Scott McKay

> > Is there doco somewhere on how/when this workflow runs? Not sure if it runs automatically or would it be part of the release process? > > It runs...

Was the change to cmake/external/onnx intentional?

We should also update the existing code to document the current behavior clearly.

Please provide the full code to reproduce and show how you are measuring performance. As you say you have two onnxruntime sessions it's not clear how/when you are creating those...

It would be best to measure the ORT performance separately with no thread pools, and without the inline call to GetInputNames(). That way you're just measuring the cost of the...

Is this an issue with onnx-tensorrt? If so https://github.com/onnx/onnx-tensorrt/issues may be a better place to ask the question.

It's important to note that there are two aspects here. One is whether the model inputs have fixed or dynamic sizes, and the other is whether the model itself supports...

The original model seemed to have a bi-directional LSTM but the conversion to onnx is a) not detecting that and using an ONNX LSTM node, and b) writing out an...

This particular model can be exported with 2 changes that @TomWildenhain-Microsoft kindly figured out. 1) Use the TF processing for TFL_While similar to TFL_IF by removing TflWhile.version_7 and adding a...

Also not clear how exactly you're measuring performance and whether external things are being included in the timing. - run the model once as a warmup. - this will initialize...