Onnx2tf

- Prevent `saved_model` (.pb) bloat
- Before
- After


What's Changed
* Prevent saved_model (.pb) bloat by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/227


**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.7.16...1.7.17

- `Softmax`
- Suppress automatic `FlexTraspose` extrapolation behavior by `Softmax` in TensorFlow
- Automatic extrapolation of `FlexTranspose` by `Softmax` occurs when the rank of the input tensor to `Softmax` is 7 or more dimensions and the `axis` is not the last dimension.
- Before
- After

What's Changed
* Suppress automatic `FlexTraspose` extrapolation behavior by `Softmax` in TensorFlow by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/226


**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.7.15...1.7.16

- `Softmax`
- Fixed `Softmax` accuracy check bug
- `Pow`
- Reduced `Pow` calculation error

- `Unimatch`

- [[TODO] [unimatch] Creation of JSON for unimatch accuracy error resolution 203](https://github.com/PINTO0309/onnx2tf/issues/203)

What's Changed
* Fixed `Softmax` accuracy check bug and reduced `Pow` calculation error by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/225


**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.7.14...1.7.15

- Improved error judgment regarding calibration data for INT8 quantization.
- Models with multiple input OPs and non-rgb-image input OPs force automatic calibration to be aborted.
- e.g. [cf_fus.onnx.zip](https://github.com/PINTO0309/onnx2tf/files/10884363/cf_fus.onnx.zip)
python
if model_input.dtype != tf.float32 \
or len(model_input.shape) != 4 \
or model_input.shape[-1] != 3:
print(
f'{Color.RED}ERROR:{Color.RESET} ' +
f'For models that have multiple input OPs and need to perform INT8 quantization calibration '+
f'using non-rgb-image input tensors, specify the calibration data with '+
f'--quant_calib_input_op_name_np_data_path. '+
f'model_input[n].shape: {model_input.shape}'
)
sys.exit(1)

- [[CenterFusion] Model Full Interger Quantize problem 222](https://github.com/PINTO0309/onnx2tf/issues/222)

What's Changed
* Improved error judgment regarding calibration data for INT8 quantization by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/224


**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.7.13...1.7.14

1. Temporarily disabled accuracy check function added as a backend function. https://github.com/PINTO0309/onnx2tf/pull/184
- Significantly reduced RAM consumption during model conversion
- Significantly improved model conversion speed
2. Backward to v1.13.1 because I noticed a bug in `sigmoid` behavior in onnxruntime v1.14.0. After updating to v1.14.0, all `sigmoid` output values are Nan.
- https://github.com/microsoft/onnxruntime/pulls?q=sigmoid

- https://github.com/microsoft/onnxruntime/issues/14885
- [[TODO] Add option to disable backend inference for accuracy checks. 208](https://github.com/PINTO0309/onnx2tf/issues/208)

What's Changed
* 1. Temporarily disabled accuracy check function added as a backend function. 2. Backward to v1.13.1 because I noticed a bug in `sigmoid` behavior in onnxruntime v1.14.0 by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/220


**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.7.12...1.7.13

- `Split`
- Disabling splits that do not make sense and preventing the generation of `FlexSplitV` were added.
- [ssdlite320_mobilenet_v3_large.onnx.zip](https://github.com/PINTO0309/onnx2tf/files/10866464/ssdlite320_mobilenet_v3_large.onnx.zip)
- `Split` that makes no sense
- `Split` converted to `FlexSplitV`.
- [Cannot reshape a tensor with 41835024 elements to shape [3234,4] (12936 elements) 217](https://github.com/PINTO0309/onnx2tf/issues/217)

What's Changed
* Disabling splits that do not make sense and preventing the generation of `FlexSplitV` were added. by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/219


**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.7.11...1.7.12