Inspiration source: @周周

Workflow name: semi-realistic and semi-illustration style

[Workflow introduction]
Open the workflow and upload a portrait image at the image loading area on the left. You can use the built-in scene prompts, reverse the Runninghub Captioner, and combine ControlNet to give it to the sampler to generate the final image. The parameters are all built-in. Click Run. Come and try it!

[Usage scenario]
You only need to upload 1 image to get a unique semi-realistic and semi-illustration style image, which is vivid and interesting. Come and use your creativity!

[Key nodes]
PULIDFLUX

[Model version]
FLUX
Model name: flux1-dev-fp8.safetensors

[ControlNet]
Strength: 0.7
Start time: 0
End time: 1.0

[LoRA model]
Model name: F.1 Lora half_illustration_V1.safetensors

[K sampler]
Scheduler: normal
Noise reduction: 1.0
Sampler: euler