SMARTRoBERTa

Reported on 5 benchmarks across 3 tasks · 1 paper · 5 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing5 results

Natural Language InferenceonMultiNLI
Dev Matched· 2019-11-08
91.1
SOTA
SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization arXiv:1911.03437
Natural Language InferenceonMultiNLI
Dev Mismatched· 2019-11-08
91.3
SOTA
SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization arXiv:1911.03437
Semantic Textual SimilarityonSTS Benchmark
Dev Pearson Correlation· 2019-11-08
92.8
SOTA
SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization arXiv:1911.03437
Semantic Textual SimilarityonSTS Benchmark
Dev Spearman Correlation· 2019-11-08
92.6
SOTA
SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization arXiv:1911.03437
Sentiment AnalysisonSST-2 Binary classification
Dev Accuracy· 2019-11-08
96.9
SOTA
SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization arXiv:1911.03437