Forum Discussion
Edit Model in Service automatical converts to Large Semantic model in Pro Workspace
We recently enabled the preview feature to be able to edit models in the service. This had been working fine for us recently.
However, beginning yesterday, when we went to edit one of our semantic models, it automatically converted the dataset to large semantic model format
After this conversion, the editing page doesn't load and posts an error message that our dataset size has exceeded the allowed size for our pro capacity. The amount is double whatever the current dataset size is. This was a fairly large dataset (809 MB) to begin with.
This semantic model is now unable to refresh, and the report pages will not load for end users.
Has anybody run into this issue and been able to resolve it. All the documentation around Large semantic models is for Premium capacity. There is no documentation that i can find around the Preview feature that indicates this behavior, and was working fine until yesterday.
Thanks
Hi, I am seeing the same issue and its happening on all datasets.
The only way i was able to resolve this was to upload a smaller version of the same model by filtering the data. Then converting it small dataset and Then uploading the full model.
This was not the case a previously as I was able to edit the model without any conversion. Not sure if this will be the intended functionality going forward or just a bug. If its intended then this feature will just become useless to the Pro users that have large models
The service does not allow you to do anything to the model once the size of the model is >1GB on Pro licence.
This is a preview feature things are likely to break. Hopefully this is just a bug they revert it back to the previous experience
5 Replies
I would add that we are also unable to revert the model to the small dataset size - we receive this message everytime we try
Hi, I am seeing the same issue and its happening on all datasets.
The only way i was able to resolve this was to upload a smaller version of the same model by filtering the data. Then converting it small dataset and Then uploading the full model.
This was not the case a previously as I was able to edit the model without any conversion. Not sure if this will be the intended functionality going forward or just a bug. If its intended then this feature will just become useless to the Pro users that have large models
The service does not allow you to do anything to the model once the size of the model is >1GB on Pro licence.
This is a preview feature things are likely to break. Hopefully this is just a bug they revert it back to the previous experienceThis is exactly what I did prior to seeing your comment and it worked. It was very frustrating - I really hope that it is a bug during the preview period
Thats great news. I tried editinng my data model yesterday and it seemed to be resolved. So it seemed like a bug.
Hi pstauduhar ,
Thank you for reaching out to the Microsoft Fabric Community. Also, thank you dkdata for your valuable input. As already mentioned, since this is a preview feature, there might be some limitations. Please refer to the documentation below:
Edit data models in the Power BI service (preview) - Power BI | Microsoft Learn
Additionally, please try these troubleshooting steps which might help resolve the issue. First, check your Workspace Settings under Power BI Service > Workspace Settings to see if Large Semantic Model is enabled. If it is, try disabling it and republishing your dataset. You can also attempt to republish the dataset from Power BI Desktop under a different name to keep it as a standard semantic model. If the issue persists, try publishing the dataset to a different Pro workspace to see if the conversion is workspace-specific.
As per Microsoft’s official documentation, Large Semantic Models are meant for Premium capacity, so if this is an intentional change, Pro users with large datasets may need to either optimize dataset size or consider Premium capacity.
I hope my suggestions give you good idea, if you need any further assistance, feel free to reach out.
If this post helps, then please give us Kudos and consider Accept it as a solution to help the other members find it more quickly.
Thank you.
