depth-nerfacto broken

[nepfaff]

Describe the bug
I'm using the default depth-nerfacto command:

ns-train depth-nerfacto --output-dir ... --data ...

I'm getting:

nerfstudio/data/datamanagers/parallel_datamanager.py", line 157, in create_train_dataset
    cache_compressed_images=self.config.cache_compressed_images,
AttributeError: 'VanillaDataManagerConfig' object has no attribute 'cache_compressed_images'

To Reproduce
As above

Expected behavior
Parallel Datamanger should not be called as the depth dataset doesn't support it yet. I didn't specify --pipeline.datamanager.load-from-disk True and depth-nerfacto doesn't even support that option.

Additional context
This is on master.

[AntonioMacaronio]

I will address this in an upcoming PR!

In the meantime, a quick fix is that inside method_configs.py please change the datamanager of method_configs["depth-nerfacto"] to the ParallelDataManagerConfig