2019 年 6 月 24–26 日

To view the English version of this schedule please go here.

Simultaneous translation will be provided for all keynote and breakout sessions.

场馆 + 赞助商展示区地图
Venue + Sponsor Showcase Map
Back To Schedule
Wednesday, June 26 • 12:05 - 12:40
将 Kubernetes 用于机器学习框架 - Arun Gupta,AWS

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Kubernetes 可以提供隔离、自动扩展、负载平衡、灵活性和 GPU 支持。这些功能对运行计算和数据密集型、且难以实现并行处理的 ML 模型至关重要。Kubernetes 部署描述符的声明性语法使集中于非运营业务的工程师可以轻松基于 Kubernetes 训练 ML 模型。在本演讲中,我们将解释 Kubernetes 为什么及怎样非常适合单节点和多节点分布式训练、将 ML 模型部署到生产系统中以及设置可视化工具,比如用于监控的 TensorBoard。具体而言,我们将展示如何基于 Kubernetes 集群设置各种开源机器学习框架,如 TensorFlow、Apache MXNet 和 Pytorch。与会者将了解到基于 Kubernetes 建立 ML 框架的分布式训练、消息传送和推理阶段,, 并获得包含完整运行样本的 GitHub repo。

avatar for Arun Gupta

Arun Gupta

Head of the Open Source Program Office, Apple
Arun Gupta is head of the Open Source Program Office at Apple and chairperson of CNCF Governing Board. Previously, he worked at Amazon, partnering  with multiple engineering teams to help define their open source strategy. He has spent over a decade at Sun Microsystems and Oracle... Read More →

Wednesday June 26, 2019 12:05 - 12:40 CST