Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

onnc-umbrella bulit onnc reports "unsupported operator" when parsing lenet.onnx #6

Open
jinchenglee opened this issue Feb 25, 2020 · 3 comments

Comments

@jinchenglee
Copy link

Built both release/debug version of onnc from onnc-umbrella, it complains:

> onnc -mquadruple nvdla ./lenet.onnx
NvDla is invoked
Fatal: Meet unsupported operator: import/conv1first/BiasAdd.```

Is this expected? 

When running the docker image tutorial (https://github.com/ONNC/onnc-tutorial/blob/master/lab_1_Environment_Setup/lab_1.md), the .onnx to .nvdla conversion is completed successfully. Seems like a onnc version differences caused issue. 
@jinchenglee
Copy link
Author

Or shall I move this issue to onnc repo?

@a127a127
Copy link

a127a127 commented Feb 26, 2020

According to your error message, I think it's due to the ONNX model you used. The error message shows there is an unsupported operator.

Different tools (ML frameworks -> ONNX) use different ONNX operator set, maybe you can try other tools. LeNet is a simple & small model, NVDLA backend supported operator should be enough.

Or you can just wait. Current open source version NVDLA backend only supports about 15 ONNX operators ( https://github.com/ONNC/onnc/blob/master/Operator.md ). I think it will support more operators in the next version.

@jinchenglee
Copy link
Author

Thanks for prompt reply. I'll just wait for next release.

BTW, just for your inference, I used the lenet.onnx provided in your lab 1. So apparently the docker version of onnc is way ahead of the git repository one. :P

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants