Trained over 20 images, featuring mostly her face and a few upper body images wich allows you to generate more interesting images. This is my second Low Rank Adaptation model.

The SD 1.5 already have trained a quite number images of Emilia Clarke so you can use the actress name to improve accuracy. You can also use this model as a style on others subjects.

Recomended model: https://huggingface.co/runwayml/stable-diffusion-v1-5

Recomended prompt start:

raw photo, emilia clarke, <lora:daeneryslorav100sd15:1> daenerys targaryen, blonde, game of thrones

Negative:

(asian:1.1), deformed, bad anatomy, disfigured, missing limb, floating limbs, twisted, blurry, fused fingers, long neck, words, logo, text, mutated hands, mutated fingers

After the generation I recomend to send the generated image to img2img and generate new one with the same settings to improve quality. Don't forget you can use inpainting to fix a heck lot of issues.

The generations can change depending on engine, sampler, model, mixed precision set, VAE, and even the accelerator you are using such as medvram and lowvram wich I hardly recommend.