No logging output during training in Google Colab

### 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).

### My Question

Hi, all.

I have been running into a logging issue lately when training KPConv on Google Colab. When I use `pipeline.run_train()`, no output is generated except for the process bar, which is not very helpful nor informative, since there is no way to know how many epochs are left or how good is model's performance:


The same occurs if I try to train RandLaNet. I have also tried to set the global Open3D verbosity level higher, but no output is generated either.

I am using the following code:

```
import open3d as o3d
import open3d.ml as _ml3d
import open3d.ml.tf as ml3d
from ml3d.datasets.customdataset import Custom3D

cfg_file = "ml3d/configs/kpconv_custom.yml"
cfg = _ml3d.utils.Config.load_from_file(cfg_file)

dataset = Custom3D(**cfg.dataset)
model = ml3d.models.KPFCNN(**cfg.model)
pipeline = ml3d.pipelines.SemanticSegmentation(model=model, dataset=dataset, **cfg.pipeline)

pipeline.run_train()
```
You can find the config file in the following [link](https://github.com/borgarpa/Open3D-ML/blob/23cae17c5993d224689fff34d15e997d1c9d5ef4/ml3d/configs/kpconv_custom.yml).

Any help is appreciated.