pytorch opacus issues - Githubissues

pytorch / opacus

Training PyTorch models with differential privacy

https://opacus.ai

Apache License 2.0

1.72k stars 345 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

[BC breaking change in torch] `weights_only` default flip for `torch.load`

#689 mikaylagawarecki opened 1 week ago
2
DP Federated Learning with different samples each round

#688 agarciall opened 1 week ago
1
can I set some model parameter's requires_grad false after make_private

#687 HuitongJin closed 1 day ago
2
DP CNN+LSTM model // keep batch_size on the first size of the input

#686 minsung-k closed 1 day ago
2
Issue with Seq2Seq model

#685 kr-ramesh opened 2 weeks ago
1
delete CircleCI configs since GithubActions CI are now live

#684 HuanyuZhang opened 2 weeks ago
1
Does Gradient Accumulation work with ghost clipping?

#683 RobRomijnders closed 3 weeks ago
3
Fix a doctest error with param_rename dictionary

#682 EnayatUllah opened 4 weeks ago
3
replace uses of np.ndarray with npt.NDArray (#680)

#681 igorsugak closed 1 month ago
2
Upgrade fbcode/pytorch to Python Scientific Stack 2

#680 igorsugak opened 1 month ago
2
Support bias correction when using Adam with DP

#679 qiaoyuet closed 3 weeks ago
1
Running Larger Models (Pipeline Parallel/Model Parallel)

#678 aflah02 closed 1 month ago
3
Website and Github update

#677 EnayatUllah opened 1 month ago
11
Invalid combnation of arguments related to empty batch

#676 SoumiDas opened 2 months ago
3
Fix the initialization function of "GradSampleModuleFastGradientClipping"

#675 HuanyuZhang closed 1 month ago
5
Fix failed github action CIs

#674 HuanyuZhang closed 2 months ago
2
GradSampleModuleFastGradientClipping ignores strict and force_functorch params

#673 anhnami closed 1 month ago
4
Model doesn't train even when ModuleValidator.validate yields no errors

#672 sarakodeiri opened 2 months ago
4
Opacus for 3D Segmentation

#671 Ziva1011 closed 1 month ago
3
Unable to Adjust Alpha Range for Privacy Bound in Latest Version of Opacus

#670 zedoul closed 2 months ago
1
How to use opacus if using DP for parameters

#669 zipckkk closed 2 months ago
2
Towards making the interface of ghost clipping same as that of PyTorch

#668 EnayatUllah closed 2 months ago
17
Updating a tutorial to include Ghost Clipping DP-SGD

#667 EnayatUllah closed 3 months ago
6
Training a simple transformer model with Opacus produces runtime error due to mismatch in dimensions

#666 nhianK closed 1 month ago
2
Add multi_gpu test for ghost clipping

#665 iden-kalemaj closed 3 months ago
9
Fixed the clipping operation in fast gradient clipping with Privacy Engine

#664 EnayatUllah closed 3 months ago
3
Opacus release v1.5.2

#663 HuanyuZhang closed 3 months ago
4
Fix DistributedDP Optimizer for Fast Gradient Clipping

#662 HuanyuZhang closed 3 months ago
8
One backward function for Ghost Clipping

#661 EnayatUllah closed 3 months ago
6
Fix issue with setting of param_group for the DPOptimizer wrapper

#660 iden-kalemaj closed 3 months ago
8
release v1.5.1

#659 HuanyuZhang closed 3 months ago
3
Explicit opt_einsum.contract import for linear.py

#658 HuanyuZhang closed 4 months ago
3
release v1.5

#657 HuanyuZhang closed 4 months ago
2
Fast Gradient Clipping and Ghost Clipping

#656 EnayatUllah closed 4 months ago
19
Microbatching Support

#655 shs037 opened 4 months ago
4
Fix the doctest due to numpy 2.0

#654 HuanyuZhang closed 4 months ago
5
Change register_backward_hook to register_full_backward_hook

#653 HuanyuZhang opened 5 months ago
1
Grad Sample Module: Use full backward hook to save activations and backprop values.

#652 ParthS007 closed 4 months ago
1
Fix gradient shape error for DPMultiheadAttention (issue 650)

#651 HuanyuZhang closed 5 months ago
5
Error in DPOptimizer: Inconsistency between batch_first argument of PrivacyEngine and DPMultiheadAttention

#650 tklausen closed 5 months ago
2
Wrapper references can be easily replaced, consider using properties instead

#649 andreatgretel closed 3 months ago
3
Pass kwargs from make_private to _prepare_optimizer

#648 eigengravy closed 5 months ago
6
BatchMemoryManager for training private GANs

#647 Caarinaaa opened 7 months ago
0
Spectral normalization in Opacus

#646 100sunnygirl closed 3 months ago
4
ValueError: Per sample gradient is not initialized. Not updated in backward pass?

#645 Lhsakfheudsj35407 opened 7 months ago
4
Making a custom transformer architecture work with opacus

#644 nhianK opened 7 months ago
3
Why GDP and RDP give different result for the same config

#643 huynhducloi00 closed 8 months ago
1
Fix BatchMemoryManager length

#641 dwahdany closed 6 months ago
7
`BatchSplittingSampler` return wrong length

#640 dwahdany closed 6 months ago
5
Integrating Opacus for a custom pytorch model raises errors but works fine on its own

#639 nhianK opened 8 months ago
2