temporarily fixed pygmsh issue that crashed the runner

1 job for !31 with sprint2 in 101 minutes and 8 seconds (queued for 2 seconds)
merge request