Spaces:
Running
Running
Models,Model Size(B),Data Source,Overall,Classification,VQA,Retrieval,Grounding | |
clip-vit-large-patch14,0.428,TIGER-Lab,37.8,42.8,9.1,53.0,51.8 | |
blip2-opt-2.7b,3.74,TIGER-Lab,25.2,27.0,4.2,33.9,47.0 | |
siglip-base-patch16-224,0.203,TIGER-Lab,34.8,40.3,8.4,31.6,59.5 | |
CLIP-ViT-H-14-laion2B-s32B-b79K,0.986,TIGER-Lab,39.7,47.8,10.9,52.3,53.3 | |
UniIR (BLIP_FF),0.247,TIGER-Lab,42.8,42.1,15.0,60.1,62.2 | |
UniIR (CLIP_SF),0.428,TIGER-Lab,44.7,44.3,16.2,61.8,65.3 | |
e5-v,8.36,TIGER-Lab,13.3,21.8,4.9,11.5,19.0 | |
Magiclens,0.428,TIGER-Lab,27.8,38.8,8.3,35.4,26.0 | |
CLIP-FullFineTuned,0.428,TIGER-Lab,45.4,55.2,19.7,53.2,62.2 | |
OpenCLIP-FullFineTuned,0.632,TIGER-Lab,47.2,56.0,21.9,55.4,64.1 | |
VLM2Vec (Phi-3.5-V-FFT),4.15,TIGER-Lab,55.9,52.8,50.3,57.8,72.3 | |
VLM2Vec (Phi-3.5-V-LoRA),4.15,TIGER-Lab,60.1,54.8,54.9,62.3,79.5 | |
VLM2Vec (LLaVA-1.6-LoRA-LowRes),7.57,TIGER-Lab,55.0,54.7,50.3,56.2,64.0 | |
VLM2Vec (LLaVA-1.6-LoRA-HighRes),7.57,TIGER-Lab,62.9,61.2,49.9,67.4,86.1 |