Posts

Like the Olympic spirit, generating high-quality images and videos faster has always been a core pursuit of the generative modeling community. Early efforts mainly followed a numerical-analysis perspective: researchers tried to build more efficient ODE solvers, such as DDIM1, DPM-Solver2, and iPNDM3, to approximate the target distribution with fewer sampling steps. At the same time, distillation is also an immediately effective acceleration method, although it usually requires additional training and higher resource cost. ...

Introduction Modern AI research is rarely done locally. Whether you are fine-tuning LLMs or training diffusion models, you will almost always work on remote servers, and often on HPC (High Performance Computing) clusters. A single server feels like a big computer without a GUI. But once you step into a Slurm cluster, the logic changes completely. Many new researchers get lost here, spending more time on setup and queue errors than on research. ...

2025 Year in Review