zhenghuanhuan
|
9c37a11036
|
[CT][MA][DP]param of DPModel:norm_clip,use default, runtime error][DP]param of DPModel:norm_clip=0, runtime error][diff_privacy][Pref] use DPModel ,the performance of lenet deteriorates too much
https://gitee.com/mindspore/dashboard?issue_id=I1IMD7
https://gitee.com/mindspore/dashboard?issue_id=I1IQ11
https://gitee.com/mindspore/dashboard?issue_id=I1IPKH
|
5 years ago |
zhenghuanhuan
|
be377fb165
|
1IKCU
fix [MA][diff_privacy][Doc] the tutorials of diff_privacy has problem
|
5 years ago |
zhenghuanhuan
|
30f6c260ee
|
update release note and lenet5_dp_model_train.py
|
5 years ago |
zheng-huanhuan
|
b2e0934bf7
|
5-month feature: add differential privacy train and optimizer.
|
5 years ago |
jin-xiulang
|
4c4c34edef
|
Modify MNIST-examples to support all devices(CPU, GPU, Ascend)
|
5 years ago |
jin-xiulang
|
aeba2b3e24
|
modify input batch_size for defenses tests
|
5 years ago |
jin-xiulang
|
243633ac6b
|
Add CPU_based examples for most mnist_examples.
|
5 years ago |
zheng-huanhuan
|
febaf7a253
|
fix pylint warning.
|
5 years ago |
mindspore-ci-bot
|
5467dbdd44
|
!13 Add a fuzzing test framework for DNN and image transform methods
Merge pull request !13 from ZhidanLiu/master
|
5 years ago |
ZhidanLiu
|
f150496e00
|
add a fuzzing test framework and image transform method
|
5 years ago |
mindspore-ci-bot
|
b21849d2c2
|
!14 Add M-DI2-FGSM method
Merge pull request !14 from zheng-huanhuan/2_master
|
5 years ago |
zheng-huanhuan
|
810df70db6
|
1. Add DI2-FGSM method in iterative_gradient_method.py
2. Add M-DI2-FGSM method in iterative_gradient_method.py
3. Add corresponding ut test.
4. Add mnist example of M_DI2_FGSM method.
|
5 years ago |
jin-xiulang
|
a4c4feca4e
|
Develop new module for model fuzz test.
|
5 years ago |
pkuliuliu
|
1765a2a63a
|
update mnist_lenet5 example
|
5 years ago |
zheng-huanhuan
|
96302d7f1b
|
initial version
|
6 years ago |