The pairlist is pre-computed, which takes about 40 minutes for the ANI2x dataset (the largest dataset in the collection). We can use the DataLoader logic to prepare these in parallel on multiple CPUs.
This is unfortunately not as simple as I initially anticipated, since we are also removing self energies in the same pass. I will expand this PR with more information as soon as I have a better understanding of the DataLoader logic.
Todos
Notable points that this PR has either accomplished or will accomplish.
Description
The pairlist is pre-computed, which takes about 40 minutes for the ANI2x dataset (the largest dataset in the collection). We can use the DataLoader logic to prepare these in parallel on multiple CPUs.
This is unfortunately not as simple as I initially anticipated, since we are also removing self energies in the same pass. I will expand this PR with more information as soon as I have a better understanding of the DataLoader logic.
Todos
Notable points that this PR has either accomplished or will accomplish.
Questions
Status