Download
1 variant available
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
✨ My First LoRA - ANYA V1 & V2 Release ✨
This is my debut in LoRA training! I created V1 in just 24 hours — pure enthusiasm and a sleepless night of work.
V1 Details:
Dataset: 126 carefully selected photos
Training: 1000 steps, 8 epochs
Base checkpoint: anima-base-v1.0
Trained resolution: 736x736 (native 736x736)
Recommended LoRA strength: 1.0
V1 turned out very lively and captures the desired style/character well even with such a modest dataset.
V2 — Improved Version:
Dataset: Expanded to 212 images (added more variety in poses, lighting, and angles)
Training: 400 steps (more optimal for the larger dataset)
Resolution: 736x736 (native 768x768)
Base: anima-base-v1.0
V2 is noticeably more stable, holds the style better, has fewer artifacts, and performs better across different prompts.
Recommended Supporting LoRAs for Best Quality: For even better results I highly recommend combining this LoRA with:
Zoda_anima
MIO-anima_b1_v5
rdbt_anima_b1 v0.39.b
These work great together and significantly boost detail, consistency and overall image quality.
Both versions (V1 + V2) are uploaded. Try them both and let me know in the comments which one you prefer! I’d really appreciate any feedback-it will help with future trainings 🔥