Version 1.2 Update: intended to improve versatility of faces - it didn't achieve that as much as I'd like. The model does quite well with male faces, but for women, there's a frequent return to very similar, round-faced and heavy-lidded, full-lipped features. It's a nice face, just a little too omnipresent. I intend to leave the model here for now - I think the primary training dataset will need to be revisited. I did at least achieve somewhat better subject-matter versatility. In particular, sci-fi and fantasy themes are coming through much stronger.
Trained using this website's tools - thank you CivitAI! Works great!
Generation preferences: My preferred sampler in ComfyUI is ipndm (for all Flux, not just this LoRA) with scheduler sgm_uniform or beta at 22+ steps. I keep the guidance between 2.8-3.5 but still haven't methodically explored this.
Flux is an interesting model for fine art style training. It has really strong photographic tendencies. And as this LoRA model is a fairly realist (not photorealist, but definitely not too abstract) style, output can sometimes be outright a photo, or too heavily influenced by Flux's photos - for instance in introducing something like bokeh where this traditional style of painting should not render such a thing. Generally, using 'painting' or 'oil painting' somewhere in your prompt will keep it on track.
After training some other illustration models, I expect to revisit this to iterate more on it. The Flux model definitely has more potential to unlock than I have so far.
Description
initial version
FAQ
Comments (6)
Uhh! Instant testing!
Knowing how good your LoRAs are and how well they mix, i'm really hopeful that it will add the soul to Flux it so desperately needs :)
It's a bit less consistent than the Classipeint model for SDXL. But when it's on, it's really lovely and I greatly appreciate the anatomical accuracy and prompt adherence. I think I'll see what it does with my annotated parchment illustration style next ... that one seems a bit further outside Flux's latent space so I'm not so sure it'll be a success.
@eldritchadam It's more responsive to training then i though and it pics up concepts pretty fast. I made a LoRA with something Flux didn't know what to do with and after like only 500 steps it worked. Lower or higher strength has much more impact then on LoRA's in SDXL and combinations are crazy.
Before i used your LoRA it kinda "tried" to make it look like a newspaper but never got there ( guess it could with better prompting ). The variations that i where surprising ( like almost a double exposure effect or it looked like the news-paper got folded vertical and horizontal ). Would have liked to post it but don't wanna float everything with the same thing.
Looking forward to your Illustration LoRA that added so much color and quality. Would like to test it again with my Vanta LoRA, maybe it work's a second time :D
@TijuanaSlumlord Oh I hadn't seen your Vanta LoRA had a Flux version! Downloading now ... I guess I had failed to follow you so that's corrected too
thanks for amazing lora as usual,
we are waiting for the standard comic/illustrations loras bro.
I'll get to them, of course. But this one is closest to my heart so it's paramount. Fortunately, Flux appears to respond very well to the same datasets SDXL did. I'm finding some differences, and working them out in this oil painting model. But I think after this one gets really dialed in, I should be able to churn out my other model styles pretty quickly.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.