Adam Pocock
Adam Pocock
I exported a logistic regression from scikit-learn using `onx = to_onnx(lr,x.astype(numpy.float32),target_opset=12,options={type(model): {'zipmap': False}})` and when inspecting that model in Java I see that the output is a tensor: ``` jshell>...
You can email the model to me at [email protected], but I'm out on vacation next week and won't look at it until the week after.
Could you test out this branch - https://github.com/Craigacp/onnxruntime/tree/java-sequence-fix. I've modified `OnnxSequence.getValue()` so it now returns either `List` or `List` depending on the sequence element type. It passes the ported C#...
I found a little more time and moved `OnnxSequence` over so that `getValue` now returns `List
Looks like the training pipeline failure was due to a CI VM error, could someone re-run it?
@fs-eire please could you kick off the orttraining-linux-ci-pipeline? The failure is unrelated to this PR, it looks like the VM filled its disk up. The JNI changes have all been...
I think I made all the requested changes here. Can @yuslepukin or @fs-eire review it again?
It does not copy on input if the buffers are direct. If they aren't direct then they need to be copied into direct ones, which are passed in to the...
The direct byte buffer abstraction in Java is different from the native pointer abstraction in C#, and I think that there is not a good solution here for outputs that...
I fixed the build error. Clang must do different exhaustiveness checking on enums than gcc or msvc as the latter two complained I wasn't assigning a value in a switch...