No Strong GPU? Train SDXL DreamBooth for Free on Kaggle, 30 GPU Hours a Week (2023 Tutorial)
Timestep sampling shift explained for diffusion models: Qwen Image uses an exponential shift and FLUX a linear one, and 2.205 matches Qwen at 1664x928.
Published Nov 24, 2023 · 19,885 views · Watch on YouTube
Get the app and files

Decoding the Timestep Shift in Diffusion Training: Qwen Image & FLUX
Timestep sampling shift explained for diffusion models: Qwen Image uses an exponential shift and FLUX a linear one, and 2.205 matches Qwen at 1664x928.

Kohya GUI Kaggle Notebook: Free SD 1.5 & SDXL Training on Dual T4 GPUs
Kohya GUI Kaggle notebook trains Stable Diffusion 1.5 and SDXL LoRA or DreamBooth models for free on a Kaggle account, on a single T4 or dual T4 GPUs.

SD 1.5 & SDXL DreamBooth and LoRA Training: Full Newbie Workflow
Stable Diffusion training workflow for newbies, built from 120+ SDXL and 70+ SD 1.5 trainings: Kohya, regularization images and ADetailer.

SDXL DreamBooth Training: Photos of Yourself in Cities and Fancy Suits
SDXL DreamBooth training on SDXL 1.0 Base lets you generate photos of yourself in different cities and fancy suits, using a config found after 120 trainings.
Chapters
- 0:00 Introduction To The Kaggle Free SDXL DreamBooth Training Tutorial
- 2:01 How to register Kaggle account and login
- 2:26 Where to and how to download Kaggle training notebook for Kohya GUI
- 2:47 How to import / load downloaded Kaggle Kohya GUI training notebook
- 3:08 How to enable GPUs and Internet on your Kaggle session
- 3:52 How to start your Kaggle session / cloud machine
- 4:02 How to see your Kaggle given free hardware features
- 4:18 How to install Kohya GUI on a Kaggle notebook
- 4:46 How to know when the Kohya GUI installation has been completed on a Kaggle notebook
- 5:00 How to download regularization images before starting training
- 5:22 Introduction to the classification dataset that I prepared
- 6:35 How to setup and enter your token to use Kohya Web UI on Kaggle
- 8:20 How to load pre-prepared configuration json file on Kohya GUI
- 8:48 How to do Dataset Preparation after configuration loaded
- 8:59 How to upload your training dataset to your Kaggle session
- 9:12 Properties of my training images dataset
- 9:22 What kind of training dataset is good and why
- 10:06 How to upload any data to Kaggle and use it on your notebook
- 10:20 How to use previously composed Kaggle dataset in your new Kaggle session
- 10:34 How to get path of session included dataset
- 10:44 Why do I train with 100 repeating and 1 epoch
- 10:54 Explanation of 1 epoch and how to calculate epochs
- 11:23 How to set path of regularization images
- 11:33 How to set instance prompt and why we set it to a rare token
- 11:46 How to set destination directory and model output into temp disk space
- 12:29 How to set Kaggle temporary models folder path
- 13:07 How many GB temporary space do Kaggle provides us for free
- 13:23 Which parameters you need to set on Kohya GUI before starting training
- 13:33 How to calculate the N number of save every N steps parameter to save checkpoints
- 13:45 How to calculate total number of steps that your Kohya Stable Diffusion going to take
- 14:10 If I want to take 5 checkpoints what number of steps I need calculation
- 14:33 How to download saved configuration json file
- 14:43 Click start training and training starts
- 14:55 Can we combine both GPU VRAM and use as a single VRAM
- 15:05 How we are setting the base model that it will do training
- 15:55 The SDXL full DreamBooth training speed we get on a free Kaggle notebook
- 16:51 Can you close your browser or computer during training
- 17:54 Can we download models during training
- 18:26 Training has been completed
- 18:57 How to prevent last checkpoint to be saved 2 times
- 19:30 How to download generated checkpoints / model files
- 21:11 How you will know the download status when downloading from Kaggle working directory
- 22:03 How to upload generated checkpoints / model files into Hugging Face for blazing fast upload and download
- 25:02 Where to find Hugging Face uploaded models after upload has been completed
- 26:54 Explanation of why generated last 2 checkpoints are duplicate
- 27:27 Hugging Face upload started and the amazing speed of the upload
- 27:49 All uploads have been completed now how to download them
- 29:02 Download speed from Hugging Face repository
- 29:17 How to terminate your Kaggle session
- 29:36 Where to see how much GPU time you have left for free on Kaggle for that week
- 29:46 How to make a fresh installation of Automatic1111 SD Web UI
- 31:05 How to download Hugging Face uploaded models with wget very fast
- 31:57 Which settings to set on a freshly installed Automatic1111 Web UI, e.g. VAE quick selection
- 32:07 How to install after detailer (adetailer) extension to improve faces automatically
- 32:51 Why you should add --no-half-vae to your command line arguments
- 33:05 How to start / restart Automatic1111 Web UI
- 33:37 How switch to the development branch of Automatic1111 Web UI to use latest version
- 34:24 Where to download amazing prompts list for DreamBooth trained models
- 35:07 How to use PNG info to quickly load prompts
- 35:52 How to do x/y/z checkpoint comparison to find the best checkpoint of your SDXL DreamBooth training
- 38:09 How to make SDXL work faster on weak GPUs
- 38:37 How to analyze results of x/y/z checkpoint comparison to decide best checkpoint
- 42:06 How to obtain better images
- 42:20 How to install TensorRT and use it to generate images very fast with same quality
- 44:41 How to use amazing prompt list as a list txt file
About this video
🌟 Master Stable Diffusion XL Training on Kaggle for Free! 🌟 Welcome to this comprehensive tutorial where I'll be guiding you through the exciting world of setting up and training Stable Diffusion XL (SDXL) with Kohya on a free Kaggle account. This video is your one-stop resource for learning everything from initiating a Kaggle session with dual T4 GPUs to fine-tuning your SDXL model for optimal performance.
Notebook ⤵️
https://www.patreon.com/posts/kohya-sdxl-lora-88397937
Tutorial GitHub Readme File ⤵️
0:00 Introduction To The Kaggle Free SDXL DreamBooth Training Tutorial
2:01 How to register Kaggle account and login
2:26 Where to and how to download Kaggle training notebook for Kohya GUI
2:47 How to import / load downloaded Kaggle Kohya GUI training notebook
3:08 How to enable GPUs and Internet on your Kaggle session
3:52 How to start your Kaggle session / cloud machine
4:02 How to see your Kaggle given free hardware features
4:18 How to install Kohya GUI on a Kaggle notebook
4:46 How to know when the Kohya GUI installation has been completed on a Kaggle notebook
5:00 How to download regularization images before starting training
5:22 Introduction to the classification dataset that I prepared
6:35 How to setup and enter your token to use Kohya Web UI on Kaggle
8:20 How to load pre-prepared configuration json file on Kohya GUI
8:48 How to do Dataset Preparation after configuration loaded
8:59 How to upload your training dataset to your Kaggle session
9:12 Properties of my training images dataset
9:22 What kind of training dataset is good and why
10:06 How to upload any data to Kaggle and use it on your notebook
10:20 How to use previously composed Kaggle dataset in your new Kaggle session
10:34 How to get path of session included dataset
10:44 Why do I train with 100 repeating and 1 epoch
10:54 Explanation of 1 epoch and how to calculate epochs
11:23 How to set path of regularization images
11:33 How to set instance prompt and why we set it to a rare token
11:46 How to set destination directory and model output into temp disk space
12:29 How to set Kaggle temporary models folder path
13:07 How many GB temporary space do Kaggle provides us for free
13:23 Which parameters you need to set on Kohya GUI before starting training
13:33 How to calculate the N number of save every N steps parameter to save checkpoints
13:45 How to calculate total number of steps that your Kohya Stable Diffusion going to take
14:10 If I want to take 5 checkpoints what number of steps I need calculation
14:33 How to download saved configuration json file
14:43 Click start training and training starts
14:55 Can we combine both GPU VRAM and use as a single VRAM
15:05 How we are setting the base model that it will do training