Views
No views yet
language="th")| Model Component | Description | URL |
|---|---|---|
| F5-TTS Thai | Flow Matching-based Thai TTS models | Link |
| F5-TTS IPA | Flow Matching-based Thai-IPA TTS models | Link |
1pip install torch cached-path librosa transformers f5-tts
2sudo apt install ffmpeggit clone https://github.com/biodatlab/thonburian-tts.git
cd thonburian-tts1from flowtts.inference import FlowTTSPipeline, ModelConfig, AudioConfig
2import torch
3
4# Configure F5-TTS model
5model_config = ModelConfig(
6 language="th",
7 model_type="F5",
8 checkpoint="hf://biodatlab/ThonburianTTS/megaF5/mega_f5_last.safetensors",
9 vocab_file="hf://biodatlab/ThonburianTTS/megaF5/mega_vocab.txt",
10 vocoder="vocos",
11 device="cuda" if torch.cuda.is_available() else "cpu"
12)
13
14# Basic audio settings
15audio_config = AudioConfig(
16 silence_threshold=-45,
17 cfg_strength=2.5,
18 speed=1.0
19)
20
21pipeline = FlowTTSPipeline(model_config, audio_config)1from flowtts.inference import FlowTTSPipeline, ModelConfig, AudioConfig
2import torch
3
4# Configure F5-TTS model
5model_config = ModelConfig(
6 model_type="F5",
7 checkpoint="hf://biodatlab/ThonburianTTS/megaIPA/model_last_prune.safetensors",
8 vocab_file="hf://biodatlab/ThonburianTTS/megaIPA/mega_vocab_ipa.txt",
9 vocoder="vocos",
10 device="cuda" if torch.cuda.is_available() else "cpu"
11)
12
13# Basic audio settings
14audio_config = AudioConfig(
15 silence_threshold=-45,
16 cfg_strength=2.5,
17 speed=1.0
18)
19
20pipeline = FlowTTSPipeline(model_config, audio_config)
 🎵 Sample 1 – Single-speaker Thai Normal Text |
 🎵 Sample 2 – Single-Speaker Thai Code-mixed Text |
 🎵 Sample 3 – Multi-Speaker Conversational Speech |
@INPROCEEDINGS{11320472,
author={Aung, Thura and Sriwirote, Panyut and Thavornmongkol, Thanachot and Pipatsrisawat, Knot and Achakulvisut, Titipat and Aung, Zaw Htet},
booktitle={2025 20th International Joint Symposium on Artificial Intelligence and Natural Language Processing (iSAI-NLP)},
title={ThonburianTTS: Enhancing Neural Flow Matching Models for Authentic Thai Text-to-Speech},
year={2025},
volume={},
number={},
pages={1-6},
keywords={Adaptation models;Codes;Accuracy;Error analysis;Phonetics;Robustness;Natural language processing;Text to speech;Noise measurement;Research and development;Thai text-to-speech;Flow matching;F5-TTS},
doi={10.1109/iSAI-NLP66160.2025.11320472}}Thura Aung, Panyut Sriwirote, Thanachot Thavornmongkol, Knot Pipatsrisawat, Titipat Achakulvisut, Zaw Htet Aung, "ThonburianTTS: Enhancing Neural Flow Matching Models for Authentic Thai Text-to-Speech", 2025 20th International Joint Symposium on Artificial Intelligence and Natural Language Processing (iSAI-NLP), Phuket, Thailand, 2025, pp. 1-6, doi: 10.1109/iSAI-NLP66160.2025.11320472.