Tasks:
- * Adapt efficient speech model training and inference frameworks
- * Build evaluation frameworks for pretraining and post-training
- * Build large-scale speech data discovery and preprocessing pipelines
- * Define and label high quality speech data
- * Develop products suited to speech model capabilities
- * Develop reinforcement learning post-training for speech dialogue
- * Explore speech-native model architectures
- * Research generalizable generative speech pretraining
Perks/Benefits:
Skills/Tech stack required:
[C++] [Data pipeline] [Data Pipeline Engineering] [Data Processing] [Distributed Training] [Inference deployment] [Large-scale] [Large-scale model] [Large-scale Model Training] [Model Evaluation] [Model Inference] [Model Inference Deployment] [Model Training] [Online Serving] [Performance optimization] [Pipeline engineering] [Python] [Reinforcement Learning] [Speech Data] [Speech data processing] [Speech modeling]
Educational requirements:
N/A
Role(s):
[Algorithm Researcher] [Machine Learning Researcher] [Researcher] [Speech Algorithm Researcher]