mori-c / tensorflow-js

Rangle.io ML Introductory Practice Using Tensorflow.js

Tensorflow

ML Introductory Practice Using Tensorflow

Overview

• tf.js is 3-5x slower than GPU tf

Workflow

1. define loss() or error()
   • of predicted output to actual output
   • checks parameter space, to min
2. create linear regression
3. apply stochastic descent
4. loss() and confusion matrices

Setup

Obs	Instruction	Tasks
1.	Clone Ryan Marchildon's repo	$ git clone [e-mail]:RyanMarchildon/tfjs-torontoai-lecture.git
2-a.	Webserver - JS	$ npm install local-web-server -g
2-b.	Webserver - Py	$ python3 -m http.server

Getting Started

1. Fundamental

   • tf.js /index.html

     <script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@1.0.0/dist/tf.min.js"></script>
     

2. Tensors

   • multi-dimensional array vector

3. CNN

   • tfjs-vis

4. Memory management

5. MNIST data with tf.js

   • loss( ) and confusion matrices

   • local host this model due to cross origin reference sharing

   • need to add script call below in html file

     <!-- Import TensorFlow.js -->
     <script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@1.0.0/dist/tf.min.js"></script>
     <!-- Import tfjs-vis -->
     <script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs-vis@1.0.2/dist/tfjs-vis.umd.min.js"></script>
     
     <!-- Import the data file -->
     <script src="data.js" type="module"></script>
     
     <!-- Import the main script file -->
     <script src="main.js" type="module"></script>
     

Futher Resources

Readings

NIP's Visual Reasoning by Progressive Module Networks - Seung Wook Kim

Audio Prototype

OpenAi MuseNet

---

Rangle.io's take on building AI applications with tensorflow.js