Patreon
Home / Apps / Research, Comparisons and Experiments / SDXL DreamBooth Captioning Comparison: WD14 vs Kosmos-2 vs O

SDXL DreamBooth Captioning Comparison: WD14 vs Kosmos-2 vs Ohwx Man

Image captioning in SDXL fine-tuning / DreamBooth of a single person reduces likeness with almost no benefit, in this OneTrainer test using 10.3 GB VRAM.

fine tuningcaptioningcomparison
SDXL DreamBooth Captioning Comparison: WD14 vs Kosmos-2 vs Ohwx Man

Highlights

Compares the effect of WD14 captions from Kohya GUI, Kosmos-2 captions and plain ohwx man on 15 images, 150 epochs and 5200 ground truth regularization images with RealVisXL V4. Captioning can still help for objects or styles; training took 4.6 hours on an RTX 3060.

Video tutorial

How To Properly Setup OneTrainer Concepts And Some Other Options - Quick Tutorial Before Big One · chapters and summary

Where to get it: the full post is free on this Patreon post by Furkan Gözükara (SECourses). Published Mar 28, 2024.

Related

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.

Kohya vs OneTrainer for FLUX Dev Fine-Tuning / DreamBooth: Comparison

Kohya vs OneTrainer for FLUX Dev Fine-Tuning / DreamBooth: Comparison

Kohya and OneTrainer compared for FLUX Dev fine-tuning: OneTrainer with EMA and a copied Kohya config gave weaker results than the best Kohya config.

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.

More in Research, Comparisons and Experiments.