感谢2noise开源项目提供的ChatTTS模型和推理框架
基于ChatTTS (https://github.com/2noise/ChatTTS) 改进的推理框架,具备流式实时语音合成能力。
近期更新更多特性,欢迎【star】优先回复
感谢bella开源项目(https://github.com/LianjiaTech/bella-openapi) 提供资源试点 https://api.bella.top/playground 语音合成(模型选择chat-tts)或者实时语音对话(选择bella-realtime模型)
https://github.com/user-attachments/assets/491ff524-e0f4-44dc-b654-59d2b5b89c81
当前只适配了Nvidia GPU,测试验证RTX8000、V100、A系列、H系列能够稳定运行。
git clone https://github.com/fengyizhu/ChatTTS
cd ChatTTS
安装依赖
pip install --upgrade -r requirements.txt
运行(参数及默认值如下,按需调整)
python -m examples/api/openai.py --host 0.0.0.0 --port 8080 --gpu_memory_utilization 0.9
同步接口用例
curl -X POST "http://localhost:8080/v1/audio/speech" \
-H "Content-Type: application/json" \
-d '{
"model": "Chat-TTS",
"input": "你好,今天天气怎么样。",
"voice": "28",
"speed": 1,
"response_format": "wav",
"stream":false
}'
流式接口用例
curl -X POST "http://localhost:8080/v1/audio/speech" \
-H "Content-Type: application/json" \
-d '{
"model": "Chat-TTS",
"input": "你好,今天天气怎么样。",
"voice": "28",
"speed": 1,
"response_format": "pcm",
"stream":true
}'
(top 30 of 44)
Python
99.6%
感谢2noise开源项目提供的ChatTTS模型和推理框架
基于ChatTTS (https://github.com/2noise/ChatTTS) 改进的推理框架,具备流式实时语音合成能力。
近期更新更多特性,欢迎【star】优先回复
感谢bella开源项目(https://github.com/LianjiaTech/bella-openapi) 提供资源试点 https://api.bella.top/playground 语音合成(模型选择chat-tts)或者实时语音对话(选择bella-realtime模型)
https://github.com/user-attachments/assets/491ff524-e0f4-44dc-b654-59d2b5b89c81
当前只适配了Nvidia GPU,测试验证RTX8000、V100、A系列、H系列能够稳定运行。
git clone https://github.com/fengyizhu/ChatTTS
cd ChatTTS
安装依赖
pip install --upgrade -r requirements.txt
运行(参数及默认值如下,按需调整)
python -m examples/api/openai.py --host 0.0.0.0 --port 8080 --gpu_memory_utilization 0.9
同步接口用例
curl -X POST "http://localhost:8080/v1/audio/speech" \
-H "Content-Type: application/json" \
-d '{
"model": "Chat-TTS",
"input": "你好,今天天气怎么样。",
"voice": "28",
"speed": 1,
"response_format": "wav",
"stream":false
}'
流式接口用例
curl -X POST "http://localhost:8080/v1/audio/speech" \
-H "Content-Type: application/json" \
-d '{
"model": "Chat-TTS",
"input": "你好,今天天气怎么样。",
"voice": "28",
"speed": 1,
"response_format": "pcm",
"stream":true
}'
(top 30 of 44)
Python
99.6%