With some of the changes pushed up recently, NeuralNetworkGPU lagged behind. This was mainly the prepending of "_" to internal-only methods, as well as moving things to a trainOpts object. This meant that the GPU was not able to work at all. This catches things up.

Author: perkyguy

browser.js

@@ -464,9 +464,9 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
 
 
   _createClass(NeuralNetworkGPU, [{
-    key: 'initialize',
-    value: function initialize(sizes) {
-      _get(NeuralNetworkGPU.prototype.__proto__ || Object.getPrototypeOf(NeuralNetworkGPU.prototype), 'initialize', this).call(this, sizes);
+    key: '_initialize',
+    value: function _initialize(sizes) {
+      _get(NeuralNetworkGPU.prototype.__proto__ || Object.getPrototypeOf(NeuralNetworkGPU.prototype), '_initialize', this).call(this, sizes);
       this.buildRunInput();
       this.buildCalculateDeltas();
       this.buildGetChanges();
@@ -485,9 +485,9 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
      */
 
   }, {
-    key: 'trainPattern',
-    value: function trainPattern(input, target, learningRate) {
-      learningRate = learningRate || this.learningRate;
+    key: '_trainPattern',
+    value: function _trainPattern(input, target, learningRate) {
+      learningRate = learningRate || this.trainOpts.learningRate;
       // forward propagate
       this.runInput(input);
 
@@ -639,7 +639,7 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
     key: 'getChanges',
     value: function getChanges(learningRate) {
       for (var layer = 1; layer <= this.outputLayer; layer++) {
-        var output = this.changesPropagate[layer](this.outputs[layer - 1], this.deltas[layer], this.weights[layer], this.changes[layer], learningRate, this.momentum);
+        var output = this.changesPropagate[layer](this.outputs[layer - 1], this.deltas[layer], this.weights[layer], this.changes[layer], learningRate, this.trainOpts.momentum);
 
         this.changes[layer] = output.changes;
         this.weights[layer] = output.weights;
@@ -701,8 +701,8 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
      */
 
   }, {
-    key: 'formatData',
-    value: function formatData(data) {
+    key: '_formatData',
+    value: function _formatData(data) {
       var _this2 = this;
 
       if (!Array.isArray(data)) {
@@ -1286,7 +1286,7 @@ var NeuralNetwork = function () {
   }, {
     key: '_prepTraining',
     value: function _prepTraining(data, options) {
-      this.updateTrainingOptions(options);
+      this._updateTrainingOptions(options);
       data = this._formatData(data);
       var endTime = Date.now() + this.trainOpts.timeout;
 

browser.min.js

@@ -22,8 +22,8 @@ export default class NeuralNetworkGPU extends NeuralNetwork {
    *
    * @param {Number[]} sizes
    */
-  initialize(sizes) {
-    super.initialize(sizes);
+  _initialize(sizes) {
+    super._initialize(sizes);
     this.buildRunInput();
     this.buildCalculateDeltas();
     this.buildGetChanges();
@@ -39,8 +39,8 @@ export default class NeuralNetworkGPU extends NeuralNetwork {
    * @param target
    * @param learningRate
    */
-  trainPattern(input, target, learningRate) {
-    learningRate = learningRate || this.learningRate;
+  _trainPattern(input, target, learningRate) {
+    learningRate = learningRate || this.trainOpts.learningRate;
     // forward propagate
     this.runInput(input);
 
@@ -205,7 +205,7 @@ export default class NeuralNetworkGPU extends NeuralNetwork {
         this.weights[layer],
         this.changes[layer],
         learningRate,
-        this.momentum
+        this.trainOpts.momentum
       );
 
       this.changes[layer] = output.changes;
@@ -264,7 +264,7 @@ export default class NeuralNetworkGPU extends NeuralNetwork {
    * @param data
    * @returns {*}
    */
-  formatData(data) {
+  _formatData(data) {
     if (!Array.isArray(data)) { // turn stream datum into array
       let tmp = [];
       tmp.push(data);