Image Classification on ImageNet-1K (Accuracy and Efficiency)
76.4Accuracy (%)OrderDP
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| OrderDPBackbone=ResNet-50, Prune ratio=40%, Epochs=90, Hardware=2-L40-GPU server2026.06 | 76.4 | 21.5 | 43 | |
| Full DataBackbone=ResNet-50, Epochs=90, Hardware=2-L40-GPU server2026.06 | 76.4 | 35.2 | 70.4 | |
| InfoBatchBackbone=ResNet-50, Prune ratio=40%, Epochs=90, Hardware=2-L40-GPU server2026.06 | 75.6 | 21.6 | 43.2 | |
| UCBBackbone=ResNet-50, Prune ratio=40%, Epochs=90, Hardware=2-L40-GPU server2026.06 | 75.4 | 21.1 | 42.2 | |
| epsilon-greedyBackbone=ResNet-50, Prune ratio=40%, Epochs=90, Hardware=2-L40-GPU server2026.06 | 75.2 | 21.1 | 42.2 | |
| RandomBackbone=ResNet-50, Prune ratio=40%, Epochs=90, Hardware=2-L40-GPU server2026.06 | 73.4 | 21.1 | 42.2 |