Chin Huang

The sys_config.device is to indicate the target runtime device. Some tensorflow ops, such as https://www.tensorflow.org/api_docs/python/tf/nn/convolution, have data_format argument and process differently based on the format. We cannot leave it as...

The current onnx-tf is to convert ONNX to Tensorflow SavedModel, obviously not targeted or optimized for TFLite. So maybe a TFLite optimizer can be developed as a new feature.

@ZHR1997 can you provide onnx file for debugging purpose?

Put it on a shared place like google drive. Please make sure it is publicly accessible. Thanks.

Thanks for bringing up the performance concerns. We are short in resources, so contributions are welcome!

Yes, please set up a github account. Thanks in advance!

There is no mapping for coordinate_transformation_mode=pytorch_half_pixel therefore a known limitation in converting Resize operator.

A couple of updates: 1. TF 1.15 is no longer supported by onnx-tf 2. We need the code that leads to the error message and the onnx file to debug

Sorry for the late reply. Took some time to get a proper env to try it out. The results are quite different for yours though. The calculation error doesn't seem...

Please try 1.5.0 which might work and 1.4.0 which should work with tf 1.14. Currently onnx-tf follows ONNX versions. Therefore, it doesn't have a release for each Tensorflow version that...