Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Learn from the best! Meet the four finalists headed to the FINALS of the Power BI Dataviz World Championships! Register now

Reply
lcarlone
New Member

Query errors after training model

Hi all,

 

I find impossible to work with Power BI Machine Learning feature! I will try to explain...

 

I loaded a table from a SQL Database into a Power BI Workspace Dataflow. I trained a ML model from that table and now I have the model trained correctly.

From this point I can't do anything else. I can't modify the original query, nor add another table, nor even change the name of the table because when I try to save in Power Query editor I get a warning which says that some queries contains errors. But if I save without modifying anything I get no errors. I can't understand what's wrong!!!

 

Please help me!

3 REPLIES 3
Anonymous
Not applicable

Hi @lcarlone,

Can you please share some more detailed information for your scenario? They will help us clarify your situations and test to troubleshoot.

How to Get Your Question Answered Quickly  

Regards,

Xiaoxin Sheng

Ok, I will try to retrace the steps!

I created a Workspace (premium). In the workspace I created a Dataflow (Immobili)

lcarlone_0-1638182484187.png

 

In the dataflow I have 3 tables with data from a SQL Database via a Data Gateway

lcarlone_1-1638182484195.png

 

On the 3 tables I trained 3 different Machine learning models

lcarlone_2-1638182484214.png

 

The model training automatically creates 2 tables (training and testing data) for each model in the dataflow and the regression reports in the workspace.

Since this point all worked correctly.

Now I would like to add one more table in the dataflow with new data from a different SQL DB on which I would like to apply the trained models.

I choose “Add table” from the dataflow, insert the connection parameters and the query and I get this screen with the data loaded correctly.

lcarlone_3-1638182484256.png

 

I click Save & Close and I get this

lcarlone_4-1638182484257.png

 

for a very long time (nearly 30 minutes). After waiting that long, I get this error that prevents saving the dataflow with the new table.

lcarlone_5-1638182484258.png

 

The query with error is not everytime the same!

Trying to save another time it succeded, but again after a lot of time.

What am I missing?

Anonymous
Not applicable

HI @lcarlone,

According to your description, it seems like the training models setting and work well before you do modify to insert other data sources into current contents, right? If that is the case, I'd like to get some more detailed information to confirm the root cause of this issue.

Gateway related issues:

Did these tables reachable from the data gateway that is used by current training modes? 

What version of the gateway are you working on? If you are working with the old versions, please update them and try again.

Data records issues:

According to the snapshots, the error message seems related to validating/committing changes failed. Did any particular in these newly added data tables? (e.g. security setting on the data source, data structures, wrong data include in tables)

The issue with specific contents:

Any change/update on your data source tables recently? 

If you create a new ML model with these new tables, did the same issue appear? 

Server/service-related issue:

If you change on other models that you created in the past, can you reproduce this issue on the old models?

Regards,

Xiaoxin Sheng

Helpful resources

Announcements
Join our Fabric User Panel

Join our Fabric User Panel

Share feedback directly with Fabric product managers, participate in targeted research studies and influence the Fabric roadmap.

February Power BI Update Carousel

Power BI Monthly Update - February 2026

Check out the February 2026 Power BI update to learn about new features.

FabCon Atlanta 2026 carousel

FabCon Atlanta 2026

Join us at FabCon Atlanta, March 16-20, for the ultimate Fabric, Power BI, AI and SQL community-led event. Save $200 with code FABCOMM.