BP - Not much. Ongoing discussion about SPICE. Some talk about new postdoc doing resubmission, but that may not be the best idea.
PB – DD will discuss SPICE with JC.
DD – Just spoke with JC. Near-term I’ll give PE direction on how to filter dataset for hdf5 creation, and he’ll upload a new version to zenodo. Medium term (not urgent) they’re thinking of offloading ML dataset stuff on OpenFF or MolSSI postdoc from NIH grant. BP, thoughts?
BP – Not sure it’ll be handy to learn about current QCF/QCP when we’re going to change that in a few months with the next release. MY postdoc is only learning the next branch.
DD – Once the next branch is out, would that postdoc be useful for datasets?
BP – To some extent, they’ll mostly be doing new feature work but some dogfooding could help them as a developer.
DD – Gotcha. JC and MS will contact you soon. I’ll spearhead resubmission. OpenFF isn’t obligated to do SPICE stuff unless the new postdoc comes online.
JW – Right, eventually openFF would get to this, but this would be in at least a few months. If it’s needed sooner then someone else would need to provide the resources.
PB – Will new postdoc help wtih porting qcsubmit?
BP – That’s on my plate, it’s about 75% done.
DD + JW – Thanks, BP!
BP – Ran backups over the weekend. Still waiting to hear about college of science grant for new equipment.
Infrastructure advances
JW – ForceBalance 1.9.4 is broken. If you use it you’ll get errors. If you somehow don’t get errors, stop using it.
PB – Yeah, I got those erros so I’m using 1.9.3.
Throughput status
OpenFF PEPCONF OptimizationDataset v1.0
PB – 4025 → 4039 (2 weeks), 156 errored out, may be they are large jobs and taking time to error out and also there are not many workers. Not a priority now, can set up workers and leave.
DD – I’ll check on this when I have time.
User issues/questions
Science needs
CC – May add protein datasets after I see how the first round of protein benchmarks does.
Add Comment