This is just Merge of https://civarchive.com/models/332111?modelVersionId=435180 with pony v6
Make the picture a little darker and more detailed
Recommended sampler: DPM++ 2M, Euler a karras
Recommended prompt: score_9, score_8_up, score_7_up, score_6_up, score_5_up,Description
this is just merge of https://civitai.com/models/264290?modelVersionId=398292
don't need score prompt (ex. score9, score8up...)
FAQ
Comments (10)
I still don't understand the point of merging a lora into a model instead of using them separately. I mean the case when it's one-to-one, not multiple models and loras with different weights and different combination. What's the benefit?
I use Lora one almost exclusively, so I had to put it in the prompt every time. I could have made it static using styles and such, but that would have caused conflicts with the plugins I use (e.g. upscaling breaks, bmab errors). In other words, I combined it with the default model because I use one Lora statically without fail.
@tkio If it's a constant part of your workflow, then it makes perfect sense to reduce number of preparation steps. Maybe I should have been more precise in the original post. I don't get why such merges are posted as separate models. It just makes more difficult to find a unique model, the space is cluttered with all possible (more often simple) variation. Don't get me wrong, it's not a rant against you. I actually should thank you for clearly stating what went into the model. Not all people do that and I usually avoid models that lack info and have a lot of buzzwords. I just think the same could be achieved with a simple post where you list resources and parameters. The biggest downside of merging one single lora for me is that you're locked-in with its weight since it's embedded and cannot tweak it for different generations.
@green_anger Oh, I see, it's a simple merge, for sure. I didn't do any additional training or merge more models like the others or Lora, but I uploaded it anyway because the results were pretty good. Also, I tried different weights (e.g. 1.0, 0.9, 0.8...), and around 0.7 was the best value (for me). Anyone can merge, but some people don't know how to do it, and some people are colab users, or new users, or people reinstalling, may have to download the model each time with the civitai link. Downloading so many different styles Lora each time is quite time consuming. So I thought it would be nice to just download a single, proven model and have it ready to go with a decent quality picture. I know civitai has a lot of cool merging models. However, I've noticed that merging too much models often don't work with Lora. I've been using Pony and this one style Lora on a regular basis, and the compatibility with Lora is good and the quality is not bad, so I hope people will try it out.
I like its oiled skin effect, thanks for good model!
When I try to remix an image, it says it's unavailable, but I don't know what that means.
Because this model is only available for downloading and not for the Civitai generator. It usually says "Create" on the right part of the page when it's available.
Nice model,but why are black and white images sometimes generated, with only the eyes having color QAQ
Add more steps + put "monochrome" in negative prompt.
Surprisingly, it works great even if you try to generate semi-realistic/grounded stuff, works with "realistic" loras too.
Great model, and extremely versatile, which is a nice surprise.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.





