Translation flask API for the Helsinki NLP models available in the Huggingface Transformers library.
First create the directory where you want to store models (and change MODEL_PATH
in config.py
accordingly). Then you can download models using the command line utility. For example...
mkdir data
python download_models.py --source en --target jap
To run with Python>=3.6:
pip install -r requirements.txt
python app.py
To run with docker:
docker build -t machine-translation-service .
docker run -p 5000:5000 -v /path/to/models:/app/data -it machine-translation-service
The front end should then become available at http://localhost:5000.
Call the service with curl:
curl --location --request POST 'http://localhost:5000/translate' \
--header 'Content-Type: application/json' \
--data-raw '{
"text":"hello",
"source":"en",
"target":"fr"
}'