List your company
ResearchFormerly WizWand

Visual Question Answering on VQA v2 (Accuracy)

78.5AccuracyLLaVA-1.5-7B
54.0660.40566.7573.095Sep 29, 2025
Updated 8mo ago

Evaluation results

MethodLinks
LLaVA-1.5-7BRetained Tokens=5762025.0978.5
ZOO-PruneRetained Tokens=1922025.0977.34
DivPruneRetained Tokens=1922025.0976.87
VisionZipRetained Tokens=1922025.0976.8
ZOO-PruneRetained Tokens=1282025.0976.57
DivPruneRetained Tokens=1282025.0975.96
SparseVLMRetained Tokens=1922025.0975.6
VisionZipRetained Tokens=1282025.0975.6
PyramidDropRetained Tokens=1922025.0975.1
ZOO-PruneRetained Tokens=642025.0975.02
HiREDRetained Tokens=1922025.0974.9
DivPruneRetained Tokens=642025.0974.11
SparseVLMRetained Tokens=1282025.0973.8
HiREDRetained Tokens=1282025.0973.4
PyramidDropRetained Tokens=1282025.0972.9
VisionZipRetained Tokens=642025.0972.4
HiREDRetained Tokens=642025.0969.7
PyramidDropRetained Tokens=642025.0969.2
SparseVLMRetained Tokens=642025.0968.2
ToMeRetained Tokens=1922025.0968
FastVRetained Tokens=1922025.0967.1
ToMeRetained Tokens=1282025.0963
FastVRetained Tokens=1282025.0961.8
ToMeRetained Tokens=642025.0957.1
FastVRetained Tokens=642025.0955