Fresh from YoinkoorLab, trained on the base motion modules using hand picked high quality data.
This motion module uses the newly released 1.5 v2 motion module as a base.
We have achieved a state of the art (SOTA) text-to-seggs motion module.
What this version offers:
Better generalizability, Insane details, good motion for realistic models.
This mm is tuned to use for yoinkooranime
-->Join the discord<<- We have an AnimateDiff bot, you can try AD without installing anything.
Whats coming: Making motion better on anime models.
( ͡° ͜ʖ ͡°)( ͡° ͜ʖ ͡°)( ͡° ͜ʖ ͡°)( ͡° ͜ʖ ͡°)
-->Join the discord<<- We have an AnimateDiff bot, you can try AD without installing anything.
( ͡° ͜ʖ ͡°)( ͡° ͜ʖ ͡°)( ͡° ͜ʖ ͡°)( ͡° ͜ʖ ͡°)
Many of our users were using vertical aspect ratios. And so,
Introducing: GODMODE
The best motion module for your *ahem* pov vertical clips.
Use whichever animeDiff UI exention that has updated to support v2. Comfyui is the best.
Do NOT use anime models, it wont have much motion. its because the new base cant produce it for some reason.
If you're not getting good results, change the sampler, or checkpoint (Movement is very reliant on the diffusion model). Also, use simple VAEs that don't affect the saturation that much.
Animediff NSFW discord server:
https://discord.gg/FpDbSkAe7m<----- join to share feedback and learn from others.
You should use 512 by 512 or 768 by 512 (landscape) at 16 context length.
Use your favourite SD keywords like: Detailed face, beautiful face etc.
girl, naked, pov, sex, cowgirl position,penis, vaginal sex,face, vertical
watermark, signature, bad quality, worst quality