Summarize the question. (e.g., "How to do inference on custom dataset ? ")

### Checklist

- [X] I have searched for [similar issues](https://github.com/isl-org/Open3D-ML/issues).
- [X] I have tested with the [latest development wheel](http://www.open3d.org/docs/latest/getting_started.html#development-version-pip).
- [X] I have checked the [release documentation](http://www.open3d.org/docs/release/) and the [latest documentation](http://www.open3d.org/docs/latest/) (for `main` branch).

### My Question

Hello. When running test in semantickitti dataset, in the test folder we get the predictions. If you get these predictions in each sequence folder of the test set as labels folder and then run test again, you will also get scores of the predictions. That at least happens in semantickitti and toronto3d datasets. But when used the same for parislille3d dataset it gives the following error:

Traceback (most recent call last):
File "/home/PycharmProjects/pythonProject1/test4.py", line 38, in
pipeline.run_test()
File "/home/anaconda3/envs/open3d/lib/python3.9/site-packages/open3d/_ml3d/torch/pipelines/semantic_segmentation.py",
line 259, in run_test
f"Overall Testing Accuracy : {self.metric_test.acc()[-1]}, mIoU : {self.metric_test.iou()[-1]}"
TypeError: 'NoneType' object is not subscriptable
test 0/1: 100%|██████████| 2901742/2901742 [02:22<00:00, 20357.49it/s]