python train.py error

[leehaining]

anaconda3 + cuda 10.0 + tensorflow 1.14.0
This error occurs when I execute the train.py . I checked the GPU usage and found that the usage rate was 0. I also changed the values of batch_ size and eopch but do not work.

[leehaining]

When running to Model.get_model(pointclouds_pl, is_training_pl, bn_decay=bn_decay), the above error will appear.
If you also encounter this situation, please tell me the solution, thank you so much.