@tensorflow/tfjs-layers - npm Package Compare versions

Comparing version 0.6.6 to 0.6.7

dist/engine/topology.d.ts

@@ -111,2 +111,4 @@ import { DataType, Scalar, serialization, Tensor } from '@tensorflow/tfjs-core';
     apply(inputs: Tensor | Tensor[] | SymbolicTensor | SymbolicTensor[], kwargs?: Kwargs): Tensor | Tensor[] | SymbolicTensor | SymbolicTensor[];
+    readonly outputShape: Shape | Shape[];
+    countParams(): number;
     build(inputShape: Shape | Shape[]): void;
@@ -113,0 +115,0 @@ getWeights(trainableOnly?: boolean): Tensor[];

dist/engine/training.d.ts

@@ -0,1 +1,2 @@
+import * as tfc from '@tensorflow/tfjs-core';
 import { io, ModelPredictConfig, Optimizer, Scalar, Tensor, Tensor1D } from '@tensorflow/tfjs-core';
@@ -59,3 +60,3 @@ import { Callback, CustomCallbackConfig, History } from '../callbacks';
 }
-export declare class Model extends Container {
+export declare class Model extends Container implements tfc.InferenceModel {
     static className: string;
@@ -81,2 +82,3 @@ optimizer: Optimizer;
     constructor(config: ContainerConfig);
+    summary(lineLength?: number, positions?: number[], printFn?: (message?: any, ...optionalParams: any[]) => void): void;
     compile(config: ModelCompileConfig): void;
@@ -86,2 +88,4 @@ private checkTrainableWeightsConsistency();
     private checkNumSamples(ins, batchSize?, steps?, stepsName?);
+    execute(inputs: Tensor | Tensor[] | NamedTensorMap, outputs: string | string[]): Tensor | Tensor[];
+    private retrieveSymbolicTensors(symbolicTensorNames);
     private predictLoop(ins, batchSize?, verbose?);
@@ -88,0 +92,0 @@ predict(x: Tensor | Tensor[], config?: ModelPredictConfig): Tensor | Tensor[];

dist/engine/training.js

@@ -63,2 +63,3 @@ "use strict";
 var generic_utils_1 = require("../utils/generic_utils");
+var layer_utils_1 = require("../utils/layer_utils");
 var math_utils_1 = require("../utils/math_utils");
@@ -354,2 +355,11 @@ var executor_1 = require("./executor");
     }
+    Model.prototype.summary = function (lineLength, positions, printFn) {
+        if (printFn === void 0) { printFn = console.log; }
+        if (!this.built) {
+            throw new errors_1.ValueError("This model has never been called, thus its weights have not been " +
+                "created yet. So no summary can be displayed. Build the model " +
+                "first (e.g., by calling it on some test data).");
+        }
+        layer_utils_1.printSummary(this, lineLength, positions, printFn);
+    };
     Model.prototype.compile = function (config) {
@@ -559,2 +569,72 @@ var _this = this;
     };
+    Model.prototype.execute = function (inputs, outputs) {
+        if (Array.isArray(outputs) && outputs.length === 0) {
+            throw new errors_1.ValueError('`outputs` is an empty Array, which is not allowed.');
+        }
+        var outputsIsArray = Array.isArray(outputs);
+        var outputNames = (outputsIsArray ? outputs :
+            [outputs]);
+        var outputSymbolicTensors = this.retrieveSymbolicTensors(outputNames);
+        var feedDict = new executor_1.FeedDict();
+        if (inputs instanceof tfjs_core_1.Tensor) {
+            inputs = [inputs];
+        }
+        if (Array.isArray(inputs)) {
+            if (inputs.length !== this.inputs.length) {
+                throw new errors_1.ValueError("The number of inputs provided (" + inputs.length + ") " +
+                    "does not match the number of inputs of this model " +
+                    ("(" + this.inputs.length + ")."));
+            }
+            for (var i = 0; i < this.inputs.length; ++i) {
+                feedDict.add(this.inputs[i], inputs[i]);
+            }
+        }
+        else {
+            for (var _i = 0, _a = this.inputs; _i < _a.length; _i++) {
+                var input = _a[_i];
+                var tensorValue = inputs[input.name];
+                if (tensorValue == null) {
+                    throw new errors_1.ValueError("No value is provided for the model's input " + input.name);
+                }
+                feedDict.add(input, tensorValue);
+            }
+        }
+        var executeOutputs = executor_1.execute(outputSymbolicTensors, feedDict);
+        return outputsIsArray ? executeOutputs : executeOutputs[0];
+    };
+    Model.prototype.retrieveSymbolicTensors = function (symbolicTensorNames) {
+        var outputSymbolicTensors = generic_utils_1.pyListRepeat(null, symbolicTensorNames.length);
+        var outputsRemaining = symbolicTensorNames.length;
+        for (var _i = 0, _a = this.layers; _i < _a.length; _i++) {
+            var layer = _a[_i];
+            var layerOutputs = Array.isArray(layer.output) ?
+                layer.output :
+                [layer.output];
+            var layerOutputNames = layerOutputs.map(function (output) { return output.name; });
+            for (var i = 0; i < symbolicTensorNames.length; ++i) {
+                var index = layerOutputNames.indexOf(symbolicTensorNames[i]);
+                if (index !== -1) {
+                    outputSymbolicTensors[i] = layerOutputs[index];
+                    outputsRemaining--;
+                }
+                if (outputsRemaining === 0) {
+                    break;
+                }
+            }
+            if (outputsRemaining === 0) {
+                break;
+            }
+        }
+        if (outputsRemaining > 0) {
+            var remainingNames_1 = [];
+            outputSymbolicTensors.forEach(function (tensor, i) {
+                if (tensor == null) {
+                    remainingNames_1.push(symbolicTensorNames[i]);
+                }
+            });
+            throw new errors_1.ValueError("Cannot find SymbolicTensors for output name(s): " +
+                ("" + JSON.stringify(remainingNames_1)));
+        }
+        return outputSymbolicTensors;
+    };
     Model.prototype.predictLoop = function (ins, batchSize, verbose) {
@@ -1076,2 +1156,5 @@ var _this = this;
     __decorate([
+        tfjs_core_1.doc({ heading: 'Models', subheading: 'Classes' })
+    ], Model.prototype, "summary", null);
+    __decorate([
         tfjs_core_1.doc({ heading: 'Models', subheading: 'Classes', configParamIndices: [0] })
@@ -1091,2 +1174,5 @@ ], Model.prototype, "compile", null);
     ], Model.prototype, "fit", null);
+    __decorate([
+        tfjs_core_1.doc({ heading: 'Models', subheading: 'Classes', configParamIndices: [1] })
+    ], Model.prototype, "save", null);
     Model = __decorate([
@@ -1093,0 +1179,0 @@ tfjs_core_1.doc({ heading: 'Models', subheading: 'Classes' })

dist/layers/convolutional_depthwise.d.ts

@@ -7,3 +7,3 @@ import { Tensor } from '@tensorflow/tfjs-core';
 import { Kwargs, Shape } from '../types';
-import { BaseConvLayerConfig, Conv2D } from './convolutional';
+import { BaseConv, BaseConvLayerConfig } from './convolutional';
 export declare function depthwiseConv2d(x: Tensor, depthwiseKernel: Tensor, strides?: [number, number], padding?: string, dataFormat?: DataFormat, dilationRate?: [number, number]): Tensor;
@@ -17,3 +17,3 @@ export interface DepthwiseConv2DLayerConfig extends BaseConvLayerConfig {
 }
-export declare class DepthwiseConv2D extends Conv2D {
+export declare class DepthwiseConv2D extends BaseConv {
     static className: string;
@@ -20,0 +20,0 @@ private readonly depthMultiplier;

dist/layers/convolutional_depthwise.js

@@ -53,3 +53,3 @@ "use strict";
     function DepthwiseConv2D(config) {
-        var _this = _super.call(this, config) || this;
+        var _this = _super.call(this, 2, config) || this;
         _this.depthwiseKernel = null;
@@ -119,5 +119,5 @@ _this.depthMultiplier =
     return DepthwiseConv2D;
-}(convolutional_1.Conv2D));
+}(convolutional_1.BaseConv));
 exports.DepthwiseConv2D = DepthwiseConv2D;
 tfjs_core_1.serialization.SerializationMap.register(DepthwiseConv2D);
 //# sourceMappingURL=convolutional_depthwise.js.map

dist/layers/convolutional.d.ts

@@ -34,5 +34,4 @@ import { serialization, Tensor } from '@tensorflow/tfjs-core';
 }
-export declare abstract class Conv extends Layer {
+export declare abstract class BaseConv extends Layer {
     protected readonly rank: number;
-    protected readonly filters: number;
     protected readonly kernelSize: number[];
@@ -42,15 +41,20 @@ protected readonly strides: number[];
     protected readonly dataFormat: DataFormat;
-    protected readonly dilationRate: number | [number] | [number, number];
     protected readonly activation: Activation;
     protected readonly useBias: boolean;
-    protected readonly kernelInitializer?: Initializer;
+    protected readonly dilationRate: number | [number] | [number, number];
     protected readonly biasInitializer?: Initializer;
-    protected readonly kernelConstraint?: Constraint;
     protected readonly biasConstraint?: Constraint;
-    protected readonly kernelRegularizer?: Regularizer;
     protected readonly biasRegularizer?: Regularizer;
-    protected kernel: LayerVariable;
     protected bias: LayerVariable;
     readonly DEFAULT_KERNEL_INITIALIZER: InitializerIdentifier;
     readonly DEFAULT_BIAS_INITIALIZER: InitializerIdentifier;
+    constructor(rank: number, config: BaseConvLayerConfig);
+    protected static verifyConfig(config: BaseConvLayerConfig): void;
+}
+export declare abstract class Conv extends BaseConv {
+    protected readonly filters: number;
+    protected kernel: LayerVariable;
+    protected readonly kernelInitializer?: Initializer;
+    protected readonly kernelConstraint?: Constraint;
+    protected readonly kernelRegularizer?: Regularizer;
     constructor(rank: number, config: ConvLayerConfig);
@@ -61,2 +65,3 @@ build(inputShape: Shape | Shape[]): void;
     getConfig(): serialization.ConfigDict;
+    protected static verifyConfig(config: ConvLayerConfig): void;
 }
@@ -67,2 +72,3 @@ export declare class Conv2D extends Conv {
     getConfig(): serialization.ConfigDict;
+    protected static verifyConfig(config: ConvLayerConfig): void;
 }
@@ -113,2 +119,3 @@ export declare class Conv2DTranspose extends Conv2D {
     getConfig(): serialization.ConfigDict;
+    static verifyConfig(config: ConvLayerConfig): void;
 }
@@ -115,0 +122,0 @@ export interface Cropping2DLayerConfig extends LayerConfig {

dist/layers/convolutional.js

@@ -125,10 +125,10 @@ "use strict";
 exports.conv2dWithBias = conv2dWithBias;
-var Conv = (function (_super) {
-    __extends(Conv, _super);
-    function Conv(rank, config) {
+var BaseConv = (function (_super) {
+    __extends(BaseConv, _super);
+    function BaseConv(rank, config) {
         var _this = _super.call(this, config) || this;
-        _this.kernel = null;
         _this.bias = null;
         _this.DEFAULT_KERNEL_INITIALIZER = 'glorotNormal';
         _this.DEFAULT_BIAS_INITIALIZER = 'zeros';
+        BaseConv.verifyConfig(config);
         _this.rank = rank;
@@ -139,3 +139,2 @@ if (_this.rank !== 1 && _this.rank !== 2) {
         }
-        _this.filters = config.filters;
         _this.kernelSize = conv_utils_1.normalizeArray(config.kernelSize, rank, 'kernelSize');
@@ -148,2 +147,9 @@ _this.strides = conv_utils_1.normalizeArray(config.strides == null ? 1 : config.strides, rank, 'strides');
         common_2.checkDataFormat(_this.dataFormat);
+        _this.activation = activations_1.getActivation(config.activation);
+        _this.useBias = config.useBias == null ? true : config.useBias;
+        _this.biasInitializer =
+            initializers_1.getInitializer(config.biasInitializer || _this.DEFAULT_BIAS_INITIALIZER);
+        _this.biasConstraint = constraints_1.getConstraint(config.biasConstraint);
+        _this.biasRegularizer = regularizers_1.getRegularizer(config.biasRegularizer);
+        _this.activityRegularizer = regularizers_1.getRegularizer(config.activityRegularizer);
         _this.dilationRate = config.dilationRate == null ? 1 : config.dilationRate;
@@ -166,12 +172,24 @@ if (_this.rank === 1 &&
         }
-        _this.activation = activations_1.getActivation(config.activation);
-        _this.useBias = config.useBias == null ? true : config.useBias;
+        return _this;
+    }
+    BaseConv.verifyConfig = function (config) {
+        generic_utils.assert('kernelSize' in config, "required key 'kernelSize' not in config");
+        if (typeof config.kernelSize !== 'number' &&
+            !generic_utils.checkArrayTypeAndLength(config.kernelSize, 'number', 1, 2))
+            throw new errors_1.ValueError("BaseConv expects config.kernelSize to be number or number[] with " +
+                ("length 1 or 2, but received " + JSON.stringify(config.kernelSize) + "."));
+    };
+    return BaseConv;
+}(topology_1.Layer));
+exports.BaseConv = BaseConv;
+var Conv = (function (_super) {
+    __extends(Conv, _super);
+    function Conv(rank, config) {
+        var _this = _super.call(this, rank, config) || this;
+        _this.kernel = null;
+        Conv.verifyConfig(config);
+        _this.filters = config.filters;
         _this.kernelInitializer = initializers_1.getInitializer(config.kernelInitializer || _this.DEFAULT_KERNEL_INITIALIZER);
-        _this.biasInitializer =
-            initializers_1.getInitializer(config.biasInitializer || _this.DEFAULT_BIAS_INITIALIZER);
         _this.kernelConstraint = constraints_1.getConstraint(config.kernelConstraint);
-        _this.biasConstraint = constraints_1.getConstraint(config.biasConstraint);
         _this.kernelRegularizer = regularizers_1.getRegularizer(config.kernelRegularizer);
-        _this.biasRegularizer = regularizers_1.getRegularizer(config.biasRegularizer);
-        _this.activityRegularizer = regularizers_1.getRegularizer(config.activityRegularizer);
         return _this;
@@ -262,4 +280,11 @@ }
     };
+    Conv.verifyConfig = function (config) {
+        if (!('filters' in config) || typeof config.filters !== 'number' ||
+            config.filters < 1) {
+            throw new errors_1.ValueError("Convolution layer expected config.filters to be a 'number' > 0 " +
+                ("but got " + JSON.stringify(config.filters)));
+        }
+    };
     return Conv;
-}(topology_1.Layer));
+}(BaseConv));
 exports.Conv = Conv;
@@ -269,3 +294,5 @@ var Conv2D = (function (_super) {
     function Conv2D(config) {
-        return _super.call(this, 2, config) || this;
+        var _this = _super.call(this, 2, config) || this;
+        Conv2D.verifyConfig(config);
+        return _this;
     }
@@ -277,2 +304,8 @@ Conv2D.prototype.getConfig = function () {
     };
+    Conv2D.verifyConfig = function (config) {
+        if ((typeof config.kernelSize !== 'number') &&
+            !generic_utils.checkArrayTypeAndLength(config.kernelSize, 'number', 1, 2))
+            throw new errors_1.ValueError("Conv2D expects config.kernelSize to be number or number[] with " +
+                ("length 1 or 2, but received " + JSON.stringify(config.kernelSize) + "."));
+    };
     Conv2D.className = 'Conv2D';
@@ -530,2 +563,3 @@ return Conv2D;
         var _this = _super.call(this, 1, config) || this;
+        Conv1D.verifyConfig(config);
         _this.inputSpec = [{ ndim: 3 }];
@@ -540,2 +574,8 @@ return _this;
     };
+    Conv1D.verifyConfig = function (config) {
+        if (typeof config.kernelSize !== 'number' &&
+            !generic_utils.checkArrayTypeAndLength(config.kernelSize, 'number', 1, 1))
+            throw new errors_1.ValueError("Conv1D expects config.kernelSize to be number or number[] with " +
+                ("length 1, but received " + JSON.stringify(config.kernelSize) + "."));
+    };
     Conv1D.className = 'Conv1D';
@@ -542,0 +582,0 @@ return Conv1D;

dist/layers/pooling.d.ts

@@ -33,3 +33,3 @@ import { serialization, Tensor } from '@tensorflow/tfjs-core';
     poolSize?: number | [number, number];
-    strides?: [number, number];
+    strides?: number | [number, number];
     padding?: PaddingMode;
@@ -36,0 +36,0 @@ dataFormat?: DataFormat;

dist/layers/pooling.js

@@ -169,3 +169,16 @@ "use strict";
             [config.poolSize, config.poolSize];
-        _this.strides = config.strides == null ? _this.poolSize : config.strides;
+        if (config.strides == null) {
+            _this.strides = _this.poolSize;
+        }
+        else if (Array.isArray(config.strides)) {
+            if (config.strides.length !== 2) {
+                throw new errors_1.ValueError("If the strides property of a 2D pooling layer is an Array, " +
+                    "it is expected to have a length of 2, but received length " +
+                    (config.strides.length + "."));
+            }
+            _this.strides = config.strides;
+        }
+        else {
+            _this.strides = [config.strides, config.strides];
+        }
         _this.padding = config.padding == null ? 'valid' : config.padding;
@@ -172,0 +185,0 @@ _this.dataFormat =

dist/layers/recurrent.d.ts

@@ -175,2 +175,3 @@ import * as tfc from '@tensorflow/tfjs-core';
 export interface GRULayerConfig extends SimpleRNNLayerConfig {
+    recurrentActivation?: string;
     implementation?: number;
@@ -184,2 +185,3 @@ }
     readonly activation: Activation;
+    readonly recurrentActivation: Activation;
     readonly useBias: boolean;
@@ -240,2 +242,3 @@ readonly kernelInitializer: Initializer;
 export interface LSTMLayerConfig extends SimpleRNNLayerConfig {
+    recurrentActivation?: string;
     unitForgetBias?: boolean;
@@ -250,2 +253,3 @@ implementation?: 1 | 2;
     readonly activation: Activation;
+    readonly recurrentActivation: Activation;
     readonly useBias: boolean;
@@ -252,0 +256,0 @@ readonly kernelInitializer: Initializer;

dist/models.d.ts

@@ -32,2 +32,4 @@ import { io, Scalar, serialization, Tensor } from '@tensorflow/tfjs-core';
     build(inputShape?: Shape | Shape[]): void;
+    countParams(): number;
+    summary(lineLength?: number, positions?: number[], printFn?: (message?: any, ...optionalParams: any[]) => void): void;
     setWeights(weights: Tensor[]): void;
@@ -34,0 +36,0 @@ updatable: boolean;

dist/models.js

@@ -216,2 +216,3 @@ "use strict";
             }
+            this.inboundNodes = [];
             new topology_1.Node({
@@ -295,2 +296,15 @@ outboundLayer: this,
     };
+    Sequential.prototype.countParams = function () {
+        if (!this.built) {
+            this.build();
+        }
+        return _super.prototype.countParams.call(this);
+    };
+    Sequential.prototype.summary = function (lineLength, positions, printFn) {
+        if (printFn === void 0) { printFn = console.log; }
+        if (!this.built) {
+            this.build();
+        }
+        _super.prototype.summary.call(this, lineLength, positions, printFn);
+    };
     Sequential.prototype.setWeights = function (weights) {
@@ -390,2 +404,5 @@ if (this.model == null) {
     __decorate([
+        tfjs_core_1.doc({ heading: 'Models', subheading: 'Classes' })
+    ], Sequential.prototype, "summary", null);
+    __decorate([
         tfjs_core_1.doc({ heading: 'Models', subheading: 'Classes', configParamIndices: [2] })
@@ -392,0 +409,0 @@ ], Sequential.prototype, "evaluate", null);

dist/optimizers.js

@@ -9,3 +9,5 @@ "use strict";
         'Adagrad': function () { return tfjs_core_1.train.adagrad(.01); },
+        'Adadelta': function () { return tfjs_core_1.train.adadelta(1.0, 0.95, K.epsilon()); },
         'Adam': function () { return tfjs_core_1.train.adam(.001, .9, .999, K.epsilon()); },
+        'Adamax': function () { return tfjs_core_1.train.adamax(0.002, .9, .999, K.epsilon(), 0.0); },
         'RMSProp': function () { return tfjs_core_1.train.rmsprop(.001, .9, null, K.epsilon()); },
@@ -15,3 +17,5 @@ 'SGD': function () { return tfjs_core_1.train.sgd(.01); }
     optimizerMap['adagrad'] = optimizerMap['Adagrad'];
+    optimizerMap['adadelta'] = optimizerMap['Adadelta'];
     optimizerMap['adam'] = optimizerMap['Adam'];
+    optimizerMap['adamax'] = optimizerMap['Adamax'];
     optimizerMap['rmsprop'] = optimizerMap['RMSProp'];
@@ -18,0 +22,0 @@ optimizerMap['sgd'] = optimizerMap['SGD'];

dist/types.d.ts

@@ -13,3 +13,3 @@ import { DataType, Scalar, Tensor } from '@tensorflow/tfjs-core';
     readonly id: number;
-    readonly name?: string;
+    readonly name: string;
     readonly originalName?: string;
@@ -16,0 +16,0 @@ readonly rank: number;

dist/utils/generic_utils.d.ts

 import { DataType, serialization, Tensor } from '@tensorflow/tfjs-core';
 import { Shape } from '../types';
+import { LayerVariable } from '../variables';
 export declare function pyListRepeat(value: any, numValues: number): any[];
@@ -28,1 +29,3 @@ export declare function assert(val: boolean, message?: string): void;
 export declare function checkStringTypeUnionValue(values: string[], label: string, value: string): void;
+export declare function checkArrayTypeAndLength(x: any, expectedType: string, minLength?: number, maxLength?: number): boolean;
+export declare function countParamsInWeights(weights: LayerVariable[]): number;

dist/utils/generic_utils.js

@@ -288,2 +288,25 @@ "use strict";
 exports.checkStringTypeUnionValue = checkStringTypeUnionValue;
+function checkArrayTypeAndLength(x, expectedType, minLength, maxLength) {
+    if (minLength === void 0) { minLength = 0; }
+    if (maxLength === void 0) { maxLength = Infinity; }
+    assert(minLength >= 0);
+    assert(maxLength >= minLength);
+    return (Array.isArray(x) && x.length >= minLength && x.length <= maxLength &&
+        x.every(function (e) { return typeof e === expectedType; }));
+}
+exports.checkArrayTypeAndLength = checkArrayTypeAndLength;
+function countParamsInWeights(weights) {
+    var count = 0;
+    for (var _i = 0, weights_1 = weights; _i < weights_1.length; _i++) {
+        var weight = weights_1[_i];
+        if (weight.shape.length === 0) {
+            count += 1;
+        }
+        else {
+            count += weight.shape.reduce(function (a, b) { return a * b; });
+        }
+    }
+    return count;
+}
+exports.countParamsInWeights = countParamsInWeights;
 //# sourceMappingURL=generic_utils.js.map

dist/utils/test_utils.js

@@ -25,3 +25,3 @@ "use strict";
 function describeMathCPUAndGPU(testName, tests) {
-    jasmine_util_1.describeWithFlags(testName, {}, function () {
+    jasmine_util_1.describeWithFlags(testName, tfjs_core_1.test_util.ALL_ENVS, function () {
         beforeEach(function () {
@@ -28,0 +28,0 @@ tfjs_backend_1.disposeScalarCache();

dist/version.d.ts

@@ -1,2 +0,2 @@
-declare const version = "0.6.6";
+declare const version = "0.6.7";
 export { version };

dist/version.js

 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-var version = '0.6.6';
+var version = '0.6.7';
 exports.version = version;
 //# sourceMappingURL=version.js.map

package.json

 {
   "name": "@tensorflow/tfjs-layers",
-  "version": "0.6.6",
+  "version": "0.6.7",
   "description": "TensorFlow layers API in JavaScript",
@@ -8,10 +8,10 @@ "private": false,
   "types": "dist/index.d.ts",
-  "jsnext:main": "dist-es6/index.js",
-  "module": "dist-es6/index.js",
+  "jsnext:main": "dist/tf-layers.esm.js",
+  "module": "dist/tf-layers.esm.js",
+  "jsdelivr": "dist/tf-layers.min.js",
+  "unpkg": "dist/tf-layers.min.js",
   "devDependencies": {
-    "@tensorflow/tfjs-core": "~0.11.6",
+    "@tensorflow/tfjs-core": "~0.11.9",
     "@types/jasmine": "~2.5.53",
-    "browserify": "~16.1.0",
     "clang-format": "~1.2.2",
-    "cross-spawn": "~5.1.0",
     "http-server": "~0.10.0",
@@ -25,5 +25,2 @@ "jasmine-core": "~3.1.0",
     "karma-typescript": "^3.0.12",
-    "karma-typescript-es6-transform": "~1.0.4",
-    "minimist": "~1.2.0",
-    "opn": "~5.1.0",
     "rimraf": "~2.6.2",
@@ -34,12 +31,9 @@ "rollup": "^0.58.2",
     "rollup-plugin-typescript2": "0.13.0",
-    "rollup-plugin-uglify": "^4.0.0",
-    "tsify": "~3.0.1",
+    "rollup-plugin-uglify": "~3.0.0",
     "tslint": "~5.6.0",
-    "typedoc": "~0.8.0",
     "typescript": "2.8.3",
-    "watchify": "~3.9.0",
     "yalc": "~1.0.0-pre.21"
   },
   "scripts": {
-    "build": "tsc --project tsconfig-es5.json && tsc",
+    "build": "tsc",
     "build-npm": "./scripts/build-npm.sh",
@@ -56,4 +50,4 @@ "format": "./tools/clang_format_ts.sh",
   "peerDependencies": {
-    "@tensorflow/tfjs-core": "~0.11.6"
+    "@tensorflow/tfjs-core": "~0.11.9"
   }
 }

tsconfig.json

 {
   "compilerOptions": {
-    "module": "ES2015",
+    "module": "commonjs",
     "moduleResolution": "node",
@@ -12,3 +12,3 @@ "noImplicitAny": true,
     "lib": ["es2015", "dom"],
-    "outDir": "./dist-es6",
+    "outDir": "./dist",
     "noUnusedLocals": true,
@@ -15,0 +15,0 @@ "noImplicitReturns": true,

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Dev dependency count

20

decreased by-28.57%

Number of medium supply chain risk alerts

26

decreased by-33.33%

Worsened metrics

Total package byte prevSize

1543465

decreased by-55.58%

Number of package files

85

decreased by-71.57%

Lines of code

26850

decreased by-48.26%

No dependency changes