Cseti's General motion LoRA trained on 32 frames for improved consistency

Model Description

I'm glad to share with you my latest experiment, a basic camera motion LoRA trained with 32-frames on an Animatediff v2 model. The training process utilized Kijai's AnimateDiff Motion Director ComfyUI nodes. This LoRA has been tested exclusively in ComfyUI with Kosinkadink's AnimateDiff-Evolved nodes, however it may be compatible with other AnimateDiff implementations.

How to use

• copy the motion lora files into your "ComfyUI\models\animatediff_motion_lora" folder

• Use the "Load Animatediff LoRA" node

• set the context_length parameter to 32 and context_overlap to 8 (it might work with other values but this is how I tested it)

• example workflow: https://civitai.com/articles/6626

Settings and recommendations

• Use it with strength: 0.8

• This motion LoRA is finetuned on a v2 motion model. It works also with AnimateLCM but don't work with v3 models

• It works with SD 1.5 checkpoints

• Animatediff v3 adapter LoRA is recommended regardless it is a v2 model.

• If you want more motion try increasing the scale multival or lower the strength of the v3 adapter LoRA

• Try playing with the strength of

  • motion LoRA

  • scale multival

  • v3_adapter lora

  like increasing the scale multival and lowering the lora strength.

• I recommend using IP-Adapter, however it reduces the effect of the mLORAs, so play with the strength. "Ease in out" weight type is a good starting point.