Package-level declarations
Types
Link copied to clipboard
Strategy for choosing which optimizable modules GEPA updates on each evolution iteration.
Link copied to clipboard
class GEPAOptimizer<Input, Output, InputLabel>(reflectionModel: LLModel, val storagePath: ResilientPath, maxIterations: Int, minibatchSize: Int, componentSelection: ComponentSelection, enableCrossover: Boolean, maxMergeInvocations: Int = 5, randomSeed: Long = 42, labelExtractor: (TrainSetItem<Input, InputLabel>) -> String) : AgentOptimizer<Input, Output, InputLabel>
GEPA (Genetic-Pareto Optimizer for Reflective Prompt Evolution).
Link copied to clipboard
object GEPAPrompts
Prompt templates used by GEPA's reflection LM for instruction proposal and crossover.
Link copied to clipboard
class GEPAReflectionProposer
Proposes improved instructions for GEPA modules by reflecting on execution failures.
Link copied to clipboard
data class ParetoCandidate(val artifact: OptimizationArtifact, val perInstanceScores: Map<Int, Double> = emptyMap(), val parentIndex: Int? = null)
A candidate in the Pareto frontier: an OptimizationArtifact with per-instance validation scores.
Link copied to clipboard
Candidate pool with Pareto-inspired selection for GEPA's evolutionary search.