nghitts-copy

unofficial mirror of NGHI-TTS models

official announcement: https://www.facebook.com/groups/binhdanhocai/permalink/921555793666509/

official repo: https://github.com/nghimestudio/nghitts

official download: https://drive.google.com/drive/folders/1f_pCpvgqfvO4fdNKM7WS4zTuXC0HBskL

this repo contains only Vietnamese voices:

• Ban Mai
• Chiếu Thành
• Duy Oryx (giọng nam siêu trầm)
• Lạc Phi
• Mai Phương
• Mạnh Dũng
• Minh Khang (Kiến Giải Mã)
• Minh Quang
• Minh Thu
• Mỹ Tâm - 2 phiên bản
• Ngọc Huyền (review phim) - 2 phiên bản
• Ngọc Ngạn
• Phương Trang
• Tài An (CD Media) - 2 phiên bản
• Thanh Phương Viettel
• Thiện Tâm
• Trấn Thành
• Việt Thảo

folder structure:

• original onnx files are in piper-tts/ folder can be used directly with Piper TTS
• converted onnx files in sherpa-onnx/ folder compatible with Sherpa ONNX
  • convert following https://k2-fsa.github.io/sherpa/onnx/tts/piper.html
    • also including generated tokens.txt file
• converted gguf files in gguf/ folder compatible with audio.cpp
  • convert following https://github.com/0xShug0/audio.cpp/blob/main/docs/models/piper_tts.md

example usage:

• Piper TTS: see piper-tts/README.md
• Sherpa ONNX: see sherpa-onnx/README.md
• audio.cpp: see gguf/README.md