[做点Lora自己偷偷用，能代替工作那就更好了，有鱼大家一起摸，早点下班少加班熬夜]
大模型：电商场景MIX_V2
采样方法：Euler a
选代步数：28-40
权重：0.8-1 推荐0.8!!!!!!!!!
实际应用：泛化性较好，可以随意输入提示词出意想不到的效果，可以生成电商人物。
放大算法：R-ESRGAN 4x+
重绘幅度：0.3-0.4
建议尺寸：512x512
可搭配其他Lora或者controlnet，希望大家多多返图，看到更多的创意和效果
优化建议请留言
Description
FAQ
Comments (7)
what?
?
GPT translation for you.
"Using Lora a bit for personal use, if it can substitute work, that would be even better. Let's all benefit from the fruits of labor together, leave work early, avoid overtime, and stay up late less.
Main model: E-commerce scenario MIX_V2
Sampling method: Euler a
Iterations: 28-40
Weights: 0.8-1 Recommend 0.8 !!!!!!!!!
Practical application: Good generalization, can produce unexpected effects by freely inputting prompt words, can generate e-commerce characters.
Magnification algorithm: R-ESRGAN 4x+
Redraw scale: 0.3-0.4
Recommended size: 512x512
Can be combined with other Lora or controlnets, hope everyone contributes more pictures, and see more creativity and effects.
Please leave optimization suggestions in the comments.
@kyon9292 thanks, I'm still confused what this is and what it was trained on. It's just for creating images like in the examples?
@outersdf In order to generate good-looking pictures in e-commerce style for use at work
这个大模型搜不到 请问搜什么呀？
你的关键词里的lora，还是抖音写实呢，搞笑不你说
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.