语音识别服务
# docker run -d -p 6379:6379 redis
docker run --name myredis -p 6379:6379 -d redis --requirepass "mypassword"
# Python 虚拟环境
conda create --name ai_asr python=3.10
pip install -r requirements.txt
cp .env.example .env
# 修改配置文件
vi .env
角色识别，使用 pyannote-audio 项目，相关配置见文档说明。
# 首次启动，初始化数据表
python -m core.models
celery -A celery_app.tasks worker --loglevel=info -Q default
celery flower -A celery_app.tasks --broker=redis://:mypassword@localhost:6379/0 --address=127.0.0.1 --port=5555
python -m server.app
- 支持模型 faster-whisper
- 支持模型 whisperX
执行出错
RuntimeError: cuDNN version incompatibility: PyTorch was compiled against (8, 9, 2) but found runtime version (8, 8, 0). PyTorch already comes bundled with cuDNN. One option to resolving this error is to ensure PyTorch can find the bundled cuDNN. Looks like your LD_LIBRARY_PATH contains incompatible version of cudnn. Please either remove it from the path or install cudnn (8, 9, 2)
解决，参考 https://discuss.pytorch.org/t/can-anyone-help-me-solve-it/175166
export LD_LIBRARY_PATH=
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent