CtrlK
BlogDocsLog inGet started
Tessl Logo

tessl/maven-org-deeplearning4j--deeplearning4j-modelimport

Keras model import functionality for DeepLearning4J

—

Pending
Overview
Eval results
Files

model-import.mddocs/

Model Import

Complete model import functionality for bringing Keras models with weights into DeepLearning4J format. This includes both Functional API models and Sequential models.

Functional API Model Import

Import complete Keras Functional API models stored in HDF5 format.

public static ComputationGraph importKerasModelAndWeights(String modelHdf5Filename) 
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

public static ComputationGraph importKerasModelAndWeights(String modelHdf5Filename, boolean enforceTrainingConfig)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

Parameters

  • modelHdf5Filename (String): Path to HDF5 file containing both model configuration and weights
  • enforceTrainingConfig (boolean): Whether to enforce training-related configurations. When true, unsupported configurations throw exceptions. When false, generates warnings but continues.

Returns

  • ComputationGraph: DeepLearning4J computation graph with imported weights

Exceptions

  • IOException: File I/O errors when reading the HDF5 file
  • InvalidKerasConfigurationException: Malformed or invalid Keras model configuration
  • UnsupportedKerasConfigurationException: Keras features not supported by DeepLearning4J

Usage Examples

import org.deeplearning4j.nn.modelimport.keras.KerasModelImport;
import org.deeplearning4j.nn.graph.ComputationGraph;

// Import with default training config enforcement
ComputationGraph model = KerasModelImport.importKerasModelAndWeights("my_keras_model.h5");

// Import with relaxed training config enforcement
ComputationGraph model = KerasModelImport.importKerasModelAndWeights("my_keras_model.h5", false);

// Use the imported model
INDArray input = Nd4j.randn(1, 224, 224, 3); // Example input
INDArray output = model.outputSingle(input);

Sequential Model Import

Import complete Keras Sequential models stored in HDF5 format.

public static MultiLayerNetwork importKerasSequentialModelAndWeights(String modelHdf5Filename)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

public static MultiLayerNetwork importKerasSequentialModelAndWeights(String modelHdf5Filename, boolean enforceTrainingConfig)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

Parameters

  • modelHdf5Filename (String): Path to HDF5 file containing Sequential model configuration and weights
  • enforceTrainingConfig (boolean): Whether to enforce training-related configurations

Returns

  • MultiLayerNetwork: DeepLearning4J multi-layer network with imported weights

Exceptions

  • IOException: File I/O errors when reading the HDF5 file
  • InvalidKerasConfigurationException: Malformed or invalid Keras model configuration
  • UnsupportedKerasConfigurationException: Keras features not supported by DeepLearning4J

Usage Examples

import org.deeplearning4j.nn.modelimport.keras.KerasModelImport;
import org.deeplearning4j.nn.multilayer.MultiLayerNetwork;

// Import Sequential model
MultiLayerNetwork model = KerasModelImport.importKerasSequentialModelAndWeights("sequential_model.h5");

// Use the imported model
INDArray input = Nd4j.randn(1, 784); // Example input for MNIST
INDArray output = model.output(input);

InputStream Import (Unsupported)

Import from InputStreams is declared but currently throws UnsupportedOperationException.

public static ComputationGraph importKerasModelAndWeights(InputStream modelHdf5Stream)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

public static ComputationGraph importKerasModelAndWeights(InputStream modelHdf5Stream, boolean enforceTrainingConfig)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

public static MultiLayerNetwork importKerasSequentialModelAndWeights(InputStream modelHdf5Stream)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

public static MultiLayerNetwork importKerasSequentialModelAndWeights(InputStream modelHdf5Stream, boolean enforceTrainingConfig)
    throws IOException, InvalidKerasConfigurationException, UnsupportedKerasConfigurationException;

These methods currently throw UnsupportedOperationException with the message "Reading HDF5 files from InputStreams currently unsupported."

Supported Model Formats

HDF5 Format Requirements

The HDF5 file must contain:

  • Model configuration as JSON string in the model_config attribute
  • Model weights in the model_weights group
  • Optional training configuration in the training_config attribute

Keras Model Types

  • Functional API Models: Complex architectures with multiple inputs/outputs, branching, and merging
  • Sequential Models: Linear stack of layers

Supported Keras Versions

This library is designed to work with Keras models from the TensorFlow/Keras ecosystem, particularly versions that were current around the 0.9.1 release timeframe.

Error Handling

When importing models, several types of errors can occur:

File Not Found

try {
    ComputationGraph model = KerasModelImport.importKerasModelAndWeights("nonexistent.h5");
} catch (IOException e) {
    System.err.println("Could not read model file: " + e.getMessage());
}

Invalid Configuration

try {
    ComputationGraph model = KerasModelImport.importKerasModelAndWeights("invalid_model.h5");
} catch (InvalidKerasConfigurationException e) {
    System.err.println("Invalid Keras configuration: " + e.getMessage());
}

Unsupported Features

try {
    ComputationGraph model = KerasModelImport.importKerasModelAndWeights("unsupported_model.h5", true);
} catch (UnsupportedKerasConfigurationException e) {
    System.err.println("Unsupported Keras feature: " + e.getMessage());
    // Try with relaxed enforcement
    ComputationGraph model = KerasModelImport.importKerasModelAndWeights("unsupported_model.h5", false);
}

Install with Tessl CLI

npx tessl i tessl/maven-org-deeplearning4j--deeplearning4j-modelimport

docs

builder-pattern.md

configuration-import.md

index.md

layer-support.md

model-import.md

pretrained-models.md

separate-files-import.md

tile.json