Job Description
📋 Description Analyze model architectures and identify bottlenecks in training and inference performance (memory Apply and develop quantization, pruning, distillation, and efficient attention techniques. Optimize model code for hardware accelerators (TPUs, GPUs) using compiler features and low-level Experiment with model partitioning and sharding (data, tensor, pipeline, expert) to improve Design and implement low-latency, high-throughput serving for generative models and optimize Build and maintain tools for performance analysis, profiling (e.g., xprof), and debugging ML models. 🎯 Requirements MS or PhD in Computer Science, ML, Robotics, or related field. 5+ years of experience with deep learning architectures (Transformers, Diffusion Models, MoEs) and Proficiency in JAX, Flax, and potentially TensorFlow/PyTorch. Experience with profiling tools (XProf, Perfetto, NVIDIA Nsight) to diagnose ML performance issues. Hands-on experience with quantization, pruning, distillation, and other model compression methods. Strong Python programming, with knowledge of C++ and software development best practices. 🎁 Benefits Salary range: $251,000 – $310,000 USD. Discretionary annual bonus program, equity incentives, and generous benefits.
