GitLab

Summary:
(See https://github.com/pytorch/audio/issues/2038 description for the overall goal.)

This commit turns the part that defines `libtorchaudio` into a function
so that it becomes easy to define libraries in the same way as `libtorchaudio`.

Built on top of https://github.com/pytorch/audio/issues/2039

Pull Request resolved: https://github.com/pytorch/audio/pull/2040

Reviewed By: hwangjeff

Differential Revision: D32851990

Pulled By: mthrok

fbshipit-source-id: a8206c62b076bc0849ada1a66c7502ae5ea35e28

Summary:
While updating the documentation in release/0.10, a HIP error was raised.
https://app.circleci.com/pipelines/github/pytorch/audio/8577/workflows/02c6ff44-a042-4f9a-8fb8-573a231f60db/jobs/452639

This happens because `pip install torchaudio -f https://...` defaults to
ROCm version while `build_doc` is supposed to pick the CPU version.

Adding suffix `+cpu` should resolve the isssue.

It is validated on https://github.com/pytorch/audio/pull/2060 https://app.circleci.com/pipelines/github/pytorch/audio/8584/workflows/25ae26e5-273f-46f8-805d-ffc7b6b8eb58/jobs/453337

Pull Request resolved: https://github.com/pytorch/audio/pull/2060

Reviewed By: carolineechen

Differential Revision: D32846765

Pulled By: mthrok

fbshipit-source-id: e6b3b32646388b8c4ba864639f8b62d8b9d39844

Summary:
(See https://github.com/pytorch/audio/issues/2038 description for the overall goal.)
This PR cleans up CMake customization logic for `libtorchaudio`.

It introduces base variables LIBTORCHAUDIO_INCLUDE_DIRS,
LIBTORCHAUDIO_LINK_LIBRARIES and LIBTORCHAUDIO_COMPILE_DEFINITIONS,
which are respectively used when calling `target_include_directories`,
`target_link_libraries` and `target_compile_definitions`.

The customization logic only modifies these variables.

The original implementation called these functions multiple times
(once par customization logic) and it is getting difficult to understand
the customization logic.

Pull Request resolved: https://github.com/pytorch/audio/pull/2039

Reviewed By: carolineechen, nateanl

Differential Revision: D32683004

Pulled By: mthrok

fbshipit-source-id: 4d41f21692ac139b1185a6ab69eb45d881ee7e73

Summary:
Addresses  https://github.com/pytorch/audio/issues/1493

cc mthrok hwangjeff

Pull Request resolved: https://github.com/pytorch/audio/pull/2034

Reviewed By: hwangjeff

Differential Revision: D32807006

Pulled By: mthrok

fbshipit-source-id: badf148646c5f768328c5a4e51bd6016b0be46f3

Summary:
Add training recipe for RNN-T Emformer ASR model to examples directory.

Pull Request resolved: https://github.com/pytorch/audio/pull/2052

Reviewed By: nateanl

Differential Revision: D32814096

Pulled By: hwangjeff

fbshipit-source-id: a5153044efc16cb39f0e6413369a6791637af76a

Summary:
1. stop&disable the windows upgrade that's the major reason of the failure of cuda installation
    https://app.circleci.com/pipelines/github/pytorch/audio/8458/workflows/feb65e3b-1093-4724-b849-1a2ac166f354/jobs/441331
     For more details please check out https://github.com/pytorch/pytorch/issues/64536
2. print the log when the cuda installation fails

Pull Request resolved: https://github.com/pytorch/audio/pull/2032

Reviewed By: mthrok

Differential Revision: D32816145

Pulled By: malfet

fbshipit-source-id: 44a2ef0dd4c43469472a6e518ed64841e2dcd5bb

Summary:
(This is a part of refactor series, followed up by https://github.com/pytorch/audio/issues/2039 and https://github.com/pytorch/audio/issues/2040.
The goal is to make it easy to add a new library artifact alongside with `libtorchudio`, as in https://github.com/pytorch/audio/pull/2048/commits/4ced990849e60f6d19e87ae22819b04d1726648e https://github.com/pytorch/audio/issues/2048 .)

We plan to add prototype/beta third party library integrations,
which could be unstable. (segfault, missing dynamic library dependencies etc...)

If we add such integrations into the existing libtorchaudio,
in the worst case, it will prevent users from just `import torchaudio`.

Instead, we would like to separate the prototype/beta integrations
into separate libraries, so that such issues would not impact all users but
users who attempt to use these prototytpe/beta features.

Say, a prototype feature `foo` is added in `torchaudio.prototype.foo`.
The following initialization procedure will achieve the above mechanism.

1. Place the library file `libtorchaudio_foo` in `torchaudio/lib`.
2. In `torchaudio.prototype.foo.__init__.py`, load the `libtorchaudio_foo`.

Note:
The approach will be slightly different for fbcode, because of how buck deploys
C++ libraries and standardized environment, but the code change here is still
applicable.

Pull Request resolved: https://github.com/pytorch/audio/pull/2038

Reviewed By: carolineechen, nateanl

Differential Revision: D32682900

Pulled By: mthrok

fbshipit-source-id: 0f402a92a366fba8c2894a0fe01f47f8cdd51376

Summary:
Our Griffin-Lim autograd tests take a long time to run. This PR adjusts some parameters to shorten the run time.

For one of the four tests:
Before:
```
test/torchaudio_unittest/transforms/autograd_cpu_test.py . [100%]

======================== 1 passed in 517.35s (0:08:37) =========================
```

After:
```
test/torchaudio_unittest/transforms/autograd_cpu_test.py . [100%]

======================== 1 passed in 104.59s (0:01:44) =========================
```

Pull Request resolved: https://github.com/pytorch/audio/pull/2037

Reviewed By: mthrok

Differential Revision: D32726213

Pulled By: hwangjeff

fbshipit-source-id: c785323ab380aea4b63fb1683b557c8ae842f54e

Summary:
Resolves https://github.com/pytorch/audio/issues/2049, https://github.com/pytorch/audio/issues/1940

Pull Request resolved: https://github.com/pytorch/audio/pull/2050

Reviewed By: nateanl

Differential Revision: D32712513

Pulled By: mthrok

fbshipit-source-id: e1db81786bcca67605ff765d27e0527e20967d1c

Summary:
Similar to https://github.com/pytorch/vision/pull/4788
Make sure the workflow could download right PyTorch with cpu or cuda in case nightly build wasn't ready at that day.

https://app.circleci.com/pipelines/github/pytorch/audio/8427/workflows/11a80738-bcdd-45e3-b37f-328be36c60ee/jobs/438285?invite=true#step-107-542

![image](https://user-images.githubusercontent.com/16190118/142969390-142df7ec-6040-40c1-9a02-17d43f5de05e.png)

Pull Request resolved: https://github.com/pytorch/audio/pull/2026

Reviewed By: hwangjeff, nateanl

Differential Revision: D32634926

Pulled By: mthrok

fbshipit-source-id: 30d6349a0a2ce174b789a5888b1c8e0544a23a37

Summary:
The previous way of detecting the merger and labels given a commit hash no longer works with ShipIt, as PRs are closed and not merged and are not associated with a commit hash. To work around this, update the script to get the merger (pulled by: ) and PR number from the commit hash message, and then collect labels from the corresponding PR.

Pull Request resolved: https://github.com/pytorch/audio/pull/2030

Reviewed By: mthrok

Differential Revision: D32634870

Pulled By: carolineechen

fbshipit-source-id: a8fcfc5912871d3cca056de43ab25b5d0acb2226

Summary:
Adds beam search decoder for RNN-T implementation ``torchaudio.prototype.RNNT`` that is TorchScript-able and supports both streaming and non-streaming inference.

Pull Request resolved: https://github.com/pytorch/audio/pull/2028

Reviewed By: mthrok

Differential Revision: D32627919

Pulled By: hwangjeff

fbshipit-source-id: aab99e346d6514a3207a9fb69d4b42978b4cdbbd

Summary:
- Remove unnecessary content list
- Remove legacy description

Pull Request resolved: https://github.com/pytorch/audio/pull/2029

Reviewed By: carolineechen

Differential Revision: D32629917

Pulled By: mthrok

fbshipit-source-id: bc9a9366c681bcf8b74907c2a6459c73fb6a7424

Summary:
The sox_effects test in `concurrent.future.ThreadPoolExecutor` started failing since couple of days. While investigate this, skipping the test.

Pull Request resolved: https://github.com/pytorch/audio/pull/2025

Reviewed By: nateanl

Differential Revision: D32615933

Pulled By: mthrok

fbshipit-source-id: 4f7301c0d3c0d11f687011e42e06d9c87ce4197f

Summary:
Allow users to use `torch.cfloat` dtype input for MVDR module. It internally convert the spectrogram into `torch.cdouble` and output the tensor with the original dtype of the spectrogram.

Pull Request resolved: https://github.com/pytorch/audio/pull/2024

Reviewed By: carolineechen

Differential Revision: D32594051

Pulled By: nateanl

fbshipit-source-id: e32609ccdc881b36300d579c90daba41c9234b46

Summary:
Fix minor typo in docs.

Pull Request resolved: https://github.com/pytorch/audio/pull/2012

Reviewed By: nateanl

Differential Revision: D32562618

Pulled By: mthrok

fbshipit-source-id: 79262a14d9b10381249602a63f400232031abaa2

Summary:
Division first, multiplication second. This helps avoid the value overflow issue. It also helps the ``stv_evd`` solution pass the gradient check.

Pull Request resolved: https://github.com/pytorch/audio/pull/2004

Reviewed By: mthrok

Differential Revision: D32539827

Pulled By: nateanl

fbshipit-source-id: 70a386608324bb6e1b1c7238c78d403698590f22

Summary:
With the introduction of tutorials, the turn around time for doc build
has become longer. By default, the tutorial is not built but SPHINXOPT=-W
treats it as error.

This commit disable the option for the local build while keeping it
for the CI.

Pull Request resolved: https://github.com/pytorch/audio/pull/2013

Reviewed By: carolineechen

Differential Revision: D32538952

Pulled By: mthrok

fbshipit-source-id: eae4ffd87100dff466f91abfe26a82aa702d605a

Summary:
Ref: https://github.com/pytorch/audio/issues/1993

Pull Request resolved: https://github.com/pytorch/audio/pull/2010

Reviewed By: mthrok

Differential Revision: D32539511

Pulled By: nateanl

fbshipit-source-id: e99be963123cbc039d79bdb514450b7e8f5a84fc

Summary: Pull Request resolved: https://github.com/pytorch/audio/pull/2018

Reviewed By: hwangjeff, mthrok

Differential Revision: D32540498

Pulled By: nateanl

fbshipit-source-id: f104fbf84c0f489906b6555f21f931b60fedb59e

Summary:
Adds streaming-capable recurrent neural network transducer (RNN-T) model that uses Emformer for its transcription network. Includes two factory functions — one that allows for building a custom model, and one that builds a preconfigured base model.

Pull Request resolved: https://github.com/pytorch/audio/pull/2003

Reviewed By: nateanl

Differential Revision: D32440879

Pulled By: hwangjeff

fbshipit-source-id: 601cb1de368427f25e3b7d120e185960595d2360

Co-authored-by: Facebook Community Bot <6422482+facebook-github-bot@users.noreply.github.com>

Summary:
Pull Request resolved: https://github.com/pytorch/audio/pull/2015

as titled

Reviewed By: hwangjeff, mthrok

Differential Revision: D32495691

fbshipit-source-id: 60d8a2337585e3147f24ca9f0b6518e30cd9134a

- Add link to index page on left
- Package Reference -> API Reference
- Update description.

* Refactor tutorial organization

* Merge tutorial subdirectoris under to examples/gallery/tutorials
* Do not use index.rst generated by Sphinx-gallery
* Instead use flat structure so that all the tutorials are listed in left menu
* Use `_assets` dir for artifacts of tutorials

It turned out that generated tutorials can embed the audio if the following conditions are met.
This commit changes how audio samples are shown in tutorials so that they become playable in doc.

1. There is only one `IPython.display.Audio` call in a cell
2. An object of `IPython.display.Audio` is the last object interpreter receives in the cell
3. Audio format is `wav`
   (`flac` can be embedded as well, but browsers (Chrome/Safari) won't play it)

Ref: https://stackoverflow.com/a/33109647