Pytorch GPT2/3 Dataloader.
Classes
GptHDF5DataProcessor
A HDF5 dataset processor for GPT pre-training.
previous
cerebras.modelzoo.data.nlp.gpt.DummyIterableDataProcessor.DummyTinyIterableDataset
next
cerebras.modelzoo.data.nlp.gpt.GptHDF5DataProcessor.GptHDF5DataProcessor