Generative Diffusion Models: Development, Training and Releasing Kandinsky 2.1
summary
Covers diffusion-based image generation from text, explaining the diffusion process, types of diffusion models, and how they are used to generate images. Describes the development, training, and release of the Kandinsky 2.1 model, including practical aspects of building and deploying that specific model.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 3:15Generative models history
- 11:39Diffusion process
- 15:50Kandinsky 2.1 architecture
- 21:09Model experiments and metrics
- 27:37Business applications
- 33:02Generative models future
- 44:25Q&A
related talks
10mDiffusion models
01h 21mGenerative AI with Diffusion Models
41mBuilding Generative Image & Video models at Scale
44mGenerative Adversarial Networks (GANs) and Stable Diffusion
27mGenerative AI: What's Next?
43mEncounters of the Third Kind With the Multimodal Generative AI Revolution
01h 04mGeneralization in Diffusion Models Arises from Geometry-Adaptive Harmonic Rrepresentations
39mDemocratizing Diffusion Models with Diffusers