It has been already, for quite a while now (at least a month or so). have a look at various twitter accounts that post AI images: e.g. https://twitter.com/PLAa

It has been already, for quite a while now (at least a month or so). have a look at various twitter accounts that post AI images: e.g. https://twitter.com/PLAawesome/media https://twitter.com/PLAawesome/media these have had progressively better and better hands over time. You can clearly see that they've been using various model merges (this Lora merging techniques like https://github.com/cloneofsimo/lora https://github.com/cloneofsimo/lora) to get two different models to combine and get the best of both. Many have done better hands and contributed it out. NSFW, but this is one i found that has very realistic hands now: https://civitai.com/models/2661 https://civitai.com/models/2661 It is faster than i can keep up. This is open source collaboration at heart. I am very glad that Stable Diffusion was released publicly. Now if only openAI would do the same with their GPT models.