Himi Ryou (Tachinomi Izakaya de Kuudere Downer-kei Joshidaisei ga Tonari ni Kuru Hanashi)/ Anima v1.0
9
64
1
8
Download
1 variant available
fp16 SafeTensor
Himi-San_Anima_v1.safetensors
Half precision, best balance ⢠63.23 MB
Verified: 12 days ago
640 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9
10 1 2 3 4 5 6 7 8 9
10 1 2 3 4 5 6 7 8 9
(9)
Aug 31, 2026

5.7K0 1 2 3 4 5 6 7 8 9.0 1 2 3 4 5 6 7 8 9K
4750 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9
3.3K0 1 2 3 4 5 6 7 8 9.0 1 2 3 4 5 6 7 8 9K
License:
AnimaThe Anima Model is licensed by CircleStone Labs LLC. Copyright CircleStone Labs LLC. IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
Built on NVIDIA Cosmos
Himi Ryou Character LoRA
Bring Himi Ryou from Tachinomi Izakaya de Kuudere Downer-kei Joshidaisei ga Tonari ni Kuru Hanashi to life with this character LoRA trained from manga artwork and a carefully curated dataset, developed using my newer training process for Anima models.
The dataset was carefully selected to capture Himi's distinctive appearance while providing variation in poses, expressions, angles, and settings. Particular attention was given to Ryou's physical appearance and recognizable features, including her short black-and-white hair, blue eyes, round glasses, and cross-shaped earrings, with the goal of preserving her identity while allowing the character to remain flexible across different generations.
Testing Uncolored Manga Panels
Another LoRA intended to test how well an Anima model could learn a character from a small dataset consisting primarily of uncolored manga panels.
Rather than relying on a large collection of fully rendered, colored illustrations, this dataset intentionally makes heavy use of black-and-white manga artwork. This provided an opportunity to test whether the training process could learn Ryou's character identity from line art, shading, and limited visual information while still producing a recognizable character in fully colored generations.
The final model was trained using my updated Anima-model training process, incorporating the lessons learned from previous character LoRAs and this experiment with predominantly uncolored source material.
Features
Character-focused LoRA for Himi Ryou
Carefully curated manga-based dataset
Dataset consisting primarily of uncolored manga panels
Experimental test of learning character identity from black-and-white source material
Focused on preserving Himi's physical appearance and distinctive features
Trained using my updated Anima-model training process
Supports a variety of poses, expressions, and environments
Designed to preserve Ryou's recognizable appearance
Works with different outfits
Manually curated dataset
This model is part of my growing collection of character and outfit LoRAs, created using careful dataset curation, experimentation, and iterative training refinement.
Have fun with Himi-San! š
If you enjoy this model, please consider leaving a rating or sharing your generations. Seeing what everyone creates is one of the best parts of releasing these projects.
Enjoy and Good Luck!
ā Daraco
NOTE: ArmetDaraco | Profile | Arc en Ciel is my official model list. The Bots are scraping these, so if you are finding them on any other site, they are reposted (hopefully with proper credit). So if you want to stay updated and support me in future releases, please follow me on Arc en Ciel!!

