Parameters for regulating hybridised generalised functions are not supported in the scf_abacus.yaml file

Exx_hybrid__alpha is set when generating high-precision data labels in the deepks init step. But it prompts the following in err.iter.

"following ars are not in the default list: ['exx_hybrid__alpha']and would be discarded"

This causes the generated data labels to be less precise. Is there any way to fix this? Or can I add this parameter when using the deepks model later to get accurate results?