colorized ink drawings - minimally cartoon or anime leaningYou don't always need to use any trigger words with the base SDXL model, but you may need to prompt for "ink art," or "line art drawing" (the former contributing more to b/w imagery) especially when using fine tune models. For one reason or other, most fine tunes initially overwhelm the ink aspect of this LoRA's style, but starting your prompt with "Ink art of..." should get you on track. Complex prompts and references to photography can likewise steer away from the drawing style.As with all of my art-style LoRAs, of paramount importance for me is versatility. You can render a wide variety of subject matter with the style trained in this LoRA. Simple object illustrations, portraits, sci-fi, fantasy ... I expect you can prompt for anything within reason.The look is of observational realist (not cartoon or anime) comic illustration, with colorized strong black line work and intricate cross-hatching.This is still an early version 1.1 - so let me know if something is deficient or could be improved and I'll look into refining the training. And if you make something great, I'd love to see it!I'm happy to share my work for free, just as the base model SDXL came free to us all. But if you felt compelled to offer some material form of thanks, you can <a target="_blank" rel="ugc" href="https://www.buymeacoffee.com/AdamDArt">buy me a coffee</a>, or you can buy my (darn near unreadable, I warn you) short story '<a target="_blank" rel="ugc" href="https://www.amazon.com/dp/B0BT8W58LT">Willem &amp; Ellene</a>'.

Ink Art XL

stronger linework, more coherent and versatile

InkArtXL_1.2.safetensors

v1.2

v1.1

v1.0

colorized ink drawings - minimally cartoon or anime leaningEdits are coming fast for this LoRA! This second update brings a few less artifacts, better coloring (to my taste at least) and, most importantly to myself, more coherent and flexible image making.You don't always need to use any trigger words with the base SDXL model, but you may need to prompt for "ink art," especially when using fine tune models. For one reason or other, most fine tunes initially overwhelm the ink aspect of this LoRA's style, but starting your prompt with "Ink art of..." should get you on track. Complex prompts and references to photography can likewise steer away from the drawing style.As with all of my art-style LoRAs, of paramount importance for me is versatility. You can render a wide variety of subject matter with the style trained in this LoRA. Simple object illustrations, portraits, sci-fi, fantasy ... I expect you can prompt for anything within reason.The look is of observational realist (not cartoon or anime) comic illustration, with colorized strong black line work and intricate cross-hatching.This is still an early version 1.1 - so let me know if something is deficient or could be improved and I'll look into refining the training. And if you make something great, I'd love to see it!I'm happy to share my work for free, just as the base model SDXL came free to us all. But if you felt compelled to offer some material form of thanks, you can <a target="_blank" rel="ugc" href="https://www.buymeacoffee.com/AdamDArt">buy me a coffee</a>, or you can buy my (darn near unreadable, I warn you) short story '<a target="_blank" rel="ugc" href="https://www.amazon.com/dp/B0BT8W58LT">Willem &amp; Ellene</a>'.

Files with SHA256: 68d31547010a1d20ccc2a8a615d4b44856dac305b7455f996b91b4f0e903a20c