High-Resolution Image Synthesis with Latent Diffusion Models
Role in this project:
ML Engineer Contributions:2 reviews, 11 commits, 1 PR in 7 months
Contributions summary:Andreas contributed to the development of the latent diffusion model, adding code related to image processing, dataset handling, and model training. They implemented a new dataset class for ImageNet and added configurations for training unconditional and class-conditional latent diffusion models. The user also integrated a k-NN search for image retrieval within the inference script and added a script for training the searcher, showcasing the development of image generation pipelines.
diffusion-modelsimage-synthesis
Contributions:40 commits, 20 pushes, 2 branches in 2 months