#8383rd of 88 in VoiceFor voice, speech and music
diffsinger
Official PyTorch code for DiffSinger singing synthesis and DiffSpeech TTS
- Stars
- 4.9k
- License
- MIT
- Last commit
- Jul 2026
- Last release
- Jan 2022
- Language
- Python
4,725 on 22 Apr 20264,899 on 11 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Research implementation of the AAAI-2022 DiffSinger paper: a shallow diffusion acoustic model that generates mel-spectrograms for singing voice synthesis, plus DiffSpeech for text-to-speech. It covers PopCS and OpenCpop pipelines (cascade, end-to-end, PNDM-accelerated) with NSF-HiFiGAN or HiFiGAN vocoders. Training and inference run from Python scripts documented per pipeline.
Who it is for: Researchers and developers experimenting with singing voice synthesis
Strengths
- Several SVS pipelines: PopCS, OpenCpop cascade, end-to-end, and PNDM variants
- Includes DiffSpeech for TTS on LJSpeech alongside singing synthesis
- Hugging Face Spaces demos for both SVS and TTS
- MIT license with official code matching the paper
Weaknesses
- Pinned to Python 3.8, torch 1.9 and CUDA 10.2/11.4 requirements files
- Last release is January 2022; README updates stop at September 2022
- No Docker image or packaged app; research-style scripts and dataset prep
- Training depends on external datasets such as OpenCpop and LJSpeech
What it needs
- Needs Python 3.8, PyTorch 1.9, PyTorch Lightning, OpenCpop dataset, LJSpeech dataset
- Models: DiffSinger, DiffSpeech, FFT-Singer, NSF-HiFiGAN, HiFiGAN
Also in Voice
See all 88| Rank | Project | Score |
|---|---|---|
| 1 | voicestudioDesktop app for voice cloning, dubbing, dictation and audiobooks | 98 out of 100 |
| 2 | handyOffline push-to-talk dictation app that pastes transcripts into any text field | 91 out of 100 |
| 3 | whisper.cppC/C++ speech-to-text engine running OpenAI Whisper models locally | 90 out of 100 |
| #4 | buzzOffline desktop app for transcribing and translating audio with Whisper | 84 out of 100 |
| #5 | xiaozhi-esp32ESP32 firmware for a voice chatbot with MCP device and cloud tool control | 83 out of 100 |