David Nicholson

There should be a `config` sub-package with the current `parseconfig` module inside of it. Currently there's a `parse` sub-package -- naming not quite explicit enough. edit: would prefer a sub-package...

core code in `FeatureExtractor` is basically unreadable, in particular the method `_from_file`. All the weird stuff I did with `if x in locals` and `np.concatenate` really obscures what's going on....

Maybe I should look to see if we already have an all-contributors section in the README :facepalm: (we do) -- thank you @mrmundt. (repeating myself from the #website channel in...

Thanks @bradleycolquitt the links are helpful. Looks like there's a PR in progress but it's stalled: https://github.com/scipy/scipy/pull/6852 Any chance you could submit a pull request where you add a subdirectory...

and plot.confusion_matrix So, make a plot module.

Sounds like the question of whether to specify both `myst-nb` and `myst-parser` in `extension` is confusing in general. But we're now at a point on this issue where people are...

@marisbasha suggested writing wrappers for loss functions so that we can easily use them as metrics -- agreed, good idea

@all-contributors add @marisbasha for ideas

I have thought about a couple of ways to do this: 1. Use introspection to literally make a dataclass directly from the training function and its type annotations -- this...

Just for context, they are close to a release candidate for Python 3.11 -- link to checklist here: https://github.com/numba/numba/issues/8304#issuecomment-1480876090