Data-to-Text Generation on E2E NLG Challenge
Metric: NIST (higher is better)
Metric: NIST (higher is better)
| # | Model↕ | NIST▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | S_1^R | 8.73 | No | Pragmatically Informative Text Generation | 2019-04-02 | Code |
| 2 | Slug | 8.613 | Yes | A Deep Ensemble Model with Slot Alignment for Se... | 2018-05-16 | - |
| 3 | TGen | 8.6094 | No | Findings of the E2E NLG Challenge | 2018-10-02 | Code |
| 4 | EDA_CS (TL) | 8.5615 | Yes | Copy mechanism and tailored training for charact... | 2019-04-26 | Code |
| 5 | EDA_CS | 8.515 | No | Copy mechanism and tailored training for charact... | 2019-04-26 | Code |
| 6 | Sys1-Primary | 8.5105 | Yes | - | - | - |
| 7 | Self-memory | 8.35 | No | Self-training from Self-memory in Data-to-text G... | 2024-01-19 | Code |
| 8 | Gong | 8.3453 | No | - | - | - |
| 9 | Zhang | 8.1804 | No | - | - | - |
| 10 | TUDA | 7.4544 | No | - | - | Code |