Become a Master of SDXL Training with Kohya SS LoRAs - Combine Power of Automatic and SDXL LoRAs

Intro
Pre-requirements of this tutorial
1-Click installer for Automatic1111 Web UI
How to install Automatic1111 Web UI for SDXL and SD 1.5 models
How to checkout and verify your installed Python version
Which Automatic1111 Web UI command line arguments you need for SDXL
Where to find all command line arguments of Automatic1111 Web UI
Where to download SDXL model files and VAE file
Which folders you need to put model and VAE files
Detailed explanation of what is VAE Variational Autoencoder of Stable Diffusion
How to set your VAE and enable quick VAE selection options in Automatic1111
What does Automatic and None options mean in SD VAE selection
Why you shouldn't use embedded VAE of SD 1.0 model
Correct resolution of SDXL - how to use SDXL
How to install Kohya SS GUI script for SDXL training
What to do if your CMD is not progressing
When you need to use FP16 instead of BF16
How to install Kohya on RunPod or on a Unix system
How to start Kohya GUI after installation
What are Stable Diffusion LoRA and DreamBooth rare token, class token, and more training
How to select SDXL model for LoRA training in Kohya GUI
How to save and load your Kohya SS training configuration
How to use my own used configuration for this tutorial video training
How to prepare your training images for Kohya LoRA or DreamBooth SDXL training
What kind of training images you should use for training
What kind of regularization images you should use? The logic of using ground truth images
What is number of repeating in Kohya SS. Which number you need to pick
Where will be your LoRA checkpoints saved
How to verify your training images dataset properly composed
How to set your generated LoRA file names
Which training parameters you should use for SDXL LoRA training
Why select train batch size 1 and gradient accumulation steps 1
The logic of number of epochs
Detailed explanation of Kohya SS training. What each parameter and option do
Which learning rate for SDXL Kohya LoRA training
Why do I use Adafactor
The rest of training settings
Which Network Rank Dimension you need to select and why
How to fix if you get out of VRAM error - not enough memory
What is Network Alpha of Kohya LoRA
Don't forget Gradient Checkpointing
How to continue training with Kohya LoRA training
What does print training command do
How to calculate number of steps for each Epoch
How to calculate how many regularization images you need
When you should increase batch size when doing Stable Diffusion training?
How number of total steps max training steps are calculated in Kohya training
How you can generate your own regularization / classification images
How to manually edit generated Kohya training command and execute it
How to start training in Kohya
How to do training on your second GPU with Kohya SS
How much VRAM is SDXL LoRA training using with Network Rank Dimension 32
SDXL LoRA training speed of RTX 3060
How to fix image file is truncated error
How to reach and contact me if you get an error
VRAM usage and speed testing of different Network Rank
How to use absolute min VRAM to make it work
When is first checkpoint generated and where they are saved
How to continue training from saved state
Auto saved configuration files
How to use LoRAs with Automatic1111 Web UI
How to assign previews to your LoRA files / checkpoints
How to do x/y/z LoRA checkpoint comparison to find best LoRA model
How to understand if your LoRA model is overtrained / cooked or not
Testing our LoRAs stylization capability
How to generate studio shot quality images that you can use on LinkedIn, Twitter, Instagram and such
How to find best generated images with using an AI tool
How to utilize ChatGPT to find very good prompts
How to utilize high-res fix and LoRA inpainting to get amazing quality distant shot images
How to fix hands and face
How to use same training command I used
When you need to reduce weight / emphasis of the rare token
How to join our Discord community for help and tips