cerebras.modelzoo.data.common.h5_map_dataset.dataset.MLMHDF5Dataset#

class cerebras.modelzoo.data.common.h5_map_dataset.dataset.MLMHDF5Dataset(*args, **kwargs)[source]#

Bases: cerebras.modelzoo.data.common.h5_map_dataset.dataset.HDF5Dataset

Dataset class to handle text preprocessing in bert mlm datasets.

Parameters: config – The config used to configure the dataset.

Methods

`generate_sample`
`load_state_dict`
`map`
`state_dict`

Attributes

by_sample

previous

cerebras.modelzoo.data.common.h5_map_dataset.dataset.HDF5DatasetConfig

next

cerebras.modelzoo.data.common.h5_map_dataset.dataset.MultiModalHDF5Dataset