Sign In

Female Argonian

29.1k

Not loaded

Download

1 variant available

bf16 SafeTensor

Argonian_V02.safetensors

BF16, good balance • 132.24 MB

Verified:

Type
LoRA
Stats

59

2

2

Generation

Not loaded

Reviews
Published

Oct 9, 2026

Base Model

Anima

Training
Steps: 1,600
Epochs: 8
Usage Tips
Strength: 0.8
Hash
AutoV2
CB8460A21E
default creator card background decoration
Followers - 249

249

Likes - 1199

1.2K

Downloads - 7839

7.8K

Project Odyssey Season 2 Participant

License:

Anima

The 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

AnimaTW_100.png

So where to start. Let's just say that initially I wanted to make high-quality Argonian art using PonyDF, but I came across the fact that PonyDF does not really understand how to draw them. More precisely, she draws them, but there are problems with rendering the texture of the face and scales. So I decided to make this model. To do this, I needed to view about 1000 images on Rule 34. It seems to be nothing complicated, at most two nervous breakdowns😭) But here is the finished model.

As for the weights of the model, then put it as you want, it won't get any worse. Personally, I drew art at a weight of 1.0. Trigger word - Female_Argonian. However, after receiving the picture, make it upscale, since there are problems with rendering faces, but they are leveled by upscale.

The model works both in the PONY version and in two iterations of ANIMA.

At the moment, I recommend using ANIMA V2.

Chronology of versions:
Pony - base model. Work on IL and Pony models. weight 0.6-1.0

V1 Old - first anima model. It works well in portrait models. Use 0.6–0.8 weight for the task. However, there is already a new version V2 for it.

V2 Current (10.10.2026)
Retrained on a cleaned set of about 170 images. The LoRA is more stable than the first Anima version: the snout, scales, claws and body shape hold at 0.75–0.85 without pushing realism tags.

Fixed incorrect tails. The tail now follows the body instead of breaking, disappearing or growing from the wrong place.

Works more willingly with other LoRAs. Outfit, style and scene LoRAs no longer wash the species out as fast as in V1.

Triggers: Argonian, Female_Argonian, Scalie, Lizard. Strength 0.8.

Still breaks into a normal face on hard prompts. Add lizard type face. Do not use 2girls in complex scenes: Anima reads it as two human girls. Describe them as two Argonian / Scalie instead. blue skin and realism still smooth the scales, use scales.