Trained at 8575 steps 332 images batch size of 4 on Civitai with Natural Language and captioning that has been checked by hand and edited on 70% of the BLIP-2 Captions.

WD-14 tagging is a custom dictionary with manual appending.

1. Small ?????? and flat ????? are out weighed by large cup sizes (Yeah Pun)
2. Use negative -"Large " when using positive +"small " or +flat chest
3. Use Natural Language with VIT tagging only where needed.