JDAI-CV/dabnn

birealnet18.onnx / birealnet18stem.onnx: how do you build them and how to run inferences with them

bcm-at-zama opened this issue · 0 comments

Hello guys,

I was very interesting by what you've been building, and wanted to ask you a bit more, if you've some time to direct me.

I wanted to know:

  • how you've been building the birealnet18.onnx and birealnet18stem.onnx you mention in your readme. I have seen ways to do .onnx to .dab, but it is mysterious to me how you've made these onnx's
  • how to launch the inference with these onnx, and check the accuracy. I have tried with onnxruntime and code like https://github.com/onnx/models/blob/master/vision/classification/imagenet_inference.ipynb , but was never able to get a significant accuracy

Thanks a lot for your help. Maybe just some pointers may be sufficient to unblock me, if you can. And congrats for your paper and tools.