Natural Language Understanding on LexGLUE
Metric: CaseHOLD (higher is better)
Metric: CaseHOLD (higher is better)
| # | Model↕ | CaseHOLD▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | CaseLaw-BERT | 75.6 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |
| 2 | Legal-BERT | 75.1 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |
| 3 | DeBERTa | 72.1 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |
| 4 | Longformer | 72 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |
| 5 | RoBERTa | 71.7 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |
| 6 | BERT | 70.7 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |
| 7 | BigBird | 70.4 | No | LexGLUE: A Benchmark Dataset for Legal Language ... | 2021-10-03 | Code |