Swift for TensorFlow

on YouTube ↗28m

" summary

Swift for TensorFlow embeds first-class differentiable programming into Swift, adds Python interoperability (NumPy, Matplotlib, OpenAI Gym) and runs in Colab with free GPUs. The team builds and debugs real projects—MiniGo is implemented in Swift, unit tests run, LLDB inspects tensor shapes, and the model runs on GPU in Colab. The 0.3 release is available for pioneers; it is experimental, not yet production-ready.

AI-written summary · about

" chapters (5)

1. 0:00Intro
2. 3:51Google Collab
3. 10:35Differentiation
4. 16:38Performance
5. 23:05Future

" related talks

• 39mSwift for TensorFlow: Machine Learning with No BoundariesPaige Bailey · Scale By The Bay 2019
• 47mSwifty TensorFlowSiamak Ashrafi · 360|iDev 2019
• 24mSwift for TensorFlow: machine learning with no boundaries.Paige Bailey · try! Swift NYC 2019
• 40mIntroduction to Swift for TensorflowBrett Koonce · AltConf 2019
• 04mTensorFlow and GoNatalie Pistunovich · dotGo 2019
• 28mTensorFlow 1.0: open source machine learningAndrew Selle · FOSSASIA Summit 2017
• 26mMachine Learning with TensorFlowRajat Monga · MLconf SF 2016
• 43mTensorflow on open source GPUsDavid Airlie · linux.conf.au 2020