TrustyAI service API
- 1. Access
- 2. Endpoints
- 2.1. DataUpload
- 2.2. DownloadEndpoint
- 2.3. DriftMetricsApproxKSTest
- 2.4. DriftMetricsFourierMMDDrift
- 2.5. DriftMetricsKSTest
- 2.6. DriftMetricsMeanshift
- 2.7. ExplainersGlobal
- 2.8. ExplainersLocal
- 2.9. FairnessMetricsGroupDisparateImpactRatio
- 2.10. FairnessMetricsGroupStatisticalParityDifference
- 2.11. IdentityEndpoint
- 2.12. InternalOnlyInferenceConsumer
- 2.13. LegacyDisparateImpactRatio
- 2.14. LegacyStatisticalParityDifference
- 2.15. MetricsInformationEndpoint
- 2.16. ServiceMetadata
- 3. Models
- 3.1. ApproxKSTestMetricRequest
- 3.2. CounterfactualExplainerConfig
- 3.3. CounterfactualExplanationConfig
- 3.4. CounterfactualExplanationRequest
- 3.5. DataRequestPayload
- 3.6. DataTagging
- 3.7. DataType
- 3.8. FourierMMDFitting
- 3.9. FourierMMDMetricRequest
- 3.10. FourierMMDParameters
- 3.11. GKSketch
- 3.12. GKSketchSummaryInner
- 3.13. GlobalExplanationRequest
- 3.14. GroupDefinitionRequest
- 3.15. GroupMetricRequest
- 3.16. IdentityMetricRequest
- 3.17. InferencePartialPayload
- 3.18. JsonNode
- 3.19. JsonNodeType
- 3.20. KSTestMetricRequest
- 3.21. LimeExplainerConfig
- 3.22. LimeExplanationConfig
- 3.23. LimeExplanationRequest
- 3.24. LinkType
- 3.25. MeanshiftMetricRequest
- 3.26. ModelConfig
- 3.27. ModelInferJointPayload
- 3.28. ModelInferRequestPayload
- 3.29. ModelInferResponsePayload
- 3.30. NameMapping
- 3.31. PartialKind
- 3.32. PartialPayloadId
- 3.33. ReconcilableFeature
- 3.34. ReconcilableOutput
- 3.35. RegularizerType
- 3.36. RowMatcher
- 3.37. SHAPExplainerConfig
- 3.38. SHAPExplanationConfig
- 3.39. SHAPExplanationRequest
- 3.40. ScheduleId
- 3.41. StatisticalSummaryValues
- 3.42. TSSaliencyExplainerConfig
- 3.43. TSSaliencyExplanationConfig
- 3.44. TSSaliencyExplanationRequest
- 3.45. TensorPayload
- 3.46. TypedValue
- 3.47. ValueNode
No description provided (generated by Openapi Generator https://github.com/openapitools/openapi-generator)
2. Endpoints
2.1. DataUpload
2.2. DownloadEndpoint
2.3. DriftMetricsApproxKSTest
2.3.1. metricsDriftApproxkstestDefinitionGet
GET /metrics/drift/approxkstest/definition
Provide a general definition of this metric.
2.3.2. metricsDriftApproxkstestPost
POST /metrics/drift/approxkstest
Compute the current value of this metric.
2.3.3. metricsDriftApproxkstestRequestDelete
DELETE /metrics/drift/approxkstest/request
Delete a recurring computation of this metric.
2.4. DriftMetricsFourierMMDDrift
2.4.1. metricsDriftFouriermmdDefinitionGet
GET /metrics/drift/fouriermmd/definition
Provide a general definition of this metric.
2.4.2. metricsDriftFouriermmdPost
POST /metrics/drift/fouriermmd
Compute the current value of this metric.
2.4.3. metricsDriftFouriermmdRequestDelete
DELETE /metrics/drift/fouriermmd/request
Delete a recurring computation of this metric.
2.5. DriftMetricsKSTest
2.5.1. metricsDriftKstestDefinitionGet
GET /metrics/drift/kstest/definition
Provide a general definition of this metric.
2.5.2. metricsDriftKstestPost
POST /metrics/drift/kstest
Compute the current value of this metric.
2.5.3. metricsDriftKstestRequestDelete
DELETE /metrics/drift/kstest/request
Delete a recurring computation of this metric.
2.6. DriftMetricsMeanshift
2.6.1. metricsDriftMeanshiftDefinitionGet
GET /metrics/drift/meanshift/definition
Provide a general definition of this metric.
2.6.2. metricsDriftMeanshiftPost
POST /metrics/drift/meanshift
Compute the current value of this metric.
2.6.3. metricsDriftMeanshiftRequestDelete
DELETE /metrics/drift/meanshift/request
Delete a recurring computation of this metric.
2.7. ExplainersGlobal
2.7.1. explainersGlobalLimePost
POST /explainers/global/lime
Compute a global LIME explanation.
2.8. ExplainersLocal
2.8.1. explainersLocalCfPost
POST /explainers/local/cf
Compute a Counterfactual explanation.
2.8.2. explainersLocalLimePost
POST /explainers/local/lime
Compute a LIME explanation.
2.8.3. explainersLocalShapPost
POST /explainers/local/shap
Compute a SHAP explanation.
2.9. FairnessMetricsGroupDisparateImpactRatio
2.9.1. metricsGroupFairnessDirDefinitionGet
GET /metrics/group/fairness/dir/definition
Provide a general definition of this metric.
2.9.2. metricsGroupFairnessDirDefinitionPost
POST /metrics/group/fairness/dir/definition
Provide a specific, plain-english interpretation of a specific value of this metric.
2.9.3. metricsGroupFairnessDirPost
POST /metrics/group/fairness/dir
Compute the current value of this metric.
2.9.4. metricsGroupFairnessDirRequestDelete
DELETE /metrics/group/fairness/dir/request
Delete a recurring computation of this metric.
2.10. FairnessMetricsGroupStatisticalParityDifference
2.10.1. metricsGroupFairnessSpdDefinitionGet
GET /metrics/group/fairness/spd/definition
Provide a general definition of this metric.
2.10.2. metricsGroupFairnessSpdDefinitionPost
POST /metrics/group/fairness/spd/definition
Provide a specific, plain-english interpretation of a specific value of this metric.
2.10.3. metricsGroupFairnessSpdPost
POST /metrics/group/fairness/spd
Compute the current value of this metric.
2.10.4. metricsGroupFairnessSpdRequestDelete
DELETE /metrics/group/fairness/spd/request
Delete a recurring computation of this metric.
2.11. IdentityEndpoint
2.11.1. metricsIdentityDefinitionGet
GET /metrics/identity/definition
Provide a general definition of this metric.
2.11.2. metricsIdentityDefinitionPost
POST /metrics/identity/definition
Provide a specific, plain-english interpretation of a specific value of this metric.
2.11.3. metricsIdentityPost
POST /metrics/identity
Provide a specific, plain-english interpretation of the current value of this metric.
2.11.4. metricsIdentityRequestDelete
DELETE /metrics/identity/request
Delete a recurring computation of this metric.
2.12. InternalOnlyInferenceConsumer
2.13. LegacyDisparateImpactRatio
2.13.1. metricsDirDefinitionGet
GET /metrics/dir/definition
Provide a general definition of this metric.
2.13.2. metricsDirDefinitionPost
POST /metrics/dir/definition
Provide a specific, plain-english interpretation of a specific value of this metric.
2.13.3. metricsDirPost
POST /metrics/dir
Compute the current value of this metric.
2.13.4. metricsDirRequestDelete
DELETE /metrics/dir/request
Delete a recurring computation of this metric.
2.14. LegacyStatisticalParityDifference
2.14.1. metricsSpdDefinitionGet
GET /metrics/spd/definition
Provide a general definition of this metric.
2.14.2. metricsSpdDefinitionPost
POST /metrics/spd/definition
Provide a specific, plain-english interpretation of a specific value of this metric.
2.14.3. metricsSpdPost
POST /metrics/spd
Compute the current value of this metric.
2.14.4. metricsSpdRequestDelete
DELETE /metrics/spd/request
Delete a recurring computation of this metric.
2.16. ServiceMetadata
2.16.1. infoGet
GET /info
Get a comprehensive overview of the model inference datasets collected by TrustyAI and the metric computations that are scheduled over those datasets.
2.16.2. infoInferenceIdsModelGet
GET /info/inference/ids/{model}
Get a list of all inference ids within a particular model inference dataset.
2.16.3. infoNamesDelete
DELETE /info/names
Remove any column names that have been applied to a particular inference model dataset.
2.16.4. infoNamesPost
POST /info/names
Apply a set of human-readable column names to a particular inference model dataset.
2.16.5. infoTagsGet
GET /info/tags
Retrieve the tags that have been applied to a particular model dataset, as well as a count of that tag’s frequency within the dataset.
3. Models
3.1. ApproxKSTestMetricRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
thresholdDelta |
Double |
double |
|||
referenceTag |
String |
||||
fitColumns |
Set of [string] |
||||
epsilon |
Double |
double |
|||
sketchFitting |
Map of GKSketch |
3.2. CounterfactualExplainerConfig
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
n_samples |
Integer |
int32 |
3.3. CounterfactualExplanationConfig
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
model |
X |
||||
explainer |
3.4. CounterfactualExplanationRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
predictionId |
X |
String |
|||
config |
X |
||||
goals |
Map of [string] |
||||
explanationConfig |
3.5. DataRequestPayload
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
matchAny |
List of RowMatcher |
||||
matchAll |
List of RowMatcher |
||||
matchNone |
List of RowMatcher |
3.6. DataTagging
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
dataTagging |
Map of [array] |
int32 |
3.8. FourierMMDFitting
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
randomSeed |
Integer |
int32 |
|||
deltaStat |
Boolean |
||||
nMode |
Integer |
int32 |
|||
scale |
List of [double] |
double |
|||
aRef |
List of [double] |
double |
|||
meanMMD |
Double |
double |
|||
stdMMD |
Double |
double |
3.9. FourierMMDMetricRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
thresholdDelta |
Double |
double |
|||
referenceTag |
String |
||||
fitColumns |
Set of [string] |
||||
parameters |
|||||
gamma |
Double |
double |
|||
fitting |
3.10. FourierMMDParameters
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
nWindow |
Integer |
int32 |
|||
nTest |
Integer |
int32 |
|||
nMode |
Integer |
int32 |
|||
randomSeed |
Integer |
int32 |
|||
sig |
Double |
double |
|||
deltaStat |
Boolean |
||||
epsilon |
Double |
double |
3.11. GKSketch
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
epsilon |
Double |
double |
|||
summary |
List of [GKSketch_summary_inner] |
||||
xmin |
Double |
double |
|||
xmax |
Double |
double |
|||
numx |
Integer |
int32 |
3.12. GKSketchSummaryInner
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
left |
Double |
double |
|||
middle |
Integer |
int32 |
|||
right |
Integer |
int32 |
3.14. GroupDefinitionRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
protectedAttribute |
String |
||||
outcomeName |
String |
||||
privilegedAttribute |
|||||
unprivilegedAttribute |
|||||
favorableOutcome |
|||||
thresholdDelta |
Double |
double |
|||
metricValue |
3.15. GroupMetricRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
protectedAttribute |
String |
||||
outcomeName |
String |
||||
privilegedAttribute |
|||||
unprivilegedAttribute |
|||||
favorableOutcome |
|||||
thresholdDelta |
Double |
double |
3.16. IdentityMetricRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
columnName |
String |
||||
lowerThreshold |
Double |
double |
|||
upperThreshold |
Double |
double |
3.17. InferencePartialPayload
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
partialPayloadId |
|||||
metadata |
Map of [string] |
||||
id |
String |
||||
kind |
request, response, |
||||
data |
String |
||||
modelid |
String |
3.18. JsonNode
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
empty |
Boolean |
||||
valueNode |
Boolean |
||||
containerNode |
Boolean |
||||
missingNode |
Boolean |
||||
array |
Boolean |
||||
object |
Boolean |
||||
nodeType |
ARRAY, BINARY, BOOLEAN, MISSING, NULL, NUMBER, OBJECT, POJO, STRING, |
||||
pojo |
Boolean |
||||
number |
Boolean |
||||
integralNumber |
Boolean |
||||
floatingPointNumber |
Boolean |
||||
short |
Boolean |
||||
int |
Boolean |
||||
long |
Boolean |
||||
float |
Boolean |
||||
double |
Boolean |
||||
bigDecimal |
Boolean |
||||
bigInteger |
Boolean |
||||
textual |
Boolean |
||||
boolean |
Boolean |
||||
null |
Boolean |
||||
binary |
Boolean |
3.20. KSTestMetricRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
thresholdDelta |
Double |
double |
|||
referenceTag |
String |
||||
fitColumns |
Set of [string] |
3.21. LimeExplainerConfig
Configuration for the LIME explainer
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
n_samples |
Integer |
Number of samples to be generated for the local linear model training |
int32 |
||
timeout |
Long |
Timeout (in seconds) for the LIME explainer |
int64 |
||
separable_dataset_ratio |
Double |
Separable dataset ration |
double |
||
retries |
Integer |
Number of retries |
int32 |
||
adaptive_variance |
Boolean |
Whether the explainer should adapt the variance in the generated (perturbed) data when it's not separable |
|||
penalize_balance_sparse |
Boolean |
Whether to penalize weights whose sparse features encoding is balanced with respect to target output |
|||
proximity_filter |
Boolean |
Whether to prefer filtering by proximity over weighting by proximity when generating samples for the linear model |
|||
proximity_threshold |
Double |
The proximity threshold used to filter samples when proximity filter is true |
double |
||
proximity_kernel_width |
Double |
The width of the kernel used to calculate proximity of sparse vector instances |
double |
||
encoding_cluster_threshold |
Double |
Encoding cluster threshold |
double |
||
encoding_gaussian_filter_width |
Double |
Encoding Gaussian filter width |
double |
||
normalize_weights |
Boolean |
Whether to normalize weights generated by LIME or not |
|||
high_score_feature_zones |
Boolean |
Whether to use high score feature zones for more accurate numeric features sampling |
|||
feature_selection |
Boolean |
Whether to operate feature selection |
|||
n_features |
Integer |
Number of features to use |
int32 |
||
track_counterfactuals |
Boolean |
Whether to track byproduct counterfactuals |
|||
use_wlr_model |
Boolean |
Whether to use a weighted linear regression model |
|||
filter_interpretable |
Boolean |
Whether to run proximity filter in the interpretable space |
3.22. LimeExplanationConfig
Configuration for the LIME explanation, including model and explainer parameters
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
model |
X |
Model configuration |
|||
explainer |
Explainer configuration |
3.23. LimeExplanationRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
predictionId |
X |
String |
|||
config |
X |
3.25. MeanshiftMetricRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
requestName |
String |
||||
metricName |
String |
||||
batchSize |
Integer |
int32 |
|||
thresholdDelta |
Double |
double |
|||
referenceTag |
String |
||||
fitColumns |
Set of [string] |
||||
fitting |
Map of StatisticalSummaryValues |
3.26. ModelConfig
Model configuration parameters
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
target |
X |
String |
Location of the model, for instance a Kubernetes service name and optionally port |
||
name |
X |
String |
Model's name |
||
version |
String |
Model's version, optional |
3.27. ModelInferJointPayload
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
model_name |
String |
||||
data_tag |
String |
||||
is_ground_truth |
Boolean |
||||
request |
|||||
response |
3.28. ModelInferRequestPayload
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
id |
String |
||||
inputs |
List of TensorPayload |
3.29. ModelInferResponsePayload
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
model_name |
String |
||||
model_version |
String |
||||
id |
String |
||||
parameters |
Map of [AnyType] |
||||
outputs |
List of TensorPayload |
3.30. NameMapping
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
modelId |
String |
||||
inputMapping |
Map of [string] |
||||
outputMapping |
Map of [string] |
3.32. PartialPayloadId
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
predictionId |
String |
||||
kind |
request, response, |
3.33. ReconcilableFeature
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
rawValueNodes |
List of ValueNode |
||||
rawValueNode |
|||||
reconciledType |
X |
List of TypedValue |
|||
multipleValued |
Boolean |
3.34. ReconcilableOutput
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
rawValueNodes |
List of ValueNode |
||||
rawValueNode |
|||||
reconciledType |
X |
List of TypedValue |
|||
multipleValued |
Boolean |
3.36. RowMatcher
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
columnName |
String |
||||
operation |
String |
||||
values |
List of ValueNode |
3.37. SHAPExplainerConfig
Configuration for the LIME explainer
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
n_samples |
Integer |
Number of samples to be generated for the local linear model training |
int32 |
||
timeout |
Long |
Timeout (in seconds) for the LIME explainer |
int64 |
||
link |
Either LOGIT or IDENTITY. If you want the SHAP values to sum to the exact model output, use IDENTITYIf your model outputs probabilities and you want the SHAP values touse log-odds units, use LOGIT |
LOGIT, IDENTITY, |
|||
regularizer |
The choice of regularizer to use when fitting data. This will select a certain fraction of features to use, based on which are most important to the regression |
AUTO, AIC, BIC, TOP_N_FEATURES, NONE, |
|||
confidence |
Double |
The size of the confidence window to use for SHAP values |
double |
||
track_counterfactuals |
Boolean |
Whether to track byproduct counterfactuals generated during explanation |
3.38. SHAPExplanationConfig
Configuration for the SHAP explanation, including model and explainer parameters
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
model |
X |
Model configuration |
|||
explainer |
Explainer configuration |
3.39. SHAPExplanationRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
predictionId |
X |
String |
|||
config |
X |
3.41. StatisticalSummaryValues
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
mean |
Double |
double |
|||
variance |
Double |
double |
|||
n |
Long |
int64 |
|||
max |
Double |
double |
|||
min |
Double |
double |
|||
sum |
Double |
double |
|||
standardDeviation |
Double |
double |
3.42. TSSaliencyExplainerConfig
Configuration for the TSSaliency explainer
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
timeout |
Long |
Timeout (in seconds) for the LIME explainer |
int64 |
||
mu |
Double |
Step size for gradient estimation |
double |
||
n_samples |
Integer |
Number of samples for gradient estimation |
int32 |
||
n_alpha |
Integer |
Number of steps in convex path |
int32 |
||
sigma |
Double |
Standard deviation |
double |
||
base_values |
X |
List of [double] |
Feature base values. Number of elements must be the same as number of features |
double |
3.43. TSSaliencyExplanationConfig
Configuration for the TSSaliency explanation, including model and explainer parameters
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
model |
X |
Model configuration |
|||
explainer |
Explainer configuration |
3.44. TSSaliencyExplanationRequest
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
predictionIds |
X |
List of [string] |
|||
config |
X |
3.45. TensorPayload
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
name |
String |
||||
shape |
List of [number] |
||||
datatype |
String |
||||
parameters |
Map of [AnyType] |
||||
data |
List of [AnyType] |
||||
executionIDs |
List of [string] |
3.46. TypedValue
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
type |
BOOL, FLOAT, DOUBLE, INT32, INT64, STRING, MAP, UNKNOWN, |
||||
value |
3.47. ValueNode
| Field Name | Required | Nullable | Type | Description | Format |
|---|---|---|---|---|---|
valueNode |
Boolean |
||||
containerNode |
Boolean |
||||
missingNode |
Boolean |
||||
array |
Boolean |
||||
object |
Boolean |
||||
nodeType |
ARRAY, BINARY, BOOLEAN, MISSING, NULL, NUMBER, OBJECT, POJO, STRING, |
||||
pojo |
Boolean |
||||
number |
Boolean |
||||
integralNumber |
Boolean |
||||
floatingPointNumber |
Boolean |
||||
short |
Boolean |
||||
int |
Boolean |
||||
long |
Boolean |
||||
float |
Boolean |
||||
double |
Boolean |
||||
bigDecimal |
Boolean |
||||
bigInteger |
Boolean |
||||
textual |
Boolean |
||||
boolean |
Boolean |
||||
null |
Boolean |
||||
binary |
Boolean |
||||
empty |
Boolean |