Haotong Qin

PT$^2$-LLM: Post-Training Ternarization for Large Language Models

$\text{S}^2$Q-VDiT: Accurate Quantized Video Diffusion Transformer with Salient Data and Sparse Token Distillation