LiteSession

View Source On Gitee

import com.mindspore.lite.LiteSession;

LiteSession defines session in MindSpore Lite for compiling Model and forwarding model.

Public Member Functions

function

boolean init(MSConfig config)

void bindThread(boolean if_bind)

boolean compileGraph(Model model)

boolean runGraph()

List<MSTensor> getInputs()

MSTensor getInputsByTensorName(String tensorName)

List<MSTensor> getOutputsByNodeName(String nodeName)

Map<String, MSTensor> getOutputMapByTensor()

List<String> getOutputTensorNames()

MSTensor getOutputByTensorName(String tensorName)

boolean resize(List<MSTensor> inputs, int[][] dims)

void free()

boolean export(String modelFilename, int model_type, int quantization_type)

boolean train()

boolean eval()

boolean isTrain()

boolean isEval()

boolean setLearningRate(float learning_rate)

boolean setupVirtualBatch(int virtualBatchMultiplier, float learningRate, float momentum)

List getFeaturesMap()

boolean updateFeatures(List features)

init

public boolean init(MSConfig config)

Initialize LiteSession.

  • Parameters

    • MSConfig: MSConfig class.

  • Returns

    Whether the initialization is successful.

bindThread

public void bindThread(boolean if_bind)

Attempt to bind or unbind threads in the thread pool to or from the specified cpu core.

  • Parameters

    • if_bind: Define whether to bind or unbind threads.

compileGraph

public boolean compileGraph(Model model)

Compile MindSpore Lite model.

  • Parameters

    • Model: Define the model to be compiled.

  • Returns

    Whether the compilation is successful.

runGraph

public boolean runGraph()

Run the session for inference.

  • Returns

    Whether the inference is successful.

getInputs

public List<MSTensor> getInputs()

Get the MSTensors input of MindSpore Lite model.

  • Returns

    The vector of MindSpore Lite MSTensor.

getInputsByTensorName

public MSTensor getInputsByTensorName(String tensorName)

Get the MSTensors input of MindSpore Lite model by the node name.

  • Parameters

    • tensorName: Define the tensor name.

  • Returns

    MindSpore Lite MSTensor.

getOutputsByNodeName

public List<MSTensor> getOutputsByNodeName(String nodeName)

Get the MSTensors output of MindSpore Lite model by the node name.

  • Parameters

    • nodeName: Define the node name.

  • Returns

    The vector of MindSpore Lite MSTensor.

getOutputMapByTensor

public Map<String, MSTensor> getOutputMapByTensor()

Get the MSTensors output of the MindSpore Lite model associated with the tensor name.

  • Returns

    The map of output tensor name and MindSpore Lite MSTensor.

getOutputTensorNames

public List<String> getOutputTensorNames()

Get the name of output tensors of the model compiled by this session.

  • Returns

    The vector of string as output tensor names in order.

getOutputByTensorName

public MSTensor getOutputByTensorName(String tensorName)

Get the MSTensors output of MindSpore Lite model by the tensor name.

  • Parameters

    • tensorName: Define the tensor name.

  • Returns

    Pointer of MindSpore Lite MSTensor.

resize

public boolean resize(List<MSTensor> inputs, int[][] dims)

Resize inputs shape.

  • Parameters

    • inputs: Model inputs.

    • dims: Define the new inputs shape.

  • Returns

    Whether the resize is successful.

free

public void free()

Free LiteSession.

export

public boolean export(String modelFilename, int model_type, int quantization_type)

Export the model.

  • Parameters

    • modelFilename: Model file name.

    • model_type: Train or Inference type.

    • quantization_type: The quant type.

  • Returns

    Whether the export is successful.

train

public void train()

Switch to the train mode

eval

public void eval()

Switch to the eval mode.

istrain

public void isTrain()

It is Train mode.

iseval

public void isEval()

It is Eval mode.

setLearningRate

public boolean setLearningRate(float learning_rate)

set learning rate.

  • Parameters

    • learning_rate: learning rate.

  • Returns

    Whether the set learning rate is successful.

setupVirtualBatch

public boolean setupVirtualBatch(int virtualBatchMultiplier, float learningRate, float momentum)

Set the virtual batch.

  • Parameters

    • virtualBatchMultiplier: virtual batch multuplier.

    • learningRate: learning rate.

    • momentum: monentum.

  • Returns

    Whether the virtual batch is successfully set.

getFeaturesMap

public List<MSTensor> getFeaturesMap()

Get the FeatureMap.

  • Returns

    FeaturesMap Tensor list.

updateFeatures

public boolean updateFeatures(List<MSTensor> features)

Update model Features.

  • Parameters

    • features: new FeatureMap Tensor List.

  • Returns

    Whether the model features is successfully update.