* Decode a (batch of) z vector into an image tensor. Z is the vector in latent * space that we want to generate an image for. * * Returns an image

Decode (i.e., translate) an encoded sentence. Args: input_seq: A `numpy.ndarray` of shape `(1, max_encoder_seq_length, num_encoder_tokens

* Discount and normalize reward values. * * This function performs two steps: * * 1. Discounts the reward values using `discountRate`. * 2. Norma

* Discount the reward values. * * @param {number[]} rewards The reward values to be discounted. * @param {number} discountRate Discount rate: a num

* Display an illustrative representation of the embeddings values * @param {*} embeddings * @param {*} parentEl

* * @param {*} tokenScores * @param {*} parentEl * @param {*} modelName

* Render the tokens * * @param {string[]} tokens the tokens * @param {Array.number[]} tokenScores model scores for each token * @param {Array.numb

* Render the tokens. * * @param {string[]} tokens tokens to display * @param {HTMLElement} parentEl parent element

* Draw a ROC curve. * * @param {tf.Tensor} targets The actual target labels, as a 1D Tensor * object consisting of only 0 and 1 values. * @param

* Draw a rectangle. * * A rectangle is a target object in the simple object detection task here. * Therefore, its bounding box is returned.

* Draw an equilateral triangle. * * A triangle are a target object in the simple object detection task here. * Therefore, its bounding box is

* Embed the input sentences using univeversal sentence encoder. * @param {string[]} sentences sentences to embed * @param {UniversalSentenceEncoder}

* Encode a string (e.g., a sentence) as a Tensor3D that can be fed directly * into the TensorFlow.js model.

* Run inference on some test Iris flower data. * * @param model The instance of `tf.Model` to run the inference with. * @param testDataset A tf.Dat

* Run inference on some test Iris flower data. * * @param model The instance of `tf.Model` to run the inference with. * @param xTest Test data feat

* Extract the first embedding matrix from a TensorFlow.js model. * * @param {tf.model} model An instance of tf.Model, assumed to contain an * Emb

Helper to display processed version of game state.

* Train the LSTM model. * * @param {number} numEpochs Number of epochs to train the model for. * @param {number} examplesPerEpoch Number of e

* Generate an image and preview it on the console. * * @param {tf.LayersModel} decoderModel Decoder portion of the VAE. * @param {number} latentDim

* Generate examples. * * Each example consists of a question, e.g., '123+456' and and an * answer, e.g., '579'. * * @param digits Maximum number

* Generate examples. * * Each example consists of a question, e.g., '123+456' and and an * answer, e.g., '579'. * * @param digits Maximum number

* Generate a single image example. * * @param {number} numCircles Number of circles (background object type) * to include. * @param {num

* Generate the example-begin indices; shuffle them randomly.

* Generates a representation of a latent space. * * Returns an array of tensors representing each dimension. Currently * each dimension is evenly s

* Generate text using the LSTM model. * * @param {number[]} sentenceIndices Seed sentence, represented as the * indices of the constituent

* Get actions based on a state-tensor input. * * @param {tf.Tensor} inputs A tf.Tensor instance of shape `[batchSize, 4]`. * @param {Float32A

* Get the set of unique characters from text.

* Returns a generator function that yields batches of data for training. * * @param {string} embeddingsPath path to ndjson file with token embedding

* Returns a list of all DOM image elements pointing to the provided srcUrl. * @param {string} srcUrl which url to search for, including 'http(s)://'

Get the required image sizes (height and width).

* Read the value of the latent-space vector fromthe sliders. * * @param {number} numRepeats Number of times to tile the single latent vector * fo

Returns the name of the main component registered from JavaScript. This is used to schedule rendering of the component.

Returns the name of the main component registered from JavaScript. This is used to schedule rendering of the component.

Get the mean and standard deviation of a data column.

* Get all training data as a data tensor and a labels tensor. * * @returns * xs: The data tensor, of shape `[numTrainExamples, 28, 28, 1]`.

* Get all training data as a data tensor and a labels tensor. * * @returns * xs: The data tensor, of shape `[numTrainExamples, 28, 28, 1]`.

Get the user's home directory (Node.js only).

* @license * Copyright 2022 Google LLC. All Rights Reserved. * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use th

Get IMDB data for training and validation. Args: vocabulary_size: Size of the vocabulary, as an `int`. max_len: Cut text after this number

* Get the sizes of the hidden layers. * * @returns {number | number[]} If the model has only one hidden layer, * return the size of the lay

Convert an iterable of word indices into words. Args: reverse_index: An `dict` mapping word index (as `int`) to word (as `str`). indices: A

* @license * Copyright 2022 Google LLC. All Rights Reserved. * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use th

* Initializes the app for the provided model metadata URL.

* Initializes the app for the provided TFLite model URL.

Loads Flipper in React Native templates. Call this in the onCreate method with something like initializeFlipper(this, getReactNativeHost().getReactIns

Loads Flipper in React Native templates. Call this in the onCreate method with something like initializeFlipper(this, getReactNativeHost().getReactIns

Loads Flipper in React Native templates. Call this in the onCreate method with something like initializeFlipper(this, getReactNativeHost().getReactIns