Statements (53)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
benchmark
|
| gptkbp:composedOf |
multiple tasks
|
| gptkbp:designedFor |
natural language understanding
|
| gptkbp:developedBy |
AI researchers
|
| gptkbp:evaluates |
model performance
|
| gptkbp:hasCitations |
overfitting
generalization task diversity |
| gptkbp:hasVariants |
SuperGLUE
Multi-GLUE |
| https://www.w3.org/2000/01/rdf-schema#label |
GLUE
|
| gptkbp:includes |
gptkb:SST-2
CoLA MNLI WNLI RTE QQP STS-B |
| gptkbp:isBasedOn |
BERT
RoBERTa XLNet |
| gptkbp:isChallengedBy |
question answering
sentiment analysis information retrieval transfer learning multi-task learning language modeling text summarization text classification named entity recognition semantic similarity domain adaptation few-shot learning zero-shot learning coreference resolution dialogue systems language inference paraphrase detection text coherence text entailment |
| gptkbp:isEvaluatedBy |
F1 score
accuracy Matthew's_correlation_coefficient |
| gptkbp:isPopularIn |
machine learning community
|
| gptkbp:isRelatedTo |
fine-tuning
language models NLP_tasks |
| gptkbp:isSupportedBy |
gptkb:PyTorch
TensorFlow Hugging Face Transformers |
| gptkbp:isUsedBy |
NLP_models
|
| gptkbp:provides |
score metrics
|
| gptkbp:releasedIn |
2018
|