Software Engineer, SystemsML, Training Frameworks & Inference Optimization
Meta

Job Info

The Core AI Model Optimization and Research (MORE) team is hiring SWEs to work on Model Optimization and Training Frameworks. The MORE team is a horizontal ML team under XRTech consisting of multiple pillars focusing on:1. Training Frameworks and Training Efficiency2. Model Optimization for On-Device Inference3. Foundational On-Device Models.Our team enables on-device models across Family of Apps, VR and AR, bringing the magic of AI to the edge. Along with our partners, we have shipped on-device SAM on IG and Full body Avatars on Quest-3. We also build and maintain the training framework (Vizard) used by multiple groups in XRTech for model development and optimization.

Software Engineer, SystemsML, Training Frameworks & Inference Optimization Responsibilities:

• Optimize models for latency and power consumption for both on-device and GPU inference.
• Work closely with partner teams to help them meet quality, latency and complexity constraints on resource constrained devices.
• Build tools to automate model optimization and compression.
• Fine tune, quantize and deploy models on phones, AR and VR devices.
• Contribute to Vizard, a new framework built for training models targeted towards on-device use-cases.
• Ideate and implement new features in Vizard to improve developer experience and efficiency.
• Drive framework adoption by supporting migration of key training workloads into the framework.
• Support periodic oncall and help resolve user issues.
• Work with the Vizard team to improve reliability and efficiency of the framework.

Minimum Qualifications:

• Currently has, or is in the process of obtaining a Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience. Degree must be completed prior to joining Meta.
• Specialized experience in the following machine learning/deep learning domains: Model quantization, compression, on-device inference, GPU inference, PyTorch

Preferred Qualifications:

• Masters/PhD degree in Computer Science, Computer Engineering or relevant technical field.
• Experience in building ML frameworks (Pytorch, Pytorch Lightning).
• Experience with distributed systems and optimizing resource utilization.
• Experience with accelerating deep learning models for on-device inference.
• Familiarity with on-device inference platforms (ARM, Qualcomm DSP).
• Optimizing machine learning model inference and training on NVIDIA GPUs.

About Meta: