This educational project was made under the guidance of Daniil Merkulov on optimization methods in MIPT. I studied the theory behind diffusion models and stable diffusion architecture. I explored methods to run and optimize training (fine-tuning) process in conditions of limited computing power and memory (in my case -- GPU in Google Colab).

Results of fine-tuning runwayml/stable-diffusion-v1-5 to the pictures of my cat Ailend:

Original image	Cartoon style	On a rocket