GTC-DC 2019: Speech Recognition and More Through Reusable Neural Components
Note: Viewing this video may require joining the NVIDIA Developer Program or login in
GTC-DC 2019: Speech Recognition and More Through Reusable Neural Components
Oleksii Kuchaiev, NVIDIA; Georg Kucsko, Kensho Technologies; Keenan Freyberg, ; Keenan Freyberg,
In this session we will present work on building a fast, state of the art transcription system in the financial domain, using modular building blocks. We will also present Neural Modules (NeMo), a framework-agnostic toolkit for speech recognition and natural language processing that facilitates creation of applications through reusable, flexible components for training state of the art models.