Generator for v2 (preactivation) ResNet models. This function generates a family of ResNet v2 models. See the resnet_v2_*() methods for specific

Test model on test data of length l using the given session.

Spatial Transformer Layer Implements a spatial transformer layer as described in [1]_. Based on [2]_ and edited by David Dao for Tensorflow.

Adds the layers defined by model_str[index:] to the model. Args: prev_layer: Input tensor. index: Position in model_str to start

0xff is in case char is signed.

The default action for a given state is assigning the most frequent tag.

Constructs a structured learning graph.

Create operations to read the MNIST input file. Args: mnist_data_file: Path of a file containing the MNIST images to process. batch_size: s

Returns a batch of inputs for seq2seq attention model. Returns: enc_batch: A batch of encoder inputs [batch_size, hps.enc_timestamps].

Returns the length (number of bytes) of the Unicode code point starting at src, based on inspecting just that one byte. This requires that src point t

Preprocesses the object. This will be called prior to calling Evaluate() or Compute() on that object.

Associates `op_name` key with `pxg_class` value. Registers `pxg_class` as the class that will be called to perform per-example differentiatio

Decodes CTC output to a string. Extracts only sequences of codes that are allowed by self.decoder. Labels that make illegal code sequences ar

Adds an LSTM layer with the given pre-parsed attributes. Always maps 4-D to 4-D regardless of summarize. Args: prev_layer: Input tensor

Pads or slices the 2nd dimension of 2-d labels to match logits_size. Covers the case of 1-d softmax output, when labels is [batch, seq] and logit

Verifies the number of parameters in the model.

Returns True if num_shards is compatible with FLAGS.num_threads.

Processes a complete data set and saves it as a TFRecord. Args: name: Unique identifier specifying the dataset. images: List of ImageMetada

Returns the argument from the feature function descriptor. It defaults to 0 if the argument has not been specified.

Compute the sanitized gradients. Args: loss: the loss tensor. var_list: the optional variables. add_noise: if true, then add no

Convolutional linear map.

Convert class labels from scalars to one-hot vectors.

Returns the i'th element.

Extract a MATLAB matrix into two numpy arrays with data and labels :param local_url: :return:

Returns the feature type for single-type feature functions.

Returns the number of feature types in the feature extractor. Invalid before Init() has been called.

Initialize data and model.

Returns a TF-Feature of int64s. Args: values: A scalar or list of values. Returns: a TF-Feature.

Returns true if the model is built for training mode.

ReportNoisyMax mechanism with noise_eps with 2*noise_eps-DP in our setting where one count can go up by one and another can go down by 1.

L2 distance between tensors true and pred. Args: true: the ground truth image. pred: the predicted image. Returns: mean squared error

Returns the number of classes in the data set.

Pushes a new element.

Sets the i'th element.

Returns the number of elements in the feature vector.

Performs beam search for decoding. Args: sess: tf.Session, session enc_inputs: ndarray of shape (enc_length, 1), the document ids to

Evaluate MNIST for a number of steps. Args: mnist_data_file: Path of a file containing the MNIST images to process. network_parameters: par

Evaluates the locator.

Returns a ragged array of SparseFeatures, for 1) each feature extractor class e, and 2) each feature f extracted by e. Underlying predicate maps will

Feature locators have an additional check that there is no intrinsic type.

Gets the new focus.

Creates a Korean senence and also initializes the token field.