Deep Learning extension RELEASE thread

pschlunderpschlunder Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, RMResearcher, MemberPosts:96RM Research
edited December 2018 inHelp

Last week we released a新深度学习extensionalong theRapidMiner Studio 9.0update.


Find an overview of the initial feature set inthis knowledge base article.


Please share problems and feedback in this thread along.


A few remarks and known issues up-front:

  • The 64-bit (standard) version of RapidMiner Studio is needed.
  • Currently only CUDA version 9.1 is supported due to extension size concerns. If you are working on a big project and need support for CUDA versions 9.0 or even 8.0, please contact us. Make sure CUDA is properly installed and its library added to your operating systems PATH variable.
  • Non-tabular data currently needs a new object (a tensor). This might be changed in the future.
  • Problems during extension loading might occur when MKL (e.g. from Anaconda) is on your system's PATH variable. If so, please remove the problematic folder (like Anaconda) for the Studio start from the PATH variable.
  • As the extension progresses we'll post learning material along the way. Until then please checkout the operator documentation and provided sample processes.

Find previous discussionshere.

We're gradualy exploring native integrations of requested features. If one isn't available, yet, stay tuned or check out the Keras extension (thread,download).

Happy testing!
Philipp

BalazsBarany Thomas_Ott jczogalla sgenzer

Answers

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, MemberPosts:1,761Unicorn

    @pschlunderreally cool guys. Is this extension supposed to replace the Keras extension or is this a new offshoot of it?

    sgenzer
  • pschlunderpschlunder Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, RMResearcher, MemberPosts:96RM Research

    It is a completely new extension not based on Keras but on the Java libraryDL4J. Eventually all needed functionality from the existing Keras extension should be available as well as additional features.

    Regards,

    Philipp

    sgenzer Thomas_Ott
  • hughesfleming68hughesfleming68 MemberPosts:323Unicorn

    Edit.... I found the explanation about why it might not install. Looking into it.

    Alex

    pschlunder
  • varunm1varunm1 Moderator, MemberPosts:1,207Unicorn
    edited November 2018
    Irrelevant thread. Moved question to keras thread
    New_Rapid.csv 415.7K
    Regards,
    Varun
    https://www.varunmandalapu.com/

    Be Safe. Follow precautions and Maintain Social Distancing

  • varunm1varunm1 Moderator, MemberPosts:1,207Unicorn
    @sgenzerany support for my issue in the previous comment? Also, in this new platform can't we edit once a comment is posted? Thanks
    Regards,
    Varun
    https://www.varunmandalapu.com/

    Be Safe. Follow precautions and Maintain Social Distancing

  • sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM ModeratorPosts:2,959Community Manager
    @varunm1sorry this is out of my KB. CC'ing the guru@pschlunderif he has time to answer...
    varunm1
  • jczogallajczogalla Employee, MemberPosts:144RM Engineering
    @varunm1sorry for being slightly off topic, but you should see a gear symbol on the top right corner of your post that let's you edit it. If you can't see it, maybe@sgenzerhas to look at it.:smile:
    varunm1 sgenzer
  • varunm1varunm1 Moderator, MemberPosts:1,207Unicorn
    @jczogallait works now.
    Regards,
    Varun
    https://www.varunmandalapu.com/

    Be Safe. Follow precautions and Maintain Social Distancing

    sgenzer
  • sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM ModeratorPosts:2,959Community Manager
    yes apologies...still working the kinks out of the permissioning of this new community platform. Please send me a DM if you have issues.

    Scott

    hughesfleming68 jczogalla
Sign InorRegisterto comment.