Sign In

Furry (Muscular)[Anima]

12.5k

Updated: Oct 21, 2026

style兽人furry malefurry

Recently loaded

Download

1 variant available

bf16 SafeTensor

XiaoYu-ANIMA-V3-step00004750.safetensors

BF16, good balance • 132.31 MB

Verified:

Type
LoRA
Stats

212

208

197

Reviews
Published

Oct 16, 2026

Base Model
Training
Steps: 4,750
Usage Tips
Strength: 0.3
Hash
AutoV2755F094A31

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

ComfyUI_00004_.png

[Anima] XiaoYu Style V3.0

🇨🇳 中文

V3.0 再次更换并优化了训练数据集。相比 V2.0,本版本的训练图片整体质量有所提升,标签质量也得到了一定改善。

与 V2.0 相比,本版本生成的背景光影效果和尘埃粒子没有那么强烈,整体表现更加克制。

推荐 LoRA 权重:0.3–0.8

个人最推荐的权重是 0.3。随着 LoRA 权重提高,角色的体型仍会逐渐变得更加壮硕。不过,权重过高可能会影响画面的整体美感,因此建议根据实际出图效果调整。

🇺🇸 English

V3.0 features another round of dataset replacement and refinement. Compared to V2.0, the training images are of better overall quality, with slightly improved tagging quality as well.

The background lighting, shadows, and dust particles are less pronounced than in V2.0, resulting in a more restrained overall look.

Recommended LoRA weight: 0.3–0.8

My personal recommendation is 0.3. As the LoRA weight increases, characters will still tend to develop a bulkier, more muscular physique. However, excessively high weights may reduce the overall aesthetic quality of the generated images, so adjust the weight according to your desired results.

🇯🇵 日本語

V3.0 では、学習データセットを再度入れ替え、改善しました。V2.0 と比較して、学習画像全体の品質が向上し、タグの品質も若干改善されています。

背景の光や影、そして埃の粒子エフェクトは V2.0 よりも控えめになり、全体的に落ち着いた仕上がりになっています。

推奨 LoRA 強度:0.3~0.8

個人的には 0.3 を最もおすすめします。LoRA の強度を上げるほど、キャラクターの体格は引き続きより大きく、筋肉質になる傾向があります。ただし、強度を上げすぎると画像全体の美しさが損なわれる場合があるため、生成結果を確認しながら調整してください。

🇰🇷 한국어

V3.0에서는 학습 데이터셋을 다시 교체하고 개선했습니다. V2.0과 비교했을 때 학습 이미지의 전반적인 품질이 향상되었으며, 태그 품질도 소폭 개선되었습니다.

배경의 조명과 그림자, 먼지 입자 효과는 V2.0보다 덜 강조되어 전체적으로 한층 절제된 결과물을 보여줍니다.

권장 LoRA 가중치: 0.3–0.8

개인적으로 가장 추천하는 가중치는 0.3입니다. LoRA 가중치를 높일수록 캐릭터의 체격이 더욱 크고 근육질로 표현되는 경향은 여전히 유지됩니다. 다만 가중치가 지나치게 높으면 이미지 전체의 미적 완성도가 떨어질 수 있으므로, 원하는 결과에 맞춰 조절하는 것을 권장합니다.

(anima V2版本我于8月15日 0:00 更换了主lora文件,从原6250步文件变为5000步文件,同时保留了原文件)

“This series is my beginner-test LoRA.”

This series of models is suitable only for large body types and chubby & muscular styles (Chubby & Muscular).

Please use the v2.0 version. The higher the LoRA strength, the larger the character’s body type and the fuller/chubbier the physique.

All example images use a strength of 1 with no additional LoRAs applied.

Setting the strength to 1 may cause changes to the character’s facial features.

It is personally recommended to slightly lower the strength in order to achieve more optimized results across various aspects.

“此系列是我的新手测试lora喵~”