Update README.md

README.md

@@ -46,6 +46,17 @@ Title: Stand As One (Stadium Chant Edition)
 
 ---
 
+---
+
+Despite not being *exactly* good for image models.
+
+It can be fairly used for *image recognition* for VL captioner models. To increase their pool of styles and detection of details , and the layout and style makes it easier for their segmentation.
+
+---
+
+
+
+
 Why not woman? Sh'mon, you have tons of women for everything. we need to balance out genders.
 
 So this is a collection of lineart spreasheets ai-generated.
@@ -66,8 +77,3 @@ It's mainly oriented for research and study.
 
 ---
 
-Despite not being *exactly* good for image models.
-
-It can be fairly used for *image recognition* for VL captioner models. To increase their pool of styles and detection of details , and the layout and style makes it easier for their segmentation.
-
----