Chatbot Evaluation on Vicuna benchmark
13,481Elo RatingGPT-4
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-42023.05 | 13,481 | |
| Guanaco 65BSize=41 GB2023.05 | 10,221 | |
| Guanaco 33BSize=21 GB2023.05 | 9,921 | |
| Vicuna 13BSize=26 GB2023.05 | 9,741 | |
| ChatGPT2023.05 | 9,661 | |
| Guanaco 13BSize=10 GB2023.05 | 9,161 | |
| Bard2023.05 | 9,021 | |
| Guanaco 7BSize=6 GB2023.05 | 8,791 |