com.redislabs.redisai

Class Model

java.lang.Object

com.redislabs.redisai.Model

public class Model
extends Object

Direct mapping to RedisAI Model

Field Summary

Fields

Modifier and Type	Field and Description
static String	BLOB_CHUNK_SIZE_PROPERTY

Constructor Summary

Constructors

Constructor and Description
Model(Backend backend, Device device, byte[] blob)
Model(Backend backend, Device device, String[] inputs, String[] outputs, byte[] blob)
Model(Backend backend, Device device, String[] inputs, String[] outputs, byte[] blob, long batchSize, long minBatchSize)
Model(Backend backend, Device device, URI modelUri)

Method Summary

Modifier and Type	Method and Description
static Model	createModelFromRespReply(List<?> reply)
Backend	getBackend()
long	getBatchSize()
byte[]	getBlob()
Device	getDevice()
String[]	getInputs()
long	getMinBatchSize()
long	getMinBatchTimeout()
protected List<byte[]>	getModelSetCommandBytes(String key) Encodes the current model properties into an AI.MODELSET command to be store in RedisAI Server
protected List<byte[]>	getModelStoreCommandArgs(String key) Encodes the current model properties into an AI.MODELSTORE command to store in RedisAI Server.
String[]	getOutputs()
String	getTag()
protected static List<byte[]>	modelExecuteCommandArgs(String key, String[] inputs, String[] outputs, long timeout, boolean includeCommandName)
protected static List<byte[]>	modelRunFlatArgs(String key, String[] inputs, String[] outputs, boolean includeCommandName)
void	setBackend(Backend backend) Deprecated. This variable will be final. Use any constructor.
Model	setBatchSize(long batchsize)
void	setBlob(byte[] blob) Deprecated. This variable will be final. Use any constructor.
void	setDevice(Device device) Deprecated. This variable will be final. Use any constructor.
Model	setInputs(String[] inputs)
Model	setMinBatchSize(long minbatchsize)
Model	setMinBatchTimeout(long minBatchTimeout)
Model	setOutputs(String[] outputs)
Model	setTag(String tag)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

BLOB_CHUNK_SIZE_PROPERTY

public static final String BLOB_CHUNK_SIZE_PROPERTY

See Also:

Constant Field Values

Constructor Detail

Model

public Model(Backend backend,
             Device device,
             URI modelUri)
      throws IOException

Parameters:

backend - - the backend for the model. can be one of TF, TFLITE, TORCH or ONNX

device - - the device that will execute the model. can be of CPU or GPU

modelUri - - filepath of the Protobuf-serialized model

Throws:

IOException

See Also:

Model(com.redislabs.redisai.Backend, com.redislabs.redisai.Device, byte[]), Files.readAllBytes(java.nio.file.Path), Paths.get(java.net.URI)

Model

public Model(Backend backend,
             Device device,
             byte[] blob)

Parameters:

backend - - the backend for the model. can be one of TF, TFLITE, TORCH or ONNX

device - - the device that will execute the model. can be of CPU or GPU

blob - - the Protobuf-serialized model

Model

public Model(Backend backend,
             Device device,
             String[] inputs,
             String[] outputs,
             byte[] blob)

Parameters:

backend - - the backend for the model. can be one of TF, TFLITE, TORCH or ONNX

device - - the device that will execute the model. can be of CPU or GPU

inputs - - one or more names of the model's input nodes (applicable only for TensorFlow models)

outputs - - one or more names of the model's output nodes (applicable only for TensorFlow models)

blob - - the Protobuf-serialized model

Model

public Model(Backend backend,
             Device device,
             String[] inputs,
             String[] outputs,
             byte[] blob,
             long batchSize,
             long minBatchSize)

Parameters:

backend - - the backend for the model. can be one of TF, TFLITE, TORCH or ONNX

device - - the device that will execute the model. can be of CPU or GPU

inputs - - one or more names of the model's input nodes (applicable only for TensorFlow models)

outputs - - one or more names of the model's output nodes (applicable only for TensorFlow models)

blob - - the Protobuf-serialized model

batchSize - - when provided with an batchsize that is greater than 0, the engine will batch incoming requests from multiple clients that use the model with input tensors of the same shape.

minBatchSize - - when provided with an minbatchsize that is greater than 0, the engine will postpone calls to AI.MODELRUN until the batch's size had reached minbatchsize

Method Detail

createModelFromRespReply

public static Model createModelFromRespReply(List<?> reply)

getTag

public String getTag()

setTag

public Model setTag(String tag)

getBlob

public byte[] getBlob()

setBlob

@Deprecated
public void setBlob(byte[] blob)

Deprecated. This variable will be final. Use any constructor.

Parameters:

blob -

getOutputs

public String[] getOutputs()

setOutputs

public Model setOutputs(String[] outputs)

getInputs

public String[] getInputs()

setInputs

public Model setInputs(String[] inputs)

getDevice

public Device getDevice()

setDevice

@Deprecated
public void setDevice(Device device)

Deprecated. This variable will be final. Use any constructor.

Parameters:

device -

getBackend

public Backend getBackend()

setBackend

@Deprecated
public void setBackend(Backend backend)

Deprecated. This variable will be final. Use any constructor.

Parameters:

backend -

getBatchSize

public long getBatchSize()

setBatchSize

public Model setBatchSize(long batchsize)

getMinBatchSize

public long getMinBatchSize()

setMinBatchSize

public Model setMinBatchSize(long minbatchsize)

getMinBatchTimeout

public long getMinBatchTimeout()

setMinBatchTimeout

public Model setMinBatchTimeout(long minBatchTimeout)

getModelSetCommandBytes

protected List<byte[]> getModelSetCommandBytes(String key)

Encodes the current model properties into an AI.MODELSET command to be store in RedisAI Server

Parameters:

key - name of key to store the Model

Returns:

getModelStoreCommandArgs

protected List<byte[]> getModelStoreCommandArgs(String key)

Encodes the current model properties into an AI.MODELSTORE command to store in RedisAI Server.

Parameters:

key -

Returns:

modelRunFlatArgs

protected static List<byte[]> modelRunFlatArgs(String key,
                                               String[] inputs,
                                               String[] outputs,
                                               boolean includeCommandName)

modelExecuteCommandArgs

protected static List<byte[]> modelExecuteCommandArgs(String key,
                                                      String[] inputs,
                                                      String[] outputs,
                                                      long timeout,
                                                      boolean includeCommandName)