SqueezeBERT: What can computer vision teach NLP about efficient neural networks?

Benchmark Model Rank Results
question-answering-on-quora-question-pairsSqueezeBERT#17Accuracy: 80.3%
semantic-textual-similarity-on-mrpcSqueezeBERT#18Accuracy: 87.8%
natural-language-inference-on-wnliSqueezeBERT#21Accuracy: 65.1
linguistic-acceptability-on-colaSqueezeBERT#35Accuracy: 46.5%
natural-language-inference-on-multinliSqueezeBERT#35Matched: 82.0Mismatched: 81.1
natural-language-inference-on-qnliSqueezeBERT#36Accuracy: 90.1%
natural-language-inference-on-rteSqueezeBERT#44Accuracy: 73.2%
sentiment-analysis-on-sst-2-binary-classificationSqueezeBERT#51Accuracy: 91.4