Data-to-text generation on E2E
0.717ROUGE-LPrefix
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PrefixBackbone=GPT-2LARGE2021.01 | 0.717 | 0.703 | 8.85 | 0.462 | 2.47 | |
| GPT-2LARGEEvaluation Protocol=Prefix (0.1%)2021.07 | 0.717 | 0.703 | 8.85 | 0.462 | 2.47 | |
| PREFIX (0.1%)Backbone=GPT-2MEDIUM2021.01 | 0.714 | 0.697 | 8.81 | 0.461 | 2.49 | |
| GPT-2MEDIUMEvaluation Protocol=Prefix (0.1%)2021.07 | 0.714 | 0.697 | 8.81 | 0.461 | 2.49 | |
| ADAPTER (3%)Backbone=GPT-2MEDIUM2021.01 | 0.713 | 0.689 | 8.71 | 0.461 | 2.47 | |
| HTLMEvaluation Protocol=Prefix (0.1%)2021.07 | 0.712 | 0.701 | 8.85 | 0.461 | 2.45 | |
| FINE-TUNEBackbone=GPT-2MEDIUM2021.01 | 0.71 | 0.682 | 8.62 | 0.462 | 2.47 | |
| GPT-2MEDIUMEvaluation Protocol=Fine-tuning2021.07 | 0.71 | 0.682 | 8.62 | 0.462 | 2.47 | |
| FT-TOP2Backbone=GPT-2MEDIUM2021.01 | 0.708 | 0.681 | 8.59 | 0.46 | 2.41 | |
| SOTA2021.01 | 0.708 | 0.686 | 8.7 | 0.453 | 2.37 | |
| HTLMEvaluation Protocol=Fine-tuning2021.07 | 0.708 | 0.703 | 8.9 | 0.463 | 2.47 | |
| FINE-TUNEBackbone=GPT-2LARGE2021.01 | 0.699 | 0.685 | 8.78 | 0.46 | 2.45 | |
| GPT-2LARGEEvaluation Protocol=Fine-tuning2021.07 | 0.699 | 0.685 | 8.78 | 0.46 | 2.45 | |
| ADAPTER (0.1%)Backbone=GPT-2MEDIUM2021.01 | 0.698 | 0.663 | 8.41 | 0.45 | 2.4 | |
| EDA_CSdelexicalization=false2019.04 | 0.6894 | 0.6705 | 8.515 | 0.4449 | 2.2355 | |
| TGendelexicalization=true2019.04 | 0.685 | 0.6593 | 8.6094 | 0.4483 | 2.2338 | |
| EDA_CSTLtransfer_learning=true2019.04 | 0.674 | 0.658 | 8.5615 | 0.4516 | 2.1803 | |
| EDA2019.04 | 0.5634 | 0.4999 | 7.1146 | 0.3369 | 1.3176 | |
| GENICLBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.551 | — | — | — | — | |
| LLM-RBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.544 | — | — | — | — | |
| BM25Backbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.538 | — | — | — | — | |
| EPRBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.536 | — | — | — | — | |
| Se²2024.02 | 0.534 | — | — | — | — | |
| E5baseBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.523 | — | — | — | — | |
| UPRISE2024.02 | 0.519 | — | — | — | — | |
| RandomBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.511 | — | — | — | — | |
| CBDSBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.508 | — | — | — | — | |
| SBERTBackbone=LLaMA-7B, Evaluation Setting=Few-shot (ICL)2025.05 | 0.506 | — | — | — | — | |
| Best-of-102024.02 | 0.487 | — | — | — | — | |
| BM252024.02 | 0.479 | — | — | — | — | |
| Instructor2024.02 | 0.433 | — | — | — | — | |
| Random2024.02 | 0.431 | — | — | — | — | |
| SBERT2024.02 | 0.426 | — | — | — | — | |
| Zero-shotBackbone=LLaMA-7B, Evaluation Setting=Zero-shot2025.05 | 0.346 | — | — | — | — | |
| HTLMEvaluation Protocol=One-Shot2021.07 | 0.316 | 0.321 | 3.35 | 0.241 | 0.78 | |
| Zero-shot2024.02 | 0.076 | — | — | — | — |