An Arrhenius plot uses the absolute temperature (in Kelvin usually), or rather its reciprocal 1/T.
The other axis needs to take some reference value that divides the measures before you take a log, usually in base 10.
I don't expect straight lines for G' nor G" on an Arrhenius plot. Only at extreme values of G". G" tends to show a maximum, not a line, around the glass transition temperature. It's even a standard way to sniff where Tg is. It's also a way to check if the polycondensation is finished.
But once you figured out what the curves do, over some limited domains the plot is linear and you can model the values with different Arrhenius expressions. After noticing the domains on the curves, you best determine the Arrhenius constants from the data, not the curves.