Exclusion list for datasets

[wiederm]

Feature Request: Exclude Data Points via exclusion Tag in dataset.toml

It would be helpful to have a way to exclude specific data points from the dataset by specifying them in an exclusion tag within the dataset.toml file. This tag would define a list of entries to skip when loading the dataset.

Proposed Implementation:
To avoid affecting the dataset split, this exclusion process could be implemented immediately after the dataset split step. This approach would allow the split to remain unaffected by the excluded entries while still filtering out unwanted data points. In that way the exclusion happens during the dataset setup phase and would have no impact on the dataset caching, which would be benefictial.