Release v1.5.4 (#763)

Summary:
Pull Request resolved: #763

As titled

Reviewed By: iden-kalemaj

Differential Revision: D75453194

fbshipit-source-id: ca64dfde7005dbc428b302ea6c1161e70c21e85b

release v.1.5.3 (#730)

Summary:
Pull Request resolved: #730

From 1.5.2 to 1.5.3

Reviewed By: HuanyuZhang

Differential Revision: D69532789

fbshipit-source-id: 2038f49cd0740fbd86068a67d7da0580fa1a893b

Opacus release v1.5.2 (#663)

Summary:
Pull Request resolved: #663

Release a new version of Opacus

Furthermore, we replace "opt_einsum.contract" by torch.einsum to avoid errors when "opt_einsum" is not available. This will not hurt the performance since torch will automatically shift to "opt_einsum" for acceleratiton when the package is available (https://pytorch.org/docs/stable/generated/torch.einsum.html)
Code pointer: https://pytorch.org/docs/stable/_modules/torch/backends/opt_einsum.html#is_available

Reviewed By: EnayatUllah

Differential Revision: D60672828

fbshipit-source-id: f8bbc0aa404e48f15ce129689a6e55af68daa5e4

release v1.5.1 (#659)

Summary:
Pull Request resolved: #659

as title

Differential Revision: D60290993

fbshipit-source-id: 86cae51208281c294dcf98bf36506a0775b76358

release v1.5 (#657)

Summary:
Pull Request resolved: #657

Releasing v1.5

Reviewed By: EnayatUllah

Differential Revision: D60058775

fbshipit-source-id: 5f4084938afcb0b8adcd83ea69a52d3847e3392d

release v1.4.1 (#628)

Summary:
Pull Request resolved: #628

Releasing v1.4.1

Differential Revision: D53626083

fbshipit-source-id: 32bf2e7b9ffb9544ed6dab8929b4bb901b3967e5

up Pytorch version for PyPI (#568)

Summary:
Deploy failed due to unavailable Python

https://github.com/pytorch/opacus/actions/runs/4241660361

Pull Request resolved: #568

Reviewed By: karthikprasad

Differential Revision: D43494621

Pulled By: ffuuugor

fbshipit-source-id: d8ae67e0133ee963715b34966ae7193156def309

release v1.4 (#563)

Summary:
subj

Pull Request resolved: #563

Reviewed By: alexandresablayrolles, Anonymani, karthikprasad

Differential Revision: D43242386

Pulled By: ffuuugor

fbshipit-source-id: 0843a298209cb424e5f4578d128307f7593835f4

Add third blog series to the website (#537)

Summary:
## Types of changes

- [ ] Bug fix (non-breaking change which fixes an issue)
- [ ] New feature (non-breaking change which adds functionality)
- [ ] Breaking change (fix or feature that would cause existing functionality to change)
- [x] Docs change / refactoring / dependency upgrade

## Motivation and Context / Related issue

Add third blog series to the website

## How Has This Been Tested (if it applies)

No test is required

## Checklist

- [x] The documentation is up-to-date with the changes I made.
- [x] I have read the **CONTRIBUTING** document and completed the CLA (see **CONTRIBUTING**).
- [x] All tests passed, and additional code has been covered with new tests.

Pull Request resolved: #537

Reviewed By: karthikprasad

Differential Revision: D41042020

Pulled By: ashkan-software

fbshipit-source-id: d10d65bb3d56620a01a57cc45da12b87041b1354

No-Op GradSampleModule (#492)

Summary:
TL;DR: Adding a No-Op GradSampleModule in case the grad samples are computed by functorch. The CIFAR10 example has been updated to show a typical use-case for that.

The neat thing about functorch is that it directly gives the per-sample gradients with a couple of lines of code. These per-sample gradients are then manually given to `p.grad_sample` by the end-user.

Pull Request resolved: #492

Reviewed By: ffuuugor

Differential Revision: D39204008

Pulled By: alexandresablayrolles

fbshipit-source-id: 22036e6c941522bba7749ef46f97d54f6ee8c551