Open apolinario opened 2 years ago
I was aware of @crowsonkb's ava aesthetic models, but these ones are new to me, what is the dataset/where the sa_0_4 come from?
ava
sa_0_4
It's a linear model trained on clip vit-l/14 embeddings Using 4000 samples annotated with https://github.com/JD-P/simulacrabot
I think the results look pretty great
I was aware of @crowsonkb's
ava
aesthetic models, but these ones are new to me, what is the dataset/where thesa_0_4
come from?