![](https://img-blog.csdnimg.cn/20201014180756928.png?x-oss-process=image/resize,m_fixed,h_64,w_64)
模型压缩
文章平均质量分 64
onion_rain
这个作者很懒,什么都没留下…
展开
-
[量化]Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constraine
Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1已复现:https://github.com/onion-rain/MCTS:trainpython trainer_exp.py --gpu 1 --arch resnet14_binary --dataset cifar10 --valuate --binary..原创 2021-01-20 18:53:50 · 318 阅读 · 0 评论 -
[剪枝]To prune, or not to prune: exploring the efficacy of pruning for model compression
To prune, or not to prune: exploring the efficacy of pruning for model compression1 Introductiongiven a bound on the model’s memory footprint, how can we arrive at the most accurate model?作者对比了两种等价的模型:(1) large-sparse(2) small-dense2 Related work原创 2021-01-18 21:42:54 · 406 阅读 · 0 评论 -
[剪枝]Learning Effcient Convolutional Networks through Network Slimming
Learning Effcient Convolutional Networks through Network Slimming已复现:https://github.com/onion-rain/MCTS:python slimmer.py --arch vgg19_bn_cifar --gpu 4 --resume checkpoints/sparsity/cifar10_vgg19_bn_cifar_sr_best.pth.tar --slim 0.7首先train ...原创 2021-01-18 21:41:34 · 168 阅读 · 0 评论 -
[剪枝]Channel Pruning for Accelerating Very Deep Neural Networks
Channel Pruning for Accelerating Very Deep Neural Networks已复现:https://github.com/onion-rain/MCTS:python thinet_exp.py --arch vgg16_bn_cifar --gpu 5 --resume checkpoints/baseline/cifar10_vgg16_bn_cifar_best.pth.tar --prune-percent 0.3 --lp-norm ..原创 2021-01-18 21:40:42 · 212 阅读 · 0 评论 -
[剪枝]Pruning Filters for Efficient ConvNets
Pruning Filters for Efficient ConvNets已复现:https://github.com/onion-rain/MCTS:python filter_pruner_exp.py --arch vgg16_bn_cifar --gpu 5 --resume checkpoints/baseline/cifar10_vgg16_bn_cifar_best.pth.tar --prune-percent 0.3 --lp-norm 1剪枝权重绝对值...原创 2021-01-18 21:34:44 · 263 阅读 · 0 评论