Data Science VM

The Data Science and Deep Learning Virtual Machines are customized VM images on Azure, loaded with data science tools used to build intelligent applications for advanced analytics. The Data Science VM is a customized VM image on Microsoft’s Azure cloud built specifically for data science work. It has many popular data science and other tools pre-installed and pre-configured to jumpstart building intelligent applications for advanced analytics. It is available on Windows Server and on Linux.
The Deep Learning VM is a specially-configured variant of the Data Science VM to make it more straightforward to use GPU-based VM instances for training deep learning models.
More details about the Data Science VMs are available in the Azure Data Science Virtual Machine Documentation. If you have a technical issue, please open a question on the developer forums through Stack Overflow.

  1. Please add top Java Machine Learning frameworks

    The DSVM is currently missing a proper environment to develop enterprise applications for Machine Learning and Deep Learning with Java. Despite of the fact that all the Python frameworks are the best tools for modeling and research, to get ready for production the backend in most case is more suitable to be C++, C# or Java, to address performances, multi-threading and backward interoperability with enterprise systems (distributed or not).
    So I would suggest to add Java frameworks for Machine and Deep Learning like

    - DeepLearning4J (see benchmarks here: https://github.com/deeplearning4j/dl4j-benchmark/blob/master/README.md)
    - Weka: this is a powerful research tool, well known…

    7 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    2 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  2. Keep the image up to date

    I deployed a Win 2016 DSVM yesterday, but have had so many issues with it i don't think it's worth continuing. So far...
    1. Can't connect with RDP due to a missing windows security patch on the image.
    2. My code won't run because python libraries are several versions behind. Tripped up so far by numpy (1.11 installed vs 1.14 current), tensorflow (1.5 installed vs 1.9 current), pandas (0.20 vs 0.23)
    3. Even on NC-series using GPU my tensorflow workload runs half the speed i get on my laptop CPU.
    This VM sounds great in theory but in practice didn't…

    6 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  3. Use JupyterLab by Default

    JupyterLab should be the default configuration when logging into the Jupyter server.

    5 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  4. DSVM improvement - NLTK data download

    I noticed that the nltk python package was installed by default in the DSVM, but the nltk data is not downloaded. The download is not difficult, but time consuming, and it would be excellent if the downloaded data was included in the DSVM package already. Instructions here: https://www.nltk.org/data.html

    4 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  5. Can you give us an idea how often the software packages in the VM image are updated?

    It would be helpful for us to know if we deploy a VM, we can expect that the software would have been updated within the last 30, 60, 90 days, etc.

    4 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    2 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  6. Make Python 3 the default

    Currently, running just "python" on the Windows DSVM starts 2.7. When using this image in scenarios like Azure Batch then this is unexpected and inconvenient as modern code typically relies on Python 3. I know I can "activate py35" (btw why not Python 3.6?), but still. Is there any reason why Python 3 can't be the default? It's time...

    2 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    2 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  7. where & how do I get a gpu???

    Why are there so many permutations... I need a GPU with the Data Science VM... I have to try all sorts of different locations/disk types just to find one. Any advice?

    2 votes
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    2 comments  ·  VM deployment  ·  Flag idea as inappropriate…  ·  Admin →
  8. Why not a Windows 10 VM image

    I would like to know If Its possible for this VM image to be in Windows 10? IS there any restrictions / limitation in doing this? I am asking this just to save some cost on big VMs. Sorry If I am missing anything thats basic.

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    1 comment  ·  VM deployment  ·  Flag idea as inappropriate…  ·  Admin →
  9. Being new to Ubuntu for Data Science workloads I have a few initial thoughts:

    Being new to Ubuntu for Data Science workloads I have a few initial thoughts:

    1) The picture of the Data Science machine seems to indicate that it has a GUI (gnome in this case). However, it does not actually appear to have a GUI unless I'm missing something... which is possible.

    2) Its on Ubuntu version 16. That's pretty old.

    3) I can't get the RDP connection to work out of the box. Maybe I'm doing something wrong?

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Getting started  ·  Flag idea as inappropriate…  ·  Admin →
  10. 1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  11. Add Scala and sbt

    Particularly useful for Spark devlopment

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  12. Bug? : The Studio failed to load. Please refresh this page.

    The "Microsoft Azure Machine Learning Studio" page header loads but no studio loads. Only some text "The Studio failed to load. Please refresh this page."

    Upon refreshing it still fails to load.

    Things I have tried:
    1. Logged out and logged in again.
    2. Created my azure resources in a different location (japan rather than SE asia)
    3. Sending feedback via smiley face widget. This is also not working .

    I know this is a defect or availability issue (not a feature suggestion) but got no where else to go!

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    4 comments  ·  VM deployment  ·  Flag idea as inappropriate…  ·  Admin →
  13. Jupyter Notebooks should be stable on Azure DSVMs/DLVMs

    Azure Data Science VMs and Deep Learning VMs should allow Jupyter Notebooks to run in a stable fashion.

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  14. Update Azure Storage Manager in Ubuntu image or at least remove current out of date version

    I want to use Azure Storage Explorer to access Azure Data Lake. The current version included does not support this nor Azure Files, tiered blob storage. This gives a very bad impression of a very useful tool.

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  15. Interactive remove command

    Include
    alias rm='rm -i'
    by default to ~/.bashrc

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  VM deployment  ·  Flag idea as inappropriate…  ·  Admin →
  16. Keras is missing in the default environments Python 3

    Need to fix the Python modules - Keras is missing from the default environments,

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    1 comment  ·  Getting started  ·  Flag idea as inappropriate…  ·  Admin →
  17. Have Tab on OS X + X2Go + Ubuntu work as intended

    If you are using OS X and X2Go with an Ubuntu DSVM (as described in the Azure DSVM docs), the TAB key does not work.

    It's a known issue for this combination, and short of remapping the keys manually, which is a process, the tab key will not work.

    When indenting code, this can be an issue.

    Windows + X2Go work fine however.

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  18. Jupyterhub set password doesn't show

    Title says it ; I wanted to setup a jupyterhub this morning (the tool is very powerfull for collaboration purposes) ; I could not do it because the "set password" doesn't show anywhere, and the port doesn't answer :(

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
  19. Useful to have script for post-provisioning config

    It would be great to have some script examples for post-provisioning the DSVM with say VS Pro, SSRS

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  VM deployment  ·  Flag idea as inappropriate…  ·  Admin →
  20. Please add top Java Machine Learning frameworks

    The DSVM is currently missing a proper environment to develop enterprise applications for Machine Learning and Deep Learning with Java. Despite of the fact that all the Python frameworks are the best tools for modeling and research, to get ready for production the backend in most case is more suitable to be C++, C# or Java, to address performances, multi-threading and backward interoperability with enterprise systems (distributed or not).
    So I would suggest to add Java frameworks for Machine and Deep Learning like

    - DeepLearning4J (see benchmarks here: https://github.com/deeplearning4j/dl4j-benchmark/blob/master/README.md)
    - Weka: this is a powerful research tool, well known…

    1 vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Tools and applications  ·  Flag idea as inappropriate…  ·  Admin →
← Previous 1
  • Don't see your idea?

Feedback and Knowledge Base