DeepEMhancer integration

Is there an environment variable which can be set in the deepEMhancer wrapper script to set a default path to the models? I don’t see one documented. Thanks!


Hi @jmh, if you installed the deepEMhancer models in the default location on the machine hosting cryosparc_master (i.e. /u/cryosparcmaster/.local/share/deepEMhancerModels/production_checkpoints), you can leave the models path param blank in CryoSPARC and deepEMhancer will find the models on its own.

Otherwise, if you’re already using a shell script with the wrapper, you can change the last line of that script to exec deepemhancer --deepLearningModelPath <path to models> $@ and also leave the models path param blank in CryoSPARC.

We’ll add the models path as a settable project-level default in a future update, but in the meantime you can try these workarounds :slight_smile:

1 Like