Ctrl+K

epbd_bert.datasets package

epbd_bert.datasets package#

Submodules#

epbd_bert.datasets.data_collators module#

class epbd_bert.datasets.data_collators.SeqLabelDataCollator(pad_token_id=0)[source]#

Bases: object

class epbd_bert.datasets.data_collators.SeqLabelEPBDDataCollator(pad_token_id=0)[source]#

Bases: object

epbd_bert.datasets.sequence_dataset module#

class epbd_bert.datasets.sequence_dataset.SequenceDataset(data_path: str, tokenizer: PreTrainedTokenizer, home_dir='')[source]#

Bases: Dataset

Dataset for supervised fine-tuning.

epbd_bert.datasets.sequence_epbd_dataset module#

class epbd_bert.datasets.sequence_epbd_dataset.SequenceEPBDDataset(data_path: str, pydnaepbd_features_path: str, tokenizer: PreTrainedTokenizer, home_dir='')[source]#

Bases: SequenceDataset

Supervised fine-tuning from seq and epbd features

epbd_bert.datasets.sequence_epbd_multimodal_dataset module#

class epbd_bert.datasets.sequence_epbd_multimodal_dataset.SequenceEPBDMultiModalDataset(data_path: str, pydnaepbd_features_path: str, tokenizer: PreTrainedTokenizer, home_dir='')[source]#

Bases: SequenceEPBDDataset

Dataset for multi-modal transformer

epbd_bert.datasets.sequence_epbd_multimodal_labelspecific_dataset module#

class epbd_bert.datasets.sequence_epbd_multimodal_labelspecific_dataset.SequenceEPBDMultiModalLabelSpecificDataset(data_path: str, pydnaepbd_features_path: str, tokenizer: PreTrainedTokenizer, label='wgEncodeAwgTfbsBroadDnd41CtcfUniPk', home_dir='')[source]#

Bases: SequenceEPBDMultiModalDataset

Dataset for multi-modal transformer

epbd_bert.datasets.sequence_randepbd_dataset module#

class epbd_bert.datasets.sequence_randepbd_dataset.SequenceRandEPBDDataset(data_path: str, pydnaepbd_features_path: str, tokenizer: PreTrainedTokenizer, home_dir='')[source]#

Bases: SequenceEPBDDataset

epbd_bert.datasets.sequence_randepbd_multimodal_dataset module#

class epbd_bert.datasets.sequence_randepbd_multimodal_dataset.SequenceRandEPBDMultiModalDataset(data_path: str, pydnaepbd_features_path: str, tokenizer: PreTrainedTokenizer, home_dir='')[source]#

Bases: SequenceEPBDDataset

Dataset for multi-modal transformer

Module contents#

Contents