SOTAVerified

Network Pruning

Network Pruning is a popular approach to reduce a heavy network to obtain a light-weight form by removing redundancy in the heavy network. In this approach, a complex over-parameterized network is first trained, then pruned based on come criterions, and finally fine-tuned to achieve comparable performance with reduced parameters.

Source: Ensemble Knowledge Distillation for Learning Improved and Efficient Networks

Papers

Showing 150 of 534 papers

TitleStatusHype
Exploring GLU Expansion Ratios: A Study of Structured Pruning in LLaMA-3.2 ModelsCode5
DepGraph: Towards Any Structural PruningCode4
Structured Pruning for Deep Convolutional Neural Networks: A surveyCode4
LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPSCode2
A Simple and Effective Pruning Approach for Large Language ModelsCode2
MaskLLM: Learnable Semi-Structured Sparsity for Large Language ModelsCode2
Pruning Filters for Efficient ConvNetsCode2
A Survey on Deep Neural Network Pruning-Taxonomy, Comparison, Analysis, and RecommendationsCode2
Is C4 Dataset Optimal for Pruning? An Investigation of Calibration Data for LLM PruningCode1
Topology-Aware Network Pruning using Multi-stage Graph Embedding and Reinforcement LearningCode1
Group Fisher Pruning for Practical Network CompressionCode1
Good Subnetworks Provably Exist: Pruning via Greedy Forward SelectionCode1
How Not to Give a FLOP: Combining Regularization and Pruning for Efficient InferenceCode1
Investigating Sparsity in Recurrent Neural NetworksCode1
Iterative Soft Shrinkage Learning for Efficient Image Super-ResolutionCode1
Filter-Pruning of Lightweight Face Detectors Using a Geometric Median CriterionCode1
Efficient Latency-Aware CNN Depth Compression via Two-Stage Dynamic ProgrammingCode1
EfficientTDNN: Efficient Architecture Search for Speaker RecognitionCode1
Filter Sketch for Network PruningCode1
M-FAC: Efficient Matrix-Free Approximations of Second-Order InformationCode1
Feather: An Elegant Solution to Effective DNN SparsificationCode1
Auto-Train-Once: Controller Network Guided Automatic Network Pruning from ScratchCode1
Finding Skill Neurons in Pre-trained Transformer-based Language ModelsCode1
Fluctuation-based Adaptive Structured Pruning for Large Language ModelsCode1
1xN Pattern for Pruning Convolutional Neural NetworksCode1
Adversarial Pruning: A Survey and Benchmark of Pruning Methods for Adversarial RobustnessCode1
AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural NetworksCode1
Channel Gating Neural NetworksCode1
How Sparse Can We Prune A Deep Network: A Fundamental Limit ViewpointCode1
HRank: Filter Pruning using High-Rank Feature MapCode1
Dynamic Sparse No Training: Training-Free Fine-tuning for Sparse LLMsCode1
Dynamic Channel Pruning: Feature Boosting and SuppressionCode1
Dynamic Sparse Training: Find Efficient Sparse Network From Scratch With Trainable Masked LayersCode1
A Gradient Flow Framework For Analyzing Network PruningCode1
DHP: Differentiable Meta Pruning via HyperNetworksCode1
DominoSearch: Find layer-wise fine-grained N:M sparse schemes from dense neural networksCode1
Aligned Structured Sparsity Learning for Efficient Image Super-ResolutionCode1
Beyond Size: How Gradients Shape Pruning Decisions in Large Language ModelsCode1
Effective Sparsification of Neural Networks with Global Sparsity ConstraintCode1
A Three-regime Model of Network PruningCode1
Automatic Neural Network Pruning that Efficiently Preserves the Model AccuracyCode1
Automatic Network Pruning via Hilbert-Schmidt Independence Criterion Lasso under Information Bottleneck PrincipleCode1
Recent Advances on Neural Network Pruning at InitializationCode1
Discovering Neural WiringsCode1
Layer-adaptive sparsity for the Magnitude-based PruningCode1
An Information Theory-inspired Strategy for Automatic Network PruningCode1
Accelerating Sparse DNN Models without Hardware-Support via Tile-Wise SparsityCode1
APP: Anytime Progressive PruningCode1
Advanced Dropout: A Model-free Methodology for Bayesian Dropout OptimizationCode1
Comparing Rewinding and Fine-tuning in Neural Network PruningCode1
Show:102550
← PrevPage 1 of 11Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1ResNet50-2.3 GFLOPsAccuracy78.79Unverified
2ResNet50-1.5 GFLOPsAccuracy78.07Unverified
3ResNet50 2.5 GFLOPSAccuracy78Unverified
4RegX-1.6GAccuracy77.97Unverified
5ResNet50 2.0 GFLOPSAccuracy77.7Unverified
6ResNet50-3G FLOPsAccuracy77.1Unverified
7ResNet50-2G FLOPsAccuracy76.4Unverified
8ResNet50-1G FLOPsAccuracy76.38Unverified
9TAS-pruned ResNet-50Accuracy76.2Unverified
10ResNet50Accuracy75.59Unverified
#ModelMetricClaimedVerifiedStatus
1FeatherTop-1 Accuracy76.93Unverified
2SpartanTop-1 Accuracy76.17Unverified
3ST-3Top-1 Accuracy76.03Unverified
4AC/DCTop-1 Accuracy75.64Unverified
5CSTop-1 Accuracy75.5Unverified
6ProbMaskTop-1 Accuracy74.68Unverified
7STRTop-1 Accuracy74.31Unverified
8DNWTop-1 Accuracy74Unverified
9GMPTop-1 Accuracy73.91Unverified
#ModelMetricClaimedVerifiedStatus
1+U-DML*Inference Time (ms)675.56Unverified
2DenseAccuracy79Unverified
3AC/DCAccuracy78.2Unverified
4Beta-RankAccuracy74.01Unverified
5TAS-pruned ResNet-110Accuracy73.16Unverified
#ModelMetricClaimedVerifiedStatus
1TAS-pruned ResNet-110Accuracy94.33Unverified
2ShuffleNet – QuantisedInference Time (ms)23.15Unverified
3AlexNet – QuantisedInference Time (ms)5.23Unverified
4MobileNet – QuantisedInference Time (ms)4.74Unverified
#ModelMetricClaimedVerifiedStatus
1FFN-ShapleyPrunedAvg #Steps12.05Unverified