Take advantage of the integration with NVTabular for seamless data preprocessing and feature engineering. You can use the building blocks to design custom architectures such as multiple towers, multiple heads and tasks, and losses.īuild sequential and session-based recommenders from any sequential tabular data. The library provides modular building blocks that are compatible with standard PyTorch modules. The Transformers4Rec library provides sequential and session-based recommendation. You can create of new models quickly and easily. Model input layer enables you to change either without impacting the other.Īssemble connectable building blocks for common RecSys architectures so that Iterate rapidly on featuring engineering and model exploration by mappingĭatasets created with NVTabular into a model input layer automatically. Loaders for TensorFlow, PyTorch, and HugeCTR. With Merlin Models, you can:Īccelerate your ranking model training by up to 10x by using performant data Models to highly-advanced deep learning models. The Merlin Models library provides standard models for recommender systems withĪn aim for high-quality implementations that range from classic machine learning Load a subset of an embedding table into a GPU in a coarse-grained, on-demand Scale embedding tables over multiple GPUs or nodes. Strategies for scaling large embedding tables beyond available memory. HugeCTR contains optimized data loaders with GPU-acceleration and provides Recommendation models by distributing training across multiple GPUs and nodes. HugeCTR is a GPU-accelerated training framework that can scale large deep learning Process datasets that exceed GPU and CPU memory without having to worry aboutįocus on what to do with the data and not how to do it by using abstraction at Prepare datasets quickly and easily for experimentation so that you can train High-level API that can define complex data transformation workflows. The library can quickly and easily manipulate terabyte-size datasets thatĪre used to train deep learning based recommender systems. NVTabular is a feature engineering and preprocessing library for tabularĭata. NVIDIA Merlin consists of the following open source libraries:
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |