Partitioned Time Series Parallelism
Isaiah
Registered Posts: 5 ✭✭
I am training a partitioned statistical models across 5 partitions on CPU containers which seems to distribute well however it does not seem to be distributing across time series within each partition. Is this a functionality in system or something I should log as an idea? I know I can likely pull this out into code but do not want to lose the visual ml tools if I don't have to.
Dataiku version used: 15
Dataiku version used: 15