Brian DeCost

@michalsmid I recently ran into this issue on 18.04, IIRC installing `apache2` and `apache2-dev` resolved it. The root cause was some perl XML library. I think only `apache2-dev` is actually...

@PanayotisManganaris thanks for contributing to refactoring! I think it is necessary because especially the dataloader helper functions are sort of outgrowing themselves and have gotten pretty complex. So to summarize...

@knc6 so in the current version for atomwise outputs, the dataset stores the forces and replicated per-atom formation energy as node features, right? So `dgl.batch` handles stacking features for us...

@PanayotisManganaris I think your point about performance is reasonable, but I think the usability is probably worth it. Maybe NamedTuples could be a solution if it ends up being a...

Hi @MiracAydin1, can you please attach the full CIF file and share the versions you're using for at least jarvis-tools and spglib? This looks like a potential issue with https://github.com/usnistgov/jarvis....

a possible workaround might be to parse the structures using pymatgen and convert to jarvis Atoms, since your augmented file seems to have been written by the pymatgen CIF io...

Hi @jiruijin, this looks like an OOM crash during the precomputation of the graphs and line graphs. We've been storing those in memory, but if you have a large dataset...

how many atoms are in a typical supercell for your dataset? The other week a collaborator was having a similar issue on a dataset with ~800 atom supercells, so I...

can you clarify exactly when the resource manager is killing your job? from your screen cap it looks like it's during the dataloader setup before any data is sent to...

to clarify a bit, if GPU memory limit is the issue, you'll get a CUDA out of memory error, and slurm will kill your job because the training script crashed....