List your company
ResearchFormerly WizWand

Image Classification on ImageNet-1k 1 (val)

75.72Top-1 AccFOAM
22.139236.049649.9663.8704Feb 16, 2022Nov 4, 2022Jul 23, 2023Apr 9, 2024Dec 26, 2024Sep 13, 2025Jun 1, 2026
Updated 5mo ago

Evaluation results

MethodLinks
FOAMf=20, τ=0.75, ϵ0=10−9, ϵmax=3 × 10−72026.0675.72——
stale Shampoof=20, τ=N/A, ϵ0=10−9, ϵmax=N/A2026.0675.13——
SOAPf=20, τ=N/A, ϵ0=10−9, ϵmax=N/A2026.0675.1——
DR-Shampoof=20, τ=0.75, ϵ0=10−9, ϵmax=N/A2026.0673.56——
AdamWf=N/A, τ=N/A, ϵ0=10−9, ϵmax=N/A2026.0672.39——
PRACTISEBackbone=ResNet-34, Training samples=1000, Pruning strategy=dropping blocks2022.0271.990.632.5
PRACTISEBackbone=ResNet-34, Training samples=500, Pruning strategy=dropping blocks2022.0271.890.532.5
PRACTISEBackbone=ResNet-34, Training samples=100, Pruning strategy=dropping blocks2022.0270.489.732.5
PRACTISEBackbone=ResNet-34, Training samples=50, Pruning strategy=dropping blocks2022.026888.232.5
BP (block)Backbone=ResNet-34, Training samples=1000, Pruning strategy=block pruning2022.0266.887.532.5
MiRBackbone=ResNet-34, Training samples=1000, Pruning strategy=filter pruning2022.0266.687.733.8
MiRBackbone=ResNet-34, Training samples=500, Pruning strategy=filter pruning2022.0265.48733.8
BP (block)Backbone=ResNet-34, Training samples=500, Pruning strategy=block pruning2022.026586.532.5
MiRBackbone=ResNet-34, Training samples=100, Pruning strategy=filter pruning2022.0262.184.833.8
BP (block)Backbone=ResNet-34, Training samples=100, Pruning strategy=block pruning2022.0261.684.332.5
BP (block)Backbone=ResNet-34, Training samples=50, Pruning strategy=block pruning2022.0260.683.532.5
MiRBackbone=ResNet-34, Training samples=50, Pruning strategy=filter pruning2022.0259.983.233.8
BP (filter)Backbone=ResNet-34, Training samples=1000, Pruning strategy=filter pruning2022.0251.276.533.8
KDBackbone=ResNet-34, Training samples=1000, Pruning strategy=filter pruning2022.0250.575.933.8
KDBackbone=ResNet-34, Training samples=500, Pruning strategy=filter pruning2022.0245.772.233.8
FSKDBackbone=ResNet-34, Training samples=1000, Pruning strategy=filter pruning2022.0244.970.533.8
BP (filter)Backbone=ResNet-34, Training samples=500, Pruning strategy=filter pruning2022.0242.970.533.8
FSKDBackbone=ResNet-34, Training samples=500, Pruning strategy=filter pruning2022.0242.869.133.8
FSKDBackbone=ResNet-34, Training samples=100, Pruning strategy=filter pruning2022.0236.663.133.8
KDBackbone=ResNet-34, Training samples=100, Pruning strategy=filter pruning2022.0233.16133.8
FSKDBackbone=ResNet-34, Training samples=50, Pruning strategy=filter pruning2022.0231.156.533.8
KDBackbone=ResNet-34, Training samples=50, Pruning strategy=filter pruning2022.0230.157.733.8
BP (filter)Backbone=ResNet-34, Training samples=100, Pruning strategy=filter pruning2022.0227.656.733.8
BP (filter)Backbone=ResNet-34, Training samples=50, Pruning strategy=filter pruning2022.0224.252.733.8