audio

audio copied to clipboard

Adopt :autosummary: to various modules

* torchaudio.compliance.kaldi
* torchaudio.sox_effects
* torchaudio.utils

overall I think this direction is good. a couple suggestions based on the pytorch core docs

• can we increase the column width of the function name (pytorch uses 50%)? currently a lot of the function names are being split into 2-3 lines which is not very readable
• for the subsections, can we also include section headers to more clearly label/separate the sections? right now there's no clear distinction, just a small gap between sections

looks good overall, and i agree with @carolineechen's recommendations. another nit is that the "torchaudio.transforms" just before the bulleted list of sections in the main body seems redundant

another comment -- I think we can also remove the bulleted list of sections right under the main title, since these are already listed in the sidebar and easy accessible for people to see/click. I think it is a bit redundant and may become too cluttered as we add more pipelines/sections, and I think at minimum we should remove the subsections "interface" and "pretrained models" components from the list there.

@mthrok has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

@mthrok has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

Hey @mthrok. You merged this PR, but labels were not properly added. Please add a primary and secondary label (See https://github.com/pytorch/audio/blob/main/.github/process_commit.py)

@mthrok is there also plans to adopt autosummary for the torchaudio, torchaudio.kaldi_io, and torchaudio.backend sections?

@mthrok is there also plans to adopt autosummary for the torchaudio, torchaudio.kaldi_io, and torchaudio.backend sections?

@carolineechen Not at the moment. It is a bit complicated for torchaudio.backend but should be easy for the rest. It's not in my immediate list of things.