Learning Machine Learning with .NET, PyTorch and the ONNX Runtime

[@1989dawit]

where do get the onnx runtime demo ?

[@AdamsTaiwan]

Would like to see the training done using Julia and FluxML.

[@tikke8511]

Big onnx model deserialization may be expensive, and for stateless REST you do it for every call. What is the "best practise" to avoid that?

[@mandauefoama8845]

Reuse the InferenceSession by caching it. Cache it per ONNX file that you opened.

[@danielkrajnik3817]

cringe