MMF Configuration System | Building Recommender Systems with PyTorch | Amanpreet Singh

In this tutorial series we show how to build deep learning recommendation systems and resolve the associated interpretability, integrity and privacy challenges. We start with an overview of the PyTorch framework, features that it offers and a brief review of the evolution of recommendation models. We delineate their typical components and build a proxy deep learning recommendation model (DLRM) in PyTorch. Then, we discuss how to interpret recommendation system results as well as how to address the corresponding integrity and quality challenges. The material for this section covers: 1. Why PyTorch? Joe Spisak/Geeta Chauhan 2. Recommender Systems using DLRM - Maxim Naumov/Dheevatsa Mudigere 3. Using Captum for Interpretability for recommender systems - Narine Kokhlikyan 𝟰. 𝗦𝗼𝗹𝘃𝗶𝗻𝗴 𝗶𝗻𝘁𝗲𝗴𝗿𝗶𝘁𝘆 / 𝗤𝗖 𝗰𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀 𝗳𝗼𝗿 𝗿𝗲𝗰𝗼𝗺𝗺𝗲𝗻𝗱𝗲𝗿 𝘀𝘆𝘀𝘁𝗲𝗺𝘀 - 𝗔𝗺𝗮𝗻𝗽𝗿𝗲𝗲𝘁 𝗦𝗶𝗻𝗴𝗵 Important references that will be covered in the tutorial: https://pytorch.org/ https://github.com/facebookresearch/dlrm https://captum.ai/ https://ai.facebook.com/blog/dlrm-an-advanced-open-source-deep-learning-recommendation-model/ Compositional Embeddings Using Complementary Partitions for Memory-Efficient Recommendation Systems (https://arxiv.org/abs/1909.02107) Mixed Dimension Embeddings with Application to Memory-Efficient Recommendation Systems (https://arxiv.org/abs/1909.11810) Deep Learning Inference in Facebook Data Centers: Characterization, Performance Optimizations and Hardware Implications (https://arxiv.org/abs/1811.09886) Deep Learning Training in Facebook Data Centers: Design of Scale-up and Scale-out Systems (https://arxiv.org/abs/2003.09518) https://paperswithcode.com/paper/the-architectural-implications-of-facebooks