Return to course: ComfyUI – Level 2 Stable Diffusion Art

ComfyUI - Level 2

LoRA

What is a LoRA?

How to use LoRA in ComfyUI

Adding 2 or more LoRAs in ComfyUI

Example: Use LoRAs to change the face and add style

Quiz - LoRA in ComfyUI

Textual inversion

What is textual inversion?

Using textual inversion in ComfyUI

Autocomplete for embeddings

Joining two strings in ComfyUI

Quiz - Textual inversion in ComfyUI

Image-to-image

Image-to-image workflow in ComfyUI

Multiple output images in image-to-image

Image variations with image-to-image workflow

Resizing an image before image-to-image

Quiz - Image-to-image in ComfyUI

Basic inpainting

Introduction - Basic inpainting in ComfyUI

Inpainting with a standard model

Inpainting with an inpainting model

Inpainting with an uploaded mask

Quiz - inpainting

Upscaling

Introduction - Upscaling in ComfyUI

Using an AI upscaler in ComfyUI

AI upscaler with a custom upscale factor in ComfyUI

Upscale with High Res Fix in ComfyUI

When to use High Res Fix

ControlNet - Part 1

Using ControlNet in ComfyUI

Copying human pose with ControlNet in ComfyUI

Copying lines and edges with ControlNet in ComfyUI

Copying depth information with ControlNet in ComfyUI

Multiple ControlNet in ComfyUI

Exercise - ControlNet in ComfyUI

Outpainting

Introduction to outpainting

Outpainting with SD 1.5 model

Outpainting with SDXL model

Outpainting with Flux Fill model

AnimateDiiff

Introduction - AnimateDiff (ComfyUI)

Setting up AnimateDiff in ComfyUI

AnimateDiff Prompt Travel

AnimateDiff video-to-video with prompt travel

Useful Tricks

Fast group muter

Aspect ratio preset

Style presets

Using style presets - SDXL

Using style presets - SD 1.5

Using style presets - Flux

Combining conditionings

Combining conditionings - Introduction (ComfyUI)

Prompt Break in ComfyUI (Conditioning Concat)

Conditioning Average and Combine (ComfyUI)

VAE

Quiz – inpainting

What is inpainting? *

It generates completely new images.

It regenerates part of the image.

It converts an image into an painting style.

It improves details of the images.

The number of sampling steps in inpainting depends on the denoising strength. *

True

False

Which node is required in inpainting but not in text-to-image? *

KSampler

VAE Decode

VAE Encode

Load Checkpoint model

You can use the same VAE Encode node for a standard and an inpainting checkpoint model. *

True

False