MY CHECKPOINT>LORA CONVERSION PIPELINE IS TEMPORARILY BROKEN, WHICH IS WHY I AM SUPPLYING THIS MODEL ONLY AS A CHECKPOINT, FOR NOW.

I am putting up my archive of trained DreamBooth models as LORAs converted via Kohya_ss - but for some reason, which I cannot yet resolve, this particular checkpoint will not convert.

I don't know how long it will be before I know more about the problem, and can maybe fix it*. In the meantime, here is the full 2GB checkpoint.

Shivam Dreambooth of US TV actress and Bewitched star Elizabeth Montgomery, using 112 ref images and 1400 (real) reg images, for 12,000 steps.

I generally use these models in complex workflows, inpainting faces after initial T2Is, and using ControlNet extensively. So if you're hoping for one-click prompt magic, my models aren't data-curated with this in mind, but rather as tools for traditional workflows that use Photoshop and other older methods.

This model may produce subject-accurate NSFW output unless countered by negative prompts.

_______________

* I would be most grateful for any community help with this issue, and for anyone else to try and convert the downloaded checkpoint in to LORA via Kohya_ss. Anyone who succeeds in converting it, please get in touch via comments.