@tensorflow/tfjs-layers
Version:
TensorFlow layers API in JavaScript
641 lines • 75.2 kB
JavaScript
/**
* @license
* Copyright 2018 Google LLC
*
* Use of this source code is governed by an MIT-style
* license that can be found in the LICENSE file or at
* https://opensource.org/licenses/MIT.
* =============================================================================
*/
/**
* deeplearn.js backend.
*/
import * as tfc from '@tensorflow/tfjs-core';
import { onesLike as coreOnesLike, scalar, tensor1d, tidy, where, zerosLike as coreZerosLike } from '@tensorflow/tfjs-core';
import { checkDataFormat } from '../common';
import { NotImplementedError, ValueError } from '../errors';
import * as math_utils from '../utils/math_utils';
import { imageDataFormat } from './common';
// tslint:enable
/* Setting and getting backend from deeplearn.js. */
// Default deeplearn.js backend is WebGL (GPU).
let backend = 'webgl';
export function setBackend(requestedBackend) {
tfc.setBackend(requestedBackend);
backend = requestedBackend;
}
export function getBackend() {
return backend;
}
/**
* Indicates whether the backend is operating symbolically.
*
* This function will be used to determine how to interpret user code. If
* it returns true, calls to the backend construct a symbolic graph; if
* it returns false, calls to the backend execute immediately.
*/
export function isBackendSymbolic() {
return false;
}
/**
* Get the number of elements in a Tensor.
* @param x The Tensor.
* @return Number of elements in `x`.
*/
export function countParams(x) {
const shape = x.shape;
if (shape.length > 0) {
return shape.reduce((a, b) => a * b);
}
else {
// Scalar.
return 1;
}
}
/**
* Casts a tensor to a different dtype and returns it.
* @param x Input tensor.
* @param dtype String: 'float32'|'int32'|'bool'.
* @returns Tensor of the specified `dtype`.
*/
export function cast(x, dtype) {
return tfc.cast(x, dtype);
}
/**
* Adds a 1-sized dimension at index "axis".
* @param x Input tensor.
* @param axis Position where to add the new axis.
* @returns Result of the dimension expansion.
*/
export function expandDims(x, axis = -1) {
const outShape = x.shape.slice();
if (axis < 0) {
axis = outShape.length + axis + 1;
}
outShape.splice(axis, 0, 1);
return tfc.reshape(x, outShape);
}
/**
* Repeats a 2D tensor.
*
* If `x` has shape `[samples, dim]` and `n` is 2, for example, the output
* will have shape `[samples, 2, dim]`.
*
* @param x Input tensor.
* @param n Integer, number of times to repeat.
* @returns The result of the repeat operation.
* @throws ValueError: If input tensor is not 2D.
*/
export function repeat(x, n) {
return tidy(() => {
if (x.shape.length !== 2) {
throw new ValueError(`repeat() expects a rank-2 tensor, but received a ` +
`rank-${x.shape.length} tensor.`);
}
const y = expandDims(x, 1);
return tile(y, [1, n, 1]);
});
}
/**
* Flatten a Tensor into 1D.
* @param x Input tensor.
* @return The result of the flattening `x`.
*/
export function flatten(x) {
const newShape = [math_utils.arrayProd(x.shape)];
return tfc.reshape(x, newShape);
}
/**
* Turn a nD tensor into a 2D tensor with same 0th dimension.
* In other words, it flattens each data samples of a batch.
*
* @param x The tensor to flatten. The rank of this tensor is required to be 2
* or higher.
* @return The result of the flattening.
*/
export function batchFlatten(x) {
if (x.rank <= 1) {
throw new ValueError(`batchFlatten requires a minimum rank of 2. Got rank: ${x.rank}.`);
}
const newShape = [x.shape[0], math_utils.arrayProd(x.shape, 1)];
return tfc.reshape(x, newShape);
}
/**
* Do slicing along the first axis.
* @param array input `tf.Tensor`.
* @param start starting index, inclusive.
* @param size size of the slice along the first axis.
* @returns result of the slicing.
* @throws ValueError: If `array` is of an unsupported subtype of `tf.Tensor`.
*/
export function sliceAlongFirstAxis(array, start, size) {
return tidy(() => {
switch (array.rank) {
case 1:
return tfc.slice1d(array, start, size);
case 2:
return tfc.slice2d(array, [start, 0], [size, array.shape[1]]);
case 3:
return tfc.slice3d(array, [start, 0, 0], [size, array.shape[1], array.shape[2]]);
case 4:
return tfc.slice4d(array, [start, 0, 0, 0], [size, array.shape[1], array.shape[2], array.shape[3]]);
case 5:
return tfc.slice(array, [start, 0, 0, 0, 0], [
size, array.shape[1], array.shape[2], array.shape[3], array.shape[4]
]);
case 6:
return tfc.slice(array, [start, 0, 0, 0, 0, 0], [
size, array.shape[1], array.shape[2], array.shape[3], array.shape[4],
array.shape[5]
]);
default:
throw new ValueError(`sliceAlongFirstAxis() received an unsupported tensor rank: ` +
`${array.rank}`);
}
});
}
/**
* Do slicing along the last axis.
* @param array input `tf.Tensor`.
* @param start starting index, inclusive.
* @param size size of the slice along the last axis.
* @returns result of the slicing.
* @throws ValueError: If `array` is of an unsupported subtype of `tf.Tensor`.
*/
export function sliceAlongLastAxis(array, start, size) {
return tidy(() => {
switch (array.rank) {
case 1:
return tfc.slice1d(array, start, size);
case 2:
return tfc.slice2d(array, [0, start], [array.shape[0], size]);
case 3:
return tfc.slice3d(array, [0, 0, start], [array.shape[0], array.shape[1], size]);
case 4:
return tfc.slice4d(array, [0, 0, 0, start], [array.shape[0], array.shape[1], array.shape[2], size]);
default:
throw new ValueError(`sliceAlongLastAxis() received an unsupported tensor rank: ` +
`${array.rank}`);
}
});
}
/**
* Do slicing along the sepcified axis.
* @param array input `tf.Tensor`.
* @param start starting index, inclusive.
* @param size of the slice along the chosen axis.
* @param choose an axis.
* @returns result of the slicing.
* @throws ValueError: If `array` is of an unsupported subtype of `tf.Tensor`.
*/
export function sliceAlongAxis(array, start, size, axis) {
return tidy(() => {
switch (array.rank) {
case 1:
return tfc.slice1d(array, start, size);
case 2:
switch (axis) {
case 1:
return sliceAlongFirstAxis(array, start, size);
case 2:
return sliceAlongLastAxis(array, start, size);
default:
throw new ValueError(`The axis is not within the rank of the tensor ` +
`${axis}`);
}
case 3:
switch (axis) {
case 1:
return sliceAlongFirstAxis(array, start, size);
case 2:
return tfc.slice3d(array, [0, start, 0], [array.shape[0], size, array.shape[2]]);
case 3:
return sliceAlongLastAxis(array, start, size);
default:
throw new ValueError(`The axis is not within the rank of the tensor ` +
`${axis}`);
}
case 4:
switch (axis) {
case 1:
return sliceAlongFirstAxis(array, start, size);
case 2:
return tfc.slice4d(array, [0, start, 0, 0], [array.shape[0], size, array.shape[2], array.shape[3]]);
case 3:
return tfc.slice4d(array, [0, 0, start, 0], [array.shape[0], array.shape[1], size, array.shape[3]]);
case 4:
return sliceAlongLastAxis(array, start, size);
default:
throw new ValueError(`The axis is not within the rank of the tensor ` +
`${axis}`);
}
default:
throw new ValueError(`sliceAlongLastAxis() received an unsupported tensor rank: ` +
`${array.rank}`);
}
});
}
/**
* Concatenates a list of tensors alongside the specified axis.
* @param tensors `Array` of tensors to concatenate.
* @param axis Concatenation axis.
* @returns The result of the concatenation.
*/
export function concatenate(tensors, axis = -1) {
let rank;
if (axis < 0) {
rank = tensors[0].rank;
if (rank !== 0) {
axis = rank;
}
else {
axis = 0;
}
}
if (axis === tensors[0].rank) {
// Porting Note: This is necessary because tfc.concat() requires axis to be
// in the interval [-rank, rank).
axis = -1;
}
// Porting Note: Sparse concat is not supported yet.
return tfc.concat(tensors, axis);
}
/**
* Concatenate two arrays along the first dimension.
* @param a The 1st `tf.Tensor` to concatenate.
* @param b The 2nd `tf.Tensor` to concatenate.
* @returns Result of the concatenation.
* @throws ValueError: If `a` is of an unsupported subtype of `tf.Tensor`.
*/
export function concatAlongFirstAxis(a, b) {
switch (a.rank) {
case 1:
return tfc.concat1d([a, b]);
case 2:
return tfc.concat2d([a, b], 0);
case 3:
return tfc.concat3d([a, b], 0);
case 4:
return tfc.concat4d([a, b], 0);
default:
throw new ValueError(`concatAlongFirstAxis() received an unsupported ` +
`tensor rank: ${a.rank}`);
}
}
/**
* Creates a tensor by tiling `x` by `n`.
* @param x A tensor.
* @param n An Array of integers or a single integer. If an Array, the length
* must be the same as the number of dimensions in `x`. If a single integer,
* it will be treated as an Array of length 1.
*/
export function tile(x, n) {
if (!Array.isArray(n)) {
n = [n];
}
if (x.rank !== n.length) {
throw new ValueError(`The length of input n (${n.length}) does not match ` +
`the number of dimensions in input x (${x.rank})`);
}
return tfc.tile(x, n);
}
/* Creation of random tensors. */
/**
* Get a tensor with normal distribution of values.
*
* @param shape Shape of the tensor.
* @param mean mean value of the normal distribution.
* @param stddev standard deviation of the normal distribution.
* @param dtype
* @param seed
* @return The normal tensor.
*/
export function randomNormal(shape, mean = 0.0, stddev = 1.0, dtype, seed) {
return tfc.randomNormal(shape, mean, stddev, dtype, seed);
}
/* Linear Algebra */
/**
* Multiply two tensors and returns the result as a tensor.
*
* For 2D tensors, this is equivalent to matrix multiplication (matMul).
* For tensors of higher ranks, it follows the Theano behavior,
* (e.g. `(2, 3) * (4, 3, 5) -> (2, 4, 5)`). From the Theano documentation:
*
* For N dimensions it is a sum product over the last axis of x and the
* second-to-last of y:
*
* @param a A tensor of at least rank 2.
* @param b A tensor of at least rank 2.
* @param activation (optional) A string identifying the activation
* function.
* @return Result of the dot operation.
*/
export function dot(a, b, activation, bias) {
if ((a.rank < 2) || (b.rank < 2)) {
throw new NotImplementedError(`dot requires both inputs to be rank >= 2` +
` but got x shape = ${a.shape} and y shape = ${b.shape}`);
}
if (b.rank >= 3) {
const xLastDim = a.shape.slice(-1)[0];
const ySecondLastDim = b.shape.slice(-2)[0];
if (xLastDim !== ySecondLastDim) {
throw new NotImplementedError(`If rank y >= 3, then the second last dim` +
` of y must equal the last dim of x but got x shape = ${a.shape} and ` +
` y shape = ${b.shape}`);
}
}
// Handle basic 2D x 2D case.
if ((a.rank === 2) && (b.rank === 2)) {
const transposeA = false;
const transposeB = false;
// tfc.fused.matMul only fuses certain activation functions. Unsupported
// activation functions are treated as 'linear' activations, which is
// equivalent to a no-op.
return tfc.fused.matMul({
a,
b: b,
transposeA,
transposeB,
bias: bias ? reshapeBias(a.rank, bias, imageDataFormat()) : null,
activation
});
}
else {
// Reshape x into the analogous 2D Tensor.
const aFirstDims = a.shape.slice(); // Holds all but the last dim of x.
const aLastDim = aFirstDims.pop();
a = tfc.reshape(a, [-1, aLastDim]);
// Reshape y into the analogous 2D Tensor, and keep track of the
// required dimensions to reproduce the output shape.
const bShape = b.shape.slice();
const bLastDim = bShape.pop();
const ySecondLastDim = bShape.pop();
const yOtherDims = [...bShape, bLastDim];
// permutation should be like [r-2, 0, 1, 2, ... r-4, r-3, r-1]
// where r is the rank of y.
const perm = Array.from({ length: b.rank }, (_, i) => {
if (i === 0) {
return b.rank - 2;
}
else if (i <= b.rank - 2) {
return i - 1;
}
return i;
});
b = tfc.reshape(tfc.transpose(b, perm), [ySecondLastDim, -1]);
// Multiply x and y as 2D Tensors, and then reshape back to original.
const outputShape = [...aFirstDims, ...yOtherDims];
const transposeA = false;
const transposeB = false;
return tfc.reshape(tfc.fused.matMul({
a,
b,
transposeA,
transposeB,
bias: bias ? reshapeBias(a.rank, bias, imageDataFormat()) : null,
activation
}), outputShape);
}
}
/**
* Compute the sign Tensor of an input Tensor.
*
* Elements of the input `tf.Tensor` that are === 0 are mapped to 0.
* Elements of the input `tf.Tensor` that are > 0 are mapped to 1.
* Elements of the input `tf.Tensor` that are < 0 are mapped to -1.
*
* @param x Input `tf.Tensor`.
* @return The sign `tf.Tensor`.
*/
export function sign(x) {
// TODO(cais): Move to the core.
return tidy(() => {
const zerosLikeX = coreZerosLike(x);
const onesLikeX = coreOnesLike(x);
return where(tfc.equal(x, zerosLikeX), zerosLikeX, where(tfc.greater(x, coreZerosLike(x)), onesLikeX, tfc.mul(-1, onesLikeX)));
});
}
/**
* Computes the one-hot representation of an integer tensor.
* @param indices nD integer tensor of shape
* `(batch_size, dim1, dim2, ... dim(n-1))`
* @param numClasses Integer, number of classes to consider.
* @returns (n + 1)D one hot representation of the input
* with shape `(batch_size, dim1, dim2, ... dim(n-1), num_classes)`
*/
export function oneHot(indices, numClasses) {
return tidy(() => {
if (indices.rank !== 1) {
throw new Error('Only 1D one-hot tensors are supported in the ' +
'deeplearn backend, at present.');
}
indices = tfc.cast(indices, 'int32');
return tfc.cast(tfc.oneHot(indices, numClasses), 'float32');
});
}
/* Elementary math functions. */
/**
* Retrieves the elements of indices `indices` in the tensor `reference`.
* @param reference A tensor.
* @param indices An integer tensor of indices or an `Array` of integers.
* @param axis Axis along which to perform the gather operation.
* @returns The result of the gathering as a tensor.
*/
export function gather(reference, indices, axis) {
return tidy(() => {
if (Array.isArray(indices)) {
indices = tensor1d(indices, 'int32');
}
else {
indices = tfc.cast(indices, 'int32');
}
return tfc.gather(reference, indices, axis);
});
}
/**
* Element-wise square.
* @param x Input tensor.
* @return element-wise x^2
*/
export function square(x) {
return tfc.mul(x, x);
}
/**
* Element-wise exponentiation.
*
* Porting Note: In PyKeras, `a` (the exponent) is a Python integer, which
* takes advatnage of the backend's (e.g., TensorFlow's) automatic
* conversion to tensor. Here we allow `a` to be either a number or a tensor.
*
* @param x The base tensor.
* @param a The exponent, tensor or number. If a number, it is rounded to the
* nearest integer and converted to a tensor.
* @returns A tensor of the same shape as `x`.
*/
export function pow(x, a) {
return tidy(() => {
if (typeof (a) === 'number') {
a = scalar(Math.round(a), 'int32');
}
if (a.dtype !== 'int32') {
throw new NotImplementedError(`Non-int32 dtype (${a.dtype}) is not supported by pow() yet`);
}
return tfc.pow(x, a);
});
}
/**
* Reshapes bias tensor according to rank of x.
*/
function reshapeBias(xRank, bias, dataFormat) {
const biasShape = bias.shape;
if (bias.rank !== 1 && bias.rank !== xRank) {
throw new ValueError(`Unexpected bias dimensions: ${bias.rank}` +
`; expected it to be 1 or ${xRank}`);
}
if (xRank === 5) {
if (dataFormat === 'channelsFirst') {
if (biasShape.length === 1) {
return tfc.reshape(bias, [1, biasShape[0], 1, 1, 1]);
}
else {
return tfc.reshape(bias, [1, biasShape[3], biasShape[0], biasShape[1], biasShape[2]]);
}
}
else if (dataFormat === 'channelsLast') {
if (biasShape.length === 1) {
return tfc.reshape(bias, [1, 1, 1, 1, biasShape[0]]);
}
else {
return tfc.reshape(bias, [1].concat(biasShape));
}
}
}
else if (xRank === 4) {
if (dataFormat === 'channelsFirst') {
if (biasShape.length === 1) {
return tfc.reshape(bias, [1, biasShape[0], 1, 1]);
}
else {
return tfc.reshape(bias, [1, biasShape[2], biasShape[0], biasShape[1]]);
}
}
else if (dataFormat === 'channelsLast') {
if (biasShape.length === 1) {
return tfc.reshape(bias, [1, 1, 1, biasShape[0]]);
}
else {
return tfc.reshape(bias, [1].concat(biasShape));
}
}
}
else if (xRank === 3) {
if (dataFormat === 'channelsFirst') {
if (biasShape.length === 1) {
return tfc.reshape(bias, [1, biasShape[0], 1]);
}
else {
return tfc.reshape(bias, [1, biasShape[1], biasShape[0]]);
}
}
else if (dataFormat === 'channelsLast') {
if (biasShape.length === 1) {
return tfc.reshape(bias, [1, 1, biasShape[0]]);
}
else {
return tfc.reshape(bias, [1].concat(biasShape));
}
}
}
else if (xRank < 3) {
return bias;
}
throw new ValueError(`Unsupported input rank by biasAdd: ${bias.rank}`);
}
/* Neural-network operations. */
/**
* Add a bias to a tensor.
*
* @param x The tensor to add the bias to.
* @param bias The bias to add to `x`. Must be 1D or the same rank as `x`.
* @return Result of the bias adding.
* @throws ValueError: If the rank of `bias` is incorrect.
*/
export function biasAdd(x, bias, dataFormat) {
return tidy(() => {
if (dataFormat == null) {
dataFormat = imageDataFormat();
}
checkDataFormat(dataFormat);
return tfc.add(x, reshapeBias(x.rank, bias, dataFormat));
});
}
/**
* Exponential linear unit (ELU).
* @param x A tensor or variable to compute the activation function for.
* @param alpha: A scalar, a scaling factor for the negative section.
* @return Output of the ELU operation.
*/
export function elu(x, alpha = 1) {
// TODO(cais): Add support for alpha values other than 1.
if (alpha !== 1) {
throw new NotImplementedError(`Support for alpha values other than 1 (${alpha}) is not implemented ` +
`yet.`);
}
return tfc.elu(x);
}
/**
* Softsign of a tensor.
*
* Defined as x / (abs(x) + 1), element-wise.
*
* @param x: Input.
* @returns Output.
*/
export function softsign(x) {
return tidy(() => tfc.div(x, tfc.add(tfc.abs(x), 1)));
}
/**
* Sets entries in `x` to zero at random, while scaling the entire tensor.
*
* @param x input tensor.
* @param level fraction of the entries in the tensor that will be set to 0.
* @param noiseShape shape of randomly generated keep/drop flags, must be
* broadcastable to the shape of `x`. Optional.
* @param seed random seed to ensure determinism. Optional.
* @returns Result of the dropout operation.
*/
export function dropout(x, level, noiseShape, seed) {
return tidy(() => tfc.dropout(x, level, noiseShape, seed));
}
/**
* Element-wise, segment-wise linear approximation of sigmoid.
*
* Returns `0.` if `x < -2.5`, `1.` if `x > 2.5`.
* In `-2.5 <= x <= 2.5`, returns `0.2 * x + 0.5`.
*
* @param x Input tensor.
* @returns Output tensor.
*/
export function hardSigmoid(x) {
return tidy(() => {
const y = tfc.add(.5, tfc.mul(.2, x));
return tfc.clipByValue(y, 0, 1);
});
}
/**
* Invoke `x` in the training phase, and `alt` otherwise.
*
* Porting Note: We do not create placeholder tensors for the `training`
* boolean flag here, because there is no such thing in the TF.js imperative
* backend.
*
* @param x The function to invoke iff `training` is `true`.
* @param alt The function to invoke iff `training` is `false`.
* @param training Boolean flag for whether training phase is active.
* @returns The return value of `x()` if `training` is `true`, or the return
* value of `alt()` if `training` is `false`.
*/
export function inTrainPhase(x, alt, training = false) {
return training ? x() : alt();
}
//# sourceMappingURL=data:application/json;base64,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