Patreon
Home / Videos / Train FLUX on Your Own Photos With Just a 6 GB GPU: Full Fin

Train FLUX on Your Own Photos With Just a 6 GB GPU: Full Fine-Tuning Course (Kohya, 2024)

Qwen Image 2512 is the new king for training: the video compares trained Base, 2512, Edit 2509 and Edit 2511 models and shows a big leap in realism.

Published Oct 22, 2024 · 38,435 views · Watch on YouTube

Get the app and files

Qwen Image 2512 Training Comparison: The Best Model to Train

Qwen Image 2512 Training Comparison: The Best Model to Train

Qwen Image 2512 is the new king for training: the video compares trained Base, 2512, Edit 2509 and Edit 2511 models and shows a big leap in realism.

SwarmUI to ComfyUI in 1-Click Tutorial: 40+ Presets, Shared Models

SwarmUI to ComfyUI in 1-Click Tutorial: 40+ Presets, Shared Models

SwarmUI exports any preset as a ComfyUI workflow in 1 click; this tutorial runs Z Image Turbo with ControlNet and Wan 2.2 image-to-video presets in ComfyUI.

Kohya FLUX Fine Tuning / DreamBooth Training Tutorial: Windows & Cloud

Kohya FLUX Fine Tuning / DreamBooth Training Tutorial: Windows & Cloud

Kohya FLUX full fine tuning / DreamBooth trains full checkpoints with configs for GPUs with as low as 8 GB VRAM, all using Torch Compile on Windows and Linux.

FLUX 2 vs FLUX SRPO & New Kohya SS GUI Premium FLUX Training App

FLUX 2 vs FLUX SRPO & New Kohya SS GUI Premium FLUX Training App

FLUX 2 Pro compared with a trained FLUX SRPO model, plus the new Kohya trainer V35 that trains FLUX on GPUs with as little as 6 GB VRAM.

Kohya FLUX DreamBooth Results: 20 Profile Images Generated Locally

Kohya FLUX DreamBooth Results: 20 Profile Images Generated Locally

Kohya FLUX DreamBooth showcase: 20 profile images generated recently in SwarmUI to change my profile photo, each 2x latent upscaled to 4 megapixels.

Kohya GUI Kaggle Notebook: Free SD 1.5 & SDXL Training on Dual T4 GPUs

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.

Chapters

  1. 0:00 Introduction
  2. 0:22 Why Fine-Tuning / DreamBooth is better than LoRA training and what is the difference
  3. 3:39 How long preparation of this tutorial taken and how i prepared it
  4. 4:09 How much time each training takes and what are the training speeds and VRAM usages of the configs
  5. 5:38 The list of things that you are going to learn from this tutorial
  6. 10:47 Technical difference between LoRA and Full Fine-Tuning / DreamBooth
  7. 12:33 Windows (private local training) part of the tutorial starts with how to install Kohya GUI for FLUX training
  8. 14:15 How to install Kohya GUI on Windows accurately
  9. 17:46 Picking the accurate configuration
  10. 22:52 Using Image Cropper
  11. 25:15 Using Image Resizer
  12. 26:42 Kohhya Workflow
  13. 27:27 Downloading Models
  14. 28:50 Changing Output Directory
  15. 30:57 Destination Path
  16. 31:32 Clikcing Prepare Training Data
  17. 34:30 Starting Training
  18. 73:49 Massed Compute (cheap private cloud training) part of the tutorial starts with how to install Kohya GUI for FLUX training
  19. 92:02 How to increase empty storage on Massed Compute and how many checkpoints you can save
  20. 121:41 RunPod (cheap private cloud training) part of the tutorial starts with how to install Kohya GUI for FLUX training

About this video

If you want to train FLUX with maximum possible quality, this is the tutorial looking for. In this comprehensive tutorial, you will learn how to install Kohya GUI and use it to fully Fine-Tune / DreamBooth FLUX model. After that how to use SwarmUI to compare generated checkpoints / models and find the very best one to generate most amazing images ever. Moreover, this tutorial will work on training everything such as training a person, an art style, a pet, an item, an object, or general Fine-Tuning (the post will be updated for this later).

🔗 Full Instructions, Configs, Installers, Information and Links Shared Post (the one used in the tutorial) ⤵️

▶️ https://www.patreon.com/posts/click-to-open-post-used-in-tutorial-112099700

The tutorial shows how to train 100% privately on Windows locally and also on the very best cheapest Cloud providers RunPod and Massed Compute same as in your local computer. Moreover the tutorial shows auxiliary tools like Ultimate Preprocessing APP for preparing training images dataset, ultra fast uploading and downloading models to Hugging Face Jupyter notebook, SUPIR upscaler to enhance images to next level and so on.

Still fixing video subtitles / captions as well

Video Chapters (still writing them)

0:00 Introduction

0:22 Why Fine-Tuning / DreamBooth is better than LoRA training and what is the difference

3:39 How long preparation of this tutorial taken and how i prepared it

4:09 How much time each training takes and what are the training speeds and VRAM usages of the configs

5:38 The list of things that you are going to learn from this tutorial

10:47 Technical difference between LoRA and Full Fine-Tuning / DreamBooth

12:33 Windows (private local training) part of the tutorial starts with how to install Kohya GUI for FLUX training

14:15 How to install Kohya GUI on Windows accurately

17:46 Picking the accurate configuration

22:52 Using Image Cropper

25:15 Using Image Resizer

26:42 Kohhya Workflow

27:27 Downloading Models

28:50 Changing Output Directory

30:57 Destination Path

31:32 Clikcing Prepare Training Data

34:30 Starting Training

1:13:49 Massed Compute (cheap private cloud training) part of the tutorial starts with how to install Kohya GUI for FLUX training

1:32:02 How to increase empty storage on Massed Compute and how many checkpoints you can save

2:01:41 RunPod (cheap private cloud training) part of the tutorial starts with how to install Kohya GUI for FLUX training