Distributed ML Infra Engineer - Large-Scale Pre-Training

The Institute of Foundation Models is seeking a distributed ML infrastructure engineer to extend and scale our training systems. You’ll work with world‑class researchers to extend distributed training frameworks (DeepSpeed, FSDP, FairScale, Horovod) and build robust multi‑node launch scripts.

You will own experiment tracking, metrics logging, and job monitoring for external visibility, and aim to improve reliability and performance of large‑scale pre‑training pipelines.