MLOps (вакансія неактивна)

We are inviting an experienced MLOps specialist to play a key role in advancing and deploying new deep learning models, and in building tools that empower our research team.

Duties

— Build machine learning tools and infrastructure for research and production in cooperation with researchers, full-stack developers, and — - DevOps professionals
— Optimize deep learning models in PyTorch
— Write beautiful and flexible Python code
— Possibly train models, keep track of performance, and tune hyperparameters


We expect you to

— Have a solid understanding of machine learning fundamentals and an interest in optimizing systems for training and inference (for both researcher/developer experience and speed)
— Understand performance issues in deep learning training and inference
— Have experience optimizing models in PyTorch or a similar deep learning library
— Write great Python code using best practices


Would be a plus

— Knowledge of CUDA / interest in developing custom CUDA kernels
— Experience with mulitprocessing and optimizing CPU-based pipelines
— Have some Deep Learning / Machine Learning pet projects


Benefits

— Take cutting-edge research on deep generative models and make it work in production for customers in Hollywood and beyond
— Freedom to try out ideas of your own