| Benchmark | Model | Rank | Results |
|---|---|---|---|
| semantic-textual-similarity-on-sts-b-glue-dev | XLM-R | #6 | Pearson/Spearman Correlation: 91.2 |
| paraphrase-identification-on-mrpc-glue-dev | XLM-R | #7 | F1: 89.5 |
| natural-language-inference-on-multinli-dev | XLM-R | #10 | Matched: 88.9 |
| natural-language-inference-on-qnli-glue-dev | XLM-R | #10 | Accuracy: 93.8 |
| sentiment-analysis-on-sst-2-binary-classification | XLM-R | #90 | Dev Accuracy: 95.0 |