* Returns a diagonal tensor with given diagonal values. * * Given a diagonal, this operation returns a tensor with the diagonal and * everything el

* Computes the grayscale dilation over the input `x`. * * @param x The input tensor, rank 3 or rank 4 of shape * `[batch, height, width, depth]

* Get the bazel target from a tfjs package's directory. * * @param dir The tfjs package's directory. * @returns The bazel target to build the tfjs

* Attempt to dispose a LayersModel's weights. * * This method decrease the reference count of the LayersModel object by 1. * * A LayersMod

* Attempt to dispose layer's weights. * * This method decreases the reference count of the Layer object by 1. * * A Layer is reference-cou

* Dispose the variables (if any) owned by this optimizer instance.

* Dispose this LayersVariable instance from memory.

* Dispose `ResourceManager`, including its hashTables and tensors in them.

* Dispose the memory if the dataId has 0 refCount. Return true if the memory * is released or delayed in this backend, false if there are still

* Dispose the memory if the dataId has 0 refCount. Return true if the memory * is released, false otherwise. * @param dataId * @oaram force O

* Dispose the memory if the dataId has 0 refCount. Return true if the memory * is released, false otherwise. * @param dataId * @oaram force O

* Dispose the weight variables that this Layer instance holds. * * @returns {number} Number of disposed variables.

* Divides two `tf.Tensor`s element-wise, A / B. Supports broadcasting. Return 0 * if denominator is 0. * * * ```js * const a = tf.tensor1d([1, 4,

* Divides two `tf.Tensor`s element-wise, A / B. Supports broadcasting. * * ```js * const a = tf.tensor1d([1, 4, 9, 16]); * const b = tf.tensor1d([

* Computes the dot product of two matrices and/or vectors, `t1` and `t2`. * * ```js * const a = tf.tensor1d([1, 2]); * const b = tf.tensor2d([[1,

* Downloads and unpacks a given tarball or zip file at a given path. * @param {string} uri The path of the compressed file to download and extract.

* Computes dropout. * * ```js * const x = tf.tensor1d([1, 2, 2, 1]); * const rate = 0.75; * const output = tf.dropout(x, rate); * output.print()

* Dump the predict results of two backends and save diffs to files. * @param model The loaded model. * @param input The actual and expected results

* Computes exponential linear element-wise: `x > 0 ? x : (e ^ x) - 1`. * * ```js * const x = tf.tensor1d([-1, 1, -3, 2]); * * x.elu().print(); /

* Encode the provided string into an array of bytes using the provided * encoding.

* Checks the input tensor mathes the given shape. * * Given an input tensor, returns a new tensor with the same values as the * input tensor with s

Returns the smallest representable number.

Returns the smallest representable number.

* Returns the truth value of (a == b) element-wise. Supports broadcasting. * * ```js * const a = tf.tensor1d([1, 2, 3]); * const b = tf.tensor1d([

* Computes Gauss error function of the input `tf.Tensor` element-wise: * `erf(x)` * * ```js * const x = tf.tensor1d([0, .1, -.1, .7]); * * x.erf

* Computes the Euclidean norm of scalar, vectors, and matrices. * * ```js * const x = tf.tensor1d([1, 2, 3, 4]); * * x.euclideanNorm().print();

* Execute the inference for the input tensors and return activation * values for specified output node names without batching. * * @param inp

* Execute internal tensors of the model with input data feed. * @param inputs Input data feed. Must match the inputs of the model. * @param outp

* Execute the inference for the input tensors and return activation * values for specified output node names without batching. * * @param inp

* Execute the inference for the input tensors and return activation * values for specified output node names without batching. * * @param inp

* Executes the inference for given input tensors. * @param inputs Tensor map for the model inputs, keyed by the input node * names. * @param

* Executes the inference for given input tensors in Async fashion. * @param inputs Tensor map for the model inputs, keyed by the input node * na