sprite-gpt / compositional.png

Commit History

Add text-conditioned model (CLIP cross-attention, 120k steps)
14295c9
verified

gmmeyer commited on