Clear gradient clipping parameters.
Clear gradient clipping parameters. In this case, gradient clipping will not be applied. In order to take effect, it needs to be called before fit.
Alternatively, one can pass in the corresponding Keras-Style string representations when calling compile.
Alternatively, one can pass in the corresponding Keras-Style string representations when calling compile.
For example: optimizer = "sgd", loss = "mse", metrics = List("accuracy")
Configure the learning process.
Configure the learning process. It MUST be called before fit or evaluate.
Optimization method to be used.
Criterion to be used.
Validation method(s) to be used. Default is null if no validation is needed.
Evaluate a model in local mode.
Evaluate a model on a given dataset.
Evaluate a model on a given dataset.
Evaluation dataset, RDD of Sample.
Number of samples per batch.
Train a model for a fixed number of epochs on a dataset.
Train a model for a fixed number of epochs on a dataset.
Training dataset, RDD of Sample.
Number of samples per gradient update. Default is 32.
Number of iterations to train. Default is 10.
RDD of Sample, or null if validation is not configured. Default is null.
Train a model for a fixed number of epochs on a dataset.
Train a model for a fixed number of epochs on a dataset.
Training dataset. If x is an instance of LocalDataSet, train in local mode.
Number of iterations to train.
Dataset for validation, or null if validation is not configured.
Freeze the model from the bottom up to the layers specified by names (inclusive).
Freeze the model from the bottom up to the layers specified by names (inclusive).
This is useful for finetune a model
Build graph: some other modules point to current module
Build graph: some other modules point to current module
upstream variables
Variable containing current module
Specify a seq of nodes as output and return a new graph using the existing nodes
Specify a node as output and return a new graph using the existing nodes
Return the node in the graph as specified by the name
Return the nodes in the graph as specified by the names
Use a model to do prediction in local mode.
Use a model to do prediction in local mode.
Prediction data, LocalDataSet.
Use a model to do prediction.
Use a model to do prediction.
Prediction data, RDD of Sample.
Number of samples per batch.
Use a model to predict for classes.
Use a model to predict for classes. By default, label predictions start from 0.
Prediction data, RDD of Sample.
Number of samples per batch. Default is 32.
Boolean. Whether result labels start from 0. Default is true. If false, result labels start from 1.
Save the current model graph to a folder, which can be displayed in TensorBoard by running the command: tensorboard --logdir logPath
Save the current model graph to a folder, which can be displayed in TensorBoard by running the command: tensorboard --logdir logPath
The path to save the model graph.
Whether to draw backward graph instead of forward.
Configure checkpoint settings to write snapshots every epoch during the training process.
Configure checkpoint settings to write snapshots every epoch during the training process. In order to take effect, it needs to be called before fit.
The path to save snapshots. Make sure this path exists beforehand.
Whether to overwrite existing snapshots in the given path. Default is true.
Set constant gradient clipping during the training process.
Set constant gradient clipping during the training process. In order to take effect, it needs to be called before fit.
The minimum value to clip by. Double.
The maximum value to clip by. Double.
Clip gradient to a maximum L2-Norm during the training process.
Clip gradient to a maximum L2-Norm during the training process. In order to take effect, it needs to be called before fit.
Gradient L2-Norm threshold. Double.
Set summary information during the training process for visualization purposes.
Set summary information during the training process for visualization purposes. Saved summary can be viewed via TensorBoard. In order to take effect, it needs to be called before fit.
Training summary will be saved to 'logDir/appName/train' and validation summary (if any) will be saved to 'logDir/appName/validation'.
The base directory path to store training and validation logs.
The name of the application.
Print out the summary information of an Analytics Zoo Keras Model.
Print out the summary information of an Analytics Zoo Keras Model.
For each layer in the model, there will be a separate row containing four columns: Layer (type) Output Shape Param # Connected to
In addition, total number of parameters of this model, separated into trainable and non-trainable counts, will be printed out after the table.
The total length of one row. Default is 120.
The maximum absolute length proportion(%) of each field. Array of Double of length 4. Usually you don't need to adjust this parameter. Default is Array(.33, .55, .67, 1), meaning that the first field will occupy up to 33% of lineLength, the second field will occupy up to (55-33)% of lineLength, the third field will occupy up to (67-55)% of lineLength, the fourth field will occupy the remaining line (100-67)%. If the field has a larger length, the remaining part will be trimmed. If the field has a smaller length, the remaining part will be white spaces.
(Since version 0.3.0) please use recommended saveModule(path, overWrite)