Huge News!Announcing our $40M Series B led by Abstract Ventures.Learn More →
Socket
Sign inDemoInstall
Socket
Sign inDemoInstall

@tensorflow/tfjs-converter

Package Overview
Dependencies
Maintainers
12
Versions
155
Alerts
File Explorer

Advanced tools

License
Socket logo

Install Socket

Detect and block malicious and high-risk dependencies

Install

@tensorflow/tfjs-converter - npm Package Compare versions

Comparing version 2.0.1 to 2.1.0

4

dist/data/types.d.ts

@@ -19,2 +19,3 @@ /**

import { TensorArray } from '../executor/tensor_array';
import { TensorList } from '../executor/tensor_list';
export declare type NamedTensorMap = {

@@ -29,2 +30,5 @@ [key: string]: Tensor;

};
export declare type TensorListMap = {
[key: number]: TensorList;
};
export interface TensorInfo {

@@ -31,0 +35,0 @@ name: string;

9

dist/executor/execution_context.d.ts

@@ -18,4 +18,5 @@ /**

import { Tensor } from '@tensorflow/tfjs-core';
import { NamedTensorsMap, TensorArrayMap } from '../data/types';
import { NamedTensorsMap, TensorArrayMap, TensorListMap } from '../data/types';
import { TensorArray } from './tensor_array';
import { TensorList } from './tensor_list';
import { FunctionExecutor } from './types';

@@ -39,2 +40,3 @@ export interface ExecutionContextInfo {

readonly tensorArrayMap: TensorArrayMap;
readonly tensorListMap: TensorListMap;
readonly functionMap: {

@@ -47,3 +49,3 @@ [key: string]: FunctionExecutor;

private _currentContextIds;
constructor(weightMap: NamedTensorsMap, tensorArrayMap: TensorArrayMap, functionMap?: {
constructor(weightMap?: NamedTensorsMap, tensorArrayMap?: TensorArrayMap, tensorListMap?: TensorListMap, functionMap?: {
[key: string]: FunctionExecutor;

@@ -87,2 +89,5 @@ });

getTensorArray(id: number): TensorArray;
addTensorList(tensorList: TensorList): void;
getTensorList(id: number): TensorList;
dispose(): void;
}

17

dist/executor/execution_context.js

@@ -11,5 +11,6 @@ /**

export class ExecutionContext {
constructor(weightMap, tensorArrayMap, functionMap = {}) {
constructor(weightMap = {}, tensorArrayMap = {}, tensorListMap = {}, functionMap = {}) {
this.weightMap = weightMap;
this.tensorArrayMap = tensorArrayMap;
this.tensorListMap = tensorListMap;
this.functionMap = functionMap;

@@ -122,3 +123,17 @@ this.rootContext = { id: 0, frameName: '', iterationId: 0 };

}
addTensorList(tensorList) {
this.tensorListMap[tensorList.id] = tensorList;
}
getTensorList(id) {
return this.tensorListMap[id];
}
dispose() {
for (const key in this.tensorArrayMap) {
this.tensorArrayMap[key].clearAndClose();
}
for (const key in this.tensorListMap) {
this.tensorListMap[key].clearAndClose();
}
}
}
//# sourceMappingURL=execution_context.js.map

25

dist/executor/graph_executor.d.ts

@@ -19,3 +19,3 @@ /**

import { ISignatureDef } from '../data/compiled_api';
import { NamedTensorsMap, TensorInfo } from '../data/types';
import { NamedTensorsMap, TensorArrayMap, TensorInfo, TensorListMap } from '../data/types';
import { Graph } from '../operations/types';

@@ -82,8 +82,21 @@ import { FunctionExecutor } from './types';

* array.
* @param disableWarning disable the no dynamic ops warning message, default
* to false
*/
executeAsync(inputs: NamedTensorMap, outputs: string[], disableWarning?: boolean): Promise<Tensor[]>;
executeFunctionAsync(inputs: Tensor[]): Promise<Tensor[]>;
executeAsync(inputs: NamedTensorMap, outputs: string[]): Promise<Tensor[]>;
/**
* Executes the inference for given input tensors in Async fashion.
* @param inputs Tensor map for the model inputs, keyed by the input node
* names.
* @param outputs output node name from the Tensorflow model, if no outputs
* are specified, the default outputs of the model would be used. You can
* inspect intermediate nodes of the model by adding them to the outputs
* array.
* @param isFunctionExecution Flag for executing a function.
* @param tensorArrayMap Optional, global TensorArray map by id. Used for
* function execution.
* @param tensorArrayMap Optinal global TensorList map by id. Used for
* function execution.
*/
private _executeAsync;
executeFunctionAsync(inputs: Tensor[], tensorArrayMap: TensorArrayMap, tensorListMap: TensorListMap): Promise<Tensor[]>;
/**
* When there are control flow nodes in the graph, the graph execution use

@@ -93,3 +106,3 @@ * ExecutionContext to keep track of the frames and loop iterators.

* @param context the execution context object for current execution.
* @param disableWarning disable no async op warning
* @param isFunctionExecution Flag for executing a function.
*/

@@ -96,0 +109,0 @@ private executeWithControlFlow;

59

dist/executor/graph_executor.js

@@ -160,4 +160,5 @@ /**

const tensorArrayMap = {};
const tensorListMap = {};
return tidy(() => {
const context = new ExecutionContext(this.weightMap, tensorArrayMap, this.functionExecutorMap);
const context = new ExecutionContext(this.weightMap, tensorArrayMap, tensorListMap, this.functionExecutorMap);
const tensorsMap = Object.assign({}, this.weightMap);

@@ -184,2 +185,6 @@ Object.keys(inputs).forEach(name => {

}
// dispose the context for the root executor
if (this.parent == null) {
context.dispose();
}
return outputs.map(name => getTensor(name, tensorsMap, context));

@@ -239,17 +244,33 @@ });

* array.
* @param disableWarning disable the no dynamic ops warning message, default
* to false
*/
async executeAsync(inputs, outputs, disableWarning = false) {
inputs = this.mapInputs(inputs);
this.checkInputs(inputs);
this.checkInputShapeAndType(inputs);
outputs = this.mapOutputs(outputs);
this.checkOutputs(outputs);
const tensorArrayMap = {};
const context = new ExecutionContext(this.weightMap, tensorArrayMap, this.functionExecutorMap);
async executeAsync(inputs, outputs) {
return this._executeAsync(inputs, outputs);
}
/**
* Executes the inference for given input tensors in Async fashion.
* @param inputs Tensor map for the model inputs, keyed by the input node
* names.
* @param outputs output node name from the Tensorflow model, if no outputs
* are specified, the default outputs of the model would be used. You can
* inspect intermediate nodes of the model by adding them to the outputs
* array.
* @param isFunctionExecution Flag for executing a function.
* @param tensorArrayMap Optional, global TensorArray map by id. Used for
* function execution.
* @param tensorArrayMap Optinal global TensorList map by id. Used for
* function execution.
*/
async _executeAsync(inputs, outputs, isFunctionExecution = false, tensorArrayMap = {}, tensorListMap = {}) {
if (!isFunctionExecution) {
inputs = this.mapInputs(inputs);
this.checkInputs(inputs);
this.checkInputShapeAndType(inputs);
outputs = this.mapOutputs(outputs);
this.checkOutputs(outputs);
}
const context = new ExecutionContext(this.weightMap, tensorArrayMap, tensorListMap, this.functionExecutorMap);
// Graph with control flow op requires runtime evaluation of the execution
// order, while without control flow the execution order is pre-determined
// in the compile method.
const tensorMap = await this.executeWithControlFlow(inputs, context, outputs, disableWarning);
const tensorMap = await this.executeWithControlFlow(inputs, context, outputs, isFunctionExecution);
const results = outputs.map(name => getTensor(name, tensorMap, context));

@@ -269,5 +290,9 @@ // dispose all the intermediate tensors

});
// dispose the context for the root executor
if (this.parent == null) {
context.dispose();
}
return results;
}
async executeFunctionAsync(inputs) {
async executeFunctionAsync(inputs, tensorArrayMap, tensorListMap) {
const mappedInputs = inputs.reduce((map, tensor, index) => {

@@ -277,3 +302,3 @@ map[this.inputs[index].name] = tensor;

}, {});
return this.executeAsync(mappedInputs, this.outputNodes, true);
return this._executeAsync(mappedInputs, this.outputNodes, true, tensorArrayMap, tensorListMap);
}

@@ -285,5 +310,5 @@ /**

* @param context the execution context object for current execution.
* @param disableWarning disable no async op warning
* @param isFunctionExecution Flag for executing a function.
*/
async executeWithControlFlow(inputs, context, outputNames, disableWarning) {
async executeWithControlFlow(inputs, context, outputNames, isFunctionExecution) {
const names = Object.keys(inputs);

@@ -310,3 +335,3 @@ const inputNodes = names.map(name => this.graph.nodes[parseNodeName(name)[0]]);

}
if (dynamicNode == null && !disableWarning) {
if (dynamicNode == null && !isFunctionExecution) {
console.warn(`This model execution did not contain any nodes with control flow ` +

@@ -313,0 +338,0 @@ `or dynamic output shapes. You can use model.execute() instead.`);

2

dist/executor/model_analysis.js

@@ -108,3 +108,3 @@ /**

'Switch', 'Merge', 'Enter', 'Exit', 'NextIteration', 'StatelessIf',
'StatelessWhile'
'StatelessWhile', 'if', 'While'
];

@@ -111,0 +111,0 @@ const DYNAMIC_SHAPE_OPS = [

6

dist/executor/tensor_array.d.ts

@@ -36,10 +36,10 @@ /**

readonly clearAfterRead: boolean;
private static nextId;
private tensors;
private closed_;
readonly idTensor: Tensor;
constructor(name: string, dtype: DataType, maxSize: number, elementShape: number[], identicalElementShapes: boolean, dynamicSize: boolean, clearAfterRead: boolean);
readonly id: number;
constructor(name: string, dtype: DataType, maxSize: number, elementShape: number[], identicalElementShapes: boolean, dynamicSize: boolean, clearAfterRead: boolean);
readonly closed: boolean;
/**
* Close the current TensorArray.
* Dispose the tensors and idTensor and mark the TensoryArray as closed.
*/

@@ -46,0 +46,0 @@ clearAndClose(): void;

17

dist/executor/tensor_array.js

@@ -17,3 +17,3 @@ /**

*/
import { concat, slice, stack, tensor, tidy, unstack } from '@tensorflow/tfjs-core';
import { concat, keep, scalar, slice, stack, tensor, tidy, unstack } from '@tensorflow/tfjs-core';
import { assertShapesMatchAllowUndefinedSize } from './tensor_utils';

@@ -35,4 +35,8 @@ /**

this.closed_ = false;
this.id = TensorArray.nextId++;
this.idTensor = scalar(0);
keep(this.idTensor);
}
get id() {
return this.idTensor.id;
}
get closed() {

@@ -42,3 +46,3 @@ return this.closed_;

/**
* Close the current TensorArray.
* Dispose the tensors and idTensor and mark the TensoryArray as closed.
*/

@@ -49,2 +53,3 @@ clearAndClose() {

this.closed_ = true;
this.idTensor.dispose();
}

@@ -105,9 +110,10 @@ size() {

assertShapesMatchAllowUndefinedSize(this.elementShape, tensor.shape, `TensorArray ${this.name}: Could not write to TensorArray index ${index}.`);
if (t && t.read) {
if (t.read) {
throw new Error(`TensorArray ${this.name}: Could not write to TensorArray index ${index}, because it has already been read.`);
}
if (t && t.written) {
if (t.written) {
throw new Error(`TensorArray ${this.name}: Could not write to TensorArray index ${index}, because it has already been written.`);
}
t.tensor = tensor;
keep(tensor);
t.written = true;

@@ -237,3 +243,2 @@ this.tensors[index] = t;

}
TensorArray.nextId = 0;
//# sourceMappingURL=tensor_array.js.map

16

dist/executor/tensor_list.d.ts

@@ -33,6 +33,8 @@ /**

export declare class TensorList {
tensors: Tensor[];
elementShape: number[];
elementDtype: DataType;
readonly tensors: Tensor[];
readonly elementShape: number[];
readonly elementDtype: DataType;
readonly idTensor: Tensor;
maxNumElements: number;
readonly id: number;
/**

@@ -52,2 +54,6 @@ *

/**
* Dispose the tensors and idTensor and clear the tensor list.
*/
clearAndClose(): void;
/**
* The size of the tensors in the tensor list.

@@ -113,3 +119,3 @@ */

*/
export declare function fromTensor(tensor: Tensor, elementShape: number[]): TensorList;
export declare function fromTensor(tensor: Tensor, elementShape: number[], elementDtype: DataType): TensorList;
/**

@@ -129,3 +135,3 @@ * Return a TensorList of the given size with empty elements.

*/
export declare function scatter(tensor: Tensor, indices: number[], elementShape: number[], numElements: number): TensorList;
export declare function scatter(tensor: Tensor, indices: number[], elementShape: number[], numElements?: number): TensorList;
/**

@@ -132,0 +138,0 @@ * Split the values of a Tensor into a TensorList.

40

dist/executor/tensor_list.js

@@ -17,3 +17,3 @@ /**

*/
import { concat, slice, stack, tensor, tidy, unstack } from '@tensorflow/tfjs-core';
import { concat, keep, scalar, slice, stack, tensor, tidy, unstack } from '@tensorflow/tfjs-core';
import { assertShapesMatchAllowUndefinedSize } from './tensor_utils';

@@ -47,4 +47,18 @@ /**

this.elementDtype = elementDtype;
if (tensors != null) {
tensors.forEach(tensor => {
if (elementDtype !== tensor.dtype) {
throw new Error(`Invalid data types; op elements ${elementDtype}, but list elements ${tensor.dtype}`);
}
assertShapesMatchAllowUndefinedSize(elementShape, tensor.shape, 'TensorList shape mismatch: ');
keep(tensor);
});
}
this.idTensor = scalar(0);
this.maxNumElements = maxNumElements;
keep(this.idTensor);
}
get id() {
return this.idTensor.id;
}
/**

@@ -57,2 +71,10 @@ * Get a new TensorList containing a copy of the underlying tensor container.

/**
* Dispose the tensors and idTensor and clear the tensor list.
*/
clearAndClose() {
this.tensors.forEach(tensor => tensor.dispose());
this.tensors.length = 0;
this.idTensor.dispose();
}
/**
* The size of the tensors in the tensor list.

@@ -111,2 +133,3 @@ */

}
keep(tensor);
this.tensors.push(tensor);

@@ -160,2 +183,3 @@ }

assertShapesMatchAllowUndefinedSize(this.elementShape, tensor.shape, 'TensorList shape mismatch: ');
keep(tensor);
this.tensors[elementIndex] = tensor;

@@ -210,3 +234,3 @@ }

*/
export function fromTensor(tensor, elementShape) {
export function fromTensor(tensor, elementShape, elementDtype) {
const dtype = tensor.dtype;

@@ -216,9 +240,8 @@ if (tensor.shape.length < 1) {

}
if (tensor.dtype !== elementDtype) {
throw new Error(`Invalid data types; op elements ${tensor.dtype}, but list elements ${elementDtype}`);
}
const outputShape = tensor.shape.slice(1);
assertShapesMatchAllowUndefinedSize(outputShape, elementShape, 'TensorList shape mismatch: ');
const tensorList = [];
for (let i = 0; i < tensor.shape[0]; ++i) {
const tmp = tensor.slice(i, i + 1).reshape(outputShape);
tensorList.push(tmp);
}
const tensorList = tensor.unstack();
return new TensorList(tensorList, elementShape, dtype);

@@ -247,3 +270,3 @@ }

const maxIndex = Math.max(...indices);
if (numElements !== -1 && maxIndex >= numElements) {
if (numElements != null && numElements !== -1 && maxIndex >= numElements) {
throw new Error(`Max index must be < array size (${maxIndex} vs. ${numElements})`);

@@ -286,2 +309,3 @@ }

}
tensor.dispose();
return tensors;

@@ -288,0 +312,0 @@ });

7

dist/executor/types.d.ts

@@ -18,6 +18,9 @@ /**

import { Tensor } from '@tensorflow/tfjs-core';
import { NamedTensorsMap } from '../data/types';
import { NamedTensorsMap, TensorArrayMap, TensorListMap } from '../data/types';
/**
*
*/
export interface FunctionExecutor {
executeFunctionAsync(inputs: Tensor[]): Promise<Tensor[]>;
executeFunctionAsync(inputs: Tensor[], tensorArrayMap: TensorArrayMap, tensorListMap: TensorListMap): Promise<Tensor[]>;
weightMap: NamedTensorsMap;
}

199

dist/operations/executors/control_executor.js

@@ -19,2 +19,3 @@ /**

import { TensorArray } from '../../executor/tensor_array';
import { fromTensor, reserve, scatter, split } from '../../executor/tensor_list';
import { getParamValue, getTensor } from './utils';

@@ -31,6 +32,6 @@ export const executeOp = async (node, tensorMap, context) => {

if (condValue[0]) {
return context.functionMap[thenFunc].executeFunctionAsync(args);
return context.functionMap[thenFunc].executeFunctionAsync(args, context.tensorArrayMap, context.tensorListMap);
}
else {
return context.functionMap[elseFunc].executeFunctionAsync(args);
return context.functionMap[elseFunc].executeFunctionAsync(args, context.tensorArrayMap, context.tensorListMap);
}

@@ -43,17 +44,45 @@ }

const args = getParamValue('args', node, tensorMap, context);
const condTensor = (await context.functionMap[condFunc].executeFunctionAsync(args))[0];
let condValue = await condTensor.data();
// Calculate the condition of the loop
const condResult = (await context.functionMap[condFunc].executeFunctionAsync(args, context.tensorArrayMap, context.tensorListMap));
const argIds = args.map(tensor => tensor.id);
let condValue = await condResult[0].data();
// Dispose the intermediate tensors for condition function
condResult.forEach(tensor => {
if (!tensor.kept && argIds.indexOf(tensor.id) === -1) {
tensor.dispose();
}
});
let result = args;
while (condValue[0]) {
result =
await context.functionMap[bodyFunc].executeFunctionAsync(result);
const condTensor = (await context.functionMap[condFunc].executeFunctionAsync(result))[0];
condValue = await condTensor.data();
// Record the previous result for intermediate tensor tracking
const origResult = result;
// Execution the body of the loop
result = await context.functionMap[bodyFunc].executeFunctionAsync(result, context.tensorArrayMap, context.tensorListMap);
const resultIds = result.map(tensor => tensor.id);
// Dispose the intermediate tensor for body function that is not global
// kept, not input/output of the body function
origResult.forEach(tensor => {
if (!tensor.kept && argIds.indexOf(tensor.id) === -1 &&
resultIds.indexOf(tensor.id) === -1) {
tensor.dispose();
}
});
// Recalcuate the condition of the loop using the latest results.
const condResult = (await context.functionMap[condFunc].executeFunctionAsync(result, context.tensorArrayMap, context.tensorListMap));
condValue = await condResult[0].data();
// Dispose the intermediate tensors for condition function
condResult.forEach(tensor => {
if (!tensor.kept && argIds.indexOf(tensor.id) === -1 &&
resultIds.indexOf(tensor.id) === -1) {
tensor.dispose();
}
});
}
return result;
}
case 'LoopCond':
case 'LoopCond': {
return [
getParamValue('pred', node, tensorMap, context).clone()
];
}
case 'Switch': {

@@ -66,7 +95,8 @@ const pred = getParamValue('pred', node, tensorMap, context);

}
case 'Merge':
case 'Merge': {
const inputName = node.inputNames.find(name => getTensor(name, tensorMap, context) !== undefined);
return inputName ? [getTensor(inputName, tensorMap, context).clone()] :
undefined;
case 'Enter':
}
case 'Enter': {
const frameId = getParamValue('frameName', node, tensorMap, context);

@@ -76,11 +106,14 @@ const data = getParamValue('tensor', node, tensorMap, context);

return [data.clone()];
case 'Exit':
}
case 'Exit': {
const tensor = getParamValue('tensor', node, tensorMap, context);
context.exitFrame();
return [tensor.clone()];
case 'NextIteration':
}
case 'NextIteration': {
const input = getParamValue('tensor', node, tensorMap, context);
context.nextIteration();
return [input.clone()];
case 'TensorArrayV3':
}
case 'TensorArrayV3': {
const size = getParamValue('size', node, tensorMap, context);

@@ -95,49 +128,145 @@ const dtype = getParamValue('dtype', node, tensorMap, context);

context.addTensorArray(tensorArray);
return [scalar(tensorArray.id), scalar(1.0)];
case 'TensorArrayWriteV3':
return [tensorArray.idTensor, scalar(1.0)];
}
case 'TensorArrayWriteV3': {
const id = getParamValue('tensorArrayId', node, tensorMap, context);
const index = getParamValue('index', node, tensorMap, context);
const writeTensor = getParamValue('tensor', node, tensorMap, context);
const writeTensorArray = context.getTensorArray(id);
const writeTensorArray = context.getTensorArray(id.id);
writeTensorArray.write(index, writeTensor);
return [scalar(1.0)];
case 'TensorArrayReadV3':
return [writeTensorArray.idTensor];
}
case 'TensorArrayReadV3': {
const readId = getParamValue('tensorArrayId', node, tensorMap, context);
const readIndex = getParamValue('index', node, tensorMap, context);
const readTensorArray = context.getTensorArray(readId);
const readTensorArray = context.getTensorArray(readId.id);
return [readTensorArray.read(readIndex)];
case 'TensorArrayGatherV3':
}
case 'TensorArrayGatherV3': {
const gatherId = getParamValue('tensorArrayId', node, tensorMap, context);
const gatherIndices = getParamValue('indices', node, tensorMap, context);
const gatherDtype = getParamValue('dtype', node, tensorMap, context);
const gatherTensorArray = context.getTensorArray(gatherId);
const gatherTensorArray = context.getTensorArray(gatherId.id);
return [gatherTensorArray.gather(gatherIndices, gatherDtype)];
case 'TensorArrayScatterV3':
}
case 'TensorArrayScatterV3': {
const scatterId = getParamValue('tensorArrayId', node, tensorMap, context);
const scatterIndices = getParamValue('indices', node, tensorMap, context);
const scatterTensor = getParamValue('tensor', node, tensorMap, context);
const scatterTensorArray = context.getTensorArray(scatterId);
const scatterTensorArray = context.getTensorArray(scatterId.id);
scatterTensorArray.scatter(scatterIndices, scatterTensor);
return [scalar(1.0)];
case 'TensorArrayConcatV3':
return [scatterTensorArray.idTensor];
}
case 'TensorArrayConcatV3': {
const concatId = getParamValue('tensorArrayId', node, tensorMap, context);
const concatTensorArray = context.getTensorArray(concatId);
const concatTensorArray = context.getTensorArray(concatId.id);
const concatDtype = getParamValue('dtype', node, tensorMap, context);
return [concatTensorArray.concat(concatDtype)];
case 'TensorArraySplitV3':
}
case 'TensorArraySplitV3': {
const splitId = getParamValue('tensorArrayId', node, tensorMap, context);
const splitTensor = getParamValue('tensor', node, tensorMap, context);
const lengths = getParamValue('lengths', node, tensorMap, context);
const splitTensorArray = context.getTensorArray(splitId);
const splitTensorArray = context.getTensorArray(splitId.id);
splitTensorArray.split(lengths, splitTensor);
return [scalar(1.0)];
case 'TensorArraySizeV3':
return [splitTensorArray.idTensor];
}
case 'TensorArraySizeV3': {
const sizeId = getParamValue('tensorArrayId', node, tensorMap, context);
const sizeTensorArray = context.getTensorArray(sizeId);
const sizeTensorArray = context.getTensorArray(sizeId.id);
return [scalar(sizeTensorArray.size(), 'int32')];
case 'TensorArrayCloseV3':
}
case 'TensorArrayCloseV3': {
const closeId = getParamValue('tensorArrayId', node, tensorMap, context);
const closeTensorArray = context.getTensorArray(closeId);
const closeTensorArray = context.getTensorArray(closeId.id);
closeTensorArray.clearAndClose();
return [scalar(0)];
return [closeTensorArray.idTensor];
}
case 'TensorListSetItem': {
const idTensor = getParamValue('tensorListId', node, tensorMap, context);
const index = getParamValue('index', node, tensorMap, context);
const writeTensor = getParamValue('tensor', node, tensorMap, context);
const tensorList = context.getTensorList(idTensor.id);
tensorList.setItem(index, writeTensor);
return [tensorList.idTensor];
}
case 'TensorListGetItem': {
const idTensor = getParamValue('tensorListId', node, tensorMap, context);
const readIndex = getParamValue('index', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const elementDType = getParamValue('elementDType', node, tensorMap, context);
const tensorList = context.getTensorList(idTensor.id);
return [tensorList.getItem(readIndex, elementShape, elementDType)];
}
case 'TensorListScatterV2':
case 'TensorListScatter': {
const scatterIndices = getParamValue('indices', node, tensorMap, context);
const scatterTensor = getParamValue('tensor', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const numElements = getParamValue('numElements', node, tensorMap, context);
const tensorList = scatter(scatterTensor, scatterIndices, elementShape, numElements);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
case 'TensorListReserve': {
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const elementDtype = getParamValue('elementDType', node, tensorMap, context);
const numElements = getParamValue('numElements', node, tensorMap, context);
const tensorList = reserve(elementShape, elementDtype, numElements);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
case 'TensorListGather': {
const gatherId = getParamValue('tensorListId', node, tensorMap, context);
const gatherIndices = getParamValue('indices', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const elementDtype = getParamValue('elementDType', node, tensorMap, context);
const tensorList = context.getTensorList(gatherId.id);
return [tensorList.gather(gatherIndices, elementDtype, elementShape)];
}
case 'TensorListStack': {
const idTensor = getParamValue('tensorListId', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const elementDtype = getParamValue('elementDType', node, tensorMap, context);
const numElements = getParamValue('numElements', node, tensorMap, context);
const tensorList = context.getTensorList(idTensor.id);
return [tensorList.stack(elementShape, elementDtype, numElements)];
}
case 'TensorListFromTensor': {
const tensor = getParamValue('tensor', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const elementDtype = getParamValue('elementDType', node, tensorMap, context);
const tensorList = fromTensor(tensor, elementShape, elementDtype);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
case 'TensorListConcat': {
const concatId = getParamValue('tensorListId', node, tensorMap, context);
const tensorList = context.getTensorList(concatId.id);
const concatDtype = getParamValue('dtype', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
return [tensorList.concat(concatDtype, elementShape)];
}
case 'TensorListPushBack': {
const idTensor = getParamValue('tensorListId', node, tensorMap, context);
const writeTensor = getParamValue('tensor', node, tensorMap, context);
const tensorList = context.getTensorList(idTensor.id);
tensorList.pushBack(writeTensor);
return [tensorList.idTensor];
}
case 'TensorListPopBack': {
const idTensor = getParamValue('tensorListId', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const elementDType = getParamValue('elementDType', node, tensorMap, context);
const tensorList = context.getTensorList(idTensor.id);
return [tensorList.popBack(elementShape, elementDType)];
}
case 'TensorListSplit': {
const splitTensor = getParamValue('tensor', node, tensorMap, context);
const elementShape = getParamValue('elementShape', node, tensorMap, context);
const lengths = getParamValue('lengths', node, tensorMap, context);
const tensorList = split(splitTensor, lengths, elementShape);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
default:

@@ -144,0 +273,0 @@ throw TypeError(`Node type ${node.op} is not implemented`);

22

dist/operations/executors/convolution_executor.js

@@ -18,3 +18,3 @@ /**

import * as tfc from '@tensorflow/tfjs-core';
import { getParamValue } from './utils';
import { getPadding, getParamValue } from './utils';
export const executeOp = (node, tensorMap, context) => {

@@ -32,3 +32,3 @@ switch (node.op) {

const stride = getParamValue('strides', node, tensorMap, context);
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
const dataFormat = getParamValue('dataFormat', node, tensorMap, context)

@@ -60,3 +60,3 @@ .toUpperCase();

const stride = getParamValue('strides', node, tensorMap, context);
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
const dataFormat = getParamValue('dataFormat', node, tensorMap, context)

@@ -85,3 +85,3 @@ .toUpperCase();

const stride = getParamValue('strides', node, tensorMap, context);
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
return [tfc.conv2dTranspose(getParamValue('x', node, tensorMap, context), getParamValue('filter', node, tensorMap, context), shape, [stride[1], stride[2]], pad)];

@@ -92,3 +92,3 @@ }

const stride = getParamValue('strides', node, tensorMap, context);
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
const dilations = getParamValue('dilations', node, tensorMap, context);

@@ -139,2 +139,14 @@ const dataFormat = getParamValue('dataFormat', node, tensorMap, context)

}
case 'Dilation2D': {
const strides = getParamValue('strides', node, tensorMap, context);
const pad = getParamValue('pad', node, tensorMap, context);
const dilations = getParamValue('dilations', node, tensorMap, context);
// strides: [1, stride_height, stride_width, 1].
const strideHeight = strides[1];
const strideWidth = strides[2];
// dilations: [1, dilation_height, dilation_width, 1].
const dilationHeight = dilations[1];
const dilationWidth = dilations[2];
return [tfc.dilation2d(getParamValue('x', node, tensorMap, context), getParamValue('filter', node, tensorMap, context), [strideHeight, strideWidth], pad, [dilationHeight, dilationWidth], 'NHWC' /* dataFormat */)];
}
default:

@@ -141,0 +153,0 @@ throw TypeError(`Node type ${node.op} is not implemented`);

2

dist/operations/executors/dynamic_executor.d.ts
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

8

dist/operations/executors/dynamic_executor.js
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -22,2 +22,3 @@ * you may not use this file except in compliance with the License.

case 'NonMaxSuppressionV5':
case 'NonMaxSuppressionV4':
case 'NonMaxSuppressionV3':

@@ -35,2 +36,7 @@ case 'NonMaxSuppressionV2': {

}
if (node.op === 'NonMaxSuppressionV4') {
const padToMaxOutputSize = getParamValue('padToMaxOutputSize', node, tensorMap, context);
const result = await tfc.image.nonMaxSuppressionPaddedAsync(boxes, scores, maxOutputSize, iouThreshold, scoreThreshold, padToMaxOutputSize);
return [result.selectedIndices, result.validOutputs];
}
return [await tfc.image.nonMaxSuppressionAsync(boxes, scores, maxOutputSize, iouThreshold, scoreThreshold)];

@@ -37,0 +43,0 @@ }

2

dist/operations/executors/evaluation_executor.d.ts
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

2

dist/operations/executors/evaluation_executor.js
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

2

dist/operations/executors/image_executor.d.ts
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

2

dist/operations/executors/image_executor.js
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

10

dist/operations/executors/slice_join_executor.js

@@ -59,9 +59,2 @@ /**

const tensor = getParamValue('x', node, tensorMap, context);
if (begin.length === 1 && tensor.shape.length > 1) {
for (let i = 1; i < tensor.shape.length; i++) {
begin.push(0);
end.push(tensor.shape[i]);
strides.push(strides[0]);
}
}
return [tfc.stridedSlice(tensor, begin, end, strides, beginMask, endMask, ellipsisMask, newAxisMask, shrinkAxisMask)];

@@ -102,3 +95,4 @@ }

const numOrSizeSplits = getParamValue('numOrSizeSplits', node, tensorMap, context);
return tfc.split(getParamValue('x', node, tensorMap, context), numOrSizeSplits, axis);
const tensor = getParamValue('x', node, tensorMap, context);
return tfc.split(tensor, numOrSizeSplits, axis);
}

@@ -105,0 +99,0 @@ case 'ScatterNd': {

1

dist/operations/executors/test_helper.d.ts

@@ -25,2 +25,3 @@ /**

export declare function createTensorShapeAttr(value: number[]): ParamValue;
export declare function createShapeAttrFromIndex(inputIndex: number): InputParamValue;
export declare function createNumericArrayAttr(value: number[]): ParamValue;

@@ -27,0 +28,0 @@ export declare function createNumericArrayAttrFromIndex(inputIndex: number): InputParamValue;

10

dist/operations/executors/test_helper.js

@@ -19,2 +19,5 @@ export function createNumberAttr(value) {

}
export function createShapeAttrFromIndex(inputIndex) {
return { inputIndexStart: inputIndex, type: 'shape' };
}
export function createNumericArrayAttr(value) {

@@ -39,3 +42,3 @@ return { value, type: 'number[]' };

opMappers.find(mapper => mapper.tfOpName === node.op);
return Object.keys(node.inputParams).every(key => {
const matched = Object.keys(node.inputParams).every(key => {
const value = node.inputParams[key];

@@ -51,3 +54,8 @@ const def = opMapper.inputs.find(param => param.name === key);

});
if (!matched) {
console.log('node = ', node);
console.log('opMapper = ', opMapper);
}
return matched;
}
//# sourceMappingURL=test_helper.js.map

8

dist/operations/executors/transformation_executor.js

@@ -18,3 +18,3 @@ /**

import * as tfc from '@tensorflow/tfjs-core';
import { getParamValue, split } from './utils';
import { getParamValue } from './utils';
export const executeOp = (node, tensorMap, context) => {

@@ -38,7 +38,7 @@ switch (node.op) {

case 'Pad': {
return [tfc.pad(getParamValue('x', node, tensorMap, context), split(getParamValue('padding', node, tensorMap, context), 2), getParamValue('constantValue', node, tensorMap, context))];
return [tfc.pad(getParamValue('x', node, tensorMap, context), getParamValue('padding', node, tensorMap, context), getParamValue('constantValue', node, tensorMap, context))];
}
case 'SpaceToBatchND': {
const blockShape = getParamValue('blockShape', node, tensorMap, context);
const paddings = split(getParamValue('paddings', node, tensorMap, context), 2);
const paddings = getParamValue('paddings', node, tensorMap, context);
return [tfc.spaceToBatchND(getParamValue('x', node, tensorMap, context), blockShape, paddings)];

@@ -48,3 +48,3 @@ }

const blockShape = getParamValue('blockShape', node, tensorMap, context);
const crops = split(getParamValue('crops', node, tensorMap, context), 2);
const crops = getParamValue('crops', node, tensorMap, context);
return [tfc.batchToSpaceND(getParamValue('x', node, tensorMap, context), blockShape, crops)];

@@ -51,0 +51,0 @@ }

1

dist/operations/executors/utils.d.ts

@@ -44,1 +44,2 @@ /**

export declare function split(arr: number[], size: number): number[][];
export declare function getPadding(node: Node, tensorMap: NamedTensorsMap, context: ExecutionContext): ValueType;

23

dist/operations/executors/utils.js

@@ -17,2 +17,3 @@ /**

*/
import * as tfc from '@tensorflow/tfjs-core';
export function getParamValue(paramName, node, tensorMap, context) {

@@ -33,5 +34,7 @@ const inputParam = node.inputParams[paramName];

}
const data = Array.prototype.slice.call(getTensor(node.inputNames.slice(start)[0], tensorMap, context)
.dataSync());
return inputParam.type === 'number' ? data[0] : data;
const tensor = getTensor(node.inputNames.slice(start)[0], tensorMap, context);
const data = tensor.dataSync();
return inputParam.type === 'number' ?
data[0] :
tfc.util.toNestedArray(tensor.shape, data);
}

@@ -95,2 +98,16 @@ const attrParam = node.attrParams[paramName];

}
export function getPadding(node, tensorMap, context) {
let pad = getParamValue('pad', node, tensorMap, context);
if (pad === 'explicit') {
// This is 1d array, we need to convert it to 2d array
pad = getParamValue('explicitPaddings', node, tensorMap, context);
const explicitPadding = [[0, 0], [0, 0], [0, 0], [0, 0]];
for (let i = 0; i < 4; i++) {
explicitPadding[i][0] = pad[i * 2];
explicitPadding[i][1] = pad[i * 2 + 1];
}
return explicitPadding;
}
return pad;
}
//# sourceMappingURL=utils.js.map

138

dist/operations/op_list/control.js

@@ -91,3 +91,3 @@ /**

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'index', 'type': 'number' },

@@ -105,3 +105,3 @@ { 'start': 2, 'name': 'tensor', 'type': 'tensor' },

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'index', 'type': 'number' },

@@ -121,3 +121,3 @@ { 'start': 2, 'name': 'flowIn', 'type': 'number' },

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'indices', 'type': 'number[]' },

@@ -135,3 +135,3 @@ { 'start': 2, 'name': 'flowIn', 'type': 'number' },

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'indices', 'type': 'number[]' },

@@ -147,3 +147,3 @@ { 'start': 2, 'name': 'tensor', 'type': 'tensor' },

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'flowIn', 'type': 'number' },

@@ -164,3 +164,3 @@ ],

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'tensor', 'type': 'tensor' },

@@ -176,3 +176,3 @@ { 'start': 2, 'name': 'lengths', 'type': 'number[]' },

'inputs': [
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' },
{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' },
{ 'start': 1, 'name': 'flowIn', 'type': 'number' }

@@ -184,3 +184,3 @@ ]

'category': 'control',
'inputs': [{ 'start': 0, 'name': 'tensorArrayId', 'type': 'number' }]
'inputs': [{ 'start': 0, 'name': 'tensorArrayId', 'type': 'tensor' }]
},

@@ -232,4 +232,124 @@ {

]
}
},
{
'tfOpName': 'TensorListScatter',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensor', 'type': 'tensor' },
{ 'start': 1, 'name': 'indices', 'type': 'number[]' },
{ 'start': 2, 'name': 'elementShape', 'type': 'shape' }
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListScatterV2',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensor', 'type': 'tensor' },
{ 'start': 1, 'name': 'indices', 'type': 'number[]' },
{ 'start': 2, 'name': 'elementShape', 'type': 'shape' },
{ 'start': 3, 'name': 'numElements', 'type': 'number' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListGather',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
{ 'start': 1, 'name': 'indices', 'type': 'number[]' },
{ 'start': 2, 'name': 'elementShape', 'type': 'shape' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListGetItem',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
{ 'start': 1, 'name': 'index', 'type': 'number' },
{ 'start': 2, 'name': 'elementShape', 'type': 'shape' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListSetItem',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
{ 'start': 1, 'name': 'index', 'type': 'number' },
{ 'start': 2, 'name': 'tensor', 'type': 'tensor' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListReserve',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'elementShape', 'type': 'shape' },
{ 'start': 1, 'name': 'numElements', 'type': 'number' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListFromTensor',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensor', 'type': 'tensor' },
{ 'start': 1, 'name': 'elementShape', 'type': 'shape' }
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListStack',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
{ 'start': 1, 'name': 'elementShape', 'type': 'shape' },
],
'attrs': [
{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' },
{ 'tfName': 'num_elements', 'name': 'numElements', 'type': 'dtype' }
]
},
{
'tfOpName': 'TensorListSplit',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensor', 'type': 'tensor' },
{ 'start': 1, 'name': 'elementShape', 'type': 'shape' },
{ 'start': 2, 'name': 'lengths', 'type': 'number[]' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListConcat',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
],
'attrs': [
{ 'tfName': 'element_shape', 'name': 'elementShape', 'type': 'shape' },
{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }
]
},
{
'tfOpName': 'TensorListPopBack',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
{ 'start': 1, 'name': 'elementShape', 'type': 'shape' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
{
'tfOpName': 'TensorListPushBack',
'category': 'control',
'inputs': [
{ 'start': 0, 'name': 'tensorListId', 'type': 'tensor' },
{ 'start': 1, 'name': 'tensor', 'type': 'tensor' },
],
'attrs': [{ 'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype' }]
},
];
//# sourceMappingURL=control.js.map

4

dist/operations/op_list/convolution.d.ts

@@ -1,5 +0,4 @@

import { OpMapper } from '../types';
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Copyright 2020 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -18,2 +17,3 @@ * you may not use this file except in compliance with the License.

*/
import { OpMapper } from '../types';
export declare const json: OpMapper[];

15

dist/operations/op_list/convolution.js
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Copyright 2020 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -329,4 +329,17 @@ * you may not use this file except in compliance with the License.

],
},
{
'tfOpName': 'Dilation2D',
'category': 'convolution',
'inputs': [
{ 'start': 0, 'name': 'x', 'type': 'tensor' },
{ 'start': 1, 'name': 'filter', 'type': 'tensor' },
],
'attrs': [
{ 'tfName': 'strides', 'name': 'strides', 'type': 'number[]' },
{ 'tfName': 'rates', 'name': 'dilations', 'type': 'number[]' },
{ 'tfName': 'padding', 'name': 'pad', 'type': 'string' }
]
}
];
//# sourceMappingURL=convolution.js.map

24

dist/operations/op_list/dynamic.js

@@ -40,2 +40,26 @@ /**

{
'tfOpName': 'NonMaxSuppressionV4',
'category': 'dynamic',
'inputs': [
{ 'start': 0, 'name': 'boxes', 'type': 'tensor' },
{ 'start': 1, 'name': 'scores', 'type': 'tensor' },
{ 'start': 2, 'name': 'maxOutputSize', 'type': 'number' },
{ 'start': 3, 'name': 'iouThreshold', 'type': 'number' },
{ 'start': 4, 'name': 'scoreThreshold', 'type': 'number' }
],
'attrs': [
{ 'tfName': 'T', 'name': 'dtype', 'type': 'dtype', 'notSupported': true }, {
'tfName': 'T_threshold',
'name': 'threshold',
'type': 'dtype',
'notSupported': true
},
{
'tfName': 'pad_to_max_output_size',
'name': 'padToMaxOutputSize',
'type': 'bool'
}
]
},
{
'tfOpName': 'NonMaxSuppressionV5',

@@ -42,0 +66,0 @@ 'category': 'dynamic',

2

dist/version.d.ts
/** @license See the LICENSE file. */
declare const version = "2.0.1";
declare const version = "2.1.0";
export { version };

2

dist/version.js
/** @license See the LICENSE file. */
// This code is auto-generated, do not modify this file!
const version = '2.0.1';
const version = '2.1.0';
export { version };
//# sourceMappingURL=version.js.map

9

package.json
{
"name": "@tensorflow/tfjs-converter",
"version": "2.0.1",
"version": "2.1.0",
"description": "Tensorflow model converter for javascript",

@@ -18,3 +18,3 @@ "main": "dist/tf-converter.node.js",

"peerDependencies": {
"@tensorflow/tfjs-core": "2.0.1"
"@tensorflow/tfjs-core": "2.1.0"
},

@@ -25,4 +25,4 @@ "devDependencies": {

"@rollup/plugin-typescript": "^3.0.0",
"@tensorflow/tfjs-backend-cpu": "2.0.1",
"@tensorflow/tfjs-core": "2.0.1",
"@tensorflow/tfjs-backend-cpu": "2.1.0",
"@tensorflow/tfjs-core": "2.1.0",
"@types/deep-equal": "^1.0.1",

@@ -81,2 +81,3 @@ "@types/jasmine": "~2.8.6",

"gen-json": "ts-node -s ./scripts/gen_json.ts",
"model-summary": "ts-node -s ./tools/model_summary.ts",
"pb2json": "ts-node -s ./tools/pb2json_converter.ts",

@@ -83,0 +84,0 @@ "build-pip-package": "yarn gen-json --test && cd python && ./build-pip-package.sh --test /tmp/tfjs-pips",

2

README.md

@@ -10,3 +10,3 @@ # Getting started

__Note__: _Session bundle and Frozen model formats have been deprecated in TensorFlow.js 1.0. Please use the TensorFlow.js 0.15.x backend to convert these formats, available in
__Note__: _Session bundle format have been deprecated in TensorFlow.js 1.0. Please use the TensorFlow.js 0.15.x backend to convert session bundle, available in
`tfjs-converter` [0.8.6](https://pypi.org/project/tensorflowjs/0.8.6/)._

@@ -13,0 +13,0 @@

2

run_tests.ts
/**
* @license
* Copyright 2020 Google Inc. All Rights Reserved.
* Copyright 2020 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

5

src/data/types.ts

@@ -19,2 +19,3 @@ /**

import {TensorArray} from '../executor/tensor_array';
import {TensorList} from '../executor/tensor_list';

@@ -33,2 +34,6 @@ export type NamedTensorMap = {

export type TensorListMap = {
[key: number]: TensorList
};
export interface TensorInfo {

@@ -35,0 +40,0 @@ name: string;

28

src/executor/execution_context.ts

@@ -19,5 +19,6 @@ /**

import {NamedTensorsMap, TensorArrayMap} from '../data/types';
import {NamedTensorsMap, TensorArrayMap, TensorListMap} from '../data/types';
import {TensorArray} from './tensor_array';
import {TensorList} from './tensor_list';
import {FunctionExecutor} from './types';

@@ -48,5 +49,6 @@

constructor(
public readonly weightMap: NamedTensorsMap,
public readonly tensorArrayMap: TensorArrayMap,
public readonly functionMap: {[key: string]: FunctionExecutor} = {}) {
readonly weightMap: NamedTensorsMap = {},
readonly tensorArrayMap: TensorArrayMap = {},
readonly tensorListMap: TensorListMap = {},
readonly functionMap: {[key: string]: FunctionExecutor} = {}) {
this.generateCurrentContextIds();

@@ -169,2 +171,20 @@ }

}
addTensorList(tensorList: TensorList) {
this.tensorListMap[tensorList.id] = tensorList;
}
getTensorList(id: number): TensorList {
return this.tensorListMap[id];
}
dispose() {
for (const key in this.tensorArrayMap) {
this.tensorArrayMap[key].clearAndClose();
}
for (const key in this.tensorListMap) {
this.tensorListMap[key].clearAndClose();
}
}
}

77

src/executor/graph_executor.ts

@@ -21,3 +21,3 @@ /**

import {ISignatureDef} from '../data/compiled_api';
import {NamedTensorsMap, TensorArrayMap, TensorInfo} from '../data/types';
import {NamedTensorsMap, TensorArrayMap, TensorInfo, TensorListMap} from '../data/types';
import {getNodeNameAndIndex, getParamValue, getTensor, getTensorsForCurrentContenxt, parseNodeName} from '../operations/executors/utils';

@@ -197,5 +197,7 @@ import {executeOp} from '../operations/operation_executor';

const tensorArrayMap: TensorArrayMap = {};
const tensorListMap: TensorListMap = {};
return tidy(() => {
const context = new ExecutionContext(
this.weightMap, tensorArrayMap, this.functionExecutorMap);
this.weightMap, tensorArrayMap, tensorListMap,
this.functionExecutorMap);
const tensorsMap: NamedTensorsMap = {...this.weightMap};

@@ -225,2 +227,6 @@ Object.keys(inputs).forEach(name => {

}
// dispose the context for the root executor
if (this.parent == null) {
context.dispose();
}
return outputs.map(name => getTensor(name, tensorsMap, context));

@@ -280,2 +286,3 @@ });

}
/**

@@ -289,16 +296,38 @@ * Executes the inference for given input tensors in Async fashion.

* array.
* @param disableWarning disable the no dynamic ops warning message, default
* to false
*/
async executeAsync(
inputs: NamedTensorMap, outputs: string[],
disableWarning = false): Promise<Tensor[]> {
inputs = this.mapInputs(inputs);
this.checkInputs(inputs);
this.checkInputShapeAndType(inputs);
outputs = this.mapOutputs(outputs);
this.checkOutputs(outputs);
const tensorArrayMap: TensorArrayMap = {};
async executeAsync(inputs: NamedTensorMap, outputs: string[]):
Promise<Tensor[]> {
return this._executeAsync(inputs, outputs);
}
/**
* Executes the inference for given input tensors in Async fashion.
* @param inputs Tensor map for the model inputs, keyed by the input node
* names.
* @param outputs output node name from the Tensorflow model, if no outputs
* are specified, the default outputs of the model would be used. You can
* inspect intermediate nodes of the model by adding them to the outputs
* array.
* @param isFunctionExecution Flag for executing a function.
* @param tensorArrayMap Optional, global TensorArray map by id. Used for
* function execution.
* @param tensorArrayMap Optinal global TensorList map by id. Used for
* function execution.
*/
private async _executeAsync(
inputs: NamedTensorMap, outputs: string[], isFunctionExecution = false,
tensorArrayMap: TensorArrayMap = {},
tensorListMap: TensorListMap = {}): Promise<Tensor[]> {
if (!isFunctionExecution) {
inputs = this.mapInputs(inputs);
this.checkInputs(inputs);
this.checkInputShapeAndType(inputs);
outputs = this.mapOutputs(outputs);
this.checkOutputs(outputs);
}
const context = new ExecutionContext(
this.weightMap, tensorArrayMap, this.functionExecutorMap);
this.weightMap, tensorArrayMap, tensorListMap,
this.functionExecutorMap);
// Graph with control flow op requires runtime evaluation of the execution

@@ -308,3 +337,3 @@ // order, while without control flow the execution order is pre-determined

const tensorMap = await this.executeWithControlFlow(
inputs, context, outputs, disableWarning);
inputs, context, outputs, isFunctionExecution);
const results = outputs.map(name => getTensor(name, tensorMap, context));

@@ -326,6 +355,13 @@

});
// dispose the context for the root executor
if (this.parent == null) {
context.dispose();
}
return results;
}
async executeFunctionAsync(inputs: Tensor[]): Promise<Tensor[]> {
async executeFunctionAsync(
inputs: Tensor[], tensorArrayMap: TensorArrayMap,
tensorListMap: TensorListMap): Promise<Tensor[]> {
const mappedInputs = inputs.reduce((map, tensor, index) => {

@@ -336,3 +372,4 @@ map[this.inputs[index].name] = tensor;

return this.executeAsync(mappedInputs, this.outputNodes, true);
return this._executeAsync(
mappedInputs, this.outputNodes, true, tensorArrayMap, tensorListMap);
}

@@ -344,7 +381,7 @@ /**

* @param context the execution context object for current execution.
* @param disableWarning disable no async op warning
* @param isFunctionExecution Flag for executing a function.
*/
private async executeWithControlFlow(
inputs: NamedTensorMap, context: ExecutionContext, outputNames: string[],
disableWarning: boolean): Promise<NamedTensorsMap> {
isFunctionExecution: boolean): Promise<NamedTensorsMap> {
const names = Object.keys(inputs);

@@ -378,3 +415,3 @@ const inputNodes =

}
if (dynamicNode == null && !disableWarning) {
if (dynamicNode == null && !isFunctionExecution) {
console.warn(

@@ -381,0 +418,0 @@ `This model execution did not contain any nodes with control flow ` +

2

src/executor/model_analysis.ts

@@ -133,3 +133,3 @@ /**

'Switch', 'Merge', 'Enter', 'Exit', 'NextIteration', 'StatelessIf',
'StatelessWhile'
'StatelessWhile', 'if', 'While'
];

@@ -136,0 +136,0 @@ const DYNAMIC_SHAPE_OPS = [

29

src/executor/tensor_array.ts

@@ -18,3 +18,4 @@ /**

import {concat, DataType, slice, stack, Tensor, tensor, tidy, unstack} from '@tensorflow/tfjs-core';
import {concat, DataType, keep, scalar, slice, stack, Tensor, tensor, tidy, unstack} from '@tensorflow/tfjs-core';
import {assertShapesMatchAllowUndefinedSize} from './tensor_utils';

@@ -33,15 +34,17 @@

export class TensorArray {
private static nextId = 0;
private tensors: TensorWithState[] = [];
private closed_ = false;
readonly id: number;
readonly idTensor: Tensor;
constructor(
public readonly name: string, public readonly dtype: DataType,
private maxSize: number, private elementShape: number[],
public readonly identicalElementShapes: boolean,
public readonly dynamicSize: boolean,
public readonly clearAfterRead: boolean) {
this.id = TensorArray.nextId++;
readonly name: string, readonly dtype: DataType, private maxSize: number,
private elementShape: number[], readonly identicalElementShapes: boolean,
readonly dynamicSize: boolean, readonly clearAfterRead: boolean) {
this.idTensor = scalar(0);
keep(this.idTensor);
}
get id() {
return this.idTensor.id;
}
get closed() {

@@ -52,3 +55,3 @@ return this.closed_;

/**
* Close the current TensorArray.
* Dispose the tensors and idTensor and mark the TensoryArray as closed.
*/

@@ -59,2 +62,3 @@ clearAndClose() {

this.closed_ = true;
this.idTensor.dispose();
}

@@ -138,3 +142,3 @@

if (t && t.read) {
if (t.read) {
throw new Error(

@@ -145,3 +149,3 @@ `TensorArray ${this.name}: Could not write to TensorArray index ${

if (t && t.written) {
if (t.written) {
throw new Error(

@@ -153,2 +157,3 @@ `TensorArray ${this.name}: Could not write to TensorArray index ${

t.tensor = tensor;
keep(tensor);
t.written = true;

@@ -155,0 +160,0 @@

58

src/executor/tensor_list.ts

@@ -18,3 +18,3 @@ /**

import {concat, DataType, slice, stack, Tensor, tensor, tidy, unstack} from '@tensorflow/tfjs-core';
import {concat, DataType, keep, scalar, slice, stack, Tensor, tensor, tidy, unstack} from '@tensorflow/tfjs-core';

@@ -39,2 +39,8 @@ import {assertShapesMatchAllowUndefinedSize} from './tensor_utils';

export class TensorList {
readonly idTensor: Tensor;
maxNumElements: number;
get id() {
return this.idTensor.id;
}
/**

@@ -49,5 +55,21 @@ *

constructor(
public tensors: Tensor[], public elementShape: number[],
public elementDtype: DataType, public maxNumElements = -1) {}
readonly tensors: Tensor[], readonly elementShape: number[],
readonly elementDtype: DataType, maxNumElements = -1) {
if (tensors != null) {
tensors.forEach(tensor => {
if (elementDtype !== tensor.dtype) {
throw new Error(`Invalid data types; op elements ${
elementDtype}, but list elements ${tensor.dtype}`);
}
assertShapesMatchAllowUndefinedSize(
elementShape, tensor.shape, 'TensorList shape mismatch: ');
keep(tensor);
});
}
this.idTensor = scalar(0);
this.maxNumElements = maxNumElements;
keep(this.idTensor);
}
/**

@@ -62,2 +84,10 @@ * Get a new TensorList containing a copy of the underlying tensor container.

/**
* Dispose the tensors and idTensor and clear the tensor list.
*/
clearAndClose() {
this.tensors.forEach(tensor => tensor.dispose());
this.tensors.length = 0;
this.idTensor.dispose();
}
/**
* The size of the tensors in the tensor list.

@@ -133,2 +163,3 @@ */

}
keep(tensor);
this.tensors.push(tensor);

@@ -201,3 +232,3 @@ }

this.elementShape, tensor.shape, 'TensorList shape mismatch: ');
keep(tensor);
this.tensors[elementIndex] = tensor;

@@ -267,3 +298,4 @@ }

*/
export function fromTensor(tensor: Tensor, elementShape: number[]) {
export function fromTensor(
tensor: Tensor, elementShape: number[], elementDtype: DataType) {
const dtype = tensor.dtype;

@@ -274,3 +306,6 @@ if (tensor.shape.length < 1) {

}
if (tensor.dtype !== elementDtype) {
throw new Error(`Invalid data types; op elements ${
tensor.dtype}, but list elements ${elementDtype}`);
}
const outputShape = tensor.shape.slice(1);

@@ -280,7 +315,3 @@ assertShapesMatchAllowUndefinedSize(

const tensorList: Tensor[] = [];
for (let i = 0; i < tensor.shape[0]; ++i) {
const tmp = tensor.slice(i, i + 1).reshape(outputShape);
tensorList.push(tmp);
}
const tensorList: Tensor[] = tensor.unstack();
return new TensorList(tensorList, elementShape, dtype);

@@ -309,3 +340,3 @@ }

tensor: Tensor, indices: number[], elementShape: number[],
numElements: number): TensorList {
numElements?: number): TensorList {
if (indices.length !== tensor.shape[0]) {

@@ -318,3 +349,3 @@ throw new Error(`Expected len(indices) == tensor.shape[0], but saw: ${

if (numElements !== -1 && maxIndex >= numElements) {
if (numElements != null && numElements !== -1 && maxIndex >= numElements) {
throw new Error(

@@ -363,2 +394,3 @@ `Max index must be < array size (${maxIndex} vs. ${numElements})`);

}
tensor.dispose();
return tensors;

@@ -365,0 +397,0 @@ });

11

src/executor/types.ts

@@ -19,6 +19,13 @@ /**

import {Tensor} from '@tensorflow/tfjs-core';
import {NamedTensorsMap} from '../data/types';
import {NamedTensorsMap, TensorArrayMap, TensorListMap} from '../data/types';
/**
*
*/
export interface FunctionExecutor {
executeFunctionAsync(inputs: Tensor[]): Promise<Tensor[]>;
executeFunctionAsync(
inputs: Tensor[], tensorArrayMap: TensorArrayMap,
tensorListMap: TensorListMap): Promise<Tensor[]>;
weightMap: NamedTensorsMap;
}

295

src/operations/executors/control_executor.ts

@@ -24,2 +24,3 @@ /**

import {TensorArray} from '../../executor/tensor_array';
import {fromTensor, reserve, scatter, split} from '../../executor/tensor_list';
import {InternalOpAsyncExecutor, Node} from '../types';

@@ -45,5 +46,7 @@

if (condValue[0]) {
return context.functionMap[thenFunc].executeFunctionAsync(args);
return context.functionMap[thenFunc].executeFunctionAsync(
args, context.tensorArrayMap, context.tensorListMap);
} else {
return context.functionMap[elseFunc].executeFunctionAsync(args);
return context.functionMap[elseFunc].executeFunctionAsync(
args, context.tensorArrayMap, context.tensorListMap);
}

@@ -59,20 +62,55 @@ }

getParamValue('args', node, tensorMap, context) as tfc.Tensor[];
const condTensor =
(await context.functionMap[condFunc].executeFunctionAsync(args))[0];
let condValue = await condTensor.data();
// Calculate the condition of the loop
const condResult =
(await context.functionMap[condFunc].executeFunctionAsync(
args, context.tensorArrayMap, context.tensorListMap));
const argIds = args.map(tensor => tensor.id);
let condValue = await condResult[0].data();
// Dispose the intermediate tensors for condition function
condResult.forEach(tensor => {
if (!tensor.kept && argIds.indexOf(tensor.id) === -1) {
tensor.dispose();
}
});
let result: tfc.Tensor[] = args;
while (condValue[0]) {
result =
await context.functionMap[bodyFunc].executeFunctionAsync(result);
const condTensor =
// Record the previous result for intermediate tensor tracking
const origResult = result;
// Execution the body of the loop
result = await context.functionMap[bodyFunc].executeFunctionAsync(
result, context.tensorArrayMap, context.tensorListMap);
const resultIds = result.map(tensor => tensor.id);
// Dispose the intermediate tensor for body function that is not global
// kept, not input/output of the body function
origResult.forEach(tensor => {
if (!tensor.kept && argIds.indexOf(tensor.id) === -1 &&
resultIds.indexOf(tensor.id) === -1) {
tensor.dispose();
}
});
// Recalcuate the condition of the loop using the latest results.
const condResult =
(await context.functionMap[condFunc].executeFunctionAsync(
result))[0];
condValue = await condTensor.data();
result, context.tensorArrayMap, context.tensorListMap));
condValue = await condResult[0].data();
// Dispose the intermediate tensors for condition function
condResult.forEach(tensor => {
if (!tensor.kept && argIds.indexOf(tensor.id) === -1 &&
resultIds.indexOf(tensor.id) === -1) {
tensor.dispose();
}
});
}
return result;
}
case 'LoopCond':
case 'LoopCond': {
return [
(getParamValue('pred', node, tensorMap, context) as tfc.Tensor).clone()
];
}
case 'Switch': {

@@ -87,3 +125,3 @@ const pred =

}
case 'Merge':
case 'Merge': {
const inputName = node.inputNames.find(

@@ -93,4 +131,4 @@ name => getTensor(name, tensorMap, context) !== undefined);

undefined;
case 'Enter':
}
case 'Enter': {
const frameId =

@@ -102,4 +140,4 @@ getParamValue('frameName', node, tensorMap, context) as string;

return [data.clone()];
case 'Exit':
}
case 'Exit': {
const tensor =

@@ -109,4 +147,4 @@ getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;

return [tensor.clone()];
case 'NextIteration':
}
case 'NextIteration': {
const input =

@@ -116,4 +154,4 @@ getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;

return [input.clone()];
case 'TensorArrayV3':
}
case 'TensorArrayV3': {
const size = getParamValue('size', node, tensorMap, context) as number;

@@ -136,25 +174,26 @@ const dtype =

context.addTensorArray(tensorArray);
return [scalar(tensorArray.id), scalar(1.0)];
case 'TensorArrayWriteV3':
const id =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
return [tensorArray.idTensor, scalar(1.0)];
}
case 'TensorArrayWriteV3': {
const id = getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const index = getParamValue('index', node, tensorMap, context) as number;
const writeTensor =
getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const writeTensorArray = context.getTensorArray(id);
const writeTensorArray = context.getTensorArray(id.id);
writeTensorArray.write(index, writeTensor);
return [scalar(1.0)];
case 'TensorArrayReadV3':
const readId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
return [writeTensorArray.idTensor];
}
case 'TensorArrayReadV3': {
const readId = getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const readIndex =
getParamValue('index', node, tensorMap, context) as number;
const readTensorArray = context.getTensorArray(readId);
const readTensorArray = context.getTensorArray(readId.id);
return [readTensorArray.read(readIndex)];
case 'TensorArrayGatherV3':
}
case 'TensorArrayGatherV3': {
const gatherId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const gatherIndices =

@@ -164,8 +203,9 @@ getParamValue('indices', node, tensorMap, context) as number[];

getParamValue('dtype', node, tensorMap, context) as tfc.DataType;
const gatherTensorArray = context.getTensorArray(gatherId);
const gatherTensorArray = context.getTensorArray(gatherId.id);
return [gatherTensorArray.gather(gatherIndices, gatherDtype)];
case 'TensorArrayScatterV3':
}
case 'TensorArrayScatterV3': {
const scatterId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const scatterIndices =

@@ -175,17 +215,19 @@ getParamValue('indices', node, tensorMap, context) as number[];

getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const scatterTensorArray = context.getTensorArray(scatterId);
const scatterTensorArray = context.getTensorArray(scatterId.id);
scatterTensorArray.scatter(scatterIndices, scatterTensor);
return [scalar(1.0)];
case 'TensorArrayConcatV3':
return [scatterTensorArray.idTensor];
}
case 'TensorArrayConcatV3': {
const concatId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
const concatTensorArray = context.getTensorArray(concatId);
getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const concatTensorArray = context.getTensorArray(concatId.id);
const concatDtype =
getParamValue('dtype', node, tensorMap, context) as tfc.DataType;
return [concatTensorArray.concat(concatDtype)];
case 'TensorArraySplitV3':
}
case 'TensorArraySplitV3': {
const splitId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const splitTensor =

@@ -195,18 +237,151 @@ getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;

getParamValue('lengths', node, tensorMap, context) as number[];
const splitTensorArray = context.getTensorArray(splitId);
const splitTensorArray = context.getTensorArray(splitId.id);
splitTensorArray.split(lengths, splitTensor);
return [scalar(1.0)];
case 'TensorArraySizeV3':
const sizeId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
const sizeTensorArray = context.getTensorArray(sizeId);
return [splitTensorArray.idTensor];
}
case 'TensorArraySizeV3': {
const sizeId = getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const sizeTensorArray = context.getTensorArray(sizeId.id);
return [scalar(sizeTensorArray.size(), 'int32')];
case 'TensorArrayCloseV3':
}
case 'TensorArrayCloseV3': {
const closeId =
getParamValue('tensorArrayId', node, tensorMap, context) as number;
const closeTensorArray = context.getTensorArray(closeId);
getParamValue('tensorArrayId', node, tensorMap, context) as
tfc.Tensor;
const closeTensorArray = context.getTensorArray(closeId.id);
closeTensorArray.clearAndClose();
return [scalar(0)];
return [closeTensorArray.idTensor];
}
case 'TensorListSetItem': {
const idTensor =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const index = getParamValue('index', node, tensorMap, context) as number;
const writeTensor =
getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const tensorList = context.getTensorList(idTensor.id);
tensorList.setItem(index, writeTensor);
return [tensorList.idTensor];
}
case 'TensorListGetItem': {
const idTensor =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const readIndex =
getParamValue('index', node, tensorMap, context) as number;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const elementDType =
getParamValue('elementDType', node, tensorMap, context) as
tfc.DataType;
const tensorList = context.getTensorList(idTensor.id);
return [tensorList.getItem(readIndex, elementShape, elementDType)];
}
case 'TensorListScatterV2':
case 'TensorListScatter': {
const scatterIndices =
getParamValue('indices', node, tensorMap, context) as number[];
const scatterTensor =
getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const numElements =
getParamValue('numElements', node, tensorMap, context) as number;
const tensorList =
scatter(scatterTensor, scatterIndices, elementShape, numElements);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
case 'TensorListReserve': {
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const elementDtype =
getParamValue('elementDType', node, tensorMap, context) as
tfc.DataType;
const numElements =
getParamValue('numElements', node, tensorMap, context) as number;
const tensorList = reserve(elementShape, elementDtype, numElements);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
case 'TensorListGather': {
const gatherId =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const gatherIndices =
getParamValue('indices', node, tensorMap, context) as number[];
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const elementDtype =
getParamValue('elementDType', node, tensorMap, context) as
tfc.DataType;
const tensorList = context.getTensorList(gatherId.id);
return [tensorList.gather(gatherIndices, elementDtype, elementShape)];
}
case 'TensorListStack': {
const idTensor =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const elementDtype =
getParamValue('elementDType', node, tensorMap, context) as
tfc.DataType;
const numElements =
getParamValue('numElements', node, tensorMap, context) as number;
const tensorList = context.getTensorList(idTensor.id);
return [tensorList.stack(elementShape, elementDtype, numElements)];
}
case 'TensorListFromTensor': {
const tensor =
getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const elementDtype =
getParamValue('elementDType', node, tensorMap, context) as
tfc.DataType;
const tensorList = fromTensor(tensor, elementShape, elementDtype);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
case 'TensorListConcat': {
const concatId =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const tensorList = context.getTensorList(concatId.id);
const concatDtype =
getParamValue('dtype', node, tensorMap, context) as tfc.DataType;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
return [tensorList.concat(concatDtype, elementShape)];
}
case 'TensorListPushBack': {
const idTensor =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const writeTensor =
getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const tensorList = context.getTensorList(idTensor.id);
tensorList.pushBack(writeTensor);
return [tensorList.idTensor];
}
case 'TensorListPopBack': {
const idTensor =
getParamValue('tensorListId', node, tensorMap, context) as tfc.Tensor;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const elementDType =
getParamValue('elementDType', node, tensorMap, context) as
tfc.DataType;
const tensorList = context.getTensorList(idTensor.id);
return [tensorList.popBack(elementShape, elementDType)];
}
case 'TensorListSplit': {
const splitTensor =
getParamValue('tensor', node, tensorMap, context) as tfc.Tensor;
const elementShape =
getParamValue('elementShape', node, tensorMap, context) as number[];
const lengths =
getParamValue('lengths', node, tensorMap, context) as number[];
const tensorList = split(splitTensor, lengths, elementShape);
context.addTensorList(tensorList);
return [tensorList.idTensor];
}
default:

@@ -213,0 +388,0 @@ throw TypeError(`Node type ${node.op} is not implemented`);

33

src/operations/executors/convolution_executor.ts

@@ -24,3 +24,3 @@ /**

import {getParamValue} from './utils';
import {getPadding, getParamValue} from './utils';

@@ -50,3 +50,3 @@ export const executeOp: InternalOpExecutor = (node: Node,

getParamValue('strides', node, tensorMap, context) as number[];
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
const dataFormat =

@@ -93,3 +93,3 @@ (getParamValue('dataFormat', node, tensorMap, context) as string)

getParamValue('strides', node, tensorMap, context) as number[];
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
const dataFormat =

@@ -127,3 +127,3 @@ (getParamValue('dataFormat', node, tensorMap, context) as string)

getParamValue('strides', node, tensorMap, context) as number[];
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
return [tfc.conv2dTranspose(

@@ -139,3 +139,3 @@ getParamValue('x', node, tensorMap, context) as tfc.Tensor3D |

getParamValue('strides', node, tensorMap, context) as number[];
const pad = getParamValue('pad', node, tensorMap, context);
const pad = getPadding(node, tensorMap, context);
const dilations =

@@ -239,2 +239,25 @@ getParamValue('dilations', node, tensorMap, context) as number[];

case 'Dilation2D': {
const strides =
getParamValue('strides', node, tensorMap, context) as number[];
const pad = getParamValue('pad', node, tensorMap, context);
const dilations =
getParamValue('dilations', node, tensorMap, context) as number[];
// strides: [1, stride_height, stride_width, 1].
const strideHeight = strides[1];
const strideWidth = strides[2];
// dilations: [1, dilation_height, dilation_width, 1].
const dilationHeight = dilations[1];
const dilationWidth = dilations[2];
return [tfc.dilation2d(
getParamValue('x', node, tensorMap, context) as tfc.Tensor3D |
tfc.Tensor4D,
getParamValue('filter', node, tensorMap, context) as tfc.Tensor3D,
[strideHeight, strideWidth], pad as 'valid' | 'same',
[dilationHeight, dilationWidth], 'NHWC' /* dataFormat */)];
}
default:

@@ -241,0 +264,0 @@ throw TypeError(`Node type ${node.op} is not implemented`);

15

src/operations/executors/dynamic_executor.ts
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -31,2 +31,3 @@ * you may not use this file except in compliance with the License.

case 'NonMaxSuppressionV5':
case 'NonMaxSuppressionV4':
case 'NonMaxSuppressionV3':

@@ -56,2 +57,14 @@ case 'NonMaxSuppressionV2': {

if (node.op === 'NonMaxSuppressionV4') {
const padToMaxOutputSize =
getParamValue('padToMaxOutputSize', node, tensorMap, context) as
boolean;
const result = await tfc.image.nonMaxSuppressionPaddedAsync(
boxes as tfc.Tensor2D, scores as tfc.Tensor1D, maxOutputSize,
iouThreshold, scoreThreshold, padToMaxOutputSize);
return [result.selectedIndices, result.validOutputs];
}
return [await tfc.image.nonMaxSuppressionAsync(

@@ -58,0 +71,0 @@ boxes as tfc.Tensor2D, scores as tfc.Tensor1D, maxOutputSize,

2

src/operations/executors/evaluation_executor.ts
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -5,0 +5,0 @@ * you may not use this file except in compliance with the License.

8

src/operations/executors/image_executor.ts
/**
* @license
* Copyright 2018 Google Inc. All Rights Reserved.
* Copyright 2018 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -27,5 +27,5 @@ * you may not use this file except in compliance with the License.

export const executeOp: InternalOpExecutor = (node: Node,
tensorMap: NamedTensorsMap,
context: ExecutionContext):
tfc.Tensor[] => {
tensorMap: NamedTensorsMap,
context: ExecutionContext):
tfc.Tensor[] => {
switch (node.op) {

@@ -32,0 +32,0 @@ case 'ResizeBilinear': {

20

src/operations/executors/slice_join_executor.ts

@@ -27,5 +27,5 @@ /**

export const executeOp: InternalOpExecutor = (node: Node,
tensorMap: NamedTensorsMap,
context: ExecutionContext):
tfc.Tensor[] => {
tensorMap: NamedTensorsMap,
context: ExecutionContext):
tfc.Tensor[] => {
switch (node.op) {

@@ -81,9 +81,3 @@ case 'ConcatV2':

const tensor = getParamValue('x', node, tensorMap, context) as tfc.Tensor;
if (begin.length === 1 && tensor.shape.length > 1) {
for (let i = 1; i < tensor.shape.length; i++) {
begin.push(0);
end.push(tensor.shape[i]);
strides.push(strides[0]);
}
}
return [tfc.stridedSlice(

@@ -131,5 +125,5 @@ tensor, begin, end, strides, beginMask, endMask, ellipsisMask,

number[];
return tfc.split(
getParamValue('x', node, tensorMap, context) as tfc.Tensor,
numOrSizeSplits, axis);
const tensor = getParamValue('x', node, tensorMap, context) as tfc.Tensor;
return tfc.split(tensor, numOrSizeSplits, axis);
}

@@ -136,0 +130,0 @@ case 'ScatterNd': {

12

src/operations/executors/test_helper.ts

@@ -43,2 +43,7 @@ /**

}
export function createShapeAttrFromIndex(inputIndex: number): InputParamValue {
return {inputIndexStart: inputIndex, type: 'shape'};
}
export function createNumericArrayAttr(value: number[]): ParamValue {

@@ -71,3 +76,3 @@ return {value, type: 'number[]'};

opMappers.find(mapper => mapper.tfOpName === node.op);
return Object.keys(node.inputParams).every(key => {
const matched = Object.keys(node.inputParams).every(key => {
const value = node.inputParams[key];

@@ -83,2 +88,7 @@ const def = opMapper.inputs.find(param => param.name === key);

});
if (!matched) {
console.log('node = ', node);
console.log('opMapper = ', opMapper);
}
return matched;
}

15

src/operations/executors/transformation_executor.ts

@@ -24,3 +24,3 @@ /**

import {getParamValue, split} from './utils';
import {getParamValue} from './utils';

@@ -58,5 +58,4 @@ export const executeOp: InternalOpExecutor = (node: Node,

getParamValue('x', node, tensorMap, context) as tfc.Tensor,
split(
getParamValue('padding', node, tensorMap, context) as number[],
2) as Array<[number, number]>,
getParamValue('padding', node, tensorMap, context) as
Array<[number, number]>,
getParamValue('constantValue', node, tensorMap, context) as number)];

@@ -67,4 +66,4 @@ }

getParamValue('blockShape', node, tensorMap, context) as number[];
const paddings = split(
getParamValue('paddings', node, tensorMap, context) as number[], 2);
const paddings =
getParamValue('paddings', node, tensorMap, context) as number[][];
return [tfc.spaceToBatchND(

@@ -77,4 +76,4 @@ getParamValue('x', node, tensorMap, context) as tfc.Tensor,

getParamValue('blockShape', node, tensorMap, context) as number[];
const crops = split(
getParamValue('crops', node, tensorMap, context) as number[], 2);
const crops =
getParamValue('crops', node, tensorMap, context) as number[][];
return [tfc.batchToSpaceND(

@@ -81,0 +80,0 @@ getParamValue('x', node, tensorMap, context) as tfc.Tensor,

28

src/operations/executors/utils.ts

@@ -43,6 +43,8 @@ /**

}
const data = Array.prototype.slice.call(
getTensor(node.inputNames.slice(start)[0], tensorMap, context)
.dataSync());
return inputParam.type === 'number' ? data[0] : data;
const tensor =
getTensor(node.inputNames.slice(start)[0], tensorMap, context);
const data = tensor.dataSync();
return inputParam.type === 'number' ?
data[0] :
tfc.util.toNestedArray(tensor.shape, data);
}

@@ -120,1 +122,19 @@ const attrParam = node.attrParams[paramName];

}
export function getPadding(
node: Node, tensorMap: NamedTensorsMap,
context: ExecutionContext): ValueType {
let pad = getParamValue('pad', node, tensorMap, context);
if (pad === 'explicit') {
// This is 1d array, we need to convert it to 2d array
pad = getParamValue('explicitPaddings', node, tensorMap, context);
const explicitPadding: [
[number, number], [number, number], [number, number], [number, number]
] = [[0, 0], [0, 0], [0, 0], [0, 0]];
for (let i = 0; i < 4; i++) {
explicitPadding[i][0] = (pad as number[])[i * 2];
explicitPadding[i][1] = (pad as number[])[i * 2 + 1];
}
return explicitPadding;
}
return pad;
}

148

src/operations/op_list/control.ts

@@ -94,3 +94,3 @@ import {OpMapper} from '../types';

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'index', 'type': 'number'},

@@ -108,3 +108,3 @@ {'start': 2, 'name': 'tensor', 'type': 'tensor'},

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'index', 'type': 'number'},

@@ -124,3 +124,3 @@ {'start': 2, 'name': 'flowIn', 'type': 'number'},

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'indices', 'type': 'number[]'},

@@ -138,3 +138,3 @@ {'start': 2, 'name': 'flowIn', 'type': 'number'},

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'indices', 'type': 'number[]'},

@@ -150,3 +150,3 @@ {'start': 2, 'name': 'tensor', 'type': 'tensor'},

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'flowIn', 'type': 'number'},

@@ -167,3 +167,3 @@ ],

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'tensor', 'type': 'tensor'},

@@ -179,3 +179,3 @@ {'start': 2, 'name': 'lengths', 'type': 'number[]'},

'inputs': [
{'start': 0, 'name': 'tensorArrayId', 'type': 'number'},
{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'},
{'start': 1, 'name': 'flowIn', 'type': 'number'}

@@ -187,3 +187,3 @@ ]

'category': 'control',
'inputs': [{'start': 0, 'name': 'tensorArrayId', 'type': 'number'}]
'inputs': [{'start': 0, 'name': 'tensorArrayId', 'type': 'tensor'}]
},

@@ -235,3 +235,133 @@ {

]
}
},
{
'tfOpName': 'TensorListScatter',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensor', 'type': 'tensor'},
{'start': 1, 'name': 'indices', 'type': 'number[]'},
{'start': 2, 'name': 'elementShape', 'type': 'shape'}
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListScatterV2',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensor', 'type': 'tensor'},
{'start': 1, 'name': 'indices', 'type': 'number[]'},
{'start': 2, 'name': 'elementShape', 'type': 'shape'},
{'start': 3, 'name': 'numElements', 'type': 'number'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListGather',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
{'start': 1, 'name': 'indices', 'type': 'number[]'},
{'start': 2, 'name': 'elementShape', 'type': 'shape'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListGetItem',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
{'start': 1, 'name': 'index', 'type': 'number'},
{'start': 2, 'name': 'elementShape', 'type': 'shape'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListSetItem',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
{'start': 1, 'name': 'index', 'type': 'number'},
{'start': 2, 'name': 'tensor', 'type': 'tensor'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListReserve',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'elementShape', 'type': 'shape'},
{'start': 1, 'name': 'numElements', 'type': 'number'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListFromTensor',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensor', 'type': 'tensor'},
{'start': 1, 'name': 'elementShape', 'type': 'shape'}
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListStack',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
{'start': 1, 'name': 'elementShape', 'type': 'shape'},
],
'attrs': [
{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'},
{'tfName': 'num_elements', 'name': 'numElements', 'type': 'dtype'}
]
},
{
'tfOpName': 'TensorListSplit',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensor', 'type': 'tensor'},
{'start': 1, 'name': 'elementShape', 'type': 'shape'},
{'start': 2, 'name': 'lengths', 'type': 'number[]'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListConcat',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
],
'attrs': [
{'tfName': 'element_shape', 'name': 'elementShape', 'type': 'shape'},
{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}
]
},
{
'tfOpName': 'TensorListPopBack',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
{'start': 1, 'name': 'elementShape', 'type': 'shape'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
{
'tfOpName': 'TensorListPushBack',
'category': 'control',
'inputs': [
{'start': 0, 'name': 'tensorListId', 'type': 'tensor'},
{'start': 1, 'name': 'tensor', 'type': 'tensor'},
],
'attrs':
[{'tfName': 'element_dtype', 'name': 'elementDType', 'type': 'dtype'}]
},
];

19

src/operations/op_list/convolution.ts

@@ -1,6 +0,4 @@

import {OpMapper} from '../types';
/**
* @license
* Copyright 2018 Google LLC. All Rights Reserved.
* Copyright 2020 Google LLC. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");

@@ -20,2 +18,4 @@ * you may not use this file except in compliance with the License.

import {OpMapper} from '../types';
export const json: OpMapper[] = [

@@ -333,3 +333,16 @@ {

],
},
{
'tfOpName': 'Dilation2D',
'category': 'convolution',
'inputs': [
{'start': 0, 'name': 'x', 'type': 'tensor'},
{'start': 1, 'name': 'filter', 'type': 'tensor'},
],
'attrs': [
{'tfName': 'strides', 'name': 'strides', 'type': 'number[]'},
{'tfName': 'rates', 'name': 'dilations', 'type': 'number[]'},
{'tfName': 'padding', 'name': 'pad', 'type': 'string'}
]
}
];

24

src/operations/op_list/dynamic.ts

@@ -43,2 +43,26 @@ import {OpMapper} from '../types';

{
'tfOpName': 'NonMaxSuppressionV4',
'category': 'dynamic',
'inputs': [
{'start': 0, 'name': 'boxes', 'type': 'tensor'},
{'start': 1, 'name': 'scores', 'type': 'tensor'},
{'start': 2, 'name': 'maxOutputSize', 'type': 'number'},
{'start': 3, 'name': 'iouThreshold', 'type': 'number'},
{'start': 4, 'name': 'scoreThreshold', 'type': 'number'}
],
'attrs': [
{'tfName': 'T', 'name': 'dtype', 'type': 'dtype', 'notSupported': true}, {
'tfName': 'T_threshold',
'name': 'threshold',
'type': 'dtype',
'notSupported': true
},
{
'tfName': 'pad_to_max_output_size',
'name': 'padToMaxOutputSize',
'type': 'bool'
}
]
},
{
'tfOpName': 'NonMaxSuppressionV5',

@@ -45,0 +69,0 @@ 'category': 'dynamic',

2

src/version.ts
/** @license See the LICENSE file. */
// This code is auto-generated, do not modify this file!
const version = '2.0.1';
const version = '2.1.0';
export {version};

34

tools/model_summary.ts

@@ -21,3 +21,3 @@ /**

if (argv.length < 3) {
console.log('Usage: ts-node pb2json.ts model_file');
console.log('Usage: yarn model-summary model_file');
return;

@@ -29,18 +29,34 @@ }

const rawdata = fs.readFileSync(sourcePath);
const nodes: Array<any> = JSON.parse(rawdata.toString())['modelTopology']['node'];
const model = JSON.parse(rawdata.toString());
if (model.format !== 'graph-model') {
console.log('This tool only supports TFJS Graph models.');
return;
}
// tslint:disable-next-line: no-any
let nodes: any[] = model['modelTopology']['node'];
const library = model['modelTopology']['library'];
if (library != null) {
const functions = library['function'];
// tslint:disable-next-line: no-any
if (functions != null) {
functions.forEach((func: any) => nodes = nodes.concat(func['nodeDef']));
}
}
const opCount: {[key: string]: number} = {};
for (const opNode of nodes) {
nodes.forEach(opNode => {
let count = 0;
const op = opNode['op'];
if (opCount[op]) {
count = opCount[op];
}
opCount[op] = count + 1;
}
count = opCount[op];
}
opCount[op] = count + 1;
});
console.log(opCount);
console.log('Total ops = ' + nodes.length);
const keys = Object.keys(opCount).sort();
keys.forEach(key => console.log(`${key}: ${opCount[key]}`));
console.log(`Total ops = ${nodes.length}`);
}
summarize(process.argv);
dist/executor/execution_context_test.js.map

Sorry, the diff of this file is not supported yet

dist/executor/execution_context.js.map

Sorry, the diff of this file is not supported yet

dist/executor/graph_executor.js.map

Sorry, the diff of this file is not supported yet

dist/executor/model_analysis.js.map

Sorry, the diff of this file is not supported yet

dist/executor/tensor_array_test.js.map

Sorry, the diff of this file is not supported yet

dist/executor/tensor_array.js.map

Sorry, the diff of this file is not supported yet

dist/executor/tensor_list_test.js.map

Sorry, the diff of this file is not supported yet

dist/executor/tensor_list.js.map

Sorry, the diff of this file is not supported yet

dist/miniprogram/index.js

Sorry, the diff of this file is too big to display

dist/miniprogram/index.js.map

Sorry, the diff of this file is not supported yet

dist/operations/custom_op/node_value_impl_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/arithmetic_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/basic_math_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/control_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/control_executor.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/convolution_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/convolution_executor.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/creation_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/dynamic_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/dynamic_executor.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/evaluation_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/graph_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/image_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/image_executor.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/logical_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/matrices_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/normalization_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/reduction_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/slice_join_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/slice_join_executor.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/spectral_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/test_helper.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/transformation_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/transformation_executor.js.map

Sorry, the diff of this file is not supported yet

dist/operations/executors/utils.js.map

Sorry, the diff of this file is not supported yet

dist/operations/op_list/control.js.map

Sorry, the diff of this file is not supported yet

dist/operations/op_list/convolution.js.map

Sorry, the diff of this file is not supported yet

dist/operations/op_list/dynamic.js.map

Sorry, the diff of this file is not supported yet

dist/operations/operation_executor_test.js.map

Sorry, the diff of this file is not supported yet

dist/tf-converter.es2017.js

Sorry, the diff of this file is too big to display

dist/tf-converter.es2017.js.map

Sorry, the diff of this file is not supported yet

dist/tf-converter.es2017.min.js

Sorry, the diff of this file is too big to display

dist/tf-converter.es2017.min.js.map

Sorry, the diff of this file is not supported yet

dist/tf-converter.js

Sorry, the diff of this file is too big to display

dist/tf-converter.js.map

Sorry, the diff of this file is not supported yet

dist/tf-converter.min.js

Sorry, the diff of this file is too big to display

dist/tf-converter.min.js.html

Sorry, the diff of this file is not supported yet

dist/tf-converter.min.js.map

Sorry, the diff of this file is not supported yet

dist/tf-converter.node.js

Sorry, the diff of this file is too big to display

dist/tf-converter.node.js.map

Sorry, the diff of this file is not supported yet

SocketSocket SOC 2 Logo

Product

  • Package Alerts
  • Integrations
  • Docs
  • Pricing
  • FAQ
  • Roadmap
  • Changelog

About

Packages

npm

Go

Maven

PyPI

Rubygems

Stay in touch

Get open source security insights delivered straight into your inbox.

  • Terms
  • Privacy
  • Security

Made with ⚡️ by Socket Inc