On the Widowmaker model the file listed under FP16 FILE is the DMD version
LOW CFG AND STEPS
Dropshot merges the aesthetics of comics and manga into a unified hybrid style with bold lines and expressive detail.
I merged my previous models, and an experimental one I just use for leverage on the merge, it doesn't offer much in comparison of others already on here (The experimental model I to merge).
I am trying to get a distinctive style in each model merge.
Description
FAQ
Comments (6)
It is an interesting visual style! Unfortunately, you can only really get one character to appear, with some details specific to your prompt. I think more variety in your training would lead to a greater utility for such a very specific model.
Thanks for the feedback! Just to clarify, this isn’t a trained model — it’s a merge of existing checkpoints. The style comes from the sources I combined rather than from a custom training. That’s why it looks locked into a certain aesthetic, I will be fine tuning it to improve composition, I may have over baked the merge.
I'm sitting here looking at your settings trying to figure out why my images aren't looking quite right when I realize this isn't a lora. 😩
This one is a checkpoint.
What style are you looking to achieve or what problem do you have with your images?
@JetlistiC Thanks, I figured it out. In my defense, I'm kind of high rn. 😁
@KorgX3 Understandable haha