v3.6.0

• Removed micromamba, updated installation scripts for Windows,Linux and Mac

v3.5.0

• Added Würstchen support
• Added "model" CLI argument
• Removed gradio deprecated style
• Pytorch 2.0.1

v3.0.0

• Added Stable Diffusion XL support
• Added Controlnet workflows-Canny, MLSD (Line control), Normal, HED, Pose, Depth, Scribble, Segmentation
• Added Pytorch 2.0 support
• Added schedulers - KDPM2DiscreteScheduler,HeunDiscreteScheduler,KDPM2AncestralDiscreteScheduler,DPMSolverSinglestepScheduler

v1.0.0-beta.1

Initial beta version of DiffusionMagic.

• Supports various Stable Diffusion workflows
  • Text to Image
  • Image to Image
  • Image variations
  • Image Inpainting
  • Depth to Image
  • Instruction based image editing
• Supports all stable diffusion Hugging Face models
• Supports Stable diffusion v1 and v2 models, derived models
• Works on Windows/Linux/Mac 64-bit
• Works on CPU,GPU,Apple Silicon M1/M2 hardware
• Supports DEIS scheduler for faster image generation (10 steps)
• Supports 7 different samplers with latest DEIS sampler
• LoRA(Low-Rank Adaptation of Large Language Models) models support (~3 MB size)
• Easy to add new diffuser model by updating stable_diffusion_models.txt
• Low VRAM mode supports GPU with RAM < 4 GB
• Fast model loading
• Supports Attention slicing and VAE slicing
• Simple installation using install.bat/install.sh