Hi,
I am interested in generating data for 2D materials with Spin Orbit Coupling so I can train ML-Hamiltonian models. I am planning on using OpenMX (which I know is supported for SOC) and FHI-aims. However, in the docs for DeepH-Dock I saw the limitations:
https://deeph-dock.readthedocs.io/en/latest/capabilities/convert/fhi_aims/demo.html#limitations
Spin Support: Only non-spin-polarized and collinear spin-polarized calculations are supported. The interface does not support SOC calculations.
If this is the case:
- Is it difficult to convert FHI-aims SOC-enabled Hamiltonians into DeepH training Hamiltonians?
- Is this a priority on the development roadmap or likely not going to be a feature any time soon?
Any more details you can provide on this would be appreciated as I would really like to train DeepH models using SOC-enabled FHI-aims data if possible.
Hi,
I am interested in generating data for 2D materials with Spin Orbit Coupling so I can train ML-Hamiltonian models. I am planning on using OpenMX (which I know is supported for SOC) and FHI-aims. However, in the docs for DeepH-Dock I saw the limitations:
https://deeph-dock.readthedocs.io/en/latest/capabilities/convert/fhi_aims/demo.html#limitations
If this is the case:
Any more details you can provide on this would be appreciated as I would really like to train DeepH models using SOC-enabled FHI-aims data if possible.