LoRa dedicated for giving a character sheet with 3 views (front-side-back)

This model was only trained with character sheets with 3 views on the character, but it may produce 2 views, 4 views and even some anomalies/artefacts (espacially with TI or LoRas of characters). To correct them, I suggest you to do those things:

• Widen your image to ~25-30% of the heigth (ex: for 512 of heigth, set the width between 512*1.25 = 640 or 512*1.30 = 666

• add simple background at the end of your prompt (with weight if necessary)

• You can also increase the weight of the model to something like 1.2 or 1.3 (you can go higher but it may alterate the image)

If the character is cropped to the knees or higher, try adding to your prompt a description of the pants or shoes the character should wear.

If you manage to have some great results or if you want to suggest some improvement for the model, you can send me reviews or feedbacks in the comment.