Welcome!

RecWizard is a comprehensive toolkit designed for Conversational Recommender System (CRS) research, based on PyTorch and Huggingface.

Getting started

Learn the basics and become familiar with launching and interacting with a RecWizard CRS model.

User guide

Learn how to use RecWizard to build your own CRS model, and contribute to RecWizard.

Resources

High-level overviews of the existing CRS datasets and trained models for RecWizard.

API reference

Thechnical descriptions of how RecWizard classes and methods work.

Installation

Note

Ensure you are using python>=3.8, then follow our instruction to install related packages. Note that we test RecWizard toolkit on Linux, and leave Windows / MacOS development in the future.

• RecWizard Installation via PyPI (Recommended)

pip install recwizard

• RecWizard Installation Locally

git clone git@github.com:McAuley-Lab/RecWizard.git
cd RecWizard; pip install -e .