List your company

Vision-Language Evaluation Suite on LLaVa 1.5

63.01GQA ScoreRADIO-ViT-H/16
49.396452.930756.46559.9993Dec 10, 2023

Evaluation results

MethodLinks
RADIO-ViT-H/16Params (M)=653, Resolution=432, Throughput=1582023.1263.0186.256.3279.28
OpenAI CLIP-L/14Params (M)=305, Resolution=336, Throughput=4142023.1262.286.0957.9278.49
DINOv2-g/14-regParams (M)=1137, Resolution=224, Throughput=2942023.1261.8885.6247.1876.23
DFN CLIP-H/14Params (M)=633, Resolution=378, Throughput=1702023.1261.7385.9156.7878.78
E-RADIO-LParams (M)=391, Resolution=512, Throughput=4682023.1261.785.0751.4776.73
Intern-ViT-6B (448)Params (M)=5537, Resolution=448, Throughput=142023.1261.1987.2360.3678.83
MetaCLIP-H/14Params (M)=632, Resolution=224, Throughput=4862023.1260.5784.7653.6575.71
Intern-ViT-6B (224)Params (M)=5902, Resolution=224, Throughput=632023.1260.1884.0252.4576.75
OpenCLIP-H/14Params (M)=632, Resolution=224, Throughput=5032023.1257.9483.6150.4872.24
SigLIP-L/14Params (M)=428, Resolution=384, Throughput=2412023.1257.784.8556.6571.94
SAM-H/16Params (M)=637, Resolution=1024, Throughput=122023.1249.9281.7643.9157.65