|
gptkbp:instanceOf
|
gptkb:Natural_language_processing_model
|
|
gptkbp:abbreviation
|
gptkb:ALBERT
|
|
gptkbp:architecture
|
gptkb:transformation
|
|
gptkbp:author
|
gptkb:Kevin_Gimpel
gptkb:Mingda_Chen
gptkb:Piyush_Sharma
gptkb:Radu_Soricut
gptkb:Sebastian_Goodman
gptkb:Zhenzhong_Lan
|
|
gptkbp:basedOn
|
gptkb:BERT
|
|
gptkbp:designedFor
|
language understanding
|
|
gptkbp:developedBy
|
gptkb:Toyota_Technological_Institute_at_Chicago
gptkb:Google_Research
|
|
gptkbp:improves
|
BERT (on some benchmarks)
|
|
gptkbp:introducedIn
|
2019
|
|
gptkbp:language
|
English
|
|
gptkbp:notableFeature
|
memory efficiency
factorized embedding parameterization
improved performance with fewer parameters
parameter sharing
|
|
gptkbp:notablePublication
|
gptkb:ALBERT:_A_Lite_BERT_for_Self-supervised_Learning_of_Language_Representations
|
|
gptkbp:openSource
|
Yes
|
|
gptkbp:pretrainedModelAvailable
|
Yes
|
|
gptkbp:publishedIn
|
gptkb:arXiv
|
|
gptkbp:usedFor
|
question answering
textual entailment
natural language inference
sentence classification
|
|
gptkbp:bfsParent
|
gptkb:ALBERT
|
|
gptkbp:bfsLayer
|
7
|
|
http://www.w3.org/2000/01/rdf-schema#label
|
A Lite BERT
|