|
gptkbp:instanceOf
|
large language model
|
|
gptkbp:architecture
|
gptkb:transformation
|
|
gptkbp:attentionMechanism
|
self-attention
|
|
gptkbp:author
|
gptkb:Jacob_Devlin
gptkb:Kenton_Lee
gptkb:Ming-Wei_Chang
gptkb:Kristina_Toutanova
|
|
gptkbp:citation
|
over 50,000 (as of 2024)
|
|
gptkbp:developedBy
|
gptkb:Google
|
|
gptkbp:fineTuning
|
task-specific
|
|
gptkbp:fullName
|
gptkb:Bidirectional_Encoder_Representations_from_Transformers
|
|
gptkbp:github
|
https://github.com/google-research/bert
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Google BERT
|
|
gptkbp:impact
|
state-of-the-art on GLUE
state-of-the-art on SQuAD
|
|
gptkbp:influenced
|
gptkb:ERNIE
gptkb:ALBERT
gptkb:DistilBERT
gptkb:RoBERTa
gptkb:XLNet
|
|
gptkbp:input
|
gptkb:text
|
|
gptkbp:inputLengthLimit
|
512 tokens
|
|
gptkbp:language
|
English
|
|
gptkbp:license
|
Apache 2.0
|
|
gptkbp:notablePublication
|
gptkb:BERT:_Pre-training_of_Deep_Bidirectional_Transformers_for_Language_Understanding
|
|
gptkbp:openSource
|
true
|
|
gptkbp:output
|
contextualized word embeddings
|
|
gptkbp:parameter
|
110 million (BERT Base)
340 million (BERT Large)
|
|
gptkbp:predecessor
|
gptkb:OpenAI_GPT
gptkb:ELMo
|
|
gptkbp:pretrainingTasks
|
masked language modeling
next sentence prediction
|
|
gptkbp:publishedIn
|
gptkb:NAACL_2019
|
|
gptkbp:releaseDate
|
2018
|
|
gptkbp:successor
|
gptkb:ALBERT
gptkb:DistilBERT
gptkb:RoBERTa
|
|
gptkbp:supports
|
transfer learning
multi-task learning
|
|
gptkbp:tokenizerType
|
gptkb:WordPiece
|
|
gptkbp:trainer
|
gptkb:Wikipedia
gptkb:BookCorpus
|
|
gptkbp:usedFor
|
question answering
sentiment analysis
text classification
named entity recognition
language understanding
|
|
gptkbp:bfsParent
|
gptkb:Retrieval-Augmented_Generation_(RAG)
|
|
gptkbp:bfsLayer
|
6
|