Stark outlines comic book style
Now with more space!
I tamed the biggest and oddest of misbehaviors from previous versions of the model - now if you prompt for 'Space' or 'aliens' or similar sci-fi elements, you won't automatically get ten planets hovering about the scene. There may still be some extra planets, and the appearance of space I taught it is rather a colorful abstraction and you may get things on a planet instead of flying in space sometimes, but all-in-all the sci-fi stuff is much more feasible to prompt for.
It still has a tendency to make a main subject a little too large in proportion to other objects in the scene, and the main subject may get repeated in background elements, but the former is generally correctible with more specific prompting and the latter isn't any worse than the base SDXL behavior.
So for now, I'm very satisfied with this style. It's consistent, coherent, and flexible. Let me know where you find it could use improvement and I'll keep it in mind for the next time I'm looking to dive back in.
Description
Now with more outer space!
FAQ
Comments (15)
amazing images! thanks for sharing!
thank you!
Wonderful, thank you!
thanks! If you make something interesting with it I'm always interested to see others' creations
Thanks for provide such a good model! Can I know how can I train my own LoRA? Thanks!
Thank you, friend! I might be able to point you in the right direction. Are you looking to train a style like this? Or something different like a character? And what model are you targeting - 1.5 or SDXL?
in any case, a good starting point is probably this guide, for training using the buzz you accrue on this very site. You hardest part of training is building your dataset. How you go about that will depend a lot on what you want to achieve
Amazing model. It can go very detailed and its also very good with photorealistic checkpoints. Thank you :)
thank you for the kind words!
This is one of my favorite models
appreciated, thanks!
Why does the meta data say this is a SD1.5 version lora?
Not a clue - maybe it was a bug in an older iteration of Kohya when I trained it? It has only ever been an SDXL model.
used to work great but on the latest forge version it no longer works. says it failed to recognize model type
sorry to hear it. I've been taking a bit of a break from the generative AI work lately and have never installed Forge - so unfortunately I have no insight to provide here. I'd search Reddit or Discord servers for Forge support.
Details
Files
EldritchComicsXL1.2.safetensors
Mirrors
EldritchComicsXL1.2.safetensors
EldritchComicsXL12.safetensors
EldritchComicsXL1.2.safetensors
hf_model_13.safetensors
EldritchComicsXL1.2.safetensors
DCcomics_SDXL_lora.safetensors
EldritchComicsXL1.2.safetensors
EldritchComicsXL1.2.safetensors
XLEldritchComics.safetensors
[comic book]EldritchComicsXL1.2.safetensors
EldritchComicsXL1.2.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.

















