AnimateDiff

⚠️ This module requires at least 16GB RAM ⚠️

Informations

• Module : AnimateDiff
• Function : Generate video from a prompt and a negative prompt using AnimateLCM or ByteDance/AnimateDiff-Lightning with Stable Diffusion Models
• Input(s) : Prompt, negative prompt
• Output(s) : Video
• HF model page :
  • emilianJR/epiCRealism
  • SG161222/Realistic_Vision_V3.0_VAE
  • digiplay/AbsoluteReality_v1.8.1
  • runwayml/stable-diffusion-v1-5
  • nitrosocke/Ghibli-Diffusion

Help

• Usage :
  • (optional) Modify the settings to use another model, modify the number of frames to generate or change dimensions of the outputs
  • Fill the prompt with what you want to see in your output video
  • Fill the negative prompt with what you DO NOT want to see in your output video
  • Click the Generate button
  • After generation, generated video is displayed in the Generated video field.
• Models :
  • You could place huggingface.co or civitai.com Stable diffusion based safetensors models in the directory /biniou/models/Stable Diffusion. Restart Biniou to see them in the models list.

Examples

• Prompt : A space rocket with trails of smoke behind it launching into space from the desert, 4k, high resolution
• Negative prompt : bad quality, worst quality, low resolution
• Output :

5718415873_1708402131_8747816.mp4