Distributed ML Engineer - Large-Scale GPU Training

Luma is seeking an engineer to build distributed training systems for its large-scale multimodal models across thousands of GPUs. You’ll work on difficult PyTorch, CUDA, and distributed-systems challenges to maximize stability and utilization in massive clusters.

Ideal candidates will have extensive experience with foundation-model training, GFP clusters, and the ability to implement advanced parallelization strategies while building robust monitoring and tooling.