Question Answering on NaturalQuestions filtered (dev)
44.3EMLlama 2
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama 2Size=70B, Shots=64-shot2023.07 | 44.3 | |
| Llama 1Size=65B, Shots=64-shot2023.07 | 39.9 | |
| Llama 2Size=34B, Shots=64-shot2023.07 | 39.9 | |
| Llama 2Size=70B, Shots=5-shot2023.07 | 39.5 | |
| Llama 1Size=33B, Shots=64-shot2023.07 | 36 | |
| FalconSize=40B, Shots=64-shot2023.07 | 35.5 | |
| Llama 1Size=65B, Shots=5-shot2023.07 | 35 | |
| Llama 2Size=13B, Shots=64-shot2023.07 | 34.6 | |
| FalconSize=40B, Shots=5-shot2023.07 | 33.5 | |
| Llama 2Size=70B, Shots=1-shot2023.07 | 33 | |
| Llama 1Size=33B, Shots=5-shot2023.07 | 32.9 | |
| Llama 2Size=34B, Shots=5-shot2023.07 | 32.8 | |
| Llama 1Size=13B, Shots=64-shot2023.07 | 31.9 | |
| Llama 2Size=13B, Shots=5-shot2023.07 | 31.2 | |
| Llama 1Size=65B, Shots=1-shot2023.07 | 31 | |
| Llama 2Size=34B, Shots=1-shot2023.07 | 30 | |
| FalconSize=40B, Shots=1-shot2023.07 | 29.5 | |
| Llama 2Size=7B, Shots=64-shot2023.07 | 29.5 | |
| MPTSize=30B, Shots=64-shot2023.07 | 29.3 | |
| Llama 1Size=33B, Shots=1-shot2023.07 | 28.3 | |
| Llama 1Size=13B, Shots=5-shot2023.07 | 28.1 | |
| Llama 2Size=13B, Shots=1-shot2023.07 | 28 | |
| MPTSize=30B, Shots=5-shot2023.07 | 26.6 | |
| FalconSize=40B, Shots=0-shot2023.07 | 26.3 | |
| Llama 1Size=7B, Shots=64-shot2023.07 | 26.1 | |
| Llama 2Size=7B, Shots=5-shot2023.07 | 25.7 | |
| Llama 2Size=70B, Shots=0-shot2023.07 | 25.3 | |
| Llama 2Size=34B, Shots=0-shot2023.07 | 25.1 | |
| Llama 1Size=33B, Shots=0-shot2023.07 | 24.9 | |
| FalconSize=7B, Shots=64-shot2023.07 | 24 | |
| Llama 1Size=65B, Shots=0-shot2023.07 | 23.8 | |
| Llama 1Size=13B, Shots=1-shot2023.07 | 23.4 | |
| MPTSize=30B, Shots=1-shot2023.07 | 23 | |
| MPTSize=7B, Shots=64-shot2023.07 | 22.7 | |
| Llama 2Size=7B, Shots=1-shot2023.07 | 22.7 | |
| Llama 1Size=7B, Shots=5-shot2023.07 | 22 | |
| FalconSize=7B, Shots=5-shot2023.07 | 21 | |
| MPTSize=7B, Shots=5-shot2023.07 | 20.8 | |
| Llama 1Size=13B, Shots=0-shot2023.07 | 20.1 | |
| Llama 1Size=7B, Shots=1-shot2023.07 | 18.7 | |
| FalconSize=7B, Shots=1-shot2023.07 | 18.1 | |
| MPTSize=7B, Shots=1-shot2023.07 | 17.8 | |
| Llama 1Size=7B, Shots=0-shot2023.07 | 16.8 | |
| Llama 2Size=7B, Shots=0-shot2023.07 | 16.4 | |
| Llama 2Size=13B, Shots=0-shot2023.07 | 16.1 | |
| MPTSize=30B, Shots=0-shot2023.07 | 15.8 | |
| FalconSize=7B, Shots=0-shot2023.07 | 15.7 | |
| MPTSize=7B, Shots=0-shot2023.07 | 11.6 |