There is a mismatch between the definitions of phi_1 and phi_2 for bilby and lalinference. This can been seen in the bilby/lalinference comparison pages for GW150914 (https://ldas-jobs.ligo.caltech.edu/~cbc/pe/bilby_lalinference_comparison/GW150914/html/Comparison_phi_2.html). The relevant lines of code in bilby are here https://git.ligo.org/lscsoft/bilby/blob/master/bilby/gw/conversion.py#L866.