Showing results for 
Search instead for 
Did you mean: 

Why is the blueprint changed?

Why is the blueprint changed?

When I click the "Copy and Edit" blueprint, it becomes changed.


blueprint1.PNG 1) This is the first blueprint in the Describe menu.



2) This is the changed blueprint, when I clicked 'Copy and Edit' button.

Why is the blueprint changed?








3) I didn't edit the blueprint and it was a default(or recommanded) blueprint, but there was a warning message "Validation for this task is not yet fully supported and training may fail". I thought that this model was completed following a default(or recommanded) blueprint. However, the message is meaning 'Train-Time Image Augmentation' task didn't work in this model, right?


Also, if I edit the blueprint and click "train" button, where can I find a result?



Labels (1)
3 Replies
DataRobot Employee
DataRobot Employee

The first blueprint, shown under Leaderboard > Blueprint tab, is the result of the data preprocessing and parameter settings testing that DataRobot did (what DataRobot chose). When you click to edit, DataRobot displays the blueprint with the branches for other data types, in case you want to modify the blueprint to accept those types too...jen

Thank you for kind reply.


You mean that the first picture is final blueprint DataRobot choose?

But I can't understand why are there numeric, categorical, date and text variable blueprint on the second picture?

Because I uploaded only image data.

DataRobot Employee
DataRobot Employee

Hi @cookie_yamyam . Composable ML is a very powerful tool--great to see you are working with it! In terms of why the blueprint changes, the documentation explains:


When a blueprint is generated, it can contain branches for data types that are not present in the current project. Unused branches are pruned (ignored) in that case. These branches are included in the copied blueprint, as they were part of the original (before the pruning) and they may be needed for future projects. For that reason, they are available and visible inside the blueprint editor.


Because of this, it is possible that one of the "pruned" branches has a validation issue. Blueprints with validation warnings are saved and can be trained, despite the warnings. 


When you train a blueprint, the resulting model is added to the Leaderboard.


Please let me know if you have any more questions or if I missed any...jen