An Alibaba researcher’s post provides insight into the life of a Large Language Model (LLM) developer in China. The daily schedule mirrors that of an OpenAI researcher, with workdays starting at 9 a.m. and ending around 1 a.m., filled with meetings, coding, model training, and brainstorming. The work regime reflects a drive to match Silicon Valley’s AI advancements. Alibaba’s LLM team, Qwen, has developed foundation models trained with English and Chinese data, with the largest having 72 billion parameters. These models have been integrated into Alibaba’s enterprise communication platform Dingtalk and online retailer Tmall.
Read more at: https://techcrunch.com