Conversational Multi-document Question Answering on Multi-doc QA specific
51.66ROUGE-1Oracle
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| OraclePredictor=LLAMA2-7B, Generator=LLAMA2-13B-chat2024.09 | 51.66 | 24.14 | 47.72 | |
| AdaCompPredictor=LLAMA2-7B, Generator=LLAMA2-13B-chat2024.09 | 46.89 | 20.56 | 43.19 | |
| Top-5Predictor=LLAMA2-7B, Generator=LLAMA2-13B-chat2024.09 | 44.94 | 18.84 | 41.29 | |
| Top-RandomPredictor=LLAMA2-7B, Generator=LLAMA2-13B-chat2024.09 | 44.89 | 18.97 | 41.24 | |
| Top-1Predictor=LLAMA2-7B, Generator=LLAMA2-13B-chat2024.09 | 43.22 | 18.05 | 39.71 | |
| RECOMPPredictor=LLAMA2-7B, Generator=LLAMA2-13B-chat2024.09 | 38.41 | 15.52 | 35.36 |