Hey,
I am wondering if there are any plans to implement a feature to restart failed jobs, similar to what is possible with RELION?
It can happen from time to time that a job fails in the later iterations due to insufficient memory/VRAM or simply just because something else causes a system crash or networking timeout. A simple resume from iteration X could save quite some time (especially if one runs heavy 2DC, high-res ab-initio, RMBC etc.).