Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Sebastian Deutsch
    @sebastiandeutsch
    hi
    Riley
    @riley-usagi
    Hey
    Sebastian Deutsch
    @sebastiandeutsch
    What's the best way to run a branch (in a fork) on CircleCI? I've tried master but it doesn't work:
    Unable to find image 'nethsix/ruby-tensorflow-ubuntu:0.0.1.a' locally
    0.0.1.a: Pulling from nethsix/ruby-tensorflow-ubuntu
    @nethsix any special configuration?
    Christian Hansen
    @chrhansen
    @sebastiandeutsch If you haven't looked here perhaps that'll give you some suggestions: https://github.com/somaticio/tensorflow.rb/pull/42/files on how to configure circle.yml
    Sebastian Deutsch
    @sebastiandeutsch
    I've build the docker image and /bin/bash into it. When I run the specs I get a strange error: symbol lookup error: /usr/local/rvm/rubies/ruby-2.2.4/lib/ruby/site_ruby/2.2.0/x86_64-linux/sciruby/Tensorflow.so: undefined symbol: TF_NewTensor
    but on the other hand the bundle goes through on the local dockerfile:
    Don't run Bundler as root. Bundler can ask for sudo if it is needed, and installing your bundle as root will break this application for all non-root users on this machine.
    Fetching gem metadata from https://rubygems.org/
    Fetching version metadata from https://rubygems.org/
    Resolving dependencies...
    Using rake 10.5.0
    Using ast 2.2.0
    Using bundler 1.12.5
    Using coderay 1.1.1
    Using diff-lcs 1.2.5
    Installing docile 1.1.5
    Using google-protoc 3.0.0.alpha.6.0.0
    Installing json 2.0.2 with native extensions
    Using method_source 0.8.2
    Using narray 0.6.1.2
    Using powerpack 0.1.1
    Using slop 3.6.0
    Using rainbow 2.0.0
    Using rspec-support 3.4.1
    Using ruby-progressbar 1.7.5
    Using unicode-display_width 1.0.5
    Installing simplecov-html 0.10.0
    Using tensorflow 0.0.1 from source at .
    Using yard 0.8.7.6
    Using parser 2.3.1.2
    Using pry 0.10.3
    Using rspec-core 3.4.2
    Using rspec-expectations 3.4.0
    Using rspec-mocks 3.4.1
    Installing simplecov 0.12.0
    Using rubocop 0.40.0
    Using rspec 3.4.0
    Installing simplecov-json 0.2
    Installing simplecov-rcov 0.2.3
    Bundle complete! 12 Gemfile dependencies, 29 gems now installed.
    Use bundle show [gemname] to see where a bundled gem is installed.
    while in circle ci it's failing to install json 2.0.2
    with the error
    Gem::InstallError: json requires Ruby version ~> 2.0.
    Installing simplecov-html 0.10.0
    An error occurred while installing json (2.0.2), and Bundler cannot continue.
    Make sure that gem install json -v '2.0.2' succeeds before bundling.
    Sebastian Deutsch
    @sebastiandeutsch
    when I ssh into the circle-ci box the ruby is outdated: ubuntu@box371:~/tensorflow.rb$ ruby --version
    ruby 1.9.3p448 (2013-06-27 revision 41675) [x86_64-linux]
    Riley
    @riley-usagi
    This is the end of the project?
    Arafat
    @Arafatk
    No Riley I am still working on a few things in my free time. I am very busy with assignments and tests.
    Christian Hansen
    @chrhansen
    Hi @riley-usagi I'm also working on the project when I have time, like @Arafatk though I also have a full-time job. Feel free to contribute! :)
    Sebastian Deutsch
    @sebastiandeutsch
    @Arafatk @chrhansen I'd love to see the changes in the image branch in master
    Arafat
    @Arafatk
    @sebastiandeutsch @chrhansen I added the changes in #62 , do take a look.
    Riley
    @riley-usagi
    Does anyone have bookmarked a good articles on the semantic neural networks?
    Christian Hansen
    @chrhansen
    @Arafatk ok I see from #62, that it means that we're using the new protocol-buffer gem.
    Christian Hansen
    @chrhansen

    Hi @/all I've added a couple of issues to the 0.1.0 milestone based on the ROADMAP.md that @sebastiandeutsch drafted and what I could see that we're missing in order to do the Basic Usage tutorial. Feel free to add missing issues, split them up if they're too big, add useful links or discuss implementations in there. Some of the helper functions zeros, random_uniform, etc. I don't think are necessary for 0.1.0, though they're probably not that hard to implement.

    I was thinking that if someone wants to work on an issue that they then assign them selves, so we can structure the work better, help each other and avoid double work. If you've started working on an issue, but are not active on it any longer, it's a good idea to un-assign yourself so someone else can pick it up.