further debugging with Garrett, the genius.

Author: robertleeplummerjr

browser.js

@@ -528,6 +528,7 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
           output: [this.sizes[layer]],
           outputToTexture: true,
           outputImmutable: true,
+          hardcodeConstants: true,
           constants: {
             size: this.sizes[layer - 1]
           }
@@ -586,7 +587,8 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
           }, {
             output: [this.sizes[layer]],
             outputToTexture: true,
-            outputImmutable: true
+            outputImmutable: true,
+            hardcodeConstants: true
           });
         } else {
           this.backwardPropagate[layer] = this.gpu.createKernelMap({
@@ -599,6 +601,7 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
             output: [this.sizes[layer]],
             outputToTexture: true,
             outputImmutable: true,
+            hardcodeConstants: true,
             constants: {
               size: this.deltas[layer + 1].length
             }
@@ -628,15 +631,18 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
         this.changesPropagate[layer] = this.gpu.createKernelMap({
           weights: _gpu2.default.alias('addWeights', addWeights),
           changes: _gpu2.default.alias('calcChanges', calcChanges)
-        }, function (previousOutputs, deltas, weights, changes, learningRate, momentum) {
-          var change = calcChanges(changes, deltas, previousOutputs, learningRate, momentum);
+        }, function (previousOutputs, deltas, weights, changes) {
+          var change = calcChanges(changes, deltas, previousOutputs);
 
           return addWeights(change, weights);
         }, {
           output: [this.sizes[layer - 1], this.sizes[layer]],
           outputToTexture: true,
+          hardcodeConstants: true,
           constants: {
-            size: this.outputs[layer - 1].length
+            size: this.outputs[layer - 1].length,
+            learningRate: this.trainOpts.learningRate,
+            momentum: this.trainOpts.momentum
           }
         });
       }
@@ -645,7 +651,7 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
     key: 'getChanges',
     value: function getChanges() {
       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], this.trainOpts.learningRate, this.trainOpts.momentum);
+        var output = this.changesPropagate[layer](this.outputs[layer - 1], this.deltas[layer], this.weights[layer], this.changes[layer]);
 
         this.changes[layer] = output.changes;
         this.weights[layer] = output.weights;
@@ -658,15 +664,19 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
         this.biasesPropagate[layer] = this.gpu.createKernel(addBiases, {
           output: [this.sizes[layer]],
           outputToTexture: true,
-          outputImmutable: true
+          outputImmutable: true,
+          hardcodeConstants: true,
+          constants: {
+            learningRate: this.trainOpts.learningRate
+          }
         });
       }
     }
   }, {
     key: 'changeBiases',
     value: function changeBiases() {
       for (var layer = 1; layer <= this.outputLayer; layer++) {
-        this.biases[layer] = this.biasesPropagate[layer](this.biases[layer], this.deltas[layer], this.trainOpts.learningRate);
+        this.biases[layer] = this.biasesPropagate[layer](this.biases[layer], this.deltas[layer]);
       }
     }
   }, {
@@ -753,14 +763,16 @@ var NeuralNetworkGPU = function (_NeuralNetwork) {
       }, {
         output: [data[0].input.length],
         outputToTexture: true,
-        outputImmutable: true
+        outputImmutable: true,
+        hardcodeConstants: true
       });
       var texturizeOutputData = this.gpu.createKernel(function (value) {
         return value[this.thread.x];
       }, {
         output: [data[0].output.length],
         outputToTexture: true,
-        outputImmutable: true
+        outputImmutable: true,
+        hardcodeConstants: true
       });
       data.forEach(function (d) {
         d.size = { input: d.input.length, output: d.output.length };
@@ -850,16 +862,16 @@ function calcError(nextWeights, nextDeltas) {
   return error;
 }
 
-function calcChanges(previousChanges, deltas, previousOutputs, learningRate, momentum) {
-  return learningRate * deltas[this.thread.y] * previousOutputs[this.thread.x] + momentum * previousChanges[this.thread.y][this.thread.x];
+function calcChanges(previousChanges, deltas, previousOutputs) {
+  return this.constants.learningRate * deltas[this.thread.y] * previousOutputs[this.thread.x] + this.constants.momentum * previousChanges[this.thread.y][this.thread.x];
 }
 
 function addWeights(change, weights) {
   return change + weights[this.thread.y][this.thread.x];
 }
 
-function addBiases(biases, deltas, learningRate) {
-  return biases[this.thread.x] + deltas[this.thread.x] * learningRate;
+function addBiases(biases, deltas) {
+  return biases[this.thread.x] + deltas[this.thread.x] * this.constants.learningRate;
 }
 
 // mean squared error, reimplemented for GPU
@@ -1324,7 +1336,7 @@ var NeuralNetwork = function () {
      *
      * @param data
      * @param options
-     * @private
+     * @protected
      * @return {{runTrainingTick: function, status: {error: number, iterations: number}}}
      */
 
@@ -1359,8 +1371,6 @@ var NeuralNetwork = function () {
   }, {
     key: 'train',
     value: function train(data) {
-      var _this4 = this;
-
       var options = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : {};
 
       var status = void 0,
@@ -1373,11 +1383,7 @@ var NeuralNetwork = function () {
       endTime = _prepTraining2.endTime;
 
 
-      var tick = function tick() {
-        _this4._trainingTick(data, status, endTime);
-        setTimeout(tick, 0);
-      };
-      tick();
+      while (this._trainingTick(data, status, endTime)) {}
       return status;
     }
 
@@ -1393,7 +1399,7 @@ var NeuralNetwork = function () {
   }, {
     key: 'trainAsync',
     value: function trainAsync(data) {
-      var _this5 = this;
+      var _this4 = this;
 
       var options = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : {};
 
@@ -1409,10 +1415,10 @@ var NeuralNetwork = function () {
 
       return new Promise(function (resolve, reject) {
         try {
-          var thawedTrain = new _thaw2.default(new Array(_this5.trainOpts.iterations), {
+          var thawedTrain = new _thaw2.default(new Array(_this4.trainOpts.iterations), {
             delay: true,
             each: function each() {
-              return _this5._trainingTick(data, status, endTime) || thawedTrain.stop();
+              return _this4._trainingTick(data, status, endTime) || thawedTrain.stop();
             },
             done: function done() {
               return resolve(status);
@@ -1593,7 +1599,7 @@ var NeuralNetwork = function () {
   }, {
     key: '_formatData',
     value: function _formatData(data) {
-      var _this6 = this;
+      var _this5 = this;
 
       if (!Array.isArray(data)) {
         // turn stream datum into array
@@ -1610,7 +1616,7 @@ var NeuralNetwork = function () {
           }));
         }
         data = data.map(function (datum) {
-          var array = _lookup2.default.toArray(_this6.inputLookup, datum.input);
+          var array = _lookup2.default.toArray(_this5.inputLookup, datum.input);
           return Object.assign({}, datum, { input: array });
         }, this);
       }
@@ -1622,7 +1628,7 @@ var NeuralNetwork = function () {
           }));
         }
         data = data.map(function (datum) {
-          var array = _lookup2.default.toArray(_this6.outputLookup, datum.output);
+          var array = _lookup2.default.toArray(_this5.outputLookup, datum.output);
           return Object.assign({}, datum, { output: array });
         }, this);
       }
@@ -1643,7 +1649,7 @@ var NeuralNetwork = function () {
   }, {
     key: 'test',
     value: function test(data) {
-      var _this7 = this;
+      var _this6 = this;
 
       data = this._formatData(data);
 
@@ -1662,13 +1668,13 @@ var NeuralNetwork = function () {
       var sum = 0;
 
       var _loop = function _loop(i) {
-        var output = _this7.runInput(data[i].input);
+        var output = _this6.runInput(data[i].input);
         var target = data[i].output;
 
         var actual = void 0,
             expected = void 0;
         if (isBinary) {
-          actual = output[0] > _this7.binaryThresh ? 1 : 0;
+          actual = output[0] > _this6.binaryThresh ? 1 : 0;
           expected = target[0];
         } else {
           actual = output.indexOf((0, _max2.default)(output));
@@ -1915,15 +1921,15 @@ var NeuralNetwork = function () {
   }, {
     key: 'isRunnable',
     get: function get() {
-      var _this8 = this;
+      var _this7 = this;
 
       if (!this.runInput) {
         console.error('Activation function has not been initialized, did you run train()?');
         return false;
       }
 
       var checkFns = ['sizes', 'outputLayer', 'biases', 'weights', 'outputs', 'deltas', 'changes', 'errors'].filter(function (c) {
-        return _this8[c] === null;
+        return _this7[c] === null;
       });
 
       if (checkFns.length > 0) {