We're training a text-to-image model from scratch and open-sourcing it

3 points| ea016 | 5 months ago |photoroom.com