nemo_rl.data.datasets.response_datasets.helpsteer3#

Module Contents#

Classes#

HelpSteer3Dataset

HelpSteer3 preference dataset for DPO training.

Functions#

API#

nemo_rl.data.datasets.response_datasets.helpsteer3.to_response_data_format(
data: dict[str, Any],
task_name: str = 'HelpSteer3',
) dict#
class nemo_rl.data.datasets.response_datasets.helpsteer3.HelpSteer3Dataset#

Bases: nemo_rl.data.datasets.raw_dataset.RawDataset

HelpSteer3 preference dataset for DPO training.

Initialization