, closeup portrait, perfect hair, hair upsweep updo, posing, (modern photo, one-shoulder Light blue gr.jpeg)
, a mature woman, perfect hair, posing, (modern photo), Graphite Gray (dress), (closeup portrait_1.1),.jpeg)
, closeup portrait, perfect hair, hair upsweep updo, posing, (modern photo, graphic tee), (polaroid fr.jpeg)
, closeup portrait, perfect hair, hair upsweep updo, posing, (vintage photo, one-shoulder dress), (pol.jpeg)
, closeup portrait, perfect hair, hair upsweep updo, posing, (modern photo, graphic tee), (polaroid fr.jpeg)
, closeup portrait, perfect hair, hair upsweep updo, posing, (modern photo, one-shoulder Taxi Cab Yell.jpeg)
, a woman with perfect hair, wearing Road Sign Brown (graphic tee_1.1), (A stylish, contemp.jpeg)
, a woman with perfect hair, wearing Nickel Gray (graphic tee_1.1), (A high-tech urban farm.jpeg)
, a woman with perfect hair, wearing Cetacean Blue (graphic tee_1.1), (A chic, contemporary.jpeg)
, a woman with perfect hair, wearing Valentine Pink (graphic tee_1.1), (A bustling, modern.jpeg)
, a woman with perfect hair, wearing Sandstone (graphic tee_1.1), (A chic, contemporary win.jpeg)
, a woman with perfect hair, wearing Lilac (graphic tee_1.1), (An innovative urban tree pla.jpeg)
, a woman with perfect hair, wearing Graphite Gray (graphic tee_1.1), (A bustling, modern c.jpeg)
, a woman with perfect hair, wearing Copper (graphic tee_1.1), (A modern floating marine co.jpeg)
, a woman with perfect hair, wearing Marines Blue (graphic tee_1.1), (An interactive augmen.jpeg)
Type | |
Stats | 1,076 |
Reviews | (66) |
Published | Dec 4, 2023 |
Base Model | |
Training | Steps: 140 Epochs: 420 |
Usage Tips | Clip Skip: 1 |
Trigger Words | dem1rmawby |
Hash | AutoV2 A6B220E4D0 |
Demi Rose Mawby is a British model and social media influencer with more than 20M followers on Instagram. I have trained her TI on about 30 high quality images and for 2000 steps. The embedding seems to work fine with the models I've tested it on (StablyDiffused, DreamShaper, Clarity, base SD 1.5). Hope you like it! :)
Update: This TI has been retrained from scratch with my current settings and a reduced dataset of 15 images. Even if I still liked the previous version (uploaded in February, such a long time ago!) very much, I was upscaling the pics back then and this resulted in oversaturation. Which the current models hide quite well, it has to be said, but even so it was time for an update. Thanks @ElizaPottinger for the timely reminder. :)
Curious about my work process? I have summarized it here.
Do you have a specific idea for a TI in mind? Visit my website and let me know.
Building a good prompt with my TIs
You're obviously free to experiment, but bear in mind that my TIs are trained with a more or less fixed phrasing, that normally starts with:
"photo of EMBEDDING_NAME, a woman"
So I recommend always starting your prompt like that and then building the rest of the prompt from there. For instance, "photo of beautiful (dem1rmawby:0.99), a woman, perfect hair, posing seductively, (modern photo, __attire_mix__), (closeup portrait), jewelry, 24mm, (analog, cinematic, film grain:1.3), (outside of Berghain Panorama Bar, nighttime, gritty), ((detailed eyes)), (epicPhoto), (color picture:1.1)".
Please also note that I'm using the "add detail" LoRA for my example pics. I recommend setting it around 0.5 for best results. CLIP Skip should always be set at 1 when generating pics with this TI.