Expected partitioned base model
We have been using partitioned models for a while. Recently (after upgrading to dataiku 9.0, but not directly after) we are started getting a lot of errors saying,
* Internal error, caused by: NestedServletException: Handler processing failed; nested exception is java.lang.AssertionError: Expected partitioned base model, caused by: AssertionError: Expected partitioned base model
* HTTP code: , type: com.dataiku.dip.exceptions.DSSInternalErrorException
when we are training our model. What is really confusing is that we train a few hundred partitions, more than half will typically succeed, while the others fail with this error.
Does anyone know why this happens? And/or any ways to mitigate the issue? I haven't found much information on the error so far, so any input is appreciated.
Answers
-
Hi,
We'll be replying to your support ticket as soon as possible.