bump to v1.0.2 (#4)

* added various fixes and improvements. bump to v0.1.1

* set fetch to global this closes issue #3

* version bump and changes to readme file

Author: bonastreyair

CHANGELOG.md

@@ -1,4 +1,24 @@
 # Changelog
+## [0.1.2] - 2020-04-12
+### Added
+ * Improvements on `README` file
+
+### Fixed
+ * Loading model error `response.arrayBuffer is not a function` found in issue [#3](https://github.com/bonastreyair/node-red-contrib-teachable-machine/issues/3)
+
+## [0.1.1] - 2020-04-11
+### Added
+ * Comments in the code
+
+### Changed
+ * Downgraded `@tensorflow/tfjs-node` from version `v1.4.0` to `v1.3.1` for better compatibility
+ * Output has changed from `checkbox` to a `list`, you can now select `Best predictions` or `All predictions`
+ * Code cleaning
+
+### Fixed
+ * WebGL loading error in JSDOM
+ * When installing the node -> `npm WARN @teachablemachine/image@0.8.4 requires a peer of @tensorflow/tfjs@1.3.1 but none is installed`
+
 
 ## [0.1.1] - 2020-04-11
 ### Added

README.md

@@ -1,6 +1,31 @@
 # node-red-contrib-teachable-machine
-A TensorFlow.js node for Node-RED that enables to run custom trained Teachable Machine image classification online models.
+A Node-RED](https://nodered.org) node based in [tensorFlow.js](https://www.tensorflow.org/js) that enables to run custom trained models using [Teachable Machine](https://teachablemachine.withgoogle.com/train/image) tool.
 
-More content to be generated...
+## Install
+Use `Manage Palette` option in `Node-RED` Menu or run the following command in your `Node-RED` user directory (typically `~/.node-red`).
+```
+npm install node-red-contrib-teachable-machine
+```
 
-Inspired by [@dceejay](https://github.com/dceejay) with [node-red-contrib-tfjs-coco-ssd](https://github.com/dceejay/tfjs-coco-ssd/)
+## Node usage
+### Step 1
+Go to [Teachable Machine](https://teachablemachine.withgoogle.com/train/image) and follow the steps to train your custom classification model.
+
+### Step 2 
+Once trained the model, click on the `Export Model` button and select `Tensorflow.js` format. Then upload your trained model for free and copy the generated URL. 
+![upload_teachable](https://user-images.githubusercontent.com/37800834/79056723-8431a100-7c59-11ea-9488-346f4f8e6004.png)
+
+### Step 3
+Paste the saved URL into the node. That URL hosts all the information to load your trained model.
+![model_URL](https://user-images.githubusercontent.com/37800834/79056644-ec33b780-7c58-11ea-9b69-8e8d4fbfda0c.png)
+
+### Step 4 
+In 1Node-RED1 Send a buffered image to the node.
+
+## Requirements
+* `Node-RED v1.0.0+`
+* Supported OS: MacOSX Catalina, Windows 10, Ubuntu 18.04
+
+**NOTE**: This not is not working using official `docker nodered/node-red` [image](https://hub.docker.com/r/nodered/node-red/) since it is based on Alpine and it has an incompatibility found in this [issue](https://github.com/tensorflow/tfjs/issues/1425).
+
+Inspired by [@dceejay](https://github.com/dceejay) with [node-red-contrib-tfjs-coco-ssd](https://github.com/dceejay/tfjs-coco-ssd/).

package.json

@@ -1,6 +1,6 @@
 {
   "name": "node-red-contrib-teachable-machine",
-  "version": "0.1.1",
+  "version": "0.1.2",
   "description": "A node to integrate teachable machine models easily.",
   "dependencies": {
     "@tensorflow/tfjs-node": "1.3.1",

teachable_machine.js

@@ -9,7 +9,7 @@ module.exports = function(RED) {
         // Require basic libraries
         tmImage = require('@teachablemachine/image');
         Canvas = require('canvas');
-        fetch  = require('node-fetch');
+        global.fetch = require('node-fetch');
         // Teachable Machine needs global scope of HTMLVideoElement class to do a check
         global.HTMLVideoElement = class HTMLVideoElement {};