These embeddings are trained on classic Sci-fi action movies from the 70s to the 2000s. I ultimately wanted something that could combine the classic scifi aesthetic and the realism of live action cinematography.
--------
To use these embeddings in AUTOMATIC1111, copy them to your embeddings folder in your stable-diffusion-webui directory. No need to restart, just invoke the embedding from your prompt. Make sure you are using a compatible model or you will get errors!
--------
The Sci-fi embeddings were all trained on 465 768x768 images that were cropped and tagged using BLIP for training.
--------
Like most other 2.1 textual inversions this works best when paired up with other embeddings and good negative prompting.
Description
Extended training to 1800 steps before I reached the point where my poor 3080 couldn't go any further. I will continue training this model till it is perfectly cohesive. First example uses some img2img to clean the face but every other example is raw txt2img.