Add declarative harness prototype
This commit is contained in:
@@ -46,6 +46,28 @@ priority 中,仍然可能只是“换皮的 rule-based harness”。因此,
|
||||
|
||||
下面的设计已经把这些 major revisions 纳入硬性要求。
|
||||
|
||||
## 2026-06-26 adversarial status
|
||||
|
||||
我们已经用 `TP=8, gmu=0.5, max-num-seqs=8` 的 bad-start case 攻击当前 production
|
||||
harness。结果显示当前 stop guard 会在 baseline 后触发
|
||||
`search_high_saturated_by_incumbent`,没有生成 topology/resource-efficiency contrast。
|
||||
这证明当前 implementation 还不是最终 contribution。
|
||||
|
||||
详细反例见
|
||||
[Bad-start stop counterexample](bad-start-stop-counterexample-20260626.md)。
|
||||
|
||||
该反例给本设计增加一个硬约束:
|
||||
|
||||
```text
|
||||
search_high_saturated_by_incumbent may be measurement evidence,
|
||||
but it cannot bypass candidate-set coverage when topology/resource efficiency
|
||||
remains tunable.
|
||||
```
|
||||
|
||||
因此新的 CoverageValidator 必须先证明没有未覆盖的 high-priority candidate,才能授权
|
||||
stop。对 `req/s/GPU` objective,未覆盖的 topology/resource-efficiency contrast 必须阻止
|
||||
stop,除非 StudySpec 明确固定 topology/GPU budget。
|
||||
|
||||
## 当前问题
|
||||
|
||||
当前 `src/aituner/harness.py` 已经具备了一些正确的抽象词汇:observation、
|
||||
|
||||
Reference in New Issue
Block a user