AIGC周刊 | 第7期
时间: 2022.12.19-12.25
最新技术:
ChatGPT还在2G冲浪?新模型「youChat」:我已能够解说2022世界杯 https://mp.weixin.qq.com/s/HQDfdgtDhRvSWSdiPQEPcg
89次实验,出错率高达40%!斯坦福首次大型调研,揭露AI写代码漏洞 https://mp.weixin.qq.com/s/tD0ssq0MYTvk7894x2eOOg
最新“3D版”DALL·E爆火,超快速度生成3D点云模型,OpenAI向谷歌新领域发起挑战丨开源 https://mp.weixin.qq.com/s/VD2zVx1HjEhzJMewFnfwSQ
PaLM-rlhf-pytorch: 在PaLM上实现RLHF https://github.com/lucidrains/PaLM-rlhf-pytorch**
Nvidia TensorRT新增对Stable Diffusion的支持 https://github.com/NVIDIA/TensorRT/commit/22d5b6e583e79a080a02ae5aa0d20993826dc4e2
Karlo:基于 OpenAI 的 unCLIP 架构的文本图像生成开源模型 https://github.com/kakaobrain/karlo
GeoCode: 生成高质量可编辑的3Dmesh https://threedle.github.io/GeoCode/
Scalable Diffusion Models with Transformers
Why Can GPT Learn In-Context?
Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generation
课程:
DreamBooth Hackathon 文生图模型微调;学技术,赢大奖 https://mp.weixin.qq.com/s/NHAXfWu_nhc1faKCvW-hBQ
商业:
AI绘画,VC投不动 https://mp.weixin.qq.com/s/3VSlk8MkhYhdLl6EAIFOZw
A16Z | 生成式AI在游戏领域的机会(市场生态&具体公司) https://mp.weixin.qq.com/s/8rWK3F3DXvgEcHJ_MCSwSg
案例:
Tome.app:自动生成PPT https://beta.tome.app/
visualhound.com: 用AI辅助时装设计 https://visualhound.com/
Image-to-Music:图片生成音乐 https://huggingface.co/spaces/fffiloni/img-to-music
Riffusion demo: https://huggingface.co/spaces/anzorq/riffusion-demo
Point*E demo: https://huggingface.co/spaces/MirageML/point-e