Global AI and Data Science

Global AI & Data Science

Train, tune and distribute models with generative AI and machine learning capabilities

 View Only

  • 1.  How to create a GPU-powered Jupyter Notebook

    Posted Mon March 16, 2020 12:54 PM
    Hello guys, how can I create a GPU-powered Jupyter Notebook with Python?
    Because I can't find how to create one. I must say that I'm on the Free (Lite?) plan now, but I can't which plan should I buy to be able to create one.
    Can someone help me with this problem?
    Thanks in advance!

    ------------------------------
    Vlad F.S.
    ------------------------------



    #AIandDSSkills
    #AIandDSSkills
    #DataandAILearning


  • 2.  RE: How to create a GPU-powered Jupyter Notebook

    Posted Tue March 17, 2020 09:44 AM
    Hey Vlad,

    You can easily switch between plans when you search the Resource list in IBM Cloud and select Watson Studio. There is a Plan menu where you can easily switch between plans. I hope this helps you. :)

    Tamas

    ------------------------------
    Tamás Stefanovits
    ------------------------------

    Original Message


  • 3.  RE: How to create a GPU-powered Jupyter Notebook

    Posted Wed March 18, 2020 05:37 AM
    Hi

    --
    Lahcene



    Original Message


  • 4.  RE: How to create a GPU-powered Jupyter Notebook

    Posted Tue March 17, 2020 09:44 AM
    Hi,
     
    There is no generally available support for GPU notebooks in Watson Studio public Cloud.
    But, you can leverage Watson Machine Learning GPU-powered training jobs. The Lite tier provides capped capacity to train on k80 GPUs.
    Paid WML accounts unlock v100 GPUs capability.
    This available for WML instances in US (Dallas/us-south) and UK (London/eu-gb)
     
     
     
    Philippe Gregoire
    IBM EcoSystem Advocacy Hub Nice/Europe - CSP & ISV technical enablement - Data Science and IoT




    Original Message


  • 5.  RE: How to create a GPU-powered Jupyter Notebook

    Posted Tue March 17, 2020 09:45 AM
    Hi Vlad,
     
    There is no generally available support for GPU notebooks in Watson Studio public Cloud.
    But, you can leverage Watson Machine Learning GPU-powered training jobs. The Lite tier provides capped capacity to train on k80 GPUs.
    Paid WML accounts unlock v100 GPUs capability.
    This available for WML instances in US (Dallas/us-south) and UK (London/eu-gb)
     
     
     
    Using WML training rather than inlined in the notebook requires a bit more setup, but it also allows for a much better use of expensive GPU resources by using GPU only during training and not while your notebook is idle or being edited.

    Philippe Gregoire
    IBM EcoSystem Advocacy Hub Nice/Europe - CSP & ISV technical enablement - Data Science and IoT


    ------------------------------
    Philippe Gregoire
    IBM
    ------------------------------

    Original Message