I've tried running a couple of newer containers to see if the issue was fixed in newer versions of the community edition, however, I'm not able to get the containers started.  

containers.intersystems.com/intersystems/irishealth-ml-community:2021.2.0.651.0

containers.intersystems.com/intersystems/irishealth-ml-community:2022.1.0.209.0

I get the following in the /store/iconfig/mgr/messages.log file when trying to start either instance.

07/07/22-20:58:55:213 (412) 0 [Utility.Event] ISC_DATA_DIRECTORY=/store/iconfig/
07/07/22-20:58:55:224 (412) 0 [Utility.Event] Executing /usr/irissys/bin/iris qlist
07/07/22-20:58:55:329 (412) 0 [Utility.Event] Versions are different
07/07/22-20:58:55:329 (412) 0 [Utility.Event] Source /usr/irissys/: 2021.2.0.651.0com
07/07/22-20:58:55:329 (412) 0 [Utility.Event] Destination /store/iconfig/: 2021.1.0.215.3com
07/07/22-20:58:55:351 (412) 0 [Utility.Event] Executing /usr/irissys/bin/iris version
07/07/22-20:58:55:354 (412) 0 [Utility.Event] Performing upgrade of /store/iconfig/ from /usr/irissys/
07/07/22-20:58:55:354 (412) 0 [Utility.Event] Upgrading from version 2021.1.0.215.3com to 2021.2.0.651.0com
07/07/22-20:58:55:354 (412) 0 [Utility.Event] Executing /usr/irissys/bin/iris qlist
07/07/22-20:58:55:380 (412) 0 [Utility.Event] Executing /usr/irissys/bin/irisdb -s/store/iconfig/mgr/ -cd
07/07/22-20:58:55:470 (412) 0 [Utility.Event] Executing rm -f shuttest.txt
07/07/22-20:58:55:540 (412) 3 [Utility.Event] Error during Upgrade,ERROR #5001: Instance IRIS was not shut down correctly, may have been forced
 

Just wondering if this issue has been fixed in a new community edition release?

@Yuri Marx 
The Predict-Maternal-Risk app linked above worked great!  No errors, and I was able to get all the way through the entire set of ML queries to build, train, validate, and predict risk using the training and validation data.