microsoft / nn-Meter

A DNN inference latency prediction toolkit for accurately modeling and predicting the latency on diverse edge devices.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

How to generate material/testmodels/mobilenetv3small_0.pb

AIxyz opened this issue · comments

commented

Hi,
I would like to generate a TF2 frozen pb model (such as material/testmodels/mobilenetv3small_0.pb), but my models(generated by TF 2.6.0 and TF 2.7.0) can't be converted to nnm-ir. I find dataset/generator/generate_model.py is used to generate the keras h5 models. Is it possible to release the reference code to generate TF2 frozen pb model? And which TF version was used to generate material/testmodels/mobilenetv3small_0.pb?

Thanks & Regards,
X. Zhang

Hi Zhang,

The frozen pb model such as material/testmodels/mobilenetv3small_0.pb were generated by tensorflow==1.12. We didn't have a well test of TF2 frozen pb model. Maybe you can try this link: https://leimao.github.io/blog/Save-Load-Inference-From-TF2-Frozen-Graph/
In dataset/generator/generate_model.py we use the keras h5 model as the final format, because we plan to release a converter from keras h5 model to nn-meter ir. This tool is blocking in one of another of our work, and I will ask my colleague if this tool is ready these days.
Sorry for the inconvenience, and I will give new comment after I test the TF2 frozen pb conversion.

Best regards,
Jiahang

Hi, I just wanted to follow up any update on TF2 frozen pb conversion?