Version 0.4.1

[0.4.1] - 2020-06-09

Added

• RAdam optimizer
• Docker build instructions

Version 0.4.0

[0.4.0] - 2019-12-18

Changed

• Upgrade base version of mxnet to mxnet v1.5.1

Fixed

• model converter test
• model visualization

Version 0.3.1

Changed

• Upgrade base version of mxnet to mxnet v1.4.1

Version 0.3.0

[0.3.0] - 2019-05-21

Added

• Model converter, which compresses weights and adapts symbol file:
  • Code is under tools/binary_converter
  • Executable can be found here after the build: build/tools/binary_converter/model-converter
  • Test/example: tests/binary/test_converter.py
• Binary inference on CPU and GPU
• Support for dilation with binary layers

Changed

• Tests now reside in tests/binary
• Code in python/mxnet/visualization.py changed:
  • print_summary now estimates amount of operations (accounts for theoretical speedup)

BMXNet-v2 Beta

[0.2.0] - 2018-12-04

First beta version.
Note that binary (and quantized) examples are maintained separately in a submodule:
example/bmxnet-examples

Added

• Functions:
  • det_sign (ada4ea1d)
  • round_ste (044f81f0)
  • approx_sign
• New operator:
  • contrib.gradcancel (see src/operator/contrib/gradient_cancel[-inl.h|.cc|.cu])
  • allows to cancel gradients (element-wise) if absolute value of input is larger than a threshold
• Binary versions of the following layers of the gluon API:
  • gluon.nn.Dense -> gluon.nn.QDense
  • gluon.nn.Conv1D -> gluon.nn.QConv1D
  • gluon.nn.Conv2D -> gluon.nn.QConv2D
  • gluon.nn.Conv3D -> gluon.nn.QConv3D
• Tests are in tests/python/unittest/test_binary.py
• Layers are in python/mxnet/gluon/nn/binary_layers.py

Changed

• Code in python/mxnet/visualization.py changed:
  • plot_network skips certain layers which clutter binary network graphs
  • print_summary calculates compressed model size and prints number of binarized/quantized and full-precision weights