TensorFlow Open Source Community And Collaboration (TF Dev Summit '19)
Key Takeaways
The TensorFlow open source community is thriving with over 50,000 commits from 1,800 contributors, and the TensorFlow 2.0 release is modular with a low-level API, utilizing Special Interest Groups and Requests for Comment (RFCs) for collaboration and development. The community is supported by various tools and platforms, including Stack Overflow, Google Developer Experts, and the TensorFlow blog.
Full Transcript
[Music] hey everybody how you doing good good excellent that was an amazing set of demos wasn't it so I work with tensorflow helping build community and collaboration around the open source project and usually I put the thank you slide at the end for listening to me but actually I want to thank you for contributing and being part of the tflow project whether you're here in this room or on the live stream there's some been some amazing talk on YouTube from people in China and India and Japan all over the world joining us virtually today thank you for your contributions the project is where it is today because of you of course in core tensor flow alone we've had this many commits uh it's now this figures out a date over 50,000 from over 1,800 contributors and much more than just code commits you know there've been 39,000 stack Overflow questions about tensorflow we have 66 machine learning Google developer experts many of whom are here with us today so welcome and thank you guys it's really great to have you with us and thank you for everything you do helping teach people about tens flow and we've had 14 guest posts to the tens oflow blog and that keeps going up there's so many ways that people are contributing whether you're organizing a Meetup whether you're teaching other people whether you're speaking at conferences thank you you're really helping build out the tensorflow ecosystem so in this talk what I want to do is discuss how we're growing the ecosystem and report back on some of the changes that we've made over the last year so I'm going to cover how we're making it easier to get involved in tens oflow how also we're trying to consult better with the users and the community and be more transparent about our development going to cover how we're empowering everybody to get involved and to do more increasing the number of uh contact points where you can get involved in the project and finally I'm going to go into a bit more depth about the conference that was announced this morning the tens flow world so let's talk about how we're making contribution easier to tensor flow one of the most important things to help people contribute to the project is increasing its modularity you heard Martin talk this morning about the low-level API and we with the move to tense flow 2.0 we're trying to make it less of a monolith both in terms of code and in terms of people organization when you come and you want to contribute to an open source project it helps to be able to find where to contribute and who to work with by splitting things out we're creating more surface area where it's easy to start building and creating new projects and our special interest groups play a big part on this and I'll talk a bit more about them later but it's not just code there's so many more places to contribute this year compared to where we were last year so I'm going to talk briefly about our documentation groups the groups getting involved in testing and people who are blogging and on YouTube and more you know I was super excited to see last week that we have published a TENS FL tutorial now in Korean and that's not a translation that we've done in our team but that has come from the community so thank you so much to Hassam park for the Korean work similarly we're able also to publish it in Russian thank you to Andrew Steen this is just so exciting to see that tensorflow is been taken to more areas around the world thanks to you I'm also really excited about the tens oflow 2.0 testing group led by Paige Bailey this is a bunch of contributors and Google developer experts who are working to give tens flow 2.0 a thorough test and you see on the screen an example of a friction log and so what's happening here is folks are going through ml workflows with tflow 2.0 documenting what they find delightful and awesome and also things that could be a little bit better if you'd like to join in this work this group uh meets weekly and often has guest talks from maintainers and Sig leaders and so on and is really helping uh bring tens flow to0 from The Cutting Edge into something that is Thoroughly tested and ready for use only mentioned we have over 14 posts uh from guests on the tent tolow blog this is a a great from a great post about a real-time person segmentation in the browser with tens flow. JS that comes from a grad student and researcher at ITP so whether it's testing whether it's documentation whether it's blogs conference talks thank you now I want to talk a little bit about tlow rfc's as you probably know RFC means requests for comments this time last year we weren't that organized about how we evolved tense flows design in terms of communicating it and I stood on this stage and told you about how we're going to launch the RFC process well now we've accepted 21 rfc's over the period of the last year this is our key way to communicate design where before code gets landed in the project we post an RFC about the design and consult widely and this isn't just about code that's coming in from the tensorflow core team outwards they can be created on and commented on by anyone and we've had several rfc's that have come from the broader community and I expect see so many more of those in the future with several for instance from the Sig groups already one of the things I'm most proud about is how the RFC process is underpinning the 2.0 transition as is mentioned earlier but all the major changes in tense flow 2 have been proposed and consulted with in rfc's you know this isn't just a great way of Consulting and getting information in feedback going forward you now have a big repository of technical documentation about why design choices were made a certain way in tensorflow and it's a great educational resource as well for people who are coming on and want to get involved in contributing to the project so I really want to give a big thanks to anyone who has authored or reviewed an RFC you've played a vital role in making tendis flow better let's talk a bit about the social structure of tensorflow you know last year I talked about how coming to a large project can be a little bit daunting you don't know where people are where people that have your interests in common are and so we created the special interest groups or sigs as a way of organizing our work there are so many uses of tensorflow so many environments so many architectures and many of them are outside of the scope that the core team can resource and what we wanted to do was enable tens flow to grow and be more sustainable by creating a way for like-minded people to collaborate around well- defined projects so this is why sigs exist they're groups of people who work in together for a defined project Focus we started last year with Sig build and now we have six of them up and running I'm going to give you a quick uh state of the sigs many in fact most of all the Sig leaders are here with us today as well so I'll give a shout out to them and hopefully you also be able to talk to them in the lunch and tomorrow so Sig add-ons first thank you to sha Morgan and Amanda Fandango for leading this group Martin mentioned uh at the beginning of the day that TF cont trip is no longer a part of tensorflow going tensorflow 2.0 and SE addons is a place where a lot of that code is going so these are parts of tend to flow um that don't fall into the core but do conform to these well- defined apis so more losses Ops layers and so on now there's already an RFC published about where you can find things that you used to find in contre that had gone into add-ons and addons are also going to publish another RFC real soon to say well how can you get involved if you have your favorite uh op or whatever that you want to step up and be a maintainer and maintain it for everybody how you can join in the project so I'd encourage you to take a look at that Sig build Sig build really is where tens oflow meets the outside world and it's not always the most glamorous piece of work but building tens oflow packaging it Distributing it is tough and so uh thank you so much to Jason Zaman and Austin Anderson who lead that Sig uh Sig build has achieved a lot in the last few months one thing it's the home for thirdparty contributed builds for architectures that we don't ship at as part of core so IBM power Intel mkl optimize builds and S build Works in improving the tens flow build and helps um us be a better neighbor in the python ecosystem as well as you can imagine machine learning generates a lot of extreme situations that needs changes in ways we evolve in packaging and distributing software CIO is a fantastic group that helps connect tens flow to other systems you know out in the real world your data exists somewhere you're using other systems it's in other formats so this group is led by y tang and Anton dimitriev and it really ships support for extra file systems extra file formats so if you're using any of these things in the Apache ecosystem or any of these file formats you can use the Sig add-ons uh module to use that data in tensor flow this group is prolific they've already dropped four releases uh last week they just created the o.4 release and they also ship our integration with with uh their module to so networking is where a lot of the alternative networking schemes that are available in contrib are going to this is led by Baron ye and donon bedo so if you're using gdr NPI this is where you can find that Sig rust led by Adam Chrome is developing idiomatic language bindings for the rust programming language if you're interested in this or any of the other sigs please talk to the leaders they really do want more help and now they're up and running they're in a great place to bring people on if you're looking for a way to get involved in contributing to tensor flow this is an ideal one finally let me touch on Sig tensor board we've rebooted Sig tensor board this year to uh really work closely with the community and so this is a great time to be involved as the tensel bo team starting to consult and figuring out how we can best enable people who using tensor Bo both in terms of creating plugins or using it at scale if you go to the demo area above and go to the tensil board stand you'll find money and gal there who'll be happy to talk to you so this is the URL for anything you want to do with joining the tens Community from docs to testing to sigs and all the other ways to be involved the develop a mailing list please head there and if you're here with us rather than on the live stream we're doing a contributor luncheon tomorrow where there'll be a little panel discussion about contributing to tensorflow and many of the core team and the Sig leaders will be there to talk to you so finally let's move on and talk about tens flow world I'm so excited about this it's our vision to bring together the amazing people people who are part of our community and give us space for everyone to connect with each other you know there's so much we can all learn from how we're all working with tens flow so working with the Riley media we're going to have this event at the end of October this year here in Santa Clara it'll be four days that really celebrate the T flow ecosystem we'll have content from talks to tutorials there'll be an expo and place for vendors to present so we understand it's tense flow gets out into the real world there is a large ecosystem um Beyond pokes in this room and we really excited that that means that we can bring everyone together you know the main point of doing something like this is to connect all the amazing users and everyone with experience to share as you heard the call for proposals is now open so if you have anything to share with your work your product your company head to tensorflow world put in a proposal we've got about four or five weeks uh while the call for proposals is open and then we'll be selecting talks a few weeks after that I really hope that you will lend your voice to this event and I'm so excited to see you all in October so once again thank you we really appreciate how much that you are part of our community honestly in my job the best thing every day is when I get to talk to folks who are using or developing tensorflow and I know that's the same for everyone in the team it really is so exciting to work with everybody here this is the way uh you can meet me please do if you have any issue or any desire to contribute or get involved in the community reach out we're really happy to talk to you so thank you very much for your attention [Music]
Original Description
TensorFlow is where it is today because of many amazing open source contributions, through GitHub, Special Interest Groups and Requests for Comment (RFCs). In addition to that, we have communities of contributors focused on documentation, and on testing the forthcoming TensorFlow 2.0 release. This talk highlights the work of those groups, and explains how you can get involved.
See the revamped dev site → https://www.tensorflow.org/
Watch all TensorFlow Dev Summit '19 sessions → http://bit.ly/TFDS19Sessions
Event homepage → http://bit.ly/TFDS19
Subscribe to the TensorFlow YouTube channel → https://bit.ly/TensorFlow1
Speakers:
Edd Wilder-James, Open Source Strategist
Music by Terra Monk → http://bit.ly/TerraMonkTFDS
Event Photo Album → http://bit.ly/TFSummit19
event: TensorFlow Dev Summit 2019; re_ty: Publish; product: TensorFlow - TensorFlow Core; fullname: Edd Wilder-James;
Watch on YouTube ↗
(saves to browser)
Sign in to unlock AI tutor explanation · ⚡30
Playlist
Uploads from TensorFlow · TensorFlow · 0 of 60
← Previous
Next →
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The TensorFlow YouTube Channel is Here!
TensorFlow
Answering Your TF Questions #AskTensorFlow
TensorFlow
Chatting With the TensorFlow Community (TensorFlow Meets)
TensorFlow
All About TensorFlow Code (Coding TensorFlow)
TensorFlow
TensorFlow: an ML platform for solving impactful and challenging problems
TensorFlow
Keynote (TensorFlow Dev Summit 2018)
TensorFlow
tf.data: Fast, flexible, and easy-to-use input pipelines (TensorFlow Dev Summit 2018)
TensorFlow
Eager Execution (TensorFlow Dev Summit 2018)
TensorFlow
Machine Learning in JavaScript (TensorFlow Dev Summit 2018)
TensorFlow
Training Performance: A user’s guide to converge faster (TensorFlow Dev Summit 2018)
TensorFlow
The Practitioner's Guide with TF High Level APIs (TensorFlow Dev Summit 2018)
TensorFlow
Distributed TensorFlow (TensorFlow Dev Summit 2018)
TensorFlow
Debugging TensorFlow with TensorBoard plugins (TensorFlow Dev Summit 2018)
TensorFlow
TensorFlow Lite (TensorFlow Dev Summit 2018)
TensorFlow
Searching Over Ideas (TensorFlow Dev Summit 2018)
TensorFlow
Reconstructing Fusion Plasmas (TensorFlow Dev Summit 2018)
TensorFlow
Nucleus: TensorFlow toolkit for Genomics (TensorFlow Dev Summit 2018)
TensorFlow
Open Source Collaboration (TensorFlow Dev Summit 2018)
TensorFlow
Swift for TensorFlow - TFiwS (TensorFlow Dev Summit 2018)
TensorFlow
TensorFlow Hub (TensorFlow Dev Summit 2018)
TensorFlow
Applied AI at The Coca-Cola Company (TensorFlow Dev Summit 2018)
TensorFlow
Real-World Robot Learning (TensorFlow Dev Summit 2018)
TensorFlow
TensorFlow Extended (TFX) (TensorFlow Dev Summit 2018)
TensorFlow
Project Magenta (TensorFlow Dev Summit 2018)
TensorFlow
TensorFlow Dev Summit 2018 - Livestream
TensorFlow
Introducing TensorFlow Lite (Coding TensorFlow)
TensorFlow
TensorFlow Dev Summit 2018 Highlights
TensorFlow
Jeff Dean, Head of AI at Google discusses the impact of ML (TensorFlow Meets)
TensorFlow
TensorFlow Mobile vs. TF Lite and More! #AskTensorFlow
TensorFlow
Using TensorFlow to enable research & production across many fields (TensorFlow Meets)
TensorFlow
Teaching TensorFlow for Deep Learning at Stanford University (TensorFlow Meets)
TensorFlow
TensorFlow Lite for Android (Coding TensorFlow)
TensorFlow
Using the tf.data API to build input pipelines (TensorFlow Meets)
TensorFlow
Training Models in the Cloud & the Benefits of AI Toolkits #AskTensorFlow
TensorFlow
Execute operations immediately with TensorFlow's Eager Execution (TensorFlow Meets)
TensorFlow
TensorFlow Lite for iOS (Coding TensorFlow)
TensorFlow
Get started with TensorFlow's High-Level APIs (Google I/O '18)
TensorFlow
TensorFlow for JavaScript (Google I/O '18)
TensorFlow
TensorFlow in production: TF Extended, TF Hub, and TF Serving (Google I/O '18)
TensorFlow
Get started with TensorFlow's High-Level APIs in 5 mins | Google I/O 2018
TensorFlow
TensorFlow and deep reinforcement learning, without a PhD (Google I/O '18)
TensorFlow
TensorFlow Lite for mobile developers (Google I/O '18)
TensorFlow
Advances in machine learning and TensorFlow (Google I/O '18)
TensorFlow
Distributed TensorFlow training (Google I/O '18)
TensorFlow
Classification using neural networks & ML regression models #AskTensorFlow
TensorFlow
TensorFlow and Keras in R - Josh Gordon meets with J.J. Allaire (TensorFlow Meets)
TensorFlow
Focus on your experiment with TensorFlow Estimators (TensorFlow Meets)
TensorFlow
How to get started with AI/ML, retraining models, & more! #AskTensorFlow
TensorFlow
TensorFlow - the deep learning solution for mobile platforms (TensorFlow Meets)
TensorFlow
MiniGo: TensorFlow Meets Andrew Jackson (TensorFlow Meets)
TensorFlow
The growth of TensorFlow with added support for JS & Swift (TensorFlow Meets)
TensorFlow
At the intersection of TensorFlow & nuclear physics (TensorFlow Meets)
TensorFlow
NVidia TensorRT: high-performance deep learning inference accelerator (TensorFlow Meets)
TensorFlow
Try TensorFlow.js in your browser (Coding TensorFlow)
TensorFlow
TensorFlow Hub: reusing machine learning modules (TensorFlow Meets)
TensorFlow
How to use TensorFlow in PyCharm (TensorFlow Tip of the Week)
TensorFlow
Training models faster with TensorFlow Hub (TensorFlow Meets)
TensorFlow
Prepare your dataset for machine learning (Coding TensorFlow)
TensorFlow
Using ML to predict insulin use for Type 1 Diabetes (TensorFlow Meets)
TensorFlow
TFX: an end-to-end machine learning platform for TensorFlow (TensorFlow Meets)
TensorFlow
More on: LLM Foundations
View skill →Related Reads
📰
📰
📰
📰
Unlock Your Potential: Mastering AI & Automation for a Brighter Future in India
Dev.to · David García
Nvidia GPU Debt Backstop Unleashes the AI Project Trinity: Capital, Offtake and Datacenters
Semi Analysis
Data From 21,000 Firms Reveals: AI Spending Is Creating Jobs, Not Killing Them
TechRepublic
Business Meditation 006
Medium · AI
🎓
Tutor Explanation
DeepCamp AI