skim

Hello @quic-ssiddego, unfortunately, I got the same error when I replaced them to fused forms by using 'fused=True'. Otherwise, I should replace them with non-fused forms to skip the bias...

@quic-ssiddego yes, I am using tensorflow.keras.layers.BatchNormalization. Thank you for the reply. I will try it again once it is updated.

It was okay to import libpymo yesterday, but I have the same error today. I am using aimet 1.14.0 for tensorflow on colab.

I tried with ubuntu 18.04, but it still occurs this error: ModuleNotFoundError: No module named 'libpymo' In the environment variables, it includes the paths as below: ``` LD_LIBRARY_PATH=/usr/l00ocal/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64:/home/skim/pythonenv/lib/python3.6/site-packages/aimet_common/x86_64-linux-gnu:/home/skim/pythonenv/lib/python3 .6/site-packages/aimet_common:/home/skim/pythonenv/lib/: ```...

fixed the issue by adding PYTHONPATH in the environment variables in pycharm configuration