Error while trying to load BERTopic model in Sagemaker end point with PyTorch

[AshokNagula]

I am trying to make a Sage maker pipeline, successfully trained the model and created end point with PyTorch framework.
when i try to make inference from the model , i'm facing the following error which seems like internal package.
using bertopic -0.14 on both training and inference instances , python - 3.9.

any thoughts @MaartenGr ?

[MaartenGr]

I believe there are a number of closed issues mentioning this specific issue. Did you already try out the suggestions there? From what I remember, it is due to some dependency problems with numpy.

[AshokNagula]

yes , Its resolved with Numpy==1.23.3 ,PyTorch == 1.10 and Python==3.8. Thanks !