Research Engineer - Audio & Speech Models

The role

You will be a core contributor on the Audio Team building open-source autoencoders, ASR, TTS, SSL, and speech-to-speech models. You will work on large-scale audio training runs, performance optimization, dataset collection, and architecture improvements.

What you will do

• Large-scale audio training runs
• Performance optimization of the training stack
• Audio dataset collection, processing, and evaluation
• Architecture and training methodology ablations and improvements

Required qualifications

• Strong research taste and intuition
• Strong implementation and prototyping ability
• Ability to work well with others in a high-paced research setting
• Excellent communication and collaboration skills
• Ability to work on research and engineering implementation at scale

Nice to have

• Expertise and intuition for training models in the audio domain, including text-to-speech, ASR, speech-to-speech, speech-emotion-recognition, or other models
• Experience in training audio autoencoders
• Understanding of signal processing, especially of audio signals
• Experience with diffusion models, consistency models, or GANs
• Experience with training on large-scale (multi-node) GPU clusters
• Strong grasp of proper experimental methodology for running rigorous ablations and other hypothesis testing
• Understanding of and interest in large-scale, highly parallel data processing pipelines
• Proficiency with PyTorch and Python
• Experience contributing to large pre-existing codebases and rapidly getting up to speed
• Previously published machine learning research in well-respected venues

Benefits

• Comprehensive medical, dental, vision, and FSA plans
• Competitive compensation and 401(k) plan
• Relocation and immigration support on a case-by-case basis
• In-office snacks and meals provided
• Unlimited PTO and company holidays

About the company

Zyphra is an artificial intelligence company based in San Francisco, California.