- Create a Virtual Environment and activate to it.
- Switch to
documentation/requirements/
and install dependencies using requirements files.
pip install -r "requirements.ver.dev.txt"
Requirement files include:
- requirements.txt: full requierments for development environment.
- requirements.dev.txt: list of requirements for development environment, without package versions.
- requirements.ver.dev.txt: list of requirements for development environment, including package versions.
- requirements.prod.txt: version list of requirements for production environment, without package versions..
- requirements.prod.txt: version list of requirements for development environment, including package versions..
- install and configure PostgreSQL.
- Install 'citext' extension for your database on psql prompt:
psql =# \c db_1
CREATE EXTENSION citext;
This allows NoSQL fields such as json and array fields which this project uses.
3. edit .env
file in project/
directory to reflect your database configuration. I.e. database credentials.
Run the following commands to create database tables:
python .\manage.py makemigrations
python .\manage.py migrate
Run following commands to import linguistics components and languages data:
python .\manage.py import-linguistics
python .\manage.py import-languages
This will populate linguistic compenent's table, as well as list of languages used to tag data.
Run setup command or manually add download required language models for your NLP tasks:
python .\manage.py setup:
This command currently includes installation of default english language resources for NLP frameworks Stanza, NLTK, and Spacy. to change NLP frameworks language models, create a copy of project/datacore/management/commands/setup.py
with a name such as setup-fr.py
and change it's parameters or write your own code. then you can call new setup file with last parameter changed to reflect your filename. E.g. python .\manage.py setup-fr
To run web server, execute following command:
python .\manage.py runserver
This will run your web and API server. on successfull execution, the site will be available on http://127.0.0.1:8000/
and APIs can be accessed via http://127.0.0.1:8000/api
.
To run Jupyter notebooks, switch to notebooks directory and execute following command.
cd ./notebooks
call python ../project/manage.py shell_plus --notebook