Run many adaptive.Learners on many cores (>10k) using MPI.
https://github.com/basnijholt/adaptive-scheduler
latest