@orq-ai/node 4.17.0 → 4.17.2
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/jsr.json +1 -1
- package/lib/config.d.ts +3 -3
- package/lib/config.js +3 -3
- package/lib/config.js.map +1 -1
- package/models/components/autorouterconfig.d.ts +21 -0
- package/models/components/autorouterconfig.d.ts.map +1 -1
- package/models/components/autorouterconfig.js.map +1 -1
- package/models/components/autorouterv2config.d.ts +54 -0
- package/models/components/autorouterv2config.d.ts.map +1 -1
- package/models/components/autorouterv2config.js.map +1 -1
- package/models/components/autorouterv2creatorconfig.d.ts +9 -0
- package/models/components/autorouterv2creatorconfig.d.ts.map +1 -1
- package/models/components/autorouterv2creatorconfig.js.map +1 -1
- package/models/components/autorouterv2effortconfig.d.ts +18 -0
- package/models/components/autorouterv2effortconfig.d.ts.map +1 -1
- package/models/components/autorouterv2effortconfig.js.map +1 -1
- package/models/components/budgetmatch.d.ts +1 -1
- package/models/components/classifyretryconfig.d.ts +4 -4
- package/models/components/classifyretryconfig.d.ts.map +1 -1
- package/models/components/classifyretryconfig.js +1 -1
- package/models/components/classifyretryconfig.js.map +1 -1
- package/models/components/embeddingretryconfig.d.ts +4 -4
- package/models/components/embeddingretryconfig.d.ts.map +1 -1
- package/models/components/embeddingretryconfig.js +1 -1
- package/models/components/embeddingretryconfig.js.map +1 -1
- package/models/components/evaluatorcalltimeout.d.ts +18 -0
- package/models/components/evaluatorcalltimeout.d.ts.map +1 -0
- package/models/components/evaluatorcalltimeout.js +54 -0
- package/models/components/evaluatorcalltimeout.js.map +1 -0
- package/models/components/evaluatorretryconfig.d.ts +20 -0
- package/models/components/evaluatorretryconfig.d.ts.map +1 -0
- package/models/components/evaluatorretryconfig.js +55 -0
- package/models/components/evaluatorretryconfig.js.map +1 -0
- package/models/components/index.d.ts +2 -0
- package/models/components/index.d.ts.map +1 -1
- package/models/components/index.js +2 -0
- package/models/components/index.js.map +1 -1
- package/models/components/invokeevaluatorrequest.d.ts +9 -0
- package/models/components/invokeevaluatorrequest.d.ts.map +1 -1
- package/models/components/invokeevaluatorrequest.js +4 -0
- package/models/components/invokeevaluatorrequest.js.map +1 -1
- package/models/components/loadbalancerconfig.d.ts +19 -1
- package/models/components/loadbalancerconfig.d.ts.map +1 -1
- package/models/components/loadbalancerconfig.js +12 -2
- package/models/components/loadbalancerconfig.js.map +1 -1
- package/models/components/reasoning.d.ts +9 -0
- package/models/components/reasoning.d.ts.map +1 -1
- package/models/components/reasoning.js +9 -1
- package/models/components/reasoning.js.map +1 -1
- package/models/components/reasoningparam.d.ts +10 -0
- package/models/components/reasoningparam.d.ts.map +1 -1
- package/models/components/reasoningparam.js +10 -1
- package/models/components/reasoningparam.js.map +1 -1
- package/models/components/reasoningpart.js +1 -1
- package/models/components/responseretryconfig.d.ts +4 -4
- package/models/components/responseretryconfig.d.ts.map +1 -1
- package/models/components/responseretryconfig.js +1 -1
- package/models/components/responseretryconfig.js.map +1 -1
- package/models/operations/createtool.js +4 -4
- package/models/operations/getalltools.js +4 -4
- package/models/operations/getv2humanevalsets.js +2 -2
- package/models/operations/getv2humanevalsetsid.js +2 -2
- package/models/operations/listdeploymentversions.js +1 -1
- package/models/operations/modelcreateawsbedrock.d.ts +6 -0
- package/models/operations/modelcreateawsbedrock.d.ts.map +1 -1
- package/models/operations/modelcreateawsbedrock.js.map +1 -1
- package/models/operations/modelupdateawsbedrock.d.ts +6 -0
- package/models/operations/modelupdateawsbedrock.d.ts.map +1 -1
- package/models/operations/modelupdateawsbedrock.js.map +1 -1
- package/models/operations/patchv2humanevalsetsid.js +2 -2
- package/models/operations/postv2humanevalsets.js +2 -2
- package/models/operations/retrievetool.js +4 -4
- package/models/operations/runagent.js +1 -1
- package/models/operations/streamrunagent.js +1 -1
- package/models/operations/updatetool.js +4 -4
- package/package.json +3 -3
- package/packages/orq-rc/src/lib/config.ts +2 -2
- package/packages/orq-rc/src/models/components/reasoningpart.ts +1 -1
- package/packages/orq-rc/src/models/operations/createtool.ts +4 -4
- package/packages/orq-rc/src/models/operations/getalltools.ts +4 -4
- package/packages/orq-rc/src/models/operations/getv2humanevalsets.ts +2 -2
- package/packages/orq-rc/src/models/operations/getv2humanevalsetsid.ts +2 -2
- package/packages/orq-rc/src/models/operations/listdeploymentversions.ts +1 -1
- package/packages/orq-rc/src/models/operations/patchv2humanevalsetsid.ts +2 -2
- package/packages/orq-rc/src/models/operations/postv2humanevalsets.ts +2 -2
- package/packages/orq-rc/src/models/operations/retrievetool.ts +4 -4
- package/packages/orq-rc/src/models/operations/runagent.ts +1 -1
- package/packages/orq-rc/src/models/operations/streamrunagent.ts +1 -1
- package/packages/orq-rc/src/models/operations/updatetool.ts +4 -4
- package/src/lib/config.ts +3 -3
- package/src/models/components/autorouterconfig.ts +21 -0
- package/src/models/components/autorouterv2config.ts +54 -0
- package/src/models/components/autorouterv2creatorconfig.ts +9 -0
- package/src/models/components/autorouterv2effortconfig.ts +18 -0
- package/src/models/components/budgetmatch.ts +1 -1
- package/src/models/components/classifyretryconfig.ts +5 -5
- package/src/models/components/embeddingretryconfig.ts +5 -5
- package/src/models/components/evaluatorcalltimeout.ts +42 -0
- package/src/models/components/evaluatorretryconfig.ts +45 -0
- package/src/models/components/index.ts +2 -0
- package/src/models/components/invokeevaluatorrequest.ts +19 -0
- package/src/models/components/loadbalancerconfig.ts +24 -2
- package/src/models/components/reasoning.ts +14 -0
- package/src/models/components/reasoningparam.ts +14 -0
- package/src/models/components/reasoningpart.ts +1 -1
- package/src/models/components/responseretryconfig.ts +5 -5
- package/src/models/operations/createtool.ts +4 -4
- package/src/models/operations/getalltools.ts +4 -4
- package/src/models/operations/getv2humanevalsets.ts +2 -2
- package/src/models/operations/getv2humanevalsetsid.ts +2 -2
- package/src/models/operations/listdeploymentversions.ts +1 -1
- package/src/models/operations/modelcreateawsbedrock.ts +6 -0
- package/src/models/operations/modelupdateawsbedrock.ts +6 -0
- package/src/models/operations/patchv2humanevalsetsid.ts +2 -2
- package/src/models/operations/postv2humanevalsets.ts +2 -2
- package/src/models/operations/retrievetool.ts +4 -4
- package/src/models/operations/runagent.ts +1 -1
- package/src/models/operations/streamrunagent.ts +1 -1
- package/src/models/operations/updatetool.ts +4 -4
|
@@ -21,14 +21,35 @@ import {
|
|
|
21
21
|
} from "./sessionaffinityconfig.js";
|
|
22
22
|
|
|
23
23
|
export type AutoRouterConfig = {
|
|
24
|
+
/**
|
|
25
|
+
* Catalog model whose benchmarks this private model copies to join Smart Router pools.
|
|
26
|
+
*/
|
|
24
27
|
benchmarkModel?: string | undefined;
|
|
28
|
+
/**
|
|
29
|
+
* Model used when the router classifies a request as simple.
|
|
30
|
+
*/
|
|
25
31
|
economicalModel?: string | undefined;
|
|
32
|
+
/**
|
|
33
|
+
* Identifier of the Smart Router configuration.
|
|
34
|
+
*/
|
|
26
35
|
id?: string | undefined;
|
|
36
|
+
/**
|
|
37
|
+
* Models the router chooses from, written as provider/model_id.
|
|
38
|
+
*/
|
|
27
39
|
models?: Array<string> | null | undefined;
|
|
40
|
+
/**
|
|
41
|
+
* Routing preference: quality, balanced, or cost.
|
|
42
|
+
*/
|
|
28
43
|
profile?: string | undefined;
|
|
29
44
|
sessionAffinity?: SessionAffinityConfig | undefined;
|
|
45
|
+
/**
|
|
46
|
+
* Model used when the router classifies a request as complex.
|
|
47
|
+
*/
|
|
30
48
|
strongModel?: string | undefined;
|
|
31
49
|
v2?: AutoRouterV2Config | undefined;
|
|
50
|
+
/**
|
|
51
|
+
* Version of the Smart Router configuration format.
|
|
52
|
+
*/
|
|
32
53
|
version?: string | undefined;
|
|
33
54
|
};
|
|
34
55
|
|
|
@@ -22,23 +22,77 @@ import {
|
|
|
22
22
|
|
|
23
23
|
export type AutoRouterV2Config = {
|
|
24
24
|
aaCreator?: AutoRouterV2CreatorConfig | undefined;
|
|
25
|
+
/**
|
|
26
|
+
* Benchmark scores reported by Artificial Analysis, keyed by evaluation.
|
|
27
|
+
*/
|
|
25
28
|
aaEvaluations?: { [k: string]: number | null } | undefined;
|
|
29
|
+
/**
|
|
30
|
+
* Whether the intelligence index is estimated rather than measured.
|
|
31
|
+
*/
|
|
26
32
|
aaIndexIsEstimated?: boolean | undefined;
|
|
33
|
+
/**
|
|
34
|
+
* Version of the intelligence index the score came from.
|
|
35
|
+
*/
|
|
27
36
|
aaIndexVersion?: string | undefined;
|
|
37
|
+
/**
|
|
38
|
+
* Whether Artificial Analysis classifies the model as a reasoning model.
|
|
39
|
+
*/
|
|
28
40
|
aaIsReasoning?: boolean | undefined;
|
|
41
|
+
/**
|
|
42
|
+
* Median output tokens per second measured by Artificial Analysis.
|
|
43
|
+
*/
|
|
29
44
|
aaMedianOutputTokensPerSecond?: number | undefined;
|
|
45
|
+
/**
|
|
46
|
+
* Median time to the first answer token measured by Artificial Analysis.
|
|
47
|
+
*/
|
|
30
48
|
aaMedianTimeToFirstAnswerToken?: number | undefined;
|
|
49
|
+
/**
|
|
50
|
+
* Median time to first token in seconds measured by Artificial Analysis.
|
|
51
|
+
*/
|
|
31
52
|
aaMedianTimeToFirstTokenSeconds?: number | undefined;
|
|
53
|
+
/**
|
|
54
|
+
* Artificial Analysis model identifier.
|
|
55
|
+
*/
|
|
32
56
|
aaModelId?: string | undefined;
|
|
57
|
+
/**
|
|
58
|
+
* Model name reported by Artificial Analysis.
|
|
59
|
+
*/
|
|
33
60
|
aaName?: string | undefined;
|
|
61
|
+
/**
|
|
62
|
+
* Token prices reported by Artificial Analysis, keyed by price type.
|
|
63
|
+
*/
|
|
34
64
|
aaPricing?: { [k: string]: number | null } | undefined;
|
|
65
|
+
/**
|
|
66
|
+
* Reasoning effort the benchmarked score applies to.
|
|
67
|
+
*/
|
|
35
68
|
aaReasoningEffort?: string | undefined;
|
|
69
|
+
/**
|
|
70
|
+
* Model release date reported by Artificial Analysis.
|
|
71
|
+
*/
|
|
36
72
|
aaReleaseDate?: string | undefined;
|
|
73
|
+
/**
|
|
74
|
+
* Artificial Analysis slug of the model.
|
|
75
|
+
*/
|
|
37
76
|
aaSlug?: string | undefined;
|
|
77
|
+
/**
|
|
78
|
+
* Reasoning effort applied when a request does not set one.
|
|
79
|
+
*/
|
|
38
80
|
defaultEffort?: string | undefined;
|
|
81
|
+
/**
|
|
82
|
+
* Whether the default effort has a benchmarked score.
|
|
83
|
+
*/
|
|
39
84
|
defaultEffortScored?: boolean | undefined;
|
|
85
|
+
/**
|
|
86
|
+
* Artificial Analysis intelligence index the router ranks the model by.
|
|
87
|
+
*/
|
|
40
88
|
intelligenceIndex: number;
|
|
89
|
+
/**
|
|
90
|
+
* Blended price per million tokens, in USD.
|
|
91
|
+
*/
|
|
41
92
|
price: number;
|
|
93
|
+
/**
|
|
94
|
+
* Benchmark data for each reasoning effort the model exposes.
|
|
95
|
+
*/
|
|
42
96
|
reasoningEfforts?: { [k: string]: AutoRouterV2EffortConfig } | undefined;
|
|
43
97
|
};
|
|
44
98
|
|
|
@@ -8,8 +8,17 @@ import { Result as SafeParseResult } from "../../types/fp.js";
|
|
|
8
8
|
import { SDKValidationError } from "../errors/sdkvalidationerror.js";
|
|
9
9
|
|
|
10
10
|
export type AutoRouterV2CreatorConfig = {
|
|
11
|
+
/**
|
|
12
|
+
* Artificial Analysis identifier of the creator.
|
|
13
|
+
*/
|
|
11
14
|
id: string;
|
|
15
|
+
/**
|
|
16
|
+
* Creator name reported by Artificial Analysis.
|
|
17
|
+
*/
|
|
12
18
|
name: string;
|
|
19
|
+
/**
|
|
20
|
+
* Artificial Analysis slug of the creator.
|
|
21
|
+
*/
|
|
13
22
|
slug?: string | undefined;
|
|
14
23
|
};
|
|
15
24
|
|
|
@@ -9,11 +9,29 @@ import { Result as SafeParseResult } from "../../types/fp.js";
|
|
|
9
9
|
import { SDKValidationError } from "../errors/sdkvalidationerror.js";
|
|
10
10
|
|
|
11
11
|
export type AutoRouterV2EffortConfig = {
|
|
12
|
+
/**
|
|
13
|
+
* Whether the score at this effort is estimated.
|
|
14
|
+
*/
|
|
12
15
|
aaIndexIsEstimated?: boolean | undefined;
|
|
16
|
+
/**
|
|
17
|
+
* Version of the intelligence index measured at this effort.
|
|
18
|
+
*/
|
|
13
19
|
aaIndexVersion?: string | undefined;
|
|
20
|
+
/**
|
|
21
|
+
* Whether the model runs as a reasoning model at this effort.
|
|
22
|
+
*/
|
|
14
23
|
aaIsReasoning?: boolean | undefined;
|
|
24
|
+
/**
|
|
25
|
+
* Reasoning effort label Artificial Analysis reports for this entry.
|
|
26
|
+
*/
|
|
15
27
|
aaReasoningEffort?: string | undefined;
|
|
28
|
+
/**
|
|
29
|
+
* Artificial Analysis slug of the model at this effort.
|
|
30
|
+
*/
|
|
16
31
|
aaSlug?: string | undefined;
|
|
32
|
+
/**
|
|
33
|
+
* Intelligence index measured at this effort.
|
|
34
|
+
*/
|
|
17
35
|
intelligenceIndex: number;
|
|
18
36
|
};
|
|
19
37
|
|
|
@@ -8,7 +8,7 @@ import { Result as SafeParseResult } from "../../types/fp.js";
|
|
|
8
8
|
import { SDKValidationError } from "../errors/sdkvalidationerror.js";
|
|
9
9
|
|
|
10
10
|
/**
|
|
11
|
-
* CEL expression that decides whether a budget applies to a request. Available variables: `model`, `provider`, `model_id`, `api_key`, `identity`, `project`, `metadata` (map), `headers` (map, lowercase keys).
|
|
11
|
+
* CEL expression that decides whether a budget applies to a request. Available variables: `model`, `provider`, `model_id`, `api_key`, `identity`, `project`, `metadata` (map), `headers` (map, lowercase keys). Required unless the budget is created from a `scope`, and rejected on update when left empty. A workspace-wide budget stores an empty expression, which matches every request in the workspace. Expressions are syntax-validated at write time.
|
|
12
12
|
*/
|
|
13
13
|
export type BudgetMatch = {
|
|
14
14
|
cel?: string | undefined;
|
|
@@ -7,19 +7,19 @@ import { remap as remap$ } from "../../lib/primitives.js";
|
|
|
7
7
|
|
|
8
8
|
export type ClassifyRetryConfig = {
|
|
9
9
|
/**
|
|
10
|
-
* Number of retries per model after the initial attempt (
|
|
10
|
+
* Number of retries per model after the initial attempt (0-5). Set 0 to disable retries. When retry is omitted, the gateway retries twice.
|
|
11
11
|
*/
|
|
12
12
|
count: number;
|
|
13
13
|
/**
|
|
14
|
-
* HTTP status codes that trigger retries, between 100 and 599. Retry-After is honored; otherwise the gateway
|
|
14
|
+
* HTTP status codes that trigger retries, between 100 and 599. Defaults to [408, 429, 500, 502, 503, 504, 529]. Retry-After is honored; otherwise the gateway backs off exponentially.
|
|
15
15
|
*/
|
|
16
|
-
onCodes
|
|
16
|
+
onCodes?: Array<number> | undefined;
|
|
17
17
|
};
|
|
18
18
|
|
|
19
19
|
/** @internal */
|
|
20
20
|
export type ClassifyRetryConfig$Outbound = {
|
|
21
21
|
count: number;
|
|
22
|
-
on_codes
|
|
22
|
+
on_codes?: Array<number> | undefined;
|
|
23
23
|
};
|
|
24
24
|
|
|
25
25
|
/** @internal */
|
|
@@ -29,7 +29,7 @@ export const ClassifyRetryConfig$outboundSchema: z.ZodType<
|
|
|
29
29
|
ClassifyRetryConfig
|
|
30
30
|
> = z.object({
|
|
31
31
|
count: z.number().int(),
|
|
32
|
-
onCodes: z.array(z.number().int()),
|
|
32
|
+
onCodes: z.array(z.number().int()).optional(),
|
|
33
33
|
}).transform((v) => {
|
|
34
34
|
return remap$(v, {
|
|
35
35
|
onCodes: "on_codes",
|
|
@@ -7,19 +7,19 @@ import { remap as remap$ } from "../../lib/primitives.js";
|
|
|
7
7
|
|
|
8
8
|
export type EmbeddingRetryConfig = {
|
|
9
9
|
/**
|
|
10
|
-
* Number of
|
|
10
|
+
* Number of retries per model after the initial attempt (0-5). Set 0 to disable retries. When retry is omitted, the gateway retries twice.
|
|
11
11
|
*/
|
|
12
12
|
count: number;
|
|
13
13
|
/**
|
|
14
|
-
* HTTP status codes that trigger retry
|
|
14
|
+
* HTTP status codes that trigger a retry. Defaults to [408, 429, 500, 502, 503, 504, 529].
|
|
15
15
|
*/
|
|
16
|
-
onCodes
|
|
16
|
+
onCodes?: Array<number> | null | undefined;
|
|
17
17
|
};
|
|
18
18
|
|
|
19
19
|
/** @internal */
|
|
20
20
|
export type EmbeddingRetryConfig$Outbound = {
|
|
21
21
|
count: number;
|
|
22
|
-
on_codes
|
|
22
|
+
on_codes?: Array<number> | null | undefined;
|
|
23
23
|
};
|
|
24
24
|
|
|
25
25
|
/** @internal */
|
|
@@ -29,7 +29,7 @@ export const EmbeddingRetryConfig$outboundSchema: z.ZodType<
|
|
|
29
29
|
EmbeddingRetryConfig
|
|
30
30
|
> = z.object({
|
|
31
31
|
count: z.number().int(),
|
|
32
|
-
onCodes: z.nullable(z.array(z.number().int())),
|
|
32
|
+
onCodes: z.nullable(z.array(z.number().int())).optional(),
|
|
33
33
|
}).transform((v) => {
|
|
34
34
|
return remap$(v, {
|
|
35
35
|
onCodes: "on_codes",
|
|
@@ -0,0 +1,42 @@
|
|
|
1
|
+
/*
|
|
2
|
+
* Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
|
|
3
|
+
*/
|
|
4
|
+
|
|
5
|
+
import * as z from "zod/v3";
|
|
6
|
+
import { remap as remap$ } from "../../lib/primitives.js";
|
|
7
|
+
|
|
8
|
+
export type EvaluatorCallTimeout = {
|
|
9
|
+
/**
|
|
10
|
+
* Maximum milliseconds per provider attempt (or execution for non-LLM evaluators).
|
|
11
|
+
*
|
|
12
|
+
* @remarks
|
|
13
|
+
* Omitted uses the server default (120000). The server caps each invocation at ten minutes.
|
|
14
|
+
*/
|
|
15
|
+
callTimeout?: number | undefined;
|
|
16
|
+
};
|
|
17
|
+
|
|
18
|
+
/** @internal */
|
|
19
|
+
export type EvaluatorCallTimeout$Outbound = {
|
|
20
|
+
call_timeout?: number | undefined;
|
|
21
|
+
};
|
|
22
|
+
|
|
23
|
+
/** @internal */
|
|
24
|
+
export const EvaluatorCallTimeout$outboundSchema: z.ZodType<
|
|
25
|
+
EvaluatorCallTimeout$Outbound,
|
|
26
|
+
z.ZodTypeDef,
|
|
27
|
+
EvaluatorCallTimeout
|
|
28
|
+
> = z.object({
|
|
29
|
+
callTimeout: z.number().int().optional(),
|
|
30
|
+
}).transform((v) => {
|
|
31
|
+
return remap$(v, {
|
|
32
|
+
callTimeout: "call_timeout",
|
|
33
|
+
});
|
|
34
|
+
});
|
|
35
|
+
|
|
36
|
+
export function evaluatorCallTimeoutToJSON(
|
|
37
|
+
evaluatorCallTimeout: EvaluatorCallTimeout,
|
|
38
|
+
): string {
|
|
39
|
+
return JSON.stringify(
|
|
40
|
+
EvaluatorCallTimeout$outboundSchema.parse(evaluatorCallTimeout),
|
|
41
|
+
);
|
|
42
|
+
}
|
|
@@ -0,0 +1,45 @@
|
|
|
1
|
+
/*
|
|
2
|
+
* Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT.
|
|
3
|
+
*/
|
|
4
|
+
|
|
5
|
+
import * as z from "zod/v3";
|
|
6
|
+
import { remap as remap$ } from "../../lib/primitives.js";
|
|
7
|
+
|
|
8
|
+
export type EvaluatorRetryConfig = {
|
|
9
|
+
/**
|
|
10
|
+
* Retries after the first attempt. Zero disables retries; omitted retry preserves existing policy.
|
|
11
|
+
*/
|
|
12
|
+
count?: number | undefined;
|
|
13
|
+
/**
|
|
14
|
+
* Omitted uses the gateway's default transient status codes.
|
|
15
|
+
*/
|
|
16
|
+
onCodes?: Array<number> | undefined;
|
|
17
|
+
};
|
|
18
|
+
|
|
19
|
+
/** @internal */
|
|
20
|
+
export type EvaluatorRetryConfig$Outbound = {
|
|
21
|
+
count?: number | undefined;
|
|
22
|
+
on_codes?: Array<number> | undefined;
|
|
23
|
+
};
|
|
24
|
+
|
|
25
|
+
/** @internal */
|
|
26
|
+
export const EvaluatorRetryConfig$outboundSchema: z.ZodType<
|
|
27
|
+
EvaluatorRetryConfig$Outbound,
|
|
28
|
+
z.ZodTypeDef,
|
|
29
|
+
EvaluatorRetryConfig
|
|
30
|
+
> = z.object({
|
|
31
|
+
count: z.number().int().optional(),
|
|
32
|
+
onCodes: z.array(z.number().int()).optional(),
|
|
33
|
+
}).transform((v) => {
|
|
34
|
+
return remap$(v, {
|
|
35
|
+
onCodes: "on_codes",
|
|
36
|
+
});
|
|
37
|
+
});
|
|
38
|
+
|
|
39
|
+
export function evaluatorRetryConfigToJSON(
|
|
40
|
+
evaluatorRetryConfig: EvaluatorRetryConfig,
|
|
41
|
+
): string {
|
|
42
|
+
return JSON.stringify(
|
|
43
|
+
EvaluatorRetryConfig$outboundSchema.parse(evaluatorRetryConfig),
|
|
44
|
+
);
|
|
45
|
+
}
|
|
@@ -229,6 +229,7 @@ export * from "./errorpart.js";
|
|
|
229
229
|
export * from "./errorstreamingevent.js";
|
|
230
230
|
export * from "./evaluationcontext.js";
|
|
231
231
|
export * from "./evaluationresult.js";
|
|
232
|
+
export * from "./evaluatorcalltimeout.js";
|
|
232
233
|
export * from "./evaluatorfilter.js";
|
|
233
234
|
export * from "./evaluatorref.js";
|
|
234
235
|
export * from "./evaluatorresponsefunction.js";
|
|
@@ -238,6 +239,7 @@ export * from "./evaluatorresponsellm.js";
|
|
|
238
239
|
export * from "./evaluatorresponsepython.js";
|
|
239
240
|
export * from "./evaluatorresponseragas.js";
|
|
240
241
|
export * from "./evaluatorresponsetypescript.js";
|
|
242
|
+
export * from "./evaluatorretryconfig.js";
|
|
241
243
|
export * from "./executionnamedstreamingevent.js";
|
|
242
244
|
export * from "./executionreviewedstreamingevent.js";
|
|
243
245
|
export * from "./executionreviewrequiredstreamingevent.js";
|
|
@@ -8,6 +8,16 @@ import {
|
|
|
8
8
|
EvaluationContext$Outbound,
|
|
9
9
|
EvaluationContext$outboundSchema,
|
|
10
10
|
} from "./evaluationcontext.js";
|
|
11
|
+
import {
|
|
12
|
+
EvaluatorCallTimeout,
|
|
13
|
+
EvaluatorCallTimeout$Outbound,
|
|
14
|
+
EvaluatorCallTimeout$outboundSchema,
|
|
15
|
+
} from "./evaluatorcalltimeout.js";
|
|
16
|
+
import {
|
|
17
|
+
EvaluatorRetryConfig,
|
|
18
|
+
EvaluatorRetryConfig$Outbound,
|
|
19
|
+
EvaluatorRetryConfig$outboundSchema,
|
|
20
|
+
} from "./evaluatorretryconfig.js";
|
|
11
21
|
|
|
12
22
|
/**
|
|
13
23
|
* Accepts two shapes. `context` names its fields after the template variables
|
|
@@ -71,6 +81,11 @@ export type InvokeEvaluatorRequest = {
|
|
|
71
81
|
* `context.variables`.
|
|
72
82
|
*/
|
|
73
83
|
variables?: { [k: string]: any } | undefined;
|
|
84
|
+
/**
|
|
85
|
+
* Overrides llm_eval provider retries for this invocation, including every jury judge.
|
|
86
|
+
*/
|
|
87
|
+
retry?: EvaluatorRetryConfig | undefined;
|
|
88
|
+
timeout?: EvaluatorCallTimeout | undefined;
|
|
74
89
|
};
|
|
75
90
|
|
|
76
91
|
/** @internal */
|
|
@@ -83,6 +98,8 @@ export type InvokeEvaluatorRequest$Outbound = {
|
|
|
83
98
|
retrievals?: Array<string> | undefined;
|
|
84
99
|
messages?: Array<{ [k: string]: any }> | undefined;
|
|
85
100
|
variables?: { [k: string]: any } | undefined;
|
|
101
|
+
retry?: EvaluatorRetryConfig$Outbound | undefined;
|
|
102
|
+
timeout?: EvaluatorCallTimeout$Outbound | undefined;
|
|
86
103
|
};
|
|
87
104
|
|
|
88
105
|
/** @internal */
|
|
@@ -99,6 +116,8 @@ export const InvokeEvaluatorRequest$outboundSchema: z.ZodType<
|
|
|
99
116
|
retrievals: z.array(z.string()).optional(),
|
|
100
117
|
messages: z.array(z.record(z.any())).optional(),
|
|
101
118
|
variables: z.record(z.any()).optional(),
|
|
119
|
+
retry: EvaluatorRetryConfig$outboundSchema.optional(),
|
|
120
|
+
timeout: EvaluatorCallTimeout$outboundSchema.optional(),
|
|
102
121
|
});
|
|
103
122
|
|
|
104
123
|
export function invokeEvaluatorRequestToJSON(
|
|
@@ -3,17 +3,39 @@
|
|
|
3
3
|
*/
|
|
4
4
|
|
|
5
5
|
import * as z from "zod/v3";
|
|
6
|
+
import { ClosedEnum } from "../../types/enums.js";
|
|
6
7
|
import {
|
|
7
8
|
LoadBalancerModelConfig,
|
|
8
9
|
LoadBalancerModelConfig$Outbound,
|
|
9
10
|
LoadBalancerModelConfig$outboundSchema,
|
|
10
11
|
} from "./loadbalancermodelconfig.js";
|
|
11
12
|
|
|
13
|
+
/**
|
|
14
|
+
* Strategy used to pick a model: weight_based splits traffic by weight, round_robin rotates through the list, latency_based favors the fastest model. An unrecognized value falls back to weight_based.
|
|
15
|
+
*/
|
|
16
|
+
export const LoadBalancerConfigType = {
|
|
17
|
+
WeightBased: "weight_based",
|
|
18
|
+
RoundRobin: "round_robin",
|
|
19
|
+
LatencyBased: "latency_based",
|
|
20
|
+
} as const;
|
|
21
|
+
/**
|
|
22
|
+
* Strategy used to pick a model: weight_based splits traffic by weight, round_robin rotates through the list, latency_based favors the fastest model. An unrecognized value falls back to weight_based.
|
|
23
|
+
*/
|
|
24
|
+
export type LoadBalancerConfigType = ClosedEnum<typeof LoadBalancerConfigType>;
|
|
25
|
+
|
|
12
26
|
export type LoadBalancerConfig = {
|
|
13
27
|
models: Array<LoadBalancerModelConfig> | null;
|
|
14
|
-
|
|
28
|
+
/**
|
|
29
|
+
* Strategy used to pick a model: weight_based splits traffic by weight, round_robin rotates through the list, latency_based favors the fastest model. An unrecognized value falls back to weight_based.
|
|
30
|
+
*/
|
|
31
|
+
type: LoadBalancerConfigType;
|
|
15
32
|
};
|
|
16
33
|
|
|
34
|
+
/** @internal */
|
|
35
|
+
export const LoadBalancerConfigType$outboundSchema: z.ZodNativeEnum<
|
|
36
|
+
typeof LoadBalancerConfigType
|
|
37
|
+
> = z.nativeEnum(LoadBalancerConfigType);
|
|
38
|
+
|
|
17
39
|
/** @internal */
|
|
18
40
|
export type LoadBalancerConfig$Outbound = {
|
|
19
41
|
models: Array<LoadBalancerModelConfig$Outbound> | null;
|
|
@@ -27,7 +49,7 @@ export const LoadBalancerConfig$outboundSchema: z.ZodType<
|
|
|
27
49
|
LoadBalancerConfig
|
|
28
50
|
> = z.object({
|
|
29
51
|
models: z.nullable(z.array(LoadBalancerModelConfig$outboundSchema)),
|
|
30
|
-
type:
|
|
52
|
+
type: LoadBalancerConfigType$outboundSchema,
|
|
31
53
|
});
|
|
32
54
|
|
|
33
55
|
export function loadBalancerConfigToJSON(
|
|
@@ -8,6 +8,13 @@ import { ClosedEnum } from "../../types/enums.js";
|
|
|
8
8
|
import { Result as SafeParseResult } from "../../types/fp.js";
|
|
9
9
|
import { SDKValidationError } from "../errors/sdkvalidationerror.js";
|
|
10
10
|
|
|
11
|
+
export const ReasoningContext = {
|
|
12
|
+
Auto: "auto",
|
|
13
|
+
CurrentTurn: "current_turn",
|
|
14
|
+
AllTurns: "all_turns",
|
|
15
|
+
} as const;
|
|
16
|
+
export type ReasoningContext = ClosedEnum<typeof ReasoningContext>;
|
|
17
|
+
|
|
11
18
|
/**
|
|
12
19
|
* Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
|
|
13
20
|
*/
|
|
@@ -40,6 +47,7 @@ export const ReasoningSummary = {
|
|
|
40
47
|
export type ReasoningSummary = ClosedEnum<typeof ReasoningSummary>;
|
|
41
48
|
|
|
42
49
|
export type Reasoning = {
|
|
50
|
+
context?: ReasoningContext | undefined;
|
|
43
51
|
/**
|
|
44
52
|
* Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
|
|
45
53
|
*/
|
|
@@ -50,6 +58,11 @@ export type Reasoning = {
|
|
|
50
58
|
summary?: ReasoningSummary | null | undefined;
|
|
51
59
|
};
|
|
52
60
|
|
|
61
|
+
/** @internal */
|
|
62
|
+
export const ReasoningContext$inboundSchema: z.ZodNativeEnum<
|
|
63
|
+
typeof ReasoningContext
|
|
64
|
+
> = z.nativeEnum(ReasoningContext);
|
|
65
|
+
|
|
53
66
|
/** @internal */
|
|
54
67
|
export const ReasoningEffort1$inboundSchema: z.ZodNativeEnum<
|
|
55
68
|
typeof ReasoningEffort1
|
|
@@ -66,6 +79,7 @@ export const Reasoning$inboundSchema: z.ZodType<
|
|
|
66
79
|
z.ZodTypeDef,
|
|
67
80
|
unknown
|
|
68
81
|
> = z.object({
|
|
82
|
+
context: ReasoningContext$inboundSchema.optional(),
|
|
69
83
|
effort: ReasoningEffort1$inboundSchema.optional(),
|
|
70
84
|
summary: z.nullable(ReasoningSummary$inboundSchema).optional(),
|
|
71
85
|
});
|
|
@@ -5,6 +5,13 @@
|
|
|
5
5
|
import * as z from "zod/v3";
|
|
6
6
|
import { ClosedEnum } from "../../types/enums.js";
|
|
7
7
|
|
|
8
|
+
export const Context = {
|
|
9
|
+
Auto: "auto",
|
|
10
|
+
CurrentTurn: "current_turn",
|
|
11
|
+
AllTurns: "all_turns",
|
|
12
|
+
} as const;
|
|
13
|
+
export type Context = ClosedEnum<typeof Context>;
|
|
14
|
+
|
|
8
15
|
/**
|
|
9
16
|
* Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
|
|
10
17
|
*/
|
|
@@ -37,6 +44,7 @@ export const Summary = {
|
|
|
37
44
|
export type Summary = ClosedEnum<typeof Summary>;
|
|
38
45
|
|
|
39
46
|
export type ReasoningParam = {
|
|
47
|
+
context?: Context | undefined;
|
|
40
48
|
/**
|
|
41
49
|
* Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
|
|
42
50
|
*/
|
|
@@ -47,6 +55,10 @@ export type ReasoningParam = {
|
|
|
47
55
|
summary?: Summary | undefined;
|
|
48
56
|
};
|
|
49
57
|
|
|
58
|
+
/** @internal */
|
|
59
|
+
export const Context$outboundSchema: z.ZodNativeEnum<typeof Context> = z
|
|
60
|
+
.nativeEnum(Context);
|
|
61
|
+
|
|
50
62
|
/** @internal */
|
|
51
63
|
export const Effort$outboundSchema: z.ZodNativeEnum<typeof Effort> = z
|
|
52
64
|
.nativeEnum(Effort);
|
|
@@ -57,6 +69,7 @@ export const Summary$outboundSchema: z.ZodNativeEnum<typeof Summary> = z
|
|
|
57
69
|
|
|
58
70
|
/** @internal */
|
|
59
71
|
export type ReasoningParam$Outbound = {
|
|
72
|
+
context?: string | undefined;
|
|
60
73
|
effort?: string | undefined;
|
|
61
74
|
summary?: string | undefined;
|
|
62
75
|
};
|
|
@@ -67,6 +80,7 @@ export const ReasoningParam$outboundSchema: z.ZodType<
|
|
|
67
80
|
z.ZodTypeDef,
|
|
68
81
|
ReasoningParam
|
|
69
82
|
> = z.object({
|
|
83
|
+
context: Context$outboundSchema.optional(),
|
|
70
84
|
effort: Effort$outboundSchema.optional(),
|
|
71
85
|
summary: Summary$outboundSchema.optional(),
|
|
72
86
|
});
|
|
@@ -40,7 +40,7 @@ export const ReasoningPart$inboundSchema: z.ZodType<
|
|
|
40
40
|
z.ZodTypeDef,
|
|
41
41
|
unknown
|
|
42
42
|
> = z.object({
|
|
43
|
-
_id: z.string().default("
|
|
43
|
+
_id: z.string().default("reasoning_01m5mt1kds4xgeb87g19tvwmz9"),
|
|
44
44
|
metadata: z.record(z.any()).optional(),
|
|
45
45
|
kind: z.literal("reasoning"),
|
|
46
46
|
reasoning: z.string(),
|
|
@@ -7,19 +7,19 @@ import { remap as remap$ } from "../../lib/primitives.js";
|
|
|
7
7
|
|
|
8
8
|
export type ResponseRetryConfig = {
|
|
9
9
|
/**
|
|
10
|
-
* Number of retries (
|
|
10
|
+
* Number of retries per model after the initial attempt (0-5). Set 0 to disable retries. When retry is omitted, the gateway retries twice.
|
|
11
11
|
*/
|
|
12
12
|
count: number;
|
|
13
13
|
/**
|
|
14
|
-
* HTTP status codes that trigger a retry
|
|
14
|
+
* HTTP status codes that trigger a retry. Defaults to [408, 429, 500, 502, 503, 504, 529].
|
|
15
15
|
*/
|
|
16
|
-
onCodes
|
|
16
|
+
onCodes?: Array<number> | null | undefined;
|
|
17
17
|
};
|
|
18
18
|
|
|
19
19
|
/** @internal */
|
|
20
20
|
export type ResponseRetryConfig$Outbound = {
|
|
21
21
|
count: number;
|
|
22
|
-
on_codes
|
|
22
|
+
on_codes?: Array<number> | null | undefined;
|
|
23
23
|
};
|
|
24
24
|
|
|
25
25
|
/** @internal */
|
|
@@ -29,7 +29,7 @@ export const ResponseRetryConfig$outboundSchema: z.ZodType<
|
|
|
29
29
|
ResponseRetryConfig
|
|
30
30
|
> = z.object({
|
|
31
31
|
count: z.number().int(),
|
|
32
|
-
onCodes: z.nullable(z.array(z.number().int())),
|
|
32
|
+
onCodes: z.nullable(z.array(z.number().int())).optional(),
|
|
33
33
|
}).transform((v) => {
|
|
34
34
|
return remap$(v, {
|
|
35
35
|
onCodes: "on_codes",
|
|
@@ -1575,7 +1575,7 @@ export const ResponseBodyCodeExecutionTool$inboundSchema: z.ZodType<
|
|
|
1575
1575
|
z.ZodTypeDef,
|
|
1576
1576
|
unknown
|
|
1577
1577
|
> = z.object({
|
|
1578
|
-
_id: z.string().default("
|
|
1578
|
+
_id: z.string().default("tool_01M5MT1KH8F83053MVCPY7Z9E1"),
|
|
1579
1579
|
path: z.string(),
|
|
1580
1580
|
key: z.string(),
|
|
1581
1581
|
display_name: z.string().optional(),
|
|
@@ -1761,7 +1761,7 @@ export const ResponseBodyHTTPTool$inboundSchema: z.ZodType<
|
|
|
1761
1761
|
z.ZodTypeDef,
|
|
1762
1762
|
unknown
|
|
1763
1763
|
> = z.object({
|
|
1764
|
-
_id: z.string().default("
|
|
1764
|
+
_id: z.string().default("tool_01M5MT1KH7BENT5RGQ3JYW851J"),
|
|
1765
1765
|
path: z.string(),
|
|
1766
1766
|
key: z.string(),
|
|
1767
1767
|
display_name: z.string().optional(),
|
|
@@ -1854,7 +1854,7 @@ export const ResponseBodyJSONSchemaTool$inboundSchema: z.ZodType<
|
|
|
1854
1854
|
z.ZodTypeDef,
|
|
1855
1855
|
unknown
|
|
1856
1856
|
> = z.object({
|
|
1857
|
-
_id: z.string().default("
|
|
1857
|
+
_id: z.string().default("tool_01M5MT1KH6TVKH3BTGX8BKJ2WR"),
|
|
1858
1858
|
path: z.string(),
|
|
1859
1859
|
key: z.string(),
|
|
1860
1860
|
display_name: z.string().optional(),
|
|
@@ -1955,7 +1955,7 @@ export const ResponseBodyFunctionTool$inboundSchema: z.ZodType<
|
|
|
1955
1955
|
z.ZodTypeDef,
|
|
1956
1956
|
unknown
|
|
1957
1957
|
> = z.object({
|
|
1958
|
-
_id: z.string().default("
|
|
1958
|
+
_id: z.string().default("tool_01M5MT1KH6GVBQ058X7AN9Z3P9"),
|
|
1959
1959
|
path: z.string(),
|
|
1960
1960
|
key: z.string(),
|
|
1961
1961
|
display_name: z.string().optional(),
|
|
@@ -639,7 +639,7 @@ export const DataCodeExecutionTool$inboundSchema: z.ZodType<
|
|
|
639
639
|
z.ZodTypeDef,
|
|
640
640
|
unknown
|
|
641
641
|
> = z.object({
|
|
642
|
-
_id: z.string().default("
|
|
642
|
+
_id: z.string().default("tool_01M5MT1KH23K7NZFT2N9C5VYH6"),
|
|
643
643
|
path: z.string(),
|
|
644
644
|
key: z.string(),
|
|
645
645
|
display_name: z.string().optional(),
|
|
@@ -821,7 +821,7 @@ export const DataHTTPTool$inboundSchema: z.ZodType<
|
|
|
821
821
|
z.ZodTypeDef,
|
|
822
822
|
unknown
|
|
823
823
|
> = z.object({
|
|
824
|
-
_id: z.string().default("
|
|
824
|
+
_id: z.string().default("tool_01M5MT1KH00SD9WX7MHSAXGMS1"),
|
|
825
825
|
path: z.string(),
|
|
826
826
|
key: z.string(),
|
|
827
827
|
display_name: z.string().optional(),
|
|
@@ -914,7 +914,7 @@ export const DataJSONSchemaTool$inboundSchema: z.ZodType<
|
|
|
914
914
|
z.ZodTypeDef,
|
|
915
915
|
unknown
|
|
916
916
|
> = z.object({
|
|
917
|
-
_id: z.string().default("
|
|
917
|
+
_id: z.string().default("tool_01M5MT1KGZCSBWF6PP4F8J1BR9"),
|
|
918
918
|
path: z.string(),
|
|
919
919
|
key: z.string(),
|
|
920
920
|
display_name: z.string().optional(),
|
|
@@ -1012,7 +1012,7 @@ export const DataFunctionTool$inboundSchema: z.ZodType<
|
|
|
1012
1012
|
z.ZodTypeDef,
|
|
1013
1013
|
unknown
|
|
1014
1014
|
> = z.object({
|
|
1015
|
-
_id: z.string().default("
|
|
1015
|
+
_id: z.string().default("tool_01M5MT1KGZSHTQ48QFJF8NT1NT"),
|
|
1016
1016
|
path: z.string(),
|
|
1017
1017
|
key: z.string(),
|
|
1018
1018
|
display_name: z.string().optional(),
|
|
@@ -140,7 +140,7 @@ export const GetV2HumanEvalSetsResponseBody2$inboundSchema: z.ZodType<
|
|
|
140
140
|
created: z.string().datetime({ offset: true }).transform(v => new Date(v))
|
|
141
141
|
.optional(),
|
|
142
142
|
updated: z.string().datetime({ offset: true }).default(
|
|
143
|
-
"2026-10-
|
|
143
|
+
"2026-10-23T16:38:22.266Z",
|
|
144
144
|
).transform(v => new Date(v)),
|
|
145
145
|
filter_type: z.literal("name"),
|
|
146
146
|
filter_value: z.string(),
|
|
@@ -185,7 +185,7 @@ export const GetV2HumanEvalSetsResponseBody1$inboundSchema: z.ZodType<
|
|
|
185
185
|
created: z.string().datetime({ offset: true }).transform(v => new Date(v))
|
|
186
186
|
.optional(),
|
|
187
187
|
updated: z.string().datetime({ offset: true }).default(
|
|
188
|
-
"2026-10-
|
|
188
|
+
"2026-10-23T16:38:22.266Z",
|
|
189
189
|
).transform(v => new Date(v)),
|
|
190
190
|
filter_type: z.literal("span_type"),
|
|
191
191
|
filter_values: z.array(z.string()),
|