Network Pruning
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 0
- 평가표
- 5
- 보존된 평가 기록
- 34
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
코드
관련 논문과 평가 결과에 연결된 구현입니다. 출처에서 확인된 저자 구현에는 공식 코드 표시가 있습니다.
https://github.com/3riccc/neural_pruning
공식 코드https://github.com/8ubpshlr23/central-filter
공식 코드https://github.com/ALFA-group/lipizzaner-ae-prune
공식 코드Guiding Evolutionary AutoEncoder Training with Activation-Based Pruning Operators
https://github.com/AamirRaihan/SWAT
공식 코드https://github.com/Alii-Ganjj/InterpretationsSteeredPruning
공식 코드Interpretations Steered Network Pruning via Amortized Inferred Saliency Maps
https://github.com/BayesWatch/pytorch-prunes
공식 코드A Closer Look at Structured Pruning for Neural Network Compression
https://github.com/CRYPTO-KU/FL-Byzantine-Library
공식 코드https://github.com/D-X-Y/NAS-Projects
공식 코드https://github.com/DT42/squeezenet_demo
공식 코드SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/DeepScale/SqueezeNet
공식 코드SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/DingXiaoH/Centripetal-SGD
공식 코드Manipulating Identical Filter Redundancy for Efficient Pruning on Deep and Complicated CNN
https://github.com/EkdeepSLubana/OrthoReg
공식 코드OrthoReg: Robust Network Pruning Using Orthonormality Regularization
https://github.com/EkdeepSLubana/flowandprune
공식 코드https://github.com/Element-Research/dpnn
공식 코드SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/ElrondL/Robust-Quantisation-Bounds
공식 코드Robust error bounds for quantised and pruned neural networks
https://github.com/Eric-mingjie/rethinking-network-pruning
공식 코드https://github.com/FraCorti/Studying-the-impact-of-magnitude-pruning-on-contrastive-learning-methods
공식 코드Studying the impact of magnitude pruning on contrastive learning methods
https://github.com/HERO-MDH/PR-DARTS
공식 코드DASS: Differentiable Architecture Search for Sparse neural networks
https://github.com/HugoTessier-lab/SWD
공식 코드Rethinking Weight Decay For Efficient Neural Network Pruning
https://github.com/IST-DASLab/ACDC
공식 코드AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks
https://github.com/IST-DASLab/M-FAC
공식 코드M-FAC: Efficient Matrix-Free Approximations of Second-Order Information
https://github.com/Jiamian-Wang/IST_for_SR_pruning
공식 코드Iterative Soft Shrinkage Learning for Efficient Image Super-Resolution
https://github.com/JingtongSu/sanity-checking-pruning
공식 코드Sanity-Checking Pruning Methods: Random Tickets can Win the Jackpot
https://github.com/KaiqiZhang/admm-pruning
공식 코드A Systematic DNN Weight Pruning Framework using Alternating Direction Method of Multipliers
https://github.com/KindXiaoming/QHMC
공식 코드Quantum-Inspired Hamiltonian Monte Carlo for Bayesian Sampling
https://github.com/MingSun-Tse/Caffe_IncReg
공식 코드Structured Pruning for Efficient ConvNets via Incremental Regularization
https://github.com/NVlabs/DeepInversion
공식 코드Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion
https://github.com/NVlabs/Taylor_pruning
공식 코드https://github.com/OmarErak/CsiNet_Optimized
공식 코드Accelerating and Compressing Deep Neural Networks for Massive MIMO CSI Feedback
https://github.com/PRIS-CV/AdvancedDropout
공식 코드Advanced Dropout: A Model-free Methodology for Bayesian Dropout Optimization
https://github.com/PaddlePaddle/PaddleOCR
공식 코드https://github.com/Qualcomm-AI-research/pruning-vs-quantization
공식 코드https://github.com/RAIVNLab/STR
공식 코드Soft Threshold Weight Reparameterization for Learnable Sparsity
https://github.com/SCUT-AILab/DCP
공식 코드Discrimination-aware Network Pruning for Deep Model Compression
https://github.com/SMKamrulHasan/Learned-Condensation-Optimizaton-Network
공식 코드출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.