FLUX LoRA vs Fine Tuning / DreamBooth Comparison: Batch Size 1 vs 7
FLUX LoRA and Fine Tuning / DreamBooth compared with 15 and 256 training images at batch size 1 and 7, for both realism and stylization.

Highlights
Huge experiment set completed and fully tested: 8 trainings, a researched learning rate for each batch size, full grids, training logs, conclusions, 28 realism and 37 stylization test prompts, plus the latest configs and Kohya GUI installers for Windows, RunPod and Massed Compute.
Video tutorial
No Powerful GPU? Train a FLUX LoRA in the Cloud in Under 1 Hour (2024 Kohya Guide) · chapters and summary
More videos: Set Up Windows 11 for AI Apps: Python, CUDA, cuDNN, C++ Build Tools, FFmpeg and Git (2024 Guide), Zero to Hero Stable Diffusion 3 Tutorial with Amazing SwarmUI SD Web UI that Utilizes ComfyUI, FLUX.1 vs Midjourney on the Same Prompt: Run FLUX Locally in SwarmUI From 6GB VRAM (2024)
Related

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.

FLUX Full Fine Tuning vs LoRA Comparison: Less Overfitting & Bleeding
FLUX full fine tuning / DreamBooth gives way better results than LoRA training, as expected, with much less overfitting and bleeding.

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.