Nvidia - Seattle, WA
posted 2 months ago
NVIDIA is seeking a Senior Machine Learning Engineer for Quantized Training to support next-generation recipes for mixed-precision training. In this role, you will be responsible for distilling large language model (LLM) research literature into its core components, translating that literature into scalable experiments, creating insights to support or refute the efficacy of various techniques, and generating reproducible training recipes. This position requires a deep understanding of the latest advancements in quantized training and the ability to apply this knowledge in practical settings. Your responsibilities will include reviewing state-of-the-art literature in quantized training, building robust, reproducible, and portable training recipes, and providing engineering support to customers using both hardware and software approaches. You will collaborate closely with hardware, software, and research teams to assess and adopt deep learning algorithmic advancements in quantization. Additionally, you will work with production software teams to implement these recipes into production workflows, ensuring that they are effective and efficient. This role is critical in shaping the future of AI at NVIDIA, as you will be at the forefront of integrating and optimizing deep learning frameworks on the most advanced GPUs. You will have the opportunity to influence the long-term opportunities that expand NVIDIA's impact on the datacenter and beyond, all while working in a creative and autonomous environment that encourages innovation.