Chin Huang comments

Results 89 comments of


                                            Chin Huang

pytoch->onnx->tensorflow how to conevrt from NCHW(onnx) to NHWC(tensorflow lite)

The data format is NCHW in ONNX. We can convert a model to either NCHW or NHWC in Tensorflow. Please make sure the device is set to "CPU", by passing...

pytoch->onnx->tensorflow how to conevrt from NCHW(onnx) to NHWC(tensorflow lite)

In that case, try --device CUDA. The model will have NCHW wherever possible.

pytoch->onnx->tensorflow how to conevrt from NCHW(onnx) to NHWC(tensorflow lite)

We have tests with all ONNX model zoo models. Please be more specific with your issue. Is the conversion not working? or the inference not working, throwing exceptions? or the...

pytoch->onnx->tensorflow how to conevrt from NCHW(onnx) to NHWC(tensorflow lite)

Currently ONNX supports NCHW only. That means the model and node inputs must be in NCHW so the operators can work according to the specs. In order to support NHWC,...

ONNX to TF conversion is removing information such as the name of operators and input/output nodes

Sorry for the late response. The exported TF saved model has the output names as you see by adding a couple lines: ``` m = tf.saved_model.load(out_dir) print('loaded model outputs =...

onnx2tf.py problem

The error seems coming from custom source code, not from this converter. Please provide complete onnx2tf.py and the onnx file if you want to get more help.

Uninitialized value error when converting tensorflow checkpoint file into a onnx model

This looks like an issue for tf2onnx. Please report it here, https://github.com/onnx/tensorflow-onnx/issues

iterating over `tf.Tensor` is not allowed: AutoGraph did convert this function. This might indicate you are trying to use an unsupported feature.

Please share FasterRCNN.onnx file for debugging.

'_UserObject' object has no attribute 'summary'

Please note the saved_model is persisted using the low-level API. So you need to use tf.saved_model.load(path_to_dir) to load, rather than the high level tf.keras.models.load_model('saved_model')

'_UserObject' object has no attribute 'summary'

I just downloaded your onnx file and ran the following with no errors. ``` import onnx_tf import onnx import tensorflow as tf onnx_path = 'digits.onnx' model_path = 'digits_savedmodel' model =...