Back to our regularly scheduled program, my comment above is an important issue. If your motor says 550V, it was made before anyone who made VFDs was even a glint in his Father's eye. That means that unless someone rebuilt it to modern winding standards, the motor insulation will have a very hard time surviving the ravages of the PWM pulses that make a VFD work. So one thing that I HIGHLY recommend is that you install a "sine wave filter" on the output of the VFD going to the motor. Not just a load reactor, you need the full blown, and most expensive, version. Failure to do so will result in a surprisingly short motor life once the VFD is implemented. Another alternative is to just bite the bullet and replace that motor with a new 575V "inverter duty" motor now, save yourself the trouble of doing two startups.
As to wiring, the output from drive to motor is no different in terms of size; 125% of FLC from the NEC tables (use the 575V table). If you are using STEEL conduit, no sharing of that conduit with ANY other cables and you are religious about grounding it, you don't need VFD cable, but don't use THHN or any other thermoplastic insulation, use XLPE insulation, as found in RHH or XHHW. Anything outside of the above conditions, you MUST use shielded VFD cable, and in this case, ground the shields at BOTH ends.
The input wiring is nothing special, but there is a different rule for it in the NEC; 430.122 saying that the wire is sized at 125% of the VFD maximum input amps, not the motor amps. That's because people can use VFDs that are bigger than the motor, so the NEC addressed the possibility of someone "turning up the dial" because the drive CAN, then smoking the input wires. So bottom line, technically you need the data sheet on that drive to pick the wire size.
I recommend line reactors ahead of VFDs, it's cheap insurance because it helps to protect the drive from spikes and surges on the line side, most of which are caused by things out of you control.
