Adds a woman saying "good boy", I found that 0.9 weight and 20 steps will give good results but yours may vary. Expression tends to be a seductive smile and messes up sometimes when the focus is on more than just the face.
Using Fix Text or Texta at weight 1.0 as well as ExpressiveH at 0.5 will give better results.
Best tags to use are: "BubbleGB, woman with speech bubble saying good boy"
This was my first LoRA so in the future I will train my models better, I also have plans for more speech bubbles with predetermined text. If you have any ideas or input for future models please let me know.