大家好,我训练lora模型已经有了一段时间,刚开始训练的时候我还处于摸索阶段,现在我终于发现我使用了错误的大模型来训练lora,这段时间我一直用的是【waiIllustriousSDXL_v14】来训练,但是我自己测试时是没问题的,我也不知道什么原因,现在我将使用【illustriousXL10_v10】来重新训练过去的lora模型,这里要感谢老师们及时发现我的错误。
https://civitai.red/articles/29309
Hello everyone, I’ve been training LoRA models for a while. When I started, I was still in the exploratory stage. Now I finally realized I’ve been using the wrong base model to train the LoRA—I’ve been using 【waiIllustriousSDXL_v14】 all this time. Strangely, my own tests seemed fine, and I’m not sure why. Now, I’m going to retrain the previous LoRA models using 【illustriousXL10_v10】. I’d like to thank the teachers for pointing out my mistake in time.