Verified: 2 years ago
SafeTensor
Type | |
Stats | 734 9,716 2 |
Reviews | (128) |
Published | Oct 12, 2023 |
Base Model | |
Hash | AutoV2 27E083B581 |
这是一个光影背景LORA。主要提供光影和背景。
针对SD不会画热气球和巨大的鸟笼做了强化训练
提示词随意写都可以不需要触发词。也可以参考我的提示词。
模型也是一样随便什么模型。例图我画了各种不同模型的效果
大家自行选喜欢的模型用就好了,
需要注意的是这个LORA开高清修复的话不要太高。高于0.4画面变化较大。
大家自行调节到需要的权重。LORA权重推荐0.8-1
This is a light background LORA. It mainly provides light and shadow and background.
Intensive training for SD's inability to draw hot air balloons and huge bird cages
Prompt words can be written anywhere you want without the trigger word. You can also refer to my hint words.
It's the same with a model any model. I drew the effects of various models
You can use whatever model you like,
It should be noted that this LORA is not too high to open HD repair. Higher than 0.4, the picture changes greatly.
Everyone adjusts to the desired weight. LORA recommended weight 0.8-1