TensorFlow 1.0: open source machine learning

on YouTube ↗28m

" summary

TensorFlow 1.0 is presented as a production-ready open source ML stack with concrete improvements: a reported 58x speed-up on 64 GPUs for Inception v3, a new C API enabling Go/Java/Haskell bindings, integrated tf.keras and Estimator higher-level APIs, and a XLA JIT compilation path for fusing ops and smaller-device performance. Community adoption shows 50k+ GitHub stars, 475+ external contributors, ~14,000 commits and 5,000 Stack Overflow answers. Added features include TensorBoard, Serving API, TPU support, and clinical-grade retinal hemorrhage detection matching expert error rates.

AI-written summary · about

" chapters (10)

1. 0:00Introduction
2. 1:05Platforms
3. 2:14Multiple languages
4. 2:43Visualization
5. 6:37Multiple models
6. 7:09How TensorFlow works
7. 8:19Progress to date
8. 11:13What can we do
9. 17:56Case Studies
10. 18:55Case Study of Medicine

" related talks

• 26mMachine Learning with TensorFlowRajat Monga · MLconf SF 2016
• 32mGetting Started with TensorFlow 2.0Paige Bailey, Josh Gordon · Google I/O 2019
• 31mTensorflow for JanitorsDániel Molnár · code.talks 2017
• 41mTensorFlow ever higherYoann Benoit · DevFest Paris 2019
• 33mWhat's new in Machine LearningKemal El Moujahid, Sarah Sirajuddin, Craig Wiley · Google I/O 2021
• 43mTensorflow on open source GPUsDavid Airlie · linux.conf.au 2020
• 04mTensorFlow and GoNatalie Pistunovich · dotGo 2019
• 18mTensorFlow.js: From prototype to production, what's new in 2022?Jason Mayes · Google I/O 2022