|
--- |
|
license: cc0-1.0 |
|
language: |
|
- en |
|
tags: |
|
- not-for-all-audiences |
|
- text-generation-inference |
|
- biology |
|
- art |
|
- motionlora |
|
--- |
|
introduction:(this is an nsfw motion lora model/Research on motion models using human bioart/trained by motion director) |
|
1.As the CC0 Shared Model Agreement:Creative Commons Zero v1.0 Universal |
|
2.any commercial legal disputes arising from the use of this model shall be borne by yourself. |
|
3.if you install this model that implies that you accept the terms. |
|
--- |
|
*example result has been released on the pixiv(https://www.pixiv.net/users/86329215) by some ai art creators(non-identity work). |
|
|
|
**motion lora model configuration |
|
1.train in 1024x576size 256rank 16framesx6 1500steps |
|
2.six videos input with folder training with 2 types of text config |
|
3.trian prompt:"a highly quality clear best 2.5D anime video of sex nsfw animation,close up in the style of anime and hyper-detailed renderings(Back view)" |
|
4.training model:aamAnyloraAnimeMixAnime_v1.safetensors, inference model:mistoonAnime_v20.safetensors |
|
5.additionally necessary specify parameters：use_offset_noise: true |
|
6.learning_rate/learning_rate_spatial/adam_weight_decay:between 5x-20xoriginal learning_rate=depending on your training set counts |
|
7.Suitable for Animatediff workflow, weight recommended in between 0.5-1 |
