astrofrog opened this issue 11 years ago · 0 comments
At the moment, each call takes 0.1s, so things are pretty slow when sampling a job with e.g. 10 sub-processes.