Hugging Face introduces aMUSEd, an efficient non-diffusion text-to-image model based on Masked Image Modeling, as an open reproduction of Google's MUSE.
A community guide combining Pivotal Tuning and Prodigy optimizer techniques for SDXL Dreambooth LoRA fine-tuning, with a new training script released on diffusers.
LG AI Research summarizes trends in AI for materials science from the 2023 MRS Fall Meeting, highlighting Google DeepMind's GNoME model and autonomous laboratory research.