
My second try to make a LoRa for Flux. This is a paint style I created in midjourney. I used it to create pictures to train this LoRa.
V1.0
This LoRa needs a trigger word, but I can't see much difference if I leave it out. “paintstyle001”
I have testing and Weight range is 1.0 or sometimes above. It depends a little on the image.
Just try it out.
V1.1
This version have some more training pics and trained to a much smaller size.
For me i like the result at strength 1.0 - 1.5. The guidance i set on 3.0. But what should i say play with the settings :)
Unfortunately, my image settings were lost during the transfer.
Settings:
Sampler: Euler
Scheduler: simple
Steps: 20
描述:
训练词语: paintstyle001
名称: paint_style_001_fluxd.safetensors
大小 (KB): 335748
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success