List your company
ResearchFormerly WizWand

Zero-shot Evaluation on OpenCLIP 38 datasets

67.3Average PerformanceEnsemble Teacher
47.33252.51657.762.884Nov 28, 2023
Updated 8mo ago

Evaluation results

MethodLinks
Ensemble TeacherTraining Dataset=DataComp-1B [18], OpenAI-400M [47], Image Encoder=ViT-L/142023.1167.3
MobileCLIP-B (LT)Training Dataset=DataCompDR-1B, Image Encoder=ViT-B/162023.1165.8
MobileCLIP-BTraining Dataset=DataCompDR-1B, Image Encoder=ViT-B/162023.1165.2
MobileCLIP-S2Training Dataset=DataCompDR-1B, Image Encoder=MCi22023.1163.7
SigLIP-B/16Training Dataset=Webli-1B, Image Encoder=ViT-B/162023.1162.3
DataComp-B/16Training Dataset=DataComp-1B [18], Image Encoder=ViT-B/162023.1161.5
MobileCLIP-S1Training Dataset=DataCompDR-1B, Image Encoder=MCi12023.1161.3
DataComp-B/32-256Training Dataset=DataComp-1B [18], Image Encoder=ViT-B/32-2562023.1160.9
DFN-B/16Training Dataset=DFN-2B [16], Image Encoder=ViT-B/162023.1160.9
EVA02-B/16Training Dataset=Merged-2B [55], Image Encoder=ViT-B/162023.1158.9
LAION-B/16Training Dataset=LAION-2B [52], Image Encoder=ViT-B/162023.1158.7
MobileCLIP-S0Training Dataset=DataCompDR-1B, Image Encoder=MCi02023.1158.1
OpenAI-B/16Training Dataset=WIT-400M [47], Image Encoder=ViT-B/162023.1158.1
DataComp-B/32Training Dataset=DataComp-1B [18], Image Encoder=ViT-B/322023.1158
LAION-B/32Training Dataset=LAION-2B [52], Image Encoder=ViT-B/322023.1154.8
TinyCLIP-61M/32Training Dataset=LAION-400M [51], Image Encoder=ViT-61M/322023.1153
OpenAI-B/32Training Dataset=OpenAI-400M [47], Image Encoder=ViT-B/322023.1152.5
TinyCLIP-40M/32Training Dataset=LAION-400M [51], Image Encoder=ViT-40M/322023.1151.2
OpenAI-RN101Training Dataset=OpenAI-400M [47], Image Encoder=ResNet-1012023.1150.3
TinyCLIP-RN30MTraining Dataset=LAION-400M [51], Image Encoder=ResNet-30M2023.1150.2
TinyCLIP-RN19MTraining Dataset=LAION-400M [51], Image Encoder=ResNet-19M2023.1148.3
OpenAI-RN50Training Dataset=OpenAI-400M [47], Image Encoder=ResNet-502023.1148.1