"Basic use", a working example config for continuing pretraining on existing modality ?

[luotsi]

Hello Clay community!

In the Getting Started document under "Basic use", there is a LightningCLI command line
python trainer.py fit --model ClayMAEModule --data ClayDataModule --config configs/config.yaml --trainer.fast_dev_run=True
Am I right that this seems to require some matching configuration not available in the main branch -- running the line as is gives:

error: Parser key "model":
Unable to load config 'ClayMAEModule'
  - Parser key "model": Unable to load config "ClayMAEModule"

Q1) Is there a matching config and sample dataset available that the first basic-use line would work on, for pretraining the encoder ?
Q2) Once that is running, I would actually like to try adding a whole new modality (image type) not previously supported in metadata.yaml . In addition to describing the image format in that file, what would I need to configure and implement, rough steps ?