predict_survival_function with baseline_hazards_ estimated with covariates at 0

[aslitoj]

Dear authors of the lifelines library,
I am new to lifelines, and already went through some discussions [https://github.com//issues/543] related to baseline hazards estimated at 0 versus mean for each covariates. I wonder whether predict_survival_function takes into account of baseline hazards while predicting the individuals survival? If so, I would like to know if there is a way to use cox ph model's predict_survival_function with user defined baseline hazards where all covariates are taken as 0 for the baseline? What I wanna have is using cox model fitted on train set, predicting the survival functions of each subjects in test sets.
I would really appreciate any help for this issue.
Regards,