Closed ahurtaud closed 2 years ago
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale
.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close
.
/lifecycle stale
Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten
.
Rotten issues close after an additional 30d of inactivity.
If this issue is safe to close now please do so with /close
.
/lifecycle rotten
Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen
.
Mark the issue as fresh with /remove-lifecycle rotten
.
/close
@sesheta: Closing this issue.
Hello, We are planning to run PAD within a kubernetes pod. However we would like to be able to keep the learning across pod deletion/restarts (which is something very common in kubernetes).
I think this is not possible today and everything is recomputed from scratch on every restarts.
Would it be possible to add a persistence within a file or a database so that a restart of PAD wont lose the training?