Role Overview
Representative responsibilities/duties will include but not limited to: optimizing core NN algorithms, integrating TTS components, implementing markups, designing and maintaining emb SDK, building automation test and release process, and working as a maintainer on codes from RD engineers.
What You Will Do
Optimize the core NN algorithms to speed up inference and generation, integrate different TTS components into a flexible pipeline, implement markups to control the TTS engine, design, implement and maintain the emb SDK, and build up the automation test and release process.
Why It Might Be a Fit
Native or close to native Japanese is a must have, plus fluent English the working language, 3+ years of hands-on experience in TTS system development, proficiency in C/C++ and Python, and mastery of ML frameworks.
Requirements
- 3+ years of hands-on experience in TTS system development
- Proficiency in C/C++ and Python
- Mastery of ML frameworks (PyTorch, TensorFlow, etc)
- Basic understanding of autoregressive / non-autoregressive acoustic models and neural vocoders
- Rich experience on software release, version control and branches maintenance
- Experience with ONNX Runtime, TensorRT, or TorchScript, etc
- Experience with zero-shot/one-shot/few-shot voice cloning or emotional TTS systems
- Native or close to native Japanese
- Fluent English
Benefits
- Remote work
- Flexible pipeline
- Automation test and release process
- Zero tolerance for workplace violence
- Basic knowledge of information security and data privacy requirements
Originally posted on