CodaLab is an open-source web-based platform that enables researchers, developers, and data scientists to collaborate, with the goal of advancing research fields where machine learning and advanced computation is used. CodaLab helps to solve many common problems in the arena of data-oriented research through its online community where people can share worksheets and participate in competitions.

To see CodaLab in action, visit www.codalab.org.

• CodaLab Wiki

The CodaLab community forum is hosted on Google Groups.

• CodaLabDev Google Groups Forum

Assume the codalab-cli repository has been checked out at the same level as this directory and is called ../codalab-cli.

Install all the required Python packages:

./dev_setup.sh

Install the standard configuration file. Edit DATABASES if want to use MySQL instead of sqlite (need to create a MySQL database separately):

cp codalab/codalab/settings/local_sample.py codalab/codalab/settings/local.py

Update the database schema:

cd codalab
./manage syncdb --migrate
./manage config_gen

If you want to use CodaLab in offline mode, run the following to download MathJax:

./manage prep_for_offline

and add this to your codalab/codalab/settings/local.py file:

LOCAL_MATHJAX = True

Start the web server:

cd codalab
./runserver 0.0.0.0:8000

Create an account for codalab by navigating to https://localhost:8000, clicking Sign In, and Sign Up. Use any email address starting with codalab@. This account is just used so we can run the following script:

source venv/bin/activate
cd codalab
python scripts/sample_cl_server_config.py

This script should print out a fragment of a JSON file with the appropriate keys, which should be added to the codalab-cli config file (usually ~/.codalab/config.json). This allows the bundle service to authenticate against the website. The JSON file looks something like this:

"auth": {
    "address": "https://localhost:8000",
    "app_id": "...",
    "app_key": "...",
    "class": "OAuthHandler"
},

Then start the bundle server:

../codalab-cli/codalab/bin/cl server

That is it!