Glid-3-xl

A 1.4B parameter text-to-image model from CompVis, finetuned on CLIP text embeds and curated data.