Statements (55)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
Transformer model
|
| gptkbp:basedOn |
BERT
|
| gptkbp:developedBy |
Hugging Face
|
| gptkbp:has |
66 million parameters
|
| https://www.w3.org/2000/01/rdf-schema#label |
DistilBERT
|
| gptkbp:isAvailableIn |
various domains
Hugging Face Model Hub |
| gptkbp:isAvenueFor |
BERT
distillation technique |
| gptkbp:isBeneficialFor |
low-resource languages
|
| gptkbp:isCompatibleWith |
Hugging Face Transformers library
|
| gptkbp:isDesignedFor |
natural language processing tasks
|
| gptkbp:isDiscussedIn |
online forums
|
| gptkbp:isDocumentedIn |
research papers
technical blogs |
| gptkbp:isEvaluatedBy |
gptkb:GLUE_benchmark
performance metrics efficiency accuracy robustness speed generalization ability SQuAD benchmark |
| gptkbp:isFocusedOn |
large text corpora
|
| gptkbp:isFoundIn |
BERT
|
| gptkbp:isIntegratedWith |
other_NLP_tools
|
| gptkbp:isOpenTo |
true
|
| gptkbp:isOptimizedFor |
faster inference
|
| gptkbp:isPartOf |
other transformer models
Hugging Face ecosystem NLP_models_family |
| gptkbp:isPopularIn |
research community
|
| gptkbp:isSupportedBy |
community contributions
|
| gptkbp:isTestedFor |
knowledge distillation
|
| gptkbp:isTrainedIn |
masked language modeling
|
| gptkbp:isUsedBy |
data scientists
developers machine learning practitioners |
| gptkbp:isUsedFor |
transfer learning
feature extraction |
| gptkbp:isUsedIn |
question answering
sentiment analysis text classification named entity recognition |
| gptkbp:isUtilizedFor |
specific tasks
|
| gptkbp:isUtilizedIn |
language translation
chatbots virtual assistants text summarization |
| gptkbp:maintains |
95% of BERT's performance
|
| gptkbp:reduces |
model size
|
| gptkbp:supports |
gptkb:PyTorch
TensorFlow |
| gptkbp:uses |
self-attention mechanism
|
| gptkbp:wasAffecting |
2019
|