AttributeError while running SemanticSegmentation training

### 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 `master` branch).

### Describe the issue

when I try to run the default code for semantic segmentation I get:

INFO - 2023-05-19 21:54:52,178 - semantic_segmentation - Initializing from scratch.
Traceback (most recent call last):
File "main.py", line 13, in
pipeline.run_train()
File ".local/lib/python3.8/site-packages/open3d/_ml3d/torch/pipelines/semantic_segmentation.py", line 354, in run_train
self.save_config(writer)
File "local/lib/python3.8/site-packages/open3d/_ml3d/torch/pipelines/semantic_segmentation.py", line 559, in save_config
writer.add_text("Description/Open3D-ML", self.cfg_tb['readme'], 0)
AttributeError: 'SemanticSegmentation' object has no attribute 'cfg_tb'

### Steps to reproduce the bug

```python
import os
import open3d.ml as _ml3d
import open3d.ml.torch as ml3d

dataset = ml3d.datasets.SemanticKITTI(dataset_path='data/kitti', use_cache=True)

# create the model with random initialization.
model = ml3d.models.RandLANet()

pipeline = ml3d.pipelines.SemanticSegmentation(model=model, dataset=dataset, max_epoch=100)

# prints training progress in the console.
pipeline.run_train()
```

### Error message

INFO - 2023-05-19 21:54:52,178 - semantic_segmentation - Initializing from scratch.
Traceback (most recent call last):
File "main.py", line 13, in
pipeline.run_train()
File ".local/lib/python3.8/site-packages/open3d/_ml3d/torch/pipelines/semantic_segmentation.py", line 354, in run_train
self.save_config(writer)
File "local/lib/python3.8/site-packages/open3d/_ml3d/torch/pipelines/semantic_segmentation.py", line 559, in save_config
writer.add_text("Description/Open3D-ML", self.cfg_tb['readme'], 0)
AttributeError: 'SemanticSegmentation' object has no attribute 'cfg_tb'

### Expected behavior

_No response_

### Open3D, Python and System information

```markdown
- Operating system: Ubuntu 20.04
- Python version: Python 3.8
- Open3D version: 0.12.0
- System type: x84
- Is this remote workstation?: yes
- How did you install Open3D?: pip
```

### Additional information

_No response_

Start with open3d/_ml3d/torch/pipelines/semantic_segmentation.py, especially run_train and save_config around the reported lines. Reproduce the default SemanticKITTI training example with the stated Open3D-ML version, then trace where cfg_tb is expected to be initialized. Done means training no longer raises this AttributeError and the relevant pipeline behavior is covered by an appropriate test.

Written by the indexing model from the issue text.