DeepState slow on linux?

[vidarsumo]

I've been using deepstate on a local Windows machine for a year now using R (through modeltime.gluonts package). It's fast and everything works fine.

Yesterday I created a docker to run my deepstate model. To my surprice it was 5-6 times slower. The only thing I found while searching online was a similar question on SO (https://stackoverflow.com/questions/67533645/forecast-based-on-gluonts-deepstate-model-is-slow-when-it-is-run-on-linux-os) but no answer.

Has anyone experienced this and is there something I can do to speed it up?

[vidarsumo]

For those of you who stumble up on this question then it seems that this had nothing to do with Linux or the gluonts package. There as a bug in the modeltime.gluonts package that caused this.