1.0.39

• Support for Keras H5 format

  -oh5, --output_h5

Full Changelog: 1.0.38...1.0.39

1.0.38

• Concat
  • Support for scalar values
  • TensorFlow does not support Concat for scalar values, so convert to tensor
• ArgMax
  • Support for FusedArgMax
  • https://github.com/tensorflow/models/tree/master/official/projects/edgetpu/vision#argmax-fusion-to-improve-segmentation-model-latency
    
  • convert option

    -rafi64, --replace_argmax_to_fused_argmax_and_indicies_is_int64
    or
    -raff32, --replace_argmax_to_fused_argmax_and_indicies_is_float32

Full Changelog: 1.0.37...1.0.38

1.0.37

• Version information display CLI option added

  -V, --version
    Show version and exit.

  $ onnx2tf -V
  
  1.0.37

1.0.36

• common_functions, TopK, ConstantOfShape
  • Bug fixes

Full Changelog: 1.0.35...1.0.36

1.0.35

• Sum
  • Bug fixes
• Softmax
  • Supports parameter substitution

Full Changelog: 1.0.34...1.0.35

1.0.34

• Dropout
  • Bug fixes
• Reshape
  • Supports combinations of zero geometry and undefined dimensions

Full Changelog: 1.0.33...1.0.34

1.0.33

• onnxsim bug fixes

Full Changelog: 1.0.32...1.0.33

1.0.32

• AveragePool
  • Bug fixes

Full Changelog: 1.0.31...1.0.32

1.0.31

• Improved channel transposition process

Full Changelog: 1.0.30...1.0.31

1.0.30

What's Changed

• Support for onnx2tf.__version__ by @PINTO0309 in #9

• Softmax
  • Bug fixes
• Dropout
  • Bug fixes

Full Changelog: 1.0.29...1.0.30