LLaVA-NeXT
Benchmarks
TaskDatasetSOTA resultTrendResults
LLaVA-NeXT Evaluation Suite (GQA, SQA-IMG, VQA-Text, POPE, MME, MMB-EN, MMB-CN, MMVet) 13B (test)
64.4GQA Score
28
LLaVA-NeXT-7B Evaluation Suite
5.6FLOPs (TFLOPs)
6
LLaVA-NeXT 8B
18.1Peak VRAM (GB)
6
LLaVA-NeXT Evaluation Suite
100Average Accuracy
6
LLaVA-NeXT textual concepts
81Detection Score
3