gptkbp:instanceOf
|
large language model
|
gptkbp:architecture
|
gptkb:transformation
|
gptkbp:author
|
gptkb:Myle_Ott
gptkb:Veselin_Stoyanov
gptkb:Francisco_Guzmán
gptkb:Guillaume_Wenzek
gptkb:Kartikay_Khandelwal
gptkb:Vishrav_Chaudhary
gptkb:Alexis_Conneau
gptkb:Luke_Zettlemoyer
gptkb:Edouard_Grave
gptkb:Naman_Goyal
|
gptkbp:basedOn
|
gptkb:RoBERTa
|
gptkbp:citation
|
2019
|
gptkbp:developedBy
|
gptkb:Facebook_AI
|
gptkbp:fullName
|
gptkb:XLM-RoBERTa
|
gptkbp:github
|
https://github.com/facebookresearch/XLM
|
https://www.w3.org/2000/01/rdf-schema#label
|
XLM-R
|
gptkbp:improves
|
gptkb:mBERT
gptkb:XLM
|
gptkbp:input
|
gptkb:text
|
gptkbp:inputSequenceLength
|
512 tokens
|
gptkbp:language
|
100
|
gptkbp:license
|
gptkb:MIT_License
|
gptkbp:notablePublication
|
gptkb:Unsupervised_Cross-lingual_Representation_Learning_at_Scale
|
gptkbp:openSource
|
true
|
gptkbp:parameter
|
550 million
|
gptkbp:releaseYear
|
2019
|
gptkbp:supports
|
transfer learning
question answering
text classification
zero-shot learning
named entity recognition
sequence labeling
sentence classification
|
gptkbp:tokenizerType
|
gptkb:SentencePiece
|
gptkbp:trainer
|
gptkb:CommonCrawl
gptkb:Wikipedia
2.5TB of filtered CommonCrawl data
|
gptkbp:type
|
multilingual model
|
gptkbp:usedFor
|
natural language processing
cross-lingual understanding
|
gptkbp:usedIn
|
gptkb:Hugging_Face_Transformers
|
gptkbp:bfsParent
|
gptkb:Meta_AI
|
gptkbp:bfsLayer
|
5
|