I'm using the Relion 3.1 GUI to run the Bayesian Polishing step in Training mode, and its taking much longer (12 hours and still running) than indicated in the 3.1 Tutorial, which says "Using 16 threads in parallel, this job took 1 hour and 15 minutes on our computer." Training mode seems to work only with 1 MPI Process, so multiple threads should speed it up. The Tutorial talks of 4000 particles, and I am running this with 5000 particles which come from 1279 micrographs (only a few particles per micrograph) but I never see more that 1 core in use on a 28-core system even though I have specified 28 threads (--j 28). It's finished preparing alignment data and is up to the 15th iteration. Is there some reason the job is not using more than 1 core?
Thanks for any pointers.
James Conway
U. Pittsburgh
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
|