Package ai.djl.translate

Class BasicTranslator<I,O>

java.lang.Object

ai.djl.translate.BasicTranslator<I,O>

Type Parameters:

I - the input class

O - the output class

All Implemented Interfaces:

PostProcessor<O>, PreProcessor<I>, Translator<I,O>

public class BasicTranslator<I,O>
extends java.lang.Object
implements Translator<I,O>

A Translator made by combining a PreProcessor and a PostProcessor.

Constructor Summary

Constructors

Constructor	Description
BasicTranslator(PreProcessor<I> preProcessor, PostProcessor<O> postProcessor)	Constructs a BasicTranslator with the default Batchifier.
BasicTranslator(PreProcessor<I> preProcessor, PostProcessor<O> postProcessor, Batchifier batchifier)	Constructs a BasicTranslator.

Method Summary

Modifier and Type	Method	Description
Batchifier	getBatchifier()	Returns the Batchifier.
void	prepare(TranslatorContext ctx)	Prepares the translator with the manager and model to use.
NDList	processInput(TranslatorContext ctx, I input)	Processes the input and converts it to NDList.
O	processOutput(TranslatorContext ctx, NDList list)	Processes the output NDList to the corresponding output object.

Methods inherited from class java.lang.Object

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

Methods inherited from interface ai.djl.translate.Translator

getExpansions, toBatchTranslator, toBatchTranslator

Constructor Detail

BasicTranslator

public BasicTranslator(PreProcessor<I> preProcessor,
                       PostProcessor<O> postProcessor)

Constructs a BasicTranslator with the default Batchifier.

Parameters:

preProcessor - the preProcessor to use for pre-processing

postProcessor - the postProcessor to use for post-processing

BasicTranslator

public BasicTranslator(PreProcessor<I> preProcessor,
                       PostProcessor<O> postProcessor,
                       Batchifier batchifier)

Constructs a BasicTranslator.

Parameters:

preProcessor - the preProcessor to use for pre-processing

postProcessor - the postProcessor to use for post-processing

batchifier - the batchifier to use

Method Detail

processOutput

public O processOutput(TranslatorContext ctx,
                       NDList list)
                throws java.lang.Exception

Processes the output NDList to the corresponding output object.

Specified by:

processOutput in interface PostProcessor<I>

Parameters:

ctx - the toolkit used for post-processing

list - the output NDList after inference, usually immutable in engines like PyTorch. @see Issue 1774

Returns:

the output object of expected type

Throws:

java.lang.Exception - if an error occurs during processing output

processInput

public NDList processInput(TranslatorContext ctx,
                           I input)
                    throws java.lang.Exception

Processes the input and converts it to NDList.

Specified by:

processInput in interface PreProcessor<I>

Parameters:

ctx - the toolkit for creating the input NDArray

input - the input object

Returns:

the NDList after pre-processing

Throws:

java.lang.Exception - if an error occurs during processing input

getBatchifier

public Batchifier getBatchifier()

Returns the Batchifier.

Specified by:

getBatchifier in interface Translator<I,O>

Returns:

the Batchifier

prepare

public void prepare(TranslatorContext ctx)
             throws java.lang.Exception

Prepares the translator with the manager and model to use.

Specified by:

prepare in interface Translator<I,O>

Parameters:

ctx - the context for the Predictor.

Throws:

java.lang.Exception - if there is an error for preparing the translator