AxLearnOptions
Defined in: https://github.com/ax-llm/ax/blob/242cf18d21de9f9d58c7c82f53305f0605497473/src/ax/dsp/learn.ts#L35
Configuration for the AxLearn agent. Combines agent configuration (name, storage) with learning configuration (teacher, budget).
Properties
| Property | Type | Description |
|---|---|---|
budget? | number | Maximum optimization rounds (default: 20) |
criteria? | string | Custom evaluation criteria for judge |
enableTracing? | boolean | Whether to log traces (default: true) |
examples? | AxTypedExample<AxGenIn>[] | Training examples (manual) |
generateExamples? | boolean | Whether to generate synthetic examples (default: true if no other data) |
judgeOptions? | Partial<AxJudgeOptions> | Judge options when auto-generating metric |
metadata? | Record<string, unknown> | Custom metadata for all traces |
metric? | AxMetricFn | Custom metric function (if not provided, auto-generates using AxJudge) |
name | string | Unique identifier/name for this agent |
onProgress? | (progress: AxLearnProgress) => void | Progress callback |
onTrace? | (trace: AxTrace) => void | Callback when a trace is logged |
storage | AxStorage | Storage backend (Required) |
synthCount? | number | Number of synthetic examples to generate |
synthOptions? | Partial<AxSynthOptions> | Synth options for data generation |
teacher | AxAIService | Teacher AI for synthetic data generation and judging (Required) |
useTraces? | boolean | Whether to use capture traces as training examples (default: true) |
validationSplit? | number | Validation split ratio (default: 0.2) |