Is there a good prompting guide for this
While I've been using Stable Diffusion since the v1.4 days (man that feels like an eon ago...), I haven't really used a model like this trained on e621 a lot (I've been using Flux too much, made my SDXL prompting skills a bit rusty). I know some in the past just did tag, tag, tag sorta thing, but with this having some natural language processing, what is a good way to handle that. Also, same thing for artist prompts, if those exist as well.
Thanks in advanced and sorry for potentially sounding like a noob in this space. 😅
Hi hi, sorry for the late response; the model is trained on e621 tags (replace underscores with spaces) and danbooru, so yes, tag, tag works just fine. There may be some limited capacity for natural language (deviating slightly from tags, but nothing on the level of flux and the likes of course).
Awesome, thank you!