List your company
ResearchFormerly WizWand

Image Classification on ImageNet 1k (Top-1, Avg Top-1, Rank)

83.5Top-1 AccSwin-B (Dense)
63.11668.40873.778.992Jun 10, 2025Aug 13, 2025Oct 17, 2025Dec 21, 2025Feb 24, 2026Apr 30, 2026Jul 4, 2026
Updated 3mo ago

Evaluation results

MethodLinks
Swin-B (Dense)Backbone=Swin-B, MACs (G)=15.4, Throughput (img/s)=690.25, Params (M)=87.8, Speedup (x)=1.002026.0783.5——
Swin-S (Dense)Backbone=Swin-S, MACs (G)=8.7, Throughput (img/s)=1106.9, Params (M)=49.6, Speedup (x)=1.002026.0783.2——
HetDPTBackbone=Swin-B, MACs (G)=9.8, Throughput (img/s)=1106.95, Params (M)=55.3, Speedup (x)=1.602026.0782.9——
SAViTBackbone=Swin-B, MACs (G)=7.7, Throughput (img/s)=1056.08, Params (M)=33.0, Speedup (x)=1.532026.0782.6——
HetDPT (10 layers)Backbone=Swin-S, MACs (G)=7.3, Throughput (img/s)=1365.9, Params (M)=43.9, Speedup (x)=1.232026.0782.4——
DW Prun. (0.2-24)Backbone=Swin-S, MACs (G)=6.8, Throughput (img/s)=1140.1, Params (M)=39.2, Speedup (x)=1.102026.0782.2——
HetDPT (8 layers)Backbone=DeiT-B, MACs (G)=11.8, Throughput (img/s)=1048.5, Params (M)=61.4, Speedup (x)=1.412026.0782.1——
DW Prun. (0.2-12)Backbone=DeiT-B, MACs (G)=13.6, Throughput (img/s)=907.6, Params (M)=68.2, Speedup (x)=1.222026.0781.9——
DeiT-B (Dense)Backbone=DeiT-B, MACs (G)=17.6, Throughput (img/s)=741.5, Params (M)=86.6, Speedup (x)=1.002026.0781.8——
NOSE (5 layers)Backbone=DeiT-B, MACs (G)=14.6, Throughput (img/s)=894.7, Params (M)=75.5, Speedup (x)=1.212026.0781.8——
HetDPT (10 layers)Backbone=DeiT-B, MACs (G)=10.5, Throughput (img/s)=1169.7, Params (M)=54.9, Speedup (x)=1.582026.0781.8——
DW Prun. (0.2-22)Backbone=Swin-S, MACs (G)=6.3, Throughput (img/s)=1273.0, Params (M)=34.4, Speedup (x)=1.152026.0781.8——
NOSE (6 layers)Backbone=DeiT-B, MACs (G)=14.1, Throughput (img/s)=921.4, Params (M)=73.2, Speedup (x)=1.242026.0781.5——
SGDModel=ResNet-101, Epoch=2002025.0681.3——
MACModel=ResNet-101, Epoch=2002025.0681.2——
KFACModel=ResNet-101, Epoch=2002025.0681.1——
EVAModel=ResNet-101, Epoch=2002025.0681.1——
FOOFModel=ResNet-101, Epoch=2002025.0681——
DW Prun. (0.2-10)Backbone=DeiT-B, MACs (G)=10.8, Throughput (img/s)=1037.9, Params (M)=59.4, Speedup (x)=1.402026.0780.9——
ADAMWModel=ResNet-101, Epoch=2002025.0680.6——
MACModel=Swin-Tiny, Epoch=2002025.0680.1——
ADAMWModel=Swin-Tiny, Epoch=2002025.0680.1——
HetDPT (6 layers)Backbone=DeiT-S, MACs (G)=3.3, Throughput (img/s)=2987.3, Params (M)=17.6, Speedup (x)=1.272026.0780.1——
FOOFModel=ResNet-101, Epoch=1002025.0680——
MACModel=ResNet-101, Epoch=1002025.0679.8——
DeiT-S (Dense)Backbone=DeiT-S, MACs (G)=4.6, Throughput (img/s)=2349.9, Params (M)=22.1, Speedup (x)=1.002026.0779.8——
MACModel=ResNet-50, Epoch=2002025.0679.7——
SGDModel=ResNet-101, Epoch=1002025.0679.7——
FOOFModel=ResNet-50, Epoch=2002025.0679.7——
HetDPT (8 layers)Backbone=DeiT-S, MACs (G)=3.0, Throughput (img/s)=3275.0, Params (M)=15.9, Speedup (x)=1.392026.0779.7——
SGDModel=ResNet-50, Epoch=2002025.0679.6——
KFACModel=ResNet-101, Epoch=1002025.0679.6——
EVAModel=ResNet-101, Epoch=1002025.0679.6——
NOSE (5 layers)Backbone=DeiT-S, MACs (G)=3.7, Throughput (img/s)=2875.3, Params (M)=19.5, Speedup (x)=1.222026.0779.6——
EVAModel=ResNet-50, Epoch=2002025.0679.4——
KFACModel=ResNet-50, Epoch=2002025.0679.3——
ADAMWModel=ResNet-50, Epoch=2002025.0679.2——
SGDModel=Swin-Tiny, Epoch=2002025.0678.9——
DW Prun. (0.3-12)Backbone=DeiT-S, MACs (G)=3.1, Throughput (img/s)=2989.8, Params (M)=15.0, Speedup (x)=1.272026.0778.6——
FOOFModel=ResNet-50, Epoch=1002025.0678.4——
KFACModel=ResNet-50, Epoch=1002025.0678.2——
SGDModel=ResNet-50, Epoch=1002025.0678.1——
MACModel=ResNet-50, Epoch=1002025.0678——
ADAMWModel=ResNet-101, Epoch=1002025.0677.9——
ADAMWModel=DeiT-Small, Epoch=2002025.0677.9——
MACModel=Swin-Tiny, Epoch=1002025.0677.7——
EVAModel=ResNet-50, Epoch=1002025.0677.7——
MACModel=DeiT-Small, Epoch=2002025.0677.4——
ADAMWModel=Swin-Tiny, Epoch=1002025.0677.4——
ADAMWModel=ResNet-50, Epoch=1002025.0676.8——
EVAModel=DeiT-Small, Epoch=2002025.0676.6——
SGDModel=Swin-Tiny, Epoch=1002025.0676——
SGDModel=DeiT-Small, Epoch=2002025.0675.3——
ADAMWModel=DeiT-Small, Epoch=1002025.0673.7——
FOOFModel=Swin-Tiny, Epoch=1002025.0673.7——
MACModel=DeiT-Small, Epoch=1002025.0673.5——
FOOFModel=Swin-Tiny, Epoch=2002025.0672.6——
KFACModel=DeiT-Small, Epoch=1002025.0669.9——
EVAModel=DeiT-Small, Epoch=1002025.0669.7——
SGDModel=DeiT-Small, Epoch=1002025.0669.1——
FOOFModel=DeiT-Small, Epoch=2002025.0667.9——
FOOFModel=DeiT-Small, Epoch=1002025.0663.9——
ADAMWSummary=All Models2025.06—782
EVASummary=All Models2025.06—585
FOOFSummary=All Models2025.06—74.74
KFACSummary=All Models2025.06—48.56
MACSummary=All Models2025.06—78.41
SGDSummary=All Models2025.06—77.33