Clone the repository and then:
# docker-compose up
With python installed (we recommend to use a venv enviroment):
# pip install -r requirements-dev.txt
# pip3 install https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.2.0/en_core_web_sm-2.2.0.tar.gz
# python
>>> import nltk
>>> nltk.download('stopwords')
>>> exit()
# ./start.sh
curl --location --request POST 'localhost:5000/classification/movements' \
--header 'Content-Type: text/plain' \
--data-raw 'walmart
uber'
To run classifier model tests:
python -m unittest
Dockerfile
copy the project instead, but still a volume has to be used to write .vector_cache
.
docker-compose -f docker-compose.yml -f docker-compose.prod.yml up --build