This Lora training is based on the 768X1024 image, training on ProtogenInfinity checkpoint.
gundamRX78_v10.safetensors
LrGundamRX78v1.safetensors