1 paper
Ji Qi, WenPeng Zhu, Li Li +6
The distributed training of foundation models, particularly large language models (LLMs), demands a high level of communication. Consequently, it is highly dependent on a centraliz…