Hannah is an experimental character LoRA created for learning, testing, and improving my LoRA training workflow.

This model is not based on any real person. All training images are synthetic and were generated with AI models.

The LoRA was trained on Z-Image Base, but it may also work with Z-Image Turbo. Both Base and Turbo are worth trying, as results can differ in facial consistency, skin texture, and overall realism.

The dataset contains 70 AI-generated images, including close-up portraits, medium shots, full-body images, and NSFW references. NSFW output is possible, but results may vary depending on the prompt, checkpoint, sampler, and LoRA strength.

For NSFW use, I recommend combining it with [ZIT] Mystic XXX.

Trigger word:

hnhF

Recommended settings:

weight: 0.8–1.0

• Works best with realistic photo prompts

• Trained on Z-Image Base

• Can also be tested with Z-Image Turbo

• Suitable for portraits, full-body images, and FaceDetailer workflows

This is an experimental release, so results may vary depending on checkpoint, prompt, sampler, and LoRA strength.

Feedback and shared results are welcome.

Training data: AI-generated images only.