Large Multi-modal Model Evaluation on LLaVA-Bench Tool Use (test)
0.893GroundingLLaVA-Plus
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| LLaVA-Plustool_usage=All Tools2023.11 | 0.893 | 0.944 | 0.967 | 0.488 | 0.823 | |
| LLaVA-Plustool_usage=Fly2023.11 | 0.886 | 0.889 | 0.902 | 0.384 | 0.765 | |
| All Tools + GPT4planner=GPT4, toolset=All Tools2023.11 | 0.775 | 0.956 | 0.952 | 0.393 | 0.769 | |
| Bing Chatversion=07302023.11 | 0.56 | 0.84 | 0.96 | 0.448 | 0.702 | |
| LLaVA2023.11 | 0.471 | 0.871 | 0.77 | 0.236 | 0.587 | |
| LLaVAevaluation_protocol=Tools in Test2023.11 | 0.417 | 0.485 | 0.72 | 0.319 | 0.485 | |
| Bardversion=07302023.11 | 0.365 | 1.053 | 1.033 | 0.6 | 0.763 | |
| MM-REACT2023.11 | 0.302 | 0.947 | 1.038 | 0.773 | 0.765 |