/camels

Primary LanguagePython

CaMeLS: Cooperative Meta-Learning Service for Recommender Systems

Instructions

Installation

  1. Clone the repository and its submodules with git clone --recurse-submodules.
  2. Recommended: Build the environment docker image with docker compose build. Alternatively: Install the requirements to a new or existing Python 3.9 virtual environment.
  3. Skip this step if you want to use this release version of CaMeLS instead of configuring your own. Instructions to host a new CaMeLS server are in the server readme. Instructions to connect with the new CaMeLS server are in the client readme.
  4. Write the CaMeLS server address TBA to the file camels/client/connection_settings.py.
  5. Optional: Use the RecSys data loader to load and process common data sets. Refer to its readme for usage.
  6. Use camels/client_functions.py to connect with the server and execute CaMeLS routines. Or call functions in camels/client/client_routine.py directly to customize more options, e.g., use custom data sets.

You can find the CaMeLS prototype that is discussed in our paper in the workshop_perspectives branch. The instructions above do not apply to that version.