List your company
ResearchFormerly WizWand

Fine-grained Image Classification on EuroSAT (test)

0.9Top-1 Test ErrorLoRA
-13.86485.793185.45285.107Mar 7, 2023
Updated 8mo ago

Evaluation results

MethodLinks
LoRABackbone=ViT-L/162023.030.9
Full FTBackbone=ViT-L/162023.031
VPTBackbone=ViT-L/162023.031.1
InCABackbone=ViT-L/162023.031.2
BitFitBackbone=ViT-L/162023.031.4
In. MLP-3Backbone=ViT-L/162023.031.5
AdaLNBackbone=ViT-L/162023.031.5
InCA (last)Backbone=ViT-L/162023.031.9
In. LPBackbone=ViT-L/162023.032.1
MLP-3Backbone=ViT-L/162023.032.5
LPBackbone=ViT-L/162023.033.7
Full FTBackbone=SWIN-L, % Trainable parameters=100%, No backbone backpropagation=false2023.0370
LoRABackbone=SWIN-L, % Trainable parameters=0.8%, No backbone backpropagation=false2023.0390
AdaLNBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=false2023.03110
InCABackbone=SWIN-L, % Trainable parameters=3.7%, No backbone backpropagation=true2023.03150
In. MLP-3Backbone=SWIN-L, % Trainable parameters=2.8%, No backbone backpropagation=true2023.03160
BitFitBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=false2023.03170
MLP-3Backbone=SWIN-L, % Trainable parameters=2.8%, No backbone backpropagation=false2023.03220
InCA (last)Backbone=SWIN-L, % Trainable parameters=3.7%, No backbone backpropagation=true2023.03240
In. LPBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=true2023.03270
LPBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=false2023.03370