This simple API is to combine Automatic Speech Recognition for Vietnamese and then translate to the other supported langagues from Google Translations client library. I utilized FastAPI for building this (Great opensource for fast and efficient building the API applications).
Please refer to the opensource https://github.com/vietai/ASR
Please refer to the opensource https://pypi.org/project/googletrans/
Please refer to the opensource https://fastapi.tiangolo.com/
git pull https://github.com/hoangtuanvu/vietnamese_asr_service.git
cd vietnamese_asr_service
Please download follow by the instruction of the source link (HuggingFace) and then copy all relevant directories and files to 'app/pretrained_models' folder.
docker build -t vi_asr_image .
docker run -d --name vi_asr -p 8000:80 vi_asr_image:latest
or do it locally
uvicorn app.service:app --host 0.0.0.0 --port 8000
pip install pytest
pytest app/test_service.py
Note: Please note that, there are some common use-cases. You then add more if needed.
Please open the WebUI followed by
http://host:port/docs
And then fill data input such as 'file' and 'data'.