Prediction Reasoning on MMBench (test)
66.5Average ScoreMerlin
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MerlinLLM Size=7B2023.11 | 66.5 | 41.3 | 51 | 83 | 97.6 | 59.7 | |
| ShikraLLM Size=7B2023.11 | 60 | 27.6 | 50 | 70.1 | 92.7 | 59.5 | |
| Kosmos-2LLM Size=1.6B2023.11 | 58.2 | 40.4 | 30 | 65.4 | 89 | 66.2 | |
| MiniGPT-4LLM Size=13B2023.11 | 48.9 | 21 | 35 | 67.3 | 90.2 | 31.1 | |
| mPLUG-OwlLLM Size=7B2023.11 | 45.9 | 16.2 | 23 | 59.8 | 91.5 | 39.2 | |
| InstructBLIPLLM Size=13B2023.11 | 44.4 | 5.7 | 24 | 67.3 | 92.7 | 32.4 | |
| LLAVALLM Size=7B2023.11 | 39.7 | 13.3 | 35 | 48.6 | 82.9 | 18.9 | |
| InstructBLIPLLM Size=7B2023.11 | 39 | 2.9 | 17 | 52.3 | 78 | 44.6 | |
| MiniGPT-4LLM Size=7B2023.11 | 27.9 | 8.6 | 13 | 29.9 | 61 | 27 | |
| MMGPTLLM Size=7B2023.11 | 16.8 | 3.8 | 13 | 12.1 | 52.4 | 2.7 | |
| OpenFlamingoLLM Size=7B2023.11 | 11.5 | 2.9 | 14 | 9.3 | 11 | 20.3 |