It would be nice to have an idea of the scalability of the NQS.
For instance, going from 2 to 16 hidden neurons takes only ~40 seconds more but the energy loss is reduced by almost an order of magnitude.
Prior to this, Issue @ #3 should be solved first.
It would be nice to have an idea of the scalability of the NQS.
For instance, going from 2 to 16 hidden neurons takes only ~40 seconds more but the energy loss is reduced by almost an order of magnitude.
Prior to this, Issue @ #3 should be solved first.