XNLI
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Natural Language Inference | XNLI (test) | Average Accuracy90 | 167 | |
| Natural Language Inference | XNLI | Accuracy87.1 | 131 | |
| Zero-Shot Cross-Lingual Transfer | XNLI | Pearson Correlation0.9639 | 48 | |
| Natural Language Inference | XNLI 1.0 (test) | Accuracy (en)89.7 | 40 | |
| Natural Language Inference | XNLI Ur (test) | Accuracy0.9643 | 26 | |
| Natural Language Inference | XNLI Ur (dev) | Accuracy70.6 | 26 | |
| Natural Language Inference | XNLI Hi (dev) | Accuracy76.91 | 26 | |
| Natural Language Inference | XNLI (dev) | Accuracy82.7 | 24 | |
| Text Classification | XNLI (test) | Accuracy (Max)81.3 | 20 | |
| Natural Language Inference | XNLI French | Accuracy59.1 | 18 | |
| Natural Language Inference | XNLI Sw (test) | Accuracy65.34 | 18 | |
| Natural Language Inference | XNLI Sw (dev) | Accuracy65.68 | 18 | |
| Natural Language Inference | XNLI Hi (test) | Accuracy71.65 | 18 | |
| Zero-shot performance prediction | XNLI | MAE1.53 | 18 | |
| Natural Language Inference | XNLI French (test) | Accuracy85.7 | 16 | |
| Natural Language Inference | XNLI 2.0 | Accuracy45.21 | 15 | |
| Sentence-pair classification | XNLI 1.1 (test) | Accuracy (EN)67.97 | 14 | |
| Cross-lingual Natural Language Inference | XNLI (test) | Accuracy (All)74.3 | 10 | |
| Natural Language Inference | XNLI Arabic | Accuracy (Normalized)33.73 | 10 | |
| Sentence Pair Classification | XNLI Chinese portion (test) | Accuracy81.3 | 9 | |
| Sentence Pair Classification | XNLI Chinese portion (dev) | Accuracy82.4 | 9 | |
| Sequence Classification | XNLI European Languages | Accuracy86.8 | 8 | |
| Sequence Classification | XNLI All Languages | Accuracy84.1 | 8 | |
| Natural Language Inference | XNLI Hindi (test) | Accuracy97.31 | 8 | |
| Natural Language Inference | XNLI Arabic (test) | Accuracy100 | 8 |
Showing 25 of 39 rows