852600df02
消融默认关,关掉时解释阶段一并跳过;选诊断实验用信息价值减成本。 eval/diagnosis/research.json 为第三块配置。vocab 字段说明的出口文字随提示词一起改,见下一次提交的 prompts/ 与同文件改动(本次含在内)。 Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
31 lines
852 B
TypeScript
31 lines
852 B
TypeScript
// tasks:任务目录里的事实和文件、怎么准备一次运行、怎么从原始产物独立判分(docs/结构.md 第二部分)。
|
|
// 不起子进程:要执行任务代码的事都只交出运行定义,由调用方交给 runner 在任务镜像里跑。
|
|
export { checkTaskFacts, TASK_KINDS, TaskFacts, type TaskKind } from "./facts.ts";
|
|
export {
|
|
diagnosisRun,
|
|
diagnosisUnversionedNames,
|
|
formalRun,
|
|
judge,
|
|
judgeDiagnosis,
|
|
lightCheck,
|
|
loadTask,
|
|
metricDefinitions,
|
|
placeBaselineArtifact,
|
|
prepareDiagnosisJudging,
|
|
prepareDiagnosisRun,
|
|
prepareJudging,
|
|
prepareRun,
|
|
protectedChanges,
|
|
readLightCheck,
|
|
} from "./task.ts";
|
|
export type {
|
|
DiagnosisEnding,
|
|
DiagnosisExperiment,
|
|
DiagnosisExperimentJudged,
|
|
DiagnosisJudged,
|
|
DiagnosisJudgingStep,
|
|
JudgingStep,
|
|
LightCheckOutcome,
|
|
Task,
|
|
} from "./types.ts";
|