
The Lora is trained on MidJourney Images that use my "Personalised Style" + Zen Ink Wash Sumi-e
This is a It is a monochromatic Ink Wash style rooted in ancient China that evolved in Japan.
Тrigger words ArsMJStyle, Zen Ink Wash Sumi-e
If you use a model with strong opinion about colors you can also use Black and White, monochrome
Start ~1 and adjust according to your preferences.
If you don't get the desired results on Pony, increase the strength..
This is a style already present in SDXL, so you can go high with the strength and still get good results.
描述:
The model produces nice effects in the 0.4 - 1.5 strength range.
I personally love using 1.5 strength on the Flux Dev GGUF Q4 and 1.2 on the regular Dev model.
You can safely use strength 2 and still get decent, though more abstract, results.
I strongly recommend using the keywords "monochrome, greyscale, black and white" if you want a classical Sumi-e effect. Without them, the LoRA can still create a beautiful "Ink Wash" style.
If you're using a low-end machine with 6/8 GB of VRAM, check out my guide:
https://civitai.com/articles/6846/running-flux-on-68-gb-vram-using-comfyui
All images in the Showcase Gallery were created using this workflow and the Flux Q4 Dev model on a machine with 8GB VRAM:
https://civitai.com/models/658639/super-simple-gguf-quantized-flux-lora-workflow
训练词语: ArsMJStyle, Zen Ink Wash Sumi-e
名称: ZenInkWashSumi-eFlux.safetensors
大小 (KB): 18819
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success