Use the activation token timeless style in your prompt (I recommend at the start)
Trained from 1.5 with VAE. - This is a dreambooth model trained on a diverse set of colourized photographs from the 1880s-1980s.
The goal of this model was to create striking images with rich tones and an anachronistic feel.
When using this model, I typically use painted illustration blur haze monochrome in my negative prompt. I encourage you to experiment and see what works well for you.
Please see this document where I share the parameters (prompt, sampler, seed, etc.) used for all example images.
I look forward to seeing everyone's cool creations!
Description
FAQ
Comments (4)
That is one dapper pooch! :D I'm definitely going to try this one out. Thanks for sharing this!
Trying to use this with Invoke, the images I'm getting are very different than your samples with Emma Stone and Rihanna generated with the exact same settings you shared in the document. Maybe you are forgetting to include some other settings? Embeddings? Also I find it weird that you don't use any commas at all in the prompt to seperate the keywords.
No, I'm not forgetting anything. Are you using Hi-Res fix like listed at the top of the document? I don't use Invoke so can't speak to anything UI specific. Beyond that, in your samples based on the tones and such the model seems to be working. Assuming you're using the same sampler and such, try generating more images?
@wavymulder Yeah, I think Invoke sets the initial image resolution automatically, that should be the reason for the difference.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.