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/AnishDelft/ModelCompression
https://github.com/Banus/caffe-demo
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/Ben-Louis/FisherPruning-Pytorch
https://github.com/BlossomingL/compression_tool
https://github.com/COMP6248-Reproducability-Challenge/REPRODUCIBILITY-REPORT-THE-LOTTERY-TICKET-HYPOTHESIS
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
https://github.com/D-X-Y/AutoDL-Projects
https://github.com/D-X-Y/GDAS
https://github.com/DNNToolBox/Net-Trim-v1
https://github.com/Dawars/SqueezeNet-tf
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/EstherBear/implementation-of-pruning-filters
https://github.com/FlorianMerkle/network-pruning-and-robustness
https://github.com/Goandwanderfaraway/squeezenet-mindspore
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/Guy-Shapira/DeepInversion
Dreaming to Distill: Data-free Knowledge Transfer via DeepInversion
https://github.com/Happy-Virus-IkBeom/LTH_Tensorflow
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
https://github.com/Hoonyyhoon/Comparison-of-Synflow_SNIP_GraSP
SNIP: Single-shot Network Pruning based on Connection Sensitivity
https://github.com/IST-DASLab/sparseprop
AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks
https://github.com/Irtza/Keras_model_compression
https://github.com/JGU-VC/progressive_stochastic_binarization
https://github.com/Jastot/Rodinka_Neural_Network
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/Kaido0/Brain-Tissue-Segment-Keras
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/KaiqiZhang/caffe-admm
A Systematic DNN Weight Pruning Framework using Alternating Direction Method of Multipliers
https://github.com/KarenUllrich/Tutorial_BayesianCompressionForDL
https://github.com/KentaItakura/Classify-crack-image-and-explain-why-using-MATLAB
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/King-Otaku/Emotion_Recognition_DNN
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/LTH14/FSKD
Few Sample Knowledge Distillation for Efficient Network Compression
https://github.com/Lucasc-99/PackNet-Continual-Learning
PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning
https://github.com/Lucasc-99/packnet_cl
PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning
https://github.com/MS-Mind/MS-Code-02/tree/main/configs/squeezenet
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/Mayurji/Image-Classification-PyTorch
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/MichaelisTrofficus/nfs
Feature Selection for Multivariate Time Series via Network Pruning
https://github.com/Mind23-2/MindCode-101/tree/main/ManiDP
https://github.com/Mind23-2/MindCode-116
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/Mind23-2/MindCode-3/tree/main/ManiDP
https://github.com/Mind23-2/MindCode-80
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
https://github.com/MindCode-4/code-9/tree/main/similarity-of-neural-network
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.