Package com.tnsai.prompt.optimizer

Class SyntheticDataGenerator

java.lang.Object

com.tnsai.prompt.optimizer.SyntheticDataGenerator

public class SyntheticDataGenerator
extends Object

Generates synthetic test data for prompt optimization.

Creates test cases based on task descriptions, including edge cases and variations.

Usage

 SyntheticDataGenerator generator = new SyntheticDataGenerator(llmClient);

 List<TestCase> testCases = generator.generate(
     "Summarize news articles",
     10,
     GenerationSpec.balanced()
 );
 

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	SyntheticDataGenerator.GenerationSpec	Configuration for test case generation.

Constructor Summary

Constructors

Constructor	Description
SyntheticDataGenerator(com.tnsai.llm.LLMClient llmClient)

Method Summary

Modifier and Type	Method	Description
List<TestCase>	generate(String taskDescription, int count)	Generates test cases for a task.
List<TestCase>	generate(String taskDescription, int count, SyntheticDataGenerator.GenerationSpec config)	Generates test cases with custom configuration.
List<TestCase>	generateEdgeCases(String taskDescription, int count)	Generates edge cases specifically.
List<TestCase>	generateGoldenDataset(String taskDescription, int count)	Generates a golden dataset (high-quality reference examples).

Methods inherited from class java.lang.Object

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

Constructor Details

SyntheticDataGenerator

public SyntheticDataGenerator(com.tnsai.llm.LLMClient llmClient)

Method Details

generate

public List<TestCase> generate(String taskDescription,
 int count)

Generates test cases for a task.

Parameters:

taskDescription - description of the task

count - number of test cases to generate

Returns:

list of generated test cases

generate

public List<TestCase> generate(String taskDescription,
 int count,
 SyntheticDataGenerator.GenerationSpec config)

Generates test cases with custom configuration.

Parameters:

taskDescription - description of the task

count - number of test cases

config - generation configuration

Returns:

list of generated test cases

generateEdgeCases

public List<TestCase> generateEdgeCases(String taskDescription,
 int count)

Generates edge cases specifically.

generateGoldenDataset

public List<TestCase> generateGoldenDataset(String taskDescription,
 int count)

Generates a golden dataset (high-quality reference examples).