2017 DevFestMN Raghu Rajah

on YouTube ↗02h 06m

" summary

A hands-on demo shows tensorflow code running from a Docker image gcr.io/tensorflow/tensorflow, authoring a Jupyter notebook, visualizing the execution graph in TensorBoard, and training a tiny linear model on 10,000 instances of synthetic data. The talk gives exact commands (docker run, tensorboard --logdir), explains session.run/placeholders/variables, warns about GPU setup and fast-changing APIs, and walks through gradient steps that converge the weights toward [1,2,3,4], proving the example actually works.

AI-written summary · about

" chapters (12)

1. 0:00Workshop introduction
2. 1:31Machine learning and TensorFlow
3. 13:17TensorFlow fundamentals
4. 22:58Setting up TensorFlow
5. 26:45First TensorFlow code
6. 36:35TensorBoard visualization
7. 44:27Graph execution in TensorFlow
8. 54:27Linear algebra primer
9. 1:07:04Machine learning core concepts
10. 1:23:59ML algorithm types
11. 1:33:17Linear regression in TensorFlow
12. 2:03:28Q&A

" related talks

• 24mDocker for Machine Learning!Haseeb Niazi, Paniraja Koppa · OpenInfra Summit 2020
• 59mTensorFlow in Three SentencesBarbara Fusinska · NDC Sydney 2017
• 01h 18mPutting Tensorflow Models into ProductionBugra Akyildiz · PyData Austin 2019
• 41mLearning TensorFlowRobert Layton · PyCon AU 2016 Science and Data Miniconf
• 37mSimplifying Training Deep & Serving Learning Models...using TensorflowHolden Karau · PyData Berlin 2018
• 01h 04mTensorFlow in your BrowserOswald Campesato · Silicon Valley Code Camp 2018
• 35mHow does Docker run machine learning on specialized AIShashank Prasanna · Docker Community All-Hands #6
• 55mTensorflow: Learning MachinesVikram Tiwari · Windy City DevFest 2018