Share via

While creating connection with Custome NER model(Language Studio) and container(Storage) we are getting an authentication error

Akshata Purshottam Palsule 21 Reputation points
2022-02-14T06:03:15.783+00:00

Hi,
We are trying to create Custom NER project using Azure Language studio while crating it ask to connect with Language resource which was success and while connecting with storage account (container) where our data for tagging is present it throws error saying container must have storage blob owner and Language resource must have reader access. We tried changing the access form IAM but facing same issue while connecting. Can you please provide solution on this .

Priority : HIGH

PFA Screenshot.173906-microsoftteams-image-11.png

Foundry Tools
Foundry Tools

Formerly known as Azure AI Services or Azure Cognitive Services is a unified collection of prebuilt AI capabilities within the Microsoft Foundry platform

0 comments No comments

Answer accepted by question author

YutongTie-9091 54,026 Reputation points Moderator
2022-02-14T23:45:50.797+00:00

Hello @Akshata Purshottam Palsule ,

Thanks for reaching out to us for this issue. I just countered the same issue as yours and lucky I have figured out the way. I hope you can solve your issue after follow my steps.

There are total 3 roles you need to assign to your blob account. They are:
owner or contributor, and
storage blob data owner or storage blob data contributor, and
reader

Here I select Contributor, Storage blob data contributor and reader as below.
174291-image.png

To accomplish this, you need to go to your Storage blob, select IAM, and add those 3 roles as below.
Click Add -> Add role assignment
174178-image.png
Select Role first, you need to do it one by one.
Let's start with reader.
Click Reader then click member to go to the next step. -> Select Managed identity under Assign access to. ->Select Members and find your resource. In the window that appears, select your subscription, and Language as the managed identity. You can search for user names in the Select field. Repeat this for all roles.
174282-image.png
174247-image.png
Click Assign. You need to do it for 3 times for every role, then everything should be fine. You should wait for 10mins then there should not be any error.

Hope this helps!

Regards,
Yutong

-Please kindly accept the answer if you feel helpful, your action help the whole community. Thanks.

Was this answer helpful?

0 comments No comments

4 additional answers

Sort by: Most helpful
  1. Akshata Purshottam Palsule 21 Reputation points
    2022-02-17T14:26:45.483+00:00

    Hello @Anonymous

    Just a question can we add more data in container while our model is trained on some data from container

    for eg our model is trained in 100 txt files so can we add new data to attached container and retrain our model ?

    Need you help!

    Was this answer helpful?


  2. Akshata Purshottam Palsule 21 Reputation points
    2022-02-17T09:06:06.52+00:00

    Hey @Anonymous
    We are facing an issue

    175219-image.png

    It says "Cannot destructure property 'documents' of 'e.assets' as it is null"

    Do you know how to resolve it ?

    Was this answer helpful?

    0 comments No comments

  3. Akshata Purshottam Palsule 21 Reputation points
    2022-02-16T05:04:24.92+00:00

    Hey @Anonymous

    Thankyou for your assist!

    Our issue is resolved :)

    Was this answer helpful?


  4. Akshata Purshottam Palsule 21 Reputation points
    2022-02-15T12:48:30.16+00:00

    Hey @Anonymous

    Thankyou for your response.

    Now we are able to connect container in custom NER model but there's one more error while clicking on create project

    Forbidden: Access to the storage account ![174521-screenshot-385.png][1]storage was denied. The Language resource must have at least 'Reader' and 'Storage Blob Data Owner' access to the storage account.

    Can you please guide us through this as well.

    Was this answer helpful?


Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.