callback_model_checkpoint
Save the model after every epoch.
MobileNetV2 model architecture
Fashion-MNIST database of fashion articles
IMDB Movie reviews sentiment classification
Instantiates a NASNet model.
callback_terminate_on_naan
Callback that terminates training when a NaN loss is encountered.
Clone a model instance.
export_savedmodel.keras.engine.training.Model
Export a Saved Model
CIFAR100 small image classification
CIFAR10 small image classification
Base R6 class for Keras constraints
Base R6 class for Keras layers
Xception V1 model for Keras.
Fits the state of the preprocessing layer to the data being passed.
Create a custom callback
Activation functions
Stop training when a monitored quantity has stopped improving.
Create a Keras Layer
flow_images_from_directory
Generates batches of data from images in a directory (with optional
augmented/normalized data)
Count the total number of scalars composing the weights.
flow_images_from_dataframe
Takes the dataframe and the path to a directory and generates batches of
augmented/normalized data.
initializer_glorot_normal
Glorot normal initializer, also called Xavier normal initializer.
Retrieve tensors for layers with multiple nodes
Keras backend tensor engine
Retrieves a layer based on either its name (unique) or index.
TensorBoard basic visualizations
Keras implementation
Callback used to stream events to a server.
fit.keras.engine.training.Model
Train a Keras model
Layer/Model configuration
initializer_glorot_uniform
Glorot uniform initializer, also called Xavier uniform initializer.
Weight constraints
compile.keras.engine.training.Model
Configure a Keras model for training
Create a Keras Wrapper
Initializer that generates tensors initialized to a constant value.
Concatenates a list of tensors alongside the specified axis.
Install Keras and the TensorFlow backend
Check if Keras is Available
Element-wise value clipping.
Cumulative sum of the values in a tensor, alongside the specified axis.
Downloads a file from a URL if it not already in the cache.
initializer_variance_scaling
Initializer capable of adapting its scale to the shape of weights.
Boston housing price regression dataset
Initializer that generates tensors initialized to 0.
Initializer that generates a random orthogonal matrix.
Base R6 class for Keras wrappers
MNIST database of handwritten digits
ResNet50 model for Keras.
Depthwise 2D convolution with separable filters.
VGG16 and VGG19 models for Keras.
Flatten a tensor.
initializer_random_normal
Initializer that generates tensors with a normal distribution.
Returns the index of the maximum value along an axis.
Element-wise absolute value.
Update tokenizer internal vocabulary based on a list of texts or list of
sequences.
Reuters newswire topics classification
Generates batches of augmented/normalized data from image data and labels
Returns the index of the minimum value along an axis.
Destroys the current TF graph and creates a new one.
Returns the dtype of a Keras tensor or variable, as a string.
Default float type
k_categorical_crossentropy
Categorical crossentropy between an output tensor and a target tensor.
Get the vocabulary for text vectorization layers
Exponential linear unit.
Fits the model on data yielded batch-by-batch by a generator.
Callback that prints metrics to stdout.
callback_reduce_lr_on_plateau
Reduce learning rate when a metric has stopped improving.
Fit image data generator internal statistics to some sample data.
Returns whether the targets
are in the top k
predictions
.
Selects x
in test phase, and alt
otherwise.
Returns the value of a variable.
Returns whether a tensor is a sparse tensor.
Returns the shape of a variable.
Returns whether x
is a symbolic tensor.
Evaluates the model on a data generator.
Element-wise log.
evaluate.keras.engine.training.Model
Evaluate a Keras model
Layer/Model weights as R arrays
Representation of HDF5 dataset to be used instead of an R array
Generate batches of image data with real-time data augmentation. The data will be
looped over (in batches).
Bitwise reduction (logical AND).
Freeze and unfreeze weights
Retrieve the next item from a generator
He uniform variance scaling initializer.
He normal initializer.
imagenet_decode_predictions
Decodes the prediction of an ImageNet model.
k_normalize_batch_in_training
Computes mean and std for batch then apply batch_normalization on batch.
Multiplies the values in a tensor, alongside the specified axis.
Returns the number of axes in a tensor, as an integer.
Returns a tensor with random binomial distribution of values.
Repeats the elements of a tensor along an axis.
Reset graph identifiers.
imagenet_preprocess_input
Preprocesses a tensor or array encoding a batch of images.
Initializer that generates the identity matrix.
Element-wise sigmoid.
LeCun normal initializer.
Turn a nD tensor into a 2D tensor with same 1st dimension.
Returns the value of more than one tensor variable.
Element-wise sign.
initializer_lecun_uniform
LeCun uniform initializer.
Computes log(sum(exp(elements across dimensions of a tensor))).
Initializer that generates tensors initialized to 1.
3D array representation of images
Loads an image into PIL format.
initializer_random_uniform
Initializer that generates tensors with a uniform distribution.
3D convolution.
Adds a bias vector to a tensor.
Creates a constant tensor.
1D convolution.
Evaluates the value of a variable.
Binary crossentropy between an output tensor and a target tensor.
Element-wise exponential.
Bitwise reduction (logical OR).
Computes sin of x element-wise.
2D Pooling.
Instantiates a Keras function
Softmax of a tensor.
Active Keras backend
Batchwise dot product.
initializer_truncated_normal
Initializer that generates a truncated normal distribution.
Sets the values of many tensor variables at once.
2D deconvolution (i.e. transposed convolution).
Creates a 1D tensor containing a sequence of integers.
2D convolution.
Applies batch normalization on x given mean, var, beta and gamma.
3D deconvolution (i.e. transposed convolution).
Casts a tensor to a different dtype and returns it.
3D Pooling.
Reshapes a tensor to the specified shape.
Runs CTC loss algorithm on each batch element.
Pads the middle dimension of a 3D tensor.
Multiplies 2 tensors (and/or variables) and returns a tensor .
Element-wise truth value of (x >= y).
Reduce elems using fn to combine them from left to right.
Reduce elems using fn to combine them from right to left.
Segment-wise linear approximation of sigmoid.
Sets entries in x
to zero at random, while scaling the entire tensor.
Decodes the output of a softmax.
Element-wise equality between two tensors.
Fuzz factor used in numeric expressions.
Selects x
in train phase, and alt
otherwise.
TF session to be used by the backend.
Cast an array to the default Keras float type.
Cumulative product of the values in a tensor, alongside the specified axis.
k_ctc_label_dense_to_sparse
Converts CTC labels from dense to sparse.
Computes cos of x element-wise.
Adds a 1-sized dimension at index axis
.
Returns a tensor with the same content as the input tensor.
Returns the static number of elements in a Keras variable or tensor.
Get the uid for the default graph.
Default image data format convention ('channels_first' or 'channels_last').
Creates a tensor by tiling x
by n
.
k_manual_variable_initialization
Sets the manual variable initialization flag.
Instantiate an identity matrix and returns it.
Returns whether x
is a placeholder.
Retrieves the elements of indices indices
in the tensor reference
.
Element-wise minimum of two tensors.
Element-wise truth value of (x <= y).
Element-wise truth value of (x < y).
Compute the moving average of a variable.
Element-wise exponentiation.
Returns whether x
is a Keras tensor.
Variance of a tensor, alongside the specified axis.
Returns the shape of tensor or variable as a list of int or NULL entries.
Prints message
and the tensor value when evaluated.
Resizes the images contained in a 4D tensor.
Apply 1D conv with un-shared weights.
Standard deviation of a tensor, alongside the specified axis.
Element-wise square.
Iterates over the time dimension of a tensor
Removes a 1-dimension from the tensor at index axis
.
Stacks a list of rank R
tensors into a rank R+1
tensor.
Element-wise rounding to the closest integer.
Instantiates an all-zeros variable and returns it.
Returns the gradients of variables
w.r.t. loss
.
Element-wise truth value of (x > y).
Apply 2D conv with un-shared weights.
Instantiates an all-zeros variable of the same shape as another tensor.
Reverse a tensor along the specified axes.
Repeats a 2D tensor.
Resizes the volume contained in a 5D tensor.
Rectified linear unit.
Softplus of a tensor.
Apply an activation function to an output.
Creates attention layer
Keras Model composed of a linear stack of layers
Map the function fn over the elements elems and return the outputs.
Layer that averages a list of inputs.
Mean of a tensor, alongside the specified axis.
2D convolution layer (e.g. spatial convolution over images).
Permutes axes in a tensor.
2D convolution with separable filters.
Maximum value in a tensor.
Element-wise maximum of two tensors.
Instantiates a placeholder tensor and returns it.
Sets the learning phase to a fixed value.
Element-wise square root.
Pads 5D tensor with zeros along the depth, height, width dimensions.
Minimum value in a tensor.
Returns a tensor with uniform distribution of values.
Normalizes a tensor wrt the L2 norm alongside the specified axis.
Softsign of a tensor.
Flattens an input
Cropping layer for 1D input (e.g. temporal sequence).
Apply multiplicative 1-centered Gaussian noise.
Rectified Linear Unit activation function
Instantiates a variable and returns it.
layer_global_max_pooling_1d
Global max pooling operation for temporal data.
layer_activation_parametric_relu
Parametric Rectified Linear Unit.
Convolutional LSTM.
k_random_uniform_variable
Instantiates a variable with values drawn from a uniform distribution.
layer_global_max_pooling_2d
Global max pooling operation for spatial data.
Returns variables
but with zero gradient w.r.t. every other variable.
Computes the one-hot representation of an integer tensor.
Instantiates an all-ones variable of the same shape as another tensor.
Instantiates a variable with values drawn from a normal distribution.
Element-wise inequality between two tensors.
Returns the learning phase flag.
Converts a sparse tensor into a dense tensor and returns it.
Transposes a tensor and returns it.
Returns a tensor with normal distribution of values.
Instantiates an all-ones tensor variable and returns it.
Sum of the values in a tensor, alongside the specified axis.
Sets the value of a variable, from an R array.
layer_activation_leaky_relu
Leaky version of a Rectified Linear Unit.
Update the value of x
by adding increment
.
Exponential Linear Unit.
Input layer
Update the value of x
by subtracting decrement
.
Wraps arbitrary expression as a layer
Average pooling for temporal data.
Average pooling operation for spatial data.
layer_locally_connected_1d
Locally-connected layer for 1D inputs.
Max pooling operation for 3D data (spatial or spatio-temporal).
Returns a tensor with truncated random normal distribution of values.
Generates a word rank-based probabilistic sampling table.
Keras array object
Returns the symbolic shape of a tensor or variable.
Depthwise separable 1D convolution.
R interface to Keras
k_sparse_categorical_crossentropy
Categorical crossentropy with integer targets.
Layer that computes the maximum (element-wise) a list of inputs.
Reshapes an output to a certain shape.
Pads the 2nd and 3rd dimensions of a 4D tensor.
Average pooling operation for 3D data (spatial or spatio-temporal).
Switches between two operations depending on a scalar value.
Applies Dropout to the input.
Transposed 2D convolution layer (sometimes called Deconvolution).
layer_global_average_pooling_2d
Global average pooling operation for spatial data.
layer_global_average_pooling_3d
Global Average pooling operation for 3D data.
Turns positive integers (indexes) into dense vectors of fixed size.
Keras Model
Create a Keras custom model
Update the value of x
to new_x
.
layer_locally_connected_2d
Locally-connected layer for 2D inputs.
Layer that adds a list of inputs.
Long Short-Term Memory unit - Hochreiter 1997.
Scaled Exponential Linear Unit.
Spatial 3D version of Dropout.
Layer that subtracts two inputs.
Masks a sequence by using a mask value to skip timesteps.
Zero-padding layer for 1D input (e.g. temporal sequence).
Assign values to names
Model performance metrics
Replicates a model on different GPUs.
Softmax activation function.
layer_activation_thresholded_relu
Thresholded Rectified Linear Unit.
Cropping layer for 2D input (e.g. picture).
Fast GRU implementation backed by CuDNN . Cropping layer for 3D data (e.g. spatial or spatio-temporal).
layer_activity_regularization
Layer that applies an update to the cost function based input activity.
Fast LSTM implementation backed by CuDNN . Text vectorization layer
layer_batch_normalization
Batch normalization layer (Ioffe and Szegedy, 2014).
Generates probability or class probability predictions for the input samples.
Pads sequences to the same length
Returns predictions for a single batch of samples.
Pipe operator
Zero-padding layer for 2D input (e.g. picture).
Max pooling operation for temporal data.
Load a Keras model from the Saved Model format
Model configuration as JSON
Max pooling operation for spatial data.
Permute the dimensions of an input according to a given pattern
Layer that concatenates a list of inputs.
predict.keras.engine.training.Model
Generate predictions from a Keras model
Applies Alpha Dropout to the input.
3D convolution layer (e.g. spatial convolution over volumes).
Upsampling layer for 1D inputs.
Upsampling layer for 2D inputs.
Repeats the input n times.
Upsampling layer for 3D inputs.
1D convolution layer (e.g. temporal convolution).
Normalize a matrix or nd-array
Adadelta optimizer.
Depthwise separable 2D convolution.
Adamax optimizer
Layer that computes a dot product between samples in two tensors.
layer_global_max_pooling_3d
Global Max pooling operation for 3D data.
Transposed 3D convolution layer (sometimes called Deconvolution).
Model configuration as YAML
Export to Saved Model format
Add a densely-connected NN layer to an output
Apply additive zero-centered Gaussian noise.
layer_global_average_pooling_1d
Global average pooling operation for temporal data.
Separable 2D convolution.
Constructs a DenseFeatures.
Remove the last layer in a model
Fully-connected RNN where the output is to be fed back to input.
Zero-padding layer for 3D data (spatial or spatio-temporal).
plot.keras_training_history
Plot training history
Model loss functions
Gated Recurrent Unit - Cho et al.
Save a text tokenizer to an external file
Convert a list of texts to a matrix.
Layer that computes the minimum (element-wise) a list of inputs.
Save model weights in the SavedModel format
Text tokenization utility
Nesterov Adam optimizer
Generates predictions for the input samples from a data generator.
Apply a layer to every temporal slice of an input.
Adagrad optimizer.
Stochastic gradient descent optimizer
RMSProp optimizer
Adam optimizer
Layer that multiplies (element-wise) a list of inputs.
Spatial 1D version of Dropout.
Utility function for generating batches of temporal data.
Spatial 2D version of Dropout.
Save/Load models using SavedModel format
Generates skipgram word pairs.
Save/Load model weights using HDF5 files
Single gradient update or model evaluation over one batch of samples.
Sets vocabulary (and optionally document frequency) data for the layer
Converts a class vector (integers) to binary class matrix.
summary.keras.engine.training.Model
Print a summary of a Keras model
Convert a list of sequences into a matrix.
L1 and L2 regularization
Provide a scope with mappings of names to custom objects
Select a Keras implementation and backend
Objects exported from other packages
Serialize a model to an R object
Reset the states for a layer
Converts a text to a sequence of indexes in a fixed-size hashing space.
Save/Load models using HDF5 files
Transform each text in texts in a sequence of integers.
texts_to_sequences_generator
Transforms each text in texts in a sequence of integers.
One-hot encode a text into a list of word indexes in a vocabulary of size n.
Convert text to a sequence of words (or tokens).
Inception V3 model, with weights pre-trained on ImageNet.
MobileNet model architecture.
Base R6 class for Keras callbacks
Instantiates the DenseNet architecture.
application_inception_resnet_v2
Inception-ResNet v2 model, with weights trained on ImageNet
callback_learning_rate_scheduler
Learning rate scheduler.
Bidirectional wrapper for RNNs.
Callback that streams epoch results to a csv file
Element-wise tanh.