Hwtagger: Automated Image Tagging for Stable Diffusion Training

HaW-Tagger provides automated image tagging data for generative AI training, with the most recent updates occurring in December 2024. Created by the HaW-Tagger organization, the system produces descriptive text labels paired with images for fine-tuning Stable Diffusion models and LoRA modules.

Use Cases

• Training Stable Diffusion checkpoints using generated text tags
• Fine-tuning LoRA models with descriptive image labels
• Automating metadata generation for generative art datasets

Strengths

• GPL-3.0 open-source license
• Tailored for Stable Diffusion metadata standards
• Active maintenance as of late 2024

Limitations

• Unknown dataset scale and row count
• Potential for automated labeling errors
• Lack of documentation on the source image domain

Provenance

Source

HaW-Tagger GitHub repository

Collection Method

Automated software tagging

Freshness

Last updated December 2024.