@miller-tech/uap 1.179.11 → 1.180.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/.tsbuildinfo +1 -1
- package/dist/benchmarks/paired/report.d.ts +86 -0
- package/dist/benchmarks/paired/report.d.ts.map +1 -1
- package/dist/benchmarks/paired/report.js +138 -0
- package/dist/benchmarks/paired/report.js.map +1 -1
- package/dist/benchmarks/paired/runner.d.ts.map +1 -1
- package/dist/benchmarks/paired/runner.js +6 -1
- package/dist/benchmarks/paired/runner.js.map +1 -1
- package/dist/bin/cli.js +2 -0
- package/dist/bin/cli.js.map +1 -1
- package/dist/cli/bench.d.ts +4 -0
- package/dist/cli/bench.d.ts.map +1 -1
- package/dist/cli/bench.js +91 -2
- package/dist/cli/bench.js.map +1 -1
- package/dist/delivery/agentic-executor.d.ts +10 -0
- package/dist/delivery/agentic-executor.d.ts.map +1 -1
- package/dist/delivery/agentic-executor.js +83 -1
- package/dist/delivery/agentic-executor.js.map +1 -1
- package/dist/self-harness/decide.d.ts +5 -0
- package/dist/self-harness/decide.d.ts.map +1 -1
- package/dist/self-harness/decide.js +19 -0
- package/dist/self-harness/decide.js.map +1 -1
- package/package.json +1 -1
- package/src/policies/enforcers/__pycache__/_common.cpython-312.pyc +0 -0
- package/templates/hooks/__pycache__/deliver_autoroute.cpython-312.pyc +0 -0
- package/templates/hooks/deliver_autoroute.py +12 -1
- package/tools/agents/scripts/__pycache__/toolcall_path_normalizer.cpython-312.pyc +0 -0
- package/tools/agents/tests/test_deliver_autoroute.py +20 -1
|
@@ -67,6 +67,13 @@ export interface AnalysisReport {
|
|
|
67
67
|
successRate: number;
|
|
68
68
|
meanTokens: number | null;
|
|
69
69
|
}[];
|
|
70
|
+
/**
|
|
71
|
+
* Whether this run could have detected a difference at all (see
|
|
72
|
+
* `assessDiscrimination`). A suite where every arm scores 100% — or 0% —
|
|
73
|
+
* yields `delta=+0.000, significant=false`, which reads like "no effect" but
|
|
74
|
+
* actually means "no measurement".
|
|
75
|
+
*/
|
|
76
|
+
discrimination: DiscriminationVerdict;
|
|
70
77
|
/**
|
|
71
78
|
* ETCSOVG disclosure card for the harness these numbers were produced by
|
|
72
79
|
* (harness plan F). Harness variance dominates model variance 7.8x
|
|
@@ -76,6 +83,83 @@ export interface AnalysisReport {
|
|
|
76
83
|
*/
|
|
77
84
|
harnessCard?: HarnessCard;
|
|
78
85
|
}
|
|
86
|
+
/**
|
|
87
|
+
* Can this run tell the arms apart at all?
|
|
88
|
+
*
|
|
89
|
+
* A paired benchmark reports `delta=+0.000, significant=false` in two very
|
|
90
|
+
* different situations: the arms genuinely perform the same, or the SUITE
|
|
91
|
+
* cannot separate them because every arm sits at the ceiling (or the floor).
|
|
92
|
+
* The numbers are identical; the conclusions are opposite. Reporting only the
|
|
93
|
+
* delta invites reading "no measurement" as "no effect" — three consecutive
|
|
94
|
+
* runs against `real-gate` (100%/100%) and `smoke` (0%/0%, wrong adapter) were
|
|
95
|
+
* each read as a null result before anyone noticed nothing had been measured.
|
|
96
|
+
*
|
|
97
|
+
* So the verdict is computed and printed alongside the delta, always.
|
|
98
|
+
*/
|
|
99
|
+
export type DiscriminationStatus = 'ok'
|
|
100
|
+
/** Every arm solved (nearly) everything — the suite cannot separate them. */
|
|
101
|
+
| 'ceiling'
|
|
102
|
+
/** Nothing was solved — suite too hard, or adapter/harness misconfigured. */
|
|
103
|
+
| 'floor'
|
|
104
|
+
/**
|
|
105
|
+
* Arms differed nowhere on CORRECTNESS: zero discordant pairs. This is the
|
|
106
|
+
* criterion McNemar tests on, and it is the honest name for the failure — a
|
|
107
|
+
* bootstrap over an all-zero delta vector returns CI [0.000, 0.000], which
|
|
108
|
+
* reads as "conclusively no effect" when it means "no information".
|
|
109
|
+
*/
|
|
110
|
+
| 'no-discordant-pairs'
|
|
111
|
+
/** The interval spans both zero and the effect worth having. */
|
|
112
|
+
| 'underpowered'
|
|
113
|
+
/** Nothing to compare — a report with no treatment arm. */
|
|
114
|
+
| 'no-comparisons';
|
|
115
|
+
export interface DiscriminationVerdict {
|
|
116
|
+
status: DiscriminationStatus;
|
|
117
|
+
/**
|
|
118
|
+
* True when a CORRECTNESS difference could have shown up. Efficiency results
|
|
119
|
+
* are judged separately — see `efficiencyUsable`.
|
|
120
|
+
*/
|
|
121
|
+
usable: boolean;
|
|
122
|
+
/**
|
|
123
|
+
* True when at least one continuous metric (tokens/turns/cost/latency) or the
|
|
124
|
+
* quality score carries a non-degenerate interval.
|
|
125
|
+
*
|
|
126
|
+
* This exists because "same accuracy, 40% fewer tokens" is a real, citable
|
|
127
|
+
* result — and the first cut of this check refused it, printing "not evidence
|
|
128
|
+
* of anything" directly above a token table reading 🟢 WIN. A saturated suite
|
|
129
|
+
* measures correctness badly and efficiency perfectly well.
|
|
130
|
+
*/
|
|
131
|
+
efficiencyUsable: boolean;
|
|
132
|
+
reason: string;
|
|
133
|
+
minSuccess: number;
|
|
134
|
+
maxSuccess: number;
|
|
135
|
+
/** Paired cells in the PRIMARY comparison. */
|
|
136
|
+
pairedCells: number;
|
|
137
|
+
/** Discordant correctness pairs in the PRIMARY comparison (McNemar b+c). */
|
|
138
|
+
discordantPairs: number;
|
|
139
|
+
}
|
|
140
|
+
export interface DiscriminationOptions {
|
|
141
|
+
/** Every arm at or above this is a ceiling. Default 0.98. */
|
|
142
|
+
ceilingAt?: number;
|
|
143
|
+
/** Every arm at or below this is a floor. Default 0.02. */
|
|
144
|
+
floorAt?: number;
|
|
145
|
+
/** Fewer paired cells than this cannot support a verdict. Default 8. */
|
|
146
|
+
minPairedCells?: number;
|
|
147
|
+
/**
|
|
148
|
+
* Smallest correctness delta worth detecting, in success-rate units. A run is
|
|
149
|
+
* inconclusive when its interval contains BOTH zero and this effect.
|
|
150
|
+
* Default 0.25.
|
|
151
|
+
*/
|
|
152
|
+
minDetectableEffect?: number;
|
|
153
|
+
}
|
|
154
|
+
/**
|
|
155
|
+
* Can this run tell the arms apart?
|
|
156
|
+
*
|
|
157
|
+
* Judged against the PRIMARY comparison (the first treatment arm), not folded
|
|
158
|
+
* across every arm: an `--ablation` run has 6 comparisons, and summing their
|
|
159
|
+
* discordance let one noisy leave-one-out arm mask a null primary, while taking
|
|
160
|
+
* the widest CI let it refuse a tight primary.
|
|
161
|
+
*/
|
|
162
|
+
export declare function assessDiscrimination(perCondition: ConditionSummary[], comparisons: Comparison[], opts?: DiscriminationOptions): DiscriminationVerdict;
|
|
79
163
|
export interface AnalyzeOptions extends PairedOptions {
|
|
80
164
|
baselineLabel?: string;
|
|
81
165
|
/** ROPE half-width for correctness (success-rate units, 0..1). A success-rate
|
|
@@ -94,6 +178,8 @@ export interface AnalyzeOptions extends PairedOptions {
|
|
|
94
178
|
* a guessed one.
|
|
95
179
|
*/
|
|
96
180
|
harness?: HarnessCardInput;
|
|
181
|
+
/** Thresholds for the ceiling/floor/underpowered verdict. */
|
|
182
|
+
discrimination?: DiscriminationOptions;
|
|
97
183
|
}
|
|
98
184
|
export declare function analyze(output: RunnerOutput, opts?: AnalyzeOptions): AnalysisReport;
|
|
99
185
|
export declare function renderMarkdown(r: AnalysisReport): string;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"report.d.ts","sourceRoot":"","sources":["../../../src/benchmarks/paired/report.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,EAGL,aAAa,EAGb,iBAAiB,EACjB,aAAa,EAEb,OAAO,EACR,MAAM,YAAY,CAAC;AACpB,OAAO,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAC3C,OAAO,EAAsB,gBAAgB,EAA2B,MAAM,YAAY,CAAC;AAC3F,OAAO,EAGL,KAAK,WAAW,EAChB,KAAK,gBAAgB,EACtB,MAAM,oBAAoB,CAAC;AAE5B,MAAM,WAAW,gBAAgB;IAC/B,KAAK,EAAE,MAAM,CAAC;IACd,CAAC,EAAE,MAAM,CAAC;IACV,WAAW,EAAE,MAAM,CAAC;IACpB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,aAAa,EAAE,MAAM,CAAC;IACtB,SAAS,EAAE,MAAM,CAAC;IAClB,sFAAsF;IACtF,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;CAC5B;AAED,MAAM,WAAW,UAAU;IACzB,KAAK,EAAE,MAAM,CAAC;IACd,QAAQ,EAAE,MAAM,CAAC;IACjB,2EAA2E;IAC3E,WAAW,EAAE;QACX,YAAY,EAAE,MAAM,CAAC;QACrB,aAAa,EAAE,MAAM,CAAC;QACtB,KAAK,EAAE,iBAAiB,CAAC;QACzB,OAAO,EAAE,aAAa,CAAC;QACvB,yEAAyE;QACzE,OAAO,EAAE,OAAO,CAAC;KAClB,CAAC;IACF,iFAAiF;IACjF,OAAO,EAAE,OAAO,CAAC,MAAM,CAAC,gBAAgB,EAAE,iBAAiB,CAAC,CAAC,CAAC;IAC9D,+EAA+E;IAC/E,cAAc,EAAE,OAAO,CAAC,MAAM,CAAC,gBAAgB,EAAE,OAAO,CAAC,CAAC,CAAC;IAC3D;;;;OAIG;IACH,OAAO,CAAC,EAAE,iBAAiB,CAAC;IAC5B,4DAA4D;IAC5D,cAAc,CAAC,EAAE,OAAO,CAAC;CAC1B;AAED,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE;QACJ,KAAK,EAAE,MAAM,CAAC;QACd,OAAO,EAAE,MAAM,CAAC;QAChB,MAAM,EAAE,MAAM,CAAC;QACf,SAAS,EAAE,MAAM,CAAC;QAClB,UAAU,EAAE,MAAM,EAAE,CAAC;QACrB,QAAQ,EAAE,MAAM,CAAC;QACjB,SAAS,EAAE,MAAM,CAAC;QAClB,UAAU,EAAE,MAAM,CAAC;KACpB,CAAC;IACF,YAAY,EAAE,gBAAgB,EAAE,CAAC;IACjC,WAAW,EAAE,UAAU,EAAE,CAAC;IAC1B,uDAAuD;IACvD,MAAM,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,GAAG,IAAI,CAAA;KAAE,EAAE,CAAC;IAC5E;;;;;;OAMG;IACH,WAAW,CAAC,EAAE,WAAW,CAAC;CAC3B;AA6DD,MAAM,WAAW,cAAe,SAAQ,aAAa;IACnD,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB;;2EAEuE;IACvE,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB;gEAC4D;IAC5D,aAAa,CAAC,EAAE,OAAO,CAAC,MAAM,CAAC,gBAAgB,EAAE,MAAM,CAAC,CAAC,CAAC;IAC1D;uEACmE;IACnE,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB;;;;OAIG;IACH,OAAO,CAAC,EAAE,gBAAgB,CAAC;
|
|
1
|
+
{"version":3,"file":"report.d.ts","sourceRoot":"","sources":["../../../src/benchmarks/paired/report.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,EAGL,aAAa,EAGb,iBAAiB,EACjB,aAAa,EAEb,OAAO,EACR,MAAM,YAAY,CAAC;AACpB,OAAO,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAC3C,OAAO,EAAsB,gBAAgB,EAA2B,MAAM,YAAY,CAAC;AAC3F,OAAO,EAGL,KAAK,WAAW,EAChB,KAAK,gBAAgB,EACtB,MAAM,oBAAoB,CAAC;AAE5B,MAAM,WAAW,gBAAgB;IAC/B,KAAK,EAAE,MAAM,CAAC;IACd,CAAC,EAAE,MAAM,CAAC;IACV,WAAW,EAAE,MAAM,CAAC;IACpB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,aAAa,EAAE,MAAM,CAAC;IACtB,SAAS,EAAE,MAAM,CAAC;IAClB,sFAAsF;IACtF,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;CAC5B;AAED,MAAM,WAAW,UAAU;IACzB,KAAK,EAAE,MAAM,CAAC;IACd,QAAQ,EAAE,MAAM,CAAC;IACjB,2EAA2E;IAC3E,WAAW,EAAE;QACX,YAAY,EAAE,MAAM,CAAC;QACrB,aAAa,EAAE,MAAM,CAAC;QACtB,KAAK,EAAE,iBAAiB,CAAC;QACzB,OAAO,EAAE,aAAa,CAAC;QACvB,yEAAyE;QACzE,OAAO,EAAE,OAAO,CAAC;KAClB,CAAC;IACF,iFAAiF;IACjF,OAAO,EAAE,OAAO,CAAC,MAAM,CAAC,gBAAgB,EAAE,iBAAiB,CAAC,CAAC,CAAC;IAC9D,+EAA+E;IAC/E,cAAc,EAAE,OAAO,CAAC,MAAM,CAAC,gBAAgB,EAAE,OAAO,CAAC,CAAC,CAAC;IAC3D;;;;OAIG;IACH,OAAO,CAAC,EAAE,iBAAiB,CAAC;IAC5B,4DAA4D;IAC5D,cAAc,CAAC,EAAE,OAAO,CAAC;CAC1B;AAED,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE;QACJ,KAAK,EAAE,MAAM,CAAC;QACd,OAAO,EAAE,MAAM,CAAC;QAChB,MAAM,EAAE,MAAM,CAAC;QACf,SAAS,EAAE,MAAM,CAAC;QAClB,UAAU,EAAE,MAAM,EAAE,CAAC;QACrB,QAAQ,EAAE,MAAM,CAAC;QACjB,SAAS,EAAE,MAAM,CAAC;QAClB,UAAU,EAAE,MAAM,CAAC;KACpB,CAAC;IACF,YAAY,EAAE,gBAAgB,EAAE,CAAC;IACjC,WAAW,EAAE,UAAU,EAAE,CAAC;IAC1B,uDAAuD;IACvD,MAAM,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,GAAG,IAAI,CAAA;KAAE,EAAE,CAAC;IAC5E;;;;;OAKG;IACH,cAAc,EAAE,qBAAqB,CAAC;IACtC;;;;;;OAMG;IACH,WAAW,CAAC,EAAE,WAAW,CAAC;CAC3B;AAED;;;;;;;;;;;;GAYG;AACH,MAAM,MAAM,oBAAoB,GAC5B,IAAI;AACN,6EAA6E;GAC3E,SAAS;AACX,6EAA6E;GAC3E,OAAO;AACT;;;;;GAKG;GACD,qBAAqB;AACvB,gEAAgE;GAC9D,cAAc;AAChB,2DAA2D;GACzD,gBAAgB,CAAC;AAErB,MAAM,WAAW,qBAAqB;IACpC,MAAM,EAAE,oBAAoB,CAAC;IAC7B;;;OAGG;IACH,MAAM,EAAE,OAAO,CAAC;IAChB;;;;;;;;OAQG;IACH,gBAAgB,EAAE,OAAO,CAAC;IAC1B,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,8CAA8C;IAC9C,WAAW,EAAE,MAAM,CAAC;IACpB,4EAA4E;IAC5E,eAAe,EAAE,MAAM,CAAC;CACzB;AAED,MAAM,WAAW,qBAAqB;IACpC,6DAA6D;IAC7D,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,2DAA2D;IAC3D,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,wEAAwE;IACxE,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB;;;;OAIG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;CAC9B;AAyBD;;;;;;;GAOG;AACH,wBAAgB,oBAAoB,CAClC,YAAY,EAAE,gBAAgB,EAAE,EAChC,WAAW,EAAE,UAAU,EAAE,EACzB,IAAI,GAAE,qBAA0B,GAC/B,qBAAqB,CAmGvB;AA6DD,MAAM,WAAW,cAAe,SAAQ,aAAa;IACnD,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB;;2EAEuE;IACvE,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB;gEAC4D;IAC5D,aAAa,CAAC,EAAE,OAAO,CAAC,MAAM,CAAC,gBAAgB,EAAE,MAAM,CAAC,CAAC,CAAC;IAC1D;uEACmE;IACnE,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB;;;;OAIG;IACH,OAAO,CAAC,EAAE,gBAAgB,CAAC;IAC3B,6DAA6D;IAC7D,cAAc,CAAC,EAAE,qBAAqB,CAAC;CACxC;AAED,wBAAgB,OAAO,CAAC,MAAM,EAAE,YAAY,EAAE,IAAI,GAAE,cAAmB,GAAG,cAAc,CA+GvF;AA4BD,wBAAgB,cAAc,CAAC,CAAC,EAAE,cAAc,GAAG,MAAM,CA+GxD"}
|
|
@@ -10,6 +10,127 @@
|
|
|
10
10
|
import { mcnemar, mean, pairedDelta, verdict, } from './stats.js';
|
|
11
11
|
import { CONTINUOUS_METRICS } from './types.js';
|
|
12
12
|
import { buildHarnessCard, renderHarnessCard, } from '../harness-card.js';
|
|
13
|
+
function finite(n) {
|
|
14
|
+
return typeof n === 'number' && Number.isFinite(n);
|
|
15
|
+
}
|
|
16
|
+
/**
|
|
17
|
+
* Does any continuous metric or quality score carry a real difference?
|
|
18
|
+
*
|
|
19
|
+
* Keyed on a NON-ZERO paired mean, not on interval width. Zero width means
|
|
20
|
+
* opposite things for the two kinds of measure: an all-zero correctness delta is
|
|
21
|
+
* no information, while a constant -1200-token delta across every cell is the
|
|
22
|
+
* strongest efficiency result there is — perfectly consistent, zero variance.
|
|
23
|
+
* Testing width alone refused exactly that.
|
|
24
|
+
*/
|
|
25
|
+
function hasEfficiencySignal(c) {
|
|
26
|
+
if (!c)
|
|
27
|
+
return false;
|
|
28
|
+
const carries = (pd) => Boolean(pd && pd.n > 0 && finite(pd.meanDelta) && pd.meanDelta !== 0);
|
|
29
|
+
for (const m of CONTINUOUS_METRICS) {
|
|
30
|
+
if (carries(c.metrics[m]))
|
|
31
|
+
return true;
|
|
32
|
+
}
|
|
33
|
+
return carries(c.quality);
|
|
34
|
+
}
|
|
35
|
+
/**
|
|
36
|
+
* Can this run tell the arms apart?
|
|
37
|
+
*
|
|
38
|
+
* Judged against the PRIMARY comparison (the first treatment arm), not folded
|
|
39
|
+
* across every arm: an `--ablation` run has 6 comparisons, and summing their
|
|
40
|
+
* discordance let one noisy leave-one-out arm mask a null primary, while taking
|
|
41
|
+
* the widest CI let it refuse a tight primary.
|
|
42
|
+
*/
|
|
43
|
+
export function assessDiscrimination(perCondition, comparisons, opts = {}) {
|
|
44
|
+
const ceilingAt = opts.ceilingAt ?? 0.98;
|
|
45
|
+
const floorAt = opts.floorAt ?? 0.02;
|
|
46
|
+
const minCells = opts.minPairedCells ?? 8;
|
|
47
|
+
const mde = opts.minDetectableEffect ?? 0.25;
|
|
48
|
+
const rates = perCondition.map((c) => c.successRate).filter(finite);
|
|
49
|
+
const minSuccess = rates.length ? Math.min(...rates) : 0;
|
|
50
|
+
const maxSuccess = rates.length ? Math.max(...rates) : 0;
|
|
51
|
+
const primary = comparisons[0];
|
|
52
|
+
const efficiencyUsable = hasEfficiencySignal(primary);
|
|
53
|
+
const d = primary?.correctness.delta;
|
|
54
|
+
const pairedCells = finite(d?.n) ? d.n : 0;
|
|
55
|
+
const discordantPairs = primary
|
|
56
|
+
? primary.correctness.mcnemar.onlyTreatment + primary.correctness.mcnemar.onlyBaseline
|
|
57
|
+
: 0;
|
|
58
|
+
const base = { minSuccess, maxSuccess, pairedCells, discordantPairs, efficiencyUsable };
|
|
59
|
+
const alsoEfficiency = efficiencyUsable
|
|
60
|
+
? ' Efficiency deltas (tokens/turns/latency) below ARE valid and may be cited.'
|
|
61
|
+
: '';
|
|
62
|
+
if (!primary) {
|
|
63
|
+
return {
|
|
64
|
+
status: 'no-comparisons',
|
|
65
|
+
usable: false,
|
|
66
|
+
reason: 'no treatment arm to compare against the baseline — nothing was measured.',
|
|
67
|
+
...base,
|
|
68
|
+
};
|
|
69
|
+
}
|
|
70
|
+
if (rates.length > 0 && minSuccess >= ceilingAt) {
|
|
71
|
+
return {
|
|
72
|
+
status: 'ceiling',
|
|
73
|
+
usable: false,
|
|
74
|
+
reason: `every condition solved ${(minSuccess * 100).toFixed(0)}%+ of cells — the suite cannot separate ` +
|
|
75
|
+
`them on correctness. Either it is too easy, or the verify command always passes (a mock adapter ` +
|
|
76
|
+
`against a mock suite looks identical). A zero correctness delta here means NO MEASUREMENT.` +
|
|
77
|
+
alsoEfficiency,
|
|
78
|
+
...base,
|
|
79
|
+
};
|
|
80
|
+
}
|
|
81
|
+
if (rates.length > 0 && maxSuccess <= floorAt) {
|
|
82
|
+
return {
|
|
83
|
+
status: 'floor',
|
|
84
|
+
usable: false,
|
|
85
|
+
reason: `no condition solved anything (max ${(maxSuccess * 100).toFixed(0)}%) — the suite is too hard, ` +
|
|
86
|
+
`or the adapter/suite are mismatched (a mock-only suite scores 0% against every real model, ` +
|
|
87
|
+
`which looks identical to "too hard"). A zero delta here means NO MEASUREMENT.` + alsoEfficiency,
|
|
88
|
+
...base,
|
|
89
|
+
};
|
|
90
|
+
}
|
|
91
|
+
if (discordantPairs === 0) {
|
|
92
|
+
return {
|
|
93
|
+
status: 'no-discordant-pairs',
|
|
94
|
+
usable: false,
|
|
95
|
+
reason: `the arms produced identical correctness on all ${pairedCells} paired cells (0 discordant), so ` +
|
|
96
|
+
`the run carries no information about a correctness difference — the CI collapses to [0, 0], ` +
|
|
97
|
+
`which LOOKS conclusive and is not.` + alsoEfficiency,
|
|
98
|
+
...base,
|
|
99
|
+
};
|
|
100
|
+
}
|
|
101
|
+
if (pairedCells > 0 && pairedCells < minCells) {
|
|
102
|
+
return {
|
|
103
|
+
status: 'underpowered',
|
|
104
|
+
usable: false,
|
|
105
|
+
reason: `only ${pairedCells} paired cells — too few to distinguish a real effect from noise. ` +
|
|
106
|
+
`Raise --epochs or add tasks before reading the delta.` + alsoEfficiency,
|
|
107
|
+
...base,
|
|
108
|
+
};
|
|
109
|
+
}
|
|
110
|
+
// Inconclusive = the interval admits both "no effect" and an effect worth
|
|
111
|
+
// having. Half-width alone got this wrong in both directions: it passed
|
|
112
|
+
// [-0.24, +0.24] (cannot tell a 24pp regression from a 24pp win) and refused
|
|
113
|
+
// [0.02, 0.55] (a proven win with a long tail).
|
|
114
|
+
const ci = d?.ci;
|
|
115
|
+
const significant = Boolean(d?.significant);
|
|
116
|
+
if (!significant && ci && finite(ci.lower) && finite(ci.upper) && ci.lower < mde && ci.upper > -mde) {
|
|
117
|
+
return {
|
|
118
|
+
status: 'underpowered',
|
|
119
|
+
usable: false,
|
|
120
|
+
reason: `the correctness CI [${ci.lower.toFixed(2)}, ${ci.upper.toFixed(2)}] contains both zero and the ` +
|
|
121
|
+
`±${mde} effect worth detecting — this run could not have told a real difference from none. ` +
|
|
122
|
+
`Raise --epochs or add tasks.` + alsoEfficiency,
|
|
123
|
+
...base,
|
|
124
|
+
};
|
|
125
|
+
}
|
|
126
|
+
return {
|
|
127
|
+
status: 'ok',
|
|
128
|
+
usable: true,
|
|
129
|
+
reason: `conditions span ${(minSuccess * 100).toFixed(0)}%-${(maxSuccess * 100).toFixed(0)}% over ` +
|
|
130
|
+
`${pairedCells} paired cells with ${discordantPairs} discordant — a difference could show up here.`,
|
|
131
|
+
...base,
|
|
132
|
+
};
|
|
133
|
+
}
|
|
13
134
|
function cellKey(r) {
|
|
14
135
|
return `${r.taskId}#${r.seed}`;
|
|
15
136
|
}
|
|
@@ -155,6 +276,7 @@ export function analyze(output, opts = {}) {
|
|
|
155
276
|
successRate: c.successRate,
|
|
156
277
|
meanTokens: c.meanTokens,
|
|
157
278
|
})),
|
|
279
|
+
discrimination: assessDiscrimination(perCondition, comparisons, opts.discrimination),
|
|
158
280
|
// Default the card's model to the one the run actually used, so the caller
|
|
159
281
|
// cannot accidentally disclose a different model than it benchmarked.
|
|
160
282
|
...(opts.harness
|
|
@@ -193,6 +315,22 @@ export function renderMarkdown(r) {
|
|
|
193
315
|
`**Baseline:** \`${r.meta.baseline}\``);
|
|
194
316
|
L.push(`*${r.meta.startedAt} → ${r.meta.finishedAt}*`);
|
|
195
317
|
L.push('');
|
|
318
|
+
// Ahead of every number, because a reader who quotes a delta from a run that
|
|
319
|
+
// could not measure anything has been misled by the report, not by the data.
|
|
320
|
+
const d = r.discrimination;
|
|
321
|
+
if (!d.usable) {
|
|
322
|
+
const what = d.efficiencyUsable ? 'NO CORRECTNESS SIGNAL' : 'NO USABLE SIGNAL';
|
|
323
|
+
L.push(`> ⚠️ **${what} (${d.status})** — ${d.reason}`);
|
|
324
|
+
L.push('>');
|
|
325
|
+
L.push(d.efficiencyUsable
|
|
326
|
+
? '> The CORRECTNESS delta below is not evidence of anything. The efficiency deltas are.'
|
|
327
|
+
: '> The deltas below are not evidence of anything. Do not cite them.');
|
|
328
|
+
L.push('');
|
|
329
|
+
}
|
|
330
|
+
else {
|
|
331
|
+
L.push(`> Signal check: ${d.reason}`);
|
|
332
|
+
L.push('');
|
|
333
|
+
}
|
|
196
334
|
L.push(`## Per-condition summary`);
|
|
197
335
|
L.push('');
|
|
198
336
|
const anyQuality = r.perCondition.some((c) => c.meanQuality != null);
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"report.js","sourceRoot":"","sources":["../../../src/benchmarks/paired/report.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,EAEL,OAAO,EAEP,IAAI,EACJ,WAAW,EAGX,OAAO,GAER,MAAM,YAAY,CAAC;AAEpB,OAAO,EAAE,kBAAkB,EAA6C,MAAM,YAAY,CAAC;AAC3F,OAAO,EACL,gBAAgB,EAChB,iBAAiB,GAGlB,MAAM,oBAAoB,CAAC;AAkE5B,SAAS,OAAO,CAAC,CAAY;IAC3B,OAAO,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;AACjC,CAAC;AAED,SAAS,WAAW,CAAC,OAAoB;IACvC,MAAM,CAAC,GAAG,IAAI,GAAG,EAAuB,CAAC;IACzC,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,GAAG,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC;QACrC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACZ,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC;IAC1B,CAAC;IACD,OAAO,CAAC,CAAC;AACX,CAAC;AAED,SAAS,SAAS,CAAC,KAAa,EAAE,IAAiB;IACjD,MAAM,OAAO,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC7D,MAAM,OAAO,GAAG,CAAC,GAAuC,EAAiB,EAAE;QACzE,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAe,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;QACnF,OAAO,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IACzC,CAAC,CAAC;IACF,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,YAAY,EAAE,SAAS,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAe,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;IAC/F,OAAO;QACL,KAAK;QACL,CAAC,EAAE,IAAI,CAAC,MAAM;QACd,WAAW,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,GAAG;QAC9C,UAAU,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC;QACpC,SAAS,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC;QAClC,WAAW,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC;QACtC,aAAa,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI,GAAG;QACjD,SAAS,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG;QAC/E,WAAW,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI;KAC/C,CAAC;AACJ,CAAC;AAED,+EAA+E;AAC/E,SAAS,SAAS,CAChB,QAAgC,EAChC,SAAiC;IAEjC,MAAM,IAAI,GAAa,EAAE,CAAC;IAC1B,MAAM,IAAI,GAAgB,EAAE,CAAC;IAC7B,MAAM,KAAK,GAAgB,EAAE,CAAC;IAC9B,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,QAAQ,EAAE,CAAC;QAC9B,MAAM,CAAC,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;QAC3B,IAAI,CAAC,EAAE,CAAC;YACN,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACb,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACb,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAChB,CAAC;IACH,CAAC;IACD,OAAO,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC;AAC/B,CAAC;AAED,SAAS,SAAS,CAAC,IAAiB;IAClC,MAAM,CAAC,GAAG,IAAI,GAAG,EAAqB,CAAC;IACvC,KAAK,MAAM,CAAC,IAAI,IAAI;QAAE,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IAC3C,OAAO,CAAC,CAAC;AACX,CAAC;AAsBD,MAAM,UAAU,OAAO,CAAC,MAAoB,EAAE,OAAuB,EAAE;IACrE,MAAM,aAAa,GAAG,IAAI,CAAC,aAAa,IAAI,UAAU,CAAC;IACvD,MAAM,OAAO,GAAG,WAAW,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;IAE5C,MAAM,UAAU,GAAG,CAAC,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC,CAAC;IACvC,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,aAAa,CAAC,EAAE,CAAC;QAChC,MAAM,IAAI,KAAK,CACb,gCAAgC,aAAa,wBAAwB,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,CAC9F,CAAC;IACJ,CAAC;IAED,MAAM,YAAY,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,SAAS,CAAC,KAAK,EAAE,OAAO,CAAC,GAAG,CAAC,KAAK,CAAE,CAAC,CAAC,CAAC;IAEtF,MAAM,OAAO,GAAG,SAAS,CAAC,OAAO,CAAC,GAAG,CAAC,aAAa,CAAE,CAAC,CAAC;IACvD,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC;IAEpE,MAAM,WAAW,GAAiB,EAAE,CAAC;IACrC,KAAK,MAAM,KAAK,IAAI,UAAU,EAAE,CAAC;QAC/B,IAAI,KAAK,KAAK,aAAa;YAAE,SAAS;QACtC,MAAM,QAAQ,GAAG,SAAS,CAAC,OAAO,CAAC,GAAG,CAAC,KAAK,CAAE,CAAC,CAAC;QAChD,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,CAAC,OAAO,EAAE,QAAQ,CAAC,CAAC;QAErD,sEAAsE;QACtE,MAAM,iBAAiB,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC;QACrG,MAAM,KAAK,GAAG,WAAW,CAAC,iBAAiB,EAAE,IAAI,CAAC,CAAC;QACnD,MAAM,EAAE,GAAG,OAAO,CAChB,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,EACnC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CACnC,CAAC;QAEF,wEAAwE;QACxE,MAAM,OAAO,GAAyD,EAAE,CAAC;QACzE,KAAK,MAAM,MAAM,IAAI,kBAAkB,EAAE,CAAC;YACxC,MAAM,MAAM,GAAa,EAAE,CAAC;YAC5B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;gBACrC,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;gBACnC,MAAM,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;gBACpC,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,IAAI,IAAI;oBAAE,MAAM,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC;YACrD,CAAC;YACD,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,CAAC,MAAM,CAAC,GAAG,WAAW,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;QACrE,CAAC;QAED,2EAA2E;QAC3E,4EAA4E;QAC5E,MAAM,aAAa,GAAa,EAAE,CAAC;QACnC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACrC,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,YAAY,EAAE,SAAS,CAAC;YAC3C,MAAM,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,YAAY,EAAE,SAAS,CAAC;YAC5C,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,IAAI,IAAI;gBAAE,aAAa,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC;QAC5D,CAAC;QACD,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,aAAa,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QACxF,MAAM,cAAc,GAAG,OAAO;YAC5B,CAAC,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,MAAM,EAAE,IAAI,CAAC,aAAa,IAAI,CAAC,EAAE,cAAc,EAAE,IAAI,EAAE,CAAC;YAC7E,CAAC,CAAC,SAAS,CAAC;QAEd,yEAAyE;QACzE,sEAAsE;QACtE,MAAM,kBAAkB,GAAG,OAAO,CAAC,KAAK,EAAE,EAAE,MAAM,EAAE,IAAI,CAAC,UAAU,IAAI,CAAC,EAAE,cAAc,EAAE,IAAI,EAAE,CAAC,CAAC;QAClG,MAAM,cAAc,GAA+C,EAAE,CAAC;QACtE,KAAK,MAAM,MAAM,IAAI,kBAAkB,EAAE,CAAC;YACxC,MAAM,EAAE,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC;YAC3B,IAAI,EAAE,EAAE,CAAC;gBACP,cAAc,CAAC,MAAM,CAAC,GAAG,OAAO,CAAC,EAAE,EAAE;oBACnC,MAAM,EAAE,IAAI,CAAC,aAAa,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC;oBACzC,cAAc,EAAE,KAAK;iBACtB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,WAAW,CAAC,IAAI,CAAC;YACf,KAAK;YACL,QAAQ,EAAE,aAAa;YACvB,WAAW,EAAE;gBACX,YAAY,EAAE,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC;gBAC3D,aAAa,EAAE,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC;gBAC7D,KAAK;gBACL,OAAO,EAAE,EAAE;gBACX,OAAO,EAAE,kBAAkB;aAC5B;YACD,OAAO;YACP,cAAc;YACd,OAAO;YACP,cAAc;SACf,CAAC,CAAC;IACL,CAAC;IAED,OAAO;QACL,IAAI,EAAE;YACJ,KAAK,EAAE,MAAM,CAAC,KAAK;YACnB,OAAO,EAAE,MAAM,CAAC,OAAO;YACvB,MAAM,EAAE,MAAM,CAAC,MAAM;YACrB,SAAS;YACT,UAAU;YACV,QAAQ,EAAE,aAAa;YACvB,SAAS,EAAE,MAAM,CAAC,SAAS;YAC3B,UAAU,EAAE,MAAM,CAAC,UAAU;SAC9B;QACD,YAAY;QACZ,WAAW;QACX,MAAM,EAAE,YAAY,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;YAC/B,KAAK,EAAE,CAAC,CAAC,KAAK;YACd,WAAW,EAAE,CAAC,CAAC,WAAW;YAC1B,UAAU,EAAE,CAAC,CAAC,UAAU;SACzB,CAAC,CAAC;QACH,2EAA2E;QAC3E,sEAAsE;QACtE,GAAG,CAAC,IAAI,CAAC,OAAO;YACd,CAAC,CAAC,EAAE,WAAW,EAAE,gBAAgB,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC,EAAE;YAC7E,CAAC,CAAC,EAAE,CAAC;KACR,CAAC;AACJ,CAAC;AAED,SAAS,GAAG,CAAC,CAAU;IACrB,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AACnB,CAAC;AAED,8EAA8E;AAC9E,oBAAoB;AACpB,8EAA8E;AAE9E,SAAS,GAAG,CAAC,CAAS;IACpB,OAAO,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC;AACjE,CAAC;AACD,SAAS,KAAK,CAAC,CAAgB,EAAE,CAAC,GAAG,CAAC;IACpC,OAAO,CAAC,IAAI,IAAI,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC;AACjE,CAAC;AACD,SAAS,KAAK,CAAC,EAAM,EAAE,CAAC,GAAG,CAAC;IAC1B,OAAO,IAAI,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC;AAC5D,CAAC;AACD,SAAS,GAAG,CAAC,CAAoB;IAC/B,OAAO,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC;AACnC,CAAC;AAED,+EAA+E;AAC/E,SAAS,MAAM,CAAC,CAAU;IACxB,OAAO,CAAC,KAAK,KAAK,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,KAAK,MAAM,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,OAAO,CAAC;AACrE,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,CAAiB;IAC9C,MAAM,CAAC,GAAa,EAAE,CAAC;IACvB,CAAC,CAAC,IAAI,CAAC,+BAA+B,CAAC,CAAC;IACxC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,CAAC,CAAC,IAAI,CACJ,cAAc,CAAC,CAAC,IAAI,CAAC,KAAK,qBAAqB,CAAC,CAAC,IAAI,CAAC,OAAO,OAAO;QAClE,cAAc,CAAC,CAAC,IAAI,CAAC,SAAS,oBAAoB,CAAC,CAAC,IAAI,CAAC,MAAM,OAAO;QACtE,mBAAmB,CAAC,CAAC,IAAI,CAAC,QAAQ,IAAI,CACzC,CAAC;IACF,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,SAAS,MAAM,CAAC,CAAC,IAAI,CAAC,UAAU,GAAG,CAAC,CAAC;IACvD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEX,CAAC,CAAC,IAAI,CAAC,0BAA0B,CAAC,CAAC;IACnC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,MAAM,UAAU,GAAG,CAAC,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,IAAI,IAAI,CAAC,CAAC;IACrE,CAAC,CAAC,IAAI,CACJ,0EAA0E;QACxE,CAAC,UAAU,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,CAAC,CACnC,CAAC;IACF,CAAC,CAAC,IAAI,CAAC,mCAAmC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IACzE,KAAK,MAAM,CAAC,IAAI,CAAC,CAAC,YAAY,EAAE,CAAC;QAC/B,CAAC,CAAC,IAAI,CACJ,OAAO,CAAC,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,WAAW,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,UAAU,CAAC,KAAK;YAC7E,GAAG,KAAK,CAAC,CAAC,CAAC,SAAS,EAAE,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,aAAa,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI;YAC3G,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CACtD,CAAC;IACJ,CAAC;IACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEX,CAAC,CAAC,IAAI,CAAC,8BAA8B,CAAC,CAAC,IAAI,CAAC,QAAQ,IAAI,CAAC,CAAC;IAC1D,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,KAAK,MAAM,GAAG,IAAI,CAAC,CAAC,WAAW,EAAE,CAAC;QAChC,CAAC,CAAC,IAAI,CAAC,SAAS,GAAG,CAAC,KAAK,IAAI,CAAC,CAAC;QAC/B,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACX,MAAM,EAAE,GAAG,GAAG,CAAC,WAAW,CAAC;QAC3B,CAAC,CAAC,IAAI,CACJ,gBAAgB,MAAM,CAAC,EAAE,CAAC,OAAO,CAAC,IAAI;YACpC,oBAAoB,GAAG,CAAC,EAAE,CAAC,YAAY,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,aAAa,CAAC,IAAI;YACvE,MAAM,CAAC,EAAE,CAAC,KAAK,CAAC,SAAS,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,cAAc,KAAK,CAAC,EAAE,CAAC,KAAK,CAAC,EAAE,CAAC,IAAI;YAC/E,KAAK,EAAE,CAAC,KAAK,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,EAAE,CAAC,KAAK,CAAC,EAAE,CACtD,CAAC;QACF,MAAM,CAAC,GAAG,EAAE,CAAC,OAAO,CAAC;QACrB,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACX,CAAC,CAAC,IAAI,CACJ,uCAAuC,CAAC,CAAC,aAAa,eAAe,CAAC,CAAC,YAAY,IAAI;YACrF,OAAO,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,OAAO,OAAO,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK;YAC3E,SAAS,CAAC,CAAC,WAAW,WAAW,CAAC,CAAC,SAAS,EAAE,CACjD,CAAC;QACF,IAAI,GAAG,CAAC,OAAO,EAAE,CAAC;YAChB,MAAM,CAAC,GAAG,GAAG,CAAC,OAAO,CAAC;YACtB,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YACX,CAAC,CAAC,IAAI,CACJ,kCAAkC,MAAM,CAAC,GAAG,CAAC,cAAc,IAAI,KAAK,CAAC,IAAI;gBACvE,KAAK,CAAC,CAAC,SAAS,CAAC,OAAO,CAAC,CAAC,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,CAAC,EAAE,CAC9F,CAAC;QACJ,CAAC;QACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACX,CAAC,CAAC,IAAI,CAAC,4CAA4C,CAAC,CAAC;QACrD,CAAC,CAAC,IAAI,CAAC,uBAAuB,CAAC,CAAC;QAChC,KAAK,MAAM,MAAM,IAAI,kBAAkB,EAAE,CAAC;YACxC,MAAM,EAAE,GAAG,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;YAC/B,IAAI,CAAC,EAAE;gBAAE,SAAS;YAClB,MAAM,CAAC,GAAG,MAAM,KAAK,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACvC,MAAM,EAAE,GAAG,GAAG,CAAC,cAAc,CAAC,MAAM,CAAC,IAAI,KAAK,CAAC;YAC/C,CAAC,CAAC,IAAI,CACJ,KAAK,MAAM,MAAM,EAAE,CAAC,SAAS,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,KAAK;gBAChE,GAAG,EAAE,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,MAAM,CAAC,EAAE,CAAC,IAAI,CAC9C,CAAC;QACJ,CAAC;QACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACb,CAAC;IAED,CAAC,CAAC,IAAI,CAAC,yBAAyB,CAAC,CAAC;IAClC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,CAAC,CAAC,IAAI,CAAC,uCAAuC,CAAC,CAAC;IAChD,CAAC,CAAC,IAAI,CAAC,eAAe,CAAC,CAAC;IACxB,KAAK,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC,WAAW,CAAC,EAAE,CAAC;QAC5E,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,QAAQ,GAAG,CAAC,CAAC,CAAC,WAAW,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC;IAChF,CAAC;IACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,CAAC,CAAC,IAAI,CACJ,4EAA4E;QAC1E,6FAA6F;QAC7F,yEAAyE,CAC5E,CAAC;IACF,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,4EAA4E;IAC5E,+DAA+D;IAC/D,IAAI,CAAC,CAAC,WAAW,EAAE,CAAC;QAClB,CAAC,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC;QACzC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACb,CAAC;IACD,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AACtB,CAAC"}
|
|
1
|
+
{"version":3,"file":"report.js","sourceRoot":"","sources":["../../../src/benchmarks/paired/report.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,EAEL,OAAO,EAEP,IAAI,EACJ,WAAW,EAGX,OAAO,GAER,MAAM,YAAY,CAAC;AAEpB,OAAO,EAAE,kBAAkB,EAA6C,MAAM,YAAY,CAAC;AAC3F,OAAO,EACL,gBAAgB,EAChB,iBAAiB,GAGlB,MAAM,oBAAoB,CAAC;AAiJ5B,SAAS,MAAM,CAAC,CAA4B;IAC1C,OAAO,OAAO,CAAC,KAAK,QAAQ,IAAI,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC;AACrD,CAAC;AAED;;;;;;;;GAQG;AACH,SAAS,mBAAmB,CAAC,CAAyB;IACpD,IAAI,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IACrB,MAAM,OAAO,GAAG,CAAC,EAAgD,EAAW,EAAE,CAC5E,OAAO,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,IAAI,MAAM,CAAC,EAAE,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC,SAAS,KAAK,CAAC,CAAC,CAAC;IACxE,KAAK,MAAM,CAAC,IAAI,kBAAkB,EAAE,CAAC;QACnC,IAAI,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC;YAAE,OAAO,IAAI,CAAC;IACzC,CAAC;IACD,OAAO,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC;AAC5B,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,UAAU,oBAAoB,CAClC,YAAgC,EAChC,WAAyB,EACzB,OAA8B,EAAE;IAEhC,MAAM,SAAS,GAAG,IAAI,CAAC,SAAS,IAAI,IAAI,CAAC;IACzC,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC;IACrC,MAAM,QAAQ,GAAG,IAAI,CAAC,cAAc,IAAI,CAAC,CAAC;IAC1C,MAAM,GAAG,GAAG,IAAI,CAAC,mBAAmB,IAAI,IAAI,CAAC;IAE7C,MAAM,KAAK,GAAG,YAAY,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;IACpE,MAAM,UAAU,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACzD,MAAM,UAAU,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAEzD,MAAM,OAAO,GAAG,WAAW,CAAC,CAAC,CAAC,CAAC;IAC/B,MAAM,gBAAgB,GAAG,mBAAmB,CAAC,OAAO,CAAC,CAAC;IACtD,MAAM,CAAC,GAAG,OAAO,EAAE,WAAW,CAAC,KAAK,CAAC;IACrC,MAAM,WAAW,GAAG,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAE,CAAmB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC9D,MAAM,eAAe,GAAG,OAAO;QAC7B,CAAC,CAAC,OAAO,CAAC,WAAW,CAAC,OAAO,CAAC,aAAa,GAAG,OAAO,CAAC,WAAW,CAAC,OAAO,CAAC,YAAY;QACtF,CAAC,CAAC,CAAC,CAAC;IACN,MAAM,IAAI,GAAG,EAAE,UAAU,EAAE,UAAU,EAAE,WAAW,EAAE,eAAe,EAAE,gBAAgB,EAAE,CAAC;IACxF,MAAM,cAAc,GAAG,gBAAgB;QACrC,CAAC,CAAC,6EAA6E;QAC/E,CAAC,CAAC,EAAE,CAAC;IAEP,IAAI,CAAC,OAAO,EAAE,CAAC;QACb,OAAO;YACL,MAAM,EAAE,gBAAgB;YACxB,MAAM,EAAE,KAAK;YACb,MAAM,EAAE,0EAA0E;YAClF,GAAG,IAAI;SACR,CAAC;IACJ,CAAC;IACD,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,IAAI,UAAU,IAAI,SAAS,EAAE,CAAC;QAChD,OAAO;YACL,MAAM,EAAE,SAAS;YACjB,MAAM,EAAE,KAAK;YACb,MAAM,EACJ,0BAA0B,CAAC,UAAU,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,0CAA0C;gBACjG,kGAAkG;gBAClG,4FAA4F;gBAC5F,cAAc;YAChB,GAAG,IAAI;SACR,CAAC;IACJ,CAAC;IACD,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,IAAI,UAAU,IAAI,OAAO,EAAE,CAAC;QAC9C,OAAO;YACL,MAAM,EAAE,OAAO;YACf,MAAM,EAAE,KAAK;YACb,MAAM,EACJ,qCAAqC,CAAC,UAAU,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,8BAA8B;gBAChG,6FAA6F;gBAC7F,+EAA+E,GAAG,cAAc;YAClG,GAAG,IAAI;SACR,CAAC;IACJ,CAAC;IACD,IAAI,eAAe,KAAK,CAAC,EAAE,CAAC;QAC1B,OAAO;YACL,MAAM,EAAE,qBAAqB;YAC7B,MAAM,EAAE,KAAK;YACb,MAAM,EACJ,kDAAkD,WAAW,mCAAmC;gBAChG,8FAA8F;gBAC9F,oCAAoC,GAAG,cAAc;YACvD,GAAG,IAAI;SACR,CAAC;IACJ,CAAC;IACD,IAAI,WAAW,GAAG,CAAC,IAAI,WAAW,GAAG,QAAQ,EAAE,CAAC;QAC9C,OAAO;YACL,MAAM,EAAE,cAAc;YACtB,MAAM,EAAE,KAAK;YACb,MAAM,EACJ,QAAQ,WAAW,mEAAmE;gBACtF,uDAAuD,GAAG,cAAc;YAC1E,GAAG,IAAI;SACR,CAAC;IACJ,CAAC;IACD,0EAA0E;IAC1E,wEAAwE;IACxE,6EAA6E;IAC7E,gDAAgD;IAChD,MAAM,EAAE,GAAG,CAAC,EAAE,EAAE,CAAC;IACjB,MAAM,WAAW,GAAG,OAAO,CAAC,CAAC,EAAE,WAAW,CAAC,CAAC;IAC5C,IAAI,CAAC,WAAW,IAAI,EAAE,IAAI,MAAM,CAAC,EAAE,CAAC,KAAK,CAAC,IAAI,MAAM,CAAC,EAAE,CAAC,KAAK,CAAC,IAAI,EAAE,CAAC,KAAK,GAAG,GAAG,IAAI,EAAE,CAAC,KAAK,GAAG,CAAC,GAAG,EAAE,CAAC;QACpG,OAAO;YACL,MAAM,EAAE,cAAc;YACtB,MAAM,EAAE,KAAK;YACb,MAAM,EACJ,uBAAuB,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,+BAA+B;gBACjG,IAAI,GAAG,sFAAsF;gBAC7F,8BAA8B,GAAG,cAAc;YACjD,GAAG,IAAI;SACR,CAAC;IACJ,CAAC;IACD,OAAO;QACL,MAAM,EAAE,IAAI;QACZ,MAAM,EAAE,IAAI;QACZ,MAAM,EACJ,mBAAmB,CAAC,UAAU,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC,UAAU,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,SAAS;YAC3F,GAAG,WAAW,sBAAsB,eAAe,gDAAgD;QACrG,GAAG,IAAI;KACR,CAAC;AACJ,CAAC;AAED,SAAS,OAAO,CAAC,CAAY;IAC3B,OAAO,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;AACjC,CAAC;AAED,SAAS,WAAW,CAAC,OAAoB;IACvC,MAAM,CAAC,GAAG,IAAI,GAAG,EAAuB,CAAC;IACzC,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,GAAG,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC;QACrC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACZ,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC;IAC1B,CAAC;IACD,OAAO,CAAC,CAAC;AACX,CAAC;AAED,SAAS,SAAS,CAAC,KAAa,EAAE,IAAiB;IACjD,MAAM,OAAO,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC7D,MAAM,OAAO,GAAG,CAAC,GAAuC,EAAiB,EAAE;QACzE,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAe,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;QACnF,OAAO,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IACzC,CAAC,CAAC;IACF,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,YAAY,EAAE,SAAS,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAe,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;IAC/F,OAAO;QACL,KAAK;QACL,CAAC,EAAE,IAAI,CAAC,MAAM;QACd,WAAW,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,GAAG;QAC9C,UAAU,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC;QACpC,SAAS,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC;QAClC,WAAW,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC;QACtC,aAAa,EAAE,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI,GAAG;QACjD,SAAS,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG;QAC/E,WAAW,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI;KAC/C,CAAC;AACJ,CAAC;AAED,+EAA+E;AAC/E,SAAS,SAAS,CAChB,QAAgC,EAChC,SAAiC;IAEjC,MAAM,IAAI,GAAa,EAAE,CAAC;IAC1B,MAAM,IAAI,GAAgB,EAAE,CAAC;IAC7B,MAAM,KAAK,GAAgB,EAAE,CAAC;IAC9B,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,QAAQ,EAAE,CAAC;QAC9B,MAAM,CAAC,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;QAC3B,IAAI,CAAC,EAAE,CAAC;YACN,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACb,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACb,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAChB,CAAC;IACH,CAAC;IACD,OAAO,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC;AAC/B,CAAC;AAED,SAAS,SAAS,CAAC,IAAiB;IAClC,MAAM,CAAC,GAAG,IAAI,GAAG,EAAqB,CAAC;IACvC,KAAK,MAAM,CAAC,IAAI,IAAI;QAAE,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IAC3C,OAAO,CAAC,CAAC;AACX,CAAC;AAwBD,MAAM,UAAU,OAAO,CAAC,MAAoB,EAAE,OAAuB,EAAE;IACrE,MAAM,aAAa,GAAG,IAAI,CAAC,aAAa,IAAI,UAAU,CAAC;IACvD,MAAM,OAAO,GAAG,WAAW,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;IAE5C,MAAM,UAAU,GAAG,CAAC,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC,CAAC;IACvC,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,aAAa,CAAC,EAAE,CAAC;QAChC,MAAM,IAAI,KAAK,CACb,gCAAgC,aAAa,wBAAwB,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,CAC9F,CAAC;IACJ,CAAC;IAED,MAAM,YAAY,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,SAAS,CAAC,KAAK,EAAE,OAAO,CAAC,GAAG,CAAC,KAAK,CAAE,CAAC,CAAC,CAAC;IAEtF,MAAM,OAAO,GAAG,SAAS,CAAC,OAAO,CAAC,GAAG,CAAC,aAAa,CAAE,CAAC,CAAC;IACvD,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC;IAEpE,MAAM,WAAW,GAAiB,EAAE,CAAC;IACrC,KAAK,MAAM,KAAK,IAAI,UAAU,EAAE,CAAC;QAC/B,IAAI,KAAK,KAAK,aAAa;YAAE,SAAS;QACtC,MAAM,QAAQ,GAAG,SAAS,CAAC,OAAO,CAAC,GAAG,CAAC,KAAK,CAAE,CAAC,CAAC;QAChD,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,CAAC,OAAO,EAAE,QAAQ,CAAC,CAAC;QAErD,sEAAsE;QACtE,MAAM,iBAAiB,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC;QACrG,MAAM,KAAK,GAAG,WAAW,CAAC,iBAAiB,EAAE,IAAI,CAAC,CAAC;QACnD,MAAM,EAAE,GAAG,OAAO,CAChB,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,EACnC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CACnC,CAAC;QAEF,wEAAwE;QACxE,MAAM,OAAO,GAAyD,EAAE,CAAC;QACzE,KAAK,MAAM,MAAM,IAAI,kBAAkB,EAAE,CAAC;YACxC,MAAM,MAAM,GAAa,EAAE,CAAC;YAC5B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;gBACrC,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;gBACnC,MAAM,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;gBACpC,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,IAAI,IAAI;oBAAE,MAAM,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC;YACrD,CAAC;YACD,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,CAAC,MAAM,CAAC,GAAG,WAAW,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;QACrE,CAAC;QAED,2EAA2E;QAC3E,4EAA4E;QAC5E,MAAM,aAAa,GAAa,EAAE,CAAC;QACnC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACrC,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,YAAY,EAAE,SAAS,CAAC;YAC3C,MAAM,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,YAAY,EAAE,SAAS,CAAC;YAC5C,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,IAAI,IAAI;gBAAE,aAAa,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC;QAC5D,CAAC;QACD,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,aAAa,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QACxF,MAAM,cAAc,GAAG,OAAO;YAC5B,CAAC,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,MAAM,EAAE,IAAI,CAAC,aAAa,IAAI,CAAC,EAAE,cAAc,EAAE,IAAI,EAAE,CAAC;YAC7E,CAAC,CAAC,SAAS,CAAC;QAEd,yEAAyE;QACzE,sEAAsE;QACtE,MAAM,kBAAkB,GAAG,OAAO,CAAC,KAAK,EAAE,EAAE,MAAM,EAAE,IAAI,CAAC,UAAU,IAAI,CAAC,EAAE,cAAc,EAAE,IAAI,EAAE,CAAC,CAAC;QAClG,MAAM,cAAc,GAA+C,EAAE,CAAC;QACtE,KAAK,MAAM,MAAM,IAAI,kBAAkB,EAAE,CAAC;YACxC,MAAM,EAAE,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC;YAC3B,IAAI,EAAE,EAAE,CAAC;gBACP,cAAc,CAAC,MAAM,CAAC,GAAG,OAAO,CAAC,EAAE,EAAE;oBACnC,MAAM,EAAE,IAAI,CAAC,aAAa,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC;oBACzC,cAAc,EAAE,KAAK;iBACtB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,WAAW,CAAC,IAAI,CAAC;YACf,KAAK;YACL,QAAQ,EAAE,aAAa;YACvB,WAAW,EAAE;gBACX,YAAY,EAAE,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC;gBAC3D,aAAa,EAAE,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC;gBAC7D,KAAK;gBACL,OAAO,EAAE,EAAE;gBACX,OAAO,EAAE,kBAAkB;aAC5B;YACD,OAAO;YACP,cAAc;YACd,OAAO;YACP,cAAc;SACf,CAAC,CAAC;IACL,CAAC;IAED,OAAO;QACL,IAAI,EAAE;YACJ,KAAK,EAAE,MAAM,CAAC,KAAK;YACnB,OAAO,EAAE,MAAM,CAAC,OAAO;YACvB,MAAM,EAAE,MAAM,CAAC,MAAM;YACrB,SAAS;YACT,UAAU;YACV,QAAQ,EAAE,aAAa;YACvB,SAAS,EAAE,MAAM,CAAC,SAAS;YAC3B,UAAU,EAAE,MAAM,CAAC,UAAU;SAC9B;QACD,YAAY;QACZ,WAAW;QACX,MAAM,EAAE,YAAY,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;YAC/B,KAAK,EAAE,CAAC,CAAC,KAAK;YACd,WAAW,EAAE,CAAC,CAAC,WAAW;YAC1B,UAAU,EAAE,CAAC,CAAC,UAAU;SACzB,CAAC,CAAC;QACH,cAAc,EAAE,oBAAoB,CAAC,YAAY,EAAE,WAAW,EAAE,IAAI,CAAC,cAAc,CAAC;QACpF,2EAA2E;QAC3E,sEAAsE;QACtE,GAAG,CAAC,IAAI,CAAC,OAAO;YACd,CAAC,CAAC,EAAE,WAAW,EAAE,gBAAgB,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC,EAAE;YAC7E,CAAC,CAAC,EAAE,CAAC;KACR,CAAC;AACJ,CAAC;AAED,SAAS,GAAG,CAAC,CAAU;IACrB,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AACnB,CAAC;AAED,8EAA8E;AAC9E,oBAAoB;AACpB,8EAA8E;AAE9E,SAAS,GAAG,CAAC,CAAS;IACpB,OAAO,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC;AACjE,CAAC;AACD,SAAS,KAAK,CAAC,CAAgB,EAAE,CAAC,GAAG,CAAC;IACpC,OAAO,CAAC,IAAI,IAAI,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC;AACjE,CAAC;AACD,SAAS,KAAK,CAAC,EAAM,EAAE,CAAC,GAAG,CAAC;IAC1B,OAAO,IAAI,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC;AAC5D,CAAC;AACD,SAAS,GAAG,CAAC,CAAoB;IAC/B,OAAO,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC;AACnC,CAAC;AAED,+EAA+E;AAC/E,SAAS,MAAM,CAAC,CAAU;IACxB,OAAO,CAAC,KAAK,KAAK,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,KAAK,MAAM,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,OAAO,CAAC;AACrE,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,CAAiB;IAC9C,MAAM,CAAC,GAAa,EAAE,CAAC;IACvB,CAAC,CAAC,IAAI,CAAC,+BAA+B,CAAC,CAAC;IACxC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,CAAC,CAAC,IAAI,CACJ,cAAc,CAAC,CAAC,IAAI,CAAC,KAAK,qBAAqB,CAAC,CAAC,IAAI,CAAC,OAAO,OAAO;QAClE,cAAc,CAAC,CAAC,IAAI,CAAC,SAAS,oBAAoB,CAAC,CAAC,IAAI,CAAC,MAAM,OAAO;QACtE,mBAAmB,CAAC,CAAC,IAAI,CAAC,QAAQ,IAAI,CACzC,CAAC;IACF,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,SAAS,MAAM,CAAC,CAAC,IAAI,CAAC,UAAU,GAAG,CAAC,CAAC;IACvD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEX,6EAA6E;IAC7E,6EAA6E;IAC7E,MAAM,CAAC,GAAG,CAAC,CAAC,cAAc,CAAC;IAC3B,IAAI,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC;QACd,MAAM,IAAI,GAAG,CAAC,CAAC,gBAAgB,CAAC,CAAC,CAAC,uBAAuB,CAAC,CAAC,CAAC,kBAAkB,CAAC;QAC/E,CAAC,CAAC,IAAI,CAAC,UAAU,IAAI,KAAK,CAAC,CAAC,MAAM,SAAS,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC;QACvD,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACZ,CAAC,CAAC,IAAI,CACJ,CAAC,CAAC,gBAAgB;YAChB,CAAC,CAAC,uFAAuF;YACzF,CAAC,CAAC,oEAAoE,CACzE,CAAC;QACF,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACb,CAAC;SAAM,CAAC;QACN,CAAC,CAAC,IAAI,CAAC,mBAAmB,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC;QACtC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACb,CAAC;IAED,CAAC,CAAC,IAAI,CAAC,0BAA0B,CAAC,CAAC;IACnC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,MAAM,UAAU,GAAG,CAAC,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,IAAI,IAAI,CAAC,CAAC;IACrE,CAAC,CAAC,IAAI,CACJ,0EAA0E;QACxE,CAAC,UAAU,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,CAAC,CACnC,CAAC;IACF,CAAC,CAAC,IAAI,CAAC,mCAAmC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IACzE,KAAK,MAAM,CAAC,IAAI,CAAC,CAAC,YAAY,EAAE,CAAC;QAC/B,CAAC,CAAC,IAAI,CACJ,OAAO,CAAC,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,WAAW,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,UAAU,CAAC,KAAK;YAC7E,GAAG,KAAK,CAAC,CAAC,CAAC,SAAS,EAAE,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,aAAa,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,IAAI;YAC3G,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CACtD,CAAC;IACJ,CAAC;IACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAEX,CAAC,CAAC,IAAI,CAAC,8BAA8B,CAAC,CAAC,IAAI,CAAC,QAAQ,IAAI,CAAC,CAAC;IAC1D,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,KAAK,MAAM,GAAG,IAAI,CAAC,CAAC,WAAW,EAAE,CAAC;QAChC,CAAC,CAAC,IAAI,CAAC,SAAS,GAAG,CAAC,KAAK,IAAI,CAAC,CAAC;QAC/B,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACX,MAAM,EAAE,GAAG,GAAG,CAAC,WAAW,CAAC;QAC3B,CAAC,CAAC,IAAI,CACJ,gBAAgB,MAAM,CAAC,EAAE,CAAC,OAAO,CAAC,IAAI;YACpC,oBAAoB,GAAG,CAAC,EAAE,CAAC,YAAY,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,aAAa,CAAC,IAAI;YACvE,MAAM,CAAC,EAAE,CAAC,KAAK,CAAC,SAAS,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,cAAc,KAAK,CAAC,EAAE,CAAC,KAAK,CAAC,EAAE,CAAC,IAAI;YAC/E,KAAK,EAAE,CAAC,KAAK,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,EAAE,CAAC,KAAK,CAAC,EAAE,CACtD,CAAC;QACF,MAAM,CAAC,GAAG,EAAE,CAAC,OAAO,CAAC;QACrB,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACX,CAAC,CAAC,IAAI,CACJ,uCAAuC,CAAC,CAAC,aAAa,eAAe,CAAC,CAAC,YAAY,IAAI;YACrF,OAAO,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,OAAO,OAAO,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK;YAC3E,SAAS,CAAC,CAAC,WAAW,WAAW,CAAC,CAAC,SAAS,EAAE,CACjD,CAAC;QACF,IAAI,GAAG,CAAC,OAAO,EAAE,CAAC;YAChB,MAAM,CAAC,GAAG,GAAG,CAAC,OAAO,CAAC;YACtB,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YACX,CAAC,CAAC,IAAI,CACJ,kCAAkC,MAAM,CAAC,GAAG,CAAC,cAAc,IAAI,KAAK,CAAC,IAAI;gBACvE,KAAK,CAAC,CAAC,SAAS,CAAC,OAAO,CAAC,CAAC,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,CAAC,EAAE,CAC9F,CAAC;QACJ,CAAC;QACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACX,CAAC,CAAC,IAAI,CAAC,4CAA4C,CAAC,CAAC;QACrD,CAAC,CAAC,IAAI,CAAC,uBAAuB,CAAC,CAAC;QAChC,KAAK,MAAM,MAAM,IAAI,kBAAkB,EAAE,CAAC;YACxC,MAAM,EAAE,GAAG,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;YAC/B,IAAI,CAAC,EAAE;gBAAE,SAAS;YAClB,MAAM,CAAC,GAAG,MAAM,KAAK,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACvC,MAAM,EAAE,GAAG,GAAG,CAAC,cAAc,CAAC,MAAM,CAAC,IAAI,KAAK,CAAC;YAC/C,CAAC,CAAC,IAAI,CACJ,KAAK,MAAM,MAAM,EAAE,CAAC,SAAS,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,KAAK;gBAChE,GAAG,EAAE,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,MAAM,CAAC,EAAE,CAAC,IAAI,CAC9C,CAAC;QACJ,CAAC;QACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACb,CAAC;IAED,CAAC,CAAC,IAAI,CAAC,yBAAyB,CAAC,CAAC;IAClC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,CAAC,CAAC,IAAI,CAAC,uCAAuC,CAAC,CAAC;IAChD,CAAC,CAAC,IAAI,CAAC,eAAe,CAAC,CAAC;IACxB,KAAK,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC,WAAW,CAAC,EAAE,CAAC;QAC5E,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,QAAQ,GAAG,CAAC,CAAC,CAAC,WAAW,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC;IAChF,CAAC;IACD,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,CAAC,CAAC,IAAI,CACJ,4EAA4E;QAC1E,6FAA6F;QAC7F,yEAAyE,CAC5E,CAAC;IACF,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACX,4EAA4E;IAC5E,+DAA+D;IAC/D,IAAI,CAAC,CAAC,WAAW,EAAE,CAAC;QAClB,CAAC,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC;QACzC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACb,CAAC;IACD,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AACtB,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"runner.d.ts","sourceRoot":"","sources":["../../../src/benchmarks/paired/runner.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAIH,OAAO,EAAgB,SAAS,EAAE,YAAY,EAAuB,MAAM,YAAY,CAAC;AAQxF,MAAM,WAAW,YAAY;IAC3B,OAAO,EAAE,SAAS,EAAE,CAAC;IACrB,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,EAAE,MAAM,CAAC;IACf,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;CACpB;AAED;;;GAGG;AACH,wBAAsB,SAAS,CAC7B,GAAG,EAAE,YAAY,EACjB,QAAQ,EAAE,MAAM,EAChB,YAAY,EAAE,MAAM,GACnB,OAAO,CAAC,YAAY,CAAC,
|
|
1
|
+
{"version":3,"file":"runner.d.ts","sourceRoot":"","sources":["../../../src/benchmarks/paired/runner.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAIH,OAAO,EAAgB,SAAS,EAAE,YAAY,EAAuB,MAAM,YAAY,CAAC;AAQxF,MAAM,WAAW,YAAY;IAC3B,OAAO,EAAE,SAAS,EAAE,CAAC;IACrB,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,EAAE,MAAM,CAAC;IACf,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;CACpB;AAED;;;GAGG;AACH,wBAAsB,SAAS,CAC7B,GAAG,EAAE,YAAY,EACjB,QAAQ,EAAE,MAAM,EAChB,YAAY,EAAE,MAAM,GACnB,OAAO,CAAC,YAAY,CAAC,CAoCvB"}
|
|
@@ -30,7 +30,12 @@ export async function runPaired(cfg, suiteDir, startedAtIso) {
|
|
|
30
30
|
done++;
|
|
31
31
|
cfg.onProgress?.(done, total, `${cell.task.id}/${cell.condition.label}#${cell.seed}`);
|
|
32
32
|
return rec;
|
|
33
|
-
},
|
|
33
|
+
},
|
|
34
|
+
// 'model', not 'cpu': every cell is an inference call, so the ceiling is the
|
|
35
|
+
// backend's slot budget, not core count. The mode only decides the fallback
|
|
36
|
+
// when the caller passes no explicit concurrency — but declaring it wrong
|
|
37
|
+
// meant an unset concurrency sized the run against the wrong resource.
|
|
38
|
+
{ maxConcurrent: cfg.concurrency, mode: 'model' });
|
|
34
39
|
return {
|
|
35
40
|
records,
|
|
36
41
|
model: cfg.model,
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"runner.js","sourceRoot":"","sources":["../../../src/benchmarks/paired/runner.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAEH,OAAO,EAAE,aAAa,EAAE,MAAM,iCAAiC,CAAC;AAChE,OAAO,EAAE,kBAAkB,EAAE,QAAQ,EAAE,SAAS,EAAE,MAAM,YAAY,CAAC;AAkBrE;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS,CAC7B,GAAiB,EACjB,QAAgB,EAChB,YAAoB;IAEpB,MAAM,KAAK,GAAW,EAAE,CAAC;IACzB,KAAK,MAAM,IAAI,IAAI,GAAG,CAAC,KAAK,EAAE,CAAC;QAC7B,KAAK,IAAI,IAAI,GAAG,CAAC,EAAE,IAAI,GAAG,GAAG,CAAC,MAAM,EAAE,IAAI,EAAE,EAAE,CAAC;YAC7C,KAAK,MAAM,SAAS,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC;gBACvC,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;YACxC,CAAC;QACH,CAAC;IACH,CAAC;IAED,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,MAAM,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC;IAE3B,MAAM,OAAO,GAAG,MAAM,aAAa,CACjC,KAAK,EACL,KAAK,EAAE,IAAI,EAAsB,EAAE;QACjC,MAAM,GAAG,GAAG,MAAM,WAAW,CAAC,IAAI,EAAE,GAAG,EAAE,QAAQ,CAAC,CAAC;QACnD,IAAI,EAAE,CAAC;QACP,GAAG,CAAC,UAAU,EAAE,CAAC,IAAI,EAAE,KAAK,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,IAAI,CAAC,SAAS,CAAC,KAAK,IAAI,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC;QACtF,OAAO,GAAG,CAAC;IACb,CAAC,
|
|
1
|
+
{"version":3,"file":"runner.js","sourceRoot":"","sources":["../../../src/benchmarks/paired/runner.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAEH,OAAO,EAAE,aAAa,EAAE,MAAM,iCAAiC,CAAC;AAChE,OAAO,EAAE,kBAAkB,EAAE,QAAQ,EAAE,SAAS,EAAE,MAAM,YAAY,CAAC;AAkBrE;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS,CAC7B,GAAiB,EACjB,QAAgB,EAChB,YAAoB;IAEpB,MAAM,KAAK,GAAW,EAAE,CAAC;IACzB,KAAK,MAAM,IAAI,IAAI,GAAG,CAAC,KAAK,EAAE,CAAC;QAC7B,KAAK,IAAI,IAAI,GAAG,CAAC,EAAE,IAAI,GAAG,GAAG,CAAC,MAAM,EAAE,IAAI,EAAE,EAAE,CAAC;YAC7C,KAAK,MAAM,SAAS,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC;gBACvC,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;YACxC,CAAC;QACH,CAAC;IACH,CAAC;IAED,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,MAAM,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC;IAE3B,MAAM,OAAO,GAAG,MAAM,aAAa,CACjC,KAAK,EACL,KAAK,EAAE,IAAI,EAAsB,EAAE;QACjC,MAAM,GAAG,GAAG,MAAM,WAAW,CAAC,IAAI,EAAE,GAAG,EAAE,QAAQ,CAAC,CAAC;QACnD,IAAI,EAAE,CAAC;QACP,GAAG,CAAC,UAAU,EAAE,CAAC,IAAI,EAAE,KAAK,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,IAAI,CAAC,SAAS,CAAC,KAAK,IAAI,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC;QACtF,OAAO,GAAG,CAAC;IACb,CAAC;IACD,6EAA6E;IAC7E,4EAA4E;IAC5E,0EAA0E;IAC1E,uEAAuE;IACvE,EAAE,aAAa,EAAE,GAAG,CAAC,WAAW,EAAE,IAAI,EAAE,OAAO,EAAE,CAClD,CAAC;IAEF,OAAO;QACL,OAAO;QACP,KAAK,EAAE,GAAG,CAAC,KAAK;QAChB,OAAO,EAAE,GAAG,CAAC,OAAO,CAAC,EAAE;QACvB,MAAM,EAAE,GAAG,CAAC,MAAM;QAClB,SAAS,EAAE,YAAY;QACvB,UAAU,EAAE,MAAM,EAAE;KACrB,CAAC;AACJ,CAAC;AAED,KAAK,UAAU,WAAW,CAAC,IAAU,EAAE,GAAiB,EAAE,QAAgB;IACxE,MAAM,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,GAAG,IAAI,CAAC;IACvC,MAAM,EAAE,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;IAEtB,MAAM,WAAW,GAAG,GAAiB,EAAE,CAAC,CAAC;QACvC,OAAO,EAAE,KAAK;QACd,MAAM,EAAE,IAAI;QACZ,OAAO,EAAE,IAAI;QACb,KAAK,EAAE,IAAI;QACX,SAAS,EAAE,IAAI;QACf,SAAS,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,EAAE;QAC1B,UAAU,EAAE,IAAI;QAChB,KAAK,EAAE,IAAI;KACZ,CAAC,CAAC;IAEH,IAAI,OAAe,CAAC;IACpB,IAAI,CAAC;QACH,OAAO,GAAG,kBAAkB,CAAC,QAAQ,EAAE,IAAI,EAAE,GAAG,CAAC,QAAQ,CAAC,CAAC;IAC7D,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,MAAM,CAAC,IAAI,EAAE,GAAG,EAAE,EAAE,GAAG,WAAW,EAAE,EAAE,KAAK,EAAE,uBAAuB,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IAC5F,CAAC;IAED,wEAAwE;IACxE,IAAI,CAAC;QACH,MAAM,KAAK,GAAG,QAAQ,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QACtC,IAAI,KAAK,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC;YAC3B,OAAO,MAAM,CAAC,IAAI,EAAE,GAAG,EAAE;gBACvB,GAAG,WAAW,EAAE;gBAChB,KAAK,EAAE,sBAAsB,KAAK,CAAC,QAAQ,GAAG;aAC/C,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,MAAM,CAAC,IAAI,EAAE,GAAG,EAAE,EAAE,GAAG,WAAW,EAAE,EAAE,KAAK,EAAE,gBAAgB,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IACrF,CAAC;IAED,oEAAoE;IACpE,IAAI,KAAK,CAAC;IACV,IAAI,CAAC;QACH,KAAK,GAAG,MAAM,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,CAAC,KAAK,EAAE,CAAC,CAAC;IACtF,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,MAAM,CAAC,IAAI,EAAE,GAAG,EAAE,EAAE,GAAG,WAAW,EAAE,EAAE,KAAK,EAAE,kBAAkB,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IACvF,CAAC;IAED,4DAA4D;IAC5D,IAAI,OAAO,GAAG,KAAK,CAAC;IACpB,IAAI,SAAS,GAAkB,KAAK,CAAC,KAAK,CAAC;IAC3C,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,SAAS,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QACnC,OAAO,GAAG,CAAC,CAAC,MAAM,CAAC;QACnB,IAAI,CAAC,OAAO,IAAI,CAAC,SAAS,EAAE,CAAC;YAC3B,SAAS,GAAG,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,kBAAkB,CAAC,CAAC,CAAC,uBAAuB,CAAC,CAAC,QAAQ,GAAG,CAAC;QACrF,CAAC;IACH,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,SAAS,GAAG,iBAAiB,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;IAC3C,CAAC;IAED,MAAM,OAAO,GAAiB;QAC5B,OAAO;QACP,MAAM,EAAE,KAAK,CAAC,MAAM;QACpB,OAAO,EAAE,KAAK,CAAC,OAAO;QACtB,KAAK,EAAE,KAAK,CAAC,KAAK;QAClB,SAAS,EAAE,KAAK,CAAC,SAAS;QAC1B,SAAS,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,EAAE;QAC1B,UAAU,EAAE,KAAK,CAAC,UAAU;QAC5B,KAAK,EAAE,SAAS;KACjB,CAAC;IACF,OAAO,MAAM,CAAC,IAAI,EAAE,GAAG,EAAE,OAAO,CAAC,CAAC;AACpC,CAAC;AAED,SAAS,MAAM,CAAC,IAAU,EAAE,GAAiB,EAAE,OAAqB;IAClE,OAAO;QACL,MAAM,EAAE,IAAI,CAAC,IAAI,CAAC,EAAE;QACpB,SAAS,EAAE,IAAI,CAAC,SAAS,CAAC,KAAK;QAC/B,IAAI,EAAE,IAAI,CAAC,IAAI;QACf,OAAO;QACP,OAAO,EAAE,GAAG,CAAC,OAAO,CAAC,EAAE;QACvB,KAAK,EAAE,GAAG,CAAC,KAAK;KACjB,CAAC;AACJ,CAAC;AAED,SAAS,MAAM,CAAC,CAAU;IACxB,OAAO,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;AACpD,CAAC;AAED,SAAS,MAAM;IACb,OAAO,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;AAClC,CAAC"}
|
package/dist/bin/cli.js
CHANGED
|
@@ -680,6 +680,8 @@ bench
|
|
|
680
680
|
.option('--rope-margin <x>', 'Practical-equivalence margin: a correctness delta within ±x (success-rate units) is a TIE, not a win', '0')
|
|
681
681
|
.option('--out <dir>', 'Artifact output directory (default: benchmark-results/paired-<ts>)')
|
|
682
682
|
.option('--json', 'Emit JSON to stdout instead of the Markdown report')
|
|
683
|
+
.option('--allow-no-signal', 'Exit 0 even when the run could not measure anything (default: exit 3, so a no-signal run cannot be mistaken for a null result)')
|
|
684
|
+
.option('--min-detectable-effect <x>', 'Smallest correctness delta the run must be able to resolve, in success-rate units (default 0.25)')
|
|
683
685
|
.action(async (options) => {
|
|
684
686
|
const cmd = await lazy.benchPaired();
|
|
685
687
|
await cmd(options);
|