Introduction to Machine Learning in Tensorflow.js
summary
TensorFlow.js runs training and inference in-browser and under Node.js, with a Python converter (pip install tensorflowjs) that turns SavedModel/Keras HDF5 into JSON plus sharded 4MB weight files; the converter implements ~120 TensorFlow ops and ~35 Keras ops. It exposes models API, layers API, and core API; out-of-the-box models include Speech Commands, PoseNet and COCO-SSD and demos like Teachable Machine and Emoji Scavenger Hunt. Chrome/WebGL inference hits ~5 ms on a GTX1080 and ~50–60 ms on an integrated laptop GPU; Node bindings reuse the native C library with CUDA.
AI-written summary · about
chapters (8)
- 0:00Start
- 8:43Core API
- 10:41Layers API
- 14:40Models repo
- 20:51Model Conversion
- 23:16Server side
- 25:10Node.js Performance
- 25:36Visualization
related talks
20mTensorFlow.js Bringing Machine Learning to the Web and Beyond
37mTensorFlow.js - Bringing ML and Linear Algebra to Node.js
48mTensorFlow.js
24mTensorFlow.js + Node
31mDeep Learning at the Edge with TensorFlow.js
28mMachine Learning on the web with TensorFlow.js
41mTensorFlow.js 101: ML in the Browser and Beyond
39mMachine Learning Magic for Your JavaScript Application