"Fantasy Impressions" is a LoRA that combines the subtleties of artistic illustration, vivid fantasy motifs and unique painting styles. This tool is ideal for creating stunning female portraits, unusual creatures, as well as fantastic scenes with exquisite detail.
This LORA can be used with or without trigger words to enhance the effect. To use WITHOUT TRIGGERS, it is necessary to increase the LORA weight to 1 or slightly higher. But for me, it seems the best result is using trigger words. Try it and experiment.
This LORA was created by combining several of my unreleased Loras and previously published ones.:
WLOP_Illustrations_FLUX https://civarchive.com/models/785576/wlopillustrationsflux
Anime Niji Style FLUX https://civarchive.com/models/838306/anime-niji-style-flux
and exclusive for TensorArt LORA Agents of Mayhem https://tensor.art/models/773788828805787685/Agents-of-Mayhem-1
Civitai (https://civarchive.com/models/785576/wlopillustrationsflux)
WLOP_Illustrations_FLUX - v1.0 | Flux LoRA | Civitai
LORA allows you to apply the Wang Ling Artist style (WLOP) to your generations models. The optimal weight is from 0.8 to 0.9, sometimes up to 1
Description
Version from SDXL, works great with LORA weights from 0.7 to 0.8, works on Juggernaut, WOW XL models
FAQ
Comments (14)
That's a great set of styles! But can't you expand the range of LORA's strength for the generations on the site? The current range of 0.7..0.8 narrows the possibilities for experimentation.
@Neuronomicon Good afternoon! Thanks! But I didn't quite understand what you were talking about. If you explain in a little more detail, maybe I can do something about it. I'm not sure if I got it right, but if you're referring to the power of LORA when generating online on CIVITAI, I can correct that.
And I've already adjusted this by setting the value from 0.7 to 1.0 If you need more, I can increase it, but then use the LORA weight at your discretion.
@speach1sdef178 Thank you very much! Ideally limits should be made 0.0..2.0 (in my opinion). Or even with negative values (useful range depends on something in the LORA properties, I'm not very knowledgeable). Sometimes I experiment with weird combinations of LORA's. Sometimes I think I just need a pinch of some style for a perfect picture. But I realize that sometimes the lack of limits can break someone's result when accidentally using a bad value.
@Neuronomicon Try it if it's good enough - ok, if not, write to me, I'll see what else I can do)))
Love this LoRA, I've been struggling for ink-style LoRAs since Flashfiring decided to remove his LoRAs, but this creates great styles
@Stu42 i'm glad that you like it. thank you
nice
Cant be installed under InvokeAI:
Unable to determine model type for FLUX_anime_Special Ink-drawing mode_merge_24_medium_3_00001_.safetensors
I love this LoRA, any plan to do it for ZIT?
Thanks for your words, I'm glad you like it. I thinking about version for ZIT, let's see :)
@Wheeman, add version for ZIT
@speach1sdef178 Wow, that was fast! Thanks a lot, I will test it tonight <3 <3
@Wheeman I hope you will like it :) have fun :)