WikiText

GPTKB entity

Statements (19)
Predicate Object
gptkbp:instanceOf gptkb:dataset
gptkbp:contains Wikipedia articles
gptkbp:creator gptkb:Salesforce_Research
gptkbp:hasVersion gptkb:WikiText-103
gptkb:WikiText-2
https://www.w3.org/2000/01/rdf-schema#label WikiText
gptkbp:language English
gptkbp:license gptkb:Creative_Commons_Attribution-ShareAlike_3.0
gptkbp:releaseYear 2016
gptkbp:size WikiText-103: ~100 million tokens
WikiText-2: ~100,000 tokens
gptkbp:usedFor gptkb:machine_learning
natural language processing
language modeling
gptkbp:usedIn GPT-2 pretraining
language model benchmarks
gptkbp:website https://blog.einstein.ai/the-wikitext-long-term-dependency-language-modeling-dataset/
gptkbp:bfsParent gptkb:NLP
gptkbp:bfsLayer 6