O'Reilly logo

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

O'Reilly TensorFlow World 2019 - Santa Clara, CA

Video Description

Open source TensorFlow 2.0 is driving the machine learning (ML) revolution around the globe. The TensorFlow World Conference Santa Clara 2019—the first international conference devoted to TensorFlow—provided the thousands who attended the conference with an extraordinary opportunity to see TensorFlow 2.0 in action, discover new ways to use it, and learn how to successfully implement it in their own enterprises. This video compilation offers you the chance to experience the TensorFlow World Conference yourself. If you get it and explore it, you’ll come away with a firm understanding of the entire machine learning stack, TensorFlow 2.0, and the reasons why companies like Spotify, LinkedIn, Amazon, Twitter, and Uber use TensorFlow to solve complex business problems.

Highlights include:

  • A front row view for all of the best keynotes, tutorials, and technical sessions from the TensorFlow World Conference Santa Clara 2019.
  • Complete presentations from some of the world’s top TensorFlow practitioners, including talks by the people and teams who developed TensorFlow.
  • Keynote addresses from TensorFlow’s leaders, such as Google Brain co-founder Jeff Dean; Theodore Summe, the head of product for Cortex, Twitter’s central ML organization; and Megan Kacholia, VP of Engineering for Google Research.
  • Deep-dive tutorials, including Laurence Moroney’s (Google Brain) primer on ML with TensorFlow; Sandeep Gupta’s (Google) review of ML in JavaScript using TensorFlow.js; and Neelima Mukiri’s (Cisco) intro to model building and optimization for TensorFlow in any Kubernetes environment.
  • Applications sessions focusing on real-world TensorFlow implementations, like Asif Hasan’s (Quantiphi) talk on using ML to both predict cancer recurrence and recommend treatment; Bhushan Jagyasi’s (Accenture) survey of TensorFlow successes in banking and insurance; and Hamel Husain’s (GitHub) review of automating developer workflows on GitHub with TensorFlow.
  • Core Technologies sessions, where you’ll hear directly from TensorFlow team members such as Paige Bailey (Google) on TensorFlow Swift, a next-generation ML platform; Raziel Alverez (Google) on TensorFlow model optimization techniques; and Robby Neale (Google) on how to build models with tf.text.
  • Accelerators sessions, including Victoria Rege (Graphcore) on how to target high-performance ML accelerators using TF XLA; Sudipta Sengupta (AWS) on the basics of integrating deep learning accelerators with TensorFlow; and Manjunath Kudlur (Cerebras Systems) on the software stack that connects users and TensorFlow to the Cerebras WSE deep learning accelerator.
  • Production pipeline sessions, such as Robert Crowe’s (Google) tutorial on using TensorFlow TFX to create ML pipelines; Animesh Singh’s (IBM) intro to TFX, hybrid cloud ML pipelines, and Kubeflow; and Shajan Dasan’s (Twitter) explanation of how Twitter builds reliable, high-scale TensorFlow inference pipelines.
  • Text, Language, and Speech sessions, where you’ll learn the basics of processing human communication from experts like Kiwisoft’s Aurélien Géron (on natural language processing using transformer architectures) and NVIDIA’s Jason Li (on end-to-end speech recognition using the OpenSeq2Seq deep learning toolkit).
  • Mobile and Edge sessions, such as Kaz Sato (Google) on using AutoML Vision to support vision recognition model training on mobile phones and Alasdair Allan (Babilim Light Industries) on using TensorFlow Lite on small, embedded devices.
  • Multiple sessions on TensorFlow in the enterprise; TensorFlow ethics, and security; and on TensorFlow in ML research.

Table of Contents

  1. Keynotes
    1. Opening keynote - Jeff Dean (Google) 00:25:39
    2. The latest from TensorFlow - Megan Kacholia (Google) 00:12:02
    3. TensorFlow, open source, and IBM (sponsored by IBM) - Frederick Reiss (IBM) 00:05:16
    4. Accelerating ML at Twitter - Theodore Summe (Twitter) 00:10:29
    5. Enterprise-ready TensorFlow in the Cloud (sponsored by Google Cloud Platform) - Craig Wiley (Google) 00:10:05
    6. TensorFlow community announcements - Kemal El Moujahid (Google) 00:13:12
    7. TFX: An end-to-end ML platform for everyone - Konstantinos Katsiapis (Google), Anusha Ramesh (Google) 00:12:28
    8. Personalization of Spotify Home and TensorFlow - Tony Jebara (Spotify) 00:08:46
    9. TensorFlow Hub: The platform to share and discover pretrained models for TensorFlow - Mike Liang (Google Research) 00:07:42
    10. Accelerating TensorFlow for research and deployment (sponsored by NVIDIA) - Ujval Kapasi (NVIDIA) 00:05:16
    11. “Human Error”: How can we help people build models that do what they expect - Anna Roth (Microsoft) 00:09:32
    12. TensorFlow Lite: ML for mobile and IoT devices - Jared Duke (Google), Sarah Sirajuddin (Google) 00:14:21
    13. Sticker recommendation and AI-driven Innovations on Hike messaging platform - Ankur Narang (Hike) 00:05:47
    14. TensorFlow.js: Bringing machine learning to JavaScript - Sandeep Gupta (Google), Joseph Paul Cohen (Mila | University of Montreal) 00:15:49
    15. MLIR: Accelerating AI - Chris Lattner (Google), Tatiana Shpeisman (Google) 00:09:56
  2. Accelerators
    1. TensorFlow on the Cerebras Wafer-Scale Engine - Manjunath Kudlur (Cerebras Systems), Andy Hock (Cerebras Systems) 00:41:28
    2. Targeting high-performance ML accelerators using XLA - Victoria Rege (Graphcore), David Norman (Graphcore) 00:35:05
    3. Fast and lean data science with TPUs - Martin Gorner (Google) 00:42:49
    4. TensorFlow and TPUs in the real world: Converting deep learning projects to train faster - Sam Witteveen (Google) 00:38:20
    5. Faster inference in TensorFlow 2.0 with TensorRT - Siddharth Sharma (NVIDIA), Joohoon Lee (NVIDIA) 00:40:00
  3. Production pipelines
    1. Introduction of Hilbert AutoML with TensorFlow Extended (TFX) at Yahoo! JAPAN - SHIN-ICHIRO OKAMOTO (Actapio f.k.a. YJ America) 00:38:31
    2. Reliable, high-scale TensorFlow inference pipelines at Twitter - Shajan Dasan (Twitter), Briac Marcatté (Twitter) 00:45:09
    3. Advanced model deployments with TensorFlow Serving - Hannes Hapke (Wunderbar.ai) 00:41:22
    4. Machine learning over real-time streaming data with TensorFlow - Yong Tang (MobileIron) 00:39:12
    5. Scaling TensorFlow at LinkedIn - Keqiu Hu (LinkedIn), Jonathan Hung (LinkedIn), Abin Shahab (Linkedin) 00:29:52
    6. How to track and manage TensorFlow 2.0 and Keras model experiments with MLflow - Juntai Zheng (Databricks) 00:37:59
    7. Running TFX end to end in hybrid clouds leveraging Kubeflow Pipelines - Animesh Singh (IBM), Pete MacKinnon (Red Hat), Tommy Li (IBM) 00:42:21
  4. Sponsored
    1. Managing the full TensorFlow training, tracking, and deployment lifecycle with MLflow (sponsored by Databricks) - Clemens Mewald (Databricks) 00:35:58
    2. Trusted AI: Bringing trust back into AI through open source (sponsored by IBM) - Animesh Singh (IBM) 00:40:09
    3. Maximizing the performance and longevity of your TensorFlow applications on Google Cloud Platform (sponsored by Google Cloud) - Karthik Ramachandran (Google Cloud), Kaz Sato (Google) 00:37:35
    4. Running TensorFlow at scale on GPUs (sponsored by NVIDIA) - Neil Truong (NVIDIA), Khoa Ho (NVIDIA) 00:45:08
  5. Core technologies
    1. Welcome & Opening Remarks - Edd Wilder-James (Google), Joana Filipa Bernardo Carrasqueira (Google) 00:04:35
    2. Great TensorFlow Research Cloud projects from around the world (and how to start your own) - Zak Stone (Google) 00:33:34
    3. Getting involved in the TensorFlow community - Joana Filipa Bernardo Carrasqueira (Google), Nicole Pang (Google) 00:31:25
    4. Building models with tf.text - Robby Neale (Google) 00:39:32
    5. Swift for TensorFlow - Paige Bailey (Google), Brennan Saeta (Google) 00:39:21
    6. TFX: Production ML pipelines with TensorFlow - Robert Crowe (Google) 00:41:45
    7. TensorFlow model optimization: Quantization and pruning - Raziel Alverez (Google) 00:40:49
    8. Unlocking the power of machine learning for your JavaScript applications with TensorFlow - Kangyi Zhang (Google), Brijesh Krishnaswami (Google), Joseph Paul Cohen (Mila | University of Montreal), Brendan Duke (ModiFace) 00:43:36
    9. TensorFlow Lite: Solution for running ML on-device - Pete Warden (Google), Nupur Garg (Google) 00:33:43
    10. Scaling TensorFlow using tf.distribute - Taylor Robie (Google), Priya Gupta (Google) 00:40:34
    11. Neural structured learning in TensorFlow - Da-Cheng Juan (Google Research), Sujith Ravi (Google AI) 00:42:30
    12. Introduction to TensorFlow 2.0: Easier for beginners and more powerful for experts - Joshua Gordon (Google) 00:40:56
  6. Ethics, security, privacy
    1. Build more inclusive TensorFlow pipelines with fairness indicators - Tulsee Doshi (Google), Christina Greer (Google) 00:35:58
    2. TensorFlow Privacy: Learning with differential privacy for training data - Ulfar Erlingsson (Google Brain) 00:41:17
    3. A journey into the world of federated learning with TensorFlow Federated - Krzysztof Ostrowski (Google) 00:44:44
  7. Meet the SIGs: Lightning Talks
    1. SIG Addons - Sean Morgan (Two Six Labs) and Yan Facai (Alibaba) 00:05:12
    2. SIG Build - Jason Zaman (Light Labs) and Austin Anderson (Google) 00:02:51
    3. SIG IO - Yong Tang (MobileIron) 00:04:28
    4. SIG JVM - Karl Lessard and Christian Tzolov (Pivotal) 00:03:00
    5. SIG Micro - Neil Tan (Antronix) 00:04:58
    6. SIG MLIR - Tatiana Shpeisman (Google) and Pankaj Kanwar (Google) 00:04:36
    7. SIG Networking - Bairen Yi (Bytedance) and Jeroen Bedorf (Leiden Observatory) 00:04:18
    8. SIG Rust - Edd Wilder-James (Google) 00:01:58
    9. SIG Swift - Brennan Saeta (Google) and Paige Bailey (Google) 00:06:21
    10. SIG TensorBoard - Mani Varadarajan (Google) and Gal Oshri (Google) 00:03:02
  8. Applications
    1. How machine learning can empower a 16-year-old to make crossing the street safer - Mikhail Szugalew (The Knowledge Society) 00:29:23
    2. Don’t beat the market; beat the bots: Adversarial networks in finance - Garrett Lander (Manceps), Al Kari (Manceps) 00:22:47
    3. TensorFlow and medicine: Using deep learning for real-time segmentation of colon polyps - Aalok Patwa (Archbishop Mitty High School) 00:35:20
    4. Personalizing the infinite jukebox: ML and the TensorFlow ecosystem at Spotify - Josh Baer (Spotify), Keshi Dai (Spotify) 00:39:40
    5. How Criteo optimized and sped up its TensorFlow models by 10x and served them under 5 ms - Nicolas kowalski (Criteo), Axel Antoniotti (Criteo) 00:38:29
    6. Tagging cancer recurrence through machine learning - Asif Hasan (Quantiphi), Adam Hammond (Quantiphi) 00:43:02
    7. TensorFlow business case study showcase - Moderated by: Deepak Bhadauria (Google) Panelists: Saurabh Mishra (Quantiphi), Upendra Sahu (Quantiphi), Bhushan Jagyasi (Accenture), David Beck (Cognizant), Rahul Sarda (Wipro Limited) 00:36:18
    8. Diagnose and explain: Neural X-Ray diagnosis with visual and textual evidence - Wisdom d'Almeida (Mila) 00:36:12
    9. Building AI to play the FIFA video game using distributed TensorFlow - Shengsheng Huang (Intel), Jason (Jinquan) Dai (Intel) 00:38:26
    10. HARP: An efficient and elastic GPU-sharing system - Lingling Jin (Alibaba), Pengfei Fan (Alibaba) 00:31:33
    11. Train and serve object detectors for autonomous driving - Pengchong Jin (Google) 00:43:16
    12. Enhance recommendations in Uber Eats with graph convolutional networks - Ankit Jain (Uber AI Labs), Piero Molino (Uber AI Labs) 00:39:27
    13. Effective sampling methods within TensorFlow input functions - William Fletcher (Datatonic), Laxmi Prajapat (Datatonic) 00:43:15
    14. Generative malware outbreak detection - Sean Park (Trend Micro) 00:46:50
  9. Mobile & Edge
    1. Working with TensorFlow Lite on Android with C++ - Joe Bowser (Adobe) 00:33:39
    2. Measuring embedded machine learning - Alasdair Allan (Babilim Light Industries) 00:38:30
    3. Deep learning for Android with TensorFlow - Margaret Maynard-Reid (Tiny Peppers) 00:28:36
    4. AutoML Vision and Edge TPU: Bringing TensorFlow Lite models to edge devices - Kaz Sato (Google) 00:43:15
  10. JavaScript
    1. From dance to diagnosis: How Tensorflow.js is shaping AI in Africa - Babusi Nyoni (Triple Black) 00:29:52
    2. Building deep learning applications using TensorFlow to combat schistosomiasis - Zac Yung-Chun Liu, Giulio De Leo, Andy Chamberlin (Stanford University), Susanne Sokolow (UCSB), Ton Ngo (IBM) 00:40:51
    3. Node-RED and TensorFlow.js: Developing deep learning IoT apps in the browser - va barbosa (IBM), Paul Van Eck (IBM) 00:36:14
    4. Handtrack.js: Building gesture-based interactions in the browser using TensorFlow.js - Victor Dibia (Cloudera Fast Forward Labs) 00:41:07
  11. Text, language, speech
    1. Improving the health of public conversations on Twitter with TensorFlow - Li Xu (Twitter), Yi Zhuang (Twitter) 00:41:49
    2. Modular convolution considered beneficial - Jack Chung (AMD), Chao Liu (AMD), Daniel Lowell (AMD) 00:49:54
    3. Natural language processing using transformer architectures - Aurélien Géron (Kiwisoft) 00:45:24
    4. A novel solution for a data augmentation and bias problem in NLP using TensorFlow - KC Tung (Microsoft) 00:40:59
  12. TensorFlow Contributor Summit
    1. TensorFlow: To 2.0 and beyond - Martin Wicke (Google) 00:16:36
    2. Why is machine learning seeing exponential growth in its communities? - Omoju Miller (GitHub) 00:32:55
    3. Getting involved in the TensorFlow community - Edd Wilder-James (Google), Joana Filipa Bernardo Carrasqueira (Google) 00:26:42
    4. How to create a perfect Pull Request and what to expect when you submit it - Chandni Shah (Google) 00:23:23
    5. Building TensorFlow: Libraries and custom-op - Jason Zaman (Light Labs), Yifei Feng (Google) 00:26:24
  13. Tutorials
    1. Accelerating training, inference, and ML applications on NVIDIA GPUs - Maggie Zhang (NVIDIA), Nathan Luehr (NVIDIA), Josh Romero (NVIDIA) - Part 1 00:59:12
    2. Accelerating training, inference, and ML applications on NVIDIA GPUs - Maggie Zhang (NVIDIA), Nathan Luehr (NVIDIA), Josh Romero (NVIDIA) - Part 2 00:56:29
    3. Accelerating training, inference, and ML applications on NVIDIA GPUs - Maggie Zhang (NVIDIA), Nathan Luehr (NVIDIA), Josh Romero (NVIDIA) - Part 3 00:50:25
    4. TensorFlow Lite: Beginner to expert - Andrew Selle (Google) - Part 1 00:55:38
    5. TensorFlow Lite: Beginner to expert - Andrew Selle (Google) - Part 2 00:48:49
    6. TensorFlow Lite: Beginner to expert - Andrew Selle (Google) - Part 3 00:53:49
    7. Zero to ML hero with TensorFlow 2.0 - Laurence Moroney (Google) - Part 1 00:46:03
    8. Zero to ML hero with TensorFlow 2.0 - Laurence Moroney (Google) - Part 2 00:49:33
    9. Zero to ML hero with TensorFlow 2.0 - Laurence Moroney (Google) - Part 3 00:49:07
    10. Zero to ML hero with TensorFlow 2.0 - Laurence Moroney (Google) - Part 4 00:51:27
    11. ML in production: Getting started with TensorFlow Extended (TFX) - Robert Crowe (Google) - Part 1 00:48:59
    12. ML in production: Getting started with TensorFlow Extended (TFX) - Robert Crowe (Google) - Part 2 00:55:27
    13. ML in production: Getting started with TensorFlow Extended (TFX) - Robert Crowe (Google) - Part 3 1:00:10
    14. Introduction to machine learning in JavaScript using TensorFlow.js - Sandeep Gupta (Google), Brijesh Krishnaswami (Google) - Part 1 1:01:34
    15. Introduction to machine learning in JavaScript using TensorFlow.js - Sandeep Gupta (Google), Brijesh Krishnaswami (Google) - Part 2 1:01:03
    16. Privacy-preserving machine learning with TensorFlow and TF Encrypted - Jason Mancuso (Dropout Labs), Yann Dupis (Dropout Labs) - Part 1 00:46:16
    17. Privacy-preserving machine learning with TensorFlow and TF Encrypted - Jason Mancuso (Dropout Labs), Yann Dupis (Dropout Labs) - Part 2 00:51:46
    18. Privacy-preserving machine learning with TensorFlow and TF Encrypted - Jason Mancuso (Dropout Labs), Yann Dupis (Dropout Labs) - Part 3 00:45:09
    19. Swift for TensorFlow in 3 hours - Mars Geldard (University of Tasmania), Tim Nugent (lonely.coffee), Paris Buttfield-Addison (Secret Lab) - Part 1 00:59:41
    20. Swift for TensorFlow in 3 hours - Mars Geldard (University of Tasmania), Tim Nugent (lonely.coffee), Paris Buttfield-Addison (Secret Lab) - Part 2 1:00:27
    21. Swift for TensorFlow in 3 hours - Mars Geldard (University of Tasmania), Tim Nugent (lonely.coffee), Paris Buttfield-Addison (Secret Lab) - Part 3 1:02:50
    22. Recurrent neural networks without a PhD - Martin Gorner (Google) - Part 1 1:02:27
    23. Recurrent neural networks without a PhD - Martin Gorner (Google) - Part 2 1:02:38
    24. Enterprise AF solution for text classification (using BERT) - Leonardo Apolonio (Clarabridge) - Part 1 00:54:36
    25. Enterprise AF solution for text classification (using BERT) - Leonardo Apolonio (Clarabridge) - Part 2 00:52:45
    26. Enterprise AF solution for text classification (using BERT) - Leonardo Apolonio (Clarabridge) - Part 3 1:01:07
    27. Hyperparameter tuning for TensorFlow using Katib and Kubeflow - Neelima Mukiri (Cisco), Meenakshi Kaushik (Cisco) - Part 1 00:45:17
    28. Hyperparameter tuning for TensorFlow using Katib and Kubeflow - Neelima Mukiri (Cisco), Meenakshi Kaushik (Cisco) - Part 2 00:44:05
    29. Hyperparameter tuning for TensorFlow using Katib and Kubeflow - Neelima Mukiri (Cisco), Meenakshi Kaushik (Cisco) - Part 3 00:45:13