UNPKG

@arizeai/phoenix-evals

Version:
101 lines 3.81 kB
"use strict"; Object.defineProperty(exports, "__esModule", { value: true }); exports.createEvaluator = createEvaluator; const openinference_core_1 = require("@arizeai/openinference-core"); const FunctionEvaluator_1 = require("../core/FunctionEvaluator"); const telemetry_1 = require("../telemetry"); const asEvaluatorFn_1 = require("./asEvaluatorFn"); function generateUniqueName() { return `evaluator-${Math.random().toString(36).substring(2, 15)}`; } /** * A factory function for creating a custom evaluator from any function. * * This function wraps a user-provided function into an evaluator that can be used * with Phoenix experiments and evaluations. The function can be synchronous or * asynchronous, and can return a number, an {@link EvaluationResult} object, or * a value that will be automatically converted to an evaluation result. * * The evaluator will automatically: * - Convert the function's return value to an {@link EvaluationResult} * - Handle both sync and async functions * - Wrap the function with OpenTelemetry spans if telemetry is enabled * - Infer the evaluator name from the function name if not provided * * @typeParam RecordType - The type of the input record that the evaluator expects. * Must extend `Record<string, unknown>`. * @typeParam Fn - The type of the function being wrapped. Must be a function that * accepts the record type and returns a value compatible with {@link EvaluationResult}. * * @param fn - The function to wrap as an evaluator. Can be synchronous or asynchronous. * The function should accept a record of type `RecordType` and return either: * - A number (will be converted to `{ score: number }`) * - An {@link EvaluationResult} object * - Any value that can be converted to an evaluation result * * @param options - Optional configuration for the evaluator. See {@link CreateEvaluatorOptions} * for details on available options. * * @returns An {@link EvaluatorInterface} that can be used with Phoenix experiments * and evaluation workflows. * * @example * Basic usage with a simple scoring function: * ```typescript * const accuracyEvaluator = createEvaluator( * ({ output, expected }) => { * return output === expected ? 1 : 0; * }, * { * name: "accuracy", * kind: "CODE", * optimizationDirection: "MAXIMIZE" * } * ); * * const result = await accuracyEvaluator.evaluate({ * output: "correct answer", * expected: "correct answer" * }); * // result: { score: 1 } * ``` * * * @example * Returning a full EvaluationResult: * ```typescript * const qualityEvaluator = createEvaluator( * ({ output }) => { * const score = calculateQuality(output); * return { * score, * label: score > 0.8 ? "high" : "low", * explanation: `Quality score: ${score}` * }; * }, * { name: "quality" } * ); * ``` */ function createEvaluator(fn, options) { var _a; const { name, kind, optimizationDirection, telemetry = { isEnabled: true }, } = options || {}; const evaluatorName = name || fn.name || generateUniqueName(); let evaluateFn = (0, asEvaluatorFn_1.asEvaluatorFn)(fn); // Add OpenTelemetry span wrapping if telemetry is enabled if (telemetry && telemetry.isEnabled) { evaluateFn = (0, openinference_core_1.withSpan)(evaluateFn, { tracer: (_a = telemetry.tracer) !== null && _a !== void 0 ? _a : telemetry_1.tracer, name: evaluatorName, kind: "EVALUATOR", }); } return new FunctionEvaluator_1.FunctionEvaluator({ evaluateFn, name: evaluatorName, kind: kind || "CODE", optimizationDirection: optimizationDirection || "MAXIMIZE", telemetry, }); } //# sourceMappingURL=createEvaluator.js.map