[SD 1.5][TextualInversion]Barbara Palvin - Embedding v1
<p>Celebrity embedding of Hungarian model Barbara Palvin. Trained on 52 various modelling and public appearance pictures found in google searches. Manually cropped and bulk-resized with PowerToys.</p><p>After copying into your stable-diffusion-webui folder, it can be triggered with just the name of the embedding (emb-babs, unless you rename the file to something you prefer). I tend to use the embedding names as part of a larger tag, like "cute emb-babs woman" Just to be sure that SD gets what I'm going for, but that's probably a placebo. It might be a hair over-trained, so if it looks a little exaggerated, try (emb-babs:0.9) and tweak from there. Samples were genned with pretty much the same prompt across several models including RinaMix2, SD1.5, Simpmaker, aEros, REA, MoistMix2, and Protogen 3.4. Possibly others. I include the model name in the metadata, so you can check by clicking the prompt info button. ①</p><p>I may have accidentally left the previous name of the embedding used during training and testing in some of the prompts, so if it isn't working, make sure it doesn't say emb-babpal2 instead of emb-babs. I think I got them all, though.</p><p></p><p>Let me know what you think, and feel free to show me your favorite images!</p>
页:
[1]