The interview process included an initial recruiter screening, followed by several technical interviews. The interviewers focused on my previous LLM projects, Transformer architecture, model training and fine-tuning, PyTorch coding, inference optimization, and distributed training. There was also a system design discussion about building a scalable LLM service and handling high-concurrency inference.
How would you optimize the training and inference efficiency of a large language model?
The following metrics were computed from 2 interview experiences for the Baidu LLM Engineer role in Beijing, China.
Baidu's interview process for their LLM Engineer roles in Beijing, China is fairly selective, failing a large portion of engineers who go through it.
Candidates reported having very good feelings for Baidu's LLM Engineer interview process in Beijing, China.