Interactive deep learning book with multi-framework code, math, and discussions. Adopted at 500 universities from 70 countries including Stanford, MIT, Harvard, and Cambridge.
GANs are the most advanced topic currently discussed in d2l.ai. However, diffusion models have taken the image generation mantle from GANs. Holding up GANs as the final chapter of the book and not discussing diffusion models at all feels like a clear indication of dated content. Additionally, fast.ai just announced that they will be adding diffusion models to their intro curriculum, as I imagine will many other intro courses. Want to make sure there's at least a discussion about this here if the content isn't already in progress or roadmapped.
GANs are the most advanced topic currently discussed in d2l.ai. However, diffusion models have taken the image generation mantle from GANs. Holding up GANs as the final chapter of the book and not discussing diffusion models at all feels like a clear indication of dated content. Additionally, fast.ai just announced that they will be adding diffusion models to their intro curriculum, as I imagine will many other intro courses. Want to make sure there's at least a discussion about this here if the content isn't already in progress or roadmapped.