Speaker
Abstract
Ray is an open source project that makes it simple to scale any compute-intensive Python workload. Industry leaders like Uber, Shopify, Spotify are building their next generation ML platforms on top of Ray. Ray is equipped with a powerful distributed scheduling mechanism which launches stateful Actors and stateless Tasks in a much more granular and lightweight fashion than existing frameworks. Meanwhile it also has an embedded distributed in-memory object store to drastically reduce data exchange overhead. These architectural advantages make Ray the ideal compute substrate for cutting-edge ML use cases including Graph Neural Networks, Online Learning, Reinforcement Learning, and so forth.
This talk will introduce the basic API and architectural concepts of Ray, as well as diving deeper into some of its innovative ML use cases.
Topics
QCon San Francisco 2022 is a three day conference for senior software engineers, architects and team leads. An international program committee of working engineers selects every session. Patterns and practices, not products and pitches.
Part of the track
MLOps Hosted by Hien Luu AI/ML Leader, Advisor, Speaker, and AuthorFrom the same track
Monday 24 October
10:35 Pacific DEKJ Session Machine Learning Ray: The Next Generation Compute Runtime for ML Applications Zhe Zhang Head of Open Source Engineering @anyscalecompute, Previously Hadoop/Spark infra Team Manager @LinkedIn Ray is an open source project that makes it simple to scale any compute-intensive Python workload. Industry leaders like Uber, Shopify, Spotify are building their next generation ML platforms on top of Ray. 11:50 Pacific DEKJ Session Machine Learning Fabricator: End-to-End Declarative Feature Engineering Platform Kunal Shah ML Platform Engineering Manager @DoorDash, Previously ML Platforms & Data Engineering frameworks @Airbnb & @YouTube At Doordash, the last year has seen a surge in applications of machine learning to various product verticals in our growing business. However, with this growth, our data scientists have had increasing bottlenecks in their development cycle because of our existing feature engineering process. 13:40 Pacific DEKJ Session Machine Learning An Open Source Infrastructure for PyTorch Mark Saroufim Applied AI Engineer @Meta In this talk we’ll go over tools and techniques to deploy PyTorch in production. The PyTorch organization maintains and supports open source tools for efficient inference like pytorch/serve, job management pytorch/torchx and streaming datasets like pytorch/data. 14:55 Pacific DEKJ Session Machine Learning Real-Time Machine Learning: Architecture and Challenges Chip Huyen Co-founder @Claypot AI, previously @Snorkel Ai & @NVIDIA Fresh data beats stale data for machine learning applications. This talk discusses the value of fresh data as well as different types of architecture and challenges of online prediction. 16:10 Pacific DEKJ Session Machine Learning Declarative Machine Learning: A Flexible, Modular and Scalable Approach for Building Production ML Models Shreya Rajpal Founder @Guardrails AI, Experienced ML Practitioner with a Decade of Experience in ML Research, Applications and Infrastructure Building ML solutions from scratch is challenging because of a variety of reasons: the long development cycles of writing low level machine learning code and the fast pace of state-of-the-art ML methods to name a few. 17:25 Seacliff D Session Unconference: MLOps Shane Hastie Global Delivery Lead for SoftEd and Lead Editor for Culture & Methods at InfoQ.com What is an unconference? At QCon SF, we’ll have unconferences in most of our tracks.