Work in progress!
- python (>= 3.9)
- mpi4py
- numpy
- scipy
- jax
- jaxlib
- flax
Add nn_eph to your PYTHONPATH and run the examples as:
mpirun python holstein_1d.py > holstein_1d.out
mpirun python holstein_2d.py > holstein_2d.out
Note that all sampling parameters are defined per process, so stochastic error should decrease if you use more processes for the same parameters.