Decoupling capacitor for h bridge circuit

I'm making an h bridge with mosfets and the ir2110 mosfet driver. What should the capacitance be for the decoupling capacitors on the 12V power supply that charges the bootstrap capacitors (between VCC and COM pins). I'm using 2 of these ics (full h bridge) so should I double the capacitance or not? Also, should I place the capacitors next to the power supply or next to the ic, or it doesnt matter? I've got both 2.2 uf and 100uf electrolytic capacitors, would that be fine?

Also, to decouple the Vdd of the ic, which will be powered with an arduino, should I add a capacitor besides the 0.1uf ceramic capacitor I will be adding? Would something like a 100 uf electrolyctic be useful or not? (I wont be always using the diver at high frequencies, if that has something to do with the capacitor value)

Datasheet for the IR2110:

The 12V rail decoupling needs to be at least an order of magnitude larger than the bootstrap capacitance, otherwise the 12V rail will be crowbarred every cycle. If the bootstrap capacitance is large consider a low
value series resistor for the bootstrap diode(s) to limit the current pulses to a few amps at most (ie 5 to 10 ohms).

Decoupling capacitors should be ceramic, and right next to the chip they are decoupling - as close as
possible, and then a bit closer (!). Minimize stray inductance between decoupling cap and the chip.

Breadboards are not really suitable for this class of circuitry I'm afraid, but will probably work at a pinch
if the layout is tight - normally large value ceramic caps are only available in SMT packages.

So I can't use an electrolyctic in paralel with a ceramic to decouple my bootstrap power supply? I've seen this layout a couple of times, where people put the very low value ceramic for high freequencies and the electrolyctic for low ones.

Why do you think you need a large bootstrap capacitor? You know this kind of bootstrapped
driver requires PWM drive?