arbox / machine-learning-with-ruby Goto Github PK

https://github.com/somaticio/tensorflow.rb says "NOTE: this doesn't work anymore" atop the README. guessing you will want to remove from your list?

https://github.com/somaticio/tensorflow.rb

cheers

Dear ML in Ruby Gist contributors,

especially @MohawkJohn, @stympy, @irfansharif, @gbuesing, @jjgh, @v0dro, @giuse,

I'm the curator of the RubyNLP list and I'm glad to welcome you all to the new community effort to collect and curate ML related resources in Ruby.

After the initial talk to @gbuesing I've decided to make the next step and convert the gist to an Awesome List.

Why it could be useful? In my opinion (please correct me) we all need a central place for information about special applications (like ML, NLP, DataScience) where Ruby can compete with e.g. Python.

@SciRuby makes great work on developing tools but there're plenty of other resources which should be collected, described, tested and represented among Rubyists and Developers using other technologies.

I can take the role of the list curator but I definitely lack all you expertise and don't have the wide horizon to find and review all the possible projects :)

So if you find this project useful please contribute, spread the word and use it :)

My plan is to support the lists for:
NLP,
ML,
Data Science,
Ruby Interoperability.

These lists are not programming projects which propel Ruby over night. They are not a new killer app like Rails. But they are still very useful as documentation and a stable basis for a steady and long term development.

Thank your for your participation! Any ideas?

Hi I'm a member and admin of the Mastodon group but I would like both of these added to the Community heading if possilbe. Thanks!

• X Ruby AI group
• Mastodon Ruby AI and Data group

Looks like a bunch of interesting gems got released recently that may be worth a look and adding to this list.

https://ankane.org/new-ml-gems

My team is writing a recommender system using collaborative filtering, and data scientists at our institution have suggest this is the algorithm we need to use. Is there one that could be included here? And if not, would you be interested in listing ours if we find we must invent our own? Thanks!

Hello.

I vote for ruby-dnn.
ruby-dnn is yet another library of deep learning in Ruby.
https://github.com/unagiootoro/ruby-dnn

Let's look into the examples directory.
https://github.com/unagiootoro/ruby-dnn/tree/master/examples
You will find that ruby-dnn's API is more Rubyish than any other deep learning library for Ruby. ruby-dnn even has a dcgan example.

The problem with ruby-dnn is the lack of GPU support. The author use Windows and cumo cannot be built. So it will take a very long time to actually run DCGAN example. (To be honest, I haven't tried it.) Ruby-dnn project appears to be a personal project and may not be intended for production.

Despite these shortcomings, ruby-dnn is a valuable project for implementing deep learning in pure Ruby.

If you feel it's too early to list it, please wait until it is mature.

Thank you.