Model of Jennifer Aniston. Took a while to figure out the best way to prepare and train but I think I figured it out. Hope you like it. It should be accurate to the yonfer Jennifer yet flexible for styling.
For even better results you can combine this LoRA with the corresponding TI by mixing at 50/50: Jennifer Anniston | Stable Diffusion TextualInversion | Civitai
Or this other TI: 90s Jennifer Aniston | Stable Diffusion TextualInversion | Civitai
Utilise the kohya-ss/sd-webui-additional-networks (github.com) in auto1111 to load the LoRA model. You can also blend it with other LoRA models for a diverse range of poses.
This model can be combined to the TI embedding I will attach with the model.
Description
For good results:
Use the trigger word
Use LoRA strength of 1
Use DPM++ SDE Karras sampling
Use hires fix with denoising strength of 0.5
FAQ
Comments (3)
you should add jennifer aniston as a tag since there already exists a dreambooth and another TI :-)
I'm of course biased as the one who did the dreambooth but even then - I still think that the TI by @scruffynerf is the best version from all those currently available (including your TI and Lora)
nonetheless, thanks for sharing, perhaps the mix of lora, ti and dreambooth can work wonders? :)
too bad this came out later, otherwise i would include it in our comparisons
can you make a LORA model for taylor swift?
Needs work- the jaw line is wrong in all these examples- she has a square jaw but it is a bit narrower than these images.