Songting Liu
Singapore, Singapore
Summary
🤩
Rockstar🎓
Top SchoolSongting Liu is an ML-focused software engineer based in Singapore with three years of hands-on experience building data pipelines and model tooling for speech synthesis and voice conversion. He is the primary contributor to a well-starred VITS fine-tuning repository that enables fast speaker adaptation and many-to-many voice conversion, where he implemented custom data loaders, spectrogram processing, and CLI utilities including Demucs-based denoising. A Nanyang Technological University alumnus with strong academic performance, Songting combines practical full-stack project work with a focus on audio preprocessing and model integration. His GitHub presence—accumulating over 10k stars—reflects both the practical impact and adoption of his open-source contributions in the speech ML community.
3 years of coding experience
4.41 / 5.00, 4.41 / 5.00 at Nanyang Technological University Singapore
Github Skills (5)
data-preprocessing10
pytorch10
machine-learning10
audio-processing10
python10
Programming languages (2)
Github contributions (5)
Plachtaa/VITS-fast-fine-tuning
Feb 2023 - Mar 2023
This repo is a pipeline of VITS finetuning for fast speaker adaptation TTS, and many-to-many voice conversion
Role in this project:
ML Engineer
ML EngineerContributions:3 releases, 146 commits, 24 PRs in 1 month
Contributions summary:Songting primarily contributed to the development of data loading and preprocessing utilities for a VITS fine-tuning pipeline, which is used for fast speaker adaptation and voice conversion. Their work involved creating custom data loaders and collate functions, along with audio and spectrogram processing, demonstrating a focus on preparing data for machine learning model training. The user also made changes to the core model structure, and added a command-line interface to facilitate audio denoising using Demucs, indicating a hands-on approach to the project's functionality and a full-stack involvement.
Plachtaa/VALL-E-X
Aug 2023 - Dec 2024
An open source implementation of Microsoft's VALL-E X zero-shot TTS model. Demo is available in https://plachtaa.github.io
Contributions:14 PRs, 96 pushes, 120 comments in 1 year 4 months
emotional-speechgpttext-to-speechtransformer-architecturetts
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.