Hello everyone!

This is a long-awaited email for sure.  Today, we released mlpack 4!
This is a significant refactoring of the library.  In addition to the
usual collection of new features, there are some major changes that I
will try to summarize quickly here:

 * Boost is no longer a dependency at all, which significantly improves
   compilation time and memory usage.

 * mlpack is header-only, so no linking with `-lmlpack` is needed.

 * mlpack now depends on cereal for serialization.

 * The neural network toolkit has been fully overhauled to use
   inheritance, instead of the boost::visitor paradigm, and is much
   simpler as a result.

The full list of changes can be seen in the changelog:

https://github.com/mlpack/mlpack/blob/4.0.0/HISTORY.md

but be warned, it is *very* long, since this is the first release in two
years. :)

As always, this release would not be possible without the hard work of
everyone in the community, including all of our Summer of Code
contributors!  As of today, the number of contributors is over
220---that's a lot, given that I remember when it was just one single
research lab at Georgia Tech!

I don't have any other cool statistics for us today, but I think we will
have them in the future.  mlpack 4 represents a complete overhaul of
mlpack and a significant step towards the vision we laid out in our
vision document: https://www.mlpack.org/papers/vision.pdf
As the months go by in the future, I'm sure we'll take more and more
steps to make that vision a reality, and I am quite excited to be a part
of that process. :)

Anyway, as usual, if you have any issues with the new version of mlpack,
please do not hesitate to open an issue on Github:

https://github.com/mlpack/mlpack/

or, you can send a chat message to the mlpack channel on Matrix:
#mlpack:matrix.org.  (See https://www.mlpack.org/community.html.)

Have a great rest of your day!  I know that mine is off to a great start
with this announcement. :)

Ryan

-- 
Ryan Curtin    | "It's... complicated."
[email protected] |  - Hobie Doyle
_______________________________________________
mlpack mailing list -- [email protected]
To unsubscribe send an email to [email protected]
%(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s

Reply via email to