Vacancy expired!
Facebook Reality Labs (FRL) focuses on delivering Facebook's vision through Virtual Reality (VR) and Augmented Reality (AR). Enabling compelling user experiences on Virtual and Augmented Reality devices requires innovation and co-design across all layers of stack from novel algorithms to custom silicon. FRL is driving the state of the art forward with breakthrough work in computer vision, speech, virtual assistant, machine learning, mixed reality, graphics, displays, sensors, and new ways to map the human body among many others.We are seeking exceptional interns with a background in developing compilers and/or performance models and/or compiler optimizations for machine learning accelerators. Our compiler targets on-device ML accelerators for mobile AR/VR applications, including VR headsets and AR glasses.You would join a team developing a 'clean slate' compiler implementation, and the work will be across the compilation stack, collaborating with teams working on the development stack from model development to accelerator architectures. We are looking for someone who can contribute to developing performance models and/or performance-targeted optimizations for the compiler. The goal is to measure and optimize the performance of ML models on a target accelerator architecture, given its compute and memory capabilities.You will work with domain experts in ML model design, NAS, compiler design and ML accelerator architectures to understand the challenges and then propose and implement models and/or optimizations in the compiler code base. You will analyze the quality of compiled output and evaluate model accuracy and/or optimization efficacy.Our internships are twelve (12) - sixteen (16) weeks long and we have various start dates throughout the year.
- Define, plan and implement performance models and/or performance-directed compiler optimizations inside the existing on-device ML accelerator compiler
- Collaborate with other research scientists and software engineers to define objective functions for compiler optimizations and quality metrics
- Benchmark compiler versions with and without optimizations
- Collaborate with software and hardware engineers to develop performance models for an ML accelerator
- Communicate optimization potential, design and results clearly, both within the group as well as to the cross-functional groups
- Currently has, or is in the process of obtaining, a PhD degree in Computer Science, Electrical Engineering or related field
- Must obtain work authorization in the country of employment at the time of hire and maintain ongoing work authorization during employment
- 2+ years experience in developing compilers, developing code performance optimization, modeling of computer accelerator architectures, or code optimization for ML accelerators
- 2+ years programming experience
- Intent to return to degree-program after the completion of the internship/co-op
- Experience with deep learning frameworks, such as PyTorch, TensorFlow, Caffe2
- Experience in modeling computer architecture performance
- Experience with machine learning models
- Publication track record in machine learning conferences and/or journals
- Experience in optimizing deep learning networks for execution constraints besides just the final accuracy
- Interpersonal experience: cross-group and cross-culture collaboration
Vacancy expired!