The project should have a .env file in its root. An example of how it can look like is:
PROJECT_NAME=fapi-transformers
BACKEND_CORS_ORIGINS=["http://localhost:8000", "https://localhost:8000", "http://localhost", "https://localhost"]
NLP_ROOT="/opt/fapi-transformers/models"
QA_MODEL={"path": "${NLP_ROOT}/path-to-qa-model", "load": true}
ZS_MODEL={"path": "${NLP_ROOT}/path-to-zero-shot-model", "load": true}
NER_MODEL={"path": "${NLP_ROOT}/path-to-ner-model", "load": true}
POSTGRES_USER=postgres
POSTGRES_PASSWORD=postgres
POSTGRES_SERVER=database
POSTGRES_DB=app
Models compatible with the transformers library (e.g. models from https://huggingface.co/models) should be placed either in a models/ folder from root of the project, or specified from somewhere else by changing services.app.volumes in docker-compose.yaml.
Should be run with docker-compose:
docker-compose up --build -d
Afterwards, API documentation will be available at http://localhost:8000/docs#/.
This project is licensed under the terms of the MIT license.