Alex Merritt
You may want to consider using the [JavaCPP Preset for ONNX](https://github.com/bytedeco/javacpp-presets/tree/master/onnx)
+1 You may find my project [ONNX-Scala](https://github.com/EmergentOrder/onnx-scala) useful here.
Ahh, I see MLflow has ONNX support now, however it seems to be limited: https://github.com/mlflow/mlflow/issues/3229
I was able to reproduce this issue, with the latest ONNX preset and a few versions of protobuf . Also confirmed I cannot reproduce the same from Python. I found...
You can serialize it to a string with `model.SerializeAsString()`, although it's not exactly readable.
Once you have the `bytes`, as per the example, try this: `com.google.protobuf.Any.parseFrom(bytes).toString()` Field names are lost, but should still be good enough to get an idea of the model structure.
@shimamoto can you please resolve the conflicts?
Looks like Travis is having some difficulties downloading dependencies...
Bumping this, this is a blocker to moving to 1.16.x for me.
Bumping this again. Still an issue as of 1.18.0