Member of Technical Staff Post Training

Skills

About the Role

You will design high-performance, scalable software for model training and post-train models for state-of-the-art performance. You will improve SFT and RL training cycles, coordinate with specialist teams, and research and implement ideas using supercompute and data infrastructure.

Requirements

• Strong software engineering skills
• Proficiency in Python, JAX, PyTorch, and XLA/MLIR
• Experience with distributed training infrastructure including Kubernetes and Slurm
• Experience with Ray
• Experience with large-scale distributed training strategies
• Experience training large models at scale
• Experience with model post-training and performance optimization

Responsibilities

• Design and write high-performance scalable software for model training
• Post-train models to achieve state-of-the-art performance
• Coordinate with specialist teams to produce broadly capable models
• Develop techniques to improve SFT and RL training cycles
• Research, implement, and experiment using supercompute and data infrastructure

Benefits

• Weekly lunch stipend of $75/£75 or local-currency equivalent
• Health and dental benefits
• Mental health budget
• RRSP matching, 401K, and pension scheme
• Parental leave top-up for up to 6 months
• Arts, culture, fitness, wellness, quality-time, and workspace-improvement benefits
• Education and learning stipend
• 6 weeks of paid vacation
• Remote travel budget and annual company offsite
• Co-working benefit
• $500 home office stipend