Machine Learning

Welcome! The Azure Machine Learning team invites you to share and vote for features to help you build, manage, and deploy custom machine learning models.

Have a technical question, or want to learn more? Please visit our documentation, MSDN forum or StackOverflow.


  1. Increase arbitrarily small metrics document limit (3000b)

    Remove the 3000b limit for metrics:

    azureml._restclient.exceptions.ServiceException: ServiceException:

    Code: 400
    
    Message: (ValidationError) Metric Document is too large, 3195b is larger than the max metric size 3000b.
    Details:
    Metric Document is too large, 3195b is larger than the max metric size 3000b.

    This is from running .log_table with only 5 integers (and associated short strings)

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  2. 1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  3. 1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  4. Need Token for words with Polish Language

    I would like to make an application that process polish sentence.

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  5. Allow to modify the bounding boxes during the review of labeled data

    We are currently working on an object identification project. We need to label a several number of different objects on one single image.
    Currently, we can only choose to accept/reject the entire image when reviewing the labeled data even there is sometimes only one mis-labeled object.
    We'd like to be able to adjust the bounding box when it is mis-labeled instead of rejecting everything else.

    6 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  6. Create pipelines inside other pipelines in the designer

    In the designer I would like to create 'high level' pipelines, in which the nodes are other pipelines. This is a standard in ETL tools.
    For example, I would like to create a 'main' pipeline with 3 nodes (data prep, model training, scoring) and clicking in one of these nodes I find a detailed pipeline. Moreover, inside a pipiline (e.g., model training) I would like to have n pipilines (e.g., an additional data preparation specific for the n models I want to train).
    In my opinion, without this possibility, the pipeline will become very confusing for big projects

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  7. Support python 3.7

    Support running Python 3.7 scripts

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  8. Bring your own Compute Cluster

    It would be great if there was a supported method to allow support for a local cluster or compute device. (such as Azure Stack Edge)

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    1 comment  ·  Flag idea as inappropriate…  ·  Admin →
  9. Request for capability to download the Bias and Feature Weights from the Trained Model

    Request for capability to download the Bias and Feature Weights from the Trained Model

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  10. Automatic Shutdown for Compute VMs

    It would be great if we could have the ability to automatically shut down Compute VM's at a specified time - similar to what you can do with VM's provisioned outside of the AML Portal.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  11. Document azure devops machine learning tasks

    To create azure devops pipelines implementing MLOps(e.g. [1], you have created pipeline tasks which can be executed in the devops pipeline, e.g ms-air-aiagility.vss-services-azureml.azureml-restApi-task.MLPublishedPipelineRestAPITask@0 and others. There clearly mirror the CLI and API, but they provide no documentation. They should.

    1: https://github.com/microsoft/MLOpsPython/blob/master/.pipelines/diabetes_regression-ci.yml

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  12. Deploy the model to Azure Functions in the Designer

    There is a request to deploy a designer-generated model to Azure Functions. Are there any plans for future development?

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  13. Can I use custom definition script in algorithms modules?

    So we can use Net# in this environment. Just like studio.azureml.net.

    2 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  14. Passthrough authentication for Azure ML Datastores and Datasets

    Currently the Datastores and Datasets in Azure ML Workspace can be used by all members of the Workspace. Only the Datastore/Dataset creator needs access to the source data. This can be problematic in cases when deciding the Workspace scope in enterprise environments.

    Please add option to use passthrough authentication in Datastores & Datasets for more granular control of access rights.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  15. How to save ContainerEvents for model-deployed ACI

    Can't save ContainerEvent to Log Analytics because Azure ML can't create logging-enabled containers. It would be better to be able to get the following logs.

    https://docs.microsoft.com/en-us/azure/container-instances/container-instances-log-analytics#view-events

    3 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  16. Visualize support for Designer's Train Model module

    Ml Studio (classic) was able to visualize the result, but Designer doesn't seem to be able to do it. To view the model parameters and feature weights, it's better to support Visualize.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  17. Please replace the UserVoice forum software used here for posting feedback

    Please replace the UserVoice forum software used here for posting feedback.

    It is user-unfriendly, a torture to work with, not maintained, lacks documentation and can likely never be fixed. It seems that it's well past the point where improvements can be made to this system.

    It's beyond me to understand why the new Microsoft (which just organized an amazingly successful MSBuild 2020) allows itself to be associated with such a poor user experience.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  18. az ml model (CLI) deploy/package without "--model" argument

    With Python SDK, you can both deploy and package models passing an empty list as argument of "models".

    Like:


    package = Model.package(ws, [], inference_config)
    package.wait_for_creation(show_output=True)

    Why can't we do the same thing with CLI?

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  19. Automated ML UI - Featurisation Setting Setting greyed out for Time Series

    I can't select the featurisation setting under automated ml from the UI for Time Series Forecasting but can for classification and regression.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
  20. In the Australia East Region, "Create New" fails with the "Set up compute target" of the Designer feature.

    Create New is fixed to Standard Dv2 family by default. In the Austoralia East region, the creation fails because the quota for dv2 family is 0.
    I want you to be able to choose the virtual machine size in the future.

    1 vote
    Vote
    Sign in
    (thinking…)
    Sign in with: Microsoft
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    0 comments  ·  Flag idea as inappropriate…  ·  Admin →
← Previous 1 3 4 5 8 9
  • Don't see your idea?

Machine Learning

Categories

Feedback and Knowledge Base