CapacityError: Unable to provision compute capacity for training. Please reduce the number of jobs and try again.

Question

I am trying to train a AWS Personalisation Model Recipe based Solution Version for the last 2 days but keep getting the below error continuously:
* CapacityError: Unable to provision compute capacity for training. Please reduce the number of jobs and try again.

I have tried the following when training a solution version but with no luck:
* Reduce the size of training data.
* Kick-off Solution Version training at odd hours (early morning, late night). 
 
I am getting this issue in the **ap-south-1** region

Answer

Please submit a ticket to AWS support if you have not already done so. In the meantime, if your solution has HPO enabled, try creating a solution with maxParallelTrainingJobs of no higher than 4.

CapacityError: Unable to provision compute capacity for training. Please reduce the number of jobs and try again.

Relevanter Inhalt