Download
2 variants available
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
I'm pleased to present my second version of the lora.
I released it fairly quickly, but it was a considerable amount of work. The lore was created using a dataset of over 100 images in my style, which resembles a standard anime screencap.
First of all, the lighting has been slightly improved in the new version. I made it a bit more saturated. Also, thanks to the larger dataset, it's much easier to generate images with various poses, and the character anatomy has been slightly improved. This has also made it possible to more effectively combine various actions with multiple characters, some examples are in the preview pictures.
Also, generating SWF images has become much easier and higher quality, in the previous version, the characters were too close, regardless of the prompt, but this has now been corrected. Again, examples are above.
I recommend that you DO NOT use positive prompts like score_9, masterpiece etc. or negative prompts, as this can sometimes degrade image quality. In some cases, you might want to try score_7 in a positive prompt. As you can see from my images, I often avoid using positive and negative prompts.
If you want an even more anime look, add an anime screencap and anime coloring.
I'd be very happy if you shared your work, and I hope you all enjoy it.
I love you all, and enjoy :3
Recommended strength: 0.8 - 1.0