MEGATRACE: Troubleshooting Hang and Slowdown in Large-Scale LLM Training Clusters.
Fangzheng Jiao, Menghao Zhang, Bolin Chen, Jiaxun Huang, Yanmin Jia, Xiaohe Hu, Bohua Xu, Bowen Han, Chunming Hu
Browse the full ICDCS paper archive.
Fangzheng Jiao, Menghao Zhang, Bolin Chen, Jiaxun Huang, Yanmin Jia, Xiaohe Hu, Bohua Xu, Bowen Han, Chunming Hu
Browse the full ICDCS paper archive.