Type | |
Stats | 26 2 |
Reviews | (1) |
Published | May 19, 2025 |
Base Model | |
Training | Epochs: 25 |
Usage Tips | Clip Skip: 1 Strength: 1 |
Trigger Words | A dipnmbrx style drawing A dipnmbrx style sketch A dipnmbrx style illustration A dipnmbrx style engraving |
Hash | AutoV2 37532E4C6F |
I post models on multiple sites, including exclusive models. For new model releases, bounties, and more, follow me on: X (Twitter) or Instagram. Also, check out LoRA Lore for information on how to train your own models. More info & links: www.darkinfinity.art
This model card is for pre-release versions of new LoRAs that I plan to post soon. I will post bounties for the best AI artists to help generate showcase images that show off the new model's full potential.
If you are participating in a bounty, please do not post your bounty images to this model card or leave reviews for the model. This is not an attempt to solicit reviews or encourage image posts on this model. To ensure the Civitai TOS is followed (and to avoid duplicate models), the version of the LoRA posted here will be deleted once the bounty is over. The LoRA will be reuploaded on its own model card with the bounty-winning images as its showcase.
Any current bounties will be listed below. Follow the link to the Civitai bounty page to learn more.
Baroque Macabre Style LoRA BOUNTY
Baroque Macabre Style is a high-contrast, engraving LoRA that emulates intricate linework and dense sketch textures. Common motifs include twisted trees, animal skulls, gothic architecture, and starry night skies. Ideal for generating monochrome illustrations with dark fantasy or occult themes, this LoRA responds well to tokens related to bones, skulls, demons, monsters, and other things that go 'bump' in the night. The model is also a big fan of fancy scrollwork, twisting patterns, and decorative knotwork.
Use the trigger word "dipnmbrx" along with a style, as in "a dipnmbrx style illustration"
Many styles work, and the trigger and LoRA together make everything from baroque-style engravings to dark fantasy photorealistic. It has been tested with sketches, illustrations, paintings, drawings, engravings, etchings, photographs, and line art.
Velvet & Shadow LoRA BETA
Note: The new version of this model has been fully developed and is available on Mage.Space as the concept Velvet & Shadow concept. Some of its output:
New model I've been working on but not quite happy with yet. Feedback would be appreciated and help me complete it. Though it's based on dark manga-style artwork, it has some other imagery mixed in. Images are supposed to be dark and shadowy, but you should also be able to prompt for interesting lighting, high contrast, and a few other artistic effects.
Use LoRA trigger phrase "a dinctrlns style photograph" or "a dinctrlns style drawing" -- it should work for other mediums as well.
Lots of words related to darkness and shadows should enhance the LoRAs effect.
Standard Flux settings should work -- Euler Simple, 20 steps, Flux.1 Dev base model.
Classic Oil Painting LoRA BOUNTY
This is a model I made for personal use, but it turned out well, so I'm going to try to post it. It has an Easter egg you may find while generating, so have fun with this one.
Use LoRA trigger phrase "a dilpntng style oil painting"
Standard Flux settings should work -- Euler Simple, 20 steps, Flux.1 Dev base model.
The bounty has ended, and this model is now published as Academic Old‑Master Oils.
Folk Art Style LoRA BOUNTY
This style is part of a new series I'm creating that produces art reminiscent of folk art or fairytale illustrations. This one reminds me of early Nordic artwork, but that may just be me.
Use LoRA trigger word "dijldflklr"
Standard Flux settings should work -- Euler Simple, 20 steps, Flux.1 Dev base model.
The bounty has ended, and this model is now published as Ink & Lore: Forgotten Tales (or will be soon).
Concept Art Sketch Style LoRA BOUNTY
This style was trained on concept art from video games and with the trigger word “digrngbrsh” but like most Flux LoRAs, the trigger word may not be necessary as long as you include associated style tokens like “a sketch of” when you use the LoRA.
Images should have a rough sketch appearance, similar to the examples on the model card.
Standard Flux settings should work -- Euler Simple, 20 steps, Flux.1 Dev base model.
The bounty has ended, and this model is now published as Sketch Pad: a Concept Art LoRA.