Hi,
I don't think it is correct to fit the effective mass using a band structure calculated along a path along the Brillouin zone, e.g. X -> G -> L, because the effective mass is defined along a specific direction of the Brillouin zone, e.g. [100].
Hence you can only fit the effective mass along a certain direction of the Brillouin zone (e.g. G->Z).
Regards,
Daniele.