Description
Our team is responsible for the quality of text-to-speech synthesis models in GigaChat. We are looking for people to work on model quality, multimodal GigaChat, and other exciting projects.
Responsibilities
- improve existing text-to-speech synthesis models and create new solutions
- ensure the superiority of our models in key performance metrics
- support the implementation of developed models into the production environment
- participate in research projects, including training acceleration, voice cloning, low-resource training, and the application of reinforcement learning methods
- read scientific papers and discuss acquired knowledge and results within the team
- present found solutions at internal seminars, make publications on Habr and in our Telegram channel.
Requirements
- commercial experience and knowledge of Python, knowledge of fundamental data structures and algorithms knowledge and understanding of mathematics and its applications in the field
- Data Science experience in developing and training deep neural networks, with special attention paid to audio processing
- successful experience integrating models into production systems broad knowledge in the fields of NLP, linguistics, Russian language, biology, and physics directly related to speech synthesis
- C++ programming skills
- publications and experience participating in scientific conferences is an advantage.
Conditions
- hybrid or remote work format (Russia)
- annual salary review and annual bonus
- corporate gym and relaxation areas
- more than 400 educational programs from SberUniversity for professional and career development extended
- VHI (Voluntary Health Insurance), preferential insurance for family, and a corporate pension program
- flexible mortgage discount equal to 1/3 of the Central Bank's key rate
- free SberPrime+ subscription, discounts on products from partner companies.
