GitLab

* Separate style check from unit test

* TEST flake8 fail

* TEST clang-format fail

* Revert "TEST flake8 fail"

This reverts commit 88d90185cefe294b9a662129a05c2e26a9844fe4.

* Revert "TEST clang-format fail"

This reverts commit 93517d1ccba650a6ccd813268fd73a4fd63b252c.

* Clean up extension build mechanism and extension location

* Add back the switch to depend on external sox

* Remove print

* Fix

* make parameter optional string.

* raise error if incorrect parameter choice.

* test with slaney.

* Add util to generate whitenoise

* Use sinusoid for pitch and revert vad

* Adopt PyTorch's test utility on kaldi compatibility test

 - Adopt PyTorch's test utility on kaldi compatibility test
 - Separate CPU test and GPU test

* Fix device/dtype conversion

1. Docker image used for smoke test is old (No Python 3.8). Updating this to the latest one. build [here](https://app.circleci.com/pipelines/github/pytorch/audio/1763/workflows/56c846a5-acaa-41a7-92f5-46ec66186c61/jobs/44484)
2. Add `latest` tag to the newly built docker so that later we can always use `latest` version and we do not have to update docker image tag manually.

* Add windows binary jobs

* Make wheel package platform-specific and python-version-specific

Currently there is no CUDA device for regular FB's infra, where `torchaudio`'s tests are ran.
Skipping these tests consistently raises a flag so we decided to split the tests into runnable and not-runnable on file-level.

This PR
 - Changes the location of C++ codes from
     - from `torchaudio/torch_sox.cpp` to `torchaudio/csrc/sox.cpp`
     - from `torchaudio/torch_sox.h` to `torchaudio/csrc/sox.h`
 - Changes the location where the resulting library is placed
     - from `_torch_sox.so` (outside of `torchaudio` module)
        to `torchaudio/_torchaudio.so`.

* add slaney normalization.

* add torchscript.

* convert to string for torchscript compatibility.

* flake8.

* use string as default.

* backend change instructions in readme.

* link to installation instructions.

* Enable CUDA tests for Windows

* Add back branch filters

Fix #628

* Make fbank support cuda

* Reduce rtol for kaldi

* fix test

* fix flake8

* Introduce common utility for defining test matrix for device/dtype

* Make resample_waveform support float64

* Mark lfilter related test as xfail when float64

* fix

Co-authored-by: Kunal Bhandari <bkunal@fb.com>

Closes #610 

* Add GPU test

* Run only on master

* Turn on tests when building through conda-build

* skip one failling test on windows.

* Switch to unittest

* Fix conda env activation

* add linux only condition

* Fix librosa installation

* conda-forge channel as a fallback

* Revert "conda-forge channel as a fallback"

This reverts commit 8c15f6638e34792640a3de3e057d9c2995425180.

* Try the pytorch way

* Use conda-forge

* Disable conda-forge

* Install future at setup env

* Move linux unittest CI job tools to dedicated directory

* Restore Linux scripts and move the updated souces to a new subdir

* Add back future

* Remove comments
Co-authored-by: Vincent Quenneville-Belair <vincentqb@gmail.com>
Co-authored-by: moto <855818+mthrok@users.noreply.github.com>

* Bust cache weekly

* Rename directory for utility

* Show 20 slowest tests

* fixup! Show 20 slowest tests

* Add one fbank compatibility test

* Update util

Signed-off-by: Bhargav Kathivarapu <bhargavkathivarapu31@gmail.com>

* Add README for test

* Apply review feedback

* Add description to test_dataloader

* lfilter: add an optional arg `clamp`

to give users control on the clamping behavior within the range [-1, 1],
which was hardcoded before.

Fixes #596

* doc string formatting

* lint

* doc string again
Co-authored-by: Vincent QB <vincentqb@users.noreply.github.com>

* Allow zero momentum in Griffin-Lim

* Skip momentum in Griffin-Lim if zero