v1(v0.5) good learning results.I have experienced poor learning with SD1, and SDXL's character learning ability is amazing..It is possible to express character with the first tag. It is possible to express basic costume with the second tag.

pennsylvania-v1.x (Mist Train Girls)

I guess I finished learning at step 250. The training data is 5 images .I'll prepare step 500 v1 for research purposes.

Files with SHA256: 13c406ba84c8319023fcb46c6ac5b0dc441887f7938bc417fd1226851752ae65