pytorch opacus issues - Githubissues

pytorch / opacus

Training PyTorch models with differential privacy

https://opacus.ai

Apache License 2.0

1.65k stars 328 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Fix the doctest due to numpy 2.0

#654 HuanyuZhang closed 1 week ago
5
Change register_backward_hook to register_full_backward_hook

#653 HuanyuZhang opened 3 weeks ago
1
Grad Sample Module: Use full backward hook to save activations and backprop values.

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

#651 HuanyuZhang closed 1 month ago
5
Error in DPOptimizer: Inconsistency between batch_first argument of PrivacyEngine and DPMultiheadAttention

#650 tklausen closed 1 month ago
2
Wrapper references can be easily replaced, consider using properties instead

#649 andreatgretel opened 2 months ago
1
Pass kwargs from make_private to _prepare_optimizer

#648 eigengravy closed 3 weeks ago
6
BatchMemoryManager for training private GANs

#647 Caarinaaa opened 2 months ago
0
Spectral normalization in Opacus

#646 100sunnygirl opened 2 months ago
4
ValueError: Per sample gradient is not initialized. Not updated in backward pass?

#645 Lhsakfheudsj35407 opened 2 months ago
4
Making a custom transformer architecture work with opacus

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

#643 huynhducloi00 closed 3 months ago
1
Error occurred when executing GroundingDinoSAMSegment (segment anything): Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument mat2 in method wrapper_CUDA_mm)

#642 Ymshkz opened 3 months ago
0
Fix BatchMemoryManager length

#641 lsc64 closed 1 month ago
7
`BatchSplittingSampler` return wrong length

#640 lsc64 closed 1 month ago
5
Integrating Opacus for a custom pytorch model raises errors but works fine on its own

#639 nhianK opened 3 months ago
2
ModuleValidator.fix() causes layer gradients to be None.

#638 zxj-c closed 1 month ago
1
OverflowError: cannot convert float infinity to integer

#637 Tnia1 closed 1 month ago
2
Error: Trying to add hooks twice to the same model

#636 sonakshiigarg opened 3 months ago
2
Some issue with loading model with 'weight' as opposed to 'pretrained=True'

#635 aymuos15 opened 4 months ago
1
Add context manager to toggle on/off privacy in training loop

#634 bnestor opened 4 months ago
0
Fix GPU-CPU device mismatch error in util filter_dilated_rows

#633 tklausen closed 4 months ago
3
LLM finetuning with Opacus

#632 KKNakkav2 closed 3 months ago
2
Fixed Opacus's Runtime error with an empty batch (issue 612)

#631 EnayatUllah closed 4 months ago
6
Fix the copyright headers

#630 HuanyuZhang closed 4 months ago
2
UnsupportedModuleError: [IllegalModuleConfigurationError('Model needs to be in training mode')]

#629 jS5t3r closed 4 months ago
1
release v1.4.1

#628 HuanyuZhang closed 4 months ago
5
Simplify `lightning` example dataloader access

#627 eonu opened 4 months ago
1
Implementing augmentation multiplicity using Functorch [Error: AttributeError: 'Tensor' object has no attribute '_forward_counter]

#626 mshubhankar closed 4 months ago
1
Fix github action CI for GPU

#625 HuanyuZhang opened 5 months ago
22
IndexError: pop from empty list

#624 shanjin2014 closed 1 month ago
1
refactor: delete unused code line

#623 helin0815 closed 5 months ago
4
Fix lint

#622 karthikprasad closed 5 months ago
4
Grant "write" permission to "github-actions[bot]"

#621 HuanyuZhang closed 5 months ago
5
Move CI to Github Actions

#620 karthikprasad closed 5 months ago
22
Fixing the corner case when the optimizer has no trainable parameters

#619 HuanyuZhang closed 5 months ago
3
opacus does not support the torch.nn.Embedding module

#618 Han-Huaqiao closed 4 months ago
4
Support for is_causal flag for forward hook of layers.DPMultiheadAttention

#617 hueck closed 5 months ago
3
Opacus uses unsafe floating-point noise generation, even with secure_mode=True

#616 TedTed closed 2 months ago
5
Can opacus only inject differential noise in only one layer of parameters or some target layers?

#615 hacker-jerry closed 5 months ago
2
Added GitHub citation file.

#614 mainrs opened 6 months ago
1
Remove requirement to be in training mode to capture activations

#613 lucmos opened 7 months ago
4
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument mat2 in method wrapper_CUDA_mm)

#612 FryLcm closed 4 months ago
15
'PRVAccountant' object has no attribute 'get_privacy_spent'

#611 feigei12346 closed 5 months ago
2
Fix the import order for #609

#610 HuanyuZhang closed 7 months ago
4
Fix Opacus's failed tests

#609 HuanyuZhang closed 7 months ago
7
Error for capture_activations_hook in grad_sample_module.py

#608 conjurer-Fan-Wu closed 7 months ago
13
Opacus on cifar10 attempts to broadcast tensors in an incorrect way

#607 jon-chuang closed 8 months ago
1
fix: make prv accountant robust to larger epsilons

#606 Solosneros closed 7 months ago
4
Move CI from CircleCI to Github Actions

#605 karthikprasad opened 9 months ago
4