by Zhou Daquan Papers
2 papers found
Conference
SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer
Enze Xie, Junsong Chen, Yuyang Zhao et al.
ICML 2025arXiv:2501.18427
111
citations
InfoBatch: Lossless Training Speed Up by Unbiased Dynamic Data Pruning
Ziheng Qin, Kai Wang, Zangwei Zheng et al.
ICLR 2024arXiv:2303.04947
80
citations