The latest release of nvlabs/sana is v2.0.0. This repository is the SANA-family codebase for efficiently generating high-resolution images and video. The repository description puts forward a linear diffusion transformer. The recorded primary language is Python, and the license is Apache-2.0. The GitHub repository records 9,407 stars.
According to the README summary, this codebase provides a complete pipeline for training and inferring image and video generation models. The included code covers SANA, SANA-1.5, SANA-Sprint, SANA-Video, SANA-Video 2.0, SANA-WM, SANA-Streaming, and Sol-RL. The scope of the repository therefore goes beyond a single model implementation and covers several SANA-family codebases together with training and inference paths. This is based on the description and README summary of the nvlabs/sana repository on GitHub.
The provided information contains no changelog for v2.0.0, so what changed in this release cannot be confirmed. The recorded momentum score is 4.3, and no novelty, technical merit, or quality scores were given. These scores alone make it hard to judge the state of the individual code or its generation performance. v2.0.0 can therefore be regarded only as the confirmed latest version, and the specific changes and performance evaluation fall outside the scope of this information.