Hierarchical text-conditional image generation with CLIP latents13 de abril, 2022 às 04:00OpenAI BlogVer notícia original
