Network Pruning on ImageNet

Metric: Accuracy (higher is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	Accuracy▼	Augmentations	Paper	Date↕	Code
1	ResNet50-2.3 GFLOPs	78.79	No	Pruning Filters for Efficient ConvNets	2016-08-31	Code
2	ResNet50-1.5 GFLOPs	78.07	No	Pruning Filters for Efficient ConvNets	2016-08-31	Code
3	ResNet50 2.5 GFLOPS	78	No	Knapsack Pruning with Inner Distillation	2020-02-19	Code
4	RegX-1.6G	77.97	No	Group Fisher Pruning for Practical Network Compr...	2021-08-02	Code
5	ResNet50 2.0 GFLOPS	77.7	No	Knapsack Pruning with Inner Distillation	2020-02-19	Code
6	ResNet50-3G FLOPs	77.1	No	EagleEye: Fast Sub-net Evaluation for Efficient ...	2020-07-06	Code
7	ResNet50-2G FLOPs	76.4	No	EagleEye: Fast Sub-net Evaluation for Efficient ...	2020-07-06	Code
8	ResNet50-1G FLOPs	76.376	No	Pruning Filters for Efficient ConvNets	2016-08-31	Code
9	TAS-pruned ResNet-50	76.2	No	Network Pruning via Transformable Architecture S...	2019-05-23	Code
10	ResNet50	75.59	No	Network Pruning That Matters: A Case Study on Re...	2021-05-07	Code
11	ResNet50-1G FLOPs	74.2	No	EagleEye: Fast Sub-net Evaluation for Efficient ...	2020-07-06	Code
12	ResNet50-1G FLOPs	74.2	No	EagleEye: Fast Sub-net Evaluation for Efficient ...	2020-07-06	Code
13	MobileNetV2	73.42	No	Group Fisher Pruning for Practical Network Compr...	2021-08-02	Code
14	ResNet50	73.14	No	AC/DC: Alternating Compressed/DeCompressed Train...	2021-06-23	Code
15	MobileNetV1-50% FLOPs	70.7	No	EagleEye: Fast Sub-net Evaluation for Efficient ...	2020-07-06	Code

#1ResNet50-2.3 GFLOPsSOTA
78.79
Accuracy· 2016-08-31
Pruning Filters for Efficient ConvNets Code
#2ResNet50-1.5 GFLOPs
78.07
Accuracy· 2016-08-31
Pruning Filters for Efficient ConvNets Code
#3ResNet50 2.5 GFLOPS
78
Accuracy· 2020-02-19
Knapsack Pruning with Inner Distillation Code
#4RegX-1.6G
77.97
Accuracy· 2021-08-02
Group Fisher Pruning for Practical Network Compression Code
#5ResNet50 2.0 GFLOPS
77.7
Accuracy· 2020-02-19
Knapsack Pruning with Inner Distillation Code
#6ResNet50-3G FLOPs
77.1
Accuracy· 2020-07-06
EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruning Code
#7ResNet50-2G FLOPs
76.4
Accuracy· 2020-07-06
EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruning Code
#8ResNet50-1G FLOPs
76.376
Accuracy· 2016-08-31
Pruning Filters for Efficient ConvNets Code
#9TAS-pruned ResNet-50
76.2
Accuracy· 2019-05-23
Network Pruning via Transformable Architecture Search Code
#10ResNet50
75.59
Accuracy· 2021-05-07
Network Pruning That Matters: A Case Study on Retraining Variants Code
#11ResNet50-1G FLOPs
74.2
Accuracy· 2020-07-06
EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruning Code
#12ResNet50-1G FLOPs
74.2
Accuracy· 2020-07-06
EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruning Code
#13MobileNetV2
73.42
Accuracy· 2021-08-02
Group Fisher Pruning for Practical Network Compression Code
#14ResNet50
73.14
Accuracy· 2021-06-23
AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks Code
#15MobileNetV1-50% FLOPs
70.7
Accuracy· 2020-07-06
EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruning Code