LLaVA-Bench
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
LLaVA-Bench
91.9Overall Score
94
LLaVA-Bench Wild
91.2LLaVA^W Score
86
LLaVA-Bench Wild
102Score
78
LLaVA-bench in-the-wild
121.88Score
73
LLaVA-Bench Wild
102.3Score
71
LLaVA-Bench
79.2LLaVA-Bench Score
48
LLaVA Bench
93.1LLaVA Bench Score
46
LLaVA Bench (val)
2.188Perplexity
44
LLaVA-Bench-Wild (LLaVA-W)
97.3Overall Score
43
LLaVA-Bench-In-The-Wild
87.1Score
38
LLaVA-Bench In-the-Wild
93.1Score
35
LLaVA-Bench In-the-Wild
67.7Score
22
LLaVA-Bench
84.3GPT-4 Score
21
LLaVA-Bench Wild (test)
97.7Score
19
LLaVA-Bench Wild
74.5GPT-4 Score
19
LLaVA-Bench
4.38Accuracy
18
LLaVA-Bench
79.1Overall Score
15
LLaVA-Bench Wilder
74.3Score
14
LLaVA-Bench In-the-Wild
109.3Score
14
LLaVA-Bench Coco
92.3Score
13
LLaVA-Bench
65.5LLaVA-B Score
12
LLaVA Bench
68.31VQA ASR
12
LLaVA-Bench Wild
92.8Relative Score
12
LLaVA-Bench In-the-Wild
78.9General Score
12
LLaVA-Bench COCO
85.76Reference Score
11