This was... frustrating to prompt and weigh. I think 0.6 is the correct weight, but I could be wrong and if the armor colors don't seem right, you can always increase the multipliers.
Part of that was due to me experimenting with using the exact same seed instead of just using a random one for each image, saving me some of the hassle.
But yes, this was a surprise to me too. Realizing that I could potentially make a LoRA for Kuuga, I decided to throw caution to the wind and just do it. Again, it was a rather frustrating process to get the images right without wonky weapons or faces. This is a very experimental LoRA, so jump into this with huge precautions.
UPDATE: Though, it could just be Anything V5 being a bother, it might work with other models just fine.