← all datasets

BPE-tokenized OpenWebText

Emerging
1papers using it
2026first seen

'BPE-tokenized OpenWebText' is a dataset that contains text data processed using Byte Pair Encoding (BPE) for tokenization, and it is used to evaluate the performance of models on various language tasks.

🔎 Find this dataset
BPE-tokenized OpenWebText — datasets — generative-models