Image Count: 6,000+ real portrait photographs
Training Duration: Approximately 720 hours
Total Iterations: 500,000+ iterations
This model is a full-parameter fine-tune of QWEN-IMAGE. A minimum of 48GB VRAM is recommended, but FP8 precision mode is supported for lower memory usage. For optimal results, try using a workflow that employs a base sampler followed by a secondary high-resolution upscaling step.
When you need to reverse-engineer or generate descriptive prompts from an image, please use the Qwen3-VL-8B-Instruct model
Description
FAQ
Comments (8)
FP8 and Q4_K_M versions are needed. This checkpoint is very big for many people.
Added to my plan, please wait patiently.
"FP8 precision mode is supported for lower memory usage"
Where can I find that version?
The UNet model is loaded with FP8 precision selected, and it is compatible with 24GB of VRAM, with actual consumption around 20GB.
how to set the K-Sampler？( Steps、Scheduler、CFG)
I need proper workflow for this model. your job seems amazing, but I really don't know how to apply this model into certain workflow.
true. i feel the same
I have Upload Simple Workflow for this mdel : https://civitai.com/models/2441829?modelVersionId=2745499