Lisbeth Zwerger (born 1954)
Lisbeth Zwerger is an award-winning Austrian illustrator renowned for her delicate, ethereal reinterpretations of classic children’s literature. Based in Vienna, she is widely considered one of the most influential contemporary artists in the genre.
Signature Style: Masterful use of watercolors, muted palettes, and expansive negative space. Her work is characterized by a dreamlike, slightly melancholic atmosphere and elegant, whimsical figures.
Key Works: Illustrated iconic editions of The Nutcracker, Alice in Wonderland, The Wizard of Oz, and the fairy tales of the Brothers Grimm and Hans Christian Andersen.
Major Honor: Awarded the Hans Christian Andersen Medal in 1990, the highest international recognition for children’s book illustration.
NOTE : Base model : Z Image Turbo.
Showcase images were created with ComfyUI and are in PNG, include metadata & workflow.
IMPORTANT NOTE: This model is a tribute to the artist and not an attempt to copy or reproduce his works.
VERY IMPORTANT NOTE :Do not like my models, do not give me buzz, do not follow me, otherwise you will be reported to Civitai for attempted corruption
Description
Steps : 8
Sampler : heunpp2
Scheduler : Normal
Weighting : around 0.75
FAQ
Comments (8)
Easily the best lora creator on this site! Do you have any plans on creating (or transfer) loras for klein 9b? It would allow for more creative control over the final result, it tends to follow base sketches from photoshop way better and better prompt adherence, and it would be possible to stack more of your amazing loras together for more interesting experiments. From a designer/artist perspective it would be a godsend!
I've done some LoRa tests with Klein9B, but so far I'm not convinced it's worth it. I find this model inferior to ZIT and ZIB.
Thank you for your feedback on my models.
@Adel_AI Probably some wrong or different setting? The results seems to be quite good and the general consensus seems to be that it trains better than zit and zib https://docs.bfl.ai/flux_2/flux2_klein_training_example
I'll also leave this reddit post just in case you might be interested, it is aimed at characters bur probably there is some knowledge to apply in style-learning also: https://www.reddit.com/r/StableDiffusion/comments/1ri65uz/basic_guide_to_creating_character_loras_for_klein/
@Buuuh12 It's possible, honestly I haven't spent much time on it. I plan to come back to it to better understand this model.
@Buuuh12 Over time, I've finalized my own approach to training LoRa models, and it's quite different from what's usually recommended.
Thanks for the links.
@Adel_AI Makes sense, as im finding a lot of contraddictory infos about lora training this days, more an art than a science. But yes the ability to stack a lot of Loras together and having a single Lora compatible with both the base and distilled is a godsend for klein, it will be worth it believe me! As soon as i will have time i'll try training it, if something good comes out (im an absolute noob in that regard) i'll reach out to you and share some knowledge! Just one last question Adel, sorry to bother, are your loras also compatible with ZiB?
@Buuuh12 LoRa ZIt are incompatible with ZIB, they don't have exactly the same architecture.