Data To Text Generation On Dart
評価指標
BLEU
BLEURT
METEOR
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | BLEU | BLEURT | METEOR | Paper Title | Repository |
|---|---|---|---|---|---|
| T5-B Baseline | 48.47 | 67.49 | 40.74 | FactSpotter: Evaluating the Factual Faithfulness of Graph-to-Text Generation | |
| self-mem + new data | 47.76 | - | - | Self-training from Self-memory in Data-to-text Generation |
0 of 2 row(s) selected.