The Merlin Models library provides standard models for recommender systems with an aim for high quality implementations that range from classic machine learning models to highly-advanced deep learning models.

The goal of this library is to make it easy for users in industry to train and deploy recommender models with the best practices that are already baked into the library. The library simplifies how users in industry can train standard models against their own dataset and put high-performance, GPU-accelerated models into production. The library also enables researchers to build custom models by incorporating standard components of deep learning recommender models and then researchers can benchmark the new models on example offline datasets.

pip install merlin-models

conda install -c nvidia merlin-models

Merlin Models is pre-installed in the NVIDIA Merlin Docker containers that are available in the NVIDIA container repository in three different containers:

To use these Docker containers, you'll first need to install the NVIDIA Container Toolkit to provide GPU support for Docker. You can use the NGC links referenced in the table above to obtain more information about how to launch and run these containers.

To learn about the core features of Merlin Models, see the Retrieval Models page. The key retrieval models are as follows:

• Matrix factorization
• YouTube DNN retrieval
• Two tower
• Ranking models such as DLRM and DCN-V2
• Multi-task learning with Mixture-of-experts or Progressive layered extraction

If you'd like to contribute to the library directly, see the CONTRIBUTING.md file. We're particularly interested in contributions or feature requests for our feature engineering and preprocessing operations. To further advance our Merlin Roadmap, we encourage you to share all the details regarding your recommender system pipeline in this survey.