DataConfig#

class stable_ssl.config.DataConfig(train_on, *args, **datasets)[source]#

Bases: object

Configuration for multiple datasets used for training the model.

Parameters:
  • train_on (str) – The dataset to train on.

  • datasets (dict[str, DatasetConfig]) – A dictionary of dataset configurations.

get_dataloaders()[source]#

Get dataloaders for the datasets.

Returns:

A dictionary containing dataloaders.

Return type:

dict

get_datasets()[source]#

Get datasets for training and validation.

Returns:

A dictionary containing datasets.

Return type:

dict