Skip to content

README.md

Latest commit

 

History

History
131 lines (76 loc) · 3.99 KB

README.md

File metadata and controls

131 lines (76 loc) · 3.99 KB

🔥 Image Generation Workflows

Welcome to my collection of advanced workflows for image generation using ComfyUI — designed for high-quality, realistic, and anime-style outputs with smart randomization features.

Screenshot

✨ Features

  • 🔀 Random Prompt Combination Generator

  • 🔀 Dynamic LoRA Loader

  • ⬆️ Upscale images up to 4x resolution

  • 📺 Supports high-resolution generation up to 1080p

  • 🧐 Realistic and consistent generations

  • 🌺 Optimized sampler and model configurations for best results

🧠 IMPORTANT TIP !!

  • All Of The Nodes Are Color Coded For Easier Understanding:

  • Orange Color Boxes Are Your Special Notes To How Each Of The Nodes Work, Read Them If You Dont Know What To Do.

  • Do not Change Anything In The Black Color Nodes They Are Functional Nodes And Tinkering Will Not Do Anything Good.

  • All The Purple Nodes Are Your Wildcard Processors, Change Them To Customise The Different Variations.

  • AlL The Green Nodes Are Realted To The Positive Prompt.

  • All The Red Nodes Are Related To The Negative Prompt.

  • The Yellow Color Coded Group Is Your Dynamic LoRA Loader.

  • The Magenta Nodes Are Bypassed Nodes Select them And Press Ctrl+B To Activate Them Again.

  • The Pink Node Is Your Latent Image Node Change It To Select The Batch Amount And Also The Image Resolution.

  • The Upscaling Node And The Vae Decode Node Are Pretty Self Explanatory.

🖼️ Screenshots

Below are sample screenshots of the workflows in action: Screenshot

🧬 The Wilrcard Processor To Generate Random Or Customised Variations

Screenshot

🏞️ Realistic/Anime Image Generation Workflow

Screenshot

🖌️ Dynamic LoRA Loader:

Screenshot

⚙️ Requirements

  • Python 3.12 (Recommended)

🔼 Recommended Upscale Models

  • 4x-AnimeSharp (for anime-style images)
  • RealESRGAN_x4plus (for general upscaling)
  • 4x-UltraSharp  (for high detail enhancement)

📦 Custom Nodes (Place inside ComfyUI/custom_nodes/)

All node-specific dependencies are mentioned in their respective repos/folders. ⚠️ Use the same virtual environment (venv) or Python version that runs ComfyUI to avoid errors.

  • ComfyUI Manager

    git clone https://github.com/Comfy-Org/ComfyUI-Manager.git custom_nodes/ComfyUI-Manager

  • ComfyUI Essentials

    git clone https://github.com/cubiq/ComfyUI_essentials.git custom_nodes/ComfyUI_essentials

  • WAS Node Suite

    git clone https://github.com/WASasquatch/was-node-suite-comfyui.git custom_nodes/was-node-suite-comfyui

  • Flow Nodes

    git clone https://github.com/gitmylo/FlowNodes.git custom_nodes/FlowNodes

  • ComfyUI-Custom-Scripts & ComfyUI-Impact-Pack From The ComfyUi-Manager

⚙️ Installation

  • Just Drag And Drop TheVariatedWorkflow.json In Your Comfyui Graph.
  • If Found Any Missing Nodes Just Install Them With Either The ComfyUI Manager Or The Link Given Above.

🧠 Resources

For models, VAEs, LoRAs, Lycoris, embeddings, wildcards, and ControlNets:

🧪 Recommended Models per Workflow

  • 🎨 Variated Anime Scene Generations

    • Works best with: Illustrious or Pony-based models (Some SD1.5 models also perform well)

  • 📸 Variated Realistic Generations

    • Best with: RealVisXL v5.0 Lightning (BakedVAE) Or Juggernaut XL Or Any Other Well Trained Stable Diffusion Model
    • Advanced option: Flux.1D or Flux.1S (for experienced users) as it might need manual configurations and installations of the Flux libraries