hao-ai-lab/FastVideo
A unified inference and post-training framework for accelerated video generation.
A star delta has been computed for 62,678 of 63,382 tracked repositories in this period, and only those can be ranked in it.
A repository missing here has not been measured far enough back yet. That does not mean it did not move.
Star counts measured between 08:09 and 17:19 (BKK).
A unified inference and post-training framework for accelerated video generation.
DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
All-in-one training for vision models (YOLO, ViTs, RT-DETR, DINOv3): pretraining, fine-tuning, distillation.
[TMLR 26] EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation
The Biorefinery Simulation and Techno-Economic Analysis Modules
SparkDiffusion: Accelerating DiT video generation by 265× via joint Sparsity, Distillation, Quantization, etc.
PMetal: high-performance Apple Silicon framework for local LLM inference, LoRA/QLoRA fine-tuning, serving, quantization, and MLX/Metal acceleration.
[ICML 2026] Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation" & Causal Forcing++
That's all 8 repositories in this period.