cerebras.modelzoo.data_preparation.nlp.hdf5_preprocessing.utils.DatasetStats#

class cerebras.modelzoo.data_preparation.nlp.hdf5_preprocessing.utils.DatasetStats[source]#

Bases: object

DatasetStats(num_sequences: int, num_tokens: int, detokenized_bytes: int, detokenized_chars: int, non_pad_tokens: int, loss_valid_tokens: int)

Methods

Attributes

num_sequences

num_tokens

detokenized_bytes

detokenized_chars

non_pad_tokens

loss_valid_tokens

__init__(num_sequences: int, num_tokens: int, detokenized_bytes: int, detokenized_chars: int, non_pad_tokens: int, loss_valid_tokens: int) → None#

previous

cerebras.modelzoo.data_preparation.nlp.hdf5_preprocessing.utils.wikitext_detokenizer

next

cerebras.modelzoo.data_preparation.nlp.hdf5_preprocessing.utils.DocObject