Onnx Model Zoo repo error when validating models

[IzanCatalan]

Hi everyone!
I'm using Onnx Model Zoo GitHub repo (https://github.com/onnx/models). I tried to execute in Jupyter Notebook and replicate in a separate python script the "imagenet_validation.ipynb" file under models/vision/classification path.

However, after installing all dependencies I get some errors no mater the onnx model file (squeeznet or resnet) (see de images).

I don't understand why these errors, apparently is because MXnet loading method because always fail when the model is loaded. It does not recognize quantize models or these models are not parsing correctly after quantization. Either way, those errors are estrange, because those models are officials files in the official GitHub repo.

Maybe someone can help me or give me some guidelines to solve o at least, replicate these imaginet validation files.
Which is the setup to make them work? which frameworks or dependencies and versions of them should be installed?

Thank you,

Izan



.