AdaptCP: Hotness-Aware Adaptive Checkpointing for Industry-Scale Recommendation Model Training
2027 ACM European Conference on Computer Systems (Eurosys) (CCF A)
Junhong Liu, Yitang Yang, Xiaoyang Sun, Jiapeng Chen, Fangzheng Li, Zheng Zheng, Menghao Zhang, Tianyu Wo, Chunming Hu, Chengru Song, Jin Ouyang, Renyu Yang