Have you confirmed that the topaz training command succeeds outside CryoSPARC?
What is the output of the command
/gpfs/share/software/tools/anaconda/3-5.2.0/bin/python -V
?
Given
I wonder whether topaz received a SIGBUS signal, which may be logged in the worker computer’s system log. Please can you
-
ask your sysadmin to check for related system log entries around 2024-01-31 14:08:47
-
try a clone of the training job were you reduce some resource-related settings from their defaults, such as
- Number of parallel processes: 2
- Number of CPUs: 2
as suggested under Topaz Preprocessing very slow