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.
omit cc22cb8 [Do not review] [Do not merge] New numpy-compatible sum
(#14739)
add 1af29e9 Fixes for wine detection tutorial (#13886)
add 3a46980 Use USE_SIGNAL_HANDLER by default set to ON in CMakeLists.txt
(#14599)
add 07aef13 Add unpooled gpu memory type (#14716)
add 1238aa0 Revert "use mkl sparse matrix to improve performance
(#14492)" (#14806)
add c7577e5 added extraction/generation of diagonal and triangonal
matrices to linalg (#14501)
add 5fda0a5 add clojure tutorials to index (#14814)
add 488fad2 Fix iterator over symbol when multiple children have the same
name (#14597)
add cdd7087 Fix Clojure BERT example's context argument (#14843)
add 84c1635 [MKLDNN] add quantized sum (#14614)
add bde1b84 [int8] Add MobileNetV2_1.0 & ResNet18 Quantization (#14823)
add 1c874cf reformat trt to use subgraph API, add fp16 support (#14040)
add 5e5a59e Upgrade Pylint version to 2.3.1 (#14807)
add 977e558 Make docblocks for Gluon BatchNorm and SyncBatchNorm
consistent with the code (#14840)
add 381a9da Print reproduction command on CI failure (#14815)
add e17b7e2 [MXNET-13578] Fix cmake installation failed (#14692)
new 3042a91 [Do not review] [Do not merge] New numpy-compatible sum
(#14739)
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 (cc22cb8)
\
N -- N -- N refs/heads/numpy (3042a91)
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 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/onnx-tensorrt | 2 +-
3rdparty/sparse-matrix/Makefile | 21 -
3rdparty/sparse-matrix/sparse_matrix.cc | 45 --
3rdparty/sparse-matrix/sparse_matrix.h | 48 --
CMakeLists.txt | 14 +-
Makefile | 36 +-
ci/build.py | 1 +
ci/docker/Dockerfile.build.ubuntu_gpu_tensorrt | 1 +
ci/docker/install/ubuntu_publish.sh | 4 +-
ci/docker/install/ubuntu_python.sh | 4 +-
ci/jenkins/Jenkins_steps.groovy | 2 +-
contrib/clojure-package/examples/bert-qa/README.md | 5 +-
.../examples/bert-qa/src/bert_qa/infer.clj | 53 +-
cpp-package/tests/travis/setup.sh | 2 +-
docs/api/python/ndarray/linalg.md | 6 +-
docs/api/python/symbol/linalg.md | 6 +-
docs/faq/env_var.md | 6 +-
docs/install/requirements.txt | 2 +-
docs/tutorials/embedded/wine_detector.md | 51 +-
docs/tutorials/index.md | 21 +-
example/quantization/README.md | 49 +-
example/quantization/imagenet_gen_qsym_mkldnn.py | 20 +-
include/mxnet/c_api.h | 1 -
python/mxnet/base.py | 2 +-
.../mxnet/contrib/onnx/mx2onnx/_export_helper.py | 2 +-
.../mxnet/contrib/onnx/mx2onnx/_op_translations.py | 2 +-
.../contrib/onnx/onnx2mx/_translation_utils.py | 2 +-
python/mxnet/contrib/quantization.py | 1 +
python/mxnet/contrib/tensorrt.py | 119 ++--
python/mxnet/contrib/text/vocab.py | 2 +-
python/mxnet/gluon/contrib/nn/basic_layers.py | 8 +-
python/mxnet/gluon/nn/basic_layers.py | 8 +-
python/mxnet/gluon/parameter.py | 2 +-
python/mxnet/gluon/trainer.py | 4 +-
python/mxnet/gluon/utils.py | 2 +-
python/mxnet/image/detection.py | 2 +-
python/mxnet/image/image.py | 2 +-
python/mxnet/io/io.py | 2 +
python/mxnet/model.py | 2 +-
python/mxnet/module/base_module.py | 2 +-
python/mxnet/module/python_module.py | 2 +-
python/mxnet/ndarray/contrib.py | 4 +-
python/mxnet/ndarray/sparse.py | 4 +-
python/mxnet/ndarray_doc.py | 2 +-
python/mxnet/operator.py | 2 +-
python/mxnet/optimizer/optimizer.py | 2 +-
python/mxnet/recordio.py | 4 +
python/mxnet/rnn/rnn_cell.py | 4 +-
python/mxnet/symbol/symbol.py | 4 +-
python/mxnet/test_utils.py | 14 +-
src/c_api/c_api_executor.cc | 86 +--
src/common/serialization.h | 318 -----------
src/executor/exec_pass.h | 12 -
src/executor/graph_executor.cc | 10 +
src/executor/graph_executor.h | 1 +
src/executor/tensorrt_pass.cc | 596 ---------------------
src/executor/trt_graph_executor.cc | 443 ---------------
src/executor/trt_graph_executor.h | 111 ----
src/initialize.cc | 4 +-
src/operator/contrib/tensorrt-inl.h | 79 ---
src/operator/contrib/tensorrt.cc | 181 -------
src/operator/nn/mkldnn/mkldnn_base.cc | 1 +
.../mkldnn/mkldnn_quantized_elemwise_add.cc | 206 +++++++
src/operator/quantization/quantization_utils.h | 1 +
.../quantization/quantized_elemwise_add-inl.h | 58 ++
.../quantization/quantized_elemwise_add.cc | 141 +++++
src/operator/subgraph/build_subgraph.cc | 16 +-
..._property.h => mkldnn_post_quantize_property.h} | 63 ++-
.../subgraph/mkldnn/mkldnn_subgraph_property.cc | 4 +-
.../tensorrt}/nnvm_to_onnx-inl.h | 131 ++---
.../{contrib => subgraph/tensorrt}/nnvm_to_onnx.cc | 281 ++++++----
.../subgraph/tensorrt}/onnx_to_tensorrt.cc | 27 +-
.../subgraph/tensorrt}/onnx_to_tensorrt.h | 18 +-
src/operator/subgraph/tensorrt/tensorrt-inl.h | 240 +++++++++
src/operator/subgraph/tensorrt/tensorrt.cc | 336 ++++++++++++
.../{contrib => subgraph/tensorrt}/tensorrt.cu | 30 +-
src/operator/tensor/dot-inl.h | 28 +-
src/operator/tensor/elemwise_binary_op_basic.cc | 3 +
src/operator/tensor/la_op-inl.h | 94 ++++
src/operator/tensor/la_op.cc | 231 ++++++++
src/operator/tensor/la_op.cu | 24 +
src/operator/tensor/la_op.h | 68 +++
src/storage/storage.cc | 10 +-
tests/cpp/misc/serialization.cc | 68 ---
tests/python/quantization/test_quantization.py | 70 ++-
tests/python/tensorrt/lenet5_train.py | 2 +
tests/python/tensorrt/test_cvnets.py | 34 +-
tests/python/tensorrt/test_cycle.py | 69 ---
tests/python/tensorrt/test_resnet18.py | 42 +-
tests/python/tensorrt/test_tensorrt_lenet5.py | 91 ++--
tests/python/tensorrt/test_training_warning.py | 70 ---
tests/python/unittest/test_operator.py | 45 ++
tests/python/unittest/test_symbol.py | 5 +
tools/caffe_converter/compare_layers.py | 5 -
tools/caffe_converter/test_converter.py | 2 -
95 files changed, 2209 insertions(+), 2752 deletions(-)
delete mode 100644 3rdparty/sparse-matrix/Makefile
delete mode 100644 3rdparty/sparse-matrix/sparse_matrix.cc
delete mode 100644 3rdparty/sparse-matrix/sparse_matrix.h
delete mode 100644 src/common/serialization.h
delete mode 100644 src/executor/tensorrt_pass.cc
delete mode 100644 src/executor/trt_graph_executor.cc
delete mode 100644 src/executor/trt_graph_executor.h
delete mode 100644 src/operator/contrib/tensorrt-inl.h
delete mode 100644 src/operator/contrib/tensorrt.cc
create mode 100644
src/operator/quantization/mkldnn/mkldnn_quantized_elemwise_add.cc
create mode 100644 src/operator/quantization/quantized_elemwise_add-inl.h
create mode 100644 src/operator/quantization/quantized_elemwise_add.cc
rename src/operator/subgraph/mkldnn/{mkldnn_conv_post_quantize_property.h =>
mkldnn_post_quantize_property.h} (68%)
rename src/operator/{contrib => subgraph/tensorrt}/nnvm_to_onnx-inl.h (61%)
rename src/operator/{contrib => subgraph/tensorrt}/nnvm_to_onnx.cc (68%)
rename src/{executor => operator/subgraph/tensorrt}/onnx_to_tensorrt.cc (89%)
rename src/{executor => operator/subgraph/tensorrt}/onnx_to_tensorrt.h (88%)
create mode 100644 src/operator/subgraph/tensorrt/tensorrt-inl.h
create mode 100644 src/operator/subgraph/tensorrt/tensorrt.cc
rename src/operator/{contrib => subgraph/tensorrt}/tensorrt.cu (69%)
delete mode 100644 tests/cpp/misc/serialization.cc
delete mode 100644 tests/python/tensorrt/test_cycle.py
delete mode 100644 tests/python/tensorrt/test_training_warning.py
