veryfront 0.1.960 → 0.1.961
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/esm/deno.js +1 -1
- package/esm/src/agent/ag-ui/browser-encoder.d.ts.map +1 -1
- package/esm/src/agent/ag-ui/browser-encoder.js +19 -0
- package/esm/src/eval/model-comparison.d.ts.map +1 -1
- package/esm/src/eval/model-comparison.js +52 -15
- package/esm/src/utils/version-constant.d.ts +1 -1
- package/esm/src/utils/version-constant.js +1 -1
- package/package.json +1 -1
package/esm/deno.js
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"browser-encoder.d.ts","sourceRoot":"","sources":["../../../../src/src/agent/ag-ui/browser-encoder.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAEjD,8CAA8C;AAC9C,MAAM,MAAM,sBAAsB,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAAG;IAAE,IAAI,EAAE,MAAM,CAAA;CAAE,CAAC;AAEhF,mEAAmE;AACnE,MAAM,WAAW,8BAA8B;IAC7C,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,iBAAiB,CAAC,EAAE,MAAM,CAAC;IAC3B,wBAAwB,CAAC,EAAE,MAAM,CAAC;IAClC,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAC9B,eAAe,CAAC,EAAE,MAAM,CAAC;IACzB,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAC7B,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAC9B,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,eAAe,CAAC,EAAE,MAAM,CAAC;IACzB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,UAAU,CAAC,EAAE,SAAS,GAAG,SAAS,GAAG,SAAS,CAAC;IAC/C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,kBAAkB,CAAC,EAAE,UAAU,GAAG,SAAS,GAAG,SAAS,CAAC;CACzD;AAED,uCAAuC;AACvC,MAAM,WAAW,uBAAuB;IACtC,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,QAAQ,EAAE,OAAO,CAAC;IAClB,mBAAmB,EAAE,MAAM,GAAG,IAAI,CAAC;IACnC,gBAAgB,EAAE,MAAM,CAAC;IACzB,kBAAkB,EAAE,MAAM,GAAG,IAAI,CAAC;IAClC,cAAc,EAAE,MAAM,GAAG,IAAI,CAAC;IAC9B,SAAS,EAAE,MAAM,CAAC;IAClB,oBAAoB,EAAE,GAAG,CAAC,MAAM,CAAC,CAAC;IAClC,gBAAgB,EAAE,OAAO,CAAC;IAC1B,gBAAgB,EAAE,OAAO,CAAC;IAC1B,QAAQ,EAAE,8BAA8B,CAAC;CAC1C;AAED,+CAA+C;AAC/C,MAAM,WAAW,uBAAuB;IACtC,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CAClC;AAED,8CAA8C;AAC9C,wBAAgB,6BAA6B,IAAI,uBAAuB,CAcvE;AAsPD,yDAAyD;AACzD,wBAAgB,gCAAgC,CAC9C,QAAQ,EAAE,8BAA8B,GACvC,aAAa,GAAG,IAAI,
|
|
1
|
+
{"version":3,"file":"browser-encoder.d.ts","sourceRoot":"","sources":["../../../../src/src/agent/ag-ui/browser-encoder.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAEjD,8CAA8C;AAC9C,MAAM,MAAM,sBAAsB,GAAG,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAAG;IAAE,IAAI,EAAE,MAAM,CAAA;CAAE,CAAC;AAEhF,mEAAmE;AACnE,MAAM,WAAW,8BAA8B;IAC7C,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,iBAAiB,CAAC,EAAE,MAAM,CAAC;IAC3B,wBAAwB,CAAC,EAAE,MAAM,CAAC;IAClC,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAC9B,eAAe,CAAC,EAAE,MAAM,CAAC;IACzB,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAC7B,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAC9B,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,eAAe,CAAC,EAAE,MAAM,CAAC;IACzB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,UAAU,CAAC,EAAE,SAAS,GAAG,SAAS,GAAG,SAAS,CAAC;IAC/C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,kBAAkB,CAAC,EAAE,UAAU,GAAG,SAAS,GAAG,SAAS,CAAC;CACzD;AAED,uCAAuC;AACvC,MAAM,WAAW,uBAAuB;IACtC,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,QAAQ,EAAE,OAAO,CAAC;IAClB,mBAAmB,EAAE,MAAM,GAAG,IAAI,CAAC;IACnC,gBAAgB,EAAE,MAAM,CAAC;IACzB,kBAAkB,EAAE,MAAM,GAAG,IAAI,CAAC;IAClC,cAAc,EAAE,MAAM,GAAG,IAAI,CAAC;IAC9B,SAAS,EAAE,MAAM,CAAC;IAClB,oBAAoB,EAAE,GAAG,CAAC,MAAM,CAAC,CAAC;IAClC,gBAAgB,EAAE,OAAO,CAAC;IAC1B,gBAAgB,EAAE,OAAO,CAAC;IAC1B,QAAQ,EAAE,8BAA8B,CAAC;CAC1C;AAED,+CAA+C;AAC/C,MAAM,WAAW,uBAAuB;IACtC,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CAClC;AAED,8CAA8C;AAC9C,wBAAgB,6BAA6B,IAAI,uBAAuB,CAcvE;AAsPD,yDAAyD;AACzD,wBAAgB,gCAAgC,CAC9C,QAAQ,EAAE,8BAA8B,GACvC,aAAa,GAAG,IAAI,CAmGtB;AAoID,wDAAwD;AACxD,wBAAgB,wCAAwC,CACtD,KAAK,EAAE,uBAAuB,EAC9B,KAAK,EAAE,sBAAsB,GAC5B,uBAAuB,EAAE,CA+N3B;AAED,4CAA4C;AAC5C,wBAAgB,yBAAyB,CACvC,KAAK,EAAE,uBAAuB,EAC9B,QAAQ,EAAE,aAAa,GAAG,IAAI,GAC7B,uBAAuB,EAAE,CA8B3B"}
|
|
@@ -280,6 +280,25 @@ export function buildAgUiBrowserFinalizeResponse(metadata) {
|
|
|
280
280
|
...(typeof metadata.reasoningTokens === "number"
|
|
281
281
|
? { reasoningTokens: metadata.reasoningTokens }
|
|
282
282
|
: {}),
|
|
283
|
+
...(typeof metadata.billableInputTokens === "number"
|
|
284
|
+
? { billableInputTokens: metadata.billableInputTokens }
|
|
285
|
+
: {}),
|
|
286
|
+
...(typeof metadata.billableOutputTokens === "number"
|
|
287
|
+
? { billableOutputTokens: metadata.billableOutputTokens }
|
|
288
|
+
: {}),
|
|
289
|
+
...(typeof metadata.costUsd === "number" ? { costUsd: metadata.costUsd } : {}),
|
|
290
|
+
...(typeof metadata.providerCostUsd === "number"
|
|
291
|
+
? { providerCostUsd: metadata.providerCostUsd }
|
|
292
|
+
: {}),
|
|
293
|
+
...(typeof metadata.veryfrontChargeUsd === "number"
|
|
294
|
+
? { veryfrontChargeUsd: metadata.veryfrontChargeUsd }
|
|
295
|
+
: {}),
|
|
296
|
+
...(typeof metadata.veryfrontBilledUsd === "number"
|
|
297
|
+
? { veryfrontBilledUsd: metadata.veryfrontBilledUsd }
|
|
298
|
+
: {}),
|
|
299
|
+
...(typeof metadata.costCredits === "number" ? { costCredits: metadata.costCredits } : {}),
|
|
300
|
+
...(metadata.costSource ? { costSource: metadata.costSource } : {}),
|
|
301
|
+
...(metadata.usageCaptureStatus ? { usageCaptureStatus: metadata.usageCaptureStatus } : {}),
|
|
283
302
|
}
|
|
284
303
|
: undefined;
|
|
285
304
|
if (!usage && Object.keys(responseMetadata).length === 0) {
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"model-comparison.d.ts","sourceRoot":"","sources":["../../../src/src/eval/model-comparison.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAGV,mBAAmB,EAEnB,0BAA0B,EAE1B,UAAU,EACX,MAAM,YAAY,CAAC;
|
|
1
|
+
{"version":3,"file":"model-comparison.d.ts","sourceRoot":"","sources":["../../../src/src/eval/model-comparison.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAGV,mBAAmB,EAEnB,0BAA0B,EAE1B,UAAU,EACX,MAAM,YAAY,CAAC;AAwiBpB,oFAAoF;AACpF,wBAAgB,uBAAuB,CACrC,OAAO,EAAE,UAAU,EAAE,EACrB,OAAO,EAAE,0BAA0B,GAClC,mBAAmB,CAqBrB;AAsBD,gFAAgF;AAChF,wBAAgB,iCAAiC,CAAC,UAAU,EAAE,mBAAmB,GAAG,MAAM,CA0DzF"}
|
|
@@ -56,13 +56,49 @@ function comparisonCost(summary) {
|
|
|
56
56
|
return summary.veryfrontBilledUsd ?? summary.veryfrontChargeUsd ?? summary.costUsd ??
|
|
57
57
|
summary.providerCostUsd;
|
|
58
58
|
}
|
|
59
|
-
function
|
|
60
|
-
|
|
61
|
-
candidateSummary.veryfrontBilledUsd !== undefined
|
|
62
|
-
|
|
63
|
-
|
|
64
|
-
|
|
65
|
-
|
|
59
|
+
function defaultCostComparison(baselineSummary, candidateSummary) {
|
|
60
|
+
if (baselineSummary.veryfrontBilledUsd !== undefined &&
|
|
61
|
+
candidateSummary.veryfrontBilledUsd !== undefined) {
|
|
62
|
+
return {
|
|
63
|
+
baseline: baselineSummary.veryfrontBilledUsd,
|
|
64
|
+
candidate: candidateSummary.veryfrontBilledUsd,
|
|
65
|
+
label: "Veryfront billed cost",
|
|
66
|
+
};
|
|
67
|
+
}
|
|
68
|
+
if (baselineSummary.costSource === "gateway" &&
|
|
69
|
+
candidateSummary.costSource === "gateway" &&
|
|
70
|
+
baselineSummary.costCredits !== undefined &&
|
|
71
|
+
candidateSummary.costCredits !== undefined) {
|
|
72
|
+
return {
|
|
73
|
+
baseline: baselineSummary.costCredits,
|
|
74
|
+
candidate: candidateSummary.costCredits,
|
|
75
|
+
label: "Veryfront credits",
|
|
76
|
+
};
|
|
77
|
+
}
|
|
78
|
+
if (baselineSummary.veryfrontChargeUsd !== undefined &&
|
|
79
|
+
candidateSummary.veryfrontChargeUsd !== undefined) {
|
|
80
|
+
return {
|
|
81
|
+
baseline: baselineSummary.veryfrontChargeUsd,
|
|
82
|
+
candidate: candidateSummary.veryfrontChargeUsd,
|
|
83
|
+
label: "Veryfront metered cost",
|
|
84
|
+
};
|
|
85
|
+
}
|
|
86
|
+
if (baselineSummary.costUsd !== undefined && candidateSummary.costUsd !== undefined) {
|
|
87
|
+
return {
|
|
88
|
+
baseline: baselineSummary.costUsd,
|
|
89
|
+
candidate: candidateSummary.costUsd,
|
|
90
|
+
label: "cost",
|
|
91
|
+
};
|
|
92
|
+
}
|
|
93
|
+
if (baselineSummary.providerCostUsd !== undefined &&
|
|
94
|
+
candidateSummary.providerCostUsd !== undefined) {
|
|
95
|
+
return {
|
|
96
|
+
baseline: baselineSummary.providerCostUsd,
|
|
97
|
+
candidate: candidateSummary.providerCostUsd,
|
|
98
|
+
label: "provider cost",
|
|
99
|
+
};
|
|
100
|
+
}
|
|
101
|
+
return undefined;
|
|
66
102
|
}
|
|
67
103
|
function metricValue(summary, metric) {
|
|
68
104
|
switch (metric) {
|
|
@@ -295,16 +331,16 @@ function createCandidateDecision(input) {
|
|
|
295
331
|
}
|
|
296
332
|
return { decision: "needs-review", objectiveScore: objective.score, reasons };
|
|
297
333
|
}
|
|
298
|
-
const
|
|
334
|
+
const costComparison = defaultCostComparison(baselineSummary, candidateSummary);
|
|
335
|
+
const costImprovement = improvementPct(costComparison?.baseline, costComparison?.candidate);
|
|
299
336
|
const tokenImprovement = improvementPct(baselineSummary.totalTokens, candidateSummary.totalTokens);
|
|
300
337
|
const latencyImprovement = improvementPct(baselineSummary.p95Ms, candidateSummary.p95Ms);
|
|
338
|
+
if (costComparison !== undefined && costImprovement !== undefined && costImprovement < 0) {
|
|
339
|
+
reasons.push(`${costComparison.label} regressed by ${formatPct(Math.abs(costImprovement))}`);
|
|
340
|
+
return { decision: "needs-review", reasons };
|
|
341
|
+
}
|
|
301
342
|
if (meetsImprovement(costImprovement, input.options.minCostImprovementPct)) {
|
|
302
|
-
|
|
303
|
-
? "Veryfront billed cost"
|
|
304
|
-
: hasGatewayMeteredCost(baselineSummary, candidateSummary)
|
|
305
|
-
? "Veryfront metered cost"
|
|
306
|
-
: "cost";
|
|
307
|
-
reasons.push(`${costLabel} improved by ${formatPct(costImprovement)}`);
|
|
343
|
+
reasons.push(`${costComparison.label} improved by ${formatPct(costImprovement)}`);
|
|
308
344
|
return { decision: "promote-candidate", reasons };
|
|
309
345
|
}
|
|
310
346
|
if (meetsImprovement(tokenImprovement, input.options.minTokenImprovementPct)) {
|
|
@@ -323,7 +359,8 @@ function compareCandidate(candidate, baseline, options) {
|
|
|
323
359
|
const candidateSummary = modelSummary(candidate, options.baselineModel);
|
|
324
360
|
const baselineSummary = modelSummary(baseline, options.baselineModel);
|
|
325
361
|
const decision = createCandidateDecision({ candidate, baseline, options });
|
|
326
|
-
const
|
|
362
|
+
const costComparison = defaultCostComparison(baselineSummary, candidateSummary);
|
|
363
|
+
const costImprovement = improvementPct(costComparison?.baseline, costComparison?.candidate);
|
|
327
364
|
const tokenImprovement = improvementPct(baselineSummary.totalTokens, candidateSummary.totalTokens);
|
|
328
365
|
const latencyImprovement = improvementPct(baselineSummary.p95Ms, candidateSummary.p95Ms);
|
|
329
366
|
return {
|