LLM compliance tests¶
The following endpoints outline how to manage AI robustness tests.
Create cost metric configuration¶
Operation path: POST /api/v2/genai/costMetricConfigurations/
Authentication requirements: BearerAuth
Create a new cost metric configuration.
Body parameter¶
{
"description": "The body of the \"Create cost metric configuration\" request.",
"properties": {
"costMetricConfigurations": {
"description": "The list of cost metric configurations to use.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
},
"playgroundId": {
"description": "The ID of the playground to associate with the cost metric configuration.",
"title": "playgroundId",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case to associate with the cost metric configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"useCaseId",
"playgroundId",
"costMetricConfigurations"
],
"title": "CreateCostMetricConfigurationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateCostMetricConfigurationRequest | true | none |
Example responses¶
201 Response
{
"description": "API response object for a single cost metric configuration.",
"properties": {
"costConfigurationId": {
"description": "The ID of the cost metric configuration.",
"title": "costConfigurationId",
"type": "string"
},
"costMetricConfigurations": {
"description": "The list of individual LLM cost configurations that constitute this cost metric configuration.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the cost metric configuration.",
"title": "playgroundId"
},
"useCaseId": {
"description": "The ID of the use case associated with the cost metric configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"costConfigurationId",
"useCaseId",
"costMetricConfigurations"
],
"title": "CostMetricConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 201 | Created | Cost configuration created successfully | CostMetricConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete cost metric configuration by cost metric configuration ID¶
Operation path: DELETE /api/v2/genai/costMetricConfigurations/{costMetricConfigurationId}/
Authentication requirements: BearerAuth
Delete an existing cost metric configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| costMetricConfigurationId | path | string | true | The ID of the cost metric configuration to delete. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Cost metric configuration successfully deleted. | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Retrieve cost metric configuration by cost metric configuration ID¶
Operation path: GET /api/v2/genai/costMetricConfigurations/{costMetricConfigurationId}/
Authentication requirements: BearerAuth
Retrieve an existing cost metric configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| costMetricConfigurationId | path | string | true | The ID of the cost metric configuration to retrieve. |
Example responses¶
200 Response
{
"description": "API response object for a single cost metric configuration.",
"properties": {
"costConfigurationId": {
"description": "The ID of the cost metric configuration.",
"title": "costConfigurationId",
"type": "string"
},
"costMetricConfigurations": {
"description": "The list of individual LLM cost configurations that constitute this cost metric configuration.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the cost metric configuration.",
"title": "playgroundId"
},
"useCaseId": {
"description": "The ID of the use case associated with the cost metric configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"costConfigurationId",
"useCaseId",
"costMetricConfigurations"
],
"title": "CostMetricConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Cost metric configuration successfully retrieved. | CostMetricConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Edit cost metric configuration by cost metric configuration ID¶
Operation path: PATCH /api/v2/genai/costMetricConfigurations/{costMetricConfigurationId}/
Authentication requirements: BearerAuth
Edit an existing cost metric configuration.
Body parameter¶
{
"description": "The body of the \"Edit cost metric configuration\" request.",
"properties": {
"costMetricConfigurations": {
"description": "The list of LLM cost configurations to apply to this cost metric configuration.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"minItems": 1,
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
}
},
"required": [
"costMetricConfigurations"
],
"title": "EditCostMetricConfigurationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| costMetricConfigurationId | path | string | true | The ID of the cost metric configuration to edit. |
| body | body | EditCostMetricConfigurationRequest | true | none |
Example responses¶
200 Response
{
"description": "API response object for a single cost metric configuration.",
"properties": {
"costConfigurationId": {
"description": "The ID of the cost metric configuration.",
"title": "costConfigurationId",
"type": "string"
},
"costMetricConfigurations": {
"description": "The list of individual LLM cost configurations that constitute this cost metric configuration.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the cost metric configuration.",
"title": "playgroundId"
},
"useCaseId": {
"description": "The ID of the use case associated with the cost metric configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"costConfigurationId",
"useCaseId",
"costMetricConfigurations"
],
"title": "CostMetricConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Cost metric configuration successfully updated. | CostMetricConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List evaluation dataset configurations¶
Operation path: GET /api/v2/genai/evaluationDatasetConfigurations/
Authentication requirements: BearerAuth
List evaluation dataset configurations.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| useCaseId | query | string | true | Only retrieve the evaluation dataset configurations associated with this use case ID. |
| playgroundId | query | string | true | Only retrieve the evaluation dataset configuration associated with this playground ID. |
| evaluationDatasetConfigurationId | query | any | false | Only retrieve the evaluation dataset configuration with this ID. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| search | query | any | false | Only retrieve the evaluation dataset configurations matching the search query. |
| sort | query | any | false | Apply this sort order to the results. Valid options are "name", "creationUserId", "creationDate", "datasetId", "userName", "datasetName", "promptColumnName", "responseColumnName". Prefix the attribute name with a dash to sort in descending order, e.g., sort=-creationDate. |
| correctnessEnabledOnly | query | boolean | false | If true, only retrieve the evaluation dataset configurations with correctness enabled. The default is false. |
| completedOnly | query | boolean | false | If true, only retrieve the evaluation dataset configurations where the evaluation dataset is in the completed status. The default is false. |
Example responses¶
200 Response
{
"description": "Paginated list of evaludation dataset configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single evaluation dataset configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "EvaluationDatasetConfigurationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Evaluation dataset configurations successfully retrieved. | ListEvaluationDatasetConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Create evaluation dataset configuration¶
Operation path: POST /api/v2/genai/evaluationDatasetConfigurations/
Authentication requirements: BearerAuth
Create a new evaluation dataset configuration.
Body parameter¶
{
"description": "The body of the \"Create evaluation dataset configuration\" request.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals. It is required to evaluate the AgentGoalAccuracyWithReference metric for agentic workflows.",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"isSyntheticDataset": {
"default": false,
"description": "Whether the evaluation dataset is synthetic.",
"title": "isSyntheticDataset",
"type": "boolean"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the evaluation dataset configuration.",
"title": "name"
},
"playgroundId": {
"description": "The ID of the playground to associate with the evaluation dataset configuration.",
"title": "playgroundId",
"type": "string"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"toolCallsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls. It is required to evaluate the ToolCallAccuracy metric for agentic workflows.",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case to associate with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"useCaseId",
"playgroundId",
"datasetId",
"promptColumnName"
],
"title": "CreateEvaluationDatasetConfigurationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateEvaluationDatasetConfigurationRequest | true | none |
Example responses¶
201 Response
{
"description": "API response object for a single evaluation dataset configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "EvaluationDatasetConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 201 | Created | Evaluation dataset configuration successfully created | EvaluationDatasetConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete evaluation dataset configuration by evaluation dataset configuration ID¶
Operation path: DELETE /api/v2/genai/evaluationDatasetConfigurations/{evaluationDatasetConfigurationId}/
Authentication requirements: BearerAuth
Delete an existing evaluation dataset configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | path | string | true | The ID of the evaluation dataset configuration to delete. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Evaluation dataset configuration successfully deleted. | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Retrieve evaluation dataset configuration by evaluation dataset configuration ID¶
Operation path: GET /api/v2/genai/evaluationDatasetConfigurations/{evaluationDatasetConfigurationId}/
Authentication requirements: BearerAuth
Retrieve an existing evaluation dataset configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | path | string | true | The ID of the evaluation dataset configuration to retrieve. |
Example responses¶
200 Response
{
"description": "API response object for a single evaluation dataset configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "EvaluationDatasetConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Evaluation dataset configuration successfully retrieved. | EvaluationDatasetConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Edit evaluation dataset configuration by evaluation dataset configuration ID¶
Operation path: PATCH /api/v2/genai/evaluationDatasetConfigurations/{evaluationDatasetConfigurationId}/
Authentication requirements: BearerAuth
Edit an existing evaluation dataset configuration.
Body parameter¶
{
"description": "The body of the \"Edit evaluation dataset configuration\" request.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the expected name of the dataset column containing expected agent goals. It is required to evaluate the AgentGoalAccuracyWithReference metric for agentic workflows.",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "If specified, enables or disables correctness for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"datasetId": {
"default": "000000000000000000000000",
"description": "If specified, updates the ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, renames the evaluation dataset configuration to this value.",
"title": "name"
},
"promptColumnName": {
"default": "None",
"description": "If specified, changes the expected name of the dataset column containing the prompt text.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the expected name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"toolCallsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the expected name of the dataset column containing expected tool calls. It is required to evaluate the ToolCallAccuracy metric for agentic workflows.",
"title": "toolCallsColumnName"
}
},
"title": "EditEvaluationDatasetConfigurationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | path | string | true | The ID of the evaluation dataset configuration to edit. |
| body | body | EditEvaluationDatasetConfigurationRequest | true | none |
Example responses¶
200 Response
{
"description": "API response object for a single evaluation dataset configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "EvaluationDatasetConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Evaluation dataset configuration successfully updated. | EvaluationDatasetConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete evaluation dataset metric aggregation¶
Operation path: DELETE /api/v2/genai/evaluationDatasetMetricAggregations/
Authentication requirements: BearerAuth
Delete the evaluation dataset metric aggregation associated with the specified LLM blueprint IDs and/or chat IDs.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmBlueprintIds | query | any | false | The IDs of the LLM blueprints to delete the associated evaluation dataset metric aggregation for. If both llmBlueprintIds and chatIds are specified, will delete the aggregation record only if it matches both criteria. |
| chatIds | query | any | false | The IDs of the chats to delete the associated evaluation dataset metric aggregation for. If both llmBlueprintIds and chatIds are specified, will delete the aggregation record only if it matches both criteria. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Evaluation dataset metric aggregation successfully deleted. | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List evaluation dataset metric aggregations¶
Operation path: GET /api/v2/genai/evaluationDatasetMetricAggregations/
Authentication requirements: BearerAuth
List evaluation dataset metric aggregations.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmBlueprintIds | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these LLM blueprint IDs. |
| chatIds | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these chat IDs. |
| evaluationDatasetConfigurationIds | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these evaluation dataset configuration IDs. |
| metricNames | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these metric names. |
| aggregationTypes | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these aggregation types. |
| currentConfigurationOnly | query | boolean | false | Only retrieve the evaluation dataset metric aggregations associated with the current configuration of the llmblueprints. |
| sort | query | any | false | Apply this sort order to the results. Valid options are "name", "creationUserId", "creationDate", "datasetId", "userName", "datasetName", "promptColumnName", "responseColumnName". Prefix the attribute name with a dash to sort in descending order, e.g., sort=-creationDate. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| nonErroredOnly | query | boolean | false | If true, only retrieve the evaluation dataset metric aggregations that are in a non-errored status. The default is false. |
Example responses¶
200 Response
{
"description": "Paginated list of evaluation dataset metric aggregations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single evaluation dataset metric aggregation.",
"properties": {
"aggregationType": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"items": {
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregated value of the metric.",
"title": "aggregationValue"
},
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
},
"customModelGuardId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model's guard the metric aggregation belongs to.",
"title": "customModelGuardId"
},
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The dataset ID of the evaluation dataset configuration.",
"title": "datasetId"
},
"datasetName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The Data Registry dataset name of the evaluation dataset configuration.",
"title": "datasetName"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration associated with the metric aggregation.",
"title": "evaluationDatasetConfigurationId"
},
"llmBlueprintId": {
"description": "The ID of the LLM blueprint associated with the metric aggregation.",
"title": "llmBlueprintId",
"type": "string"
},
"metricName": {
"description": "The name of the metric associated with the metric aggregation.",
"title": "metricName",
"type": "string"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the out-of-the-box dataset."
},
"tenantId": {
"description": "The ID of the tenant the metric aggregation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName",
"llmBlueprintId",
"evaluationDatasetConfigurationId",
"ootbDatasetName",
"datasetId",
"datasetName",
"metricName",
"aggregationValue",
"aggregationType",
"tenantId",
"customModelGuardId"
],
"title": "EvaluationDatasetMetricAggregationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetMetricAggregationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Evaluation dataset metric aggregations successfully retrieved. | ListEvaluationDatasetMetricAggregationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Create evaluation dataset metric aggregation¶
Operation path: POST /api/v2/genai/evaluationDatasetMetricAggregations/
Authentication requirements: BearerAuth
Create a new evaluation dataset metric aggregation.
Body parameter¶
{
"description": "The body of the \"Create evaluation dataset metric aggregation\" request.",
"properties": {
"chatName": {
"default": "Aggregated chat",
"description": "The name for the new chat that will contain the associated prompts and responses.",
"maxLength": 5000,
"title": "chatName",
"type": "string"
},
"evaluationDatasetConfigurationId": {
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId",
"type": "string"
},
"insightsConfiguration": {
"description": "The configuration of insights for the metric aggregation.",
"items": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"minItems": 1,
"title": "insightsConfiguration",
"type": "array"
},
"llmBlueprintIds": {
"description": "The IDs of the LLM blueprints to use for the metric aggregation.",
"items": {
"type": "string"
},
"maxItems": 3,
"minItems": 1,
"title": "llmBlueprintIds",
"type": "array"
}
},
"required": [
"llmBlueprintIds",
"evaluationDatasetConfigurationId",
"insightsConfiguration"
],
"title": "CreateEvaluationDatasetMetricAggregationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateEvaluationDatasetMetricAggregationRequest | true | none |
Example responses¶
202 Response
{
"description": "The body of the \"Create evaluation dataset metric aggregation\" response.",
"properties": {
"chatIds": {
"description": "The IDs of the chats associated with the metric aggregation.",
"items": {
"type": "string"
},
"title": "chatIds",
"type": "array"
},
"jobId": {
"description": "The ID of the evaluation dataset metric aggregation job.",
"format": "uuid4",
"title": "jobId",
"type": "string"
}
},
"required": [
"jobId",
"chatIds"
],
"title": "CreateEvaluationDatasetMetricAggregationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 202 | Accepted | Evaluation dataset metric aggregation job successfully accepted. Follow the Location header to poll for job execution status. |
CreateEvaluationDatasetMetricAggregationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List evaluation dataset metric aggregations aggregated by llm blueprint¶
Operation path: GET /api/v2/genai/evaluationDatasetMetricAggregations/aggregateByLLMBlueprint/
Authentication requirements: BearerAuth
List evaluation dataset metric aggregations aggregated by llm blueprint.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmBlueprintIds | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these LLM blueprint IDs. |
| chatIds | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these chat IDs. |
| evaluationDatasetConfigurationIds | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these evaluation dataset configuration IDs. |
| metricNames | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these metric names. |
| aggregationTypes | query | any | false | Only retrieve the evaluation dataset metric aggregations associated with these aggregation types. |
| currentConfigurationOnly | query | boolean | false | Only retrieve the evaluation dataset metric aggregations associated with the current configuration of the llmblueprints. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| nonErroredOnly | query | boolean | false | If true, only retrieve the evaluation dataset metric aggregations that are in a non-errored status. The default is false. |
Example responses¶
200 Response
{
"description": "Paginated list of evaluation dataset metric aggregations, aggregated by LLM blueprint.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for multiple evaluation dataset metric aggregation\naggregated by llm blueprint.",
"properties": {
"aggregatedItemCount": {
"description": "Number of items aggregated.",
"title": "aggregatedItemCount",
"type": "integer"
},
"aggregatedItemDetails": {
"description": "List of details for aggregated items.",
"items": {
"description": "Details for aggregated items.",
"properties": {
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "EvaluationDatasetMetricAggregationChatDetails",
"type": "object"
},
"title": "aggregatedItemDetails",
"type": "array"
},
"aggregationType": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"items": {
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregated value of the metric.",
"title": "aggregationValue"
},
"customModelGuardId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model's guard the metric aggregation belongs to.",
"title": "customModelGuardId"
},
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The dataset ID of the evaluation dataset configuration.",
"title": "datasetId"
},
"datasetName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The Data Registry dataset name of the evaluation dataset configuration.",
"title": "datasetName"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration associated with the metric aggregation.",
"title": "evaluationDatasetConfigurationId"
},
"llmBlueprintId": {
"description": "The ID of the LLM blueprint associated with the metric aggregation.",
"title": "llmBlueprintId",
"type": "string"
},
"metricName": {
"description": "The name of the metric associated with the metric aggregation.",
"title": "metricName",
"type": "string"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the out-of-the-box dataset."
},
"tenantId": {
"description": "The ID of the tenant the metric aggregation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
}
},
"required": [
"llmBlueprintId",
"evaluationDatasetConfigurationId",
"ootbDatasetName",
"datasetId",
"datasetName",
"metricName",
"aggregationValue",
"aggregationType",
"tenantId",
"customModelGuardId",
"aggregatedItemDetails",
"aggregatedItemCount"
],
"title": "EvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Evaluation dataset metric aggregations aggregated by llm blueprint successfully retrieved. | ListEvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List evaluation dataset metric aggregations unique computed metrics by uniquefield¶
Operation path: GET /api/v2/genai/evaluationDatasetMetricAggregations/uniqueFieldValues/{uniqueField}/
Authentication requirements: BearerAuth
List evaluation dataset metric aggregations unique computed metrics.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| uniqueField | path | EvaluationDatasetMetricAggregationFieldQueryParam | true | Retrieve the list of this unique field. |
| llmBlueprintIds | query | any | false | Only retrieve the list of the unique field associated with these LLM blueprint IDs. |
| metricNames | query | any | false | Only retrieve the list of the unique field associated with these metric names. |
| chatIds | query | any | false | Only retrieve the list of the unique field associated with these chat IDs. |
| evaluationDatasetConfigurationIds | query | any | false | Only retrieve the list of the unique field associated with these evaluation dataset configuration IDs. |
| aggregationTypes | query | any | false | Only retrieve the list of the unique field associated with these aggregation types. |
| currentConfigurationOnly | query | boolean | false | Only retrieve the evaluation dataset metric aggregations associated with the current configuration of the llmblueprints. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| nonErroredOnly | query | boolean | false | If true, only retrieve the list of the unique field for aggregation records that are in a non-errored status. The default is false. |
Enumerated Values¶
| Parameter | Value |
|---|---|
| uniqueField | [metricName, llmBlueprintId, aggregationType, evaluationDatasetConfigurationId] |
Example responses¶
200 Response
{
"description": "Paginated list of evaluation dataset metric aggregations with unique computed metrics.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single unique computed metric.",
"properties": {
"uniqueFieldValue": {
"description": "The unique value associated with the metric aggregation.",
"title": "uniqueFieldValue",
"type": "string"
}
},
"required": [
"uniqueFieldValue"
],
"title": "EvaluationDatasetMetricAggregationUniqueFieldValuesResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetMetricAggregationUniqueFieldValuesResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Evaluation dataset metric aggregations unique computed metrics successfully retrieved. | ListEvaluationDatasetMetricAggregationUniqueFieldValuesResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List LLM test configuration¶
Operation path: GET /api/v2/genai/llmTestConfigurations/
Authentication requirements: BearerAuth
List LLM test configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| useCaseId | query | any | false | Use Case ID. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| testConfigType | query | LLMTestConfigurationType | false | Whether to return out-of-the-box (ootb) or custom LLM test configurations in the response. |
Enumerated Values¶
| Parameter | Value |
|---|---|
| testConfigType | [ootb, custom] |
Example responses¶
200 Response
{
"description": "Paginated list of LLM test configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single LLMTestConfiguration.",
"properties": {
"creationDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationDate"
},
"creationUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationUserId"
},
"datasetEvaluations": {
"description": "The LLM test dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
},
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"description": "The description of the LLM Test configuration.",
"title": "description",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the LLM test configuration.",
"title": "errorMessage"
},
"id": {
"description": "The ID of the LLM Test configuration.",
"title": "id",
"type": "string"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"lastUpdateDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateDate"
},
"lastUpdateUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateUserId"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "The name of the LLM Test configuration.",
"title": "name",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, the use case ID associated with the LLM Test configuration.",
"title": "useCaseId"
},
"warnings": {
"description": "Warnings for this LLM test configuration.",
"items": {
"additionalProperties": {
"type": "string"
},
"propertyNames": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"type": "object"
},
"title": "warnings",
"type": "array"
}
},
"required": [
"id",
"name",
"description",
"datasetEvaluations",
"llmTestGradingCriteria",
"isOutOfTheBoxTestConfiguration",
"warnings"
],
"title": "LLMTestConfigurationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestConfigurationsResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | ListLLMTestConfigurationsResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Create LLM test configuration¶
Operation path: POST /api/v2/genai/llmTestConfigurations/
Authentication requirements: BearerAuth
Create a new LLM test configuration.
Body parameter¶
{
"description": "Request object for creating a LLMTestConfiguration.",
"properties": {
"datasetEvaluations": {
"description": "Dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 0,
"description": "The max number of prompts to evaluate.",
"maximum": 5000,
"minimum": 0,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset name. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria"
],
"title": "DatasetEvaluationRequest",
"type": "object"
},
"maxItems": 10,
"minItems": 1,
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"default": "",
"description": "LLM test configuration description.",
"maxLength": 5000,
"title": "description",
"type": "string"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "LLM test configuration name.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The use case ID associated with the LLM Test configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"name",
"useCaseId",
"datasetEvaluations",
"llmTestGradingCriteria"
],
"title": "CreateLLMTestConfigurationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateLLMTestConfigurationRequest | true | none |
Example responses¶
201 Response
{
"description": "API response object for a single LLMTestConfiguration.",
"properties": {
"creationDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationDate"
},
"creationUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationUserId"
},
"datasetEvaluations": {
"description": "The LLM test dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
},
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"description": "The description of the LLM Test configuration.",
"title": "description",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the LLM test configuration.",
"title": "errorMessage"
},
"id": {
"description": "The ID of the LLM Test configuration.",
"title": "id",
"type": "string"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"lastUpdateDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateDate"
},
"lastUpdateUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateUserId"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "The name of the LLM Test configuration.",
"title": "name",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, the use case ID associated with the LLM Test configuration.",
"title": "useCaseId"
},
"warnings": {
"description": "Warnings for this LLM test configuration.",
"items": {
"additionalProperties": {
"type": "string"
},
"propertyNames": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"type": "object"
},
"title": "warnings",
"type": "array"
}
},
"required": [
"id",
"name",
"description",
"datasetEvaluations",
"llmTestGradingCriteria",
"isOutOfTheBoxTestConfiguration",
"warnings"
],
"title": "LLMTestConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 201 | Created | Successful Response | LLMTestConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List non out-of-the-box datasets¶
Operation path: GET /api/v2/genai/llmTestConfigurations/nonOotbDatasets/
Authentication requirements: BearerAuth
List the supported non out-of-the-box datasets that can be used with an LLM test configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| useCaseId | query | string | true | Use Case ID. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| search | query | any | false | Only retrieve the datasets with names matching the search query. |
Example responses¶
200 Response
{
"description": "Paginated list of non-OOTB datasets for use with LLM test configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "Non out-of-the-box dataset used with an LLM test configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "LLMTestConfigurationNonOOTBDatasetResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestConfigurationNonOOTBDatasetsResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Non out-of-the-box datasets successfully retrieved. | ListLLMTestConfigurationNonOOTBDatasetsResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List out-of-the-box datasets¶
Operation path: GET /api/v2/genai/llmTestConfigurations/ootbDatasets/
Authentication requirements: BearerAuth
List the supported out-of-the-box datasets that can be used with an LLM test configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| search | query | any | false | Only retrieve the datasets with names matching the search query. |
Example responses¶
200 Response
{
"description": "Paginated list of OOTB datasets for use with LLM test configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "Out-of-the-box dataset used with an LLM test configuration.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "LLMTestConfigurationOOTBDatasetResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestConfigurationOOTBDatasetsResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Out-of-the-box datasets successfully retrieved. | ListLLMTestConfigurationOOTBDatasetsResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List supported insights¶
Operation path: GET /api/v2/genai/llmTestConfigurations/supportedInsights/
Authentication requirements: BearerAuth
List the supported LLM test insight configurations for the specified use case.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| useCaseId | query | any | false | If specified, only retrieve the insights supported by this use case ID. |
| playgroundId | query | any | false | If specified, only retrieve the insights supported by the use case for which the playgroundId belongs. |
Example responses¶
200 Response
{
"description": "Response model for supported insights.",
"properties": {
"datasetsCompatibility": {
"description": "The list of insight to evaluation datasets compatibility.",
"items": {
"description": "Insight to evaluation datasets compatibility.",
"properties": {
"incompatibleDatasets": {
"description": "The list of incompatible datasets.",
"items": {
"description": "Dataset identifier.",
"properties": {
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the dataset, if any.",
"title": "datasetId"
},
"datasetName": {
"description": "The name of the dataset.",
"title": "datasetName",
"type": "string"
}
},
"required": [
"datasetName",
"datasetId"
],
"title": "DatasetIdentifier",
"type": "object"
},
"title": "incompatibleDatasets",
"type": "array"
},
"insightName": {
"description": "The name of the insight.",
"title": "insightName",
"type": "string"
}
},
"required": [
"insightName",
"incompatibleDatasets"
],
"title": "InsightToEvalDatasetsCompatibility",
"type": "object"
},
"title": "datasetsCompatibility",
"type": "array"
},
"supportedInsightConfigurations": {
"description": "The list of supported insight configurations for the LLM Tests.",
"items": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"title": "supportedInsightConfigurations",
"type": "array"
}
},
"required": [
"supportedInsightConfigurations",
"datasetsCompatibility"
],
"title": "LLMTestConfigurationSupportedInsightsResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | LLM test supported insight configurations successfully retrieved. | LLMTestConfigurationSupportedInsightsResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete LLM test configuration by LLM test configuration ID¶
Operation path: DELETE /api/v2/genai/llmTestConfigurations/{llmTestConfigurationId}/
Authentication requirements: BearerAuth
Delete an existing LLM test configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestConfigurationId | path | string | true | The ID of the LLM Test Configuration to delete. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Successful Response | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Retrieve LLM test configuration by LLM test configuration ID¶
Operation path: GET /api/v2/genai/llmTestConfigurations/{llmTestConfigurationId}/
Authentication requirements: BearerAuth
Retrieve an existing LLM test configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestConfigurationId | path | string | true | The ID of the LLM Test Configuration to retrieve. |
Example responses¶
200 Response
{
"description": "API response object for a single LLMTestConfiguration.",
"properties": {
"creationDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationDate"
},
"creationUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationUserId"
},
"datasetEvaluations": {
"description": "The LLM test dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
},
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"description": "The description of the LLM Test configuration.",
"title": "description",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the LLM test configuration.",
"title": "errorMessage"
},
"id": {
"description": "The ID of the LLM Test configuration.",
"title": "id",
"type": "string"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"lastUpdateDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateDate"
},
"lastUpdateUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateUserId"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "The name of the LLM Test configuration.",
"title": "name",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, the use case ID associated with the LLM Test configuration.",
"title": "useCaseId"
},
"warnings": {
"description": "Warnings for this LLM test configuration.",
"items": {
"additionalProperties": {
"type": "string"
},
"propertyNames": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"type": "object"
},
"title": "warnings",
"type": "array"
}
},
"required": [
"id",
"name",
"description",
"datasetEvaluations",
"llmTestGradingCriteria",
"isOutOfTheBoxTestConfiguration",
"warnings"
],
"title": "LLMTestConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | LLMTestConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Edit LLM test configuration by LLM test configuration ID¶
Operation path: PATCH /api/v2/genai/llmTestConfigurations/{llmTestConfigurationId}/
Authentication requirements: BearerAuth
Edit an existing LLM test configuration.
Body parameter¶
{
"description": "Request object for editing a LLMTestConfiguration.",
"properties": {
"datasetEvaluations": {
"anyOf": [
{
"items": {
"description": "Dataset evaluation.",
"properties": {
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 0,
"description": "The max number of prompts to evaluate.",
"maximum": 5000,
"minimum": 0,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset name. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria"
],
"title": "DatasetEvaluationRequest",
"type": "object"
},
"maxItems": 10,
"minItems": 1,
"type": "array"
},
{
"type": "null"
}
],
"description": "New Dataset evaluations.",
"title": "datasetEvaluations"
},
"description": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "New LLM test configuration description.",
"title": "description"
},
"llmTestGradingCriteria": {
"anyOf": [
{
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
{
"type": "null"
}
],
"description": "New LLM test grading criteria."
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "New LLM test configuration name.",
"title": "name"
}
},
"title": "EditLLMTestConfigurationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestConfigurationId | path | string | true | The ID of the LLM Test Configuration to update. |
| body | body | EditLLMTestConfigurationRequest | true | none |
Example responses¶
200 Response
{
"description": "API response object for a single LLMTestConfiguration.",
"properties": {
"creationDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationDate"
},
"creationUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationUserId"
},
"datasetEvaluations": {
"description": "The LLM test dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
},
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"description": "The description of the LLM Test configuration.",
"title": "description",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the LLM test configuration.",
"title": "errorMessage"
},
"id": {
"description": "The ID of the LLM Test configuration.",
"title": "id",
"type": "string"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"lastUpdateDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateDate"
},
"lastUpdateUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateUserId"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "The name of the LLM Test configuration.",
"title": "name",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, the use case ID associated with the LLM Test configuration.",
"title": "useCaseId"
},
"warnings": {
"description": "Warnings for this LLM test configuration.",
"items": {
"additionalProperties": {
"type": "string"
},
"propertyNames": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"type": "object"
},
"title": "warnings",
"type": "array"
}
},
"required": [
"id",
"name",
"description",
"datasetEvaluations",
"llmTestGradingCriteria",
"isOutOfTheBoxTestConfiguration",
"warnings"
],
"title": "LLMTestConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | LLMTestConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List LLM test results¶
Operation path: GET /api/v2/genai/llmTestResults/
Authentication requirements: BearerAuth
List LLM test results.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestConfigurationId | query | any | false | LLM Test Configuration ID. |
| llmBlueprintId | query | any | false | LLM Blueprint ID. |
| llmTestSuiteId | query | any | false | LLM Test Suite ID. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
Example responses¶
200 Response
{
"description": "Paginated list of LLM test results.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single LLMTestResult.",
"properties": {
"creationDate": {
"description": "LLM test result creation date (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "ID of the user that created this LLM test result.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user who created this LLM result.",
"title": "creationUserName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message if the LLM Test Result failed.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error resolution message if the LLM Test Result failed.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result based on the llm test grading criteria. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "LLM test result ID.",
"title": "id",
"type": "string"
},
"insightEvaluationResults": {
"description": "The Insight evaluation results.",
"items": {
"description": "API response object for a single InsightEvaluationResult.",
"properties": {
"aggregationType": {
"anyOf": [
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Aggregation type."
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "Aggregation value. None indicates that the aggregation failed.",
"title": "aggregationValue"
},
"chatId": {
"description": "Chat ID.",
"title": "chatId",
"type": "string"
},
"chatName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Chat name.",
"title": "chatName"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Custom Model LLM Validation ID if using custom model LLM.",
"title": "customModelLLMValidationId"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset configuration ID.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "Evaluation name.",
"maxLength": 5000,
"title": "evaluationName",
"type": "string"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "Insight evaluation result ID.",
"title": "id",
"type": "string"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"lastUpdateDate": {
"description": "Last update date of the insight evaluation result (ISO 8601 formatted).",
"format": "date-time",
"title": "lastUpdateDate",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM ID used for this insight evaluation result.",
"title": "llmId"
},
"llmTestResultId": {
"description": "LLM test result ID this insight evaluation result is associated to.",
"title": "llmTestResultId",
"type": "string"
},
"maxNumPrompts": {
"description": "Number of prompts used in evaluation.",
"title": "maxNumPrompts",
"type": "integer"
},
"metricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the metric.",
"title": "metricName"
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"id",
"llmTestResultId",
"maxNumPrompts",
"promptSamplingStrategy",
"chatId",
"chatName",
"evaluationName",
"insightGradingCriteria",
"lastUpdateDate"
],
"title": "InsightEvaluationResultResponse",
"type": "object"
},
"title": "insightEvaluationResults",
"type": "array"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"llmBlueprintId": {
"description": "LLM Blueprint ID.",
"title": "llmBlueprintId",
"type": "string"
},
"llmBlueprintSnapshot": {
"description": "A snapshot in time of a LLMBlueprint's functional parameters.",
"properties": {
"description": {
"description": "The description of the LLMBlueprint at the time of snapshotting.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLMBlueprint for which the snapshot was produced.",
"title": "id",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM selected for this LLM blueprint.",
"title": "llmId"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of LLM settings.",
"title": "llmSettings"
},
"name": {
"description": "The name of the LLMBlueprint at the time of snapshotting.",
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The playground id of the LLMBlueprint.",
"title": "playgroundId",
"type": "string"
},
"promptType": {
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
},
"snapshotDate": {
"description": "The date when the snapshot was produced.",
"format": "date-time",
"title": "snapshotDate",
"type": "string"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database linked to this LLM blueprint.",
"title": "vectorDatabaseId"
},
"vectorDatabaseSettings": {
"anyOf": [
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of vector database settings."
}
},
"required": [
"id",
"name",
"description",
"playgroundId",
"promptType"
],
"title": "LLMBlueprintSnapshot",
"type": "object"
},
"llmTestConfigurationId": {
"description": "LLM test configuration ID this LLM result is associated to.",
"title": "llmTestConfigurationId",
"type": "string"
},
"llmTestConfigurationName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the LLM test configuration this LLM result is associated to.",
"title": "llmTestConfigurationName"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"llmTestSuiteId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM test suite ID to which the LLM test configuration is associated to.",
"title": "llmTestSuiteId"
},
"passPercentage": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "The percentage of underlying insight evaluation results that have a PASS grading result. If not specified, execution status is not COMPLETED.",
"title": "passPercentage"
},
"useCaseId": {
"description": "Use case ID this LLM test result belongs to.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"llmTestConfigurationId",
"llmTestConfigurationName",
"isOutOfTheBoxTestConfiguration",
"useCaseId",
"llmBlueprintId",
"llmBlueprintSnapshot",
"llmTestGradingCriteria",
"executionStatus",
"insightEvaluationResults",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "LLMTestResultResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestResultResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | ListLLMTestResultResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Create LLM test result¶
Operation path: POST /api/v2/genai/llmTestResults/
Authentication requirements: BearerAuth
Create a new LLM test result.
Body parameter¶
{
"description": "Request object for creating a LLMTestResult.",
"properties": {
"llmBlueprintId": {
"description": "The LLM Blueprint ID associated with the LLM Test result.",
"title": "llmBlueprintId",
"type": "string"
},
"llmTestConfigurationId": {
"description": "The use case ID associated with the LLM Test result.",
"title": "llmTestConfigurationId",
"type": "string"
}
},
"required": [
"llmTestConfigurationId",
"llmBlueprintId"
],
"title": "CreateLLMTestResultRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateLLMTestResultRequest | true | none |
Example responses¶
202 Response
{
"description": "API response object for a single LLMTestResult.",
"properties": {
"creationDate": {
"description": "LLM test result creation date (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "ID of the user that created this LLM test result.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user who created this LLM result.",
"title": "creationUserName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message if the LLM Test Result failed.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error resolution message if the LLM Test Result failed.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result based on the llm test grading criteria. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "LLM test result ID.",
"title": "id",
"type": "string"
},
"insightEvaluationResults": {
"description": "The Insight evaluation results.",
"items": {
"description": "API response object for a single InsightEvaluationResult.",
"properties": {
"aggregationType": {
"anyOf": [
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Aggregation type."
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "Aggregation value. None indicates that the aggregation failed.",
"title": "aggregationValue"
},
"chatId": {
"description": "Chat ID.",
"title": "chatId",
"type": "string"
},
"chatName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Chat name.",
"title": "chatName"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Custom Model LLM Validation ID if using custom model LLM.",
"title": "customModelLLMValidationId"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset configuration ID.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "Evaluation name.",
"maxLength": 5000,
"title": "evaluationName",
"type": "string"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "Insight evaluation result ID.",
"title": "id",
"type": "string"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"lastUpdateDate": {
"description": "Last update date of the insight evaluation result (ISO 8601 formatted).",
"format": "date-time",
"title": "lastUpdateDate",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM ID used for this insight evaluation result.",
"title": "llmId"
},
"llmTestResultId": {
"description": "LLM test result ID this insight evaluation result is associated to.",
"title": "llmTestResultId",
"type": "string"
},
"maxNumPrompts": {
"description": "Number of prompts used in evaluation.",
"title": "maxNumPrompts",
"type": "integer"
},
"metricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the metric.",
"title": "metricName"
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"id",
"llmTestResultId",
"maxNumPrompts",
"promptSamplingStrategy",
"chatId",
"chatName",
"evaluationName",
"insightGradingCriteria",
"lastUpdateDate"
],
"title": "InsightEvaluationResultResponse",
"type": "object"
},
"title": "insightEvaluationResults",
"type": "array"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"llmBlueprintId": {
"description": "LLM Blueprint ID.",
"title": "llmBlueprintId",
"type": "string"
},
"llmBlueprintSnapshot": {
"description": "A snapshot in time of a LLMBlueprint's functional parameters.",
"properties": {
"description": {
"description": "The description of the LLMBlueprint at the time of snapshotting.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLMBlueprint for which the snapshot was produced.",
"title": "id",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM selected for this LLM blueprint.",
"title": "llmId"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of LLM settings.",
"title": "llmSettings"
},
"name": {
"description": "The name of the LLMBlueprint at the time of snapshotting.",
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The playground id of the LLMBlueprint.",
"title": "playgroundId",
"type": "string"
},
"promptType": {
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
},
"snapshotDate": {
"description": "The date when the snapshot was produced.",
"format": "date-time",
"title": "snapshotDate",
"type": "string"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database linked to this LLM blueprint.",
"title": "vectorDatabaseId"
},
"vectorDatabaseSettings": {
"anyOf": [
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of vector database settings."
}
},
"required": [
"id",
"name",
"description",
"playgroundId",
"promptType"
],
"title": "LLMBlueprintSnapshot",
"type": "object"
},
"llmTestConfigurationId": {
"description": "LLM test configuration ID this LLM result is associated to.",
"title": "llmTestConfigurationId",
"type": "string"
},
"llmTestConfigurationName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the LLM test configuration this LLM result is associated to.",
"title": "llmTestConfigurationName"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"llmTestSuiteId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM test suite ID to which the LLM test configuration is associated to.",
"title": "llmTestSuiteId"
},
"passPercentage": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "The percentage of underlying insight evaluation results that have a PASS grading result. If not specified, execution status is not COMPLETED.",
"title": "passPercentage"
},
"useCaseId": {
"description": "Use case ID this LLM test result belongs to.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"llmTestConfigurationId",
"llmTestConfigurationName",
"isOutOfTheBoxTestConfiguration",
"useCaseId",
"llmBlueprintId",
"llmBlueprintSnapshot",
"llmTestGradingCriteria",
"executionStatus",
"insightEvaluationResults",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "LLMTestResultResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 202 | Accepted | Successful Response | LLMTestResultResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete LLM test result by LLM test result ID¶
Operation path: DELETE /api/v2/genai/llmTestResults/{llmTestResultId}/
Authentication requirements: BearerAuth
Delete an existing LLM test result.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestResultId | path | string | true | The ID of the LLM Test Result to delete. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Successful Response | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Retrieve LLM test result by LLM test result ID¶
Operation path: GET /api/v2/genai/llmTestResults/{llmTestResultId}/
Authentication requirements: BearerAuth
Retrieve an existing LLM test result.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestResultId | path | string | true | The ID of the LLM Test Result to retrieve. |
Example responses¶
200 Response
{
"description": "API response object for a single LLMTestResult.",
"properties": {
"creationDate": {
"description": "LLM test result creation date (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "ID of the user that created this LLM test result.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user who created this LLM result.",
"title": "creationUserName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message if the LLM Test Result failed.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error resolution message if the LLM Test Result failed.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result based on the llm test grading criteria. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "LLM test result ID.",
"title": "id",
"type": "string"
},
"insightEvaluationResults": {
"description": "The Insight evaluation results.",
"items": {
"description": "API response object for a single InsightEvaluationResult.",
"properties": {
"aggregationType": {
"anyOf": [
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Aggregation type."
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "Aggregation value. None indicates that the aggregation failed.",
"title": "aggregationValue"
},
"chatId": {
"description": "Chat ID.",
"title": "chatId",
"type": "string"
},
"chatName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Chat name.",
"title": "chatName"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Custom Model LLM Validation ID if using custom model LLM.",
"title": "customModelLLMValidationId"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset configuration ID.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "Evaluation name.",
"maxLength": 5000,
"title": "evaluationName",
"type": "string"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "Insight evaluation result ID.",
"title": "id",
"type": "string"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"lastUpdateDate": {
"description": "Last update date of the insight evaluation result (ISO 8601 formatted).",
"format": "date-time",
"title": "lastUpdateDate",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM ID used for this insight evaluation result.",
"title": "llmId"
},
"llmTestResultId": {
"description": "LLM test result ID this insight evaluation result is associated to.",
"title": "llmTestResultId",
"type": "string"
},
"maxNumPrompts": {
"description": "Number of prompts used in evaluation.",
"title": "maxNumPrompts",
"type": "integer"
},
"metricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the metric.",
"title": "metricName"
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"id",
"llmTestResultId",
"maxNumPrompts",
"promptSamplingStrategy",
"chatId",
"chatName",
"evaluationName",
"insightGradingCriteria",
"lastUpdateDate"
],
"title": "InsightEvaluationResultResponse",
"type": "object"
},
"title": "insightEvaluationResults",
"type": "array"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"llmBlueprintId": {
"description": "LLM Blueprint ID.",
"title": "llmBlueprintId",
"type": "string"
},
"llmBlueprintSnapshot": {
"description": "A snapshot in time of a LLMBlueprint's functional parameters.",
"properties": {
"description": {
"description": "The description of the LLMBlueprint at the time of snapshotting.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLMBlueprint for which the snapshot was produced.",
"title": "id",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM selected for this LLM blueprint.",
"title": "llmId"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of LLM settings.",
"title": "llmSettings"
},
"name": {
"description": "The name of the LLMBlueprint at the time of snapshotting.",
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The playground id of the LLMBlueprint.",
"title": "playgroundId",
"type": "string"
},
"promptType": {
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
},
"snapshotDate": {
"description": "The date when the snapshot was produced.",
"format": "date-time",
"title": "snapshotDate",
"type": "string"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database linked to this LLM blueprint.",
"title": "vectorDatabaseId"
},
"vectorDatabaseSettings": {
"anyOf": [
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of vector database settings."
}
},
"required": [
"id",
"name",
"description",
"playgroundId",
"promptType"
],
"title": "LLMBlueprintSnapshot",
"type": "object"
},
"llmTestConfigurationId": {
"description": "LLM test configuration ID this LLM result is associated to.",
"title": "llmTestConfigurationId",
"type": "string"
},
"llmTestConfigurationName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the LLM test configuration this LLM result is associated to.",
"title": "llmTestConfigurationName"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"llmTestSuiteId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM test suite ID to which the LLM test configuration is associated to.",
"title": "llmTestSuiteId"
},
"passPercentage": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "The percentage of underlying insight evaluation results that have a PASS grading result. If not specified, execution status is not COMPLETED.",
"title": "passPercentage"
},
"useCaseId": {
"description": "Use case ID this LLM test result belongs to.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"llmTestConfigurationId",
"llmTestConfigurationName",
"isOutOfTheBoxTestConfiguration",
"useCaseId",
"llmBlueprintId",
"llmBlueprintSnapshot",
"llmTestGradingCriteria",
"executionStatus",
"insightEvaluationResults",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "LLMTestResultResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | LLMTestResultResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List LLM test suites¶
Operation path: GET /api/v2/genai/llmTestSuites/
Authentication requirements: BearerAuth
List LLM test suites.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| useCaseId | query | any | false | Only retrieve the LLM test suites associated with this use case ID. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| sort | query | any | false | Apply this sort order to the results. Valid options are "name" and "creationDate". Prefix the attribute name with a dash to sort in descending order, e.g., sort=-creationDate. |
Example responses¶
200 Response
{
"description": "Paginated list of LLM test suites.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "LLMTestSuite object formatted for API output.",
"properties": {
"creationDate": {
"description": "The creation date of the chat (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the chat.",
"title": "creationUserId",
"type": "string"
},
"description": {
"description": "The description of the LLM test suite.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLM test suite.",
"title": "id",
"type": "string"
},
"llmTestConfigurationIds": {
"description": "The IDs of the LLM test configurations in this LLM test suite.",
"items": {
"type": "string"
},
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case associated with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"name",
"description",
"useCaseId",
"llmTestConfigurationIds",
"creationDate",
"creationUserId"
],
"title": "LLMTestSuiteResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestSuitesResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | ListLLMTestSuitesResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Create LLM test suite¶
Operation path: POST /api/v2/genai/llmTestSuites/
Authentication requirements: BearerAuth
Create a new LLM test suite.
Body parameter¶
{
"description": "The body of the \"Create LLM test suite\" request.",
"properties": {
"description": {
"default": "",
"description": "The description of the LLM test suite.",
"maxLength": 5000,
"title": "description",
"type": "string"
},
"llmTestConfigurationIds": {
"default": [],
"description": "The IDs of the LLM test configurations in the LLM test suite.",
"items": {
"type": "string"
},
"maxItems": 100,
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case to associate with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"name",
"useCaseId"
],
"title": "CreateLLMTestSuiteRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateLLMTestSuiteRequest | true | none |
Example responses¶
201 Response
{
"description": "LLMTestSuite object formatted for API output.",
"properties": {
"creationDate": {
"description": "The creation date of the chat (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the chat.",
"title": "creationUserId",
"type": "string"
},
"description": {
"description": "The description of the LLM test suite.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLM test suite.",
"title": "id",
"type": "string"
},
"llmTestConfigurationIds": {
"description": "The IDs of the LLM test configurations in this LLM test suite.",
"items": {
"type": "string"
},
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case associated with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"name",
"description",
"useCaseId",
"llmTestConfigurationIds",
"creationDate",
"creationUserId"
],
"title": "LLMTestSuiteResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 201 | Created | Successful Response | LLMTestSuiteResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete LLM test suite by LLM test suite ID¶
Operation path: DELETE /api/v2/genai/llmTestSuites/{llmTestSuiteId}/
Authentication requirements: BearerAuth
Delete an existing LLM test suite.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestSuiteId | path | string | true | The ID of the LLM test suite to delete. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Successful Response | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Retrieve LLM test suite by LLM test suite ID¶
Operation path: GET /api/v2/genai/llmTestSuites/{llmTestSuiteId}/
Authentication requirements: BearerAuth
Retrieve an existing LLM test suite.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestSuiteId | path | string | true | The ID of the LLM test suite to retrieve. |
Example responses¶
200 Response
{
"description": "LLMTestSuite object formatted for API output.",
"properties": {
"creationDate": {
"description": "The creation date of the chat (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the chat.",
"title": "creationUserId",
"type": "string"
},
"description": {
"description": "The description of the LLM test suite.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLM test suite.",
"title": "id",
"type": "string"
},
"llmTestConfigurationIds": {
"description": "The IDs of the LLM test configurations in this LLM test suite.",
"items": {
"type": "string"
},
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case associated with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"name",
"description",
"useCaseId",
"llmTestConfigurationIds",
"creationDate",
"creationUserId"
],
"title": "LLMTestSuiteResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | LLMTestSuiteResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Edit LLM test suite by LLM test suite ID¶
Operation path: PATCH /api/v2/genai/llmTestSuites/{llmTestSuiteId}/
Authentication requirements: BearerAuth
Edit an existing LLM test suite.
Body parameter¶
{
"description": "The body of the \"Edit LLM test suite\" request.",
"properties": {
"description": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The description of the LLM test suite.",
"title": "description"
},
"llmTestConfigurationIds": {
"anyOf": [
{
"items": {
"type": "string"
},
"maxItems": 100,
"type": "array"
},
{
"type": "null"
}
],
"description": "The IDs of the LLM test configurations in the LLM test suite.",
"title": "llmTestConfigurationIds"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the LLM test suite.",
"title": "name"
}
},
"title": "EditLLMTestSuiteRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| llmTestSuiteId | path | string | true | The ID of the LLM test suite to edit. |
| body | body | EditLLMTestSuiteRequest | true | none |
Example responses¶
200 Response
{
"description": "LLMTestSuite object formatted for API output.",
"properties": {
"creationDate": {
"description": "The creation date of the chat (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the chat.",
"title": "creationUserId",
"type": "string"
},
"description": {
"description": "The description of the LLM test suite.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLM test suite.",
"title": "id",
"type": "string"
},
"llmTestConfigurationIds": {
"description": "The IDs of the LLM test configurations in this LLM test suite.",
"items": {
"type": "string"
},
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case associated with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"name",
"description",
"useCaseId",
"llmTestConfigurationIds",
"creationDate",
"creationUserId"
],
"title": "LLMTestSuiteResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Successful Response | LLMTestSuiteResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete OOTB metric configuration by ootb metric configuration ID¶
Operation path: DELETE /api/v2/genai/ootbMetricConfigurations/{ootbMetricConfigurationId}/
Authentication requirements: BearerAuth
Delete single OOTB metric configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| ootbMetricConfigurationId | path | string | true | The ID of the metric configuration. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | OOTB metric configuration successfully deleted. | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Get OOTB metric configuration by ootb metric configuration ID¶
Operation path: GET /api/v2/genai/ootbMetricConfigurations/{ootbMetricConfigurationId}/
Authentication requirements: BearerAuth
Get OOTB metric configuration from the configuration.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| ootbMetricConfigurationId | path | string | true | The ID of the metric configuration. |
Example responses¶
200 Response
{
"description": "API response object for a single OOTB metric.",
"properties": {
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"customOotbMetricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The custom OOTB metric name to be associated with the OOTB metric.",
"title": "customOotbMetricName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the OOTB metric configuration.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"isAgentic": {
"default": false,
"description": "Whether the OOTB metric configuration is specific to agentic workflows.",
"title": "isAgentic",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM to use for `correctness` and `faithfulness` metrics.",
"title": "llmId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration to be associated with the OOTB metric."
},
"ootbMetricConfigurationId": {
"description": "The ID of OOTB metric.",
"title": "ootbMetricConfigurationId",
"type": "string"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
}
],
"description": "The name of the OOTB metric.",
"title": "ootbMetricName"
}
},
"required": [
"ootbMetricName",
"ootbMetricConfigurationId",
"executionStatus"
],
"title": "OOTBMetricConfigurationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | OOTB metric configuration | OOTBMetricConfigurationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
List sidecar model metric validations¶
Operation path: GET /api/v2/genai/sidecarModelMetricValidations/
Authentication requirements: BearerAuth
List sidecar model metric validations.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| useCaseId | query | any | false | Only retrieve the sidecar model metric validations associated with these use case IDs. |
| offset | query | integer | false | Skip the specified number of values. |
| limit | query | integer | false | Retrieve only the specified number of values. |
| search | query | any | false | Only retrieve the sidecar model metric validations matching the search query. |
| sort | query | any | false | Apply this sort order to the results. Valid options are "name", "deploymentName", "userName", "creationDate". Prefix the attribute name with a dash to sort in descending order, e.g., sort=-creationDate. |
| completedOnly | query | boolean | false | If true, only retrieve the completed sidecar model metric validations. The default is false. |
| deploymentId | query | any | false | Only retrieve the sidecar model metric validations associated with this deployment ID. |
| modelId | query | any | false | Only retrieve the sidecar model metric validations associated with this model ID. |
| promptColumnName | query | any | false | Only retrieve the sidecar model metric validations where the custom model uses this column name for prompt input. |
| targetColumnName | query | any | false | Only retrieve the sidecar model metric validations where the custom model uses this column name for prediction output. |
| citationsPrefixColumnName | query | any | false | Only retrieve the sidecar model metric validations where the custom model uses this column name prefix for citation inputs. |
Example responses¶
200 Response
{
"description": "Paginated list of sidecar model metric validations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListSidecarModelMetricValidationnResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Sidecar model metric validations successfully retrieved. | ListSidecarModelMetricValidationnResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Validate sidecar model metric¶
Operation path: POST /api/v2/genai/sidecarModelMetricValidations/
Authentication requirements: BearerAuth
Validate a metric hosted in a custom model deployment (also known as a sidecar model metric) for use in the playground.
Body parameter¶
{
"description": "The body of the \"Validate sidecar model metric\" request.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"expectedResponseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for the expected response text input.",
"title": "expectedResponseColumnName"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model used in the deployment.",
"title": "modelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration to be associated with the sidecar model metric."
},
"name": {
"default": "Untitled",
"description": "The name to use for the validated custom model.",
"maxLength": 5000,
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The ID of the playground to associate with the validated custom model.",
"title": "playgroundId",
"type": "string"
},
"predictionTimeout": {
"default": 300,
"description": "The timeout in seconds for the prediction when validating a custom model. Defaults to 300.",
"maximum": 600,
"minimum": 1,
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"maxLength": 5000,
"title": "targetColumnName",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case to associate with the validated custom model.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"deploymentId",
"useCaseId",
"playgroundId",
"targetColumnName"
],
"title": "CreateSidecarModelMetricValidationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | CreateSidecarModelMetricValidationRequest | true | none |
Example responses¶
202 Response
{
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 202 | Accepted | Sidecar model metric validation job successfully accepted. Follow the Location header to poll for job execution status. |
SidecarModelMetricValidationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Delete sidecar model metric validation by validation ID¶
Operation path: DELETE /api/v2/genai/sidecarModelMetricValidations/{validationId}/
Authentication requirements: BearerAuth
Delete an existing sidecar model metric validation.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| validationId | path | string | true | The ID of the sidecar model metric validation to delete. |
Example responses¶
422 Response
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 204 | No Content | Sidecar model metric validation successfully deleted. | None |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Retrieve sidecar model metric validation status by validation ID¶
Operation path: GET /api/v2/genai/sidecarModelMetricValidations/{validationId}/
Authentication requirements: BearerAuth
Retrieve the status of validating a sidecar model metric.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| validationId | path | string | true | The ID of the sidecar model metric validation to retrieve. |
Example responses¶
200 Response
{
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Sidecar model metric validation status successfully retrieved. | SidecarModelMetricValidationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Edit sidecar model metric validation by validation ID¶
Operation path: PATCH /api/v2/genai/sidecarModelMetricValidations/{validationId}/
Authentication requirements: BearerAuth
Edit an existing sidecar model metric validation.
Body parameter¶
{
"description": "The body of the \"Edit sidecar model metric validation\" request.",
"properties": {
"chatModelId": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The model ID to specify when calling the OpenAI chat completion API of the deployment. If this parameter is specified, the deployment must support the OpenAI chat completion API.",
"title": "chatModelId"
},
"citationsPrefixColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the column name prefix that will be used to submit the citation inputs to the sidecar model.",
"title": "citationsPrefixColumnName"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the ID of the deployment associated with this custom model validation.",
"title": "deploymentId"
},
"expectedResponseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to submit the expected response text input to the sidecar model.",
"title": "expectedResponseColumnName"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the ID of the model associated with this custom model validation.",
"title": "modelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration to be associated with the sidecar model metric."
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, renames the custom model validation to this value.",
"title": "name"
},
"predictionTimeout": {
"anyOf": [
{
"maximum": 600,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "If specified, sets the timeout in seconds for the prediction when validating a custom model.",
"title": "predictionTimeout"
},
"promptColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to format the prompt text input for the custom model deployment.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to submit the response text input to the sidecar model.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to extract the prediction response from the custom model deployment.",
"title": "targetColumnName"
}
},
"title": "EditSidecarModelMetricValidationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| validationId | path | string | true | The ID of the sidecar model metric validation to edit. |
| body | body | EditSidecarModelMetricValidationRequest | true | none |
Example responses¶
200 Response
{
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Sidecar model metric validation successfully updated. | SidecarModelMetricValidationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Revalidate sidecar model metric by validation ID¶
Operation path: POST /api/v2/genai/sidecarModelMetricValidations/{validationId}/revalidate/
Authentication requirements: BearerAuth
Revalidate an existing sidecar model metric validation.
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| validationId | path | string | true | The ID of the sidecar model metric validation to revalidate. |
Example responses¶
200 Response
{
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Sidecar model metric successfully revalidated. | SidecarModelMetricValidationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Generate synthetic evaluation dataset¶
Operation path: POST /api/v2/genai/syntheticEvaluationDatasetGenerations/
Authentication requirements: BearerAuth
Generate a synthetic evaluation dataset.
Body parameter¶
{
"description": "The body of the \"Generate synthetic evaluation dataset\" request.",
"properties": {
"datasetName": {
"anyOf": [
{
"maxLength": 255,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the generated dataset.",
"title": "datasetName"
},
"language": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The language to use for the generated dataset.",
"title": "language"
},
"llmId": {
"description": "The ID of the LLM to use for synthetic dataset generation.",
"title": "llmId",
"type": "string"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "If specified, uses these LLM settings for the prompt and updates the settings of the corresponding chat or LLM blueprint to use these LLM settings.",
"title": "llmSettings"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database to use for synthetic dataset generation.",
"title": "vectorDatabaseId"
}
},
"required": [
"llmId"
],
"title": "SyntheticEvaluationDatasetGenerationRequest",
"type": "object"
}
Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| body | body | SyntheticEvaluationDatasetGenerationRequest | true | none |
Example responses¶
202 Response
{
"description": "The body of the \"Create synthetic evaluation dataset\" response.",
"properties": {
"datasetId": {
"description": "The ID of the created dataset.",
"title": "datasetId",
"type": "string"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName",
"type": "string"
}
},
"required": [
"datasetId",
"promptColumnName",
"responseColumnName"
],
"title": "SyntheticEvaluationDatasetGenerationResponse",
"type": "object"
}
Responses¶
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 202 | Accepted | Synthetic evaluation data generation job successfully accepted. Follow the Location header to poll for job execution status. |
SyntheticEvaluationDatasetGenerationResponse |
| 422 | Unprocessable Entity | Validation Error | HTTPValidationErrorResponse |
Schemas¶
AggregatedAggregationValue
{
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
}
AggregatedAggregationValue
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of metric aggregation items aggregated. | |
| item | string | true | The name of the item. | |
| value | number | true | The value associated with the item. |
AggregationType
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
}
AggregationType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| AggregationType | string | false | The type of the metric aggregation. |
Enumerated Values¶
| Property | Value |
|---|---|
| AggregationType | [average, percentYes, classPercentCoverage, ngramImportance, guardConditionPercentYes] |
AggregationValue
{
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
}
AggregationValue
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| item | string | true | The name of the item. | |
| value | number | true | The value associated with the item. |
ArgumentMatchMode
{
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
ArgumentMatchMode
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ArgumentMatchMode | string | false | The different modes for comparing the arguments of tool calls. |
Enumerated Values¶
| Property | Value |
|---|---|
| ArgumentMatchMode | [exact_match, ignore_arguments] |
CommonLLMSettings
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
}
CommonLLMSettings
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| maxCompletionLength | any | false | Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | integer | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| systemPrompt | any | false | System prompt guides the style of the LLM response. It is a "universal" prompt, prepended to all individual prompts. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 500000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
CostMetricConfigurationResponse
{
"description": "API response object for a single cost metric configuration.",
"properties": {
"costConfigurationId": {
"description": "The ID of the cost metric configuration.",
"title": "costConfigurationId",
"type": "string"
},
"costMetricConfigurations": {
"description": "The list of individual LLM cost configurations that constitute this cost metric configuration.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the cost metric configuration.",
"title": "playgroundId"
},
"useCaseId": {
"description": "The ID of the use case associated with the cost metric configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"costConfigurationId",
"useCaseId",
"costMetricConfigurations"
],
"title": "CostMetricConfigurationResponse",
"type": "object"
}
CostMetricConfigurationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| costConfigurationId | string | true | The ID of the cost metric configuration. | |
| costMetricConfigurations | [LLMCostConfigurationResponse] | true | The list of individual LLM cost configurations that constitute this cost metric configuration. | |
| name | any | false | The name to use for the cost configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| playgroundId | any | false | The ID of the playground associated with the cost metric configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| useCaseId | string | true | The ID of the use case associated with the cost metric configuration. |
CreateCostMetricConfigurationRequest
{
"description": "The body of the \"Create cost metric configuration\" request.",
"properties": {
"costMetricConfigurations": {
"description": "The list of cost metric configurations to use.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
},
"playgroundId": {
"description": "The ID of the playground to associate with the cost metric configuration.",
"title": "playgroundId",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case to associate with the cost metric configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"useCaseId",
"playgroundId",
"costMetricConfigurations"
],
"title": "CreateCostMetricConfigurationRequest",
"type": "object"
}
CreateCostMetricConfigurationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| costMetricConfigurations | [LLMCostConfigurationResponse] | true | The list of cost metric configurations to use. | |
| name | any | false | The name to use for the cost configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| playgroundId | string | true | The ID of the playground to associate with the cost metric configuration. | |
| useCaseId | string | true | The ID of the use case to associate with the cost metric configuration. |
CreateEvaluationDatasetConfigurationRequest
{
"description": "The body of the \"Create evaluation dataset configuration\" request.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals. It is required to evaluate the AgentGoalAccuracyWithReference metric for agentic workflows.",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"isSyntheticDataset": {
"default": false,
"description": "Whether the evaluation dataset is synthetic.",
"title": "isSyntheticDataset",
"type": "boolean"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the evaluation dataset configuration.",
"title": "name"
},
"playgroundId": {
"description": "The ID of the playground to associate with the evaluation dataset configuration.",
"title": "playgroundId",
"type": "string"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"toolCallsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls. It is required to evaluate the ToolCallAccuracy metric for agentic workflows.",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case to associate with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"useCaseId",
"playgroundId",
"datasetId",
"promptColumnName"
],
"title": "CreateEvaluationDatasetConfigurationRequest",
"type": "object"
}
CreateEvaluationDatasetConfigurationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| agentGoalsColumnName | any | false | The name of the dataset column containing expected agent goals. It is required to evaluate the AgentGoalAccuracyWithReference metric for agentic workflows. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| correctnessEnabled | any | false | Whether correctness is enabled for the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetId | string | true | The ID of the evaluation dataset. | |
| isSyntheticDataset | boolean | false | Whether the evaluation dataset is synthetic. | |
| name | any | false | The name of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| playgroundId | string | true | The ID of the playground to associate with the evaluation dataset configuration. | |
| promptColumnName | string | true | The name of the dataset column containing the prompt text. | |
| responseColumnName | any | false | The name of the dataset column containing the response text. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| toolCallsColumnName | any | false | The name of the dataset column containing expected tool calls. It is required to evaluate the ToolCallAccuracy metric for agentic workflows. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| useCaseId | string | true | The ID of the use case to associate with the evaluation dataset configuration. |
CreateEvaluationDatasetMetricAggregationRequest
{
"description": "The body of the \"Create evaluation dataset metric aggregation\" request.",
"properties": {
"chatName": {
"default": "Aggregated chat",
"description": "The name for the new chat that will contain the associated prompts and responses.",
"maxLength": 5000,
"title": "chatName",
"type": "string"
},
"evaluationDatasetConfigurationId": {
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId",
"type": "string"
},
"insightsConfiguration": {
"description": "The configuration of insights for the metric aggregation.",
"items": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"minItems": 1,
"title": "insightsConfiguration",
"type": "array"
},
"llmBlueprintIds": {
"description": "The IDs of the LLM blueprints to use for the metric aggregation.",
"items": {
"type": "string"
},
"maxItems": 3,
"minItems": 1,
"title": "llmBlueprintIds",
"type": "array"
}
},
"required": [
"llmBlueprintIds",
"evaluationDatasetConfigurationId",
"insightsConfiguration"
],
"title": "CreateEvaluationDatasetMetricAggregationRequest",
"type": "object"
}
CreateEvaluationDatasetMetricAggregationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| chatName | string | false | maxLength: 5000 |
The name for the new chat that will contain the associated prompts and responses. |
| evaluationDatasetConfigurationId | string | true | The ID of the evaluation dataset configuration. | |
| insightsConfiguration | [InsightsConfigurationWithAdditionalData] | true | minItems: 1 |
The configuration of insights for the metric aggregation. |
| llmBlueprintIds | [string] | true | maxItems: 3 minItems: 1 |
The IDs of the LLM blueprints to use for the metric aggregation. |
CreateEvaluationDatasetMetricAggregationResponse
{
"description": "The body of the \"Create evaluation dataset metric aggregation\" response.",
"properties": {
"chatIds": {
"description": "The IDs of the chats associated with the metric aggregation.",
"items": {
"type": "string"
},
"title": "chatIds",
"type": "array"
},
"jobId": {
"description": "The ID of the evaluation dataset metric aggregation job.",
"format": "uuid4",
"title": "jobId",
"type": "string"
}
},
"required": [
"jobId",
"chatIds"
],
"title": "CreateEvaluationDatasetMetricAggregationResponse",
"type": "object"
}
CreateEvaluationDatasetMetricAggregationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| chatIds | [string] | true | The IDs of the chats associated with the metric aggregation. | |
| jobId | string(uuid4) | true | The ID of the evaluation dataset metric aggregation job. |
CreateLLMTestConfigurationRequest
{
"description": "Request object for creating a LLMTestConfiguration.",
"properties": {
"datasetEvaluations": {
"description": "Dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 0,
"description": "The max number of prompts to evaluate.",
"maximum": 5000,
"minimum": 0,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset name. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria"
],
"title": "DatasetEvaluationRequest",
"type": "object"
},
"maxItems": 10,
"minItems": 1,
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"default": "",
"description": "LLM test configuration description.",
"maxLength": 5000,
"title": "description",
"type": "string"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "LLM test configuration name.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The use case ID associated with the LLM Test configuration.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"name",
"useCaseId",
"datasetEvaluations",
"llmTestGradingCriteria"
],
"title": "CreateLLMTestConfigurationRequest",
"type": "object"
}
CreateLLMTestConfigurationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetEvaluations | [DatasetEvaluationRequest] | true | maxItems: 10 minItems: 1 |
Dataset evaluations. |
| description | string | false | maxLength: 5000 |
LLM test configuration description. |
| llmTestGradingCriteria | LLMTestGradingCriteria | true | LLM test grading criteria. | |
| name | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
LLM test configuration name. |
| useCaseId | string | true | The use case ID associated with the LLM Test configuration. |
CreateLLMTestResultRequest
{
"description": "Request object for creating a LLMTestResult.",
"properties": {
"llmBlueprintId": {
"description": "The LLM Blueprint ID associated with the LLM Test result.",
"title": "llmBlueprintId",
"type": "string"
},
"llmTestConfigurationId": {
"description": "The use case ID associated with the LLM Test result.",
"title": "llmTestConfigurationId",
"type": "string"
}
},
"required": [
"llmTestConfigurationId",
"llmBlueprintId"
],
"title": "CreateLLMTestResultRequest",
"type": "object"
}
CreateLLMTestResultRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmBlueprintId | string | true | The LLM Blueprint ID associated with the LLM Test result. | |
| llmTestConfigurationId | string | true | The use case ID associated with the LLM Test result. |
CreateLLMTestSuiteRequest
{
"description": "The body of the \"Create LLM test suite\" request.",
"properties": {
"description": {
"default": "",
"description": "The description of the LLM test suite.",
"maxLength": 5000,
"title": "description",
"type": "string"
},
"llmTestConfigurationIds": {
"default": [],
"description": "The IDs of the LLM test configurations in the LLM test suite.",
"items": {
"type": "string"
},
"maxItems": 100,
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case to associate with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"name",
"useCaseId"
],
"title": "CreateLLMTestSuiteRequest",
"type": "object"
}
CreateLLMTestSuiteRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| description | string | false | maxLength: 5000 |
The description of the LLM test suite. |
| llmTestConfigurationIds | [string] | false | maxItems: 100 |
The IDs of the LLM test configurations in the LLM test suite. |
| name | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the LLM test suite. |
| useCaseId | string | true | The ID of the use case to associate with the LLM test suite. |
CreateSidecarModelMetricValidationRequest
{
"description": "The body of the \"Validate sidecar model metric\" request.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"expectedResponseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for the expected response text input.",
"title": "expectedResponseColumnName"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model used in the deployment.",
"title": "modelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration to be associated with the sidecar model metric."
},
"name": {
"default": "Untitled",
"description": "The name to use for the validated custom model.",
"maxLength": 5000,
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The ID of the playground to associate with the validated custom model.",
"title": "playgroundId",
"type": "string"
},
"predictionTimeout": {
"default": 300,
"description": "The timeout in seconds for the prediction when validating a custom model. Defaults to 300.",
"maximum": 600,
"minimum": 1,
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"maxLength": 5000,
"title": "targetColumnName",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case to associate with the validated custom model.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"deploymentId",
"useCaseId",
"playgroundId",
"targetColumnName"
],
"title": "CreateSidecarModelMetricValidationRequest",
"type": "object"
}
CreateSidecarModelMetricValidationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| citationsPrefixColumnName | any | false | The column name prefix the custom model uses for citation inputs. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| deploymentId | string | true | The ID of the custom model deployment. | |
| expectedResponseColumnName | any | false | The name of the column the custom model uses for the expected response text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| modelId | any | false | The ID of the model used in the deployment. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| moderationConfiguration | any | false | The moderation configuration to be associated with the sidecar model metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ModerationConfigurationWithoutID | false | Moderation Configuration associated with an insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | string | false | maxLength: 5000 |
The name to use for the validated custom model. |
| playgroundId | string | true | The ID of the playground to associate with the validated custom model. | |
| predictionTimeout | integer | false | maximum: 600 minimum: 1 |
The timeout in seconds for the prediction when validating a custom model. Defaults to 300. |
| promptColumnName | any | false | The name of the column the custom model uses for prompt text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| responseColumnName | any | false | The name of the column the custom model uses for response text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| targetColumnName | string | true | maxLength: 5000 |
The name of the column the custom model uses for prediction output. |
| useCaseId | string | true | The ID of the use case to associate with the validated custom model. |
CustomModelChatLLMSettings
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
}
CustomModelChatLLMSettings
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customModelId | string | true | The ID of the custom model used via chat completion interface. | |
| systemPrompt | any | false | System prompt guides the style of the LLM response. It is a "universal" prompt, prepended to all individual prompts. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 500000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
CustomModelGuard
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
}
CustomModelGuard
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the guard. |
| ootbType | any | false | Out of the box type of the guard. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CustomModelGuardOOTBType | false | OOTB type as used in the moderation_config.yaml file of the custom model. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| stage | CustomModelGuardStage | true | Stage on which the guard gets applied. | |
| type | CustomModelGuardType | true | Type of the guard. |
CustomModelGuardOOTBType
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
}
CustomModelGuardOOTBType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| CustomModelGuardOOTBType | string | false | OOTB type as used in the moderation_config.yaml file of the custom model. |
Enumerated Values¶
| Property | Value |
|---|---|
| CustomModelGuardOOTBType | [token_count, rouge_1, faithfulness, agent_goal_accuracy, custom_metric, cost, task_adherence] |
CustomModelGuardStage
{
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
}
CustomModelGuardStage
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| CustomModelGuardStage | string | false | Guard stage as used in the moderation_config.yaml file of the custom model. |
Enumerated Values¶
| Property | Value |
|---|---|
| CustomModelGuardStage | [prompt, response] |
CustomModelGuardType
{
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
CustomModelGuardType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| CustomModelGuardType | string | false | Guard type as used in the moderation_config.yaml file of the custom model. |
Enumerated Values¶
| Property | Value |
|---|---|
| CustomModelGuardType | [ootb, model, nemo_guardrails] |
CustomModelLLMSettings
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
}
CustomModelLLMSettings
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| externalLlmContextSize | any | false | The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | integer | false | maximum: 128000 minimum: 128 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| systemPrompt | any | false | System prompt guides the style of the LLM response. It is a "universal" prompt, prepended to all individual prompts. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 500000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| validationId | any | false | The validation ID of the custom model LLM. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
CustomModelValidationStatus
{
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
CustomModelValidationStatus
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| CustomModelValidationStatus | string | false | Status of custom model validation. |
Enumerated Values¶
| Property | Value |
|---|---|
| CustomModelValidationStatus | [TESTING, PASSED, FAILED] |
DatasetEvaluationRequest
{
"description": "Dataset evaluation.",
"properties": {
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 0,
"description": "The max number of prompts to evaluate.",
"maximum": 5000,
"minimum": 0,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset name. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria"
],
"title": "DatasetEvaluationRequest",
"type": "object"
}
DatasetEvaluationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | any | false | The ID of the evaluation dataset configuration for this dataset evaluation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationName | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the evaluation. This name should provide context regarding what is being evaluated. |
| insightConfiguration | InsightsConfigurationWithAdditionalData | true | The configuration of insights with extra data. | |
| insightGradingCriteria | InsightGradingCriteria | true | Grading criteria for an insight. | |
| maxNumPrompts | integer | false | maximum: 5000 minimum: 0 |
The max number of prompts to evaluate. |
| ootbDatasetName | any | false | Out-of-the-box evaluation dataset name. This applies only to our predefined public evaluation datasets. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBDatasetName | false | Out-of-the-box dataset name. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptSamplingStrategy | PromptSamplingStrategy | false | The prompt sampling strategy. Controls how max_num_prompts are sampled. |
DatasetEvaluationResponse
{
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
}
DatasetEvaluationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorMessage | any | false | The error message associated with the dataset evaluation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | any | false | The ID of the evaluation dataset configuration for this dataset evaluation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetName | any | true | Evaluation dataset name. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationName | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the evaluation. This name should provide context regarding what is being evaluated. |
| insightConfiguration | InsightsConfigurationWithAdditionalData | true | The configuration of insights with extra data. | |
| insightGradingCriteria | InsightGradingCriteria | true | Grading criteria for an insight. | |
| maxNumPrompts | integer | false | maximum: 5000 |
The max number of prompts to evaluate. |
| ootbDataset | any | false | Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBDataset | false | Out-of-the-box dataset. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptSamplingStrategy | PromptSamplingStrategy | false | The prompt sampling strategy. Controls how max_num_prompts are sampled. |
DatasetIdentifier
{
"description": "Dataset identifier.",
"properties": {
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the dataset, if any.",
"title": "datasetId"
},
"datasetName": {
"description": "The name of the dataset.",
"title": "datasetName",
"type": "string"
}
},
"required": [
"datasetName",
"datasetId"
],
"title": "DatasetIdentifier",
"type": "object"
}
DatasetIdentifier
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetId | any | true | The ID of the dataset, if any. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetName | string | true | The name of the dataset. |
DeploymentAccessData
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
}
DeploymentAccessData
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| authorizationHeader | string | false | The Authorization header to use for the deployment. |
|
| chatApiUrl | any | false | The URL of the deployment's chat API. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datarobotKey | any | true | The server key associated with the prediction API. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| inputType | DeploymentInputType | true | The format of the input data. | |
| modelType | SupportedDeploymentType | true | The type of the target output the deployment produces. | |
| predictionApiUrl | string | true | The URL of the deployment's prediction API. |
DeploymentInputType
{
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
}
DeploymentInputType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| DeploymentInputType | string | false | The format of the input data submitted to a DataRobot deployment. |
Enumerated Values¶
| Property | Value |
|---|---|
| DeploymentInputType | [CSV, JSON] |
EditCostMetricConfigurationRequest
{
"description": "The body of the \"Edit cost metric configuration\" request.",
"properties": {
"costMetricConfigurations": {
"description": "The list of LLM cost configurations to apply to this cost metric configuration.",
"items": {
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
},
"minItems": 1,
"title": "costMetricConfigurations",
"type": "array"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the cost configuration.",
"title": "name"
}
},
"required": [
"costMetricConfigurations"
],
"title": "EditCostMetricConfigurationRequest",
"type": "object"
}
EditCostMetricConfigurationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| costMetricConfigurations | [LLMCostConfigurationResponse] | true | minItems: 1 |
The list of LLM cost configurations to apply to this cost metric configuration. |
| name | any | false | The name to use for the cost configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
EditEvaluationDatasetConfigurationRequest
{
"description": "The body of the \"Edit evaluation dataset configuration\" request.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the expected name of the dataset column containing expected agent goals. It is required to evaluate the AgentGoalAccuracyWithReference metric for agentic workflows.",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "If specified, enables or disables correctness for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"datasetId": {
"default": "000000000000000000000000",
"description": "If specified, updates the ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, renames the evaluation dataset configuration to this value.",
"title": "name"
},
"promptColumnName": {
"default": "None",
"description": "If specified, changes the expected name of the dataset column containing the prompt text.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the expected name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"toolCallsColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the expected name of the dataset column containing expected tool calls. It is required to evaluate the ToolCallAccuracy metric for agentic workflows.",
"title": "toolCallsColumnName"
}
},
"title": "EditEvaluationDatasetConfigurationRequest",
"type": "object"
}
EditEvaluationDatasetConfigurationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| agentGoalsColumnName | any | false | If specified, changes the expected name of the dataset column containing expected agent goals. It is required to evaluate the AgentGoalAccuracyWithReference metric for agentic workflows. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| correctnessEnabled | any | false | If specified, enables or disables correctness for the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetId | string | false | If specified, updates the ID of the evaluation dataset. | |
| name | any | false | If specified, renames the evaluation dataset configuration to this value. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
If specified, changes the expected name of the dataset column containing the prompt text. |
| responseColumnName | any | false | If specified, changes the expected name of the dataset column containing the response text. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| toolCallsColumnName | any | false | If specified, changes the expected name of the dataset column containing expected tool calls. It is required to evaluate the ToolCallAccuracy metric for agentic workflows. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
EditLLMTestConfigurationRequest
{
"description": "Request object for editing a LLMTestConfiguration.",
"properties": {
"datasetEvaluations": {
"anyOf": [
{
"items": {
"description": "Dataset evaluation.",
"properties": {
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 0,
"description": "The max number of prompts to evaluate.",
"maximum": 5000,
"minimum": 0,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset name. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria"
],
"title": "DatasetEvaluationRequest",
"type": "object"
},
"maxItems": 10,
"minItems": 1,
"type": "array"
},
{
"type": "null"
}
],
"description": "New Dataset evaluations.",
"title": "datasetEvaluations"
},
"description": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "New LLM test configuration description.",
"title": "description"
},
"llmTestGradingCriteria": {
"anyOf": [
{
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
{
"type": "null"
}
],
"description": "New LLM test grading criteria."
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "New LLM test configuration name.",
"title": "name"
}
},
"title": "EditLLMTestConfigurationRequest",
"type": "object"
}
EditLLMTestConfigurationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetEvaluations | any | false | New Dataset evaluations. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [DatasetEvaluationRequest] | false | maxItems: 10 minItems: 1 |
[Dataset evaluation.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| description | any | false | New LLM test configuration description. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmTestGradingCriteria | any | false | New LLM test grading criteria. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | LLMTestGradingCriteria | false | Grading criteria for the LLM Test configuration. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | any | false | New LLM test configuration name. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
EditLLMTestSuiteRequest
{
"description": "The body of the \"Edit LLM test suite\" request.",
"properties": {
"description": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The description of the LLM test suite.",
"title": "description"
},
"llmTestConfigurationIds": {
"anyOf": [
{
"items": {
"type": "string"
},
"maxItems": 100,
"type": "array"
},
{
"type": "null"
}
],
"description": "The IDs of the LLM test configurations in the LLM test suite.",
"title": "llmTestConfigurationIds"
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the LLM test suite.",
"title": "name"
}
},
"title": "EditLLMTestSuiteRequest",
"type": "object"
}
EditLLMTestSuiteRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| description | any | false | The description of the LLM test suite. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmTestConfigurationIds | any | false | The IDs of the LLM test configurations in the LLM test suite. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [string] | false | maxItems: 100 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | any | false | The name of the LLM test suite. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
EditSidecarModelMetricValidationRequest
{
"description": "The body of the \"Edit sidecar model metric validation\" request.",
"properties": {
"chatModelId": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "The model ID to specify when calling the OpenAI chat completion API of the deployment. If this parameter is specified, the deployment must support the OpenAI chat completion API.",
"title": "chatModelId"
},
"citationsPrefixColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the column name prefix that will be used to submit the citation inputs to the sidecar model.",
"title": "citationsPrefixColumnName"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the ID of the deployment associated with this custom model validation.",
"title": "deploymentId"
},
"expectedResponseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to submit the expected response text input to the sidecar model.",
"title": "expectedResponseColumnName"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the ID of the model associated with this custom model validation.",
"title": "modelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration to be associated with the sidecar model metric."
},
"name": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, renames the custom model validation to this value.",
"title": "name"
},
"predictionTimeout": {
"anyOf": [
{
"maximum": 600,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "If specified, sets the timeout in seconds for the prediction when validating a custom model.",
"title": "predictionTimeout"
},
"promptColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to format the prompt text input for the custom model deployment.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to submit the response text input to the sidecar model.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, changes the name of the column that will be used to extract the prediction response from the custom model deployment.",
"title": "targetColumnName"
}
},
"title": "EditSidecarModelMetricValidationRequest",
"type": "object"
}
EditSidecarModelMetricValidationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| chatModelId | any | false | The model ID to specify when calling the OpenAI chat completion API of the deployment. If this parameter is specified, the deployment must support the OpenAI chat completion API. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| citationsPrefixColumnName | any | false | If specified, changes the column name prefix that will be used to submit the citation inputs to the sidecar model. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| deploymentId | any | false | If specified, changes the ID of the deployment associated with this custom model validation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| expectedResponseColumnName | any | false | If specified, changes the name of the column that will be used to submit the expected response text input to the sidecar model. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| modelId | any | false | If specified, changes the ID of the model associated with this custom model validation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| moderationConfiguration | any | false | The moderation configuration to be associated with the sidecar model metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ModerationConfigurationWithoutID | false | Moderation Configuration associated with an insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | any | false | If specified, renames the custom model validation to this value. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| predictionTimeout | any | false | If specified, sets the timeout in seconds for the prediction when validating a custom model. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | integer | false | maximum: 600 minimum: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | any | false | If specified, changes the name of the column that will be used to format the prompt text input for the custom model deployment. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| responseColumnName | any | false | If specified, changes the name of the column that will be used to submit the response text input to the sidecar model. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| targetColumnName | any | false | If specified, changes the name of the column that will be used to extract the prediction response from the custom model deployment. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
EvaluationDatasetConfigurationResponse
{
"description": "API response object for a single evaluation dataset configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "EvaluationDatasetConfigurationResponse",
"type": "object"
}
EvaluationDatasetConfigurationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| agentGoalsColumnName | any | true | The name of the dataset column containing expected agent goals (For agentic workflows). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| correctnessEnabled | any | true | Whether correctness is enabled for the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationDate | string(date-time) | true | The creation date of the evaluation dataset configuration (ISO 8601 formatted). | |
| creationUserId | string | true | The ID of the user that created the evaluation dataset configuration. | |
| datasetId | string | true | The ID of the evaluation dataset. | |
| datasetName | string | true | The name of the evaluation dataset. | |
| errorMessage | any | false | The error message associated with the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| executionStatus | ExecutionStatus | true | The execution status of the evaluation dataset. | |
| id | string | true | The ID of the evaluation dataset configuration. | |
| name | string | true | The name of the evaluation dataset configuration. | |
| playgroundId | any | true | The ID of the playground associated with the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | string | true | The name of the dataset column containing the prompt text. | |
| responseColumnName | any | true | The name of the dataset column containing the response text. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| rowsCount | integer | true | The rows count of the evaluation dataset. | |
| size | integer | true | The size of the evaluation dataset (in bytes). | |
| tenantId | string(uuid4) | true | The ID of the DataRobot tenant this evaluation dataset configuration belongs to. | |
| toolCallsColumnName | any | true | The name of the dataset column containing expected tool calls (For agentic workflows). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| useCaseId | string | true | The ID of the use case associated with the evaluation dataset configuration. | |
| userName | string | true | The name of the user that created the evaluation dataset configuration. |
EvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse
{
"description": "API response object for multiple evaluation dataset metric aggregation\naggregated by llm blueprint.",
"properties": {
"aggregatedItemCount": {
"description": "Number of items aggregated.",
"title": "aggregatedItemCount",
"type": "integer"
},
"aggregatedItemDetails": {
"description": "List of details for aggregated items.",
"items": {
"description": "Details for aggregated items.",
"properties": {
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "EvaluationDatasetMetricAggregationChatDetails",
"type": "object"
},
"title": "aggregatedItemDetails",
"type": "array"
},
"aggregationType": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"items": {
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregated value of the metric.",
"title": "aggregationValue"
},
"customModelGuardId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model's guard the metric aggregation belongs to.",
"title": "customModelGuardId"
},
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The dataset ID of the evaluation dataset configuration.",
"title": "datasetId"
},
"datasetName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The Data Registry dataset name of the evaluation dataset configuration.",
"title": "datasetName"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration associated with the metric aggregation.",
"title": "evaluationDatasetConfigurationId"
},
"llmBlueprintId": {
"description": "The ID of the LLM blueprint associated with the metric aggregation.",
"title": "llmBlueprintId",
"type": "string"
},
"metricName": {
"description": "The name of the metric associated with the metric aggregation.",
"title": "metricName",
"type": "string"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the out-of-the-box dataset."
},
"tenantId": {
"description": "The ID of the tenant the metric aggregation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
}
},
"required": [
"llmBlueprintId",
"evaluationDatasetConfigurationId",
"ootbDatasetName",
"datasetId",
"datasetName",
"metricName",
"aggregationValue",
"aggregationType",
"tenantId",
"customModelGuardId",
"aggregatedItemDetails",
"aggregatedItemCount"
],
"title": "EvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse",
"type": "object"
}
EvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| aggregatedItemCount | integer | true | Number of items aggregated. | |
| aggregatedItemDetails | [EvaluationDatasetMetricAggregationChatDetails] | true | List of details for aggregated items. | |
| aggregationType | AggregationType | true | The type of metric aggregation. | |
| aggregationValue | any | true | The aggregated value of the metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | number | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [AggregationValue] | false | [An individual record in an itemized metric aggregation.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [AggregatedAggregationValue] | false | [Aggregated record of multiple of the same item across different metric aggregation runs.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customModelGuardId | any | true | The ID of the custom model's guard the metric aggregation belongs to. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetId | any | true | The dataset ID of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetName | any | true | The Data Registry dataset name of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | any | true | The ID of the evaluation dataset configuration associated with the metric aggregation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmBlueprintId | string | true | The ID of the LLM blueprint associated with the metric aggregation. | |
| metricName | string | true | The name of the metric associated with the metric aggregation. | |
| ootbDatasetName | any | true | The name of the out-of-the-box dataset. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBDatasetName | false | Out-of-the-box dataset name. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| tenantId | string(uuid4) | true | The ID of the tenant the metric aggregation belongs to. |
EvaluationDatasetMetricAggregationChatDetails
{
"description": "Details for aggregated items.",
"properties": {
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "EvaluationDatasetMetricAggregationChatDetails",
"type": "object"
}
EvaluationDatasetMetricAggregationChatDetails
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| chatId | string | true | The ID of the chat associated with the metric aggregation. | |
| chatLink | string | true | The link to the chat associated with the metric aggregation. | |
| chatName | string | true | The name of the chat associated with the metric aggregation. | |
| creationDate | string(date-time) | true | The creation date of the metric aggregation (ISO 8601 formatted). | |
| creationUserId | string | true | The ID of the user that created the metric aggregation. | |
| creationUserName | string | true | The name of the user that created the metric aggregation. |
EvaluationDatasetMetricAggregationFieldQueryParam
{
"description": "Field used for aggregation when listing evaluation dataset metric aggregations.",
"enum": [
"metricName",
"llmBlueprintId",
"aggregationType",
"evaluationDatasetConfigurationId"
],
"title": "EvaluationDatasetMetricAggregationFieldQueryParam",
"type": "string"
}
EvaluationDatasetMetricAggregationFieldQueryParam
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| EvaluationDatasetMetricAggregationFieldQueryParam | string | false | Field used for aggregation when listing evaluation dataset metric aggregations. |
Enumerated Values¶
| Property | Value |
|---|---|
| EvaluationDatasetMetricAggregationFieldQueryParam | [metricName, llmBlueprintId, aggregationType, evaluationDatasetConfigurationId] |
EvaluationDatasetMetricAggregationResponse
{
"description": "API response object for a single evaluation dataset metric aggregation.",
"properties": {
"aggregationType": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"items": {
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregated value of the metric.",
"title": "aggregationValue"
},
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
},
"customModelGuardId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model's guard the metric aggregation belongs to.",
"title": "customModelGuardId"
},
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The dataset ID of the evaluation dataset configuration.",
"title": "datasetId"
},
"datasetName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The Data Registry dataset name of the evaluation dataset configuration.",
"title": "datasetName"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration associated with the metric aggregation.",
"title": "evaluationDatasetConfigurationId"
},
"llmBlueprintId": {
"description": "The ID of the LLM blueprint associated with the metric aggregation.",
"title": "llmBlueprintId",
"type": "string"
},
"metricName": {
"description": "The name of the metric associated with the metric aggregation.",
"title": "metricName",
"type": "string"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the out-of-the-box dataset."
},
"tenantId": {
"description": "The ID of the tenant the metric aggregation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName",
"llmBlueprintId",
"evaluationDatasetConfigurationId",
"ootbDatasetName",
"datasetId",
"datasetName",
"metricName",
"aggregationValue",
"aggregationType",
"tenantId",
"customModelGuardId"
],
"title": "EvaluationDatasetMetricAggregationResponse",
"type": "object"
}
EvaluationDatasetMetricAggregationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| aggregationType | AggregationType | true | The type of metric aggregation. | |
| aggregationValue | any | true | The aggregated value of the metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | number | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [AggregationValue] | false | [An individual record in an itemized metric aggregation.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [AggregatedAggregationValue] | false | [Aggregated record of multiple of the same item across different metric aggregation runs.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| chatId | string | true | The ID of the chat associated with the metric aggregation. | |
| chatLink | string | true | The link to the chat associated with the metric aggregation. | |
| chatName | string | true | The name of the chat associated with the metric aggregation. | |
| creationDate | string(date-time) | true | The creation date of the metric aggregation (ISO 8601 formatted). | |
| creationUserId | string | true | The ID of the user that created the metric aggregation. | |
| creationUserName | string | true | The name of the user that created the metric aggregation. | |
| customModelGuardId | any | true | The ID of the custom model's guard the metric aggregation belongs to. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetId | any | true | The dataset ID of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetName | any | true | The Data Registry dataset name of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | any | true | The ID of the evaluation dataset configuration associated with the metric aggregation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmBlueprintId | string | true | The ID of the LLM blueprint associated with the metric aggregation. | |
| metricName | string | true | The name of the metric associated with the metric aggregation. | |
| ootbDatasetName | any | true | The name of the out-of-the-box dataset. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBDatasetName | false | Out-of-the-box dataset name. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| tenantId | string(uuid4) | true | The ID of the tenant the metric aggregation belongs to. |
EvaluationDatasetMetricAggregationUniqueFieldValuesResponse
{
"description": "API response object for a single unique computed metric.",
"properties": {
"uniqueFieldValue": {
"description": "The unique value associated with the metric aggregation.",
"title": "uniqueFieldValue",
"type": "string"
}
},
"required": [
"uniqueFieldValue"
],
"title": "EvaluationDatasetMetricAggregationUniqueFieldValuesResponse",
"type": "object"
}
EvaluationDatasetMetricAggregationUniqueFieldValuesResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| uniqueFieldValue | string | true | The unique value associated with the metric aggregation. |
ExecutionStatus
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
}
ExecutionStatus
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ExecutionStatus | string | false | Job and entity execution status. |
Enumerated Values¶
| Property | Value |
|---|---|
| ExecutionStatus | [NEW, RUNNING, COMPLETED, REQUIRES_USER_INPUT, SKIPPED, ERROR] |
ExtraMetricSettings
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
}
ExtraMetricSettings
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| toolCallAccuracy | any | false | Extra settings for the tool call accuracy metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ToolCallAccuracySettings | false | Additional arguments for the tool call accuracy metric. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
GradingResult
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
}
GradingResult
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| GradingResult | string | false | Grading result. |
Enumerated Values¶
| Property | Value |
|---|---|
| GradingResult | [PASS, FAIL] |
GuardCondition
{
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
}
GuardCondition
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| comparand | any | true | The comparand(s) used in the guard condition. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | number | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [string] | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| comparator | GuardConditionComparator | true | The comparator used in the guard condition. |
GuardConditionComparator
{
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
GuardConditionComparator
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| GuardConditionComparator | string | false | The comparator used in a guard condition. |
Enumerated Values¶
| Property | Value |
|---|---|
| GuardConditionComparator | [greaterThan, lessThan, equals, notEquals, is, isNot, matches, doesNotMatch, contains, doesNotContain] |
HTTPValidationErrorResponse
{
"properties": {
"detail": {
"items": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
},
"title": "detail",
"type": "array"
}
},
"title": "HTTPValidationErrorResponse",
"type": "object"
}
HTTPValidationErrorResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| detail | [ValidationError] | false | none |
InsightErrorResolution
{
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
}
InsightErrorResolution
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| InsightErrorResolution | string | false | Error type linking directly to the field name that is related to the error. |
Enumerated Values¶
| Property | Value |
|---|---|
| InsightErrorResolution | [ootbMetricName, intervention, guardCondition, sidecarOverall, sidecarRevalidate, sidecarDeploymentId, sidecarInputColumnName, sidecarOutputColumnName, promptPipelineFiles, promptPipelineTemplateId, responsePipelineFiles, responsePipelineTemplateId] |
InsightEvaluationResultResponse
{
"description": "API response object for a single InsightEvaluationResult.",
"properties": {
"aggregationType": {
"anyOf": [
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Aggregation type."
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "Aggregation value. None indicates that the aggregation failed.",
"title": "aggregationValue"
},
"chatId": {
"description": "Chat ID.",
"title": "chatId",
"type": "string"
},
"chatName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Chat name.",
"title": "chatName"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Custom Model LLM Validation ID if using custom model LLM.",
"title": "customModelLLMValidationId"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset configuration ID.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "Evaluation name.",
"maxLength": 5000,
"title": "evaluationName",
"type": "string"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "Insight evaluation result ID.",
"title": "id",
"type": "string"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"lastUpdateDate": {
"description": "Last update date of the insight evaluation result (ISO 8601 formatted).",
"format": "date-time",
"title": "lastUpdateDate",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM ID used for this insight evaluation result.",
"title": "llmId"
},
"llmTestResultId": {
"description": "LLM test result ID this insight evaluation result is associated to.",
"title": "llmTestResultId",
"type": "string"
},
"maxNumPrompts": {
"description": "Number of prompts used in evaluation.",
"title": "maxNumPrompts",
"type": "integer"
},
"metricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the metric.",
"title": "metricName"
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"id",
"llmTestResultId",
"maxNumPrompts",
"promptSamplingStrategy",
"chatId",
"chatName",
"evaluationName",
"insightGradingCriteria",
"lastUpdateDate"
],
"title": "InsightEvaluationResultResponse",
"type": "object"
}
InsightEvaluationResultResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| aggregationType | any | false | Aggregation type. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | AggregationType | false | The type of the metric aggregation. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| aggregationValue | any | false | Aggregation value. None indicates that the aggregation failed. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | number | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [AggregationValue] | false | [An individual record in an itemized metric aggregation.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| chatId | string | true | Chat ID. | |
| chatName | any | true | Chat name. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customModelLLMValidationId | any | false | Custom Model LLM Validation ID if using custom model LLM. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | any | false | Evaluation dataset configuration ID. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetName | any | false | Evaluation dataset name. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationName | string | true | maxLength: 5000 |
Evaluation name. |
| executionStatus | ExecutionStatus | false | The execution status of the insight evaluation result. | |
| gradingResult | any | false | The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | GradingResult | false | Grading result. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| id | string | true | Insight evaluation result ID. | |
| insightGradingCriteria | InsightGradingCriteria | true | Insight grading criteria. | |
| lastUpdateDate | string(date-time) | true | Last update date of the insight evaluation result (ISO 8601 formatted). | |
| llmId | any | false | LLM ID used for this insight evaluation result. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmTestResultId | string | true | LLM test result ID this insight evaluation result is associated to. | |
| maxNumPrompts | integer | true | Number of prompts used in evaluation. | |
| metricName | any | false | Name of the metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptSamplingStrategy | PromptSamplingStrategy | true | Prompt sampling strategy for maxNumPrompts. |
InsightGradingCriteria
{
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
}
InsightGradingCriteria
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| passThreshold | integer | true | maximum: 100 minimum: 0 |
The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass. |
InsightToEvalDatasetsCompatibility
{
"description": "Insight to evaluation datasets compatibility.",
"properties": {
"incompatibleDatasets": {
"description": "The list of incompatible datasets.",
"items": {
"description": "Dataset identifier.",
"properties": {
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the dataset, if any.",
"title": "datasetId"
},
"datasetName": {
"description": "The name of the dataset.",
"title": "datasetName",
"type": "string"
}
},
"required": [
"datasetName",
"datasetId"
],
"title": "DatasetIdentifier",
"type": "object"
},
"title": "incompatibleDatasets",
"type": "array"
},
"insightName": {
"description": "The name of the insight.",
"title": "insightName",
"type": "string"
}
},
"required": [
"insightName",
"incompatibleDatasets"
],
"title": "InsightToEvalDatasetsCompatibility",
"type": "object"
}
InsightToEvalDatasetsCompatibility
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| incompatibleDatasets | [DatasetIdentifier] | true | The list of incompatible datasets. | |
| insightName | string | true | The name of the insight. |
InsightTypes
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
}
InsightTypes
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| InsightTypes | string | false | The type of insight. |
Enumerated Values¶
| Property | Value |
|---|---|
| InsightTypes | [Reference, Quality metric, Operational metric, Evaluation deployment, Custom metric, Nemo] |
InsightsConfigurationWithAdditionalData
{
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
}
InsightsConfigurationWithAdditionalData
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| aggregationTypes | any | true | The aggregation types used in the insights configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [AggregationType] | false | [The type of the metric aggregation.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| costConfigurationId | any | false | The ID of the cost configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customMetricId | any | false | The ID of the custom metric (if using a custom metric). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customModelGuard | any | false | Guard as configured in the custom model. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CustomModelGuard | false | Details of a guard as defined for the custom model. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customModelLLMValidationId | any | false | The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| deploymentId | any | false | The ID of the custom model deployment associated with the insight. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorMessage | any | false | The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorResolution | any | false | The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [string] | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| evaluationDatasetConfigurationId | any | false | The ID of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| executionStatus | any | false | The execution status of the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ExecutionStatus | false | Job and entity execution status. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| extraMetricSettings | any | false | Extra settings for the metric that do not reference other entities. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ExtraMetricSettings | false | Extra settings for the metric that do not reference other entities. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| insightName | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the insight. |
| insightType | any | false | The type of the insight. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | InsightTypes | false | The type of insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| isTransferable | boolean | false | Indicates if insight can be transferred to production. | |
| llmId | any | false | The LLM ID for OOTB metrics that use LLMs. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmIsActive | any | false | Whether the LLM is active. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmIsDeprecated | any | false | Whether the LLM is deprecated and will be removed in a future release. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| modelId | any | false | The ID of the model associated with deploymentId. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| modelPackageRegisteredModelId | any | false | The ID of the registered model package associated with deploymentId. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| moderationConfiguration | any | false | The moderation configuration associated with the insight configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ModerationConfigurationWithID | false | Moderation Configuration associated with an insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ModerationConfigurationWithoutID | false | Moderation Configuration associated with an insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| nemoMetricId | any | false | The ID of the Nemo configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ootbMetricId | any | false | The ID of the ootb metric (if using an ootb metric). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ootbMetricName | any | false | The OOTB metric name. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBMetricInsightNames | false | The Out-Of-The-Box metric name that can be used in the playground. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBAgenticMetricInsightNames | false | The Out-Of-The-Box metric name that can be used in an Agentic playground. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OTELMetricInsightNames | false | Metrics that can only be calculated using OTEL Trace/metric data. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| resultUnit | any | false | The unit of measurement associated with the insight result. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | MetricUnit | false | The unit of measurement associated with a metric. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| sidecarModelMetricMetadata | any | false | The metadata of the sidecar model metric (if using a sidecar model metric). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | SidecarModelMetricMetadata | false | The metadata of a sidecar model metric. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| sidecarModelMetricValidationId | any | false | The ID of the sidecar model metric validation (if using a sidecar model metric). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| stage | any | false | The stage (prompt or response) where insight is calculated at. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | PipelineStage | false | Enum that describes at which stage the metric may be calculated. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
Intervention
{
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
Intervention
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| action | ModerationAction | true | The intervention strategy. | |
| message | string | true | minLength: 1 minLength: 1 |
The intervention message to replace the prediction when a guard condition is satisfied. |
LLMBlueprintSnapshot
{
"description": "A snapshot in time of a LLMBlueprint's functional parameters.",
"properties": {
"description": {
"description": "The description of the LLMBlueprint at the time of snapshotting.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLMBlueprint for which the snapshot was produced.",
"title": "id",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM selected for this LLM blueprint.",
"title": "llmId"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of LLM settings.",
"title": "llmSettings"
},
"name": {
"description": "The name of the LLMBlueprint at the time of snapshotting.",
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The playground id of the LLMBlueprint.",
"title": "playgroundId",
"type": "string"
},
"promptType": {
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
},
"snapshotDate": {
"description": "The date when the snapshot was produced.",
"format": "date-time",
"title": "snapshotDate",
"type": "string"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database linked to this LLM blueprint.",
"title": "vectorDatabaseId"
},
"vectorDatabaseSettings": {
"anyOf": [
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of vector database settings."
}
},
"required": [
"id",
"name",
"description",
"playgroundId",
"promptType"
],
"title": "LLMBlueprintSnapshot",
"type": "object"
}
LLMBlueprintSnapshot
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| description | string | true | The description of the LLMBlueprint at the time of snapshotting. | |
| id | string | true | The ID of the LLMBlueprint for which the snapshot was produced. | |
| llmId | any | false | The ID of the LLM selected for this LLM blueprint. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmSettings | any | false | A key/value dictionary of LLM settings. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CommonLLMSettings | false | The settings that are available for all non-custom LLMs. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CustomModelLLMSettings | false | The settings that are available for custom model LLMs. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CustomModelChatLLMSettings | false | The settings that are available for custom model LLMs used via chat completion interface. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | string | true | The name of the LLMBlueprint at the time of snapshotting. | |
| playgroundId | string | true | The playground id of the LLMBlueprint. | |
| promptType | PromptType | true | The prompting type of the LLMBlueprint at the time of snapshotting. | |
| snapshotDate | string(date-time) | false | The date when the snapshot was produced. | |
| vectorDatabaseId | any | false | The ID of the vector database linked to this LLM blueprint. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| vectorDatabaseSettings | any | false | A key/value dictionary of vector database settings. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | VectorDatabaseSettings | false | Vector database retrieval settings. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
LLMCostConfigurationResponse
{
"description": "API request/response object for a cost configuration of a single LLM.",
"properties": {
"currencyCode": {
"default": "USD",
"description": "The arbitrary code code of the currency of `inputTokenPrice` and `outputTokenPrice`.",
"maxLength": 7,
"title": "currencyCode",
"type": "string"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"inputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceInputTokenCount` input tokens.",
"minimum": 0,
"title": "inputTokenPrice",
"type": "number"
},
"llmId": {
"description": "The ID of the LLM associated with this cost configuration.",
"title": "llmId",
"type": "string"
},
"outputTokenPrice": {
"default": 0.01,
"description": "The price of processing `referenceOutputTokenCount` output tokens.",
"minimum": 0,
"title": "outputTokenPrice",
"type": "number"
},
"referenceInputTokenCount": {
"default": 1000,
"description": "The number of input tokens corresponding to `inputTokenPrice`.",
"minimum": 0,
"title": "referenceInputTokenCount",
"type": "integer"
},
"referenceOutputTokenCount": {
"default": 1000,
"description": "The number of output tokens corresponding to `outputTokenPrice`.",
"minimum": 0,
"title": "referenceOutputTokenCount",
"type": "integer"
}
},
"required": [
"llmId"
],
"title": "LLMCostConfigurationResponse",
"type": "object"
}
LLMCostConfigurationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| currencyCode | string | false | maxLength: 7 |
The arbitrary code code of the currency of inputTokenPrice and outputTokenPrice. |
| customModelLLMValidationId | any | false | The ID of the custom model LLM validation (if using a custom model LLM). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| inputTokenPrice | number | false | minimum: 0 |
The price of processing referenceInputTokenCount input tokens. |
| llmId | string | true | The ID of the LLM associated with this cost configuration. | |
| outputTokenPrice | number | false | minimum: 0 |
The price of processing referenceOutputTokenCount output tokens. |
| referenceInputTokenCount | integer | false | minimum: 0 |
The number of input tokens corresponding to inputTokenPrice. |
| referenceOutputTokenCount | integer | false | minimum: 0 |
The number of output tokens corresponding to outputTokenPrice. |
LLMTestConfigurationNonOOTBDatasetResponse
{
"description": "Non out-of-the-box dataset used with an LLM test configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "LLMTestConfigurationNonOOTBDatasetResponse",
"type": "object"
}
LLMTestConfigurationNonOOTBDatasetResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| agentGoalsColumnName | any | true | The name of the dataset column containing expected agent goals (For agentic workflows). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| correctnessEnabled | any | true | Whether correctness is enabled for the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | boolean | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationDate | string(date-time) | true | The creation date of the evaluation dataset configuration (ISO 8601 formatted). | |
| creationUserId | string | true | The ID of the user that created the evaluation dataset configuration. | |
| datasetId | string | true | The ID of the evaluation dataset. | |
| datasetName | string | true | The name of the evaluation dataset. | |
| errorMessage | any | false | The error message associated with the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| executionStatus | ExecutionStatus | true | The execution status of the evaluation dataset. | |
| id | string | true | The ID of the evaluation dataset configuration. | |
| name | string | true | The name of the evaluation dataset configuration. | |
| playgroundId | any | true | The ID of the playground associated with the evaluation dataset configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | string | true | The name of the dataset column containing the prompt text. | |
| responseColumnName | any | true | The name of the dataset column containing the response text. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| rowsCount | integer | true | The rows count of the evaluation dataset. | |
| size | integer | true | The size of the evaluation dataset (in bytes). | |
| tenantId | string(uuid4) | true | The ID of the DataRobot tenant this evaluation dataset configuration belongs to. | |
| toolCallsColumnName | any | true | The name of the dataset column containing expected tool calls (For agentic workflows). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| useCaseId | string | true | The ID of the use case associated with the evaluation dataset configuration. | |
| userName | string | true | The name of the user that created the evaluation dataset configuration. |
LLMTestConfigurationOOTBDatasetResponse
{
"description": "Out-of-the-box dataset used with an LLM test configuration.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "LLMTestConfigurationOOTBDatasetResponse",
"type": "object"
}
LLMTestConfigurationOOTBDatasetResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetName | OOTBDatasetName | true | The name of the evaluation dataset. | |
| datasetUrl | any | true | The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBDatasetUrl | false | Out-of-the-box dataset URL. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the prompt column. |
| responseColumnName | any | true | The name of the response column, if present. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| rowsCount | integer | true | The number rows in the dataset. | |
| warning | any | false | Warning about the content of the dataset. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
LLMTestConfigurationResponse
{
"description": "API response object for a single LLMTestConfiguration.",
"properties": {
"creationDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationDate"
},
"creationUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationUserId"
},
"datasetEvaluations": {
"description": "The LLM test dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
},
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"description": "The description of the LLM Test configuration.",
"title": "description",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the LLM test configuration.",
"title": "errorMessage"
},
"id": {
"description": "The ID of the LLM Test configuration.",
"title": "id",
"type": "string"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"lastUpdateDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateDate"
},
"lastUpdateUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateUserId"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "The name of the LLM Test configuration.",
"title": "name",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, the use case ID associated with the LLM Test configuration.",
"title": "useCaseId"
},
"warnings": {
"description": "Warnings for this LLM test configuration.",
"items": {
"additionalProperties": {
"type": "string"
},
"propertyNames": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"type": "object"
},
"title": "warnings",
"type": "array"
}
},
"required": [
"id",
"name",
"description",
"datasetEvaluations",
"llmTestGradingCriteria",
"isOutOfTheBoxTestConfiguration",
"warnings"
],
"title": "LLMTestConfigurationResponse",
"type": "object"
}
LLMTestConfigurationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationDate | any | false | The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string(date-time) | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationUserId | any | false | The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetEvaluations | [DatasetEvaluationResponse] | true | The LLM test dataset evaluations. | |
| description | string | true | The description of the LLM Test configuration. | |
| errorMessage | any | false | The error message associated with the LLM test configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| id | string | true | The ID of the LLM Test configuration. | |
| isOutOfTheBoxTestConfiguration | boolean | true | Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration. | |
| lastUpdateDate | any | false | The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string(date-time) | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| lastUpdateUserId | any | false | The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmTestGradingCriteria | LLMTestGradingCriteria | true | The LLM test grading criteria. | |
| name | string | true | The name of the LLM Test configuration. | |
| useCaseId | any | false | If specified, the use case ID associated with the LLM Test configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| warnings | [object] | true | Warnings for this LLM test configuration. | |
| » additionalProperties | string | false | none |
LLMTestConfigurationSupportedInsightsResponse
{
"description": "Response model for supported insights.",
"properties": {
"datasetsCompatibility": {
"description": "The list of insight to evaluation datasets compatibility.",
"items": {
"description": "Insight to evaluation datasets compatibility.",
"properties": {
"incompatibleDatasets": {
"description": "The list of incompatible datasets.",
"items": {
"description": "Dataset identifier.",
"properties": {
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the dataset, if any.",
"title": "datasetId"
},
"datasetName": {
"description": "The name of the dataset.",
"title": "datasetName",
"type": "string"
}
},
"required": [
"datasetName",
"datasetId"
],
"title": "DatasetIdentifier",
"type": "object"
},
"title": "incompatibleDatasets",
"type": "array"
},
"insightName": {
"description": "The name of the insight.",
"title": "insightName",
"type": "string"
}
},
"required": [
"insightName",
"incompatibleDatasets"
],
"title": "InsightToEvalDatasetsCompatibility",
"type": "object"
},
"title": "datasetsCompatibility",
"type": "array"
},
"supportedInsightConfigurations": {
"description": "The list of supported insight configurations for the LLM Tests.",
"items": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"title": "supportedInsightConfigurations",
"type": "array"
}
},
"required": [
"supportedInsightConfigurations",
"datasetsCompatibility"
],
"title": "LLMTestConfigurationSupportedInsightsResponse",
"type": "object"
}
LLMTestConfigurationSupportedInsightsResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetsCompatibility | [InsightToEvalDatasetsCompatibility] | true | The list of insight to evaluation datasets compatibility. | |
| supportedInsightConfigurations | [InsightsConfigurationWithAdditionalData] | true | The list of supported insight configurations for the LLM Tests. |
LLMTestConfigurationType
{
"description": "Type of LLMTestConfiguration.",
"enum": [
"ootb",
"custom"
],
"title": "LLMTestConfigurationType",
"type": "string"
}
LLMTestConfigurationType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| LLMTestConfigurationType | string | false | Type of LLMTestConfiguration. |
Enumerated Values¶
| Property | Value |
|---|---|
| LLMTestConfigurationType | [ootb, custom] |
LLMTestGradingCriteria
{
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
}
LLMTestGradingCriteria
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| passThreshold | integer | true | maximum: 100 minimum: 0 |
The percentage threshold for Pass results across dataset-insight pairs. |
LLMTestResultResponse
{
"description": "API response object for a single LLMTestResult.",
"properties": {
"creationDate": {
"description": "LLM test result creation date (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "ID of the user that created this LLM test result.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user who created this LLM result.",
"title": "creationUserName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message if the LLM Test Result failed.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error resolution message if the LLM Test Result failed.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result based on the llm test grading criteria. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "LLM test result ID.",
"title": "id",
"type": "string"
},
"insightEvaluationResults": {
"description": "The Insight evaluation results.",
"items": {
"description": "API response object for a single InsightEvaluationResult.",
"properties": {
"aggregationType": {
"anyOf": [
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Aggregation type."
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "Aggregation value. None indicates that the aggregation failed.",
"title": "aggregationValue"
},
"chatId": {
"description": "Chat ID.",
"title": "chatId",
"type": "string"
},
"chatName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Chat name.",
"title": "chatName"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Custom Model LLM Validation ID if using custom model LLM.",
"title": "customModelLLMValidationId"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset configuration ID.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "Evaluation name.",
"maxLength": 5000,
"title": "evaluationName",
"type": "string"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "Insight evaluation result ID.",
"title": "id",
"type": "string"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"lastUpdateDate": {
"description": "Last update date of the insight evaluation result (ISO 8601 formatted).",
"format": "date-time",
"title": "lastUpdateDate",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM ID used for this insight evaluation result.",
"title": "llmId"
},
"llmTestResultId": {
"description": "LLM test result ID this insight evaluation result is associated to.",
"title": "llmTestResultId",
"type": "string"
},
"maxNumPrompts": {
"description": "Number of prompts used in evaluation.",
"title": "maxNumPrompts",
"type": "integer"
},
"metricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the metric.",
"title": "metricName"
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"id",
"llmTestResultId",
"maxNumPrompts",
"promptSamplingStrategy",
"chatId",
"chatName",
"evaluationName",
"insightGradingCriteria",
"lastUpdateDate"
],
"title": "InsightEvaluationResultResponse",
"type": "object"
},
"title": "insightEvaluationResults",
"type": "array"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"llmBlueprintId": {
"description": "LLM Blueprint ID.",
"title": "llmBlueprintId",
"type": "string"
},
"llmBlueprintSnapshot": {
"description": "A snapshot in time of a LLMBlueprint's functional parameters.",
"properties": {
"description": {
"description": "The description of the LLMBlueprint at the time of snapshotting.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLMBlueprint for which the snapshot was produced.",
"title": "id",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM selected for this LLM blueprint.",
"title": "llmId"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of LLM settings.",
"title": "llmSettings"
},
"name": {
"description": "The name of the LLMBlueprint at the time of snapshotting.",
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The playground id of the LLMBlueprint.",
"title": "playgroundId",
"type": "string"
},
"promptType": {
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
},
"snapshotDate": {
"description": "The date when the snapshot was produced.",
"format": "date-time",
"title": "snapshotDate",
"type": "string"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database linked to this LLM blueprint.",
"title": "vectorDatabaseId"
},
"vectorDatabaseSettings": {
"anyOf": [
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of vector database settings."
}
},
"required": [
"id",
"name",
"description",
"playgroundId",
"promptType"
],
"title": "LLMBlueprintSnapshot",
"type": "object"
},
"llmTestConfigurationId": {
"description": "LLM test configuration ID this LLM result is associated to.",
"title": "llmTestConfigurationId",
"type": "string"
},
"llmTestConfigurationName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the LLM test configuration this LLM result is associated to.",
"title": "llmTestConfigurationName"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"llmTestSuiteId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM test suite ID to which the LLM test configuration is associated to.",
"title": "llmTestSuiteId"
},
"passPercentage": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "The percentage of underlying insight evaluation results that have a PASS grading result. If not specified, execution status is not COMPLETED.",
"title": "passPercentage"
},
"useCaseId": {
"description": "Use case ID this LLM test result belongs to.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"llmTestConfigurationId",
"llmTestConfigurationName",
"isOutOfTheBoxTestConfiguration",
"useCaseId",
"llmBlueprintId",
"llmBlueprintSnapshot",
"llmTestGradingCriteria",
"executionStatus",
"insightEvaluationResults",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "LLMTestResultResponse",
"type": "object"
}
LLMTestResultResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationDate | string(date-time) | true | LLM test result creation date (ISO 8601 formatted). | |
| creationUserId | string | true | ID of the user that created this LLM test result. | |
| creationUserName | string | true | The name of the user who created this LLM result. | |
| errorMessage | any | false | The error message if the LLM Test Result failed. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorResolution | any | false | The error resolution message if the LLM Test Result failed. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| executionStatus | ExecutionStatus | true | The LLM Test execution status. | |
| gradingResult | any | false | The grading result based on the llm test grading criteria. If not specified, execution status is not COMPLETED. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | GradingResult | false | Grading result. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| id | string | true | LLM test result ID. | |
| insightEvaluationResults | [InsightEvaluationResultResponse] | true | The Insight evaluation results. | |
| isOutOfTheBoxTestConfiguration | boolean | true | Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration. | |
| llmBlueprintId | string | true | LLM Blueprint ID. | |
| llmBlueprintSnapshot | LLMBlueprintSnapshot | true | A snapshot of the llm blueprint entity at the time of LLM Test execution. | |
| llmTestConfigurationId | string | true | LLM test configuration ID this LLM result is associated to. | |
| llmTestConfigurationName | any | true | Name of the LLM test configuration this LLM result is associated to. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmTestGradingCriteria | LLMTestGradingCriteria | true | LLM test grading criteria. | |
| llmTestSuiteId | any | false | LLM test suite ID to which the LLM test configuration is associated to. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| passPercentage | any | false | The percentage of underlying insight evaluation results that have a PASS grading result. If not specified, execution status is not COMPLETED. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | number | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| useCaseId | string | true | Use case ID this LLM test result belongs to. |
LLMTestSuiteResponse
{
"description": "LLMTestSuite object formatted for API output.",
"properties": {
"creationDate": {
"description": "The creation date of the chat (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the chat.",
"title": "creationUserId",
"type": "string"
},
"description": {
"description": "The description of the LLM test suite.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLM test suite.",
"title": "id",
"type": "string"
},
"llmTestConfigurationIds": {
"description": "The IDs of the LLM test configurations in this LLM test suite.",
"items": {
"type": "string"
},
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case associated with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"name",
"description",
"useCaseId",
"llmTestConfigurationIds",
"creationDate",
"creationUserId"
],
"title": "LLMTestSuiteResponse",
"type": "object"
}
LLMTestSuiteResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationDate | string(date-time) | true | The creation date of the chat (ISO 8601 formatted). | |
| creationUserId | string | true | The ID of the user that created the chat. | |
| description | string | true | The description of the LLM test suite. | |
| id | string | true | The ID of the LLM test suite. | |
| llmTestConfigurationIds | [string] | true | The IDs of the LLM test configurations in this LLM test suite. | |
| name | string | true | The name of the LLM test suite. | |
| useCaseId | string | true | The ID of the use case associated with the LLM test suite. |
ListCustomModelValidationSortQueryParam
{
"description": "Sort order values for listing custom model validations.",
"enum": [
"name",
"-name",
"deploymentName",
"-deploymentName",
"userName",
"-userName",
"creationDate",
"-creationDate"
],
"title": "ListCustomModelValidationSortQueryParam",
"type": "string"
}
ListCustomModelValidationSortQueryParam
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ListCustomModelValidationSortQueryParam | string | false | Sort order values for listing custom model validations. |
Enumerated Values¶
| Property | Value |
|---|---|
| ListCustomModelValidationSortQueryParam | [name, -name, deploymentName, -deploymentName, userName, -userName, creationDate, -creationDate] |
ListEvaluationDatasetConfigurationResponse
{
"description": "Paginated list of evaludation dataset configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single evaluation dataset configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "EvaluationDatasetConfigurationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetConfigurationResponse",
"type": "object"
}
ListEvaluationDatasetConfigurationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [EvaluationDatasetConfigurationResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListEvaluationDatasetConfigurationsQueryParam
{
"description": "Sort order values for listing evaluation dataset configurations.",
"enum": [
"name",
"-name",
"creationUserId",
"-creationUserId",
"creationDate",
"-creationDate",
"datasetId",
"-datasetId",
"userName",
"-userName",
"datasetName",
"-datasetName",
"promptColumnName",
"-promptColumnName",
"responseColumnName",
"-responseColumnName"
],
"title": "ListEvaluationDatasetConfigurationsQueryParam",
"type": "string"
}
ListEvaluationDatasetConfigurationsQueryParam
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ListEvaluationDatasetConfigurationsQueryParam | string | false | Sort order values for listing evaluation dataset configurations. |
Enumerated Values¶
| Property | Value |
|---|---|
| ListEvaluationDatasetConfigurationsQueryParam | [name, -name, creationUserId, -creationUserId, creationDate, -creationDate, datasetId, -datasetId, userName, -userName, datasetName, -datasetName, promptColumnName, -promptColumnName, responseColumnName, -responseColumnName] |
ListEvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse
{
"description": "Paginated list of evaluation dataset metric aggregations, aggregated by LLM blueprint.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for multiple evaluation dataset metric aggregation\naggregated by llm blueprint.",
"properties": {
"aggregatedItemCount": {
"description": "Number of items aggregated.",
"title": "aggregatedItemCount",
"type": "integer"
},
"aggregatedItemDetails": {
"description": "List of details for aggregated items.",
"items": {
"description": "Details for aggregated items.",
"properties": {
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "EvaluationDatasetMetricAggregationChatDetails",
"type": "object"
},
"title": "aggregatedItemDetails",
"type": "array"
},
"aggregationType": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"items": {
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregated value of the metric.",
"title": "aggregationValue"
},
"customModelGuardId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model's guard the metric aggregation belongs to.",
"title": "customModelGuardId"
},
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The dataset ID of the evaluation dataset configuration.",
"title": "datasetId"
},
"datasetName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The Data Registry dataset name of the evaluation dataset configuration.",
"title": "datasetName"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration associated with the metric aggregation.",
"title": "evaluationDatasetConfigurationId"
},
"llmBlueprintId": {
"description": "The ID of the LLM blueprint associated with the metric aggregation.",
"title": "llmBlueprintId",
"type": "string"
},
"metricName": {
"description": "The name of the metric associated with the metric aggregation.",
"title": "metricName",
"type": "string"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the out-of-the-box dataset."
},
"tenantId": {
"description": "The ID of the tenant the metric aggregation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
}
},
"required": [
"llmBlueprintId",
"evaluationDatasetConfigurationId",
"ootbDatasetName",
"datasetId",
"datasetName",
"metricName",
"aggregationValue",
"aggregationType",
"tenantId",
"customModelGuardId",
"aggregatedItemDetails",
"aggregatedItemCount"
],
"title": "EvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse",
"type": "object"
}
ListEvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [EvaluationDatasetMetricAggregationAggregatedByLLMBlueprintResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListEvaluationDatasetMetricAggregationResponse
{
"description": "Paginated list of evaluation dataset metric aggregations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single evaluation dataset metric aggregation.",
"properties": {
"aggregationType": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"items": {
"description": "Aggregated record of multiple of the same item across different metric aggregation runs.",
"properties": {
"count": {
"description": "The number of metric aggregation items aggregated.",
"title": "count",
"type": "integer"
},
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value",
"count"
],
"title": "AggregatedAggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregated value of the metric.",
"title": "aggregationValue"
},
"chatId": {
"description": "The ID of the chat associated with the metric aggregation.",
"title": "chatId",
"type": "string"
},
"chatLink": {
"description": "The link to the chat associated with the metric aggregation.",
"title": "chatLink",
"type": "string"
},
"chatName": {
"description": "The name of the chat associated with the metric aggregation.",
"title": "chatName",
"type": "string"
},
"creationDate": {
"description": "The creation date of the metric aggregation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the metric aggregation.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user that created the metric aggregation.",
"title": "creationUserName",
"type": "string"
},
"customModelGuardId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model's guard the metric aggregation belongs to.",
"title": "customModelGuardId"
},
"datasetId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The dataset ID of the evaluation dataset configuration.",
"title": "datasetId"
},
"datasetName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The Data Registry dataset name of the evaluation dataset configuration.",
"title": "datasetName"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration associated with the metric aggregation.",
"title": "evaluationDatasetConfigurationId"
},
"llmBlueprintId": {
"description": "The ID of the LLM blueprint associated with the metric aggregation.",
"title": "llmBlueprintId",
"type": "string"
},
"metricName": {
"description": "The name of the metric associated with the metric aggregation.",
"title": "metricName",
"type": "string"
},
"ootbDatasetName": {
"anyOf": [
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the out-of-the-box dataset."
},
"tenantId": {
"description": "The ID of the tenant the metric aggregation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
}
},
"required": [
"chatId",
"chatName",
"chatLink",
"creationDate",
"creationUserId",
"creationUserName",
"llmBlueprintId",
"evaluationDatasetConfigurationId",
"ootbDatasetName",
"datasetId",
"datasetName",
"metricName",
"aggregationValue",
"aggregationType",
"tenantId",
"customModelGuardId"
],
"title": "EvaluationDatasetMetricAggregationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetMetricAggregationResponse",
"type": "object"
}
ListEvaluationDatasetMetricAggregationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [EvaluationDatasetMetricAggregationResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListEvaluationDatasetMetricAggregationSortQueryParam
{
"description": "Sort order values for listing evaluation dataset metric aggregations.",
"enum": [
"metricName",
"-metricName",
"aggregationValue",
"-aggregationValue",
"datasetName",
"-datasetName",
"datasetId",
"-datasetId",
"creationUserId",
"-creationUserId",
"creationUserName",
"-creationUserName",
"creationDate",
"-creationDate",
"evaluationDatasetConfigurationId",
"-evaluationDatasetConfigurationId"
],
"title": "ListEvaluationDatasetMetricAggregationSortQueryParam",
"type": "string"
}
ListEvaluationDatasetMetricAggregationSortQueryParam
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ListEvaluationDatasetMetricAggregationSortQueryParam | string | false | Sort order values for listing evaluation dataset metric aggregations. |
Enumerated Values¶
| Property | Value |
|---|---|
| ListEvaluationDatasetMetricAggregationSortQueryParam | [metricName, -metricName, aggregationValue, -aggregationValue, datasetName, -datasetName, datasetId, -datasetId, creationUserId, -creationUserId, creationUserName, -creationUserName, creationDate, -creationDate, evaluationDatasetConfigurationId, -evaluationDatasetConfigurationId] |
ListEvaluationDatasetMetricAggregationUniqueFieldValuesResponse
{
"description": "Paginated list of evaluation dataset metric aggregations with unique computed metrics.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single unique computed metric.",
"properties": {
"uniqueFieldValue": {
"description": "The unique value associated with the metric aggregation.",
"title": "uniqueFieldValue",
"type": "string"
}
},
"required": [
"uniqueFieldValue"
],
"title": "EvaluationDatasetMetricAggregationUniqueFieldValuesResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListEvaluationDatasetMetricAggregationUniqueFieldValuesResponse",
"type": "object"
}
ListEvaluationDatasetMetricAggregationUniqueFieldValuesResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [EvaluationDatasetMetricAggregationUniqueFieldValuesResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListLLMTestConfigurationNonOOTBDatasetsResponse
{
"description": "Paginated list of non-OOTB datasets for use with LLM test configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "Non out-of-the-box dataset used with an LLM test configuration.",
"properties": {
"agentGoalsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected agent goals (For agentic workflows).",
"title": "agentGoalsColumnName"
},
"correctnessEnabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"deprecated": true,
"description": "Whether correctness is enabled for the evaluation dataset configuration.",
"title": "correctnessEnabled"
},
"creationDate": {
"description": "The creation date of the evaluation dataset configuration (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the evaluation dataset configuration.",
"title": "creationUserId",
"type": "string"
},
"datasetId": {
"description": "The ID of the evaluation dataset.",
"title": "datasetId",
"type": "string"
},
"datasetName": {
"description": "The name of the evaluation dataset.",
"title": "datasetName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration.",
"title": "errorMessage"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"id": {
"description": "The ID of the evaluation dataset configuration.",
"title": "id",
"type": "string"
},
"name": {
"description": "The name of the evaluation dataset configuration.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the evaluation dataset configuration.",
"title": "playgroundId"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The rows count of the evaluation dataset.",
"title": "rowsCount",
"type": "integer"
},
"size": {
"description": "The size of the evaluation dataset (in bytes).",
"title": "size",
"type": "integer"
},
"tenantId": {
"description": "The ID of the DataRobot tenant this evaluation dataset configuration belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"toolCallsColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the dataset column containing expected tool calls (For agentic workflows).",
"title": "toolCallsColumnName"
},
"useCaseId": {
"description": "The ID of the use case associated with the evaluation dataset configuration.",
"title": "useCaseId",
"type": "string"
},
"userName": {
"description": "The name of the user that created the evaluation dataset configuration.",
"title": "userName",
"type": "string"
}
},
"required": [
"id",
"name",
"size",
"rowsCount",
"useCaseId",
"playgroundId",
"datasetId",
"datasetName",
"promptColumnName",
"responseColumnName",
"toolCallsColumnName",
"agentGoalsColumnName",
"userName",
"correctnessEnabled",
"creationUserId",
"creationDate",
"tenantId",
"executionStatus"
],
"title": "LLMTestConfigurationNonOOTBDatasetResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestConfigurationNonOOTBDatasetsResponse",
"type": "object"
}
ListLLMTestConfigurationNonOOTBDatasetsResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [LLMTestConfigurationNonOOTBDatasetResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListLLMTestConfigurationOOTBDatasetsResponse
{
"description": "Paginated list of OOTB datasets for use with LLM test configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "Out-of-the-box dataset used with an LLM test configuration.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "LLMTestConfigurationOOTBDatasetResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestConfigurationOOTBDatasetsResponse",
"type": "object"
}
ListLLMTestConfigurationOOTBDatasetsResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [LLMTestConfigurationOOTBDatasetResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListLLMTestConfigurationsResponse
{
"description": "Paginated list of LLM test configurations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single LLMTestConfiguration.",
"properties": {
"creationDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The creation date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationDate"
},
"creationUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who created the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "creationUserId"
},
"datasetEvaluations": {
"description": "The LLM test dataset evaluations.",
"items": {
"description": "Dataset evaluation.",
"properties": {
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the dataset evaluation.",
"title": "errorMessage"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration for this dataset evaluation.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "The name of the evaluation. This name should provide context regarding what is being evaluated.",
"maxLength": 5000,
"minLength": 1,
"title": "evaluationName",
"type": "string"
},
"insightConfiguration": {
"description": "The configuration of insights with extra data.",
"properties": {
"aggregationTypes": {
"anyOf": [
{
"items": {
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The aggregation types used in the insights configuration.",
"title": "aggregationTypes"
},
"costConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the cost configuration.",
"title": "costConfigurationId"
},
"customMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom metric (if using a custom metric).",
"title": "customMetricId"
},
"customModelGuard": {
"anyOf": [
{
"description": "Details of a guard as defined for the custom model.",
"properties": {
"name": {
"description": "The name of the guard.",
"maxLength": 5000,
"minLength": 1,
"title": "name",
"type": "string"
},
"ootbType": {
"anyOf": [
{
"description": "OOTB type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"token_count",
"rouge_1",
"faithfulness",
"agent_goal_accuracy",
"custom_metric",
"cost",
"task_adherence"
],
"title": "CustomModelGuardOOTBType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Out of the box type of the guard."
},
"stage": {
"description": "Guard stage as used in the moderation_config.yaml file of the custom model.",
"enum": [
"prompt",
"response"
],
"title": "CustomModelGuardStage",
"type": "string"
},
"type": {
"description": "Guard type as used in the moderation_config.yaml file of the custom model.",
"enum": [
"ootb",
"model",
"nemo_guardrails"
],
"title": "CustomModelGuardType",
"type": "string"
}
},
"required": [
"type",
"stage",
"name"
],
"title": "CustomModelGuard",
"type": "object"
},
{
"type": "null"
}
],
"description": "Guard as configured in the custom model."
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation if using a custom model LLM for OOTB metrics.",
"title": "customModelLLMValidationId"
},
"deploymentId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model deployment associated with the insight.",
"title": "deploymentId"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the evaluation dataset configuration or sidecar model metric validation or OOTB metric.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the evaluation dataset configuration.",
"title": "evaluationDatasetConfigurationId"
},
"executionStatus": {
"anyOf": [
{
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
{
"type": "null"
}
],
"description": "The execution status of the evaluation dataset configuration."
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"insightName": {
"description": "The name of the insight.",
"maxLength": 5000,
"minLength": 1,
"title": "insightName",
"type": "string"
},
"insightType": {
"anyOf": [
{
"description": "The type of insight.",
"enum": [
"Reference",
"Quality metric",
"Operational metric",
"Evaluation deployment",
"Custom metric",
"Nemo"
],
"title": "InsightTypes",
"type": "string"
},
{
"type": "null"
}
],
"description": "The type of the insight."
},
"isTransferable": {
"default": false,
"description": "Indicates if insight can be transferred to production.",
"title": "isTransferable",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The LLM ID for OOTB metrics that use LLMs.",
"title": "llmId"
},
"llmIsActive": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is active.",
"title": "llmIsActive"
},
"llmIsDeprecated": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"description": "Whether the LLM is deprecated and will be removed in a future release.",
"title": "llmIsDeprecated"
},
"modelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the model associated with `deploymentId`.",
"title": "modelId"
},
"modelPackageRegisteredModelId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the registered model package associated with `deploymentId`.",
"title": "modelPackageRegisteredModelId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
},
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the insight configuration.",
"title": "moderationConfiguration"
},
"nemoMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the Nemo configuration.",
"title": "nemoMetricId"
},
"ootbMetricId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the ootb metric (if using an ootb metric).",
"title": "ootbMetricId"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
},
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
},
{
"type": "null"
}
],
"description": "The OOTB metric name.",
"title": "ootbMetricName"
},
"resultUnit": {
"anyOf": [
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
},
{
"type": "null"
}
],
"description": "The unit of measurement associated with the insight result."
},
"sidecarModelMetricMetadata": {
"anyOf": [
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
},
{
"type": "null"
}
],
"description": "The metadata of the sidecar model metric (if using a sidecar model metric)."
},
"sidecarModelMetricValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the sidecar model metric validation (if using a sidecar model metric).",
"title": "sidecarModelMetricValidationId"
},
"stage": {
"anyOf": [
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
},
{
"type": "null"
}
],
"description": "The stage (prompt or response) where insight is calculated at."
}
},
"required": [
"insightName",
"aggregationTypes"
],
"title": "InsightsConfigurationWithAdditionalData",
"type": "object"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"maxNumPrompts": {
"default": 100,
"description": "The max number of prompts to evaluate.",
"exclusiveMinimum": 0,
"maximum": 5000,
"title": "maxNumPrompts",
"type": "integer"
},
"ootbDataset": {
"anyOf": [
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
},
{
"type": "null"
}
],
"description": "Out-of-the-box evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"evaluationName",
"insightConfiguration",
"insightGradingCriteria",
"evaluationDatasetName"
],
"title": "DatasetEvaluationResponse",
"type": "object"
},
"title": "datasetEvaluations",
"type": "array"
},
"description": {
"description": "The description of the LLM Test configuration.",
"title": "description",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the LLM test configuration.",
"title": "errorMessage"
},
"id": {
"description": "The ID of the LLM Test configuration.",
"title": "id",
"type": "string"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"lastUpdateDate": {
"anyOf": [
{
"format": "date-time",
"type": "string"
},
{
"type": "null"
}
],
"description": "The last update date of the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateDate"
},
"lastUpdateUserId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the user who last updated the LLM Test configuration. For OOTB LLM Test configurations this is null.",
"title": "lastUpdateUserId"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"name": {
"description": "The name of the LLM Test configuration.",
"title": "name",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "If specified, the use case ID associated with the LLM Test configuration.",
"title": "useCaseId"
},
"warnings": {
"description": "Warnings for this LLM test configuration.",
"items": {
"additionalProperties": {
"type": "string"
},
"propertyNames": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"type": "object"
},
"title": "warnings",
"type": "array"
}
},
"required": [
"id",
"name",
"description",
"datasetEvaluations",
"llmTestGradingCriteria",
"isOutOfTheBoxTestConfiguration",
"warnings"
],
"title": "LLMTestConfigurationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestConfigurationsResponse",
"type": "object"
}
ListLLMTestConfigurationsResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [LLMTestConfigurationResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListLLMTestResultResponse
{
"description": "Paginated list of LLM test results.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single LLMTestResult.",
"properties": {
"creationDate": {
"description": "LLM test result creation date (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "ID of the user that created this LLM test result.",
"title": "creationUserId",
"type": "string"
},
"creationUserName": {
"description": "The name of the user who created this LLM result.",
"title": "creationUserName",
"type": "string"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message if the LLM Test Result failed.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error resolution message if the LLM Test Result failed.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result based on the llm test grading criteria. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "LLM test result ID.",
"title": "id",
"type": "string"
},
"insightEvaluationResults": {
"description": "The Insight evaluation results.",
"items": {
"description": "API response object for a single InsightEvaluationResult.",
"properties": {
"aggregationType": {
"anyOf": [
{
"description": "The type of the metric aggregation.",
"enum": [
"average",
"percentYes",
"classPercentCoverage",
"ngramImportance",
"guardConditionPercentYes"
],
"title": "AggregationType",
"type": "string"
},
{
"type": "null"
}
],
"description": "Aggregation type."
},
"aggregationValue": {
"anyOf": [
{
"type": "number"
},
{
"items": {
"description": "An individual record in an itemized metric aggregation.",
"properties": {
"item": {
"description": "The name of the item.",
"title": "item",
"type": "string"
},
"value": {
"description": "The value associated with the item.",
"title": "value",
"type": "number"
}
},
"required": [
"item",
"value"
],
"title": "AggregationValue",
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "Aggregation value. None indicates that the aggregation failed.",
"title": "aggregationValue"
},
"chatId": {
"description": "Chat ID.",
"title": "chatId",
"type": "string"
},
"chatName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Chat name.",
"title": "chatName"
},
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Custom Model LLM Validation ID if using custom model LLM.",
"title": "customModelLLMValidationId"
},
"evaluationDatasetConfigurationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset configuration ID.",
"title": "evaluationDatasetConfigurationId"
},
"evaluationDatasetName": {
"anyOf": [
{
"maxLength": 5000,
"type": "string"
},
{
"type": "null"
}
],
"description": "Evaluation dataset name.",
"title": "evaluationDatasetName"
},
"evaluationName": {
"description": "Evaluation name.",
"maxLength": 5000,
"title": "evaluationName",
"type": "string"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"gradingResult": {
"anyOf": [
{
"description": "Grading result.",
"enum": [
"PASS",
"FAIL"
],
"title": "GradingResult",
"type": "string"
},
{
"type": "null"
}
],
"description": "The grading result for this insight evaluation result. If not specified, execution status is not COMPLETED."
},
"id": {
"description": "Insight evaluation result ID.",
"title": "id",
"type": "string"
},
"insightGradingCriteria": {
"description": "Grading criteria for an insight.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass result. Greater than or equal to this threshold indicates a Pass.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "InsightGradingCriteria",
"type": "object"
},
"lastUpdateDate": {
"description": "Last update date of the insight evaluation result (ISO 8601 formatted).",
"format": "date-time",
"title": "lastUpdateDate",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM ID used for this insight evaluation result.",
"title": "llmId"
},
"llmTestResultId": {
"description": "LLM test result ID this insight evaluation result is associated to.",
"title": "llmTestResultId",
"type": "string"
},
"maxNumPrompts": {
"description": "Number of prompts used in evaluation.",
"title": "maxNumPrompts",
"type": "integer"
},
"metricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the metric.",
"title": "metricName"
},
"promptSamplingStrategy": {
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
},
"required": [
"id",
"llmTestResultId",
"maxNumPrompts",
"promptSamplingStrategy",
"chatId",
"chatName",
"evaluationName",
"insightGradingCriteria",
"lastUpdateDate"
],
"title": "InsightEvaluationResultResponse",
"type": "object"
},
"title": "insightEvaluationResults",
"type": "array"
},
"isOutOfTheBoxTestConfiguration": {
"description": "Identifies the LLM Test configuration as an out-of-the-box (OOTB) test configuration.",
"title": "isOutOfTheBoxTestConfiguration",
"type": "boolean"
},
"llmBlueprintId": {
"description": "LLM Blueprint ID.",
"title": "llmBlueprintId",
"type": "string"
},
"llmBlueprintSnapshot": {
"description": "A snapshot in time of a LLMBlueprint's functional parameters.",
"properties": {
"description": {
"description": "The description of the LLMBlueprint at the time of snapshotting.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLMBlueprint for which the snapshot was produced.",
"title": "id",
"type": "string"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM selected for this LLM blueprint.",
"title": "llmId"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of LLM settings.",
"title": "llmSettings"
},
"name": {
"description": "The name of the LLMBlueprint at the time of snapshotting.",
"title": "name",
"type": "string"
},
"playgroundId": {
"description": "The playground id of the LLMBlueprint.",
"title": "playgroundId",
"type": "string"
},
"promptType": {
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
},
"snapshotDate": {
"description": "The date when the snapshot was produced.",
"format": "date-time",
"title": "snapshotDate",
"type": "string"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database linked to this LLM blueprint.",
"title": "vectorDatabaseId"
},
"vectorDatabaseSettings": {
"anyOf": [
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "A key/value dictionary of vector database settings."
}
},
"required": [
"id",
"name",
"description",
"playgroundId",
"promptType"
],
"title": "LLMBlueprintSnapshot",
"type": "object"
},
"llmTestConfigurationId": {
"description": "LLM test configuration ID this LLM result is associated to.",
"title": "llmTestConfigurationId",
"type": "string"
},
"llmTestConfigurationName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "Name of the LLM test configuration this LLM result is associated to.",
"title": "llmTestConfigurationName"
},
"llmTestGradingCriteria": {
"description": "Grading criteria for the LLM Test configuration.",
"properties": {
"passThreshold": {
"description": "The percentage threshold for Pass results across dataset-insight pairs.",
"maximum": 100,
"minimum": 0,
"title": "passThreshold",
"type": "integer"
}
},
"required": [
"passThreshold"
],
"title": "LLMTestGradingCriteria",
"type": "object"
},
"llmTestSuiteId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "LLM test suite ID to which the LLM test configuration is associated to.",
"title": "llmTestSuiteId"
},
"passPercentage": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "The percentage of underlying insight evaluation results that have a PASS grading result. If not specified, execution status is not COMPLETED.",
"title": "passPercentage"
},
"useCaseId": {
"description": "Use case ID this LLM test result belongs to.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"llmTestConfigurationId",
"llmTestConfigurationName",
"isOutOfTheBoxTestConfiguration",
"useCaseId",
"llmBlueprintId",
"llmBlueprintSnapshot",
"llmTestGradingCriteria",
"executionStatus",
"insightEvaluationResults",
"creationDate",
"creationUserId",
"creationUserName"
],
"title": "LLMTestResultResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestResultResponse",
"type": "object"
}
ListLLMTestResultResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [LLMTestResultResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListLLMTestSuiteSortQueryParam
{
"description": "Sort order values for listing chats.",
"enum": [
"name",
"-name",
"creationDate",
"-creationDate"
],
"title": "ListLLMTestSuiteSortQueryParam",
"type": "string"
}
ListLLMTestSuiteSortQueryParam
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ListLLMTestSuiteSortQueryParam | string | false | Sort order values for listing chats. |
Enumerated Values¶
| Property | Value |
|---|---|
| ListLLMTestSuiteSortQueryParam | [name, -name, creationDate, -creationDate] |
ListLLMTestSuitesResponse
{
"description": "Paginated list of LLM test suites.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "LLMTestSuite object formatted for API output.",
"properties": {
"creationDate": {
"description": "The creation date of the chat (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"creationUserId": {
"description": "The ID of the user that created the chat.",
"title": "creationUserId",
"type": "string"
},
"description": {
"description": "The description of the LLM test suite.",
"title": "description",
"type": "string"
},
"id": {
"description": "The ID of the LLM test suite.",
"title": "id",
"type": "string"
},
"llmTestConfigurationIds": {
"description": "The IDs of the LLM test configurations in this LLM test suite.",
"items": {
"type": "string"
},
"title": "llmTestConfigurationIds",
"type": "array"
},
"name": {
"description": "The name of the LLM test suite.",
"title": "name",
"type": "string"
},
"useCaseId": {
"description": "The ID of the use case associated with the LLM test suite.",
"title": "useCaseId",
"type": "string"
}
},
"required": [
"id",
"name",
"description",
"useCaseId",
"llmTestConfigurationIds",
"creationDate",
"creationUserId"
],
"title": "LLMTestSuiteResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListLLMTestSuitesResponse",
"type": "object"
}
ListLLMTestSuitesResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [LLMTestSuiteResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
ListSidecarModelMetricValidationnResponse
{
"description": "Paginated list of sidecar model metric validations.",
"properties": {
"count": {
"description": "The number of records on this page.",
"title": "count",
"type": "integer"
},
"data": {
"description": "The list of records.",
"items": {
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
},
"title": "data",
"type": "array"
},
"next": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the next page, or `null` if there is no such page.",
"title": "next"
},
"previous": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL to the previous page, or `null` if there is no such page.",
"title": "previous"
},
"totalCount": {
"description": "The total number of records.",
"title": "totalCount",
"type": "integer"
}
},
"required": [
"totalCount",
"count",
"next",
"previous",
"data"
],
"title": "ListSidecarModelMetricValidationnResponse",
"type": "object"
}
ListSidecarModelMetricValidationnResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| count | integer | true | The number of records on this page. | |
| data | [SidecarModelMetricValidationResponse] | true | The list of records. | |
| next | any | true | The URL to the next page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| previous | any | true | The URL to the previous page, or null if there is no such page. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| totalCount | integer | true | The total number of records. |
MetricUnit
{
"description": "The unit of measurement associated with a metric.",
"enum": [
"s",
"ms",
"%"
],
"title": "MetricUnit",
"type": "string"
}
MetricUnit
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| MetricUnit | string | false | The unit of measurement associated with a metric. |
Enumerated Values¶
| Property | Value |
|---|---|
| MetricUnit | [s, ms, %] |
ModerationAction
{
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
}
ModerationAction
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ModerationAction | string | false | The moderation strategy. |
Enumerated Values¶
| Property | Value |
|---|---|
| ModerationAction | [block, report, reportAndBlock] |
ModerationConfigurationWithID
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithID",
"type": "object"
}
ModerationConfigurationWithID
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| guardConditions | [GuardCondition] | true | maxItems: 1 minItems: 1 |
The guard conditions associated with a metric. |
| intervention | Intervention | true | The intervention specific moderation configuration. |
ModerationConfigurationWithoutID
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
}
ModerationConfigurationWithoutID
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| guardConditions | [GuardCondition] | true | maxItems: 1 minItems: 1 |
The guard conditions associated with a metric. |
| intervention | Intervention | true | The intervention specific moderation configuration. |
OOTBAgenticMetricInsightNames
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
}
OOTBAgenticMetricInsightNames
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| OOTBAgenticMetricInsightNames | string | false | The Out-Of-The-Box metric name that can be used in an Agentic playground. |
Enumerated Values¶
| Property | Value |
|---|---|
| OOTBAgenticMetricInsightNames | [tool_call_accuracy, agent_goal_accuracy_with_reference] |
OOTBDataset
{
"description": "Out-of-the-box dataset.",
"properties": {
"datasetName": {
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
},
"datasetUrl": {
"anyOf": [
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
},
{
"type": "null"
}
],
"description": "The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets."
},
"promptColumnName": {
"description": "The name of the prompt column.",
"maxLength": 5000,
"minLength": 1,
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"anyOf": [
{
"maxLength": 5000,
"minLength": 1,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the response column, if present.",
"title": "responseColumnName"
},
"rowsCount": {
"description": "The number rows in the dataset.",
"title": "rowsCount",
"type": "integer"
},
"warning": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Warning about the content of the dataset.",
"title": "warning"
}
},
"required": [
"datasetName",
"datasetUrl",
"promptColumnName",
"responseColumnName",
"rowsCount"
],
"title": "OOTBDataset",
"type": "object"
}
OOTBDataset
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetName | OOTBDatasetName | true | The name of the evaluation dataset. | |
| datasetUrl | any | true | The public URL of the evaluation dataset. This applies only to our predefined public evaluation datasets. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBDatasetUrl | false | Out-of-the-box dataset URL. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | string | true | maxLength: 5000 minLength: 1 minLength: 1 |
The name of the prompt column. |
| responseColumnName | any | true | The name of the response column, if present. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 5000 minLength: 1 minLength: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| rowsCount | integer | true | The number rows in the dataset. | |
| warning | any | false | Warning about the content of the dataset. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
OOTBDatasetName
{
"description": "Out-of-the-box dataset name.",
"enum": [
"jailbreak-v1.csv",
"bbq-lite-age-v1.csv",
"bbq-lite-gender-v1.csv",
"bbq-lite-race-ethnicity-v1.csv",
"bbq-lite-religion-v1.csv",
"bbq-lite-disability-status-v1.csv",
"bbq-lite-sexual-orientation-v1.csv",
"bbq-lite-nationality-v1.csv",
"bbq-lite-ses-v1.csv",
"completeness-parent-v1.csv",
"completeness-grandparent-v1.csv",
"completeness-great-grandparent-v1.csv",
"pii-v1.csv",
"toxicity-v2.csv",
"jbbq-age-v1.csv",
"jbbq-gender-identity-v1.csv",
"jbbq-physical-appearance-v1.csv",
"jbbq-disability-status-v1.csv",
"jbbq-sexual-orientation-v1.csv"
],
"title": "OOTBDatasetName",
"type": "string"
}
OOTBDatasetName
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| OOTBDatasetName | string | false | Out-of-the-box dataset name. |
Enumerated Values¶
| Property | Value |
|---|---|
| OOTBDatasetName | [jailbreak-v1.csv, bbq-lite-age-v1.csv, bbq-lite-gender-v1.csv, bbq-lite-race-ethnicity-v1.csv, bbq-lite-religion-v1.csv, bbq-lite-disability-status-v1.csv, bbq-lite-sexual-orientation-v1.csv, bbq-lite-nationality-v1.csv, bbq-lite-ses-v1.csv, completeness-parent-v1.csv, completeness-grandparent-v1.csv, completeness-great-grandparent-v1.csv, pii-v1.csv, toxicity-v2.csv, jbbq-age-v1.csv, jbbq-gender-identity-v1.csv, jbbq-physical-appearance-v1.csv, jbbq-disability-status-v1.csv, jbbq-sexual-orientation-v1.csv] |
OOTBDatasetUrl
{
"description": "Out-of-the-box dataset URL.",
"enum": [
"https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv",
"https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv"
],
"title": "OOTBDatasetUrl",
"type": "string"
}
OOTBDatasetUrl
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| OOTBDatasetUrl | string | false | Out-of-the-box dataset URL. |
Enumerated Values¶
| Property | Value |
|---|---|
| OOTBDatasetUrl | [https://s3.amazonaws.com/datarobot_public_datasets/genai/jailbreak-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-age-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-gender-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-race-ethnicity-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-religion-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-disability-status-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-sexual-orientation-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-nationality-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/bbq-lite-ses-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-parent-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-grandparent-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/completeness-great-grandparent-v1.csv, https://s3.amazonaws.com/datarobot_public_datasets/genai/pii-v1.csv] |
OOTBMetricConfigurationResponse
{
"description": "API response object for a single OOTB metric.",
"properties": {
"customModelLLMValidationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the custom model LLM validation (if using a custom model LLM).",
"title": "customModelLLMValidationId"
},
"customOotbMetricName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The custom OOTB metric name to be associated with the OOTB metric.",
"title": "customOotbMetricName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the OOTB metric configuration.",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"executionStatus": {
"description": "Job and entity execution status.",
"enum": [
"NEW",
"RUNNING",
"COMPLETED",
"REQUIRES_USER_INPUT",
"SKIPPED",
"ERROR"
],
"title": "ExecutionStatus",
"type": "string"
},
"extraMetricSettings": {
"anyOf": [
{
"description": "Extra settings for the metric that do not reference other entities.",
"properties": {
"toolCallAccuracy": {
"anyOf": [
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the tool call accuracy metric."
}
},
"title": "ExtraMetricSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "Extra settings for the metric that do not reference other entities."
},
"isAgentic": {
"default": false,
"description": "Whether the OOTB metric configuration is specific to agentic workflows.",
"title": "isAgentic",
"type": "boolean"
},
"llmId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the LLM to use for `correctness` and `faithfulness` metrics.",
"title": "llmId"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration to be associated with the OOTB metric."
},
"ootbMetricConfigurationId": {
"description": "The ID of OOTB metric.",
"title": "ootbMetricConfigurationId",
"type": "string"
},
"ootbMetricName": {
"anyOf": [
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
},
{
"description": "The Out-Of-The-Box metric name that can be used in an Agentic playground.",
"enum": [
"tool_call_accuracy",
"agent_goal_accuracy_with_reference"
],
"title": "OOTBAgenticMetricInsightNames",
"type": "string"
}
],
"description": "The name of the OOTB metric.",
"title": "ootbMetricName"
}
},
"required": [
"ootbMetricName",
"ootbMetricConfigurationId",
"executionStatus"
],
"title": "OOTBMetricConfigurationResponse",
"type": "object"
}
OOTBMetricConfigurationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customModelLLMValidationId | any | false | The ID of the custom model LLM validation (if using a custom model LLM). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| customOotbMetricName | any | false | The custom OOTB metric name to be associated with the OOTB metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorMessage | any | false | The error message associated with the OOTB metric configuration. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorResolution | any | false | The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [InsightErrorResolution] | false | [Error type linking directly to the field name that is related to the error.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| executionStatus | ExecutionStatus | true | The execution status of the OOTB metric configuration. | |
| extraMetricSettings | any | false | Extra settings for the metric that do not reference other entities. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ExtraMetricSettings | false | Extra settings for the metric that do not reference other entities. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| isAgentic | boolean | false | Whether the OOTB metric configuration is specific to agentic workflows. | |
| llmId | any | false | The ID of the LLM to use for correctness and faithfulness metrics. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| moderationConfiguration | any | false | The moderation configuration to be associated with the OOTB metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ModerationConfigurationWithoutID | false | Moderation Configuration associated with an insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| ootbMetricConfigurationId | string | true | The ID of OOTB metric. | |
| ootbMetricName | any | true | The name of the OOTB metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBMetricInsightNames | false | The Out-Of-The-Box metric name that can be used in the playground. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | OOTBAgenticMetricInsightNames | false | The Out-Of-The-Box metric name that can be used in an Agentic playground. |
OOTBMetricInsightNames
{
"description": "The Out-Of-The-Box metric name that can be used in the playground.",
"enum": [
"latency",
"citations",
"rouge_1",
"faithfulness",
"correctness",
"prompt_tokens",
"response_tokens",
"document_tokens",
"all_tokens",
"jailbreak_violation",
"toxicity_violation",
"pii_violation",
"exact_match",
"starts_with",
"contains"
],
"title": "OOTBMetricInsightNames",
"type": "string"
}
OOTBMetricInsightNames
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| OOTBMetricInsightNames | string | false | The Out-Of-The-Box metric name that can be used in the playground. |
Enumerated Values¶
| Property | Value |
|---|---|
| OOTBMetricInsightNames | [latency, citations, rouge_1, faithfulness, correctness, prompt_tokens, response_tokens, document_tokens, all_tokens, jailbreak_violation, toxicity_violation, pii_violation, exact_match, starts_with, contains] |
OTELMetricInsightNames
{
"description": "Metrics that can only be calculated using OTEL Trace/metric data.",
"enum": [
"agent_latency",
"agent_tokens"
],
"title": "OTELMetricInsightNames",
"type": "string"
}
OTELMetricInsightNames
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| OTELMetricInsightNames | string | false | Metrics that can only be calculated using OTEL Trace/metric data. |
Enumerated Values¶
| Property | Value |
|---|---|
| OTELMetricInsightNames | [agent_latency, agent_tokens] |
PipelineStage
{
"description": "Enum that describes at which stage the metric may be calculated.",
"enum": [
"prompt_pipeline",
"response_pipeline"
],
"title": "PipelineStage",
"type": "string"
}
PipelineStage
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| PipelineStage | string | false | Enum that describes at which stage the metric may be calculated. |
Enumerated Values¶
| Property | Value |
|---|---|
| PipelineStage | [prompt_pipeline, response_pipeline] |
PromptSamplingStrategy
{
"description": "The prompt sampling strategy for the evaluation dataset configuration.",
"enum": [
"random_without_replacement",
"first_n_rows"
],
"title": "PromptSamplingStrategy",
"type": "string"
}
PromptSamplingStrategy
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| PromptSamplingStrategy | string | false | The prompt sampling strategy for the evaluation dataset configuration. |
Enumerated Values¶
| Property | Value |
|---|---|
| PromptSamplingStrategy | [random_without_replacement, first_n_rows] |
PromptType
{
"description": "Determines whether chat history is submitted as context to the user prompt.",
"enum": [
"CHAT_HISTORY_AWARE",
"ONE_TIME_PROMPT"
],
"title": "PromptType",
"type": "string"
}
PromptType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| PromptType | string | false | Determines whether chat history is submitted as context to the user prompt. |
Enumerated Values¶
| Property | Value |
|---|---|
| PromptType | [CHAT_HISTORY_AWARE, ONE_TIME_PROMPT] |
RetrievalMode
{
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
}
RetrievalMode
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| RetrievalMode | string | false | Retrieval modes for vector databases. |
Enumerated Values¶
| Property | Value |
|---|---|
| RetrievalMode | [similarity, maximal_marginal_relevance] |
SidecarModelMetricMetadata
{
"description": "The metadata of a sidecar model metric.",
"properties": {
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName"
}
},
"required": [
"targetColumnName"
],
"title": "SidecarModelMetricMetadata",
"type": "object"
}
SidecarModelMetricMetadata
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| expectedResponseColumnName | any | false | The name of the column the custom model uses for expected response text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| promptColumnName | any | false | The name of the column the custom model uses for prompt text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| responseColumnName | any | false | The name of the column the custom model uses for response text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| targetColumnName | any | true | The name of the column the custom model uses for prediction output. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
SidecarModelMetricValidationResponse
{
"description": "API response object for a single sidecar model metric validation.",
"properties": {
"citationsPrefixColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The column name prefix the custom model uses for citation inputs.",
"title": "citationsPrefixColumnName"
},
"creationDate": {
"description": "The creation date of the custom model validation (ISO 8601 formatted).",
"format": "date-time",
"title": "creationDate",
"type": "string"
},
"deploymentAccessData": {
"anyOf": [
{
"description": "Add authorization_header to avoid breaking change to API.",
"properties": {
"authorizationHeader": {
"default": "[REDACTED]",
"description": "The `Authorization` header to use for the deployment.",
"title": "authorizationHeader",
"type": "string"
},
"chatApiUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The URL of the deployment's chat API.",
"title": "chatApiUrl"
},
"datarobotKey": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The server key associated with the prediction API.",
"title": "datarobotKey"
},
"inputType": {
"description": "The format of the input data submitted to a DataRobot deployment.",
"enum": [
"CSV",
"JSON"
],
"title": "DeploymentInputType",
"type": "string"
},
"modelType": {
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
},
"predictionApiUrl": {
"description": "The URL of the deployment's prediction API.",
"title": "predictionApiUrl",
"type": "string"
}
},
"required": [
"predictionApiUrl",
"datarobotKey",
"inputType",
"modelType"
],
"title": "DeploymentAccessData",
"type": "object"
},
{
"type": "null"
}
],
"description": "The parameters used for accessing the deployment."
},
"deploymentId": {
"description": "The ID of the custom model deployment.",
"title": "deploymentId",
"type": "string"
},
"deploymentName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the custom model deployment.",
"title": "deploymentName"
},
"errorMessage": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The error message associated with the validation error (if the validation failed).",
"title": "errorMessage"
},
"errorResolution": {
"anyOf": [
{
"items": {
"description": "Error type linking directly to the field name that is related to the error.",
"enum": [
"ootbMetricName",
"intervention",
"guardCondition",
"sidecarOverall",
"sidecarRevalidate",
"sidecarDeploymentId",
"sidecarInputColumnName",
"sidecarOutputColumnName",
"promptPipelineFiles",
"promptPipelineTemplateId",
"responsePipelineFiles",
"responsePipelineTemplateId"
],
"title": "InsightErrorResolution",
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"description": "The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any.",
"title": "errorResolution"
},
"expectedResponseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for expected response text input.",
"title": "expectedResponseColumnName"
},
"id": {
"description": "The ID of the custom model validation.",
"title": "id",
"type": "string"
},
"modelId": {
"description": "The ID of the model used in the deployment.",
"title": "modelId",
"type": "string"
},
"moderationConfiguration": {
"anyOf": [
{
"description": "Moderation Configuration associated with an insight.",
"properties": {
"guardConditions": {
"description": "The guard conditions associated with a metric.",
"items": {
"description": "The guard condition for a metric.",
"properties": {
"comparand": {
"anyOf": [
{
"type": "number"
},
{
"type": "string"
},
{
"type": "boolean"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"description": "The comparand(s) used in the guard condition.",
"title": "comparand"
},
"comparator": {
"description": "The comparator used in a guard condition.",
"enum": [
"greaterThan",
"lessThan",
"equals",
"notEquals",
"is",
"isNot",
"matches",
"doesNotMatch",
"contains",
"doesNotContain"
],
"title": "GuardConditionComparator",
"type": "string"
}
},
"required": [
"comparator",
"comparand"
],
"title": "GuardCondition",
"type": "object"
},
"maxItems": 1,
"minItems": 1,
"title": "guardConditions",
"type": "array"
},
"intervention": {
"description": "The intervention configuration for a metric.",
"properties": {
"action": {
"description": "The moderation strategy.",
"enum": [
"block",
"report",
"reportAndBlock"
],
"title": "ModerationAction",
"type": "string"
},
"message": {
"description": "The intervention message to replace the prediction when a guard condition is satisfied.",
"minLength": 1,
"title": "message",
"type": "string"
}
},
"required": [
"action",
"message"
],
"title": "Intervention",
"type": "object"
}
},
"required": [
"guardConditions",
"intervention"
],
"title": "ModerationConfigurationWithoutID",
"type": "object"
},
{
"type": "null"
}
],
"description": "The moderation configuration associated with the sidecar model metric."
},
"name": {
"description": "The name of the validated custom model.",
"title": "name",
"type": "string"
},
"playgroundId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the playground associated with the sidecar model metric validation.",
"title": "playgroundId"
},
"predictionTimeout": {
"description": "The timeout in seconds for the prediction API used in this custom model validation.",
"title": "predictionTimeout",
"type": "integer"
},
"promptColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for prompt text input.",
"title": "promptColumnName"
},
"responseColumnName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the column the custom model uses for response text input.",
"title": "responseColumnName"
},
"targetColumnName": {
"description": "The name of the column the custom model uses for prediction output.",
"title": "targetColumnName",
"type": "string"
},
"tenantId": {
"description": "The ID of the tenant the custom model validation belongs to.",
"format": "uuid4",
"title": "tenantId",
"type": "string"
},
"useCaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the use case associated with the validated custom model.",
"title": "useCaseId"
},
"userId": {
"description": "The ID of the user that created this custom model validation.",
"title": "userId",
"type": "string"
},
"userName": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The name of the user that created this custom model validation.",
"title": "userName"
},
"validationStatus": {
"description": "Status of custom model validation.",
"enum": [
"TESTING",
"PASSED",
"FAILED"
],
"title": "CustomModelValidationStatus",
"type": "string"
}
},
"required": [
"id",
"deploymentId",
"targetColumnName",
"validationStatus",
"modelId",
"deploymentAccessData",
"tenantId",
"name",
"useCaseId",
"creationDate",
"userId",
"predictionTimeout",
"playgroundId",
"citationsPrefixColumnName",
"promptColumnName",
"responseColumnName",
"expectedResponseColumnName"
],
"title": "SidecarModelMetricValidationResponse",
"type": "object"
}
SidecarModelMetricValidationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| citationsPrefixColumnName | any | true | The column name prefix the custom model uses for citation inputs. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| creationDate | string(date-time) | true | The creation date of the custom model validation (ISO 8601 formatted). | |
| deploymentAccessData | any | true | The parameters used for accessing the deployment. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | DeploymentAccessData | false | Add authorization_header to avoid breaking change to API. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| deploymentId | string | true | The ID of the custom model deployment. | |
| deploymentName | any | false | The name of the custom model deployment. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorMessage | any | false | The error message associated with the validation error (if the validation failed). |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| errorResolution | any | false | The error type associated with the insight error status and error message as an indicator of what fields needs to be edited if any. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | [InsightErrorResolution] | false | [Error type linking directly to the field name that is related to the error.] |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| expectedResponseColumnName | any | true | The name of the column the custom model uses for expected response text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| id | string | true | The ID of the custom model validation. | |
| modelId | string | true | The ID of the model used in the deployment. | |
| moderationConfiguration | any | false | The moderation configuration associated with the sidecar model metric. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | ModerationConfigurationWithoutID | false | Moderation Configuration associated with an insight. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| name | string | true | The name of the validated custom model. | |
| playgroundId | any | true | The ID of the playground associated with the sidecar model metric validation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| predictionTimeout | integer | true | The timeout in seconds for the prediction API used in this custom model validation. | |
| promptColumnName | any | true | The name of the column the custom model uses for prompt text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| responseColumnName | any | true | The name of the column the custom model uses for response text input. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| targetColumnName | string | true | The name of the column the custom model uses for prediction output. | |
| tenantId | string(uuid4) | true | The ID of the tenant the custom model validation belongs to. | |
| useCaseId | any | true | The ID of the use case associated with the validated custom model. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| userId | string | true | The ID of the user that created this custom model validation. | |
| userName | any | false | The name of the user that created this custom model validation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| validationStatus | CustomModelValidationStatus | true | The status of the custom model validation. |
SupportedDeploymentType
{
"description": "The type of the target output a DataRobot deployment produces.",
"enum": [
"TEXT_GENERATION",
"VECTOR_DATABASE",
"UNSTRUCTURED",
"REGRESSION",
"MULTICLASS",
"BINARY",
"NOT_SUPPORTED"
],
"title": "SupportedDeploymentType",
"type": "string"
}
SupportedDeploymentType
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| SupportedDeploymentType | string | false | The type of the target output a DataRobot deployment produces. |
Enumerated Values¶
| Property | Value |
|---|---|
| SupportedDeploymentType | [TEXT_GENERATION, VECTOR_DATABASE, UNSTRUCTURED, REGRESSION, MULTICLASS, BINARY, NOT_SUPPORTED] |
SyntheticEvaluationDatasetGenerationRequest
{
"description": "The body of the \"Generate synthetic evaluation dataset\" request.",
"properties": {
"datasetName": {
"anyOf": [
{
"maxLength": 255,
"type": "string"
},
{
"type": "null"
}
],
"description": "The name to use for the generated dataset.",
"title": "datasetName"
},
"language": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The language to use for the generated dataset.",
"title": "language"
},
"llmId": {
"description": "The ID of the LLM to use for synthetic dataset generation.",
"title": "llmId",
"type": "string"
},
"llmSettings": {
"anyOf": [
{
"additionalProperties": true,
"description": "The settings that are available for all non-custom LLMs.",
"properties": {
"maxCompletionLength": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"description": "Maximum number of tokens allowed in the chat completion. Use this value to, for example, control costs on token-based charges or manage response length for chat text limits.",
"title": "maxCompletionLength"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"title": "CommonLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs.",
"properties": {
"externalLlmContextSize": {
"anyOf": [
{
"maximum": 128000,
"minimum": 128,
"type": "integer"
},
{
"type": "null"
}
],
"default": 4096,
"description": "The external LLM's context size, in tokens. This value is only used for pruning documents supplied to the LLM when a vector database is associated with the LLM blueprint. It does not affect the external LLM's actual context size in any way and is not supplied to the LLM.",
"title": "externalLlmContextSize"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
},
"validationId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The validation ID of the custom model LLM.",
"title": "validationId"
}
},
"title": "CustomModelLLMSettings",
"type": "object"
},
{
"additionalProperties": false,
"description": "The settings that are available for custom model LLMs used via chat completion interface.",
"properties": {
"customModelId": {
"description": "The ID of the custom model used via chat completion interface.",
"title": "customModelId",
"type": "string"
},
"systemPrompt": {
"anyOf": [
{
"maxLength": 500000,
"type": "string"
},
{
"type": "null"
}
],
"description": "System prompt guides the style of the LLM response. It is a \"universal\" prompt, prepended to all individual prompts.",
"title": "systemPrompt"
}
},
"required": [
"customModelId"
],
"title": "CustomModelChatLLMSettings",
"type": "object"
},
{
"type": "null"
}
],
"description": "If specified, uses these LLM settings for the prompt and updates the settings of the corresponding chat or LLM blueprint to use these LLM settings.",
"title": "llmSettings"
},
"vectorDatabaseId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "The ID of the vector database to use for synthetic dataset generation.",
"title": "vectorDatabaseId"
}
},
"required": [
"llmId"
],
"title": "SyntheticEvaluationDatasetGenerationRequest",
"type": "object"
}
SyntheticEvaluationDatasetGenerationRequest
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetName | any | false | The name to use for the generated dataset. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | maxLength: 255 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| language | any | false | The language to use for the generated dataset. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| llmId | string | true | The ID of the LLM to use for synthetic dataset generation. | |
| llmSettings | any | false | If specified, uses these LLM settings for the prompt and updates the settings of the corresponding chat or LLM blueprint to use these LLM settings. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CommonLLMSettings | false | The settings that are available for all non-custom LLMs. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CustomModelLLMSettings | false | The settings that are available for custom model LLMs. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | CustomModelChatLLMSettings | false | The settings that are available for custom model LLMs used via chat completion interface. |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| vectorDatabaseId | any | false | The ID of the vector database to use for synthetic dataset generation. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
SyntheticEvaluationDatasetGenerationResponse
{
"description": "The body of the \"Create synthetic evaluation dataset\" response.",
"properties": {
"datasetId": {
"description": "The ID of the created dataset.",
"title": "datasetId",
"type": "string"
},
"promptColumnName": {
"description": "The name of the dataset column containing the prompt text.",
"title": "promptColumnName",
"type": "string"
},
"responseColumnName": {
"description": "The name of the dataset column containing the response text.",
"title": "responseColumnName",
"type": "string"
}
},
"required": [
"datasetId",
"promptColumnName",
"responseColumnName"
],
"title": "SyntheticEvaluationDatasetGenerationResponse",
"type": "object"
}
SyntheticEvaluationDatasetGenerationResponse
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| datasetId | string | true | The ID of the created dataset. | |
| promptColumnName | string | true | The name of the dataset column containing the prompt text. | |
| responseColumnName | string | true | The name of the dataset column containing the response text. |
ToolCallAccuracySettings
{
"description": "Additional arguments for the tool call accuracy metric.",
"properties": {
"argumentComparison": {
"description": "The different modes for comparing the arguments of tool calls.",
"enum": [
"exact_match",
"ignore_arguments"
],
"title": "ArgumentMatchMode",
"type": "string"
}
},
"required": [
"argumentComparison"
],
"title": "ToolCallAccuracySettings",
"type": "object"
}
ToolCallAccuracySettings
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| argumentComparison | ArgumentMatchMode | true | Setting defining how arguments of tool calls should be compared during the metric computation. |
ValidationError
{
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"title": "loc",
"type": "array"
},
"msg": {
"title": "msg",
"type": "string"
},
"type": {
"title": "type",
"type": "string"
}
},
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError",
"type": "object"
}
ValidationError
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| loc | [anyOf] | true | none |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | string | false | none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | integer | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| msg | string | true | none | |
| type | string | true | none |
VectorDatabaseRetrievers
{
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
VectorDatabaseRetrievers
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| VectorDatabaseRetrievers | string | false | The method used to retrieve relevant chunks from the vector database. |
Enumerated Values¶
| Property | Value |
|---|---|
| VectorDatabaseRetrievers | [SINGLE_LOOKUP_RETRIEVER, CONVERSATIONAL_RETRIEVER, MULTI_STEP_RETRIEVER] |
VectorDatabaseSettings
{
"description": "Vector database retrieval settings.",
"properties": {
"addNeighborChunks": {
"default": false,
"description": "Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1.",
"title": "addNeighborChunks",
"type": "boolean"
},
"maxDocumentsRetrievedPerPrompt": {
"anyOf": [
{
"maximum": 10,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of chunks to retrieve from the vector database.",
"title": "maxDocumentsRetrievedPerPrompt"
},
"maxTokens": {
"anyOf": [
{
"maximum": 51200,
"minimum": 1,
"type": "integer"
},
{
"type": "null"
}
],
"description": "The maximum number of tokens to retrieve from the vector database.",
"title": "maxTokens"
},
"maximalMarginalRelevanceLambda": {
"default": 0.5,
"description": "Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0).",
"maximum": 1,
"minimum": 0,
"title": "maximalMarginalRelevanceLambda",
"type": "number"
},
"retrievalMode": {
"description": "Retrieval modes for vector databases.",
"enum": [
"similarity",
"maximal_marginal_relevance"
],
"title": "RetrievalMode",
"type": "string"
},
"retriever": {
"description": "The method used to retrieve relevant chunks from the vector database.",
"enum": [
"SINGLE_LOOKUP_RETRIEVER",
"CONVERSATIONAL_RETRIEVER",
"MULTI_STEP_RETRIEVER"
],
"title": "VectorDatabaseRetrievers",
"type": "string"
}
},
"title": "VectorDatabaseSettings",
"type": "object"
}
VectorDatabaseSettings
Properties¶
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| addNeighborChunks | boolean | false | Add neighboring chunks to those that the similarity search retrieves, such that when selected, search returns i, i-1, and i+1. | |
| maxDocumentsRetrievedPerPrompt | any | false | The maximum number of chunks to retrieve from the vector database. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | integer | false | maximum: 10 minimum: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| maxTokens | any | false | The maximum number of tokens to retrieve from the vector database. |
anyOf
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | integer | false | maximum: 51200 minimum: 1 |
none |
or
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| » anonymous | null | false | none |
continued
| Name | Type | Required | Restrictions | Description |
|---|---|---|---|---|
| maximalMarginalRelevanceLambda | number | false | maximum: 1 minimum: 0 |
Adjust the retrieval of chunks when using maximal marginal relevance to favor diversity (0.0) or similarity (1.0). |
| retrievalMode | RetrievalMode | false | The retrieval mode to use. Similarity search or maximal marginal relevance. | |
| retriever | VectorDatabaseRetrievers | false | The method used to retrieve relevant chunks from the vector database. |