Spaces:

Imoksan
/

z-image-lora-trainer

Running

App Files Files Community

z-image-lora-trainer / README.md

Imoksan

Upload README.md with huggingface_hub

194aaa9 verified 5 months ago

preview code

raw

history blame contribute delete

620 Bytes

A newer version of the Gradio SDK is available: 6.18.0

Upgrade

metadata

title: Z-Image LoRA Trainer
emoji: 🎨
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 5.29.0
app_file: app.py
pinned: false
license: apache-2.0
suggested_hardware: l4x1
suggested_storage: large

Z-Image LoRA Trainer

Train custom LoRA models for Z-Image Turbo with a simple interface.

Usage

Upload 6-20 training images
Set your trigger word
Configure training parameters
Start training (requires L4 GPU - $0.80/hr)
Download your LoRA

Recommended Settings

Resolution: 512-768px for L4 GPU
Steps: 2000-3000
Learning Rate: 0.0001-0.0002
LoRA Rank: 8-16