Ahsoka Tano (Star Wars the Clone Wars | Season 7) | LoCon
Ahsoka Tano (Star Wars the Clone Wars | Season 7) | LoCon
- TypeLoCon
- ModelSD 1.5
- Trained WordsAhsokaS7, colored skin,blue eyes,AhsokaS7Outfit, blue shirt, vambraces, fingerless gloves, hip armor, blue skirt, blue leggings, knee boots
- TagLoCon3d ahsoka ahsoka tano animated anime blue eyes blue lightsaber cartoon character colored skin face paint female jedi lightsaber movie orange skin season 7 series character star wars star wars the clone wars western woman
Ahsoka Tano for Star Wars the Clone Wars Season 7. Works best around 1.0. Training done on NAI.
Thank you all for 100 Reviews! I plan to open up to requests for more fun project ideas soon.
This is a LoCon which can be placed in LoRA folder on Automatic1111 1.5+ so make sure to update otherwise place in Lycoris folder.
Trigger: “AhsokaS7, colored skin“ always needed, add “blue eyes” for stability.
Suggestions/notes:
Outfit: “AhsokaS7Outfit, blue shirt, vambraces, fingerless gloves, hip armor, blue skirt, blue leggings, knee boots“
The shirt design can be a bit random along with the “hip armor” due to small inconsistencies in the fanart from the data.
Outfit design bleeds into shirts/dresses a little but adding brackets “(dress)” can help.
Flexible clothing wise but the more words use to describe outfits the less bleeding there will be.
I plan to make Ahsoka’s other versions one by one and eventually mix them into one LoRA if possible once all are made.
I was not really happy with the Ahsoka models on the site as it seemed several of her designs were mixed into one and produced inaccurate designs. I made this with accuracy as the goal and spent the better part of the last month manually editing the images/tags to get a decent result and I think it came out really well. I hope you all enjoy.
Feedback and reviews are always appreciated.
Nerdy training numbers:
Trained on D8Dreambooth trainer
Optimizer: AdawW Dadaptation
Training resolution: 768
Unet LR: 1
Tnec LR: 1
Unet weight decay: 0.016
Tenc weight decay: 0.032
12 Epochs – 8256 Steps
Trained on 344 images using Reg images.