End-to-end Question Answering on Bird
20.6AccuracyARM
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ARMGeneration model=Llama3.1-8B-It2025.01 | 20.6 | 44 | 51.8 | |
| DR@5Generation model=Llama3.1-8B-It, top-k=52025.01 | 17.5 | 34.2 | 40.6 | |
| DRR@5Generation model=Llama3.1-8B-It, top-k=52025.01 | 16.8 | 39.8 | 46.9 | |
| DRR-D@5Generation model=Llama3.1-8B-It, top-k=5, query decomposition=true2025.01 | 15.9 | 38.8 | 46.3 | |
| DR-D@5Generation model=Llama3.1-8B-It, top-k=5, query decomposition=true2025.01 | 13.9 | 27.2 | 33.3 | |
| ReActGeneration model=Llama3.1-8B-It2025.01 | 4.7 | 27 | 32.5 |