Learning-Rate-Dropout
Pytorch implementation of Learning Rate Dropout.
Paper Link: https://arxiv.org/pdf/1912.00144.pdf
Train ResNet-34 for Cifar10:
run:
python main.py --model=resnet --optim=adam_lrd --lr=0.001 --LRD_p=0.5
python main.py --model=resnet --optim=adam --lr=0.001
python main.py --model=resnet --optim=sgd_lrd --lr=0.1 --LRD_p=0.5
python main.py --model=resnet --optim=sgd --lr=0.1
python main.py --model=resnet --optim=rmsprop_lrd --lr=0.001 --LRD_p=0.5
python main.py --model=resnet --optim=rmsprop --lr=0.001
After training, run "plot.py" to show the learning curves.
56 Jul 15, 2022
2 Dec 02, 2022
5 Dec 04, 2022
2 Nov 11, 2022
11 Dec 17, 2022
130 Jan 06, 2023
99 Dec 26, 2022
38 Sep 16, 2022
18 May 04, 2022
1 May 31, 2022
165 Nov 22, 2022
1 Sep 11, 2021
32 Dec 21, 2022
1 Jan 15, 2022
573 Jan 04, 2023
26 Dec 03, 2022
78 Dec 08, 2022
25 Dec 04, 2022
1 Apr 17, 2022
72 Jan 02, 2023