Staff Machine Learning Engineer (Generative Model Performance & Efficiency)

Join Waymo, a leader in autonomous driving technology. As a Staff Machine Learning Engineer, you will work on cutting-edge simulations for testing and training the Waymo Driver. You will analyze model architectures, optimize code for hardware accelerators, and design low-latency serving solutions. Enjoy a comprehensive benefits package, including medical, dental, and vision insurance, competitive compensation, and a hybrid work model.

Responsibilities

• Analyser les architectures de modèles et identifier les goulets d'étranglement dans les performances d'entraînement et d'inférence.
• Appliquer et développer des techniques telles que la quantification, l'élagage, la distillation des connaissances et les mécanismes d'attention efficaces.
• Optimiser le code du modèle pour des accélérateurs matériels spécifiques (TPUs, GPUs), en tirant parti des fonctionnalités du compilateur et des bibliothèques de bas niveau.

Requirements

• Strong programming skills in Python and potentially C++, with experience in software development best practices

Proficiency in JAX, Flax, and potentially TensorFlow/PyTorch

• 5+ years of experience with deep learning architectures (especially Transformers, Diffusion Models, MoEs), algorithms, and optimization techniques
• MS or PhD in Computer Science, Machine Learning, Robotics, or a related field
• Hands-on experience with quantization, pruning, distillation, and other model compression methods
• Familiarity with ML compilers like XLA and an understanding of how they translate high-level code to efficient hardware instructions
• Knowledge of TPU and GPU architectures and how to optimize code for them
• Experience contributing to frameworks and libraries that improve training speed and scalability (e.g., JAX, Gemax, XManager)
• Understanding of concepts related to training and serving models across multiple devices and machines

Preferred

• Expertise in using profiling tools (e.g., XProf, Perfetto, NVIDIA Nsight) to diagnose performance issues in ML workloads

Pay

USD 251,000 - 310,000