ai-forever/kandinsky-2
text2img model trained on LAION HighRes and fine-tuned on internal datasets
Public
6.2M
runs