increase resemblance by experimenting with step count in training #avatarai

figure out how to input a sample image then use my photoreal model and put the user on top #avatarai

try outpainting of high res pics with Stable Diffusion on my own model #avatarai

train higher step model to increase resemblance #avatarai

try Tower of Belem and Ibiza on model #avatarai

try airport renders #avatarai

make step count demo to show how AI imaging works #avatarai

compare Stable Diffusion base with my model #avatarai

do more renders with my merged model #avatarai

add clean input folder to clean worker #avatarai to save more disk space

see progress of moving all old images to R2 #avatarai now at 87% done

fix #avatarai downloads were not showing if not upscaling finished yet and images were already put on R2 for new jobs

add text in 5 languages for people not to just upload the same pic 20 times #avatarai

upload images straight from GPU API to R2 to avoid file storage and add support for upscaling from R2 to Replicate API after #avatarai

redirect R2 bucket r2.avatarai.me to main site #avatarai

make script to move #avatarai old images to R2

make R2 bucket for #avatarai to move images

try Textual Embeddings first time in SD2.1 #avatarai

manage to make SD2.1 work on MacBook Pro #avatarai

go hot bath to think how to fix SD2.1 on local MacBook Pro #avatarai