Datasetpublic

gmongaras/BERT_Base_Cased_512_Dataset2

Dataset using the bert-cased tokenizer, cutoff sentences to 512 length (not sentence pairs), all sentence pairs extracted. Original datasets: https://huggingface.co/datasets/bookcorpus Original dataset: https://huggingface.co/datasets/wikipedia Variant: 20220301.en

sourceHugging Faceupdated 3y agoView on Hugging Face

0likes236downloads

Dataset Card

Dataset using the bert-cased tokenizer, cutoff sentences to 512 length (not sentence pairs), all sentence pairs extracted.

Original datasets: https://huggingface.co/datasets/bookcorpus Original dataset: https://huggingface.co/datasets/wikipedia Variant: 20220301.en