldnn.py | 9 +-
include/mxnet/imperative.h | 23 +-
julia/NEWS.md | 7 +-
julia/deps/build.jl | 7 +-
julia/src/MXNet.jl | 3 +-
julia/src/base.jl | 7 +-
julia/src/deprecated.jl | 8 +-
julia/src/ndarray/arithmetic.jl | 52 +-
julia/src/ndarray/remap.jl | 12 +
julia/src/optimizer.jl | 2 +-
julia/src/runtime.jl | 76 ++
julia/src/symbolic-node.jl | 987 +--------------------
julia/src/symbolic-node/arithmetic.jl | 127 +++
julia/src/symbolic-node/array.jl | 122 +++
julia/src/symbolic-node/autodiff.jl | 178 ++++
julia/src/symbolic-node/io.jl | 58 ++
julia/src/symbolic-node/op.jl | 444 +++++++++
julia/src/symbolic-node/show.jl | 62 ++
julia/src/symbolic-node/type.jl | 123 +++
julia/test/unittest/ndarray.jl | 12 +-
make/config.mk | 4 +
make/crosscompile.jetson.mk | 4 +
make/maven/maven_darwin_mkl.mk | 4 +
make/maven/maven_linux_cu90mkl.mk | 4 +
make/maven/maven_linux_cu92mkl.mk | 4 +
make/maven/maven_linux_mkl.mk | 4 +
make/osx.mk | 4 +
make/pip/pip_darwin_cpu.mk | 4 +
make/pip/pip_darwin_mkl.mk | 4 +
make/pip/pip_linux_cpu.mk | 4 +
make/pip/pip_linux_cu100.mk | 4 +
make/pip/pip_linux_cu100mkl.mk | 4 +
make/pip/pip_linux_cu75.mk | 4 +
make/pip/pip_linux_cu75mkl.mk | 4 +
make/pip/pip_linux_cu80.mk | 4 +
make/pip/pip_linux_cu80mkl.mk | 4 +
make/pip/pip_linux_cu90.mk | 4 +
make/pip/pip_linux_cu90mkl.mk | 4 +
make/pip/pip_linux_cu91.mk | 4 +
make/pip/pip_linux_cu91mkl.mk | 4 +
make/pip/pip_linux_cu92.mk | 4 +
make/pip/pip_linux_cu92mkl.mk | 4 +
make/pip/pip_linux_mkl.mk | 4 +
make/readthedocs.mk | 4 +
python/mxnet/gluon/loss.py | 85 +-
python/mxnet/gluon/rnn/rnn_layer.py | 59 +-
python/mxnet/initializer.py | 13 +
python/mxnet/ndarray/ndarray.py | 7 +-
python/mxnet/symbol/symbol.py | 10 +-
python/mxnet/test_utils.py | 11 +-
python/mxnet/util.py | 16 +
.../assembly/src/main/assembly/assembly.xml | 2 +
.../src/main/scala/org/apache/mxnet/NDArray.scala | 8 +-
.../main/scala/org/apache/mxnet/Optimizer.scala | 16 +-
.../test/scala/org/apache/mxnet/CheckUtils.scala | 4 +-
.../test/scala/org/apache/mxnet/NDArraySuite.scala | 27 +-
scala-package/deploy/src/main/deploy/deploy.xml | 10 +
.../org/apache/mxnet/infer/ImageClassifier.scala | 2 +
.../org/apache/mxnet/infer/ObjectDetector.scala | 25 +-
.../scala/org/apache/mxnet/infer/Predictor.scala | 7 +
src/engine/threaded_engine.h | 9 +
src/executor/graph_executor.cc | 140 ++-
src/executor/graph_executor.h | 6 +-
src/imperative/cached_op.cc | 115 ++-
src/imperative/cached_op.h | 11 +-
src/imperative/imperative.cc | 3 +-
src/imperative/imperative_utils.cc | 235 +++--
src/imperative/imperative_utils.h | 15 +-
src/io/image_aug_default.cc | 1 +
src/io/image_det_aug_default.cc | 1 +
src/io/image_io.cc | 1 +
src/io/opencv_compatibility.h | 59 ++
src/operator/contrib/bounding_box-common.h | 30 +
src/operator/contrib/bounding_box-inl.cuh | 70 +-
src/operator/contrib/bounding_box-inl.h | 76 +-
src/operator/contrib/bounding_box.cc | 8 +-
src/operator/mshadow_op.h | 48 +-
src/operator/nn/cudnn/cudnn_convolution-inl.h | 8 +-
.../nn/mkldnn/mkldnn_fully_connected-inl.h | 133 +++
src/operator/nn/mkldnn/mkldnn_fully_connected.cc | 343 ++++---
src/operator/quantization/dequantize.cc | 3 +
.../mkldnn/mkldnn_quantized_fully_connected.cc | 134 +++
...tized_pooling.cc => mkldnn_quantized_ops-inl.h} | 35 +-
src/operator/quantization/quantize-inl.h | 4 +
src/operator/quantization/quantize.cc | 3 +
src/operator/quantization/quantize_v2-inl.h | 4 +
src/operator/quantization/quantize_v2.cc | 3 +
src/operator/quantization/quantized_concat.cc | 3 +
src/operator/quantization/quantized_conv.cc | 3 +
src/operator/quantization/quantized_flatten.cc | 3 +
.../quantization/quantized_fully_connected.cc | 129 ++-
src/operator/quantization/quantized_pooling.cc | 3 +
src/operator/quantization/requantize.cc | 3 +
src/operator/subgraph/mkldnn/mkldnn_conv.cc | 3 +
.../mkldnn/mkldnn_conv_post_quantize_property.cc | 6 +-
.../subgraph/mkldnn/mkldnn_conv_property.cc | 5 +-
src/operator/subgraph/mkldnn/mkldnn_fc.cc | 442 +++++++++
.../mkldnn/mkldnn_fc_post_quantize_property.cc | 217 +++++
src/operator/subgraph/mkldnn/mkldnn_fc_property.cc | 193 ++++
src/operator/subgraph/subgraph_property.h | 7 +
src/operator/tensor/init_op.h | 5 +
src/operator/tensor/la_op.cc | 16 +-
src/operator/tensor/ravel.cc | 6 +-
src/operator/tensor/ravel.h | 3 +-
tests/python/gpu/test_gluon_gpu.py | 78 ++
tests/python/gpu/test_kvstore_gpu.py | 1 +
tests/python/gpu/test_operator_gpu.py | 88 +-
tests/python/mkl/test_subgraph.py | 214 ++++-
tests/python/quantization/test_quantization.py | 80 +-
tests/python/unittest/common.py | 50 ++
tests/python/unittest/test_contrib_operator.py | 41 +-
tests/python/unittest/test_dynamic_shape.py | 54 ++
tests/python/unittest/test_gluon_rnn.py | 98 +-
tests/python/unittest/test_loss.py | 35 +-
tests/python/unittest/test_ndarray.py | 29 +
tests/python/unittest/test_operator.py | 13 +-
tests/python/unittest/test_optimizer.py | 1 +
tests/requirements.txt | 3 +
tools/bandwidth/test_measure.py | 10 +-
tools/coreml/converter/_layers.py | 46 +-
tools/coreml/converter/_mxnet_converter.py | 3 +
tools/coreml/test/test_mxnet_converter.py | 67 ++
tools/im2rec.cc | 1 +
tools/pip/MANIFEST.in | 1 +
tools/pip/setup.py | 2 +
tools/staticbuild/build_lib.sh | 4 +
192 files changed, 5417 insertions(+), 2192 deletions(-)
create mode 100644 ci/qemu/test_requirements.txt
copy docker/docker-python/{Dockerfile.mxnet.python.cpu => Dockerfile.mxnet.python3.cpu} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.cpu.mkl => Dockerfile.mxnet.python3.cpu.mkl} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.gpu.cu80 => Dockerfile.mxnet.python3.gpu.cu80} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.gpu.cu80.mkl => Dockerfile.mxnet.python3.gpu.cu80.mkl} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.gpu.cu90 => Dockerfile.mxnet.python3.gpu.cu90} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.gpu.cu90.mkl => Dockerfile.mxnet.python3.gpu.cu90.mkl} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.gpu.cu92 => Dockerfile.mxnet.python3.gpu.cu92} (81%)
copy docker/docker-python/{Dockerfile.mxnet.python.gpu.cu92.mkl => Dockerfile.mxnet.python3.gpu.cu92.mkl} (81%)
create mode 100644 julia/src/runtime.jl
create mode 100644 julia/src/symbolic-node/arithmetic.jl
create mode 100644 julia/src/symbolic-node/array.jl
create mode 100644 julia/src/symbolic-node/autodiff.jl
create mode 100644 julia/src/symbolic-node/io.jl
create mode 100644 julia/src/symbolic-node/op.jl
create mode 100644 julia/src/symbolic-node/show.jl
create mode 100644 julia/src/symbolic-node/type.jl
create mode 100644 src/io/opencv_compatibility.h
create mode 100644 src/operator/nn/mkldnn/mkldnn_fully_connected-inl.h
create mode 100644 src/operator/quantization/mkldnn/mkldnn_quantized_fully_connected.cc
copy src/operator/quantization/mkldnn/{mkldnn_quantized_pooling.cc => mkldnn_quantized_ops-inl.h} (55%)
create mode 100644 src/operator/subgraph/mkldnn/mkldnn_fc.cc
create mode 100644 src/operator/subgraph/mkldnn/mkldnn_fc_post_quantize_property.cc
create mode 100644 src/operator/subgraph/mkldnn/mkldnn_fc_property.cc
create mode 100644 tests/python/unittest/test_dynamic_shape.py